diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,115256 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 16459,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.075703262652652e-05,
+      "grad_norm": 0.17771193385124207,
+      "learning_rate": 0.0,
+      "loss": 1.1983,
+      "step": 1
+    },
+    {
+      "epoch": 0.00012151406525305304,
+      "grad_norm": 0.6649877429008484,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.434,
+      "step": 2
+    },
+    {
+      "epoch": 0.00018227109787957957,
+      "grad_norm": 0.1839398741722107,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.2097,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002430281305061061,
+      "grad_norm": 0.3777402341365814,
+      "learning_rate": 6e-06,
+      "loss": 1.4124,
+      "step": 4
+    },
+    {
+      "epoch": 0.0003037851631326326,
+      "grad_norm": 0.2884521484375,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.42,
+      "step": 5
+    },
+    {
+      "epoch": 0.00036454219575915913,
+      "grad_norm": 0.18382205069065094,
+      "learning_rate": 1e-05,
+      "loss": 1.2358,
+      "step": 6
+    },
+    {
+      "epoch": 0.00042529922838568566,
+      "grad_norm": 0.8757884502410889,
+      "learning_rate": 1.2e-05,
+      "loss": 1.4333,
+      "step": 7
+    },
+    {
+      "epoch": 0.0004860562610122122,
+      "grad_norm": 0.2029954046010971,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.2286,
+      "step": 8
+    },
+    {
+      "epoch": 0.0005468132936387387,
+      "grad_norm": 0.6275981068611145,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.3836,
+      "step": 9
+    },
+    {
+      "epoch": 0.0006075703262652652,
+      "grad_norm": 0.2643689513206482,
+      "learning_rate": 1.8e-05,
+      "loss": 1.2047,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006683273588917917,
+      "grad_norm": 0.2377251386642456,
+      "learning_rate": 2e-05,
+      "loss": 1.2234,
+      "step": 11
+    },
+    {
+      "epoch": 0.0007290843915183183,
+      "grad_norm": 0.27713561058044434,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.2786,
+      "step": 12
+    },
+    {
+      "epoch": 0.0007898414241448448,
+      "grad_norm": 0.25062233209609985,
+      "learning_rate": 2.4e-05,
+      "loss": 1.2505,
+      "step": 13
+    },
+    {
+      "epoch": 0.0008505984567713713,
+      "grad_norm": 0.2383335530757904,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.358,
+      "step": 14
+    },
+    {
+      "epoch": 0.0009113554893978978,
+      "grad_norm": 0.3477482795715332,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.4309,
+      "step": 15
+    },
+    {
+      "epoch": 0.0009721125220244244,
+      "grad_norm": 0.24556300044059753,
+      "learning_rate": 3e-05,
+      "loss": 1.2673,
+      "step": 16
+    },
+    {
+      "epoch": 0.0010328695546509509,
+      "grad_norm": 1.1669777631759644,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.3584,
+      "step": 17
+    },
+    {
+      "epoch": 0.0010936265872774774,
+      "grad_norm": 0.26149648427963257,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.3964,
+      "step": 18
+    },
+    {
+      "epoch": 0.001154383619904004,
+      "grad_norm": 0.30708009004592896,
+      "learning_rate": 3.6e-05,
+      "loss": 1.3155,
+      "step": 19
+    },
+    {
+      "epoch": 0.0012151406525305304,
+      "grad_norm": 0.23968864977359772,
+      "learning_rate": 3.8e-05,
+      "loss": 1.2109,
+      "step": 20
+    },
+    {
+      "epoch": 0.001275897685157057,
+      "grad_norm": 0.2607705593109131,
+      "learning_rate": 4e-05,
+      "loss": 1.3489,
+      "step": 21
+    },
+    {
+      "epoch": 0.0013366547177835835,
+      "grad_norm": 0.2762501835823059,
+      "learning_rate": 4.2e-05,
+      "loss": 1.2211,
+      "step": 22
+    },
+    {
+      "epoch": 0.00139741175041011,
+      "grad_norm": 0.3239821195602417,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.4063,
+      "step": 23
+    },
+    {
+      "epoch": 0.0014581687830366365,
+      "grad_norm": 0.27468156814575195,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.2292,
+      "step": 24
+    },
+    {
+      "epoch": 0.001518925815663163,
+      "grad_norm": 0.34043198823928833,
+      "learning_rate": 4.8e-05,
+      "loss": 1.2227,
+      "step": 25
+    },
+    {
+      "epoch": 0.0015796828482896896,
+      "grad_norm": 0.23437948524951935,
+      "learning_rate": 5e-05,
+      "loss": 1.2727,
+      "step": 26
+    },
+    {
+      "epoch": 0.001640439880916216,
+      "grad_norm": 0.2744443416595459,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.2517,
+      "step": 27
+    },
+    {
+      "epoch": 0.0017011969135427426,
+      "grad_norm": 0.3189443051815033,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.1987,
+      "step": 28
+    },
+    {
+      "epoch": 0.0017619539461692691,
+      "grad_norm": 0.36634373664855957,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.2931,
+      "step": 29
+    },
+    {
+      "epoch": 0.0018227109787957957,
+      "grad_norm": 0.21552708745002747,
+      "learning_rate": 5.8e-05,
+      "loss": 1.19,
+      "step": 30
+    },
+    {
+      "epoch": 0.0018834680114223222,
+      "grad_norm": 0.22331136465072632,
+      "learning_rate": 6e-05,
+      "loss": 1.166,
+      "step": 31
+    },
+    {
+      "epoch": 0.0019442250440488487,
+      "grad_norm": 0.27115359902381897,
+      "learning_rate": 6.2e-05,
+      "loss": 1.2394,
+      "step": 32
+    },
+    {
+      "epoch": 0.002004982076675375,
+      "grad_norm": 0.24723933637142181,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.298,
+      "step": 33
+    },
+    {
+      "epoch": 0.0020657391093019018,
+      "grad_norm": 0.6054807305335999,
+      "learning_rate": 6.6e-05,
+      "loss": 1.2711,
+      "step": 34
+    },
+    {
+      "epoch": 0.002126496141928428,
+      "grad_norm": 0.2657296657562256,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.1846,
+      "step": 35
+    },
+    {
+      "epoch": 0.002187253174554955,
+      "grad_norm": 0.5996095538139343,
+      "learning_rate": 7e-05,
+      "loss": 1.4208,
+      "step": 36
+    },
+    {
+      "epoch": 0.002248010207181481,
+      "grad_norm": 0.20939430594444275,
+      "learning_rate": 7.2e-05,
+      "loss": 1.21,
+      "step": 37
+    },
+    {
+      "epoch": 0.002308767239808008,
+      "grad_norm": 0.5714825987815857,
+      "learning_rate": 7.4e-05,
+      "loss": 1.2827,
+      "step": 38
+    },
+    {
+      "epoch": 0.002369524272434534,
+      "grad_norm": 0.9815083742141724,
+      "learning_rate": 7.6e-05,
+      "loss": 1.1884,
+      "step": 39
+    },
+    {
+      "epoch": 0.002430281305061061,
+      "grad_norm": 1.6414822340011597,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.4738,
+      "step": 40
+    },
+    {
+      "epoch": 0.002491038337687587,
+      "grad_norm": 0.3299960792064667,
+      "learning_rate": 8e-05,
+      "loss": 1.2818,
+      "step": 41
+    },
+    {
+      "epoch": 0.002551795370314114,
+      "grad_norm": 0.2786610424518585,
+      "learning_rate": 8.2e-05,
+      "loss": 1.3226,
+      "step": 42
+    },
+    {
+      "epoch": 0.0026125524029406402,
+      "grad_norm": 0.48672959208488464,
+      "learning_rate": 8.4e-05,
+      "loss": 1.3591,
+      "step": 43
+    },
+    {
+      "epoch": 0.002673309435567167,
+      "grad_norm": 0.2563237249851227,
+      "learning_rate": 8.6e-05,
+      "loss": 1.2016,
+      "step": 44
+    },
+    {
+      "epoch": 0.0027340664681936933,
+      "grad_norm": 0.31436944007873535,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.3627,
+      "step": 45
+    },
+    {
+      "epoch": 0.00279482350082022,
+      "grad_norm": 0.4340488016605377,
+      "learning_rate": 9e-05,
+      "loss": 1.3916,
+      "step": 46
+    },
+    {
+      "epoch": 0.0028555805334467463,
+      "grad_norm": 0.30021172761917114,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.1985,
+      "step": 47
+    },
+    {
+      "epoch": 0.002916337566073273,
+      "grad_norm": 0.587597668170929,
+      "learning_rate": 9.4e-05,
+      "loss": 1.3836,
+      "step": 48
+    },
+    {
+      "epoch": 0.0029770945986997994,
+      "grad_norm": 0.2984046936035156,
+      "learning_rate": 9.6e-05,
+      "loss": 1.2871,
+      "step": 49
+    },
+    {
+      "epoch": 0.003037851631326326,
+      "grad_norm": 1.8778438568115234,
+      "learning_rate": 9.8e-05,
+      "loss": 1.2312,
+      "step": 50
+    },
+    {
+      "epoch": 0.0030986086639528524,
+      "grad_norm": 0.548705518245697,
+      "learning_rate": 0.0001,
+      "loss": 1.2312,
+      "step": 51
+    },
+    {
+      "epoch": 0.003159365696579379,
+      "grad_norm": 0.5713818073272705,
+      "learning_rate": 9.999999908362017e-05,
+      "loss": 1.2052,
+      "step": 52
+    },
+    {
+      "epoch": 0.0032201227292059055,
+      "grad_norm": 0.4120612144470215,
+      "learning_rate": 9.999999633448071e-05,
+      "loss": 1.3071,
+      "step": 53
+    },
+    {
+      "epoch": 0.003280879761832432,
+      "grad_norm": 0.48224759101867676,
+      "learning_rate": 9.999999175258171e-05,
+      "loss": 1.2323,
+      "step": 54
+    },
+    {
+      "epoch": 0.0033416367944589585,
+      "grad_norm": 0.5659412741661072,
+      "learning_rate": 9.999998533792336e-05,
+      "loss": 1.203,
+      "step": 55
+    },
+    {
+      "epoch": 0.0034023938270854853,
+      "grad_norm": 0.34445783495903015,
+      "learning_rate": 9.999997709050588e-05,
+      "loss": 1.2022,
+      "step": 56
+    },
+    {
+      "epoch": 0.0034631508597120116,
+      "grad_norm": 0.957758903503418,
+      "learning_rate": 9.999996701032956e-05,
+      "loss": 1.2781,
+      "step": 57
+    },
+    {
+      "epoch": 0.0035239078923385383,
+      "grad_norm": 0.4782249331474304,
+      "learning_rate": 9.99999550973948e-05,
+      "loss": 1.166,
+      "step": 58
+    },
+    {
+      "epoch": 0.0035846649249650646,
+      "grad_norm": 0.4631136655807495,
+      "learning_rate": 9.999994135170202e-05,
+      "loss": 1.1847,
+      "step": 59
+    },
+    {
+      "epoch": 0.0036454219575915913,
+      "grad_norm": 0.49536243081092834,
+      "learning_rate": 9.999992577325172e-05,
+      "loss": 1.3205,
+      "step": 60
+    },
+    {
+      "epoch": 0.0037061789902181176,
+      "grad_norm": 0.6254922151565552,
+      "learning_rate": 9.999990836204447e-05,
+      "loss": 1.175,
+      "step": 61
+    },
+    {
+      "epoch": 0.0037669360228446444,
+      "grad_norm": 0.4348151385784149,
+      "learning_rate": 9.999988911808093e-05,
+      "loss": 1.306,
+      "step": 62
+    },
+    {
+      "epoch": 0.0038276930554711707,
+      "grad_norm": 0.3686026930809021,
+      "learning_rate": 9.999986804136179e-05,
+      "loss": 1.2343,
+      "step": 63
+    },
+    {
+      "epoch": 0.0038884500880976974,
+      "grad_norm": 0.26751378178596497,
+      "learning_rate": 9.99998451318878e-05,
+      "loss": 1.266,
+      "step": 64
+    },
+    {
+      "epoch": 0.003949207120724224,
+      "grad_norm": 0.555131733417511,
+      "learning_rate": 9.999982038965985e-05,
+      "loss": 1.1944,
+      "step": 65
+    },
+    {
+      "epoch": 0.00400996415335075,
+      "grad_norm": 0.2945740520954132,
+      "learning_rate": 9.999979381467879e-05,
+      "loss": 1.1458,
+      "step": 66
+    },
+    {
+      "epoch": 0.004070721185977277,
+      "grad_norm": 0.5865735411643982,
+      "learning_rate": 9.999976540694564e-05,
+      "loss": 1.2297,
+      "step": 67
+    },
+    {
+      "epoch": 0.0041314782186038035,
+      "grad_norm": 0.5849476456642151,
+      "learning_rate": 9.999973516646142e-05,
+      "loss": 1.3176,
+      "step": 68
+    },
+    {
+      "epoch": 0.00419223525123033,
+      "grad_norm": 0.3707224726676941,
+      "learning_rate": 9.999970309322725e-05,
+      "loss": 1.2165,
+      "step": 69
+    },
+    {
+      "epoch": 0.004252992283856856,
+      "grad_norm": 0.6674245595932007,
+      "learning_rate": 9.999966918724429e-05,
+      "loss": 1.248,
+      "step": 70
+    },
+    {
+      "epoch": 0.004313749316483383,
+      "grad_norm": 0.33753934502601624,
+      "learning_rate": 9.999963344851379e-05,
+      "loss": 1.3501,
+      "step": 71
+    },
+    {
+      "epoch": 0.00437450634910991,
+      "grad_norm": 0.6501384377479553,
+      "learning_rate": 9.999959587703706e-05,
+      "loss": 1.2387,
+      "step": 72
+    },
+    {
+      "epoch": 0.004435263381736436,
+      "grad_norm": 0.380902498960495,
+      "learning_rate": 9.999955647281549e-05,
+      "loss": 1.1812,
+      "step": 73
+    },
+    {
+      "epoch": 0.004496020414362962,
+      "grad_norm": 0.5059203505516052,
+      "learning_rate": 9.999951523585051e-05,
+      "loss": 1.2051,
+      "step": 74
+    },
+    {
+      "epoch": 0.004556777446989489,
+      "grad_norm": 0.21010912954807281,
+      "learning_rate": 9.999947216614363e-05,
+      "loss": 1.1709,
+      "step": 75
+    },
+    {
+      "epoch": 0.004617534479616016,
+      "grad_norm": 1.1988906860351562,
+      "learning_rate": 9.999942726369644e-05,
+      "loss": 1.1707,
+      "step": 76
+    },
+    {
+      "epoch": 0.004678291512242542,
+      "grad_norm": 0.3643610179424286,
+      "learning_rate": 9.999938052851057e-05,
+      "loss": 1.2246,
+      "step": 77
+    },
+    {
+      "epoch": 0.004739048544869068,
+      "grad_norm": 0.2887507379055023,
+      "learning_rate": 9.999933196058775e-05,
+      "loss": 1.2145,
+      "step": 78
+    },
+    {
+      "epoch": 0.0047998055774955955,
+      "grad_norm": 0.8529403209686279,
+      "learning_rate": 9.999928155992975e-05,
+      "loss": 1.2869,
+      "step": 79
+    },
+    {
+      "epoch": 0.004860562610122122,
+      "grad_norm": 0.4480118155479431,
+      "learning_rate": 9.99992293265384e-05,
+      "loss": 1.3645,
+      "step": 80
+    },
+    {
+      "epoch": 0.004921319642748648,
+      "grad_norm": 0.5300185680389404,
+      "learning_rate": 9.999917526041567e-05,
+      "loss": 1.3036,
+      "step": 81
+    },
+    {
+      "epoch": 0.004982076675375174,
+      "grad_norm": 0.5703990459442139,
+      "learning_rate": 9.999911936156348e-05,
+      "loss": 1.1698,
+      "step": 82
+    },
+    {
+      "epoch": 0.005042833708001702,
+      "grad_norm": 0.4402383267879486,
+      "learning_rate": 9.999906162998392e-05,
+      "loss": 1.2819,
+      "step": 83
+    },
+    {
+      "epoch": 0.005103590740628228,
+      "grad_norm": 0.3625530004501343,
+      "learning_rate": 9.999900206567906e-05,
+      "loss": 1.1667,
+      "step": 84
+    },
+    {
+      "epoch": 0.005164347773254754,
+      "grad_norm": 0.31713375449180603,
+      "learning_rate": 9.999894066865115e-05,
+      "loss": 1.1938,
+      "step": 85
+    },
+    {
+      "epoch": 0.0052251048058812805,
+      "grad_norm": 0.49715203046798706,
+      "learning_rate": 9.999887743890239e-05,
+      "loss": 1.3341,
+      "step": 86
+    },
+    {
+      "epoch": 0.005285861838507808,
+      "grad_norm": 0.3231082856655121,
+      "learning_rate": 9.999881237643512e-05,
+      "loss": 1.3648,
+      "step": 87
+    },
+    {
+      "epoch": 0.005346618871134334,
+      "grad_norm": 0.22051608562469482,
+      "learning_rate": 9.999874548125172e-05,
+      "loss": 1.218,
+      "step": 88
+    },
+    {
+      "epoch": 0.00540737590376086,
+      "grad_norm": 0.7233831882476807,
+      "learning_rate": 9.999867675335464e-05,
+      "loss": 1.181,
+      "step": 89
+    },
+    {
+      "epoch": 0.005468132936387387,
+      "grad_norm": 0.24952976405620575,
+      "learning_rate": 9.999860619274639e-05,
+      "loss": 1.2167,
+      "step": 90
+    },
+    {
+      "epoch": 0.005528889969013914,
+      "grad_norm": 0.3820326030254364,
+      "learning_rate": 9.999853379942956e-05,
+      "loss": 1.304,
+      "step": 91
+    },
+    {
+      "epoch": 0.00558964700164044,
+      "grad_norm": 0.28423139452934265,
+      "learning_rate": 9.999845957340684e-05,
+      "loss": 1.1986,
+      "step": 92
+    },
+    {
+      "epoch": 0.005650404034266966,
+      "grad_norm": 10.655237197875977,
+      "learning_rate": 9.999838351468088e-05,
+      "loss": 1.1906,
+      "step": 93
+    },
+    {
+      "epoch": 0.005711161066893493,
+      "grad_norm": 0.5088858604431152,
+      "learning_rate": 9.999830562325453e-05,
+      "loss": 1.3004,
+      "step": 94
+    },
+    {
+      "epoch": 0.00577191809952002,
+      "grad_norm": 0.4481561481952667,
+      "learning_rate": 9.999822589913062e-05,
+      "loss": 1.2209,
+      "step": 95
+    },
+    {
+      "epoch": 0.005832675132146546,
+      "grad_norm": 0.23201344907283783,
+      "learning_rate": 9.999814434231207e-05,
+      "loss": 1.1614,
+      "step": 96
+    },
+    {
+      "epoch": 0.0058934321647730725,
+      "grad_norm": 0.5318001508712769,
+      "learning_rate": 9.999806095280189e-05,
+      "loss": 1.1645,
+      "step": 97
+    },
+    {
+      "epoch": 0.005954189197399599,
+      "grad_norm": 0.19209295511245728,
+      "learning_rate": 9.99979757306031e-05,
+      "loss": 1.1684,
+      "step": 98
+    },
+    {
+      "epoch": 0.006014946230026126,
+      "grad_norm": 0.4901897609233856,
+      "learning_rate": 9.999788867571887e-05,
+      "loss": 1.1833,
+      "step": 99
+    },
+    {
+      "epoch": 0.006075703262652652,
+      "grad_norm": 0.24492308497428894,
+      "learning_rate": 9.999779978815233e-05,
+      "loss": 1.2093,
+      "step": 100
+    },
+    {
+      "epoch": 0.0061364602952791785,
+      "grad_norm": 0.3037416636943817,
+      "learning_rate": 9.99977090679068e-05,
+      "loss": 1.1944,
+      "step": 101
+    },
+    {
+      "epoch": 0.006197217327905705,
+      "grad_norm": 0.38511979579925537,
+      "learning_rate": 9.999761651498559e-05,
+      "loss": 1.2449,
+      "step": 102
+    },
+    {
+      "epoch": 0.006257974360532232,
+      "grad_norm": 45.19050979614258,
+      "learning_rate": 9.999752212939206e-05,
+      "loss": 1.3232,
+      "step": 103
+    },
+    {
+      "epoch": 0.006318731393158758,
+      "grad_norm": 0.3553546369075775,
+      "learning_rate": 9.99974259111297e-05,
+      "loss": 1.2232,
+      "step": 104
+    },
+    {
+      "epoch": 0.006379488425785285,
+      "grad_norm": 0.23319929838180542,
+      "learning_rate": 9.999732786020203e-05,
+      "loss": 1.1975,
+      "step": 105
+    },
+    {
+      "epoch": 0.006440245458411811,
+      "grad_norm": 0.21722131967544556,
+      "learning_rate": 9.999722797661264e-05,
+      "loss": 1.2408,
+      "step": 106
+    },
+    {
+      "epoch": 0.006501002491038338,
+      "grad_norm": 0.3070448338985443,
+      "learning_rate": 9.99971262603652e-05,
+      "loss": 1.1694,
+      "step": 107
+    },
+    {
+      "epoch": 0.006561759523664864,
+      "grad_norm": 0.20213374495506287,
+      "learning_rate": 9.999702271146343e-05,
+      "loss": 1.2252,
+      "step": 108
+    },
+    {
+      "epoch": 0.006622516556291391,
+      "grad_norm": 0.872248649597168,
+      "learning_rate": 9.999691732991112e-05,
+      "loss": 1.1857,
+      "step": 109
+    },
+    {
+      "epoch": 0.006683273588917917,
+      "grad_norm": 0.2687833309173584,
+      "learning_rate": 9.999681011571215e-05,
+      "loss": 1.1348,
+      "step": 110
+    },
+    {
+      "epoch": 0.006744030621544444,
+      "grad_norm": 0.22979538142681122,
+      "learning_rate": 9.999670106887045e-05,
+      "loss": 1.319,
+      "step": 111
+    },
+    {
+      "epoch": 0.0068047876541709705,
+      "grad_norm": 0.3538469970226288,
+      "learning_rate": 9.999659018938998e-05,
+      "loss": 1.3412,
+      "step": 112
+    },
+    {
+      "epoch": 0.006865544686797497,
+      "grad_norm": 0.23582477867603302,
+      "learning_rate": 9.999647747727485e-05,
+      "loss": 1.1568,
+      "step": 113
+    },
+    {
+      "epoch": 0.006926301719424023,
+      "grad_norm": 0.2528184652328491,
+      "learning_rate": 9.999636293252919e-05,
+      "loss": 1.1557,
+      "step": 114
+    },
+    {
+      "epoch": 0.006987058752050549,
+      "grad_norm": 0.22521843016147614,
+      "learning_rate": 9.999624655515716e-05,
+      "loss": 1.1874,
+      "step": 115
+    },
+    {
+      "epoch": 0.007047815784677077,
+      "grad_norm": 0.26984483003616333,
+      "learning_rate": 9.999612834516305e-05,
+      "loss": 1.3546,
+      "step": 116
+    },
+    {
+      "epoch": 0.007108572817303603,
+      "grad_norm": 0.357096403837204,
+      "learning_rate": 9.99960083025512e-05,
+      "loss": 1.3785,
+      "step": 117
+    },
+    {
+      "epoch": 0.007169329849930129,
+      "grad_norm": 0.41239115595817566,
+      "learning_rate": 9.9995886427326e-05,
+      "loss": 1.1813,
+      "step": 118
+    },
+    {
+      "epoch": 0.0072300868825566555,
+      "grad_norm": 0.19457732141017914,
+      "learning_rate": 9.999576271949192e-05,
+      "loss": 1.2222,
+      "step": 119
+    },
+    {
+      "epoch": 0.007290843915183183,
+      "grad_norm": 0.298249214887619,
+      "learning_rate": 9.999563717905349e-05,
+      "loss": 1.2833,
+      "step": 120
+    },
+    {
+      "epoch": 0.007351600947809709,
+      "grad_norm": 0.3192598223686218,
+      "learning_rate": 9.999550980601531e-05,
+      "loss": 1.2762,
+      "step": 121
+    },
+    {
+      "epoch": 0.007412357980436235,
+      "grad_norm": 0.2641714811325073,
+      "learning_rate": 9.999538060038207e-05,
+      "loss": 1.2404,
+      "step": 122
+    },
+    {
+      "epoch": 0.007473115013062762,
+      "grad_norm": 0.21821774542331696,
+      "learning_rate": 9.999524956215848e-05,
+      "loss": 1.1696,
+      "step": 123
+    },
+    {
+      "epoch": 0.007533872045689289,
+      "grad_norm": 0.21299785375595093,
+      "learning_rate": 9.999511669134935e-05,
+      "loss": 1.1902,
+      "step": 124
+    },
+    {
+      "epoch": 0.007594629078315815,
+      "grad_norm": 0.2785874605178833,
+      "learning_rate": 9.999498198795955e-05,
+      "loss": 1.2195,
+      "step": 125
+    },
+    {
+      "epoch": 0.007655386110942341,
+      "grad_norm": 0.293491929769516,
+      "learning_rate": 9.999484545199405e-05,
+      "loss": 1.2029,
+      "step": 126
+    },
+    {
+      "epoch": 0.007716143143568868,
+      "grad_norm": 0.3122383952140808,
+      "learning_rate": 9.999470708345779e-05,
+      "loss": 1.1709,
+      "step": 127
+    },
+    {
+      "epoch": 0.007776900176195395,
+      "grad_norm": 0.3170028328895569,
+      "learning_rate": 9.99945668823559e-05,
+      "loss": 1.1759,
+      "step": 128
+    },
+    {
+      "epoch": 0.007837657208821921,
+      "grad_norm": 0.3421311676502228,
+      "learning_rate": 9.999442484869348e-05,
+      "loss": 1.3083,
+      "step": 129
+    },
+    {
+      "epoch": 0.007898414241448447,
+      "grad_norm": 0.2029871642589569,
+      "learning_rate": 9.999428098247576e-05,
+      "loss": 1.2978,
+      "step": 130
+    },
+    {
+      "epoch": 0.007959171274074974,
+      "grad_norm": 0.17570683360099792,
+      "learning_rate": 9.999413528370803e-05,
+      "loss": 1.1686,
+      "step": 131
+    },
+    {
+      "epoch": 0.0080199283067015,
+      "grad_norm": 0.1681767851114273,
+      "learning_rate": 9.999398775239559e-05,
+      "loss": 1.1594,
+      "step": 132
+    },
+    {
+      "epoch": 0.008080685339328026,
+      "grad_norm": 0.22254763543605804,
+      "learning_rate": 9.999383838854385e-05,
+      "loss": 1.1966,
+      "step": 133
+    },
+    {
+      "epoch": 0.008141442371954554,
+      "grad_norm": 0.39164310693740845,
+      "learning_rate": 9.999368719215831e-05,
+      "loss": 1.2692,
+      "step": 134
+    },
+    {
+      "epoch": 0.00820219940458108,
+      "grad_norm": 0.2754465341567993,
+      "learning_rate": 9.999353416324452e-05,
+      "loss": 1.2483,
+      "step": 135
+    },
+    {
+      "epoch": 0.008262956437207607,
+      "grad_norm": 0.33278337121009827,
+      "learning_rate": 9.999337930180805e-05,
+      "loss": 1.2654,
+      "step": 136
+    },
+    {
+      "epoch": 0.008323713469834133,
+      "grad_norm": 0.5331653356552124,
+      "learning_rate": 9.999322260785461e-05,
+      "loss": 1.3173,
+      "step": 137
+    },
+    {
+      "epoch": 0.00838447050246066,
+      "grad_norm": 0.387779176235199,
+      "learning_rate": 9.999306408138994e-05,
+      "loss": 1.1948,
+      "step": 138
+    },
+    {
+      "epoch": 0.008445227535087186,
+      "grad_norm": 0.434356153011322,
+      "learning_rate": 9.999290372241982e-05,
+      "loss": 1.1932,
+      "step": 139
+    },
+    {
+      "epoch": 0.008505984567713712,
+      "grad_norm": 53.942237854003906,
+      "learning_rate": 9.999274153095016e-05,
+      "loss": 1.2247,
+      "step": 140
+    },
+    {
+      "epoch": 0.008566741600340239,
+      "grad_norm": 0.5120449066162109,
+      "learning_rate": 9.99925775069869e-05,
+      "loss": 1.2842,
+      "step": 141
+    },
+    {
+      "epoch": 0.008627498632966767,
+      "grad_norm": 0.40591004490852356,
+      "learning_rate": 9.999241165053605e-05,
+      "loss": 1.2218,
+      "step": 142
+    },
+    {
+      "epoch": 0.008688255665593293,
+      "grad_norm": 0.35961678624153137,
+      "learning_rate": 9.999224396160367e-05,
+      "loss": 1.3417,
+      "step": 143
+    },
+    {
+      "epoch": 0.00874901269821982,
+      "grad_norm": 0.2422172874212265,
+      "learning_rate": 9.999207444019594e-05,
+      "loss": 1.2508,
+      "step": 144
+    },
+    {
+      "epoch": 0.008809769730846346,
+      "grad_norm": 0.21179278194904327,
+      "learning_rate": 9.999190308631906e-05,
+      "loss": 1.2071,
+      "step": 145
+    },
+    {
+      "epoch": 0.008870526763472872,
+      "grad_norm": 0.22785666584968567,
+      "learning_rate": 9.999172989997931e-05,
+      "loss": 1.2072,
+      "step": 146
+    },
+    {
+      "epoch": 0.008931283796099398,
+      "grad_norm": 0.22191546857357025,
+      "learning_rate": 9.999155488118302e-05,
+      "loss": 1.1528,
+      "step": 147
+    },
+    {
+      "epoch": 0.008992040828725924,
+      "grad_norm": 0.4478309750556946,
+      "learning_rate": 9.999137802993664e-05,
+      "loss": 1.168,
+      "step": 148
+    },
+    {
+      "epoch": 0.00905279786135245,
+      "grad_norm": 0.2741773724555969,
+      "learning_rate": 9.999119934624663e-05,
+      "loss": 1.2232,
+      "step": 149
+    },
+    {
+      "epoch": 0.009113554893978979,
+      "grad_norm": 0.2959984242916107,
+      "learning_rate": 9.999101883011954e-05,
+      "loss": 1.2353,
+      "step": 150
+    },
+    {
+      "epoch": 0.009174311926605505,
+      "grad_norm": 0.24366505444049835,
+      "learning_rate": 9.9990836481562e-05,
+      "loss": 1.201,
+      "step": 151
+    },
+    {
+      "epoch": 0.009235068959232031,
+      "grad_norm": 0.27018189430236816,
+      "learning_rate": 9.99906523005807e-05,
+      "loss": 1.2382,
+      "step": 152
+    },
+    {
+      "epoch": 0.009295825991858558,
+      "grad_norm": 0.5289650559425354,
+      "learning_rate": 9.999046628718234e-05,
+      "loss": 1.2516,
+      "step": 153
+    },
+    {
+      "epoch": 0.009356583024485084,
+      "grad_norm": 0.23208869993686676,
+      "learning_rate": 9.99902784413738e-05,
+      "loss": 1.185,
+      "step": 154
+    },
+    {
+      "epoch": 0.00941734005711161,
+      "grad_norm": 0.28683531284332275,
+      "learning_rate": 9.999008876316194e-05,
+      "loss": 1.2603,
+      "step": 155
+    },
+    {
+      "epoch": 0.009478097089738137,
+      "grad_norm": 0.2525317072868347,
+      "learning_rate": 9.99898972525537e-05,
+      "loss": 1.1543,
+      "step": 156
+    },
+    {
+      "epoch": 0.009538854122364663,
+      "grad_norm": 0.5868335366249084,
+      "learning_rate": 9.998970390955613e-05,
+      "loss": 1.232,
+      "step": 157
+    },
+    {
+      "epoch": 0.009599611154991191,
+      "grad_norm": 0.2296774834394455,
+      "learning_rate": 9.99895087341763e-05,
+      "loss": 1.1516,
+      "step": 158
+    },
+    {
+      "epoch": 0.009660368187617717,
+      "grad_norm": 0.3657951354980469,
+      "learning_rate": 9.998931172642134e-05,
+      "loss": 1.1734,
+      "step": 159
+    },
+    {
+      "epoch": 0.009721125220244244,
+      "grad_norm": 0.2197127789258957,
+      "learning_rate": 9.998911288629852e-05,
+      "loss": 1.2657,
+      "step": 160
+    },
+    {
+      "epoch": 0.00978188225287077,
+      "grad_norm": 0.18814058601856232,
+      "learning_rate": 9.99889122138151e-05,
+      "loss": 1.181,
+      "step": 161
+    },
+    {
+      "epoch": 0.009842639285497296,
+      "grad_norm": 0.19529810547828674,
+      "learning_rate": 9.998870970897844e-05,
+      "loss": 1.1395,
+      "step": 162
+    },
+    {
+      "epoch": 0.009903396318123822,
+      "grad_norm": 2.752793788909912,
+      "learning_rate": 9.998850537179596e-05,
+      "loss": 1.1917,
+      "step": 163
+    },
+    {
+      "epoch": 0.009964153350750349,
+      "grad_norm": 0.29561033844947815,
+      "learning_rate": 9.998829920227514e-05,
+      "loss": 1.2864,
+      "step": 164
+    },
+    {
+      "epoch": 0.010024910383376875,
+      "grad_norm": 0.4728116989135742,
+      "learning_rate": 9.998809120042358e-05,
+      "loss": 1.1622,
+      "step": 165
+    },
+    {
+      "epoch": 0.010085667416003403,
+      "grad_norm": 0.3741132318973541,
+      "learning_rate": 9.998788136624885e-05,
+      "loss": 1.2486,
+      "step": 166
+    },
+    {
+      "epoch": 0.01014642444862993,
+      "grad_norm": 0.32585614919662476,
+      "learning_rate": 9.998766969975868e-05,
+      "loss": 1.1889,
+      "step": 167
+    },
+    {
+      "epoch": 0.010207181481256456,
+      "grad_norm": 0.38850677013397217,
+      "learning_rate": 9.99874562009608e-05,
+      "loss": 1.139,
+      "step": 168
+    },
+    {
+      "epoch": 0.010267938513882982,
+      "grad_norm": 0.23338811099529266,
+      "learning_rate": 9.998724086986307e-05,
+      "loss": 1.1367,
+      "step": 169
+    },
+    {
+      "epoch": 0.010328695546509508,
+      "grad_norm": 0.3551323711872101,
+      "learning_rate": 9.998702370647335e-05,
+      "loss": 1.2448,
+      "step": 170
+    },
+    {
+      "epoch": 0.010389452579136035,
+      "grad_norm": 0.3741970956325531,
+      "learning_rate": 9.998680471079962e-05,
+      "loss": 1.2648,
+      "step": 171
+    },
+    {
+      "epoch": 0.010450209611762561,
+      "grad_norm": 0.31997615098953247,
+      "learning_rate": 9.99865838828499e-05,
+      "loss": 1.2447,
+      "step": 172
+    },
+    {
+      "epoch": 0.010510966644389087,
+      "grad_norm": 0.24562925100326538,
+      "learning_rate": 9.998636122263228e-05,
+      "loss": 1.1752,
+      "step": 173
+    },
+    {
+      "epoch": 0.010571723677015615,
+      "grad_norm": 0.44380733370780945,
+      "learning_rate": 9.998613673015494e-05,
+      "loss": 1.263,
+      "step": 174
+    },
+    {
+      "epoch": 0.010632480709642142,
+      "grad_norm": 0.2343125343322754,
+      "learning_rate": 9.998591040542608e-05,
+      "loss": 1.193,
+      "step": 175
+    },
+    {
+      "epoch": 0.010693237742268668,
+      "grad_norm": 0.3768921196460724,
+      "learning_rate": 9.998568224845402e-05,
+      "loss": 1.2294,
+      "step": 176
+    },
+    {
+      "epoch": 0.010753994774895194,
+      "grad_norm": 0.2599649131298065,
+      "learning_rate": 9.998545225924711e-05,
+      "loss": 1.211,
+      "step": 177
+    },
+    {
+      "epoch": 0.01081475180752172,
+      "grad_norm": 0.35759758949279785,
+      "learning_rate": 9.99852204378138e-05,
+      "loss": 1.2009,
+      "step": 178
+    },
+    {
+      "epoch": 0.010875508840148247,
+      "grad_norm": 0.18380869925022125,
+      "learning_rate": 9.998498678416254e-05,
+      "loss": 1.1803,
+      "step": 179
+    },
+    {
+      "epoch": 0.010936265872774773,
+      "grad_norm": 0.2130039930343628,
+      "learning_rate": 9.998475129830196e-05,
+      "loss": 1.1625,
+      "step": 180
+    },
+    {
+      "epoch": 0.0109970229054013,
+      "grad_norm": 0.23620057106018066,
+      "learning_rate": 9.998451398024066e-05,
+      "loss": 1.2213,
+      "step": 181
+    },
+    {
+      "epoch": 0.011057779938027828,
+      "grad_norm": 0.3125370740890503,
+      "learning_rate": 9.998427482998732e-05,
+      "loss": 1.1596,
+      "step": 182
+    },
+    {
+      "epoch": 0.011118536970654354,
+      "grad_norm": 0.21282735466957092,
+      "learning_rate": 9.998403384755074e-05,
+      "loss": 1.1924,
+      "step": 183
+    },
+    {
+      "epoch": 0.01117929400328088,
+      "grad_norm": 0.19769485294818878,
+      "learning_rate": 9.998379103293973e-05,
+      "loss": 1.1965,
+      "step": 184
+    },
+    {
+      "epoch": 0.011240051035907406,
+      "grad_norm": 0.21187029778957367,
+      "learning_rate": 9.99835463861632e-05,
+      "loss": 1.1526,
+      "step": 185
+    },
+    {
+      "epoch": 0.011300808068533933,
+      "grad_norm": 0.39983147382736206,
+      "learning_rate": 9.998329990723011e-05,
+      "loss": 1.2193,
+      "step": 186
+    },
+    {
+      "epoch": 0.011361565101160459,
+      "grad_norm": 0.3108305335044861,
+      "learning_rate": 9.998305159614951e-05,
+      "loss": 1.2,
+      "step": 187
+    },
+    {
+      "epoch": 0.011422322133786985,
+      "grad_norm": 3.6537888050079346,
+      "learning_rate": 9.998280145293049e-05,
+      "loss": 1.2986,
+      "step": 188
+    },
+    {
+      "epoch": 0.011483079166413512,
+      "grad_norm": 0.378499835729599,
+      "learning_rate": 9.998254947758221e-05,
+      "loss": 1.2226,
+      "step": 189
+    },
+    {
+      "epoch": 0.01154383619904004,
+      "grad_norm": 0.2172241359949112,
+      "learning_rate": 9.998229567011393e-05,
+      "loss": 1.1442,
+      "step": 190
+    },
+    {
+      "epoch": 0.011604593231666566,
+      "grad_norm": 0.2371995598077774,
+      "learning_rate": 9.998204003053492e-05,
+      "loss": 1.2858,
+      "step": 191
+    },
+    {
+      "epoch": 0.011665350264293092,
+      "grad_norm": 0.2743389904499054,
+      "learning_rate": 9.99817825588546e-05,
+      "loss": 1.2653,
+      "step": 192
+    },
+    {
+      "epoch": 0.011726107296919619,
+      "grad_norm": 0.4419936239719391,
+      "learning_rate": 9.998152325508236e-05,
+      "loss": 1.241,
+      "step": 193
+    },
+    {
+      "epoch": 0.011786864329546145,
+      "grad_norm": 2.6996798515319824,
+      "learning_rate": 9.998126211922773e-05,
+      "loss": 1.2827,
+      "step": 194
+    },
+    {
+      "epoch": 0.011847621362172671,
+      "grad_norm": 0.5182858109474182,
+      "learning_rate": 9.998099915130029e-05,
+      "loss": 1.248,
+      "step": 195
+    },
+    {
+      "epoch": 0.011908378394799198,
+      "grad_norm": 0.22894130647182465,
+      "learning_rate": 9.998073435130963e-05,
+      "loss": 1.1438,
+      "step": 196
+    },
+    {
+      "epoch": 0.011969135427425724,
+      "grad_norm": 0.592929482460022,
+      "learning_rate": 9.998046771926553e-05,
+      "loss": 1.3242,
+      "step": 197
+    },
+    {
+      "epoch": 0.012029892460052252,
+      "grad_norm": 0.21945858001708984,
+      "learning_rate": 9.99801992551777e-05,
+      "loss": 1.1592,
+      "step": 198
+    },
+    {
+      "epoch": 0.012090649492678778,
+      "grad_norm": 0.40205755829811096,
+      "learning_rate": 9.997992895905601e-05,
+      "loss": 1.2715,
+      "step": 199
+    },
+    {
+      "epoch": 0.012151406525305304,
+      "grad_norm": 31.714807510375977,
+      "learning_rate": 9.997965683091037e-05,
+      "loss": 1.1869,
+      "step": 200
+    },
+    {
+      "epoch": 0.01221216355793183,
+      "grad_norm": 0.3603743016719818,
+      "learning_rate": 9.997938287075076e-05,
+      "loss": 1.2231,
+      "step": 201
+    },
+    {
+      "epoch": 0.012272920590558357,
+      "grad_norm": 0.33039894700050354,
+      "learning_rate": 9.997910707858718e-05,
+      "loss": 1.2133,
+      "step": 202
+    },
+    {
+      "epoch": 0.012333677623184883,
+      "grad_norm": 0.22338835895061493,
+      "learning_rate": 9.997882945442978e-05,
+      "loss": 1.1859,
+      "step": 203
+    },
+    {
+      "epoch": 0.01239443465581141,
+      "grad_norm": 0.41773277521133423,
+      "learning_rate": 9.997854999828873e-05,
+      "loss": 1.2061,
+      "step": 204
+    },
+    {
+      "epoch": 0.012455191688437936,
+      "grad_norm": 0.2586667537689209,
+      "learning_rate": 9.997826871017427e-05,
+      "loss": 1.2613,
+      "step": 205
+    },
+    {
+      "epoch": 0.012515948721064464,
+      "grad_norm": 0.4832109212875366,
+      "learning_rate": 9.99779855900967e-05,
+      "loss": 1.1931,
+      "step": 206
+    },
+    {
+      "epoch": 0.01257670575369099,
+      "grad_norm": 0.19160710275173187,
+      "learning_rate": 9.997770063806641e-05,
+      "loss": 1.1408,
+      "step": 207
+    },
+    {
+      "epoch": 0.012637462786317517,
+      "grad_norm": 0.36751052737236023,
+      "learning_rate": 9.997741385409385e-05,
+      "loss": 1.2612,
+      "step": 208
+    },
+    {
+      "epoch": 0.012698219818944043,
+      "grad_norm": 1.9435746669769287,
+      "learning_rate": 9.997712523818952e-05,
+      "loss": 1.1441,
+      "step": 209
+    },
+    {
+      "epoch": 0.01275897685157057,
+      "grad_norm": 0.3401590585708618,
+      "learning_rate": 9.997683479036401e-05,
+      "loss": 1.1435,
+      "step": 210
+    },
+    {
+      "epoch": 0.012819733884197096,
+      "grad_norm": 2.203845977783203,
+      "learning_rate": 9.997654251062795e-05,
+      "loss": 1.1754,
+      "step": 211
+    },
+    {
+      "epoch": 0.012880490916823622,
+      "grad_norm": 2.6514179706573486,
+      "learning_rate": 9.997624839899208e-05,
+      "loss": 1.1724,
+      "step": 212
+    },
+    {
+      "epoch": 0.012941247949450148,
+      "grad_norm": 0.378422349691391,
+      "learning_rate": 9.997595245546716e-05,
+      "loss": 1.1723,
+      "step": 213
+    },
+    {
+      "epoch": 0.013002004982076676,
+      "grad_norm": 0.21302397549152374,
+      "learning_rate": 9.997565468006404e-05,
+      "loss": 1.2132,
+      "step": 214
+    },
+    {
+      "epoch": 0.013062762014703203,
+      "grad_norm": 0.5641165971755981,
+      "learning_rate": 9.997535507279363e-05,
+      "loss": 1.2206,
+      "step": 215
+    },
+    {
+      "epoch": 0.013123519047329729,
+      "grad_norm": 0.22931545972824097,
+      "learning_rate": 9.997505363366693e-05,
+      "loss": 1.1798,
+      "step": 216
+    },
+    {
+      "epoch": 0.013184276079956255,
+      "grad_norm": 0.2969755530357361,
+      "learning_rate": 9.997475036269497e-05,
+      "loss": 1.1653,
+      "step": 217
+    },
+    {
+      "epoch": 0.013245033112582781,
+      "grad_norm": 0.318168580532074,
+      "learning_rate": 9.997444525988888e-05,
+      "loss": 1.2642,
+      "step": 218
+    },
+    {
+      "epoch": 0.013305790145209308,
+      "grad_norm": 0.22343435883522034,
+      "learning_rate": 9.997413832525984e-05,
+      "loss": 1.2654,
+      "step": 219
+    },
+    {
+      "epoch": 0.013366547177835834,
+      "grad_norm": 0.25345805287361145,
+      "learning_rate": 9.997382955881911e-05,
+      "loss": 1.2097,
+      "step": 220
+    },
+    {
+      "epoch": 0.01342730421046236,
+      "grad_norm": 0.24147434532642365,
+      "learning_rate": 9.997351896057799e-05,
+      "loss": 1.1422,
+      "step": 221
+    },
+    {
+      "epoch": 0.013488061243088888,
+      "grad_norm": 0.29657846689224243,
+      "learning_rate": 9.997320653054787e-05,
+      "loss": 1.1875,
+      "step": 222
+    },
+    {
+      "epoch": 0.013548818275715415,
+      "grad_norm": 0.22879545390605927,
+      "learning_rate": 9.99728922687402e-05,
+      "loss": 1.2242,
+      "step": 223
+    },
+    {
+      "epoch": 0.013609575308341941,
+      "grad_norm": 0.19556359946727753,
+      "learning_rate": 9.997257617516651e-05,
+      "loss": 1.189,
+      "step": 224
+    },
+    {
+      "epoch": 0.013670332340968467,
+      "grad_norm": 0.3813997507095337,
+      "learning_rate": 9.997225824983838e-05,
+      "loss": 1.2457,
+      "step": 225
+    },
+    {
+      "epoch": 0.013731089373594994,
+      "grad_norm": 0.17610788345336914,
+      "learning_rate": 9.997193849276746e-05,
+      "loss": 1.1453,
+      "step": 226
+    },
+    {
+      "epoch": 0.01379184640622152,
+      "grad_norm": 0.38481786847114563,
+      "learning_rate": 9.997161690396549e-05,
+      "loss": 1.2695,
+      "step": 227
+    },
+    {
+      "epoch": 0.013852603438848046,
+      "grad_norm": 0.19644692540168762,
+      "learning_rate": 9.997129348344423e-05,
+      "loss": 1.1992,
+      "step": 228
+    },
+    {
+      "epoch": 0.013913360471474573,
+      "grad_norm": 0.2646336853504181,
+      "learning_rate": 9.997096823121555e-05,
+      "loss": 1.2049,
+      "step": 229
+    },
+    {
+      "epoch": 0.013974117504101099,
+      "grad_norm": 0.22506456077098846,
+      "learning_rate": 9.997064114729138e-05,
+      "loss": 1.1622,
+      "step": 230
+    },
+    {
+      "epoch": 0.014034874536727627,
+      "grad_norm": 0.3764677345752716,
+      "learning_rate": 9.997031223168368e-05,
+      "loss": 1.2147,
+      "step": 231
+    },
+    {
+      "epoch": 0.014095631569354153,
+      "grad_norm": 0.3747536242008209,
+      "learning_rate": 9.996998148440454e-05,
+      "loss": 1.1736,
+      "step": 232
+    },
+    {
+      "epoch": 0.01415638860198068,
+      "grad_norm": 0.26718929409980774,
+      "learning_rate": 9.996964890546607e-05,
+      "loss": 1.1771,
+      "step": 233
+    },
+    {
+      "epoch": 0.014217145634607206,
+      "grad_norm": 0.6982794404029846,
+      "learning_rate": 9.996931449488045e-05,
+      "loss": 1.2514,
+      "step": 234
+    },
+    {
+      "epoch": 0.014277902667233732,
+      "grad_norm": 0.4008491635322571,
+      "learning_rate": 9.996897825265996e-05,
+      "loss": 1.2765,
+      "step": 235
+    },
+    {
+      "epoch": 0.014338659699860258,
+      "grad_norm": 0.21541957557201385,
+      "learning_rate": 9.99686401788169e-05,
+      "loss": 1.1965,
+      "step": 236
+    },
+    {
+      "epoch": 0.014399416732486785,
+      "grad_norm": 0.20242953300476074,
+      "learning_rate": 9.996830027336368e-05,
+      "loss": 1.2024,
+      "step": 237
+    },
+    {
+      "epoch": 0.014460173765113311,
+      "grad_norm": 0.41439831256866455,
+      "learning_rate": 9.996795853631278e-05,
+      "loss": 1.2789,
+      "step": 238
+    },
+    {
+      "epoch": 0.014520930797739839,
+      "grad_norm": 0.23059511184692383,
+      "learning_rate": 9.996761496767668e-05,
+      "loss": 1.1712,
+      "step": 239
+    },
+    {
+      "epoch": 0.014581687830366365,
+      "grad_norm": 0.19941109418869019,
+      "learning_rate": 9.9967269567468e-05,
+      "loss": 1.1966,
+      "step": 240
+    },
+    {
+      "epoch": 0.014642444862992892,
+      "grad_norm": 0.26879042387008667,
+      "learning_rate": 9.996692233569938e-05,
+      "loss": 1.2386,
+      "step": 241
+    },
+    {
+      "epoch": 0.014703201895619418,
+      "grad_norm": 0.19818615913391113,
+      "learning_rate": 9.996657327238357e-05,
+      "loss": 1.1954,
+      "step": 242
+    },
+    {
+      "epoch": 0.014763958928245944,
+      "grad_norm": 0.22947187721729279,
+      "learning_rate": 9.996622237753336e-05,
+      "loss": 1.1885,
+      "step": 243
+    },
+    {
+      "epoch": 0.01482471596087247,
+      "grad_norm": 0.458009272813797,
+      "learning_rate": 9.996586965116161e-05,
+      "loss": 1.3335,
+      "step": 244
+    },
+    {
+      "epoch": 0.014885472993498997,
+      "grad_norm": 0.30162313580513,
+      "learning_rate": 9.996551509328127e-05,
+      "loss": 1.23,
+      "step": 245
+    },
+    {
+      "epoch": 0.014946230026125523,
+      "grad_norm": 0.2162998914718628,
+      "learning_rate": 9.996515870390529e-05,
+      "loss": 1.1318,
+      "step": 246
+    },
+    {
+      "epoch": 0.015006987058752051,
+      "grad_norm": 0.2238757312297821,
+      "learning_rate": 9.996480048304677e-05,
+      "loss": 1.1285,
+      "step": 247
+    },
+    {
+      "epoch": 0.015067744091378578,
+      "grad_norm": 0.20800887048244476,
+      "learning_rate": 9.996444043071883e-05,
+      "loss": 1.1329,
+      "step": 248
+    },
+    {
+      "epoch": 0.015128501124005104,
+      "grad_norm": 0.18456752598285675,
+      "learning_rate": 9.996407854693467e-05,
+      "loss": 1.1904,
+      "step": 249
+    },
+    {
+      "epoch": 0.01518925815663163,
+      "grad_norm": 0.20270679891109467,
+      "learning_rate": 9.996371483170754e-05,
+      "loss": 1.1364,
+      "step": 250
+    },
+    {
+      "epoch": 0.015250015189258156,
+      "grad_norm": 0.3495061993598938,
+      "learning_rate": 9.99633492850508e-05,
+      "loss": 1.3883,
+      "step": 251
+    },
+    {
+      "epoch": 0.015310772221884683,
+      "grad_norm": 0.26616501808166504,
+      "learning_rate": 9.996298190697783e-05,
+      "loss": 1.2393,
+      "step": 252
+    },
+    {
+      "epoch": 0.015371529254511209,
+      "grad_norm": 0.20503906905651093,
+      "learning_rate": 9.996261269750211e-05,
+      "loss": 1.1654,
+      "step": 253
+    },
+    {
+      "epoch": 0.015432286287137735,
+      "grad_norm": 0.26730602979660034,
+      "learning_rate": 9.996224165663714e-05,
+      "loss": 1.2569,
+      "step": 254
+    },
+    {
+      "epoch": 0.015493043319764263,
+      "grad_norm": 0.47674667835235596,
+      "learning_rate": 9.996186878439657e-05,
+      "loss": 1.3208,
+      "step": 255
+    },
+    {
+      "epoch": 0.01555380035239079,
+      "grad_norm": 0.26248225569725037,
+      "learning_rate": 9.996149408079403e-05,
+      "loss": 1.2583,
+      "step": 256
+    },
+    {
+      "epoch": 0.015614557385017316,
+      "grad_norm": 0.4549104571342468,
+      "learning_rate": 9.996111754584327e-05,
+      "loss": 1.241,
+      "step": 257
+    },
+    {
+      "epoch": 0.015675314417643842,
+      "grad_norm": 0.21723133325576782,
+      "learning_rate": 9.996073917955808e-05,
+      "loss": 1.1343,
+      "step": 258
+    },
+    {
+      "epoch": 0.01573607145027037,
+      "grad_norm": 0.48765337467193604,
+      "learning_rate": 9.996035898195236e-05,
+      "loss": 1.1317,
+      "step": 259
+    },
+    {
+      "epoch": 0.015796828482896895,
+      "grad_norm": 0.26611945033073425,
+      "learning_rate": 9.995997695304e-05,
+      "loss": 1.1473,
+      "step": 260
+    },
+    {
+      "epoch": 0.015857585515523423,
+      "grad_norm": 0.4465886354446411,
+      "learning_rate": 9.995959309283505e-05,
+      "loss": 1.3175,
+      "step": 261
+    },
+    {
+      "epoch": 0.015918342548149948,
+      "grad_norm": 2.0013575553894043,
+      "learning_rate": 9.995920740135154e-05,
+      "loss": 1.1756,
+      "step": 262
+    },
+    {
+      "epoch": 0.015979099580776476,
+      "grad_norm": 0.2904813587665558,
+      "learning_rate": 9.995881987860363e-05,
+      "loss": 1.14,
+      "step": 263
+    },
+    {
+      "epoch": 0.016039856613403,
+      "grad_norm": 0.27571505308151245,
+      "learning_rate": 9.995843052460552e-05,
+      "loss": 1.2668,
+      "step": 264
+    },
+    {
+      "epoch": 0.016100613646029528,
+      "grad_norm": 10.14664077758789,
+      "learning_rate": 9.995803933937149e-05,
+      "loss": 1.233,
+      "step": 265
+    },
+    {
+      "epoch": 0.016161370678656053,
+      "grad_norm": 0.2483414113521576,
+      "learning_rate": 9.995764632291586e-05,
+      "loss": 1.1974,
+      "step": 266
+    },
+    {
+      "epoch": 0.01622212771128258,
+      "grad_norm": 0.2465965896844864,
+      "learning_rate": 9.995725147525306e-05,
+      "loss": 1.2868,
+      "step": 267
+    },
+    {
+      "epoch": 0.01628288474390911,
+      "grad_norm": 0.3032623827457428,
+      "learning_rate": 9.995685479639754e-05,
+      "loss": 1.1648,
+      "step": 268
+    },
+    {
+      "epoch": 0.016343641776535633,
+      "grad_norm": 0.17899607121944427,
+      "learning_rate": 9.995645628636385e-05,
+      "loss": 1.1919,
+      "step": 269
+    },
+    {
+      "epoch": 0.01640439880916216,
+      "grad_norm": 0.19266390800476074,
+      "learning_rate": 9.995605594516661e-05,
+      "loss": 1.1668,
+      "step": 270
+    },
+    {
+      "epoch": 0.016465155841788686,
+      "grad_norm": 0.3554879128932953,
+      "learning_rate": 9.995565377282048e-05,
+      "loss": 1.1938,
+      "step": 271
+    },
+    {
+      "epoch": 0.016525912874415214,
+      "grad_norm": 0.23894625902175903,
+      "learning_rate": 9.995524976934018e-05,
+      "loss": 1.1835,
+      "step": 272
+    },
+    {
+      "epoch": 0.01658666990704174,
+      "grad_norm": 0.17239512503147125,
+      "learning_rate": 9.995484393474057e-05,
+      "loss": 1.1438,
+      "step": 273
+    },
+    {
+      "epoch": 0.016647426939668267,
+      "grad_norm": 0.23845449090003967,
+      "learning_rate": 9.995443626903651e-05,
+      "loss": 1.2587,
+      "step": 274
+    },
+    {
+      "epoch": 0.016708183972294795,
+      "grad_norm": 0.31088948249816895,
+      "learning_rate": 9.995402677224291e-05,
+      "loss": 1.1911,
+      "step": 275
+    },
+    {
+      "epoch": 0.01676894100492132,
+      "grad_norm": 0.2854573428630829,
+      "learning_rate": 9.995361544437481e-05,
+      "loss": 1.2585,
+      "step": 276
+    },
+    {
+      "epoch": 0.016829698037547847,
+      "grad_norm": 0.5123035311698914,
+      "learning_rate": 9.995320228544727e-05,
+      "loss": 1.4816,
+      "step": 277
+    },
+    {
+      "epoch": 0.016890455070174372,
+      "grad_norm": 0.2312437742948532,
+      "learning_rate": 9.995278729547546e-05,
+      "loss": 1.1743,
+      "step": 278
+    },
+    {
+      "epoch": 0.0169512121028009,
+      "grad_norm": 0.6658782958984375,
+      "learning_rate": 9.995237047447458e-05,
+      "loss": 1.3736,
+      "step": 279
+    },
+    {
+      "epoch": 0.017011969135427425,
+      "grad_norm": 0.45647940039634705,
+      "learning_rate": 9.995195182245988e-05,
+      "loss": 1.2589,
+      "step": 280
+    },
+    {
+      "epoch": 0.017072726168053953,
+      "grad_norm": 0.37576019763946533,
+      "learning_rate": 9.995153133944676e-05,
+      "loss": 1.3273,
+      "step": 281
+    },
+    {
+      "epoch": 0.017133483200680477,
+      "grad_norm": 0.3769761323928833,
+      "learning_rate": 9.995110902545058e-05,
+      "loss": 1.2635,
+      "step": 282
+    },
+    {
+      "epoch": 0.017194240233307005,
+      "grad_norm": 0.7584442496299744,
+      "learning_rate": 9.995068488048685e-05,
+      "loss": 1.1655,
+      "step": 283
+    },
+    {
+      "epoch": 0.017254997265933533,
+      "grad_norm": 0.5158006548881531,
+      "learning_rate": 9.995025890457113e-05,
+      "loss": 1.1605,
+      "step": 284
+    },
+    {
+      "epoch": 0.017315754298560058,
+      "grad_norm": 0.3699159026145935,
+      "learning_rate": 9.994983109771898e-05,
+      "loss": 1.1622,
+      "step": 285
+    },
+    {
+      "epoch": 0.017376511331186586,
+      "grad_norm": 0.23761245608329773,
+      "learning_rate": 9.994940145994614e-05,
+      "loss": 1.1539,
+      "step": 286
+    },
+    {
+      "epoch": 0.01743726836381311,
+      "grad_norm": 0.47396376729011536,
+      "learning_rate": 9.994896999126832e-05,
+      "loss": 1.127,
+      "step": 287
+    },
+    {
+      "epoch": 0.01749802539643964,
+      "grad_norm": 0.20256276428699493,
+      "learning_rate": 9.994853669170136e-05,
+      "loss": 1.1673,
+      "step": 288
+    },
+    {
+      "epoch": 0.017558782429066163,
+      "grad_norm": 0.4062318205833435,
+      "learning_rate": 9.994810156126114e-05,
+      "loss": 1.1567,
+      "step": 289
+    },
+    {
+      "epoch": 0.01761953946169269,
+      "grad_norm": 0.41786348819732666,
+      "learning_rate": 9.994766459996359e-05,
+      "loss": 1.2449,
+      "step": 290
+    },
+    {
+      "epoch": 0.01768029649431922,
+      "grad_norm": 0.28585192561149597,
+      "learning_rate": 9.994722580782475e-05,
+      "loss": 1.168,
+      "step": 291
+    },
+    {
+      "epoch": 0.017741053526945744,
+      "grad_norm": 0.3699677586555481,
+      "learning_rate": 9.994678518486067e-05,
+      "loss": 1.1523,
+      "step": 292
+    },
+    {
+      "epoch": 0.01780181055957227,
+      "grad_norm": 0.300614595413208,
+      "learning_rate": 9.994634273108755e-05,
+      "loss": 1.1457,
+      "step": 293
+    },
+    {
+      "epoch": 0.017862567592198796,
+      "grad_norm": 0.33786723017692566,
+      "learning_rate": 9.994589844652157e-05,
+      "loss": 1.154,
+      "step": 294
+    },
+    {
+      "epoch": 0.017923324624825324,
+      "grad_norm": 0.43694934248924255,
+      "learning_rate": 9.994545233117903e-05,
+      "loss": 1.3118,
+      "step": 295
+    },
+    {
+      "epoch": 0.01798408165745185,
+      "grad_norm": 0.5097596645355225,
+      "learning_rate": 9.994500438507628e-05,
+      "loss": 1.3259,
+      "step": 296
+    },
+    {
+      "epoch": 0.018044838690078377,
+      "grad_norm": 0.3348499536514282,
+      "learning_rate": 9.994455460822975e-05,
+      "loss": 1.1655,
+      "step": 297
+    },
+    {
+      "epoch": 0.0181055957227049,
+      "grad_norm": 33.572723388671875,
+      "learning_rate": 9.994410300065591e-05,
+      "loss": 1.1179,
+      "step": 298
+    },
+    {
+      "epoch": 0.01816635275533143,
+      "grad_norm": 0.3271271288394928,
+      "learning_rate": 9.994364956237133e-05,
+      "loss": 1.3444,
+      "step": 299
+    },
+    {
+      "epoch": 0.018227109787957958,
+      "grad_norm": 0.27192381024360657,
+      "learning_rate": 9.994319429339259e-05,
+      "loss": 1.3966,
+      "step": 300
+    },
+    {
+      "epoch": 0.018287866820584482,
+      "grad_norm": 0.2065957635641098,
+      "learning_rate": 9.994273719373643e-05,
+      "loss": 1.1691,
+      "step": 301
+    },
+    {
+      "epoch": 0.01834862385321101,
+      "grad_norm": 0.20475585758686066,
+      "learning_rate": 9.99422782634196e-05,
+      "loss": 1.1543,
+      "step": 302
+    },
+    {
+      "epoch": 0.018409380885837535,
+      "grad_norm": 0.2624063193798065,
+      "learning_rate": 9.994181750245887e-05,
+      "loss": 1.199,
+      "step": 303
+    },
+    {
+      "epoch": 0.018470137918464063,
+      "grad_norm": 2.3660545349121094,
+      "learning_rate": 9.99413549108712e-05,
+      "loss": 1.1645,
+      "step": 304
+    },
+    {
+      "epoch": 0.018530894951090587,
+      "grad_norm": 0.3732457160949707,
+      "learning_rate": 9.99408904886735e-05,
+      "loss": 1.2981,
+      "step": 305
+    },
+    {
+      "epoch": 0.018591651983717115,
+      "grad_norm": 0.23340417444705963,
+      "learning_rate": 9.99404242358828e-05,
+      "loss": 1.1844,
+      "step": 306
+    },
+    {
+      "epoch": 0.018652409016343643,
+      "grad_norm": 0.8010504245758057,
+      "learning_rate": 9.993995615251622e-05,
+      "loss": 1.1564,
+      "step": 307
+    },
+    {
+      "epoch": 0.018713166048970168,
+      "grad_norm": 0.33755290508270264,
+      "learning_rate": 9.993948623859086e-05,
+      "loss": 1.2193,
+      "step": 308
+    },
+    {
+      "epoch": 0.018773923081596696,
+      "grad_norm": 0.7690300345420837,
+      "learning_rate": 9.993901449412399e-05,
+      "loss": 1.3164,
+      "step": 309
+    },
+    {
+      "epoch": 0.01883468011422322,
+      "grad_norm": 0.22133295238018036,
+      "learning_rate": 9.993854091913291e-05,
+      "loss": 1.1487,
+      "step": 310
+    },
+    {
+      "epoch": 0.01889543714684975,
+      "grad_norm": 0.42820850014686584,
+      "learning_rate": 9.993806551363494e-05,
+      "loss": 1.1758,
+      "step": 311
+    },
+    {
+      "epoch": 0.018956194179476273,
+      "grad_norm": 0.5701407194137573,
+      "learning_rate": 9.993758827764752e-05,
+      "loss": 1.41,
+      "step": 312
+    },
+    {
+      "epoch": 0.0190169512121028,
+      "grad_norm": 0.2954898178577423,
+      "learning_rate": 9.993710921118816e-05,
+      "loss": 1.3422,
+      "step": 313
+    },
+    {
+      "epoch": 0.019077708244729326,
+      "grad_norm": 0.34081703424453735,
+      "learning_rate": 9.993662831427442e-05,
+      "loss": 1.1599,
+      "step": 314
+    },
+    {
+      "epoch": 0.019138465277355854,
+      "grad_norm": 0.47216105461120605,
+      "learning_rate": 9.99361455869239e-05,
+      "loss": 1.291,
+      "step": 315
+    },
+    {
+      "epoch": 0.019199222309982382,
+      "grad_norm": 0.2706839144229889,
+      "learning_rate": 9.993566102915432e-05,
+      "loss": 1.1446,
+      "step": 316
+    },
+    {
+      "epoch": 0.019259979342608907,
+      "grad_norm": 0.5909138917922974,
+      "learning_rate": 9.993517464098344e-05,
+      "loss": 1.2405,
+      "step": 317
+    },
+    {
+      "epoch": 0.019320736375235435,
+      "grad_norm": 0.2069237381219864,
+      "learning_rate": 9.993468642242905e-05,
+      "loss": 1.1611,
+      "step": 318
+    },
+    {
+      "epoch": 0.01938149340786196,
+      "grad_norm": 0.3580915331840515,
+      "learning_rate": 9.99341963735091e-05,
+      "loss": 1.1563,
+      "step": 319
+    },
+    {
+      "epoch": 0.019442250440488487,
+      "grad_norm": 0.33375483751296997,
+      "learning_rate": 9.993370449424153e-05,
+      "loss": 1.2603,
+      "step": 320
+    },
+    {
+      "epoch": 0.01950300747311501,
+      "grad_norm": 0.4273125231266022,
+      "learning_rate": 9.993321078464437e-05,
+      "loss": 1.4076,
+      "step": 321
+    },
+    {
+      "epoch": 0.01956376450574154,
+      "grad_norm": 0.38784006237983704,
+      "learning_rate": 9.99327152447357e-05,
+      "loss": 1.1459,
+      "step": 322
+    },
+    {
+      "epoch": 0.019624521538368068,
+      "grad_norm": 0.34369736909866333,
+      "learning_rate": 9.993221787453373e-05,
+      "loss": 1.2008,
+      "step": 323
+    },
+    {
+      "epoch": 0.019685278570994592,
+      "grad_norm": 0.3320353627204895,
+      "learning_rate": 9.993171867405664e-05,
+      "loss": 1.1748,
+      "step": 324
+    },
+    {
+      "epoch": 0.01974603560362112,
+      "grad_norm": 0.31289252638816833,
+      "learning_rate": 9.993121764332276e-05,
+      "loss": 1.1769,
+      "step": 325
+    },
+    {
+      "epoch": 0.019806792636247645,
+      "grad_norm": 0.44145724177360535,
+      "learning_rate": 9.993071478235044e-05,
+      "loss": 1.2588,
+      "step": 326
+    },
+    {
+      "epoch": 0.019867549668874173,
+      "grad_norm": 0.7167911529541016,
+      "learning_rate": 9.993021009115812e-05,
+      "loss": 1.1599,
+      "step": 327
+    },
+    {
+      "epoch": 0.019928306701500698,
+      "grad_norm": 0.4565293490886688,
+      "learning_rate": 9.99297035697643e-05,
+      "loss": 1.1647,
+      "step": 328
+    },
+    {
+      "epoch": 0.019989063734127226,
+      "grad_norm": 0.2835087776184082,
+      "learning_rate": 9.992919521818756e-05,
+      "loss": 1.2806,
+      "step": 329
+    },
+    {
+      "epoch": 0.02004982076675375,
+      "grad_norm": 1.6273994445800781,
+      "learning_rate": 9.99286850364465e-05,
+      "loss": 1.1873,
+      "step": 330
+    },
+    {
+      "epoch": 0.020110577799380278,
+      "grad_norm": 0.3168855905532837,
+      "learning_rate": 9.992817302455983e-05,
+      "loss": 1.1684,
+      "step": 331
+    },
+    {
+      "epoch": 0.020171334832006806,
+      "grad_norm": 0.5479095578193665,
+      "learning_rate": 9.992765918254634e-05,
+      "loss": 1.1279,
+      "step": 332
+    },
+    {
+      "epoch": 0.02023209186463333,
+      "grad_norm": 0.5789514780044556,
+      "learning_rate": 9.992714351042485e-05,
+      "loss": 1.4497,
+      "step": 333
+    },
+    {
+      "epoch": 0.02029284889725986,
+      "grad_norm": 0.2367178201675415,
+      "learning_rate": 9.992662600821427e-05,
+      "loss": 1.2091,
+      "step": 334
+    },
+    {
+      "epoch": 0.020353605929886383,
+      "grad_norm": 0.3046312630176544,
+      "learning_rate": 9.992610667593355e-05,
+      "loss": 1.2742,
+      "step": 335
+    },
+    {
+      "epoch": 0.02041436296251291,
+      "grad_norm": 0.32870152592658997,
+      "learning_rate": 9.992558551360174e-05,
+      "loss": 1.2338,
+      "step": 336
+    },
+    {
+      "epoch": 0.020475119995139436,
+      "grad_norm": 0.21014073491096497,
+      "learning_rate": 9.992506252123796e-05,
+      "loss": 1.1617,
+      "step": 337
+    },
+    {
+      "epoch": 0.020535877027765964,
+      "grad_norm": 0.2694036662578583,
+      "learning_rate": 9.992453769886135e-05,
+      "loss": 1.1296,
+      "step": 338
+    },
+    {
+      "epoch": 0.02059663406039249,
+      "grad_norm": 0.21450552344322205,
+      "learning_rate": 9.992401104649116e-05,
+      "loss": 1.2217,
+      "step": 339
+    },
+    {
+      "epoch": 0.020657391093019017,
+      "grad_norm": 0.25186142325401306,
+      "learning_rate": 9.99234825641467e-05,
+      "loss": 1.3127,
+      "step": 340
+    },
+    {
+      "epoch": 0.020718148125645545,
+      "grad_norm": 0.3109813928604126,
+      "learning_rate": 9.992295225184733e-05,
+      "loss": 1.2242,
+      "step": 341
+    },
+    {
+      "epoch": 0.02077890515827207,
+      "grad_norm": 0.21320192515850067,
+      "learning_rate": 9.99224201096125e-05,
+      "loss": 1.262,
+      "step": 342
+    },
+    {
+      "epoch": 0.020839662190898597,
+      "grad_norm": 0.3940616250038147,
+      "learning_rate": 9.99218861374617e-05,
+      "loss": 1.2604,
+      "step": 343
+    },
+    {
+      "epoch": 0.020900419223525122,
+      "grad_norm": 0.4618244767189026,
+      "learning_rate": 9.992135033541454e-05,
+      "loss": 1.2799,
+      "step": 344
+    },
+    {
+      "epoch": 0.02096117625615165,
+      "grad_norm": 0.22235806286334991,
+      "learning_rate": 9.992081270349061e-05,
+      "loss": 1.144,
+      "step": 345
+    },
+    {
+      "epoch": 0.021021933288778175,
+      "grad_norm": 0.27402716875076294,
+      "learning_rate": 9.992027324170965e-05,
+      "loss": 1.2279,
+      "step": 346
+    },
+    {
+      "epoch": 0.021082690321404703,
+      "grad_norm": 0.24074974656105042,
+      "learning_rate": 9.991973195009143e-05,
+      "loss": 1.1355,
+      "step": 347
+    },
+    {
+      "epoch": 0.02114344735403123,
+      "grad_norm": 0.21464583277702332,
+      "learning_rate": 9.991918882865576e-05,
+      "loss": 1.2164,
+      "step": 348
+    },
+    {
+      "epoch": 0.021204204386657755,
+      "grad_norm": 0.2532672882080078,
+      "learning_rate": 9.99186438774226e-05,
+      "loss": 1.1653,
+      "step": 349
+    },
+    {
+      "epoch": 0.021264961419284283,
+      "grad_norm": 0.19243644177913666,
+      "learning_rate": 9.991809709641189e-05,
+      "loss": 1.1504,
+      "step": 350
+    },
+    {
+      "epoch": 0.021325718451910808,
+      "grad_norm": 0.17330235242843628,
+      "learning_rate": 9.99175484856437e-05,
+      "loss": 1.1787,
+      "step": 351
+    },
+    {
+      "epoch": 0.021386475484537336,
+      "grad_norm": 0.44046974182128906,
+      "learning_rate": 9.99169980451381e-05,
+      "loss": 1.2375,
+      "step": 352
+    },
+    {
+      "epoch": 0.02144723251716386,
+      "grad_norm": 0.27522218227386475,
+      "learning_rate": 9.99164457749153e-05,
+      "loss": 1.261,
+      "step": 353
+    },
+    {
+      "epoch": 0.02150798954979039,
+      "grad_norm": 0.24426224827766418,
+      "learning_rate": 9.991589167499553e-05,
+      "loss": 1.2295,
+      "step": 354
+    },
+    {
+      "epoch": 0.021568746582416913,
+      "grad_norm": 0.1877678632736206,
+      "learning_rate": 9.991533574539909e-05,
+      "loss": 1.1229,
+      "step": 355
+    },
+    {
+      "epoch": 0.02162950361504344,
+      "grad_norm": 0.18022331595420837,
+      "learning_rate": 9.991477798614638e-05,
+      "loss": 1.1579,
+      "step": 356
+    },
+    {
+      "epoch": 0.02169026064766997,
+      "grad_norm": 0.2182506024837494,
+      "learning_rate": 9.991421839725783e-05,
+      "loss": 1.2418,
+      "step": 357
+    },
+    {
+      "epoch": 0.021751017680296494,
+      "grad_norm": 0.20350974798202515,
+      "learning_rate": 9.991365697875397e-05,
+      "loss": 1.1571,
+      "step": 358
+    },
+    {
+      "epoch": 0.021811774712923022,
+      "grad_norm": 0.17868244647979736,
+      "learning_rate": 9.991309373065533e-05,
+      "loss": 1.1354,
+      "step": 359
+    },
+    {
+      "epoch": 0.021872531745549546,
+      "grad_norm": 0.23201099038124084,
+      "learning_rate": 9.991252865298262e-05,
+      "loss": 1.2203,
+      "step": 360
+    },
+    {
+      "epoch": 0.021933288778176074,
+      "grad_norm": 0.9139298796653748,
+      "learning_rate": 9.991196174575653e-05,
+      "loss": 1.3402,
+      "step": 361
+    },
+    {
+      "epoch": 0.0219940458108026,
+      "grad_norm": 0.3012075424194336,
+      "learning_rate": 9.991139300899784e-05,
+      "loss": 1.4223,
+      "step": 362
+    },
+    {
+      "epoch": 0.022054802843429127,
+      "grad_norm": 0.43885961174964905,
+      "learning_rate": 9.991082244272736e-05,
+      "loss": 1.2401,
+      "step": 363
+    },
+    {
+      "epoch": 0.022115559876055655,
+      "grad_norm": 0.2859572768211365,
+      "learning_rate": 9.991025004696605e-05,
+      "loss": 1.1391,
+      "step": 364
+    },
+    {
+      "epoch": 0.02217631690868218,
+      "grad_norm": 0.3498673737049103,
+      "learning_rate": 9.990967582173489e-05,
+      "loss": 1.2427,
+      "step": 365
+    },
+    {
+      "epoch": 0.022237073941308708,
+      "grad_norm": 0.3147677183151245,
+      "learning_rate": 9.99090997670549e-05,
+      "loss": 1.1636,
+      "step": 366
+    },
+    {
+      "epoch": 0.022297830973935232,
+      "grad_norm": 0.19915129244327545,
+      "learning_rate": 9.990852188294721e-05,
+      "loss": 1.1531,
+      "step": 367
+    },
+    {
+      "epoch": 0.02235858800656176,
+      "grad_norm": 0.22790472209453583,
+      "learning_rate": 9.990794216943302e-05,
+      "loss": 1.144,
+      "step": 368
+    },
+    {
+      "epoch": 0.022419345039188285,
+      "grad_norm": 2.5829055309295654,
+      "learning_rate": 9.990736062653354e-05,
+      "loss": 1.1795,
+      "step": 369
+    },
+    {
+      "epoch": 0.022480102071814813,
+      "grad_norm": 0.263721764087677,
+      "learning_rate": 9.990677725427013e-05,
+      "loss": 1.2126,
+      "step": 370
+    },
+    {
+      "epoch": 0.022540859104441337,
+      "grad_norm": 0.8752437829971313,
+      "learning_rate": 9.990619205266414e-05,
+      "loss": 1.1374,
+      "step": 371
+    },
+    {
+      "epoch": 0.022601616137067865,
+      "grad_norm": 0.25581830739974976,
+      "learning_rate": 9.990560502173704e-05,
+      "loss": 1.2511,
+      "step": 372
+    },
+    {
+      "epoch": 0.022662373169694393,
+      "grad_norm": 0.2347150593996048,
+      "learning_rate": 9.990501616151034e-05,
+      "loss": 1.1987,
+      "step": 373
+    },
+    {
+      "epoch": 0.022723130202320918,
+      "grad_norm": 0.25042441487312317,
+      "learning_rate": 9.990442547200562e-05,
+      "loss": 1.1607,
+      "step": 374
+    },
+    {
+      "epoch": 0.022783887234947446,
+      "grad_norm": 0.2432336062192917,
+      "learning_rate": 9.990383295324454e-05,
+      "loss": 1.4697,
+      "step": 375
+    },
+    {
+      "epoch": 0.02284464426757397,
+      "grad_norm": 0.20277556777000427,
+      "learning_rate": 9.990323860524882e-05,
+      "loss": 1.1355,
+      "step": 376
+    },
+    {
+      "epoch": 0.0229054013002005,
+      "grad_norm": 0.2945151925086975,
+      "learning_rate": 9.990264242804026e-05,
+      "loss": 1.3467,
+      "step": 377
+    },
+    {
+      "epoch": 0.022966158332827023,
+      "grad_norm": 0.2620909810066223,
+      "learning_rate": 9.990204442164066e-05,
+      "loss": 1.2339,
+      "step": 378
+    },
+    {
+      "epoch": 0.02302691536545355,
+      "grad_norm": 0.3506776988506317,
+      "learning_rate": 9.9901444586072e-05,
+      "loss": 1.2258,
+      "step": 379
+    },
+    {
+      "epoch": 0.02308767239808008,
+      "grad_norm": 0.3582307994365692,
+      "learning_rate": 9.990084292135624e-05,
+      "loss": 1.1352,
+      "step": 380
+    },
+    {
+      "epoch": 0.023148429430706604,
+      "grad_norm": 14.608827590942383,
+      "learning_rate": 9.990023942751543e-05,
+      "loss": 1.3026,
+      "step": 381
+    },
+    {
+      "epoch": 0.023209186463333132,
+      "grad_norm": 0.32785990834236145,
+      "learning_rate": 9.989963410457171e-05,
+      "loss": 1.2796,
+      "step": 382
+    },
+    {
+      "epoch": 0.023269943495959657,
+      "grad_norm": 0.24141253530979156,
+      "learning_rate": 9.989902695254725e-05,
+      "loss": 1.2581,
+      "step": 383
+    },
+    {
+      "epoch": 0.023330700528586185,
+      "grad_norm": 0.32767367362976074,
+      "learning_rate": 9.989841797146431e-05,
+      "loss": 1.211,
+      "step": 384
+    },
+    {
+      "epoch": 0.02339145756121271,
+      "grad_norm": 0.28206735849380493,
+      "learning_rate": 9.989780716134522e-05,
+      "loss": 1.2436,
+      "step": 385
+    },
+    {
+      "epoch": 0.023452214593839237,
+      "grad_norm": 0.2510630190372467,
+      "learning_rate": 9.989719452221235e-05,
+      "loss": 1.235,
+      "step": 386
+    },
+    {
+      "epoch": 0.023512971626465762,
+      "grad_norm": 0.36102914810180664,
+      "learning_rate": 9.989658005408818e-05,
+      "loss": 1.2017,
+      "step": 387
+    },
+    {
+      "epoch": 0.02357372865909229,
+      "grad_norm": 0.3127259612083435,
+      "learning_rate": 9.989596375699522e-05,
+      "loss": 1.1618,
+      "step": 388
+    },
+    {
+      "epoch": 0.023634485691718818,
+      "grad_norm": 2.716749429702759,
+      "learning_rate": 9.989534563095607e-05,
+      "loss": 1.362,
+      "step": 389
+    },
+    {
+      "epoch": 0.023695242724345342,
+      "grad_norm": 0.2603667974472046,
+      "learning_rate": 9.989472567599338e-05,
+      "loss": 1.2619,
+      "step": 390
+    },
+    {
+      "epoch": 0.02375599975697187,
+      "grad_norm": 0.37386372685432434,
+      "learning_rate": 9.989410389212987e-05,
+      "loss": 1.2361,
+      "step": 391
+    },
+    {
+      "epoch": 0.023816756789598395,
+      "grad_norm": 0.3464028239250183,
+      "learning_rate": 9.989348027938834e-05,
+      "loss": 1.146,
+      "step": 392
+    },
+    {
+      "epoch": 0.023877513822224923,
+      "grad_norm": 0.40520069003105164,
+      "learning_rate": 9.989285483779164e-05,
+      "loss": 1.3619,
+      "step": 393
+    },
+    {
+      "epoch": 0.023938270854851448,
+      "grad_norm": 0.18270514905452728,
+      "learning_rate": 9.989222756736271e-05,
+      "loss": 1.1226,
+      "step": 394
+    },
+    {
+      "epoch": 0.023999027887477976,
+      "grad_norm": 0.278956800699234,
+      "learning_rate": 9.989159846812454e-05,
+      "loss": 1.2004,
+      "step": 395
+    },
+    {
+      "epoch": 0.024059784920104504,
+      "grad_norm": 0.38345643877983093,
+      "learning_rate": 9.989096754010018e-05,
+      "loss": 1.1555,
+      "step": 396
+    },
+    {
+      "epoch": 0.02412054195273103,
+      "grad_norm": 0.297656774520874,
+      "learning_rate": 9.989033478331276e-05,
+      "loss": 1.2252,
+      "step": 397
+    },
+    {
+      "epoch": 0.024181298985357556,
+      "grad_norm": 23.014734268188477,
+      "learning_rate": 9.988970019778547e-05,
+      "loss": 1.217,
+      "step": 398
+    },
+    {
+      "epoch": 0.02424205601798408,
+      "grad_norm": 0.3520590364933014,
+      "learning_rate": 9.988906378354158e-05,
+      "loss": 1.1651,
+      "step": 399
+    },
+    {
+      "epoch": 0.02430281305061061,
+      "grad_norm": 0.8311512470245361,
+      "learning_rate": 9.988842554060441e-05,
+      "loss": 1.1466,
+      "step": 400
+    },
+    {
+      "epoch": 0.024363570083237134,
+      "grad_norm": 0.7144541144371033,
+      "learning_rate": 9.988778546899738e-05,
+      "loss": 1.125,
+      "step": 401
+    },
+    {
+      "epoch": 0.02442432711586366,
+      "grad_norm": 0.40504345297813416,
+      "learning_rate": 9.98871435687439e-05,
+      "loss": 1.1774,
+      "step": 402
+    },
+    {
+      "epoch": 0.024485084148490186,
+      "grad_norm": 0.2774827778339386,
+      "learning_rate": 9.988649983986755e-05,
+      "loss": 1.1722,
+      "step": 403
+    },
+    {
+      "epoch": 0.024545841181116714,
+      "grad_norm": 0.3524899184703827,
+      "learning_rate": 9.98858542823919e-05,
+      "loss": 1.2614,
+      "step": 404
+    },
+    {
+      "epoch": 0.024606598213743242,
+      "grad_norm": 0.2749498188495636,
+      "learning_rate": 9.98852068963406e-05,
+      "loss": 1.1467,
+      "step": 405
+    },
+    {
+      "epoch": 0.024667355246369767,
+      "grad_norm": 0.2550564110279083,
+      "learning_rate": 9.988455768173741e-05,
+      "loss": 1.2173,
+      "step": 406
+    },
+    {
+      "epoch": 0.024728112278996295,
+      "grad_norm": 0.32169127464294434,
+      "learning_rate": 9.988390663860613e-05,
+      "loss": 1.256,
+      "step": 407
+    },
+    {
+      "epoch": 0.02478886931162282,
+      "grad_norm": 0.26087912917137146,
+      "learning_rate": 9.988325376697059e-05,
+      "loss": 1.1355,
+      "step": 408
+    },
+    {
+      "epoch": 0.024849626344249347,
+      "grad_norm": 0.3460897207260132,
+      "learning_rate": 9.988259906685475e-05,
+      "loss": 1.147,
+      "step": 409
+    },
+    {
+      "epoch": 0.024910383376875872,
+      "grad_norm": 0.23557282984256744,
+      "learning_rate": 9.98819425382826e-05,
+      "loss": 1.2472,
+      "step": 410
+    },
+    {
+      "epoch": 0.0249711404095024,
+      "grad_norm": 0.3029102683067322,
+      "learning_rate": 9.988128418127822e-05,
+      "loss": 1.1614,
+      "step": 411
+    },
+    {
+      "epoch": 0.025031897442128928,
+      "grad_norm": 0.401030957698822,
+      "learning_rate": 9.98806239958657e-05,
+      "loss": 1.2113,
+      "step": 412
+    },
+    {
+      "epoch": 0.025092654474755453,
+      "grad_norm": 0.21606552600860596,
+      "learning_rate": 9.987996198206928e-05,
+      "loss": 1.0921,
+      "step": 413
+    },
+    {
+      "epoch": 0.02515341150738198,
+      "grad_norm": 0.30670490860939026,
+      "learning_rate": 9.987929813991321e-05,
+      "loss": 1.2477,
+      "step": 414
+    },
+    {
+      "epoch": 0.025214168540008505,
+      "grad_norm": 0.2819117605686188,
+      "learning_rate": 9.987863246942181e-05,
+      "loss": 1.1571,
+      "step": 415
+    },
+    {
+      "epoch": 0.025274925572635033,
+      "grad_norm": 0.44100797176361084,
+      "learning_rate": 9.987796497061952e-05,
+      "loss": 1.2278,
+      "step": 416
+    },
+    {
+      "epoch": 0.025335682605261558,
+      "grad_norm": 0.3788920044898987,
+      "learning_rate": 9.987729564353078e-05,
+      "loss": 1.2841,
+      "step": 417
+    },
+    {
+      "epoch": 0.025396439637888086,
+      "grad_norm": 0.6525524258613586,
+      "learning_rate": 9.987662448818011e-05,
+      "loss": 1.1318,
+      "step": 418
+    },
+    {
+      "epoch": 0.02545719667051461,
+      "grad_norm": 0.5220158696174622,
+      "learning_rate": 9.987595150459215e-05,
+      "loss": 1.2277,
+      "step": 419
+    },
+    {
+      "epoch": 0.02551795370314114,
+      "grad_norm": 0.43092164397239685,
+      "learning_rate": 9.987527669279152e-05,
+      "loss": 1.16,
+      "step": 420
+    },
+    {
+      "epoch": 0.025578710735767667,
+      "grad_norm": 0.5233555436134338,
+      "learning_rate": 9.9874600052803e-05,
+      "loss": 1.2413,
+      "step": 421
+    },
+    {
+      "epoch": 0.02563946776839419,
+      "grad_norm": 0.20341508090496063,
+      "learning_rate": 9.987392158465138e-05,
+      "loss": 1.1414,
+      "step": 422
+    },
+    {
+      "epoch": 0.02570022480102072,
+      "grad_norm": 0.5091098546981812,
+      "learning_rate": 9.987324128836151e-05,
+      "loss": 1.1614,
+      "step": 423
+    },
+    {
+      "epoch": 0.025760981833647244,
+      "grad_norm": 0.2747238576412201,
+      "learning_rate": 9.987255916395835e-05,
+      "loss": 1.2224,
+      "step": 424
+    },
+    {
+      "epoch": 0.025821738866273772,
+      "grad_norm": 0.4567953646183014,
+      "learning_rate": 9.987187521146688e-05,
+      "loss": 1.224,
+      "step": 425
+    },
+    {
+      "epoch": 0.025882495898900296,
+      "grad_norm": 0.37994566559791565,
+      "learning_rate": 9.98711894309122e-05,
+      "loss": 1.1301,
+      "step": 426
+    },
+    {
+      "epoch": 0.025943252931526824,
+      "grad_norm": 0.22085264325141907,
+      "learning_rate": 9.987050182231943e-05,
+      "loss": 1.1586,
+      "step": 427
+    },
+    {
+      "epoch": 0.026004009964153352,
+      "grad_norm": 0.41410860419273376,
+      "learning_rate": 9.986981238571376e-05,
+      "loss": 1.2118,
+      "step": 428
+    },
+    {
+      "epoch": 0.026064766996779877,
+      "grad_norm": 2.435889959335327,
+      "learning_rate": 9.986912112112049e-05,
+      "loss": 1.1195,
+      "step": 429
+    },
+    {
+      "epoch": 0.026125524029406405,
+      "grad_norm": 0.35806381702423096,
+      "learning_rate": 9.986842802856494e-05,
+      "loss": 1.1783,
+      "step": 430
+    },
+    {
+      "epoch": 0.02618628106203293,
+      "grad_norm": 0.2825682461261749,
+      "learning_rate": 9.986773310807254e-05,
+      "loss": 1.2659,
+      "step": 431
+    },
+    {
+      "epoch": 0.026247038094659458,
+      "grad_norm": 0.17404058575630188,
+      "learning_rate": 9.986703635966872e-05,
+      "loss": 1.1721,
+      "step": 432
+    },
+    {
+      "epoch": 0.026307795127285982,
+      "grad_norm": 0.27772650122642517,
+      "learning_rate": 9.986633778337905e-05,
+      "loss": 1.2585,
+      "step": 433
+    },
+    {
+      "epoch": 0.02636855215991251,
+      "grad_norm": 0.17797604203224182,
+      "learning_rate": 9.986563737922914e-05,
+      "loss": 1.138,
+      "step": 434
+    },
+    {
+      "epoch": 0.026429309192539035,
+      "grad_norm": 0.319357305765152,
+      "learning_rate": 9.986493514724464e-05,
+      "loss": 1.2481,
+      "step": 435
+    },
+    {
+      "epoch": 0.026490066225165563,
+      "grad_norm": 0.24233309924602509,
+      "learning_rate": 9.98642310874513e-05,
+      "loss": 1.2086,
+      "step": 436
+    },
+    {
+      "epoch": 0.02655082325779209,
+      "grad_norm": 44.90375900268555,
+      "learning_rate": 9.986352519987494e-05,
+      "loss": 1.2702,
+      "step": 437
+    },
+    {
+      "epoch": 0.026611580290418616,
+      "grad_norm": 0.24607115983963013,
+      "learning_rate": 9.986281748454143e-05,
+      "loss": 1.1617,
+      "step": 438
+    },
+    {
+      "epoch": 0.026672337323045144,
+      "grad_norm": 0.21750274300575256,
+      "learning_rate": 9.986210794147671e-05,
+      "loss": 1.1653,
+      "step": 439
+    },
+    {
+      "epoch": 0.026733094355671668,
+      "grad_norm": 0.23555466532707214,
+      "learning_rate": 9.986139657070677e-05,
+      "loss": 1.1698,
+      "step": 440
+    },
+    {
+      "epoch": 0.026793851388298196,
+      "grad_norm": 0.1842479705810547,
+      "learning_rate": 9.98606833722577e-05,
+      "loss": 1.1573,
+      "step": 441
+    },
+    {
+      "epoch": 0.02685460842092472,
+      "grad_norm": 0.21280412375926971,
+      "learning_rate": 9.985996834615566e-05,
+      "loss": 1.1245,
+      "step": 442
+    },
+    {
+      "epoch": 0.02691536545355125,
+      "grad_norm": 0.24020196497440338,
+      "learning_rate": 9.985925149242683e-05,
+      "loss": 1.2301,
+      "step": 443
+    },
+    {
+      "epoch": 0.026976122486177777,
+      "grad_norm": 0.4025554656982422,
+      "learning_rate": 9.98585328110975e-05,
+      "loss": 1.3422,
+      "step": 444
+    },
+    {
+      "epoch": 0.0270368795188043,
+      "grad_norm": 0.20919737219810486,
+      "learning_rate": 9.985781230219402e-05,
+      "loss": 1.222,
+      "step": 445
+    },
+    {
+      "epoch": 0.02709763655143083,
+      "grad_norm": 0.2558465301990509,
+      "learning_rate": 9.985708996574278e-05,
+      "loss": 1.2962,
+      "step": 446
+    },
+    {
+      "epoch": 0.027158393584057354,
+      "grad_norm": 0.30489349365234375,
+      "learning_rate": 9.985636580177026e-05,
+      "loss": 1.4212,
+      "step": 447
+    },
+    {
+      "epoch": 0.027219150616683882,
+      "grad_norm": 0.17291104793548584,
+      "learning_rate": 9.985563981030304e-05,
+      "loss": 1.1398,
+      "step": 448
+    },
+    {
+      "epoch": 0.027279907649310407,
+      "grad_norm": 0.2797357141971588,
+      "learning_rate": 9.98549119913677e-05,
+      "loss": 1.194,
+      "step": 449
+    },
+    {
+      "epoch": 0.027340664681936935,
+      "grad_norm": 0.20370079576969147,
+      "learning_rate": 9.985418234499094e-05,
+      "loss": 1.156,
+      "step": 450
+    },
+    {
+      "epoch": 0.02740142171456346,
+      "grad_norm": 0.19331230223178864,
+      "learning_rate": 9.985345087119946e-05,
+      "loss": 1.1384,
+      "step": 451
+    },
+    {
+      "epoch": 0.027462178747189987,
+      "grad_norm": 0.3319239616394043,
+      "learning_rate": 9.98527175700201e-05,
+      "loss": 1.266,
+      "step": 452
+    },
+    {
+      "epoch": 0.027522935779816515,
+      "grad_norm": 0.30400246381759644,
+      "learning_rate": 9.985198244147978e-05,
+      "loss": 1.1891,
+      "step": 453
+    },
+    {
+      "epoch": 0.02758369281244304,
+      "grad_norm": 0.21529674530029297,
+      "learning_rate": 9.985124548560536e-05,
+      "loss": 1.1589,
+      "step": 454
+    },
+    {
+      "epoch": 0.027644449845069568,
+      "grad_norm": 0.605178952217102,
+      "learning_rate": 9.985050670242393e-05,
+      "loss": 1.2537,
+      "step": 455
+    },
+    {
+      "epoch": 0.027705206877696092,
+      "grad_norm": 0.27420905232429504,
+      "learning_rate": 9.984976609196253e-05,
+      "loss": 1.234,
+      "step": 456
+    },
+    {
+      "epoch": 0.02776596391032262,
+      "grad_norm": 0.19000773131847382,
+      "learning_rate": 9.984902365424833e-05,
+      "loss": 1.1405,
+      "step": 457
+    },
+    {
+      "epoch": 0.027826720942949145,
+      "grad_norm": 0.19945290684700012,
+      "learning_rate": 9.984827938930852e-05,
+      "loss": 1.2119,
+      "step": 458
+    },
+    {
+      "epoch": 0.027887477975575673,
+      "grad_norm": 0.22797727584838867,
+      "learning_rate": 9.98475332971704e-05,
+      "loss": 1.2138,
+      "step": 459
+    },
+    {
+      "epoch": 0.027948235008202198,
+      "grad_norm": 0.26123619079589844,
+      "learning_rate": 9.98467853778613e-05,
+      "loss": 1.1181,
+      "step": 460
+    },
+    {
+      "epoch": 0.028008992040828726,
+      "grad_norm": 0.19567619264125824,
+      "learning_rate": 9.984603563140866e-05,
+      "loss": 1.1587,
+      "step": 461
+    },
+    {
+      "epoch": 0.028069749073455254,
+      "grad_norm": 0.3084138333797455,
+      "learning_rate": 9.984528405783994e-05,
+      "loss": 1.2768,
+      "step": 462
+    },
+    {
+      "epoch": 0.02813050610608178,
+      "grad_norm": 0.24002857506275177,
+      "learning_rate": 9.984453065718271e-05,
+      "loss": 1.1468,
+      "step": 463
+    },
+    {
+      "epoch": 0.028191263138708306,
+      "grad_norm": 0.2294125258922577,
+      "learning_rate": 9.984377542946458e-05,
+      "loss": 1.2839,
+      "step": 464
+    },
+    {
+      "epoch": 0.02825202017133483,
+      "grad_norm": 0.26755377650260925,
+      "learning_rate": 9.98430183747132e-05,
+      "loss": 1.1732,
+      "step": 465
+    },
+    {
+      "epoch": 0.02831277720396136,
+      "grad_norm": 0.22083140909671783,
+      "learning_rate": 9.984225949295636e-05,
+      "loss": 1.1504,
+      "step": 466
+    },
+    {
+      "epoch": 0.028373534236587884,
+      "grad_norm": 0.15084177255630493,
+      "learning_rate": 9.984149878422186e-05,
+      "loss": 1.1279,
+      "step": 467
+    },
+    {
+      "epoch": 0.02843429126921441,
+      "grad_norm": 0.32440218329429626,
+      "learning_rate": 9.98407362485376e-05,
+      "loss": 1.1775,
+      "step": 468
+    },
+    {
+      "epoch": 0.02849504830184094,
+      "grad_norm": 0.22640341520309448,
+      "learning_rate": 9.98399718859315e-05,
+      "loss": 1.2064,
+      "step": 469
+    },
+    {
+      "epoch": 0.028555805334467464,
+      "grad_norm": 0.2065805196762085,
+      "learning_rate": 9.983920569643161e-05,
+      "loss": 1.1692,
+      "step": 470
+    },
+    {
+      "epoch": 0.028616562367093992,
+      "grad_norm": 0.5118774175643921,
+      "learning_rate": 9.9838437680066e-05,
+      "loss": 1.2046,
+      "step": 471
+    },
+    {
+      "epoch": 0.028677319399720517,
+      "grad_norm": 0.3265882432460785,
+      "learning_rate": 9.983766783686284e-05,
+      "loss": 1.2811,
+      "step": 472
+    },
+    {
+      "epoch": 0.028738076432347045,
+      "grad_norm": 0.377768337726593,
+      "learning_rate": 9.983689616685031e-05,
+      "loss": 1.3418,
+      "step": 473
+    },
+    {
+      "epoch": 0.02879883346497357,
+      "grad_norm": 0.3566953241825104,
+      "learning_rate": 9.983612267005671e-05,
+      "loss": 1.2245,
+      "step": 474
+    },
+    {
+      "epoch": 0.028859590497600097,
+      "grad_norm": 0.3818965554237366,
+      "learning_rate": 9.983534734651042e-05,
+      "loss": 1.1651,
+      "step": 475
+    },
+    {
+      "epoch": 0.028920347530226622,
+      "grad_norm": 0.3248099684715271,
+      "learning_rate": 9.983457019623983e-05,
+      "loss": 1.1557,
+      "step": 476
+    },
+    {
+      "epoch": 0.02898110456285315,
+      "grad_norm": 0.22260965406894684,
+      "learning_rate": 9.983379121927345e-05,
+      "loss": 1.1751,
+      "step": 477
+    },
+    {
+      "epoch": 0.029041861595479678,
+      "grad_norm": 0.29320624470710754,
+      "learning_rate": 9.98330104156398e-05,
+      "loss": 1.2332,
+      "step": 478
+    },
+    {
+      "epoch": 0.029102618628106203,
+      "grad_norm": 0.2943146526813507,
+      "learning_rate": 9.983222778536754e-05,
+      "loss": 1.1922,
+      "step": 479
+    },
+    {
+      "epoch": 0.02916337566073273,
+      "grad_norm": 0.2977176010608673,
+      "learning_rate": 9.983144332848533e-05,
+      "loss": 1.265,
+      "step": 480
+    },
+    {
+      "epoch": 0.029224132693359255,
+      "grad_norm": 2.0611252784729004,
+      "learning_rate": 9.983065704502193e-05,
+      "loss": 1.2488,
+      "step": 481
+    },
+    {
+      "epoch": 0.029284889725985783,
+      "grad_norm": 0.325751394033432,
+      "learning_rate": 9.982986893500617e-05,
+      "loss": 1.1734,
+      "step": 482
+    },
+    {
+      "epoch": 0.029345646758612308,
+      "grad_norm": 0.20634117722511292,
+      "learning_rate": 9.982907899846692e-05,
+      "loss": 1.1138,
+      "step": 483
+    },
+    {
+      "epoch": 0.029406403791238836,
+      "grad_norm": 0.27968963980674744,
+      "learning_rate": 9.982828723543316e-05,
+      "loss": 1.34,
+      "step": 484
+    },
+    {
+      "epoch": 0.029467160823865364,
+      "grad_norm": 0.23482371866703033,
+      "learning_rate": 9.98274936459339e-05,
+      "loss": 1.2572,
+      "step": 485
+    },
+    {
+      "epoch": 0.02952791785649189,
+      "grad_norm": 0.2467680126428604,
+      "learning_rate": 9.982669822999822e-05,
+      "loss": 1.144,
+      "step": 486
+    },
+    {
+      "epoch": 0.029588674889118417,
+      "grad_norm": 0.2581479549407959,
+      "learning_rate": 9.982590098765528e-05,
+      "loss": 1.3201,
+      "step": 487
+    },
+    {
+      "epoch": 0.02964943192174494,
+      "grad_norm": 0.528778612613678,
+      "learning_rate": 9.982510191893433e-05,
+      "loss": 1.201,
+      "step": 488
+    },
+    {
+      "epoch": 0.02971018895437147,
+      "grad_norm": 0.194417804479599,
+      "learning_rate": 9.982430102386462e-05,
+      "loss": 1.1689,
+      "step": 489
+    },
+    {
+      "epoch": 0.029770945986997994,
+      "grad_norm": 0.7083895802497864,
+      "learning_rate": 9.982349830247553e-05,
+      "loss": 1.2267,
+      "step": 490
+    },
+    {
+      "epoch": 0.029831703019624522,
+      "grad_norm": 0.2740737199783325,
+      "learning_rate": 9.982269375479648e-05,
+      "loss": 1.1455,
+      "step": 491
+    },
+    {
+      "epoch": 0.029892460052251046,
+      "grad_norm": 0.3193904757499695,
+      "learning_rate": 9.982188738085697e-05,
+      "loss": 1.274,
+      "step": 492
+    },
+    {
+      "epoch": 0.029953217084877574,
+      "grad_norm": 0.24167534708976746,
+      "learning_rate": 9.982107918068652e-05,
+      "loss": 1.2912,
+      "step": 493
+    },
+    {
+      "epoch": 0.030013974117504102,
+      "grad_norm": 0.3258008062839508,
+      "learning_rate": 9.98202691543148e-05,
+      "loss": 1.1588,
+      "step": 494
+    },
+    {
+      "epoch": 0.030074731150130627,
+      "grad_norm": 0.2044263482093811,
+      "learning_rate": 9.98194573017715e-05,
+      "loss": 1.1273,
+      "step": 495
+    },
+    {
+      "epoch": 0.030135488182757155,
+      "grad_norm": 0.17838792502880096,
+      "learning_rate": 9.981864362308634e-05,
+      "loss": 1.1111,
+      "step": 496
+    },
+    {
+      "epoch": 0.03019624521538368,
+      "grad_norm": 0.4175994396209717,
+      "learning_rate": 9.981782811828918e-05,
+      "loss": 1.1661,
+      "step": 497
+    },
+    {
+      "epoch": 0.030257002248010208,
+      "grad_norm": 0.4278128445148468,
+      "learning_rate": 9.981701078740988e-05,
+      "loss": 1.1671,
+      "step": 498
+    },
+    {
+      "epoch": 0.030317759280636732,
+      "grad_norm": 0.3493889272212982,
+      "learning_rate": 9.981619163047843e-05,
+      "loss": 1.2339,
+      "step": 499
+    },
+    {
+      "epoch": 0.03037851631326326,
+      "grad_norm": 0.37411755323410034,
+      "learning_rate": 9.981537064752486e-05,
+      "loss": 1.3125,
+      "step": 500
+    },
+    {
+      "epoch": 0.03043927334588979,
+      "grad_norm": 0.2385791838169098,
+      "learning_rate": 9.981454783857925e-05,
+      "loss": 1.2324,
+      "step": 501
+    },
+    {
+      "epoch": 0.030500030378516313,
+      "grad_norm": 0.20261655747890472,
+      "learning_rate": 9.981372320367175e-05,
+      "loss": 1.1406,
+      "step": 502
+    },
+    {
+      "epoch": 0.03056078741114284,
+      "grad_norm": 0.2751544713973999,
+      "learning_rate": 9.98128967428326e-05,
+      "loss": 1.1737,
+      "step": 503
+    },
+    {
+      "epoch": 0.030621544443769366,
+      "grad_norm": 0.4848704934120178,
+      "learning_rate": 9.981206845609209e-05,
+      "loss": 1.3857,
+      "step": 504
+    },
+    {
+      "epoch": 0.030682301476395894,
+      "grad_norm": 0.48026302456855774,
+      "learning_rate": 9.981123834348058e-05,
+      "loss": 1.2668,
+      "step": 505
+    },
+    {
+      "epoch": 0.030743058509022418,
+      "grad_norm": 0.33266085386276245,
+      "learning_rate": 9.98104064050285e-05,
+      "loss": 1.1248,
+      "step": 506
+    },
+    {
+      "epoch": 0.030803815541648946,
+      "grad_norm": 0.25252965092658997,
+      "learning_rate": 9.980957264076635e-05,
+      "loss": 1.2272,
+      "step": 507
+    },
+    {
+      "epoch": 0.03086457257427547,
+      "grad_norm": 0.3843928873538971,
+      "learning_rate": 9.980873705072468e-05,
+      "loss": 1.2356,
+      "step": 508
+    },
+    {
+      "epoch": 0.030925329606902,
+      "grad_norm": 0.33188146352767944,
+      "learning_rate": 9.980789963493415e-05,
+      "loss": 1.2464,
+      "step": 509
+    },
+    {
+      "epoch": 0.030986086639528527,
+      "grad_norm": 0.20660185813903809,
+      "learning_rate": 9.98070603934254e-05,
+      "loss": 1.1742,
+      "step": 510
+    },
+    {
+      "epoch": 0.03104684367215505,
+      "grad_norm": 0.23412880301475525,
+      "learning_rate": 9.980621932622924e-05,
+      "loss": 1.1094,
+      "step": 511
+    },
+    {
+      "epoch": 0.03110760070478158,
+      "grad_norm": 0.27089810371398926,
+      "learning_rate": 9.980537643337648e-05,
+      "loss": 1.1472,
+      "step": 512
+    },
+    {
+      "epoch": 0.031168357737408104,
+      "grad_norm": 0.6027984023094177,
+      "learning_rate": 9.980453171489802e-05,
+      "loss": 1.1834,
+      "step": 513
+    },
+    {
+      "epoch": 0.031229114770034632,
+      "grad_norm": 0.22745102643966675,
+      "learning_rate": 9.980368517082481e-05,
+      "loss": 1.1557,
+      "step": 514
+    },
+    {
+      "epoch": 0.03128987180266116,
+      "grad_norm": 0.31312257051467896,
+      "learning_rate": 9.980283680118792e-05,
+      "loss": 1.388,
+      "step": 515
+    },
+    {
+      "epoch": 0.031350628835287685,
+      "grad_norm": 2.9024300575256348,
+      "learning_rate": 9.98019866060184e-05,
+      "loss": 1.2529,
+      "step": 516
+    },
+    {
+      "epoch": 0.03141138586791421,
+      "grad_norm": 0.2904795706272125,
+      "learning_rate": 9.980113458534744e-05,
+      "loss": 1.1286,
+      "step": 517
+    },
+    {
+      "epoch": 0.03147214290054074,
+      "grad_norm": 18.508773803710938,
+      "learning_rate": 9.980028073920627e-05,
+      "loss": 1.1558,
+      "step": 518
+    },
+    {
+      "epoch": 0.03153289993316726,
+      "grad_norm": 0.5735429525375366,
+      "learning_rate": 9.979942506762617e-05,
+      "loss": 1.3104,
+      "step": 519
+    },
+    {
+      "epoch": 0.03159365696579379,
+      "grad_norm": 0.885894775390625,
+      "learning_rate": 9.979856757063853e-05,
+      "loss": 1.1641,
+      "step": 520
+    },
+    {
+      "epoch": 0.03165441399842032,
+      "grad_norm": 0.30875757336616516,
+      "learning_rate": 9.979770824827477e-05,
+      "loss": 1.1308,
+      "step": 521
+    },
+    {
+      "epoch": 0.031715171031046846,
+      "grad_norm": 0.39021605253219604,
+      "learning_rate": 9.979684710056639e-05,
+      "loss": 1.1799,
+      "step": 522
+    },
+    {
+      "epoch": 0.03177592806367337,
+      "grad_norm": 0.23148266971111298,
+      "learning_rate": 9.979598412754493e-05,
+      "loss": 1.1633,
+      "step": 523
+    },
+    {
+      "epoch": 0.031836685096299895,
+      "grad_norm": 0.2037530392408371,
+      "learning_rate": 9.979511932924209e-05,
+      "loss": 1.2202,
+      "step": 524
+    },
+    {
+      "epoch": 0.03189744212892642,
+      "grad_norm": 0.23318496346473694,
+      "learning_rate": 9.979425270568949e-05,
+      "loss": 1.1552,
+      "step": 525
+    },
+    {
+      "epoch": 0.03195819916155295,
+      "grad_norm": 0.22681386768817902,
+      "learning_rate": 9.979338425691894e-05,
+      "loss": 1.2651,
+      "step": 526
+    },
+    {
+      "epoch": 0.03201895619417948,
+      "grad_norm": 0.2016787976026535,
+      "learning_rate": 9.979251398296228e-05,
+      "loss": 1.2703,
+      "step": 527
+    },
+    {
+      "epoch": 0.032079713226806,
+      "grad_norm": 0.3406883478164673,
+      "learning_rate": 9.979164188385138e-05,
+      "loss": 1.1342,
+      "step": 528
+    },
+    {
+      "epoch": 0.03214047025943253,
+      "grad_norm": 0.19631391763687134,
+      "learning_rate": 9.979076795961821e-05,
+      "loss": 1.1635,
+      "step": 529
+    },
+    {
+      "epoch": 0.032201227292059056,
+      "grad_norm": 0.6007776856422424,
+      "learning_rate": 9.978989221029484e-05,
+      "loss": 1.1395,
+      "step": 530
+    },
+    {
+      "epoch": 0.032261984324685584,
+      "grad_norm": 0.17776396870613098,
+      "learning_rate": 9.978901463591335e-05,
+      "loss": 1.1325,
+      "step": 531
+    },
+    {
+      "epoch": 0.032322741357312106,
+      "grad_norm": 0.21798866987228394,
+      "learning_rate": 9.978813523650589e-05,
+      "loss": 1.1477,
+      "step": 532
+    },
+    {
+      "epoch": 0.032383498389938634,
+      "grad_norm": 0.28931427001953125,
+      "learning_rate": 9.978725401210471e-05,
+      "loss": 1.171,
+      "step": 533
+    },
+    {
+      "epoch": 0.03244425542256516,
+      "grad_norm": 0.28167280554771423,
+      "learning_rate": 9.97863709627421e-05,
+      "loss": 1.2423,
+      "step": 534
+    },
+    {
+      "epoch": 0.03250501245519169,
+      "grad_norm": 0.27362319827079773,
+      "learning_rate": 9.978548608845046e-05,
+      "loss": 1.1502,
+      "step": 535
+    },
+    {
+      "epoch": 0.03256576948781822,
+      "grad_norm": 0.2425500899553299,
+      "learning_rate": 9.97845993892622e-05,
+      "loss": 1.1928,
+      "step": 536
+    },
+    {
+      "epoch": 0.03262652652044474,
+      "grad_norm": 0.3945496380329132,
+      "learning_rate": 9.978371086520983e-05,
+      "loss": 1.1624,
+      "step": 537
+    },
+    {
+      "epoch": 0.03268728355307127,
+      "grad_norm": 0.3749012351036072,
+      "learning_rate": 9.97828205163259e-05,
+      "loss": 1.1653,
+      "step": 538
+    },
+    {
+      "epoch": 0.032748040585697795,
+      "grad_norm": 0.23593762516975403,
+      "learning_rate": 9.978192834264306e-05,
+      "loss": 1.1626,
+      "step": 539
+    },
+    {
+      "epoch": 0.03280879761832432,
+      "grad_norm": 0.3769228756427765,
+      "learning_rate": 9.978103434419403e-05,
+      "loss": 1.1896,
+      "step": 540
+    },
+    {
+      "epoch": 0.03286955465095085,
+      "grad_norm": 0.1705779880285263,
+      "learning_rate": 9.978013852101155e-05,
+      "loss": 1.1551,
+      "step": 541
+    },
+    {
+      "epoch": 0.03293031168357737,
+      "grad_norm": 0.20579193532466888,
+      "learning_rate": 9.977924087312849e-05,
+      "loss": 1.1785,
+      "step": 542
+    },
+    {
+      "epoch": 0.0329910687162039,
+      "grad_norm": 0.22284187376499176,
+      "learning_rate": 9.977834140057771e-05,
+      "loss": 1.2733,
+      "step": 543
+    },
+    {
+      "epoch": 0.03305182574883043,
+      "grad_norm": 0.23286156356334686,
+      "learning_rate": 9.977744010339222e-05,
+      "loss": 1.2651,
+      "step": 544
+    },
+    {
+      "epoch": 0.033112582781456956,
+      "grad_norm": 0.20310372114181519,
+      "learning_rate": 9.977653698160503e-05,
+      "loss": 1.1101,
+      "step": 545
+    },
+    {
+      "epoch": 0.03317333981408348,
+      "grad_norm": 0.3925718367099762,
+      "learning_rate": 9.977563203524926e-05,
+      "loss": 1.3441,
+      "step": 546
+    },
+    {
+      "epoch": 0.033234096846710005,
+      "grad_norm": 0.36995238065719604,
+      "learning_rate": 9.977472526435809e-05,
+      "loss": 1.222,
+      "step": 547
+    },
+    {
+      "epoch": 0.03329485387933653,
+      "grad_norm": 0.269410640001297,
+      "learning_rate": 9.977381666896472e-05,
+      "loss": 1.1506,
+      "step": 548
+    },
+    {
+      "epoch": 0.03335561091196306,
+      "grad_norm": 0.3384246230125427,
+      "learning_rate": 9.977290624910249e-05,
+      "loss": 1.1627,
+      "step": 549
+    },
+    {
+      "epoch": 0.03341636794458959,
+      "grad_norm": 1.7741787433624268,
+      "learning_rate": 9.977199400480477e-05,
+      "loss": 1.1497,
+      "step": 550
+    },
+    {
+      "epoch": 0.03347712497721611,
+      "grad_norm": 0.5351099371910095,
+      "learning_rate": 9.977107993610497e-05,
+      "loss": 1.1993,
+      "step": 551
+    },
+    {
+      "epoch": 0.03353788200984264,
+      "grad_norm": 0.4378606379032135,
+      "learning_rate": 9.977016404303663e-05,
+      "loss": 1.232,
+      "step": 552
+    },
+    {
+      "epoch": 0.03359863904246917,
+      "grad_norm": 0.4066349267959595,
+      "learning_rate": 9.97692463256333e-05,
+      "loss": 1.1599,
+      "step": 553
+    },
+    {
+      "epoch": 0.033659396075095695,
+      "grad_norm": 1.2139794826507568,
+      "learning_rate": 9.976832678392862e-05,
+      "loss": 1.3566,
+      "step": 554
+    },
+    {
+      "epoch": 0.033720153107722216,
+      "grad_norm": 0.38028234243392944,
+      "learning_rate": 9.976740541795632e-05,
+      "loss": 1.2715,
+      "step": 555
+    },
+    {
+      "epoch": 0.033780910140348744,
+      "grad_norm": 0.3841686248779297,
+      "learning_rate": 9.976648222775013e-05,
+      "loss": 1.1234,
+      "step": 556
+    },
+    {
+      "epoch": 0.03384166717297527,
+      "grad_norm": 0.28865474462509155,
+      "learning_rate": 9.976555721334392e-05,
+      "loss": 1.1951,
+      "step": 557
+    },
+    {
+      "epoch": 0.0339024242056018,
+      "grad_norm": 0.46802064776420593,
+      "learning_rate": 9.976463037477162e-05,
+      "loss": 1.1814,
+      "step": 558
+    },
+    {
+      "epoch": 0.03396318123822833,
+      "grad_norm": 0.3400562107563019,
+      "learning_rate": 9.976370171206714e-05,
+      "loss": 1.2694,
+      "step": 559
+    },
+    {
+      "epoch": 0.03402393827085485,
+      "grad_norm": 0.24644635617733002,
+      "learning_rate": 9.976277122526457e-05,
+      "loss": 1.2131,
+      "step": 560
+    },
+    {
+      "epoch": 0.03408469530348138,
+      "grad_norm": 0.35238033533096313,
+      "learning_rate": 9.976183891439801e-05,
+      "loss": 1.2933,
+      "step": 561
+    },
+    {
+      "epoch": 0.034145452336107905,
+      "grad_norm": 0.5878807902336121,
+      "learning_rate": 9.976090477950161e-05,
+      "loss": 1.1863,
+      "step": 562
+    },
+    {
+      "epoch": 0.03420620936873443,
+      "grad_norm": 0.23976045846939087,
+      "learning_rate": 9.975996882060963e-05,
+      "loss": 1.1726,
+      "step": 563
+    },
+    {
+      "epoch": 0.034266966401360954,
+      "grad_norm": 0.45904335379600525,
+      "learning_rate": 9.975903103775638e-05,
+      "loss": 1.256,
+      "step": 564
+    },
+    {
+      "epoch": 0.03432772343398748,
+      "grad_norm": 0.22525691986083984,
+      "learning_rate": 9.975809143097622e-05,
+      "loss": 1.1364,
+      "step": 565
+    },
+    {
+      "epoch": 0.03438848046661401,
+      "grad_norm": 0.31694698333740234,
+      "learning_rate": 9.975715000030361e-05,
+      "loss": 1.2514,
+      "step": 566
+    },
+    {
+      "epoch": 0.03444923749924054,
+      "grad_norm": 0.46745648980140686,
+      "learning_rate": 9.975620674577303e-05,
+      "loss": 1.1186,
+      "step": 567
+    },
+    {
+      "epoch": 0.034509994531867066,
+      "grad_norm": 0.5622209310531616,
+      "learning_rate": 9.97552616674191e-05,
+      "loss": 1.2989,
+      "step": 568
+    },
+    {
+      "epoch": 0.03457075156449359,
+      "grad_norm": 0.2631570100784302,
+      "learning_rate": 9.975431476527642e-05,
+      "loss": 1.1795,
+      "step": 569
+    },
+    {
+      "epoch": 0.034631508597120116,
+      "grad_norm": 0.706419825553894,
+      "learning_rate": 9.975336603937972e-05,
+      "loss": 1.1334,
+      "step": 570
+    },
+    {
+      "epoch": 0.034692265629746644,
+      "grad_norm": 0.234999418258667,
+      "learning_rate": 9.975241548976377e-05,
+      "loss": 1.1608,
+      "step": 571
+    },
+    {
+      "epoch": 0.03475302266237317,
+      "grad_norm": 0.3765067458152771,
+      "learning_rate": 9.975146311646342e-05,
+      "loss": 1.2703,
+      "step": 572
+    },
+    {
+      "epoch": 0.03481377969499969,
+      "grad_norm": 0.3500179946422577,
+      "learning_rate": 9.975050891951356e-05,
+      "loss": 1.1921,
+      "step": 573
+    },
+    {
+      "epoch": 0.03487453672762622,
+      "grad_norm": 0.3293748199939728,
+      "learning_rate": 9.97495528989492e-05,
+      "loss": 1.1633,
+      "step": 574
+    },
+    {
+      "epoch": 0.03493529376025275,
+      "grad_norm": 0.24921773374080658,
+      "learning_rate": 9.974859505480534e-05,
+      "loss": 1.1629,
+      "step": 575
+    },
+    {
+      "epoch": 0.03499605079287928,
+      "grad_norm": 0.553653359413147,
+      "learning_rate": 9.974763538711713e-05,
+      "loss": 1.1635,
+      "step": 576
+    },
+    {
+      "epoch": 0.035056807825505805,
+      "grad_norm": 0.25096818804740906,
+      "learning_rate": 9.97466738959197e-05,
+      "loss": 1.1521,
+      "step": 577
+    },
+    {
+      "epoch": 0.035117564858132326,
+      "grad_norm": 0.7093678712844849,
+      "learning_rate": 9.974571058124836e-05,
+      "loss": 1.1843,
+      "step": 578
+    },
+    {
+      "epoch": 0.035178321890758854,
+      "grad_norm": 0.4011084735393524,
+      "learning_rate": 9.974474544313835e-05,
+      "loss": 1.2213,
+      "step": 579
+    },
+    {
+      "epoch": 0.03523907892338538,
+      "grad_norm": 0.2578686773777008,
+      "learning_rate": 9.974377848162511e-05,
+      "loss": 1.2518,
+      "step": 580
+    },
+    {
+      "epoch": 0.03529983595601191,
+      "grad_norm": 0.5746710896492004,
+      "learning_rate": 9.974280969674403e-05,
+      "loss": 1.1599,
+      "step": 581
+    },
+    {
+      "epoch": 0.03536059298863844,
+      "grad_norm": 0.46072524785995483,
+      "learning_rate": 9.974183908853065e-05,
+      "loss": 1.1874,
+      "step": 582
+    },
+    {
+      "epoch": 0.03542135002126496,
+      "grad_norm": 0.3364895284175873,
+      "learning_rate": 9.974086665702055e-05,
+      "loss": 1.2154,
+      "step": 583
+    },
+    {
+      "epoch": 0.03548210705389149,
+      "grad_norm": 0.6634140610694885,
+      "learning_rate": 9.973989240224936e-05,
+      "loss": 1.2645,
+      "step": 584
+    },
+    {
+      "epoch": 0.035542864086518015,
+      "grad_norm": 0.18739502131938934,
+      "learning_rate": 9.97389163242528e-05,
+      "loss": 1.1574,
+      "step": 585
+    },
+    {
+      "epoch": 0.03560362111914454,
+      "grad_norm": 0.3704637587070465,
+      "learning_rate": 9.973793842306667e-05,
+      "loss": 1.1632,
+      "step": 586
+    },
+    {
+      "epoch": 0.035664378151771065,
+      "grad_norm": 0.406078964471817,
+      "learning_rate": 9.973695869872676e-05,
+      "loss": 1.1416,
+      "step": 587
+    },
+    {
+      "epoch": 0.03572513518439759,
+      "grad_norm": 0.22447341680526733,
+      "learning_rate": 9.973597715126904e-05,
+      "loss": 1.1105,
+      "step": 588
+    },
+    {
+      "epoch": 0.03578589221702412,
+      "grad_norm": 0.20255105197429657,
+      "learning_rate": 9.973499378072945e-05,
+      "loss": 1.1235,
+      "step": 589
+    },
+    {
+      "epoch": 0.03584664924965065,
+      "grad_norm": 0.35011032223701477,
+      "learning_rate": 9.973400858714406e-05,
+      "loss": 1.1137,
+      "step": 590
+    },
+    {
+      "epoch": 0.03590740628227718,
+      "grad_norm": 0.33853498101234436,
+      "learning_rate": 9.973302157054897e-05,
+      "loss": 1.3758,
+      "step": 591
+    },
+    {
+      "epoch": 0.0359681633149037,
+      "grad_norm": 0.2971729040145874,
+      "learning_rate": 9.973203273098035e-05,
+      "loss": 1.1714,
+      "step": 592
+    },
+    {
+      "epoch": 0.036028920347530226,
+      "grad_norm": 0.35486486554145813,
+      "learning_rate": 9.973104206847446e-05,
+      "loss": 1.2644,
+      "step": 593
+    },
+    {
+      "epoch": 0.036089677380156754,
+      "grad_norm": 0.382726788520813,
+      "learning_rate": 9.973004958306761e-05,
+      "loss": 1.1486,
+      "step": 594
+    },
+    {
+      "epoch": 0.03615043441278328,
+      "grad_norm": 0.28380724787712097,
+      "learning_rate": 9.972905527479621e-05,
+      "loss": 1.1434,
+      "step": 595
+    },
+    {
+      "epoch": 0.0362111914454098,
+      "grad_norm": 0.27264660596847534,
+      "learning_rate": 9.972805914369665e-05,
+      "loss": 1.1639,
+      "step": 596
+    },
+    {
+      "epoch": 0.03627194847803633,
+      "grad_norm": 0.30375102162361145,
+      "learning_rate": 9.972706118980546e-05,
+      "loss": 1.1345,
+      "step": 597
+    },
+    {
+      "epoch": 0.03633270551066286,
+      "grad_norm": 0.20119057595729828,
+      "learning_rate": 9.972606141315924e-05,
+      "loss": 1.1644,
+      "step": 598
+    },
+    {
+      "epoch": 0.03639346254328939,
+      "grad_norm": 1.647966980934143,
+      "learning_rate": 9.972505981379464e-05,
+      "loss": 1.1781,
+      "step": 599
+    },
+    {
+      "epoch": 0.036454219575915915,
+      "grad_norm": 0.41259098052978516,
+      "learning_rate": 9.972405639174834e-05,
+      "loss": 1.2058,
+      "step": 600
+    },
+    {
+      "epoch": 0.036514976608542436,
+      "grad_norm": 10.903308868408203,
+      "learning_rate": 9.972305114705715e-05,
+      "loss": 1.2472,
+      "step": 601
+    },
+    {
+      "epoch": 0.036575733641168964,
+      "grad_norm": 0.8497883677482605,
+      "learning_rate": 9.972204407975792e-05,
+      "loss": 1.3055,
+      "step": 602
+    },
+    {
+      "epoch": 0.03663649067379549,
+      "grad_norm": 0.2843227684497833,
+      "learning_rate": 9.972103518988753e-05,
+      "loss": 1.3834,
+      "step": 603
+    },
+    {
+      "epoch": 0.03669724770642202,
+      "grad_norm": 0.24960339069366455,
+      "learning_rate": 9.9720024477483e-05,
+      "loss": 1.1357,
+      "step": 604
+    },
+    {
+      "epoch": 0.03675800473904854,
+      "grad_norm": 0.3230237662792206,
+      "learning_rate": 9.971901194258135e-05,
+      "loss": 1.2343,
+      "step": 605
+    },
+    {
+      "epoch": 0.03681876177167507,
+      "grad_norm": 0.4179176092147827,
+      "learning_rate": 9.971799758521972e-05,
+      "loss": 1.1919,
+      "step": 606
+    },
+    {
+      "epoch": 0.0368795188043016,
+      "grad_norm": 0.1874077171087265,
+      "learning_rate": 9.971698140543528e-05,
+      "loss": 1.1471,
+      "step": 607
+    },
+    {
+      "epoch": 0.036940275836928126,
+      "grad_norm": 0.28784072399139404,
+      "learning_rate": 9.971596340326526e-05,
+      "loss": 1.2477,
+      "step": 608
+    },
+    {
+      "epoch": 0.037001032869554654,
+      "grad_norm": 0.44791775941848755,
+      "learning_rate": 9.971494357874701e-05,
+      "loss": 1.2427,
+      "step": 609
+    },
+    {
+      "epoch": 0.037061789902181175,
+      "grad_norm": 0.19704169034957886,
+      "learning_rate": 9.971392193191789e-05,
+      "loss": 1.1317,
+      "step": 610
+    },
+    {
+      "epoch": 0.0371225469348077,
+      "grad_norm": 0.760901689529419,
+      "learning_rate": 9.971289846281535e-05,
+      "loss": 1.2597,
+      "step": 611
+    },
+    {
+      "epoch": 0.03718330396743423,
+      "grad_norm": 0.24059328436851501,
+      "learning_rate": 9.97118731714769e-05,
+      "loss": 1.3124,
+      "step": 612
+    },
+    {
+      "epoch": 0.03724406100006076,
+      "grad_norm": 0.46775710582733154,
+      "learning_rate": 9.971084605794013e-05,
+      "loss": 1.1538,
+      "step": 613
+    },
+    {
+      "epoch": 0.03730481803268729,
+      "grad_norm": 0.2856523096561432,
+      "learning_rate": 9.97098171222427e-05,
+      "loss": 1.2097,
+      "step": 614
+    },
+    {
+      "epoch": 0.03736557506531381,
+      "grad_norm": 0.21430863440036774,
+      "learning_rate": 9.970878636442231e-05,
+      "loss": 1.1704,
+      "step": 615
+    },
+    {
+      "epoch": 0.037426332097940336,
+      "grad_norm": 0.2931556701660156,
+      "learning_rate": 9.970775378451674e-05,
+      "loss": 1.1479,
+      "step": 616
+    },
+    {
+      "epoch": 0.037487089130566864,
+      "grad_norm": 0.45134103298187256,
+      "learning_rate": 9.970671938256386e-05,
+      "loss": 1.4385,
+      "step": 617
+    },
+    {
+      "epoch": 0.03754784616319339,
+      "grad_norm": 0.28096574544906616,
+      "learning_rate": 9.970568315860156e-05,
+      "loss": 1.189,
+      "step": 618
+    },
+    {
+      "epoch": 0.03760860319581991,
+      "grad_norm": 0.5512173175811768,
+      "learning_rate": 9.970464511266784e-05,
+      "loss": 1.2531,
+      "step": 619
+    },
+    {
+      "epoch": 0.03766936022844644,
+      "grad_norm": 0.16908709704875946,
+      "learning_rate": 9.970360524480076e-05,
+      "loss": 1.1393,
+      "step": 620
+    },
+    {
+      "epoch": 0.03773011726107297,
+      "grad_norm": 0.4691760540008545,
+      "learning_rate": 9.97025635550384e-05,
+      "loss": 1.2468,
+      "step": 621
+    },
+    {
+      "epoch": 0.0377908742936995,
+      "grad_norm": 0.4211539924144745,
+      "learning_rate": 9.970152004341897e-05,
+      "loss": 1.1406,
+      "step": 622
+    },
+    {
+      "epoch": 0.037851631326326025,
+      "grad_norm": 0.2903730869293213,
+      "learning_rate": 9.970047470998071e-05,
+      "loss": 1.2194,
+      "step": 623
+    },
+    {
+      "epoch": 0.037912388358952546,
+      "grad_norm": 0.7190700173377991,
+      "learning_rate": 9.969942755476198e-05,
+      "loss": 1.2489,
+      "step": 624
+    },
+    {
+      "epoch": 0.037973145391579075,
+      "grad_norm": 1.3949562311172485,
+      "learning_rate": 9.969837857780109e-05,
+      "loss": 1.4563,
+      "step": 625
+    },
+    {
+      "epoch": 0.0380339024242056,
+      "grad_norm": 0.250702828168869,
+      "learning_rate": 9.969732777913653e-05,
+      "loss": 1.1457,
+      "step": 626
+    },
+    {
+      "epoch": 0.03809465945683213,
+      "grad_norm": 0.24044571816921234,
+      "learning_rate": 9.969627515880682e-05,
+      "loss": 1.1348,
+      "step": 627
+    },
+    {
+      "epoch": 0.03815541648945865,
+      "grad_norm": 0.23596933484077454,
+      "learning_rate": 9.969522071685054e-05,
+      "loss": 1.1084,
+      "step": 628
+    },
+    {
+      "epoch": 0.03821617352208518,
+      "grad_norm": 0.39057591557502747,
+      "learning_rate": 9.969416445330634e-05,
+      "loss": 1.1658,
+      "step": 629
+    },
+    {
+      "epoch": 0.03827693055471171,
+      "grad_norm": 0.19264733791351318,
+      "learning_rate": 9.969310636821294e-05,
+      "loss": 1.2475,
+      "step": 630
+    },
+    {
+      "epoch": 0.038337687587338236,
+      "grad_norm": 0.23034612834453583,
+      "learning_rate": 9.96920464616091e-05,
+      "loss": 1.1769,
+      "step": 631
+    },
+    {
+      "epoch": 0.038398444619964764,
+      "grad_norm": 0.21617116034030914,
+      "learning_rate": 9.96909847335337e-05,
+      "loss": 1.1668,
+      "step": 632
+    },
+    {
+      "epoch": 0.038459201652591285,
+      "grad_norm": 0.2674920856952667,
+      "learning_rate": 9.968992118402567e-05,
+      "loss": 1.2422,
+      "step": 633
+    },
+    {
+      "epoch": 0.03851995868521781,
+      "grad_norm": 0.30965131521224976,
+      "learning_rate": 9.968885581312397e-05,
+      "loss": 1.1773,
+      "step": 634
+    },
+    {
+      "epoch": 0.03858071571784434,
+      "grad_norm": 0.20929965376853943,
+      "learning_rate": 9.968778862086763e-05,
+      "loss": 1.2276,
+      "step": 635
+    },
+    {
+      "epoch": 0.03864147275047087,
+      "grad_norm": 0.28453242778778076,
+      "learning_rate": 9.968671960729581e-05,
+      "loss": 1.1192,
+      "step": 636
+    },
+    {
+      "epoch": 0.03870222978309739,
+      "grad_norm": 0.29931753873825073,
+      "learning_rate": 9.968564877244768e-05,
+      "loss": 1.2148,
+      "step": 637
+    },
+    {
+      "epoch": 0.03876298681572392,
+      "grad_norm": 0.4244219660758972,
+      "learning_rate": 9.968457611636248e-05,
+      "loss": 1.2368,
+      "step": 638
+    },
+    {
+      "epoch": 0.038823743848350446,
+      "grad_norm": 0.27939629554748535,
+      "learning_rate": 9.968350163907956e-05,
+      "loss": 1.0929,
+      "step": 639
+    },
+    {
+      "epoch": 0.038884500880976974,
+      "grad_norm": 0.23533456027507782,
+      "learning_rate": 9.968242534063827e-05,
+      "loss": 1.1855,
+      "step": 640
+    },
+    {
+      "epoch": 0.0389452579136035,
+      "grad_norm": 0.22506773471832275,
+      "learning_rate": 9.968134722107807e-05,
+      "loss": 1.319,
+      "step": 641
+    },
+    {
+      "epoch": 0.03900601494623002,
+      "grad_norm": 0.3768306374549866,
+      "learning_rate": 9.968026728043849e-05,
+      "loss": 1.1154,
+      "step": 642
+    },
+    {
+      "epoch": 0.03906677197885655,
+      "grad_norm": 0.2744412422180176,
+      "learning_rate": 9.96791855187591e-05,
+      "loss": 1.3277,
+      "step": 643
+    },
+    {
+      "epoch": 0.03912752901148308,
+      "grad_norm": 0.6380423307418823,
+      "learning_rate": 9.967810193607958e-05,
+      "loss": 1.1232,
+      "step": 644
+    },
+    {
+      "epoch": 0.03918828604410961,
+      "grad_norm": 0.2580445408821106,
+      "learning_rate": 9.967701653243963e-05,
+      "loss": 1.2685,
+      "step": 645
+    },
+    {
+      "epoch": 0.039249043076736136,
+      "grad_norm": 0.33161017298698425,
+      "learning_rate": 9.967592930787902e-05,
+      "loss": 1.1529,
+      "step": 646
+    },
+    {
+      "epoch": 0.03930980010936266,
+      "grad_norm": 0.21009348332881927,
+      "learning_rate": 9.967484026243763e-05,
+      "loss": 1.1526,
+      "step": 647
+    },
+    {
+      "epoch": 0.039370557141989185,
+      "grad_norm": 0.22496499121189117,
+      "learning_rate": 9.967374939615537e-05,
+      "loss": 1.1729,
+      "step": 648
+    },
+    {
+      "epoch": 0.03943131417461571,
+      "grad_norm": 0.2177799940109253,
+      "learning_rate": 9.967265670907222e-05,
+      "loss": 1.2111,
+      "step": 649
+    },
+    {
+      "epoch": 0.03949207120724224,
+      "grad_norm": 0.19962714612483978,
+      "learning_rate": 9.967156220122824e-05,
+      "loss": 1.1794,
+      "step": 650
+    },
+    {
+      "epoch": 0.03955282823986876,
+      "grad_norm": 0.3001613914966583,
+      "learning_rate": 9.967046587266354e-05,
+      "loss": 1.2756,
+      "step": 651
+    },
+    {
+      "epoch": 0.03961358527249529,
+      "grad_norm": 0.17096281051635742,
+      "learning_rate": 9.966936772341832e-05,
+      "loss": 1.2438,
+      "step": 652
+    },
+    {
+      "epoch": 0.03967434230512182,
+      "grad_norm": 0.3118416666984558,
+      "learning_rate": 9.966826775353281e-05,
+      "loss": 1.2722,
+      "step": 653
+    },
+    {
+      "epoch": 0.039735099337748346,
+      "grad_norm": 0.1986151784658432,
+      "learning_rate": 9.966716596304737e-05,
+      "loss": 1.1658,
+      "step": 654
+    },
+    {
+      "epoch": 0.039795856370374874,
+      "grad_norm": 0.2683010697364807,
+      "learning_rate": 9.966606235200234e-05,
+      "loss": 1.1902,
+      "step": 655
+    },
+    {
+      "epoch": 0.039856613403001395,
+      "grad_norm": 0.20436552166938782,
+      "learning_rate": 9.96649569204382e-05,
+      "loss": 1.2356,
+      "step": 656
+    },
+    {
+      "epoch": 0.03991737043562792,
+      "grad_norm": 0.31484872102737427,
+      "learning_rate": 9.966384966839547e-05,
+      "loss": 1.2013,
+      "step": 657
+    },
+    {
+      "epoch": 0.03997812746825445,
+      "grad_norm": 0.1810244768857956,
+      "learning_rate": 9.966274059591472e-05,
+      "loss": 1.1815,
+      "step": 658
+    },
+    {
+      "epoch": 0.04003888450088098,
+      "grad_norm": 0.21846987307071686,
+      "learning_rate": 9.966162970303663e-05,
+      "loss": 1.1703,
+      "step": 659
+    },
+    {
+      "epoch": 0.0400996415335075,
+      "grad_norm": 0.17850124835968018,
+      "learning_rate": 9.96605169898019e-05,
+      "loss": 1.1477,
+      "step": 660
+    },
+    {
+      "epoch": 0.04016039856613403,
+      "grad_norm": 0.23468884825706482,
+      "learning_rate": 9.965940245625131e-05,
+      "loss": 1.1484,
+      "step": 661
+    },
+    {
+      "epoch": 0.040221155598760557,
+      "grad_norm": 0.1681394726037979,
+      "learning_rate": 9.965828610242574e-05,
+      "loss": 1.1739,
+      "step": 662
+    },
+    {
+      "epoch": 0.040281912631387085,
+      "grad_norm": 0.3256566822528839,
+      "learning_rate": 9.965716792836606e-05,
+      "loss": 1.3391,
+      "step": 663
+    },
+    {
+      "epoch": 0.04034266966401361,
+      "grad_norm": 0.31519389152526855,
+      "learning_rate": 9.965604793411331e-05,
+      "loss": 1.3248,
+      "step": 664
+    },
+    {
+      "epoch": 0.040403426696640134,
+      "grad_norm": 0.38986799120903015,
+      "learning_rate": 9.965492611970854e-05,
+      "loss": 1.1911,
+      "step": 665
+    },
+    {
+      "epoch": 0.04046418372926666,
+      "grad_norm": 0.2666741907596588,
+      "learning_rate": 9.965380248519284e-05,
+      "loss": 1.1738,
+      "step": 666
+    },
+    {
+      "epoch": 0.04052494076189319,
+      "grad_norm": 4.861932277679443,
+      "learning_rate": 9.965267703060743e-05,
+      "loss": 1.2675,
+      "step": 667
+    },
+    {
+      "epoch": 0.04058569779451972,
+      "grad_norm": 0.2190086990594864,
+      "learning_rate": 9.965154975599352e-05,
+      "loss": 1.2847,
+      "step": 668
+    },
+    {
+      "epoch": 0.04064645482714624,
+      "grad_norm": 0.300801545381546,
+      "learning_rate": 9.965042066139248e-05,
+      "loss": 1.1476,
+      "step": 669
+    },
+    {
+      "epoch": 0.04070721185977277,
+      "grad_norm": 0.1510702222585678,
+      "learning_rate": 9.964928974684565e-05,
+      "loss": 1.1339,
+      "step": 670
+    },
+    {
+      "epoch": 0.040767968892399295,
+      "grad_norm": 0.250924289226532,
+      "learning_rate": 9.964815701239452e-05,
+      "loss": 1.2956,
+      "step": 671
+    },
+    {
+      "epoch": 0.04082872592502582,
+      "grad_norm": 0.3033612072467804,
+      "learning_rate": 9.964702245808059e-05,
+      "loss": 1.0902,
+      "step": 672
+    },
+    {
+      "epoch": 0.04088948295765235,
+      "grad_norm": 0.22391510009765625,
+      "learning_rate": 9.964588608394547e-05,
+      "loss": 1.3219,
+      "step": 673
+    },
+    {
+      "epoch": 0.04095023999027887,
+      "grad_norm": 0.28877395391464233,
+      "learning_rate": 9.96447478900308e-05,
+      "loss": 1.1541,
+      "step": 674
+    },
+    {
+      "epoch": 0.0410109970229054,
+      "grad_norm": 0.21077710390090942,
+      "learning_rate": 9.964360787637827e-05,
+      "loss": 1.2123,
+      "step": 675
+    },
+    {
+      "epoch": 0.04107175405553193,
+      "grad_norm": 0.2088579684495926,
+      "learning_rate": 9.964246604302974e-05,
+      "loss": 1.1197,
+      "step": 676
+    },
+    {
+      "epoch": 0.041132511088158456,
+      "grad_norm": 0.2369183748960495,
+      "learning_rate": 9.964132239002699e-05,
+      "loss": 1.2081,
+      "step": 677
+    },
+    {
+      "epoch": 0.04119326812078498,
+      "grad_norm": 0.35517820715904236,
+      "learning_rate": 9.964017691741198e-05,
+      "loss": 1.1635,
+      "step": 678
+    },
+    {
+      "epoch": 0.041254025153411505,
+      "grad_norm": 0.2329184114933014,
+      "learning_rate": 9.96390296252267e-05,
+      "loss": 1.1153,
+      "step": 679
+    },
+    {
+      "epoch": 0.041314782186038033,
+      "grad_norm": 0.249697744846344,
+      "learning_rate": 9.963788051351318e-05,
+      "loss": 1.308,
+      "step": 680
+    },
+    {
+      "epoch": 0.04137553921866456,
+      "grad_norm": 0.2712596356868744,
+      "learning_rate": 9.963672958231356e-05,
+      "loss": 1.2595,
+      "step": 681
+    },
+    {
+      "epoch": 0.04143629625129109,
+      "grad_norm": 0.15390761196613312,
+      "learning_rate": 9.963557683167003e-05,
+      "loss": 1.133,
+      "step": 682
+    },
+    {
+      "epoch": 0.04149705328391761,
+      "grad_norm": 0.31179946660995483,
+      "learning_rate": 9.963442226162482e-05,
+      "loss": 1.1638,
+      "step": 683
+    },
+    {
+      "epoch": 0.04155781031654414,
+      "grad_norm": 0.25089114904403687,
+      "learning_rate": 9.963326587222028e-05,
+      "loss": 1.1421,
+      "step": 684
+    },
+    {
+      "epoch": 0.04161856734917067,
+      "grad_norm": 0.2573951184749603,
+      "learning_rate": 9.96321076634988e-05,
+      "loss": 1.1881,
+      "step": 685
+    },
+    {
+      "epoch": 0.041679324381797195,
+      "grad_norm": 12.048274993896484,
+      "learning_rate": 9.96309476355028e-05,
+      "loss": 1.1963,
+      "step": 686
+    },
+    {
+      "epoch": 0.04174008141442372,
+      "grad_norm": 0.5792553424835205,
+      "learning_rate": 9.962978578827483e-05,
+      "loss": 1.1683,
+      "step": 687
+    },
+    {
+      "epoch": 0.041800838447050244,
+      "grad_norm": 0.26494306325912476,
+      "learning_rate": 9.962862212185746e-05,
+      "loss": 1.3472,
+      "step": 688
+    },
+    {
+      "epoch": 0.04186159547967677,
+      "grad_norm": 0.28178900480270386,
+      "learning_rate": 9.962745663629335e-05,
+      "loss": 1.2154,
+      "step": 689
+    },
+    {
+      "epoch": 0.0419223525123033,
+      "grad_norm": 0.340414434671402,
+      "learning_rate": 9.962628933162524e-05,
+      "loss": 1.2354,
+      "step": 690
+    },
+    {
+      "epoch": 0.04198310954492983,
+      "grad_norm": 0.24672427773475647,
+      "learning_rate": 9.962512020789588e-05,
+      "loss": 1.1686,
+      "step": 691
+    },
+    {
+      "epoch": 0.04204386657755635,
+      "grad_norm": 0.22035415470600128,
+      "learning_rate": 9.962394926514817e-05,
+      "loss": 1.1783,
+      "step": 692
+    },
+    {
+      "epoch": 0.04210462361018288,
+      "grad_norm": 0.4015302062034607,
+      "learning_rate": 9.9622776503425e-05,
+      "loss": 1.2413,
+      "step": 693
+    },
+    {
+      "epoch": 0.042165380642809405,
+      "grad_norm": 0.18768566846847534,
+      "learning_rate": 9.962160192276935e-05,
+      "loss": 1.1275,
+      "step": 694
+    },
+    {
+      "epoch": 0.04222613767543593,
+      "grad_norm": 0.3100888729095459,
+      "learning_rate": 9.962042552322431e-05,
+      "loss": 1.1383,
+      "step": 695
+    },
+    {
+      "epoch": 0.04228689470806246,
+      "grad_norm": 0.2774471938610077,
+      "learning_rate": 9.961924730483297e-05,
+      "loss": 1.2108,
+      "step": 696
+    },
+    {
+      "epoch": 0.04234765174068898,
+      "grad_norm": 0.510339617729187,
+      "learning_rate": 9.961806726763854e-05,
+      "loss": 1.2698,
+      "step": 697
+    },
+    {
+      "epoch": 0.04240840877331551,
+      "grad_norm": 0.4789173901081085,
+      "learning_rate": 9.961688541168424e-05,
+      "loss": 1.1169,
+      "step": 698
+    },
+    {
+      "epoch": 0.04246916580594204,
+      "grad_norm": 0.20484888553619385,
+      "learning_rate": 9.961570173701344e-05,
+      "loss": 1.1606,
+      "step": 699
+    },
+    {
+      "epoch": 0.04252992283856857,
+      "grad_norm": 0.30656594038009644,
+      "learning_rate": 9.961451624366949e-05,
+      "loss": 1.13,
+      "step": 700
+    },
+    {
+      "epoch": 0.04259067987119509,
+      "grad_norm": 0.33675429224967957,
+      "learning_rate": 9.961332893169586e-05,
+      "loss": 1.3475,
+      "step": 701
+    },
+    {
+      "epoch": 0.042651436903821616,
+      "grad_norm": 0.3515656888484955,
+      "learning_rate": 9.961213980113606e-05,
+      "loss": 1.1764,
+      "step": 702
+    },
+    {
+      "epoch": 0.042712193936448144,
+      "grad_norm": 0.43863317370414734,
+      "learning_rate": 9.96109488520337e-05,
+      "loss": 1.1448,
+      "step": 703
+    },
+    {
+      "epoch": 0.04277295096907467,
+      "grad_norm": 0.3262401521205902,
+      "learning_rate": 9.960975608443241e-05,
+      "loss": 1.1955,
+      "step": 704
+    },
+    {
+      "epoch": 0.0428337080017012,
+      "grad_norm": 0.28044047951698303,
+      "learning_rate": 9.960856149837592e-05,
+      "loss": 1.1635,
+      "step": 705
+    },
+    {
+      "epoch": 0.04289446503432772,
+      "grad_norm": 0.2565014362335205,
+      "learning_rate": 9.960736509390804e-05,
+      "loss": 1.1502,
+      "step": 706
+    },
+    {
+      "epoch": 0.04295522206695425,
+      "grad_norm": 0.2596074640750885,
+      "learning_rate": 9.960616687107258e-05,
+      "loss": 1.1355,
+      "step": 707
+    },
+    {
+      "epoch": 0.04301597909958078,
+      "grad_norm": 0.3365881145000458,
+      "learning_rate": 9.96049668299135e-05,
+      "loss": 1.1826,
+      "step": 708
+    },
+    {
+      "epoch": 0.043076736132207305,
+      "grad_norm": 0.20425531268119812,
+      "learning_rate": 9.960376497047476e-05,
+      "loss": 1.1858,
+      "step": 709
+    },
+    {
+      "epoch": 0.043137493164833826,
+      "grad_norm": 0.27772635221481323,
+      "learning_rate": 9.960256129280043e-05,
+      "loss": 1.1995,
+      "step": 710
+    },
+    {
+      "epoch": 0.043198250197460354,
+      "grad_norm": 0.9875108003616333,
+      "learning_rate": 9.960135579693462e-05,
+      "loss": 1.1236,
+      "step": 711
+    },
+    {
+      "epoch": 0.04325900723008688,
+      "grad_norm": 0.21602267026901245,
+      "learning_rate": 9.960014848292154e-05,
+      "loss": 1.1126,
+      "step": 712
+    },
+    {
+      "epoch": 0.04331976426271341,
+      "grad_norm": 0.19521121680736542,
+      "learning_rate": 9.959893935080542e-05,
+      "loss": 1.1219,
+      "step": 713
+    },
+    {
+      "epoch": 0.04338052129533994,
+      "grad_norm": 0.19386346638202667,
+      "learning_rate": 9.95977284006306e-05,
+      "loss": 1.2103,
+      "step": 714
+    },
+    {
+      "epoch": 0.04344127832796646,
+      "grad_norm": 0.2295754998922348,
+      "learning_rate": 9.959651563244146e-05,
+      "loss": 1.1571,
+      "step": 715
+    },
+    {
+      "epoch": 0.04350203536059299,
+      "grad_norm": 0.20171459019184113,
+      "learning_rate": 9.959530104628245e-05,
+      "loss": 1.1399,
+      "step": 716
+    },
+    {
+      "epoch": 0.043562792393219515,
+      "grad_norm": 0.216019868850708,
+      "learning_rate": 9.959408464219809e-05,
+      "loss": 1.248,
+      "step": 717
+    },
+    {
+      "epoch": 0.043623549425846044,
+      "grad_norm": 0.386947900056839,
+      "learning_rate": 9.959286642023297e-05,
+      "loss": 1.1545,
+      "step": 718
+    },
+    {
+      "epoch": 0.04368430645847257,
+      "grad_norm": 0.33577942848205566,
+      "learning_rate": 9.959164638043175e-05,
+      "loss": 1.1484,
+      "step": 719
+    },
+    {
+      "epoch": 0.04374506349109909,
+      "grad_norm": 0.24220316112041473,
+      "learning_rate": 9.959042452283914e-05,
+      "loss": 1.3661,
+      "step": 720
+    },
+    {
+      "epoch": 0.04380582052372562,
+      "grad_norm": 0.2601158916950226,
+      "learning_rate": 9.958920084749993e-05,
+      "loss": 1.2403,
+      "step": 721
+    },
+    {
+      "epoch": 0.04386657755635215,
+      "grad_norm": 0.29173552989959717,
+      "learning_rate": 9.958797535445899e-05,
+      "loss": 1.1332,
+      "step": 722
+    },
+    {
+      "epoch": 0.04392733458897868,
+      "grad_norm": 0.17891642451286316,
+      "learning_rate": 9.958674804376123e-05,
+      "loss": 1.1169,
+      "step": 723
+    },
+    {
+      "epoch": 0.0439880916216052,
+      "grad_norm": 0.3143637478351593,
+      "learning_rate": 9.958551891545161e-05,
+      "loss": 1.227,
+      "step": 724
+    },
+    {
+      "epoch": 0.044048848654231726,
+      "grad_norm": 0.22817392647266388,
+      "learning_rate": 9.958428796957523e-05,
+      "loss": 1.1594,
+      "step": 725
+    },
+    {
+      "epoch": 0.044109605686858254,
+      "grad_norm": 0.14670951664447784,
+      "learning_rate": 9.95830552061772e-05,
+      "loss": 1.1391,
+      "step": 726
+    },
+    {
+      "epoch": 0.04417036271948478,
+      "grad_norm": 0.1939353346824646,
+      "learning_rate": 9.958182062530268e-05,
+      "loss": 1.137,
+      "step": 727
+    },
+    {
+      "epoch": 0.04423111975211131,
+      "grad_norm": 0.22217488288879395,
+      "learning_rate": 9.958058422699695e-05,
+      "loss": 1.1166,
+      "step": 728
+    },
+    {
+      "epoch": 0.04429187678473783,
+      "grad_norm": 0.3429369032382965,
+      "learning_rate": 9.957934601130529e-05,
+      "loss": 1.337,
+      "step": 729
+    },
+    {
+      "epoch": 0.04435263381736436,
+      "grad_norm": 0.3005068898200989,
+      "learning_rate": 9.957810597827316e-05,
+      "loss": 1.262,
+      "step": 730
+    },
+    {
+      "epoch": 0.04441339084999089,
+      "grad_norm": 0.3939972519874573,
+      "learning_rate": 9.957686412794595e-05,
+      "loss": 1.1021,
+      "step": 731
+    },
+    {
+      "epoch": 0.044474147882617415,
+      "grad_norm": 0.20464569330215454,
+      "learning_rate": 9.957562046036921e-05,
+      "loss": 1.2641,
+      "step": 732
+    },
+    {
+      "epoch": 0.044534904915243936,
+      "grad_norm": 0.3335719406604767,
+      "learning_rate": 9.957437497558852e-05,
+      "loss": 1.1632,
+      "step": 733
+    },
+    {
+      "epoch": 0.044595661947870464,
+      "grad_norm": 0.35726863145828247,
+      "learning_rate": 9.957312767364953e-05,
+      "loss": 1.2758,
+      "step": 734
+    },
+    {
+      "epoch": 0.04465641898049699,
+      "grad_norm": 0.2093924731016159,
+      "learning_rate": 9.957187855459798e-05,
+      "loss": 1.1587,
+      "step": 735
+    },
+    {
+      "epoch": 0.04471717601312352,
+      "grad_norm": 0.3535689115524292,
+      "learning_rate": 9.957062761847962e-05,
+      "loss": 1.2929,
+      "step": 736
+    },
+    {
+      "epoch": 0.04477793304575005,
+      "grad_norm": 0.1873035728931427,
+      "learning_rate": 9.956937486534033e-05,
+      "loss": 1.1691,
+      "step": 737
+    },
+    {
+      "epoch": 0.04483869007837657,
+      "grad_norm": 0.22934308648109436,
+      "learning_rate": 9.956812029522602e-05,
+      "loss": 1.1608,
+      "step": 738
+    },
+    {
+      "epoch": 0.0448994471110031,
+      "grad_norm": 0.4806212782859802,
+      "learning_rate": 9.956686390818268e-05,
+      "loss": 1.3248,
+      "step": 739
+    },
+    {
+      "epoch": 0.044960204143629626,
+      "grad_norm": 0.38668379187583923,
+      "learning_rate": 9.956560570425635e-05,
+      "loss": 1.358,
+      "step": 740
+    },
+    {
+      "epoch": 0.045020961176256154,
+      "grad_norm": 0.3260784149169922,
+      "learning_rate": 9.956434568349318e-05,
+      "loss": 1.1787,
+      "step": 741
+    },
+    {
+      "epoch": 0.045081718208882675,
+      "grad_norm": 0.20555247366428375,
+      "learning_rate": 9.956308384593933e-05,
+      "loss": 1.1206,
+      "step": 742
+    },
+    {
+      "epoch": 0.0451424752415092,
+      "grad_norm": 0.21444964408874512,
+      "learning_rate": 9.956182019164106e-05,
+      "loss": 1.2573,
+      "step": 743
+    },
+    {
+      "epoch": 0.04520323227413573,
+      "grad_norm": 0.20223313570022583,
+      "learning_rate": 9.956055472064469e-05,
+      "loss": 1.1326,
+      "step": 744
+    },
+    {
+      "epoch": 0.04526398930676226,
+      "grad_norm": 0.17892780900001526,
+      "learning_rate": 9.955928743299662e-05,
+      "loss": 1.1954,
+      "step": 745
+    },
+    {
+      "epoch": 0.04532474633938879,
+      "grad_norm": 0.1672462522983551,
+      "learning_rate": 9.955801832874328e-05,
+      "loss": 1.1484,
+      "step": 746
+    },
+    {
+      "epoch": 0.04538550337201531,
+      "grad_norm": 0.36391040682792664,
+      "learning_rate": 9.955674740793119e-05,
+      "loss": 1.2516,
+      "step": 747
+    },
+    {
+      "epoch": 0.045446260404641836,
+      "grad_norm": 0.38535046577453613,
+      "learning_rate": 9.955547467060696e-05,
+      "loss": 1.2676,
+      "step": 748
+    },
+    {
+      "epoch": 0.045507017437268364,
+      "grad_norm": 0.27136051654815674,
+      "learning_rate": 9.955420011681722e-05,
+      "loss": 1.1743,
+      "step": 749
+    },
+    {
+      "epoch": 0.04556777446989489,
+      "grad_norm": 0.19951479136943817,
+      "learning_rate": 9.955292374660869e-05,
+      "loss": 1.1381,
+      "step": 750
+    },
+    {
+      "epoch": 0.04562853150252142,
+      "grad_norm": 0.1706264764070511,
+      "learning_rate": 9.955164556002819e-05,
+      "loss": 1.1095,
+      "step": 751
+    },
+    {
+      "epoch": 0.04568928853514794,
+      "grad_norm": 0.2307135909795761,
+      "learning_rate": 9.955036555712252e-05,
+      "loss": 1.1944,
+      "step": 752
+    },
+    {
+      "epoch": 0.04575004556777447,
+      "grad_norm": 0.5656225085258484,
+      "learning_rate": 9.954908373793861e-05,
+      "loss": 1.5714,
+      "step": 753
+    },
+    {
+      "epoch": 0.045810802600401,
+      "grad_norm": 0.26978978514671326,
+      "learning_rate": 9.954780010252347e-05,
+      "loss": 1.1083,
+      "step": 754
+    },
+    {
+      "epoch": 0.045871559633027525,
+      "grad_norm": 0.21326936781406403,
+      "learning_rate": 9.954651465092414e-05,
+      "loss": 1.124,
+      "step": 755
+    },
+    {
+      "epoch": 0.04593231666565405,
+      "grad_norm": 0.4203796088695526,
+      "learning_rate": 9.954522738318776e-05,
+      "loss": 1.2392,
+      "step": 756
+    },
+    {
+      "epoch": 0.045993073698280575,
+      "grad_norm": 0.27247968316078186,
+      "learning_rate": 9.954393829936146e-05,
+      "loss": 1.1137,
+      "step": 757
+    },
+    {
+      "epoch": 0.0460538307309071,
+      "grad_norm": 0.31686800718307495,
+      "learning_rate": 9.954264739949255e-05,
+      "loss": 1.2118,
+      "step": 758
+    },
+    {
+      "epoch": 0.04611458776353363,
+      "grad_norm": 0.3327178359031677,
+      "learning_rate": 9.954135468362831e-05,
+      "loss": 1.1996,
+      "step": 759
+    },
+    {
+      "epoch": 0.04617534479616016,
+      "grad_norm": 0.228616401553154,
+      "learning_rate": 9.954006015181614e-05,
+      "loss": 1.272,
+      "step": 760
+    },
+    {
+      "epoch": 0.04623610182878668,
+      "grad_norm": 0.5685213208198547,
+      "learning_rate": 9.95387638041035e-05,
+      "loss": 1.0946,
+      "step": 761
+    },
+    {
+      "epoch": 0.04629685886141321,
+      "grad_norm": 0.40294498205184937,
+      "learning_rate": 9.953746564053788e-05,
+      "loss": 1.1519,
+      "step": 762
+    },
+    {
+      "epoch": 0.046357615894039736,
+      "grad_norm": 0.3475242853164673,
+      "learning_rate": 9.953616566116689e-05,
+      "loss": 1.2288,
+      "step": 763
+    },
+    {
+      "epoch": 0.046418372926666264,
+      "grad_norm": 0.2769295275211334,
+      "learning_rate": 9.953486386603816e-05,
+      "loss": 1.107,
+      "step": 764
+    },
+    {
+      "epoch": 0.046479129959292785,
+      "grad_norm": 0.19877159595489502,
+      "learning_rate": 9.953356025519945e-05,
+      "loss": 1.1911,
+      "step": 765
+    },
+    {
+      "epoch": 0.04653988699191931,
+      "grad_norm": 0.32342344522476196,
+      "learning_rate": 9.953225482869849e-05,
+      "loss": 1.2127,
+      "step": 766
+    },
+    {
+      "epoch": 0.04660064402454584,
+      "grad_norm": 0.4708094596862793,
+      "learning_rate": 9.953094758658315e-05,
+      "loss": 1.1604,
+      "step": 767
+    },
+    {
+      "epoch": 0.04666140105717237,
+      "grad_norm": 1.619902491569519,
+      "learning_rate": 9.952963852890137e-05,
+      "loss": 1.1045,
+      "step": 768
+    },
+    {
+      "epoch": 0.0467221580897989,
+      "grad_norm": 0.6743023991584778,
+      "learning_rate": 9.95283276557011e-05,
+      "loss": 1.2388,
+      "step": 769
+    },
+    {
+      "epoch": 0.04678291512242542,
+      "grad_norm": 0.2565208375453949,
+      "learning_rate": 9.952701496703043e-05,
+      "loss": 1.1911,
+      "step": 770
+    },
+    {
+      "epoch": 0.046843672155051946,
+      "grad_norm": 0.5599068999290466,
+      "learning_rate": 9.952570046293743e-05,
+      "loss": 1.189,
+      "step": 771
+    },
+    {
+      "epoch": 0.046904429187678474,
+      "grad_norm": 0.22475752234458923,
+      "learning_rate": 9.952438414347032e-05,
+      "loss": 1.1731,
+      "step": 772
+    },
+    {
+      "epoch": 0.046965186220305,
+      "grad_norm": 0.18163125216960907,
+      "learning_rate": 9.952306600867734e-05,
+      "loss": 1.1635,
+      "step": 773
+    },
+    {
+      "epoch": 0.047025943252931524,
+      "grad_norm": 0.23896318674087524,
+      "learning_rate": 9.952174605860678e-05,
+      "loss": 1.2031,
+      "step": 774
+    },
+    {
+      "epoch": 0.04708670028555805,
+      "grad_norm": 0.23537391424179077,
+      "learning_rate": 9.952042429330705e-05,
+      "loss": 1.1295,
+      "step": 775
+    },
+    {
+      "epoch": 0.04714745731818458,
+      "grad_norm": 0.17851340770721436,
+      "learning_rate": 9.951910071282662e-05,
+      "loss": 1.138,
+      "step": 776
+    },
+    {
+      "epoch": 0.04720821435081111,
+      "grad_norm": 0.25157350301742554,
+      "learning_rate": 9.951777531721396e-05,
+      "loss": 1.1467,
+      "step": 777
+    },
+    {
+      "epoch": 0.047268971383437636,
+      "grad_norm": 0.19467604160308838,
+      "learning_rate": 9.951644810651767e-05,
+      "loss": 1.2312,
+      "step": 778
+    },
+    {
+      "epoch": 0.04732972841606416,
+      "grad_norm": 0.34383106231689453,
+      "learning_rate": 9.951511908078644e-05,
+      "loss": 1.3084,
+      "step": 779
+    },
+    {
+      "epoch": 0.047390485448690685,
+      "grad_norm": 0.2738543450832367,
+      "learning_rate": 9.951378824006891e-05,
+      "loss": 1.2193,
+      "step": 780
+    },
+    {
+      "epoch": 0.04745124248131721,
+      "grad_norm": 0.3195165991783142,
+      "learning_rate": 9.951245558441391e-05,
+      "loss": 1.2006,
+      "step": 781
+    },
+    {
+      "epoch": 0.04751199951394374,
+      "grad_norm": 0.20014366507530212,
+      "learning_rate": 9.951112111387029e-05,
+      "loss": 1.148,
+      "step": 782
+    },
+    {
+      "epoch": 0.04757275654657027,
+      "grad_norm": 0.7697231769561768,
+      "learning_rate": 9.950978482848693e-05,
+      "loss": 1.2827,
+      "step": 783
+    },
+    {
+      "epoch": 0.04763351357919679,
+      "grad_norm": 0.21268193423748016,
+      "learning_rate": 9.950844672831288e-05,
+      "loss": 1.2048,
+      "step": 784
+    },
+    {
+      "epoch": 0.04769427061182332,
+      "grad_norm": 0.20504657924175262,
+      "learning_rate": 9.950710681339712e-05,
+      "loss": 1.114,
+      "step": 785
+    },
+    {
+      "epoch": 0.047755027644449846,
+      "grad_norm": 0.2534680664539337,
+      "learning_rate": 9.95057650837888e-05,
+      "loss": 1.1585,
+      "step": 786
+    },
+    {
+      "epoch": 0.047815784677076374,
+      "grad_norm": 0.23410208523273468,
+      "learning_rate": 9.950442153953707e-05,
+      "loss": 1.1851,
+      "step": 787
+    },
+    {
+      "epoch": 0.047876541709702895,
+      "grad_norm": 0.21911850571632385,
+      "learning_rate": 9.950307618069122e-05,
+      "loss": 1.1967,
+      "step": 788
+    },
+    {
+      "epoch": 0.04793729874232942,
+      "grad_norm": 0.2831370234489441,
+      "learning_rate": 9.950172900730053e-05,
+      "loss": 1.2854,
+      "step": 789
+    },
+    {
+      "epoch": 0.04799805577495595,
+      "grad_norm": 0.20728982985019684,
+      "learning_rate": 9.950038001941441e-05,
+      "loss": 1.2317,
+      "step": 790
+    },
+    {
+      "epoch": 0.04805881280758248,
+      "grad_norm": 0.32236918807029724,
+      "learning_rate": 9.949902921708228e-05,
+      "loss": 1.2141,
+      "step": 791
+    },
+    {
+      "epoch": 0.04811956984020901,
+      "grad_norm": 0.2485806941986084,
+      "learning_rate": 9.949767660035368e-05,
+      "loss": 1.1434,
+      "step": 792
+    },
+    {
+      "epoch": 0.04818032687283553,
+      "grad_norm": 0.19720609486103058,
+      "learning_rate": 9.949632216927815e-05,
+      "loss": 1.1489,
+      "step": 793
+    },
+    {
+      "epoch": 0.04824108390546206,
+      "grad_norm": 0.2985354959964752,
+      "learning_rate": 9.949496592390539e-05,
+      "loss": 1.2711,
+      "step": 794
+    },
+    {
+      "epoch": 0.048301840938088585,
+      "grad_norm": 0.2601931393146515,
+      "learning_rate": 9.949360786428508e-05,
+      "loss": 1.1099,
+      "step": 795
+    },
+    {
+      "epoch": 0.04836259797071511,
+      "grad_norm": 0.21803413331508636,
+      "learning_rate": 9.949224799046699e-05,
+      "loss": 1.3067,
+      "step": 796
+    },
+    {
+      "epoch": 0.048423355003341634,
+      "grad_norm": 0.1801166832447052,
+      "learning_rate": 9.9490886302501e-05,
+      "loss": 1.1214,
+      "step": 797
+    },
+    {
+      "epoch": 0.04848411203596816,
+      "grad_norm": 0.16227313876152039,
+      "learning_rate": 9.948952280043701e-05,
+      "loss": 1.1485,
+      "step": 798
+    },
+    {
+      "epoch": 0.04854486906859469,
+      "grad_norm": 0.1986556202173233,
+      "learning_rate": 9.948815748432499e-05,
+      "loss": 1.1811,
+      "step": 799
+    },
+    {
+      "epoch": 0.04860562610122122,
+      "grad_norm": 0.2063964605331421,
+      "learning_rate": 9.948679035421499e-05,
+      "loss": 1.1029,
+      "step": 800
+    },
+    {
+      "epoch": 0.048666383133847746,
+      "grad_norm": 0.21714679896831512,
+      "learning_rate": 9.948542141015714e-05,
+      "loss": 1.2234,
+      "step": 801
+    },
+    {
+      "epoch": 0.04872714016647427,
+      "grad_norm": 0.19539695978164673,
+      "learning_rate": 9.948405065220156e-05,
+      "loss": 1.1159,
+      "step": 802
+    },
+    {
+      "epoch": 0.048787897199100795,
+      "grad_norm": 0.8304261565208435,
+      "learning_rate": 9.948267808039857e-05,
+      "loss": 1.1999,
+      "step": 803
+    },
+    {
+      "epoch": 0.04884865423172732,
+      "grad_norm": 0.2534937858581543,
+      "learning_rate": 9.948130369479844e-05,
+      "loss": 1.1281,
+      "step": 804
+    },
+    {
+      "epoch": 0.04890941126435385,
+      "grad_norm": 0.6079438328742981,
+      "learning_rate": 9.947992749545158e-05,
+      "loss": 1.1047,
+      "step": 805
+    },
+    {
+      "epoch": 0.04897016829698037,
+      "grad_norm": 0.7486149668693542,
+      "learning_rate": 9.94785494824084e-05,
+      "loss": 1.3219,
+      "step": 806
+    },
+    {
+      "epoch": 0.0490309253296069,
+      "grad_norm": 0.40212923288345337,
+      "learning_rate": 9.947716965571941e-05,
+      "loss": 1.1158,
+      "step": 807
+    },
+    {
+      "epoch": 0.04909168236223343,
+      "grad_norm": 0.32530421018600464,
+      "learning_rate": 9.947578801543522e-05,
+      "loss": 1.1595,
+      "step": 808
+    },
+    {
+      "epoch": 0.049152439394859956,
+      "grad_norm": 0.3300507962703705,
+      "learning_rate": 9.947440456160645e-05,
+      "loss": 1.1426,
+      "step": 809
+    },
+    {
+      "epoch": 0.049213196427486484,
+      "grad_norm": 0.2577906548976898,
+      "learning_rate": 9.947301929428384e-05,
+      "loss": 1.1482,
+      "step": 810
+    },
+    {
+      "epoch": 0.049273953460113006,
+      "grad_norm": 93.94365692138672,
+      "learning_rate": 9.94716322135181e-05,
+      "loss": 1.1666,
+      "step": 811
+    },
+    {
+      "epoch": 0.049334710492739534,
+      "grad_norm": 1.0377168655395508,
+      "learning_rate": 9.947024331936015e-05,
+      "loss": 1.1194,
+      "step": 812
+    },
+    {
+      "epoch": 0.04939546752536606,
+      "grad_norm": 0.2052767127752304,
+      "learning_rate": 9.946885261186086e-05,
+      "loss": 1.1368,
+      "step": 813
+    },
+    {
+      "epoch": 0.04945622455799259,
+      "grad_norm": 0.28331485390663147,
+      "learning_rate": 9.946746009107121e-05,
+      "loss": 1.126,
+      "step": 814
+    },
+    {
+      "epoch": 0.04951698159061911,
+      "grad_norm": 0.2663123309612274,
+      "learning_rate": 9.946606575704227e-05,
+      "loss": 1.1422,
+      "step": 815
+    },
+    {
+      "epoch": 0.04957773862324564,
+      "grad_norm": 0.3345920741558075,
+      "learning_rate": 9.946466960982512e-05,
+      "loss": 1.225,
+      "step": 816
+    },
+    {
+      "epoch": 0.04963849565587217,
+      "grad_norm": 0.27361583709716797,
+      "learning_rate": 9.946327164947094e-05,
+      "loss": 1.1195,
+      "step": 817
+    },
+    {
+      "epoch": 0.049699252688498695,
+      "grad_norm": 0.42706793546676636,
+      "learning_rate": 9.946187187603097e-05,
+      "loss": 1.157,
+      "step": 818
+    },
+    {
+      "epoch": 0.04976000972112522,
+      "grad_norm": 0.20006580650806427,
+      "learning_rate": 9.946047028955651e-05,
+      "loss": 1.2072,
+      "step": 819
+    },
+    {
+      "epoch": 0.049820766753751744,
+      "grad_norm": 0.33667150139808655,
+      "learning_rate": 9.945906689009898e-05,
+      "loss": 1.2,
+      "step": 820
+    },
+    {
+      "epoch": 0.04988152378637827,
+      "grad_norm": 0.1654760092496872,
+      "learning_rate": 9.94576616777098e-05,
+      "loss": 1.1349,
+      "step": 821
+    },
+    {
+      "epoch": 0.0499422808190048,
+      "grad_norm": 0.2234649658203125,
+      "learning_rate": 9.945625465244044e-05,
+      "loss": 1.1722,
+      "step": 822
+    },
+    {
+      "epoch": 0.05000303785163133,
+      "grad_norm": 0.2198670357465744,
+      "learning_rate": 9.945484581434253e-05,
+      "loss": 1.1624,
+      "step": 823
+    },
+    {
+      "epoch": 0.050063794884257856,
+      "grad_norm": 0.3813166916370392,
+      "learning_rate": 9.945343516346768e-05,
+      "loss": 1.1479,
+      "step": 824
+    },
+    {
+      "epoch": 0.05012455191688438,
+      "grad_norm": 0.17954643070697784,
+      "learning_rate": 9.94520226998676e-05,
+      "loss": 1.168,
+      "step": 825
+    },
+    {
+      "epoch": 0.050185308949510905,
+      "grad_norm": 0.3486316204071045,
+      "learning_rate": 9.945060842359408e-05,
+      "loss": 1.1609,
+      "step": 826
+    },
+    {
+      "epoch": 0.05024606598213743,
+      "grad_norm": 0.18670342862606049,
+      "learning_rate": 9.944919233469894e-05,
+      "loss": 1.2275,
+      "step": 827
+    },
+    {
+      "epoch": 0.05030682301476396,
+      "grad_norm": 0.42580702900886536,
+      "learning_rate": 9.944777443323409e-05,
+      "loss": 1.2689,
+      "step": 828
+    },
+    {
+      "epoch": 0.05036758004739048,
+      "grad_norm": 0.1723891943693161,
+      "learning_rate": 9.944635471925153e-05,
+      "loss": 1.1456,
+      "step": 829
+    },
+    {
+      "epoch": 0.05042833708001701,
+      "grad_norm": 0.4415867328643799,
+      "learning_rate": 9.944493319280326e-05,
+      "loss": 1.171,
+      "step": 830
+    },
+    {
+      "epoch": 0.05048909411264354,
+      "grad_norm": 0.18282508850097656,
+      "learning_rate": 9.944350985394142e-05,
+      "loss": 1.1425,
+      "step": 831
+    },
+    {
+      "epoch": 0.05054985114527007,
+      "grad_norm": 5.300233364105225,
+      "learning_rate": 9.944208470271816e-05,
+      "loss": 1.1379,
+      "step": 832
+    },
+    {
+      "epoch": 0.050610608177896595,
+      "grad_norm": 0.4158828854560852,
+      "learning_rate": 9.944065773918574e-05,
+      "loss": 1.1716,
+      "step": 833
+    },
+    {
+      "epoch": 0.050671365210523116,
+      "grad_norm": 0.9802979230880737,
+      "learning_rate": 9.943922896339644e-05,
+      "loss": 1.2526,
+      "step": 834
+    },
+    {
+      "epoch": 0.050732122243149644,
+      "grad_norm": 0.15631850063800812,
+      "learning_rate": 9.943779837540266e-05,
+      "loss": 1.1664,
+      "step": 835
+    },
+    {
+      "epoch": 0.05079287927577617,
+      "grad_norm": 0.39770379662513733,
+      "learning_rate": 9.94363659752568e-05,
+      "loss": 1.1575,
+      "step": 836
+    },
+    {
+      "epoch": 0.0508536363084027,
+      "grad_norm": 0.2634415030479431,
+      "learning_rate": 9.943493176301142e-05,
+      "loss": 1.1118,
+      "step": 837
+    },
+    {
+      "epoch": 0.05091439334102922,
+      "grad_norm": 0.4317116439342499,
+      "learning_rate": 9.943349573871905e-05,
+      "loss": 1.4348,
+      "step": 838
+    },
+    {
+      "epoch": 0.05097515037365575,
+      "grad_norm": 0.4688839316368103,
+      "learning_rate": 9.943205790243233e-05,
+      "loss": 1.0981,
+      "step": 839
+    },
+    {
+      "epoch": 0.05103590740628228,
+      "grad_norm": 0.3700367510318756,
+      "learning_rate": 9.943061825420398e-05,
+      "loss": 1.168,
+      "step": 840
+    },
+    {
+      "epoch": 0.051096664438908805,
+      "grad_norm": 0.18428249657154083,
+      "learning_rate": 9.942917679408676e-05,
+      "loss": 1.2174,
+      "step": 841
+    },
+    {
+      "epoch": 0.05115742147153533,
+      "grad_norm": 0.1904333531856537,
+      "learning_rate": 9.94277335221335e-05,
+      "loss": 1.1403,
+      "step": 842
+    },
+    {
+      "epoch": 0.051218178504161854,
+      "grad_norm": 0.23415088653564453,
+      "learning_rate": 9.942628843839714e-05,
+      "loss": 1.1146,
+      "step": 843
+    },
+    {
+      "epoch": 0.05127893553678838,
+      "grad_norm": 0.17599831521511078,
+      "learning_rate": 9.94248415429306e-05,
+      "loss": 1.1422,
+      "step": 844
+    },
+    {
+      "epoch": 0.05133969256941491,
+      "grad_norm": 0.3157031238079071,
+      "learning_rate": 9.942339283578694e-05,
+      "loss": 1.1025,
+      "step": 845
+    },
+    {
+      "epoch": 0.05140044960204144,
+      "grad_norm": 0.27406126260757446,
+      "learning_rate": 9.942194231701929e-05,
+      "loss": 1.3205,
+      "step": 846
+    },
+    {
+      "epoch": 0.05146120663466796,
+      "grad_norm": 0.534218966960907,
+      "learning_rate": 9.942048998668077e-05,
+      "loss": 1.2346,
+      "step": 847
+    },
+    {
+      "epoch": 0.05152196366729449,
+      "grad_norm": 0.24832171201705933,
+      "learning_rate": 9.941903584482463e-05,
+      "loss": 1.2309,
+      "step": 848
+    },
+    {
+      "epoch": 0.051582720699921016,
+      "grad_norm": 0.3907499313354492,
+      "learning_rate": 9.941757989150419e-05,
+      "loss": 1.2425,
+      "step": 849
+    },
+    {
+      "epoch": 0.051643477732547544,
+      "grad_norm": 0.17559818923473358,
+      "learning_rate": 9.94161221267728e-05,
+      "loss": 1.1128,
+      "step": 850
+    },
+    {
+      "epoch": 0.05170423476517407,
+      "grad_norm": 0.7275607585906982,
+      "learning_rate": 9.94146625506839e-05,
+      "loss": 1.3044,
+      "step": 851
+    },
+    {
+      "epoch": 0.05176499179780059,
+      "grad_norm": 0.37362074851989746,
+      "learning_rate": 9.9413201163291e-05,
+      "loss": 1.2876,
+      "step": 852
+    },
+    {
+      "epoch": 0.05182574883042712,
+      "grad_norm": 0.3065381348133087,
+      "learning_rate": 9.941173796464767e-05,
+      "loss": 1.1509,
+      "step": 853
+    },
+    {
+      "epoch": 0.05188650586305365,
+      "grad_norm": 0.19520434737205505,
+      "learning_rate": 9.941027295480752e-05,
+      "loss": 1.1182,
+      "step": 854
+    },
+    {
+      "epoch": 0.05194726289568018,
+      "grad_norm": 0.26167190074920654,
+      "learning_rate": 9.940880613382427e-05,
+      "loss": 1.217,
+      "step": 855
+    },
+    {
+      "epoch": 0.052008019928306705,
+      "grad_norm": 0.3808404505252838,
+      "learning_rate": 9.940733750175166e-05,
+      "loss": 1.2523,
+      "step": 856
+    },
+    {
+      "epoch": 0.052068776960933226,
+      "grad_norm": 0.3537200391292572,
+      "learning_rate": 9.940586705864357e-05,
+      "loss": 1.2519,
+      "step": 857
+    },
+    {
+      "epoch": 0.052129533993559754,
+      "grad_norm": 0.2563754916191101,
+      "learning_rate": 9.940439480455386e-05,
+      "loss": 1.1331,
+      "step": 858
+    },
+    {
+      "epoch": 0.05219029102618628,
+      "grad_norm": 0.3466266393661499,
+      "learning_rate": 9.940292073953652e-05,
+      "loss": 1.1245,
+      "step": 859
+    },
+    {
+      "epoch": 0.05225104805881281,
+      "grad_norm": 0.4410223662853241,
+      "learning_rate": 9.940144486364556e-05,
+      "loss": 1.0973,
+      "step": 860
+    },
+    {
+      "epoch": 0.05231180509143933,
+      "grad_norm": 0.15505509078502655,
+      "learning_rate": 9.93999671769351e-05,
+      "loss": 1.1085,
+      "step": 861
+    },
+    {
+      "epoch": 0.05237256212406586,
+      "grad_norm": 0.3102237582206726,
+      "learning_rate": 9.939848767945927e-05,
+      "loss": 1.1551,
+      "step": 862
+    },
+    {
+      "epoch": 0.05243331915669239,
+      "grad_norm": 0.2113533914089203,
+      "learning_rate": 9.939700637127235e-05,
+      "loss": 1.1252,
+      "step": 863
+    },
+    {
+      "epoch": 0.052494076189318915,
+      "grad_norm": 0.37543654441833496,
+      "learning_rate": 9.93955232524286e-05,
+      "loss": 1.0799,
+      "step": 864
+    },
+    {
+      "epoch": 0.05255483322194544,
+      "grad_norm": 0.18159475922584534,
+      "learning_rate": 9.93940383229824e-05,
+      "loss": 1.1994,
+      "step": 865
+    },
+    {
+      "epoch": 0.052615590254571964,
+      "grad_norm": 0.2303076982498169,
+      "learning_rate": 9.939255158298819e-05,
+      "loss": 1.1605,
+      "step": 866
+    },
+    {
+      "epoch": 0.05267634728719849,
+      "grad_norm": 0.20846877992153168,
+      "learning_rate": 9.939106303250044e-05,
+      "loss": 1.1441,
+      "step": 867
+    },
+    {
+      "epoch": 0.05273710431982502,
+      "grad_norm": 0.215975821018219,
+      "learning_rate": 9.938957267157374e-05,
+      "loss": 1.1633,
+      "step": 868
+    },
+    {
+      "epoch": 0.05279786135245155,
+      "grad_norm": 0.18180370330810547,
+      "learning_rate": 9.93880805002627e-05,
+      "loss": 1.1145,
+      "step": 869
+    },
+    {
+      "epoch": 0.05285861838507807,
+      "grad_norm": 0.4241054654121399,
+      "learning_rate": 9.938658651862201e-05,
+      "loss": 1.2631,
+      "step": 870
+    },
+    {
+      "epoch": 0.0529193754177046,
+      "grad_norm": 0.48412010073661804,
+      "learning_rate": 9.938509072670646e-05,
+      "loss": 1.1602,
+      "step": 871
+    },
+    {
+      "epoch": 0.052980132450331126,
+      "grad_norm": 0.24033743143081665,
+      "learning_rate": 9.938359312457087e-05,
+      "loss": 1.2714,
+      "step": 872
+    },
+    {
+      "epoch": 0.053040889482957654,
+      "grad_norm": 0.2354423701763153,
+      "learning_rate": 9.938209371227011e-05,
+      "loss": 1.1154,
+      "step": 873
+    },
+    {
+      "epoch": 0.05310164651558418,
+      "grad_norm": 0.19644786417484283,
+      "learning_rate": 9.938059248985916e-05,
+      "loss": 1.1415,
+      "step": 874
+    },
+    {
+      "epoch": 0.0531624035482107,
+      "grad_norm": 0.1432892233133316,
+      "learning_rate": 9.937908945739307e-05,
+      "loss": 1.114,
+      "step": 875
+    },
+    {
+      "epoch": 0.05322316058083723,
+      "grad_norm": 0.30041730403900146,
+      "learning_rate": 9.93775846149269e-05,
+      "loss": 1.2071,
+      "step": 876
+    },
+    {
+      "epoch": 0.05328391761346376,
+      "grad_norm": 0.24715879559516907,
+      "learning_rate": 9.937607796251582e-05,
+      "loss": 1.2008,
+      "step": 877
+    },
+    {
+      "epoch": 0.05334467464609029,
+      "grad_norm": 0.2856390178203583,
+      "learning_rate": 9.937456950021504e-05,
+      "loss": 1.1197,
+      "step": 878
+    },
+    {
+      "epoch": 0.05340543167871681,
+      "grad_norm": 0.589713454246521,
+      "learning_rate": 9.937305922807989e-05,
+      "loss": 1.1317,
+      "step": 879
+    },
+    {
+      "epoch": 0.053466188711343336,
+      "grad_norm": 0.2540833055973053,
+      "learning_rate": 9.937154714616572e-05,
+      "loss": 1.2142,
+      "step": 880
+    },
+    {
+      "epoch": 0.053526945743969864,
+      "grad_norm": 0.37999507784843445,
+      "learning_rate": 9.937003325452792e-05,
+      "loss": 1.1927,
+      "step": 881
+    },
+    {
+      "epoch": 0.05358770277659639,
+      "grad_norm": 0.2898063659667969,
+      "learning_rate": 9.936851755322203e-05,
+      "loss": 1.1661,
+      "step": 882
+    },
+    {
+      "epoch": 0.05364845980922292,
+      "grad_norm": 0.20970571041107178,
+      "learning_rate": 9.936700004230358e-05,
+      "loss": 1.0903,
+      "step": 883
+    },
+    {
+      "epoch": 0.05370921684184944,
+      "grad_norm": 0.5133906602859497,
+      "learning_rate": 9.93654807218282e-05,
+      "loss": 1.194,
+      "step": 884
+    },
+    {
+      "epoch": 0.05376997387447597,
+      "grad_norm": 0.2385426014661789,
+      "learning_rate": 9.936395959185158e-05,
+      "loss": 1.2504,
+      "step": 885
+    },
+    {
+      "epoch": 0.0538307309071025,
+      "grad_norm": 1.2981534004211426,
+      "learning_rate": 9.936243665242948e-05,
+      "loss": 1.1405,
+      "step": 886
+    },
+    {
+      "epoch": 0.053891487939729026,
+      "grad_norm": 0.4661676287651062,
+      "learning_rate": 9.936091190361772e-05,
+      "loss": 1.1215,
+      "step": 887
+    },
+    {
+      "epoch": 0.053952244972355554,
+      "grad_norm": 0.20790480077266693,
+      "learning_rate": 9.93593853454722e-05,
+      "loss": 1.2572,
+      "step": 888
+    },
+    {
+      "epoch": 0.054013002004982075,
+      "grad_norm": 0.23346084356307983,
+      "learning_rate": 9.935785697804887e-05,
+      "loss": 1.1955,
+      "step": 889
+    },
+    {
+      "epoch": 0.0540737590376086,
+      "grad_norm": 0.31828176975250244,
+      "learning_rate": 9.935632680140374e-05,
+      "loss": 1.1165,
+      "step": 890
+    },
+    {
+      "epoch": 0.05413451607023513,
+      "grad_norm": 0.2110651433467865,
+      "learning_rate": 9.935479481559291e-05,
+      "loss": 1.2794,
+      "step": 891
+    },
+    {
+      "epoch": 0.05419527310286166,
+      "grad_norm": 0.2790307402610779,
+      "learning_rate": 9.935326102067255e-05,
+      "loss": 1.131,
+      "step": 892
+    },
+    {
+      "epoch": 0.05425603013548818,
+      "grad_norm": 0.2054932862520218,
+      "learning_rate": 9.935172541669886e-05,
+      "loss": 1.0983,
+      "step": 893
+    },
+    {
+      "epoch": 0.05431678716811471,
+      "grad_norm": 0.23212817311286926,
+      "learning_rate": 9.935018800372813e-05,
+      "loss": 1.1096,
+      "step": 894
+    },
+    {
+      "epoch": 0.054377544200741236,
+      "grad_norm": 0.19015608727931976,
+      "learning_rate": 9.934864878181673e-05,
+      "loss": 1.1842,
+      "step": 895
+    },
+    {
+      "epoch": 0.054438301233367764,
+      "grad_norm": 0.48252880573272705,
+      "learning_rate": 9.934710775102107e-05,
+      "loss": 1.2139,
+      "step": 896
+    },
+    {
+      "epoch": 0.05449905826599429,
+      "grad_norm": 0.2064215987920761,
+      "learning_rate": 9.934556491139764e-05,
+      "loss": 1.2348,
+      "step": 897
+    },
+    {
+      "epoch": 0.05455981529862081,
+      "grad_norm": 0.15281763672828674,
+      "learning_rate": 9.934402026300298e-05,
+      "loss": 1.1138,
+      "step": 898
+    },
+    {
+      "epoch": 0.05462057233124734,
+      "grad_norm": 0.32750558853149414,
+      "learning_rate": 9.934247380589371e-05,
+      "loss": 1.2344,
+      "step": 899
+    },
+    {
+      "epoch": 0.05468132936387387,
+      "grad_norm": 0.21798034012317657,
+      "learning_rate": 9.934092554012655e-05,
+      "loss": 1.1157,
+      "step": 900
+    },
+    {
+      "epoch": 0.0547420863965004,
+      "grad_norm": 0.20733627676963806,
+      "learning_rate": 9.933937546575822e-05,
+      "loss": 1.132,
+      "step": 901
+    },
+    {
+      "epoch": 0.05480284342912692,
+      "grad_norm": 0.2235470861196518,
+      "learning_rate": 9.933782358284553e-05,
+      "loss": 1.1446,
+      "step": 902
+    },
+    {
+      "epoch": 0.054863600461753446,
+      "grad_norm": 0.1842838078737259,
+      "learning_rate": 9.93362698914454e-05,
+      "loss": 1.1237,
+      "step": 903
+    },
+    {
+      "epoch": 0.054924357494379974,
+      "grad_norm": 0.42488864064216614,
+      "learning_rate": 9.933471439161475e-05,
+      "loss": 1.1635,
+      "step": 904
+    },
+    {
+      "epoch": 0.0549851145270065,
+      "grad_norm": 0.15811462700366974,
+      "learning_rate": 9.93331570834106e-05,
+      "loss": 1.1065,
+      "step": 905
+    },
+    {
+      "epoch": 0.05504587155963303,
+      "grad_norm": 0.39768218994140625,
+      "learning_rate": 9.933159796689005e-05,
+      "loss": 1.2619,
+      "step": 906
+    },
+    {
+      "epoch": 0.05510662859225955,
+      "grad_norm": 0.29817453026771545,
+      "learning_rate": 9.933003704211024e-05,
+      "loss": 1.1998,
+      "step": 907
+    },
+    {
+      "epoch": 0.05516738562488608,
+      "grad_norm": 0.2882938086986542,
+      "learning_rate": 9.93284743091284e-05,
+      "loss": 1.132,
+      "step": 908
+    },
+    {
+      "epoch": 0.05522814265751261,
+      "grad_norm": 0.21322005987167358,
+      "learning_rate": 9.932690976800177e-05,
+      "loss": 1.1996,
+      "step": 909
+    },
+    {
+      "epoch": 0.055288899690139136,
+      "grad_norm": 0.29061540961265564,
+      "learning_rate": 9.932534341878774e-05,
+      "loss": 1.3895,
+      "step": 910
+    },
+    {
+      "epoch": 0.05534965672276566,
+      "grad_norm": 0.37948229908943176,
+      "learning_rate": 9.932377526154373e-05,
+      "loss": 1.2223,
+      "step": 911
+    },
+    {
+      "epoch": 0.055410413755392185,
+      "grad_norm": 0.17468896508216858,
+      "learning_rate": 9.932220529632719e-05,
+      "loss": 1.0767,
+      "step": 912
+    },
+    {
+      "epoch": 0.05547117078801871,
+      "grad_norm": 0.432624876499176,
+      "learning_rate": 9.932063352319567e-05,
+      "loss": 1.1254,
+      "step": 913
+    },
+    {
+      "epoch": 0.05553192782064524,
+      "grad_norm": 0.25776907801628113,
+      "learning_rate": 9.931905994220681e-05,
+      "loss": 1.2517,
+      "step": 914
+    },
+    {
+      "epoch": 0.05559268485327177,
+      "grad_norm": 0.19674532115459442,
+      "learning_rate": 9.931748455341828e-05,
+      "loss": 1.1216,
+      "step": 915
+    },
+    {
+      "epoch": 0.05565344188589829,
+      "grad_norm": 0.23850758373737335,
+      "learning_rate": 9.93159073568878e-05,
+      "loss": 1.1144,
+      "step": 916
+    },
+    {
+      "epoch": 0.05571419891852482,
+      "grad_norm": 0.18916510045528412,
+      "learning_rate": 9.931432835267322e-05,
+      "loss": 1.1824,
+      "step": 917
+    },
+    {
+      "epoch": 0.055774955951151346,
+      "grad_norm": 0.35642287135124207,
+      "learning_rate": 9.931274754083239e-05,
+      "loss": 1.1282,
+      "step": 918
+    },
+    {
+      "epoch": 0.055835712983777874,
+      "grad_norm": 0.22358252108097076,
+      "learning_rate": 9.931116492142328e-05,
+      "loss": 1.1392,
+      "step": 919
+    },
+    {
+      "epoch": 0.055896470016404395,
+      "grad_norm": 0.35258379578590393,
+      "learning_rate": 9.930958049450389e-05,
+      "loss": 1.195,
+      "step": 920
+    },
+    {
+      "epoch": 0.05595722704903092,
+      "grad_norm": 2.5283114910125732,
+      "learning_rate": 9.930799426013229e-05,
+      "loss": 1.1529,
+      "step": 921
+    },
+    {
+      "epoch": 0.05601798408165745,
+      "grad_norm": 0.38866716623306274,
+      "learning_rate": 9.930640621836662e-05,
+      "loss": 1.122,
+      "step": 922
+    },
+    {
+      "epoch": 0.05607874111428398,
+      "grad_norm": 0.2504761815071106,
+      "learning_rate": 9.93048163692651e-05,
+      "loss": 1.1247,
+      "step": 923
+    },
+    {
+      "epoch": 0.05613949814691051,
+      "grad_norm": 0.31465429067611694,
+      "learning_rate": 9.930322471288603e-05,
+      "loss": 1.1066,
+      "step": 924
+    },
+    {
+      "epoch": 0.05620025517953703,
+      "grad_norm": 0.38542166352272034,
+      "learning_rate": 9.93016312492877e-05,
+      "loss": 1.1648,
+      "step": 925
+    },
+    {
+      "epoch": 0.05626101221216356,
+      "grad_norm": 0.2616014778614044,
+      "learning_rate": 9.930003597852855e-05,
+      "loss": 1.1363,
+      "step": 926
+    },
+    {
+      "epoch": 0.056321769244790085,
+      "grad_norm": 0.4364415109157562,
+      "learning_rate": 9.929843890066705e-05,
+      "loss": 1.2609,
+      "step": 927
+    },
+    {
+      "epoch": 0.05638252627741661,
+      "grad_norm": 0.9745305776596069,
+      "learning_rate": 9.929684001576176e-05,
+      "loss": 1.2179,
+      "step": 928
+    },
+    {
+      "epoch": 0.05644328331004314,
+      "grad_norm": 0.28503185510635376,
+      "learning_rate": 9.929523932387126e-05,
+      "loss": 1.0999,
+      "step": 929
+    },
+    {
+      "epoch": 0.05650404034266966,
+      "grad_norm": 0.2756706476211548,
+      "learning_rate": 9.929363682505424e-05,
+      "loss": 1.2154,
+      "step": 930
+    },
+    {
+      "epoch": 0.05656479737529619,
+      "grad_norm": 0.22424040734767914,
+      "learning_rate": 9.929203251936942e-05,
+      "loss": 1.1516,
+      "step": 931
+    },
+    {
+      "epoch": 0.05662555440792272,
+      "grad_norm": 0.2764267027378082,
+      "learning_rate": 9.929042640687565e-05,
+      "loss": 1.1456,
+      "step": 932
+    },
+    {
+      "epoch": 0.056686311440549246,
+      "grad_norm": 0.2318948656320572,
+      "learning_rate": 9.928881848763174e-05,
+      "loss": 1.2542,
+      "step": 933
+    },
+    {
+      "epoch": 0.05674706847317577,
+      "grad_norm": 0.1942901760339737,
+      "learning_rate": 9.928720876169668e-05,
+      "loss": 1.1262,
+      "step": 934
+    },
+    {
+      "epoch": 0.056807825505802295,
+      "grad_norm": 0.2438351958990097,
+      "learning_rate": 9.928559722912944e-05,
+      "loss": 1.1353,
+      "step": 935
+    },
+    {
+      "epoch": 0.05686858253842882,
+      "grad_norm": 0.2117772400379181,
+      "learning_rate": 9.928398388998911e-05,
+      "loss": 1.1261,
+      "step": 936
+    },
+    {
+      "epoch": 0.05692933957105535,
+      "grad_norm": 2.341944932937622,
+      "learning_rate": 9.928236874433485e-05,
+      "loss": 1.1439,
+      "step": 937
+    },
+    {
+      "epoch": 0.05699009660368188,
+      "grad_norm": 0.3281532824039459,
+      "learning_rate": 9.92807517922258e-05,
+      "loss": 1.1972,
+      "step": 938
+    },
+    {
+      "epoch": 0.0570508536363084,
+      "grad_norm": 0.17223651707172394,
+      "learning_rate": 9.92791330337213e-05,
+      "loss": 1.0931,
+      "step": 939
+    },
+    {
+      "epoch": 0.05711161066893493,
+      "grad_norm": 0.5078747868537903,
+      "learning_rate": 9.927751246888064e-05,
+      "loss": 1.4061,
+      "step": 940
+    },
+    {
+      "epoch": 0.057172367701561456,
+      "grad_norm": 0.2017948180437088,
+      "learning_rate": 9.927589009776322e-05,
+      "loss": 1.1117,
+      "step": 941
+    },
+    {
+      "epoch": 0.057233124734187985,
+      "grad_norm": 0.2222599983215332,
+      "learning_rate": 9.927426592042854e-05,
+      "loss": 1.1915,
+      "step": 942
+    },
+    {
+      "epoch": 0.057293881766814506,
+      "grad_norm": 0.19689585268497467,
+      "learning_rate": 9.927263993693612e-05,
+      "loss": 1.1632,
+      "step": 943
+    },
+    {
+      "epoch": 0.057354638799441034,
+      "grad_norm": 0.38269299268722534,
+      "learning_rate": 9.927101214734555e-05,
+      "loss": 1.2244,
+      "step": 944
+    },
+    {
+      "epoch": 0.05741539583206756,
+      "grad_norm": 0.2522115707397461,
+      "learning_rate": 9.926938255171652e-05,
+      "loss": 1.2768,
+      "step": 945
+    },
+    {
+      "epoch": 0.05747615286469409,
+      "grad_norm": 0.22336071729660034,
+      "learning_rate": 9.926775115010873e-05,
+      "loss": 1.1996,
+      "step": 946
+    },
+    {
+      "epoch": 0.05753690989732062,
+      "grad_norm": 0.27844613790512085,
+      "learning_rate": 9.9266117942582e-05,
+      "loss": 1.1432,
+      "step": 947
+    },
+    {
+      "epoch": 0.05759766692994714,
+      "grad_norm": 0.23788277804851532,
+      "learning_rate": 9.926448292919622e-05,
+      "loss": 1.2991,
+      "step": 948
+    },
+    {
+      "epoch": 0.05765842396257367,
+      "grad_norm": 0.24779294431209564,
+      "learning_rate": 9.926284611001126e-05,
+      "loss": 1.2684,
+      "step": 949
+    },
+    {
+      "epoch": 0.057719180995200195,
+      "grad_norm": 0.7948906421661377,
+      "learning_rate": 9.926120748508719e-05,
+      "loss": 1.3878,
+      "step": 950
+    },
+    {
+      "epoch": 0.05777993802782672,
+      "grad_norm": 0.24091163277626038,
+      "learning_rate": 9.925956705448401e-05,
+      "loss": 1.1803,
+      "step": 951
+    },
+    {
+      "epoch": 0.057840695060453244,
+      "grad_norm": 0.4433269500732422,
+      "learning_rate": 9.925792481826188e-05,
+      "loss": 1.2907,
+      "step": 952
+    },
+    {
+      "epoch": 0.05790145209307977,
+      "grad_norm": 0.1771276295185089,
+      "learning_rate": 9.9256280776481e-05,
+      "loss": 1.1229,
+      "step": 953
+    },
+    {
+      "epoch": 0.0579622091257063,
+      "grad_norm": 0.17962491512298584,
+      "learning_rate": 9.925463492920163e-05,
+      "loss": 1.1335,
+      "step": 954
+    },
+    {
+      "epoch": 0.05802296615833283,
+      "grad_norm": 0.1824798583984375,
+      "learning_rate": 9.92529872764841e-05,
+      "loss": 1.1585,
+      "step": 955
+    },
+    {
+      "epoch": 0.058083723190959356,
+      "grad_norm": 0.34185677766799927,
+      "learning_rate": 9.925133781838878e-05,
+      "loss": 1.1331,
+      "step": 956
+    },
+    {
+      "epoch": 0.05814448022358588,
+      "grad_norm": 0.1740175038576126,
+      "learning_rate": 9.924968655497617e-05,
+      "loss": 1.1418,
+      "step": 957
+    },
+    {
+      "epoch": 0.058205237256212405,
+      "grad_norm": 0.4179982841014862,
+      "learning_rate": 9.924803348630678e-05,
+      "loss": 1.2057,
+      "step": 958
+    },
+    {
+      "epoch": 0.05826599428883893,
+      "grad_norm": 0.3665924370288849,
+      "learning_rate": 9.924637861244121e-05,
+      "loss": 1.3906,
+      "step": 959
+    },
+    {
+      "epoch": 0.05832675132146546,
+      "grad_norm": 0.6215278506278992,
+      "learning_rate": 9.92447219334401e-05,
+      "loss": 1.269,
+      "step": 960
+    },
+    {
+      "epoch": 0.05838750835409199,
+      "grad_norm": 0.24217715859413147,
+      "learning_rate": 9.924306344936418e-05,
+      "loss": 1.1126,
+      "step": 961
+    },
+    {
+      "epoch": 0.05844826538671851,
+      "grad_norm": 0.5626233220100403,
+      "learning_rate": 9.924140316027426e-05,
+      "loss": 1.277,
+      "step": 962
+    },
+    {
+      "epoch": 0.05850902241934504,
+      "grad_norm": 0.1770382672548294,
+      "learning_rate": 9.92397410662312e-05,
+      "loss": 1.1785,
+      "step": 963
+    },
+    {
+      "epoch": 0.05856977945197157,
+      "grad_norm": 2.129831314086914,
+      "learning_rate": 9.92380771672959e-05,
+      "loss": 1.188,
+      "step": 964
+    },
+    {
+      "epoch": 0.058630536484598095,
+      "grad_norm": 0.3123109042644501,
+      "learning_rate": 9.923641146352938e-05,
+      "loss": 1.2407,
+      "step": 965
+    },
+    {
+      "epoch": 0.058691293517224616,
+      "grad_norm": 0.22422635555267334,
+      "learning_rate": 9.923474395499265e-05,
+      "loss": 1.2625,
+      "step": 966
+    },
+    {
+      "epoch": 0.058752050549851144,
+      "grad_norm": 0.6242901682853699,
+      "learning_rate": 9.923307464174689e-05,
+      "loss": 1.1412,
+      "step": 967
+    },
+    {
+      "epoch": 0.05881280758247767,
+      "grad_norm": 0.17358358204364777,
+      "learning_rate": 9.923140352385325e-05,
+      "loss": 1.0924,
+      "step": 968
+    },
+    {
+      "epoch": 0.0588735646151042,
+      "grad_norm": 0.20891009271144867,
+      "learning_rate": 9.922973060137301e-05,
+      "loss": 1.17,
+      "step": 969
+    },
+    {
+      "epoch": 0.05893432164773073,
+      "grad_norm": 0.4106835126876831,
+      "learning_rate": 9.922805587436748e-05,
+      "loss": 1.1235,
+      "step": 970
+    },
+    {
+      "epoch": 0.05899507868035725,
+      "grad_norm": 0.23005026578903198,
+      "learning_rate": 9.922637934289802e-05,
+      "loss": 1.1308,
+      "step": 971
+    },
+    {
+      "epoch": 0.05905583571298378,
+      "grad_norm": 0.19174528121948242,
+      "learning_rate": 9.922470100702615e-05,
+      "loss": 1.1512,
+      "step": 972
+    },
+    {
+      "epoch": 0.059116592745610305,
+      "grad_norm": 0.1874125599861145,
+      "learning_rate": 9.922302086681332e-05,
+      "loss": 1.1523,
+      "step": 973
+    },
+    {
+      "epoch": 0.05917734977823683,
+      "grad_norm": 0.5835013389587402,
+      "learning_rate": 9.922133892232115e-05,
+      "loss": 1.1181,
+      "step": 974
+    },
+    {
+      "epoch": 0.059238106810863354,
+      "grad_norm": 0.2347194254398346,
+      "learning_rate": 9.921965517361129e-05,
+      "loss": 1.1478,
+      "step": 975
+    },
+    {
+      "epoch": 0.05929886384348988,
+      "grad_norm": 0.40479937195777893,
+      "learning_rate": 9.921796962074545e-05,
+      "loss": 1.1101,
+      "step": 976
+    },
+    {
+      "epoch": 0.05935962087611641,
+      "grad_norm": 0.6876659393310547,
+      "learning_rate": 9.921628226378543e-05,
+      "loss": 1.1141,
+      "step": 977
+    },
+    {
+      "epoch": 0.05942037790874294,
+      "grad_norm": 0.3584234416484833,
+      "learning_rate": 9.921459310279306e-05,
+      "loss": 1.2632,
+      "step": 978
+    },
+    {
+      "epoch": 0.059481134941369466,
+      "grad_norm": 0.2755407989025116,
+      "learning_rate": 9.921290213783027e-05,
+      "loss": 1.2233,
+      "step": 979
+    },
+    {
+      "epoch": 0.05954189197399599,
+      "grad_norm": 0.3321934640407562,
+      "learning_rate": 9.921120936895904e-05,
+      "loss": 1.1421,
+      "step": 980
+    },
+    {
+      "epoch": 0.059602649006622516,
+      "grad_norm": 0.4014555513858795,
+      "learning_rate": 9.920951479624143e-05,
+      "loss": 1.1824,
+      "step": 981
+    },
+    {
+      "epoch": 0.059663406039249044,
+      "grad_norm": 0.24272578954696655,
+      "learning_rate": 9.920781841973955e-05,
+      "loss": 1.1869,
+      "step": 982
+    },
+    {
+      "epoch": 0.05972416307187557,
+      "grad_norm": 0.24495117366313934,
+      "learning_rate": 9.920612023951554e-05,
+      "loss": 1.1139,
+      "step": 983
+    },
+    {
+      "epoch": 0.05978492010450209,
+      "grad_norm": 1.0873308181762695,
+      "learning_rate": 9.920442025563171e-05,
+      "loss": 1.2191,
+      "step": 984
+    },
+    {
+      "epoch": 0.05984567713712862,
+      "grad_norm": 0.20630523562431335,
+      "learning_rate": 9.920271846815032e-05,
+      "loss": 1.1433,
+      "step": 985
+    },
+    {
+      "epoch": 0.05990643416975515,
+      "grad_norm": 0.36086443066596985,
+      "learning_rate": 9.92010148771338e-05,
+      "loss": 1.2207,
+      "step": 986
+    },
+    {
+      "epoch": 0.05996719120238168,
+      "grad_norm": 0.2526913583278656,
+      "learning_rate": 9.919930948264456e-05,
+      "loss": 1.1345,
+      "step": 987
+    },
+    {
+      "epoch": 0.060027948235008205,
+      "grad_norm": 0.3324904143810272,
+      "learning_rate": 9.91976022847451e-05,
+      "loss": 1.105,
+      "step": 988
+    },
+    {
+      "epoch": 0.060088705267634726,
+      "grad_norm": 0.2521546185016632,
+      "learning_rate": 9.919589328349805e-05,
+      "loss": 1.2172,
+      "step": 989
+    },
+    {
+      "epoch": 0.060149462300261254,
+      "grad_norm": 0.2850136160850525,
+      "learning_rate": 9.919418247896602e-05,
+      "loss": 1.3185,
+      "step": 990
+    },
+    {
+      "epoch": 0.06021021933288778,
+      "grad_norm": 0.24153576791286469,
+      "learning_rate": 9.919246987121169e-05,
+      "loss": 1.1532,
+      "step": 991
+    },
+    {
+      "epoch": 0.06027097636551431,
+      "grad_norm": 0.17544732987880707,
+      "learning_rate": 9.91907554602979e-05,
+      "loss": 1.0686,
+      "step": 992
+    },
+    {
+      "epoch": 0.06033173339814084,
+      "grad_norm": 1.1323429346084595,
+      "learning_rate": 9.918903924628744e-05,
+      "loss": 1.1567,
+      "step": 993
+    },
+    {
+      "epoch": 0.06039249043076736,
+      "grad_norm": 0.23179759085178375,
+      "learning_rate": 9.918732122924326e-05,
+      "loss": 1.1906,
+      "step": 994
+    },
+    {
+      "epoch": 0.06045324746339389,
+      "grad_norm": 0.20516924560070038,
+      "learning_rate": 9.91856014092283e-05,
+      "loss": 1.1259,
+      "step": 995
+    },
+    {
+      "epoch": 0.060514004496020415,
+      "grad_norm": 0.18624304234981537,
+      "learning_rate": 9.91838797863056e-05,
+      "loss": 1.1802,
+      "step": 996
+    },
+    {
+      "epoch": 0.06057476152864694,
+      "grad_norm": 0.4013606607913971,
+      "learning_rate": 9.91821563605383e-05,
+      "loss": 1.0975,
+      "step": 997
+    },
+    {
+      "epoch": 0.060635518561273465,
+      "grad_norm": 0.5464007258415222,
+      "learning_rate": 9.918043113198955e-05,
+      "loss": 1.3132,
+      "step": 998
+    },
+    {
+      "epoch": 0.06069627559389999,
+      "grad_norm": 0.18759983777999878,
+      "learning_rate": 9.917870410072258e-05,
+      "loss": 1.2185,
+      "step": 999
+    },
+    {
+      "epoch": 0.06075703262652652,
+      "grad_norm": 0.1758163571357727,
+      "learning_rate": 9.917697526680072e-05,
+      "loss": 1.1618,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06081778965915305,
+      "grad_norm": 0.18332824110984802,
+      "learning_rate": 9.917524463028731e-05,
+      "loss": 1.1097,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06087854669177958,
+      "grad_norm": 0.16281358897686005,
+      "learning_rate": 9.91735121912458e-05,
+      "loss": 1.139,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0609393037244061,
+      "grad_norm": 0.1521851271390915,
+      "learning_rate": 9.91717779497397e-05,
+      "loss": 1.1511,
+      "step": 1003
+    },
+    {
+      "epoch": 0.061000060757032626,
+      "grad_norm": 0.17558541893959045,
+      "learning_rate": 9.917004190583259e-05,
+      "loss": 1.0927,
+      "step": 1004
+    },
+    {
+      "epoch": 0.061060817789659154,
+      "grad_norm": 0.17117518186569214,
+      "learning_rate": 9.916830405958808e-05,
+      "loss": 1.1405,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06112157482228568,
+      "grad_norm": 0.3612566888332367,
+      "learning_rate": 9.916656441106989e-05,
+      "loss": 1.2507,
+      "step": 1006
+    },
+    {
+      "epoch": 0.0611823318549122,
+      "grad_norm": 0.17844192683696747,
+      "learning_rate": 9.916482296034175e-05,
+      "loss": 1.1921,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06124308888753873,
+      "grad_norm": 0.29219844937324524,
+      "learning_rate": 9.916307970746754e-05,
+      "loss": 1.2054,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06130384592016526,
+      "grad_norm": 0.16926388442516327,
+      "learning_rate": 9.916133465251114e-05,
+      "loss": 1.1033,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06136460295279179,
+      "grad_norm": 0.22389206290245056,
+      "learning_rate": 9.915958779553652e-05,
+      "loss": 1.1324,
+      "step": 1010
+    },
+    {
+      "epoch": 0.061425359985418315,
+      "grad_norm": 0.21128493547439575,
+      "learning_rate": 9.91578391366077e-05,
+      "loss": 1.2314,
+      "step": 1011
+    },
+    {
+      "epoch": 0.061486117018044836,
+      "grad_norm": 0.20081967115402222,
+      "learning_rate": 9.915608867578878e-05,
+      "loss": 1.1144,
+      "step": 1012
+    },
+    {
+      "epoch": 0.061546874050671364,
+      "grad_norm": 0.25086820125579834,
+      "learning_rate": 9.915433641314394e-05,
+      "loss": 1.1498,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06160763108329789,
+      "grad_norm": 0.14907516539096832,
+      "learning_rate": 9.915258234873739e-05,
+      "loss": 1.1583,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06166838811592442,
+      "grad_norm": 0.2024078518152237,
+      "learning_rate": 9.915082648263343e-05,
+      "loss": 1.1158,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06172914514855094,
+      "grad_norm": 0.29856154322624207,
+      "learning_rate": 9.914906881489643e-05,
+      "loss": 1.2516,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06178990218117747,
+      "grad_norm": 5.346932888031006,
+      "learning_rate": 9.914730934559083e-05,
+      "loss": 1.0974,
+      "step": 1017
+    },
+    {
+      "epoch": 0.061850659213804,
+      "grad_norm": 0.35687682032585144,
+      "learning_rate": 9.914554807478108e-05,
+      "loss": 1.1446,
+      "step": 1018
+    },
+    {
+      "epoch": 0.061911416246430526,
+      "grad_norm": 0.1732306331396103,
+      "learning_rate": 9.914378500253176e-05,
+      "loss": 1.2398,
+      "step": 1019
+    },
+    {
+      "epoch": 0.061972173279057054,
+      "grad_norm": 0.2185167372226715,
+      "learning_rate": 9.914202012890751e-05,
+      "loss": 1.14,
+      "step": 1020
+    },
+    {
+      "epoch": 0.062032930311683575,
+      "grad_norm": 0.19063043594360352,
+      "learning_rate": 9.914025345397304e-05,
+      "loss": 1.1288,
+      "step": 1021
+    },
+    {
+      "epoch": 0.0620936873443101,
+      "grad_norm": 0.14763686060905457,
+      "learning_rate": 9.913848497779306e-05,
+      "loss": 1.1335,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06215444437693663,
+      "grad_norm": 0.23980973660945892,
+      "learning_rate": 9.91367147004324e-05,
+      "loss": 1.183,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06221520140956316,
+      "grad_norm": 0.212006077170372,
+      "learning_rate": 9.913494262195597e-05,
+      "loss": 1.2123,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06227595844218968,
+      "grad_norm": 5.338240623474121,
+      "learning_rate": 9.913316874242874e-05,
+      "loss": 1.1678,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06233671547481621,
+      "grad_norm": 0.25872495770454407,
+      "learning_rate": 9.91313930619157e-05,
+      "loss": 1.2411,
+      "step": 1026
+    },
+    {
+      "epoch": 0.062397472507442736,
+      "grad_norm": 0.3396739363670349,
+      "learning_rate": 9.912961558048195e-05,
+      "loss": 1.1413,
+      "step": 1027
+    },
+    {
+      "epoch": 0.062458229540069264,
+      "grad_norm": 0.22687095403671265,
+      "learning_rate": 9.912783629819264e-05,
+      "loss": 1.2663,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06251898657269579,
+      "grad_norm": 0.43820720911026,
+      "learning_rate": 9.912605521511301e-05,
+      "loss": 1.2936,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06257974360532231,
+      "grad_norm": 0.30604350566864014,
+      "learning_rate": 9.912427233130832e-05,
+      "loss": 1.1777,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06264050063794885,
+      "grad_norm": 0.2056499719619751,
+      "learning_rate": 9.912248764684393e-05,
+      "loss": 1.1757,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06270125767057537,
+      "grad_norm": 0.43428412079811096,
+      "learning_rate": 9.912070116178526e-05,
+      "loss": 1.0841,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06276201470320189,
+      "grad_norm": 0.219242662191391,
+      "learning_rate": 9.91189128761978e-05,
+      "loss": 1.138,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06282277173582843,
+      "grad_norm": 0.20234309136867523,
+      "learning_rate": 9.911712279014709e-05,
+      "loss": 1.1664,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06288352876845495,
+      "grad_norm": 0.2271445393562317,
+      "learning_rate": 9.911533090369876e-05,
+      "loss": 1.323,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06294428580108148,
+      "grad_norm": 0.2031880021095276,
+      "learning_rate": 9.911353721691848e-05,
+      "loss": 1.1477,
+      "step": 1036
+    },
+    {
+      "epoch": 0.063005042833708,
+      "grad_norm": 0.1386653035879135,
+      "learning_rate": 9.911174172987199e-05,
+      "loss": 1.111,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06306579986633452,
+      "grad_norm": 0.23107148706912994,
+      "learning_rate": 9.910994444262512e-05,
+      "loss": 1.2201,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06312655689896106,
+      "grad_norm": 0.14069907367229462,
+      "learning_rate": 9.910814535524372e-05,
+      "loss": 1.1235,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06318731393158758,
+      "grad_norm": 0.18750937283039093,
+      "learning_rate": 9.910634446779379e-05,
+      "loss": 1.1225,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0632480709642141,
+      "grad_norm": 0.2907664179801941,
+      "learning_rate": 9.91045417803413e-05,
+      "loss": 1.1408,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06330882799684064,
+      "grad_norm": 0.22893452644348145,
+      "learning_rate": 9.910273729295235e-05,
+      "loss": 1.1074,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06336958502946716,
+      "grad_norm": 0.7455631494522095,
+      "learning_rate": 9.910093100569305e-05,
+      "loss": 1.1824,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06343034206209369,
+      "grad_norm": 0.17006048560142517,
+      "learning_rate": 9.909912291862966e-05,
+      "loss": 1.1108,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06349109909472021,
+      "grad_norm": 0.17600016295909882,
+      "learning_rate": 9.909731303182841e-05,
+      "loss": 1.1277,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06355185612734673,
+      "grad_norm": 0.2076251208782196,
+      "learning_rate": 9.909550134535566e-05,
+      "loss": 1.1285,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06361261315997327,
+      "grad_norm": 0.2192397117614746,
+      "learning_rate": 9.909368785927783e-05,
+      "loss": 1.4415,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06367337019259979,
+      "grad_norm": 0.4094569683074951,
+      "learning_rate": 9.909187257366137e-05,
+      "loss": 1.1842,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06373412722522633,
+      "grad_norm": 0.1654943972826004,
+      "learning_rate": 9.909005548857284e-05,
+      "loss": 1.1334,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06379488425785285,
+      "grad_norm": 0.31593313813209534,
+      "learning_rate": 9.908823660407884e-05,
+      "loss": 1.1099,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06385564129047937,
+      "grad_norm": 0.5052706003189087,
+      "learning_rate": 9.908641592024603e-05,
+      "loss": 1.1977,
+      "step": 1051
+    },
+    {
+      "epoch": 0.0639163983231059,
+      "grad_norm": 0.17280252277851105,
+      "learning_rate": 9.908459343714116e-05,
+      "loss": 1.1342,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06397715535573242,
+      "grad_norm": 0.34011808037757874,
+      "learning_rate": 9.908276915483104e-05,
+      "loss": 1.0999,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06403791238835896,
+      "grad_norm": 0.20396362245082855,
+      "learning_rate": 9.908094307338251e-05,
+      "loss": 1.1962,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06409866942098548,
+      "grad_norm": 0.21117782592773438,
+      "learning_rate": 9.907911519286254e-05,
+      "loss": 1.1473,
+      "step": 1055
+    },
+    {
+      "epoch": 0.064159426453612,
+      "grad_norm": 0.22127699851989746,
+      "learning_rate": 9.90772855133381e-05,
+      "loss": 1.1186,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06422018348623854,
+      "grad_norm": 0.2519165575504303,
+      "learning_rate": 9.907545403487629e-05,
+      "loss": 1.2315,
+      "step": 1057
+    },
+    {
+      "epoch": 0.06428094051886506,
+      "grad_norm": 0.17791754007339478,
+      "learning_rate": 9.907362075754422e-05,
+      "loss": 1.1335,
+      "step": 1058
+    },
+    {
+      "epoch": 0.06434169755149159,
+      "grad_norm": 0.45089244842529297,
+      "learning_rate": 9.907178568140909e-05,
+      "loss": 1.1355,
+      "step": 1059
+    },
+    {
+      "epoch": 0.06440245458411811,
+      "grad_norm": 0.19349536299705505,
+      "learning_rate": 9.906994880653818e-05,
+      "loss": 1.1226,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06446321161674463,
+      "grad_norm": 0.41839855909347534,
+      "learning_rate": 9.90681101329988e-05,
+      "loss": 1.308,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06452396864937117,
+      "grad_norm": 0.2293994426727295,
+      "learning_rate": 9.906626966085837e-05,
+      "loss": 1.1184,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06458472568199769,
+      "grad_norm": 0.19593217968940735,
+      "learning_rate": 9.906442739018433e-05,
+      "loss": 1.1379,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06464548271462421,
+      "grad_norm": 0.26556774973869324,
+      "learning_rate": 9.906258332104424e-05,
+      "loss": 1.1278,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06470623974725075,
+      "grad_norm": 0.37029293179512024,
+      "learning_rate": 9.906073745350565e-05,
+      "loss": 1.1897,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06476699677987727,
+      "grad_norm": 0.29654473066329956,
+      "learning_rate": 9.905888978763627e-05,
+      "loss": 1.1712,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0648277538125038,
+      "grad_norm": 0.24746425449848175,
+      "learning_rate": 9.905704032350378e-05,
+      "loss": 1.1696,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06488851084513032,
+      "grad_norm": 0.2603473961353302,
+      "learning_rate": 9.9055189061176e-05,
+      "loss": 1.1313,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06494926787775684,
+      "grad_norm": 0.21180987358093262,
+      "learning_rate": 9.905333600072079e-05,
+      "loss": 1.1054,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06501002491038338,
+      "grad_norm": 0.23979009687900543,
+      "learning_rate": 9.905148114220606e-05,
+      "loss": 1.2503,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0650707819430099,
+      "grad_norm": 0.19146323204040527,
+      "learning_rate": 9.904962448569981e-05,
+      "loss": 1.1472,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06513153897563644,
+      "grad_norm": 0.1779710352420807,
+      "learning_rate": 9.904776603127009e-05,
+      "loss": 1.1426,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06519229600826296,
+      "grad_norm": 0.21693828701972961,
+      "learning_rate": 9.904590577898503e-05,
+      "loss": 1.157,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06525305304088948,
+      "grad_norm": 0.16904886066913605,
+      "learning_rate": 9.904404372891281e-05,
+      "loss": 1.0952,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06531381007351601,
+      "grad_norm": 0.18482241034507751,
+      "learning_rate": 9.90421798811217e-05,
+      "loss": 1.1501,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06537456710614253,
+      "grad_norm": 0.19291286170482635,
+      "learning_rate": 9.904031423567998e-05,
+      "loss": 1.1276,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06543532413876907,
+      "grad_norm": 0.16024720668792725,
+      "learning_rate": 9.903844679265608e-05,
+      "loss": 1.0988,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06549608117139559,
+      "grad_norm": 0.27875852584838867,
+      "learning_rate": 9.903657755211844e-05,
+      "loss": 1.2271,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06555683820402211,
+      "grad_norm": 0.31460678577423096,
+      "learning_rate": 9.903470651413555e-05,
+      "loss": 1.2298,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06561759523664865,
+      "grad_norm": 0.35740309953689575,
+      "learning_rate": 9.903283367877603e-05,
+      "loss": 1.1621,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06567835226927517,
+      "grad_norm": 0.4617520570755005,
+      "learning_rate": 9.903095904610851e-05,
+      "loss": 1.1541,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0657391093019017,
+      "grad_norm": 0.2352830469608307,
+      "learning_rate": 9.902908261620172e-05,
+      "loss": 1.1506,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06579986633452822,
+      "grad_norm": 0.5313760638237,
+      "learning_rate": 9.902720438912441e-05,
+      "loss": 1.317,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06586062336715474,
+      "grad_norm": 0.332553505897522,
+      "learning_rate": 9.902532436494547e-05,
+      "loss": 1.1067,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06592138039978128,
+      "grad_norm": 0.20040388405323029,
+      "learning_rate": 9.902344254373377e-05,
+      "loss": 1.1594,
+      "step": 1085
+    },
+    {
+      "epoch": 0.0659821374324078,
+      "grad_norm": 0.2770216763019562,
+      "learning_rate": 9.902155892555831e-05,
+      "loss": 1.2575,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06604289446503432,
+      "grad_norm": 0.3603607416152954,
+      "learning_rate": 9.901967351048813e-05,
+      "loss": 1.1843,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06610365149766086,
+      "grad_norm": 0.16854630410671234,
+      "learning_rate": 9.901778629859235e-05,
+      "loss": 1.1679,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06616440853028738,
+      "grad_norm": 0.3734396696090698,
+      "learning_rate": 9.901589728994014e-05,
+      "loss": 1.1973,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06622516556291391,
+      "grad_norm": 0.396348237991333,
+      "learning_rate": 9.901400648460074e-05,
+      "loss": 1.1766,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06628592259554043,
+      "grad_norm": 0.2696155607700348,
+      "learning_rate": 9.901211388264345e-05,
+      "loss": 1.1673,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06634667962816695,
+      "grad_norm": 0.3086203634738922,
+      "learning_rate": 9.901021948413765e-05,
+      "loss": 1.1411,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06640743666079349,
+      "grad_norm": 0.17181465029716492,
+      "learning_rate": 9.90083232891528e-05,
+      "loss": 1.1331,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06646819369342001,
+      "grad_norm": 0.18774396181106567,
+      "learning_rate": 9.900642529775836e-05,
+      "loss": 1.1222,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06652895072604655,
+      "grad_norm": 0.4674195349216461,
+      "learning_rate": 9.900452551002395e-05,
+      "loss": 1.4539,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06658970775867307,
+      "grad_norm": 0.2495420128107071,
+      "learning_rate": 9.900262392601918e-05,
+      "loss": 1.2424,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06665046479129959,
+      "grad_norm": 0.22591935098171234,
+      "learning_rate": 9.900072054581375e-05,
+      "loss": 1.1439,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06671122182392612,
+      "grad_norm": 0.2648542523384094,
+      "learning_rate": 9.899881536947744e-05,
+      "loss": 1.2879,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06677197885655264,
+      "grad_norm": 0.2803839147090912,
+      "learning_rate": 9.899690839708008e-05,
+      "loss": 1.2386,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06683273588917918,
+      "grad_norm": 0.18792013823986053,
+      "learning_rate": 9.899499962869157e-05,
+      "loss": 1.224,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0668934929218057,
+      "grad_norm": 0.211750790476799,
+      "learning_rate": 9.899308906438189e-05,
+      "loss": 1.2128,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06695424995443222,
+      "grad_norm": 0.2044948935508728,
+      "learning_rate": 9.899117670422104e-05,
+      "loss": 1.122,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06701500698705876,
+      "grad_norm": 0.2009739726781845,
+      "learning_rate": 9.898926254827916e-05,
+      "loss": 1.2161,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06707576401968528,
+      "grad_norm": 0.2129359394311905,
+      "learning_rate": 9.898734659662638e-05,
+      "loss": 1.1392,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0671365210523118,
+      "grad_norm": 0.2559523582458496,
+      "learning_rate": 9.898542884933295e-05,
+      "loss": 1.1593,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06719727808493833,
+      "grad_norm": 0.2548045516014099,
+      "learning_rate": 9.898350930646916e-05,
+      "loss": 1.2174,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06725803511756485,
+      "grad_norm": 0.17254139482975006,
+      "learning_rate": 9.898158796810535e-05,
+      "loss": 1.1374,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06731879215019139,
+      "grad_norm": 0.17324994504451752,
+      "learning_rate": 9.897966483431198e-05,
+      "loss": 1.1529,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06737954918281791,
+      "grad_norm": 0.22996747493743896,
+      "learning_rate": 9.897773990515952e-05,
+      "loss": 1.1929,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06744030621544443,
+      "grad_norm": 0.18637020885944366,
+      "learning_rate": 9.897581318071853e-05,
+      "loss": 1.1107,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06750106324807097,
+      "grad_norm": 0.17861661314964294,
+      "learning_rate": 9.897388466105966e-05,
+      "loss": 1.1971,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06756182028069749,
+      "grad_norm": 0.5430263876914978,
+      "learning_rate": 9.897195434625356e-05,
+      "loss": 1.206,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06762257731332402,
+      "grad_norm": 0.16525183618068695,
+      "learning_rate": 9.897002223637103e-05,
+      "loss": 1.1195,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06768333434595054,
+      "grad_norm": 0.17113755643367767,
+      "learning_rate": 9.896808833148285e-05,
+      "loss": 1.1206,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06774409137857706,
+      "grad_norm": 0.20011554658412933,
+      "learning_rate": 9.896615263165994e-05,
+      "loss": 1.1031,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0678048484112036,
+      "grad_norm": 0.2237466424703598,
+      "learning_rate": 9.896421513697323e-05,
+      "loss": 1.1701,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06786560544383012,
+      "grad_norm": 0.7923113107681274,
+      "learning_rate": 9.896227584749375e-05,
+      "loss": 1.3552,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06792636247645666,
+      "grad_norm": 0.16993427276611328,
+      "learning_rate": 9.89603347632926e-05,
+      "loss": 1.1334,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06798711950908318,
+      "grad_norm": 0.4024191200733185,
+      "learning_rate": 9.89583918844409e-05,
+      "loss": 1.1952,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0680478765417097,
+      "grad_norm": 0.5233056545257568,
+      "learning_rate": 9.895644721100989e-05,
+      "loss": 1.2922,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06810863357433623,
+      "grad_norm": 0.24080172181129456,
+      "learning_rate": 9.895450074307084e-05,
+      "loss": 1.0954,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06816939060696275,
+      "grad_norm": 0.2148400843143463,
+      "learning_rate": 9.89525524806951e-05,
+      "loss": 1.1461,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06823014763958929,
+      "grad_norm": 0.21003541350364685,
+      "learning_rate": 9.895060242395408e-05,
+      "loss": 1.1404,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06829090467221581,
+      "grad_norm": 0.27357712388038635,
+      "learning_rate": 9.894865057291928e-05,
+      "loss": 1.1488,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06835166170484233,
+      "grad_norm": 0.19512055814266205,
+      "learning_rate": 9.894669692766223e-05,
+      "loss": 1.2444,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06841241873746887,
+      "grad_norm": 0.2192995697259903,
+      "learning_rate": 9.894474148825454e-05,
+      "loss": 1.1388,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06847317577009539,
+      "grad_norm": 0.6291006207466125,
+      "learning_rate": 9.89427842547679e-05,
+      "loss": 1.2148,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06853393280272191,
+      "grad_norm": 0.22833654284477234,
+      "learning_rate": 9.894082522727403e-05,
+      "loss": 1.1385,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06859468983534844,
+      "grad_norm": 0.389053076505661,
+      "learning_rate": 9.893886440584476e-05,
+      "loss": 1.2045,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06865544686797496,
+      "grad_norm": 0.3422658443450928,
+      "learning_rate": 9.893690179055194e-05,
+      "loss": 1.0926,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0687162039006015,
+      "grad_norm": 0.297583669424057,
+      "learning_rate": 9.893493738146754e-05,
+      "loss": 1.2369,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06877696093322802,
+      "grad_norm": 0.34056973457336426,
+      "learning_rate": 9.893297117866355e-05,
+      "loss": 1.1381,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06883771796585454,
+      "grad_norm": 0.1922212690114975,
+      "learning_rate": 9.893100318221204e-05,
+      "loss": 1.1479,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06889847499848108,
+      "grad_norm": 0.26063457131385803,
+      "learning_rate": 9.892903339218515e-05,
+      "loss": 1.1112,
+      "step": 1134
+    },
+    {
+      "epoch": 0.0689592320311076,
+      "grad_norm": 0.2509213387966156,
+      "learning_rate": 9.892706180865509e-05,
+      "loss": 1.1531,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06901998906373413,
+      "grad_norm": 0.25920236110687256,
+      "learning_rate": 9.89250884316941e-05,
+      "loss": 1.2092,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06908074609636065,
+      "grad_norm": 0.28422781825065613,
+      "learning_rate": 9.892311326137455e-05,
+      "loss": 1.1242,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06914150312898718,
+      "grad_norm": 0.30794256925582886,
+      "learning_rate": 9.892113629776884e-05,
+      "loss": 1.2653,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06920226016161371,
+      "grad_norm": 0.3708270788192749,
+      "learning_rate": 9.891915754094941e-05,
+      "loss": 1.1174,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06926301719424023,
+      "grad_norm": 0.284943550825119,
+      "learning_rate": 9.891717699098882e-05,
+      "loss": 1.3187,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06932377422686677,
+      "grad_norm": 0.15996769070625305,
+      "learning_rate": 9.891519464795961e-05,
+      "loss": 1.1042,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06938453125949329,
+      "grad_norm": 0.24439401924610138,
+      "learning_rate": 9.891321051193452e-05,
+      "loss": 1.1945,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06944528829211981,
+      "grad_norm": 3.1704530715942383,
+      "learning_rate": 9.891122458298626e-05,
+      "loss": 1.2771,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06950604532474634,
+      "grad_norm": 0.47090238332748413,
+      "learning_rate": 9.890923686118758e-05,
+      "loss": 1.0811,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06956680235737286,
+      "grad_norm": 0.194844052195549,
+      "learning_rate": 9.890724734661137e-05,
+      "loss": 1.1444,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06962755938999939,
+      "grad_norm": 0.2976624071598053,
+      "learning_rate": 9.890525603933057e-05,
+      "loss": 1.1794,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06968831642262592,
+      "grad_norm": 0.42262864112854004,
+      "learning_rate": 9.890326293941813e-05,
+      "loss": 1.0882,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06974907345525244,
+      "grad_norm": 0.418404757976532,
+      "learning_rate": 9.890126804694715e-05,
+      "loss": 1.2888,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06980983048787898,
+      "grad_norm": 0.26866504549980164,
+      "learning_rate": 9.889927136199074e-05,
+      "loss": 1.3024,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0698705875205055,
+      "grad_norm": 0.27440404891967773,
+      "learning_rate": 9.889727288462208e-05,
+      "loss": 1.149,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06993134455313202,
+      "grad_norm": 0.18479019403457642,
+      "learning_rate": 9.889527261491444e-05,
+      "loss": 1.1183,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06999210158575855,
+      "grad_norm": 0.22347456216812134,
+      "learning_rate": 9.889327055294112e-05,
+      "loss": 1.3339,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07005285861838507,
+      "grad_norm": 0.21663208305835724,
+      "learning_rate": 9.889126669877553e-05,
+      "loss": 1.1024,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07011361565101161,
+      "grad_norm": 0.1680421233177185,
+      "learning_rate": 9.88892610524911e-05,
+      "loss": 1.1227,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07017437268363813,
+      "grad_norm": 0.42199599742889404,
+      "learning_rate": 9.888725361416136e-05,
+      "loss": 1.2934,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07023512971626465,
+      "grad_norm": 0.18065239489078522,
+      "learning_rate": 9.888524438385988e-05,
+      "loss": 1.1768,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07029588674889119,
+      "grad_norm": 0.2552657127380371,
+      "learning_rate": 9.888323336166033e-05,
+      "loss": 1.2412,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07035664378151771,
+      "grad_norm": 0.23475800454616547,
+      "learning_rate": 9.88812205476364e-05,
+      "loss": 1.0921,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07041740081414424,
+      "grad_norm": 0.18801848590373993,
+      "learning_rate": 9.88792059418619e-05,
+      "loss": 1.1756,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07047815784677076,
+      "grad_norm": 0.2849925756454468,
+      "learning_rate": 9.887718954441064e-05,
+      "loss": 1.2694,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07053891487939729,
+      "grad_norm": 0.2758830189704895,
+      "learning_rate": 9.887517135535656e-05,
+      "loss": 1.0912,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07059967191202382,
+      "grad_norm": 0.1720307469367981,
+      "learning_rate": 9.887315137477362e-05,
+      "loss": 1.1549,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07066042894465034,
+      "grad_norm": 0.2587137222290039,
+      "learning_rate": 9.887112960273587e-05,
+      "loss": 1.1093,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07072118597727688,
+      "grad_norm": 0.17992408573627472,
+      "learning_rate": 9.886910603931743e-05,
+      "loss": 1.0915,
+      "step": 1164
+    },
+    {
+      "epoch": 0.0707819430099034,
+      "grad_norm": 0.5324072241783142,
+      "learning_rate": 9.886708068459244e-05,
+      "loss": 1.1718,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07084270004252992,
+      "grad_norm": 0.19721661508083344,
+      "learning_rate": 9.886505353863518e-05,
+      "loss": 1.1445,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07090345707515645,
+      "grad_norm": 0.3066786527633667,
+      "learning_rate": 9.886302460151992e-05,
+      "loss": 1.2077,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07096421410778297,
+      "grad_norm": 0.4767993688583374,
+      "learning_rate": 9.886099387332107e-05,
+      "loss": 1.4113,
+      "step": 1168
+    },
+    {
+      "epoch": 0.0710249711404095,
+      "grad_norm": 0.35383138060569763,
+      "learning_rate": 9.885896135411304e-05,
+      "loss": 1.1748,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07108572817303603,
+      "grad_norm": 0.23907434940338135,
+      "learning_rate": 9.885692704397032e-05,
+      "loss": 1.209,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07114648520566255,
+      "grad_norm": 0.5726443529129028,
+      "learning_rate": 9.885489094296751e-05,
+      "loss": 1.2388,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07120724223828909,
+      "grad_norm": 0.19819119572639465,
+      "learning_rate": 9.885285305117922e-05,
+      "loss": 1.2159,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07126799927091561,
+      "grad_norm": 0.2896987497806549,
+      "learning_rate": 9.885081336868018e-05,
+      "loss": 1.2073,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07132875630354213,
+      "grad_norm": 0.20259325206279755,
+      "learning_rate": 9.884877189554511e-05,
+      "loss": 1.166,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07138951333616866,
+      "grad_norm": 0.21030500531196594,
+      "learning_rate": 9.884672863184888e-05,
+      "loss": 1.2611,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07145027036879519,
+      "grad_norm": 0.19030259549617767,
+      "learning_rate": 9.884468357766636e-05,
+      "loss": 1.1742,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07151102740142172,
+      "grad_norm": 0.3621237277984619,
+      "learning_rate": 9.884263673307253e-05,
+      "loss": 1.1848,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07157178443404824,
+      "grad_norm": 0.2247137874364853,
+      "learning_rate": 9.884058809814239e-05,
+      "loss": 1.1659,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07163254146667476,
+      "grad_norm": 0.2524459660053253,
+      "learning_rate": 9.883853767295108e-05,
+      "loss": 1.1831,
+      "step": 1179
+    },
+    {
+      "epoch": 0.0716932984993013,
+      "grad_norm": 0.29188045859336853,
+      "learning_rate": 9.883648545757372e-05,
+      "loss": 1.1499,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07175405553192782,
+      "grad_norm": 0.21408967673778534,
+      "learning_rate": 9.883443145208553e-05,
+      "loss": 1.196,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07181481256455435,
+      "grad_norm": 0.5212968587875366,
+      "learning_rate": 9.883237565656182e-05,
+      "loss": 1.1737,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07187556959718087,
+      "grad_norm": 0.24218496680259705,
+      "learning_rate": 9.883031807107797e-05,
+      "loss": 1.2154,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0719363266298074,
+      "grad_norm": 2.173280715942383,
+      "learning_rate": 9.882825869570934e-05,
+      "loss": 1.2281,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07199708366243393,
+      "grad_norm": 0.2711932361125946,
+      "learning_rate": 9.882619753053146e-05,
+      "loss": 1.1877,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07205784069506045,
+      "grad_norm": 0.2696588337421417,
+      "learning_rate": 9.882413457561988e-05,
+      "loss": 1.2172,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07211859772768699,
+      "grad_norm": 0.2311020791530609,
+      "learning_rate": 9.88220698310502e-05,
+      "loss": 1.2176,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07217935476031351,
+      "grad_norm": 0.2556701898574829,
+      "learning_rate": 9.88200032968981e-05,
+      "loss": 1.0911,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07224011179294003,
+      "grad_norm": 0.30819210410118103,
+      "learning_rate": 9.881793497323937e-05,
+      "loss": 1.2522,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07230086882556656,
+      "grad_norm": 0.3643665611743927,
+      "learning_rate": 9.881586486014979e-05,
+      "loss": 1.0996,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07236162585819308,
+      "grad_norm": 0.37965983152389526,
+      "learning_rate": 9.881379295770524e-05,
+      "loss": 1.1983,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0724223828908196,
+      "grad_norm": 0.21066147089004517,
+      "learning_rate": 9.881171926598167e-05,
+      "loss": 1.1675,
+      "step": 1192
+    },
+    {
+      "epoch": 0.07248313992344614,
+      "grad_norm": 0.41477102041244507,
+      "learning_rate": 9.88096437850551e-05,
+      "loss": 1.1229,
+      "step": 1193
+    },
+    {
+      "epoch": 0.07254389695607266,
+      "grad_norm": 0.25262150168418884,
+      "learning_rate": 9.880756651500162e-05,
+      "loss": 1.1047,
+      "step": 1194
+    },
+    {
+      "epoch": 0.0726046539886992,
+      "grad_norm": 0.3302343487739563,
+      "learning_rate": 9.880548745589733e-05,
+      "loss": 1.0916,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07266541102132572,
+      "grad_norm": 0.1993912160396576,
+      "learning_rate": 9.880340660781848e-05,
+      "loss": 1.2562,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07272616805395224,
+      "grad_norm": 0.215809166431427,
+      "learning_rate": 9.880132397084132e-05,
+      "loss": 1.1034,
+      "step": 1197
+    },
+    {
+      "epoch": 0.07278692508657877,
+      "grad_norm": 1.1583436727523804,
+      "learning_rate": 9.87992395450422e-05,
+      "loss": 1.1674,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0728476821192053,
+      "grad_norm": 0.2466304749250412,
+      "learning_rate": 9.87971533304975e-05,
+      "loss": 1.1824,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07290843915183183,
+      "grad_norm": 0.3802541494369507,
+      "learning_rate": 9.879506532728375e-05,
+      "loss": 1.1907,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07296919618445835,
+      "grad_norm": 0.25770261883735657,
+      "learning_rate": 9.879297553547742e-05,
+      "loss": 1.1257,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07302995321708487,
+      "grad_norm": 0.18711112439632416,
+      "learning_rate": 9.879088395515516e-05,
+      "loss": 1.1677,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07309071024971141,
+      "grad_norm": 0.22568635642528534,
+      "learning_rate": 9.87887905863936e-05,
+      "loss": 1.1684,
+      "step": 1203
+    },
+    {
+      "epoch": 0.07315146728233793,
+      "grad_norm": 0.28566041588783264,
+      "learning_rate": 9.878669542926951e-05,
+      "loss": 1.2044,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07321222431496446,
+      "grad_norm": 0.2669213116168976,
+      "learning_rate": 9.878459848385965e-05,
+      "loss": 1.2526,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07327298134759098,
+      "grad_norm": 0.2798711061477661,
+      "learning_rate": 9.87824997502409e-05,
+      "loss": 1.1711,
+      "step": 1206
+    },
+    {
+      "epoch": 0.0733337383802175,
+      "grad_norm": 0.2689701318740845,
+      "learning_rate": 9.87803992284902e-05,
+      "loss": 1.1288,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07339449541284404,
+      "grad_norm": 0.21430884301662445,
+      "learning_rate": 9.877829691868454e-05,
+      "loss": 1.1165,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07345525244547056,
+      "grad_norm": 0.18101859092712402,
+      "learning_rate": 9.877619282090099e-05,
+      "loss": 1.1032,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07351600947809708,
+      "grad_norm": 0.9025177359580994,
+      "learning_rate": 9.877408693521663e-05,
+      "loss": 1.1539,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07357676651072362,
+      "grad_norm": 0.37415724992752075,
+      "learning_rate": 9.877197926170872e-05,
+      "loss": 1.1166,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07363752354335014,
+      "grad_norm": 0.8708227872848511,
+      "learning_rate": 9.876986980045446e-05,
+      "loss": 1.2109,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07369828057597667,
+      "grad_norm": 0.4024532735347748,
+      "learning_rate": 9.876775855153122e-05,
+      "loss": 1.2499,
+      "step": 1213
+    },
+    {
+      "epoch": 0.0737590376086032,
+      "grad_norm": 0.21525508165359497,
+      "learning_rate": 9.876564551501635e-05,
+      "loss": 1.1584,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07381979464122972,
+      "grad_norm": 0.28530144691467285,
+      "learning_rate": 9.87635306909873e-05,
+      "loss": 1.2519,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07388055167385625,
+      "grad_norm": 0.17295902967453003,
+      "learning_rate": 9.876141407952163e-05,
+      "loss": 1.1868,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07394130870648277,
+      "grad_norm": 12.117770195007324,
+      "learning_rate": 9.87592956806969e-05,
+      "loss": 1.2399,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07400206573910931,
+      "grad_norm": 0.23098962008953094,
+      "learning_rate": 9.875717549459075e-05,
+      "loss": 1.132,
+      "step": 1218
+    },
+    {
+      "epoch": 0.07406282277173583,
+      "grad_norm": 0.5400723814964294,
+      "learning_rate": 9.875505352128091e-05,
+      "loss": 1.0978,
+      "step": 1219
+    },
+    {
+      "epoch": 0.07412357980436235,
+      "grad_norm": 0.17768548429012299,
+      "learning_rate": 9.875292976084516e-05,
+      "loss": 1.1184,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07418433683698888,
+      "grad_norm": 0.1565571129322052,
+      "learning_rate": 9.875080421336135e-05,
+      "loss": 1.1045,
+      "step": 1221
+    },
+    {
+      "epoch": 0.0742450938696154,
+      "grad_norm": 0.1982605904340744,
+      "learning_rate": 9.874867687890739e-05,
+      "loss": 1.2224,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07430585090224194,
+      "grad_norm": 0.1820371150970459,
+      "learning_rate": 9.874654775756125e-05,
+      "loss": 1.1732,
+      "step": 1223
+    },
+    {
+      "epoch": 0.07436660793486846,
+      "grad_norm": 0.1819758266210556,
+      "learning_rate": 9.874441684940099e-05,
+      "loss": 1.1459,
+      "step": 1224
+    },
+    {
+      "epoch": 0.07442736496749498,
+      "grad_norm": 0.15922798216342926,
+      "learning_rate": 9.874228415450469e-05,
+      "loss": 1.1319,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07448812200012152,
+      "grad_norm": 0.29691576957702637,
+      "learning_rate": 9.874014967295055e-05,
+      "loss": 1.4323,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07454887903274804,
+      "grad_norm": 0.2321590930223465,
+      "learning_rate": 9.873801340481681e-05,
+      "loss": 1.2083,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07460963606537457,
+      "grad_norm": 0.22122740745544434,
+      "learning_rate": 9.873587535018177e-05,
+      "loss": 1.1489,
+      "step": 1228
+    },
+    {
+      "epoch": 0.0746703930980011,
+      "grad_norm": 0.24191616475582123,
+      "learning_rate": 9.87337355091238e-05,
+      "loss": 1.2007,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07473115013062762,
+      "grad_norm": 0.1998981088399887,
+      "learning_rate": 9.873159388172132e-05,
+      "loss": 1.1998,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07479190716325415,
+      "grad_norm": 0.22300809621810913,
+      "learning_rate": 9.872945046805284e-05,
+      "loss": 1.1431,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07485266419588067,
+      "grad_norm": 0.17642222344875336,
+      "learning_rate": 9.872730526819696e-05,
+      "loss": 1.1344,
+      "step": 1232
+    },
+    {
+      "epoch": 0.0749134212285072,
+      "grad_norm": 0.2079276293516159,
+      "learning_rate": 9.872515828223228e-05,
+      "loss": 1.2673,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07497417826113373,
+      "grad_norm": 0.19985605776309967,
+      "learning_rate": 9.87230095102375e-05,
+      "loss": 1.106,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07503493529376025,
+      "grad_norm": 0.22975324094295502,
+      "learning_rate": 9.872085895229138e-05,
+      "loss": 1.2679,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07509569232638678,
+      "grad_norm": 0.41138115525245667,
+      "learning_rate": 9.871870660847276e-05,
+      "loss": 1.1804,
+      "step": 1236
+    },
+    {
+      "epoch": 0.0751564493590133,
+      "grad_norm": 0.23521828651428223,
+      "learning_rate": 9.871655247886053e-05,
+      "loss": 1.198,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07521720639163983,
+      "grad_norm": 0.1866956353187561,
+      "learning_rate": 9.871439656353367e-05,
+      "loss": 1.1982,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07527796342426636,
+      "grad_norm": 0.27496084570884705,
+      "learning_rate": 9.871223886257117e-05,
+      "loss": 1.2967,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07533872045689288,
+      "grad_norm": 0.3111632764339447,
+      "learning_rate": 9.871007937605214e-05,
+      "loss": 1.101,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07539947748951942,
+      "grad_norm": 0.16673867404460907,
+      "learning_rate": 9.870791810405575e-05,
+      "loss": 1.1584,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07546023452214594,
+      "grad_norm": 0.538335919380188,
+      "learning_rate": 9.870575504666119e-05,
+      "loss": 1.1885,
+      "step": 1242
+    },
+    {
+      "epoch": 0.07552099155477246,
+      "grad_norm": 0.18732623755931854,
+      "learning_rate": 9.870359020394779e-05,
+      "loss": 1.1032,
+      "step": 1243
+    },
+    {
+      "epoch": 0.075581748587399,
+      "grad_norm": 0.15791022777557373,
+      "learning_rate": 9.870142357599485e-05,
+      "loss": 1.1374,
+      "step": 1244
+    },
+    {
+      "epoch": 0.07564250562002552,
+      "grad_norm": 0.19111013412475586,
+      "learning_rate": 9.869925516288182e-05,
+      "loss": 1.1167,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07570326265265205,
+      "grad_norm": 0.22212238609790802,
+      "learning_rate": 9.869708496468819e-05,
+      "loss": 1.1865,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07576401968527857,
+      "grad_norm": 0.20679518580436707,
+      "learning_rate": 9.869491298149348e-05,
+      "loss": 1.1974,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07582477671790509,
+      "grad_norm": 0.40185847878456116,
+      "learning_rate": 9.869273921337734e-05,
+      "loss": 1.2046,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07588553375053163,
+      "grad_norm": 0.2737756371498108,
+      "learning_rate": 9.869056366041942e-05,
+      "loss": 1.1058,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07594629078315815,
+      "grad_norm": 0.24144308269023895,
+      "learning_rate": 9.868838632269947e-05,
+      "loss": 1.165,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07600704781578467,
+      "grad_norm": 0.4306756854057312,
+      "learning_rate": 9.868620720029734e-05,
+      "loss": 1.2316,
+      "step": 1251
+    },
+    {
+      "epoch": 0.0760678048484112,
+      "grad_norm": 0.18132753670215607,
+      "learning_rate": 9.868402629329284e-05,
+      "loss": 1.1462,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07612856188103773,
+      "grad_norm": 0.1788402497768402,
+      "learning_rate": 9.868184360176596e-05,
+      "loss": 1.1362,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07618931891366426,
+      "grad_norm": 0.2483174353837967,
+      "learning_rate": 9.86796591257967e-05,
+      "loss": 1.1175,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07625007594629078,
+      "grad_norm": 0.2149389535188675,
+      "learning_rate": 9.867747286546512e-05,
+      "loss": 1.2059,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0763108329789173,
+      "grad_norm": 0.20216621458530426,
+      "learning_rate": 9.867528482085135e-05,
+      "loss": 1.1647,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07637159001154384,
+      "grad_norm": 0.24809569120407104,
+      "learning_rate": 9.867309499203562e-05,
+      "loss": 1.138,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07643234704417036,
+      "grad_norm": 0.2221045196056366,
+      "learning_rate": 9.867090337909819e-05,
+      "loss": 1.1432,
+      "step": 1258
+    },
+    {
+      "epoch": 0.0764931040767969,
+      "grad_norm": 0.3315385580062866,
+      "learning_rate": 9.866870998211937e-05,
+      "loss": 1.1068,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07655386110942342,
+      "grad_norm": 0.18087199330329895,
+      "learning_rate": 9.866651480117958e-05,
+      "loss": 1.221,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07661461814204994,
+      "grad_norm": 0.14501695334911346,
+      "learning_rate": 9.86643178363593e-05,
+      "loss": 1.1216,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07667537517467647,
+      "grad_norm": 0.19891618192195892,
+      "learning_rate": 9.866211908773903e-05,
+      "loss": 1.212,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07673613220730299,
+      "grad_norm": 0.21249403059482574,
+      "learning_rate": 9.865991855539939e-05,
+      "loss": 1.1338,
+      "step": 1263
+    },
+    {
+      "epoch": 0.07679688923992953,
+      "grad_norm": 0.16505753993988037,
+      "learning_rate": 9.865771623942102e-05,
+      "loss": 1.1192,
+      "step": 1264
+    },
+    {
+      "epoch": 0.07685764627255605,
+      "grad_norm": 0.3122229278087616,
+      "learning_rate": 9.865551213988467e-05,
+      "loss": 1.1952,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07691840330518257,
+      "grad_norm": 0.18771880865097046,
+      "learning_rate": 9.86533062568711e-05,
+      "loss": 1.2009,
+      "step": 1266
+    },
+    {
+      "epoch": 0.0769791603378091,
+      "grad_norm": 0.21376119554042816,
+      "learning_rate": 9.86510985904612e-05,
+      "loss": 1.1342,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07703991737043563,
+      "grad_norm": 0.16033343970775604,
+      "learning_rate": 9.864888914073586e-05,
+      "loss": 1.1222,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07710067440306216,
+      "grad_norm": 1.6344711780548096,
+      "learning_rate": 9.864667790777609e-05,
+      "loss": 1.121,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07716143143568868,
+      "grad_norm": 0.2040167599916458,
+      "learning_rate": 9.864446489166294e-05,
+      "loss": 1.1286,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0772221884683152,
+      "grad_norm": 0.3274246156215668,
+      "learning_rate": 9.864225009247751e-05,
+      "loss": 1.1375,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07728294550094174,
+      "grad_norm": 0.14763733744621277,
+      "learning_rate": 9.864003351030104e-05,
+      "loss": 1.1291,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07734370253356826,
+      "grad_norm": 0.2939620018005371,
+      "learning_rate": 9.863781514521471e-05,
+      "loss": 1.0642,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07740445956619478,
+      "grad_norm": 0.20359748601913452,
+      "learning_rate": 9.863559499729988e-05,
+      "loss": 1.1845,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07746521659882132,
+      "grad_norm": 0.17832833528518677,
+      "learning_rate": 9.86333730666379e-05,
+      "loss": 1.1146,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07752597363144784,
+      "grad_norm": 0.18401280045509338,
+      "learning_rate": 9.863114935331023e-05,
+      "loss": 1.228,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07758673066407437,
+      "grad_norm": 0.46494609117507935,
+      "learning_rate": 9.86289238573984e-05,
+      "loss": 1.255,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07764748769670089,
+      "grad_norm": 0.251230925321579,
+      "learning_rate": 9.862669657898395e-05,
+      "loss": 1.2977,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07770824472932741,
+      "grad_norm": 0.23302753269672394,
+      "learning_rate": 9.862446751814854e-05,
+      "loss": 1.2235,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07776900176195395,
+      "grad_norm": 0.16729407012462616,
+      "learning_rate": 9.862223667497388e-05,
+      "loss": 1.1208,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07782975879458047,
+      "grad_norm": 0.21287155151367188,
+      "learning_rate": 9.862000404954172e-05,
+      "loss": 1.1885,
+      "step": 1281
+    },
+    {
+      "epoch": 0.077890515827207,
+      "grad_norm": 0.37683218717575073,
+      "learning_rate": 9.861776964193392e-05,
+      "loss": 1.3295,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07795127285983353,
+      "grad_norm": 0.23647919297218323,
+      "learning_rate": 9.861553345223238e-05,
+      "loss": 1.0999,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07801202989246005,
+      "grad_norm": 0.24196778237819672,
+      "learning_rate": 9.861329548051908e-05,
+      "loss": 1.1301,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07807278692508658,
+      "grad_norm": 0.22322338819503784,
+      "learning_rate": 9.861105572687602e-05,
+      "loss": 1.1164,
+      "step": 1285
+    },
+    {
+      "epoch": 0.0781335439577131,
+      "grad_norm": 0.17207689583301544,
+      "learning_rate": 9.860881419138531e-05,
+      "loss": 1.1383,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07819430099033964,
+      "grad_norm": 0.19567154347896576,
+      "learning_rate": 9.860657087412912e-05,
+      "loss": 1.1383,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07825505802296616,
+      "grad_norm": 0.20552359521389008,
+      "learning_rate": 9.86043257751897e-05,
+      "loss": 1.1251,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07831581505559268,
+      "grad_norm": 0.16334939002990723,
+      "learning_rate": 9.860207889464929e-05,
+      "loss": 1.1959,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07837657208821922,
+      "grad_norm": 0.196927011013031,
+      "learning_rate": 9.859983023259031e-05,
+      "loss": 1.2013,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07843732912084574,
+      "grad_norm": 0.1954898089170456,
+      "learning_rate": 9.859757978909516e-05,
+      "loss": 1.2278,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07849808615347227,
+      "grad_norm": 0.18258541822433472,
+      "learning_rate": 9.859532756424632e-05,
+      "loss": 1.2141,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07855884318609879,
+      "grad_norm": 9.932856559753418,
+      "learning_rate": 9.859307355812637e-05,
+      "loss": 1.1842,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07861960021872531,
+      "grad_norm": 0.31945687532424927,
+      "learning_rate": 9.859081777081789e-05,
+      "loss": 1.2044,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07868035725135185,
+      "grad_norm": 0.2624160647392273,
+      "learning_rate": 9.858856020240362e-05,
+      "loss": 1.071,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07874111428397837,
+      "grad_norm": 0.257998526096344,
+      "learning_rate": 9.858630085296625e-05,
+      "loss": 1.1729,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07880187131660489,
+      "grad_norm": 0.3425275385379791,
+      "learning_rate": 9.858403972258865e-05,
+      "loss": 1.1558,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07886262834923143,
+      "grad_norm": 0.2115568369626999,
+      "learning_rate": 9.858177681135368e-05,
+      "loss": 1.1376,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07892338538185795,
+      "grad_norm": 0.1707857847213745,
+      "learning_rate": 9.85795121193443e-05,
+      "loss": 1.0875,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07898414241448448,
+      "grad_norm": 0.38808855414390564,
+      "learning_rate": 9.85772456466435e-05,
+      "loss": 1.0925,
+      "step": 1300
+    },
+    {
+      "epoch": 0.079044899447111,
+      "grad_norm": 0.17025871574878693,
+      "learning_rate": 9.85749773933344e-05,
+      "loss": 1.0948,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07910565647973752,
+      "grad_norm": 0.23683875799179077,
+      "learning_rate": 9.857270735950008e-05,
+      "loss": 1.1932,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07916641351236406,
+      "grad_norm": 42.265953063964844,
+      "learning_rate": 9.857043554522379e-05,
+      "loss": 1.1412,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07922717054499058,
+      "grad_norm": 0.22707150876522064,
+      "learning_rate": 9.85681619505888e-05,
+      "loss": 1.128,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07928792757761711,
+      "grad_norm": 0.23970550298690796,
+      "learning_rate": 9.856588657567845e-05,
+      "loss": 1.1108,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07934868461024364,
+      "grad_norm": 0.16838587820529938,
+      "learning_rate": 9.856360942057615e-05,
+      "loss": 1.1313,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07940944164287016,
+      "grad_norm": 0.21601083874702454,
+      "learning_rate": 9.856133048536533e-05,
+      "loss": 1.1223,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07947019867549669,
+      "grad_norm": 0.2647416293621063,
+      "learning_rate": 9.855904977012958e-05,
+      "loss": 1.2414,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07953095570812321,
+      "grad_norm": 0.2626563012599945,
+      "learning_rate": 9.855676727495246e-05,
+      "loss": 1.0582,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07959171274074975,
+      "grad_norm": 0.2018452137708664,
+      "learning_rate": 9.855448299991765e-05,
+      "loss": 1.127,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07965246977337627,
+      "grad_norm": 0.21179260313510895,
+      "learning_rate": 9.855219694510889e-05,
+      "loss": 1.1301,
+      "step": 1311
+    },
+    {
+      "epoch": 0.07971322680600279,
+      "grad_norm": 0.20376284420490265,
+      "learning_rate": 9.854990911060995e-05,
+      "loss": 1.086,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07977398383862933,
+      "grad_norm": 0.1992858648300171,
+      "learning_rate": 9.854761949650473e-05,
+      "loss": 1.1166,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07983474087125585,
+      "grad_norm": 0.4276573956012726,
+      "learning_rate": 9.854532810287711e-05,
+      "loss": 1.2091,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07989549790388237,
+      "grad_norm": 0.9072816371917725,
+      "learning_rate": 9.854303492981111e-05,
+      "loss": 1.1063,
+      "step": 1315
+    },
+    {
+      "epoch": 0.0799562549365089,
+      "grad_norm": 0.29177671670913696,
+      "learning_rate": 9.854073997739079e-05,
+      "loss": 1.1673,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08001701196913542,
+      "grad_norm": 0.2633453607559204,
+      "learning_rate": 9.853844324570027e-05,
+      "loss": 1.1838,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08007776900176196,
+      "grad_norm": 0.349187970161438,
+      "learning_rate": 9.853614473482371e-05,
+      "loss": 1.2875,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08013852603438848,
+      "grad_norm": 0.22905027866363525,
+      "learning_rate": 9.85338444448454e-05,
+      "loss": 1.153,
+      "step": 1319
+    },
+    {
+      "epoch": 0.080199283067015,
+      "grad_norm": 0.17674872279167175,
+      "learning_rate": 9.853154237584965e-05,
+      "loss": 1.1263,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08026004009964154,
+      "grad_norm": 0.16252605617046356,
+      "learning_rate": 9.852923852792081e-05,
+      "loss": 1.0811,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08032079713226806,
+      "grad_norm": 0.2902568578720093,
+      "learning_rate": 9.852693290114335e-05,
+      "loss": 1.2009,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08038155416489459,
+      "grad_norm": 0.20993369817733765,
+      "learning_rate": 9.85246254956018e-05,
+      "loss": 1.1819,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08044231119752111,
+      "grad_norm": 0.36674124002456665,
+      "learning_rate": 9.852231631138071e-05,
+      "loss": 1.0939,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08050306823014763,
+      "grad_norm": 0.2898842394351959,
+      "learning_rate": 9.852000534856474e-05,
+      "loss": 1.1931,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08056382526277417,
+      "grad_norm": 0.36723166704177856,
+      "learning_rate": 9.851769260723862e-05,
+      "loss": 1.1553,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08062458229540069,
+      "grad_norm": 0.2708528935909271,
+      "learning_rate": 9.851537808748706e-05,
+      "loss": 1.139,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08068533932802723,
+      "grad_norm": 0.2334737330675125,
+      "learning_rate": 9.851306178939497e-05,
+      "loss": 1.095,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08074609636065375,
+      "grad_norm": 0.17848482728004456,
+      "learning_rate": 9.85107437130472e-05,
+      "loss": 1.1491,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08080685339328027,
+      "grad_norm": 0.22593189775943756,
+      "learning_rate": 9.850842385852876e-05,
+      "loss": 1.088,
+      "step": 1330
+    },
+    {
+      "epoch": 0.0808676104259068,
+      "grad_norm": 0.21066004037857056,
+      "learning_rate": 9.850610222592465e-05,
+      "loss": 1.3048,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08092836745853332,
+      "grad_norm": 0.24408243596553802,
+      "learning_rate": 9.850377881531999e-05,
+      "loss": 1.1438,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08098912449115986,
+      "grad_norm": 0.18779738247394562,
+      "learning_rate": 9.850145362679994e-05,
+      "loss": 1.1229,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08104988152378638,
+      "grad_norm": 0.2079487442970276,
+      "learning_rate": 9.849912666044975e-05,
+      "loss": 1.1505,
+      "step": 1334
+    },
+    {
+      "epoch": 0.0811106385564129,
+      "grad_norm": 0.4151386022567749,
+      "learning_rate": 9.84967979163547e-05,
+      "loss": 1.4287,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08117139558903944,
+      "grad_norm": 0.14635898172855377,
+      "learning_rate": 9.849446739460012e-05,
+      "loss": 1.0805,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08123215262166596,
+      "grad_norm": 0.17209403216838837,
+      "learning_rate": 9.849213509527148e-05,
+      "loss": 1.0971,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08129290965429248,
+      "grad_norm": 0.17887446284294128,
+      "learning_rate": 9.848980101845426e-05,
+      "loss": 1.1353,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08135366668691901,
+      "grad_norm": 0.30196571350097656,
+      "learning_rate": 9.848746516423399e-05,
+      "loss": 1.1899,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08141442371954553,
+      "grad_norm": 0.2856433391571045,
+      "learning_rate": 9.848512753269635e-05,
+      "loss": 1.2318,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08147518075217207,
+      "grad_norm": 0.3166598677635193,
+      "learning_rate": 9.848278812392696e-05,
+      "loss": 1.1151,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08153593778479859,
+      "grad_norm": 0.165154367685318,
+      "learning_rate": 9.848044693801159e-05,
+      "loss": 1.0753,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08159669481742511,
+      "grad_norm": 0.18315227329730988,
+      "learning_rate": 9.847810397503609e-05,
+      "loss": 1.1308,
+      "step": 1343
+    },
+    {
+      "epoch": 0.08165745185005165,
+      "grad_norm": 0.20149187743663788,
+      "learning_rate": 9.84757592350863e-05,
+      "loss": 1.1014,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08171820888267817,
+      "grad_norm": 0.1828075647354126,
+      "learning_rate": 9.84734127182482e-05,
+      "loss": 1.1288,
+      "step": 1345
+    },
+    {
+      "epoch": 0.0817789659153047,
+      "grad_norm": 0.36319729685783386,
+      "learning_rate": 9.847106442460778e-05,
+      "loss": 1.1111,
+      "step": 1346
+    },
+    {
+      "epoch": 0.08183972294793122,
+      "grad_norm": 0.389381468296051,
+      "learning_rate": 9.846871435425112e-05,
+      "loss": 1.2138,
+      "step": 1347
+    },
+    {
+      "epoch": 0.08190047998055774,
+      "grad_norm": 0.2722962498664856,
+      "learning_rate": 9.846636250726438e-05,
+      "loss": 1.1577,
+      "step": 1348
+    },
+    {
+      "epoch": 0.08196123701318428,
+      "grad_norm": 0.16517791152000427,
+      "learning_rate": 9.846400888373374e-05,
+      "loss": 1.1192,
+      "step": 1349
+    },
+    {
+      "epoch": 0.0820219940458108,
+      "grad_norm": 0.2437630146741867,
+      "learning_rate": 9.84616534837455e-05,
+      "loss": 1.1106,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08208275107843734,
+      "grad_norm": 0.21548691391944885,
+      "learning_rate": 9.845929630738598e-05,
+      "loss": 1.1218,
+      "step": 1351
+    },
+    {
+      "epoch": 0.08214350811106386,
+      "grad_norm": 0.374189168214798,
+      "learning_rate": 9.845693735474158e-05,
+      "loss": 1.1204,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08220426514369038,
+      "grad_norm": 0.201692134141922,
+      "learning_rate": 9.845457662589877e-05,
+      "loss": 1.2067,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08226502217631691,
+      "grad_norm": 0.23874618113040924,
+      "learning_rate": 9.845221412094409e-05,
+      "loss": 1.1763,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08232577920894343,
+      "grad_norm": 0.15880019962787628,
+      "learning_rate": 9.844984983996415e-05,
+      "loss": 1.1304,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08238653624156995,
+      "grad_norm": 0.17314967513084412,
+      "learning_rate": 9.84474837830456e-05,
+      "loss": 1.1418,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08244729327419649,
+      "grad_norm": 0.21444889903068542,
+      "learning_rate": 9.844511595027515e-05,
+      "loss": 1.1605,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08250805030682301,
+      "grad_norm": 0.17126399278640747,
+      "learning_rate": 9.844274634173962e-05,
+      "loss": 1.1315,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08256880733944955,
+      "grad_norm": 0.7597689032554626,
+      "learning_rate": 9.844037495752584e-05,
+      "loss": 1.1185,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08262956437207607,
+      "grad_norm": 0.17652194201946259,
+      "learning_rate": 9.843800179772078e-05,
+      "loss": 1.1937,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08269032140470259,
+      "grad_norm": 0.22796188294887543,
+      "learning_rate": 9.843562686241139e-05,
+      "loss": 1.2028,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08275107843732912,
+      "grad_norm": 0.20764288306236267,
+      "learning_rate": 9.843325015168474e-05,
+      "loss": 1.2556,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08281183546995564,
+      "grad_norm": 0.25892022252082825,
+      "learning_rate": 9.843087166562795e-05,
+      "loss": 1.3303,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08287259250258218,
+      "grad_norm": 0.19510486721992493,
+      "learning_rate": 9.842849140432817e-05,
+      "loss": 1.2397,
+      "step": 1364
+    },
+    {
+      "epoch": 0.0829333495352087,
+      "grad_norm": 0.23131868243217468,
+      "learning_rate": 9.84261093678727e-05,
+      "loss": 1.2398,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08299410656783522,
+      "grad_norm": 0.3368610739707947,
+      "learning_rate": 9.842372555634882e-05,
+      "loss": 1.1568,
+      "step": 1366
+    },
+    {
+      "epoch": 0.08305486360046176,
+      "grad_norm": 0.16682179272174835,
+      "learning_rate": 9.842133996984393e-05,
+      "loss": 1.0998,
+      "step": 1367
+    },
+    {
+      "epoch": 0.08311562063308828,
+      "grad_norm": 0.2443765252828598,
+      "learning_rate": 9.841895260844546e-05,
+      "loss": 1.1174,
+      "step": 1368
+    },
+    {
+      "epoch": 0.08317637766571481,
+      "grad_norm": 0.5775284171104431,
+      "learning_rate": 9.841656347224091e-05,
+      "loss": 1.323,
+      "step": 1369
+    },
+    {
+      "epoch": 0.08323713469834133,
+      "grad_norm": 1.681067705154419,
+      "learning_rate": 9.841417256131788e-05,
+      "loss": 1.107,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08329789173096785,
+      "grad_norm": 0.3213003873825073,
+      "learning_rate": 9.8411779875764e-05,
+      "loss": 1.1204,
+      "step": 1371
+    },
+    {
+      "epoch": 0.08335864876359439,
+      "grad_norm": 0.2900693714618683,
+      "learning_rate": 9.840938541566698e-05,
+      "loss": 1.2616,
+      "step": 1372
+    },
+    {
+      "epoch": 0.08341940579622091,
+      "grad_norm": 0.767281711101532,
+      "learning_rate": 9.840698918111455e-05,
+      "loss": 1.1139,
+      "step": 1373
+    },
+    {
+      "epoch": 0.08348016282884745,
+      "grad_norm": 0.2874467372894287,
+      "learning_rate": 9.84045911721946e-05,
+      "loss": 1.1888,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08354091986147397,
+      "grad_norm": 2.1407675743103027,
+      "learning_rate": 9.8402191388995e-05,
+      "loss": 1.2438,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08360167689410049,
+      "grad_norm": 0.2052859365940094,
+      "learning_rate": 9.83997898316037e-05,
+      "loss": 1.1505,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08366243392672702,
+      "grad_norm": 0.25272366404533386,
+      "learning_rate": 9.839738650010877e-05,
+      "loss": 1.0833,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08372319095935354,
+      "grad_norm": 0.25053277611732483,
+      "learning_rate": 9.839498139459827e-05,
+      "loss": 1.1944,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08378394799198007,
+      "grad_norm": 0.2881149649620056,
+      "learning_rate": 9.839257451516037e-05,
+      "loss": 1.3879,
+      "step": 1379
+    },
+    {
+      "epoch": 0.0838447050246066,
+      "grad_norm": 0.262215256690979,
+      "learning_rate": 9.839016586188331e-05,
+      "loss": 1.1336,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08390546205723312,
+      "grad_norm": 0.20073488354682922,
+      "learning_rate": 9.838775543485537e-05,
+      "loss": 1.1795,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08396621908985966,
+      "grad_norm": 0.2712678611278534,
+      "learning_rate": 9.838534323416489e-05,
+      "loss": 1.1929,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08402697612248618,
+      "grad_norm": 0.18539045751094818,
+      "learning_rate": 9.838292925990029e-05,
+      "loss": 1.1202,
+      "step": 1383
+    },
+    {
+      "epoch": 0.0840877331551127,
+      "grad_norm": 0.22492966055870056,
+      "learning_rate": 9.83805135121501e-05,
+      "loss": 1.1616,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08414849018773923,
+      "grad_norm": 0.26752761006355286,
+      "learning_rate": 9.837809599100281e-05,
+      "loss": 1.2089,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08420924722036575,
+      "grad_norm": 0.16586409509181976,
+      "learning_rate": 9.837567669654706e-05,
+      "loss": 1.0907,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08427000425299229,
+      "grad_norm": 0.34646153450012207,
+      "learning_rate": 9.837325562887155e-05,
+      "loss": 1.1581,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08433076128561881,
+      "grad_norm": 0.16896191239356995,
+      "learning_rate": 9.837083278806497e-05,
+      "loss": 1.095,
+      "step": 1388
+    },
+    {
+      "epoch": 0.08439151831824533,
+      "grad_norm": 111.09854125976562,
+      "learning_rate": 9.836840817421619e-05,
+      "loss": 1.1189,
+      "step": 1389
+    },
+    {
+      "epoch": 0.08445227535087187,
+      "grad_norm": 0.6128570437431335,
+      "learning_rate": 9.836598178741405e-05,
+      "loss": 1.1847,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08451303238349839,
+      "grad_norm": 0.25501441955566406,
+      "learning_rate": 9.83635536277475e-05,
+      "loss": 1.0972,
+      "step": 1391
+    },
+    {
+      "epoch": 0.08457378941612492,
+      "grad_norm": 0.17326213419437408,
+      "learning_rate": 9.836112369530555e-05,
+      "loss": 1.1662,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08463454644875144,
+      "grad_norm": 0.6709007024765015,
+      "learning_rate": 9.835869199017725e-05,
+      "loss": 1.0639,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08469530348137796,
+      "grad_norm": 0.4226175844669342,
+      "learning_rate": 9.835625851245177e-05,
+      "loss": 1.1667,
+      "step": 1394
+    },
+    {
+      "epoch": 0.0847560605140045,
+      "grad_norm": 0.5547512173652649,
+      "learning_rate": 9.835382326221826e-05,
+      "loss": 1.197,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08481681754663102,
+      "grad_norm": 0.27206289768218994,
+      "learning_rate": 9.835138623956603e-05,
+      "loss": 1.1978,
+      "step": 1396
+    },
+    {
+      "epoch": 0.08487757457925756,
+      "grad_norm": 0.2442013919353485,
+      "learning_rate": 9.834894744458438e-05,
+      "loss": 1.2557,
+      "step": 1397
+    },
+    {
+      "epoch": 0.08493833161188408,
+      "grad_norm": 0.30892157554626465,
+      "learning_rate": 9.834650687736272e-05,
+      "loss": 1.1426,
+      "step": 1398
+    },
+    {
+      "epoch": 0.0849990886445106,
+      "grad_norm": 0.1691623330116272,
+      "learning_rate": 9.83440645379905e-05,
+      "loss": 1.1941,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08505984567713713,
+      "grad_norm": 0.17318879067897797,
+      "learning_rate": 9.834162042655726e-05,
+      "loss": 1.1134,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08512060270976365,
+      "grad_norm": 0.21950559318065643,
+      "learning_rate": 9.833917454315258e-05,
+      "loss": 1.1139,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08518135974239018,
+      "grad_norm": 0.1323520839214325,
+      "learning_rate": 9.83367268878661e-05,
+      "loss": 1.0788,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08524211677501671,
+      "grad_norm": 0.15223145484924316,
+      "learning_rate": 9.833427746078756e-05,
+      "loss": 1.0954,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08530287380764323,
+      "grad_norm": 0.15274764597415924,
+      "learning_rate": 9.833182626200675e-05,
+      "loss": 1.117,
+      "step": 1404
+    },
+    {
+      "epoch": 0.08536363084026977,
+      "grad_norm": 0.1928999423980713,
+      "learning_rate": 9.83293732916135e-05,
+      "loss": 1.0919,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08542438787289629,
+      "grad_norm": 0.2198067456483841,
+      "learning_rate": 9.832691854969771e-05,
+      "loss": 1.1946,
+      "step": 1406
+    },
+    {
+      "epoch": 0.08548514490552281,
+      "grad_norm": 0.1498374044895172,
+      "learning_rate": 9.83244620363494e-05,
+      "loss": 1.1198,
+      "step": 1407
+    },
+    {
+      "epoch": 0.08554590193814934,
+      "grad_norm": 0.2062024027109146,
+      "learning_rate": 9.832200375165858e-05,
+      "loss": 1.2049,
+      "step": 1408
+    },
+    {
+      "epoch": 0.08560665897077586,
+      "grad_norm": 0.35424208641052246,
+      "learning_rate": 9.831954369571538e-05,
+      "loss": 1.1415,
+      "step": 1409
+    },
+    {
+      "epoch": 0.0856674160034024,
+      "grad_norm": 0.19721995294094086,
+      "learning_rate": 9.831708186860996e-05,
+      "loss": 1.1542,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08572817303602892,
+      "grad_norm": 0.31192660331726074,
+      "learning_rate": 9.831461827043257e-05,
+      "loss": 1.1151,
+      "step": 1411
+    },
+    {
+      "epoch": 0.08578893006865544,
+      "grad_norm": 0.228634312748909,
+      "learning_rate": 9.831215290127351e-05,
+      "loss": 1.2332,
+      "step": 1412
+    },
+    {
+      "epoch": 0.08584968710128198,
+      "grad_norm": 0.25092756748199463,
+      "learning_rate": 9.830968576122315e-05,
+      "loss": 1.385,
+      "step": 1413
+    },
+    {
+      "epoch": 0.0859104441339085,
+      "grad_norm": 0.16668741405010223,
+      "learning_rate": 9.830721685037191e-05,
+      "loss": 1.1658,
+      "step": 1414
+    },
+    {
+      "epoch": 0.08597120116653503,
+      "grad_norm": 0.1422152817249298,
+      "learning_rate": 9.830474616881029e-05,
+      "loss": 1.0965,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08603195819916155,
+      "grad_norm": 0.20908355712890625,
+      "learning_rate": 9.830227371662888e-05,
+      "loss": 1.057,
+      "step": 1416
+    },
+    {
+      "epoch": 0.08609271523178808,
+      "grad_norm": 0.262549489736557,
+      "learning_rate": 9.829979949391828e-05,
+      "loss": 1.1591,
+      "step": 1417
+    },
+    {
+      "epoch": 0.08615347226441461,
+      "grad_norm": 0.964087188243866,
+      "learning_rate": 9.82973235007692e-05,
+      "loss": 1.3327,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08621422929704113,
+      "grad_norm": 0.34942829608917236,
+      "learning_rate": 9.82948457372724e-05,
+      "loss": 1.1333,
+      "step": 1419
+    },
+    {
+      "epoch": 0.08627498632966765,
+      "grad_norm": 0.1866210699081421,
+      "learning_rate": 9.829236620351868e-05,
+      "loss": 1.1201,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08633574336229419,
+      "grad_norm": 0.4157951772212982,
+      "learning_rate": 9.828988489959894e-05,
+      "loss": 1.101,
+      "step": 1421
+    },
+    {
+      "epoch": 0.08639650039492071,
+      "grad_norm": 0.16485145688056946,
+      "learning_rate": 9.828740182560415e-05,
+      "loss": 1.1931,
+      "step": 1422
+    },
+    {
+      "epoch": 0.08645725742754724,
+      "grad_norm": 0.4265294075012207,
+      "learning_rate": 9.828491698162531e-05,
+      "loss": 1.1155,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08651801446017376,
+      "grad_norm": 0.20402522385120392,
+      "learning_rate": 9.828243036775348e-05,
+      "loss": 1.1142,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08657877149280029,
+      "grad_norm": 0.2030715048313141,
+      "learning_rate": 9.827994198407986e-05,
+      "loss": 1.1302,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08663952852542682,
+      "grad_norm": 0.2869037687778473,
+      "learning_rate": 9.827745183069563e-05,
+      "loss": 1.1581,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08670028555805334,
+      "grad_norm": 0.2697184383869171,
+      "learning_rate": 9.827495990769207e-05,
+      "loss": 1.2221,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08676104259067988,
+      "grad_norm": 0.22515903413295746,
+      "learning_rate": 9.827246621516051e-05,
+      "loss": 1.1929,
+      "step": 1428
+    },
+    {
+      "epoch": 0.0868217996233064,
+      "grad_norm": 0.20299787819385529,
+      "learning_rate": 9.826997075319238e-05,
+      "loss": 1.1041,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08688255665593292,
+      "grad_norm": 0.542178750038147,
+      "learning_rate": 9.826747352187915e-05,
+      "loss": 1.1418,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08694331368855945,
+      "grad_norm": 1.657292366027832,
+      "learning_rate": 9.826497452131234e-05,
+      "loss": 1.1563,
+      "step": 1431
+    },
+    {
+      "epoch": 0.08700407072118597,
+      "grad_norm": 0.28403112292289734,
+      "learning_rate": 9.826247375158355e-05,
+      "loss": 1.1742,
+      "step": 1432
+    },
+    {
+      "epoch": 0.08706482775381251,
+      "grad_norm": 0.20294104516506195,
+      "learning_rate": 9.825997121278448e-05,
+      "loss": 1.1269,
+      "step": 1433
+    },
+    {
+      "epoch": 0.08712558478643903,
+      "grad_norm": 0.3333057165145874,
+      "learning_rate": 9.825746690500681e-05,
+      "loss": 1.0829,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08718634181906555,
+      "grad_norm": 0.258637398481369,
+      "learning_rate": 9.825496082834238e-05,
+      "loss": 1.1133,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08724709885169209,
+      "grad_norm": 0.18659645318984985,
+      "learning_rate": 9.825245298288301e-05,
+      "loss": 1.142,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08730785588431861,
+      "grad_norm": 0.7480716705322266,
+      "learning_rate": 9.824994336872068e-05,
+      "loss": 1.1573,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08736861291694514,
+      "grad_norm": 0.17065827548503876,
+      "learning_rate": 9.824743198594734e-05,
+      "loss": 1.1086,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08742936994957166,
+      "grad_norm": 0.2011638879776001,
+      "learning_rate": 9.824491883465504e-05,
+      "loss": 1.2348,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08749012698219819,
+      "grad_norm": 0.29686489701271057,
+      "learning_rate": 9.824240391493593e-05,
+      "loss": 1.1317,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08755088401482472,
+      "grad_norm": 0.24617424607276917,
+      "learning_rate": 9.823988722688218e-05,
+      "loss": 1.182,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08761164104745124,
+      "grad_norm": 0.3014746904373169,
+      "learning_rate": 9.823736877058602e-05,
+      "loss": 1.26,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08767239808007776,
+      "grad_norm": 0.2874422073364258,
+      "learning_rate": 9.82348485461398e-05,
+      "loss": 1.0887,
+      "step": 1443
+    },
+    {
+      "epoch": 0.0877331551127043,
+      "grad_norm": 0.2400795817375183,
+      "learning_rate": 9.823232655363588e-05,
+      "loss": 1.266,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08779391214533082,
+      "grad_norm": 0.18708005547523499,
+      "learning_rate": 9.822980279316671e-05,
+      "loss": 1.2317,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08785466917795735,
+      "grad_norm": 0.26659032702445984,
+      "learning_rate": 9.82272772648248e-05,
+      "loss": 1.2477,
+      "step": 1446
+    },
+    {
+      "epoch": 0.08791542621058387,
+      "grad_norm": 0.18628554046154022,
+      "learning_rate": 9.82247499687027e-05,
+      "loss": 1.1306,
+      "step": 1447
+    },
+    {
+      "epoch": 0.0879761832432104,
+      "grad_norm": 0.2454700917005539,
+      "learning_rate": 9.822222090489307e-05,
+      "loss": 1.2603,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08803694027583693,
+      "grad_norm": 0.18475788831710815,
+      "learning_rate": 9.821969007348863e-05,
+      "loss": 1.1994,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08809769730846345,
+      "grad_norm": 3.490493059158325,
+      "learning_rate": 9.821715747458213e-05,
+      "loss": 1.1901,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08815845434108999,
+      "grad_norm": 0.36890801787376404,
+      "learning_rate": 9.821462310826639e-05,
+      "loss": 1.2076,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08821921137371651,
+      "grad_norm": 0.1761927753686905,
+      "learning_rate": 9.821208697463434e-05,
+      "loss": 1.2016,
+      "step": 1452
+    },
+    {
+      "epoch": 0.08827996840634303,
+      "grad_norm": 0.17505629360675812,
+      "learning_rate": 9.820954907377891e-05,
+      "loss": 1.1376,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08834072543896956,
+      "grad_norm": 0.5224376916885376,
+      "learning_rate": 9.820700940579312e-05,
+      "loss": 1.1075,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08840148247159609,
+      "grad_norm": 0.2143557369709015,
+      "learning_rate": 9.82044679707701e-05,
+      "loss": 1.1989,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08846223950422262,
+      "grad_norm": 0.31814083456993103,
+      "learning_rate": 9.820192476880299e-05,
+      "loss": 1.2507,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08852299653684914,
+      "grad_norm": 0.27409324049949646,
+      "learning_rate": 9.819937979998501e-05,
+      "loss": 1.1286,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08858375356947566,
+      "grad_norm": 0.16574041545391083,
+      "learning_rate": 9.819683306440945e-05,
+      "loss": 1.0965,
+      "step": 1458
+    },
+    {
+      "epoch": 0.0886445106021022,
+      "grad_norm": 0.2751779556274414,
+      "learning_rate": 9.819428456216966e-05,
+      "loss": 1.2349,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08870526763472872,
+      "grad_norm": 0.2155376523733139,
+      "learning_rate": 9.819173429335904e-05,
+      "loss": 1.1434,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08876602466735524,
+      "grad_norm": 0.18793049454689026,
+      "learning_rate": 9.81891822580711e-05,
+      "loss": 1.1261,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08882678169998177,
+      "grad_norm": 0.7266697883605957,
+      "learning_rate": 9.818662845639936e-05,
+      "loss": 1.1947,
+      "step": 1462
+    },
+    {
+      "epoch": 0.0888875387326083,
+      "grad_norm": 0.2719860374927521,
+      "learning_rate": 9.818407288843743e-05,
+      "loss": 1.3699,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08894829576523483,
+      "grad_norm": 0.32996392250061035,
+      "learning_rate": 9.818151555427901e-05,
+      "loss": 1.1435,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08900905279786135,
+      "grad_norm": 0.4552542269229889,
+      "learning_rate": 9.817895645401783e-05,
+      "loss": 1.2124,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08906980983048787,
+      "grad_norm": 0.16603194177150726,
+      "learning_rate": 9.817639558774768e-05,
+      "loss": 1.1151,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08913056686311441,
+      "grad_norm": 0.25212615728378296,
+      "learning_rate": 9.817383295556242e-05,
+      "loss": 1.2246,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08919132389574093,
+      "grad_norm": 0.15637518465518951,
+      "learning_rate": 9.817126855755601e-05,
+      "loss": 1.1218,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08925208092836746,
+      "grad_norm": 0.17580582201480865,
+      "learning_rate": 9.816870239382245e-05,
+      "loss": 1.1458,
+      "step": 1469
+    },
+    {
+      "epoch": 0.08931283796099398,
+      "grad_norm": 0.3365723490715027,
+      "learning_rate": 9.81661344644558e-05,
+      "loss": 1.1947,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0893735949936205,
+      "grad_norm": 0.26322677731513977,
+      "learning_rate": 9.816356476955015e-05,
+      "loss": 1.158,
+      "step": 1471
+    },
+    {
+      "epoch": 0.08943435202624704,
+      "grad_norm": 0.14961378276348114,
+      "learning_rate": 9.816099330919975e-05,
+      "loss": 1.1121,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08949510905887356,
+      "grad_norm": 0.2173193097114563,
+      "learning_rate": 9.815842008349882e-05,
+      "loss": 1.1972,
+      "step": 1473
+    },
+    {
+      "epoch": 0.0895558660915001,
+      "grad_norm": 0.18777205049991608,
+      "learning_rate": 9.815584509254169e-05,
+      "loss": 1.1275,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08961662312412662,
+      "grad_norm": 0.23447135090827942,
+      "learning_rate": 9.815326833642278e-05,
+      "loss": 1.0941,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08967738015675314,
+      "grad_norm": 0.26662781834602356,
+      "learning_rate": 9.815068981523647e-05,
+      "loss": 1.2104,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08973813718937967,
+      "grad_norm": 2.838395357131958,
+      "learning_rate": 9.814810952907735e-05,
+      "loss": 1.2111,
+      "step": 1477
+    },
+    {
+      "epoch": 0.0897988942220062,
+      "grad_norm": 0.17212064564228058,
+      "learning_rate": 9.814552747803996e-05,
+      "loss": 1.1448,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08985965125463273,
+      "grad_norm": 0.38833194971084595,
+      "learning_rate": 9.814294366221895e-05,
+      "loss": 1.182,
+      "step": 1479
+    },
+    {
+      "epoch": 0.08992040828725925,
+      "grad_norm": 0.44230154156684875,
+      "learning_rate": 9.814035808170902e-05,
+      "loss": 1.1896,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08998116531988577,
+      "grad_norm": 0.24242104589939117,
+      "learning_rate": 9.813777073660498e-05,
+      "loss": 1.1827,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09004192235251231,
+      "grad_norm": 0.1917129009962082,
+      "learning_rate": 9.813518162700164e-05,
+      "loss": 1.154,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09010267938513883,
+      "grad_norm": 0.34480440616607666,
+      "learning_rate": 9.81325907529939e-05,
+      "loss": 1.1323,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09016343641776535,
+      "grad_norm": 0.2701270580291748,
+      "learning_rate": 9.812999811467676e-05,
+      "loss": 1.1819,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09022419345039188,
+      "grad_norm": 0.31518250703811646,
+      "learning_rate": 9.812740371214523e-05,
+      "loss": 1.0967,
+      "step": 1485
+    },
+    {
+      "epoch": 0.0902849504830184,
+      "grad_norm": 0.19874174892902374,
+      "learning_rate": 9.812480754549442e-05,
+      "loss": 1.1014,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09034570751564494,
+      "grad_norm": 0.21445833146572113,
+      "learning_rate": 9.812220961481949e-05,
+      "loss": 1.172,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09040646454827146,
+      "grad_norm": 0.21220004558563232,
+      "learning_rate": 9.811960992021565e-05,
+      "loss": 1.1239,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09046722158089798,
+      "grad_norm": 0.23813070356845856,
+      "learning_rate": 9.811700846177823e-05,
+      "loss": 1.2943,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09052797861352452,
+      "grad_norm": 0.2615263760089874,
+      "learning_rate": 9.811440523960254e-05,
+      "loss": 1.19,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09058873564615104,
+      "grad_norm": 0.36815667152404785,
+      "learning_rate": 9.811180025378404e-05,
+      "loss": 1.1033,
+      "step": 1491
+    },
+    {
+      "epoch": 0.09064949267877757,
+      "grad_norm": 0.3342699706554413,
+      "learning_rate": 9.810919350441818e-05,
+      "loss": 1.1374,
+      "step": 1492
+    },
+    {
+      "epoch": 0.0907102497114041,
+      "grad_norm": 0.37105441093444824,
+      "learning_rate": 9.810658499160056e-05,
+      "loss": 1.2355,
+      "step": 1493
+    },
+    {
+      "epoch": 0.09077100674403062,
+      "grad_norm": 0.20406359434127808,
+      "learning_rate": 9.810397471542676e-05,
+      "loss": 1.2127,
+      "step": 1494
+    },
+    {
+      "epoch": 0.09083176377665715,
+      "grad_norm": 0.21318334341049194,
+      "learning_rate": 9.810136267599246e-05,
+      "loss": 1.125,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09089252080928367,
+      "grad_norm": 0.2612029016017914,
+      "learning_rate": 9.809874887339343e-05,
+      "loss": 1.1731,
+      "step": 1496
+    },
+    {
+      "epoch": 0.09095327784191021,
+      "grad_norm": 0.23621846735477448,
+      "learning_rate": 9.809613330772544e-05,
+      "loss": 1.1748,
+      "step": 1497
+    },
+    {
+      "epoch": 0.09101403487453673,
+      "grad_norm": 0.2109217643737793,
+      "learning_rate": 9.809351597908441e-05,
+      "loss": 1.1202,
+      "step": 1498
+    },
+    {
+      "epoch": 0.09107479190716325,
+      "grad_norm": 0.24618826806545258,
+      "learning_rate": 9.809089688756623e-05,
+      "loss": 1.1282,
+      "step": 1499
+    },
+    {
+      "epoch": 0.09113554893978978,
+      "grad_norm": 0.15508413314819336,
+      "learning_rate": 9.808827603326695e-05,
+      "loss": 1.1079,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0911963059724163,
+      "grad_norm": 0.24064220488071442,
+      "learning_rate": 9.808565341628261e-05,
+      "loss": 1.1677,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09125706300504284,
+      "grad_norm": 0.16915874183177948,
+      "learning_rate": 9.808302903670935e-05,
+      "loss": 1.1159,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09131782003766936,
+      "grad_norm": 0.2899004817008972,
+      "learning_rate": 9.808040289464336e-05,
+      "loss": 1.0971,
+      "step": 1503
+    },
+    {
+      "epoch": 0.09137857707029588,
+      "grad_norm": 0.13940022885799408,
+      "learning_rate": 9.807777499018092e-05,
+      "loss": 1.1236,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09143933410292242,
+      "grad_norm": 0.16495467722415924,
+      "learning_rate": 9.807514532341832e-05,
+      "loss": 1.085,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09150009113554894,
+      "grad_norm": 0.19143767654895782,
+      "learning_rate": 9.8072513894452e-05,
+      "loss": 1.1395,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09156084816817546,
+      "grad_norm": 0.2714698612689972,
+      "learning_rate": 9.806988070337839e-05,
+      "loss": 1.2117,
+      "step": 1507
+    },
+    {
+      "epoch": 0.091621605200802,
+      "grad_norm": 0.18536613881587982,
+      "learning_rate": 9.8067245750294e-05,
+      "loss": 1.1944,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09168236223342852,
+      "grad_norm": 0.3692673444747925,
+      "learning_rate": 9.806460903529544e-05,
+      "loss": 1.1712,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 0.1665329784154892,
+      "learning_rate": 9.806197055847934e-05,
+      "loss": 1.1485,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09180387629868157,
+      "grad_norm": 0.24269534647464752,
+      "learning_rate": 9.805933031994241e-05,
+      "loss": 1.2609,
+      "step": 1511
+    },
+    {
+      "epoch": 0.0918646333313081,
+      "grad_norm": 0.19015014171600342,
+      "learning_rate": 9.805668831978145e-05,
+      "loss": 1.2201,
+      "step": 1512
+    },
+    {
+      "epoch": 0.09192539036393463,
+      "grad_norm": 2.2654175758361816,
+      "learning_rate": 9.805404455809328e-05,
+      "loss": 1.1309,
+      "step": 1513
+    },
+    {
+      "epoch": 0.09198614739656115,
+      "grad_norm": 0.24054457247257233,
+      "learning_rate": 9.805139903497482e-05,
+      "loss": 1.3016,
+      "step": 1514
+    },
+    {
+      "epoch": 0.09204690442918768,
+      "grad_norm": 0.1855025440454483,
+      "learning_rate": 9.804875175052304e-05,
+      "loss": 1.11,
+      "step": 1515
+    },
+    {
+      "epoch": 0.0921076614618142,
+      "grad_norm": 0.15947149693965912,
+      "learning_rate": 9.804610270483498e-05,
+      "loss": 1.075,
+      "step": 1516
+    },
+    {
+      "epoch": 0.09216841849444073,
+      "grad_norm": 0.27962183952331543,
+      "learning_rate": 9.804345189800774e-05,
+      "loss": 1.29,
+      "step": 1517
+    },
+    {
+      "epoch": 0.09222917552706726,
+      "grad_norm": 0.1883019655942917,
+      "learning_rate": 9.804079933013847e-05,
+      "loss": 1.1871,
+      "step": 1518
+    },
+    {
+      "epoch": 0.09228993255969378,
+      "grad_norm": 0.43887776136398315,
+      "learning_rate": 9.803814500132445e-05,
+      "loss": 1.1545,
+      "step": 1519
+    },
+    {
+      "epoch": 0.09235068959232032,
+      "grad_norm": 0.21325574815273285,
+      "learning_rate": 9.80354889116629e-05,
+      "loss": 1.0737,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09241144662494684,
+      "grad_norm": 0.2204999029636383,
+      "learning_rate": 9.803283106125123e-05,
+      "loss": 1.166,
+      "step": 1521
+    },
+    {
+      "epoch": 0.09247220365757336,
+      "grad_norm": 0.5710244178771973,
+      "learning_rate": 9.803017145018687e-05,
+      "loss": 1.1305,
+      "step": 1522
+    },
+    {
+      "epoch": 0.0925329606901999,
+      "grad_norm": 0.5452705025672913,
+      "learning_rate": 9.802751007856728e-05,
+      "loss": 1.2947,
+      "step": 1523
+    },
+    {
+      "epoch": 0.09259371772282642,
+      "grad_norm": 0.28894880414009094,
+      "learning_rate": 9.802484694649001e-05,
+      "loss": 1.1546,
+      "step": 1524
+    },
+    {
+      "epoch": 0.09265447475545294,
+      "grad_norm": 0.25489091873168945,
+      "learning_rate": 9.802218205405272e-05,
+      "loss": 1.1156,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09271523178807947,
+      "grad_norm": 0.2926221787929535,
+      "learning_rate": 9.801951540135304e-05,
+      "loss": 1.1505,
+      "step": 1526
+    },
+    {
+      "epoch": 0.09277598882070599,
+      "grad_norm": 0.2533493936061859,
+      "learning_rate": 9.801684698848874e-05,
+      "loss": 1.1537,
+      "step": 1527
+    },
+    {
+      "epoch": 0.09283674585333253,
+      "grad_norm": 0.2091417759656906,
+      "learning_rate": 9.801417681555765e-05,
+      "loss": 1.2596,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09289750288595905,
+      "grad_norm": 0.3001818358898163,
+      "learning_rate": 9.801150488265762e-05,
+      "loss": 1.2709,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09295825991858557,
+      "grad_norm": 0.23405836522579193,
+      "learning_rate": 9.80088311898866e-05,
+      "loss": 1.1304,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0930190169512121,
+      "grad_norm": 0.20990322530269623,
+      "learning_rate": 9.80061557373426e-05,
+      "loss": 1.1991,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09307977398383863,
+      "grad_norm": 0.198495551943779,
+      "learning_rate": 9.800347852512366e-05,
+      "loss": 1.087,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09314053101646516,
+      "grad_norm": 0.17552435398101807,
+      "learning_rate": 9.800079955332795e-05,
+      "loss": 1.1108,
+      "step": 1533
+    },
+    {
+      "epoch": 0.09320128804909168,
+      "grad_norm": 0.36946913599967957,
+      "learning_rate": 9.799811882205366e-05,
+      "loss": 1.1175,
+      "step": 1534
+    },
+    {
+      "epoch": 0.0932620450817182,
+      "grad_norm": 0.33861300349235535,
+      "learning_rate": 9.799543633139904e-05,
+      "loss": 1.2174,
+      "step": 1535
+    },
+    {
+      "epoch": 0.09332280211434474,
+      "grad_norm": 0.1550939530134201,
+      "learning_rate": 9.799275208146242e-05,
+      "loss": 1.1529,
+      "step": 1536
+    },
+    {
+      "epoch": 0.09338355914697126,
+      "grad_norm": 0.26320332288742065,
+      "learning_rate": 9.79900660723422e-05,
+      "loss": 1.2943,
+      "step": 1537
+    },
+    {
+      "epoch": 0.0934443161795978,
+      "grad_norm": 0.24720866978168488,
+      "learning_rate": 9.798737830413683e-05,
+      "loss": 1.1566,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09350507321222432,
+      "grad_norm": 0.17509174346923828,
+      "learning_rate": 9.798468877694484e-05,
+      "loss": 1.0469,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09356583024485084,
+      "grad_norm": 0.16493506729602814,
+      "learning_rate": 9.79819974908648e-05,
+      "loss": 1.1515,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09362658727747737,
+      "grad_norm": 0.18030470609664917,
+      "learning_rate": 9.797930444599537e-05,
+      "loss": 1.1241,
+      "step": 1541
+    },
+    {
+      "epoch": 0.09368734431010389,
+      "grad_norm": 0.3736095428466797,
+      "learning_rate": 9.797660964243527e-05,
+      "loss": 1.2003,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09374810134273043,
+      "grad_norm": 0.5499204993247986,
+      "learning_rate": 9.797391308028325e-05,
+      "loss": 1.1973,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09380885837535695,
+      "grad_norm": 0.17458085715770721,
+      "learning_rate": 9.797121475963819e-05,
+      "loss": 1.1444,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09386961540798347,
+      "grad_norm": 2.378434658050537,
+      "learning_rate": 9.796851468059897e-05,
+      "loss": 1.1095,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09393037244061,
+      "grad_norm": 0.20916523039340973,
+      "learning_rate": 9.796581284326458e-05,
+      "loss": 1.1683,
+      "step": 1546
+    },
+    {
+      "epoch": 0.09399112947323653,
+      "grad_norm": 0.22721506655216217,
+      "learning_rate": 9.796310924773404e-05,
+      "loss": 1.0673,
+      "step": 1547
+    },
+    {
+      "epoch": 0.09405188650586305,
+      "grad_norm": 0.19539077579975128,
+      "learning_rate": 9.796040389410645e-05,
+      "loss": 1.0904,
+      "step": 1548
+    },
+    {
+      "epoch": 0.09411264353848958,
+      "grad_norm": 0.23285514116287231,
+      "learning_rate": 9.795769678248101e-05,
+      "loss": 1.1746,
+      "step": 1549
+    },
+    {
+      "epoch": 0.0941734005711161,
+      "grad_norm": 0.36843621730804443,
+      "learning_rate": 9.795498791295692e-05,
+      "loss": 1.2558,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09423415760374264,
+      "grad_norm": 0.24364718794822693,
+      "learning_rate": 9.795227728563346e-05,
+      "loss": 1.1049,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09429491463636916,
+      "grad_norm": 0.1703343242406845,
+      "learning_rate": 9.794956490061002e-05,
+      "loss": 1.1199,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09435567166899568,
+      "grad_norm": 0.3334558308124542,
+      "learning_rate": 9.794685075798602e-05,
+      "loss": 1.166,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09441642870162222,
+      "grad_norm": 0.16733412444591522,
+      "learning_rate": 9.794413485786092e-05,
+      "loss": 1.1018,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09447718573424874,
+      "grad_norm": 0.1812572330236435,
+      "learning_rate": 9.79414172003343e-05,
+      "loss": 1.1022,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09453794276687527,
+      "grad_norm": 0.24859310686588287,
+      "learning_rate": 9.793869778550575e-05,
+      "loss": 1.2903,
+      "step": 1556
+    },
+    {
+      "epoch": 0.09459869979950179,
+      "grad_norm": 0.4420587718486786,
+      "learning_rate": 9.793597661347498e-05,
+      "loss": 1.1542,
+      "step": 1557
+    },
+    {
+      "epoch": 0.09465945683212831,
+      "grad_norm": 0.20983166992664337,
+      "learning_rate": 9.793325368434172e-05,
+      "loss": 1.2063,
+      "step": 1558
+    },
+    {
+      "epoch": 0.09472021386475485,
+      "grad_norm": 0.3158051669597626,
+      "learning_rate": 9.793052899820579e-05,
+      "loss": 1.1,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09478097089738137,
+      "grad_norm": 0.2202339470386505,
+      "learning_rate": 9.792780255516704e-05,
+      "loss": 1.1121,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0948417279300079,
+      "grad_norm": 0.3246478736400604,
+      "learning_rate": 9.792507435532543e-05,
+      "loss": 1.2683,
+      "step": 1561
+    },
+    {
+      "epoch": 0.09490248496263443,
+      "grad_norm": 0.5433441400527954,
+      "learning_rate": 9.792234439878095e-05,
+      "loss": 1.205,
+      "step": 1562
+    },
+    {
+      "epoch": 0.09496324199526095,
+      "grad_norm": 0.20413453876972198,
+      "learning_rate": 9.791961268563368e-05,
+      "loss": 1.2,
+      "step": 1563
+    },
+    {
+      "epoch": 0.09502399902788748,
+      "grad_norm": 0.20412945747375488,
+      "learning_rate": 9.791687921598374e-05,
+      "loss": 1.1271,
+      "step": 1564
+    },
+    {
+      "epoch": 0.095084756060514,
+      "grad_norm": 0.3125613033771515,
+      "learning_rate": 9.791414398993133e-05,
+      "loss": 1.0784,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09514551309314054,
+      "grad_norm": 0.22349762916564941,
+      "learning_rate": 9.791140700757671e-05,
+      "loss": 1.2507,
+      "step": 1566
+    },
+    {
+      "epoch": 0.09520627012576706,
+      "grad_norm": 0.30813756585121155,
+      "learning_rate": 9.790866826902021e-05,
+      "loss": 1.0591,
+      "step": 1567
+    },
+    {
+      "epoch": 0.09526702715839358,
+      "grad_norm": 0.13589349389076233,
+      "learning_rate": 9.790592777436221e-05,
+      "loss": 1.1089,
+      "step": 1568
+    },
+    {
+      "epoch": 0.09532778419102012,
+      "grad_norm": 0.3689309358596802,
+      "learning_rate": 9.790318552370317e-05,
+      "loss": 1.263,
+      "step": 1569
+    },
+    {
+      "epoch": 0.09538854122364664,
+      "grad_norm": 0.26732349395751953,
+      "learning_rate": 9.79004415171436e-05,
+      "loss": 1.1818,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09544929825627316,
+      "grad_norm": 0.16413229703903198,
+      "learning_rate": 9.78976957547841e-05,
+      "loss": 1.2115,
+      "step": 1571
+    },
+    {
+      "epoch": 0.09551005528889969,
+      "grad_norm": 0.49256080389022827,
+      "learning_rate": 9.789494823672529e-05,
+      "loss": 1.1012,
+      "step": 1572
+    },
+    {
+      "epoch": 0.09557081232152621,
+      "grad_norm": 0.23552145063877106,
+      "learning_rate": 9.78921989630679e-05,
+      "loss": 1.109,
+      "step": 1573
+    },
+    {
+      "epoch": 0.09563156935415275,
+      "grad_norm": 0.36763328313827515,
+      "learning_rate": 9.78894479339127e-05,
+      "loss": 1.1065,
+      "step": 1574
+    },
+    {
+      "epoch": 0.09569232638677927,
+      "grad_norm": 0.4475506544113159,
+      "learning_rate": 9.788669514936052e-05,
+      "loss": 1.2324,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09575308341940579,
+      "grad_norm": 0.21985404193401337,
+      "learning_rate": 9.788394060951229e-05,
+      "loss": 1.0976,
+      "step": 1576
+    },
+    {
+      "epoch": 0.09581384045203233,
+      "grad_norm": 0.22121183574199677,
+      "learning_rate": 9.788118431446894e-05,
+      "loss": 1.1328,
+      "step": 1577
+    },
+    {
+      "epoch": 0.09587459748465885,
+      "grad_norm": 0.25006103515625,
+      "learning_rate": 9.787842626433154e-05,
+      "loss": 1.1504,
+      "step": 1578
+    },
+    {
+      "epoch": 0.09593535451728538,
+      "grad_norm": 0.18385586142539978,
+      "learning_rate": 9.787566645920116e-05,
+      "loss": 1.0625,
+      "step": 1579
+    },
+    {
+      "epoch": 0.0959961115499119,
+      "grad_norm": 0.2686677873134613,
+      "learning_rate": 9.787290489917899e-05,
+      "loss": 1.2092,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09605686858253842,
+      "grad_norm": 0.14898186922073364,
+      "learning_rate": 9.787014158436623e-05,
+      "loss": 1.0999,
+      "step": 1581
+    },
+    {
+      "epoch": 0.09611762561516496,
+      "grad_norm": 0.2643880546092987,
+      "learning_rate": 9.786737651486416e-05,
+      "loss": 1.196,
+      "step": 1582
+    },
+    {
+      "epoch": 0.09617838264779148,
+      "grad_norm": 0.32479721307754517,
+      "learning_rate": 9.786460969077417e-05,
+      "loss": 1.2581,
+      "step": 1583
+    },
+    {
+      "epoch": 0.09623913968041801,
+      "grad_norm": 0.2819264829158783,
+      "learning_rate": 9.786184111219765e-05,
+      "loss": 1.0928,
+      "step": 1584
+    },
+    {
+      "epoch": 0.09629989671304454,
+      "grad_norm": 0.24512705206871033,
+      "learning_rate": 9.78590707792361e-05,
+      "loss": 1.2701,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09636065374567106,
+      "grad_norm": 0.8442322015762329,
+      "learning_rate": 9.785629869199104e-05,
+      "loss": 1.1323,
+      "step": 1586
+    },
+    {
+      "epoch": 0.09642141077829759,
+      "grad_norm": 2.7841343879699707,
+      "learning_rate": 9.785352485056413e-05,
+      "loss": 1.0889,
+      "step": 1587
+    },
+    {
+      "epoch": 0.09648216781092411,
+      "grad_norm": 0.2092188447713852,
+      "learning_rate": 9.7850749255057e-05,
+      "loss": 1.2605,
+      "step": 1588
+    },
+    {
+      "epoch": 0.09654292484355063,
+      "grad_norm": 0.3294709324836731,
+      "learning_rate": 9.784797190557142e-05,
+      "loss": 1.1841,
+      "step": 1589
+    },
+    {
+      "epoch": 0.09660368187617717,
+      "grad_norm": 0.18616704642772675,
+      "learning_rate": 9.784519280220916e-05,
+      "loss": 1.0728,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09666443890880369,
+      "grad_norm": 0.2453993260860443,
+      "learning_rate": 9.784241194507212e-05,
+      "loss": 1.1221,
+      "step": 1591
+    },
+    {
+      "epoch": 0.09672519594143023,
+      "grad_norm": 0.1521400660276413,
+      "learning_rate": 9.783962933426223e-05,
+      "loss": 1.117,
+      "step": 1592
+    },
+    {
+      "epoch": 0.09678595297405675,
+      "grad_norm": 0.33559900522232056,
+      "learning_rate": 9.783684496988147e-05,
+      "loss": 1.1055,
+      "step": 1593
+    },
+    {
+      "epoch": 0.09684671000668327,
+      "grad_norm": 0.4022634029388428,
+      "learning_rate": 9.783405885203192e-05,
+      "loss": 1.1465,
+      "step": 1594
+    },
+    {
+      "epoch": 0.0969074670393098,
+      "grad_norm": 0.34860312938690186,
+      "learning_rate": 9.783127098081569e-05,
+      "loss": 1.2242,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09696822407193632,
+      "grad_norm": 0.3028814196586609,
+      "learning_rate": 9.782848135633497e-05,
+      "loss": 1.1923,
+      "step": 1596
+    },
+    {
+      "epoch": 0.09702898110456286,
+      "grad_norm": 0.1872156411409378,
+      "learning_rate": 9.782568997869203e-05,
+      "loss": 1.1541,
+      "step": 1597
+    },
+    {
+      "epoch": 0.09708973813718938,
+      "grad_norm": 0.2270052582025528,
+      "learning_rate": 9.782289684798917e-05,
+      "loss": 1.194,
+      "step": 1598
+    },
+    {
+      "epoch": 0.0971504951698159,
+      "grad_norm": 0.27103304862976074,
+      "learning_rate": 9.78201019643288e-05,
+      "loss": 1.1189,
+      "step": 1599
+    },
+    {
+      "epoch": 0.09721125220244244,
+      "grad_norm": 0.18456096947193146,
+      "learning_rate": 9.781730532781333e-05,
+      "loss": 1.1643,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09727200923506896,
+      "grad_norm": 0.2039758712053299,
+      "learning_rate": 9.781450693854531e-05,
+      "loss": 1.1189,
+      "step": 1601
+    },
+    {
+      "epoch": 0.09733276626769549,
+      "grad_norm": 0.5236639976501465,
+      "learning_rate": 9.781170679662728e-05,
+      "loss": 1.1399,
+      "step": 1602
+    },
+    {
+      "epoch": 0.09739352330032201,
+      "grad_norm": 0.8051347136497498,
+      "learning_rate": 9.78089049021619e-05,
+      "loss": 1.2496,
+      "step": 1603
+    },
+    {
+      "epoch": 0.09745428033294853,
+      "grad_norm": 0.35441306233406067,
+      "learning_rate": 9.780610125525187e-05,
+      "loss": 1.2011,
+      "step": 1604
+    },
+    {
+      "epoch": 0.09751503736557507,
+      "grad_norm": 0.32944169640541077,
+      "learning_rate": 9.780329585599995e-05,
+      "loss": 1.088,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09757579439820159,
+      "grad_norm": 0.42272084951400757,
+      "learning_rate": 9.7800488704509e-05,
+      "loss": 1.0473,
+      "step": 1606
+    },
+    {
+      "epoch": 0.09763655143082813,
+      "grad_norm": 0.19113437831401825,
+      "learning_rate": 9.779767980088189e-05,
+      "loss": 1.1319,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09769730846345465,
+      "grad_norm": 0.30906257033348083,
+      "learning_rate": 9.779486914522157e-05,
+      "loss": 1.1885,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09775806549608117,
+      "grad_norm": 0.29779139161109924,
+      "learning_rate": 9.779205673763109e-05,
+      "loss": 1.2956,
+      "step": 1609
+    },
+    {
+      "epoch": 0.0978188225287077,
+      "grad_norm": 0.22943831980228424,
+      "learning_rate": 9.778924257821353e-05,
+      "loss": 1.0899,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09787957956133422,
+      "grad_norm": 0.15976007282733917,
+      "learning_rate": 9.778642666707206e-05,
+      "loss": 1.1268,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09794033659396074,
+      "grad_norm": 0.2557831108570099,
+      "learning_rate": 9.778360900430987e-05,
+      "loss": 1.2081,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09800109362658728,
+      "grad_norm": 0.16285130381584167,
+      "learning_rate": 9.778078959003027e-05,
+      "loss": 1.1314,
+      "step": 1613
+    },
+    {
+      "epoch": 0.0980618506592138,
+      "grad_norm": 0.3343181908130646,
+      "learning_rate": 9.777796842433658e-05,
+      "loss": 1.3532,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09812260769184034,
+      "grad_norm": 0.7450612783432007,
+      "learning_rate": 9.777514550733224e-05,
+      "loss": 1.0902,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09818336472446686,
+      "grad_norm": 0.25191569328308105,
+      "learning_rate": 9.777232083912069e-05,
+      "loss": 1.2048,
+      "step": 1616
+    },
+    {
+      "epoch": 0.09824412175709338,
+      "grad_norm": 1.6702781915664673,
+      "learning_rate": 9.776949441980549e-05,
+      "loss": 1.1222,
+      "step": 1617
+    },
+    {
+      "epoch": 0.09830487878971991,
+      "grad_norm": 0.1711432784795761,
+      "learning_rate": 9.776666624949023e-05,
+      "loss": 1.1741,
+      "step": 1618
+    },
+    {
+      "epoch": 0.09836563582234643,
+      "grad_norm": 0.392383337020874,
+      "learning_rate": 9.77638363282786e-05,
+      "loss": 1.1959,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09842639285497297,
+      "grad_norm": 0.26078468561172485,
+      "learning_rate": 9.77610046562743e-05,
+      "loss": 1.1689,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09848714988759949,
+      "grad_norm": 0.2494201809167862,
+      "learning_rate": 9.775817123358116e-05,
+      "loss": 1.2252,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09854790692022601,
+      "grad_norm": 0.3646562099456787,
+      "learning_rate": 9.7755336060303e-05,
+      "loss": 1.186,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09860866395285255,
+      "grad_norm": 0.240766704082489,
+      "learning_rate": 9.775249913654378e-05,
+      "loss": 1.2895,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09866942098547907,
+      "grad_norm": 1.1815179586410522,
+      "learning_rate": 9.774966046240748e-05,
+      "loss": 1.251,
+      "step": 1624
+    },
+    {
+      "epoch": 0.0987301780181056,
+      "grad_norm": 0.5118951797485352,
+      "learning_rate": 9.774682003799813e-05,
+      "loss": 1.1115,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09879093505073212,
+      "grad_norm": 0.28984978795051575,
+      "learning_rate": 9.774397786341988e-05,
+      "loss": 1.1508,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09885169208335864,
+      "grad_norm": 0.2386028915643692,
+      "learning_rate": 9.774113393877688e-05,
+      "loss": 1.0675,
+      "step": 1627
+    },
+    {
+      "epoch": 0.09891244911598518,
+      "grad_norm": 0.2303244024515152,
+      "learning_rate": 9.77382882641734e-05,
+      "loss": 1.2194,
+      "step": 1628
+    },
+    {
+      "epoch": 0.0989732061486117,
+      "grad_norm": 0.2386283427476883,
+      "learning_rate": 9.773544083971372e-05,
+      "loss": 1.2006,
+      "step": 1629
+    },
+    {
+      "epoch": 0.09903396318123822,
+      "grad_norm": 0.36446717381477356,
+      "learning_rate": 9.773259166550224e-05,
+      "loss": 1.1489,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09909472021386476,
+      "grad_norm": 0.6474947929382324,
+      "learning_rate": 9.772974074164338e-05,
+      "loss": 1.1415,
+      "step": 1631
+    },
+    {
+      "epoch": 0.09915547724649128,
+      "grad_norm": 0.3799828886985779,
+      "learning_rate": 9.772688806824164e-05,
+      "loss": 1.1391,
+      "step": 1632
+    },
+    {
+      "epoch": 0.09921623427911781,
+      "grad_norm": 0.26079803705215454,
+      "learning_rate": 9.772403364540161e-05,
+      "loss": 1.2086,
+      "step": 1633
+    },
+    {
+      "epoch": 0.09927699131174433,
+      "grad_norm": 3.875704050064087,
+      "learning_rate": 9.77211774732279e-05,
+      "loss": 1.1272,
+      "step": 1634
+    },
+    {
+      "epoch": 0.09933774834437085,
+      "grad_norm": 0.5074887871742249,
+      "learning_rate": 9.77183195518252e-05,
+      "loss": 1.1026,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09939850537699739,
+      "grad_norm": 0.21204863488674164,
+      "learning_rate": 9.771545988129827e-05,
+      "loss": 1.2959,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09945926240962391,
+      "grad_norm": 0.2473514825105667,
+      "learning_rate": 9.771259846175193e-05,
+      "loss": 1.1836,
+      "step": 1637
+    },
+    {
+      "epoch": 0.09952001944225045,
+      "grad_norm": 0.26578426361083984,
+      "learning_rate": 9.770973529329108e-05,
+      "loss": 1.0606,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09958077647487697,
+      "grad_norm": 0.20983055233955383,
+      "learning_rate": 9.770687037602066e-05,
+      "loss": 1.0774,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09964153350750349,
+      "grad_norm": 0.23893332481384277,
+      "learning_rate": 9.770400371004568e-05,
+      "loss": 1.1362,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09970229054013002,
+      "grad_norm": 0.2511257827281952,
+      "learning_rate": 9.770113529547123e-05,
+      "loss": 1.2297,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09976304757275654,
+      "grad_norm": 0.18219970166683197,
+      "learning_rate": 9.769826513240244e-05,
+      "loss": 1.208,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09982380460538308,
+      "grad_norm": 0.3269709050655365,
+      "learning_rate": 9.769539322094453e-05,
+      "loss": 1.3274,
+      "step": 1643
+    },
+    {
+      "epoch": 0.0998845616380096,
+      "grad_norm": 0.3039088547229767,
+      "learning_rate": 9.769251956120276e-05,
+      "loss": 1.2201,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09994531867063612,
+      "grad_norm": 0.29194459319114685,
+      "learning_rate": 9.768964415328245e-05,
+      "loss": 1.1301,
+      "step": 1645
+    },
+    {
+      "epoch": 0.10000607570326266,
+      "grad_norm": 0.2589316964149475,
+      "learning_rate": 9.768676699728903e-05,
+      "loss": 1.0832,
+      "step": 1646
+    },
+    {
+      "epoch": 0.10006683273588918,
+      "grad_norm": 0.2882067561149597,
+      "learning_rate": 9.768388809332796e-05,
+      "loss": 1.1025,
+      "step": 1647
+    },
+    {
+      "epoch": 0.10012758976851571,
+      "grad_norm": 0.21772560477256775,
+      "learning_rate": 9.768100744150473e-05,
+      "loss": 1.2394,
+      "step": 1648
+    },
+    {
+      "epoch": 0.10018834680114223,
+      "grad_norm": 0.206012561917305,
+      "learning_rate": 9.767812504192497e-05,
+      "loss": 1.0577,
+      "step": 1649
+    },
+    {
+      "epoch": 0.10024910383376875,
+      "grad_norm": 0.5202322006225586,
+      "learning_rate": 9.767524089469432e-05,
+      "loss": 1.2404,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10030986086639529,
+      "grad_norm": 0.20388440787792206,
+      "learning_rate": 9.76723549999185e-05,
+      "loss": 1.064,
+      "step": 1651
+    },
+    {
+      "epoch": 0.10037061789902181,
+      "grad_norm": 0.16354244947433472,
+      "learning_rate": 9.766946735770328e-05,
+      "loss": 1.0964,
+      "step": 1652
+    },
+    {
+      "epoch": 0.10043137493164833,
+      "grad_norm": 0.17277511954307556,
+      "learning_rate": 9.766657796815453e-05,
+      "loss": 1.0597,
+      "step": 1653
+    },
+    {
+      "epoch": 0.10049213196427487,
+      "grad_norm": 0.2964755892753601,
+      "learning_rate": 9.766368683137815e-05,
+      "loss": 1.2554,
+      "step": 1654
+    },
+    {
+      "epoch": 0.10055288899690139,
+      "grad_norm": 0.16489538550376892,
+      "learning_rate": 9.766079394748012e-05,
+      "loss": 1.1278,
+      "step": 1655
+    },
+    {
+      "epoch": 0.10061364602952792,
+      "grad_norm": 0.30312579870224,
+      "learning_rate": 9.765789931656646e-05,
+      "loss": 1.2997,
+      "step": 1656
+    },
+    {
+      "epoch": 0.10067440306215444,
+      "grad_norm": 0.3645249307155609,
+      "learning_rate": 9.765500293874331e-05,
+      "loss": 1.3426,
+      "step": 1657
+    },
+    {
+      "epoch": 0.10073516009478097,
+      "grad_norm": 0.19458957016468048,
+      "learning_rate": 9.765210481411681e-05,
+      "loss": 1.1525,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1007959171274075,
+      "grad_norm": 0.17949829995632172,
+      "learning_rate": 9.764920494279319e-05,
+      "loss": 1.123,
+      "step": 1659
+    },
+    {
+      "epoch": 0.10085667416003402,
+      "grad_norm": 0.25160884857177734,
+      "learning_rate": 9.764630332487876e-05,
+      "loss": 1.1265,
+      "step": 1660
+    },
+    {
+      "epoch": 0.10091743119266056,
+      "grad_norm": 0.2752400040626526,
+      "learning_rate": 9.764339996047987e-05,
+      "loss": 1.3065,
+      "step": 1661
+    },
+    {
+      "epoch": 0.10097818822528708,
+      "grad_norm": 0.2502821087837219,
+      "learning_rate": 9.764049484970293e-05,
+      "loss": 1.2764,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1010389452579136,
+      "grad_norm": 0.20508714020252228,
+      "learning_rate": 9.763758799265446e-05,
+      "loss": 1.1136,
+      "step": 1663
+    },
+    {
+      "epoch": 0.10109970229054013,
+      "grad_norm": 0.2746257483959198,
+      "learning_rate": 9.763467938944098e-05,
+      "loss": 1.0847,
+      "step": 1664
+    },
+    {
+      "epoch": 0.10116045932316665,
+      "grad_norm": 0.21187755465507507,
+      "learning_rate": 9.763176904016913e-05,
+      "loss": 1.1526,
+      "step": 1665
+    },
+    {
+      "epoch": 0.10122121635579319,
+      "grad_norm": 0.27146077156066895,
+      "learning_rate": 9.762885694494558e-05,
+      "loss": 1.1298,
+      "step": 1666
+    },
+    {
+      "epoch": 0.10128197338841971,
+      "grad_norm": 0.19881564378738403,
+      "learning_rate": 9.762594310387707e-05,
+      "loss": 1.1272,
+      "step": 1667
+    },
+    {
+      "epoch": 0.10134273042104623,
+      "grad_norm": 0.23094390332698822,
+      "learning_rate": 9.762302751707039e-05,
+      "loss": 1.0704,
+      "step": 1668
+    },
+    {
+      "epoch": 0.10140348745367277,
+      "grad_norm": 0.18011099100112915,
+      "learning_rate": 9.762011018463246e-05,
+      "loss": 1.1111,
+      "step": 1669
+    },
+    {
+      "epoch": 0.10146424448629929,
+      "grad_norm": 0.15692360699176788,
+      "learning_rate": 9.761719110667019e-05,
+      "loss": 1.1067,
+      "step": 1670
+    },
+    {
+      "epoch": 0.10152500151892582,
+      "grad_norm": 0.23953652381896973,
+      "learning_rate": 9.761427028329056e-05,
+      "loss": 1.2114,
+      "step": 1671
+    },
+    {
+      "epoch": 0.10158575855155234,
+      "grad_norm": 0.2880992889404297,
+      "learning_rate": 9.761134771460064e-05,
+      "loss": 1.0508,
+      "step": 1672
+    },
+    {
+      "epoch": 0.10164651558417886,
+      "grad_norm": 1.0627996921539307,
+      "learning_rate": 9.760842340070758e-05,
+      "loss": 1.0736,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1017072726168054,
+      "grad_norm": 0.19745668768882751,
+      "learning_rate": 9.760549734171856e-05,
+      "loss": 1.1869,
+      "step": 1674
+    },
+    {
+      "epoch": 0.10176802964943192,
+      "grad_norm": 0.23604899644851685,
+      "learning_rate": 9.760256953774083e-05,
+      "loss": 1.1255,
+      "step": 1675
+    },
+    {
+      "epoch": 0.10182878668205844,
+      "grad_norm": 0.2554061710834503,
+      "learning_rate": 9.759963998888171e-05,
+      "loss": 1.1129,
+      "step": 1676
+    },
+    {
+      "epoch": 0.10188954371468498,
+      "grad_norm": 0.35546228289604187,
+      "learning_rate": 9.759670869524858e-05,
+      "loss": 1.181,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1019503007473115,
+      "grad_norm": 0.3503586947917938,
+      "learning_rate": 9.75937756569489e-05,
+      "loss": 1.25,
+      "step": 1678
+    },
+    {
+      "epoch": 0.10201105777993803,
+      "grad_norm": 0.20203061401844025,
+      "learning_rate": 9.759084087409017e-05,
+      "loss": 1.1654,
+      "step": 1679
+    },
+    {
+      "epoch": 0.10207181481256455,
+      "grad_norm": 0.36170125007629395,
+      "learning_rate": 9.758790434677998e-05,
+      "loss": 1.1405,
+      "step": 1680
+    },
+    {
+      "epoch": 0.10213257184519108,
+      "grad_norm": 0.4543344974517822,
+      "learning_rate": 9.758496607512594e-05,
+      "loss": 1.2559,
+      "step": 1681
+    },
+    {
+      "epoch": 0.10219332887781761,
+      "grad_norm": 0.20456837117671967,
+      "learning_rate": 9.758202605923579e-05,
+      "loss": 1.1705,
+      "step": 1682
+    },
+    {
+      "epoch": 0.10225408591044413,
+      "grad_norm": 0.22422146797180176,
+      "learning_rate": 9.757908429921725e-05,
+      "loss": 1.1844,
+      "step": 1683
+    },
+    {
+      "epoch": 0.10231484294307067,
+      "grad_norm": 0.2442871630191803,
+      "learning_rate": 9.757614079517821e-05,
+      "loss": 1.1151,
+      "step": 1684
+    },
+    {
+      "epoch": 0.10237559997569719,
+      "grad_norm": 0.1535910964012146,
+      "learning_rate": 9.75731955472265e-05,
+      "loss": 1.1399,
+      "step": 1685
+    },
+    {
+      "epoch": 0.10243635700832371,
+      "grad_norm": 0.25678959488868713,
+      "learning_rate": 9.757024855547014e-05,
+      "loss": 1.1288,
+      "step": 1686
+    },
+    {
+      "epoch": 0.10249711404095024,
+      "grad_norm": 0.3460509777069092,
+      "learning_rate": 9.756729982001712e-05,
+      "loss": 1.2216,
+      "step": 1687
+    },
+    {
+      "epoch": 0.10255787107357676,
+      "grad_norm": 0.17896756529808044,
+      "learning_rate": 9.756434934097552e-05,
+      "loss": 1.1684,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1026186281062033,
+      "grad_norm": 0.2536429166793823,
+      "learning_rate": 9.756139711845349e-05,
+      "loss": 1.1846,
+      "step": 1689
+    },
+    {
+      "epoch": 0.10267938513882982,
+      "grad_norm": 0.33528488874435425,
+      "learning_rate": 9.755844315255928e-05,
+      "loss": 1.2888,
+      "step": 1690
+    },
+    {
+      "epoch": 0.10274014217145634,
+      "grad_norm": 0.18117205798625946,
+      "learning_rate": 9.755548744340112e-05,
+      "loss": 1.08,
+      "step": 1691
+    },
+    {
+      "epoch": 0.10280089920408288,
+      "grad_norm": 0.1998588591814041,
+      "learning_rate": 9.75525299910874e-05,
+      "loss": 1.0944,
+      "step": 1692
+    },
+    {
+      "epoch": 0.1028616562367094,
+      "grad_norm": 0.9771398901939392,
+      "learning_rate": 9.754957079572649e-05,
+      "loss": 1.1342,
+      "step": 1693
+    },
+    {
+      "epoch": 0.10292241326933592,
+      "grad_norm": 0.5298376679420471,
+      "learning_rate": 9.754660985742688e-05,
+      "loss": 1.2322,
+      "step": 1694
+    },
+    {
+      "epoch": 0.10298317030196245,
+      "grad_norm": 0.2937504053115845,
+      "learning_rate": 9.754364717629707e-05,
+      "loss": 1.0599,
+      "step": 1695
+    },
+    {
+      "epoch": 0.10304392733458898,
+      "grad_norm": 0.2285371869802475,
+      "learning_rate": 9.75406827524457e-05,
+      "loss": 1.1776,
+      "step": 1696
+    },
+    {
+      "epoch": 0.10310468436721551,
+      "grad_norm": 0.3292471468448639,
+      "learning_rate": 9.753771658598141e-05,
+      "loss": 1.134,
+      "step": 1697
+    },
+    {
+      "epoch": 0.10316544139984203,
+      "grad_norm": 0.2330576777458191,
+      "learning_rate": 9.753474867701294e-05,
+      "loss": 1.1176,
+      "step": 1698
+    },
+    {
+      "epoch": 0.10322619843246855,
+      "grad_norm": 0.24973610043525696,
+      "learning_rate": 9.753177902564905e-05,
+      "loss": 1.1582,
+      "step": 1699
+    },
+    {
+      "epoch": 0.10328695546509509,
+      "grad_norm": 0.9256431460380554,
+      "learning_rate": 9.752880763199864e-05,
+      "loss": 1.1076,
+      "step": 1700
+    },
+    {
+      "epoch": 0.10334771249772161,
+      "grad_norm": 0.26770707964897156,
+      "learning_rate": 9.752583449617058e-05,
+      "loss": 1.1696,
+      "step": 1701
+    },
+    {
+      "epoch": 0.10340846953034814,
+      "grad_norm": 4.23089075088501,
+      "learning_rate": 9.752285961827388e-05,
+      "loss": 1.1308,
+      "step": 1702
+    },
+    {
+      "epoch": 0.10346922656297466,
+      "grad_norm": 0.3109228312969208,
+      "learning_rate": 9.751988299841756e-05,
+      "loss": 1.1146,
+      "step": 1703
+    },
+    {
+      "epoch": 0.10352998359560119,
+      "grad_norm": 0.1922626793384552,
+      "learning_rate": 9.751690463671075e-05,
+      "loss": 1.0804,
+      "step": 1704
+    },
+    {
+      "epoch": 0.10359074062822772,
+      "grad_norm": 0.316914439201355,
+      "learning_rate": 9.751392453326262e-05,
+      "loss": 1.1219,
+      "step": 1705
+    },
+    {
+      "epoch": 0.10365149766085424,
+      "grad_norm": 0.5654424428939819,
+      "learning_rate": 9.751094268818239e-05,
+      "loss": 1.1409,
+      "step": 1706
+    },
+    {
+      "epoch": 0.10371225469348078,
+      "grad_norm": 0.31209778785705566,
+      "learning_rate": 9.750795910157939e-05,
+      "loss": 1.1141,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1037730117261073,
+      "grad_norm": 0.19559776782989502,
+      "learning_rate": 9.750497377356296e-05,
+      "loss": 1.1685,
+      "step": 1708
+    },
+    {
+      "epoch": 0.10383376875873382,
+      "grad_norm": 0.19794543087482452,
+      "learning_rate": 9.750198670424254e-05,
+      "loss": 1.1754,
+      "step": 1709
+    },
+    {
+      "epoch": 0.10389452579136035,
+      "grad_norm": 0.23751486837863922,
+      "learning_rate": 9.749899789372759e-05,
+      "loss": 1.1905,
+      "step": 1710
+    },
+    {
+      "epoch": 0.10395528282398687,
+      "grad_norm": 0.27798542380332947,
+      "learning_rate": 9.749600734212771e-05,
+      "loss": 1.1134,
+      "step": 1711
+    },
+    {
+      "epoch": 0.10401603985661341,
+      "grad_norm": 0.237484872341156,
+      "learning_rate": 9.74930150495525e-05,
+      "loss": 1.0911,
+      "step": 1712
+    },
+    {
+      "epoch": 0.10407679688923993,
+      "grad_norm": 6.3565354347229,
+      "learning_rate": 9.749002101611164e-05,
+      "loss": 1.1332,
+      "step": 1713
+    },
+    {
+      "epoch": 0.10413755392186645,
+      "grad_norm": 0.19158251583576202,
+      "learning_rate": 9.748702524191489e-05,
+      "loss": 1.1536,
+      "step": 1714
+    },
+    {
+      "epoch": 0.10419831095449299,
+      "grad_norm": 0.19259385764598846,
+      "learning_rate": 9.748402772707204e-05,
+      "loss": 1.0904,
+      "step": 1715
+    },
+    {
+      "epoch": 0.10425906798711951,
+      "grad_norm": 0.28275877237319946,
+      "learning_rate": 9.748102847169296e-05,
+      "loss": 1.1084,
+      "step": 1716
+    },
+    {
+      "epoch": 0.10431982501974603,
+      "grad_norm": 0.19667786359786987,
+      "learning_rate": 9.747802747588763e-05,
+      "loss": 1.2508,
+      "step": 1717
+    },
+    {
+      "epoch": 0.10438058205237256,
+      "grad_norm": 0.4985334873199463,
+      "learning_rate": 9.747502473976602e-05,
+      "loss": 1.2777,
+      "step": 1718
+    },
+    {
+      "epoch": 0.10444133908499909,
+      "grad_norm": 0.16044367849826813,
+      "learning_rate": 9.74720202634382e-05,
+      "loss": 1.1079,
+      "step": 1719
+    },
+    {
+      "epoch": 0.10450209611762562,
+      "grad_norm": 0.21739163994789124,
+      "learning_rate": 9.74690140470143e-05,
+      "loss": 1.2065,
+      "step": 1720
+    },
+    {
+      "epoch": 0.10456285315025214,
+      "grad_norm": 0.22656330466270447,
+      "learning_rate": 9.746600609060451e-05,
+      "loss": 1.1168,
+      "step": 1721
+    },
+    {
+      "epoch": 0.10462361018287866,
+      "grad_norm": 0.18773066997528076,
+      "learning_rate": 9.746299639431909e-05,
+      "loss": 1.1428,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1046843672155052,
+      "grad_norm": 0.2294616550207138,
+      "learning_rate": 9.745998495826836e-05,
+      "loss": 1.1423,
+      "step": 1723
+    },
+    {
+      "epoch": 0.10474512424813172,
+      "grad_norm": 0.2877761125564575,
+      "learning_rate": 9.745697178256273e-05,
+      "loss": 1.1063,
+      "step": 1724
+    },
+    {
+      "epoch": 0.10480588128075825,
+      "grad_norm": 0.34316951036453247,
+      "learning_rate": 9.74539568673126e-05,
+      "loss": 1.2422,
+      "step": 1725
+    },
+    {
+      "epoch": 0.10486663831338477,
+      "grad_norm": 13.001538276672363,
+      "learning_rate": 9.745094021262852e-05,
+      "loss": 1.0724,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1049273953460113,
+      "grad_norm": 0.5038268566131592,
+      "learning_rate": 9.744792181862106e-05,
+      "loss": 1.2204,
+      "step": 1727
+    },
+    {
+      "epoch": 0.10498815237863783,
+      "grad_norm": 0.38262248039245605,
+      "learning_rate": 9.744490168540083e-05,
+      "loss": 1.1701,
+      "step": 1728
+    },
+    {
+      "epoch": 0.10504890941126435,
+      "grad_norm": 0.2696792483329773,
+      "learning_rate": 9.74418798130786e-05,
+      "loss": 1.2236,
+      "step": 1729
+    },
+    {
+      "epoch": 0.10510966644389089,
+      "grad_norm": 0.5042509436607361,
+      "learning_rate": 9.743885620176506e-05,
+      "loss": 1.1234,
+      "step": 1730
+    },
+    {
+      "epoch": 0.10517042347651741,
+      "grad_norm": 0.21859440207481384,
+      "learning_rate": 9.743583085157109e-05,
+      "loss": 1.2039,
+      "step": 1731
+    },
+    {
+      "epoch": 0.10523118050914393,
+      "grad_norm": 0.2642357051372528,
+      "learning_rate": 9.743280376260758e-05,
+      "loss": 1.2029,
+      "step": 1732
+    },
+    {
+      "epoch": 0.10529193754177046,
+      "grad_norm": 0.4985358417034149,
+      "learning_rate": 9.742977493498546e-05,
+      "loss": 1.102,
+      "step": 1733
+    },
+    {
+      "epoch": 0.10535269457439699,
+      "grad_norm": 0.17374558746814728,
+      "learning_rate": 9.742674436881578e-05,
+      "loss": 1.0923,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1054134516070235,
+      "grad_norm": 0.35733723640441895,
+      "learning_rate": 9.742371206420962e-05,
+      "loss": 1.1821,
+      "step": 1735
+    },
+    {
+      "epoch": 0.10547420863965004,
+      "grad_norm": 0.4453296363353729,
+      "learning_rate": 9.742067802127812e-05,
+      "loss": 1.0994,
+      "step": 1736
+    },
+    {
+      "epoch": 0.10553496567227656,
+      "grad_norm": 0.193424254655838,
+      "learning_rate": 9.74176422401325e-05,
+      "loss": 1.2039,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1055957227049031,
+      "grad_norm": 0.48401230573654175,
+      "learning_rate": 9.741460472088404e-05,
+      "loss": 1.2039,
+      "step": 1738
+    },
+    {
+      "epoch": 0.10565647973752962,
+      "grad_norm": 0.2332983911037445,
+      "learning_rate": 9.74115654636441e-05,
+      "loss": 1.1414,
+      "step": 1739
+    },
+    {
+      "epoch": 0.10571723677015614,
+      "grad_norm": 0.2843562662601471,
+      "learning_rate": 9.740852446852403e-05,
+      "loss": 1.092,
+      "step": 1740
+    },
+    {
+      "epoch": 0.10577799380278267,
+      "grad_norm": 0.21252241730690002,
+      "learning_rate": 9.740548173563536e-05,
+      "loss": 1.1014,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1058387508354092,
+      "grad_norm": 0.3645320236682892,
+      "learning_rate": 9.740243726508957e-05,
+      "loss": 1.0982,
+      "step": 1742
+    },
+    {
+      "epoch": 0.10589950786803573,
+      "grad_norm": 0.2132883071899414,
+      "learning_rate": 9.739939105699829e-05,
+      "loss": 1.0891,
+      "step": 1743
+    },
+    {
+      "epoch": 0.10596026490066225,
+      "grad_norm": 0.19943320751190186,
+      "learning_rate": 9.739634311147318e-05,
+      "loss": 1.1056,
+      "step": 1744
+    },
+    {
+      "epoch": 0.10602102193328877,
+      "grad_norm": 0.19694775342941284,
+      "learning_rate": 9.739329342862593e-05,
+      "loss": 1.1578,
+      "step": 1745
+    },
+    {
+      "epoch": 0.10608177896591531,
+      "grad_norm": 0.32192084193229675,
+      "learning_rate": 9.739024200856836e-05,
+      "loss": 1.2166,
+      "step": 1746
+    },
+    {
+      "epoch": 0.10614253599854183,
+      "grad_norm": 0.29021573066711426,
+      "learning_rate": 9.738718885141231e-05,
+      "loss": 1.0897,
+      "step": 1747
+    },
+    {
+      "epoch": 0.10620329303116836,
+      "grad_norm": 0.18188691139221191,
+      "learning_rate": 9.73841339572697e-05,
+      "loss": 1.1592,
+      "step": 1748
+    },
+    {
+      "epoch": 0.10626405006379488,
+      "grad_norm": 0.4569520354270935,
+      "learning_rate": 9.738107732625249e-05,
+      "loss": 1.1454,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1063248070964214,
+      "grad_norm": 0.27050137519836426,
+      "learning_rate": 9.737801895847273e-05,
+      "loss": 1.1302,
+      "step": 1750
+    },
+    {
+      "epoch": 0.10638556412904794,
+      "grad_norm": 0.16040939092636108,
+      "learning_rate": 9.737495885404253e-05,
+      "loss": 1.1544,
+      "step": 1751
+    },
+    {
+      "epoch": 0.10644632116167446,
+      "grad_norm": 0.15606291592121124,
+      "learning_rate": 9.737189701307405e-05,
+      "loss": 1.1086,
+      "step": 1752
+    },
+    {
+      "epoch": 0.106507078194301,
+      "grad_norm": 0.1992669552564621,
+      "learning_rate": 9.736883343567955e-05,
+      "loss": 1.0962,
+      "step": 1753
+    },
+    {
+      "epoch": 0.10656783522692752,
+      "grad_norm": 0.23568962514400482,
+      "learning_rate": 9.736576812197129e-05,
+      "loss": 1.2839,
+      "step": 1754
+    },
+    {
+      "epoch": 0.10662859225955404,
+      "grad_norm": 0.21118885278701782,
+      "learning_rate": 9.736270107206164e-05,
+      "loss": 1.345,
+      "step": 1755
+    },
+    {
+      "epoch": 0.10668934929218057,
+      "grad_norm": 0.3307506740093231,
+      "learning_rate": 9.735963228606303e-05,
+      "loss": 1.0882,
+      "step": 1756
+    },
+    {
+      "epoch": 0.1067501063248071,
+      "grad_norm": 0.19812536239624023,
+      "learning_rate": 9.735656176408793e-05,
+      "loss": 1.1122,
+      "step": 1757
+    },
+    {
+      "epoch": 0.10681086335743362,
+      "grad_norm": 0.24365496635437012,
+      "learning_rate": 9.735348950624895e-05,
+      "loss": 1.0807,
+      "step": 1758
+    },
+    {
+      "epoch": 0.10687162039006015,
+      "grad_norm": 0.18023565411567688,
+      "learning_rate": 9.735041551265862e-05,
+      "loss": 1.147,
+      "step": 1759
+    },
+    {
+      "epoch": 0.10693237742268667,
+      "grad_norm": 0.18322375416755676,
+      "learning_rate": 9.734733978342965e-05,
+      "loss": 1.114,
+      "step": 1760
+    },
+    {
+      "epoch": 0.10699313445531321,
+      "grad_norm": 0.1496008038520813,
+      "learning_rate": 9.734426231867482e-05,
+      "loss": 1.1278,
+      "step": 1761
+    },
+    {
+      "epoch": 0.10705389148793973,
+      "grad_norm": 0.23670625686645508,
+      "learning_rate": 9.734118311850688e-05,
+      "loss": 1.1621,
+      "step": 1762
+    },
+    {
+      "epoch": 0.10711464852056625,
+      "grad_norm": 0.5887247323989868,
+      "learning_rate": 9.733810218303873e-05,
+      "loss": 1.3251,
+      "step": 1763
+    },
+    {
+      "epoch": 0.10717540555319278,
+      "grad_norm": 0.35178157687187195,
+      "learning_rate": 9.733501951238329e-05,
+      "loss": 1.2094,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1072361625858193,
+      "grad_norm": 0.19476966559886932,
+      "learning_rate": 9.733193510665358e-05,
+      "loss": 1.0805,
+      "step": 1765
+    },
+    {
+      "epoch": 0.10729691961844584,
+      "grad_norm": 0.20517274737358093,
+      "learning_rate": 9.732884896596262e-05,
+      "loss": 1.1121,
+      "step": 1766
+    },
+    {
+      "epoch": 0.10735767665107236,
+      "grad_norm": 0.1934758722782135,
+      "learning_rate": 9.732576109042355e-05,
+      "loss": 1.1029,
+      "step": 1767
+    },
+    {
+      "epoch": 0.10741843368369888,
+      "grad_norm": 0.28425684571266174,
+      "learning_rate": 9.732267148014955e-05,
+      "loss": 1.2216,
+      "step": 1768
+    },
+    {
+      "epoch": 0.10747919071632542,
+      "grad_norm": 0.9072685837745667,
+      "learning_rate": 9.73195801352539e-05,
+      "loss": 1.1507,
+      "step": 1769
+    },
+    {
+      "epoch": 0.10753994774895194,
+      "grad_norm": 0.4446607232093811,
+      "learning_rate": 9.731648705584989e-05,
+      "loss": 1.1356,
+      "step": 1770
+    },
+    {
+      "epoch": 0.10760070478157847,
+      "grad_norm": 1.4298312664031982,
+      "learning_rate": 9.73133922420509e-05,
+      "loss": 1.2759,
+      "step": 1771
+    },
+    {
+      "epoch": 0.107661461814205,
+      "grad_norm": 0.2696684002876282,
+      "learning_rate": 9.731029569397037e-05,
+      "loss": 1.2024,
+      "step": 1772
+    },
+    {
+      "epoch": 0.10772221884683152,
+      "grad_norm": 0.20106251537799835,
+      "learning_rate": 9.730719741172181e-05,
+      "loss": 1.2146,
+      "step": 1773
+    },
+    {
+      "epoch": 0.10778297587945805,
+      "grad_norm": 0.26841217279434204,
+      "learning_rate": 9.730409739541877e-05,
+      "loss": 1.3705,
+      "step": 1774
+    },
+    {
+      "epoch": 0.10784373291208457,
+      "grad_norm": 0.1646181046962738,
+      "learning_rate": 9.730099564517491e-05,
+      "loss": 1.1204,
+      "step": 1775
+    },
+    {
+      "epoch": 0.10790448994471111,
+      "grad_norm": 0.25667041540145874,
+      "learning_rate": 9.729789216110392e-05,
+      "loss": 1.0921,
+      "step": 1776
+    },
+    {
+      "epoch": 0.10796524697733763,
+      "grad_norm": 0.33628520369529724,
+      "learning_rate": 9.729478694331954e-05,
+      "loss": 1.1826,
+      "step": 1777
+    },
+    {
+      "epoch": 0.10802600400996415,
+      "grad_norm": 0.21840974688529968,
+      "learning_rate": 9.72916799919356e-05,
+      "loss": 1.2281,
+      "step": 1778
+    },
+    {
+      "epoch": 0.10808676104259068,
+      "grad_norm": 0.8014854788780212,
+      "learning_rate": 9.7288571307066e-05,
+      "loss": 1.2282,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1081475180752172,
+      "grad_norm": 0.2203182727098465,
+      "learning_rate": 9.728546088882466e-05,
+      "loss": 1.1345,
+      "step": 1780
+    },
+    {
+      "epoch": 0.10820827510784373,
+      "grad_norm": 0.463958740234375,
+      "learning_rate": 9.728234873732562e-05,
+      "loss": 1.109,
+      "step": 1781
+    },
+    {
+      "epoch": 0.10826903214047026,
+      "grad_norm": 0.32588332891464233,
+      "learning_rate": 9.727923485268296e-05,
+      "loss": 1.165,
+      "step": 1782
+    },
+    {
+      "epoch": 0.10832978917309678,
+      "grad_norm": 0.2140774130821228,
+      "learning_rate": 9.727611923501079e-05,
+      "loss": 1.2313,
+      "step": 1783
+    },
+    {
+      "epoch": 0.10839054620572332,
+      "grad_norm": 0.2978186309337616,
+      "learning_rate": 9.727300188442334e-05,
+      "loss": 1.1756,
+      "step": 1784
+    },
+    {
+      "epoch": 0.10845130323834984,
+      "grad_norm": 0.18642206490039825,
+      "learning_rate": 9.726988280103485e-05,
+      "loss": 1.0946,
+      "step": 1785
+    },
+    {
+      "epoch": 0.10851206027097636,
+      "grad_norm": 0.2392071634531021,
+      "learning_rate": 9.726676198495968e-05,
+      "loss": 1.2295,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1085728173036029,
+      "grad_norm": 0.3410801887512207,
+      "learning_rate": 9.726363943631223e-05,
+      "loss": 1.1238,
+      "step": 1787
+    },
+    {
+      "epoch": 0.10863357433622942,
+      "grad_norm": 0.19988597929477692,
+      "learning_rate": 9.726051515520692e-05,
+      "loss": 1.0833,
+      "step": 1788
+    },
+    {
+      "epoch": 0.10869433136885595,
+      "grad_norm": 0.48140695691108704,
+      "learning_rate": 9.725738914175831e-05,
+      "loss": 1.1641,
+      "step": 1789
+    },
+    {
+      "epoch": 0.10875508840148247,
+      "grad_norm": 0.25766366720199585,
+      "learning_rate": 9.725426139608094e-05,
+      "loss": 1.1248,
+      "step": 1790
+    },
+    {
+      "epoch": 0.108815845434109,
+      "grad_norm": 0.7419375777244568,
+      "learning_rate": 9.72511319182895e-05,
+      "loss": 1.2262,
+      "step": 1791
+    },
+    {
+      "epoch": 0.10887660246673553,
+      "grad_norm": 0.3252478241920471,
+      "learning_rate": 9.724800070849869e-05,
+      "loss": 1.1827,
+      "step": 1792
+    },
+    {
+      "epoch": 0.10893735949936205,
+      "grad_norm": 0.3237375020980835,
+      "learning_rate": 9.724486776682328e-05,
+      "loss": 1.3831,
+      "step": 1793
+    },
+    {
+      "epoch": 0.10899811653198858,
+      "grad_norm": 0.1870550662279129,
+      "learning_rate": 9.72417330933781e-05,
+      "loss": 1.1907,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1090588735646151,
+      "grad_norm": 0.27788177132606506,
+      "learning_rate": 9.723859668827809e-05,
+      "loss": 1.1254,
+      "step": 1795
+    },
+    {
+      "epoch": 0.10911963059724163,
+      "grad_norm": 0.49869078397750854,
+      "learning_rate": 9.723545855163816e-05,
+      "loss": 1.2561,
+      "step": 1796
+    },
+    {
+      "epoch": 0.10918038762986816,
+      "grad_norm": 0.7547779083251953,
+      "learning_rate": 9.723231868357337e-05,
+      "loss": 1.1466,
+      "step": 1797
+    },
+    {
+      "epoch": 0.10924114466249468,
+      "grad_norm": 0.5552695393562317,
+      "learning_rate": 9.722917708419881e-05,
+      "loss": 1.111,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1093019016951212,
+      "grad_norm": 0.17805175483226776,
+      "learning_rate": 9.722603375362963e-05,
+      "loss": 1.1214,
+      "step": 1799
+    },
+    {
+      "epoch": 0.10936265872774774,
+      "grad_norm": 0.4798164963722229,
+      "learning_rate": 9.722288869198106e-05,
+      "loss": 1.1838,
+      "step": 1800
+    },
+    {
+      "epoch": 0.10942341576037426,
+      "grad_norm": 0.5813899040222168,
+      "learning_rate": 9.721974189936837e-05,
+      "loss": 1.2165,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1094841727930008,
+      "grad_norm": 0.17604577541351318,
+      "learning_rate": 9.721659337590693e-05,
+      "loss": 1.1176,
+      "step": 1802
+    },
+    {
+      "epoch": 0.10954492982562732,
+      "grad_norm": 0.9933179616928101,
+      "learning_rate": 9.721344312171213e-05,
+      "loss": 1.2117,
+      "step": 1803
+    },
+    {
+      "epoch": 0.10960568685825384,
+      "grad_norm": 0.3179173767566681,
+      "learning_rate": 9.721029113689943e-05,
+      "loss": 1.0823,
+      "step": 1804
+    },
+    {
+      "epoch": 0.10966644389088037,
+      "grad_norm": 1.0834643840789795,
+      "learning_rate": 9.72071374215844e-05,
+      "loss": 1.1301,
+      "step": 1805
+    },
+    {
+      "epoch": 0.10972720092350689,
+      "grad_norm": 0.29624152183532715,
+      "learning_rate": 9.72039819758826e-05,
+      "loss": 1.1041,
+      "step": 1806
+    },
+    {
+      "epoch": 0.10978795795613343,
+      "grad_norm": 0.3164976239204407,
+      "learning_rate": 9.720082479990974e-05,
+      "loss": 1.2096,
+      "step": 1807
+    },
+    {
+      "epoch": 0.10984871498875995,
+      "grad_norm": 0.25658395886421204,
+      "learning_rate": 9.719766589378152e-05,
+      "loss": 1.0777,
+      "step": 1808
+    },
+    {
+      "epoch": 0.10990947202138647,
+      "grad_norm": 0.3463870882987976,
+      "learning_rate": 9.719450525761373e-05,
+      "loss": 1.1936,
+      "step": 1809
+    },
+    {
+      "epoch": 0.109970229054013,
+      "grad_norm": 0.2338787317276001,
+      "learning_rate": 9.719134289152224e-05,
+      "loss": 1.1548,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11003098608663953,
+      "grad_norm": 0.16285766661167145,
+      "learning_rate": 9.718817879562293e-05,
+      "loss": 1.1017,
+      "step": 1811
+    },
+    {
+      "epoch": 0.11009174311926606,
+      "grad_norm": 0.4239717423915863,
+      "learning_rate": 9.718501297003181e-05,
+      "loss": 1.2376,
+      "step": 1812
+    },
+    {
+      "epoch": 0.11015250015189258,
+      "grad_norm": 0.24120941758155823,
+      "learning_rate": 9.718184541486492e-05,
+      "loss": 1.097,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1102132571845191,
+      "grad_norm": 0.3054282069206238,
+      "learning_rate": 9.717867613023838e-05,
+      "loss": 1.1179,
+      "step": 1814
+    },
+    {
+      "epoch": 0.11027401421714564,
+      "grad_norm": 0.2704208493232727,
+      "learning_rate": 9.717550511626834e-05,
+      "loss": 1.1473,
+      "step": 1815
+    },
+    {
+      "epoch": 0.11033477124977216,
+      "grad_norm": 0.500048816204071,
+      "learning_rate": 9.717233237307104e-05,
+      "loss": 1.2067,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1103955282823987,
+      "grad_norm": 0.2887870967388153,
+      "learning_rate": 9.716915790076278e-05,
+      "loss": 1.2325,
+      "step": 1817
+    },
+    {
+      "epoch": 0.11045628531502522,
+      "grad_norm": 0.2627791166305542,
+      "learning_rate": 9.71659816994599e-05,
+      "loss": 1.1396,
+      "step": 1818
+    },
+    {
+      "epoch": 0.11051704234765174,
+      "grad_norm": 0.20322081446647644,
+      "learning_rate": 9.716280376927887e-05,
+      "loss": 1.1051,
+      "step": 1819
+    },
+    {
+      "epoch": 0.11057779938027827,
+      "grad_norm": 0.18145835399627686,
+      "learning_rate": 9.715962411033615e-05,
+      "loss": 1.1262,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11063855641290479,
+      "grad_norm": 0.31508344411849976,
+      "learning_rate": 9.715644272274828e-05,
+      "loss": 1.2081,
+      "step": 1821
+    },
+    {
+      "epoch": 0.11069931344553131,
+      "grad_norm": 0.560475766658783,
+      "learning_rate": 9.715325960663188e-05,
+      "loss": 1.3645,
+      "step": 1822
+    },
+    {
+      "epoch": 0.11076007047815785,
+      "grad_norm": 0.1997685581445694,
+      "learning_rate": 9.715007476210364e-05,
+      "loss": 1.1891,
+      "step": 1823
+    },
+    {
+      "epoch": 0.11082082751078437,
+      "grad_norm": 0.29143574833869934,
+      "learning_rate": 9.714688818928031e-05,
+      "loss": 1.0803,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1108815845434109,
+      "grad_norm": 0.2193433791399002,
+      "learning_rate": 9.714369988827867e-05,
+      "loss": 1.0965,
+      "step": 1825
+    },
+    {
+      "epoch": 0.11094234157603743,
+      "grad_norm": 0.6419479250907898,
+      "learning_rate": 9.71405098592156e-05,
+      "loss": 1.095,
+      "step": 1826
+    },
+    {
+      "epoch": 0.11100309860866395,
+      "grad_norm": 0.2430139183998108,
+      "learning_rate": 9.713731810220803e-05,
+      "loss": 1.0786,
+      "step": 1827
+    },
+    {
+      "epoch": 0.11106385564129048,
+      "grad_norm": 0.22759728133678436,
+      "learning_rate": 9.713412461737296e-05,
+      "loss": 1.1457,
+      "step": 1828
+    },
+    {
+      "epoch": 0.111124612673917,
+      "grad_norm": 0.3639973998069763,
+      "learning_rate": 9.713092940482743e-05,
+      "loss": 1.1163,
+      "step": 1829
+    },
+    {
+      "epoch": 0.11118536970654354,
+      "grad_norm": 8.10069465637207,
+      "learning_rate": 9.712773246468859e-05,
+      "loss": 1.2446,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11124612673917006,
+      "grad_norm": 0.19557404518127441,
+      "learning_rate": 9.712453379707357e-05,
+      "loss": 1.1389,
+      "step": 1831
+    },
+    {
+      "epoch": 0.11130688377179658,
+      "grad_norm": 0.23012775182724,
+      "learning_rate": 9.712133340209969e-05,
+      "loss": 1.2565,
+      "step": 1832
+    },
+    {
+      "epoch": 0.11136764080442312,
+      "grad_norm": 0.19063954055309296,
+      "learning_rate": 9.711813127988422e-05,
+      "loss": 1.0712,
+      "step": 1833
+    },
+    {
+      "epoch": 0.11142839783704964,
+      "grad_norm": 0.44537022709846497,
+      "learning_rate": 9.711492743054453e-05,
+      "loss": 1.1444,
+      "step": 1834
+    },
+    {
+      "epoch": 0.11148915486967617,
+      "grad_norm": 0.146654412150383,
+      "learning_rate": 9.711172185419808e-05,
+      "loss": 1.1223,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11154991190230269,
+      "grad_norm": 0.1751268208026886,
+      "learning_rate": 9.710851455096235e-05,
+      "loss": 1.1225,
+      "step": 1836
+    },
+    {
+      "epoch": 0.11161066893492921,
+      "grad_norm": 0.34237268567085266,
+      "learning_rate": 9.71053055209549e-05,
+      "loss": 1.0778,
+      "step": 1837
+    },
+    {
+      "epoch": 0.11167142596755575,
+      "grad_norm": 0.2107909619808197,
+      "learning_rate": 9.71020947642934e-05,
+      "loss": 1.1847,
+      "step": 1838
+    },
+    {
+      "epoch": 0.11173218300018227,
+      "grad_norm": 0.6270161867141724,
+      "learning_rate": 9.70988822810955e-05,
+      "loss": 1.0963,
+      "step": 1839
+    },
+    {
+      "epoch": 0.11179294003280879,
+      "grad_norm": 0.21378874778747559,
+      "learning_rate": 9.709566807147895e-05,
+      "loss": 1.1671,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11185369706543533,
+      "grad_norm": 0.1499761939048767,
+      "learning_rate": 9.70924521355616e-05,
+      "loss": 1.0603,
+      "step": 1841
+    },
+    {
+      "epoch": 0.11191445409806185,
+      "grad_norm": 0.1642175018787384,
+      "learning_rate": 9.70892344734613e-05,
+      "loss": 1.1341,
+      "step": 1842
+    },
+    {
+      "epoch": 0.11197521113068838,
+      "grad_norm": 0.56465744972229,
+      "learning_rate": 9.708601508529602e-05,
+      "loss": 1.1082,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1120359681633149,
+      "grad_norm": 0.19741006195545197,
+      "learning_rate": 9.708279397118376e-05,
+      "loss": 1.2651,
+      "step": 1844
+    },
+    {
+      "epoch": 0.11209672519594142,
+      "grad_norm": 0.18394073843955994,
+      "learning_rate": 9.707957113124257e-05,
+      "loss": 1.1114,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11215748222856796,
+      "grad_norm": 0.2483065277338028,
+      "learning_rate": 9.707634656559061e-05,
+      "loss": 1.1527,
+      "step": 1846
+    },
+    {
+      "epoch": 0.11221823926119448,
+      "grad_norm": 0.2128303050994873,
+      "learning_rate": 9.707312027434608e-05,
+      "loss": 1.2226,
+      "step": 1847
+    },
+    {
+      "epoch": 0.11227899629382102,
+      "grad_norm": 0.22518791258335114,
+      "learning_rate": 9.706989225762721e-05,
+      "loss": 1.1515,
+      "step": 1848
+    },
+    {
+      "epoch": 0.11233975332644754,
+      "grad_norm": 0.31041276454925537,
+      "learning_rate": 9.706666251555234e-05,
+      "loss": 1.2707,
+      "step": 1849
+    },
+    {
+      "epoch": 0.11240051035907406,
+      "grad_norm": 0.18872512876987457,
+      "learning_rate": 9.706343104823987e-05,
+      "loss": 1.1886,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11246126739170059,
+      "grad_norm": 0.4778657853603363,
+      "learning_rate": 9.706019785580823e-05,
+      "loss": 1.1597,
+      "step": 1851
+    },
+    {
+      "epoch": 0.11252202442432711,
+      "grad_norm": 0.20142699778079987,
+      "learning_rate": 9.705696293837594e-05,
+      "loss": 1.1902,
+      "step": 1852
+    },
+    {
+      "epoch": 0.11258278145695365,
+      "grad_norm": 0.179465651512146,
+      "learning_rate": 9.705372629606158e-05,
+      "loss": 1.1218,
+      "step": 1853
+    },
+    {
+      "epoch": 0.11264353848958017,
+      "grad_norm": 0.22530245780944824,
+      "learning_rate": 9.705048792898378e-05,
+      "loss": 1.1999,
+      "step": 1854
+    },
+    {
+      "epoch": 0.11270429552220669,
+      "grad_norm": 0.2807729244232178,
+      "learning_rate": 9.704724783726127e-05,
+      "loss": 1.1244,
+      "step": 1855
+    },
+    {
+      "epoch": 0.11276505255483323,
+      "grad_norm": 0.305794894695282,
+      "learning_rate": 9.704400602101278e-05,
+      "loss": 1.1369,
+      "step": 1856
+    },
+    {
+      "epoch": 0.11282580958745975,
+      "grad_norm": 0.37672796845436096,
+      "learning_rate": 9.704076248035717e-05,
+      "loss": 1.2643,
+      "step": 1857
+    },
+    {
+      "epoch": 0.11288656662008628,
+      "grad_norm": 0.19682687520980835,
+      "learning_rate": 9.703751721541329e-05,
+      "loss": 1.1059,
+      "step": 1858
+    },
+    {
+      "epoch": 0.1129473236527128,
+      "grad_norm": 0.15786024928092957,
+      "learning_rate": 9.703427022630015e-05,
+      "loss": 1.2051,
+      "step": 1859
+    },
+    {
+      "epoch": 0.11300808068533932,
+      "grad_norm": 0.5112740397453308,
+      "learning_rate": 9.703102151313674e-05,
+      "loss": 1.2165,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11306883771796586,
+      "grad_norm": 0.3164134621620178,
+      "learning_rate": 9.702777107604215e-05,
+      "loss": 1.1781,
+      "step": 1861
+    },
+    {
+      "epoch": 0.11312959475059238,
+      "grad_norm": 0.18649594485759735,
+      "learning_rate": 9.70245189151355e-05,
+      "loss": 1.0863,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1131903517832189,
+      "grad_norm": 0.19943934679031372,
+      "learning_rate": 9.702126503053603e-05,
+      "loss": 1.2801,
+      "step": 1863
+    },
+    {
+      "epoch": 0.11325110881584544,
+      "grad_norm": 0.29414230585098267,
+      "learning_rate": 9.701800942236301e-05,
+      "loss": 1.1029,
+      "step": 1864
+    },
+    {
+      "epoch": 0.11331186584847196,
+      "grad_norm": 0.2480049878358841,
+      "learning_rate": 9.701475209073576e-05,
+      "loss": 1.1546,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11337262288109849,
+      "grad_norm": 0.2536844313144684,
+      "learning_rate": 9.701149303577368e-05,
+      "loss": 1.2017,
+      "step": 1866
+    },
+    {
+      "epoch": 0.11343337991372501,
+      "grad_norm": 0.20147861540317535,
+      "learning_rate": 9.700823225759623e-05,
+      "loss": 1.1049,
+      "step": 1867
+    },
+    {
+      "epoch": 0.11349413694635153,
+      "grad_norm": 5.159074306488037,
+      "learning_rate": 9.700496975632295e-05,
+      "loss": 1.1675,
+      "step": 1868
+    },
+    {
+      "epoch": 0.11355489397897807,
+      "grad_norm": 1.5400784015655518,
+      "learning_rate": 9.700170553207341e-05,
+      "loss": 1.1369,
+      "step": 1869
+    },
+    {
+      "epoch": 0.11361565101160459,
+      "grad_norm": 0.6182984709739685,
+      "learning_rate": 9.699843958496729e-05,
+      "loss": 1.146,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11367640804423113,
+      "grad_norm": 0.19991667568683624,
+      "learning_rate": 9.699517191512426e-05,
+      "loss": 1.1584,
+      "step": 1871
+    },
+    {
+      "epoch": 0.11373716507685765,
+      "grad_norm": 0.35074588656425476,
+      "learning_rate": 9.699190252266414e-05,
+      "loss": 1.124,
+      "step": 1872
+    },
+    {
+      "epoch": 0.11379792210948417,
+      "grad_norm": 0.243534654378891,
+      "learning_rate": 9.698863140770673e-05,
+      "loss": 1.2255,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1138586791421107,
+      "grad_norm": 0.20570048689842224,
+      "learning_rate": 9.698535857037197e-05,
+      "loss": 1.1179,
+      "step": 1874
+    },
+    {
+      "epoch": 0.11391943617473722,
+      "grad_norm": 0.2860731780529022,
+      "learning_rate": 9.698208401077979e-05,
+      "loss": 1.1203,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11398019320736376,
+      "grad_norm": 0.4588428735733032,
+      "learning_rate": 9.697880772905026e-05,
+      "loss": 1.3248,
+      "step": 1876
+    },
+    {
+      "epoch": 0.11404095023999028,
+      "grad_norm": 0.3027544319629669,
+      "learning_rate": 9.697552972530344e-05,
+      "loss": 1.3643,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1141017072726168,
+      "grad_norm": 0.24429982900619507,
+      "learning_rate": 9.697224999965948e-05,
+      "loss": 1.2928,
+      "step": 1878
+    },
+    {
+      "epoch": 0.11416246430524334,
+      "grad_norm": 0.2989386022090912,
+      "learning_rate": 9.696896855223865e-05,
+      "loss": 1.0822,
+      "step": 1879
+    },
+    {
+      "epoch": 0.11422322133786986,
+      "grad_norm": 0.2681278586387634,
+      "learning_rate": 9.696568538316117e-05,
+      "loss": 1.1431,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11428397837049639,
+      "grad_norm": 14.011563301086426,
+      "learning_rate": 9.696240049254743e-05,
+      "loss": 1.1651,
+      "step": 1881
+    },
+    {
+      "epoch": 0.11434473540312291,
+      "grad_norm": 0.5634395480155945,
+      "learning_rate": 9.695911388051783e-05,
+      "loss": 1.1015,
+      "step": 1882
+    },
+    {
+      "epoch": 0.11440549243574943,
+      "grad_norm": 0.42623788118362427,
+      "learning_rate": 9.695582554719283e-05,
+      "loss": 1.0696,
+      "step": 1883
+    },
+    {
+      "epoch": 0.11446624946837597,
+      "grad_norm": 0.7773213386535645,
+      "learning_rate": 9.695253549269296e-05,
+      "loss": 1.1993,
+      "step": 1884
+    },
+    {
+      "epoch": 0.11452700650100249,
+      "grad_norm": 1.0263391733169556,
+      "learning_rate": 9.694924371713883e-05,
+      "loss": 1.1557,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11458776353362901,
+      "grad_norm": 0.4339512288570404,
+      "learning_rate": 9.694595022065109e-05,
+      "loss": 1.2667,
+      "step": 1886
+    },
+    {
+      "epoch": 0.11464852056625555,
+      "grad_norm": 0.3500816524028778,
+      "learning_rate": 9.694265500335046e-05,
+      "loss": 1.2477,
+      "step": 1887
+    },
+    {
+      "epoch": 0.11470927759888207,
+      "grad_norm": 0.885793924331665,
+      "learning_rate": 9.693935806535776e-05,
+      "loss": 1.1568,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1147700346315086,
+      "grad_norm": 0.6202738285064697,
+      "learning_rate": 9.693605940679381e-05,
+      "loss": 1.1341,
+      "step": 1889
+    },
+    {
+      "epoch": 0.11483079166413512,
+      "grad_norm": 0.3628444969654083,
+      "learning_rate": 9.693275902777953e-05,
+      "loss": 1.1351,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11489154869676164,
+      "grad_norm": 0.6494534611701965,
+      "learning_rate": 9.692945692843589e-05,
+      "loss": 1.2245,
+      "step": 1891
+    },
+    {
+      "epoch": 0.11495230572938818,
+      "grad_norm": 0.6455859541893005,
+      "learning_rate": 9.692615310888394e-05,
+      "loss": 1.1446,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1150130627620147,
+      "grad_norm": 0.48541146516799927,
+      "learning_rate": 9.692284756924477e-05,
+      "loss": 1.1475,
+      "step": 1893
+    },
+    {
+      "epoch": 0.11507381979464124,
+      "grad_norm": 0.4792094826698303,
+      "learning_rate": 9.691954030963957e-05,
+      "loss": 1.227,
+      "step": 1894
+    },
+    {
+      "epoch": 0.11513457682726776,
+      "grad_norm": 0.7279345393180847,
+      "learning_rate": 9.691623133018955e-05,
+      "loss": 1.3239,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11519533385989428,
+      "grad_norm": 0.2750690281391144,
+      "learning_rate": 9.691292063101598e-05,
+      "loss": 1.1426,
+      "step": 1896
+    },
+    {
+      "epoch": 0.11525609089252081,
+      "grad_norm": 0.40976056456565857,
+      "learning_rate": 9.690960821224025e-05,
+      "loss": 1.106,
+      "step": 1897
+    },
+    {
+      "epoch": 0.11531684792514733,
+      "grad_norm": 0.6517239212989807,
+      "learning_rate": 9.690629407398376e-05,
+      "loss": 1.2699,
+      "step": 1898
+    },
+    {
+      "epoch": 0.11537760495777387,
+      "grad_norm": 0.2929058372974396,
+      "learning_rate": 9.6902978216368e-05,
+      "loss": 1.2669,
+      "step": 1899
+    },
+    {
+      "epoch": 0.11543836199040039,
+      "grad_norm": 0.4887823760509491,
+      "learning_rate": 9.68996606395145e-05,
+      "loss": 1.1478,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11549911902302691,
+      "grad_norm": 0.2126196324825287,
+      "learning_rate": 9.689634134354489e-05,
+      "loss": 1.0822,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11555987605565345,
+      "grad_norm": 0.19483217597007751,
+      "learning_rate": 9.68930203285808e-05,
+      "loss": 1.1242,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11562063308827997,
+      "grad_norm": 0.37179771065711975,
+      "learning_rate": 9.688969759474399e-05,
+      "loss": 1.1922,
+      "step": 1903
+    },
+    {
+      "epoch": 0.11568139012090649,
+      "grad_norm": 0.37623727321624756,
+      "learning_rate": 9.688637314215625e-05,
+      "loss": 1.0962,
+      "step": 1904
+    },
+    {
+      "epoch": 0.11574214715353302,
+      "grad_norm": 0.25979217886924744,
+      "learning_rate": 9.688304697093947e-05,
+      "loss": 1.1665,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11580290418615954,
+      "grad_norm": 0.15539799630641937,
+      "learning_rate": 9.68797190812155e-05,
+      "loss": 1.1105,
+      "step": 1906
+    },
+    {
+      "epoch": 0.11586366121878608,
+      "grad_norm": 0.1922324299812317,
+      "learning_rate": 9.687638947310638e-05,
+      "loss": 1.1682,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1159244182514126,
+      "grad_norm": 1.510488510131836,
+      "learning_rate": 9.687305814673415e-05,
+      "loss": 1.117,
+      "step": 1908
+    },
+    {
+      "epoch": 0.11598517528403912,
+      "grad_norm": 14.29343032836914,
+      "learning_rate": 9.686972510222091e-05,
+      "loss": 1.1978,
+      "step": 1909
+    },
+    {
+      "epoch": 0.11604593231666566,
+      "grad_norm": 1.5797078609466553,
+      "learning_rate": 9.686639033968883e-05,
+      "loss": 1.1014,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11610668934929218,
+      "grad_norm": 0.2675571143627167,
+      "learning_rate": 9.686305385926016e-05,
+      "loss": 1.1589,
+      "step": 1911
+    },
+    {
+      "epoch": 0.11616744638191871,
+      "grad_norm": 0.21127727627754211,
+      "learning_rate": 9.685971566105719e-05,
+      "loss": 1.094,
+      "step": 1912
+    },
+    {
+      "epoch": 0.11622820341454523,
+      "grad_norm": 0.3406842350959778,
+      "learning_rate": 9.685637574520228e-05,
+      "loss": 1.1041,
+      "step": 1913
+    },
+    {
+      "epoch": 0.11628896044717175,
+      "grad_norm": 0.23028457164764404,
+      "learning_rate": 9.685303411181788e-05,
+      "loss": 1.1838,
+      "step": 1914
+    },
+    {
+      "epoch": 0.11634971747979829,
+      "grad_norm": 0.19239498674869537,
+      "learning_rate": 9.684969076102643e-05,
+      "loss": 1.1053,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11641047451242481,
+      "grad_norm": 0.3043873906135559,
+      "learning_rate": 9.684634569295053e-05,
+      "loss": 1.1838,
+      "step": 1916
+    },
+    {
+      "epoch": 0.11647123154505135,
+      "grad_norm": 0.24072453379631042,
+      "learning_rate": 9.684299890771275e-05,
+      "loss": 1.101,
+      "step": 1917
+    },
+    {
+      "epoch": 0.11653198857767787,
+      "grad_norm": 0.2098928540945053,
+      "learning_rate": 9.683965040543582e-05,
+      "loss": 1.2147,
+      "step": 1918
+    },
+    {
+      "epoch": 0.11659274561030439,
+      "grad_norm": 0.2557503283023834,
+      "learning_rate": 9.683630018624243e-05,
+      "loss": 1.0932,
+      "step": 1919
+    },
+    {
+      "epoch": 0.11665350264293092,
+      "grad_norm": 0.2447894662618637,
+      "learning_rate": 9.683294825025541e-05,
+      "loss": 1.2172,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11671425967555744,
+      "grad_norm": 0.2724704146385193,
+      "learning_rate": 9.682959459759763e-05,
+      "loss": 1.1753,
+      "step": 1921
+    },
+    {
+      "epoch": 0.11677501670818398,
+      "grad_norm": 1.4174976348876953,
+      "learning_rate": 9.6826239228392e-05,
+      "loss": 1.1319,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1168357737408105,
+      "grad_norm": 0.2176712453365326,
+      "learning_rate": 9.682288214276152e-05,
+      "loss": 1.1378,
+      "step": 1923
+    },
+    {
+      "epoch": 0.11689653077343702,
+      "grad_norm": 0.1698758900165558,
+      "learning_rate": 9.681952334082923e-05,
+      "loss": 1.0954,
+      "step": 1924
+    },
+    {
+      "epoch": 0.11695728780606356,
+      "grad_norm": 0.15976884961128235,
+      "learning_rate": 9.681616282271828e-05,
+      "loss": 1.0782,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11701804483869008,
+      "grad_norm": 0.2583989202976227,
+      "learning_rate": 9.681280058855183e-05,
+      "loss": 1.0959,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1170788018713166,
+      "grad_norm": 0.30633655190467834,
+      "learning_rate": 9.680943663845312e-05,
+      "loss": 1.1164,
+      "step": 1927
+    },
+    {
+      "epoch": 0.11713955890394313,
+      "grad_norm": 0.4438948333263397,
+      "learning_rate": 9.680607097254546e-05,
+      "loss": 1.1459,
+      "step": 1928
+    },
+    {
+      "epoch": 0.11720031593656965,
+      "grad_norm": 0.17000195384025574,
+      "learning_rate": 9.680270359095222e-05,
+      "loss": 1.0975,
+      "step": 1929
+    },
+    {
+      "epoch": 0.11726107296919619,
+      "grad_norm": 0.20205284655094147,
+      "learning_rate": 9.679933449379684e-05,
+      "loss": 1.2877,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11732183000182271,
+      "grad_norm": 0.20819339156150818,
+      "learning_rate": 9.67959636812028e-05,
+      "loss": 1.137,
+      "step": 1931
+    },
+    {
+      "epoch": 0.11738258703444923,
+      "grad_norm": 0.20226801931858063,
+      "learning_rate": 9.679259115329369e-05,
+      "loss": 1.095,
+      "step": 1932
+    },
+    {
+      "epoch": 0.11744334406707577,
+      "grad_norm": 0.26550427079200745,
+      "learning_rate": 9.678921691019308e-05,
+      "loss": 1.2601,
+      "step": 1933
+    },
+    {
+      "epoch": 0.11750410109970229,
+      "grad_norm": 0.1841442584991455,
+      "learning_rate": 9.678584095202468e-05,
+      "loss": 1.161,
+      "step": 1934
+    },
+    {
+      "epoch": 0.11756485813232882,
+      "grad_norm": 0.2929319739341736,
+      "learning_rate": 9.678246327891224e-05,
+      "loss": 1.1133,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11762561516495534,
+      "grad_norm": 0.3009735345840454,
+      "learning_rate": 9.677908389097959e-05,
+      "loss": 1.0937,
+      "step": 1936
+    },
+    {
+      "epoch": 0.11768637219758186,
+      "grad_norm": 0.2584429681301117,
+      "learning_rate": 9.677570278835055e-05,
+      "loss": 1.1119,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1177471292302084,
+      "grad_norm": 0.3425711989402771,
+      "learning_rate": 9.677231997114909e-05,
+      "loss": 1.1299,
+      "step": 1938
+    },
+    {
+      "epoch": 0.11780788626283492,
+      "grad_norm": 0.23567362129688263,
+      "learning_rate": 9.676893543949921e-05,
+      "loss": 1.1459,
+      "step": 1939
+    },
+    {
+      "epoch": 0.11786864329546146,
+      "grad_norm": 0.19330216944217682,
+      "learning_rate": 9.676554919352494e-05,
+      "loss": 1.1361,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11792940032808798,
+      "grad_norm": 0.5115233659744263,
+      "learning_rate": 9.676216123335045e-05,
+      "loss": 1.3027,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1179901573607145,
+      "grad_norm": 0.18392927944660187,
+      "learning_rate": 9.675877155909988e-05,
+      "loss": 1.1172,
+      "step": 1942
+    },
+    {
+      "epoch": 0.11805091439334103,
+      "grad_norm": 0.3016204535961151,
+      "learning_rate": 9.675538017089752e-05,
+      "loss": 1.0995,
+      "step": 1943
+    },
+    {
+      "epoch": 0.11811167142596755,
+      "grad_norm": 0.48495960235595703,
+      "learning_rate": 9.675198706886765e-05,
+      "loss": 1.1967,
+      "step": 1944
+    },
+    {
+      "epoch": 0.11817242845859408,
+      "grad_norm": 0.20842774212360382,
+      "learning_rate": 9.674859225313465e-05,
+      "loss": 1.1686,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11823318549122061,
+      "grad_norm": 0.5592312216758728,
+      "learning_rate": 9.674519572382297e-05,
+      "loss": 1.1757,
+      "step": 1946
+    },
+    {
+      "epoch": 0.11829394252384713,
+      "grad_norm": 1.581610083580017,
+      "learning_rate": 9.674179748105712e-05,
+      "loss": 1.3027,
+      "step": 1947
+    },
+    {
+      "epoch": 0.11835469955647367,
+      "grad_norm": 0.27288055419921875,
+      "learning_rate": 9.673839752496163e-05,
+      "loss": 1.3086,
+      "step": 1948
+    },
+    {
+      "epoch": 0.11841545658910019,
+      "grad_norm": 2.6161465644836426,
+      "learning_rate": 9.673499585566114e-05,
+      "loss": 1.0706,
+      "step": 1949
+    },
+    {
+      "epoch": 0.11847621362172671,
+      "grad_norm": 0.3142314553260803,
+      "learning_rate": 9.673159247328036e-05,
+      "loss": 1.1208,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11853697065435324,
+      "grad_norm": 0.1905544102191925,
+      "learning_rate": 9.672818737794403e-05,
+      "loss": 1.1175,
+      "step": 1951
+    },
+    {
+      "epoch": 0.11859772768697976,
+      "grad_norm": 0.19226299226284027,
+      "learning_rate": 9.672478056977694e-05,
+      "loss": 1.1381,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1186584847196063,
+      "grad_norm": 0.4684247672557831,
+      "learning_rate": 9.672137204890398e-05,
+      "loss": 1.1744,
+      "step": 1953
+    },
+    {
+      "epoch": 0.11871924175223282,
+      "grad_norm": 0.17875200510025024,
+      "learning_rate": 9.671796181545012e-05,
+      "loss": 1.1189,
+      "step": 1954
+    },
+    {
+      "epoch": 0.11877999878485934,
+      "grad_norm": 1.081351399421692,
+      "learning_rate": 9.671454986954035e-05,
+      "loss": 1.1078,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11884075581748588,
+      "grad_norm": 0.19021007418632507,
+      "learning_rate": 9.671113621129969e-05,
+      "loss": 1.1212,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1189015128501124,
+      "grad_norm": 0.22308015823364258,
+      "learning_rate": 9.670772084085331e-05,
+      "loss": 1.2717,
+      "step": 1957
+    },
+    {
+      "epoch": 0.11896226988273893,
+      "grad_norm": 0.21829305589199066,
+      "learning_rate": 9.670430375832641e-05,
+      "loss": 1.1336,
+      "step": 1958
+    },
+    {
+      "epoch": 0.11902302691536545,
+      "grad_norm": 0.19864550232887268,
+      "learning_rate": 9.670088496384422e-05,
+      "loss": 1.0512,
+      "step": 1959
+    },
+    {
+      "epoch": 0.11908378394799198,
+      "grad_norm": 0.14482282102108002,
+      "learning_rate": 9.669746445753207e-05,
+      "loss": 1.1212,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11914454098061851,
+      "grad_norm": 0.1603849083185196,
+      "learning_rate": 9.669404223951533e-05,
+      "loss": 1.1138,
+      "step": 1961
+    },
+    {
+      "epoch": 0.11920529801324503,
+      "grad_norm": 0.2783275246620178,
+      "learning_rate": 9.669061830991943e-05,
+      "loss": 1.1294,
+      "step": 1962
+    },
+    {
+      "epoch": 0.11926605504587157,
+      "grad_norm": 0.17078639566898346,
+      "learning_rate": 9.668719266886992e-05,
+      "loss": 1.1165,
+      "step": 1963
+    },
+    {
+      "epoch": 0.11932681207849809,
+      "grad_norm": 0.35063353180885315,
+      "learning_rate": 9.668376531649233e-05,
+      "loss": 1.1364,
+      "step": 1964
+    },
+    {
+      "epoch": 0.11938756911112461,
+      "grad_norm": 0.2500370144844055,
+      "learning_rate": 9.668033625291229e-05,
+      "loss": 1.2439,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11944832614375114,
+      "grad_norm": 0.18498098850250244,
+      "learning_rate": 9.667690547825549e-05,
+      "loss": 1.1212,
+      "step": 1966
+    },
+    {
+      "epoch": 0.11950908317637766,
+      "grad_norm": 0.3198544681072235,
+      "learning_rate": 9.667347299264772e-05,
+      "loss": 1.1064,
+      "step": 1967
+    },
+    {
+      "epoch": 0.11956984020900419,
+      "grad_norm": 0.1935877799987793,
+      "learning_rate": 9.667003879621477e-05,
+      "loss": 1.1498,
+      "step": 1968
+    },
+    {
+      "epoch": 0.11963059724163072,
+      "grad_norm": 0.2101842761039734,
+      "learning_rate": 9.666660288908251e-05,
+      "loss": 1.0786,
+      "step": 1969
+    },
+    {
+      "epoch": 0.11969135427425724,
+      "grad_norm": 0.204456627368927,
+      "learning_rate": 9.666316527137693e-05,
+      "loss": 1.0552,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11975211130688378,
+      "grad_norm": 0.1915925294160843,
+      "learning_rate": 9.665972594322399e-05,
+      "loss": 1.0912,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1198128683395103,
+      "grad_norm": 0.7430527806282043,
+      "learning_rate": 9.665628490474977e-05,
+      "loss": 1.2088,
+      "step": 1972
+    },
+    {
+      "epoch": 0.11987362537213682,
+      "grad_norm": 1.4529510736465454,
+      "learning_rate": 9.665284215608041e-05,
+      "loss": 1.1075,
+      "step": 1973
+    },
+    {
+      "epoch": 0.11993438240476335,
+      "grad_norm": 0.26790040731430054,
+      "learning_rate": 9.664939769734212e-05,
+      "loss": 1.2762,
+      "step": 1974
+    },
+    {
+      "epoch": 0.11999513943738987,
+      "grad_norm": 0.256966769695282,
+      "learning_rate": 9.664595152866113e-05,
+      "loss": 1.1003,
+      "step": 1975
+    },
+    {
+      "epoch": 0.12005589647001641,
+      "grad_norm": 1.6612577438354492,
+      "learning_rate": 9.664250365016376e-05,
+      "loss": 1.2115,
+      "step": 1976
+    },
+    {
+      "epoch": 0.12011665350264293,
+      "grad_norm": 0.26674968004226685,
+      "learning_rate": 9.663905406197642e-05,
+      "loss": 1.2234,
+      "step": 1977
+    },
+    {
+      "epoch": 0.12017741053526945,
+      "grad_norm": 0.24348914623260498,
+      "learning_rate": 9.663560276422553e-05,
+      "loss": 1.0615,
+      "step": 1978
+    },
+    {
+      "epoch": 0.12023816756789599,
+      "grad_norm": 0.18058128654956818,
+      "learning_rate": 9.66321497570376e-05,
+      "loss": 1.1416,
+      "step": 1979
+    },
+    {
+      "epoch": 0.12029892460052251,
+      "grad_norm": 0.14928728342056274,
+      "learning_rate": 9.662869504053923e-05,
+      "loss": 1.2116,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12035968163314904,
+      "grad_norm": 0.4173123240470886,
+      "learning_rate": 9.6625238614857e-05,
+      "loss": 1.2366,
+      "step": 1981
+    },
+    {
+      "epoch": 0.12042043866577556,
+      "grad_norm": 0.26743918657302856,
+      "learning_rate": 9.662178048011765e-05,
+      "loss": 1.1249,
+      "step": 1982
+    },
+    {
+      "epoch": 0.12048119569840209,
+      "grad_norm": 0.25977104902267456,
+      "learning_rate": 9.661832063644794e-05,
+      "loss": 1.1114,
+      "step": 1983
+    },
+    {
+      "epoch": 0.12054195273102862,
+      "grad_norm": 6.254899024963379,
+      "learning_rate": 9.661485908397465e-05,
+      "loss": 1.1036,
+      "step": 1984
+    },
+    {
+      "epoch": 0.12060270976365514,
+      "grad_norm": 0.32930925488471985,
+      "learning_rate": 9.66113958228247e-05,
+      "loss": 1.0942,
+      "step": 1985
+    },
+    {
+      "epoch": 0.12066346679628168,
+      "grad_norm": 0.2594299912452698,
+      "learning_rate": 9.660793085312503e-05,
+      "loss": 1.294,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1207242238289082,
+      "grad_norm": 0.21070104837417603,
+      "learning_rate": 9.660446417500265e-05,
+      "loss": 1.176,
+      "step": 1987
+    },
+    {
+      "epoch": 0.12078498086153472,
+      "grad_norm": 0.2816389203071594,
+      "learning_rate": 9.660099578858461e-05,
+      "loss": 1.1136,
+      "step": 1988
+    },
+    {
+      "epoch": 0.12084573789416125,
+      "grad_norm": 0.18064992129802704,
+      "learning_rate": 9.659752569399807e-05,
+      "loss": 1.0706,
+      "step": 1989
+    },
+    {
+      "epoch": 0.12090649492678777,
+      "grad_norm": 0.19540466368198395,
+      "learning_rate": 9.659405389137021e-05,
+      "loss": 1.0981,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1209672519594143,
+      "grad_norm": 0.23018568754196167,
+      "learning_rate": 9.659058038082833e-05,
+      "loss": 1.1232,
+      "step": 1991
+    },
+    {
+      "epoch": 0.12102800899204083,
+      "grad_norm": 0.42158737778663635,
+      "learning_rate": 9.65871051624997e-05,
+      "loss": 1.2598,
+      "step": 1992
+    },
+    {
+      "epoch": 0.12108876602466735,
+      "grad_norm": 0.3112030625343323,
+      "learning_rate": 9.658362823651172e-05,
+      "loss": 1.186,
+      "step": 1993
+    },
+    {
+      "epoch": 0.12114952305729389,
+      "grad_norm": 0.4480117857456207,
+      "learning_rate": 9.658014960299185e-05,
+      "loss": 1.159,
+      "step": 1994
+    },
+    {
+      "epoch": 0.12121028008992041,
+      "grad_norm": 0.29857707023620605,
+      "learning_rate": 9.657666926206758e-05,
+      "loss": 1.1786,
+      "step": 1995
+    },
+    {
+      "epoch": 0.12127103712254693,
+      "grad_norm": 2.4238693714141846,
+      "learning_rate": 9.65731872138665e-05,
+      "loss": 1.164,
+      "step": 1996
+    },
+    {
+      "epoch": 0.12133179415517346,
+      "grad_norm": 0.29271450638771057,
+      "learning_rate": 9.656970345851625e-05,
+      "loss": 1.3141,
+      "step": 1997
+    },
+    {
+      "epoch": 0.12139255118779999,
+      "grad_norm": 0.14524228870868683,
+      "learning_rate": 9.656621799614452e-05,
+      "loss": 1.1148,
+      "step": 1998
+    },
+    {
+      "epoch": 0.12145330822042652,
+      "grad_norm": 0.18711404502391815,
+      "learning_rate": 9.656273082687905e-05,
+      "loss": 1.1084,
+      "step": 1999
+    },
+    {
+      "epoch": 0.12151406525305304,
+      "grad_norm": 0.36704376339912415,
+      "learning_rate": 9.655924195084768e-05,
+      "loss": 1.1146,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12157482228567956,
+      "grad_norm": 0.17453031241893768,
+      "learning_rate": 9.655575136817831e-05,
+      "loss": 1.3484,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1216355793183061,
+      "grad_norm": 0.4107412099838257,
+      "learning_rate": 9.655225907899888e-05,
+      "loss": 1.0793,
+      "step": 2002
+    },
+    {
+      "epoch": 0.12169633635093262,
+      "grad_norm": 0.24669001996517181,
+      "learning_rate": 9.654876508343738e-05,
+      "loss": 1.0473,
+      "step": 2003
+    },
+    {
+      "epoch": 0.12175709338355915,
+      "grad_norm": 0.9859693050384521,
+      "learning_rate": 9.654526938162191e-05,
+      "loss": 1.1558,
+      "step": 2004
+    },
+    {
+      "epoch": 0.12181785041618567,
+      "grad_norm": 0.7443386316299438,
+      "learning_rate": 9.65417719736806e-05,
+      "loss": 1.1378,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1218786074488122,
+      "grad_norm": 0.21646425127983093,
+      "learning_rate": 9.653827285974163e-05,
+      "loss": 1.1407,
+      "step": 2006
+    },
+    {
+      "epoch": 0.12193936448143873,
+      "grad_norm": 0.2408502846956253,
+      "learning_rate": 9.653477203993327e-05,
+      "loss": 1.1871,
+      "step": 2007
+    },
+    {
+      "epoch": 0.12200012151406525,
+      "grad_norm": 0.21233363449573517,
+      "learning_rate": 9.653126951438385e-05,
+      "loss": 1.1506,
+      "step": 2008
+    },
+    {
+      "epoch": 0.12206087854669177,
+      "grad_norm": 0.19248034060001373,
+      "learning_rate": 9.652776528322175e-05,
+      "loss": 1.1267,
+      "step": 2009
+    },
+    {
+      "epoch": 0.12212163557931831,
+      "grad_norm": 0.17353682219982147,
+      "learning_rate": 9.652425934657542e-05,
+      "loss": 1.2013,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12218239261194483,
+      "grad_norm": 0.16145093739032745,
+      "learning_rate": 9.652075170457337e-05,
+      "loss": 1.1187,
+      "step": 2011
+    },
+    {
+      "epoch": 0.12224314964457136,
+      "grad_norm": 0.17668648064136505,
+      "learning_rate": 9.651724235734418e-05,
+      "loss": 1.1471,
+      "step": 2012
+    },
+    {
+      "epoch": 0.12230390667719789,
+      "grad_norm": 0.1906491369009018,
+      "learning_rate": 9.651373130501649e-05,
+      "loss": 1.1935,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1223646637098244,
+      "grad_norm": 0.2117723822593689,
+      "learning_rate": 9.651021854771897e-05,
+      "loss": 1.1284,
+      "step": 2014
+    },
+    {
+      "epoch": 0.12242542074245094,
+      "grad_norm": 0.33802300691604614,
+      "learning_rate": 9.650670408558041e-05,
+      "loss": 1.1806,
+      "step": 2015
+    },
+    {
+      "epoch": 0.12248617777507746,
+      "grad_norm": 0.4124891459941864,
+      "learning_rate": 9.65031879187296e-05,
+      "loss": 1.2326,
+      "step": 2016
+    },
+    {
+      "epoch": 0.122546934807704,
+      "grad_norm": 3.347226858139038,
+      "learning_rate": 9.649967004729548e-05,
+      "loss": 1.1085,
+      "step": 2017
+    },
+    {
+      "epoch": 0.12260769184033052,
+      "grad_norm": 0.3927828371524811,
+      "learning_rate": 9.649615047140695e-05,
+      "loss": 1.1895,
+      "step": 2018
+    },
+    {
+      "epoch": 0.12266844887295704,
+      "grad_norm": 0.27255886793136597,
+      "learning_rate": 9.649262919119303e-05,
+      "loss": 1.1501,
+      "step": 2019
+    },
+    {
+      "epoch": 0.12272920590558357,
+      "grad_norm": 0.2866484522819519,
+      "learning_rate": 9.64891062067828e-05,
+      "loss": 1.1284,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1227899629382101,
+      "grad_norm": 0.24547314643859863,
+      "learning_rate": 9.64855815183054e-05,
+      "loss": 1.1576,
+      "step": 2021
+    },
+    {
+      "epoch": 0.12285071997083663,
+      "grad_norm": 0.46214425563812256,
+      "learning_rate": 9.648205512589003e-05,
+      "loss": 1.0792,
+      "step": 2022
+    },
+    {
+      "epoch": 0.12291147700346315,
+      "grad_norm": 0.23865380883216858,
+      "learning_rate": 9.647852702966596e-05,
+      "loss": 1.1023,
+      "step": 2023
+    },
+    {
+      "epoch": 0.12297223403608967,
+      "grad_norm": 0.48760348558425903,
+      "learning_rate": 9.647499722976247e-05,
+      "loss": 1.1602,
+      "step": 2024
+    },
+    {
+      "epoch": 0.12303299106871621,
+      "grad_norm": 0.2786392569541931,
+      "learning_rate": 9.647146572630899e-05,
+      "loss": 1.1064,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12309374810134273,
+      "grad_norm": 0.2121281921863556,
+      "learning_rate": 9.646793251943493e-05,
+      "loss": 1.1417,
+      "step": 2026
+    },
+    {
+      "epoch": 0.12315450513396926,
+      "grad_norm": 1.844231367111206,
+      "learning_rate": 9.646439760926984e-05,
+      "loss": 1.0822,
+      "step": 2027
+    },
+    {
+      "epoch": 0.12321526216659578,
+      "grad_norm": 0.2326965034008026,
+      "learning_rate": 9.646086099594326e-05,
+      "loss": 1.175,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1232760191992223,
+      "grad_norm": 0.3406953811645508,
+      "learning_rate": 9.645732267958485e-05,
+      "loss": 1.3142,
+      "step": 2029
+    },
+    {
+      "epoch": 0.12333677623184884,
+      "grad_norm": 0.40433642268180847,
+      "learning_rate": 9.645378266032429e-05,
+      "loss": 1.2599,
+      "step": 2030
+    },
+    {
+      "epoch": 0.12339753326447536,
+      "grad_norm": 0.228359192609787,
+      "learning_rate": 9.645024093829136e-05,
+      "loss": 1.0942,
+      "step": 2031
+    },
+    {
+      "epoch": 0.12345829029710188,
+      "grad_norm": 0.2280920147895813,
+      "learning_rate": 9.644669751361587e-05,
+      "loss": 1.1415,
+      "step": 2032
+    },
+    {
+      "epoch": 0.12351904732972842,
+      "grad_norm": 0.45911288261413574,
+      "learning_rate": 9.644315238642769e-05,
+      "loss": 1.0738,
+      "step": 2033
+    },
+    {
+      "epoch": 0.12357980436235494,
+      "grad_norm": 0.21488778293132782,
+      "learning_rate": 9.643960555685679e-05,
+      "loss": 1.0921,
+      "step": 2034
+    },
+    {
+      "epoch": 0.12364056139498147,
+      "grad_norm": 0.5509289503097534,
+      "learning_rate": 9.643605702503318e-05,
+      "loss": 1.1341,
+      "step": 2035
+    },
+    {
+      "epoch": 0.123701318427608,
+      "grad_norm": 0.16387374699115753,
+      "learning_rate": 9.643250679108691e-05,
+      "loss": 1.1103,
+      "step": 2036
+    },
+    {
+      "epoch": 0.12376207546023452,
+      "grad_norm": 0.2071555256843567,
+      "learning_rate": 9.642895485514813e-05,
+      "loss": 1.1346,
+      "step": 2037
+    },
+    {
+      "epoch": 0.12382283249286105,
+      "grad_norm": 0.22651129961013794,
+      "learning_rate": 9.642540121734705e-05,
+      "loss": 1.1164,
+      "step": 2038
+    },
+    {
+      "epoch": 0.12388358952548757,
+      "grad_norm": 0.17861492931842804,
+      "learning_rate": 9.64218458778139e-05,
+      "loss": 1.1708,
+      "step": 2039
+    },
+    {
+      "epoch": 0.12394434655811411,
+      "grad_norm": 0.26989665627479553,
+      "learning_rate": 9.641828883667902e-05,
+      "loss": 1.1775,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12400510359074063,
+      "grad_norm": 0.29572224617004395,
+      "learning_rate": 9.641473009407278e-05,
+      "loss": 1.1271,
+      "step": 2041
+    },
+    {
+      "epoch": 0.12406586062336715,
+      "grad_norm": 0.19515860080718994,
+      "learning_rate": 9.641116965012565e-05,
+      "loss": 1.1297,
+      "step": 2042
+    },
+    {
+      "epoch": 0.12412661765599368,
+      "grad_norm": 0.5475988984107971,
+      "learning_rate": 9.640760750496812e-05,
+      "loss": 1.3058,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1241873746886202,
+      "grad_norm": 0.1982891708612442,
+      "learning_rate": 9.640404365873077e-05,
+      "loss": 1.0944,
+      "step": 2044
+    },
+    {
+      "epoch": 0.12424813172124674,
+      "grad_norm": 0.20374320447444916,
+      "learning_rate": 9.640047811154424e-05,
+      "loss": 1.2383,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12430888875387326,
+      "grad_norm": 0.22014868259429932,
+      "learning_rate": 9.639691086353919e-05,
+      "loss": 1.1166,
+      "step": 2046
+    },
+    {
+      "epoch": 0.12436964578649978,
+      "grad_norm": 0.1554224193096161,
+      "learning_rate": 9.639334191484642e-05,
+      "loss": 1.1428,
+      "step": 2047
+    },
+    {
+      "epoch": 0.12443040281912632,
+      "grad_norm": 0.20734219253063202,
+      "learning_rate": 9.638977126559675e-05,
+      "loss": 1.1305,
+      "step": 2048
+    },
+    {
+      "epoch": 0.12449115985175284,
+      "grad_norm": 0.2932432293891907,
+      "learning_rate": 9.638619891592102e-05,
+      "loss": 1.1506,
+      "step": 2049
+    },
+    {
+      "epoch": 0.12455191688437936,
+      "grad_norm": 0.2742735743522644,
+      "learning_rate": 9.638262486595022e-05,
+      "loss": 1.0932,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1246126739170059,
+      "grad_norm": 0.5033144354820251,
+      "learning_rate": 9.637904911581532e-05,
+      "loss": 1.171,
+      "step": 2051
+    },
+    {
+      "epoch": 0.12467343094963242,
+      "grad_norm": 0.22795073688030243,
+      "learning_rate": 9.637547166564741e-05,
+      "loss": 1.1873,
+      "step": 2052
+    },
+    {
+      "epoch": 0.12473418798225895,
+      "grad_norm": 0.2395406812429428,
+      "learning_rate": 9.637189251557766e-05,
+      "loss": 1.1115,
+      "step": 2053
+    },
+    {
+      "epoch": 0.12479494501488547,
+      "grad_norm": 0.31437185406684875,
+      "learning_rate": 9.636831166573719e-05,
+      "loss": 1.0799,
+      "step": 2054
+    },
+    {
+      "epoch": 0.124855702047512,
+      "grad_norm": 0.24858003854751587,
+      "learning_rate": 9.63647291162573e-05,
+      "loss": 1.1045,
+      "step": 2055
+    },
+    {
+      "epoch": 0.12491645908013853,
+      "grad_norm": 0.2202848494052887,
+      "learning_rate": 9.63611448672693e-05,
+      "loss": 1.1087,
+      "step": 2056
+    },
+    {
+      "epoch": 0.12497721611276505,
+      "grad_norm": 0.2237938791513443,
+      "learning_rate": 9.635755891890459e-05,
+      "loss": 1.1379,
+      "step": 2057
+    },
+    {
+      "epoch": 0.12503797314539158,
+      "grad_norm": 0.259045273065567,
+      "learning_rate": 9.635397127129458e-05,
+      "loss": 1.1036,
+      "step": 2058
+    },
+    {
+      "epoch": 0.12509873017801812,
+      "grad_norm": 0.42144864797592163,
+      "learning_rate": 9.63503819245708e-05,
+      "loss": 1.2534,
+      "step": 2059
+    },
+    {
+      "epoch": 0.12515948721064463,
+      "grad_norm": 0.3736537992954254,
+      "learning_rate": 9.634679087886481e-05,
+      "loss": 1.2059,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12522024424327116,
+      "grad_norm": 0.3590710759162903,
+      "learning_rate": 9.634319813430826e-05,
+      "loss": 1.1077,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1252810012758977,
+      "grad_norm": 0.21131914854049683,
+      "learning_rate": 9.633960369103279e-05,
+      "loss": 1.117,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1253417583085242,
+      "grad_norm": 2.375375270843506,
+      "learning_rate": 9.63360075491702e-05,
+      "loss": 1.2209,
+      "step": 2063
+    },
+    {
+      "epoch": 0.12540251534115074,
+      "grad_norm": 0.25554004311561584,
+      "learning_rate": 9.63324097088523e-05,
+      "loss": 1.2378,
+      "step": 2064
+    },
+    {
+      "epoch": 0.12546327237377727,
+      "grad_norm": 0.18050119280815125,
+      "learning_rate": 9.632881017021098e-05,
+      "loss": 1.1429,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12552402940640378,
+      "grad_norm": 0.3133840560913086,
+      "learning_rate": 9.632520893337815e-05,
+      "loss": 1.1511,
+      "step": 2066
+    },
+    {
+      "epoch": 0.12558478643903032,
+      "grad_norm": 0.16320054233074188,
+      "learning_rate": 9.632160599848585e-05,
+      "loss": 1.1112,
+      "step": 2067
+    },
+    {
+      "epoch": 0.12564554347165685,
+      "grad_norm": 0.1702062040567398,
+      "learning_rate": 9.63180013656661e-05,
+      "loss": 1.1993,
+      "step": 2068
+    },
+    {
+      "epoch": 0.12570630050428336,
+      "grad_norm": 0.18877971172332764,
+      "learning_rate": 9.631439503505108e-05,
+      "loss": 1.2219,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1257670575369099,
+      "grad_norm": 0.21893955767154694,
+      "learning_rate": 9.631078700677296e-05,
+      "loss": 1.1868,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12582781456953643,
+      "grad_norm": 0.186843603849411,
+      "learning_rate": 9.6307177280964e-05,
+      "loss": 1.095,
+      "step": 2071
+    },
+    {
+      "epoch": 0.12588857160216296,
+      "grad_norm": 0.21330058574676514,
+      "learning_rate": 9.630356585775647e-05,
+      "loss": 1.2843,
+      "step": 2072
+    },
+    {
+      "epoch": 0.12594932863478947,
+      "grad_norm": 2.1996421813964844,
+      "learning_rate": 9.62999527372828e-05,
+      "loss": 1.1012,
+      "step": 2073
+    },
+    {
+      "epoch": 0.126010085667416,
+      "grad_norm": 0.21556246280670166,
+      "learning_rate": 9.629633791967541e-05,
+      "loss": 1.0712,
+      "step": 2074
+    },
+    {
+      "epoch": 0.12607084270004254,
+      "grad_norm": 1.677614450454712,
+      "learning_rate": 9.629272140506682e-05,
+      "loss": 1.0929,
+      "step": 2075
+    },
+    {
+      "epoch": 0.12613159973266905,
+      "grad_norm": 0.1669522523880005,
+      "learning_rate": 9.628910319358956e-05,
+      "loss": 1.2191,
+      "step": 2076
+    },
+    {
+      "epoch": 0.12619235676529558,
+      "grad_norm": 0.21765844523906708,
+      "learning_rate": 9.62854832853763e-05,
+      "loss": 1.1005,
+      "step": 2077
+    },
+    {
+      "epoch": 0.12625311379792212,
+      "grad_norm": 14.662498474121094,
+      "learning_rate": 9.628186168055968e-05,
+      "loss": 1.1484,
+      "step": 2078
+    },
+    {
+      "epoch": 0.12631387083054862,
+      "grad_norm": 1.579979419708252,
+      "learning_rate": 9.627823837927247e-05,
+      "loss": 1.1292,
+      "step": 2079
+    },
+    {
+      "epoch": 0.12637462786317516,
+      "grad_norm": 0.18985305726528168,
+      "learning_rate": 9.627461338164751e-05,
+      "loss": 1.0702,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1264353848958017,
+      "grad_norm": 0.25318366289138794,
+      "learning_rate": 9.627098668781763e-05,
+      "loss": 1.2087,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1264961419284282,
+      "grad_norm": 0.20284001529216766,
+      "learning_rate": 9.62673582979158e-05,
+      "loss": 1.1284,
+      "step": 2082
+    },
+    {
+      "epoch": 0.12655689896105474,
+      "grad_norm": 0.22754044830799103,
+      "learning_rate": 9.626372821207501e-05,
+      "loss": 1.1194,
+      "step": 2083
+    },
+    {
+      "epoch": 0.12661765599368127,
+      "grad_norm": 0.2247520387172699,
+      "learning_rate": 9.62600964304283e-05,
+      "loss": 1.2736,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1266784130263078,
+      "grad_norm": 0.1490245908498764,
+      "learning_rate": 9.625646295310884e-05,
+      "loss": 1.0571,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12673917005893431,
+      "grad_norm": 0.22191564738750458,
+      "learning_rate": 9.625282778024978e-05,
+      "loss": 1.2504,
+      "step": 2086
+    },
+    {
+      "epoch": 0.12679992709156085,
+      "grad_norm": 0.26560288667678833,
+      "learning_rate": 9.624919091198437e-05,
+      "loss": 1.1647,
+      "step": 2087
+    },
+    {
+      "epoch": 0.12686068412418738,
+      "grad_norm": 1.2738986015319824,
+      "learning_rate": 9.624555234844595e-05,
+      "loss": 1.0918,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1269214411568139,
+      "grad_norm": 0.16314783692359924,
+      "learning_rate": 9.624191208976785e-05,
+      "loss": 1.1497,
+      "step": 2089
+    },
+    {
+      "epoch": 0.12698219818944043,
+      "grad_norm": 0.2169427126646042,
+      "learning_rate": 9.623827013608351e-05,
+      "loss": 1.075,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12704295522206696,
+      "grad_norm": 0.2520436942577362,
+      "learning_rate": 9.623462648752648e-05,
+      "loss": 1.1826,
+      "step": 2091
+    },
+    {
+      "epoch": 0.12710371225469347,
+      "grad_norm": 0.16898399591445923,
+      "learning_rate": 9.623098114423024e-05,
+      "loss": 1.1824,
+      "step": 2092
+    },
+    {
+      "epoch": 0.12716446928732,
+      "grad_norm": 0.1813206523656845,
+      "learning_rate": 9.622733410632847e-05,
+      "loss": 1.1155,
+      "step": 2093
+    },
+    {
+      "epoch": 0.12722522631994654,
+      "grad_norm": 0.16200028359889984,
+      "learning_rate": 9.622368537395483e-05,
+      "loss": 1.1357,
+      "step": 2094
+    },
+    {
+      "epoch": 0.12728598335257307,
+      "grad_norm": 0.14974044263362885,
+      "learning_rate": 9.622003494724307e-05,
+      "loss": 1.067,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12734674038519958,
+      "grad_norm": 1.432485818862915,
+      "learning_rate": 9.621638282632698e-05,
+      "loss": 1.2379,
+      "step": 2096
+    },
+    {
+      "epoch": 0.12740749741782612,
+      "grad_norm": 0.38108500838279724,
+      "learning_rate": 9.621272901134046e-05,
+      "loss": 1.1973,
+      "step": 2097
+    },
+    {
+      "epoch": 0.12746825445045265,
+      "grad_norm": 0.19689838588237762,
+      "learning_rate": 9.620907350241742e-05,
+      "loss": 1.1828,
+      "step": 2098
+    },
+    {
+      "epoch": 0.12752901148307916,
+      "grad_norm": 0.17784243822097778,
+      "learning_rate": 9.620541629969187e-05,
+      "loss": 1.1185,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1275897685157057,
+      "grad_norm": 0.21143406629562378,
+      "learning_rate": 9.620175740329784e-05,
+      "loss": 1.2074,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12765052554833223,
+      "grad_norm": 0.18847842514514923,
+      "learning_rate": 9.619809681336947e-05,
+      "loss": 1.0765,
+      "step": 2101
+    },
+    {
+      "epoch": 0.12771128258095873,
+      "grad_norm": 1.8641302585601807,
+      "learning_rate": 9.619443453004092e-05,
+      "loss": 1.231,
+      "step": 2102
+    },
+    {
+      "epoch": 0.12777203961358527,
+      "grad_norm": 0.23724384605884552,
+      "learning_rate": 9.619077055344645e-05,
+      "loss": 1.1779,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1278327966462118,
+      "grad_norm": 0.8766135573387146,
+      "learning_rate": 9.618710488372035e-05,
+      "loss": 1.3835,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1278935536788383,
+      "grad_norm": 1.4700849056243896,
+      "learning_rate": 9.618343752099701e-05,
+      "loss": 1.2594,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12795431071146485,
+      "grad_norm": 0.22571195662021637,
+      "learning_rate": 9.617976846541084e-05,
+      "loss": 1.1178,
+      "step": 2106
+    },
+    {
+      "epoch": 0.12801506774409138,
+      "grad_norm": 0.21347974240779877,
+      "learning_rate": 9.617609771709632e-05,
+      "loss": 1.1699,
+      "step": 2107
+    },
+    {
+      "epoch": 0.12807582477671792,
+      "grad_norm": 0.2740815281867981,
+      "learning_rate": 9.617242527618802e-05,
+      "loss": 1.1872,
+      "step": 2108
+    },
+    {
+      "epoch": 0.12813658180934442,
+      "grad_norm": 0.1986178308725357,
+      "learning_rate": 9.616875114282053e-05,
+      "loss": 1.1545,
+      "step": 2109
+    },
+    {
+      "epoch": 0.12819733884197096,
+      "grad_norm": 0.30879130959510803,
+      "learning_rate": 9.616507531712855e-05,
+      "loss": 1.1211,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1282580958745975,
+      "grad_norm": 0.5088827013969421,
+      "learning_rate": 9.616139779924682e-05,
+      "loss": 1.2997,
+      "step": 2111
+    },
+    {
+      "epoch": 0.128318852907224,
+      "grad_norm": 0.33232057094573975,
+      "learning_rate": 9.615771858931013e-05,
+      "loss": 1.0808,
+      "step": 2112
+    },
+    {
+      "epoch": 0.12837960993985054,
+      "grad_norm": 8.197784423828125,
+      "learning_rate": 9.615403768745332e-05,
+      "loss": 1.2707,
+      "step": 2113
+    },
+    {
+      "epoch": 0.12844036697247707,
+      "grad_norm": 0.2203400433063507,
+      "learning_rate": 9.615035509381136e-05,
+      "loss": 1.1625,
+      "step": 2114
+    },
+    {
+      "epoch": 0.12850112400510358,
+      "grad_norm": 0.1596183031797409,
+      "learning_rate": 9.614667080851922e-05,
+      "loss": 1.1055,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1285618810377301,
+      "grad_norm": 0.30698826909065247,
+      "learning_rate": 9.614298483171192e-05,
+      "loss": 1.2276,
+      "step": 2116
+    },
+    {
+      "epoch": 0.12862263807035665,
+      "grad_norm": 0.32907792925834656,
+      "learning_rate": 9.61392971635246e-05,
+      "loss": 1.1992,
+      "step": 2117
+    },
+    {
+      "epoch": 0.12868339510298318,
+      "grad_norm": 0.15932609140872955,
+      "learning_rate": 9.613560780409243e-05,
+      "loss": 1.1416,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1287441521356097,
+      "grad_norm": 0.1595773547887802,
+      "learning_rate": 9.613191675355063e-05,
+      "loss": 1.1133,
+      "step": 2119
+    },
+    {
+      "epoch": 0.12880490916823623,
+      "grad_norm": 0.17283505201339722,
+      "learning_rate": 9.612822401203452e-05,
+      "loss": 1.1745,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12886566620086276,
+      "grad_norm": 0.19520869851112366,
+      "learning_rate": 9.612452957967943e-05,
+      "loss": 1.0721,
+      "step": 2121
+    },
+    {
+      "epoch": 0.12892642323348927,
+      "grad_norm": 0.1592271476984024,
+      "learning_rate": 9.612083345662078e-05,
+      "loss": 1.0574,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1289871802661158,
+      "grad_norm": 0.16798152029514313,
+      "learning_rate": 9.611713564299408e-05,
+      "loss": 1.2063,
+      "step": 2123
+    },
+    {
+      "epoch": 0.12904793729874234,
+      "grad_norm": 0.22492818534374237,
+      "learning_rate": 9.611343613893487e-05,
+      "loss": 1.2212,
+      "step": 2124
+    },
+    {
+      "epoch": 0.12910869433136885,
+      "grad_norm": 0.17656837403774261,
+      "learning_rate": 9.610973494457872e-05,
+      "loss": 1.1992,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12916945136399538,
+      "grad_norm": 0.45715394616127014,
+      "learning_rate": 9.610603206006133e-05,
+      "loss": 1.1674,
+      "step": 2126
+    },
+    {
+      "epoch": 0.12923020839662192,
+      "grad_norm": 0.26578259468078613,
+      "learning_rate": 9.610232748551844e-05,
+      "loss": 1.2901,
+      "step": 2127
+    },
+    {
+      "epoch": 0.12929096542924842,
+      "grad_norm": 0.8173021078109741,
+      "learning_rate": 9.60986212210858e-05,
+      "loss": 1.1727,
+      "step": 2128
+    },
+    {
+      "epoch": 0.12935172246187496,
+      "grad_norm": 0.43204405903816223,
+      "learning_rate": 9.60949132668993e-05,
+      "loss": 1.2081,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1294124794945015,
+      "grad_norm": 0.25303393602371216,
+      "learning_rate": 9.609120362309485e-05,
+      "loss": 1.078,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12947323652712803,
+      "grad_norm": 1.1481194496154785,
+      "learning_rate": 9.608749228980841e-05,
+      "loss": 1.2162,
+      "step": 2131
+    },
+    {
+      "epoch": 0.12953399355975453,
+      "grad_norm": 0.9371265172958374,
+      "learning_rate": 9.608377926717604e-05,
+      "loss": 1.1349,
+      "step": 2132
+    },
+    {
+      "epoch": 0.12959475059238107,
+      "grad_norm": 0.33399972319602966,
+      "learning_rate": 9.608006455533383e-05,
+      "loss": 1.2676,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1296555076250076,
+      "grad_norm": 0.1876756250858307,
+      "learning_rate": 9.607634815441795e-05,
+      "loss": 1.1356,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1297162646576341,
+      "grad_norm": 0.22510530054569244,
+      "learning_rate": 9.607263006456462e-05,
+      "loss": 1.1162,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12977702169026065,
+      "grad_norm": 0.1685524880886078,
+      "learning_rate": 9.606891028591013e-05,
+      "loss": 1.0784,
+      "step": 2136
+    },
+    {
+      "epoch": 0.12983777872288718,
+      "grad_norm": 0.1984611600637436,
+      "learning_rate": 9.606518881859083e-05,
+      "loss": 1.1671,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1298985357555137,
+      "grad_norm": 0.16086313128471375,
+      "learning_rate": 9.606146566274312e-05,
+      "loss": 1.1053,
+      "step": 2138
+    },
+    {
+      "epoch": 0.12995929278814022,
+      "grad_norm": 0.19444610178470612,
+      "learning_rate": 9.60577408185035e-05,
+      "loss": 1.1507,
+      "step": 2139
+    },
+    {
+      "epoch": 0.13002004982076676,
+      "grad_norm": 0.22640952467918396,
+      "learning_rate": 9.605401428600846e-05,
+      "loss": 1.0466,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1300808068533933,
+      "grad_norm": 0.5249062180519104,
+      "learning_rate": 9.605028606539466e-05,
+      "loss": 1.314,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1301415638860198,
+      "grad_norm": 0.2367764264345169,
+      "learning_rate": 9.60465561567987e-05,
+      "loss": 1.2181,
+      "step": 2142
+    },
+    {
+      "epoch": 0.13020232091864634,
+      "grad_norm": 0.4680385887622833,
+      "learning_rate": 9.604282456035733e-05,
+      "loss": 1.2948,
+      "step": 2143
+    },
+    {
+      "epoch": 0.13026307795127287,
+      "grad_norm": 0.18987536430358887,
+      "learning_rate": 9.603909127620732e-05,
+      "loss": 1.1559,
+      "step": 2144
+    },
+    {
+      "epoch": 0.13032383498389938,
+      "grad_norm": 0.9202283024787903,
+      "learning_rate": 9.603535630448552e-05,
+      "loss": 1.085,
+      "step": 2145
+    },
+    {
+      "epoch": 0.1303845920165259,
+      "grad_norm": 0.25142955780029297,
+      "learning_rate": 9.603161964532885e-05,
+      "loss": 1.1354,
+      "step": 2146
+    },
+    {
+      "epoch": 0.13044534904915245,
+      "grad_norm": 0.34282171726226807,
+      "learning_rate": 9.602788129887426e-05,
+      "loss": 1.2372,
+      "step": 2147
+    },
+    {
+      "epoch": 0.13050610608177896,
+      "grad_norm": 0.3733201026916504,
+      "learning_rate": 9.602414126525876e-05,
+      "loss": 1.2886,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1305668631144055,
+      "grad_norm": 0.2155725210905075,
+      "learning_rate": 9.60203995446195e-05,
+      "loss": 1.0738,
+      "step": 2149
+    },
+    {
+      "epoch": 0.13062762014703203,
+      "grad_norm": 0.2998858392238617,
+      "learning_rate": 9.601665613709357e-05,
+      "loss": 1.201,
+      "step": 2150
+    },
+    {
+      "epoch": 0.13068837717965853,
+      "grad_norm": 0.1856468766927719,
+      "learning_rate": 9.601291104281823e-05,
+      "loss": 1.1018,
+      "step": 2151
+    },
+    {
+      "epoch": 0.13074913421228507,
+      "grad_norm": 0.17658545076847076,
+      "learning_rate": 9.600916426193074e-05,
+      "loss": 1.1235,
+      "step": 2152
+    },
+    {
+      "epoch": 0.1308098912449116,
+      "grad_norm": 0.1622500866651535,
+      "learning_rate": 9.600541579456844e-05,
+      "loss": 1.1145,
+      "step": 2153
+    },
+    {
+      "epoch": 0.13087064827753814,
+      "grad_norm": 0.17708943784236908,
+      "learning_rate": 9.600166564086874e-05,
+      "loss": 1.0893,
+      "step": 2154
+    },
+    {
+      "epoch": 0.13093140531016464,
+      "grad_norm": 0.14311006665229797,
+      "learning_rate": 9.599791380096908e-05,
+      "loss": 1.1153,
+      "step": 2155
+    },
+    {
+      "epoch": 0.13099216234279118,
+      "grad_norm": 0.1791146993637085,
+      "learning_rate": 9.5994160275007e-05,
+      "loss": 1.2152,
+      "step": 2156
+    },
+    {
+      "epoch": 0.13105291937541771,
+      "grad_norm": 0.5301501750946045,
+      "learning_rate": 9.59904050631201e-05,
+      "loss": 1.1234,
+      "step": 2157
+    },
+    {
+      "epoch": 0.13111367640804422,
+      "grad_norm": 0.21681536734104156,
+      "learning_rate": 9.5986648165446e-05,
+      "loss": 1.2372,
+      "step": 2158
+    },
+    {
+      "epoch": 0.13117443344067076,
+      "grad_norm": 0.3805176317691803,
+      "learning_rate": 9.598288958212242e-05,
+      "loss": 1.1109,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1312351904732973,
+      "grad_norm": 0.19702747464179993,
+      "learning_rate": 9.597912931328714e-05,
+      "loss": 1.2659,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1312959475059238,
+      "grad_norm": 0.20077842473983765,
+      "learning_rate": 9.597536735907797e-05,
+      "loss": 1.1336,
+      "step": 2161
+    },
+    {
+      "epoch": 0.13135670453855033,
+      "grad_norm": 0.36625880002975464,
+      "learning_rate": 9.597160371963286e-05,
+      "loss": 1.2307,
+      "step": 2162
+    },
+    {
+      "epoch": 0.13141746157117687,
+      "grad_norm": 0.38198354840278625,
+      "learning_rate": 9.596783839508969e-05,
+      "loss": 1.2184,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1314782186038034,
+      "grad_norm": 0.5571491122245789,
+      "learning_rate": 9.596407138558653e-05,
+      "loss": 1.0653,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1315389756364299,
+      "grad_norm": 0.15978047251701355,
+      "learning_rate": 9.596030269126146e-05,
+      "loss": 1.1472,
+      "step": 2165
+    },
+    {
+      "epoch": 0.13159973266905645,
+      "grad_norm": 0.2738722562789917,
+      "learning_rate": 9.595653231225261e-05,
+      "loss": 1.2037,
+      "step": 2166
+    },
+    {
+      "epoch": 0.13166048970168298,
+      "grad_norm": 0.2074570208787918,
+      "learning_rate": 9.595276024869817e-05,
+      "loss": 1.1429,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1317212467343095,
+      "grad_norm": 0.29405421018600464,
+      "learning_rate": 9.594898650073644e-05,
+      "loss": 1.1024,
+      "step": 2168
+    },
+    {
+      "epoch": 0.13178200376693602,
+      "grad_norm": 0.21341866254806519,
+      "learning_rate": 9.594521106850571e-05,
+      "loss": 1.0728,
+      "step": 2169
+    },
+    {
+      "epoch": 0.13184276079956256,
+      "grad_norm": 0.1805810034275055,
+      "learning_rate": 9.594143395214441e-05,
+      "loss": 1.1444,
+      "step": 2170
+    },
+    {
+      "epoch": 0.13190351783218907,
+      "grad_norm": 0.2784469723701477,
+      "learning_rate": 9.593765515179095e-05,
+      "loss": 1.1289,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1319642748648156,
+      "grad_norm": 0.2613830268383026,
+      "learning_rate": 9.593387466758386e-05,
+      "loss": 1.1165,
+      "step": 2172
+    },
+    {
+      "epoch": 0.13202503189744214,
+      "grad_norm": 0.19762438535690308,
+      "learning_rate": 9.593009249966172e-05,
+      "loss": 1.0818,
+      "step": 2173
+    },
+    {
+      "epoch": 0.13208578893006864,
+      "grad_norm": 0.353282630443573,
+      "learning_rate": 9.592630864816317e-05,
+      "loss": 1.1068,
+      "step": 2174
+    },
+    {
+      "epoch": 0.13214654596269518,
+      "grad_norm": 0.1910041719675064,
+      "learning_rate": 9.59225231132269e-05,
+      "loss": 1.1169,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1322073029953217,
+      "grad_norm": 0.19785964488983154,
+      "learning_rate": 9.591873589499166e-05,
+      "loss": 1.2027,
+      "step": 2176
+    },
+    {
+      "epoch": 0.13226806002794825,
+      "grad_norm": 0.3330136835575104,
+      "learning_rate": 9.591494699359627e-05,
+      "loss": 1.0565,
+      "step": 2177
+    },
+    {
+      "epoch": 0.13232881706057475,
+      "grad_norm": 0.23309539258480072,
+      "learning_rate": 9.591115640917963e-05,
+      "loss": 1.3634,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1323895740932013,
+      "grad_norm": 0.743053138256073,
+      "learning_rate": 9.590736414188069e-05,
+      "loss": 1.2602,
+      "step": 2179
+    },
+    {
+      "epoch": 0.13245033112582782,
+      "grad_norm": 0.16662898659706116,
+      "learning_rate": 9.590357019183843e-05,
+      "loss": 1.119,
+      "step": 2180
+    },
+    {
+      "epoch": 0.13251108815845433,
+      "grad_norm": 0.4592370390892029,
+      "learning_rate": 9.589977455919193e-05,
+      "loss": 1.099,
+      "step": 2181
+    },
+    {
+      "epoch": 0.13257184519108087,
+      "grad_norm": 0.2229791283607483,
+      "learning_rate": 9.589597724408032e-05,
+      "loss": 1.1117,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1326326022237074,
+      "grad_norm": 0.22165454924106598,
+      "learning_rate": 9.589217824664281e-05,
+      "loss": 1.0925,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1326933592563339,
+      "grad_norm": 0.2257971316576004,
+      "learning_rate": 9.588837756701861e-05,
+      "loss": 1.2189,
+      "step": 2184
+    },
+    {
+      "epoch": 0.13275411628896044,
+      "grad_norm": 0.21456827223300934,
+      "learning_rate": 9.588457520534708e-05,
+      "loss": 1.1333,
+      "step": 2185
+    },
+    {
+      "epoch": 0.13281487332158698,
+      "grad_norm": 0.18152987957000732,
+      "learning_rate": 9.588077116176756e-05,
+      "loss": 1.1679,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1328756303542135,
+      "grad_norm": 0.8997235298156738,
+      "learning_rate": 9.587696543641953e-05,
+      "loss": 1.2548,
+      "step": 2187
+    },
+    {
+      "epoch": 0.13293638738684002,
+      "grad_norm": 0.30472829937934875,
+      "learning_rate": 9.587315802944245e-05,
+      "loss": 1.2025,
+      "step": 2188
+    },
+    {
+      "epoch": 0.13299714441946656,
+      "grad_norm": 0.24143661558628082,
+      "learning_rate": 9.58693489409759e-05,
+      "loss": 1.1173,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1330579014520931,
+      "grad_norm": 0.24073444306850433,
+      "learning_rate": 9.586553817115948e-05,
+      "loss": 1.0794,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1331186584847196,
+      "grad_norm": 0.1910742074251175,
+      "learning_rate": 9.586172572013291e-05,
+      "loss": 1.1332,
+      "step": 2191
+    },
+    {
+      "epoch": 0.13317941551734613,
+      "grad_norm": 0.19061161577701569,
+      "learning_rate": 9.585791158803592e-05,
+      "loss": 1.1671,
+      "step": 2192
+    },
+    {
+      "epoch": 0.13324017254997267,
+      "grad_norm": 0.23281781375408173,
+      "learning_rate": 9.58540957750083e-05,
+      "loss": 1.1418,
+      "step": 2193
+    },
+    {
+      "epoch": 0.13330092958259918,
+      "grad_norm": 0.1959117352962494,
+      "learning_rate": 9.585027828118995e-05,
+      "loss": 1.125,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1333616866152257,
+      "grad_norm": 0.13621953129768372,
+      "learning_rate": 9.584645910672078e-05,
+      "loss": 1.0938,
+      "step": 2195
+    },
+    {
+      "epoch": 0.13342244364785225,
+      "grad_norm": 0.31365418434143066,
+      "learning_rate": 9.58426382517408e-05,
+      "loss": 1.1188,
+      "step": 2196
+    },
+    {
+      "epoch": 0.13348320068047875,
+      "grad_norm": 0.2360733449459076,
+      "learning_rate": 9.583881571639005e-05,
+      "loss": 1.2033,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1335439577131053,
+      "grad_norm": 0.5398380160331726,
+      "learning_rate": 9.583499150080865e-05,
+      "loss": 1.1213,
+      "step": 2198
+    },
+    {
+      "epoch": 0.13360471474573182,
+      "grad_norm": 0.2268095314502716,
+      "learning_rate": 9.583116560513678e-05,
+      "loss": 1.0736,
+      "step": 2199
+    },
+    {
+      "epoch": 0.13366547177835836,
+      "grad_norm": 0.2760641276836395,
+      "learning_rate": 9.582733802951467e-05,
+      "loss": 1.1001,
+      "step": 2200
+    },
+    {
+      "epoch": 0.13372622881098487,
+      "grad_norm": 0.2718813419342041,
+      "learning_rate": 9.582350877408262e-05,
+      "loss": 1.1135,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1337869858436114,
+      "grad_norm": 0.16864360868930817,
+      "learning_rate": 9.5819677838981e-05,
+      "loss": 1.1534,
+      "step": 2202
+    },
+    {
+      "epoch": 0.13384774287623794,
+      "grad_norm": 0.1424243301153183,
+      "learning_rate": 9.581584522435024e-05,
+      "loss": 1.0753,
+      "step": 2203
+    },
+    {
+      "epoch": 0.13390849990886444,
+      "grad_norm": 0.33547699451446533,
+      "learning_rate": 9.581201093033082e-05,
+      "loss": 1.1696,
+      "step": 2204
+    },
+    {
+      "epoch": 0.13396925694149098,
+      "grad_norm": 0.1380401849746704,
+      "learning_rate": 9.580817495706327e-05,
+      "loss": 1.0863,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1340300139741175,
+      "grad_norm": 0.3057708740234375,
+      "learning_rate": 9.580433730468823e-05,
+      "loss": 1.1465,
+      "step": 2206
+    },
+    {
+      "epoch": 0.13409077100674402,
+      "grad_norm": 0.16703514754772186,
+      "learning_rate": 9.580049797334634e-05,
+      "loss": 1.1367,
+      "step": 2207
+    },
+    {
+      "epoch": 0.13415152803937055,
+      "grad_norm": 0.2826164960861206,
+      "learning_rate": 9.579665696317835e-05,
+      "loss": 1.2703,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1342122850719971,
+      "grad_norm": 0.19026175141334534,
+      "learning_rate": 9.579281427432504e-05,
+      "loss": 1.1272,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1342730421046236,
+      "grad_norm": 0.33325132727622986,
+      "learning_rate": 9.578896990692727e-05,
+      "loss": 1.1676,
+      "step": 2210
+    },
+    {
+      "epoch": 0.13433379913725013,
+      "grad_norm": 0.2609262764453888,
+      "learning_rate": 9.578512386112596e-05,
+      "loss": 1.0918,
+      "step": 2211
+    },
+    {
+      "epoch": 0.13439455616987667,
+      "grad_norm": 0.26915979385375977,
+      "learning_rate": 9.578127613706209e-05,
+      "loss": 1.1017,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1344553132025032,
+      "grad_norm": 0.3109450936317444,
+      "learning_rate": 9.577742673487668e-05,
+      "loss": 1.1867,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1345160702351297,
+      "grad_norm": 0.38755548000335693,
+      "learning_rate": 9.577357565471086e-05,
+      "loss": 1.1065,
+      "step": 2214
+    },
+    {
+      "epoch": 0.13457682726775624,
+      "grad_norm": 0.18848983943462372,
+      "learning_rate": 9.576972289670578e-05,
+      "loss": 1.1081,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13463758430038278,
+      "grad_norm": 0.5497241020202637,
+      "learning_rate": 9.576586846100264e-05,
+      "loss": 1.1253,
+      "step": 2216
+    },
+    {
+      "epoch": 0.13469834133300929,
+      "grad_norm": 0.3082449436187744,
+      "learning_rate": 9.576201234774275e-05,
+      "loss": 1.0867,
+      "step": 2217
+    },
+    {
+      "epoch": 0.13475909836563582,
+      "grad_norm": 0.26647523045539856,
+      "learning_rate": 9.575815455706748e-05,
+      "loss": 1.0775,
+      "step": 2218
+    },
+    {
+      "epoch": 0.13481985539826236,
+      "grad_norm": 0.30094531178474426,
+      "learning_rate": 9.575429508911817e-05,
+      "loss": 1.198,
+      "step": 2219
+    },
+    {
+      "epoch": 0.13488061243088886,
+      "grad_norm": 0.33741235733032227,
+      "learning_rate": 9.575043394403635e-05,
+      "loss": 1.2459,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1349413694635154,
+      "grad_norm": 0.16895785927772522,
+      "learning_rate": 9.574657112196354e-05,
+      "loss": 1.1017,
+      "step": 2221
+    },
+    {
+      "epoch": 0.13500212649614193,
+      "grad_norm": 5.2270708084106445,
+      "learning_rate": 9.57427066230413e-05,
+      "loss": 1.0723,
+      "step": 2222
+    },
+    {
+      "epoch": 0.13506288352876847,
+      "grad_norm": 0.4301412105560303,
+      "learning_rate": 9.573884044741133e-05,
+      "loss": 1.1206,
+      "step": 2223
+    },
+    {
+      "epoch": 0.13512364056139498,
+      "grad_norm": 1.5419443845748901,
+      "learning_rate": 9.57349725952153e-05,
+      "loss": 1.1332,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1351843975940215,
+      "grad_norm": 0.25238433480262756,
+      "learning_rate": 9.573110306659504e-05,
+      "loss": 1.1976,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13524515462664805,
+      "grad_norm": 0.5974071025848389,
+      "learning_rate": 9.572723186169232e-05,
+      "loss": 1.1163,
+      "step": 2226
+    },
+    {
+      "epoch": 0.13530591165927455,
+      "grad_norm": 0.6074419617652893,
+      "learning_rate": 9.57233589806491e-05,
+      "loss": 1.0886,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1353666686919011,
+      "grad_norm": 0.3090222179889679,
+      "learning_rate": 9.57194844236073e-05,
+      "loss": 1.0706,
+      "step": 2228
+    },
+    {
+      "epoch": 0.13542742572452762,
+      "grad_norm": 0.35570967197418213,
+      "learning_rate": 9.571560819070898e-05,
+      "loss": 1.1781,
+      "step": 2229
+    },
+    {
+      "epoch": 0.13548818275715413,
+      "grad_norm": 0.165644571185112,
+      "learning_rate": 9.57117302820962e-05,
+      "loss": 1.1643,
+      "step": 2230
+    },
+    {
+      "epoch": 0.13554893978978066,
+      "grad_norm": 6.318443775177002,
+      "learning_rate": 9.57078506979111e-05,
+      "loss": 1.1198,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1356096968224072,
+      "grad_norm": 0.587719202041626,
+      "learning_rate": 9.570396943829592e-05,
+      "loss": 1.1641,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1356704538550337,
+      "grad_norm": 0.34268298745155334,
+      "learning_rate": 9.570008650339288e-05,
+      "loss": 1.1181,
+      "step": 2233
+    },
+    {
+      "epoch": 0.13573121088766024,
+      "grad_norm": 0.49453502893447876,
+      "learning_rate": 9.569620189334435e-05,
+      "loss": 1.2803,
+      "step": 2234
+    },
+    {
+      "epoch": 0.13579196792028678,
+      "grad_norm": 0.7012412548065186,
+      "learning_rate": 9.56923156082927e-05,
+      "loss": 1.1048,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1358527249529133,
+      "grad_norm": 0.8635398745536804,
+      "learning_rate": 9.568842764838041e-05,
+      "loss": 1.0926,
+      "step": 2236
+    },
+    {
+      "epoch": 0.13591348198553982,
+      "grad_norm": 0.42028212547302246,
+      "learning_rate": 9.568453801374995e-05,
+      "loss": 1.2891,
+      "step": 2237
+    },
+    {
+      "epoch": 0.13597423901816635,
+      "grad_norm": 0.851941704750061,
+      "learning_rate": 9.568064670454393e-05,
+      "loss": 1.0685,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1360349960507929,
+      "grad_norm": 0.20819708704948425,
+      "learning_rate": 9.567675372090497e-05,
+      "loss": 1.1275,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1360957530834194,
+      "grad_norm": 0.4963993430137634,
+      "learning_rate": 9.567285906297577e-05,
+      "loss": 1.2704,
+      "step": 2240
+    },
+    {
+      "epoch": 0.13615651011604593,
+      "grad_norm": 0.5887370109558105,
+      "learning_rate": 9.56689627308991e-05,
+      "loss": 1.1956,
+      "step": 2241
+    },
+    {
+      "epoch": 0.13621726714867247,
+      "grad_norm": 0.33166342973709106,
+      "learning_rate": 9.566506472481778e-05,
+      "loss": 1.0992,
+      "step": 2242
+    },
+    {
+      "epoch": 0.13627802418129897,
+      "grad_norm": 0.3141010105609894,
+      "learning_rate": 9.566116504487466e-05,
+      "loss": 1.1705,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1363387812139255,
+      "grad_norm": 0.34887564182281494,
+      "learning_rate": 9.565726369121273e-05,
+      "loss": 1.1057,
+      "step": 2244
+    },
+    {
+      "epoch": 0.13639953824655204,
+      "grad_norm": 0.2421671450138092,
+      "learning_rate": 9.565336066397496e-05,
+      "loss": 1.0938,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13646029527917858,
+      "grad_norm": 0.41549810767173767,
+      "learning_rate": 9.564945596330446e-05,
+      "loss": 1.3261,
+      "step": 2246
+    },
+    {
+      "epoch": 0.13652105231180509,
+      "grad_norm": 0.44896894693374634,
+      "learning_rate": 9.564554958934431e-05,
+      "loss": 1.2324,
+      "step": 2247
+    },
+    {
+      "epoch": 0.13658180934443162,
+      "grad_norm": 0.46312132477760315,
+      "learning_rate": 9.56416415422377e-05,
+      "loss": 1.1076,
+      "step": 2248
+    },
+    {
+      "epoch": 0.13664256637705816,
+      "grad_norm": 0.22303961217403412,
+      "learning_rate": 9.563773182212791e-05,
+      "loss": 1.1594,
+      "step": 2249
+    },
+    {
+      "epoch": 0.13670332340968466,
+      "grad_norm": 0.5063765645027161,
+      "learning_rate": 9.563382042915826e-05,
+      "loss": 1.0987,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1367640804423112,
+      "grad_norm": 0.31590166687965393,
+      "learning_rate": 9.562990736347207e-05,
+      "loss": 1.1698,
+      "step": 2251
+    },
+    {
+      "epoch": 0.13682483747493773,
+      "grad_norm": 0.23123250901699066,
+      "learning_rate": 9.562599262521281e-05,
+      "loss": 1.1019,
+      "step": 2252
+    },
+    {
+      "epoch": 0.13688559450756424,
+      "grad_norm": 0.4999677538871765,
+      "learning_rate": 9.562207621452398e-05,
+      "loss": 1.2709,
+      "step": 2253
+    },
+    {
+      "epoch": 0.13694635154019077,
+      "grad_norm": 0.24811404943466187,
+      "learning_rate": 9.561815813154913e-05,
+      "loss": 1.1109,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1370071085728173,
+      "grad_norm": 0.16323330998420715,
+      "learning_rate": 9.561423837643186e-05,
+      "loss": 1.1211,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13706786560544382,
+      "grad_norm": 0.6281250715255737,
+      "learning_rate": 9.561031694931589e-05,
+      "loss": 1.1062,
+      "step": 2256
+    },
+    {
+      "epoch": 0.13712862263807035,
+      "grad_norm": 0.24023033678531647,
+      "learning_rate": 9.560639385034491e-05,
+      "loss": 1.1081,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1371893796706969,
+      "grad_norm": 0.21369919180870056,
+      "learning_rate": 9.560246907966276e-05,
+      "loss": 1.0844,
+      "step": 2258
+    },
+    {
+      "epoch": 0.13725013670332342,
+      "grad_norm": 0.2748388350009918,
+      "learning_rate": 9.55985426374133e-05,
+      "loss": 1.0781,
+      "step": 2259
+    },
+    {
+      "epoch": 0.13731089373594993,
+      "grad_norm": 0.23953093588352203,
+      "learning_rate": 9.559461452374043e-05,
+      "loss": 1.1211,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13737165076857646,
+      "grad_norm": 0.26806795597076416,
+      "learning_rate": 9.559068473878816e-05,
+      "loss": 1.0397,
+      "step": 2261
+    },
+    {
+      "epoch": 0.137432407801203,
+      "grad_norm": 0.5642307996749878,
+      "learning_rate": 9.558675328270053e-05,
+      "loss": 1.2112,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1374931648338295,
+      "grad_norm": 0.23219093680381775,
+      "learning_rate": 9.558282015562165e-05,
+      "loss": 1.0983,
+      "step": 2263
+    },
+    {
+      "epoch": 0.13755392186645604,
+      "grad_norm": 0.4778085947036743,
+      "learning_rate": 9.557888535769568e-05,
+      "loss": 1.0925,
+      "step": 2264
+    },
+    {
+      "epoch": 0.13761467889908258,
+      "grad_norm": 0.24083127081394196,
+      "learning_rate": 9.557494888906685e-05,
+      "loss": 1.095,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13767543593170908,
+      "grad_norm": 0.16870884597301483,
+      "learning_rate": 9.557101074987946e-05,
+      "loss": 1.1156,
+      "step": 2266
+    },
+    {
+      "epoch": 0.13773619296433562,
+      "grad_norm": 0.15064562857151031,
+      "learning_rate": 9.556707094027786e-05,
+      "loss": 1.1584,
+      "step": 2267
+    },
+    {
+      "epoch": 0.13779694999696215,
+      "grad_norm": 0.24823467433452606,
+      "learning_rate": 9.556312946040649e-05,
+      "loss": 1.3144,
+      "step": 2268
+    },
+    {
+      "epoch": 0.1378577070295887,
+      "grad_norm": 0.15248197317123413,
+      "learning_rate": 9.555918631040979e-05,
+      "loss": 1.1203,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1379184640622152,
+      "grad_norm": 0.28081315755844116,
+      "learning_rate": 9.55552414904323e-05,
+      "loss": 1.1326,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13797922109484173,
+      "grad_norm": 0.16592133045196533,
+      "learning_rate": 9.555129500061864e-05,
+      "loss": 1.1351,
+      "step": 2271
+    },
+    {
+      "epoch": 0.13803997812746827,
+      "grad_norm": 0.24708499014377594,
+      "learning_rate": 9.554734684111345e-05,
+      "loss": 1.1884,
+      "step": 2272
+    },
+    {
+      "epoch": 0.13810073516009477,
+      "grad_norm": 0.1949533224105835,
+      "learning_rate": 9.554339701206145e-05,
+      "loss": 1.2322,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1381614921927213,
+      "grad_norm": 0.17027607560157776,
+      "learning_rate": 9.553944551360745e-05,
+      "loss": 1.1543,
+      "step": 2274
+    },
+    {
+      "epoch": 0.13822224922534784,
+      "grad_norm": 0.38415852189064026,
+      "learning_rate": 9.553549234589624e-05,
+      "loss": 1.2894,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13828300625797435,
+      "grad_norm": 0.31658172607421875,
+      "learning_rate": 9.553153750907279e-05,
+      "loss": 1.105,
+      "step": 2276
+    },
+    {
+      "epoch": 0.13834376329060089,
+      "grad_norm": 0.22498324513435364,
+      "learning_rate": 9.552758100328202e-05,
+      "loss": 1.1804,
+      "step": 2277
+    },
+    {
+      "epoch": 0.13840452032322742,
+      "grad_norm": 0.5629101395606995,
+      "learning_rate": 9.552362282866896e-05,
+      "loss": 1.0719,
+      "step": 2278
+    },
+    {
+      "epoch": 0.13846527735585393,
+      "grad_norm": 0.3577674925327301,
+      "learning_rate": 9.55196629853787e-05,
+      "loss": 1.1355,
+      "step": 2279
+    },
+    {
+      "epoch": 0.13852603438848046,
+      "grad_norm": 0.16320045292377472,
+      "learning_rate": 9.551570147355642e-05,
+      "loss": 1.2234,
+      "step": 2280
+    },
+    {
+      "epoch": 0.138586791421107,
+      "grad_norm": 0.2317439317703247,
+      "learning_rate": 9.55117382933473e-05,
+      "loss": 1.1686,
+      "step": 2281
+    },
+    {
+      "epoch": 0.13864754845373353,
+      "grad_norm": 0.4535728693008423,
+      "learning_rate": 9.55077734448966e-05,
+      "loss": 1.2012,
+      "step": 2282
+    },
+    {
+      "epoch": 0.13870830548636004,
+      "grad_norm": 0.19711743295192719,
+      "learning_rate": 9.550380692834967e-05,
+      "loss": 1.1153,
+      "step": 2283
+    },
+    {
+      "epoch": 0.13876906251898657,
+      "grad_norm": 0.3493136465549469,
+      "learning_rate": 9.549983874385191e-05,
+      "loss": 1.1343,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1388298195516131,
+      "grad_norm": 1.0998477935791016,
+      "learning_rate": 9.549586889154875e-05,
+      "loss": 1.1199,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13889057658423962,
+      "grad_norm": 0.2433186024427414,
+      "learning_rate": 9.549189737158574e-05,
+      "loss": 1.0883,
+      "step": 2286
+    },
+    {
+      "epoch": 0.13895133361686615,
+      "grad_norm": 0.3650475740432739,
+      "learning_rate": 9.548792418410844e-05,
+      "loss": 1.1727,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1390120906494927,
+      "grad_norm": 0.49776846170425415,
+      "learning_rate": 9.548394932926248e-05,
+      "loss": 1.2129,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1390728476821192,
+      "grad_norm": 0.34604546427726746,
+      "learning_rate": 9.547997280719355e-05,
+      "loss": 1.1886,
+      "step": 2289
+    },
+    {
+      "epoch": 0.13913360471474573,
+      "grad_norm": 0.17261995375156403,
+      "learning_rate": 9.547599461804743e-05,
+      "loss": 1.1657,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13919436174737226,
+      "grad_norm": 0.1966523379087448,
+      "learning_rate": 9.547201476196996e-05,
+      "loss": 1.1179,
+      "step": 2291
+    },
+    {
+      "epoch": 0.13925511877999877,
+      "grad_norm": 0.2130250483751297,
+      "learning_rate": 9.546803323910698e-05,
+      "loss": 1.0747,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1393158758126253,
+      "grad_norm": 0.8317158222198486,
+      "learning_rate": 9.546405004960445e-05,
+      "loss": 1.0561,
+      "step": 2293
+    },
+    {
+      "epoch": 0.13937663284525184,
+      "grad_norm": 0.15411683917045593,
+      "learning_rate": 9.546006519360839e-05,
+      "loss": 1.0869,
+      "step": 2294
+    },
+    {
+      "epoch": 0.13943738987787838,
+      "grad_norm": 1.1167824268341064,
+      "learning_rate": 9.545607867126484e-05,
+      "loss": 1.1704,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13949814691050488,
+      "grad_norm": 0.9352914690971375,
+      "learning_rate": 9.545209048271997e-05,
+      "loss": 1.1141,
+      "step": 2296
+    },
+    {
+      "epoch": 0.13955890394313142,
+      "grad_norm": 0.3974536061286926,
+      "learning_rate": 9.54481006281199e-05,
+      "loss": 1.2302,
+      "step": 2297
+    },
+    {
+      "epoch": 0.13961966097575795,
+      "grad_norm": 0.8662892580032349,
+      "learning_rate": 9.544410910761094e-05,
+      "loss": 1.15,
+      "step": 2298
+    },
+    {
+      "epoch": 0.13968041800838446,
+      "grad_norm": 0.1656741201877594,
+      "learning_rate": 9.544011592133936e-05,
+      "loss": 1.0683,
+      "step": 2299
+    },
+    {
+      "epoch": 0.139741175041011,
+      "grad_norm": 0.5774582028388977,
+      "learning_rate": 9.543612106945157e-05,
+      "loss": 1.076,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13980193207363753,
+      "grad_norm": 0.24073489010334015,
+      "learning_rate": 9.543212455209396e-05,
+      "loss": 1.0723,
+      "step": 2301
+    },
+    {
+      "epoch": 0.13986268910626404,
+      "grad_norm": 0.6168996691703796,
+      "learning_rate": 9.542812636941306e-05,
+      "loss": 1.099,
+      "step": 2302
+    },
+    {
+      "epoch": 0.13992344613889057,
+      "grad_norm": 0.3081745207309723,
+      "learning_rate": 9.542412652155539e-05,
+      "loss": 1.1008,
+      "step": 2303
+    },
+    {
+      "epoch": 0.1399842031715171,
+      "grad_norm": 0.1635631024837494,
+      "learning_rate": 9.542012500866759e-05,
+      "loss": 1.1042,
+      "step": 2304
+    },
+    {
+      "epoch": 0.14004496020414364,
+      "grad_norm": 0.36980193853378296,
+      "learning_rate": 9.541612183089632e-05,
+      "loss": 1.191,
+      "step": 2305
+    },
+    {
+      "epoch": 0.14010571723677015,
+      "grad_norm": 0.277229905128479,
+      "learning_rate": 9.541211698838834e-05,
+      "loss": 1.056,
+      "step": 2306
+    },
+    {
+      "epoch": 0.14016647426939668,
+      "grad_norm": 0.20339277386665344,
+      "learning_rate": 9.540811048129041e-05,
+      "loss": 1.1423,
+      "step": 2307
+    },
+    {
+      "epoch": 0.14022723130202322,
+      "grad_norm": 0.21004325151443481,
+      "learning_rate": 9.540410230974943e-05,
+      "loss": 1.3987,
+      "step": 2308
+    },
+    {
+      "epoch": 0.14028798833464973,
+      "grad_norm": 0.30993664264678955,
+      "learning_rate": 9.54000924739123e-05,
+      "loss": 1.1818,
+      "step": 2309
+    },
+    {
+      "epoch": 0.14034874536727626,
+      "grad_norm": 1.4932667016983032,
+      "learning_rate": 9.539608097392602e-05,
+      "loss": 1.1422,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1404095023999028,
+      "grad_norm": 0.3883330523967743,
+      "learning_rate": 9.539206780993761e-05,
+      "loss": 1.1709,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1404702594325293,
+      "grad_norm": 0.24175414443016052,
+      "learning_rate": 9.538805298209416e-05,
+      "loss": 1.129,
+      "step": 2312
+    },
+    {
+      "epoch": 0.14053101646515584,
+      "grad_norm": 0.2146836668252945,
+      "learning_rate": 9.538403649054288e-05,
+      "loss": 1.0827,
+      "step": 2313
+    },
+    {
+      "epoch": 0.14059177349778237,
+      "grad_norm": 0.3535554111003876,
+      "learning_rate": 9.538001833543095e-05,
+      "loss": 1.1411,
+      "step": 2314
+    },
+    {
+      "epoch": 0.14065253053040888,
+      "grad_norm": 0.18057365715503693,
+      "learning_rate": 9.537599851690569e-05,
+      "loss": 1.1256,
+      "step": 2315
+    },
+    {
+      "epoch": 0.14071328756303542,
+      "grad_norm": 0.3745160400867462,
+      "learning_rate": 9.537197703511443e-05,
+      "loss": 1.1264,
+      "step": 2316
+    },
+    {
+      "epoch": 0.14077404459566195,
+      "grad_norm": 0.5218973159790039,
+      "learning_rate": 9.536795389020457e-05,
+      "loss": 1.2895,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1408348016282885,
+      "grad_norm": 0.17800843715667725,
+      "learning_rate": 9.536392908232361e-05,
+      "loss": 1.1102,
+      "step": 2318
+    },
+    {
+      "epoch": 0.140895558660915,
+      "grad_norm": 0.22896233201026917,
+      "learning_rate": 9.535990261161904e-05,
+      "loss": 1.1925,
+      "step": 2319
+    },
+    {
+      "epoch": 0.14095631569354153,
+      "grad_norm": 0.34376412630081177,
+      "learning_rate": 9.535587447823849e-05,
+      "loss": 1.1925,
+      "step": 2320
+    },
+    {
+      "epoch": 0.14101707272616806,
+      "grad_norm": 0.147017240524292,
+      "learning_rate": 9.535184468232958e-05,
+      "loss": 1.132,
+      "step": 2321
+    },
+    {
+      "epoch": 0.14107782975879457,
+      "grad_norm": 0.3092877268791199,
+      "learning_rate": 9.534781322404005e-05,
+      "loss": 1.1574,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1411385867914211,
+      "grad_norm": 0.15649278461933136,
+      "learning_rate": 9.534378010351766e-05,
+      "loss": 1.066,
+      "step": 2323
+    },
+    {
+      "epoch": 0.14119934382404764,
+      "grad_norm": 0.21558749675750732,
+      "learning_rate": 9.533974532091023e-05,
+      "loss": 1.0897,
+      "step": 2324
+    },
+    {
+      "epoch": 0.14126010085667415,
+      "grad_norm": 0.2511991560459137,
+      "learning_rate": 9.53357088763657e-05,
+      "loss": 1.1565,
+      "step": 2325
+    },
+    {
+      "epoch": 0.14132085788930068,
+      "grad_norm": 0.3122481405735016,
+      "learning_rate": 9.533167077003198e-05,
+      "loss": 1.0551,
+      "step": 2326
+    },
+    {
+      "epoch": 0.14138161492192722,
+      "grad_norm": 0.23617467284202576,
+      "learning_rate": 9.532763100205709e-05,
+      "loss": 1.1767,
+      "step": 2327
+    },
+    {
+      "epoch": 0.14144237195455375,
+      "grad_norm": 0.26131677627563477,
+      "learning_rate": 9.532358957258915e-05,
+      "loss": 1.208,
+      "step": 2328
+    },
+    {
+      "epoch": 0.14150312898718026,
+      "grad_norm": 0.1727125346660614,
+      "learning_rate": 9.531954648177625e-05,
+      "loss": 1.0769,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1415638860198068,
+      "grad_norm": 2.5413877964019775,
+      "learning_rate": 9.531550172976664e-05,
+      "loss": 1.2453,
+      "step": 2330
+    },
+    {
+      "epoch": 0.14162464305243333,
+      "grad_norm": 0.22673800587654114,
+      "learning_rate": 9.531145531670853e-05,
+      "loss": 1.1197,
+      "step": 2331
+    },
+    {
+      "epoch": 0.14168540008505984,
+      "grad_norm": 0.16913120448589325,
+      "learning_rate": 9.530740724275028e-05,
+      "loss": 1.1916,
+      "step": 2332
+    },
+    {
+      "epoch": 0.14174615711768637,
+      "grad_norm": 0.31697878241539,
+      "learning_rate": 9.530335750804025e-05,
+      "loss": 1.2311,
+      "step": 2333
+    },
+    {
+      "epoch": 0.1418069141503129,
+      "grad_norm": 0.26434314250946045,
+      "learning_rate": 9.52993061127269e-05,
+      "loss": 1.1003,
+      "step": 2334
+    },
+    {
+      "epoch": 0.14186767118293941,
+      "grad_norm": 0.4128108322620392,
+      "learning_rate": 9.529525305695874e-05,
+      "loss": 1.1663,
+      "step": 2335
+    },
+    {
+      "epoch": 0.14192842821556595,
+      "grad_norm": 0.6479545831680298,
+      "learning_rate": 9.529119834088432e-05,
+      "loss": 1.0976,
+      "step": 2336
+    },
+    {
+      "epoch": 0.14198918524819248,
+      "grad_norm": 0.5405136942863464,
+      "learning_rate": 9.528714196465224e-05,
+      "loss": 1.2017,
+      "step": 2337
+    },
+    {
+      "epoch": 0.142049942280819,
+      "grad_norm": 0.7159753441810608,
+      "learning_rate": 9.528308392841126e-05,
+      "loss": 1.2827,
+      "step": 2338
+    },
+    {
+      "epoch": 0.14211069931344553,
+      "grad_norm": 0.26535531878471375,
+      "learning_rate": 9.527902423231005e-05,
+      "loss": 1.1856,
+      "step": 2339
+    },
+    {
+      "epoch": 0.14217145634607206,
+      "grad_norm": 1.685245156288147,
+      "learning_rate": 9.527496287649748e-05,
+      "loss": 1.1421,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1422322133786986,
+      "grad_norm": 0.21378839015960693,
+      "learning_rate": 9.527089986112239e-05,
+      "loss": 1.1778,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1422929704113251,
+      "grad_norm": 0.22931383550167084,
+      "learning_rate": 9.52668351863337e-05,
+      "loss": 1.1815,
+      "step": 2342
+    },
+    {
+      "epoch": 0.14235372744395164,
+      "grad_norm": 0.2110852152109146,
+      "learning_rate": 9.526276885228044e-05,
+      "loss": 1.0991,
+      "step": 2343
+    },
+    {
+      "epoch": 0.14241448447657817,
+      "grad_norm": 0.18488439917564392,
+      "learning_rate": 9.525870085911162e-05,
+      "loss": 1.3852,
+      "step": 2344
+    },
+    {
+      "epoch": 0.14247524150920468,
+      "grad_norm": 0.1488366723060608,
+      "learning_rate": 9.525463120697637e-05,
+      "loss": 1.0614,
+      "step": 2345
+    },
+    {
+      "epoch": 0.14253599854183122,
+      "grad_norm": 0.19828559458255768,
+      "learning_rate": 9.525055989602388e-05,
+      "loss": 1.298,
+      "step": 2346
+    },
+    {
+      "epoch": 0.14259675557445775,
+      "grad_norm": 0.16060969233512878,
+      "learning_rate": 9.524648692640336e-05,
+      "loss": 1.1147,
+      "step": 2347
+    },
+    {
+      "epoch": 0.14265751260708426,
+      "grad_norm": 1.2450884580612183,
+      "learning_rate": 9.524241229826411e-05,
+      "loss": 1.1099,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1427182696397108,
+      "grad_norm": 0.28205931186676025,
+      "learning_rate": 9.523833601175551e-05,
+      "loss": 1.0847,
+      "step": 2349
+    },
+    {
+      "epoch": 0.14277902667233733,
+      "grad_norm": 0.206913024187088,
+      "learning_rate": 9.523425806702695e-05,
+      "loss": 1.0987,
+      "step": 2350
+    },
+    {
+      "epoch": 0.14283978370496386,
+      "grad_norm": 0.16962188482284546,
+      "learning_rate": 9.523017846422793e-05,
+      "loss": 1.1007,
+      "step": 2351
+    },
+    {
+      "epoch": 0.14290054073759037,
+      "grad_norm": 0.3351297676563263,
+      "learning_rate": 9.522609720350795e-05,
+      "loss": 1.2933,
+      "step": 2352
+    },
+    {
+      "epoch": 0.1429612977702169,
+      "grad_norm": 0.26097571849823,
+      "learning_rate": 9.522201428501664e-05,
+      "loss": 1.0871,
+      "step": 2353
+    },
+    {
+      "epoch": 0.14302205480284344,
+      "grad_norm": 0.23117025196552277,
+      "learning_rate": 9.521792970890366e-05,
+      "loss": 1.1786,
+      "step": 2354
+    },
+    {
+      "epoch": 0.14308281183546995,
+      "grad_norm": 0.29111966490745544,
+      "learning_rate": 9.521384347531874e-05,
+      "loss": 1.2099,
+      "step": 2355
+    },
+    {
+      "epoch": 0.14314356886809648,
+      "grad_norm": 0.18806366622447968,
+      "learning_rate": 9.520975558441164e-05,
+      "loss": 1.0685,
+      "step": 2356
+    },
+    {
+      "epoch": 0.14320432590072302,
+      "grad_norm": 0.1590515524148941,
+      "learning_rate": 9.520566603633221e-05,
+      "loss": 1.0795,
+      "step": 2357
+    },
+    {
+      "epoch": 0.14326508293334952,
+      "grad_norm": 0.172119602560997,
+      "learning_rate": 9.520157483123034e-05,
+      "loss": 1.18,
+      "step": 2358
+    },
+    {
+      "epoch": 0.14332583996597606,
+      "grad_norm": 0.1784110963344574,
+      "learning_rate": 9.519748196925601e-05,
+      "loss": 1.2058,
+      "step": 2359
+    },
+    {
+      "epoch": 0.1433865969986026,
+      "grad_norm": 0.17967002093791962,
+      "learning_rate": 9.519338745055926e-05,
+      "loss": 1.0634,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1434473540312291,
+      "grad_norm": 0.30833303928375244,
+      "learning_rate": 9.518929127529016e-05,
+      "loss": 1.1298,
+      "step": 2361
+    },
+    {
+      "epoch": 0.14350811106385564,
+      "grad_norm": 0.162101611495018,
+      "learning_rate": 9.518519344359882e-05,
+      "loss": 1.1489,
+      "step": 2362
+    },
+    {
+      "epoch": 0.14356886809648217,
+      "grad_norm": 0.22001470625400543,
+      "learning_rate": 9.51810939556355e-05,
+      "loss": 1.1747,
+      "step": 2363
+    },
+    {
+      "epoch": 0.1436296251291087,
+      "grad_norm": 0.20032212138175964,
+      "learning_rate": 9.517699281155046e-05,
+      "loss": 1.1118,
+      "step": 2364
+    },
+    {
+      "epoch": 0.14369038216173521,
+      "grad_norm": 0.206923246383667,
+      "learning_rate": 9.5172890011494e-05,
+      "loss": 1.2583,
+      "step": 2365
+    },
+    {
+      "epoch": 0.14375113919436175,
+      "grad_norm": 0.14203689992427826,
+      "learning_rate": 9.516878555561652e-05,
+      "loss": 1.0626,
+      "step": 2366
+    },
+    {
+      "epoch": 0.14381189622698828,
+      "grad_norm": 2.838843822479248,
+      "learning_rate": 9.516467944406848e-05,
+      "loss": 1.1281,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1438726532596148,
+      "grad_norm": 0.16107012331485748,
+      "learning_rate": 9.516057167700039e-05,
+      "loss": 1.1291,
+      "step": 2368
+    },
+    {
+      "epoch": 0.14393341029224133,
+      "grad_norm": 0.23750454187393188,
+      "learning_rate": 9.515646225456282e-05,
+      "loss": 1.1163,
+      "step": 2369
+    },
+    {
+      "epoch": 0.14399416732486786,
+      "grad_norm": 0.7830845713615417,
+      "learning_rate": 9.51523511769064e-05,
+      "loss": 1.1287,
+      "step": 2370
+    },
+    {
+      "epoch": 0.14405492435749437,
+      "grad_norm": 0.30437442660331726,
+      "learning_rate": 9.51482384441818e-05,
+      "loss": 1.046,
+      "step": 2371
+    },
+    {
+      "epoch": 0.1441156813901209,
+      "grad_norm": 0.17047496140003204,
+      "learning_rate": 9.51441240565398e-05,
+      "loss": 1.1627,
+      "step": 2372
+    },
+    {
+      "epoch": 0.14417643842274744,
+      "grad_norm": 0.18535156548023224,
+      "learning_rate": 9.514000801413122e-05,
+      "loss": 1.0805,
+      "step": 2373
+    },
+    {
+      "epoch": 0.14423719545537397,
+      "grad_norm": 0.21000000834465027,
+      "learning_rate": 9.513589031710691e-05,
+      "loss": 1.1576,
+      "step": 2374
+    },
+    {
+      "epoch": 0.14429795248800048,
+      "grad_norm": 0.22640864551067352,
+      "learning_rate": 9.513177096561781e-05,
+      "loss": 1.1064,
+      "step": 2375
+    },
+    {
+      "epoch": 0.14435870952062702,
+      "grad_norm": 0.20231229066848755,
+      "learning_rate": 9.512764995981493e-05,
+      "loss": 1.1338,
+      "step": 2376
+    },
+    {
+      "epoch": 0.14441946655325355,
+      "grad_norm": 0.1628715693950653,
+      "learning_rate": 9.512352729984932e-05,
+      "loss": 1.1069,
+      "step": 2377
+    },
+    {
+      "epoch": 0.14448022358588006,
+      "grad_norm": 0.3245583772659302,
+      "learning_rate": 9.51194029858721e-05,
+      "loss": 1.1173,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1445409806185066,
+      "grad_norm": 0.21512292325496674,
+      "learning_rate": 9.511527701803442e-05,
+      "loss": 1.0938,
+      "step": 2379
+    },
+    {
+      "epoch": 0.14460173765113313,
+      "grad_norm": 0.184569850564003,
+      "learning_rate": 9.511114939648755e-05,
+      "loss": 1.1926,
+      "step": 2380
+    },
+    {
+      "epoch": 0.14466249468375963,
+      "grad_norm": 0.175813227891922,
+      "learning_rate": 9.51070201213828e-05,
+      "loss": 1.0953,
+      "step": 2381
+    },
+    {
+      "epoch": 0.14472325171638617,
+      "grad_norm": 0.21451465785503387,
+      "learning_rate": 9.510288919287149e-05,
+      "loss": 1.1846,
+      "step": 2382
+    },
+    {
+      "epoch": 0.1447840087490127,
+      "grad_norm": 0.15214553475379944,
+      "learning_rate": 9.509875661110506e-05,
+      "loss": 1.1163,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1448447657816392,
+      "grad_norm": 0.1464489847421646,
+      "learning_rate": 9.509462237623499e-05,
+      "loss": 1.1324,
+      "step": 2384
+    },
+    {
+      "epoch": 0.14490552281426575,
+      "grad_norm": 0.14790543913841248,
+      "learning_rate": 9.509048648841281e-05,
+      "loss": 1.1133,
+      "step": 2385
+    },
+    {
+      "epoch": 0.14496627984689228,
+      "grad_norm": 0.23965315520763397,
+      "learning_rate": 9.508634894779014e-05,
+      "loss": 1.0818,
+      "step": 2386
+    },
+    {
+      "epoch": 0.14502703687951882,
+      "grad_norm": 0.18583054840564728,
+      "learning_rate": 9.508220975451864e-05,
+      "loss": 1.1308,
+      "step": 2387
+    },
+    {
+      "epoch": 0.14508779391214532,
+      "grad_norm": 0.960493266582489,
+      "learning_rate": 9.507806890875002e-05,
+      "loss": 1.173,
+      "step": 2388
+    },
+    {
+      "epoch": 0.14514855094477186,
+      "grad_norm": 0.16508913040161133,
+      "learning_rate": 9.507392641063607e-05,
+      "loss": 1.1809,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1452093079773984,
+      "grad_norm": 0.233011394739151,
+      "learning_rate": 9.506978226032864e-05,
+      "loss": 1.1236,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1452700650100249,
+      "grad_norm": 0.1477642059326172,
+      "learning_rate": 9.506563645797963e-05,
+      "loss": 1.1063,
+      "step": 2391
+    },
+    {
+      "epoch": 0.14533082204265144,
+      "grad_norm": 0.14913016557693481,
+      "learning_rate": 9.5061489003741e-05,
+      "loss": 1.132,
+      "step": 2392
+    },
+    {
+      "epoch": 0.14539157907527797,
+      "grad_norm": 0.1605728417634964,
+      "learning_rate": 9.505733989776478e-05,
+      "loss": 1.1114,
+      "step": 2393
+    },
+    {
+      "epoch": 0.14545233610790448,
+      "grad_norm": 0.24527214467525482,
+      "learning_rate": 9.505318914020307e-05,
+      "loss": 1.0602,
+      "step": 2394
+    },
+    {
+      "epoch": 0.145513093140531,
+      "grad_norm": 0.18539302051067352,
+      "learning_rate": 9.504903673120799e-05,
+      "loss": 1.1224,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14557385017315755,
+      "grad_norm": 0.35246172547340393,
+      "learning_rate": 9.504488267093175e-05,
+      "loss": 1.0643,
+      "step": 2396
+    },
+    {
+      "epoch": 0.14563460720578406,
+      "grad_norm": 0.18054181337356567,
+      "learning_rate": 9.504072695952664e-05,
+      "loss": 1.1607,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1456953642384106,
+      "grad_norm": 0.34346428513526917,
+      "learning_rate": 9.503656959714499e-05,
+      "loss": 1.1836,
+      "step": 2398
+    },
+    {
+      "epoch": 0.14575612127103713,
+      "grad_norm": 0.24999122321605682,
+      "learning_rate": 9.503241058393917e-05,
+      "loss": 1.1015,
+      "step": 2399
+    },
+    {
+      "epoch": 0.14581687830366366,
+      "grad_norm": 0.23386499285697937,
+      "learning_rate": 9.502824992006163e-05,
+      "loss": 1.1424,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14587763533629017,
+      "grad_norm": 0.43365755677223206,
+      "learning_rate": 9.502408760566489e-05,
+      "loss": 1.3813,
+      "step": 2401
+    },
+    {
+      "epoch": 0.1459383923689167,
+      "grad_norm": 0.9056212306022644,
+      "learning_rate": 9.501992364090151e-05,
+      "loss": 1.1485,
+      "step": 2402
+    },
+    {
+      "epoch": 0.14599914940154324,
+      "grad_norm": 0.34531742334365845,
+      "learning_rate": 9.501575802592413e-05,
+      "loss": 1.1166,
+      "step": 2403
+    },
+    {
+      "epoch": 0.14605990643416975,
+      "grad_norm": 0.20710986852645874,
+      "learning_rate": 9.501159076088543e-05,
+      "loss": 1.1896,
+      "step": 2404
+    },
+    {
+      "epoch": 0.14612066346679628,
+      "grad_norm": 0.19179341197013855,
+      "learning_rate": 9.500742184593818e-05,
+      "loss": 1.1641,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14618142049942282,
+      "grad_norm": 0.208840012550354,
+      "learning_rate": 9.500325128123517e-05,
+      "loss": 1.1568,
+      "step": 2406
+    },
+    {
+      "epoch": 0.14624217753204932,
+      "grad_norm": 0.3734762370586395,
+      "learning_rate": 9.49990790669293e-05,
+      "loss": 1.1741,
+      "step": 2407
+    },
+    {
+      "epoch": 0.14630293456467586,
+      "grad_norm": 0.19922985136508942,
+      "learning_rate": 9.499490520317349e-05,
+      "loss": 1.1229,
+      "step": 2408
+    },
+    {
+      "epoch": 0.1463636915973024,
+      "grad_norm": 0.22177734971046448,
+      "learning_rate": 9.499072969012073e-05,
+      "loss": 1.0884,
+      "step": 2409
+    },
+    {
+      "epoch": 0.14642444862992893,
+      "grad_norm": 0.17181159555912018,
+      "learning_rate": 9.498655252792406e-05,
+      "loss": 1.0948,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14648520566255543,
+      "grad_norm": 0.2920275330543518,
+      "learning_rate": 9.498237371673663e-05,
+      "loss": 1.0722,
+      "step": 2411
+    },
+    {
+      "epoch": 0.14654596269518197,
+      "grad_norm": 0.15354323387145996,
+      "learning_rate": 9.497819325671159e-05,
+      "loss": 1.081,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1466067197278085,
+      "grad_norm": 0.2565551996231079,
+      "learning_rate": 9.497401114800219e-05,
+      "loss": 1.233,
+      "step": 2413
+    },
+    {
+      "epoch": 0.146667476760435,
+      "grad_norm": 0.21972645819187164,
+      "learning_rate": 9.49698273907617e-05,
+      "loss": 1.0564,
+      "step": 2414
+    },
+    {
+      "epoch": 0.14672823379306155,
+      "grad_norm": 0.27179932594299316,
+      "learning_rate": 9.496564198514351e-05,
+      "loss": 1.263,
+      "step": 2415
+    },
+    {
+      "epoch": 0.14678899082568808,
+      "grad_norm": 0.21390116214752197,
+      "learning_rate": 9.496145493130102e-05,
+      "loss": 1.1002,
+      "step": 2416
+    },
+    {
+      "epoch": 0.1468497478583146,
+      "grad_norm": 0.19426123797893524,
+      "learning_rate": 9.49572662293877e-05,
+      "loss": 1.0568,
+      "step": 2417
+    },
+    {
+      "epoch": 0.14691050489094112,
+      "grad_norm": 3.3889048099517822,
+      "learning_rate": 9.495307587955711e-05,
+      "loss": 1.0866,
+      "step": 2418
+    },
+    {
+      "epoch": 0.14697126192356766,
+      "grad_norm": 0.18671445548534393,
+      "learning_rate": 9.494888388196282e-05,
+      "loss": 1.0914,
+      "step": 2419
+    },
+    {
+      "epoch": 0.14703201895619417,
+      "grad_norm": 0.17855724692344666,
+      "learning_rate": 9.494469023675851e-05,
+      "loss": 1.1418,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1470927759888207,
+      "grad_norm": 0.1991187036037445,
+      "learning_rate": 9.49404949440979e-05,
+      "loss": 1.0787,
+      "step": 2421
+    },
+    {
+      "epoch": 0.14715353302144724,
+      "grad_norm": 0.21052385866641998,
+      "learning_rate": 9.493629800413474e-05,
+      "loss": 1.2927,
+      "step": 2422
+    },
+    {
+      "epoch": 0.14721429005407377,
+      "grad_norm": 0.25615546107292175,
+      "learning_rate": 9.49320994170229e-05,
+      "loss": 1.1977,
+      "step": 2423
+    },
+    {
+      "epoch": 0.14727504708670028,
+      "grad_norm": 0.3489145040512085,
+      "learning_rate": 9.492789918291627e-05,
+      "loss": 1.166,
+      "step": 2424
+    },
+    {
+      "epoch": 0.1473358041193268,
+      "grad_norm": 0.13200810551643372,
+      "learning_rate": 9.492369730196882e-05,
+      "loss": 1.0489,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14739656115195335,
+      "grad_norm": 0.16536767780780792,
+      "learning_rate": 9.491949377433456e-05,
+      "loss": 1.1017,
+      "step": 2426
+    },
+    {
+      "epoch": 0.14745731818457986,
+      "grad_norm": 0.19354049861431122,
+      "learning_rate": 9.491528860016756e-05,
+      "loss": 1.1395,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1475180752172064,
+      "grad_norm": 0.2011626660823822,
+      "learning_rate": 9.491108177962199e-05,
+      "loss": 1.1231,
+      "step": 2428
+    },
+    {
+      "epoch": 0.14757883224983293,
+      "grad_norm": 0.17097003757953644,
+      "learning_rate": 9.490687331285202e-05,
+      "loss": 1.0884,
+      "step": 2429
+    },
+    {
+      "epoch": 0.14763958928245943,
+      "grad_norm": 0.35095110535621643,
+      "learning_rate": 9.490266320001195e-05,
+      "loss": 1.0691,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14770034631508597,
+      "grad_norm": 0.19442760944366455,
+      "learning_rate": 9.489845144125607e-05,
+      "loss": 1.1133,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1477611033477125,
+      "grad_norm": 0.17445772886276245,
+      "learning_rate": 9.489423803673877e-05,
+      "loss": 1.0856,
+      "step": 2432
+    },
+    {
+      "epoch": 0.14782186038033904,
+      "grad_norm": 0.21538683772087097,
+      "learning_rate": 9.489002298661449e-05,
+      "loss": 1.1704,
+      "step": 2433
+    },
+    {
+      "epoch": 0.14788261741296554,
+      "grad_norm": 0.22413907945156097,
+      "learning_rate": 9.488580629103776e-05,
+      "loss": 1.0727,
+      "step": 2434
+    },
+    {
+      "epoch": 0.14794337444559208,
+      "grad_norm": 0.23690840601921082,
+      "learning_rate": 9.488158795016311e-05,
+      "loss": 1.1026,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14800413147821861,
+      "grad_norm": 1.9819402694702148,
+      "learning_rate": 9.487736796414519e-05,
+      "loss": 1.1287,
+      "step": 2436
+    },
+    {
+      "epoch": 0.14806488851084512,
+      "grad_norm": 0.31615498661994934,
+      "learning_rate": 9.487314633313867e-05,
+      "loss": 1.2106,
+      "step": 2437
+    },
+    {
+      "epoch": 0.14812564554347166,
+      "grad_norm": 0.2629581093788147,
+      "learning_rate": 9.486892305729828e-05,
+      "loss": 1.1235,
+      "step": 2438
+    },
+    {
+      "epoch": 0.1481864025760982,
+      "grad_norm": 0.13540659844875336,
+      "learning_rate": 9.486469813677885e-05,
+      "loss": 1.0822,
+      "step": 2439
+    },
+    {
+      "epoch": 0.1482471596087247,
+      "grad_norm": 0.2806120812892914,
+      "learning_rate": 9.486047157173526e-05,
+      "loss": 1.0551,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14830791664135123,
+      "grad_norm": 0.16001953184604645,
+      "learning_rate": 9.485624336232239e-05,
+      "loss": 1.1186,
+      "step": 2441
+    },
+    {
+      "epoch": 0.14836867367397777,
+      "grad_norm": 1.1844679117202759,
+      "learning_rate": 9.485201350869525e-05,
+      "loss": 1.1247,
+      "step": 2442
+    },
+    {
+      "epoch": 0.14842943070660428,
+      "grad_norm": 0.16722291707992554,
+      "learning_rate": 9.48477820110089e-05,
+      "loss": 1.1587,
+      "step": 2443
+    },
+    {
+      "epoch": 0.1484901877392308,
+      "grad_norm": 0.9222338795661926,
+      "learning_rate": 9.484354886941842e-05,
+      "loss": 1.1013,
+      "step": 2444
+    },
+    {
+      "epoch": 0.14855094477185735,
+      "grad_norm": 0.4270765781402588,
+      "learning_rate": 9.483931408407898e-05,
+      "loss": 1.129,
+      "step": 2445
+    },
+    {
+      "epoch": 0.14861170180448388,
+      "grad_norm": 0.25752943754196167,
+      "learning_rate": 9.483507765514583e-05,
+      "loss": 1.2061,
+      "step": 2446
+    },
+    {
+      "epoch": 0.1486724588371104,
+      "grad_norm": 0.2032468020915985,
+      "learning_rate": 9.483083958277423e-05,
+      "loss": 1.2155,
+      "step": 2447
+    },
+    {
+      "epoch": 0.14873321586973692,
+      "grad_norm": 0.17130106687545776,
+      "learning_rate": 9.482659986711955e-05,
+      "loss": 1.1266,
+      "step": 2448
+    },
+    {
+      "epoch": 0.14879397290236346,
+      "grad_norm": 0.17864754796028137,
+      "learning_rate": 9.482235850833719e-05,
+      "loss": 1.1025,
+      "step": 2449
+    },
+    {
+      "epoch": 0.14885472993498997,
+      "grad_norm": 0.2654895782470703,
+      "learning_rate": 9.48181155065826e-05,
+      "loss": 1.1251,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1489154869676165,
+      "grad_norm": 0.13779211044311523,
+      "learning_rate": 9.481387086201132e-05,
+      "loss": 1.0993,
+      "step": 2451
+    },
+    {
+      "epoch": 0.14897624400024304,
+      "grad_norm": 0.17782364785671234,
+      "learning_rate": 9.480962457477896e-05,
+      "loss": 1.0823,
+      "step": 2452
+    },
+    {
+      "epoch": 0.14903700103286954,
+      "grad_norm": 0.2756407856941223,
+      "learning_rate": 9.480537664504115e-05,
+      "loss": 1.1324,
+      "step": 2453
+    },
+    {
+      "epoch": 0.14909775806549608,
+      "grad_norm": 0.22156783938407898,
+      "learning_rate": 9.480112707295359e-05,
+      "loss": 1.158,
+      "step": 2454
+    },
+    {
+      "epoch": 0.1491585150981226,
+      "grad_norm": 0.15597949922084808,
+      "learning_rate": 9.479687585867205e-05,
+      "loss": 1.1511,
+      "step": 2455
+    },
+    {
+      "epoch": 0.14921927213074915,
+      "grad_norm": 0.5653387308120728,
+      "learning_rate": 9.479262300235238e-05,
+      "loss": 1.0808,
+      "step": 2456
+    },
+    {
+      "epoch": 0.14928002916337565,
+      "grad_norm": 0.16475999355316162,
+      "learning_rate": 9.478836850415045e-05,
+      "loss": 1.2192,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1493407861960022,
+      "grad_norm": 0.20751553773880005,
+      "learning_rate": 9.478411236422221e-05,
+      "loss": 1.0879,
+      "step": 2458
+    },
+    {
+      "epoch": 0.14940154322862872,
+      "grad_norm": 0.9479624032974243,
+      "learning_rate": 9.477985458272369e-05,
+      "loss": 1.1913,
+      "step": 2459
+    },
+    {
+      "epoch": 0.14946230026125523,
+      "grad_norm": 0.1658661663532257,
+      "learning_rate": 9.477559515981092e-05,
+      "loss": 1.1942,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14952305729388177,
+      "grad_norm": 0.16715334355831146,
+      "learning_rate": 9.477133409564007e-05,
+      "loss": 1.2325,
+      "step": 2461
+    },
+    {
+      "epoch": 0.1495838143265083,
+      "grad_norm": 6.42448091506958,
+      "learning_rate": 9.476707139036731e-05,
+      "loss": 1.1289,
+      "step": 2462
+    },
+    {
+      "epoch": 0.1496445713591348,
+      "grad_norm": 0.20347461104393005,
+      "learning_rate": 9.47628070441489e-05,
+      "loss": 1.0632,
+      "step": 2463
+    },
+    {
+      "epoch": 0.14970532839176134,
+      "grad_norm": 0.27208903431892395,
+      "learning_rate": 9.475854105714115e-05,
+      "loss": 1.111,
+      "step": 2464
+    },
+    {
+      "epoch": 0.14976608542438788,
+      "grad_norm": 0.16360872983932495,
+      "learning_rate": 9.475427342950042e-05,
+      "loss": 1.1489,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1498268424570144,
+      "grad_norm": 0.13911795616149902,
+      "learning_rate": 9.475000416138314e-05,
+      "loss": 1.0844,
+      "step": 2466
+    },
+    {
+      "epoch": 0.14988759948964092,
+      "grad_norm": 0.3688059151172638,
+      "learning_rate": 9.474573325294582e-05,
+      "loss": 1.084,
+      "step": 2467
+    },
+    {
+      "epoch": 0.14994835652226746,
+      "grad_norm": 0.17356690764427185,
+      "learning_rate": 9.4741460704345e-05,
+      "loss": 1.1703,
+      "step": 2468
+    },
+    {
+      "epoch": 0.150009113554894,
+      "grad_norm": 0.194352388381958,
+      "learning_rate": 9.473718651573726e-05,
+      "loss": 1.1525,
+      "step": 2469
+    },
+    {
+      "epoch": 0.1500698705875205,
+      "grad_norm": 0.1539538949728012,
+      "learning_rate": 9.473291068727933e-05,
+      "loss": 1.107,
+      "step": 2470
+    },
+    {
+      "epoch": 0.15013062762014703,
+      "grad_norm": 0.2254716455936432,
+      "learning_rate": 9.472863321912791e-05,
+      "loss": 1.2808,
+      "step": 2471
+    },
+    {
+      "epoch": 0.15019138465277357,
+      "grad_norm": 0.20095323026180267,
+      "learning_rate": 9.472435411143978e-05,
+      "loss": 1.1233,
+      "step": 2472
+    },
+    {
+      "epoch": 0.15025214168540008,
+      "grad_norm": 0.18226878345012665,
+      "learning_rate": 9.47200733643718e-05,
+      "loss": 1.2446,
+      "step": 2473
+    },
+    {
+      "epoch": 0.1503128987180266,
+      "grad_norm": 0.5279532074928284,
+      "learning_rate": 9.471579097808092e-05,
+      "loss": 1.1541,
+      "step": 2474
+    },
+    {
+      "epoch": 0.15037365575065315,
+      "grad_norm": 0.19655056297779083,
+      "learning_rate": 9.471150695272406e-05,
+      "loss": 1.1113,
+      "step": 2475
+    },
+    {
+      "epoch": 0.15043441278327965,
+      "grad_norm": 0.20901861786842346,
+      "learning_rate": 9.470722128845826e-05,
+      "loss": 1.0721,
+      "step": 2476
+    },
+    {
+      "epoch": 0.1504951698159062,
+      "grad_norm": 0.22774384915828705,
+      "learning_rate": 9.470293398544064e-05,
+      "loss": 1.2215,
+      "step": 2477
+    },
+    {
+      "epoch": 0.15055592684853272,
+      "grad_norm": 0.20482629537582397,
+      "learning_rate": 9.469864504382832e-05,
+      "loss": 1.114,
+      "step": 2478
+    },
+    {
+      "epoch": 0.15061668388115926,
+      "grad_norm": 1.307689905166626,
+      "learning_rate": 9.469435446377852e-05,
+      "loss": 1.1239,
+      "step": 2479
+    },
+    {
+      "epoch": 0.15067744091378577,
+      "grad_norm": 0.2381213754415512,
+      "learning_rate": 9.469006224544854e-05,
+      "loss": 1.131,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1507381979464123,
+      "grad_norm": 0.14561453461647034,
+      "learning_rate": 9.468576838899567e-05,
+      "loss": 1.0768,
+      "step": 2481
+    },
+    {
+      "epoch": 0.15079895497903884,
+      "grad_norm": 0.20880720019340515,
+      "learning_rate": 9.468147289457732e-05,
+      "loss": 1.1832,
+      "step": 2482
+    },
+    {
+      "epoch": 0.15085971201166534,
+      "grad_norm": 0.6424425840377808,
+      "learning_rate": 9.467717576235095e-05,
+      "loss": 1.2518,
+      "step": 2483
+    },
+    {
+      "epoch": 0.15092046904429188,
+      "grad_norm": 0.20040066540241241,
+      "learning_rate": 9.467287699247409e-05,
+      "loss": 1.249,
+      "step": 2484
+    },
+    {
+      "epoch": 0.1509812260769184,
+      "grad_norm": 0.22147944569587708,
+      "learning_rate": 9.466857658510427e-05,
+      "loss": 1.1468,
+      "step": 2485
+    },
+    {
+      "epoch": 0.15104198310954492,
+      "grad_norm": 0.21854613721370697,
+      "learning_rate": 9.466427454039914e-05,
+      "loss": 1.083,
+      "step": 2486
+    },
+    {
+      "epoch": 0.15110274014217145,
+      "grad_norm": 0.2647652328014374,
+      "learning_rate": 9.465997085851639e-05,
+      "loss": 1.1411,
+      "step": 2487
+    },
+    {
+      "epoch": 0.151163497174798,
+      "grad_norm": 0.24156533181667328,
+      "learning_rate": 9.465566553961378e-05,
+      "loss": 1.1711,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1512242542074245,
+      "grad_norm": 0.15108446776866913,
+      "learning_rate": 9.465135858384911e-05,
+      "loss": 1.0747,
+      "step": 2489
+    },
+    {
+      "epoch": 0.15128501124005103,
+      "grad_norm": 1.964840054512024,
+      "learning_rate": 9.464704999138027e-05,
+      "loss": 1.2311,
+      "step": 2490
+    },
+    {
+      "epoch": 0.15134576827267757,
+      "grad_norm": 0.3127102851867676,
+      "learning_rate": 9.464273976236517e-05,
+      "loss": 1.1105,
+      "step": 2491
+    },
+    {
+      "epoch": 0.1514065253053041,
+      "grad_norm": 0.2378004938364029,
+      "learning_rate": 9.463842789696184e-05,
+      "loss": 1.0764,
+      "step": 2492
+    },
+    {
+      "epoch": 0.1514672823379306,
+      "grad_norm": 0.14329063892364502,
+      "learning_rate": 9.463411439532828e-05,
+      "loss": 1.1309,
+      "step": 2493
+    },
+    {
+      "epoch": 0.15152803937055714,
+      "grad_norm": 0.23911572992801666,
+      "learning_rate": 9.462979925762266e-05,
+      "loss": 1.2079,
+      "step": 2494
+    },
+    {
+      "epoch": 0.15158879640318368,
+      "grad_norm": 0.19386433064937592,
+      "learning_rate": 9.462548248400309e-05,
+      "loss": 1.1623,
+      "step": 2495
+    },
+    {
+      "epoch": 0.15164955343581019,
+      "grad_norm": 0.21827779710292816,
+      "learning_rate": 9.462116407462784e-05,
+      "loss": 1.1149,
+      "step": 2496
+    },
+    {
+      "epoch": 0.15171031046843672,
+      "grad_norm": 0.2998878061771393,
+      "learning_rate": 9.46168440296552e-05,
+      "loss": 1.1089,
+      "step": 2497
+    },
+    {
+      "epoch": 0.15177106750106326,
+      "grad_norm": 0.3298884332180023,
+      "learning_rate": 9.461252234924351e-05,
+      "loss": 1.152,
+      "step": 2498
+    },
+    {
+      "epoch": 0.15183182453368976,
+      "grad_norm": 0.229386568069458,
+      "learning_rate": 9.460819903355121e-05,
+      "loss": 1.3153,
+      "step": 2499
+    },
+    {
+      "epoch": 0.1518925815663163,
+      "grad_norm": 0.27866312861442566,
+      "learning_rate": 9.460387408273675e-05,
+      "loss": 1.1676,
+      "step": 2500
+    },
+    {
+      "epoch": 0.15195333859894283,
+      "grad_norm": 0.2763281464576721,
+      "learning_rate": 9.459954749695865e-05,
+      "loss": 1.1258,
+      "step": 2501
+    },
+    {
+      "epoch": 0.15201409563156934,
+      "grad_norm": 0.34656262397766113,
+      "learning_rate": 9.459521927637551e-05,
+      "loss": 1.0938,
+      "step": 2502
+    },
+    {
+      "epoch": 0.15207485266419588,
+      "grad_norm": 1.4510091543197632,
+      "learning_rate": 9.4590889421146e-05,
+      "loss": 1.1494,
+      "step": 2503
+    },
+    {
+      "epoch": 0.1521356096968224,
+      "grad_norm": 3.499176263809204,
+      "learning_rate": 9.458655793142883e-05,
+      "loss": 1.1702,
+      "step": 2504
+    },
+    {
+      "epoch": 0.15219636672944895,
+      "grad_norm": 0.16883930563926697,
+      "learning_rate": 9.458222480738275e-05,
+      "loss": 1.1477,
+      "step": 2505
+    },
+    {
+      "epoch": 0.15225712376207545,
+      "grad_norm": 0.2272227555513382,
+      "learning_rate": 9.45778900491666e-05,
+      "loss": 1.1914,
+      "step": 2506
+    },
+    {
+      "epoch": 0.152317880794702,
+      "grad_norm": 0.3026867210865021,
+      "learning_rate": 9.457355365693927e-05,
+      "loss": 1.1802,
+      "step": 2507
+    },
+    {
+      "epoch": 0.15237863782732852,
+      "grad_norm": 0.14221596717834473,
+      "learning_rate": 9.456921563085972e-05,
+      "loss": 1.1192,
+      "step": 2508
+    },
+    {
+      "epoch": 0.15243939485995503,
+      "grad_norm": 0.20842139422893524,
+      "learning_rate": 9.456487597108694e-05,
+      "loss": 1.076,
+      "step": 2509
+    },
+    {
+      "epoch": 0.15250015189258156,
+      "grad_norm": 0.22788575291633606,
+      "learning_rate": 9.456053467778003e-05,
+      "loss": 1.0718,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1525609089252081,
+      "grad_norm": 0.22824859619140625,
+      "learning_rate": 9.455619175109811e-05,
+      "loss": 1.1488,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1526216659578346,
+      "grad_norm": 0.2251524180173874,
+      "learning_rate": 9.455184719120038e-05,
+      "loss": 1.1245,
+      "step": 2512
+    },
+    {
+      "epoch": 0.15268242299046114,
+      "grad_norm": 0.26367315649986267,
+      "learning_rate": 9.454750099824606e-05,
+      "loss": 1.0966,
+      "step": 2513
+    },
+    {
+      "epoch": 0.15274318002308768,
+      "grad_norm": 2.280456066131592,
+      "learning_rate": 9.454315317239448e-05,
+      "loss": 1.1042,
+      "step": 2514
+    },
+    {
+      "epoch": 0.1528039370557142,
+      "grad_norm": 0.23190902173519135,
+      "learning_rate": 9.453880371380503e-05,
+      "loss": 1.1163,
+      "step": 2515
+    },
+    {
+      "epoch": 0.15286469408834072,
+      "grad_norm": 0.2745177745819092,
+      "learning_rate": 9.453445262263711e-05,
+      "loss": 1.1081,
+      "step": 2516
+    },
+    {
+      "epoch": 0.15292545112096725,
+      "grad_norm": 1.4664905071258545,
+      "learning_rate": 9.453009989905022e-05,
+      "loss": 1.12,
+      "step": 2517
+    },
+    {
+      "epoch": 0.1529862081535938,
+      "grad_norm": 0.5669666528701782,
+      "learning_rate": 9.452574554320391e-05,
+      "loss": 1.1778,
+      "step": 2518
+    },
+    {
+      "epoch": 0.1530469651862203,
+      "grad_norm": 0.824489414691925,
+      "learning_rate": 9.452138955525779e-05,
+      "loss": 1.1576,
+      "step": 2519
+    },
+    {
+      "epoch": 0.15310772221884683,
+      "grad_norm": 0.6353117227554321,
+      "learning_rate": 9.451703193537153e-05,
+      "loss": 1.1855,
+      "step": 2520
+    },
+    {
+      "epoch": 0.15316847925147337,
+      "grad_norm": 1.1025758981704712,
+      "learning_rate": 9.451267268370486e-05,
+      "loss": 1.1288,
+      "step": 2521
+    },
+    {
+      "epoch": 0.15322923628409987,
+      "grad_norm": 0.4223443567752838,
+      "learning_rate": 9.450831180041757e-05,
+      "loss": 1.1644,
+      "step": 2522
+    },
+    {
+      "epoch": 0.1532899933167264,
+      "grad_norm": 0.2359962910413742,
+      "learning_rate": 9.450394928566951e-05,
+      "loss": 1.077,
+      "step": 2523
+    },
+    {
+      "epoch": 0.15335075034935294,
+      "grad_norm": 0.5190343856811523,
+      "learning_rate": 9.449958513962059e-05,
+      "loss": 1.0871,
+      "step": 2524
+    },
+    {
+      "epoch": 0.15341150738197945,
+      "grad_norm": 0.6135958433151245,
+      "learning_rate": 9.449521936243075e-05,
+      "loss": 1.1836,
+      "step": 2525
+    },
+    {
+      "epoch": 0.15347226441460599,
+      "grad_norm": 0.5102086663246155,
+      "learning_rate": 9.449085195426007e-05,
+      "loss": 1.1572,
+      "step": 2526
+    },
+    {
+      "epoch": 0.15353302144723252,
+      "grad_norm": 0.2285347729921341,
+      "learning_rate": 9.448648291526861e-05,
+      "loss": 1.1021,
+      "step": 2527
+    },
+    {
+      "epoch": 0.15359377847985906,
+      "grad_norm": 0.16917438805103302,
+      "learning_rate": 9.448211224561651e-05,
+      "loss": 1.0787,
+      "step": 2528
+    },
+    {
+      "epoch": 0.15365453551248556,
+      "grad_norm": 0.2284194976091385,
+      "learning_rate": 9.447773994546399e-05,
+      "loss": 1.1107,
+      "step": 2529
+    },
+    {
+      "epoch": 0.1537152925451121,
+      "grad_norm": 0.9874343276023865,
+      "learning_rate": 9.44733660149713e-05,
+      "loss": 1.2561,
+      "step": 2530
+    },
+    {
+      "epoch": 0.15377604957773863,
+      "grad_norm": 0.46491485834121704,
+      "learning_rate": 9.446899045429881e-05,
+      "loss": 1.272,
+      "step": 2531
+    },
+    {
+      "epoch": 0.15383680661036514,
+      "grad_norm": 0.18492628633975983,
+      "learning_rate": 9.446461326360687e-05,
+      "loss": 1.1473,
+      "step": 2532
+    },
+    {
+      "epoch": 0.15389756364299167,
+      "grad_norm": 0.26542723178863525,
+      "learning_rate": 9.446023444305592e-05,
+      "loss": 1.3215,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1539583206756182,
+      "grad_norm": 0.32704147696495056,
+      "learning_rate": 9.44558539928065e-05,
+      "loss": 1.0896,
+      "step": 2534
+    },
+    {
+      "epoch": 0.15401907770824472,
+      "grad_norm": 0.3065110743045807,
+      "learning_rate": 9.445147191301917e-05,
+      "loss": 1.1356,
+      "step": 2535
+    },
+    {
+      "epoch": 0.15407983474087125,
+      "grad_norm": 0.23968765139579773,
+      "learning_rate": 9.444708820385451e-05,
+      "loss": 1.1387,
+      "step": 2536
+    },
+    {
+      "epoch": 0.1541405917734978,
+      "grad_norm": 0.26323044300079346,
+      "learning_rate": 9.444270286547328e-05,
+      "loss": 1.0891,
+      "step": 2537
+    },
+    {
+      "epoch": 0.15420134880612432,
+      "grad_norm": 0.21186158061027527,
+      "learning_rate": 9.443831589803616e-05,
+      "loss": 1.2058,
+      "step": 2538
+    },
+    {
+      "epoch": 0.15426210583875083,
+      "grad_norm": 0.2574732005596161,
+      "learning_rate": 9.443392730170398e-05,
+      "loss": 1.1316,
+      "step": 2539
+    },
+    {
+      "epoch": 0.15432286287137736,
+      "grad_norm": 0.22741980850696564,
+      "learning_rate": 9.442953707663763e-05,
+      "loss": 1.2031,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1543836199040039,
+      "grad_norm": 0.3144490420818329,
+      "learning_rate": 9.442514522299797e-05,
+      "loss": 1.0981,
+      "step": 2541
+    },
+    {
+      "epoch": 0.1544443769366304,
+      "grad_norm": 0.2632023096084595,
+      "learning_rate": 9.442075174094606e-05,
+      "loss": 1.0804,
+      "step": 2542
+    },
+    {
+      "epoch": 0.15450513396925694,
+      "grad_norm": 0.279153972864151,
+      "learning_rate": 9.441635663064289e-05,
+      "loss": 1.0267,
+      "step": 2543
+    },
+    {
+      "epoch": 0.15456589100188348,
+      "grad_norm": 0.17136962711811066,
+      "learning_rate": 9.441195989224958e-05,
+      "loss": 1.1099,
+      "step": 2544
+    },
+    {
+      "epoch": 0.15462664803450998,
+      "grad_norm": 0.2738109827041626,
+      "learning_rate": 9.440756152592731e-05,
+      "loss": 1.144,
+      "step": 2545
+    },
+    {
+      "epoch": 0.15468740506713652,
+      "grad_norm": 0.3377268612384796,
+      "learning_rate": 9.440316153183729e-05,
+      "loss": 1.221,
+      "step": 2546
+    },
+    {
+      "epoch": 0.15474816209976305,
+      "grad_norm": 0.6182905435562134,
+      "learning_rate": 9.43987599101408e-05,
+      "loss": 1.3155,
+      "step": 2547
+    },
+    {
+      "epoch": 0.15480891913238956,
+      "grad_norm": 0.25432828068733215,
+      "learning_rate": 9.439435666099916e-05,
+      "loss": 1.1005,
+      "step": 2548
+    },
+    {
+      "epoch": 0.1548696761650161,
+      "grad_norm": 0.28780269622802734,
+      "learning_rate": 9.438995178457381e-05,
+      "loss": 1.2799,
+      "step": 2549
+    },
+    {
+      "epoch": 0.15493043319764263,
+      "grad_norm": 0.3011924922466278,
+      "learning_rate": 9.438554528102621e-05,
+      "loss": 1.1175,
+      "step": 2550
+    },
+    {
+      "epoch": 0.15499119023026917,
+      "grad_norm": 0.24541157484054565,
+      "learning_rate": 9.438113715051785e-05,
+      "loss": 1.1392,
+      "step": 2551
+    },
+    {
+      "epoch": 0.15505194726289567,
+      "grad_norm": 0.2185327410697937,
+      "learning_rate": 9.437672739321034e-05,
+      "loss": 1.0796,
+      "step": 2552
+    },
+    {
+      "epoch": 0.1551127042955222,
+      "grad_norm": 0.1773419827222824,
+      "learning_rate": 9.43723160092653e-05,
+      "loss": 1.0839,
+      "step": 2553
+    },
+    {
+      "epoch": 0.15517346132814874,
+      "grad_norm": 0.2325228601694107,
+      "learning_rate": 9.436790299884445e-05,
+      "loss": 1.1762,
+      "step": 2554
+    },
+    {
+      "epoch": 0.15523421836077525,
+      "grad_norm": 0.3388921916484833,
+      "learning_rate": 9.436348836210953e-05,
+      "loss": 1.1389,
+      "step": 2555
+    },
+    {
+      "epoch": 0.15529497539340179,
+      "grad_norm": 0.18111151456832886,
+      "learning_rate": 9.435907209922238e-05,
+      "loss": 1.0575,
+      "step": 2556
+    },
+    {
+      "epoch": 0.15535573242602832,
+      "grad_norm": 0.3273448050022125,
+      "learning_rate": 9.435465421034486e-05,
+      "loss": 1.5042,
+      "step": 2557
+    },
+    {
+      "epoch": 0.15541648945865483,
+      "grad_norm": 0.27869129180908203,
+      "learning_rate": 9.435023469563892e-05,
+      "loss": 1.1753,
+      "step": 2558
+    },
+    {
+      "epoch": 0.15547724649128136,
+      "grad_norm": 0.21914230287075043,
+      "learning_rate": 9.434581355526655e-05,
+      "loss": 1.1313,
+      "step": 2559
+    },
+    {
+      "epoch": 0.1555380035239079,
+      "grad_norm": 0.2693517506122589,
+      "learning_rate": 9.434139078938982e-05,
+      "loss": 1.196,
+      "step": 2560
+    },
+    {
+      "epoch": 0.15559876055653443,
+      "grad_norm": 0.31690874695777893,
+      "learning_rate": 9.433696639817082e-05,
+      "loss": 1.1136,
+      "step": 2561
+    },
+    {
+      "epoch": 0.15565951758916094,
+      "grad_norm": 0.28778785467147827,
+      "learning_rate": 9.433254038177178e-05,
+      "loss": 1.1675,
+      "step": 2562
+    },
+    {
+      "epoch": 0.15572027462178747,
+      "grad_norm": 0.32684969902038574,
+      "learning_rate": 9.432811274035486e-05,
+      "loss": 1.1399,
+      "step": 2563
+    },
+    {
+      "epoch": 0.155781031654414,
+      "grad_norm": 0.152581125497818,
+      "learning_rate": 9.432368347408244e-05,
+      "loss": 1.0687,
+      "step": 2564
+    },
+    {
+      "epoch": 0.15584178868704052,
+      "grad_norm": 0.22406163811683655,
+      "learning_rate": 9.431925258311682e-05,
+      "loss": 1.1608,
+      "step": 2565
+    },
+    {
+      "epoch": 0.15590254571966705,
+      "grad_norm": 0.19227343797683716,
+      "learning_rate": 9.431482006762042e-05,
+      "loss": 1.1161,
+      "step": 2566
+    },
+    {
+      "epoch": 0.1559633027522936,
+      "grad_norm": 0.22506600618362427,
+      "learning_rate": 9.431038592775575e-05,
+      "loss": 1.1232,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1560240597849201,
+      "grad_norm": 0.43968236446380615,
+      "learning_rate": 9.430595016368528e-05,
+      "loss": 1.2479,
+      "step": 2568
+    },
+    {
+      "epoch": 0.15608481681754663,
+      "grad_norm": 0.6710606813430786,
+      "learning_rate": 9.430151277557167e-05,
+      "loss": 1.1688,
+      "step": 2569
+    },
+    {
+      "epoch": 0.15614557385017316,
+      "grad_norm": 0.2122555822134018,
+      "learning_rate": 9.429707376357753e-05,
+      "loss": 1.167,
+      "step": 2570
+    },
+    {
+      "epoch": 0.15620633088279967,
+      "grad_norm": 0.20122987031936646,
+      "learning_rate": 9.42926331278656e-05,
+      "loss": 1.1036,
+      "step": 2571
+    },
+    {
+      "epoch": 0.1562670879154262,
+      "grad_norm": 0.19168125092983246,
+      "learning_rate": 9.428819086859865e-05,
+      "loss": 1.1438,
+      "step": 2572
+    },
+    {
+      "epoch": 0.15632784494805274,
+      "grad_norm": 0.18105441331863403,
+      "learning_rate": 9.428374698593949e-05,
+      "loss": 1.236,
+      "step": 2573
+    },
+    {
+      "epoch": 0.15638860198067928,
+      "grad_norm": 0.1889175921678543,
+      "learning_rate": 9.427930148005102e-05,
+      "loss": 1.1755,
+      "step": 2574
+    },
+    {
+      "epoch": 0.15644935901330578,
+      "grad_norm": 0.17544624209403992,
+      "learning_rate": 9.42748543510962e-05,
+      "loss": 1.1175,
+      "step": 2575
+    },
+    {
+      "epoch": 0.15651011604593232,
+      "grad_norm": 0.21776503324508667,
+      "learning_rate": 9.427040559923804e-05,
+      "loss": 1.0939,
+      "step": 2576
+    },
+    {
+      "epoch": 0.15657087307855885,
+      "grad_norm": 0.24412862956523895,
+      "learning_rate": 9.426595522463961e-05,
+      "loss": 1.0695,
+      "step": 2577
+    },
+    {
+      "epoch": 0.15663163011118536,
+      "grad_norm": 0.16117967665195465,
+      "learning_rate": 9.426150322746404e-05,
+      "loss": 1.0902,
+      "step": 2578
+    },
+    {
+      "epoch": 0.1566923871438119,
+      "grad_norm": 0.15866626799106598,
+      "learning_rate": 9.425704960787449e-05,
+      "loss": 1.0782,
+      "step": 2579
+    },
+    {
+      "epoch": 0.15675314417643843,
+      "grad_norm": 0.19047687947750092,
+      "learning_rate": 9.425259436603425e-05,
+      "loss": 1.1106,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15681390120906494,
+      "grad_norm": 0.19610057771205902,
+      "learning_rate": 9.42481375021066e-05,
+      "loss": 1.148,
+      "step": 2581
+    },
+    {
+      "epoch": 0.15687465824169147,
+      "grad_norm": 2.3571863174438477,
+      "learning_rate": 9.424367901625491e-05,
+      "loss": 1.0905,
+      "step": 2582
+    },
+    {
+      "epoch": 0.156935415274318,
+      "grad_norm": 0.17872482538223267,
+      "learning_rate": 9.423921890864263e-05,
+      "loss": 1.1368,
+      "step": 2583
+    },
+    {
+      "epoch": 0.15699617230694454,
+      "grad_norm": 0.2017214000225067,
+      "learning_rate": 9.423475717943321e-05,
+      "loss": 1.1529,
+      "step": 2584
+    },
+    {
+      "epoch": 0.15705692933957105,
+      "grad_norm": 0.15281355381011963,
+      "learning_rate": 9.423029382879023e-05,
+      "loss": 1.1106,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15711768637219758,
+      "grad_norm": 0.42306381464004517,
+      "learning_rate": 9.422582885687728e-05,
+      "loss": 1.1726,
+      "step": 2586
+    },
+    {
+      "epoch": 0.15717844340482412,
+      "grad_norm": 0.17120933532714844,
+      "learning_rate": 9.422136226385802e-05,
+      "loss": 1.1587,
+      "step": 2587
+    },
+    {
+      "epoch": 0.15723920043745063,
+      "grad_norm": 0.26416444778442383,
+      "learning_rate": 9.42168940498962e-05,
+      "loss": 1.2959,
+      "step": 2588
+    },
+    {
+      "epoch": 0.15729995747007716,
+      "grad_norm": 0.7534836530685425,
+      "learning_rate": 9.421242421515554e-05,
+      "loss": 1.0588,
+      "step": 2589
+    },
+    {
+      "epoch": 0.1573607145027037,
+      "grad_norm": 0.2690413296222687,
+      "learning_rate": 9.420795275979997e-05,
+      "loss": 1.1023,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1574214715353302,
+      "grad_norm": 0.1511676013469696,
+      "learning_rate": 9.420347968399332e-05,
+      "loss": 1.0809,
+      "step": 2591
+    },
+    {
+      "epoch": 0.15748222856795674,
+      "grad_norm": 0.14605583250522614,
+      "learning_rate": 9.419900498789958e-05,
+      "loss": 1.1054,
+      "step": 2592
+    },
+    {
+      "epoch": 0.15754298560058327,
+      "grad_norm": 0.1760273426771164,
+      "learning_rate": 9.419452867168277e-05,
+      "loss": 1.1226,
+      "step": 2593
+    },
+    {
+      "epoch": 0.15760374263320978,
+      "grad_norm": 7.571736812591553,
+      "learning_rate": 9.419005073550697e-05,
+      "loss": 1.2598,
+      "step": 2594
+    },
+    {
+      "epoch": 0.15766449966583632,
+      "grad_norm": 0.16633674502372742,
+      "learning_rate": 9.418557117953632e-05,
+      "loss": 1.0656,
+      "step": 2595
+    },
+    {
+      "epoch": 0.15772525669846285,
+      "grad_norm": 0.15302538871765137,
+      "learning_rate": 9.418109000393501e-05,
+      "loss": 1.0378,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1577860137310894,
+      "grad_norm": 0.16779173910617828,
+      "learning_rate": 9.417660720886732e-05,
+      "loss": 1.1809,
+      "step": 2597
+    },
+    {
+      "epoch": 0.1578467707637159,
+      "grad_norm": 0.14601485431194305,
+      "learning_rate": 9.417212279449754e-05,
+      "loss": 1.0573,
+      "step": 2598
+    },
+    {
+      "epoch": 0.15790752779634243,
+      "grad_norm": 0.4659806489944458,
+      "learning_rate": 9.416763676099007e-05,
+      "loss": 1.2161,
+      "step": 2599
+    },
+    {
+      "epoch": 0.15796828482896896,
+      "grad_norm": 0.1553792655467987,
+      "learning_rate": 9.416314910850931e-05,
+      "loss": 1.1836,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15802904186159547,
+      "grad_norm": 0.3346526026725769,
+      "learning_rate": 9.415865983721982e-05,
+      "loss": 1.1007,
+      "step": 2601
+    },
+    {
+      "epoch": 0.158089798894222,
+      "grad_norm": 0.18200501799583435,
+      "learning_rate": 9.41541689472861e-05,
+      "loss": 1.1573,
+      "step": 2602
+    },
+    {
+      "epoch": 0.15815055592684854,
+      "grad_norm": 0.19293729960918427,
+      "learning_rate": 9.414967643887279e-05,
+      "loss": 1.1278,
+      "step": 2603
+    },
+    {
+      "epoch": 0.15821131295947505,
+      "grad_norm": 0.21648606657981873,
+      "learning_rate": 9.414518231214455e-05,
+      "loss": 1.129,
+      "step": 2604
+    },
+    {
+      "epoch": 0.15827206999210158,
+      "grad_norm": 0.1847764402627945,
+      "learning_rate": 9.414068656726612e-05,
+      "loss": 1.104,
+      "step": 2605
+    },
+    {
+      "epoch": 0.15833282702472812,
+      "grad_norm": 0.2570077180862427,
+      "learning_rate": 9.41361892044023e-05,
+      "loss": 1.1204,
+      "step": 2606
+    },
+    {
+      "epoch": 0.15839358405735462,
+      "grad_norm": 0.14342860877513885,
+      "learning_rate": 9.413169022371793e-05,
+      "loss": 1.0479,
+      "step": 2607
+    },
+    {
+      "epoch": 0.15845434108998116,
+      "grad_norm": 0.781670093536377,
+      "learning_rate": 9.412718962537792e-05,
+      "loss": 1.0611,
+      "step": 2608
+    },
+    {
+      "epoch": 0.1585150981226077,
+      "grad_norm": 0.23704147338867188,
+      "learning_rate": 9.412268740954724e-05,
+      "loss": 1.2103,
+      "step": 2609
+    },
+    {
+      "epoch": 0.15857585515523423,
+      "grad_norm": 0.24569956958293915,
+      "learning_rate": 9.411818357639093e-05,
+      "loss": 1.1834,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15863661218786074,
+      "grad_norm": 0.20215216279029846,
+      "learning_rate": 9.411367812607408e-05,
+      "loss": 1.1998,
+      "step": 2611
+    },
+    {
+      "epoch": 0.15869736922048727,
+      "grad_norm": 0.6925764679908752,
+      "learning_rate": 9.410917105876183e-05,
+      "loss": 1.2567,
+      "step": 2612
+    },
+    {
+      "epoch": 0.1587581262531138,
+      "grad_norm": 0.18702685832977295,
+      "learning_rate": 9.410466237461937e-05,
+      "loss": 1.1092,
+      "step": 2613
+    },
+    {
+      "epoch": 0.15881888328574031,
+      "grad_norm": 0.822773277759552,
+      "learning_rate": 9.4100152073812e-05,
+      "loss": 1.1646,
+      "step": 2614
+    },
+    {
+      "epoch": 0.15887964031836685,
+      "grad_norm": 0.27991822361946106,
+      "learning_rate": 9.409564015650502e-05,
+      "loss": 1.1806,
+      "step": 2615
+    },
+    {
+      "epoch": 0.15894039735099338,
+      "grad_norm": 0.2208004593849182,
+      "learning_rate": 9.409112662286385e-05,
+      "loss": 1.201,
+      "step": 2616
+    },
+    {
+      "epoch": 0.1590011543836199,
+      "grad_norm": 0.28630945086479187,
+      "learning_rate": 9.408661147305389e-05,
+      "loss": 1.1783,
+      "step": 2617
+    },
+    {
+      "epoch": 0.15906191141624643,
+      "grad_norm": 0.16344986855983734,
+      "learning_rate": 9.408209470724068e-05,
+      "loss": 1.074,
+      "step": 2618
+    },
+    {
+      "epoch": 0.15912266844887296,
+      "grad_norm": 0.9132686853408813,
+      "learning_rate": 9.407757632558975e-05,
+      "loss": 1.1918,
+      "step": 2619
+    },
+    {
+      "epoch": 0.1591834254814995,
+      "grad_norm": 0.24981817603111267,
+      "learning_rate": 9.407305632826675e-05,
+      "loss": 1.2166,
+      "step": 2620
+    },
+    {
+      "epoch": 0.159244182514126,
+      "grad_norm": 0.14491136372089386,
+      "learning_rate": 9.406853471543735e-05,
+      "loss": 1.0663,
+      "step": 2621
+    },
+    {
+      "epoch": 0.15930493954675254,
+      "grad_norm": 0.7831388711929321,
+      "learning_rate": 9.40640114872673e-05,
+      "loss": 1.0644,
+      "step": 2622
+    },
+    {
+      "epoch": 0.15936569657937907,
+      "grad_norm": 0.3581639230251312,
+      "learning_rate": 9.405948664392239e-05,
+      "loss": 1.0748,
+      "step": 2623
+    },
+    {
+      "epoch": 0.15942645361200558,
+      "grad_norm": 0.3864647150039673,
+      "learning_rate": 9.405496018556847e-05,
+      "loss": 1.075,
+      "step": 2624
+    },
+    {
+      "epoch": 0.15948721064463212,
+      "grad_norm": 0.21236863732337952,
+      "learning_rate": 9.405043211237147e-05,
+      "loss": 1.1033,
+      "step": 2625
+    },
+    {
+      "epoch": 0.15954796767725865,
+      "grad_norm": 0.24810928106307983,
+      "learning_rate": 9.404590242449738e-05,
+      "loss": 1.2299,
+      "step": 2626
+    },
+    {
+      "epoch": 0.15960872470988516,
+      "grad_norm": 0.16409023106098175,
+      "learning_rate": 9.40413711221122e-05,
+      "loss": 1.1018,
+      "step": 2627
+    },
+    {
+      "epoch": 0.1596694817425117,
+      "grad_norm": 0.319889098405838,
+      "learning_rate": 9.403683820538208e-05,
+      "loss": 1.2046,
+      "step": 2628
+    },
+    {
+      "epoch": 0.15973023877513823,
+      "grad_norm": 0.2668726146221161,
+      "learning_rate": 9.403230367447311e-05,
+      "loss": 1.1948,
+      "step": 2629
+    },
+    {
+      "epoch": 0.15979099580776474,
+      "grad_norm": 0.254044771194458,
+      "learning_rate": 9.402776752955155e-05,
+      "loss": 1.2068,
+      "step": 2630
+    },
+    {
+      "epoch": 0.15985175284039127,
+      "grad_norm": 0.23893888294696808,
+      "learning_rate": 9.402322977078366e-05,
+      "loss": 1.1305,
+      "step": 2631
+    },
+    {
+      "epoch": 0.1599125098730178,
+      "grad_norm": 0.18897725641727448,
+      "learning_rate": 9.401869039833577e-05,
+      "loss": 1.1491,
+      "step": 2632
+    },
+    {
+      "epoch": 0.15997326690564434,
+      "grad_norm": 0.2109520137310028,
+      "learning_rate": 9.40141494123743e-05,
+      "loss": 1.0789,
+      "step": 2633
+    },
+    {
+      "epoch": 0.16003402393827085,
+      "grad_norm": 0.417977899312973,
+      "learning_rate": 9.400960681306566e-05,
+      "loss": 1.4091,
+      "step": 2634
+    },
+    {
+      "epoch": 0.16009478097089738,
+      "grad_norm": 0.2826853394508362,
+      "learning_rate": 9.400506260057635e-05,
+      "loss": 1.149,
+      "step": 2635
+    },
+    {
+      "epoch": 0.16015553800352392,
+      "grad_norm": 1.1785639524459839,
+      "learning_rate": 9.4000516775073e-05,
+      "loss": 1.0705,
+      "step": 2636
+    },
+    {
+      "epoch": 0.16021629503615042,
+      "grad_norm": 0.15625526010990143,
+      "learning_rate": 9.399596933672217e-05,
+      "loss": 1.1168,
+      "step": 2637
+    },
+    {
+      "epoch": 0.16027705206877696,
+      "grad_norm": 0.3457401692867279,
+      "learning_rate": 9.399142028569061e-05,
+      "loss": 1.0788,
+      "step": 2638
+    },
+    {
+      "epoch": 0.1603378091014035,
+      "grad_norm": 0.2573912739753723,
+      "learning_rate": 9.398686962214502e-05,
+      "loss": 1.2239,
+      "step": 2639
+    },
+    {
+      "epoch": 0.16039856613403,
+      "grad_norm": 0.18996167182922363,
+      "learning_rate": 9.398231734625222e-05,
+      "loss": 1.0599,
+      "step": 2640
+    },
+    {
+      "epoch": 0.16045932316665654,
+      "grad_norm": 0.1542142927646637,
+      "learning_rate": 9.39777634581791e-05,
+      "loss": 1.0834,
+      "step": 2641
+    },
+    {
+      "epoch": 0.16052008019928307,
+      "grad_norm": 0.26509547233581543,
+      "learning_rate": 9.397320795809254e-05,
+      "loss": 1.3499,
+      "step": 2642
+    },
+    {
+      "epoch": 0.1605808372319096,
+      "grad_norm": 0.16394969820976257,
+      "learning_rate": 9.396865084615954e-05,
+      "loss": 1.0985,
+      "step": 2643
+    },
+    {
+      "epoch": 0.16064159426453611,
+      "grad_norm": 0.24615415930747986,
+      "learning_rate": 9.396409212254715e-05,
+      "loss": 1.1233,
+      "step": 2644
+    },
+    {
+      "epoch": 0.16070235129716265,
+      "grad_norm": 0.2274923473596573,
+      "learning_rate": 9.395953178742246e-05,
+      "loss": 1.1726,
+      "step": 2645
+    },
+    {
+      "epoch": 0.16076310832978918,
+      "grad_norm": 0.2049041986465454,
+      "learning_rate": 9.395496984095264e-05,
+      "loss": 1.0847,
+      "step": 2646
+    },
+    {
+      "epoch": 0.1608238653624157,
+      "grad_norm": 0.16849271953105927,
+      "learning_rate": 9.39504062833049e-05,
+      "loss": 1.2479,
+      "step": 2647
+    },
+    {
+      "epoch": 0.16088462239504223,
+      "grad_norm": 1.42002534866333,
+      "learning_rate": 9.394584111464653e-05,
+      "loss": 1.1433,
+      "step": 2648
+    },
+    {
+      "epoch": 0.16094537942766876,
+      "grad_norm": 0.21978677809238434,
+      "learning_rate": 9.394127433514485e-05,
+      "loss": 1.2329,
+      "step": 2649
+    },
+    {
+      "epoch": 0.16100613646029527,
+      "grad_norm": 0.2311258316040039,
+      "learning_rate": 9.393670594496727e-05,
+      "loss": 1.1341,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1610668934929218,
+      "grad_norm": 0.21976394951343536,
+      "learning_rate": 9.393213594428124e-05,
+      "loss": 1.1124,
+      "step": 2651
+    },
+    {
+      "epoch": 0.16112765052554834,
+      "grad_norm": 0.13613387942314148,
+      "learning_rate": 9.392756433325426e-05,
+      "loss": 1.0568,
+      "step": 2652
+    },
+    {
+      "epoch": 0.16118840755817485,
+      "grad_norm": 0.4879782795906067,
+      "learning_rate": 9.392299111205396e-05,
+      "loss": 1.3004,
+      "step": 2653
+    },
+    {
+      "epoch": 0.16124916459080138,
+      "grad_norm": 0.32171717286109924,
+      "learning_rate": 9.39184162808479e-05,
+      "loss": 1.1321,
+      "step": 2654
+    },
+    {
+      "epoch": 0.16130992162342792,
+      "grad_norm": 0.24977348744869232,
+      "learning_rate": 9.39138398398038e-05,
+      "loss": 1.1745,
+      "step": 2655
+    },
+    {
+      "epoch": 0.16137067865605445,
+      "grad_norm": 0.21851065754890442,
+      "learning_rate": 9.390926178908943e-05,
+      "loss": 1.2122,
+      "step": 2656
+    },
+    {
+      "epoch": 0.16143143568868096,
+      "grad_norm": 0.34062761068344116,
+      "learning_rate": 9.390468212887255e-05,
+      "loss": 1.0949,
+      "step": 2657
+    },
+    {
+      "epoch": 0.1614921927213075,
+      "grad_norm": 0.3069348633289337,
+      "learning_rate": 9.39001008593211e-05,
+      "loss": 1.0964,
+      "step": 2658
+    },
+    {
+      "epoch": 0.16155294975393403,
+      "grad_norm": 0.19767002761363983,
+      "learning_rate": 9.389551798060294e-05,
+      "loss": 1.087,
+      "step": 2659
+    },
+    {
+      "epoch": 0.16161370678656053,
+      "grad_norm": 0.32879358530044556,
+      "learning_rate": 9.38909334928861e-05,
+      "loss": 1.1335,
+      "step": 2660
+    },
+    {
+      "epoch": 0.16167446381918707,
+      "grad_norm": 0.5914991497993469,
+      "learning_rate": 9.388634739633859e-05,
+      "loss": 1.092,
+      "step": 2661
+    },
+    {
+      "epoch": 0.1617352208518136,
+      "grad_norm": 1.1937721967697144,
+      "learning_rate": 9.388175969112856e-05,
+      "loss": 1.0999,
+      "step": 2662
+    },
+    {
+      "epoch": 0.1617959778844401,
+      "grad_norm": 0.2987099587917328,
+      "learning_rate": 9.387717037742412e-05,
+      "loss": 1.1449,
+      "step": 2663
+    },
+    {
+      "epoch": 0.16185673491706665,
+      "grad_norm": 0.18860803544521332,
+      "learning_rate": 9.387257945539354e-05,
+      "loss": 1.0914,
+      "step": 2664
+    },
+    {
+      "epoch": 0.16191749194969318,
+      "grad_norm": 0.29970529675483704,
+      "learning_rate": 9.386798692520507e-05,
+      "loss": 1.1046,
+      "step": 2665
+    },
+    {
+      "epoch": 0.16197824898231972,
+      "grad_norm": 0.24461442232131958,
+      "learning_rate": 9.386339278702704e-05,
+      "loss": 1.1449,
+      "step": 2666
+    },
+    {
+      "epoch": 0.16203900601494622,
+      "grad_norm": 0.22977766394615173,
+      "learning_rate": 9.385879704102789e-05,
+      "loss": 1.0855,
+      "step": 2667
+    },
+    {
+      "epoch": 0.16209976304757276,
+      "grad_norm": 0.2146521955728531,
+      "learning_rate": 9.385419968737605e-05,
+      "loss": 1.114,
+      "step": 2668
+    },
+    {
+      "epoch": 0.1621605200801993,
+      "grad_norm": 0.18897241353988647,
+      "learning_rate": 9.384960072624005e-05,
+      "loss": 1.1655,
+      "step": 2669
+    },
+    {
+      "epoch": 0.1622212771128258,
+      "grad_norm": 0.16305464506149292,
+      "learning_rate": 9.384500015778846e-05,
+      "loss": 1.1466,
+      "step": 2670
+    },
+    {
+      "epoch": 0.16228203414545234,
+      "grad_norm": 0.20688949525356293,
+      "learning_rate": 9.384039798218991e-05,
+      "loss": 1.1434,
+      "step": 2671
+    },
+    {
+      "epoch": 0.16234279117807887,
+      "grad_norm": 0.18620194494724274,
+      "learning_rate": 9.38357941996131e-05,
+      "loss": 1.1145,
+      "step": 2672
+    },
+    {
+      "epoch": 0.16240354821070538,
+      "grad_norm": 0.2123836874961853,
+      "learning_rate": 9.383118881022679e-05,
+      "loss": 1.1196,
+      "step": 2673
+    },
+    {
+      "epoch": 0.1624643052433319,
+      "grad_norm": 0.23215189576148987,
+      "learning_rate": 9.382658181419977e-05,
+      "loss": 1.2262,
+      "step": 2674
+    },
+    {
+      "epoch": 0.16252506227595845,
+      "grad_norm": 2.866469383239746,
+      "learning_rate": 9.382197321170092e-05,
+      "loss": 1.1225,
+      "step": 2675
+    },
+    {
+      "epoch": 0.16258581930858496,
+      "grad_norm": 0.2372300624847412,
+      "learning_rate": 9.381736300289917e-05,
+      "loss": 1.1578,
+      "step": 2676
+    },
+    {
+      "epoch": 0.1626465763412115,
+      "grad_norm": 0.30157554149627686,
+      "learning_rate": 9.38127511879635e-05,
+      "loss": 1.1298,
+      "step": 2677
+    },
+    {
+      "epoch": 0.16270733337383803,
+      "grad_norm": 2.6343443393707275,
+      "learning_rate": 9.380813776706298e-05,
+      "loss": 1.1643,
+      "step": 2678
+    },
+    {
+      "epoch": 0.16276809040646456,
+      "grad_norm": 0.24585436284542084,
+      "learning_rate": 9.380352274036672e-05,
+      "loss": 1.2431,
+      "step": 2679
+    },
+    {
+      "epoch": 0.16282884743909107,
+      "grad_norm": 0.24145565927028656,
+      "learning_rate": 9.379890610804385e-05,
+      "loss": 1.212,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1628896044717176,
+      "grad_norm": 0.23217593133449554,
+      "learning_rate": 9.379428787026361e-05,
+      "loss": 1.1875,
+      "step": 2681
+    },
+    {
+      "epoch": 0.16295036150434414,
+      "grad_norm": 0.21196480095386505,
+      "learning_rate": 9.378966802719528e-05,
+      "loss": 1.0943,
+      "step": 2682
+    },
+    {
+      "epoch": 0.16301111853697065,
+      "grad_norm": 0.33148467540740967,
+      "learning_rate": 9.378504657900821e-05,
+      "loss": 1.0745,
+      "step": 2683
+    },
+    {
+      "epoch": 0.16307187556959718,
+      "grad_norm": 0.3658837378025055,
+      "learning_rate": 9.37804235258718e-05,
+      "loss": 1.1366,
+      "step": 2684
+    },
+    {
+      "epoch": 0.16313263260222371,
+      "grad_norm": 0.2708093225955963,
+      "learning_rate": 9.377579886795549e-05,
+      "loss": 1.1838,
+      "step": 2685
+    },
+    {
+      "epoch": 0.16319338963485022,
+      "grad_norm": 0.8164376616477966,
+      "learning_rate": 9.37711726054288e-05,
+      "loss": 1.2086,
+      "step": 2686
+    },
+    {
+      "epoch": 0.16325414666747676,
+      "grad_norm": 0.17520476877689362,
+      "learning_rate": 9.376654473846134e-05,
+      "loss": 1.1692,
+      "step": 2687
+    },
+    {
+      "epoch": 0.1633149037001033,
+      "grad_norm": 1.635270357131958,
+      "learning_rate": 9.376191526722272e-05,
+      "loss": 1.2304,
+      "step": 2688
+    },
+    {
+      "epoch": 0.16337566073272983,
+      "grad_norm": 0.19752848148345947,
+      "learning_rate": 9.375728419188264e-05,
+      "loss": 1.1778,
+      "step": 2689
+    },
+    {
+      "epoch": 0.16343641776535633,
+      "grad_norm": 0.5907459259033203,
+      "learning_rate": 9.375265151261083e-05,
+      "loss": 1.1286,
+      "step": 2690
+    },
+    {
+      "epoch": 0.16349717479798287,
+      "grad_norm": 0.5910320281982422,
+      "learning_rate": 9.374801722957714e-05,
+      "loss": 1.0852,
+      "step": 2691
+    },
+    {
+      "epoch": 0.1635579318306094,
+      "grad_norm": 0.19244112074375153,
+      "learning_rate": 9.374338134295141e-05,
+      "loss": 1.061,
+      "step": 2692
+    },
+    {
+      "epoch": 0.1636186888632359,
+      "grad_norm": 0.1832464039325714,
+      "learning_rate": 9.373874385290358e-05,
+      "loss": 1.1541,
+      "step": 2693
+    },
+    {
+      "epoch": 0.16367944589586245,
+      "grad_norm": 0.2991447150707245,
+      "learning_rate": 9.373410475960364e-05,
+      "loss": 1.0924,
+      "step": 2694
+    },
+    {
+      "epoch": 0.16374020292848898,
+      "grad_norm": 0.25456613302230835,
+      "learning_rate": 9.372946406322164e-05,
+      "loss": 1.0989,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1638009599611155,
+      "grad_norm": 0.21387642621994019,
+      "learning_rate": 9.37248217639277e-05,
+      "loss": 1.1089,
+      "step": 2696
+    },
+    {
+      "epoch": 0.16386171699374202,
+      "grad_norm": 0.26866665482521057,
+      "learning_rate": 9.372017786189195e-05,
+      "loss": 1.2257,
+      "step": 2697
+    },
+    {
+      "epoch": 0.16392247402636856,
+      "grad_norm": 0.2666727602481842,
+      "learning_rate": 9.371553235728463e-05,
+      "loss": 1.1979,
+      "step": 2698
+    },
+    {
+      "epoch": 0.16398323105899507,
+      "grad_norm": 0.20347638428211212,
+      "learning_rate": 9.371088525027602e-05,
+      "loss": 1.1374,
+      "step": 2699
+    },
+    {
+      "epoch": 0.1640439880916216,
+      "grad_norm": 0.3014492690563202,
+      "learning_rate": 9.370623654103646e-05,
+      "loss": 1.1903,
+      "step": 2700
+    },
+    {
+      "epoch": 0.16410474512424814,
+      "grad_norm": 0.7942011952400208,
+      "learning_rate": 9.370158622973637e-05,
+      "loss": 1.1665,
+      "step": 2701
+    },
+    {
+      "epoch": 0.16416550215687467,
+      "grad_norm": 0.3673689067363739,
+      "learning_rate": 9.369693431654617e-05,
+      "loss": 1.1997,
+      "step": 2702
+    },
+    {
+      "epoch": 0.16422625918950118,
+      "grad_norm": 0.22094377875328064,
+      "learning_rate": 9.36922808016364e-05,
+      "loss": 1.1707,
+      "step": 2703
+    },
+    {
+      "epoch": 0.1642870162221277,
+      "grad_norm": 0.22264212369918823,
+      "learning_rate": 9.368762568517763e-05,
+      "loss": 1.0806,
+      "step": 2704
+    },
+    {
+      "epoch": 0.16434777325475425,
+      "grad_norm": 0.3930180072784424,
+      "learning_rate": 9.36829689673405e-05,
+      "loss": 1.0871,
+      "step": 2705
+    },
+    {
+      "epoch": 0.16440853028738076,
+      "grad_norm": 0.1813138872385025,
+      "learning_rate": 9.367831064829571e-05,
+      "loss": 1.127,
+      "step": 2706
+    },
+    {
+      "epoch": 0.1644692873200073,
+      "grad_norm": 2.13755464553833,
+      "learning_rate": 9.367365072821397e-05,
+      "loss": 1.167,
+      "step": 2707
+    },
+    {
+      "epoch": 0.16453004435263383,
+      "grad_norm": 0.2634475529193878,
+      "learning_rate": 9.366898920726615e-05,
+      "loss": 1.1304,
+      "step": 2708
+    },
+    {
+      "epoch": 0.16459080138526033,
+      "grad_norm": 0.1927611529827118,
+      "learning_rate": 9.366432608562308e-05,
+      "loss": 1.0681,
+      "step": 2709
+    },
+    {
+      "epoch": 0.16465155841788687,
+      "grad_norm": 0.1683140993118286,
+      "learning_rate": 9.365966136345569e-05,
+      "loss": 1.1105,
+      "step": 2710
+    },
+    {
+      "epoch": 0.1647123154505134,
+      "grad_norm": 0.23627561330795288,
+      "learning_rate": 9.365499504093498e-05,
+      "loss": 1.2114,
+      "step": 2711
+    },
+    {
+      "epoch": 0.1647730724831399,
+      "grad_norm": 0.696429967880249,
+      "learning_rate": 9.365032711823198e-05,
+      "loss": 1.3981,
+      "step": 2712
+    },
+    {
+      "epoch": 0.16483382951576644,
+      "grad_norm": 0.32340294122695923,
+      "learning_rate": 9.364565759551782e-05,
+      "loss": 1.1709,
+      "step": 2713
+    },
+    {
+      "epoch": 0.16489458654839298,
+      "grad_norm": 0.39964428544044495,
+      "learning_rate": 9.364098647296364e-05,
+      "loss": 1.1139,
+      "step": 2714
+    },
+    {
+      "epoch": 0.16495534358101951,
+      "grad_norm": 0.44005492329597473,
+      "learning_rate": 9.363631375074065e-05,
+      "loss": 1.131,
+      "step": 2715
+    },
+    {
+      "epoch": 0.16501610061364602,
+      "grad_norm": 0.7380490899085999,
+      "learning_rate": 9.363163942902016e-05,
+      "loss": 1.1971,
+      "step": 2716
+    },
+    {
+      "epoch": 0.16507685764627256,
+      "grad_norm": 0.45174384117126465,
+      "learning_rate": 9.362696350797348e-05,
+      "loss": 1.1194,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1651376146788991,
+      "grad_norm": 0.31846198439598083,
+      "learning_rate": 9.362228598777203e-05,
+      "loss": 1.2118,
+      "step": 2718
+    },
+    {
+      "epoch": 0.1651983717115256,
+      "grad_norm": 0.24986636638641357,
+      "learning_rate": 9.361760686858725e-05,
+      "loss": 1.1776,
+      "step": 2719
+    },
+    {
+      "epoch": 0.16525912874415213,
+      "grad_norm": 0.20426741242408752,
+      "learning_rate": 9.361292615059065e-05,
+      "loss": 1.1528,
+      "step": 2720
+    },
+    {
+      "epoch": 0.16531988577677867,
+      "grad_norm": 0.19015654921531677,
+      "learning_rate": 9.360824383395383e-05,
+      "loss": 1.1477,
+      "step": 2721
+    },
+    {
+      "epoch": 0.16538064280940518,
+      "grad_norm": 0.2341787964105606,
+      "learning_rate": 9.360355991884839e-05,
+      "loss": 1.0444,
+      "step": 2722
+    },
+    {
+      "epoch": 0.1654413998420317,
+      "grad_norm": 0.34809285402297974,
+      "learning_rate": 9.359887440544605e-05,
+      "loss": 1.2855,
+      "step": 2723
+    },
+    {
+      "epoch": 0.16550215687465825,
+      "grad_norm": 0.20901837944984436,
+      "learning_rate": 9.359418729391852e-05,
+      "loss": 1.2398,
+      "step": 2724
+    },
+    {
+      "epoch": 0.16556291390728478,
+      "grad_norm": 0.20196299254894257,
+      "learning_rate": 9.358949858443764e-05,
+      "loss": 1.1097,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1656236709399113,
+      "grad_norm": 0.3473243713378906,
+      "learning_rate": 9.358480827717526e-05,
+      "loss": 1.2743,
+      "step": 2726
+    },
+    {
+      "epoch": 0.16568442797253782,
+      "grad_norm": 0.22709813714027405,
+      "learning_rate": 9.358011637230331e-05,
+      "loss": 1.1176,
+      "step": 2727
+    },
+    {
+      "epoch": 0.16574518500516436,
+      "grad_norm": 0.9128461480140686,
+      "learning_rate": 9.357542286999377e-05,
+      "loss": 1.1398,
+      "step": 2728
+    },
+    {
+      "epoch": 0.16580594203779087,
+      "grad_norm": 0.31940165162086487,
+      "learning_rate": 9.357072777041869e-05,
+      "loss": 1.0853,
+      "step": 2729
+    },
+    {
+      "epoch": 0.1658666990704174,
+      "grad_norm": 0.3543567657470703,
+      "learning_rate": 9.356603107375015e-05,
+      "loss": 1.3009,
+      "step": 2730
+    },
+    {
+      "epoch": 0.16592745610304394,
+      "grad_norm": 0.3245181441307068,
+      "learning_rate": 9.356133278016033e-05,
+      "loss": 1.1683,
+      "step": 2731
+    },
+    {
+      "epoch": 0.16598821313567044,
+      "grad_norm": 0.2407148778438568,
+      "learning_rate": 9.355663288982141e-05,
+      "loss": 1.0895,
+      "step": 2732
+    },
+    {
+      "epoch": 0.16604897016829698,
+      "grad_norm": 0.476728230714798,
+      "learning_rate": 9.355193140290573e-05,
+      "loss": 1.3047,
+      "step": 2733
+    },
+    {
+      "epoch": 0.1661097272009235,
+      "grad_norm": 0.13437965512275696,
+      "learning_rate": 9.354722831958556e-05,
+      "loss": 1.0877,
+      "step": 2734
+    },
+    {
+      "epoch": 0.16617048423355002,
+      "grad_norm": 0.6698725819587708,
+      "learning_rate": 9.354252364003333e-05,
+      "loss": 1.2105,
+      "step": 2735
+    },
+    {
+      "epoch": 0.16623124126617655,
+      "grad_norm": 0.3112938702106476,
+      "learning_rate": 9.353781736442146e-05,
+      "loss": 1.0774,
+      "step": 2736
+    },
+    {
+      "epoch": 0.1662919982988031,
+      "grad_norm": 0.27897047996520996,
+      "learning_rate": 9.353310949292251e-05,
+      "loss": 1.132,
+      "step": 2737
+    },
+    {
+      "epoch": 0.16635275533142962,
+      "grad_norm": 0.26251503825187683,
+      "learning_rate": 9.3528400025709e-05,
+      "loss": 1.1084,
+      "step": 2738
+    },
+    {
+      "epoch": 0.16641351236405613,
+      "grad_norm": 1.4709620475769043,
+      "learning_rate": 9.352368896295356e-05,
+      "loss": 1.1027,
+      "step": 2739
+    },
+    {
+      "epoch": 0.16647426939668267,
+      "grad_norm": 0.2612626850605011,
+      "learning_rate": 9.351897630482891e-05,
+      "loss": 1.2431,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1665350264293092,
+      "grad_norm": 0.5787556171417236,
+      "learning_rate": 9.351426205150777e-05,
+      "loss": 1.1367,
+      "step": 2741
+    },
+    {
+      "epoch": 0.1665957834619357,
+      "grad_norm": 0.34013068675994873,
+      "learning_rate": 9.350954620316293e-05,
+      "loss": 1.1198,
+      "step": 2742
+    },
+    {
+      "epoch": 0.16665654049456224,
+      "grad_norm": 0.5343618392944336,
+      "learning_rate": 9.350482875996726e-05,
+      "loss": 1.2055,
+      "step": 2743
+    },
+    {
+      "epoch": 0.16671729752718878,
+      "grad_norm": 0.49730050563812256,
+      "learning_rate": 9.350010972209371e-05,
+      "loss": 1.2903,
+      "step": 2744
+    },
+    {
+      "epoch": 0.1667780545598153,
+      "grad_norm": 0.2530938982963562,
+      "learning_rate": 9.349538908971521e-05,
+      "loss": 1.1995,
+      "step": 2745
+    },
+    {
+      "epoch": 0.16683881159244182,
+      "grad_norm": 1.39519464969635,
+      "learning_rate": 9.349066686300483e-05,
+      "loss": 1.2048,
+      "step": 2746
+    },
+    {
+      "epoch": 0.16689956862506836,
+      "grad_norm": 0.37702691555023193,
+      "learning_rate": 9.348594304213562e-05,
+      "loss": 1.3088,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1669603256576949,
+      "grad_norm": 0.2868204712867737,
+      "learning_rate": 9.348121762728079e-05,
+      "loss": 1.1414,
+      "step": 2748
+    },
+    {
+      "epoch": 0.1670210826903214,
+      "grad_norm": 0.40916576981544495,
+      "learning_rate": 9.347649061861352e-05,
+      "loss": 1.2217,
+      "step": 2749
+    },
+    {
+      "epoch": 0.16708183972294793,
+      "grad_norm": 0.2935818135738373,
+      "learning_rate": 9.347176201630707e-05,
+      "loss": 1.1218,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16714259675557447,
+      "grad_norm": 0.21246002614498138,
+      "learning_rate": 9.346703182053479e-05,
+      "loss": 1.1828,
+      "step": 2751
+    },
+    {
+      "epoch": 0.16720335378820098,
+      "grad_norm": 0.2406328022480011,
+      "learning_rate": 9.346230003147005e-05,
+      "loss": 1.2039,
+      "step": 2752
+    },
+    {
+      "epoch": 0.1672641108208275,
+      "grad_norm": 0.3806989789009094,
+      "learning_rate": 9.345756664928631e-05,
+      "loss": 1.1006,
+      "step": 2753
+    },
+    {
+      "epoch": 0.16732486785345405,
+      "grad_norm": 0.15987083315849304,
+      "learning_rate": 9.345283167415706e-05,
+      "loss": 1.1102,
+      "step": 2754
+    },
+    {
+      "epoch": 0.16738562488608055,
+      "grad_norm": 0.16697901487350464,
+      "learning_rate": 9.344809510625586e-05,
+      "loss": 1.0585,
+      "step": 2755
+    },
+    {
+      "epoch": 0.1674463819187071,
+      "grad_norm": 0.3083542585372925,
+      "learning_rate": 9.344335694575634e-05,
+      "loss": 1.1051,
+      "step": 2756
+    },
+    {
+      "epoch": 0.16750713895133362,
+      "grad_norm": 0.26503217220306396,
+      "learning_rate": 9.343861719283217e-05,
+      "loss": 1.0784,
+      "step": 2757
+    },
+    {
+      "epoch": 0.16756789598396013,
+      "grad_norm": 0.16039438545703888,
+      "learning_rate": 9.34338758476571e-05,
+      "loss": 1.1258,
+      "step": 2758
+    },
+    {
+      "epoch": 0.16762865301658667,
+      "grad_norm": 0.20033229887485504,
+      "learning_rate": 9.342913291040491e-05,
+      "loss": 1.1395,
+      "step": 2759
+    },
+    {
+      "epoch": 0.1676894100492132,
+      "grad_norm": 0.23209625482559204,
+      "learning_rate": 9.342438838124945e-05,
+      "loss": 1.0981,
+      "step": 2760
+    },
+    {
+      "epoch": 0.16775016708183974,
+      "grad_norm": 0.6559396386146545,
+      "learning_rate": 9.341964226036466e-05,
+      "loss": 1.0904,
+      "step": 2761
+    },
+    {
+      "epoch": 0.16781092411446624,
+      "grad_norm": 0.26381775736808777,
+      "learning_rate": 9.341489454792447e-05,
+      "loss": 1.2064,
+      "step": 2762
+    },
+    {
+      "epoch": 0.16787168114709278,
+      "grad_norm": 0.2045595496892929,
+      "learning_rate": 9.341014524410295e-05,
+      "loss": 1.1783,
+      "step": 2763
+    },
+    {
+      "epoch": 0.1679324381797193,
+      "grad_norm": 0.2016468644142151,
+      "learning_rate": 9.340539434907414e-05,
+      "loss": 1.112,
+      "step": 2764
+    },
+    {
+      "epoch": 0.16799319521234582,
+      "grad_norm": 0.18968884646892548,
+      "learning_rate": 9.340064186301221e-05,
+      "loss": 1.0729,
+      "step": 2765
+    },
+    {
+      "epoch": 0.16805395224497235,
+      "grad_norm": 0.12279792875051498,
+      "learning_rate": 9.339588778609138e-05,
+      "loss": 1.0543,
+      "step": 2766
+    },
+    {
+      "epoch": 0.1681147092775989,
+      "grad_norm": 0.210651233792305,
+      "learning_rate": 9.33911321184859e-05,
+      "loss": 1.204,
+      "step": 2767
+    },
+    {
+      "epoch": 0.1681754663102254,
+      "grad_norm": 0.30051079392433167,
+      "learning_rate": 9.338637486037007e-05,
+      "loss": 1.2788,
+      "step": 2768
+    },
+    {
+      "epoch": 0.16823622334285193,
+      "grad_norm": 0.21153174340724945,
+      "learning_rate": 9.338161601191828e-05,
+      "loss": 1.1877,
+      "step": 2769
+    },
+    {
+      "epoch": 0.16829698037547847,
+      "grad_norm": 0.18878796696662903,
+      "learning_rate": 9.337685557330497e-05,
+      "loss": 1.0766,
+      "step": 2770
+    },
+    {
+      "epoch": 0.168357737408105,
+      "grad_norm": 0.1966601461172104,
+      "learning_rate": 9.337209354470467e-05,
+      "loss": 1.117,
+      "step": 2771
+    },
+    {
+      "epoch": 0.1684184944407315,
+      "grad_norm": 0.3076675236225128,
+      "learning_rate": 9.336732992629186e-05,
+      "loss": 1.1717,
+      "step": 2772
+    },
+    {
+      "epoch": 0.16847925147335804,
+      "grad_norm": 0.15472185611724854,
+      "learning_rate": 9.33625647182412e-05,
+      "loss": 1.0859,
+      "step": 2773
+    },
+    {
+      "epoch": 0.16854000850598458,
+      "grad_norm": 0.1461690366268158,
+      "learning_rate": 9.335779792072734e-05,
+      "loss": 1.0672,
+      "step": 2774
+    },
+    {
+      "epoch": 0.16860076553861109,
+      "grad_norm": 0.577778697013855,
+      "learning_rate": 9.335302953392501e-05,
+      "loss": 1.2297,
+      "step": 2775
+    },
+    {
+      "epoch": 0.16866152257123762,
+      "grad_norm": 2.478369951248169,
+      "learning_rate": 9.334825955800904e-05,
+      "loss": 1.0548,
+      "step": 2776
+    },
+    {
+      "epoch": 0.16872227960386416,
+      "grad_norm": 0.3680042624473572,
+      "learning_rate": 9.33434879931542e-05,
+      "loss": 1.2124,
+      "step": 2777
+    },
+    {
+      "epoch": 0.16878303663649066,
+      "grad_norm": 0.23315078020095825,
+      "learning_rate": 9.333871483953544e-05,
+      "loss": 1.1652,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1688437936691172,
+      "grad_norm": 9.248869895935059,
+      "learning_rate": 9.333394009732772e-05,
+      "loss": 1.3086,
+      "step": 2779
+    },
+    {
+      "epoch": 0.16890455070174373,
+      "grad_norm": 0.1792900711297989,
+      "learning_rate": 9.332916376670603e-05,
+      "loss": 1.124,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16896530773437024,
+      "grad_norm": 0.29281532764434814,
+      "learning_rate": 9.332438584784549e-05,
+      "loss": 1.1318,
+      "step": 2781
+    },
+    {
+      "epoch": 0.16902606476699678,
+      "grad_norm": 0.14941686391830444,
+      "learning_rate": 9.33196063409212e-05,
+      "loss": 1.0851,
+      "step": 2782
+    },
+    {
+      "epoch": 0.1690868217996233,
+      "grad_norm": 0.30175021290779114,
+      "learning_rate": 9.331482524610837e-05,
+      "loss": 1.0731,
+      "step": 2783
+    },
+    {
+      "epoch": 0.16914757883224985,
+      "grad_norm": 0.2641928493976593,
+      "learning_rate": 9.331004256358224e-05,
+      "loss": 1.0891,
+      "step": 2784
+    },
+    {
+      "epoch": 0.16920833586487635,
+      "grad_norm": 0.15149296820163727,
+      "learning_rate": 9.330525829351813e-05,
+      "loss": 1.0991,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1692690928975029,
+      "grad_norm": 0.2485985904932022,
+      "learning_rate": 9.330047243609142e-05,
+      "loss": 1.2373,
+      "step": 2786
+    },
+    {
+      "epoch": 0.16932984993012942,
+      "grad_norm": 0.2151057869195938,
+      "learning_rate": 9.32956849914775e-05,
+      "loss": 1.1282,
+      "step": 2787
+    },
+    {
+      "epoch": 0.16939060696275593,
+      "grad_norm": 0.2067212611436844,
+      "learning_rate": 9.329089595985189e-05,
+      "loss": 1.2167,
+      "step": 2788
+    },
+    {
+      "epoch": 0.16945136399538246,
+      "grad_norm": 0.19152012467384338,
+      "learning_rate": 9.328610534139013e-05,
+      "loss": 1.1732,
+      "step": 2789
+    },
+    {
+      "epoch": 0.169512121028009,
+      "grad_norm": 0.26229506731033325,
+      "learning_rate": 9.32813131362678e-05,
+      "loss": 1.1492,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1695728780606355,
+      "grad_norm": 0.15380549430847168,
+      "learning_rate": 9.327651934466057e-05,
+      "loss": 1.0588,
+      "step": 2791
+    },
+    {
+      "epoch": 0.16963363509326204,
+      "grad_norm": 0.8333106637001038,
+      "learning_rate": 9.327172396674417e-05,
+      "loss": 1.109,
+      "step": 2792
+    },
+    {
+      "epoch": 0.16969439212588858,
+      "grad_norm": 0.16392317414283752,
+      "learning_rate": 9.326692700269434e-05,
+      "loss": 1.0875,
+      "step": 2793
+    },
+    {
+      "epoch": 0.1697551491585151,
+      "grad_norm": 0.14444230496883392,
+      "learning_rate": 9.326212845268695e-05,
+      "loss": 1.0835,
+      "step": 2794
+    },
+    {
+      "epoch": 0.16981590619114162,
+      "grad_norm": 0.13465823233127594,
+      "learning_rate": 9.325732831689791e-05,
+      "loss": 1.0591,
+      "step": 2795
+    },
+    {
+      "epoch": 0.16987666322376815,
+      "grad_norm": 0.244447261095047,
+      "learning_rate": 9.325252659550309e-05,
+      "loss": 1.0747,
+      "step": 2796
+    },
+    {
+      "epoch": 0.1699374202563947,
+      "grad_norm": 0.2744373679161072,
+      "learning_rate": 9.324772328867858e-05,
+      "loss": 1.0442,
+      "step": 2797
+    },
+    {
+      "epoch": 0.1699981772890212,
+      "grad_norm": 0.9067546725273132,
+      "learning_rate": 9.32429183966004e-05,
+      "loss": 1.0712,
+      "step": 2798
+    },
+    {
+      "epoch": 0.17005893432164773,
+      "grad_norm": 0.1323540359735489,
+      "learning_rate": 9.323811191944469e-05,
+      "loss": 1.0656,
+      "step": 2799
+    },
+    {
+      "epoch": 0.17011969135427427,
+      "grad_norm": 0.1757102906703949,
+      "learning_rate": 9.323330385738762e-05,
+      "loss": 1.1109,
+      "step": 2800
+    },
+    {
+      "epoch": 0.17018044838690077,
+      "grad_norm": 0.4943138659000397,
+      "learning_rate": 9.322849421060546e-05,
+      "loss": 1.1901,
+      "step": 2801
+    },
+    {
+      "epoch": 0.1702412054195273,
+      "grad_norm": 8.577896118164062,
+      "learning_rate": 9.322368297927448e-05,
+      "loss": 1.1125,
+      "step": 2802
+    },
+    {
+      "epoch": 0.17030196245215384,
+      "grad_norm": 0.20444080233573914,
+      "learning_rate": 9.321887016357104e-05,
+      "loss": 1.1183,
+      "step": 2803
+    },
+    {
+      "epoch": 0.17036271948478035,
+      "grad_norm": 0.24041442573070526,
+      "learning_rate": 9.321405576367157e-05,
+      "loss": 1.3482,
+      "step": 2804
+    },
+    {
+      "epoch": 0.17042347651740689,
+      "grad_norm": 0.5964112281799316,
+      "learning_rate": 9.320923977975253e-05,
+      "loss": 1.2177,
+      "step": 2805
+    },
+    {
+      "epoch": 0.17048423355003342,
+      "grad_norm": 7.097474575042725,
+      "learning_rate": 9.320442221199044e-05,
+      "loss": 1.2228,
+      "step": 2806
+    },
+    {
+      "epoch": 0.17054499058265996,
+      "grad_norm": 0.2913483679294586,
+      "learning_rate": 9.319960306056192e-05,
+      "loss": 1.1019,
+      "step": 2807
+    },
+    {
+      "epoch": 0.17060574761528646,
+      "grad_norm": 0.2946627736091614,
+      "learning_rate": 9.319478232564359e-05,
+      "loss": 1.0749,
+      "step": 2808
+    },
+    {
+      "epoch": 0.170666504647913,
+      "grad_norm": 0.2581668198108673,
+      "learning_rate": 9.318996000741215e-05,
+      "loss": 1.2164,
+      "step": 2809
+    },
+    {
+      "epoch": 0.17072726168053953,
+      "grad_norm": 0.3176397681236267,
+      "learning_rate": 9.31851361060444e-05,
+      "loss": 1.1545,
+      "step": 2810
+    },
+    {
+      "epoch": 0.17078801871316604,
+      "grad_norm": 0.7890453338623047,
+      "learning_rate": 9.318031062171715e-05,
+      "loss": 1.304,
+      "step": 2811
+    },
+    {
+      "epoch": 0.17084877574579257,
+      "grad_norm": 0.3896808326244354,
+      "learning_rate": 9.317548355460725e-05,
+      "loss": 1.3385,
+      "step": 2812
+    },
+    {
+      "epoch": 0.1709095327784191,
+      "grad_norm": 0.20673814415931702,
+      "learning_rate": 9.317065490489165e-05,
+      "loss": 1.149,
+      "step": 2813
+    },
+    {
+      "epoch": 0.17097028981104562,
+      "grad_norm": 4.286653995513916,
+      "learning_rate": 9.316582467274735e-05,
+      "loss": 1.1895,
+      "step": 2814
+    },
+    {
+      "epoch": 0.17103104684367215,
+      "grad_norm": 0.24091972410678864,
+      "learning_rate": 9.316099285835141e-05,
+      "loss": 1.1407,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1710918038762987,
+      "grad_norm": 0.23865258693695068,
+      "learning_rate": 9.315615946188093e-05,
+      "loss": 1.1443,
+      "step": 2816
+    },
+    {
+      "epoch": 0.1711525609089252,
+      "grad_norm": 0.25635668635368347,
+      "learning_rate": 9.315132448351309e-05,
+      "loss": 1.1026,
+      "step": 2817
+    },
+    {
+      "epoch": 0.17121331794155173,
+      "grad_norm": 0.2651605010032654,
+      "learning_rate": 9.314648792342511e-05,
+      "loss": 1.0955,
+      "step": 2818
+    },
+    {
+      "epoch": 0.17127407497417826,
+      "grad_norm": 0.3693333864212036,
+      "learning_rate": 9.314164978179427e-05,
+      "loss": 1.1378,
+      "step": 2819
+    },
+    {
+      "epoch": 0.1713348320068048,
+      "grad_norm": 3.0835492610931396,
+      "learning_rate": 9.313681005879791e-05,
+      "loss": 1.0815,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1713955890394313,
+      "grad_norm": 0.3346802890300751,
+      "learning_rate": 9.313196875461345e-05,
+      "loss": 1.1228,
+      "step": 2821
+    },
+    {
+      "epoch": 0.17145634607205784,
+      "grad_norm": 53.88895797729492,
+      "learning_rate": 9.312712586941834e-05,
+      "loss": 1.1496,
+      "step": 2822
+    },
+    {
+      "epoch": 0.17151710310468438,
+      "grad_norm": 0.1832890808582306,
+      "learning_rate": 9.312228140339009e-05,
+      "loss": 1.0758,
+      "step": 2823
+    },
+    {
+      "epoch": 0.17157786013731088,
+      "grad_norm": 0.2312437742948532,
+      "learning_rate": 9.311743535670628e-05,
+      "loss": 1.1014,
+      "step": 2824
+    },
+    {
+      "epoch": 0.17163861716993742,
+      "grad_norm": 0.1523139476776123,
+      "learning_rate": 9.311258772954454e-05,
+      "loss": 1.0823,
+      "step": 2825
+    },
+    {
+      "epoch": 0.17169937420256395,
+      "grad_norm": 0.20451554656028748,
+      "learning_rate": 9.310773852208256e-05,
+      "loss": 1.1055,
+      "step": 2826
+    },
+    {
+      "epoch": 0.17176013123519046,
+      "grad_norm": 0.16789209842681885,
+      "learning_rate": 9.310288773449811e-05,
+      "loss": 1.1059,
+      "step": 2827
+    },
+    {
+      "epoch": 0.171820888267817,
+      "grad_norm": 0.21686002612113953,
+      "learning_rate": 9.309803536696897e-05,
+      "loss": 1.2283,
+      "step": 2828
+    },
+    {
+      "epoch": 0.17188164530044353,
+      "grad_norm": 0.18368317186832428,
+      "learning_rate": 9.3093181419673e-05,
+      "loss": 1.1682,
+      "step": 2829
+    },
+    {
+      "epoch": 0.17194240233307007,
+      "grad_norm": 0.19566959142684937,
+      "learning_rate": 9.308832589278815e-05,
+      "loss": 1.1125,
+      "step": 2830
+    },
+    {
+      "epoch": 0.17200315936569657,
+      "grad_norm": 0.17353063821792603,
+      "learning_rate": 9.308346878649238e-05,
+      "loss": 1.1333,
+      "step": 2831
+    },
+    {
+      "epoch": 0.1720639163983231,
+      "grad_norm": 0.1296449452638626,
+      "learning_rate": 9.307861010096372e-05,
+      "loss": 1.0655,
+      "step": 2832
+    },
+    {
+      "epoch": 0.17212467343094964,
+      "grad_norm": 0.17133085429668427,
+      "learning_rate": 9.30737498363803e-05,
+      "loss": 1.1367,
+      "step": 2833
+    },
+    {
+      "epoch": 0.17218543046357615,
+      "grad_norm": 0.16282299160957336,
+      "learning_rate": 9.306888799292027e-05,
+      "loss": 1.077,
+      "step": 2834
+    },
+    {
+      "epoch": 0.17224618749620269,
+      "grad_norm": 0.20644213259220123,
+      "learning_rate": 9.306402457076178e-05,
+      "loss": 1.148,
+      "step": 2835
+    },
+    {
+      "epoch": 0.17230694452882922,
+      "grad_norm": 0.14916478097438812,
+      "learning_rate": 9.305915957008317e-05,
+      "loss": 1.0827,
+      "step": 2836
+    },
+    {
+      "epoch": 0.17236770156145573,
+      "grad_norm": 0.2374756783246994,
+      "learning_rate": 9.305429299106275e-05,
+      "loss": 1.0482,
+      "step": 2837
+    },
+    {
+      "epoch": 0.17242845859408226,
+      "grad_norm": 0.1406780183315277,
+      "learning_rate": 9.304942483387889e-05,
+      "loss": 1.0751,
+      "step": 2838
+    },
+    {
+      "epoch": 0.1724892156267088,
+      "grad_norm": 0.26799002289772034,
+      "learning_rate": 9.304455509871004e-05,
+      "loss": 1.1306,
+      "step": 2839
+    },
+    {
+      "epoch": 0.1725499726593353,
+      "grad_norm": 0.43531063199043274,
+      "learning_rate": 9.303968378573471e-05,
+      "loss": 1.1339,
+      "step": 2840
+    },
+    {
+      "epoch": 0.17261072969196184,
+      "grad_norm": 0.24498429894447327,
+      "learning_rate": 9.303481089513144e-05,
+      "loss": 1.2976,
+      "step": 2841
+    },
+    {
+      "epoch": 0.17267148672458837,
+      "grad_norm": 0.3665919899940491,
+      "learning_rate": 9.302993642707887e-05,
+      "loss": 1.1735,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1727322437572149,
+      "grad_norm": 1.2711642980575562,
+      "learning_rate": 9.302506038175565e-05,
+      "loss": 1.2278,
+      "step": 2843
+    },
+    {
+      "epoch": 0.17279300078984142,
+      "grad_norm": 0.21386763453483582,
+      "learning_rate": 9.302018275934053e-05,
+      "loss": 1.0788,
+      "step": 2844
+    },
+    {
+      "epoch": 0.17285375782246795,
+      "grad_norm": 0.2421104460954666,
+      "learning_rate": 9.30153035600123e-05,
+      "loss": 1.1543,
+      "step": 2845
+    },
+    {
+      "epoch": 0.1729145148550945,
+      "grad_norm": 0.21388423442840576,
+      "learning_rate": 9.30104227839498e-05,
+      "loss": 1.1734,
+      "step": 2846
+    },
+    {
+      "epoch": 0.172975271887721,
+      "grad_norm": 0.21080902218818665,
+      "learning_rate": 9.300554043133194e-05,
+      "loss": 1.0814,
+      "step": 2847
+    },
+    {
+      "epoch": 0.17303602892034753,
+      "grad_norm": 0.2629931569099426,
+      "learning_rate": 9.300065650233767e-05,
+      "loss": 1.0584,
+      "step": 2848
+    },
+    {
+      "epoch": 0.17309678595297406,
+      "grad_norm": 0.22488714754581451,
+      "learning_rate": 9.299577099714605e-05,
+      "loss": 1.258,
+      "step": 2849
+    },
+    {
+      "epoch": 0.17315754298560057,
+      "grad_norm": 0.28502562642097473,
+      "learning_rate": 9.299088391593611e-05,
+      "loss": 1.0994,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1732183000182271,
+      "grad_norm": 0.4765513241291046,
+      "learning_rate": 9.298599525888703e-05,
+      "loss": 1.1479,
+      "step": 2851
+    },
+    {
+      "epoch": 0.17327905705085364,
+      "grad_norm": 0.20497465133666992,
+      "learning_rate": 9.298110502617796e-05,
+      "loss": 1.1424,
+      "step": 2852
+    },
+    {
+      "epoch": 0.17333981408348018,
+      "grad_norm": 0.34637463092803955,
+      "learning_rate": 9.29762132179882e-05,
+      "loss": 1.1249,
+      "step": 2853
+    },
+    {
+      "epoch": 0.17340057111610668,
+      "grad_norm": 0.8920005559921265,
+      "learning_rate": 9.297131983449702e-05,
+      "loss": 1.2085,
+      "step": 2854
+    },
+    {
+      "epoch": 0.17346132814873322,
+      "grad_norm": 0.24549148976802826,
+      "learning_rate": 9.296642487588381e-05,
+      "loss": 1.1641,
+      "step": 2855
+    },
+    {
+      "epoch": 0.17352208518135975,
+      "grad_norm": 0.23367923498153687,
+      "learning_rate": 9.296152834232801e-05,
+      "loss": 1.0847,
+      "step": 2856
+    },
+    {
+      "epoch": 0.17358284221398626,
+      "grad_norm": 0.39331212639808655,
+      "learning_rate": 9.295663023400907e-05,
+      "loss": 1.2004,
+      "step": 2857
+    },
+    {
+      "epoch": 0.1736435992466128,
+      "grad_norm": 0.3297134339809418,
+      "learning_rate": 9.295173055110653e-05,
+      "loss": 1.3704,
+      "step": 2858
+    },
+    {
+      "epoch": 0.17370435627923933,
+      "grad_norm": 0.2672659456729889,
+      "learning_rate": 9.294682929380003e-05,
+      "loss": 1.274,
+      "step": 2859
+    },
+    {
+      "epoch": 0.17376511331186584,
+      "grad_norm": 0.2202669084072113,
+      "learning_rate": 9.29419264622692e-05,
+      "loss": 1.0505,
+      "step": 2860
+    },
+    {
+      "epoch": 0.17382587034449237,
+      "grad_norm": 0.22581006586551666,
+      "learning_rate": 9.293702205669373e-05,
+      "loss": 1.2409,
+      "step": 2861
+    },
+    {
+      "epoch": 0.1738866273771189,
+      "grad_norm": 0.33781954646110535,
+      "learning_rate": 9.293211607725342e-05,
+      "loss": 1.2296,
+      "step": 2862
+    },
+    {
+      "epoch": 0.17394738440974541,
+      "grad_norm": 0.22288081049919128,
+      "learning_rate": 9.292720852412811e-05,
+      "loss": 1.1575,
+      "step": 2863
+    },
+    {
+      "epoch": 0.17400814144237195,
+      "grad_norm": 0.17195765674114227,
+      "learning_rate": 9.292229939749768e-05,
+      "loss": 1.1396,
+      "step": 2864
+    },
+    {
+      "epoch": 0.17406889847499848,
+      "grad_norm": 0.2843230068683624,
+      "learning_rate": 9.291738869754207e-05,
+      "loss": 1.1374,
+      "step": 2865
+    },
+    {
+      "epoch": 0.17412965550762502,
+      "grad_norm": 0.22741979360580444,
+      "learning_rate": 9.291247642444127e-05,
+      "loss": 1.1021,
+      "step": 2866
+    },
+    {
+      "epoch": 0.17419041254025153,
+      "grad_norm": 0.31253963708877563,
+      "learning_rate": 9.290756257837535e-05,
+      "loss": 1.2492,
+      "step": 2867
+    },
+    {
+      "epoch": 0.17425116957287806,
+      "grad_norm": 0.3234567642211914,
+      "learning_rate": 9.290264715952444e-05,
+      "loss": 1.1904,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1743119266055046,
+      "grad_norm": 0.19920937716960907,
+      "learning_rate": 9.28977301680687e-05,
+      "loss": 1.1446,
+      "step": 2869
+    },
+    {
+      "epoch": 0.1743726836381311,
+      "grad_norm": 0.2732466161251068,
+      "learning_rate": 9.289281160418837e-05,
+      "loss": 1.3202,
+      "step": 2870
+    },
+    {
+      "epoch": 0.17443344067075764,
+      "grad_norm": 0.15174159407615662,
+      "learning_rate": 9.288789146806374e-05,
+      "loss": 1.0677,
+      "step": 2871
+    },
+    {
+      "epoch": 0.17449419770338417,
+      "grad_norm": 0.26488688588142395,
+      "learning_rate": 9.288296975987515e-05,
+      "loss": 1.3555,
+      "step": 2872
+    },
+    {
+      "epoch": 0.17455495473601068,
+      "grad_norm": 0.3000930845737457,
+      "learning_rate": 9.287804647980303e-05,
+      "loss": 1.2565,
+      "step": 2873
+    },
+    {
+      "epoch": 0.17461571176863722,
+      "grad_norm": 0.1789645552635193,
+      "learning_rate": 9.287312162802782e-05,
+      "loss": 1.118,
+      "step": 2874
+    },
+    {
+      "epoch": 0.17467646880126375,
+      "grad_norm": 0.22553816437721252,
+      "learning_rate": 9.286819520473004e-05,
+      "loss": 1.1472,
+      "step": 2875
+    },
+    {
+      "epoch": 0.1747372258338903,
+      "grad_norm": 0.5480632185935974,
+      "learning_rate": 9.286326721009029e-05,
+      "loss": 1.1658,
+      "step": 2876
+    },
+    {
+      "epoch": 0.1747979828665168,
+      "grad_norm": 0.2610057294368744,
+      "learning_rate": 9.28583376442892e-05,
+      "loss": 1.181,
+      "step": 2877
+    },
+    {
+      "epoch": 0.17485873989914333,
+      "grad_norm": 0.5123583674430847,
+      "learning_rate": 9.285340650750745e-05,
+      "loss": 1.2391,
+      "step": 2878
+    },
+    {
+      "epoch": 0.17491949693176986,
+      "grad_norm": 0.18001104891300201,
+      "learning_rate": 9.284847379992578e-05,
+      "loss": 1.0335,
+      "step": 2879
+    },
+    {
+      "epoch": 0.17498025396439637,
+      "grad_norm": 0.261791467666626,
+      "learning_rate": 9.284353952172506e-05,
+      "loss": 1.2783,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1750410109970229,
+      "grad_norm": 0.3688179552555084,
+      "learning_rate": 9.28386036730861e-05,
+      "loss": 1.1358,
+      "step": 2881
+    },
+    {
+      "epoch": 0.17510176802964944,
+      "grad_norm": 0.34989133477211,
+      "learning_rate": 9.283366625418984e-05,
+      "loss": 1.1603,
+      "step": 2882
+    },
+    {
+      "epoch": 0.17516252506227595,
+      "grad_norm": 0.30162036418914795,
+      "learning_rate": 9.282872726521726e-05,
+      "loss": 1.1486,
+      "step": 2883
+    },
+    {
+      "epoch": 0.17522328209490248,
+      "grad_norm": 0.38653531670570374,
+      "learning_rate": 9.282378670634941e-05,
+      "loss": 1.0619,
+      "step": 2884
+    },
+    {
+      "epoch": 0.17528403912752902,
+      "grad_norm": 0.16464611887931824,
+      "learning_rate": 9.281884457776739e-05,
+      "loss": 1.1325,
+      "step": 2885
+    },
+    {
+      "epoch": 0.17534479616015552,
+      "grad_norm": 0.3459137976169586,
+      "learning_rate": 9.281390087965232e-05,
+      "loss": 1.1303,
+      "step": 2886
+    },
+    {
+      "epoch": 0.17540555319278206,
+      "grad_norm": 0.4004635214805603,
+      "learning_rate": 9.280895561218546e-05,
+      "loss": 1.1083,
+      "step": 2887
+    },
+    {
+      "epoch": 0.1754663102254086,
+      "grad_norm": 0.20689265429973602,
+      "learning_rate": 9.280400877554805e-05,
+      "loss": 1.0448,
+      "step": 2888
+    },
+    {
+      "epoch": 0.17552706725803513,
+      "grad_norm": 0.5363435745239258,
+      "learning_rate": 9.279906036992143e-05,
+      "loss": 1.0712,
+      "step": 2889
+    },
+    {
+      "epoch": 0.17558782429066164,
+      "grad_norm": 0.5169433355331421,
+      "learning_rate": 9.279411039548697e-05,
+      "loss": 1.1009,
+      "step": 2890
+    },
+    {
+      "epoch": 0.17564858132328817,
+      "grad_norm": 0.22366322576999664,
+      "learning_rate": 9.278915885242612e-05,
+      "loss": 1.1869,
+      "step": 2891
+    },
+    {
+      "epoch": 0.1757093383559147,
+      "grad_norm": 0.5242984294891357,
+      "learning_rate": 9.278420574092039e-05,
+      "loss": 1.1001,
+      "step": 2892
+    },
+    {
+      "epoch": 0.17577009538854121,
+      "grad_norm": 0.5860154032707214,
+      "learning_rate": 9.277925106115133e-05,
+      "loss": 1.056,
+      "step": 2893
+    },
+    {
+      "epoch": 0.17583085242116775,
+      "grad_norm": 0.24157953262329102,
+      "learning_rate": 9.277429481330054e-05,
+      "loss": 1.1536,
+      "step": 2894
+    },
+    {
+      "epoch": 0.17589160945379428,
+      "grad_norm": 2.524925708770752,
+      "learning_rate": 9.276933699754973e-05,
+      "loss": 1.1161,
+      "step": 2895
+    },
+    {
+      "epoch": 0.1759523664864208,
+      "grad_norm": 0.9460970163345337,
+      "learning_rate": 9.276437761408058e-05,
+      "loss": 1.1056,
+      "step": 2896
+    },
+    {
+      "epoch": 0.17601312351904733,
+      "grad_norm": 0.19609951972961426,
+      "learning_rate": 9.27594166630749e-05,
+      "loss": 1.0992,
+      "step": 2897
+    },
+    {
+      "epoch": 0.17607388055167386,
+      "grad_norm": 0.3549923300743103,
+      "learning_rate": 9.275445414471455e-05,
+      "loss": 1.2927,
+      "step": 2898
+    },
+    {
+      "epoch": 0.1761346375843004,
+      "grad_norm": 0.6184433102607727,
+      "learning_rate": 9.274949005918143e-05,
+      "loss": 1.1187,
+      "step": 2899
+    },
+    {
+      "epoch": 0.1761953946169269,
+      "grad_norm": 0.21333909034729004,
+      "learning_rate": 9.274452440665748e-05,
+      "loss": 1.0563,
+      "step": 2900
+    },
+    {
+      "epoch": 0.17625615164955344,
+      "grad_norm": 0.48458486795425415,
+      "learning_rate": 9.273955718732472e-05,
+      "loss": 1.1032,
+      "step": 2901
+    },
+    {
+      "epoch": 0.17631690868217997,
+      "grad_norm": 0.3009887635707855,
+      "learning_rate": 9.273458840136523e-05,
+      "loss": 1.1468,
+      "step": 2902
+    },
+    {
+      "epoch": 0.17637766571480648,
+      "grad_norm": 0.207598015666008,
+      "learning_rate": 9.272961804896114e-05,
+      "loss": 1.1485,
+      "step": 2903
+    },
+    {
+      "epoch": 0.17643842274743302,
+      "grad_norm": 0.5576958060264587,
+      "learning_rate": 9.272464613029466e-05,
+      "loss": 1.1241,
+      "step": 2904
+    },
+    {
+      "epoch": 0.17649917978005955,
+      "grad_norm": 0.36373335123062134,
+      "learning_rate": 9.271967264554799e-05,
+      "loss": 1.1535,
+      "step": 2905
+    },
+    {
+      "epoch": 0.17655993681268606,
+      "grad_norm": 0.3215940296649933,
+      "learning_rate": 9.271469759490348e-05,
+      "loss": 1.1741,
+      "step": 2906
+    },
+    {
+      "epoch": 0.1766206938453126,
+      "grad_norm": 0.22963079810142517,
+      "learning_rate": 9.270972097854346e-05,
+      "loss": 1.0935,
+      "step": 2907
+    },
+    {
+      "epoch": 0.17668145087793913,
+      "grad_norm": 0.37165361642837524,
+      "learning_rate": 9.270474279665037e-05,
+      "loss": 1.2967,
+      "step": 2908
+    },
+    {
+      "epoch": 0.17674220791056564,
+      "grad_norm": 0.31660640239715576,
+      "learning_rate": 9.269976304940669e-05,
+      "loss": 1.0709,
+      "step": 2909
+    },
+    {
+      "epoch": 0.17680296494319217,
+      "grad_norm": 0.18994460999965668,
+      "learning_rate": 9.269478173699492e-05,
+      "loss": 1.0734,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1768637219758187,
+      "grad_norm": 0.3182249963283539,
+      "learning_rate": 9.268979885959767e-05,
+      "loss": 1.1146,
+      "step": 2911
+    },
+    {
+      "epoch": 0.17692447900844524,
+      "grad_norm": 0.2843031585216522,
+      "learning_rate": 9.26848144173976e-05,
+      "loss": 1.2331,
+      "step": 2912
+    },
+    {
+      "epoch": 0.17698523604107175,
+      "grad_norm": 0.2277602106332779,
+      "learning_rate": 9.267982841057743e-05,
+      "loss": 1.2026,
+      "step": 2913
+    },
+    {
+      "epoch": 0.17704599307369828,
+      "grad_norm": 3.008589029312134,
+      "learning_rate": 9.267484083931987e-05,
+      "loss": 1.1579,
+      "step": 2914
+    },
+    {
+      "epoch": 0.17710675010632482,
+      "grad_norm": 0.21989019215106964,
+      "learning_rate": 9.266985170380777e-05,
+      "loss": 1.1289,
+      "step": 2915
+    },
+    {
+      "epoch": 0.17716750713895132,
+      "grad_norm": 3.9978926181793213,
+      "learning_rate": 9.266486100422402e-05,
+      "loss": 1.0819,
+      "step": 2916
+    },
+    {
+      "epoch": 0.17722826417157786,
+      "grad_norm": 0.18277336657047272,
+      "learning_rate": 9.265986874075155e-05,
+      "loss": 1.1572,
+      "step": 2917
+    },
+    {
+      "epoch": 0.1772890212042044,
+      "grad_norm": 0.19926071166992188,
+      "learning_rate": 9.265487491357334e-05,
+      "loss": 1.2134,
+      "step": 2918
+    },
+    {
+      "epoch": 0.1773497782368309,
+      "grad_norm": 0.19073501229286194,
+      "learning_rate": 9.264987952287243e-05,
+      "loss": 1.0866,
+      "step": 2919
+    },
+    {
+      "epoch": 0.17741053526945744,
+      "grad_norm": 0.1900995820760727,
+      "learning_rate": 9.264488256883194e-05,
+      "loss": 1.0677,
+      "step": 2920
+    },
+    {
+      "epoch": 0.17747129230208397,
+      "grad_norm": 0.23190531134605408,
+      "learning_rate": 9.263988405163506e-05,
+      "loss": 1.3003,
+      "step": 2921
+    },
+    {
+      "epoch": 0.17753204933471048,
+      "grad_norm": 0.3566952049732208,
+      "learning_rate": 9.263488397146498e-05,
+      "loss": 1.1074,
+      "step": 2922
+    },
+    {
+      "epoch": 0.177592806367337,
+      "grad_norm": 0.4597981572151184,
+      "learning_rate": 9.262988232850497e-05,
+      "loss": 1.1799,
+      "step": 2923
+    },
+    {
+      "epoch": 0.17765356339996355,
+      "grad_norm": 0.20072588324546814,
+      "learning_rate": 9.262487912293841e-05,
+      "loss": 1.1389,
+      "step": 2924
+    },
+    {
+      "epoch": 0.17771432043259008,
+      "grad_norm": 0.28654614090919495,
+      "learning_rate": 9.261987435494866e-05,
+      "loss": 1.219,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1777750774652166,
+      "grad_norm": 5.362672328948975,
+      "learning_rate": 9.261486802471916e-05,
+      "loss": 1.1247,
+      "step": 2926
+    },
+    {
+      "epoch": 0.17783583449784313,
+      "grad_norm": 0.2221587747335434,
+      "learning_rate": 9.260986013243345e-05,
+      "loss": 1.1279,
+      "step": 2927
+    },
+    {
+      "epoch": 0.17789659153046966,
+      "grad_norm": 0.8303794860839844,
+      "learning_rate": 9.260485067827508e-05,
+      "loss": 1.0504,
+      "step": 2928
+    },
+    {
+      "epoch": 0.17795734856309617,
+      "grad_norm": 0.22825317084789276,
+      "learning_rate": 9.259983966242767e-05,
+      "loss": 1.1048,
+      "step": 2929
+    },
+    {
+      "epoch": 0.1780181055957227,
+      "grad_norm": 0.21876034140586853,
+      "learning_rate": 9.25948270850749e-05,
+      "loss": 1.196,
+      "step": 2930
+    },
+    {
+      "epoch": 0.17807886262834924,
+      "grad_norm": 0.18129688501358032,
+      "learning_rate": 9.258981294640051e-05,
+      "loss": 1.1931,
+      "step": 2931
+    },
+    {
+      "epoch": 0.17813961966097575,
+      "grad_norm": 0.2925190329551697,
+      "learning_rate": 9.25847972465883e-05,
+      "loss": 1.0909,
+      "step": 2932
+    },
+    {
+      "epoch": 0.17820037669360228,
+      "grad_norm": 0.19749584794044495,
+      "learning_rate": 9.25797799858221e-05,
+      "loss": 1.0849,
+      "step": 2933
+    },
+    {
+      "epoch": 0.17826113372622882,
+      "grad_norm": 0.2684710919857025,
+      "learning_rate": 9.257476116428586e-05,
+      "loss": 1.1571,
+      "step": 2934
+    },
+    {
+      "epoch": 0.17832189075885535,
+      "grad_norm": 0.19014763832092285,
+      "learning_rate": 9.256974078216349e-05,
+      "loss": 1.1072,
+      "step": 2935
+    },
+    {
+      "epoch": 0.17838264779148186,
+      "grad_norm": 0.3571806848049164,
+      "learning_rate": 9.256471883963906e-05,
+      "loss": 1.0365,
+      "step": 2936
+    },
+    {
+      "epoch": 0.1784434048241084,
+      "grad_norm": 11.446825981140137,
+      "learning_rate": 9.255969533689664e-05,
+      "loss": 1.0546,
+      "step": 2937
+    },
+    {
+      "epoch": 0.17850416185673493,
+      "grad_norm": 0.2527330219745636,
+      "learning_rate": 9.255467027412034e-05,
+      "loss": 1.1649,
+      "step": 2938
+    },
+    {
+      "epoch": 0.17856491888936143,
+      "grad_norm": 0.19866426289081573,
+      "learning_rate": 9.25496436514944e-05,
+      "loss": 1.1168,
+      "step": 2939
+    },
+    {
+      "epoch": 0.17862567592198797,
+      "grad_norm": 0.1877591460943222,
+      "learning_rate": 9.254461546920303e-05,
+      "loss": 1.0906,
+      "step": 2940
+    },
+    {
+      "epoch": 0.1786864329546145,
+      "grad_norm": 0.17339429259300232,
+      "learning_rate": 9.253958572743058e-05,
+      "loss": 1.1445,
+      "step": 2941
+    },
+    {
+      "epoch": 0.178747189987241,
+      "grad_norm": 3.7702085971832275,
+      "learning_rate": 9.253455442636137e-05,
+      "loss": 1.1286,
+      "step": 2942
+    },
+    {
+      "epoch": 0.17880794701986755,
+      "grad_norm": 2.1460444927215576,
+      "learning_rate": 9.252952156617985e-05,
+      "loss": 1.1913,
+      "step": 2943
+    },
+    {
+      "epoch": 0.17886870405249408,
+      "grad_norm": 0.31069785356521606,
+      "learning_rate": 9.25244871470705e-05,
+      "loss": 1.2596,
+      "step": 2944
+    },
+    {
+      "epoch": 0.1789294610851206,
+      "grad_norm": 0.34162893891334534,
+      "learning_rate": 9.251945116921786e-05,
+      "loss": 1.0846,
+      "step": 2945
+    },
+    {
+      "epoch": 0.17899021811774712,
+      "grad_norm": 0.2573460638523102,
+      "learning_rate": 9.25144136328065e-05,
+      "loss": 1.1143,
+      "step": 2946
+    },
+    {
+      "epoch": 0.17905097515037366,
+      "grad_norm": 0.24816922843456268,
+      "learning_rate": 9.250937453802109e-05,
+      "loss": 1.2445,
+      "step": 2947
+    },
+    {
+      "epoch": 0.1791117321830002,
+      "grad_norm": 0.7603325247764587,
+      "learning_rate": 9.250433388504635e-05,
+      "loss": 1.1113,
+      "step": 2948
+    },
+    {
+      "epoch": 0.1791724892156267,
+      "grad_norm": 0.28203895688056946,
+      "learning_rate": 9.249929167406703e-05,
+      "loss": 1.2478,
+      "step": 2949
+    },
+    {
+      "epoch": 0.17923324624825324,
+      "grad_norm": 0.5692694783210754,
+      "learning_rate": 9.249424790526797e-05,
+      "loss": 1.2256,
+      "step": 2950
+    },
+    {
+      "epoch": 0.17929400328087977,
+      "grad_norm": 1.2878317832946777,
+      "learning_rate": 9.248920257883404e-05,
+      "loss": 1.2078,
+      "step": 2951
+    },
+    {
+      "epoch": 0.17935476031350628,
+      "grad_norm": 0.30191266536712646,
+      "learning_rate": 9.248415569495016e-05,
+      "loss": 1.0396,
+      "step": 2952
+    },
+    {
+      "epoch": 0.1794155173461328,
+      "grad_norm": 0.2611783444881439,
+      "learning_rate": 9.247910725380136e-05,
+      "loss": 1.1523,
+      "step": 2953
+    },
+    {
+      "epoch": 0.17947627437875935,
+      "grad_norm": 1.5869423151016235,
+      "learning_rate": 9.247405725557266e-05,
+      "loss": 1.2259,
+      "step": 2954
+    },
+    {
+      "epoch": 0.17953703141138586,
+      "grad_norm": 0.35859984159469604,
+      "learning_rate": 9.246900570044917e-05,
+      "loss": 1.1144,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1795977884440124,
+      "grad_norm": 0.26638662815093994,
+      "learning_rate": 9.246395258861609e-05,
+      "loss": 1.1497,
+      "step": 2956
+    },
+    {
+      "epoch": 0.17965854547663893,
+      "grad_norm": 0.15947876870632172,
+      "learning_rate": 9.245889792025862e-05,
+      "loss": 1.0957,
+      "step": 2957
+    },
+    {
+      "epoch": 0.17971930250926546,
+      "grad_norm": 0.2812144160270691,
+      "learning_rate": 9.245384169556202e-05,
+      "loss": 1.129,
+      "step": 2958
+    },
+    {
+      "epoch": 0.17978005954189197,
+      "grad_norm": 0.2027055025100708,
+      "learning_rate": 9.244878391471166e-05,
+      "loss": 1.2172,
+      "step": 2959
+    },
+    {
+      "epoch": 0.1798408165745185,
+      "grad_norm": 0.15256880223751068,
+      "learning_rate": 9.244372457789291e-05,
+      "loss": 1.1098,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17990157360714504,
+      "grad_norm": 0.276745080947876,
+      "learning_rate": 9.243866368529124e-05,
+      "loss": 1.1624,
+      "step": 2961
+    },
+    {
+      "epoch": 0.17996233063977154,
+      "grad_norm": 0.3117915987968445,
+      "learning_rate": 9.243360123709215e-05,
+      "loss": 1.1335,
+      "step": 2962
+    },
+    {
+      "epoch": 0.18002308767239808,
+      "grad_norm": 0.15853925049304962,
+      "learning_rate": 9.242853723348119e-05,
+      "loss": 1.1877,
+      "step": 2963
+    },
+    {
+      "epoch": 0.18008384470502461,
+      "grad_norm": 0.24287500977516174,
+      "learning_rate": 9.242347167464402e-05,
+      "loss": 1.1927,
+      "step": 2964
+    },
+    {
+      "epoch": 0.18014460173765112,
+      "grad_norm": 0.19106827676296234,
+      "learning_rate": 9.241840456076628e-05,
+      "loss": 1.097,
+      "step": 2965
+    },
+    {
+      "epoch": 0.18020535877027766,
+      "grad_norm": 0.18172281980514526,
+      "learning_rate": 9.241333589203372e-05,
+      "loss": 1.2125,
+      "step": 2966
+    },
+    {
+      "epoch": 0.1802661158029042,
+      "grad_norm": 0.1866511106491089,
+      "learning_rate": 9.240826566863214e-05,
+      "loss": 1.1017,
+      "step": 2967
+    },
+    {
+      "epoch": 0.1803268728355307,
+      "grad_norm": 0.14384622871875763,
+      "learning_rate": 9.240319389074739e-05,
+      "loss": 1.0846,
+      "step": 2968
+    },
+    {
+      "epoch": 0.18038762986815723,
+      "grad_norm": 5.6967010498046875,
+      "learning_rate": 9.239812055856536e-05,
+      "loss": 1.1679,
+      "step": 2969
+    },
+    {
+      "epoch": 0.18044838690078377,
+      "grad_norm": 0.26043370366096497,
+      "learning_rate": 9.239304567227204e-05,
+      "loss": 1.0847,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1805091439334103,
+      "grad_norm": 0.2230168581008911,
+      "learning_rate": 9.238796923205343e-05,
+      "loss": 1.1662,
+      "step": 2971
+    },
+    {
+      "epoch": 0.1805699009660368,
+      "grad_norm": 0.270250529050827,
+      "learning_rate": 9.238289123809561e-05,
+      "loss": 1.1246,
+      "step": 2972
+    },
+    {
+      "epoch": 0.18063065799866335,
+      "grad_norm": 0.7367274761199951,
+      "learning_rate": 9.237781169058473e-05,
+      "loss": 1.2119,
+      "step": 2973
+    },
+    {
+      "epoch": 0.18069141503128988,
+      "grad_norm": 0.31884264945983887,
+      "learning_rate": 9.237273058970694e-05,
+      "loss": 1.1147,
+      "step": 2974
+    },
+    {
+      "epoch": 0.1807521720639164,
+      "grad_norm": 0.1792020946741104,
+      "learning_rate": 9.236764793564857e-05,
+      "loss": 1.1491,
+      "step": 2975
+    },
+    {
+      "epoch": 0.18081292909654292,
+      "grad_norm": 0.8205214738845825,
+      "learning_rate": 9.236256372859583e-05,
+      "loss": 1.2785,
+      "step": 2976
+    },
+    {
+      "epoch": 0.18087368612916946,
+      "grad_norm": 0.3323436975479126,
+      "learning_rate": 9.235747796873514e-05,
+      "loss": 1.1044,
+      "step": 2977
+    },
+    {
+      "epoch": 0.18093444316179597,
+      "grad_norm": 0.22030521929264069,
+      "learning_rate": 9.235239065625292e-05,
+      "loss": 1.0525,
+      "step": 2978
+    },
+    {
+      "epoch": 0.1809952001944225,
+      "grad_norm": 0.2923031151294708,
+      "learning_rate": 9.234730179133564e-05,
+      "loss": 1.0687,
+      "step": 2979
+    },
+    {
+      "epoch": 0.18105595722704904,
+      "grad_norm": 0.16640354692935944,
+      "learning_rate": 9.234221137416981e-05,
+      "loss": 1.0725,
+      "step": 2980
+    },
+    {
+      "epoch": 0.18111671425967557,
+      "grad_norm": 0.408065527677536,
+      "learning_rate": 9.233711940494204e-05,
+      "loss": 1.1525,
+      "step": 2981
+    },
+    {
+      "epoch": 0.18117747129230208,
+      "grad_norm": 0.37556928396224976,
+      "learning_rate": 9.233202588383897e-05,
+      "loss": 1.2737,
+      "step": 2982
+    },
+    {
+      "epoch": 0.1812382283249286,
+      "grad_norm": 0.20571397244930267,
+      "learning_rate": 9.232693081104734e-05,
+      "loss": 1.1488,
+      "step": 2983
+    },
+    {
+      "epoch": 0.18129898535755515,
+      "grad_norm": 0.36000514030456543,
+      "learning_rate": 9.232183418675384e-05,
+      "loss": 1.1286,
+      "step": 2984
+    },
+    {
+      "epoch": 0.18135974239018166,
+      "grad_norm": 0.35860154032707214,
+      "learning_rate": 9.231673601114534e-05,
+      "loss": 1.0556,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1814204994228082,
+      "grad_norm": 0.31243014335632324,
+      "learning_rate": 9.231163628440871e-05,
+      "loss": 1.221,
+      "step": 2986
+    },
+    {
+      "epoch": 0.18148125645543473,
+      "grad_norm": 0.22265656292438507,
+      "learning_rate": 9.230653500673086e-05,
+      "loss": 1.0512,
+      "step": 2987
+    },
+    {
+      "epoch": 0.18154201348806123,
+      "grad_norm": 0.44037526845932007,
+      "learning_rate": 9.230143217829881e-05,
+      "loss": 1.2424,
+      "step": 2988
+    },
+    {
+      "epoch": 0.18160277052068777,
+      "grad_norm": 0.18836897611618042,
+      "learning_rate": 9.229632779929957e-05,
+      "loss": 1.0907,
+      "step": 2989
+    },
+    {
+      "epoch": 0.1816635275533143,
+      "grad_norm": 0.680586040019989,
+      "learning_rate": 9.229122186992026e-05,
+      "loss": 1.0429,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1817242845859408,
+      "grad_norm": 0.22910404205322266,
+      "learning_rate": 9.228611439034804e-05,
+      "loss": 1.3046,
+      "step": 2991
+    },
+    {
+      "epoch": 0.18178504161856734,
+      "grad_norm": 0.18243393301963806,
+      "learning_rate": 9.228100536077013e-05,
+      "loss": 1.1163,
+      "step": 2992
+    },
+    {
+      "epoch": 0.18184579865119388,
+      "grad_norm": 0.20854873955249786,
+      "learning_rate": 9.22758947813738e-05,
+      "loss": 1.121,
+      "step": 2993
+    },
+    {
+      "epoch": 0.18190655568382041,
+      "grad_norm": 17.038894653320312,
+      "learning_rate": 9.227078265234637e-05,
+      "loss": 1.1108,
+      "step": 2994
+    },
+    {
+      "epoch": 0.18196731271644692,
+      "grad_norm": 0.23351851105690002,
+      "learning_rate": 9.226566897387522e-05,
+      "loss": 1.1237,
+      "step": 2995
+    },
+    {
+      "epoch": 0.18202806974907346,
+      "grad_norm": 0.18451769649982452,
+      "learning_rate": 9.226055374614781e-05,
+      "loss": 1.1559,
+      "step": 2996
+    },
+    {
+      "epoch": 0.1820888267817,
+      "grad_norm": 0.248361274600029,
+      "learning_rate": 9.225543696935164e-05,
+      "loss": 1.0858,
+      "step": 2997
+    },
+    {
+      "epoch": 0.1821495838143265,
+      "grad_norm": 0.18257442116737366,
+      "learning_rate": 9.225031864367425e-05,
+      "loss": 1.1095,
+      "step": 2998
+    },
+    {
+      "epoch": 0.18221034084695303,
+      "grad_norm": 0.29653942584991455,
+      "learning_rate": 9.224519876930327e-05,
+      "loss": 1.2073,
+      "step": 2999
+    },
+    {
+      "epoch": 0.18227109787957957,
+      "grad_norm": 0.1599666327238083,
+      "learning_rate": 9.224007734642637e-05,
+      "loss": 1.0709,
+      "step": 3000
+    },
+    {
+      "epoch": 0.18233185491220608,
+      "grad_norm": 0.3649294078350067,
+      "learning_rate": 9.223495437523127e-05,
+      "loss": 1.1197,
+      "step": 3001
+    },
+    {
+      "epoch": 0.1823926119448326,
+      "grad_norm": 0.2119244933128357,
+      "learning_rate": 9.222982985590574e-05,
+      "loss": 1.2567,
+      "step": 3002
+    },
+    {
+      "epoch": 0.18245336897745915,
+      "grad_norm": 0.21541699767112732,
+      "learning_rate": 9.222470378863763e-05,
+      "loss": 1.1022,
+      "step": 3003
+    },
+    {
+      "epoch": 0.18251412601008568,
+      "grad_norm": 0.30862703919410706,
+      "learning_rate": 9.221957617361485e-05,
+      "loss": 1.1673,
+      "step": 3004
+    },
+    {
+      "epoch": 0.1825748830427122,
+      "grad_norm": 0.3197075426578522,
+      "learning_rate": 9.221444701102534e-05,
+      "loss": 1.1253,
+      "step": 3005
+    },
+    {
+      "epoch": 0.18263564007533872,
+      "grad_norm": 0.18545778095722198,
+      "learning_rate": 9.220931630105712e-05,
+      "loss": 1.0923,
+      "step": 3006
+    },
+    {
+      "epoch": 0.18269639710796526,
+      "grad_norm": 0.14189809560775757,
+      "learning_rate": 9.220418404389826e-05,
+      "loss": 1.0865,
+      "step": 3007
+    },
+    {
+      "epoch": 0.18275715414059177,
+      "grad_norm": 0.15887649357318878,
+      "learning_rate": 9.219905023973687e-05,
+      "loss": 1.0706,
+      "step": 3008
+    },
+    {
+      "epoch": 0.1828179111732183,
+      "grad_norm": 0.18498313426971436,
+      "learning_rate": 9.219391488876112e-05,
+      "loss": 1.1515,
+      "step": 3009
+    },
+    {
+      "epoch": 0.18287866820584484,
+      "grad_norm": 27.192461013793945,
+      "learning_rate": 9.218877799115928e-05,
+      "loss": 1.1109,
+      "step": 3010
+    },
+    {
+      "epoch": 0.18293942523847134,
+      "grad_norm": 0.2581998109817505,
+      "learning_rate": 9.218363954711964e-05,
+      "loss": 1.1937,
+      "step": 3011
+    },
+    {
+      "epoch": 0.18300018227109788,
+      "grad_norm": 0.26588180661201477,
+      "learning_rate": 9.217849955683051e-05,
+      "loss": 1.261,
+      "step": 3012
+    },
+    {
+      "epoch": 0.1830609393037244,
+      "grad_norm": 0.23390227556228638,
+      "learning_rate": 9.217335802048035e-05,
+      "loss": 1.1572,
+      "step": 3013
+    },
+    {
+      "epoch": 0.18312169633635092,
+      "grad_norm": 0.20528551936149597,
+      "learning_rate": 9.21682149382576e-05,
+      "loss": 1.1749,
+      "step": 3014
+    },
+    {
+      "epoch": 0.18318245336897745,
+      "grad_norm": 0.308460533618927,
+      "learning_rate": 9.216307031035077e-05,
+      "loss": 1.0873,
+      "step": 3015
+    },
+    {
+      "epoch": 0.183243210401604,
+      "grad_norm": 0.26810169219970703,
+      "learning_rate": 9.215792413694844e-05,
+      "loss": 1.2088,
+      "step": 3016
+    },
+    {
+      "epoch": 0.18330396743423052,
+      "grad_norm": 0.8332200050354004,
+      "learning_rate": 9.215277641823925e-05,
+      "loss": 1.1477,
+      "step": 3017
+    },
+    {
+      "epoch": 0.18336472446685703,
+      "grad_norm": 0.16305555403232574,
+      "learning_rate": 9.214762715441192e-05,
+      "loss": 1.0813,
+      "step": 3018
+    },
+    {
+      "epoch": 0.18342548149948357,
+      "grad_norm": 0.20657306909561157,
+      "learning_rate": 9.214247634565514e-05,
+      "loss": 1.1083,
+      "step": 3019
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 0.23661090433597565,
+      "learning_rate": 9.213732399215776e-05,
+      "loss": 1.2504,
+      "step": 3020
+    },
+    {
+      "epoch": 0.1835469955647366,
+      "grad_norm": 0.22513361275196075,
+      "learning_rate": 9.213217009410862e-05,
+      "loss": 1.0823,
+      "step": 3021
+    },
+    {
+      "epoch": 0.18360775259736314,
+      "grad_norm": 0.20578782260417938,
+      "learning_rate": 9.212701465169664e-05,
+      "loss": 1.1396,
+      "step": 3022
+    },
+    {
+      "epoch": 0.18366850962998968,
+      "grad_norm": 2.9302711486816406,
+      "learning_rate": 9.21218576651108e-05,
+      "loss": 1.1487,
+      "step": 3023
+    },
+    {
+      "epoch": 0.1837292666626162,
+      "grad_norm": 0.19550593197345734,
+      "learning_rate": 9.211669913454013e-05,
+      "loss": 1.0833,
+      "step": 3024
+    },
+    {
+      "epoch": 0.18379002369524272,
+      "grad_norm": 0.1986549347639084,
+      "learning_rate": 9.21115390601737e-05,
+      "loss": 1.1655,
+      "step": 3025
+    },
+    {
+      "epoch": 0.18385078072786926,
+      "grad_norm": 0.24554631114006042,
+      "learning_rate": 9.210637744220067e-05,
+      "loss": 1.0549,
+      "step": 3026
+    },
+    {
+      "epoch": 0.18391153776049576,
+      "grad_norm": 0.29472488164901733,
+      "learning_rate": 9.210121428081022e-05,
+      "loss": 1.2964,
+      "step": 3027
+    },
+    {
+      "epoch": 0.1839722947931223,
+      "grad_norm": 0.26621246337890625,
+      "learning_rate": 9.209604957619164e-05,
+      "loss": 1.0757,
+      "step": 3028
+    },
+    {
+      "epoch": 0.18403305182574883,
+      "grad_norm": 1.509060263633728,
+      "learning_rate": 9.209088332853424e-05,
+      "loss": 1.1352,
+      "step": 3029
+    },
+    {
+      "epoch": 0.18409380885837537,
+      "grad_norm": 0.15099680423736572,
+      "learning_rate": 9.208571553802734e-05,
+      "loss": 1.1149,
+      "step": 3030
+    },
+    {
+      "epoch": 0.18415456589100188,
+      "grad_norm": 0.3969057500362396,
+      "learning_rate": 9.208054620486043e-05,
+      "loss": 1.1671,
+      "step": 3031
+    },
+    {
+      "epoch": 0.1842153229236284,
+      "grad_norm": 28.691776275634766,
+      "learning_rate": 9.207537532922296e-05,
+      "loss": 1.2,
+      "step": 3032
+    },
+    {
+      "epoch": 0.18427607995625495,
+      "grad_norm": 0.2952224910259247,
+      "learning_rate": 9.207020291130448e-05,
+      "loss": 1.2385,
+      "step": 3033
+    },
+    {
+      "epoch": 0.18433683698888145,
+      "grad_norm": 0.16883589327335358,
+      "learning_rate": 9.206502895129456e-05,
+      "loss": 1.1068,
+      "step": 3034
+    },
+    {
+      "epoch": 0.184397594021508,
+      "grad_norm": 0.19729213416576385,
+      "learning_rate": 9.205985344938288e-05,
+      "loss": 1.1054,
+      "step": 3035
+    },
+    {
+      "epoch": 0.18445835105413452,
+      "grad_norm": 0.435029536485672,
+      "learning_rate": 9.205467640575916e-05,
+      "loss": 1.1658,
+      "step": 3036
+    },
+    {
+      "epoch": 0.18451910808676103,
+      "grad_norm": 0.15445058047771454,
+      "learning_rate": 9.204949782061311e-05,
+      "loss": 1.038,
+      "step": 3037
+    },
+    {
+      "epoch": 0.18457986511938757,
+      "grad_norm": 0.19000063836574554,
+      "learning_rate": 9.204431769413462e-05,
+      "loss": 1.2583,
+      "step": 3038
+    },
+    {
+      "epoch": 0.1846406221520141,
+      "grad_norm": 0.16063889861106873,
+      "learning_rate": 9.203913602651354e-05,
+      "loss": 1.1036,
+      "step": 3039
+    },
+    {
+      "epoch": 0.18470137918464063,
+      "grad_norm": 0.47722092270851135,
+      "learning_rate": 9.203395281793979e-05,
+      "loss": 1.0754,
+      "step": 3040
+    },
+    {
+      "epoch": 0.18476213621726714,
+      "grad_norm": 0.22657538950443268,
+      "learning_rate": 9.202876806860338e-05,
+      "loss": 1.0867,
+      "step": 3041
+    },
+    {
+      "epoch": 0.18482289324989368,
+      "grad_norm": 0.2091595083475113,
+      "learning_rate": 9.202358177869435e-05,
+      "loss": 1.2261,
+      "step": 3042
+    },
+    {
+      "epoch": 0.1848836502825202,
+      "grad_norm": 0.1922287940979004,
+      "learning_rate": 9.20183939484028e-05,
+      "loss": 1.179,
+      "step": 3043
+    },
+    {
+      "epoch": 0.18494440731514672,
+      "grad_norm": 0.2977611720561981,
+      "learning_rate": 9.201320457791891e-05,
+      "loss": 1.0848,
+      "step": 3044
+    },
+    {
+      "epoch": 0.18500516434777325,
+      "grad_norm": 0.14819122850894928,
+      "learning_rate": 9.200801366743289e-05,
+      "loss": 1.1249,
+      "step": 3045
+    },
+    {
+      "epoch": 0.1850659213803998,
+      "grad_norm": 0.20349188148975372,
+      "learning_rate": 9.200282121713499e-05,
+      "loss": 1.1914,
+      "step": 3046
+    },
+    {
+      "epoch": 0.1851266784130263,
+      "grad_norm": 0.25512316823005676,
+      "learning_rate": 9.199762722721557e-05,
+      "loss": 1.0801,
+      "step": 3047
+    },
+    {
+      "epoch": 0.18518743544565283,
+      "grad_norm": 0.3465666174888611,
+      "learning_rate": 9.1992431697865e-05,
+      "loss": 1.3186,
+      "step": 3048
+    },
+    {
+      "epoch": 0.18524819247827937,
+      "grad_norm": 0.202579066157341,
+      "learning_rate": 9.198723462927375e-05,
+      "loss": 1.0727,
+      "step": 3049
+    },
+    {
+      "epoch": 0.18530894951090587,
+      "grad_norm": 0.27540209889411926,
+      "learning_rate": 9.198203602163227e-05,
+      "loss": 1.0662,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1853697065435324,
+      "grad_norm": 0.21132218837738037,
+      "learning_rate": 9.197683587513116e-05,
+      "loss": 1.1831,
+      "step": 3051
+    },
+    {
+      "epoch": 0.18543046357615894,
+      "grad_norm": 0.1563459038734436,
+      "learning_rate": 9.197163418996101e-05,
+      "loss": 1.0974,
+      "step": 3052
+    },
+    {
+      "epoch": 0.18549122060878548,
+      "grad_norm": 0.4012366235256195,
+      "learning_rate": 9.19664309663125e-05,
+      "loss": 1.0923,
+      "step": 3053
+    },
+    {
+      "epoch": 0.18555197764141199,
+      "grad_norm": 0.14499591290950775,
+      "learning_rate": 9.196122620437635e-05,
+      "loss": 1.0746,
+      "step": 3054
+    },
+    {
+      "epoch": 0.18561273467403852,
+      "grad_norm": 0.21168065071105957,
+      "learning_rate": 9.195601990434335e-05,
+      "loss": 1.2122,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18567349170666506,
+      "grad_norm": 0.2861902415752411,
+      "learning_rate": 9.195081206640433e-05,
+      "loss": 1.1907,
+      "step": 3056
+    },
+    {
+      "epoch": 0.18573424873929156,
+      "grad_norm": 0.35591644048690796,
+      "learning_rate": 9.194560269075016e-05,
+      "loss": 1.0346,
+      "step": 3057
+    },
+    {
+      "epoch": 0.1857950057719181,
+      "grad_norm": 0.29218196868896484,
+      "learning_rate": 9.194039177757184e-05,
+      "loss": 1.0847,
+      "step": 3058
+    },
+    {
+      "epoch": 0.18585576280454463,
+      "grad_norm": 0.4533981680870056,
+      "learning_rate": 9.193517932706036e-05,
+      "loss": 1.101,
+      "step": 3059
+    },
+    {
+      "epoch": 0.18591651983717114,
+      "grad_norm": 0.20430108904838562,
+      "learning_rate": 9.192996533940675e-05,
+      "loss": 1.1202,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18597727686979768,
+      "grad_norm": 0.3417561650276184,
+      "learning_rate": 9.192474981480216e-05,
+      "loss": 1.1267,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1860380339024242,
+      "grad_norm": 0.22390389442443848,
+      "learning_rate": 9.191953275343779e-05,
+      "loss": 1.3037,
+      "step": 3062
+    },
+    {
+      "epoch": 0.18609879093505075,
+      "grad_norm": 0.18954911828041077,
+      "learning_rate": 9.191431415550482e-05,
+      "loss": 1.0919,
+      "step": 3063
+    },
+    {
+      "epoch": 0.18615954796767725,
+      "grad_norm": 0.15950433909893036,
+      "learning_rate": 9.190909402119456e-05,
+      "loss": 1.1088,
+      "step": 3064
+    },
+    {
+      "epoch": 0.1862203050003038,
+      "grad_norm": 0.28802913427352905,
+      "learning_rate": 9.190387235069837e-05,
+      "loss": 1.1666,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18628106203293032,
+      "grad_norm": 0.31233733892440796,
+      "learning_rate": 9.189864914420763e-05,
+      "loss": 1.0677,
+      "step": 3066
+    },
+    {
+      "epoch": 0.18634181906555683,
+      "grad_norm": 0.1936078518629074,
+      "learning_rate": 9.189342440191382e-05,
+      "loss": 1.1203,
+      "step": 3067
+    },
+    {
+      "epoch": 0.18640257609818336,
+      "grad_norm": 0.20946606993675232,
+      "learning_rate": 9.188819812400842e-05,
+      "loss": 1.1818,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1864633331308099,
+      "grad_norm": 0.4461238980293274,
+      "learning_rate": 9.188297031068304e-05,
+      "loss": 1.0657,
+      "step": 3069
+    },
+    {
+      "epoch": 0.1865240901634364,
+      "grad_norm": 0.17963141202926636,
+      "learning_rate": 9.187774096212926e-05,
+      "loss": 1.1849,
+      "step": 3070
+    },
+    {
+      "epoch": 0.18658484719606294,
+      "grad_norm": 0.3228558897972107,
+      "learning_rate": 9.187251007853883e-05,
+      "loss": 1.2921,
+      "step": 3071
+    },
+    {
+      "epoch": 0.18664560422868948,
+      "grad_norm": 0.17975732684135437,
+      "learning_rate": 9.186727766010343e-05,
+      "loss": 1.1752,
+      "step": 3072
+    },
+    {
+      "epoch": 0.18670636126131598,
+      "grad_norm": 0.228794127702713,
+      "learning_rate": 9.186204370701486e-05,
+      "loss": 1.1413,
+      "step": 3073
+    },
+    {
+      "epoch": 0.18676711829394252,
+      "grad_norm": 1.022456407546997,
+      "learning_rate": 9.185680821946501e-05,
+      "loss": 1.3289,
+      "step": 3074
+    },
+    {
+      "epoch": 0.18682787532656905,
+      "grad_norm": 0.15229488909244537,
+      "learning_rate": 9.185157119764575e-05,
+      "loss": 1.1194,
+      "step": 3075
+    },
+    {
+      "epoch": 0.1868886323591956,
+      "grad_norm": 0.33429181575775146,
+      "learning_rate": 9.184633264174907e-05,
+      "loss": 1.2345,
+      "step": 3076
+    },
+    {
+      "epoch": 0.1869493893918221,
+      "grad_norm": 0.31576576828956604,
+      "learning_rate": 9.184109255196696e-05,
+      "loss": 1.1097,
+      "step": 3077
+    },
+    {
+      "epoch": 0.18701014642444863,
+      "grad_norm": 0.5714771747589111,
+      "learning_rate": 9.183585092849153e-05,
+      "loss": 1.1388,
+      "step": 3078
+    },
+    {
+      "epoch": 0.18707090345707517,
+      "grad_norm": 0.32460519671440125,
+      "learning_rate": 9.183060777151489e-05,
+      "loss": 1.0879,
+      "step": 3079
+    },
+    {
+      "epoch": 0.18713166048970167,
+      "grad_norm": 0.2623949944972992,
+      "learning_rate": 9.182536308122925e-05,
+      "loss": 1.0947,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1871924175223282,
+      "grad_norm": 0.1441493034362793,
+      "learning_rate": 9.182011685782683e-05,
+      "loss": 1.0689,
+      "step": 3081
+    },
+    {
+      "epoch": 0.18725317455495474,
+      "grad_norm": 0.32306981086730957,
+      "learning_rate": 9.181486910149994e-05,
+      "loss": 1.0787,
+      "step": 3082
+    },
+    {
+      "epoch": 0.18731393158758125,
+      "grad_norm": 0.19291339814662933,
+      "learning_rate": 9.180961981244094e-05,
+      "loss": 1.1086,
+      "step": 3083
+    },
+    {
+      "epoch": 0.18737468862020779,
+      "grad_norm": 0.19064386188983917,
+      "learning_rate": 9.180436899084225e-05,
+      "loss": 1.171,
+      "step": 3084
+    },
+    {
+      "epoch": 0.18743544565283432,
+      "grad_norm": 0.2563016712665558,
+      "learning_rate": 9.179911663689633e-05,
+      "loss": 1.1988,
+      "step": 3085
+    },
+    {
+      "epoch": 0.18749620268546086,
+      "grad_norm": 0.2070801556110382,
+      "learning_rate": 9.179386275079571e-05,
+      "loss": 1.2355,
+      "step": 3086
+    },
+    {
+      "epoch": 0.18755695971808736,
+      "grad_norm": 0.18820656836032867,
+      "learning_rate": 9.178860733273298e-05,
+      "loss": 1.1071,
+      "step": 3087
+    },
+    {
+      "epoch": 0.1876177167507139,
+      "grad_norm": 0.20887188613414764,
+      "learning_rate": 9.178335038290075e-05,
+      "loss": 1.107,
+      "step": 3088
+    },
+    {
+      "epoch": 0.18767847378334043,
+      "grad_norm": 0.18751654028892517,
+      "learning_rate": 9.177809190149176e-05,
+      "loss": 1.1717,
+      "step": 3089
+    },
+    {
+      "epoch": 0.18773923081596694,
+      "grad_norm": 0.19147314131259918,
+      "learning_rate": 9.177283188869873e-05,
+      "loss": 1.0788,
+      "step": 3090
+    },
+    {
+      "epoch": 0.18779998784859347,
+      "grad_norm": 0.3001182973384857,
+      "learning_rate": 9.176757034471447e-05,
+      "loss": 1.1573,
+      "step": 3091
+    },
+    {
+      "epoch": 0.18786074488122,
+      "grad_norm": 0.1793118417263031,
+      "learning_rate": 9.176230726973185e-05,
+      "loss": 1.1376,
+      "step": 3092
+    },
+    {
+      "epoch": 0.18792150191384652,
+      "grad_norm": 0.2062307745218277,
+      "learning_rate": 9.175704266394378e-05,
+      "loss": 1.1352,
+      "step": 3093
+    },
+    {
+      "epoch": 0.18798225894647305,
+      "grad_norm": 0.19472725689411163,
+      "learning_rate": 9.175177652754323e-05,
+      "loss": 1.0824,
+      "step": 3094
+    },
+    {
+      "epoch": 0.1880430159790996,
+      "grad_norm": 0.1955190896987915,
+      "learning_rate": 9.174650886072325e-05,
+      "loss": 1.165,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1881037730117261,
+      "grad_norm": 0.1639154702425003,
+      "learning_rate": 9.174123966367691e-05,
+      "loss": 1.2194,
+      "step": 3096
+    },
+    {
+      "epoch": 0.18816453004435263,
+      "grad_norm": 0.3118985593318939,
+      "learning_rate": 9.173596893659738e-05,
+      "loss": 1.2413,
+      "step": 3097
+    },
+    {
+      "epoch": 0.18822528707697916,
+      "grad_norm": 0.16718478500843048,
+      "learning_rate": 9.173069667967785e-05,
+      "loss": 1.1035,
+      "step": 3098
+    },
+    {
+      "epoch": 0.1882860441096057,
+      "grad_norm": 0.2691071629524231,
+      "learning_rate": 9.172542289311153e-05,
+      "loss": 1.0747,
+      "step": 3099
+    },
+    {
+      "epoch": 0.1883468011422322,
+      "grad_norm": 0.2464882880449295,
+      "learning_rate": 9.172014757709179e-05,
+      "loss": 1.1528,
+      "step": 3100
+    },
+    {
+      "epoch": 0.18840755817485874,
+      "grad_norm": 0.472747802734375,
+      "learning_rate": 9.171487073181198e-05,
+      "loss": 1.1142,
+      "step": 3101
+    },
+    {
+      "epoch": 0.18846831520748528,
+      "grad_norm": 0.2910997271537781,
+      "learning_rate": 9.17095923574655e-05,
+      "loss": 1.0531,
+      "step": 3102
+    },
+    {
+      "epoch": 0.18852907224011178,
+      "grad_norm": 0.21663829684257507,
+      "learning_rate": 9.170431245424587e-05,
+      "loss": 1.0939,
+      "step": 3103
+    },
+    {
+      "epoch": 0.18858982927273832,
+      "grad_norm": 0.17455355823040009,
+      "learning_rate": 9.169903102234659e-05,
+      "loss": 1.1838,
+      "step": 3104
+    },
+    {
+      "epoch": 0.18865058630536485,
+      "grad_norm": 0.2848408818244934,
+      "learning_rate": 9.169374806196127e-05,
+      "loss": 1.1643,
+      "step": 3105
+    },
+    {
+      "epoch": 0.18871134333799136,
+      "grad_norm": 0.21605317294597626,
+      "learning_rate": 9.168846357328357e-05,
+      "loss": 1.1257,
+      "step": 3106
+    },
+    {
+      "epoch": 0.1887721003706179,
+      "grad_norm": 0.3147536516189575,
+      "learning_rate": 9.168317755650717e-05,
+      "loss": 1.3585,
+      "step": 3107
+    },
+    {
+      "epoch": 0.18883285740324443,
+      "grad_norm": 0.37438318133354187,
+      "learning_rate": 9.167789001182584e-05,
+      "loss": 1.1881,
+      "step": 3108
+    },
+    {
+      "epoch": 0.18889361443587097,
+      "grad_norm": 0.26369765400886536,
+      "learning_rate": 9.16726009394334e-05,
+      "loss": 1.0964,
+      "step": 3109
+    },
+    {
+      "epoch": 0.18895437146849747,
+      "grad_norm": 0.15647491812705994,
+      "learning_rate": 9.166731033952372e-05,
+      "loss": 1.1093,
+      "step": 3110
+    },
+    {
+      "epoch": 0.189015128501124,
+      "grad_norm": 0.15289811789989471,
+      "learning_rate": 9.166201821229072e-05,
+      "loss": 1.0353,
+      "step": 3111
+    },
+    {
+      "epoch": 0.18907588553375054,
+      "grad_norm": 0.2987431287765503,
+      "learning_rate": 9.165672455792838e-05,
+      "loss": 1.0674,
+      "step": 3112
+    },
+    {
+      "epoch": 0.18913664256637705,
+      "grad_norm": 0.21291735768318176,
+      "learning_rate": 9.165142937663076e-05,
+      "loss": 1.118,
+      "step": 3113
+    },
+    {
+      "epoch": 0.18919739959900359,
+      "grad_norm": 0.3091643452644348,
+      "learning_rate": 9.164613266859193e-05,
+      "loss": 1.2288,
+      "step": 3114
+    },
+    {
+      "epoch": 0.18925815663163012,
+      "grad_norm": 0.22075144946575165,
+      "learning_rate": 9.164083443400609e-05,
+      "loss": 1.053,
+      "step": 3115
+    },
+    {
+      "epoch": 0.18931891366425663,
+      "grad_norm": 0.3806934356689453,
+      "learning_rate": 9.163553467306738e-05,
+      "loss": 1.191,
+      "step": 3116
+    },
+    {
+      "epoch": 0.18937967069688316,
+      "grad_norm": 0.37007635831832886,
+      "learning_rate": 9.163023338597011e-05,
+      "loss": 1.1147,
+      "step": 3117
+    },
+    {
+      "epoch": 0.1894404277295097,
+      "grad_norm": 0.1943754106760025,
+      "learning_rate": 9.16249305729086e-05,
+      "loss": 1.1602,
+      "step": 3118
+    },
+    {
+      "epoch": 0.1895011847621362,
+      "grad_norm": 0.16483408212661743,
+      "learning_rate": 9.16196262340772e-05,
+      "loss": 1.0826,
+      "step": 3119
+    },
+    {
+      "epoch": 0.18956194179476274,
+      "grad_norm": 0.2626393437385559,
+      "learning_rate": 9.161432036967036e-05,
+      "loss": 1.0773,
+      "step": 3120
+    },
+    {
+      "epoch": 0.18962269882738927,
+      "grad_norm": 0.29525965452194214,
+      "learning_rate": 9.160901297988256e-05,
+      "loss": 1.1494,
+      "step": 3121
+    },
+    {
+      "epoch": 0.1896834558600158,
+      "grad_norm": 0.22029602527618408,
+      "learning_rate": 9.160370406490834e-05,
+      "loss": 1.111,
+      "step": 3122
+    },
+    {
+      "epoch": 0.18974421289264232,
+      "grad_norm": 0.21798256039619446,
+      "learning_rate": 9.159839362494232e-05,
+      "loss": 1.196,
+      "step": 3123
+    },
+    {
+      "epoch": 0.18980496992526885,
+      "grad_norm": 2.241982936859131,
+      "learning_rate": 9.159308166017915e-05,
+      "loss": 1.1281,
+      "step": 3124
+    },
+    {
+      "epoch": 0.1898657269578954,
+      "grad_norm": 0.2661694288253784,
+      "learning_rate": 9.158776817081351e-05,
+      "loss": 1.1396,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1899264839905219,
+      "grad_norm": 0.17097768187522888,
+      "learning_rate": 9.158245315704019e-05,
+      "loss": 1.0994,
+      "step": 3126
+    },
+    {
+      "epoch": 0.18998724102314843,
+      "grad_norm": 0.16301696002483368,
+      "learning_rate": 9.157713661905402e-05,
+      "loss": 1.0794,
+      "step": 3127
+    },
+    {
+      "epoch": 0.19004799805577496,
+      "grad_norm": 0.24085086584091187,
+      "learning_rate": 9.157181855704988e-05,
+      "loss": 1.1211,
+      "step": 3128
+    },
+    {
+      "epoch": 0.19010875508840147,
+      "grad_norm": 0.18063335120677948,
+      "learning_rate": 9.15664989712227e-05,
+      "loss": 1.1099,
+      "step": 3129
+    },
+    {
+      "epoch": 0.190169512121028,
+      "grad_norm": 0.18649516999721527,
+      "learning_rate": 9.156117786176743e-05,
+      "loss": 1.1835,
+      "step": 3130
+    },
+    {
+      "epoch": 0.19023026915365454,
+      "grad_norm": 0.20546311140060425,
+      "learning_rate": 9.155585522887917e-05,
+      "loss": 1.1672,
+      "step": 3131
+    },
+    {
+      "epoch": 0.19029102618628108,
+      "grad_norm": 0.19135670363903046,
+      "learning_rate": 9.155053107275301e-05,
+      "loss": 1.0869,
+      "step": 3132
+    },
+    {
+      "epoch": 0.19035178321890758,
+      "grad_norm": 0.1522771567106247,
+      "learning_rate": 9.15452053935841e-05,
+      "loss": 1.0961,
+      "step": 3133
+    },
+    {
+      "epoch": 0.19041254025153412,
+      "grad_norm": 0.1502159982919693,
+      "learning_rate": 9.153987819156766e-05,
+      "loss": 1.0879,
+      "step": 3134
+    },
+    {
+      "epoch": 0.19047329728416065,
+      "grad_norm": 0.21839258074760437,
+      "learning_rate": 9.153454946689895e-05,
+      "loss": 1.0473,
+      "step": 3135
+    },
+    {
+      "epoch": 0.19053405431678716,
+      "grad_norm": 0.16287802159786224,
+      "learning_rate": 9.152921921977329e-05,
+      "loss": 1.1494,
+      "step": 3136
+    },
+    {
+      "epoch": 0.1905948113494137,
+      "grad_norm": 0.20858292281627655,
+      "learning_rate": 9.15238874503861e-05,
+      "loss": 1.129,
+      "step": 3137
+    },
+    {
+      "epoch": 0.19065556838204023,
+      "grad_norm": 0.3644711375236511,
+      "learning_rate": 9.151855415893276e-05,
+      "loss": 1.0882,
+      "step": 3138
+    },
+    {
+      "epoch": 0.19071632541466674,
+      "grad_norm": 0.2145756036043167,
+      "learning_rate": 9.15132193456088e-05,
+      "loss": 1.0757,
+      "step": 3139
+    },
+    {
+      "epoch": 0.19077708244729327,
+      "grad_norm": 0.17800205945968628,
+      "learning_rate": 9.150788301060977e-05,
+      "loss": 1.1829,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1908378394799198,
+      "grad_norm": 0.1862039417028427,
+      "learning_rate": 9.150254515413126e-05,
+      "loss": 1.1993,
+      "step": 3141
+    },
+    {
+      "epoch": 0.19089859651254631,
+      "grad_norm": 0.1509667932987213,
+      "learning_rate": 9.149720577636893e-05,
+      "loss": 1.1101,
+      "step": 3142
+    },
+    {
+      "epoch": 0.19095935354517285,
+      "grad_norm": 0.15213097631931305,
+      "learning_rate": 9.14918648775185e-05,
+      "loss": 1.1364,
+      "step": 3143
+    },
+    {
+      "epoch": 0.19102011057779938,
+      "grad_norm": 0.20119433104991913,
+      "learning_rate": 9.148652245777576e-05,
+      "loss": 1.0713,
+      "step": 3144
+    },
+    {
+      "epoch": 0.19108086761042592,
+      "grad_norm": 0.5513084530830383,
+      "learning_rate": 9.14811785173365e-05,
+      "loss": 1.339,
+      "step": 3145
+    },
+    {
+      "epoch": 0.19114162464305243,
+      "grad_norm": 0.15572676062583923,
+      "learning_rate": 9.147583305639663e-05,
+      "loss": 1.1066,
+      "step": 3146
+    },
+    {
+      "epoch": 0.19120238167567896,
+      "grad_norm": 0.21261997520923615,
+      "learning_rate": 9.14704860751521e-05,
+      "loss": 1.0429,
+      "step": 3147
+    },
+    {
+      "epoch": 0.1912631387083055,
+      "grad_norm": 0.3659953474998474,
+      "learning_rate": 9.146513757379886e-05,
+      "loss": 1.1101,
+      "step": 3148
+    },
+    {
+      "epoch": 0.191323895740932,
+      "grad_norm": 0.2075212597846985,
+      "learning_rate": 9.145978755253299e-05,
+      "loss": 1.2015,
+      "step": 3149
+    },
+    {
+      "epoch": 0.19138465277355854,
+      "grad_norm": 1.9673792123794556,
+      "learning_rate": 9.14544360115506e-05,
+      "loss": 1.0626,
+      "step": 3150
+    },
+    {
+      "epoch": 0.19144540980618507,
+      "grad_norm": 0.2159821093082428,
+      "learning_rate": 9.144908295104786e-05,
+      "loss": 1.1379,
+      "step": 3151
+    },
+    {
+      "epoch": 0.19150616683881158,
+      "grad_norm": 0.2122391015291214,
+      "learning_rate": 9.144372837122095e-05,
+      "loss": 1.0993,
+      "step": 3152
+    },
+    {
+      "epoch": 0.19156692387143812,
+      "grad_norm": 0.2374678999185562,
+      "learning_rate": 9.143837227226617e-05,
+      "loss": 1.2314,
+      "step": 3153
+    },
+    {
+      "epoch": 0.19162768090406465,
+      "grad_norm": 0.16333802044391632,
+      "learning_rate": 9.143301465437986e-05,
+      "loss": 1.0771,
+      "step": 3154
+    },
+    {
+      "epoch": 0.19168843793669116,
+      "grad_norm": 0.31628501415252686,
+      "learning_rate": 9.142765551775838e-05,
+      "loss": 1.1685,
+      "step": 3155
+    },
+    {
+      "epoch": 0.1917491949693177,
+      "grad_norm": 0.20512613654136658,
+      "learning_rate": 9.142229486259817e-05,
+      "loss": 1.2663,
+      "step": 3156
+    },
+    {
+      "epoch": 0.19180995200194423,
+      "grad_norm": 0.3468573987483978,
+      "learning_rate": 9.141693268909575e-05,
+      "loss": 1.0842,
+      "step": 3157
+    },
+    {
+      "epoch": 0.19187070903457076,
+      "grad_norm": 1.7431859970092773,
+      "learning_rate": 9.141156899744764e-05,
+      "loss": 1.1773,
+      "step": 3158
+    },
+    {
+      "epoch": 0.19193146606719727,
+      "grad_norm": 0.31735095381736755,
+      "learning_rate": 9.140620378785047e-05,
+      "loss": 1.1777,
+      "step": 3159
+    },
+    {
+      "epoch": 0.1919922230998238,
+      "grad_norm": 0.19960574805736542,
+      "learning_rate": 9.14008370605009e-05,
+      "loss": 1.1418,
+      "step": 3160
+    },
+    {
+      "epoch": 0.19205298013245034,
+      "grad_norm": 0.14055337011814117,
+      "learning_rate": 9.139546881559564e-05,
+      "loss": 1.0975,
+      "step": 3161
+    },
+    {
+      "epoch": 0.19211373716507685,
+      "grad_norm": 0.13902729749679565,
+      "learning_rate": 9.139009905333147e-05,
+      "loss": 1.049,
+      "step": 3162
+    },
+    {
+      "epoch": 0.19217449419770338,
+      "grad_norm": 0.19508013129234314,
+      "learning_rate": 9.138472777390523e-05,
+      "loss": 1.1961,
+      "step": 3163
+    },
+    {
+      "epoch": 0.19223525123032992,
+      "grad_norm": 0.2648068964481354,
+      "learning_rate": 9.137935497751378e-05,
+      "loss": 1.1785,
+      "step": 3164
+    },
+    {
+      "epoch": 0.19229600826295642,
+      "grad_norm": 1.4964137077331543,
+      "learning_rate": 9.137398066435409e-05,
+      "loss": 1.1217,
+      "step": 3165
+    },
+    {
+      "epoch": 0.19235676529558296,
+      "grad_norm": 0.1764449030160904,
+      "learning_rate": 9.136860483462312e-05,
+      "loss": 1.0715,
+      "step": 3166
+    },
+    {
+      "epoch": 0.1924175223282095,
+      "grad_norm": 0.13910436630249023,
+      "learning_rate": 9.136322748851796e-05,
+      "loss": 1.0677,
+      "step": 3167
+    },
+    {
+      "epoch": 0.19247827936083603,
+      "grad_norm": 0.23355336487293243,
+      "learning_rate": 9.135784862623569e-05,
+      "loss": 1.0591,
+      "step": 3168
+    },
+    {
+      "epoch": 0.19253903639346254,
+      "grad_norm": 0.16057777404785156,
+      "learning_rate": 9.13524682479735e-05,
+      "loss": 1.1288,
+      "step": 3169
+    },
+    {
+      "epoch": 0.19259979342608907,
+      "grad_norm": 0.19113057851791382,
+      "learning_rate": 9.134708635392858e-05,
+      "loss": 1.1947,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1926605504587156,
+      "grad_norm": 0.2551184296607971,
+      "learning_rate": 9.134170294429822e-05,
+      "loss": 1.1336,
+      "step": 3171
+    },
+    {
+      "epoch": 0.19272130749134211,
+      "grad_norm": 0.2899314761161804,
+      "learning_rate": 9.133631801927975e-05,
+      "loss": 1.1371,
+      "step": 3172
+    },
+    {
+      "epoch": 0.19278206452396865,
+      "grad_norm": 0.20496691763401031,
+      "learning_rate": 9.133093157907055e-05,
+      "loss": 1.17,
+      "step": 3173
+    },
+    {
+      "epoch": 0.19284282155659518,
+      "grad_norm": 0.20662060379981995,
+      "learning_rate": 9.132554362386807e-05,
+      "loss": 1.2085,
+      "step": 3174
+    },
+    {
+      "epoch": 0.1929035785892217,
+      "grad_norm": 0.18490777909755707,
+      "learning_rate": 9.132015415386979e-05,
+      "loss": 1.047,
+      "step": 3175
+    },
+    {
+      "epoch": 0.19296433562184823,
+      "grad_norm": 0.25164735317230225,
+      "learning_rate": 9.131476316927327e-05,
+      "loss": 1.2809,
+      "step": 3176
+    },
+    {
+      "epoch": 0.19302509265447476,
+      "grad_norm": 0.16715030372142792,
+      "learning_rate": 9.130937067027615e-05,
+      "loss": 1.1962,
+      "step": 3177
+    },
+    {
+      "epoch": 0.19308584968710127,
+      "grad_norm": 0.46896255016326904,
+      "learning_rate": 9.130397665707603e-05,
+      "loss": 1.1489,
+      "step": 3178
+    },
+    {
+      "epoch": 0.1931466067197278,
+      "grad_norm": 0.1492065042257309,
+      "learning_rate": 9.129858112987066e-05,
+      "loss": 1.0597,
+      "step": 3179
+    },
+    {
+      "epoch": 0.19320736375235434,
+      "grad_norm": 0.24866871535778046,
+      "learning_rate": 9.129318408885784e-05,
+      "loss": 1.0984,
+      "step": 3180
+    },
+    {
+      "epoch": 0.19326812078498087,
+      "grad_norm": 0.2036585658788681,
+      "learning_rate": 9.128778553423536e-05,
+      "loss": 1.1276,
+      "step": 3181
+    },
+    {
+      "epoch": 0.19332887781760738,
+      "grad_norm": 0.3559352457523346,
+      "learning_rate": 9.128238546620114e-05,
+      "loss": 1.188,
+      "step": 3182
+    },
+    {
+      "epoch": 0.19338963485023392,
+      "grad_norm": 0.22386854887008667,
+      "learning_rate": 9.127698388495307e-05,
+      "loss": 1.0944,
+      "step": 3183
+    },
+    {
+      "epoch": 0.19345039188286045,
+      "grad_norm": 0.17664191126823425,
+      "learning_rate": 9.12715807906892e-05,
+      "loss": 1.1793,
+      "step": 3184
+    },
+    {
+      "epoch": 0.19351114891548696,
+      "grad_norm": 0.27878689765930176,
+      "learning_rate": 9.126617618360756e-05,
+      "loss": 1.2249,
+      "step": 3185
+    },
+    {
+      "epoch": 0.1935719059481135,
+      "grad_norm": 0.1589423269033432,
+      "learning_rate": 9.126077006390623e-05,
+      "loss": 1.0777,
+      "step": 3186
+    },
+    {
+      "epoch": 0.19363266298074003,
+      "grad_norm": 0.1803324818611145,
+      "learning_rate": 9.125536243178342e-05,
+      "loss": 1.1675,
+      "step": 3187
+    },
+    {
+      "epoch": 0.19369342001336654,
+      "grad_norm": 0.22235427796840668,
+      "learning_rate": 9.124995328743732e-05,
+      "loss": 1.2067,
+      "step": 3188
+    },
+    {
+      "epoch": 0.19375417704599307,
+      "grad_norm": 0.15766115486621857,
+      "learning_rate": 9.124454263106622e-05,
+      "loss": 1.1238,
+      "step": 3189
+    },
+    {
+      "epoch": 0.1938149340786196,
+      "grad_norm": 0.2703237235546112,
+      "learning_rate": 9.123913046286841e-05,
+      "loss": 1.2034,
+      "step": 3190
+    },
+    {
+      "epoch": 0.19387569111124614,
+      "grad_norm": 0.13763552904129028,
+      "learning_rate": 9.123371678304232e-05,
+      "loss": 1.1076,
+      "step": 3191
+    },
+    {
+      "epoch": 0.19393644814387265,
+      "grad_norm": 0.1967121958732605,
+      "learning_rate": 9.122830159178636e-05,
+      "loss": 1.1278,
+      "step": 3192
+    },
+    {
+      "epoch": 0.19399720517649918,
+      "grad_norm": 0.23657777905464172,
+      "learning_rate": 9.122288488929906e-05,
+      "loss": 1.0981,
+      "step": 3193
+    },
+    {
+      "epoch": 0.19405796220912572,
+      "grad_norm": 1.0357747077941895,
+      "learning_rate": 9.121746667577893e-05,
+      "loss": 1.0994,
+      "step": 3194
+    },
+    {
+      "epoch": 0.19411871924175222,
+      "grad_norm": 0.15961074829101562,
+      "learning_rate": 9.12120469514246e-05,
+      "loss": 1.1181,
+      "step": 3195
+    },
+    {
+      "epoch": 0.19417947627437876,
+      "grad_norm": 0.23469798266887665,
+      "learning_rate": 9.12066257164347e-05,
+      "loss": 1.1791,
+      "step": 3196
+    },
+    {
+      "epoch": 0.1942402333070053,
+      "grad_norm": 0.3526883125305176,
+      "learning_rate": 9.120120297100799e-05,
+      "loss": 1.163,
+      "step": 3197
+    },
+    {
+      "epoch": 0.1943009903396318,
+      "grad_norm": 1.5761868953704834,
+      "learning_rate": 9.119577871534322e-05,
+      "loss": 1.0717,
+      "step": 3198
+    },
+    {
+      "epoch": 0.19436174737225834,
+      "grad_norm": 0.41292497515678406,
+      "learning_rate": 9.119035294963921e-05,
+      "loss": 1.13,
+      "step": 3199
+    },
+    {
+      "epoch": 0.19442250440488487,
+      "grad_norm": 0.2322758138179779,
+      "learning_rate": 9.118492567409486e-05,
+      "loss": 1.1806,
+      "step": 3200
+    },
+    {
+      "epoch": 0.19448326143751138,
+      "grad_norm": 0.30504292249679565,
+      "learning_rate": 9.117949688890909e-05,
+      "loss": 1.2634,
+      "step": 3201
+    },
+    {
+      "epoch": 0.1945440184701379,
+      "grad_norm": 0.2333499789237976,
+      "learning_rate": 9.11740665942809e-05,
+      "loss": 1.0816,
+      "step": 3202
+    },
+    {
+      "epoch": 0.19460477550276445,
+      "grad_norm": 0.2496776282787323,
+      "learning_rate": 9.116863479040936e-05,
+      "loss": 1.0963,
+      "step": 3203
+    },
+    {
+      "epoch": 0.19466553253539098,
+      "grad_norm": 0.1488950103521347,
+      "learning_rate": 9.116320147749354e-05,
+      "loss": 1.1188,
+      "step": 3204
+    },
+    {
+      "epoch": 0.1947262895680175,
+      "grad_norm": 0.1364702731370926,
+      "learning_rate": 9.11577666557326e-05,
+      "loss": 1.0807,
+      "step": 3205
+    },
+    {
+      "epoch": 0.19478704660064403,
+      "grad_norm": 0.2220604419708252,
+      "learning_rate": 9.115233032532578e-05,
+      "loss": 1.0772,
+      "step": 3206
+    },
+    {
+      "epoch": 0.19484780363327056,
+      "grad_norm": 0.1982111632823944,
+      "learning_rate": 9.114689248647235e-05,
+      "loss": 1.1418,
+      "step": 3207
+    },
+    {
+      "epoch": 0.19490856066589707,
+      "grad_norm": 0.16218845546245575,
+      "learning_rate": 9.114145313937161e-05,
+      "loss": 1.0943,
+      "step": 3208
+    },
+    {
+      "epoch": 0.1949693176985236,
+      "grad_norm": 0.19595873355865479,
+      "learning_rate": 9.113601228422294e-05,
+      "loss": 1.2117,
+      "step": 3209
+    },
+    {
+      "epoch": 0.19503007473115014,
+      "grad_norm": 0.20046143233776093,
+      "learning_rate": 9.11305699212258e-05,
+      "loss": 1.1831,
+      "step": 3210
+    },
+    {
+      "epoch": 0.19509083176377665,
+      "grad_norm": 0.19938664138317108,
+      "learning_rate": 9.112512605057968e-05,
+      "loss": 1.1029,
+      "step": 3211
+    },
+    {
+      "epoch": 0.19515158879640318,
+      "grad_norm": 0.29923155903816223,
+      "learning_rate": 9.111968067248411e-05,
+      "loss": 1.2037,
+      "step": 3212
+    },
+    {
+      "epoch": 0.19521234582902972,
+      "grad_norm": 0.12634147703647614,
+      "learning_rate": 9.111423378713868e-05,
+      "loss": 1.0746,
+      "step": 3213
+    },
+    {
+      "epoch": 0.19527310286165625,
+      "grad_norm": 0.14267168939113617,
+      "learning_rate": 9.110878539474308e-05,
+      "loss": 1.1066,
+      "step": 3214
+    },
+    {
+      "epoch": 0.19533385989428276,
+      "grad_norm": 0.18030843138694763,
+      "learning_rate": 9.1103335495497e-05,
+      "loss": 1.1082,
+      "step": 3215
+    },
+    {
+      "epoch": 0.1953946169269093,
+      "grad_norm": 0.12984958291053772,
+      "learning_rate": 9.10978840896002e-05,
+      "loss": 1.0789,
+      "step": 3216
+    },
+    {
+      "epoch": 0.19545537395953583,
+      "grad_norm": 0.4353140592575073,
+      "learning_rate": 9.109243117725254e-05,
+      "loss": 1.1851,
+      "step": 3217
+    },
+    {
+      "epoch": 0.19551613099216233,
+      "grad_norm": 0.2476755976676941,
+      "learning_rate": 9.108697675865385e-05,
+      "loss": 1.3506,
+      "step": 3218
+    },
+    {
+      "epoch": 0.19557688802478887,
+      "grad_norm": 0.19675281643867493,
+      "learning_rate": 9.10815208340041e-05,
+      "loss": 1.1011,
+      "step": 3219
+    },
+    {
+      "epoch": 0.1956376450574154,
+      "grad_norm": 0.1628730148077011,
+      "learning_rate": 9.107606340350325e-05,
+      "loss": 1.152,
+      "step": 3220
+    },
+    {
+      "epoch": 0.1956984020900419,
+      "grad_norm": 0.3081006705760956,
+      "learning_rate": 9.107060446735138e-05,
+      "loss": 1.0779,
+      "step": 3221
+    },
+    {
+      "epoch": 0.19575915912266845,
+      "grad_norm": 0.15976488590240479,
+      "learning_rate": 9.106514402574854e-05,
+      "loss": 1.0751,
+      "step": 3222
+    },
+    {
+      "epoch": 0.19581991615529498,
+      "grad_norm": 0.16874565184116364,
+      "learning_rate": 9.105968207889492e-05,
+      "loss": 1.0956,
+      "step": 3223
+    },
+    {
+      "epoch": 0.1958806731879215,
+      "grad_norm": 0.17898161709308624,
+      "learning_rate": 9.105421862699071e-05,
+      "loss": 1.1343,
+      "step": 3224
+    },
+    {
+      "epoch": 0.19594143022054802,
+      "grad_norm": 0.16883991658687592,
+      "learning_rate": 9.104875367023619e-05,
+      "loss": 1.1453,
+      "step": 3225
+    },
+    {
+      "epoch": 0.19600218725317456,
+      "grad_norm": 0.3435705006122589,
+      "learning_rate": 9.104328720883166e-05,
+      "loss": 1.1495,
+      "step": 3226
+    },
+    {
+      "epoch": 0.1960629442858011,
+      "grad_norm": 0.16361819207668304,
+      "learning_rate": 9.103781924297751e-05,
+      "loss": 1.1403,
+      "step": 3227
+    },
+    {
+      "epoch": 0.1961237013184276,
+      "grad_norm": 0.25067466497421265,
+      "learning_rate": 9.103234977287416e-05,
+      "loss": 1.1475,
+      "step": 3228
+    },
+    {
+      "epoch": 0.19618445835105414,
+      "grad_norm": 1.0418776273727417,
+      "learning_rate": 9.10268787987221e-05,
+      "loss": 1.1273,
+      "step": 3229
+    },
+    {
+      "epoch": 0.19624521538368067,
+      "grad_norm": 0.17021557688713074,
+      "learning_rate": 9.102140632072187e-05,
+      "loss": 1.0359,
+      "step": 3230
+    },
+    {
+      "epoch": 0.19630597241630718,
+      "grad_norm": 0.24031531810760498,
+      "learning_rate": 9.101593233907405e-05,
+      "loss": 1.092,
+      "step": 3231
+    },
+    {
+      "epoch": 0.1963667294489337,
+      "grad_norm": 5.0789055824279785,
+      "learning_rate": 9.101045685397931e-05,
+      "loss": 1.0906,
+      "step": 3232
+    },
+    {
+      "epoch": 0.19642748648156025,
+      "grad_norm": 0.26952677965164185,
+      "learning_rate": 9.100497986563833e-05,
+      "loss": 1.1626,
+      "step": 3233
+    },
+    {
+      "epoch": 0.19648824351418676,
+      "grad_norm": 0.2004709094762802,
+      "learning_rate": 9.099950137425193e-05,
+      "loss": 1.3217,
+      "step": 3234
+    },
+    {
+      "epoch": 0.1965490005468133,
+      "grad_norm": 0.17830704152584076,
+      "learning_rate": 9.099402138002084e-05,
+      "loss": 1.0553,
+      "step": 3235
+    },
+    {
+      "epoch": 0.19660975757943983,
+      "grad_norm": 0.2199990600347519,
+      "learning_rate": 9.098853988314599e-05,
+      "loss": 1.0785,
+      "step": 3236
+    },
+    {
+      "epoch": 0.19667051461206636,
+      "grad_norm": 0.20517028868198395,
+      "learning_rate": 9.098305688382829e-05,
+      "loss": 1.1831,
+      "step": 3237
+    },
+    {
+      "epoch": 0.19673127164469287,
+      "grad_norm": 0.30997705459594727,
+      "learning_rate": 9.097757238226872e-05,
+      "loss": 1.1353,
+      "step": 3238
+    },
+    {
+      "epoch": 0.1967920286773194,
+      "grad_norm": 0.19933241605758667,
+      "learning_rate": 9.097208637866829e-05,
+      "loss": 1.1661,
+      "step": 3239
+    },
+    {
+      "epoch": 0.19685278570994594,
+      "grad_norm": 0.16712698340415955,
+      "learning_rate": 9.096659887322813e-05,
+      "loss": 1.0909,
+      "step": 3240
+    },
+    {
+      "epoch": 0.19691354274257244,
+      "grad_norm": 2.340427875518799,
+      "learning_rate": 9.096110986614938e-05,
+      "loss": 1.0552,
+      "step": 3241
+    },
+    {
+      "epoch": 0.19697429977519898,
+      "grad_norm": 42.843692779541016,
+      "learning_rate": 9.095561935763321e-05,
+      "loss": 1.1362,
+      "step": 3242
+    },
+    {
+      "epoch": 0.19703505680782551,
+      "grad_norm": 0.3771587312221527,
+      "learning_rate": 9.095012734788093e-05,
+      "loss": 1.1772,
+      "step": 3243
+    },
+    {
+      "epoch": 0.19709581384045202,
+      "grad_norm": 0.3147752285003662,
+      "learning_rate": 9.094463383709379e-05,
+      "loss": 1.1349,
+      "step": 3244
+    },
+    {
+      "epoch": 0.19715657087307856,
+      "grad_norm": 0.31446409225463867,
+      "learning_rate": 9.093913882547318e-05,
+      "loss": 1.0958,
+      "step": 3245
+    },
+    {
+      "epoch": 0.1972173279057051,
+      "grad_norm": 0.3560512661933899,
+      "learning_rate": 9.093364231322052e-05,
+      "loss": 1.0741,
+      "step": 3246
+    },
+    {
+      "epoch": 0.1972780849383316,
+      "grad_norm": 0.16011124849319458,
+      "learning_rate": 9.09281443005373e-05,
+      "loss": 1.0644,
+      "step": 3247
+    },
+    {
+      "epoch": 0.19733884197095813,
+      "grad_norm": 0.21780221164226532,
+      "learning_rate": 9.092264478762505e-05,
+      "loss": 1.0863,
+      "step": 3248
+    },
+    {
+      "epoch": 0.19739959900358467,
+      "grad_norm": 0.2751162052154541,
+      "learning_rate": 9.091714377468532e-05,
+      "loss": 1.1093,
+      "step": 3249
+    },
+    {
+      "epoch": 0.1974603560362112,
+      "grad_norm": 0.17367391288280487,
+      "learning_rate": 9.09116412619198e-05,
+      "loss": 1.115,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1975211130688377,
+      "grad_norm": 1.4182554483413696,
+      "learning_rate": 9.090613724953014e-05,
+      "loss": 1.256,
+      "step": 3251
+    },
+    {
+      "epoch": 0.19758187010146425,
+      "grad_norm": 0.3456888794898987,
+      "learning_rate": 9.090063173771813e-05,
+      "loss": 1.1735,
+      "step": 3252
+    },
+    {
+      "epoch": 0.19764262713409078,
+      "grad_norm": 0.21287770569324493,
+      "learning_rate": 9.089512472668554e-05,
+      "loss": 1.064,
+      "step": 3253
+    },
+    {
+      "epoch": 0.1977033841667173,
+      "grad_norm": 0.49738410115242004,
+      "learning_rate": 9.088961621663425e-05,
+      "loss": 1.1764,
+      "step": 3254
+    },
+    {
+      "epoch": 0.19776414119934382,
+      "grad_norm": 0.19693298637866974,
+      "learning_rate": 9.08841062077662e-05,
+      "loss": 1.2348,
+      "step": 3255
+    },
+    {
+      "epoch": 0.19782489823197036,
+      "grad_norm": 0.1314329206943512,
+      "learning_rate": 9.08785947002833e-05,
+      "loss": 1.1,
+      "step": 3256
+    },
+    {
+      "epoch": 0.19788565526459687,
+      "grad_norm": 0.23774775862693787,
+      "learning_rate": 9.087308169438765e-05,
+      "loss": 1.1191,
+      "step": 3257
+    },
+    {
+      "epoch": 0.1979464122972234,
+      "grad_norm": 0.25573259592056274,
+      "learning_rate": 9.086756719028124e-05,
+      "loss": 1.1807,
+      "step": 3258
+    },
+    {
+      "epoch": 0.19800716932984994,
+      "grad_norm": 0.5008425116539001,
+      "learning_rate": 9.086205118816628e-05,
+      "loss": 1.2152,
+      "step": 3259
+    },
+    {
+      "epoch": 0.19806792636247644,
+      "grad_norm": 0.19056855142116547,
+      "learning_rate": 9.085653368824494e-05,
+      "loss": 1.1101,
+      "step": 3260
+    },
+    {
+      "epoch": 0.19812868339510298,
+      "grad_norm": 1.883497714996338,
+      "learning_rate": 9.085101469071945e-05,
+      "loss": 1.0974,
+      "step": 3261
+    },
+    {
+      "epoch": 0.1981894404277295,
+      "grad_norm": 0.18575043976306915,
+      "learning_rate": 9.084549419579213e-05,
+      "loss": 1.075,
+      "step": 3262
+    },
+    {
+      "epoch": 0.19825019746035605,
+      "grad_norm": 0.1883113831281662,
+      "learning_rate": 9.083997220366533e-05,
+      "loss": 1.1047,
+      "step": 3263
+    },
+    {
+      "epoch": 0.19831095449298256,
+      "grad_norm": 0.1844778209924698,
+      "learning_rate": 9.083444871454144e-05,
+      "loss": 1.1149,
+      "step": 3264
+    },
+    {
+      "epoch": 0.1983717115256091,
+      "grad_norm": 0.16545872390270233,
+      "learning_rate": 9.082892372862295e-05,
+      "loss": 1.1239,
+      "step": 3265
+    },
+    {
+      "epoch": 0.19843246855823563,
+      "grad_norm": 0.21504627168178558,
+      "learning_rate": 9.082339724611236e-05,
+      "loss": 1.0891,
+      "step": 3266
+    },
+    {
+      "epoch": 0.19849322559086213,
+      "grad_norm": 0.2592552602291107,
+      "learning_rate": 9.081786926721226e-05,
+      "loss": 1.0709,
+      "step": 3267
+    },
+    {
+      "epoch": 0.19855398262348867,
+      "grad_norm": 0.3487328290939331,
+      "learning_rate": 9.081233979212526e-05,
+      "loss": 1.1722,
+      "step": 3268
+    },
+    {
+      "epoch": 0.1986147396561152,
+      "grad_norm": 0.28977230191230774,
+      "learning_rate": 9.080680882105406e-05,
+      "loss": 1.1047,
+      "step": 3269
+    },
+    {
+      "epoch": 0.1986754966887417,
+      "grad_norm": 0.4293149411678314,
+      "learning_rate": 9.080127635420142e-05,
+      "loss": 1.1267,
+      "step": 3270
+    },
+    {
+      "epoch": 0.19873625372136824,
+      "grad_norm": 0.18525612354278564,
+      "learning_rate": 9.079574239177008e-05,
+      "loss": 1.1272,
+      "step": 3271
+    },
+    {
+      "epoch": 0.19879701075399478,
+      "grad_norm": 0.25222426652908325,
+      "learning_rate": 9.079020693396292e-05,
+      "loss": 1.134,
+      "step": 3272
+    },
+    {
+      "epoch": 0.19885776778662131,
+      "grad_norm": 0.2567189931869507,
+      "learning_rate": 9.078466998098285e-05,
+      "loss": 1.1383,
+      "step": 3273
+    },
+    {
+      "epoch": 0.19891852481924782,
+      "grad_norm": 0.560010552406311,
+      "learning_rate": 9.077913153303282e-05,
+      "loss": 1.2302,
+      "step": 3274
+    },
+    {
+      "epoch": 0.19897928185187436,
+      "grad_norm": 0.2820112109184265,
+      "learning_rate": 9.077359159031582e-05,
+      "loss": 1.1325,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1990400388845009,
+      "grad_norm": 0.15497444570064545,
+      "learning_rate": 9.076805015303496e-05,
+      "loss": 1.0117,
+      "step": 3276
+    },
+    {
+      "epoch": 0.1991007959171274,
+      "grad_norm": 0.21076422929763794,
+      "learning_rate": 9.076250722139333e-05,
+      "loss": 1.1599,
+      "step": 3277
+    },
+    {
+      "epoch": 0.19916155294975393,
+      "grad_norm": 0.2797802984714508,
+      "learning_rate": 9.075696279559414e-05,
+      "loss": 1.2243,
+      "step": 3278
+    },
+    {
+      "epoch": 0.19922230998238047,
+      "grad_norm": 0.2437891662120819,
+      "learning_rate": 9.075141687584057e-05,
+      "loss": 1.1717,
+      "step": 3279
+    },
+    {
+      "epoch": 0.19928306701500698,
+      "grad_norm": 0.7717027068138123,
+      "learning_rate": 9.074586946233595e-05,
+      "loss": 1.1687,
+      "step": 3280
+    },
+    {
+      "epoch": 0.1993438240476335,
+      "grad_norm": 0.2633930742740631,
+      "learning_rate": 9.074032055528361e-05,
+      "loss": 1.1992,
+      "step": 3281
+    },
+    {
+      "epoch": 0.19940458108026005,
+      "grad_norm": 0.2726195454597473,
+      "learning_rate": 9.073477015488696e-05,
+      "loss": 1.1327,
+      "step": 3282
+    },
+    {
+      "epoch": 0.19946533811288655,
+      "grad_norm": 0.1584884077310562,
+      "learning_rate": 9.072921826134942e-05,
+      "loss": 1.1177,
+      "step": 3283
+    },
+    {
+      "epoch": 0.1995260951455131,
+      "grad_norm": 0.25239813327789307,
+      "learning_rate": 9.072366487487451e-05,
+      "loss": 1.0695,
+      "step": 3284
+    },
+    {
+      "epoch": 0.19958685217813962,
+      "grad_norm": 0.18151557445526123,
+      "learning_rate": 9.07181099956658e-05,
+      "loss": 1.1048,
+      "step": 3285
+    },
+    {
+      "epoch": 0.19964760921076616,
+      "grad_norm": 0.526943027973175,
+      "learning_rate": 9.071255362392688e-05,
+      "loss": 1.114,
+      "step": 3286
+    },
+    {
+      "epoch": 0.19970836624339267,
+      "grad_norm": 0.28988730907440186,
+      "learning_rate": 9.070699575986146e-05,
+      "loss": 1.1565,
+      "step": 3287
+    },
+    {
+      "epoch": 0.1997691232760192,
+      "grad_norm": 0.19109733402729034,
+      "learning_rate": 9.070143640367323e-05,
+      "loss": 1.1704,
+      "step": 3288
+    },
+    {
+      "epoch": 0.19982988030864574,
+      "grad_norm": 0.7349984049797058,
+      "learning_rate": 9.069587555556598e-05,
+      "loss": 1.2812,
+      "step": 3289
+    },
+    {
+      "epoch": 0.19989063734127224,
+      "grad_norm": 0.22742979228496552,
+      "learning_rate": 9.069031321574354e-05,
+      "loss": 1.1273,
+      "step": 3290
+    },
+    {
+      "epoch": 0.19995139437389878,
+      "grad_norm": 0.22230419516563416,
+      "learning_rate": 9.068474938440983e-05,
+      "loss": 1.108,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2000121514065253,
+      "grad_norm": 0.3051619529724121,
+      "learning_rate": 9.067918406176875e-05,
+      "loss": 1.1005,
+      "step": 3292
+    },
+    {
+      "epoch": 0.20007290843915182,
+      "grad_norm": 0.1708410084247589,
+      "learning_rate": 9.06736172480243e-05,
+      "loss": 1.1002,
+      "step": 3293
+    },
+    {
+      "epoch": 0.20013366547177835,
+      "grad_norm": 0.1780860424041748,
+      "learning_rate": 9.066804894338056e-05,
+      "loss": 1.0511,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2001944225044049,
+      "grad_norm": 0.18345783650875092,
+      "learning_rate": 9.066247914804164e-05,
+      "loss": 1.0988,
+      "step": 3295
+    },
+    {
+      "epoch": 0.20025517953703142,
+      "grad_norm": 0.2013816386461258,
+      "learning_rate": 9.065690786221168e-05,
+      "loss": 1.2174,
+      "step": 3296
+    },
+    {
+      "epoch": 0.20031593656965793,
+      "grad_norm": 0.17284561693668365,
+      "learning_rate": 9.06513350860949e-05,
+      "loss": 1.0817,
+      "step": 3297
+    },
+    {
+      "epoch": 0.20037669360228447,
+      "grad_norm": 0.14359714090824127,
+      "learning_rate": 9.064576081989557e-05,
+      "loss": 1.0861,
+      "step": 3298
+    },
+    {
+      "epoch": 0.200437450634911,
+      "grad_norm": 0.7734746932983398,
+      "learning_rate": 9.064018506381802e-05,
+      "loss": 1.1187,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2004982076675375,
+      "grad_norm": 0.15058156847953796,
+      "learning_rate": 9.063460781806663e-05,
+      "loss": 1.1217,
+      "step": 3300
+    },
+    {
+      "epoch": 0.20055896470016404,
+      "grad_norm": 0.2474808543920517,
+      "learning_rate": 9.062902908284585e-05,
+      "loss": 1.1728,
+      "step": 3301
+    },
+    {
+      "epoch": 0.20061972173279058,
+      "grad_norm": 0.14751644432544708,
+      "learning_rate": 9.062344885836014e-05,
+      "loss": 1.1342,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2006804787654171,
+      "grad_norm": 0.2249361276626587,
+      "learning_rate": 9.061786714481406e-05,
+      "loss": 1.1267,
+      "step": 3303
+    },
+    {
+      "epoch": 0.20074123579804362,
+      "grad_norm": 0.23503537476062775,
+      "learning_rate": 9.061228394241222e-05,
+      "loss": 1.1269,
+      "step": 3304
+    },
+    {
+      "epoch": 0.20080199283067016,
+      "grad_norm": 0.24838240444660187,
+      "learning_rate": 9.060669925135925e-05,
+      "loss": 1.0512,
+      "step": 3305
+    },
+    {
+      "epoch": 0.20086274986329666,
+      "grad_norm": 0.17550477385520935,
+      "learning_rate": 9.060111307185988e-05,
+      "loss": 1.0657,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2009235068959232,
+      "grad_norm": 0.2066640853881836,
+      "learning_rate": 9.059552540411886e-05,
+      "loss": 1.1623,
+      "step": 3307
+    },
+    {
+      "epoch": 0.20098426392854973,
+      "grad_norm": 0.16343830525875092,
+      "learning_rate": 9.0589936248341e-05,
+      "loss": 1.1079,
+      "step": 3308
+    },
+    {
+      "epoch": 0.20104502096117627,
+      "grad_norm": 0.33369284868240356,
+      "learning_rate": 9.058434560473119e-05,
+      "loss": 1.0754,
+      "step": 3309
+    },
+    {
+      "epoch": 0.20110577799380278,
+      "grad_norm": 0.4190661311149597,
+      "learning_rate": 9.057875347349434e-05,
+      "loss": 1.125,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2011665350264293,
+      "grad_norm": 1.5997530221939087,
+      "learning_rate": 9.057315985483545e-05,
+      "loss": 1.1197,
+      "step": 3311
+    },
+    {
+      "epoch": 0.20122729205905585,
+      "grad_norm": 0.1344272345304489,
+      "learning_rate": 9.056756474895954e-05,
+      "loss": 1.0992,
+      "step": 3312
+    },
+    {
+      "epoch": 0.20128804909168235,
+      "grad_norm": 0.14829017221927643,
+      "learning_rate": 9.05619681560717e-05,
+      "loss": 1.0901,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2013488061243089,
+      "grad_norm": 0.20510458946228027,
+      "learning_rate": 9.055637007637707e-05,
+      "loss": 1.1626,
+      "step": 3314
+    },
+    {
+      "epoch": 0.20140956315693542,
+      "grad_norm": 0.16008488833904266,
+      "learning_rate": 9.055077051008087e-05,
+      "loss": 1.0767,
+      "step": 3315
+    },
+    {
+      "epoch": 0.20147032018956193,
+      "grad_norm": 0.23861490190029144,
+      "learning_rate": 9.054516945738833e-05,
+      "loss": 1.1874,
+      "step": 3316
+    },
+    {
+      "epoch": 0.20153107722218846,
+      "grad_norm": 0.16540761291980743,
+      "learning_rate": 9.053956691850476e-05,
+      "loss": 1.0869,
+      "step": 3317
+    },
+    {
+      "epoch": 0.201591834254815,
+      "grad_norm": 0.3213448226451874,
+      "learning_rate": 9.053396289363553e-05,
+      "loss": 1.4313,
+      "step": 3318
+    },
+    {
+      "epoch": 0.20165259128744153,
+      "grad_norm": 0.22906824946403503,
+      "learning_rate": 9.052835738298606e-05,
+      "loss": 1.1131,
+      "step": 3319
+    },
+    {
+      "epoch": 0.20171334832006804,
+      "grad_norm": 0.19157810509204865,
+      "learning_rate": 9.052275038676183e-05,
+      "loss": 1.057,
+      "step": 3320
+    },
+    {
+      "epoch": 0.20177410535269458,
+      "grad_norm": 0.31121575832366943,
+      "learning_rate": 9.051714190516834e-05,
+      "loss": 1.2415,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2018348623853211,
+      "grad_norm": 0.4245741367340088,
+      "learning_rate": 9.051153193841118e-05,
+      "loss": 1.318,
+      "step": 3322
+    },
+    {
+      "epoch": 0.20189561941794762,
+      "grad_norm": 0.30429720878601074,
+      "learning_rate": 9.050592048669597e-05,
+      "loss": 1.0931,
+      "step": 3323
+    },
+    {
+      "epoch": 0.20195637645057415,
+      "grad_norm": 0.16724011301994324,
+      "learning_rate": 9.050030755022844e-05,
+      "loss": 1.103,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2020171334832007,
+      "grad_norm": 0.16439297795295715,
+      "learning_rate": 9.04946931292143e-05,
+      "loss": 1.085,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2020778905158272,
+      "grad_norm": 0.24683915078639984,
+      "learning_rate": 9.048907722385935e-05,
+      "loss": 1.2018,
+      "step": 3326
+    },
+    {
+      "epoch": 0.20213864754845373,
+      "grad_norm": 0.16863733530044556,
+      "learning_rate": 9.048345983436946e-05,
+      "loss": 1.1387,
+      "step": 3327
+    },
+    {
+      "epoch": 0.20219940458108027,
+      "grad_norm": 0.9390217065811157,
+      "learning_rate": 9.047784096095052e-05,
+      "loss": 1.1491,
+      "step": 3328
+    },
+    {
+      "epoch": 0.20226016161370677,
+      "grad_norm": 0.2639472484588623,
+      "learning_rate": 9.047222060380849e-05,
+      "loss": 1.0758,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2023209186463333,
+      "grad_norm": 0.16340938210487366,
+      "learning_rate": 9.046659876314938e-05,
+      "loss": 1.0856,
+      "step": 3330
+    },
+    {
+      "epoch": 0.20238167567895984,
+      "grad_norm": 0.2741747498512268,
+      "learning_rate": 9.046097543917928e-05,
+      "loss": 1.1686,
+      "step": 3331
+    },
+    {
+      "epoch": 0.20244243271158638,
+      "grad_norm": 0.22821323573589325,
+      "learning_rate": 9.045535063210432e-05,
+      "loss": 1.1587,
+      "step": 3332
+    },
+    {
+      "epoch": 0.20250318974421289,
+      "grad_norm": 4.2999491691589355,
+      "learning_rate": 9.044972434213064e-05,
+      "loss": 1.1362,
+      "step": 3333
+    },
+    {
+      "epoch": 0.20256394677683942,
+      "grad_norm": 0.3759039044380188,
+      "learning_rate": 9.04440965694645e-05,
+      "loss": 1.2306,
+      "step": 3334
+    },
+    {
+      "epoch": 0.20262470380946596,
+      "grad_norm": 1.0817079544067383,
+      "learning_rate": 9.04384673143122e-05,
+      "loss": 1.1433,
+      "step": 3335
+    },
+    {
+      "epoch": 0.20268546084209246,
+      "grad_norm": 0.2255803495645523,
+      "learning_rate": 9.043283657688005e-05,
+      "loss": 1.0925,
+      "step": 3336
+    },
+    {
+      "epoch": 0.202746217874719,
+      "grad_norm": 0.2970569431781769,
+      "learning_rate": 9.042720435737447e-05,
+      "loss": 1.156,
+      "step": 3337
+    },
+    {
+      "epoch": 0.20280697490734553,
+      "grad_norm": 0.1731918901205063,
+      "learning_rate": 9.042157065600188e-05,
+      "loss": 1.0979,
+      "step": 3338
+    },
+    {
+      "epoch": 0.20286773193997204,
+      "grad_norm": 0.19859391450881958,
+      "learning_rate": 9.041593547296883e-05,
+      "loss": 1.0874,
+      "step": 3339
+    },
+    {
+      "epoch": 0.20292848897259858,
+      "grad_norm": 0.21392711997032166,
+      "learning_rate": 9.041029880848183e-05,
+      "loss": 1.2161,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2029892460052251,
+      "grad_norm": 0.1796914041042328,
+      "learning_rate": 9.040466066274753e-05,
+      "loss": 1.1418,
+      "step": 3341
+    },
+    {
+      "epoch": 0.20305000303785165,
+      "grad_norm": 0.2005152553319931,
+      "learning_rate": 9.039902103597258e-05,
+      "loss": 1.1152,
+      "step": 3342
+    },
+    {
+      "epoch": 0.20311076007047815,
+      "grad_norm": 0.18333251774311066,
+      "learning_rate": 9.03933799283637e-05,
+      "loss": 1.1705,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2031715171031047,
+      "grad_norm": 0.5630664229393005,
+      "learning_rate": 9.038773734012768e-05,
+      "loss": 1.0914,
+      "step": 3344
+    },
+    {
+      "epoch": 0.20323227413573122,
+      "grad_norm": 0.35749906301498413,
+      "learning_rate": 9.038209327147133e-05,
+      "loss": 1.2042,
+      "step": 3345
+    },
+    {
+      "epoch": 0.20329303116835773,
+      "grad_norm": 0.18422846496105194,
+      "learning_rate": 9.037644772260155e-05,
+      "loss": 1.108,
+      "step": 3346
+    },
+    {
+      "epoch": 0.20335378820098426,
+      "grad_norm": 0.179338276386261,
+      "learning_rate": 9.037080069372527e-05,
+      "loss": 1.1074,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2034145452336108,
+      "grad_norm": 0.22576642036437988,
+      "learning_rate": 9.03651521850495e-05,
+      "loss": 1.134,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2034753022662373,
+      "grad_norm": 0.1988999992609024,
+      "learning_rate": 9.035950219678126e-05,
+      "loss": 1.1069,
+      "step": 3349
+    },
+    {
+      "epoch": 0.20353605929886384,
+      "grad_norm": 0.44660013914108276,
+      "learning_rate": 9.035385072912768e-05,
+      "loss": 1.264,
+      "step": 3350
+    },
+    {
+      "epoch": 0.20359681633149038,
+      "grad_norm": 0.31285372376441956,
+      "learning_rate": 9.034819778229588e-05,
+      "loss": 1.2506,
+      "step": 3351
+    },
+    {
+      "epoch": 0.20365757336411688,
+      "grad_norm": 0.16578587889671326,
+      "learning_rate": 9.034254335649311e-05,
+      "loss": 1.0862,
+      "step": 3352
+    },
+    {
+      "epoch": 0.20371833039674342,
+      "grad_norm": 0.14541122317314148,
+      "learning_rate": 9.033688745192661e-05,
+      "loss": 1.0819,
+      "step": 3353
+    },
+    {
+      "epoch": 0.20377908742936995,
+      "grad_norm": 0.17592382431030273,
+      "learning_rate": 9.03312300688037e-05,
+      "loss": 1.0943,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2038398444619965,
+      "grad_norm": 0.2473861128091812,
+      "learning_rate": 9.032557120733176e-05,
+      "loss": 1.1806,
+      "step": 3355
+    },
+    {
+      "epoch": 0.203900601494623,
+      "grad_norm": 0.22891579568386078,
+      "learning_rate": 9.031991086771823e-05,
+      "loss": 1.1701,
+      "step": 3356
+    },
+    {
+      "epoch": 0.20396135852724953,
+      "grad_norm": 0.1464105099439621,
+      "learning_rate": 9.031424905017054e-05,
+      "loss": 1.0917,
+      "step": 3357
+    },
+    {
+      "epoch": 0.20402211555987607,
+      "grad_norm": 0.18087196350097656,
+      "learning_rate": 9.030858575489628e-05,
+      "loss": 1.1303,
+      "step": 3358
+    },
+    {
+      "epoch": 0.20408287259250257,
+      "grad_norm": 0.14956440031528473,
+      "learning_rate": 9.030292098210302e-05,
+      "loss": 1.0915,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2041436296251291,
+      "grad_norm": 0.2575131356716156,
+      "learning_rate": 9.029725473199839e-05,
+      "loss": 1.1665,
+      "step": 3360
+    },
+    {
+      "epoch": 0.20420438665775564,
+      "grad_norm": 0.17843805253505707,
+      "learning_rate": 9.02915870047901e-05,
+      "loss": 1.0882,
+      "step": 3361
+    },
+    {
+      "epoch": 0.20426514369038215,
+      "grad_norm": 0.20931629836559296,
+      "learning_rate": 9.02859178006859e-05,
+      "loss": 1.1664,
+      "step": 3362
+    },
+    {
+      "epoch": 0.20432590072300869,
+      "grad_norm": 0.21923457086086273,
+      "learning_rate": 9.02802471198936e-05,
+      "loss": 1.0871,
+      "step": 3363
+    },
+    {
+      "epoch": 0.20438665775563522,
+      "grad_norm": 0.4388730823993683,
+      "learning_rate": 9.027457496262104e-05,
+      "loss": 1.1021,
+      "step": 3364
+    },
+    {
+      "epoch": 0.20444741478826173,
+      "grad_norm": 0.23835933208465576,
+      "learning_rate": 9.026890132907618e-05,
+      "loss": 1.1708,
+      "step": 3365
+    },
+    {
+      "epoch": 0.20450817182088826,
+      "grad_norm": 0.2694539427757263,
+      "learning_rate": 9.026322621946692e-05,
+      "loss": 1.0946,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2045689288535148,
+      "grad_norm": 1.1554162502288818,
+      "learning_rate": 9.025754963400134e-05,
+      "loss": 1.1365,
+      "step": 3367
+    },
+    {
+      "epoch": 0.20462968588614133,
+      "grad_norm": 0.1602601557970047,
+      "learning_rate": 9.02518715728875e-05,
+      "loss": 1.11,
+      "step": 3368
+    },
+    {
+      "epoch": 0.20469044291876784,
+      "grad_norm": 0.2515504062175751,
+      "learning_rate": 9.024619203633353e-05,
+      "loss": 1.1114,
+      "step": 3369
+    },
+    {
+      "epoch": 0.20475119995139437,
+      "grad_norm": 0.2663159966468811,
+      "learning_rate": 9.02405110245476e-05,
+      "loss": 1.144,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2048119569840209,
+      "grad_norm": 0.1740141659975052,
+      "learning_rate": 9.023482853773797e-05,
+      "loss": 1.0782,
+      "step": 3371
+    },
+    {
+      "epoch": 0.20487271401664742,
+      "grad_norm": 1.3628073930740356,
+      "learning_rate": 9.022914457611293e-05,
+      "loss": 1.0954,
+      "step": 3372
+    },
+    {
+      "epoch": 0.20493347104927395,
+      "grad_norm": 1.5277669429779053,
+      "learning_rate": 9.02234591398808e-05,
+      "loss": 1.1113,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2049942280819005,
+      "grad_norm": 0.22233106195926666,
+      "learning_rate": 9.021777222925001e-05,
+      "loss": 1.1034,
+      "step": 3374
+    },
+    {
+      "epoch": 0.205054985114527,
+      "grad_norm": 0.18932431936264038,
+      "learning_rate": 9.021208384442901e-05,
+      "loss": 1.1093,
+      "step": 3375
+    },
+    {
+      "epoch": 0.20511574214715353,
+      "grad_norm": 0.28633913397789,
+      "learning_rate": 9.020639398562629e-05,
+      "loss": 1.138,
+      "step": 3376
+    },
+    {
+      "epoch": 0.20517649917978006,
+      "grad_norm": 0.2889212965965271,
+      "learning_rate": 9.020070265305043e-05,
+      "loss": 1.0839,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2052372562124066,
+      "grad_norm": 0.2777705192565918,
+      "learning_rate": 9.019500984691004e-05,
+      "loss": 1.3459,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2052980132450331,
+      "grad_norm": 0.26010599732398987,
+      "learning_rate": 9.018931556741381e-05,
+      "loss": 1.1585,
+      "step": 3379
+    },
+    {
+      "epoch": 0.20535877027765964,
+      "grad_norm": 0.2716197669506073,
+      "learning_rate": 9.018361981477045e-05,
+      "loss": 1.0921,
+      "step": 3380
+    },
+    {
+      "epoch": 0.20541952731028618,
+      "grad_norm": 0.14544996619224548,
+      "learning_rate": 9.017792258918873e-05,
+      "loss": 1.1272,
+      "step": 3381
+    },
+    {
+      "epoch": 0.20548028434291268,
+      "grad_norm": 0.1978648602962494,
+      "learning_rate": 9.01722238908775e-05,
+      "loss": 1.1856,
+      "step": 3382
+    },
+    {
+      "epoch": 0.20554104137553922,
+      "grad_norm": 0.18912170827388763,
+      "learning_rate": 9.016652372004562e-05,
+      "loss": 1.0779,
+      "step": 3383
+    },
+    {
+      "epoch": 0.20560179840816575,
+      "grad_norm": 0.20028752088546753,
+      "learning_rate": 9.016082207690207e-05,
+      "loss": 1.118,
+      "step": 3384
+    },
+    {
+      "epoch": 0.20566255544079226,
+      "grad_norm": 2.3095862865448,
+      "learning_rate": 9.01551189616558e-05,
+      "loss": 1.1394,
+      "step": 3385
+    },
+    {
+      "epoch": 0.2057233124734188,
+      "grad_norm": 0.21716797351837158,
+      "learning_rate": 9.01494143745159e-05,
+      "loss": 1.1877,
+      "step": 3386
+    },
+    {
+      "epoch": 0.20578406950604533,
+      "grad_norm": 0.15557648241519928,
+      "learning_rate": 9.014370831569144e-05,
+      "loss": 1.1201,
+      "step": 3387
+    },
+    {
+      "epoch": 0.20584482653867184,
+      "grad_norm": 0.1555710881948471,
+      "learning_rate": 9.01380007853916e-05,
+      "loss": 1.0736,
+      "step": 3388
+    },
+    {
+      "epoch": 0.20590558357129837,
+      "grad_norm": 0.18824075162410736,
+      "learning_rate": 9.01322917838256e-05,
+      "loss": 1.0712,
+      "step": 3389
+    },
+    {
+      "epoch": 0.2059663406039249,
+      "grad_norm": 0.26053574681282043,
+      "learning_rate": 9.012658131120266e-05,
+      "loss": 1.0993,
+      "step": 3390
+    },
+    {
+      "epoch": 0.20602709763655144,
+      "grad_norm": 0.210182785987854,
+      "learning_rate": 9.012086936773214e-05,
+      "loss": 1.0759,
+      "step": 3391
+    },
+    {
+      "epoch": 0.20608785466917795,
+      "grad_norm": 0.17331752181053162,
+      "learning_rate": 9.011515595362342e-05,
+      "loss": 1.0844,
+      "step": 3392
+    },
+    {
+      "epoch": 0.20614861170180449,
+      "grad_norm": 0.14416228234767914,
+      "learning_rate": 9.010944106908586e-05,
+      "loss": 1.0983,
+      "step": 3393
+    },
+    {
+      "epoch": 0.20620936873443102,
+      "grad_norm": 0.3085445463657379,
+      "learning_rate": 9.010372471432902e-05,
+      "loss": 1.0516,
+      "step": 3394
+    },
+    {
+      "epoch": 0.20627012576705753,
+      "grad_norm": 0.1775248795747757,
+      "learning_rate": 9.009800688956237e-05,
+      "loss": 1.1444,
+      "step": 3395
+    },
+    {
+      "epoch": 0.20633088279968406,
+      "grad_norm": 0.24787196516990662,
+      "learning_rate": 9.009228759499555e-05,
+      "loss": 1.0822,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2063916398323106,
+      "grad_norm": 0.4156123399734497,
+      "learning_rate": 9.008656683083816e-05,
+      "loss": 1.0482,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2064523968649371,
+      "grad_norm": 0.3085361123085022,
+      "learning_rate": 9.008084459729993e-05,
+      "loss": 1.1109,
+      "step": 3398
+    },
+    {
+      "epoch": 0.20651315389756364,
+      "grad_norm": 0.22662469744682312,
+      "learning_rate": 9.007512089459058e-05,
+      "loss": 1.1664,
+      "step": 3399
+    },
+    {
+      "epoch": 0.20657391093019017,
+      "grad_norm": 0.24231384694576263,
+      "learning_rate": 9.006939572291994e-05,
+      "loss": 1.0466,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2066346679628167,
+      "grad_norm": 0.1650387942790985,
+      "learning_rate": 9.006366908249784e-05,
+      "loss": 1.1078,
+      "step": 3401
+    },
+    {
+      "epoch": 0.20669542499544322,
+      "grad_norm": 0.3309670388698578,
+      "learning_rate": 9.005794097353422e-05,
+      "loss": 1.0796,
+      "step": 3402
+    },
+    {
+      "epoch": 0.20675618202806975,
+      "grad_norm": 0.2354893833398819,
+      "learning_rate": 9.005221139623903e-05,
+      "loss": 1.1447,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2068169390606963,
+      "grad_norm": 0.25864624977111816,
+      "learning_rate": 9.004648035082228e-05,
+      "loss": 1.1428,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2068776960933228,
+      "grad_norm": 0.2769456207752228,
+      "learning_rate": 9.004074783749405e-05,
+      "loss": 1.2156,
+      "step": 3405
+    },
+    {
+      "epoch": 0.20693845312594933,
+      "grad_norm": 0.3436751365661621,
+      "learning_rate": 9.003501385646449e-05,
+      "loss": 1.1811,
+      "step": 3406
+    },
+    {
+      "epoch": 0.20699921015857586,
+      "grad_norm": 0.18414154648780823,
+      "learning_rate": 9.002927840794374e-05,
+      "loss": 1.0763,
+      "step": 3407
+    },
+    {
+      "epoch": 0.20705996719120237,
+      "grad_norm": 0.2883095443248749,
+      "learning_rate": 9.002354149214205e-05,
+      "loss": 1.1309,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2071207242238289,
+      "grad_norm": 0.8748995661735535,
+      "learning_rate": 9.001780310926972e-05,
+      "loss": 1.1212,
+      "step": 3409
+    },
+    {
+      "epoch": 0.20718148125645544,
+      "grad_norm": 0.16728021204471588,
+      "learning_rate": 9.001206325953709e-05,
+      "loss": 1.0571,
+      "step": 3410
+    },
+    {
+      "epoch": 0.20724223828908195,
+      "grad_norm": 0.16911764442920685,
+      "learning_rate": 9.000632194315453e-05,
+      "loss": 1.1535,
+      "step": 3411
+    },
+    {
+      "epoch": 0.20730299532170848,
+      "grad_norm": 0.23415647447109222,
+      "learning_rate": 9.00005791603325e-05,
+      "loss": 1.1268,
+      "step": 3412
+    },
+    {
+      "epoch": 0.20736375235433502,
+      "grad_norm": 0.1691472828388214,
+      "learning_rate": 8.999483491128153e-05,
+      "loss": 1.0652,
+      "step": 3413
+    },
+    {
+      "epoch": 0.20742450938696155,
+      "grad_norm": 0.1758832484483719,
+      "learning_rate": 8.998908919621214e-05,
+      "loss": 1.1133,
+      "step": 3414
+    },
+    {
+      "epoch": 0.20748526641958806,
+      "grad_norm": 0.301696240901947,
+      "learning_rate": 8.998334201533495e-05,
+      "loss": 1.1943,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2075460234522146,
+      "grad_norm": 0.34288668632507324,
+      "learning_rate": 8.997759336886066e-05,
+      "loss": 1.188,
+      "step": 3416
+    },
+    {
+      "epoch": 0.20760678048484113,
+      "grad_norm": 0.6763338446617126,
+      "learning_rate": 8.997184325699994e-05,
+      "loss": 1.0859,
+      "step": 3417
+    },
+    {
+      "epoch": 0.20766753751746764,
+      "grad_norm": 0.39041468501091003,
+      "learning_rate": 8.996609167996357e-05,
+      "loss": 1.1999,
+      "step": 3418
+    },
+    {
+      "epoch": 0.20772829455009417,
+      "grad_norm": 0.14516863226890564,
+      "learning_rate": 8.99603386379624e-05,
+      "loss": 1.0651,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2077890515827207,
+      "grad_norm": 12.497024536132812,
+      "learning_rate": 8.995458413120727e-05,
+      "loss": 1.1849,
+      "step": 3420
+    },
+    {
+      "epoch": 0.20784980861534721,
+      "grad_norm": 0.1711440533399582,
+      "learning_rate": 8.994882815990917e-05,
+      "loss": 1.1478,
+      "step": 3421
+    },
+    {
+      "epoch": 0.20791056564797375,
+      "grad_norm": 0.19265122711658478,
+      "learning_rate": 8.994307072427902e-05,
+      "loss": 1.0803,
+      "step": 3422
+    },
+    {
+      "epoch": 0.20797132268060028,
+      "grad_norm": 0.22108204662799835,
+      "learning_rate": 8.993731182452791e-05,
+      "loss": 1.3449,
+      "step": 3423
+    },
+    {
+      "epoch": 0.20803207971322682,
+      "grad_norm": 0.1360526978969574,
+      "learning_rate": 8.993155146086691e-05,
+      "loss": 1.0814,
+      "step": 3424
+    },
+    {
+      "epoch": 0.20809283674585333,
+      "grad_norm": 0.18838582932949066,
+      "learning_rate": 8.992578963350717e-05,
+      "loss": 1.156,
+      "step": 3425
+    },
+    {
+      "epoch": 0.20815359377847986,
+      "grad_norm": 0.4641799330711365,
+      "learning_rate": 8.99200263426599e-05,
+      "loss": 1.167,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2082143508111064,
+      "grad_norm": 0.1729332059621811,
+      "learning_rate": 8.991426158853635e-05,
+      "loss": 1.0738,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2082751078437329,
+      "grad_norm": 0.1480157971382141,
+      "learning_rate": 8.990849537134782e-05,
+      "loss": 1.1109,
+      "step": 3428
+    },
+    {
+      "epoch": 0.20833586487635944,
+      "grad_norm": 0.20335061848163605,
+      "learning_rate": 8.990272769130567e-05,
+      "loss": 1.2222,
+      "step": 3429
+    },
+    {
+      "epoch": 0.20839662190898597,
+      "grad_norm": 0.18276669085025787,
+      "learning_rate": 8.989695854862132e-05,
+      "loss": 1.0947,
+      "step": 3430
+    },
+    {
+      "epoch": 0.20845737894161248,
+      "grad_norm": 3.883676290512085,
+      "learning_rate": 8.989118794350625e-05,
+      "loss": 1.2349,
+      "step": 3431
+    },
+    {
+      "epoch": 0.20851813597423902,
+      "grad_norm": 0.2777820825576782,
+      "learning_rate": 8.9885415876172e-05,
+      "loss": 1.1044,
+      "step": 3432
+    },
+    {
+      "epoch": 0.20857889300686555,
+      "grad_norm": 0.1472485512495041,
+      "learning_rate": 8.987964234683009e-05,
+      "loss": 1.0502,
+      "step": 3433
+    },
+    {
+      "epoch": 0.20863965003949206,
+      "grad_norm": 0.16455063223838806,
+      "learning_rate": 8.98738673556922e-05,
+      "loss": 1.102,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2087004070721186,
+      "grad_norm": 0.2353052943944931,
+      "learning_rate": 8.986809090296998e-05,
+      "loss": 1.1555,
+      "step": 3435
+    },
+    {
+      "epoch": 0.20876116410474513,
+      "grad_norm": 0.14562679827213287,
+      "learning_rate": 8.98623129888752e-05,
+      "loss": 1.0827,
+      "step": 3436
+    },
+    {
+      "epoch": 0.20882192113737166,
+      "grad_norm": 0.2758413851261139,
+      "learning_rate": 8.985653361361961e-05,
+      "loss": 1.0943,
+      "step": 3437
+    },
+    {
+      "epoch": 0.20888267816999817,
+      "grad_norm": 0.14613857865333557,
+      "learning_rate": 8.98507527774151e-05,
+      "loss": 1.279,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2089434352026247,
+      "grad_norm": 0.11982880532741547,
+      "learning_rate": 8.98449704804735e-05,
+      "loss": 1.0714,
+      "step": 3439
+    },
+    {
+      "epoch": 0.20900419223525124,
+      "grad_norm": 0.9307172894477844,
+      "learning_rate": 8.983918672300687e-05,
+      "loss": 1.0974,
+      "step": 3440
+    },
+    {
+      "epoch": 0.20906494926787775,
+      "grad_norm": 0.23882558941841125,
+      "learning_rate": 8.98334015052271e-05,
+      "loss": 1.2063,
+      "step": 3441
+    },
+    {
+      "epoch": 0.20912570630050428,
+      "grad_norm": 0.14072297513484955,
+      "learning_rate": 8.982761482734632e-05,
+      "loss": 1.0889,
+      "step": 3442
+    },
+    {
+      "epoch": 0.20918646333313082,
+      "grad_norm": 0.15262438356876373,
+      "learning_rate": 8.982182668957661e-05,
+      "loss": 1.082,
+      "step": 3443
+    },
+    {
+      "epoch": 0.20924722036575732,
+      "grad_norm": 0.2728678584098816,
+      "learning_rate": 8.981603709213014e-05,
+      "loss": 1.1356,
+      "step": 3444
+    },
+    {
+      "epoch": 0.20930797739838386,
+      "grad_norm": 0.2701585292816162,
+      "learning_rate": 8.981024603521914e-05,
+      "loss": 1.2097,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2093687344310104,
+      "grad_norm": 0.142396479845047,
+      "learning_rate": 8.980445351905588e-05,
+      "loss": 1.071,
+      "step": 3446
+    },
+    {
+      "epoch": 0.20942949146363693,
+      "grad_norm": 0.13167715072631836,
+      "learning_rate": 8.979865954385269e-05,
+      "loss": 1.2422,
+      "step": 3447
+    },
+    {
+      "epoch": 0.20949024849626344,
+      "grad_norm": 0.3174823224544525,
+      "learning_rate": 8.979286410982194e-05,
+      "loss": 1.1203,
+      "step": 3448
+    },
+    {
+      "epoch": 0.20955100552888997,
+      "grad_norm": 0.20211239159107208,
+      "learning_rate": 8.978706721717605e-05,
+      "loss": 1.1399,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2096117625615165,
+      "grad_norm": 0.16633735597133636,
+      "learning_rate": 8.978126886612754e-05,
+      "loss": 1.1647,
+      "step": 3450
+    },
+    {
+      "epoch": 0.20967251959414301,
+      "grad_norm": 0.24114300310611725,
+      "learning_rate": 8.977546905688891e-05,
+      "loss": 1.085,
+      "step": 3451
+    },
+    {
+      "epoch": 0.20973327662676955,
+      "grad_norm": 0.3945278823375702,
+      "learning_rate": 8.976966778967279e-05,
+      "loss": 1.0882,
+      "step": 3452
+    },
+    {
+      "epoch": 0.20979403365939608,
+      "grad_norm": 0.33712682127952576,
+      "learning_rate": 8.976386506469178e-05,
+      "loss": 1.2284,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2098547906920226,
+      "grad_norm": 0.2718818187713623,
+      "learning_rate": 8.975806088215864e-05,
+      "loss": 1.1495,
+      "step": 3454
+    },
+    {
+      "epoch": 0.20991554772464913,
+      "grad_norm": 0.15609952807426453,
+      "learning_rate": 8.975225524228609e-05,
+      "loss": 1.1036,
+      "step": 3455
+    },
+    {
+      "epoch": 0.20997630475727566,
+      "grad_norm": 0.1752764880657196,
+      "learning_rate": 8.974644814528694e-05,
+      "loss": 1.1146,
+      "step": 3456
+    },
+    {
+      "epoch": 0.21003706178990217,
+      "grad_norm": 0.21618975698947906,
+      "learning_rate": 8.974063959137405e-05,
+      "loss": 1.065,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2100978188225287,
+      "grad_norm": 0.20082253217697144,
+      "learning_rate": 8.973482958076032e-05,
+      "loss": 1.0912,
+      "step": 3458
+    },
+    {
+      "epoch": 0.21015857585515524,
+      "grad_norm": 0.13558422029018402,
+      "learning_rate": 8.972901811365875e-05,
+      "loss": 1.083,
+      "step": 3459
+    },
+    {
+      "epoch": 0.21021933288778177,
+      "grad_norm": 17.55464744567871,
+      "learning_rate": 8.972320519028232e-05,
+      "loss": 1.1911,
+      "step": 3460
+    },
+    {
+      "epoch": 0.21028008992040828,
+      "grad_norm": 0.1939544528722763,
+      "learning_rate": 8.971739081084414e-05,
+      "loss": 1.0666,
+      "step": 3461
+    },
+    {
+      "epoch": 0.21034084695303482,
+      "grad_norm": 0.2793985605239868,
+      "learning_rate": 8.971157497555733e-05,
+      "loss": 1.2245,
+      "step": 3462
+    },
+    {
+      "epoch": 0.21040160398566135,
+      "grad_norm": 0.6017389893531799,
+      "learning_rate": 8.970575768463505e-05,
+      "loss": 1.278,
+      "step": 3463
+    },
+    {
+      "epoch": 0.21046236101828786,
+      "grad_norm": 12.897541999816895,
+      "learning_rate": 8.969993893829054e-05,
+      "loss": 1.1187,
+      "step": 3464
+    },
+    {
+      "epoch": 0.2105231180509144,
+      "grad_norm": 0.2867054343223572,
+      "learning_rate": 8.969411873673712e-05,
+      "loss": 1.1324,
+      "step": 3465
+    },
+    {
+      "epoch": 0.21058387508354093,
+      "grad_norm": 0.1979219615459442,
+      "learning_rate": 8.968829708018809e-05,
+      "loss": 1.1689,
+      "step": 3466
+    },
+    {
+      "epoch": 0.21064463211616744,
+      "grad_norm": 0.3768339157104492,
+      "learning_rate": 8.968247396885685e-05,
+      "loss": 1.0789,
+      "step": 3467
+    },
+    {
+      "epoch": 0.21070538914879397,
+      "grad_norm": 0.17232266068458557,
+      "learning_rate": 8.967664940295685e-05,
+      "loss": 1.1094,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2107661461814205,
+      "grad_norm": 0.17694714665412903,
+      "learning_rate": 8.967082338270163e-05,
+      "loss": 1.1441,
+      "step": 3469
+    },
+    {
+      "epoch": 0.210826903214047,
+      "grad_norm": 0.35034406185150146,
+      "learning_rate": 8.966499590830468e-05,
+      "loss": 1.1235,
+      "step": 3470
+    },
+    {
+      "epoch": 0.21088766024667355,
+      "grad_norm": 0.40152639150619507,
+      "learning_rate": 8.965916697997966e-05,
+      "loss": 1.1417,
+      "step": 3471
+    },
+    {
+      "epoch": 0.21094841727930008,
+      "grad_norm": 0.17416897416114807,
+      "learning_rate": 8.96533365979402e-05,
+      "loss": 1.0615,
+      "step": 3472
+    },
+    {
+      "epoch": 0.21100917431192662,
+      "grad_norm": 0.3140276372432709,
+      "learning_rate": 8.964750476240003e-05,
+      "loss": 1.1857,
+      "step": 3473
+    },
+    {
+      "epoch": 0.21106993134455312,
+      "grad_norm": 0.30841904878616333,
+      "learning_rate": 8.964167147357289e-05,
+      "loss": 1.0743,
+      "step": 3474
+    },
+    {
+      "epoch": 0.21113068837717966,
+      "grad_norm": 0.1996213048696518,
+      "learning_rate": 8.963583673167263e-05,
+      "loss": 1.1743,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2111914454098062,
+      "grad_norm": 0.2031165212392807,
+      "learning_rate": 8.96300005369131e-05,
+      "loss": 1.1261,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2112522024424327,
+      "grad_norm": 0.8546181321144104,
+      "learning_rate": 8.962416288950827e-05,
+      "loss": 1.3121,
+      "step": 3477
+    },
+    {
+      "epoch": 0.21131295947505924,
+      "grad_norm": 0.22097034752368927,
+      "learning_rate": 8.961832378967207e-05,
+      "loss": 1.0704,
+      "step": 3478
+    },
+    {
+      "epoch": 0.21137371650768577,
+      "grad_norm": 0.1509115993976593,
+      "learning_rate": 8.961248323761855e-05,
+      "loss": 1.0723,
+      "step": 3479
+    },
+    {
+      "epoch": 0.21143447354031228,
+      "grad_norm": 0.24100449681282043,
+      "learning_rate": 8.960664123356182e-05,
+      "loss": 1.1231,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2114952305729388,
+      "grad_norm": 0.18507619202136993,
+      "learning_rate": 8.960079777771598e-05,
+      "loss": 1.1152,
+      "step": 3481
+    },
+    {
+      "epoch": 0.21155598760556535,
+      "grad_norm": 0.4750421345233917,
+      "learning_rate": 8.959495287029525e-05,
+      "loss": 1.0421,
+      "step": 3482
+    },
+    {
+      "epoch": 0.21161674463819188,
+      "grad_norm": 0.18387670814990997,
+      "learning_rate": 8.958910651151387e-05,
+      "loss": 1.0501,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2116775016708184,
+      "grad_norm": 6.725606441497803,
+      "learning_rate": 8.958325870158613e-05,
+      "loss": 1.0825,
+      "step": 3484
+    },
+    {
+      "epoch": 0.21173825870344493,
+      "grad_norm": 0.3028864562511444,
+      "learning_rate": 8.95774094407264e-05,
+      "loss": 1.0771,
+      "step": 3485
+    },
+    {
+      "epoch": 0.21179901573607146,
+      "grad_norm": 0.15914765000343323,
+      "learning_rate": 8.957155872914907e-05,
+      "loss": 1.0584,
+      "step": 3486
+    },
+    {
+      "epoch": 0.21185977276869797,
+      "grad_norm": 0.2881225645542145,
+      "learning_rate": 8.956570656706862e-05,
+      "loss": 1.1153,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2119205298013245,
+      "grad_norm": 0.3317762315273285,
+      "learning_rate": 8.955985295469953e-05,
+      "loss": 1.0941,
+      "step": 3488
+    },
+    {
+      "epoch": 0.21198128683395104,
+      "grad_norm": 0.33383673429489136,
+      "learning_rate": 8.95539978922564e-05,
+      "loss": 1.0717,
+      "step": 3489
+    },
+    {
+      "epoch": 0.21204204386657755,
+      "grad_norm": 0.16560310125350952,
+      "learning_rate": 8.954814137995382e-05,
+      "loss": 1.1082,
+      "step": 3490
+    },
+    {
+      "epoch": 0.21210280089920408,
+      "grad_norm": 0.2758764624595642,
+      "learning_rate": 8.954228341800648e-05,
+      "loss": 1.069,
+      "step": 3491
+    },
+    {
+      "epoch": 0.21216355793183062,
+      "grad_norm": 0.3274565041065216,
+      "learning_rate": 8.953642400662909e-05,
+      "loss": 1.0943,
+      "step": 3492
+    },
+    {
+      "epoch": 0.21222431496445712,
+      "grad_norm": 0.2620927691459656,
+      "learning_rate": 8.953056314603644e-05,
+      "loss": 1.1549,
+      "step": 3493
+    },
+    {
+      "epoch": 0.21228507199708366,
+      "grad_norm": 0.14102204144001007,
+      "learning_rate": 8.952470083644336e-05,
+      "loss": 1.1133,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2123458290297102,
+      "grad_norm": 0.16233918070793152,
+      "learning_rate": 8.951883707806472e-05,
+      "loss": 1.0934,
+      "step": 3495
+    },
+    {
+      "epoch": 0.21240658606233673,
+      "grad_norm": 0.22172996401786804,
+      "learning_rate": 8.951297187111547e-05,
+      "loss": 1.0896,
+      "step": 3496
+    },
+    {
+      "epoch": 0.21246734309496323,
+      "grad_norm": 2.096376657485962,
+      "learning_rate": 8.95071052158106e-05,
+      "loss": 1.1297,
+      "step": 3497
+    },
+    {
+      "epoch": 0.21252810012758977,
+      "grad_norm": 0.23521722853183746,
+      "learning_rate": 8.950123711236517e-05,
+      "loss": 1.0986,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2125888571602163,
+      "grad_norm": 0.206913024187088,
+      "learning_rate": 8.949536756099424e-05,
+      "loss": 1.1535,
+      "step": 3499
+    },
+    {
+      "epoch": 0.2126496141928428,
+      "grad_norm": 0.5676190257072449,
+      "learning_rate": 8.948949656191299e-05,
+      "loss": 1.1095,
+      "step": 3500
+    },
+    {
+      "epoch": 0.21271037122546935,
+      "grad_norm": 0.23115776479244232,
+      "learning_rate": 8.94836241153366e-05,
+      "loss": 1.0747,
+      "step": 3501
+    },
+    {
+      "epoch": 0.21277112825809588,
+      "grad_norm": 0.14194437861442566,
+      "learning_rate": 8.947775022148033e-05,
+      "loss": 1.0772,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2128318852907224,
+      "grad_norm": 0.2450140416622162,
+      "learning_rate": 8.94718748805595e-05,
+      "loss": 1.2935,
+      "step": 3503
+    },
+    {
+      "epoch": 0.21289264232334892,
+      "grad_norm": 0.2724868357181549,
+      "learning_rate": 8.946599809278949e-05,
+      "loss": 1.1514,
+      "step": 3504
+    },
+    {
+      "epoch": 0.21295339935597546,
+      "grad_norm": 0.23426179587841034,
+      "learning_rate": 8.946011985838565e-05,
+      "loss": 1.3099,
+      "step": 3505
+    },
+    {
+      "epoch": 0.213014156388602,
+      "grad_norm": 0.3953523337841034,
+      "learning_rate": 8.945424017756352e-05,
+      "loss": 1.1211,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2130749134212285,
+      "grad_norm": 0.19598069787025452,
+      "learning_rate": 8.944835905053858e-05,
+      "loss": 1.0589,
+      "step": 3507
+    },
+    {
+      "epoch": 0.21313567045385504,
+      "grad_norm": 0.19311000406742096,
+      "learning_rate": 8.944247647752642e-05,
+      "loss": 1.1569,
+      "step": 3508
+    },
+    {
+      "epoch": 0.21319642748648157,
+      "grad_norm": 2.1531529426574707,
+      "learning_rate": 8.943659245874266e-05,
+      "loss": 1.2014,
+      "step": 3509
+    },
+    {
+      "epoch": 0.21325718451910808,
+      "grad_norm": 0.3676398694515228,
+      "learning_rate": 8.943070699440295e-05,
+      "loss": 1.3526,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2133179415517346,
+      "grad_norm": 0.354640394449234,
+      "learning_rate": 8.942482008472309e-05,
+      "loss": 1.1498,
+      "step": 3511
+    },
+    {
+      "epoch": 0.21337869858436115,
+      "grad_norm": 0.37891027331352234,
+      "learning_rate": 8.941893172991882e-05,
+      "loss": 1.073,
+      "step": 3512
+    },
+    {
+      "epoch": 0.21343945561698766,
+      "grad_norm": 0.418806254863739,
+      "learning_rate": 8.941304193020599e-05,
+      "loss": 1.327,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2135002126496142,
+      "grad_norm": 2.414013624191284,
+      "learning_rate": 8.940715068580047e-05,
+      "loss": 1.2974,
+      "step": 3514
+    },
+    {
+      "epoch": 0.21356096968224073,
+      "grad_norm": 0.2315111607313156,
+      "learning_rate": 8.940125799691826e-05,
+      "loss": 1.0818,
+      "step": 3515
+    },
+    {
+      "epoch": 0.21362172671486723,
+      "grad_norm": 0.25556647777557373,
+      "learning_rate": 8.93953638637753e-05,
+      "loss": 1.0616,
+      "step": 3516
+    },
+    {
+      "epoch": 0.21368248374749377,
+      "grad_norm": 0.17034971714019775,
+      "learning_rate": 8.938946828658766e-05,
+      "loss": 1.0675,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2137432407801203,
+      "grad_norm": 0.3051464557647705,
+      "learning_rate": 8.938357126557144e-05,
+      "loss": 1.265,
+      "step": 3518
+    },
+    {
+      "epoch": 0.21380399781274684,
+      "grad_norm": 0.29688090085983276,
+      "learning_rate": 8.937767280094281e-05,
+      "loss": 1.081,
+      "step": 3519
+    },
+    {
+      "epoch": 0.21386475484537334,
+      "grad_norm": 0.1980140060186386,
+      "learning_rate": 8.937177289291798e-05,
+      "loss": 1.0786,
+      "step": 3520
+    },
+    {
+      "epoch": 0.21392551187799988,
+      "grad_norm": 0.19571496546268463,
+      "learning_rate": 8.93658715417132e-05,
+      "loss": 1.0961,
+      "step": 3521
+    },
+    {
+      "epoch": 0.21398626891062641,
+      "grad_norm": 0.21203705668449402,
+      "learning_rate": 8.935996874754478e-05,
+      "loss": 1.2127,
+      "step": 3522
+    },
+    {
+      "epoch": 0.21404702594325292,
+      "grad_norm": 0.3411729037761688,
+      "learning_rate": 8.935406451062911e-05,
+      "loss": 1.2697,
+      "step": 3523
+    },
+    {
+      "epoch": 0.21410778297587946,
+      "grad_norm": 0.17787259817123413,
+      "learning_rate": 8.93481588311826e-05,
+      "loss": 1.103,
+      "step": 3524
+    },
+    {
+      "epoch": 0.214168540008506,
+      "grad_norm": 0.14897394180297852,
+      "learning_rate": 8.934225170942172e-05,
+      "loss": 1.079,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2142292970411325,
+      "grad_norm": 0.19840256869792938,
+      "learning_rate": 8.933634314556299e-05,
+      "loss": 1.1951,
+      "step": 3526
+    },
+    {
+      "epoch": 0.21429005407375903,
+      "grad_norm": 0.25057125091552734,
+      "learning_rate": 8.9330433139823e-05,
+      "loss": 1.0791,
+      "step": 3527
+    },
+    {
+      "epoch": 0.21435081110638557,
+      "grad_norm": 0.21097703278064728,
+      "learning_rate": 8.932452169241838e-05,
+      "loss": 1.1514,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2144115681390121,
+      "grad_norm": 0.42289677262306213,
+      "learning_rate": 8.931860880356583e-05,
+      "loss": 1.1388,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2144723251716386,
+      "grad_norm": 0.18627946078777313,
+      "learning_rate": 8.931269447348206e-05,
+      "loss": 1.1116,
+      "step": 3530
+    },
+    {
+      "epoch": 0.21453308220426515,
+      "grad_norm": 0.2027835249900818,
+      "learning_rate": 8.930677870238388e-05,
+      "loss": 1.126,
+      "step": 3531
+    },
+    {
+      "epoch": 0.21459383923689168,
+      "grad_norm": 0.5285890102386475,
+      "learning_rate": 8.930086149048813e-05,
+      "loss": 1.0493,
+      "step": 3532
+    },
+    {
+      "epoch": 0.2146545962695182,
+      "grad_norm": 0.22224541008472443,
+      "learning_rate": 8.92949428380117e-05,
+      "loss": 1.2482,
+      "step": 3533
+    },
+    {
+      "epoch": 0.21471535330214472,
+      "grad_norm": 0.22938747704029083,
+      "learning_rate": 8.928902274517154e-05,
+      "loss": 1.138,
+      "step": 3534
+    },
+    {
+      "epoch": 0.21477611033477126,
+      "grad_norm": 0.241376131772995,
+      "learning_rate": 8.928310121218469e-05,
+      "loss": 1.1248,
+      "step": 3535
+    },
+    {
+      "epoch": 0.21483686736739777,
+      "grad_norm": 0.21722251176834106,
+      "learning_rate": 8.927717823926813e-05,
+      "loss": 1.139,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2148976244000243,
+      "grad_norm": 0.6935981512069702,
+      "learning_rate": 8.927125382663903e-05,
+      "loss": 1.1276,
+      "step": 3537
+    },
+    {
+      "epoch": 0.21495838143265084,
+      "grad_norm": 0.16614855825901031,
+      "learning_rate": 8.926532797451452e-05,
+      "loss": 1.1664,
+      "step": 3538
+    },
+    {
+      "epoch": 0.21501913846527734,
+      "grad_norm": 0.21620385348796844,
+      "learning_rate": 8.925940068311183e-05,
+      "loss": 1.2537,
+      "step": 3539
+    },
+    {
+      "epoch": 0.21507989549790388,
+      "grad_norm": 0.21182186901569366,
+      "learning_rate": 8.925347195264823e-05,
+      "loss": 1.0652,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2151406525305304,
+      "grad_norm": 0.15097814798355103,
+      "learning_rate": 8.924754178334102e-05,
+      "loss": 1.0744,
+      "step": 3541
+    },
+    {
+      "epoch": 0.21520140956315695,
+      "grad_norm": 0.16458383202552795,
+      "learning_rate": 8.924161017540758e-05,
+      "loss": 1.093,
+      "step": 3542
+    },
+    {
+      "epoch": 0.21526216659578346,
+      "grad_norm": 0.5575851798057556,
+      "learning_rate": 8.923567712906534e-05,
+      "loss": 1.1397,
+      "step": 3543
+    },
+    {
+      "epoch": 0.21532292362841,
+      "grad_norm": 0.20459336042404175,
+      "learning_rate": 8.922974264453175e-05,
+      "loss": 1.1298,
+      "step": 3544
+    },
+    {
+      "epoch": 0.21538368066103653,
+      "grad_norm": 0.20945602655410767,
+      "learning_rate": 8.922380672202437e-05,
+      "loss": 1.2046,
+      "step": 3545
+    },
+    {
+      "epoch": 0.21544443769366303,
+      "grad_norm": 0.24499641358852386,
+      "learning_rate": 8.921786936176077e-05,
+      "loss": 1.1787,
+      "step": 3546
+    },
+    {
+      "epoch": 0.21550519472628957,
+      "grad_norm": 0.2540135085582733,
+      "learning_rate": 8.92119305639586e-05,
+      "loss": 1.1748,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2155659517589161,
+      "grad_norm": 0.20312140882015228,
+      "learning_rate": 8.920599032883554e-05,
+      "loss": 1.0987,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2156267087915426,
+      "grad_norm": 0.1747300773859024,
+      "learning_rate": 8.92000486566093e-05,
+      "loss": 1.0825,
+      "step": 3549
+    },
+    {
+      "epoch": 0.21568746582416914,
+      "grad_norm": 0.1898849755525589,
+      "learning_rate": 8.919410554749772e-05,
+      "loss": 1.1206,
+      "step": 3550
+    },
+    {
+      "epoch": 0.21574822285679568,
+      "grad_norm": 0.21884649991989136,
+      "learning_rate": 8.918816100171862e-05,
+      "loss": 1.1316,
+      "step": 3551
+    },
+    {
+      "epoch": 0.21580897988942221,
+      "grad_norm": 0.15853926539421082,
+      "learning_rate": 8.91822150194899e-05,
+      "loss": 1.1037,
+      "step": 3552
+    },
+    {
+      "epoch": 0.21586973692204872,
+      "grad_norm": 0.29001468420028687,
+      "learning_rate": 8.917626760102951e-05,
+      "loss": 1.1606,
+      "step": 3553
+    },
+    {
+      "epoch": 0.21593049395467526,
+      "grad_norm": 0.1479199230670929,
+      "learning_rate": 8.917031874655546e-05,
+      "loss": 1.0943,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2159912509873018,
+      "grad_norm": 0.16203229129314423,
+      "learning_rate": 8.916436845628582e-05,
+      "loss": 1.1757,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2160520080199283,
+      "grad_norm": 0.2718237340450287,
+      "learning_rate": 8.915841673043868e-05,
+      "loss": 1.1535,
+      "step": 3556
+    },
+    {
+      "epoch": 0.21611276505255483,
+      "grad_norm": 0.24396225810050964,
+      "learning_rate": 8.91524635692322e-05,
+      "loss": 1.1568,
+      "step": 3557
+    },
+    {
+      "epoch": 0.21617352208518137,
+      "grad_norm": 0.2197808027267456,
+      "learning_rate": 8.91465089728846e-05,
+      "loss": 1.14,
+      "step": 3558
+    },
+    {
+      "epoch": 0.21623427911780788,
+      "grad_norm": 0.35085004568099976,
+      "learning_rate": 8.914055294161414e-05,
+      "loss": 1.1216,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2162950361504344,
+      "grad_norm": 0.2695600986480713,
+      "learning_rate": 8.913459547563917e-05,
+      "loss": 1.2923,
+      "step": 3560
+    },
+    {
+      "epoch": 0.21635579318306095,
+      "grad_norm": 0.2137691229581833,
+      "learning_rate": 8.9128636575178e-05,
+      "loss": 1.2772,
+      "step": 3561
+    },
+    {
+      "epoch": 0.21641655021568745,
+      "grad_norm": 0.2682209312915802,
+      "learning_rate": 8.912267624044913e-05,
+      "loss": 1.0875,
+      "step": 3562
+    },
+    {
+      "epoch": 0.216477307248314,
+      "grad_norm": 2.038889169692993,
+      "learning_rate": 8.911671447167099e-05,
+      "loss": 1.2319,
+      "step": 3563
+    },
+    {
+      "epoch": 0.21653806428094052,
+      "grad_norm": 0.19789953529834747,
+      "learning_rate": 8.91107512690621e-05,
+      "loss": 1.0811,
+      "step": 3564
+    },
+    {
+      "epoch": 0.21659882131356706,
+      "grad_norm": 0.3618752360343933,
+      "learning_rate": 8.910478663284108e-05,
+      "loss": 1.1574,
+      "step": 3565
+    },
+    {
+      "epoch": 0.21665957834619357,
+      "grad_norm": 0.20511986315250397,
+      "learning_rate": 8.909882056322653e-05,
+      "loss": 1.1594,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2167203353788201,
+      "grad_norm": 0.4069550931453705,
+      "learning_rate": 8.909285306043717e-05,
+      "loss": 1.1093,
+      "step": 3567
+    },
+    {
+      "epoch": 0.21678109241144664,
+      "grad_norm": 1.0260331630706787,
+      "learning_rate": 8.908688412469173e-05,
+      "loss": 1.0382,
+      "step": 3568
+    },
+    {
+      "epoch": 0.21684184944407314,
+      "grad_norm": 0.17617836594581604,
+      "learning_rate": 8.9080913756209e-05,
+      "loss": 1.1001,
+      "step": 3569
+    },
+    {
+      "epoch": 0.21690260647669968,
+      "grad_norm": 0.2915533781051636,
+      "learning_rate": 8.90749419552078e-05,
+      "loss": 1.0576,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2169633635093262,
+      "grad_norm": 0.1760842353105545,
+      "learning_rate": 8.906896872190706e-05,
+      "loss": 1.1451,
+      "step": 3571
+    },
+    {
+      "epoch": 0.21702412054195272,
+      "grad_norm": 0.2882809042930603,
+      "learning_rate": 8.906299405652573e-05,
+      "loss": 1.1624,
+      "step": 3572
+    },
+    {
+      "epoch": 0.21708487757457925,
+      "grad_norm": 0.21022208034992218,
+      "learning_rate": 8.90570179592828e-05,
+      "loss": 1.0747,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2171456346072058,
+      "grad_norm": 0.1654147058725357,
+      "learning_rate": 8.905104043039731e-05,
+      "loss": 1.0847,
+      "step": 3574
+    },
+    {
+      "epoch": 0.2172063916398323,
+      "grad_norm": 0.17736396193504333,
+      "learning_rate": 8.90450614700884e-05,
+      "loss": 1.0945,
+      "step": 3575
+    },
+    {
+      "epoch": 0.21726714867245883,
+      "grad_norm": 0.1518404185771942,
+      "learning_rate": 8.903908107857521e-05,
+      "loss": 1.1247,
+      "step": 3576
+    },
+    {
+      "epoch": 0.21732790570508537,
+      "grad_norm": 0.2741888761520386,
+      "learning_rate": 8.903309925607696e-05,
+      "loss": 1.088,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2173886627377119,
+      "grad_norm": 0.7856914401054382,
+      "learning_rate": 8.90271160028129e-05,
+      "loss": 1.3599,
+      "step": 3578
+    },
+    {
+      "epoch": 0.2174494197703384,
+      "grad_norm": 0.5083143711090088,
+      "learning_rate": 8.902113131900238e-05,
+      "loss": 1.137,
+      "step": 3579
+    },
+    {
+      "epoch": 0.21751017680296494,
+      "grad_norm": 0.37476813793182373,
+      "learning_rate": 8.901514520486473e-05,
+      "loss": 1.0453,
+      "step": 3580
+    },
+    {
+      "epoch": 0.21757093383559148,
+      "grad_norm": 0.31943994760513306,
+      "learning_rate": 8.90091576606194e-05,
+      "loss": 1.2407,
+      "step": 3581
+    },
+    {
+      "epoch": 0.217631690868218,
+      "grad_norm": 0.26216375827789307,
+      "learning_rate": 8.900316868648586e-05,
+      "loss": 1.1617,
+      "step": 3582
+    },
+    {
+      "epoch": 0.21769244790084452,
+      "grad_norm": 0.2337801158428192,
+      "learning_rate": 8.899717828268361e-05,
+      "loss": 1.0981,
+      "step": 3583
+    },
+    {
+      "epoch": 0.21775320493347106,
+      "grad_norm": 0.24762462079524994,
+      "learning_rate": 8.899118644943227e-05,
+      "loss": 1.1134,
+      "step": 3584
+    },
+    {
+      "epoch": 0.21781396196609756,
+      "grad_norm": 0.35765013098716736,
+      "learning_rate": 8.898519318695146e-05,
+      "loss": 1.0458,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2178747189987241,
+      "grad_norm": 0.1970292031764984,
+      "learning_rate": 8.897919849546086e-05,
+      "loss": 1.0927,
+      "step": 3586
+    },
+    {
+      "epoch": 0.21793547603135063,
+      "grad_norm": 0.31617265939712524,
+      "learning_rate": 8.897320237518019e-05,
+      "loss": 1.0529,
+      "step": 3587
+    },
+    {
+      "epoch": 0.21799623306397717,
+      "grad_norm": 0.15339192748069763,
+      "learning_rate": 8.896720482632925e-05,
+      "loss": 1.0535,
+      "step": 3588
+    },
+    {
+      "epoch": 0.21805699009660368,
+      "grad_norm": 0.19203001260757446,
+      "learning_rate": 8.896120584912791e-05,
+      "loss": 1.0831,
+      "step": 3589
+    },
+    {
+      "epoch": 0.2181177471292302,
+      "grad_norm": 0.2581885755062103,
+      "learning_rate": 8.895520544379601e-05,
+      "loss": 1.1794,
+      "step": 3590
+    },
+    {
+      "epoch": 0.21817850416185675,
+      "grad_norm": 0.38130053877830505,
+      "learning_rate": 8.894920361055354e-05,
+      "loss": 1.112,
+      "step": 3591
+    },
+    {
+      "epoch": 0.21823926119448325,
+      "grad_norm": 0.18617768585681915,
+      "learning_rate": 8.894320034962047e-05,
+      "loss": 1.2227,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2183000182271098,
+      "grad_norm": 0.37853309512138367,
+      "learning_rate": 8.893719566121688e-05,
+      "loss": 1.1876,
+      "step": 3593
+    },
+    {
+      "epoch": 0.21836077525973632,
+      "grad_norm": 0.3068560063838959,
+      "learning_rate": 8.893118954556286e-05,
+      "loss": 1.0832,
+      "step": 3594
+    },
+    {
+      "epoch": 0.21842153229236283,
+      "grad_norm": 0.1947592943906784,
+      "learning_rate": 8.892518200287854e-05,
+      "loss": 1.073,
+      "step": 3595
+    },
+    {
+      "epoch": 0.21848228932498936,
+      "grad_norm": 0.3098341226577759,
+      "learning_rate": 8.891917303338416e-05,
+      "loss": 1.0463,
+      "step": 3596
+    },
+    {
+      "epoch": 0.2185430463576159,
+      "grad_norm": 0.16980788111686707,
+      "learning_rate": 8.891316263729997e-05,
+      "loss": 1.1057,
+      "step": 3597
+    },
+    {
+      "epoch": 0.2186038033902424,
+      "grad_norm": 0.22771577537059784,
+      "learning_rate": 8.890715081484628e-05,
+      "loss": 1.1578,
+      "step": 3598
+    },
+    {
+      "epoch": 0.21866456042286894,
+      "grad_norm": 0.46696770191192627,
+      "learning_rate": 8.890113756624347e-05,
+      "loss": 1.1605,
+      "step": 3599
+    },
+    {
+      "epoch": 0.21872531745549548,
+      "grad_norm": 0.27803370356559753,
+      "learning_rate": 8.889512289171194e-05,
+      "loss": 1.1542,
+      "step": 3600
+    },
+    {
+      "epoch": 0.218786074488122,
+      "grad_norm": 0.21509814262390137,
+      "learning_rate": 8.888910679147215e-05,
+      "loss": 1.0754,
+      "step": 3601
+    },
+    {
+      "epoch": 0.21884683152074852,
+      "grad_norm": 0.181820347905159,
+      "learning_rate": 8.888308926574463e-05,
+      "loss": 1.2084,
+      "step": 3602
+    },
+    {
+      "epoch": 0.21890758855337505,
+      "grad_norm": 0.22318626940250397,
+      "learning_rate": 8.887707031474996e-05,
+      "loss": 1.1644,
+      "step": 3603
+    },
+    {
+      "epoch": 0.2189683455860016,
+      "grad_norm": 0.25211507081985474,
+      "learning_rate": 8.887104993870878e-05,
+      "loss": 1.1419,
+      "step": 3604
+    },
+    {
+      "epoch": 0.2190291026186281,
+      "grad_norm": 0.3897862732410431,
+      "learning_rate": 8.886502813784173e-05,
+      "loss": 1.1857,
+      "step": 3605
+    },
+    {
+      "epoch": 0.21908985965125463,
+      "grad_norm": 0.17802387475967407,
+      "learning_rate": 8.885900491236957e-05,
+      "loss": 1.1071,
+      "step": 3606
+    },
+    {
+      "epoch": 0.21915061668388117,
+      "grad_norm": 0.42088744044303894,
+      "learning_rate": 8.885298026251307e-05,
+      "loss": 1.1424,
+      "step": 3607
+    },
+    {
+      "epoch": 0.21921137371650767,
+      "grad_norm": 0.3917873203754425,
+      "learning_rate": 8.884695418849307e-05,
+      "loss": 1.0925,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2192721307491342,
+      "grad_norm": 0.23588791489601135,
+      "learning_rate": 8.884092669053046e-05,
+      "loss": 1.1482,
+      "step": 3609
+    },
+    {
+      "epoch": 0.21933288778176074,
+      "grad_norm": 0.22042417526245117,
+      "learning_rate": 8.883489776884617e-05,
+      "loss": 1.1998,
+      "step": 3610
+    },
+    {
+      "epoch": 0.21939364481438728,
+      "grad_norm": 0.15223854780197144,
+      "learning_rate": 8.88288674236612e-05,
+      "loss": 1.1156,
+      "step": 3611
+    },
+    {
+      "epoch": 0.21945440184701379,
+      "grad_norm": 0.1689809411764145,
+      "learning_rate": 8.882283565519658e-05,
+      "loss": 1.0694,
+      "step": 3612
+    },
+    {
+      "epoch": 0.21951515887964032,
+      "grad_norm": 0.17674720287322998,
+      "learning_rate": 8.881680246367343e-05,
+      "loss": 1.1402,
+      "step": 3613
+    },
+    {
+      "epoch": 0.21957591591226686,
+      "grad_norm": 3.3900296688079834,
+      "learning_rate": 8.881076784931287e-05,
+      "loss": 1.0622,
+      "step": 3614
+    },
+    {
+      "epoch": 0.21963667294489336,
+      "grad_norm": 0.17992424964904785,
+      "learning_rate": 8.880473181233611e-05,
+      "loss": 1.0617,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2196974299775199,
+      "grad_norm": 0.2613798677921295,
+      "learning_rate": 8.879869435296442e-05,
+      "loss": 1.1863,
+      "step": 3616
+    },
+    {
+      "epoch": 0.21975818701014643,
+      "grad_norm": 0.3601166009902954,
+      "learning_rate": 8.879265547141909e-05,
+      "loss": 1.1004,
+      "step": 3617
+    },
+    {
+      "epoch": 0.21981894404277294,
+      "grad_norm": 0.1415201872587204,
+      "learning_rate": 8.878661516792144e-05,
+      "loss": 1.0756,
+      "step": 3618
+    },
+    {
+      "epoch": 0.21987970107539948,
+      "grad_norm": 1.9572412967681885,
+      "learning_rate": 8.878057344269295e-05,
+      "loss": 1.1808,
+      "step": 3619
+    },
+    {
+      "epoch": 0.219940458108026,
+      "grad_norm": 0.326018363237381,
+      "learning_rate": 8.877453029595504e-05,
+      "loss": 1.0405,
+      "step": 3620
+    },
+    {
+      "epoch": 0.22000121514065252,
+      "grad_norm": 0.22868525981903076,
+      "learning_rate": 8.876848572792922e-05,
+      "loss": 1.0882,
+      "step": 3621
+    },
+    {
+      "epoch": 0.22006197217327905,
+      "grad_norm": 0.2570403814315796,
+      "learning_rate": 8.876243973883706e-05,
+      "loss": 1.277,
+      "step": 3622
+    },
+    {
+      "epoch": 0.2201227292059056,
+      "grad_norm": 0.16984650492668152,
+      "learning_rate": 8.875639232890018e-05,
+      "loss": 1.1458,
+      "step": 3623
+    },
+    {
+      "epoch": 0.22018348623853212,
+      "grad_norm": 0.16814802587032318,
+      "learning_rate": 8.875034349834023e-05,
+      "loss": 1.0737,
+      "step": 3624
+    },
+    {
+      "epoch": 0.22024424327115863,
+      "grad_norm": 0.19076281785964966,
+      "learning_rate": 8.874429324737898e-05,
+      "loss": 1.1215,
+      "step": 3625
+    },
+    {
+      "epoch": 0.22030500030378516,
+      "grad_norm": 0.1807945966720581,
+      "learning_rate": 8.873824157623816e-05,
+      "loss": 1.2199,
+      "step": 3626
+    },
+    {
+      "epoch": 0.2203657573364117,
+      "grad_norm": 0.36208876967430115,
+      "learning_rate": 8.87321884851396e-05,
+      "loss": 1.2129,
+      "step": 3627
+    },
+    {
+      "epoch": 0.2204265143690382,
+      "grad_norm": 0.2234572470188141,
+      "learning_rate": 8.872613397430519e-05,
+      "loss": 1.1678,
+      "step": 3628
+    },
+    {
+      "epoch": 0.22048727140166474,
+      "grad_norm": 0.38514724373817444,
+      "learning_rate": 8.872007804395685e-05,
+      "loss": 1.1732,
+      "step": 3629
+    },
+    {
+      "epoch": 0.22054802843429128,
+      "grad_norm": 0.31021639704704285,
+      "learning_rate": 8.871402069431659e-05,
+      "loss": 1.3114,
+      "step": 3630
+    },
+    {
+      "epoch": 0.22060878546691778,
+      "grad_norm": 0.2765832543373108,
+      "learning_rate": 8.870796192560639e-05,
+      "loss": 1.0864,
+      "step": 3631
+    },
+    {
+      "epoch": 0.22066954249954432,
+      "grad_norm": 0.13771745562553406,
+      "learning_rate": 8.870190173804837e-05,
+      "loss": 1.0653,
+      "step": 3632
+    },
+    {
+      "epoch": 0.22073029953217085,
+      "grad_norm": 0.14678744971752167,
+      "learning_rate": 8.869584013186465e-05,
+      "loss": 1.0717,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2207910565647974,
+      "grad_norm": 0.28426042199134827,
+      "learning_rate": 8.868977710727744e-05,
+      "loss": 1.2044,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2208518135974239,
+      "grad_norm": 0.18499046564102173,
+      "learning_rate": 8.8683712664509e-05,
+      "loss": 1.1538,
+      "step": 3635
+    },
+    {
+      "epoch": 0.22091257063005043,
+      "grad_norm": 0.16694340109825134,
+      "learning_rate": 8.867764680378155e-05,
+      "loss": 1.0826,
+      "step": 3636
+    },
+    {
+      "epoch": 0.22097332766267697,
+      "grad_norm": 0.3892214000225067,
+      "learning_rate": 8.867157952531751e-05,
+      "loss": 1.3207,
+      "step": 3637
+    },
+    {
+      "epoch": 0.22103408469530347,
+      "grad_norm": 0.1352841705083847,
+      "learning_rate": 8.866551082933923e-05,
+      "loss": 1.0926,
+      "step": 3638
+    },
+    {
+      "epoch": 0.22109484172793,
+      "grad_norm": 0.27010589838027954,
+      "learning_rate": 8.865944071606918e-05,
+      "loss": 1.1157,
+      "step": 3639
+    },
+    {
+      "epoch": 0.22115559876055654,
+      "grad_norm": 0.20642493665218353,
+      "learning_rate": 8.865336918572986e-05,
+      "loss": 1.1863,
+      "step": 3640
+    },
+    {
+      "epoch": 0.22121635579318305,
+      "grad_norm": 0.17077583074569702,
+      "learning_rate": 8.864729623854382e-05,
+      "loss": 1.117,
+      "step": 3641
+    },
+    {
+      "epoch": 0.22127711282580959,
+      "grad_norm": 0.1897292584180832,
+      "learning_rate": 8.864122187473367e-05,
+      "loss": 1.1191,
+      "step": 3642
+    },
+    {
+      "epoch": 0.22133786985843612,
+      "grad_norm": 0.17921572923660278,
+      "learning_rate": 8.863514609452205e-05,
+      "loss": 1.1178,
+      "step": 3643
+    },
+    {
+      "epoch": 0.22139862689106263,
+      "grad_norm": 0.15557968616485596,
+      "learning_rate": 8.86290688981317e-05,
+      "loss": 1.1214,
+      "step": 3644
+    },
+    {
+      "epoch": 0.22145938392368916,
+      "grad_norm": 0.21869108080863953,
+      "learning_rate": 8.862299028578536e-05,
+      "loss": 1.0722,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2215201409563157,
+      "grad_norm": 1.952772855758667,
+      "learning_rate": 8.861691025770583e-05,
+      "loss": 1.1345,
+      "step": 3646
+    },
+    {
+      "epoch": 0.22158089798894223,
+      "grad_norm": 0.1759057492017746,
+      "learning_rate": 8.861082881411599e-05,
+      "loss": 1.1673,
+      "step": 3647
+    },
+    {
+      "epoch": 0.22164165502156874,
+      "grad_norm": 0.15730935335159302,
+      "learning_rate": 8.860474595523878e-05,
+      "loss": 1.1144,
+      "step": 3648
+    },
+    {
+      "epoch": 0.22170241205419527,
+      "grad_norm": 0.1722322702407837,
+      "learning_rate": 8.859866168129714e-05,
+      "loss": 1.0729,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2217631690868218,
+      "grad_norm": 0.3860315680503845,
+      "learning_rate": 8.859257599251407e-05,
+      "loss": 1.1058,
+      "step": 3650
+    },
+    {
+      "epoch": 0.22182392611944832,
+      "grad_norm": 2.915440320968628,
+      "learning_rate": 8.85864888891127e-05,
+      "loss": 1.1387,
+      "step": 3651
+    },
+    {
+      "epoch": 0.22188468315207485,
+      "grad_norm": 0.2498331218957901,
+      "learning_rate": 8.858040037131609e-05,
+      "loss": 1.0924,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2219454401847014,
+      "grad_norm": 0.16406171023845673,
+      "learning_rate": 8.857431043934747e-05,
+      "loss": 1.0686,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2220061972173279,
+      "grad_norm": 0.20061764121055603,
+      "learning_rate": 8.856821909343004e-05,
+      "loss": 1.1362,
+      "step": 3654
+    },
+    {
+      "epoch": 0.22206695424995443,
+      "grad_norm": 0.17562153935432434,
+      "learning_rate": 8.856212633378707e-05,
+      "loss": 1.0865,
+      "step": 3655
+    },
+    {
+      "epoch": 0.22212771128258096,
+      "grad_norm": 0.1663135141134262,
+      "learning_rate": 8.855603216064192e-05,
+      "loss": 1.1391,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2221884683152075,
+      "grad_norm": 0.5653345584869385,
+      "learning_rate": 8.854993657421796e-05,
+      "loss": 1.5107,
+      "step": 3657
+    },
+    {
+      "epoch": 0.222249225347834,
+      "grad_norm": 0.3727596402168274,
+      "learning_rate": 8.854383957473863e-05,
+      "loss": 1.1561,
+      "step": 3658
+    },
+    {
+      "epoch": 0.22230998238046054,
+      "grad_norm": 0.18117567896842957,
+      "learning_rate": 8.85377411624274e-05,
+      "loss": 1.106,
+      "step": 3659
+    },
+    {
+      "epoch": 0.22237073941308708,
+      "grad_norm": 0.23197732865810394,
+      "learning_rate": 8.85316413375078e-05,
+      "loss": 1.1812,
+      "step": 3660
+    },
+    {
+      "epoch": 0.22243149644571358,
+      "grad_norm": 0.1865433156490326,
+      "learning_rate": 8.852554010020346e-05,
+      "loss": 1.0467,
+      "step": 3661
+    },
+    {
+      "epoch": 0.22249225347834012,
+      "grad_norm": 0.24907264113426208,
+      "learning_rate": 8.8519437450738e-05,
+      "loss": 1.0755,
+      "step": 3662
+    },
+    {
+      "epoch": 0.22255301051096665,
+      "grad_norm": 0.44304803013801575,
+      "learning_rate": 8.85133333893351e-05,
+      "loss": 1.0524,
+      "step": 3663
+    },
+    {
+      "epoch": 0.22261376754359316,
+      "grad_norm": 0.24000073969364166,
+      "learning_rate": 8.850722791621853e-05,
+      "loss": 1.3545,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2226745245762197,
+      "grad_norm": 0.18592676520347595,
+      "learning_rate": 8.850112103161207e-05,
+      "loss": 1.1556,
+      "step": 3665
+    },
+    {
+      "epoch": 0.22273528160884623,
+      "grad_norm": 0.2611141502857208,
+      "learning_rate": 8.849501273573959e-05,
+      "loss": 1.2765,
+      "step": 3666
+    },
+    {
+      "epoch": 0.22279603864147274,
+      "grad_norm": 0.2191876918077469,
+      "learning_rate": 8.848890302882497e-05,
+      "loss": 1.0758,
+      "step": 3667
+    },
+    {
+      "epoch": 0.22285679567409927,
+      "grad_norm": 0.524749755859375,
+      "learning_rate": 8.848279191109217e-05,
+      "loss": 1.2897,
+      "step": 3668
+    },
+    {
+      "epoch": 0.2229175527067258,
+      "grad_norm": 0.20052844285964966,
+      "learning_rate": 8.84766793827652e-05,
+      "loss": 1.1826,
+      "step": 3669
+    },
+    {
+      "epoch": 0.22297830973935234,
+      "grad_norm": 0.3286508321762085,
+      "learning_rate": 8.847056544406811e-05,
+      "loss": 1.1109,
+      "step": 3670
+    },
+    {
+      "epoch": 0.22303906677197885,
+      "grad_norm": 0.13770145177841187,
+      "learning_rate": 8.8464450095225e-05,
+      "loss": 1.0742,
+      "step": 3671
+    },
+    {
+      "epoch": 0.22309982380460538,
+      "grad_norm": 0.19726188480854034,
+      "learning_rate": 8.845833333646005e-05,
+      "loss": 1.0962,
+      "step": 3672
+    },
+    {
+      "epoch": 0.22316058083723192,
+      "grad_norm": 0.21102310717105865,
+      "learning_rate": 8.845221516799745e-05,
+      "loss": 1.074,
+      "step": 3673
+    },
+    {
+      "epoch": 0.22322133786985843,
+      "grad_norm": 0.19460909068584442,
+      "learning_rate": 8.844609559006145e-05,
+      "loss": 1.1209,
+      "step": 3674
+    },
+    {
+      "epoch": 0.22328209490248496,
+      "grad_norm": 0.1998390257358551,
+      "learning_rate": 8.843997460287641e-05,
+      "loss": 1.0577,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2233428519351115,
+      "grad_norm": 0.19281549751758575,
+      "learning_rate": 8.843385220666667e-05,
+      "loss": 1.0754,
+      "step": 3676
+    },
+    {
+      "epoch": 0.223403608967738,
+      "grad_norm": 0.18167705833911896,
+      "learning_rate": 8.842772840165663e-05,
+      "loss": 1.0898,
+      "step": 3677
+    },
+    {
+      "epoch": 0.22346436600036454,
+      "grad_norm": 0.1706853061914444,
+      "learning_rate": 8.84216031880708e-05,
+      "loss": 1.116,
+      "step": 3678
+    },
+    {
+      "epoch": 0.22352512303299107,
+      "grad_norm": 0.19648264348506927,
+      "learning_rate": 8.841547656613367e-05,
+      "loss": 1.1903,
+      "step": 3679
+    },
+    {
+      "epoch": 0.22358588006561758,
+      "grad_norm": 0.19444352388381958,
+      "learning_rate": 8.840934853606981e-05,
+      "loss": 1.0587,
+      "step": 3680
+    },
+    {
+      "epoch": 0.22364663709824412,
+      "grad_norm": 0.2148924469947815,
+      "learning_rate": 8.840321909810386e-05,
+      "loss": 1.1402,
+      "step": 3681
+    },
+    {
+      "epoch": 0.22370739413087065,
+      "grad_norm": 0.35384291410446167,
+      "learning_rate": 8.839708825246049e-05,
+      "loss": 1.1711,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2237681511634972,
+      "grad_norm": 0.30873680114746094,
+      "learning_rate": 8.839095599936444e-05,
+      "loss": 1.3454,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2238289081961237,
+      "grad_norm": 0.3413650691509247,
+      "learning_rate": 8.838482233904046e-05,
+      "loss": 1.0719,
+      "step": 3684
+    },
+    {
+      "epoch": 0.22388966522875023,
+      "grad_norm": 0.4800724685192108,
+      "learning_rate": 8.837868727171342e-05,
+      "loss": 1.1946,
+      "step": 3685
+    },
+    {
+      "epoch": 0.22395042226137676,
+      "grad_norm": 0.1908242404460907,
+      "learning_rate": 8.837255079760815e-05,
+      "loss": 1.1637,
+      "step": 3686
+    },
+    {
+      "epoch": 0.22401117929400327,
+      "grad_norm": 0.4864121973514557,
+      "learning_rate": 8.836641291694964e-05,
+      "loss": 1.1185,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2240719363266298,
+      "grad_norm": 0.9131503701210022,
+      "learning_rate": 8.836027362996285e-05,
+      "loss": 1.1372,
+      "step": 3688
+    },
+    {
+      "epoch": 0.22413269335925634,
+      "grad_norm": 0.197661355137825,
+      "learning_rate": 8.83541329368728e-05,
+      "loss": 1.0941,
+      "step": 3689
+    },
+    {
+      "epoch": 0.22419345039188285,
+      "grad_norm": 0.20966699719429016,
+      "learning_rate": 8.834799083790459e-05,
+      "loss": 1.0455,
+      "step": 3690
+    },
+    {
+      "epoch": 0.22425420742450938,
+      "grad_norm": 0.21055859327316284,
+      "learning_rate": 8.834184733328337e-05,
+      "loss": 1.1266,
+      "step": 3691
+    },
+    {
+      "epoch": 0.22431496445713592,
+      "grad_norm": 0.4950893819332123,
+      "learning_rate": 8.833570242323432e-05,
+      "loss": 1.0672,
+      "step": 3692
+    },
+    {
+      "epoch": 0.22437572148976245,
+      "grad_norm": 0.28170105814933777,
+      "learning_rate": 8.83295561079827e-05,
+      "loss": 1.226,
+      "step": 3693
+    },
+    {
+      "epoch": 0.22443647852238896,
+      "grad_norm": 0.3324810266494751,
+      "learning_rate": 8.832340838775378e-05,
+      "loss": 1.1796,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2244972355550155,
+      "grad_norm": 0.1564098298549652,
+      "learning_rate": 8.831725926277293e-05,
+      "loss": 1.1227,
+      "step": 3695
+    },
+    {
+      "epoch": 0.22455799258764203,
+      "grad_norm": 0.18331372737884521,
+      "learning_rate": 8.831110873326552e-05,
+      "loss": 1.1592,
+      "step": 3696
+    },
+    {
+      "epoch": 0.22461874962026854,
+      "grad_norm": 0.22116729617118835,
+      "learning_rate": 8.830495679945703e-05,
+      "loss": 1.1262,
+      "step": 3697
+    },
+    {
+      "epoch": 0.22467950665289507,
+      "grad_norm": 0.16599038243293762,
+      "learning_rate": 8.829880346157294e-05,
+      "loss": 1.1148,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2247402636855216,
+      "grad_norm": 0.13323058187961578,
+      "learning_rate": 8.82926487198388e-05,
+      "loss": 1.0943,
+      "step": 3699
+    },
+    {
+      "epoch": 0.22480102071814811,
+      "grad_norm": 0.2711448669433594,
+      "learning_rate": 8.82864925744802e-05,
+      "loss": 1.0608,
+      "step": 3700
+    },
+    {
+      "epoch": 0.22486177775077465,
+      "grad_norm": 0.19824278354644775,
+      "learning_rate": 8.828033502572285e-05,
+      "loss": 1.0967,
+      "step": 3701
+    },
+    {
+      "epoch": 0.22492253478340118,
+      "grad_norm": 0.24330858886241913,
+      "learning_rate": 8.82741760737924e-05,
+      "loss": 1.1607,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2249832918160277,
+      "grad_norm": 0.18452030420303345,
+      "learning_rate": 8.826801571891463e-05,
+      "loss": 1.0607,
+      "step": 3703
+    },
+    {
+      "epoch": 0.22504404884865423,
+      "grad_norm": 0.2032107263803482,
+      "learning_rate": 8.826185396131536e-05,
+      "loss": 1.098,
+      "step": 3704
+    },
+    {
+      "epoch": 0.22510480588128076,
+      "grad_norm": 0.12496750056743622,
+      "learning_rate": 8.825569080122043e-05,
+      "loss": 1.0273,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2251655629139073,
+      "grad_norm": 0.19655679166316986,
+      "learning_rate": 8.824952623885576e-05,
+      "loss": 1.0811,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2252263199465338,
+      "grad_norm": 0.4549234211444855,
+      "learning_rate": 8.82433602744473e-05,
+      "loss": 1.0751,
+      "step": 3707
+    },
+    {
+      "epoch": 0.22528707697916034,
+      "grad_norm": 0.18192775547504425,
+      "learning_rate": 8.823719290822108e-05,
+      "loss": 1.1389,
+      "step": 3708
+    },
+    {
+      "epoch": 0.22534783401178687,
+      "grad_norm": 0.18826892971992493,
+      "learning_rate": 8.823102414040317e-05,
+      "loss": 1.1737,
+      "step": 3709
+    },
+    {
+      "epoch": 0.22540859104441338,
+      "grad_norm": 0.16363559663295746,
+      "learning_rate": 8.822485397121968e-05,
+      "loss": 1.124,
+      "step": 3710
+    },
+    {
+      "epoch": 0.22546934807703992,
+      "grad_norm": 0.18880683183670044,
+      "learning_rate": 8.821868240089676e-05,
+      "loss": 1.1377,
+      "step": 3711
+    },
+    {
+      "epoch": 0.22553010510966645,
+      "grad_norm": 0.17491993308067322,
+      "learning_rate": 8.821250942966069e-05,
+      "loss": 1.1446,
+      "step": 3712
+    },
+    {
+      "epoch": 0.22559086214229296,
+      "grad_norm": 0.18036296963691711,
+      "learning_rate": 8.820633505773766e-05,
+      "loss": 1.0953,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2256516191749195,
+      "grad_norm": 0.19422069191932678,
+      "learning_rate": 8.820015928535406e-05,
+      "loss": 1.1226,
+      "step": 3714
+    },
+    {
+      "epoch": 0.22571237620754603,
+      "grad_norm": 0.1913684606552124,
+      "learning_rate": 8.819398211273623e-05,
+      "loss": 1.098,
+      "step": 3715
+    },
+    {
+      "epoch": 0.22577313324017256,
+      "grad_norm": 0.18528629839420319,
+      "learning_rate": 8.81878035401106e-05,
+      "loss": 1.1686,
+      "step": 3716
+    },
+    {
+      "epoch": 0.22583389027279907,
+      "grad_norm": 0.13923168182373047,
+      "learning_rate": 8.818162356770366e-05,
+      "loss": 1.0793,
+      "step": 3717
+    },
+    {
+      "epoch": 0.2258946473054256,
+      "grad_norm": 0.3453364074230194,
+      "learning_rate": 8.81754421957419e-05,
+      "loss": 1.1074,
+      "step": 3718
+    },
+    {
+      "epoch": 0.22595540433805214,
+      "grad_norm": 0.3160702586174011,
+      "learning_rate": 8.816925942445196e-05,
+      "loss": 1.0595,
+      "step": 3719
+    },
+    {
+      "epoch": 0.22601616137067865,
+      "grad_norm": 0.47938573360443115,
+      "learning_rate": 8.816307525406042e-05,
+      "loss": 1.2385,
+      "step": 3720
+    },
+    {
+      "epoch": 0.22607691840330518,
+      "grad_norm": 0.9228609800338745,
+      "learning_rate": 8.815688968479399e-05,
+      "loss": 1.1115,
+      "step": 3721
+    },
+    {
+      "epoch": 0.22613767543593172,
+      "grad_norm": 0.32599395513534546,
+      "learning_rate": 8.815070271687937e-05,
+      "loss": 1.1696,
+      "step": 3722
+    },
+    {
+      "epoch": 0.22619843246855822,
+      "grad_norm": 0.26398760080337524,
+      "learning_rate": 8.814451435054339e-05,
+      "loss": 1.112,
+      "step": 3723
+    },
+    {
+      "epoch": 0.22625918950118476,
+      "grad_norm": 0.2827211618423462,
+      "learning_rate": 8.813832458601285e-05,
+      "loss": 1.1057,
+      "step": 3724
+    },
+    {
+      "epoch": 0.2263199465338113,
+      "grad_norm": 0.9766664505004883,
+      "learning_rate": 8.813213342351466e-05,
+      "loss": 1.2191,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2263807035664378,
+      "grad_norm": 0.17697346210479736,
+      "learning_rate": 8.812594086327574e-05,
+      "loss": 1.1614,
+      "step": 3726
+    },
+    {
+      "epoch": 0.22644146059906434,
+      "grad_norm": 0.35698121786117554,
+      "learning_rate": 8.811974690552311e-05,
+      "loss": 1.1805,
+      "step": 3727
+    },
+    {
+      "epoch": 0.22650221763169087,
+      "grad_norm": 0.2610389292240143,
+      "learning_rate": 8.811355155048377e-05,
+      "loss": 1.1218,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2265629746643174,
+      "grad_norm": 0.13635700941085815,
+      "learning_rate": 8.810735479838483e-05,
+      "loss": 1.0754,
+      "step": 3729
+    },
+    {
+      "epoch": 0.22662373169694391,
+      "grad_norm": 0.29382216930389404,
+      "learning_rate": 8.810115664945343e-05,
+      "loss": 1.1801,
+      "step": 3730
+    },
+    {
+      "epoch": 0.22668448872957045,
+      "grad_norm": 0.3335505723953247,
+      "learning_rate": 8.809495710391678e-05,
+      "loss": 1.2342,
+      "step": 3731
+    },
+    {
+      "epoch": 0.22674524576219698,
+      "grad_norm": 0.4739501476287842,
+      "learning_rate": 8.808875616200211e-05,
+      "loss": 1.0965,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2268060027948235,
+      "grad_norm": 0.5466656684875488,
+      "learning_rate": 8.808255382393672e-05,
+      "loss": 1.1186,
+      "step": 3733
+    },
+    {
+      "epoch": 0.22686675982745003,
+      "grad_norm": 0.2992178201675415,
+      "learning_rate": 8.807635008994796e-05,
+      "loss": 1.1411,
+      "step": 3734
+    },
+    {
+      "epoch": 0.22692751686007656,
+      "grad_norm": 0.20904937386512756,
+      "learning_rate": 8.807014496026322e-05,
+      "loss": 1.1973,
+      "step": 3735
+    },
+    {
+      "epoch": 0.22698827389270307,
+      "grad_norm": 0.2990994155406952,
+      "learning_rate": 8.806393843510996e-05,
+      "loss": 1.1496,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2270490309253296,
+      "grad_norm": 0.28559526801109314,
+      "learning_rate": 8.805773051471568e-05,
+      "loss": 1.071,
+      "step": 3737
+    },
+    {
+      "epoch": 0.22710978795795614,
+      "grad_norm": 0.1629854291677475,
+      "learning_rate": 8.805152119930796e-05,
+      "loss": 1.1273,
+      "step": 3738
+    },
+    {
+      "epoch": 0.22717054499058267,
+      "grad_norm": 1.2423242330551147,
+      "learning_rate": 8.804531048911433e-05,
+      "loss": 1.1299,
+      "step": 3739
+    },
+    {
+      "epoch": 0.22723130202320918,
+      "grad_norm": 0.39676475524902344,
+      "learning_rate": 8.803909838436251e-05,
+      "loss": 1.1738,
+      "step": 3740
+    },
+    {
+      "epoch": 0.22729205905583572,
+      "grad_norm": 0.3079032599925995,
+      "learning_rate": 8.803288488528019e-05,
+      "loss": 1.0961,
+      "step": 3741
+    },
+    {
+      "epoch": 0.22735281608846225,
+      "grad_norm": 0.1665235161781311,
+      "learning_rate": 8.802666999209512e-05,
+      "loss": 1.0557,
+      "step": 3742
+    },
+    {
+      "epoch": 0.22741357312108876,
+      "grad_norm": 0.15917164087295532,
+      "learning_rate": 8.802045370503512e-05,
+      "loss": 1.0823,
+      "step": 3743
+    },
+    {
+      "epoch": 0.2274743301537153,
+      "grad_norm": 0.7063826322555542,
+      "learning_rate": 8.801423602432801e-05,
+      "loss": 1.1571,
+      "step": 3744
+    },
+    {
+      "epoch": 0.22753508718634183,
+      "grad_norm": 0.24075105786323547,
+      "learning_rate": 8.800801695020176e-05,
+      "loss": 1.1807,
+      "step": 3745
+    },
+    {
+      "epoch": 0.22759584421896834,
+      "grad_norm": 27.179767608642578,
+      "learning_rate": 8.800179648288429e-05,
+      "loss": 1.0843,
+      "step": 3746
+    },
+    {
+      "epoch": 0.22765660125159487,
+      "grad_norm": 0.18558628857135773,
+      "learning_rate": 8.799557462260361e-05,
+      "loss": 1.0655,
+      "step": 3747
+    },
+    {
+      "epoch": 0.2277173582842214,
+      "grad_norm": 0.24082210659980774,
+      "learning_rate": 8.798935136958782e-05,
+      "loss": 1.0474,
+      "step": 3748
+    },
+    {
+      "epoch": 0.2277781153168479,
+      "grad_norm": 0.21904778480529785,
+      "learning_rate": 8.798312672406501e-05,
+      "loss": 1.1789,
+      "step": 3749
+    },
+    {
+      "epoch": 0.22783887234947445,
+      "grad_norm": 0.38997456431388855,
+      "learning_rate": 8.797690068626334e-05,
+      "loss": 1.3074,
+      "step": 3750
+    },
+    {
+      "epoch": 0.22789962938210098,
+      "grad_norm": 0.2737247049808502,
+      "learning_rate": 8.797067325641104e-05,
+      "loss": 1.1582,
+      "step": 3751
+    },
+    {
+      "epoch": 0.22796038641472752,
+      "grad_norm": 0.3012070059776306,
+      "learning_rate": 8.796444443473637e-05,
+      "loss": 1.1549,
+      "step": 3752
+    },
+    {
+      "epoch": 0.22802114344735402,
+      "grad_norm": 0.16187265515327454,
+      "learning_rate": 8.795821422146766e-05,
+      "loss": 1.042,
+      "step": 3753
+    },
+    {
+      "epoch": 0.22808190047998056,
+      "grad_norm": 0.23431642353534698,
+      "learning_rate": 8.795198261683326e-05,
+      "loss": 1.2678,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2281426575126071,
+      "grad_norm": 0.2818852961063385,
+      "learning_rate": 8.794574962106161e-05,
+      "loss": 1.1846,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2282034145452336,
+      "grad_norm": 0.161614328622818,
+      "learning_rate": 8.793951523438116e-05,
+      "loss": 1.103,
+      "step": 3756
+    },
+    {
+      "epoch": 0.22826417157786014,
+      "grad_norm": 0.2946518361568451,
+      "learning_rate": 8.793327945702046e-05,
+      "loss": 1.0485,
+      "step": 3757
+    },
+    {
+      "epoch": 0.22832492861048667,
+      "grad_norm": 0.22220878303050995,
+      "learning_rate": 8.792704228920805e-05,
+      "loss": 1.0509,
+      "step": 3758
+    },
+    {
+      "epoch": 0.22838568564311318,
+      "grad_norm": 0.18830649554729462,
+      "learning_rate": 8.79208037311726e-05,
+      "loss": 1.12,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2284464426757397,
+      "grad_norm": 0.18957538902759552,
+      "learning_rate": 8.791456378314273e-05,
+      "loss": 1.1573,
+      "step": 3760
+    },
+    {
+      "epoch": 0.22850719970836625,
+      "grad_norm": 0.3290649652481079,
+      "learning_rate": 8.790832244534721e-05,
+      "loss": 1.1844,
+      "step": 3761
+    },
+    {
+      "epoch": 0.22856795674099278,
+      "grad_norm": 0.34733831882476807,
+      "learning_rate": 8.790207971801481e-05,
+      "loss": 1.2929,
+      "step": 3762
+    },
+    {
+      "epoch": 0.2286287137736193,
+      "grad_norm": 0.33905866742134094,
+      "learning_rate": 8.789583560137435e-05,
+      "loss": 1.1612,
+      "step": 3763
+    },
+    {
+      "epoch": 0.22868947080624583,
+      "grad_norm": 0.3364482522010803,
+      "learning_rate": 8.78895900956547e-05,
+      "loss": 1.0785,
+      "step": 3764
+    },
+    {
+      "epoch": 0.22875022783887236,
+      "grad_norm": 0.23005352914333344,
+      "learning_rate": 8.78833432010848e-05,
+      "loss": 1.1679,
+      "step": 3765
+    },
+    {
+      "epoch": 0.22881098487149887,
+      "grad_norm": 0.25035086274147034,
+      "learning_rate": 8.787709491789364e-05,
+      "loss": 1.0954,
+      "step": 3766
+    },
+    {
+      "epoch": 0.2288717419041254,
+      "grad_norm": 0.26574721932411194,
+      "learning_rate": 8.787084524631025e-05,
+      "loss": 1.1232,
+      "step": 3767
+    },
+    {
+      "epoch": 0.22893249893675194,
+      "grad_norm": 0.4117295742034912,
+      "learning_rate": 8.786459418656369e-05,
+      "loss": 1.172,
+      "step": 3768
+    },
+    {
+      "epoch": 0.22899325596937845,
+      "grad_norm": 0.23607565462589264,
+      "learning_rate": 8.785834173888315e-05,
+      "loss": 1.1597,
+      "step": 3769
+    },
+    {
+      "epoch": 0.22905401300200498,
+      "grad_norm": 0.2233234941959381,
+      "learning_rate": 8.785208790349773e-05,
+      "loss": 1.1454,
+      "step": 3770
+    },
+    {
+      "epoch": 0.22911477003463152,
+      "grad_norm": 0.20566169917583466,
+      "learning_rate": 8.784583268063673e-05,
+      "loss": 1.151,
+      "step": 3771
+    },
+    {
+      "epoch": 0.22917552706725802,
+      "grad_norm": 0.1566246598958969,
+      "learning_rate": 8.783957607052941e-05,
+      "loss": 1.0916,
+      "step": 3772
+    },
+    {
+      "epoch": 0.22923628409988456,
+      "grad_norm": 0.30101585388183594,
+      "learning_rate": 8.783331807340514e-05,
+      "loss": 1.2645,
+      "step": 3773
+    },
+    {
+      "epoch": 0.2292970411325111,
+      "grad_norm": 0.5011516213417053,
+      "learning_rate": 8.782705868949326e-05,
+      "loss": 1.2457,
+      "step": 3774
+    },
+    {
+      "epoch": 0.22935779816513763,
+      "grad_norm": 0.1741447001695633,
+      "learning_rate": 8.782079791902322e-05,
+      "loss": 1.0693,
+      "step": 3775
+    },
+    {
+      "epoch": 0.22941855519776413,
+      "grad_norm": 0.41780203580856323,
+      "learning_rate": 8.781453576222455e-05,
+      "loss": 1.0987,
+      "step": 3776
+    },
+    {
+      "epoch": 0.22947931223039067,
+      "grad_norm": 0.23424886167049408,
+      "learning_rate": 8.780827221932675e-05,
+      "loss": 1.1255,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2295400692630172,
+      "grad_norm": 0.20799852907657623,
+      "learning_rate": 8.780200729055943e-05,
+      "loss": 1.1814,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2296008262956437,
+      "grad_norm": 0.26352182030677795,
+      "learning_rate": 8.77957409761522e-05,
+      "loss": 1.1503,
+      "step": 3779
+    },
+    {
+      "epoch": 0.22966158332827025,
+      "grad_norm": 0.19761155545711517,
+      "learning_rate": 8.778947327633481e-05,
+      "loss": 1.2301,
+      "step": 3780
+    },
+    {
+      "epoch": 0.22972234036089678,
+      "grad_norm": 0.4628106355667114,
+      "learning_rate": 8.778320419133697e-05,
+      "loss": 1.1296,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2297830973935233,
+      "grad_norm": 0.19939939677715302,
+      "learning_rate": 8.777693372138846e-05,
+      "loss": 1.1382,
+      "step": 3782
+    },
+    {
+      "epoch": 0.22984385442614982,
+      "grad_norm": 0.16180074214935303,
+      "learning_rate": 8.777066186671915e-05,
+      "loss": 1.0956,
+      "step": 3783
+    },
+    {
+      "epoch": 0.22990461145877636,
+      "grad_norm": 0.19431500136852264,
+      "learning_rate": 8.776438862755893e-05,
+      "loss": 1.1338,
+      "step": 3784
+    },
+    {
+      "epoch": 0.22996536849140287,
+      "grad_norm": 0.28697606921195984,
+      "learning_rate": 8.775811400413774e-05,
+      "loss": 1.1386,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2300261255240294,
+      "grad_norm": 0.196644127368927,
+      "learning_rate": 8.775183799668559e-05,
+      "loss": 1.1499,
+      "step": 3786
+    },
+    {
+      "epoch": 0.23008688255665594,
+      "grad_norm": 0.26397594809532166,
+      "learning_rate": 8.774556060543253e-05,
+      "loss": 1.1589,
+      "step": 3787
+    },
+    {
+      "epoch": 0.23014763958928247,
+      "grad_norm": 0.20761661231517792,
+      "learning_rate": 8.773928183060862e-05,
+      "loss": 1.1331,
+      "step": 3788
+    },
+    {
+      "epoch": 0.23020839662190898,
+      "grad_norm": 0.19499783217906952,
+      "learning_rate": 8.773300167244407e-05,
+      "loss": 1.0759,
+      "step": 3789
+    },
+    {
+      "epoch": 0.2302691536545355,
+      "grad_norm": 0.24904966354370117,
+      "learning_rate": 8.772672013116903e-05,
+      "loss": 1.0893,
+      "step": 3790
+    },
+    {
+      "epoch": 0.23032991068716205,
+      "grad_norm": 0.2817010283470154,
+      "learning_rate": 8.772043720701378e-05,
+      "loss": 1.0732,
+      "step": 3791
+    },
+    {
+      "epoch": 0.23039066771978856,
+      "grad_norm": 0.2399549037218094,
+      "learning_rate": 8.771415290020862e-05,
+      "loss": 1.1777,
+      "step": 3792
+    },
+    {
+      "epoch": 0.2304514247524151,
+      "grad_norm": 0.4989120066165924,
+      "learning_rate": 8.770786721098387e-05,
+      "loss": 1.2487,
+      "step": 3793
+    },
+    {
+      "epoch": 0.23051218178504163,
+      "grad_norm": 0.21616044640541077,
+      "learning_rate": 8.770158013956999e-05,
+      "loss": 1.0881,
+      "step": 3794
+    },
+    {
+      "epoch": 0.23057293881766813,
+      "grad_norm": 0.19007226824760437,
+      "learning_rate": 8.769529168619739e-05,
+      "loss": 1.0648,
+      "step": 3795
+    },
+    {
+      "epoch": 0.23063369585029467,
+      "grad_norm": 0.25410953164100647,
+      "learning_rate": 8.768900185109658e-05,
+      "loss": 1.1141,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2306944528829212,
+      "grad_norm": 0.28295817971229553,
+      "learning_rate": 8.768271063449812e-05,
+      "loss": 1.0692,
+      "step": 3797
+    },
+    {
+      "epoch": 0.23075520991554774,
+      "grad_norm": 0.2932504415512085,
+      "learning_rate": 8.767641803663262e-05,
+      "loss": 1.1745,
+      "step": 3798
+    },
+    {
+      "epoch": 0.23081596694817424,
+      "grad_norm": 2.4574594497680664,
+      "learning_rate": 8.767012405773074e-05,
+      "loss": 1.2215,
+      "step": 3799
+    },
+    {
+      "epoch": 0.23087672398080078,
+      "grad_norm": 0.2336098849773407,
+      "learning_rate": 8.766382869802318e-05,
+      "loss": 1.3222,
+      "step": 3800
+    },
+    {
+      "epoch": 0.23093748101342731,
+      "grad_norm": 0.522935152053833,
+      "learning_rate": 8.76575319577407e-05,
+      "loss": 1.3722,
+      "step": 3801
+    },
+    {
+      "epoch": 0.23099823804605382,
+      "grad_norm": 0.17931655049324036,
+      "learning_rate": 8.76512338371141e-05,
+      "loss": 1.0627,
+      "step": 3802
+    },
+    {
+      "epoch": 0.23105899507868036,
+      "grad_norm": 0.24202129244804382,
+      "learning_rate": 8.764493433637426e-05,
+      "loss": 1.216,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2311197521113069,
+      "grad_norm": 0.1920606642961502,
+      "learning_rate": 8.763863345575208e-05,
+      "loss": 1.2506,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2311805091439334,
+      "grad_norm": 0.4114784300327301,
+      "learning_rate": 8.763233119547849e-05,
+      "loss": 1.0882,
+      "step": 3805
+    },
+    {
+      "epoch": 0.23124126617655993,
+      "grad_norm": 0.21371228992938995,
+      "learning_rate": 8.762602755578454e-05,
+      "loss": 1.1809,
+      "step": 3806
+    },
+    {
+      "epoch": 0.23130202320918647,
+      "grad_norm": 0.31484973430633545,
+      "learning_rate": 8.761972253690127e-05,
+      "loss": 1.0815,
+      "step": 3807
+    },
+    {
+      "epoch": 0.23136278024181298,
+      "grad_norm": 0.24163700640201569,
+      "learning_rate": 8.76134161390598e-05,
+      "loss": 1.1532,
+      "step": 3808
+    },
+    {
+      "epoch": 0.2314235372744395,
+      "grad_norm": 0.1594315767288208,
+      "learning_rate": 8.76071083624913e-05,
+      "loss": 1.1058,
+      "step": 3809
+    },
+    {
+      "epoch": 0.23148429430706605,
+      "grad_norm": 0.2613028883934021,
+      "learning_rate": 8.760079920742698e-05,
+      "loss": 1.1137,
+      "step": 3810
+    },
+    {
+      "epoch": 0.23154505133969258,
+      "grad_norm": 0.2661849856376648,
+      "learning_rate": 8.759448867409807e-05,
+      "loss": 1.029,
+      "step": 3811
+    },
+    {
+      "epoch": 0.2316058083723191,
+      "grad_norm": 0.2608145475387573,
+      "learning_rate": 8.758817676273593e-05,
+      "loss": 1.177,
+      "step": 3812
+    },
+    {
+      "epoch": 0.23166656540494562,
+      "grad_norm": 0.31724369525909424,
+      "learning_rate": 8.758186347357191e-05,
+      "loss": 1.2088,
+      "step": 3813
+    },
+    {
+      "epoch": 0.23172732243757216,
+      "grad_norm": 0.21606537699699402,
+      "learning_rate": 8.75755488068374e-05,
+      "loss": 1.1477,
+      "step": 3814
+    },
+    {
+      "epoch": 0.23178807947019867,
+      "grad_norm": 20.253334045410156,
+      "learning_rate": 8.75692327627639e-05,
+      "loss": 1.0873,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2318488365028252,
+      "grad_norm": 0.1502893716096878,
+      "learning_rate": 8.756291534158292e-05,
+      "loss": 1.0762,
+      "step": 3816
+    },
+    {
+      "epoch": 0.23190959353545174,
+      "grad_norm": 0.14848195016384125,
+      "learning_rate": 8.755659654352599e-05,
+      "loss": 1.0858,
+      "step": 3817
+    },
+    {
+      "epoch": 0.23197035056807824,
+      "grad_norm": 0.15895536541938782,
+      "learning_rate": 8.755027636882478e-05,
+      "loss": 1.0642,
+      "step": 3818
+    },
+    {
+      "epoch": 0.23203110760070478,
+      "grad_norm": 0.1896950900554657,
+      "learning_rate": 8.754395481771091e-05,
+      "loss": 1.1298,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2320918646333313,
+      "grad_norm": 0.18929994106292725,
+      "learning_rate": 8.753763189041614e-05,
+      "loss": 1.0555,
+      "step": 3820
+    },
+    {
+      "epoch": 0.23215262166595785,
+      "grad_norm": 0.21950852870941162,
+      "learning_rate": 8.75313075871722e-05,
+      "loss": 1.1665,
+      "step": 3821
+    },
+    {
+      "epoch": 0.23221337869858436,
+      "grad_norm": 0.16273915767669678,
+      "learning_rate": 8.752498190821095e-05,
+      "loss": 1.0839,
+      "step": 3822
+    },
+    {
+      "epoch": 0.2322741357312109,
+      "grad_norm": 0.15747126936912537,
+      "learning_rate": 8.75186548537642e-05,
+      "loss": 1.0848,
+      "step": 3823
+    },
+    {
+      "epoch": 0.23233489276383743,
+      "grad_norm": 0.23414598405361176,
+      "learning_rate": 8.751232642406392e-05,
+      "loss": 1.1249,
+      "step": 3824
+    },
+    {
+      "epoch": 0.23239564979646393,
+      "grad_norm": 4.0957560539245605,
+      "learning_rate": 8.750599661934206e-05,
+      "loss": 1.1789,
+      "step": 3825
+    },
+    {
+      "epoch": 0.23245640682909047,
+      "grad_norm": 0.3437539339065552,
+      "learning_rate": 8.749966543983066e-05,
+      "loss": 1.1209,
+      "step": 3826
+    },
+    {
+      "epoch": 0.232517163861717,
+      "grad_norm": 0.3358255624771118,
+      "learning_rate": 8.749333288576177e-05,
+      "loss": 1.2534,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2325779208943435,
+      "grad_norm": 0.18156570196151733,
+      "learning_rate": 8.748699895736752e-05,
+      "loss": 1.067,
+      "step": 3828
+    },
+    {
+      "epoch": 0.23263867792697004,
+      "grad_norm": 0.26023340225219727,
+      "learning_rate": 8.748066365488006e-05,
+      "loss": 1.1019,
+      "step": 3829
+    },
+    {
+      "epoch": 0.23269943495959658,
+      "grad_norm": 0.19864824414253235,
+      "learning_rate": 8.747432697853164e-05,
+      "loss": 1.0978,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2327601919922231,
+      "grad_norm": 0.1605607122182846,
+      "learning_rate": 8.74679889285545e-05,
+      "loss": 1.1128,
+      "step": 3831
+    },
+    {
+      "epoch": 0.23282094902484962,
+      "grad_norm": 0.2532506287097931,
+      "learning_rate": 8.746164950518101e-05,
+      "loss": 1.2199,
+      "step": 3832
+    },
+    {
+      "epoch": 0.23288170605747616,
+      "grad_norm": 0.15219800174236298,
+      "learning_rate": 8.745530870864351e-05,
+      "loss": 1.0595,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2329424630901027,
+      "grad_norm": 0.22808650135993958,
+      "learning_rate": 8.744896653917443e-05,
+      "loss": 1.0397,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2330032201227292,
+      "grad_norm": 1.2698999643325806,
+      "learning_rate": 8.744262299700623e-05,
+      "loss": 1.0765,
+      "step": 3835
+    },
+    {
+      "epoch": 0.23306397715535573,
+      "grad_norm": 0.134122833609581,
+      "learning_rate": 8.743627808237146e-05,
+      "loss": 1.249,
+      "step": 3836
+    },
+    {
+      "epoch": 0.23312473418798227,
+      "grad_norm": 0.22514206171035767,
+      "learning_rate": 8.742993179550267e-05,
+      "loss": 1.1934,
+      "step": 3837
+    },
+    {
+      "epoch": 0.23318549122060878,
+      "grad_norm": 0.18784482777118683,
+      "learning_rate": 8.74235841366325e-05,
+      "loss": 1.2562,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2332462482532353,
+      "grad_norm": 0.24985751509666443,
+      "learning_rate": 8.741723510599363e-05,
+      "loss": 1.0411,
+      "step": 3839
+    },
+    {
+      "epoch": 0.23330700528586185,
+      "grad_norm": 0.17648372054100037,
+      "learning_rate": 8.741088470381877e-05,
+      "loss": 1.0666,
+      "step": 3840
+    },
+    {
+      "epoch": 0.23336776231848835,
+      "grad_norm": 0.2940845787525177,
+      "learning_rate": 8.740453293034069e-05,
+      "loss": 1.0762,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2334285193511149,
+      "grad_norm": 0.13416394591331482,
+      "learning_rate": 8.739817978579223e-05,
+      "loss": 1.0697,
+      "step": 3842
+    },
+    {
+      "epoch": 0.23348927638374142,
+      "grad_norm": 0.1555926352739334,
+      "learning_rate": 8.739182527040626e-05,
+      "loss": 1.1161,
+      "step": 3843
+    },
+    {
+      "epoch": 0.23355003341636796,
+      "grad_norm": 0.23895837366580963,
+      "learning_rate": 8.738546938441571e-05,
+      "loss": 1.2879,
+      "step": 3844
+    },
+    {
+      "epoch": 0.23361079044899447,
+      "grad_norm": 0.19918054342269897,
+      "learning_rate": 8.737911212805355e-05,
+      "loss": 1.173,
+      "step": 3845
+    },
+    {
+      "epoch": 0.233671547481621,
+      "grad_norm": 0.30027905106544495,
+      "learning_rate": 8.737275350155281e-05,
+      "loss": 1.2347,
+      "step": 3846
+    },
+    {
+      "epoch": 0.23373230451424754,
+      "grad_norm": 0.25957414507865906,
+      "learning_rate": 8.736639350514658e-05,
+      "loss": 1.1053,
+      "step": 3847
+    },
+    {
+      "epoch": 0.23379306154687404,
+      "grad_norm": 0.24853013455867767,
+      "learning_rate": 8.736003213906795e-05,
+      "loss": 1.0713,
+      "step": 3848
+    },
+    {
+      "epoch": 0.23385381857950058,
+      "grad_norm": 0.15455150604248047,
+      "learning_rate": 8.735366940355014e-05,
+      "loss": 1.0926,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2339145756121271,
+      "grad_norm": 0.5605828166007996,
+      "learning_rate": 8.734730529882634e-05,
+      "loss": 1.1265,
+      "step": 3850
+    },
+    {
+      "epoch": 0.23397533264475362,
+      "grad_norm": 0.1810554563999176,
+      "learning_rate": 8.734093982512986e-05,
+      "loss": 1.2079,
+      "step": 3851
+    },
+    {
+      "epoch": 0.23403608967738015,
+      "grad_norm": 0.21177291870117188,
+      "learning_rate": 8.733457298269401e-05,
+      "loss": 1.1391,
+      "step": 3852
+    },
+    {
+      "epoch": 0.2340968467100067,
+      "grad_norm": 0.16143867373466492,
+      "learning_rate": 8.732820477175217e-05,
+      "loss": 1.1609,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2341576037426332,
+      "grad_norm": 0.1898384839296341,
+      "learning_rate": 8.732183519253776e-05,
+      "loss": 1.1843,
+      "step": 3854
+    },
+    {
+      "epoch": 0.23421836077525973,
+      "grad_norm": 0.17061768472194672,
+      "learning_rate": 8.731546424528428e-05,
+      "loss": 1.1317,
+      "step": 3855
+    },
+    {
+      "epoch": 0.23427911780788627,
+      "grad_norm": 0.17413118481636047,
+      "learning_rate": 8.730909193022523e-05,
+      "loss": 1.0467,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2343398748405128,
+      "grad_norm": 0.39212051033973694,
+      "learning_rate": 8.730271824759421e-05,
+      "loss": 1.1719,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2344006318731393,
+      "grad_norm": 0.21302811801433563,
+      "learning_rate": 8.729634319762486e-05,
+      "loss": 1.2754,
+      "step": 3858
+    },
+    {
+      "epoch": 0.23446138890576584,
+      "grad_norm": 23.89767837524414,
+      "learning_rate": 8.728996678055081e-05,
+      "loss": 1.0702,
+      "step": 3859
+    },
+    {
+      "epoch": 0.23452214593839238,
+      "grad_norm": 0.1730816662311554,
+      "learning_rate": 8.728358899660585e-05,
+      "loss": 1.0778,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2345829029710189,
+      "grad_norm": 7.371491432189941,
+      "learning_rate": 8.727720984602371e-05,
+      "loss": 1.0689,
+      "step": 3861
+    },
+    {
+      "epoch": 0.23464366000364542,
+      "grad_norm": 0.1834903359413147,
+      "learning_rate": 8.727082932903823e-05,
+      "loss": 1.0628,
+      "step": 3862
+    },
+    {
+      "epoch": 0.23470441703627196,
+      "grad_norm": 0.1326923370361328,
+      "learning_rate": 8.72644474458833e-05,
+      "loss": 1.0816,
+      "step": 3863
+    },
+    {
+      "epoch": 0.23476517406889846,
+      "grad_norm": 0.19308006763458252,
+      "learning_rate": 8.725806419679287e-05,
+      "loss": 1.1026,
+      "step": 3864
+    },
+    {
+      "epoch": 0.234825931101525,
+      "grad_norm": 2.243105173110962,
+      "learning_rate": 8.725167958200088e-05,
+      "loss": 1.0872,
+      "step": 3865
+    },
+    {
+      "epoch": 0.23488668813415153,
+      "grad_norm": 0.3164255619049072,
+      "learning_rate": 8.724529360174137e-05,
+      "loss": 1.1596,
+      "step": 3866
+    },
+    {
+      "epoch": 0.23494744516677807,
+      "grad_norm": 0.4158472716808319,
+      "learning_rate": 8.723890625624844e-05,
+      "loss": 1.2297,
+      "step": 3867
+    },
+    {
+      "epoch": 0.23500820219940458,
+      "grad_norm": 0.2862829566001892,
+      "learning_rate": 8.723251754575619e-05,
+      "loss": 1.2261,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2350689592320311,
+      "grad_norm": 0.14094950258731842,
+      "learning_rate": 8.72261274704988e-05,
+      "loss": 1.0538,
+      "step": 3869
+    },
+    {
+      "epoch": 0.23512971626465765,
+      "grad_norm": 0.2651413679122925,
+      "learning_rate": 8.721973603071053e-05,
+      "loss": 1.0878,
+      "step": 3870
+    },
+    {
+      "epoch": 0.23519047329728415,
+      "grad_norm": 0.19096867740154266,
+      "learning_rate": 8.721334322662564e-05,
+      "loss": 1.2233,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2352512303299107,
+      "grad_norm": 0.21207264065742493,
+      "learning_rate": 8.720694905847844e-05,
+      "loss": 1.127,
+      "step": 3872
+    },
+    {
+      "epoch": 0.23531198736253722,
+      "grad_norm": 0.16888320446014404,
+      "learning_rate": 8.720055352650337e-05,
+      "loss": 1.1311,
+      "step": 3873
+    },
+    {
+      "epoch": 0.23537274439516373,
+      "grad_norm": 0.25098758935928345,
+      "learning_rate": 8.719415663093479e-05,
+      "loss": 1.1042,
+      "step": 3874
+    },
+    {
+      "epoch": 0.23543350142779026,
+      "grad_norm": 0.18126602470874786,
+      "learning_rate": 8.71877583720072e-05,
+      "loss": 1.133,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2354942584604168,
+      "grad_norm": 0.19230689108371735,
+      "learning_rate": 8.718135874995516e-05,
+      "loss": 1.1127,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2355550154930433,
+      "grad_norm": 0.15672238171100616,
+      "learning_rate": 8.717495776501323e-05,
+      "loss": 1.0651,
+      "step": 3877
+    },
+    {
+      "epoch": 0.23561577252566984,
+      "grad_norm": 0.16875606775283813,
+      "learning_rate": 8.716855541741603e-05,
+      "loss": 1.0939,
+      "step": 3878
+    },
+    {
+      "epoch": 0.23567652955829638,
+      "grad_norm": 0.2219727784395218,
+      "learning_rate": 8.716215170739825e-05,
+      "loss": 1.1563,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2357372865909229,
+      "grad_norm": 0.33217132091522217,
+      "learning_rate": 8.715574663519462e-05,
+      "loss": 1.2994,
+      "step": 3880
+    },
+    {
+      "epoch": 0.23579804362354942,
+      "grad_norm": 0.21564796566963196,
+      "learning_rate": 8.71493402010399e-05,
+      "loss": 1.244,
+      "step": 3881
+    },
+    {
+      "epoch": 0.23585880065617595,
+      "grad_norm": 0.8253178596496582,
+      "learning_rate": 8.714293240516894e-05,
+      "loss": 1.1492,
+      "step": 3882
+    },
+    {
+      "epoch": 0.2359195576888025,
+      "grad_norm": 0.18198446929454803,
+      "learning_rate": 8.713652324781662e-05,
+      "loss": 1.0882,
+      "step": 3883
+    },
+    {
+      "epoch": 0.235980314721429,
+      "grad_norm": 0.20700703561306,
+      "learning_rate": 8.713011272921785e-05,
+      "loss": 1.2974,
+      "step": 3884
+    },
+    {
+      "epoch": 0.23604107175405553,
+      "grad_norm": 0.12848831713199615,
+      "learning_rate": 8.712370084960764e-05,
+      "loss": 1.1016,
+      "step": 3885
+    },
+    {
+      "epoch": 0.23610182878668207,
+      "grad_norm": 0.16396000981330872,
+      "learning_rate": 8.711728760922097e-05,
+      "loss": 1.0422,
+      "step": 3886
+    },
+    {
+      "epoch": 0.23616258581930857,
+      "grad_norm": 0.16817036271095276,
+      "learning_rate": 8.711087300829297e-05,
+      "loss": 1.1538,
+      "step": 3887
+    },
+    {
+      "epoch": 0.2362233428519351,
+      "grad_norm": 0.19173792004585266,
+      "learning_rate": 8.710445704705874e-05,
+      "loss": 1.0392,
+      "step": 3888
+    },
+    {
+      "epoch": 0.23628409988456164,
+      "grad_norm": 0.14542429149150848,
+      "learning_rate": 8.709803972575346e-05,
+      "loss": 1.1001,
+      "step": 3889
+    },
+    {
+      "epoch": 0.23634485691718815,
+      "grad_norm": 0.21353450417518616,
+      "learning_rate": 8.709162104461238e-05,
+      "loss": 1.0637,
+      "step": 3890
+    },
+    {
+      "epoch": 0.23640561394981469,
+      "grad_norm": 0.1834249198436737,
+      "learning_rate": 8.708520100387074e-05,
+      "loss": 1.1117,
+      "step": 3891
+    },
+    {
+      "epoch": 0.23646637098244122,
+      "grad_norm": 0.15393726527690887,
+      "learning_rate": 8.707877960376392e-05,
+      "loss": 1.0971,
+      "step": 3892
+    },
+    {
+      "epoch": 0.23652712801506776,
+      "grad_norm": 0.5606473088264465,
+      "learning_rate": 8.707235684452723e-05,
+      "loss": 1.0863,
+      "step": 3893
+    },
+    {
+      "epoch": 0.23658788504769426,
+      "grad_norm": 0.13707295060157776,
+      "learning_rate": 8.706593272639615e-05,
+      "loss": 1.0518,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2366486420803208,
+      "grad_norm": 0.1525871753692627,
+      "learning_rate": 8.705950724960614e-05,
+      "loss": 1.0698,
+      "step": 3895
+    },
+    {
+      "epoch": 0.23670939911294733,
+      "grad_norm": 0.2221229523420334,
+      "learning_rate": 8.705308041439272e-05,
+      "loss": 1.1103,
+      "step": 3896
+    },
+    {
+      "epoch": 0.23677015614557384,
+      "grad_norm": 0.1364176869392395,
+      "learning_rate": 8.70466522209915e-05,
+      "loss": 1.0736,
+      "step": 3897
+    },
+    {
+      "epoch": 0.23683091317820038,
+      "grad_norm": 0.18070849776268005,
+      "learning_rate": 8.704022266963807e-05,
+      "loss": 1.2127,
+      "step": 3898
+    },
+    {
+      "epoch": 0.2368916702108269,
+      "grad_norm": 0.2248815894126892,
+      "learning_rate": 8.703379176056811e-05,
+      "loss": 1.1609,
+      "step": 3899
+    },
+    {
+      "epoch": 0.23695242724345342,
+      "grad_norm": 0.5836701393127441,
+      "learning_rate": 8.702735949401736e-05,
+      "loss": 1.1479,
+      "step": 3900
+    },
+    {
+      "epoch": 0.23701318427607995,
+      "grad_norm": 0.13758723437786102,
+      "learning_rate": 8.702092587022159e-05,
+      "loss": 1.0971,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2370739413087065,
+      "grad_norm": 0.2805210053920746,
+      "learning_rate": 8.701449088941662e-05,
+      "loss": 1.0365,
+      "step": 3902
+    },
+    {
+      "epoch": 0.23713469834133302,
+      "grad_norm": 0.16959865391254425,
+      "learning_rate": 8.700805455183835e-05,
+      "loss": 1.155,
+      "step": 3903
+    },
+    {
+      "epoch": 0.23719545537395953,
+      "grad_norm": 0.14467813074588776,
+      "learning_rate": 8.700161685772267e-05,
+      "loss": 1.0939,
+      "step": 3904
+    },
+    {
+      "epoch": 0.23725621240658606,
+      "grad_norm": 0.156314417719841,
+      "learning_rate": 8.699517780730557e-05,
+      "loss": 1.1059,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2373169694392126,
+      "grad_norm": 0.29469722509384155,
+      "learning_rate": 8.698873740082309e-05,
+      "loss": 1.1633,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2373777264718391,
+      "grad_norm": 13.932208061218262,
+      "learning_rate": 8.698229563851129e-05,
+      "loss": 1.1156,
+      "step": 3907
+    },
+    {
+      "epoch": 0.23743848350446564,
+      "grad_norm": 0.1899506002664566,
+      "learning_rate": 8.697585252060629e-05,
+      "loss": 1.117,
+      "step": 3908
+    },
+    {
+      "epoch": 0.23749924053709218,
+      "grad_norm": 0.16111107170581818,
+      "learning_rate": 8.696940804734426e-05,
+      "loss": 1.0811,
+      "step": 3909
+    },
+    {
+      "epoch": 0.23755999756971868,
+      "grad_norm": 0.1723010540008545,
+      "learning_rate": 8.696296221896144e-05,
+      "loss": 1.0966,
+      "step": 3910
+    },
+    {
+      "epoch": 0.23762075460234522,
+      "grad_norm": 0.18821151554584503,
+      "learning_rate": 8.695651503569409e-05,
+      "loss": 1.1135,
+      "step": 3911
+    },
+    {
+      "epoch": 0.23768151163497175,
+      "grad_norm": 0.1950039565563202,
+      "learning_rate": 8.695006649777854e-05,
+      "loss": 1.1129,
+      "step": 3912
+    },
+    {
+      "epoch": 0.23774226866759826,
+      "grad_norm": 0.14388521015644073,
+      "learning_rate": 8.694361660545116e-05,
+      "loss": 1.128,
+      "step": 3913
+    },
+    {
+      "epoch": 0.2378030257002248,
+      "grad_norm": 0.2002483457326889,
+      "learning_rate": 8.693716535894837e-05,
+      "loss": 1.0713,
+      "step": 3914
+    },
+    {
+      "epoch": 0.23786378273285133,
+      "grad_norm": 0.36832526326179504,
+      "learning_rate": 8.693071275850666e-05,
+      "loss": 1.1553,
+      "step": 3915
+    },
+    {
+      "epoch": 0.23792453976547787,
+      "grad_norm": 2.9972379207611084,
+      "learning_rate": 8.692425880436253e-05,
+      "loss": 1.1144,
+      "step": 3916
+    },
+    {
+      "epoch": 0.23798529679810437,
+      "grad_norm": 0.17029139399528503,
+      "learning_rate": 8.691780349675255e-05,
+      "loss": 1.0681,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2380460538307309,
+      "grad_norm": 0.6380923390388489,
+      "learning_rate": 8.691134683591334e-05,
+      "loss": 1.1038,
+      "step": 3918
+    },
+    {
+      "epoch": 0.23810681086335744,
+      "grad_norm": 0.19650183618068695,
+      "learning_rate": 8.690488882208159e-05,
+      "loss": 1.1262,
+      "step": 3919
+    },
+    {
+      "epoch": 0.23816756789598395,
+      "grad_norm": 0.19674339890480042,
+      "learning_rate": 8.6898429455494e-05,
+      "loss": 1.1344,
+      "step": 3920
+    },
+    {
+      "epoch": 0.23822832492861049,
+      "grad_norm": 0.1724308282136917,
+      "learning_rate": 8.689196873638734e-05,
+      "loss": 1.114,
+      "step": 3921
+    },
+    {
+      "epoch": 0.23828908196123702,
+      "grad_norm": 0.15377965569496155,
+      "learning_rate": 8.688550666499844e-05,
+      "loss": 1.0968,
+      "step": 3922
+    },
+    {
+      "epoch": 0.23834983899386353,
+      "grad_norm": 0.2071285992860794,
+      "learning_rate": 8.687904324156417e-05,
+      "loss": 1.1281,
+      "step": 3923
+    },
+    {
+      "epoch": 0.23841059602649006,
+      "grad_norm": 0.15540525317192078,
+      "learning_rate": 8.687257846632142e-05,
+      "loss": 1.0709,
+      "step": 3924
+    },
+    {
+      "epoch": 0.2384713530591166,
+      "grad_norm": 0.2519376277923584,
+      "learning_rate": 8.686611233950721e-05,
+      "loss": 1.1394,
+      "step": 3925
+    },
+    {
+      "epoch": 0.23853211009174313,
+      "grad_norm": 0.16704869270324707,
+      "learning_rate": 8.685964486135852e-05,
+      "loss": 1.0773,
+      "step": 3926
+    },
+    {
+      "epoch": 0.23859286712436964,
+      "grad_norm": 0.175711989402771,
+      "learning_rate": 8.68531760321124e-05,
+      "loss": 1.0482,
+      "step": 3927
+    },
+    {
+      "epoch": 0.23865362415699617,
+      "grad_norm": 0.21318276226520538,
+      "learning_rate": 8.684670585200601e-05,
+      "loss": 1.1116,
+      "step": 3928
+    },
+    {
+      "epoch": 0.2387143811896227,
+      "grad_norm": 0.15422223508358002,
+      "learning_rate": 8.684023432127648e-05,
+      "loss": 1.1867,
+      "step": 3929
+    },
+    {
+      "epoch": 0.23877513822224922,
+      "grad_norm": 0.19644825160503387,
+      "learning_rate": 8.683376144016106e-05,
+      "loss": 1.0807,
+      "step": 3930
+    },
+    {
+      "epoch": 0.23883589525487575,
+      "grad_norm": 0.14864318072795868,
+      "learning_rate": 8.682728720889697e-05,
+      "loss": 1.104,
+      "step": 3931
+    },
+    {
+      "epoch": 0.2388966522875023,
+      "grad_norm": 0.1880117654800415,
+      "learning_rate": 8.682081162772156e-05,
+      "loss": 1.0858,
+      "step": 3932
+    },
+    {
+      "epoch": 0.2389574093201288,
+      "grad_norm": 0.1523965448141098,
+      "learning_rate": 8.681433469687219e-05,
+      "loss": 1.1366,
+      "step": 3933
+    },
+    {
+      "epoch": 0.23901816635275533,
+      "grad_norm": 0.8888483643531799,
+      "learning_rate": 8.680785641658626e-05,
+      "loss": 1.111,
+      "step": 3934
+    },
+    {
+      "epoch": 0.23907892338538186,
+      "grad_norm": 0.23758070170879364,
+      "learning_rate": 8.680137678710125e-05,
+      "loss": 1.0727,
+      "step": 3935
+    },
+    {
+      "epoch": 0.23913968041800837,
+      "grad_norm": 0.19169849157333374,
+      "learning_rate": 8.679489580865463e-05,
+      "loss": 1.1219,
+      "step": 3936
+    },
+    {
+      "epoch": 0.2392004374506349,
+      "grad_norm": 0.20867425203323364,
+      "learning_rate": 8.6788413481484e-05,
+      "loss": 1.1069,
+      "step": 3937
+    },
+    {
+      "epoch": 0.23926119448326144,
+      "grad_norm": 0.1340804100036621,
+      "learning_rate": 8.678192980582697e-05,
+      "loss": 1.099,
+      "step": 3938
+    },
+    {
+      "epoch": 0.23932195151588798,
+      "grad_norm": 0.3417584002017975,
+      "learning_rate": 8.677544478192121e-05,
+      "loss": 1.1643,
+      "step": 3939
+    },
+    {
+      "epoch": 0.23938270854851448,
+      "grad_norm": 0.16733908653259277,
+      "learning_rate": 8.676895841000441e-05,
+      "loss": 1.0706,
+      "step": 3940
+    },
+    {
+      "epoch": 0.23944346558114102,
+      "grad_norm": 0.13786454498767853,
+      "learning_rate": 8.676247069031431e-05,
+      "loss": 1.0377,
+      "step": 3941
+    },
+    {
+      "epoch": 0.23950422261376755,
+      "grad_norm": 0.22302298247814178,
+      "learning_rate": 8.675598162308875e-05,
+      "loss": 1.2489,
+      "step": 3942
+    },
+    {
+      "epoch": 0.23956497964639406,
+      "grad_norm": 0.19295349717140198,
+      "learning_rate": 8.67494912085656e-05,
+      "loss": 1.1547,
+      "step": 3943
+    },
+    {
+      "epoch": 0.2396257366790206,
+      "grad_norm": 0.3624386489391327,
+      "learning_rate": 8.674299944698271e-05,
+      "loss": 1.0695,
+      "step": 3944
+    },
+    {
+      "epoch": 0.23968649371164713,
+      "grad_norm": 0.15758883953094482,
+      "learning_rate": 8.67365063385781e-05,
+      "loss": 1.1164,
+      "step": 3945
+    },
+    {
+      "epoch": 0.23974725074427364,
+      "grad_norm": 0.14956587553024292,
+      "learning_rate": 8.673001188358974e-05,
+      "loss": 1.0498,
+      "step": 3946
+    },
+    {
+      "epoch": 0.23980800777690017,
+      "grad_norm": 0.1457321047782898,
+      "learning_rate": 8.672351608225568e-05,
+      "loss": 1.1066,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2398687648095267,
+      "grad_norm": 0.724364161491394,
+      "learning_rate": 8.671701893481405e-05,
+      "loss": 1.1812,
+      "step": 3948
+    },
+    {
+      "epoch": 0.23992952184215324,
+      "grad_norm": 0.2393360584974289,
+      "learning_rate": 8.6710520441503e-05,
+      "loss": 1.1985,
+      "step": 3949
+    },
+    {
+      "epoch": 0.23999027887477975,
+      "grad_norm": 0.23643238842487335,
+      "learning_rate": 8.670402060256073e-05,
+      "loss": 1.1031,
+      "step": 3950
+    },
+    {
+      "epoch": 0.24005103590740628,
+      "grad_norm": 0.17582935094833374,
+      "learning_rate": 8.669751941822549e-05,
+      "loss": 1.0795,
+      "step": 3951
+    },
+    {
+      "epoch": 0.24011179294003282,
+      "grad_norm": 0.15575046837329865,
+      "learning_rate": 8.669101688873558e-05,
+      "loss": 1.0899,
+      "step": 3952
+    },
+    {
+      "epoch": 0.24017254997265933,
+      "grad_norm": 0.12399573624134064,
+      "learning_rate": 8.668451301432934e-05,
+      "loss": 1.0854,
+      "step": 3953
+    },
+    {
+      "epoch": 0.24023330700528586,
+      "grad_norm": 0.15120410919189453,
+      "learning_rate": 8.66780077952452e-05,
+      "loss": 1.113,
+      "step": 3954
+    },
+    {
+      "epoch": 0.2402940640379124,
+      "grad_norm": 3.1159369945526123,
+      "learning_rate": 8.667150123172158e-05,
+      "loss": 1.0955,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2403548210705389,
+      "grad_norm": 0.31155282258987427,
+      "learning_rate": 8.6664993323997e-05,
+      "loss": 1.2229,
+      "step": 3956
+    },
+    {
+      "epoch": 0.24041557810316544,
+      "grad_norm": 0.20258402824401855,
+      "learning_rate": 8.665848407231001e-05,
+      "loss": 1.0941,
+      "step": 3957
+    },
+    {
+      "epoch": 0.24047633513579197,
+      "grad_norm": 0.2539091408252716,
+      "learning_rate": 8.66519734768992e-05,
+      "loss": 1.1218,
+      "step": 3958
+    },
+    {
+      "epoch": 0.24053709216841848,
+      "grad_norm": 0.15777987241744995,
+      "learning_rate": 8.664546153800322e-05,
+      "loss": 1.1047,
+      "step": 3959
+    },
+    {
+      "epoch": 0.24059784920104502,
+      "grad_norm": 0.15768924355506897,
+      "learning_rate": 8.663894825586073e-05,
+      "loss": 1.1152,
+      "step": 3960
+    },
+    {
+      "epoch": 0.24065860623367155,
+      "grad_norm": 0.15072941780090332,
+      "learning_rate": 8.663243363071056e-05,
+      "loss": 1.1198,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2407193632662981,
+      "grad_norm": 0.18585556745529175,
+      "learning_rate": 8.662591766279142e-05,
+      "loss": 1.2109,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2407801202989246,
+      "grad_norm": 0.14197374880313873,
+      "learning_rate": 8.66194003523422e-05,
+      "loss": 1.0695,
+      "step": 3963
+    },
+    {
+      "epoch": 0.24084087733155113,
+      "grad_norm": 0.15779241919517517,
+      "learning_rate": 8.661288169960177e-05,
+      "loss": 1.1455,
+      "step": 3964
+    },
+    {
+      "epoch": 0.24090163436417766,
+      "grad_norm": 0.6295207142829895,
+      "learning_rate": 8.660636170480912e-05,
+      "loss": 1.1348,
+      "step": 3965
+    },
+    {
+      "epoch": 0.24096239139680417,
+      "grad_norm": 0.12053091078996658,
+      "learning_rate": 8.659984036820317e-05,
+      "loss": 1.0692,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2410231484294307,
+      "grad_norm": 0.1854286789894104,
+      "learning_rate": 8.659331769002301e-05,
+      "loss": 1.2086,
+      "step": 3967
+    },
+    {
+      "epoch": 0.24108390546205724,
+      "grad_norm": 0.5717991590499878,
+      "learning_rate": 8.65867936705077e-05,
+      "loss": 1.2266,
+      "step": 3968
+    },
+    {
+      "epoch": 0.24114466249468375,
+      "grad_norm": 0.19806058704853058,
+      "learning_rate": 8.658026830989642e-05,
+      "loss": 1.1149,
+      "step": 3969
+    },
+    {
+      "epoch": 0.24120541952731028,
+      "grad_norm": 0.24884095788002014,
+      "learning_rate": 8.657374160842833e-05,
+      "loss": 1.0465,
+      "step": 3970
+    },
+    {
+      "epoch": 0.24126617655993682,
+      "grad_norm": 0.19447007775306702,
+      "learning_rate": 8.656721356634267e-05,
+      "loss": 1.1684,
+      "step": 3971
+    },
+    {
+      "epoch": 0.24132693359256335,
+      "grad_norm": 0.28897160291671753,
+      "learning_rate": 8.656068418387873e-05,
+      "loss": 1.105,
+      "step": 3972
+    },
+    {
+      "epoch": 0.24138769062518986,
+      "grad_norm": 0.14791648089885712,
+      "learning_rate": 8.655415346127582e-05,
+      "loss": 1.1586,
+      "step": 3973
+    },
+    {
+      "epoch": 0.2414484476578164,
+      "grad_norm": 0.2585868537425995,
+      "learning_rate": 8.654762139877338e-05,
+      "loss": 1.0975,
+      "step": 3974
+    },
+    {
+      "epoch": 0.24150920469044293,
+      "grad_norm": 0.19113487005233765,
+      "learning_rate": 8.654108799661081e-05,
+      "loss": 1.0899,
+      "step": 3975
+    },
+    {
+      "epoch": 0.24156996172306944,
+      "grad_norm": 0.3172217905521393,
+      "learning_rate": 8.653455325502757e-05,
+      "loss": 1.0752,
+      "step": 3976
+    },
+    {
+      "epoch": 0.24163071875569597,
+      "grad_norm": 2.1464030742645264,
+      "learning_rate": 8.652801717426324e-05,
+      "loss": 1.1344,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2416914757883225,
+      "grad_norm": 0.27316781878471375,
+      "learning_rate": 8.652147975455738e-05,
+      "loss": 1.2896,
+      "step": 3978
+    },
+    {
+      "epoch": 0.24175223282094901,
+      "grad_norm": 0.30954229831695557,
+      "learning_rate": 8.651494099614963e-05,
+      "loss": 1.1448,
+      "step": 3979
+    },
+    {
+      "epoch": 0.24181298985357555,
+      "grad_norm": 0.14505885541439056,
+      "learning_rate": 8.650840089927964e-05,
+      "loss": 1.067,
+      "step": 3980
+    },
+    {
+      "epoch": 0.24187374688620208,
+      "grad_norm": 0.20532795786857605,
+      "learning_rate": 8.650185946418716e-05,
+      "loss": 1.1173,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2419345039188286,
+      "grad_norm": 0.43968433141708374,
+      "learning_rate": 8.649531669111198e-05,
+      "loss": 1.054,
+      "step": 3982
+    },
+    {
+      "epoch": 0.24199526095145513,
+      "grad_norm": 0.16732043027877808,
+      "learning_rate": 8.648877258029389e-05,
+      "loss": 1.0866,
+      "step": 3983
+    },
+    {
+      "epoch": 0.24205601798408166,
+      "grad_norm": 0.3106529116630554,
+      "learning_rate": 8.64822271319728e-05,
+      "loss": 1.161,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2421167750167082,
+      "grad_norm": 0.2534385621547699,
+      "learning_rate": 8.647568034638862e-05,
+      "loss": 1.1392,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2421775320493347,
+      "grad_norm": 0.1997934877872467,
+      "learning_rate": 8.646913222378134e-05,
+      "loss": 1.0776,
+      "step": 3986
+    },
+    {
+      "epoch": 0.24223828908196124,
+      "grad_norm": 0.24092431366443634,
+      "learning_rate": 8.646258276439096e-05,
+      "loss": 1.0539,
+      "step": 3987
+    },
+    {
+      "epoch": 0.24229904611458777,
+      "grad_norm": 0.3611429035663605,
+      "learning_rate": 8.645603196845756e-05,
+      "loss": 1.0804,
+      "step": 3988
+    },
+    {
+      "epoch": 0.24235980314721428,
+      "grad_norm": 0.18255358934402466,
+      "learning_rate": 8.644947983622125e-05,
+      "loss": 1.0668,
+      "step": 3989
+    },
+    {
+      "epoch": 0.24242056017984082,
+      "grad_norm": 0.1695895493030548,
+      "learning_rate": 8.644292636792221e-05,
+      "loss": 1.0798,
+      "step": 3990
+    },
+    {
+      "epoch": 0.24248131721246735,
+      "grad_norm": 0.13713617622852325,
+      "learning_rate": 8.64363715638007e-05,
+      "loss": 1.0878,
+      "step": 3991
+    },
+    {
+      "epoch": 0.24254207424509386,
+      "grad_norm": 0.1993829607963562,
+      "learning_rate": 8.64298154240969e-05,
+      "loss": 1.1505,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2426028312777204,
+      "grad_norm": 0.2562691271305084,
+      "learning_rate": 8.64232579490512e-05,
+      "loss": 1.1111,
+      "step": 3993
+    },
+    {
+      "epoch": 0.24266358831034693,
+      "grad_norm": 2.1048951148986816,
+      "learning_rate": 8.641669913890394e-05,
+      "loss": 1.1707,
+      "step": 3994
+    },
+    {
+      "epoch": 0.24272434534297344,
+      "grad_norm": 0.21180875599384308,
+      "learning_rate": 8.64101389938955e-05,
+      "loss": 1.0976,
+      "step": 3995
+    },
+    {
+      "epoch": 0.24278510237559997,
+      "grad_norm": 0.2694220244884491,
+      "learning_rate": 8.640357751426642e-05,
+      "loss": 1.1383,
+      "step": 3996
+    },
+    {
+      "epoch": 0.2428458594082265,
+      "grad_norm": 6.604887962341309,
+      "learning_rate": 8.639701470025714e-05,
+      "loss": 1.1034,
+      "step": 3997
+    },
+    {
+      "epoch": 0.24290661644085304,
+      "grad_norm": 0.5258581042289734,
+      "learning_rate": 8.639045055210827e-05,
+      "loss": 1.1047,
+      "step": 3998
+    },
+    {
+      "epoch": 0.24296737347347955,
+      "grad_norm": 0.18104997277259827,
+      "learning_rate": 8.638388507006039e-05,
+      "loss": 1.1544,
+      "step": 3999
+    },
+    {
+      "epoch": 0.24302813050610608,
+      "grad_norm": 0.21292586624622345,
+      "learning_rate": 8.637731825435418e-05,
+      "loss": 1.1266,
+      "step": 4000
+    },
+    {
+      "epoch": 0.24308888753873262,
+      "grad_norm": 0.3438781499862671,
+      "learning_rate": 8.637075010523032e-05,
+      "loss": 1.1126,
+      "step": 4001
+    },
+    {
+      "epoch": 0.24314964457135912,
+      "grad_norm": 0.14674068987369537,
+      "learning_rate": 8.636418062292962e-05,
+      "loss": 1.1072,
+      "step": 4002
+    },
+    {
+      "epoch": 0.24321040160398566,
+      "grad_norm": 0.38550257682800293,
+      "learning_rate": 8.635760980769281e-05,
+      "loss": 1.1461,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2432711586366122,
+      "grad_norm": 0.16236717998981476,
+      "learning_rate": 8.635103765976081e-05,
+      "loss": 1.0789,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2433319156692387,
+      "grad_norm": 0.5983213186264038,
+      "learning_rate": 8.634446417937449e-05,
+      "loss": 1.1615,
+      "step": 4005
+    },
+    {
+      "epoch": 0.24339267270186524,
+      "grad_norm": 0.3105098605155945,
+      "learning_rate": 8.63378893667748e-05,
+      "loss": 1.1373,
+      "step": 4006
+    },
+    {
+      "epoch": 0.24345342973449177,
+      "grad_norm": 0.18556448817253113,
+      "learning_rate": 8.633131322220276e-05,
+      "loss": 1.0749,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2435141867671183,
+      "grad_norm": 0.26110541820526123,
+      "learning_rate": 8.632473574589941e-05,
+      "loss": 1.0877,
+      "step": 4008
+    },
+    {
+      "epoch": 0.24357494379974481,
+      "grad_norm": 0.46966123580932617,
+      "learning_rate": 8.631815693810586e-05,
+      "loss": 1.1334,
+      "step": 4009
+    },
+    {
+      "epoch": 0.24363570083237135,
+      "grad_norm": 0.23228362202644348,
+      "learning_rate": 8.631157679906323e-05,
+      "loss": 1.1264,
+      "step": 4010
+    },
+    {
+      "epoch": 0.24369645786499788,
+      "grad_norm": 0.2246306836605072,
+      "learning_rate": 8.630499532901275e-05,
+      "loss": 1.157,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2437572148976244,
+      "grad_norm": 0.24068474769592285,
+      "learning_rate": 8.629841252819564e-05,
+      "loss": 1.1571,
+      "step": 4012
+    },
+    {
+      "epoch": 0.24381797193025093,
+      "grad_norm": 0.2753877639770508,
+      "learning_rate": 8.629182839685321e-05,
+      "loss": 1.1175,
+      "step": 4013
+    },
+    {
+      "epoch": 0.24387872896287746,
+      "grad_norm": 0.19838157296180725,
+      "learning_rate": 8.628524293522679e-05,
+      "loss": 1.1656,
+      "step": 4014
+    },
+    {
+      "epoch": 0.24393948599550397,
+      "grad_norm": 0.24955883622169495,
+      "learning_rate": 8.627865614355777e-05,
+      "loss": 1.0274,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2440002430281305,
+      "grad_norm": 0.1909952610731125,
+      "learning_rate": 8.62720680220876e-05,
+      "loss": 1.1304,
+      "step": 4016
+    },
+    {
+      "epoch": 0.24406100006075704,
+      "grad_norm": 0.1548011600971222,
+      "learning_rate": 8.626547857105777e-05,
+      "loss": 1.0564,
+      "step": 4017
+    },
+    {
+      "epoch": 0.24412175709338355,
+      "grad_norm": 0.6794400811195374,
+      "learning_rate": 8.62588877907098e-05,
+      "loss": 1.144,
+      "step": 4018
+    },
+    {
+      "epoch": 0.24418251412601008,
+      "grad_norm": 0.2033773809671402,
+      "learning_rate": 8.625229568128531e-05,
+      "loss": 1.2283,
+      "step": 4019
+    },
+    {
+      "epoch": 0.24424327115863662,
+      "grad_norm": 0.19142000377178192,
+      "learning_rate": 8.624570224302591e-05,
+      "loss": 1.0659,
+      "step": 4020
+    },
+    {
+      "epoch": 0.24430402819126315,
+      "grad_norm": 0.20253171026706696,
+      "learning_rate": 8.623910747617328e-05,
+      "loss": 1.0685,
+      "step": 4021
+    },
+    {
+      "epoch": 0.24436478522388966,
+      "grad_norm": 0.15959037840366364,
+      "learning_rate": 8.623251138096916e-05,
+      "loss": 1.1282,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2444255422565162,
+      "grad_norm": 0.2258574664592743,
+      "learning_rate": 8.622591395765533e-05,
+      "loss": 1.1014,
+      "step": 4023
+    },
+    {
+      "epoch": 0.24448629928914273,
+      "grad_norm": 0.3244194984436035,
+      "learning_rate": 8.621931520647363e-05,
+      "loss": 1.1262,
+      "step": 4024
+    },
+    {
+      "epoch": 0.24454705632176924,
+      "grad_norm": 0.22501277923583984,
+      "learning_rate": 8.621271512766593e-05,
+      "loss": 1.2468,
+      "step": 4025
+    },
+    {
+      "epoch": 0.24460781335439577,
+      "grad_norm": 0.3266246020793915,
+      "learning_rate": 8.620611372147414e-05,
+      "loss": 1.0979,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2446685703870223,
+      "grad_norm": 0.1458095908164978,
+      "learning_rate": 8.619951098814028e-05,
+      "loss": 1.0966,
+      "step": 4027
+    },
+    {
+      "epoch": 0.2447293274196488,
+      "grad_norm": 0.7091784477233887,
+      "learning_rate": 8.619290692790633e-05,
+      "loss": 1.1805,
+      "step": 4028
+    },
+    {
+      "epoch": 0.24479008445227535,
+      "grad_norm": 0.3023465573787689,
+      "learning_rate": 8.618630154101438e-05,
+      "loss": 1.1671,
+      "step": 4029
+    },
+    {
+      "epoch": 0.24485084148490188,
+      "grad_norm": 0.3327714204788208,
+      "learning_rate": 8.617969482770656e-05,
+      "loss": 1.0544,
+      "step": 4030
+    },
+    {
+      "epoch": 0.24491159851752842,
+      "grad_norm": 0.3605053424835205,
+      "learning_rate": 8.617308678822502e-05,
+      "loss": 1.1492,
+      "step": 4031
+    },
+    {
+      "epoch": 0.24497235555015492,
+      "grad_norm": 18.277729034423828,
+      "learning_rate": 8.6166477422812e-05,
+      "loss": 1.127,
+      "step": 4032
+    },
+    {
+      "epoch": 0.24503311258278146,
+      "grad_norm": 0.34724563360214233,
+      "learning_rate": 8.615986673170977e-05,
+      "loss": 1.0869,
+      "step": 4033
+    },
+    {
+      "epoch": 0.245093869615408,
+      "grad_norm": 0.23811745643615723,
+      "learning_rate": 8.615325471516063e-05,
+      "loss": 1.1345,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2451546266480345,
+      "grad_norm": 0.2629840075969696,
+      "learning_rate": 8.614664137340694e-05,
+      "loss": 1.0795,
+      "step": 4035
+    },
+    {
+      "epoch": 0.24521538368066104,
+      "grad_norm": 0.5882347822189331,
+      "learning_rate": 8.614002670669114e-05,
+      "loss": 1.3391,
+      "step": 4036
+    },
+    {
+      "epoch": 0.24527614071328757,
+      "grad_norm": 0.17361603677272797,
+      "learning_rate": 8.613341071525565e-05,
+      "loss": 1.1735,
+      "step": 4037
+    },
+    {
+      "epoch": 0.24533689774591408,
+      "grad_norm": 0.17132027447223663,
+      "learning_rate": 8.612679339934302e-05,
+      "loss": 1.0885,
+      "step": 4038
+    },
+    {
+      "epoch": 0.2453976547785406,
+      "grad_norm": 0.31367743015289307,
+      "learning_rate": 8.61201747591958e-05,
+      "loss": 1.2087,
+      "step": 4039
+    },
+    {
+      "epoch": 0.24545841181116715,
+      "grad_norm": 0.30873334407806396,
+      "learning_rate": 8.611355479505658e-05,
+      "loss": 1.0547,
+      "step": 4040
+    },
+    {
+      "epoch": 0.24551916884379366,
+      "grad_norm": 0.2688838541507721,
+      "learning_rate": 8.610693350716806e-05,
+      "loss": 1.1403,
+      "step": 4041
+    },
+    {
+      "epoch": 0.2455799258764202,
+      "grad_norm": 0.23757126927375793,
+      "learning_rate": 8.610031089577289e-05,
+      "loss": 1.0852,
+      "step": 4042
+    },
+    {
+      "epoch": 0.24564068290904673,
+      "grad_norm": 0.16408149898052216,
+      "learning_rate": 8.609368696111383e-05,
+      "loss": 1.0152,
+      "step": 4043
+    },
+    {
+      "epoch": 0.24570143994167326,
+      "grad_norm": 0.14532369375228882,
+      "learning_rate": 8.608706170343371e-05,
+      "loss": 1.0695,
+      "step": 4044
+    },
+    {
+      "epoch": 0.24576219697429977,
+      "grad_norm": 0.2933334410190582,
+      "learning_rate": 8.608043512297537e-05,
+      "loss": 1.1332,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2458229540069263,
+      "grad_norm": 0.2835923433303833,
+      "learning_rate": 8.607380721998171e-05,
+      "loss": 1.0596,
+      "step": 4046
+    },
+    {
+      "epoch": 0.24588371103955284,
+      "grad_norm": 0.16383066773414612,
+      "learning_rate": 8.606717799469568e-05,
+      "loss": 1.1277,
+      "step": 4047
+    },
+    {
+      "epoch": 0.24594446807217935,
+      "grad_norm": 0.32419484853744507,
+      "learning_rate": 8.606054744736026e-05,
+      "loss": 1.1274,
+      "step": 4048
+    },
+    {
+      "epoch": 0.24600522510480588,
+      "grad_norm": 0.24372728168964386,
+      "learning_rate": 8.605391557821849e-05,
+      "loss": 1.1013,
+      "step": 4049
+    },
+    {
+      "epoch": 0.24606598213743242,
+      "grad_norm": 0.4960094094276428,
+      "learning_rate": 8.604728238751349e-05,
+      "loss": 1.2317,
+      "step": 4050
+    },
+    {
+      "epoch": 0.24612673917005892,
+      "grad_norm": 0.17118974030017853,
+      "learning_rate": 8.604064787548839e-05,
+      "loss": 1.1164,
+      "step": 4051
+    },
+    {
+      "epoch": 0.24618749620268546,
+      "grad_norm": 0.20355871319770813,
+      "learning_rate": 8.603401204238637e-05,
+      "loss": 1.1536,
+      "step": 4052
+    },
+    {
+      "epoch": 0.246248253235312,
+      "grad_norm": 0.1677030324935913,
+      "learning_rate": 8.602737488845067e-05,
+      "loss": 1.1402,
+      "step": 4053
+    },
+    {
+      "epoch": 0.24630901026793853,
+      "grad_norm": 0.1634184718132019,
+      "learning_rate": 8.602073641392458e-05,
+      "loss": 1.0338,
+      "step": 4054
+    },
+    {
+      "epoch": 0.24636976730056503,
+      "grad_norm": 0.19681808352470398,
+      "learning_rate": 8.601409661905144e-05,
+      "loss": 1.1602,
+      "step": 4055
+    },
+    {
+      "epoch": 0.24643052433319157,
+      "grad_norm": 0.17885880172252655,
+      "learning_rate": 8.600745550407464e-05,
+      "loss": 1.1191,
+      "step": 4056
+    },
+    {
+      "epoch": 0.2464912813658181,
+      "grad_norm": 0.20643024146556854,
+      "learning_rate": 8.600081306923758e-05,
+      "loss": 1.1407,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2465520383984446,
+      "grad_norm": 0.20853880047798157,
+      "learning_rate": 8.599416931478374e-05,
+      "loss": 1.2485,
+      "step": 4058
+    },
+    {
+      "epoch": 0.24661279543107115,
+      "grad_norm": 0.15868984162807465,
+      "learning_rate": 8.598752424095669e-05,
+      "loss": 1.0705,
+      "step": 4059
+    },
+    {
+      "epoch": 0.24667355246369768,
+      "grad_norm": 0.8027739524841309,
+      "learning_rate": 8.598087784799998e-05,
+      "loss": 1.2232,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2467343094963242,
+      "grad_norm": 0.13661803305149078,
+      "learning_rate": 8.597423013615724e-05,
+      "loss": 1.0464,
+      "step": 4061
+    },
+    {
+      "epoch": 0.24679506652895072,
+      "grad_norm": 0.25538215041160583,
+      "learning_rate": 8.596758110567214e-05,
+      "loss": 1.164,
+      "step": 4062
+    },
+    {
+      "epoch": 0.24685582356157726,
+      "grad_norm": 0.2539287507534027,
+      "learning_rate": 8.59609307567884e-05,
+      "loss": 1.0747,
+      "step": 4063
+    },
+    {
+      "epoch": 0.24691658059420377,
+      "grad_norm": 0.7665238976478577,
+      "learning_rate": 8.595427908974979e-05,
+      "loss": 1.1589,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2469773376268303,
+      "grad_norm": 0.19996581971645355,
+      "learning_rate": 8.594762610480012e-05,
+      "loss": 1.2087,
+      "step": 4065
+    },
+    {
+      "epoch": 0.24703809465945684,
+      "grad_norm": 5.7554497718811035,
+      "learning_rate": 8.594097180218329e-05,
+      "loss": 1.1282,
+      "step": 4066
+    },
+    {
+      "epoch": 0.24709885169208337,
+      "grad_norm": 0.2446056306362152,
+      "learning_rate": 8.593431618214318e-05,
+      "loss": 1.2048,
+      "step": 4067
+    },
+    {
+      "epoch": 0.24715960872470988,
+      "grad_norm": 0.1952815055847168,
+      "learning_rate": 8.592765924492377e-05,
+      "loss": 1.1804,
+      "step": 4068
+    },
+    {
+      "epoch": 0.2472203657573364,
+      "grad_norm": 0.20773836970329285,
+      "learning_rate": 8.592100099076905e-05,
+      "loss": 1.0441,
+      "step": 4069
+    },
+    {
+      "epoch": 0.24728112278996295,
+      "grad_norm": 0.4322708547115326,
+      "learning_rate": 8.59143414199231e-05,
+      "loss": 1.2189,
+      "step": 4070
+    },
+    {
+      "epoch": 0.24734187982258946,
+      "grad_norm": 0.18040430545806885,
+      "learning_rate": 8.590768053263005e-05,
+      "loss": 1.1488,
+      "step": 4071
+    },
+    {
+      "epoch": 0.247402636855216,
+      "grad_norm": 0.28984418511390686,
+      "learning_rate": 8.5901018329134e-05,
+      "loss": 1.1822,
+      "step": 4072
+    },
+    {
+      "epoch": 0.24746339388784253,
+      "grad_norm": 0.24308879673480988,
+      "learning_rate": 8.589435480967918e-05,
+      "loss": 1.2023,
+      "step": 4073
+    },
+    {
+      "epoch": 0.24752415092046903,
+      "grad_norm": 0.1683763712644577,
+      "learning_rate": 8.588768997450987e-05,
+      "loss": 1.2014,
+      "step": 4074
+    },
+    {
+      "epoch": 0.24758490795309557,
+      "grad_norm": 0.2185237556695938,
+      "learning_rate": 8.588102382387033e-05,
+      "loss": 1.1085,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2476456649857221,
+      "grad_norm": 0.3711228668689728,
+      "learning_rate": 8.587435635800494e-05,
+      "loss": 1.1659,
+      "step": 4076
+    },
+    {
+      "epoch": 0.24770642201834864,
+      "grad_norm": 0.14320625364780426,
+      "learning_rate": 8.586768757715806e-05,
+      "loss": 1.0488,
+      "step": 4077
+    },
+    {
+      "epoch": 0.24776717905097514,
+      "grad_norm": 0.21578729152679443,
+      "learning_rate": 8.586101748157418e-05,
+      "loss": 1.0628,
+      "step": 4078
+    },
+    {
+      "epoch": 0.24782793608360168,
+      "grad_norm": 0.15633291006088257,
+      "learning_rate": 8.585434607149776e-05,
+      "loss": 1.0938,
+      "step": 4079
+    },
+    {
+      "epoch": 0.24788869311622821,
+      "grad_norm": 0.18531322479248047,
+      "learning_rate": 8.584767334717337e-05,
+      "loss": 1.0921,
+      "step": 4080
+    },
+    {
+      "epoch": 0.24794945014885472,
+      "grad_norm": 0.23115776479244232,
+      "learning_rate": 8.584099930884557e-05,
+      "loss": 1.135,
+      "step": 4081
+    },
+    {
+      "epoch": 0.24801020718148126,
+      "grad_norm": 0.14599020779132843,
+      "learning_rate": 8.583432395675901e-05,
+      "loss": 1.0658,
+      "step": 4082
+    },
+    {
+      "epoch": 0.2480709642141078,
+      "grad_norm": 0.14802511036396027,
+      "learning_rate": 8.582764729115838e-05,
+      "loss": 1.0776,
+      "step": 4083
+    },
+    {
+      "epoch": 0.2481317212467343,
+      "grad_norm": 0.20576101541519165,
+      "learning_rate": 8.582096931228842e-05,
+      "loss": 1.2338,
+      "step": 4084
+    },
+    {
+      "epoch": 0.24819247827936083,
+      "grad_norm": 0.238702192902565,
+      "learning_rate": 8.581429002039392e-05,
+      "loss": 1.2124,
+      "step": 4085
+    },
+    {
+      "epoch": 0.24825323531198737,
+      "grad_norm": 0.2632155418395996,
+      "learning_rate": 8.580760941571967e-05,
+      "loss": 1.1338,
+      "step": 4086
+    },
+    {
+      "epoch": 0.24831399234461388,
+      "grad_norm": 0.18265683948993683,
+      "learning_rate": 8.580092749851058e-05,
+      "loss": 1.1111,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2483747493772404,
+      "grad_norm": 0.1798429787158966,
+      "learning_rate": 8.579424426901159e-05,
+      "loss": 1.1069,
+      "step": 4088
+    },
+    {
+      "epoch": 0.24843550640986695,
+      "grad_norm": 1.7813425064086914,
+      "learning_rate": 8.578755972746763e-05,
+      "loss": 1.244,
+      "step": 4089
+    },
+    {
+      "epoch": 0.24849626344249348,
+      "grad_norm": 0.19594089686870575,
+      "learning_rate": 8.578087387412377e-05,
+      "loss": 1.0355,
+      "step": 4090
+    },
+    {
+      "epoch": 0.24855702047512,
+      "grad_norm": 0.18351732194423676,
+      "learning_rate": 8.577418670922506e-05,
+      "loss": 1.1217,
+      "step": 4091
+    },
+    {
+      "epoch": 0.24861777750774652,
+      "grad_norm": 0.15555503964424133,
+      "learning_rate": 8.576749823301662e-05,
+      "loss": 1.1005,
+      "step": 4092
+    },
+    {
+      "epoch": 0.24867853454037306,
+      "grad_norm": 0.15773619711399078,
+      "learning_rate": 8.576080844574362e-05,
+      "loss": 1.1117,
+      "step": 4093
+    },
+    {
+      "epoch": 0.24873929157299957,
+      "grad_norm": 0.24765901267528534,
+      "learning_rate": 8.575411734765128e-05,
+      "loss": 1.0443,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2488000486056261,
+      "grad_norm": 0.16648131608963013,
+      "learning_rate": 8.574742493898484e-05,
+      "loss": 1.0965,
+      "step": 4095
+    },
+    {
+      "epoch": 0.24886080563825264,
+      "grad_norm": 0.14474792778491974,
+      "learning_rate": 8.574073121998964e-05,
+      "loss": 1.1223,
+      "step": 4096
+    },
+    {
+      "epoch": 0.24892156267087914,
+      "grad_norm": 0.1604004204273224,
+      "learning_rate": 8.573403619091103e-05,
+      "loss": 1.0575,
+      "step": 4097
+    },
+    {
+      "epoch": 0.24898231970350568,
+      "grad_norm": 0.2689473330974579,
+      "learning_rate": 8.572733985199442e-05,
+      "loss": 1.1346,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2490430767361322,
+      "grad_norm": 0.16532671451568604,
+      "learning_rate": 8.572064220348525e-05,
+      "loss": 1.1929,
+      "step": 4099
+    },
+    {
+      "epoch": 0.24910383376875872,
+      "grad_norm": 0.12528428435325623,
+      "learning_rate": 8.571394324562905e-05,
+      "loss": 1.0594,
+      "step": 4100
+    },
+    {
+      "epoch": 0.24916459080138526,
+      "grad_norm": 0.316062331199646,
+      "learning_rate": 8.570724297867135e-05,
+      "loss": 1.0969,
+      "step": 4101
+    },
+    {
+      "epoch": 0.2492253478340118,
+      "grad_norm": 0.17280779778957367,
+      "learning_rate": 8.570054140285775e-05,
+      "loss": 1.1598,
+      "step": 4102
+    },
+    {
+      "epoch": 0.24928610486663833,
+      "grad_norm": 0.23015287518501282,
+      "learning_rate": 8.56938385184339e-05,
+      "loss": 1.2757,
+      "step": 4103
+    },
+    {
+      "epoch": 0.24934686189926483,
+      "grad_norm": 0.32703980803489685,
+      "learning_rate": 8.568713432564552e-05,
+      "loss": 1.103,
+      "step": 4104
+    },
+    {
+      "epoch": 0.24940761893189137,
+      "grad_norm": 0.24935807287693024,
+      "learning_rate": 8.56804288247383e-05,
+      "loss": 1.0897,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2494683759645179,
+      "grad_norm": 0.24203459918498993,
+      "learning_rate": 8.56737220159581e-05,
+      "loss": 1.1474,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2495291329971444,
+      "grad_norm": 0.1566619873046875,
+      "learning_rate": 8.56670138995507e-05,
+      "loss": 1.126,
+      "step": 4107
+    },
+    {
+      "epoch": 0.24958989002977094,
+      "grad_norm": 0.26541823148727417,
+      "learning_rate": 8.566030447576203e-05,
+      "loss": 1.1483,
+      "step": 4108
+    },
+    {
+      "epoch": 0.24965064706239748,
+      "grad_norm": 0.289373517036438,
+      "learning_rate": 8.565359374483799e-05,
+      "loss": 1.2045,
+      "step": 4109
+    },
+    {
+      "epoch": 0.249711404095024,
+      "grad_norm": 0.17153102159500122,
+      "learning_rate": 8.56468817070246e-05,
+      "loss": 1.1377,
+      "step": 4110
+    },
+    {
+      "epoch": 0.24977216112765052,
+      "grad_norm": 0.28802037239074707,
+      "learning_rate": 8.564016836256785e-05,
+      "loss": 1.2679,
+      "step": 4111
+    },
+    {
+      "epoch": 0.24983291816027706,
+      "grad_norm": 1.233233094215393,
+      "learning_rate": 8.563345371171385e-05,
+      "loss": 1.1001,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2498936751929036,
+      "grad_norm": 0.38556793332099915,
+      "learning_rate": 8.56267377547087e-05,
+      "loss": 1.0648,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2499544322255301,
+      "grad_norm": 0.19476401805877686,
+      "learning_rate": 8.562002049179862e-05,
+      "loss": 1.1667,
+      "step": 4114
+    },
+    {
+      "epoch": 0.25001518925815663,
+      "grad_norm": 0.19244134426116943,
+      "learning_rate": 8.561330192322978e-05,
+      "loss": 1.1302,
+      "step": 4115
+    },
+    {
+      "epoch": 0.25007594629078317,
+      "grad_norm": 0.4519463777542114,
+      "learning_rate": 8.560658204924849e-05,
+      "loss": 1.0912,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2501367033234097,
+      "grad_norm": 0.24016918241977692,
+      "learning_rate": 8.559986087010104e-05,
+      "loss": 1.1388,
+      "step": 4117
+    },
+    {
+      "epoch": 0.25019746035603624,
+      "grad_norm": 0.3761630952358246,
+      "learning_rate": 8.559313838603381e-05,
+      "loss": 1.2035,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2502582173886627,
+      "grad_norm": 0.3559887707233429,
+      "learning_rate": 8.558641459729321e-05,
+      "loss": 1.131,
+      "step": 4119
+    },
+    {
+      "epoch": 0.25031897442128925,
+      "grad_norm": 0.20240308344364166,
+      "learning_rate": 8.557968950412571e-05,
+      "loss": 1.0828,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2503797314539158,
+      "grad_norm": 0.2807188332080841,
+      "learning_rate": 8.557296310677783e-05,
+      "loss": 1.1277,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2504404884865423,
+      "grad_norm": 0.29699328541755676,
+      "learning_rate": 8.556623540549608e-05,
+      "loss": 1.039,
+      "step": 4122
+    },
+    {
+      "epoch": 0.25050124551916886,
+      "grad_norm": 0.22494173049926758,
+      "learning_rate": 8.555950640052713e-05,
+      "loss": 1.0894,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2505620025517954,
+      "grad_norm": 0.3755808472633362,
+      "learning_rate": 8.555277609211757e-05,
+      "loss": 1.0686,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2506227595844219,
+      "grad_norm": 0.28935179114341736,
+      "learning_rate": 8.554604448051413e-05,
+      "loss": 1.0538,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2506835166170484,
+      "grad_norm": 0.32570749521255493,
+      "learning_rate": 8.553931156596358e-05,
+      "loss": 1.2517,
+      "step": 4126
+    },
+    {
+      "epoch": 0.25074427364967494,
+      "grad_norm": 0.41436418890953064,
+      "learning_rate": 8.553257734871267e-05,
+      "loss": 1.0494,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2508050306823015,
+      "grad_norm": 0.23903585970401764,
+      "learning_rate": 8.55258418290083e-05,
+      "loss": 1.0858,
+      "step": 4128
+    },
+    {
+      "epoch": 0.250865787714928,
+      "grad_norm": 0.29696860909461975,
+      "learning_rate": 8.551910500709732e-05,
+      "loss": 1.1465,
+      "step": 4129
+    },
+    {
+      "epoch": 0.25092654474755455,
+      "grad_norm": 0.3306981027126312,
+      "learning_rate": 8.551236688322666e-05,
+      "loss": 1.2771,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2509873017801811,
+      "grad_norm": 2.7766425609588623,
+      "learning_rate": 8.550562745764334e-05,
+      "loss": 1.0992,
+      "step": 4131
+    },
+    {
+      "epoch": 0.25104805881280756,
+      "grad_norm": 0.20471929013729095,
+      "learning_rate": 8.549888673059439e-05,
+      "loss": 1.0946,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2511088158454341,
+      "grad_norm": 0.22637993097305298,
+      "learning_rate": 8.549214470232688e-05,
+      "loss": 1.1798,
+      "step": 4133
+    },
+    {
+      "epoch": 0.25116957287806063,
+      "grad_norm": 0.19550451636314392,
+      "learning_rate": 8.548540137308795e-05,
+      "loss": 1.1265,
+      "step": 4134
+    },
+    {
+      "epoch": 0.25123032991068717,
+      "grad_norm": 0.2085788995027542,
+      "learning_rate": 8.547865674312476e-05,
+      "loss": 1.1003,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2512910869433137,
+      "grad_norm": 0.1719086915254593,
+      "learning_rate": 8.547191081268456e-05,
+      "loss": 1.1077,
+      "step": 4136
+    },
+    {
+      "epoch": 0.25135184397594024,
+      "grad_norm": 0.21202941238880157,
+      "learning_rate": 8.546516358201462e-05,
+      "loss": 1.109,
+      "step": 4137
+    },
+    {
+      "epoch": 0.2514126010085667,
+      "grad_norm": 0.3908483386039734,
+      "learning_rate": 8.545841505136224e-05,
+      "loss": 1.3424,
+      "step": 4138
+    },
+    {
+      "epoch": 0.25147335804119325,
+      "grad_norm": 0.192022442817688,
+      "learning_rate": 8.545166522097481e-05,
+      "loss": 1.1281,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2515341150738198,
+      "grad_norm": 0.20203399658203125,
+      "learning_rate": 8.544491409109973e-05,
+      "loss": 1.1444,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2515948721064463,
+      "grad_norm": 0.19474977254867554,
+      "learning_rate": 8.543816166198449e-05,
+      "loss": 1.1229,
+      "step": 4141
+    },
+    {
+      "epoch": 0.25165562913907286,
+      "grad_norm": 6.34874963760376,
+      "learning_rate": 8.543140793387657e-05,
+      "loss": 1.1531,
+      "step": 4142
+    },
+    {
+      "epoch": 0.2517163861716994,
+      "grad_norm": 1.0578254461288452,
+      "learning_rate": 8.542465290702353e-05,
+      "loss": 1.1249,
+      "step": 4143
+    },
+    {
+      "epoch": 0.2517771432043259,
+      "grad_norm": 0.3454793393611908,
+      "learning_rate": 8.541789658167301e-05,
+      "loss": 1.0144,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2518379002369524,
+      "grad_norm": 0.244012251496315,
+      "learning_rate": 8.541113895807264e-05,
+      "loss": 1.1771,
+      "step": 4145
+    },
+    {
+      "epoch": 0.25189865726957894,
+      "grad_norm": 0.3643401265144348,
+      "learning_rate": 8.540438003647013e-05,
+      "loss": 1.0516,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2519594143022055,
+      "grad_norm": 0.1746654510498047,
+      "learning_rate": 8.539761981711321e-05,
+      "loss": 1.0609,
+      "step": 4147
+    },
+    {
+      "epoch": 0.252020171334832,
+      "grad_norm": 0.17908631265163422,
+      "learning_rate": 8.539085830024971e-05,
+      "loss": 1.166,
+      "step": 4148
+    },
+    {
+      "epoch": 0.25208092836745855,
+      "grad_norm": 0.29916009306907654,
+      "learning_rate": 8.538409548612746e-05,
+      "loss": 1.2271,
+      "step": 4149
+    },
+    {
+      "epoch": 0.2521416854000851,
+      "grad_norm": 0.17699405550956726,
+      "learning_rate": 8.537733137499435e-05,
+      "loss": 1.1746,
+      "step": 4150
+    },
+    {
+      "epoch": 0.25220244243271156,
+      "grad_norm": 0.45046988129615784,
+      "learning_rate": 8.537056596709832e-05,
+      "loss": 1.1261,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2522631994653381,
+      "grad_norm": 0.13706566393375397,
+      "learning_rate": 8.536379926268734e-05,
+      "loss": 1.0986,
+      "step": 4152
+    },
+    {
+      "epoch": 0.25232395649796463,
+      "grad_norm": 3.5723299980163574,
+      "learning_rate": 8.535703126200948e-05,
+      "loss": 1.105,
+      "step": 4153
+    },
+    {
+      "epoch": 0.25238471353059116,
+      "grad_norm": 0.255502849817276,
+      "learning_rate": 8.535026196531279e-05,
+      "loss": 1.137,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2524454705632177,
+      "grad_norm": 0.5019230246543884,
+      "learning_rate": 8.534349137284543e-05,
+      "loss": 1.2151,
+      "step": 4155
+    },
+    {
+      "epoch": 0.25250622759584423,
+      "grad_norm": 0.15695320069789886,
+      "learning_rate": 8.533671948485555e-05,
+      "loss": 1.0984,
+      "step": 4156
+    },
+    {
+      "epoch": 0.25256698462847077,
+      "grad_norm": 0.23317253589630127,
+      "learning_rate": 8.53299463015914e-05,
+      "loss": 1.0763,
+      "step": 4157
+    },
+    {
+      "epoch": 0.25262774166109725,
+      "grad_norm": 0.1939711719751358,
+      "learning_rate": 8.532317182330123e-05,
+      "loss": 1.2722,
+      "step": 4158
+    },
+    {
+      "epoch": 0.2526884986937238,
+      "grad_norm": 0.2504059374332428,
+      "learning_rate": 8.531639605023339e-05,
+      "loss": 1.1656,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2527492557263503,
+      "grad_norm": 0.1749381721019745,
+      "learning_rate": 8.53096189826362e-05,
+      "loss": 1.0923,
+      "step": 4160
+    },
+    {
+      "epoch": 0.25281001275897685,
+      "grad_norm": 0.21789798140525818,
+      "learning_rate": 8.53028406207581e-05,
+      "loss": 1.1331,
+      "step": 4161
+    },
+    {
+      "epoch": 0.2528707697916034,
+      "grad_norm": 0.17451786994934082,
+      "learning_rate": 8.529606096484757e-05,
+      "loss": 1.0753,
+      "step": 4162
+    },
+    {
+      "epoch": 0.2529315268242299,
+      "grad_norm": 0.2880437672138214,
+      "learning_rate": 8.528928001515309e-05,
+      "loss": 1.1936,
+      "step": 4163
+    },
+    {
+      "epoch": 0.2529922838568564,
+      "grad_norm": 0.15425370633602142,
+      "learning_rate": 8.528249777192324e-05,
+      "loss": 1.1206,
+      "step": 4164
+    },
+    {
+      "epoch": 0.25305304088948294,
+      "grad_norm": 0.14943325519561768,
+      "learning_rate": 8.527571423540662e-05,
+      "loss": 1.057,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2531137979221095,
+      "grad_norm": 0.17268097400665283,
+      "learning_rate": 8.526892940585185e-05,
+      "loss": 1.2054,
+      "step": 4166
+    },
+    {
+      "epoch": 0.253174554954736,
+      "grad_norm": 3.0992114543914795,
+      "learning_rate": 8.526214328350767e-05,
+      "loss": 1.065,
+      "step": 4167
+    },
+    {
+      "epoch": 0.25323531198736254,
+      "grad_norm": 0.2546354830265045,
+      "learning_rate": 8.525535586862281e-05,
+      "loss": 1.1614,
+      "step": 4168
+    },
+    {
+      "epoch": 0.2532960690199891,
+      "grad_norm": 0.16679732501506805,
+      "learning_rate": 8.524856716144607e-05,
+      "loss": 1.0882,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2533568260526156,
+      "grad_norm": 0.19214554131031036,
+      "learning_rate": 8.524177716222628e-05,
+      "loss": 1.1004,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2534175830852421,
+      "grad_norm": 0.18400248885154724,
+      "learning_rate": 8.523498587121235e-05,
+      "loss": 1.0916,
+      "step": 4171
+    },
+    {
+      "epoch": 0.25347834011786863,
+      "grad_norm": 0.2818697392940521,
+      "learning_rate": 8.52281932886532e-05,
+      "loss": 1.2159,
+      "step": 4172
+    },
+    {
+      "epoch": 0.25353909715049516,
+      "grad_norm": 0.41420456767082214,
+      "learning_rate": 8.52213994147978e-05,
+      "loss": 1.1639,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2535998541831217,
+      "grad_norm": 1.186070203781128,
+      "learning_rate": 8.52146042498952e-05,
+      "loss": 1.1032,
+      "step": 4174
+    },
+    {
+      "epoch": 0.25366061121574823,
+      "grad_norm": 0.34048500657081604,
+      "learning_rate": 8.520780779419448e-05,
+      "loss": 1.0767,
+      "step": 4175
+    },
+    {
+      "epoch": 0.25372136824837477,
+      "grad_norm": 0.21824349462985992,
+      "learning_rate": 8.520101004794476e-05,
+      "loss": 1.2362,
+      "step": 4176
+    },
+    {
+      "epoch": 0.2537821252810013,
+      "grad_norm": 0.31136640906333923,
+      "learning_rate": 8.519421101139522e-05,
+      "loss": 1.1252,
+      "step": 4177
+    },
+    {
+      "epoch": 0.2538428823136278,
+      "grad_norm": 0.17634399235248566,
+      "learning_rate": 8.518741068479506e-05,
+      "loss": 1.1509,
+      "step": 4178
+    },
+    {
+      "epoch": 0.2539036393462543,
+      "grad_norm": 0.1874326914548874,
+      "learning_rate": 8.518060906839356e-05,
+      "loss": 1.1848,
+      "step": 4179
+    },
+    {
+      "epoch": 0.25396439637888085,
+      "grad_norm": 0.2158823311328888,
+      "learning_rate": 8.517380616244005e-05,
+      "loss": 1.1532,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2540251534115074,
+      "grad_norm": 0.21650990843772888,
+      "learning_rate": 8.516700196718385e-05,
+      "loss": 1.1425,
+      "step": 4181
+    },
+    {
+      "epoch": 0.2540859104441339,
+      "grad_norm": 0.27733394503593445,
+      "learning_rate": 8.516019648287441e-05,
+      "loss": 1.116,
+      "step": 4182
+    },
+    {
+      "epoch": 0.25414666747676046,
+      "grad_norm": 4.468677043914795,
+      "learning_rate": 8.515338970976117e-05,
+      "loss": 1.1007,
+      "step": 4183
+    },
+    {
+      "epoch": 0.25420742450938694,
+      "grad_norm": 0.22979959845542908,
+      "learning_rate": 8.514658164809364e-05,
+      "loss": 1.0845,
+      "step": 4184
+    },
+    {
+      "epoch": 0.25426818154201347,
+      "grad_norm": 0.22904656827449799,
+      "learning_rate": 8.513977229812135e-05,
+      "loss": 1.0828,
+      "step": 4185
+    },
+    {
+      "epoch": 0.25432893857464,
+      "grad_norm": 0.20850655436515808,
+      "learning_rate": 8.513296166009395e-05,
+      "loss": 1.0891,
+      "step": 4186
+    },
+    {
+      "epoch": 0.25438969560726654,
+      "grad_norm": 0.1685955971479416,
+      "learning_rate": 8.512614973426101e-05,
+      "loss": 1.1043,
+      "step": 4187
+    },
+    {
+      "epoch": 0.2544504526398931,
+      "grad_norm": 0.2577073276042938,
+      "learning_rate": 8.511933652087227e-05,
+      "loss": 1.099,
+      "step": 4188
+    },
+    {
+      "epoch": 0.2545112096725196,
+      "grad_norm": 0.7813034653663635,
+      "learning_rate": 8.511252202017747e-05,
+      "loss": 1.1036,
+      "step": 4189
+    },
+    {
+      "epoch": 0.25457196670514615,
+      "grad_norm": 0.18157364428043365,
+      "learning_rate": 8.51057062324264e-05,
+      "loss": 1.2152,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2546327237377726,
+      "grad_norm": 0.23822636902332306,
+      "learning_rate": 8.509888915786886e-05,
+      "loss": 1.1606,
+      "step": 4191
+    },
+    {
+      "epoch": 0.25469348077039916,
+      "grad_norm": 0.24963602423667908,
+      "learning_rate": 8.509207079675478e-05,
+      "loss": 1.2555,
+      "step": 4192
+    },
+    {
+      "epoch": 0.2547542378030257,
+      "grad_norm": 0.16131234169006348,
+      "learning_rate": 8.508525114933404e-05,
+      "loss": 1.0895,
+      "step": 4193
+    },
+    {
+      "epoch": 0.25481499483565223,
+      "grad_norm": 0.16613486409187317,
+      "learning_rate": 8.507843021585665e-05,
+      "loss": 1.0926,
+      "step": 4194
+    },
+    {
+      "epoch": 0.25487575186827877,
+      "grad_norm": 0.24539700150489807,
+      "learning_rate": 8.50716079965726e-05,
+      "loss": 1.0889,
+      "step": 4195
+    },
+    {
+      "epoch": 0.2549365089009053,
+      "grad_norm": 0.16711640357971191,
+      "learning_rate": 8.5064784491732e-05,
+      "loss": 1.0738,
+      "step": 4196
+    },
+    {
+      "epoch": 0.2549972659335318,
+      "grad_norm": 0.20085102319717407,
+      "learning_rate": 8.505795970158493e-05,
+      "loss": 1.0931,
+      "step": 4197
+    },
+    {
+      "epoch": 0.2550580229661583,
+      "grad_norm": 0.2539522647857666,
+      "learning_rate": 8.505113362638158e-05,
+      "loss": 1.1715,
+      "step": 4198
+    },
+    {
+      "epoch": 0.25511877999878485,
+      "grad_norm": 0.39260950684547424,
+      "learning_rate": 8.504430626637215e-05,
+      "loss": 1.1744,
+      "step": 4199
+    },
+    {
+      "epoch": 0.2551795370314114,
+      "grad_norm": 0.27809226512908936,
+      "learning_rate": 8.503747762180691e-05,
+      "loss": 1.1154,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2552402940640379,
+      "grad_norm": 0.21992912888526917,
+      "learning_rate": 8.503064769293614e-05,
+      "loss": 1.1283,
+      "step": 4201
+    },
+    {
+      "epoch": 0.25530105109666446,
+      "grad_norm": 0.26056337356567383,
+      "learning_rate": 8.50238164800102e-05,
+      "loss": 1.1155,
+      "step": 4202
+    },
+    {
+      "epoch": 0.255361808129291,
+      "grad_norm": 0.15038877725601196,
+      "learning_rate": 8.501698398327951e-05,
+      "loss": 1.0914,
+      "step": 4203
+    },
+    {
+      "epoch": 0.25542256516191747,
+      "grad_norm": 0.24054546654224396,
+      "learning_rate": 8.501015020299452e-05,
+      "loss": 1.3729,
+      "step": 4204
+    },
+    {
+      "epoch": 0.255483322194544,
+      "grad_norm": 0.17988722026348114,
+      "learning_rate": 8.500331513940569e-05,
+      "loss": 1.1602,
+      "step": 4205
+    },
+    {
+      "epoch": 0.25554407922717054,
+      "grad_norm": 0.12434462457895279,
+      "learning_rate": 8.499647879276359e-05,
+      "loss": 1.0818,
+      "step": 4206
+    },
+    {
+      "epoch": 0.2556048362597971,
+      "grad_norm": 0.18362249433994293,
+      "learning_rate": 8.498964116331876e-05,
+      "loss": 1.1345,
+      "step": 4207
+    },
+    {
+      "epoch": 0.2556655932924236,
+      "grad_norm": 0.22588826715946198,
+      "learning_rate": 8.49828022513219e-05,
+      "loss": 1.2381,
+      "step": 4208
+    },
+    {
+      "epoch": 0.25572635032505014,
+      "grad_norm": 0.1283007562160492,
+      "learning_rate": 8.497596205702366e-05,
+      "loss": 1.1278,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2557871073576766,
+      "grad_norm": 0.9142860174179077,
+      "learning_rate": 8.496912058067476e-05,
+      "loss": 1.2001,
+      "step": 4210
+    },
+    {
+      "epoch": 0.25584786439030316,
+      "grad_norm": 0.17875459790229797,
+      "learning_rate": 8.496227782252599e-05,
+      "loss": 1.0761,
+      "step": 4211
+    },
+    {
+      "epoch": 0.2559086214229297,
+      "grad_norm": 0.12321389466524124,
+      "learning_rate": 8.495543378282818e-05,
+      "loss": 1.0366,
+      "step": 4212
+    },
+    {
+      "epoch": 0.25596937845555623,
+      "grad_norm": 0.1925307661294937,
+      "learning_rate": 8.494858846183217e-05,
+      "loss": 1.0533,
+      "step": 4213
+    },
+    {
+      "epoch": 0.25603013548818276,
+      "grad_norm": 0.868889331817627,
+      "learning_rate": 8.494174185978889e-05,
+      "loss": 1.1255,
+      "step": 4214
+    },
+    {
+      "epoch": 0.2560908925208093,
+      "grad_norm": 0.7886120080947876,
+      "learning_rate": 8.493489397694932e-05,
+      "loss": 1.1666,
+      "step": 4215
+    },
+    {
+      "epoch": 0.25615164955343583,
+      "grad_norm": 0.2437831610441208,
+      "learning_rate": 8.492804481356446e-05,
+      "loss": 1.2113,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2562124065860623,
+      "grad_norm": 0.255470871925354,
+      "learning_rate": 8.492119436988538e-05,
+      "loss": 1.0551,
+      "step": 4217
+    },
+    {
+      "epoch": 0.25627316361868885,
+      "grad_norm": 0.240932896733284,
+      "learning_rate": 8.491434264616313e-05,
+      "loss": 1.1391,
+      "step": 4218
+    },
+    {
+      "epoch": 0.2563339206513154,
+      "grad_norm": 0.220208078622818,
+      "learning_rate": 8.490748964264894e-05,
+      "loss": 1.1684,
+      "step": 4219
+    },
+    {
+      "epoch": 0.2563946776839419,
+      "grad_norm": 0.21479731798171997,
+      "learning_rate": 8.490063535959396e-05,
+      "loss": 1.1756,
+      "step": 4220
+    },
+    {
+      "epoch": 0.25645543471656845,
+      "grad_norm": 0.38608306646347046,
+      "learning_rate": 8.489377979724944e-05,
+      "loss": 1.1017,
+      "step": 4221
+    },
+    {
+      "epoch": 0.256516191749195,
+      "grad_norm": 0.2426665723323822,
+      "learning_rate": 8.488692295586668e-05,
+      "loss": 1.0529,
+      "step": 4222
+    },
+    {
+      "epoch": 0.2565769487818215,
+      "grad_norm": 0.1743011623620987,
+      "learning_rate": 8.4880064835697e-05,
+      "loss": 1.0883,
+      "step": 4223
+    },
+    {
+      "epoch": 0.256637705814448,
+      "grad_norm": 0.8394467234611511,
+      "learning_rate": 8.487320543699183e-05,
+      "loss": 1.1242,
+      "step": 4224
+    },
+    {
+      "epoch": 0.25669846284707454,
+      "grad_norm": 0.21643461287021637,
+      "learning_rate": 8.486634476000255e-05,
+      "loss": 1.2507,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2567592198797011,
+      "grad_norm": 0.14840495586395264,
+      "learning_rate": 8.485948280498067e-05,
+      "loss": 1.0298,
+      "step": 4226
+    },
+    {
+      "epoch": 0.2568199769123276,
+      "grad_norm": 0.20398449897766113,
+      "learning_rate": 8.485261957217771e-05,
+      "loss": 1.1437,
+      "step": 4227
+    },
+    {
+      "epoch": 0.25688073394495414,
+      "grad_norm": 0.19567550718784332,
+      "learning_rate": 8.484575506184525e-05,
+      "loss": 1.1123,
+      "step": 4228
+    },
+    {
+      "epoch": 0.2569414909775807,
+      "grad_norm": 0.27100253105163574,
+      "learning_rate": 8.483888927423489e-05,
+      "loss": 1.093,
+      "step": 4229
+    },
+    {
+      "epoch": 0.25700224801020716,
+      "grad_norm": 0.1475774496793747,
+      "learning_rate": 8.483202220959831e-05,
+      "loss": 1.0509,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2570630050428337,
+      "grad_norm": 0.13975857198238373,
+      "learning_rate": 8.482515386818724e-05,
+      "loss": 1.0791,
+      "step": 4231
+    },
+    {
+      "epoch": 0.2571237620754602,
+      "grad_norm": 0.8387301564216614,
+      "learning_rate": 8.48182842502534e-05,
+      "loss": 1.1257,
+      "step": 4232
+    },
+    {
+      "epoch": 0.25718451910808676,
+      "grad_norm": 0.14440058171749115,
+      "learning_rate": 8.481141335604863e-05,
+      "loss": 1.0842,
+      "step": 4233
+    },
+    {
+      "epoch": 0.2572452761407133,
+      "grad_norm": 0.16440021991729736,
+      "learning_rate": 8.480454118582479e-05,
+      "loss": 1.0561,
+      "step": 4234
+    },
+    {
+      "epoch": 0.25730603317333983,
+      "grad_norm": 1.0560884475708008,
+      "learning_rate": 8.479766773983374e-05,
+      "loss": 1.0607,
+      "step": 4235
+    },
+    {
+      "epoch": 0.25736679020596637,
+      "grad_norm": 0.4002532958984375,
+      "learning_rate": 8.479079301832746e-05,
+      "loss": 1.1793,
+      "step": 4236
+    },
+    {
+      "epoch": 0.25742754723859285,
+      "grad_norm": 0.1388968974351883,
+      "learning_rate": 8.478391702155795e-05,
+      "loss": 1.1066,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2574883042712194,
+      "grad_norm": 0.31794053316116333,
+      "learning_rate": 8.477703974977722e-05,
+      "loss": 1.1935,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2575490613038459,
+      "grad_norm": 0.1266797035932541,
+      "learning_rate": 8.47701612032374e-05,
+      "loss": 1.0229,
+      "step": 4239
+    },
+    {
+      "epoch": 0.25760981833647245,
+      "grad_norm": 1.874619483947754,
+      "learning_rate": 8.476328138219058e-05,
+      "loss": 1.1506,
+      "step": 4240
+    },
+    {
+      "epoch": 0.257670575369099,
+      "grad_norm": 0.40473389625549316,
+      "learning_rate": 8.475640028688896e-05,
+      "loss": 1.2329,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2577313324017255,
+      "grad_norm": 0.13560648262500763,
+      "learning_rate": 8.474951791758478e-05,
+      "loss": 1.0643,
+      "step": 4242
+    },
+    {
+      "epoch": 0.257792089434352,
+      "grad_norm": 0.16973525285720825,
+      "learning_rate": 8.474263427453031e-05,
+      "loss": 1.1239,
+      "step": 4243
+    },
+    {
+      "epoch": 0.25785284646697854,
+      "grad_norm": 0.24408873915672302,
+      "learning_rate": 8.473574935797786e-05,
+      "loss": 1.2423,
+      "step": 4244
+    },
+    {
+      "epoch": 0.25791360349960507,
+      "grad_norm": 0.2804044187068939,
+      "learning_rate": 8.472886316817979e-05,
+      "loss": 1.1492,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2579743605322316,
+      "grad_norm": 0.2863284945487976,
+      "learning_rate": 8.472197570538853e-05,
+      "loss": 1.0617,
+      "step": 4246
+    },
+    {
+      "epoch": 0.25803511756485814,
+      "grad_norm": 0.17607618868350983,
+      "learning_rate": 8.471508696985655e-05,
+      "loss": 1.0959,
+      "step": 4247
+    },
+    {
+      "epoch": 0.2580958745974847,
+      "grad_norm": 0.23290187120437622,
+      "learning_rate": 8.470819696183634e-05,
+      "loss": 1.2404,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2581566316301112,
+      "grad_norm": 0.4581441879272461,
+      "learning_rate": 8.470130568158047e-05,
+      "loss": 1.1571,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2582173886627377,
+      "grad_norm": 0.16368238627910614,
+      "learning_rate": 8.469441312934153e-05,
+      "loss": 1.098,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2582781456953642,
+      "grad_norm": 0.18559230864048004,
+      "learning_rate": 8.468751930537215e-05,
+      "loss": 1.0555,
+      "step": 4251
+    },
+    {
+      "epoch": 0.25833890272799076,
+      "grad_norm": 0.24541161954402924,
+      "learning_rate": 8.468062420992506e-05,
+      "loss": 1.1215,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2583996597606173,
+      "grad_norm": 0.17333202064037323,
+      "learning_rate": 8.467372784325299e-05,
+      "loss": 1.1694,
+      "step": 4253
+    },
+    {
+      "epoch": 0.25846041679324383,
+      "grad_norm": 0.2998686730861664,
+      "learning_rate": 8.466683020560873e-05,
+      "loss": 1.1827,
+      "step": 4254
+    },
+    {
+      "epoch": 0.25852117382587037,
+      "grad_norm": 0.23007234930992126,
+      "learning_rate": 8.465993129724508e-05,
+      "loss": 1.1592,
+      "step": 4255
+    },
+    {
+      "epoch": 0.25858193085849684,
+      "grad_norm": 0.20353084802627563,
+      "learning_rate": 8.465303111841498e-05,
+      "loss": 1.2069,
+      "step": 4256
+    },
+    {
+      "epoch": 0.2586426878911234,
+      "grad_norm": 0.47833624482154846,
+      "learning_rate": 8.464612966937131e-05,
+      "loss": 1.2244,
+      "step": 4257
+    },
+    {
+      "epoch": 0.2587034449237499,
+      "grad_norm": 0.16635839641094208,
+      "learning_rate": 8.463922695036706e-05,
+      "loss": 1.0911,
+      "step": 4258
+    },
+    {
+      "epoch": 0.25876420195637645,
+      "grad_norm": 0.639962911605835,
+      "learning_rate": 8.463232296165526e-05,
+      "loss": 1.0513,
+      "step": 4259
+    },
+    {
+      "epoch": 0.258824958989003,
+      "grad_norm": 0.32631582021713257,
+      "learning_rate": 8.462541770348895e-05,
+      "loss": 1.2023,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2588857160216295,
+      "grad_norm": 4.5949225425720215,
+      "learning_rate": 8.461851117612127e-05,
+      "loss": 1.0874,
+      "step": 4261
+    },
+    {
+      "epoch": 0.25894647305425605,
+      "grad_norm": 0.2057335525751114,
+      "learning_rate": 8.461160337980538e-05,
+      "loss": 1.1289,
+      "step": 4262
+    },
+    {
+      "epoch": 0.25900723008688253,
+      "grad_norm": 0.32290101051330566,
+      "learning_rate": 8.460469431479448e-05,
+      "loss": 1.125,
+      "step": 4263
+    },
+    {
+      "epoch": 0.25906798711950907,
+      "grad_norm": 0.16816361248493195,
+      "learning_rate": 8.45977839813418e-05,
+      "loss": 1.083,
+      "step": 4264
+    },
+    {
+      "epoch": 0.2591287441521356,
+      "grad_norm": 0.20611028373241425,
+      "learning_rate": 8.459087237970068e-05,
+      "loss": 1.1246,
+      "step": 4265
+    },
+    {
+      "epoch": 0.25918950118476214,
+      "grad_norm": 0.18841104209423065,
+      "learning_rate": 8.458395951012444e-05,
+      "loss": 1.1281,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2592502582173887,
+      "grad_norm": 0.8322115540504456,
+      "learning_rate": 8.457704537286647e-05,
+      "loss": 1.1157,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2593110152500152,
+      "grad_norm": 0.20937062799930573,
+      "learning_rate": 8.457012996818023e-05,
+      "loss": 1.1715,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2593717722826417,
+      "grad_norm": 0.34700995683670044,
+      "learning_rate": 8.45632132963192e-05,
+      "loss": 1.1975,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2594325293152682,
+      "grad_norm": 0.44231733679771423,
+      "learning_rate": 8.45562953575369e-05,
+      "loss": 1.0327,
+      "step": 4270
+    },
+    {
+      "epoch": 0.25949328634789476,
+      "grad_norm": 0.2505379319190979,
+      "learning_rate": 8.454937615208692e-05,
+      "loss": 1.1351,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2595540433805213,
+      "grad_norm": 0.3006500005722046,
+      "learning_rate": 8.454245568022285e-05,
+      "loss": 1.0673,
+      "step": 4272
+    },
+    {
+      "epoch": 0.25961480041314783,
+      "grad_norm": 0.19378769397735596,
+      "learning_rate": 8.453553394219841e-05,
+      "loss": 1.1147,
+      "step": 4273
+    },
+    {
+      "epoch": 0.25967555744577436,
+      "grad_norm": 0.20149783790111542,
+      "learning_rate": 8.45286109382673e-05,
+      "loss": 1.1276,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2597363144784009,
+      "grad_norm": 0.2344742715358734,
+      "learning_rate": 8.452168666868327e-05,
+      "loss": 1.096,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2597970715110274,
+      "grad_norm": 0.23637834191322327,
+      "learning_rate": 8.451476113370016e-05,
+      "loss": 1.0784,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2598578285436539,
+      "grad_norm": 0.4054419696331024,
+      "learning_rate": 8.450783433357182e-05,
+      "loss": 1.1277,
+      "step": 4277
+    },
+    {
+      "epoch": 0.25991858557628045,
+      "grad_norm": 0.18842460215091705,
+      "learning_rate": 8.45009062685521e-05,
+      "loss": 1.1831,
+      "step": 4278
+    },
+    {
+      "epoch": 0.259979342608907,
+      "grad_norm": 0.5069484710693359,
+      "learning_rate": 8.449397693889502e-05,
+      "loss": 1.1123,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2600400996415335,
+      "grad_norm": 0.25176990032196045,
+      "learning_rate": 8.448704634485454e-05,
+      "loss": 1.1092,
+      "step": 4280
+    },
+    {
+      "epoch": 0.26010085667416005,
+      "grad_norm": 0.13377034664154053,
+      "learning_rate": 8.448011448668472e-05,
+      "loss": 1.0943,
+      "step": 4281
+    },
+    {
+      "epoch": 0.2601616137067866,
+      "grad_norm": 0.9178459048271179,
+      "learning_rate": 8.447318136463962e-05,
+      "loss": 1.1304,
+      "step": 4282
+    },
+    {
+      "epoch": 0.26022237073941307,
+      "grad_norm": 0.20416206121444702,
+      "learning_rate": 8.446624697897342e-05,
+      "loss": 1.077,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2602831277720396,
+      "grad_norm": 0.1249048113822937,
+      "learning_rate": 8.445931132994025e-05,
+      "loss": 1.0849,
+      "step": 4284
+    },
+    {
+      "epoch": 0.26034388480466614,
+      "grad_norm": 1.0604158639907837,
+      "learning_rate": 8.445237441779437e-05,
+      "loss": 1.1111,
+      "step": 4285
+    },
+    {
+      "epoch": 0.26040464183729267,
+      "grad_norm": 0.20574629306793213,
+      "learning_rate": 8.444543624279006e-05,
+      "loss": 1.0639,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2604653988699192,
+      "grad_norm": 0.27051597833633423,
+      "learning_rate": 8.443849680518161e-05,
+      "loss": 1.0105,
+      "step": 4287
+    },
+    {
+      "epoch": 0.26052615590254574,
+      "grad_norm": 0.16592645645141602,
+      "learning_rate": 8.44315561052234e-05,
+      "loss": 1.1072,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2605869129351722,
+      "grad_norm": 0.22462166845798492,
+      "learning_rate": 8.442461414316986e-05,
+      "loss": 1.1015,
+      "step": 4289
+    },
+    {
+      "epoch": 0.26064766996779876,
+      "grad_norm": 0.3104848861694336,
+      "learning_rate": 8.441767091927543e-05,
+      "loss": 1.093,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2607084270004253,
+      "grad_norm": 0.9585015177726746,
+      "learning_rate": 8.441072643379461e-05,
+      "loss": 1.1744,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2607691840330518,
+      "grad_norm": 0.3315548896789551,
+      "learning_rate": 8.440378068698196e-05,
+      "loss": 1.1473,
+      "step": 4292
+    },
+    {
+      "epoch": 0.26082994106567836,
+      "grad_norm": 0.2847521901130676,
+      "learning_rate": 8.43968336790921e-05,
+      "loss": 1.1992,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2608906980983049,
+      "grad_norm": 0.20233559608459473,
+      "learning_rate": 8.438988541037964e-05,
+      "loss": 1.2571,
+      "step": 4294
+    },
+    {
+      "epoch": 0.26095145513093143,
+      "grad_norm": 0.2187298834323883,
+      "learning_rate": 8.438293588109928e-05,
+      "loss": 1.3668,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2610122121635579,
+      "grad_norm": 0.26863613724708557,
+      "learning_rate": 8.437598509150575e-05,
+      "loss": 1.0784,
+      "step": 4296
+    },
+    {
+      "epoch": 0.26107296919618445,
+      "grad_norm": 0.1813523918390274,
+      "learning_rate": 8.436903304185386e-05,
+      "loss": 1.0299,
+      "step": 4297
+    },
+    {
+      "epoch": 0.261133726228811,
+      "grad_norm": 0.18379293382167816,
+      "learning_rate": 8.436207973239842e-05,
+      "loss": 1.088,
+      "step": 4298
+    },
+    {
+      "epoch": 0.2611944832614375,
+      "grad_norm": 0.19964168965816498,
+      "learning_rate": 8.43551251633943e-05,
+      "loss": 1.0851,
+      "step": 4299
+    },
+    {
+      "epoch": 0.26125524029406405,
+      "grad_norm": 0.21964845061302185,
+      "learning_rate": 8.434816933509641e-05,
+      "loss": 1.1641,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2613159973266906,
+      "grad_norm": 0.18196550011634827,
+      "learning_rate": 8.434121224775977e-05,
+      "loss": 1.0875,
+      "step": 4301
+    },
+    {
+      "epoch": 0.26137675435931707,
+      "grad_norm": 0.2147371768951416,
+      "learning_rate": 8.433425390163932e-05,
+      "loss": 1.1425,
+      "step": 4302
+    },
+    {
+      "epoch": 0.2614375113919436,
+      "grad_norm": 0.22804512083530426,
+      "learning_rate": 8.432729429699017e-05,
+      "loss": 1.2779,
+      "step": 4303
+    },
+    {
+      "epoch": 0.26149826842457013,
+      "grad_norm": 0.17448338866233826,
+      "learning_rate": 8.432033343406742e-05,
+      "loss": 1.1232,
+      "step": 4304
+    },
+    {
+      "epoch": 0.26155902545719667,
+      "grad_norm": 0.1834874004125595,
+      "learning_rate": 8.431337131312622e-05,
+      "loss": 1.0752,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2616197824898232,
+      "grad_norm": 0.14007149636745453,
+      "learning_rate": 8.430640793442175e-05,
+      "loss": 1.0383,
+      "step": 4306
+    },
+    {
+      "epoch": 0.26168053952244974,
+      "grad_norm": 0.1365254521369934,
+      "learning_rate": 8.429944329820927e-05,
+      "loss": 1.0877,
+      "step": 4307
+    },
+    {
+      "epoch": 0.2617412965550763,
+      "grad_norm": 2.537677049636841,
+      "learning_rate": 8.429247740474406e-05,
+      "loss": 1.1223,
+      "step": 4308
+    },
+    {
+      "epoch": 0.26180205358770275,
+      "grad_norm": 0.18881793320178986,
+      "learning_rate": 8.428551025428146e-05,
+      "loss": 1.1266,
+      "step": 4309
+    },
+    {
+      "epoch": 0.2618628106203293,
+      "grad_norm": 0.1634659767150879,
+      "learning_rate": 8.427854184707686e-05,
+      "loss": 1.0816,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2619235676529558,
+      "grad_norm": 0.1854061633348465,
+      "learning_rate": 8.427157218338568e-05,
+      "loss": 1.199,
+      "step": 4311
+    },
+    {
+      "epoch": 0.26198432468558236,
+      "grad_norm": 0.3584587872028351,
+      "learning_rate": 8.426460126346341e-05,
+      "loss": 1.2258,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2620450817182089,
+      "grad_norm": 0.20669493079185486,
+      "learning_rate": 8.425762908756553e-05,
+      "loss": 1.0801,
+      "step": 4313
+    },
+    {
+      "epoch": 0.26210583875083543,
+      "grad_norm": 0.23939336836338043,
+      "learning_rate": 8.425065565594767e-05,
+      "loss": 1.0658,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2621665957834619,
+      "grad_norm": 0.2513979375362396,
+      "learning_rate": 8.424368096886538e-05,
+      "loss": 1.1918,
+      "step": 4315
+    },
+    {
+      "epoch": 0.26222735281608844,
+      "grad_norm": 0.15272289514541626,
+      "learning_rate": 8.423670502657435e-05,
+      "loss": 1.0491,
+      "step": 4316
+    },
+    {
+      "epoch": 0.262288109848715,
+      "grad_norm": 0.20569448173046112,
+      "learning_rate": 8.422972782933029e-05,
+      "loss": 1.2418,
+      "step": 4317
+    },
+    {
+      "epoch": 0.2623488668813415,
+      "grad_norm": 0.17268654704093933,
+      "learning_rate": 8.422274937738893e-05,
+      "loss": 1.0728,
+      "step": 4318
+    },
+    {
+      "epoch": 0.26240962391396805,
+      "grad_norm": 0.12413746118545532,
+      "learning_rate": 8.421576967100607e-05,
+      "loss": 1.0836,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2624703809465946,
+      "grad_norm": 0.19145981967449188,
+      "learning_rate": 8.420878871043757e-05,
+      "loss": 1.1253,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2625311379792211,
+      "grad_norm": 0.1925724893808365,
+      "learning_rate": 8.420180649593929e-05,
+      "loss": 1.1103,
+      "step": 4321
+    },
+    {
+      "epoch": 0.2625918950118476,
+      "grad_norm": 0.2260303944349289,
+      "learning_rate": 8.419482302776721e-05,
+      "loss": 1.1659,
+      "step": 4322
+    },
+    {
+      "epoch": 0.26265265204447413,
+      "grad_norm": 0.1830778419971466,
+      "learning_rate": 8.418783830617727e-05,
+      "loss": 1.1641,
+      "step": 4323
+    },
+    {
+      "epoch": 0.26271340907710067,
+      "grad_norm": 0.1560674011707306,
+      "learning_rate": 8.418085233142551e-05,
+      "loss": 1.0825,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2627741661097272,
+      "grad_norm": 0.18006204068660736,
+      "learning_rate": 8.4173865103768e-05,
+      "loss": 1.0678,
+      "step": 4325
+    },
+    {
+      "epoch": 0.26283492314235374,
+      "grad_norm": 0.35525408387184143,
+      "learning_rate": 8.416687662346085e-05,
+      "loss": 1.1184,
+      "step": 4326
+    },
+    {
+      "epoch": 0.2628956801749803,
+      "grad_norm": 0.1623912751674652,
+      "learning_rate": 8.415988689076025e-05,
+      "loss": 1.1097,
+      "step": 4327
+    },
+    {
+      "epoch": 0.2629564372076068,
+      "grad_norm": 0.11738666146993637,
+      "learning_rate": 8.415289590592237e-05,
+      "loss": 1.1073,
+      "step": 4328
+    },
+    {
+      "epoch": 0.2630171942402333,
+      "grad_norm": 0.24086801707744598,
+      "learning_rate": 8.41459036692035e-05,
+      "loss": 1.0626,
+      "step": 4329
+    },
+    {
+      "epoch": 0.2630779512728598,
+      "grad_norm": 0.22589251399040222,
+      "learning_rate": 8.413891018085993e-05,
+      "loss": 1.174,
+      "step": 4330
+    },
+    {
+      "epoch": 0.26313870830548636,
+      "grad_norm": 0.14131517708301544,
+      "learning_rate": 8.413191544114803e-05,
+      "loss": 1.0931,
+      "step": 4331
+    },
+    {
+      "epoch": 0.2631994653381129,
+      "grad_norm": 0.22869080305099487,
+      "learning_rate": 8.412491945032414e-05,
+      "loss": 1.0693,
+      "step": 4332
+    },
+    {
+      "epoch": 0.2632602223707394,
+      "grad_norm": 0.2333141565322876,
+      "learning_rate": 8.411792220864474e-05,
+      "loss": 1.0574,
+      "step": 4333
+    },
+    {
+      "epoch": 0.26332097940336596,
+      "grad_norm": 0.1286310851573944,
+      "learning_rate": 8.41109237163663e-05,
+      "loss": 1.104,
+      "step": 4334
+    },
+    {
+      "epoch": 0.26338173643599244,
+      "grad_norm": 0.19540579617023468,
+      "learning_rate": 8.410392397374538e-05,
+      "loss": 1.0828,
+      "step": 4335
+    },
+    {
+      "epoch": 0.263442493468619,
+      "grad_norm": 0.11774860322475433,
+      "learning_rate": 8.409692298103852e-05,
+      "loss": 1.0604,
+      "step": 4336
+    },
+    {
+      "epoch": 0.2635032505012455,
+      "grad_norm": 0.24366101622581482,
+      "learning_rate": 8.408992073850237e-05,
+      "loss": 1.1778,
+      "step": 4337
+    },
+    {
+      "epoch": 0.26356400753387205,
+      "grad_norm": 0.17126290500164032,
+      "learning_rate": 8.408291724639357e-05,
+      "loss": 1.1507,
+      "step": 4338
+    },
+    {
+      "epoch": 0.2636247645664986,
+      "grad_norm": 0.1923706829547882,
+      "learning_rate": 8.407591250496885e-05,
+      "loss": 1.0667,
+      "step": 4339
+    },
+    {
+      "epoch": 0.2636855215991251,
+      "grad_norm": 0.1331157386302948,
+      "learning_rate": 8.406890651448499e-05,
+      "loss": 1.0635,
+      "step": 4340
+    },
+    {
+      "epoch": 0.26374627863175165,
+      "grad_norm": 0.1537851095199585,
+      "learning_rate": 8.406189927519876e-05,
+      "loss": 1.178,
+      "step": 4341
+    },
+    {
+      "epoch": 0.26380703566437813,
+      "grad_norm": 0.17169001698493958,
+      "learning_rate": 8.405489078736704e-05,
+      "loss": 1.1051,
+      "step": 4342
+    },
+    {
+      "epoch": 0.26386779269700467,
+      "grad_norm": 0.1729595959186554,
+      "learning_rate": 8.40478810512467e-05,
+      "loss": 1.0372,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2639285497296312,
+      "grad_norm": 0.28310850262641907,
+      "learning_rate": 8.404087006709471e-05,
+      "loss": 1.1145,
+      "step": 4344
+    },
+    {
+      "epoch": 0.26398930676225774,
+      "grad_norm": 0.23747852444648743,
+      "learning_rate": 8.403385783516804e-05,
+      "loss": 1.2195,
+      "step": 4345
+    },
+    {
+      "epoch": 0.26405006379488427,
+      "grad_norm": 0.8016255497932434,
+      "learning_rate": 8.402684435572374e-05,
+      "loss": 1.1465,
+      "step": 4346
+    },
+    {
+      "epoch": 0.2641108208275108,
+      "grad_norm": 0.134220153093338,
+      "learning_rate": 8.401982962901887e-05,
+      "loss": 1.0676,
+      "step": 4347
+    },
+    {
+      "epoch": 0.2641715778601373,
+      "grad_norm": 0.273553729057312,
+      "learning_rate": 8.401281365531058e-05,
+      "loss": 1.0634,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2642323348927638,
+      "grad_norm": 2.3341445922851562,
+      "learning_rate": 8.400579643485602e-05,
+      "loss": 1.1043,
+      "step": 4349
+    },
+    {
+      "epoch": 0.26429309192539036,
+      "grad_norm": 0.21785476803779602,
+      "learning_rate": 8.399877796791244e-05,
+      "loss": 1.2355,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2643538489580169,
+      "grad_norm": 0.1721855103969574,
+      "learning_rate": 8.399175825473706e-05,
+      "loss": 1.1752,
+      "step": 4351
+    },
+    {
+      "epoch": 0.2644146059906434,
+      "grad_norm": 0.3808407783508301,
+      "learning_rate": 8.398473729558723e-05,
+      "loss": 1.1659,
+      "step": 4352
+    },
+    {
+      "epoch": 0.26447536302326996,
+      "grad_norm": 0.20280729234218597,
+      "learning_rate": 8.397771509072028e-05,
+      "loss": 1.0915,
+      "step": 4353
+    },
+    {
+      "epoch": 0.2645361200558965,
+      "grad_norm": 0.20858359336853027,
+      "learning_rate": 8.39706916403936e-05,
+      "loss": 1.1146,
+      "step": 4354
+    },
+    {
+      "epoch": 0.264596877088523,
+      "grad_norm": 0.20127947628498077,
+      "learning_rate": 8.396366694486466e-05,
+      "loss": 1.1034,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2646576341211495,
+      "grad_norm": 0.1488337367773056,
+      "learning_rate": 8.395664100439096e-05,
+      "loss": 1.0257,
+      "step": 4356
+    },
+    {
+      "epoch": 0.26471839115377604,
+      "grad_norm": 0.18887291848659515,
+      "learning_rate": 8.394961381922999e-05,
+      "loss": 1.1087,
+      "step": 4357
+    },
+    {
+      "epoch": 0.2647791481864026,
+      "grad_norm": 0.19009460508823395,
+      "learning_rate": 8.394258538963937e-05,
+      "loss": 1.1282,
+      "step": 4358
+    },
+    {
+      "epoch": 0.2648399052190291,
+      "grad_norm": 0.2440362423658371,
+      "learning_rate": 8.393555571587673e-05,
+      "loss": 1.0398,
+      "step": 4359
+    },
+    {
+      "epoch": 0.26490066225165565,
+      "grad_norm": 1.8505592346191406,
+      "learning_rate": 8.392852479819974e-05,
+      "loss": 1.1352,
+      "step": 4360
+    },
+    {
+      "epoch": 0.26496141928428213,
+      "grad_norm": 0.22668585181236267,
+      "learning_rate": 8.392149263686611e-05,
+      "loss": 1.0999,
+      "step": 4361
+    },
+    {
+      "epoch": 0.26502217631690866,
+      "grad_norm": 0.21716921031475067,
+      "learning_rate": 8.391445923213361e-05,
+      "loss": 1.0966,
+      "step": 4362
+    },
+    {
+      "epoch": 0.2650829333495352,
+      "grad_norm": 0.1901271641254425,
+      "learning_rate": 8.390742458426007e-05,
+      "loss": 1.2186,
+      "step": 4363
+    },
+    {
+      "epoch": 0.26514369038216173,
+      "grad_norm": 0.24133965373039246,
+      "learning_rate": 8.39003886935033e-05,
+      "loss": 1.1675,
+      "step": 4364
+    },
+    {
+      "epoch": 0.26520444741478827,
+      "grad_norm": 0.22141315042972565,
+      "learning_rate": 8.389335156012124e-05,
+      "loss": 1.1113,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2652652044474148,
+      "grad_norm": 0.14368648827075958,
+      "learning_rate": 8.388631318437183e-05,
+      "loss": 1.0755,
+      "step": 4366
+    },
+    {
+      "epoch": 0.26532596148004134,
+      "grad_norm": 0.22720423340797424,
+      "learning_rate": 8.387927356651306e-05,
+      "loss": 1.1467,
+      "step": 4367
+    },
+    {
+      "epoch": 0.2653867185126678,
+      "grad_norm": 0.2000090777873993,
+      "learning_rate": 8.387223270680296e-05,
+      "loss": 1.0808,
+      "step": 4368
+    },
+    {
+      "epoch": 0.26544747554529435,
+      "grad_norm": 0.20596404373645782,
+      "learning_rate": 8.386519060549962e-05,
+      "loss": 1.1044,
+      "step": 4369
+    },
+    {
+      "epoch": 0.2655082325779209,
+      "grad_norm": 0.1902557611465454,
+      "learning_rate": 8.38581472628612e-05,
+      "loss": 1.1099,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2655689896105474,
+      "grad_norm": 0.1719650775194168,
+      "learning_rate": 8.385110267914582e-05,
+      "loss": 1.1367,
+      "step": 4371
+    },
+    {
+      "epoch": 0.26562974664317396,
+      "grad_norm": 0.16829217970371246,
+      "learning_rate": 8.384405685461174e-05,
+      "loss": 1.1559,
+      "step": 4372
+    },
+    {
+      "epoch": 0.2656905036758005,
+      "grad_norm": 0.9026603102684021,
+      "learning_rate": 8.383700978951722e-05,
+      "loss": 1.153,
+      "step": 4373
+    },
+    {
+      "epoch": 0.265751260708427,
+      "grad_norm": 0.8896664977073669,
+      "learning_rate": 8.382996148412055e-05,
+      "loss": 1.0729,
+      "step": 4374
+    },
+    {
+      "epoch": 0.2658120177410535,
+      "grad_norm": 0.20693574845790863,
+      "learning_rate": 8.38229119386801e-05,
+      "loss": 1.1608,
+      "step": 4375
+    },
+    {
+      "epoch": 0.26587277477368004,
+      "grad_norm": 0.6755214333534241,
+      "learning_rate": 8.38158611534543e-05,
+      "loss": 1.1074,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2659335318063066,
+      "grad_norm": 0.3992058038711548,
+      "learning_rate": 8.380880912870155e-05,
+      "loss": 1.0855,
+      "step": 4377
+    },
+    {
+      "epoch": 0.2659942888389331,
+      "grad_norm": 0.1554999202489853,
+      "learning_rate": 8.380175586468038e-05,
+      "loss": 1.0975,
+      "step": 4378
+    },
+    {
+      "epoch": 0.26605504587155965,
+      "grad_norm": 0.15274523198604584,
+      "learning_rate": 8.37947013616493e-05,
+      "loss": 1.0872,
+      "step": 4379
+    },
+    {
+      "epoch": 0.2661158029041862,
+      "grad_norm": 0.1353253722190857,
+      "learning_rate": 8.378764561986691e-05,
+      "loss": 0.9963,
+      "step": 4380
+    },
+    {
+      "epoch": 0.26617655993681266,
+      "grad_norm": 0.20015887916088104,
+      "learning_rate": 8.378058863959186e-05,
+      "loss": 1.1271,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2662373169694392,
+      "grad_norm": 0.192811518907547,
+      "learning_rate": 8.377353042108277e-05,
+      "loss": 1.1003,
+      "step": 4382
+    },
+    {
+      "epoch": 0.26629807400206573,
+      "grad_norm": 0.5832599997520447,
+      "learning_rate": 8.376647096459843e-05,
+      "loss": 1.0796,
+      "step": 4383
+    },
+    {
+      "epoch": 0.26635883103469227,
+      "grad_norm": 0.1367943435907364,
+      "learning_rate": 8.375941027039755e-05,
+      "loss": 1.0546,
+      "step": 4384
+    },
+    {
+      "epoch": 0.2664195880673188,
+      "grad_norm": 0.42363548278808594,
+      "learning_rate": 8.375234833873898e-05,
+      "loss": 1.0708,
+      "step": 4385
+    },
+    {
+      "epoch": 0.26648034509994534,
+      "grad_norm": 4.4285173416137695,
+      "learning_rate": 8.374528516988154e-05,
+      "loss": 1.0485,
+      "step": 4386
+    },
+    {
+      "epoch": 0.26654110213257187,
+      "grad_norm": 0.1950882524251938,
+      "learning_rate": 8.373822076408415e-05,
+      "loss": 1.0633,
+      "step": 4387
+    },
+    {
+      "epoch": 0.26660185916519835,
+      "grad_norm": 0.15844829380512238,
+      "learning_rate": 8.373115512160576e-05,
+      "loss": 1.1573,
+      "step": 4388
+    },
+    {
+      "epoch": 0.2666626161978249,
+      "grad_norm": 0.16829128563404083,
+      "learning_rate": 8.372408824270534e-05,
+      "loss": 1.1345,
+      "step": 4389
+    },
+    {
+      "epoch": 0.2667233732304514,
+      "grad_norm": 0.5832083225250244,
+      "learning_rate": 8.371702012764198e-05,
+      "loss": 1.1448,
+      "step": 4390
+    },
+    {
+      "epoch": 0.26678413026307796,
+      "grad_norm": 0.16852720081806183,
+      "learning_rate": 8.370995077667471e-05,
+      "loss": 1.0671,
+      "step": 4391
+    },
+    {
+      "epoch": 0.2668448872957045,
+      "grad_norm": 0.1869085729122162,
+      "learning_rate": 8.370288019006269e-05,
+      "loss": 1.0775,
+      "step": 4392
+    },
+    {
+      "epoch": 0.266905644328331,
+      "grad_norm": 1.2228291034698486,
+      "learning_rate": 8.369580836806508e-05,
+      "loss": 1.0945,
+      "step": 4393
+    },
+    {
+      "epoch": 0.2669664013609575,
+      "grad_norm": 0.9789943099021912,
+      "learning_rate": 8.368873531094109e-05,
+      "loss": 1.2168,
+      "step": 4394
+    },
+    {
+      "epoch": 0.26702715839358404,
+      "grad_norm": 0.22951073944568634,
+      "learning_rate": 8.368166101894999e-05,
+      "loss": 1.1413,
+      "step": 4395
+    },
+    {
+      "epoch": 0.2670879154262106,
+      "grad_norm": 0.17168264091014862,
+      "learning_rate": 8.36745854923511e-05,
+      "loss": 1.1657,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2671486724588371,
+      "grad_norm": 0.21820928156375885,
+      "learning_rate": 8.366750873140379e-05,
+      "loss": 1.1243,
+      "step": 4397
+    },
+    {
+      "epoch": 0.26720942949146365,
+      "grad_norm": 0.35053229331970215,
+      "learning_rate": 8.366043073636743e-05,
+      "loss": 1.1848,
+      "step": 4398
+    },
+    {
+      "epoch": 0.2672701865240902,
+      "grad_norm": 0.1923595815896988,
+      "learning_rate": 8.365335150750147e-05,
+      "loss": 1.1164,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2673309435567167,
+      "grad_norm": 0.29366567730903625,
+      "learning_rate": 8.36462710450654e-05,
+      "loss": 1.0871,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2673917005893432,
+      "grad_norm": 0.17893652617931366,
+      "learning_rate": 8.363918934931875e-05,
+      "loss": 1.1154,
+      "step": 4401
+    },
+    {
+      "epoch": 0.26745245762196973,
+      "grad_norm": 0.18954449892044067,
+      "learning_rate": 8.363210642052113e-05,
+      "loss": 1.0414,
+      "step": 4402
+    },
+    {
+      "epoch": 0.26751321465459627,
+      "grad_norm": 0.23722869157791138,
+      "learning_rate": 8.362502225893216e-05,
+      "loss": 1.0774,
+      "step": 4403
+    },
+    {
+      "epoch": 0.2675739716872228,
+      "grad_norm": 0.1975167691707611,
+      "learning_rate": 8.361793686481149e-05,
+      "loss": 1.2206,
+      "step": 4404
+    },
+    {
+      "epoch": 0.26763472871984934,
+      "grad_norm": 0.218455970287323,
+      "learning_rate": 8.361085023841884e-05,
+      "loss": 1.2528,
+      "step": 4405
+    },
+    {
+      "epoch": 0.26769548575247587,
+      "grad_norm": 0.27026262879371643,
+      "learning_rate": 8.360376238001397e-05,
+      "loss": 1.1444,
+      "step": 4406
+    },
+    {
+      "epoch": 0.26775624278510235,
+      "grad_norm": 0.16529852151870728,
+      "learning_rate": 8.359667328985672e-05,
+      "loss": 1.0959,
+      "step": 4407
+    },
+    {
+      "epoch": 0.2678169998177289,
+      "grad_norm": 1.2387454509735107,
+      "learning_rate": 8.358958296820689e-05,
+      "loss": 1.1062,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2678777568503554,
+      "grad_norm": 0.33961236476898193,
+      "learning_rate": 8.358249141532442e-05,
+      "loss": 1.0911,
+      "step": 4409
+    },
+    {
+      "epoch": 0.26793851388298195,
+      "grad_norm": 0.16197733581066132,
+      "learning_rate": 8.357539863146922e-05,
+      "loss": 1.1403,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2679992709156085,
+      "grad_norm": 0.47110915184020996,
+      "learning_rate": 8.356830461690131e-05,
+      "loss": 1.0413,
+      "step": 4411
+    },
+    {
+      "epoch": 0.268060027948235,
+      "grad_norm": 0.1606959104537964,
+      "learning_rate": 8.35612093718807e-05,
+      "loss": 1.0596,
+      "step": 4412
+    },
+    {
+      "epoch": 0.26812078498086156,
+      "grad_norm": 0.383693665266037,
+      "learning_rate": 8.355411289666749e-05,
+      "loss": 1.0932,
+      "step": 4413
+    },
+    {
+      "epoch": 0.26818154201348804,
+      "grad_norm": 0.16229099035263062,
+      "learning_rate": 8.354701519152175e-05,
+      "loss": 1.1067,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2682422990461146,
+      "grad_norm": 0.1542864590883255,
+      "learning_rate": 8.35399162567037e-05,
+      "loss": 1.0703,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2683030560787411,
+      "grad_norm": 0.5054668188095093,
+      "learning_rate": 8.353281609247355e-05,
+      "loss": 1.1897,
+      "step": 4416
+    },
+    {
+      "epoch": 0.26836381311136764,
+      "grad_norm": 0.2539095878601074,
+      "learning_rate": 8.352571469909152e-05,
+      "loss": 1.0821,
+      "step": 4417
+    },
+    {
+      "epoch": 0.2684245701439942,
+      "grad_norm": 0.23482497036457062,
+      "learning_rate": 8.351861207681795e-05,
+      "loss": 1.1568,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2684853271766207,
+      "grad_norm": 0.21159839630126953,
+      "learning_rate": 8.351150822591318e-05,
+      "loss": 1.1703,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2685460842092472,
+      "grad_norm": 0.17760232090950012,
+      "learning_rate": 8.350440314663759e-05,
+      "loss": 1.0928,
+      "step": 4420
+    },
+    {
+      "epoch": 0.26860684124187373,
+      "grad_norm": 0.15669623017311096,
+      "learning_rate": 8.349729683925163e-05,
+      "loss": 1.2026,
+      "step": 4421
+    },
+    {
+      "epoch": 0.26866759827450026,
+      "grad_norm": 0.20074933767318726,
+      "learning_rate": 8.349018930401577e-05,
+      "loss": 1.0896,
+      "step": 4422
+    },
+    {
+      "epoch": 0.2687283553071268,
+      "grad_norm": 0.22339986264705658,
+      "learning_rate": 8.348308054119058e-05,
+      "loss": 1.1512,
+      "step": 4423
+    },
+    {
+      "epoch": 0.26878911233975333,
+      "grad_norm": 0.21061965823173523,
+      "learning_rate": 8.347597055103658e-05,
+      "loss": 1.1121,
+      "step": 4424
+    },
+    {
+      "epoch": 0.26884986937237987,
+      "grad_norm": 0.3132881820201874,
+      "learning_rate": 8.346885933381439e-05,
+      "loss": 1.1502,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2689106264050064,
+      "grad_norm": 0.3162994384765625,
+      "learning_rate": 8.346174688978471e-05,
+      "loss": 1.1553,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2689713834376329,
+      "grad_norm": 0.29460978507995605,
+      "learning_rate": 8.345463321920823e-05,
+      "loss": 1.0553,
+      "step": 4427
+    },
+    {
+      "epoch": 0.2690321404702594,
+      "grad_norm": 0.285300612449646,
+      "learning_rate": 8.34475183223457e-05,
+      "loss": 1.0708,
+      "step": 4428
+    },
+    {
+      "epoch": 0.26909289750288595,
+      "grad_norm": 0.41496700048446655,
+      "learning_rate": 8.344040219945793e-05,
+      "loss": 1.1075,
+      "step": 4429
+    },
+    {
+      "epoch": 0.2691536545355125,
+      "grad_norm": 0.15282705426216125,
+      "learning_rate": 8.343328485080574e-05,
+      "loss": 1.0922,
+      "step": 4430
+    },
+    {
+      "epoch": 0.269214411568139,
+      "grad_norm": 0.2773056626319885,
+      "learning_rate": 8.342616627665003e-05,
+      "loss": 1.1538,
+      "step": 4431
+    },
+    {
+      "epoch": 0.26927516860076556,
+      "grad_norm": 0.36892804503440857,
+      "learning_rate": 8.341904647725174e-05,
+      "loss": 1.1762,
+      "step": 4432
+    },
+    {
+      "epoch": 0.2693359256333921,
+      "grad_norm": 0.1873452365398407,
+      "learning_rate": 8.341192545287185e-05,
+      "loss": 1.1327,
+      "step": 4433
+    },
+    {
+      "epoch": 0.26939668266601857,
+      "grad_norm": 0.2599208652973175,
+      "learning_rate": 8.340480320377136e-05,
+      "loss": 1.0998,
+      "step": 4434
+    },
+    {
+      "epoch": 0.2694574396986451,
+      "grad_norm": 0.1900174915790558,
+      "learning_rate": 8.339767973021135e-05,
+      "loss": 1.0994,
+      "step": 4435
+    },
+    {
+      "epoch": 0.26951819673127164,
+      "grad_norm": 0.3114914894104004,
+      "learning_rate": 8.339055503245294e-05,
+      "loss": 1.1037,
+      "step": 4436
+    },
+    {
+      "epoch": 0.2695789537638982,
+      "grad_norm": 0.21054331958293915,
+      "learning_rate": 8.338342911075729e-05,
+      "loss": 1.1417,
+      "step": 4437
+    },
+    {
+      "epoch": 0.2696397107965247,
+      "grad_norm": 0.2541572153568268,
+      "learning_rate": 8.33763019653856e-05,
+      "loss": 1.1443,
+      "step": 4438
+    },
+    {
+      "epoch": 0.26970046782915125,
+      "grad_norm": 0.3991122245788574,
+      "learning_rate": 8.336917359659909e-05,
+      "loss": 1.1564,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2697612248617777,
+      "grad_norm": 0.19234372675418854,
+      "learning_rate": 8.336204400465908e-05,
+      "loss": 1.0291,
+      "step": 4440
+    },
+    {
+      "epoch": 0.26982198189440426,
+      "grad_norm": 0.5077095031738281,
+      "learning_rate": 8.335491318982689e-05,
+      "loss": 1.2786,
+      "step": 4441
+    },
+    {
+      "epoch": 0.2698827389270308,
+      "grad_norm": 0.28668439388275146,
+      "learning_rate": 8.334778115236392e-05,
+      "loss": 1.1063,
+      "step": 4442
+    },
+    {
+      "epoch": 0.26994349595965733,
+      "grad_norm": 0.2900209426879883,
+      "learning_rate": 8.334064789253157e-05,
+      "loss": 1.1019,
+      "step": 4443
+    },
+    {
+      "epoch": 0.27000425299228387,
+      "grad_norm": 0.18565015494823456,
+      "learning_rate": 8.333351341059134e-05,
+      "loss": 1.0789,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2700650100249104,
+      "grad_norm": 0.2625967562198639,
+      "learning_rate": 8.332637770680474e-05,
+      "loss": 1.0648,
+      "step": 4445
+    },
+    {
+      "epoch": 0.27012576705753694,
+      "grad_norm": 0.26043277978897095,
+      "learning_rate": 8.331924078143329e-05,
+      "loss": 1.1473,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2701865240901634,
+      "grad_norm": 0.1826217621564865,
+      "learning_rate": 8.331210263473867e-05,
+      "loss": 1.1316,
+      "step": 4447
+    },
+    {
+      "epoch": 0.27024728112278995,
+      "grad_norm": 0.223603293299675,
+      "learning_rate": 8.330496326698247e-05,
+      "loss": 1.0609,
+      "step": 4448
+    },
+    {
+      "epoch": 0.2703080381554165,
+      "grad_norm": 0.20060575008392334,
+      "learning_rate": 8.329782267842641e-05,
+      "loss": 1.1495,
+      "step": 4449
+    },
+    {
+      "epoch": 0.270368795188043,
+      "grad_norm": 0.2720164954662323,
+      "learning_rate": 8.329068086933221e-05,
+      "loss": 1.0722,
+      "step": 4450
+    },
+    {
+      "epoch": 0.27042955222066956,
+      "grad_norm": 0.14457330107688904,
+      "learning_rate": 8.328353783996168e-05,
+      "loss": 1.0962,
+      "step": 4451
+    },
+    {
+      "epoch": 0.2704903092532961,
+      "grad_norm": 0.27922821044921875,
+      "learning_rate": 8.327639359057663e-05,
+      "loss": 1.0457,
+      "step": 4452
+    },
+    {
+      "epoch": 0.27055106628592257,
+      "grad_norm": 0.23993876576423645,
+      "learning_rate": 8.326924812143894e-05,
+      "loss": 1.1135,
+      "step": 4453
+    },
+    {
+      "epoch": 0.2706118233185491,
+      "grad_norm": 0.11556341499090195,
+      "learning_rate": 8.326210143281055e-05,
+      "loss": 1.0577,
+      "step": 4454
+    },
+    {
+      "epoch": 0.27067258035117564,
+      "grad_norm": 0.13276229798793793,
+      "learning_rate": 8.325495352495339e-05,
+      "loss": 1.0455,
+      "step": 4455
+    },
+    {
+      "epoch": 0.2707333373838022,
+      "grad_norm": 0.35786738991737366,
+      "learning_rate": 8.324780439812946e-05,
+      "loss": 1.1242,
+      "step": 4456
+    },
+    {
+      "epoch": 0.2707940944164287,
+      "grad_norm": 0.16986536979675293,
+      "learning_rate": 8.324065405260085e-05,
+      "loss": 1.1377,
+      "step": 4457
+    },
+    {
+      "epoch": 0.27085485144905525,
+      "grad_norm": 0.36732205748558044,
+      "learning_rate": 8.323350248862964e-05,
+      "loss": 1.1018,
+      "step": 4458
+    },
+    {
+      "epoch": 0.2709156084816818,
+      "grad_norm": 0.20355327427387238,
+      "learning_rate": 8.322634970647797e-05,
+      "loss": 1.1698,
+      "step": 4459
+    },
+    {
+      "epoch": 0.27097636551430826,
+      "grad_norm": 0.4248415231704712,
+      "learning_rate": 8.321919570640803e-05,
+      "loss": 1.3169,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2710371225469348,
+      "grad_norm": 0.2988456189632416,
+      "learning_rate": 8.321204048868206e-05,
+      "loss": 1.0989,
+      "step": 4461
+    },
+    {
+      "epoch": 0.27109787957956133,
+      "grad_norm": 0.241630420088768,
+      "learning_rate": 8.32048840535623e-05,
+      "loss": 1.111,
+      "step": 4462
+    },
+    {
+      "epoch": 0.27115863661218786,
+      "grad_norm": 0.2592075765132904,
+      "learning_rate": 8.319772640131113e-05,
+      "loss": 1.2772,
+      "step": 4463
+    },
+    {
+      "epoch": 0.2712193936448144,
+      "grad_norm": 0.23523442447185516,
+      "learning_rate": 8.319056753219086e-05,
+      "loss": 1.1463,
+      "step": 4464
+    },
+    {
+      "epoch": 0.27128015067744093,
+      "grad_norm": 0.15017908811569214,
+      "learning_rate": 8.318340744646393e-05,
+      "loss": 1.2114,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2713409077100674,
+      "grad_norm": 0.4901198148727417,
+      "learning_rate": 8.317624614439278e-05,
+      "loss": 1.3439,
+      "step": 4466
+    },
+    {
+      "epoch": 0.27140166474269395,
+      "grad_norm": 0.1556641310453415,
+      "learning_rate": 8.316908362623991e-05,
+      "loss": 1.0707,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2714624217753205,
+      "grad_norm": 0.2285921275615692,
+      "learning_rate": 8.31619198922679e-05,
+      "loss": 1.2014,
+      "step": 4468
+    },
+    {
+      "epoch": 0.271523178807947,
+      "grad_norm": 0.1972893625497818,
+      "learning_rate": 8.315475494273927e-05,
+      "loss": 1.1493,
+      "step": 4469
+    },
+    {
+      "epoch": 0.27158393584057355,
+      "grad_norm": 0.3628298044204712,
+      "learning_rate": 8.31475887779167e-05,
+      "loss": 1.1442,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2716446928732001,
+      "grad_norm": 0.16546091437339783,
+      "learning_rate": 8.314042139806287e-05,
+      "loss": 1.0411,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2717054499058266,
+      "grad_norm": 0.5137971043586731,
+      "learning_rate": 8.313325280344046e-05,
+      "loss": 1.0667,
+      "step": 4472
+    },
+    {
+      "epoch": 0.2717662069384531,
+      "grad_norm": 0.17597582936286926,
+      "learning_rate": 8.312608299431229e-05,
+      "loss": 1.0752,
+      "step": 4473
+    },
+    {
+      "epoch": 0.27182696397107964,
+      "grad_norm": 0.227156400680542,
+      "learning_rate": 8.311891197094113e-05,
+      "loss": 1.1466,
+      "step": 4474
+    },
+    {
+      "epoch": 0.2718877210037062,
+      "grad_norm": 0.18647707998752594,
+      "learning_rate": 8.311173973358985e-05,
+      "loss": 1.0874,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2719484780363327,
+      "grad_norm": 0.14235243201255798,
+      "learning_rate": 8.310456628252136e-05,
+      "loss": 1.0651,
+      "step": 4476
+    },
+    {
+      "epoch": 0.27200923506895924,
+      "grad_norm": 0.3674888014793396,
+      "learning_rate": 8.309739161799859e-05,
+      "loss": 1.0871,
+      "step": 4477
+    },
+    {
+      "epoch": 0.2720699921015858,
+      "grad_norm": 0.25865834951400757,
+      "learning_rate": 8.309021574028453e-05,
+      "loss": 1.1204,
+      "step": 4478
+    },
+    {
+      "epoch": 0.27213074913421226,
+      "grad_norm": 0.27623218297958374,
+      "learning_rate": 8.30830386496422e-05,
+      "loss": 1.0639,
+      "step": 4479
+    },
+    {
+      "epoch": 0.2721915061668388,
+      "grad_norm": 0.19091758131980896,
+      "learning_rate": 8.307586034633472e-05,
+      "loss": 1.2216,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2722522631994653,
+      "grad_norm": 0.2135348618030548,
+      "learning_rate": 8.306868083062518e-05,
+      "loss": 1.0923,
+      "step": 4481
+    },
+    {
+      "epoch": 0.27231302023209186,
+      "grad_norm": 0.34670794010162354,
+      "learning_rate": 8.306150010277675e-05,
+      "loss": 1.1129,
+      "step": 4482
+    },
+    {
+      "epoch": 0.2723737772647184,
+      "grad_norm": 0.14597612619400024,
+      "learning_rate": 8.305431816305262e-05,
+      "loss": 1.107,
+      "step": 4483
+    },
+    {
+      "epoch": 0.27243453429734493,
+      "grad_norm": 0.25984859466552734,
+      "learning_rate": 8.30471350117161e-05,
+      "loss": 1.0409,
+      "step": 4484
+    },
+    {
+      "epoch": 0.27249529132997147,
+      "grad_norm": 0.1626632660627365,
+      "learning_rate": 8.303995064903045e-05,
+      "loss": 1.0621,
+      "step": 4485
+    },
+    {
+      "epoch": 0.27255604836259795,
+      "grad_norm": 0.20829637348651886,
+      "learning_rate": 8.303276507525902e-05,
+      "loss": 1.0655,
+      "step": 4486
+    },
+    {
+      "epoch": 0.2726168053952245,
+      "grad_norm": 0.31719571352005005,
+      "learning_rate": 8.30255782906652e-05,
+      "loss": 1.1134,
+      "step": 4487
+    },
+    {
+      "epoch": 0.272677562427851,
+      "grad_norm": 0.14328913390636444,
+      "learning_rate": 8.301839029551243e-05,
+      "loss": 1.0976,
+      "step": 4488
+    },
+    {
+      "epoch": 0.27273831946047755,
+      "grad_norm": 0.422942578792572,
+      "learning_rate": 8.301120109006418e-05,
+      "loss": 1.1305,
+      "step": 4489
+    },
+    {
+      "epoch": 0.2727990764931041,
+      "grad_norm": 0.12200777977705002,
+      "learning_rate": 8.300401067458396e-05,
+      "loss": 1.0784,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2728598335257306,
+      "grad_norm": 0.2534925043582916,
+      "learning_rate": 8.299681904933536e-05,
+      "loss": 1.1613,
+      "step": 4491
+    },
+    {
+      "epoch": 0.27292059055835716,
+      "grad_norm": 0.1515333652496338,
+      "learning_rate": 8.298962621458197e-05,
+      "loss": 1.0507,
+      "step": 4492
+    },
+    {
+      "epoch": 0.27298134759098364,
+      "grad_norm": 0.2900865375995636,
+      "learning_rate": 8.298243217058747e-05,
+      "loss": 1.0689,
+      "step": 4493
+    },
+    {
+      "epoch": 0.27304210462361017,
+      "grad_norm": 0.17336629331111908,
+      "learning_rate": 8.297523691761554e-05,
+      "loss": 1.0488,
+      "step": 4494
+    },
+    {
+      "epoch": 0.2731028616562367,
+      "grad_norm": 0.18693529069423676,
+      "learning_rate": 8.296804045592992e-05,
+      "loss": 1.0673,
+      "step": 4495
+    },
+    {
+      "epoch": 0.27316361868886324,
+      "grad_norm": 0.2472897171974182,
+      "learning_rate": 8.296084278579442e-05,
+      "loss": 1.2907,
+      "step": 4496
+    },
+    {
+      "epoch": 0.2732243757214898,
+      "grad_norm": 0.1313284933567047,
+      "learning_rate": 8.295364390747284e-05,
+      "loss": 1.0733,
+      "step": 4497
+    },
+    {
+      "epoch": 0.2732851327541163,
+      "grad_norm": 0.15705497562885284,
+      "learning_rate": 8.294644382122906e-05,
+      "loss": 1.2581,
+      "step": 4498
+    },
+    {
+      "epoch": 0.2733458897867428,
+      "grad_norm": 0.1942000836133957,
+      "learning_rate": 8.293924252732704e-05,
+      "loss": 1.1025,
+      "step": 4499
+    },
+    {
+      "epoch": 0.2734066468193693,
+      "grad_norm": 0.2068057507276535,
+      "learning_rate": 8.29320400260307e-05,
+      "loss": 1.2565,
+      "step": 4500
+    },
+    {
+      "epoch": 0.27346740385199586,
+      "grad_norm": 0.17843692004680634,
+      "learning_rate": 8.292483631760408e-05,
+      "loss": 1.2376,
+      "step": 4501
+    },
+    {
+      "epoch": 0.2735281608846224,
+      "grad_norm": 0.16642257571220398,
+      "learning_rate": 8.291763140231122e-05,
+      "loss": 1.1336,
+      "step": 4502
+    },
+    {
+      "epoch": 0.27358891791724893,
+      "grad_norm": 0.179488867521286,
+      "learning_rate": 8.291042528041622e-05,
+      "loss": 1.1094,
+      "step": 4503
+    },
+    {
+      "epoch": 0.27364967494987547,
+      "grad_norm": 0.22454987466335297,
+      "learning_rate": 8.290321795218321e-05,
+      "loss": 1.0633,
+      "step": 4504
+    },
+    {
+      "epoch": 0.273710431982502,
+      "grad_norm": 0.26553380489349365,
+      "learning_rate": 8.289600941787639e-05,
+      "loss": 1.081,
+      "step": 4505
+    },
+    {
+      "epoch": 0.2737711890151285,
+      "grad_norm": 0.16264106333255768,
+      "learning_rate": 8.288879967775997e-05,
+      "loss": 1.0821,
+      "step": 4506
+    },
+    {
+      "epoch": 0.273831946047755,
+      "grad_norm": 0.41247719526290894,
+      "learning_rate": 8.288158873209826e-05,
+      "loss": 1.1986,
+      "step": 4507
+    },
+    {
+      "epoch": 0.27389270308038155,
+      "grad_norm": 0.24013249576091766,
+      "learning_rate": 8.287437658115556e-05,
+      "loss": 1.1142,
+      "step": 4508
+    },
+    {
+      "epoch": 0.2739534601130081,
+      "grad_norm": 0.4005536437034607,
+      "learning_rate": 8.286716322519624e-05,
+      "loss": 1.0716,
+      "step": 4509
+    },
+    {
+      "epoch": 0.2740142171456346,
+      "grad_norm": 0.34069719910621643,
+      "learning_rate": 8.285994866448468e-05,
+      "loss": 1.0756,
+      "step": 4510
+    },
+    {
+      "epoch": 0.27407497417826115,
+      "grad_norm": 0.2507462203502655,
+      "learning_rate": 8.285273289928535e-05,
+      "loss": 1.1925,
+      "step": 4511
+    },
+    {
+      "epoch": 0.27413573121088763,
+      "grad_norm": 0.2653174102306366,
+      "learning_rate": 8.284551592986277e-05,
+      "loss": 1.1208,
+      "step": 4512
+    },
+    {
+      "epoch": 0.27419648824351417,
+      "grad_norm": 0.17350877821445465,
+      "learning_rate": 8.283829775648143e-05,
+      "loss": 1.1445,
+      "step": 4513
+    },
+    {
+      "epoch": 0.2742572452761407,
+      "grad_norm": 3.2536861896514893,
+      "learning_rate": 8.283107837940593e-05,
+      "loss": 1.1013,
+      "step": 4514
+    },
+    {
+      "epoch": 0.27431800230876724,
+      "grad_norm": 0.6003816723823547,
+      "learning_rate": 8.282385779890092e-05,
+      "loss": 1.1052,
+      "step": 4515
+    },
+    {
+      "epoch": 0.2743787593413938,
+      "grad_norm": 1.152153491973877,
+      "learning_rate": 8.281663601523107e-05,
+      "loss": 1.2417,
+      "step": 4516
+    },
+    {
+      "epoch": 0.2744395163740203,
+      "grad_norm": 0.39114993810653687,
+      "learning_rate": 8.280941302866108e-05,
+      "loss": 1.117,
+      "step": 4517
+    },
+    {
+      "epoch": 0.27450027340664684,
+      "grad_norm": 0.26109611988067627,
+      "learning_rate": 8.28021888394557e-05,
+      "loss": 1.1302,
+      "step": 4518
+    },
+    {
+      "epoch": 0.2745610304392733,
+      "grad_norm": 0.4179966151714325,
+      "learning_rate": 8.279496344787975e-05,
+      "loss": 1.0478,
+      "step": 4519
+    },
+    {
+      "epoch": 0.27462178747189986,
+      "grad_norm": 0.19528590142726898,
+      "learning_rate": 8.278773685419808e-05,
+      "loss": 1.1558,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2746825445045264,
+      "grad_norm": 0.326621413230896,
+      "learning_rate": 8.278050905867559e-05,
+      "loss": 1.1047,
+      "step": 4521
+    },
+    {
+      "epoch": 0.27474330153715293,
+      "grad_norm": 0.2561730444431305,
+      "learning_rate": 8.277328006157718e-05,
+      "loss": 1.1224,
+      "step": 4522
+    },
+    {
+      "epoch": 0.27480405856977946,
+      "grad_norm": 0.29456570744514465,
+      "learning_rate": 8.276604986316785e-05,
+      "loss": 1.0875,
+      "step": 4523
+    },
+    {
+      "epoch": 0.274864815602406,
+      "grad_norm": 0.19292910397052765,
+      "learning_rate": 8.275881846371265e-05,
+      "loss": 1.0527,
+      "step": 4524
+    },
+    {
+      "epoch": 0.2749255726350325,
+      "grad_norm": 0.2293853908777237,
+      "learning_rate": 8.275158586347662e-05,
+      "loss": 1.0718,
+      "step": 4525
+    },
+    {
+      "epoch": 0.274986329667659,
+      "grad_norm": 0.2319045066833496,
+      "learning_rate": 8.274435206272489e-05,
+      "loss": 1.0412,
+      "step": 4526
+    },
+    {
+      "epoch": 0.27504708670028555,
+      "grad_norm": 0.19103401899337769,
+      "learning_rate": 8.27371170617226e-05,
+      "loss": 1.0694,
+      "step": 4527
+    },
+    {
+      "epoch": 0.2751078437329121,
+      "grad_norm": 0.2501087784767151,
+      "learning_rate": 8.272988086073494e-05,
+      "loss": 1.1286,
+      "step": 4528
+    },
+    {
+      "epoch": 0.2751686007655386,
+      "grad_norm": 0.32802215218544006,
+      "learning_rate": 8.272264346002719e-05,
+      "loss": 1.2307,
+      "step": 4529
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 0.14455841481685638,
+      "learning_rate": 8.271540485986461e-05,
+      "loss": 1.1745,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2752901148307917,
+      "grad_norm": 0.24390700459480286,
+      "learning_rate": 8.270816506051253e-05,
+      "loss": 1.1297,
+      "step": 4531
+    },
+    {
+      "epoch": 0.27535087186341817,
+      "grad_norm": 0.3448156416416168,
+      "learning_rate": 8.270092406223635e-05,
+      "loss": 1.112,
+      "step": 4532
+    },
+    {
+      "epoch": 0.2754116288960447,
+      "grad_norm": 0.15369431674480438,
+      "learning_rate": 8.269368186530149e-05,
+      "loss": 1.1169,
+      "step": 4533
+    },
+    {
+      "epoch": 0.27547238592867124,
+      "grad_norm": 0.18233084678649902,
+      "learning_rate": 8.26864384699734e-05,
+      "loss": 1.1651,
+      "step": 4534
+    },
+    {
+      "epoch": 0.2755331429612978,
+      "grad_norm": 0.17538852989673615,
+      "learning_rate": 8.267919387651757e-05,
+      "loss": 1.1014,
+      "step": 4535
+    },
+    {
+      "epoch": 0.2755938999939243,
+      "grad_norm": 0.1329255998134613,
+      "learning_rate": 8.267194808519958e-05,
+      "loss": 1.0677,
+      "step": 4536
+    },
+    {
+      "epoch": 0.27565465702655084,
+      "grad_norm": 0.4503363072872162,
+      "learning_rate": 8.266470109628501e-05,
+      "loss": 1.0506,
+      "step": 4537
+    },
+    {
+      "epoch": 0.2757154140591774,
+      "grad_norm": 0.1340859830379486,
+      "learning_rate": 8.265745291003953e-05,
+      "loss": 1.042,
+      "step": 4538
+    },
+    {
+      "epoch": 0.27577617109180386,
+      "grad_norm": 0.13064821064472198,
+      "learning_rate": 8.265020352672877e-05,
+      "loss": 1.0345,
+      "step": 4539
+    },
+    {
+      "epoch": 0.2758369281244304,
+      "grad_norm": 0.15410307049751282,
+      "learning_rate": 8.26429529466185e-05,
+      "loss": 1.091,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2758976851570569,
+      "grad_norm": 0.13411352038383484,
+      "learning_rate": 8.26357011699745e-05,
+      "loss": 1.0297,
+      "step": 4541
+    },
+    {
+      "epoch": 0.27595844218968346,
+      "grad_norm": 0.1289222687482834,
+      "learning_rate": 8.262844819706255e-05,
+      "loss": 1.101,
+      "step": 4542
+    },
+    {
+      "epoch": 0.27601919922231,
+      "grad_norm": 0.16975997388362885,
+      "learning_rate": 8.262119402814852e-05,
+      "loss": 1.0901,
+      "step": 4543
+    },
+    {
+      "epoch": 0.27607995625493653,
+      "grad_norm": 0.740832507610321,
+      "learning_rate": 8.261393866349832e-05,
+      "loss": 1.2342,
+      "step": 4544
+    },
+    {
+      "epoch": 0.276140713287563,
+      "grad_norm": 0.16148574650287628,
+      "learning_rate": 8.260668210337789e-05,
+      "loss": 1.1655,
+      "step": 4545
+    },
+    {
+      "epoch": 0.27620147032018955,
+      "grad_norm": 0.13259842991828918,
+      "learning_rate": 8.259942434805324e-05,
+      "loss": 1.0706,
+      "step": 4546
+    },
+    {
+      "epoch": 0.2762622273528161,
+      "grad_norm": 0.34875038266181946,
+      "learning_rate": 8.259216539779038e-05,
+      "loss": 1.0904,
+      "step": 4547
+    },
+    {
+      "epoch": 0.2763229843854426,
+      "grad_norm": 0.16057588160037994,
+      "learning_rate": 8.25849052528554e-05,
+      "loss": 1.0704,
+      "step": 4548
+    },
+    {
+      "epoch": 0.27638374141806915,
+      "grad_norm": 0.17092905938625336,
+      "learning_rate": 8.25776439135144e-05,
+      "loss": 1.0683,
+      "step": 4549
+    },
+    {
+      "epoch": 0.2764444984506957,
+      "grad_norm": 0.21016310155391693,
+      "learning_rate": 8.25703813800336e-05,
+      "loss": 1.1103,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2765052554833222,
+      "grad_norm": 0.22097277641296387,
+      "learning_rate": 8.256311765267916e-05,
+      "loss": 1.1482,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2765660125159487,
+      "grad_norm": 0.160888209939003,
+      "learning_rate": 8.255585273171735e-05,
+      "loss": 1.0122,
+      "step": 4552
+    },
+    {
+      "epoch": 0.27662676954857524,
+      "grad_norm": 0.2927054464817047,
+      "learning_rate": 8.254858661741444e-05,
+      "loss": 1.1903,
+      "step": 4553
+    },
+    {
+      "epoch": 0.27668752658120177,
+      "grad_norm": 0.15477386116981506,
+      "learning_rate": 8.254131931003683e-05,
+      "loss": 1.0771,
+      "step": 4554
+    },
+    {
+      "epoch": 0.2767482836138283,
+      "grad_norm": 0.1971157342195511,
+      "learning_rate": 8.253405080985087e-05,
+      "loss": 1.1034,
+      "step": 4555
+    },
+    {
+      "epoch": 0.27680904064645484,
+      "grad_norm": 0.6341344714164734,
+      "learning_rate": 8.252678111712296e-05,
+      "loss": 1.3019,
+      "step": 4556
+    },
+    {
+      "epoch": 0.2768697976790814,
+      "grad_norm": 0.2178129106760025,
+      "learning_rate": 8.251951023211962e-05,
+      "loss": 1.1374,
+      "step": 4557
+    },
+    {
+      "epoch": 0.27693055471170785,
+      "grad_norm": 0.1752653867006302,
+      "learning_rate": 8.251223815510734e-05,
+      "loss": 1.1242,
+      "step": 4558
+    },
+    {
+      "epoch": 0.2769913117443344,
+      "grad_norm": 0.2596944570541382,
+      "learning_rate": 8.250496488635268e-05,
+      "loss": 1.1055,
+      "step": 4559
+    },
+    {
+      "epoch": 0.2770520687769609,
+      "grad_norm": 0.14579010009765625,
+      "learning_rate": 8.249769042612227e-05,
+      "loss": 1.0866,
+      "step": 4560
+    },
+    {
+      "epoch": 0.27711282580958746,
+      "grad_norm": 0.25169453024864197,
+      "learning_rate": 8.249041477468273e-05,
+      "loss": 1.1961,
+      "step": 4561
+    },
+    {
+      "epoch": 0.277173582842214,
+      "grad_norm": 6.168807506561279,
+      "learning_rate": 8.248313793230074e-05,
+      "loss": 1.0726,
+      "step": 4562
+    },
+    {
+      "epoch": 0.27723433987484053,
+      "grad_norm": 1.6058295965194702,
+      "learning_rate": 8.247585989924306e-05,
+      "loss": 1.3326,
+      "step": 4563
+    },
+    {
+      "epoch": 0.27729509690746706,
+      "grad_norm": 0.16555973887443542,
+      "learning_rate": 8.246858067577648e-05,
+      "loss": 1.117,
+      "step": 4564
+    },
+    {
+      "epoch": 0.27735585394009354,
+      "grad_norm": 0.1868455559015274,
+      "learning_rate": 8.246130026216776e-05,
+      "loss": 1.1559,
+      "step": 4565
+    },
+    {
+      "epoch": 0.2774166109727201,
+      "grad_norm": 0.3097016513347626,
+      "learning_rate": 8.245401865868383e-05,
+      "loss": 1.1407,
+      "step": 4566
+    },
+    {
+      "epoch": 0.2774773680053466,
+      "grad_norm": 0.16543899476528168,
+      "learning_rate": 8.244673586559157e-05,
+      "loss": 1.1081,
+      "step": 4567
+    },
+    {
+      "epoch": 0.27753812503797315,
+      "grad_norm": 0.1566673070192337,
+      "learning_rate": 8.243945188315791e-05,
+      "loss": 1.0997,
+      "step": 4568
+    },
+    {
+      "epoch": 0.2775988820705997,
+      "grad_norm": 0.2237454652786255,
+      "learning_rate": 8.243216671164991e-05,
+      "loss": 1.1184,
+      "step": 4569
+    },
+    {
+      "epoch": 0.2776596391032262,
+      "grad_norm": 0.19540287554264069,
+      "learning_rate": 8.242488035133454e-05,
+      "loss": 1.1303,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2777203961358527,
+      "grad_norm": 0.16383953392505646,
+      "learning_rate": 8.241759280247893e-05,
+      "loss": 1.0531,
+      "step": 4571
+    },
+    {
+      "epoch": 0.27778115316847923,
+      "grad_norm": 0.14741382002830505,
+      "learning_rate": 8.241030406535017e-05,
+      "loss": 1.0493,
+      "step": 4572
+    },
+    {
+      "epoch": 0.27784191020110577,
+      "grad_norm": 0.20527935028076172,
+      "learning_rate": 8.240301414021548e-05,
+      "loss": 1.0885,
+      "step": 4573
+    },
+    {
+      "epoch": 0.2779026672337323,
+      "grad_norm": 0.3679577112197876,
+      "learning_rate": 8.239572302734202e-05,
+      "loss": 1.6023,
+      "step": 4574
+    },
+    {
+      "epoch": 0.27796342426635884,
+      "grad_norm": 0.19015346467494965,
+      "learning_rate": 8.238843072699708e-05,
+      "loss": 1.0688,
+      "step": 4575
+    },
+    {
+      "epoch": 0.2780241812989854,
+      "grad_norm": 0.31583908200263977,
+      "learning_rate": 8.238113723944795e-05,
+      "loss": 1.1412,
+      "step": 4576
+    },
+    {
+      "epoch": 0.2780849383316119,
+      "grad_norm": 0.18067505955696106,
+      "learning_rate": 8.237384256496197e-05,
+      "loss": 1.0639,
+      "step": 4577
+    },
+    {
+      "epoch": 0.2781456953642384,
+      "grad_norm": 0.27275148034095764,
+      "learning_rate": 8.236654670380655e-05,
+      "loss": 1.1374,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2782064523968649,
+      "grad_norm": 0.3129608929157257,
+      "learning_rate": 8.235924965624909e-05,
+      "loss": 1.0456,
+      "step": 4579
+    },
+    {
+      "epoch": 0.27826720942949146,
+      "grad_norm": 0.3833840489387512,
+      "learning_rate": 8.235195142255707e-05,
+      "loss": 1.0886,
+      "step": 4580
+    },
+    {
+      "epoch": 0.278327966462118,
+      "grad_norm": 0.20252899825572968,
+      "learning_rate": 8.234465200299803e-05,
+      "loss": 1.2008,
+      "step": 4581
+    },
+    {
+      "epoch": 0.2783887234947445,
+      "grad_norm": 0.23847143352031708,
+      "learning_rate": 8.233735139783952e-05,
+      "loss": 1.1262,
+      "step": 4582
+    },
+    {
+      "epoch": 0.27844948052737106,
+      "grad_norm": 0.16970612108707428,
+      "learning_rate": 8.233004960734913e-05,
+      "loss": 1.1091,
+      "step": 4583
+    },
+    {
+      "epoch": 0.27851023755999754,
+      "grad_norm": 0.21134312450885773,
+      "learning_rate": 8.232274663179453e-05,
+      "loss": 1.1954,
+      "step": 4584
+    },
+    {
+      "epoch": 0.2785709945926241,
+      "grad_norm": 0.16985607147216797,
+      "learning_rate": 8.231544247144341e-05,
+      "loss": 1.0927,
+      "step": 4585
+    },
+    {
+      "epoch": 0.2786317516252506,
+      "grad_norm": 0.13938719034194946,
+      "learning_rate": 8.23081371265635e-05,
+      "loss": 1.0526,
+      "step": 4586
+    },
+    {
+      "epoch": 0.27869250865787715,
+      "grad_norm": 0.19115585088729858,
+      "learning_rate": 8.230083059742257e-05,
+      "loss": 1.1016,
+      "step": 4587
+    },
+    {
+      "epoch": 0.2787532656905037,
+      "grad_norm": 0.14090359210968018,
+      "learning_rate": 8.229352288428846e-05,
+      "loss": 1.035,
+      "step": 4588
+    },
+    {
+      "epoch": 0.2788140227231302,
+      "grad_norm": 0.2454357147216797,
+      "learning_rate": 8.228621398742902e-05,
+      "loss": 1.1749,
+      "step": 4589
+    },
+    {
+      "epoch": 0.27887477975575675,
+      "grad_norm": 0.21398983895778656,
+      "learning_rate": 8.227890390711216e-05,
+      "loss": 1.1485,
+      "step": 4590
+    },
+    {
+      "epoch": 0.27893553678838323,
+      "grad_norm": 0.23139554262161255,
+      "learning_rate": 8.227159264360586e-05,
+      "loss": 1.0546,
+      "step": 4591
+    },
+    {
+      "epoch": 0.27899629382100977,
+      "grad_norm": 0.16300325095653534,
+      "learning_rate": 8.226428019717807e-05,
+      "loss": 1.1082,
+      "step": 4592
+    },
+    {
+      "epoch": 0.2790570508536363,
+      "grad_norm": 1.25884211063385,
+      "learning_rate": 8.225696656809688e-05,
+      "loss": 1.0853,
+      "step": 4593
+    },
+    {
+      "epoch": 0.27911780788626284,
+      "grad_norm": 0.18098796904087067,
+      "learning_rate": 8.224965175663032e-05,
+      "loss": 1.1053,
+      "step": 4594
+    },
+    {
+      "epoch": 0.27917856491888937,
+      "grad_norm": 0.35015469789505005,
+      "learning_rate": 8.224233576304655e-05,
+      "loss": 1.0849,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2792393219515159,
+      "grad_norm": 0.14025036990642548,
+      "learning_rate": 8.223501858761372e-05,
+      "loss": 1.0583,
+      "step": 4596
+    },
+    {
+      "epoch": 0.27930007898414244,
+      "grad_norm": 0.15127043426036835,
+      "learning_rate": 8.222770023060006e-05,
+      "loss": 1.0861,
+      "step": 4597
+    },
+    {
+      "epoch": 0.2793608360167689,
+      "grad_norm": 0.17315421998500824,
+      "learning_rate": 8.222038069227383e-05,
+      "loss": 1.1027,
+      "step": 4598
+    },
+    {
+      "epoch": 0.27942159304939546,
+      "grad_norm": 0.29059749841690063,
+      "learning_rate": 8.221305997290329e-05,
+      "loss": 1.0508,
+      "step": 4599
+    },
+    {
+      "epoch": 0.279482350082022,
+      "grad_norm": 0.20000159740447998,
+      "learning_rate": 8.220573807275685e-05,
+      "loss": 1.1479,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2795431071146485,
+      "grad_norm": 0.14069730043411255,
+      "learning_rate": 8.219841499210283e-05,
+      "loss": 1.0821,
+      "step": 4601
+    },
+    {
+      "epoch": 0.27960386414727506,
+      "grad_norm": 0.22600381076335907,
+      "learning_rate": 8.219109073120967e-05,
+      "loss": 1.0802,
+      "step": 4602
+    },
+    {
+      "epoch": 0.2796646211799016,
+      "grad_norm": 0.47676920890808105,
+      "learning_rate": 8.218376529034589e-05,
+      "loss": 1.1481,
+      "step": 4603
+    },
+    {
+      "epoch": 0.2797253782125281,
+      "grad_norm": 0.178662970662117,
+      "learning_rate": 8.217643866977995e-05,
+      "loss": 1.1206,
+      "step": 4604
+    },
+    {
+      "epoch": 0.2797861352451546,
+      "grad_norm": 0.5735136270523071,
+      "learning_rate": 8.216911086978046e-05,
+      "loss": 1.1023,
+      "step": 4605
+    },
+    {
+      "epoch": 0.27984689227778115,
+      "grad_norm": 0.18864288926124573,
+      "learning_rate": 8.216178189061595e-05,
+      "loss": 1.219,
+      "step": 4606
+    },
+    {
+      "epoch": 0.2799076493104077,
+      "grad_norm": 0.22152753174304962,
+      "learning_rate": 8.215445173255514e-05,
+      "loss": 1.1029,
+      "step": 4607
+    },
+    {
+      "epoch": 0.2799684063430342,
+      "grad_norm": 0.28739672899246216,
+      "learning_rate": 8.214712039586667e-05,
+      "loss": 1.1797,
+      "step": 4608
+    },
+    {
+      "epoch": 0.28002916337566075,
+      "grad_norm": 0.26197290420532227,
+      "learning_rate": 8.21397878808193e-05,
+      "loss": 1.2171,
+      "step": 4609
+    },
+    {
+      "epoch": 0.2800899204082873,
+      "grad_norm": 0.4948214292526245,
+      "learning_rate": 8.213245418768179e-05,
+      "loss": 1.2254,
+      "step": 4610
+    },
+    {
+      "epoch": 0.28015067744091376,
+      "grad_norm": 0.260137140750885,
+      "learning_rate": 8.212511931672296e-05,
+      "loss": 1.073,
+      "step": 4611
+    },
+    {
+      "epoch": 0.2802114344735403,
+      "grad_norm": 0.19344522058963776,
+      "learning_rate": 8.211778326821166e-05,
+      "loss": 1.2006,
+      "step": 4612
+    },
+    {
+      "epoch": 0.28027219150616683,
+      "grad_norm": 0.23450690507888794,
+      "learning_rate": 8.211044604241681e-05,
+      "loss": 1.1887,
+      "step": 4613
+    },
+    {
+      "epoch": 0.28033294853879337,
+      "grad_norm": 0.24577519297599792,
+      "learning_rate": 8.210310763960737e-05,
+      "loss": 1.1533,
+      "step": 4614
+    },
+    {
+      "epoch": 0.2803937055714199,
+      "grad_norm": 0.3236617147922516,
+      "learning_rate": 8.20957680600523e-05,
+      "loss": 1.1662,
+      "step": 4615
+    },
+    {
+      "epoch": 0.28045446260404644,
+      "grad_norm": 0.29707232117652893,
+      "learning_rate": 8.208842730402065e-05,
+      "loss": 1.184,
+      "step": 4616
+    },
+    {
+      "epoch": 0.2805152196366729,
+      "grad_norm": 0.24683883786201477,
+      "learning_rate": 8.208108537178149e-05,
+      "loss": 1.1164,
+      "step": 4617
+    },
+    {
+      "epoch": 0.28057597666929945,
+      "grad_norm": 0.1817791759967804,
+      "learning_rate": 8.207374226360396e-05,
+      "loss": 1.1824,
+      "step": 4618
+    },
+    {
+      "epoch": 0.280636733701926,
+      "grad_norm": 0.22569699585437775,
+      "learning_rate": 8.206639797975719e-05,
+      "loss": 1.0921,
+      "step": 4619
+    },
+    {
+      "epoch": 0.2806974907345525,
+      "grad_norm": 0.6512724757194519,
+      "learning_rate": 8.205905252051041e-05,
+      "loss": 1.0864,
+      "step": 4620
+    },
+    {
+      "epoch": 0.28075824776717906,
+      "grad_norm": 0.3657459318637848,
+      "learning_rate": 8.205170588613286e-05,
+      "loss": 1.1494,
+      "step": 4621
+    },
+    {
+      "epoch": 0.2808190047998056,
+      "grad_norm": 0.1689261794090271,
+      "learning_rate": 8.204435807689385e-05,
+      "loss": 1.0679,
+      "step": 4622
+    },
+    {
+      "epoch": 0.28087976183243213,
+      "grad_norm": 0.18063753843307495,
+      "learning_rate": 8.203700909306269e-05,
+      "loss": 1.0988,
+      "step": 4623
+    },
+    {
+      "epoch": 0.2809405188650586,
+      "grad_norm": 1.233905553817749,
+      "learning_rate": 8.202965893490878e-05,
+      "loss": 1.0576,
+      "step": 4624
+    },
+    {
+      "epoch": 0.28100127589768514,
+      "grad_norm": 0.8604289889335632,
+      "learning_rate": 8.202230760270151e-05,
+      "loss": 1.0996,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2810620329303117,
+      "grad_norm": 0.28963378071784973,
+      "learning_rate": 8.201495509671037e-05,
+      "loss": 1.0629,
+      "step": 4626
+    },
+    {
+      "epoch": 0.2811227899629382,
+      "grad_norm": 0.2162565141916275,
+      "learning_rate": 8.200760141720486e-05,
+      "loss": 1.2243,
+      "step": 4627
+    },
+    {
+      "epoch": 0.28118354699556475,
+      "grad_norm": 0.30542391538619995,
+      "learning_rate": 8.200024656445455e-05,
+      "loss": 1.0582,
+      "step": 4628
+    },
+    {
+      "epoch": 0.2812443040281913,
+      "grad_norm": 0.32875868678092957,
+      "learning_rate": 8.1992890538729e-05,
+      "loss": 1.1191,
+      "step": 4629
+    },
+    {
+      "epoch": 0.28130506106081776,
+      "grad_norm": 0.47611168026924133,
+      "learning_rate": 8.198553334029786e-05,
+      "loss": 1.0689,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2813658180934443,
+      "grad_norm": 0.38822394609451294,
+      "learning_rate": 8.197817496943084e-05,
+      "loss": 1.0782,
+      "step": 4631
+    },
+    {
+      "epoch": 0.28142657512607083,
+      "grad_norm": 0.2146041989326477,
+      "learning_rate": 8.197081542639762e-05,
+      "loss": 1.0429,
+      "step": 4632
+    },
+    {
+      "epoch": 0.28148733215869737,
+      "grad_norm": 0.2552999258041382,
+      "learning_rate": 8.196345471146797e-05,
+      "loss": 1.1339,
+      "step": 4633
+    },
+    {
+      "epoch": 0.2815480891913239,
+      "grad_norm": 0.2688206136226654,
+      "learning_rate": 8.195609282491172e-05,
+      "loss": 1.0717,
+      "step": 4634
+    },
+    {
+      "epoch": 0.28160884622395044,
+      "grad_norm": 0.5100186467170715,
+      "learning_rate": 8.194872976699871e-05,
+      "loss": 1.1656,
+      "step": 4635
+    },
+    {
+      "epoch": 0.281669603256577,
+      "grad_norm": 0.18688523769378662,
+      "learning_rate": 8.194136553799883e-05,
+      "loss": 1.0769,
+      "step": 4636
+    },
+    {
+      "epoch": 0.28173036028920345,
+      "grad_norm": 7.792951583862305,
+      "learning_rate": 8.193400013818202e-05,
+      "loss": 1.206,
+      "step": 4637
+    },
+    {
+      "epoch": 0.28179111732183,
+      "grad_norm": 0.29357993602752686,
+      "learning_rate": 8.192663356781827e-05,
+      "loss": 1.2129,
+      "step": 4638
+    },
+    {
+      "epoch": 0.2818518743544565,
+      "grad_norm": 0.26707616448402405,
+      "learning_rate": 8.19192658271776e-05,
+      "loss": 1.1004,
+      "step": 4639
+    },
+    {
+      "epoch": 0.28191263138708306,
+      "grad_norm": 0.18493105471134186,
+      "learning_rate": 8.191189691653006e-05,
+      "loss": 1.0836,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2819733884197096,
+      "grad_norm": 0.2140301764011383,
+      "learning_rate": 8.190452683614577e-05,
+      "loss": 1.1302,
+      "step": 4641
+    },
+    {
+      "epoch": 0.2820341454523361,
+      "grad_norm": 0.21024712920188904,
+      "learning_rate": 8.18971555862949e-05,
+      "loss": 1.0932,
+      "step": 4642
+    },
+    {
+      "epoch": 0.28209490248496266,
+      "grad_norm": 0.2261425405740738,
+      "learning_rate": 8.188978316724761e-05,
+      "loss": 1.0946,
+      "step": 4643
+    },
+    {
+      "epoch": 0.28215565951758914,
+      "grad_norm": 0.18732121586799622,
+      "learning_rate": 8.188240957927416e-05,
+      "loss": 1.1945,
+      "step": 4644
+    },
+    {
+      "epoch": 0.2822164165502157,
+      "grad_norm": 0.2324189692735672,
+      "learning_rate": 8.187503482264482e-05,
+      "loss": 1.3628,
+      "step": 4645
+    },
+    {
+      "epoch": 0.2822771735828422,
+      "grad_norm": 0.1965351551771164,
+      "learning_rate": 8.18676588976299e-05,
+      "loss": 1.1673,
+      "step": 4646
+    },
+    {
+      "epoch": 0.28233793061546875,
+      "grad_norm": 0.17932644486427307,
+      "learning_rate": 8.186028180449981e-05,
+      "loss": 1.0744,
+      "step": 4647
+    },
+    {
+      "epoch": 0.2823986876480953,
+      "grad_norm": 0.4422328472137451,
+      "learning_rate": 8.185290354352493e-05,
+      "loss": 1.2956,
+      "step": 4648
+    },
+    {
+      "epoch": 0.2824594446807218,
+      "grad_norm": 0.24107696115970612,
+      "learning_rate": 8.18455241149757e-05,
+      "loss": 1.1298,
+      "step": 4649
+    },
+    {
+      "epoch": 0.2825202017133483,
+      "grad_norm": 0.22162525355815887,
+      "learning_rate": 8.183814351912265e-05,
+      "loss": 1.1265,
+      "step": 4650
+    },
+    {
+      "epoch": 0.28258095874597483,
+      "grad_norm": 0.39656251668930054,
+      "learning_rate": 8.183076175623628e-05,
+      "loss": 1.2503,
+      "step": 4651
+    },
+    {
+      "epoch": 0.28264171577860137,
+      "grad_norm": 0.2623554766178131,
+      "learning_rate": 8.182337882658717e-05,
+      "loss": 1.1311,
+      "step": 4652
+    },
+    {
+      "epoch": 0.2827024728112279,
+      "grad_norm": 0.18301978707313538,
+      "learning_rate": 8.181599473044598e-05,
+      "loss": 1.2532,
+      "step": 4653
+    },
+    {
+      "epoch": 0.28276322984385444,
+      "grad_norm": 1.0720328092575073,
+      "learning_rate": 8.180860946808336e-05,
+      "loss": 1.5338,
+      "step": 4654
+    },
+    {
+      "epoch": 0.28282398687648097,
+      "grad_norm": 0.38734033703804016,
+      "learning_rate": 8.180122303976999e-05,
+      "loss": 1.1591,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2828847439091075,
+      "grad_norm": 0.5955361723899841,
+      "learning_rate": 8.179383544577664e-05,
+      "loss": 1.2536,
+      "step": 4656
+    },
+    {
+      "epoch": 0.282945500941734,
+      "grad_norm": 0.19823609292507172,
+      "learning_rate": 8.178644668637412e-05,
+      "loss": 1.0633,
+      "step": 4657
+    },
+    {
+      "epoch": 0.2830062579743605,
+      "grad_norm": 0.4541545510292053,
+      "learning_rate": 8.177905676183325e-05,
+      "loss": 1.1511,
+      "step": 4658
+    },
+    {
+      "epoch": 0.28306701500698705,
+      "grad_norm": 0.2795678675174713,
+      "learning_rate": 8.177166567242491e-05,
+      "loss": 1.1396,
+      "step": 4659
+    },
+    {
+      "epoch": 0.2831277720396136,
+      "grad_norm": 0.2806314527988434,
+      "learning_rate": 8.176427341842001e-05,
+      "loss": 1.1468,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2831885290722401,
+      "grad_norm": 0.3850073218345642,
+      "learning_rate": 8.175688000008956e-05,
+      "loss": 1.1084,
+      "step": 4661
+    },
+    {
+      "epoch": 0.28324928610486666,
+      "grad_norm": 0.23455089330673218,
+      "learning_rate": 8.17494854177045e-05,
+      "loss": 1.1676,
+      "step": 4662
+    },
+    {
+      "epoch": 0.28331004313749314,
+      "grad_norm": 0.28819727897644043,
+      "learning_rate": 8.174208967153593e-05,
+      "loss": 1.2905,
+      "step": 4663
+    },
+    {
+      "epoch": 0.2833708001701197,
+      "grad_norm": 0.28520601987838745,
+      "learning_rate": 8.173469276185495e-05,
+      "loss": 1.1198,
+      "step": 4664
+    },
+    {
+      "epoch": 0.2834315572027462,
+      "grad_norm": 0.16577650606632233,
+      "learning_rate": 8.172729468893264e-05,
+      "loss": 1.1215,
+      "step": 4665
+    },
+    {
+      "epoch": 0.28349231423537274,
+      "grad_norm": 0.35518643260002136,
+      "learning_rate": 8.171989545304022e-05,
+      "loss": 1.3198,
+      "step": 4666
+    },
+    {
+      "epoch": 0.2835530712679993,
+      "grad_norm": 0.19905826449394226,
+      "learning_rate": 8.171249505444891e-05,
+      "loss": 1.1921,
+      "step": 4667
+    },
+    {
+      "epoch": 0.2836138283006258,
+      "grad_norm": 0.1744869202375412,
+      "learning_rate": 8.170509349342996e-05,
+      "loss": 1.081,
+      "step": 4668
+    },
+    {
+      "epoch": 0.28367458533325235,
+      "grad_norm": 0.26516008377075195,
+      "learning_rate": 8.169769077025466e-05,
+      "loss": 1.0577,
+      "step": 4669
+    },
+    {
+      "epoch": 0.28373534236587883,
+      "grad_norm": 0.17361098527908325,
+      "learning_rate": 8.16902868851944e-05,
+      "loss": 1.1093,
+      "step": 4670
+    },
+    {
+      "epoch": 0.28379609939850536,
+      "grad_norm": 0.29466310143470764,
+      "learning_rate": 8.168288183852055e-05,
+      "loss": 1.0544,
+      "step": 4671
+    },
+    {
+      "epoch": 0.2838568564311319,
+      "grad_norm": 0.2904646098613739,
+      "learning_rate": 8.167547563050452e-05,
+      "loss": 1.2464,
+      "step": 4672
+    },
+    {
+      "epoch": 0.28391761346375843,
+      "grad_norm": 0.3171013295650482,
+      "learning_rate": 8.166806826141784e-05,
+      "loss": 1.0834,
+      "step": 4673
+    },
+    {
+      "epoch": 0.28397837049638497,
+      "grad_norm": 0.20559333264827728,
+      "learning_rate": 8.166065973153199e-05,
+      "loss": 1.1319,
+      "step": 4674
+    },
+    {
+      "epoch": 0.2840391275290115,
+      "grad_norm": 1.1493704319000244,
+      "learning_rate": 8.165325004111852e-05,
+      "loss": 1.1405,
+      "step": 4675
+    },
+    {
+      "epoch": 0.284099884561638,
+      "grad_norm": 0.477880597114563,
+      "learning_rate": 8.164583919044906e-05,
+      "loss": 1.075,
+      "step": 4676
+    },
+    {
+      "epoch": 0.2841606415942645,
+      "grad_norm": 0.16275012493133545,
+      "learning_rate": 8.163842717979525e-05,
+      "loss": 1.0794,
+      "step": 4677
+    },
+    {
+      "epoch": 0.28422139862689105,
+      "grad_norm": 0.21456238627433777,
+      "learning_rate": 8.163101400942877e-05,
+      "loss": 1.1095,
+      "step": 4678
+    },
+    {
+      "epoch": 0.2842821556595176,
+      "grad_norm": 0.17840400338172913,
+      "learning_rate": 8.162359967962134e-05,
+      "loss": 1.1096,
+      "step": 4679
+    },
+    {
+      "epoch": 0.2843429126921441,
+      "grad_norm": 0.20673146843910217,
+      "learning_rate": 8.161618419064478e-05,
+      "loss": 1.1351,
+      "step": 4680
+    },
+    {
+      "epoch": 0.28440366972477066,
+      "grad_norm": 0.185395747423172,
+      "learning_rate": 8.160876754277087e-05,
+      "loss": 1.0876,
+      "step": 4681
+    },
+    {
+      "epoch": 0.2844644267573972,
+      "grad_norm": 0.2899877727031708,
+      "learning_rate": 8.160134973627147e-05,
+      "loss": 1.1328,
+      "step": 4682
+    },
+    {
+      "epoch": 0.2845251837900237,
+      "grad_norm": 0.13526462018489838,
+      "learning_rate": 8.159393077141849e-05,
+      "loss": 1.0842,
+      "step": 4683
+    },
+    {
+      "epoch": 0.2845859408226502,
+      "grad_norm": 0.19481588900089264,
+      "learning_rate": 8.158651064848387e-05,
+      "loss": 1.249,
+      "step": 4684
+    },
+    {
+      "epoch": 0.28464669785527674,
+      "grad_norm": 0.3543314039707184,
+      "learning_rate": 8.157908936773961e-05,
+      "loss": 1.0935,
+      "step": 4685
+    },
+    {
+      "epoch": 0.2847074548879033,
+      "grad_norm": 0.28960728645324707,
+      "learning_rate": 8.15716669294577e-05,
+      "loss": 1.1669,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2847682119205298,
+      "grad_norm": 0.22876624763011932,
+      "learning_rate": 8.156424333391025e-05,
+      "loss": 1.1913,
+      "step": 4687
+    },
+    {
+      "epoch": 0.28482896895315635,
+      "grad_norm": 0.40697211027145386,
+      "learning_rate": 8.155681858136937e-05,
+      "loss": 1.0249,
+      "step": 4688
+    },
+    {
+      "epoch": 0.2848897259857828,
+      "grad_norm": 0.22622931003570557,
+      "learning_rate": 8.154939267210719e-05,
+      "loss": 1.0453,
+      "step": 4689
+    },
+    {
+      "epoch": 0.28495048301840936,
+      "grad_norm": 0.1916540116071701,
+      "learning_rate": 8.154196560639592e-05,
+      "loss": 1.1253,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2850112400510359,
+      "grad_norm": 2.1343138217926025,
+      "learning_rate": 8.153453738450782e-05,
+      "loss": 1.0774,
+      "step": 4691
+    },
+    {
+      "epoch": 0.28507199708366243,
+      "grad_norm": 0.1893221139907837,
+      "learning_rate": 8.152710800671515e-05,
+      "loss": 1.1148,
+      "step": 4692
+    },
+    {
+      "epoch": 0.28513275411628897,
+      "grad_norm": 0.18746380507946014,
+      "learning_rate": 8.151967747329024e-05,
+      "loss": 1.1496,
+      "step": 4693
+    },
+    {
+      "epoch": 0.2851935111489155,
+      "grad_norm": 0.35583555698394775,
+      "learning_rate": 8.151224578450545e-05,
+      "loss": 1.1697,
+      "step": 4694
+    },
+    {
+      "epoch": 0.28525426818154204,
+      "grad_norm": 0.27671295404434204,
+      "learning_rate": 8.150481294063321e-05,
+      "loss": 1.0999,
+      "step": 4695
+    },
+    {
+      "epoch": 0.2853150252141685,
+      "grad_norm": 0.2466990202665329,
+      "learning_rate": 8.149737894194596e-05,
+      "loss": 1.0794,
+      "step": 4696
+    },
+    {
+      "epoch": 0.28537578224679505,
+      "grad_norm": 0.213019460439682,
+      "learning_rate": 8.14899437887162e-05,
+      "loss": 1.1749,
+      "step": 4697
+    },
+    {
+      "epoch": 0.2854365392794216,
+      "grad_norm": 0.26015639305114746,
+      "learning_rate": 8.148250748121646e-05,
+      "loss": 1.1433,
+      "step": 4698
+    },
+    {
+      "epoch": 0.2854972963120481,
+      "grad_norm": 0.2658534049987793,
+      "learning_rate": 8.147507001971933e-05,
+      "loss": 1.2109,
+      "step": 4699
+    },
+    {
+      "epoch": 0.28555805334467466,
+      "grad_norm": 0.35617485642433167,
+      "learning_rate": 8.14676314044974e-05,
+      "loss": 1.0989,
+      "step": 4700
+    },
+    {
+      "epoch": 0.2856188103773012,
+      "grad_norm": 0.18339228630065918,
+      "learning_rate": 8.146019163582339e-05,
+      "loss": 1.1722,
+      "step": 4701
+    },
+    {
+      "epoch": 0.2856795674099277,
+      "grad_norm": 0.24170809984207153,
+      "learning_rate": 8.145275071396996e-05,
+      "loss": 1.1733,
+      "step": 4702
+    },
+    {
+      "epoch": 0.2857403244425542,
+      "grad_norm": 0.19103199243545532,
+      "learning_rate": 8.144530863920987e-05,
+      "loss": 1.2067,
+      "step": 4703
+    },
+    {
+      "epoch": 0.28580108147518074,
+      "grad_norm": 0.20725975930690765,
+      "learning_rate": 8.143786541181591e-05,
+      "loss": 1.0453,
+      "step": 4704
+    },
+    {
+      "epoch": 0.2858618385078073,
+      "grad_norm": 0.17264272272586823,
+      "learning_rate": 8.143042103206094e-05,
+      "loss": 1.0208,
+      "step": 4705
+    },
+    {
+      "epoch": 0.2859225955404338,
+      "grad_norm": 0.1455267369747162,
+      "learning_rate": 8.142297550021777e-05,
+      "loss": 1.0559,
+      "step": 4706
+    },
+    {
+      "epoch": 0.28598335257306035,
+      "grad_norm": 0.24420267343521118,
+      "learning_rate": 8.14155288165594e-05,
+      "loss": 1.0617,
+      "step": 4707
+    },
+    {
+      "epoch": 0.2860441096056869,
+      "grad_norm": 0.23790127038955688,
+      "learning_rate": 8.140808098135872e-05,
+      "loss": 1.0679,
+      "step": 4708
+    },
+    {
+      "epoch": 0.28610486663831336,
+      "grad_norm": 0.16236484050750732,
+      "learning_rate": 8.140063199488878e-05,
+      "loss": 1.0838,
+      "step": 4709
+    },
+    {
+      "epoch": 0.2861656236709399,
+      "grad_norm": 1.5107526779174805,
+      "learning_rate": 8.139318185742259e-05,
+      "loss": 1.198,
+      "step": 4710
+    },
+    {
+      "epoch": 0.28622638070356643,
+      "grad_norm": 0.29380860924720764,
+      "learning_rate": 8.138573056923328e-05,
+      "loss": 1.1327,
+      "step": 4711
+    },
+    {
+      "epoch": 0.28628713773619296,
+      "grad_norm": 2.3719253540039062,
+      "learning_rate": 8.137827813059392e-05,
+      "loss": 1.1012,
+      "step": 4712
+    },
+    {
+      "epoch": 0.2863478947688195,
+      "grad_norm": 0.33579131960868835,
+      "learning_rate": 8.137082454177771e-05,
+      "loss": 1.2244,
+      "step": 4713
+    },
+    {
+      "epoch": 0.28640865180144603,
+      "grad_norm": 0.19710390269756317,
+      "learning_rate": 8.136336980305787e-05,
+      "loss": 1.057,
+      "step": 4714
+    },
+    {
+      "epoch": 0.28646940883407257,
+      "grad_norm": 0.2228107899427414,
+      "learning_rate": 8.135591391470766e-05,
+      "loss": 1.2096,
+      "step": 4715
+    },
+    {
+      "epoch": 0.28653016586669905,
+      "grad_norm": 0.19869717955589294,
+      "learning_rate": 8.134845687700035e-05,
+      "loss": 1.1458,
+      "step": 4716
+    },
+    {
+      "epoch": 0.2865909228993256,
+      "grad_norm": 0.19826026260852814,
+      "learning_rate": 8.13409986902093e-05,
+      "loss": 1.0752,
+      "step": 4717
+    },
+    {
+      "epoch": 0.2866516799319521,
+      "grad_norm": 0.24265635013580322,
+      "learning_rate": 8.133353935460788e-05,
+      "loss": 1.1447,
+      "step": 4718
+    },
+    {
+      "epoch": 0.28671243696457865,
+      "grad_norm": 0.295247346162796,
+      "learning_rate": 8.132607887046952e-05,
+      "loss": 1.1352,
+      "step": 4719
+    },
+    {
+      "epoch": 0.2867731939972052,
+      "grad_norm": 0.563098669052124,
+      "learning_rate": 8.131861723806769e-05,
+      "loss": 1.2249,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2868339510298317,
+      "grad_norm": 0.22320379316806793,
+      "learning_rate": 8.131115445767588e-05,
+      "loss": 1.0678,
+      "step": 4721
+    },
+    {
+      "epoch": 0.2868947080624582,
+      "grad_norm": 0.23385919630527496,
+      "learning_rate": 8.130369052956765e-05,
+      "loss": 1.0941,
+      "step": 4722
+    },
+    {
+      "epoch": 0.28695546509508474,
+      "grad_norm": 0.22694692015647888,
+      "learning_rate": 8.129622545401661e-05,
+      "loss": 1.062,
+      "step": 4723
+    },
+    {
+      "epoch": 0.2870162221277113,
+      "grad_norm": 0.205569326877594,
+      "learning_rate": 8.128875923129635e-05,
+      "loss": 1.0572,
+      "step": 4724
+    },
+    {
+      "epoch": 0.2870769791603378,
+      "grad_norm": 0.7134333252906799,
+      "learning_rate": 8.12812918616806e-05,
+      "loss": 1.0946,
+      "step": 4725
+    },
+    {
+      "epoch": 0.28713773619296434,
+      "grad_norm": 0.1782463788986206,
+      "learning_rate": 8.127382334544302e-05,
+      "loss": 1.0619,
+      "step": 4726
+    },
+    {
+      "epoch": 0.2871984932255909,
+      "grad_norm": 0.2327854186296463,
+      "learning_rate": 8.126635368285742e-05,
+      "loss": 1.2124,
+      "step": 4727
+    },
+    {
+      "epoch": 0.2872592502582174,
+      "grad_norm": 0.18850277364253998,
+      "learning_rate": 8.125888287419759e-05,
+      "loss": 1.1462,
+      "step": 4728
+    },
+    {
+      "epoch": 0.2873200072908439,
+      "grad_norm": 0.20325693488121033,
+      "learning_rate": 8.125141091973736e-05,
+      "loss": 1.24,
+      "step": 4729
+    },
+    {
+      "epoch": 0.28738076432347043,
+      "grad_norm": 0.23426564037799835,
+      "learning_rate": 8.124393781975062e-05,
+      "loss": 1.3633,
+      "step": 4730
+    },
+    {
+      "epoch": 0.28744152135609696,
+      "grad_norm": 0.17329590022563934,
+      "learning_rate": 8.123646357451129e-05,
+      "loss": 1.0398,
+      "step": 4731
+    },
+    {
+      "epoch": 0.2875022783887235,
+      "grad_norm": 0.2696012258529663,
+      "learning_rate": 8.122898818429336e-05,
+      "loss": 1.281,
+      "step": 4732
+    },
+    {
+      "epoch": 0.28756303542135003,
+      "grad_norm": 0.333278626203537,
+      "learning_rate": 8.122151164937083e-05,
+      "loss": 1.2321,
+      "step": 4733
+    },
+    {
+      "epoch": 0.28762379245397657,
+      "grad_norm": 0.22068017721176147,
+      "learning_rate": 8.121403397001774e-05,
+      "loss": 1.0782,
+      "step": 4734
+    },
+    {
+      "epoch": 0.28768454948660305,
+      "grad_norm": 0.19823525846004486,
+      "learning_rate": 8.120655514650824e-05,
+      "loss": 1.0738,
+      "step": 4735
+    },
+    {
+      "epoch": 0.2877453065192296,
+      "grad_norm": 0.3010410964488983,
+      "learning_rate": 8.11990751791164e-05,
+      "loss": 1.1292,
+      "step": 4736
+    },
+    {
+      "epoch": 0.2878060635518561,
+      "grad_norm": 0.1889469176530838,
+      "learning_rate": 8.119159406811643e-05,
+      "loss": 1.0807,
+      "step": 4737
+    },
+    {
+      "epoch": 0.28786682058448265,
+      "grad_norm": 0.21690398454666138,
+      "learning_rate": 8.118411181378256e-05,
+      "loss": 1.103,
+      "step": 4738
+    },
+    {
+      "epoch": 0.2879275776171092,
+      "grad_norm": 0.5330593585968018,
+      "learning_rate": 8.117662841638903e-05,
+      "loss": 1.344,
+      "step": 4739
+    },
+    {
+      "epoch": 0.2879883346497357,
+      "grad_norm": 0.19033487141132355,
+      "learning_rate": 8.116914387621018e-05,
+      "loss": 1.1315,
+      "step": 4740
+    },
+    {
+      "epoch": 0.28804909168236226,
+      "grad_norm": 0.293365478515625,
+      "learning_rate": 8.116165819352031e-05,
+      "loss": 1.1059,
+      "step": 4741
+    },
+    {
+      "epoch": 0.28810984871498874,
+      "grad_norm": 0.19638319313526154,
+      "learning_rate": 8.115417136859385e-05,
+      "loss": 1.1785,
+      "step": 4742
+    },
+    {
+      "epoch": 0.28817060574761527,
+      "grad_norm": 0.2521298825740814,
+      "learning_rate": 8.114668340170522e-05,
+      "loss": 1.085,
+      "step": 4743
+    },
+    {
+      "epoch": 0.2882313627802418,
+      "grad_norm": 0.1999048888683319,
+      "learning_rate": 8.113919429312888e-05,
+      "loss": 1.1324,
+      "step": 4744
+    },
+    {
+      "epoch": 0.28829211981286834,
+      "grad_norm": 0.49342015385627747,
+      "learning_rate": 8.113170404313937e-05,
+      "loss": 1.0628,
+      "step": 4745
+    },
+    {
+      "epoch": 0.2883528768454949,
+      "grad_norm": 0.1688714176416397,
+      "learning_rate": 8.112421265201121e-05,
+      "loss": 1.1498,
+      "step": 4746
+    },
+    {
+      "epoch": 0.2884136338781214,
+      "grad_norm": 0.6101064682006836,
+      "learning_rate": 8.111672012001904e-05,
+      "loss": 1.243,
+      "step": 4747
+    },
+    {
+      "epoch": 0.28847439091074795,
+      "grad_norm": 1.2826509475708008,
+      "learning_rate": 8.110922644743746e-05,
+      "loss": 1.1068,
+      "step": 4748
+    },
+    {
+      "epoch": 0.2885351479433744,
+      "grad_norm": 0.22259414196014404,
+      "learning_rate": 8.110173163454118e-05,
+      "loss": 1.1576,
+      "step": 4749
+    },
+    {
+      "epoch": 0.28859590497600096,
+      "grad_norm": 0.18059051036834717,
+      "learning_rate": 8.109423568160491e-05,
+      "loss": 1.0945,
+      "step": 4750
+    },
+    {
+      "epoch": 0.2886566620086275,
+      "grad_norm": 0.16577550768852234,
+      "learning_rate": 8.108673858890342e-05,
+      "loss": 1.1079,
+      "step": 4751
+    },
+    {
+      "epoch": 0.28871741904125403,
+      "grad_norm": 0.5138909220695496,
+      "learning_rate": 8.107924035671153e-05,
+      "loss": 1.1859,
+      "step": 4752
+    },
+    {
+      "epoch": 0.28877817607388057,
+      "grad_norm": 0.18410277366638184,
+      "learning_rate": 8.107174098530406e-05,
+      "loss": 1.1216,
+      "step": 4753
+    },
+    {
+      "epoch": 0.2888389331065071,
+      "grad_norm": 0.2318829894065857,
+      "learning_rate": 8.106424047495593e-05,
+      "loss": 1.1795,
+      "step": 4754
+    },
+    {
+      "epoch": 0.2888996901391336,
+      "grad_norm": 0.14606933295726776,
+      "learning_rate": 8.105673882594207e-05,
+      "loss": 1.0248,
+      "step": 4755
+    },
+    {
+      "epoch": 0.2889604471717601,
+      "grad_norm": 0.14447882771492004,
+      "learning_rate": 8.104923603853743e-05,
+      "loss": 1.0765,
+      "step": 4756
+    },
+    {
+      "epoch": 0.28902120420438665,
+      "grad_norm": 0.1729961633682251,
+      "learning_rate": 8.104173211301705e-05,
+      "loss": 1.0789,
+      "step": 4757
+    },
+    {
+      "epoch": 0.2890819612370132,
+      "grad_norm": 0.1393560916185379,
+      "learning_rate": 8.103422704965597e-05,
+      "loss": 1.0379,
+      "step": 4758
+    },
+    {
+      "epoch": 0.2891427182696397,
+      "grad_norm": 0.2100219875574112,
+      "learning_rate": 8.10267208487293e-05,
+      "loss": 1.3307,
+      "step": 4759
+    },
+    {
+      "epoch": 0.28920347530226626,
+      "grad_norm": 0.3587530553340912,
+      "learning_rate": 8.10192135105122e-05,
+      "loss": 1.3124,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2892642323348928,
+      "grad_norm": 0.19232696294784546,
+      "learning_rate": 8.10117050352798e-05,
+      "loss": 1.0873,
+      "step": 4761
+    },
+    {
+      "epoch": 0.28932498936751927,
+      "grad_norm": 0.16133850812911987,
+      "learning_rate": 8.100419542330738e-05,
+      "loss": 1.0677,
+      "step": 4762
+    },
+    {
+      "epoch": 0.2893857464001458,
+      "grad_norm": 0.12450205534696579,
+      "learning_rate": 8.099668467487018e-05,
+      "loss": 1.0577,
+      "step": 4763
+    },
+    {
+      "epoch": 0.28944650343277234,
+      "grad_norm": 0.9260332584381104,
+      "learning_rate": 8.09891727902435e-05,
+      "loss": 1.1521,
+      "step": 4764
+    },
+    {
+      "epoch": 0.2895072604653989,
+      "grad_norm": 0.1659279763698578,
+      "learning_rate": 8.098165976970273e-05,
+      "loss": 1.0509,
+      "step": 4765
+    },
+    {
+      "epoch": 0.2895680174980254,
+      "grad_norm": 0.18438123166561127,
+      "learning_rate": 8.097414561352321e-05,
+      "loss": 1.0343,
+      "step": 4766
+    },
+    {
+      "epoch": 0.28962877453065194,
+      "grad_norm": 0.14608722925186157,
+      "learning_rate": 8.09666303219804e-05,
+      "loss": 1.115,
+      "step": 4767
+    },
+    {
+      "epoch": 0.2896895315632784,
+      "grad_norm": 0.14206336438655853,
+      "learning_rate": 8.095911389534977e-05,
+      "loss": 1.0897,
+      "step": 4768
+    },
+    {
+      "epoch": 0.28975028859590496,
+      "grad_norm": 0.12960398197174072,
+      "learning_rate": 8.095159633390683e-05,
+      "loss": 1.0864,
+      "step": 4769
+    },
+    {
+      "epoch": 0.2898110456285315,
+      "grad_norm": 0.1357732117176056,
+      "learning_rate": 8.094407763792714e-05,
+      "loss": 1.0866,
+      "step": 4770
+    },
+    {
+      "epoch": 0.28987180266115803,
+      "grad_norm": 0.16963078081607819,
+      "learning_rate": 8.093655780768632e-05,
+      "loss": 1.1873,
+      "step": 4771
+    },
+    {
+      "epoch": 0.28993255969378456,
+      "grad_norm": 0.17405863106250763,
+      "learning_rate": 8.092903684345999e-05,
+      "loss": 1.0914,
+      "step": 4772
+    },
+    {
+      "epoch": 0.2899933167264111,
+      "grad_norm": 0.20222699642181396,
+      "learning_rate": 8.092151474552381e-05,
+      "loss": 1.0646,
+      "step": 4773
+    },
+    {
+      "epoch": 0.29005407375903763,
+      "grad_norm": 0.19266146421432495,
+      "learning_rate": 8.091399151415355e-05,
+      "loss": 1.0609,
+      "step": 4774
+    },
+    {
+      "epoch": 0.2901148307916641,
+      "grad_norm": 0.145595520734787,
+      "learning_rate": 8.090646714962497e-05,
+      "loss": 1.0504,
+      "step": 4775
+    },
+    {
+      "epoch": 0.29017558782429065,
+      "grad_norm": 0.17000088095664978,
+      "learning_rate": 8.089894165221386e-05,
+      "loss": 1.0952,
+      "step": 4776
+    },
+    {
+      "epoch": 0.2902363448569172,
+      "grad_norm": 0.17131289839744568,
+      "learning_rate": 8.089141502219603e-05,
+      "loss": 1.0528,
+      "step": 4777
+    },
+    {
+      "epoch": 0.2902971018895437,
+      "grad_norm": 0.1466439813375473,
+      "learning_rate": 8.088388725984745e-05,
+      "loss": 1.0521,
+      "step": 4778
+    },
+    {
+      "epoch": 0.29035785892217025,
+      "grad_norm": 0.16580042243003845,
+      "learning_rate": 8.0876358365444e-05,
+      "loss": 1.1028,
+      "step": 4779
+    },
+    {
+      "epoch": 0.2904186159547968,
+      "grad_norm": 0.24699246883392334,
+      "learning_rate": 8.086882833926168e-05,
+      "loss": 1.1469,
+      "step": 4780
+    },
+    {
+      "epoch": 0.29047937298742327,
+      "grad_norm": 0.18186621367931366,
+      "learning_rate": 8.086129718157648e-05,
+      "loss": 1.1523,
+      "step": 4781
+    },
+    {
+      "epoch": 0.2905401300200498,
+      "grad_norm": 0.15710754692554474,
+      "learning_rate": 8.085376489266447e-05,
+      "loss": 1.0283,
+      "step": 4782
+    },
+    {
+      "epoch": 0.29060088705267634,
+      "grad_norm": 0.2273917943239212,
+      "learning_rate": 8.084623147280173e-05,
+      "loss": 1.1856,
+      "step": 4783
+    },
+    {
+      "epoch": 0.2906616440853029,
+      "grad_norm": 0.21383728086948395,
+      "learning_rate": 8.083869692226442e-05,
+      "loss": 1.1517,
+      "step": 4784
+    },
+    {
+      "epoch": 0.2907224011179294,
+      "grad_norm": 0.22727559506893158,
+      "learning_rate": 8.083116124132873e-05,
+      "loss": 1.1662,
+      "step": 4785
+    },
+    {
+      "epoch": 0.29078315815055594,
+      "grad_norm": 0.18458192050457,
+      "learning_rate": 8.082362443027085e-05,
+      "loss": 1.1141,
+      "step": 4786
+    },
+    {
+      "epoch": 0.2908439151831825,
+      "grad_norm": 2.501112699508667,
+      "learning_rate": 8.081608648936707e-05,
+      "loss": 1.2255,
+      "step": 4787
+    },
+    {
+      "epoch": 0.29090467221580896,
+      "grad_norm": 0.2777290642261505,
+      "learning_rate": 8.080854741889367e-05,
+      "loss": 1.114,
+      "step": 4788
+    },
+    {
+      "epoch": 0.2909654292484355,
+      "grad_norm": 0.13057801127433777,
+      "learning_rate": 8.080100721912702e-05,
+      "loss": 1.0318,
+      "step": 4789
+    },
+    {
+      "epoch": 0.291026186281062,
+      "grad_norm": 0.14138005673885345,
+      "learning_rate": 8.079346589034347e-05,
+      "loss": 1.0856,
+      "step": 4790
+    },
+    {
+      "epoch": 0.29108694331368856,
+      "grad_norm": 0.628367006778717,
+      "learning_rate": 8.07859234328195e-05,
+      "loss": 1.0691,
+      "step": 4791
+    },
+    {
+      "epoch": 0.2911477003463151,
+      "grad_norm": 0.1583787351846695,
+      "learning_rate": 8.077837984683155e-05,
+      "loss": 1.061,
+      "step": 4792
+    },
+    {
+      "epoch": 0.29120845737894163,
+      "grad_norm": 0.1693117320537567,
+      "learning_rate": 8.077083513265613e-05,
+      "loss": 1.1425,
+      "step": 4793
+    },
+    {
+      "epoch": 0.2912692144115681,
+      "grad_norm": 0.19208115339279175,
+      "learning_rate": 8.076328929056982e-05,
+      "loss": 1.1139,
+      "step": 4794
+    },
+    {
+      "epoch": 0.29132997144419465,
+      "grad_norm": 0.14601236581802368,
+      "learning_rate": 8.075574232084918e-05,
+      "loss": 1.0764,
+      "step": 4795
+    },
+    {
+      "epoch": 0.2913907284768212,
+      "grad_norm": 0.1709860861301422,
+      "learning_rate": 8.074819422377086e-05,
+      "loss": 1.1293,
+      "step": 4796
+    },
+    {
+      "epoch": 0.2914514855094477,
+      "grad_norm": 0.16490888595581055,
+      "learning_rate": 8.074064499961153e-05,
+      "loss": 1.0842,
+      "step": 4797
+    },
+    {
+      "epoch": 0.29151224254207425,
+      "grad_norm": 0.29872190952301025,
+      "learning_rate": 8.073309464864794e-05,
+      "loss": 1.2023,
+      "step": 4798
+    },
+    {
+      "epoch": 0.2915729995747008,
+      "grad_norm": 0.21467891335487366,
+      "learning_rate": 8.072554317115681e-05,
+      "loss": 1.0675,
+      "step": 4799
+    },
+    {
+      "epoch": 0.2916337566073273,
+      "grad_norm": 0.3493388295173645,
+      "learning_rate": 8.071799056741495e-05,
+      "loss": 1.1085,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2916945136399538,
+      "grad_norm": 0.14783447980880737,
+      "learning_rate": 8.071043683769922e-05,
+      "loss": 1.1098,
+      "step": 4801
+    },
+    {
+      "epoch": 0.29175527067258034,
+      "grad_norm": 0.1892547458410263,
+      "learning_rate": 8.070288198228648e-05,
+      "loss": 1.1149,
+      "step": 4802
+    },
+    {
+      "epoch": 0.29181602770520687,
+      "grad_norm": 0.1804693639278412,
+      "learning_rate": 8.069532600145368e-05,
+      "loss": 1.2042,
+      "step": 4803
+    },
+    {
+      "epoch": 0.2918767847378334,
+      "grad_norm": 0.17891071736812592,
+      "learning_rate": 8.068776889547777e-05,
+      "loss": 1.1429,
+      "step": 4804
+    },
+    {
+      "epoch": 0.29193754177045994,
+      "grad_norm": 0.1375022828578949,
+      "learning_rate": 8.068021066463575e-05,
+      "loss": 1.0472,
+      "step": 4805
+    },
+    {
+      "epoch": 0.2919982988030865,
+      "grad_norm": 0.14984142780303955,
+      "learning_rate": 8.067265130920468e-05,
+      "loss": 1.0789,
+      "step": 4806
+    },
+    {
+      "epoch": 0.292059055835713,
+      "grad_norm": 0.2128085345029831,
+      "learning_rate": 8.066509082946166e-05,
+      "loss": 1.1318,
+      "step": 4807
+    },
+    {
+      "epoch": 0.2921198128683395,
+      "grad_norm": 0.15035560727119446,
+      "learning_rate": 8.06575292256838e-05,
+      "loss": 1.0993,
+      "step": 4808
+    },
+    {
+      "epoch": 0.292180569900966,
+      "grad_norm": 0.17657846212387085,
+      "learning_rate": 8.064996649814827e-05,
+      "loss": 1.1598,
+      "step": 4809
+    },
+    {
+      "epoch": 0.29224132693359256,
+      "grad_norm": 0.16894984245300293,
+      "learning_rate": 8.06424026471323e-05,
+      "loss": 1.0638,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2923020839662191,
+      "grad_norm": 0.4684833288192749,
+      "learning_rate": 8.063483767291313e-05,
+      "loss": 1.1031,
+      "step": 4811
+    },
+    {
+      "epoch": 0.29236284099884563,
+      "grad_norm": 0.2029121220111847,
+      "learning_rate": 8.062727157576808e-05,
+      "loss": 1.0987,
+      "step": 4812
+    },
+    {
+      "epoch": 0.29242359803147217,
+      "grad_norm": 0.1746753752231598,
+      "learning_rate": 8.061970435597447e-05,
+      "loss": 1.105,
+      "step": 4813
+    },
+    {
+      "epoch": 0.29248435506409864,
+      "grad_norm": 0.16286218166351318,
+      "learning_rate": 8.061213601380966e-05,
+      "loss": 1.0918,
+      "step": 4814
+    },
+    {
+      "epoch": 0.2925451120967252,
+      "grad_norm": 0.12893499433994293,
+      "learning_rate": 8.06045665495511e-05,
+      "loss": 1.0189,
+      "step": 4815
+    },
+    {
+      "epoch": 0.2926058691293517,
+      "grad_norm": 0.19703014194965363,
+      "learning_rate": 8.059699596347625e-05,
+      "loss": 1.1905,
+      "step": 4816
+    },
+    {
+      "epoch": 0.29266662616197825,
+      "grad_norm": 0.16384656727313995,
+      "learning_rate": 8.058942425586258e-05,
+      "loss": 1.0969,
+      "step": 4817
+    },
+    {
+      "epoch": 0.2927273831946048,
+      "grad_norm": 0.18788295984268188,
+      "learning_rate": 8.058185142698765e-05,
+      "loss": 1.0691,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2927881402272313,
+      "grad_norm": 0.1627199947834015,
+      "learning_rate": 8.057427747712905e-05,
+      "loss": 1.0965,
+      "step": 4819
+    },
+    {
+      "epoch": 0.29284889725985785,
+      "grad_norm": 0.24151472747325897,
+      "learning_rate": 8.05667024065644e-05,
+      "loss": 1.0922,
+      "step": 4820
+    },
+    {
+      "epoch": 0.29290965429248433,
+      "grad_norm": 0.16679595410823822,
+      "learning_rate": 8.055912621557136e-05,
+      "loss": 1.1382,
+      "step": 4821
+    },
+    {
+      "epoch": 0.29297041132511087,
+      "grad_norm": 0.32155275344848633,
+      "learning_rate": 8.055154890442766e-05,
+      "loss": 1.0255,
+      "step": 4822
+    },
+    {
+      "epoch": 0.2930311683577374,
+      "grad_norm": 0.11976300179958344,
+      "learning_rate": 8.0543970473411e-05,
+      "loss": 1.0184,
+      "step": 4823
+    },
+    {
+      "epoch": 0.29309192539036394,
+      "grad_norm": 0.17672741413116455,
+      "learning_rate": 8.053639092279922e-05,
+      "loss": 1.1088,
+      "step": 4824
+    },
+    {
+      "epoch": 0.2931526824229905,
+      "grad_norm": 0.27308014035224915,
+      "learning_rate": 8.052881025287014e-05,
+      "loss": 1.0774,
+      "step": 4825
+    },
+    {
+      "epoch": 0.293213439455617,
+      "grad_norm": 0.12782545387744904,
+      "learning_rate": 8.052122846390159e-05,
+      "loss": 1.0579,
+      "step": 4826
+    },
+    {
+      "epoch": 0.2932741964882435,
+      "grad_norm": 1.653658390045166,
+      "learning_rate": 8.051364555617151e-05,
+      "loss": 1.1211,
+      "step": 4827
+    },
+    {
+      "epoch": 0.29333495352087,
+      "grad_norm": 0.17118819057941437,
+      "learning_rate": 8.050606152995788e-05,
+      "loss": 1.1686,
+      "step": 4828
+    },
+    {
+      "epoch": 0.29339571055349656,
+      "grad_norm": 0.12530478835105896,
+      "learning_rate": 8.049847638553864e-05,
+      "loss": 1.0382,
+      "step": 4829
+    },
+    {
+      "epoch": 0.2934564675861231,
+      "grad_norm": 0.30110740661621094,
+      "learning_rate": 8.049089012319188e-05,
+      "loss": 1.1891,
+      "step": 4830
+    },
+    {
+      "epoch": 0.29351722461874963,
+      "grad_norm": 0.21907773613929749,
+      "learning_rate": 8.048330274319562e-05,
+      "loss": 1.1105,
+      "step": 4831
+    },
+    {
+      "epoch": 0.29357798165137616,
+      "grad_norm": 0.1301761120557785,
+      "learning_rate": 8.047571424582803e-05,
+      "loss": 1.0864,
+      "step": 4832
+    },
+    {
+      "epoch": 0.2936387386840027,
+      "grad_norm": 0.2238079011440277,
+      "learning_rate": 8.046812463136724e-05,
+      "loss": 1.051,
+      "step": 4833
+    },
+    {
+      "epoch": 0.2936994957166292,
+      "grad_norm": 0.34450873732566833,
+      "learning_rate": 8.046053390009144e-05,
+      "loss": 1.1267,
+      "step": 4834
+    },
+    {
+      "epoch": 0.2937602527492557,
+      "grad_norm": 0.14757469296455383,
+      "learning_rate": 8.045294205227889e-05,
+      "loss": 1.0877,
+      "step": 4835
+    },
+    {
+      "epoch": 0.29382100978188225,
+      "grad_norm": 0.13920970261096954,
+      "learning_rate": 8.044534908820786e-05,
+      "loss": 1.0596,
+      "step": 4836
+    },
+    {
+      "epoch": 0.2938817668145088,
+      "grad_norm": 0.28844472765922546,
+      "learning_rate": 8.043775500815668e-05,
+      "loss": 1.2527,
+      "step": 4837
+    },
+    {
+      "epoch": 0.2939425238471353,
+      "grad_norm": 0.18626797199249268,
+      "learning_rate": 8.04301598124037e-05,
+      "loss": 1.1656,
+      "step": 4838
+    },
+    {
+      "epoch": 0.29400328087976185,
+      "grad_norm": 0.21400314569473267,
+      "learning_rate": 8.042256350122734e-05,
+      "loss": 1.1937,
+      "step": 4839
+    },
+    {
+      "epoch": 0.29406403791238833,
+      "grad_norm": 0.22018057107925415,
+      "learning_rate": 8.041496607490602e-05,
+      "loss": 1.1395,
+      "step": 4840
+    },
+    {
+      "epoch": 0.29412479494501487,
+      "grad_norm": 0.17825329303741455,
+      "learning_rate": 8.040736753371825e-05,
+      "loss": 1.1149,
+      "step": 4841
+    },
+    {
+      "epoch": 0.2941855519776414,
+      "grad_norm": 0.1618984341621399,
+      "learning_rate": 8.039976787794255e-05,
+      "loss": 1.0572,
+      "step": 4842
+    },
+    {
+      "epoch": 0.29424630901026794,
+      "grad_norm": 0.15700997412204742,
+      "learning_rate": 8.039216710785746e-05,
+      "loss": 1.0678,
+      "step": 4843
+    },
+    {
+      "epoch": 0.29430706604289447,
+      "grad_norm": 0.3226756751537323,
+      "learning_rate": 8.038456522374163e-05,
+      "loss": 1.3166,
+      "step": 4844
+    },
+    {
+      "epoch": 0.294367823075521,
+      "grad_norm": 0.11979656666517258,
+      "learning_rate": 8.037696222587368e-05,
+      "loss": 1.0468,
+      "step": 4845
+    },
+    {
+      "epoch": 0.29442858010814754,
+      "grad_norm": 0.12319060415029526,
+      "learning_rate": 8.036935811453231e-05,
+      "loss": 1.0336,
+      "step": 4846
+    },
+    {
+      "epoch": 0.294489337140774,
+      "grad_norm": 0.17491278052330017,
+      "learning_rate": 8.036175288999625e-05,
+      "loss": 1.1986,
+      "step": 4847
+    },
+    {
+      "epoch": 0.29455009417340056,
+      "grad_norm": 0.25195613503456116,
+      "learning_rate": 8.035414655254427e-05,
+      "loss": 1.231,
+      "step": 4848
+    },
+    {
+      "epoch": 0.2946108512060271,
+      "grad_norm": 0.1673683524131775,
+      "learning_rate": 8.034653910245517e-05,
+      "loss": 1.0813,
+      "step": 4849
+    },
+    {
+      "epoch": 0.2946716082386536,
+      "grad_norm": 0.20356975495815277,
+      "learning_rate": 8.033893054000783e-05,
+      "loss": 1.1635,
+      "step": 4850
+    },
+    {
+      "epoch": 0.29473236527128016,
+      "grad_norm": 0.6214486956596375,
+      "learning_rate": 8.033132086548112e-05,
+      "loss": 1.1949,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2947931223039067,
+      "grad_norm": 0.20783737301826477,
+      "learning_rate": 8.032371007915397e-05,
+      "loss": 1.1779,
+      "step": 4852
+    },
+    {
+      "epoch": 0.29485387933653323,
+      "grad_norm": 0.15741893649101257,
+      "learning_rate": 8.031609818130538e-05,
+      "loss": 1.0474,
+      "step": 4853
+    },
+    {
+      "epoch": 0.2949146363691597,
+      "grad_norm": 0.2226976752281189,
+      "learning_rate": 8.030848517221434e-05,
+      "loss": 1.0723,
+      "step": 4854
+    },
+    {
+      "epoch": 0.29497539340178625,
+      "grad_norm": 0.46578341722488403,
+      "learning_rate": 8.030087105215993e-05,
+      "loss": 1.0787,
+      "step": 4855
+    },
+    {
+      "epoch": 0.2950361504344128,
+      "grad_norm": 0.16051632165908813,
+      "learning_rate": 8.029325582142121e-05,
+      "loss": 1.1115,
+      "step": 4856
+    },
+    {
+      "epoch": 0.2950969074670393,
+      "grad_norm": 0.28879329562187195,
+      "learning_rate": 8.028563948027737e-05,
+      "loss": 1.1211,
+      "step": 4857
+    },
+    {
+      "epoch": 0.29515766449966585,
+      "grad_norm": 0.24414896965026855,
+      "learning_rate": 8.027802202900756e-05,
+      "loss": 1.0423,
+      "step": 4858
+    },
+    {
+      "epoch": 0.2952184215322924,
+      "grad_norm": 0.15892767906188965,
+      "learning_rate": 8.027040346789099e-05,
+      "loss": 1.1272,
+      "step": 4859
+    },
+    {
+      "epoch": 0.29527917856491886,
+      "grad_norm": 0.1327177882194519,
+      "learning_rate": 8.026278379720693e-05,
+      "loss": 1.0385,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2953399355975454,
+      "grad_norm": 0.14075778424739838,
+      "learning_rate": 8.025516301723468e-05,
+      "loss": 1.0746,
+      "step": 4861
+    },
+    {
+      "epoch": 0.29540069263017193,
+      "grad_norm": 0.11402823776006699,
+      "learning_rate": 8.02475411282536e-05,
+      "loss": 1.0574,
+      "step": 4862
+    },
+    {
+      "epoch": 0.29546144966279847,
+      "grad_norm": 0.18731290102005005,
+      "learning_rate": 8.023991813054302e-05,
+      "loss": 1.0603,
+      "step": 4863
+    },
+    {
+      "epoch": 0.295522206695425,
+      "grad_norm": 0.20494194328784943,
+      "learning_rate": 8.023229402438241e-05,
+      "loss": 1.1572,
+      "step": 4864
+    },
+    {
+      "epoch": 0.29558296372805154,
+      "grad_norm": 0.23749224841594696,
+      "learning_rate": 8.022466881005121e-05,
+      "loss": 1.0468,
+      "step": 4865
+    },
+    {
+      "epoch": 0.2956437207606781,
+      "grad_norm": 0.17095661163330078,
+      "learning_rate": 8.021704248782895e-05,
+      "loss": 1.1394,
+      "step": 4866
+    },
+    {
+      "epoch": 0.29570447779330455,
+      "grad_norm": 0.5161401629447937,
+      "learning_rate": 8.020941505799515e-05,
+      "loss": 1.2939,
+      "step": 4867
+    },
+    {
+      "epoch": 0.2957652348259311,
+      "grad_norm": 0.1483251452445984,
+      "learning_rate": 8.02017865208294e-05,
+      "loss": 1.0967,
+      "step": 4868
+    },
+    {
+      "epoch": 0.2958259918585576,
+      "grad_norm": 0.31485968828201294,
+      "learning_rate": 8.019415687661134e-05,
+      "loss": 1.0865,
+      "step": 4869
+    },
+    {
+      "epoch": 0.29588674889118416,
+      "grad_norm": 0.5022217631340027,
+      "learning_rate": 8.01865261256206e-05,
+      "loss": 1.1151,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2959475059238107,
+      "grad_norm": 0.2569407522678375,
+      "learning_rate": 8.017889426813692e-05,
+      "loss": 1.1146,
+      "step": 4871
+    },
+    {
+      "epoch": 0.29600826295643723,
+      "grad_norm": 0.32475942373275757,
+      "learning_rate": 8.017126130444004e-05,
+      "loss": 1.1545,
+      "step": 4872
+    },
+    {
+      "epoch": 0.2960690199890637,
+      "grad_norm": 0.32882651686668396,
+      "learning_rate": 8.016362723480974e-05,
+      "loss": 1.0458,
+      "step": 4873
+    },
+    {
+      "epoch": 0.29612977702169024,
+      "grad_norm": 0.28669607639312744,
+      "learning_rate": 8.015599205952586e-05,
+      "loss": 1.108,
+      "step": 4874
+    },
+    {
+      "epoch": 0.2961905340543168,
+      "grad_norm": 0.26534077525138855,
+      "learning_rate": 8.014835577886826e-05,
+      "loss": 1.0918,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2962512910869433,
+      "grad_norm": 0.19056156277656555,
+      "learning_rate": 8.014071839311686e-05,
+      "loss": 1.1559,
+      "step": 4876
+    },
+    {
+      "epoch": 0.29631204811956985,
+      "grad_norm": 0.1659012883901596,
+      "learning_rate": 8.013307990255159e-05,
+      "loss": 1.0685,
+      "step": 4877
+    },
+    {
+      "epoch": 0.2963728051521964,
+      "grad_norm": 0.5224549174308777,
+      "learning_rate": 8.012544030745244e-05,
+      "loss": 1.1839,
+      "step": 4878
+    },
+    {
+      "epoch": 0.2964335621848229,
+      "grad_norm": 0.2468225061893463,
+      "learning_rate": 8.011779960809947e-05,
+      "loss": 1.1048,
+      "step": 4879
+    },
+    {
+      "epoch": 0.2964943192174494,
+      "grad_norm": 0.14499223232269287,
+      "learning_rate": 8.011015780477274e-05,
+      "loss": 1.092,
+      "step": 4880
+    },
+    {
+      "epoch": 0.29655507625007593,
+      "grad_norm": 0.1538555920124054,
+      "learning_rate": 8.010251489775234e-05,
+      "loss": 1.0888,
+      "step": 4881
+    },
+    {
+      "epoch": 0.29661583328270247,
+      "grad_norm": 0.22946538031101227,
+      "learning_rate": 8.009487088731845e-05,
+      "loss": 1.1833,
+      "step": 4882
+    },
+    {
+      "epoch": 0.296676590315329,
+      "grad_norm": 0.17426064610481262,
+      "learning_rate": 8.008722577375122e-05,
+      "loss": 1.1255,
+      "step": 4883
+    },
+    {
+      "epoch": 0.29673734734795554,
+      "grad_norm": 0.22603633999824524,
+      "learning_rate": 8.007957955733094e-05,
+      "loss": 1.1739,
+      "step": 4884
+    },
+    {
+      "epoch": 0.2967981043805821,
+      "grad_norm": 0.3607141077518463,
+      "learning_rate": 8.007193223833786e-05,
+      "loss": 1.1017,
+      "step": 4885
+    },
+    {
+      "epoch": 0.29685886141320855,
+      "grad_norm": 0.2515600621700287,
+      "learning_rate": 8.006428381705228e-05,
+      "loss": 1.1138,
+      "step": 4886
+    },
+    {
+      "epoch": 0.2969196184458351,
+      "grad_norm": 0.11034715920686722,
+      "learning_rate": 8.005663429375457e-05,
+      "loss": 1.0618,
+      "step": 4887
+    },
+    {
+      "epoch": 0.2969803754784616,
+      "grad_norm": 0.24948768317699432,
+      "learning_rate": 8.004898366872512e-05,
+      "loss": 1.0942,
+      "step": 4888
+    },
+    {
+      "epoch": 0.29704113251108816,
+      "grad_norm": 0.20381641387939453,
+      "learning_rate": 8.004133194224436e-05,
+      "loss": 1.0782,
+      "step": 4889
+    },
+    {
+      "epoch": 0.2971018895437147,
+      "grad_norm": 0.20852461457252502,
+      "learning_rate": 8.003367911459279e-05,
+      "loss": 1.135,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2971626465763412,
+      "grad_norm": 2.300133228302002,
+      "learning_rate": 8.002602518605089e-05,
+      "loss": 1.0878,
+      "step": 4891
+    },
+    {
+      "epoch": 0.29722340360896776,
+      "grad_norm": 0.22455871105194092,
+      "learning_rate": 8.001837015689924e-05,
+      "loss": 1.1789,
+      "step": 4892
+    },
+    {
+      "epoch": 0.29728416064159424,
+      "grad_norm": 0.15111899375915527,
+      "learning_rate": 8.001071402741842e-05,
+      "loss": 1.085,
+      "step": 4893
+    },
+    {
+      "epoch": 0.2973449176742208,
+      "grad_norm": 0.21998457610607147,
+      "learning_rate": 8.000305679788908e-05,
+      "loss": 1.0692,
+      "step": 4894
+    },
+    {
+      "epoch": 0.2974056747068473,
+      "grad_norm": 0.34050309658050537,
+      "learning_rate": 7.99953984685919e-05,
+      "loss": 1.1033,
+      "step": 4895
+    },
+    {
+      "epoch": 0.29746643173947385,
+      "grad_norm": 0.24851812422275543,
+      "learning_rate": 7.998773903980759e-05,
+      "loss": 1.2472,
+      "step": 4896
+    },
+    {
+      "epoch": 0.2975271887721004,
+      "grad_norm": 0.1403234750032425,
+      "learning_rate": 7.99800785118169e-05,
+      "loss": 1.0835,
+      "step": 4897
+    },
+    {
+      "epoch": 0.2975879458047269,
+      "grad_norm": 0.14398226141929626,
+      "learning_rate": 7.997241688490062e-05,
+      "loss": 1.0664,
+      "step": 4898
+    },
+    {
+      "epoch": 0.2976487028373534,
+      "grad_norm": 0.2252487987279892,
+      "learning_rate": 7.996475415933964e-05,
+      "loss": 1.1443,
+      "step": 4899
+    },
+    {
+      "epoch": 0.29770945986997993,
+      "grad_norm": 0.15963275730609894,
+      "learning_rate": 7.99570903354148e-05,
+      "loss": 1.0689,
+      "step": 4900
+    },
+    {
+      "epoch": 0.29777021690260647,
+      "grad_norm": 0.14332592487335205,
+      "learning_rate": 7.9949425413407e-05,
+      "loss": 1.1201,
+      "step": 4901
+    },
+    {
+      "epoch": 0.297830973935233,
+      "grad_norm": 0.25667598843574524,
+      "learning_rate": 7.994175939359725e-05,
+      "loss": 1.2122,
+      "step": 4902
+    },
+    {
+      "epoch": 0.29789173096785954,
+      "grad_norm": 0.24005277454853058,
+      "learning_rate": 7.993409227626651e-05,
+      "loss": 1.1038,
+      "step": 4903
+    },
+    {
+      "epoch": 0.29795248800048607,
+      "grad_norm": 0.45693233609199524,
+      "learning_rate": 7.992642406169584e-05,
+      "loss": 1.061,
+      "step": 4904
+    },
+    {
+      "epoch": 0.2980132450331126,
+      "grad_norm": 0.2885357737541199,
+      "learning_rate": 7.991875475016629e-05,
+      "loss": 1.0943,
+      "step": 4905
+    },
+    {
+      "epoch": 0.2980740020657391,
+      "grad_norm": 0.15787114202976227,
+      "learning_rate": 7.991108434195902e-05,
+      "loss": 1.0836,
+      "step": 4906
+    },
+    {
+      "epoch": 0.2981347590983656,
+      "grad_norm": 0.25007563829421997,
+      "learning_rate": 7.990341283735516e-05,
+      "loss": 1.1384,
+      "step": 4907
+    },
+    {
+      "epoch": 0.29819551613099216,
+      "grad_norm": 0.15230529010295868,
+      "learning_rate": 7.989574023663592e-05,
+      "loss": 1.0359,
+      "step": 4908
+    },
+    {
+      "epoch": 0.2982562731636187,
+      "grad_norm": 0.49063366651535034,
+      "learning_rate": 7.988806654008256e-05,
+      "loss": 1.0633,
+      "step": 4909
+    },
+    {
+      "epoch": 0.2983170301962452,
+      "grad_norm": 0.24013467133045197,
+      "learning_rate": 7.988039174797633e-05,
+      "loss": 1.1554,
+      "step": 4910
+    },
+    {
+      "epoch": 0.29837778722887176,
+      "grad_norm": 0.42186757922172546,
+      "learning_rate": 7.987271586059857e-05,
+      "loss": 1.1717,
+      "step": 4911
+    },
+    {
+      "epoch": 0.2984385442614983,
+      "grad_norm": 0.29849252104759216,
+      "learning_rate": 7.986503887823064e-05,
+      "loss": 1.0789,
+      "step": 4912
+    },
+    {
+      "epoch": 0.2984993012941248,
+      "grad_norm": 0.19035378098487854,
+      "learning_rate": 7.985736080115393e-05,
+      "loss": 1.1628,
+      "step": 4913
+    },
+    {
+      "epoch": 0.2985600583267513,
+      "grad_norm": 0.13357847929000854,
+      "learning_rate": 7.98496816296499e-05,
+      "loss": 1.0374,
+      "step": 4914
+    },
+    {
+      "epoch": 0.29862081535937784,
+      "grad_norm": 0.25546133518218994,
+      "learning_rate": 7.984200136399999e-05,
+      "loss": 1.0679,
+      "step": 4915
+    },
+    {
+      "epoch": 0.2986815723920044,
+      "grad_norm": 0.21756219863891602,
+      "learning_rate": 7.983432000448578e-05,
+      "loss": 1.0823,
+      "step": 4916
+    },
+    {
+      "epoch": 0.2987423294246309,
+      "grad_norm": 0.17571885883808136,
+      "learning_rate": 7.982663755138879e-05,
+      "loss": 1.0591,
+      "step": 4917
+    },
+    {
+      "epoch": 0.29880308645725745,
+      "grad_norm": 0.2558269500732422,
+      "learning_rate": 7.981895400499065e-05,
+      "loss": 1.0669,
+      "step": 4918
+    },
+    {
+      "epoch": 0.29886384348988393,
+      "grad_norm": 0.4805607199668884,
+      "learning_rate": 7.981126936557298e-05,
+      "loss": 1.2421,
+      "step": 4919
+    },
+    {
+      "epoch": 0.29892460052251046,
+      "grad_norm": 0.7044974565505981,
+      "learning_rate": 7.980358363341747e-05,
+      "loss": 1.1244,
+      "step": 4920
+    },
+    {
+      "epoch": 0.298985357555137,
+      "grad_norm": 0.3205346465110779,
+      "learning_rate": 7.979589680880584e-05,
+      "loss": 1.1731,
+      "step": 4921
+    },
+    {
+      "epoch": 0.29904611458776353,
+      "grad_norm": 0.18769443035125732,
+      "learning_rate": 7.978820889201985e-05,
+      "loss": 1.0607,
+      "step": 4922
+    },
+    {
+      "epoch": 0.29910687162039007,
+      "grad_norm": 0.18973058462142944,
+      "learning_rate": 7.97805198833413e-05,
+      "loss": 1.1534,
+      "step": 4923
+    },
+    {
+      "epoch": 0.2991676286530166,
+      "grad_norm": 0.26898670196533203,
+      "learning_rate": 7.977282978305205e-05,
+      "loss": 1.1502,
+      "step": 4924
+    },
+    {
+      "epoch": 0.29922838568564314,
+      "grad_norm": 0.13413062691688538,
+      "learning_rate": 7.976513859143395e-05,
+      "loss": 1.0372,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2992891427182696,
+      "grad_norm": 0.22134798765182495,
+      "learning_rate": 7.975744630876896e-05,
+      "loss": 1.0807,
+      "step": 4926
+    },
+    {
+      "epoch": 0.29934989975089615,
+      "grad_norm": 0.760003387928009,
+      "learning_rate": 7.974975293533902e-05,
+      "loss": 1.098,
+      "step": 4927
+    },
+    {
+      "epoch": 0.2994106567835227,
+      "grad_norm": 0.1347123384475708,
+      "learning_rate": 7.974205847142613e-05,
+      "loss": 1.1281,
+      "step": 4928
+    },
+    {
+      "epoch": 0.2994714138161492,
+      "grad_norm": 0.16674065589904785,
+      "learning_rate": 7.973436291731234e-05,
+      "loss": 1.0827,
+      "step": 4929
+    },
+    {
+      "epoch": 0.29953217084877576,
+      "grad_norm": 0.1286884993314743,
+      "learning_rate": 7.972666627327974e-05,
+      "loss": 1.052,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2995929278814023,
+      "grad_norm": 0.172612726688385,
+      "learning_rate": 7.971896853961042e-05,
+      "loss": 1.1248,
+      "step": 4931
+    },
+    {
+      "epoch": 0.2996536849140288,
+      "grad_norm": 0.5562714338302612,
+      "learning_rate": 7.971126971658659e-05,
+      "loss": 1.3952,
+      "step": 4932
+    },
+    {
+      "epoch": 0.2997144419466553,
+      "grad_norm": 0.25090518593788147,
+      "learning_rate": 7.970356980449041e-05,
+      "loss": 1.055,
+      "step": 4933
+    },
+    {
+      "epoch": 0.29977519897928184,
+      "grad_norm": 0.1664423793554306,
+      "learning_rate": 7.969586880360413e-05,
+      "loss": 1.0859,
+      "step": 4934
+    },
+    {
+      "epoch": 0.2998359560119084,
+      "grad_norm": 0.3369622230529785,
+      "learning_rate": 7.968816671421005e-05,
+      "loss": 1.231,
+      "step": 4935
+    },
+    {
+      "epoch": 0.2998967130445349,
+      "grad_norm": 0.20732329785823822,
+      "learning_rate": 7.968046353659048e-05,
+      "loss": 1.2489,
+      "step": 4936
+    },
+    {
+      "epoch": 0.29995747007716145,
+      "grad_norm": 0.1416826993227005,
+      "learning_rate": 7.967275927102777e-05,
+      "loss": 1.0499,
+      "step": 4937
+    },
+    {
+      "epoch": 0.300018227109788,
+      "grad_norm": 0.1614678055047989,
+      "learning_rate": 7.966505391780436e-05,
+      "loss": 1.1482,
+      "step": 4938
+    },
+    {
+      "epoch": 0.30007898414241446,
+      "grad_norm": 0.20388440787792206,
+      "learning_rate": 7.965734747720264e-05,
+      "loss": 1.098,
+      "step": 4939
+    },
+    {
+      "epoch": 0.300139741175041,
+      "grad_norm": 0.15187425911426544,
+      "learning_rate": 7.964963994950513e-05,
+      "loss": 1.0801,
+      "step": 4940
+    },
+    {
+      "epoch": 0.30020049820766753,
+      "grad_norm": 0.14990165829658508,
+      "learning_rate": 7.964193133499432e-05,
+      "loss": 1.0409,
+      "step": 4941
+    },
+    {
+      "epoch": 0.30026125524029407,
+      "grad_norm": 0.24640433490276337,
+      "learning_rate": 7.963422163395281e-05,
+      "loss": 1.1354,
+      "step": 4942
+    },
+    {
+      "epoch": 0.3003220122729206,
+      "grad_norm": 0.16749872267246246,
+      "learning_rate": 7.962651084666316e-05,
+      "loss": 1.0518,
+      "step": 4943
+    },
+    {
+      "epoch": 0.30038276930554714,
+      "grad_norm": 0.11995523422956467,
+      "learning_rate": 7.961879897340803e-05,
+      "loss": 1.0149,
+      "step": 4944
+    },
+    {
+      "epoch": 0.3004435263381736,
+      "grad_norm": 0.22328971326351166,
+      "learning_rate": 7.96110860144701e-05,
+      "loss": 1.1297,
+      "step": 4945
+    },
+    {
+      "epoch": 0.30050428337080015,
+      "grad_norm": 0.2527675926685333,
+      "learning_rate": 7.96033719701321e-05,
+      "loss": 0.9945,
+      "step": 4946
+    },
+    {
+      "epoch": 0.3005650404034267,
+      "grad_norm": 0.43052467703819275,
+      "learning_rate": 7.959565684067675e-05,
+      "loss": 1.2694,
+      "step": 4947
+    },
+    {
+      "epoch": 0.3006257974360532,
+      "grad_norm": 0.28732234239578247,
+      "learning_rate": 7.95879406263869e-05,
+      "loss": 1.0808,
+      "step": 4948
+    },
+    {
+      "epoch": 0.30068655446867976,
+      "grad_norm": 3.940929889678955,
+      "learning_rate": 7.958022332754537e-05,
+      "loss": 1.0754,
+      "step": 4949
+    },
+    {
+      "epoch": 0.3007473115013063,
+      "grad_norm": 0.33323633670806885,
+      "learning_rate": 7.957250494443504e-05,
+      "loss": 1.1545,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3008080685339328,
+      "grad_norm": 0.15152540802955627,
+      "learning_rate": 7.956478547733881e-05,
+      "loss": 1.0185,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3008688255665593,
+      "grad_norm": 0.1421220600605011,
+      "learning_rate": 7.955706492653966e-05,
+      "loss": 1.1018,
+      "step": 4952
+    },
+    {
+      "epoch": 0.30092958259918584,
+      "grad_norm": 0.39070144295692444,
+      "learning_rate": 7.954934329232058e-05,
+      "loss": 1.0973,
+      "step": 4953
+    },
+    {
+      "epoch": 0.3009903396318124,
+      "grad_norm": 0.5201708078384399,
+      "learning_rate": 7.954162057496462e-05,
+      "loss": 1.1615,
+      "step": 4954
+    },
+    {
+      "epoch": 0.3010510966644389,
+      "grad_norm": 0.1853858232498169,
+      "learning_rate": 7.953389677475482e-05,
+      "loss": 1.1546,
+      "step": 4955
+    },
+    {
+      "epoch": 0.30111185369706545,
+      "grad_norm": 0.13055375218391418,
+      "learning_rate": 7.952617189197435e-05,
+      "loss": 1.0911,
+      "step": 4956
+    },
+    {
+      "epoch": 0.301172610729692,
+      "grad_norm": 0.16457238793373108,
+      "learning_rate": 7.951844592690634e-05,
+      "loss": 1.0622,
+      "step": 4957
+    },
+    {
+      "epoch": 0.3012333677623185,
+      "grad_norm": 0.19498306512832642,
+      "learning_rate": 7.951071887983397e-05,
+      "loss": 1.1219,
+      "step": 4958
+    },
+    {
+      "epoch": 0.301294124794945,
+      "grad_norm": 0.1648094356060028,
+      "learning_rate": 7.950299075104052e-05,
+      "loss": 1.075,
+      "step": 4959
+    },
+    {
+      "epoch": 0.30135488182757153,
+      "grad_norm": 0.16813687980175018,
+      "learning_rate": 7.949526154080922e-05,
+      "loss": 1.1705,
+      "step": 4960
+    },
+    {
+      "epoch": 0.30141563886019807,
+      "grad_norm": 0.15543539822101593,
+      "learning_rate": 7.948753124942343e-05,
+      "loss": 1.0736,
+      "step": 4961
+    },
+    {
+      "epoch": 0.3014763958928246,
+      "grad_norm": 0.2634231448173523,
+      "learning_rate": 7.947979987716647e-05,
+      "loss": 1.1821,
+      "step": 4962
+    },
+    {
+      "epoch": 0.30153715292545114,
+      "grad_norm": 0.1685730218887329,
+      "learning_rate": 7.947206742432174e-05,
+      "loss": 1.163,
+      "step": 4963
+    },
+    {
+      "epoch": 0.30159790995807767,
+      "grad_norm": 0.21360357105731964,
+      "learning_rate": 7.946433389117268e-05,
+      "loss": 1.0712,
+      "step": 4964
+    },
+    {
+      "epoch": 0.30165866699070415,
+      "grad_norm": 0.17472833395004272,
+      "learning_rate": 7.945659927800277e-05,
+      "loss": 1.0628,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3017194240233307,
+      "grad_norm": 0.23866689205169678,
+      "learning_rate": 7.944886358509554e-05,
+      "loss": 1.0556,
+      "step": 4966
+    },
+    {
+      "epoch": 0.3017801810559572,
+      "grad_norm": 0.20201341807842255,
+      "learning_rate": 7.94411268127345e-05,
+      "loss": 1.036,
+      "step": 4967
+    },
+    {
+      "epoch": 0.30184093808858375,
+      "grad_norm": 0.21411119401454926,
+      "learning_rate": 7.943338896120328e-05,
+      "loss": 1.2082,
+      "step": 4968
+    },
+    {
+      "epoch": 0.3019016951212103,
+      "grad_norm": 0.19133217632770538,
+      "learning_rate": 7.942565003078548e-05,
+      "loss": 1.1021,
+      "step": 4969
+    },
+    {
+      "epoch": 0.3019624521538368,
+      "grad_norm": 0.5134078860282898,
+      "learning_rate": 7.94179100217648e-05,
+      "loss": 1.1174,
+      "step": 4970
+    },
+    {
+      "epoch": 0.30202320918646336,
+      "grad_norm": 0.26805537939071655,
+      "learning_rate": 7.941016893442495e-05,
+      "loss": 1.1782,
+      "step": 4971
+    },
+    {
+      "epoch": 0.30208396621908984,
+      "grad_norm": 0.23270122706890106,
+      "learning_rate": 7.940242676904965e-05,
+      "loss": 1.3901,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3021447232517164,
+      "grad_norm": 0.3333161771297455,
+      "learning_rate": 7.939468352592274e-05,
+      "loss": 1.1211,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3022054802843429,
+      "grad_norm": 0.17558227479457855,
+      "learning_rate": 7.938693920532801e-05,
+      "loss": 1.078,
+      "step": 4974
+    },
+    {
+      "epoch": 0.30226623731696944,
+      "grad_norm": 0.499388188123703,
+      "learning_rate": 7.937919380754934e-05,
+      "loss": 1.0234,
+      "step": 4975
+    },
+    {
+      "epoch": 0.302326994349596,
+      "grad_norm": 7.581668376922607,
+      "learning_rate": 7.937144733287065e-05,
+      "loss": 1.0464,
+      "step": 4976
+    },
+    {
+      "epoch": 0.3023877513822225,
+      "grad_norm": 0.23260857164859772,
+      "learning_rate": 7.936369978157589e-05,
+      "loss": 1.1839,
+      "step": 4977
+    },
+    {
+      "epoch": 0.302448508414849,
+      "grad_norm": 0.15449270606040955,
+      "learning_rate": 7.935595115394902e-05,
+      "loss": 1.0658,
+      "step": 4978
+    },
+    {
+      "epoch": 0.30250926544747553,
+      "grad_norm": 0.18905365467071533,
+      "learning_rate": 7.93482014502741e-05,
+      "loss": 1.05,
+      "step": 4979
+    },
+    {
+      "epoch": 0.30257002248010206,
+      "grad_norm": 0.13975360989570618,
+      "learning_rate": 7.934045067083517e-05,
+      "loss": 1.0636,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3026307795127286,
+      "grad_norm": 0.13250093162059784,
+      "learning_rate": 7.933269881591635e-05,
+      "loss": 1.0972,
+      "step": 4981
+    },
+    {
+      "epoch": 0.30269153654535513,
+      "grad_norm": 0.15861931443214417,
+      "learning_rate": 7.932494588580179e-05,
+      "loss": 1.0982,
+      "step": 4982
+    },
+    {
+      "epoch": 0.30275229357798167,
+      "grad_norm": 0.24535703659057617,
+      "learning_rate": 7.931719188077567e-05,
+      "loss": 1.1169,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3028130506106082,
+      "grad_norm": 0.19138747453689575,
+      "learning_rate": 7.930943680112222e-05,
+      "loss": 1.1768,
+      "step": 4984
+    },
+    {
+      "epoch": 0.3028738076432347,
+      "grad_norm": 0.39421167969703674,
+      "learning_rate": 7.930168064712569e-05,
+      "loss": 1.0769,
+      "step": 4985
+    },
+    {
+      "epoch": 0.3029345646758612,
+      "grad_norm": 0.23053671419620514,
+      "learning_rate": 7.92939234190704e-05,
+      "loss": 1.1584,
+      "step": 4986
+    },
+    {
+      "epoch": 0.30299532170848775,
+      "grad_norm": 0.6366480588912964,
+      "learning_rate": 7.928616511724068e-05,
+      "loss": 1.1671,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3030560787411143,
+      "grad_norm": 0.3631538152694702,
+      "learning_rate": 7.927840574192092e-05,
+      "loss": 1.1417,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3031168357737408,
+      "grad_norm": 0.27150318026542664,
+      "learning_rate": 7.927064529339555e-05,
+      "loss": 1.136,
+      "step": 4989
+    },
+    {
+      "epoch": 0.30317759280636736,
+      "grad_norm": 0.4584693908691406,
+      "learning_rate": 7.926288377194899e-05,
+      "loss": 1.2183,
+      "step": 4990
+    },
+    {
+      "epoch": 0.30323834983899384,
+      "grad_norm": 0.20481142401695251,
+      "learning_rate": 7.925512117786578e-05,
+      "loss": 1.1193,
+      "step": 4991
+    },
+    {
+      "epoch": 0.30329910687162037,
+      "grad_norm": 0.35838496685028076,
+      "learning_rate": 7.924735751143045e-05,
+      "loss": 1.2194,
+      "step": 4992
+    },
+    {
+      "epoch": 0.3033598639042469,
+      "grad_norm": 0.36893385648727417,
+      "learning_rate": 7.923959277292759e-05,
+      "loss": 1.186,
+      "step": 4993
+    },
+    {
+      "epoch": 0.30342062093687344,
+      "grad_norm": 0.15746700763702393,
+      "learning_rate": 7.923182696264177e-05,
+      "loss": 1.0833,
+      "step": 4994
+    },
+    {
+      "epoch": 0.3034813779695,
+      "grad_norm": 0.3159339725971222,
+      "learning_rate": 7.922406008085771e-05,
+      "loss": 1.2208,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3035421350021265,
+      "grad_norm": 0.2524588704109192,
+      "learning_rate": 7.921629212786008e-05,
+      "loss": 1.1383,
+      "step": 4996
+    },
+    {
+      "epoch": 0.30360289203475305,
+      "grad_norm": 0.4001350402832031,
+      "learning_rate": 7.92085231039336e-05,
+      "loss": 1.1228,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3036636490673795,
+      "grad_norm": 0.39483341574668884,
+      "learning_rate": 7.920075300936308e-05,
+      "loss": 1.1851,
+      "step": 4998
+    },
+    {
+      "epoch": 0.30372440610000606,
+      "grad_norm": 0.5460163950920105,
+      "learning_rate": 7.919298184443331e-05,
+      "loss": 1.0453,
+      "step": 4999
+    },
+    {
+      "epoch": 0.3037851631326326,
+      "grad_norm": 0.2757039964199066,
+      "learning_rate": 7.918520960942913e-05,
+      "loss": 1.2287,
+      "step": 5000
+    },
+    {
+      "epoch": 0.30384592016525913,
+      "grad_norm": 0.21803759038448334,
+      "learning_rate": 7.917743630463546e-05,
+      "loss": 1.129,
+      "step": 5001
+    },
+    {
+      "epoch": 0.30390667719788567,
+      "grad_norm": 0.21053986251354218,
+      "learning_rate": 7.916966193033723e-05,
+      "loss": 1.1218,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3039674342305122,
+      "grad_norm": 0.23183192312717438,
+      "learning_rate": 7.916188648681939e-05,
+      "loss": 1.1931,
+      "step": 5003
+    },
+    {
+      "epoch": 0.3040281912631387,
+      "grad_norm": 0.14881351590156555,
+      "learning_rate": 7.915410997436697e-05,
+      "loss": 1.0757,
+      "step": 5004
+    },
+    {
+      "epoch": 0.3040889482957652,
+      "grad_norm": 0.17829740047454834,
+      "learning_rate": 7.914633239326502e-05,
+      "loss": 1.0884,
+      "step": 5005
+    },
+    {
+      "epoch": 0.30414970532839175,
+      "grad_norm": 0.938035249710083,
+      "learning_rate": 7.913855374379862e-05,
+      "loss": 1.2527,
+      "step": 5006
+    },
+    {
+      "epoch": 0.3042104623610183,
+      "grad_norm": 0.1620355099439621,
+      "learning_rate": 7.913077402625289e-05,
+      "loss": 1.0918,
+      "step": 5007
+    },
+    {
+      "epoch": 0.3042712193936448,
+      "grad_norm": 0.1912277787923813,
+      "learning_rate": 7.912299324091301e-05,
+      "loss": 1.1789,
+      "step": 5008
+    },
+    {
+      "epoch": 0.30433197642627136,
+      "grad_norm": 0.15307871997356415,
+      "learning_rate": 7.911521138806419e-05,
+      "loss": 1.0679,
+      "step": 5009
+    },
+    {
+      "epoch": 0.3043927334588979,
+      "grad_norm": 0.22374896705150604,
+      "learning_rate": 7.910742846799169e-05,
+      "loss": 1.0754,
+      "step": 5010
+    },
+    {
+      "epoch": 0.30445349049152437,
+      "grad_norm": 0.16303293406963348,
+      "learning_rate": 7.909964448098073e-05,
+      "loss": 1.0597,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3045142475241509,
+      "grad_norm": 0.39754754304885864,
+      "learning_rate": 7.90918594273167e-05,
+      "loss": 1.3041,
+      "step": 5012
+    },
+    {
+      "epoch": 0.30457500455677744,
+      "grad_norm": 0.15414276719093323,
+      "learning_rate": 7.908407330728493e-05,
+      "loss": 1.0549,
+      "step": 5013
+    },
+    {
+      "epoch": 0.304635761589404,
+      "grad_norm": 1.6768600940704346,
+      "learning_rate": 7.907628612117084e-05,
+      "loss": 1.218,
+      "step": 5014
+    },
+    {
+      "epoch": 0.3046965186220305,
+      "grad_norm": 0.20556695759296417,
+      "learning_rate": 7.906849786925987e-05,
+      "loss": 1.1323,
+      "step": 5015
+    },
+    {
+      "epoch": 0.30475727565465704,
+      "grad_norm": 0.8451787829399109,
+      "learning_rate": 7.906070855183747e-05,
+      "loss": 1.1674,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3048180326872836,
+      "grad_norm": 0.5405980944633484,
+      "learning_rate": 7.90529181691892e-05,
+      "loss": 1.2248,
+      "step": 5017
+    },
+    {
+      "epoch": 0.30487878971991006,
+      "grad_norm": 0.1711481809616089,
+      "learning_rate": 7.904512672160058e-05,
+      "loss": 1.1353,
+      "step": 5018
+    },
+    {
+      "epoch": 0.3049395467525366,
+      "grad_norm": 0.19523300230503082,
+      "learning_rate": 7.903733420935723e-05,
+      "loss": 1.0708,
+      "step": 5019
+    },
+    {
+      "epoch": 0.30500030378516313,
+      "grad_norm": 1.2843191623687744,
+      "learning_rate": 7.90295406327448e-05,
+      "loss": 1.0794,
+      "step": 5020
+    },
+    {
+      "epoch": 0.30506106081778966,
+      "grad_norm": 0.22761689126491547,
+      "learning_rate": 7.902174599204892e-05,
+      "loss": 1.1177,
+      "step": 5021
+    },
+    {
+      "epoch": 0.3051218178504162,
+      "grad_norm": 1.1813548803329468,
+      "learning_rate": 7.901395028755536e-05,
+      "loss": 1.3213,
+      "step": 5022
+    },
+    {
+      "epoch": 0.30518257488304273,
+      "grad_norm": 0.4325633943080902,
+      "learning_rate": 7.900615351954982e-05,
+      "loss": 1.1314,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3052433319156692,
+      "grad_norm": 0.6260473728179932,
+      "learning_rate": 7.899835568831813e-05,
+      "loss": 1.0374,
+      "step": 5024
+    },
+    {
+      "epoch": 0.30530408894829575,
+      "grad_norm": 0.18695132434368134,
+      "learning_rate": 7.89905567941461e-05,
+      "loss": 1.113,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3053648459809223,
+      "grad_norm": 0.3728717863559723,
+      "learning_rate": 7.898275683731962e-05,
+      "loss": 1.1198,
+      "step": 5026
+    },
+    {
+      "epoch": 0.3054256030135488,
+      "grad_norm": 0.44424402713775635,
+      "learning_rate": 7.897495581812457e-05,
+      "loss": 1.0337,
+      "step": 5027
+    },
+    {
+      "epoch": 0.30548636004617535,
+      "grad_norm": 0.1744932383298874,
+      "learning_rate": 7.89671537368469e-05,
+      "loss": 1.1343,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3055471170788019,
+      "grad_norm": 0.7512226104736328,
+      "learning_rate": 7.895935059377264e-05,
+      "loss": 1.2605,
+      "step": 5029
+    },
+    {
+      "epoch": 0.3056078741114284,
+      "grad_norm": 0.309328556060791,
+      "learning_rate": 7.895154638918778e-05,
+      "loss": 1.1397,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3056686311440549,
+      "grad_norm": 0.20712530612945557,
+      "learning_rate": 7.894374112337839e-05,
+      "loss": 1.1313,
+      "step": 5031
+    },
+    {
+      "epoch": 0.30572938817668144,
+      "grad_norm": 0.36675825715065,
+      "learning_rate": 7.893593479663057e-05,
+      "loss": 1.0413,
+      "step": 5032
+    },
+    {
+      "epoch": 0.305790145209308,
+      "grad_norm": 0.5124492645263672,
+      "learning_rate": 7.892812740923045e-05,
+      "loss": 1.0542,
+      "step": 5033
+    },
+    {
+      "epoch": 0.3058509022419345,
+      "grad_norm": 0.2213868945837021,
+      "learning_rate": 7.892031896146425e-05,
+      "loss": 1.0665,
+      "step": 5034
+    },
+    {
+      "epoch": 0.30591165927456104,
+      "grad_norm": 0.36965662240982056,
+      "learning_rate": 7.891250945361817e-05,
+      "loss": 1.0811,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3059724163071876,
+      "grad_norm": 0.29335683584213257,
+      "learning_rate": 7.890469888597847e-05,
+      "loss": 1.037,
+      "step": 5036
+    },
+    {
+      "epoch": 0.30603317333981406,
+      "grad_norm": 0.21715201437473297,
+      "learning_rate": 7.889688725883143e-05,
+      "loss": 1.0774,
+      "step": 5037
+    },
+    {
+      "epoch": 0.3060939303724406,
+      "grad_norm": 0.16601771116256714,
+      "learning_rate": 7.88890745724634e-05,
+      "loss": 1.0936,
+      "step": 5038
+    },
+    {
+      "epoch": 0.3061546874050671,
+      "grad_norm": 0.1503555029630661,
+      "learning_rate": 7.888126082716077e-05,
+      "loss": 1.0898,
+      "step": 5039
+    },
+    {
+      "epoch": 0.30621544443769366,
+      "grad_norm": 0.13763649761676788,
+      "learning_rate": 7.887344602320994e-05,
+      "loss": 1.089,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3062762014703202,
+      "grad_norm": 0.2543410658836365,
+      "learning_rate": 7.886563016089737e-05,
+      "loss": 1.1757,
+      "step": 5041
+    },
+    {
+      "epoch": 0.30633695850294673,
+      "grad_norm": 0.37459924817085266,
+      "learning_rate": 7.885781324050953e-05,
+      "loss": 1.1547,
+      "step": 5042
+    },
+    {
+      "epoch": 0.30639771553557327,
+      "grad_norm": 0.16601353883743286,
+      "learning_rate": 7.884999526233297e-05,
+      "loss": 1.0775,
+      "step": 5043
+    },
+    {
+      "epoch": 0.30645847256819975,
+      "grad_norm": 0.18259654939174652,
+      "learning_rate": 7.884217622665426e-05,
+      "loss": 1.0922,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3065192296008263,
+      "grad_norm": 0.26253360509872437,
+      "learning_rate": 7.883435613376003e-05,
+      "loss": 1.1165,
+      "step": 5045
+    },
+    {
+      "epoch": 0.3065799866334528,
+      "grad_norm": 0.7012448906898499,
+      "learning_rate": 7.882653498393688e-05,
+      "loss": 1.1857,
+      "step": 5046
+    },
+    {
+      "epoch": 0.30664074366607935,
+      "grad_norm": 0.306253045797348,
+      "learning_rate": 7.88187127774715e-05,
+      "loss": 1.1296,
+      "step": 5047
+    },
+    {
+      "epoch": 0.3067015006987059,
+      "grad_norm": 0.19283844530582428,
+      "learning_rate": 7.881088951465066e-05,
+      "loss": 1.0603,
+      "step": 5048
+    },
+    {
+      "epoch": 0.3067622577313324,
+      "grad_norm": 0.1580544412136078,
+      "learning_rate": 7.880306519576109e-05,
+      "loss": 1.1252,
+      "step": 5049
+    },
+    {
+      "epoch": 0.3068230147639589,
+      "grad_norm": 0.16325531899929047,
+      "learning_rate": 7.87952398210896e-05,
+      "loss": 1.0204,
+      "step": 5050
+    },
+    {
+      "epoch": 0.30688377179658544,
+      "grad_norm": 0.17274509370326996,
+      "learning_rate": 7.878741339092302e-05,
+      "loss": 1.0666,
+      "step": 5051
+    },
+    {
+      "epoch": 0.30694452882921197,
+      "grad_norm": 0.2777055501937866,
+      "learning_rate": 7.877958590554824e-05,
+      "loss": 1.2514,
+      "step": 5052
+    },
+    {
+      "epoch": 0.3070052858618385,
+      "grad_norm": 0.15812616050243378,
+      "learning_rate": 7.877175736525217e-05,
+      "loss": 1.0664,
+      "step": 5053
+    },
+    {
+      "epoch": 0.30706604289446504,
+      "grad_norm": 0.6769807934761047,
+      "learning_rate": 7.87639277703218e-05,
+      "loss": 1.3595,
+      "step": 5054
+    },
+    {
+      "epoch": 0.3071267999270916,
+      "grad_norm": 0.19761823117733002,
+      "learning_rate": 7.875609712104407e-05,
+      "loss": 1.0974,
+      "step": 5055
+    },
+    {
+      "epoch": 0.3071875569597181,
+      "grad_norm": 0.1579507738351822,
+      "learning_rate": 7.874826541770604e-05,
+      "loss": 1.0592,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3072483139923446,
+      "grad_norm": 0.15911029279232025,
+      "learning_rate": 7.874043266059479e-05,
+      "loss": 1.1017,
+      "step": 5057
+    },
+    {
+      "epoch": 0.3073090710249711,
+      "grad_norm": 0.12055608630180359,
+      "learning_rate": 7.873259884999744e-05,
+      "loss": 1.0086,
+      "step": 5058
+    },
+    {
+      "epoch": 0.30736982805759766,
+      "grad_norm": 0.1480165421962738,
+      "learning_rate": 7.87247639862011e-05,
+      "loss": 1.021,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3074305850902242,
+      "grad_norm": 0.12465279549360275,
+      "learning_rate": 7.871692806949301e-05,
+      "loss": 1.057,
+      "step": 5060
+    },
+    {
+      "epoch": 0.30749134212285073,
+      "grad_norm": 0.1836155503988266,
+      "learning_rate": 7.870909110016036e-05,
+      "loss": 1.0774,
+      "step": 5061
+    },
+    {
+      "epoch": 0.30755209915547727,
+      "grad_norm": 0.13721024990081787,
+      "learning_rate": 7.870125307849042e-05,
+      "loss": 1.0415,
+      "step": 5062
+    },
+    {
+      "epoch": 0.3076128561881038,
+      "grad_norm": 0.1414865106344223,
+      "learning_rate": 7.86934140047705e-05,
+      "loss": 1.0651,
+      "step": 5063
+    },
+    {
+      "epoch": 0.3076736132207303,
+      "grad_norm": 0.2621670961380005,
+      "learning_rate": 7.868557387928796e-05,
+      "loss": 1.1895,
+      "step": 5064
+    },
+    {
+      "epoch": 0.3077343702533568,
+      "grad_norm": 0.6092076301574707,
+      "learning_rate": 7.867773270233014e-05,
+      "loss": 1.0737,
+      "step": 5065
+    },
+    {
+      "epoch": 0.30779512728598335,
+      "grad_norm": 0.3646189570426941,
+      "learning_rate": 7.866989047418452e-05,
+      "loss": 1.2532,
+      "step": 5066
+    },
+    {
+      "epoch": 0.3078558843186099,
+      "grad_norm": 0.18351449072360992,
+      "learning_rate": 7.86620471951385e-05,
+      "loss": 1.0589,
+      "step": 5067
+    },
+    {
+      "epoch": 0.3079166413512364,
+      "grad_norm": 0.23390978574752808,
+      "learning_rate": 7.865420286547959e-05,
+      "loss": 1.0713,
+      "step": 5068
+    },
+    {
+      "epoch": 0.30797739838386295,
+      "grad_norm": 0.15962259471416473,
+      "learning_rate": 7.864635748549534e-05,
+      "loss": 1.0818,
+      "step": 5069
+    },
+    {
+      "epoch": 0.30803815541648943,
+      "grad_norm": 0.17449022829532623,
+      "learning_rate": 7.863851105547333e-05,
+      "loss": 1.0783,
+      "step": 5070
+    },
+    {
+      "epoch": 0.30809891244911597,
+      "grad_norm": 0.13715773820877075,
+      "learning_rate": 7.863066357570115e-05,
+      "loss": 1.0409,
+      "step": 5071
+    },
+    {
+      "epoch": 0.3081596694817425,
+      "grad_norm": 0.20138972997665405,
+      "learning_rate": 7.862281504646647e-05,
+      "loss": 1.0236,
+      "step": 5072
+    },
+    {
+      "epoch": 0.30822042651436904,
+      "grad_norm": 0.8320518732070923,
+      "learning_rate": 7.861496546805697e-05,
+      "loss": 1.1709,
+      "step": 5073
+    },
+    {
+      "epoch": 0.3082811835469956,
+      "grad_norm": 0.2535605728626251,
+      "learning_rate": 7.860711484076038e-05,
+      "loss": 1.1913,
+      "step": 5074
+    },
+    {
+      "epoch": 0.3083419405796221,
+      "grad_norm": 0.2105361521244049,
+      "learning_rate": 7.859926316486448e-05,
+      "loss": 1.1223,
+      "step": 5075
+    },
+    {
+      "epoch": 0.30840269761224864,
+      "grad_norm": 0.26760080456733704,
+      "learning_rate": 7.859141044065704e-05,
+      "loss": 1.1247,
+      "step": 5076
+    },
+    {
+      "epoch": 0.3084634546448751,
+      "grad_norm": 0.14098025858402252,
+      "learning_rate": 7.858355666842594e-05,
+      "loss": 1.0657,
+      "step": 5077
+    },
+    {
+      "epoch": 0.30852421167750166,
+      "grad_norm": 0.25043994188308716,
+      "learning_rate": 7.857570184845903e-05,
+      "loss": 1.1209,
+      "step": 5078
+    },
+    {
+      "epoch": 0.3085849687101282,
+      "grad_norm": 0.23554249107837677,
+      "learning_rate": 7.856784598104426e-05,
+      "loss": 1.2225,
+      "step": 5079
+    },
+    {
+      "epoch": 0.30864572574275473,
+      "grad_norm": 0.19218409061431885,
+      "learning_rate": 7.855998906646956e-05,
+      "loss": 1.1355,
+      "step": 5080
+    },
+    {
+      "epoch": 0.30870648277538126,
+      "grad_norm": 0.20556126534938812,
+      "learning_rate": 7.855213110502293e-05,
+      "loss": 1.1302,
+      "step": 5081
+    },
+    {
+      "epoch": 0.3087672398080078,
+      "grad_norm": 0.16717664897441864,
+      "learning_rate": 7.854427209699244e-05,
+      "loss": 1.1274,
+      "step": 5082
+    },
+    {
+      "epoch": 0.3088279968406343,
+      "grad_norm": 0.1822149008512497,
+      "learning_rate": 7.853641204266614e-05,
+      "loss": 1.1169,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3088887538732608,
+      "grad_norm": 0.1569221466779709,
+      "learning_rate": 7.852855094233211e-05,
+      "loss": 1.1279,
+      "step": 5084
+    },
+    {
+      "epoch": 0.30894951090588735,
+      "grad_norm": 0.25345590710639954,
+      "learning_rate": 7.852068879627855e-05,
+      "loss": 1.1864,
+      "step": 5085
+    },
+    {
+      "epoch": 0.3090102679385139,
+      "grad_norm": 0.16206392645835876,
+      "learning_rate": 7.851282560479363e-05,
+      "loss": 1.0941,
+      "step": 5086
+    },
+    {
+      "epoch": 0.3090710249711404,
+      "grad_norm": 0.17163822054862976,
+      "learning_rate": 7.850496136816558e-05,
+      "loss": 1.0964,
+      "step": 5087
+    },
+    {
+      "epoch": 0.30913178200376695,
+      "grad_norm": 0.2909422814846039,
+      "learning_rate": 7.849709608668265e-05,
+      "loss": 1.2,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3091925390363935,
+      "grad_norm": 0.47417882084846497,
+      "learning_rate": 7.848922976063317e-05,
+      "loss": 1.0497,
+      "step": 5089
+    },
+    {
+      "epoch": 0.30925329606901997,
+      "grad_norm": 0.18791179358959198,
+      "learning_rate": 7.848136239030546e-05,
+      "loss": 1.1051,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3093140531016465,
+      "grad_norm": 0.20847441256046295,
+      "learning_rate": 7.84734939759879e-05,
+      "loss": 1.1509,
+      "step": 5091
+    },
+    {
+      "epoch": 0.30937481013427304,
+      "grad_norm": 0.25191542506217957,
+      "learning_rate": 7.846562451796892e-05,
+      "loss": 1.0694,
+      "step": 5092
+    },
+    {
+      "epoch": 0.30943556716689957,
+      "grad_norm": 0.32054251432418823,
+      "learning_rate": 7.845775401653696e-05,
+      "loss": 1.0594,
+      "step": 5093
+    },
+    {
+      "epoch": 0.3094963241995261,
+      "grad_norm": 0.20492412149906158,
+      "learning_rate": 7.844988247198055e-05,
+      "loss": 1.2071,
+      "step": 5094
+    },
+    {
+      "epoch": 0.30955708123215264,
+      "grad_norm": 0.20347364246845245,
+      "learning_rate": 7.844200988458818e-05,
+      "loss": 1.1415,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3096178382647791,
+      "grad_norm": 0.2791954576969147,
+      "learning_rate": 7.843413625464843e-05,
+      "loss": 1.0529,
+      "step": 5096
+    },
+    {
+      "epoch": 0.30967859529740566,
+      "grad_norm": 0.15032117068767548,
+      "learning_rate": 7.842626158244994e-05,
+      "loss": 1.076,
+      "step": 5097
+    },
+    {
+      "epoch": 0.3097393523300322,
+      "grad_norm": 0.5204444527626038,
+      "learning_rate": 7.841838586828134e-05,
+      "loss": 1.0288,
+      "step": 5098
+    },
+    {
+      "epoch": 0.3098001093626587,
+      "grad_norm": 0.26252487301826477,
+      "learning_rate": 7.841050911243129e-05,
+      "loss": 1.1046,
+      "step": 5099
+    },
+    {
+      "epoch": 0.30986086639528526,
+      "grad_norm": 0.21715553104877472,
+      "learning_rate": 7.840263131518856e-05,
+      "loss": 1.1699,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3099216234279118,
+      "grad_norm": 0.400338351726532,
+      "learning_rate": 7.839475247684187e-05,
+      "loss": 1.0411,
+      "step": 5101
+    },
+    {
+      "epoch": 0.30998238046053833,
+      "grad_norm": 0.16862495243549347,
+      "learning_rate": 7.838687259768004e-05,
+      "loss": 1.1446,
+      "step": 5102
+    },
+    {
+      "epoch": 0.3100431374931648,
+      "grad_norm": 0.188529372215271,
+      "learning_rate": 7.837899167799194e-05,
+      "loss": 1.0879,
+      "step": 5103
+    },
+    {
+      "epoch": 0.31010389452579135,
+      "grad_norm": 0.20136164128780365,
+      "learning_rate": 7.837110971806638e-05,
+      "loss": 1.1205,
+      "step": 5104
+    },
+    {
+      "epoch": 0.3101646515584179,
+      "grad_norm": 0.27363207936286926,
+      "learning_rate": 7.836322671819232e-05,
+      "loss": 1.111,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3102254085910444,
+      "grad_norm": 0.16804730892181396,
+      "learning_rate": 7.83553426786587e-05,
+      "loss": 1.1572,
+      "step": 5106
+    },
+    {
+      "epoch": 0.31028616562367095,
+      "grad_norm": 0.26315054297447205,
+      "learning_rate": 7.834745759975453e-05,
+      "loss": 1.0769,
+      "step": 5107
+    },
+    {
+      "epoch": 0.3103469226562975,
+      "grad_norm": 0.23740170896053314,
+      "learning_rate": 7.83395714817688e-05,
+      "loss": 1.1112,
+      "step": 5108
+    },
+    {
+      "epoch": 0.31040767968892397,
+      "grad_norm": 0.17278525233268738,
+      "learning_rate": 7.83316843249906e-05,
+      "loss": 1.1308,
+      "step": 5109
+    },
+    {
+      "epoch": 0.3104684367215505,
+      "grad_norm": 0.3111562728881836,
+      "learning_rate": 7.832379612970905e-05,
+      "loss": 1.1484,
+      "step": 5110
+    },
+    {
+      "epoch": 0.31052919375417704,
+      "grad_norm": 0.1663617044687271,
+      "learning_rate": 7.831590689621326e-05,
+      "loss": 1.0541,
+      "step": 5111
+    },
+    {
+      "epoch": 0.31058995078680357,
+      "grad_norm": 0.22036807239055634,
+      "learning_rate": 7.830801662479244e-05,
+      "loss": 1.2443,
+      "step": 5112
+    },
+    {
+      "epoch": 0.3106507078194301,
+      "grad_norm": 0.1357625424861908,
+      "learning_rate": 7.830012531573581e-05,
+      "loss": 1.0797,
+      "step": 5113
+    },
+    {
+      "epoch": 0.31071146485205664,
+      "grad_norm": 0.19941739737987518,
+      "learning_rate": 7.829223296933259e-05,
+      "loss": 1.1688,
+      "step": 5114
+    },
+    {
+      "epoch": 0.3107722218846832,
+      "grad_norm": 0.17946669459342957,
+      "learning_rate": 7.828433958587213e-05,
+      "loss": 1.1471,
+      "step": 5115
+    },
+    {
+      "epoch": 0.31083297891730965,
+      "grad_norm": 0.527148962020874,
+      "learning_rate": 7.827644516564371e-05,
+      "loss": 1.1591,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3108937359499362,
+      "grad_norm": 0.17529942095279694,
+      "learning_rate": 7.826854970893674e-05,
+      "loss": 1.2252,
+      "step": 5117
+    },
+    {
+      "epoch": 0.3109544929825627,
+      "grad_norm": 0.16480781137943268,
+      "learning_rate": 7.826065321604061e-05,
+      "loss": 1.0701,
+      "step": 5118
+    },
+    {
+      "epoch": 0.31101525001518926,
+      "grad_norm": 0.15994659066200256,
+      "learning_rate": 7.825275568724478e-05,
+      "loss": 1.1142,
+      "step": 5119
+    },
+    {
+      "epoch": 0.3110760070478158,
+      "grad_norm": 0.15284310281276703,
+      "learning_rate": 7.824485712283873e-05,
+      "loss": 1.1072,
+      "step": 5120
+    },
+    {
+      "epoch": 0.31113676408044233,
+      "grad_norm": 0.17039936780929565,
+      "learning_rate": 7.823695752311198e-05,
+      "loss": 1.0696,
+      "step": 5121
+    },
+    {
+      "epoch": 0.31119752111306886,
+      "grad_norm": 0.1231837347149849,
+      "learning_rate": 7.82290568883541e-05,
+      "loss": 1.0322,
+      "step": 5122
+    },
+    {
+      "epoch": 0.31125827814569534,
+      "grad_norm": 0.1808963119983673,
+      "learning_rate": 7.822115521885466e-05,
+      "loss": 1.1309,
+      "step": 5123
+    },
+    {
+      "epoch": 0.3113190351783219,
+      "grad_norm": 0.1607855260372162,
+      "learning_rate": 7.821325251490335e-05,
+      "loss": 1.0626,
+      "step": 5124
+    },
+    {
+      "epoch": 0.3113797922109484,
+      "grad_norm": 0.17335842549800873,
+      "learning_rate": 7.82053487767898e-05,
+      "loss": 1.0651,
+      "step": 5125
+    },
+    {
+      "epoch": 0.31144054924357495,
+      "grad_norm": 0.2586134374141693,
+      "learning_rate": 7.819744400480373e-05,
+      "loss": 1.0913,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3115013062762015,
+      "grad_norm": 0.2554481029510498,
+      "learning_rate": 7.818953819923491e-05,
+      "loss": 1.2143,
+      "step": 5127
+    },
+    {
+      "epoch": 0.311562063308828,
+      "grad_norm": 0.1633116602897644,
+      "learning_rate": 7.818163136037314e-05,
+      "loss": 1.1459,
+      "step": 5128
+    },
+    {
+      "epoch": 0.3116228203414545,
+      "grad_norm": 0.18200641870498657,
+      "learning_rate": 7.817372348850818e-05,
+      "loss": 1.0786,
+      "step": 5129
+    },
+    {
+      "epoch": 0.31168357737408103,
+      "grad_norm": 0.2773749530315399,
+      "learning_rate": 7.816581458392998e-05,
+      "loss": 1.1695,
+      "step": 5130
+    },
+    {
+      "epoch": 0.31174433440670757,
+      "grad_norm": 0.3059140741825104,
+      "learning_rate": 7.815790464692838e-05,
+      "loss": 1.1431,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3118050914393341,
+      "grad_norm": 0.1893763691186905,
+      "learning_rate": 7.814999367779336e-05,
+      "loss": 1.0852,
+      "step": 5132
+    },
+    {
+      "epoch": 0.31186584847196064,
+      "grad_norm": 0.34205004572868347,
+      "learning_rate": 7.814208167681486e-05,
+      "loss": 1.0825,
+      "step": 5133
+    },
+    {
+      "epoch": 0.3119266055045872,
+      "grad_norm": 0.20849581062793732,
+      "learning_rate": 7.813416864428294e-05,
+      "loss": 1.0395,
+      "step": 5134
+    },
+    {
+      "epoch": 0.3119873625372137,
+      "grad_norm": 0.2009948343038559,
+      "learning_rate": 7.81262545804876e-05,
+      "loss": 1.1489,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3120481195698402,
+      "grad_norm": 0.2520940899848938,
+      "learning_rate": 7.811833948571898e-05,
+      "loss": 1.2503,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3121088766024667,
+      "grad_norm": 0.8176343441009521,
+      "learning_rate": 7.81104233602672e-05,
+      "loss": 1.1063,
+      "step": 5137
+    },
+    {
+      "epoch": 0.31216963363509326,
+      "grad_norm": 0.20283618569374084,
+      "learning_rate": 7.810250620442239e-05,
+      "loss": 1.1712,
+      "step": 5138
+    },
+    {
+      "epoch": 0.3122303906677198,
+      "grad_norm": 0.31284841895103455,
+      "learning_rate": 7.80945880184748e-05,
+      "loss": 1.085,
+      "step": 5139
+    },
+    {
+      "epoch": 0.3122911477003463,
+      "grad_norm": 0.31324079632759094,
+      "learning_rate": 7.808666880271466e-05,
+      "loss": 1.193,
+      "step": 5140
+    },
+    {
+      "epoch": 0.31235190473297286,
+      "grad_norm": 0.1671028733253479,
+      "learning_rate": 7.807874855743223e-05,
+      "loss": 1.0407,
+      "step": 5141
+    },
+    {
+      "epoch": 0.31241266176559934,
+      "grad_norm": 0.25555145740509033,
+      "learning_rate": 7.807082728291786e-05,
+      "loss": 1.0618,
+      "step": 5142
+    },
+    {
+      "epoch": 0.3124734187982259,
+      "grad_norm": 0.5085744261741638,
+      "learning_rate": 7.806290497946189e-05,
+      "loss": 1.3183,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3125341758308524,
+      "grad_norm": 0.13113556802272797,
+      "learning_rate": 7.805498164735469e-05,
+      "loss": 1.1089,
+      "step": 5144
+    },
+    {
+      "epoch": 0.31259493286347895,
+      "grad_norm": 0.2609107196331024,
+      "learning_rate": 7.804705728688675e-05,
+      "loss": 1.1589,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3126556898961055,
+      "grad_norm": 0.33784806728363037,
+      "learning_rate": 7.803913189834849e-05,
+      "loss": 1.3375,
+      "step": 5146
+    },
+    {
+      "epoch": 0.312716446928732,
+      "grad_norm": 0.16786982119083405,
+      "learning_rate": 7.803120548203042e-05,
+      "loss": 1.0685,
+      "step": 5147
+    },
+    {
+      "epoch": 0.31277720396135855,
+      "grad_norm": 0.19983232021331787,
+      "learning_rate": 7.80232780382231e-05,
+      "loss": 1.1291,
+      "step": 5148
+    },
+    {
+      "epoch": 0.31283796099398503,
+      "grad_norm": 0.1912095844745636,
+      "learning_rate": 7.80153495672171e-05,
+      "loss": 1.0716,
+      "step": 5149
+    },
+    {
+      "epoch": 0.31289871802661157,
+      "grad_norm": 0.1497095674276352,
+      "learning_rate": 7.800742006930306e-05,
+      "loss": 1.0912,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3129594750592381,
+      "grad_norm": 0.15939418971538544,
+      "learning_rate": 7.799948954477163e-05,
+      "loss": 1.07,
+      "step": 5151
+    },
+    {
+      "epoch": 0.31302023209186464,
+      "grad_norm": 0.19541843235492706,
+      "learning_rate": 7.799155799391349e-05,
+      "loss": 1.014,
+      "step": 5152
+    },
+    {
+      "epoch": 0.31308098912449117,
+      "grad_norm": 0.20815326273441315,
+      "learning_rate": 7.798362541701938e-05,
+      "loss": 1.1034,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3131417461571177,
+      "grad_norm": 0.2816903591156006,
+      "learning_rate": 7.797569181438007e-05,
+      "loss": 1.0955,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3132025031897442,
+      "grad_norm": 0.16486649215221405,
+      "learning_rate": 7.796775718628636e-05,
+      "loss": 1.1044,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3132632602223707,
+      "grad_norm": 0.22881430387496948,
+      "learning_rate": 7.795982153302911e-05,
+      "loss": 1.1176,
+      "step": 5156
+    },
+    {
+      "epoch": 0.31332401725499726,
+      "grad_norm": 0.18074339628219604,
+      "learning_rate": 7.79518848548992e-05,
+      "loss": 1.09,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3133847742876238,
+      "grad_norm": 0.18104323744773865,
+      "learning_rate": 7.794394715218755e-05,
+      "loss": 1.139,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3134455313202503,
+      "grad_norm": 0.17354963719844818,
+      "learning_rate": 7.793600842518512e-05,
+      "loss": 1.142,
+      "step": 5159
+    },
+    {
+      "epoch": 0.31350628835287686,
+      "grad_norm": 3.4281907081604004,
+      "learning_rate": 7.79280686741829e-05,
+      "loss": 1.1442,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3135670453855034,
+      "grad_norm": 0.2361103743314743,
+      "learning_rate": 7.792012789947192e-05,
+      "loss": 1.119,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3136278024181299,
+      "grad_norm": 0.20149557292461395,
+      "learning_rate": 7.791218610134323e-05,
+      "loss": 1.1081,
+      "step": 5162
+    },
+    {
+      "epoch": 0.3136885594507564,
+      "grad_norm": 0.13651804625988007,
+      "learning_rate": 7.7904243280088e-05,
+      "loss": 1.0802,
+      "step": 5163
+    },
+    {
+      "epoch": 0.31374931648338295,
+      "grad_norm": 0.18828308582305908,
+      "learning_rate": 7.789629943599732e-05,
+      "loss": 1.1238,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3138100735160095,
+      "grad_norm": 0.25163477659225464,
+      "learning_rate": 7.78883545693624e-05,
+      "loss": 1.2072,
+      "step": 5165
+    },
+    {
+      "epoch": 0.313870830548636,
+      "grad_norm": 0.17760860919952393,
+      "learning_rate": 7.788040868047445e-05,
+      "loss": 1.0718,
+      "step": 5166
+    },
+    {
+      "epoch": 0.31393158758126255,
+      "grad_norm": 0.16761569678783417,
+      "learning_rate": 7.787246176962474e-05,
+      "loss": 1.1063,
+      "step": 5167
+    },
+    {
+      "epoch": 0.3139923446138891,
+      "grad_norm": 0.27144646644592285,
+      "learning_rate": 7.786451383710455e-05,
+      "loss": 1.1528,
+      "step": 5168
+    },
+    {
+      "epoch": 0.31405310164651556,
+      "grad_norm": 0.15995943546295166,
+      "learning_rate": 7.78565648832052e-05,
+      "loss": 1.1231,
+      "step": 5169
+    },
+    {
+      "epoch": 0.3141138586791421,
+      "grad_norm": 0.15376867353916168,
+      "learning_rate": 7.784861490821811e-05,
+      "loss": 1.0839,
+      "step": 5170
+    },
+    {
+      "epoch": 0.31417461571176863,
+      "grad_norm": 0.18771442770957947,
+      "learning_rate": 7.784066391243465e-05,
+      "loss": 1.1233,
+      "step": 5171
+    },
+    {
+      "epoch": 0.31423537274439517,
+      "grad_norm": 1.352339267730713,
+      "learning_rate": 7.783271189614626e-05,
+      "loss": 1.0791,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3142961297770217,
+      "grad_norm": 0.12809929251670837,
+      "learning_rate": 7.782475885964444e-05,
+      "loss": 1.0845,
+      "step": 5173
+    },
+    {
+      "epoch": 0.31435688680964824,
+      "grad_norm": 0.1668839007616043,
+      "learning_rate": 7.78168048032207e-05,
+      "loss": 1.1146,
+      "step": 5174
+    },
+    {
+      "epoch": 0.3144176438422747,
+      "grad_norm": 0.42424067854881287,
+      "learning_rate": 7.780884972716662e-05,
+      "loss": 1.0524,
+      "step": 5175
+    },
+    {
+      "epoch": 0.31447840087490125,
+      "grad_norm": 0.14323019981384277,
+      "learning_rate": 7.780089363177376e-05,
+      "loss": 1.0056,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3145391579075278,
+      "grad_norm": 0.19844859838485718,
+      "learning_rate": 7.779293651733379e-05,
+      "loss": 1.1222,
+      "step": 5177
+    },
+    {
+      "epoch": 0.3145999149401543,
+      "grad_norm": 0.26833683252334595,
+      "learning_rate": 7.778497838413834e-05,
+      "loss": 1.1296,
+      "step": 5178
+    },
+    {
+      "epoch": 0.31466067197278086,
+      "grad_norm": 0.1764657348394394,
+      "learning_rate": 7.777701923247916e-05,
+      "loss": 1.0837,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3147214290054074,
+      "grad_norm": 0.24348610639572144,
+      "learning_rate": 7.776905906264795e-05,
+      "loss": 1.1094,
+      "step": 5180
+    },
+    {
+      "epoch": 0.31478218603803393,
+      "grad_norm": 0.404840886592865,
+      "learning_rate": 7.776109787493651e-05,
+      "loss": 1.1243,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3148429430706604,
+      "grad_norm": 0.19809971749782562,
+      "learning_rate": 7.775313566963667e-05,
+      "loss": 1.104,
+      "step": 5182
+    },
+    {
+      "epoch": 0.31490370010328694,
+      "grad_norm": 0.1251288503408432,
+      "learning_rate": 7.77451724470403e-05,
+      "loss": 1.0254,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3149644571359135,
+      "grad_norm": 0.18405430018901825,
+      "learning_rate": 7.773720820743922e-05,
+      "loss": 1.109,
+      "step": 5184
+    },
+    {
+      "epoch": 0.31502521416854,
+      "grad_norm": 0.6925203204154968,
+      "learning_rate": 7.772924295112545e-05,
+      "loss": 1.0947,
+      "step": 5185
+    },
+    {
+      "epoch": 0.31508597120116655,
+      "grad_norm": 0.26931998133659363,
+      "learning_rate": 7.772127667839092e-05,
+      "loss": 1.074,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3151467282337931,
+      "grad_norm": 0.14279203116893768,
+      "learning_rate": 7.771330938952761e-05,
+      "loss": 1.0851,
+      "step": 5187
+    },
+    {
+      "epoch": 0.31520748526641956,
+      "grad_norm": 0.26213353872299194,
+      "learning_rate": 7.770534108482761e-05,
+      "loss": 1.0812,
+      "step": 5188
+    },
+    {
+      "epoch": 0.3152682422990461,
+      "grad_norm": 0.4701729714870453,
+      "learning_rate": 7.769737176458296e-05,
+      "loss": 1.1508,
+      "step": 5189
+    },
+    {
+      "epoch": 0.31532899933167263,
+      "grad_norm": 0.4827955961227417,
+      "learning_rate": 7.768940142908581e-05,
+      "loss": 1.0391,
+      "step": 5190
+    },
+    {
+      "epoch": 0.31538975636429917,
+      "grad_norm": 0.1750054508447647,
+      "learning_rate": 7.768143007862829e-05,
+      "loss": 1.0941,
+      "step": 5191
+    },
+    {
+      "epoch": 0.3154505133969257,
+      "grad_norm": 0.19122281670570374,
+      "learning_rate": 7.767345771350261e-05,
+      "loss": 1.1639,
+      "step": 5192
+    },
+    {
+      "epoch": 0.31551127042955224,
+      "grad_norm": 6.284471035003662,
+      "learning_rate": 7.766548433400096e-05,
+      "loss": 1.1548,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3155720274621788,
+      "grad_norm": 0.14756184816360474,
+      "learning_rate": 7.765750994041566e-05,
+      "loss": 1.0822,
+      "step": 5194
+    },
+    {
+      "epoch": 0.31563278449480525,
+      "grad_norm": 0.16128914058208466,
+      "learning_rate": 7.764953453303898e-05,
+      "loss": 1.1433,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3156935415274318,
+      "grad_norm": 0.13844360411167145,
+      "learning_rate": 7.764155811216327e-05,
+      "loss": 1.0279,
+      "step": 5196
+    },
+    {
+      "epoch": 0.3157542985600583,
+      "grad_norm": 0.43880733847618103,
+      "learning_rate": 7.76335806780809e-05,
+      "loss": 1.2821,
+      "step": 5197
+    },
+    {
+      "epoch": 0.31581505559268486,
+      "grad_norm": 0.2243174910545349,
+      "learning_rate": 7.76256022310843e-05,
+      "loss": 1.1521,
+      "step": 5198
+    },
+    {
+      "epoch": 0.3158758126253114,
+      "grad_norm": 0.27900493144989014,
+      "learning_rate": 7.761762277146589e-05,
+      "loss": 1.1816,
+      "step": 5199
+    },
+    {
+      "epoch": 0.3159365696579379,
+      "grad_norm": 0.24200977385044098,
+      "learning_rate": 7.76096422995182e-05,
+      "loss": 1.1435,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3159973266905644,
+      "grad_norm": 8.08497142791748,
+      "learning_rate": 7.760166081553373e-05,
+      "loss": 1.1547,
+      "step": 5201
+    },
+    {
+      "epoch": 0.31605808372319094,
+      "grad_norm": 0.17424650490283966,
+      "learning_rate": 7.759367831980504e-05,
+      "loss": 1.0744,
+      "step": 5202
+    },
+    {
+      "epoch": 0.3161188407558175,
+      "grad_norm": 0.2214883267879486,
+      "learning_rate": 7.758569481262475e-05,
+      "loss": 1.1992,
+      "step": 5203
+    },
+    {
+      "epoch": 0.316179597788444,
+      "grad_norm": 0.4513641893863678,
+      "learning_rate": 7.757771029428547e-05,
+      "loss": 1.1952,
+      "step": 5204
+    },
+    {
+      "epoch": 0.31624035482107055,
+      "grad_norm": 0.24038437008857727,
+      "learning_rate": 7.756972476507989e-05,
+      "loss": 1.1015,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3163011118536971,
+      "grad_norm": 0.1677442342042923,
+      "learning_rate": 7.756173822530071e-05,
+      "loss": 1.1296,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3163618688863236,
+      "grad_norm": 0.14486092329025269,
+      "learning_rate": 7.755375067524071e-05,
+      "loss": 1.0801,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3164226259189501,
+      "grad_norm": 0.17801958322525024,
+      "learning_rate": 7.754576211519264e-05,
+      "loss": 1.1745,
+      "step": 5208
+    },
+    {
+      "epoch": 0.31648338295157663,
+      "grad_norm": 0.18977457284927368,
+      "learning_rate": 7.753777254544934e-05,
+      "loss": 1.127,
+      "step": 5209
+    },
+    {
+      "epoch": 0.31654413998420317,
+      "grad_norm": 0.17817537486553192,
+      "learning_rate": 7.752978196630368e-05,
+      "loss": 1.1358,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3166048970168297,
+      "grad_norm": 0.18756113946437836,
+      "learning_rate": 7.752179037804851e-05,
+      "loss": 1.0939,
+      "step": 5211
+    },
+    {
+      "epoch": 0.31666565404945624,
+      "grad_norm": 0.14955630898475647,
+      "learning_rate": 7.75137977809768e-05,
+      "loss": 1.0801,
+      "step": 5212
+    },
+    {
+      "epoch": 0.31672641108208277,
+      "grad_norm": 5.002853870391846,
+      "learning_rate": 7.750580417538153e-05,
+      "loss": 1.1666,
+      "step": 5213
+    },
+    {
+      "epoch": 0.31678716811470925,
+      "grad_norm": 0.21667887270450592,
+      "learning_rate": 7.749780956155568e-05,
+      "loss": 1.0649,
+      "step": 5214
+    },
+    {
+      "epoch": 0.3168479251473358,
+      "grad_norm": 0.4345817267894745,
+      "learning_rate": 7.74898139397923e-05,
+      "loss": 1.0416,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3169086821799623,
+      "grad_norm": 0.3247835338115692,
+      "learning_rate": 7.748181731038449e-05,
+      "loss": 1.1617,
+      "step": 5216
+    },
+    {
+      "epoch": 0.31696943921258885,
+      "grad_norm": 0.18982625007629395,
+      "learning_rate": 7.747381967362536e-05,
+      "loss": 1.0961,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3170301962452154,
+      "grad_norm": 37.097442626953125,
+      "learning_rate": 7.746582102980804e-05,
+      "loss": 1.1801,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3170909532778419,
+      "grad_norm": 0.8889961242675781,
+      "learning_rate": 7.745782137922574e-05,
+      "loss": 1.1257,
+      "step": 5219
+    },
+    {
+      "epoch": 0.31715171031046846,
+      "grad_norm": 0.2838846743106842,
+      "learning_rate": 7.744982072217171e-05,
+      "loss": 1.1328,
+      "step": 5220
+    },
+    {
+      "epoch": 0.31721246734309494,
+      "grad_norm": 0.16100822389125824,
+      "learning_rate": 7.74418190589392e-05,
+      "loss": 1.0843,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3172732243757215,
+      "grad_norm": 0.3542579710483551,
+      "learning_rate": 7.743381638982148e-05,
+      "loss": 1.0543,
+      "step": 5222
+    },
+    {
+      "epoch": 0.317333981408348,
+      "grad_norm": 0.13324719667434692,
+      "learning_rate": 7.742581271511192e-05,
+      "loss": 1.0876,
+      "step": 5223
+    },
+    {
+      "epoch": 0.31739473844097454,
+      "grad_norm": 0.1612965613603592,
+      "learning_rate": 7.741780803510391e-05,
+      "loss": 1.1505,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3174554954736011,
+      "grad_norm": 0.1852015107870102,
+      "learning_rate": 7.740980235009084e-05,
+      "loss": 1.1412,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3175162525062276,
+      "grad_norm": 0.15180452167987823,
+      "learning_rate": 7.740179566036617e-05,
+      "loss": 1.0735,
+      "step": 5226
+    },
+    {
+      "epoch": 0.31757700953885415,
+      "grad_norm": 0.14446836709976196,
+      "learning_rate": 7.739378796622338e-05,
+      "loss": 1.0564,
+      "step": 5227
+    },
+    {
+      "epoch": 0.31763776657148063,
+      "grad_norm": 1.1163607835769653,
+      "learning_rate": 7.738577926795601e-05,
+      "loss": 1.0983,
+      "step": 5228
+    },
+    {
+      "epoch": 0.31769852360410716,
+      "grad_norm": 2.0002286434173584,
+      "learning_rate": 7.737776956585759e-05,
+      "loss": 1.1263,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3177592806367337,
+      "grad_norm": 0.14760617911815643,
+      "learning_rate": 7.736975886022173e-05,
+      "loss": 1.02,
+      "step": 5230
+    },
+    {
+      "epoch": 0.31782003766936023,
+      "grad_norm": 0.20231235027313232,
+      "learning_rate": 7.736174715134208e-05,
+      "loss": 1.0735,
+      "step": 5231
+    },
+    {
+      "epoch": 0.31788079470198677,
+      "grad_norm": 0.12804539501667023,
+      "learning_rate": 7.735373443951229e-05,
+      "loss": 1.0808,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3179415517346133,
+      "grad_norm": 0.22206659615039825,
+      "learning_rate": 7.734572072502609e-05,
+      "loss": 1.1571,
+      "step": 5233
+    },
+    {
+      "epoch": 0.3180023087672398,
+      "grad_norm": 0.2367912232875824,
+      "learning_rate": 7.733770600817721e-05,
+      "loss": 1.0904,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3180630657998663,
+      "grad_norm": 0.5960977077484131,
+      "learning_rate": 7.732969028925943e-05,
+      "loss": 1.0733,
+      "step": 5235
+    },
+    {
+      "epoch": 0.31812382283249285,
+      "grad_norm": 0.219915509223938,
+      "learning_rate": 7.732167356856655e-05,
+      "loss": 1.2214,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3181845798651194,
+      "grad_norm": 0.43676573038101196,
+      "learning_rate": 7.731365584639248e-05,
+      "loss": 1.1628,
+      "step": 5237
+    },
+    {
+      "epoch": 0.3182453368977459,
+      "grad_norm": 0.45794686675071716,
+      "learning_rate": 7.730563712303104e-05,
+      "loss": 1.1118,
+      "step": 5238
+    },
+    {
+      "epoch": 0.31830609393037246,
+      "grad_norm": 0.13144971430301666,
+      "learning_rate": 7.729761739877622e-05,
+      "loss": 1.0708,
+      "step": 5239
+    },
+    {
+      "epoch": 0.318366850962999,
+      "grad_norm": 0.23507849872112274,
+      "learning_rate": 7.728959667392195e-05,
+      "loss": 1.1135,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3184276079956255,
+      "grad_norm": 0.2360433042049408,
+      "learning_rate": 7.728157494876223e-05,
+      "loss": 1.0795,
+      "step": 5241
+    },
+    {
+      "epoch": 0.318488365028252,
+      "grad_norm": 0.2307671159505844,
+      "learning_rate": 7.727355222359112e-05,
+      "loss": 1.1673,
+      "step": 5242
+    },
+    {
+      "epoch": 0.31854912206087854,
+      "grad_norm": 0.20978204905986786,
+      "learning_rate": 7.726552849870267e-05,
+      "loss": 1.1977,
+      "step": 5243
+    },
+    {
+      "epoch": 0.3186098790935051,
+      "grad_norm": 0.22503332793712616,
+      "learning_rate": 7.7257503774391e-05,
+      "loss": 1.2425,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3186706361261316,
+      "grad_norm": 0.15387806296348572,
+      "learning_rate": 7.724947805095027e-05,
+      "loss": 1.0151,
+      "step": 5245
+    },
+    {
+      "epoch": 0.31873139315875815,
+      "grad_norm": 0.13361290097236633,
+      "learning_rate": 7.724145132867466e-05,
+      "loss": 1.067,
+      "step": 5246
+    },
+    {
+      "epoch": 0.3187921501913846,
+      "grad_norm": 0.13001136481761932,
+      "learning_rate": 7.723342360785837e-05,
+      "loss": 1.0508,
+      "step": 5247
+    },
+    {
+      "epoch": 0.31885290722401116,
+      "grad_norm": 0.22706355154514313,
+      "learning_rate": 7.722539488879569e-05,
+      "loss": 1.2687,
+      "step": 5248
+    },
+    {
+      "epoch": 0.3189136642566377,
+      "grad_norm": 0.16564755141735077,
+      "learning_rate": 7.721736517178089e-05,
+      "loss": 1.0756,
+      "step": 5249
+    },
+    {
+      "epoch": 0.31897442128926423,
+      "grad_norm": 0.2275359034538269,
+      "learning_rate": 7.72093344571083e-05,
+      "loss": 1.0813,
+      "step": 5250
+    },
+    {
+      "epoch": 0.31903517832189077,
+      "grad_norm": 0.12234020233154297,
+      "learning_rate": 7.72013027450723e-05,
+      "loss": 1.0399,
+      "step": 5251
+    },
+    {
+      "epoch": 0.3190959353545173,
+      "grad_norm": 0.19603148102760315,
+      "learning_rate": 7.71932700359673e-05,
+      "loss": 1.1804,
+      "step": 5252
+    },
+    {
+      "epoch": 0.31915669238714384,
+      "grad_norm": 0.19225366413593292,
+      "learning_rate": 7.718523633008772e-05,
+      "loss": 1.1531,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3192174494197703,
+      "grad_norm": 0.1957905888557434,
+      "learning_rate": 7.717720162772806e-05,
+      "loss": 1.1132,
+      "step": 5254
+    },
+    {
+      "epoch": 0.31927820645239685,
+      "grad_norm": 0.4529777765274048,
+      "learning_rate": 7.716916592918281e-05,
+      "loss": 1.2141,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3193389634850234,
+      "grad_norm": 0.2204008251428604,
+      "learning_rate": 7.716112923474654e-05,
+      "loss": 1.1488,
+      "step": 5256
+    },
+    {
+      "epoch": 0.3193997205176499,
+      "grad_norm": 0.24447494745254517,
+      "learning_rate": 7.715309154471382e-05,
+      "loss": 1.256,
+      "step": 5257
+    },
+    {
+      "epoch": 0.31946047755027646,
+      "grad_norm": 0.16351109743118286,
+      "learning_rate": 7.714505285937928e-05,
+      "loss": 1.0119,
+      "step": 5258
+    },
+    {
+      "epoch": 0.319521234582903,
+      "grad_norm": 0.27826523780822754,
+      "learning_rate": 7.713701317903759e-05,
+      "loss": 1.1552,
+      "step": 5259
+    },
+    {
+      "epoch": 0.31958199161552947,
+      "grad_norm": 0.4236544668674469,
+      "learning_rate": 7.712897250398343e-05,
+      "loss": 1.0764,
+      "step": 5260
+    },
+    {
+      "epoch": 0.319642748648156,
+      "grad_norm": 0.19925330579280853,
+      "learning_rate": 7.712093083451153e-05,
+      "loss": 1.1677,
+      "step": 5261
+    },
+    {
+      "epoch": 0.31970350568078254,
+      "grad_norm": 0.1551828682422638,
+      "learning_rate": 7.711288817091669e-05,
+      "loss": 1.0962,
+      "step": 5262
+    },
+    {
+      "epoch": 0.3197642627134091,
+      "grad_norm": 0.24796655774116516,
+      "learning_rate": 7.710484451349367e-05,
+      "loss": 1.1001,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3198250197460356,
+      "grad_norm": 0.17019613087177277,
+      "learning_rate": 7.709679986253735e-05,
+      "loss": 1.0776,
+      "step": 5264
+    },
+    {
+      "epoch": 0.31988577677866215,
+      "grad_norm": 0.1914503425359726,
+      "learning_rate": 7.708875421834258e-05,
+      "loss": 1.069,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3199465338112887,
+      "grad_norm": 0.287005215883255,
+      "learning_rate": 7.70807075812043e-05,
+      "loss": 1.2286,
+      "step": 5266
+    },
+    {
+      "epoch": 0.32000729084391516,
+      "grad_norm": 0.4092259407043457,
+      "learning_rate": 7.707265995141744e-05,
+      "loss": 1.1158,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3200680478765417,
+      "grad_norm": 0.21153295040130615,
+      "learning_rate": 7.7064611329277e-05,
+      "loss": 1.1045,
+      "step": 5268
+    },
+    {
+      "epoch": 0.32012880490916823,
+      "grad_norm": 0.1494741588830948,
+      "learning_rate": 7.705656171507802e-05,
+      "loss": 1.0573,
+      "step": 5269
+    },
+    {
+      "epoch": 0.32018956194179476,
+      "grad_norm": 0.26257410645484924,
+      "learning_rate": 7.70485111091155e-05,
+      "loss": 1.1391,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3202503189744213,
+      "grad_norm": 0.17116105556488037,
+      "learning_rate": 7.704045951168461e-05,
+      "loss": 1.0949,
+      "step": 5271
+    },
+    {
+      "epoch": 0.32031107600704783,
+      "grad_norm": 0.17433473467826843,
+      "learning_rate": 7.703240692308044e-05,
+      "loss": 1.1713,
+      "step": 5272
+    },
+    {
+      "epoch": 0.32037183303967437,
+      "grad_norm": 0.162774920463562,
+      "learning_rate": 7.702435334359817e-05,
+      "loss": 1.1501,
+      "step": 5273
+    },
+    {
+      "epoch": 0.32043259007230085,
+      "grad_norm": 0.1266329288482666,
+      "learning_rate": 7.7016298773533e-05,
+      "loss": 1.0956,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3204933471049274,
+      "grad_norm": 0.19025348126888275,
+      "learning_rate": 7.700824321318019e-05,
+      "loss": 1.0888,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3205541041375539,
+      "grad_norm": 0.18125171959400177,
+      "learning_rate": 7.7000186662835e-05,
+      "loss": 1.1075,
+      "step": 5276
+    },
+    {
+      "epoch": 0.32061486117018045,
+      "grad_norm": 0.11788712441921234,
+      "learning_rate": 7.699212912279274e-05,
+      "loss": 1.0304,
+      "step": 5277
+    },
+    {
+      "epoch": 0.320675618202807,
+      "grad_norm": 0.4160124957561493,
+      "learning_rate": 7.698407059334877e-05,
+      "loss": 1.2273,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3207363752354335,
+      "grad_norm": 1.816224217414856,
+      "learning_rate": 7.697601107479846e-05,
+      "loss": 1.1682,
+      "step": 5279
+    },
+    {
+      "epoch": 0.32079713226806,
+      "grad_norm": 0.17960700392723083,
+      "learning_rate": 7.696795056743727e-05,
+      "loss": 1.155,
+      "step": 5280
+    },
+    {
+      "epoch": 0.32085788930068654,
+      "grad_norm": 0.22117215394973755,
+      "learning_rate": 7.695988907156066e-05,
+      "loss": 1.2285,
+      "step": 5281
+    },
+    {
+      "epoch": 0.3209186463333131,
+      "grad_norm": 0.22642168402671814,
+      "learning_rate": 7.695182658746407e-05,
+      "loss": 1.1356,
+      "step": 5282
+    },
+    {
+      "epoch": 0.3209794033659396,
+      "grad_norm": 2.557623863220215,
+      "learning_rate": 7.694376311544307e-05,
+      "loss": 1.1825,
+      "step": 5283
+    },
+    {
+      "epoch": 0.32104016039856614,
+      "grad_norm": 0.19348691403865814,
+      "learning_rate": 7.693569865579325e-05,
+      "loss": 1.1193,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3211009174311927,
+      "grad_norm": 0.1772070974111557,
+      "learning_rate": 7.692763320881015e-05,
+      "loss": 1.2338,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3211616744638192,
+      "grad_norm": 3.7062413692474365,
+      "learning_rate": 7.691956677478948e-05,
+      "loss": 1.2231,
+      "step": 5286
+    },
+    {
+      "epoch": 0.3212224314964457,
+      "grad_norm": 0.14694485068321228,
+      "learning_rate": 7.691149935402689e-05,
+      "loss": 1.0527,
+      "step": 5287
+    },
+    {
+      "epoch": 0.32128318852907223,
+      "grad_norm": 0.35769063234329224,
+      "learning_rate": 7.690343094681807e-05,
+      "loss": 1.1529,
+      "step": 5288
+    },
+    {
+      "epoch": 0.32134394556169876,
+      "grad_norm": 0.19548434019088745,
+      "learning_rate": 7.68953615534588e-05,
+      "loss": 0.9972,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3214047025943253,
+      "grad_norm": 0.2896392345428467,
+      "learning_rate": 7.688729117424484e-05,
+      "loss": 1.0352,
+      "step": 5290
+    },
+    {
+      "epoch": 0.32146545962695183,
+      "grad_norm": 0.1843344122171402,
+      "learning_rate": 7.687921980947206e-05,
+      "loss": 1.063,
+      "step": 5291
+    },
+    {
+      "epoch": 0.32152621665957837,
+      "grad_norm": 0.23084795475006104,
+      "learning_rate": 7.687114745943624e-05,
+      "loss": 1.2774,
+      "step": 5292
+    },
+    {
+      "epoch": 0.32158697369220485,
+      "grad_norm": 0.24386334419250488,
+      "learning_rate": 7.686307412443334e-05,
+      "loss": 1.1572,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3216477307248314,
+      "grad_norm": 0.18706533312797546,
+      "learning_rate": 7.685499980475927e-05,
+      "loss": 1.1078,
+      "step": 5294
+    },
+    {
+      "epoch": 0.3217084877574579,
+      "grad_norm": 0.16719487309455872,
+      "learning_rate": 7.684692450070996e-05,
+      "loss": 1.0529,
+      "step": 5295
+    },
+    {
+      "epoch": 0.32176924479008445,
+      "grad_norm": 0.1591726392507553,
+      "learning_rate": 7.683884821258147e-05,
+      "loss": 1.0388,
+      "step": 5296
+    },
+    {
+      "epoch": 0.321830001822711,
+      "grad_norm": 0.22185593843460083,
+      "learning_rate": 7.68307709406698e-05,
+      "loss": 1.0579,
+      "step": 5297
+    },
+    {
+      "epoch": 0.3218907588553375,
+      "grad_norm": 0.14457620680332184,
+      "learning_rate": 7.682269268527104e-05,
+      "loss": 1.056,
+      "step": 5298
+    },
+    {
+      "epoch": 0.32195151588796406,
+      "grad_norm": 0.22602593898773193,
+      "learning_rate": 7.681461344668131e-05,
+      "loss": 1.1887,
+      "step": 5299
+    },
+    {
+      "epoch": 0.32201227292059054,
+      "grad_norm": 0.1864214390516281,
+      "learning_rate": 7.680653322519671e-05,
+      "loss": 1.048,
+      "step": 5300
+    },
+    {
+      "epoch": 0.32207302995321707,
+      "grad_norm": 0.19677239656448364,
+      "learning_rate": 7.679845202111347e-05,
+      "loss": 1.1757,
+      "step": 5301
+    },
+    {
+      "epoch": 0.3221337869858436,
+      "grad_norm": 0.3017561435699463,
+      "learning_rate": 7.67903698347278e-05,
+      "loss": 1.1122,
+      "step": 5302
+    },
+    {
+      "epoch": 0.32219454401847014,
+      "grad_norm": 0.12074954807758331,
+      "learning_rate": 7.678228666633592e-05,
+      "loss": 1.0468,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3222553010510967,
+      "grad_norm": 0.17102929949760437,
+      "learning_rate": 7.677420251623418e-05,
+      "loss": 1.1208,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3223160580837232,
+      "grad_norm": 0.3344148099422455,
+      "learning_rate": 7.676611738471882e-05,
+      "loss": 1.0861,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3223768151163497,
+      "grad_norm": 0.30935409665107727,
+      "learning_rate": 7.675803127208629e-05,
+      "loss": 1.0522,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3224375721489762,
+      "grad_norm": 0.12636761367321014,
+      "learning_rate": 7.674994417863295e-05,
+      "loss": 1.0914,
+      "step": 5307
+    },
+    {
+      "epoch": 0.32249832918160276,
+      "grad_norm": 0.20348799228668213,
+      "learning_rate": 7.674185610465523e-05,
+      "loss": 1.1769,
+      "step": 5308
+    },
+    {
+      "epoch": 0.3225590862142293,
+      "grad_norm": 0.21296432614326477,
+      "learning_rate": 7.673376705044962e-05,
+      "loss": 1.0786,
+      "step": 5309
+    },
+    {
+      "epoch": 0.32261984324685583,
+      "grad_norm": 1.714321494102478,
+      "learning_rate": 7.67256770163126e-05,
+      "loss": 1.0796,
+      "step": 5310
+    },
+    {
+      "epoch": 0.32268060027948237,
+      "grad_norm": 0.27067720890045166,
+      "learning_rate": 7.671758600254071e-05,
+      "loss": 1.1471,
+      "step": 5311
+    },
+    {
+      "epoch": 0.3227413573121089,
+      "grad_norm": 0.14980441331863403,
+      "learning_rate": 7.670949400943054e-05,
+      "loss": 1.1106,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3228021143447354,
+      "grad_norm": 0.2398141771554947,
+      "learning_rate": 7.670140103727871e-05,
+      "loss": 1.2155,
+      "step": 5313
+    },
+    {
+      "epoch": 0.3228628713773619,
+      "grad_norm": 0.3515942096710205,
+      "learning_rate": 7.66933070863819e-05,
+      "loss": 1.0888,
+      "step": 5314
+    },
+    {
+      "epoch": 0.32292362840998845,
+      "grad_norm": 0.1338648647069931,
+      "learning_rate": 7.668521215703672e-05,
+      "loss": 1.0556,
+      "step": 5315
+    },
+    {
+      "epoch": 0.322984385442615,
+      "grad_norm": 0.40135011076927185,
+      "learning_rate": 7.667711624953993e-05,
+      "loss": 1.0619,
+      "step": 5316
+    },
+    {
+      "epoch": 0.3230451424752415,
+      "grad_norm": 0.2750815749168396,
+      "learning_rate": 7.66690193641883e-05,
+      "loss": 1.035,
+      "step": 5317
+    },
+    {
+      "epoch": 0.32310589950786806,
+      "grad_norm": 0.1642073690891266,
+      "learning_rate": 7.66609215012786e-05,
+      "loss": 1.1097,
+      "step": 5318
+    },
+    {
+      "epoch": 0.32316665654049453,
+      "grad_norm": 0.16084831953048706,
+      "learning_rate": 7.665282266110768e-05,
+      "loss": 1.074,
+      "step": 5319
+    },
+    {
+      "epoch": 0.32322741357312107,
+      "grad_norm": 0.25044238567352295,
+      "learning_rate": 7.664472284397239e-05,
+      "loss": 1.1129,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3232881706057476,
+      "grad_norm": 0.1886386126279831,
+      "learning_rate": 7.663662205016964e-05,
+      "loss": 1.1071,
+      "step": 5321
+    },
+    {
+      "epoch": 0.32334892763837414,
+      "grad_norm": 0.2613964378833771,
+      "learning_rate": 7.662852027999636e-05,
+      "loss": 1.2313,
+      "step": 5322
+    },
+    {
+      "epoch": 0.3234096846710007,
+      "grad_norm": 0.3531823456287384,
+      "learning_rate": 7.662041753374949e-05,
+      "loss": 1.0519,
+      "step": 5323
+    },
+    {
+      "epoch": 0.3234704417036272,
+      "grad_norm": 0.1669054925441742,
+      "learning_rate": 7.661231381172611e-05,
+      "loss": 1.0758,
+      "step": 5324
+    },
+    {
+      "epoch": 0.32353119873625374,
+      "grad_norm": 0.1835552304983139,
+      "learning_rate": 7.660420911422321e-05,
+      "loss": 1.2275,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3235919557688802,
+      "grad_norm": 0.1822025626897812,
+      "learning_rate": 7.659610344153787e-05,
+      "loss": 1.2218,
+      "step": 5326
+    },
+    {
+      "epoch": 0.32365271280150676,
+      "grad_norm": 1.3519262075424194,
+      "learning_rate": 7.658799679396724e-05,
+      "loss": 1.0978,
+      "step": 5327
+    },
+    {
+      "epoch": 0.3237134698341333,
+      "grad_norm": 0.19118796288967133,
+      "learning_rate": 7.657988917180843e-05,
+      "loss": 1.1404,
+      "step": 5328
+    },
+    {
+      "epoch": 0.32377422686675983,
+      "grad_norm": 25.419490814208984,
+      "learning_rate": 7.657178057535866e-05,
+      "loss": 1.3757,
+      "step": 5329
+    },
+    {
+      "epoch": 0.32383498389938636,
+      "grad_norm": 0.20100799202919006,
+      "learning_rate": 7.656367100491512e-05,
+      "loss": 1.0809,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3238957409320129,
+      "grad_norm": 0.3121403753757477,
+      "learning_rate": 7.65555604607751e-05,
+      "loss": 1.1042,
+      "step": 5331
+    },
+    {
+      "epoch": 0.32395649796463943,
+      "grad_norm": 0.16733300685882568,
+      "learning_rate": 7.654744894323586e-05,
+      "loss": 1.1085,
+      "step": 5332
+    },
+    {
+      "epoch": 0.3240172549972659,
+      "grad_norm": 0.1692768931388855,
+      "learning_rate": 7.653933645259475e-05,
+      "loss": 1.1277,
+      "step": 5333
+    },
+    {
+      "epoch": 0.32407801202989245,
+      "grad_norm": 0.3412485122680664,
+      "learning_rate": 7.653122298914915e-05,
+      "loss": 1.0951,
+      "step": 5334
+    },
+    {
+      "epoch": 0.324138769062519,
+      "grad_norm": 0.15294376015663147,
+      "learning_rate": 7.652310855319643e-05,
+      "loss": 1.1336,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3241995260951455,
+      "grad_norm": 0.20286858081817627,
+      "learning_rate": 7.651499314503403e-05,
+      "loss": 1.0978,
+      "step": 5336
+    },
+    {
+      "epoch": 0.32426028312777205,
+      "grad_norm": 0.14353255927562714,
+      "learning_rate": 7.650687676495945e-05,
+      "loss": 1.0742,
+      "step": 5337
+    },
+    {
+      "epoch": 0.3243210401603986,
+      "grad_norm": 0.1851993054151535,
+      "learning_rate": 7.649875941327018e-05,
+      "loss": 1.0608,
+      "step": 5338
+    },
+    {
+      "epoch": 0.32438179719302507,
+      "grad_norm": 0.15749365091323853,
+      "learning_rate": 7.649064109026374e-05,
+      "loss": 1.1293,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3244425542256516,
+      "grad_norm": 0.14843803644180298,
+      "learning_rate": 7.648252179623775e-05,
+      "loss": 1.124,
+      "step": 5340
+    },
+    {
+      "epoch": 0.32450331125827814,
+      "grad_norm": 0.15867199003696442,
+      "learning_rate": 7.647440153148979e-05,
+      "loss": 1.125,
+      "step": 5341
+    },
+    {
+      "epoch": 0.3245640682909047,
+      "grad_norm": 0.19077855348587036,
+      "learning_rate": 7.646628029631754e-05,
+      "loss": 1.1323,
+      "step": 5342
+    },
+    {
+      "epoch": 0.3246248253235312,
+      "grad_norm": 0.2224782556295395,
+      "learning_rate": 7.645815809101866e-05,
+      "loss": 1.2911,
+      "step": 5343
+    },
+    {
+      "epoch": 0.32468558235615774,
+      "grad_norm": 0.13767606019973755,
+      "learning_rate": 7.645003491589089e-05,
+      "loss": 1.0544,
+      "step": 5344
+    },
+    {
+      "epoch": 0.3247463393887843,
+      "grad_norm": 0.9019991755485535,
+      "learning_rate": 7.644191077123197e-05,
+      "loss": 1.202,
+      "step": 5345
+    },
+    {
+      "epoch": 0.32480709642141076,
+      "grad_norm": 0.8234071135520935,
+      "learning_rate": 7.64337856573397e-05,
+      "loss": 1.1843,
+      "step": 5346
+    },
+    {
+      "epoch": 0.3248678534540373,
+      "grad_norm": 0.14335016906261444,
+      "learning_rate": 7.64256595745119e-05,
+      "loss": 1.0796,
+      "step": 5347
+    },
+    {
+      "epoch": 0.3249286104866638,
+      "grad_norm": 0.19248712062835693,
+      "learning_rate": 7.641753252304646e-05,
+      "loss": 1.116,
+      "step": 5348
+    },
+    {
+      "epoch": 0.32498936751929036,
+      "grad_norm": 0.2020735740661621,
+      "learning_rate": 7.640940450324124e-05,
+      "loss": 1.1848,
+      "step": 5349
+    },
+    {
+      "epoch": 0.3250501245519169,
+      "grad_norm": 0.14655005931854248,
+      "learning_rate": 7.640127551539419e-05,
+      "loss": 1.1154,
+      "step": 5350
+    },
+    {
+      "epoch": 0.32511088158454343,
+      "grad_norm": 0.2234189659357071,
+      "learning_rate": 7.63931455598033e-05,
+      "loss": 1.1787,
+      "step": 5351
+    },
+    {
+      "epoch": 0.3251716386171699,
+      "grad_norm": 0.37811279296875,
+      "learning_rate": 7.638501463676655e-05,
+      "loss": 1.0986,
+      "step": 5352
+    },
+    {
+      "epoch": 0.32523239564979645,
+      "grad_norm": 0.15809832513332367,
+      "learning_rate": 7.6376882746582e-05,
+      "loss": 1.0922,
+      "step": 5353
+    },
+    {
+      "epoch": 0.325293152682423,
+      "grad_norm": 0.16970260441303253,
+      "learning_rate": 7.63687498895477e-05,
+      "loss": 1.104,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3253539097150495,
+      "grad_norm": 0.17243817448616028,
+      "learning_rate": 7.636061606596178e-05,
+      "loss": 1.067,
+      "step": 5355
+    },
+    {
+      "epoch": 0.32541466674767605,
+      "grad_norm": 0.15289516746997833,
+      "learning_rate": 7.635248127612238e-05,
+      "loss": 1.0662,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3254754237803026,
+      "grad_norm": 0.17811495065689087,
+      "learning_rate": 7.634434552032768e-05,
+      "loss": 1.0967,
+      "step": 5357
+    },
+    {
+      "epoch": 0.3255361808129291,
+      "grad_norm": 0.14455220103263855,
+      "learning_rate": 7.63362087988759e-05,
+      "loss": 1.0499,
+      "step": 5358
+    },
+    {
+      "epoch": 0.3255969378455556,
+      "grad_norm": 0.1708948016166687,
+      "learning_rate": 7.632807111206531e-05,
+      "loss": 1.1196,
+      "step": 5359
+    },
+    {
+      "epoch": 0.32565769487818214,
+      "grad_norm": 0.16589440405368805,
+      "learning_rate": 7.631993246019418e-05,
+      "loss": 1.0705,
+      "step": 5360
+    },
+    {
+      "epoch": 0.32571845191080867,
+      "grad_norm": 0.12406129390001297,
+      "learning_rate": 7.631179284356085e-05,
+      "loss": 1.08,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3257792089434352,
+      "grad_norm": 0.19620484113693237,
+      "learning_rate": 7.630365226246365e-05,
+      "loss": 1.0555,
+      "step": 5362
+    },
+    {
+      "epoch": 0.32583996597606174,
+      "grad_norm": 0.24526768922805786,
+      "learning_rate": 7.6295510717201e-05,
+      "loss": 1.1506,
+      "step": 5363
+    },
+    {
+      "epoch": 0.3259007230086883,
+      "grad_norm": 0.38127610087394714,
+      "learning_rate": 7.628736820807132e-05,
+      "loss": 1.1142,
+      "step": 5364
+    },
+    {
+      "epoch": 0.32596148004131476,
+      "grad_norm": 0.19456841051578522,
+      "learning_rate": 7.627922473537307e-05,
+      "loss": 1.1398,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3260222370739413,
+      "grad_norm": 0.8136919736862183,
+      "learning_rate": 7.627108029940477e-05,
+      "loss": 1.0773,
+      "step": 5366
+    },
+    {
+      "epoch": 0.3260829941065678,
+      "grad_norm": 0.17009076476097107,
+      "learning_rate": 7.626293490046493e-05,
+      "loss": 1.172,
+      "step": 5367
+    },
+    {
+      "epoch": 0.32614375113919436,
+      "grad_norm": 0.24989072978496552,
+      "learning_rate": 7.625478853885215e-05,
+      "loss": 1.0367,
+      "step": 5368
+    },
+    {
+      "epoch": 0.3262045081718209,
+      "grad_norm": 0.22486184537410736,
+      "learning_rate": 7.6246641214865e-05,
+      "loss": 1.0291,
+      "step": 5369
+    },
+    {
+      "epoch": 0.32626526520444743,
+      "grad_norm": 0.25649169087409973,
+      "learning_rate": 7.623849292880216e-05,
+      "loss": 1.1169,
+      "step": 5370
+    },
+    {
+      "epoch": 0.32632602223707396,
+      "grad_norm": 0.22672656178474426,
+      "learning_rate": 7.623034368096229e-05,
+      "loss": 1.004,
+      "step": 5371
+    },
+    {
+      "epoch": 0.32638677926970044,
+      "grad_norm": 0.13147930800914764,
+      "learning_rate": 7.62221934716441e-05,
+      "loss": 1.0593,
+      "step": 5372
+    },
+    {
+      "epoch": 0.326447536302327,
+      "grad_norm": 0.2682172954082489,
+      "learning_rate": 7.621404230114632e-05,
+      "loss": 1.0627,
+      "step": 5373
+    },
+    {
+      "epoch": 0.3265082933349535,
+      "grad_norm": 0.15371541678905487,
+      "learning_rate": 7.62058901697678e-05,
+      "loss": 1.0906,
+      "step": 5374
+    },
+    {
+      "epoch": 0.32656905036758005,
+      "grad_norm": 6.770089149475098,
+      "learning_rate": 7.619773707780727e-05,
+      "loss": 1.0783,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3266298074002066,
+      "grad_norm": 0.222050741314888,
+      "learning_rate": 7.618958302556363e-05,
+      "loss": 1.0888,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3266905644328331,
+      "grad_norm": 0.18792663514614105,
+      "learning_rate": 7.618142801333576e-05,
+      "loss": 1.1686,
+      "step": 5377
+    },
+    {
+      "epoch": 0.32675132146545965,
+      "grad_norm": 0.47172272205352783,
+      "learning_rate": 7.61732720414226e-05,
+      "loss": 1.2057,
+      "step": 5378
+    },
+    {
+      "epoch": 0.32681207849808613,
+      "grad_norm": 0.3117204010486603,
+      "learning_rate": 7.616511511012308e-05,
+      "loss": 1.0713,
+      "step": 5379
+    },
+    {
+      "epoch": 0.32687283553071267,
+      "grad_norm": 0.1713021695613861,
+      "learning_rate": 7.615695721973621e-05,
+      "loss": 1.1123,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3269335925633392,
+      "grad_norm": 0.13214349746704102,
+      "learning_rate": 7.614879837056103e-05,
+      "loss": 1.0323,
+      "step": 5381
+    },
+    {
+      "epoch": 0.32699434959596574,
+      "grad_norm": 0.17013302445411682,
+      "learning_rate": 7.614063856289658e-05,
+      "loss": 1.0746,
+      "step": 5382
+    },
+    {
+      "epoch": 0.3270551066285923,
+      "grad_norm": 0.1985173225402832,
+      "learning_rate": 7.613247779704197e-05,
+      "loss": 1.0644,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3271158636612188,
+      "grad_norm": 0.21263551712036133,
+      "learning_rate": 7.612431607329635e-05,
+      "loss": 1.0573,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3271766206938453,
+      "grad_norm": 0.6282607316970825,
+      "learning_rate": 7.611615339195886e-05,
+      "loss": 1.0843,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3272373777264718,
+      "grad_norm": 0.33215171098709106,
+      "learning_rate": 7.610798975332871e-05,
+      "loss": 1.1194,
+      "step": 5386
+    },
+    {
+      "epoch": 0.32729813475909836,
+      "grad_norm": 0.2752797603607178,
+      "learning_rate": 7.609982515770516e-05,
+      "loss": 1.1742,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3273588917917249,
+      "grad_norm": 0.21759207546710968,
+      "learning_rate": 7.609165960538748e-05,
+      "loss": 1.1583,
+      "step": 5388
+    },
+    {
+      "epoch": 0.32741964882435143,
+      "grad_norm": 0.30800655484199524,
+      "learning_rate": 7.608349309667498e-05,
+      "loss": 1.2168,
+      "step": 5389
+    },
+    {
+      "epoch": 0.32748040585697796,
+      "grad_norm": 0.17498065531253815,
+      "learning_rate": 7.607532563186696e-05,
+      "loss": 1.1263,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3275411628896045,
+      "grad_norm": 0.15697449445724487,
+      "learning_rate": 7.606715721126288e-05,
+      "loss": 1.086,
+      "step": 5391
+    },
+    {
+      "epoch": 0.327601919922231,
+      "grad_norm": 0.16416527330875397,
+      "learning_rate": 7.60589878351621e-05,
+      "loss": 1.0964,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3276626769548575,
+      "grad_norm": 0.21482136845588684,
+      "learning_rate": 7.605081750386407e-05,
+      "loss": 1.1328,
+      "step": 5393
+    },
+    {
+      "epoch": 0.32772343398748405,
+      "grad_norm": 0.27260974049568176,
+      "learning_rate": 7.604264621766833e-05,
+      "loss": 1.1123,
+      "step": 5394
+    },
+    {
+      "epoch": 0.3277841910201106,
+      "grad_norm": 0.23360304534435272,
+      "learning_rate": 7.603447397687432e-05,
+      "loss": 1.0687,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3278449480527371,
+      "grad_norm": 0.13089804351329803,
+      "learning_rate": 7.602630078178164e-05,
+      "loss": 1.0495,
+      "step": 5396
+    },
+    {
+      "epoch": 0.32790570508536365,
+      "grad_norm": 0.13218490779399872,
+      "learning_rate": 7.601812663268988e-05,
+      "loss": 1.0475,
+      "step": 5397
+    },
+    {
+      "epoch": 0.32796646211799013,
+      "grad_norm": 0.375641793012619,
+      "learning_rate": 7.600995152989865e-05,
+      "loss": 1.0799,
+      "step": 5398
+    },
+    {
+      "epoch": 0.32802721915061667,
+      "grad_norm": 0.18387757241725922,
+      "learning_rate": 7.600177547370763e-05,
+      "loss": 1.1275,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3280879761832432,
+      "grad_norm": 0.1416109949350357,
+      "learning_rate": 7.599359846441651e-05,
+      "loss": 1.0789,
+      "step": 5400
+    },
+    {
+      "epoch": 0.32814873321586974,
+      "grad_norm": 0.20161789655685425,
+      "learning_rate": 7.5985420502325e-05,
+      "loss": 1.0724,
+      "step": 5401
+    },
+    {
+      "epoch": 0.32820949024849627,
+      "grad_norm": 0.13243705034255981,
+      "learning_rate": 7.597724158773289e-05,
+      "loss": 1.0758,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3282702472811228,
+      "grad_norm": 0.24550196528434753,
+      "learning_rate": 7.596906172093995e-05,
+      "loss": 1.0983,
+      "step": 5403
+    },
+    {
+      "epoch": 0.32833100431374934,
+      "grad_norm": 0.2173246145248413,
+      "learning_rate": 7.596088090224604e-05,
+      "loss": 1.1317,
+      "step": 5404
+    },
+    {
+      "epoch": 0.3283917613463758,
+      "grad_norm": 0.23761998116970062,
+      "learning_rate": 7.595269913195104e-05,
+      "loss": 1.1346,
+      "step": 5405
+    },
+    {
+      "epoch": 0.32845251837900236,
+      "grad_norm": 0.42454764246940613,
+      "learning_rate": 7.594451641035482e-05,
+      "loss": 1.0699,
+      "step": 5406
+    },
+    {
+      "epoch": 0.3285132754116289,
+      "grad_norm": 0.159896120429039,
+      "learning_rate": 7.593633273775734e-05,
+      "loss": 1.0873,
+      "step": 5407
+    },
+    {
+      "epoch": 0.3285740324442554,
+      "grad_norm": 0.13794861733913422,
+      "learning_rate": 7.592814811445856e-05,
+      "loss": 1.1247,
+      "step": 5408
+    },
+    {
+      "epoch": 0.32863478947688196,
+      "grad_norm": 0.20439006388187408,
+      "learning_rate": 7.591996254075851e-05,
+      "loss": 1.0699,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3286955465095085,
+      "grad_norm": 0.1942063421010971,
+      "learning_rate": 7.591177601695722e-05,
+      "loss": 1.0832,
+      "step": 5410
+    },
+    {
+      "epoch": 0.328756303542135,
+      "grad_norm": 0.19093358516693115,
+      "learning_rate": 7.59035885433548e-05,
+      "loss": 1.0546,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3288170605747615,
+      "grad_norm": 0.14206360280513763,
+      "learning_rate": 7.58954001202513e-05,
+      "loss": 1.0647,
+      "step": 5412
+    },
+    {
+      "epoch": 0.32887781760738805,
+      "grad_norm": 0.1793767511844635,
+      "learning_rate": 7.588721074794691e-05,
+      "loss": 1.0647,
+      "step": 5413
+    },
+    {
+      "epoch": 0.3289385746400146,
+      "grad_norm": 0.2375302016735077,
+      "learning_rate": 7.58790204267418e-05,
+      "loss": 1.128,
+      "step": 5414
+    },
+    {
+      "epoch": 0.3289993316726411,
+      "grad_norm": 0.17666825652122498,
+      "learning_rate": 7.58708291569362e-05,
+      "loss": 1.1092,
+      "step": 5415
+    },
+    {
+      "epoch": 0.32906008870526765,
+      "grad_norm": 0.27784040570259094,
+      "learning_rate": 7.586263693883036e-05,
+      "loss": 1.0649,
+      "step": 5416
+    },
+    {
+      "epoch": 0.3291208457378942,
+      "grad_norm": 0.18404880166053772,
+      "learning_rate": 7.585444377272456e-05,
+      "loss": 1.1707,
+      "step": 5417
+    },
+    {
+      "epoch": 0.32918160277052066,
+      "grad_norm": 0.42867541313171387,
+      "learning_rate": 7.584624965891913e-05,
+      "loss": 1.1083,
+      "step": 5418
+    },
+    {
+      "epoch": 0.3292423598031472,
+      "grad_norm": 0.26918578147888184,
+      "learning_rate": 7.583805459771443e-05,
+      "loss": 1.1006,
+      "step": 5419
+    },
+    {
+      "epoch": 0.32930311683577373,
+      "grad_norm": 0.12470528483390808,
+      "learning_rate": 7.582985858941083e-05,
+      "loss": 1.0458,
+      "step": 5420
+    },
+    {
+      "epoch": 0.32936387386840027,
+      "grad_norm": 0.21023140847682953,
+      "learning_rate": 7.582166163430877e-05,
+      "loss": 1.0358,
+      "step": 5421
+    },
+    {
+      "epoch": 0.3294246309010268,
+      "grad_norm": 0.1549159735441208,
+      "learning_rate": 7.581346373270873e-05,
+      "loss": 1.1289,
+      "step": 5422
+    },
+    {
+      "epoch": 0.32948538793365334,
+      "grad_norm": 0.1667061597108841,
+      "learning_rate": 7.580526488491116e-05,
+      "loss": 1.0213,
+      "step": 5423
+    },
+    {
+      "epoch": 0.3295461449662798,
+      "grad_norm": 0.49645593762397766,
+      "learning_rate": 7.579706509121663e-05,
+      "loss": 1.3674,
+      "step": 5424
+    },
+    {
+      "epoch": 0.32960690199890635,
+      "grad_norm": 0.16318374872207642,
+      "learning_rate": 7.57888643519257e-05,
+      "loss": 1.0536,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3296676590315329,
+      "grad_norm": 0.15818971395492554,
+      "learning_rate": 7.578066266733897e-05,
+      "loss": 1.0969,
+      "step": 5426
+    },
+    {
+      "epoch": 0.3297284160641594,
+      "grad_norm": 0.2040669322013855,
+      "learning_rate": 7.577246003775704e-05,
+      "loss": 1.0815,
+      "step": 5427
+    },
+    {
+      "epoch": 0.32978917309678596,
+      "grad_norm": 1.1522833108901978,
+      "learning_rate": 7.576425646348062e-05,
+      "loss": 1.043,
+      "step": 5428
+    },
+    {
+      "epoch": 0.3298499301294125,
+      "grad_norm": 0.1402571201324463,
+      "learning_rate": 7.575605194481039e-05,
+      "loss": 1.0392,
+      "step": 5429
+    },
+    {
+      "epoch": 0.32991068716203903,
+      "grad_norm": 0.23267050087451935,
+      "learning_rate": 7.574784648204711e-05,
+      "loss": 1.2631,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3299714441946655,
+      "grad_norm": 0.16812367737293243,
+      "learning_rate": 7.573964007549155e-05,
+      "loss": 1.1362,
+      "step": 5431
+    },
+    {
+      "epoch": 0.33003220122729204,
+      "grad_norm": 5.5949835777282715,
+      "learning_rate": 7.573143272544449e-05,
+      "loss": 1.0358,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3300929582599186,
+      "grad_norm": 0.31999778747558594,
+      "learning_rate": 7.572322443220678e-05,
+      "loss": 1.0605,
+      "step": 5433
+    },
+    {
+      "epoch": 0.3301537152925451,
+      "grad_norm": 0.14096160233020782,
+      "learning_rate": 7.571501519607933e-05,
+      "loss": 1.0424,
+      "step": 5434
+    },
+    {
+      "epoch": 0.33021447232517165,
+      "grad_norm": 0.23999495804309845,
+      "learning_rate": 7.570680501736303e-05,
+      "loss": 1.0938,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3302752293577982,
+      "grad_norm": 0.2573355436325073,
+      "learning_rate": 7.56985938963588e-05,
+      "loss": 1.129,
+      "step": 5436
+    },
+    {
+      "epoch": 0.3303359863904247,
+      "grad_norm": 0.3784105181694031,
+      "learning_rate": 7.569038183336765e-05,
+      "loss": 1.2554,
+      "step": 5437
+    },
+    {
+      "epoch": 0.3303967434230512,
+      "grad_norm": 0.2289574295282364,
+      "learning_rate": 7.56821688286906e-05,
+      "loss": 1.1632,
+      "step": 5438
+    },
+    {
+      "epoch": 0.33045750045567773,
+      "grad_norm": 0.25765928626060486,
+      "learning_rate": 7.567395488262868e-05,
+      "loss": 1.1846,
+      "step": 5439
+    },
+    {
+      "epoch": 0.33051825748830427,
+      "grad_norm": 0.13408100605010986,
+      "learning_rate": 7.566573999548297e-05,
+      "loss": 1.0368,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3305790145209308,
+      "grad_norm": 0.13287340104579926,
+      "learning_rate": 7.565752416755461e-05,
+      "loss": 1.0991,
+      "step": 5441
+    },
+    {
+      "epoch": 0.33063977155355734,
+      "grad_norm": 0.18916484713554382,
+      "learning_rate": 7.564930739914473e-05,
+      "loss": 1.127,
+      "step": 5442
+    },
+    {
+      "epoch": 0.3307005285861839,
+      "grad_norm": 0.23863853514194489,
+      "learning_rate": 7.564108969055455e-05,
+      "loss": 1.1935,
+      "step": 5443
+    },
+    {
+      "epoch": 0.33076128561881035,
+      "grad_norm": 0.2178945392370224,
+      "learning_rate": 7.563287104208526e-05,
+      "loss": 1.1345,
+      "step": 5444
+    },
+    {
+      "epoch": 0.3308220426514369,
+      "grad_norm": 0.3364085257053375,
+      "learning_rate": 7.562465145403813e-05,
+      "loss": 1.1097,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3308827996840634,
+      "grad_norm": 0.19464822113513947,
+      "learning_rate": 7.561643092671444e-05,
+      "loss": 1.0984,
+      "step": 5446
+    },
+    {
+      "epoch": 0.33094355671668996,
+      "grad_norm": 0.20432879030704498,
+      "learning_rate": 7.560820946041554e-05,
+      "loss": 1.0489,
+      "step": 5447
+    },
+    {
+      "epoch": 0.3310043137493165,
+      "grad_norm": 0.3350868225097656,
+      "learning_rate": 7.559998705544277e-05,
+      "loss": 1.0911,
+      "step": 5448
+    },
+    {
+      "epoch": 0.331065070781943,
+      "grad_norm": 0.13863138854503632,
+      "learning_rate": 7.559176371209751e-05,
+      "loss": 1.0284,
+      "step": 5449
+    },
+    {
+      "epoch": 0.33112582781456956,
+      "grad_norm": 0.37578779458999634,
+      "learning_rate": 7.558353943068122e-05,
+      "loss": 1.1807,
+      "step": 5450
+    },
+    {
+      "epoch": 0.33118658484719604,
+      "grad_norm": 0.2311204969882965,
+      "learning_rate": 7.557531421149534e-05,
+      "loss": 1.1274,
+      "step": 5451
+    },
+    {
+      "epoch": 0.3312473418798226,
+      "grad_norm": 0.2685815393924713,
+      "learning_rate": 7.556708805484138e-05,
+      "loss": 1.0295,
+      "step": 5452
+    },
+    {
+      "epoch": 0.3313080989124491,
+      "grad_norm": 0.2088678628206253,
+      "learning_rate": 7.555886096102084e-05,
+      "loss": 1.1626,
+      "step": 5453
+    },
+    {
+      "epoch": 0.33136885594507565,
+      "grad_norm": 5.203686714172363,
+      "learning_rate": 7.555063293033533e-05,
+      "loss": 1.1302,
+      "step": 5454
+    },
+    {
+      "epoch": 0.3314296129777022,
+      "grad_norm": 0.2517106533050537,
+      "learning_rate": 7.554240396308643e-05,
+      "loss": 1.0603,
+      "step": 5455
+    },
+    {
+      "epoch": 0.3314903700103287,
+      "grad_norm": 0.19543345272541046,
+      "learning_rate": 7.553417405957579e-05,
+      "loss": 1.2118,
+      "step": 5456
+    },
+    {
+      "epoch": 0.3315511270429552,
+      "grad_norm": 0.1629648655653,
+      "learning_rate": 7.552594322010504e-05,
+      "loss": 1.1508,
+      "step": 5457
+    },
+    {
+      "epoch": 0.33161188407558173,
+      "grad_norm": 0.21248915791511536,
+      "learning_rate": 7.551771144497592e-05,
+      "loss": 1.0847,
+      "step": 5458
+    },
+    {
+      "epoch": 0.33167264110820827,
+      "grad_norm": 0.20575113594532013,
+      "learning_rate": 7.550947873449016e-05,
+      "loss": 1.1877,
+      "step": 5459
+    },
+    {
+      "epoch": 0.3317333981408348,
+      "grad_norm": 0.14594343304634094,
+      "learning_rate": 7.550124508894951e-05,
+      "loss": 1.0824,
+      "step": 5460
+    },
+    {
+      "epoch": 0.33179415517346134,
+      "grad_norm": 0.19440650939941406,
+      "learning_rate": 7.54930105086558e-05,
+      "loss": 1.0894,
+      "step": 5461
+    },
+    {
+      "epoch": 0.33185491220608787,
+      "grad_norm": 0.16481439769268036,
+      "learning_rate": 7.548477499391087e-05,
+      "loss": 1.1211,
+      "step": 5462
+    },
+    {
+      "epoch": 0.3319156692387144,
+      "grad_norm": 0.24972113966941833,
+      "learning_rate": 7.547653854501657e-05,
+      "loss": 1.0991,
+      "step": 5463
+    },
+    {
+      "epoch": 0.3319764262713409,
+      "grad_norm": 0.18423761427402496,
+      "learning_rate": 7.546830116227484e-05,
+      "loss": 1.1164,
+      "step": 5464
+    },
+    {
+      "epoch": 0.3320371833039674,
+      "grad_norm": 0.2262442708015442,
+      "learning_rate": 7.54600628459876e-05,
+      "loss": 1.1068,
+      "step": 5465
+    },
+    {
+      "epoch": 0.33209794033659396,
+      "grad_norm": 0.17784802615642548,
+      "learning_rate": 7.545182359645684e-05,
+      "loss": 1.113,
+      "step": 5466
+    },
+    {
+      "epoch": 0.3321586973692205,
+      "grad_norm": 0.2207585722208023,
+      "learning_rate": 7.544358341398457e-05,
+      "loss": 1.178,
+      "step": 5467
+    },
+    {
+      "epoch": 0.332219454401847,
+      "grad_norm": 0.19200880825519562,
+      "learning_rate": 7.543534229887281e-05,
+      "loss": 1.0855,
+      "step": 5468
+    },
+    {
+      "epoch": 0.33228021143447356,
+      "grad_norm": 0.1824488341808319,
+      "learning_rate": 7.54271002514237e-05,
+      "loss": 1.3626,
+      "step": 5469
+    },
+    {
+      "epoch": 0.33234096846710004,
+      "grad_norm": 0.22011467814445496,
+      "learning_rate": 7.541885727193928e-05,
+      "loss": 1.1189,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3324017254997266,
+      "grad_norm": 0.33231163024902344,
+      "learning_rate": 7.541061336072175e-05,
+      "loss": 1.0944,
+      "step": 5471
+    },
+    {
+      "epoch": 0.3324624825323531,
+      "grad_norm": 0.18799728155136108,
+      "learning_rate": 7.540236851807327e-05,
+      "loss": 1.0891,
+      "step": 5472
+    },
+    {
+      "epoch": 0.33252323956497964,
+      "grad_norm": 0.5314054489135742,
+      "learning_rate": 7.539412274429605e-05,
+      "loss": 1.0746,
+      "step": 5473
+    },
+    {
+      "epoch": 0.3325839965976062,
+      "grad_norm": 0.13672508299350739,
+      "learning_rate": 7.538587603969238e-05,
+      "loss": 1.0563,
+      "step": 5474
+    },
+    {
+      "epoch": 0.3326447536302327,
+      "grad_norm": 0.30675119161605835,
+      "learning_rate": 7.53776284045645e-05,
+      "loss": 1.0091,
+      "step": 5475
+    },
+    {
+      "epoch": 0.33270551066285925,
+      "grad_norm": 0.3291470408439636,
+      "learning_rate": 7.536937983921474e-05,
+      "loss": 1.1303,
+      "step": 5476
+    },
+    {
+      "epoch": 0.33276626769548573,
+      "grad_norm": 0.1538793444633484,
+      "learning_rate": 7.536113034394547e-05,
+      "loss": 1.0574,
+      "step": 5477
+    },
+    {
+      "epoch": 0.33282702472811226,
+      "grad_norm": 0.2535040080547333,
+      "learning_rate": 7.535287991905906e-05,
+      "loss": 1.0811,
+      "step": 5478
+    },
+    {
+      "epoch": 0.3328877817607388,
+      "grad_norm": 0.17257514595985413,
+      "learning_rate": 7.534462856485793e-05,
+      "loss": 1.1203,
+      "step": 5479
+    },
+    {
+      "epoch": 0.33294853879336533,
+      "grad_norm": 0.22970615327358246,
+      "learning_rate": 7.533637628164455e-05,
+      "loss": 1.1531,
+      "step": 5480
+    },
+    {
+      "epoch": 0.33300929582599187,
+      "grad_norm": 0.4710615873336792,
+      "learning_rate": 7.532812306972139e-05,
+      "loss": 1.0932,
+      "step": 5481
+    },
+    {
+      "epoch": 0.3330700528586184,
+      "grad_norm": 0.17988619208335876,
+      "learning_rate": 7.531986892939101e-05,
+      "loss": 1.0801,
+      "step": 5482
+    },
+    {
+      "epoch": 0.33313080989124494,
+      "grad_norm": 0.19155313074588776,
+      "learning_rate": 7.531161386095591e-05,
+      "loss": 1.1033,
+      "step": 5483
+    },
+    {
+      "epoch": 0.3331915669238714,
+      "grad_norm": 0.17571823298931122,
+      "learning_rate": 7.530335786471871e-05,
+      "loss": 1.2097,
+      "step": 5484
+    },
+    {
+      "epoch": 0.33325232395649795,
+      "grad_norm": 3.874110221862793,
+      "learning_rate": 7.529510094098207e-05,
+      "loss": 1.0823,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3333130809891245,
+      "grad_norm": 1.1709591150283813,
+      "learning_rate": 7.528684309004859e-05,
+      "loss": 1.2051,
+      "step": 5486
+    },
+    {
+      "epoch": 0.333373838021751,
+      "grad_norm": 0.22638088464736938,
+      "learning_rate": 7.527858431222098e-05,
+      "loss": 1.1015,
+      "step": 5487
+    },
+    {
+      "epoch": 0.33343459505437756,
+      "grad_norm": 0.15556208789348602,
+      "learning_rate": 7.527032460780201e-05,
+      "loss": 1.085,
+      "step": 5488
+    },
+    {
+      "epoch": 0.3334953520870041,
+      "grad_norm": 0.18091131746768951,
+      "learning_rate": 7.526206397709437e-05,
+      "loss": 1.1757,
+      "step": 5489
+    },
+    {
+      "epoch": 0.3335561091196306,
+      "grad_norm": 0.1496092528104782,
+      "learning_rate": 7.525380242040092e-05,
+      "loss": 1.0332,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3336168661522571,
+      "grad_norm": 0.15536093711853027,
+      "learning_rate": 7.524553993802443e-05,
+      "loss": 1.1971,
+      "step": 5491
+    },
+    {
+      "epoch": 0.33367762318488364,
+      "grad_norm": 0.25183531641960144,
+      "learning_rate": 7.523727653026783e-05,
+      "loss": 1.1268,
+      "step": 5492
+    },
+    {
+      "epoch": 0.3337383802175102,
+      "grad_norm": 0.15332481265068054,
+      "learning_rate": 7.522901219743396e-05,
+      "loss": 1.1097,
+      "step": 5493
+    },
+    {
+      "epoch": 0.3337991372501367,
+      "grad_norm": 0.16587772965431213,
+      "learning_rate": 7.522074693982577e-05,
+      "loss": 1.1238,
+      "step": 5494
+    },
+    {
+      "epoch": 0.33385989428276325,
+      "grad_norm": 0.21817606687545776,
+      "learning_rate": 7.521248075774623e-05,
+      "loss": 1.0806,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3339206513153898,
+      "grad_norm": 0.371785044670105,
+      "learning_rate": 7.520421365149834e-05,
+      "loss": 1.1514,
+      "step": 5496
+    },
+    {
+      "epoch": 0.33398140834801626,
+      "grad_norm": 0.13117551803588867,
+      "learning_rate": 7.519594562138511e-05,
+      "loss": 1.0224,
+      "step": 5497
+    },
+    {
+      "epoch": 0.3340421653806428,
+      "grad_norm": 0.13245713710784912,
+      "learning_rate": 7.518767666770965e-05,
+      "loss": 1.0559,
+      "step": 5498
+    },
+    {
+      "epoch": 0.33410292241326933,
+      "grad_norm": 0.1396120935678482,
+      "learning_rate": 7.5179406790775e-05,
+      "loss": 1.0532,
+      "step": 5499
+    },
+    {
+      "epoch": 0.33416367944589587,
+      "grad_norm": 0.15593959391117096,
+      "learning_rate": 7.517113599088434e-05,
+      "loss": 1.0197,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3342244364785224,
+      "grad_norm": 0.11808733642101288,
+      "learning_rate": 7.516286426834084e-05,
+      "loss": 1.0396,
+      "step": 5501
+    },
+    {
+      "epoch": 0.33428519351114894,
+      "grad_norm": 0.21038290858268738,
+      "learning_rate": 7.515459162344767e-05,
+      "loss": 1.1596,
+      "step": 5502
+    },
+    {
+      "epoch": 0.3343459505437754,
+      "grad_norm": 0.1895390748977661,
+      "learning_rate": 7.51463180565081e-05,
+      "loss": 1.1275,
+      "step": 5503
+    },
+    {
+      "epoch": 0.33440670757640195,
+      "grad_norm": 0.2504735291004181,
+      "learning_rate": 7.513804356782536e-05,
+      "loss": 1.0489,
+      "step": 5504
+    },
+    {
+      "epoch": 0.3344674646090285,
+      "grad_norm": 2.3111376762390137,
+      "learning_rate": 7.512976815770279e-05,
+      "loss": 1.0479,
+      "step": 5505
+    },
+    {
+      "epoch": 0.334528221641655,
+      "grad_norm": 12.76346206665039,
+      "learning_rate": 7.512149182644371e-05,
+      "loss": 1.0611,
+      "step": 5506
+    },
+    {
+      "epoch": 0.33458897867428156,
+      "grad_norm": 0.2044251561164856,
+      "learning_rate": 7.511321457435145e-05,
+      "loss": 1.0564,
+      "step": 5507
+    },
+    {
+      "epoch": 0.3346497357069081,
+      "grad_norm": 0.22237356007099152,
+      "learning_rate": 7.51049364017295e-05,
+      "loss": 1.1788,
+      "step": 5508
+    },
+    {
+      "epoch": 0.3347104927395346,
+      "grad_norm": 0.28337711095809937,
+      "learning_rate": 7.509665730888125e-05,
+      "loss": 1.1912,
+      "step": 5509
+    },
+    {
+      "epoch": 0.3347712497721611,
+      "grad_norm": 0.20663613080978394,
+      "learning_rate": 7.508837729611016e-05,
+      "loss": 1.0964,
+      "step": 5510
+    },
+    {
+      "epoch": 0.33483200680478764,
+      "grad_norm": 0.23966573178768158,
+      "learning_rate": 7.508009636371976e-05,
+      "loss": 1.0412,
+      "step": 5511
+    },
+    {
+      "epoch": 0.3348927638374142,
+      "grad_norm": 0.25408032536506653,
+      "learning_rate": 7.507181451201357e-05,
+      "loss": 1.1001,
+      "step": 5512
+    },
+    {
+      "epoch": 0.3349535208700407,
+      "grad_norm": 1.6409603357315063,
+      "learning_rate": 7.506353174129518e-05,
+      "loss": 1.1754,
+      "step": 5513
+    },
+    {
+      "epoch": 0.33501427790266725,
+      "grad_norm": 0.13990652561187744,
+      "learning_rate": 7.505524805186821e-05,
+      "loss": 1.0497,
+      "step": 5514
+    },
+    {
+      "epoch": 0.3350750349352938,
+      "grad_norm": 0.20568786561489105,
+      "learning_rate": 7.504696344403626e-05,
+      "loss": 1.0818,
+      "step": 5515
+    },
+    {
+      "epoch": 0.33513579196792026,
+      "grad_norm": 0.17626319825649261,
+      "learning_rate": 7.503867791810302e-05,
+      "loss": 1.1946,
+      "step": 5516
+    },
+    {
+      "epoch": 0.3351965490005468,
+      "grad_norm": 0.478924959897995,
+      "learning_rate": 7.503039147437218e-05,
+      "loss": 1.1956,
+      "step": 5517
+    },
+    {
+      "epoch": 0.33525730603317333,
+      "grad_norm": 0.30180832743644714,
+      "learning_rate": 7.502210411314755e-05,
+      "loss": 1.0471,
+      "step": 5518
+    },
+    {
+      "epoch": 0.33531806306579987,
+      "grad_norm": 0.2921055257320404,
+      "learning_rate": 7.501381583473284e-05,
+      "loss": 1.0678,
+      "step": 5519
+    },
+    {
+      "epoch": 0.3353788200984264,
+      "grad_norm": 0.2026301473379135,
+      "learning_rate": 7.500552663943186e-05,
+      "loss": 1.1336,
+      "step": 5520
+    },
+    {
+      "epoch": 0.33543957713105294,
+      "grad_norm": 0.1825183480978012,
+      "learning_rate": 7.499723652754847e-05,
+      "loss": 1.1394,
+      "step": 5521
+    },
+    {
+      "epoch": 0.33550033416367947,
+      "grad_norm": 0.1588670313358307,
+      "learning_rate": 7.498894549938656e-05,
+      "loss": 1.0613,
+      "step": 5522
+    },
+    {
+      "epoch": 0.33556109119630595,
+      "grad_norm": 0.17803405225276947,
+      "learning_rate": 7.498065355525002e-05,
+      "loss": 1.0756,
+      "step": 5523
+    },
+    {
+      "epoch": 0.3356218482289325,
+      "grad_norm": 0.15975968539714813,
+      "learning_rate": 7.497236069544278e-05,
+      "loss": 1.1229,
+      "step": 5524
+    },
+    {
+      "epoch": 0.335682605261559,
+      "grad_norm": 0.21353621780872345,
+      "learning_rate": 7.496406692026884e-05,
+      "loss": 1.0469,
+      "step": 5525
+    },
+    {
+      "epoch": 0.33574336229418555,
+      "grad_norm": 0.2091209590435028,
+      "learning_rate": 7.49557722300322e-05,
+      "loss": 1.1089,
+      "step": 5526
+    },
+    {
+      "epoch": 0.3358041193268121,
+      "grad_norm": 0.35379528999328613,
+      "learning_rate": 7.494747662503691e-05,
+      "loss": 1.0785,
+      "step": 5527
+    },
+    {
+      "epoch": 0.3358648763594386,
+      "grad_norm": 0.16948559880256653,
+      "learning_rate": 7.493918010558704e-05,
+      "loss": 1.0616,
+      "step": 5528
+    },
+    {
+      "epoch": 0.3359256333920651,
+      "grad_norm": 0.13938705623149872,
+      "learning_rate": 7.49308826719867e-05,
+      "loss": 1.0721,
+      "step": 5529
+    },
+    {
+      "epoch": 0.33598639042469164,
+      "grad_norm": 0.1933734118938446,
+      "learning_rate": 7.492258432454002e-05,
+      "loss": 1.1133,
+      "step": 5530
+    },
+    {
+      "epoch": 0.3360471474573182,
+      "grad_norm": 0.15812847018241882,
+      "learning_rate": 7.491428506355122e-05,
+      "loss": 1.0783,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3361079044899447,
+      "grad_norm": 0.14268743991851807,
+      "learning_rate": 7.490598488932449e-05,
+      "loss": 1.0934,
+      "step": 5532
+    },
+    {
+      "epoch": 0.33616866152257124,
+      "grad_norm": 0.21588250994682312,
+      "learning_rate": 7.489768380216405e-05,
+      "loss": 1.2146,
+      "step": 5533
+    },
+    {
+      "epoch": 0.3362294185551978,
+      "grad_norm": 0.11810000240802765,
+      "learning_rate": 7.48893818023742e-05,
+      "loss": 1.0105,
+      "step": 5534
+    },
+    {
+      "epoch": 0.3362901755878243,
+      "grad_norm": 0.13542866706848145,
+      "learning_rate": 7.488107889025926e-05,
+      "loss": 1.0928,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3363509326204508,
+      "grad_norm": 0.3417040705680847,
+      "learning_rate": 7.487277506612356e-05,
+      "loss": 1.1799,
+      "step": 5536
+    },
+    {
+      "epoch": 0.33641168965307733,
+      "grad_norm": 0.25541195273399353,
+      "learning_rate": 7.486447033027148e-05,
+      "loss": 1.1153,
+      "step": 5537
+    },
+    {
+      "epoch": 0.33647244668570386,
+      "grad_norm": 0.27999597787857056,
+      "learning_rate": 7.485616468300743e-05,
+      "loss": 1.1655,
+      "step": 5538
+    },
+    {
+      "epoch": 0.3365332037183304,
+      "grad_norm": 0.3260159492492676,
+      "learning_rate": 7.484785812463587e-05,
+      "loss": 1.1765,
+      "step": 5539
+    },
+    {
+      "epoch": 0.33659396075095693,
+      "grad_norm": 0.1697273850440979,
+      "learning_rate": 7.483955065546126e-05,
+      "loss": 1.1354,
+      "step": 5540
+    },
+    {
+      "epoch": 0.33665471778358347,
+      "grad_norm": 0.2634475827217102,
+      "learning_rate": 7.483124227578811e-05,
+      "loss": 1.2169,
+      "step": 5541
+    },
+    {
+      "epoch": 0.33671547481621,
+      "grad_norm": 0.18889199197292328,
+      "learning_rate": 7.482293298592098e-05,
+      "loss": 1.0853,
+      "step": 5542
+    },
+    {
+      "epoch": 0.3367762318488365,
+      "grad_norm": 0.12441056221723557,
+      "learning_rate": 7.481462278616444e-05,
+      "loss": 1.0495,
+      "step": 5543
+    },
+    {
+      "epoch": 0.336836988881463,
+      "grad_norm": 0.18259049952030182,
+      "learning_rate": 7.480631167682312e-05,
+      "loss": 1.2154,
+      "step": 5544
+    },
+    {
+      "epoch": 0.33689774591408955,
+      "grad_norm": 0.17367684841156006,
+      "learning_rate": 7.479799965820164e-05,
+      "loss": 1.0803,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3369585029467161,
+      "grad_norm": 0.20125249028205872,
+      "learning_rate": 7.478968673060469e-05,
+      "loss": 1.1825,
+      "step": 5546
+    },
+    {
+      "epoch": 0.3370192599793426,
+      "grad_norm": 0.1303924173116684,
+      "learning_rate": 7.478137289433699e-05,
+      "loss": 1.0362,
+      "step": 5547
+    },
+    {
+      "epoch": 0.33708001701196916,
+      "grad_norm": 0.16618074476718903,
+      "learning_rate": 7.477305814970327e-05,
+      "loss": 1.0664,
+      "step": 5548
+    },
+    {
+      "epoch": 0.33714077404459564,
+      "grad_norm": 0.2883400022983551,
+      "learning_rate": 7.476474249700831e-05,
+      "loss": 1.1497,
+      "step": 5549
+    },
+    {
+      "epoch": 0.33720153107722217,
+      "grad_norm": 0.2781234383583069,
+      "learning_rate": 7.475642593655694e-05,
+      "loss": 1.1587,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3372622881098487,
+      "grad_norm": 0.2012518048286438,
+      "learning_rate": 7.474810846865396e-05,
+      "loss": 1.1081,
+      "step": 5551
+    },
+    {
+      "epoch": 0.33732304514247524,
+      "grad_norm": 0.1505262851715088,
+      "learning_rate": 7.473979009360433e-05,
+      "loss": 1.0729,
+      "step": 5552
+    },
+    {
+      "epoch": 0.3373838021751018,
+      "grad_norm": 0.17856548726558685,
+      "learning_rate": 7.473147081171287e-05,
+      "loss": 1.1449,
+      "step": 5553
+    },
+    {
+      "epoch": 0.3374445592077283,
+      "grad_norm": 0.21609683334827423,
+      "learning_rate": 7.472315062328459e-05,
+      "loss": 1.2114,
+      "step": 5554
+    },
+    {
+      "epoch": 0.33750531624035485,
+      "grad_norm": 0.14103642106056213,
+      "learning_rate": 7.471482952862444e-05,
+      "loss": 1.0796,
+      "step": 5555
+    },
+    {
+      "epoch": 0.3375660732729813,
+      "grad_norm": 0.13603602349758148,
+      "learning_rate": 7.470650752803743e-05,
+      "loss": 1.0575,
+      "step": 5556
+    },
+    {
+      "epoch": 0.33762683030560786,
+      "grad_norm": 0.30538976192474365,
+      "learning_rate": 7.46981846218286e-05,
+      "loss": 1.2169,
+      "step": 5557
+    },
+    {
+      "epoch": 0.3376875873382344,
+      "grad_norm": 1.272532343864441,
+      "learning_rate": 7.468986081030307e-05,
+      "loss": 1.0851,
+      "step": 5558
+    },
+    {
+      "epoch": 0.33774834437086093,
+      "grad_norm": 0.33729514479637146,
+      "learning_rate": 7.468153609376591e-05,
+      "loss": 1.0894,
+      "step": 5559
+    },
+    {
+      "epoch": 0.33780910140348747,
+      "grad_norm": 0.2312454879283905,
+      "learning_rate": 7.467321047252226e-05,
+      "loss": 1.1492,
+      "step": 5560
+    },
+    {
+      "epoch": 0.337869858436114,
+      "grad_norm": 0.16546347737312317,
+      "learning_rate": 7.466488394687731e-05,
+      "loss": 1.0618,
+      "step": 5561
+    },
+    {
+      "epoch": 0.3379306154687405,
+      "grad_norm": 0.5239700078964233,
+      "learning_rate": 7.465655651713627e-05,
+      "loss": 1.1053,
+      "step": 5562
+    },
+    {
+      "epoch": 0.337991372501367,
+      "grad_norm": 0.22329890727996826,
+      "learning_rate": 7.464822818360441e-05,
+      "loss": 1.0917,
+      "step": 5563
+    },
+    {
+      "epoch": 0.33805212953399355,
+      "grad_norm": 0.27938318252563477,
+      "learning_rate": 7.463989894658696e-05,
+      "loss": 1.2364,
+      "step": 5564
+    },
+    {
+      "epoch": 0.3381128865666201,
+      "grad_norm": 0.15419381856918335,
+      "learning_rate": 7.463156880638926e-05,
+      "loss": 1.0473,
+      "step": 5565
+    },
+    {
+      "epoch": 0.3381736435992466,
+      "grad_norm": 1.407528042793274,
+      "learning_rate": 7.462323776331664e-05,
+      "loss": 1.1546,
+      "step": 5566
+    },
+    {
+      "epoch": 0.33823440063187316,
+      "grad_norm": 0.31853002309799194,
+      "learning_rate": 7.461490581767448e-05,
+      "loss": 1.09,
+      "step": 5567
+    },
+    {
+      "epoch": 0.3382951576644997,
+      "grad_norm": 0.1421099603176117,
+      "learning_rate": 7.460657296976819e-05,
+      "loss": 1.057,
+      "step": 5568
+    },
+    {
+      "epoch": 0.33835591469712617,
+      "grad_norm": 0.1373155266046524,
+      "learning_rate": 7.459823921990321e-05,
+      "loss": 1.0674,
+      "step": 5569
+    },
+    {
+      "epoch": 0.3384166717297527,
+      "grad_norm": 0.15903204679489136,
+      "learning_rate": 7.458990456838501e-05,
+      "loss": 1.0691,
+      "step": 5570
+    },
+    {
+      "epoch": 0.33847742876237924,
+      "grad_norm": 0.14871279895305634,
+      "learning_rate": 7.458156901551912e-05,
+      "loss": 1.0473,
+      "step": 5571
+    },
+    {
+      "epoch": 0.3385381857950058,
+      "grad_norm": 0.33478689193725586,
+      "learning_rate": 7.457323256161106e-05,
+      "loss": 1.1369,
+      "step": 5572
+    },
+    {
+      "epoch": 0.3385989428276323,
+      "grad_norm": 0.1887255311012268,
+      "learning_rate": 7.456489520696642e-05,
+      "loss": 1.0479,
+      "step": 5573
+    },
+    {
+      "epoch": 0.33865969986025884,
+      "grad_norm": 0.2735128700733185,
+      "learning_rate": 7.455655695189077e-05,
+      "loss": 1.104,
+      "step": 5574
+    },
+    {
+      "epoch": 0.3387204568928853,
+      "grad_norm": 0.23236620426177979,
+      "learning_rate": 7.454821779668981e-05,
+      "loss": 1.0628,
+      "step": 5575
+    },
+    {
+      "epoch": 0.33878121392551186,
+      "grad_norm": 0.16397589445114136,
+      "learning_rate": 7.453987774166917e-05,
+      "loss": 1.1087,
+      "step": 5576
+    },
+    {
+      "epoch": 0.3388419709581384,
+      "grad_norm": 0.18313263356685638,
+      "learning_rate": 7.453153678713455e-05,
+      "loss": 1.1311,
+      "step": 5577
+    },
+    {
+      "epoch": 0.33890272799076493,
+      "grad_norm": 0.8418211340904236,
+      "learning_rate": 7.452319493339173e-05,
+      "loss": 1.2568,
+      "step": 5578
+    },
+    {
+      "epoch": 0.33896348502339146,
+      "grad_norm": 0.23657409846782684,
+      "learning_rate": 7.451485218074646e-05,
+      "loss": 1.083,
+      "step": 5579
+    },
+    {
+      "epoch": 0.339024242056018,
+      "grad_norm": 2.726797580718994,
+      "learning_rate": 7.450650852950452e-05,
+      "loss": 1.0701,
+      "step": 5580
+    },
+    {
+      "epoch": 0.33908499908864453,
+      "grad_norm": 0.23161150515079498,
+      "learning_rate": 7.44981639799718e-05,
+      "loss": 1.1538,
+      "step": 5581
+    },
+    {
+      "epoch": 0.339145756121271,
+      "grad_norm": 1.5758064985275269,
+      "learning_rate": 7.448981853245414e-05,
+      "loss": 1.309,
+      "step": 5582
+    },
+    {
+      "epoch": 0.33920651315389755,
+      "grad_norm": 0.3868856728076935,
+      "learning_rate": 7.448147218725744e-05,
+      "loss": 1.169,
+      "step": 5583
+    },
+    {
+      "epoch": 0.3392672701865241,
+      "grad_norm": 0.166421577334404,
+      "learning_rate": 7.447312494468764e-05,
+      "loss": 1.0814,
+      "step": 5584
+    },
+    {
+      "epoch": 0.3393280272191506,
+      "grad_norm": 1.2591252326965332,
+      "learning_rate": 7.446477680505073e-05,
+      "loss": 1.1229,
+      "step": 5585
+    },
+    {
+      "epoch": 0.33938878425177715,
+      "grad_norm": 0.18639180064201355,
+      "learning_rate": 7.445642776865269e-05,
+      "loss": 1.0656,
+      "step": 5586
+    },
+    {
+      "epoch": 0.3394495412844037,
+      "grad_norm": 0.22079963982105255,
+      "learning_rate": 7.444807783579955e-05,
+      "loss": 1.0502,
+      "step": 5587
+    },
+    {
+      "epoch": 0.3395102983170302,
+      "grad_norm": 0.1429029405117035,
+      "learning_rate": 7.44397270067974e-05,
+      "loss": 1.1084,
+      "step": 5588
+    },
+    {
+      "epoch": 0.3395710553496567,
+      "grad_norm": 0.2187453806400299,
+      "learning_rate": 7.443137528195234e-05,
+      "loss": 1.0849,
+      "step": 5589
+    },
+    {
+      "epoch": 0.33963181238228324,
+      "grad_norm": 0.2235606610774994,
+      "learning_rate": 7.442302266157048e-05,
+      "loss": 1.0794,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3396925694149098,
+      "grad_norm": 0.4305715560913086,
+      "learning_rate": 7.441466914595803e-05,
+      "loss": 1.096,
+      "step": 5591
+    },
+    {
+      "epoch": 0.3397533264475363,
+      "grad_norm": 0.19101032614707947,
+      "learning_rate": 7.440631473542115e-05,
+      "loss": 1.203,
+      "step": 5592
+    },
+    {
+      "epoch": 0.33981408348016284,
+      "grad_norm": 0.17606954276561737,
+      "learning_rate": 7.439795943026608e-05,
+      "loss": 1.051,
+      "step": 5593
+    },
+    {
+      "epoch": 0.3398748405127894,
+      "grad_norm": 0.4550893306732178,
+      "learning_rate": 7.43896032307991e-05,
+      "loss": 1.1026,
+      "step": 5594
+    },
+    {
+      "epoch": 0.33993559754541586,
+      "grad_norm": 0.12969236075878143,
+      "learning_rate": 7.438124613732648e-05,
+      "loss": 1.0219,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3399963545780424,
+      "grad_norm": 0.2768893837928772,
+      "learning_rate": 7.43728881501546e-05,
+      "loss": 1.1329,
+      "step": 5596
+    },
+    {
+      "epoch": 0.3400571116106689,
+      "grad_norm": 0.24179022014141083,
+      "learning_rate": 7.436452926958978e-05,
+      "loss": 1.1184,
+      "step": 5597
+    },
+    {
+      "epoch": 0.34011786864329546,
+      "grad_norm": 0.8925986886024475,
+      "learning_rate": 7.435616949593842e-05,
+      "loss": 1.0757,
+      "step": 5598
+    },
+    {
+      "epoch": 0.340178625675922,
+      "grad_norm": 0.13775081932544708,
+      "learning_rate": 7.434780882950698e-05,
+      "loss": 1.0459,
+      "step": 5599
+    },
+    {
+      "epoch": 0.34023938270854853,
+      "grad_norm": 0.28351345658302307,
+      "learning_rate": 7.433944727060187e-05,
+      "loss": 1.1884,
+      "step": 5600
+    },
+    {
+      "epoch": 0.34030013974117507,
+      "grad_norm": 0.7399532794952393,
+      "learning_rate": 7.433108481952965e-05,
+      "loss": 1.2486,
+      "step": 5601
+    },
+    {
+      "epoch": 0.34036089677380155,
+      "grad_norm": 0.16441327333450317,
+      "learning_rate": 7.432272147659678e-05,
+      "loss": 1.0957,
+      "step": 5602
+    },
+    {
+      "epoch": 0.3404216538064281,
+      "grad_norm": 0.163296177983284,
+      "learning_rate": 7.431435724210988e-05,
+      "loss": 1.1066,
+      "step": 5603
+    },
+    {
+      "epoch": 0.3404824108390546,
+      "grad_norm": 0.15724974870681763,
+      "learning_rate": 7.43059921163755e-05,
+      "loss": 1.0814,
+      "step": 5604
+    },
+    {
+      "epoch": 0.34054316787168115,
+      "grad_norm": 0.18417984247207642,
+      "learning_rate": 7.42976260997003e-05,
+      "loss": 1.0665,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3406039249043077,
+      "grad_norm": 0.2278042435646057,
+      "learning_rate": 7.428925919239089e-05,
+      "loss": 1.0832,
+      "step": 5606
+    },
+    {
+      "epoch": 0.3406646819369342,
+      "grad_norm": 0.25080230832099915,
+      "learning_rate": 7.4280891394754e-05,
+      "loss": 1.1882,
+      "step": 5607
+    },
+    {
+      "epoch": 0.3407254389695607,
+      "grad_norm": 0.12096163630485535,
+      "learning_rate": 7.427252270709635e-05,
+      "loss": 1.0625,
+      "step": 5608
+    },
+    {
+      "epoch": 0.34078619600218724,
+      "grad_norm": 0.13884446024894714,
+      "learning_rate": 7.426415312972469e-05,
+      "loss": 1.0871,
+      "step": 5609
+    },
+    {
+      "epoch": 0.34084695303481377,
+      "grad_norm": 0.19591939449310303,
+      "learning_rate": 7.425578266294579e-05,
+      "loss": 1.1022,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3409077100674403,
+      "grad_norm": 0.23972727358341217,
+      "learning_rate": 7.42474113070665e-05,
+      "loss": 1.1176,
+      "step": 5611
+    },
+    {
+      "epoch": 0.34096846710006684,
+      "grad_norm": 0.19567696750164032,
+      "learning_rate": 7.423903906239367e-05,
+      "loss": 1.0735,
+      "step": 5612
+    },
+    {
+      "epoch": 0.3410292241326934,
+      "grad_norm": 0.18039634823799133,
+      "learning_rate": 7.423066592923417e-05,
+      "loss": 1.0964,
+      "step": 5613
+    },
+    {
+      "epoch": 0.3410899811653199,
+      "grad_norm": 0.24264463782310486,
+      "learning_rate": 7.422229190789492e-05,
+      "loss": 1.0596,
+      "step": 5614
+    },
+    {
+      "epoch": 0.3411507381979464,
+      "grad_norm": 0.2057594507932663,
+      "learning_rate": 7.421391699868288e-05,
+      "loss": 1.0339,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3412114952305729,
+      "grad_norm": 0.1615353524684906,
+      "learning_rate": 7.420554120190504e-05,
+      "loss": 1.0968,
+      "step": 5616
+    },
+    {
+      "epoch": 0.34127225226319946,
+      "grad_norm": 0.22651933133602142,
+      "learning_rate": 7.41971645178684e-05,
+      "loss": 1.0786,
+      "step": 5617
+    },
+    {
+      "epoch": 0.341333009295826,
+      "grad_norm": 0.19685249030590057,
+      "learning_rate": 7.418878694688002e-05,
+      "loss": 1.1282,
+      "step": 5618
+    },
+    {
+      "epoch": 0.34139376632845253,
+      "grad_norm": 0.1910761445760727,
+      "learning_rate": 7.418040848924698e-05,
+      "loss": 1.1786,
+      "step": 5619
+    },
+    {
+      "epoch": 0.34145452336107907,
+      "grad_norm": 0.20940546691417694,
+      "learning_rate": 7.41720291452764e-05,
+      "loss": 1.1485,
+      "step": 5620
+    },
+    {
+      "epoch": 0.34151528039370554,
+      "grad_norm": 0.13719259202480316,
+      "learning_rate": 7.416364891527541e-05,
+      "loss": 1.0778,
+      "step": 5621
+    },
+    {
+      "epoch": 0.3415760374263321,
+      "grad_norm": 0.13285192847251892,
+      "learning_rate": 7.415526779955121e-05,
+      "loss": 1.0748,
+      "step": 5622
+    },
+    {
+      "epoch": 0.3416367944589586,
+      "grad_norm": 3.7748587131500244,
+      "learning_rate": 7.414688579841098e-05,
+      "loss": 1.1041,
+      "step": 5623
+    },
+    {
+      "epoch": 0.34169755149158515,
+      "grad_norm": 0.14596666395664215,
+      "learning_rate": 7.4138502912162e-05,
+      "loss": 1.1143,
+      "step": 5624
+    },
+    {
+      "epoch": 0.3417583085242117,
+      "grad_norm": 0.15469378232955933,
+      "learning_rate": 7.413011914111152e-05,
+      "loss": 1.0945,
+      "step": 5625
+    },
+    {
+      "epoch": 0.3418190655568382,
+      "grad_norm": 0.808948814868927,
+      "learning_rate": 7.412173448556687e-05,
+      "loss": 1.2254,
+      "step": 5626
+    },
+    {
+      "epoch": 0.34187982258946475,
+      "grad_norm": 0.5553398132324219,
+      "learning_rate": 7.411334894583538e-05,
+      "loss": 1.138,
+      "step": 5627
+    },
+    {
+      "epoch": 0.34194057962209123,
+      "grad_norm": 0.146551713347435,
+      "learning_rate": 7.41049625222244e-05,
+      "loss": 1.0603,
+      "step": 5628
+    },
+    {
+      "epoch": 0.34200133665471777,
+      "grad_norm": 0.4100880026817322,
+      "learning_rate": 7.409657521504137e-05,
+      "loss": 1.1189,
+      "step": 5629
+    },
+    {
+      "epoch": 0.3420620936873443,
+      "grad_norm": 0.16061285138130188,
+      "learning_rate": 7.408818702459374e-05,
+      "loss": 1.0913,
+      "step": 5630
+    },
+    {
+      "epoch": 0.34212285071997084,
+      "grad_norm": 0.27554699778556824,
+      "learning_rate": 7.407979795118892e-05,
+      "loss": 1.2068,
+      "step": 5631
+    },
+    {
+      "epoch": 0.3421836077525974,
+      "grad_norm": 0.14766065776348114,
+      "learning_rate": 7.407140799513449e-05,
+      "loss": 1.0972,
+      "step": 5632
+    },
+    {
+      "epoch": 0.3422443647852239,
+      "grad_norm": 0.19677849113941193,
+      "learning_rate": 7.406301715673793e-05,
+      "loss": 1.0682,
+      "step": 5633
+    },
+    {
+      "epoch": 0.3423051218178504,
+      "grad_norm": 0.26291149854660034,
+      "learning_rate": 7.405462543630681e-05,
+      "loss": 1.159,
+      "step": 5634
+    },
+    {
+      "epoch": 0.3423658788504769,
+      "grad_norm": 0.18334637582302094,
+      "learning_rate": 7.404623283414876e-05,
+      "loss": 1.0782,
+      "step": 5635
+    },
+    {
+      "epoch": 0.34242663588310346,
+      "grad_norm": 0.16423004865646362,
+      "learning_rate": 7.40378393505714e-05,
+      "loss": 1.0463,
+      "step": 5636
+    },
+    {
+      "epoch": 0.34248739291573,
+      "grad_norm": 0.2510235607624054,
+      "learning_rate": 7.402944498588239e-05,
+      "loss": 1.1497,
+      "step": 5637
+    },
+    {
+      "epoch": 0.34254814994835653,
+      "grad_norm": 0.24117280542850494,
+      "learning_rate": 7.402104974038943e-05,
+      "loss": 1.173,
+      "step": 5638
+    },
+    {
+      "epoch": 0.34260890698098306,
+      "grad_norm": 0.16989333927631378,
+      "learning_rate": 7.401265361440026e-05,
+      "loss": 1.1391,
+      "step": 5639
+    },
+    {
+      "epoch": 0.3426696640136096,
+      "grad_norm": 0.2275887280702591,
+      "learning_rate": 7.40042566082226e-05,
+      "loss": 1.0677,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3427304210462361,
+      "grad_norm": 0.17539876699447632,
+      "learning_rate": 7.39958587221643e-05,
+      "loss": 1.1916,
+      "step": 5641
+    },
+    {
+      "epoch": 0.3427911780788626,
+      "grad_norm": 3.434662342071533,
+      "learning_rate": 7.398745995653316e-05,
+      "loss": 1.0947,
+      "step": 5642
+    },
+    {
+      "epoch": 0.34285193511148915,
+      "grad_norm": 0.3420221507549286,
+      "learning_rate": 7.397906031163702e-05,
+      "loss": 1.1176,
+      "step": 5643
+    },
+    {
+      "epoch": 0.3429126921441157,
+      "grad_norm": 4.487304210662842,
+      "learning_rate": 7.397065978778379e-05,
+      "loss": 1.151,
+      "step": 5644
+    },
+    {
+      "epoch": 0.3429734491767422,
+      "grad_norm": 0.17696045339107513,
+      "learning_rate": 7.396225838528141e-05,
+      "loss": 1.0733,
+      "step": 5645
+    },
+    {
+      "epoch": 0.34303420620936875,
+      "grad_norm": 0.22626429796218872,
+      "learning_rate": 7.395385610443782e-05,
+      "loss": 1.1691,
+      "step": 5646
+    },
+    {
+      "epoch": 0.3430949632419953,
+      "grad_norm": 0.1350768655538559,
+      "learning_rate": 7.394545294556098e-05,
+      "loss": 1.0289,
+      "step": 5647
+    },
+    {
+      "epoch": 0.34315572027462177,
+      "grad_norm": 0.17733575403690338,
+      "learning_rate": 7.393704890895895e-05,
+      "loss": 1.0973,
+      "step": 5648
+    },
+    {
+      "epoch": 0.3432164773072483,
+      "grad_norm": 0.2509138584136963,
+      "learning_rate": 7.392864399493976e-05,
+      "loss": 1.2118,
+      "step": 5649
+    },
+    {
+      "epoch": 0.34327723433987484,
+      "grad_norm": 0.3420547544956207,
+      "learning_rate": 7.39202382038115e-05,
+      "loss": 1.2192,
+      "step": 5650
+    },
+    {
+      "epoch": 0.34333799137250137,
+      "grad_norm": 0.1549650877714157,
+      "learning_rate": 7.39118315358823e-05,
+      "loss": 1.1308,
+      "step": 5651
+    },
+    {
+      "epoch": 0.3433987484051279,
+      "grad_norm": 0.2076931744813919,
+      "learning_rate": 7.390342399146027e-05,
+      "loss": 1.1029,
+      "step": 5652
+    },
+    {
+      "epoch": 0.34345950543775444,
+      "grad_norm": 0.14477044343948364,
+      "learning_rate": 7.389501557085362e-05,
+      "loss": 1.0735,
+      "step": 5653
+    },
+    {
+      "epoch": 0.3435202624703809,
+      "grad_norm": 0.2162570059299469,
+      "learning_rate": 7.388660627437055e-05,
+      "loss": 1.1981,
+      "step": 5654
+    },
+    {
+      "epoch": 0.34358101950300746,
+      "grad_norm": 0.1991337090730667,
+      "learning_rate": 7.387819610231932e-05,
+      "loss": 1.1393,
+      "step": 5655
+    },
+    {
+      "epoch": 0.343641776535634,
+      "grad_norm": 0.34177350997924805,
+      "learning_rate": 7.386978505500822e-05,
+      "loss": 1.0783,
+      "step": 5656
+    },
+    {
+      "epoch": 0.3437025335682605,
+      "grad_norm": 0.16313883662223816,
+      "learning_rate": 7.386137313274548e-05,
+      "loss": 1.1165,
+      "step": 5657
+    },
+    {
+      "epoch": 0.34376329060088706,
+      "grad_norm": 0.21096296608448029,
+      "learning_rate": 7.385296033583953e-05,
+      "loss": 1.1561,
+      "step": 5658
+    },
+    {
+      "epoch": 0.3438240476335136,
+      "grad_norm": 0.33268654346466064,
+      "learning_rate": 7.38445466645987e-05,
+      "loss": 1.067,
+      "step": 5659
+    },
+    {
+      "epoch": 0.34388480466614013,
+      "grad_norm": 0.14179377257823944,
+      "learning_rate": 7.383613211933141e-05,
+      "loss": 1.0873,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3439455616987666,
+      "grad_norm": 0.19833813607692719,
+      "learning_rate": 7.382771670034608e-05,
+      "loss": 1.1134,
+      "step": 5661
+    },
+    {
+      "epoch": 0.34400631873139315,
+      "grad_norm": 0.4791829288005829,
+      "learning_rate": 7.381930040795119e-05,
+      "loss": 1.2737,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3440670757640197,
+      "grad_norm": 0.4864122271537781,
+      "learning_rate": 7.381088324245525e-05,
+      "loss": 1.0829,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3441278327966462,
+      "grad_norm": 0.16897287964820862,
+      "learning_rate": 7.380246520416678e-05,
+      "loss": 1.0861,
+      "step": 5664
+    },
+    {
+      "epoch": 0.34418858982927275,
+      "grad_norm": 0.2143501490354538,
+      "learning_rate": 7.379404629339433e-05,
+      "loss": 1.0881,
+      "step": 5665
+    },
+    {
+      "epoch": 0.3442493468618993,
+      "grad_norm": 0.3365273177623749,
+      "learning_rate": 7.378562651044653e-05,
+      "loss": 1.1237,
+      "step": 5666
+    },
+    {
+      "epoch": 0.34431010389452577,
+      "grad_norm": 0.8202298879623413,
+      "learning_rate": 7.377720585563199e-05,
+      "loss": 1.1335,
+      "step": 5667
+    },
+    {
+      "epoch": 0.3443708609271523,
+      "grad_norm": 0.16889657080173492,
+      "learning_rate": 7.376878432925938e-05,
+      "loss": 1.1591,
+      "step": 5668
+    },
+    {
+      "epoch": 0.34443161795977884,
+      "grad_norm": 0.15688326954841614,
+      "learning_rate": 7.376036193163738e-05,
+      "loss": 1.0354,
+      "step": 5669
+    },
+    {
+      "epoch": 0.34449237499240537,
+      "grad_norm": 0.21872282028198242,
+      "learning_rate": 7.375193866307471e-05,
+      "loss": 1.1424,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3445531320250319,
+      "grad_norm": 0.20044030249118805,
+      "learning_rate": 7.374351452388016e-05,
+      "loss": 1.1177,
+      "step": 5671
+    },
+    {
+      "epoch": 0.34461388905765844,
+      "grad_norm": 0.47033408284187317,
+      "learning_rate": 7.373508951436247e-05,
+      "loss": 1.2517,
+      "step": 5672
+    },
+    {
+      "epoch": 0.344674646090285,
+      "grad_norm": 1.573832631111145,
+      "learning_rate": 7.37266636348305e-05,
+      "loss": 1.0547,
+      "step": 5673
+    },
+    {
+      "epoch": 0.34473540312291145,
+      "grad_norm": 0.1649778038263321,
+      "learning_rate": 7.371823688559308e-05,
+      "loss": 1.0617,
+      "step": 5674
+    },
+    {
+      "epoch": 0.344796160155538,
+      "grad_norm": 0.30305325984954834,
+      "learning_rate": 7.37098092669591e-05,
+      "loss": 1.1152,
+      "step": 5675
+    },
+    {
+      "epoch": 0.3448569171881645,
+      "grad_norm": 0.19876450300216675,
+      "learning_rate": 7.37013807792375e-05,
+      "loss": 1.0866,
+      "step": 5676
+    },
+    {
+      "epoch": 0.34491767422079106,
+      "grad_norm": 0.27401667833328247,
+      "learning_rate": 7.36929514227372e-05,
+      "loss": 1.0587,
+      "step": 5677
+    },
+    {
+      "epoch": 0.3449784312534176,
+      "grad_norm": 0.1567569524049759,
+      "learning_rate": 7.368452119776716e-05,
+      "loss": 1.0622,
+      "step": 5678
+    },
+    {
+      "epoch": 0.34503918828604413,
+      "grad_norm": 0.202568918466568,
+      "learning_rate": 7.367609010463646e-05,
+      "loss": 1.0361,
+      "step": 5679
+    },
+    {
+      "epoch": 0.3450999453186706,
+      "grad_norm": 0.17404919862747192,
+      "learning_rate": 7.366765814365406e-05,
+      "loss": 1.0743,
+      "step": 5680
+    },
+    {
+      "epoch": 0.34516070235129714,
+      "grad_norm": 1.4974126815795898,
+      "learning_rate": 7.36592253151291e-05,
+      "loss": 1.1646,
+      "step": 5681
+    },
+    {
+      "epoch": 0.3452214593839237,
+      "grad_norm": 0.2991856634616852,
+      "learning_rate": 7.365079161937066e-05,
+      "loss": 1.1555,
+      "step": 5682
+    },
+    {
+      "epoch": 0.3452822164165502,
+      "grad_norm": 0.25850579142570496,
+      "learning_rate": 7.364235705668787e-05,
+      "loss": 1.0941,
+      "step": 5683
+    },
+    {
+      "epoch": 0.34534297344917675,
+      "grad_norm": 0.20604000985622406,
+      "learning_rate": 7.363392162738992e-05,
+      "loss": 1.1276,
+      "step": 5684
+    },
+    {
+      "epoch": 0.3454037304818033,
+      "grad_norm": 0.15421925485134125,
+      "learning_rate": 7.362548533178599e-05,
+      "loss": 1.0567,
+      "step": 5685
+    },
+    {
+      "epoch": 0.3454644875144298,
+      "grad_norm": 0.2226591855287552,
+      "learning_rate": 7.361704817018535e-05,
+      "loss": 1.218,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3455252445470563,
+      "grad_norm": 0.2017289251089096,
+      "learning_rate": 7.360861014289724e-05,
+      "loss": 1.1461,
+      "step": 5687
+    },
+    {
+      "epoch": 0.34558600157968283,
+      "grad_norm": 0.1717287003993988,
+      "learning_rate": 7.360017125023096e-05,
+      "loss": 1.161,
+      "step": 5688
+    },
+    {
+      "epoch": 0.34564675861230937,
+      "grad_norm": 0.34899288415908813,
+      "learning_rate": 7.359173149249584e-05,
+      "loss": 1.2619,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3457075156449359,
+      "grad_norm": 0.15625520050525665,
+      "learning_rate": 7.358329087000124e-05,
+      "loss": 1.0627,
+      "step": 5690
+    },
+    {
+      "epoch": 0.34576827267756244,
+      "grad_norm": 0.13026641309261322,
+      "learning_rate": 7.357484938305655e-05,
+      "loss": 1.0959,
+      "step": 5691
+    },
+    {
+      "epoch": 0.345829029710189,
+      "grad_norm": 0.6139512658119202,
+      "learning_rate": 7.356640703197119e-05,
+      "loss": 1.1515,
+      "step": 5692
+    },
+    {
+      "epoch": 0.3458897867428155,
+      "grad_norm": 0.16386525332927704,
+      "learning_rate": 7.355796381705465e-05,
+      "loss": 1.1346,
+      "step": 5693
+    },
+    {
+      "epoch": 0.345950543775442,
+      "grad_norm": 0.21870790421962738,
+      "learning_rate": 7.354951973861639e-05,
+      "loss": 1.1361,
+      "step": 5694
+    },
+    {
+      "epoch": 0.3460113008080685,
+      "grad_norm": 0.17906992137432098,
+      "learning_rate": 7.354107479696591e-05,
+      "loss": 1.1116,
+      "step": 5695
+    },
+    {
+      "epoch": 0.34607205784069506,
+      "grad_norm": 0.16407795250415802,
+      "learning_rate": 7.35326289924128e-05,
+      "loss": 1.0573,
+      "step": 5696
+    },
+    {
+      "epoch": 0.3461328148733216,
+      "grad_norm": 0.3035754859447479,
+      "learning_rate": 7.352418232526663e-05,
+      "loss": 1.1312,
+      "step": 5697
+    },
+    {
+      "epoch": 0.3461935719059481,
+      "grad_norm": 0.12715089321136475,
+      "learning_rate": 7.3515734795837e-05,
+      "loss": 1.0728,
+      "step": 5698
+    },
+    {
+      "epoch": 0.34625432893857466,
+      "grad_norm": 0.14146849513053894,
+      "learning_rate": 7.350728640443358e-05,
+      "loss": 1.0633,
+      "step": 5699
+    },
+    {
+      "epoch": 0.34631508597120114,
+      "grad_norm": 0.2877689301967621,
+      "learning_rate": 7.3498837151366e-05,
+      "loss": 1.0632,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3463758430038277,
+      "grad_norm": 0.2097448706626892,
+      "learning_rate": 7.349038703694401e-05,
+      "loss": 1.0886,
+      "step": 5701
+    },
+    {
+      "epoch": 0.3464366000364542,
+      "grad_norm": 0.21265172958374023,
+      "learning_rate": 7.348193606147738e-05,
+      "loss": 1.1493,
+      "step": 5702
+    },
+    {
+      "epoch": 0.34649735706908075,
+      "grad_norm": 0.12911838293075562,
+      "learning_rate": 7.347348422527581e-05,
+      "loss": 1.0571,
+      "step": 5703
+    },
+    {
+      "epoch": 0.3465581141017073,
+      "grad_norm": 0.13353310525417328,
+      "learning_rate": 7.346503152864916e-05,
+      "loss": 1.0911,
+      "step": 5704
+    },
+    {
+      "epoch": 0.3466188711343338,
+      "grad_norm": 0.14378947019577026,
+      "learning_rate": 7.345657797190723e-05,
+      "loss": 1.0687,
+      "step": 5705
+    },
+    {
+      "epoch": 0.34667962816696035,
+      "grad_norm": 0.14005614817142487,
+      "learning_rate": 7.344812355535989e-05,
+      "loss": 1.108,
+      "step": 5706
+    },
+    {
+      "epoch": 0.34674038519958683,
+      "grad_norm": 1.9512763023376465,
+      "learning_rate": 7.343966827931707e-05,
+      "loss": 1.0834,
+      "step": 5707
+    },
+    {
+      "epoch": 0.34680114223221337,
+      "grad_norm": 0.29365137219429016,
+      "learning_rate": 7.343121214408867e-05,
+      "loss": 1.1683,
+      "step": 5708
+    },
+    {
+      "epoch": 0.3468618992648399,
+      "grad_norm": 0.6108081936836243,
+      "learning_rate": 7.342275514998466e-05,
+      "loss": 1.0346,
+      "step": 5709
+    },
+    {
+      "epoch": 0.34692265629746644,
+      "grad_norm": 0.2702333331108093,
+      "learning_rate": 7.341429729731503e-05,
+      "loss": 1.1872,
+      "step": 5710
+    },
+    {
+      "epoch": 0.34698341333009297,
+      "grad_norm": 0.2966519296169281,
+      "learning_rate": 7.34058385863898e-05,
+      "loss": 1.2099,
+      "step": 5711
+    },
+    {
+      "epoch": 0.3470441703627195,
+      "grad_norm": 1.0424479246139526,
+      "learning_rate": 7.339737901751905e-05,
+      "loss": 1.251,
+      "step": 5712
+    },
+    {
+      "epoch": 0.347104927395346,
+      "grad_norm": 0.5929242372512817,
+      "learning_rate": 7.338891859101285e-05,
+      "loss": 1.2657,
+      "step": 5713
+    },
+    {
+      "epoch": 0.3471656844279725,
+      "grad_norm": 0.17844748497009277,
+      "learning_rate": 7.338045730718129e-05,
+      "loss": 1.132,
+      "step": 5714
+    },
+    {
+      "epoch": 0.34722644146059906,
+      "grad_norm": 0.31952664256095886,
+      "learning_rate": 7.337199516633458e-05,
+      "loss": 1.1326,
+      "step": 5715
+    },
+    {
+      "epoch": 0.3472871984932256,
+      "grad_norm": 0.24436403810977936,
+      "learning_rate": 7.336353216878285e-05,
+      "loss": 1.0437,
+      "step": 5716
+    },
+    {
+      "epoch": 0.3473479555258521,
+      "grad_norm": 0.2095416635274887,
+      "learning_rate": 7.335506831483634e-05,
+      "loss": 1.2193,
+      "step": 5717
+    },
+    {
+      "epoch": 0.34740871255847866,
+      "grad_norm": 0.19626320898532867,
+      "learning_rate": 7.334660360480527e-05,
+      "loss": 1.0921,
+      "step": 5718
+    },
+    {
+      "epoch": 0.3474694695911052,
+      "grad_norm": 0.17072322964668274,
+      "learning_rate": 7.333813803899996e-05,
+      "loss": 1.1532,
+      "step": 5719
+    },
+    {
+      "epoch": 0.3475302266237317,
+      "grad_norm": 0.18080373108386993,
+      "learning_rate": 7.332967161773068e-05,
+      "loss": 1.1549,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3475909836563582,
+      "grad_norm": 0.19815124571323395,
+      "learning_rate": 7.332120434130778e-05,
+      "loss": 1.1661,
+      "step": 5721
+    },
+    {
+      "epoch": 0.34765174068898475,
+      "grad_norm": 1.6336404085159302,
+      "learning_rate": 7.33127362100416e-05,
+      "loss": 1.1206,
+      "step": 5722
+    },
+    {
+      "epoch": 0.3477124977216113,
+      "grad_norm": 0.16710756719112396,
+      "learning_rate": 7.33042672242426e-05,
+      "loss": 1.128,
+      "step": 5723
+    },
+    {
+      "epoch": 0.3477732547542378,
+      "grad_norm": 0.14641173183918,
+      "learning_rate": 7.329579738422116e-05,
+      "loss": 1.0815,
+      "step": 5724
+    },
+    {
+      "epoch": 0.34783401178686435,
+      "grad_norm": 0.29195988178253174,
+      "learning_rate": 7.328732669028777e-05,
+      "loss": 1.1457,
+      "step": 5725
+    },
+    {
+      "epoch": 0.34789476881949083,
+      "grad_norm": 0.13369852304458618,
+      "learning_rate": 7.327885514275294e-05,
+      "loss": 1.0279,
+      "step": 5726
+    },
+    {
+      "epoch": 0.34795552585211736,
+      "grad_norm": 0.15933088958263397,
+      "learning_rate": 7.327038274192714e-05,
+      "loss": 1.0546,
+      "step": 5727
+    },
+    {
+      "epoch": 0.3480162828847439,
+      "grad_norm": 0.14478226006031036,
+      "learning_rate": 7.326190948812097e-05,
+      "loss": 1.0877,
+      "step": 5728
+    },
+    {
+      "epoch": 0.34807703991737043,
+      "grad_norm": 0.2454705834388733,
+      "learning_rate": 7.325343538164503e-05,
+      "loss": 1.0718,
+      "step": 5729
+    },
+    {
+      "epoch": 0.34813779694999697,
+      "grad_norm": 0.1678120344877243,
+      "learning_rate": 7.32449604228099e-05,
+      "loss": 1.0125,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3481985539826235,
+      "grad_norm": 0.18025526404380798,
+      "learning_rate": 7.323648461192627e-05,
+      "loss": 1.1315,
+      "step": 5731
+    },
+    {
+      "epoch": 0.34825931101525004,
+      "grad_norm": 0.2066357582807541,
+      "learning_rate": 7.32280079493048e-05,
+      "loss": 1.1235,
+      "step": 5732
+    },
+    {
+      "epoch": 0.3483200680478765,
+      "grad_norm": 0.1464719921350479,
+      "learning_rate": 7.321953043525622e-05,
+      "loss": 1.0697,
+      "step": 5733
+    },
+    {
+      "epoch": 0.34838082508050305,
+      "grad_norm": 0.16893261671066284,
+      "learning_rate": 7.321105207009124e-05,
+      "loss": 1.1194,
+      "step": 5734
+    },
+    {
+      "epoch": 0.3484415821131296,
+      "grad_norm": 3.2623672485351562,
+      "learning_rate": 7.320257285412066e-05,
+      "loss": 1.1092,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3485023391457561,
+      "grad_norm": 0.19582438468933105,
+      "learning_rate": 7.319409278765532e-05,
+      "loss": 1.2521,
+      "step": 5736
+    },
+    {
+      "epoch": 0.34856309617838266,
+      "grad_norm": 0.5753293037414551,
+      "learning_rate": 7.318561187100598e-05,
+      "loss": 1.1917,
+      "step": 5737
+    },
+    {
+      "epoch": 0.3486238532110092,
+      "grad_norm": 0.17827434837818146,
+      "learning_rate": 7.317713010448359e-05,
+      "loss": 1.0871,
+      "step": 5738
+    },
+    {
+      "epoch": 0.3486846102436357,
+      "grad_norm": 0.1499662697315216,
+      "learning_rate": 7.316864748839901e-05,
+      "loss": 1.0344,
+      "step": 5739
+    },
+    {
+      "epoch": 0.3487453672762622,
+      "grad_norm": 0.15069326758384705,
+      "learning_rate": 7.316016402306315e-05,
+      "loss": 1.0608,
+      "step": 5740
+    },
+    {
+      "epoch": 0.34880612430888874,
+      "grad_norm": 0.2449057549238205,
+      "learning_rate": 7.315167970878705e-05,
+      "loss": 1.0996,
+      "step": 5741
+    },
+    {
+      "epoch": 0.3488668813415153,
+      "grad_norm": 0.7468262910842896,
+      "learning_rate": 7.314319454588162e-05,
+      "loss": 1.2121,
+      "step": 5742
+    },
+    {
+      "epoch": 0.3489276383741418,
+      "grad_norm": 0.6195400953292847,
+      "learning_rate": 7.313470853465794e-05,
+      "loss": 1.1891,
+      "step": 5743
+    },
+    {
+      "epoch": 0.34898839540676835,
+      "grad_norm": 0.157234787940979,
+      "learning_rate": 7.312622167542702e-05,
+      "loss": 1.1093,
+      "step": 5744
+    },
+    {
+      "epoch": 0.3490491524393949,
+      "grad_norm": 3.5099685192108154,
+      "learning_rate": 7.311773396849999e-05,
+      "loss": 1.0526,
+      "step": 5745
+    },
+    {
+      "epoch": 0.34910990947202136,
+      "grad_norm": 0.22053498029708862,
+      "learning_rate": 7.310924541418796e-05,
+      "loss": 1.1293,
+      "step": 5746
+    },
+    {
+      "epoch": 0.3491706665046479,
+      "grad_norm": 0.8692840933799744,
+      "learning_rate": 7.310075601280206e-05,
+      "loss": 1.3386,
+      "step": 5747
+    },
+    {
+      "epoch": 0.34923142353727443,
+      "grad_norm": 0.23964044451713562,
+      "learning_rate": 7.309226576465349e-05,
+      "loss": 1.2406,
+      "step": 5748
+    },
+    {
+      "epoch": 0.34929218056990097,
+      "grad_norm": 0.6238002181053162,
+      "learning_rate": 7.308377467005347e-05,
+      "loss": 1.2017,
+      "step": 5749
+    },
+    {
+      "epoch": 0.3493529376025275,
+      "grad_norm": 0.24637416005134583,
+      "learning_rate": 7.307528272931319e-05,
+      "loss": 1.0706,
+      "step": 5750
+    },
+    {
+      "epoch": 0.34941369463515404,
+      "grad_norm": 0.21685929596424103,
+      "learning_rate": 7.306678994274398e-05,
+      "loss": 1.0615,
+      "step": 5751
+    },
+    {
+      "epoch": 0.3494744516677806,
+      "grad_norm": 0.1708822250366211,
+      "learning_rate": 7.305829631065714e-05,
+      "loss": 1.1175,
+      "step": 5752
+    },
+    {
+      "epoch": 0.34953520870040705,
+      "grad_norm": 0.23879919946193695,
+      "learning_rate": 7.304980183336397e-05,
+      "loss": 1.1185,
+      "step": 5753
+    },
+    {
+      "epoch": 0.3495959657330336,
+      "grad_norm": 0.18171828985214233,
+      "learning_rate": 7.304130651117588e-05,
+      "loss": 1.1051,
+      "step": 5754
+    },
+    {
+      "epoch": 0.3496567227656601,
+      "grad_norm": 0.2343205064535141,
+      "learning_rate": 7.303281034440421e-05,
+      "loss": 1.102,
+      "step": 5755
+    },
+    {
+      "epoch": 0.34971747979828666,
+      "grad_norm": 0.36213231086730957,
+      "learning_rate": 7.302431333336048e-05,
+      "loss": 1.0919,
+      "step": 5756
+    },
+    {
+      "epoch": 0.3497782368309132,
+      "grad_norm": 0.24571682512760162,
+      "learning_rate": 7.301581547835606e-05,
+      "loss": 1.189,
+      "step": 5757
+    },
+    {
+      "epoch": 0.3498389938635397,
+      "grad_norm": 0.1978072226047516,
+      "learning_rate": 7.300731677970246e-05,
+      "loss": 1.1187,
+      "step": 5758
+    },
+    {
+      "epoch": 0.3498997508961662,
+      "grad_norm": 0.53170245885849,
+      "learning_rate": 7.299881723771124e-05,
+      "loss": 1.208,
+      "step": 5759
+    },
+    {
+      "epoch": 0.34996050792879274,
+      "grad_norm": 0.2298247069120407,
+      "learning_rate": 7.299031685269392e-05,
+      "loss": 1.0478,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3500212649614193,
+      "grad_norm": 0.6664319634437561,
+      "learning_rate": 7.298181562496209e-05,
+      "loss": 1.159,
+      "step": 5761
+    },
+    {
+      "epoch": 0.3500820219940458,
+      "grad_norm": 0.18363624811172485,
+      "learning_rate": 7.297331355482737e-05,
+      "loss": 1.0679,
+      "step": 5762
+    },
+    {
+      "epoch": 0.35014277902667235,
+      "grad_norm": 0.4725208878517151,
+      "learning_rate": 7.296481064260141e-05,
+      "loss": 1.0559,
+      "step": 5763
+    },
+    {
+      "epoch": 0.3502035360592989,
+      "grad_norm": 0.39487212896347046,
+      "learning_rate": 7.295630688859587e-05,
+      "loss": 1.0848,
+      "step": 5764
+    },
+    {
+      "epoch": 0.3502642930919254,
+      "grad_norm": 0.33195677399635315,
+      "learning_rate": 7.294780229312246e-05,
+      "loss": 1.2254,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3503250501245519,
+      "grad_norm": 0.16192245483398438,
+      "learning_rate": 7.293929685649292e-05,
+      "loss": 1.0606,
+      "step": 5766
+    },
+    {
+      "epoch": 0.35038580715717843,
+      "grad_norm": 0.18568256497383118,
+      "learning_rate": 7.293079057901902e-05,
+      "loss": 1.1535,
+      "step": 5767
+    },
+    {
+      "epoch": 0.35044656418980497,
+      "grad_norm": 0.17248280346393585,
+      "learning_rate": 7.292228346101255e-05,
+      "loss": 1.0904,
+      "step": 5768
+    },
+    {
+      "epoch": 0.3505073212224315,
+      "grad_norm": 0.29796722531318665,
+      "learning_rate": 7.291377550278537e-05,
+      "loss": 1.128,
+      "step": 5769
+    },
+    {
+      "epoch": 0.35056807825505804,
+      "grad_norm": 0.32810768485069275,
+      "learning_rate": 7.290526670464933e-05,
+      "loss": 1.2998,
+      "step": 5770
+    },
+    {
+      "epoch": 0.35062883528768457,
+      "grad_norm": 0.9751858115196228,
+      "learning_rate": 7.289675706691628e-05,
+      "loss": 1.1057,
+      "step": 5771
+    },
+    {
+      "epoch": 0.35068959232031105,
+      "grad_norm": 0.321365624666214,
+      "learning_rate": 7.288824658989818e-05,
+      "loss": 1.0544,
+      "step": 5772
+    },
+    {
+      "epoch": 0.3507503493529376,
+      "grad_norm": 1.1922931671142578,
+      "learning_rate": 7.287973527390699e-05,
+      "loss": 1.1017,
+      "step": 5773
+    },
+    {
+      "epoch": 0.3508111063855641,
+      "grad_norm": 0.14850562810897827,
+      "learning_rate": 7.287122311925467e-05,
+      "loss": 1.1196,
+      "step": 5774
+    },
+    {
+      "epoch": 0.35087186341819065,
+      "grad_norm": 0.16037853062152863,
+      "learning_rate": 7.286271012625326e-05,
+      "loss": 1.0693,
+      "step": 5775
+    },
+    {
+      "epoch": 0.3509326204508172,
+      "grad_norm": 0.18525104224681854,
+      "learning_rate": 7.285419629521477e-05,
+      "loss": 1.0864,
+      "step": 5776
+    },
+    {
+      "epoch": 0.3509933774834437,
+      "grad_norm": 0.41641688346862793,
+      "learning_rate": 7.284568162645131e-05,
+      "loss": 1.0468,
+      "step": 5777
+    },
+    {
+      "epoch": 0.35105413451607026,
+      "grad_norm": 0.30591610074043274,
+      "learning_rate": 7.283716612027497e-05,
+      "loss": 1.2554,
+      "step": 5778
+    },
+    {
+      "epoch": 0.35111489154869674,
+      "grad_norm": 0.13217495381832123,
+      "learning_rate": 7.282864977699789e-05,
+      "loss": 1.0358,
+      "step": 5779
+    },
+    {
+      "epoch": 0.3511756485813233,
+      "grad_norm": 15.487896919250488,
+      "learning_rate": 7.282013259693225e-05,
+      "loss": 1.0538,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3512364056139498,
+      "grad_norm": 0.15854030847549438,
+      "learning_rate": 7.281161458039022e-05,
+      "loss": 1.0876,
+      "step": 5781
+    },
+    {
+      "epoch": 0.35129716264657634,
+      "grad_norm": 0.1775718480348587,
+      "learning_rate": 7.280309572768407e-05,
+      "loss": 1.1245,
+      "step": 5782
+    },
+    {
+      "epoch": 0.3513579196792029,
+      "grad_norm": 0.653881311416626,
+      "learning_rate": 7.279457603912603e-05,
+      "loss": 1.2809,
+      "step": 5783
+    },
+    {
+      "epoch": 0.3514186767118294,
+      "grad_norm": 0.23063676059246063,
+      "learning_rate": 7.27860555150284e-05,
+      "loss": 1.0354,
+      "step": 5784
+    },
+    {
+      "epoch": 0.3514794337444559,
+      "grad_norm": 0.17329175770282745,
+      "learning_rate": 7.277753415570349e-05,
+      "loss": 1.1186,
+      "step": 5785
+    },
+    {
+      "epoch": 0.35154019077708243,
+      "grad_norm": 0.15079350769519806,
+      "learning_rate": 7.276901196146367e-05,
+      "loss": 1.1657,
+      "step": 5786
+    },
+    {
+      "epoch": 0.35160094780970896,
+      "grad_norm": 0.23822999000549316,
+      "learning_rate": 7.276048893262132e-05,
+      "loss": 1.0697,
+      "step": 5787
+    },
+    {
+      "epoch": 0.3516617048423355,
+      "grad_norm": 0.20611199736595154,
+      "learning_rate": 7.275196506948885e-05,
+      "loss": 1.0403,
+      "step": 5788
+    },
+    {
+      "epoch": 0.35172246187496203,
+      "grad_norm": 0.25846919417381287,
+      "learning_rate": 7.274344037237868e-05,
+      "loss": 1.259,
+      "step": 5789
+    },
+    {
+      "epoch": 0.35178321890758857,
+      "grad_norm": 0.23814713954925537,
+      "learning_rate": 7.273491484160334e-05,
+      "loss": 1.075,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3518439759402151,
+      "grad_norm": 0.5330560207366943,
+      "learning_rate": 7.272638847747528e-05,
+      "loss": 1.055,
+      "step": 5791
+    },
+    {
+      "epoch": 0.3519047329728416,
+      "grad_norm": 0.4811760187149048,
+      "learning_rate": 7.271786128030706e-05,
+      "loss": 1.2015,
+      "step": 5792
+    },
+    {
+      "epoch": 0.3519654900054681,
+      "grad_norm": 0.13954052329063416,
+      "learning_rate": 7.270933325041125e-05,
+      "loss": 1.0401,
+      "step": 5793
+    },
+    {
+      "epoch": 0.35202624703809465,
+      "grad_norm": 0.13437391817569733,
+      "learning_rate": 7.270080438810043e-05,
+      "loss": 1.0669,
+      "step": 5794
+    },
+    {
+      "epoch": 0.3520870040707212,
+      "grad_norm": 0.12584757804870605,
+      "learning_rate": 7.269227469368725e-05,
+      "loss": 1.0786,
+      "step": 5795
+    },
+    {
+      "epoch": 0.3521477611033477,
+      "grad_norm": 0.16621801257133484,
+      "learning_rate": 7.268374416748434e-05,
+      "loss": 1.12,
+      "step": 5796
+    },
+    {
+      "epoch": 0.35220851813597426,
+      "grad_norm": 0.15626905858516693,
+      "learning_rate": 7.267521280980442e-05,
+      "loss": 1.1097,
+      "step": 5797
+    },
+    {
+      "epoch": 0.3522692751686008,
+      "grad_norm": 0.43506526947021484,
+      "learning_rate": 7.266668062096018e-05,
+      "loss": 1.3063,
+      "step": 5798
+    },
+    {
+      "epoch": 0.3523300322012273,
+      "grad_norm": 0.32517826557159424,
+      "learning_rate": 7.265814760126439e-05,
+      "loss": 1.1496,
+      "step": 5799
+    },
+    {
+      "epoch": 0.3523907892338538,
+      "grad_norm": 0.26677462458610535,
+      "learning_rate": 7.26496137510298e-05,
+      "loss": 1.0512,
+      "step": 5800
+    },
+    {
+      "epoch": 0.35245154626648034,
+      "grad_norm": 0.2610955536365509,
+      "learning_rate": 7.264107907056928e-05,
+      "loss": 1.1183,
+      "step": 5801
+    },
+    {
+      "epoch": 0.3525123032991069,
+      "grad_norm": 0.16532549262046814,
+      "learning_rate": 7.263254356019558e-05,
+      "loss": 1.071,
+      "step": 5802
+    },
+    {
+      "epoch": 0.3525730603317334,
+      "grad_norm": 0.18929168581962585,
+      "learning_rate": 7.262400722022165e-05,
+      "loss": 1.1551,
+      "step": 5803
+    },
+    {
+      "epoch": 0.35263381736435995,
+      "grad_norm": 0.19736342132091522,
+      "learning_rate": 7.261547005096036e-05,
+      "loss": 1.1133,
+      "step": 5804
+    },
+    {
+      "epoch": 0.3526945743969864,
+      "grad_norm": 0.270641565322876,
+      "learning_rate": 7.260693205272463e-05,
+      "loss": 1.1173,
+      "step": 5805
+    },
+    {
+      "epoch": 0.35275533142961296,
+      "grad_norm": 0.14757870137691498,
+      "learning_rate": 7.259839322582747e-05,
+      "loss": 1.0433,
+      "step": 5806
+    },
+    {
+      "epoch": 0.3528160884622395,
+      "grad_norm": 0.15412771701812744,
+      "learning_rate": 7.25898535705818e-05,
+      "loss": 1.0538,
+      "step": 5807
+    },
+    {
+      "epoch": 0.35287684549486603,
+      "grad_norm": 0.2050395905971527,
+      "learning_rate": 7.258131308730069e-05,
+      "loss": 1.13,
+      "step": 5808
+    },
+    {
+      "epoch": 0.35293760252749257,
+      "grad_norm": 0.18094982206821442,
+      "learning_rate": 7.25727717762972e-05,
+      "loss": 1.1525,
+      "step": 5809
+    },
+    {
+      "epoch": 0.3529983595601191,
+      "grad_norm": 0.1964893490076065,
+      "learning_rate": 7.256422963788438e-05,
+      "loss": 1.2286,
+      "step": 5810
+    },
+    {
+      "epoch": 0.35305911659274564,
+      "grad_norm": 0.21823422610759735,
+      "learning_rate": 7.255568667237537e-05,
+      "loss": 1.2311,
+      "step": 5811
+    },
+    {
+      "epoch": 0.3531198736253721,
+      "grad_norm": 0.14053837954998016,
+      "learning_rate": 7.25471428800833e-05,
+      "loss": 1.1162,
+      "step": 5812
+    },
+    {
+      "epoch": 0.35318063065799865,
+      "grad_norm": 0.17161621153354645,
+      "learning_rate": 7.253859826132137e-05,
+      "loss": 1.1776,
+      "step": 5813
+    },
+    {
+      "epoch": 0.3532413876906252,
+      "grad_norm": 0.34896978735923767,
+      "learning_rate": 7.253005281640275e-05,
+      "loss": 1.217,
+      "step": 5814
+    },
+    {
+      "epoch": 0.3533021447232517,
+      "grad_norm": 0.14858554303646088,
+      "learning_rate": 7.252150654564067e-05,
+      "loss": 1.0859,
+      "step": 5815
+    },
+    {
+      "epoch": 0.35336290175587826,
+      "grad_norm": 0.215206116437912,
+      "learning_rate": 7.251295944934843e-05,
+      "loss": 1.2089,
+      "step": 5816
+    },
+    {
+      "epoch": 0.3534236587885048,
+      "grad_norm": 1.9434363842010498,
+      "learning_rate": 7.250441152783931e-05,
+      "loss": 1.1805,
+      "step": 5817
+    },
+    {
+      "epoch": 0.35348441582113127,
+      "grad_norm": 0.1971471905708313,
+      "learning_rate": 7.249586278142664e-05,
+      "loss": 1.1045,
+      "step": 5818
+    },
+    {
+      "epoch": 0.3535451728537578,
+      "grad_norm": 0.25170570611953735,
+      "learning_rate": 7.248731321042377e-05,
+      "loss": 1.1289,
+      "step": 5819
+    },
+    {
+      "epoch": 0.35360592988638434,
+      "grad_norm": 0.20548677444458008,
+      "learning_rate": 7.247876281514408e-05,
+      "loss": 1.0986,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3536666869190109,
+      "grad_norm": 0.37013623118400574,
+      "learning_rate": 7.2470211595901e-05,
+      "loss": 1.1252,
+      "step": 5821
+    },
+    {
+      "epoch": 0.3537274439516374,
+      "grad_norm": 0.16794206202030182,
+      "learning_rate": 7.246165955300796e-05,
+      "loss": 1.0588,
+      "step": 5822
+    },
+    {
+      "epoch": 0.35378820098426395,
+      "grad_norm": 0.13325662910938263,
+      "learning_rate": 7.245310668677848e-05,
+      "loss": 1.0545,
+      "step": 5823
+    },
+    {
+      "epoch": 0.3538489580168905,
+      "grad_norm": 0.18467111885547638,
+      "learning_rate": 7.2444552997526e-05,
+      "loss": 1.0946,
+      "step": 5824
+    },
+    {
+      "epoch": 0.35390971504951696,
+      "grad_norm": 0.18164540827274323,
+      "learning_rate": 7.24359984855641e-05,
+      "loss": 1.1457,
+      "step": 5825
+    },
+    {
+      "epoch": 0.3539704720821435,
+      "grad_norm": 0.9037814736366272,
+      "learning_rate": 7.242744315120634e-05,
+      "loss": 1.0995,
+      "step": 5826
+    },
+    {
+      "epoch": 0.35403122911477003,
+      "grad_norm": 0.3074457049369812,
+      "learning_rate": 7.241888699476632e-05,
+      "loss": 1.1664,
+      "step": 5827
+    },
+    {
+      "epoch": 0.35409198614739656,
+      "grad_norm": 0.24883440136909485,
+      "learning_rate": 7.241033001655765e-05,
+      "loss": 1.0164,
+      "step": 5828
+    },
+    {
+      "epoch": 0.3541527431800231,
+      "grad_norm": 0.1257573962211609,
+      "learning_rate": 7.240177221689401e-05,
+      "loss": 1.076,
+      "step": 5829
+    },
+    {
+      "epoch": 0.35421350021264963,
+      "grad_norm": 0.14039850234985352,
+      "learning_rate": 7.239321359608908e-05,
+      "loss": 1.0648,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3542742572452761,
+      "grad_norm": 0.13982249796390533,
+      "learning_rate": 7.238465415445659e-05,
+      "loss": 1.0384,
+      "step": 5831
+    },
+    {
+      "epoch": 0.35433501427790265,
+      "grad_norm": 0.1574956178665161,
+      "learning_rate": 7.237609389231025e-05,
+      "loss": 1.1417,
+      "step": 5832
+    },
+    {
+      "epoch": 0.3543957713105292,
+      "grad_norm": 0.15349221229553223,
+      "learning_rate": 7.23675328099639e-05,
+      "loss": 1.1083,
+      "step": 5833
+    },
+    {
+      "epoch": 0.3544565283431557,
+      "grad_norm": 0.2074791043996811,
+      "learning_rate": 7.235897090773127e-05,
+      "loss": 1.0862,
+      "step": 5834
+    },
+    {
+      "epoch": 0.35451728537578225,
+      "grad_norm": 0.2703007757663727,
+      "learning_rate": 7.235040818592627e-05,
+      "loss": 1.2171,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3545780424084088,
+      "grad_norm": 0.2722099721431732,
+      "learning_rate": 7.234184464486271e-05,
+      "loss": 1.0861,
+      "step": 5836
+    },
+    {
+      "epoch": 0.3546387994410353,
+      "grad_norm": 0.1791897863149643,
+      "learning_rate": 7.233328028485454e-05,
+      "loss": 1.0108,
+      "step": 5837
+    },
+    {
+      "epoch": 0.3546995564736618,
+      "grad_norm": 0.19273117184638977,
+      "learning_rate": 7.232471510621565e-05,
+      "loss": 1.0425,
+      "step": 5838
+    },
+    {
+      "epoch": 0.35476031350628834,
+      "grad_norm": 0.14207425713539124,
+      "learning_rate": 7.231614910926002e-05,
+      "loss": 1.0942,
+      "step": 5839
+    },
+    {
+      "epoch": 0.3548210705389149,
+      "grad_norm": 0.24160359799861908,
+      "learning_rate": 7.230758229430164e-05,
+      "loss": 1.4506,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3548818275715414,
+      "grad_norm": 0.20541724562644958,
+      "learning_rate": 7.229901466165449e-05,
+      "loss": 1.1123,
+      "step": 5841
+    },
+    {
+      "epoch": 0.35494258460416794,
+      "grad_norm": 0.20765239000320435,
+      "learning_rate": 7.229044621163267e-05,
+      "loss": 1.0419,
+      "step": 5842
+    },
+    {
+      "epoch": 0.3550033416367945,
+      "grad_norm": 0.23890089988708496,
+      "learning_rate": 7.228187694455023e-05,
+      "loss": 1.1002,
+      "step": 5843
+    },
+    {
+      "epoch": 0.35506409866942096,
+      "grad_norm": 0.20605246722698212,
+      "learning_rate": 7.227330686072129e-05,
+      "loss": 1.0974,
+      "step": 5844
+    },
+    {
+      "epoch": 0.3551248557020475,
+      "grad_norm": 0.2058846801519394,
+      "learning_rate": 7.226473596045997e-05,
+      "loss": 1.0605,
+      "step": 5845
+    },
+    {
+      "epoch": 0.355185612734674,
+      "grad_norm": 0.14860503375530243,
+      "learning_rate": 7.225616424408045e-05,
+      "loss": 1.1214,
+      "step": 5846
+    },
+    {
+      "epoch": 0.35524636976730056,
+      "grad_norm": 0.271267831325531,
+      "learning_rate": 7.224759171189694e-05,
+      "loss": 1.1043,
+      "step": 5847
+    },
+    {
+      "epoch": 0.3553071267999271,
+      "grad_norm": 0.17829258739948273,
+      "learning_rate": 7.223901836422366e-05,
+      "loss": 1.085,
+      "step": 5848
+    },
+    {
+      "epoch": 0.35536788383255363,
+      "grad_norm": 0.2951408326625824,
+      "learning_rate": 7.223044420137484e-05,
+      "loss": 1.0586,
+      "step": 5849
+    },
+    {
+      "epoch": 0.35542864086518017,
+      "grad_norm": 0.2137591689825058,
+      "learning_rate": 7.22218692236648e-05,
+      "loss": 1.2399,
+      "step": 5850
+    },
+    {
+      "epoch": 0.35548939789780665,
+      "grad_norm": 0.23036998510360718,
+      "learning_rate": 7.221329343140784e-05,
+      "loss": 1.1197,
+      "step": 5851
+    },
+    {
+      "epoch": 0.3555501549304332,
+      "grad_norm": 0.15717172622680664,
+      "learning_rate": 7.220471682491834e-05,
+      "loss": 1.1348,
+      "step": 5852
+    },
+    {
+      "epoch": 0.3556109119630597,
+      "grad_norm": 0.13252952694892883,
+      "learning_rate": 7.219613940451064e-05,
+      "loss": 1.0578,
+      "step": 5853
+    },
+    {
+      "epoch": 0.35567166899568625,
+      "grad_norm": 0.1493271291255951,
+      "learning_rate": 7.218756117049915e-05,
+      "loss": 1.1174,
+      "step": 5854
+    },
+    {
+      "epoch": 0.3557324260283128,
+      "grad_norm": 0.19335941970348358,
+      "learning_rate": 7.217898212319833e-05,
+      "loss": 1.181,
+      "step": 5855
+    },
+    {
+      "epoch": 0.3557931830609393,
+      "grad_norm": 0.16902048885822296,
+      "learning_rate": 7.217040226292262e-05,
+      "loss": 1.1299,
+      "step": 5856
+    },
+    {
+      "epoch": 0.35585394009356586,
+      "grad_norm": 0.21546825766563416,
+      "learning_rate": 7.216182158998654e-05,
+      "loss": 1.1415,
+      "step": 5857
+    },
+    {
+      "epoch": 0.35591469712619234,
+      "grad_norm": 0.14659367501735687,
+      "learning_rate": 7.215324010470462e-05,
+      "loss": 1.0742,
+      "step": 5858
+    },
+    {
+      "epoch": 0.35597545415881887,
+      "grad_norm": 0.14131975173950195,
+      "learning_rate": 7.214465780739137e-05,
+      "loss": 1.0485,
+      "step": 5859
+    },
+    {
+      "epoch": 0.3560362111914454,
+      "grad_norm": 0.15080803632736206,
+      "learning_rate": 7.213607469836143e-05,
+      "loss": 1.1213,
+      "step": 5860
+    },
+    {
+      "epoch": 0.35609696822407194,
+      "grad_norm": 0.11239893734455109,
+      "learning_rate": 7.212749077792937e-05,
+      "loss": 1.0296,
+      "step": 5861
+    },
+    {
+      "epoch": 0.3561577252566985,
+      "grad_norm": 0.17335954308509827,
+      "learning_rate": 7.211890604640989e-05,
+      "loss": 1.1704,
+      "step": 5862
+    },
+    {
+      "epoch": 0.356218482289325,
+      "grad_norm": 0.1877175122499466,
+      "learning_rate": 7.211032050411763e-05,
+      "loss": 1.1369,
+      "step": 5863
+    },
+    {
+      "epoch": 0.3562792393219515,
+      "grad_norm": 0.46775996685028076,
+      "learning_rate": 7.210173415136727e-05,
+      "loss": 1.1046,
+      "step": 5864
+    },
+    {
+      "epoch": 0.356339996354578,
+      "grad_norm": 0.2042836993932724,
+      "learning_rate": 7.209314698847359e-05,
+      "loss": 1.2355,
+      "step": 5865
+    },
+    {
+      "epoch": 0.35640075338720456,
+      "grad_norm": 0.2687236964702606,
+      "learning_rate": 7.208455901575133e-05,
+      "loss": 1.2335,
+      "step": 5866
+    },
+    {
+      "epoch": 0.3564615104198311,
+      "grad_norm": 0.2034236043691635,
+      "learning_rate": 7.20759702335153e-05,
+      "loss": 1.143,
+      "step": 5867
+    },
+    {
+      "epoch": 0.35652226745245763,
+      "grad_norm": 0.1796652227640152,
+      "learning_rate": 7.206738064208031e-05,
+      "loss": 1.0573,
+      "step": 5868
+    },
+    {
+      "epoch": 0.35658302448508417,
+      "grad_norm": 0.147388756275177,
+      "learning_rate": 7.205879024176121e-05,
+      "loss": 1.0968,
+      "step": 5869
+    },
+    {
+      "epoch": 0.3566437815177107,
+      "grad_norm": 0.15133580565452576,
+      "learning_rate": 7.205019903287291e-05,
+      "loss": 1.0879,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3567045385503372,
+      "grad_norm": 0.19186252355575562,
+      "learning_rate": 7.204160701573028e-05,
+      "loss": 1.2921,
+      "step": 5871
+    },
+    {
+      "epoch": 0.3567652955829637,
+      "grad_norm": 0.18380627036094666,
+      "learning_rate": 7.203301419064828e-05,
+      "loss": 1.1138,
+      "step": 5872
+    },
+    {
+      "epoch": 0.35682605261559025,
+      "grad_norm": 0.16337788105010986,
+      "learning_rate": 7.202442055794192e-05,
+      "loss": 1.1027,
+      "step": 5873
+    },
+    {
+      "epoch": 0.3568868096482168,
+      "grad_norm": 0.12010836601257324,
+      "learning_rate": 7.201582611792614e-05,
+      "loss": 1.0696,
+      "step": 5874
+    },
+    {
+      "epoch": 0.3569475666808433,
+      "grad_norm": 0.20754846930503845,
+      "learning_rate": 7.2007230870916e-05,
+      "loss": 1.1449,
+      "step": 5875
+    },
+    {
+      "epoch": 0.35700832371346986,
+      "grad_norm": 0.426717609167099,
+      "learning_rate": 7.199863481722656e-05,
+      "loss": 1.2325,
+      "step": 5876
+    },
+    {
+      "epoch": 0.35706908074609633,
+      "grad_norm": 0.16668668389320374,
+      "learning_rate": 7.199003795717292e-05,
+      "loss": 1.0529,
+      "step": 5877
+    },
+    {
+      "epoch": 0.35712983777872287,
+      "grad_norm": 0.149159774184227,
+      "learning_rate": 7.198144029107018e-05,
+      "loss": 1.0929,
+      "step": 5878
+    },
+    {
+      "epoch": 0.3571905948113494,
+      "grad_norm": 0.3591291904449463,
+      "learning_rate": 7.19728418192335e-05,
+      "loss": 1.0927,
+      "step": 5879
+    },
+    {
+      "epoch": 0.35725135184397594,
+      "grad_norm": 0.17408284544944763,
+      "learning_rate": 7.196424254197804e-05,
+      "loss": 1.1429,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3573121088766025,
+      "grad_norm": 0.815989077091217,
+      "learning_rate": 7.195564245961904e-05,
+      "loss": 1.1256,
+      "step": 5881
+    },
+    {
+      "epoch": 0.357372865909229,
+      "grad_norm": 22.102617263793945,
+      "learning_rate": 7.194704157247172e-05,
+      "loss": 1.0558,
+      "step": 5882
+    },
+    {
+      "epoch": 0.35743362294185554,
+      "grad_norm": 0.33766666054725647,
+      "learning_rate": 7.193843988085135e-05,
+      "loss": 1.1757,
+      "step": 5883
+    },
+    {
+      "epoch": 0.357494379974482,
+      "grad_norm": 1.3121384382247925,
+      "learning_rate": 7.192983738507322e-05,
+      "loss": 1.0954,
+      "step": 5884
+    },
+    {
+      "epoch": 0.35755513700710856,
+      "grad_norm": 0.1580880731344223,
+      "learning_rate": 7.192123408545266e-05,
+      "loss": 1.1042,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3576158940397351,
+      "grad_norm": 2.1086742877960205,
+      "learning_rate": 7.191262998230504e-05,
+      "loss": 1.2362,
+      "step": 5886
+    },
+    {
+      "epoch": 0.35767665107236163,
+      "grad_norm": 0.23374570906162262,
+      "learning_rate": 7.190402507594571e-05,
+      "loss": 1.1557,
+      "step": 5887
+    },
+    {
+      "epoch": 0.35773740810498816,
+      "grad_norm": 0.17526526749134064,
+      "learning_rate": 7.189541936669013e-05,
+      "loss": 1.1567,
+      "step": 5888
+    },
+    {
+      "epoch": 0.3577981651376147,
+      "grad_norm": 0.18390122056007385,
+      "learning_rate": 7.188681285485371e-05,
+      "loss": 1.0668,
+      "step": 5889
+    },
+    {
+      "epoch": 0.3578589221702412,
+      "grad_norm": 0.2820643484592438,
+      "learning_rate": 7.187820554075193e-05,
+      "loss": 1.2079,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3579196792028677,
+      "grad_norm": 0.17928802967071533,
+      "learning_rate": 7.186959742470031e-05,
+      "loss": 1.1336,
+      "step": 5891
+    },
+    {
+      "epoch": 0.35798043623549425,
+      "grad_norm": 0.22715944051742554,
+      "learning_rate": 7.186098850701435e-05,
+      "loss": 1.1369,
+      "step": 5892
+    },
+    {
+      "epoch": 0.3580411932681208,
+      "grad_norm": 0.22867751121520996,
+      "learning_rate": 7.185237878800966e-05,
+      "loss": 1.2372,
+      "step": 5893
+    },
+    {
+      "epoch": 0.3581019503007473,
+      "grad_norm": 0.26421797275543213,
+      "learning_rate": 7.184376826800178e-05,
+      "loss": 1.0644,
+      "step": 5894
+    },
+    {
+      "epoch": 0.35816270733337385,
+      "grad_norm": 0.40871885418891907,
+      "learning_rate": 7.183515694730635e-05,
+      "loss": 1.2616,
+      "step": 5895
+    },
+    {
+      "epoch": 0.3582234643660004,
+      "grad_norm": 0.174764946103096,
+      "learning_rate": 7.182654482623903e-05,
+      "loss": 1.1292,
+      "step": 5896
+    },
+    {
+      "epoch": 0.35828422139862687,
+      "grad_norm": 0.1459951102733612,
+      "learning_rate": 7.181793190511549e-05,
+      "loss": 1.0934,
+      "step": 5897
+    },
+    {
+      "epoch": 0.3583449784312534,
+      "grad_norm": 0.2506260275840759,
+      "learning_rate": 7.180931818425142e-05,
+      "loss": 1.2078,
+      "step": 5898
+    },
+    {
+      "epoch": 0.35840573546387994,
+      "grad_norm": 0.21106983721256256,
+      "learning_rate": 7.180070366396261e-05,
+      "loss": 1.158,
+      "step": 5899
+    },
+    {
+      "epoch": 0.3584664924965065,
+      "grad_norm": 0.2122010886669159,
+      "learning_rate": 7.179208834456477e-05,
+      "loss": 1.0334,
+      "step": 5900
+    },
+    {
+      "epoch": 0.358527249529133,
+      "grad_norm": 0.2114807516336441,
+      "learning_rate": 7.178347222637373e-05,
+      "loss": 1.0905,
+      "step": 5901
+    },
+    {
+      "epoch": 0.35858800656175954,
+      "grad_norm": 0.21153010427951813,
+      "learning_rate": 7.17748553097053e-05,
+      "loss": 1.0293,
+      "step": 5902
+    },
+    {
+      "epoch": 0.3586487635943861,
+      "grad_norm": 0.22813719511032104,
+      "learning_rate": 7.176623759487534e-05,
+      "loss": 1.0809,
+      "step": 5903
+    },
+    {
+      "epoch": 0.35870952062701256,
+      "grad_norm": 0.1508573740720749,
+      "learning_rate": 7.175761908219974e-05,
+      "loss": 1.0913,
+      "step": 5904
+    },
+    {
+      "epoch": 0.3587702776596391,
+      "grad_norm": 0.20379769802093506,
+      "learning_rate": 7.174899977199439e-05,
+      "loss": 1.2287,
+      "step": 5905
+    },
+    {
+      "epoch": 0.3588310346922656,
+      "grad_norm": 0.2024427056312561,
+      "learning_rate": 7.174037966457525e-05,
+      "loss": 1.1184,
+      "step": 5906
+    },
+    {
+      "epoch": 0.35889179172489216,
+      "grad_norm": 0.14341942965984344,
+      "learning_rate": 7.173175876025831e-05,
+      "loss": 1.0739,
+      "step": 5907
+    },
+    {
+      "epoch": 0.3589525487575187,
+      "grad_norm": 0.18194203078746796,
+      "learning_rate": 7.172313705935955e-05,
+      "loss": 1.0635,
+      "step": 5908
+    },
+    {
+      "epoch": 0.35901330579014523,
+      "grad_norm": 0.4800781309604645,
+      "learning_rate": 7.171451456219499e-05,
+      "loss": 1.0819,
+      "step": 5909
+    },
+    {
+      "epoch": 0.3590740628227717,
+      "grad_norm": 0.18123412132263184,
+      "learning_rate": 7.17058912690807e-05,
+      "loss": 1.126,
+      "step": 5910
+    },
+    {
+      "epoch": 0.35913481985539825,
+      "grad_norm": 0.14279356598854065,
+      "learning_rate": 7.169726718033278e-05,
+      "loss": 1.2625,
+      "step": 5911
+    },
+    {
+      "epoch": 0.3591955768880248,
+      "grad_norm": 0.22185450792312622,
+      "learning_rate": 7.168864229626734e-05,
+      "loss": 1.0829,
+      "step": 5912
+    },
+    {
+      "epoch": 0.3592563339206513,
+      "grad_norm": 0.13260206580162048,
+      "learning_rate": 7.168001661720051e-05,
+      "loss": 1.0462,
+      "step": 5913
+    },
+    {
+      "epoch": 0.35931709095327785,
+      "grad_norm": 0.18639041483402252,
+      "learning_rate": 7.16713901434485e-05,
+      "loss": 1.1881,
+      "step": 5914
+    },
+    {
+      "epoch": 0.3593778479859044,
+      "grad_norm": 0.24409301578998566,
+      "learning_rate": 7.166276287532748e-05,
+      "loss": 1.1298,
+      "step": 5915
+    },
+    {
+      "epoch": 0.3594386050185309,
+      "grad_norm": 0.17557953298091888,
+      "learning_rate": 7.16541348131537e-05,
+      "loss": 1.0667,
+      "step": 5916
+    },
+    {
+      "epoch": 0.3594993620511574,
+      "grad_norm": 0.2696946859359741,
+      "learning_rate": 7.164550595724344e-05,
+      "loss": 1.0989,
+      "step": 5917
+    },
+    {
+      "epoch": 0.35956011908378394,
+      "grad_norm": 0.18718624114990234,
+      "learning_rate": 7.163687630791294e-05,
+      "loss": 1.0972,
+      "step": 5918
+    },
+    {
+      "epoch": 0.35962087611641047,
+      "grad_norm": 0.14111928641796112,
+      "learning_rate": 7.162824586547858e-05,
+      "loss": 1.1031,
+      "step": 5919
+    },
+    {
+      "epoch": 0.359681633149037,
+      "grad_norm": 0.20912785828113556,
+      "learning_rate": 7.161961463025669e-05,
+      "loss": 1.1477,
+      "step": 5920
+    },
+    {
+      "epoch": 0.35974239018166354,
+      "grad_norm": 0.19308944046497345,
+      "learning_rate": 7.161098260256363e-05,
+      "loss": 1.1293,
+      "step": 5921
+    },
+    {
+      "epoch": 0.3598031472142901,
+      "grad_norm": 0.33384910225868225,
+      "learning_rate": 7.160234978271585e-05,
+      "loss": 1.1733,
+      "step": 5922
+    },
+    {
+      "epoch": 0.35986390424691655,
+      "grad_norm": 0.2403935194015503,
+      "learning_rate": 7.159371617102975e-05,
+      "loss": 1.0779,
+      "step": 5923
+    },
+    {
+      "epoch": 0.3599246612795431,
+      "grad_norm": 0.15773074328899384,
+      "learning_rate": 7.15850817678218e-05,
+      "loss": 1.1289,
+      "step": 5924
+    },
+    {
+      "epoch": 0.3599854183121696,
+      "grad_norm": 0.15671879053115845,
+      "learning_rate": 7.157644657340852e-05,
+      "loss": 1.0892,
+      "step": 5925
+    },
+    {
+      "epoch": 0.36004617534479616,
+      "grad_norm": 0.17868761718273163,
+      "learning_rate": 7.15678105881064e-05,
+      "loss": 1.1468,
+      "step": 5926
+    },
+    {
+      "epoch": 0.3601069323774227,
+      "grad_norm": 0.18014110624790192,
+      "learning_rate": 7.155917381223205e-05,
+      "loss": 1.135,
+      "step": 5927
+    },
+    {
+      "epoch": 0.36016768941004923,
+      "grad_norm": 3.4692015647888184,
+      "learning_rate": 7.1550536246102e-05,
+      "loss": 1.0914,
+      "step": 5928
+    },
+    {
+      "epoch": 0.36022844644267576,
+      "grad_norm": 0.19155573844909668,
+      "learning_rate": 7.154189789003286e-05,
+      "loss": 1.0702,
+      "step": 5929
+    },
+    {
+      "epoch": 0.36028920347530224,
+      "grad_norm": 0.1417994350194931,
+      "learning_rate": 7.153325874434131e-05,
+      "loss": 1.0433,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3603499605079288,
+      "grad_norm": 0.2555970251560211,
+      "learning_rate": 7.152461880934399e-05,
+      "loss": 1.0917,
+      "step": 5931
+    },
+    {
+      "epoch": 0.3604107175405553,
+      "grad_norm": 0.16115422546863556,
+      "learning_rate": 7.151597808535761e-05,
+      "loss": 1.0477,
+      "step": 5932
+    },
+    {
+      "epoch": 0.36047147457318185,
+      "grad_norm": 0.2958734929561615,
+      "learning_rate": 7.150733657269889e-05,
+      "loss": 1.0699,
+      "step": 5933
+    },
+    {
+      "epoch": 0.3605322316058084,
+      "grad_norm": 0.18945914506912231,
+      "learning_rate": 7.14986942716846e-05,
+      "loss": 1.1249,
+      "step": 5934
+    },
+    {
+      "epoch": 0.3605929886384349,
+      "grad_norm": 0.3092036843299866,
+      "learning_rate": 7.14900511826315e-05,
+      "loss": 1.092,
+      "step": 5935
+    },
+    {
+      "epoch": 0.3606537456710614,
+      "grad_norm": 0.1366962194442749,
+      "learning_rate": 7.148140730585644e-05,
+      "loss": 1.0881,
+      "step": 5936
+    },
+    {
+      "epoch": 0.36071450270368793,
+      "grad_norm": 0.17601732909679413,
+      "learning_rate": 7.147276264167623e-05,
+      "loss": 1.1401,
+      "step": 5937
+    },
+    {
+      "epoch": 0.36077525973631447,
+      "grad_norm": 0.20743726193904877,
+      "learning_rate": 7.146411719040775e-05,
+      "loss": 1.0814,
+      "step": 5938
+    },
+    {
+      "epoch": 0.360836016768941,
+      "grad_norm": 0.1295817494392395,
+      "learning_rate": 7.145547095236791e-05,
+      "loss": 1.0473,
+      "step": 5939
+    },
+    {
+      "epoch": 0.36089677380156754,
+      "grad_norm": 0.1812259554862976,
+      "learning_rate": 7.144682392787365e-05,
+      "loss": 1.102,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3609575308341941,
+      "grad_norm": 0.2486310601234436,
+      "learning_rate": 7.14381761172419e-05,
+      "loss": 1.2024,
+      "step": 5941
+    },
+    {
+      "epoch": 0.3610182878668206,
+      "grad_norm": 0.15354949235916138,
+      "learning_rate": 7.142952752078965e-05,
+      "loss": 1.0609,
+      "step": 5942
+    },
+    {
+      "epoch": 0.3610790448994471,
+      "grad_norm": 0.532869815826416,
+      "learning_rate": 7.142087813883395e-05,
+      "loss": 1.2216,
+      "step": 5943
+    },
+    {
+      "epoch": 0.3611398019320736,
+      "grad_norm": 0.1500968188047409,
+      "learning_rate": 7.14122279716918e-05,
+      "loss": 1.067,
+      "step": 5944
+    },
+    {
+      "epoch": 0.36120055896470016,
+      "grad_norm": 0.1807074397802353,
+      "learning_rate": 7.140357701968031e-05,
+      "loss": 1.1021,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3612613159973267,
+      "grad_norm": 0.20345276594161987,
+      "learning_rate": 7.139492528311657e-05,
+      "loss": 1.1096,
+      "step": 5946
+    },
+    {
+      "epoch": 0.36132207302995323,
+      "grad_norm": 0.39479169249534607,
+      "learning_rate": 7.138627276231768e-05,
+      "loss": 1.3074,
+      "step": 5947
+    },
+    {
+      "epoch": 0.36138283006257976,
+      "grad_norm": 0.4613039195537567,
+      "learning_rate": 7.137761945760086e-05,
+      "loss": 1.1521,
+      "step": 5948
+    },
+    {
+      "epoch": 0.36144358709520624,
+      "grad_norm": 0.4169823229312897,
+      "learning_rate": 7.136896536928326e-05,
+      "loss": 1.1755,
+      "step": 5949
+    },
+    {
+      "epoch": 0.3615043441278328,
+      "grad_norm": 0.19304074347019196,
+      "learning_rate": 7.13603104976821e-05,
+      "loss": 1.102,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3615651011604593,
+      "grad_norm": 0.3686462640762329,
+      "learning_rate": 7.135165484311462e-05,
+      "loss": 1.252,
+      "step": 5951
+    },
+    {
+      "epoch": 0.36162585819308585,
+      "grad_norm": 0.26206162571907043,
+      "learning_rate": 7.134299840589811e-05,
+      "loss": 1.0926,
+      "step": 5952
+    },
+    {
+      "epoch": 0.3616866152257124,
+      "grad_norm": 0.34802401065826416,
+      "learning_rate": 7.133434118634986e-05,
+      "loss": 1.0846,
+      "step": 5953
+    },
+    {
+      "epoch": 0.3617473722583389,
+      "grad_norm": 0.26187098026275635,
+      "learning_rate": 7.132568318478722e-05,
+      "loss": 1.2137,
+      "step": 5954
+    },
+    {
+      "epoch": 0.36180812929096545,
+      "grad_norm": 0.5490225553512573,
+      "learning_rate": 7.131702440152755e-05,
+      "loss": 1.0896,
+      "step": 5955
+    },
+    {
+      "epoch": 0.36186888632359193,
+      "grad_norm": 0.32293233275413513,
+      "learning_rate": 7.130836483688821e-05,
+      "loss": 1.0628,
+      "step": 5956
+    },
+    {
+      "epoch": 0.36192964335621847,
+      "grad_norm": 0.2808167636394501,
+      "learning_rate": 7.129970449118666e-05,
+      "loss": 1.0765,
+      "step": 5957
+    },
+    {
+      "epoch": 0.361990400388845,
+      "grad_norm": 0.24105970561504364,
+      "learning_rate": 7.129104336474031e-05,
+      "loss": 1.1343,
+      "step": 5958
+    },
+    {
+      "epoch": 0.36205115742147154,
+      "grad_norm": 0.30015817284584045,
+      "learning_rate": 7.128238145786665e-05,
+      "loss": 1.1138,
+      "step": 5959
+    },
+    {
+      "epoch": 0.36211191445409807,
+      "grad_norm": 0.17284353077411652,
+      "learning_rate": 7.12737187708832e-05,
+      "loss": 1.0238,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3621726714867246,
+      "grad_norm": 0.18822729587554932,
+      "learning_rate": 7.126505530410746e-05,
+      "loss": 1.1787,
+      "step": 5961
+    },
+    {
+      "epoch": 0.36223342851935114,
+      "grad_norm": 0.2871028184890747,
+      "learning_rate": 7.125639105785699e-05,
+      "loss": 1.1398,
+      "step": 5962
+    },
+    {
+      "epoch": 0.3622941855519776,
+      "grad_norm": 0.21566082537174225,
+      "learning_rate": 7.124772603244943e-05,
+      "loss": 1.1501,
+      "step": 5963
+    },
+    {
+      "epoch": 0.36235494258460416,
+      "grad_norm": 0.16293512284755707,
+      "learning_rate": 7.123906022820238e-05,
+      "loss": 1.096,
+      "step": 5964
+    },
+    {
+      "epoch": 0.3624156996172307,
+      "grad_norm": 0.14977681636810303,
+      "learning_rate": 7.123039364543342e-05,
+      "loss": 1.0984,
+      "step": 5965
+    },
+    {
+      "epoch": 0.3624764566498572,
+      "grad_norm": 0.12604397535324097,
+      "learning_rate": 7.122172628446031e-05,
+      "loss": 0.9909,
+      "step": 5966
+    },
+    {
+      "epoch": 0.36253721368248376,
+      "grad_norm": 0.7872484922409058,
+      "learning_rate": 7.12130581456007e-05,
+      "loss": 1.196,
+      "step": 5967
+    },
+    {
+      "epoch": 0.3625979707151103,
+      "grad_norm": 0.19841915369033813,
+      "learning_rate": 7.120438922917236e-05,
+      "loss": 1.1695,
+      "step": 5968
+    },
+    {
+      "epoch": 0.3626587277477368,
+      "grad_norm": 0.20634128153324127,
+      "learning_rate": 7.119571953549305e-05,
+      "loss": 1.1998,
+      "step": 5969
+    },
+    {
+      "epoch": 0.3627194847803633,
+      "grad_norm": 0.12938696146011353,
+      "learning_rate": 7.118704906488052e-05,
+      "loss": 1.06,
+      "step": 5970
+    },
+    {
+      "epoch": 0.36278024181298985,
+      "grad_norm": 0.16447527706623077,
+      "learning_rate": 7.117837781765261e-05,
+      "loss": 1.1027,
+      "step": 5971
+    },
+    {
+      "epoch": 0.3628409988456164,
+      "grad_norm": 3.4606661796569824,
+      "learning_rate": 7.116970579412717e-05,
+      "loss": 1.1739,
+      "step": 5972
+    },
+    {
+      "epoch": 0.3629017558782429,
+      "grad_norm": 0.2508723735809326,
+      "learning_rate": 7.116103299462207e-05,
+      "loss": 1.1552,
+      "step": 5973
+    },
+    {
+      "epoch": 0.36296251291086945,
+      "grad_norm": 0.28259187936782837,
+      "learning_rate": 7.115235941945524e-05,
+      "loss": 1.1627,
+      "step": 5974
+    },
+    {
+      "epoch": 0.363023269943496,
+      "grad_norm": 0.20259036123752594,
+      "learning_rate": 7.114368506894456e-05,
+      "loss": 1.0997,
+      "step": 5975
+    },
+    {
+      "epoch": 0.36308402697612246,
+      "grad_norm": 0.20449164509773254,
+      "learning_rate": 7.113500994340801e-05,
+      "loss": 1.1848,
+      "step": 5976
+    },
+    {
+      "epoch": 0.363144784008749,
+      "grad_norm": 0.15127572417259216,
+      "learning_rate": 7.112633404316361e-05,
+      "loss": 1.0675,
+      "step": 5977
+    },
+    {
+      "epoch": 0.36320554104137553,
+      "grad_norm": 0.29695481061935425,
+      "learning_rate": 7.111765736852934e-05,
+      "loss": 1.192,
+      "step": 5978
+    },
+    {
+      "epoch": 0.36326629807400207,
+      "grad_norm": 1.2648652791976929,
+      "learning_rate": 7.110897991982326e-05,
+      "loss": 1.1737,
+      "step": 5979
+    },
+    {
+      "epoch": 0.3633270551066286,
+      "grad_norm": 0.31794071197509766,
+      "learning_rate": 7.110030169736344e-05,
+      "loss": 1.1979,
+      "step": 5980
+    },
+    {
+      "epoch": 0.36338781213925514,
+      "grad_norm": 0.21422503888607025,
+      "learning_rate": 7.109162270146797e-05,
+      "loss": 1.1712,
+      "step": 5981
+    },
+    {
+      "epoch": 0.3634485691718816,
+      "grad_norm": 0.33230382204055786,
+      "learning_rate": 7.108294293245501e-05,
+      "loss": 1.1507,
+      "step": 5982
+    },
+    {
+      "epoch": 0.36350932620450815,
+      "grad_norm": 0.22521817684173584,
+      "learning_rate": 7.107426239064269e-05,
+      "loss": 1.0411,
+      "step": 5983
+    },
+    {
+      "epoch": 0.3635700832371347,
+      "grad_norm": 0.22728776931762695,
+      "learning_rate": 7.106558107634923e-05,
+      "loss": 1.2453,
+      "step": 5984
+    },
+    {
+      "epoch": 0.3636308402697612,
+      "grad_norm": 0.13818351924419403,
+      "learning_rate": 7.10568989898928e-05,
+      "loss": 1.0527,
+      "step": 5985
+    },
+    {
+      "epoch": 0.36369159730238776,
+      "grad_norm": 0.25233694911003113,
+      "learning_rate": 7.104821613159167e-05,
+      "loss": 1.2351,
+      "step": 5986
+    },
+    {
+      "epoch": 0.3637523543350143,
+      "grad_norm": 0.19004298746585846,
+      "learning_rate": 7.103953250176413e-05,
+      "loss": 1.1024,
+      "step": 5987
+    },
+    {
+      "epoch": 0.36381311136764083,
+      "grad_norm": 0.37169086933135986,
+      "learning_rate": 7.103084810072843e-05,
+      "loss": 1.1874,
+      "step": 5988
+    },
+    {
+      "epoch": 0.3638738684002673,
+      "grad_norm": 0.15443316102027893,
+      "learning_rate": 7.102216292880295e-05,
+      "loss": 1.1087,
+      "step": 5989
+    },
+    {
+      "epoch": 0.36393462543289384,
+      "grad_norm": 0.1607917994260788,
+      "learning_rate": 7.101347698630601e-05,
+      "loss": 1.1258,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3639953824655204,
+      "grad_norm": 0.1835775226354599,
+      "learning_rate": 7.100479027355603e-05,
+      "loss": 1.0674,
+      "step": 5991
+    },
+    {
+      "epoch": 0.3640561394981469,
+      "grad_norm": 0.5484330058097839,
+      "learning_rate": 7.099610279087138e-05,
+      "loss": 1.1876,
+      "step": 5992
+    },
+    {
+      "epoch": 0.36411689653077345,
+      "grad_norm": 0.4057673513889313,
+      "learning_rate": 7.098741453857053e-05,
+      "loss": 1.1698,
+      "step": 5993
+    },
+    {
+      "epoch": 0.3641776535634,
+      "grad_norm": 0.27871638536453247,
+      "learning_rate": 7.097872551697195e-05,
+      "loss": 1.051,
+      "step": 5994
+    },
+    {
+      "epoch": 0.36423841059602646,
+      "grad_norm": 0.16202746331691742,
+      "learning_rate": 7.097003572639413e-05,
+      "loss": 1.1082,
+      "step": 5995
+    },
+    {
+      "epoch": 0.364299167628653,
+      "grad_norm": 0.2087109088897705,
+      "learning_rate": 7.096134516715557e-05,
+      "loss": 1.0972,
+      "step": 5996
+    },
+    {
+      "epoch": 0.36435992466127953,
+      "grad_norm": 0.187360480427742,
+      "learning_rate": 7.09526538395749e-05,
+      "loss": 1.0991,
+      "step": 5997
+    },
+    {
+      "epoch": 0.36442068169390607,
+      "grad_norm": 0.13905178010463715,
+      "learning_rate": 7.094396174397061e-05,
+      "loss": 1.0998,
+      "step": 5998
+    },
+    {
+      "epoch": 0.3644814387265326,
+      "grad_norm": 0.17264042794704437,
+      "learning_rate": 7.093526888066137e-05,
+      "loss": 1.1434,
+      "step": 5999
+    },
+    {
+      "epoch": 0.36454219575915914,
+      "grad_norm": 0.2895245850086212,
+      "learning_rate": 7.09265752499658e-05,
+      "loss": 1.2232,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3646029527917857,
+      "grad_norm": 0.14069925248622894,
+      "learning_rate": 7.091788085220259e-05,
+      "loss": 1.1153,
+      "step": 6001
+    },
+    {
+      "epoch": 0.36466370982441215,
+      "grad_norm": 0.15252617001533508,
+      "learning_rate": 7.09091856876904e-05,
+      "loss": 1.112,
+      "step": 6002
+    },
+    {
+      "epoch": 0.3647244668570387,
+      "grad_norm": 0.3880089521408081,
+      "learning_rate": 7.090048975674796e-05,
+      "loss": 1.2197,
+      "step": 6003
+    },
+    {
+      "epoch": 0.3647852238896652,
+      "grad_norm": 0.7139638066291809,
+      "learning_rate": 7.089179305969403e-05,
+      "loss": 1.0787,
+      "step": 6004
+    },
+    {
+      "epoch": 0.36484598092229176,
+      "grad_norm": 0.2053522914648056,
+      "learning_rate": 7.088309559684739e-05,
+      "loss": 1.0356,
+      "step": 6005
+    },
+    {
+      "epoch": 0.3649067379549183,
+      "grad_norm": 0.18879061937332153,
+      "learning_rate": 7.087439736852684e-05,
+      "loss": 1.0274,
+      "step": 6006
+    },
+    {
+      "epoch": 0.3649674949875448,
+      "grad_norm": 0.15123805403709412,
+      "learning_rate": 7.086569837505122e-05,
+      "loss": 1.1557,
+      "step": 6007
+    },
+    {
+      "epoch": 0.36502825202017136,
+      "grad_norm": 0.1728934943675995,
+      "learning_rate": 7.08569986167394e-05,
+      "loss": 1.1208,
+      "step": 6008
+    },
+    {
+      "epoch": 0.36508900905279784,
+      "grad_norm": 0.1747850626707077,
+      "learning_rate": 7.084829809391026e-05,
+      "loss": 1.1182,
+      "step": 6009
+    },
+    {
+      "epoch": 0.3651497660854244,
+      "grad_norm": 0.17299675941467285,
+      "learning_rate": 7.083959680688273e-05,
+      "loss": 1.0947,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3652105231180509,
+      "grad_norm": 0.18715403974056244,
+      "learning_rate": 7.083089475597574e-05,
+      "loss": 1.089,
+      "step": 6011
+    },
+    {
+      "epoch": 0.36527128015067745,
+      "grad_norm": 0.23310427367687225,
+      "learning_rate": 7.082219194150828e-05,
+      "loss": 1.0448,
+      "step": 6012
+    },
+    {
+      "epoch": 0.365332037183304,
+      "grad_norm": 0.12462879717350006,
+      "learning_rate": 7.081348836379935e-05,
+      "loss": 1.0576,
+      "step": 6013
+    },
+    {
+      "epoch": 0.3653927942159305,
+      "grad_norm": 0.5681056380271912,
+      "learning_rate": 7.080478402316798e-05,
+      "loss": 1.0825,
+      "step": 6014
+    },
+    {
+      "epoch": 0.365453551248557,
+      "grad_norm": 4.0196919441223145,
+      "learning_rate": 7.079607891993323e-05,
+      "loss": 1.1142,
+      "step": 6015
+    },
+    {
+      "epoch": 0.36551430828118353,
+      "grad_norm": 0.27448785305023193,
+      "learning_rate": 7.07873730544142e-05,
+      "loss": 1.0766,
+      "step": 6016
+    },
+    {
+      "epoch": 0.36557506531381007,
+      "grad_norm": 0.1731339395046234,
+      "learning_rate": 7.077866642692998e-05,
+      "loss": 1.0986,
+      "step": 6017
+    },
+    {
+      "epoch": 0.3656358223464366,
+      "grad_norm": 0.22972512245178223,
+      "learning_rate": 7.076995903779972e-05,
+      "loss": 1.1725,
+      "step": 6018
+    },
+    {
+      "epoch": 0.36569657937906314,
+      "grad_norm": 4.182675838470459,
+      "learning_rate": 7.07612508873426e-05,
+      "loss": 1.0308,
+      "step": 6019
+    },
+    {
+      "epoch": 0.36575733641168967,
+      "grad_norm": 0.15066596865653992,
+      "learning_rate": 7.075254197587782e-05,
+      "loss": 1.0312,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3658180934443162,
+      "grad_norm": 0.205636665225029,
+      "learning_rate": 7.074383230372461e-05,
+      "loss": 1.0671,
+      "step": 6021
+    },
+    {
+      "epoch": 0.3658788504769427,
+      "grad_norm": 0.1955956667661667,
+      "learning_rate": 7.073512187120221e-05,
+      "loss": 1.1173,
+      "step": 6022
+    },
+    {
+      "epoch": 0.3659396075095692,
+      "grad_norm": 0.25524693727493286,
+      "learning_rate": 7.072641067862991e-05,
+      "loss": 1.2482,
+      "step": 6023
+    },
+    {
+      "epoch": 0.36600036454219576,
+      "grad_norm": 0.2876506447792053,
+      "learning_rate": 7.071769872632701e-05,
+      "loss": 1.3079,
+      "step": 6024
+    },
+    {
+      "epoch": 0.3660611215748223,
+      "grad_norm": 0.20998498797416687,
+      "learning_rate": 7.070898601461287e-05,
+      "loss": 1.0703,
+      "step": 6025
+    },
+    {
+      "epoch": 0.3661218786074488,
+      "grad_norm": 0.16320553421974182,
+      "learning_rate": 7.070027254380684e-05,
+      "loss": 1.0442,
+      "step": 6026
+    },
+    {
+      "epoch": 0.36618263564007536,
+      "grad_norm": 0.2894010841846466,
+      "learning_rate": 7.069155831422832e-05,
+      "loss": 1.0691,
+      "step": 6027
+    },
+    {
+      "epoch": 0.36624339267270184,
+      "grad_norm": 0.2230319380760193,
+      "learning_rate": 7.068284332619673e-05,
+      "loss": 1.069,
+      "step": 6028
+    },
+    {
+      "epoch": 0.3663041497053284,
+      "grad_norm": 0.3279269337654114,
+      "learning_rate": 7.067412758003154e-05,
+      "loss": 1.1033,
+      "step": 6029
+    },
+    {
+      "epoch": 0.3663649067379549,
+      "grad_norm": 0.16777008771896362,
+      "learning_rate": 7.066541107605217e-05,
+      "loss": 1.1586,
+      "step": 6030
+    },
+    {
+      "epoch": 0.36642566377058144,
+      "grad_norm": 0.23910661041736603,
+      "learning_rate": 7.065669381457819e-05,
+      "loss": 1.127,
+      "step": 6031
+    },
+    {
+      "epoch": 0.366486420803208,
+      "grad_norm": 0.6004482507705688,
+      "learning_rate": 7.064797579592909e-05,
+      "loss": 1.0522,
+      "step": 6032
+    },
+    {
+      "epoch": 0.3665471778358345,
+      "grad_norm": 0.22792962193489075,
+      "learning_rate": 7.063925702042446e-05,
+      "loss": 1.0699,
+      "step": 6033
+    },
+    {
+      "epoch": 0.36660793486846105,
+      "grad_norm": 0.17332828044891357,
+      "learning_rate": 7.063053748838386e-05,
+      "loss": 1.1035,
+      "step": 6034
+    },
+    {
+      "epoch": 0.36666869190108753,
+      "grad_norm": 0.24174825847148895,
+      "learning_rate": 7.062181720012693e-05,
+      "loss": 1.1761,
+      "step": 6035
+    },
+    {
+      "epoch": 0.36672944893371406,
+      "grad_norm": 0.19885759055614471,
+      "learning_rate": 7.06130961559733e-05,
+      "loss": 1.0727,
+      "step": 6036
+    },
+    {
+      "epoch": 0.3667902059663406,
+      "grad_norm": 0.19328734278678894,
+      "learning_rate": 7.060437435624265e-05,
+      "loss": 1.0582,
+      "step": 6037
+    },
+    {
+      "epoch": 0.36685096299896713,
+      "grad_norm": 0.19828392565250397,
+      "learning_rate": 7.059565180125466e-05,
+      "loss": 1.1848,
+      "step": 6038
+    },
+    {
+      "epoch": 0.36691172003159367,
+      "grad_norm": 0.20452788472175598,
+      "learning_rate": 7.05869284913291e-05,
+      "loss": 1.0519,
+      "step": 6039
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 0.2678563892841339,
+      "learning_rate": 7.057820442678567e-05,
+      "loss": 1.2584,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3670332340968467,
+      "grad_norm": 0.2401776760816574,
+      "learning_rate": 7.05694796079442e-05,
+      "loss": 1.0755,
+      "step": 6041
+    },
+    {
+      "epoch": 0.3670939911294732,
+      "grad_norm": 0.21597108244895935,
+      "learning_rate": 7.056075403512448e-05,
+      "loss": 1.2046,
+      "step": 6042
+    },
+    {
+      "epoch": 0.36715474816209975,
+      "grad_norm": 0.20854249596595764,
+      "learning_rate": 7.055202770864633e-05,
+      "loss": 1.0752,
+      "step": 6043
+    },
+    {
+      "epoch": 0.3672155051947263,
+      "grad_norm": 0.1955675333738327,
+      "learning_rate": 7.054330062882964e-05,
+      "loss": 1.0651,
+      "step": 6044
+    },
+    {
+      "epoch": 0.3672762622273528,
+      "grad_norm": 0.26309099793434143,
+      "learning_rate": 7.05345727959943e-05,
+      "loss": 1.271,
+      "step": 6045
+    },
+    {
+      "epoch": 0.36733701925997936,
+      "grad_norm": 0.19660969078540802,
+      "learning_rate": 7.05258442104602e-05,
+      "loss": 1.1057,
+      "step": 6046
+    },
+    {
+      "epoch": 0.3673977762926059,
+      "grad_norm": 0.2614695131778717,
+      "learning_rate": 7.051711487254734e-05,
+      "loss": 1.092,
+      "step": 6047
+    },
+    {
+      "epoch": 0.3674585333252324,
+      "grad_norm": 0.16886524856090546,
+      "learning_rate": 7.050838478257566e-05,
+      "loss": 1.0392,
+      "step": 6048
+    },
+    {
+      "epoch": 0.3675192903578589,
+      "grad_norm": 1.104804277420044,
+      "learning_rate": 7.049965394086518e-05,
+      "loss": 1.0514,
+      "step": 6049
+    },
+    {
+      "epoch": 0.36758004739048544,
+      "grad_norm": 0.18418674170970917,
+      "learning_rate": 7.04909223477359e-05,
+      "loss": 1.0766,
+      "step": 6050
+    },
+    {
+      "epoch": 0.367640804423112,
+      "grad_norm": 0.18491162359714508,
+      "learning_rate": 7.048219000350792e-05,
+      "loss": 1.1927,
+      "step": 6051
+    },
+    {
+      "epoch": 0.3677015614557385,
+      "grad_norm": 0.27615538239479065,
+      "learning_rate": 7.04734569085013e-05,
+      "loss": 1.0177,
+      "step": 6052
+    },
+    {
+      "epoch": 0.36776231848836505,
+      "grad_norm": 0.22296200692653656,
+      "learning_rate": 7.046472306303616e-05,
+      "loss": 1.1007,
+      "step": 6053
+    },
+    {
+      "epoch": 0.3678230755209915,
+      "grad_norm": 0.22059759497642517,
+      "learning_rate": 7.045598846743264e-05,
+      "loss": 1.2174,
+      "step": 6054
+    },
+    {
+      "epoch": 0.36788383255361806,
+      "grad_norm": 0.5893834829330444,
+      "learning_rate": 7.04472531220109e-05,
+      "loss": 1.3108,
+      "step": 6055
+    },
+    {
+      "epoch": 0.3679445895862446,
+      "grad_norm": 0.2247239053249359,
+      "learning_rate": 7.043851702709115e-05,
+      "loss": 1.141,
+      "step": 6056
+    },
+    {
+      "epoch": 0.36800534661887113,
+      "grad_norm": 0.3982189893722534,
+      "learning_rate": 7.042978018299361e-05,
+      "loss": 1.1064,
+      "step": 6057
+    },
+    {
+      "epoch": 0.36806610365149767,
+      "grad_norm": 0.2997436225414276,
+      "learning_rate": 7.042104259003853e-05,
+      "loss": 1.1572,
+      "step": 6058
+    },
+    {
+      "epoch": 0.3681268606841242,
+      "grad_norm": 0.20508989691734314,
+      "learning_rate": 7.041230424854619e-05,
+      "loss": 1.0877,
+      "step": 6059
+    },
+    {
+      "epoch": 0.36818761771675074,
+      "grad_norm": 0.1397322565317154,
+      "learning_rate": 7.04035651588369e-05,
+      "loss": 1.0579,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3682483747493772,
+      "grad_norm": 0.1818924993276596,
+      "learning_rate": 7.039482532123096e-05,
+      "loss": 1.1172,
+      "step": 6061
+    },
+    {
+      "epoch": 0.36830913178200375,
+      "grad_norm": 0.5817762613296509,
+      "learning_rate": 7.038608473604877e-05,
+      "loss": 1.1744,
+      "step": 6062
+    },
+    {
+      "epoch": 0.3683698888146303,
+      "grad_norm": 1.063594937324524,
+      "learning_rate": 7.037734340361069e-05,
+      "loss": 1.1224,
+      "step": 6063
+    },
+    {
+      "epoch": 0.3684306458472568,
+      "grad_norm": 0.15738733112812042,
+      "learning_rate": 7.036860132423718e-05,
+      "loss": 1.3237,
+      "step": 6064
+    },
+    {
+      "epoch": 0.36849140287988336,
+      "grad_norm": 0.15839505195617676,
+      "learning_rate": 7.035985849824863e-05,
+      "loss": 1.0783,
+      "step": 6065
+    },
+    {
+      "epoch": 0.3685521599125099,
+      "grad_norm": 0.19259479641914368,
+      "learning_rate": 7.035111492596552e-05,
+      "loss": 1.0934,
+      "step": 6066
+    },
+    {
+      "epoch": 0.3686129169451364,
+      "grad_norm": 0.3592526316642761,
+      "learning_rate": 7.034237060770838e-05,
+      "loss": 1.1451,
+      "step": 6067
+    },
+    {
+      "epoch": 0.3686736739777629,
+      "grad_norm": 0.6591197848320007,
+      "learning_rate": 7.03336255437977e-05,
+      "loss": 1.0937,
+      "step": 6068
+    },
+    {
+      "epoch": 0.36873443101038944,
+      "grad_norm": 0.15160061419010162,
+      "learning_rate": 7.032487973455406e-05,
+      "loss": 1.1105,
+      "step": 6069
+    },
+    {
+      "epoch": 0.368795188043016,
+      "grad_norm": 0.20214974880218506,
+      "learning_rate": 7.031613318029802e-05,
+      "loss": 1.0484,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3688559450756425,
+      "grad_norm": 0.24280185997486115,
+      "learning_rate": 7.030738588135018e-05,
+      "loss": 1.0632,
+      "step": 6071
+    },
+    {
+      "epoch": 0.36891670210826905,
+      "grad_norm": 0.15089461207389832,
+      "learning_rate": 7.02986378380312e-05,
+      "loss": 1.0494,
+      "step": 6072
+    },
+    {
+      "epoch": 0.3689774591408956,
+      "grad_norm": 0.20249488949775696,
+      "learning_rate": 7.028988905066171e-05,
+      "loss": 1.078,
+      "step": 6073
+    },
+    {
+      "epoch": 0.36903821617352206,
+      "grad_norm": 0.15266983211040497,
+      "learning_rate": 7.028113951956242e-05,
+      "loss": 1.0337,
+      "step": 6074
+    },
+    {
+      "epoch": 0.3690989732061486,
+      "grad_norm": 0.13306894898414612,
+      "learning_rate": 7.027238924505404e-05,
+      "loss": 1.0445,
+      "step": 6075
+    },
+    {
+      "epoch": 0.36915973023877513,
+      "grad_norm": 0.17093436419963837,
+      "learning_rate": 7.026363822745732e-05,
+      "loss": 1.1148,
+      "step": 6076
+    },
+    {
+      "epoch": 0.36922048727140167,
+      "grad_norm": 0.22219501435756683,
+      "learning_rate": 7.025488646709302e-05,
+      "loss": 1.1146,
+      "step": 6077
+    },
+    {
+      "epoch": 0.3692812443040282,
+      "grad_norm": 0.21313917636871338,
+      "learning_rate": 7.024613396428195e-05,
+      "loss": 1.1769,
+      "step": 6078
+    },
+    {
+      "epoch": 0.36934200133665473,
+      "grad_norm": 0.23730236291885376,
+      "learning_rate": 7.023738071934491e-05,
+      "loss": 1.0662,
+      "step": 6079
+    },
+    {
+      "epoch": 0.36940275836928127,
+      "grad_norm": 0.23709143698215485,
+      "learning_rate": 7.022862673260278e-05,
+      "loss": 1.0675,
+      "step": 6080
+    },
+    {
+      "epoch": 0.36946351540190775,
+      "grad_norm": 0.14758968353271484,
+      "learning_rate": 7.021987200437642e-05,
+      "loss": 1.0876,
+      "step": 6081
+    },
+    {
+      "epoch": 0.3695242724345343,
+      "grad_norm": 0.26542794704437256,
+      "learning_rate": 7.021111653498675e-05,
+      "loss": 1.0414,
+      "step": 6082
+    },
+    {
+      "epoch": 0.3695850294671608,
+      "grad_norm": 0.21081890165805817,
+      "learning_rate": 7.02023603247547e-05,
+      "loss": 1.0966,
+      "step": 6083
+    },
+    {
+      "epoch": 0.36964578649978735,
+      "grad_norm": 0.3085448443889618,
+      "learning_rate": 7.019360337400122e-05,
+      "loss": 1.3571,
+      "step": 6084
+    },
+    {
+      "epoch": 0.3697065435324139,
+      "grad_norm": 0.15007925033569336,
+      "learning_rate": 7.01848456830473e-05,
+      "loss": 1.0917,
+      "step": 6085
+    },
+    {
+      "epoch": 0.3697673005650404,
+      "grad_norm": 0.1844521462917328,
+      "learning_rate": 7.017608725221397e-05,
+      "loss": 1.138,
+      "step": 6086
+    },
+    {
+      "epoch": 0.3698280575976669,
+      "grad_norm": 0.12841415405273438,
+      "learning_rate": 7.016732808182224e-05,
+      "loss": 1.0582,
+      "step": 6087
+    },
+    {
+      "epoch": 0.36988881463029344,
+      "grad_norm": 0.3314495384693146,
+      "learning_rate": 7.015856817219324e-05,
+      "loss": 1.2464,
+      "step": 6088
+    },
+    {
+      "epoch": 0.36994957166292,
+      "grad_norm": 0.15241481363773346,
+      "learning_rate": 7.014980752364799e-05,
+      "loss": 1.0571,
+      "step": 6089
+    },
+    {
+      "epoch": 0.3700103286955465,
+      "grad_norm": 0.5650461316108704,
+      "learning_rate": 7.014104613650766e-05,
+      "loss": 1.1379,
+      "step": 6090
+    },
+    {
+      "epoch": 0.37007108572817304,
+      "grad_norm": 0.16181419789791107,
+      "learning_rate": 7.013228401109341e-05,
+      "loss": 1.042,
+      "step": 6091
+    },
+    {
+      "epoch": 0.3701318427607996,
+      "grad_norm": 0.12752079963684082,
+      "learning_rate": 7.012352114772638e-05,
+      "loss": 1.047,
+      "step": 6092
+    },
+    {
+      "epoch": 0.3701925997934261,
+      "grad_norm": 0.20489177107810974,
+      "learning_rate": 7.011475754672781e-05,
+      "loss": 1.2822,
+      "step": 6093
+    },
+    {
+      "epoch": 0.3702533568260526,
+      "grad_norm": 0.13086888194084167,
+      "learning_rate": 7.010599320841889e-05,
+      "loss": 1.0124,
+      "step": 6094
+    },
+    {
+      "epoch": 0.37031411385867913,
+      "grad_norm": 0.2657249867916107,
+      "learning_rate": 7.009722813312092e-05,
+      "loss": 1.1187,
+      "step": 6095
+    },
+    {
+      "epoch": 0.37037487089130566,
+      "grad_norm": 0.14157021045684814,
+      "learning_rate": 7.008846232115515e-05,
+      "loss": 1.1032,
+      "step": 6096
+    },
+    {
+      "epoch": 0.3704356279239322,
+      "grad_norm": 0.19336383044719696,
+      "learning_rate": 7.007969577284292e-05,
+      "loss": 1.1428,
+      "step": 6097
+    },
+    {
+      "epoch": 0.37049638495655873,
+      "grad_norm": 0.11828788369894028,
+      "learning_rate": 7.007092848850559e-05,
+      "loss": 1.0689,
+      "step": 6098
+    },
+    {
+      "epoch": 0.37055714198918527,
+      "grad_norm": 0.24823912978172302,
+      "learning_rate": 7.006216046846444e-05,
+      "loss": 1.0898,
+      "step": 6099
+    },
+    {
+      "epoch": 0.37061789902181175,
+      "grad_norm": 0.2931252121925354,
+      "learning_rate": 7.005339171304095e-05,
+      "loss": 1.0534,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3706786560544383,
+      "grad_norm": 0.14766299724578857,
+      "learning_rate": 7.004462222255653e-05,
+      "loss": 1.0924,
+      "step": 6101
+    },
+    {
+      "epoch": 0.3707394130870648,
+      "grad_norm": 0.14781631529331207,
+      "learning_rate": 7.003585199733258e-05,
+      "loss": 1.0715,
+      "step": 6102
+    },
+    {
+      "epoch": 0.37080017011969135,
+      "grad_norm": 0.1661837100982666,
+      "learning_rate": 7.00270810376906e-05,
+      "loss": 1.0693,
+      "step": 6103
+    },
+    {
+      "epoch": 0.3708609271523179,
+      "grad_norm": 0.157626673579216,
+      "learning_rate": 7.00183093439521e-05,
+      "loss": 1.0911,
+      "step": 6104
+    },
+    {
+      "epoch": 0.3709216841849444,
+      "grad_norm": 0.1764076054096222,
+      "learning_rate": 7.000953691643861e-05,
+      "loss": 1.1437,
+      "step": 6105
+    },
+    {
+      "epoch": 0.37098244121757096,
+      "grad_norm": 0.18832449615001678,
+      "learning_rate": 7.000076375547168e-05,
+      "loss": 1.048,
+      "step": 6106
+    },
+    {
+      "epoch": 0.37104319825019744,
+      "grad_norm": 0.3083008825778961,
+      "learning_rate": 6.999198986137287e-05,
+      "loss": 1.1835,
+      "step": 6107
+    },
+    {
+      "epoch": 0.37110395528282397,
+      "grad_norm": 0.13549351692199707,
+      "learning_rate": 6.998321523446383e-05,
+      "loss": 1.05,
+      "step": 6108
+    },
+    {
+      "epoch": 0.3711647123154505,
+      "grad_norm": 0.3009432256221771,
+      "learning_rate": 6.997443987506616e-05,
+      "loss": 1.2666,
+      "step": 6109
+    },
+    {
+      "epoch": 0.37122546934807704,
+      "grad_norm": 0.16966791450977325,
+      "learning_rate": 6.996566378350152e-05,
+      "loss": 1.1162,
+      "step": 6110
+    },
+    {
+      "epoch": 0.3712862263807036,
+      "grad_norm": 0.1756017506122589,
+      "learning_rate": 6.995688696009164e-05,
+      "loss": 1.2071,
+      "step": 6111
+    },
+    {
+      "epoch": 0.3713469834133301,
+      "grad_norm": 2.539581060409546,
+      "learning_rate": 6.994810940515819e-05,
+      "loss": 1.1902,
+      "step": 6112
+    },
+    {
+      "epoch": 0.37140774044595665,
+      "grad_norm": 0.3390590250492096,
+      "learning_rate": 6.993933111902296e-05,
+      "loss": 1.2008,
+      "step": 6113
+    },
+    {
+      "epoch": 0.3714684974785831,
+      "grad_norm": 0.21464766561985016,
+      "learning_rate": 6.993055210200768e-05,
+      "loss": 1.085,
+      "step": 6114
+    },
+    {
+      "epoch": 0.37152925451120966,
+      "grad_norm": 0.470664381980896,
+      "learning_rate": 6.992177235443416e-05,
+      "loss": 1.2072,
+      "step": 6115
+    },
+    {
+      "epoch": 0.3715900115438362,
+      "grad_norm": 0.2351076900959015,
+      "learning_rate": 6.991299187662421e-05,
+      "loss": 1.1909,
+      "step": 6116
+    },
+    {
+      "epoch": 0.37165076857646273,
+      "grad_norm": 0.19528645277023315,
+      "learning_rate": 6.990421066889971e-05,
+      "loss": 1.0534,
+      "step": 6117
+    },
+    {
+      "epoch": 0.37171152560908927,
+      "grad_norm": 2.4312665462493896,
+      "learning_rate": 6.989542873158252e-05,
+      "loss": 1.1163,
+      "step": 6118
+    },
+    {
+      "epoch": 0.3717722826417158,
+      "grad_norm": 0.18483975529670715,
+      "learning_rate": 6.988664606499454e-05,
+      "loss": 1.1588,
+      "step": 6119
+    },
+    {
+      "epoch": 0.3718330396743423,
+      "grad_norm": 0.2241080403327942,
+      "learning_rate": 6.98778626694577e-05,
+      "loss": 1.0895,
+      "step": 6120
+    },
+    {
+      "epoch": 0.3718937967069688,
+      "grad_norm": 0.1853787750005722,
+      "learning_rate": 6.986907854529397e-05,
+      "loss": 1.0141,
+      "step": 6121
+    },
+    {
+      "epoch": 0.37195455373959535,
+      "grad_norm": 0.2961101233959198,
+      "learning_rate": 6.986029369282533e-05,
+      "loss": 1.1629,
+      "step": 6122
+    },
+    {
+      "epoch": 0.3720153107722219,
+      "grad_norm": 0.2024242877960205,
+      "learning_rate": 6.985150811237377e-05,
+      "loss": 1.0389,
+      "step": 6123
+    },
+    {
+      "epoch": 0.3720760678048484,
+      "grad_norm": 0.1818472295999527,
+      "learning_rate": 6.984272180426134e-05,
+      "loss": 1.0895,
+      "step": 6124
+    },
+    {
+      "epoch": 0.37213682483747496,
+      "grad_norm": 0.16849341988563538,
+      "learning_rate": 6.983393476881013e-05,
+      "loss": 1.0816,
+      "step": 6125
+    },
+    {
+      "epoch": 0.3721975818701015,
+      "grad_norm": 0.21810662746429443,
+      "learning_rate": 6.98251470063422e-05,
+      "loss": 1.0274,
+      "step": 6126
+    },
+    {
+      "epoch": 0.37225833890272797,
+      "grad_norm": 0.3145095109939575,
+      "learning_rate": 6.981635851717966e-05,
+      "loss": 1.2075,
+      "step": 6127
+    },
+    {
+      "epoch": 0.3723190959353545,
+      "grad_norm": 0.30857306718826294,
+      "learning_rate": 6.980756930164468e-05,
+      "loss": 1.1858,
+      "step": 6128
+    },
+    {
+      "epoch": 0.37237985296798104,
+      "grad_norm": 0.3091966509819031,
+      "learning_rate": 6.979877936005942e-05,
+      "loss": 1.1373,
+      "step": 6129
+    },
+    {
+      "epoch": 0.3724406100006076,
+      "grad_norm": 0.2408003807067871,
+      "learning_rate": 6.978998869274606e-05,
+      "loss": 1.0775,
+      "step": 6130
+    },
+    {
+      "epoch": 0.3725013670332341,
+      "grad_norm": 0.13595938682556152,
+      "learning_rate": 6.978119730002686e-05,
+      "loss": 1.0285,
+      "step": 6131
+    },
+    {
+      "epoch": 0.37256212406586064,
+      "grad_norm": 0.16466747224330902,
+      "learning_rate": 6.977240518222403e-05,
+      "loss": 1.0682,
+      "step": 6132
+    },
+    {
+      "epoch": 0.3726228810984871,
+      "grad_norm": 0.28831911087036133,
+      "learning_rate": 6.976361233965988e-05,
+      "loss": 1.1288,
+      "step": 6133
+    },
+    {
+      "epoch": 0.37268363813111366,
+      "grad_norm": 0.236700639128685,
+      "learning_rate": 6.97548187726567e-05,
+      "loss": 1.1594,
+      "step": 6134
+    },
+    {
+      "epoch": 0.3727443951637402,
+      "grad_norm": 0.12770244479179382,
+      "learning_rate": 6.974602448153683e-05,
+      "loss": 1.0815,
+      "step": 6135
+    },
+    {
+      "epoch": 0.37280515219636673,
+      "grad_norm": 0.1409512460231781,
+      "learning_rate": 6.973722946662258e-05,
+      "loss": 1.1057,
+      "step": 6136
+    },
+    {
+      "epoch": 0.37286590922899326,
+      "grad_norm": 0.1878613382577896,
+      "learning_rate": 6.972843372823638e-05,
+      "loss": 1.1584,
+      "step": 6137
+    },
+    {
+      "epoch": 0.3729266662616198,
+      "grad_norm": 0.20101282000541687,
+      "learning_rate": 6.971963726670065e-05,
+      "loss": 1.0868,
+      "step": 6138
+    },
+    {
+      "epoch": 0.37298742329424633,
+      "grad_norm": 0.17033053934574127,
+      "learning_rate": 6.97108400823378e-05,
+      "loss": 1.0962,
+      "step": 6139
+    },
+    {
+      "epoch": 0.3730481803268728,
+      "grad_norm": 0.18454821407794952,
+      "learning_rate": 6.970204217547027e-05,
+      "loss": 1.1581,
+      "step": 6140
+    },
+    {
+      "epoch": 0.37310893735949935,
+      "grad_norm": 0.22834104299545288,
+      "learning_rate": 6.96932435464206e-05,
+      "loss": 1.0749,
+      "step": 6141
+    },
+    {
+      "epoch": 0.3731696943921259,
+      "grad_norm": 0.15021520853042603,
+      "learning_rate": 6.968444419551128e-05,
+      "loss": 1.1433,
+      "step": 6142
+    },
+    {
+      "epoch": 0.3732304514247524,
+      "grad_norm": 0.2514025866985321,
+      "learning_rate": 6.967564412306485e-05,
+      "loss": 0.9998,
+      "step": 6143
+    },
+    {
+      "epoch": 0.37329120845737895,
+      "grad_norm": 0.18249507248401642,
+      "learning_rate": 6.966684332940388e-05,
+      "loss": 1.1168,
+      "step": 6144
+    },
+    {
+      "epoch": 0.3733519654900055,
+      "grad_norm": 0.20229239761829376,
+      "learning_rate": 6.965804181485097e-05,
+      "loss": 1.0933,
+      "step": 6145
+    },
+    {
+      "epoch": 0.37341272252263197,
+      "grad_norm": 7.055782318115234,
+      "learning_rate": 6.964923957972872e-05,
+      "loss": 1.0567,
+      "step": 6146
+    },
+    {
+      "epoch": 0.3734734795552585,
+      "grad_norm": 0.18048037588596344,
+      "learning_rate": 6.964043662435981e-05,
+      "loss": 1.1217,
+      "step": 6147
+    },
+    {
+      "epoch": 0.37353423658788504,
+      "grad_norm": 0.25209271907806396,
+      "learning_rate": 6.963163294906691e-05,
+      "loss": 1.0352,
+      "step": 6148
+    },
+    {
+      "epoch": 0.3735949936205116,
+      "grad_norm": 0.17877323925495148,
+      "learning_rate": 6.962282855417269e-05,
+      "loss": 1.1214,
+      "step": 6149
+    },
+    {
+      "epoch": 0.3736557506531381,
+      "grad_norm": 0.33786898851394653,
+      "learning_rate": 6.96140234399999e-05,
+      "loss": 1.0787,
+      "step": 6150
+    },
+    {
+      "epoch": 0.37371650768576464,
+      "grad_norm": 0.25326746702194214,
+      "learning_rate": 6.960521760687129e-05,
+      "loss": 1.0764,
+      "step": 6151
+    },
+    {
+      "epoch": 0.3737772647183912,
+      "grad_norm": 0.1220151036977768,
+      "learning_rate": 6.959641105510966e-05,
+      "loss": 1.0211,
+      "step": 6152
+    },
+    {
+      "epoch": 0.37383802175101766,
+      "grad_norm": 0.18178948760032654,
+      "learning_rate": 6.958760378503776e-05,
+      "loss": 1.0548,
+      "step": 6153
+    },
+    {
+      "epoch": 0.3738987787836442,
+      "grad_norm": 0.17142456769943237,
+      "learning_rate": 6.957879579697848e-05,
+      "loss": 1.1817,
+      "step": 6154
+    },
+    {
+      "epoch": 0.3739595358162707,
+      "grad_norm": 0.21144050359725952,
+      "learning_rate": 6.956998709125467e-05,
+      "loss": 1.1075,
+      "step": 6155
+    },
+    {
+      "epoch": 0.37402029284889726,
+      "grad_norm": 4.238260746002197,
+      "learning_rate": 6.956117766818915e-05,
+      "loss": 1.0603,
+      "step": 6156
+    },
+    {
+      "epoch": 0.3740810498815238,
+      "grad_norm": 0.3379259705543518,
+      "learning_rate": 6.955236752810493e-05,
+      "loss": 1.078,
+      "step": 6157
+    },
+    {
+      "epoch": 0.37414180691415033,
+      "grad_norm": 0.27843043208122253,
+      "learning_rate": 6.954355667132489e-05,
+      "loss": 1.2243,
+      "step": 6158
+    },
+    {
+      "epoch": 0.3742025639467768,
+      "grad_norm": 0.2400258481502533,
+      "learning_rate": 6.953474509817197e-05,
+      "loss": 1.1211,
+      "step": 6159
+    },
+    {
+      "epoch": 0.37426332097940335,
+      "grad_norm": 0.35528701543807983,
+      "learning_rate": 6.952593280896922e-05,
+      "loss": 1.0683,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3743240780120299,
+      "grad_norm": 0.1909889131784439,
+      "learning_rate": 6.951711980403963e-05,
+      "loss": 1.0956,
+      "step": 6161
+    },
+    {
+      "epoch": 0.3743848350446564,
+      "grad_norm": 0.24124369025230408,
+      "learning_rate": 6.950830608370623e-05,
+      "loss": 1.1332,
+      "step": 6162
+    },
+    {
+      "epoch": 0.37444559207728295,
+      "grad_norm": 0.29295313358306885,
+      "learning_rate": 6.94994916482921e-05,
+      "loss": 1.2311,
+      "step": 6163
+    },
+    {
+      "epoch": 0.3745063491099095,
+      "grad_norm": 0.13519087433815002,
+      "learning_rate": 6.949067649812033e-05,
+      "loss": 1.0519,
+      "step": 6164
+    },
+    {
+      "epoch": 0.374567106142536,
+      "grad_norm": 0.45620399713516235,
+      "learning_rate": 6.948186063351407e-05,
+      "loss": 1.3742,
+      "step": 6165
+    },
+    {
+      "epoch": 0.3746278631751625,
+      "grad_norm": 0.1934356540441513,
+      "learning_rate": 6.947304405479644e-05,
+      "loss": 1.0893,
+      "step": 6166
+    },
+    {
+      "epoch": 0.37468862020778904,
+      "grad_norm": 0.32619509100914,
+      "learning_rate": 6.946422676229058e-05,
+      "loss": 1.231,
+      "step": 6167
+    },
+    {
+      "epoch": 0.37474937724041557,
+      "grad_norm": 3.7163872718811035,
+      "learning_rate": 6.945540875631977e-05,
+      "loss": 1.1135,
+      "step": 6168
+    },
+    {
+      "epoch": 0.3748101342730421,
+      "grad_norm": 0.19481046497821808,
+      "learning_rate": 6.944659003720716e-05,
+      "loss": 1.0907,
+      "step": 6169
+    },
+    {
+      "epoch": 0.37487089130566864,
+      "grad_norm": 0.18538278341293335,
+      "learning_rate": 6.943777060527605e-05,
+      "loss": 1.1023,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3749316483382952,
+      "grad_norm": 0.1559235006570816,
+      "learning_rate": 6.94289504608497e-05,
+      "loss": 1.0195,
+      "step": 6171
+    },
+    {
+      "epoch": 0.3749924053709217,
+      "grad_norm": 0.17050620913505554,
+      "learning_rate": 6.94201296042514e-05,
+      "loss": 1.0486,
+      "step": 6172
+    },
+    {
+      "epoch": 0.3750531624035482,
+      "grad_norm": 0.17463983595371246,
+      "learning_rate": 6.941130803580451e-05,
+      "loss": 1.0774,
+      "step": 6173
+    },
+    {
+      "epoch": 0.3751139194361747,
+      "grad_norm": 0.16148193180561066,
+      "learning_rate": 6.940248575583236e-05,
+      "loss": 1.0274,
+      "step": 6174
+    },
+    {
+      "epoch": 0.37517467646880126,
+      "grad_norm": 0.14038436114788055,
+      "learning_rate": 6.939366276465836e-05,
+      "loss": 1.0671,
+      "step": 6175
+    },
+    {
+      "epoch": 0.3752354335014278,
+      "grad_norm": 0.17784066498279572,
+      "learning_rate": 6.938483906260589e-05,
+      "loss": 1.108,
+      "step": 6176
+    },
+    {
+      "epoch": 0.37529619053405433,
+      "grad_norm": 3.010763168334961,
+      "learning_rate": 6.93760146499984e-05,
+      "loss": 1.0644,
+      "step": 6177
+    },
+    {
+      "epoch": 0.37535694756668087,
+      "grad_norm": 0.15693561732769012,
+      "learning_rate": 6.936718952715936e-05,
+      "loss": 1.0948,
+      "step": 6178
+    },
+    {
+      "epoch": 0.37541770459930734,
+      "grad_norm": 0.21804840862751007,
+      "learning_rate": 6.935836369441223e-05,
+      "loss": 1.1942,
+      "step": 6179
+    },
+    {
+      "epoch": 0.3754784616319339,
+      "grad_norm": 0.295316219329834,
+      "learning_rate": 6.934953715208054e-05,
+      "loss": 1.2403,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3755392186645604,
+      "grad_norm": 0.13129793107509613,
+      "learning_rate": 6.934070990048784e-05,
+      "loss": 1.0554,
+      "step": 6181
+    },
+    {
+      "epoch": 0.37559997569718695,
+      "grad_norm": 0.19522185623645782,
+      "learning_rate": 6.933188193995766e-05,
+      "loss": 1.0771,
+      "step": 6182
+    },
+    {
+      "epoch": 0.3756607327298135,
+      "grad_norm": 0.15450729429721832,
+      "learning_rate": 6.932305327081361e-05,
+      "loss": 1.1766,
+      "step": 6183
+    },
+    {
+      "epoch": 0.37572148976244,
+      "grad_norm": 0.29695799946784973,
+      "learning_rate": 6.931422389337932e-05,
+      "loss": 1.1088,
+      "step": 6184
+    },
+    {
+      "epoch": 0.37578224679506655,
+      "grad_norm": 0.14404602348804474,
+      "learning_rate": 6.93053938079784e-05,
+      "loss": 1.0977,
+      "step": 6185
+    },
+    {
+      "epoch": 0.37584300382769303,
+      "grad_norm": 0.13579857349395752,
+      "learning_rate": 6.929656301493455e-05,
+      "loss": 1.0769,
+      "step": 6186
+    },
+    {
+      "epoch": 0.37590376086031957,
+      "grad_norm": 0.14122948050498962,
+      "learning_rate": 6.928773151457147e-05,
+      "loss": 1.0465,
+      "step": 6187
+    },
+    {
+      "epoch": 0.3759645178929461,
+      "grad_norm": 0.8193957805633545,
+      "learning_rate": 6.927889930721286e-05,
+      "loss": 1.1652,
+      "step": 6188
+    },
+    {
+      "epoch": 0.37602527492557264,
+      "grad_norm": 0.16174200177192688,
+      "learning_rate": 6.927006639318245e-05,
+      "loss": 1.2036,
+      "step": 6189
+    },
+    {
+      "epoch": 0.3760860319581992,
+      "grad_norm": 0.16232994198799133,
+      "learning_rate": 6.926123277280403e-05,
+      "loss": 1.0831,
+      "step": 6190
+    },
+    {
+      "epoch": 0.3761467889908257,
+      "grad_norm": 0.12621040642261505,
+      "learning_rate": 6.925239844640141e-05,
+      "loss": 1.0879,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3762075460234522,
+      "grad_norm": 0.17994782328605652,
+      "learning_rate": 6.92435634142984e-05,
+      "loss": 1.1079,
+      "step": 6192
+    },
+    {
+      "epoch": 0.3762683030560787,
+      "grad_norm": 0.15814127027988434,
+      "learning_rate": 6.923472767681886e-05,
+      "loss": 1.1201,
+      "step": 6193
+    },
+    {
+      "epoch": 0.37632906008870526,
+      "grad_norm": 0.17147469520568848,
+      "learning_rate": 6.922589123428665e-05,
+      "loss": 1.0973,
+      "step": 6194
+    },
+    {
+      "epoch": 0.3763898171213318,
+      "grad_norm": 0.17876200377941132,
+      "learning_rate": 6.92170540870257e-05,
+      "loss": 1.0564,
+      "step": 6195
+    },
+    {
+      "epoch": 0.37645057415395833,
+      "grad_norm": 0.2009342461824417,
+      "learning_rate": 6.920821623535989e-05,
+      "loss": 1.1544,
+      "step": 6196
+    },
+    {
+      "epoch": 0.37651133118658486,
+      "grad_norm": 0.1431100219488144,
+      "learning_rate": 6.919937767961322e-05,
+      "loss": 1.0564,
+      "step": 6197
+    },
+    {
+      "epoch": 0.3765720882192114,
+      "grad_norm": 0.18452730774879456,
+      "learning_rate": 6.919053842010964e-05,
+      "loss": 1.1519,
+      "step": 6198
+    },
+    {
+      "epoch": 0.3766328452518379,
+      "grad_norm": 0.31846579909324646,
+      "learning_rate": 6.918169845717318e-05,
+      "loss": 1.1067,
+      "step": 6199
+    },
+    {
+      "epoch": 0.3766936022844644,
+      "grad_norm": 0.1929236501455307,
+      "learning_rate": 6.917285779112784e-05,
+      "loss": 1.0778,
+      "step": 6200
+    },
+    {
+      "epoch": 0.37675435931709095,
+      "grad_norm": 0.4140916168689728,
+      "learning_rate": 6.916401642229769e-05,
+      "loss": 1.0852,
+      "step": 6201
+    },
+    {
+      "epoch": 0.3768151163497175,
+      "grad_norm": 0.17804791033267975,
+      "learning_rate": 6.915517435100683e-05,
+      "loss": 1.1071,
+      "step": 6202
+    },
+    {
+      "epoch": 0.376875873382344,
+      "grad_norm": 0.15515582263469696,
+      "learning_rate": 6.914633157757932e-05,
+      "loss": 1.1028,
+      "step": 6203
+    },
+    {
+      "epoch": 0.37693663041497055,
+      "grad_norm": 0.2886927127838135,
+      "learning_rate": 6.913748810233935e-05,
+      "loss": 1.0972,
+      "step": 6204
+    },
+    {
+      "epoch": 0.37699738744759703,
+      "grad_norm": 0.17595957219600677,
+      "learning_rate": 6.912864392561105e-05,
+      "loss": 1.1183,
+      "step": 6205
+    },
+    {
+      "epoch": 0.37705814448022357,
+      "grad_norm": 1.979278802871704,
+      "learning_rate": 6.911979904771862e-05,
+      "loss": 1.1344,
+      "step": 6206
+    },
+    {
+      "epoch": 0.3771189015128501,
+      "grad_norm": 0.22587543725967407,
+      "learning_rate": 6.911095346898624e-05,
+      "loss": 1.0265,
+      "step": 6207
+    },
+    {
+      "epoch": 0.37717965854547664,
+      "grad_norm": 0.22240546345710754,
+      "learning_rate": 6.910210718973818e-05,
+      "loss": 1.2302,
+      "step": 6208
+    },
+    {
+      "epoch": 0.37724041557810317,
+      "grad_norm": 0.14650443196296692,
+      "learning_rate": 6.90932602102987e-05,
+      "loss": 1.1186,
+      "step": 6209
+    },
+    {
+      "epoch": 0.3773011726107297,
+      "grad_norm": 0.1838058978319168,
+      "learning_rate": 6.908441253099205e-05,
+      "loss": 1.142,
+      "step": 6210
+    },
+    {
+      "epoch": 0.37736192964335624,
+      "grad_norm": 0.23207156360149384,
+      "learning_rate": 6.907556415214258e-05,
+      "loss": 1.0898,
+      "step": 6211
+    },
+    {
+      "epoch": 0.3774226866759827,
+      "grad_norm": 0.14599956572055817,
+      "learning_rate": 6.906671507407463e-05,
+      "loss": 1.0784,
+      "step": 6212
+    },
+    {
+      "epoch": 0.37748344370860926,
+      "grad_norm": 0.23296137154102325,
+      "learning_rate": 6.905786529711254e-05,
+      "loss": 1.0722,
+      "step": 6213
+    },
+    {
+      "epoch": 0.3775442007412358,
+      "grad_norm": 0.19169026613235474,
+      "learning_rate": 6.90490148215807e-05,
+      "loss": 1.2047,
+      "step": 6214
+    },
+    {
+      "epoch": 0.3776049577738623,
+      "grad_norm": 0.27794721722602844,
+      "learning_rate": 6.904016364780358e-05,
+      "loss": 1.0987,
+      "step": 6215
+    },
+    {
+      "epoch": 0.37766571480648886,
+      "grad_norm": 0.31110909581184387,
+      "learning_rate": 6.903131177610554e-05,
+      "loss": 1.0693,
+      "step": 6216
+    },
+    {
+      "epoch": 0.3777264718391154,
+      "grad_norm": 0.19506366550922394,
+      "learning_rate": 6.902245920681112e-05,
+      "loss": 1.0664,
+      "step": 6217
+    },
+    {
+      "epoch": 0.37778722887174193,
+      "grad_norm": 1.2798144817352295,
+      "learning_rate": 6.901360594024477e-05,
+      "loss": 1.1754,
+      "step": 6218
+    },
+    {
+      "epoch": 0.3778479859043684,
+      "grad_norm": 0.19944436848163605,
+      "learning_rate": 6.900475197673101e-05,
+      "loss": 1.09,
+      "step": 6219
+    },
+    {
+      "epoch": 0.37790874293699495,
+      "grad_norm": 0.1938268095254898,
+      "learning_rate": 6.89958973165944e-05,
+      "loss": 1.1191,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3779694999696215,
+      "grad_norm": 0.1337699145078659,
+      "learning_rate": 6.898704196015949e-05,
+      "loss": 1.0946,
+      "step": 6221
+    },
+    {
+      "epoch": 0.378030257002248,
+      "grad_norm": 0.1706361025571823,
+      "learning_rate": 6.897818590775092e-05,
+      "loss": 1.1659,
+      "step": 6222
+    },
+    {
+      "epoch": 0.37809101403487455,
+      "grad_norm": 0.20918989181518555,
+      "learning_rate": 6.896932915969325e-05,
+      "loss": 1.1382,
+      "step": 6223
+    },
+    {
+      "epoch": 0.3781517710675011,
+      "grad_norm": 1.891122579574585,
+      "learning_rate": 6.896047171631114e-05,
+      "loss": 1.2807,
+      "step": 6224
+    },
+    {
+      "epoch": 0.37821252810012757,
+      "grad_norm": 0.20246365666389465,
+      "learning_rate": 6.895161357792931e-05,
+      "loss": 1.1135,
+      "step": 6225
+    },
+    {
+      "epoch": 0.3782732851327541,
+      "grad_norm": 0.16818103194236755,
+      "learning_rate": 6.89427547448724e-05,
+      "loss": 1.0715,
+      "step": 6226
+    },
+    {
+      "epoch": 0.37833404216538064,
+      "grad_norm": 0.19742870330810547,
+      "learning_rate": 6.893389521746515e-05,
+      "loss": 1.0905,
+      "step": 6227
+    },
+    {
+      "epoch": 0.37839479919800717,
+      "grad_norm": 0.14059999585151672,
+      "learning_rate": 6.892503499603233e-05,
+      "loss": 1.047,
+      "step": 6228
+    },
+    {
+      "epoch": 0.3784555562306337,
+      "grad_norm": 0.26940542459487915,
+      "learning_rate": 6.891617408089866e-05,
+      "loss": 1.1264,
+      "step": 6229
+    },
+    {
+      "epoch": 0.37851631326326024,
+      "grad_norm": 0.1259632408618927,
+      "learning_rate": 6.8907312472389e-05,
+      "loss": 1.0456,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3785770702958868,
+      "grad_norm": 0.17143326997756958,
+      "learning_rate": 6.889845017082815e-05,
+      "loss": 1.1082,
+      "step": 6231
+    },
+    {
+      "epoch": 0.37863782732851325,
+      "grad_norm": 0.17943157255649567,
+      "learning_rate": 6.888958717654095e-05,
+      "loss": 1.0741,
+      "step": 6232
+    },
+    {
+      "epoch": 0.3786985843611398,
+      "grad_norm": 0.22239373624324799,
+      "learning_rate": 6.888072348985227e-05,
+      "loss": 1.0426,
+      "step": 6233
+    },
+    {
+      "epoch": 0.3787593413937663,
+      "grad_norm": 0.19490393996238708,
+      "learning_rate": 6.887185911108702e-05,
+      "loss": 1.098,
+      "step": 6234
+    },
+    {
+      "epoch": 0.37882009842639286,
+      "grad_norm": 0.15172097086906433,
+      "learning_rate": 6.886299404057015e-05,
+      "loss": 1.0932,
+      "step": 6235
+    },
+    {
+      "epoch": 0.3788808554590194,
+      "grad_norm": 0.2695910334587097,
+      "learning_rate": 6.885412827862657e-05,
+      "loss": 1.1831,
+      "step": 6236
+    },
+    {
+      "epoch": 0.37894161249164593,
+      "grad_norm": 0.17541678249835968,
+      "learning_rate": 6.884526182558126e-05,
+      "loss": 1.157,
+      "step": 6237
+    },
+    {
+      "epoch": 0.3790023695242724,
+      "grad_norm": 0.23607981204986572,
+      "learning_rate": 6.883639468175927e-05,
+      "loss": 1.0669,
+      "step": 6238
+    },
+    {
+      "epoch": 0.37906312655689894,
+      "grad_norm": 0.14976032078266144,
+      "learning_rate": 6.882752684748556e-05,
+      "loss": 1.1239,
+      "step": 6239
+    },
+    {
+      "epoch": 0.3791238835895255,
+      "grad_norm": 3.4635632038116455,
+      "learning_rate": 6.881865832308523e-05,
+      "loss": 1.1176,
+      "step": 6240
+    },
+    {
+      "epoch": 0.379184640622152,
+      "grad_norm": 0.20082558691501617,
+      "learning_rate": 6.880978910888334e-05,
+      "loss": 1.1138,
+      "step": 6241
+    },
+    {
+      "epoch": 0.37924539765477855,
+      "grad_norm": 0.43915224075317383,
+      "learning_rate": 6.8800919205205e-05,
+      "loss": 1.1007,
+      "step": 6242
+    },
+    {
+      "epoch": 0.3793061546874051,
+      "grad_norm": 0.17540255188941956,
+      "learning_rate": 6.879204861237532e-05,
+      "loss": 1.0648,
+      "step": 6243
+    },
+    {
+      "epoch": 0.3793669117200316,
+      "grad_norm": 0.3072577714920044,
+      "learning_rate": 6.878317733071947e-05,
+      "loss": 1.0377,
+      "step": 6244
+    },
+    {
+      "epoch": 0.3794276687526581,
+      "grad_norm": 0.13999001681804657,
+      "learning_rate": 6.877430536056263e-05,
+      "loss": 1.0418,
+      "step": 6245
+    },
+    {
+      "epoch": 0.37948842578528463,
+      "grad_norm": 0.263920396566391,
+      "learning_rate": 6.876543270222999e-05,
+      "loss": 1.1796,
+      "step": 6246
+    },
+    {
+      "epoch": 0.37954918281791117,
+      "grad_norm": 0.212153822183609,
+      "learning_rate": 6.875655935604678e-05,
+      "loss": 1.0772,
+      "step": 6247
+    },
+    {
+      "epoch": 0.3796099398505377,
+      "grad_norm": 2.7500085830688477,
+      "learning_rate": 6.874768532233828e-05,
+      "loss": 1.064,
+      "step": 6248
+    },
+    {
+      "epoch": 0.37967069688316424,
+      "grad_norm": 0.21294887363910675,
+      "learning_rate": 6.873881060142973e-05,
+      "loss": 1.1994,
+      "step": 6249
+    },
+    {
+      "epoch": 0.3797314539157908,
+      "grad_norm": 0.2289019525051117,
+      "learning_rate": 6.872993519364648e-05,
+      "loss": 1.0736,
+      "step": 6250
+    },
+    {
+      "epoch": 0.37979221094841725,
+      "grad_norm": 0.18465881049633026,
+      "learning_rate": 6.872105909931382e-05,
+      "loss": 1.0748,
+      "step": 6251
+    },
+    {
+      "epoch": 0.3798529679810438,
+      "grad_norm": 0.12557293474674225,
+      "learning_rate": 6.871218231875713e-05,
+      "loss": 1.0902,
+      "step": 6252
+    },
+    {
+      "epoch": 0.3799137250136703,
+      "grad_norm": 0.28713440895080566,
+      "learning_rate": 6.870330485230176e-05,
+      "loss": 1.1982,
+      "step": 6253
+    },
+    {
+      "epoch": 0.37997448204629686,
+      "grad_norm": 0.2273520678281784,
+      "learning_rate": 6.869442670027315e-05,
+      "loss": 1.0376,
+      "step": 6254
+    },
+    {
+      "epoch": 0.3800352390789234,
+      "grad_norm": 0.21783767640590668,
+      "learning_rate": 6.868554786299672e-05,
+      "loss": 1.1128,
+      "step": 6255
+    },
+    {
+      "epoch": 0.3800959961115499,
+      "grad_norm": 0.2760343551635742,
+      "learning_rate": 6.867666834079791e-05,
+      "loss": 1.04,
+      "step": 6256
+    },
+    {
+      "epoch": 0.38015675314417646,
+      "grad_norm": 0.3612003028392792,
+      "learning_rate": 6.86677881340022e-05,
+      "loss": 1.0864,
+      "step": 6257
+    },
+    {
+      "epoch": 0.38021751017680294,
+      "grad_norm": 0.5285005569458008,
+      "learning_rate": 6.865890724293515e-05,
+      "loss": 1.0548,
+      "step": 6258
+    },
+    {
+      "epoch": 0.3802782672094295,
+      "grad_norm": 0.22248300909996033,
+      "learning_rate": 6.865002566792223e-05,
+      "loss": 1.133,
+      "step": 6259
+    },
+    {
+      "epoch": 0.380339024242056,
+      "grad_norm": 0.1370825320482254,
+      "learning_rate": 6.864114340928898e-05,
+      "loss": 1.0439,
+      "step": 6260
+    },
+    {
+      "epoch": 0.38039978127468255,
+      "grad_norm": 0.42779216170310974,
+      "learning_rate": 6.863226046736105e-05,
+      "loss": 1.0953,
+      "step": 6261
+    },
+    {
+      "epoch": 0.3804605383073091,
+      "grad_norm": 0.16123263537883759,
+      "learning_rate": 6.862337684246401e-05,
+      "loss": 1.0761,
+      "step": 6262
+    },
+    {
+      "epoch": 0.3805212953399356,
+      "grad_norm": 0.15944361686706543,
+      "learning_rate": 6.861449253492348e-05,
+      "loss": 1.1203,
+      "step": 6263
+    },
+    {
+      "epoch": 0.38058205237256215,
+      "grad_norm": 0.13039962947368622,
+      "learning_rate": 6.860560754506513e-05,
+      "loss": 1.0614,
+      "step": 6264
+    },
+    {
+      "epoch": 0.38064280940518863,
+      "grad_norm": 0.32936879992485046,
+      "learning_rate": 6.859672187321464e-05,
+      "loss": 1.1507,
+      "step": 6265
+    },
+    {
+      "epoch": 0.38070356643781517,
+      "grad_norm": 0.17140968143939972,
+      "learning_rate": 6.858783551969773e-05,
+      "loss": 1.0751,
+      "step": 6266
+    },
+    {
+      "epoch": 0.3807643234704417,
+      "grad_norm": 0.19645406305789948,
+      "learning_rate": 6.85789484848401e-05,
+      "loss": 1.1237,
+      "step": 6267
+    },
+    {
+      "epoch": 0.38082508050306824,
+      "grad_norm": 0.14797905087471008,
+      "learning_rate": 6.857006076896751e-05,
+      "loss": 1.082,
+      "step": 6268
+    },
+    {
+      "epoch": 0.38088583753569477,
+      "grad_norm": 0.34342846274375916,
+      "learning_rate": 6.856117237240577e-05,
+      "loss": 1.2068,
+      "step": 6269
+    },
+    {
+      "epoch": 0.3809465945683213,
+      "grad_norm": 0.1380472481250763,
+      "learning_rate": 6.855228329548067e-05,
+      "loss": 1.0892,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3810073516009478,
+      "grad_norm": 0.48161420226097107,
+      "learning_rate": 6.854339353851804e-05,
+      "loss": 1.2117,
+      "step": 6271
+    },
+    {
+      "epoch": 0.3810681086335743,
+      "grad_norm": 0.11848028004169464,
+      "learning_rate": 6.853450310184375e-05,
+      "loss": 1.0409,
+      "step": 6272
+    },
+    {
+      "epoch": 0.38112886566620086,
+      "grad_norm": 0.24575193226337433,
+      "learning_rate": 6.852561198578364e-05,
+      "loss": 1.0887,
+      "step": 6273
+    },
+    {
+      "epoch": 0.3811896226988274,
+      "grad_norm": 0.24208222329616547,
+      "learning_rate": 6.851672019066364e-05,
+      "loss": 1.2321,
+      "step": 6274
+    },
+    {
+      "epoch": 0.3812503797314539,
+      "grad_norm": 0.18586929142475128,
+      "learning_rate": 6.85078277168097e-05,
+      "loss": 1.0546,
+      "step": 6275
+    },
+    {
+      "epoch": 0.38131113676408046,
+      "grad_norm": 0.1334799975156784,
+      "learning_rate": 6.849893456454775e-05,
+      "loss": 1.0723,
+      "step": 6276
+    },
+    {
+      "epoch": 0.381371893796707,
+      "grad_norm": 0.15104424953460693,
+      "learning_rate": 6.849004073420377e-05,
+      "loss": 1.0679,
+      "step": 6277
+    },
+    {
+      "epoch": 0.3814326508293335,
+      "grad_norm": 0.24542753398418427,
+      "learning_rate": 6.848114622610377e-05,
+      "loss": 1.2248,
+      "step": 6278
+    },
+    {
+      "epoch": 0.38149340786196,
+      "grad_norm": 0.161895751953125,
+      "learning_rate": 6.847225104057382e-05,
+      "loss": 1.0902,
+      "step": 6279
+    },
+    {
+      "epoch": 0.38155416489458654,
+      "grad_norm": 0.19862933456897736,
+      "learning_rate": 6.84633551779399e-05,
+      "loss": 1.1896,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3816149219272131,
+      "grad_norm": 0.21482138335704803,
+      "learning_rate": 6.845445863852814e-05,
+      "loss": 1.1489,
+      "step": 6281
+    },
+    {
+      "epoch": 0.3816756789598396,
+      "grad_norm": 0.1661219298839569,
+      "learning_rate": 6.844556142266464e-05,
+      "loss": 1.0921,
+      "step": 6282
+    },
+    {
+      "epoch": 0.38173643599246615,
+      "grad_norm": 0.12064141780138016,
+      "learning_rate": 6.843666353067549e-05,
+      "loss": 1.0664,
+      "step": 6283
+    },
+    {
+      "epoch": 0.38179719302509263,
+      "grad_norm": 2.6466097831726074,
+      "learning_rate": 6.84277649628869e-05,
+      "loss": 1.1072,
+      "step": 6284
+    },
+    {
+      "epoch": 0.38185795005771916,
+      "grad_norm": 0.13032706081867218,
+      "learning_rate": 6.841886571962501e-05,
+      "loss": 1.0948,
+      "step": 6285
+    },
+    {
+      "epoch": 0.3819187070903457,
+      "grad_norm": 0.12376551330089569,
+      "learning_rate": 6.840996580121606e-05,
+      "loss": 1.0807,
+      "step": 6286
+    },
+    {
+      "epoch": 0.38197946412297223,
+      "grad_norm": 0.43306705355644226,
+      "learning_rate": 6.840106520798624e-05,
+      "loss": 1.0655,
+      "step": 6287
+    },
+    {
+      "epoch": 0.38204022115559877,
+      "grad_norm": 0.458124577999115,
+      "learning_rate": 6.839216394026182e-05,
+      "loss": 1.0614,
+      "step": 6288
+    },
+    {
+      "epoch": 0.3821009781882253,
+      "grad_norm": 0.1400168240070343,
+      "learning_rate": 6.838326199836906e-05,
+      "loss": 1.093,
+      "step": 6289
+    },
+    {
+      "epoch": 0.38216173522085184,
+      "grad_norm": 0.3852221965789795,
+      "learning_rate": 6.83743593826343e-05,
+      "loss": 1.0747,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3822224922534783,
+      "grad_norm": 0.17086315155029297,
+      "learning_rate": 6.836545609338383e-05,
+      "loss": 1.0572,
+      "step": 6291
+    },
+    {
+      "epoch": 0.38228324928610485,
+      "grad_norm": 0.31543099880218506,
+      "learning_rate": 6.835655213094403e-05,
+      "loss": 1.1819,
+      "step": 6292
+    },
+    {
+      "epoch": 0.3823440063187314,
+      "grad_norm": 0.1580762267112732,
+      "learning_rate": 6.834764749564125e-05,
+      "loss": 1.1262,
+      "step": 6293
+    },
+    {
+      "epoch": 0.3824047633513579,
+      "grad_norm": 0.15375079214572906,
+      "learning_rate": 6.833874218780192e-05,
+      "loss": 1.0844,
+      "step": 6294
+    },
+    {
+      "epoch": 0.38246552038398446,
+      "grad_norm": 0.1681925356388092,
+      "learning_rate": 6.832983620775244e-05,
+      "loss": 1.1153,
+      "step": 6295
+    },
+    {
+      "epoch": 0.382526277416611,
+      "grad_norm": 0.13680367171764374,
+      "learning_rate": 6.832092955581926e-05,
+      "loss": 1.0719,
+      "step": 6296
+    },
+    {
+      "epoch": 0.3825870344492375,
+      "grad_norm": 3.1433639526367188,
+      "learning_rate": 6.831202223232889e-05,
+      "loss": 1.0424,
+      "step": 6297
+    },
+    {
+      "epoch": 0.382647791481864,
+      "grad_norm": 0.22310005128383636,
+      "learning_rate": 6.830311423760779e-05,
+      "loss": 1.0496,
+      "step": 6298
+    },
+    {
+      "epoch": 0.38270854851449054,
+      "grad_norm": 0.13245375454425812,
+      "learning_rate": 6.829420557198251e-05,
+      "loss": 1.0397,
+      "step": 6299
+    },
+    {
+      "epoch": 0.3827693055471171,
+      "grad_norm": 0.3128804862499237,
+      "learning_rate": 6.828529623577957e-05,
+      "loss": 1.1724,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3828300625797436,
+      "grad_norm": 0.15887482464313507,
+      "learning_rate": 6.827638622932558e-05,
+      "loss": 1.0794,
+      "step": 6301
+    },
+    {
+      "epoch": 0.38289081961237015,
+      "grad_norm": 0.1257995218038559,
+      "learning_rate": 6.826747555294711e-05,
+      "loss": 1.0546,
+      "step": 6302
+    },
+    {
+      "epoch": 0.3829515766449967,
+      "grad_norm": 0.23607231676578522,
+      "learning_rate": 6.825856420697081e-05,
+      "loss": 1.1403,
+      "step": 6303
+    },
+    {
+      "epoch": 0.38301233367762316,
+      "grad_norm": 0.14890529215335846,
+      "learning_rate": 6.824965219172328e-05,
+      "loss": 1.0259,
+      "step": 6304
+    },
+    {
+      "epoch": 0.3830730907102497,
+      "grad_norm": 0.5186467170715332,
+      "learning_rate": 6.824073950753124e-05,
+      "loss": 1.0646,
+      "step": 6305
+    },
+    {
+      "epoch": 0.38313384774287623,
+      "grad_norm": 0.1965055614709854,
+      "learning_rate": 6.823182615472137e-05,
+      "loss": 1.067,
+      "step": 6306
+    },
+    {
+      "epoch": 0.38319460477550277,
+      "grad_norm": 0.3759694993495941,
+      "learning_rate": 6.822291213362038e-05,
+      "loss": 1.1792,
+      "step": 6307
+    },
+    {
+      "epoch": 0.3832553618081293,
+      "grad_norm": 0.4658016264438629,
+      "learning_rate": 6.821399744455503e-05,
+      "loss": 1.0957,
+      "step": 6308
+    },
+    {
+      "epoch": 0.38331611884075584,
+      "grad_norm": 0.16722477972507477,
+      "learning_rate": 6.820508208785207e-05,
+      "loss": 1.0696,
+      "step": 6309
+    },
+    {
+      "epoch": 0.3833768758733823,
+      "grad_norm": 0.14305900037288666,
+      "learning_rate": 6.819616606383832e-05,
+      "loss": 1.0703,
+      "step": 6310
+    },
+    {
+      "epoch": 0.38343763290600885,
+      "grad_norm": 0.24095800518989563,
+      "learning_rate": 6.818724937284059e-05,
+      "loss": 1.0699,
+      "step": 6311
+    },
+    {
+      "epoch": 0.3834983899386354,
+      "grad_norm": 0.17182044684886932,
+      "learning_rate": 6.817833201518571e-05,
+      "loss": 1.0756,
+      "step": 6312
+    },
+    {
+      "epoch": 0.3835591469712619,
+      "grad_norm": 0.1402570754289627,
+      "learning_rate": 6.816941399120056e-05,
+      "loss": 1.0917,
+      "step": 6313
+    },
+    {
+      "epoch": 0.38361990400388846,
+      "grad_norm": 0.21883484721183777,
+      "learning_rate": 6.816049530121202e-05,
+      "loss": 1.1313,
+      "step": 6314
+    },
+    {
+      "epoch": 0.383680661036515,
+      "grad_norm": 0.4477614462375641,
+      "learning_rate": 6.815157594554702e-05,
+      "loss": 1.0917,
+      "step": 6315
+    },
+    {
+      "epoch": 0.3837414180691415,
+      "grad_norm": 0.1858288198709488,
+      "learning_rate": 6.81426559245325e-05,
+      "loss": 1.124,
+      "step": 6316
+    },
+    {
+      "epoch": 0.383802175101768,
+      "grad_norm": 0.22513240575790405,
+      "learning_rate": 6.81337352384954e-05,
+      "loss": 1.1854,
+      "step": 6317
+    },
+    {
+      "epoch": 0.38386293213439454,
+      "grad_norm": 0.2168305367231369,
+      "learning_rate": 6.812481388776275e-05,
+      "loss": 1.0872,
+      "step": 6318
+    },
+    {
+      "epoch": 0.3839236891670211,
+      "grad_norm": 0.17464782297611237,
+      "learning_rate": 6.811589187266155e-05,
+      "loss": 1.1634,
+      "step": 6319
+    },
+    {
+      "epoch": 0.3839844461996476,
+      "grad_norm": 0.14811238646507263,
+      "learning_rate": 6.81069691935188e-05,
+      "loss": 1.0339,
+      "step": 6320
+    },
+    {
+      "epoch": 0.38404520323227415,
+      "grad_norm": 0.22146408259868622,
+      "learning_rate": 6.809804585066162e-05,
+      "loss": 1.2309,
+      "step": 6321
+    },
+    {
+      "epoch": 0.3841059602649007,
+      "grad_norm": 0.12117231637239456,
+      "learning_rate": 6.808912184441707e-05,
+      "loss": 1.0786,
+      "step": 6322
+    },
+    {
+      "epoch": 0.3841667172975272,
+      "grad_norm": 0.21472381055355072,
+      "learning_rate": 6.808019717511226e-05,
+      "loss": 1.0601,
+      "step": 6323
+    },
+    {
+      "epoch": 0.3842274743301537,
+      "grad_norm": 0.12870797514915466,
+      "learning_rate": 6.807127184307433e-05,
+      "loss": 1.0582,
+      "step": 6324
+    },
+    {
+      "epoch": 0.38428823136278023,
+      "grad_norm": 0.18556331098079681,
+      "learning_rate": 6.806234584863043e-05,
+      "loss": 1.0549,
+      "step": 6325
+    },
+    {
+      "epoch": 0.38434898839540677,
+      "grad_norm": 0.1558101624250412,
+      "learning_rate": 6.805341919210775e-05,
+      "loss": 1.0464,
+      "step": 6326
+    },
+    {
+      "epoch": 0.3844097454280333,
+      "grad_norm": 0.1820998340845108,
+      "learning_rate": 6.80444918738335e-05,
+      "loss": 1.1543,
+      "step": 6327
+    },
+    {
+      "epoch": 0.38447050246065984,
+      "grad_norm": 0.15699411928653717,
+      "learning_rate": 6.803556389413491e-05,
+      "loss": 1.0545,
+      "step": 6328
+    },
+    {
+      "epoch": 0.38453125949328637,
+      "grad_norm": 0.17958933115005493,
+      "learning_rate": 6.802663525333926e-05,
+      "loss": 1.103,
+      "step": 6329
+    },
+    {
+      "epoch": 0.38459201652591285,
+      "grad_norm": 0.11888808757066727,
+      "learning_rate": 6.801770595177379e-05,
+      "loss": 1.0349,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3846527735585394,
+      "grad_norm": 0.2576075494289398,
+      "learning_rate": 6.800877598976583e-05,
+      "loss": 1.182,
+      "step": 6331
+    },
+    {
+      "epoch": 0.3847135305911659,
+      "grad_norm": 0.228237584233284,
+      "learning_rate": 6.79998453676427e-05,
+      "loss": 1.1383,
+      "step": 6332
+    },
+    {
+      "epoch": 0.38477428762379245,
+      "grad_norm": 0.1752501130104065,
+      "learning_rate": 6.799091408573177e-05,
+      "loss": 1.1117,
+      "step": 6333
+    },
+    {
+      "epoch": 0.384835044656419,
+      "grad_norm": 0.18376250565052032,
+      "learning_rate": 6.79819821443604e-05,
+      "loss": 1.1039,
+      "step": 6334
+    },
+    {
+      "epoch": 0.3848958016890455,
+      "grad_norm": 0.15483279526233673,
+      "learning_rate": 6.7973049543856e-05,
+      "loss": 1.0891,
+      "step": 6335
+    },
+    {
+      "epoch": 0.38495655872167206,
+      "grad_norm": 0.26706215739250183,
+      "learning_rate": 6.7964116284546e-05,
+      "loss": 1.0494,
+      "step": 6336
+    },
+    {
+      "epoch": 0.38501731575429854,
+      "grad_norm": 0.20220930874347687,
+      "learning_rate": 6.795518236675784e-05,
+      "loss": 1.205,
+      "step": 6337
+    },
+    {
+      "epoch": 0.3850780727869251,
+      "grad_norm": 0.4418732523918152,
+      "learning_rate": 6.7946247790819e-05,
+      "loss": 1.0601,
+      "step": 6338
+    },
+    {
+      "epoch": 0.3851388298195516,
+      "grad_norm": 0.16609244048595428,
+      "learning_rate": 6.793731255705699e-05,
+      "loss": 1.0968,
+      "step": 6339
+    },
+    {
+      "epoch": 0.38519958685217814,
+      "grad_norm": 0.2317761927843094,
+      "learning_rate": 6.79283766657993e-05,
+      "loss": 1.0485,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3852603438848047,
+      "grad_norm": 0.16373886168003082,
+      "learning_rate": 6.79194401173735e-05,
+      "loss": 1.0558,
+      "step": 6341
+    },
+    {
+      "epoch": 0.3853211009174312,
+      "grad_norm": 0.18133790791034698,
+      "learning_rate": 6.791050291210716e-05,
+      "loss": 1.0689,
+      "step": 6342
+    },
+    {
+      "epoch": 0.3853818579500577,
+      "grad_norm": 0.39687323570251465,
+      "learning_rate": 6.79015650503279e-05,
+      "loss": 1.1413,
+      "step": 6343
+    },
+    {
+      "epoch": 0.38544261498268423,
+      "grad_norm": 0.1871715635061264,
+      "learning_rate": 6.78926265323633e-05,
+      "loss": 1.0097,
+      "step": 6344
+    },
+    {
+      "epoch": 0.38550337201531076,
+      "grad_norm": 0.18519176542758942,
+      "learning_rate": 6.788368735854099e-05,
+      "loss": 1.0522,
+      "step": 6345
+    },
+    {
+      "epoch": 0.3855641290479373,
+      "grad_norm": 0.13306421041488647,
+      "learning_rate": 6.787474752918869e-05,
+      "loss": 1.0624,
+      "step": 6346
+    },
+    {
+      "epoch": 0.38562488608056383,
+      "grad_norm": 0.4891233444213867,
+      "learning_rate": 6.786580704463406e-05,
+      "loss": 1.237,
+      "step": 6347
+    },
+    {
+      "epoch": 0.38568564311319037,
+      "grad_norm": 0.1669224500656128,
+      "learning_rate": 6.785686590520481e-05,
+      "loss": 1.1104,
+      "step": 6348
+    },
+    {
+      "epoch": 0.3857464001458169,
+      "grad_norm": 0.19904066622257233,
+      "learning_rate": 6.78479241112287e-05,
+      "loss": 1.2674,
+      "step": 6349
+    },
+    {
+      "epoch": 0.3858071571784434,
+      "grad_norm": 0.28875163197517395,
+      "learning_rate": 6.783898166303346e-05,
+      "loss": 1.1573,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3858679142110699,
+      "grad_norm": 1.2536693811416626,
+      "learning_rate": 6.783003856094692e-05,
+      "loss": 1.0715,
+      "step": 6351
+    },
+    {
+      "epoch": 0.38592867124369645,
+      "grad_norm": 3.2727980613708496,
+      "learning_rate": 6.782109480529689e-05,
+      "loss": 1.09,
+      "step": 6352
+    },
+    {
+      "epoch": 0.385989428276323,
+      "grad_norm": 0.23191440105438232,
+      "learning_rate": 6.781215039641114e-05,
+      "loss": 1.0536,
+      "step": 6353
+    },
+    {
+      "epoch": 0.3860501853089495,
+      "grad_norm": 0.16875076293945312,
+      "learning_rate": 6.78032053346176e-05,
+      "loss": 1.065,
+      "step": 6354
+    },
+    {
+      "epoch": 0.38611094234157606,
+      "grad_norm": 0.2235843539237976,
+      "learning_rate": 6.779425962024411e-05,
+      "loss": 1.2331,
+      "step": 6355
+    },
+    {
+      "epoch": 0.38617169937420254,
+      "grad_norm": 0.1509760618209839,
+      "learning_rate": 6.77853132536186e-05,
+      "loss": 1.0639,
+      "step": 6356
+    },
+    {
+      "epoch": 0.38623245640682907,
+      "grad_norm": 0.21987676620483398,
+      "learning_rate": 6.777636623506901e-05,
+      "loss": 1.0701,
+      "step": 6357
+    },
+    {
+      "epoch": 0.3862932134394556,
+      "grad_norm": 0.3164826035499573,
+      "learning_rate": 6.776741856492326e-05,
+      "loss": 1.1628,
+      "step": 6358
+    },
+    {
+      "epoch": 0.38635397047208214,
+      "grad_norm": 0.32382747530937195,
+      "learning_rate": 6.775847024350936e-05,
+      "loss": 1.1056,
+      "step": 6359
+    },
+    {
+      "epoch": 0.3864147275047087,
+      "grad_norm": 0.47914010286331177,
+      "learning_rate": 6.77495212711553e-05,
+      "loss": 1.1833,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3864754845373352,
+      "grad_norm": 5.246538162231445,
+      "learning_rate": 6.77405716481891e-05,
+      "loss": 1.0786,
+      "step": 6361
+    },
+    {
+      "epoch": 0.38653624156996175,
+      "grad_norm": 0.452112078666687,
+      "learning_rate": 6.773162137493884e-05,
+      "loss": 1.0889,
+      "step": 6362
+    },
+    {
+      "epoch": 0.3865969986025882,
+      "grad_norm": 0.149632528424263,
+      "learning_rate": 6.772267045173254e-05,
+      "loss": 1.0516,
+      "step": 6363
+    },
+    {
+      "epoch": 0.38665775563521476,
+      "grad_norm": 0.16905269026756287,
+      "learning_rate": 6.771371887889835e-05,
+      "loss": 1.0975,
+      "step": 6364
+    },
+    {
+      "epoch": 0.3867185126678413,
+      "grad_norm": 0.16825300455093384,
+      "learning_rate": 6.770476665676435e-05,
+      "loss": 1.0974,
+      "step": 6365
+    },
+    {
+      "epoch": 0.38677926970046783,
+      "grad_norm": 0.18294303119182587,
+      "learning_rate": 6.769581378565873e-05,
+      "loss": 1.1711,
+      "step": 6366
+    },
+    {
+      "epoch": 0.38684002673309437,
+      "grad_norm": 0.14837855100631714,
+      "learning_rate": 6.768686026590963e-05,
+      "loss": 1.0737,
+      "step": 6367
+    },
+    {
+      "epoch": 0.3869007837657209,
+      "grad_norm": 0.23875705897808075,
+      "learning_rate": 6.767790609784523e-05,
+      "loss": 1.1615,
+      "step": 6368
+    },
+    {
+      "epoch": 0.38696154079834744,
+      "grad_norm": 3.9483509063720703,
+      "learning_rate": 6.766895128179378e-05,
+      "loss": 1.0591,
+      "step": 6369
+    },
+    {
+      "epoch": 0.3870222978309739,
+      "grad_norm": 0.16219587624073029,
+      "learning_rate": 6.765999581808351e-05,
+      "loss": 1.065,
+      "step": 6370
+    },
+    {
+      "epoch": 0.38708305486360045,
+      "grad_norm": 0.875992476940155,
+      "learning_rate": 6.765103970704266e-05,
+      "loss": 1.1725,
+      "step": 6371
+    },
+    {
+      "epoch": 0.387143811896227,
+      "grad_norm": 1.1835606098175049,
+      "learning_rate": 6.764208294899956e-05,
+      "loss": 1.068,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3872045689288535,
+      "grad_norm": 0.43274930119514465,
+      "learning_rate": 6.76331255442825e-05,
+      "loss": 1.0657,
+      "step": 6373
+    },
+    {
+      "epoch": 0.38726532596148006,
+      "grad_norm": 0.13690155744552612,
+      "learning_rate": 6.762416749321978e-05,
+      "loss": 1.0966,
+      "step": 6374
+    },
+    {
+      "epoch": 0.3873260829941066,
+      "grad_norm": 0.16002944111824036,
+      "learning_rate": 6.761520879613982e-05,
+      "loss": 1.1208,
+      "step": 6375
+    },
+    {
+      "epoch": 0.38738684002673307,
+      "grad_norm": 0.2330576479434967,
+      "learning_rate": 6.760624945337097e-05,
+      "loss": 1.1549,
+      "step": 6376
+    },
+    {
+      "epoch": 0.3874475970593596,
+      "grad_norm": 0.12631060183048248,
+      "learning_rate": 6.759728946524165e-05,
+      "loss": 1.0978,
+      "step": 6377
+    },
+    {
+      "epoch": 0.38750835409198614,
+      "grad_norm": 0.1642490178346634,
+      "learning_rate": 6.758832883208029e-05,
+      "loss": 1.0364,
+      "step": 6378
+    },
+    {
+      "epoch": 0.3875691111246127,
+      "grad_norm": 0.18020714819431305,
+      "learning_rate": 6.757936755421532e-05,
+      "loss": 1.1049,
+      "step": 6379
+    },
+    {
+      "epoch": 0.3876298681572392,
+      "grad_norm": 0.14700233936309814,
+      "learning_rate": 6.757040563197525e-05,
+      "loss": 1.0716,
+      "step": 6380
+    },
+    {
+      "epoch": 0.38769062518986575,
+      "grad_norm": 0.26052945852279663,
+      "learning_rate": 6.756144306568856e-05,
+      "loss": 1.2431,
+      "step": 6381
+    },
+    {
+      "epoch": 0.3877513822224923,
+      "grad_norm": 0.20602093636989594,
+      "learning_rate": 6.755247985568379e-05,
+      "loss": 1.0884,
+      "step": 6382
+    },
+    {
+      "epoch": 0.38781213925511876,
+      "grad_norm": 0.16676735877990723,
+      "learning_rate": 6.754351600228945e-05,
+      "loss": 1.1352,
+      "step": 6383
+    },
+    {
+      "epoch": 0.3878728962877453,
+      "grad_norm": 0.24783259630203247,
+      "learning_rate": 6.753455150583415e-05,
+      "loss": 1.1362,
+      "step": 6384
+    },
+    {
+      "epoch": 0.38793365332037183,
+      "grad_norm": 0.17209668457508087,
+      "learning_rate": 6.752558636664648e-05,
+      "loss": 1.1415,
+      "step": 6385
+    },
+    {
+      "epoch": 0.38799441035299836,
+      "grad_norm": 28.459001541137695,
+      "learning_rate": 6.751662058505506e-05,
+      "loss": 1.0554,
+      "step": 6386
+    },
+    {
+      "epoch": 0.3880551673856249,
+      "grad_norm": 0.2686154246330261,
+      "learning_rate": 6.750765416138851e-05,
+      "loss": 1.2096,
+      "step": 6387
+    },
+    {
+      "epoch": 0.38811592441825143,
+      "grad_norm": 0.46046197414398193,
+      "learning_rate": 6.749868709597553e-05,
+      "loss": 1.0291,
+      "step": 6388
+    },
+    {
+      "epoch": 0.3881766814508779,
+      "grad_norm": 0.2041414976119995,
+      "learning_rate": 6.748971938914479e-05,
+      "loss": 1.1269,
+      "step": 6389
+    },
+    {
+      "epoch": 0.38823743848350445,
+      "grad_norm": 0.2749987244606018,
+      "learning_rate": 6.7480751041225e-05,
+      "loss": 1.156,
+      "step": 6390
+    },
+    {
+      "epoch": 0.388298195516131,
+      "grad_norm": 0.26438790559768677,
+      "learning_rate": 6.74717820525449e-05,
+      "loss": 1.2141,
+      "step": 6391
+    },
+    {
+      "epoch": 0.3883589525487575,
+      "grad_norm": 0.5964327454566956,
+      "learning_rate": 6.746281242343325e-05,
+      "loss": 1.1732,
+      "step": 6392
+    },
+    {
+      "epoch": 0.38841970958138405,
+      "grad_norm": 0.18892355263233185,
+      "learning_rate": 6.745384215421884e-05,
+      "loss": 1.0508,
+      "step": 6393
+    },
+    {
+      "epoch": 0.3884804666140106,
+      "grad_norm": 0.13409344851970673,
+      "learning_rate": 6.744487124523047e-05,
+      "loss": 1.0793,
+      "step": 6394
+    },
+    {
+      "epoch": 0.3885412236466371,
+      "grad_norm": 0.2899182140827179,
+      "learning_rate": 6.743589969679697e-05,
+      "loss": 1.1475,
+      "step": 6395
+    },
+    {
+      "epoch": 0.3886019806792636,
+      "grad_norm": 0.1661696881055832,
+      "learning_rate": 6.742692750924722e-05,
+      "loss": 1.0464,
+      "step": 6396
+    },
+    {
+      "epoch": 0.38866273771189014,
+      "grad_norm": 0.1817939281463623,
+      "learning_rate": 6.741795468291003e-05,
+      "loss": 1.1267,
+      "step": 6397
+    },
+    {
+      "epoch": 0.3887234947445167,
+      "grad_norm": 0.14239276945590973,
+      "learning_rate": 6.740898121811437e-05,
+      "loss": 1.0579,
+      "step": 6398
+    },
+    {
+      "epoch": 0.3887842517771432,
+      "grad_norm": 0.177519291639328,
+      "learning_rate": 6.740000711518915e-05,
+      "loss": 1.0958,
+      "step": 6399
+    },
+    {
+      "epoch": 0.38884500880976974,
+      "grad_norm": 0.22885555028915405,
+      "learning_rate": 6.739103237446328e-05,
+      "loss": 1.1554,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3889057658423963,
+      "grad_norm": 0.19650842249393463,
+      "learning_rate": 6.738205699626578e-05,
+      "loss": 1.0324,
+      "step": 6401
+    },
+    {
+      "epoch": 0.38896652287502276,
+      "grad_norm": 0.21699388325214386,
+      "learning_rate": 6.737308098092561e-05,
+      "loss": 1.1588,
+      "step": 6402
+    },
+    {
+      "epoch": 0.3890272799076493,
+      "grad_norm": 0.21354782581329346,
+      "learning_rate": 6.736410432877179e-05,
+      "loss": 1.1649,
+      "step": 6403
+    },
+    {
+      "epoch": 0.3890880369402758,
+      "grad_norm": 0.2252529114484787,
+      "learning_rate": 6.735512704013339e-05,
+      "loss": 1.1875,
+      "step": 6404
+    },
+    {
+      "epoch": 0.38914879397290236,
+      "grad_norm": 0.18823717534542084,
+      "learning_rate": 6.734614911533944e-05,
+      "loss": 1.1775,
+      "step": 6405
+    },
+    {
+      "epoch": 0.3892095510055289,
+      "grad_norm": 0.12210594862699509,
+      "learning_rate": 6.733717055471904e-05,
+      "loss": 1.051,
+      "step": 6406
+    },
+    {
+      "epoch": 0.38927030803815543,
+      "grad_norm": 0.23055042326450348,
+      "learning_rate": 6.732819135860129e-05,
+      "loss": 1.1588,
+      "step": 6407
+    },
+    {
+      "epoch": 0.38933106507078197,
+      "grad_norm": 0.1257196068763733,
+      "learning_rate": 6.731921152731535e-05,
+      "loss": 1.0678,
+      "step": 6408
+    },
+    {
+      "epoch": 0.38939182210340845,
+      "grad_norm": 0.2573351562023163,
+      "learning_rate": 6.731023106119038e-05,
+      "loss": 1.1496,
+      "step": 6409
+    },
+    {
+      "epoch": 0.389452579136035,
+      "grad_norm": 0.15273867547512054,
+      "learning_rate": 6.730124996055552e-05,
+      "loss": 1.1913,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3895133361686615,
+      "grad_norm": 0.17081992328166962,
+      "learning_rate": 6.729226822574e-05,
+      "loss": 1.1167,
+      "step": 6411
+    },
+    {
+      "epoch": 0.38957409320128805,
+      "grad_norm": 0.14605049788951874,
+      "learning_rate": 6.728328585707306e-05,
+      "loss": 1.0909,
+      "step": 6412
+    },
+    {
+      "epoch": 0.3896348502339146,
+      "grad_norm": 0.19868595898151398,
+      "learning_rate": 6.727430285488393e-05,
+      "loss": 1.0877,
+      "step": 6413
+    },
+    {
+      "epoch": 0.3896956072665411,
+      "grad_norm": 0.3197658360004425,
+      "learning_rate": 6.726531921950189e-05,
+      "loss": 1.1905,
+      "step": 6414
+    },
+    {
+      "epoch": 0.3897563642991676,
+      "grad_norm": 0.11486510932445526,
+      "learning_rate": 6.725633495125623e-05,
+      "loss": 1.0434,
+      "step": 6415
+    },
+    {
+      "epoch": 0.38981712133179414,
+      "grad_norm": 0.16571666300296783,
+      "learning_rate": 6.724735005047629e-05,
+      "loss": 1.0855,
+      "step": 6416
+    },
+    {
+      "epoch": 0.38987787836442067,
+      "grad_norm": 0.278359591960907,
+      "learning_rate": 6.72383645174914e-05,
+      "loss": 1.297,
+      "step": 6417
+    },
+    {
+      "epoch": 0.3899386353970472,
+      "grad_norm": 0.4895019233226776,
+      "learning_rate": 6.72293783526309e-05,
+      "loss": 1.0789,
+      "step": 6418
+    },
+    {
+      "epoch": 0.38999939242967374,
+      "grad_norm": 0.13772952556610107,
+      "learning_rate": 6.722039155622424e-05,
+      "loss": 1.1092,
+      "step": 6419
+    },
+    {
+      "epoch": 0.3900601494623003,
+      "grad_norm": 0.21690237522125244,
+      "learning_rate": 6.721140412860077e-05,
+      "loss": 1.0926,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3901209064949268,
+      "grad_norm": 0.8746984601020813,
+      "learning_rate": 6.720241607008997e-05,
+      "loss": 1.0559,
+      "step": 6421
+    },
+    {
+      "epoch": 0.3901816635275533,
+      "grad_norm": 0.13369816541671753,
+      "learning_rate": 6.719342738102128e-05,
+      "loss": 1.0433,
+      "step": 6422
+    },
+    {
+      "epoch": 0.3902424205601798,
+      "grad_norm": 0.15946729481220245,
+      "learning_rate": 6.71844380617242e-05,
+      "loss": 1.1068,
+      "step": 6423
+    },
+    {
+      "epoch": 0.39030317759280636,
+      "grad_norm": 0.2904299795627594,
+      "learning_rate": 6.717544811252821e-05,
+      "loss": 1.1724,
+      "step": 6424
+    },
+    {
+      "epoch": 0.3903639346254329,
+      "grad_norm": 0.13413147628307343,
+      "learning_rate": 6.716645753376284e-05,
+      "loss": 1.0318,
+      "step": 6425
+    },
+    {
+      "epoch": 0.39042469165805943,
+      "grad_norm": 0.1666967272758484,
+      "learning_rate": 6.715746632575767e-05,
+      "loss": 1.0444,
+      "step": 6426
+    },
+    {
+      "epoch": 0.39048544869068597,
+      "grad_norm": 0.20521396398544312,
+      "learning_rate": 6.714847448884225e-05,
+      "loss": 1.0385,
+      "step": 6427
+    },
+    {
+      "epoch": 0.3905462057233125,
+      "grad_norm": 0.33713576197624207,
+      "learning_rate": 6.713948202334617e-05,
+      "loss": 1.1237,
+      "step": 6428
+    },
+    {
+      "epoch": 0.390606962755939,
+      "grad_norm": 0.23071159422397614,
+      "learning_rate": 6.713048892959908e-05,
+      "loss": 1.1186,
+      "step": 6429
+    },
+    {
+      "epoch": 0.3906677197885655,
+      "grad_norm": 6.779727935791016,
+      "learning_rate": 6.71214952079306e-05,
+      "loss": 1.1391,
+      "step": 6430
+    },
+    {
+      "epoch": 0.39072847682119205,
+      "grad_norm": 0.3514596223831177,
+      "learning_rate": 6.711250085867039e-05,
+      "loss": 1.1629,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3907892338538186,
+      "grad_norm": 0.18650712072849274,
+      "learning_rate": 6.710350588214817e-05,
+      "loss": 1.1439,
+      "step": 6432
+    },
+    {
+      "epoch": 0.3908499908864451,
+      "grad_norm": 0.4680578410625458,
+      "learning_rate": 6.709451027869364e-05,
+      "loss": 1.0795,
+      "step": 6433
+    },
+    {
+      "epoch": 0.39091074791907165,
+      "grad_norm": 0.19026999175548553,
+      "learning_rate": 6.708551404863653e-05,
+      "loss": 1.0439,
+      "step": 6434
+    },
+    {
+      "epoch": 0.39097150495169813,
+      "grad_norm": 0.2945087254047394,
+      "learning_rate": 6.707651719230658e-05,
+      "loss": 1.0689,
+      "step": 6435
+    },
+    {
+      "epoch": 0.39103226198432467,
+      "grad_norm": 0.15697064995765686,
+      "learning_rate": 6.70675197100336e-05,
+      "loss": 1.0379,
+      "step": 6436
+    },
+    {
+      "epoch": 0.3910930190169512,
+      "grad_norm": 0.527032732963562,
+      "learning_rate": 6.70585216021474e-05,
+      "loss": 1.1346,
+      "step": 6437
+    },
+    {
+      "epoch": 0.39115377604957774,
+      "grad_norm": 1.2023156881332397,
+      "learning_rate": 6.704952286897778e-05,
+      "loss": 1.0391,
+      "step": 6438
+    },
+    {
+      "epoch": 0.3912145330822043,
+      "grad_norm": 0.15741731226444244,
+      "learning_rate": 6.704052351085462e-05,
+      "loss": 1.0807,
+      "step": 6439
+    },
+    {
+      "epoch": 0.3912752901148308,
+      "grad_norm": 0.15480320155620575,
+      "learning_rate": 6.703152352810775e-05,
+      "loss": 1.0479,
+      "step": 6440
+    },
+    {
+      "epoch": 0.39133604714745734,
+      "grad_norm": 0.16746895015239716,
+      "learning_rate": 6.70225229210671e-05,
+      "loss": 1.1609,
+      "step": 6441
+    },
+    {
+      "epoch": 0.3913968041800838,
+      "grad_norm": 0.12224670499563217,
+      "learning_rate": 6.70135216900626e-05,
+      "loss": 1.0185,
+      "step": 6442
+    },
+    {
+      "epoch": 0.39145756121271036,
+      "grad_norm": 0.17947405576705933,
+      "learning_rate": 6.700451983542415e-05,
+      "loss": 1.1291,
+      "step": 6443
+    },
+    {
+      "epoch": 0.3915183182453369,
+      "grad_norm": 0.1567447930574417,
+      "learning_rate": 6.699551735748176e-05,
+      "loss": 1.0741,
+      "step": 6444
+    },
+    {
+      "epoch": 0.39157907527796343,
+      "grad_norm": 0.2524197995662689,
+      "learning_rate": 6.698651425656539e-05,
+      "loss": 1.2205,
+      "step": 6445
+    },
+    {
+      "epoch": 0.39163983231058996,
+      "grad_norm": 0.20584231615066528,
+      "learning_rate": 6.697751053300505e-05,
+      "loss": 1.138,
+      "step": 6446
+    },
+    {
+      "epoch": 0.3917005893432165,
+      "grad_norm": 0.21095509827136993,
+      "learning_rate": 6.696850618713079e-05,
+      "loss": 1.1867,
+      "step": 6447
+    },
+    {
+      "epoch": 0.391761346375843,
+      "grad_norm": 0.1479436755180359,
+      "learning_rate": 6.695950121927265e-05,
+      "loss": 1.0653,
+      "step": 6448
+    },
+    {
+      "epoch": 0.3918221034084695,
+      "grad_norm": 0.6414020657539368,
+      "learning_rate": 6.695049562976072e-05,
+      "loss": 1.2842,
+      "step": 6449
+    },
+    {
+      "epoch": 0.39188286044109605,
+      "grad_norm": 0.15622633695602417,
+      "learning_rate": 6.694148941892509e-05,
+      "loss": 1.0771,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3919436174737226,
+      "grad_norm": 0.34224677085876465,
+      "learning_rate": 6.693248258709588e-05,
+      "loss": 1.1831,
+      "step": 6451
+    },
+    {
+      "epoch": 0.3920043745063491,
+      "grad_norm": 0.49429529905319214,
+      "learning_rate": 6.692347513460326e-05,
+      "loss": 1.0159,
+      "step": 6452
+    },
+    {
+      "epoch": 0.39206513153897565,
+      "grad_norm": 0.15195602178573608,
+      "learning_rate": 6.69144670617774e-05,
+      "loss": 1.0711,
+      "step": 6453
+    },
+    {
+      "epoch": 0.3921258885716022,
+      "grad_norm": 0.22439368069171906,
+      "learning_rate": 6.690545836894845e-05,
+      "loss": 1.254,
+      "step": 6454
+    },
+    {
+      "epoch": 0.39218664560422867,
+      "grad_norm": 0.1413053423166275,
+      "learning_rate": 6.689644905644668e-05,
+      "loss": 1.0396,
+      "step": 6455
+    },
+    {
+      "epoch": 0.3922474026368552,
+      "grad_norm": 0.18424078822135925,
+      "learning_rate": 6.688743912460229e-05,
+      "loss": 1.0853,
+      "step": 6456
+    },
+    {
+      "epoch": 0.39230815966948174,
+      "grad_norm": 0.20230261981487274,
+      "learning_rate": 6.687842857374556e-05,
+      "loss": 1.1808,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3923689167021083,
+      "grad_norm": 0.19519254565238953,
+      "learning_rate": 6.686941740420678e-05,
+      "loss": 1.1793,
+      "step": 6458
+    },
+    {
+      "epoch": 0.3924296737347348,
+      "grad_norm": 0.179273784160614,
+      "learning_rate": 6.686040561631623e-05,
+      "loss": 1.1164,
+      "step": 6459
+    },
+    {
+      "epoch": 0.39249043076736134,
+      "grad_norm": 0.1486324667930603,
+      "learning_rate": 6.685139321040426e-05,
+      "loss": 1.1227,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3925511877999878,
+      "grad_norm": 0.1647394597530365,
+      "learning_rate": 6.684238018680121e-05,
+      "loss": 1.0897,
+      "step": 6461
+    },
+    {
+      "epoch": 0.39261194483261436,
+      "grad_norm": 0.1527169793844223,
+      "learning_rate": 6.683336654583745e-05,
+      "loss": 1.0465,
+      "step": 6462
+    },
+    {
+      "epoch": 0.3926727018652409,
+      "grad_norm": 0.13533085584640503,
+      "learning_rate": 6.682435228784341e-05,
+      "loss": 1.0504,
+      "step": 6463
+    },
+    {
+      "epoch": 0.3927334588978674,
+      "grad_norm": 0.1908823698759079,
+      "learning_rate": 6.681533741314945e-05,
+      "loss": 1.1417,
+      "step": 6464
+    },
+    {
+      "epoch": 0.39279421593049396,
+      "grad_norm": 0.20578359067440033,
+      "learning_rate": 6.680632192208609e-05,
+      "loss": 1.1132,
+      "step": 6465
+    },
+    {
+      "epoch": 0.3928549729631205,
+      "grad_norm": 0.2384302318096161,
+      "learning_rate": 6.679730581498372e-05,
+      "loss": 1.1656,
+      "step": 6466
+    },
+    {
+      "epoch": 0.39291572999574703,
+      "grad_norm": 0.16097606718540192,
+      "learning_rate": 6.678828909217286e-05,
+      "loss": 1.1049,
+      "step": 6467
+    },
+    {
+      "epoch": 0.3929764870283735,
+      "grad_norm": 0.13006088137626648,
+      "learning_rate": 6.677927175398403e-05,
+      "loss": 1.0118,
+      "step": 6468
+    },
+    {
+      "epoch": 0.39303724406100005,
+      "grad_norm": 0.29412567615509033,
+      "learning_rate": 6.677025380074775e-05,
+      "loss": 1.1508,
+      "step": 6469
+    },
+    {
+      "epoch": 0.3930980010936266,
+      "grad_norm": 0.42197516560554504,
+      "learning_rate": 6.676123523279457e-05,
+      "loss": 1.0784,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3931587581262531,
+      "grad_norm": 0.32488706707954407,
+      "learning_rate": 6.675221605045507e-05,
+      "loss": 1.2118,
+      "step": 6471
+    },
+    {
+      "epoch": 0.39321951515887965,
+      "grad_norm": 0.15832357108592987,
+      "learning_rate": 6.674319625405985e-05,
+      "loss": 1.0735,
+      "step": 6472
+    },
+    {
+      "epoch": 0.3932802721915062,
+      "grad_norm": 0.17760059237480164,
+      "learning_rate": 6.673417584393955e-05,
+      "loss": 1.1496,
+      "step": 6473
+    },
+    {
+      "epoch": 0.3933410292241327,
+      "grad_norm": 0.18103596568107605,
+      "learning_rate": 6.67251548204248e-05,
+      "loss": 1.089,
+      "step": 6474
+    },
+    {
+      "epoch": 0.3934017862567592,
+      "grad_norm": 0.28530701994895935,
+      "learning_rate": 6.671613318384624e-05,
+      "loss": 1.1169,
+      "step": 6475
+    },
+    {
+      "epoch": 0.39346254328938574,
+      "grad_norm": 0.16695894300937653,
+      "learning_rate": 6.670711093453461e-05,
+      "loss": 1.068,
+      "step": 6476
+    },
+    {
+      "epoch": 0.39352330032201227,
+      "grad_norm": 0.14309263229370117,
+      "learning_rate": 6.669808807282058e-05,
+      "loss": 1.0534,
+      "step": 6477
+    },
+    {
+      "epoch": 0.3935840573546388,
+      "grad_norm": 0.21720516681671143,
+      "learning_rate": 6.668906459903493e-05,
+      "loss": 1.0586,
+      "step": 6478
+    },
+    {
+      "epoch": 0.39364481438726534,
+      "grad_norm": 0.3170429468154907,
+      "learning_rate": 6.668004051350839e-05,
+      "loss": 1.2139,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3937055714198919,
+      "grad_norm": 0.24288886785507202,
+      "learning_rate": 6.667101581657172e-05,
+      "loss": 1.1619,
+      "step": 6480
+    },
+    {
+      "epoch": 0.39376632845251835,
+      "grad_norm": 0.21614795923233032,
+      "learning_rate": 6.666199050855575e-05,
+      "loss": 1.0565,
+      "step": 6481
+    },
+    {
+      "epoch": 0.3938270854851449,
+      "grad_norm": 0.4317033886909485,
+      "learning_rate": 6.665296458979131e-05,
+      "loss": 1.089,
+      "step": 6482
+    },
+    {
+      "epoch": 0.3938878425177714,
+      "grad_norm": 0.15566758811473846,
+      "learning_rate": 6.664393806060922e-05,
+      "loss": 1.0807,
+      "step": 6483
+    },
+    {
+      "epoch": 0.39394859955039796,
+      "grad_norm": 0.18097607791423798,
+      "learning_rate": 6.663491092134038e-05,
+      "loss": 1.0569,
+      "step": 6484
+    },
+    {
+      "epoch": 0.3940093565830245,
+      "grad_norm": 0.15195651352405548,
+      "learning_rate": 6.662588317231564e-05,
+      "loss": 1.141,
+      "step": 6485
+    },
+    {
+      "epoch": 0.39407011361565103,
+      "grad_norm": 0.40401092171669006,
+      "learning_rate": 6.661685481386596e-05,
+      "loss": 1.1663,
+      "step": 6486
+    },
+    {
+      "epoch": 0.39413087064827756,
+      "grad_norm": 0.17111338675022125,
+      "learning_rate": 6.660782584632223e-05,
+      "loss": 1.1666,
+      "step": 6487
+    },
+    {
+      "epoch": 0.39419162768090404,
+      "grad_norm": 0.24630577862262726,
+      "learning_rate": 6.659879627001545e-05,
+      "loss": 1.1749,
+      "step": 6488
+    },
+    {
+      "epoch": 0.3942523847135306,
+      "grad_norm": 0.13845933973789215,
+      "learning_rate": 6.658976608527658e-05,
+      "loss": 1.0276,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3943131417461571,
+      "grad_norm": 0.18594589829444885,
+      "learning_rate": 6.658073529243664e-05,
+      "loss": 1.1567,
+      "step": 6490
+    },
+    {
+      "epoch": 0.39437389877878365,
+      "grad_norm": 0.14208967983722687,
+      "learning_rate": 6.657170389182663e-05,
+      "loss": 1.0712,
+      "step": 6491
+    },
+    {
+      "epoch": 0.3944346558114102,
+      "grad_norm": 0.3039787709712982,
+      "learning_rate": 6.656267188377762e-05,
+      "loss": 1.059,
+      "step": 6492
+    },
+    {
+      "epoch": 0.3944954128440367,
+      "grad_norm": 0.2600259780883789,
+      "learning_rate": 6.655363926862066e-05,
+      "loss": 1.1182,
+      "step": 6493
+    },
+    {
+      "epoch": 0.3945561698766632,
+      "grad_norm": 0.17212441563606262,
+      "learning_rate": 6.654460604668687e-05,
+      "loss": 1.1503,
+      "step": 6494
+    },
+    {
+      "epoch": 0.39461692690928973,
+      "grad_norm": 0.3530206084251404,
+      "learning_rate": 6.653557221830732e-05,
+      "loss": 1.223,
+      "step": 6495
+    },
+    {
+      "epoch": 0.39467768394191627,
+      "grad_norm": 0.18086989223957062,
+      "learning_rate": 6.652653778381319e-05,
+      "loss": 1.0511,
+      "step": 6496
+    },
+    {
+      "epoch": 0.3947384409745428,
+      "grad_norm": 0.14229127764701843,
+      "learning_rate": 6.651750274353563e-05,
+      "loss": 1.0806,
+      "step": 6497
+    },
+    {
+      "epoch": 0.39479919800716934,
+      "grad_norm": 2.082559108734131,
+      "learning_rate": 6.650846709780581e-05,
+      "loss": 1.0685,
+      "step": 6498
+    },
+    {
+      "epoch": 0.3948599550397959,
+      "grad_norm": 0.15213042497634888,
+      "learning_rate": 6.649943084695493e-05,
+      "loss": 1.0686,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3949207120724224,
+      "grad_norm": 0.12228461354970932,
+      "learning_rate": 6.649039399131423e-05,
+      "loss": 1.0573,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3949814691050489,
+      "grad_norm": 4.972786903381348,
+      "learning_rate": 6.648135653121494e-05,
+      "loss": 1.0946,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3950422261376754,
+      "grad_norm": 0.17178978025913239,
+      "learning_rate": 6.647231846698835e-05,
+      "loss": 1.0982,
+      "step": 6502
+    },
+    {
+      "epoch": 0.39510298317030196,
+      "grad_norm": 0.6011859178543091,
+      "learning_rate": 6.646327979896574e-05,
+      "loss": 1.076,
+      "step": 6503
+    },
+    {
+      "epoch": 0.3951637402029285,
+      "grad_norm": 0.24459262192249298,
+      "learning_rate": 6.645424052747843e-05,
+      "loss": 1.0556,
+      "step": 6504
+    },
+    {
+      "epoch": 0.39522449723555503,
+      "grad_norm": 0.15008269250392914,
+      "learning_rate": 6.644520065285775e-05,
+      "loss": 1.0435,
+      "step": 6505
+    },
+    {
+      "epoch": 0.39528525426818156,
+      "grad_norm": 0.20823614299297333,
+      "learning_rate": 6.643616017543506e-05,
+      "loss": 1.0401,
+      "step": 6506
+    },
+    {
+      "epoch": 0.39534601130080804,
+      "grad_norm": 0.16204649209976196,
+      "learning_rate": 6.642711909554174e-05,
+      "loss": 1.058,
+      "step": 6507
+    },
+    {
+      "epoch": 0.3954067683334346,
+      "grad_norm": 0.32570192217826843,
+      "learning_rate": 6.64180774135092e-05,
+      "loss": 1.0567,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3954675253660611,
+      "grad_norm": 0.14407196640968323,
+      "learning_rate": 6.640903512966885e-05,
+      "loss": 1.0811,
+      "step": 6509
+    },
+    {
+      "epoch": 0.39552828239868765,
+      "grad_norm": 0.16951946914196014,
+      "learning_rate": 6.639999224435217e-05,
+      "loss": 1.1062,
+      "step": 6510
+    },
+    {
+      "epoch": 0.3955890394313142,
+      "grad_norm": 5.580693244934082,
+      "learning_rate": 6.639094875789058e-05,
+      "loss": 1.1061,
+      "step": 6511
+    },
+    {
+      "epoch": 0.3956497964639407,
+      "grad_norm": 0.2052382528781891,
+      "learning_rate": 6.638190467061559e-05,
+      "loss": 1.2548,
+      "step": 6512
+    },
+    {
+      "epoch": 0.39571055349656725,
+      "grad_norm": 0.17212602496147156,
+      "learning_rate": 6.637285998285874e-05,
+      "loss": 1.0813,
+      "step": 6513
+    },
+    {
+      "epoch": 0.39577131052919373,
+      "grad_norm": 0.434367299079895,
+      "learning_rate": 6.636381469495152e-05,
+      "loss": 1.0953,
+      "step": 6514
+    },
+    {
+      "epoch": 0.39583206756182027,
+      "grad_norm": 0.12933877110481262,
+      "learning_rate": 6.635476880722553e-05,
+      "loss": 1.0243,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3958928245944468,
+      "grad_norm": 0.14470084011554718,
+      "learning_rate": 6.634572232001232e-05,
+      "loss": 1.0763,
+      "step": 6516
+    },
+    {
+      "epoch": 0.39595358162707334,
+      "grad_norm": 0.17749479413032532,
+      "learning_rate": 6.633667523364349e-05,
+      "loss": 1.129,
+      "step": 6517
+    },
+    {
+      "epoch": 0.39601433865969987,
+      "grad_norm": 0.19330710172653198,
+      "learning_rate": 6.63276275484507e-05,
+      "loss": 1.0798,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3960750956923264,
+      "grad_norm": 0.14301753044128418,
+      "learning_rate": 6.631857926476552e-05,
+      "loss": 1.0933,
+      "step": 6519
+    },
+    {
+      "epoch": 0.3961358527249529,
+      "grad_norm": 0.23434649407863617,
+      "learning_rate": 6.63095303829197e-05,
+      "loss": 1.1547,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3961966097575794,
+      "grad_norm": 0.1665913462638855,
+      "learning_rate": 6.630048090324488e-05,
+      "loss": 1.1146,
+      "step": 6521
+    },
+    {
+      "epoch": 0.39625736679020596,
+      "grad_norm": 0.15681713819503784,
+      "learning_rate": 6.629143082607278e-05,
+      "loss": 1.0566,
+      "step": 6522
+    },
+    {
+      "epoch": 0.3963181238228325,
+      "grad_norm": 0.16899584233760834,
+      "learning_rate": 6.628238015173514e-05,
+      "loss": 1.1174,
+      "step": 6523
+    },
+    {
+      "epoch": 0.396378880855459,
+      "grad_norm": 0.30240699648857117,
+      "learning_rate": 6.627332888056369e-05,
+      "loss": 1.0942,
+      "step": 6524
+    },
+    {
+      "epoch": 0.39643963788808556,
+      "grad_norm": 0.14742599427700043,
+      "learning_rate": 6.626427701289025e-05,
+      "loss": 1.1393,
+      "step": 6525
+    },
+    {
+      "epoch": 0.3965003949207121,
+      "grad_norm": 0.17558152973651886,
+      "learning_rate": 6.625522454904657e-05,
+      "loss": 1.0943,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3965611519533386,
+      "grad_norm": 0.17351263761520386,
+      "learning_rate": 6.62461714893645e-05,
+      "loss": 1.0987,
+      "step": 6527
+    },
+    {
+      "epoch": 0.3966219089859651,
+      "grad_norm": 0.14238056540489197,
+      "learning_rate": 6.623711783417587e-05,
+      "loss": 1.051,
+      "step": 6528
+    },
+    {
+      "epoch": 0.39668266601859165,
+      "grad_norm": 0.24085108935832977,
+      "learning_rate": 6.622806358381255e-05,
+      "loss": 1.1409,
+      "step": 6529
+    },
+    {
+      "epoch": 0.3967434230512182,
+      "grad_norm": 2.1114413738250732,
+      "learning_rate": 6.621900873860645e-05,
+      "loss": 1.1023,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3968041800838447,
+      "grad_norm": 0.23471508920192719,
+      "learning_rate": 6.620995329888942e-05,
+      "loss": 1.0999,
+      "step": 6531
+    },
+    {
+      "epoch": 0.39686493711647125,
+      "grad_norm": 8.686838150024414,
+      "learning_rate": 6.620089726499341e-05,
+      "loss": 1.1093,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3969256941490978,
+      "grad_norm": 0.24466344714164734,
+      "learning_rate": 6.61918406372504e-05,
+      "loss": 1.1733,
+      "step": 6533
+    },
+    {
+      "epoch": 0.39698645118172426,
+      "grad_norm": 0.150930255651474,
+      "learning_rate": 6.618278341599233e-05,
+      "loss": 1.0391,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3970472082143508,
+      "grad_norm": 0.16410765051841736,
+      "learning_rate": 6.61737256015512e-05,
+      "loss": 1.0606,
+      "step": 6535
+    },
+    {
+      "epoch": 0.39710796524697733,
+      "grad_norm": 0.16845226287841797,
+      "learning_rate": 6.616466719425905e-05,
+      "loss": 1.0905,
+      "step": 6536
+    },
+    {
+      "epoch": 0.39716872227960387,
+      "grad_norm": 0.23544728755950928,
+      "learning_rate": 6.61556081944479e-05,
+      "loss": 1.1362,
+      "step": 6537
+    },
+    {
+      "epoch": 0.3972294793122304,
+      "grad_norm": 0.12614917755126953,
+      "learning_rate": 6.61465486024498e-05,
+      "loss": 1.0817,
+      "step": 6538
+    },
+    {
+      "epoch": 0.39729023634485694,
+      "grad_norm": 0.16425821185112,
+      "learning_rate": 6.613748841859683e-05,
+      "loss": 1.1852,
+      "step": 6539
+    },
+    {
+      "epoch": 0.3973509933774834,
+      "grad_norm": 0.12663373351097107,
+      "learning_rate": 6.612842764322113e-05,
+      "loss": 1.0299,
+      "step": 6540
+    },
+    {
+      "epoch": 0.39741175041010995,
+      "grad_norm": 0.25890493392944336,
+      "learning_rate": 6.611936627665477e-05,
+      "loss": 1.1362,
+      "step": 6541
+    },
+    {
+      "epoch": 0.3974725074427365,
+      "grad_norm": 0.13185185194015503,
+      "learning_rate": 6.611030431922993e-05,
+      "loss": 1.072,
+      "step": 6542
+    },
+    {
+      "epoch": 0.397533264475363,
+      "grad_norm": 0.11852778494358063,
+      "learning_rate": 6.610124177127879e-05,
+      "loss": 1.0077,
+      "step": 6543
+    },
+    {
+      "epoch": 0.39759402150798956,
+      "grad_norm": 0.17963242530822754,
+      "learning_rate": 6.609217863313349e-05,
+      "loss": 1.0591,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3976547785406161,
+      "grad_norm": 0.12181134521961212,
+      "learning_rate": 6.608311490512629e-05,
+      "loss": 1.0382,
+      "step": 6545
+    },
+    {
+      "epoch": 0.39771553557324263,
+      "grad_norm": 0.18284626305103302,
+      "learning_rate": 6.60740505875894e-05,
+      "loss": 1.2584,
+      "step": 6546
+    },
+    {
+      "epoch": 0.3977762926058691,
+      "grad_norm": 0.13705766201019287,
+      "learning_rate": 6.606498568085508e-05,
+      "loss": 1.0503,
+      "step": 6547
+    },
+    {
+      "epoch": 0.39783704963849564,
+      "grad_norm": 0.18914468586444855,
+      "learning_rate": 6.60559201852556e-05,
+      "loss": 1.0995,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3978978066711222,
+      "grad_norm": 0.35786524415016174,
+      "learning_rate": 6.604685410112327e-05,
+      "loss": 1.1095,
+      "step": 6549
+    },
+    {
+      "epoch": 0.3979585637037487,
+      "grad_norm": 0.16245368123054504,
+      "learning_rate": 6.60377874287904e-05,
+      "loss": 1.0953,
+      "step": 6550
+    },
+    {
+      "epoch": 0.39801932073637525,
+      "grad_norm": 0.27149254083633423,
+      "learning_rate": 6.602872016858933e-05,
+      "loss": 1.0592,
+      "step": 6551
+    },
+    {
+      "epoch": 0.3980800777690018,
+      "grad_norm": 0.3707484006881714,
+      "learning_rate": 6.601965232085241e-05,
+      "loss": 1.081,
+      "step": 6552
+    },
+    {
+      "epoch": 0.39814083480162826,
+      "grad_norm": 0.30184659361839294,
+      "learning_rate": 6.601058388591206e-05,
+      "loss": 1.1586,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3982015918342548,
+      "grad_norm": 0.11948718875646591,
+      "learning_rate": 6.600151486410066e-05,
+      "loss": 1.0004,
+      "step": 6554
+    },
+    {
+      "epoch": 0.39826234886688133,
+      "grad_norm": 0.2058251053094864,
+      "learning_rate": 6.599244525575062e-05,
+      "loss": 1.0685,
+      "step": 6555
+    },
+    {
+      "epoch": 0.39832310589950787,
+      "grad_norm": 0.13420431315898895,
+      "learning_rate": 6.598337506119443e-05,
+      "loss": 1.1134,
+      "step": 6556
+    },
+    {
+      "epoch": 0.3983838629321344,
+      "grad_norm": 0.45619717240333557,
+      "learning_rate": 6.597430428076451e-05,
+      "loss": 1.1429,
+      "step": 6557
+    },
+    {
+      "epoch": 0.39844461996476094,
+      "grad_norm": 0.16612450778484344,
+      "learning_rate": 6.596523291479342e-05,
+      "loss": 1.1198,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3985053769973875,
+      "grad_norm": 0.23507781326770782,
+      "learning_rate": 6.59561609636136e-05,
+      "loss": 1.2081,
+      "step": 6559
+    },
+    {
+      "epoch": 0.39856613403001395,
+      "grad_norm": 0.1710253208875656,
+      "learning_rate": 6.594708842755762e-05,
+      "loss": 1.1253,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3986268910626405,
+      "grad_norm": 0.28977149724960327,
+      "learning_rate": 6.593801530695804e-05,
+      "loss": 1.1377,
+      "step": 6561
+    },
+    {
+      "epoch": 0.398687648095267,
+      "grad_norm": 0.16601623594760895,
+      "learning_rate": 6.592894160214743e-05,
+      "loss": 1.1475,
+      "step": 6562
+    },
+    {
+      "epoch": 0.39874840512789356,
+      "grad_norm": 0.3298317790031433,
+      "learning_rate": 6.591986731345839e-05,
+      "loss": 1.184,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3988091621605201,
+      "grad_norm": 0.139494389295578,
+      "learning_rate": 6.591079244122352e-05,
+      "loss": 1.0064,
+      "step": 6564
+    },
+    {
+      "epoch": 0.3988699191931466,
+      "grad_norm": 0.2404806911945343,
+      "learning_rate": 6.590171698577547e-05,
+      "loss": 1.1143,
+      "step": 6565
+    },
+    {
+      "epoch": 0.3989306762257731,
+      "grad_norm": 0.3489436209201813,
+      "learning_rate": 6.589264094744693e-05,
+      "loss": 1.0959,
+      "step": 6566
+    },
+    {
+      "epoch": 0.39899143325839964,
+      "grad_norm": 0.13315507769584656,
+      "learning_rate": 6.588356432657057e-05,
+      "loss": 1.1232,
+      "step": 6567
+    },
+    {
+      "epoch": 0.3990521902910262,
+      "grad_norm": 8.759673118591309,
+      "learning_rate": 6.587448712347907e-05,
+      "loss": 1.0895,
+      "step": 6568
+    },
+    {
+      "epoch": 0.3991129473236527,
+      "grad_norm": 0.21881161630153656,
+      "learning_rate": 6.586540933850518e-05,
+      "loss": 1.11,
+      "step": 6569
+    },
+    {
+      "epoch": 0.39917370435627925,
+      "grad_norm": 0.2691590189933777,
+      "learning_rate": 6.585633097198165e-05,
+      "loss": 1.0763,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3992344613889058,
+      "grad_norm": 0.20333918929100037,
+      "learning_rate": 6.584725202424125e-05,
+      "loss": 1.1471,
+      "step": 6571
+    },
+    {
+      "epoch": 0.3992952184215323,
+      "grad_norm": 0.2324604094028473,
+      "learning_rate": 6.583817249561675e-05,
+      "loss": 1.1277,
+      "step": 6572
+    },
+    {
+      "epoch": 0.3993559754541588,
+      "grad_norm": 0.2534354031085968,
+      "learning_rate": 6.582909238644098e-05,
+      "loss": 1.0799,
+      "step": 6573
+    },
+    {
+      "epoch": 0.39941673248678533,
+      "grad_norm": 0.23539762198925018,
+      "learning_rate": 6.582001169704676e-05,
+      "loss": 1.0613,
+      "step": 6574
+    },
+    {
+      "epoch": 0.39947748951941187,
+      "grad_norm": 0.14437204599380493,
+      "learning_rate": 6.581093042776699e-05,
+      "loss": 1.1095,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3995382465520384,
+      "grad_norm": 0.17713621258735657,
+      "learning_rate": 6.580184857893446e-05,
+      "loss": 1.0469,
+      "step": 6576
+    },
+    {
+      "epoch": 0.39959900358466494,
+      "grad_norm": 0.1882336437702179,
+      "learning_rate": 6.579276615088214e-05,
+      "loss": 1.0508,
+      "step": 6577
+    },
+    {
+      "epoch": 0.39965976061729147,
+      "grad_norm": 0.14026860892772675,
+      "learning_rate": 6.578368314394292e-05,
+      "loss": 1.0861,
+      "step": 6578
+    },
+    {
+      "epoch": 0.399720517649918,
+      "grad_norm": 0.2194347381591797,
+      "learning_rate": 6.577459955844976e-05,
+      "loss": 1.1215,
+      "step": 6579
+    },
+    {
+      "epoch": 0.3997812746825445,
+      "grad_norm": 0.22031378746032715,
+      "learning_rate": 6.576551539473559e-05,
+      "loss": 1.0821,
+      "step": 6580
+    },
+    {
+      "epoch": 0.399842031715171,
+      "grad_norm": 0.1323220133781433,
+      "learning_rate": 6.575643065313339e-05,
+      "loss": 1.0822,
+      "step": 6581
+    },
+    {
+      "epoch": 0.39990278874779756,
+      "grad_norm": 0.17923913896083832,
+      "learning_rate": 6.574734533397621e-05,
+      "loss": 1.0931,
+      "step": 6582
+    },
+    {
+      "epoch": 0.3999635457804241,
+      "grad_norm": 0.2591210603713989,
+      "learning_rate": 6.573825943759702e-05,
+      "loss": 1.0764,
+      "step": 6583
+    },
+    {
+      "epoch": 0.4000243028130506,
+      "grad_norm": 0.24421407282352448,
+      "learning_rate": 6.57291729643289e-05,
+      "loss": 1.038,
+      "step": 6584
+    },
+    {
+      "epoch": 0.40008505984567716,
+      "grad_norm": 0.2938361167907715,
+      "learning_rate": 6.57200859145049e-05,
+      "loss": 1.1638,
+      "step": 6585
+    },
+    {
+      "epoch": 0.40014581687830364,
+      "grad_norm": 0.3407941162586212,
+      "learning_rate": 6.57109982884581e-05,
+      "loss": 1.1792,
+      "step": 6586
+    },
+    {
+      "epoch": 0.4002065739109302,
+      "grad_norm": 0.15962469577789307,
+      "learning_rate": 6.570191008652165e-05,
+      "loss": 1.0725,
+      "step": 6587
+    },
+    {
+      "epoch": 0.4002673309435567,
+      "grad_norm": 0.202643483877182,
+      "learning_rate": 6.569282130902862e-05,
+      "loss": 1.1958,
+      "step": 6588
+    },
+    {
+      "epoch": 0.40032808797618324,
+      "grad_norm": 0.24373379349708557,
+      "learning_rate": 6.568373195631221e-05,
+      "loss": 1.085,
+      "step": 6589
+    },
+    {
+      "epoch": 0.4003888450088098,
+      "grad_norm": 0.250884085893631,
+      "learning_rate": 6.567464202870559e-05,
+      "loss": 1.0494,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4004496020414363,
+      "grad_norm": 0.14808085560798645,
+      "learning_rate": 6.566555152654189e-05,
+      "loss": 1.0627,
+      "step": 6591
+    },
+    {
+      "epoch": 0.40051035907406285,
+      "grad_norm": 0.1772744208574295,
+      "learning_rate": 6.56564604501544e-05,
+      "loss": 1.0672,
+      "step": 6592
+    },
+    {
+      "epoch": 0.40057111610668933,
+      "grad_norm": 0.15409553050994873,
+      "learning_rate": 6.564736879987632e-05,
+      "loss": 1.0838,
+      "step": 6593
+    },
+    {
+      "epoch": 0.40063187313931586,
+      "grad_norm": 0.22324049472808838,
+      "learning_rate": 6.563827657604092e-05,
+      "loss": 1.1507,
+      "step": 6594
+    },
+    {
+      "epoch": 0.4006926301719424,
+      "grad_norm": 0.15755148231983185,
+      "learning_rate": 6.562918377898146e-05,
+      "loss": 1.1539,
+      "step": 6595
+    },
+    {
+      "epoch": 0.40075338720456893,
+      "grad_norm": 0.14027974009513855,
+      "learning_rate": 6.562009040903124e-05,
+      "loss": 1.0186,
+      "step": 6596
+    },
+    {
+      "epoch": 0.40081414423719547,
+      "grad_norm": 0.20236752927303314,
+      "learning_rate": 6.561099646652361e-05,
+      "loss": 1.1087,
+      "step": 6597
+    },
+    {
+      "epoch": 0.400874901269822,
+      "grad_norm": 0.11700024455785751,
+      "learning_rate": 6.560190195179188e-05,
+      "loss": 1.0545,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4009356583024485,
+      "grad_norm": 0.20795938372612,
+      "learning_rate": 6.55928068651694e-05,
+      "loss": 1.1491,
+      "step": 6599
+    },
+    {
+      "epoch": 0.400996415335075,
+      "grad_norm": 0.4421660602092743,
+      "learning_rate": 6.55837112069896e-05,
+      "loss": 1.0629,
+      "step": 6600
+    },
+    {
+      "epoch": 0.40105717236770155,
+      "grad_norm": 0.2056204080581665,
+      "learning_rate": 6.557461497758583e-05,
+      "loss": 1.0939,
+      "step": 6601
+    },
+    {
+      "epoch": 0.4011179294003281,
+      "grad_norm": 0.366139680147171,
+      "learning_rate": 6.556551817729153e-05,
+      "loss": 1.1591,
+      "step": 6602
+    },
+    {
+      "epoch": 0.4011786864329546,
+      "grad_norm": 0.14360135793685913,
+      "learning_rate": 6.555642080644018e-05,
+      "loss": 1.0735,
+      "step": 6603
+    },
+    {
+      "epoch": 0.40123944346558116,
+      "grad_norm": 0.1836598813533783,
+      "learning_rate": 6.55473228653652e-05,
+      "loss": 1.1618,
+      "step": 6604
+    },
+    {
+      "epoch": 0.4013002004982077,
+      "grad_norm": 0.26345452666282654,
+      "learning_rate": 6.55382243544001e-05,
+      "loss": 1.0369,
+      "step": 6605
+    },
+    {
+      "epoch": 0.4013609575308342,
+      "grad_norm": 0.15187640488147736,
+      "learning_rate": 6.552912527387838e-05,
+      "loss": 1.099,
+      "step": 6606
+    },
+    {
+      "epoch": 0.4014217145634607,
+      "grad_norm": 0.16972950100898743,
+      "learning_rate": 6.552002562413358e-05,
+      "loss": 1.0588,
+      "step": 6607
+    },
+    {
+      "epoch": 0.40148247159608724,
+      "grad_norm": 0.25443100929260254,
+      "learning_rate": 6.551092540549923e-05,
+      "loss": 1.0541,
+      "step": 6608
+    },
+    {
+      "epoch": 0.4015432286287138,
+      "grad_norm": 0.19845768809318542,
+      "learning_rate": 6.550182461830892e-05,
+      "loss": 1.0847,
+      "step": 6609
+    },
+    {
+      "epoch": 0.4016039856613403,
+      "grad_norm": 0.16475890576839447,
+      "learning_rate": 6.549272326289623e-05,
+      "loss": 1.1807,
+      "step": 6610
+    },
+    {
+      "epoch": 0.40166474269396685,
+      "grad_norm": 0.22839957475662231,
+      "learning_rate": 6.548362133959478e-05,
+      "loss": 1.1024,
+      "step": 6611
+    },
+    {
+      "epoch": 0.4017254997265933,
+      "grad_norm": 0.2388787716627121,
+      "learning_rate": 6.547451884873817e-05,
+      "loss": 1.1277,
+      "step": 6612
+    },
+    {
+      "epoch": 0.40178625675921986,
+      "grad_norm": 0.18775707483291626,
+      "learning_rate": 6.54654157906601e-05,
+      "loss": 1.0862,
+      "step": 6613
+    },
+    {
+      "epoch": 0.4018470137918464,
+      "grad_norm": 0.32869967818260193,
+      "learning_rate": 6.545631216569423e-05,
+      "loss": 1.1068,
+      "step": 6614
+    },
+    {
+      "epoch": 0.40190777082447293,
+      "grad_norm": 0.18017113208770752,
+      "learning_rate": 6.544720797417426e-05,
+      "loss": 1.0704,
+      "step": 6615
+    },
+    {
+      "epoch": 0.40196852785709947,
+      "grad_norm": 0.3624647855758667,
+      "learning_rate": 6.543810321643386e-05,
+      "loss": 1.0809,
+      "step": 6616
+    },
+    {
+      "epoch": 0.402029284889726,
+      "grad_norm": 0.1427358090877533,
+      "learning_rate": 6.542899789280684e-05,
+      "loss": 1.0692,
+      "step": 6617
+    },
+    {
+      "epoch": 0.40209004192235254,
+      "grad_norm": 0.20363053679466248,
+      "learning_rate": 6.541989200362689e-05,
+      "loss": 1.1052,
+      "step": 6618
+    },
+    {
+      "epoch": 0.402150798954979,
+      "grad_norm": 0.360845685005188,
+      "learning_rate": 6.541078554922784e-05,
+      "loss": 1.1304,
+      "step": 6619
+    },
+    {
+      "epoch": 0.40221155598760555,
+      "grad_norm": 0.13105599582195282,
+      "learning_rate": 6.540167852994346e-05,
+      "loss": 1.0792,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4022723130202321,
+      "grad_norm": 0.14443010091781616,
+      "learning_rate": 6.539257094610757e-05,
+      "loss": 1.0527,
+      "step": 6621
+    },
+    {
+      "epoch": 0.4023330700528586,
+      "grad_norm": 0.16193968057632446,
+      "learning_rate": 6.538346279805401e-05,
+      "loss": 1.108,
+      "step": 6622
+    },
+    {
+      "epoch": 0.40239382708548516,
+      "grad_norm": 0.22881177067756653,
+      "learning_rate": 6.537435408611667e-05,
+      "loss": 1.1628,
+      "step": 6623
+    },
+    {
+      "epoch": 0.4024545841181117,
+      "grad_norm": 0.24028386175632477,
+      "learning_rate": 6.53652448106294e-05,
+      "loss": 1.0347,
+      "step": 6624
+    },
+    {
+      "epoch": 0.40251534115073817,
+      "grad_norm": 0.1537184864282608,
+      "learning_rate": 6.53561349719261e-05,
+      "loss": 1.1427,
+      "step": 6625
+    },
+    {
+      "epoch": 0.4025760981833647,
+      "grad_norm": 0.28651416301727295,
+      "learning_rate": 6.534702457034071e-05,
+      "loss": 1.1191,
+      "step": 6626
+    },
+    {
+      "epoch": 0.40263685521599124,
+      "grad_norm": 0.23378139734268188,
+      "learning_rate": 6.533791360620717e-05,
+      "loss": 1.0672,
+      "step": 6627
+    },
+    {
+      "epoch": 0.4026976122486178,
+      "grad_norm": 0.19313105940818787,
+      "learning_rate": 6.532880207985945e-05,
+      "loss": 1.1301,
+      "step": 6628
+    },
+    {
+      "epoch": 0.4027583692812443,
+      "grad_norm": 0.20975559949874878,
+      "learning_rate": 6.531968999163151e-05,
+      "loss": 1.0649,
+      "step": 6629
+    },
+    {
+      "epoch": 0.40281912631387085,
+      "grad_norm": 0.2829676568508148,
+      "learning_rate": 6.53105773418574e-05,
+      "loss": 1.2322,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4028798833464974,
+      "grad_norm": 0.15994608402252197,
+      "learning_rate": 6.53014641308711e-05,
+      "loss": 1.0822,
+      "step": 6631
+    },
+    {
+      "epoch": 0.40294064037912386,
+      "grad_norm": 0.22853371500968933,
+      "learning_rate": 6.529235035900667e-05,
+      "loss": 1.0704,
+      "step": 6632
+    },
+    {
+      "epoch": 0.4030013974117504,
+      "grad_norm": 0.1920643150806427,
+      "learning_rate": 6.528323602659817e-05,
+      "loss": 1.1006,
+      "step": 6633
+    },
+    {
+      "epoch": 0.40306215444437693,
+      "grad_norm": 0.17940761148929596,
+      "learning_rate": 6.527412113397973e-05,
+      "loss": 1.142,
+      "step": 6634
+    },
+    {
+      "epoch": 0.40312291147700346,
+      "grad_norm": 0.13779808580875397,
+      "learning_rate": 6.52650056814854e-05,
+      "loss": 1.0438,
+      "step": 6635
+    },
+    {
+      "epoch": 0.40318366850963,
+      "grad_norm": 0.21858450770378113,
+      "learning_rate": 6.525588966944936e-05,
+      "loss": 1.0789,
+      "step": 6636
+    },
+    {
+      "epoch": 0.40324442554225653,
+      "grad_norm": 0.34694162011146545,
+      "learning_rate": 6.524677309820571e-05,
+      "loss": 1.0641,
+      "step": 6637
+    },
+    {
+      "epoch": 0.40330518257488307,
+      "grad_norm": 0.12775734066963196,
+      "learning_rate": 6.523765596808866e-05,
+      "loss": 1.0867,
+      "step": 6638
+    },
+    {
+      "epoch": 0.40336593960750955,
+      "grad_norm": 0.3817265033721924,
+      "learning_rate": 6.52285382794324e-05,
+      "loss": 1.0058,
+      "step": 6639
+    },
+    {
+      "epoch": 0.4034266966401361,
+      "grad_norm": 0.2208240032196045,
+      "learning_rate": 6.52194200325711e-05,
+      "loss": 1.0304,
+      "step": 6640
+    },
+    {
+      "epoch": 0.4034874536727626,
+      "grad_norm": 0.13443036377429962,
+      "learning_rate": 6.521030122783903e-05,
+      "loss": 1.0481,
+      "step": 6641
+    },
+    {
+      "epoch": 0.40354821070538915,
+      "grad_norm": 0.20626302063465118,
+      "learning_rate": 6.520118186557043e-05,
+      "loss": 1.0605,
+      "step": 6642
+    },
+    {
+      "epoch": 0.4036089677380157,
+      "grad_norm": 0.20362652838230133,
+      "learning_rate": 6.519206194609954e-05,
+      "loss": 1.0659,
+      "step": 6643
+    },
+    {
+      "epoch": 0.4036697247706422,
+      "grad_norm": 0.14869822561740875,
+      "learning_rate": 6.518294146976073e-05,
+      "loss": 1.0318,
+      "step": 6644
+    },
+    {
+      "epoch": 0.4037304818032687,
+      "grad_norm": 0.149115189909935,
+      "learning_rate": 6.517382043688823e-05,
+      "loss": 1.0876,
+      "step": 6645
+    },
+    {
+      "epoch": 0.40379123883589524,
+      "grad_norm": 0.1736573874950409,
+      "learning_rate": 6.516469884781644e-05,
+      "loss": 1.0794,
+      "step": 6646
+    },
+    {
+      "epoch": 0.4038519958685218,
+      "grad_norm": 0.2508244514465332,
+      "learning_rate": 6.515557670287966e-05,
+      "loss": 1.1283,
+      "step": 6647
+    },
+    {
+      "epoch": 0.4039127529011483,
+      "grad_norm": 0.20715117454528809,
+      "learning_rate": 6.514645400241229e-05,
+      "loss": 1.0799,
+      "step": 6648
+    },
+    {
+      "epoch": 0.40397350993377484,
+      "grad_norm": 0.6394730806350708,
+      "learning_rate": 6.513733074674872e-05,
+      "loss": 1.2992,
+      "step": 6649
+    },
+    {
+      "epoch": 0.4040342669664014,
+      "grad_norm": 0.21538077294826508,
+      "learning_rate": 6.512820693622337e-05,
+      "loss": 1.159,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4040950239990279,
+      "grad_norm": 0.43514516949653625,
+      "learning_rate": 6.511908257117067e-05,
+      "loss": 1.1446,
+      "step": 6651
+    },
+    {
+      "epoch": 0.4041557810316544,
+      "grad_norm": 0.21478870511054993,
+      "learning_rate": 6.510995765192508e-05,
+      "loss": 1.0499,
+      "step": 6652
+    },
+    {
+      "epoch": 0.40421653806428093,
+      "grad_norm": 0.49485352635383606,
+      "learning_rate": 6.510083217882106e-05,
+      "loss": 1.211,
+      "step": 6653
+    },
+    {
+      "epoch": 0.40427729509690746,
+      "grad_norm": 0.17889845371246338,
+      "learning_rate": 6.509170615219313e-05,
+      "loss": 1.1264,
+      "step": 6654
+    },
+    {
+      "epoch": 0.404338052129534,
+      "grad_norm": 0.15409870445728302,
+      "learning_rate": 6.508257957237578e-05,
+      "loss": 1.2703,
+      "step": 6655
+    },
+    {
+      "epoch": 0.40439880916216053,
+      "grad_norm": 0.16098731756210327,
+      "learning_rate": 6.507345243970356e-05,
+      "loss": 1.0317,
+      "step": 6656
+    },
+    {
+      "epoch": 0.40445956619478707,
+      "grad_norm": 2.5118050575256348,
+      "learning_rate": 6.506432475451106e-05,
+      "loss": 1.0563,
+      "step": 6657
+    },
+    {
+      "epoch": 0.40452032322741355,
+      "grad_norm": 0.17978206276893616,
+      "learning_rate": 6.505519651713277e-05,
+      "loss": 1.0711,
+      "step": 6658
+    },
+    {
+      "epoch": 0.4045810802600401,
+      "grad_norm": 0.2846367657184601,
+      "learning_rate": 6.504606772790338e-05,
+      "loss": 1.1987,
+      "step": 6659
+    },
+    {
+      "epoch": 0.4046418372926666,
+      "grad_norm": 0.12307152152061462,
+      "learning_rate": 6.503693838715748e-05,
+      "loss": 1.0401,
+      "step": 6660
+    },
+    {
+      "epoch": 0.40470259432529315,
+      "grad_norm": 0.2261846661567688,
+      "learning_rate": 6.502780849522965e-05,
+      "loss": 1.1381,
+      "step": 6661
+    },
+    {
+      "epoch": 0.4047633513579197,
+      "grad_norm": 0.16556531190872192,
+      "learning_rate": 6.501867805245463e-05,
+      "loss": 1.0702,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4048241083905462,
+      "grad_norm": 0.42099320888519287,
+      "learning_rate": 6.500954705916706e-05,
+      "loss": 1.1616,
+      "step": 6663
+    },
+    {
+      "epoch": 0.40488486542317276,
+      "grad_norm": 0.17864783108234406,
+      "learning_rate": 6.500041551570163e-05,
+      "loss": 1.1111,
+      "step": 6664
+    },
+    {
+      "epoch": 0.40494562245579924,
+      "grad_norm": 0.19972102344036102,
+      "learning_rate": 6.499128342239307e-05,
+      "loss": 1.2066,
+      "step": 6665
+    },
+    {
+      "epoch": 0.40500637948842577,
+      "grad_norm": 0.1650223284959793,
+      "learning_rate": 6.498215077957612e-05,
+      "loss": 1.1443,
+      "step": 6666
+    },
+    {
+      "epoch": 0.4050671365210523,
+      "grad_norm": 0.28008827567100525,
+      "learning_rate": 6.497301758758557e-05,
+      "loss": 1.2032,
+      "step": 6667
+    },
+    {
+      "epoch": 0.40512789355367884,
+      "grad_norm": 0.21093466877937317,
+      "learning_rate": 6.496388384675613e-05,
+      "loss": 1.1249,
+      "step": 6668
+    },
+    {
+      "epoch": 0.4051886505863054,
+      "grad_norm": 0.22220522165298462,
+      "learning_rate": 6.495474955742263e-05,
+      "loss": 1.1426,
+      "step": 6669
+    },
+    {
+      "epoch": 0.4052494076189319,
+      "grad_norm": 0.13069680333137512,
+      "learning_rate": 6.494561471991991e-05,
+      "loss": 1.0801,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4053101646515584,
+      "grad_norm": 0.15735410153865814,
+      "learning_rate": 6.493647933458278e-05,
+      "loss": 1.0833,
+      "step": 6671
+    },
+    {
+      "epoch": 0.4053709216841849,
+      "grad_norm": 0.1487436592578888,
+      "learning_rate": 6.492734340174614e-05,
+      "loss": 1.0503,
+      "step": 6672
+    },
+    {
+      "epoch": 0.40543167871681146,
+      "grad_norm": 0.13360805809497833,
+      "learning_rate": 6.491820692174481e-05,
+      "loss": 1.013,
+      "step": 6673
+    },
+    {
+      "epoch": 0.405492435749438,
+      "grad_norm": 0.17869211733341217,
+      "learning_rate": 6.490906989491375e-05,
+      "loss": 1.06,
+      "step": 6674
+    },
+    {
+      "epoch": 0.40555319278206453,
+      "grad_norm": 0.1763959378004074,
+      "learning_rate": 6.489993232158783e-05,
+      "loss": 1.1478,
+      "step": 6675
+    },
+    {
+      "epoch": 0.40561394981469107,
+      "grad_norm": 0.2103027105331421,
+      "learning_rate": 6.489079420210203e-05,
+      "loss": 1.2345,
+      "step": 6676
+    },
+    {
+      "epoch": 0.4056747068473176,
+      "grad_norm": 0.20595324039459229,
+      "learning_rate": 6.488165553679128e-05,
+      "loss": 1.144,
+      "step": 6677
+    },
+    {
+      "epoch": 0.4057354638799441,
+      "grad_norm": 0.21653783321380615,
+      "learning_rate": 6.487251632599056e-05,
+      "loss": 1.1657,
+      "step": 6678
+    },
+    {
+      "epoch": 0.4057962209125706,
+      "grad_norm": 0.37403324246406555,
+      "learning_rate": 6.486337657003489e-05,
+      "loss": 1.1102,
+      "step": 6679
+    },
+    {
+      "epoch": 0.40585697794519715,
+      "grad_norm": 0.8291086554527283,
+      "learning_rate": 6.485423626925928e-05,
+      "loss": 1.1297,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4059177349778237,
+      "grad_norm": 0.19228830933570862,
+      "learning_rate": 6.484509542399876e-05,
+      "loss": 1.0228,
+      "step": 6681
+    },
+    {
+      "epoch": 0.4059784920104502,
+      "grad_norm": 0.21620532870292664,
+      "learning_rate": 6.48359540345884e-05,
+      "loss": 1.0831,
+      "step": 6682
+    },
+    {
+      "epoch": 0.40603924904307676,
+      "grad_norm": 0.19082558155059814,
+      "learning_rate": 6.482681210136328e-05,
+      "loss": 1.2095,
+      "step": 6683
+    },
+    {
+      "epoch": 0.4061000060757033,
+      "grad_norm": 0.1522844433784485,
+      "learning_rate": 6.48176696246585e-05,
+      "loss": 1.1178,
+      "step": 6684
+    },
+    {
+      "epoch": 0.40616076310832977,
+      "grad_norm": 3.2175731658935547,
+      "learning_rate": 6.480852660480918e-05,
+      "loss": 1.1549,
+      "step": 6685
+    },
+    {
+      "epoch": 0.4062215201409563,
+      "grad_norm": 0.1618947833776474,
+      "learning_rate": 6.479938304215043e-05,
+      "loss": 1.0719,
+      "step": 6686
+    },
+    {
+      "epoch": 0.40628227717358284,
+      "grad_norm": 0.16067153215408325,
+      "learning_rate": 6.479023893701745e-05,
+      "loss": 1.0778,
+      "step": 6687
+    },
+    {
+      "epoch": 0.4063430342062094,
+      "grad_norm": 0.1437070518732071,
+      "learning_rate": 6.478109428974542e-05,
+      "loss": 1.1159,
+      "step": 6688
+    },
+    {
+      "epoch": 0.4064037912388359,
+      "grad_norm": 0.3020385503768921,
+      "learning_rate": 6.47719491006695e-05,
+      "loss": 0.999,
+      "step": 6689
+    },
+    {
+      "epoch": 0.40646454827146244,
+      "grad_norm": 0.15864762663841248,
+      "learning_rate": 6.476280337012494e-05,
+      "loss": 1.1149,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4065253053040889,
+      "grad_norm": 1.2274796962738037,
+      "learning_rate": 6.475365709844697e-05,
+      "loss": 1.0734,
+      "step": 6691
+    },
+    {
+      "epoch": 0.40658606233671546,
+      "grad_norm": 0.17850178480148315,
+      "learning_rate": 6.474451028597083e-05,
+      "loss": 0.9927,
+      "step": 6692
+    },
+    {
+      "epoch": 0.406646819369342,
+      "grad_norm": 0.1862597018480301,
+      "learning_rate": 6.473536293303182e-05,
+      "loss": 1.0516,
+      "step": 6693
+    },
+    {
+      "epoch": 0.40670757640196853,
+      "grad_norm": 0.30372729897499084,
+      "learning_rate": 6.472621503996525e-05,
+      "loss": 1.1004,
+      "step": 6694
+    },
+    {
+      "epoch": 0.40676833343459506,
+      "grad_norm": 0.17860114574432373,
+      "learning_rate": 6.471706660710641e-05,
+      "loss": 1.1441,
+      "step": 6695
+    },
+    {
+      "epoch": 0.4068290904672216,
+      "grad_norm": 0.14600655436515808,
+      "learning_rate": 6.470791763479065e-05,
+      "loss": 1.0836,
+      "step": 6696
+    },
+    {
+      "epoch": 0.40688984749984813,
+      "grad_norm": 0.45386990904808044,
+      "learning_rate": 6.469876812335333e-05,
+      "loss": 1.1657,
+      "step": 6697
+    },
+    {
+      "epoch": 0.4069506045324746,
+      "grad_norm": 0.197322815656662,
+      "learning_rate": 6.468961807312982e-05,
+      "loss": 1.1366,
+      "step": 6698
+    },
+    {
+      "epoch": 0.40701136156510115,
+      "grad_norm": 0.1556454449892044,
+      "learning_rate": 6.468046748445552e-05,
+      "loss": 1.0545,
+      "step": 6699
+    },
+    {
+      "epoch": 0.4070721185977277,
+      "grad_norm": 0.12340997159481049,
+      "learning_rate": 6.467131635766584e-05,
+      "loss": 1.1206,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4071328756303542,
+      "grad_norm": 0.2479964643716812,
+      "learning_rate": 6.466216469309624e-05,
+      "loss": 1.0821,
+      "step": 6701
+    },
+    {
+      "epoch": 0.40719363266298075,
+      "grad_norm": 0.15532584488391876,
+      "learning_rate": 6.465301249108214e-05,
+      "loss": 1.0491,
+      "step": 6702
+    },
+    {
+      "epoch": 0.4072543896956073,
+      "grad_norm": 0.21279531717300415,
+      "learning_rate": 6.464385975195905e-05,
+      "loss": 1.1038,
+      "step": 6703
+    },
+    {
+      "epoch": 0.40731514672823377,
+      "grad_norm": 0.19126805663108826,
+      "learning_rate": 6.463470647606244e-05,
+      "loss": 1.1087,
+      "step": 6704
+    },
+    {
+      "epoch": 0.4073759037608603,
+      "grad_norm": 0.22247013449668884,
+      "learning_rate": 6.462555266372784e-05,
+      "loss": 1.1314,
+      "step": 6705
+    },
+    {
+      "epoch": 0.40743666079348684,
+      "grad_norm": 0.22831468284130096,
+      "learning_rate": 6.461639831529078e-05,
+      "loss": 1.0675,
+      "step": 6706
+    },
+    {
+      "epoch": 0.4074974178261134,
+      "grad_norm": 0.12370583415031433,
+      "learning_rate": 6.460724343108681e-05,
+      "loss": 1.031,
+      "step": 6707
+    },
+    {
+      "epoch": 0.4075581748587399,
+      "grad_norm": 0.1619083732366562,
+      "learning_rate": 6.459808801145151e-05,
+      "loss": 1.096,
+      "step": 6708
+    },
+    {
+      "epoch": 0.40761893189136644,
+      "grad_norm": 0.19678902626037598,
+      "learning_rate": 6.458893205672047e-05,
+      "loss": 1.0405,
+      "step": 6709
+    },
+    {
+      "epoch": 0.407679688923993,
+      "grad_norm": 0.2028418779373169,
+      "learning_rate": 6.457977556722931e-05,
+      "loss": 1.0651,
+      "step": 6710
+    },
+    {
+      "epoch": 0.40774044595661946,
+      "grad_norm": 0.16231292486190796,
+      "learning_rate": 6.457061854331368e-05,
+      "loss": 1.0691,
+      "step": 6711
+    },
+    {
+      "epoch": 0.407801202989246,
+      "grad_norm": 0.8556142449378967,
+      "learning_rate": 6.456146098530918e-05,
+      "loss": 1.2452,
+      "step": 6712
+    },
+    {
+      "epoch": 0.4078619600218725,
+      "grad_norm": 0.15859106183052063,
+      "learning_rate": 6.455230289355151e-05,
+      "loss": 1.027,
+      "step": 6713
+    },
+    {
+      "epoch": 0.40792271705449906,
+      "grad_norm": 0.329671710729599,
+      "learning_rate": 6.454314426837639e-05,
+      "loss": 1.1167,
+      "step": 6714
+    },
+    {
+      "epoch": 0.4079834740871256,
+      "grad_norm": 0.3965393304824829,
+      "learning_rate": 6.453398511011948e-05,
+      "loss": 1.1326,
+      "step": 6715
+    },
+    {
+      "epoch": 0.40804423111975213,
+      "grad_norm": 0.15830573439598083,
+      "learning_rate": 6.452482541911656e-05,
+      "loss": 1.1237,
+      "step": 6716
+    },
+    {
+      "epoch": 0.4081049881523786,
+      "grad_norm": 0.1571277678012848,
+      "learning_rate": 6.451566519570336e-05,
+      "loss": 1.0272,
+      "step": 6717
+    },
+    {
+      "epoch": 0.40816574518500515,
+      "grad_norm": 0.1618984192609787,
+      "learning_rate": 6.450650444021561e-05,
+      "loss": 1.1627,
+      "step": 6718
+    },
+    {
+      "epoch": 0.4082265022176317,
+      "grad_norm": 0.19462212920188904,
+      "learning_rate": 6.449734315298917e-05,
+      "loss": 1.0825,
+      "step": 6719
+    },
+    {
+      "epoch": 0.4082872592502582,
+      "grad_norm": 0.24113252758979797,
+      "learning_rate": 6.448818133435978e-05,
+      "loss": 1.0641,
+      "step": 6720
+    },
+    {
+      "epoch": 0.40834801628288475,
+      "grad_norm": 0.15695524215698242,
+      "learning_rate": 6.447901898466335e-05,
+      "loss": 1.1029,
+      "step": 6721
+    },
+    {
+      "epoch": 0.4084087733155113,
+      "grad_norm": 0.2879861295223236,
+      "learning_rate": 6.446985610423565e-05,
+      "loss": 1.1403,
+      "step": 6722
+    },
+    {
+      "epoch": 0.4084695303481378,
+      "grad_norm": 0.15667253732681274,
+      "learning_rate": 6.446069269341257e-05,
+      "loss": 1.1269,
+      "step": 6723
+    },
+    {
+      "epoch": 0.4085302873807643,
+      "grad_norm": 1.1046885251998901,
+      "learning_rate": 6.445152875253002e-05,
+      "loss": 1.0691,
+      "step": 6724
+    },
+    {
+      "epoch": 0.40859104441339084,
+      "grad_norm": 0.2713603675365448,
+      "learning_rate": 6.44423642819239e-05,
+      "loss": 1.1142,
+      "step": 6725
+    },
+    {
+      "epoch": 0.40865180144601737,
+      "grad_norm": 0.22601619362831116,
+      "learning_rate": 6.443319928193011e-05,
+      "loss": 1.0528,
+      "step": 6726
+    },
+    {
+      "epoch": 0.4087125584786439,
+      "grad_norm": 0.16097596287727356,
+      "learning_rate": 6.442403375288462e-05,
+      "loss": 1.0478,
+      "step": 6727
+    },
+    {
+      "epoch": 0.40877331551127044,
+      "grad_norm": 0.30255916714668274,
+      "learning_rate": 6.441486769512336e-05,
+      "loss": 1.0845,
+      "step": 6728
+    },
+    {
+      "epoch": 0.408834072543897,
+      "grad_norm": 0.20107612013816833,
+      "learning_rate": 6.440570110898236e-05,
+      "loss": 1.1634,
+      "step": 6729
+    },
+    {
+      "epoch": 0.40889482957652346,
+      "grad_norm": 0.20882728695869446,
+      "learning_rate": 6.43965339947976e-05,
+      "loss": 1.2285,
+      "step": 6730
+    },
+    {
+      "epoch": 0.40895558660915,
+      "grad_norm": 0.2033892571926117,
+      "learning_rate": 6.43873663529051e-05,
+      "loss": 1.144,
+      "step": 6731
+    },
+    {
+      "epoch": 0.4090163436417765,
+      "grad_norm": 0.34708642959594727,
+      "learning_rate": 6.43781981836409e-05,
+      "loss": 1.1474,
+      "step": 6732
+    },
+    {
+      "epoch": 0.40907710067440306,
+      "grad_norm": 0.20012971758842468,
+      "learning_rate": 6.436902948734107e-05,
+      "loss": 1.0641,
+      "step": 6733
+    },
+    {
+      "epoch": 0.4091378577070296,
+      "grad_norm": 0.11697956919670105,
+      "learning_rate": 6.43598602643417e-05,
+      "loss": 1.0262,
+      "step": 6734
+    },
+    {
+      "epoch": 0.40919861473965613,
+      "grad_norm": 0.21744292974472046,
+      "learning_rate": 6.435069051497886e-05,
+      "loss": 1.136,
+      "step": 6735
+    },
+    {
+      "epoch": 0.40925937177228267,
+      "grad_norm": 0.3735305070877075,
+      "learning_rate": 6.434152023958869e-05,
+      "loss": 1.0168,
+      "step": 6736
+    },
+    {
+      "epoch": 0.40932012880490914,
+      "grad_norm": 0.2032550722360611,
+      "learning_rate": 6.433234943850733e-05,
+      "loss": 1.0121,
+      "step": 6737
+    },
+    {
+      "epoch": 0.4093808858375357,
+      "grad_norm": 18.607328414916992,
+      "learning_rate": 6.432317811207091e-05,
+      "loss": 1.1331,
+      "step": 6738
+    },
+    {
+      "epoch": 0.4094416428701622,
+      "grad_norm": 0.3275555372238159,
+      "learning_rate": 6.431400626061565e-05,
+      "loss": 1.144,
+      "step": 6739
+    },
+    {
+      "epoch": 0.40950239990278875,
+      "grad_norm": 0.2409260869026184,
+      "learning_rate": 6.430483388447772e-05,
+      "loss": 1.0839,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4095631569354153,
+      "grad_norm": 0.1464911252260208,
+      "learning_rate": 6.429566098399334e-05,
+      "loss": 1.0603,
+      "step": 6741
+    },
+    {
+      "epoch": 0.4096239139680418,
+      "grad_norm": 0.2618921995162964,
+      "learning_rate": 6.428648755949873e-05,
+      "loss": 1.1141,
+      "step": 6742
+    },
+    {
+      "epoch": 0.40968467100066835,
+      "grad_norm": 0.35685738921165466,
+      "learning_rate": 6.427731361133016e-05,
+      "loss": 1.1225,
+      "step": 6743
+    },
+    {
+      "epoch": 0.40974542803329483,
+      "grad_norm": 0.2647813558578491,
+      "learning_rate": 6.426813913982392e-05,
+      "loss": 1.2056,
+      "step": 6744
+    },
+    {
+      "epoch": 0.40980618506592137,
+      "grad_norm": 0.15585041046142578,
+      "learning_rate": 6.425896414531627e-05,
+      "loss": 1.1543,
+      "step": 6745
+    },
+    {
+      "epoch": 0.4098669420985479,
+      "grad_norm": 0.19152897596359253,
+      "learning_rate": 6.424978862814352e-05,
+      "loss": 1.1875,
+      "step": 6746
+    },
+    {
+      "epoch": 0.40992769913117444,
+      "grad_norm": 0.1646856963634491,
+      "learning_rate": 6.424061258864203e-05,
+      "loss": 1.0461,
+      "step": 6747
+    },
+    {
+      "epoch": 0.409988456163801,
+      "grad_norm": 0.17298170924186707,
+      "learning_rate": 6.423143602714813e-05,
+      "loss": 1.0966,
+      "step": 6748
+    },
+    {
+      "epoch": 0.4100492131964275,
+      "grad_norm": 0.12542447447776794,
+      "learning_rate": 6.422225894399818e-05,
+      "loss": 1.0465,
+      "step": 6749
+    },
+    {
+      "epoch": 0.410109970229054,
+      "grad_norm": 0.1578681766986847,
+      "learning_rate": 6.421308133952858e-05,
+      "loss": 1.0543,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4101707272616805,
+      "grad_norm": 0.30167078971862793,
+      "learning_rate": 6.420390321407575e-05,
+      "loss": 1.1853,
+      "step": 6751
+    },
+    {
+      "epoch": 0.41023148429430706,
+      "grad_norm": 0.24929563701152802,
+      "learning_rate": 6.41947245679761e-05,
+      "loss": 1.1201,
+      "step": 6752
+    },
+    {
+      "epoch": 0.4102922413269336,
+      "grad_norm": 0.24105627834796906,
+      "learning_rate": 6.418554540156607e-05,
+      "loss": 1.1356,
+      "step": 6753
+    },
+    {
+      "epoch": 0.41035299835956013,
+      "grad_norm": 0.15548865497112274,
+      "learning_rate": 6.417636571518213e-05,
+      "loss": 1.0628,
+      "step": 6754
+    },
+    {
+      "epoch": 0.41041375539218666,
+      "grad_norm": 0.17175735533237457,
+      "learning_rate": 6.416718550916078e-05,
+      "loss": 1.1799,
+      "step": 6755
+    },
+    {
+      "epoch": 0.4104745124248132,
+      "grad_norm": 0.24108925461769104,
+      "learning_rate": 6.415800478383849e-05,
+      "loss": 1.0849,
+      "step": 6756
+    },
+    {
+      "epoch": 0.4105352694574397,
+      "grad_norm": 2.870900869369507,
+      "learning_rate": 6.41488235395518e-05,
+      "loss": 1.1703,
+      "step": 6757
+    },
+    {
+      "epoch": 0.4105960264900662,
+      "grad_norm": 0.19000382721424103,
+      "learning_rate": 6.413964177663726e-05,
+      "loss": 1.0626,
+      "step": 6758
+    },
+    {
+      "epoch": 0.41065678352269275,
+      "grad_norm": 0.28872933983802795,
+      "learning_rate": 6.41304594954314e-05,
+      "loss": 1.2691,
+      "step": 6759
+    },
+    {
+      "epoch": 0.4107175405553193,
+      "grad_norm": 0.15360146760940552,
+      "learning_rate": 6.412127669627083e-05,
+      "loss": 1.0329,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4107782975879458,
+      "grad_norm": 0.3661043345928192,
+      "learning_rate": 6.411209337949214e-05,
+      "loss": 1.1655,
+      "step": 6761
+    },
+    {
+      "epoch": 0.41083905462057235,
+      "grad_norm": 0.13706107437610626,
+      "learning_rate": 6.410290954543191e-05,
+      "loss": 1.0234,
+      "step": 6762
+    },
+    {
+      "epoch": 0.41089981165319883,
+      "grad_norm": 0.2353687286376953,
+      "learning_rate": 6.409372519442682e-05,
+      "loss": 1.2065,
+      "step": 6763
+    },
+    {
+      "epoch": 0.41096056868582537,
+      "grad_norm": 0.2996273338794708,
+      "learning_rate": 6.408454032681352e-05,
+      "loss": 1.0825,
+      "step": 6764
+    },
+    {
+      "epoch": 0.4110213257184519,
+      "grad_norm": 0.18953882157802582,
+      "learning_rate": 6.407535494292867e-05,
+      "loss": 1.1542,
+      "step": 6765
+    },
+    {
+      "epoch": 0.41108208275107844,
+      "grad_norm": 0.2067681849002838,
+      "learning_rate": 6.406616904310894e-05,
+      "loss": 1.1228,
+      "step": 6766
+    },
+    {
+      "epoch": 0.41114283978370497,
+      "grad_norm": 0.15057958662509918,
+      "learning_rate": 6.405698262769109e-05,
+      "loss": 1.0869,
+      "step": 6767
+    },
+    {
+      "epoch": 0.4112035968163315,
+      "grad_norm": 0.25824567675590515,
+      "learning_rate": 6.404779569701182e-05,
+      "loss": 1.0192,
+      "step": 6768
+    },
+    {
+      "epoch": 0.41126435384895804,
+      "grad_norm": 0.13443215191364288,
+      "learning_rate": 6.403860825140789e-05,
+      "loss": 1.0788,
+      "step": 6769
+    },
+    {
+      "epoch": 0.4113251108815845,
+      "grad_norm": 0.2732512950897217,
+      "learning_rate": 6.402942029121603e-05,
+      "loss": 1.2363,
+      "step": 6770
+    },
+    {
+      "epoch": 0.41138586791421106,
+      "grad_norm": 1.3949735164642334,
+      "learning_rate": 6.40202318167731e-05,
+      "loss": 1.1352,
+      "step": 6771
+    },
+    {
+      "epoch": 0.4114466249468376,
+      "grad_norm": 0.2742675840854645,
+      "learning_rate": 6.401104282841583e-05,
+      "loss": 1.1955,
+      "step": 6772
+    },
+    {
+      "epoch": 0.4115073819794641,
+      "grad_norm": 0.22874107956886292,
+      "learning_rate": 6.400185332648108e-05,
+      "loss": 1.15,
+      "step": 6773
+    },
+    {
+      "epoch": 0.41156813901209066,
+      "grad_norm": 0.2986707091331482,
+      "learning_rate": 6.399266331130571e-05,
+      "loss": 1.1729,
+      "step": 6774
+    },
+    {
+      "epoch": 0.4116288960447172,
+      "grad_norm": 0.21805629134178162,
+      "learning_rate": 6.398347278322654e-05,
+      "loss": 1.1117,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4116896530773437,
+      "grad_norm": 0.17125198245048523,
+      "learning_rate": 6.397428174258047e-05,
+      "loss": 1.134,
+      "step": 6776
+    },
+    {
+      "epoch": 0.4117504101099702,
+      "grad_norm": 0.13878971338272095,
+      "learning_rate": 6.396509018970442e-05,
+      "loss": 1.053,
+      "step": 6777
+    },
+    {
+      "epoch": 0.41181116714259675,
+      "grad_norm": 0.2812134325504303,
+      "learning_rate": 6.395589812493528e-05,
+      "loss": 1.1214,
+      "step": 6778
+    },
+    {
+      "epoch": 0.4118719241752233,
+      "grad_norm": 0.12149051576852798,
+      "learning_rate": 6.394670554861002e-05,
+      "loss": 1.0445,
+      "step": 6779
+    },
+    {
+      "epoch": 0.4119326812078498,
+      "grad_norm": 22.036266326904297,
+      "learning_rate": 6.393751246106553e-05,
+      "loss": 1.1176,
+      "step": 6780
+    },
+    {
+      "epoch": 0.41199343824047635,
+      "grad_norm": 0.6228098273277283,
+      "learning_rate": 6.392831886263886e-05,
+      "loss": 1.0863,
+      "step": 6781
+    },
+    {
+      "epoch": 0.4120541952731029,
+      "grad_norm": 0.7305749654769897,
+      "learning_rate": 6.391912475366695e-05,
+      "loss": 1.0611,
+      "step": 6782
+    },
+    {
+      "epoch": 0.41211495230572937,
+      "grad_norm": 0.41448134183883667,
+      "learning_rate": 6.390993013448683e-05,
+      "loss": 1.1178,
+      "step": 6783
+    },
+    {
+      "epoch": 0.4121757093383559,
+      "grad_norm": 0.16076204180717468,
+      "learning_rate": 6.390073500543556e-05,
+      "loss": 1.1146,
+      "step": 6784
+    },
+    {
+      "epoch": 0.41223646637098244,
+      "grad_norm": 0.1853996217250824,
+      "learning_rate": 6.389153936685012e-05,
+      "loss": 1.0825,
+      "step": 6785
+    },
+    {
+      "epoch": 0.41229722340360897,
+      "grad_norm": 0.17406095564365387,
+      "learning_rate": 6.388234321906765e-05,
+      "loss": 1.0645,
+      "step": 6786
+    },
+    {
+      "epoch": 0.4123579804362355,
+      "grad_norm": 0.28208327293395996,
+      "learning_rate": 6.38731465624252e-05,
+      "loss": 1.1606,
+      "step": 6787
+    },
+    {
+      "epoch": 0.41241873746886204,
+      "grad_norm": 0.13500328361988068,
+      "learning_rate": 6.386394939725989e-05,
+      "loss": 1.0769,
+      "step": 6788
+    },
+    {
+      "epoch": 0.4124794945014886,
+      "grad_norm": 0.1835724264383316,
+      "learning_rate": 6.385475172390882e-05,
+      "loss": 1.0882,
+      "step": 6789
+    },
+    {
+      "epoch": 0.41254025153411505,
+      "grad_norm": 0.16545453667640686,
+      "learning_rate": 6.384555354270913e-05,
+      "loss": 1.0934,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4126010085667416,
+      "grad_norm": 11.786956787109375,
+      "learning_rate": 6.383635485399804e-05,
+      "loss": 1.0899,
+      "step": 6791
+    },
+    {
+      "epoch": 0.4126617655993681,
+      "grad_norm": 0.17428408563137054,
+      "learning_rate": 6.382715565811267e-05,
+      "loss": 1.1009,
+      "step": 6792
+    },
+    {
+      "epoch": 0.41272252263199466,
+      "grad_norm": 0.2572706937789917,
+      "learning_rate": 6.381795595539023e-05,
+      "loss": 1.0628,
+      "step": 6793
+    },
+    {
+      "epoch": 0.4127832796646212,
+      "grad_norm": 0.1567520946264267,
+      "learning_rate": 6.380875574616795e-05,
+      "loss": 1.0409,
+      "step": 6794
+    },
+    {
+      "epoch": 0.41284403669724773,
+      "grad_norm": 0.21802347898483276,
+      "learning_rate": 6.379955503078306e-05,
+      "loss": 0.9951,
+      "step": 6795
+    },
+    {
+      "epoch": 0.4129047937298742,
+      "grad_norm": 0.16666068136692047,
+      "learning_rate": 6.379035380957281e-05,
+      "loss": 1.0834,
+      "step": 6796
+    },
+    {
+      "epoch": 0.41296555076250074,
+      "grad_norm": 0.15167683362960815,
+      "learning_rate": 6.378115208287447e-05,
+      "loss": 1.093,
+      "step": 6797
+    },
+    {
+      "epoch": 0.4130263077951273,
+      "grad_norm": 0.21332664787769318,
+      "learning_rate": 6.377194985102534e-05,
+      "loss": 1.0815,
+      "step": 6798
+    },
+    {
+      "epoch": 0.4130870648277538,
+      "grad_norm": 0.41679686307907104,
+      "learning_rate": 6.376274711436274e-05,
+      "loss": 1.3194,
+      "step": 6799
+    },
+    {
+      "epoch": 0.41314782186038035,
+      "grad_norm": 0.22373071312904358,
+      "learning_rate": 6.375354387322397e-05,
+      "loss": 1.2072,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4132085788930069,
+      "grad_norm": 0.14026744663715363,
+      "learning_rate": 6.37443401279464e-05,
+      "loss": 1.0749,
+      "step": 6801
+    },
+    {
+      "epoch": 0.4132693359256334,
+      "grad_norm": 0.14397601783275604,
+      "learning_rate": 6.373513587886737e-05,
+      "loss": 1.0595,
+      "step": 6802
+    },
+    {
+      "epoch": 0.4133300929582599,
+      "grad_norm": 0.419059157371521,
+      "learning_rate": 6.37259311263243e-05,
+      "loss": 1.3038,
+      "step": 6803
+    },
+    {
+      "epoch": 0.41339084999088643,
+      "grad_norm": 6.196994304656982,
+      "learning_rate": 6.371672587065456e-05,
+      "loss": 1.1054,
+      "step": 6804
+    },
+    {
+      "epoch": 0.41345160702351297,
+      "grad_norm": 0.2972855865955353,
+      "learning_rate": 6.370752011219559e-05,
+      "loss": 1.0455,
+      "step": 6805
+    },
+    {
+      "epoch": 0.4135123640561395,
+      "grad_norm": 0.1494392305612564,
+      "learning_rate": 6.369831385128479e-05,
+      "loss": 1.0755,
+      "step": 6806
+    },
+    {
+      "epoch": 0.41357312108876604,
+      "grad_norm": 0.31748634576797485,
+      "learning_rate": 6.368910708825968e-05,
+      "loss": 1.1824,
+      "step": 6807
+    },
+    {
+      "epoch": 0.4136338781213926,
+      "grad_norm": 0.21726271510124207,
+      "learning_rate": 6.36798998234577e-05,
+      "loss": 1.0903,
+      "step": 6808
+    },
+    {
+      "epoch": 0.41369463515401905,
+      "grad_norm": 0.16084003448486328,
+      "learning_rate": 6.367069205721634e-05,
+      "loss": 1.0449,
+      "step": 6809
+    },
+    {
+      "epoch": 0.4137553921866456,
+      "grad_norm": 0.13093261420726776,
+      "learning_rate": 6.366148378987312e-05,
+      "loss": 1.0608,
+      "step": 6810
+    },
+    {
+      "epoch": 0.4138161492192721,
+      "grad_norm": 0.2830486297607422,
+      "learning_rate": 6.365227502176557e-05,
+      "loss": 1.1464,
+      "step": 6811
+    },
+    {
+      "epoch": 0.41387690625189866,
+      "grad_norm": 0.12262752652168274,
+      "learning_rate": 6.364306575323124e-05,
+      "loss": 1.0376,
+      "step": 6812
+    },
+    {
+      "epoch": 0.4139376632845252,
+      "grad_norm": 0.3101465106010437,
+      "learning_rate": 6.363385598460772e-05,
+      "loss": 1.076,
+      "step": 6813
+    },
+    {
+      "epoch": 0.4139984203171517,
+      "grad_norm": 0.17994682490825653,
+      "learning_rate": 6.362464571623254e-05,
+      "loss": 1.1801,
+      "step": 6814
+    },
+    {
+      "epoch": 0.41405917734977826,
+      "grad_norm": 0.2440444678068161,
+      "learning_rate": 6.361543494844337e-05,
+      "loss": 1.1403,
+      "step": 6815
+    },
+    {
+      "epoch": 0.41411993438240474,
+      "grad_norm": 0.1671774983406067,
+      "learning_rate": 6.360622368157776e-05,
+      "loss": 1.1513,
+      "step": 6816
+    },
+    {
+      "epoch": 0.4141806914150313,
+      "grad_norm": 0.4949195981025696,
+      "learning_rate": 6.359701191597342e-05,
+      "loss": 1.1362,
+      "step": 6817
+    },
+    {
+      "epoch": 0.4142414484476578,
+      "grad_norm": 0.1731252670288086,
+      "learning_rate": 6.358779965196799e-05,
+      "loss": 1.0316,
+      "step": 6818
+    },
+    {
+      "epoch": 0.41430220548028435,
+      "grad_norm": 0.18434561789035797,
+      "learning_rate": 6.35785868898991e-05,
+      "loss": 1.1337,
+      "step": 6819
+    },
+    {
+      "epoch": 0.4143629625129109,
+      "grad_norm": 0.15021756291389465,
+      "learning_rate": 6.356937363010452e-05,
+      "loss": 1.0661,
+      "step": 6820
+    },
+    {
+      "epoch": 0.4144237195455374,
+      "grad_norm": 0.1818803995847702,
+      "learning_rate": 6.35601598729219e-05,
+      "loss": 1.029,
+      "step": 6821
+    },
+    {
+      "epoch": 0.4144844765781639,
+      "grad_norm": 0.22433963418006897,
+      "learning_rate": 6.355094561868901e-05,
+      "loss": 1.1532,
+      "step": 6822
+    },
+    {
+      "epoch": 0.41454523361079043,
+      "grad_norm": 0.18789389729499817,
+      "learning_rate": 6.35417308677436e-05,
+      "loss": 1.055,
+      "step": 6823
+    },
+    {
+      "epoch": 0.41460599064341697,
+      "grad_norm": 0.218994140625,
+      "learning_rate": 6.353251562042342e-05,
+      "loss": 1.0583,
+      "step": 6824
+    },
+    {
+      "epoch": 0.4146667476760435,
+      "grad_norm": 0.21378548443317413,
+      "learning_rate": 6.352329987706625e-05,
+      "loss": 1.2208,
+      "step": 6825
+    },
+    {
+      "epoch": 0.41472750470867004,
+      "grad_norm": 0.12409744411706924,
+      "learning_rate": 6.351408363800993e-05,
+      "loss": 1.0516,
+      "step": 6826
+    },
+    {
+      "epoch": 0.41478826174129657,
+      "grad_norm": 0.3805442452430725,
+      "learning_rate": 6.350486690359223e-05,
+      "loss": 1.0773,
+      "step": 6827
+    },
+    {
+      "epoch": 0.4148490187739231,
+      "grad_norm": 0.23587088286876678,
+      "learning_rate": 6.349564967415105e-05,
+      "loss": 1.1544,
+      "step": 6828
+    },
+    {
+      "epoch": 0.4149097758065496,
+      "grad_norm": 0.1841599941253662,
+      "learning_rate": 6.348643195002422e-05,
+      "loss": 1.2298,
+      "step": 6829
+    },
+    {
+      "epoch": 0.4149705328391761,
+      "grad_norm": 0.1608675718307495,
+      "learning_rate": 6.34772137315496e-05,
+      "loss": 1.0988,
+      "step": 6830
+    },
+    {
+      "epoch": 0.41503128987180266,
+      "grad_norm": 0.13292677700519562,
+      "learning_rate": 6.346799501906513e-05,
+      "loss": 1.0107,
+      "step": 6831
+    },
+    {
+      "epoch": 0.4150920469044292,
+      "grad_norm": 0.18974928557872772,
+      "learning_rate": 6.34587758129087e-05,
+      "loss": 1.2204,
+      "step": 6832
+    },
+    {
+      "epoch": 0.4151528039370557,
+      "grad_norm": 0.19476690888404846,
+      "learning_rate": 6.344955611341822e-05,
+      "loss": 1.1685,
+      "step": 6833
+    },
+    {
+      "epoch": 0.41521356096968226,
+      "grad_norm": 0.14657852053642273,
+      "learning_rate": 6.344033592093169e-05,
+      "loss": 1.0884,
+      "step": 6834
+    },
+    {
+      "epoch": 0.41527431800230874,
+      "grad_norm": 0.21968382596969604,
+      "learning_rate": 6.343111523578703e-05,
+      "loss": 1.0708,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4153350750349353,
+      "grad_norm": 0.14353160560131073,
+      "learning_rate": 6.342189405832225e-05,
+      "loss": 1.0546,
+      "step": 6836
+    },
+    {
+      "epoch": 0.4153958320675618,
+      "grad_norm": 0.15635627508163452,
+      "learning_rate": 6.341267238887534e-05,
+      "loss": 1.0967,
+      "step": 6837
+    },
+    {
+      "epoch": 0.41545658910018834,
+      "grad_norm": 0.1640235334634781,
+      "learning_rate": 6.340345022778435e-05,
+      "loss": 1.0507,
+      "step": 6838
+    },
+    {
+      "epoch": 0.4155173461328149,
+      "grad_norm": 0.20663829147815704,
+      "learning_rate": 6.339422757538727e-05,
+      "loss": 1.1422,
+      "step": 6839
+    },
+    {
+      "epoch": 0.4155781031654414,
+      "grad_norm": 0.15276584029197693,
+      "learning_rate": 6.33850044320222e-05,
+      "loss": 1.0889,
+      "step": 6840
+    },
+    {
+      "epoch": 0.41563886019806795,
+      "grad_norm": 0.2972036302089691,
+      "learning_rate": 6.337578079802724e-05,
+      "loss": 1.0401,
+      "step": 6841
+    },
+    {
+      "epoch": 0.41569961723069443,
+      "grad_norm": 0.2572774887084961,
+      "learning_rate": 6.33665566737404e-05,
+      "loss": 1.2376,
+      "step": 6842
+    },
+    {
+      "epoch": 0.41576037426332096,
+      "grad_norm": 0.18544059991836548,
+      "learning_rate": 6.335733205949988e-05,
+      "loss": 1.0772,
+      "step": 6843
+    },
+    {
+      "epoch": 0.4158211312959475,
+      "grad_norm": 0.14605402946472168,
+      "learning_rate": 6.334810695564375e-05,
+      "loss": 1.1364,
+      "step": 6844
+    },
+    {
+      "epoch": 0.41588188832857403,
+      "grad_norm": 0.12782181799411774,
+      "learning_rate": 6.333888136251018e-05,
+      "loss": 1.0563,
+      "step": 6845
+    },
+    {
+      "epoch": 0.41594264536120057,
+      "grad_norm": 0.22031638026237488,
+      "learning_rate": 6.332965528043735e-05,
+      "loss": 1.1313,
+      "step": 6846
+    },
+    {
+      "epoch": 0.4160034023938271,
+      "grad_norm": 0.11370088160037994,
+      "learning_rate": 6.33204287097634e-05,
+      "loss": 1.0812,
+      "step": 6847
+    },
+    {
+      "epoch": 0.41606415942645364,
+      "grad_norm": 0.14567701518535614,
+      "learning_rate": 6.331120165082659e-05,
+      "loss": 1.0732,
+      "step": 6848
+    },
+    {
+      "epoch": 0.4161249164590801,
+      "grad_norm": 0.26609399914741516,
+      "learning_rate": 6.33019741039651e-05,
+      "loss": 1.1244,
+      "step": 6849
+    },
+    {
+      "epoch": 0.41618567349170665,
+      "grad_norm": 0.2305385023355484,
+      "learning_rate": 6.329274606951718e-05,
+      "loss": 1.0898,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4162464305243332,
+      "grad_norm": 0.1748591810464859,
+      "learning_rate": 6.32835175478211e-05,
+      "loss": 1.0617,
+      "step": 6851
+    },
+    {
+      "epoch": 0.4163071875569597,
+      "grad_norm": 0.12653739750385284,
+      "learning_rate": 6.327428853921508e-05,
+      "loss": 1.0546,
+      "step": 6852
+    },
+    {
+      "epoch": 0.41636794458958626,
+      "grad_norm": 0.5172829031944275,
+      "learning_rate": 6.326505904403745e-05,
+      "loss": 1.2402,
+      "step": 6853
+    },
+    {
+      "epoch": 0.4164287016222128,
+      "grad_norm": 0.16379953920841217,
+      "learning_rate": 6.325582906262655e-05,
+      "loss": 1.0003,
+      "step": 6854
+    },
+    {
+      "epoch": 0.4164894586548393,
+      "grad_norm": 0.23378968238830566,
+      "learning_rate": 6.324659859532065e-05,
+      "loss": 1.1527,
+      "step": 6855
+    },
+    {
+      "epoch": 0.4165502156874658,
+      "grad_norm": 0.28856486082077026,
+      "learning_rate": 6.323736764245813e-05,
+      "loss": 1.077,
+      "step": 6856
+    },
+    {
+      "epoch": 0.41661097272009234,
+      "grad_norm": 1.1506011486053467,
+      "learning_rate": 6.322813620437732e-05,
+      "loss": 1.1055,
+      "step": 6857
+    },
+    {
+      "epoch": 0.4166717297527189,
+      "grad_norm": 0.5873069763183594,
+      "learning_rate": 6.321890428141665e-05,
+      "loss": 1.1425,
+      "step": 6858
+    },
+    {
+      "epoch": 0.4167324867853454,
+      "grad_norm": 0.23544001579284668,
+      "learning_rate": 6.320967187391447e-05,
+      "loss": 1.1597,
+      "step": 6859
+    },
+    {
+      "epoch": 0.41679324381797195,
+      "grad_norm": 0.3136921525001526,
+      "learning_rate": 6.32004389822092e-05,
+      "loss": 1.2012,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4168540008505985,
+      "grad_norm": 0.2939581871032715,
+      "learning_rate": 6.319120560663932e-05,
+      "loss": 1.1314,
+      "step": 6861
+    },
+    {
+      "epoch": 0.41691475788322496,
+      "grad_norm": 0.17378170788288116,
+      "learning_rate": 6.318197174754325e-05,
+      "loss": 1.105,
+      "step": 6862
+    },
+    {
+      "epoch": 0.4169755149158515,
+      "grad_norm": 0.25381678342819214,
+      "learning_rate": 6.317273740525942e-05,
+      "loss": 1.1491,
+      "step": 6863
+    },
+    {
+      "epoch": 0.41703627194847803,
+      "grad_norm": 0.15965205430984497,
+      "learning_rate": 6.316350258012638e-05,
+      "loss": 1.1048,
+      "step": 6864
+    },
+    {
+      "epoch": 0.41709702898110457,
+      "grad_norm": 0.1606505811214447,
+      "learning_rate": 6.31542672724826e-05,
+      "loss": 1.1133,
+      "step": 6865
+    },
+    {
+      "epoch": 0.4171577860137311,
+      "grad_norm": 0.208512544631958,
+      "learning_rate": 6.314503148266662e-05,
+      "loss": 1.1683,
+      "step": 6866
+    },
+    {
+      "epoch": 0.41721854304635764,
+      "grad_norm": 0.3423088788986206,
+      "learning_rate": 6.313579521101697e-05,
+      "loss": 1.1472,
+      "step": 6867
+    },
+    {
+      "epoch": 0.4172793000789841,
+      "grad_norm": 0.23983359336853027,
+      "learning_rate": 6.31265584578722e-05,
+      "loss": 1.0281,
+      "step": 6868
+    },
+    {
+      "epoch": 0.41734005711161065,
+      "grad_norm": 0.1681239902973175,
+      "learning_rate": 6.31173212235709e-05,
+      "loss": 1.0925,
+      "step": 6869
+    },
+    {
+      "epoch": 0.4174008141442372,
+      "grad_norm": 7.762028217315674,
+      "learning_rate": 6.310808350845164e-05,
+      "loss": 1.0839,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4174615711768637,
+      "grad_norm": 0.1919071525335312,
+      "learning_rate": 6.309884531285307e-05,
+      "loss": 1.2627,
+      "step": 6871
+    },
+    {
+      "epoch": 0.41752232820949026,
+      "grad_norm": 0.21320335566997528,
+      "learning_rate": 6.308960663711376e-05,
+      "loss": 1.0737,
+      "step": 6872
+    },
+    {
+      "epoch": 0.4175830852421168,
+      "grad_norm": 0.19712430238723755,
+      "learning_rate": 6.30803674815724e-05,
+      "loss": 1.0324,
+      "step": 6873
+    },
+    {
+      "epoch": 0.4176438422747433,
+      "grad_norm": 0.21667392551898956,
+      "learning_rate": 6.307112784656766e-05,
+      "loss": 1.0627,
+      "step": 6874
+    },
+    {
+      "epoch": 0.4177045993073698,
+      "grad_norm": 0.17969444394111633,
+      "learning_rate": 6.306188773243819e-05,
+      "loss": 0.9994,
+      "step": 6875
+    },
+    {
+      "epoch": 0.41776535633999634,
+      "grad_norm": 0.23410826921463013,
+      "learning_rate": 6.305264713952269e-05,
+      "loss": 1.1836,
+      "step": 6876
+    },
+    {
+      "epoch": 0.4178261133726229,
+      "grad_norm": 0.20407861471176147,
+      "learning_rate": 6.30434060681599e-05,
+      "loss": 1.2992,
+      "step": 6877
+    },
+    {
+      "epoch": 0.4178868704052494,
+      "grad_norm": 0.5336781144142151,
+      "learning_rate": 6.303416451868853e-05,
+      "loss": 1.1372,
+      "step": 6878
+    },
+    {
+      "epoch": 0.41794762743787595,
+      "grad_norm": 0.14235033094882965,
+      "learning_rate": 6.302492249144736e-05,
+      "loss": 1.0444,
+      "step": 6879
+    },
+    {
+      "epoch": 0.4180083844705025,
+      "grad_norm": 27.98836326599121,
+      "learning_rate": 6.301567998677512e-05,
+      "loss": 1.1922,
+      "step": 6880
+    },
+    {
+      "epoch": 0.41806914150312896,
+      "grad_norm": 0.31793883442878723,
+      "learning_rate": 6.300643700501061e-05,
+      "loss": 1.2796,
+      "step": 6881
+    },
+    {
+      "epoch": 0.4181298985357555,
+      "grad_norm": 0.16187778115272522,
+      "learning_rate": 6.299719354649266e-05,
+      "loss": 1.0702,
+      "step": 6882
+    },
+    {
+      "epoch": 0.41819065556838203,
+      "grad_norm": 0.5476409196853638,
+      "learning_rate": 6.298794961156004e-05,
+      "loss": 1.319,
+      "step": 6883
+    },
+    {
+      "epoch": 0.41825141260100857,
+      "grad_norm": 0.2328275740146637,
+      "learning_rate": 6.297870520055163e-05,
+      "loss": 1.1273,
+      "step": 6884
+    },
+    {
+      "epoch": 0.4183121696336351,
+      "grad_norm": 0.12698839604854584,
+      "learning_rate": 6.296946031380628e-05,
+      "loss": 1.0523,
+      "step": 6885
+    },
+    {
+      "epoch": 0.41837292666626164,
+      "grad_norm": 0.14996226131916046,
+      "learning_rate": 6.296021495166284e-05,
+      "loss": 1.0498,
+      "step": 6886
+    },
+    {
+      "epoch": 0.41843368369888817,
+      "grad_norm": 0.20301690697669983,
+      "learning_rate": 6.295096911446023e-05,
+      "loss": 1.1121,
+      "step": 6887
+    },
+    {
+      "epoch": 0.41849444073151465,
+      "grad_norm": 0.19202876091003418,
+      "learning_rate": 6.294172280253734e-05,
+      "loss": 1.091,
+      "step": 6888
+    },
+    {
+      "epoch": 0.4185551977641412,
+      "grad_norm": 19.74171257019043,
+      "learning_rate": 6.29324760162331e-05,
+      "loss": 1.0966,
+      "step": 6889
+    },
+    {
+      "epoch": 0.4186159547967677,
+      "grad_norm": 0.2416202574968338,
+      "learning_rate": 6.292322875588645e-05,
+      "loss": 1.1811,
+      "step": 6890
+    },
+    {
+      "epoch": 0.41867671182939425,
+      "grad_norm": 0.18140235543251038,
+      "learning_rate": 6.291398102183634e-05,
+      "loss": 1.0655,
+      "step": 6891
+    },
+    {
+      "epoch": 0.4187374688620208,
+      "grad_norm": 0.18717539310455322,
+      "learning_rate": 6.290473281442177e-05,
+      "loss": 1.1467,
+      "step": 6892
+    },
+    {
+      "epoch": 0.4187982258946473,
+      "grad_norm": 0.1472216099500656,
+      "learning_rate": 6.289548413398172e-05,
+      "loss": 1.0387,
+      "step": 6893
+    },
+    {
+      "epoch": 0.41885898292727386,
+      "grad_norm": 0.18043498694896698,
+      "learning_rate": 6.288623498085522e-05,
+      "loss": 1.0747,
+      "step": 6894
+    },
+    {
+      "epoch": 0.41891973995990034,
+      "grad_norm": 0.15664604306221008,
+      "learning_rate": 6.287698535538129e-05,
+      "loss": 1.112,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4189804969925269,
+      "grad_norm": 0.2661375403404236,
+      "learning_rate": 6.286773525789894e-05,
+      "loss": 1.0764,
+      "step": 6896
+    },
+    {
+      "epoch": 0.4190412540251534,
+      "grad_norm": 0.14460201561450958,
+      "learning_rate": 6.285848468874728e-05,
+      "loss": 1.0316,
+      "step": 6897
+    },
+    {
+      "epoch": 0.41910201105777994,
+      "grad_norm": 0.22918212413787842,
+      "learning_rate": 6.28492336482654e-05,
+      "loss": 1.1167,
+      "step": 6898
+    },
+    {
+      "epoch": 0.4191627680904065,
+      "grad_norm": 0.1595105230808258,
+      "learning_rate": 6.283998213679236e-05,
+      "loss": 1.065,
+      "step": 6899
+    },
+    {
+      "epoch": 0.419223525123033,
+      "grad_norm": 1.2316800355911255,
+      "learning_rate": 6.28307301546673e-05,
+      "loss": 1.0789,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4192842821556595,
+      "grad_norm": 0.287421315908432,
+      "learning_rate": 6.282147770222932e-05,
+      "loss": 1.0824,
+      "step": 6901
+    },
+    {
+      "epoch": 0.41934503918828603,
+      "grad_norm": 0.1340104341506958,
+      "learning_rate": 6.281222477981762e-05,
+      "loss": 1.0551,
+      "step": 6902
+    },
+    {
+      "epoch": 0.41940579622091256,
+      "grad_norm": 0.35405653715133667,
+      "learning_rate": 6.280297138777135e-05,
+      "loss": 1.0831,
+      "step": 6903
+    },
+    {
+      "epoch": 0.4194665532535391,
+      "grad_norm": 0.13373813033103943,
+      "learning_rate": 6.279371752642968e-05,
+      "loss": 1.0476,
+      "step": 6904
+    },
+    {
+      "epoch": 0.41952731028616563,
+      "grad_norm": 0.14785555005073547,
+      "learning_rate": 6.278446319613183e-05,
+      "loss": 1.0999,
+      "step": 6905
+    },
+    {
+      "epoch": 0.41958806731879217,
+      "grad_norm": 0.11077269911766052,
+      "learning_rate": 6.2775208397217e-05,
+      "loss": 1.0273,
+      "step": 6906
+    },
+    {
+      "epoch": 0.4196488243514187,
+      "grad_norm": 0.11789675801992416,
+      "learning_rate": 6.276595313002442e-05,
+      "loss": 1.059,
+      "step": 6907
+    },
+    {
+      "epoch": 0.4197095813840452,
+      "grad_norm": 0.16572049260139465,
+      "learning_rate": 6.275669739489339e-05,
+      "loss": 1.1243,
+      "step": 6908
+    },
+    {
+      "epoch": 0.4197703384166717,
+      "grad_norm": 2.3128082752227783,
+      "learning_rate": 6.274744119216314e-05,
+      "loss": 1.0493,
+      "step": 6909
+    },
+    {
+      "epoch": 0.41983109544929825,
+      "grad_norm": 0.27944597601890564,
+      "learning_rate": 6.273818452217295e-05,
+      "loss": 1.0743,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4198918524819248,
+      "grad_norm": 0.3292635977268219,
+      "learning_rate": 6.27289273852622e-05,
+      "loss": 1.1443,
+      "step": 6911
+    },
+    {
+      "epoch": 0.4199526095145513,
+      "grad_norm": 0.1531735509634018,
+      "learning_rate": 6.271966978177009e-05,
+      "loss": 1.061,
+      "step": 6912
+    },
+    {
+      "epoch": 0.42001336654717786,
+      "grad_norm": 0.1380385309457779,
+      "learning_rate": 6.271041171203606e-05,
+      "loss": 1.0429,
+      "step": 6913
+    },
+    {
+      "epoch": 0.42007412357980434,
+      "grad_norm": 0.18009182810783386,
+      "learning_rate": 6.270115317639943e-05,
+      "loss": 1.1294,
+      "step": 6914
+    },
+    {
+      "epoch": 0.42013488061243087,
+      "grad_norm": 5.402449607849121,
+      "learning_rate": 6.269189417519957e-05,
+      "loss": 1.1245,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4201956376450574,
+      "grad_norm": 0.1931043416261673,
+      "learning_rate": 6.268263470877587e-05,
+      "loss": 1.2073,
+      "step": 6916
+    },
+    {
+      "epoch": 0.42025639467768394,
+      "grad_norm": 6.12284517288208,
+      "learning_rate": 6.267337477746776e-05,
+      "loss": 1.121,
+      "step": 6917
+    },
+    {
+      "epoch": 0.4203171517103105,
+      "grad_norm": 0.1379459798336029,
+      "learning_rate": 6.266411438161465e-05,
+      "loss": 1.0877,
+      "step": 6918
+    },
+    {
+      "epoch": 0.420377908742937,
+      "grad_norm": 0.18237358331680298,
+      "learning_rate": 6.265485352155595e-05,
+      "loss": 1.1493,
+      "step": 6919
+    },
+    {
+      "epoch": 0.42043866577556355,
+      "grad_norm": 0.1345060020685196,
+      "learning_rate": 6.264559219763116e-05,
+      "loss": 1.0716,
+      "step": 6920
+    },
+    {
+      "epoch": 0.42049942280819,
+      "grad_norm": 0.14291448891162872,
+      "learning_rate": 6.263633041017975e-05,
+      "loss": 1.1124,
+      "step": 6921
+    },
+    {
+      "epoch": 0.42056017984081656,
+      "grad_norm": 0.1362290382385254,
+      "learning_rate": 6.262706815954121e-05,
+      "loss": 1.0575,
+      "step": 6922
+    },
+    {
+      "epoch": 0.4206209368734431,
+      "grad_norm": 0.12679123878479004,
+      "learning_rate": 6.261780544605504e-05,
+      "loss": 1.0844,
+      "step": 6923
+    },
+    {
+      "epoch": 0.42068169390606963,
+      "grad_norm": 0.20512786507606506,
+      "learning_rate": 6.260854227006079e-05,
+      "loss": 1.123,
+      "step": 6924
+    },
+    {
+      "epoch": 0.42074245093869617,
+      "grad_norm": 0.21317705512046814,
+      "learning_rate": 6.259927863189796e-05,
+      "loss": 1.1117,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4208032079713227,
+      "grad_norm": 0.15350694954395294,
+      "learning_rate": 6.259001453190615e-05,
+      "loss": 1.0852,
+      "step": 6926
+    },
+    {
+      "epoch": 0.4208639650039492,
+      "grad_norm": 0.18139959871768951,
+      "learning_rate": 6.258074997042493e-05,
+      "loss": 1.1731,
+      "step": 6927
+    },
+    {
+      "epoch": 0.4209247220365757,
+      "grad_norm": 0.19312353432178497,
+      "learning_rate": 6.257148494779387e-05,
+      "loss": 1.2615,
+      "step": 6928
+    },
+    {
+      "epoch": 0.42098547906920225,
+      "grad_norm": 0.2791098654270172,
+      "learning_rate": 6.256221946435262e-05,
+      "loss": 1.0902,
+      "step": 6929
+    },
+    {
+      "epoch": 0.4210462361018288,
+      "grad_norm": 0.1469917893409729,
+      "learning_rate": 6.255295352044075e-05,
+      "loss": 1.0227,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4211069931344553,
+      "grad_norm": 0.1947174370288849,
+      "learning_rate": 6.254368711639797e-05,
+      "loss": 1.1611,
+      "step": 6931
+    },
+    {
+      "epoch": 0.42116775016708186,
+      "grad_norm": 0.19812403619289398,
+      "learning_rate": 6.253442025256392e-05,
+      "loss": 1.0874,
+      "step": 6932
+    },
+    {
+      "epoch": 0.4212285071997084,
+      "grad_norm": 0.29190686345100403,
+      "learning_rate": 6.252515292927827e-05,
+      "loss": 1.1326,
+      "step": 6933
+    },
+    {
+      "epoch": 0.42128926423233487,
+      "grad_norm": 0.40126460790634155,
+      "learning_rate": 6.25158851468807e-05,
+      "loss": 1.079,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4213500212649614,
+      "grad_norm": 0.23086629807949066,
+      "learning_rate": 6.250661690571095e-05,
+      "loss": 1.0901,
+      "step": 6935
+    },
+    {
+      "epoch": 0.42141077829758794,
+      "grad_norm": 0.4902494251728058,
+      "learning_rate": 6.249734820610876e-05,
+      "loss": 1.0133,
+      "step": 6936
+    },
+    {
+      "epoch": 0.4214715353302145,
+      "grad_norm": 0.1629909873008728,
+      "learning_rate": 6.248807904841383e-05,
+      "loss": 1.0541,
+      "step": 6937
+    },
+    {
+      "epoch": 0.421532292362841,
+      "grad_norm": 0.43007004261016846,
+      "learning_rate": 6.247880943296596e-05,
+      "loss": 1.1121,
+      "step": 6938
+    },
+    {
+      "epoch": 0.42159304939546755,
+      "grad_norm": 0.1909950226545334,
+      "learning_rate": 6.24695393601049e-05,
+      "loss": 1.0356,
+      "step": 6939
+    },
+    {
+      "epoch": 0.421653806428094,
+      "grad_norm": 0.3682610094547272,
+      "learning_rate": 6.246026883017048e-05,
+      "loss": 1.1733,
+      "step": 6940
+    },
+    {
+      "epoch": 0.42171456346072056,
+      "grad_norm": 0.2998441457748413,
+      "learning_rate": 6.24509978435025e-05,
+      "loss": 1.0487,
+      "step": 6941
+    },
+    {
+      "epoch": 0.4217753204933471,
+      "grad_norm": 0.16706110537052155,
+      "learning_rate": 6.244172640044081e-05,
+      "loss": 1.1036,
+      "step": 6942
+    },
+    {
+      "epoch": 0.42183607752597363,
+      "grad_norm": 0.18088893592357635,
+      "learning_rate": 6.24324545013252e-05,
+      "loss": 1.0899,
+      "step": 6943
+    },
+    {
+      "epoch": 0.42189683455860016,
+      "grad_norm": 0.21891342103481293,
+      "learning_rate": 6.242318214649557e-05,
+      "loss": 1.0751,
+      "step": 6944
+    },
+    {
+      "epoch": 0.4219575915912267,
+      "grad_norm": 0.1899438351392746,
+      "learning_rate": 6.24139093362918e-05,
+      "loss": 1.0458,
+      "step": 6945
+    },
+    {
+      "epoch": 0.42201834862385323,
+      "grad_norm": 0.14089667797088623,
+      "learning_rate": 6.240463607105377e-05,
+      "loss": 1.0152,
+      "step": 6946
+    },
+    {
+      "epoch": 0.4220791056564797,
+      "grad_norm": 0.14117169380187988,
+      "learning_rate": 6.239536235112144e-05,
+      "loss": 1.0353,
+      "step": 6947
+    },
+    {
+      "epoch": 0.42213986268910625,
+      "grad_norm": 0.1743745058774948,
+      "learning_rate": 6.238608817683468e-05,
+      "loss": 1.0739,
+      "step": 6948
+    },
+    {
+      "epoch": 0.4222006197217328,
+      "grad_norm": 0.21024319529533386,
+      "learning_rate": 6.237681354853349e-05,
+      "loss": 1.1628,
+      "step": 6949
+    },
+    {
+      "epoch": 0.4222613767543593,
+      "grad_norm": 0.20903562009334564,
+      "learning_rate": 6.236753846655778e-05,
+      "loss": 1.1894,
+      "step": 6950
+    },
+    {
+      "epoch": 0.42232213378698585,
+      "grad_norm": 0.1602024883031845,
+      "learning_rate": 6.235826293124756e-05,
+      "loss": 1.034,
+      "step": 6951
+    },
+    {
+      "epoch": 0.4223828908196124,
+      "grad_norm": 0.19580984115600586,
+      "learning_rate": 6.234898694294284e-05,
+      "loss": 1.1636,
+      "step": 6952
+    },
+    {
+      "epoch": 0.4224436478522389,
+      "grad_norm": 0.13619299232959747,
+      "learning_rate": 6.233971050198361e-05,
+      "loss": 1.0798,
+      "step": 6953
+    },
+    {
+      "epoch": 0.4225044048848654,
+      "grad_norm": 0.14674031734466553,
+      "learning_rate": 6.23304336087099e-05,
+      "loss": 1.1592,
+      "step": 6954
+    },
+    {
+      "epoch": 0.42256516191749194,
+      "grad_norm": 0.169845312833786,
+      "learning_rate": 6.232115626346178e-05,
+      "loss": 1.1067,
+      "step": 6955
+    },
+    {
+      "epoch": 0.4226259189501185,
+      "grad_norm": 0.17236468195915222,
+      "learning_rate": 6.231187846657927e-05,
+      "loss": 1.0417,
+      "step": 6956
+    },
+    {
+      "epoch": 0.422686675982745,
+      "grad_norm": 0.16715261340141296,
+      "learning_rate": 6.23026002184025e-05,
+      "loss": 1.1337,
+      "step": 6957
+    },
+    {
+      "epoch": 0.42274743301537154,
+      "grad_norm": 0.42817223072052,
+      "learning_rate": 6.229332151927151e-05,
+      "loss": 1.373,
+      "step": 6958
+    },
+    {
+      "epoch": 0.4228081900479981,
+      "grad_norm": 0.1281808465719223,
+      "learning_rate": 6.228404236952647e-05,
+      "loss": 1.0387,
+      "step": 6959
+    },
+    {
+      "epoch": 0.42286894708062456,
+      "grad_norm": 0.13899502158164978,
+      "learning_rate": 6.227476276950748e-05,
+      "loss": 1.0434,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4229297041132511,
+      "grad_norm": 0.18265396356582642,
+      "learning_rate": 6.226548271955466e-05,
+      "loss": 1.0292,
+      "step": 6961
+    },
+    {
+      "epoch": 0.4229904611458776,
+      "grad_norm": 0.161569744348526,
+      "learning_rate": 6.225620222000826e-05,
+      "loss": 1.1229,
+      "step": 6962
+    },
+    {
+      "epoch": 0.42305121817850416,
+      "grad_norm": 0.1807115226984024,
+      "learning_rate": 6.224692127120834e-05,
+      "loss": 1.0406,
+      "step": 6963
+    },
+    {
+      "epoch": 0.4231119752111307,
+      "grad_norm": 0.17286376655101776,
+      "learning_rate": 6.223763987349518e-05,
+      "loss": 1.1447,
+      "step": 6964
+    },
+    {
+      "epoch": 0.42317273224375723,
+      "grad_norm": 0.24327607452869415,
+      "learning_rate": 6.222835802720897e-05,
+      "loss": 1.048,
+      "step": 6965
+    },
+    {
+      "epoch": 0.42323348927638377,
+      "grad_norm": 0.20642633736133575,
+      "learning_rate": 6.22190757326899e-05,
+      "loss": 1.1765,
+      "step": 6966
+    },
+    {
+      "epoch": 0.42329424630901025,
+      "grad_norm": 2.7735562324523926,
+      "learning_rate": 6.220979299027829e-05,
+      "loss": 1.1822,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4233550033416368,
+      "grad_norm": 0.152414932847023,
+      "learning_rate": 6.220050980031433e-05,
+      "loss": 1.141,
+      "step": 6968
+    },
+    {
+      "epoch": 0.4234157603742633,
+      "grad_norm": 0.2437283843755722,
+      "learning_rate": 6.219122616313832e-05,
+      "loss": 1.1929,
+      "step": 6969
+    },
+    {
+      "epoch": 0.42347651740688985,
+      "grad_norm": 0.21336692571640015,
+      "learning_rate": 6.218194207909058e-05,
+      "loss": 1.1662,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4235372744395164,
+      "grad_norm": 0.20766429603099823,
+      "learning_rate": 6.21726575485114e-05,
+      "loss": 1.0903,
+      "step": 6971
+    },
+    {
+      "epoch": 0.4235980314721429,
+      "grad_norm": 0.1277417540550232,
+      "learning_rate": 6.216337257174109e-05,
+      "loss": 1.1028,
+      "step": 6972
+    },
+    {
+      "epoch": 0.4236587885047694,
+      "grad_norm": 0.18504175543785095,
+      "learning_rate": 6.215408714912001e-05,
+      "loss": 1.1471,
+      "step": 6973
+    },
+    {
+      "epoch": 0.42371954553739594,
+      "grad_norm": 0.1630711406469345,
+      "learning_rate": 6.214480128098851e-05,
+      "loss": 1.0445,
+      "step": 6974
+    },
+    {
+      "epoch": 0.42378030257002247,
+      "grad_norm": 0.1733415424823761,
+      "learning_rate": 6.2135514967687e-05,
+      "loss": 1.0522,
+      "step": 6975
+    },
+    {
+      "epoch": 0.423841059602649,
+      "grad_norm": 0.19768063724040985,
+      "learning_rate": 6.212622820955582e-05,
+      "loss": 1.1364,
+      "step": 6976
+    },
+    {
+      "epoch": 0.42390181663527554,
+      "grad_norm": 0.15273819863796234,
+      "learning_rate": 6.211694100693538e-05,
+      "loss": 1.1118,
+      "step": 6977
+    },
+    {
+      "epoch": 0.4239625736679021,
+      "grad_norm": 0.17157956957817078,
+      "learning_rate": 6.210765336016616e-05,
+      "loss": 1.0561,
+      "step": 6978
+    },
+    {
+      "epoch": 0.4240233307005286,
+      "grad_norm": 0.36483943462371826,
+      "learning_rate": 6.209836526958854e-05,
+      "loss": 1.1137,
+      "step": 6979
+    },
+    {
+      "epoch": 0.4240840877331551,
+      "grad_norm": 0.1682959347963333,
+      "learning_rate": 6.208907673554303e-05,
+      "loss": 1.0244,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4241448447657816,
+      "grad_norm": 0.3107626438140869,
+      "learning_rate": 6.207978775837005e-05,
+      "loss": 1.1828,
+      "step": 6981
+    },
+    {
+      "epoch": 0.42420560179840816,
+      "grad_norm": 0.1407303810119629,
+      "learning_rate": 6.207049833841014e-05,
+      "loss": 1.1248,
+      "step": 6982
+    },
+    {
+      "epoch": 0.4242663588310347,
+      "grad_norm": 0.16782920062541962,
+      "learning_rate": 6.206120847600377e-05,
+      "loss": 1.0773,
+      "step": 6983
+    },
+    {
+      "epoch": 0.42432711586366123,
+      "grad_norm": 0.19861558079719543,
+      "learning_rate": 6.205191817149145e-05,
+      "loss": 1.1012,
+      "step": 6984
+    },
+    {
+      "epoch": 0.42438787289628777,
+      "grad_norm": 0.15361498296260834,
+      "learning_rate": 6.204262742521377e-05,
+      "loss": 1.0886,
+      "step": 6985
+    },
+    {
+      "epoch": 0.42444862992891425,
+      "grad_norm": 0.14355316758155823,
+      "learning_rate": 6.203333623751124e-05,
+      "loss": 1.0522,
+      "step": 6986
+    },
+    {
+      "epoch": 0.4245093869615408,
+      "grad_norm": 0.21123307943344116,
+      "learning_rate": 6.202404460872444e-05,
+      "loss": 1.1052,
+      "step": 6987
+    },
+    {
+      "epoch": 0.4245701439941673,
+      "grad_norm": 0.20525924861431122,
+      "learning_rate": 6.201475253919397e-05,
+      "loss": 1.0438,
+      "step": 6988
+    },
+    {
+      "epoch": 0.42463090102679385,
+      "grad_norm": 46.03693389892578,
+      "learning_rate": 6.200546002926043e-05,
+      "loss": 1.0986,
+      "step": 6989
+    },
+    {
+      "epoch": 0.4246916580594204,
+      "grad_norm": 0.19850599765777588,
+      "learning_rate": 6.199616707926441e-05,
+      "loss": 1.1396,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4247524150920469,
+      "grad_norm": 0.1901119500398636,
+      "learning_rate": 6.198687368954659e-05,
+      "loss": 1.1318,
+      "step": 6991
+    },
+    {
+      "epoch": 0.42481317212467345,
+      "grad_norm": 0.1583910584449768,
+      "learning_rate": 6.197757986044759e-05,
+      "loss": 1.1032,
+      "step": 6992
+    },
+    {
+      "epoch": 0.42487392915729993,
+      "grad_norm": 0.15106023848056793,
+      "learning_rate": 6.196828559230809e-05,
+      "loss": 1.0996,
+      "step": 6993
+    },
+    {
+      "epoch": 0.42493468618992647,
+      "grad_norm": 0.2375694066286087,
+      "learning_rate": 6.195899088546875e-05,
+      "loss": 1.1119,
+      "step": 6994
+    },
+    {
+      "epoch": 0.424995443222553,
+      "grad_norm": 0.2967996895313263,
+      "learning_rate": 6.19496957402703e-05,
+      "loss": 1.0543,
+      "step": 6995
+    },
+    {
+      "epoch": 0.42505620025517954,
+      "grad_norm": 0.4244559705257416,
+      "learning_rate": 6.194040015705344e-05,
+      "loss": 1.2376,
+      "step": 6996
+    },
+    {
+      "epoch": 0.4251169572878061,
+      "grad_norm": 0.3686823844909668,
+      "learning_rate": 6.19311041361589e-05,
+      "loss": 1.0578,
+      "step": 6997
+    },
+    {
+      "epoch": 0.4251777143204326,
+      "grad_norm": 0.1373891681432724,
+      "learning_rate": 6.192180767792743e-05,
+      "loss": 1.0618,
+      "step": 6998
+    },
+    {
+      "epoch": 0.42523847135305914,
+      "grad_norm": 0.16797654330730438,
+      "learning_rate": 6.191251078269981e-05,
+      "loss": 1.0495,
+      "step": 6999
+    },
+    {
+      "epoch": 0.4252992283856856,
+      "grad_norm": 0.1643468737602234,
+      "learning_rate": 6.19032134508168e-05,
+      "loss": 1.0902,
+      "step": 7000
+    },
+    {
+      "epoch": 0.42535998541831216,
+      "grad_norm": 0.3246098458766937,
+      "learning_rate": 6.189391568261919e-05,
+      "loss": 1.1858,
+      "step": 7001
+    },
+    {
+      "epoch": 0.4254207424509387,
+      "grad_norm": 0.12454608082771301,
+      "learning_rate": 6.188461747844781e-05,
+      "loss": 1.0969,
+      "step": 7002
+    },
+    {
+      "epoch": 0.42548149948356523,
+      "grad_norm": 0.16250506043434143,
+      "learning_rate": 6.187531883864348e-05,
+      "loss": 1.1429,
+      "step": 7003
+    },
+    {
+      "epoch": 0.42554225651619176,
+      "grad_norm": 0.17062848806381226,
+      "learning_rate": 6.186601976354705e-05,
+      "loss": 1.1037,
+      "step": 7004
+    },
+    {
+      "epoch": 0.4256030135488183,
+      "grad_norm": 2.5766096115112305,
+      "learning_rate": 6.185672025349935e-05,
+      "loss": 1.1161,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4256637705814448,
+      "grad_norm": 0.19814927875995636,
+      "learning_rate": 6.18474203088413e-05,
+      "loss": 1.2225,
+      "step": 7006
+    },
+    {
+      "epoch": 0.4257245276140713,
+      "grad_norm": 0.2760142385959625,
+      "learning_rate": 6.183811992991375e-05,
+      "loss": 1.0919,
+      "step": 7007
+    },
+    {
+      "epoch": 0.42578528464669785,
+      "grad_norm": 0.6109820008277893,
+      "learning_rate": 6.182881911705762e-05,
+      "loss": 1.2268,
+      "step": 7008
+    },
+    {
+      "epoch": 0.4258460416793244,
+      "grad_norm": 0.20365901291370392,
+      "learning_rate": 6.181951787061388e-05,
+      "loss": 1.1179,
+      "step": 7009
+    },
+    {
+      "epoch": 0.4259067987119509,
+      "grad_norm": 0.15513215959072113,
+      "learning_rate": 6.181021619092339e-05,
+      "loss": 1.1462,
+      "step": 7010
+    },
+    {
+      "epoch": 0.42596755574457745,
+      "grad_norm": 0.1652374416589737,
+      "learning_rate": 6.180091407832715e-05,
+      "loss": 1.0216,
+      "step": 7011
+    },
+    {
+      "epoch": 0.426028312777204,
+      "grad_norm": 0.13578563928604126,
+      "learning_rate": 6.179161153316614e-05,
+      "loss": 1.0311,
+      "step": 7012
+    },
+    {
+      "epoch": 0.42608906980983047,
+      "grad_norm": 0.20253820717334747,
+      "learning_rate": 6.17823085557813e-05,
+      "loss": 1.1829,
+      "step": 7013
+    },
+    {
+      "epoch": 0.426149826842457,
+      "grad_norm": 0.42810219526290894,
+      "learning_rate": 6.177300514651367e-05,
+      "loss": 1.262,
+      "step": 7014
+    },
+    {
+      "epoch": 0.42621058387508354,
+      "grad_norm": 0.18049682676792145,
+      "learning_rate": 6.176370130570428e-05,
+      "loss": 1.1112,
+      "step": 7015
+    },
+    {
+      "epoch": 0.4262713409077101,
+      "grad_norm": 0.26479053497314453,
+      "learning_rate": 6.175439703369414e-05,
+      "loss": 1.1471,
+      "step": 7016
+    },
+    {
+      "epoch": 0.4263320979403366,
+      "grad_norm": 0.13780274987220764,
+      "learning_rate": 6.17450923308243e-05,
+      "loss": 1.0598,
+      "step": 7017
+    },
+    {
+      "epoch": 0.42639285497296314,
+      "grad_norm": 0.2020138055086136,
+      "learning_rate": 6.173578719743581e-05,
+      "loss": 1.0978,
+      "step": 7018
+    },
+    {
+      "epoch": 0.4264536120055896,
+      "grad_norm": 0.17220433056354523,
+      "learning_rate": 6.17264816338698e-05,
+      "loss": 1.0873,
+      "step": 7019
+    },
+    {
+      "epoch": 0.42651436903821616,
+      "grad_norm": 0.26138007640838623,
+      "learning_rate": 6.171717564046735e-05,
+      "loss": 1.0651,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4265751260708427,
+      "grad_norm": 0.13816316425800323,
+      "learning_rate": 6.170786921756951e-05,
+      "loss": 1.0139,
+      "step": 7021
+    },
+    {
+      "epoch": 0.4266358831034692,
+      "grad_norm": 0.32481956481933594,
+      "learning_rate": 6.169856236551751e-05,
+      "loss": 1.1706,
+      "step": 7022
+    },
+    {
+      "epoch": 0.42669664013609576,
+      "grad_norm": 0.14297835528850555,
+      "learning_rate": 6.168925508465241e-05,
+      "loss": 1.0174,
+      "step": 7023
+    },
+    {
+      "epoch": 0.4267573971687223,
+      "grad_norm": 0.15055494010448456,
+      "learning_rate": 6.167994737531543e-05,
+      "loss": 1.0698,
+      "step": 7024
+    },
+    {
+      "epoch": 0.42681815420134883,
+      "grad_norm": 0.24918074905872345,
+      "learning_rate": 6.167063923784771e-05,
+      "loss": 1.0262,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4268789112339753,
+      "grad_norm": 0.19334658980369568,
+      "learning_rate": 6.166133067259047e-05,
+      "loss": 1.0675,
+      "step": 7026
+    },
+    {
+      "epoch": 0.42693966826660185,
+      "grad_norm": 0.22851526737213135,
+      "learning_rate": 6.165202167988488e-05,
+      "loss": 1.1112,
+      "step": 7027
+    },
+    {
+      "epoch": 0.4270004252992284,
+      "grad_norm": 13.368069648742676,
+      "learning_rate": 6.16427122600722e-05,
+      "loss": 1.1536,
+      "step": 7028
+    },
+    {
+      "epoch": 0.4270611823318549,
+      "grad_norm": 0.5032018423080444,
+      "learning_rate": 6.163340241349365e-05,
+      "loss": 1.2312,
+      "step": 7029
+    },
+    {
+      "epoch": 0.42712193936448145,
+      "grad_norm": 0.33650246262550354,
+      "learning_rate": 6.162409214049047e-05,
+      "loss": 1.067,
+      "step": 7030
+    },
+    {
+      "epoch": 0.427182696397108,
+      "grad_norm": 0.19822436571121216,
+      "learning_rate": 6.161478144140395e-05,
+      "loss": 1.1227,
+      "step": 7031
+    },
+    {
+      "epoch": 0.42724345342973447,
+      "grad_norm": 0.24417605996131897,
+      "learning_rate": 6.160547031657538e-05,
+      "loss": 1.1816,
+      "step": 7032
+    },
+    {
+      "epoch": 0.427304210462361,
+      "grad_norm": 0.20915569365024567,
+      "learning_rate": 6.159615876634603e-05,
+      "loss": 1.1855,
+      "step": 7033
+    },
+    {
+      "epoch": 0.42736496749498754,
+      "grad_norm": 0.16673614084720612,
+      "learning_rate": 6.158684679105727e-05,
+      "loss": 1.041,
+      "step": 7034
+    },
+    {
+      "epoch": 0.42742572452761407,
+      "grad_norm": 0.13468198478221893,
+      "learning_rate": 6.15775343910504e-05,
+      "loss": 1.0355,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4274864815602406,
+      "grad_norm": 0.3226775527000427,
+      "learning_rate": 6.156822156666674e-05,
+      "loss": 1.1822,
+      "step": 7036
+    },
+    {
+      "epoch": 0.42754723859286714,
+      "grad_norm": 0.575821578502655,
+      "learning_rate": 6.155890831824769e-05,
+      "loss": 1.0471,
+      "step": 7037
+    },
+    {
+      "epoch": 0.4276079956254937,
+      "grad_norm": 0.2560516595840454,
+      "learning_rate": 6.154959464613464e-05,
+      "loss": 0.9903,
+      "step": 7038
+    },
+    {
+      "epoch": 0.42766875265812015,
+      "grad_norm": 0.13638198375701904,
+      "learning_rate": 6.154028055066896e-05,
+      "loss": 1.0836,
+      "step": 7039
+    },
+    {
+      "epoch": 0.4277295096907467,
+      "grad_norm": 0.19872820377349854,
+      "learning_rate": 6.153096603219207e-05,
+      "loss": 1.1425,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4277902667233732,
+      "grad_norm": 0.18162383139133453,
+      "learning_rate": 6.152165109104536e-05,
+      "loss": 1.1514,
+      "step": 7041
+    },
+    {
+      "epoch": 0.42785102375599976,
+      "grad_norm": 0.1593349277973175,
+      "learning_rate": 6.151233572757034e-05,
+      "loss": 1.1199,
+      "step": 7042
+    },
+    {
+      "epoch": 0.4279117807886263,
+      "grad_norm": 0.23361849784851074,
+      "learning_rate": 6.150301994210841e-05,
+      "loss": 1.2252,
+      "step": 7043
+    },
+    {
+      "epoch": 0.42797253782125283,
+      "grad_norm": 0.13178184628486633,
+      "learning_rate": 6.149370373500105e-05,
+      "loss": 1.0471,
+      "step": 7044
+    },
+    {
+      "epoch": 0.4280332948538793,
+      "grad_norm": 0.2979399859905243,
+      "learning_rate": 6.14843871065898e-05,
+      "loss": 1.1882,
+      "step": 7045
+    },
+    {
+      "epoch": 0.42809405188650584,
+      "grad_norm": 0.20199747383594513,
+      "learning_rate": 6.147507005721607e-05,
+      "loss": 1.1594,
+      "step": 7046
+    },
+    {
+      "epoch": 0.4281548089191324,
+      "grad_norm": 0.21474188566207886,
+      "learning_rate": 6.146575258722146e-05,
+      "loss": 1.1894,
+      "step": 7047
+    },
+    {
+      "epoch": 0.4282155659517589,
+      "grad_norm": 0.17794427275657654,
+      "learning_rate": 6.145643469694747e-05,
+      "loss": 1.105,
+      "step": 7048
+    },
+    {
+      "epoch": 0.42827632298438545,
+      "grad_norm": 0.5070486068725586,
+      "learning_rate": 6.144711638673564e-05,
+      "loss": 1.1564,
+      "step": 7049
+    },
+    {
+      "epoch": 0.428337080017012,
+      "grad_norm": 0.17088961601257324,
+      "learning_rate": 6.143779765692756e-05,
+      "loss": 1.0613,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4283978370496385,
+      "grad_norm": 0.1595560908317566,
+      "learning_rate": 6.142847850786477e-05,
+      "loss": 1.2335,
+      "step": 7051
+    },
+    {
+      "epoch": 0.428458594082265,
+      "grad_norm": 0.2286539524793625,
+      "learning_rate": 6.141915893988892e-05,
+      "loss": 1.1241,
+      "step": 7052
+    },
+    {
+      "epoch": 0.42851935111489153,
+      "grad_norm": 0.23533058166503906,
+      "learning_rate": 6.140983895334158e-05,
+      "loss": 1.1091,
+      "step": 7053
+    },
+    {
+      "epoch": 0.42858010814751807,
+      "grad_norm": 0.19321134686470032,
+      "learning_rate": 6.140051854856437e-05,
+      "loss": 1.2191,
+      "step": 7054
+    },
+    {
+      "epoch": 0.4286408651801446,
+      "grad_norm": 0.21221891045570374,
+      "learning_rate": 6.139119772589899e-05,
+      "loss": 1.1174,
+      "step": 7055
+    },
+    {
+      "epoch": 0.42870162221277114,
+      "grad_norm": 0.3473241329193115,
+      "learning_rate": 6.138187648568704e-05,
+      "loss": 1.2597,
+      "step": 7056
+    },
+    {
+      "epoch": 0.4287623792453977,
+      "grad_norm": 0.21167616546154022,
+      "learning_rate": 6.137255482827018e-05,
+      "loss": 1.1508,
+      "step": 7057
+    },
+    {
+      "epoch": 0.4288231362780242,
+      "grad_norm": 0.1301553100347519,
+      "learning_rate": 6.136323275399016e-05,
+      "loss": 1.0411,
+      "step": 7058
+    },
+    {
+      "epoch": 0.4288838933106507,
+      "grad_norm": 0.4767073690891266,
+      "learning_rate": 6.135391026318862e-05,
+      "loss": 1.3891,
+      "step": 7059
+    },
+    {
+      "epoch": 0.4289446503432772,
+      "grad_norm": 0.14260655641555786,
+      "learning_rate": 6.134458735620732e-05,
+      "loss": 1.0142,
+      "step": 7060
+    },
+    {
+      "epoch": 0.42900540737590376,
+      "grad_norm": 0.2777833044528961,
+      "learning_rate": 6.133526403338798e-05,
+      "loss": 1.0565,
+      "step": 7061
+    },
+    {
+      "epoch": 0.4290661644085303,
+      "grad_norm": 0.15962089598178864,
+      "learning_rate": 6.132594029507234e-05,
+      "loss": 1.0886,
+      "step": 7062
+    },
+    {
+      "epoch": 0.42912692144115683,
+      "grad_norm": 0.20868034660816193,
+      "learning_rate": 6.131661614160217e-05,
+      "loss": 1.1047,
+      "step": 7063
+    },
+    {
+      "epoch": 0.42918767847378336,
+      "grad_norm": 0.16495519876480103,
+      "learning_rate": 6.130729157331924e-05,
+      "loss": 1.151,
+      "step": 7064
+    },
+    {
+      "epoch": 0.42924843550640984,
+      "grad_norm": 0.17820709943771362,
+      "learning_rate": 6.129796659056535e-05,
+      "loss": 1.1386,
+      "step": 7065
+    },
+    {
+      "epoch": 0.4293091925390364,
+      "grad_norm": 0.1497243046760559,
+      "learning_rate": 6.128864119368234e-05,
+      "loss": 1.0805,
+      "step": 7066
+    },
+    {
+      "epoch": 0.4293699495716629,
+      "grad_norm": 0.3838287889957428,
+      "learning_rate": 6.127931538301198e-05,
+      "loss": 1.0966,
+      "step": 7067
+    },
+    {
+      "epoch": 0.42943070660428945,
+      "grad_norm": 0.16190454363822937,
+      "learning_rate": 6.126998915889612e-05,
+      "loss": 1.1048,
+      "step": 7068
+    },
+    {
+      "epoch": 0.429491463636916,
+      "grad_norm": 0.17607471346855164,
+      "learning_rate": 6.126066252167667e-05,
+      "loss": 1.0616,
+      "step": 7069
+    },
+    {
+      "epoch": 0.4295522206695425,
+      "grad_norm": 0.2324267476797104,
+      "learning_rate": 6.125133547169543e-05,
+      "loss": 1.2412,
+      "step": 7070
+    },
+    {
+      "epoch": 0.42961297770216905,
+      "grad_norm": 0.13477569818496704,
+      "learning_rate": 6.124200800929431e-05,
+      "loss": 1.0669,
+      "step": 7071
+    },
+    {
+      "epoch": 0.42967373473479553,
+      "grad_norm": 0.13471132516860962,
+      "learning_rate": 6.123268013481523e-05,
+      "loss": 1.0692,
+      "step": 7072
+    },
+    {
+      "epoch": 0.42973449176742207,
+      "grad_norm": 0.15257640182971954,
+      "learning_rate": 6.122335184860008e-05,
+      "loss": 1.091,
+      "step": 7073
+    },
+    {
+      "epoch": 0.4297952488000486,
+      "grad_norm": 0.15090148150920868,
+      "learning_rate": 6.121402315099081e-05,
+      "loss": 1.1209,
+      "step": 7074
+    },
+    {
+      "epoch": 0.42985600583267514,
+      "grad_norm": 0.11312448978424072,
+      "learning_rate": 6.120469404232932e-05,
+      "loss": 1.0342,
+      "step": 7075
+    },
+    {
+      "epoch": 0.42991676286530167,
+      "grad_norm": 0.17266380786895752,
+      "learning_rate": 6.119536452295765e-05,
+      "loss": 1.1304,
+      "step": 7076
+    },
+    {
+      "epoch": 0.4299775198979282,
+      "grad_norm": 0.2930562198162079,
+      "learning_rate": 6.118603459321772e-05,
+      "loss": 1.1513,
+      "step": 7077
+    },
+    {
+      "epoch": 0.4300382769305547,
+      "grad_norm": 0.10981795191764832,
+      "learning_rate": 6.11767042534515e-05,
+      "loss": 1.0226,
+      "step": 7078
+    },
+    {
+      "epoch": 0.4300990339631812,
+      "grad_norm": 0.13515910506248474,
+      "learning_rate": 6.116737350400105e-05,
+      "loss": 1.0393,
+      "step": 7079
+    },
+    {
+      "epoch": 0.43015979099580776,
+      "grad_norm": 0.1950734704732895,
+      "learning_rate": 6.115804234520834e-05,
+      "loss": 1.0281,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4302205480284343,
+      "grad_norm": 0.1276920735836029,
+      "learning_rate": 6.114871077741546e-05,
+      "loss": 1.0492,
+      "step": 7081
+    },
+    {
+      "epoch": 0.4302813050610608,
+      "grad_norm": 0.17948336899280548,
+      "learning_rate": 6.113937880096442e-05,
+      "loss": 1.1086,
+      "step": 7082
+    },
+    {
+      "epoch": 0.43034206209368736,
+      "grad_norm": 0.11978979408740997,
+      "learning_rate": 6.11300464161973e-05,
+      "loss": 1.0521,
+      "step": 7083
+    },
+    {
+      "epoch": 0.4304028191263139,
+      "grad_norm": 0.20154602825641632,
+      "learning_rate": 6.112071362345617e-05,
+      "loss": 1.1547,
+      "step": 7084
+    },
+    {
+      "epoch": 0.4304635761589404,
+      "grad_norm": 0.11545953154563904,
+      "learning_rate": 6.111138042308314e-05,
+      "loss": 1.03,
+      "step": 7085
+    },
+    {
+      "epoch": 0.4305243331915669,
+      "grad_norm": 0.14451618492603302,
+      "learning_rate": 6.110204681542031e-05,
+      "loss": 1.0605,
+      "step": 7086
+    },
+    {
+      "epoch": 0.43058509022419345,
+      "grad_norm": 0.2848051190376282,
+      "learning_rate": 6.109271280080981e-05,
+      "loss": 1.2883,
+      "step": 7087
+    },
+    {
+      "epoch": 0.43064584725682,
+      "grad_norm": 0.17974348366260529,
+      "learning_rate": 6.108337837959377e-05,
+      "loss": 1.115,
+      "step": 7088
+    },
+    {
+      "epoch": 0.4307066042894465,
+      "grad_norm": 0.16068078577518463,
+      "learning_rate": 6.107404355211437e-05,
+      "loss": 1.0828,
+      "step": 7089
+    },
+    {
+      "epoch": 0.43076736132207305,
+      "grad_norm": 0.11544747650623322,
+      "learning_rate": 6.106470831871374e-05,
+      "loss": 1.0812,
+      "step": 7090
+    },
+    {
+      "epoch": 0.43082811835469953,
+      "grad_norm": 0.11844844371080399,
+      "learning_rate": 6.10553726797341e-05,
+      "loss": 1.0738,
+      "step": 7091
+    },
+    {
+      "epoch": 0.43088887538732606,
+      "grad_norm": 0.3195345103740692,
+      "learning_rate": 6.104603663551764e-05,
+      "loss": 1.1275,
+      "step": 7092
+    },
+    {
+      "epoch": 0.4309496324199526,
+      "grad_norm": 0.12494722753763199,
+      "learning_rate": 6.103670018640657e-05,
+      "loss": 1.0839,
+      "step": 7093
+    },
+    {
+      "epoch": 0.43101038945257913,
+      "grad_norm": 0.16662168502807617,
+      "learning_rate": 6.102736333274313e-05,
+      "loss": 1.1831,
+      "step": 7094
+    },
+    {
+      "epoch": 0.43107114648520567,
+      "grad_norm": 0.21858850121498108,
+      "learning_rate": 6.1018026074869536e-05,
+      "loss": 1.1509,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4311319035178322,
+      "grad_norm": 0.17780764400959015,
+      "learning_rate": 6.100868841312808e-05,
+      "loss": 1.0821,
+      "step": 7096
+    },
+    {
+      "epoch": 0.43119266055045874,
+      "grad_norm": 0.231523796916008,
+      "learning_rate": 6.0999350347861016e-05,
+      "loss": 1.0561,
+      "step": 7097
+    },
+    {
+      "epoch": 0.4312534175830852,
+      "grad_norm": 0.1594240963459015,
+      "learning_rate": 6.099001187941063e-05,
+      "loss": 1.1148,
+      "step": 7098
+    },
+    {
+      "epoch": 0.43131417461571175,
+      "grad_norm": 0.2271859347820282,
+      "learning_rate": 6.098067300811927e-05,
+      "loss": 1.1758,
+      "step": 7099
+    },
+    {
+      "epoch": 0.4313749316483383,
+      "grad_norm": 0.1946006417274475,
+      "learning_rate": 6.09713337343292e-05,
+      "loss": 1.1624,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4314356886809648,
+      "grad_norm": 0.12236620485782623,
+      "learning_rate": 6.096199405838274e-05,
+      "loss": 1.0467,
+      "step": 7101
+    },
+    {
+      "epoch": 0.43149644571359136,
+      "grad_norm": 0.18780270218849182,
+      "learning_rate": 6.095265398062231e-05,
+      "loss": 1.0469,
+      "step": 7102
+    },
+    {
+      "epoch": 0.4315572027462179,
+      "grad_norm": 0.2614232301712036,
+      "learning_rate": 6.094331350139021e-05,
+      "loss": 1.0238,
+      "step": 7103
+    },
+    {
+      "epoch": 0.43161795977884443,
+      "grad_norm": 0.14401251077651978,
+      "learning_rate": 6.093397262102884e-05,
+      "loss": 1.0763,
+      "step": 7104
+    },
+    {
+      "epoch": 0.4316787168114709,
+      "grad_norm": 0.33712613582611084,
+      "learning_rate": 6.09246313398806e-05,
+      "loss": 1.1354,
+      "step": 7105
+    },
+    {
+      "epoch": 0.43173947384409744,
+      "grad_norm": 0.40387997031211853,
+      "learning_rate": 6.0915289658287877e-05,
+      "loss": 1.0825,
+      "step": 7106
+    },
+    {
+      "epoch": 0.431800230876724,
+      "grad_norm": 0.13862301409244537,
+      "learning_rate": 6.090594757659311e-05,
+      "loss": 1.043,
+      "step": 7107
+    },
+    {
+      "epoch": 0.4318609879093505,
+      "grad_norm": 0.263348251581192,
+      "learning_rate": 6.089660509513872e-05,
+      "loss": 1.0176,
+      "step": 7108
+    },
+    {
+      "epoch": 0.43192174494197705,
+      "grad_norm": 0.3066016435623169,
+      "learning_rate": 6.0887262214267174e-05,
+      "loss": 1.0681,
+      "step": 7109
+    },
+    {
+      "epoch": 0.4319825019746036,
+      "grad_norm": 0.13966190814971924,
+      "learning_rate": 6.0877918934320924e-05,
+      "loss": 1.0875,
+      "step": 7110
+    },
+    {
+      "epoch": 0.43204325900723006,
+      "grad_norm": 0.27148106694221497,
+      "learning_rate": 6.086857525564244e-05,
+      "loss": 1.002,
+      "step": 7111
+    },
+    {
+      "epoch": 0.4321040160398566,
+      "grad_norm": 0.18690867722034454,
+      "learning_rate": 6.085923117857424e-05,
+      "loss": 1.0406,
+      "step": 7112
+    },
+    {
+      "epoch": 0.43216477307248313,
+      "grad_norm": 0.21091775596141815,
+      "learning_rate": 6.0849886703458826e-05,
+      "loss": 1.2286,
+      "step": 7113
+    },
+    {
+      "epoch": 0.43222553010510967,
+      "grad_norm": 0.2024928480386734,
+      "learning_rate": 6.08405418306387e-05,
+      "loss": 1.0651,
+      "step": 7114
+    },
+    {
+      "epoch": 0.4322862871377362,
+      "grad_norm": 0.20826666057109833,
+      "learning_rate": 6.083119656045643e-05,
+      "loss": 1.1842,
+      "step": 7115
+    },
+    {
+      "epoch": 0.43234704417036274,
+      "grad_norm": 0.22445645928382874,
+      "learning_rate": 6.082185089325455e-05,
+      "loss": 1.1479,
+      "step": 7116
+    },
+    {
+      "epoch": 0.4324078012029893,
+      "grad_norm": 0.20395037531852722,
+      "learning_rate": 6.0812504829375636e-05,
+      "loss": 1.103,
+      "step": 7117
+    },
+    {
+      "epoch": 0.43246855823561575,
+      "grad_norm": 0.2830505669116974,
+      "learning_rate": 6.080315836916228e-05,
+      "loss": 1.0743,
+      "step": 7118
+    },
+    {
+      "epoch": 0.4325293152682423,
+      "grad_norm": 1.192375659942627,
+      "learning_rate": 6.079381151295706e-05,
+      "loss": 1.0642,
+      "step": 7119
+    },
+    {
+      "epoch": 0.4325900723008688,
+      "grad_norm": 0.1309729963541031,
+      "learning_rate": 6.0784464261102584e-05,
+      "loss": 1.0456,
+      "step": 7120
+    },
+    {
+      "epoch": 0.43265082933349536,
+      "grad_norm": 0.24944883584976196,
+      "learning_rate": 6.07751166139415e-05,
+      "loss": 1.1335,
+      "step": 7121
+    },
+    {
+      "epoch": 0.4327115863661219,
+      "grad_norm": 0.13679952919483185,
+      "learning_rate": 6.076576857181642e-05,
+      "loss": 1.0541,
+      "step": 7122
+    },
+    {
+      "epoch": 0.4327723433987484,
+      "grad_norm": 0.17177702486515045,
+      "learning_rate": 6.0756420135070016e-05,
+      "loss": 1.0998,
+      "step": 7123
+    },
+    {
+      "epoch": 0.4328331004313749,
+      "grad_norm": 0.11775720864534378,
+      "learning_rate": 6.0747071304044943e-05,
+      "loss": 1.0794,
+      "step": 7124
+    },
+    {
+      "epoch": 0.43289385746400144,
+      "grad_norm": 0.16918975114822388,
+      "learning_rate": 6.07377220790839e-05,
+      "loss": 1.1037,
+      "step": 7125
+    },
+    {
+      "epoch": 0.432954614496628,
+      "grad_norm": 0.16255351901054382,
+      "learning_rate": 6.0728372460529594e-05,
+      "loss": 1.1624,
+      "step": 7126
+    },
+    {
+      "epoch": 0.4330153715292545,
+      "grad_norm": 0.1941896378993988,
+      "learning_rate": 6.07190224487247e-05,
+      "loss": 1.1485,
+      "step": 7127
+    },
+    {
+      "epoch": 0.43307612856188105,
+      "grad_norm": 0.22777590155601501,
+      "learning_rate": 6.070967204401199e-05,
+      "loss": 1.1124,
+      "step": 7128
+    },
+    {
+      "epoch": 0.4331368855945076,
+      "grad_norm": 0.4661976397037506,
+      "learning_rate": 6.070032124673416e-05,
+      "loss": 1.1488,
+      "step": 7129
+    },
+    {
+      "epoch": 0.4331976426271341,
+      "grad_norm": 0.12373591214418411,
+      "learning_rate": 6.0690970057234e-05,
+      "loss": 1.0287,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4332583996597606,
+      "grad_norm": 0.15416377782821655,
+      "learning_rate": 6.068161847585426e-05,
+      "loss": 1.0552,
+      "step": 7131
+    },
+    {
+      "epoch": 0.43331915669238713,
+      "grad_norm": 0.13549821078777313,
+      "learning_rate": 6.067226650293774e-05,
+      "loss": 1.1038,
+      "step": 7132
+    },
+    {
+      "epoch": 0.43337991372501367,
+      "grad_norm": 0.921561062335968,
+      "learning_rate": 6.066291413882721e-05,
+      "loss": 1.0443,
+      "step": 7133
+    },
+    {
+      "epoch": 0.4334406707576402,
+      "grad_norm": 3.5042123794555664,
+      "learning_rate": 6.06535613838655e-05,
+      "loss": 1.1176,
+      "step": 7134
+    },
+    {
+      "epoch": 0.43350142779026674,
+      "grad_norm": 0.15883924067020416,
+      "learning_rate": 6.064420823839545e-05,
+      "loss": 1.0985,
+      "step": 7135
+    },
+    {
+      "epoch": 0.43356218482289327,
+      "grad_norm": 0.19465826451778412,
+      "learning_rate": 6.0634854702759904e-05,
+      "loss": 1.1411,
+      "step": 7136
+    },
+    {
+      "epoch": 0.43362294185551975,
+      "grad_norm": 0.1418951004743576,
+      "learning_rate": 6.062550077730167e-05,
+      "loss": 1.0655,
+      "step": 7137
+    },
+    {
+      "epoch": 0.4336836988881463,
+      "grad_norm": 0.15734487771987915,
+      "learning_rate": 6.0616146462363664e-05,
+      "loss": 1.046,
+      "step": 7138
+    },
+    {
+      "epoch": 0.4337444559207728,
+      "grad_norm": 0.2207980901002884,
+      "learning_rate": 6.060679175828877e-05,
+      "loss": 1.0857,
+      "step": 7139
+    },
+    {
+      "epoch": 0.43380521295339936,
+      "grad_norm": 0.16430488228797913,
+      "learning_rate": 6.059743666541986e-05,
+      "loss": 1.1034,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4338659699860259,
+      "grad_norm": 0.19490373134613037,
+      "learning_rate": 6.0588081184099865e-05,
+      "loss": 1.1176,
+      "step": 7141
+    },
+    {
+      "epoch": 0.4339267270186524,
+      "grad_norm": 0.15159621834754944,
+      "learning_rate": 6.05787253146717e-05,
+      "loss": 1.147,
+      "step": 7142
+    },
+    {
+      "epoch": 0.43398748405127896,
+      "grad_norm": 0.14486467838287354,
+      "learning_rate": 6.056936905747834e-05,
+      "loss": 1.0935,
+      "step": 7143
+    },
+    {
+      "epoch": 0.43404824108390544,
+      "grad_norm": 0.15860125422477722,
+      "learning_rate": 6.05600124128627e-05,
+      "loss": 1.0545,
+      "step": 7144
+    },
+    {
+      "epoch": 0.434108998116532,
+      "grad_norm": 0.1415642946958542,
+      "learning_rate": 6.055065538116774e-05,
+      "loss": 1.1636,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4341697551491585,
+      "grad_norm": 0.13268671929836273,
+      "learning_rate": 6.054129796273651e-05,
+      "loss": 1.0779,
+      "step": 7146
+    },
+    {
+      "epoch": 0.43423051218178504,
+      "grad_norm": 0.21452026069164276,
+      "learning_rate": 6.0531940157911924e-05,
+      "loss": 1.1252,
+      "step": 7147
+    },
+    {
+      "epoch": 0.4342912692144116,
+      "grad_norm": 68.19952392578125,
+      "learning_rate": 6.052258196703706e-05,
+      "loss": 1.0644,
+      "step": 7148
+    },
+    {
+      "epoch": 0.4343520262470381,
+      "grad_norm": 0.4241727888584137,
+      "learning_rate": 6.051322339045493e-05,
+      "loss": 1.2822,
+      "step": 7149
+    },
+    {
+      "epoch": 0.4344127832796646,
+      "grad_norm": 0.15386973321437836,
+      "learning_rate": 6.050386442850854e-05,
+      "loss": 1.0769,
+      "step": 7150
+    },
+    {
+      "epoch": 0.43447354031229113,
+      "grad_norm": 0.14582303166389465,
+      "learning_rate": 6.049450508154098e-05,
+      "loss": 1.0454,
+      "step": 7151
+    },
+    {
+      "epoch": 0.43453429734491766,
+      "grad_norm": 0.17524971067905426,
+      "learning_rate": 6.04851453498953e-05,
+      "loss": 1.1636,
+      "step": 7152
+    },
+    {
+      "epoch": 0.4345950543775442,
+      "grad_norm": 0.22474323213100433,
+      "learning_rate": 6.047578523391459e-05,
+      "loss": 1.1451,
+      "step": 7153
+    },
+    {
+      "epoch": 0.43465581141017073,
+      "grad_norm": 0.15816301107406616,
+      "learning_rate": 6.046642473394196e-05,
+      "loss": 1.1828,
+      "step": 7154
+    },
+    {
+      "epoch": 0.43471656844279727,
+      "grad_norm": 0.3579353094100952,
+      "learning_rate": 6.045706385032049e-05,
+      "loss": 1.2668,
+      "step": 7155
+    },
+    {
+      "epoch": 0.4347773254754238,
+      "grad_norm": 0.1534872055053711,
+      "learning_rate": 6.044770258339335e-05,
+      "loss": 1.0896,
+      "step": 7156
+    },
+    {
+      "epoch": 0.4348380825080503,
+      "grad_norm": 0.15883764624595642,
+      "learning_rate": 6.0438340933503644e-05,
+      "loss": 1.0624,
+      "step": 7157
+    },
+    {
+      "epoch": 0.4348988395406768,
+      "grad_norm": 0.11349524557590485,
+      "learning_rate": 6.0428978900994516e-05,
+      "loss": 1.0198,
+      "step": 7158
+    },
+    {
+      "epoch": 0.43495959657330335,
+      "grad_norm": 0.1767425686120987,
+      "learning_rate": 6.0419616486209174e-05,
+      "loss": 1.0026,
+      "step": 7159
+    },
+    {
+      "epoch": 0.4350203536059299,
+      "grad_norm": 0.13428759574890137,
+      "learning_rate": 6.041025368949076e-05,
+      "loss": 1.0105,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4350811106385564,
+      "grad_norm": 0.15892168879508972,
+      "learning_rate": 6.0400890511182485e-05,
+      "loss": 1.1038,
+      "step": 7161
+    },
+    {
+      "epoch": 0.43514186767118296,
+      "grad_norm": 0.14943519234657288,
+      "learning_rate": 6.039152695162756e-05,
+      "loss": 1.0742,
+      "step": 7162
+    },
+    {
+      "epoch": 0.4352026247038095,
+      "grad_norm": 0.602253258228302,
+      "learning_rate": 6.038216301116921e-05,
+      "loss": 1.2142,
+      "step": 7163
+    },
+    {
+      "epoch": 0.435263381736436,
+      "grad_norm": 0.14757990837097168,
+      "learning_rate": 6.037279869015067e-05,
+      "loss": 1.0739,
+      "step": 7164
+    },
+    {
+      "epoch": 0.4353241387690625,
+      "grad_norm": 0.14695501327514648,
+      "learning_rate": 6.0363433988915176e-05,
+      "loss": 1.0311,
+      "step": 7165
+    },
+    {
+      "epoch": 0.43538489580168904,
+      "grad_norm": 1.7526659965515137,
+      "learning_rate": 6.0354068907806014e-05,
+      "loss": 1.0223,
+      "step": 7166
+    },
+    {
+      "epoch": 0.4354456528343156,
+      "grad_norm": 0.1997620165348053,
+      "learning_rate": 6.034470344716646e-05,
+      "loss": 1.0764,
+      "step": 7167
+    },
+    {
+      "epoch": 0.4355064098669421,
+      "grad_norm": 0.17104572057724,
+      "learning_rate": 6.033533760733978e-05,
+      "loss": 1.0475,
+      "step": 7168
+    },
+    {
+      "epoch": 0.43556716689956865,
+      "grad_norm": 0.19382449984550476,
+      "learning_rate": 6.0325971388669334e-05,
+      "loss": 1.1337,
+      "step": 7169
+    },
+    {
+      "epoch": 0.4356279239321951,
+      "grad_norm": 0.18365032970905304,
+      "learning_rate": 6.03166047914984e-05,
+      "loss": 1.0475,
+      "step": 7170
+    },
+    {
+      "epoch": 0.43568868096482166,
+      "grad_norm": 0.16334949433803558,
+      "learning_rate": 6.0307237816170304e-05,
+      "loss": 1.1123,
+      "step": 7171
+    },
+    {
+      "epoch": 0.4357494379974482,
+      "grad_norm": 0.5135619640350342,
+      "learning_rate": 6.029787046302843e-05,
+      "loss": 1.2474,
+      "step": 7172
+    },
+    {
+      "epoch": 0.43581019503007473,
+      "grad_norm": 0.13248133659362793,
+      "learning_rate": 6.028850273241612e-05,
+      "loss": 1.0637,
+      "step": 7173
+    },
+    {
+      "epoch": 0.43587095206270127,
+      "grad_norm": 2.4597551822662354,
+      "learning_rate": 6.027913462467676e-05,
+      "loss": 1.0462,
+      "step": 7174
+    },
+    {
+      "epoch": 0.4359317090953278,
+      "grad_norm": 0.2982158362865448,
+      "learning_rate": 6.026976614015373e-05,
+      "loss": 1.0594,
+      "step": 7175
+    },
+    {
+      "epoch": 0.43599246612795434,
+      "grad_norm": 0.29003480076789856,
+      "learning_rate": 6.0260397279190437e-05,
+      "loss": 1.1222,
+      "step": 7176
+    },
+    {
+      "epoch": 0.4360532231605808,
+      "grad_norm": 0.33089935779571533,
+      "learning_rate": 6.02510280421303e-05,
+      "loss": 1.2227,
+      "step": 7177
+    },
+    {
+      "epoch": 0.43611398019320735,
+      "grad_norm": 0.5047960877418518,
+      "learning_rate": 6.0241658429316747e-05,
+      "loss": 1.1323,
+      "step": 7178
+    },
+    {
+      "epoch": 0.4361747372258339,
+      "grad_norm": 0.1691085547208786,
+      "learning_rate": 6.0232288441093224e-05,
+      "loss": 1.0994,
+      "step": 7179
+    },
+    {
+      "epoch": 0.4362354942584604,
+      "grad_norm": 0.1709735244512558,
+      "learning_rate": 6.022291807780319e-05,
+      "loss": 1.0786,
+      "step": 7180
+    },
+    {
+      "epoch": 0.43629625129108696,
+      "grad_norm": 0.25987717509269714,
+      "learning_rate": 6.0213547339790097e-05,
+      "loss": 1.0558,
+      "step": 7181
+    },
+    {
+      "epoch": 0.4363570083237135,
+      "grad_norm": 0.14235125482082367,
+      "learning_rate": 6.020417622739748e-05,
+      "loss": 1.0835,
+      "step": 7182
+    },
+    {
+      "epoch": 0.43641776535633997,
+      "grad_norm": 0.9190377593040466,
+      "learning_rate": 6.01948047409688e-05,
+      "loss": 1.1797,
+      "step": 7183
+    },
+    {
+      "epoch": 0.4364785223889665,
+      "grad_norm": 0.33326396346092224,
+      "learning_rate": 6.018543288084759e-05,
+      "loss": 1.2175,
+      "step": 7184
+    },
+    {
+      "epoch": 0.43653927942159304,
+      "grad_norm": 0.20607006549835205,
+      "learning_rate": 6.0176060647377367e-05,
+      "loss": 1.0854,
+      "step": 7185
+    },
+    {
+      "epoch": 0.4366000364542196,
+      "grad_norm": 0.15370653569698334,
+      "learning_rate": 6.016668804090168e-05,
+      "loss": 1.08,
+      "step": 7186
+    },
+    {
+      "epoch": 0.4366607934868461,
+      "grad_norm": 0.3283064663410187,
+      "learning_rate": 6.0157315061764066e-05,
+      "loss": 1.2804,
+      "step": 7187
+    },
+    {
+      "epoch": 0.43672155051947265,
+      "grad_norm": 0.14727094769477844,
+      "learning_rate": 6.014794171030811e-05,
+      "loss": 1.0919,
+      "step": 7188
+    },
+    {
+      "epoch": 0.4367823075520992,
+      "grad_norm": 0.12322274595499039,
+      "learning_rate": 6.013856798687738e-05,
+      "loss": 1.0545,
+      "step": 7189
+    },
+    {
+      "epoch": 0.43684306458472566,
+      "grad_norm": 0.11874996870756149,
+      "learning_rate": 6.012919389181549e-05,
+      "loss": 1.0469,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4369038216173522,
+      "grad_norm": 0.9322015047073364,
+      "learning_rate": 6.011981942546603e-05,
+      "loss": 1.1716,
+      "step": 7191
+    },
+    {
+      "epoch": 0.43696457864997873,
+      "grad_norm": 0.177250936627388,
+      "learning_rate": 6.0110444588172645e-05,
+      "loss": 1.0946,
+      "step": 7192
+    },
+    {
+      "epoch": 0.43702533568260526,
+      "grad_norm": 0.18527120351791382,
+      "learning_rate": 6.010106938027896e-05,
+      "loss": 1.0729,
+      "step": 7193
+    },
+    {
+      "epoch": 0.4370860927152318,
+      "grad_norm": 0.24348853528499603,
+      "learning_rate": 6.0091693802128614e-05,
+      "loss": 1.0784,
+      "step": 7194
+    },
+    {
+      "epoch": 0.43714684974785833,
+      "grad_norm": 0.1880393922328949,
+      "learning_rate": 6.0082317854065285e-05,
+      "loss": 1.1799,
+      "step": 7195
+    },
+    {
+      "epoch": 0.4372076067804848,
+      "grad_norm": 0.30407989025115967,
+      "learning_rate": 6.0072941536432644e-05,
+      "loss": 1.1176,
+      "step": 7196
+    },
+    {
+      "epoch": 0.43726836381311135,
+      "grad_norm": 0.1593843698501587,
+      "learning_rate": 6.0063564849574385e-05,
+      "loss": 1.1005,
+      "step": 7197
+    },
+    {
+      "epoch": 0.4373291208457379,
+      "grad_norm": 0.12760384380817413,
+      "learning_rate": 6.005418779383421e-05,
+      "loss": 1.0543,
+      "step": 7198
+    },
+    {
+      "epoch": 0.4373898778783644,
+      "grad_norm": 0.1750766932964325,
+      "learning_rate": 6.0044810369555846e-05,
+      "loss": 1.1043,
+      "step": 7199
+    },
+    {
+      "epoch": 0.43745063491099095,
+      "grad_norm": 0.12246454507112503,
+      "learning_rate": 6.003543257708302e-05,
+      "loss": 1.0177,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4375113919436175,
+      "grad_norm": 0.14238403737545013,
+      "learning_rate": 6.002605441675946e-05,
+      "loss": 1.0319,
+      "step": 7201
+    },
+    {
+      "epoch": 0.437572148976244,
+      "grad_norm": 0.2159753292798996,
+      "learning_rate": 6.001667588892893e-05,
+      "loss": 1.1293,
+      "step": 7202
+    },
+    {
+      "epoch": 0.4376329060088705,
+      "grad_norm": 0.26794809103012085,
+      "learning_rate": 6.000729699393523e-05,
+      "loss": 1.0496,
+      "step": 7203
+    },
+    {
+      "epoch": 0.43769366304149704,
+      "grad_norm": 0.18779681622982025,
+      "learning_rate": 5.999791773212211e-05,
+      "loss": 1.1013,
+      "step": 7204
+    },
+    {
+      "epoch": 0.4377544200741236,
+      "grad_norm": 0.22233515977859497,
+      "learning_rate": 5.998853810383339e-05,
+      "loss": 0.9845,
+      "step": 7205
+    },
+    {
+      "epoch": 0.4378151771067501,
+      "grad_norm": 1.4402379989624023,
+      "learning_rate": 5.9979158109412894e-05,
+      "loss": 1.1515,
+      "step": 7206
+    },
+    {
+      "epoch": 0.43787593413937664,
+      "grad_norm": 0.2819216251373291,
+      "learning_rate": 5.996977774920439e-05,
+      "loss": 1.0863,
+      "step": 7207
+    },
+    {
+      "epoch": 0.4379366911720032,
+      "grad_norm": 0.1818607747554779,
+      "learning_rate": 5.996039702355177e-05,
+      "loss": 1.1223,
+      "step": 7208
+    },
+    {
+      "epoch": 0.4379974482046297,
+      "grad_norm": 0.14225588738918304,
+      "learning_rate": 5.995101593279888e-05,
+      "loss": 1.0412,
+      "step": 7209
+    },
+    {
+      "epoch": 0.4380582052372562,
+      "grad_norm": 0.1902163326740265,
+      "learning_rate": 5.9941634477289576e-05,
+      "loss": 1.0985,
+      "step": 7210
+    },
+    {
+      "epoch": 0.43811896226988273,
+      "grad_norm": 0.1476496011018753,
+      "learning_rate": 5.993225265736774e-05,
+      "loss": 1.1104,
+      "step": 7211
+    },
+    {
+      "epoch": 0.43817971930250926,
+      "grad_norm": 0.26813390851020813,
+      "learning_rate": 5.9922870473377244e-05,
+      "loss": 1.0151,
+      "step": 7212
+    },
+    {
+      "epoch": 0.4382404763351358,
+      "grad_norm": 0.15807177126407623,
+      "learning_rate": 5.991348792566204e-05,
+      "loss": 1.0277,
+      "step": 7213
+    },
+    {
+      "epoch": 0.43830123336776233,
+      "grad_norm": 0.3268468976020813,
+      "learning_rate": 5.9904105014566e-05,
+      "loss": 1.1809,
+      "step": 7214
+    },
+    {
+      "epoch": 0.43836199040038887,
+      "grad_norm": 0.37448757886886597,
+      "learning_rate": 5.989472174043308e-05,
+      "loss": 1.0742,
+      "step": 7215
+    },
+    {
+      "epoch": 0.43842274743301535,
+      "grad_norm": 0.23370249569416046,
+      "learning_rate": 5.988533810360724e-05,
+      "loss": 1.0567,
+      "step": 7216
+    },
+    {
+      "epoch": 0.4384835044656419,
+      "grad_norm": 0.5218729376792908,
+      "learning_rate": 5.9875954104432394e-05,
+      "loss": 1.1128,
+      "step": 7217
+    },
+    {
+      "epoch": 0.4385442614982684,
+      "grad_norm": 0.20440687239170074,
+      "learning_rate": 5.9866569743252555e-05,
+      "loss": 1.1062,
+      "step": 7218
+    },
+    {
+      "epoch": 0.43860501853089495,
+      "grad_norm": 0.21167246997356415,
+      "learning_rate": 5.98571850204117e-05,
+      "loss": 1.0505,
+      "step": 7219
+    },
+    {
+      "epoch": 0.4386657755635215,
+      "grad_norm": 0.22053633630275726,
+      "learning_rate": 5.984779993625381e-05,
+      "loss": 1.0248,
+      "step": 7220
+    },
+    {
+      "epoch": 0.438726532596148,
+      "grad_norm": 0.1635219156742096,
+      "learning_rate": 5.983841449112292e-05,
+      "loss": 1.0654,
+      "step": 7221
+    },
+    {
+      "epoch": 0.43878728962877456,
+      "grad_norm": 0.23663997650146484,
+      "learning_rate": 5.9829028685363044e-05,
+      "loss": 1.0579,
+      "step": 7222
+    },
+    {
+      "epoch": 0.43884804666140104,
+      "grad_norm": 0.1273733675479889,
+      "learning_rate": 5.981964251931822e-05,
+      "loss": 1.0797,
+      "step": 7223
+    },
+    {
+      "epoch": 0.43890880369402757,
+      "grad_norm": 0.2503994405269623,
+      "learning_rate": 5.98102559933325e-05,
+      "loss": 1.1365,
+      "step": 7224
+    },
+    {
+      "epoch": 0.4389695607266541,
+      "grad_norm": 0.24734438955783844,
+      "learning_rate": 5.980086910774994e-05,
+      "loss": 1.1173,
+      "step": 7225
+    },
+    {
+      "epoch": 0.43903031775928064,
+      "grad_norm": 0.12965643405914307,
+      "learning_rate": 5.9791481862914645e-05,
+      "loss": 1.0925,
+      "step": 7226
+    },
+    {
+      "epoch": 0.4390910747919072,
+      "grad_norm": 0.21391692757606506,
+      "learning_rate": 5.978209425917067e-05,
+      "loss": 1.2036,
+      "step": 7227
+    },
+    {
+      "epoch": 0.4391518318245337,
+      "grad_norm": 0.3929983079433441,
+      "learning_rate": 5.977270629686216e-05,
+      "loss": 1.0999,
+      "step": 7228
+    },
+    {
+      "epoch": 0.4392125888571602,
+      "grad_norm": 0.18287576735019684,
+      "learning_rate": 5.976331797633321e-05,
+      "loss": 1.0769,
+      "step": 7229
+    },
+    {
+      "epoch": 0.4392733458897867,
+      "grad_norm": 0.16073887050151825,
+      "learning_rate": 5.9753929297927945e-05,
+      "loss": 1.1016,
+      "step": 7230
+    },
+    {
+      "epoch": 0.43933410292241326,
+      "grad_norm": 0.20056360960006714,
+      "learning_rate": 5.974454026199051e-05,
+      "loss": 1.1041,
+      "step": 7231
+    },
+    {
+      "epoch": 0.4393948599550398,
+      "grad_norm": 0.30369260907173157,
+      "learning_rate": 5.9735150868865086e-05,
+      "loss": 1.1303,
+      "step": 7232
+    },
+    {
+      "epoch": 0.43945561698766633,
+      "grad_norm": 0.17466102540493011,
+      "learning_rate": 5.9725761118895816e-05,
+      "loss": 1.1051,
+      "step": 7233
+    },
+    {
+      "epoch": 0.43951637402029287,
+      "grad_norm": 0.1584279090166092,
+      "learning_rate": 5.9716371012426885e-05,
+      "loss": 1.085,
+      "step": 7234
+    },
+    {
+      "epoch": 0.4395771310529194,
+      "grad_norm": 0.25505560636520386,
+      "learning_rate": 5.9706980549802494e-05,
+      "loss": 1.2213,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4396378880855459,
+      "grad_norm": 0.1809869259595871,
+      "learning_rate": 5.969758973136688e-05,
+      "loss": 1.1755,
+      "step": 7236
+    },
+    {
+      "epoch": 0.4396986451181724,
+      "grad_norm": 0.13349549472332,
+      "learning_rate": 5.968819855746423e-05,
+      "loss": 1.0584,
+      "step": 7237
+    },
+    {
+      "epoch": 0.43975940215079895,
+      "grad_norm": 0.2590322494506836,
+      "learning_rate": 5.9678807028438766e-05,
+      "loss": 1.1261,
+      "step": 7238
+    },
+    {
+      "epoch": 0.4398201591834255,
+      "grad_norm": 0.268155038356781,
+      "learning_rate": 5.9669415144634775e-05,
+      "loss": 1.0605,
+      "step": 7239
+    },
+    {
+      "epoch": 0.439880916216052,
+      "grad_norm": 0.14229999482631683,
+      "learning_rate": 5.966002290639651e-05,
+      "loss": 1.0538,
+      "step": 7240
+    },
+    {
+      "epoch": 0.43994167324867856,
+      "grad_norm": 0.17587120831012726,
+      "learning_rate": 5.9650630314068226e-05,
+      "loss": 1.0914,
+      "step": 7241
+    },
+    {
+      "epoch": 0.44000243028130503,
+      "grad_norm": 0.29001328349113464,
+      "learning_rate": 5.964123736799423e-05,
+      "loss": 1.0644,
+      "step": 7242
+    },
+    {
+      "epoch": 0.44006318731393157,
+      "grad_norm": 0.19917793571949005,
+      "learning_rate": 5.9631844068518813e-05,
+      "loss": 1.1887,
+      "step": 7243
+    },
+    {
+      "epoch": 0.4401239443465581,
+      "grad_norm": 0.20727324485778809,
+      "learning_rate": 5.962245041598629e-05,
+      "loss": 1.0929,
+      "step": 7244
+    },
+    {
+      "epoch": 0.44018470137918464,
+      "grad_norm": 0.22194448113441467,
+      "learning_rate": 5.961305641074099e-05,
+      "loss": 1.1479,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4402454584118112,
+      "grad_norm": 0.17089296877384186,
+      "learning_rate": 5.960366205312725e-05,
+      "loss": 1.0884,
+      "step": 7246
+    },
+    {
+      "epoch": 0.4403062154444377,
+      "grad_norm": 0.18409684300422668,
+      "learning_rate": 5.959426734348942e-05,
+      "loss": 1.1724,
+      "step": 7247
+    },
+    {
+      "epoch": 0.44036697247706424,
+      "grad_norm": 0.27629512548446655,
+      "learning_rate": 5.9584872282171845e-05,
+      "loss": 1.2123,
+      "step": 7248
+    },
+    {
+      "epoch": 0.4404277295096907,
+      "grad_norm": 0.18911081552505493,
+      "learning_rate": 5.957547686951894e-05,
+      "loss": 1.0732,
+      "step": 7249
+    },
+    {
+      "epoch": 0.44048848654231726,
+      "grad_norm": 0.19058875739574432,
+      "learning_rate": 5.9566081105875093e-05,
+      "loss": 1.0903,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4405492435749438,
+      "grad_norm": 0.15656369924545288,
+      "learning_rate": 5.955668499158468e-05,
+      "loss": 1.0886,
+      "step": 7251
+    },
+    {
+      "epoch": 0.44061000060757033,
+      "grad_norm": 0.17857365310192108,
+      "learning_rate": 5.954728852699213e-05,
+      "loss": 1.0921,
+      "step": 7252
+    },
+    {
+      "epoch": 0.44067075764019686,
+      "grad_norm": 0.16066035628318787,
+      "learning_rate": 5.9537891712441885e-05,
+      "loss": 1.1337,
+      "step": 7253
+    },
+    {
+      "epoch": 0.4407315146728234,
+      "grad_norm": 0.22143623232841492,
+      "learning_rate": 5.952849454827837e-05,
+      "loss": 1.3762,
+      "step": 7254
+    },
+    {
+      "epoch": 0.4407922717054499,
+      "grad_norm": 0.12572744488716125,
+      "learning_rate": 5.951909703484604e-05,
+      "loss": 1.0206,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4408530287380764,
+      "grad_norm": 0.12908640503883362,
+      "learning_rate": 5.950969917248938e-05,
+      "loss": 1.0384,
+      "step": 7256
+    },
+    {
+      "epoch": 0.44091378577070295,
+      "grad_norm": 0.15913233160972595,
+      "learning_rate": 5.950030096155285e-05,
+      "loss": 1.1174,
+      "step": 7257
+    },
+    {
+      "epoch": 0.4409745428033295,
+      "grad_norm": 0.1637459397315979,
+      "learning_rate": 5.9490902402380944e-05,
+      "loss": 1.0997,
+      "step": 7258
+    },
+    {
+      "epoch": 0.441035299835956,
+      "grad_norm": 0.13719452917575836,
+      "learning_rate": 5.948150349531818e-05,
+      "loss": 1.05,
+      "step": 7259
+    },
+    {
+      "epoch": 0.44109605686858255,
+      "grad_norm": 0.14094413816928864,
+      "learning_rate": 5.947210424070908e-05,
+      "loss": 1.0559,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4411568139012091,
+      "grad_norm": 0.11847657710313797,
+      "learning_rate": 5.946270463889816e-05,
+      "loss": 1.0521,
+      "step": 7261
+    },
+    {
+      "epoch": 0.44121757093383557,
+      "grad_norm": 0.14528438448905945,
+      "learning_rate": 5.945330469022997e-05,
+      "loss": 1.1106,
+      "step": 7262
+    },
+    {
+      "epoch": 0.4412783279664621,
+      "grad_norm": 0.14616583287715912,
+      "learning_rate": 5.94439043950491e-05,
+      "loss": 1.1743,
+      "step": 7263
+    },
+    {
+      "epoch": 0.44133908499908864,
+      "grad_norm": 0.3158397078514099,
+      "learning_rate": 5.943450375370006e-05,
+      "loss": 1.1681,
+      "step": 7264
+    },
+    {
+      "epoch": 0.4413998420317152,
+      "grad_norm": 0.22226400673389435,
+      "learning_rate": 5.942510276652746e-05,
+      "loss": 1.2765,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4414605990643417,
+      "grad_norm": 0.4645533859729767,
+      "learning_rate": 5.94157014338759e-05,
+      "loss": 1.1982,
+      "step": 7266
+    },
+    {
+      "epoch": 0.44152135609696824,
+      "grad_norm": 0.20253103971481323,
+      "learning_rate": 5.940629975608999e-05,
+      "loss": 1.0946,
+      "step": 7267
+    },
+    {
+      "epoch": 0.4415821131295948,
+      "grad_norm": 0.3862606883049011,
+      "learning_rate": 5.939689773351434e-05,
+      "loss": 1.0629,
+      "step": 7268
+    },
+    {
+      "epoch": 0.44164287016222126,
+      "grad_norm": 0.1699996292591095,
+      "learning_rate": 5.938749536649357e-05,
+      "loss": 1.0495,
+      "step": 7269
+    },
+    {
+      "epoch": 0.4417036271948478,
+      "grad_norm": 0.16358983516693115,
+      "learning_rate": 5.937809265537239e-05,
+      "loss": 1.0575,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4417643842274743,
+      "grad_norm": 0.3102813959121704,
+      "learning_rate": 5.936868960049538e-05,
+      "loss": 1.1185,
+      "step": 7271
+    },
+    {
+      "epoch": 0.44182514126010086,
+      "grad_norm": 0.20680458843708038,
+      "learning_rate": 5.935928620220723e-05,
+      "loss": 1.1294,
+      "step": 7272
+    },
+    {
+      "epoch": 0.4418858982927274,
+      "grad_norm": 0.4974614083766937,
+      "learning_rate": 5.934988246085266e-05,
+      "loss": 1.0614,
+      "step": 7273
+    },
+    {
+      "epoch": 0.44194665532535393,
+      "grad_norm": 0.1325724571943283,
+      "learning_rate": 5.934047837677632e-05,
+      "loss": 1.0715,
+      "step": 7274
+    },
+    {
+      "epoch": 0.4420074123579804,
+      "grad_norm": 0.2840862572193146,
+      "learning_rate": 5.933107395032295e-05,
+      "loss": 1.1105,
+      "step": 7275
+    },
+    {
+      "epoch": 0.44206816939060695,
+      "grad_norm": 0.1518881767988205,
+      "learning_rate": 5.932166918183726e-05,
+      "loss": 1.07,
+      "step": 7276
+    },
+    {
+      "epoch": 0.4421289264232335,
+      "grad_norm": 0.2648848295211792,
+      "learning_rate": 5.931226407166398e-05,
+      "loss": 1.1245,
+      "step": 7277
+    },
+    {
+      "epoch": 0.44218968345586,
+      "grad_norm": 0.1568232923746109,
+      "learning_rate": 5.930285862014787e-05,
+      "loss": 1.0539,
+      "step": 7278
+    },
+    {
+      "epoch": 0.44225044048848655,
+      "grad_norm": 0.159840390086174,
+      "learning_rate": 5.929345282763367e-05,
+      "loss": 1.0897,
+      "step": 7279
+    },
+    {
+      "epoch": 0.4423111975211131,
+      "grad_norm": 0.2896788418292999,
+      "learning_rate": 5.928404669446617e-05,
+      "loss": 1.0601,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4423719545537396,
+      "grad_norm": 0.11926238238811493,
+      "learning_rate": 5.927464022099014e-05,
+      "loss": 1.0444,
+      "step": 7281
+    },
+    {
+      "epoch": 0.4424327115863661,
+      "grad_norm": 0.589303731918335,
+      "learning_rate": 5.926523340755037e-05,
+      "loss": 1.136,
+      "step": 7282
+    },
+    {
+      "epoch": 0.44249346861899264,
+      "grad_norm": 0.26594099402427673,
+      "learning_rate": 5.92558262544917e-05,
+      "loss": 1.0513,
+      "step": 7283
+    },
+    {
+      "epoch": 0.44255422565161917,
+      "grad_norm": 0.2291601002216339,
+      "learning_rate": 5.924641876215891e-05,
+      "loss": 1.0835,
+      "step": 7284
+    },
+    {
+      "epoch": 0.4426149826842457,
+      "grad_norm": 0.15537233650684357,
+      "learning_rate": 5.9237010930896856e-05,
+      "loss": 1.0609,
+      "step": 7285
+    },
+    {
+      "epoch": 0.44267573971687224,
+      "grad_norm": 0.2580200135707855,
+      "learning_rate": 5.922760276105038e-05,
+      "loss": 1.1255,
+      "step": 7286
+    },
+    {
+      "epoch": 0.4427364967494988,
+      "grad_norm": 0.26946792006492615,
+      "learning_rate": 5.9218194252964355e-05,
+      "loss": 1.097,
+      "step": 7287
+    },
+    {
+      "epoch": 0.44279725378212526,
+      "grad_norm": 0.25447866320610046,
+      "learning_rate": 5.9208785406983624e-05,
+      "loss": 1.0824,
+      "step": 7288
+    },
+    {
+      "epoch": 0.4428580108147518,
+      "grad_norm": 0.22378745675086975,
+      "learning_rate": 5.919937622345309e-05,
+      "loss": 1.1728,
+      "step": 7289
+    },
+    {
+      "epoch": 0.4429187678473783,
+      "grad_norm": 0.17891372740268707,
+      "learning_rate": 5.9189966702717636e-05,
+      "loss": 1.0486,
+      "step": 7290
+    },
+    {
+      "epoch": 0.44297952488000486,
+      "grad_norm": 0.1861187368631363,
+      "learning_rate": 5.918055684512218e-05,
+      "loss": 1.1243,
+      "step": 7291
+    },
+    {
+      "epoch": 0.4430402819126314,
+      "grad_norm": 0.20694777369499207,
+      "learning_rate": 5.917114665101163e-05,
+      "loss": 1.1523,
+      "step": 7292
+    },
+    {
+      "epoch": 0.44310103894525793,
+      "grad_norm": 0.13934335112571716,
+      "learning_rate": 5.916173612073095e-05,
+      "loss": 1.1038,
+      "step": 7293
+    },
+    {
+      "epoch": 0.44316179597788447,
+      "grad_norm": 0.14663487672805786,
+      "learning_rate": 5.915232525462505e-05,
+      "loss": 1.0764,
+      "step": 7294
+    },
+    {
+      "epoch": 0.44322255301051094,
+      "grad_norm": 0.45416945219039917,
+      "learning_rate": 5.914291405303889e-05,
+      "loss": 1.0441,
+      "step": 7295
+    },
+    {
+      "epoch": 0.4432833100431375,
+      "grad_norm": 0.2938636541366577,
+      "learning_rate": 5.913350251631745e-05,
+      "loss": 1.1389,
+      "step": 7296
+    },
+    {
+      "epoch": 0.443344067075764,
+      "grad_norm": 0.4333260953426361,
+      "learning_rate": 5.9124090644805706e-05,
+      "loss": 1.0706,
+      "step": 7297
+    },
+    {
+      "epoch": 0.44340482410839055,
+      "grad_norm": 0.15775185823440552,
+      "learning_rate": 5.911467843884867e-05,
+      "loss": 1.1443,
+      "step": 7298
+    },
+    {
+      "epoch": 0.4434655811410171,
+      "grad_norm": 0.1447594314813614,
+      "learning_rate": 5.910526589879133e-05,
+      "loss": 1.0979,
+      "step": 7299
+    },
+    {
+      "epoch": 0.4435263381736436,
+      "grad_norm": 0.14317794144153595,
+      "learning_rate": 5.9095853024978706e-05,
+      "loss": 1.0893,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4435870952062701,
+      "grad_norm": 0.20676583051681519,
+      "learning_rate": 5.9086439817755833e-05,
+      "loss": 1.1735,
+      "step": 7301
+    },
+    {
+      "epoch": 0.44364785223889663,
+      "grad_norm": 0.14047355949878693,
+      "learning_rate": 5.907702627746775e-05,
+      "loss": 1.044,
+      "step": 7302
+    },
+    {
+      "epoch": 0.44370860927152317,
+      "grad_norm": 0.16433227062225342,
+      "learning_rate": 5.9067612404459516e-05,
+      "loss": 1.0795,
+      "step": 7303
+    },
+    {
+      "epoch": 0.4437693663041497,
+      "grad_norm": 0.17662374675273895,
+      "learning_rate": 5.90581981990762e-05,
+      "loss": 1.0974,
+      "step": 7304
+    },
+    {
+      "epoch": 0.44383012333677624,
+      "grad_norm": 0.42414590716362,
+      "learning_rate": 5.904878366166285e-05,
+      "loss": 1.1943,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4438908803694028,
+      "grad_norm": 0.2576986849308014,
+      "learning_rate": 5.903936879256462e-05,
+      "loss": 1.1524,
+      "step": 7306
+    },
+    {
+      "epoch": 0.4439516374020293,
+      "grad_norm": 0.21064572036266327,
+      "learning_rate": 5.9029953592126575e-05,
+      "loss": 1.0453,
+      "step": 7307
+    },
+    {
+      "epoch": 0.4440123944346558,
+      "grad_norm": 0.1555577516555786,
+      "learning_rate": 5.902053806069382e-05,
+      "loss": 1.0352,
+      "step": 7308
+    },
+    {
+      "epoch": 0.4440731514672823,
+      "grad_norm": 0.1901993304491043,
+      "learning_rate": 5.9011122198611515e-05,
+      "loss": 1.1059,
+      "step": 7309
+    },
+    {
+      "epoch": 0.44413390849990886,
+      "grad_norm": 0.3174474537372589,
+      "learning_rate": 5.900170600622477e-05,
+      "loss": 1.1541,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4441946655325354,
+      "grad_norm": 0.18357430398464203,
+      "learning_rate": 5.899228948387876e-05,
+      "loss": 1.1525,
+      "step": 7311
+    },
+    {
+      "epoch": 0.44425542256516193,
+      "grad_norm": 0.1494145542383194,
+      "learning_rate": 5.898287263191864e-05,
+      "loss": 1.0441,
+      "step": 7312
+    },
+    {
+      "epoch": 0.44431617959778846,
+      "grad_norm": 0.18338635563850403,
+      "learning_rate": 5.8973455450689586e-05,
+      "loss": 1.1381,
+      "step": 7313
+    },
+    {
+      "epoch": 0.444376936630415,
+      "grad_norm": 0.19906257092952728,
+      "learning_rate": 5.896403794053679e-05,
+      "loss": 1.1647,
+      "step": 7314
+    },
+    {
+      "epoch": 0.4444376936630415,
+      "grad_norm": 0.30824077129364014,
+      "learning_rate": 5.895462010180545e-05,
+      "loss": 1.106,
+      "step": 7315
+    },
+    {
+      "epoch": 0.444498450695668,
+      "grad_norm": 0.14091327786445618,
+      "learning_rate": 5.894520193484077e-05,
+      "loss": 1.0421,
+      "step": 7316
+    },
+    {
+      "epoch": 0.44455920772829455,
+      "grad_norm": 0.30853745341300964,
+      "learning_rate": 5.8935783439988004e-05,
+      "loss": 1.0531,
+      "step": 7317
+    },
+    {
+      "epoch": 0.4446199647609211,
+      "grad_norm": 0.15057380497455597,
+      "learning_rate": 5.892636461759234e-05,
+      "loss": 1.0464,
+      "step": 7318
+    },
+    {
+      "epoch": 0.4446807217935476,
+      "grad_norm": 0.14891479909420013,
+      "learning_rate": 5.891694546799909e-05,
+      "loss": 1.1184,
+      "step": 7319
+    },
+    {
+      "epoch": 0.44474147882617415,
+      "grad_norm": 0.28520748019218445,
+      "learning_rate": 5.8907525991553444e-05,
+      "loss": 1.2007,
+      "step": 7320
+    },
+    {
+      "epoch": 0.44480223585880063,
+      "grad_norm": 0.17570620775222778,
+      "learning_rate": 5.889810618860074e-05,
+      "loss": 1.0749,
+      "step": 7321
+    },
+    {
+      "epoch": 0.44486299289142717,
+      "grad_norm": 0.19225281476974487,
+      "learning_rate": 5.888868605948622e-05,
+      "loss": 1.039,
+      "step": 7322
+    },
+    {
+      "epoch": 0.4449237499240537,
+      "grad_norm": 0.2215781956911087,
+      "learning_rate": 5.88792656045552e-05,
+      "loss": 1.0872,
+      "step": 7323
+    },
+    {
+      "epoch": 0.44498450695668024,
+      "grad_norm": 1.9763462543487549,
+      "learning_rate": 5.8869844824152985e-05,
+      "loss": 1.1749,
+      "step": 7324
+    },
+    {
+      "epoch": 0.44504526398930677,
+      "grad_norm": 0.15236996114253998,
+      "learning_rate": 5.886042371862489e-05,
+      "loss": 0.9973,
+      "step": 7325
+    },
+    {
+      "epoch": 0.4451060210219333,
+      "grad_norm": 0.169319286942482,
+      "learning_rate": 5.8851002288316234e-05,
+      "loss": 1.098,
+      "step": 7326
+    },
+    {
+      "epoch": 0.44516677805455984,
+      "grad_norm": 0.9842188358306885,
+      "learning_rate": 5.884158053357241e-05,
+      "loss": 1.1303,
+      "step": 7327
+    },
+    {
+      "epoch": 0.4452275350871863,
+      "grad_norm": 0.17401878535747528,
+      "learning_rate": 5.883215845473872e-05,
+      "loss": 1.049,
+      "step": 7328
+    },
+    {
+      "epoch": 0.44528829211981286,
+      "grad_norm": 0.2303812950849533,
+      "learning_rate": 5.8822736052160553e-05,
+      "loss": 1.147,
+      "step": 7329
+    },
+    {
+      "epoch": 0.4453490491524394,
+      "grad_norm": 0.21957561373710632,
+      "learning_rate": 5.881331332618332e-05,
+      "loss": 1.1082,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4454098061850659,
+      "grad_norm": 1.7288386821746826,
+      "learning_rate": 5.880389027715235e-05,
+      "loss": 1.3549,
+      "step": 7331
+    },
+    {
+      "epoch": 0.44547056321769246,
+      "grad_norm": 0.27854982018470764,
+      "learning_rate": 5.8794466905413095e-05,
+      "loss": 1.0973,
+      "step": 7332
+    },
+    {
+      "epoch": 0.445531320250319,
+      "grad_norm": 0.1918107122182846,
+      "learning_rate": 5.8785043211310954e-05,
+      "loss": 1.039,
+      "step": 7333
+    },
+    {
+      "epoch": 0.4455920772829455,
+      "grad_norm": 0.1668788641691208,
+      "learning_rate": 5.8775619195191355e-05,
+      "loss": 1.1035,
+      "step": 7334
+    },
+    {
+      "epoch": 0.445652834315572,
+      "grad_norm": 0.9830353260040283,
+      "learning_rate": 5.876619485739974e-05,
+      "loss": 1.1331,
+      "step": 7335
+    },
+    {
+      "epoch": 0.44571359134819855,
+      "grad_norm": 0.17448976635932922,
+      "learning_rate": 5.875677019828154e-05,
+      "loss": 1.0597,
+      "step": 7336
+    },
+    {
+      "epoch": 0.4457743483808251,
+      "grad_norm": 0.17785082757472992,
+      "learning_rate": 5.874734521818227e-05,
+      "loss": 1.0808,
+      "step": 7337
+    },
+    {
+      "epoch": 0.4458351054134516,
+      "grad_norm": 0.12727798521518707,
+      "learning_rate": 5.873791991744735e-05,
+      "loss": 1.0992,
+      "step": 7338
+    },
+    {
+      "epoch": 0.44589586244607815,
+      "grad_norm": 0.2568991482257843,
+      "learning_rate": 5.872849429642228e-05,
+      "loss": 0.9957,
+      "step": 7339
+    },
+    {
+      "epoch": 0.4459566194787047,
+      "grad_norm": 0.19845077395439148,
+      "learning_rate": 5.871906835545259e-05,
+      "loss": 1.0614,
+      "step": 7340
+    },
+    {
+      "epoch": 0.44601737651133117,
+      "grad_norm": 0.13526736199855804,
+      "learning_rate": 5.8709642094883736e-05,
+      "loss": 1.019,
+      "step": 7341
+    },
+    {
+      "epoch": 0.4460781335439577,
+      "grad_norm": 0.20291964709758759,
+      "learning_rate": 5.870021551506127e-05,
+      "loss": 1.0623,
+      "step": 7342
+    },
+    {
+      "epoch": 0.44613889057658423,
+      "grad_norm": 0.19378912448883057,
+      "learning_rate": 5.869078861633075e-05,
+      "loss": 1.184,
+      "step": 7343
+    },
+    {
+      "epoch": 0.44619964760921077,
+      "grad_norm": 0.1606374830007553,
+      "learning_rate": 5.868136139903766e-05,
+      "loss": 1.102,
+      "step": 7344
+    },
+    {
+      "epoch": 0.4462604046418373,
+      "grad_norm": 2.80609393119812,
+      "learning_rate": 5.8671933863527606e-05,
+      "loss": 1.0534,
+      "step": 7345
+    },
+    {
+      "epoch": 0.44632116167446384,
+      "grad_norm": 0.13264694809913635,
+      "learning_rate": 5.8662506010146146e-05,
+      "loss": 1.0595,
+      "step": 7346
+    },
+    {
+      "epoch": 0.4463819187070903,
+      "grad_norm": 0.5925062298774719,
+      "learning_rate": 5.865307783923885e-05,
+      "loss": 1.0506,
+      "step": 7347
+    },
+    {
+      "epoch": 0.44644267573971685,
+      "grad_norm": 0.14357206225395203,
+      "learning_rate": 5.86436493511513e-05,
+      "loss": 1.0716,
+      "step": 7348
+    },
+    {
+      "epoch": 0.4465034327723434,
+      "grad_norm": 0.2187967151403427,
+      "learning_rate": 5.8634220546229115e-05,
+      "loss": 1.1221,
+      "step": 7349
+    },
+    {
+      "epoch": 0.4465641898049699,
+      "grad_norm": 0.11231549084186554,
+      "learning_rate": 5.8624791424817926e-05,
+      "loss": 1.0283,
+      "step": 7350
+    },
+    {
+      "epoch": 0.44662494683759646,
+      "grad_norm": 0.2309916913509369,
+      "learning_rate": 5.8615361987263325e-05,
+      "loss": 1.0409,
+      "step": 7351
+    },
+    {
+      "epoch": 0.446685703870223,
+      "grad_norm": 0.11349146068096161,
+      "learning_rate": 5.860593223391096e-05,
+      "loss": 1.0145,
+      "step": 7352
+    },
+    {
+      "epoch": 0.44674646090284953,
+      "grad_norm": 0.1541716605424881,
+      "learning_rate": 5.8596502165106494e-05,
+      "loss": 1.0077,
+      "step": 7353
+    },
+    {
+      "epoch": 0.446807217935476,
+      "grad_norm": 0.1144527867436409,
+      "learning_rate": 5.858707178119558e-05,
+      "loss": 1.0385,
+      "step": 7354
+    },
+    {
+      "epoch": 0.44686797496810254,
+      "grad_norm": 0.17247967422008514,
+      "learning_rate": 5.857764108252389e-05,
+      "loss": 1.038,
+      "step": 7355
+    },
+    {
+      "epoch": 0.4469287320007291,
+      "grad_norm": 0.1535053551197052,
+      "learning_rate": 5.856821006943711e-05,
+      "loss": 1.0024,
+      "step": 7356
+    },
+    {
+      "epoch": 0.4469894890333556,
+      "grad_norm": 0.14443626999855042,
+      "learning_rate": 5.855877874228093e-05,
+      "loss": 1.0887,
+      "step": 7357
+    },
+    {
+      "epoch": 0.44705024606598215,
+      "grad_norm": 0.18524456024169922,
+      "learning_rate": 5.8549347101401066e-05,
+      "loss": 1.1141,
+      "step": 7358
+    },
+    {
+      "epoch": 0.4471110030986087,
+      "grad_norm": 0.1365940123796463,
+      "learning_rate": 5.853991514714322e-05,
+      "loss": 1.0602,
+      "step": 7359
+    },
+    {
+      "epoch": 0.44717176013123516,
+      "grad_norm": 0.13903789222240448,
+      "learning_rate": 5.8530482879853145e-05,
+      "loss": 0.9976,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4472325171638617,
+      "grad_norm": 0.19113962352275848,
+      "learning_rate": 5.8521050299876566e-05,
+      "loss": 1.182,
+      "step": 7361
+    },
+    {
+      "epoch": 0.44729327419648823,
+      "grad_norm": 0.18511556088924408,
+      "learning_rate": 5.851161740755924e-05,
+      "loss": 1.0859,
+      "step": 7362
+    },
+    {
+      "epoch": 0.44735403122911477,
+      "grad_norm": 0.23406803607940674,
+      "learning_rate": 5.8502184203246924e-05,
+      "loss": 1.0925,
+      "step": 7363
+    },
+    {
+      "epoch": 0.4474147882617413,
+      "grad_norm": 0.19588711857795715,
+      "learning_rate": 5.849275068728543e-05,
+      "loss": 1.1043,
+      "step": 7364
+    },
+    {
+      "epoch": 0.44747554529436784,
+      "grad_norm": 0.2184978872537613,
+      "learning_rate": 5.848331686002049e-05,
+      "loss": 1.1085,
+      "step": 7365
+    },
+    {
+      "epoch": 0.4475363023269944,
+      "grad_norm": 0.19283030927181244,
+      "learning_rate": 5.847388272179795e-05,
+      "loss": 1.1215,
+      "step": 7366
+    },
+    {
+      "epoch": 0.44759705935962085,
+      "grad_norm": 0.14727739989757538,
+      "learning_rate": 5.846444827296359e-05,
+      "loss": 1.0655,
+      "step": 7367
+    },
+    {
+      "epoch": 0.4476578163922474,
+      "grad_norm": 0.22092007100582123,
+      "learning_rate": 5.845501351386324e-05,
+      "loss": 1.1196,
+      "step": 7368
+    },
+    {
+      "epoch": 0.4477185734248739,
+      "grad_norm": 0.2107936292886734,
+      "learning_rate": 5.844557844484274e-05,
+      "loss": 1.0024,
+      "step": 7369
+    },
+    {
+      "epoch": 0.44777933045750046,
+      "grad_norm": 0.1994905322790146,
+      "learning_rate": 5.8436143066247936e-05,
+      "loss": 1.0843,
+      "step": 7370
+    },
+    {
+      "epoch": 0.447840087490127,
+      "grad_norm": 0.11246003210544586,
+      "learning_rate": 5.8426707378424675e-05,
+      "loss": 1.0078,
+      "step": 7371
+    },
+    {
+      "epoch": 0.4479008445227535,
+      "grad_norm": 0.17775876820087433,
+      "learning_rate": 5.841727138171882e-05,
+      "loss": 1.0651,
+      "step": 7372
+    },
+    {
+      "epoch": 0.44796160155538006,
+      "grad_norm": 0.1904994547367096,
+      "learning_rate": 5.840783507647625e-05,
+      "loss": 1.1008,
+      "step": 7373
+    },
+    {
+      "epoch": 0.44802235858800654,
+      "grad_norm": 0.12441320717334747,
+      "learning_rate": 5.839839846304287e-05,
+      "loss": 1.0822,
+      "step": 7374
+    },
+    {
+      "epoch": 0.4480831156206331,
+      "grad_norm": 0.16783198714256287,
+      "learning_rate": 5.838896154176456e-05,
+      "loss": 1.0587,
+      "step": 7375
+    },
+    {
+      "epoch": 0.4481438726532596,
+      "grad_norm": 0.1952381432056427,
+      "learning_rate": 5.837952431298726e-05,
+      "loss": 1.0509,
+      "step": 7376
+    },
+    {
+      "epoch": 0.44820462968588615,
+      "grad_norm": 0.995356023311615,
+      "learning_rate": 5.8370086777056866e-05,
+      "loss": 1.0485,
+      "step": 7377
+    },
+    {
+      "epoch": 0.4482653867185127,
+      "grad_norm": 0.14168232679367065,
+      "learning_rate": 5.836064893431933e-05,
+      "loss": 1.0604,
+      "step": 7378
+    },
+    {
+      "epoch": 0.4483261437511392,
+      "grad_norm": 0.17297153174877167,
+      "learning_rate": 5.835121078512058e-05,
+      "loss": 1.1616,
+      "step": 7379
+    },
+    {
+      "epoch": 0.4483869007837657,
+      "grad_norm": 0.15718460083007812,
+      "learning_rate": 5.83417723298066e-05,
+      "loss": 1.13,
+      "step": 7380
+    },
+    {
+      "epoch": 0.44844765781639223,
+      "grad_norm": 0.12464726716279984,
+      "learning_rate": 5.833233356872333e-05,
+      "loss": 1.0326,
+      "step": 7381
+    },
+    {
+      "epoch": 0.44850841484901877,
+      "grad_norm": 0.19435523450374603,
+      "learning_rate": 5.832289450221677e-05,
+      "loss": 1.0759,
+      "step": 7382
+    },
+    {
+      "epoch": 0.4485691718816453,
+      "grad_norm": 0.1311478614807129,
+      "learning_rate": 5.831345513063289e-05,
+      "loss": 1.0834,
+      "step": 7383
+    },
+    {
+      "epoch": 0.44862992891427184,
+      "grad_norm": 0.2084561586380005,
+      "learning_rate": 5.8304015454317726e-05,
+      "loss": 1.0589,
+      "step": 7384
+    },
+    {
+      "epoch": 0.44869068594689837,
+      "grad_norm": 0.1819782555103302,
+      "learning_rate": 5.8294575473617254e-05,
+      "loss": 1.2061,
+      "step": 7385
+    },
+    {
+      "epoch": 0.4487514429795249,
+      "grad_norm": 0.1541939675807953,
+      "learning_rate": 5.8285135188877517e-05,
+      "loss": 1.0991,
+      "step": 7386
+    },
+    {
+      "epoch": 0.4488122000121514,
+      "grad_norm": 0.34547290205955505,
+      "learning_rate": 5.8275694600444565e-05,
+      "loss": 1.253,
+      "step": 7387
+    },
+    {
+      "epoch": 0.4488729570447779,
+      "grad_norm": 0.30192795395851135,
+      "learning_rate": 5.8266253708664396e-05,
+      "loss": 1.1729,
+      "step": 7388
+    },
+    {
+      "epoch": 0.44893371407740446,
+      "grad_norm": 0.1295192986726761,
+      "learning_rate": 5.825681251388313e-05,
+      "loss": 1.0909,
+      "step": 7389
+    },
+    {
+      "epoch": 0.448994471110031,
+      "grad_norm": 0.16431687772274017,
+      "learning_rate": 5.824737101644679e-05,
+      "loss": 1.1382,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4490552281426575,
+      "grad_norm": 0.17481756210327148,
+      "learning_rate": 5.823792921670147e-05,
+      "loss": 1.1314,
+      "step": 7391
+    },
+    {
+      "epoch": 0.44911598517528406,
+      "grad_norm": 0.22288189828395844,
+      "learning_rate": 5.822848711499327e-05,
+      "loss": 1.0858,
+      "step": 7392
+    },
+    {
+      "epoch": 0.44917674220791054,
+      "grad_norm": 10.75539493560791,
+      "learning_rate": 5.8219044711668266e-05,
+      "loss": 1.0969,
+      "step": 7393
+    },
+    {
+      "epoch": 0.4492374992405371,
+      "grad_norm": 0.15337887406349182,
+      "learning_rate": 5.8209602007072615e-05,
+      "loss": 1.0685,
+      "step": 7394
+    },
+    {
+      "epoch": 0.4492982562731636,
+      "grad_norm": 0.1962725669145584,
+      "learning_rate": 5.8200159001552415e-05,
+      "loss": 1.0642,
+      "step": 7395
+    },
+    {
+      "epoch": 0.44935901330579014,
+      "grad_norm": 0.19740870594978333,
+      "learning_rate": 5.819071569545378e-05,
+      "loss": 1.0856,
+      "step": 7396
+    },
+    {
+      "epoch": 0.4494197703384167,
+      "grad_norm": 0.12731491029262543,
+      "learning_rate": 5.81812720891229e-05,
+      "loss": 1.0393,
+      "step": 7397
+    },
+    {
+      "epoch": 0.4494805273710432,
+      "grad_norm": 0.1905568540096283,
+      "learning_rate": 5.817182818290588e-05,
+      "loss": 1.0499,
+      "step": 7398
+    },
+    {
+      "epoch": 0.44954128440366975,
+      "grad_norm": 0.18223793804645538,
+      "learning_rate": 5.816238397714895e-05,
+      "loss": 1.0821,
+      "step": 7399
+    },
+    {
+      "epoch": 0.44960204143629623,
+      "grad_norm": 0.15613962709903717,
+      "learning_rate": 5.815293947219825e-05,
+      "loss": 1.0468,
+      "step": 7400
+    },
+    {
+      "epoch": 0.44966279846892276,
+      "grad_norm": 0.16405601799488068,
+      "learning_rate": 5.814349466839997e-05,
+      "loss": 1.1184,
+      "step": 7401
+    },
+    {
+      "epoch": 0.4497235555015493,
+      "grad_norm": 0.16086284816265106,
+      "learning_rate": 5.8134049566100325e-05,
+      "loss": 1.0652,
+      "step": 7402
+    },
+    {
+      "epoch": 0.44978431253417583,
+      "grad_norm": 0.1482958048582077,
+      "learning_rate": 5.812460416564551e-05,
+      "loss": 1.0386,
+      "step": 7403
+    },
+    {
+      "epoch": 0.44984506956680237,
+      "grad_norm": 0.16720061004161835,
+      "learning_rate": 5.811515846738177e-05,
+      "loss": 1.056,
+      "step": 7404
+    },
+    {
+      "epoch": 0.4499058265994289,
+      "grad_norm": 0.10713080316781998,
+      "learning_rate": 5.810571247165533e-05,
+      "loss": 1.0274,
+      "step": 7405
+    },
+    {
+      "epoch": 0.4499665836320554,
+      "grad_norm": 0.11764936149120331,
+      "learning_rate": 5.809626617881241e-05,
+      "loss": 0.9931,
+      "step": 7406
+    },
+    {
+      "epoch": 0.4500273406646819,
+      "grad_norm": 0.17225749790668488,
+      "learning_rate": 5.808681958919932e-05,
+      "loss": 1.1447,
+      "step": 7407
+    },
+    {
+      "epoch": 0.45008809769730845,
+      "grad_norm": 1.5785449743270874,
+      "learning_rate": 5.807737270316228e-05,
+      "loss": 1.0848,
+      "step": 7408
+    },
+    {
+      "epoch": 0.450148854729935,
+      "grad_norm": 0.11203330010175705,
+      "learning_rate": 5.806792552104757e-05,
+      "loss": 1.0138,
+      "step": 7409
+    },
+    {
+      "epoch": 0.4502096117625615,
+      "grad_norm": 0.12918095290660858,
+      "learning_rate": 5.8058478043201526e-05,
+      "loss": 1.0493,
+      "step": 7410
+    },
+    {
+      "epoch": 0.45027036879518806,
+      "grad_norm": 0.5991353988647461,
+      "learning_rate": 5.8049030269970374e-05,
+      "loss": 1.0833,
+      "step": 7411
+    },
+    {
+      "epoch": 0.4503311258278146,
+      "grad_norm": 0.39992600679397583,
+      "learning_rate": 5.803958220170047e-05,
+      "loss": 1.1555,
+      "step": 7412
+    },
+    {
+      "epoch": 0.4503918828604411,
+      "grad_norm": 0.2323329597711563,
+      "learning_rate": 5.8030133838738146e-05,
+      "loss": 1.126,
+      "step": 7413
+    },
+    {
+      "epoch": 0.4504526398930676,
+      "grad_norm": 0.1528501659631729,
+      "learning_rate": 5.802068518142969e-05,
+      "loss": 1.0292,
+      "step": 7414
+    },
+    {
+      "epoch": 0.45051339692569414,
+      "grad_norm": 1.09116530418396,
+      "learning_rate": 5.801123623012148e-05,
+      "loss": 1.056,
+      "step": 7415
+    },
+    {
+      "epoch": 0.4505741539583207,
+      "grad_norm": 0.1937636435031891,
+      "learning_rate": 5.8001786985159856e-05,
+      "loss": 1.0361,
+      "step": 7416
+    },
+    {
+      "epoch": 0.4506349109909472,
+      "grad_norm": 0.4029299318790436,
+      "learning_rate": 5.7992337446891186e-05,
+      "loss": 1.1597,
+      "step": 7417
+    },
+    {
+      "epoch": 0.45069566802357375,
+      "grad_norm": 0.47917649149894714,
+      "learning_rate": 5.798288761566184e-05,
+      "loss": 1.215,
+      "step": 7418
+    },
+    {
+      "epoch": 0.4507564250562003,
+      "grad_norm": 0.23846226930618286,
+      "learning_rate": 5.797343749181819e-05,
+      "loss": 1.1331,
+      "step": 7419
+    },
+    {
+      "epoch": 0.45081718208882676,
+      "grad_norm": 0.2682127356529236,
+      "learning_rate": 5.796398707570668e-05,
+      "loss": 1.0445,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4508779391214533,
+      "grad_norm": 0.16246290504932404,
+      "learning_rate": 5.795453636767366e-05,
+      "loss": 1.0531,
+      "step": 7421
+    },
+    {
+      "epoch": 0.45093869615407983,
+      "grad_norm": 0.15183283388614655,
+      "learning_rate": 5.794508536806558e-05,
+      "loss": 1.0492,
+      "step": 7422
+    },
+    {
+      "epoch": 0.45099945318670637,
+      "grad_norm": 0.14682994782924652,
+      "learning_rate": 5.793563407722886e-05,
+      "loss": 1.0382,
+      "step": 7423
+    },
+    {
+      "epoch": 0.4510602102193329,
+      "grad_norm": 0.1822127252817154,
+      "learning_rate": 5.792618249550994e-05,
+      "loss": 1.1113,
+      "step": 7424
+    },
+    {
+      "epoch": 0.45112096725195944,
+      "grad_norm": 0.30428874492645264,
+      "learning_rate": 5.791673062325527e-05,
+      "loss": 1.1447,
+      "step": 7425
+    },
+    {
+      "epoch": 0.4511817242845859,
+      "grad_norm": 0.19708821177482605,
+      "learning_rate": 5.79072784608113e-05,
+      "loss": 1.1603,
+      "step": 7426
+    },
+    {
+      "epoch": 0.45124248131721245,
+      "grad_norm": 0.1499456912279129,
+      "learning_rate": 5.789782600852452e-05,
+      "loss": 1.0522,
+      "step": 7427
+    },
+    {
+      "epoch": 0.451303238349839,
+      "grad_norm": 0.12563274800777435,
+      "learning_rate": 5.7888373266741394e-05,
+      "loss": 1.0535,
+      "step": 7428
+    },
+    {
+      "epoch": 0.4513639953824655,
+      "grad_norm": 0.12478446215391159,
+      "learning_rate": 5.787892023580841e-05,
+      "loss": 1.0219,
+      "step": 7429
+    },
+    {
+      "epoch": 0.45142475241509206,
+      "grad_norm": 0.21333153545856476,
+      "learning_rate": 5.786946691607209e-05,
+      "loss": 1.0769,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4514855094477186,
+      "grad_norm": 0.22072742879390717,
+      "learning_rate": 5.786001330787896e-05,
+      "loss": 1.083,
+      "step": 7431
+    },
+    {
+      "epoch": 0.4515462664803451,
+      "grad_norm": 0.16860386729240417,
+      "learning_rate": 5.78505594115755e-05,
+      "loss": 1.0954,
+      "step": 7432
+    },
+    {
+      "epoch": 0.4516070235129716,
+      "grad_norm": 0.16723968088626862,
+      "learning_rate": 5.784110522750828e-05,
+      "loss": 1.0567,
+      "step": 7433
+    },
+    {
+      "epoch": 0.45166778054559814,
+      "grad_norm": 0.11265820264816284,
+      "learning_rate": 5.783165075602381e-05,
+      "loss": 1.1,
+      "step": 7434
+    },
+    {
+      "epoch": 0.4517285375782247,
+      "grad_norm": 0.2543773055076599,
+      "learning_rate": 5.78221959974687e-05,
+      "loss": 1.2925,
+      "step": 7435
+    },
+    {
+      "epoch": 0.4517892946108512,
+      "grad_norm": 0.2166946679353714,
+      "learning_rate": 5.781274095218946e-05,
+      "loss": 1.0519,
+      "step": 7436
+    },
+    {
+      "epoch": 0.45185005164347775,
+      "grad_norm": 0.14822642505168915,
+      "learning_rate": 5.78032856205327e-05,
+      "loss": 1.1237,
+      "step": 7437
+    },
+    {
+      "epoch": 0.4519108086761043,
+      "grad_norm": 0.13509392738342285,
+      "learning_rate": 5.7793830002845e-05,
+      "loss": 1.0998,
+      "step": 7438
+    },
+    {
+      "epoch": 0.45197156570873076,
+      "grad_norm": 0.13430723547935486,
+      "learning_rate": 5.778437409947295e-05,
+      "loss": 1.0393,
+      "step": 7439
+    },
+    {
+      "epoch": 0.4520323227413573,
+      "grad_norm": 0.30174896121025085,
+      "learning_rate": 5.777491791076315e-05,
+      "loss": 1.2745,
+      "step": 7440
+    },
+    {
+      "epoch": 0.45209307977398383,
+      "grad_norm": 0.14422793686389923,
+      "learning_rate": 5.776546143706225e-05,
+      "loss": 1.1172,
+      "step": 7441
+    },
+    {
+      "epoch": 0.45215383680661037,
+      "grad_norm": 2.848076105117798,
+      "learning_rate": 5.7756004678716847e-05,
+      "loss": 1.0586,
+      "step": 7442
+    },
+    {
+      "epoch": 0.4522145938392369,
+      "grad_norm": 0.1597025990486145,
+      "learning_rate": 5.77465476360736e-05,
+      "loss": 1.0813,
+      "step": 7443
+    },
+    {
+      "epoch": 0.45227535087186344,
+      "grad_norm": 0.14917469024658203,
+      "learning_rate": 5.773709030947916e-05,
+      "loss": 1.0286,
+      "step": 7444
+    },
+    {
+      "epoch": 0.45233610790448997,
+      "grad_norm": 0.19608931243419647,
+      "learning_rate": 5.7727632699280166e-05,
+      "loss": 1.13,
+      "step": 7445
+    },
+    {
+      "epoch": 0.45239686493711645,
+      "grad_norm": 0.14525218307971954,
+      "learning_rate": 5.771817480582331e-05,
+      "loss": 1.1098,
+      "step": 7446
+    },
+    {
+      "epoch": 0.452457621969743,
+      "grad_norm": 0.17831704020500183,
+      "learning_rate": 5.7708716629455264e-05,
+      "loss": 1.0069,
+      "step": 7447
+    },
+    {
+      "epoch": 0.4525183790023695,
+      "grad_norm": 0.1452292799949646,
+      "learning_rate": 5.7699258170522705e-05,
+      "loss": 1.0734,
+      "step": 7448
+    },
+    {
+      "epoch": 0.45257913603499605,
+      "grad_norm": 0.11639715731143951,
+      "learning_rate": 5.768979942937237e-05,
+      "loss": 1.0521,
+      "step": 7449
+    },
+    {
+      "epoch": 0.4526398930676226,
+      "grad_norm": 0.2177213877439499,
+      "learning_rate": 5.768034040635093e-05,
+      "loss": 1.1612,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4527006501002491,
+      "grad_norm": 0.17206627130508423,
+      "learning_rate": 5.767088110180515e-05,
+      "loss": 1.11,
+      "step": 7451
+    },
+    {
+      "epoch": 0.4527614071328756,
+      "grad_norm": 0.14323192834854126,
+      "learning_rate": 5.766142151608174e-05,
+      "loss": 1.0841,
+      "step": 7452
+    },
+    {
+      "epoch": 0.45282216416550214,
+      "grad_norm": 0.13353809714317322,
+      "learning_rate": 5.765196164952742e-05,
+      "loss": 1.0435,
+      "step": 7453
+    },
+    {
+      "epoch": 0.4528829211981287,
+      "grad_norm": 0.16537903249263763,
+      "learning_rate": 5.7642501502488986e-05,
+      "loss": 1.1244,
+      "step": 7454
+    },
+    {
+      "epoch": 0.4529436782307552,
+      "grad_norm": 0.12540307641029358,
+      "learning_rate": 5.7633041075313176e-05,
+      "loss": 1.0307,
+      "step": 7455
+    },
+    {
+      "epoch": 0.45300443526338174,
+      "grad_norm": 0.2685820460319519,
+      "learning_rate": 5.7623580368346765e-05,
+      "loss": 1.2347,
+      "step": 7456
+    },
+    {
+      "epoch": 0.4530651922960083,
+      "grad_norm": 0.11319020390510559,
+      "learning_rate": 5.761411938193656e-05,
+      "loss": 1.0311,
+      "step": 7457
+    },
+    {
+      "epoch": 0.4531259493286348,
+      "grad_norm": 0.19712980091571808,
+      "learning_rate": 5.760465811642932e-05,
+      "loss": 1.1076,
+      "step": 7458
+    },
+    {
+      "epoch": 0.4531867063612613,
+      "grad_norm": 1.0889067649841309,
+      "learning_rate": 5.759519657217187e-05,
+      "loss": 1.0751,
+      "step": 7459
+    },
+    {
+      "epoch": 0.45324746339388783,
+      "grad_norm": 0.16383759677410126,
+      "learning_rate": 5.758573474951102e-05,
+      "loss": 1.0686,
+      "step": 7460
+    },
+    {
+      "epoch": 0.45330822042651436,
+      "grad_norm": 10.906392097473145,
+      "learning_rate": 5.75762726487936e-05,
+      "loss": 1.1086,
+      "step": 7461
+    },
+    {
+      "epoch": 0.4533689774591409,
+      "grad_norm": 0.4894454777240753,
+      "learning_rate": 5.7566810270366445e-05,
+      "loss": 1.0495,
+      "step": 7462
+    },
+    {
+      "epoch": 0.45342973449176743,
+      "grad_norm": 0.20121070742607117,
+      "learning_rate": 5.7557347614576384e-05,
+      "loss": 1.2094,
+      "step": 7463
+    },
+    {
+      "epoch": 0.45349049152439397,
+      "grad_norm": 0.19886942207813263,
+      "learning_rate": 5.75478846817703e-05,
+      "loss": 1.0844,
+      "step": 7464
+    },
+    {
+      "epoch": 0.45355124855702045,
+      "grad_norm": 0.19015374779701233,
+      "learning_rate": 5.753842147229504e-05,
+      "loss": 1.1203,
+      "step": 7465
+    },
+    {
+      "epoch": 0.453612005589647,
+      "grad_norm": 0.24247772991657257,
+      "learning_rate": 5.752895798649748e-05,
+      "loss": 1.0626,
+      "step": 7466
+    },
+    {
+      "epoch": 0.4536727626222735,
+      "grad_norm": 0.18095257878303528,
+      "learning_rate": 5.751949422472452e-05,
+      "loss": 1.1557,
+      "step": 7467
+    },
+    {
+      "epoch": 0.45373351965490005,
+      "grad_norm": 0.1507074534893036,
+      "learning_rate": 5.751003018732303e-05,
+      "loss": 1.0508,
+      "step": 7468
+    },
+    {
+      "epoch": 0.4537942766875266,
+      "grad_norm": 0.1635165959596634,
+      "learning_rate": 5.750056587463995e-05,
+      "loss": 1.0823,
+      "step": 7469
+    },
+    {
+      "epoch": 0.4538550337201531,
+      "grad_norm": 0.16673457622528076,
+      "learning_rate": 5.7491101287022154e-05,
+      "loss": 1.0986,
+      "step": 7470
+    },
+    {
+      "epoch": 0.45391579075277966,
+      "grad_norm": 0.17124459147453308,
+      "learning_rate": 5.748163642481661e-05,
+      "loss": 1.0737,
+      "step": 7471
+    },
+    {
+      "epoch": 0.45397654778540614,
+      "grad_norm": 0.24389341473579407,
+      "learning_rate": 5.7472171288370235e-05,
+      "loss": 1.1192,
+      "step": 7472
+    },
+    {
+      "epoch": 0.45403730481803267,
+      "grad_norm": 0.5769345760345459,
+      "learning_rate": 5.746270587802998e-05,
+      "loss": 1.2811,
+      "step": 7473
+    },
+    {
+      "epoch": 0.4540980618506592,
+      "grad_norm": 0.15695974230766296,
+      "learning_rate": 5.745324019414279e-05,
+      "loss": 1.0716,
+      "step": 7474
+    },
+    {
+      "epoch": 0.45415881888328574,
+      "grad_norm": 3.043384552001953,
+      "learning_rate": 5.7443774237055644e-05,
+      "loss": 1.0556,
+      "step": 7475
+    },
+    {
+      "epoch": 0.4542195759159123,
+      "grad_norm": 0.2506196200847626,
+      "learning_rate": 5.74343080071155e-05,
+      "loss": 1.1081,
+      "step": 7476
+    },
+    {
+      "epoch": 0.4542803329485388,
+      "grad_norm": 0.21501842141151428,
+      "learning_rate": 5.742484150466939e-05,
+      "loss": 1.1938,
+      "step": 7477
+    },
+    {
+      "epoch": 0.45434108998116535,
+      "grad_norm": 0.15379853546619415,
+      "learning_rate": 5.741537473006425e-05,
+      "loss": 1.1079,
+      "step": 7478
+    },
+    {
+      "epoch": 0.4544018470137918,
+      "grad_norm": 0.658745527267456,
+      "learning_rate": 5.7405907683647134e-05,
+      "loss": 1.1059,
+      "step": 7479
+    },
+    {
+      "epoch": 0.45446260404641836,
+      "grad_norm": 0.21668709814548492,
+      "learning_rate": 5.739644036576502e-05,
+      "loss": 1.0096,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4545233610790449,
+      "grad_norm": 0.14211909472942352,
+      "learning_rate": 5.738697277676498e-05,
+      "loss": 1.0324,
+      "step": 7481
+    },
+    {
+      "epoch": 0.45458411811167143,
+      "grad_norm": 1.3377307653427124,
+      "learning_rate": 5.737750491699402e-05,
+      "loss": 1.2082,
+      "step": 7482
+    },
+    {
+      "epoch": 0.45464487514429797,
+      "grad_norm": 0.1769695281982422,
+      "learning_rate": 5.736803678679918e-05,
+      "loss": 1.053,
+      "step": 7483
+    },
+    {
+      "epoch": 0.4547056321769245,
+      "grad_norm": 0.15515007078647614,
+      "learning_rate": 5.735856838652754e-05,
+      "loss": 1.0571,
+      "step": 7484
+    },
+    {
+      "epoch": 0.454766389209551,
+      "grad_norm": 0.24353644251823425,
+      "learning_rate": 5.734909971652614e-05,
+      "loss": 1.0264,
+      "step": 7485
+    },
+    {
+      "epoch": 0.4548271462421775,
+      "grad_norm": 0.19325301051139832,
+      "learning_rate": 5.733963077714206e-05,
+      "loss": 1.2901,
+      "step": 7486
+    },
+    {
+      "epoch": 0.45488790327480405,
+      "grad_norm": 0.16973786056041718,
+      "learning_rate": 5.7330161568722426e-05,
+      "loss": 1.0374,
+      "step": 7487
+    },
+    {
+      "epoch": 0.4549486603074306,
+      "grad_norm": 0.25033992528915405,
+      "learning_rate": 5.732069209161429e-05,
+      "loss": 1.2324,
+      "step": 7488
+    },
+    {
+      "epoch": 0.4550094173400571,
+      "grad_norm": 49.92509460449219,
+      "learning_rate": 5.731122234616476e-05,
+      "loss": 1.0527,
+      "step": 7489
+    },
+    {
+      "epoch": 0.45507017437268366,
+      "grad_norm": 0.21506592631340027,
+      "learning_rate": 5.730175233272097e-05,
+      "loss": 1.0632,
+      "step": 7490
+    },
+    {
+      "epoch": 0.4551309314053102,
+      "grad_norm": 3.1655020713806152,
+      "learning_rate": 5.729228205163003e-05,
+      "loss": 1.0467,
+      "step": 7491
+    },
+    {
+      "epoch": 0.45519168843793667,
+      "grad_norm": 0.2095041573047638,
+      "learning_rate": 5.7282811503239096e-05,
+      "loss": 1.1275,
+      "step": 7492
+    },
+    {
+      "epoch": 0.4552524454705632,
+      "grad_norm": 0.3740250766277313,
+      "learning_rate": 5.727334068789529e-05,
+      "loss": 1.1432,
+      "step": 7493
+    },
+    {
+      "epoch": 0.45531320250318974,
+      "grad_norm": 0.12217546254396439,
+      "learning_rate": 5.7263869605945775e-05,
+      "loss": 1.0468,
+      "step": 7494
+    },
+    {
+      "epoch": 0.4553739595358163,
+      "grad_norm": 0.16230589151382446,
+      "learning_rate": 5.725439825773772e-05,
+      "loss": 1.0893,
+      "step": 7495
+    },
+    {
+      "epoch": 0.4554347165684428,
+      "grad_norm": 0.13122041523456573,
+      "learning_rate": 5.724492664361829e-05,
+      "loss": 1.052,
+      "step": 7496
+    },
+    {
+      "epoch": 0.45549547360106935,
+      "grad_norm": 0.23191942274570465,
+      "learning_rate": 5.7235454763934657e-05,
+      "loss": 1.0789,
+      "step": 7497
+    },
+    {
+      "epoch": 0.4555562306336958,
+      "grad_norm": 0.13227961957454681,
+      "learning_rate": 5.722598261903407e-05,
+      "loss": 1.0579,
+      "step": 7498
+    },
+    {
+      "epoch": 0.45561698766632236,
+      "grad_norm": 0.1584184467792511,
+      "learning_rate": 5.721651020926366e-05,
+      "loss": 1.0694,
+      "step": 7499
+    },
+    {
+      "epoch": 0.4556777446989489,
+      "grad_norm": 4.345352649688721,
+      "learning_rate": 5.7207037534970686e-05,
+      "loss": 1.1363,
+      "step": 7500
+    },
+    {
+      "epoch": 0.45573850173157543,
+      "grad_norm": 0.3370049297809601,
+      "learning_rate": 5.719756459650237e-05,
+      "loss": 1.0893,
+      "step": 7501
+    },
+    {
+      "epoch": 0.45579925876420196,
+      "grad_norm": 0.1856507509946823,
+      "learning_rate": 5.718809139420591e-05,
+      "loss": 1.2368,
+      "step": 7502
+    },
+    {
+      "epoch": 0.4558600157968285,
+      "grad_norm": 0.4172630310058594,
+      "learning_rate": 5.717861792842858e-05,
+      "loss": 1.3132,
+      "step": 7503
+    },
+    {
+      "epoch": 0.45592077282945503,
+      "grad_norm": 0.22857438027858734,
+      "learning_rate": 5.716914419951762e-05,
+      "loss": 1.0564,
+      "step": 7504
+    },
+    {
+      "epoch": 0.4559815298620815,
+      "grad_norm": 0.15569455921649933,
+      "learning_rate": 5.7159670207820305e-05,
+      "loss": 1.0379,
+      "step": 7505
+    },
+    {
+      "epoch": 0.45604228689470805,
+      "grad_norm": 0.134469136595726,
+      "learning_rate": 5.715019595368388e-05,
+      "loss": 1.0775,
+      "step": 7506
+    },
+    {
+      "epoch": 0.4561030439273346,
+      "grad_norm": 0.11792246997356415,
+      "learning_rate": 5.714072143745564e-05,
+      "loss": 1.0335,
+      "step": 7507
+    },
+    {
+      "epoch": 0.4561638009599611,
+      "grad_norm": 0.18644648790359497,
+      "learning_rate": 5.71312466594829e-05,
+      "loss": 1.1581,
+      "step": 7508
+    },
+    {
+      "epoch": 0.45622455799258765,
+      "grad_norm": 0.11870857328176498,
+      "learning_rate": 5.712177162011291e-05,
+      "loss": 1.0161,
+      "step": 7509
+    },
+    {
+      "epoch": 0.4562853150252142,
+      "grad_norm": 0.13469670712947845,
+      "learning_rate": 5.7112296319693004e-05,
+      "loss": 1.0613,
+      "step": 7510
+    },
+    {
+      "epoch": 0.45634607205784067,
+      "grad_norm": 0.3815622329711914,
+      "learning_rate": 5.710282075857052e-05,
+      "loss": 1.1575,
+      "step": 7511
+    },
+    {
+      "epoch": 0.4564068290904672,
+      "grad_norm": 0.22992122173309326,
+      "learning_rate": 5.709334493709274e-05,
+      "loss": 1.0239,
+      "step": 7512
+    },
+    {
+      "epoch": 0.45646758612309374,
+      "grad_norm": 0.18595284223556519,
+      "learning_rate": 5.7083868855607045e-05,
+      "loss": 1.0607,
+      "step": 7513
+    },
+    {
+      "epoch": 0.4565283431557203,
+      "grad_norm": 0.1342955082654953,
+      "learning_rate": 5.707439251446077e-05,
+      "loss": 1.077,
+      "step": 7514
+    },
+    {
+      "epoch": 0.4565891001883468,
+      "grad_norm": 0.28029853105545044,
+      "learning_rate": 5.706491591400128e-05,
+      "loss": 1.0669,
+      "step": 7515
+    },
+    {
+      "epoch": 0.45664985722097334,
+      "grad_norm": 0.2969526946544647,
+      "learning_rate": 5.705543905457592e-05,
+      "loss": 1.1455,
+      "step": 7516
+    },
+    {
+      "epoch": 0.4567106142535999,
+      "grad_norm": 0.32625094056129456,
+      "learning_rate": 5.704596193653208e-05,
+      "loss": 1.1397,
+      "step": 7517
+    },
+    {
+      "epoch": 0.45677137128622636,
+      "grad_norm": 0.13720357418060303,
+      "learning_rate": 5.703648456021714e-05,
+      "loss": 1.0565,
+      "step": 7518
+    },
+    {
+      "epoch": 0.4568321283188529,
+      "grad_norm": 0.21891836822032928,
+      "learning_rate": 5.702700692597851e-05,
+      "loss": 1.1714,
+      "step": 7519
+    },
+    {
+      "epoch": 0.4568928853514794,
+      "grad_norm": 0.3338230550289154,
+      "learning_rate": 5.701752903416356e-05,
+      "loss": 1.1148,
+      "step": 7520
+    },
+    {
+      "epoch": 0.45695364238410596,
+      "grad_norm": 0.2304871827363968,
+      "learning_rate": 5.7008050885119754e-05,
+      "loss": 1.0731,
+      "step": 7521
+    },
+    {
+      "epoch": 0.4570143994167325,
+      "grad_norm": 0.2524801194667816,
+      "learning_rate": 5.699857247919447e-05,
+      "loss": 1.1334,
+      "step": 7522
+    },
+    {
+      "epoch": 0.45707515644935903,
+      "grad_norm": 0.25523653626441956,
+      "learning_rate": 5.698909381673516e-05,
+      "loss": 1.0527,
+      "step": 7523
+    },
+    {
+      "epoch": 0.45713591348198557,
+      "grad_norm": 0.24774916470050812,
+      "learning_rate": 5.697961489808927e-05,
+      "loss": 1.1022,
+      "step": 7524
+    },
+    {
+      "epoch": 0.45719667051461205,
+      "grad_norm": 0.4545222222805023,
+      "learning_rate": 5.697013572360424e-05,
+      "loss": 1.0904,
+      "step": 7525
+    },
+    {
+      "epoch": 0.4572574275472386,
+      "grad_norm": 0.2186569720506668,
+      "learning_rate": 5.696065629362755e-05,
+      "loss": 1.2372,
+      "step": 7526
+    },
+    {
+      "epoch": 0.4573181845798651,
+      "grad_norm": 0.44130274653434753,
+      "learning_rate": 5.695117660850665e-05,
+      "loss": 1.0495,
+      "step": 7527
+    },
+    {
+      "epoch": 0.45737894161249165,
+      "grad_norm": 0.1566094607114792,
+      "learning_rate": 5.694169666858903e-05,
+      "loss": 1.0535,
+      "step": 7528
+    },
+    {
+      "epoch": 0.4574396986451182,
+      "grad_norm": 0.25207868218421936,
+      "learning_rate": 5.693221647422218e-05,
+      "loss": 1.0542,
+      "step": 7529
+    },
+    {
+      "epoch": 0.4575004556777447,
+      "grad_norm": 0.2549058198928833,
+      "learning_rate": 5.692273602575358e-05,
+      "loss": 1.0901,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4575612127103712,
+      "grad_norm": 0.2928144037723541,
+      "learning_rate": 5.691325532353078e-05,
+      "loss": 1.0308,
+      "step": 7531
+    },
+    {
+      "epoch": 0.45762196974299774,
+      "grad_norm": 1.2693920135498047,
+      "learning_rate": 5.690377436790125e-05,
+      "loss": 1.1426,
+      "step": 7532
+    },
+    {
+      "epoch": 0.45768272677562427,
+      "grad_norm": 2.5480690002441406,
+      "learning_rate": 5.689429315921253e-05,
+      "loss": 1.0497,
+      "step": 7533
+    },
+    {
+      "epoch": 0.4577434838082508,
+      "grad_norm": 0.18954262137413025,
+      "learning_rate": 5.688481169781218e-05,
+      "loss": 1.1384,
+      "step": 7534
+    },
+    {
+      "epoch": 0.45780424084087734,
+      "grad_norm": 0.23400700092315674,
+      "learning_rate": 5.687532998404771e-05,
+      "loss": 1.1508,
+      "step": 7535
+    },
+    {
+      "epoch": 0.4578649978735039,
+      "grad_norm": 0.15430310368537903,
+      "learning_rate": 5.68658480182667e-05,
+      "loss": 1.0806,
+      "step": 7536
+    },
+    {
+      "epoch": 0.4579257549061304,
+      "grad_norm": 0.24934610724449158,
+      "learning_rate": 5.68563658008167e-05,
+      "loss": 1.0373,
+      "step": 7537
+    },
+    {
+      "epoch": 0.4579865119387569,
+      "grad_norm": 0.3733794391155243,
+      "learning_rate": 5.6846883332045284e-05,
+      "loss": 1.1144,
+      "step": 7538
+    },
+    {
+      "epoch": 0.4580472689713834,
+      "grad_norm": 0.15330719947814941,
+      "learning_rate": 5.683740061230005e-05,
+      "loss": 1.0456,
+      "step": 7539
+    },
+    {
+      "epoch": 0.45810802600400996,
+      "grad_norm": 0.2150934636592865,
+      "learning_rate": 5.6827917641928566e-05,
+      "loss": 1.1329,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4581687830366365,
+      "grad_norm": 0.2593918740749359,
+      "learning_rate": 5.681843442127844e-05,
+      "loss": 1.165,
+      "step": 7541
+    },
+    {
+      "epoch": 0.45822954006926303,
+      "grad_norm": 0.3031356930732727,
+      "learning_rate": 5.680895095069728e-05,
+      "loss": 1.0436,
+      "step": 7542
+    },
+    {
+      "epoch": 0.45829029710188957,
+      "grad_norm": 0.2594084143638611,
+      "learning_rate": 5.67994672305327e-05,
+      "loss": 1.2377,
+      "step": 7543
+    },
+    {
+      "epoch": 0.45835105413451604,
+      "grad_norm": 0.2219531387090683,
+      "learning_rate": 5.678998326113236e-05,
+      "loss": 1.0988,
+      "step": 7544
+    },
+    {
+      "epoch": 0.4584118111671426,
+      "grad_norm": 0.18685947358608246,
+      "learning_rate": 5.678049904284385e-05,
+      "loss": 1.0624,
+      "step": 7545
+    },
+    {
+      "epoch": 0.4584725681997691,
+      "grad_norm": 0.21842002868652344,
+      "learning_rate": 5.6771014576014836e-05,
+      "loss": 1.0707,
+      "step": 7546
+    },
+    {
+      "epoch": 0.45853332523239565,
+      "grad_norm": 0.3374932110309601,
+      "learning_rate": 5.676152986099298e-05,
+      "loss": 1.2428,
+      "step": 7547
+    },
+    {
+      "epoch": 0.4585940822650222,
+      "grad_norm": 0.1564943939447403,
+      "learning_rate": 5.675204489812594e-05,
+      "loss": 1.0611,
+      "step": 7548
+    },
+    {
+      "epoch": 0.4586548392976487,
+      "grad_norm": 0.2009422332048416,
+      "learning_rate": 5.674255968776139e-05,
+      "loss": 1.1097,
+      "step": 7549
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 0.29928141832351685,
+      "learning_rate": 5.673307423024701e-05,
+      "loss": 1.0432,
+      "step": 7550
+    },
+    {
+      "epoch": 0.45877635336290173,
+      "grad_norm": 0.17595785856246948,
+      "learning_rate": 5.6723588525930496e-05,
+      "loss": 1.1462,
+      "step": 7551
+    },
+    {
+      "epoch": 0.45883711039552827,
+      "grad_norm": 0.34310516715049744,
+      "learning_rate": 5.671410257515955e-05,
+      "loss": 1.2845,
+      "step": 7552
+    },
+    {
+      "epoch": 0.4588978674281548,
+      "grad_norm": 0.13096091151237488,
+      "learning_rate": 5.670461637828186e-05,
+      "loss": 1.0821,
+      "step": 7553
+    },
+    {
+      "epoch": 0.45895862446078134,
+      "grad_norm": 0.17904110252857208,
+      "learning_rate": 5.669512993564516e-05,
+      "loss": 1.1369,
+      "step": 7554
+    },
+    {
+      "epoch": 0.4590193814934079,
+      "grad_norm": 0.20433752238750458,
+      "learning_rate": 5.66856432475972e-05,
+      "loss": 1.1922,
+      "step": 7555
+    },
+    {
+      "epoch": 0.4590801385260344,
+      "grad_norm": 0.17755264043807983,
+      "learning_rate": 5.6676156314485665e-05,
+      "loss": 1.0175,
+      "step": 7556
+    },
+    {
+      "epoch": 0.4591408955586609,
+      "grad_norm": 0.1238652840256691,
+      "learning_rate": 5.6666669136658343e-05,
+      "loss": 1.0429,
+      "step": 7557
+    },
+    {
+      "epoch": 0.4592016525912874,
+      "grad_norm": 0.1593741774559021,
+      "learning_rate": 5.665718171446299e-05,
+      "loss": 1.079,
+      "step": 7558
+    },
+    {
+      "epoch": 0.45926240962391396,
+      "grad_norm": 0.14473868906497955,
+      "learning_rate": 5.6647694048247335e-05,
+      "loss": 1.0854,
+      "step": 7559
+    },
+    {
+      "epoch": 0.4593231666565405,
+      "grad_norm": 0.2751701772212982,
+      "learning_rate": 5.6638206138359185e-05,
+      "loss": 1.2336,
+      "step": 7560
+    },
+    {
+      "epoch": 0.45938392368916703,
+      "grad_norm": 0.15657395124435425,
+      "learning_rate": 5.662871798514629e-05,
+      "loss": 1.0734,
+      "step": 7561
+    },
+    {
+      "epoch": 0.45944468072179356,
+      "grad_norm": 0.39454635977745056,
+      "learning_rate": 5.661922958895647e-05,
+      "loss": 1.1759,
+      "step": 7562
+    },
+    {
+      "epoch": 0.4595054377544201,
+      "grad_norm": 0.2491476684808731,
+      "learning_rate": 5.66097409501375e-05,
+      "loss": 1.046,
+      "step": 7563
+    },
+    {
+      "epoch": 0.4595661947870466,
+      "grad_norm": 0.24620307981967926,
+      "learning_rate": 5.6600252069037206e-05,
+      "loss": 1.0304,
+      "step": 7564
+    },
+    {
+      "epoch": 0.4596269518196731,
+      "grad_norm": 0.16522061824798584,
+      "learning_rate": 5.65907629460034e-05,
+      "loss": 1.0829,
+      "step": 7565
+    },
+    {
+      "epoch": 0.45968770885229965,
+      "grad_norm": 0.23070620000362396,
+      "learning_rate": 5.65812735813839e-05,
+      "loss": 1.0888,
+      "step": 7566
+    },
+    {
+      "epoch": 0.4597484658849262,
+      "grad_norm": 0.13351227343082428,
+      "learning_rate": 5.6571783975526523e-05,
+      "loss": 1.1077,
+      "step": 7567
+    },
+    {
+      "epoch": 0.4598092229175527,
+      "grad_norm": 0.5957766771316528,
+      "learning_rate": 5.656229412877917e-05,
+      "loss": 1.0399,
+      "step": 7568
+    },
+    {
+      "epoch": 0.45986997995017925,
+      "grad_norm": 0.19671595096588135,
+      "learning_rate": 5.655280404148963e-05,
+      "loss": 1.0529,
+      "step": 7569
+    },
+    {
+      "epoch": 0.45993073698280573,
+      "grad_norm": 0.19638168811798096,
+      "learning_rate": 5.65433137140058e-05,
+      "loss": 1.1049,
+      "step": 7570
+    },
+    {
+      "epoch": 0.45999149401543227,
+      "grad_norm": 0.1827968806028366,
+      "learning_rate": 5.6533823146675546e-05,
+      "loss": 1.0553,
+      "step": 7571
+    },
+    {
+      "epoch": 0.4600522510480588,
+      "grad_norm": 0.6725801825523376,
+      "learning_rate": 5.652433233984673e-05,
+      "loss": 1.0822,
+      "step": 7572
+    },
+    {
+      "epoch": 0.46011300808068534,
+      "grad_norm": 0.1218404695391655,
+      "learning_rate": 5.651484129386726e-05,
+      "loss": 1.0449,
+      "step": 7573
+    },
+    {
+      "epoch": 0.4601737651133119,
+      "grad_norm": 0.2364734709262848,
+      "learning_rate": 5.6505350009085024e-05,
+      "loss": 1.0068,
+      "step": 7574
+    },
+    {
+      "epoch": 0.4602345221459384,
+      "grad_norm": 1.5241683721542358,
+      "learning_rate": 5.649585848584793e-05,
+      "loss": 1.1521,
+      "step": 7575
+    },
+    {
+      "epoch": 0.46029527917856494,
+      "grad_norm": 0.14631372690200806,
+      "learning_rate": 5.648636672450388e-05,
+      "loss": 1.0765,
+      "step": 7576
+    },
+    {
+      "epoch": 0.4603560362111914,
+      "grad_norm": 0.3273763060569763,
+      "learning_rate": 5.6476874725400795e-05,
+      "loss": 1.1211,
+      "step": 7577
+    },
+    {
+      "epoch": 0.46041679324381796,
+      "grad_norm": 0.15713606774806976,
+      "learning_rate": 5.646738248888663e-05,
+      "loss": 1.0777,
+      "step": 7578
+    },
+    {
+      "epoch": 0.4604775502764445,
+      "grad_norm": 0.29719293117523193,
+      "learning_rate": 5.64578900153093e-05,
+      "loss": 1.2381,
+      "step": 7579
+    },
+    {
+      "epoch": 0.460538307309071,
+      "grad_norm": 0.4006941318511963,
+      "learning_rate": 5.6448397305016766e-05,
+      "loss": 1.0431,
+      "step": 7580
+    },
+    {
+      "epoch": 0.46059906434169756,
+      "grad_norm": 0.3481408953666687,
+      "learning_rate": 5.6438904358357004e-05,
+      "loss": 1.0488,
+      "step": 7581
+    },
+    {
+      "epoch": 0.4606598213743241,
+      "grad_norm": 0.18954573571681976,
+      "learning_rate": 5.642941117567793e-05,
+      "loss": 1.0892,
+      "step": 7582
+    },
+    {
+      "epoch": 0.46072057840695063,
+      "grad_norm": 0.32978764176368713,
+      "learning_rate": 5.6419917757327555e-05,
+      "loss": 1.078,
+      "step": 7583
+    },
+    {
+      "epoch": 0.4607813354395771,
+      "grad_norm": 0.17325016856193542,
+      "learning_rate": 5.6410424103653844e-05,
+      "loss": 1.1239,
+      "step": 7584
+    },
+    {
+      "epoch": 0.46084209247220365,
+      "grad_norm": 0.19177311658859253,
+      "learning_rate": 5.640093021500481e-05,
+      "loss": 1.0542,
+      "step": 7585
+    },
+    {
+      "epoch": 0.4609028495048302,
+      "grad_norm": 0.18840324878692627,
+      "learning_rate": 5.6391436091728446e-05,
+      "loss": 1.1353,
+      "step": 7586
+    },
+    {
+      "epoch": 0.4609636065374567,
+      "grad_norm": 0.152621328830719,
+      "learning_rate": 5.638194173417274e-05,
+      "loss": 1.0428,
+      "step": 7587
+    },
+    {
+      "epoch": 0.46102436357008325,
+      "grad_norm": 0.17865662276744843,
+      "learning_rate": 5.637244714268575e-05,
+      "loss": 1.1341,
+      "step": 7588
+    },
+    {
+      "epoch": 0.4610851206027098,
+      "grad_norm": 1.0702911615371704,
+      "learning_rate": 5.636295231761547e-05,
+      "loss": 1.0964,
+      "step": 7589
+    },
+    {
+      "epoch": 0.46114587763533627,
+      "grad_norm": 0.18569421768188477,
+      "learning_rate": 5.6353457259309925e-05,
+      "loss": 1.0688,
+      "step": 7590
+    },
+    {
+      "epoch": 0.4612066346679628,
+      "grad_norm": 0.15140080451965332,
+      "learning_rate": 5.6343961968117196e-05,
+      "loss": 1.062,
+      "step": 7591
+    },
+    {
+      "epoch": 0.46126739170058934,
+      "grad_norm": 0.516845703125,
+      "learning_rate": 5.63344664443853e-05,
+      "loss": 1.0543,
+      "step": 7592
+    },
+    {
+      "epoch": 0.46132814873321587,
+      "grad_norm": 0.19516772031784058,
+      "learning_rate": 5.632497068846232e-05,
+      "loss": 1.0706,
+      "step": 7593
+    },
+    {
+      "epoch": 0.4613889057658424,
+      "grad_norm": 0.1170029565691948,
+      "learning_rate": 5.631547470069631e-05,
+      "loss": 1.0205,
+      "step": 7594
+    },
+    {
+      "epoch": 0.46144966279846894,
+      "grad_norm": 0.4323507249355316,
+      "learning_rate": 5.630597848143537e-05,
+      "loss": 1.0735,
+      "step": 7595
+    },
+    {
+      "epoch": 0.4615104198310955,
+      "grad_norm": 0.32684460282325745,
+      "learning_rate": 5.6296482031027565e-05,
+      "loss": 1.157,
+      "step": 7596
+    },
+    {
+      "epoch": 0.46157117686372195,
+      "grad_norm": 0.29347795248031616,
+      "learning_rate": 5.6286985349820986e-05,
+      "loss": 1.1309,
+      "step": 7597
+    },
+    {
+      "epoch": 0.4616319338963485,
+      "grad_norm": 0.18857432901859283,
+      "learning_rate": 5.627748843816375e-05,
+      "loss": 1.1816,
+      "step": 7598
+    },
+    {
+      "epoch": 0.461692690928975,
+      "grad_norm": 7.915004253387451,
+      "learning_rate": 5.626799129640395e-05,
+      "loss": 1.1834,
+      "step": 7599
+    },
+    {
+      "epoch": 0.46175344796160156,
+      "grad_norm": 0.2563534677028656,
+      "learning_rate": 5.625849392488972e-05,
+      "loss": 1.2312,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4618142049942281,
+      "grad_norm": 0.18239572644233704,
+      "learning_rate": 5.62489963239692e-05,
+      "loss": 1.1721,
+      "step": 7601
+    },
+    {
+      "epoch": 0.46187496202685463,
+      "grad_norm": 0.14203126728534698,
+      "learning_rate": 5.623949849399052e-05,
+      "loss": 1.0179,
+      "step": 7602
+    },
+    {
+      "epoch": 0.4619357190594811,
+      "grad_norm": 0.12551337480545044,
+      "learning_rate": 5.623000043530179e-05,
+      "loss": 1.0602,
+      "step": 7603
+    },
+    {
+      "epoch": 0.46199647609210764,
+      "grad_norm": 0.1708313375711441,
+      "learning_rate": 5.6220502148251197e-05,
+      "loss": 1.1203,
+      "step": 7604
+    },
+    {
+      "epoch": 0.4620572331247342,
+      "grad_norm": 0.12785588204860687,
+      "learning_rate": 5.62110036331869e-05,
+      "loss": 1.0581,
+      "step": 7605
+    },
+    {
+      "epoch": 0.4621179901573607,
+      "grad_norm": 0.25426387786865234,
+      "learning_rate": 5.620150489045706e-05,
+      "loss": 1.1265,
+      "step": 7606
+    },
+    {
+      "epoch": 0.46217874718998725,
+      "grad_norm": 0.1477641463279724,
+      "learning_rate": 5.619200592040986e-05,
+      "loss": 1.0383,
+      "step": 7607
+    },
+    {
+      "epoch": 0.4622395042226138,
+      "grad_norm": 0.17011034488677979,
+      "learning_rate": 5.618250672339349e-05,
+      "loss": 1.0614,
+      "step": 7608
+    },
+    {
+      "epoch": 0.4623002612552403,
+      "grad_norm": 0.5043832063674927,
+      "learning_rate": 5.617300729975614e-05,
+      "loss": 1.1171,
+      "step": 7609
+    },
+    {
+      "epoch": 0.4623610182878668,
+      "grad_norm": 0.16781552135944366,
+      "learning_rate": 5.6163507649846026e-05,
+      "loss": 1.0651,
+      "step": 7610
+    },
+    {
+      "epoch": 0.46242177532049333,
+      "grad_norm": 0.1682317703962326,
+      "learning_rate": 5.615400777401133e-05,
+      "loss": 1.1196,
+      "step": 7611
+    },
+    {
+      "epoch": 0.46248253235311987,
+      "grad_norm": 0.1677902340888977,
+      "learning_rate": 5.61445076726003e-05,
+      "loss": 1.1295,
+      "step": 7612
+    },
+    {
+      "epoch": 0.4625432893857464,
+      "grad_norm": 3.1851911544799805,
+      "learning_rate": 5.613500734596113e-05,
+      "loss": 1.0841,
+      "step": 7613
+    },
+    {
+      "epoch": 0.46260404641837294,
+      "grad_norm": 0.19222573935985565,
+      "learning_rate": 5.612550679444211e-05,
+      "loss": 1.0268,
+      "step": 7614
+    },
+    {
+      "epoch": 0.4626648034509995,
+      "grad_norm": 0.30901357531547546,
+      "learning_rate": 5.611600601839144e-05,
+      "loss": 1.2294,
+      "step": 7615
+    },
+    {
+      "epoch": 0.46272556048362595,
+      "grad_norm": 0.1423233449459076,
+      "learning_rate": 5.610650501815739e-05,
+      "loss": 1.0767,
+      "step": 7616
+    },
+    {
+      "epoch": 0.4627863175162525,
+      "grad_norm": 0.1298656016588211,
+      "learning_rate": 5.609700379408822e-05,
+      "loss": 1.075,
+      "step": 7617
+    },
+    {
+      "epoch": 0.462847074548879,
+      "grad_norm": 0.27051857113838196,
+      "learning_rate": 5.60875023465322e-05,
+      "loss": 1.2075,
+      "step": 7618
+    },
+    {
+      "epoch": 0.46290783158150556,
+      "grad_norm": 0.12271862477064133,
+      "learning_rate": 5.6078000675837594e-05,
+      "loss": 1.0499,
+      "step": 7619
+    },
+    {
+      "epoch": 0.4629685886141321,
+      "grad_norm": 0.1604657918214798,
+      "learning_rate": 5.6068498782352707e-05,
+      "loss": 1.0606,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4630293456467586,
+      "grad_norm": 0.14413392543792725,
+      "learning_rate": 5.605899666642582e-05,
+      "loss": 1.0435,
+      "step": 7621
+    },
+    {
+      "epoch": 0.46309010267938516,
+      "grad_norm": 0.2147381454706192,
+      "learning_rate": 5.604949432840525e-05,
+      "loss": 1.0417,
+      "step": 7622
+    },
+    {
+      "epoch": 0.46315085971201164,
+      "grad_norm": 0.15062421560287476,
+      "learning_rate": 5.6039991768639276e-05,
+      "loss": 1.1245,
+      "step": 7623
+    },
+    {
+      "epoch": 0.4632116167446382,
+      "grad_norm": 0.15868224203586578,
+      "learning_rate": 5.603048898747625e-05,
+      "loss": 1.1355,
+      "step": 7624
+    },
+    {
+      "epoch": 0.4632723737772647,
+      "grad_norm": 0.19187390804290771,
+      "learning_rate": 5.6020985985264486e-05,
+      "loss": 1.1011,
+      "step": 7625
+    },
+    {
+      "epoch": 0.46333313080989125,
+      "grad_norm": 0.20561785995960236,
+      "learning_rate": 5.601148276235231e-05,
+      "loss": 1.1224,
+      "step": 7626
+    },
+    {
+      "epoch": 0.4633938878425178,
+      "grad_norm": 0.17380262911319733,
+      "learning_rate": 5.600197931908807e-05,
+      "loss": 1.147,
+      "step": 7627
+    },
+    {
+      "epoch": 0.4634546448751443,
+      "grad_norm": 0.1822991967201233,
+      "learning_rate": 5.599247565582012e-05,
+      "loss": 1.1022,
+      "step": 7628
+    },
+    {
+      "epoch": 0.46351540190777085,
+      "grad_norm": 0.45203229784965515,
+      "learning_rate": 5.598297177289682e-05,
+      "loss": 1.1577,
+      "step": 7629
+    },
+    {
+      "epoch": 0.46357615894039733,
+      "grad_norm": 0.14951534569263458,
+      "learning_rate": 5.597346767066654e-05,
+      "loss": 1.0283,
+      "step": 7630
+    },
+    {
+      "epoch": 0.46363691597302387,
+      "grad_norm": 0.3066413104534149,
+      "learning_rate": 5.5963963349477636e-05,
+      "loss": 1.0693,
+      "step": 7631
+    },
+    {
+      "epoch": 0.4636976730056504,
+      "grad_norm": 0.8103142976760864,
+      "learning_rate": 5.59544588096785e-05,
+      "loss": 1.073,
+      "step": 7632
+    },
+    {
+      "epoch": 0.46375843003827694,
+      "grad_norm": 0.34123125672340393,
+      "learning_rate": 5.594495405161754e-05,
+      "loss": 1.1859,
+      "step": 7633
+    },
+    {
+      "epoch": 0.46381918707090347,
+      "grad_norm": 0.1669309288263321,
+      "learning_rate": 5.593544907564312e-05,
+      "loss": 1.0837,
+      "step": 7634
+    },
+    {
+      "epoch": 0.46387994410353,
+      "grad_norm": 0.22812524437904358,
+      "learning_rate": 5.592594388210369e-05,
+      "loss": 1.0743,
+      "step": 7635
+    },
+    {
+      "epoch": 0.4639407011361565,
+      "grad_norm": 1.2388594150543213,
+      "learning_rate": 5.591643847134762e-05,
+      "loss": 1.0607,
+      "step": 7636
+    },
+    {
+      "epoch": 0.464001458168783,
+      "grad_norm": 1.0885109901428223,
+      "learning_rate": 5.590693284372336e-05,
+      "loss": 1.0916,
+      "step": 7637
+    },
+    {
+      "epoch": 0.46406221520140956,
+      "grad_norm": 0.13822364807128906,
+      "learning_rate": 5.589742699957935e-05,
+      "loss": 1.0739,
+      "step": 7638
+    },
+    {
+      "epoch": 0.4641229722340361,
+      "grad_norm": 0.1727365404367447,
+      "learning_rate": 5.588792093926399e-05,
+      "loss": 1.0989,
+      "step": 7639
+    },
+    {
+      "epoch": 0.4641837292666626,
+      "grad_norm": 0.2509419322013855,
+      "learning_rate": 5.5878414663125765e-05,
+      "loss": 1.124,
+      "step": 7640
+    },
+    {
+      "epoch": 0.46424448629928916,
+      "grad_norm": 0.2248162031173706,
+      "learning_rate": 5.586890817151311e-05,
+      "loss": 1.0429,
+      "step": 7641
+    },
+    {
+      "epoch": 0.4643052433319157,
+      "grad_norm": 0.18183498084545135,
+      "learning_rate": 5.58594014647745e-05,
+      "loss": 1.0876,
+      "step": 7642
+    },
+    {
+      "epoch": 0.4643660003645422,
+      "grad_norm": 0.19940079748630524,
+      "learning_rate": 5.5849894543258396e-05,
+      "loss": 1.1727,
+      "step": 7643
+    },
+    {
+      "epoch": 0.4644267573971687,
+      "grad_norm": 0.12469612061977386,
+      "learning_rate": 5.5840387407313254e-05,
+      "loss": 1.0311,
+      "step": 7644
+    },
+    {
+      "epoch": 0.46448751442979525,
+      "grad_norm": 0.17444831132888794,
+      "learning_rate": 5.5830880057287616e-05,
+      "loss": 1.084,
+      "step": 7645
+    },
+    {
+      "epoch": 0.4645482714624218,
+      "grad_norm": 2.5646579265594482,
+      "learning_rate": 5.5821372493529924e-05,
+      "loss": 1.0094,
+      "step": 7646
+    },
+    {
+      "epoch": 0.4646090284950483,
+      "grad_norm": 0.12972000241279602,
+      "learning_rate": 5.581186471638868e-05,
+      "loss": 1.0733,
+      "step": 7647
+    },
+    {
+      "epoch": 0.46466978552767485,
+      "grad_norm": 0.4693816900253296,
+      "learning_rate": 5.580235672621246e-05,
+      "loss": 1.133,
+      "step": 7648
+    },
+    {
+      "epoch": 0.46473054256030133,
+      "grad_norm": 0.18356047570705414,
+      "learning_rate": 5.5792848523349684e-05,
+      "loss": 1.181,
+      "step": 7649
+    },
+    {
+      "epoch": 0.46479129959292786,
+      "grad_norm": 0.24481543898582458,
+      "learning_rate": 5.578334010814895e-05,
+      "loss": 1.0815,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4648520566255544,
+      "grad_norm": 0.3973977565765381,
+      "learning_rate": 5.577383148095876e-05,
+      "loss": 1.072,
+      "step": 7651
+    },
+    {
+      "epoch": 0.46491281365818093,
+      "grad_norm": 0.29410886764526367,
+      "learning_rate": 5.5764322642127665e-05,
+      "loss": 1.1073,
+      "step": 7652
+    },
+    {
+      "epoch": 0.46497357069080747,
+      "grad_norm": 0.27756214141845703,
+      "learning_rate": 5.575481359200421e-05,
+      "loss": 1.1719,
+      "step": 7653
+    },
+    {
+      "epoch": 0.465034327723434,
+      "grad_norm": 0.24933436512947083,
+      "learning_rate": 5.5745304330936945e-05,
+      "loss": 1.3256,
+      "step": 7654
+    },
+    {
+      "epoch": 0.46509508475606054,
+      "grad_norm": 0.12517014145851135,
+      "learning_rate": 5.5735794859274436e-05,
+      "loss": 1.015,
+      "step": 7655
+    },
+    {
+      "epoch": 0.465155841788687,
+      "grad_norm": 0.16356950998306274,
+      "learning_rate": 5.5726285177365265e-05,
+      "loss": 1.1986,
+      "step": 7656
+    },
+    {
+      "epoch": 0.46521659882131355,
+      "grad_norm": 0.1978103518486023,
+      "learning_rate": 5.5716775285558e-05,
+      "loss": 1.0759,
+      "step": 7657
+    },
+    {
+      "epoch": 0.4652773558539401,
+      "grad_norm": 0.17229701578617096,
+      "learning_rate": 5.570726518420124e-05,
+      "loss": 1.1028,
+      "step": 7658
+    },
+    {
+      "epoch": 0.4653381128865666,
+      "grad_norm": 0.23430059850215912,
+      "learning_rate": 5.569775487364357e-05,
+      "loss": 1.1268,
+      "step": 7659
+    },
+    {
+      "epoch": 0.46539886991919316,
+      "grad_norm": 0.14466476440429688,
+      "learning_rate": 5.5688244354233577e-05,
+      "loss": 1.0849,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4654596269518197,
+      "grad_norm": 0.165469229221344,
+      "learning_rate": 5.56787336263199e-05,
+      "loss": 1.0958,
+      "step": 7661
+    },
+    {
+      "epoch": 0.4655203839844462,
+      "grad_norm": 0.13033407926559448,
+      "learning_rate": 5.5669222690251146e-05,
+      "loss": 1.0625,
+      "step": 7662
+    },
+    {
+      "epoch": 0.4655811410170727,
+      "grad_norm": 0.13559192419052124,
+      "learning_rate": 5.5659711546375935e-05,
+      "loss": 1.0696,
+      "step": 7663
+    },
+    {
+      "epoch": 0.46564189804969924,
+      "grad_norm": 0.23872654139995575,
+      "learning_rate": 5.5650200195042914e-05,
+      "loss": 1.1407,
+      "step": 7664
+    },
+    {
+      "epoch": 0.4657026550823258,
+      "grad_norm": 0.2885741591453552,
+      "learning_rate": 5.56406886366007e-05,
+      "loss": 1.1699,
+      "step": 7665
+    },
+    {
+      "epoch": 0.4657634121149523,
+      "grad_norm": 0.2483404576778412,
+      "learning_rate": 5.563117687139796e-05,
+      "loss": 1.1875,
+      "step": 7666
+    },
+    {
+      "epoch": 0.46582416914757885,
+      "grad_norm": 0.17019130289554596,
+      "learning_rate": 5.5621664899783345e-05,
+      "loss": 1.1328,
+      "step": 7667
+    },
+    {
+      "epoch": 0.4658849261802054,
+      "grad_norm": 0.19060491025447845,
+      "learning_rate": 5.561215272210551e-05,
+      "loss": 1.1019,
+      "step": 7668
+    },
+    {
+      "epoch": 0.46594568321283186,
+      "grad_norm": 0.30430859327316284,
+      "learning_rate": 5.560264033871313e-05,
+      "loss": 1.1902,
+      "step": 7669
+    },
+    {
+      "epoch": 0.4660064402454584,
+      "grad_norm": 0.20313583314418793,
+      "learning_rate": 5.559312774995489e-05,
+      "loss": 1.1339,
+      "step": 7670
+    },
+    {
+      "epoch": 0.46606719727808493,
+      "grad_norm": 0.12355385720729828,
+      "learning_rate": 5.5583614956179464e-05,
+      "loss": 1.0667,
+      "step": 7671
+    },
+    {
+      "epoch": 0.46612795431071147,
+      "grad_norm": 0.13360348343849182,
+      "learning_rate": 5.5574101957735556e-05,
+      "loss": 1.0758,
+      "step": 7672
+    },
+    {
+      "epoch": 0.466188711343338,
+      "grad_norm": 0.16960491240024567,
+      "learning_rate": 5.556458875497187e-05,
+      "loss": 1.0487,
+      "step": 7673
+    },
+    {
+      "epoch": 0.46624946837596454,
+      "grad_norm": 0.14360320568084717,
+      "learning_rate": 5.555507534823711e-05,
+      "loss": 1.0692,
+      "step": 7674
+    },
+    {
+      "epoch": 0.466310225408591,
+      "grad_norm": 0.11860251426696777,
+      "learning_rate": 5.554556173787998e-05,
+      "loss": 1.0709,
+      "step": 7675
+    },
+    {
+      "epoch": 0.46637098244121755,
+      "grad_norm": 0.3663264513015747,
+      "learning_rate": 5.553604792424922e-05,
+      "loss": 1.072,
+      "step": 7676
+    },
+    {
+      "epoch": 0.4664317394738441,
+      "grad_norm": 0.39243003726005554,
+      "learning_rate": 5.552653390769356e-05,
+      "loss": 1.0582,
+      "step": 7677
+    },
+    {
+      "epoch": 0.4664924965064706,
+      "grad_norm": 0.15361706912517548,
+      "learning_rate": 5.551701968856171e-05,
+      "loss": 1.0509,
+      "step": 7678
+    },
+    {
+      "epoch": 0.46655325353909716,
+      "grad_norm": 0.21406547725200653,
+      "learning_rate": 5.5507505267202456e-05,
+      "loss": 1.1269,
+      "step": 7679
+    },
+    {
+      "epoch": 0.4666140105717237,
+      "grad_norm": 0.23460249602794647,
+      "learning_rate": 5.5497990643964505e-05,
+      "loss": 1.1227,
+      "step": 7680
+    },
+    {
+      "epoch": 0.4666747676043502,
+      "grad_norm": 0.14040905237197876,
+      "learning_rate": 5.5488475819196674e-05,
+      "loss": 1.1328,
+      "step": 7681
+    },
+    {
+      "epoch": 0.4667355246369767,
+      "grad_norm": 0.21273450553417206,
+      "learning_rate": 5.547896079324769e-05,
+      "loss": 1.0463,
+      "step": 7682
+    },
+    {
+      "epoch": 0.46679628166960324,
+      "grad_norm": 0.12193840742111206,
+      "learning_rate": 5.546944556646634e-05,
+      "loss": 1.0542,
+      "step": 7683
+    },
+    {
+      "epoch": 0.4668570387022298,
+      "grad_norm": 0.34236642718315125,
+      "learning_rate": 5.54599301392014e-05,
+      "loss": 1.0547,
+      "step": 7684
+    },
+    {
+      "epoch": 0.4669177957348563,
+      "grad_norm": 0.16691485047340393,
+      "learning_rate": 5.545041451180167e-05,
+      "loss": 1.127,
+      "step": 7685
+    },
+    {
+      "epoch": 0.46697855276748285,
+      "grad_norm": 0.19637003540992737,
+      "learning_rate": 5.544089868461594e-05,
+      "loss": 1.075,
+      "step": 7686
+    },
+    {
+      "epoch": 0.4670393098001094,
+      "grad_norm": 0.20217357575893402,
+      "learning_rate": 5.5431382657993025e-05,
+      "loss": 1.1611,
+      "step": 7687
+    },
+    {
+      "epoch": 0.4671000668327359,
+      "grad_norm": 0.14192472398281097,
+      "learning_rate": 5.542186643228172e-05,
+      "loss": 1.0641,
+      "step": 7688
+    },
+    {
+      "epoch": 0.4671608238653624,
+      "grad_norm": 0.2243974655866623,
+      "learning_rate": 5.541235000783086e-05,
+      "loss": 1.0851,
+      "step": 7689
+    },
+    {
+      "epoch": 0.46722158089798893,
+      "grad_norm": 0.15024195611476898,
+      "learning_rate": 5.540283338498926e-05,
+      "loss": 1.0835,
+      "step": 7690
+    },
+    {
+      "epoch": 0.46728233793061547,
+      "grad_norm": 0.14921025931835175,
+      "learning_rate": 5.539331656410575e-05,
+      "loss": 1.0471,
+      "step": 7691
+    },
+    {
+      "epoch": 0.467343094963242,
+      "grad_norm": 0.20188233256340027,
+      "learning_rate": 5.5383799545529205e-05,
+      "loss": 1.0741,
+      "step": 7692
+    },
+    {
+      "epoch": 0.46740385199586854,
+      "grad_norm": 0.17332637310028076,
+      "learning_rate": 5.5374282329608404e-05,
+      "loss": 1.0258,
+      "step": 7693
+    },
+    {
+      "epoch": 0.46746460902849507,
+      "grad_norm": 0.317302942276001,
+      "learning_rate": 5.5364764916692277e-05,
+      "loss": 1.1503,
+      "step": 7694
+    },
+    {
+      "epoch": 0.46752536606112155,
+      "grad_norm": 0.23128226399421692,
+      "learning_rate": 5.5355247307129664e-05,
+      "loss": 1.0749,
+      "step": 7695
+    },
+    {
+      "epoch": 0.4675861230937481,
+      "grad_norm": 0.21716414391994476,
+      "learning_rate": 5.5345729501269396e-05,
+      "loss": 1.2077,
+      "step": 7696
+    },
+    {
+      "epoch": 0.4676468801263746,
+      "grad_norm": 0.41507014632225037,
+      "learning_rate": 5.53362114994604e-05,
+      "loss": 1.0666,
+      "step": 7697
+    },
+    {
+      "epoch": 0.46770763715900115,
+      "grad_norm": 0.14391931891441345,
+      "learning_rate": 5.5326693302051537e-05,
+      "loss": 1.0611,
+      "step": 7698
+    },
+    {
+      "epoch": 0.4677683941916277,
+      "grad_norm": 0.15689924359321594,
+      "learning_rate": 5.531717490939169e-05,
+      "loss": 1.0617,
+      "step": 7699
+    },
+    {
+      "epoch": 0.4678291512242542,
+      "grad_norm": 0.15438872575759888,
+      "learning_rate": 5.530765632182978e-05,
+      "loss": 1.1145,
+      "step": 7700
+    },
+    {
+      "epoch": 0.46788990825688076,
+      "grad_norm": 0.2057807892560959,
+      "learning_rate": 5.529813753971469e-05,
+      "loss": 1.1581,
+      "step": 7701
+    },
+    {
+      "epoch": 0.46795066528950724,
+      "grad_norm": 29.648632049560547,
+      "learning_rate": 5.5288618563395366e-05,
+      "loss": 1.0382,
+      "step": 7702
+    },
+    {
+      "epoch": 0.4680114223221338,
+      "grad_norm": 0.15054571628570557,
+      "learning_rate": 5.527909939322069e-05,
+      "loss": 1.0575,
+      "step": 7703
+    },
+    {
+      "epoch": 0.4680721793547603,
+      "grad_norm": 0.1688757836818695,
+      "learning_rate": 5.5269580029539605e-05,
+      "loss": 1.1171,
+      "step": 7704
+    },
+    {
+      "epoch": 0.46813293638738684,
+      "grad_norm": 0.20824523270130157,
+      "learning_rate": 5.526006047270106e-05,
+      "loss": 1.0807,
+      "step": 7705
+    },
+    {
+      "epoch": 0.4681936934200134,
+      "grad_norm": 0.1984352320432663,
+      "learning_rate": 5.525054072305395e-05,
+      "loss": 1.0547,
+      "step": 7706
+    },
+    {
+      "epoch": 0.4682544504526399,
+      "grad_norm": 0.11433682590723038,
+      "learning_rate": 5.5241020780947284e-05,
+      "loss": 1.0176,
+      "step": 7707
+    },
+    {
+      "epoch": 0.4683152074852664,
+      "grad_norm": 0.24109522998332977,
+      "learning_rate": 5.523150064672999e-05,
+      "loss": 1.2032,
+      "step": 7708
+    },
+    {
+      "epoch": 0.46837596451789293,
+      "grad_norm": 0.17710252106189728,
+      "learning_rate": 5.522198032075101e-05,
+      "loss": 1.0757,
+      "step": 7709
+    },
+    {
+      "epoch": 0.46843672155051946,
+      "grad_norm": 0.1847608983516693,
+      "learning_rate": 5.521245980335934e-05,
+      "loss": 1.0615,
+      "step": 7710
+    },
+    {
+      "epoch": 0.468497478583146,
+      "grad_norm": 0.12425071746110916,
+      "learning_rate": 5.520293909490396e-05,
+      "loss": 1.0648,
+      "step": 7711
+    },
+    {
+      "epoch": 0.46855823561577253,
+      "grad_norm": 0.22357232868671417,
+      "learning_rate": 5.5193418195733825e-05,
+      "loss": 1.0918,
+      "step": 7712
+    },
+    {
+      "epoch": 0.46861899264839907,
+      "grad_norm": 0.19007642567157745,
+      "learning_rate": 5.5183897106197945e-05,
+      "loss": 1.1689,
+      "step": 7713
+    },
+    {
+      "epoch": 0.4686797496810256,
+      "grad_norm": 0.2071160525083542,
+      "learning_rate": 5.517437582664531e-05,
+      "loss": 1.0804,
+      "step": 7714
+    },
+    {
+      "epoch": 0.4687405067136521,
+      "grad_norm": 0.14910295605659485,
+      "learning_rate": 5.516485435742496e-05,
+      "loss": 1.1555,
+      "step": 7715
+    },
+    {
+      "epoch": 0.4688012637462786,
+      "grad_norm": 0.12315890938043594,
+      "learning_rate": 5.515533269888583e-05,
+      "loss": 1.0108,
+      "step": 7716
+    },
+    {
+      "epoch": 0.46886202077890515,
+      "grad_norm": 0.18095339834690094,
+      "learning_rate": 5.514581085137701e-05,
+      "loss": 1.0851,
+      "step": 7717
+    },
+    {
+      "epoch": 0.4689227778115317,
+      "grad_norm": 0.20210176706314087,
+      "learning_rate": 5.51362888152475e-05,
+      "loss": 1.0758,
+      "step": 7718
+    },
+    {
+      "epoch": 0.4689835348441582,
+      "grad_norm": 0.1383674144744873,
+      "learning_rate": 5.512676659084631e-05,
+      "loss": 1.0767,
+      "step": 7719
+    },
+    {
+      "epoch": 0.46904429187678476,
+      "grad_norm": 0.1770932674407959,
+      "learning_rate": 5.511724417852252e-05,
+      "loss": 1.1355,
+      "step": 7720
+    },
+    {
+      "epoch": 0.46910504890941124,
+      "grad_norm": 0.1914380043745041,
+      "learning_rate": 5.510772157862515e-05,
+      "loss": 1.0826,
+      "step": 7721
+    },
+    {
+      "epoch": 0.4691658059420378,
+      "grad_norm": 0.13793520629405975,
+      "learning_rate": 5.5098198791503265e-05,
+      "loss": 1.0382,
+      "step": 7722
+    },
+    {
+      "epoch": 0.4692265629746643,
+      "grad_norm": 0.18655921518802643,
+      "learning_rate": 5.5088675817505905e-05,
+      "loss": 1.1243,
+      "step": 7723
+    },
+    {
+      "epoch": 0.46928732000729084,
+      "grad_norm": 2.703308343887329,
+      "learning_rate": 5.507915265698215e-05,
+      "loss": 1.0956,
+      "step": 7724
+    },
+    {
+      "epoch": 0.4693480770399174,
+      "grad_norm": 0.5491136312484741,
+      "learning_rate": 5.506962931028109e-05,
+      "loss": 1.0964,
+      "step": 7725
+    },
+    {
+      "epoch": 0.4694088340725439,
+      "grad_norm": 1.9153352975845337,
+      "learning_rate": 5.506010577775177e-05,
+      "loss": 1.0614,
+      "step": 7726
+    },
+    {
+      "epoch": 0.46946959110517045,
+      "grad_norm": 0.22011426091194153,
+      "learning_rate": 5.5050582059743296e-05,
+      "loss": 1.1205,
+      "step": 7727
+    },
+    {
+      "epoch": 0.4695303481377969,
+      "grad_norm": 0.2554408311843872,
+      "learning_rate": 5.5041058156604766e-05,
+      "loss": 1.0518,
+      "step": 7728
+    },
+    {
+      "epoch": 0.46959110517042346,
+      "grad_norm": 0.25837868452072144,
+      "learning_rate": 5.503153406868527e-05,
+      "loss": 1.1819,
+      "step": 7729
+    },
+    {
+      "epoch": 0.46965186220305,
+      "grad_norm": 0.1197812557220459,
+      "learning_rate": 5.502200979633392e-05,
+      "loss": 1.0153,
+      "step": 7730
+    },
+    {
+      "epoch": 0.46971261923567653,
+      "grad_norm": 0.1912653148174286,
+      "learning_rate": 5.5012485339899835e-05,
+      "loss": 1.1036,
+      "step": 7731
+    },
+    {
+      "epoch": 0.46977337626830307,
+      "grad_norm": 0.17191733419895172,
+      "learning_rate": 5.5002960699732135e-05,
+      "loss": 1.1622,
+      "step": 7732
+    },
+    {
+      "epoch": 0.4698341333009296,
+      "grad_norm": 0.14717574417591095,
+      "learning_rate": 5.499343587617993e-05,
+      "loss": 1.0963,
+      "step": 7733
+    },
+    {
+      "epoch": 0.46989489033355614,
+      "grad_norm": 0.16133275628089905,
+      "learning_rate": 5.4983910869592384e-05,
+      "loss": 1.0896,
+      "step": 7734
+    },
+    {
+      "epoch": 0.4699556473661826,
+      "grad_norm": 0.15688002109527588,
+      "learning_rate": 5.497438568031861e-05,
+      "loss": 1.0214,
+      "step": 7735
+    },
+    {
+      "epoch": 0.47001640439880915,
+      "grad_norm": 0.24089637398719788,
+      "learning_rate": 5.496486030870777e-05,
+      "loss": 1.1374,
+      "step": 7736
+    },
+    {
+      "epoch": 0.4700771614314357,
+      "grad_norm": 0.1316508948802948,
+      "learning_rate": 5.4955334755109014e-05,
+      "loss": 1.0939,
+      "step": 7737
+    },
+    {
+      "epoch": 0.4701379184640622,
+      "grad_norm": 2.5219810009002686,
+      "learning_rate": 5.49458090198715e-05,
+      "loss": 1.0275,
+      "step": 7738
+    },
+    {
+      "epoch": 0.47019867549668876,
+      "grad_norm": 0.284229576587677,
+      "learning_rate": 5.493628310334442e-05,
+      "loss": 1.2056,
+      "step": 7739
+    },
+    {
+      "epoch": 0.4702594325293153,
+      "grad_norm": 0.1665724366903305,
+      "learning_rate": 5.4926757005876904e-05,
+      "loss": 1.1052,
+      "step": 7740
+    },
+    {
+      "epoch": 0.47032018956194177,
+      "grad_norm": 0.14781175553798676,
+      "learning_rate": 5.491723072781817e-05,
+      "loss": 1.1326,
+      "step": 7741
+    },
+    {
+      "epoch": 0.4703809465945683,
+      "grad_norm": 0.15880252420902252,
+      "learning_rate": 5.4907704269517394e-05,
+      "loss": 1.1092,
+      "step": 7742
+    },
+    {
+      "epoch": 0.47044170362719484,
+      "grad_norm": 0.22209881246089935,
+      "learning_rate": 5.489817763132377e-05,
+      "loss": 1.107,
+      "step": 7743
+    },
+    {
+      "epoch": 0.4705024606598214,
+      "grad_norm": 0.2783887982368469,
+      "learning_rate": 5.4888650813586494e-05,
+      "loss": 1.1081,
+      "step": 7744
+    },
+    {
+      "epoch": 0.4705632176924479,
+      "grad_norm": 0.22827184200286865,
+      "learning_rate": 5.4879123816654786e-05,
+      "loss": 1.1623,
+      "step": 7745
+    },
+    {
+      "epoch": 0.47062397472507445,
+      "grad_norm": 0.4204244911670685,
+      "learning_rate": 5.486959664087784e-05,
+      "loss": 1.0821,
+      "step": 7746
+    },
+    {
+      "epoch": 0.470684731757701,
+      "grad_norm": 0.12838798761367798,
+      "learning_rate": 5.4860069286604896e-05,
+      "loss": 1.0201,
+      "step": 7747
+    },
+    {
+      "epoch": 0.47074548879032746,
+      "grad_norm": 0.1261807531118393,
+      "learning_rate": 5.485054175418516e-05,
+      "loss": 1.0269,
+      "step": 7748
+    },
+    {
+      "epoch": 0.470806245822954,
+      "grad_norm": 0.25944215059280396,
+      "learning_rate": 5.48410140439679e-05,
+      "loss": 1.0849,
+      "step": 7749
+    },
+    {
+      "epoch": 0.47086700285558053,
+      "grad_norm": 0.19042851030826569,
+      "learning_rate": 5.483148615630231e-05,
+      "loss": 1.1666,
+      "step": 7750
+    },
+    {
+      "epoch": 0.47092775988820706,
+      "grad_norm": 0.6102296710014343,
+      "learning_rate": 5.4821958091537675e-05,
+      "loss": 1.0643,
+      "step": 7751
+    },
+    {
+      "epoch": 0.4709885169208336,
+      "grad_norm": 0.17072387039661407,
+      "learning_rate": 5.481242985002325e-05,
+      "loss": 1.1393,
+      "step": 7752
+    },
+    {
+      "epoch": 0.47104927395346013,
+      "grad_norm": 0.17193695902824402,
+      "learning_rate": 5.480290143210825e-05,
+      "loss": 1.1388,
+      "step": 7753
+    },
+    {
+      "epoch": 0.4711100309860866,
+      "grad_norm": 0.19930435717105865,
+      "learning_rate": 5.4793372838141974e-05,
+      "loss": 1.0579,
+      "step": 7754
+    },
+    {
+      "epoch": 0.47117078801871315,
+      "grad_norm": 0.20226283371448517,
+      "learning_rate": 5.47838440684737e-05,
+      "loss": 1.1862,
+      "step": 7755
+    },
+    {
+      "epoch": 0.4712315450513397,
+      "grad_norm": 0.22341449558734894,
+      "learning_rate": 5.4774315123452703e-05,
+      "loss": 1.19,
+      "step": 7756
+    },
+    {
+      "epoch": 0.4712923020839662,
+      "grad_norm": 0.14023500680923462,
+      "learning_rate": 5.476478600342825e-05,
+      "loss": 1.0576,
+      "step": 7757
+    },
+    {
+      "epoch": 0.47135305911659275,
+      "grad_norm": 0.17156977951526642,
+      "learning_rate": 5.4755256708749635e-05,
+      "loss": 1.0887,
+      "step": 7758
+    },
+    {
+      "epoch": 0.4714138161492193,
+      "grad_norm": 0.14423759281635284,
+      "learning_rate": 5.4745727239766195e-05,
+      "loss": 1.0629,
+      "step": 7759
+    },
+    {
+      "epoch": 0.4714745731818458,
+      "grad_norm": 0.12208879739046097,
+      "learning_rate": 5.4736197596827185e-05,
+      "loss": 1.0569,
+      "step": 7760
+    },
+    {
+      "epoch": 0.4715353302144723,
+      "grad_norm": 0.16867636144161224,
+      "learning_rate": 5.4726667780281935e-05,
+      "loss": 1.113,
+      "step": 7761
+    },
+    {
+      "epoch": 0.47159608724709884,
+      "grad_norm": 0.2090156078338623,
+      "learning_rate": 5.471713779047979e-05,
+      "loss": 1.07,
+      "step": 7762
+    },
+    {
+      "epoch": 0.4716568442797254,
+      "grad_norm": 0.5601214170455933,
+      "learning_rate": 5.470760762777001e-05,
+      "loss": 1.1139,
+      "step": 7763
+    },
+    {
+      "epoch": 0.4717176013123519,
+      "grad_norm": 0.14180058240890503,
+      "learning_rate": 5.469807729250198e-05,
+      "loss": 1.0703,
+      "step": 7764
+    },
+    {
+      "epoch": 0.47177835834497844,
+      "grad_norm": 0.20316222310066223,
+      "learning_rate": 5.468854678502503e-05,
+      "loss": 1.1128,
+      "step": 7765
+    },
+    {
+      "epoch": 0.471839115377605,
+      "grad_norm": 0.1542683094739914,
+      "learning_rate": 5.4679016105688475e-05,
+      "loss": 1.0391,
+      "step": 7766
+    },
+    {
+      "epoch": 0.47189987241023146,
+      "grad_norm": 0.20195795595645905,
+      "learning_rate": 5.466948525484169e-05,
+      "loss": 1.0544,
+      "step": 7767
+    },
+    {
+      "epoch": 0.471960629442858,
+      "grad_norm": 0.18277931213378906,
+      "learning_rate": 5.465995423283401e-05,
+      "loss": 1.1079,
+      "step": 7768
+    },
+    {
+      "epoch": 0.47202138647548453,
+      "grad_norm": 0.21681025624275208,
+      "learning_rate": 5.465042304001482e-05,
+      "loss": 1.1719,
+      "step": 7769
+    },
+    {
+      "epoch": 0.47208214350811106,
+      "grad_norm": 0.32704997062683105,
+      "learning_rate": 5.4640891676733466e-05,
+      "loss": 1.1155,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4721429005407376,
+      "grad_norm": 0.25209930539131165,
+      "learning_rate": 5.463136014333933e-05,
+      "loss": 1.1418,
+      "step": 7771
+    },
+    {
+      "epoch": 0.47220365757336413,
+      "grad_norm": 0.13654296100139618,
+      "learning_rate": 5.46218284401818e-05,
+      "loss": 1.0462,
+      "step": 7772
+    },
+    {
+      "epoch": 0.47226441460599067,
+      "grad_norm": 0.1953367441892624,
+      "learning_rate": 5.4612296567610244e-05,
+      "loss": 1.1316,
+      "step": 7773
+    },
+    {
+      "epoch": 0.47232517163861715,
+      "grad_norm": 6.810418128967285,
+      "learning_rate": 5.4602764525974074e-05,
+      "loss": 1.1277,
+      "step": 7774
+    },
+    {
+      "epoch": 0.4723859286712437,
+      "grad_norm": 0.20527994632720947,
+      "learning_rate": 5.459323231562269e-05,
+      "loss": 1.0892,
+      "step": 7775
+    },
+    {
+      "epoch": 0.4724466857038702,
+      "grad_norm": 0.2602483034133911,
+      "learning_rate": 5.458369993690546e-05,
+      "loss": 1.1021,
+      "step": 7776
+    },
+    {
+      "epoch": 0.47250744273649675,
+      "grad_norm": 0.19205524027347565,
+      "learning_rate": 5.4574167390171836e-05,
+      "loss": 1.1201,
+      "step": 7777
+    },
+    {
+      "epoch": 0.4725681997691233,
+      "grad_norm": 0.14140817523002625,
+      "learning_rate": 5.456463467577123e-05,
+      "loss": 1.0448,
+      "step": 7778
+    },
+    {
+      "epoch": 0.4726289568017498,
+      "grad_norm": 0.20182453095912933,
+      "learning_rate": 5.4555101794053054e-05,
+      "loss": 1.1289,
+      "step": 7779
+    },
+    {
+      "epoch": 0.4726897138343763,
+      "grad_norm": 0.1880880892276764,
+      "learning_rate": 5.4545568745366734e-05,
+      "loss": 1.0591,
+      "step": 7780
+    },
+    {
+      "epoch": 0.47275047086700284,
+      "grad_norm": 0.23056897521018982,
+      "learning_rate": 5.4536035530061716e-05,
+      "loss": 1.0786,
+      "step": 7781
+    },
+    {
+      "epoch": 0.47281122789962937,
+      "grad_norm": 0.1171664372086525,
+      "learning_rate": 5.452650214848746e-05,
+      "loss": 0.9748,
+      "step": 7782
+    },
+    {
+      "epoch": 0.4728719849322559,
+      "grad_norm": 0.2350465953350067,
+      "learning_rate": 5.451696860099337e-05,
+      "loss": 1.0683,
+      "step": 7783
+    },
+    {
+      "epoch": 0.47293274196488244,
+      "grad_norm": 0.1455424576997757,
+      "learning_rate": 5.450743488792892e-05,
+      "loss": 1.1746,
+      "step": 7784
+    },
+    {
+      "epoch": 0.472993498997509,
+      "grad_norm": 0.12110444903373718,
+      "learning_rate": 5.4497901009643594e-05,
+      "loss": 1.0514,
+      "step": 7785
+    },
+    {
+      "epoch": 0.4730542560301355,
+      "grad_norm": 0.8733172416687012,
+      "learning_rate": 5.4488366966486804e-05,
+      "loss": 1.1994,
+      "step": 7786
+    },
+    {
+      "epoch": 0.473115013062762,
+      "grad_norm": 0.16160254180431366,
+      "learning_rate": 5.447883275880809e-05,
+      "loss": 1.0621,
+      "step": 7787
+    },
+    {
+      "epoch": 0.4731757700953885,
+      "grad_norm": 0.18750369548797607,
+      "learning_rate": 5.446929838695687e-05,
+      "loss": 1.2648,
+      "step": 7788
+    },
+    {
+      "epoch": 0.47323652712801506,
+      "grad_norm": 0.13489563763141632,
+      "learning_rate": 5.445976385128266e-05,
+      "loss": 1.0345,
+      "step": 7789
+    },
+    {
+      "epoch": 0.4732972841606416,
+      "grad_norm": 0.23016774654388428,
+      "learning_rate": 5.445022915213496e-05,
+      "loss": 1.1304,
+      "step": 7790
+    },
+    {
+      "epoch": 0.47335804119326813,
+      "grad_norm": 0.15345828235149384,
+      "learning_rate": 5.444069428986324e-05,
+      "loss": 1.0549,
+      "step": 7791
+    },
+    {
+      "epoch": 0.47341879822589467,
+      "grad_norm": 0.25145938992500305,
+      "learning_rate": 5.4431159264817e-05,
+      "loss": 1.0497,
+      "step": 7792
+    },
+    {
+      "epoch": 0.4734795552585212,
+      "grad_norm": 0.19060321152210236,
+      "learning_rate": 5.442162407734578e-05,
+      "loss": 1.0812,
+      "step": 7793
+    },
+    {
+      "epoch": 0.4735403122911477,
+      "grad_norm": 0.2530083656311035,
+      "learning_rate": 5.441208872779905e-05,
+      "loss": 1.1228,
+      "step": 7794
+    },
+    {
+      "epoch": 0.4736010693237742,
+      "grad_norm": 0.12563717365264893,
+      "learning_rate": 5.440255321652637e-05,
+      "loss": 1.0195,
+      "step": 7795
+    },
+    {
+      "epoch": 0.47366182635640075,
+      "grad_norm": 0.19678513705730438,
+      "learning_rate": 5.4393017543877266e-05,
+      "loss": 1.1552,
+      "step": 7796
+    },
+    {
+      "epoch": 0.4737225833890273,
+      "grad_norm": 0.28398069739341736,
+      "learning_rate": 5.438348171020123e-05,
+      "loss": 1.091,
+      "step": 7797
+    },
+    {
+      "epoch": 0.4737833404216538,
+      "grad_norm": 0.20876304805278778,
+      "learning_rate": 5.437394571584784e-05,
+      "loss": 1.0837,
+      "step": 7798
+    },
+    {
+      "epoch": 0.47384409745428036,
+      "grad_norm": 0.45880165696144104,
+      "learning_rate": 5.436440956116662e-05,
+      "loss": 1.3407,
+      "step": 7799
+    },
+    {
+      "epoch": 0.47390485448690683,
+      "grad_norm": 0.23414337635040283,
+      "learning_rate": 5.435487324650712e-05,
+      "loss": 1.0597,
+      "step": 7800
+    },
+    {
+      "epoch": 0.47396561151953337,
+      "grad_norm": 0.1596781462430954,
+      "learning_rate": 5.43453367722189e-05,
+      "loss": 1.0568,
+      "step": 7801
+    },
+    {
+      "epoch": 0.4740263685521599,
+      "grad_norm": 0.19410710036754608,
+      "learning_rate": 5.433580013865153e-05,
+      "loss": 1.0466,
+      "step": 7802
+    },
+    {
+      "epoch": 0.47408712558478644,
+      "grad_norm": 0.2315683811903,
+      "learning_rate": 5.432626334615456e-05,
+      "loss": 1.0954,
+      "step": 7803
+    },
+    {
+      "epoch": 0.474147882617413,
+      "grad_norm": 0.18427301943302155,
+      "learning_rate": 5.431672639507758e-05,
+      "loss": 1.0862,
+      "step": 7804
+    },
+    {
+      "epoch": 0.4742086396500395,
+      "grad_norm": 0.15700361132621765,
+      "learning_rate": 5.430718928577013e-05,
+      "loss": 1.065,
+      "step": 7805
+    },
+    {
+      "epoch": 0.47426939668266604,
+      "grad_norm": 0.19870778918266296,
+      "learning_rate": 5.429765201858186e-05,
+      "loss": 1.0405,
+      "step": 7806
+    },
+    {
+      "epoch": 0.4743301537152925,
+      "grad_norm": 0.18470308184623718,
+      "learning_rate": 5.4288114593862305e-05,
+      "loss": 1.1483,
+      "step": 7807
+    },
+    {
+      "epoch": 0.47439091074791906,
+      "grad_norm": 0.24553771317005157,
+      "learning_rate": 5.4278577011961084e-05,
+      "loss": 1.0676,
+      "step": 7808
+    },
+    {
+      "epoch": 0.4744516677805456,
+      "grad_norm": 0.1649901121854782,
+      "learning_rate": 5.4269039273227796e-05,
+      "loss": 1.1483,
+      "step": 7809
+    },
+    {
+      "epoch": 0.47451242481317213,
+      "grad_norm": 0.5021433234214783,
+      "learning_rate": 5.4259501378012055e-05,
+      "loss": 1.0552,
+      "step": 7810
+    },
+    {
+      "epoch": 0.47457318184579866,
+      "grad_norm": 0.17254818975925446,
+      "learning_rate": 5.424996332666347e-05,
+      "loss": 1.1915,
+      "step": 7811
+    },
+    {
+      "epoch": 0.4746339388784252,
+      "grad_norm": 0.21085822582244873,
+      "learning_rate": 5.424042511953165e-05,
+      "loss": 1.0015,
+      "step": 7812
+    },
+    {
+      "epoch": 0.4746946959110517,
+      "grad_norm": 0.12940984964370728,
+      "learning_rate": 5.4230886756966224e-05,
+      "loss": 1.0798,
+      "step": 7813
+    },
+    {
+      "epoch": 0.4747554529436782,
+      "grad_norm": 0.15791188180446625,
+      "learning_rate": 5.4221348239316826e-05,
+      "loss": 1.0185,
+      "step": 7814
+    },
+    {
+      "epoch": 0.47481620997630475,
+      "grad_norm": 0.19145667552947998,
+      "learning_rate": 5.421180956693309e-05,
+      "loss": 1.0688,
+      "step": 7815
+    },
+    {
+      "epoch": 0.4748769670089313,
+      "grad_norm": 0.1844419240951538,
+      "learning_rate": 5.4202270740164676e-05,
+      "loss": 1.0806,
+      "step": 7816
+    },
+    {
+      "epoch": 0.4749377240415578,
+      "grad_norm": 0.1794002801179886,
+      "learning_rate": 5.4192731759361194e-05,
+      "loss": 1.1276,
+      "step": 7817
+    },
+    {
+      "epoch": 0.47499848107418435,
+      "grad_norm": 0.19025595486164093,
+      "learning_rate": 5.418319262487234e-05,
+      "loss": 1.1032,
+      "step": 7818
+    },
+    {
+      "epoch": 0.4750592381068109,
+      "grad_norm": 0.16529054939746857,
+      "learning_rate": 5.417365333704775e-05,
+      "loss": 1.0487,
+      "step": 7819
+    },
+    {
+      "epoch": 0.47511999513943737,
+      "grad_norm": 0.3196588456630707,
+      "learning_rate": 5.4164113896237065e-05,
+      "loss": 1.1987,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4751807521720639,
+      "grad_norm": 0.1594213843345642,
+      "learning_rate": 5.4154574302790006e-05,
+      "loss": 1.0485,
+      "step": 7821
+    },
+    {
+      "epoch": 0.47524150920469044,
+      "grad_norm": 0.23622652888298035,
+      "learning_rate": 5.4145034557056216e-05,
+      "loss": 1.0181,
+      "step": 7822
+    },
+    {
+      "epoch": 0.475302266237317,
+      "grad_norm": 0.14866940677165985,
+      "learning_rate": 5.4135494659385375e-05,
+      "loss": 1.0821,
+      "step": 7823
+    },
+    {
+      "epoch": 0.4753630232699435,
+      "grad_norm": 0.16754929721355438,
+      "learning_rate": 5.4125954610127194e-05,
+      "loss": 1.0843,
+      "step": 7824
+    },
+    {
+      "epoch": 0.47542378030257004,
+      "grad_norm": 0.15314020216464996,
+      "learning_rate": 5.4116414409631334e-05,
+      "loss": 1.1154,
+      "step": 7825
+    },
+    {
+      "epoch": 0.4754845373351965,
+      "grad_norm": 0.12049468606710434,
+      "learning_rate": 5.4106874058247526e-05,
+      "loss": 1.0753,
+      "step": 7826
+    },
+    {
+      "epoch": 0.47554529436782306,
+      "grad_norm": 0.16232872009277344,
+      "learning_rate": 5.409733355632544e-05,
+      "loss": 1.1097,
+      "step": 7827
+    },
+    {
+      "epoch": 0.4756060514004496,
+      "grad_norm": 0.13334311544895172,
+      "learning_rate": 5.40877929042148e-05,
+      "loss": 1.0695,
+      "step": 7828
+    },
+    {
+      "epoch": 0.4756668084330761,
+      "grad_norm": 0.23005901277065277,
+      "learning_rate": 5.407825210226534e-05,
+      "loss": 1.0816,
+      "step": 7829
+    },
+    {
+      "epoch": 0.47572756546570266,
+      "grad_norm": 1.7344071865081787,
+      "learning_rate": 5.406871115082672e-05,
+      "loss": 1.1305,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4757883224983292,
+      "grad_norm": 0.22466632723808289,
+      "learning_rate": 5.405917005024874e-05,
+      "loss": 1.1343,
+      "step": 7831
+    },
+    {
+      "epoch": 0.47584907953095573,
+      "grad_norm": 0.16500264406204224,
+      "learning_rate": 5.4049628800881105e-05,
+      "loss": 1.0241,
+      "step": 7832
+    },
+    {
+      "epoch": 0.4759098365635822,
+      "grad_norm": 0.11504673957824707,
+      "learning_rate": 5.404008740307351e-05,
+      "loss": 1.0506,
+      "step": 7833
+    },
+    {
+      "epoch": 0.47597059359620875,
+      "grad_norm": 0.2061557024717331,
+      "learning_rate": 5.4030545857175754e-05,
+      "loss": 1.091,
+      "step": 7834
+    },
+    {
+      "epoch": 0.4760313506288353,
+      "grad_norm": 0.14971303939819336,
+      "learning_rate": 5.402100416353756e-05,
+      "loss": 1.1017,
+      "step": 7835
+    },
+    {
+      "epoch": 0.4760921076614618,
+      "grad_norm": 0.13605548441410065,
+      "learning_rate": 5.4011462322508674e-05,
+      "loss": 1.0577,
+      "step": 7836
+    },
+    {
+      "epoch": 0.47615286469408835,
+      "grad_norm": 0.13173580169677734,
+      "learning_rate": 5.400192033443886e-05,
+      "loss": 1.0498,
+      "step": 7837
+    },
+    {
+      "epoch": 0.4762136217267149,
+      "grad_norm": 0.16944468021392822,
+      "learning_rate": 5.399237819967788e-05,
+      "loss": 1.0513,
+      "step": 7838
+    },
+    {
+      "epoch": 0.4762743787593414,
+      "grad_norm": 0.12402139604091644,
+      "learning_rate": 5.398283591857552e-05,
+      "loss": 1.0867,
+      "step": 7839
+    },
+    {
+      "epoch": 0.4763351357919679,
+      "grad_norm": 0.7062398791313171,
+      "learning_rate": 5.397329349148154e-05,
+      "loss": 1.1476,
+      "step": 7840
+    },
+    {
+      "epoch": 0.47639589282459444,
+      "grad_norm": 0.9580641388893127,
+      "learning_rate": 5.3963750918745694e-05,
+      "loss": 1.0157,
+      "step": 7841
+    },
+    {
+      "epoch": 0.47645664985722097,
+      "grad_norm": 0.24378855526447296,
+      "learning_rate": 5.3954208200717814e-05,
+      "loss": 1.2185,
+      "step": 7842
+    },
+    {
+      "epoch": 0.4765174068898475,
+      "grad_norm": 0.22034133970737457,
+      "learning_rate": 5.3944665337747655e-05,
+      "loss": 1.0968,
+      "step": 7843
+    },
+    {
+      "epoch": 0.47657816392247404,
+      "grad_norm": 0.2139805108308792,
+      "learning_rate": 5.393512233018504e-05,
+      "loss": 1.0286,
+      "step": 7844
+    },
+    {
+      "epoch": 0.4766389209551006,
+      "grad_norm": 0.14700251817703247,
+      "learning_rate": 5.3925579178379745e-05,
+      "loss": 1.1194,
+      "step": 7845
+    },
+    {
+      "epoch": 0.47669967798772706,
+      "grad_norm": 0.20244181156158447,
+      "learning_rate": 5.3916035882681595e-05,
+      "loss": 1.1879,
+      "step": 7846
+    },
+    {
+      "epoch": 0.4767604350203536,
+      "grad_norm": 0.12988512217998505,
+      "learning_rate": 5.3906492443440384e-05,
+      "loss": 1.111,
+      "step": 7847
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 0.12920653820037842,
+      "learning_rate": 5.3896948861005935e-05,
+      "loss": 1.0324,
+      "step": 7848
+    },
+    {
+      "epoch": 0.47688194908560666,
+      "grad_norm": 0.13387957215309143,
+      "learning_rate": 5.388740513572809e-05,
+      "loss": 1.0631,
+      "step": 7849
+    },
+    {
+      "epoch": 0.4769427061182332,
+      "grad_norm": 0.24146319925785065,
+      "learning_rate": 5.3877861267956645e-05,
+      "loss": 1.2698,
+      "step": 7850
+    },
+    {
+      "epoch": 0.47700346315085973,
+      "grad_norm": 0.12807698547840118,
+      "learning_rate": 5.386831725804143e-05,
+      "loss": 1.0056,
+      "step": 7851
+    },
+    {
+      "epoch": 0.47706422018348627,
+      "grad_norm": 0.33459898829460144,
+      "learning_rate": 5.385877310633233e-05,
+      "loss": 1.2747,
+      "step": 7852
+    },
+    {
+      "epoch": 0.47712497721611274,
+      "grad_norm": 0.16809044778347015,
+      "learning_rate": 5.3849228813179144e-05,
+      "loss": 1.1117,
+      "step": 7853
+    },
+    {
+      "epoch": 0.4771857342487393,
+      "grad_norm": 0.20247362554073334,
+      "learning_rate": 5.3839684378931724e-05,
+      "loss": 1.0077,
+      "step": 7854
+    },
+    {
+      "epoch": 0.4772464912813658,
+      "grad_norm": 7.45745849609375,
+      "learning_rate": 5.383013980393993e-05,
+      "loss": 1.0652,
+      "step": 7855
+    },
+    {
+      "epoch": 0.47730724831399235,
+      "grad_norm": 0.2583889067173004,
+      "learning_rate": 5.3820595088553626e-05,
+      "loss": 1.1445,
+      "step": 7856
+    },
+    {
+      "epoch": 0.4773680053466189,
+      "grad_norm": 0.15075862407684326,
+      "learning_rate": 5.381105023312267e-05,
+      "loss": 1.0751,
+      "step": 7857
+    },
+    {
+      "epoch": 0.4774287623792454,
+      "grad_norm": 0.5770914554595947,
+      "learning_rate": 5.380150523799693e-05,
+      "loss": 1.1802,
+      "step": 7858
+    },
+    {
+      "epoch": 0.4774895194118719,
+      "grad_norm": 0.3534625172615051,
+      "learning_rate": 5.3791960103526284e-05,
+      "loss": 1.1225,
+      "step": 7859
+    },
+    {
+      "epoch": 0.47755027644449843,
+      "grad_norm": 0.2707405686378479,
+      "learning_rate": 5.378241483006061e-05,
+      "loss": 1.3073,
+      "step": 7860
+    },
+    {
+      "epoch": 0.47761103347712497,
+      "grad_norm": 0.180998295545578,
+      "learning_rate": 5.377286941794979e-05,
+      "loss": 1.0496,
+      "step": 7861
+    },
+    {
+      "epoch": 0.4776717905097515,
+      "grad_norm": 0.2080940157175064,
+      "learning_rate": 5.376332386754369e-05,
+      "loss": 1.1915,
+      "step": 7862
+    },
+    {
+      "epoch": 0.47773254754237804,
+      "grad_norm": 0.17377610504627228,
+      "learning_rate": 5.375377817919226e-05,
+      "loss": 1.017,
+      "step": 7863
+    },
+    {
+      "epoch": 0.4777933045750046,
+      "grad_norm": 0.29675769805908203,
+      "learning_rate": 5.374423235324534e-05,
+      "loss": 1.15,
+      "step": 7864
+    },
+    {
+      "epoch": 0.4778540616076311,
+      "grad_norm": 0.24133987724781036,
+      "learning_rate": 5.3734686390052866e-05,
+      "loss": 1.1266,
+      "step": 7865
+    },
+    {
+      "epoch": 0.4779148186402576,
+      "grad_norm": 0.16240110993385315,
+      "learning_rate": 5.372514028996475e-05,
+      "loss": 1.0719,
+      "step": 7866
+    },
+    {
+      "epoch": 0.4779755756728841,
+      "grad_norm": 0.27527379989624023,
+      "learning_rate": 5.371559405333089e-05,
+      "loss": 1.0623,
+      "step": 7867
+    },
+    {
+      "epoch": 0.47803633270551066,
+      "grad_norm": 0.17591215670108795,
+      "learning_rate": 5.370604768050121e-05,
+      "loss": 1.1256,
+      "step": 7868
+    },
+    {
+      "epoch": 0.4780970897381372,
+      "grad_norm": 0.21628862619400024,
+      "learning_rate": 5.369650117182564e-05,
+      "loss": 1.0643,
+      "step": 7869
+    },
+    {
+      "epoch": 0.47815784677076373,
+      "grad_norm": 0.1805424839258194,
+      "learning_rate": 5.3686954527654096e-05,
+      "loss": 1.1736,
+      "step": 7870
+    },
+    {
+      "epoch": 0.47821860380339026,
+      "grad_norm": 2.398637533187866,
+      "learning_rate": 5.367740774833653e-05,
+      "loss": 1.1197,
+      "step": 7871
+    },
+    {
+      "epoch": 0.47827936083601674,
+      "grad_norm": 0.18604539334774017,
+      "learning_rate": 5.366786083422286e-05,
+      "loss": 1.1382,
+      "step": 7872
+    },
+    {
+      "epoch": 0.4783401178686433,
+      "grad_norm": 0.3984617590904236,
+      "learning_rate": 5.365831378566306e-05,
+      "loss": 1.3101,
+      "step": 7873
+    },
+    {
+      "epoch": 0.4784008749012698,
+      "grad_norm": 0.1727215200662613,
+      "learning_rate": 5.364876660300705e-05,
+      "loss": 1.0559,
+      "step": 7874
+    },
+    {
+      "epoch": 0.47846163193389635,
+      "grad_norm": 0.19871488213539124,
+      "learning_rate": 5.363921928660479e-05,
+      "loss": 1.0094,
+      "step": 7875
+    },
+    {
+      "epoch": 0.4785223889665229,
+      "grad_norm": 0.2542242407798767,
+      "learning_rate": 5.362967183680627e-05,
+      "loss": 1.1125,
+      "step": 7876
+    },
+    {
+      "epoch": 0.4785831459991494,
+      "grad_norm": 0.1417936384677887,
+      "learning_rate": 5.36201242539614e-05,
+      "loss": 1.0357,
+      "step": 7877
+    },
+    {
+      "epoch": 0.47864390303177595,
+      "grad_norm": 0.2642741799354553,
+      "learning_rate": 5.3610576538420185e-05,
+      "loss": 1.1574,
+      "step": 7878
+    },
+    {
+      "epoch": 0.47870466006440243,
+      "grad_norm": 0.21164394915103912,
+      "learning_rate": 5.3601028690532586e-05,
+      "loss": 1.1746,
+      "step": 7879
+    },
+    {
+      "epoch": 0.47876541709702897,
+      "grad_norm": 0.32209643721580505,
+      "learning_rate": 5.359148071064859e-05,
+      "loss": 1.0919,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4788261741296555,
+      "grad_norm": 0.2014092057943344,
+      "learning_rate": 5.358193259911817e-05,
+      "loss": 1.0956,
+      "step": 7881
+    },
+    {
+      "epoch": 0.47888693116228204,
+      "grad_norm": 0.1983584761619568,
+      "learning_rate": 5.35723843562913e-05,
+      "loss": 1.0658,
+      "step": 7882
+    },
+    {
+      "epoch": 0.47894768819490857,
+      "grad_norm": 0.2146424651145935,
+      "learning_rate": 5.356283598251803e-05,
+      "loss": 1.2607,
+      "step": 7883
+    },
+    {
+      "epoch": 0.4790084452275351,
+      "grad_norm": 0.2187325656414032,
+      "learning_rate": 5.3553287478148295e-05,
+      "loss": 1.012,
+      "step": 7884
+    },
+    {
+      "epoch": 0.4790692022601616,
+      "grad_norm": 0.202541783452034,
+      "learning_rate": 5.3543738843532124e-05,
+      "loss": 1.1371,
+      "step": 7885
+    },
+    {
+      "epoch": 0.4791299592927881,
+      "grad_norm": 0.17969128489494324,
+      "learning_rate": 5.353419007901953e-05,
+      "loss": 1.0692,
+      "step": 7886
+    },
+    {
+      "epoch": 0.47919071632541466,
+      "grad_norm": 0.4171610176563263,
+      "learning_rate": 5.352464118496049e-05,
+      "loss": 1.3031,
+      "step": 7887
+    },
+    {
+      "epoch": 0.4792514733580412,
+      "grad_norm": 0.7410226464271545,
+      "learning_rate": 5.351509216170507e-05,
+      "loss": 1.0556,
+      "step": 7888
+    },
+    {
+      "epoch": 0.4793122303906677,
+      "grad_norm": 0.1526513546705246,
+      "learning_rate": 5.350554300960326e-05,
+      "loss": 1.1034,
+      "step": 7889
+    },
+    {
+      "epoch": 0.47937298742329426,
+      "grad_norm": 0.17666850984096527,
+      "learning_rate": 5.349599372900509e-05,
+      "loss": 1.1811,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4794337444559208,
+      "grad_norm": 1.730309009552002,
+      "learning_rate": 5.34864443202606e-05,
+      "loss": 1.077,
+      "step": 7891
+    },
+    {
+      "epoch": 0.4794945014885473,
+      "grad_norm": 0.19253453612327576,
+      "learning_rate": 5.347689478371982e-05,
+      "loss": 1.0733,
+      "step": 7892
+    },
+    {
+      "epoch": 0.4795552585211738,
+      "grad_norm": 0.4330640137195587,
+      "learning_rate": 5.346734511973278e-05,
+      "loss": 1.1161,
+      "step": 7893
+    },
+    {
+      "epoch": 0.47961601555380035,
+      "grad_norm": 0.1686244010925293,
+      "learning_rate": 5.345779532864954e-05,
+      "loss": 1.0456,
+      "step": 7894
+    },
+    {
+      "epoch": 0.4796767725864269,
+      "grad_norm": 0.13274863362312317,
+      "learning_rate": 5.344824541082014e-05,
+      "loss": 1.0379,
+      "step": 7895
+    },
+    {
+      "epoch": 0.4797375296190534,
+      "grad_norm": 0.13108132779598236,
+      "learning_rate": 5.343869536659465e-05,
+      "loss": 1.0185,
+      "step": 7896
+    },
+    {
+      "epoch": 0.47979828665167995,
+      "grad_norm": 0.24932914972305298,
+      "learning_rate": 5.342914519632312e-05,
+      "loss": 1.0402,
+      "step": 7897
+    },
+    {
+      "epoch": 0.4798590436843065,
+      "grad_norm": 0.17208437621593475,
+      "learning_rate": 5.3419594900355585e-05,
+      "loss": 1.0514,
+      "step": 7898
+    },
+    {
+      "epoch": 0.47991980071693296,
+      "grad_norm": 0.1338982880115509,
+      "learning_rate": 5.3410044479042166e-05,
+      "loss": 1.0036,
+      "step": 7899
+    },
+    {
+      "epoch": 0.4799805577495595,
+      "grad_norm": 0.2246599793434143,
+      "learning_rate": 5.3400493932732873e-05,
+      "loss": 1.1542,
+      "step": 7900
+    },
+    {
+      "epoch": 0.48004131478218603,
+      "grad_norm": 0.2111661583185196,
+      "learning_rate": 5.339094326177785e-05,
+      "loss": 1.1145,
+      "step": 7901
+    },
+    {
+      "epoch": 0.48010207181481257,
+      "grad_norm": 0.2084105759859085,
+      "learning_rate": 5.3381392466527135e-05,
+      "loss": 1.0655,
+      "step": 7902
+    },
+    {
+      "epoch": 0.4801628288474391,
+      "grad_norm": 0.20062394440174103,
+      "learning_rate": 5.3371841547330825e-05,
+      "loss": 1.0884,
+      "step": 7903
+    },
+    {
+      "epoch": 0.48022358588006564,
+      "grad_norm": 0.3038314878940582,
+      "learning_rate": 5.3362290504539005e-05,
+      "loss": 1.0506,
+      "step": 7904
+    },
+    {
+      "epoch": 0.4802843429126921,
+      "grad_norm": 0.23090730607509613,
+      "learning_rate": 5.335273933850179e-05,
+      "loss": 1.0537,
+      "step": 7905
+    },
+    {
+      "epoch": 0.48034509994531865,
+      "grad_norm": 0.1430114209651947,
+      "learning_rate": 5.334318804956926e-05,
+      "loss": 1.0321,
+      "step": 7906
+    },
+    {
+      "epoch": 0.4804058569779452,
+      "grad_norm": 0.30413711071014404,
+      "learning_rate": 5.333363663809152e-05,
+      "loss": 1.0682,
+      "step": 7907
+    },
+    {
+      "epoch": 0.4804666140105717,
+      "grad_norm": 0.21721361577510834,
+      "learning_rate": 5.3324085104418683e-05,
+      "loss": 1.0291,
+      "step": 7908
+    },
+    {
+      "epoch": 0.48052737104319826,
+      "grad_norm": 0.1294948160648346,
+      "learning_rate": 5.331453344890088e-05,
+      "loss": 1.0913,
+      "step": 7909
+    },
+    {
+      "epoch": 0.4805881280758248,
+      "grad_norm": 0.2155667543411255,
+      "learning_rate": 5.33049816718882e-05,
+      "loss": 1.2055,
+      "step": 7910
+    },
+    {
+      "epoch": 0.48064888510845133,
+      "grad_norm": 0.3649980425834656,
+      "learning_rate": 5.3295429773730774e-05,
+      "loss": 1.2326,
+      "step": 7911
+    },
+    {
+      "epoch": 0.4807096421410778,
+      "grad_norm": 0.2715924084186554,
+      "learning_rate": 5.328587775477875e-05,
+      "loss": 1.1139,
+      "step": 7912
+    },
+    {
+      "epoch": 0.48077039917370434,
+      "grad_norm": 0.13731925189495087,
+      "learning_rate": 5.327632561538223e-05,
+      "loss": 1.0574,
+      "step": 7913
+    },
+    {
+      "epoch": 0.4808311562063309,
+      "grad_norm": 0.15858833491802216,
+      "learning_rate": 5.3266773355891365e-05,
+      "loss": 1.0083,
+      "step": 7914
+    },
+    {
+      "epoch": 0.4808919132389574,
+      "grad_norm": 0.2582811415195465,
+      "learning_rate": 5.325722097665629e-05,
+      "loss": 1.0731,
+      "step": 7915
+    },
+    {
+      "epoch": 0.48095267027158395,
+      "grad_norm": 0.24416813254356384,
+      "learning_rate": 5.3247668478027144e-05,
+      "loss": 1.043,
+      "step": 7916
+    },
+    {
+      "epoch": 0.4810134273042105,
+      "grad_norm": 0.2177872210741043,
+      "learning_rate": 5.32381158603541e-05,
+      "loss": 1.0251,
+      "step": 7917
+    },
+    {
+      "epoch": 0.48107418433683696,
+      "grad_norm": 0.23981410264968872,
+      "learning_rate": 5.3228563123987266e-05,
+      "loss": 1.1534,
+      "step": 7918
+    },
+    {
+      "epoch": 0.4811349413694635,
+      "grad_norm": 0.24031716585159302,
+      "learning_rate": 5.321901026927685e-05,
+      "loss": 1.0805,
+      "step": 7919
+    },
+    {
+      "epoch": 0.48119569840209003,
+      "grad_norm": 0.2082665115594864,
+      "learning_rate": 5.320945729657298e-05,
+      "loss": 1.1479,
+      "step": 7920
+    },
+    {
+      "epoch": 0.48125645543471657,
+      "grad_norm": 0.1649162769317627,
+      "learning_rate": 5.319990420622583e-05,
+      "loss": 1.0763,
+      "step": 7921
+    },
+    {
+      "epoch": 0.4813172124673431,
+      "grad_norm": 0.1867951601743698,
+      "learning_rate": 5.319035099858558e-05,
+      "loss": 1.2107,
+      "step": 7922
+    },
+    {
+      "epoch": 0.48137796949996964,
+      "grad_norm": 0.19225385785102844,
+      "learning_rate": 5.3180797674002394e-05,
+      "loss": 1.3577,
+      "step": 7923
+    },
+    {
+      "epoch": 0.4814387265325962,
+      "grad_norm": 0.29202041029930115,
+      "learning_rate": 5.317124423282646e-05,
+      "loss": 1.0761,
+      "step": 7924
+    },
+    {
+      "epoch": 0.48149948356522265,
+      "grad_norm": 0.1566685438156128,
+      "learning_rate": 5.316169067540795e-05,
+      "loss": 1.0715,
+      "step": 7925
+    },
+    {
+      "epoch": 0.4815602405978492,
+      "grad_norm": 0.1629745215177536,
+      "learning_rate": 5.315213700209706e-05,
+      "loss": 1.0888,
+      "step": 7926
+    },
+    {
+      "epoch": 0.4816209976304757,
+      "grad_norm": 0.14564397931098938,
+      "learning_rate": 5.314258321324398e-05,
+      "loss": 1.0502,
+      "step": 7927
+    },
+    {
+      "epoch": 0.48168175466310226,
+      "grad_norm": 0.15911243855953217,
+      "learning_rate": 5.31330293091989e-05,
+      "loss": 1.0768,
+      "step": 7928
+    },
+    {
+      "epoch": 0.4817425116957288,
+      "grad_norm": 0.18059316277503967,
+      "learning_rate": 5.312347529031203e-05,
+      "loss": 1.176,
+      "step": 7929
+    },
+    {
+      "epoch": 0.4818032687283553,
+      "grad_norm": 0.15501949191093445,
+      "learning_rate": 5.3113921156933566e-05,
+      "loss": 1.0504,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4818640257609818,
+      "grad_norm": 0.21993504464626312,
+      "learning_rate": 5.3104366909413705e-05,
+      "loss": 1.2641,
+      "step": 7931
+    },
+    {
+      "epoch": 0.48192478279360834,
+      "grad_norm": 0.20999452471733093,
+      "learning_rate": 5.309481254810269e-05,
+      "loss": 1.0181,
+      "step": 7932
+    },
+    {
+      "epoch": 0.4819855398262349,
+      "grad_norm": 0.24557897448539734,
+      "learning_rate": 5.308525807335074e-05,
+      "loss": 1.0617,
+      "step": 7933
+    },
+    {
+      "epoch": 0.4820462968588614,
+      "grad_norm": 0.16740214824676514,
+      "learning_rate": 5.3075703485508036e-05,
+      "loss": 1.0498,
+      "step": 7934
+    },
+    {
+      "epoch": 0.48210705389148795,
+      "grad_norm": 0.14449343085289001,
+      "learning_rate": 5.306614878492483e-05,
+      "loss": 1.0816,
+      "step": 7935
+    },
+    {
+      "epoch": 0.4821678109241145,
+      "grad_norm": 0.14462348818778992,
+      "learning_rate": 5.305659397195135e-05,
+      "loss": 1.0324,
+      "step": 7936
+    },
+    {
+      "epoch": 0.482228567956741,
+      "grad_norm": 0.24806468188762665,
+      "learning_rate": 5.304703904693783e-05,
+      "loss": 1.0069,
+      "step": 7937
+    },
+    {
+      "epoch": 0.4822893249893675,
+      "grad_norm": 0.3005404472351074,
+      "learning_rate": 5.303748401023451e-05,
+      "loss": 1.1536,
+      "step": 7938
+    },
+    {
+      "epoch": 0.48235008202199403,
+      "grad_norm": 0.1694445013999939,
+      "learning_rate": 5.30279288621916e-05,
+      "loss": 1.0587,
+      "step": 7939
+    },
+    {
+      "epoch": 0.48241083905462057,
+      "grad_norm": 0.49404528737068176,
+      "learning_rate": 5.30183736031594e-05,
+      "loss": 1.1646,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4824715960872471,
+      "grad_norm": 0.16141681373119354,
+      "learning_rate": 5.3008818233488124e-05,
+      "loss": 1.0195,
+      "step": 7941
+    },
+    {
+      "epoch": 0.48253235311987364,
+      "grad_norm": 0.14614440500736237,
+      "learning_rate": 5.299926275352802e-05,
+      "loss": 1.0414,
+      "step": 7942
+    },
+    {
+      "epoch": 0.48259311015250017,
+      "grad_norm": 0.3665587902069092,
+      "learning_rate": 5.298970716362939e-05,
+      "loss": 1.0628,
+      "step": 7943
+    },
+    {
+      "epoch": 0.4826538671851267,
+      "grad_norm": 0.19485582411289215,
+      "learning_rate": 5.2980151464142434e-05,
+      "loss": 1.1274,
+      "step": 7944
+    },
+    {
+      "epoch": 0.4827146242177532,
+      "grad_norm": 0.2236337959766388,
+      "learning_rate": 5.297059565541746e-05,
+      "loss": 1.1642,
+      "step": 7945
+    },
+    {
+      "epoch": 0.4827753812503797,
+      "grad_norm": 0.470614492893219,
+      "learning_rate": 5.296103973780475e-05,
+      "loss": 1.0924,
+      "step": 7946
+    },
+    {
+      "epoch": 0.48283613828300626,
+      "grad_norm": 0.7623069882392883,
+      "learning_rate": 5.2951483711654534e-05,
+      "loss": 1.0472,
+      "step": 7947
+    },
+    {
+      "epoch": 0.4828968953156328,
+      "grad_norm": 0.19647791981697083,
+      "learning_rate": 5.294192757731712e-05,
+      "loss": 1.1392,
+      "step": 7948
+    },
+    {
+      "epoch": 0.4829576523482593,
+      "grad_norm": 0.23711782693862915,
+      "learning_rate": 5.2932371335142784e-05,
+      "loss": 1.0503,
+      "step": 7949
+    },
+    {
+      "epoch": 0.48301840938088586,
+      "grad_norm": 0.2014499455690384,
+      "learning_rate": 5.29228149854818e-05,
+      "loss": 1.0746,
+      "step": 7950
+    },
+    {
+      "epoch": 0.48307916641351234,
+      "grad_norm": 0.2196071445941925,
+      "learning_rate": 5.2913258528684476e-05,
+      "loss": 1.1907,
+      "step": 7951
+    },
+    {
+      "epoch": 0.4831399234461389,
+      "grad_norm": 0.2745140492916107,
+      "learning_rate": 5.290370196510108e-05,
+      "loss": 1.2252,
+      "step": 7952
+    },
+    {
+      "epoch": 0.4832006804787654,
+      "grad_norm": 0.19129444658756256,
+      "learning_rate": 5.2894145295081954e-05,
+      "loss": 1.0874,
+      "step": 7953
+    },
+    {
+      "epoch": 0.48326143751139194,
+      "grad_norm": 0.25906902551651,
+      "learning_rate": 5.2884588518977364e-05,
+      "loss": 1.1451,
+      "step": 7954
+    },
+    {
+      "epoch": 0.4833221945440185,
+      "grad_norm": 0.16583764553070068,
+      "learning_rate": 5.287503163713761e-05,
+      "loss": 1.1273,
+      "step": 7955
+    },
+    {
+      "epoch": 0.483382951576645,
+      "grad_norm": 0.9115603566169739,
+      "learning_rate": 5.286547464991304e-05,
+      "loss": 1.0457,
+      "step": 7956
+    },
+    {
+      "epoch": 0.48344370860927155,
+      "grad_norm": 0.11355254054069519,
+      "learning_rate": 5.285591755765391e-05,
+      "loss": 1.0668,
+      "step": 7957
+    },
+    {
+      "epoch": 0.48350446564189803,
+      "grad_norm": 0.9551360607147217,
+      "learning_rate": 5.28463603607106e-05,
+      "loss": 1.094,
+      "step": 7958
+    },
+    {
+      "epoch": 0.48356522267452456,
+      "grad_norm": 0.18994790315628052,
+      "learning_rate": 5.2836803059433385e-05,
+      "loss": 1.0461,
+      "step": 7959
+    },
+    {
+      "epoch": 0.4836259797071511,
+      "grad_norm": 0.1270156353712082,
+      "learning_rate": 5.282724565417261e-05,
+      "loss": 1.0836,
+      "step": 7960
+    },
+    {
+      "epoch": 0.48368673673977763,
+      "grad_norm": 0.22180868685245514,
+      "learning_rate": 5.28176881452786e-05,
+      "loss": 1.1258,
+      "step": 7961
+    },
+    {
+      "epoch": 0.48374749377240417,
+      "grad_norm": 0.2829572260379791,
+      "learning_rate": 5.280813053310168e-05,
+      "loss": 1.2545,
+      "step": 7962
+    },
+    {
+      "epoch": 0.4838082508050307,
+      "grad_norm": 0.12600886821746826,
+      "learning_rate": 5.279857281799221e-05,
+      "loss": 1.052,
+      "step": 7963
+    },
+    {
+      "epoch": 0.4838690078376572,
+      "grad_norm": 0.2003079205751419,
+      "learning_rate": 5.27890150003005e-05,
+      "loss": 1.0105,
+      "step": 7964
+    },
+    {
+      "epoch": 0.4839297648702837,
+      "grad_norm": 0.9814311265945435,
+      "learning_rate": 5.2779457080376895e-05,
+      "loss": 1.1166,
+      "step": 7965
+    },
+    {
+      "epoch": 0.48399052190291025,
+      "grad_norm": 0.1546151489019394,
+      "learning_rate": 5.276989905857178e-05,
+      "loss": 1.0787,
+      "step": 7966
+    },
+    {
+      "epoch": 0.4840512789355368,
+      "grad_norm": 0.5525694489479065,
+      "learning_rate": 5.276034093523545e-05,
+      "loss": 1.0378,
+      "step": 7967
+    },
+    {
+      "epoch": 0.4841120359681633,
+      "grad_norm": 0.2193448692560196,
+      "learning_rate": 5.27507827107183e-05,
+      "loss": 1.1484,
+      "step": 7968
+    },
+    {
+      "epoch": 0.48417279300078986,
+      "grad_norm": 0.18941007554531097,
+      "learning_rate": 5.274122438537069e-05,
+      "loss": 1.0993,
+      "step": 7969
+    },
+    {
+      "epoch": 0.4842335500334164,
+      "grad_norm": 0.38067716360092163,
+      "learning_rate": 5.2731665959542964e-05,
+      "loss": 1.095,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4842943070660429,
+      "grad_norm": 0.1646842211484909,
+      "learning_rate": 5.272210743358549e-05,
+      "loss": 1.0901,
+      "step": 7971
+    },
+    {
+      "epoch": 0.4843550640986694,
+      "grad_norm": 0.1659906655550003,
+      "learning_rate": 5.271254880784865e-05,
+      "loss": 1.0534,
+      "step": 7972
+    },
+    {
+      "epoch": 0.48441582113129594,
+      "grad_norm": 0.16462163627147675,
+      "learning_rate": 5.27029900826828e-05,
+      "loss": 1.1142,
+      "step": 7973
+    },
+    {
+      "epoch": 0.4844765781639225,
+      "grad_norm": 0.12224703282117844,
+      "learning_rate": 5.269343125843833e-05,
+      "loss": 1.0569,
+      "step": 7974
+    },
+    {
+      "epoch": 0.484537335196549,
+      "grad_norm": 0.21103839576244354,
+      "learning_rate": 5.268387233546561e-05,
+      "loss": 1.0869,
+      "step": 7975
+    },
+    {
+      "epoch": 0.48459809222917555,
+      "grad_norm": 0.14903800189495087,
+      "learning_rate": 5.267431331411504e-05,
+      "loss": 1.04,
+      "step": 7976
+    },
+    {
+      "epoch": 0.484658849261802,
+      "grad_norm": 0.2757299244403839,
+      "learning_rate": 5.2664754194737007e-05,
+      "loss": 1.0386,
+      "step": 7977
+    },
+    {
+      "epoch": 0.48471960629442856,
+      "grad_norm": 0.11410175263881683,
+      "learning_rate": 5.2655194977681864e-05,
+      "loss": 1.0371,
+      "step": 7978
+    },
+    {
+      "epoch": 0.4847803633270551,
+      "grad_norm": 0.12369665503501892,
+      "learning_rate": 5.264563566330006e-05,
+      "loss": 1.0136,
+      "step": 7979
+    },
+    {
+      "epoch": 0.48484112035968163,
+      "grad_norm": 0.23373769223690033,
+      "learning_rate": 5.263607625194197e-05,
+      "loss": 1.0418,
+      "step": 7980
+    },
+    {
+      "epoch": 0.48490187739230817,
+      "grad_norm": 0.18790516257286072,
+      "learning_rate": 5.262651674395799e-05,
+      "loss": 1.1498,
+      "step": 7981
+    },
+    {
+      "epoch": 0.4849626344249347,
+      "grad_norm": 0.13051916658878326,
+      "learning_rate": 5.261695713969853e-05,
+      "loss": 1.0668,
+      "step": 7982
+    },
+    {
+      "epoch": 0.48502339145756124,
+      "grad_norm": 0.1773120015859604,
+      "learning_rate": 5.2607397439514e-05,
+      "loss": 1.1004,
+      "step": 7983
+    },
+    {
+      "epoch": 0.4850841484901877,
+      "grad_norm": 0.21541526913642883,
+      "learning_rate": 5.259783764375481e-05,
+      "loss": 1.2231,
+      "step": 7984
+    },
+    {
+      "epoch": 0.48514490552281425,
+      "grad_norm": 0.15957030653953552,
+      "learning_rate": 5.2588277752771383e-05,
+      "loss": 1.1902,
+      "step": 7985
+    },
+    {
+      "epoch": 0.4852056625554408,
+      "grad_norm": 0.12087242305278778,
+      "learning_rate": 5.257871776691413e-05,
+      "loss": 1.0611,
+      "step": 7986
+    },
+    {
+      "epoch": 0.4852664195880673,
+      "grad_norm": 0.1997787058353424,
+      "learning_rate": 5.2569157686533486e-05,
+      "loss": 1.1269,
+      "step": 7987
+    },
+    {
+      "epoch": 0.48532717662069386,
+      "grad_norm": 0.29204103350639343,
+      "learning_rate": 5.2559597511979865e-05,
+      "loss": 1.2454,
+      "step": 7988
+    },
+    {
+      "epoch": 0.4853879336533204,
+      "grad_norm": 0.12689681351184845,
+      "learning_rate": 5.25500372436037e-05,
+      "loss": 1.034,
+      "step": 7989
+    },
+    {
+      "epoch": 0.48544869068594687,
+      "grad_norm": 0.234115332365036,
+      "learning_rate": 5.254047688175544e-05,
+      "loss": 1.0758,
+      "step": 7990
+    },
+    {
+      "epoch": 0.4855094477185734,
+      "grad_norm": 0.20598378777503967,
+      "learning_rate": 5.253091642678549e-05,
+      "loss": 1.139,
+      "step": 7991
+    },
+    {
+      "epoch": 0.48557020475119994,
+      "grad_norm": 0.5636376142501831,
+      "learning_rate": 5.252135587904432e-05,
+      "loss": 1.1992,
+      "step": 7992
+    },
+    {
+      "epoch": 0.4856309617838265,
+      "grad_norm": 0.19981490075588226,
+      "learning_rate": 5.251179523888237e-05,
+      "loss": 1.13,
+      "step": 7993
+    },
+    {
+      "epoch": 0.485691718816453,
+      "grad_norm": 0.14565733075141907,
+      "learning_rate": 5.2502234506650074e-05,
+      "loss": 1.0073,
+      "step": 7994
+    },
+    {
+      "epoch": 0.48575247584907955,
+      "grad_norm": 0.2524283230304718,
+      "learning_rate": 5.249267368269788e-05,
+      "loss": 1.1495,
+      "step": 7995
+    },
+    {
+      "epoch": 0.4858132328817061,
+      "grad_norm": 0.2057437300682068,
+      "learning_rate": 5.2483112767376267e-05,
+      "loss": 1.0412,
+      "step": 7996
+    },
+    {
+      "epoch": 0.48587398991433256,
+      "grad_norm": 0.1583324372768402,
+      "learning_rate": 5.247355176103567e-05,
+      "loss": 1.275,
+      "step": 7997
+    },
+    {
+      "epoch": 0.4859347469469591,
+      "grad_norm": 0.14873246848583221,
+      "learning_rate": 5.246399066402654e-05,
+      "loss": 1.07,
+      "step": 7998
+    },
+    {
+      "epoch": 0.48599550397958563,
+      "grad_norm": 0.20360715687274933,
+      "learning_rate": 5.2454429476699365e-05,
+      "loss": 1.2321,
+      "step": 7999
+    },
+    {
+      "epoch": 0.48605626101221217,
+      "grad_norm": 0.1450817734003067,
+      "learning_rate": 5.2444868199404616e-05,
+      "loss": 1.0631,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4861170180448387,
+      "grad_norm": 0.28071364760398865,
+      "learning_rate": 5.2435306832492724e-05,
+      "loss": 1.1646,
+      "step": 8001
+    },
+    {
+      "epoch": 0.48617777507746524,
+      "grad_norm": 0.1492142379283905,
+      "learning_rate": 5.2425745376314206e-05,
+      "loss": 1.0385,
+      "step": 8002
+    },
+    {
+      "epoch": 0.48623853211009177,
+      "grad_norm": 0.1308092325925827,
+      "learning_rate": 5.2416183831219535e-05,
+      "loss": 1.0746,
+      "step": 8003
+    },
+    {
+      "epoch": 0.48629928914271825,
+      "grad_norm": 0.16332940757274628,
+      "learning_rate": 5.240662219755915e-05,
+      "loss": 1.1104,
+      "step": 8004
+    },
+    {
+      "epoch": 0.4863600461753448,
+      "grad_norm": 0.13730299472808838,
+      "learning_rate": 5.239706047568358e-05,
+      "loss": 1.1203,
+      "step": 8005
+    },
+    {
+      "epoch": 0.4864208032079713,
+      "grad_norm": 0.22228753566741943,
+      "learning_rate": 5.2387498665943294e-05,
+      "loss": 1.2661,
+      "step": 8006
+    },
+    {
+      "epoch": 0.48648156024059785,
+      "grad_norm": 0.20831681787967682,
+      "learning_rate": 5.237793676868879e-05,
+      "loss": 1.1509,
+      "step": 8007
+    },
+    {
+      "epoch": 0.4865423172732244,
+      "grad_norm": 1.1074758768081665,
+      "learning_rate": 5.2368374784270544e-05,
+      "loss": 1.145,
+      "step": 8008
+    },
+    {
+      "epoch": 0.4866030743058509,
+      "grad_norm": 0.18163339793682098,
+      "learning_rate": 5.2358812713039054e-05,
+      "loss": 1.1492,
+      "step": 8009
+    },
+    {
+      "epoch": 0.4866638313384774,
+      "grad_norm": 0.17313796281814575,
+      "learning_rate": 5.234925055534485e-05,
+      "loss": 1.0973,
+      "step": 8010
+    },
+    {
+      "epoch": 0.48672458837110394,
+      "grad_norm": 0.14672242105007172,
+      "learning_rate": 5.233968831153839e-05,
+      "loss": 1.0585,
+      "step": 8011
+    },
+    {
+      "epoch": 0.4867853454037305,
+      "grad_norm": 0.2614966332912445,
+      "learning_rate": 5.233012598197021e-05,
+      "loss": 1.0867,
+      "step": 8012
+    },
+    {
+      "epoch": 0.486846102436357,
+      "grad_norm": 0.16797687113285065,
+      "learning_rate": 5.232056356699083e-05,
+      "loss": 1.0453,
+      "step": 8013
+    },
+    {
+      "epoch": 0.48690685946898354,
+      "grad_norm": 0.1362597644329071,
+      "learning_rate": 5.2311001066950715e-05,
+      "loss": 1.0838,
+      "step": 8014
+    },
+    {
+      "epoch": 0.4869676165016101,
+      "grad_norm": 0.1593841165304184,
+      "learning_rate": 5.2301438482200425e-05,
+      "loss": 1.0238,
+      "step": 8015
+    },
+    {
+      "epoch": 0.4870283735342366,
+      "grad_norm": 0.13067477941513062,
+      "learning_rate": 5.2291875813090465e-05,
+      "loss": 1.0082,
+      "step": 8016
+    },
+    {
+      "epoch": 0.4870891305668631,
+      "grad_norm": 0.2266821265220642,
+      "learning_rate": 5.228231305997136e-05,
+      "loss": 1.1389,
+      "step": 8017
+    },
+    {
+      "epoch": 0.48714988759948963,
+      "grad_norm": 0.11394955962896347,
+      "learning_rate": 5.227275022319361e-05,
+      "loss": 1.0428,
+      "step": 8018
+    },
+    {
+      "epoch": 0.48721064463211616,
+      "grad_norm": 0.16107165813446045,
+      "learning_rate": 5.226318730310776e-05,
+      "loss": 1.06,
+      "step": 8019
+    },
+    {
+      "epoch": 0.4872714016647427,
+      "grad_norm": 0.2283700853586197,
+      "learning_rate": 5.2253624300064365e-05,
+      "loss": 1.0908,
+      "step": 8020
+    },
+    {
+      "epoch": 0.48733215869736923,
+      "grad_norm": 0.1669009029865265,
+      "learning_rate": 5.224406121441392e-05,
+      "loss": 1.118,
+      "step": 8021
+    },
+    {
+      "epoch": 0.48739291572999577,
+      "grad_norm": 0.36684319376945496,
+      "learning_rate": 5.223449804650697e-05,
+      "loss": 1.006,
+      "step": 8022
+    },
+    {
+      "epoch": 0.48745367276262225,
+      "grad_norm": 0.15023773908615112,
+      "learning_rate": 5.222493479669409e-05,
+      "loss": 1.1483,
+      "step": 8023
+    },
+    {
+      "epoch": 0.4875144297952488,
+      "grad_norm": 0.6638458371162415,
+      "learning_rate": 5.221537146532577e-05,
+      "loss": 1.1516,
+      "step": 8024
+    },
+    {
+      "epoch": 0.4875751868278753,
+      "grad_norm": 0.2511345148086548,
+      "learning_rate": 5.220580805275258e-05,
+      "loss": 1.0713,
+      "step": 8025
+    },
+    {
+      "epoch": 0.48763594386050185,
+      "grad_norm": 0.14232602715492249,
+      "learning_rate": 5.219624455932508e-05,
+      "loss": 1.0901,
+      "step": 8026
+    },
+    {
+      "epoch": 0.4876967008931284,
+      "grad_norm": 0.205320805311203,
+      "learning_rate": 5.218668098539381e-05,
+      "loss": 1.2123,
+      "step": 8027
+    },
+    {
+      "epoch": 0.4877574579257549,
+      "grad_norm": 0.17856544256210327,
+      "learning_rate": 5.217711733130932e-05,
+      "loss": 1.0312,
+      "step": 8028
+    },
+    {
+      "epoch": 0.48781821495838146,
+      "grad_norm": 0.5738333463668823,
+      "learning_rate": 5.216755359742218e-05,
+      "loss": 1.0911,
+      "step": 8029
+    },
+    {
+      "epoch": 0.48787897199100794,
+      "grad_norm": 0.17153413593769073,
+      "learning_rate": 5.215798978408295e-05,
+      "loss": 1.0853,
+      "step": 8030
+    },
+    {
+      "epoch": 0.48793972902363447,
+      "grad_norm": 0.16926462948322296,
+      "learning_rate": 5.214842589164218e-05,
+      "loss": 1.0512,
+      "step": 8031
+    },
+    {
+      "epoch": 0.488000486056261,
+      "grad_norm": 0.14639419317245483,
+      "learning_rate": 5.213886192045043e-05,
+      "loss": 1.0356,
+      "step": 8032
+    },
+    {
+      "epoch": 0.48806124308888754,
+      "grad_norm": 0.12281551212072372,
+      "learning_rate": 5.2129297870858304e-05,
+      "loss": 1.0345,
+      "step": 8033
+    },
+    {
+      "epoch": 0.4881220001215141,
+      "grad_norm": 0.3668050467967987,
+      "learning_rate": 5.2119733743216346e-05,
+      "loss": 1.1403,
+      "step": 8034
+    },
+    {
+      "epoch": 0.4881827571541406,
+      "grad_norm": 0.18001694977283478,
+      "learning_rate": 5.2110169537875134e-05,
+      "loss": 1.2287,
+      "step": 8035
+    },
+    {
+      "epoch": 0.4882435141867671,
+      "grad_norm": 0.12598273158073425,
+      "learning_rate": 5.210060525518525e-05,
+      "loss": 1.0753,
+      "step": 8036
+    },
+    {
+      "epoch": 0.4883042712193936,
+      "grad_norm": 4.260048866271973,
+      "learning_rate": 5.209104089549728e-05,
+      "loss": 1.1117,
+      "step": 8037
+    },
+    {
+      "epoch": 0.48836502825202016,
+      "grad_norm": 0.20910169184207916,
+      "learning_rate": 5.2081476459161796e-05,
+      "loss": 1.0553,
+      "step": 8038
+    },
+    {
+      "epoch": 0.4884257852846467,
+      "grad_norm": 0.34444794058799744,
+      "learning_rate": 5.20719119465294e-05,
+      "loss": 1.1254,
+      "step": 8039
+    },
+    {
+      "epoch": 0.48848654231727323,
+      "grad_norm": 0.21555787324905396,
+      "learning_rate": 5.2062347357950655e-05,
+      "loss": 1.2044,
+      "step": 8040
+    },
+    {
+      "epoch": 0.48854729934989977,
+      "grad_norm": 0.38426005840301514,
+      "learning_rate": 5.205278269377618e-05,
+      "loss": 1.0867,
+      "step": 8041
+    },
+    {
+      "epoch": 0.4886080563825263,
+      "grad_norm": 0.17334817349910736,
+      "learning_rate": 5.204321795435656e-05,
+      "loss": 1.1405,
+      "step": 8042
+    },
+    {
+      "epoch": 0.4886688134151528,
+      "grad_norm": 0.1316339522600174,
+      "learning_rate": 5.2033653140042384e-05,
+      "loss": 1.0939,
+      "step": 8043
+    },
+    {
+      "epoch": 0.4887295704477793,
+      "grad_norm": 0.3123212158679962,
+      "learning_rate": 5.202408825118427e-05,
+      "loss": 1.1619,
+      "step": 8044
+    },
+    {
+      "epoch": 0.48879032748040585,
+      "grad_norm": 0.22630399465560913,
+      "learning_rate": 5.201452328813279e-05,
+      "loss": 1.046,
+      "step": 8045
+    },
+    {
+      "epoch": 0.4888510845130324,
+      "grad_norm": 0.15358251333236694,
+      "learning_rate": 5.200495825123858e-05,
+      "loss": 1.0724,
+      "step": 8046
+    },
+    {
+      "epoch": 0.4889118415456589,
+      "grad_norm": 0.14148572087287903,
+      "learning_rate": 5.199539314085224e-05,
+      "loss": 1.0066,
+      "step": 8047
+    },
+    {
+      "epoch": 0.48897259857828546,
+      "grad_norm": 0.20466063916683197,
+      "learning_rate": 5.1985827957324376e-05,
+      "loss": 1.1558,
+      "step": 8048
+    },
+    {
+      "epoch": 0.489033355610912,
+      "grad_norm": 0.5848605036735535,
+      "learning_rate": 5.197626270100561e-05,
+      "loss": 1.1226,
+      "step": 8049
+    },
+    {
+      "epoch": 0.48909411264353847,
+      "grad_norm": 0.17252375185489655,
+      "learning_rate": 5.196669737224654e-05,
+      "loss": 1.1444,
+      "step": 8050
+    },
+    {
+      "epoch": 0.489154869676165,
+      "grad_norm": 0.1706843376159668,
+      "learning_rate": 5.195713197139781e-05,
+      "loss": 1.1242,
+      "step": 8051
+    },
+    {
+      "epoch": 0.48921562670879154,
+      "grad_norm": 3.653329372406006,
+      "learning_rate": 5.194756649881003e-05,
+      "loss": 1.1025,
+      "step": 8052
+    },
+    {
+      "epoch": 0.4892763837414181,
+      "grad_norm": 0.368230402469635,
+      "learning_rate": 5.1938000954833824e-05,
+      "loss": 1.227,
+      "step": 8053
+    },
+    {
+      "epoch": 0.4893371407740446,
+      "grad_norm": 0.21818876266479492,
+      "learning_rate": 5.192843533981981e-05,
+      "loss": 1.1089,
+      "step": 8054
+    },
+    {
+      "epoch": 0.48939789780667114,
+      "grad_norm": 0.16938568651676178,
+      "learning_rate": 5.191886965411863e-05,
+      "loss": 1.0654,
+      "step": 8055
+    },
+    {
+      "epoch": 0.4894586548392976,
+      "grad_norm": 0.4583264887332916,
+      "learning_rate": 5.190930389808091e-05,
+      "loss": 1.2066,
+      "step": 8056
+    },
+    {
+      "epoch": 0.48951941187192416,
+      "grad_norm": 0.26365500688552856,
+      "learning_rate": 5.1899738072057294e-05,
+      "loss": 1.097,
+      "step": 8057
+    },
+    {
+      "epoch": 0.4895801689045507,
+      "grad_norm": 0.7178860306739807,
+      "learning_rate": 5.18901721763984e-05,
+      "loss": 1.0628,
+      "step": 8058
+    },
+    {
+      "epoch": 0.48964092593717723,
+      "grad_norm": 0.2661334276199341,
+      "learning_rate": 5.1880606211454886e-05,
+      "loss": 1.0743,
+      "step": 8059
+    },
+    {
+      "epoch": 0.48970168296980376,
+      "grad_norm": 0.2064826935529709,
+      "learning_rate": 5.187104017757738e-05,
+      "loss": 1.0727,
+      "step": 8060
+    },
+    {
+      "epoch": 0.4897624400024303,
+      "grad_norm": 0.1711578071117401,
+      "learning_rate": 5.186147407511654e-05,
+      "loss": 1.0394,
+      "step": 8061
+    },
+    {
+      "epoch": 0.48982319703505683,
+      "grad_norm": 0.25302138924598694,
+      "learning_rate": 5.185190790442301e-05,
+      "loss": 1.0506,
+      "step": 8062
+    },
+    {
+      "epoch": 0.4898839540676833,
+      "grad_norm": 0.14483009278774261,
+      "learning_rate": 5.184234166584745e-05,
+      "loss": 1.0901,
+      "step": 8063
+    },
+    {
+      "epoch": 0.48994471110030985,
+      "grad_norm": 0.2093842625617981,
+      "learning_rate": 5.1832775359740484e-05,
+      "loss": 1.0458,
+      "step": 8064
+    },
+    {
+      "epoch": 0.4900054681329364,
+      "grad_norm": 0.14583586156368256,
+      "learning_rate": 5.1823208986452786e-05,
+      "loss": 0.9877,
+      "step": 8065
+    },
+    {
+      "epoch": 0.4900662251655629,
+      "grad_norm": 0.14035290479660034,
+      "learning_rate": 5.181364254633501e-05,
+      "loss": 1.0456,
+      "step": 8066
+    },
+    {
+      "epoch": 0.49012698219818945,
+      "grad_norm": 0.4639998972415924,
+      "learning_rate": 5.180407603973783e-05,
+      "loss": 1.0946,
+      "step": 8067
+    },
+    {
+      "epoch": 0.490187739230816,
+      "grad_norm": 0.1791357845067978,
+      "learning_rate": 5.1794509467011864e-05,
+      "loss": 1.0854,
+      "step": 8068
+    },
+    {
+      "epoch": 0.49024849626344247,
+      "grad_norm": 0.15634910762310028,
+      "learning_rate": 5.178494282850782e-05,
+      "loss": 1.0984,
+      "step": 8069
+    },
+    {
+      "epoch": 0.490309253296069,
+      "grad_norm": 1.635263442993164,
+      "learning_rate": 5.177537612457637e-05,
+      "loss": 1.2299,
+      "step": 8070
+    },
+    {
+      "epoch": 0.49037001032869554,
+      "grad_norm": 0.13857577741146088,
+      "learning_rate": 5.176580935556814e-05,
+      "loss": 1.0622,
+      "step": 8071
+    },
+    {
+      "epoch": 0.4904307673613221,
+      "grad_norm": 0.12869401276111603,
+      "learning_rate": 5.175624252183383e-05,
+      "loss": 1.0787,
+      "step": 8072
+    },
+    {
+      "epoch": 0.4904915243939486,
+      "grad_norm": 0.15388746559619904,
+      "learning_rate": 5.174667562372413e-05,
+      "loss": 1.0931,
+      "step": 8073
+    },
+    {
+      "epoch": 0.49055228142657514,
+      "grad_norm": 0.17299987375736237,
+      "learning_rate": 5.1737108661589685e-05,
+      "loss": 1.1027,
+      "step": 8074
+    },
+    {
+      "epoch": 0.4906130384592017,
+      "grad_norm": 0.1280059665441513,
+      "learning_rate": 5.172754163578118e-05,
+      "loss": 1.0775,
+      "step": 8075
+    },
+    {
+      "epoch": 0.49067379549182816,
+      "grad_norm": 0.14672118425369263,
+      "learning_rate": 5.17179745466493e-05,
+      "loss": 1.1298,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4907345525244547,
+      "grad_norm": 0.25295767188072205,
+      "learning_rate": 5.1708407394544755e-05,
+      "loss": 1.1434,
+      "step": 8077
+    },
+    {
+      "epoch": 0.4907953095570812,
+      "grad_norm": 0.16360440850257874,
+      "learning_rate": 5.1698840179818186e-05,
+      "loss": 1.0919,
+      "step": 8078
+    },
+    {
+      "epoch": 0.49085606658970776,
+      "grad_norm": 0.36694246530532837,
+      "learning_rate": 5.1689272902820296e-05,
+      "loss": 1.0782,
+      "step": 8079
+    },
+    {
+      "epoch": 0.4909168236223343,
+      "grad_norm": 0.2163536101579666,
+      "learning_rate": 5.16797055639018e-05,
+      "loss": 1.2255,
+      "step": 8080
+    },
+    {
+      "epoch": 0.49097758065496083,
+      "grad_norm": 0.2372448742389679,
+      "learning_rate": 5.167013816341336e-05,
+      "loss": 1.0121,
+      "step": 8081
+    },
+    {
+      "epoch": 0.4910383376875873,
+      "grad_norm": 0.23705677688121796,
+      "learning_rate": 5.166057070170569e-05,
+      "loss": 1.0877,
+      "step": 8082
+    },
+    {
+      "epoch": 0.49109909472021385,
+      "grad_norm": 0.1316308081150055,
+      "learning_rate": 5.1651003179129476e-05,
+      "loss": 1.0561,
+      "step": 8083
+    },
+    {
+      "epoch": 0.4911598517528404,
+      "grad_norm": 0.2815787196159363,
+      "learning_rate": 5.164143559603543e-05,
+      "loss": 1.0579,
+      "step": 8084
+    },
+    {
+      "epoch": 0.4912206087854669,
+      "grad_norm": 0.20056988298892975,
+      "learning_rate": 5.163186795277423e-05,
+      "loss": 1.1034,
+      "step": 8085
+    },
+    {
+      "epoch": 0.49128136581809345,
+      "grad_norm": 0.22818082571029663,
+      "learning_rate": 5.162230024969661e-05,
+      "loss": 1.0575,
+      "step": 8086
+    },
+    {
+      "epoch": 0.49134212285072,
+      "grad_norm": 0.28945690393447876,
+      "learning_rate": 5.161273248715325e-05,
+      "loss": 1.099,
+      "step": 8087
+    },
+    {
+      "epoch": 0.4914028798833465,
+      "grad_norm": 0.2143135666847229,
+      "learning_rate": 5.1603164665494877e-05,
+      "loss": 1.1362,
+      "step": 8088
+    },
+    {
+      "epoch": 0.491463636915973,
+      "grad_norm": 0.1453467160463333,
+      "learning_rate": 5.159359678507217e-05,
+      "loss": 1.0261,
+      "step": 8089
+    },
+    {
+      "epoch": 0.49152439394859954,
+      "grad_norm": 0.15251703560352325,
+      "learning_rate": 5.15840288462359e-05,
+      "loss": 1.0341,
+      "step": 8090
+    },
+    {
+      "epoch": 0.49158515098122607,
+      "grad_norm": 0.20740902423858643,
+      "learning_rate": 5.1574460849336724e-05,
+      "loss": 1.0742,
+      "step": 8091
+    },
+    {
+      "epoch": 0.4916459080138526,
+      "grad_norm": 0.48251283168792725,
+      "learning_rate": 5.156489279472537e-05,
+      "loss": 1.0298,
+      "step": 8092
+    },
+    {
+      "epoch": 0.49170666504647914,
+      "grad_norm": 0.31824442744255066,
+      "learning_rate": 5.155532468275259e-05,
+      "loss": 1.1727,
+      "step": 8093
+    },
+    {
+      "epoch": 0.4917674220791057,
+      "grad_norm": 0.14958001673221588,
+      "learning_rate": 5.1545756513769084e-05,
+      "loss": 1.0608,
+      "step": 8094
+    },
+    {
+      "epoch": 0.49182817911173216,
+      "grad_norm": 0.16572943329811096,
+      "learning_rate": 5.1536188288125565e-05,
+      "loss": 1.0245,
+      "step": 8095
+    },
+    {
+      "epoch": 0.4918889361443587,
+      "grad_norm": 0.20816749334335327,
+      "learning_rate": 5.152662000617277e-05,
+      "loss": 1.1165,
+      "step": 8096
+    },
+    {
+      "epoch": 0.4919496931769852,
+      "grad_norm": 4.0026726722717285,
+      "learning_rate": 5.151705166826143e-05,
+      "loss": 1.1173,
+      "step": 8097
+    },
+    {
+      "epoch": 0.49201045020961176,
+      "grad_norm": 1.2023611068725586,
+      "learning_rate": 5.1507483274742266e-05,
+      "loss": 1.0872,
+      "step": 8098
+    },
+    {
+      "epoch": 0.4920712072422383,
+      "grad_norm": 0.23394490778446198,
+      "learning_rate": 5.1497914825965996e-05,
+      "loss": 1.1161,
+      "step": 8099
+    },
+    {
+      "epoch": 0.49213196427486483,
+      "grad_norm": 0.23241440951824188,
+      "learning_rate": 5.1488346322283384e-05,
+      "loss": 1.1167,
+      "step": 8100
+    },
+    {
+      "epoch": 0.49219272130749137,
+      "grad_norm": 0.182684987783432,
+      "learning_rate": 5.147877776404515e-05,
+      "loss": 1.0474,
+      "step": 8101
+    },
+    {
+      "epoch": 0.49225347834011784,
+      "grad_norm": 0.185987189412117,
+      "learning_rate": 5.1469209151602014e-05,
+      "loss": 1.1285,
+      "step": 8102
+    },
+    {
+      "epoch": 0.4923142353727444,
+      "grad_norm": 0.18138177692890167,
+      "learning_rate": 5.145964048530475e-05,
+      "loss": 1.0419,
+      "step": 8103
+    },
+    {
+      "epoch": 0.4923749924053709,
+      "grad_norm": 0.1576085239648819,
+      "learning_rate": 5.145007176550408e-05,
+      "loss": 1.0747,
+      "step": 8104
+    },
+    {
+      "epoch": 0.49243574943799745,
+      "grad_norm": 0.4329274594783783,
+      "learning_rate": 5.144050299255074e-05,
+      "loss": 1.3283,
+      "step": 8105
+    },
+    {
+      "epoch": 0.492496506470624,
+      "grad_norm": 0.12247934192419052,
+      "learning_rate": 5.143093416679549e-05,
+      "loss": 1.0534,
+      "step": 8106
+    },
+    {
+      "epoch": 0.4925572635032505,
+      "grad_norm": 0.19341446459293365,
+      "learning_rate": 5.142136528858907e-05,
+      "loss": 1.095,
+      "step": 8107
+    },
+    {
+      "epoch": 0.49261802053587705,
+      "grad_norm": 0.3289068043231964,
+      "learning_rate": 5.1411796358282235e-05,
+      "loss": 1.1127,
+      "step": 8108
+    },
+    {
+      "epoch": 0.49267877756850353,
+      "grad_norm": 0.11702846735715866,
+      "learning_rate": 5.140222737622573e-05,
+      "loss": 1.0267,
+      "step": 8109
+    },
+    {
+      "epoch": 0.49273953460113007,
+      "grad_norm": 0.43668898940086365,
+      "learning_rate": 5.139265834277032e-05,
+      "loss": 1.1228,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4928002916337566,
+      "grad_norm": 0.12550920248031616,
+      "learning_rate": 5.1383089258266725e-05,
+      "loss": 1.0298,
+      "step": 8111
+    },
+    {
+      "epoch": 0.49286104866638314,
+      "grad_norm": 0.13258221745491028,
+      "learning_rate": 5.1373520123065735e-05,
+      "loss": 1.0591,
+      "step": 8112
+    },
+    {
+      "epoch": 0.4929218056990097,
+      "grad_norm": 0.13505661487579346,
+      "learning_rate": 5.1363950937518104e-05,
+      "loss": 1.0526,
+      "step": 8113
+    },
+    {
+      "epoch": 0.4929825627316362,
+      "grad_norm": 0.1507914811372757,
+      "learning_rate": 5.1354381701974597e-05,
+      "loss": 1.0821,
+      "step": 8114
+    },
+    {
+      "epoch": 0.4930433197642627,
+      "grad_norm": 0.1936265230178833,
+      "learning_rate": 5.1344812416785946e-05,
+      "loss": 1.0794,
+      "step": 8115
+    },
+    {
+      "epoch": 0.4931040767968892,
+      "grad_norm": 0.23144285380840302,
+      "learning_rate": 5.133524308230295e-05,
+      "loss": 1.0622,
+      "step": 8116
+    },
+    {
+      "epoch": 0.49316483382951576,
+      "grad_norm": 0.20834168791770935,
+      "learning_rate": 5.1325673698876356e-05,
+      "loss": 1.1006,
+      "step": 8117
+    },
+    {
+      "epoch": 0.4932255908621423,
+      "grad_norm": 0.1457124948501587,
+      "learning_rate": 5.1316104266856936e-05,
+      "loss": 1.1049,
+      "step": 8118
+    },
+    {
+      "epoch": 0.49328634789476883,
+      "grad_norm": 0.13936744630336761,
+      "learning_rate": 5.130653478659546e-05,
+      "loss": 1.091,
+      "step": 8119
+    },
+    {
+      "epoch": 0.49334710492739536,
+      "grad_norm": 0.12499111145734787,
+      "learning_rate": 5.1296965258442706e-05,
+      "loss": 1.0565,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4934078619600219,
+      "grad_norm": 0.12328989058732986,
+      "learning_rate": 5.128739568274944e-05,
+      "loss": 1.1194,
+      "step": 8121
+    },
+    {
+      "epoch": 0.4934686189926484,
+      "grad_norm": 0.11281615495681763,
+      "learning_rate": 5.127782605986643e-05,
+      "loss": 1.0856,
+      "step": 8122
+    },
+    {
+      "epoch": 0.4935293760252749,
+      "grad_norm": 0.19786526262760162,
+      "learning_rate": 5.1268256390144444e-05,
+      "loss": 1.1847,
+      "step": 8123
+    },
+    {
+      "epoch": 0.49359013305790145,
+      "grad_norm": 0.2402956634759903,
+      "learning_rate": 5.125868667393431e-05,
+      "loss": 1.1225,
+      "step": 8124
+    },
+    {
+      "epoch": 0.493650890090528,
+      "grad_norm": 0.21540434658527374,
+      "learning_rate": 5.124911691158674e-05,
+      "loss": 1.1308,
+      "step": 8125
+    },
+    {
+      "epoch": 0.4937116471231545,
+      "grad_norm": 0.12889930605888367,
+      "learning_rate": 5.123954710345256e-05,
+      "loss": 1.0247,
+      "step": 8126
+    },
+    {
+      "epoch": 0.49377240415578105,
+      "grad_norm": 0.20945194363594055,
+      "learning_rate": 5.122997724988255e-05,
+      "loss": 1.0621,
+      "step": 8127
+    },
+    {
+      "epoch": 0.49383316118840753,
+      "grad_norm": 0.22293387353420258,
+      "learning_rate": 5.122040735122747e-05,
+      "loss": 1.2814,
+      "step": 8128
+    },
+    {
+      "epoch": 0.49389391822103407,
+      "grad_norm": 0.31326350569725037,
+      "learning_rate": 5.121083740783813e-05,
+      "loss": 1.2136,
+      "step": 8129
+    },
+    {
+      "epoch": 0.4939546752536606,
+      "grad_norm": 0.1952691674232483,
+      "learning_rate": 5.120126742006531e-05,
+      "loss": 1.109,
+      "step": 8130
+    },
+    {
+      "epoch": 0.49401543228628714,
+      "grad_norm": 0.1779486984014511,
+      "learning_rate": 5.119169738825981e-05,
+      "loss": 1.1342,
+      "step": 8131
+    },
+    {
+      "epoch": 0.49407618931891367,
+      "grad_norm": 0.19886529445648193,
+      "learning_rate": 5.11821273127724e-05,
+      "loss": 1.1375,
+      "step": 8132
+    },
+    {
+      "epoch": 0.4941369463515402,
+      "grad_norm": 0.1656958907842636,
+      "learning_rate": 5.117255719395388e-05,
+      "loss": 1.0374,
+      "step": 8133
+    },
+    {
+      "epoch": 0.49419770338416674,
+      "grad_norm": 0.36190176010131836,
+      "learning_rate": 5.1162987032155074e-05,
+      "loss": 1.1571,
+      "step": 8134
+    },
+    {
+      "epoch": 0.4942584604167932,
+      "grad_norm": 0.27930328249931335,
+      "learning_rate": 5.115341682772674e-05,
+      "loss": 1.0269,
+      "step": 8135
+    },
+    {
+      "epoch": 0.49431921744941976,
+      "grad_norm": 0.2375454604625702,
+      "learning_rate": 5.114384658101967e-05,
+      "loss": 1.0736,
+      "step": 8136
+    },
+    {
+      "epoch": 0.4943799744820463,
+      "grad_norm": 0.2661468982696533,
+      "learning_rate": 5.113427629238472e-05,
+      "loss": 1.1042,
+      "step": 8137
+    },
+    {
+      "epoch": 0.4944407315146728,
+      "grad_norm": 0.18297788500785828,
+      "learning_rate": 5.112470596217261e-05,
+      "loss": 1.122,
+      "step": 8138
+    },
+    {
+      "epoch": 0.49450148854729936,
+      "grad_norm": 0.12859390676021576,
+      "learning_rate": 5.111513559073422e-05,
+      "loss": 1.0424,
+      "step": 8139
+    },
+    {
+      "epoch": 0.4945622455799259,
+      "grad_norm": 0.20417475700378418,
+      "learning_rate": 5.110556517842031e-05,
+      "loss": 1.1031,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4946230026125524,
+      "grad_norm": 0.4640609323978424,
+      "learning_rate": 5.109599472558169e-05,
+      "loss": 1.1813,
+      "step": 8141
+    },
+    {
+      "epoch": 0.4946837596451789,
+      "grad_norm": 0.3602086901664734,
+      "learning_rate": 5.108642423256918e-05,
+      "loss": 1.2167,
+      "step": 8142
+    },
+    {
+      "epoch": 0.49474451667780545,
+      "grad_norm": 0.1720881164073944,
+      "learning_rate": 5.1076853699733576e-05,
+      "loss": 1.134,
+      "step": 8143
+    },
+    {
+      "epoch": 0.494805273710432,
+      "grad_norm": 0.16503708064556122,
+      "learning_rate": 5.106728312742568e-05,
+      "loss": 1.1311,
+      "step": 8144
+    },
+    {
+      "epoch": 0.4948660307430585,
+      "grad_norm": 1.6363108158111572,
+      "learning_rate": 5.105771251599633e-05,
+      "loss": 1.1103,
+      "step": 8145
+    },
+    {
+      "epoch": 0.49492678777568505,
+      "grad_norm": 0.1599103808403015,
+      "learning_rate": 5.104814186579631e-05,
+      "loss": 1.0433,
+      "step": 8146
+    },
+    {
+      "epoch": 0.4949875448083116,
+      "grad_norm": 0.135761097073555,
+      "learning_rate": 5.103857117717647e-05,
+      "loss": 1.0645,
+      "step": 8147
+    },
+    {
+      "epoch": 0.49504830184093807,
+      "grad_norm": 2.425630807876587,
+      "learning_rate": 5.102900045048758e-05,
+      "loss": 1.0743,
+      "step": 8148
+    },
+    {
+      "epoch": 0.4951090588735646,
+      "grad_norm": 0.20664967596530914,
+      "learning_rate": 5.101942968608048e-05,
+      "loss": 1.086,
+      "step": 8149
+    },
+    {
+      "epoch": 0.49516981590619114,
+      "grad_norm": 0.20991453528404236,
+      "learning_rate": 5.1009858884305995e-05,
+      "loss": 1.1199,
+      "step": 8150
+    },
+    {
+      "epoch": 0.49523057293881767,
+      "grad_norm": 0.21037231385707855,
+      "learning_rate": 5.100028804551493e-05,
+      "loss": 1.0924,
+      "step": 8151
+    },
+    {
+      "epoch": 0.4952913299714442,
+      "grad_norm": 0.2317184954881668,
+      "learning_rate": 5.099071717005812e-05,
+      "loss": 1.1425,
+      "step": 8152
+    },
+    {
+      "epoch": 0.49535208700407074,
+      "grad_norm": 0.20401878654956818,
+      "learning_rate": 5.0981146258286385e-05,
+      "loss": 1.0787,
+      "step": 8153
+    },
+    {
+      "epoch": 0.4954128440366973,
+      "grad_norm": 0.47571122646331787,
+      "learning_rate": 5.0971575310550546e-05,
+      "loss": 1.1633,
+      "step": 8154
+    },
+    {
+      "epoch": 0.49547360106932375,
+      "grad_norm": 0.18967245519161224,
+      "learning_rate": 5.0962004327201416e-05,
+      "loss": 1.0949,
+      "step": 8155
+    },
+    {
+      "epoch": 0.4955343581019503,
+      "grad_norm": 2.0704503059387207,
+      "learning_rate": 5.0952433308589844e-05,
+      "loss": 1.1116,
+      "step": 8156
+    },
+    {
+      "epoch": 0.4955951151345768,
+      "grad_norm": 0.2872239351272583,
+      "learning_rate": 5.094286225506664e-05,
+      "loss": 1.0119,
+      "step": 8157
+    },
+    {
+      "epoch": 0.49565587216720336,
+      "grad_norm": 0.14626392722129822,
+      "learning_rate": 5.093329116698263e-05,
+      "loss": 1.0656,
+      "step": 8158
+    },
+    {
+      "epoch": 0.4957166291998299,
+      "grad_norm": 0.22473354637622833,
+      "learning_rate": 5.092372004468865e-05,
+      "loss": 1.0813,
+      "step": 8159
+    },
+    {
+      "epoch": 0.49577738623245643,
+      "grad_norm": 0.25407880544662476,
+      "learning_rate": 5.091414888853554e-05,
+      "loss": 1.0723,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4958381432650829,
+      "grad_norm": 0.14181669056415558,
+      "learning_rate": 5.090457769887412e-05,
+      "loss": 1.044,
+      "step": 8161
+    },
+    {
+      "epoch": 0.49589890029770944,
+      "grad_norm": 0.29317861795425415,
+      "learning_rate": 5.089500647605524e-05,
+      "loss": 1.1828,
+      "step": 8162
+    },
+    {
+      "epoch": 0.495959657330336,
+      "grad_norm": 0.1615181714296341,
+      "learning_rate": 5.088543522042972e-05,
+      "loss": 1.0371,
+      "step": 8163
+    },
+    {
+      "epoch": 0.4960204143629625,
+      "grad_norm": 0.16157956421375275,
+      "learning_rate": 5.08758639323484e-05,
+      "loss": 1.0826,
+      "step": 8164
+    },
+    {
+      "epoch": 0.49608117139558905,
+      "grad_norm": 0.14580108225345612,
+      "learning_rate": 5.086629261216211e-05,
+      "loss": 1.0761,
+      "step": 8165
+    },
+    {
+      "epoch": 0.4961419284282156,
+      "grad_norm": 0.14426328241825104,
+      "learning_rate": 5.0856721260221716e-05,
+      "loss": 1.0669,
+      "step": 8166
+    },
+    {
+      "epoch": 0.4962026854608421,
+      "grad_norm": 0.1830640435218811,
+      "learning_rate": 5.084714987687803e-05,
+      "loss": 1.0907,
+      "step": 8167
+    },
+    {
+      "epoch": 0.4962634424934686,
+      "grad_norm": 0.23111943900585175,
+      "learning_rate": 5.08375784624819e-05,
+      "loss": 1.144,
+      "step": 8168
+    },
+    {
+      "epoch": 0.49632419952609513,
+      "grad_norm": 0.26133355498313904,
+      "learning_rate": 5.082800701738416e-05,
+      "loss": 1.0952,
+      "step": 8169
+    },
+    {
+      "epoch": 0.49638495655872167,
+      "grad_norm": 0.20605742931365967,
+      "learning_rate": 5.0818435541935675e-05,
+      "loss": 1.12,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4964457135913482,
+      "grad_norm": 0.1918751299381256,
+      "learning_rate": 5.080886403648728e-05,
+      "loss": 1.142,
+      "step": 8171
+    },
+    {
+      "epoch": 0.49650647062397474,
+      "grad_norm": 1.250995397567749,
+      "learning_rate": 5.079929250138981e-05,
+      "loss": 1.0772,
+      "step": 8172
+    },
+    {
+      "epoch": 0.4965672276566013,
+      "grad_norm": 0.2142866849899292,
+      "learning_rate": 5.0789720936994115e-05,
+      "loss": 1.0732,
+      "step": 8173
+    },
+    {
+      "epoch": 0.49662798468922775,
+      "grad_norm": 0.29276618361473083,
+      "learning_rate": 5.078014934365105e-05,
+      "loss": 1.0331,
+      "step": 8174
+    },
+    {
+      "epoch": 0.4966887417218543,
+      "grad_norm": 0.2521336078643799,
+      "learning_rate": 5.077057772171146e-05,
+      "loss": 1.1482,
+      "step": 8175
+    },
+    {
+      "epoch": 0.4967494987544808,
+      "grad_norm": 0.22795985639095306,
+      "learning_rate": 5.076100607152621e-05,
+      "loss": 1.031,
+      "step": 8176
+    },
+    {
+      "epoch": 0.49681025578710736,
+      "grad_norm": 0.26999640464782715,
+      "learning_rate": 5.0751434393446116e-05,
+      "loss": 1.01,
+      "step": 8177
+    },
+    {
+      "epoch": 0.4968710128197339,
+      "grad_norm": 0.1529611051082611,
+      "learning_rate": 5.074186268782205e-05,
+      "loss": 1.0345,
+      "step": 8178
+    },
+    {
+      "epoch": 0.4969317698523604,
+      "grad_norm": 0.21202926337718964,
+      "learning_rate": 5.073229095500488e-05,
+      "loss": 1.1232,
+      "step": 8179
+    },
+    {
+      "epoch": 0.49699252688498696,
+      "grad_norm": 0.1875039041042328,
+      "learning_rate": 5.0722719195345415e-05,
+      "loss": 1.1163,
+      "step": 8180
+    },
+    {
+      "epoch": 0.49705328391761344,
+      "grad_norm": 0.19611190259456635,
+      "learning_rate": 5.071314740919457e-05,
+      "loss": 1.1717,
+      "step": 8181
+    },
+    {
+      "epoch": 0.49711404095024,
+      "grad_norm": 0.21463800966739655,
+      "learning_rate": 5.070357559690314e-05,
+      "loss": 1.0346,
+      "step": 8182
+    },
+    {
+      "epoch": 0.4971747979828665,
+      "grad_norm": 0.21034328639507294,
+      "learning_rate": 5.069400375882203e-05,
+      "loss": 1.1268,
+      "step": 8183
+    },
+    {
+      "epoch": 0.49723555501549305,
+      "grad_norm": 0.18212296068668365,
+      "learning_rate": 5.0684431895302086e-05,
+      "loss": 1.064,
+      "step": 8184
+    },
+    {
+      "epoch": 0.4972963120481196,
+      "grad_norm": 0.2803708612918854,
+      "learning_rate": 5.067486000669414e-05,
+      "loss": 1.155,
+      "step": 8185
+    },
+    {
+      "epoch": 0.4973570690807461,
+      "grad_norm": 0.2102777659893036,
+      "learning_rate": 5.066528809334907e-05,
+      "loss": 1.1148,
+      "step": 8186
+    },
+    {
+      "epoch": 0.4974178261133726,
+      "grad_norm": 0.17017783224582672,
+      "learning_rate": 5.065571615561774e-05,
+      "loss": 1.0931,
+      "step": 8187
+    },
+    {
+      "epoch": 0.49747858314599913,
+      "grad_norm": 0.2427312582731247,
+      "learning_rate": 5.064614419385101e-05,
+      "loss": 1.2314,
+      "step": 8188
+    },
+    {
+      "epoch": 0.49753934017862567,
+      "grad_norm": 3.5309224128723145,
+      "learning_rate": 5.0636572208399743e-05,
+      "loss": 1.0487,
+      "step": 8189
+    },
+    {
+      "epoch": 0.4976000972112522,
+      "grad_norm": 0.747995138168335,
+      "learning_rate": 5.062700019961478e-05,
+      "loss": 1.1525,
+      "step": 8190
+    },
+    {
+      "epoch": 0.49766085424387874,
+      "grad_norm": 0.23654013872146606,
+      "learning_rate": 5.061742816784704e-05,
+      "loss": 1.1189,
+      "step": 8191
+    },
+    {
+      "epoch": 0.49772161127650527,
+      "grad_norm": 0.1926470249891281,
+      "learning_rate": 5.0607856113447324e-05,
+      "loss": 1.1093,
+      "step": 8192
+    },
+    {
+      "epoch": 0.4977823683091318,
+      "grad_norm": 0.2659144997596741,
+      "learning_rate": 5.059828403676652e-05,
+      "loss": 1.0888,
+      "step": 8193
+    },
+    {
+      "epoch": 0.4978431253417583,
+      "grad_norm": 0.17904962599277496,
+      "learning_rate": 5.058871193815552e-05,
+      "loss": 1.0981,
+      "step": 8194
+    },
+    {
+      "epoch": 0.4979038823743848,
+      "grad_norm": 0.16388067603111267,
+      "learning_rate": 5.0579139817965146e-05,
+      "loss": 1.0259,
+      "step": 8195
+    },
+    {
+      "epoch": 0.49796463940701136,
+      "grad_norm": 0.8273473381996155,
+      "learning_rate": 5.0569567676546304e-05,
+      "loss": 1.0974,
+      "step": 8196
+    },
+    {
+      "epoch": 0.4980253964396379,
+      "grad_norm": 0.9225310683250427,
+      "learning_rate": 5.055999551424986e-05,
+      "loss": 1.0881,
+      "step": 8197
+    },
+    {
+      "epoch": 0.4980861534722644,
+      "grad_norm": 0.20532894134521484,
+      "learning_rate": 5.055042333142666e-05,
+      "loss": 1.0351,
+      "step": 8198
+    },
+    {
+      "epoch": 0.49814691050489096,
+      "grad_norm": 0.13360542058944702,
+      "learning_rate": 5.054085112842758e-05,
+      "loss": 1.0417,
+      "step": 8199
+    },
+    {
+      "epoch": 0.49820766753751744,
+      "grad_norm": 0.2290326952934265,
+      "learning_rate": 5.053127890560351e-05,
+      "loss": 1.1253,
+      "step": 8200
+    },
+    {
+      "epoch": 0.498268424570144,
+      "grad_norm": 0.17761391401290894,
+      "learning_rate": 5.05217066633053e-05,
+      "loss": 1.3029,
+      "step": 8201
+    },
+    {
+      "epoch": 0.4983291816027705,
+      "grad_norm": 0.17378677427768707,
+      "learning_rate": 5.051213440188383e-05,
+      "loss": 1.1049,
+      "step": 8202
+    },
+    {
+      "epoch": 0.49838993863539705,
+      "grad_norm": 0.153955340385437,
+      "learning_rate": 5.050256212168997e-05,
+      "loss": 1.0613,
+      "step": 8203
+    },
+    {
+      "epoch": 0.4984506956680236,
+      "grad_norm": 0.2369835525751114,
+      "learning_rate": 5.049298982307462e-05,
+      "loss": 1.1353,
+      "step": 8204
+    },
+    {
+      "epoch": 0.4985114527006501,
+      "grad_norm": 0.3454437851905823,
+      "learning_rate": 5.048341750638862e-05,
+      "loss": 1.0689,
+      "step": 8205
+    },
+    {
+      "epoch": 0.49857220973327665,
+      "grad_norm": 0.35425442457199097,
+      "learning_rate": 5.0473845171982844e-05,
+      "loss": 1.043,
+      "step": 8206
+    },
+    {
+      "epoch": 0.49863296676590313,
+      "grad_norm": 0.17250806093215942,
+      "learning_rate": 5.0464272820208214e-05,
+      "loss": 1.0134,
+      "step": 8207
+    },
+    {
+      "epoch": 0.49869372379852966,
+      "grad_norm": 0.9371998906135559,
+      "learning_rate": 5.045470045141554e-05,
+      "loss": 1.0978,
+      "step": 8208
+    },
+    {
+      "epoch": 0.4987544808311562,
+      "grad_norm": 0.12351718544960022,
+      "learning_rate": 5.044512806595575e-05,
+      "loss": 1.0731,
+      "step": 8209
+    },
+    {
+      "epoch": 0.49881523786378273,
+      "grad_norm": 0.19571208953857422,
+      "learning_rate": 5.04355556641797e-05,
+      "loss": 1.1871,
+      "step": 8210
+    },
+    {
+      "epoch": 0.49887599489640927,
+      "grad_norm": 0.13838797807693481,
+      "learning_rate": 5.042598324643828e-05,
+      "loss": 1.078,
+      "step": 8211
+    },
+    {
+      "epoch": 0.4989367519290358,
+      "grad_norm": 0.3813706636428833,
+      "learning_rate": 5.041641081308236e-05,
+      "loss": 1.0104,
+      "step": 8212
+    },
+    {
+      "epoch": 0.49899750896166234,
+      "grad_norm": 0.14353570342063904,
+      "learning_rate": 5.0406838364462804e-05,
+      "loss": 1.0368,
+      "step": 8213
+    },
+    {
+      "epoch": 0.4990582659942888,
+      "grad_norm": 0.21878044307231903,
+      "learning_rate": 5.039726590093053e-05,
+      "loss": 1.2278,
+      "step": 8214
+    },
+    {
+      "epoch": 0.49911902302691535,
+      "grad_norm": 1.1359318494796753,
+      "learning_rate": 5.03876934228364e-05,
+      "loss": 1.0577,
+      "step": 8215
+    },
+    {
+      "epoch": 0.4991797800595419,
+      "grad_norm": 0.15227779746055603,
+      "learning_rate": 5.037812093053127e-05,
+      "loss": 1.0808,
+      "step": 8216
+    },
+    {
+      "epoch": 0.4992405370921684,
+      "grad_norm": 0.21359845995903015,
+      "learning_rate": 5.036854842436608e-05,
+      "loss": 1.1212,
+      "step": 8217
+    },
+    {
+      "epoch": 0.49930129412479496,
+      "grad_norm": 0.1925724446773529,
+      "learning_rate": 5.035897590469164e-05,
+      "loss": 1.1241,
+      "step": 8218
+    },
+    {
+      "epoch": 0.4993620511574215,
+      "grad_norm": 0.20035238564014435,
+      "learning_rate": 5.0349403371858886e-05,
+      "loss": 1.0131,
+      "step": 8219
+    },
+    {
+      "epoch": 0.499422808190048,
+      "grad_norm": 0.18301323056221008,
+      "learning_rate": 5.033983082621868e-05,
+      "loss": 1.1172,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4994835652226745,
+      "grad_norm": 0.17572705447673798,
+      "learning_rate": 5.033025826812191e-05,
+      "loss": 1.0583,
+      "step": 8221
+    },
+    {
+      "epoch": 0.49954432225530104,
+      "grad_norm": 0.19760817289352417,
+      "learning_rate": 5.0320685697919465e-05,
+      "loss": 1.1087,
+      "step": 8222
+    },
+    {
+      "epoch": 0.4996050792879276,
+      "grad_norm": 0.26191139221191406,
+      "learning_rate": 5.0311113115962215e-05,
+      "loss": 1.0983,
+      "step": 8223
+    },
+    {
+      "epoch": 0.4996658363205541,
+      "grad_norm": 0.30545851588249207,
+      "learning_rate": 5.030154052260106e-05,
+      "loss": 1.1286,
+      "step": 8224
+    },
+    {
+      "epoch": 0.49972659335318065,
+      "grad_norm": 0.24228700995445251,
+      "learning_rate": 5.0291967918186875e-05,
+      "loss": 1.1084,
+      "step": 8225
+    },
+    {
+      "epoch": 0.4997873503858072,
+      "grad_norm": 0.16840411722660065,
+      "learning_rate": 5.028239530307054e-05,
+      "loss": 1.0668,
+      "step": 8226
+    },
+    {
+      "epoch": 0.49984810741843366,
+      "grad_norm": 0.16302527487277985,
+      "learning_rate": 5.027282267760296e-05,
+      "loss": 0.9963,
+      "step": 8227
+    },
+    {
+      "epoch": 0.4999088644510602,
+      "grad_norm": 0.14595066010951996,
+      "learning_rate": 5.026325004213504e-05,
+      "loss": 1.0778,
+      "step": 8228
+    },
+    {
+      "epoch": 0.49996962148368673,
+      "grad_norm": 0.5399507284164429,
+      "learning_rate": 5.02536773970176e-05,
+      "loss": 1.0661,
+      "step": 8229
+    },
+    {
+      "epoch": 0.5000303785163133,
+      "grad_norm": 0.14924471080303192,
+      "learning_rate": 5.024410474260157e-05,
+      "loss": 1.0671,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5000911355489398,
+      "grad_norm": 0.8931715488433838,
+      "learning_rate": 5.023453207923784e-05,
+      "loss": 1.4328,
+      "step": 8231
+    },
+    {
+      "epoch": 0.5001518925815663,
+      "grad_norm": 0.16851367056369781,
+      "learning_rate": 5.0224959407277296e-05,
+      "loss": 1.0402,
+      "step": 8232
+    },
+    {
+      "epoch": 0.5002126496141929,
+      "grad_norm": 0.13297845423221588,
+      "learning_rate": 5.0215386727070824e-05,
+      "loss": 1.0906,
+      "step": 8233
+    },
+    {
+      "epoch": 0.5002734066468194,
+      "grad_norm": 0.42528483271598816,
+      "learning_rate": 5.02058140389693e-05,
+      "loss": 1.078,
+      "step": 8234
+    },
+    {
+      "epoch": 0.5003341636794459,
+      "grad_norm": 0.1264801323413849,
+      "learning_rate": 5.019624134332363e-05,
+      "loss": 1.048,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5003949207120725,
+      "grad_norm": 0.2873716950416565,
+      "learning_rate": 5.018666864048469e-05,
+      "loss": 1.1872,
+      "step": 8236
+    },
+    {
+      "epoch": 0.5004556777446989,
+      "grad_norm": 0.16146931052207947,
+      "learning_rate": 5.017709593080338e-05,
+      "loss": 1.0365,
+      "step": 8237
+    },
+    {
+      "epoch": 0.5005164347773254,
+      "grad_norm": 0.35578781366348267,
+      "learning_rate": 5.0167523214630586e-05,
+      "loss": 1.0896,
+      "step": 8238
+    },
+    {
+      "epoch": 0.500577191809952,
+      "grad_norm": 0.16928507387638092,
+      "learning_rate": 5.015795049231717e-05,
+      "loss": 1.047,
+      "step": 8239
+    },
+    {
+      "epoch": 0.5006379488425785,
+      "grad_norm": 0.3534541130065918,
+      "learning_rate": 5.0148377764214084e-05,
+      "loss": 1.0634,
+      "step": 8240
+    },
+    {
+      "epoch": 0.500698705875205,
+      "grad_norm": 0.1801421344280243,
+      "learning_rate": 5.0138805030672176e-05,
+      "loss": 1.0528,
+      "step": 8241
+    },
+    {
+      "epoch": 0.5007594629078316,
+      "grad_norm": 0.32417336106300354,
+      "learning_rate": 5.0129232292042325e-05,
+      "loss": 1.0255,
+      "step": 8242
+    },
+    {
+      "epoch": 0.5008202199404581,
+      "grad_norm": 0.1729975789785385,
+      "learning_rate": 5.0119659548675446e-05,
+      "loss": 1.0761,
+      "step": 8243
+    },
+    {
+      "epoch": 0.5008809769730846,
+      "grad_norm": 0.20362064242362976,
+      "learning_rate": 5.011008680092243e-05,
+      "loss": 1.102,
+      "step": 8244
+    },
+    {
+      "epoch": 0.5009417340057112,
+      "grad_norm": 0.42174428701400757,
+      "learning_rate": 5.010051404913416e-05,
+      "loss": 1.0797,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5010024910383377,
+      "grad_norm": 0.13610652089118958,
+      "learning_rate": 5.009094129366153e-05,
+      "loss": 1.1002,
+      "step": 8246
+    },
+    {
+      "epoch": 0.5010632480709643,
+      "grad_norm": 0.14378514885902405,
+      "learning_rate": 5.008136853485541e-05,
+      "loss": 1.0406,
+      "step": 8247
+    },
+    {
+      "epoch": 0.5011240051035908,
+      "grad_norm": 0.14423348009586334,
+      "learning_rate": 5.007179577306674e-05,
+      "loss": 1.0981,
+      "step": 8248
+    },
+    {
+      "epoch": 0.5011847621362173,
+      "grad_norm": 0.22621090710163116,
+      "learning_rate": 5.006222300864637e-05,
+      "loss": 1.075,
+      "step": 8249
+    },
+    {
+      "epoch": 0.5012455191688437,
+      "grad_norm": 0.20006893575191498,
+      "learning_rate": 5.005265024194519e-05,
+      "loss": 1.1468,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5013062762014703,
+      "grad_norm": 0.18724669516086578,
+      "learning_rate": 5.004307747331413e-05,
+      "loss": 1.0676,
+      "step": 8251
+    },
+    {
+      "epoch": 0.5013670332340968,
+      "grad_norm": 0.4504494369029999,
+      "learning_rate": 5.003350470310403e-05,
+      "loss": 1.198,
+      "step": 8252
+    },
+    {
+      "epoch": 0.5014277902667233,
+      "grad_norm": 0.20448404550552368,
+      "learning_rate": 5.002393193166582e-05,
+      "loss": 1.1066,
+      "step": 8253
+    },
+    {
+      "epoch": 0.5014885472993499,
+      "grad_norm": 0.26566049456596375,
+      "learning_rate": 5.0014359159350385e-05,
+      "loss": 1.0701,
+      "step": 8254
+    },
+    {
+      "epoch": 0.5015493043319764,
+      "grad_norm": 0.1513049602508545,
+      "learning_rate": 5.0004786386508616e-05,
+      "loss": 1.1191,
+      "step": 8255
+    },
+    {
+      "epoch": 0.501610061364603,
+      "grad_norm": 0.11535308510065079,
+      "learning_rate": 4.9995213613491396e-05,
+      "loss": 1.0038,
+      "step": 8256
+    },
+    {
+      "epoch": 0.5016708183972295,
+      "grad_norm": 0.18777461349964142,
+      "learning_rate": 4.998564084064962e-05,
+      "loss": 1.0201,
+      "step": 8257
+    },
+    {
+      "epoch": 0.501731575429856,
+      "grad_norm": 0.31867924332618713,
+      "learning_rate": 4.9976068068334176e-05,
+      "loss": 1.0506,
+      "step": 8258
+    },
+    {
+      "epoch": 0.5017923324624826,
+      "grad_norm": 0.14262045919895172,
+      "learning_rate": 4.9966495296895974e-05,
+      "loss": 1.136,
+      "step": 8259
+    },
+    {
+      "epoch": 0.5018530894951091,
+      "grad_norm": 0.1401367485523224,
+      "learning_rate": 4.995692252668589e-05,
+      "loss": 1.0649,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5019138465277356,
+      "grad_norm": 0.1476631611585617,
+      "learning_rate": 4.9947349758054816e-05,
+      "loss": 1.1078,
+      "step": 8261
+    },
+    {
+      "epoch": 0.5019746035603622,
+      "grad_norm": 0.1467423439025879,
+      "learning_rate": 4.993777699135364e-05,
+      "loss": 1.078,
+      "step": 8262
+    },
+    {
+      "epoch": 0.5020353605929886,
+      "grad_norm": 0.36051467061042786,
+      "learning_rate": 4.9928204226933264e-05,
+      "loss": 1.2199,
+      "step": 8263
+    },
+    {
+      "epoch": 0.5020961176256151,
+      "grad_norm": 0.42102688550949097,
+      "learning_rate": 4.991863146514459e-05,
+      "loss": 1.045,
+      "step": 8264
+    },
+    {
+      "epoch": 0.5021568746582417,
+      "grad_norm": 0.12762510776519775,
+      "learning_rate": 4.990905870633848e-05,
+      "loss": 1.0754,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5022176316908682,
+      "grad_norm": 0.18396292626857758,
+      "learning_rate": 4.989948595086586e-05,
+      "loss": 1.2735,
+      "step": 8266
+    },
+    {
+      "epoch": 0.5022783887234947,
+      "grad_norm": 0.13720965385437012,
+      "learning_rate": 4.988991319907758e-05,
+      "loss": 1.0712,
+      "step": 8267
+    },
+    {
+      "epoch": 0.5023391457561213,
+      "grad_norm": 0.14273101091384888,
+      "learning_rate": 4.988034045132456e-05,
+      "loss": 1.0345,
+      "step": 8268
+    },
+    {
+      "epoch": 0.5023999027887478,
+      "grad_norm": 1.1191110610961914,
+      "learning_rate": 4.9870767707957686e-05,
+      "loss": 1.172,
+      "step": 8269
+    },
+    {
+      "epoch": 0.5024606598213743,
+      "grad_norm": 0.1882486194372177,
+      "learning_rate": 4.9861194969327835e-05,
+      "loss": 1.0999,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5025214168540009,
+      "grad_norm": 0.15251389145851135,
+      "learning_rate": 4.985162223578592e-05,
+      "loss": 1.0377,
+      "step": 8271
+    },
+    {
+      "epoch": 0.5025821738866274,
+      "grad_norm": 0.2699245810508728,
+      "learning_rate": 4.984204950768282e-05,
+      "loss": 1.1296,
+      "step": 8272
+    },
+    {
+      "epoch": 0.5026429309192539,
+      "grad_norm": 1.4700466394424438,
+      "learning_rate": 4.983247678536944e-05,
+      "loss": 1.1121,
+      "step": 8273
+    },
+    {
+      "epoch": 0.5027036879518805,
+      "grad_norm": 0.9751190543174744,
+      "learning_rate": 4.982290406919664e-05,
+      "loss": 1.0971,
+      "step": 8274
+    },
+    {
+      "epoch": 0.502764444984507,
+      "grad_norm": 0.29272234439849854,
+      "learning_rate": 4.981333135951532e-05,
+      "loss": 1.1349,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5028252020171334,
+      "grad_norm": 0.31304648518562317,
+      "learning_rate": 4.980375865667638e-05,
+      "loss": 1.072,
+      "step": 8276
+    },
+    {
+      "epoch": 0.50288595904976,
+      "grad_norm": 0.20816628634929657,
+      "learning_rate": 4.979418596103071e-05,
+      "loss": 1.0687,
+      "step": 8277
+    },
+    {
+      "epoch": 0.5029467160823865,
+      "grad_norm": 0.11504765599966049,
+      "learning_rate": 4.978461327292918e-05,
+      "loss": 1.067,
+      "step": 8278
+    },
+    {
+      "epoch": 0.503007473115013,
+      "grad_norm": 0.1644960343837738,
+      "learning_rate": 4.977504059272273e-05,
+      "loss": 1.1067,
+      "step": 8279
+    },
+    {
+      "epoch": 0.5030682301476396,
+      "grad_norm": 0.17996364831924438,
+      "learning_rate": 4.9765467920762164e-05,
+      "loss": 1.1757,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5031289871802661,
+      "grad_norm": 0.24800243973731995,
+      "learning_rate": 4.975589525739844e-05,
+      "loss": 1.1103,
+      "step": 8281
+    },
+    {
+      "epoch": 0.5031897442128926,
+      "grad_norm": 0.12489693611860275,
+      "learning_rate": 4.974632260298242e-05,
+      "loss": 1.0354,
+      "step": 8282
+    },
+    {
+      "epoch": 0.5032505012455192,
+      "grad_norm": 0.2894044518470764,
+      "learning_rate": 4.973674995786498e-05,
+      "loss": 1.1608,
+      "step": 8283
+    },
+    {
+      "epoch": 0.5033112582781457,
+      "grad_norm": 0.1456850916147232,
+      "learning_rate": 4.9727177322397034e-05,
+      "loss": 1.0766,
+      "step": 8284
+    },
+    {
+      "epoch": 0.5033720153107722,
+      "grad_norm": 0.1728762984275818,
+      "learning_rate": 4.971760469692946e-05,
+      "loss": 1.1459,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5034327723433988,
+      "grad_norm": 0.17219896614551544,
+      "learning_rate": 4.9708032081813144e-05,
+      "loss": 1.0324,
+      "step": 8286
+    },
+    {
+      "epoch": 0.5034935293760253,
+      "grad_norm": 0.1751197874546051,
+      "learning_rate": 4.969845947739896e-05,
+      "loss": 1.0958,
+      "step": 8287
+    },
+    {
+      "epoch": 0.5035542864086519,
+      "grad_norm": 0.1838088184595108,
+      "learning_rate": 4.96888868840378e-05,
+      "loss": 1.2078,
+      "step": 8288
+    },
+    {
+      "epoch": 0.5036150434412783,
+      "grad_norm": 0.20298756659030914,
+      "learning_rate": 4.967931430208055e-05,
+      "loss": 1.064,
+      "step": 8289
+    },
+    {
+      "epoch": 0.5036758004739048,
+      "grad_norm": 0.182186096906662,
+      "learning_rate": 4.96697417318781e-05,
+      "loss": 1.0071,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5037365575065313,
+      "grad_norm": 0.22610075771808624,
+      "learning_rate": 4.9660169173781316e-05,
+      "loss": 1.2349,
+      "step": 8291
+    },
+    {
+      "epoch": 0.5037973145391579,
+      "grad_norm": 0.15375541150569916,
+      "learning_rate": 4.965059662814112e-05,
+      "loss": 1.0728,
+      "step": 8292
+    },
+    {
+      "epoch": 0.5038580715717844,
+      "grad_norm": 0.23978929221630096,
+      "learning_rate": 4.964102409530837e-05,
+      "loss": 1.004,
+      "step": 8293
+    },
+    {
+      "epoch": 0.503918828604411,
+      "grad_norm": 0.1773739457130432,
+      "learning_rate": 4.963145157563395e-05,
+      "loss": 1.0872,
+      "step": 8294
+    },
+    {
+      "epoch": 0.5039795856370375,
+      "grad_norm": 0.11503232270479202,
+      "learning_rate": 4.962187906946873e-05,
+      "loss": 1.0313,
+      "step": 8295
+    },
+    {
+      "epoch": 0.504040342669664,
+      "grad_norm": 0.14824290573596954,
+      "learning_rate": 4.961230657716361e-05,
+      "loss": 1.0599,
+      "step": 8296
+    },
+    {
+      "epoch": 0.5041010997022906,
+      "grad_norm": 0.21134766936302185,
+      "learning_rate": 4.960273409906947e-05,
+      "loss": 1.0274,
+      "step": 8297
+    },
+    {
+      "epoch": 0.5041618567349171,
+      "grad_norm": 0.2069268524646759,
+      "learning_rate": 4.9593161635537194e-05,
+      "loss": 1.1218,
+      "step": 8298
+    },
+    {
+      "epoch": 0.5042226137675436,
+      "grad_norm": 2.4647109508514404,
+      "learning_rate": 4.958358918691764e-05,
+      "loss": 1.0774,
+      "step": 8299
+    },
+    {
+      "epoch": 0.5042833708001702,
+      "grad_norm": 1.0594382286071777,
+      "learning_rate": 4.9574016753561744e-05,
+      "loss": 1.086,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5043441278327967,
+      "grad_norm": 0.39914965629577637,
+      "learning_rate": 4.956444433582031e-05,
+      "loss": 1.0667,
+      "step": 8301
+    },
+    {
+      "epoch": 0.5044048848654231,
+      "grad_norm": 8.242327690124512,
+      "learning_rate": 4.955487193404426e-05,
+      "loss": 1.0377,
+      "step": 8302
+    },
+    {
+      "epoch": 0.5044656418980497,
+      "grad_norm": 0.35576769709587097,
+      "learning_rate": 4.954529954858447e-05,
+      "loss": 1.1649,
+      "step": 8303
+    },
+    {
+      "epoch": 0.5045263989306762,
+      "grad_norm": 0.2670916020870209,
+      "learning_rate": 4.9535727179791804e-05,
+      "loss": 1.0443,
+      "step": 8304
+    },
+    {
+      "epoch": 0.5045871559633027,
+      "grad_norm": 13.714424133300781,
+      "learning_rate": 4.9526154828017154e-05,
+      "loss": 1.2174,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5046479129959293,
+      "grad_norm": 0.15306521952152252,
+      "learning_rate": 4.95165824936114e-05,
+      "loss": 1.0686,
+      "step": 8306
+    },
+    {
+      "epoch": 0.5047086700285558,
+      "grad_norm": 0.4864562451839447,
+      "learning_rate": 4.950701017692541e-05,
+      "loss": 1.1131,
+      "step": 8307
+    },
+    {
+      "epoch": 0.5047694270611823,
+      "grad_norm": 0.12344438582658768,
+      "learning_rate": 4.9497437878310045e-05,
+      "loss": 1.0559,
+      "step": 8308
+    },
+    {
+      "epoch": 0.5048301840938089,
+      "grad_norm": 0.17158269882202148,
+      "learning_rate": 4.9487865598116177e-05,
+      "loss": 1.0564,
+      "step": 8309
+    },
+    {
+      "epoch": 0.5048909411264354,
+      "grad_norm": 0.14548590779304504,
+      "learning_rate": 4.947829333669471e-05,
+      "loss": 1.0802,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5049516981590619,
+      "grad_norm": 0.1790032833814621,
+      "learning_rate": 4.946872109439651e-05,
+      "loss": 1.029,
+      "step": 8311
+    },
+    {
+      "epoch": 0.5050124551916885,
+      "grad_norm": 2.055565357208252,
+      "learning_rate": 4.945914887157242e-05,
+      "loss": 1.1672,
+      "step": 8312
+    },
+    {
+      "epoch": 0.505073212224315,
+      "grad_norm": 0.1323356032371521,
+      "learning_rate": 4.9449576668573374e-05,
+      "loss": 1.0071,
+      "step": 8313
+    },
+    {
+      "epoch": 0.5051339692569415,
+      "grad_norm": 1.1997883319854736,
+      "learning_rate": 4.9440004485750165e-05,
+      "loss": 1.0596,
+      "step": 8314
+    },
+    {
+      "epoch": 0.505194726289568,
+      "grad_norm": 5.971543788909912,
+      "learning_rate": 4.943043232345371e-05,
+      "loss": 1.1915,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5052554833221945,
+      "grad_norm": 0.6543163657188416,
+      "learning_rate": 4.942086018203486e-05,
+      "loss": 1.1988,
+      "step": 8316
+    },
+    {
+      "epoch": 0.505316240354821,
+      "grad_norm": 0.17534765601158142,
+      "learning_rate": 4.941128806184449e-05,
+      "loss": 1.0399,
+      "step": 8317
+    },
+    {
+      "epoch": 0.5053769973874476,
+      "grad_norm": 0.15541040897369385,
+      "learning_rate": 4.940171596323348e-05,
+      "loss": 1.085,
+      "step": 8318
+    },
+    {
+      "epoch": 0.5054377544200741,
+      "grad_norm": 0.44979262351989746,
+      "learning_rate": 4.939214388655269e-05,
+      "loss": 1.2206,
+      "step": 8319
+    },
+    {
+      "epoch": 0.5054985114527006,
+      "grad_norm": 0.1236925721168518,
+      "learning_rate": 4.938257183215299e-05,
+      "loss": 1.0607,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5055592684853272,
+      "grad_norm": 0.17472024261951447,
+      "learning_rate": 4.937299980038523e-05,
+      "loss": 1.1329,
+      "step": 8321
+    },
+    {
+      "epoch": 0.5056200255179537,
+      "grad_norm": 0.1306833028793335,
+      "learning_rate": 4.936342779160027e-05,
+      "loss": 1.1088,
+      "step": 8322
+    },
+    {
+      "epoch": 0.5056807825505802,
+      "grad_norm": 0.14495404064655304,
+      "learning_rate": 4.9353855806149003e-05,
+      "loss": 1.0233,
+      "step": 8323
+    },
+    {
+      "epoch": 0.5057415395832068,
+      "grad_norm": 0.13575270771980286,
+      "learning_rate": 4.934428384438227e-05,
+      "loss": 1.0718,
+      "step": 8324
+    },
+    {
+      "epoch": 0.5058022966158333,
+      "grad_norm": 0.26696622371673584,
+      "learning_rate": 4.933471190665093e-05,
+      "loss": 1.1328,
+      "step": 8325
+    },
+    {
+      "epoch": 0.5058630536484598,
+      "grad_norm": 0.18755652010440826,
+      "learning_rate": 4.9325139993305894e-05,
+      "loss": 1.1105,
+      "step": 8326
+    },
+    {
+      "epoch": 0.5059238106810864,
+      "grad_norm": 0.2036897838115692,
+      "learning_rate": 4.931556810469794e-05,
+      "loss": 1.0683,
+      "step": 8327
+    },
+    {
+      "epoch": 0.5059845677137128,
+      "grad_norm": 0.1344156563282013,
+      "learning_rate": 4.930599624117798e-05,
+      "loss": 1.0316,
+      "step": 8328
+    },
+    {
+      "epoch": 0.5060453247463393,
+      "grad_norm": 0.11509817093610764,
+      "learning_rate": 4.929642440309687e-05,
+      "loss": 1.0582,
+      "step": 8329
+    },
+    {
+      "epoch": 0.5061060817789659,
+      "grad_norm": 0.33605095744132996,
+      "learning_rate": 4.9286852590805446e-05,
+      "loss": 1.0665,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5061668388115924,
+      "grad_norm": 0.22083233296871185,
+      "learning_rate": 4.927728080465458e-05,
+      "loss": 1.1513,
+      "step": 8331
+    },
+    {
+      "epoch": 0.506227595844219,
+      "grad_norm": 0.2308490127325058,
+      "learning_rate": 4.9267709044995136e-05,
+      "loss": 1.0228,
+      "step": 8332
+    },
+    {
+      "epoch": 0.5062883528768455,
+      "grad_norm": 0.23748978972434998,
+      "learning_rate": 4.925813731217797e-05,
+      "loss": 1.2818,
+      "step": 8333
+    },
+    {
+      "epoch": 0.506349109909472,
+      "grad_norm": 0.41256120800971985,
+      "learning_rate": 4.92485656065539e-05,
+      "loss": 1.2997,
+      "step": 8334
+    },
+    {
+      "epoch": 0.5064098669420986,
+      "grad_norm": 0.18186990916728973,
+      "learning_rate": 4.923899392847381e-05,
+      "loss": 1.1078,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5064706239747251,
+      "grad_norm": 0.26134192943573,
+      "learning_rate": 4.922942227828855e-05,
+      "loss": 1.1419,
+      "step": 8336
+    },
+    {
+      "epoch": 0.5065313810073516,
+      "grad_norm": 0.33839961886405945,
+      "learning_rate": 4.921985065634896e-05,
+      "loss": 1.0407,
+      "step": 8337
+    },
+    {
+      "epoch": 0.5065921380399782,
+      "grad_norm": 0.1901620626449585,
+      "learning_rate": 4.921027906300588e-05,
+      "loss": 1.0358,
+      "step": 8338
+    },
+    {
+      "epoch": 0.5066528950726047,
+      "grad_norm": 0.23898373544216156,
+      "learning_rate": 4.920070749861019e-05,
+      "loss": 1.0535,
+      "step": 8339
+    },
+    {
+      "epoch": 0.5067136521052312,
+      "grad_norm": 0.17903980612754822,
+      "learning_rate": 4.9191135963512744e-05,
+      "loss": 1.047,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5067744091378578,
+      "grad_norm": 0.12277766317129135,
+      "learning_rate": 4.918156445806434e-05,
+      "loss": 1.0323,
+      "step": 8341
+    },
+    {
+      "epoch": 0.5068351661704842,
+      "grad_norm": 1.125687599182129,
+      "learning_rate": 4.917199298261585e-05,
+      "loss": 1.175,
+      "step": 8342
+    },
+    {
+      "epoch": 0.5068959232031107,
+      "grad_norm": 0.12285816669464111,
+      "learning_rate": 4.916242153751811e-05,
+      "loss": 1.0709,
+      "step": 8343
+    },
+    {
+      "epoch": 0.5069566802357373,
+      "grad_norm": 0.15012361109256744,
+      "learning_rate": 4.9152850123121977e-05,
+      "loss": 1.0389,
+      "step": 8344
+    },
+    {
+      "epoch": 0.5070174372683638,
+      "grad_norm": 0.8612470030784607,
+      "learning_rate": 4.9143278739778296e-05,
+      "loss": 1.1423,
+      "step": 8345
+    },
+    {
+      "epoch": 0.5070781943009903,
+      "grad_norm": 0.22222037613391876,
+      "learning_rate": 4.91337073878379e-05,
+      "loss": 1.068,
+      "step": 8346
+    },
+    {
+      "epoch": 0.5071389513336169,
+      "grad_norm": 40.13809585571289,
+      "learning_rate": 4.912413606765161e-05,
+      "loss": 1.2587,
+      "step": 8347
+    },
+    {
+      "epoch": 0.5071997083662434,
+      "grad_norm": 0.25332459807395935,
+      "learning_rate": 4.911456477957029e-05,
+      "loss": 1.0471,
+      "step": 8348
+    },
+    {
+      "epoch": 0.5072604653988699,
+      "grad_norm": 0.20554561913013458,
+      "learning_rate": 4.910499352394478e-05,
+      "loss": 1.0618,
+      "step": 8349
+    },
+    {
+      "epoch": 0.5073212224314965,
+      "grad_norm": 0.20227853953838348,
+      "learning_rate": 4.909542230112589e-05,
+      "loss": 1.1739,
+      "step": 8350
+    },
+    {
+      "epoch": 0.507381979464123,
+      "grad_norm": 0.16589759290218353,
+      "learning_rate": 4.9085851111464464e-05,
+      "loss": 1.0338,
+      "step": 8351
+    },
+    {
+      "epoch": 0.5074427364967495,
+      "grad_norm": 0.25207895040512085,
+      "learning_rate": 4.9076279955311357e-05,
+      "loss": 1.0719,
+      "step": 8352
+    },
+    {
+      "epoch": 0.5075034935293761,
+      "grad_norm": 0.5040931105613708,
+      "learning_rate": 4.9066708833017396e-05,
+      "loss": 1.1578,
+      "step": 8353
+    },
+    {
+      "epoch": 0.5075642505620026,
+      "grad_norm": 0.238718181848526,
+      "learning_rate": 4.905713774493339e-05,
+      "loss": 1.1963,
+      "step": 8354
+    },
+    {
+      "epoch": 0.507625007594629,
+      "grad_norm": 0.24442973732948303,
+      "learning_rate": 4.9047566691410174e-05,
+      "loss": 1.0836,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5076857646272556,
+      "grad_norm": 0.6389700770378113,
+      "learning_rate": 4.9037995672798595e-05,
+      "loss": 1.0674,
+      "step": 8356
+    },
+    {
+      "epoch": 0.5077465216598821,
+      "grad_norm": 0.16048896312713623,
+      "learning_rate": 4.9028424689449466e-05,
+      "loss": 1.0854,
+      "step": 8357
+    },
+    {
+      "epoch": 0.5078072786925086,
+      "grad_norm": 0.23621916770935059,
+      "learning_rate": 4.901885374171362e-05,
+      "loss": 1.0831,
+      "step": 8358
+    },
+    {
+      "epoch": 0.5078680357251352,
+      "grad_norm": 0.1800125539302826,
+      "learning_rate": 4.900928282994187e-05,
+      "loss": 1.0643,
+      "step": 8359
+    },
+    {
+      "epoch": 0.5079287927577617,
+      "grad_norm": 0.23138806223869324,
+      "learning_rate": 4.8999711954485075e-05,
+      "loss": 1.1036,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5079895497903882,
+      "grad_norm": 0.13819749653339386,
+      "learning_rate": 4.8990141115694024e-05,
+      "loss": 1.0323,
+      "step": 8361
+    },
+    {
+      "epoch": 0.5080503068230148,
+      "grad_norm": 0.14696162939071655,
+      "learning_rate": 4.898057031391953e-05,
+      "loss": 1.1208,
+      "step": 8362
+    },
+    {
+      "epoch": 0.5081110638556413,
+      "grad_norm": 0.12947984039783478,
+      "learning_rate": 4.897099954951243e-05,
+      "loss": 1.0478,
+      "step": 8363
+    },
+    {
+      "epoch": 0.5081718208882678,
+      "grad_norm": 0.1823125034570694,
+      "learning_rate": 4.896142882282355e-05,
+      "loss": 1.1095,
+      "step": 8364
+    },
+    {
+      "epoch": 0.5082325779208944,
+      "grad_norm": 0.5740890502929688,
+      "learning_rate": 4.895185813420369e-05,
+      "loss": 1.1337,
+      "step": 8365
+    },
+    {
+      "epoch": 0.5082933349535209,
+      "grad_norm": 0.18487799167633057,
+      "learning_rate": 4.8942287484003684e-05,
+      "loss": 1.1344,
+      "step": 8366
+    },
+    {
+      "epoch": 0.5083540919861474,
+      "grad_norm": 0.14665530622005463,
+      "learning_rate": 4.893271687257434e-05,
+      "loss": 1.0444,
+      "step": 8367
+    },
+    {
+      "epoch": 0.5084148490187739,
+      "grad_norm": 0.18262310326099396,
+      "learning_rate": 4.892314630026644e-05,
+      "loss": 1.1154,
+      "step": 8368
+    },
+    {
+      "epoch": 0.5084756060514004,
+      "grad_norm": 0.16470301151275635,
+      "learning_rate": 4.8913575767430834e-05,
+      "loss": 1.066,
+      "step": 8369
+    },
+    {
+      "epoch": 0.5085363630840269,
+      "grad_norm": 0.12337921559810638,
+      "learning_rate": 4.890400527441832e-05,
+      "loss": 1.0216,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5085971201166535,
+      "grad_norm": 0.42846640944480896,
+      "learning_rate": 4.889443482157969e-05,
+      "loss": 1.2121,
+      "step": 8371
+    },
+    {
+      "epoch": 0.50865787714928,
+      "grad_norm": 0.332369863986969,
+      "learning_rate": 4.8884864409265786e-05,
+      "loss": 1.113,
+      "step": 8372
+    },
+    {
+      "epoch": 0.5087186341819065,
+      "grad_norm": 0.39648205041885376,
+      "learning_rate": 4.887529403782739e-05,
+      "loss": 1.0549,
+      "step": 8373
+    },
+    {
+      "epoch": 0.5087793912145331,
+      "grad_norm": 0.1325053721666336,
+      "learning_rate": 4.886572370761531e-05,
+      "loss": 1.0496,
+      "step": 8374
+    },
+    {
+      "epoch": 0.5088401482471596,
+      "grad_norm": 0.38651201128959656,
+      "learning_rate": 4.885615341898034e-05,
+      "loss": 1.0689,
+      "step": 8375
+    },
+    {
+      "epoch": 0.5089009052797862,
+      "grad_norm": 0.1649765968322754,
+      "learning_rate": 4.8846583172273275e-05,
+      "loss": 1.0749,
+      "step": 8376
+    },
+    {
+      "epoch": 0.5089616623124127,
+      "grad_norm": 0.11553426831960678,
+      "learning_rate": 4.8837012967844944e-05,
+      "loss": 1.0396,
+      "step": 8377
+    },
+    {
+      "epoch": 0.5090224193450392,
+      "grad_norm": 0.12951795756816864,
+      "learning_rate": 4.882744280604612e-05,
+      "loss": 1.0914,
+      "step": 8378
+    },
+    {
+      "epoch": 0.5090831763776658,
+      "grad_norm": 0.20966719090938568,
+      "learning_rate": 4.8817872687227594e-05,
+      "loss": 1.088,
+      "step": 8379
+    },
+    {
+      "epoch": 0.5091439334102923,
+      "grad_norm": 0.21130825579166412,
+      "learning_rate": 4.8808302611740215e-05,
+      "loss": 1.0573,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5092046904429187,
+      "grad_norm": 0.19537219405174255,
+      "learning_rate": 4.87987325799347e-05,
+      "loss": 1.0652,
+      "step": 8381
+    },
+    {
+      "epoch": 0.5092654474755453,
+      "grad_norm": 0.2663423418998718,
+      "learning_rate": 4.878916259216188e-05,
+      "loss": 1.0784,
+      "step": 8382
+    },
+    {
+      "epoch": 0.5093262045081718,
+      "grad_norm": 0.6188851594924927,
+      "learning_rate": 4.877959264877254e-05,
+      "loss": 1.0337,
+      "step": 8383
+    },
+    {
+      "epoch": 0.5093869615407983,
+      "grad_norm": 0.2325194925069809,
+      "learning_rate": 4.877002275011746e-05,
+      "loss": 1.1417,
+      "step": 8384
+    },
+    {
+      "epoch": 0.5094477185734249,
+      "grad_norm": 0.13020442426204681,
+      "learning_rate": 4.876045289654744e-05,
+      "loss": 1.0677,
+      "step": 8385
+    },
+    {
+      "epoch": 0.5095084756060514,
+      "grad_norm": 0.1499943882226944,
+      "learning_rate": 4.875088308841326e-05,
+      "loss": 1.0139,
+      "step": 8386
+    },
+    {
+      "epoch": 0.5095692326386779,
+      "grad_norm": 0.18690651655197144,
+      "learning_rate": 4.8741313326065715e-05,
+      "loss": 1.0417,
+      "step": 8387
+    },
+    {
+      "epoch": 0.5096299896713045,
+      "grad_norm": 0.21670256555080414,
+      "learning_rate": 4.873174360985557e-05,
+      "loss": 1.0929,
+      "step": 8388
+    },
+    {
+      "epoch": 0.509690746703931,
+      "grad_norm": 0.15587611496448517,
+      "learning_rate": 4.872217394013359e-05,
+      "loss": 1.0635,
+      "step": 8389
+    },
+    {
+      "epoch": 0.5097515037365575,
+      "grad_norm": 0.6354724168777466,
+      "learning_rate": 4.8712604317250576e-05,
+      "loss": 1.1001,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5098122607691841,
+      "grad_norm": 0.15006624162197113,
+      "learning_rate": 4.87030347415573e-05,
+      "loss": 1.1248,
+      "step": 8391
+    },
+    {
+      "epoch": 0.5098730178018106,
+      "grad_norm": 0.12928888201713562,
+      "learning_rate": 4.869346521340453e-05,
+      "loss": 1.0064,
+      "step": 8392
+    },
+    {
+      "epoch": 0.5099337748344371,
+      "grad_norm": 0.18467842042446136,
+      "learning_rate": 4.868389573314308e-05,
+      "loss": 1.1236,
+      "step": 8393
+    },
+    {
+      "epoch": 0.5099945318670636,
+      "grad_norm": 0.74225914478302,
+      "learning_rate": 4.867432630112365e-05,
+      "loss": 1.0268,
+      "step": 8394
+    },
+    {
+      "epoch": 0.5100552888996901,
+      "grad_norm": 0.1632264256477356,
+      "learning_rate": 4.8664756917697066e-05,
+      "loss": 1.1005,
+      "step": 8395
+    },
+    {
+      "epoch": 0.5101160459323166,
+      "grad_norm": 0.17403481900691986,
+      "learning_rate": 4.8655187583214066e-05,
+      "loss": 1.0288,
+      "step": 8396
+    },
+    {
+      "epoch": 0.5101768029649432,
+      "grad_norm": 0.14741408824920654,
+      "learning_rate": 4.8645618298025415e-05,
+      "loss": 1.0774,
+      "step": 8397
+    },
+    {
+      "epoch": 0.5102375599975697,
+      "grad_norm": 0.20694880187511444,
+      "learning_rate": 4.86360490624819e-05,
+      "loss": 1.1291,
+      "step": 8398
+    },
+    {
+      "epoch": 0.5102983170301962,
+      "grad_norm": 0.15443269908428192,
+      "learning_rate": 4.862647987693427e-05,
+      "loss": 1.0092,
+      "step": 8399
+    },
+    {
+      "epoch": 0.5103590740628228,
+      "grad_norm": 0.17777439951896667,
+      "learning_rate": 4.8616910741733294e-05,
+      "loss": 1.0872,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5104198310954493,
+      "grad_norm": 0.12636573612689972,
+      "learning_rate": 4.860734165722971e-05,
+      "loss": 1.0549,
+      "step": 8401
+    },
+    {
+      "epoch": 0.5104805881280758,
+      "grad_norm": 0.21173891425132751,
+      "learning_rate": 4.859777262377428e-05,
+      "loss": 1.0293,
+      "step": 8402
+    },
+    {
+      "epoch": 0.5105413451607024,
+      "grad_norm": 0.2152070701122284,
+      "learning_rate": 4.858820364171777e-05,
+      "loss": 1.0677,
+      "step": 8403
+    },
+    {
+      "epoch": 0.5106021021933289,
+      "grad_norm": 0.2382003664970398,
+      "learning_rate": 4.8578634711410933e-05,
+      "loss": 1.0634,
+      "step": 8404
+    },
+    {
+      "epoch": 0.5106628592259554,
+      "grad_norm": 0.310569703578949,
+      "learning_rate": 4.856906583320451e-05,
+      "loss": 1.0424,
+      "step": 8405
+    },
+    {
+      "epoch": 0.510723616258582,
+      "grad_norm": 0.2448827475309372,
+      "learning_rate": 4.855949700744926e-05,
+      "loss": 1.0607,
+      "step": 8406
+    },
+    {
+      "epoch": 0.5107843732912084,
+      "grad_norm": 0.59648197889328,
+      "learning_rate": 4.854992823449594e-05,
+      "loss": 1.1614,
+      "step": 8407
+    },
+    {
+      "epoch": 0.5108451303238349,
+      "grad_norm": 0.13383783400058746,
+      "learning_rate": 4.854035951469527e-05,
+      "loss": 1.0891,
+      "step": 8408
+    },
+    {
+      "epoch": 0.5109058873564615,
+      "grad_norm": 0.1836208999156952,
+      "learning_rate": 4.8530790848398e-05,
+      "loss": 1.1025,
+      "step": 8409
+    },
+    {
+      "epoch": 0.510966644389088,
+      "grad_norm": 0.16916318237781525,
+      "learning_rate": 4.8521222235954864e-05,
+      "loss": 1.1377,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5110274014217145,
+      "grad_norm": 0.22267036139965057,
+      "learning_rate": 4.851165367771663e-05,
+      "loss": 1.0696,
+      "step": 8411
+    },
+    {
+      "epoch": 0.5110881584543411,
+      "grad_norm": 0.12919865548610687,
+      "learning_rate": 4.8502085174034015e-05,
+      "loss": 1.0257,
+      "step": 8412
+    },
+    {
+      "epoch": 0.5111489154869676,
+      "grad_norm": 0.16686958074569702,
+      "learning_rate": 4.8492516725257766e-05,
+      "loss": 1.0933,
+      "step": 8413
+    },
+    {
+      "epoch": 0.5112096725195941,
+      "grad_norm": 0.9743154048919678,
+      "learning_rate": 4.84829483317386e-05,
+      "loss": 1.0628,
+      "step": 8414
+    },
+    {
+      "epoch": 0.5112704295522207,
+      "grad_norm": 0.20970475673675537,
+      "learning_rate": 4.847337999382724e-05,
+      "loss": 1.1368,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5113311865848472,
+      "grad_norm": 0.6008890867233276,
+      "learning_rate": 4.8463811711874446e-05,
+      "loss": 1.0215,
+      "step": 8416
+    },
+    {
+      "epoch": 0.5113919436174738,
+      "grad_norm": 0.15277495980262756,
+      "learning_rate": 4.8454243486230934e-05,
+      "loss": 1.1052,
+      "step": 8417
+    },
+    {
+      "epoch": 0.5114527006501003,
+      "grad_norm": 0.278396874666214,
+      "learning_rate": 4.844467531724741e-05,
+      "loss": 1.2837,
+      "step": 8418
+    },
+    {
+      "epoch": 0.5115134576827268,
+      "grad_norm": 0.13862337172031403,
+      "learning_rate": 4.843510720527462e-05,
+      "loss": 1.0813,
+      "step": 8419
+    },
+    {
+      "epoch": 0.5115742147153532,
+      "grad_norm": 0.13305699825286865,
+      "learning_rate": 4.8425539150663294e-05,
+      "loss": 1.0853,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5116349717479798,
+      "grad_norm": 0.10755746066570282,
+      "learning_rate": 4.8415971153764126e-05,
+      "loss": 1.0058,
+      "step": 8421
+    },
+    {
+      "epoch": 0.5116957287806063,
+      "grad_norm": 0.1701010763645172,
+      "learning_rate": 4.840640321492784e-05,
+      "loss": 1.1213,
+      "step": 8422
+    },
+    {
+      "epoch": 0.5117564858132329,
+      "grad_norm": 0.20484969019889832,
+      "learning_rate": 4.839683533450514e-05,
+      "loss": 1.0533,
+      "step": 8423
+    },
+    {
+      "epoch": 0.5118172428458594,
+      "grad_norm": 0.15652777254581451,
+      "learning_rate": 4.8387267512846764e-05,
+      "loss": 1.0716,
+      "step": 8424
+    },
+    {
+      "epoch": 0.5118779998784859,
+      "grad_norm": 0.1799188107252121,
+      "learning_rate": 4.8377699750303404e-05,
+      "loss": 1.0912,
+      "step": 8425
+    },
+    {
+      "epoch": 0.5119387569111125,
+      "grad_norm": 0.3735145628452301,
+      "learning_rate": 4.836813204722577e-05,
+      "loss": 1.2903,
+      "step": 8426
+    },
+    {
+      "epoch": 0.511999513943739,
+      "grad_norm": 0.14639516174793243,
+      "learning_rate": 4.83585644039646e-05,
+      "loss": 1.0672,
+      "step": 8427
+    },
+    {
+      "epoch": 0.5120602709763655,
+      "grad_norm": 0.17273227870464325,
+      "learning_rate": 4.8348996820870535e-05,
+      "loss": 1.1334,
+      "step": 8428
+    },
+    {
+      "epoch": 0.5121210280089921,
+      "grad_norm": 0.14567801356315613,
+      "learning_rate": 4.833942929829433e-05,
+      "loss": 1.0344,
+      "step": 8429
+    },
+    {
+      "epoch": 0.5121817850416186,
+      "grad_norm": 0.13478966057300568,
+      "learning_rate": 4.8329861836586645e-05,
+      "loss": 1.0883,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5122425420742451,
+      "grad_norm": 0.23374485969543457,
+      "learning_rate": 4.832029443609821e-05,
+      "loss": 1.011,
+      "step": 8431
+    },
+    {
+      "epoch": 0.5123032991068717,
+      "grad_norm": 0.13496573269367218,
+      "learning_rate": 4.83107270971797e-05,
+      "loss": 1.0662,
+      "step": 8432
+    },
+    {
+      "epoch": 0.5123640561394981,
+      "grad_norm": 0.17448122799396515,
+      "learning_rate": 4.830115982018183e-05,
+      "loss": 1.0729,
+      "step": 8433
+    },
+    {
+      "epoch": 0.5124248131721246,
+      "grad_norm": 0.4021522104740143,
+      "learning_rate": 4.829159260545527e-05,
+      "loss": 1.104,
+      "step": 8434
+    },
+    {
+      "epoch": 0.5124855702047512,
+      "grad_norm": 0.29043328762054443,
+      "learning_rate": 4.828202545335071e-05,
+      "loss": 1.1243,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5125463272373777,
+      "grad_norm": 0.13032442331314087,
+      "learning_rate": 4.827245836421883e-05,
+      "loss": 1.063,
+      "step": 8436
+    },
+    {
+      "epoch": 0.5126070842700042,
+      "grad_norm": 0.17315416038036346,
+      "learning_rate": 4.8262891338410334e-05,
+      "loss": 1.0945,
+      "step": 8437
+    },
+    {
+      "epoch": 0.5126678413026308,
+      "grad_norm": 0.3445857763290405,
+      "learning_rate": 4.825332437627589e-05,
+      "loss": 1.1487,
+      "step": 8438
+    },
+    {
+      "epoch": 0.5127285983352573,
+      "grad_norm": 0.1359899640083313,
+      "learning_rate": 4.824375747816616e-05,
+      "loss": 1.0677,
+      "step": 8439
+    },
+    {
+      "epoch": 0.5127893553678838,
+      "grad_norm": 0.2265770435333252,
+      "learning_rate": 4.8234190644431874e-05,
+      "loss": 1.1656,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5128501124005104,
+      "grad_norm": 0.14991723001003265,
+      "learning_rate": 4.822462387542365e-05,
+      "loss": 1.095,
+      "step": 8441
+    },
+    {
+      "epoch": 0.5129108694331369,
+      "grad_norm": 0.26123762130737305,
+      "learning_rate": 4.821505717149219e-05,
+      "loss": 1.1039,
+      "step": 8442
+    },
+    {
+      "epoch": 0.5129716264657634,
+      "grad_norm": 0.38255685567855835,
+      "learning_rate": 4.8205490532988154e-05,
+      "loss": 1.0928,
+      "step": 8443
+    },
+    {
+      "epoch": 0.51303238349839,
+      "grad_norm": 0.31451302766799927,
+      "learning_rate": 4.819592396026219e-05,
+      "loss": 1.118,
+      "step": 8444
+    },
+    {
+      "epoch": 0.5130931405310165,
+      "grad_norm": 0.2046777606010437,
+      "learning_rate": 4.8186357453665e-05,
+      "loss": 1.0964,
+      "step": 8445
+    },
+    {
+      "epoch": 0.513153897563643,
+      "grad_norm": 0.18283958733081818,
+      "learning_rate": 4.8176791013547226e-05,
+      "loss": 1.0803,
+      "step": 8446
+    },
+    {
+      "epoch": 0.5132146545962695,
+      "grad_norm": 0.11106229573488235,
+      "learning_rate": 4.816722464025954e-05,
+      "loss": 1.0044,
+      "step": 8447
+    },
+    {
+      "epoch": 0.513275411628896,
+      "grad_norm": 0.25317859649658203,
+      "learning_rate": 4.8157658334152566e-05,
+      "loss": 1.2103,
+      "step": 8448
+    },
+    {
+      "epoch": 0.5133361686615225,
+      "grad_norm": 0.20891964435577393,
+      "learning_rate": 4.8148092095576994e-05,
+      "loss": 1.1547,
+      "step": 8449
+    },
+    {
+      "epoch": 0.5133969256941491,
+      "grad_norm": 0.21446585655212402,
+      "learning_rate": 4.813852592488346e-05,
+      "loss": 1.3115,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5134576827267756,
+      "grad_norm": 0.1956537365913391,
+      "learning_rate": 4.812895982242262e-05,
+      "loss": 1.1286,
+      "step": 8451
+    },
+    {
+      "epoch": 0.5135184397594021,
+      "grad_norm": 0.15341708064079285,
+      "learning_rate": 4.811939378854512e-05,
+      "loss": 1.058,
+      "step": 8452
+    },
+    {
+      "epoch": 0.5135791967920287,
+      "grad_norm": 0.14769281446933746,
+      "learning_rate": 4.81098278236016e-05,
+      "loss": 1.0884,
+      "step": 8453
+    },
+    {
+      "epoch": 0.5136399538246552,
+      "grad_norm": 0.21183790266513824,
+      "learning_rate": 4.810026192794273e-05,
+      "loss": 1.0245,
+      "step": 8454
+    },
+    {
+      "epoch": 0.5137007108572818,
+      "grad_norm": 0.14863301813602448,
+      "learning_rate": 4.809069610191911e-05,
+      "loss": 1.1093,
+      "step": 8455
+    },
+    {
+      "epoch": 0.5137614678899083,
+      "grad_norm": 0.12465561926364899,
+      "learning_rate": 4.808113034588138e-05,
+      "loss": 1.0241,
+      "step": 8456
+    },
+    {
+      "epoch": 0.5138222249225348,
+      "grad_norm": 0.16065973043441772,
+      "learning_rate": 4.80715646601802e-05,
+      "loss": 1.0863,
+      "step": 8457
+    },
+    {
+      "epoch": 0.5138829819551614,
+      "grad_norm": 0.1811482161283493,
+      "learning_rate": 4.806199904516619e-05,
+      "loss": 1.05,
+      "step": 8458
+    },
+    {
+      "epoch": 0.5139437389877879,
+      "grad_norm": 0.13222797214984894,
+      "learning_rate": 4.805243350118998e-05,
+      "loss": 1.0773,
+      "step": 8459
+    },
+    {
+      "epoch": 0.5140044960204143,
+      "grad_norm": 0.24850741028785706,
+      "learning_rate": 4.804286802860221e-05,
+      "loss": 1.2645,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5140652530530408,
+      "grad_norm": 0.13444441556930542,
+      "learning_rate": 4.803330262775346e-05,
+      "loss": 1.085,
+      "step": 8461
+    },
+    {
+      "epoch": 0.5141260100856674,
+      "grad_norm": 1.200520396232605,
+      "learning_rate": 4.8023737298994404e-05,
+      "loss": 1.1363,
+      "step": 8462
+    },
+    {
+      "epoch": 0.5141867671182939,
+      "grad_norm": 0.22339344024658203,
+      "learning_rate": 4.8014172042675636e-05,
+      "loss": 1.1247,
+      "step": 8463
+    },
+    {
+      "epoch": 0.5142475241509205,
+      "grad_norm": 0.12136029452085495,
+      "learning_rate": 4.800460685914776e-05,
+      "loss": 1.0682,
+      "step": 8464
+    },
+    {
+      "epoch": 0.514308281183547,
+      "grad_norm": 0.16660837829113007,
+      "learning_rate": 4.799504174876142e-05,
+      "loss": 1.0777,
+      "step": 8465
+    },
+    {
+      "epoch": 0.5143690382161735,
+      "grad_norm": 0.24574171006679535,
+      "learning_rate": 4.7985476711867215e-05,
+      "loss": 1.2804,
+      "step": 8466
+    },
+    {
+      "epoch": 0.5144297952488001,
+      "grad_norm": 0.26523008942604065,
+      "learning_rate": 4.797591174881576e-05,
+      "loss": 1.1038,
+      "step": 8467
+    },
+    {
+      "epoch": 0.5144905522814266,
+      "grad_norm": 0.9609251618385315,
+      "learning_rate": 4.796634685995764e-05,
+      "loss": 1.1237,
+      "step": 8468
+    },
+    {
+      "epoch": 0.5145513093140531,
+      "grad_norm": 0.2267611175775528,
+      "learning_rate": 4.795678204564345e-05,
+      "loss": 1.0742,
+      "step": 8469
+    },
+    {
+      "epoch": 0.5146120663466797,
+      "grad_norm": 0.1811976581811905,
+      "learning_rate": 4.794721730622383e-05,
+      "loss": 1.0544,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5146728233793062,
+      "grad_norm": 0.2098594605922699,
+      "learning_rate": 4.793765264204935e-05,
+      "loss": 1.13,
+      "step": 8471
+    },
+    {
+      "epoch": 0.5147335804119327,
+      "grad_norm": 0.19684365391731262,
+      "learning_rate": 4.79280880534706e-05,
+      "loss": 1.0976,
+      "step": 8472
+    },
+    {
+      "epoch": 0.5147943374445592,
+      "grad_norm": 0.12757156789302826,
+      "learning_rate": 4.791852354083821e-05,
+      "loss": 1.0299,
+      "step": 8473
+    },
+    {
+      "epoch": 0.5148550944771857,
+      "grad_norm": 0.1435091644525528,
+      "learning_rate": 4.790895910450273e-05,
+      "loss": 1.1124,
+      "step": 8474
+    },
+    {
+      "epoch": 0.5149158515098122,
+      "grad_norm": 0.19999675452709198,
+      "learning_rate": 4.789939474481476e-05,
+      "loss": 1.0936,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5149766085424388,
+      "grad_norm": 0.14922617375850677,
+      "learning_rate": 4.788983046212488e-05,
+      "loss": 1.0786,
+      "step": 8476
+    },
+    {
+      "epoch": 0.5150373655750653,
+      "grad_norm": 0.21757040917873383,
+      "learning_rate": 4.7880266256783666e-05,
+      "loss": 1.0763,
+      "step": 8477
+    },
+    {
+      "epoch": 0.5150981226076918,
+      "grad_norm": 0.12935255467891693,
+      "learning_rate": 4.78707021291417e-05,
+      "loss": 1.0149,
+      "step": 8478
+    },
+    {
+      "epoch": 0.5151588796403184,
+      "grad_norm": 0.14098288118839264,
+      "learning_rate": 4.7861138079549566e-05,
+      "loss": 1.0358,
+      "step": 8479
+    },
+    {
+      "epoch": 0.5152196366729449,
+      "grad_norm": 0.12675568461418152,
+      "learning_rate": 4.7851574108357845e-05,
+      "loss": 1.0817,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5152803937055714,
+      "grad_norm": 0.16993513703346252,
+      "learning_rate": 4.784201021591707e-05,
+      "loss": 1.0767,
+      "step": 8481
+    },
+    {
+      "epoch": 0.515341150738198,
+      "grad_norm": 0.9628027677536011,
+      "learning_rate": 4.7832446402577825e-05,
+      "loss": 1.0496,
+      "step": 8482
+    },
+    {
+      "epoch": 0.5154019077708245,
+      "grad_norm": 0.16150809824466705,
+      "learning_rate": 4.7822882668690685e-05,
+      "loss": 1.1095,
+      "step": 8483
+    },
+    {
+      "epoch": 0.515462664803451,
+      "grad_norm": 0.1148170679807663,
+      "learning_rate": 4.78133190146062e-05,
+      "loss": 0.9828,
+      "step": 8484
+    },
+    {
+      "epoch": 0.5155234218360776,
+      "grad_norm": 0.15926486253738403,
+      "learning_rate": 4.7803755440674916e-05,
+      "loss": 1.1507,
+      "step": 8485
+    },
+    {
+      "epoch": 0.515584178868704,
+      "grad_norm": 0.1537747085094452,
+      "learning_rate": 4.7794191947247414e-05,
+      "loss": 1.0779,
+      "step": 8486
+    },
+    {
+      "epoch": 0.5156449359013305,
+      "grad_norm": 0.17183583974838257,
+      "learning_rate": 4.778462853467425e-05,
+      "loss": 1.1285,
+      "step": 8487
+    },
+    {
+      "epoch": 0.5157056929339571,
+      "grad_norm": 0.11602478474378586,
+      "learning_rate": 4.777506520330593e-05,
+      "loss": 1.0038,
+      "step": 8488
+    },
+    {
+      "epoch": 0.5157664499665836,
+      "grad_norm": 0.22357726097106934,
+      "learning_rate": 4.7765501953493033e-05,
+      "loss": 1.156,
+      "step": 8489
+    },
+    {
+      "epoch": 0.5158272069992101,
+      "grad_norm": 0.1359436810016632,
+      "learning_rate": 4.7755938785586086e-05,
+      "loss": 1.0895,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5158879640318367,
+      "grad_norm": 0.14483585953712463,
+      "learning_rate": 4.7746375699935646e-05,
+      "loss": 1.0647,
+      "step": 8491
+    },
+    {
+      "epoch": 0.5159487210644632,
+      "grad_norm": 0.1440819501876831,
+      "learning_rate": 4.773681269689224e-05,
+      "loss": 1.0651,
+      "step": 8492
+    },
+    {
+      "epoch": 0.5160094780970897,
+      "grad_norm": 0.7805384397506714,
+      "learning_rate": 4.772724977680639e-05,
+      "loss": 1.2545,
+      "step": 8493
+    },
+    {
+      "epoch": 0.5160702351297163,
+      "grad_norm": 0.14209496974945068,
+      "learning_rate": 4.7717686940028675e-05,
+      "loss": 1.0452,
+      "step": 8494
+    },
+    {
+      "epoch": 0.5161309921623428,
+      "grad_norm": 0.13533027470111847,
+      "learning_rate": 4.770812418690954e-05,
+      "loss": 1.0473,
+      "step": 8495
+    },
+    {
+      "epoch": 0.5161917491949694,
+      "grad_norm": 0.4063666760921478,
+      "learning_rate": 4.769856151779959e-05,
+      "loss": 1.1717,
+      "step": 8496
+    },
+    {
+      "epoch": 0.5162525062275959,
+      "grad_norm": 0.16200560331344604,
+      "learning_rate": 4.768899893304929e-05,
+      "loss": 1.1014,
+      "step": 8497
+    },
+    {
+      "epoch": 0.5163132632602224,
+      "grad_norm": 2.216160774230957,
+      "learning_rate": 4.767943643300918e-05,
+      "loss": 1.0507,
+      "step": 8498
+    },
+    {
+      "epoch": 0.5163740202928488,
+      "grad_norm": 0.19600948691368103,
+      "learning_rate": 4.7669874018029785e-05,
+      "loss": 1.154,
+      "step": 8499
+    },
+    {
+      "epoch": 0.5164347773254754,
+      "grad_norm": 0.2035873681306839,
+      "learning_rate": 4.766031168846162e-05,
+      "loss": 1.1135,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5164955343581019,
+      "grad_norm": 0.1573156863451004,
+      "learning_rate": 4.765074944465517e-05,
+      "loss": 1.0793,
+      "step": 8501
+    },
+    {
+      "epoch": 0.5165562913907285,
+      "grad_norm": 0.17408792674541473,
+      "learning_rate": 4.764118728696096e-05,
+      "loss": 1.0975,
+      "step": 8502
+    },
+    {
+      "epoch": 0.516617048423355,
+      "grad_norm": 0.14286479353904724,
+      "learning_rate": 4.763162521572947e-05,
+      "loss": 0.995,
+      "step": 8503
+    },
+    {
+      "epoch": 0.5166778054559815,
+      "grad_norm": 0.15472820401191711,
+      "learning_rate": 4.7622063231311225e-05,
+      "loss": 1.0947,
+      "step": 8504
+    },
+    {
+      "epoch": 0.516738562488608,
+      "grad_norm": 0.2096824198961258,
+      "learning_rate": 4.761250133405671e-05,
+      "loss": 1.0966,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5167993195212346,
+      "grad_norm": 0.13473306596279144,
+      "learning_rate": 4.760293952431642e-05,
+      "loss": 1.0508,
+      "step": 8506
+    },
+    {
+      "epoch": 0.5168600765538611,
+      "grad_norm": 0.20115388929843903,
+      "learning_rate": 4.759337780244087e-05,
+      "loss": 1.0652,
+      "step": 8507
+    },
+    {
+      "epoch": 0.5169208335864877,
+      "grad_norm": 0.14727814495563507,
+      "learning_rate": 4.758381616878049e-05,
+      "loss": 1.1178,
+      "step": 8508
+    },
+    {
+      "epoch": 0.5169815906191142,
+      "grad_norm": 0.36924317479133606,
+      "learning_rate": 4.7574254623685806e-05,
+      "loss": 1.3663,
+      "step": 8509
+    },
+    {
+      "epoch": 0.5170423476517407,
+      "grad_norm": 0.3887315094470978,
+      "learning_rate": 4.756469316750729e-05,
+      "loss": 1.0526,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5171031046843673,
+      "grad_norm": 0.2359502613544464,
+      "learning_rate": 4.7555131800595395e-05,
+      "loss": 1.0543,
+      "step": 8511
+    },
+    {
+      "epoch": 0.5171638617169937,
+      "grad_norm": 0.19352594017982483,
+      "learning_rate": 4.754557052330063e-05,
+      "loss": 1.0995,
+      "step": 8512
+    },
+    {
+      "epoch": 0.5172246187496202,
+      "grad_norm": 0.15768180787563324,
+      "learning_rate": 4.753600933597346e-05,
+      "loss": 1.0783,
+      "step": 8513
+    },
+    {
+      "epoch": 0.5172853757822468,
+      "grad_norm": 0.12654650211334229,
+      "learning_rate": 4.752644823896436e-05,
+      "loss": 1.0843,
+      "step": 8514
+    },
+    {
+      "epoch": 0.5173461328148733,
+      "grad_norm": 0.12636877596378326,
+      "learning_rate": 4.751688723262376e-05,
+      "loss": 1.0595,
+      "step": 8515
+    },
+    {
+      "epoch": 0.5174068898474998,
+      "grad_norm": 0.14403897523880005,
+      "learning_rate": 4.750732631730212e-05,
+      "loss": 1.0554,
+      "step": 8516
+    },
+    {
+      "epoch": 0.5174676468801264,
+      "grad_norm": 0.15285925567150116,
+      "learning_rate": 4.749776549334994e-05,
+      "loss": 1.1014,
+      "step": 8517
+    },
+    {
+      "epoch": 0.5175284039127529,
+      "grad_norm": 0.1536620855331421,
+      "learning_rate": 4.7488204761117645e-05,
+      "loss": 1.0784,
+      "step": 8518
+    },
+    {
+      "epoch": 0.5175891609453794,
+      "grad_norm": 0.15283985435962677,
+      "learning_rate": 4.747864412095567e-05,
+      "loss": 1.0696,
+      "step": 8519
+    },
+    {
+      "epoch": 0.517649917978006,
+      "grad_norm": 0.12930315732955933,
+      "learning_rate": 4.7469083573214507e-05,
+      "loss": 1.0957,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5177106750106325,
+      "grad_norm": 0.18128079175949097,
+      "learning_rate": 4.745952311824458e-05,
+      "loss": 1.2746,
+      "step": 8521
+    },
+    {
+      "epoch": 0.517771432043259,
+      "grad_norm": 0.20560994744300842,
+      "learning_rate": 4.744996275639631e-05,
+      "loss": 1.1442,
+      "step": 8522
+    },
+    {
+      "epoch": 0.5178321890758856,
+      "grad_norm": 0.1893935650587082,
+      "learning_rate": 4.744040248802015e-05,
+      "loss": 1.1339,
+      "step": 8523
+    },
+    {
+      "epoch": 0.5178929461085121,
+      "grad_norm": 0.2803191542625427,
+      "learning_rate": 4.7430842313466526e-05,
+      "loss": 1.1397,
+      "step": 8524
+    },
+    {
+      "epoch": 0.5179537031411385,
+      "grad_norm": 0.14333286881446838,
+      "learning_rate": 4.742128223308587e-05,
+      "loss": 1.0146,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5180144601737651,
+      "grad_norm": 0.11949789524078369,
+      "learning_rate": 4.741172224722863e-05,
+      "loss": 1.0436,
+      "step": 8526
+    },
+    {
+      "epoch": 0.5180752172063916,
+      "grad_norm": 0.7034077048301697,
+      "learning_rate": 4.740216235624521e-05,
+      "loss": 1.1816,
+      "step": 8527
+    },
+    {
+      "epoch": 0.5181359742390181,
+      "grad_norm": 0.12296085804700851,
+      "learning_rate": 4.739260256048603e-05,
+      "loss": 1.0484,
+      "step": 8528
+    },
+    {
+      "epoch": 0.5181967312716447,
+      "grad_norm": 0.2202768325805664,
+      "learning_rate": 4.7383042860301486e-05,
+      "loss": 1.1118,
+      "step": 8529
+    },
+    {
+      "epoch": 0.5182574883042712,
+      "grad_norm": 0.13064312934875488,
+      "learning_rate": 4.737348325604203e-05,
+      "loss": 1.0228,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5183182453368977,
+      "grad_norm": 0.21044398844242096,
+      "learning_rate": 4.736392374805805e-05,
+      "loss": 1.2567,
+      "step": 8531
+    },
+    {
+      "epoch": 0.5183790023695243,
+      "grad_norm": 0.14555367827415466,
+      "learning_rate": 4.7354364336699944e-05,
+      "loss": 1.0687,
+      "step": 8532
+    },
+    {
+      "epoch": 0.5184397594021508,
+      "grad_norm": 0.17461712658405304,
+      "learning_rate": 4.734480502231813e-05,
+      "loss": 1.2131,
+      "step": 8533
+    },
+    {
+      "epoch": 0.5185005164347773,
+      "grad_norm": 0.6179195642471313,
+      "learning_rate": 4.733524580526302e-05,
+      "loss": 1.3834,
+      "step": 8534
+    },
+    {
+      "epoch": 0.5185612734674039,
+      "grad_norm": 0.15406516194343567,
+      "learning_rate": 4.732568668588497e-05,
+      "loss": 1.0398,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5186220305000304,
+      "grad_norm": 0.3225206434726715,
+      "learning_rate": 4.7316127664534406e-05,
+      "loss": 1.1191,
+      "step": 8536
+    },
+    {
+      "epoch": 0.518682787532657,
+      "grad_norm": 0.16026699542999268,
+      "learning_rate": 4.730656874156168e-05,
+      "loss": 1.0895,
+      "step": 8537
+    },
+    {
+      "epoch": 0.5187435445652834,
+      "grad_norm": 0.13030384480953217,
+      "learning_rate": 4.729700991731721e-05,
+      "loss": 0.9668,
+      "step": 8538
+    },
+    {
+      "epoch": 0.5188043015979099,
+      "grad_norm": 0.9277085661888123,
+      "learning_rate": 4.728745119215136e-05,
+      "loss": 1.0626,
+      "step": 8539
+    },
+    {
+      "epoch": 0.5188650586305364,
+      "grad_norm": 0.155452698469162,
+      "learning_rate": 4.7277892566414505e-05,
+      "loss": 1.0366,
+      "step": 8540
+    },
+    {
+      "epoch": 0.518925815663163,
+      "grad_norm": 0.19903744757175446,
+      "learning_rate": 4.726833404045705e-05,
+      "loss": 1.1415,
+      "step": 8541
+    },
+    {
+      "epoch": 0.5189865726957895,
+      "grad_norm": 0.13731133937835693,
+      "learning_rate": 4.725877561462932e-05,
+      "loss": 1.1048,
+      "step": 8542
+    },
+    {
+      "epoch": 0.519047329728416,
+      "grad_norm": 0.14890733361244202,
+      "learning_rate": 4.72492172892817e-05,
+      "loss": 1.0721,
+      "step": 8543
+    },
+    {
+      "epoch": 0.5191080867610426,
+      "grad_norm": 1.7920825481414795,
+      "learning_rate": 4.7239659064764555e-05,
+      "loss": 1.1146,
+      "step": 8544
+    },
+    {
+      "epoch": 0.5191688437936691,
+      "grad_norm": 2.9327590465545654,
+      "learning_rate": 4.723010094142824e-05,
+      "loss": 1.0877,
+      "step": 8545
+    },
+    {
+      "epoch": 0.5192296008262957,
+      "grad_norm": 0.27135559916496277,
+      "learning_rate": 4.722054291962311e-05,
+      "loss": 1.1143,
+      "step": 8546
+    },
+    {
+      "epoch": 0.5192903578589222,
+      "grad_norm": 0.25059840083122253,
+      "learning_rate": 4.7210984999699525e-05,
+      "loss": 1.0972,
+      "step": 8547
+    },
+    {
+      "epoch": 0.5193511148915487,
+      "grad_norm": 0.12423308193683624,
+      "learning_rate": 4.720142718200781e-05,
+      "loss": 1.0428,
+      "step": 8548
+    },
+    {
+      "epoch": 0.5194118719241753,
+      "grad_norm": 0.22610218822956085,
+      "learning_rate": 4.719186946689832e-05,
+      "loss": 1.0909,
+      "step": 8549
+    },
+    {
+      "epoch": 0.5194726289568018,
+      "grad_norm": 0.211140975356102,
+      "learning_rate": 4.7182311854721405e-05,
+      "loss": 1.1883,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5195333859894283,
+      "grad_norm": 0.13000492751598358,
+      "learning_rate": 4.7172754345827395e-05,
+      "loss": 1.0572,
+      "step": 8551
+    },
+    {
+      "epoch": 0.5195941430220548,
+      "grad_norm": 0.32262375950813293,
+      "learning_rate": 4.716319694056662e-05,
+      "loss": 1.1451,
+      "step": 8552
+    },
+    {
+      "epoch": 0.5196549000546813,
+      "grad_norm": 0.14049126207828522,
+      "learning_rate": 4.7153639639289405e-05,
+      "loss": 1.0213,
+      "step": 8553
+    },
+    {
+      "epoch": 0.5197156570873078,
+      "grad_norm": 0.23663240671157837,
+      "learning_rate": 4.7144082442346096e-05,
+      "loss": 1.1021,
+      "step": 8554
+    },
+    {
+      "epoch": 0.5197764141199344,
+      "grad_norm": 0.15460443496704102,
+      "learning_rate": 4.713452535008698e-05,
+      "loss": 1.0654,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5198371711525609,
+      "grad_norm": 0.14666052162647247,
+      "learning_rate": 4.71249683628624e-05,
+      "loss": 1.1444,
+      "step": 8556
+    },
+    {
+      "epoch": 0.5198979281851874,
+      "grad_norm": 0.13057032227516174,
+      "learning_rate": 4.711541148102265e-05,
+      "loss": 1.014,
+      "step": 8557
+    },
+    {
+      "epoch": 0.519958685217814,
+      "grad_norm": 0.32210874557495117,
+      "learning_rate": 4.710585470491806e-05,
+      "loss": 1.117,
+      "step": 8558
+    },
+    {
+      "epoch": 0.5200194422504405,
+      "grad_norm": 0.28795522451400757,
+      "learning_rate": 4.7096298034898916e-05,
+      "loss": 1.0833,
+      "step": 8559
+    },
+    {
+      "epoch": 0.520080199283067,
+      "grad_norm": 0.14907464385032654,
+      "learning_rate": 4.7086741471315535e-05,
+      "loss": 1.0744,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5201409563156936,
+      "grad_norm": 0.18301212787628174,
+      "learning_rate": 4.707718501451822e-05,
+      "loss": 1.1399,
+      "step": 8561
+    },
+    {
+      "epoch": 0.5202017133483201,
+      "grad_norm": 0.24117273092269897,
+      "learning_rate": 4.7067628664857234e-05,
+      "loss": 1.1392,
+      "step": 8562
+    },
+    {
+      "epoch": 0.5202624703809466,
+      "grad_norm": 0.47974660992622375,
+      "learning_rate": 4.705807242268289e-05,
+      "loss": 1.053,
+      "step": 8563
+    },
+    {
+      "epoch": 0.5203232274135732,
+      "grad_norm": 0.1926921308040619,
+      "learning_rate": 4.704851628834548e-05,
+      "loss": 1.1231,
+      "step": 8564
+    },
+    {
+      "epoch": 0.5203839844461996,
+      "grad_norm": 0.12540577352046967,
+      "learning_rate": 4.703896026219525e-05,
+      "loss": 1.0977,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5204447414788261,
+      "grad_norm": 0.8470072746276855,
+      "learning_rate": 4.702940434458253e-05,
+      "loss": 1.0427,
+      "step": 8566
+    },
+    {
+      "epoch": 0.5205054985114527,
+      "grad_norm": 0.2913875877857208,
+      "learning_rate": 4.701984853585757e-05,
+      "loss": 1.1211,
+      "step": 8567
+    },
+    {
+      "epoch": 0.5205662555440792,
+      "grad_norm": 0.18532782793045044,
+      "learning_rate": 4.7010292836370636e-05,
+      "loss": 1.2509,
+      "step": 8568
+    },
+    {
+      "epoch": 0.5206270125767057,
+      "grad_norm": 0.1036534383893013,
+      "learning_rate": 4.7000737246471985e-05,
+      "loss": 1.0159,
+      "step": 8569
+    },
+    {
+      "epoch": 0.5206877696093323,
+      "grad_norm": 0.258127897977829,
+      "learning_rate": 4.699118176651188e-05,
+      "loss": 1.0991,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5207485266419588,
+      "grad_norm": 0.14909648895263672,
+      "learning_rate": 4.69816263968406e-05,
+      "loss": 1.0391,
+      "step": 8571
+    },
+    {
+      "epoch": 0.5208092836745853,
+      "grad_norm": 0.18307402729988098,
+      "learning_rate": 4.6972071137808396e-05,
+      "loss": 1.0576,
+      "step": 8572
+    },
+    {
+      "epoch": 0.5208700407072119,
+      "grad_norm": 0.18096312880516052,
+      "learning_rate": 4.69625159897655e-05,
+      "loss": 1.1762,
+      "step": 8573
+    },
+    {
+      "epoch": 0.5209307977398384,
+      "grad_norm": 0.14437976479530334,
+      "learning_rate": 4.6952960953062194e-05,
+      "loss": 1.0641,
+      "step": 8574
+    },
+    {
+      "epoch": 0.520991554772465,
+      "grad_norm": 0.1318669617176056,
+      "learning_rate": 4.694340602804866e-05,
+      "loss": 1.0619,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5210523118050915,
+      "grad_norm": 0.3191777765750885,
+      "learning_rate": 4.693385121507518e-05,
+      "loss": 1.1428,
+      "step": 8576
+    },
+    {
+      "epoch": 0.521113068837718,
+      "grad_norm": 0.16258472204208374,
+      "learning_rate": 4.6924296514491976e-05,
+      "loss": 1.1067,
+      "step": 8577
+    },
+    {
+      "epoch": 0.5211738258703444,
+      "grad_norm": 0.19094589352607727,
+      "learning_rate": 4.691474192664927e-05,
+      "loss": 1.1186,
+      "step": 8578
+    },
+    {
+      "epoch": 0.521234582902971,
+      "grad_norm": 0.1727590262889862,
+      "learning_rate": 4.6905187451897306e-05,
+      "loss": 1.0734,
+      "step": 8579
+    },
+    {
+      "epoch": 0.5212953399355975,
+      "grad_norm": 0.20080527663230896,
+      "learning_rate": 4.6895633090586287e-05,
+      "loss": 1.0286,
+      "step": 8580
+    },
+    {
+      "epoch": 0.521356096968224,
+      "grad_norm": 0.3128844201564789,
+      "learning_rate": 4.688607884306645e-05,
+      "loss": 1.1375,
+      "step": 8581
+    },
+    {
+      "epoch": 0.5214168540008506,
+      "grad_norm": 0.30854642391204834,
+      "learning_rate": 4.6876524709687995e-05,
+      "loss": 1.036,
+      "step": 8582
+    },
+    {
+      "epoch": 0.5214776110334771,
+      "grad_norm": 0.11264657229185104,
+      "learning_rate": 4.686697069080111e-05,
+      "loss": 1.0065,
+      "step": 8583
+    },
+    {
+      "epoch": 0.5215383680661037,
+      "grad_norm": 0.22289027273654938,
+      "learning_rate": 4.6857416786756036e-05,
+      "loss": 1.1451,
+      "step": 8584
+    },
+    {
+      "epoch": 0.5215991250987302,
+      "grad_norm": 0.15465950965881348,
+      "learning_rate": 4.684786299790295e-05,
+      "loss": 1.0474,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5216598821313567,
+      "grad_norm": 0.21442024409770966,
+      "learning_rate": 4.683830932459205e-05,
+      "loss": 1.1101,
+      "step": 8586
+    },
+    {
+      "epoch": 0.5217206391639833,
+      "grad_norm": 0.16656796634197235,
+      "learning_rate": 4.682875576717354e-05,
+      "loss": 1.1451,
+      "step": 8587
+    },
+    {
+      "epoch": 0.5217813961966098,
+      "grad_norm": 0.40059763193130493,
+      "learning_rate": 4.681920232599761e-05,
+      "loss": 1.0725,
+      "step": 8588
+    },
+    {
+      "epoch": 0.5218421532292363,
+      "grad_norm": 0.24076776206493378,
+      "learning_rate": 4.680964900141443e-05,
+      "loss": 1.0504,
+      "step": 8589
+    },
+    {
+      "epoch": 0.5219029102618629,
+      "grad_norm": 0.3196810781955719,
+      "learning_rate": 4.680009579377418e-05,
+      "loss": 1.0874,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5219636672944893,
+      "grad_norm": 0.1243172362446785,
+      "learning_rate": 4.679054270342703e-05,
+      "loss": 1.0737,
+      "step": 8591
+    },
+    {
+      "epoch": 0.5220244243271158,
+      "grad_norm": 0.15369655191898346,
+      "learning_rate": 4.6780989730723155e-05,
+      "loss": 1.0562,
+      "step": 8592
+    },
+    {
+      "epoch": 0.5220851813597424,
+      "grad_norm": 0.15701650083065033,
+      "learning_rate": 4.677143687601274e-05,
+      "loss": 1.0065,
+      "step": 8593
+    },
+    {
+      "epoch": 0.5221459383923689,
+      "grad_norm": 0.2327444851398468,
+      "learning_rate": 4.676188413964593e-05,
+      "loss": 1.0102,
+      "step": 8594
+    },
+    {
+      "epoch": 0.5222066954249954,
+      "grad_norm": 0.20564907789230347,
+      "learning_rate": 4.675233152197287e-05,
+      "loss": 1.0518,
+      "step": 8595
+    },
+    {
+      "epoch": 0.522267452457622,
+      "grad_norm": 0.14837133884429932,
+      "learning_rate": 4.6742779023343725e-05,
+      "loss": 1.0464,
+      "step": 8596
+    },
+    {
+      "epoch": 0.5223282094902485,
+      "grad_norm": 0.13209976255893707,
+      "learning_rate": 4.673322664410865e-05,
+      "loss": 1.0408,
+      "step": 8597
+    },
+    {
+      "epoch": 0.522388966522875,
+      "grad_norm": 0.1632174402475357,
+      "learning_rate": 4.672367438461778e-05,
+      "loss": 1.1276,
+      "step": 8598
+    },
+    {
+      "epoch": 0.5224497235555016,
+      "grad_norm": 0.247183695435524,
+      "learning_rate": 4.671412224522126e-05,
+      "loss": 1.0894,
+      "step": 8599
+    },
+    {
+      "epoch": 0.5225104805881281,
+      "grad_norm": 0.1803322732448578,
+      "learning_rate": 4.670457022626922e-05,
+      "loss": 1.0554,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5225712376207546,
+      "grad_norm": 0.43978366255760193,
+      "learning_rate": 4.6695018328111814e-05,
+      "loss": 1.1711,
+      "step": 8601
+    },
+    {
+      "epoch": 0.5226319946533812,
+      "grad_norm": 0.18564847111701965,
+      "learning_rate": 4.668546655109914e-05,
+      "loss": 1.1763,
+      "step": 8602
+    },
+    {
+      "epoch": 0.5226927516860077,
+      "grad_norm": 0.16568569839000702,
+      "learning_rate": 4.667591489558133e-05,
+      "loss": 1.0638,
+      "step": 8603
+    },
+    {
+      "epoch": 0.5227535087186341,
+      "grad_norm": 0.14004436135292053,
+      "learning_rate": 4.666636336190849e-05,
+      "loss": 1.0573,
+      "step": 8604
+    },
+    {
+      "epoch": 0.5228142657512607,
+      "grad_norm": 0.1846337616443634,
+      "learning_rate": 4.6656811950430754e-05,
+      "loss": 1.044,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5228750227838872,
+      "grad_norm": 0.12017328292131424,
+      "learning_rate": 4.664726066149822e-05,
+      "loss": 1.0496,
+      "step": 8606
+    },
+    {
+      "epoch": 0.5229357798165137,
+      "grad_norm": 0.24024394154548645,
+      "learning_rate": 4.663770949546099e-05,
+      "loss": 1.1732,
+      "step": 8607
+    },
+    {
+      "epoch": 0.5229965368491403,
+      "grad_norm": 0.22270292043685913,
+      "learning_rate": 4.66281584526692e-05,
+      "loss": 1.2911,
+      "step": 8608
+    },
+    {
+      "epoch": 0.5230572938817668,
+      "grad_norm": 0.15339328348636627,
+      "learning_rate": 4.661860753347288e-05,
+      "loss": 1.0564,
+      "step": 8609
+    },
+    {
+      "epoch": 0.5231180509143933,
+      "grad_norm": 0.2351030707359314,
+      "learning_rate": 4.660905673822217e-05,
+      "loss": 1.0272,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5231788079470199,
+      "grad_norm": 0.2053232491016388,
+      "learning_rate": 4.659950606726713e-05,
+      "loss": 1.1677,
+      "step": 8611
+    },
+    {
+      "epoch": 0.5232395649796464,
+      "grad_norm": 0.14381182193756104,
+      "learning_rate": 4.658995552095785e-05,
+      "loss": 1.0556,
+      "step": 8612
+    },
+    {
+      "epoch": 0.523300322012273,
+      "grad_norm": 0.867560863494873,
+      "learning_rate": 4.658040509964441e-05,
+      "loss": 1.1115,
+      "step": 8613
+    },
+    {
+      "epoch": 0.5233610790448995,
+      "grad_norm": 0.22009895741939545,
+      "learning_rate": 4.657085480367691e-05,
+      "loss": 1.0878,
+      "step": 8614
+    },
+    {
+      "epoch": 0.523421836077526,
+      "grad_norm": 0.12190107256174088,
+      "learning_rate": 4.6561304633405365e-05,
+      "loss": 1.0959,
+      "step": 8615
+    },
+    {
+      "epoch": 0.5234825931101525,
+      "grad_norm": 0.17178240418434143,
+      "learning_rate": 4.6551754589179867e-05,
+      "loss": 1.0113,
+      "step": 8616
+    },
+    {
+      "epoch": 0.523543350142779,
+      "grad_norm": 0.18270833790302277,
+      "learning_rate": 4.654220467135047e-05,
+      "loss": 1.1768,
+      "step": 8617
+    },
+    {
+      "epoch": 0.5236041071754055,
+      "grad_norm": 0.21078373491764069,
+      "learning_rate": 4.653265488026723e-05,
+      "loss": 1.2156,
+      "step": 8618
+    },
+    {
+      "epoch": 0.523664864208032,
+      "grad_norm": 0.16738082468509674,
+      "learning_rate": 4.6523105216280196e-05,
+      "loss": 1.1635,
+      "step": 8619
+    },
+    {
+      "epoch": 0.5237256212406586,
+      "grad_norm": 0.15675511956214905,
+      "learning_rate": 4.65135556797394e-05,
+      "loss": 1.0399,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5237863782732851,
+      "grad_norm": 0.12584108114242554,
+      "learning_rate": 4.6504006270994934e-05,
+      "loss": 1.026,
+      "step": 8621
+    },
+    {
+      "epoch": 0.5238471353059116,
+      "grad_norm": 0.13266484439373016,
+      "learning_rate": 4.649445699039675e-05,
+      "loss": 1.052,
+      "step": 8622
+    },
+    {
+      "epoch": 0.5239078923385382,
+      "grad_norm": 0.1908525973558426,
+      "learning_rate": 4.648490783829495e-05,
+      "loss": 1.0591,
+      "step": 8623
+    },
+    {
+      "epoch": 0.5239686493711647,
+      "grad_norm": 0.17659196257591248,
+      "learning_rate": 4.6475358815039514e-05,
+      "loss": 1.0764,
+      "step": 8624
+    },
+    {
+      "epoch": 0.5240294064037913,
+      "grad_norm": 0.14164158701896667,
+      "learning_rate": 4.6465809920980485e-05,
+      "loss": 1.1083,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5240901634364178,
+      "grad_norm": 0.15035857260227203,
+      "learning_rate": 4.645626115646788e-05,
+      "loss": 0.9924,
+      "step": 8626
+    },
+    {
+      "epoch": 0.5241509204690443,
+      "grad_norm": 0.16510003805160522,
+      "learning_rate": 4.644671252185171e-05,
+      "loss": 1.1492,
+      "step": 8627
+    },
+    {
+      "epoch": 0.5242116775016709,
+      "grad_norm": 0.19746075570583344,
+      "learning_rate": 4.643716401748199e-05,
+      "loss": 1.068,
+      "step": 8628
+    },
+    {
+      "epoch": 0.5242724345342974,
+      "grad_norm": 0.13315987586975098,
+      "learning_rate": 4.64276156437087e-05,
+      "loss": 1.0826,
+      "step": 8629
+    },
+    {
+      "epoch": 0.5243331915669238,
+      "grad_norm": 0.1473248302936554,
+      "learning_rate": 4.641806740088184e-05,
+      "loss": 1.064,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5243939485995504,
+      "grad_norm": 0.21956190466880798,
+      "learning_rate": 4.640851928935142e-05,
+      "loss": 1.0046,
+      "step": 8631
+    },
+    {
+      "epoch": 0.5244547056321769,
+      "grad_norm": 0.12775710225105286,
+      "learning_rate": 4.6398971309467426e-05,
+      "loss": 1.0448,
+      "step": 8632
+    },
+    {
+      "epoch": 0.5245154626648034,
+      "grad_norm": 0.13934054970741272,
+      "learning_rate": 4.638942346157982e-05,
+      "loss": 1.0606,
+      "step": 8633
+    },
+    {
+      "epoch": 0.52457621969743,
+      "grad_norm": 0.85642409324646,
+      "learning_rate": 4.6379875746038625e-05,
+      "loss": 1.2726,
+      "step": 8634
+    },
+    {
+      "epoch": 0.5246369767300565,
+      "grad_norm": 0.21907411515712738,
+      "learning_rate": 4.6370328163193754e-05,
+      "loss": 1.0919,
+      "step": 8635
+    },
+    {
+      "epoch": 0.524697733762683,
+      "grad_norm": 0.20063483715057373,
+      "learning_rate": 4.6360780713395214e-05,
+      "loss": 1.1787,
+      "step": 8636
+    },
+    {
+      "epoch": 0.5247584907953096,
+      "grad_norm": 0.15670740604400635,
+      "learning_rate": 4.6351233396992966e-05,
+      "loss": 1.1677,
+      "step": 8637
+    },
+    {
+      "epoch": 0.5248192478279361,
+      "grad_norm": 0.2139139324426651,
+      "learning_rate": 4.634168621433694e-05,
+      "loss": 1.0672,
+      "step": 8638
+    },
+    {
+      "epoch": 0.5248800048605626,
+      "grad_norm": 0.1647370159626007,
+      "learning_rate": 4.633213916577714e-05,
+      "loss": 1.0713,
+      "step": 8639
+    },
+    {
+      "epoch": 0.5249407618931892,
+      "grad_norm": 0.1305478811264038,
+      "learning_rate": 4.632259225166348e-05,
+      "loss": 1.033,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5250015189258157,
+      "grad_norm": 0.15238302946090698,
+      "learning_rate": 4.631304547234592e-05,
+      "loss": 1.1097,
+      "step": 8641
+    },
+    {
+      "epoch": 0.5250622759584422,
+      "grad_norm": 0.20471568405628204,
+      "learning_rate": 4.630349882817437e-05,
+      "loss": 1.1534,
+      "step": 8642
+    },
+    {
+      "epoch": 0.5251230329910687,
+      "grad_norm": 0.18734177947044373,
+      "learning_rate": 4.62939523194988e-05,
+      "loss": 1.1485,
+      "step": 8643
+    },
+    {
+      "epoch": 0.5251837900236952,
+      "grad_norm": 0.14287249743938446,
+      "learning_rate": 4.628440594666912e-05,
+      "loss": 1.1792,
+      "step": 8644
+    },
+    {
+      "epoch": 0.5252445470563217,
+      "grad_norm": 0.1861511468887329,
+      "learning_rate": 4.627485971003527e-05,
+      "loss": 0.9983,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5253053040889483,
+      "grad_norm": 0.1467745006084442,
+      "learning_rate": 4.626531360994713e-05,
+      "loss": 1.0541,
+      "step": 8646
+    },
+    {
+      "epoch": 0.5253660611215748,
+      "grad_norm": 0.7447397112846375,
+      "learning_rate": 4.625576764675466e-05,
+      "loss": 1.1406,
+      "step": 8647
+    },
+    {
+      "epoch": 0.5254268181542013,
+      "grad_norm": 0.20171573758125305,
+      "learning_rate": 4.624622182080776e-05,
+      "loss": 1.0727,
+      "step": 8648
+    },
+    {
+      "epoch": 0.5254875751868279,
+      "grad_norm": 0.24454587697982788,
+      "learning_rate": 4.623667613245632e-05,
+      "loss": 1.097,
+      "step": 8649
+    },
+    {
+      "epoch": 0.5255483322194544,
+      "grad_norm": 0.33716270327568054,
+      "learning_rate": 4.622713058205023e-05,
+      "loss": 1.0666,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5256090892520809,
+      "grad_norm": 3.97310733795166,
+      "learning_rate": 4.62175851699394e-05,
+      "loss": 1.0461,
+      "step": 8651
+    },
+    {
+      "epoch": 0.5256698462847075,
+      "grad_norm": 0.1582850217819214,
+      "learning_rate": 4.620803989647373e-05,
+      "loss": 1.1161,
+      "step": 8652
+    },
+    {
+      "epoch": 0.525730603317334,
+      "grad_norm": 0.2617885172367096,
+      "learning_rate": 4.6198494762003076e-05,
+      "loss": 1.0628,
+      "step": 8653
+    },
+    {
+      "epoch": 0.5257913603499605,
+      "grad_norm": 0.17769746482372284,
+      "learning_rate": 4.618894976687733e-05,
+      "loss": 1.1235,
+      "step": 8654
+    },
+    {
+      "epoch": 0.5258521173825871,
+      "grad_norm": 0.33753904700279236,
+      "learning_rate": 4.6179404911446386e-05,
+      "loss": 1.1388,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5259128744152136,
+      "grad_norm": 0.18174488842487335,
+      "learning_rate": 4.6169860196060084e-05,
+      "loss": 1.1416,
+      "step": 8656
+    },
+    {
+      "epoch": 0.52597363144784,
+      "grad_norm": 0.1649746149778366,
+      "learning_rate": 4.6160315621068294e-05,
+      "loss": 1.0226,
+      "step": 8657
+    },
+    {
+      "epoch": 0.5260343884804666,
+      "grad_norm": 0.1672050803899765,
+      "learning_rate": 4.615077118682087e-05,
+      "loss": 1.0787,
+      "step": 8658
+    },
+    {
+      "epoch": 0.5260951455130931,
+      "grad_norm": 0.17580434679985046,
+      "learning_rate": 4.6141226893667684e-05,
+      "loss": 1.0415,
+      "step": 8659
+    },
+    {
+      "epoch": 0.5261559025457196,
+      "grad_norm": 0.5573168396949768,
+      "learning_rate": 4.6131682741958566e-05,
+      "loss": 1.092,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5262166595783462,
+      "grad_norm": 0.3986499607563019,
+      "learning_rate": 4.612213873204338e-05,
+      "loss": 1.1489,
+      "step": 8661
+    },
+    {
+      "epoch": 0.5262774166109727,
+      "grad_norm": 0.21384546160697937,
+      "learning_rate": 4.611259486427194e-05,
+      "loss": 1.098,
+      "step": 8662
+    },
+    {
+      "epoch": 0.5263381736435992,
+      "grad_norm": 0.17524658143520355,
+      "learning_rate": 4.610305113899407e-05,
+      "loss": 1.0754,
+      "step": 8663
+    },
+    {
+      "epoch": 0.5263989306762258,
+      "grad_norm": 0.1727045774459839,
+      "learning_rate": 4.609350755655963e-05,
+      "loss": 1.0928,
+      "step": 8664
+    },
+    {
+      "epoch": 0.5264596877088523,
+      "grad_norm": 0.1485423445701599,
+      "learning_rate": 4.608396411731842e-05,
+      "loss": 1.0425,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5265204447414789,
+      "grad_norm": 0.2882673442363739,
+      "learning_rate": 4.607442082162026e-05,
+      "loss": 1.2019,
+      "step": 8666
+    },
+    {
+      "epoch": 0.5265812017741054,
+      "grad_norm": 0.1373392641544342,
+      "learning_rate": 4.606487766981496e-05,
+      "loss": 1.0457,
+      "step": 8667
+    },
+    {
+      "epoch": 0.5266419588067319,
+      "grad_norm": 0.5117956399917603,
+      "learning_rate": 4.605533466225235e-05,
+      "loss": 1.2386,
+      "step": 8668
+    },
+    {
+      "epoch": 0.5267027158393585,
+      "grad_norm": 0.7526136636734009,
+      "learning_rate": 4.60457917992822e-05,
+      "loss": 1.1093,
+      "step": 8669
+    },
+    {
+      "epoch": 0.5267634728719849,
+      "grad_norm": 0.31235063076019287,
+      "learning_rate": 4.603624908125432e-05,
+      "loss": 1.1961,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5268242299046114,
+      "grad_norm": 0.4177379906177521,
+      "learning_rate": 4.6026706508518474e-05,
+      "loss": 1.1936,
+      "step": 8671
+    },
+    {
+      "epoch": 0.526884986937238,
+      "grad_norm": 0.15043529868125916,
+      "learning_rate": 4.601716408142448e-05,
+      "loss": 1.0563,
+      "step": 8672
+    },
+    {
+      "epoch": 0.5269457439698645,
+      "grad_norm": 0.17881150543689728,
+      "learning_rate": 4.600762180032212e-05,
+      "loss": 1.0621,
+      "step": 8673
+    },
+    {
+      "epoch": 0.527006501002491,
+      "grad_norm": 0.2779545783996582,
+      "learning_rate": 4.599807966556113e-05,
+      "loss": 1.0724,
+      "step": 8674
+    },
+    {
+      "epoch": 0.5270672580351176,
+      "grad_norm": 1.8141210079193115,
+      "learning_rate": 4.598853767749135e-05,
+      "loss": 1.0501,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5271280150677441,
+      "grad_norm": 0.20107562839984894,
+      "learning_rate": 4.597899583646245e-05,
+      "loss": 1.0175,
+      "step": 8676
+    },
+    {
+      "epoch": 0.5271887721003706,
+      "grad_norm": 0.19702781736850739,
+      "learning_rate": 4.596945414282425e-05,
+      "loss": 1.1646,
+      "step": 8677
+    },
+    {
+      "epoch": 0.5272495291329972,
+      "grad_norm": 0.3360294997692108,
+      "learning_rate": 4.59599125969265e-05,
+      "loss": 1.2875,
+      "step": 8678
+    },
+    {
+      "epoch": 0.5273102861656237,
+      "grad_norm": 0.2003525346517563,
+      "learning_rate": 4.5950371199118906e-05,
+      "loss": 1.0849,
+      "step": 8679
+    },
+    {
+      "epoch": 0.5273710431982502,
+      "grad_norm": 0.1472960114479065,
+      "learning_rate": 4.5940829949751255e-05,
+      "loss": 0.9855,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5274318002308768,
+      "grad_norm": 0.19163544476032257,
+      "learning_rate": 4.593128884917328e-05,
+      "loss": 1.0969,
+      "step": 8681
+    },
+    {
+      "epoch": 0.5274925572635033,
+      "grad_norm": 0.16724400222301483,
+      "learning_rate": 4.592174789773469e-05,
+      "loss": 1.0859,
+      "step": 8682
+    },
+    {
+      "epoch": 0.5275533142961297,
+      "grad_norm": 15.174552917480469,
+      "learning_rate": 4.591220709578521e-05,
+      "loss": 1.0845,
+      "step": 8683
+    },
+    {
+      "epoch": 0.5276140713287563,
+      "grad_norm": 0.14118710160255432,
+      "learning_rate": 4.590266644367457e-05,
+      "loss": 1.0497,
+      "step": 8684
+    },
+    {
+      "epoch": 0.5276748283613828,
+      "grad_norm": 0.2985968291759491,
+      "learning_rate": 4.589312594175249e-05,
+      "loss": 1.1774,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5277355853940093,
+      "grad_norm": 0.4103105962276459,
+      "learning_rate": 4.588358559036867e-05,
+      "loss": 1.1844,
+      "step": 8686
+    },
+    {
+      "epoch": 0.5277963424266359,
+      "grad_norm": 0.417360782623291,
+      "learning_rate": 4.5874045389872804e-05,
+      "loss": 1.2113,
+      "step": 8687
+    },
+    {
+      "epoch": 0.5278570994592624,
+      "grad_norm": 0.24194343388080597,
+      "learning_rate": 4.5864505340614636e-05,
+      "loss": 1.3424,
+      "step": 8688
+    },
+    {
+      "epoch": 0.5279178564918889,
+      "grad_norm": 0.1654786765575409,
+      "learning_rate": 4.5854965442943796e-05,
+      "loss": 1.088,
+      "step": 8689
+    },
+    {
+      "epoch": 0.5279786135245155,
+      "grad_norm": 0.1660112887620926,
+      "learning_rate": 4.5845425697210005e-05,
+      "loss": 1.0636,
+      "step": 8690
+    },
+    {
+      "epoch": 0.528039370557142,
+      "grad_norm": 1.0987192392349243,
+      "learning_rate": 4.583588610376294e-05,
+      "loss": 1.2099,
+      "step": 8691
+    },
+    {
+      "epoch": 0.5281001275897685,
+      "grad_norm": 1.8631782531738281,
+      "learning_rate": 4.582634666295226e-05,
+      "loss": 1.1296,
+      "step": 8692
+    },
+    {
+      "epoch": 0.5281608846223951,
+      "grad_norm": 0.24905526638031006,
+      "learning_rate": 4.581680737512767e-05,
+      "loss": 1.0537,
+      "step": 8693
+    },
+    {
+      "epoch": 0.5282216416550216,
+      "grad_norm": 0.2177335023880005,
+      "learning_rate": 4.58072682406388e-05,
+      "loss": 1.0183,
+      "step": 8694
+    },
+    {
+      "epoch": 0.5282823986876481,
+      "grad_norm": 0.26563701033592224,
+      "learning_rate": 4.579772925983535e-05,
+      "loss": 1.1127,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5283431557202746,
+      "grad_norm": 0.4449535310268402,
+      "learning_rate": 4.578819043306692e-05,
+      "loss": 1.144,
+      "step": 8696
+    },
+    {
+      "epoch": 0.5284039127529011,
+      "grad_norm": 0.22405296564102173,
+      "learning_rate": 4.5778651760683185e-05,
+      "loss": 1.0336,
+      "step": 8697
+    },
+    {
+      "epoch": 0.5284646697855276,
+      "grad_norm": 0.5492637157440186,
+      "learning_rate": 4.576911324303379e-05,
+      "loss": 1.2332,
+      "step": 8698
+    },
+    {
+      "epoch": 0.5285254268181542,
+      "grad_norm": 0.6914724111557007,
+      "learning_rate": 4.575957488046837e-05,
+      "loss": 1.1727,
+      "step": 8699
+    },
+    {
+      "epoch": 0.5285861838507807,
+      "grad_norm": 0.25794705748558044,
+      "learning_rate": 4.5750036673336535e-05,
+      "loss": 1.0396,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5286469408834072,
+      "grad_norm": 0.2574668824672699,
+      "learning_rate": 4.574049862198797e-05,
+      "loss": 1.0947,
+      "step": 8701
+    },
+    {
+      "epoch": 0.5287076979160338,
+      "grad_norm": 0.4927690625190735,
+      "learning_rate": 4.573096072677221e-05,
+      "loss": 1.0905,
+      "step": 8702
+    },
+    {
+      "epoch": 0.5287684549486603,
+      "grad_norm": 0.25256097316741943,
+      "learning_rate": 4.572142298803893e-05,
+      "loss": 1.017,
+      "step": 8703
+    },
+    {
+      "epoch": 0.5288292119812869,
+      "grad_norm": 0.25896692276000977,
+      "learning_rate": 4.571188540613771e-05,
+      "loss": 1.1333,
+      "step": 8704
+    },
+    {
+      "epoch": 0.5288899690139134,
+      "grad_norm": 0.2356635481119156,
+      "learning_rate": 4.5702347981418146e-05,
+      "loss": 1.118,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5289507260465399,
+      "grad_norm": 0.27731621265411377,
+      "learning_rate": 4.569281071422987e-05,
+      "loss": 1.094,
+      "step": 8706
+    },
+    {
+      "epoch": 0.5290114830791665,
+      "grad_norm": 0.23638027906417847,
+      "learning_rate": 4.568327360492244e-05,
+      "loss": 1.1346,
+      "step": 8707
+    },
+    {
+      "epoch": 0.529072240111793,
+      "grad_norm": 0.1733100712299347,
+      "learning_rate": 4.567373665384546e-05,
+      "loss": 1.0479,
+      "step": 8708
+    },
+    {
+      "epoch": 0.5291329971444194,
+      "grad_norm": 0.18118803203105927,
+      "learning_rate": 4.5664199861348495e-05,
+      "loss": 1.0569,
+      "step": 8709
+    },
+    {
+      "epoch": 0.529193754177046,
+      "grad_norm": 0.20814651250839233,
+      "learning_rate": 4.5654663227781106e-05,
+      "loss": 1.0691,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5292545112096725,
+      "grad_norm": 0.27674707770347595,
+      "learning_rate": 4.564512675349289e-05,
+      "loss": 1.0877,
+      "step": 8711
+    },
+    {
+      "epoch": 0.529315268242299,
+      "grad_norm": 0.2000836879014969,
+      "learning_rate": 4.563559043883339e-05,
+      "loss": 1.1641,
+      "step": 8712
+    },
+    {
+      "epoch": 0.5293760252749256,
+      "grad_norm": 0.30237486958503723,
+      "learning_rate": 4.5626054284152165e-05,
+      "loss": 1.187,
+      "step": 8713
+    },
+    {
+      "epoch": 0.5294367823075521,
+      "grad_norm": 0.21813370287418365,
+      "learning_rate": 4.561651828979877e-05,
+      "loss": 1.1632,
+      "step": 8714
+    },
+    {
+      "epoch": 0.5294975393401786,
+      "grad_norm": 0.13433311879634857,
+      "learning_rate": 4.560698245612275e-05,
+      "loss": 1.0518,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5295582963728052,
+      "grad_norm": 0.14734883606433868,
+      "learning_rate": 4.559744678347363e-05,
+      "loss": 1.0382,
+      "step": 8716
+    },
+    {
+      "epoch": 0.5296190534054317,
+      "grad_norm": 0.11894255131483078,
+      "learning_rate": 4.558791127220096e-05,
+      "loss": 1.007,
+      "step": 8717
+    },
+    {
+      "epoch": 0.5296798104380582,
+      "grad_norm": 0.7833021879196167,
+      "learning_rate": 4.5578375922654235e-05,
+      "loss": 1.109,
+      "step": 8718
+    },
+    {
+      "epoch": 0.5297405674706848,
+      "grad_norm": 0.14893975853919983,
+      "learning_rate": 4.5568840735183005e-05,
+      "loss": 1.0708,
+      "step": 8719
+    },
+    {
+      "epoch": 0.5298013245033113,
+      "grad_norm": 0.139544278383255,
+      "learning_rate": 4.555930571013678e-05,
+      "loss": 1.1002,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5298620815359378,
+      "grad_norm": 0.19169317185878754,
+      "learning_rate": 4.554977084786507e-05,
+      "loss": 1.0939,
+      "step": 8721
+    },
+    {
+      "epoch": 0.5299228385685643,
+      "grad_norm": 0.14398300647735596,
+      "learning_rate": 4.5540236148717356e-05,
+      "loss": 1.0891,
+      "step": 8722
+    },
+    {
+      "epoch": 0.5299835956011908,
+      "grad_norm": 0.1250181943178177,
+      "learning_rate": 4.553070161304314e-05,
+      "loss": 1.0149,
+      "step": 8723
+    },
+    {
+      "epoch": 0.5300443526338173,
+      "grad_norm": 0.28524330258369446,
+      "learning_rate": 4.552116724119193e-05,
+      "loss": 1.0837,
+      "step": 8724
+    },
+    {
+      "epoch": 0.5301051096664439,
+      "grad_norm": 0.13696500658988953,
+      "learning_rate": 4.55116330335132e-05,
+      "loss": 1.045,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5301658666990704,
+      "grad_norm": 0.1360851377248764,
+      "learning_rate": 4.5502098990356425e-05,
+      "loss": 1.0537,
+      "step": 8726
+    },
+    {
+      "epoch": 0.5302266237316969,
+      "grad_norm": 0.1495683640241623,
+      "learning_rate": 4.549256511207108e-05,
+      "loss": 1.0636,
+      "step": 8727
+    },
+    {
+      "epoch": 0.5302873807643235,
+      "grad_norm": 0.15629592537879944,
+      "learning_rate": 4.548303139900666e-05,
+      "loss": 1.0077,
+      "step": 8728
+    },
+    {
+      "epoch": 0.53034813779695,
+      "grad_norm": 0.14614929258823395,
+      "learning_rate": 4.547349785151257e-05,
+      "loss": 1.0294,
+      "step": 8729
+    },
+    {
+      "epoch": 0.5304088948295765,
+      "grad_norm": 0.1470760852098465,
+      "learning_rate": 4.546396446993829e-05,
+      "loss": 1.0315,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5304696518622031,
+      "grad_norm": 0.15036186575889587,
+      "learning_rate": 4.545443125463327e-05,
+      "loss": 1.0804,
+      "step": 8731
+    },
+    {
+      "epoch": 0.5305304088948296,
+      "grad_norm": 0.5035548806190491,
+      "learning_rate": 4.544489820594696e-05,
+      "loss": 1.1199,
+      "step": 8732
+    },
+    {
+      "epoch": 0.5305911659274561,
+      "grad_norm": 0.23608504235744476,
+      "learning_rate": 4.543536532422878e-05,
+      "loss": 1.158,
+      "step": 8733
+    },
+    {
+      "epoch": 0.5306519229600827,
+      "grad_norm": 0.2268620878458023,
+      "learning_rate": 4.5425832609828156e-05,
+      "loss": 1.1468,
+      "step": 8734
+    },
+    {
+      "epoch": 0.5307126799927091,
+      "grad_norm": 0.14344771206378937,
+      "learning_rate": 4.541630006309456e-05,
+      "loss": 1.0013,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5307734370253356,
+      "grad_norm": 0.1313079297542572,
+      "learning_rate": 4.5406767684377336e-05,
+      "loss": 1.0578,
+      "step": 8736
+    },
+    {
+      "epoch": 0.5308341940579622,
+      "grad_norm": 0.17482100427150726,
+      "learning_rate": 4.539723547402594e-05,
+      "loss": 0.9898,
+      "step": 8737
+    },
+    {
+      "epoch": 0.5308949510905887,
+      "grad_norm": 0.1623125970363617,
+      "learning_rate": 4.5387703432389775e-05,
+      "loss": 1.1636,
+      "step": 8738
+    },
+    {
+      "epoch": 0.5309557081232152,
+      "grad_norm": 0.1889430284500122,
+      "learning_rate": 4.537817155981821e-05,
+      "loss": 1.1606,
+      "step": 8739
+    },
+    {
+      "epoch": 0.5310164651558418,
+      "grad_norm": 0.20790928602218628,
+      "learning_rate": 4.5368639856660674e-05,
+      "loss": 1.2408,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5310772221884683,
+      "grad_norm": 0.1391948163509369,
+      "learning_rate": 4.535910832326654e-05,
+      "loss": 1.1023,
+      "step": 8741
+    },
+    {
+      "epoch": 0.5311379792210948,
+      "grad_norm": 0.13725924491882324,
+      "learning_rate": 4.53495769599852e-05,
+      "loss": 1.055,
+      "step": 8742
+    },
+    {
+      "epoch": 0.5311987362537214,
+      "grad_norm": 0.19517844915390015,
+      "learning_rate": 4.534004576716599e-05,
+      "loss": 1.1158,
+      "step": 8743
+    },
+    {
+      "epoch": 0.5312594932863479,
+      "grad_norm": 5.072870254516602,
+      "learning_rate": 4.5330514745158324e-05,
+      "loss": 1.1402,
+      "step": 8744
+    },
+    {
+      "epoch": 0.5313202503189745,
+      "grad_norm": 0.24879388511180878,
+      "learning_rate": 4.532098389431154e-05,
+      "loss": 1.1199,
+      "step": 8745
+    },
+    {
+      "epoch": 0.531381007351601,
+      "grad_norm": 0.17267988622188568,
+      "learning_rate": 4.531145321497498e-05,
+      "loss": 1.0712,
+      "step": 8746
+    },
+    {
+      "epoch": 0.5314417643842275,
+      "grad_norm": 0.14354780316352844,
+      "learning_rate": 4.530192270749801e-05,
+      "loss": 1.0917,
+      "step": 8747
+    },
+    {
+      "epoch": 0.531502521416854,
+      "grad_norm": 0.12416389584541321,
+      "learning_rate": 4.5292392372229996e-05,
+      "loss": 1.0396,
+      "step": 8748
+    },
+    {
+      "epoch": 0.5315632784494805,
+      "grad_norm": 0.17430990934371948,
+      "learning_rate": 4.5282862209520236e-05,
+      "loss": 1.0602,
+      "step": 8749
+    },
+    {
+      "epoch": 0.531624035482107,
+      "grad_norm": 0.14102205634117126,
+      "learning_rate": 4.5273332219718076e-05,
+      "loss": 1.0283,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5316847925147336,
+      "grad_norm": 0.1749001443386078,
+      "learning_rate": 4.526380240317282e-05,
+      "loss": 1.1014,
+      "step": 8751
+    },
+    {
+      "epoch": 0.5317455495473601,
+      "grad_norm": 0.17561006546020508,
+      "learning_rate": 4.525427276023382e-05,
+      "loss": 1.1112,
+      "step": 8752
+    },
+    {
+      "epoch": 0.5318063065799866,
+      "grad_norm": 0.15018552541732788,
+      "learning_rate": 4.524474329125036e-05,
+      "loss": 1.0339,
+      "step": 8753
+    },
+    {
+      "epoch": 0.5318670636126132,
+      "grad_norm": 0.17170754075050354,
+      "learning_rate": 4.523521399657176e-05,
+      "loss": 1.0689,
+      "step": 8754
+    },
+    {
+      "epoch": 0.5319278206452397,
+      "grad_norm": 0.19590263068675995,
+      "learning_rate": 4.522568487654733e-05,
+      "loss": 1.2279,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5319885776778662,
+      "grad_norm": 0.18176360428333282,
+      "learning_rate": 4.5216155931526306e-05,
+      "loss": 1.041,
+      "step": 8756
+    },
+    {
+      "epoch": 0.5320493347104928,
+      "grad_norm": 0.1707795262336731,
+      "learning_rate": 4.520662716185803e-05,
+      "loss": 1.0765,
+      "step": 8757
+    },
+    {
+      "epoch": 0.5321100917431193,
+      "grad_norm": 0.1737699806690216,
+      "learning_rate": 4.519709856789176e-05,
+      "loss": 1.1436,
+      "step": 8758
+    },
+    {
+      "epoch": 0.5321708487757458,
+      "grad_norm": 0.20430248975753784,
+      "learning_rate": 4.5187570149976764e-05,
+      "loss": 1.1429,
+      "step": 8759
+    },
+    {
+      "epoch": 0.5322316058083724,
+      "grad_norm": 0.14655305445194244,
+      "learning_rate": 4.517804190846232e-05,
+      "loss": 1.0815,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5322923628409989,
+      "grad_norm": 0.12511833012104034,
+      "learning_rate": 4.5168513843697686e-05,
+      "loss": 1.0549,
+      "step": 8761
+    },
+    {
+      "epoch": 0.5323531198736253,
+      "grad_norm": 0.1219484955072403,
+      "learning_rate": 4.5158985956032127e-05,
+      "loss": 1.0129,
+      "step": 8762
+    },
+    {
+      "epoch": 0.5324138769062519,
+      "grad_norm": 0.5499448180198669,
+      "learning_rate": 4.5149458245814854e-05,
+      "loss": 1.1436,
+      "step": 8763
+    },
+    {
+      "epoch": 0.5324746339388784,
+      "grad_norm": 0.1544366329908371,
+      "learning_rate": 4.513993071339511e-05,
+      "loss": 1.1108,
+      "step": 8764
+    },
+    {
+      "epoch": 0.5325353909715049,
+      "grad_norm": 0.14449705183506012,
+      "learning_rate": 4.513040335912217e-05,
+      "loss": 1.0283,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5325961480041315,
+      "grad_norm": 0.1301535815000534,
+      "learning_rate": 4.5120876183345226e-05,
+      "loss": 1.0468,
+      "step": 8766
+    },
+    {
+      "epoch": 0.532656905036758,
+      "grad_norm": 0.25364330410957336,
+      "learning_rate": 4.5111349186413504e-05,
+      "loss": 1.1584,
+      "step": 8767
+    },
+    {
+      "epoch": 0.5327176620693845,
+      "grad_norm": 0.14642125368118286,
+      "learning_rate": 4.510182236867625e-05,
+      "loss": 1.0952,
+      "step": 8768
+    },
+    {
+      "epoch": 0.5327784191020111,
+      "grad_norm": 0.21068856120109558,
+      "learning_rate": 4.509229573048261e-05,
+      "loss": 1.0406,
+      "step": 8769
+    },
+    {
+      "epoch": 0.5328391761346376,
+      "grad_norm": 0.11859681457281113,
+      "learning_rate": 4.5082769272181844e-05,
+      "loss": 1.0723,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5328999331672641,
+      "grad_norm": 0.12620341777801514,
+      "learning_rate": 4.507324299412311e-05,
+      "loss": 1.0329,
+      "step": 8771
+    },
+    {
+      "epoch": 0.5329606901998907,
+      "grad_norm": 0.2836582064628601,
+      "learning_rate": 4.50637168966556e-05,
+      "loss": 1.0677,
+      "step": 8772
+    },
+    {
+      "epoch": 0.5330214472325172,
+      "grad_norm": 2.033245086669922,
+      "learning_rate": 4.5054190980128494e-05,
+      "loss": 1.1083,
+      "step": 8773
+    },
+    {
+      "epoch": 0.5330822042651437,
+      "grad_norm": 0.1591397076845169,
+      "learning_rate": 4.504466524489099e-05,
+      "loss": 1.0262,
+      "step": 8774
+    },
+    {
+      "epoch": 0.5331429612977702,
+      "grad_norm": 0.18666645884513855,
+      "learning_rate": 4.503513969129225e-05,
+      "loss": 1.133,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5332037183303967,
+      "grad_norm": 0.20969098806381226,
+      "learning_rate": 4.502561431968141e-05,
+      "loss": 1.0419,
+      "step": 8776
+    },
+    {
+      "epoch": 0.5332644753630232,
+      "grad_norm": 0.19253887236118317,
+      "learning_rate": 4.501608913040763e-05,
+      "loss": 1.043,
+      "step": 8777
+    },
+    {
+      "epoch": 0.5333252323956498,
+      "grad_norm": 0.15176169574260712,
+      "learning_rate": 4.5006564123820074e-05,
+      "loss": 1.0511,
+      "step": 8778
+    },
+    {
+      "epoch": 0.5333859894282763,
+      "grad_norm": 2.6306838989257812,
+      "learning_rate": 4.499703930026788e-05,
+      "loss": 1.0436,
+      "step": 8779
+    },
+    {
+      "epoch": 0.5334467464609028,
+      "grad_norm": 0.22063086926937103,
+      "learning_rate": 4.498751466010017e-05,
+      "loss": 1.0458,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5335075034935294,
+      "grad_norm": 0.17193427681922913,
+      "learning_rate": 4.497799020366608e-05,
+      "loss": 1.1182,
+      "step": 8781
+    },
+    {
+      "epoch": 0.5335682605261559,
+      "grad_norm": 0.8149873614311218,
+      "learning_rate": 4.496846593131474e-05,
+      "loss": 1.0971,
+      "step": 8782
+    },
+    {
+      "epoch": 0.5336290175587824,
+      "grad_norm": 0.16165894269943237,
+      "learning_rate": 4.495894184339525e-05,
+      "loss": 1.036,
+      "step": 8783
+    },
+    {
+      "epoch": 0.533689774591409,
+      "grad_norm": 0.5210099816322327,
+      "learning_rate": 4.494941794025672e-05,
+      "loss": 1.0178,
+      "step": 8784
+    },
+    {
+      "epoch": 0.5337505316240355,
+      "grad_norm": 0.11411449313163757,
+      "learning_rate": 4.493989422224824e-05,
+      "loss": 1.0394,
+      "step": 8785
+    },
+    {
+      "epoch": 0.533811288656662,
+      "grad_norm": 0.13185474276542664,
+      "learning_rate": 4.493037068971892e-05,
+      "loss": 1.0346,
+      "step": 8786
+    },
+    {
+      "epoch": 0.5338720456892886,
+      "grad_norm": 1.2379601001739502,
+      "learning_rate": 4.492084734301785e-05,
+      "loss": 1.1314,
+      "step": 8787
+    },
+    {
+      "epoch": 0.533932802721915,
+      "grad_norm": 0.18772807717323303,
+      "learning_rate": 4.491132418249411e-05,
+      "loss": 1.0133,
+      "step": 8788
+    },
+    {
+      "epoch": 0.5339935597545415,
+      "grad_norm": 0.22782091796398163,
+      "learning_rate": 4.490180120849676e-05,
+      "loss": 1.172,
+      "step": 8789
+    },
+    {
+      "epoch": 0.5340543167871681,
+      "grad_norm": 4.431173801422119,
+      "learning_rate": 4.489227842137485e-05,
+      "loss": 1.1203,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5341150738197946,
+      "grad_norm": 0.16473138332366943,
+      "learning_rate": 4.488275582147749e-05,
+      "loss": 1.0887,
+      "step": 8791
+    },
+    {
+      "epoch": 0.5341758308524212,
+      "grad_norm": 0.17537109553813934,
+      "learning_rate": 4.4873233409153695e-05,
+      "loss": 1.093,
+      "step": 8792
+    },
+    {
+      "epoch": 0.5342365878850477,
+      "grad_norm": 0.13717277348041534,
+      "learning_rate": 4.486371118475251e-05,
+      "loss": 1.0312,
+      "step": 8793
+    },
+    {
+      "epoch": 0.5342973449176742,
+      "grad_norm": 0.14005355536937714,
+      "learning_rate": 4.4854189148622994e-05,
+      "loss": 1.045,
+      "step": 8794
+    },
+    {
+      "epoch": 0.5343581019503008,
+      "grad_norm": 0.14777666330337524,
+      "learning_rate": 4.484466730111418e-05,
+      "loss": 1.1127,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5344188589829273,
+      "grad_norm": 0.1594071239233017,
+      "learning_rate": 4.4835145642575074e-05,
+      "loss": 1.0486,
+      "step": 8796
+    },
+    {
+      "epoch": 0.5344796160155538,
+      "grad_norm": 0.7868282794952393,
+      "learning_rate": 4.48256241733547e-05,
+      "loss": 1.0866,
+      "step": 8797
+    },
+    {
+      "epoch": 0.5345403730481804,
+      "grad_norm": 0.1698208898305893,
+      "learning_rate": 4.481610289380206e-05,
+      "loss": 1.1494,
+      "step": 8798
+    },
+    {
+      "epoch": 0.5346011300808069,
+      "grad_norm": 2.915067195892334,
+      "learning_rate": 4.4806581804266186e-05,
+      "loss": 1.1736,
+      "step": 8799
+    },
+    {
+      "epoch": 0.5346618871134334,
+      "grad_norm": 0.19487714767456055,
+      "learning_rate": 4.4797060905096055e-05,
+      "loss": 1.1308,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5347226441460599,
+      "grad_norm": 0.42165955901145935,
+      "learning_rate": 4.478754019664065e-05,
+      "loss": 1.114,
+      "step": 8801
+    },
+    {
+      "epoch": 0.5347834011786864,
+      "grad_norm": 0.13892394304275513,
+      "learning_rate": 4.4778019679249004e-05,
+      "loss": 1.1049,
+      "step": 8802
+    },
+    {
+      "epoch": 0.5348441582113129,
+      "grad_norm": 0.17071881890296936,
+      "learning_rate": 4.4768499353270026e-05,
+      "loss": 1.1136,
+      "step": 8803
+    },
+    {
+      "epoch": 0.5349049152439395,
+      "grad_norm": 0.12368696182966232,
+      "learning_rate": 4.475897921905272e-05,
+      "loss": 1.0697,
+      "step": 8804
+    },
+    {
+      "epoch": 0.534965672276566,
+      "grad_norm": 0.12244225293397903,
+      "learning_rate": 4.474945927694605e-05,
+      "loss": 1.0742,
+      "step": 8805
+    },
+    {
+      "epoch": 0.5350264293091925,
+      "grad_norm": 0.11974979192018509,
+      "learning_rate": 4.473993952729895e-05,
+      "loss": 1.066,
+      "step": 8806
+    },
+    {
+      "epoch": 0.5350871863418191,
+      "grad_norm": 0.11680539697408676,
+      "learning_rate": 4.47304199704604e-05,
+      "loss": 1.0343,
+      "step": 8807
+    },
+    {
+      "epoch": 0.5351479433744456,
+      "grad_norm": 0.1846323311328888,
+      "learning_rate": 4.472090060677932e-05,
+      "loss": 1.0626,
+      "step": 8808
+    },
+    {
+      "epoch": 0.5352087004070721,
+      "grad_norm": 1.0687099695205688,
+      "learning_rate": 4.471138143660465e-05,
+      "loss": 1.0661,
+      "step": 8809
+    },
+    {
+      "epoch": 0.5352694574396987,
+      "grad_norm": 0.12311764806509018,
+      "learning_rate": 4.470186246028532e-05,
+      "loss": 1.0031,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5353302144723252,
+      "grad_norm": 0.3085007071495056,
+      "learning_rate": 4.469234367817023e-05,
+      "loss": 1.1088,
+      "step": 8811
+    },
+    {
+      "epoch": 0.5353909715049517,
+      "grad_norm": 0.17489825189113617,
+      "learning_rate": 4.4682825090608314e-05,
+      "loss": 1.1509,
+      "step": 8812
+    },
+    {
+      "epoch": 0.5354517285375783,
+      "grad_norm": 0.7380678653717041,
+      "learning_rate": 4.467330669794848e-05,
+      "loss": 1.0371,
+      "step": 8813
+    },
+    {
+      "epoch": 0.5355124855702047,
+      "grad_norm": 0.15315717458724976,
+      "learning_rate": 4.4663788500539603e-05,
+      "loss": 1.103,
+      "step": 8814
+    },
+    {
+      "epoch": 0.5355732426028312,
+      "grad_norm": 0.21286073327064514,
+      "learning_rate": 4.465427049873062e-05,
+      "loss": 1.0372,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5356339996354578,
+      "grad_norm": 0.47610294818878174,
+      "learning_rate": 4.464475269287036e-05,
+      "loss": 1.0699,
+      "step": 8816
+    },
+    {
+      "epoch": 0.5356947566680843,
+      "grad_norm": 0.1400757133960724,
+      "learning_rate": 4.463523508330773e-05,
+      "loss": 1.0679,
+      "step": 8817
+    },
+    {
+      "epoch": 0.5357555137007108,
+      "grad_norm": 0.2324218899011612,
+      "learning_rate": 4.46257176703916e-05,
+      "loss": 1.0718,
+      "step": 8818
+    },
+    {
+      "epoch": 0.5358162707333374,
+      "grad_norm": 0.17135950922966003,
+      "learning_rate": 4.461620045447081e-05,
+      "loss": 1.1083,
+      "step": 8819
+    },
+    {
+      "epoch": 0.5358770277659639,
+      "grad_norm": 0.2737594246864319,
+      "learning_rate": 4.4606683435894255e-05,
+      "loss": 1.1433,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5359377847985904,
+      "grad_norm": 0.22625966370105743,
+      "learning_rate": 4.459716661501075e-05,
+      "loss": 1.0643,
+      "step": 8821
+    },
+    {
+      "epoch": 0.535998541831217,
+      "grad_norm": 0.3096715211868286,
+      "learning_rate": 4.458764999216916e-05,
+      "loss": 1.1407,
+      "step": 8822
+    },
+    {
+      "epoch": 0.5360592988638435,
+      "grad_norm": 0.20227383077144623,
+      "learning_rate": 4.45781335677183e-05,
+      "loss": 1.1686,
+      "step": 8823
+    },
+    {
+      "epoch": 0.53612005589647,
+      "grad_norm": 0.15743187069892883,
+      "learning_rate": 4.4568617342006986e-05,
+      "loss": 1.1052,
+      "step": 8824
+    },
+    {
+      "epoch": 0.5361808129290966,
+      "grad_norm": 0.16367174685001373,
+      "learning_rate": 4.4559101315384066e-05,
+      "loss": 1.0887,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5362415699617231,
+      "grad_norm": 1.1347402334213257,
+      "learning_rate": 4.454958548819834e-05,
+      "loss": 1.135,
+      "step": 8826
+    },
+    {
+      "epoch": 0.5363023269943495,
+      "grad_norm": 0.2929964065551758,
+      "learning_rate": 4.45400698607986e-05,
+      "loss": 1.2995,
+      "step": 8827
+    },
+    {
+      "epoch": 0.5363630840269761,
+      "grad_norm": 0.15751983225345612,
+      "learning_rate": 4.4530554433533666e-05,
+      "loss": 1.1234,
+      "step": 8828
+    },
+    {
+      "epoch": 0.5364238410596026,
+      "grad_norm": 0.25171127915382385,
+      "learning_rate": 4.452103920675232e-05,
+      "loss": 1.1662,
+      "step": 8829
+    },
+    {
+      "epoch": 0.5364845980922291,
+      "grad_norm": 0.20062485337257385,
+      "learning_rate": 4.4511524180803345e-05,
+      "loss": 1.1854,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5365453551248557,
+      "grad_norm": 0.15805010497570038,
+      "learning_rate": 4.45020093560355e-05,
+      "loss": 1.1129,
+      "step": 8831
+    },
+    {
+      "epoch": 0.5366061121574822,
+      "grad_norm": 0.17778664827346802,
+      "learning_rate": 4.4492494732797556e-05,
+      "loss": 1.0276,
+      "step": 8832
+    },
+    {
+      "epoch": 0.5366668691901088,
+      "grad_norm": 0.15899530053138733,
+      "learning_rate": 4.44829803114383e-05,
+      "loss": 1.1698,
+      "step": 8833
+    },
+    {
+      "epoch": 0.5367276262227353,
+      "grad_norm": 0.16488373279571533,
+      "learning_rate": 4.447346609230646e-05,
+      "loss": 1.0858,
+      "step": 8834
+    },
+    {
+      "epoch": 0.5367883832553618,
+      "grad_norm": 0.12887528538703918,
+      "learning_rate": 4.44639520757508e-05,
+      "loss": 1.0697,
+      "step": 8835
+    },
+    {
+      "epoch": 0.5368491402879884,
+      "grad_norm": 0.21981944143772125,
+      "learning_rate": 4.445443826212003e-05,
+      "loss": 1.1858,
+      "step": 8836
+    },
+    {
+      "epoch": 0.5369098973206149,
+      "grad_norm": 0.29560816287994385,
+      "learning_rate": 4.44449246517629e-05,
+      "loss": 1.1876,
+      "step": 8837
+    },
+    {
+      "epoch": 0.5369706543532414,
+      "grad_norm": 0.14979951083660126,
+      "learning_rate": 4.4435411245028134e-05,
+      "loss": 1.0613,
+      "step": 8838
+    },
+    {
+      "epoch": 0.537031411385868,
+      "grad_norm": 1.7466373443603516,
+      "learning_rate": 4.442589804226445e-05,
+      "loss": 1.0611,
+      "step": 8839
+    },
+    {
+      "epoch": 0.5370921684184944,
+      "grad_norm": 0.21137700974941254,
+      "learning_rate": 4.4416385043820534e-05,
+      "loss": 1.0438,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5371529254511209,
+      "grad_norm": 0.21308887004852295,
+      "learning_rate": 4.4406872250045115e-05,
+      "loss": 1.0869,
+      "step": 8841
+    },
+    {
+      "epoch": 0.5372136824837475,
+      "grad_norm": 0.3430609405040741,
+      "learning_rate": 4.4397359661286886e-05,
+      "loss": 1.1257,
+      "step": 8842
+    },
+    {
+      "epoch": 0.537274439516374,
+      "grad_norm": 0.14277590811252594,
+      "learning_rate": 4.438784727789451e-05,
+      "loss": 1.1124,
+      "step": 8843
+    },
+    {
+      "epoch": 0.5373351965490005,
+      "grad_norm": 0.21899861097335815,
+      "learning_rate": 4.437833510021667e-05,
+      "loss": 1.097,
+      "step": 8844
+    },
+    {
+      "epoch": 0.5373959535816271,
+      "grad_norm": 0.15400643646717072,
+      "learning_rate": 4.436882312860205e-05,
+      "loss": 1.0893,
+      "step": 8845
+    },
+    {
+      "epoch": 0.5374567106142536,
+      "grad_norm": 0.21397428214550018,
+      "learning_rate": 4.435931136339931e-05,
+      "loss": 1.115,
+      "step": 8846
+    },
+    {
+      "epoch": 0.5375174676468801,
+      "grad_norm": 0.16637156903743744,
+      "learning_rate": 4.43497998049571e-05,
+      "loss": 1.0701,
+      "step": 8847
+    },
+    {
+      "epoch": 0.5375782246795067,
+      "grad_norm": 0.17225094139575958,
+      "learning_rate": 4.4340288453624057e-05,
+      "loss": 1.1407,
+      "step": 8848
+    },
+    {
+      "epoch": 0.5376389817121332,
+      "grad_norm": 0.20830106735229492,
+      "learning_rate": 4.433077730974886e-05,
+      "loss": 1.1055,
+      "step": 8849
+    },
+    {
+      "epoch": 0.5376997387447597,
+      "grad_norm": 0.16150915622711182,
+      "learning_rate": 4.432126637368011e-05,
+      "loss": 1.0615,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5377604957773863,
+      "grad_norm": 0.13671712577342987,
+      "learning_rate": 4.4311755645766435e-05,
+      "loss": 1.0434,
+      "step": 8851
+    },
+    {
+      "epoch": 0.5378212528100128,
+      "grad_norm": 0.15423385798931122,
+      "learning_rate": 4.430224512635644e-05,
+      "loss": 1.0734,
+      "step": 8852
+    },
+    {
+      "epoch": 0.5378820098426392,
+      "grad_norm": 0.1592642217874527,
+      "learning_rate": 4.4292734815798767e-05,
+      "loss": 1.1297,
+      "step": 8853
+    },
+    {
+      "epoch": 0.5379427668752658,
+      "grad_norm": 0.2576007843017578,
+      "learning_rate": 4.4283224714442e-05,
+      "loss": 1.0747,
+      "step": 8854
+    },
+    {
+      "epoch": 0.5380035239078923,
+      "grad_norm": 0.285372793674469,
+      "learning_rate": 4.4273714822634754e-05,
+      "loss": 1.1583,
+      "step": 8855
+    },
+    {
+      "epoch": 0.5380642809405188,
+      "grad_norm": 0.13003991544246674,
+      "learning_rate": 4.426420514072558e-05,
+      "loss": 1.0867,
+      "step": 8856
+    },
+    {
+      "epoch": 0.5381250379731454,
+      "grad_norm": 0.21347947418689728,
+      "learning_rate": 4.4254695669063066e-05,
+      "loss": 1.1144,
+      "step": 8857
+    },
+    {
+      "epoch": 0.5381857950057719,
+      "grad_norm": 0.1835072636604309,
+      "learning_rate": 4.4245186407995804e-05,
+      "loss": 1.1567,
+      "step": 8858
+    },
+    {
+      "epoch": 0.5382465520383984,
+      "grad_norm": 0.2373637557029724,
+      "learning_rate": 4.4235677357872346e-05,
+      "loss": 1.1071,
+      "step": 8859
+    },
+    {
+      "epoch": 0.538307309071025,
+      "grad_norm": 0.12043152749538422,
+      "learning_rate": 4.422616851904124e-05,
+      "loss": 1.0436,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5383680661036515,
+      "grad_norm": 0.19565355777740479,
+      "learning_rate": 4.4216659891851054e-05,
+      "loss": 1.0849,
+      "step": 8861
+    },
+    {
+      "epoch": 0.538428823136278,
+      "grad_norm": 0.12992137670516968,
+      "learning_rate": 4.420715147665033e-05,
+      "loss": 1.0452,
+      "step": 8862
+    },
+    {
+      "epoch": 0.5384895801689046,
+      "grad_norm": 0.15971948206424713,
+      "learning_rate": 4.419764327378757e-05,
+      "loss": 1.1674,
+      "step": 8863
+    },
+    {
+      "epoch": 0.5385503372015311,
+      "grad_norm": 17.72208023071289,
+      "learning_rate": 4.4188135283611324e-05,
+      "loss": 1.051,
+      "step": 8864
+    },
+    {
+      "epoch": 0.5386110942341577,
+      "grad_norm": 0.14163905382156372,
+      "learning_rate": 4.417862750647009e-05,
+      "loss": 1.0574,
+      "step": 8865
+    },
+    {
+      "epoch": 0.5386718512667842,
+      "grad_norm": 0.27367568016052246,
+      "learning_rate": 4.41691199427124e-05,
+      "loss": 1.0457,
+      "step": 8866
+    },
+    {
+      "epoch": 0.5387326082994106,
+      "grad_norm": 0.18459880352020264,
+      "learning_rate": 4.4159612592686744e-05,
+      "loss": 1.1731,
+      "step": 8867
+    },
+    {
+      "epoch": 0.5387933653320371,
+      "grad_norm": 0.12797421216964722,
+      "learning_rate": 4.415010545674161e-05,
+      "loss": 1.0332,
+      "step": 8868
+    },
+    {
+      "epoch": 0.5388541223646637,
+      "grad_norm": 0.1594666838645935,
+      "learning_rate": 4.414059853522552e-05,
+      "loss": 1.151,
+      "step": 8869
+    },
+    {
+      "epoch": 0.5389148793972902,
+      "grad_norm": 0.22706584632396698,
+      "learning_rate": 4.4131091828486894e-05,
+      "loss": 1.0859,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5389756364299167,
+      "grad_norm": 0.21176759898662567,
+      "learning_rate": 4.412158533687424e-05,
+      "loss": 1.2333,
+      "step": 8871
+    },
+    {
+      "epoch": 0.5390363934625433,
+      "grad_norm": 0.17553776502609253,
+      "learning_rate": 4.411207906073601e-05,
+      "loss": 1.1174,
+      "step": 8872
+    },
+    {
+      "epoch": 0.5390971504951698,
+      "grad_norm": 0.2820688486099243,
+      "learning_rate": 4.4102573000420657e-05,
+      "loss": 1.0816,
+      "step": 8873
+    },
+    {
+      "epoch": 0.5391579075277964,
+      "grad_norm": 0.13691022992134094,
+      "learning_rate": 4.409306715627663e-05,
+      "loss": 1.0404,
+      "step": 8874
+    },
+    {
+      "epoch": 0.5392186645604229,
+      "grad_norm": 0.15847517549991608,
+      "learning_rate": 4.4083561528652384e-05,
+      "loss": 1.0776,
+      "step": 8875
+    },
+    {
+      "epoch": 0.5392794215930494,
+      "grad_norm": 0.2586681842803955,
+      "learning_rate": 4.407405611789633e-05,
+      "loss": 1.0797,
+      "step": 8876
+    },
+    {
+      "epoch": 0.539340178625676,
+      "grad_norm": 0.18156175315380096,
+      "learning_rate": 4.406455092435689e-05,
+      "loss": 1.0736,
+      "step": 8877
+    },
+    {
+      "epoch": 0.5394009356583025,
+      "grad_norm": 0.24516716599464417,
+      "learning_rate": 4.405504594838247e-05,
+      "loss": 1.1073,
+      "step": 8878
+    },
+    {
+      "epoch": 0.539461692690929,
+      "grad_norm": 0.19556109607219696,
+      "learning_rate": 4.40455411903215e-05,
+      "loss": 1.1026,
+      "step": 8879
+    },
+    {
+      "epoch": 0.5395224497235555,
+      "grad_norm": 0.3313329219818115,
+      "learning_rate": 4.4036036650522376e-05,
+      "loss": 1.0721,
+      "step": 8880
+    },
+    {
+      "epoch": 0.539583206756182,
+      "grad_norm": 0.15644875168800354,
+      "learning_rate": 4.402653232933346e-05,
+      "loss": 1.1154,
+      "step": 8881
+    },
+    {
+      "epoch": 0.5396439637888085,
+      "grad_norm": 0.30278852581977844,
+      "learning_rate": 4.40170282271032e-05,
+      "loss": 1.1149,
+      "step": 8882
+    },
+    {
+      "epoch": 0.5397047208214351,
+      "grad_norm": 0.12426209449768066,
+      "learning_rate": 4.400752434417989e-05,
+      "loss": 1.0553,
+      "step": 8883
+    },
+    {
+      "epoch": 0.5397654778540616,
+      "grad_norm": 0.12561176717281342,
+      "learning_rate": 4.399802068091195e-05,
+      "loss": 1.0521,
+      "step": 8884
+    },
+    {
+      "epoch": 0.5398262348866881,
+      "grad_norm": 0.4633282721042633,
+      "learning_rate": 4.398851723764771e-05,
+      "loss": 1.3691,
+      "step": 8885
+    },
+    {
+      "epoch": 0.5398869919193147,
+      "grad_norm": 0.1240839883685112,
+      "learning_rate": 4.397901401473552e-05,
+      "loss": 1.0433,
+      "step": 8886
+    },
+    {
+      "epoch": 0.5399477489519412,
+      "grad_norm": 0.14549250900745392,
+      "learning_rate": 4.396951101252376e-05,
+      "loss": 1.0515,
+      "step": 8887
+    },
+    {
+      "epoch": 0.5400085059845677,
+      "grad_norm": 0.19717401266098022,
+      "learning_rate": 4.396000823136073e-05,
+      "loss": 1.1299,
+      "step": 8888
+    },
+    {
+      "epoch": 0.5400692630171943,
+      "grad_norm": 0.15542292594909668,
+      "learning_rate": 4.395050567159478e-05,
+      "loss": 1.0373,
+      "step": 8889
+    },
+    {
+      "epoch": 0.5401300200498208,
+      "grad_norm": 0.24425075948238373,
+      "learning_rate": 4.39410033335742e-05,
+      "loss": 1.0766,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5401907770824473,
+      "grad_norm": 0.26645809412002563,
+      "learning_rate": 4.3931501217647305e-05,
+      "loss": 1.1077,
+      "step": 8891
+    },
+    {
+      "epoch": 0.5402515341150739,
+      "grad_norm": 0.15452319383621216,
+      "learning_rate": 4.392199932416242e-05,
+      "loss": 1.0618,
+      "step": 8892
+    },
+    {
+      "epoch": 0.5403122911477003,
+      "grad_norm": 0.170376256108284,
+      "learning_rate": 4.3912497653467815e-05,
+      "loss": 1.0709,
+      "step": 8893
+    },
+    {
+      "epoch": 0.5403730481803268,
+      "grad_norm": 0.2439974695444107,
+      "learning_rate": 4.3902996205911784e-05,
+      "loss": 1.1064,
+      "step": 8894
+    },
+    {
+      "epoch": 0.5404338052129534,
+      "grad_norm": 0.1813730001449585,
+      "learning_rate": 4.389349498184261e-05,
+      "loss": 1.1105,
+      "step": 8895
+    },
+    {
+      "epoch": 0.5404945622455799,
+      "grad_norm": 0.1947130560874939,
+      "learning_rate": 4.3883993981608576e-05,
+      "loss": 1.0493,
+      "step": 8896
+    },
+    {
+      "epoch": 0.5405553192782064,
+      "grad_norm": 0.20357480645179749,
+      "learning_rate": 4.387449320555791e-05,
+      "loss": 1.0732,
+      "step": 8897
+    },
+    {
+      "epoch": 0.540616076310833,
+      "grad_norm": 0.15034297108650208,
+      "learning_rate": 4.3864992654038875e-05,
+      "loss": 1.0392,
+      "step": 8898
+    },
+    {
+      "epoch": 0.5406768333434595,
+      "grad_norm": 0.15569765865802765,
+      "learning_rate": 4.3855492327399713e-05,
+      "loss": 1.1124,
+      "step": 8899
+    },
+    {
+      "epoch": 0.540737590376086,
+      "grad_norm": 0.15072257816791534,
+      "learning_rate": 4.384599222598868e-05,
+      "loss": 1.1082,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5407983474087126,
+      "grad_norm": 0.19490168988704681,
+      "learning_rate": 4.383649235015399e-05,
+      "loss": 1.0999,
+      "step": 8901
+    },
+    {
+      "epoch": 0.5408591044413391,
+      "grad_norm": 0.16578565537929535,
+      "learning_rate": 4.3826992700243876e-05,
+      "loss": 1.1026,
+      "step": 8902
+    },
+    {
+      "epoch": 0.5409198614739656,
+      "grad_norm": 0.21928803622722626,
+      "learning_rate": 4.381749327660652e-05,
+      "loss": 1.0478,
+      "step": 8903
+    },
+    {
+      "epoch": 0.5409806185065922,
+      "grad_norm": 0.12841108441352844,
+      "learning_rate": 4.3807994079590145e-05,
+      "loss": 1.0183,
+      "step": 8904
+    },
+    {
+      "epoch": 0.5410413755392187,
+      "grad_norm": 0.18991781771183014,
+      "learning_rate": 4.3798495109542946e-05,
+      "loss": 1.0409,
+      "step": 8905
+    },
+    {
+      "epoch": 0.5411021325718451,
+      "grad_norm": 0.1959805190563202,
+      "learning_rate": 4.378899636681311e-05,
+      "loss": 1.1746,
+      "step": 8906
+    },
+    {
+      "epoch": 0.5411628896044717,
+      "grad_norm": 0.17164644598960876,
+      "learning_rate": 4.37794978517488e-05,
+      "loss": 1.135,
+      "step": 8907
+    },
+    {
+      "epoch": 0.5412236466370982,
+      "grad_norm": 0.2689679265022278,
+      "learning_rate": 4.376999956469822e-05,
+      "loss": 1.0805,
+      "step": 8908
+    },
+    {
+      "epoch": 0.5412844036697247,
+      "grad_norm": 0.22434532642364502,
+      "learning_rate": 4.376050150600951e-05,
+      "loss": 1.0775,
+      "step": 8909
+    },
+    {
+      "epoch": 0.5413451607023513,
+      "grad_norm": 0.22144941985607147,
+      "learning_rate": 4.375100367603081e-05,
+      "loss": 1.1149,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5414059177349778,
+      "grad_norm": 0.22280794382095337,
+      "learning_rate": 4.3741506075110285e-05,
+      "loss": 1.1546,
+      "step": 8911
+    },
+    {
+      "epoch": 0.5414666747676044,
+      "grad_norm": 0.1654522567987442,
+      "learning_rate": 4.373200870359605e-05,
+      "loss": 1.056,
+      "step": 8912
+    },
+    {
+      "epoch": 0.5415274318002309,
+      "grad_norm": 0.24848440289497375,
+      "learning_rate": 4.372251156183626e-05,
+      "loss": 1.1273,
+      "step": 8913
+    },
+    {
+      "epoch": 0.5415881888328574,
+      "grad_norm": 0.25273504853248596,
+      "learning_rate": 4.3713014650179026e-05,
+      "loss": 1.0653,
+      "step": 8914
+    },
+    {
+      "epoch": 0.541648945865484,
+      "grad_norm": 0.15259887278079987,
+      "learning_rate": 4.370351796897244e-05,
+      "loss": 1.0733,
+      "step": 8915
+    },
+    {
+      "epoch": 0.5417097028981105,
+      "grad_norm": 0.16894663870334625,
+      "learning_rate": 4.369402151856465e-05,
+      "loss": 1.0846,
+      "step": 8916
+    },
+    {
+      "epoch": 0.541770459930737,
+      "grad_norm": 0.5015411376953125,
+      "learning_rate": 4.368452529930369e-05,
+      "loss": 1.2142,
+      "step": 8917
+    },
+    {
+      "epoch": 0.5418312169633636,
+      "grad_norm": 0.1331171989440918,
+      "learning_rate": 4.367502931153769e-05,
+      "loss": 1.0917,
+      "step": 8918
+    },
+    {
+      "epoch": 0.54189197399599,
+      "grad_norm": 0.16969387233257294,
+      "learning_rate": 4.366553355561471e-05,
+      "loss": 1.084,
+      "step": 8919
+    },
+    {
+      "epoch": 0.5419527310286165,
+      "grad_norm": 0.19851133227348328,
+      "learning_rate": 4.3656038031882816e-05,
+      "loss": 1.187,
+      "step": 8920
+    },
+    {
+      "epoch": 0.542013488061243,
+      "grad_norm": 0.17347542941570282,
+      "learning_rate": 4.364654274069008e-05,
+      "loss": 1.0572,
+      "step": 8921
+    },
+    {
+      "epoch": 0.5420742450938696,
+      "grad_norm": 0.2708236277103424,
+      "learning_rate": 4.363704768238456e-05,
+      "loss": 1.1133,
+      "step": 8922
+    },
+    {
+      "epoch": 0.5421350021264961,
+      "grad_norm": 0.7941281199455261,
+      "learning_rate": 4.362755285731427e-05,
+      "loss": 0.9987,
+      "step": 8923
+    },
+    {
+      "epoch": 0.5421957591591227,
+      "grad_norm": 0.21143792569637299,
+      "learning_rate": 4.3618058265827264e-05,
+      "loss": 1.1074,
+      "step": 8924
+    },
+    {
+      "epoch": 0.5422565161917492,
+      "grad_norm": 0.16356296837329865,
+      "learning_rate": 4.360856390827156e-05,
+      "loss": 1.1182,
+      "step": 8925
+    },
+    {
+      "epoch": 0.5423172732243757,
+      "grad_norm": 0.1267087757587433,
+      "learning_rate": 4.359906978499519e-05,
+      "loss": 1.0231,
+      "step": 8926
+    },
+    {
+      "epoch": 0.5423780302570023,
+      "grad_norm": 0.12112586200237274,
+      "learning_rate": 4.358957589634616e-05,
+      "loss": 1.082,
+      "step": 8927
+    },
+    {
+      "epoch": 0.5424387872896288,
+      "grad_norm": 0.8716529011726379,
+      "learning_rate": 4.358008224267245e-05,
+      "loss": 1.207,
+      "step": 8928
+    },
+    {
+      "epoch": 0.5424995443222553,
+      "grad_norm": 2.495262384414673,
+      "learning_rate": 4.3570588824322096e-05,
+      "loss": 1.0556,
+      "step": 8929
+    },
+    {
+      "epoch": 0.5425603013548819,
+      "grad_norm": 0.145907923579216,
+      "learning_rate": 4.356109564164302e-05,
+      "loss": 1.0894,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5426210583875084,
+      "grad_norm": 0.13720855116844177,
+      "learning_rate": 4.355160269498324e-05,
+      "loss": 1.0515,
+      "step": 8931
+    },
+    {
+      "epoch": 0.5426818154201348,
+      "grad_norm": 0.24630962312221527,
+      "learning_rate": 4.354210998469071e-05,
+      "loss": 1.1706,
+      "step": 8932
+    },
+    {
+      "epoch": 0.5427425724527614,
+      "grad_norm": 0.12788785994052887,
+      "learning_rate": 4.353261751111337e-05,
+      "loss": 1.0484,
+      "step": 8933
+    },
+    {
+      "epoch": 0.5428033294853879,
+      "grad_norm": 0.17123116552829742,
+      "learning_rate": 4.352312527459921e-05,
+      "loss": 1.0325,
+      "step": 8934
+    },
+    {
+      "epoch": 0.5428640865180144,
+      "grad_norm": 0.30883681774139404,
+      "learning_rate": 4.351363327549613e-05,
+      "loss": 1.3499,
+      "step": 8935
+    },
+    {
+      "epoch": 0.542924843550641,
+      "grad_norm": 0.2096819281578064,
+      "learning_rate": 4.35041415141521e-05,
+      "loss": 1.2215,
+      "step": 8936
+    },
+    {
+      "epoch": 0.5429856005832675,
+      "grad_norm": 0.158378005027771,
+      "learning_rate": 4.349464999091498e-05,
+      "loss": 1.1658,
+      "step": 8937
+    },
+    {
+      "epoch": 0.543046357615894,
+      "grad_norm": 0.5119233131408691,
+      "learning_rate": 4.3485158706132744e-05,
+      "loss": 1.0704,
+      "step": 8938
+    },
+    {
+      "epoch": 0.5431071146485206,
+      "grad_norm": 0.2655501961708069,
+      "learning_rate": 4.347566766015327e-05,
+      "loss": 1.1108,
+      "step": 8939
+    },
+    {
+      "epoch": 0.5431678716811471,
+      "grad_norm": 0.13762451708316803,
+      "learning_rate": 4.346617685332447e-05,
+      "loss": 1.0482,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5432286287137736,
+      "grad_norm": 0.18351638317108154,
+      "learning_rate": 4.3456686285994205e-05,
+      "loss": 1.0011,
+      "step": 8941
+    },
+    {
+      "epoch": 0.5432893857464002,
+      "grad_norm": 0.17542748153209686,
+      "learning_rate": 4.3447195958510384e-05,
+      "loss": 1.0055,
+      "step": 8942
+    },
+    {
+      "epoch": 0.5433501427790267,
+      "grad_norm": 0.16803203523159027,
+      "learning_rate": 4.343770587122086e-05,
+      "loss": 1.1914,
+      "step": 8943
+    },
+    {
+      "epoch": 0.5434108998116532,
+      "grad_norm": 0.11520200222730637,
+      "learning_rate": 4.342821602447348e-05,
+      "loss": 1.0222,
+      "step": 8944
+    },
+    {
+      "epoch": 0.5434716568442797,
+      "grad_norm": 0.14066025614738464,
+      "learning_rate": 4.341872641861612e-05,
+      "loss": 1.1117,
+      "step": 8945
+    },
+    {
+      "epoch": 0.5435324138769062,
+      "grad_norm": 0.13849855959415436,
+      "learning_rate": 4.3409237053996615e-05,
+      "loss": 1.0811,
+      "step": 8946
+    },
+    {
+      "epoch": 0.5435931709095327,
+      "grad_norm": 0.12274722754955292,
+      "learning_rate": 4.3399747930962805e-05,
+      "loss": 1.0202,
+      "step": 8947
+    },
+    {
+      "epoch": 0.5436539279421593,
+      "grad_norm": 0.1560252159833908,
+      "learning_rate": 4.3390259049862503e-05,
+      "loss": 1.0629,
+      "step": 8948
+    },
+    {
+      "epoch": 0.5437146849747858,
+      "grad_norm": 0.28013524413108826,
+      "learning_rate": 4.338077041104355e-05,
+      "loss": 1.0796,
+      "step": 8949
+    },
+    {
+      "epoch": 0.5437754420074123,
+      "grad_norm": 0.15444281697273254,
+      "learning_rate": 4.337128201485371e-05,
+      "loss": 1.1592,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5438361990400389,
+      "grad_norm": 0.26170778274536133,
+      "learning_rate": 4.336179386164083e-05,
+      "loss": 1.2249,
+      "step": 8951
+    },
+    {
+      "epoch": 0.5438969560726654,
+      "grad_norm": 0.17421099543571472,
+      "learning_rate": 4.335230595175267e-05,
+      "loss": 1.1489,
+      "step": 8952
+    },
+    {
+      "epoch": 0.543957713105292,
+      "grad_norm": 0.2767316699028015,
+      "learning_rate": 4.334281828553701e-05,
+      "loss": 1.14,
+      "step": 8953
+    },
+    {
+      "epoch": 0.5440184701379185,
+      "grad_norm": 0.19828414916992188,
+      "learning_rate": 4.333333086334165e-05,
+      "loss": 1.1125,
+      "step": 8954
+    },
+    {
+      "epoch": 0.544079227170545,
+      "grad_norm": 0.3742527365684509,
+      "learning_rate": 4.332384368551433e-05,
+      "loss": 1.0893,
+      "step": 8955
+    },
+    {
+      "epoch": 0.5441399842031716,
+      "grad_norm": 0.22007344663143158,
+      "learning_rate": 4.3314356752402826e-05,
+      "loss": 1.1368,
+      "step": 8956
+    },
+    {
+      "epoch": 0.5442007412357981,
+      "grad_norm": 0.23593324422836304,
+      "learning_rate": 4.330487006435485e-05,
+      "loss": 1.036,
+      "step": 8957
+    },
+    {
+      "epoch": 0.5442614982684245,
+      "grad_norm": 0.2624753415584564,
+      "learning_rate": 4.329538362171815e-05,
+      "loss": 1.0548,
+      "step": 8958
+    },
+    {
+      "epoch": 0.544322255301051,
+      "grad_norm": 0.44120582938194275,
+      "learning_rate": 4.328589742484047e-05,
+      "loss": 1.2736,
+      "step": 8959
+    },
+    {
+      "epoch": 0.5443830123336776,
+      "grad_norm": 0.16598977148532867,
+      "learning_rate": 4.3276411474069515e-05,
+      "loss": 1.1385,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5444437693663041,
+      "grad_norm": 0.13173899054527283,
+      "learning_rate": 4.326692576975299e-05,
+      "loss": 1.0448,
+      "step": 8961
+    },
+    {
+      "epoch": 0.5445045263989307,
+      "grad_norm": 0.9805957078933716,
+      "learning_rate": 4.325744031223861e-05,
+      "loss": 1.1726,
+      "step": 8962
+    },
+    {
+      "epoch": 0.5445652834315572,
+      "grad_norm": 0.21744363009929657,
+      "learning_rate": 4.324795510187407e-05,
+      "loss": 1.146,
+      "step": 8963
+    },
+    {
+      "epoch": 0.5446260404641837,
+      "grad_norm": 0.11330989003181458,
+      "learning_rate": 4.323847013900703e-05,
+      "loss": 1.0502,
+      "step": 8964
+    },
+    {
+      "epoch": 0.5446867974968103,
+      "grad_norm": 0.22659528255462646,
+      "learning_rate": 4.3228985423985176e-05,
+      "loss": 1.0651,
+      "step": 8965
+    },
+    {
+      "epoch": 0.5447475545294368,
+      "grad_norm": 0.16389329731464386,
+      "learning_rate": 4.3219500957156157e-05,
+      "loss": 1.0999,
+      "step": 8966
+    },
+    {
+      "epoch": 0.5448083115620633,
+      "grad_norm": 0.17884069681167603,
+      "learning_rate": 4.321001673886765e-05,
+      "loss": 1.1442,
+      "step": 8967
+    },
+    {
+      "epoch": 0.5448690685946899,
+      "grad_norm": 0.11576376855373383,
+      "learning_rate": 4.32005327694673e-05,
+      "loss": 1.004,
+      "step": 8968
+    },
+    {
+      "epoch": 0.5449298256273164,
+      "grad_norm": 0.19715000689029694,
+      "learning_rate": 4.319104904930274e-05,
+      "loss": 1.1733,
+      "step": 8969
+    },
+    {
+      "epoch": 0.5449905826599429,
+      "grad_norm": 0.191041499376297,
+      "learning_rate": 4.3181565578721584e-05,
+      "loss": 1.1468,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5450513396925695,
+      "grad_norm": 0.30083873867988586,
+      "learning_rate": 4.3172082358071446e-05,
+      "loss": 1.2533,
+      "step": 8971
+    },
+    {
+      "epoch": 0.5451120967251959,
+      "grad_norm": 0.2145347148180008,
+      "learning_rate": 4.3162599387699966e-05,
+      "loss": 1.1295,
+      "step": 8972
+    },
+    {
+      "epoch": 0.5451728537578224,
+      "grad_norm": 0.12549899518489838,
+      "learning_rate": 4.315311666795472e-05,
+      "loss": 1.0227,
+      "step": 8973
+    },
+    {
+      "epoch": 0.545233610790449,
+      "grad_norm": 0.12783335149288177,
+      "learning_rate": 4.31436341991833e-05,
+      "loss": 1.0562,
+      "step": 8974
+    },
+    {
+      "epoch": 0.5452943678230755,
+      "grad_norm": 0.1302906572818756,
+      "learning_rate": 4.31341519817333e-05,
+      "loss": 1.0648,
+      "step": 8975
+    },
+    {
+      "epoch": 0.545355124855702,
+      "grad_norm": 0.12566597759723663,
+      "learning_rate": 4.31246700159523e-05,
+      "loss": 1.04,
+      "step": 8976
+    },
+    {
+      "epoch": 0.5454158818883286,
+      "grad_norm": 0.17287570238113403,
+      "learning_rate": 4.3115188302187844e-05,
+      "loss": 1.0558,
+      "step": 8977
+    },
+    {
+      "epoch": 0.5454766389209551,
+      "grad_norm": 0.20188038051128387,
+      "learning_rate": 4.310570684078749e-05,
+      "loss": 1.1919,
+      "step": 8978
+    },
+    {
+      "epoch": 0.5455373959535816,
+      "grad_norm": 0.2005751132965088,
+      "learning_rate": 4.309622563209876e-05,
+      "loss": 1.0822,
+      "step": 8979
+    },
+    {
+      "epoch": 0.5455981529862082,
+      "grad_norm": 0.2326764315366745,
+      "learning_rate": 4.308674467646924e-05,
+      "loss": 1.0903,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5456589100188347,
+      "grad_norm": 0.12649978697299957,
+      "learning_rate": 4.307726397424642e-05,
+      "loss": 1.044,
+      "step": 8981
+    },
+    {
+      "epoch": 0.5457196670514612,
+      "grad_norm": 0.1572638750076294,
+      "learning_rate": 4.306778352577783e-05,
+      "loss": 1.061,
+      "step": 8982
+    },
+    {
+      "epoch": 0.5457804240840878,
+      "grad_norm": 0.14660786092281342,
+      "learning_rate": 4.3058303331410994e-05,
+      "loss": 1.0848,
+      "step": 8983
+    },
+    {
+      "epoch": 0.5458411811167143,
+      "grad_norm": 0.12938623130321503,
+      "learning_rate": 4.3048823391493366e-05,
+      "loss": 1.0715,
+      "step": 8984
+    },
+    {
+      "epoch": 0.5459019381493407,
+      "grad_norm": 0.25859758257865906,
+      "learning_rate": 4.3039343706372465e-05,
+      "loss": 1.0771,
+      "step": 8985
+    },
+    {
+      "epoch": 0.5459626951819673,
+      "grad_norm": 0.16211803257465363,
+      "learning_rate": 4.302986427639577e-05,
+      "loss": 1.1068,
+      "step": 8986
+    },
+    {
+      "epoch": 0.5460234522145938,
+      "grad_norm": 0.32541561126708984,
+      "learning_rate": 4.302038510191073e-05,
+      "loss": 1.0978,
+      "step": 8987
+    },
+    {
+      "epoch": 0.5460842092472203,
+      "grad_norm": 0.11973492801189423,
+      "learning_rate": 4.301090618326484e-05,
+      "loss": 0.9997,
+      "step": 8988
+    },
+    {
+      "epoch": 0.5461449662798469,
+      "grad_norm": 0.17381417751312256,
+      "learning_rate": 4.300142752080555e-05,
+      "loss": 1.1196,
+      "step": 8989
+    },
+    {
+      "epoch": 0.5462057233124734,
+      "grad_norm": 0.2164960354566574,
+      "learning_rate": 4.299194911488027e-05,
+      "loss": 1.0808,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5462664803451,
+      "grad_norm": 0.18794627487659454,
+      "learning_rate": 4.2982470965836455e-05,
+      "loss": 1.018,
+      "step": 8991
+    },
+    {
+      "epoch": 0.5463272373777265,
+      "grad_norm": 0.33291032910346985,
+      "learning_rate": 4.297299307402151e-05,
+      "loss": 1.1558,
+      "step": 8992
+    },
+    {
+      "epoch": 0.546387994410353,
+      "grad_norm": 0.21408331394195557,
+      "learning_rate": 4.2963515439782864e-05,
+      "loss": 1.0628,
+      "step": 8993
+    },
+    {
+      "epoch": 0.5464487514429796,
+      "grad_norm": 0.1658337414264679,
+      "learning_rate": 4.295403806346793e-05,
+      "loss": 1.1061,
+      "step": 8994
+    },
+    {
+      "epoch": 0.5465095084756061,
+      "grad_norm": 0.17583677172660828,
+      "learning_rate": 4.294456094542408e-05,
+      "loss": 1.0672,
+      "step": 8995
+    },
+    {
+      "epoch": 0.5465702655082326,
+      "grad_norm": 0.24074803292751312,
+      "learning_rate": 4.2935084085998745e-05,
+      "loss": 1.1618,
+      "step": 8996
+    },
+    {
+      "epoch": 0.5466310225408592,
+      "grad_norm": 0.1395314484834671,
+      "learning_rate": 4.2925607485539235e-05,
+      "loss": 1.0544,
+      "step": 8997
+    },
+    {
+      "epoch": 0.5466917795734856,
+      "grad_norm": 0.23153254389762878,
+      "learning_rate": 4.291613114439296e-05,
+      "loss": 1.1302,
+      "step": 8998
+    },
+    {
+      "epoch": 0.5467525366061121,
+      "grad_norm": 0.20346395671367645,
+      "learning_rate": 4.290665506290727e-05,
+      "loss": 1.1132,
+      "step": 8999
+    },
+    {
+      "epoch": 0.5468132936387387,
+      "grad_norm": 5.548253536224365,
+      "learning_rate": 4.28971792414295e-05,
+      "loss": 1.0077,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5468740506713652,
+      "grad_norm": 0.23009063303470612,
+      "learning_rate": 4.2887703680307e-05,
+      "loss": 1.0841,
+      "step": 9001
+    },
+    {
+      "epoch": 0.5469348077039917,
+      "grad_norm": 1.0316535234451294,
+      "learning_rate": 4.28782283798871e-05,
+      "loss": 1.1485,
+      "step": 9002
+    },
+    {
+      "epoch": 0.5469955647366183,
+      "grad_norm": 0.23427841067314148,
+      "learning_rate": 4.2868753340517135e-05,
+      "loss": 1.1046,
+      "step": 9003
+    },
+    {
+      "epoch": 0.5470563217692448,
+      "grad_norm": 2.872511625289917,
+      "learning_rate": 4.2859278562544373e-05,
+      "loss": 1.0494,
+      "step": 9004
+    },
+    {
+      "epoch": 0.5471170788018713,
+      "grad_norm": 0.7402791380882263,
+      "learning_rate": 4.284980404631613e-05,
+      "loss": 1.0541,
+      "step": 9005
+    },
+    {
+      "epoch": 0.5471778358344979,
+      "grad_norm": 0.15071167051792145,
+      "learning_rate": 4.2840329792179713e-05,
+      "loss": 1.1449,
+      "step": 9006
+    },
+    {
+      "epoch": 0.5472385928671244,
+      "grad_norm": 0.24968582391738892,
+      "learning_rate": 4.283085580048239e-05,
+      "loss": 1.2006,
+      "step": 9007
+    },
+    {
+      "epoch": 0.5472993498997509,
+      "grad_norm": 0.31403857469558716,
+      "learning_rate": 4.282138207157142e-05,
+      "loss": 1.121,
+      "step": 9008
+    },
+    {
+      "epoch": 0.5473601069323775,
+      "grad_norm": 0.32856741547584534,
+      "learning_rate": 4.281190860579411e-05,
+      "loss": 1.1303,
+      "step": 9009
+    },
+    {
+      "epoch": 0.547420863965004,
+      "grad_norm": 0.16440510749816895,
+      "learning_rate": 4.2802435403497654e-05,
+      "loss": 1.051,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5474816209976304,
+      "grad_norm": 0.1508004516363144,
+      "learning_rate": 4.2792962465029326e-05,
+      "loss": 1.0438,
+      "step": 9011
+    },
+    {
+      "epoch": 0.547542378030257,
+      "grad_norm": 0.1534392535686493,
+      "learning_rate": 4.278348979073635e-05,
+      "loss": 1.0537,
+      "step": 9012
+    },
+    {
+      "epoch": 0.5476031350628835,
+      "grad_norm": 0.11253668367862701,
+      "learning_rate": 4.2774017380965944e-05,
+      "loss": 1.0282,
+      "step": 9013
+    },
+    {
+      "epoch": 0.54766389209551,
+      "grad_norm": 0.21295097470283508,
+      "learning_rate": 4.2764545236065335e-05,
+      "loss": 1.0956,
+      "step": 9014
+    },
+    {
+      "epoch": 0.5477246491281366,
+      "grad_norm": 0.17431528866291046,
+      "learning_rate": 4.275507335638172e-05,
+      "loss": 1.1214,
+      "step": 9015
+    },
+    {
+      "epoch": 0.5477854061607631,
+      "grad_norm": 0.19457079470157623,
+      "learning_rate": 4.2745601742262306e-05,
+      "loss": 1.1526,
+      "step": 9016
+    },
+    {
+      "epoch": 0.5478461631933896,
+      "grad_norm": 0.20628224313259125,
+      "learning_rate": 4.273613039405425e-05,
+      "loss": 1.1626,
+      "step": 9017
+    },
+    {
+      "epoch": 0.5479069202260162,
+      "grad_norm": 0.27337104082107544,
+      "learning_rate": 4.272665931210472e-05,
+      "loss": 1.109,
+      "step": 9018
+    },
+    {
+      "epoch": 0.5479676772586427,
+      "grad_norm": 0.14324428141117096,
+      "learning_rate": 4.2717188496760916e-05,
+      "loss": 1.0948,
+      "step": 9019
+    },
+    {
+      "epoch": 0.5480284342912692,
+      "grad_norm": 0.24862165749073029,
+      "learning_rate": 4.2707717948369974e-05,
+      "loss": 1.0717,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5480891913238958,
+      "grad_norm": 0.20165401697158813,
+      "learning_rate": 4.2698247667279036e-05,
+      "loss": 1.0216,
+      "step": 9021
+    },
+    {
+      "epoch": 0.5481499483565223,
+      "grad_norm": 0.133982315659523,
+      "learning_rate": 4.268877765383524e-05,
+      "loss": 1.0226,
+      "step": 9022
+    },
+    {
+      "epoch": 0.5482107053891488,
+      "grad_norm": 0.2255735695362091,
+      "learning_rate": 4.267930790838573e-05,
+      "loss": 1.0981,
+      "step": 9023
+    },
+    {
+      "epoch": 0.5482714624217753,
+      "grad_norm": 0.15962889790534973,
+      "learning_rate": 4.266983843127759e-05,
+      "loss": 1.0329,
+      "step": 9024
+    },
+    {
+      "epoch": 0.5483322194544018,
+      "grad_norm": 0.11453757435083389,
+      "learning_rate": 4.2660369222857944e-05,
+      "loss": 1.0353,
+      "step": 9025
+    },
+    {
+      "epoch": 0.5483929764870283,
+      "grad_norm": 0.1269068419933319,
+      "learning_rate": 4.265090028347387e-05,
+      "loss": 1.051,
+      "step": 9026
+    },
+    {
+      "epoch": 0.5484537335196549,
+      "grad_norm": 0.19441090524196625,
+      "learning_rate": 4.2641431613472475e-05,
+      "loss": 1.0557,
+      "step": 9027
+    },
+    {
+      "epoch": 0.5485144905522814,
+      "grad_norm": 0.13083462417125702,
+      "learning_rate": 4.263196321320083e-05,
+      "loss": 1.0419,
+      "step": 9028
+    },
+    {
+      "epoch": 0.5485752475849079,
+      "grad_norm": 0.14303791522979736,
+      "learning_rate": 4.262249508300598e-05,
+      "loss": 1.0674,
+      "step": 9029
+    },
+    {
+      "epoch": 0.5486360046175345,
+      "grad_norm": 0.1453258991241455,
+      "learning_rate": 4.261302722323504e-05,
+      "loss": 1.1258,
+      "step": 9030
+    },
+    {
+      "epoch": 0.548696761650161,
+      "grad_norm": 0.12882418930530548,
+      "learning_rate": 4.260355963423498e-05,
+      "loss": 1.0553,
+      "step": 9031
+    },
+    {
+      "epoch": 0.5487575186827875,
+      "grad_norm": 0.18207204341888428,
+      "learning_rate": 4.2594092316352885e-05,
+      "loss": 1.1068,
+      "step": 9032
+    },
+    {
+      "epoch": 0.5488182757154141,
+      "grad_norm": 0.16854768991470337,
+      "learning_rate": 4.258462526993576e-05,
+      "loss": 1.0493,
+      "step": 9033
+    },
+    {
+      "epoch": 0.5488790327480406,
+      "grad_norm": 0.14280124008655548,
+      "learning_rate": 4.257515849533063e-05,
+      "loss": 1.079,
+      "step": 9034
+    },
+    {
+      "epoch": 0.5489397897806672,
+      "grad_norm": 0.13142609596252441,
+      "learning_rate": 4.25656919928845e-05,
+      "loss": 1.0689,
+      "step": 9035
+    },
+    {
+      "epoch": 0.5490005468132937,
+      "grad_norm": 0.22240585088729858,
+      "learning_rate": 4.255622576294438e-05,
+      "loss": 1.1404,
+      "step": 9036
+    },
+    {
+      "epoch": 0.5490613038459201,
+      "grad_norm": 0.1691696047782898,
+      "learning_rate": 4.254675980585723e-05,
+      "loss": 1.1479,
+      "step": 9037
+    },
+    {
+      "epoch": 0.5491220608785466,
+      "grad_norm": 0.22538401186466217,
+      "learning_rate": 4.253729412197004e-05,
+      "loss": 1.1245,
+      "step": 9038
+    },
+    {
+      "epoch": 0.5491828179111732,
+      "grad_norm": 0.1962536722421646,
+      "learning_rate": 4.2527828711629784e-05,
+      "loss": 1.0722,
+      "step": 9039
+    },
+    {
+      "epoch": 0.5492435749437997,
+      "grad_norm": 0.1319868564605713,
+      "learning_rate": 4.25183635751834e-05,
+      "loss": 1.0729,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5493043319764263,
+      "grad_norm": 0.17093893885612488,
+      "learning_rate": 4.250889871297785e-05,
+      "loss": 1.1271,
+      "step": 9041
+    },
+    {
+      "epoch": 0.5493650890090528,
+      "grad_norm": 0.132630854845047,
+      "learning_rate": 4.249943412536006e-05,
+      "loss": 1.0778,
+      "step": 9042
+    },
+    {
+      "epoch": 0.5494258460416793,
+      "grad_norm": 0.11211725324392319,
+      "learning_rate": 4.248996981267698e-05,
+      "loss": 1.0427,
+      "step": 9043
+    },
+    {
+      "epoch": 0.5494866030743059,
+      "grad_norm": 0.13270297646522522,
+      "learning_rate": 4.24805057752755e-05,
+      "loss": 1.0222,
+      "step": 9044
+    },
+    {
+      "epoch": 0.5495473601069324,
+      "grad_norm": 0.14900323748588562,
+      "learning_rate": 4.2471042013502534e-05,
+      "loss": 1.0251,
+      "step": 9045
+    },
+    {
+      "epoch": 0.5496081171395589,
+      "grad_norm": 0.866405189037323,
+      "learning_rate": 4.246157852770496e-05,
+      "loss": 1.2466,
+      "step": 9046
+    },
+    {
+      "epoch": 0.5496688741721855,
+      "grad_norm": 0.24184077978134155,
+      "learning_rate": 4.245211531822971e-05,
+      "loss": 1.1705,
+      "step": 9047
+    },
+    {
+      "epoch": 0.549729631204812,
+      "grad_norm": 0.19942255318164825,
+      "learning_rate": 4.244265238542362e-05,
+      "loss": 1.1178,
+      "step": 9048
+    },
+    {
+      "epoch": 0.5497903882374385,
+      "grad_norm": 0.11792849749326706,
+      "learning_rate": 4.2433189729633566e-05,
+      "loss": 1.0116,
+      "step": 9049
+    },
+    {
+      "epoch": 0.549851145270065,
+      "grad_norm": 0.14112327992916107,
+      "learning_rate": 4.242372735120642e-05,
+      "loss": 1.0749,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5499119023026915,
+      "grad_norm": 0.1418844312429428,
+      "learning_rate": 4.241426525048899e-05,
+      "loss": 1.0325,
+      "step": 9051
+    },
+    {
+      "epoch": 0.549972659335318,
+      "grad_norm": 0.15489812195301056,
+      "learning_rate": 4.240480342782814e-05,
+      "loss": 1.1121,
+      "step": 9052
+    },
+    {
+      "epoch": 0.5500334163679446,
+      "grad_norm": 0.14163701236248016,
+      "learning_rate": 4.239534188357069e-05,
+      "loss": 1.0602,
+      "step": 9053
+    },
+    {
+      "epoch": 0.5500941734005711,
+      "grad_norm": 0.11680836230516434,
+      "learning_rate": 4.238588061806345e-05,
+      "loss": 1.0147,
+      "step": 9054
+    },
+    {
+      "epoch": 0.5501549304331976,
+      "grad_norm": 0.20463886857032776,
+      "learning_rate": 4.237641963165323e-05,
+      "loss": 1.0429,
+      "step": 9055
+    },
+    {
+      "epoch": 0.5502156874658242,
+      "grad_norm": 0.4137899577617645,
+      "learning_rate": 4.236695892468684e-05,
+      "loss": 1.0282,
+      "step": 9056
+    },
+    {
+      "epoch": 0.5502764444984507,
+      "grad_norm": 48.9248161315918,
+      "learning_rate": 4.2357498497511026e-05,
+      "loss": 1.2152,
+      "step": 9057
+    },
+    {
+      "epoch": 0.5503372015310772,
+      "grad_norm": 0.3811236023902893,
+      "learning_rate": 4.234803835047259e-05,
+      "loss": 1.1267,
+      "step": 9058
+    },
+    {
+      "epoch": 0.5503979585637038,
+      "grad_norm": 0.12349487096071243,
+      "learning_rate": 4.233857848391828e-05,
+      "loss": 1.037,
+      "step": 9059
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 0.2883005738258362,
+      "learning_rate": 4.232911889819486e-05,
+      "loss": 1.1034,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5505194726289568,
+      "grad_norm": 0.21447651088237762,
+      "learning_rate": 4.231965959364907e-05,
+      "loss": 1.0715,
+      "step": 9061
+    },
+    {
+      "epoch": 0.5505802296615834,
+      "grad_norm": 0.1179421916604042,
+      "learning_rate": 4.231020057062763e-05,
+      "loss": 1.0245,
+      "step": 9062
+    },
+    {
+      "epoch": 0.5506409866942098,
+      "grad_norm": 0.1727469563484192,
+      "learning_rate": 4.230074182947731e-05,
+      "loss": 1.1492,
+      "step": 9063
+    },
+    {
+      "epoch": 0.5507017437268363,
+      "grad_norm": 1.4222290515899658,
+      "learning_rate": 4.2291283370544754e-05,
+      "loss": 1.0728,
+      "step": 9064
+    },
+    {
+      "epoch": 0.5507625007594629,
+      "grad_norm": 0.1375548392534256,
+      "learning_rate": 4.228182519417671e-05,
+      "loss": 1.0859,
+      "step": 9065
+    },
+    {
+      "epoch": 0.5508232577920894,
+      "grad_norm": 0.17478792369365692,
+      "learning_rate": 4.2272367300719846e-05,
+      "loss": 1.1092,
+      "step": 9066
+    },
+    {
+      "epoch": 0.5508840148247159,
+      "grad_norm": 0.163072407245636,
+      "learning_rate": 4.2262909690520845e-05,
+      "loss": 1.0759,
+      "step": 9067
+    },
+    {
+      "epoch": 0.5509447718573425,
+      "grad_norm": 0.1808890849351883,
+      "learning_rate": 4.22534523639264e-05,
+      "loss": 1.0488,
+      "step": 9068
+    },
+    {
+      "epoch": 0.551005528889969,
+      "grad_norm": 1.1838573217391968,
+      "learning_rate": 4.224399532128315e-05,
+      "loss": 1.0856,
+      "step": 9069
+    },
+    {
+      "epoch": 0.5510662859225955,
+      "grad_norm": 0.13643936812877655,
+      "learning_rate": 4.223453856293777e-05,
+      "loss": 1.0473,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5511270429552221,
+      "grad_norm": 0.36972951889038086,
+      "learning_rate": 4.222508208923686e-05,
+      "loss": 1.0227,
+      "step": 9071
+    },
+    {
+      "epoch": 0.5511877999878486,
+      "grad_norm": 0.1670450121164322,
+      "learning_rate": 4.221562590052707e-05,
+      "loss": 1.1461,
+      "step": 9072
+    },
+    {
+      "epoch": 0.5512485570204751,
+      "grad_norm": 0.20135442912578583,
+      "learning_rate": 4.220616999715501e-05,
+      "loss": 1.0979,
+      "step": 9073
+    },
+    {
+      "epoch": 0.5513093140531017,
+      "grad_norm": 0.12519875168800354,
+      "learning_rate": 4.219671437946731e-05,
+      "loss": 1.0521,
+      "step": 9074
+    },
+    {
+      "epoch": 0.5513700710857282,
+      "grad_norm": 0.7594452500343323,
+      "learning_rate": 4.218725904781054e-05,
+      "loss": 1.1522,
+      "step": 9075
+    },
+    {
+      "epoch": 0.5514308281183548,
+      "grad_norm": 0.15588246285915375,
+      "learning_rate": 4.2177804002531326e-05,
+      "loss": 1.1389,
+      "step": 9076
+    },
+    {
+      "epoch": 0.5514915851509812,
+      "grad_norm": 4.16948127746582,
+      "learning_rate": 4.2168349243976194e-05,
+      "loss": 1.1676,
+      "step": 9077
+    },
+    {
+      "epoch": 0.5515523421836077,
+      "grad_norm": 0.6780422329902649,
+      "learning_rate": 4.215889477249174e-05,
+      "loss": 1.1329,
+      "step": 9078
+    },
+    {
+      "epoch": 0.5516130992162342,
+      "grad_norm": 0.2095414400100708,
+      "learning_rate": 4.2149440588424515e-05,
+      "loss": 1.0194,
+      "step": 9079
+    },
+    {
+      "epoch": 0.5516738562488608,
+      "grad_norm": 0.1834208220243454,
+      "learning_rate": 4.2139986692121045e-05,
+      "loss": 1.162,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5517346132814873,
+      "grad_norm": 0.18039526045322418,
+      "learning_rate": 4.21305330839279e-05,
+      "loss": 1.0437,
+      "step": 9081
+    },
+    {
+      "epoch": 0.5517953703141139,
+      "grad_norm": 0.22322294116020203,
+      "learning_rate": 4.212107976419159e-05,
+      "loss": 1.0523,
+      "step": 9082
+    },
+    {
+      "epoch": 0.5518561273467404,
+      "grad_norm": 0.2128293216228485,
+      "learning_rate": 4.211162673325863e-05,
+      "loss": 1.1181,
+      "step": 9083
+    },
+    {
+      "epoch": 0.5519168843793669,
+      "grad_norm": 0.17772603034973145,
+      "learning_rate": 4.210217399147551e-05,
+      "loss": 1.1684,
+      "step": 9084
+    },
+    {
+      "epoch": 0.5519776414119935,
+      "grad_norm": 0.3734113872051239,
+      "learning_rate": 4.209272153918871e-05,
+      "loss": 1.0578,
+      "step": 9085
+    },
+    {
+      "epoch": 0.55203839844462,
+      "grad_norm": 0.2782934308052063,
+      "learning_rate": 4.208326937674475e-05,
+      "loss": 1.1397,
+      "step": 9086
+    },
+    {
+      "epoch": 0.5520991554772465,
+      "grad_norm": 0.6208287477493286,
+      "learning_rate": 4.2073817504490076e-05,
+      "loss": 1.1481,
+      "step": 9087
+    },
+    {
+      "epoch": 0.5521599125098731,
+      "grad_norm": 0.2628735899925232,
+      "learning_rate": 4.2064365922771145e-05,
+      "loss": 1.0267,
+      "step": 9088
+    },
+    {
+      "epoch": 0.5522206695424996,
+      "grad_norm": 0.19605694711208344,
+      "learning_rate": 4.205491463193442e-05,
+      "loss": 1.0894,
+      "step": 9089
+    },
+    {
+      "epoch": 0.552281426575126,
+      "grad_norm": 0.16692741215229034,
+      "learning_rate": 4.204546363232635e-05,
+      "loss": 1.0414,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5523421836077526,
+      "grad_norm": 0.11422184854745865,
+      "learning_rate": 4.203601292429334e-05,
+      "loss": 1.0077,
+      "step": 9091
+    },
+    {
+      "epoch": 0.5524029406403791,
+      "grad_norm": 0.19897115230560303,
+      "learning_rate": 4.202656250818181e-05,
+      "loss": 1.0989,
+      "step": 9092
+    },
+    {
+      "epoch": 0.5524636976730056,
+      "grad_norm": 0.1422666758298874,
+      "learning_rate": 4.201711238433817e-05,
+      "loss": 1.0557,
+      "step": 9093
+    },
+    {
+      "epoch": 0.5525244547056322,
+      "grad_norm": 0.20156043767929077,
+      "learning_rate": 4.2007662553108825e-05,
+      "loss": 1.029,
+      "step": 9094
+    },
+    {
+      "epoch": 0.5525852117382587,
+      "grad_norm": 0.19875234365463257,
+      "learning_rate": 4.199821301484015e-05,
+      "loss": 1.013,
+      "step": 9095
+    },
+    {
+      "epoch": 0.5526459687708852,
+      "grad_norm": 0.1583237498998642,
+      "learning_rate": 4.198876376987851e-05,
+      "loss": 1.0777,
+      "step": 9096
+    },
+    {
+      "epoch": 0.5527067258035118,
+      "grad_norm": 0.12764251232147217,
+      "learning_rate": 4.197931481857033e-05,
+      "loss": 1.0936,
+      "step": 9097
+    },
+    {
+      "epoch": 0.5527674828361383,
+      "grad_norm": 0.1488335132598877,
+      "learning_rate": 4.196986616126187e-05,
+      "loss": 1.0507,
+      "step": 9098
+    },
+    {
+      "epoch": 0.5528282398687648,
+      "grad_norm": 0.16056792438030243,
+      "learning_rate": 4.1960417798299536e-05,
+      "loss": 1.0813,
+      "step": 9099
+    },
+    {
+      "epoch": 0.5528889969013914,
+      "grad_norm": 0.1478900909423828,
+      "learning_rate": 4.195096973002964e-05,
+      "loss": 1.0775,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5529497539340179,
+      "grad_norm": 0.17586633563041687,
+      "learning_rate": 4.194152195679849e-05,
+      "loss": 1.105,
+      "step": 9101
+    },
+    {
+      "epoch": 0.5530105109666444,
+      "grad_norm": 0.12817417085170746,
+      "learning_rate": 4.193207447895243e-05,
+      "loss": 1.084,
+      "step": 9102
+    },
+    {
+      "epoch": 0.5530712679992709,
+      "grad_norm": 0.1798471212387085,
+      "learning_rate": 4.192262729683774e-05,
+      "loss": 1.1244,
+      "step": 9103
+    },
+    {
+      "epoch": 0.5531320250318974,
+      "grad_norm": 0.39058050513267517,
+      "learning_rate": 4.19131804108007e-05,
+      "loss": 1.172,
+      "step": 9104
+    },
+    {
+      "epoch": 0.5531927820645239,
+      "grad_norm": 0.1737455278635025,
+      "learning_rate": 4.1903733821187594e-05,
+      "loss": 1.0815,
+      "step": 9105
+    },
+    {
+      "epoch": 0.5532535390971505,
+      "grad_norm": 0.21475253999233246,
+      "learning_rate": 4.189428752834468e-05,
+      "loss": 1.1136,
+      "step": 9106
+    },
+    {
+      "epoch": 0.553314296129777,
+      "grad_norm": 0.6660194993019104,
+      "learning_rate": 4.188484153261824e-05,
+      "loss": 1.41,
+      "step": 9107
+    },
+    {
+      "epoch": 0.5533750531624035,
+      "grad_norm": 1.6338199377059937,
+      "learning_rate": 4.18753958343545e-05,
+      "loss": 1.0513,
+      "step": 9108
+    },
+    {
+      "epoch": 0.5534358101950301,
+      "grad_norm": 0.2118992656469345,
+      "learning_rate": 4.186595043389968e-05,
+      "loss": 1.0665,
+      "step": 9109
+    },
+    {
+      "epoch": 0.5534965672276566,
+      "grad_norm": 0.20369698107242584,
+      "learning_rate": 4.185650533160005e-05,
+      "loss": 1.0997,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5535573242602831,
+      "grad_norm": 0.13237892091274261,
+      "learning_rate": 4.1847060527801766e-05,
+      "loss": 1.0507,
+      "step": 9111
+    },
+    {
+      "epoch": 0.5536180812929097,
+      "grad_norm": 0.29496198892593384,
+      "learning_rate": 4.183761602285106e-05,
+      "loss": 1.0301,
+      "step": 9112
+    },
+    {
+      "epoch": 0.5536788383255362,
+      "grad_norm": 0.1493644416332245,
+      "learning_rate": 4.182817181709412e-05,
+      "loss": 1.0194,
+      "step": 9113
+    },
+    {
+      "epoch": 0.5537395953581628,
+      "grad_norm": 0.1697172373533249,
+      "learning_rate": 4.181872791087711e-05,
+      "loss": 1.091,
+      "step": 9114
+    },
+    {
+      "epoch": 0.5538003523907893,
+      "grad_norm": 0.1491381973028183,
+      "learning_rate": 4.180928430454622e-05,
+      "loss": 1.0115,
+      "step": 9115
+    },
+    {
+      "epoch": 0.5538611094234157,
+      "grad_norm": 0.15601514279842377,
+      "learning_rate": 4.17998409984476e-05,
+      "loss": 1.0507,
+      "step": 9116
+    },
+    {
+      "epoch": 0.5539218664560422,
+      "grad_norm": 0.14003917574882507,
+      "learning_rate": 4.17903979929274e-05,
+      "loss": 1.0675,
+      "step": 9117
+    },
+    {
+      "epoch": 0.5539826234886688,
+      "grad_norm": 0.14665600657463074,
+      "learning_rate": 4.178095528833174e-05,
+      "loss": 1.0589,
+      "step": 9118
+    },
+    {
+      "epoch": 0.5540433805212953,
+      "grad_norm": 0.14360341429710388,
+      "learning_rate": 4.177151288500674e-05,
+      "loss": 1.0291,
+      "step": 9119
+    },
+    {
+      "epoch": 0.5541041375539218,
+      "grad_norm": 0.1472918689250946,
+      "learning_rate": 4.176207078329853e-05,
+      "loss": 1.0682,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5541648945865484,
+      "grad_norm": 0.21565595269203186,
+      "learning_rate": 4.175262898355322e-05,
+      "loss": 1.0591,
+      "step": 9121
+    },
+    {
+      "epoch": 0.5542256516191749,
+      "grad_norm": 0.2663908898830414,
+      "learning_rate": 4.174318748611687e-05,
+      "loss": 1.1274,
+      "step": 9122
+    },
+    {
+      "epoch": 0.5542864086518015,
+      "grad_norm": 0.16110442578792572,
+      "learning_rate": 4.1733746291335616e-05,
+      "loss": 1.0243,
+      "step": 9123
+    },
+    {
+      "epoch": 0.554347165684428,
+      "grad_norm": 0.11592574417591095,
+      "learning_rate": 4.172430539955546e-05,
+      "loss": 1.0567,
+      "step": 9124
+    },
+    {
+      "epoch": 0.5544079227170545,
+      "grad_norm": 0.1892186403274536,
+      "learning_rate": 4.1714864811122495e-05,
+      "loss": 1.21,
+      "step": 9125
+    },
+    {
+      "epoch": 0.5544686797496811,
+      "grad_norm": 0.15835748612880707,
+      "learning_rate": 4.1705424526382765e-05,
+      "loss": 1.0296,
+      "step": 9126
+    },
+    {
+      "epoch": 0.5545294367823076,
+      "grad_norm": 0.7885303497314453,
+      "learning_rate": 4.1695984545682285e-05,
+      "loss": 1.0375,
+      "step": 9127
+    },
+    {
+      "epoch": 0.5545901938149341,
+      "grad_norm": 0.37598615884780884,
+      "learning_rate": 4.1686544869367114e-05,
+      "loss": 1.0165,
+      "step": 9128
+    },
+    {
+      "epoch": 0.5546509508475606,
+      "grad_norm": 0.14977487921714783,
+      "learning_rate": 4.167710549778324e-05,
+      "loss": 1.0525,
+      "step": 9129
+    },
+    {
+      "epoch": 0.5547117078801871,
+      "grad_norm": 0.12446727603673935,
+      "learning_rate": 4.166766643127669e-05,
+      "loss": 1.0609,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5547724649128136,
+      "grad_norm": 1.2196428775787354,
+      "learning_rate": 4.165822767019343e-05,
+      "loss": 1.1014,
+      "step": 9131
+    },
+    {
+      "epoch": 0.5548332219454402,
+      "grad_norm": 0.18719960749149323,
+      "learning_rate": 4.164878921487943e-05,
+      "loss": 1.0826,
+      "step": 9132
+    },
+    {
+      "epoch": 0.5548939789780667,
+      "grad_norm": 0.15722621977329254,
+      "learning_rate": 4.163935106568069e-05,
+      "loss": 1.1298,
+      "step": 9133
+    },
+    {
+      "epoch": 0.5549547360106932,
+      "grad_norm": 0.17909014225006104,
+      "learning_rate": 4.1629913222943146e-05,
+      "loss": 1.1346,
+      "step": 9134
+    },
+    {
+      "epoch": 0.5550154930433198,
+      "grad_norm": 0.14320820569992065,
+      "learning_rate": 4.162047568701275e-05,
+      "loss": 1.1234,
+      "step": 9135
+    },
+    {
+      "epoch": 0.5550762500759463,
+      "grad_norm": 0.1808628886938095,
+      "learning_rate": 4.1611038458235435e-05,
+      "loss": 1.2227,
+      "step": 9136
+    },
+    {
+      "epoch": 0.5551370071085728,
+      "grad_norm": 0.12627863883972168,
+      "learning_rate": 4.1601601536957144e-05,
+      "loss": 1.0702,
+      "step": 9137
+    },
+    {
+      "epoch": 0.5551977641411994,
+      "grad_norm": 0.18002478778362274,
+      "learning_rate": 4.159216492352377e-05,
+      "loss": 1.0489,
+      "step": 9138
+    },
+    {
+      "epoch": 0.5552585211738259,
+      "grad_norm": 0.14645852148532867,
+      "learning_rate": 4.15827286182812e-05,
+      "loss": 1.0984,
+      "step": 9139
+    },
+    {
+      "epoch": 0.5553192782064524,
+      "grad_norm": 0.11596376448869705,
+      "learning_rate": 4.1573292621575344e-05,
+      "loss": 1.0727,
+      "step": 9140
+    },
+    {
+      "epoch": 0.555380035239079,
+      "grad_norm": 0.1481286734342575,
+      "learning_rate": 4.1563856933752075e-05,
+      "loss": 1.0694,
+      "step": 9141
+    },
+    {
+      "epoch": 0.5554407922717054,
+      "grad_norm": 0.16426613926887512,
+      "learning_rate": 4.1554421555157266e-05,
+      "loss": 1.047,
+      "step": 9142
+    },
+    {
+      "epoch": 0.5555015493043319,
+      "grad_norm": 0.1251041740179062,
+      "learning_rate": 4.1544986486136784e-05,
+      "loss": 1.0583,
+      "step": 9143
+    },
+    {
+      "epoch": 0.5555623063369585,
+      "grad_norm": 0.6372851729393005,
+      "learning_rate": 4.153555172703643e-05,
+      "loss": 1.0703,
+      "step": 9144
+    },
+    {
+      "epoch": 0.555623063369585,
+      "grad_norm": 0.6742607951164246,
+      "learning_rate": 4.152611727820207e-05,
+      "loss": 1.2313,
+      "step": 9145
+    },
+    {
+      "epoch": 0.5556838204022115,
+      "grad_norm": 0.2668459117412567,
+      "learning_rate": 4.151668313997952e-05,
+      "loss": 1.0941,
+      "step": 9146
+    },
+    {
+      "epoch": 0.5557445774348381,
+      "grad_norm": 0.1434379518032074,
+      "learning_rate": 4.150724931271458e-05,
+      "loss": 1.0497,
+      "step": 9147
+    },
+    {
+      "epoch": 0.5558053344674646,
+      "grad_norm": 0.18460677564144135,
+      "learning_rate": 4.149781579675307e-05,
+      "loss": 1.1759,
+      "step": 9148
+    },
+    {
+      "epoch": 0.5558660915000911,
+      "grad_norm": 0.33510252833366394,
+      "learning_rate": 4.148838259244077e-05,
+      "loss": 1.0058,
+      "step": 9149
+    },
+    {
+      "epoch": 0.5559268485327177,
+      "grad_norm": 0.14448833465576172,
+      "learning_rate": 4.147894970012345e-05,
+      "loss": 1.0497,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5559876055653442,
+      "grad_norm": 0.1655113399028778,
+      "learning_rate": 4.146951712014688e-05,
+      "loss": 1.0536,
+      "step": 9151
+    },
+    {
+      "epoch": 0.5560483625979707,
+      "grad_norm": 0.12680092453956604,
+      "learning_rate": 4.146008485285678e-05,
+      "loss": 1.0814,
+      "step": 9152
+    },
+    {
+      "epoch": 0.5561091196305973,
+      "grad_norm": 0.24189986288547516,
+      "learning_rate": 4.145065289859895e-05,
+      "loss": 1.177,
+      "step": 9153
+    },
+    {
+      "epoch": 0.5561698766632238,
+      "grad_norm": 0.1563245952129364,
+      "learning_rate": 4.144122125771908e-05,
+      "loss": 1.0078,
+      "step": 9154
+    },
+    {
+      "epoch": 0.5562306336958502,
+      "grad_norm": 0.119784876704216,
+      "learning_rate": 4.143178993056289e-05,
+      "loss": 1.066,
+      "step": 9155
+    },
+    {
+      "epoch": 0.5562913907284768,
+      "grad_norm": 0.17714856564998627,
+      "learning_rate": 4.1422358917476114e-05,
+      "loss": 1.1419,
+      "step": 9156
+    },
+    {
+      "epoch": 0.5563521477611033,
+      "grad_norm": 0.13944889605045319,
+      "learning_rate": 4.1412928218804434e-05,
+      "loss": 1.063,
+      "step": 9157
+    },
+    {
+      "epoch": 0.5564129047937298,
+      "grad_norm": 0.15701355040073395,
+      "learning_rate": 4.140349783489352e-05,
+      "loss": 1.1095,
+      "step": 9158
+    },
+    {
+      "epoch": 0.5564736618263564,
+      "grad_norm": 0.1993296891450882,
+      "learning_rate": 4.139406776608905e-05,
+      "loss": 1.1329,
+      "step": 9159
+    },
+    {
+      "epoch": 0.5565344188589829,
+      "grad_norm": 1.117037296295166,
+      "learning_rate": 4.138463801273668e-05,
+      "loss": 1.0801,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5565951758916095,
+      "grad_norm": 0.14102678000926971,
+      "learning_rate": 4.1375208575182085e-05,
+      "loss": 1.1481,
+      "step": 9161
+    },
+    {
+      "epoch": 0.556655932924236,
+      "grad_norm": 0.19037580490112305,
+      "learning_rate": 4.136577945377088e-05,
+      "loss": 1.0346,
+      "step": 9162
+    },
+    {
+      "epoch": 0.5567166899568625,
+      "grad_norm": 0.48535385727882385,
+      "learning_rate": 4.13563506488487e-05,
+      "loss": 1.1743,
+      "step": 9163
+    },
+    {
+      "epoch": 0.556777446989489,
+      "grad_norm": 0.20781023800373077,
+      "learning_rate": 4.1346922160761184e-05,
+      "loss": 1.1715,
+      "step": 9164
+    },
+    {
+      "epoch": 0.5568382040221156,
+      "grad_norm": 0.15947094559669495,
+      "learning_rate": 4.133749398985387e-05,
+      "loss": 1.1245,
+      "step": 9165
+    },
+    {
+      "epoch": 0.5568989610547421,
+      "grad_norm": 0.17786350846290588,
+      "learning_rate": 4.1328066136472405e-05,
+      "loss": 1.2078,
+      "step": 9166
+    },
+    {
+      "epoch": 0.5569597180873687,
+      "grad_norm": 0.16088789701461792,
+      "learning_rate": 4.1318638600962346e-05,
+      "loss": 1.127,
+      "step": 9167
+    },
+    {
+      "epoch": 0.5570204751199951,
+      "grad_norm": 0.12699060142040253,
+      "learning_rate": 4.130921138366926e-05,
+      "loss": 1.0874,
+      "step": 9168
+    },
+    {
+      "epoch": 0.5570812321526216,
+      "grad_norm": 0.12009410560131073,
+      "learning_rate": 4.1299784484938726e-05,
+      "loss": 1.0541,
+      "step": 9169
+    },
+    {
+      "epoch": 0.5571419891852482,
+      "grad_norm": 0.12249850481748581,
+      "learning_rate": 4.1290357905116276e-05,
+      "loss": 1.0293,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5572027462178747,
+      "grad_norm": 0.22208324074745178,
+      "learning_rate": 4.1280931644547435e-05,
+      "loss": 1.1401,
+      "step": 9171
+    },
+    {
+      "epoch": 0.5572635032505012,
+      "grad_norm": 0.12204176187515259,
+      "learning_rate": 4.1271505703577726e-05,
+      "loss": 1.067,
+      "step": 9172
+    },
+    {
+      "epoch": 0.5573242602831278,
+      "grad_norm": 0.11667222529649734,
+      "learning_rate": 4.1262080082552656e-05,
+      "loss": 1.0292,
+      "step": 9173
+    },
+    {
+      "epoch": 0.5573850173157543,
+      "grad_norm": 0.24544478952884674,
+      "learning_rate": 4.1252654781817746e-05,
+      "loss": 1.2436,
+      "step": 9174
+    },
+    {
+      "epoch": 0.5574457743483808,
+      "grad_norm": 0.17343677580356598,
+      "learning_rate": 4.124322980171845e-05,
+      "loss": 1.1382,
+      "step": 9175
+    },
+    {
+      "epoch": 0.5575065313810074,
+      "grad_norm": 0.19015738368034363,
+      "learning_rate": 4.1233805142600265e-05,
+      "loss": 1.0339,
+      "step": 9176
+    },
+    {
+      "epoch": 0.5575672884136339,
+      "grad_norm": 0.18894977867603302,
+      "learning_rate": 4.122438080480867e-05,
+      "loss": 1.0983,
+      "step": 9177
+    },
+    {
+      "epoch": 0.5576280454462604,
+      "grad_norm": 0.23062144219875336,
+      "learning_rate": 4.121495678868906e-05,
+      "loss": 1.0913,
+      "step": 9178
+    },
+    {
+      "epoch": 0.557688802478887,
+      "grad_norm": 0.18861408531665802,
+      "learning_rate": 4.1205533094586916e-05,
+      "loss": 1.1221,
+      "step": 9179
+    },
+    {
+      "epoch": 0.5577495595115135,
+      "grad_norm": 0.18230387568473816,
+      "learning_rate": 4.1196109722847666e-05,
+      "loss": 1.1506,
+      "step": 9180
+    },
+    {
+      "epoch": 0.55781031654414,
+      "grad_norm": 0.5075845718383789,
+      "learning_rate": 4.118668667381669e-05,
+      "loss": 1.0723,
+      "step": 9181
+    },
+    {
+      "epoch": 0.5578710735767665,
+      "grad_norm": 0.1514529436826706,
+      "learning_rate": 4.117726394783944e-05,
+      "loss": 1.0702,
+      "step": 9182
+    },
+    {
+      "epoch": 0.557931830609393,
+      "grad_norm": 0.27151474356651306,
+      "learning_rate": 4.116784154526128e-05,
+      "loss": 1.1091,
+      "step": 9183
+    },
+    {
+      "epoch": 0.5579925876420195,
+      "grad_norm": 2.193075180053711,
+      "learning_rate": 4.115841946642761e-05,
+      "loss": 1.0424,
+      "step": 9184
+    },
+    {
+      "epoch": 0.5580533446746461,
+      "grad_norm": 0.16766516864299774,
+      "learning_rate": 4.114899771168377e-05,
+      "loss": 1.1557,
+      "step": 9185
+    },
+    {
+      "epoch": 0.5581141017072726,
+      "grad_norm": 0.2055845707654953,
+      "learning_rate": 4.1139576281375125e-05,
+      "loss": 1.1078,
+      "step": 9186
+    },
+    {
+      "epoch": 0.5581748587398991,
+      "grad_norm": 0.1501486748456955,
+      "learning_rate": 4.1130155175847026e-05,
+      "loss": 1.0784,
+      "step": 9187
+    },
+    {
+      "epoch": 0.5582356157725257,
+      "grad_norm": 0.7084944248199463,
+      "learning_rate": 4.112073439544481e-05,
+      "loss": 1.0515,
+      "step": 9188
+    },
+    {
+      "epoch": 0.5582963728051522,
+      "grad_norm": 0.16138140857219696,
+      "learning_rate": 4.1111313940513776e-05,
+      "loss": 1.0415,
+      "step": 9189
+    },
+    {
+      "epoch": 0.5583571298377787,
+      "grad_norm": 0.23819054663181305,
+      "learning_rate": 4.110189381139928e-05,
+      "loss": 1.1669,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5584178868704053,
+      "grad_norm": 0.13890103995800018,
+      "learning_rate": 4.109247400844656e-05,
+      "loss": 1.076,
+      "step": 9191
+    },
+    {
+      "epoch": 0.5584786439030318,
+      "grad_norm": 0.23677076399326324,
+      "learning_rate": 4.108305453200093e-05,
+      "loss": 1.1462,
+      "step": 9192
+    },
+    {
+      "epoch": 0.5585394009356583,
+      "grad_norm": 0.21250909566879272,
+      "learning_rate": 4.107363538240767e-05,
+      "loss": 1.15,
+      "step": 9193
+    },
+    {
+      "epoch": 0.5586001579682849,
+      "grad_norm": 0.11922785639762878,
+      "learning_rate": 4.106421656001201e-05,
+      "loss": 0.9903,
+      "step": 9194
+    },
+    {
+      "epoch": 0.5586609150009113,
+      "grad_norm": 0.1227865144610405,
+      "learning_rate": 4.105479806515923e-05,
+      "loss": 1.0515,
+      "step": 9195
+    },
+    {
+      "epoch": 0.5587216720335378,
+      "grad_norm": 0.19078896939754486,
+      "learning_rate": 4.104537989819456e-05,
+      "loss": 1.1567,
+      "step": 9196
+    },
+    {
+      "epoch": 0.5587824290661644,
+      "grad_norm": 0.24832095205783844,
+      "learning_rate": 4.103596205946323e-05,
+      "loss": 1.0544,
+      "step": 9197
+    },
+    {
+      "epoch": 0.5588431860987909,
+      "grad_norm": 0.2160288244485855,
+      "learning_rate": 4.102654454931043e-05,
+      "loss": 1.1281,
+      "step": 9198
+    },
+    {
+      "epoch": 0.5589039431314174,
+      "grad_norm": 0.1606760025024414,
+      "learning_rate": 4.101712736808137e-05,
+      "loss": 1.014,
+      "step": 9199
+    },
+    {
+      "epoch": 0.558964700164044,
+      "grad_norm": 2.069169282913208,
+      "learning_rate": 4.100771051612125e-05,
+      "loss": 1.1125,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5590254571966705,
+      "grad_norm": 0.28553861379623413,
+      "learning_rate": 4.0998293993775237e-05,
+      "loss": 1.1188,
+      "step": 9201
+    },
+    {
+      "epoch": 0.559086214229297,
+      "grad_norm": 0.21476872265338898,
+      "learning_rate": 4.098887780138849e-05,
+      "loss": 1.0517,
+      "step": 9202
+    },
+    {
+      "epoch": 0.5591469712619236,
+      "grad_norm": 0.2278563678264618,
+      "learning_rate": 4.097946193930618e-05,
+      "loss": 1.1718,
+      "step": 9203
+    },
+    {
+      "epoch": 0.5592077282945501,
+      "grad_norm": 0.9798266291618347,
+      "learning_rate": 4.0970046407873444e-05,
+      "loss": 1.2012,
+      "step": 9204
+    },
+    {
+      "epoch": 0.5592684853271767,
+      "grad_norm": 1.392189621925354,
+      "learning_rate": 4.0960631207435394e-05,
+      "loss": 1.0678,
+      "step": 9205
+    },
+    {
+      "epoch": 0.5593292423598032,
+      "grad_norm": 0.19495056569576263,
+      "learning_rate": 4.095121633833715e-05,
+      "loss": 1.1045,
+      "step": 9206
+    },
+    {
+      "epoch": 0.5593899993924297,
+      "grad_norm": 0.18605756759643555,
+      "learning_rate": 4.094180180092382e-05,
+      "loss": 1.0737,
+      "step": 9207
+    },
+    {
+      "epoch": 0.5594507564250562,
+      "grad_norm": 0.13544100522994995,
+      "learning_rate": 4.0932387595540496e-05,
+      "loss": 1.2418,
+      "step": 9208
+    },
+    {
+      "epoch": 0.5595115134576827,
+      "grad_norm": 0.1522100865840912,
+      "learning_rate": 4.0922973722532264e-05,
+      "loss": 1.1121,
+      "step": 9209
+    },
+    {
+      "epoch": 0.5595722704903092,
+      "grad_norm": 0.17265434563159943,
+      "learning_rate": 4.091356018224419e-05,
+      "loss": 1.0583,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5596330275229358,
+      "grad_norm": 0.18241240084171295,
+      "learning_rate": 4.090414697502132e-05,
+      "loss": 1.1417,
+      "step": 9211
+    },
+    {
+      "epoch": 0.5596937845555623,
+      "grad_norm": 0.13621163368225098,
+      "learning_rate": 4.089473410120868e-05,
+      "loss": 1.0356,
+      "step": 9212
+    },
+    {
+      "epoch": 0.5597545415881888,
+      "grad_norm": 0.20651018619537354,
+      "learning_rate": 4.088532156115134e-05,
+      "loss": 1.0771,
+      "step": 9213
+    },
+    {
+      "epoch": 0.5598152986208154,
+      "grad_norm": 1.1205204725265503,
+      "learning_rate": 4.08759093551943e-05,
+      "loss": 1.0865,
+      "step": 9214
+    },
+    {
+      "epoch": 0.5598760556534419,
+      "grad_norm": 0.15205805003643036,
+      "learning_rate": 4.0866497483682556e-05,
+      "loss": 1.0847,
+      "step": 9215
+    },
+    {
+      "epoch": 0.5599368126860684,
+      "grad_norm": 0.29775378108024597,
+      "learning_rate": 4.0857085946961115e-05,
+      "loss": 1.2864,
+      "step": 9216
+    },
+    {
+      "epoch": 0.559997569718695,
+      "grad_norm": 0.13234169781208038,
+      "learning_rate": 4.0847674745374974e-05,
+      "loss": 1.0042,
+      "step": 9217
+    },
+    {
+      "epoch": 0.5600583267513215,
+      "grad_norm": 0.16491706669330597,
+      "learning_rate": 4.0838263879269073e-05,
+      "loss": 1.0903,
+      "step": 9218
+    },
+    {
+      "epoch": 0.560119083783948,
+      "grad_norm": 0.1890789419412613,
+      "learning_rate": 4.082885334898838e-05,
+      "loss": 1.0531,
+      "step": 9219
+    },
+    {
+      "epoch": 0.5601798408165746,
+      "grad_norm": 0.16434350609779358,
+      "learning_rate": 4.081944315487783e-05,
+      "loss": 1.1153,
+      "step": 9220
+    },
+    {
+      "epoch": 0.560240597849201,
+      "grad_norm": 0.7292875051498413,
+      "learning_rate": 4.0810033297282375e-05,
+      "loss": 1.0567,
+      "step": 9221
+    },
+    {
+      "epoch": 0.5603013548818275,
+      "grad_norm": 0.16977371275424957,
+      "learning_rate": 4.080062377654692e-05,
+      "loss": 1.1081,
+      "step": 9222
+    },
+    {
+      "epoch": 0.5603621119144541,
+      "grad_norm": 0.18402603268623352,
+      "learning_rate": 4.079121459301638e-05,
+      "loss": 1.1288,
+      "step": 9223
+    },
+    {
+      "epoch": 0.5604228689470806,
+      "grad_norm": 0.12298519909381866,
+      "learning_rate": 4.078180574703567e-05,
+      "loss": 1.0805,
+      "step": 9224
+    },
+    {
+      "epoch": 0.5604836259797071,
+      "grad_norm": 0.1397022157907486,
+      "learning_rate": 4.077239723894963e-05,
+      "loss": 1.0769,
+      "step": 9225
+    },
+    {
+      "epoch": 0.5605443830123337,
+      "grad_norm": 0.2981606125831604,
+      "learning_rate": 4.0762989069103156e-05,
+      "loss": 1.0848,
+      "step": 9226
+    },
+    {
+      "epoch": 0.5606051400449602,
+      "grad_norm": 0.1549970656633377,
+      "learning_rate": 4.075358123784111e-05,
+      "loss": 1.1055,
+      "step": 9227
+    },
+    {
+      "epoch": 0.5606658970775867,
+      "grad_norm": 0.19356030225753784,
+      "learning_rate": 4.074417374550831e-05,
+      "loss": 1.1577,
+      "step": 9228
+    },
+    {
+      "epoch": 0.5607266541102133,
+      "grad_norm": 0.278775155544281,
+      "learning_rate": 4.073476659244963e-05,
+      "loss": 1.0363,
+      "step": 9229
+    },
+    {
+      "epoch": 0.5607874111428398,
+      "grad_norm": 0.26943546533584595,
+      "learning_rate": 4.072535977900988e-05,
+      "loss": 1.0827,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5608481681754663,
+      "grad_norm": 0.19701923429965973,
+      "learning_rate": 4.071595330553385e-05,
+      "loss": 1.1149,
+      "step": 9231
+    },
+    {
+      "epoch": 0.5609089252080929,
+      "grad_norm": 0.13717126846313477,
+      "learning_rate": 4.070654717236634e-05,
+      "loss": 1.0747,
+      "step": 9232
+    },
+    {
+      "epoch": 0.5609696822407194,
+      "grad_norm": 0.15929082036018372,
+      "learning_rate": 4.069714137985214e-05,
+      "loss": 1.0392,
+      "step": 9233
+    },
+    {
+      "epoch": 0.5610304392733458,
+      "grad_norm": 0.2341483235359192,
+      "learning_rate": 4.0687735928336024e-05,
+      "loss": 1.1599,
+      "step": 9234
+    },
+    {
+      "epoch": 0.5610911963059724,
+      "grad_norm": 0.24456939101219177,
+      "learning_rate": 4.067833081816275e-05,
+      "loss": 0.9986,
+      "step": 9235
+    },
+    {
+      "epoch": 0.5611519533385989,
+      "grad_norm": 0.1813986897468567,
+      "learning_rate": 4.066892604967705e-05,
+      "loss": 1.1335,
+      "step": 9236
+    },
+    {
+      "epoch": 0.5612127103712254,
+      "grad_norm": 0.13461871445178986,
+      "learning_rate": 4.06595216232237e-05,
+      "loss": 1.0855,
+      "step": 9237
+    },
+    {
+      "epoch": 0.561273467403852,
+      "grad_norm": 0.2357327789068222,
+      "learning_rate": 4.0650117539147366e-05,
+      "loss": 1.1281,
+      "step": 9238
+    },
+    {
+      "epoch": 0.5613342244364785,
+      "grad_norm": 0.9234658479690552,
+      "learning_rate": 4.064071379779278e-05,
+      "loss": 1.1923,
+      "step": 9239
+    },
+    {
+      "epoch": 0.561394981469105,
+      "grad_norm": 0.42659205198287964,
+      "learning_rate": 4.0631310399504634e-05,
+      "loss": 1.2212,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5614557385017316,
+      "grad_norm": 0.15744329988956451,
+      "learning_rate": 4.062190734462763e-05,
+      "loss": 1.1453,
+      "step": 9241
+    },
+    {
+      "epoch": 0.5615164955343581,
+      "grad_norm": 0.22790741920471191,
+      "learning_rate": 4.061250463350642e-05,
+      "loss": 1.2416,
+      "step": 9242
+    },
+    {
+      "epoch": 0.5615772525669847,
+      "grad_norm": 0.20907776057720184,
+      "learning_rate": 4.060310226648567e-05,
+      "loss": 1.0399,
+      "step": 9243
+    },
+    {
+      "epoch": 0.5616380095996112,
+      "grad_norm": 0.16667909920215607,
+      "learning_rate": 4.059370024391003e-05,
+      "loss": 1.1031,
+      "step": 9244
+    },
+    {
+      "epoch": 0.5616987666322377,
+      "grad_norm": 0.3520148992538452,
+      "learning_rate": 4.058429856612411e-05,
+      "loss": 1.1062,
+      "step": 9245
+    },
+    {
+      "epoch": 0.5617595236648643,
+      "grad_norm": 0.21632179617881775,
+      "learning_rate": 4.057489723347255e-05,
+      "loss": 1.126,
+      "step": 9246
+    },
+    {
+      "epoch": 0.5618202806974907,
+      "grad_norm": 0.2084217667579651,
+      "learning_rate": 4.056549624629996e-05,
+      "loss": 1.0224,
+      "step": 9247
+    },
+    {
+      "epoch": 0.5618810377301172,
+      "grad_norm": 0.2183791548013687,
+      "learning_rate": 4.055609560495092e-05,
+      "loss": 1.0627,
+      "step": 9248
+    },
+    {
+      "epoch": 0.5619417947627438,
+      "grad_norm": 8.142712593078613,
+      "learning_rate": 4.054669530977002e-05,
+      "loss": 1.0978,
+      "step": 9249
+    },
+    {
+      "epoch": 0.5620025517953703,
+      "grad_norm": 0.13680914044380188,
+      "learning_rate": 4.053729536110184e-05,
+      "loss": 1.0367,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5620633088279968,
+      "grad_norm": 0.21016928553581238,
+      "learning_rate": 4.052789575929094e-05,
+      "loss": 1.0802,
+      "step": 9251
+    },
+    {
+      "epoch": 0.5621240658606234,
+      "grad_norm": 0.13444884121418,
+      "learning_rate": 4.051849650468184e-05,
+      "loss": 1.0312,
+      "step": 9252
+    },
+    {
+      "epoch": 0.5621848228932499,
+      "grad_norm": 0.5965390205383301,
+      "learning_rate": 4.050909759761907e-05,
+      "loss": 1.0374,
+      "step": 9253
+    },
+    {
+      "epoch": 0.5622455799258764,
+      "grad_norm": 0.23923075199127197,
+      "learning_rate": 4.049969903844717e-05,
+      "loss": 1.1398,
+      "step": 9254
+    },
+    {
+      "epoch": 0.562306336958503,
+      "grad_norm": 0.15571463108062744,
+      "learning_rate": 4.049030082751064e-05,
+      "loss": 1.0744,
+      "step": 9255
+    },
+    {
+      "epoch": 0.5623670939911295,
+      "grad_norm": 0.19940346479415894,
+      "learning_rate": 4.048090296515396e-05,
+      "loss": 1.1872,
+      "step": 9256
+    },
+    {
+      "epoch": 0.562427851023756,
+      "grad_norm": 0.1466224640607834,
+      "learning_rate": 4.047150545172166e-05,
+      "loss": 1.0977,
+      "step": 9257
+    },
+    {
+      "epoch": 0.5624886080563826,
+      "grad_norm": 0.1474834829568863,
+      "learning_rate": 4.046210828755813e-05,
+      "loss": 1.1366,
+      "step": 9258
+    },
+    {
+      "epoch": 0.5625493650890091,
+      "grad_norm": 0.13714270293712616,
+      "learning_rate": 4.0452711473007873e-05,
+      "loss": 1.0169,
+      "step": 9259
+    },
+    {
+      "epoch": 0.5626101221216355,
+      "grad_norm": 0.15613627433776855,
+      "learning_rate": 4.0443315008415334e-05,
+      "loss": 1.1062,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5626708791542621,
+      "grad_norm": 1.8506195545196533,
+      "learning_rate": 4.043391889412491e-05,
+      "loss": 1.0487,
+      "step": 9261
+    },
+    {
+      "epoch": 0.5627316361868886,
+      "grad_norm": 0.1302403211593628,
+      "learning_rate": 4.042452313048105e-05,
+      "loss": 1.0356,
+      "step": 9262
+    },
+    {
+      "epoch": 0.5627923932195151,
+      "grad_norm": 0.21262824535369873,
+      "learning_rate": 4.041512771782815e-05,
+      "loss": 1.2119,
+      "step": 9263
+    },
+    {
+      "epoch": 0.5628531502521417,
+      "grad_norm": 0.15733003616333008,
+      "learning_rate": 4.040573265651061e-05,
+      "loss": 1.1157,
+      "step": 9264
+    },
+    {
+      "epoch": 0.5629139072847682,
+      "grad_norm": 0.24077053368091583,
+      "learning_rate": 4.0396337946872774e-05,
+      "loss": 1.1276,
+      "step": 9265
+    },
+    {
+      "epoch": 0.5629746643173947,
+      "grad_norm": 0.16773982346057892,
+      "learning_rate": 4.038694358925902e-05,
+      "loss": 1.0722,
+      "step": 9266
+    },
+    {
+      "epoch": 0.5630354213500213,
+      "grad_norm": 0.15182141959667206,
+      "learning_rate": 4.037754958401372e-05,
+      "loss": 1.0635,
+      "step": 9267
+    },
+    {
+      "epoch": 0.5630961783826478,
+      "grad_norm": 0.4050828218460083,
+      "learning_rate": 4.036815593148119e-05,
+      "loss": 1.1695,
+      "step": 9268
+    },
+    {
+      "epoch": 0.5631569354152743,
+      "grad_norm": 0.12399948388338089,
+      "learning_rate": 4.035876263200577e-05,
+      "loss": 1.0449,
+      "step": 9269
+    },
+    {
+      "epoch": 0.5632176924479009,
+      "grad_norm": 0.12388220429420471,
+      "learning_rate": 4.034936968593177e-05,
+      "loss": 1.0941,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5632784494805274,
+      "grad_norm": 0.16216684877872467,
+      "learning_rate": 4.033997709360351e-05,
+      "loss": 1.0491,
+      "step": 9271
+    },
+    {
+      "epoch": 0.563339206513154,
+      "grad_norm": 0.5428144931793213,
+      "learning_rate": 4.033058485536524e-05,
+      "loss": 1.1478,
+      "step": 9272
+    },
+    {
+      "epoch": 0.5633999635457804,
+      "grad_norm": 0.1771230399608612,
+      "learning_rate": 4.0321192971561245e-05,
+      "loss": 1.1018,
+      "step": 9273
+    },
+    {
+      "epoch": 0.5634607205784069,
+      "grad_norm": 9.403618812561035,
+      "learning_rate": 4.031180144253579e-05,
+      "loss": 1.0903,
+      "step": 9274
+    },
+    {
+      "epoch": 0.5635214776110334,
+      "grad_norm": 0.2488657385110855,
+      "learning_rate": 4.0302410268633134e-05,
+      "loss": 1.0739,
+      "step": 9275
+    },
+    {
+      "epoch": 0.56358223464366,
+      "grad_norm": 0.14782527089118958,
+      "learning_rate": 4.02930194501975e-05,
+      "loss": 1.1432,
+      "step": 9276
+    },
+    {
+      "epoch": 0.5636429916762865,
+      "grad_norm": 0.14933501183986664,
+      "learning_rate": 4.028362898757313e-05,
+      "loss": 1.0251,
+      "step": 9277
+    },
+    {
+      "epoch": 0.563703748708913,
+      "grad_norm": 0.13341885805130005,
+      "learning_rate": 4.027423888110421e-05,
+      "loss": 1.0581,
+      "step": 9278
+    },
+    {
+      "epoch": 0.5637645057415396,
+      "grad_norm": 0.3196311891078949,
+      "learning_rate": 4.026484913113493e-05,
+      "loss": 1.0843,
+      "step": 9279
+    },
+    {
+      "epoch": 0.5638252627741661,
+      "grad_norm": 1.8453270196914673,
+      "learning_rate": 4.0255459738009496e-05,
+      "loss": 1.1265,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5638860198067926,
+      "grad_norm": 0.15659084916114807,
+      "learning_rate": 4.0246070702072066e-05,
+      "loss": 1.0409,
+      "step": 9281
+    },
+    {
+      "epoch": 0.5639467768394192,
+      "grad_norm": 0.125941663980484,
+      "learning_rate": 4.0236682023666796e-05,
+      "loss": 1.0557,
+      "step": 9282
+    },
+    {
+      "epoch": 0.5640075338720457,
+      "grad_norm": 0.1870090216398239,
+      "learning_rate": 4.022729370313784e-05,
+      "loss": 1.162,
+      "step": 9283
+    },
+    {
+      "epoch": 0.5640682909046723,
+      "grad_norm": 0.17505106329917908,
+      "learning_rate": 4.0217905740829334e-05,
+      "loss": 0.9839,
+      "step": 9284
+    },
+    {
+      "epoch": 0.5641290479372988,
+      "grad_norm": 1.1048821210861206,
+      "learning_rate": 4.020851813708537e-05,
+      "loss": 1.026,
+      "step": 9285
+    },
+    {
+      "epoch": 0.5641898049699253,
+      "grad_norm": 0.15395544469356537,
+      "learning_rate": 4.0199130892250075e-05,
+      "loss": 1.1568,
+      "step": 9286
+    },
+    {
+      "epoch": 0.5642505620025517,
+      "grad_norm": 0.12581133842468262,
+      "learning_rate": 4.018974400666751e-05,
+      "loss": 1.0514,
+      "step": 9287
+    },
+    {
+      "epoch": 0.5643113190351783,
+      "grad_norm": 0.2663700580596924,
+      "learning_rate": 4.0180357480681794e-05,
+      "loss": 1.1256,
+      "step": 9288
+    },
+    {
+      "epoch": 0.5643720760678048,
+      "grad_norm": 0.19487228989601135,
+      "learning_rate": 4.0170971314636974e-05,
+      "loss": 1.2091,
+      "step": 9289
+    },
+    {
+      "epoch": 0.5644328331004314,
+      "grad_norm": 0.17007489502429962,
+      "learning_rate": 4.0161585508877084e-05,
+      "loss": 1.1693,
+      "step": 9290
+    },
+    {
+      "epoch": 0.5644935901330579,
+      "grad_norm": 1.9008747339248657,
+      "learning_rate": 4.015220006374621e-05,
+      "loss": 1.0776,
+      "step": 9291
+    },
+    {
+      "epoch": 0.5645543471656844,
+      "grad_norm": 0.13507528603076935,
+      "learning_rate": 4.0142814979588314e-05,
+      "loss": 1.0141,
+      "step": 9292
+    },
+    {
+      "epoch": 0.564615104198311,
+      "grad_norm": 0.18734753131866455,
+      "learning_rate": 4.0133430256747456e-05,
+      "loss": 1.1283,
+      "step": 9293
+    },
+    {
+      "epoch": 0.5646758612309375,
+      "grad_norm": 0.1442510187625885,
+      "learning_rate": 4.012404589556762e-05,
+      "loss": 1.0857,
+      "step": 9294
+    },
+    {
+      "epoch": 0.564736618263564,
+      "grad_norm": 0.1793176680803299,
+      "learning_rate": 4.011466189639277e-05,
+      "loss": 1.0161,
+      "step": 9295
+    },
+    {
+      "epoch": 0.5647973752961906,
+      "grad_norm": 0.17933893203735352,
+      "learning_rate": 4.010527825956692e-05,
+      "loss": 1.1322,
+      "step": 9296
+    },
+    {
+      "epoch": 0.5648581323288171,
+      "grad_norm": 0.20198547840118408,
+      "learning_rate": 4.009589498543401e-05,
+      "loss": 1.0565,
+      "step": 9297
+    },
+    {
+      "epoch": 0.5649188893614436,
+      "grad_norm": 0.28964295983314514,
+      "learning_rate": 4.008651207433798e-05,
+      "loss": 1.1322,
+      "step": 9298
+    },
+    {
+      "epoch": 0.5649796463940702,
+      "grad_norm": 0.3297141194343567,
+      "learning_rate": 4.007712952662277e-05,
+      "loss": 1.1392,
+      "step": 9299
+    },
+    {
+      "epoch": 0.5650404034266966,
+      "grad_norm": 0.18128494918346405,
+      "learning_rate": 4.006774734263227e-05,
+      "loss": 1.1262,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5651011604593231,
+      "grad_norm": 0.25018155574798584,
+      "learning_rate": 4.0058365522710435e-05,
+      "loss": 1.0916,
+      "step": 9301
+    },
+    {
+      "epoch": 0.5651619174919497,
+      "grad_norm": 0.2777166962623596,
+      "learning_rate": 4.0048984067201126e-05,
+      "loss": 1.3092,
+      "step": 9302
+    },
+    {
+      "epoch": 0.5652226745245762,
+      "grad_norm": 0.16668292880058289,
+      "learning_rate": 4.003960297644822e-05,
+      "loss": 1.0368,
+      "step": 9303
+    },
+    {
+      "epoch": 0.5652834315572027,
+      "grad_norm": 0.214456707239151,
+      "learning_rate": 4.003022225079563e-05,
+      "loss": 1.1412,
+      "step": 9304
+    },
+    {
+      "epoch": 0.5653441885898293,
+      "grad_norm": 0.14797136187553406,
+      "learning_rate": 4.002084189058714e-05,
+      "loss": 1.0916,
+      "step": 9305
+    },
+    {
+      "epoch": 0.5654049456224558,
+      "grad_norm": 0.12118365615606308,
+      "learning_rate": 4.001146189616662e-05,
+      "loss": 1.0001,
+      "step": 9306
+    },
+    {
+      "epoch": 0.5654657026550823,
+      "grad_norm": 0.13804593682289124,
+      "learning_rate": 4.00020822678779e-05,
+      "loss": 1.0719,
+      "step": 9307
+    },
+    {
+      "epoch": 0.5655264596877089,
+      "grad_norm": 0.1874879151582718,
+      "learning_rate": 3.999270300606477e-05,
+      "loss": 1.1724,
+      "step": 9308
+    },
+    {
+      "epoch": 0.5655872167203354,
+      "grad_norm": 0.11761140078306198,
+      "learning_rate": 3.9983324111071064e-05,
+      "loss": 1.0454,
+      "step": 9309
+    },
+    {
+      "epoch": 0.5656479737529619,
+      "grad_norm": 0.8463904857635498,
+      "learning_rate": 3.9973945583240545e-05,
+      "loss": 1.0446,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5657087307855885,
+      "grad_norm": 0.30176177620887756,
+      "learning_rate": 3.9964567422917e-05,
+      "loss": 1.1424,
+      "step": 9311
+    },
+    {
+      "epoch": 0.565769487818215,
+      "grad_norm": 0.18117481470108032,
+      "learning_rate": 3.995518963044417e-05,
+      "loss": 1.1793,
+      "step": 9312
+    },
+    {
+      "epoch": 0.5658302448508414,
+      "grad_norm": 0.20154625177383423,
+      "learning_rate": 3.994581220616579e-05,
+      "loss": 1.0541,
+      "step": 9313
+    },
+    {
+      "epoch": 0.565891001883468,
+      "grad_norm": 0.15985903143882751,
+      "learning_rate": 3.993643515042562e-05,
+      "loss": 1.0832,
+      "step": 9314
+    },
+    {
+      "epoch": 0.5659517589160945,
+      "grad_norm": 0.4219246506690979,
+      "learning_rate": 3.992705846356737e-05,
+      "loss": 1.0448,
+      "step": 9315
+    },
+    {
+      "epoch": 0.566012515948721,
+      "grad_norm": 0.13238303363323212,
+      "learning_rate": 3.991768214593472e-05,
+      "loss": 1.0338,
+      "step": 9316
+    },
+    {
+      "epoch": 0.5660732729813476,
+      "grad_norm": 0.15834259986877441,
+      "learning_rate": 3.990830619787139e-05,
+      "loss": 1.0897,
+      "step": 9317
+    },
+    {
+      "epoch": 0.5661340300139741,
+      "grad_norm": 7.561474800109863,
+      "learning_rate": 3.9898930619721065e-05,
+      "loss": 1.0701,
+      "step": 9318
+    },
+    {
+      "epoch": 0.5661947870466006,
+      "grad_norm": 0.3598170876502991,
+      "learning_rate": 3.9889555411827374e-05,
+      "loss": 1.0967,
+      "step": 9319
+    },
+    {
+      "epoch": 0.5662555440792272,
+      "grad_norm": 0.23852452635765076,
+      "learning_rate": 3.988018057453398e-05,
+      "loss": 1.0428,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5663163011118537,
+      "grad_norm": 0.13089148700237274,
+      "learning_rate": 3.987080610818452e-05,
+      "loss": 1.0212,
+      "step": 9321
+    },
+    {
+      "epoch": 0.5663770581444802,
+      "grad_norm": 0.15455998480319977,
+      "learning_rate": 3.986143201312263e-05,
+      "loss": 1.0467,
+      "step": 9322
+    },
+    {
+      "epoch": 0.5664378151771068,
+      "grad_norm": 0.1878483146429062,
+      "learning_rate": 3.9852058289691905e-05,
+      "loss": 1.0567,
+      "step": 9323
+    },
+    {
+      "epoch": 0.5664985722097333,
+      "grad_norm": 0.21425405144691467,
+      "learning_rate": 3.9842684938235966e-05,
+      "loss": 1.0102,
+      "step": 9324
+    },
+    {
+      "epoch": 0.5665593292423599,
+      "grad_norm": 0.16678059101104736,
+      "learning_rate": 3.9833311959098354e-05,
+      "loss": 1.0462,
+      "step": 9325
+    },
+    {
+      "epoch": 0.5666200862749863,
+      "grad_norm": 0.45588287711143494,
+      "learning_rate": 3.982393935262264e-05,
+      "loss": 1.2509,
+      "step": 9326
+    },
+    {
+      "epoch": 0.5666808433076128,
+      "grad_norm": 0.14263463020324707,
+      "learning_rate": 3.9814567119152416e-05,
+      "loss": 1.0678,
+      "step": 9327
+    },
+    {
+      "epoch": 0.5667416003402393,
+      "grad_norm": 0.13871221244335175,
+      "learning_rate": 3.98051952590312e-05,
+      "loss": 1.1122,
+      "step": 9328
+    },
+    {
+      "epoch": 0.5668023573728659,
+      "grad_norm": 0.41435712575912476,
+      "learning_rate": 3.979582377260252e-05,
+      "loss": 1.0951,
+      "step": 9329
+    },
+    {
+      "epoch": 0.5668631144054924,
+      "grad_norm": 0.15884922444820404,
+      "learning_rate": 3.9786452660209895e-05,
+      "loss": 1.1005,
+      "step": 9330
+    },
+    {
+      "epoch": 0.566923871438119,
+      "grad_norm": 0.3118976950645447,
+      "learning_rate": 3.977708192219683e-05,
+      "loss": 1.1567,
+      "step": 9331
+    },
+    {
+      "epoch": 0.5669846284707455,
+      "grad_norm": 0.13443267345428467,
+      "learning_rate": 3.97677115589068e-05,
+      "loss": 1.0394,
+      "step": 9332
+    },
+    {
+      "epoch": 0.567045385503372,
+      "grad_norm": 0.13821937143802643,
+      "learning_rate": 3.9758341570683265e-05,
+      "loss": 1.0507,
+      "step": 9333
+    },
+    {
+      "epoch": 0.5671061425359986,
+      "grad_norm": 0.16394339501857758,
+      "learning_rate": 3.974897195786971e-05,
+      "loss": 1.0217,
+      "step": 9334
+    },
+    {
+      "epoch": 0.5671668995686251,
+      "grad_norm": 0.20001637935638428,
+      "learning_rate": 3.9739602720809575e-05,
+      "loss": 1.1347,
+      "step": 9335
+    },
+    {
+      "epoch": 0.5672276566012516,
+      "grad_norm": 0.15204745531082153,
+      "learning_rate": 3.9730233859846284e-05,
+      "loss": 1.0489,
+      "step": 9336
+    },
+    {
+      "epoch": 0.5672884136338782,
+      "grad_norm": 0.27701836824417114,
+      "learning_rate": 3.972086537532324e-05,
+      "loss": 1.0803,
+      "step": 9337
+    },
+    {
+      "epoch": 0.5673491706665047,
+      "grad_norm": 0.28854477405548096,
+      "learning_rate": 3.971149726758389e-05,
+      "loss": 1.1778,
+      "step": 9338
+    },
+    {
+      "epoch": 0.5674099276991311,
+      "grad_norm": 0.15160194039344788,
+      "learning_rate": 3.9702129536971577e-05,
+      "loss": 1.1792,
+      "step": 9339
+    },
+    {
+      "epoch": 0.5674706847317577,
+      "grad_norm": 0.1626320481300354,
+      "learning_rate": 3.96927621838297e-05,
+      "loss": 1.067,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5675314417643842,
+      "grad_norm": 0.21114283800125122,
+      "learning_rate": 3.968339520850161e-05,
+      "loss": 1.1389,
+      "step": 9341
+    },
+    {
+      "epoch": 0.5675921987970107,
+      "grad_norm": 0.13761281967163086,
+      "learning_rate": 3.967402861133068e-05,
+      "loss": 1.0398,
+      "step": 9342
+    },
+    {
+      "epoch": 0.5676529558296373,
+      "grad_norm": 0.1349709928035736,
+      "learning_rate": 3.966466239266021e-05,
+      "loss": 1.1146,
+      "step": 9343
+    },
+    {
+      "epoch": 0.5677137128622638,
+      "grad_norm": 0.14251908659934998,
+      "learning_rate": 3.9655296552833565e-05,
+      "loss": 1.0173,
+      "step": 9344
+    },
+    {
+      "epoch": 0.5677744698948903,
+      "grad_norm": 0.15189701318740845,
+      "learning_rate": 3.9645931092194004e-05,
+      "loss": 1.0556,
+      "step": 9345
+    },
+    {
+      "epoch": 0.5678352269275169,
+      "grad_norm": 0.16033907234668732,
+      "learning_rate": 3.963656601108483e-05,
+      "loss": 1.0844,
+      "step": 9346
+    },
+    {
+      "epoch": 0.5678959839601434,
+      "grad_norm": 0.12061890214681625,
+      "learning_rate": 3.962720130984935e-05,
+      "loss": 1.0828,
+      "step": 9347
+    },
+    {
+      "epoch": 0.5679567409927699,
+      "grad_norm": 0.3561362028121948,
+      "learning_rate": 3.96178369888308e-05,
+      "loss": 1.1843,
+      "step": 9348
+    },
+    {
+      "epoch": 0.5680174980253965,
+      "grad_norm": 0.20875217020511627,
+      "learning_rate": 3.9608473048372444e-05,
+      "loss": 1.111,
+      "step": 9349
+    },
+    {
+      "epoch": 0.568078255058023,
+      "grad_norm": 0.22136706113815308,
+      "learning_rate": 3.959910948881751e-05,
+      "loss": 1.0714,
+      "step": 9350
+    },
+    {
+      "epoch": 0.5681390120906495,
+      "grad_norm": 0.14425520598888397,
+      "learning_rate": 3.958974631050926e-05,
+      "loss": 1.067,
+      "step": 9351
+    },
+    {
+      "epoch": 0.568199769123276,
+      "grad_norm": 0.1793167144060135,
+      "learning_rate": 3.9580383513790845e-05,
+      "loss": 1.0601,
+      "step": 9352
+    },
+    {
+      "epoch": 0.5682605261559025,
+      "grad_norm": 0.1604180485010147,
+      "learning_rate": 3.9571021099005495e-05,
+      "loss": 1.0436,
+      "step": 9353
+    },
+    {
+      "epoch": 0.568321283188529,
+      "grad_norm": 0.12525637447834015,
+      "learning_rate": 3.956165906649637e-05,
+      "loss": 1.1014,
+      "step": 9354
+    },
+    {
+      "epoch": 0.5683820402211556,
+      "grad_norm": 0.3861655294895172,
+      "learning_rate": 3.955229741660665e-05,
+      "loss": 1.1394,
+      "step": 9355
+    },
+    {
+      "epoch": 0.5684427972537821,
+      "grad_norm": 0.3664751946926117,
+      "learning_rate": 3.95429361496795e-05,
+      "loss": 1.2092,
+      "step": 9356
+    },
+    {
+      "epoch": 0.5685035542864086,
+      "grad_norm": 0.1381242573261261,
+      "learning_rate": 3.953357526605803e-05,
+      "loss": 1.0252,
+      "step": 9357
+    },
+    {
+      "epoch": 0.5685643113190352,
+      "grad_norm": 0.11836809664964676,
+      "learning_rate": 3.952421476608542e-05,
+      "loss": 1.0592,
+      "step": 9358
+    },
+    {
+      "epoch": 0.5686250683516617,
+      "grad_norm": 0.15251173079013824,
+      "learning_rate": 3.95148546501047e-05,
+      "loss": 1.1237,
+      "step": 9359
+    },
+    {
+      "epoch": 0.5686858253842882,
+      "grad_norm": 0.27714094519615173,
+      "learning_rate": 3.9505494918459034e-05,
+      "loss": 1.0587,
+      "step": 9360
+    },
+    {
+      "epoch": 0.5687465824169148,
+      "grad_norm": 0.1492842435836792,
+      "learning_rate": 3.9496135571491475e-05,
+      "loss": 1.0524,
+      "step": 9361
+    },
+    {
+      "epoch": 0.5688073394495413,
+      "grad_norm": 0.13584201037883759,
+      "learning_rate": 3.9486776609545086e-05,
+      "loss": 1.1166,
+      "step": 9362
+    },
+    {
+      "epoch": 0.5688680964821679,
+      "grad_norm": 0.20638935267925262,
+      "learning_rate": 3.947741803296294e-05,
+      "loss": 1.1507,
+      "step": 9363
+    },
+    {
+      "epoch": 0.5689288535147944,
+      "grad_norm": 0.16127735376358032,
+      "learning_rate": 3.946805984208808e-05,
+      "loss": 1.049,
+      "step": 9364
+    },
+    {
+      "epoch": 0.5689896105474208,
+      "grad_norm": 0.3948340117931366,
+      "learning_rate": 3.9458702037263515e-05,
+      "loss": 1.0238,
+      "step": 9365
+    },
+    {
+      "epoch": 0.5690503675800473,
+      "grad_norm": 0.13667432963848114,
+      "learning_rate": 3.9449344618832265e-05,
+      "loss": 1.1048,
+      "step": 9366
+    },
+    {
+      "epoch": 0.5691111246126739,
+      "grad_norm": 0.14048372209072113,
+      "learning_rate": 3.943998758713732e-05,
+      "loss": 1.061,
+      "step": 9367
+    },
+    {
+      "epoch": 0.5691718816453004,
+      "grad_norm": 0.2314092516899109,
+      "learning_rate": 3.943063094252167e-05,
+      "loss": 1.2434,
+      "step": 9368
+    },
+    {
+      "epoch": 0.569232638677927,
+      "grad_norm": 0.20214729011058807,
+      "learning_rate": 3.94212746853283e-05,
+      "loss": 1.0621,
+      "step": 9369
+    },
+    {
+      "epoch": 0.5692933957105535,
+      "grad_norm": 0.20059621334075928,
+      "learning_rate": 3.9411918815900126e-05,
+      "loss": 1.0898,
+      "step": 9370
+    },
+    {
+      "epoch": 0.56935415274318,
+      "grad_norm": 0.2255912721157074,
+      "learning_rate": 3.9402563334580155e-05,
+      "loss": 1.2409,
+      "step": 9371
+    },
+    {
+      "epoch": 0.5694149097758066,
+      "grad_norm": 0.1617220938205719,
+      "learning_rate": 3.939320824171124e-05,
+      "loss": 1.1072,
+      "step": 9372
+    },
+    {
+      "epoch": 0.5694756668084331,
+      "grad_norm": 0.19971393048763275,
+      "learning_rate": 3.938385353763634e-05,
+      "loss": 1.0481,
+      "step": 9373
+    },
+    {
+      "epoch": 0.5695364238410596,
+      "grad_norm": 0.16343924403190613,
+      "learning_rate": 3.937449922269834e-05,
+      "loss": 1.0222,
+      "step": 9374
+    },
+    {
+      "epoch": 0.5695971808736862,
+      "grad_norm": 0.18706795573234558,
+      "learning_rate": 3.9365145297240114e-05,
+      "loss": 1.1051,
+      "step": 9375
+    },
+    {
+      "epoch": 0.5696579379063127,
+      "grad_norm": 0.16287298500537872,
+      "learning_rate": 3.935579176160455e-05,
+      "loss": 1.1133,
+      "step": 9376
+    },
+    {
+      "epoch": 0.5697186949389392,
+      "grad_norm": 0.35146385431289673,
+      "learning_rate": 3.93464386161345e-05,
+      "loss": 1.1841,
+      "step": 9377
+    },
+    {
+      "epoch": 0.5697794519715657,
+      "grad_norm": 0.11296513676643372,
+      "learning_rate": 3.933708586117281e-05,
+      "loss": 1.0019,
+      "step": 9378
+    },
+    {
+      "epoch": 0.5698402090041922,
+      "grad_norm": 0.17266689240932465,
+      "learning_rate": 3.9327733497062294e-05,
+      "loss": 1.049,
+      "step": 9379
+    },
+    {
+      "epoch": 0.5699009660368187,
+      "grad_norm": 0.2204691767692566,
+      "learning_rate": 3.931838152414575e-05,
+      "loss": 1.091,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5699617230694453,
+      "grad_norm": 0.19983020424842834,
+      "learning_rate": 3.930902994276601e-05,
+      "loss": 1.082,
+      "step": 9381
+    },
+    {
+      "epoch": 0.5700224801020718,
+      "grad_norm": 0.20253503322601318,
+      "learning_rate": 3.929967875326584e-05,
+      "loss": 1.0606,
+      "step": 9382
+    },
+    {
+      "epoch": 0.5700832371346983,
+      "grad_norm": 0.1941344290971756,
+      "learning_rate": 3.9290327955988014e-05,
+      "loss": 1.1429,
+      "step": 9383
+    },
+    {
+      "epoch": 0.5701439941673249,
+      "grad_norm": 0.3806479573249817,
+      "learning_rate": 3.9280977551275296e-05,
+      "loss": 1.1604,
+      "step": 9384
+    },
+    {
+      "epoch": 0.5702047511999514,
+      "grad_norm": 0.2293899804353714,
+      "learning_rate": 3.9271627539470425e-05,
+      "loss": 1.0977,
+      "step": 9385
+    },
+    {
+      "epoch": 0.5702655082325779,
+      "grad_norm": 0.20892605185508728,
+      "learning_rate": 3.926227792091611e-05,
+      "loss": 1.0988,
+      "step": 9386
+    },
+    {
+      "epoch": 0.5703262652652045,
+      "grad_norm": 0.14298801124095917,
+      "learning_rate": 3.925292869595507e-05,
+      "loss": 1.0489,
+      "step": 9387
+    },
+    {
+      "epoch": 0.570387022297831,
+      "grad_norm": 0.29442140460014343,
+      "learning_rate": 3.9243579864929995e-05,
+      "loss": 1.0525,
+      "step": 9388
+    },
+    {
+      "epoch": 0.5704477793304575,
+      "grad_norm": 0.23393961787223816,
+      "learning_rate": 3.9234231428183594e-05,
+      "loss": 1.0799,
+      "step": 9389
+    },
+    {
+      "epoch": 0.5705085363630841,
+      "grad_norm": 0.16348634660243988,
+      "learning_rate": 3.922488338605852e-05,
+      "loss": 1.0522,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5705692933957106,
+      "grad_norm": 0.5351850986480713,
+      "learning_rate": 3.9215535738897435e-05,
+      "loss": 1.0977,
+      "step": 9391
+    },
+    {
+      "epoch": 0.570630050428337,
+      "grad_norm": 0.18496620655059814,
+      "learning_rate": 3.920618848704297e-05,
+      "loss": 1.1478,
+      "step": 9392
+    },
+    {
+      "epoch": 0.5706908074609636,
+      "grad_norm": 0.20554541051387787,
+      "learning_rate": 3.919684163083773e-05,
+      "loss": 1.1469,
+      "step": 9393
+    },
+    {
+      "epoch": 0.5707515644935901,
+      "grad_norm": 0.43232595920562744,
+      "learning_rate": 3.918749517062437e-05,
+      "loss": 1.2087,
+      "step": 9394
+    },
+    {
+      "epoch": 0.5708123215262166,
+      "grad_norm": 0.2660805583000183,
+      "learning_rate": 3.9178149106745454e-05,
+      "loss": 1.0644,
+      "step": 9395
+    },
+    {
+      "epoch": 0.5708730785588432,
+      "grad_norm": 0.1719880998134613,
+      "learning_rate": 3.9168803439543575e-05,
+      "loss": 1.05,
+      "step": 9396
+    },
+    {
+      "epoch": 0.5709338355914697,
+      "grad_norm": 0.1952056735754013,
+      "learning_rate": 3.91594581693613e-05,
+      "loss": 1.1386,
+      "step": 9397
+    },
+    {
+      "epoch": 0.5709945926240962,
+      "grad_norm": 0.18509529531002045,
+      "learning_rate": 3.91501132965412e-05,
+      "loss": 1.052,
+      "step": 9398
+    },
+    {
+      "epoch": 0.5710553496567228,
+      "grad_norm": 0.20501284301280975,
+      "learning_rate": 3.9140768821425775e-05,
+      "loss": 1.0618,
+      "step": 9399
+    },
+    {
+      "epoch": 0.5711161066893493,
+      "grad_norm": 0.11503874510526657,
+      "learning_rate": 3.9131424744357574e-05,
+      "loss": 1.0255,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5711768637219758,
+      "grad_norm": 0.23324188590049744,
+      "learning_rate": 3.912208106567909e-05,
+      "loss": 1.1138,
+      "step": 9401
+    },
+    {
+      "epoch": 0.5712376207546024,
+      "grad_norm": 0.12903693318367004,
+      "learning_rate": 3.911273778573284e-05,
+      "loss": 1.0259,
+      "step": 9402
+    },
+    {
+      "epoch": 0.5712983777872289,
+      "grad_norm": 0.23123428225517273,
+      "learning_rate": 3.910339490486128e-05,
+      "loss": 1.1218,
+      "step": 9403
+    },
+    {
+      "epoch": 0.5713591348198555,
+      "grad_norm": 0.18961231410503387,
+      "learning_rate": 3.9094052423406886e-05,
+      "loss": 1.0625,
+      "step": 9404
+    },
+    {
+      "epoch": 0.5714198918524819,
+      "grad_norm": 0.13013260066509247,
+      "learning_rate": 3.908471034171214e-05,
+      "loss": 1.0603,
+      "step": 9405
+    },
+    {
+      "epoch": 0.5714806488851084,
+      "grad_norm": 0.13224995136260986,
+      "learning_rate": 3.907536866011941e-05,
+      "loss": 1.1031,
+      "step": 9406
+    },
+    {
+      "epoch": 0.571541405917735,
+      "grad_norm": 0.1757730394601822,
+      "learning_rate": 3.906602737897116e-05,
+      "loss": 1.0995,
+      "step": 9407
+    },
+    {
+      "epoch": 0.5716021629503615,
+      "grad_norm": 1.349341630935669,
+      "learning_rate": 3.9056686498609803e-05,
+      "loss": 1.2023,
+      "step": 9408
+    },
+    {
+      "epoch": 0.571662919982988,
+      "grad_norm": 0.13470570743083954,
+      "learning_rate": 3.9047346019377695e-05,
+      "loss": 1.0785,
+      "step": 9409
+    },
+    {
+      "epoch": 0.5717236770156146,
+      "grad_norm": 0.1447269320487976,
+      "learning_rate": 3.903800594161725e-05,
+      "loss": 1.1064,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5717844340482411,
+      "grad_norm": 0.22651201486587524,
+      "learning_rate": 3.9028666265670835e-05,
+      "loss": 1.0325,
+      "step": 9411
+    },
+    {
+      "epoch": 0.5718451910808676,
+      "grad_norm": 0.14072097837924957,
+      "learning_rate": 3.901932699188076e-05,
+      "loss": 1.0215,
+      "step": 9412
+    },
+    {
+      "epoch": 0.5719059481134942,
+      "grad_norm": 0.15580034255981445,
+      "learning_rate": 3.900998812058937e-05,
+      "loss": 1.0087,
+      "step": 9413
+    },
+    {
+      "epoch": 0.5719667051461207,
+      "grad_norm": 0.13120998442173004,
+      "learning_rate": 3.900064965213899e-05,
+      "loss": 1.0517,
+      "step": 9414
+    },
+    {
+      "epoch": 0.5720274621787472,
+      "grad_norm": 0.16143788397312164,
+      "learning_rate": 3.8991311586871925e-05,
+      "loss": 1.1034,
+      "step": 9415
+    },
+    {
+      "epoch": 0.5720882192113738,
+      "grad_norm": 0.1284245252609253,
+      "learning_rate": 3.898197392513047e-05,
+      "loss": 1.0483,
+      "step": 9416
+    },
+    {
+      "epoch": 0.5721489762440003,
+      "grad_norm": 0.1260843425989151,
+      "learning_rate": 3.897263666725688e-05,
+      "loss": 1.0085,
+      "step": 9417
+    },
+    {
+      "epoch": 0.5722097332766267,
+      "grad_norm": 0.14079874753952026,
+      "learning_rate": 3.896329981359345e-05,
+      "loss": 1.0296,
+      "step": 9418
+    },
+    {
+      "epoch": 0.5722704903092533,
+      "grad_norm": 0.15053793787956238,
+      "learning_rate": 3.895396336448237e-05,
+      "loss": 1.0672,
+      "step": 9419
+    },
+    {
+      "epoch": 0.5723312473418798,
+      "grad_norm": 0.6577506065368652,
+      "learning_rate": 3.894462732026591e-05,
+      "loss": 1.2259,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5723920043745063,
+      "grad_norm": 0.1847575306892395,
+      "learning_rate": 3.893529168128627e-05,
+      "loss": 1.1215,
+      "step": 9421
+    },
+    {
+      "epoch": 0.5724527614071329,
+      "grad_norm": 0.2351631075143814,
+      "learning_rate": 3.8925956447885645e-05,
+      "loss": 1.1323,
+      "step": 9422
+    },
+    {
+      "epoch": 0.5725135184397594,
+      "grad_norm": 0.19972263276576996,
+      "learning_rate": 3.8916621620406236e-05,
+      "loss": 1.0736,
+      "step": 9423
+    },
+    {
+      "epoch": 0.5725742754723859,
+      "grad_norm": 0.13498350977897644,
+      "learning_rate": 3.89072871991902e-05,
+      "loss": 1.0143,
+      "step": 9424
+    },
+    {
+      "epoch": 0.5726350325050125,
+      "grad_norm": 0.30430951714515686,
+      "learning_rate": 3.889795318457971e-05,
+      "loss": 1.2341,
+      "step": 9425
+    },
+    {
+      "epoch": 0.572695789537639,
+      "grad_norm": 0.14694227278232574,
+      "learning_rate": 3.888861957691688e-05,
+      "loss": 1.0595,
+      "step": 9426
+    },
+    {
+      "epoch": 0.5727565465702655,
+      "grad_norm": 0.257284551858902,
+      "learning_rate": 3.8879286376543835e-05,
+      "loss": 1.173,
+      "step": 9427
+    },
+    {
+      "epoch": 0.5728173036028921,
+      "grad_norm": 0.4094320237636566,
+      "learning_rate": 3.886995358380272e-05,
+      "loss": 1.0723,
+      "step": 9428
+    },
+    {
+      "epoch": 0.5728780606355186,
+      "grad_norm": 0.15119077265262604,
+      "learning_rate": 3.886062119903559e-05,
+      "loss": 1.0735,
+      "step": 9429
+    },
+    {
+      "epoch": 0.5729388176681451,
+      "grad_norm": 0.15245527029037476,
+      "learning_rate": 3.885128922258454e-05,
+      "loss": 1.0173,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5729995747007716,
+      "grad_norm": 0.17560645937919617,
+      "learning_rate": 3.884195765479166e-05,
+      "loss": 1.0854,
+      "step": 9431
+    },
+    {
+      "epoch": 0.5730603317333981,
+      "grad_norm": 1.8061186075210571,
+      "learning_rate": 3.8832626495998976e-05,
+      "loss": 1.044,
+      "step": 9432
+    },
+    {
+      "epoch": 0.5731210887660246,
+      "grad_norm": 0.15292610228061676,
+      "learning_rate": 3.882329574654852e-05,
+      "loss": 1.0392,
+      "step": 9433
+    },
+    {
+      "epoch": 0.5731818457986512,
+      "grad_norm": 0.4316026568412781,
+      "learning_rate": 3.88139654067823e-05,
+      "loss": 1.1387,
+      "step": 9434
+    },
+    {
+      "epoch": 0.5732426028312777,
+      "grad_norm": 0.17155036330223083,
+      "learning_rate": 3.880463547704236e-05,
+      "loss": 1.1373,
+      "step": 9435
+    },
+    {
+      "epoch": 0.5733033598639042,
+      "grad_norm": 0.14839480817317963,
+      "learning_rate": 3.879530595767067e-05,
+      "loss": 1.0884,
+      "step": 9436
+    },
+    {
+      "epoch": 0.5733641168965308,
+      "grad_norm": 0.13140985369682312,
+      "learning_rate": 3.878597684900921e-05,
+      "loss": 1.0844,
+      "step": 9437
+    },
+    {
+      "epoch": 0.5734248739291573,
+      "grad_norm": 0.472240686416626,
+      "learning_rate": 3.877664815139994e-05,
+      "loss": 1.0934,
+      "step": 9438
+    },
+    {
+      "epoch": 0.5734856309617838,
+      "grad_norm": 0.1632692962884903,
+      "learning_rate": 3.8767319865184785e-05,
+      "loss": 1.0508,
+      "step": 9439
+    },
+    {
+      "epoch": 0.5735463879944104,
+      "grad_norm": 0.24705441296100616,
+      "learning_rate": 3.875799199070569e-05,
+      "loss": 1.0675,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5736071450270369,
+      "grad_norm": 0.22303584218025208,
+      "learning_rate": 3.8748664528304585e-05,
+      "loss": 1.1092,
+      "step": 9441
+    },
+    {
+      "epoch": 0.5736679020596634,
+      "grad_norm": 0.9875159859657288,
+      "learning_rate": 3.8739337478323346e-05,
+      "loss": 1.374,
+      "step": 9442
+    },
+    {
+      "epoch": 0.57372865909229,
+      "grad_norm": 0.2708924114704132,
+      "learning_rate": 3.873001084110387e-05,
+      "loss": 1.0546,
+      "step": 9443
+    },
+    {
+      "epoch": 0.5737894161249164,
+      "grad_norm": 0.16349385678768158,
+      "learning_rate": 3.8720684616988035e-05,
+      "loss": 1.0983,
+      "step": 9444
+    },
+    {
+      "epoch": 0.5738501731575429,
+      "grad_norm": 0.17874056100845337,
+      "learning_rate": 3.871135880631769e-05,
+      "loss": 1.1012,
+      "step": 9445
+    },
+    {
+      "epoch": 0.5739109301901695,
+      "grad_norm": 0.14569275081157684,
+      "learning_rate": 3.870203340943466e-05,
+      "loss": 1.0513,
+      "step": 9446
+    },
+    {
+      "epoch": 0.573971687222796,
+      "grad_norm": 0.19844374060630798,
+      "learning_rate": 3.869270842668077e-05,
+      "loss": 1.0311,
+      "step": 9447
+    },
+    {
+      "epoch": 0.5740324442554225,
+      "grad_norm": 0.13414768874645233,
+      "learning_rate": 3.868338385839785e-05,
+      "loss": 0.9853,
+      "step": 9448
+    },
+    {
+      "epoch": 0.5740932012880491,
+      "grad_norm": 0.17769423127174377,
+      "learning_rate": 3.867405970492767e-05,
+      "loss": 1.0348,
+      "step": 9449
+    },
+    {
+      "epoch": 0.5741539583206756,
+      "grad_norm": 0.20692425966262817,
+      "learning_rate": 3.866473596661202e-05,
+      "loss": 1.067,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5742147153533022,
+      "grad_norm": 0.1440236121416092,
+      "learning_rate": 3.865541264379268e-05,
+      "loss": 1.0152,
+      "step": 9451
+    },
+    {
+      "epoch": 0.5742754723859287,
+      "grad_norm": 0.16100431978702545,
+      "learning_rate": 3.864608973681139e-05,
+      "loss": 1.0976,
+      "step": 9452
+    },
+    {
+      "epoch": 0.5743362294185552,
+      "grad_norm": 0.12844322621822357,
+      "learning_rate": 3.863676724600986e-05,
+      "loss": 1.0736,
+      "step": 9453
+    },
+    {
+      "epoch": 0.5743969864511818,
+      "grad_norm": 0.19452571868896484,
+      "learning_rate": 3.862744517172982e-05,
+      "loss": 1.1387,
+      "step": 9454
+    },
+    {
+      "epoch": 0.5744577434838083,
+      "grad_norm": 0.15475571155548096,
+      "learning_rate": 3.8618123514312976e-05,
+      "loss": 1.0415,
+      "step": 9455
+    },
+    {
+      "epoch": 0.5745185005164348,
+      "grad_norm": 0.1372871994972229,
+      "learning_rate": 3.860880227410102e-05,
+      "loss": 1.0488,
+      "step": 9456
+    },
+    {
+      "epoch": 0.5745792575490613,
+      "grad_norm": 0.15570376813411713,
+      "learning_rate": 3.859948145143562e-05,
+      "loss": 1.1063,
+      "step": 9457
+    },
+    {
+      "epoch": 0.5746400145816878,
+      "grad_norm": 0.22972489893436432,
+      "learning_rate": 3.859016104665844e-05,
+      "loss": 1.0534,
+      "step": 9458
+    },
+    {
+      "epoch": 0.5747007716143143,
+      "grad_norm": 0.12764061987400055,
+      "learning_rate": 3.85808410601111e-05,
+      "loss": 1.0705,
+      "step": 9459
+    },
+    {
+      "epoch": 0.5747615286469409,
+      "grad_norm": 0.25844812393188477,
+      "learning_rate": 3.8571521492135235e-05,
+      "loss": 1.0179,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5748222856795674,
+      "grad_norm": 0.26326504349708557,
+      "learning_rate": 3.856220234307246e-05,
+      "loss": 0.9958,
+      "step": 9461
+    },
+    {
+      "epoch": 0.5748830427121939,
+      "grad_norm": 0.18367226421833038,
+      "learning_rate": 3.855288361326437e-05,
+      "loss": 1.0969,
+      "step": 9462
+    },
+    {
+      "epoch": 0.5749437997448205,
+      "grad_norm": 0.17877331376075745,
+      "learning_rate": 3.8543565303052545e-05,
+      "loss": 1.0697,
+      "step": 9463
+    },
+    {
+      "epoch": 0.575004556777447,
+      "grad_norm": 0.2133195698261261,
+      "learning_rate": 3.8534247412778545e-05,
+      "loss": 1.0795,
+      "step": 9464
+    },
+    {
+      "epoch": 0.5750653138100735,
+      "grad_norm": 0.1494211107492447,
+      "learning_rate": 3.852492994278394e-05,
+      "loss": 1.1312,
+      "step": 9465
+    },
+    {
+      "epoch": 0.5751260708427001,
+      "grad_norm": 0.18412959575653076,
+      "learning_rate": 3.851561289341023e-05,
+      "loss": 1.0814,
+      "step": 9466
+    },
+    {
+      "epoch": 0.5751868278753266,
+      "grad_norm": 0.16899055242538452,
+      "learning_rate": 3.850629626499895e-05,
+      "loss": 1.1157,
+      "step": 9467
+    },
+    {
+      "epoch": 0.5752475849079531,
+      "grad_norm": 0.11114361137151718,
+      "learning_rate": 3.84969800578916e-05,
+      "loss": 1.0274,
+      "step": 9468
+    },
+    {
+      "epoch": 0.5753083419405797,
+      "grad_norm": 0.15761998295783997,
+      "learning_rate": 3.848766427242967e-05,
+      "loss": 1.0926,
+      "step": 9469
+    },
+    {
+      "epoch": 0.5753690989732061,
+      "grad_norm": 1.3900434970855713,
+      "learning_rate": 3.847834890895464e-05,
+      "loss": 1.1706,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5754298560058326,
+      "grad_norm": 0.17364750802516937,
+      "learning_rate": 3.8469033967807944e-05,
+      "loss": 1.1114,
+      "step": 9471
+    },
+    {
+      "epoch": 0.5754906130384592,
+      "grad_norm": 0.17179855704307556,
+      "learning_rate": 3.845971944933107e-05,
+      "loss": 1.1155,
+      "step": 9472
+    },
+    {
+      "epoch": 0.5755513700710857,
+      "grad_norm": 0.24840620160102844,
+      "learning_rate": 3.845040535386537e-05,
+      "loss": 1.1415,
+      "step": 9473
+    },
+    {
+      "epoch": 0.5756121271037122,
+      "grad_norm": 0.13832220435142517,
+      "learning_rate": 3.844109168175231e-05,
+      "loss": 1.0976,
+      "step": 9474
+    },
+    {
+      "epoch": 0.5756728841363388,
+      "grad_norm": 0.14361411333084106,
+      "learning_rate": 3.8431778433333274e-05,
+      "loss": 1.0617,
+      "step": 9475
+    },
+    {
+      "epoch": 0.5757336411689653,
+      "grad_norm": 0.39725133776664734,
+      "learning_rate": 3.842246560894962e-05,
+      "loss": 1.3927,
+      "step": 9476
+    },
+    {
+      "epoch": 0.5757943982015918,
+      "grad_norm": 1.257039189338684,
+      "learning_rate": 3.841315320894274e-05,
+      "loss": 1.1913,
+      "step": 9477
+    },
+    {
+      "epoch": 0.5758551552342184,
+      "grad_norm": 0.24458445608615875,
+      "learning_rate": 3.840384123365397e-05,
+      "loss": 1.0635,
+      "step": 9478
+    },
+    {
+      "epoch": 0.5759159122668449,
+      "grad_norm": 0.1508876085281372,
+      "learning_rate": 3.839452968342464e-05,
+      "loss": 1.0651,
+      "step": 9479
+    },
+    {
+      "epoch": 0.5759766692994714,
+      "grad_norm": 0.20164048671722412,
+      "learning_rate": 3.838521855859607e-05,
+      "loss": 1.1319,
+      "step": 9480
+    },
+    {
+      "epoch": 0.576037426332098,
+      "grad_norm": 0.2970738112926483,
+      "learning_rate": 3.837590785950954e-05,
+      "loss": 1.2593,
+      "step": 9481
+    },
+    {
+      "epoch": 0.5760981833647245,
+      "grad_norm": 0.1766209453344345,
+      "learning_rate": 3.836659758650637e-05,
+      "loss": 1.0936,
+      "step": 9482
+    },
+    {
+      "epoch": 0.5761589403973509,
+      "grad_norm": 0.18700683116912842,
+      "learning_rate": 3.835728773992781e-05,
+      "loss": 1.158,
+      "step": 9483
+    },
+    {
+      "epoch": 0.5762196974299775,
+      "grad_norm": 0.2654655873775482,
+      "learning_rate": 3.8347978320115116e-05,
+      "loss": 1.0342,
+      "step": 9484
+    },
+    {
+      "epoch": 0.576280454462604,
+      "grad_norm": 0.20301446318626404,
+      "learning_rate": 3.8338669327409555e-05,
+      "loss": 1.0669,
+      "step": 9485
+    },
+    {
+      "epoch": 0.5763412114952305,
+      "grad_norm": 0.12816840410232544,
+      "learning_rate": 3.832936076215229e-05,
+      "loss": 1.0647,
+      "step": 9486
+    },
+    {
+      "epoch": 0.5764019685278571,
+      "grad_norm": 0.19805611670017242,
+      "learning_rate": 3.832005262468458e-05,
+      "loss": 1.0501,
+      "step": 9487
+    },
+    {
+      "epoch": 0.5764627255604836,
+      "grad_norm": 0.23758670687675476,
+      "learning_rate": 3.831074491534759e-05,
+      "loss": 1.0757,
+      "step": 9488
+    },
+    {
+      "epoch": 0.5765234825931101,
+      "grad_norm": 0.12797591090202332,
+      "learning_rate": 3.83014376344825e-05,
+      "loss": 1.0124,
+      "step": 9489
+    },
+    {
+      "epoch": 0.5765842396257367,
+      "grad_norm": 0.19223390519618988,
+      "learning_rate": 3.829213078243048e-05,
+      "loss": 1.1857,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5766449966583632,
+      "grad_norm": 0.16237464547157288,
+      "learning_rate": 3.828282435953267e-05,
+      "loss": 1.0556,
+      "step": 9491
+    },
+    {
+      "epoch": 0.5767057536909898,
+      "grad_norm": 0.1579771488904953,
+      "learning_rate": 3.827351836613021e-05,
+      "loss": 1.0399,
+      "step": 9492
+    },
+    {
+      "epoch": 0.5767665107236163,
+      "grad_norm": 2.5476765632629395,
+      "learning_rate": 3.82642128025642e-05,
+      "loss": 1.0881,
+      "step": 9493
+    },
+    {
+      "epoch": 0.5768272677562428,
+      "grad_norm": 0.33077988028526306,
+      "learning_rate": 3.825490766917572e-05,
+      "loss": 1.0575,
+      "step": 9494
+    },
+    {
+      "epoch": 0.5768880247888694,
+      "grad_norm": 0.17774169147014618,
+      "learning_rate": 3.8245602966305876e-05,
+      "loss": 1.0979,
+      "step": 9495
+    },
+    {
+      "epoch": 0.5769487818214959,
+      "grad_norm": 0.14809660613536835,
+      "learning_rate": 3.823629869429573e-05,
+      "loss": 1.0374,
+      "step": 9496
+    },
+    {
+      "epoch": 0.5770095388541223,
+      "grad_norm": 0.1711939573287964,
+      "learning_rate": 3.822699485348632e-05,
+      "loss": 1.1024,
+      "step": 9497
+    },
+    {
+      "epoch": 0.5770702958867489,
+      "grad_norm": 0.3284071683883667,
+      "learning_rate": 3.8217691444218725e-05,
+      "loss": 1.0745,
+      "step": 9498
+    },
+    {
+      "epoch": 0.5771310529193754,
+      "grad_norm": 0.1721469610929489,
+      "learning_rate": 3.8208388466833886e-05,
+      "loss": 1.114,
+      "step": 9499
+    },
+    {
+      "epoch": 0.5771918099520019,
+      "grad_norm": 0.14575400948524475,
+      "learning_rate": 3.8199085921672864e-05,
+      "loss": 1.036,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5772525669846285,
+      "grad_norm": 0.14909052848815918,
+      "learning_rate": 3.818978380907663e-05,
+      "loss": 1.0338,
+      "step": 9501
+    },
+    {
+      "epoch": 0.577313324017255,
+      "grad_norm": 0.4301423132419586,
+      "learning_rate": 3.818048212938614e-05,
+      "loss": 1.1253,
+      "step": 9502
+    },
+    {
+      "epoch": 0.5773740810498815,
+      "grad_norm": 0.6998014450073242,
+      "learning_rate": 3.817118088294237e-05,
+      "loss": 1.0611,
+      "step": 9503
+    },
+    {
+      "epoch": 0.5774348380825081,
+      "grad_norm": 0.16323049366474152,
+      "learning_rate": 3.8161880070086254e-05,
+      "loss": 1.0187,
+      "step": 9504
+    },
+    {
+      "epoch": 0.5774955951151346,
+      "grad_norm": 0.18943989276885986,
+      "learning_rate": 3.8152579691158726e-05,
+      "loss": 1.1636,
+      "step": 9505
+    },
+    {
+      "epoch": 0.5775563521477611,
+      "grad_norm": 0.14730489253997803,
+      "learning_rate": 3.814327974650067e-05,
+      "loss": 1.0257,
+      "step": 9506
+    },
+    {
+      "epoch": 0.5776171091803877,
+      "grad_norm": 0.17674776911735535,
+      "learning_rate": 3.813398023645297e-05,
+      "loss": 1.039,
+      "step": 9507
+    },
+    {
+      "epoch": 0.5776778662130142,
+      "grad_norm": 0.1440037190914154,
+      "learning_rate": 3.812468116135653e-05,
+      "loss": 1.0839,
+      "step": 9508
+    },
+    {
+      "epoch": 0.5777386232456407,
+      "grad_norm": 0.15408442914485931,
+      "learning_rate": 3.81153825215522e-05,
+      "loss": 1.0754,
+      "step": 9509
+    },
+    {
+      "epoch": 0.5777993802782672,
+      "grad_norm": 0.19904424250125885,
+      "learning_rate": 3.8106084317380806e-05,
+      "loss": 1.1259,
+      "step": 9510
+    },
+    {
+      "epoch": 0.5778601373108937,
+      "grad_norm": 0.17540007829666138,
+      "learning_rate": 3.80967865491832e-05,
+      "loss": 1.2089,
+      "step": 9511
+    },
+    {
+      "epoch": 0.5779208943435202,
+      "grad_norm": 0.13899482786655426,
+      "learning_rate": 3.8087489217300205e-05,
+      "loss": 1.0441,
+      "step": 9512
+    },
+    {
+      "epoch": 0.5779816513761468,
+      "grad_norm": 0.14518341422080994,
+      "learning_rate": 3.807819232207258e-05,
+      "loss": 1.0838,
+      "step": 9513
+    },
+    {
+      "epoch": 0.5780424084087733,
+      "grad_norm": 0.13078469038009644,
+      "learning_rate": 3.8068895863841114e-05,
+      "loss": 1.052,
+      "step": 9514
+    },
+    {
+      "epoch": 0.5781031654413998,
+      "grad_norm": 0.13291147351264954,
+      "learning_rate": 3.805959984294657e-05,
+      "loss": 1.0534,
+      "step": 9515
+    },
+    {
+      "epoch": 0.5781639224740264,
+      "grad_norm": 0.15113776922225952,
+      "learning_rate": 3.8050304259729705e-05,
+      "loss": 1.0779,
+      "step": 9516
+    },
+    {
+      "epoch": 0.5782246795066529,
+      "grad_norm": 0.1307903677225113,
+      "learning_rate": 3.804100911453126e-05,
+      "loss": 1.0643,
+      "step": 9517
+    },
+    {
+      "epoch": 0.5782854365392794,
+      "grad_norm": 0.29275310039520264,
+      "learning_rate": 3.803171440769194e-05,
+      "loss": 1.0843,
+      "step": 9518
+    },
+    {
+      "epoch": 0.578346193571906,
+      "grad_norm": 0.2584379315376282,
+      "learning_rate": 3.802242013955243e-05,
+      "loss": 1.0756,
+      "step": 9519
+    },
+    {
+      "epoch": 0.5784069506045325,
+      "grad_norm": 0.14966893196105957,
+      "learning_rate": 3.8013126310453426e-05,
+      "loss": 1.0374,
+      "step": 9520
+    },
+    {
+      "epoch": 0.578467707637159,
+      "grad_norm": 0.1569487303495407,
+      "learning_rate": 3.800383292073559e-05,
+      "loss": 1.1168,
+      "step": 9521
+    },
+    {
+      "epoch": 0.5785284646697856,
+      "grad_norm": 6.4563188552856445,
+      "learning_rate": 3.799453997073959e-05,
+      "loss": 1.1016,
+      "step": 9522
+    },
+    {
+      "epoch": 0.578589221702412,
+      "grad_norm": 0.25884515047073364,
+      "learning_rate": 3.798524746080603e-05,
+      "loss": 1.073,
+      "step": 9523
+    },
+    {
+      "epoch": 0.5786499787350385,
+      "grad_norm": 0.18514546751976013,
+      "learning_rate": 3.797595539127556e-05,
+      "loss": 1.1444,
+      "step": 9524
+    },
+    {
+      "epoch": 0.5787107357676651,
+      "grad_norm": 0.18107715249061584,
+      "learning_rate": 3.7966663762488775e-05,
+      "loss": 1.0531,
+      "step": 9525
+    },
+    {
+      "epoch": 0.5787714928002916,
+      "grad_norm": 0.19950293004512787,
+      "learning_rate": 3.795737257478625e-05,
+      "loss": 1.0847,
+      "step": 9526
+    },
+    {
+      "epoch": 0.5788322498329181,
+      "grad_norm": 0.16721925139427185,
+      "learning_rate": 3.794808182850856e-05,
+      "loss": 1.0912,
+      "step": 9527
+    },
+    {
+      "epoch": 0.5788930068655447,
+      "grad_norm": 0.11909819394350052,
+      "learning_rate": 3.793879152399625e-05,
+      "loss": 1.0477,
+      "step": 9528
+    },
+    {
+      "epoch": 0.5789537638981712,
+      "grad_norm": 0.2933345139026642,
+      "learning_rate": 3.7929501661589875e-05,
+      "loss": 1.0891,
+      "step": 9529
+    },
+    {
+      "epoch": 0.5790145209307977,
+      "grad_norm": 0.1371058076620102,
+      "learning_rate": 3.7920212241629954e-05,
+      "loss": 1.108,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5790752779634243,
+      "grad_norm": 0.14041735231876373,
+      "learning_rate": 3.7910923264456974e-05,
+      "loss": 1.0633,
+      "step": 9531
+    },
+    {
+      "epoch": 0.5791360349960508,
+      "grad_norm": 0.10737396031618118,
+      "learning_rate": 3.790163473041146e-05,
+      "loss": 1.0046,
+      "step": 9532
+    },
+    {
+      "epoch": 0.5791967920286774,
+      "grad_norm": 0.18709294497966766,
+      "learning_rate": 3.789234663983386e-05,
+      "loss": 1.0852,
+      "step": 9533
+    },
+    {
+      "epoch": 0.5792575490613039,
+      "grad_norm": 0.45671042799949646,
+      "learning_rate": 3.788305899306463e-05,
+      "loss": 1.1648,
+      "step": 9534
+    },
+    {
+      "epoch": 0.5793183060939304,
+      "grad_norm": 0.1386062651872635,
+      "learning_rate": 3.7873771790444205e-05,
+      "loss": 1.049,
+      "step": 9535
+    },
+    {
+      "epoch": 0.5793790631265568,
+      "grad_norm": 0.16956038773059845,
+      "learning_rate": 3.786448503231302e-05,
+      "loss": 1.0065,
+      "step": 9536
+    },
+    {
+      "epoch": 0.5794398201591834,
+      "grad_norm": 0.15613463521003723,
+      "learning_rate": 3.785519871901149e-05,
+      "loss": 1.1221,
+      "step": 9537
+    },
+    {
+      "epoch": 0.5795005771918099,
+      "grad_norm": 0.12265809625387192,
+      "learning_rate": 3.784591285088e-05,
+      "loss": 1.0162,
+      "step": 9538
+    },
+    {
+      "epoch": 0.5795613342244365,
+      "grad_norm": 0.18768829107284546,
+      "learning_rate": 3.783662742825893e-05,
+      "loss": 1.1512,
+      "step": 9539
+    },
+    {
+      "epoch": 0.579622091257063,
+      "grad_norm": 0.13276711106300354,
+      "learning_rate": 3.7827342451488616e-05,
+      "loss": 1.0593,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5796828482896895,
+      "grad_norm": 0.37872061133384705,
+      "learning_rate": 3.7818057920909425e-05,
+      "loss": 1.0451,
+      "step": 9541
+    },
+    {
+      "epoch": 0.5797436053223161,
+      "grad_norm": 0.33937597274780273,
+      "learning_rate": 3.7808773836861676e-05,
+      "loss": 1.0868,
+      "step": 9542
+    },
+    {
+      "epoch": 0.5798043623549426,
+      "grad_norm": 0.16435493528842926,
+      "learning_rate": 3.7799490199685675e-05,
+      "loss": 1.0944,
+      "step": 9543
+    },
+    {
+      "epoch": 0.5798651193875691,
+      "grad_norm": 0.23212830722332,
+      "learning_rate": 3.7790207009721724e-05,
+      "loss": 1.1096,
+      "step": 9544
+    },
+    {
+      "epoch": 0.5799258764201957,
+      "grad_norm": 0.1821647435426712,
+      "learning_rate": 3.77809242673101e-05,
+      "loss": 1.1037,
+      "step": 9545
+    },
+    {
+      "epoch": 0.5799866334528222,
+      "grad_norm": 0.3730841875076294,
+      "learning_rate": 3.7771641972791056e-05,
+      "loss": 1.008,
+      "step": 9546
+    },
+    {
+      "epoch": 0.5800473904854487,
+      "grad_norm": 0.1395702064037323,
+      "learning_rate": 3.776236012650484e-05,
+      "loss": 1.0522,
+      "step": 9547
+    },
+    {
+      "epoch": 0.5801081475180753,
+      "grad_norm": 0.13790613412857056,
+      "learning_rate": 3.7753078728791663e-05,
+      "loss": 1.0399,
+      "step": 9548
+    },
+    {
+      "epoch": 0.5801689045507017,
+      "grad_norm": 0.22079253196716309,
+      "learning_rate": 3.774379777999176e-05,
+      "loss": 1.1603,
+      "step": 9549
+    },
+    {
+      "epoch": 0.5802296615833282,
+      "grad_norm": 0.18304385244846344,
+      "learning_rate": 3.773451728044533e-05,
+      "loss": 1.0375,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5802904186159548,
+      "grad_norm": 0.14945372939109802,
+      "learning_rate": 3.772523723049252e-05,
+      "loss": 1.0865,
+      "step": 9551
+    },
+    {
+      "epoch": 0.5803511756485813,
+      "grad_norm": 0.1659063994884491,
+      "learning_rate": 3.771595763047355e-05,
+      "loss": 1.1383,
+      "step": 9552
+    },
+    {
+      "epoch": 0.5804119326812078,
+      "grad_norm": 0.2993048429489136,
+      "learning_rate": 3.770667848072848e-05,
+      "loss": 1.1015,
+      "step": 9553
+    },
+    {
+      "epoch": 0.5804726897138344,
+      "grad_norm": 0.21550501883029938,
+      "learning_rate": 3.769739978159752e-05,
+      "loss": 1.1564,
+      "step": 9554
+    },
+    {
+      "epoch": 0.5805334467464609,
+      "grad_norm": 0.14772237837314606,
+      "learning_rate": 3.768812153342074e-05,
+      "loss": 1.1567,
+      "step": 9555
+    },
+    {
+      "epoch": 0.5805942037790874,
+      "grad_norm": 0.14872927963733673,
+      "learning_rate": 3.7678843736538224e-05,
+      "loss": 1.0764,
+      "step": 9556
+    },
+    {
+      "epoch": 0.580654960811714,
+      "grad_norm": 0.23177458345890045,
+      "learning_rate": 3.76695663912901e-05,
+      "loss": 1.2135,
+      "step": 9557
+    },
+    {
+      "epoch": 0.5807157178443405,
+      "grad_norm": 0.21273943781852722,
+      "learning_rate": 3.7660289498016395e-05,
+      "loss": 1.0559,
+      "step": 9558
+    },
+    {
+      "epoch": 0.580776474876967,
+      "grad_norm": 0.18735463917255402,
+      "learning_rate": 3.7651013057057174e-05,
+      "loss": 1.1357,
+      "step": 9559
+    },
+    {
+      "epoch": 0.5808372319095936,
+      "grad_norm": 0.33305954933166504,
+      "learning_rate": 3.764173706875245e-05,
+      "loss": 1.2752,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5808979889422201,
+      "grad_norm": 0.19825394451618195,
+      "learning_rate": 3.7632461533442223e-05,
+      "loss": 1.1545,
+      "step": 9561
+    },
+    {
+      "epoch": 0.5809587459748465,
+      "grad_norm": 0.2466171681880951,
+      "learning_rate": 3.7623186451466526e-05,
+      "loss": 1.1579,
+      "step": 9562
+    },
+    {
+      "epoch": 0.5810195030074731,
+      "grad_norm": 0.2018802911043167,
+      "learning_rate": 3.761391182316532e-05,
+      "loss": 1.0114,
+      "step": 9563
+    },
+    {
+      "epoch": 0.5810802600400996,
+      "grad_norm": 0.15394359827041626,
+      "learning_rate": 3.760463764887856e-05,
+      "loss": 1.0898,
+      "step": 9564
+    },
+    {
+      "epoch": 0.5811410170727261,
+      "grad_norm": 0.16687077283859253,
+      "learning_rate": 3.759536392894624e-05,
+      "loss": 1.0372,
+      "step": 9565
+    },
+    {
+      "epoch": 0.5812017741053527,
+      "grad_norm": 0.15514753758907318,
+      "learning_rate": 3.758609066370821e-05,
+      "loss": 1.1511,
+      "step": 9566
+    },
+    {
+      "epoch": 0.5812625311379792,
+      "grad_norm": 0.16094811260700226,
+      "learning_rate": 3.757681785350445e-05,
+      "loss": 1.1445,
+      "step": 9567
+    },
+    {
+      "epoch": 0.5813232881706057,
+      "grad_norm": 0.1270761787891388,
+      "learning_rate": 3.7567545498674824e-05,
+      "loss": 1.0394,
+      "step": 9568
+    },
+    {
+      "epoch": 0.5813840452032323,
+      "grad_norm": 0.32100990414619446,
+      "learning_rate": 3.7558273599559216e-05,
+      "loss": 1.1557,
+      "step": 9569
+    },
+    {
+      "epoch": 0.5814448022358588,
+      "grad_norm": 0.16068701446056366,
+      "learning_rate": 3.75490021564975e-05,
+      "loss": 1.0526,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5815055592684854,
+      "grad_norm": 0.21692292392253876,
+      "learning_rate": 3.753973116982952e-05,
+      "loss": 1.0261,
+      "step": 9571
+    },
+    {
+      "epoch": 0.5815663163011119,
+      "grad_norm": 0.2852819859981537,
+      "learning_rate": 3.7530460639895106e-05,
+      "loss": 1.1158,
+      "step": 9572
+    },
+    {
+      "epoch": 0.5816270733337384,
+      "grad_norm": 0.1359856277704239,
+      "learning_rate": 3.7521190567034065e-05,
+      "loss": 0.9819,
+      "step": 9573
+    },
+    {
+      "epoch": 0.581687830366365,
+      "grad_norm": 0.3075811564922333,
+      "learning_rate": 3.7511920951586185e-05,
+      "loss": 1.2604,
+      "step": 9574
+    },
+    {
+      "epoch": 0.5817485873989914,
+      "grad_norm": 0.30809861421585083,
+      "learning_rate": 3.750265179389126e-05,
+      "loss": 1.2348,
+      "step": 9575
+    },
+    {
+      "epoch": 0.5818093444316179,
+      "grad_norm": 2.467280626296997,
+      "learning_rate": 3.7493383094289054e-05,
+      "loss": 1.0171,
+      "step": 9576
+    },
+    {
+      "epoch": 0.5818701014642444,
+      "grad_norm": 0.19968968629837036,
+      "learning_rate": 3.74841148531193e-05,
+      "loss": 1.0411,
+      "step": 9577
+    },
+    {
+      "epoch": 0.581930858496871,
+      "grad_norm": 0.16926982998847961,
+      "learning_rate": 3.747484707072174e-05,
+      "loss": 1.0171,
+      "step": 9578
+    },
+    {
+      "epoch": 0.5819916155294975,
+      "grad_norm": 0.19006602466106415,
+      "learning_rate": 3.7465579747436094e-05,
+      "loss": 1.2035,
+      "step": 9579
+    },
+    {
+      "epoch": 0.582052372562124,
+      "grad_norm": 0.13153447210788727,
+      "learning_rate": 3.745631288360204e-05,
+      "loss": 1.0737,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5821131295947506,
+      "grad_norm": 0.13281063735485077,
+      "learning_rate": 3.744704647955926e-05,
+      "loss": 1.0744,
+      "step": 9581
+    },
+    {
+      "epoch": 0.5821738866273771,
+      "grad_norm": 0.18007522821426392,
+      "learning_rate": 3.74377805356474e-05,
+      "loss": 1.057,
+      "step": 9582
+    },
+    {
+      "epoch": 0.5822346436600037,
+      "grad_norm": 0.2089676558971405,
+      "learning_rate": 3.7428515052206134e-05,
+      "loss": 1.1189,
+      "step": 9583
+    },
+    {
+      "epoch": 0.5822954006926302,
+      "grad_norm": 0.11739115417003632,
+      "learning_rate": 3.741925002957509e-05,
+      "loss": 1.0181,
+      "step": 9584
+    },
+    {
+      "epoch": 0.5823561577252567,
+      "grad_norm": 0.1435522735118866,
+      "learning_rate": 3.740998546809387e-05,
+      "loss": 1.0409,
+      "step": 9585
+    },
+    {
+      "epoch": 0.5824169147578833,
+      "grad_norm": 0.16813106834888458,
+      "learning_rate": 3.7400721368102054e-05,
+      "loss": 1.0666,
+      "step": 9586
+    },
+    {
+      "epoch": 0.5824776717905098,
+      "grad_norm": 0.30288517475128174,
+      "learning_rate": 3.7391457729939226e-05,
+      "loss": 1.0876,
+      "step": 9587
+    },
+    {
+      "epoch": 0.5825384288231362,
+      "grad_norm": 0.13969860970973969,
+      "learning_rate": 3.738219455394496e-05,
+      "loss": 1.0342,
+      "step": 9588
+    },
+    {
+      "epoch": 0.5825991858557628,
+      "grad_norm": 0.18582023680210114,
+      "learning_rate": 3.737293184045879e-05,
+      "loss": 1.1074,
+      "step": 9589
+    },
+    {
+      "epoch": 0.5826599428883893,
+      "grad_norm": 0.18492519855499268,
+      "learning_rate": 3.736366958982024e-05,
+      "loss": 1.1016,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5827206999210158,
+      "grad_norm": 0.1521170288324356,
+      "learning_rate": 3.735440780236883e-05,
+      "loss": 1.1054,
+      "step": 9591
+    },
+    {
+      "epoch": 0.5827814569536424,
+      "grad_norm": 0.12238254398107529,
+      "learning_rate": 3.734514647844406e-05,
+      "loss": 1.0188,
+      "step": 9592
+    },
+    {
+      "epoch": 0.5828422139862689,
+      "grad_norm": 0.1778760552406311,
+      "learning_rate": 3.733588561838538e-05,
+      "loss": 1.1207,
+      "step": 9593
+    },
+    {
+      "epoch": 0.5829029710188954,
+      "grad_norm": 0.33001652359962463,
+      "learning_rate": 3.7326625222532256e-05,
+      "loss": 1.242,
+      "step": 9594
+    },
+    {
+      "epoch": 0.582963728051522,
+      "grad_norm": 0.22192518413066864,
+      "learning_rate": 3.7317365291224124e-05,
+      "loss": 1.0981,
+      "step": 9595
+    },
+    {
+      "epoch": 0.5830244850841485,
+      "grad_norm": 0.18330062925815582,
+      "learning_rate": 3.7308105824800436e-05,
+      "loss": 1.0298,
+      "step": 9596
+    },
+    {
+      "epoch": 0.583085242116775,
+      "grad_norm": 0.12442557513713837,
+      "learning_rate": 3.729884682360058e-05,
+      "loss": 1.0617,
+      "step": 9597
+    },
+    {
+      "epoch": 0.5831459991494016,
+      "grad_norm": 0.14353442192077637,
+      "learning_rate": 3.728958828796394e-05,
+      "loss": 1.0985,
+      "step": 9598
+    },
+    {
+      "epoch": 0.5832067561820281,
+      "grad_norm": 0.16315993666648865,
+      "learning_rate": 3.728033021822992e-05,
+      "loss": 1.1067,
+      "step": 9599
+    },
+    {
+      "epoch": 0.5832675132146546,
+      "grad_norm": 0.32369089126586914,
+      "learning_rate": 3.7271072614737834e-05,
+      "loss": 1.1238,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5833282702472812,
+      "grad_norm": 0.14030922949314117,
+      "learning_rate": 3.7261815477827044e-05,
+      "loss": 1.0612,
+      "step": 9601
+    },
+    {
+      "epoch": 0.5833890272799076,
+      "grad_norm": 0.1400366872549057,
+      "learning_rate": 3.725255880783688e-05,
+      "loss": 1.0251,
+      "step": 9602
+    },
+    {
+      "epoch": 0.5834497843125341,
+      "grad_norm": 0.221337229013443,
+      "learning_rate": 3.724330260510662e-05,
+      "loss": 1.1251,
+      "step": 9603
+    },
+    {
+      "epoch": 0.5835105413451607,
+      "grad_norm": 0.22027553617954254,
+      "learning_rate": 3.7234046869975575e-05,
+      "loss": 1.0546,
+      "step": 9604
+    },
+    {
+      "epoch": 0.5835712983777872,
+      "grad_norm": 0.15704768896102905,
+      "learning_rate": 3.722479160278301e-05,
+      "loss": 1.1272,
+      "step": 9605
+    },
+    {
+      "epoch": 0.5836320554104137,
+      "grad_norm": 0.3534877896308899,
+      "learning_rate": 3.7215536803868204e-05,
+      "loss": 1.034,
+      "step": 9606
+    },
+    {
+      "epoch": 0.5836928124430403,
+      "grad_norm": 2.277421236038208,
+      "learning_rate": 3.720628247357034e-05,
+      "loss": 1.0541,
+      "step": 9607
+    },
+    {
+      "epoch": 0.5837535694756668,
+      "grad_norm": 0.17706996202468872,
+      "learning_rate": 3.719702861222866e-05,
+      "loss": 1.0062,
+      "step": 9608
+    },
+    {
+      "epoch": 0.5838143265082933,
+      "grad_norm": 0.12530963122844696,
+      "learning_rate": 3.718777522018238e-05,
+      "loss": 1.0443,
+      "step": 9609
+    },
+    {
+      "epoch": 0.5838750835409199,
+      "grad_norm": 0.2009391486644745,
+      "learning_rate": 3.717852229777068e-05,
+      "loss": 1.074,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5839358405735464,
+      "grad_norm": 0.23084455728530884,
+      "learning_rate": 3.7169269845332714e-05,
+      "loss": 1.0247,
+      "step": 9611
+    },
+    {
+      "epoch": 0.583996597606173,
+      "grad_norm": 0.1966819018125534,
+      "learning_rate": 3.716001786320766e-05,
+      "loss": 1.0508,
+      "step": 9612
+    },
+    {
+      "epoch": 0.5840573546387995,
+      "grad_norm": 0.10911506414413452,
+      "learning_rate": 3.7150766351734616e-05,
+      "loss": 1.0649,
+      "step": 9613
+    },
+    {
+      "epoch": 0.584118111671426,
+      "grad_norm": 0.24463769793510437,
+      "learning_rate": 3.7141515311252724e-05,
+      "loss": 1.08,
+      "step": 9614
+    },
+    {
+      "epoch": 0.5841788687040524,
+      "grad_norm": 0.6713007092475891,
+      "learning_rate": 3.713226474210107e-05,
+      "loss": 1.0709,
+      "step": 9615
+    },
+    {
+      "epoch": 0.584239625736679,
+      "grad_norm": 0.17672321200370789,
+      "learning_rate": 3.7123014644618726e-05,
+      "loss": 1.0346,
+      "step": 9616
+    },
+    {
+      "epoch": 0.5843003827693055,
+      "grad_norm": 0.19808872044086456,
+      "learning_rate": 3.711376501914479e-05,
+      "loss": 1.1261,
+      "step": 9617
+    },
+    {
+      "epoch": 0.584361139801932,
+      "grad_norm": 0.2292851060628891,
+      "learning_rate": 3.7104515866018274e-05,
+      "loss": 1.1067,
+      "step": 9618
+    },
+    {
+      "epoch": 0.5844218968345586,
+      "grad_norm": 0.1645694226026535,
+      "learning_rate": 3.7095267185578244e-05,
+      "loss": 1.1376,
+      "step": 9619
+    },
+    {
+      "epoch": 0.5844826538671851,
+      "grad_norm": 0.13933876156806946,
+      "learning_rate": 3.708601897816367e-05,
+      "loss": 1.0809,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5845434108998117,
+      "grad_norm": 0.17309273779392242,
+      "learning_rate": 3.7076771244113565e-05,
+      "loss": 1.1881,
+      "step": 9621
+    },
+    {
+      "epoch": 0.5846041679324382,
+      "grad_norm": 0.15303075313568115,
+      "learning_rate": 3.7067523983766915e-05,
+      "loss": 1.0673,
+      "step": 9622
+    },
+    {
+      "epoch": 0.5846649249650647,
+      "grad_norm": 0.2508925795555115,
+      "learning_rate": 3.705827719746267e-05,
+      "loss": 1.0556,
+      "step": 9623
+    },
+    {
+      "epoch": 0.5847256819976913,
+      "grad_norm": 0.382703572511673,
+      "learning_rate": 3.704903088553977e-05,
+      "loss": 1.1719,
+      "step": 9624
+    },
+    {
+      "epoch": 0.5847864390303178,
+      "grad_norm": 0.300736665725708,
+      "learning_rate": 3.703978504833716e-05,
+      "loss": 1.2541,
+      "step": 9625
+    },
+    {
+      "epoch": 0.5848471960629443,
+      "grad_norm": 0.15248076617717743,
+      "learning_rate": 3.7030539686193734e-05,
+      "loss": 1.0604,
+      "step": 9626
+    },
+    {
+      "epoch": 0.5849079530955709,
+      "grad_norm": 0.998816192150116,
+      "learning_rate": 3.702129479944838e-05,
+      "loss": 1.0413,
+      "step": 9627
+    },
+    {
+      "epoch": 0.5849687101281973,
+      "grad_norm": 0.18758432567119598,
+      "learning_rate": 3.7012050388439966e-05,
+      "loss": 1.3046,
+      "step": 9628
+    },
+    {
+      "epoch": 0.5850294671608238,
+      "grad_norm": 0.22952404618263245,
+      "learning_rate": 3.7002806453507356e-05,
+      "loss": 1.0774,
+      "step": 9629
+    },
+    {
+      "epoch": 0.5850902241934504,
+      "grad_norm": 0.1458398550748825,
+      "learning_rate": 3.69935629949894e-05,
+      "loss": 1.0456,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5851509812260769,
+      "grad_norm": 0.11816703528165817,
+      "learning_rate": 3.6984320013224894e-05,
+      "loss": 1.0271,
+      "step": 9631
+    },
+    {
+      "epoch": 0.5852117382587034,
+      "grad_norm": 1.8252034187316895,
+      "learning_rate": 3.697507750855267e-05,
+      "loss": 1.0597,
+      "step": 9632
+    },
+    {
+      "epoch": 0.58527249529133,
+      "grad_norm": 0.21289880573749542,
+      "learning_rate": 3.696583548131147e-05,
+      "loss": 1.2213,
+      "step": 9633
+    },
+    {
+      "epoch": 0.5853332523239565,
+      "grad_norm": 0.11876389384269714,
+      "learning_rate": 3.6956593931840114e-05,
+      "loss": 1.0309,
+      "step": 9634
+    },
+    {
+      "epoch": 0.585394009356583,
+      "grad_norm": 0.17165471613407135,
+      "learning_rate": 3.694735286047731e-05,
+      "loss": 1.0515,
+      "step": 9635
+    },
+    {
+      "epoch": 0.5854547663892096,
+      "grad_norm": 0.2886955440044403,
+      "learning_rate": 3.693811226756182e-05,
+      "loss": 1.0805,
+      "step": 9636
+    },
+    {
+      "epoch": 0.5855155234218361,
+      "grad_norm": 0.1725347489118576,
+      "learning_rate": 3.692887215343235e-05,
+      "loss": 1.1446,
+      "step": 9637
+    },
+    {
+      "epoch": 0.5855762804544626,
+      "grad_norm": 0.16695545613765717,
+      "learning_rate": 3.691963251842759e-05,
+      "loss": 1.0156,
+      "step": 9638
+    },
+    {
+      "epoch": 0.5856370374870892,
+      "grad_norm": 0.11789147555828094,
+      "learning_rate": 3.6910393362886254e-05,
+      "loss": 1.0422,
+      "step": 9639
+    },
+    {
+      "epoch": 0.5856977945197157,
+      "grad_norm": 0.10687675327062607,
+      "learning_rate": 3.690115468714697e-05,
+      "loss": 1.0517,
+      "step": 9640
+    },
+    {
+      "epoch": 0.5857585515523421,
+      "grad_norm": 0.20679590106010437,
+      "learning_rate": 3.6891916491548366e-05,
+      "loss": 1.1821,
+      "step": 9641
+    },
+    {
+      "epoch": 0.5858193085849687,
+      "grad_norm": 0.35516971349716187,
+      "learning_rate": 3.688267877642912e-05,
+      "loss": 1.2186,
+      "step": 9642
+    },
+    {
+      "epoch": 0.5858800656175952,
+      "grad_norm": 0.167977437376976,
+      "learning_rate": 3.687344154212781e-05,
+      "loss": 1.1288,
+      "step": 9643
+    },
+    {
+      "epoch": 0.5859408226502217,
+      "grad_norm": 0.22867953777313232,
+      "learning_rate": 3.686420478898304e-05,
+      "loss": 1.0763,
+      "step": 9644
+    },
+    {
+      "epoch": 0.5860015796828483,
+      "grad_norm": 0.16651295125484467,
+      "learning_rate": 3.685496851733338e-05,
+      "loss": 1.0758,
+      "step": 9645
+    },
+    {
+      "epoch": 0.5860623367154748,
+      "grad_norm": 0.13324297964572906,
+      "learning_rate": 3.6845732727517404e-05,
+      "loss": 1.0628,
+      "step": 9646
+    },
+    {
+      "epoch": 0.5861230937481013,
+      "grad_norm": 0.13318656384944916,
+      "learning_rate": 3.6836497419873626e-05,
+      "loss": 1.0798,
+      "step": 9647
+    },
+    {
+      "epoch": 0.5861838507807279,
+      "grad_norm": 0.4447791874408722,
+      "learning_rate": 3.682726259474059e-05,
+      "loss": 1.0934,
+      "step": 9648
+    },
+    {
+      "epoch": 0.5862446078133544,
+      "grad_norm": 0.1209213063120842,
+      "learning_rate": 3.681802825245677e-05,
+      "loss": 1.0489,
+      "step": 9649
+    },
+    {
+      "epoch": 0.586305364845981,
+      "grad_norm": 0.21711570024490356,
+      "learning_rate": 3.680879439336069e-05,
+      "loss": 1.3267,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5863661218786075,
+      "grad_norm": 0.22704561054706573,
+      "learning_rate": 3.679956101779079e-05,
+      "loss": 1.2459,
+      "step": 9651
+    },
+    {
+      "epoch": 0.586426878911234,
+      "grad_norm": 0.17967826128005981,
+      "learning_rate": 3.679032812608555e-05,
+      "loss": 1.0259,
+      "step": 9652
+    },
+    {
+      "epoch": 0.5864876359438606,
+      "grad_norm": 0.13895003497600555,
+      "learning_rate": 3.678109571858337e-05,
+      "loss": 1.0909,
+      "step": 9653
+    },
+    {
+      "epoch": 0.586548392976487,
+      "grad_norm": 0.10685412585735321,
+      "learning_rate": 3.6771863795622674e-05,
+      "loss": 1.0384,
+      "step": 9654
+    },
+    {
+      "epoch": 0.5866091500091135,
+      "grad_norm": 0.12601348757743835,
+      "learning_rate": 3.6762632357541884e-05,
+      "loss": 1.0736,
+      "step": 9655
+    },
+    {
+      "epoch": 0.58666990704174,
+      "grad_norm": 0.16080918908119202,
+      "learning_rate": 3.675340140467936e-05,
+      "loss": 1.0346,
+      "step": 9656
+    },
+    {
+      "epoch": 0.5867306640743666,
+      "grad_norm": 0.14532466232776642,
+      "learning_rate": 3.6744170937373454e-05,
+      "loss": 1.0835,
+      "step": 9657
+    },
+    {
+      "epoch": 0.5867914211069931,
+      "grad_norm": 0.14446935057640076,
+      "learning_rate": 3.673494095596254e-05,
+      "loss": 1.1205,
+      "step": 9658
+    },
+    {
+      "epoch": 0.5868521781396197,
+      "grad_norm": 0.1682351976633072,
+      "learning_rate": 3.672571146078494e-05,
+      "loss": 1.095,
+      "step": 9659
+    },
+    {
+      "epoch": 0.5869129351722462,
+      "grad_norm": 0.29009154438972473,
+      "learning_rate": 3.6716482452178934e-05,
+      "loss": 1.1829,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5869736922048727,
+      "grad_norm": 0.15475969016551971,
+      "learning_rate": 3.6707253930482835e-05,
+      "loss": 1.1304,
+      "step": 9661
+    },
+    {
+      "epoch": 0.5870344492374993,
+      "grad_norm": 0.1583895981311798,
+      "learning_rate": 3.66980258960349e-05,
+      "loss": 1.0853,
+      "step": 9662
+    },
+    {
+      "epoch": 0.5870952062701258,
+      "grad_norm": 0.17046944797039032,
+      "learning_rate": 3.6688798349173416e-05,
+      "loss": 1.0419,
+      "step": 9663
+    },
+    {
+      "epoch": 0.5871559633027523,
+      "grad_norm": 0.140001580119133,
+      "learning_rate": 3.66795712902366e-05,
+      "loss": 1.0849,
+      "step": 9664
+    },
+    {
+      "epoch": 0.5872167203353789,
+      "grad_norm": 0.1780005395412445,
+      "learning_rate": 3.6670344719562655e-05,
+      "loss": 1.2036,
+      "step": 9665
+    },
+    {
+      "epoch": 0.5872774773680054,
+      "grad_norm": 0.25022220611572266,
+      "learning_rate": 3.666111863748984e-05,
+      "loss": 1.0155,
+      "step": 9666
+    },
+    {
+      "epoch": 0.5873382344006318,
+      "grad_norm": 0.17986616492271423,
+      "learning_rate": 3.6651893044356264e-05,
+      "loss": 1.1346,
+      "step": 9667
+    },
+    {
+      "epoch": 0.5873989914332584,
+      "grad_norm": 0.2608937621116638,
+      "learning_rate": 3.664266794050014e-05,
+      "loss": 1.0366,
+      "step": 9668
+    },
+    {
+      "epoch": 0.5874597484658849,
+      "grad_norm": 0.2139994502067566,
+      "learning_rate": 3.66334433262596e-05,
+      "loss": 1.1001,
+      "step": 9669
+    },
+    {
+      "epoch": 0.5875205054985114,
+      "grad_norm": 0.26043063402175903,
+      "learning_rate": 3.6624219201972776e-05,
+      "loss": 1.1152,
+      "step": 9670
+    },
+    {
+      "epoch": 0.587581262531138,
+      "grad_norm": 0.48465073108673096,
+      "learning_rate": 3.661499556797779e-05,
+      "loss": 1.1041,
+      "step": 9671
+    },
+    {
+      "epoch": 0.5876420195637645,
+      "grad_norm": 0.16609665751457214,
+      "learning_rate": 3.660577242461273e-05,
+      "loss": 1.0641,
+      "step": 9672
+    },
+    {
+      "epoch": 0.587702776596391,
+      "grad_norm": 0.1469862461090088,
+      "learning_rate": 3.6596549772215674e-05,
+      "loss": 1.0916,
+      "step": 9673
+    },
+    {
+      "epoch": 0.5877635336290176,
+      "grad_norm": 0.254740446805954,
+      "learning_rate": 3.658732761112468e-05,
+      "loss": 1.1355,
+      "step": 9674
+    },
+    {
+      "epoch": 0.5878242906616441,
+      "grad_norm": 0.2856753468513489,
+      "learning_rate": 3.657810594167777e-05,
+      "loss": 1.0557,
+      "step": 9675
+    },
+    {
+      "epoch": 0.5878850476942706,
+      "grad_norm": 0.13355521857738495,
+      "learning_rate": 3.656888476421298e-05,
+      "loss": 1.0546,
+      "step": 9676
+    },
+    {
+      "epoch": 0.5879458047268972,
+      "grad_norm": 0.1959685981273651,
+      "learning_rate": 3.655966407906832e-05,
+      "loss": 1.0779,
+      "step": 9677
+    },
+    {
+      "epoch": 0.5880065617595237,
+      "grad_norm": 0.15109848976135254,
+      "learning_rate": 3.655044388658176e-05,
+      "loss": 1.0887,
+      "step": 9678
+    },
+    {
+      "epoch": 0.5880673187921502,
+      "grad_norm": 0.15172675251960754,
+      "learning_rate": 3.6541224187091324e-05,
+      "loss": 1.1211,
+      "step": 9679
+    },
+    {
+      "epoch": 0.5881280758247767,
+      "grad_norm": 0.15554292500019073,
+      "learning_rate": 3.6532004980934876e-05,
+      "loss": 1.0886,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5881888328574032,
+      "grad_norm": 0.16062697768211365,
+      "learning_rate": 3.65227862684504e-05,
+      "loss": 1.0527,
+      "step": 9681
+    },
+    {
+      "epoch": 0.5882495898900297,
+      "grad_norm": 12.580662727355957,
+      "learning_rate": 3.65135680499758e-05,
+      "loss": 1.0685,
+      "step": 9682
+    },
+    {
+      "epoch": 0.5883103469226563,
+      "grad_norm": 0.26436224579811096,
+      "learning_rate": 3.650435032584895e-05,
+      "loss": 1.1265,
+      "step": 9683
+    },
+    {
+      "epoch": 0.5883711039552828,
+      "grad_norm": 0.16329185664653778,
+      "learning_rate": 3.649513309640777e-05,
+      "loss": 1.1335,
+      "step": 9684
+    },
+    {
+      "epoch": 0.5884318609879093,
+      "grad_norm": 3.0906193256378174,
+      "learning_rate": 3.6485916361990094e-05,
+      "loss": 1.0864,
+      "step": 9685
+    },
+    {
+      "epoch": 0.5884926180205359,
+      "grad_norm": 0.23503564298152924,
+      "learning_rate": 3.647670012293377e-05,
+      "loss": 1.2349,
+      "step": 9686
+    },
+    {
+      "epoch": 0.5885533750531624,
+      "grad_norm": 0.18491913378238678,
+      "learning_rate": 3.646748437957661e-05,
+      "loss": 1.0454,
+      "step": 9687
+    },
+    {
+      "epoch": 0.5886141320857889,
+      "grad_norm": 0.14460882544517517,
+      "learning_rate": 3.645826913225642e-05,
+      "loss": 1.0482,
+      "step": 9688
+    },
+    {
+      "epoch": 0.5886748891184155,
+      "grad_norm": 0.16084350645542145,
+      "learning_rate": 3.6449054381311e-05,
+      "loss": 1.1088,
+      "step": 9689
+    },
+    {
+      "epoch": 0.588735646151042,
+      "grad_norm": 0.32970014214515686,
+      "learning_rate": 3.643984012707811e-05,
+      "loss": 1.1329,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5887964031836685,
+      "grad_norm": 0.14513203501701355,
+      "learning_rate": 3.643062636989549e-05,
+      "loss": 1.0926,
+      "step": 9691
+    },
+    {
+      "epoch": 0.5888571602162951,
+      "grad_norm": 0.13109713792800903,
+      "learning_rate": 3.642141311010089e-05,
+      "loss": 1.0657,
+      "step": 9692
+    },
+    {
+      "epoch": 0.5889179172489215,
+      "grad_norm": 0.12235511839389801,
+      "learning_rate": 3.641220034803203e-05,
+      "loss": 1.0773,
+      "step": 9693
+    },
+    {
+      "epoch": 0.588978674281548,
+      "grad_norm": 0.15629084408283234,
+      "learning_rate": 3.640298808402659e-05,
+      "loss": 1.0427,
+      "step": 9694
+    },
+    {
+      "epoch": 0.5890394313141746,
+      "grad_norm": 0.15765614807605743,
+      "learning_rate": 3.639377631842225e-05,
+      "loss": 1.0123,
+      "step": 9695
+    },
+    {
+      "epoch": 0.5891001883468011,
+      "grad_norm": 6.2862348556518555,
+      "learning_rate": 3.6384565051556654e-05,
+      "loss": 1.0683,
+      "step": 9696
+    },
+    {
+      "epoch": 0.5891609453794276,
+      "grad_norm": 0.18151108920574188,
+      "learning_rate": 3.637535428376746e-05,
+      "loss": 1.1443,
+      "step": 9697
+    },
+    {
+      "epoch": 0.5892217024120542,
+      "grad_norm": 0.136151984333992,
+      "learning_rate": 3.63661440153923e-05,
+      "loss": 1.0831,
+      "step": 9698
+    },
+    {
+      "epoch": 0.5892824594446807,
+      "grad_norm": 0.1270492523908615,
+      "learning_rate": 3.635693424676877e-05,
+      "loss": 1.0711,
+      "step": 9699
+    },
+    {
+      "epoch": 0.5893432164773073,
+      "grad_norm": 0.2637779116630554,
+      "learning_rate": 3.6347724978234444e-05,
+      "loss": 1.2085,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5894039735099338,
+      "grad_norm": 0.1644567996263504,
+      "learning_rate": 3.633851621012689e-05,
+      "loss": 1.0329,
+      "step": 9701
+    },
+    {
+      "epoch": 0.5894647305425603,
+      "grad_norm": 0.12515617907047272,
+      "learning_rate": 3.6329307942783676e-05,
+      "loss": 1.0566,
+      "step": 9702
+    },
+    {
+      "epoch": 0.5895254875751869,
+      "grad_norm": 0.19501769542694092,
+      "learning_rate": 3.632010017654232e-05,
+      "loss": 1.1064,
+      "step": 9703
+    },
+    {
+      "epoch": 0.5895862446078134,
+      "grad_norm": 0.17413097620010376,
+      "learning_rate": 3.6310892911740325e-05,
+      "loss": 1.1467,
+      "step": 9704
+    },
+    {
+      "epoch": 0.5896470016404399,
+      "grad_norm": 0.1883806437253952,
+      "learning_rate": 3.63016861487152e-05,
+      "loss": 1.0681,
+      "step": 9705
+    },
+    {
+      "epoch": 0.5897077586730665,
+      "grad_norm": 0.11197453737258911,
+      "learning_rate": 3.629247988780444e-05,
+      "loss": 1.0395,
+      "step": 9706
+    },
+    {
+      "epoch": 0.5897685157056929,
+      "grad_norm": 0.129778653383255,
+      "learning_rate": 3.628327412934546e-05,
+      "loss": 1.1012,
+      "step": 9707
+    },
+    {
+      "epoch": 0.5898292727383194,
+      "grad_norm": 0.13288386166095734,
+      "learning_rate": 3.627406887367572e-05,
+      "loss": 1.0753,
+      "step": 9708
+    },
+    {
+      "epoch": 0.589890029770946,
+      "grad_norm": 0.15064646303653717,
+      "learning_rate": 3.626486412113264e-05,
+      "loss": 1.1315,
+      "step": 9709
+    },
+    {
+      "epoch": 0.5899507868035725,
+      "grad_norm": 0.11687958240509033,
+      "learning_rate": 3.6255659872053616e-05,
+      "loss": 1.0245,
+      "step": 9710
+    },
+    {
+      "epoch": 0.590011543836199,
+      "grad_norm": 0.13441386818885803,
+      "learning_rate": 3.6246456126776044e-05,
+      "loss": 1.2249,
+      "step": 9711
+    },
+    {
+      "epoch": 0.5900723008688256,
+      "grad_norm": 0.45595914125442505,
+      "learning_rate": 3.623725288563726e-05,
+      "loss": 1.1587,
+      "step": 9712
+    },
+    {
+      "epoch": 0.5901330579014521,
+      "grad_norm": 0.17567183077335358,
+      "learning_rate": 3.6228050148974674e-05,
+      "loss": 1.0741,
+      "step": 9713
+    },
+    {
+      "epoch": 0.5901938149340786,
+      "grad_norm": 0.16142162680625916,
+      "learning_rate": 3.6218847917125535e-05,
+      "loss": 1.0423,
+      "step": 9714
+    },
+    {
+      "epoch": 0.5902545719667052,
+      "grad_norm": 0.21234942972660065,
+      "learning_rate": 3.62096461904272e-05,
+      "loss": 1.133,
+      "step": 9715
+    },
+    {
+      "epoch": 0.5903153289993317,
+      "grad_norm": 0.24975788593292236,
+      "learning_rate": 3.620044496921695e-05,
+      "loss": 1.2539,
+      "step": 9716
+    },
+    {
+      "epoch": 0.5903760860319582,
+      "grad_norm": 0.17137493193149567,
+      "learning_rate": 3.619124425383206e-05,
+      "loss": 1.036,
+      "step": 9717
+    },
+    {
+      "epoch": 0.5904368430645848,
+      "grad_norm": 0.19582246243953705,
+      "learning_rate": 3.618204404460978e-05,
+      "loss": 1.1645,
+      "step": 9718
+    },
+    {
+      "epoch": 0.5904976000972113,
+      "grad_norm": 0.2093893587589264,
+      "learning_rate": 3.617284434188735e-05,
+      "loss": 1.1117,
+      "step": 9719
+    },
+    {
+      "epoch": 0.5905583571298377,
+      "grad_norm": 0.12619788944721222,
+      "learning_rate": 3.616364514600198e-05,
+      "loss": 1.0607,
+      "step": 9720
+    },
+    {
+      "epoch": 0.5906191141624643,
+      "grad_norm": 0.21111953258514404,
+      "learning_rate": 3.615444645729087e-05,
+      "loss": 1.1281,
+      "step": 9721
+    },
+    {
+      "epoch": 0.5906798711950908,
+      "grad_norm": 0.3733457922935486,
+      "learning_rate": 3.6145248276091194e-05,
+      "loss": 1.0778,
+      "step": 9722
+    },
+    {
+      "epoch": 0.5907406282277173,
+      "grad_norm": 0.14821258187294006,
+      "learning_rate": 3.613605060274013e-05,
+      "loss": 1.0515,
+      "step": 9723
+    },
+    {
+      "epoch": 0.5908013852603439,
+      "grad_norm": 0.18887345492839813,
+      "learning_rate": 3.612685343757481e-05,
+      "loss": 1.0697,
+      "step": 9724
+    },
+    {
+      "epoch": 0.5908621422929704,
+      "grad_norm": 0.1876954287290573,
+      "learning_rate": 3.611765678093235e-05,
+      "loss": 1.0477,
+      "step": 9725
+    },
+    {
+      "epoch": 0.5909228993255969,
+      "grad_norm": 0.23776161670684814,
+      "learning_rate": 3.610846063314988e-05,
+      "loss": 1.1025,
+      "step": 9726
+    },
+    {
+      "epoch": 0.5909836563582235,
+      "grad_norm": 0.19449111819267273,
+      "learning_rate": 3.609926499456446e-05,
+      "loss": 1.0667,
+      "step": 9727
+    },
+    {
+      "epoch": 0.59104441339085,
+      "grad_norm": 0.21076911687850952,
+      "learning_rate": 3.609006986551318e-05,
+      "loss": 1.03,
+      "step": 9728
+    },
+    {
+      "epoch": 0.5911051704234765,
+      "grad_norm": 0.18143071234226227,
+      "learning_rate": 3.6080875246333065e-05,
+      "loss": 1.0281,
+      "step": 9729
+    },
+    {
+      "epoch": 0.5911659274561031,
+      "grad_norm": 0.1307360976934433,
+      "learning_rate": 3.6071681137361155e-05,
+      "loss": 1.081,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5912266844887296,
+      "grad_norm": 0.22783619165420532,
+      "learning_rate": 3.606248753893448e-05,
+      "loss": 1.0565,
+      "step": 9731
+    },
+    {
+      "epoch": 0.5912874415213561,
+      "grad_norm": 0.1415441781282425,
+      "learning_rate": 3.605329445139001e-05,
+      "loss": 1.0448,
+      "step": 9732
+    },
+    {
+      "epoch": 0.5913481985539826,
+      "grad_norm": 4.329713344573975,
+      "learning_rate": 3.604410187506474e-05,
+      "loss": 1.0548,
+      "step": 9733
+    },
+    {
+      "epoch": 0.5914089555866091,
+      "grad_norm": 0.16169285774230957,
+      "learning_rate": 3.603490981029559e-05,
+      "loss": 1.071,
+      "step": 9734
+    },
+    {
+      "epoch": 0.5914697126192356,
+      "grad_norm": 0.1344682276248932,
+      "learning_rate": 3.602571825741953e-05,
+      "loss": 1.0402,
+      "step": 9735
+    },
+    {
+      "epoch": 0.5915304696518622,
+      "grad_norm": 0.17793437838554382,
+      "learning_rate": 3.6016527216773475e-05,
+      "loss": 1.0594,
+      "step": 9736
+    },
+    {
+      "epoch": 0.5915912266844887,
+      "grad_norm": 0.10395093262195587,
+      "learning_rate": 3.60073366886943e-05,
+      "loss": 1.0068,
+      "step": 9737
+    },
+    {
+      "epoch": 0.5916519837171152,
+      "grad_norm": 0.13307540118694305,
+      "learning_rate": 3.5998146673518915e-05,
+      "loss": 1.0896,
+      "step": 9738
+    },
+    {
+      "epoch": 0.5917127407497418,
+      "grad_norm": 0.2078952044248581,
+      "learning_rate": 3.598895717158418e-05,
+      "loss": 1.09,
+      "step": 9739
+    },
+    {
+      "epoch": 0.5917734977823683,
+      "grad_norm": 0.19054830074310303,
+      "learning_rate": 3.597976818322694e-05,
+      "loss": 1.1146,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5918342548149949,
+      "grad_norm": 0.1147465705871582,
+      "learning_rate": 3.597057970878398e-05,
+      "loss": 1.0438,
+      "step": 9741
+    },
+    {
+      "epoch": 0.5918950118476214,
+      "grad_norm": 0.15593543648719788,
+      "learning_rate": 3.596139174859213e-05,
+      "loss": 1.055,
+      "step": 9742
+    },
+    {
+      "epoch": 0.5919557688802479,
+      "grad_norm": 0.16292649507522583,
+      "learning_rate": 3.5952204302988194e-05,
+      "loss": 1.0582,
+      "step": 9743
+    },
+    {
+      "epoch": 0.5920165259128745,
+      "grad_norm": 0.1694033443927765,
+      "learning_rate": 3.594301737230892e-05,
+      "loss": 1.0674,
+      "step": 9744
+    },
+    {
+      "epoch": 0.592077282945501,
+      "grad_norm": 0.12481171637773514,
+      "learning_rate": 3.5933830956891054e-05,
+      "loss": 1.0597,
+      "step": 9745
+    },
+    {
+      "epoch": 0.5921380399781274,
+      "grad_norm": 0.1326330453157425,
+      "learning_rate": 3.592464505707136e-05,
+      "loss": 1.0534,
+      "step": 9746
+    },
+    {
+      "epoch": 0.592198797010754,
+      "grad_norm": 0.9225171208381653,
+      "learning_rate": 3.591545967318649e-05,
+      "loss": 1.0924,
+      "step": 9747
+    },
+    {
+      "epoch": 0.5922595540433805,
+      "grad_norm": 0.13176541030406952,
+      "learning_rate": 3.590627480557318e-05,
+      "loss": 1.0137,
+      "step": 9748
+    },
+    {
+      "epoch": 0.592320311076007,
+      "grad_norm": 0.12434348464012146,
+      "learning_rate": 3.5897090454568097e-05,
+      "loss": 1.0294,
+      "step": 9749
+    },
+    {
+      "epoch": 0.5923810681086336,
+      "grad_norm": 0.21365907788276672,
+      "learning_rate": 3.588790662050788e-05,
+      "loss": 1.0716,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5924418251412601,
+      "grad_norm": 0.11712610721588135,
+      "learning_rate": 3.587872330372918e-05,
+      "loss": 1.0333,
+      "step": 9751
+    },
+    {
+      "epoch": 0.5925025821738866,
+      "grad_norm": 0.1176094338297844,
+      "learning_rate": 3.58695405045686e-05,
+      "loss": 1.0008,
+      "step": 9752
+    },
+    {
+      "epoch": 0.5925633392065132,
+      "grad_norm": 0.14124080538749695,
+      "learning_rate": 3.586035822336277e-05,
+      "loss": 1.0401,
+      "step": 9753
+    },
+    {
+      "epoch": 0.5926240962391397,
+      "grad_norm": 0.11916171759366989,
+      "learning_rate": 3.5851176460448216e-05,
+      "loss": 1.0361,
+      "step": 9754
+    },
+    {
+      "epoch": 0.5926848532717662,
+      "grad_norm": 0.13915252685546875,
+      "learning_rate": 3.5841995216161514e-05,
+      "loss": 1.1064,
+      "step": 9755
+    },
+    {
+      "epoch": 0.5927456103043928,
+      "grad_norm": 0.10760362446308136,
+      "learning_rate": 3.583281449083924e-05,
+      "loss": 1.0268,
+      "step": 9756
+    },
+    {
+      "epoch": 0.5928063673370193,
+      "grad_norm": 0.21113231778144836,
+      "learning_rate": 3.582363428481788e-05,
+      "loss": 1.0781,
+      "step": 9757
+    },
+    {
+      "epoch": 0.5928671243696458,
+      "grad_norm": 0.41020911931991577,
+      "learning_rate": 3.5814454598433936e-05,
+      "loss": 1.2678,
+      "step": 9758
+    },
+    {
+      "epoch": 0.5929278814022723,
+      "grad_norm": 0.11867465078830719,
+      "learning_rate": 3.58052754320239e-05,
+      "loss": 1.0106,
+      "step": 9759
+    },
+    {
+      "epoch": 0.5929886384348988,
+      "grad_norm": 0.18245571851730347,
+      "learning_rate": 3.579609678592426e-05,
+      "loss": 1.0353,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5930493954675253,
+      "grad_norm": 0.1418563574552536,
+      "learning_rate": 3.5786918660471424e-05,
+      "loss": 1.0813,
+      "step": 9761
+    },
+    {
+      "epoch": 0.5931101525001519,
+      "grad_norm": 0.2041768878698349,
+      "learning_rate": 3.5777741056001834e-05,
+      "loss": 1.0359,
+      "step": 9762
+    },
+    {
+      "epoch": 0.5931709095327784,
+      "grad_norm": 0.13422484695911407,
+      "learning_rate": 3.5768563972851886e-05,
+      "loss": 1.0514,
+      "step": 9763
+    },
+    {
+      "epoch": 0.5932316665654049,
+      "grad_norm": 0.30140751600265503,
+      "learning_rate": 3.575938741135798e-05,
+      "loss": 1.0959,
+      "step": 9764
+    },
+    {
+      "epoch": 0.5932924235980315,
+      "grad_norm": 0.2591914236545563,
+      "learning_rate": 3.575021137185649e-05,
+      "loss": 1.0814,
+      "step": 9765
+    },
+    {
+      "epoch": 0.593353180630658,
+      "grad_norm": 0.6737829446792603,
+      "learning_rate": 3.574103585468376e-05,
+      "loss": 1.1226,
+      "step": 9766
+    },
+    {
+      "epoch": 0.5934139376632845,
+      "grad_norm": 0.17212356626987457,
+      "learning_rate": 3.57318608601761e-05,
+      "loss": 1.1293,
+      "step": 9767
+    },
+    {
+      "epoch": 0.5934746946959111,
+      "grad_norm": 0.12765397131443024,
+      "learning_rate": 3.572268638866985e-05,
+      "loss": 1.0669,
+      "step": 9768
+    },
+    {
+      "epoch": 0.5935354517285376,
+      "grad_norm": 1.4577455520629883,
+      "learning_rate": 3.5713512440501276e-05,
+      "loss": 1.1219,
+      "step": 9769
+    },
+    {
+      "epoch": 0.5935962087611641,
+      "grad_norm": 0.15666432678699493,
+      "learning_rate": 3.570433901600668e-05,
+      "loss": 1.0548,
+      "step": 9770
+    },
+    {
+      "epoch": 0.5936569657937907,
+      "grad_norm": 0.1695893555879593,
+      "learning_rate": 3.569516611552228e-05,
+      "loss": 1.0468,
+      "step": 9771
+    },
+    {
+      "epoch": 0.5937177228264171,
+      "grad_norm": 0.13557523488998413,
+      "learning_rate": 3.568599373938435e-05,
+      "loss": 1.0182,
+      "step": 9772
+    },
+    {
+      "epoch": 0.5937784798590436,
+      "grad_norm": 0.1633175015449524,
+      "learning_rate": 3.5676821887929095e-05,
+      "loss": 1.0461,
+      "step": 9773
+    },
+    {
+      "epoch": 0.5938392368916702,
+      "grad_norm": 0.21911802887916565,
+      "learning_rate": 3.566765056149269e-05,
+      "loss": 1.2169,
+      "step": 9774
+    },
+    {
+      "epoch": 0.5938999939242967,
+      "grad_norm": 0.17607736587524414,
+      "learning_rate": 3.5658479760411326e-05,
+      "loss": 1.08,
+      "step": 9775
+    },
+    {
+      "epoch": 0.5939607509569232,
+      "grad_norm": 0.3913068175315857,
+      "learning_rate": 3.564930948502115e-05,
+      "loss": 1.1745,
+      "step": 9776
+    },
+    {
+      "epoch": 0.5940215079895498,
+      "grad_norm": 0.14123575389385223,
+      "learning_rate": 3.564013973565832e-05,
+      "loss": 1.0913,
+      "step": 9777
+    },
+    {
+      "epoch": 0.5940822650221763,
+      "grad_norm": 0.18459638953208923,
+      "learning_rate": 3.5630970512658935e-05,
+      "loss": 1.0579,
+      "step": 9778
+    },
+    {
+      "epoch": 0.5941430220548028,
+      "grad_norm": 0.2039920538663864,
+      "learning_rate": 3.56218018163591e-05,
+      "loss": 1.0352,
+      "step": 9779
+    },
+    {
+      "epoch": 0.5942037790874294,
+      "grad_norm": 0.16813644766807556,
+      "learning_rate": 3.561263364709493e-05,
+      "loss": 1.1772,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5942645361200559,
+      "grad_norm": 0.2098015695810318,
+      "learning_rate": 3.560346600520242e-05,
+      "loss": 1.1171,
+      "step": 9781
+    },
+    {
+      "epoch": 0.5943252931526825,
+      "grad_norm": 0.1300709992647171,
+      "learning_rate": 3.5594298891017654e-05,
+      "loss": 1.0222,
+      "step": 9782
+    },
+    {
+      "epoch": 0.594386050185309,
+      "grad_norm": 0.11240896582603455,
+      "learning_rate": 3.5585132304876655e-05,
+      "loss": 1.0746,
+      "step": 9783
+    },
+    {
+      "epoch": 0.5944468072179355,
+      "grad_norm": 0.27615344524383545,
+      "learning_rate": 3.55759662471154e-05,
+      "loss": 1.0439,
+      "step": 9784
+    },
+    {
+      "epoch": 0.594507564250562,
+      "grad_norm": 0.47950446605682373,
+      "learning_rate": 3.5566800718069904e-05,
+      "loss": 1.1026,
+      "step": 9785
+    },
+    {
+      "epoch": 0.5945683212831885,
+      "grad_norm": 0.14027227461338043,
+      "learning_rate": 3.555763571807613e-05,
+      "loss": 1.0509,
+      "step": 9786
+    },
+    {
+      "epoch": 0.594629078315815,
+      "grad_norm": 0.11497654020786285,
+      "learning_rate": 3.5548471247469994e-05,
+      "loss": 1.0836,
+      "step": 9787
+    },
+    {
+      "epoch": 0.5946898353484416,
+      "grad_norm": 0.19415542483329773,
+      "learning_rate": 3.5539307306587444e-05,
+      "loss": 1.0364,
+      "step": 9788
+    },
+    {
+      "epoch": 0.5947505923810681,
+      "grad_norm": 0.15232017636299133,
+      "learning_rate": 3.553014389576437e-05,
+      "loss": 1.0243,
+      "step": 9789
+    },
+    {
+      "epoch": 0.5948113494136946,
+      "grad_norm": 0.13926400244235992,
+      "learning_rate": 3.552098101533667e-05,
+      "loss": 1.0284,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5948721064463212,
+      "grad_norm": 0.1758466362953186,
+      "learning_rate": 3.5511818665640215e-05,
+      "loss": 1.0182,
+      "step": 9791
+    },
+    {
+      "epoch": 0.5949328634789477,
+      "grad_norm": 0.1317562609910965,
+      "learning_rate": 3.550265684701084e-05,
+      "loss": 1.0125,
+      "step": 9792
+    },
+    {
+      "epoch": 0.5949936205115742,
+      "grad_norm": 0.1492239236831665,
+      "learning_rate": 3.5493495559784405e-05,
+      "loss": 0.9802,
+      "step": 9793
+    },
+    {
+      "epoch": 0.5950543775442008,
+      "grad_norm": 0.1287379264831543,
+      "learning_rate": 3.5484334804296674e-05,
+      "loss": 1.0813,
+      "step": 9794
+    },
+    {
+      "epoch": 0.5951151345768273,
+      "grad_norm": 0.10461851209402084,
+      "learning_rate": 3.5475174580883456e-05,
+      "loss": 1.0211,
+      "step": 9795
+    },
+    {
+      "epoch": 0.5951758916094538,
+      "grad_norm": 0.22647830843925476,
+      "learning_rate": 3.546601488988053e-05,
+      "loss": 1.1452,
+      "step": 9796
+    },
+    {
+      "epoch": 0.5952366486420804,
+      "grad_norm": 0.259226530790329,
+      "learning_rate": 3.545685573162362e-05,
+      "loss": 1.0191,
+      "step": 9797
+    },
+    {
+      "epoch": 0.5952974056747068,
+      "grad_norm": 0.11776480823755264,
+      "learning_rate": 3.544769710644849e-05,
+      "loss": 1.0224,
+      "step": 9798
+    },
+    {
+      "epoch": 0.5953581627073333,
+      "grad_norm": 0.12759307026863098,
+      "learning_rate": 3.543853901469083e-05,
+      "loss": 1.0401,
+      "step": 9799
+    },
+    {
+      "epoch": 0.5954189197399599,
+      "grad_norm": 0.17199723422527313,
+      "learning_rate": 3.542938145668635e-05,
+      "loss": 1.1237,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5954796767725864,
+      "grad_norm": 0.11512981355190277,
+      "learning_rate": 3.54202244327707e-05,
+      "loss": 1.0424,
+      "step": 9801
+    },
+    {
+      "epoch": 0.5955404338052129,
+      "grad_norm": 0.11472322791814804,
+      "learning_rate": 3.541106794327953e-05,
+      "loss": 1.0312,
+      "step": 9802
+    },
+    {
+      "epoch": 0.5956011908378395,
+      "grad_norm": 0.11672493070363998,
+      "learning_rate": 3.54019119885485e-05,
+      "loss": 1.0568,
+      "step": 9803
+    },
+    {
+      "epoch": 0.595661947870466,
+      "grad_norm": 0.2715204954147339,
+      "learning_rate": 3.53927565689132e-05,
+      "loss": 1.0753,
+      "step": 9804
+    },
+    {
+      "epoch": 0.5957227049030925,
+      "grad_norm": 0.1581743210554123,
+      "learning_rate": 3.538360168470922e-05,
+      "loss": 1.0758,
+      "step": 9805
+    },
+    {
+      "epoch": 0.5957834619357191,
+      "grad_norm": 0.1087142825126648,
+      "learning_rate": 3.537444733627219e-05,
+      "loss": 1.0104,
+      "step": 9806
+    },
+    {
+      "epoch": 0.5958442189683456,
+      "grad_norm": 0.12890246510505676,
+      "learning_rate": 3.536529352393757e-05,
+      "loss": 1.0681,
+      "step": 9807
+    },
+    {
+      "epoch": 0.5959049760009721,
+      "grad_norm": 0.12123747169971466,
+      "learning_rate": 3.535614024804097e-05,
+      "loss": 1.0417,
+      "step": 9808
+    },
+    {
+      "epoch": 0.5959657330335987,
+      "grad_norm": 0.2113116830587387,
+      "learning_rate": 3.5346987508917876e-05,
+      "loss": 1.0876,
+      "step": 9809
+    },
+    {
+      "epoch": 0.5960264900662252,
+      "grad_norm": 0.36501821875572205,
+      "learning_rate": 3.533783530690377e-05,
+      "loss": 1.2459,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5960872470988517,
+      "grad_norm": 0.141757994890213,
+      "learning_rate": 3.532868364233416e-05,
+      "loss": 1.0997,
+      "step": 9811
+    },
+    {
+      "epoch": 0.5961480041314782,
+      "grad_norm": 0.22064721584320068,
+      "learning_rate": 3.531953251554449e-05,
+      "loss": 1.2755,
+      "step": 9812
+    },
+    {
+      "epoch": 0.5962087611641047,
+      "grad_norm": 0.2296675741672516,
+      "learning_rate": 3.531038192687021e-05,
+      "loss": 1.1259,
+      "step": 9813
+    },
+    {
+      "epoch": 0.5962695181967312,
+      "grad_norm": 0.13480280339717865,
+      "learning_rate": 3.530123187664669e-05,
+      "loss": 1.0737,
+      "step": 9814
+    },
+    {
+      "epoch": 0.5963302752293578,
+      "grad_norm": 0.1989314705133438,
+      "learning_rate": 3.529208236520936e-05,
+      "loss": 1.0736,
+      "step": 9815
+    },
+    {
+      "epoch": 0.5963910322619843,
+      "grad_norm": 0.2590467035770416,
+      "learning_rate": 3.52829333928936e-05,
+      "loss": 0.9889,
+      "step": 9816
+    },
+    {
+      "epoch": 0.5964517892946108,
+      "grad_norm": 0.19125080108642578,
+      "learning_rate": 3.527378496003476e-05,
+      "loss": 1.1253,
+      "step": 9817
+    },
+    {
+      "epoch": 0.5965125463272374,
+      "grad_norm": 0.12208377569913864,
+      "learning_rate": 3.5264637066968175e-05,
+      "loss": 1.0019,
+      "step": 9818
+    },
+    {
+      "epoch": 0.5965733033598639,
+      "grad_norm": 0.18308749794960022,
+      "learning_rate": 3.5255489714029176e-05,
+      "loss": 1.0547,
+      "step": 9819
+    },
+    {
+      "epoch": 0.5966340603924905,
+      "grad_norm": 0.22826820611953735,
+      "learning_rate": 3.524634290155306e-05,
+      "loss": 1.1794,
+      "step": 9820
+    },
+    {
+      "epoch": 0.596694817425117,
+      "grad_norm": 0.12926438450813293,
+      "learning_rate": 3.523719662987508e-05,
+      "loss": 1.0652,
+      "step": 9821
+    },
+    {
+      "epoch": 0.5967555744577435,
+      "grad_norm": 0.23061497509479523,
+      "learning_rate": 3.522805089933052e-05,
+      "loss": 1.1796,
+      "step": 9822
+    },
+    {
+      "epoch": 0.5968163314903701,
+      "grad_norm": 0.18201753497123718,
+      "learning_rate": 3.52189057102546e-05,
+      "loss": 1.1523,
+      "step": 9823
+    },
+    {
+      "epoch": 0.5968770885229966,
+      "grad_norm": 0.39602550864219666,
+      "learning_rate": 3.520976106298255e-05,
+      "loss": 1.132,
+      "step": 9824
+    },
+    {
+      "epoch": 0.596937845555623,
+      "grad_norm": 0.9689731597900391,
+      "learning_rate": 3.520061695784958e-05,
+      "loss": 1.0133,
+      "step": 9825
+    },
+    {
+      "epoch": 0.5969986025882495,
+      "grad_norm": 0.1577349752187729,
+      "learning_rate": 3.5191473395190834e-05,
+      "loss": 1.0683,
+      "step": 9826
+    },
+    {
+      "epoch": 0.5970593596208761,
+      "grad_norm": 0.12769648432731628,
+      "learning_rate": 3.518233037534151e-05,
+      "loss": 1.075,
+      "step": 9827
+    },
+    {
+      "epoch": 0.5971201166535026,
+      "grad_norm": 8.29349422454834,
+      "learning_rate": 3.517318789863673e-05,
+      "loss": 1.039,
+      "step": 9828
+    },
+    {
+      "epoch": 0.5971808736861292,
+      "grad_norm": 1.2124171257019043,
+      "learning_rate": 3.516404596541161e-05,
+      "loss": 1.0738,
+      "step": 9829
+    },
+    {
+      "epoch": 0.5972416307187557,
+      "grad_norm": 0.1338082104921341,
+      "learning_rate": 3.515490457600126e-05,
+      "loss": 1.0306,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5973023877513822,
+      "grad_norm": 0.12332873791456223,
+      "learning_rate": 3.5145763730740735e-05,
+      "loss": 1.0568,
+      "step": 9831
+    },
+    {
+      "epoch": 0.5973631447840088,
+      "grad_norm": 0.1417461782693863,
+      "learning_rate": 3.513662342996512e-05,
+      "loss": 1.0009,
+      "step": 9832
+    },
+    {
+      "epoch": 0.5974239018166353,
+      "grad_norm": 0.16883361339569092,
+      "learning_rate": 3.512748367400945e-05,
+      "loss": 1.0215,
+      "step": 9833
+    },
+    {
+      "epoch": 0.5974846588492618,
+      "grad_norm": 0.1710953414440155,
+      "learning_rate": 3.511834446320875e-05,
+      "loss": 1.1178,
+      "step": 9834
+    },
+    {
+      "epoch": 0.5975454158818884,
+      "grad_norm": 0.1153365820646286,
+      "learning_rate": 3.510920579789799e-05,
+      "loss": 1.0419,
+      "step": 9835
+    },
+    {
+      "epoch": 0.5976061729145149,
+      "grad_norm": 0.2657296061515808,
+      "learning_rate": 3.510006767841218e-05,
+      "loss": 1.1383,
+      "step": 9836
+    },
+    {
+      "epoch": 0.5976669299471414,
+      "grad_norm": 0.15079621970653534,
+      "learning_rate": 3.5090930105086264e-05,
+      "loss": 1.0884,
+      "step": 9837
+    },
+    {
+      "epoch": 0.5977276869797679,
+      "grad_norm": 0.1657806783914566,
+      "learning_rate": 3.508179307825518e-05,
+      "loss": 1.1339,
+      "step": 9838
+    },
+    {
+      "epoch": 0.5977884440123944,
+      "grad_norm": 0.22033043205738068,
+      "learning_rate": 3.5072656598253875e-05,
+      "loss": 1.0688,
+      "step": 9839
+    },
+    {
+      "epoch": 0.5978492010450209,
+      "grad_norm": 0.19386543333530426,
+      "learning_rate": 3.506352066541723e-05,
+      "loss": 1.1037,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5979099580776475,
+      "grad_norm": 0.18420758843421936,
+      "learning_rate": 3.5054385280080104e-05,
+      "loss": 1.132,
+      "step": 9841
+    },
+    {
+      "epoch": 0.597970715110274,
+      "grad_norm": 0.13463185727596283,
+      "learning_rate": 3.504525044257738e-05,
+      "loss": 1.1056,
+      "step": 9842
+    },
+    {
+      "epoch": 0.5980314721429005,
+      "grad_norm": 0.1635916382074356,
+      "learning_rate": 3.503611615324388e-05,
+      "loss": 1.087,
+      "step": 9843
+    },
+    {
+      "epoch": 0.5980922291755271,
+      "grad_norm": 0.2804458439350128,
+      "learning_rate": 3.502698241241445e-05,
+      "loss": 1.0993,
+      "step": 9844
+    },
+    {
+      "epoch": 0.5981529862081536,
+      "grad_norm": 0.1420588195323944,
+      "learning_rate": 3.501784922042387e-05,
+      "loss": 1.0947,
+      "step": 9845
+    },
+    {
+      "epoch": 0.5982137432407801,
+      "grad_norm": 0.18791642785072327,
+      "learning_rate": 3.500871657760692e-05,
+      "loss": 1.0317,
+      "step": 9846
+    },
+    {
+      "epoch": 0.5982745002734067,
+      "grad_norm": 0.25479090213775635,
+      "learning_rate": 3.4999584484298385e-05,
+      "loss": 1.012,
+      "step": 9847
+    },
+    {
+      "epoch": 0.5983352573060332,
+      "grad_norm": 0.22574377059936523,
+      "learning_rate": 3.4990452940832954e-05,
+      "loss": 1.112,
+      "step": 9848
+    },
+    {
+      "epoch": 0.5983960143386597,
+      "grad_norm": 0.16704055666923523,
+      "learning_rate": 3.4981321947545374e-05,
+      "loss": 1.062,
+      "step": 9849
+    },
+    {
+      "epoch": 0.5984567713712863,
+      "grad_norm": 0.15069197118282318,
+      "learning_rate": 3.497219150477035e-05,
+      "loss": 1.0945,
+      "step": 9850
+    },
+    {
+      "epoch": 0.5985175284039127,
+      "grad_norm": 0.2017909288406372,
+      "learning_rate": 3.496306161284254e-05,
+      "loss": 1.1254,
+      "step": 9851
+    },
+    {
+      "epoch": 0.5985782854365392,
+      "grad_norm": 0.1575460433959961,
+      "learning_rate": 3.495393227209662e-05,
+      "loss": 1.095,
+      "step": 9852
+    },
+    {
+      "epoch": 0.5986390424691658,
+      "grad_norm": 0.19820109009742737,
+      "learning_rate": 3.4944803482867225e-05,
+      "loss": 1.0684,
+      "step": 9853
+    },
+    {
+      "epoch": 0.5986997995017923,
+      "grad_norm": 0.1419120877981186,
+      "learning_rate": 3.4935675245488976e-05,
+      "loss": 1.1057,
+      "step": 9854
+    },
+    {
+      "epoch": 0.5987605565344188,
+      "grad_norm": 0.22121503949165344,
+      "learning_rate": 3.4926547560296445e-05,
+      "loss": 1.0533,
+      "step": 9855
+    },
+    {
+      "epoch": 0.5988213135670454,
+      "grad_norm": 0.22658799588680267,
+      "learning_rate": 3.4917420427624224e-05,
+      "loss": 1.0575,
+      "step": 9856
+    },
+    {
+      "epoch": 0.5988820705996719,
+      "grad_norm": 0.13354939222335815,
+      "learning_rate": 3.4908293847806885e-05,
+      "loss": 1.092,
+      "step": 9857
+    },
+    {
+      "epoch": 0.5989428276322984,
+      "grad_norm": 0.18550904095172882,
+      "learning_rate": 3.489916782117895e-05,
+      "loss": 1.096,
+      "step": 9858
+    },
+    {
+      "epoch": 0.599003584664925,
+      "grad_norm": 0.14349329471588135,
+      "learning_rate": 3.489004234807492e-05,
+      "loss": 1.0385,
+      "step": 9859
+    },
+    {
+      "epoch": 0.5990643416975515,
+      "grad_norm": 0.11136182397603989,
+      "learning_rate": 3.488091742882935e-05,
+      "loss": 1.066,
+      "step": 9860
+    },
+    {
+      "epoch": 0.599125098730178,
+      "grad_norm": 0.17109361290931702,
+      "learning_rate": 3.4871793063776636e-05,
+      "loss": 1.0444,
+      "step": 9861
+    },
+    {
+      "epoch": 0.5991858557628046,
+      "grad_norm": 0.4811965227127075,
+      "learning_rate": 3.486266925325129e-05,
+      "loss": 1.1447,
+      "step": 9862
+    },
+    {
+      "epoch": 0.5992466127954311,
+      "grad_norm": 0.15174521505832672,
+      "learning_rate": 3.4853545997587725e-05,
+      "loss": 1.1018,
+      "step": 9863
+    },
+    {
+      "epoch": 0.5993073698280575,
+      "grad_norm": 0.17021965980529785,
+      "learning_rate": 3.4844423297120345e-05,
+      "loss": 1.1157,
+      "step": 9864
+    },
+    {
+      "epoch": 0.5993681268606841,
+      "grad_norm": 10.048443794250488,
+      "learning_rate": 3.4835301152183575e-05,
+      "loss": 1.1272,
+      "step": 9865
+    },
+    {
+      "epoch": 0.5994288838933106,
+      "grad_norm": 0.16117344796657562,
+      "learning_rate": 3.4826179563111764e-05,
+      "loss": 1.0734,
+      "step": 9866
+    },
+    {
+      "epoch": 0.5994896409259372,
+      "grad_norm": 0.14404891431331635,
+      "learning_rate": 3.4817058530239285e-05,
+      "loss": 1.0993,
+      "step": 9867
+    },
+    {
+      "epoch": 0.5995503979585637,
+      "grad_norm": 0.14917537569999695,
+      "learning_rate": 3.480793805390046e-05,
+      "loss": 1.036,
+      "step": 9868
+    },
+    {
+      "epoch": 0.5996111549911902,
+      "grad_norm": 0.2056112140417099,
+      "learning_rate": 3.479881813442959e-05,
+      "loss": 1.0609,
+      "step": 9869
+    },
+    {
+      "epoch": 0.5996719120238168,
+      "grad_norm": 0.15713055431842804,
+      "learning_rate": 3.4789698772160986e-05,
+      "loss": 0.9829,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5997326690564433,
+      "grad_norm": 0.13040582835674286,
+      "learning_rate": 3.478057996742891e-05,
+      "loss": 1.0383,
+      "step": 9871
+    },
+    {
+      "epoch": 0.5997934260890698,
+      "grad_norm": 0.32370424270629883,
+      "learning_rate": 3.477146172056761e-05,
+      "loss": 1.0757,
+      "step": 9872
+    },
+    {
+      "epoch": 0.5998541831216964,
+      "grad_norm": 0.2965570390224457,
+      "learning_rate": 3.4762344031911354e-05,
+      "loss": 1.2418,
+      "step": 9873
+    },
+    {
+      "epoch": 0.5999149401543229,
+      "grad_norm": 0.21298748254776,
+      "learning_rate": 3.4753226901794294e-05,
+      "loss": 1.2357,
+      "step": 9874
+    },
+    {
+      "epoch": 0.5999756971869494,
+      "grad_norm": 0.2686533033847809,
+      "learning_rate": 3.474411033055066e-05,
+      "loss": 1.141,
+      "step": 9875
+    },
+    {
+      "epoch": 0.600036454219576,
+      "grad_norm": 0.2947900593280792,
+      "learning_rate": 3.473499431851461e-05,
+      "loss": 1.1134,
+      "step": 9876
+    },
+    {
+      "epoch": 0.6000972112522024,
+      "grad_norm": 0.1630922257900238,
+      "learning_rate": 3.472587886602028e-05,
+      "loss": 1.0854,
+      "step": 9877
+    },
+    {
+      "epoch": 0.6001579682848289,
+      "grad_norm": 0.27238163352012634,
+      "learning_rate": 3.471676397340183e-05,
+      "loss": 1.1999,
+      "step": 9878
+    },
+    {
+      "epoch": 0.6002187253174555,
+      "grad_norm": 0.23262737691402435,
+      "learning_rate": 3.470764964099334e-05,
+      "loss": 1.092,
+      "step": 9879
+    },
+    {
+      "epoch": 0.600279482350082,
+      "grad_norm": 0.15897086262702942,
+      "learning_rate": 3.4698535869128936e-05,
+      "loss": 1.0942,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6003402393827085,
+      "grad_norm": 0.20158566534519196,
+      "learning_rate": 3.4689422658142635e-05,
+      "loss": 1.0326,
+      "step": 9881
+    },
+    {
+      "epoch": 0.6004009964153351,
+      "grad_norm": 0.13681265711784363,
+      "learning_rate": 3.46803100083685e-05,
+      "loss": 1.0768,
+      "step": 9882
+    },
+    {
+      "epoch": 0.6004617534479616,
+      "grad_norm": 0.1403191089630127,
+      "learning_rate": 3.4671197920140564e-05,
+      "loss": 1.0179,
+      "step": 9883
+    },
+    {
+      "epoch": 0.6005225104805881,
+      "grad_norm": 0.14691096544265747,
+      "learning_rate": 3.466208639379284e-05,
+      "loss": 1.1037,
+      "step": 9884
+    },
+    {
+      "epoch": 0.6005832675132147,
+      "grad_norm": 0.1478205919265747,
+      "learning_rate": 3.465297542965929e-05,
+      "loss": 1.0709,
+      "step": 9885
+    },
+    {
+      "epoch": 0.6006440245458412,
+      "grad_norm": 1.6108773946762085,
+      "learning_rate": 3.464386502807391e-05,
+      "loss": 1.0479,
+      "step": 9886
+    },
+    {
+      "epoch": 0.6007047815784677,
+      "grad_norm": 0.10805761814117432,
+      "learning_rate": 3.463475518937063e-05,
+      "loss": 1.0305,
+      "step": 9887
+    },
+    {
+      "epoch": 0.6007655386110943,
+      "grad_norm": 0.14876708388328552,
+      "learning_rate": 3.4625645913883355e-05,
+      "loss": 1.1031,
+      "step": 9888
+    },
+    {
+      "epoch": 0.6008262956437208,
+      "grad_norm": 0.2512546479701996,
+      "learning_rate": 3.4616537201946e-05,
+      "loss": 1.1995,
+      "step": 9889
+    },
+    {
+      "epoch": 0.6008870526763472,
+      "grad_norm": 0.1403788924217224,
+      "learning_rate": 3.460742905389244e-05,
+      "loss": 1.0758,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6009478097089738,
+      "grad_norm": 0.12643319368362427,
+      "learning_rate": 3.459832147005656e-05,
+      "loss": 1.0355,
+      "step": 9891
+    },
+    {
+      "epoch": 0.6010085667416003,
+      "grad_norm": 0.2717328667640686,
+      "learning_rate": 3.458921445077218e-05,
+      "loss": 1.0379,
+      "step": 9892
+    },
+    {
+      "epoch": 0.6010693237742268,
+      "grad_norm": 0.14643657207489014,
+      "learning_rate": 3.4580107996373106e-05,
+      "loss": 1.0712,
+      "step": 9893
+    },
+    {
+      "epoch": 0.6011300808068534,
+      "grad_norm": 0.11006568372249603,
+      "learning_rate": 3.4571002107193195e-05,
+      "loss": 1.0195,
+      "step": 9894
+    },
+    {
+      "epoch": 0.6011908378394799,
+      "grad_norm": 0.1936243176460266,
+      "learning_rate": 3.4561896783566145e-05,
+      "loss": 1.1409,
+      "step": 9895
+    },
+    {
+      "epoch": 0.6012515948721064,
+      "grad_norm": 0.16525816917419434,
+      "learning_rate": 3.455279202582577e-05,
+      "loss": 1.1443,
+      "step": 9896
+    },
+    {
+      "epoch": 0.601312351904733,
+      "grad_norm": 0.2687567472457886,
+      "learning_rate": 3.454368783430578e-05,
+      "loss": 1.236,
+      "step": 9897
+    },
+    {
+      "epoch": 0.6013731089373595,
+      "grad_norm": 0.19922126829624176,
+      "learning_rate": 3.453458420933989e-05,
+      "loss": 1.0721,
+      "step": 9898
+    },
+    {
+      "epoch": 0.601433865969986,
+      "grad_norm": 0.1761442869901657,
+      "learning_rate": 3.452548115126183e-05,
+      "loss": 1.0249,
+      "step": 9899
+    },
+    {
+      "epoch": 0.6014946230026126,
+      "grad_norm": 0.3242648243904114,
+      "learning_rate": 3.451637866040524e-05,
+      "loss": 1.2274,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6015553800352391,
+      "grad_norm": 0.17694208025932312,
+      "learning_rate": 3.450727673710379e-05,
+      "loss": 1.1644,
+      "step": 9901
+    },
+    {
+      "epoch": 0.6016161370678657,
+      "grad_norm": 0.14571638405323029,
+      "learning_rate": 3.44981753816911e-05,
+      "loss": 1.0457,
+      "step": 9902
+    },
+    {
+      "epoch": 0.6016768941004921,
+      "grad_norm": 0.2872547209262848,
+      "learning_rate": 3.448907459450078e-05,
+      "loss": 1.1491,
+      "step": 9903
+    },
+    {
+      "epoch": 0.6017376511331186,
+      "grad_norm": 0.11442230641841888,
+      "learning_rate": 3.447997437586643e-05,
+      "loss": 1.0131,
+      "step": 9904
+    },
+    {
+      "epoch": 0.6017984081657451,
+      "grad_norm": 0.1830238550901413,
+      "learning_rate": 3.447087472612163e-05,
+      "loss": 1.0277,
+      "step": 9905
+    },
+    {
+      "epoch": 0.6018591651983717,
+      "grad_norm": 0.12001951783895493,
+      "learning_rate": 3.44617756455999e-05,
+      "loss": 1.0789,
+      "step": 9906
+    },
+    {
+      "epoch": 0.6019199222309982,
+      "grad_norm": 0.21640633046627045,
+      "learning_rate": 3.445267713463482e-05,
+      "loss": 1.1854,
+      "step": 9907
+    },
+    {
+      "epoch": 0.6019806792636248,
+      "grad_norm": 0.2306807041168213,
+      "learning_rate": 3.444357919355984e-05,
+      "loss": 1.1046,
+      "step": 9908
+    },
+    {
+      "epoch": 0.6020414362962513,
+      "grad_norm": 0.22462685406208038,
+      "learning_rate": 3.443448182270848e-05,
+      "loss": 1.1133,
+      "step": 9909
+    },
+    {
+      "epoch": 0.6021021933288778,
+      "grad_norm": 0.1434704214334488,
+      "learning_rate": 3.442538502241419e-05,
+      "loss": 1.061,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6021629503615044,
+      "grad_norm": 0.17077481746673584,
+      "learning_rate": 3.441628879301042e-05,
+      "loss": 1.0613,
+      "step": 9911
+    },
+    {
+      "epoch": 0.6022237073941309,
+      "grad_norm": 0.25269031524658203,
+      "learning_rate": 3.44071931348306e-05,
+      "loss": 1.0108,
+      "step": 9912
+    },
+    {
+      "epoch": 0.6022844644267574,
+      "grad_norm": 0.17380735278129578,
+      "learning_rate": 3.439809804820814e-05,
+      "loss": 1.0152,
+      "step": 9913
+    },
+    {
+      "epoch": 0.602345221459384,
+      "grad_norm": 0.13962875306606293,
+      "learning_rate": 3.438900353347641e-05,
+      "loss": 1.0166,
+      "step": 9914
+    },
+    {
+      "epoch": 0.6024059784920105,
+      "grad_norm": 0.1959177702665329,
+      "learning_rate": 3.437990959096877e-05,
+      "loss": 1.0248,
+      "step": 9915
+    },
+    {
+      "epoch": 0.602466735524637,
+      "grad_norm": 0.1793917566537857,
+      "learning_rate": 3.437081622101855e-05,
+      "loss": 1.0951,
+      "step": 9916
+    },
+    {
+      "epoch": 0.6025274925572635,
+      "grad_norm": 0.16855919361114502,
+      "learning_rate": 3.436172342395909e-05,
+      "loss": 1.075,
+      "step": 9917
+    },
+    {
+      "epoch": 0.60258824958989,
+      "grad_norm": 0.18728770315647125,
+      "learning_rate": 3.435263120012369e-05,
+      "loss": 1.0791,
+      "step": 9918
+    },
+    {
+      "epoch": 0.6026490066225165,
+      "grad_norm": 0.16851933300495148,
+      "learning_rate": 3.4343539549845604e-05,
+      "loss": 1.0714,
+      "step": 9919
+    },
+    {
+      "epoch": 0.6027097636551431,
+      "grad_norm": 0.20317442715168,
+      "learning_rate": 3.433444847345812e-05,
+      "loss": 1.1057,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6027705206877696,
+      "grad_norm": 0.1339462697505951,
+      "learning_rate": 3.4325357971294447e-05,
+      "loss": 1.0451,
+      "step": 9921
+    },
+    {
+      "epoch": 0.6028312777203961,
+      "grad_norm": 0.6338938474655151,
+      "learning_rate": 3.4316268043687806e-05,
+      "loss": 1.0922,
+      "step": 9922
+    },
+    {
+      "epoch": 0.6028920347530227,
+      "grad_norm": 0.24797533452510834,
+      "learning_rate": 3.430717869097139e-05,
+      "loss": 1.0616,
+      "step": 9923
+    },
+    {
+      "epoch": 0.6029527917856492,
+      "grad_norm": 0.16027776896953583,
+      "learning_rate": 3.4298089913478365e-05,
+      "loss": 1.0415,
+      "step": 9924
+    },
+    {
+      "epoch": 0.6030135488182757,
+      "grad_norm": 3.4841666221618652,
+      "learning_rate": 3.42890017115419e-05,
+      "loss": 1.0353,
+      "step": 9925
+    },
+    {
+      "epoch": 0.6030743058509023,
+      "grad_norm": 0.1936003565788269,
+      "learning_rate": 3.427991408549511e-05,
+      "loss": 1.0495,
+      "step": 9926
+    },
+    {
+      "epoch": 0.6031350628835288,
+      "grad_norm": 0.26577481627464294,
+      "learning_rate": 3.427082703567113e-05,
+      "loss": 1.1489,
+      "step": 9927
+    },
+    {
+      "epoch": 0.6031958199161553,
+      "grad_norm": 0.1490364819765091,
+      "learning_rate": 3.426174056240299e-05,
+      "loss": 1.1031,
+      "step": 9928
+    },
+    {
+      "epoch": 0.6032565769487819,
+      "grad_norm": 0.1860506385564804,
+      "learning_rate": 3.425265466602381e-05,
+      "loss": 1.0159,
+      "step": 9929
+    },
+    {
+      "epoch": 0.6033173339814083,
+      "grad_norm": 0.22505608201026917,
+      "learning_rate": 3.424356934686661e-05,
+      "loss": 1.1298,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6033780910140348,
+      "grad_norm": 0.18013067543506622,
+      "learning_rate": 3.4234484605264416e-05,
+      "loss": 1.1182,
+      "step": 9931
+    },
+    {
+      "epoch": 0.6034388480466614,
+      "grad_norm": 0.12454739958047867,
+      "learning_rate": 3.422540044155025e-05,
+      "loss": 1.0455,
+      "step": 9932
+    },
+    {
+      "epoch": 0.6034996050792879,
+      "grad_norm": 0.1451205462217331,
+      "learning_rate": 3.421631685605707e-05,
+      "loss": 1.0411,
+      "step": 9933
+    },
+    {
+      "epoch": 0.6035603621119144,
+      "grad_norm": 0.18871988356113434,
+      "learning_rate": 3.420723384911787e-05,
+      "loss": 1.0932,
+      "step": 9934
+    },
+    {
+      "epoch": 0.603621119144541,
+      "grad_norm": 0.17945139110088348,
+      "learning_rate": 3.4198151421065546e-05,
+      "loss": 1.1294,
+      "step": 9935
+    },
+    {
+      "epoch": 0.6036818761771675,
+      "grad_norm": 0.13867826759815216,
+      "learning_rate": 3.418906957223303e-05,
+      "loss": 1.0384,
+      "step": 9936
+    },
+    {
+      "epoch": 0.603742633209794,
+      "grad_norm": 0.1369640976190567,
+      "learning_rate": 3.4179988302953235e-05,
+      "loss": 1.0742,
+      "step": 9937
+    },
+    {
+      "epoch": 0.6038033902424206,
+      "grad_norm": 0.6449360847473145,
+      "learning_rate": 3.417090761355902e-05,
+      "loss": 1.0601,
+      "step": 9938
+    },
+    {
+      "epoch": 0.6038641472750471,
+      "grad_norm": 0.17826220393180847,
+      "learning_rate": 3.416182750438325e-05,
+      "loss": 1.0608,
+      "step": 9939
+    },
+    {
+      "epoch": 0.6039249043076736,
+      "grad_norm": 0.14630477130413055,
+      "learning_rate": 3.415274797575878e-05,
+      "loss": 1.1097,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6039856613403002,
+      "grad_norm": 0.11577188223600388,
+      "learning_rate": 3.414366902801836e-05,
+      "loss": 1.0183,
+      "step": 9941
+    },
+    {
+      "epoch": 0.6040464183729267,
+      "grad_norm": 0.1503218412399292,
+      "learning_rate": 3.413459066149483e-05,
+      "loss": 1.0807,
+      "step": 9942
+    },
+    {
+      "epoch": 0.6041071754055531,
+      "grad_norm": 0.15518291294574738,
+      "learning_rate": 3.412551287652094e-05,
+      "loss": 1.0353,
+      "step": 9943
+    },
+    {
+      "epoch": 0.6041679324381797,
+      "grad_norm": 0.9498729705810547,
+      "learning_rate": 3.411643567342944e-05,
+      "loss": 1.078,
+      "step": 9944
+    },
+    {
+      "epoch": 0.6042286894708062,
+      "grad_norm": 0.5091332197189331,
+      "learning_rate": 3.410735905255307e-05,
+      "loss": 1.1926,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6042894465034327,
+      "grad_norm": 1.295035719871521,
+      "learning_rate": 3.409828301422453e-05,
+      "loss": 1.1119,
+      "step": 9946
+    },
+    {
+      "epoch": 0.6043502035360593,
+      "grad_norm": 0.13873019814491272,
+      "learning_rate": 3.408920755877651e-05,
+      "loss": 1.0396,
+      "step": 9947
+    },
+    {
+      "epoch": 0.6044109605686858,
+      "grad_norm": 0.19527703523635864,
+      "learning_rate": 3.4080132686541644e-05,
+      "loss": 1.113,
+      "step": 9948
+    },
+    {
+      "epoch": 0.6044717176013124,
+      "grad_norm": 0.3224121332168579,
+      "learning_rate": 3.4071058397852586e-05,
+      "loss": 1.1502,
+      "step": 9949
+    },
+    {
+      "epoch": 0.6045324746339389,
+      "grad_norm": 0.13711580634117126,
+      "learning_rate": 3.406198469304197e-05,
+      "loss": 1.0632,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6045932316665654,
+      "grad_norm": 0.28841862082481384,
+      "learning_rate": 3.405291157244238e-05,
+      "loss": 1.14,
+      "step": 9951
+    },
+    {
+      "epoch": 0.604653988699192,
+      "grad_norm": 0.15258243680000305,
+      "learning_rate": 3.40438390363864e-05,
+      "loss": 1.033,
+      "step": 9952
+    },
+    {
+      "epoch": 0.6047147457318185,
+      "grad_norm": 0.40781161189079285,
+      "learning_rate": 3.4034767085206586e-05,
+      "loss": 1.1834,
+      "step": 9953
+    },
+    {
+      "epoch": 0.604775502764445,
+      "grad_norm": 0.13136610388755798,
+      "learning_rate": 3.4025695719235484e-05,
+      "loss": 0.9972,
+      "step": 9954
+    },
+    {
+      "epoch": 0.6048362597970716,
+      "grad_norm": 0.13902992010116577,
+      "learning_rate": 3.401662493880559e-05,
+      "loss": 1.1063,
+      "step": 9955
+    },
+    {
+      "epoch": 0.604897016829698,
+      "grad_norm": 0.7766736149787903,
+      "learning_rate": 3.400755474424939e-05,
+      "loss": 1.1366,
+      "step": 9956
+    },
+    {
+      "epoch": 0.6049577738623245,
+      "grad_norm": 0.2598614990711212,
+      "learning_rate": 3.399848513589935e-05,
+      "loss": 1.0349,
+      "step": 9957
+    },
+    {
+      "epoch": 0.6050185308949511,
+      "grad_norm": 0.1781376600265503,
+      "learning_rate": 3.3989416114087946e-05,
+      "loss": 1.0489,
+      "step": 9958
+    },
+    {
+      "epoch": 0.6050792879275776,
+      "grad_norm": 0.11530880630016327,
+      "learning_rate": 3.398034767914759e-05,
+      "loss": 1.0068,
+      "step": 9959
+    },
+    {
+      "epoch": 0.6051400449602041,
+      "grad_norm": 0.1621524542570114,
+      "learning_rate": 3.397127983141067e-05,
+      "loss": 1.0955,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6052008019928307,
+      "grad_norm": 0.21690423786640167,
+      "learning_rate": 3.396221257120962e-05,
+      "loss": 1.0444,
+      "step": 9961
+    },
+    {
+      "epoch": 0.6052615590254572,
+      "grad_norm": 0.9953030943870544,
+      "learning_rate": 3.3953145898876746e-05,
+      "loss": 1.0638,
+      "step": 9962
+    },
+    {
+      "epoch": 0.6053223160580837,
+      "grad_norm": 0.13792288303375244,
+      "learning_rate": 3.394407981474441e-05,
+      "loss": 1.095,
+      "step": 9963
+    },
+    {
+      "epoch": 0.6053830730907103,
+      "grad_norm": 0.18514662981033325,
+      "learning_rate": 3.393501431914493e-05,
+      "loss": 1.0625,
+      "step": 9964
+    },
+    {
+      "epoch": 0.6054438301233368,
+      "grad_norm": 0.21107499301433563,
+      "learning_rate": 3.3925949412410605e-05,
+      "loss": 1.0457,
+      "step": 9965
+    },
+    {
+      "epoch": 0.6055045871559633,
+      "grad_norm": 0.20659013092517853,
+      "learning_rate": 3.391688509487372e-05,
+      "loss": 1.0561,
+      "step": 9966
+    },
+    {
+      "epoch": 0.6055653441885899,
+      "grad_norm": 0.15880298614501953,
+      "learning_rate": 3.390782136686652e-05,
+      "loss": 1.1561,
+      "step": 9967
+    },
+    {
+      "epoch": 0.6056261012212164,
+      "grad_norm": 0.16636249423027039,
+      "learning_rate": 3.389875822872124e-05,
+      "loss": 1.094,
+      "step": 9968
+    },
+    {
+      "epoch": 0.6056868582538428,
+      "grad_norm": 0.1496499478816986,
+      "learning_rate": 3.3889695680770085e-05,
+      "loss": 1.0398,
+      "step": 9969
+    },
+    {
+      "epoch": 0.6057476152864694,
+      "grad_norm": 0.1958138793706894,
+      "learning_rate": 3.388063372334524e-05,
+      "loss": 1.0476,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6058083723190959,
+      "grad_norm": 0.14606082439422607,
+      "learning_rate": 3.387157235677889e-05,
+      "loss": 1.0819,
+      "step": 9971
+    },
+    {
+      "epoch": 0.6058691293517224,
+      "grad_norm": 0.9565188884735107,
+      "learning_rate": 3.386251158140317e-05,
+      "loss": 1.0515,
+      "step": 9972
+    },
+    {
+      "epoch": 0.605929886384349,
+      "grad_norm": 0.18806388974189758,
+      "learning_rate": 3.385345139755021e-05,
+      "loss": 1.1578,
+      "step": 9973
+    },
+    {
+      "epoch": 0.6059906434169755,
+      "grad_norm": 0.21009288728237152,
+      "learning_rate": 3.3844391805552124e-05,
+      "loss": 1.1221,
+      "step": 9974
+    },
+    {
+      "epoch": 0.606051400449602,
+      "grad_norm": 0.17453421652317047,
+      "learning_rate": 3.383533280574096e-05,
+      "loss": 1.1219,
+      "step": 9975
+    },
+    {
+      "epoch": 0.6061121574822286,
+      "grad_norm": 0.133493110537529,
+      "learning_rate": 3.38262743984488e-05,
+      "loss": 1.0319,
+      "step": 9976
+    },
+    {
+      "epoch": 0.6061729145148551,
+      "grad_norm": 0.21481958031654358,
+      "learning_rate": 3.3817216584007685e-05,
+      "loss": 1.1577,
+      "step": 9977
+    },
+    {
+      "epoch": 0.6062336715474816,
+      "grad_norm": 0.12405280768871307,
+      "learning_rate": 3.3808159362749605e-05,
+      "loss": 1.051,
+      "step": 9978
+    },
+    {
+      "epoch": 0.6062944285801082,
+      "grad_norm": 0.20788033306598663,
+      "learning_rate": 3.379910273500659e-05,
+      "loss": 1.1289,
+      "step": 9979
+    },
+    {
+      "epoch": 0.6063551856127347,
+      "grad_norm": 0.12333488464355469,
+      "learning_rate": 3.3790046701110596e-05,
+      "loss": 1.1233,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6064159426453613,
+      "grad_norm": 0.23280581831932068,
+      "learning_rate": 3.378099126139358e-05,
+      "loss": 1.0311,
+      "step": 9981
+    },
+    {
+      "epoch": 0.6064766996779877,
+      "grad_norm": 0.17630217969417572,
+      "learning_rate": 3.3771936416187456e-05,
+      "loss": 1.1104,
+      "step": 9982
+    },
+    {
+      "epoch": 0.6065374567106142,
+      "grad_norm": 0.15698201954364777,
+      "learning_rate": 3.376288216582413e-05,
+      "loss": 1.077,
+      "step": 9983
+    },
+    {
+      "epoch": 0.6065982137432407,
+      "grad_norm": 0.13801109790802002,
+      "learning_rate": 3.375382851063551e-05,
+      "loss": 1.052,
+      "step": 9984
+    },
+    {
+      "epoch": 0.6066589707758673,
+      "grad_norm": 0.12702766060829163,
+      "learning_rate": 3.374477545095344e-05,
+      "loss": 1.0186,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6067197278084938,
+      "grad_norm": 0.21862299740314484,
+      "learning_rate": 3.3735722987109764e-05,
+      "loss": 1.0767,
+      "step": 9986
+    },
+    {
+      "epoch": 0.6067804848411203,
+      "grad_norm": 2.080496311187744,
+      "learning_rate": 3.3726671119436325e-05,
+      "loss": 1.074,
+      "step": 9987
+    },
+    {
+      "epoch": 0.6068412418737469,
+      "grad_norm": 0.14575068652629852,
+      "learning_rate": 3.371761984826488e-05,
+      "loss": 1.0791,
+      "step": 9988
+    },
+    {
+      "epoch": 0.6069019989063734,
+      "grad_norm": 0.4230603277683258,
+      "learning_rate": 3.370856917392724e-05,
+      "loss": 1.1887,
+      "step": 9989
+    },
+    {
+      "epoch": 0.606962755939,
+      "grad_norm": 0.11740444600582123,
+      "learning_rate": 3.369951909675514e-05,
+      "loss": 1.024,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6070235129716265,
+      "grad_norm": 0.14493657648563385,
+      "learning_rate": 3.3690469617080315e-05,
+      "loss": 1.084,
+      "step": 9991
+    },
+    {
+      "epoch": 0.607084270004253,
+      "grad_norm": 4.207291603088379,
+      "learning_rate": 3.368142073523448e-05,
+      "loss": 1.0914,
+      "step": 9992
+    },
+    {
+      "epoch": 0.6071450270368796,
+      "grad_norm": 0.20452556014060974,
+      "learning_rate": 3.367237245154933e-05,
+      "loss": 1.0837,
+      "step": 9993
+    },
+    {
+      "epoch": 0.6072057840695061,
+      "grad_norm": 0.22807788848876953,
+      "learning_rate": 3.366332476635653e-05,
+      "loss": 1.0819,
+      "step": 9994
+    },
+    {
+      "epoch": 0.6072665411021325,
+      "grad_norm": 0.15235096216201782,
+      "learning_rate": 3.36542776799877e-05,
+      "loss": 1.0217,
+      "step": 9995
+    },
+    {
+      "epoch": 0.607327298134759,
+      "grad_norm": 0.13611027598381042,
+      "learning_rate": 3.3645231192774486e-05,
+      "loss": 1.0813,
+      "step": 9996
+    },
+    {
+      "epoch": 0.6073880551673856,
+      "grad_norm": 0.14824730157852173,
+      "learning_rate": 3.363618530504848e-05,
+      "loss": 1.0464,
+      "step": 9997
+    },
+    {
+      "epoch": 0.6074488122000121,
+      "grad_norm": 0.23028042912483215,
+      "learning_rate": 3.362714001714128e-05,
+      "loss": 1.1059,
+      "step": 9998
+    },
+    {
+      "epoch": 0.6075095692326387,
+      "grad_norm": 0.1377759426832199,
+      "learning_rate": 3.361809532938441e-05,
+      "loss": 1.0555,
+      "step": 9999
+    },
+    {
+      "epoch": 0.6075703262652652,
+      "grad_norm": 0.31320837140083313,
+      "learning_rate": 3.360905124210943e-05,
+      "loss": 1.0533,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6076310832978917,
+      "grad_norm": 0.1362789273262024,
+      "learning_rate": 3.360000775564785e-05,
+      "loss": 0.9948,
+      "step": 10001
+    },
+    {
+      "epoch": 0.6076918403305183,
+      "grad_norm": 0.13190710544586182,
+      "learning_rate": 3.3590964870331157e-05,
+      "loss": 1.0373,
+      "step": 10002
+    },
+    {
+      "epoch": 0.6077525973631448,
+      "grad_norm": 0.14964397251605988,
+      "learning_rate": 3.3581922586490815e-05,
+      "loss": 1.0544,
+      "step": 10003
+    },
+    {
+      "epoch": 0.6078133543957713,
+      "grad_norm": 0.21665766835212708,
+      "learning_rate": 3.357288090445827e-05,
+      "loss": 1.157,
+      "step": 10004
+    },
+    {
+      "epoch": 0.6078741114283979,
+      "grad_norm": 0.15179190039634705,
+      "learning_rate": 3.356383982456495e-05,
+      "loss": 1.0442,
+      "step": 10005
+    },
+    {
+      "epoch": 0.6079348684610244,
+      "grad_norm": 0.1554936021566391,
+      "learning_rate": 3.355479934714226e-05,
+      "loss": 1.0531,
+      "step": 10006
+    },
+    {
+      "epoch": 0.6079956254936509,
+      "grad_norm": 0.16013725101947784,
+      "learning_rate": 3.354575947252159e-05,
+      "loss": 1.034,
+      "step": 10007
+    },
+    {
+      "epoch": 0.6080563825262774,
+      "grad_norm": 0.16513048112392426,
+      "learning_rate": 3.353672020103428e-05,
+      "loss": 1.0643,
+      "step": 10008
+    },
+    {
+      "epoch": 0.6081171395589039,
+      "grad_norm": 0.1938750445842743,
+      "learning_rate": 3.352768153301166e-05,
+      "loss": 1.1527,
+      "step": 10009
+    },
+    {
+      "epoch": 0.6081778965915304,
+      "grad_norm": 0.1725156307220459,
+      "learning_rate": 3.351864346878507e-05,
+      "loss": 1.1187,
+      "step": 10010
+    },
+    {
+      "epoch": 0.608238653624157,
+      "grad_norm": 0.1336262822151184,
+      "learning_rate": 3.350960600868579e-05,
+      "loss": 1.0731,
+      "step": 10011
+    },
+    {
+      "epoch": 0.6082994106567835,
+      "grad_norm": 0.38251498341560364,
+      "learning_rate": 3.350056915304508e-05,
+      "loss": 1.0273,
+      "step": 10012
+    },
+    {
+      "epoch": 0.60836016768941,
+      "grad_norm": 0.14297010004520416,
+      "learning_rate": 3.34915329021942e-05,
+      "loss": 1.0649,
+      "step": 10013
+    },
+    {
+      "epoch": 0.6084209247220366,
+      "grad_norm": 0.12223240733146667,
+      "learning_rate": 3.348249725646439e-05,
+      "loss": 1.0048,
+      "step": 10014
+    },
+    {
+      "epoch": 0.6084816817546631,
+      "grad_norm": 0.123337522149086,
+      "learning_rate": 3.347346221618682e-05,
+      "loss": 1.1941,
+      "step": 10015
+    },
+    {
+      "epoch": 0.6085424387872896,
+      "grad_norm": 0.12456902861595154,
+      "learning_rate": 3.3464427781692695e-05,
+      "loss": 1.0506,
+      "step": 10016
+    },
+    {
+      "epoch": 0.6086031958199162,
+      "grad_norm": 0.5937848091125488,
+      "learning_rate": 3.345539395331315e-05,
+      "loss": 1.165,
+      "step": 10017
+    },
+    {
+      "epoch": 0.6086639528525427,
+      "grad_norm": 0.15157192945480347,
+      "learning_rate": 3.344636073137934e-05,
+      "loss": 1.0648,
+      "step": 10018
+    },
+    {
+      "epoch": 0.6087247098851692,
+      "grad_norm": 0.2541963458061218,
+      "learning_rate": 3.343732811622239e-05,
+      "loss": 1.0739,
+      "step": 10019
+    },
+    {
+      "epoch": 0.6087854669177958,
+      "grad_norm": 0.11770328879356384,
+      "learning_rate": 3.342829610817337e-05,
+      "loss": 1.0604,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6088462239504223,
+      "grad_norm": 0.19118331372737885,
+      "learning_rate": 3.341926470756337e-05,
+      "loss": 1.1141,
+      "step": 10021
+    },
+    {
+      "epoch": 0.6089069809830487,
+      "grad_norm": 0.14297187328338623,
+      "learning_rate": 3.341023391472342e-05,
+      "loss": 1.0649,
+      "step": 10022
+    },
+    {
+      "epoch": 0.6089677380156753,
+      "grad_norm": 0.15951859951019287,
+      "learning_rate": 3.340120372998455e-05,
+      "loss": 1.0616,
+      "step": 10023
+    },
+    {
+      "epoch": 0.6090284950483018,
+      "grad_norm": 0.11807980388402939,
+      "learning_rate": 3.339217415367778e-05,
+      "loss": 1.0619,
+      "step": 10024
+    },
+    {
+      "epoch": 0.6090892520809283,
+      "grad_norm": 0.2047649472951889,
+      "learning_rate": 3.338314518613406e-05,
+      "loss": 1.2266,
+      "step": 10025
+    },
+    {
+      "epoch": 0.6091500091135549,
+      "grad_norm": 0.11830107122659683,
+      "learning_rate": 3.337411682768436e-05,
+      "loss": 1.0404,
+      "step": 10026
+    },
+    {
+      "epoch": 0.6092107661461814,
+      "grad_norm": 0.12327075004577637,
+      "learning_rate": 3.336508907865964e-05,
+      "loss": 1.0397,
+      "step": 10027
+    },
+    {
+      "epoch": 0.609271523178808,
+      "grad_norm": 0.12254289537668228,
+      "learning_rate": 3.3356061939390795e-05,
+      "loss": 1.0602,
+      "step": 10028
+    },
+    {
+      "epoch": 0.6093322802114345,
+      "grad_norm": 0.16362264752388,
+      "learning_rate": 3.3347035410208705e-05,
+      "loss": 1.2008,
+      "step": 10029
+    },
+    {
+      "epoch": 0.609393037244061,
+      "grad_norm": 0.23142939805984497,
+      "learning_rate": 3.333800949144426e-05,
+      "loss": 1.019,
+      "step": 10030
+    },
+    {
+      "epoch": 0.6094537942766876,
+      "grad_norm": 0.118656225502491,
+      "learning_rate": 3.332898418342829e-05,
+      "loss": 1.0602,
+      "step": 10031
+    },
+    {
+      "epoch": 0.6095145513093141,
+      "grad_norm": 0.15101879835128784,
+      "learning_rate": 3.331995948649161e-05,
+      "loss": 1.0141,
+      "step": 10032
+    },
+    {
+      "epoch": 0.6095753083419406,
+      "grad_norm": 0.2564082443714142,
+      "learning_rate": 3.331093540096507e-05,
+      "loss": 1.08,
+      "step": 10033
+    },
+    {
+      "epoch": 0.6096360653745672,
+      "grad_norm": 0.23222987353801727,
+      "learning_rate": 3.330191192717942e-05,
+      "loss": 1.191,
+      "step": 10034
+    },
+    {
+      "epoch": 0.6096968224071936,
+      "grad_norm": 0.3385956287384033,
+      "learning_rate": 3.3292889065465405e-05,
+      "loss": 1.2905,
+      "step": 10035
+    },
+    {
+      "epoch": 0.6097575794398201,
+      "grad_norm": 0.1074041798710823,
+      "learning_rate": 3.328386681615377e-05,
+      "loss": 1.0584,
+      "step": 10036
+    },
+    {
+      "epoch": 0.6098183364724467,
+      "grad_norm": 0.12528693675994873,
+      "learning_rate": 3.327484517957522e-05,
+      "loss": 1.0424,
+      "step": 10037
+    },
+    {
+      "epoch": 0.6098790935050732,
+      "grad_norm": 0.18149995803833008,
+      "learning_rate": 3.326582415606046e-05,
+      "loss": 1.031,
+      "step": 10038
+    },
+    {
+      "epoch": 0.6099398505376997,
+      "grad_norm": 0.13242071866989136,
+      "learning_rate": 3.325680374594015e-05,
+      "loss": 1.2254,
+      "step": 10039
+    },
+    {
+      "epoch": 0.6100006075703263,
+      "grad_norm": 0.175632044672966,
+      "learning_rate": 3.324778394954493e-05,
+      "loss": 1.0641,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6100613646029528,
+      "grad_norm": 0.18444722890853882,
+      "learning_rate": 3.323876476720546e-05,
+      "loss": 1.113,
+      "step": 10041
+    },
+    {
+      "epoch": 0.6101221216355793,
+      "grad_norm": 0.15609140694141388,
+      "learning_rate": 3.322974619925227e-05,
+      "loss": 1.0359,
+      "step": 10042
+    },
+    {
+      "epoch": 0.6101828786682059,
+      "grad_norm": 0.1381586790084839,
+      "learning_rate": 3.322072824601598e-05,
+      "loss": 1.0405,
+      "step": 10043
+    },
+    {
+      "epoch": 0.6102436357008324,
+      "grad_norm": 0.4765521287918091,
+      "learning_rate": 3.321171090782715e-05,
+      "loss": 1.2477,
+      "step": 10044
+    },
+    {
+      "epoch": 0.6103043927334589,
+      "grad_norm": 0.18282434344291687,
+      "learning_rate": 3.320269418501628e-05,
+      "loss": 1.0574,
+      "step": 10045
+    },
+    {
+      "epoch": 0.6103651497660855,
+      "grad_norm": 0.19166181981563568,
+      "learning_rate": 3.3193678077913926e-05,
+      "loss": 1.084,
+      "step": 10046
+    },
+    {
+      "epoch": 0.610425906798712,
+      "grad_norm": 0.13756051659584045,
+      "learning_rate": 3.3184662586850545e-05,
+      "loss": 1.0684,
+      "step": 10047
+    },
+    {
+      "epoch": 0.6104866638313384,
+      "grad_norm": 0.2356358766555786,
+      "learning_rate": 3.317564771215661e-05,
+      "loss": 1.0379,
+      "step": 10048
+    },
+    {
+      "epoch": 0.610547420863965,
+      "grad_norm": 0.15451419353485107,
+      "learning_rate": 3.316663345416256e-05,
+      "loss": 1.1197,
+      "step": 10049
+    },
+    {
+      "epoch": 0.6106081778965915,
+      "grad_norm": 3.828979730606079,
+      "learning_rate": 3.315761981319881e-05,
+      "loss": 1.1231,
+      "step": 10050
+    },
+    {
+      "epoch": 0.610668934929218,
+      "grad_norm": 0.2020713984966278,
+      "learning_rate": 3.314860678959575e-05,
+      "loss": 1.036,
+      "step": 10051
+    },
+    {
+      "epoch": 0.6107296919618446,
+      "grad_norm": 0.1029885783791542,
+      "learning_rate": 3.313959438368378e-05,
+      "loss": 1.0235,
+      "step": 10052
+    },
+    {
+      "epoch": 0.6107904489944711,
+      "grad_norm": 0.15715427696704865,
+      "learning_rate": 3.3130582595793226e-05,
+      "loss": 1.1494,
+      "step": 10053
+    },
+    {
+      "epoch": 0.6108512060270976,
+      "grad_norm": 0.1653818041086197,
+      "learning_rate": 3.3121571426254456e-05,
+      "loss": 1.0726,
+      "step": 10054
+    },
+    {
+      "epoch": 0.6109119630597242,
+      "grad_norm": 0.33791765570640564,
+      "learning_rate": 3.311256087539771e-05,
+      "loss": 1.1035,
+      "step": 10055
+    },
+    {
+      "epoch": 0.6109727200923507,
+      "grad_norm": 0.2315925508737564,
+      "learning_rate": 3.310355094355333e-05,
+      "loss": 1.1706,
+      "step": 10056
+    },
+    {
+      "epoch": 0.6110334771249772,
+      "grad_norm": 0.24786247313022614,
+      "learning_rate": 3.309454163105156e-05,
+      "loss": 1.0292,
+      "step": 10057
+    },
+    {
+      "epoch": 0.6110942341576038,
+      "grad_norm": 0.48100122809410095,
+      "learning_rate": 3.308553293822262e-05,
+      "loss": 1.1718,
+      "step": 10058
+    },
+    {
+      "epoch": 0.6111549911902303,
+      "grad_norm": 0.1851678490638733,
+      "learning_rate": 3.307652486539674e-05,
+      "loss": 1.0945,
+      "step": 10059
+    },
+    {
+      "epoch": 0.6112157482228568,
+      "grad_norm": 0.2552066445350647,
+      "learning_rate": 3.306751741290413e-05,
+      "loss": 1.1719,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6112765052554833,
+      "grad_norm": 0.204861581325531,
+      "learning_rate": 3.305851058107494e-05,
+      "loss": 1.1591,
+      "step": 10061
+    },
+    {
+      "epoch": 0.6113372622881098,
+      "grad_norm": 0.13126778602600098,
+      "learning_rate": 3.30495043702393e-05,
+      "loss": 1.0683,
+      "step": 10062
+    },
+    {
+      "epoch": 0.6113980193207363,
+      "grad_norm": 0.21355962753295898,
+      "learning_rate": 3.3040498780727355e-05,
+      "loss": 1.0148,
+      "step": 10063
+    },
+    {
+      "epoch": 0.6114587763533629,
+      "grad_norm": 0.1705038696527481,
+      "learning_rate": 3.303149381286922e-05,
+      "loss": 1.0905,
+      "step": 10064
+    },
+    {
+      "epoch": 0.6115195333859894,
+      "grad_norm": 0.14262165129184723,
+      "learning_rate": 3.3022489466994954e-05,
+      "loss": 1.052,
+      "step": 10065
+    },
+    {
+      "epoch": 0.611580290418616,
+      "grad_norm": 0.26569920778274536,
+      "learning_rate": 3.3013485743434615e-05,
+      "loss": 1.2143,
+      "step": 10066
+    },
+    {
+      "epoch": 0.6116410474512425,
+      "grad_norm": 0.17504224181175232,
+      "learning_rate": 3.300448264251824e-05,
+      "loss": 1.0714,
+      "step": 10067
+    },
+    {
+      "epoch": 0.611701804483869,
+      "grad_norm": 0.31215140223503113,
+      "learning_rate": 3.299548016457586e-05,
+      "loss": 1.0429,
+      "step": 10068
+    },
+    {
+      "epoch": 0.6117625615164956,
+      "grad_norm": 0.11999605596065521,
+      "learning_rate": 3.298647830993742e-05,
+      "loss": 0.9838,
+      "step": 10069
+    },
+    {
+      "epoch": 0.6118233185491221,
+      "grad_norm": 0.22586029767990112,
+      "learning_rate": 3.2977477078932904e-05,
+      "loss": 1.0422,
+      "step": 10070
+    },
+    {
+      "epoch": 0.6118840755817486,
+      "grad_norm": 0.1299959421157837,
+      "learning_rate": 3.296847647189225e-05,
+      "loss": 1.0724,
+      "step": 10071
+    },
+    {
+      "epoch": 0.6119448326143752,
+      "grad_norm": 0.13999871909618378,
+      "learning_rate": 3.29594764891454e-05,
+      "loss": 1.0004,
+      "step": 10072
+    },
+    {
+      "epoch": 0.6120055896470017,
+      "grad_norm": 0.13911172747612,
+      "learning_rate": 3.295047713102223e-05,
+      "loss": 1.0468,
+      "step": 10073
+    },
+    {
+      "epoch": 0.6120663466796281,
+      "grad_norm": 0.1452186107635498,
+      "learning_rate": 3.294147839785262e-05,
+      "loss": 1.0418,
+      "step": 10074
+    },
+    {
+      "epoch": 0.6121271037122546,
+      "grad_norm": 0.16490453481674194,
+      "learning_rate": 3.293248028996642e-05,
+      "loss": 1.0837,
+      "step": 10075
+    },
+    {
+      "epoch": 0.6121878607448812,
+      "grad_norm": 0.12922118604183197,
+      "learning_rate": 3.292348280769343e-05,
+      "loss": 0.9894,
+      "step": 10076
+    },
+    {
+      "epoch": 0.6122486177775077,
+      "grad_norm": 0.13093970715999603,
+      "learning_rate": 3.29144859513635e-05,
+      "loss": 1.0333,
+      "step": 10077
+    },
+    {
+      "epoch": 0.6123093748101343,
+      "grad_norm": 0.15148712694644928,
+      "learning_rate": 3.2905489721306385e-05,
+      "loss": 1.0701,
+      "step": 10078
+    },
+    {
+      "epoch": 0.6123701318427608,
+      "grad_norm": 0.25300851464271545,
+      "learning_rate": 3.289649411785183e-05,
+      "loss": 1.1538,
+      "step": 10079
+    },
+    {
+      "epoch": 0.6124308888753873,
+      "grad_norm": 0.13811343908309937,
+      "learning_rate": 3.2887499141329606e-05,
+      "loss": 1.0605,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6124916459080139,
+      "grad_norm": 0.20048639178276062,
+      "learning_rate": 3.287850479206942e-05,
+      "loss": 1.0471,
+      "step": 10081
+    },
+    {
+      "epoch": 0.6125524029406404,
+      "grad_norm": 0.22141355276107788,
+      "learning_rate": 3.286951107040094e-05,
+      "loss": 1.0536,
+      "step": 10082
+    },
+    {
+      "epoch": 0.6126131599732669,
+      "grad_norm": 0.42613619565963745,
+      "learning_rate": 3.2860517976653845e-05,
+      "loss": 0.9982,
+      "step": 10083
+    },
+    {
+      "epoch": 0.6126739170058935,
+      "grad_norm": 0.5643616318702698,
+      "learning_rate": 3.2851525511157764e-05,
+      "loss": 1.0455,
+      "step": 10084
+    },
+    {
+      "epoch": 0.61273467403852,
+      "grad_norm": 0.14203448593616486,
+      "learning_rate": 3.284253367424234e-05,
+      "loss": 1.0232,
+      "step": 10085
+    },
+    {
+      "epoch": 0.6127954310711465,
+      "grad_norm": 0.23522542417049408,
+      "learning_rate": 3.283354246623716e-05,
+      "loss": 1.035,
+      "step": 10086
+    },
+    {
+      "epoch": 0.612856188103773,
+      "grad_norm": 0.1335030496120453,
+      "learning_rate": 3.282455188747179e-05,
+      "loss": 1.0774,
+      "step": 10087
+    },
+    {
+      "epoch": 0.6129169451363995,
+      "grad_norm": 0.1179598867893219,
+      "learning_rate": 3.2815561938275815e-05,
+      "loss": 1.0189,
+      "step": 10088
+    },
+    {
+      "epoch": 0.612977702169026,
+      "grad_norm": 0.2610608637332916,
+      "learning_rate": 3.280657261897872e-05,
+      "loss": 1.1764,
+      "step": 10089
+    },
+    {
+      "epoch": 0.6130384592016526,
+      "grad_norm": 1.9669466018676758,
+      "learning_rate": 3.279758392991004e-05,
+      "loss": 1.0661,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6130992162342791,
+      "grad_norm": 0.4825950860977173,
+      "learning_rate": 3.278859587139923e-05,
+      "loss": 1.02,
+      "step": 10091
+    },
+    {
+      "epoch": 0.6131599732669056,
+      "grad_norm": 0.4671441316604614,
+      "learning_rate": 3.277960844377577e-05,
+      "loss": 1.2157,
+      "step": 10092
+    },
+    {
+      "epoch": 0.6132207302995322,
+      "grad_norm": 0.11444177478551865,
+      "learning_rate": 3.277062164736909e-05,
+      "loss": 1.0302,
+      "step": 10093
+    },
+    {
+      "epoch": 0.6132814873321587,
+      "grad_norm": 0.16415248811244965,
+      "learning_rate": 3.276163548250862e-05,
+      "loss": 1.0196,
+      "step": 10094
+    },
+    {
+      "epoch": 0.6133422443647852,
+      "grad_norm": 0.30551454424858093,
+      "learning_rate": 3.275264994952373e-05,
+      "loss": 1.1052,
+      "step": 10095
+    },
+    {
+      "epoch": 0.6134030013974118,
+      "grad_norm": 0.15775832533836365,
+      "learning_rate": 3.2743665048743785e-05,
+      "loss": 1.1031,
+      "step": 10096
+    },
+    {
+      "epoch": 0.6134637584300383,
+      "grad_norm": 0.3249801695346832,
+      "learning_rate": 3.273468078049812e-05,
+      "loss": 1.0651,
+      "step": 10097
+    },
+    {
+      "epoch": 0.6135245154626648,
+      "grad_norm": 0.12081403285264969,
+      "learning_rate": 3.272569714511607e-05,
+      "loss": 1.0351,
+      "step": 10098
+    },
+    {
+      "epoch": 0.6135852724952914,
+      "grad_norm": 0.13086235523223877,
+      "learning_rate": 3.2716714142926953e-05,
+      "loss": 1.0114,
+      "step": 10099
+    },
+    {
+      "epoch": 0.6136460295279178,
+      "grad_norm": 0.17534758150577545,
+      "learning_rate": 3.270773177425999e-05,
+      "loss": 1.1343,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6137067865605443,
+      "grad_norm": 0.15806272625923157,
+      "learning_rate": 3.2698750039444504e-05,
+      "loss": 1.0546,
+      "step": 10101
+    },
+    {
+      "epoch": 0.6137675435931709,
+      "grad_norm": 0.12268459796905518,
+      "learning_rate": 3.268976893880964e-05,
+      "loss": 1.0859,
+      "step": 10102
+    },
+    {
+      "epoch": 0.6138283006257974,
+      "grad_norm": 0.2153148502111435,
+      "learning_rate": 3.268078847268465e-05,
+      "loss": 1.1813,
+      "step": 10103
+    },
+    {
+      "epoch": 0.6138890576584239,
+      "grad_norm": 0.1714666187763214,
+      "learning_rate": 3.2671808641398714e-05,
+      "loss": 1.0427,
+      "step": 10104
+    },
+    {
+      "epoch": 0.6139498146910505,
+      "grad_norm": 0.231259286403656,
+      "learning_rate": 3.266282944528097e-05,
+      "loss": 1.0358,
+      "step": 10105
+    },
+    {
+      "epoch": 0.614010571723677,
+      "grad_norm": 0.1366434097290039,
+      "learning_rate": 3.2653850884660576e-05,
+      "loss": 1.079,
+      "step": 10106
+    },
+    {
+      "epoch": 0.6140713287563035,
+      "grad_norm": 0.3615018129348755,
+      "learning_rate": 3.2644872959866626e-05,
+      "loss": 1.2342,
+      "step": 10107
+    },
+    {
+      "epoch": 0.6141320857889301,
+      "grad_norm": 0.11281843483448029,
+      "learning_rate": 3.263589567122823e-05,
+      "loss": 1.0384,
+      "step": 10108
+    },
+    {
+      "epoch": 0.6141928428215566,
+      "grad_norm": 0.25387972593307495,
+      "learning_rate": 3.262691901907442e-05,
+      "loss": 1.2153,
+      "step": 10109
+    },
+    {
+      "epoch": 0.6142535998541832,
+      "grad_norm": 0.1566266417503357,
+      "learning_rate": 3.261794300373424e-05,
+      "loss": 1.0753,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6143143568868097,
+      "grad_norm": 0.15186652541160583,
+      "learning_rate": 3.260896762553673e-05,
+      "loss": 1.1173,
+      "step": 10111
+    },
+    {
+      "epoch": 0.6143751139194362,
+      "grad_norm": 0.384333074092865,
+      "learning_rate": 3.259999288481087e-05,
+      "loss": 1.2207,
+      "step": 10112
+    },
+    {
+      "epoch": 0.6144358709520626,
+      "grad_norm": 0.18710604310035706,
+      "learning_rate": 3.259101878188563e-05,
+      "loss": 1.04,
+      "step": 10113
+    },
+    {
+      "epoch": 0.6144966279846892,
+      "grad_norm": 0.11207471042871475,
+      "learning_rate": 3.2582045317089966e-05,
+      "loss": 1.0117,
+      "step": 10114
+    },
+    {
+      "epoch": 0.6145573850173157,
+      "grad_norm": 0.3282966911792755,
+      "learning_rate": 3.257307249075281e-05,
+      "loss": 1.2205,
+      "step": 10115
+    },
+    {
+      "epoch": 0.6146181420499423,
+      "grad_norm": 0.20154400169849396,
+      "learning_rate": 3.2564100303203035e-05,
+      "loss": 1.0794,
+      "step": 10116
+    },
+    {
+      "epoch": 0.6146788990825688,
+      "grad_norm": 0.14722199738025665,
+      "learning_rate": 3.2555128754769544e-05,
+      "loss": 1.0998,
+      "step": 10117
+    },
+    {
+      "epoch": 0.6147396561151953,
+      "grad_norm": 0.12835252285003662,
+      "learning_rate": 3.254615784578117e-05,
+      "loss": 1.0498,
+      "step": 10118
+    },
+    {
+      "epoch": 0.6148004131478219,
+      "grad_norm": 0.15189583599567413,
+      "learning_rate": 3.253718757656675e-05,
+      "loss": 1.0448,
+      "step": 10119
+    },
+    {
+      "epoch": 0.6148611701804484,
+      "grad_norm": 0.16218426823616028,
+      "learning_rate": 3.252821794745511e-05,
+      "loss": 1.101,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6149219272130749,
+      "grad_norm": 0.13713757693767548,
+      "learning_rate": 3.251924895877502e-05,
+      "loss": 1.0704,
+      "step": 10121
+    },
+    {
+      "epoch": 0.6149826842457015,
+      "grad_norm": 0.17553897202014923,
+      "learning_rate": 3.251028061085522e-05,
+      "loss": 1.1297,
+      "step": 10122
+    },
+    {
+      "epoch": 0.615043441278328,
+      "grad_norm": 0.14513979852199554,
+      "learning_rate": 3.250131290402448e-05,
+      "loss": 1.0314,
+      "step": 10123
+    },
+    {
+      "epoch": 0.6151041983109545,
+      "grad_norm": 0.2382880449295044,
+      "learning_rate": 3.249234583861149e-05,
+      "loss": 1.0434,
+      "step": 10124
+    },
+    {
+      "epoch": 0.6151649553435811,
+      "grad_norm": 0.1594681739807129,
+      "learning_rate": 3.248337941494496e-05,
+      "loss": 1.0743,
+      "step": 10125
+    },
+    {
+      "epoch": 0.6152257123762076,
+      "grad_norm": 0.14009036123752594,
+      "learning_rate": 3.2474413633353515e-05,
+      "loss": 1.0436,
+      "step": 10126
+    },
+    {
+      "epoch": 0.615286469408834,
+      "grad_norm": 0.40217599272727966,
+      "learning_rate": 3.246544849416585e-05,
+      "loss": 1.0653,
+      "step": 10127
+    },
+    {
+      "epoch": 0.6153472264414606,
+      "grad_norm": 0.7947143912315369,
+      "learning_rate": 3.245648399771057e-05,
+      "loss": 1.0742,
+      "step": 10128
+    },
+    {
+      "epoch": 0.6154079834740871,
+      "grad_norm": 0.142560213804245,
+      "learning_rate": 3.2447520144316246e-05,
+      "loss": 1.1069,
+      "step": 10129
+    },
+    {
+      "epoch": 0.6154687405067136,
+      "grad_norm": 0.11180705577135086,
+      "learning_rate": 3.243855693431146e-05,
+      "loss": 1.0368,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6155294975393402,
+      "grad_norm": 0.274832159280777,
+      "learning_rate": 3.2429594368024766e-05,
+      "loss": 1.0994,
+      "step": 10131
+    },
+    {
+      "epoch": 0.6155902545719667,
+      "grad_norm": 0.12490338087081909,
+      "learning_rate": 3.2420632445784685e-05,
+      "loss": 1.0175,
+      "step": 10132
+    },
+    {
+      "epoch": 0.6156510116045932,
+      "grad_norm": 0.17404744029045105,
+      "learning_rate": 3.241167116791971e-05,
+      "loss": 1.1929,
+      "step": 10133
+    },
+    {
+      "epoch": 0.6157117686372198,
+      "grad_norm": 0.16612984240055084,
+      "learning_rate": 3.240271053475834e-05,
+      "loss": 1.0293,
+      "step": 10134
+    },
+    {
+      "epoch": 0.6157725256698463,
+      "grad_norm": 0.16274787485599518,
+      "learning_rate": 3.239375054662903e-05,
+      "loss": 1.06,
+      "step": 10135
+    },
+    {
+      "epoch": 0.6158332827024728,
+      "grad_norm": 0.16788017749786377,
+      "learning_rate": 3.238479120386019e-05,
+      "loss": 1.0531,
+      "step": 10136
+    },
+    {
+      "epoch": 0.6158940397350994,
+      "grad_norm": 0.15702417492866516,
+      "learning_rate": 3.237583250678022e-05,
+      "loss": 1.037,
+      "step": 10137
+    },
+    {
+      "epoch": 0.6159547967677259,
+      "grad_norm": 0.18851259350776672,
+      "learning_rate": 3.236687445571752e-05,
+      "loss": 1.0325,
+      "step": 10138
+    },
+    {
+      "epoch": 0.6160155538003524,
+      "grad_norm": 0.21135739982128143,
+      "learning_rate": 3.2357917051000454e-05,
+      "loss": 1.3347,
+      "step": 10139
+    },
+    {
+      "epoch": 0.6160763108329789,
+      "grad_norm": 0.23360225558280945,
+      "learning_rate": 3.2348960292957345e-05,
+      "loss": 1.1451,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6161370678656054,
+      "grad_norm": 0.11978757381439209,
+      "learning_rate": 3.234000418191652e-05,
+      "loss": 1.0201,
+      "step": 10141
+    },
+    {
+      "epoch": 0.6161978248982319,
+      "grad_norm": 0.8820341229438782,
+      "learning_rate": 3.233104871820623e-05,
+      "loss": 1.2746,
+      "step": 10142
+    },
+    {
+      "epoch": 0.6162585819308585,
+      "grad_norm": 0.1710776388645172,
+      "learning_rate": 3.2322093902154776e-05,
+      "loss": 1.0617,
+      "step": 10143
+    },
+    {
+      "epoch": 0.616319338963485,
+      "grad_norm": 0.1668907105922699,
+      "learning_rate": 3.2313139734090395e-05,
+      "loss": 1.0229,
+      "step": 10144
+    },
+    {
+      "epoch": 0.6163800959961115,
+      "grad_norm": 17.193498611450195,
+      "learning_rate": 3.2304186214341284e-05,
+      "loss": 1.2036,
+      "step": 10145
+    },
+    {
+      "epoch": 0.6164408530287381,
+      "grad_norm": 0.16899652779102325,
+      "learning_rate": 3.2295233343235645e-05,
+      "loss": 1.1058,
+      "step": 10146
+    },
+    {
+      "epoch": 0.6165016100613646,
+      "grad_norm": 0.11913230270147324,
+      "learning_rate": 3.228628112110166e-05,
+      "loss": 1.0108,
+      "step": 10147
+    },
+    {
+      "epoch": 0.6165623670939911,
+      "grad_norm": 0.15516933798789978,
+      "learning_rate": 3.227732954826747e-05,
+      "loss": 1.0876,
+      "step": 10148
+    },
+    {
+      "epoch": 0.6166231241266177,
+      "grad_norm": 0.23348292708396912,
+      "learning_rate": 3.2268378625061185e-05,
+      "loss": 1.1455,
+      "step": 10149
+    },
+    {
+      "epoch": 0.6166838811592442,
+      "grad_norm": 0.39024418592453003,
+      "learning_rate": 3.225942835181091e-05,
+      "loss": 1.2577,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6167446381918708,
+      "grad_norm": 0.14650747179985046,
+      "learning_rate": 3.2250478728844704e-05,
+      "loss": 1.0989,
+      "step": 10151
+    },
+    {
+      "epoch": 0.6168053952244973,
+      "grad_norm": 0.17481721937656403,
+      "learning_rate": 3.224152975649064e-05,
+      "loss": 1.022,
+      "step": 10152
+    },
+    {
+      "epoch": 0.6168661522571237,
+      "grad_norm": 0.14200538396835327,
+      "learning_rate": 3.223258143507674e-05,
+      "loss": 1.0547,
+      "step": 10153
+    },
+    {
+      "epoch": 0.6169269092897502,
+      "grad_norm": 0.1690562665462494,
+      "learning_rate": 3.2223633764931e-05,
+      "loss": 1.1528,
+      "step": 10154
+    },
+    {
+      "epoch": 0.6169876663223768,
+      "grad_norm": 0.14175823330879211,
+      "learning_rate": 3.2214686746381415e-05,
+      "loss": 1.04,
+      "step": 10155
+    },
+    {
+      "epoch": 0.6170484233550033,
+      "grad_norm": 0.16546085476875305,
+      "learning_rate": 3.22057403797559e-05,
+      "loss": 1.1701,
+      "step": 10156
+    },
+    {
+      "epoch": 0.6171091803876299,
+      "grad_norm": 0.12204727530479431,
+      "learning_rate": 3.219679466538242e-05,
+      "loss": 1.0554,
+      "step": 10157
+    },
+    {
+      "epoch": 0.6171699374202564,
+      "grad_norm": 0.19949981570243835,
+      "learning_rate": 3.218784960358888e-05,
+      "loss": 1.019,
+      "step": 10158
+    },
+    {
+      "epoch": 0.6172306944528829,
+      "grad_norm": 0.2087552398443222,
+      "learning_rate": 3.217890519470313e-05,
+      "loss": 1.0144,
+      "step": 10159
+    },
+    {
+      "epoch": 0.6172914514855095,
+      "grad_norm": 0.19289065897464752,
+      "learning_rate": 3.216996143905308e-05,
+      "loss": 1.0775,
+      "step": 10160
+    },
+    {
+      "epoch": 0.617352208518136,
+      "grad_norm": 0.14146707952022552,
+      "learning_rate": 3.216101833696655e-05,
+      "loss": 1.0725,
+      "step": 10161
+    },
+    {
+      "epoch": 0.6174129655507625,
+      "grad_norm": 0.10160516202449799,
+      "learning_rate": 3.215207588877133e-05,
+      "loss": 1.0466,
+      "step": 10162
+    },
+    {
+      "epoch": 0.6174737225833891,
+      "grad_norm": 0.16701847314834595,
+      "learning_rate": 3.214313409479521e-05,
+      "loss": 1.0712,
+      "step": 10163
+    },
+    {
+      "epoch": 0.6175344796160156,
+      "grad_norm": 0.14244811236858368,
+      "learning_rate": 3.2134192955365954e-05,
+      "loss": 1.0711,
+      "step": 10164
+    },
+    {
+      "epoch": 0.6175952366486421,
+      "grad_norm": 0.11480254679918289,
+      "learning_rate": 3.2125252470811315e-05,
+      "loss": 1.0228,
+      "step": 10165
+    },
+    {
+      "epoch": 0.6176559936812686,
+      "grad_norm": 0.10029196739196777,
+      "learning_rate": 3.211631264145901e-05,
+      "loss": 1.0352,
+      "step": 10166
+    },
+    {
+      "epoch": 0.6177167507138951,
+      "grad_norm": 0.11403120309114456,
+      "learning_rate": 3.2107373467636715e-05,
+      "loss": 1.0478,
+      "step": 10167
+    },
+    {
+      "epoch": 0.6177775077465216,
+      "grad_norm": 0.1351073831319809,
+      "learning_rate": 3.209843494967213e-05,
+      "loss": 1.0764,
+      "step": 10168
+    },
+    {
+      "epoch": 0.6178382647791482,
+      "grad_norm": 0.10452364385128021,
+      "learning_rate": 3.208949708789284e-05,
+      "loss": 1.0281,
+      "step": 10169
+    },
+    {
+      "epoch": 0.6178990218117747,
+      "grad_norm": 0.27388057112693787,
+      "learning_rate": 3.2080559882626505e-05,
+      "loss": 1.0349,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6179597788444012,
+      "grad_norm": 0.13104680180549622,
+      "learning_rate": 3.2071623334200714e-05,
+      "loss": 1.023,
+      "step": 10171
+    },
+    {
+      "epoch": 0.6180205358770278,
+      "grad_norm": 0.13915517926216125,
+      "learning_rate": 3.2062687442943023e-05,
+      "loss": 1.0893,
+      "step": 10172
+    },
+    {
+      "epoch": 0.6180812929096543,
+      "grad_norm": 0.16732257604599,
+      "learning_rate": 3.2053752209181004e-05,
+      "loss": 1.1302,
+      "step": 10173
+    },
+    {
+      "epoch": 0.6181420499422808,
+      "grad_norm": 0.1678365021944046,
+      "learning_rate": 3.2044817633242165e-05,
+      "loss": 1.1527,
+      "step": 10174
+    },
+    {
+      "epoch": 0.6182028069749074,
+      "grad_norm": 0.15535657107830048,
+      "learning_rate": 3.203588371545402e-05,
+      "loss": 0.997,
+      "step": 10175
+    },
+    {
+      "epoch": 0.6182635640075339,
+      "grad_norm": 0.2423597276210785,
+      "learning_rate": 3.2026950456144013e-05,
+      "loss": 1.0559,
+      "step": 10176
+    },
+    {
+      "epoch": 0.6183243210401604,
+      "grad_norm": 0.11685813963413239,
+      "learning_rate": 3.2018017855639606e-05,
+      "loss": 1.0282,
+      "step": 10177
+    },
+    {
+      "epoch": 0.618385078072787,
+      "grad_norm": 0.16012737154960632,
+      "learning_rate": 3.200908591426823e-05,
+      "loss": 1.0619,
+      "step": 10178
+    },
+    {
+      "epoch": 0.6184458351054134,
+      "grad_norm": 0.10786806046962738,
+      "learning_rate": 3.2000154632357305e-05,
+      "loss": 1.0494,
+      "step": 10179
+    },
+    {
+      "epoch": 0.6185065921380399,
+      "grad_norm": 0.1822032630443573,
+      "learning_rate": 3.199122401023417e-05,
+      "loss": 1.159,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6185673491706665,
+      "grad_norm": 0.20492428541183472,
+      "learning_rate": 3.198229404822621e-05,
+      "loss": 1.0545,
+      "step": 10181
+    },
+    {
+      "epoch": 0.618628106203293,
+      "grad_norm": 4.97695779800415,
+      "learning_rate": 3.197336474666076e-05,
+      "loss": 1.1523,
+      "step": 10182
+    },
+    {
+      "epoch": 0.6186888632359195,
+      "grad_norm": 0.16445136070251465,
+      "learning_rate": 3.196443610586509e-05,
+      "loss": 1.1553,
+      "step": 10183
+    },
+    {
+      "epoch": 0.6187496202685461,
+      "grad_norm": 0.31077760457992554,
+      "learning_rate": 3.195550812616651e-05,
+      "loss": 1.1558,
+      "step": 10184
+    },
+    {
+      "epoch": 0.6188103773011726,
+      "grad_norm": 0.1357533186674118,
+      "learning_rate": 3.1946580807892255e-05,
+      "loss": 1.0948,
+      "step": 10185
+    },
+    {
+      "epoch": 0.6188711343337991,
+      "grad_norm": 0.11395826190710068,
+      "learning_rate": 3.1937654151369584e-05,
+      "loss": 1.0557,
+      "step": 10186
+    },
+    {
+      "epoch": 0.6189318913664257,
+      "grad_norm": 0.2083105444908142,
+      "learning_rate": 3.192872815692569e-05,
+      "loss": 1.0904,
+      "step": 10187
+    },
+    {
+      "epoch": 0.6189926483990522,
+      "grad_norm": 0.6378744840621948,
+      "learning_rate": 3.191980282488776e-05,
+      "loss": 1.0577,
+      "step": 10188
+    },
+    {
+      "epoch": 0.6190534054316787,
+      "grad_norm": 0.3138388395309448,
+      "learning_rate": 3.191087815558295e-05,
+      "loss": 1.0547,
+      "step": 10189
+    },
+    {
+      "epoch": 0.6191141624643053,
+      "grad_norm": 0.2703898549079895,
+      "learning_rate": 3.190195414933839e-05,
+      "loss": 1.1554,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6191749194969318,
+      "grad_norm": 0.1886947900056839,
+      "learning_rate": 3.18930308064812e-05,
+      "loss": 1.1593,
+      "step": 10191
+    },
+    {
+      "epoch": 0.6192356765295582,
+      "grad_norm": 0.1358928382396698,
+      "learning_rate": 3.188410812733847e-05,
+      "loss": 1.0587,
+      "step": 10192
+    },
+    {
+      "epoch": 0.6192964335621848,
+      "grad_norm": 0.1611601561307907,
+      "learning_rate": 3.187518611223725e-05,
+      "loss": 1.0332,
+      "step": 10193
+    },
+    {
+      "epoch": 0.6193571905948113,
+      "grad_norm": 0.22348247468471527,
+      "learning_rate": 3.186626476150459e-05,
+      "loss": 1.0699,
+      "step": 10194
+    },
+    {
+      "epoch": 0.6194179476274378,
+      "grad_norm": 0.16034063696861267,
+      "learning_rate": 3.185734407546752e-05,
+      "loss": 1.0695,
+      "step": 10195
+    },
+    {
+      "epoch": 0.6194787046600644,
+      "grad_norm": 0.1393730640411377,
+      "learning_rate": 3.1848424054452994e-05,
+      "loss": 1.0583,
+      "step": 10196
+    },
+    {
+      "epoch": 0.6195394616926909,
+      "grad_norm": 0.15678107738494873,
+      "learning_rate": 3.1839504698787995e-05,
+      "loss": 1.0268,
+      "step": 10197
+    },
+    {
+      "epoch": 0.6196002187253175,
+      "grad_norm": 1.5592325925827026,
+      "learning_rate": 3.1830586008799454e-05,
+      "loss": 1.0568,
+      "step": 10198
+    },
+    {
+      "epoch": 0.619660975757944,
+      "grad_norm": 0.3552964925765991,
+      "learning_rate": 3.1821667984814305e-05,
+      "loss": 1.0747,
+      "step": 10199
+    },
+    {
+      "epoch": 0.6197217327905705,
+      "grad_norm": 0.12567734718322754,
+      "learning_rate": 3.181275062715943e-05,
+      "loss": 1.0094,
+      "step": 10200
+    },
+    {
+      "epoch": 0.6197824898231971,
+      "grad_norm": 0.21178902685642242,
+      "learning_rate": 3.1803833936161676e-05,
+      "loss": 1.1242,
+      "step": 10201
+    },
+    {
+      "epoch": 0.6198432468558236,
+      "grad_norm": 0.20213253796100616,
+      "learning_rate": 3.179491791214795e-05,
+      "loss": 1.0352,
+      "step": 10202
+    },
+    {
+      "epoch": 0.6199040038884501,
+      "grad_norm": 0.4800195097923279,
+      "learning_rate": 3.178600255544498e-05,
+      "loss": 1.027,
+      "step": 10203
+    },
+    {
+      "epoch": 0.6199647609210767,
+      "grad_norm": 0.23955950140953064,
+      "learning_rate": 3.177708786637963e-05,
+      "loss": 1.1343,
+      "step": 10204
+    },
+    {
+      "epoch": 0.6200255179537031,
+      "grad_norm": 0.15821386873722076,
+      "learning_rate": 3.1768173845278645e-05,
+      "loss": 1.1236,
+      "step": 10205
+    },
+    {
+      "epoch": 0.6200862749863296,
+      "grad_norm": 0.14776530861854553,
+      "learning_rate": 3.1759260492468754e-05,
+      "loss": 1.0658,
+      "step": 10206
+    },
+    {
+      "epoch": 0.6201470320189562,
+      "grad_norm": 0.1479075849056244,
+      "learning_rate": 3.175034780827672e-05,
+      "loss": 0.9996,
+      "step": 10207
+    },
+    {
+      "epoch": 0.6202077890515827,
+      "grad_norm": 6.934096336364746,
+      "learning_rate": 3.174143579302922e-05,
+      "loss": 1.0476,
+      "step": 10208
+    },
+    {
+      "epoch": 0.6202685460842092,
+      "grad_norm": 0.17443256080150604,
+      "learning_rate": 3.1732524447052904e-05,
+      "loss": 1.0868,
+      "step": 10209
+    },
+    {
+      "epoch": 0.6203293031168358,
+      "grad_norm": 0.3047315180301666,
+      "learning_rate": 3.1723613770674436e-05,
+      "loss": 1.1481,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6203900601494623,
+      "grad_norm": 0.20682884752750397,
+      "learning_rate": 3.171470376422043e-05,
+      "loss": 1.0244,
+      "step": 10211
+    },
+    {
+      "epoch": 0.6204508171820888,
+      "grad_norm": 0.18574631214141846,
+      "learning_rate": 3.1705794428017506e-05,
+      "loss": 1.0227,
+      "step": 10212
+    },
+    {
+      "epoch": 0.6205115742147154,
+      "grad_norm": 0.11193668842315674,
+      "learning_rate": 3.169688576239222e-05,
+      "loss": 1.04,
+      "step": 10213
+    },
+    {
+      "epoch": 0.6205723312473419,
+      "grad_norm": 0.42144519090652466,
+      "learning_rate": 3.168797776767112e-05,
+      "loss": 1.1116,
+      "step": 10214
+    },
+    {
+      "epoch": 0.6206330882799684,
+      "grad_norm": 0.15046969056129456,
+      "learning_rate": 3.167907044418074e-05,
+      "loss": 1.0039,
+      "step": 10215
+    },
+    {
+      "epoch": 0.620693845312595,
+      "grad_norm": 0.14643318951129913,
+      "learning_rate": 3.1670163792247574e-05,
+      "loss": 1.0977,
+      "step": 10216
+    },
+    {
+      "epoch": 0.6207546023452215,
+      "grad_norm": 0.7566734552383423,
+      "learning_rate": 3.16612578121981e-05,
+      "loss": 1.0646,
+      "step": 10217
+    },
+    {
+      "epoch": 0.6208153593778479,
+      "grad_norm": 0.17360030114650726,
+      "learning_rate": 3.165235250435876e-05,
+      "loss": 1.1453,
+      "step": 10218
+    },
+    {
+      "epoch": 0.6208761164104745,
+      "grad_norm": 0.2574431598186493,
+      "learning_rate": 3.164344786905598e-05,
+      "loss": 1.0381,
+      "step": 10219
+    },
+    {
+      "epoch": 0.620936873443101,
+      "grad_norm": 0.17490963637828827,
+      "learning_rate": 3.163454390661617e-05,
+      "loss": 1.0717,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6209976304757275,
+      "grad_norm": 0.3749448359012604,
+      "learning_rate": 3.162564061736571e-05,
+      "loss": 1.0627,
+      "step": 10221
+    },
+    {
+      "epoch": 0.6210583875083541,
+      "grad_norm": 1.0792803764343262,
+      "learning_rate": 3.1616738001630954e-05,
+      "loss": 1.1325,
+      "step": 10222
+    },
+    {
+      "epoch": 0.6211191445409806,
+      "grad_norm": 0.19147220253944397,
+      "learning_rate": 3.1607836059738196e-05,
+      "loss": 1.0971,
+      "step": 10223
+    },
+    {
+      "epoch": 0.6211799015736071,
+      "grad_norm": 0.3345095217227936,
+      "learning_rate": 3.159893479201377e-05,
+      "loss": 1.18,
+      "step": 10224
+    },
+    {
+      "epoch": 0.6212406586062337,
+      "grad_norm": 0.2410873919725418,
+      "learning_rate": 3.1590034198783956e-05,
+      "loss": 1.0383,
+      "step": 10225
+    },
+    {
+      "epoch": 0.6213014156388602,
+      "grad_norm": 0.24107584357261658,
+      "learning_rate": 3.158113428037499e-05,
+      "loss": 1.1225,
+      "step": 10226
+    },
+    {
+      "epoch": 0.6213621726714867,
+      "grad_norm": 0.19483482837677002,
+      "learning_rate": 3.15722350371131e-05,
+      "loss": 1.0414,
+      "step": 10227
+    },
+    {
+      "epoch": 0.6214229297041133,
+      "grad_norm": 0.14700210094451904,
+      "learning_rate": 3.156333646932451e-05,
+      "loss": 1.0837,
+      "step": 10228
+    },
+    {
+      "epoch": 0.6214836867367398,
+      "grad_norm": 1.29800546169281,
+      "learning_rate": 3.155443857733539e-05,
+      "loss": 1.0978,
+      "step": 10229
+    },
+    {
+      "epoch": 0.6215444437693664,
+      "grad_norm": 0.27275630831718445,
+      "learning_rate": 3.1545541361471876e-05,
+      "loss": 1.0979,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6216052008019929,
+      "grad_norm": 0.16258209943771362,
+      "learning_rate": 3.1536644822060105e-05,
+      "loss": 1.0831,
+      "step": 10231
+    },
+    {
+      "epoch": 0.6216659578346193,
+      "grad_norm": 0.1788981854915619,
+      "learning_rate": 3.1527748959426194e-05,
+      "loss": 1.156,
+      "step": 10232
+    },
+    {
+      "epoch": 0.6217267148672458,
+      "grad_norm": 0.12015987187623978,
+      "learning_rate": 3.151885377389622e-05,
+      "loss": 1.0424,
+      "step": 10233
+    },
+    {
+      "epoch": 0.6217874718998724,
+      "grad_norm": 0.16199100017547607,
+      "learning_rate": 3.150995926579622e-05,
+      "loss": 1.0768,
+      "step": 10234
+    },
+    {
+      "epoch": 0.6218482289324989,
+      "grad_norm": 0.3051406443119049,
+      "learning_rate": 3.150106543545227e-05,
+      "loss": 1.13,
+      "step": 10235
+    },
+    {
+      "epoch": 0.6219089859651254,
+      "grad_norm": 0.12980064749717712,
+      "learning_rate": 3.1492172283190314e-05,
+      "loss": 1.056,
+      "step": 10236
+    },
+    {
+      "epoch": 0.621969742997752,
+      "grad_norm": 0.1398184895515442,
+      "learning_rate": 3.148327980933636e-05,
+      "loss": 1.0652,
+      "step": 10237
+    },
+    {
+      "epoch": 0.6220305000303785,
+      "grad_norm": 0.2077093869447708,
+      "learning_rate": 3.1474388014216376e-05,
+      "loss": 1.1068,
+      "step": 10238
+    },
+    {
+      "epoch": 0.622091257063005,
+      "grad_norm": 0.150908961892128,
+      "learning_rate": 3.1465496898156266e-05,
+      "loss": 1.0443,
+      "step": 10239
+    },
+    {
+      "epoch": 0.6221520140956316,
+      "grad_norm": 0.1392572820186615,
+      "learning_rate": 3.145660646148196e-05,
+      "loss": 1.0423,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6222127711282581,
+      "grad_norm": 0.4874022603034973,
+      "learning_rate": 3.1447716704519326e-05,
+      "loss": 1.1987,
+      "step": 10241
+    },
+    {
+      "epoch": 0.6222735281608847,
+      "grad_norm": 2.573227643966675,
+      "learning_rate": 3.143882762759424e-05,
+      "loss": 1.2781,
+      "step": 10242
+    },
+    {
+      "epoch": 0.6223342851935112,
+      "grad_norm": 0.21928462386131287,
+      "learning_rate": 3.1429939231032494e-05,
+      "loss": 1.1183,
+      "step": 10243
+    },
+    {
+      "epoch": 0.6223950422261377,
+      "grad_norm": 0.24733304977416992,
+      "learning_rate": 3.142105151515991e-05,
+      "loss": 1.131,
+      "step": 10244
+    },
+    {
+      "epoch": 0.6224557992587642,
+      "grad_norm": 0.28779125213623047,
+      "learning_rate": 3.1412164480302286e-05,
+      "loss": 1.1175,
+      "step": 10245
+    },
+    {
+      "epoch": 0.6225165562913907,
+      "grad_norm": 0.21045023202896118,
+      "learning_rate": 3.140327812678536e-05,
+      "loss": 1.0034,
+      "step": 10246
+    },
+    {
+      "epoch": 0.6225773133240172,
+      "grad_norm": 0.9853264689445496,
+      "learning_rate": 3.139439245493486e-05,
+      "loss": 1.0436,
+      "step": 10247
+    },
+    {
+      "epoch": 0.6226380703566438,
+      "grad_norm": 0.14474140107631683,
+      "learning_rate": 3.138550746507652e-05,
+      "loss": 1.0312,
+      "step": 10248
+    },
+    {
+      "epoch": 0.6226988273892703,
+      "grad_norm": 0.14982201159000397,
+      "learning_rate": 3.137662315753601e-05,
+      "loss": 1.0111,
+      "step": 10249
+    },
+    {
+      "epoch": 0.6227595844218968,
+      "grad_norm": 0.17669539153575897,
+      "learning_rate": 3.1367739532638964e-05,
+      "loss": 1.1114,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6228203414545234,
+      "grad_norm": 0.5805287957191467,
+      "learning_rate": 3.135885659071102e-05,
+      "loss": 1.28,
+      "step": 10251
+    },
+    {
+      "epoch": 0.6228810984871499,
+      "grad_norm": 0.19784559309482574,
+      "learning_rate": 3.134997433207779e-05,
+      "loss": 1.0695,
+      "step": 10252
+    },
+    {
+      "epoch": 0.6229418555197764,
+      "grad_norm": 0.1529422402381897,
+      "learning_rate": 3.1341092757064864e-05,
+      "loss": 1.1249,
+      "step": 10253
+    },
+    {
+      "epoch": 0.623002612552403,
+      "grad_norm": 0.11288879811763763,
+      "learning_rate": 3.1332211865997786e-05,
+      "loss": 1.0073,
+      "step": 10254
+    },
+    {
+      "epoch": 0.6230633695850295,
+      "grad_norm": 0.2541508972644806,
+      "learning_rate": 3.132333165920211e-05,
+      "loss": 1.0554,
+      "step": 10255
+    },
+    {
+      "epoch": 0.623124126617656,
+      "grad_norm": 1.5954954624176025,
+      "learning_rate": 3.1314452137003305e-05,
+      "loss": 1.0949,
+      "step": 10256
+    },
+    {
+      "epoch": 0.6231848836502826,
+      "grad_norm": 0.1637292206287384,
+      "learning_rate": 3.130557329972685e-05,
+      "loss": 1.1201,
+      "step": 10257
+    },
+    {
+      "epoch": 0.623245640682909,
+      "grad_norm": 0.15132907032966614,
+      "learning_rate": 3.129669514769825e-05,
+      "loss": 1.0161,
+      "step": 10258
+    },
+    {
+      "epoch": 0.6233063977155355,
+      "grad_norm": 0.20064134895801544,
+      "learning_rate": 3.128781768124289e-05,
+      "loss": 1.1348,
+      "step": 10259
+    },
+    {
+      "epoch": 0.6233671547481621,
+      "grad_norm": 1.8573087453842163,
+      "learning_rate": 3.127894090068618e-05,
+      "loss": 1.1136,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6234279117807886,
+      "grad_norm": 0.17055903375148773,
+      "learning_rate": 3.1270064806353525e-05,
+      "loss": 1.0872,
+      "step": 10261
+    },
+    {
+      "epoch": 0.6234886688134151,
+      "grad_norm": 0.34187719225883484,
+      "learning_rate": 3.126118939857027e-05,
+      "loss": 1.091,
+      "step": 10262
+    },
+    {
+      "epoch": 0.6235494258460417,
+      "grad_norm": 0.12974116206169128,
+      "learning_rate": 3.1252314677661734e-05,
+      "loss": 1.1349,
+      "step": 10263
+    },
+    {
+      "epoch": 0.6236101828786682,
+      "grad_norm": 0.15073318779468536,
+      "learning_rate": 3.124344064395322e-05,
+      "loss": 1.0491,
+      "step": 10264
+    },
+    {
+      "epoch": 0.6236709399112947,
+      "grad_norm": 0.12552976608276367,
+      "learning_rate": 3.1234567297770025e-05,
+      "loss": 1.0792,
+      "step": 10265
+    },
+    {
+      "epoch": 0.6237316969439213,
+      "grad_norm": 0.15528784692287445,
+      "learning_rate": 3.122569463943739e-05,
+      "loss": 1.1446,
+      "step": 10266
+    },
+    {
+      "epoch": 0.6237924539765478,
+      "grad_norm": 0.1845795214176178,
+      "learning_rate": 3.1216822669280545e-05,
+      "loss": 1.1041,
+      "step": 10267
+    },
+    {
+      "epoch": 0.6238532110091743,
+      "grad_norm": 0.5201460123062134,
+      "learning_rate": 3.120795138762468e-05,
+      "loss": 1.1157,
+      "step": 10268
+    },
+    {
+      "epoch": 0.6239139680418009,
+      "grad_norm": 0.221136212348938,
+      "learning_rate": 3.119908079479503e-05,
+      "loss": 1.1601,
+      "step": 10269
+    },
+    {
+      "epoch": 0.6239747250744274,
+      "grad_norm": 0.7188616394996643,
+      "learning_rate": 3.1190210891116674e-05,
+      "loss": 1.0447,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6240354821070538,
+      "grad_norm": 0.14929260313510895,
+      "learning_rate": 3.118134167691478e-05,
+      "loss": 1.0258,
+      "step": 10271
+    },
+    {
+      "epoch": 0.6240962391396804,
+      "grad_norm": 0.19001224637031555,
+      "learning_rate": 3.117247315251445e-05,
+      "loss": 1.0106,
+      "step": 10272
+    },
+    {
+      "epoch": 0.6241569961723069,
+      "grad_norm": 0.3410162329673767,
+      "learning_rate": 3.116360531824074e-05,
+      "loss": 1.1658,
+      "step": 10273
+    },
+    {
+      "epoch": 0.6242177532049334,
+      "grad_norm": 0.15914678573608398,
+      "learning_rate": 3.115473817441873e-05,
+      "loss": 1.0854,
+      "step": 10274
+    },
+    {
+      "epoch": 0.62427851023756,
+      "grad_norm": 0.15766802430152893,
+      "learning_rate": 3.114587172137345e-05,
+      "loss": 1.0955,
+      "step": 10275
+    },
+    {
+      "epoch": 0.6243392672701865,
+      "grad_norm": 0.1829780787229538,
+      "learning_rate": 3.113700595942988e-05,
+      "loss": 1.1395,
+      "step": 10276
+    },
+    {
+      "epoch": 0.624400024302813,
+      "grad_norm": 0.4062417149543762,
+      "learning_rate": 3.1128140888912994e-05,
+      "loss": 1.3405,
+      "step": 10277
+    },
+    {
+      "epoch": 0.6244607813354396,
+      "grad_norm": 0.45306307077407837,
+      "learning_rate": 3.1119276510147744e-05,
+      "loss": 1.0387,
+      "step": 10278
+    },
+    {
+      "epoch": 0.6245215383680661,
+      "grad_norm": 0.1936022788286209,
+      "learning_rate": 3.111041282345907e-05,
+      "loss": 1.1977,
+      "step": 10279
+    },
+    {
+      "epoch": 0.6245822954006927,
+      "grad_norm": 0.12848900258541107,
+      "learning_rate": 3.110154982917187e-05,
+      "loss": 1.0389,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6246430524333192,
+      "grad_norm": 0.16689981520175934,
+      "learning_rate": 3.1092687527610994e-05,
+      "loss": 1.0021,
+      "step": 10281
+    },
+    {
+      "epoch": 0.6247038094659457,
+      "grad_norm": 0.18329235911369324,
+      "learning_rate": 3.1083825919101354e-05,
+      "loss": 1.0268,
+      "step": 10282
+    },
+    {
+      "epoch": 0.6247645664985723,
+      "grad_norm": 0.11465802043676376,
+      "learning_rate": 3.107496500396769e-05,
+      "loss": 1.0389,
+      "step": 10283
+    },
+    {
+      "epoch": 0.6248253235311987,
+      "grad_norm": 0.13188806176185608,
+      "learning_rate": 3.1066104782534866e-05,
+      "loss": 1.0612,
+      "step": 10284
+    },
+    {
+      "epoch": 0.6248860805638252,
+      "grad_norm": 0.13338439166545868,
+      "learning_rate": 3.105724525512762e-05,
+      "loss": 1.0394,
+      "step": 10285
+    },
+    {
+      "epoch": 0.6249468375964518,
+      "grad_norm": 0.13666728138923645,
+      "learning_rate": 3.1048386422070704e-05,
+      "loss": 1.0665,
+      "step": 10286
+    },
+    {
+      "epoch": 0.6250075946290783,
+      "grad_norm": 0.6837540864944458,
+      "learning_rate": 3.1039528283688854e-05,
+      "loss": 1.0765,
+      "step": 10287
+    },
+    {
+      "epoch": 0.6250683516617048,
+      "grad_norm": 0.1986069679260254,
+      "learning_rate": 3.1030670840306766e-05,
+      "loss": 1.161,
+      "step": 10288
+    },
+    {
+      "epoch": 0.6251291086943314,
+      "grad_norm": 0.8365297913551331,
+      "learning_rate": 3.102181409224911e-05,
+      "loss": 1.0698,
+      "step": 10289
+    },
+    {
+      "epoch": 0.6251898657269579,
+      "grad_norm": 0.14574310183525085,
+      "learning_rate": 3.101295803984052e-05,
+      "loss": 1.0741,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6252506227595844,
+      "grad_norm": 0.19370320439338684,
+      "learning_rate": 3.100410268340561e-05,
+      "loss": 1.0286,
+      "step": 10291
+    },
+    {
+      "epoch": 0.625311379792211,
+      "grad_norm": 0.13562780618667603,
+      "learning_rate": 3.0995248023269e-05,
+      "loss": 1.0542,
+      "step": 10292
+    },
+    {
+      "epoch": 0.6253721368248375,
+      "grad_norm": 0.1542542725801468,
+      "learning_rate": 3.098639405975525e-05,
+      "loss": 1.0348,
+      "step": 10293
+    },
+    {
+      "epoch": 0.625432893857464,
+      "grad_norm": 0.13755856454372406,
+      "learning_rate": 3.097754079318889e-05,
+      "loss": 1.0213,
+      "step": 10294
+    },
+    {
+      "epoch": 0.6254936508900906,
+      "grad_norm": 0.13255921006202698,
+      "learning_rate": 3.096868822389448e-05,
+      "loss": 1.0832,
+      "step": 10295
+    },
+    {
+      "epoch": 0.6255544079227171,
+      "grad_norm": 0.11111794412136078,
+      "learning_rate": 3.095983635219645e-05,
+      "loss": 1.0273,
+      "step": 10296
+    },
+    {
+      "epoch": 0.6256151649553435,
+      "grad_norm": 0.12085297703742981,
+      "learning_rate": 3.09509851784193e-05,
+      "loss": 1.0675,
+      "step": 10297
+    },
+    {
+      "epoch": 0.6256759219879701,
+      "grad_norm": 0.12704680860042572,
+      "learning_rate": 3.0942134702887485e-05,
+      "loss": 1.0619,
+      "step": 10298
+    },
+    {
+      "epoch": 0.6257366790205966,
+      "grad_norm": 0.14826908707618713,
+      "learning_rate": 3.0933284925925386e-05,
+      "loss": 1.1516,
+      "step": 10299
+    },
+    {
+      "epoch": 0.6257974360532231,
+      "grad_norm": 0.25654202699661255,
+      "learning_rate": 3.092443584785742e-05,
+      "loss": 1.0558,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6258581930858497,
+      "grad_norm": 0.14167040586471558,
+      "learning_rate": 3.0915587469007954e-05,
+      "loss": 1.1049,
+      "step": 10301
+    },
+    {
+      "epoch": 0.6259189501184762,
+      "grad_norm": 0.13542619347572327,
+      "learning_rate": 3.090673978970133e-05,
+      "loss": 1.1049,
+      "step": 10302
+    },
+    {
+      "epoch": 0.6259797071511027,
+      "grad_norm": 0.9043248891830444,
+      "learning_rate": 3.0897892810261844e-05,
+      "loss": 1.0825,
+      "step": 10303
+    },
+    {
+      "epoch": 0.6260404641837293,
+      "grad_norm": 0.19542792439460754,
+      "learning_rate": 3.0889046531013765e-05,
+      "loss": 1.1743,
+      "step": 10304
+    },
+    {
+      "epoch": 0.6261012212163558,
+      "grad_norm": 0.12149197608232498,
+      "learning_rate": 3.0880200952281395e-05,
+      "loss": 1.0514,
+      "step": 10305
+    },
+    {
+      "epoch": 0.6261619782489823,
+      "grad_norm": 0.13701939582824707,
+      "learning_rate": 3.087135607438896e-05,
+      "loss": 1.033,
+      "step": 10306
+    },
+    {
+      "epoch": 0.6262227352816089,
+      "grad_norm": 0.14510488510131836,
+      "learning_rate": 3.086251189766065e-05,
+      "loss": 1.0288,
+      "step": 10307
+    },
+    {
+      "epoch": 0.6262834923142354,
+      "grad_norm": 0.2603748142719269,
+      "learning_rate": 3.0853668422420675e-05,
+      "loss": 1.0656,
+      "step": 10308
+    },
+    {
+      "epoch": 0.626344249346862,
+      "grad_norm": 0.18288971483707428,
+      "learning_rate": 3.08448256489932e-05,
+      "loss": 1.0943,
+      "step": 10309
+    },
+    {
+      "epoch": 0.6264050063794884,
+      "grad_norm": 1.9514892101287842,
+      "learning_rate": 3.083598357770233e-05,
+      "loss": 1.0917,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6264657634121149,
+      "grad_norm": 0.15267646312713623,
+      "learning_rate": 3.082714220887218e-05,
+      "loss": 1.0294,
+      "step": 10311
+    },
+    {
+      "epoch": 0.6265265204447414,
+      "grad_norm": 0.1306448131799698,
+      "learning_rate": 3.081830154282684e-05,
+      "loss": 1.048,
+      "step": 10312
+    },
+    {
+      "epoch": 0.626587277477368,
+      "grad_norm": 0.2210254669189453,
+      "learning_rate": 3.080946157989037e-05,
+      "loss": 1.1473,
+      "step": 10313
+    },
+    {
+      "epoch": 0.6266480345099945,
+      "grad_norm": 0.13626231253147125,
+      "learning_rate": 3.0800622320386795e-05,
+      "loss": 1.0451,
+      "step": 10314
+    },
+    {
+      "epoch": 0.626708791542621,
+      "grad_norm": 0.1890278458595276,
+      "learning_rate": 3.0791783764640134e-05,
+      "loss": 1.0647,
+      "step": 10315
+    },
+    {
+      "epoch": 0.6267695485752476,
+      "grad_norm": 0.13382858037948608,
+      "learning_rate": 3.078294591297432e-05,
+      "loss": 1.0338,
+      "step": 10316
+    },
+    {
+      "epoch": 0.6268303056078741,
+      "grad_norm": 0.17762376368045807,
+      "learning_rate": 3.077410876571336e-05,
+      "loss": 1.0598,
+      "step": 10317
+    },
+    {
+      "epoch": 0.6268910626405007,
+      "grad_norm": 0.17865866422653198,
+      "learning_rate": 3.076527232318116e-05,
+      "loss": 1.1162,
+      "step": 10318
+    },
+    {
+      "epoch": 0.6269518196731272,
+      "grad_norm": 0.20925891399383545,
+      "learning_rate": 3.075643658570161e-05,
+      "loss": 1.1072,
+      "step": 10319
+    },
+    {
+      "epoch": 0.6270125767057537,
+      "grad_norm": 0.1224282830953598,
+      "learning_rate": 3.0747601553598596e-05,
+      "loss": 1.0703,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6270733337383803,
+      "grad_norm": 0.16769208014011383,
+      "learning_rate": 3.0738767227195976e-05,
+      "loss": 1.1299,
+      "step": 10321
+    },
+    {
+      "epoch": 0.6271340907710068,
+      "grad_norm": 0.18002669513225555,
+      "learning_rate": 3.072993360681757e-05,
+      "loss": 1.0029,
+      "step": 10322
+    },
+    {
+      "epoch": 0.6271948478036332,
+      "grad_norm": 0.16207602620124817,
+      "learning_rate": 3.0721100692787177e-05,
+      "loss": 1.0968,
+      "step": 10323
+    },
+    {
+      "epoch": 0.6272556048362598,
+      "grad_norm": 0.14748461544513702,
+      "learning_rate": 3.0712268485428544e-05,
+      "loss": 1.1545,
+      "step": 10324
+    },
+    {
+      "epoch": 0.6273163618688863,
+      "grad_norm": 0.6007484197616577,
+      "learning_rate": 3.070343698506545e-05,
+      "loss": 1.1102,
+      "step": 10325
+    },
+    {
+      "epoch": 0.6273771189015128,
+      "grad_norm": 0.6533710360527039,
+      "learning_rate": 3.06946061920216e-05,
+      "loss": 1.2175,
+      "step": 10326
+    },
+    {
+      "epoch": 0.6274378759341394,
+      "grad_norm": 0.248569518327713,
+      "learning_rate": 3.0685776106620704e-05,
+      "loss": 1.2088,
+      "step": 10327
+    },
+    {
+      "epoch": 0.6274986329667659,
+      "grad_norm": 0.17653091251850128,
+      "learning_rate": 3.0676946729186395e-05,
+      "loss": 1.0316,
+      "step": 10328
+    },
+    {
+      "epoch": 0.6275593899993924,
+      "grad_norm": 0.21529167890548706,
+      "learning_rate": 3.066811806004236e-05,
+      "loss": 1.013,
+      "step": 10329
+    },
+    {
+      "epoch": 0.627620147032019,
+      "grad_norm": 0.24814148247241974,
+      "learning_rate": 3.065929009951219e-05,
+      "loss": 1.0888,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6276809040646455,
+      "grad_norm": 0.16948087513446808,
+      "learning_rate": 3.0650462847919476e-05,
+      "loss": 1.0208,
+      "step": 10331
+    },
+    {
+      "epoch": 0.627741661097272,
+      "grad_norm": 0.25379952788352966,
+      "learning_rate": 3.0641636305587776e-05,
+      "loss": 1.0766,
+      "step": 10332
+    },
+    {
+      "epoch": 0.6278024181298986,
+      "grad_norm": 0.4796024560928345,
+      "learning_rate": 3.063281047284065e-05,
+      "loss": 1.3597,
+      "step": 10333
+    },
+    {
+      "epoch": 0.6278631751625251,
+      "grad_norm": 0.13805541396141052,
+      "learning_rate": 3.0623985350001606e-05,
+      "loss": 1.0247,
+      "step": 10334
+    },
+    {
+      "epoch": 0.6279239321951516,
+      "grad_norm": 0.2653367519378662,
+      "learning_rate": 3.0615160937394105e-05,
+      "loss": 1.1992,
+      "step": 10335
+    },
+    {
+      "epoch": 0.6279846892277782,
+      "grad_norm": 0.2275865077972412,
+      "learning_rate": 3.060633723534167e-05,
+      "loss": 1.1838,
+      "step": 10336
+    },
+    {
+      "epoch": 0.6280454462604046,
+      "grad_norm": 0.2066417783498764,
+      "learning_rate": 3.0597514244167647e-05,
+      "loss": 1.0618,
+      "step": 10337
+    },
+    {
+      "epoch": 0.6281062032930311,
+      "grad_norm": 0.15471604466438293,
+      "learning_rate": 3.05886919641955e-05,
+      "loss": 1.0433,
+      "step": 10338
+    },
+    {
+      "epoch": 0.6281669603256577,
+      "grad_norm": 0.2105868011713028,
+      "learning_rate": 3.057987039574861e-05,
+      "loss": 1.0913,
+      "step": 10339
+    },
+    {
+      "epoch": 0.6282277173582842,
+      "grad_norm": 0.196090430021286,
+      "learning_rate": 3.057104953915031e-05,
+      "loss": 1.1626,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6282884743909107,
+      "grad_norm": 0.12826426327228546,
+      "learning_rate": 3.0562229394723956e-05,
+      "loss": 1.0358,
+      "step": 10341
+    },
+    {
+      "epoch": 0.6283492314235373,
+      "grad_norm": 0.17824378609657288,
+      "learning_rate": 3.055340996279285e-05,
+      "loss": 1.0445,
+      "step": 10342
+    },
+    {
+      "epoch": 0.6284099884561638,
+      "grad_norm": 0.14610296487808228,
+      "learning_rate": 3.054459124368025e-05,
+      "loss": 1.0577,
+      "step": 10343
+    },
+    {
+      "epoch": 0.6284707454887903,
+      "grad_norm": 0.18150828778743744,
+      "learning_rate": 3.053577323770942e-05,
+      "loss": 1.0247,
+      "step": 10344
+    },
+    {
+      "epoch": 0.6285315025214169,
+      "grad_norm": 0.2705344557762146,
+      "learning_rate": 3.0526955945203576e-05,
+      "loss": 1.0826,
+      "step": 10345
+    },
+    {
+      "epoch": 0.6285922595540434,
+      "grad_norm": 0.19752287864685059,
+      "learning_rate": 3.051813936648593e-05,
+      "loss": 1.1792,
+      "step": 10346
+    },
+    {
+      "epoch": 0.6286530165866699,
+      "grad_norm": 0.1566229909658432,
+      "learning_rate": 3.050932350187966e-05,
+      "loss": 1.0608,
+      "step": 10347
+    },
+    {
+      "epoch": 0.6287137736192965,
+      "grad_norm": 0.13781581819057465,
+      "learning_rate": 3.050050835170789e-05,
+      "loss": 1.0477,
+      "step": 10348
+    },
+    {
+      "epoch": 0.628774530651923,
+      "grad_norm": 0.5694911479949951,
+      "learning_rate": 3.0491693916293783e-05,
+      "loss": 1.1422,
+      "step": 10349
+    },
+    {
+      "epoch": 0.6288352876845494,
+      "grad_norm": 0.14116372168064117,
+      "learning_rate": 3.0482880195960382e-05,
+      "loss": 1.0572,
+      "step": 10350
+    },
+    {
+      "epoch": 0.628896044717176,
+      "grad_norm": 0.18642696738243103,
+      "learning_rate": 3.0474067191030785e-05,
+      "loss": 1.0727,
+      "step": 10351
+    },
+    {
+      "epoch": 0.6289568017498025,
+      "grad_norm": 0.1293736845254898,
+      "learning_rate": 3.0465254901828032e-05,
+      "loss": 1.0733,
+      "step": 10352
+    },
+    {
+      "epoch": 0.629017558782429,
+      "grad_norm": 0.2832860052585602,
+      "learning_rate": 3.0456443328675133e-05,
+      "loss": 1.0717,
+      "step": 10353
+    },
+    {
+      "epoch": 0.6290783158150556,
+      "grad_norm": 0.18724596500396729,
+      "learning_rate": 3.0447632471895083e-05,
+      "loss": 1.1637,
+      "step": 10354
+    },
+    {
+      "epoch": 0.6291390728476821,
+      "grad_norm": 0.1231849417090416,
+      "learning_rate": 3.0438822331810846e-05,
+      "loss": 1.0353,
+      "step": 10355
+    },
+    {
+      "epoch": 0.6291998298803086,
+      "grad_norm": 0.13783642649650574,
+      "learning_rate": 3.0430012908745363e-05,
+      "loss": 1.0222,
+      "step": 10356
+    },
+    {
+      "epoch": 0.6292605869129352,
+      "grad_norm": 0.12766417860984802,
+      "learning_rate": 3.0421204203021536e-05,
+      "loss": 1.007,
+      "step": 10357
+    },
+    {
+      "epoch": 0.6293213439455617,
+      "grad_norm": 0.2922574579715729,
+      "learning_rate": 3.041239621496224e-05,
+      "loss": 1.311,
+      "step": 10358
+    },
+    {
+      "epoch": 0.6293821009781883,
+      "grad_norm": 0.16196677088737488,
+      "learning_rate": 3.040358894489036e-05,
+      "loss": 1.0634,
+      "step": 10359
+    },
+    {
+      "epoch": 0.6294428580108148,
+      "grad_norm": 0.1410250961780548,
+      "learning_rate": 3.0394782393128712e-05,
+      "loss": 1.0788,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6295036150434413,
+      "grad_norm": 0.16530294716358185,
+      "learning_rate": 3.0385976560000096e-05,
+      "loss": 1.1272,
+      "step": 10361
+    },
+    {
+      "epoch": 0.6295643720760679,
+      "grad_norm": 0.12859342992305756,
+      "learning_rate": 3.037717144582733e-05,
+      "loss": 1.0155,
+      "step": 10362
+    },
+    {
+      "epoch": 0.6296251291086943,
+      "grad_norm": 0.16427485644817352,
+      "learning_rate": 3.036836705093311e-05,
+      "loss": 1.0501,
+      "step": 10363
+    },
+    {
+      "epoch": 0.6296858861413208,
+      "grad_norm": 0.1354958713054657,
+      "learning_rate": 3.0359563375640198e-05,
+      "loss": 1.0097,
+      "step": 10364
+    },
+    {
+      "epoch": 0.6297466431739474,
+      "grad_norm": 0.15866374969482422,
+      "learning_rate": 3.0350760420271285e-05,
+      "loss": 1.1355,
+      "step": 10365
+    },
+    {
+      "epoch": 0.6298074002065739,
+      "grad_norm": 0.21070945262908936,
+      "learning_rate": 3.0341958185149043e-05,
+      "loss": 1.151,
+      "step": 10366
+    },
+    {
+      "epoch": 0.6298681572392004,
+      "grad_norm": 3.1443984508514404,
+      "learning_rate": 3.033315667059613e-05,
+      "loss": 1.0752,
+      "step": 10367
+    },
+    {
+      "epoch": 0.629928914271827,
+      "grad_norm": 0.18908613920211792,
+      "learning_rate": 3.0324355876935163e-05,
+      "loss": 1.0556,
+      "step": 10368
+    },
+    {
+      "epoch": 0.6299896713044535,
+      "grad_norm": 0.22113458812236786,
+      "learning_rate": 3.0315555804488737e-05,
+      "loss": 1.045,
+      "step": 10369
+    },
+    {
+      "epoch": 0.63005042833708,
+      "grad_norm": 0.16768483817577362,
+      "learning_rate": 3.0306756453579417e-05,
+      "loss": 0.9816,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6301111853697066,
+      "grad_norm": 0.12660662829875946,
+      "learning_rate": 3.029795782452973e-05,
+      "loss": 1.1073,
+      "step": 10371
+    },
+    {
+      "epoch": 0.6301719424023331,
+      "grad_norm": 0.15312685072422028,
+      "learning_rate": 3.0289159917662223e-05,
+      "loss": 1.0493,
+      "step": 10372
+    },
+    {
+      "epoch": 0.6302326994349596,
+      "grad_norm": 0.2500297725200653,
+      "learning_rate": 3.0280362733299362e-05,
+      "loss": 1.1539,
+      "step": 10373
+    },
+    {
+      "epoch": 0.6302934564675862,
+      "grad_norm": 0.17541056871414185,
+      "learning_rate": 3.0271566271763613e-05,
+      "loss": 1.1442,
+      "step": 10374
+    },
+    {
+      "epoch": 0.6303542135002127,
+      "grad_norm": 0.24468013644218445,
+      "learning_rate": 3.026277053337742e-05,
+      "loss": 1.1894,
+      "step": 10375
+    },
+    {
+      "epoch": 0.6304149705328391,
+      "grad_norm": 0.1420433223247528,
+      "learning_rate": 3.0253975518463206e-05,
+      "loss": 1.0121,
+      "step": 10376
+    },
+    {
+      "epoch": 0.6304757275654657,
+      "grad_norm": 0.5652739405632019,
+      "learning_rate": 3.0245181227343322e-05,
+      "loss": 1.0751,
+      "step": 10377
+    },
+    {
+      "epoch": 0.6305364845980922,
+      "grad_norm": 0.22571349143981934,
+      "learning_rate": 3.023638766034014e-05,
+      "loss": 1.1303,
+      "step": 10378
+    },
+    {
+      "epoch": 0.6305972416307187,
+      "grad_norm": 0.14158114790916443,
+      "learning_rate": 3.0227594817775974e-05,
+      "loss": 1.0511,
+      "step": 10379
+    },
+    {
+      "epoch": 0.6306579986633453,
+      "grad_norm": 0.11062348634004593,
+      "learning_rate": 3.0218802699973152e-05,
+      "loss": 1.0611,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6307187556959718,
+      "grad_norm": 0.14190226793289185,
+      "learning_rate": 3.0210011307253945e-05,
+      "loss": 1.0565,
+      "step": 10381
+    },
+    {
+      "epoch": 0.6307795127285983,
+      "grad_norm": 0.24531146883964539,
+      "learning_rate": 3.0201220639940608e-05,
+      "loss": 1.1492,
+      "step": 10382
+    },
+    {
+      "epoch": 0.6308402697612249,
+      "grad_norm": 0.11381656676530838,
+      "learning_rate": 3.0192430698355346e-05,
+      "loss": 1.0532,
+      "step": 10383
+    },
+    {
+      "epoch": 0.6309010267938514,
+      "grad_norm": 0.15549519658088684,
+      "learning_rate": 3.0183641482820353e-05,
+      "loss": 1.1218,
+      "step": 10384
+    },
+    {
+      "epoch": 0.6309617838264779,
+      "grad_norm": 0.5387110710144043,
+      "learning_rate": 3.017485299365782e-05,
+      "loss": 1.17,
+      "step": 10385
+    },
+    {
+      "epoch": 0.6310225408591045,
+      "grad_norm": 0.11693233996629715,
+      "learning_rate": 3.0166065231189888e-05,
+      "loss": 1.0525,
+      "step": 10386
+    },
+    {
+      "epoch": 0.631083297891731,
+      "grad_norm": 0.2796156704425812,
+      "learning_rate": 3.015727819573865e-05,
+      "loss": 1.1601,
+      "step": 10387
+    },
+    {
+      "epoch": 0.6311440549243575,
+      "grad_norm": 0.1062614768743515,
+      "learning_rate": 3.0148491887626235e-05,
+      "loss": 1.0041,
+      "step": 10388
+    },
+    {
+      "epoch": 0.631204811956984,
+      "grad_norm": 0.19674816727638245,
+      "learning_rate": 3.0139706307174698e-05,
+      "loss": 1.1256,
+      "step": 10389
+    },
+    {
+      "epoch": 0.6312655689896105,
+      "grad_norm": 0.24845124781131744,
+      "learning_rate": 3.013092145470605e-05,
+      "loss": 1.0809,
+      "step": 10390
+    },
+    {
+      "epoch": 0.631326326022237,
+      "grad_norm": 0.1545766294002533,
+      "learning_rate": 3.0122137330542318e-05,
+      "loss": 1.065,
+      "step": 10391
+    },
+    {
+      "epoch": 0.6313870830548636,
+      "grad_norm": 0.1449037492275238,
+      "learning_rate": 3.0113353935005474e-05,
+      "loss": 1.0079,
+      "step": 10392
+    },
+    {
+      "epoch": 0.6314478400874901,
+      "grad_norm": 0.2765040993690491,
+      "learning_rate": 3.010457126841749e-05,
+      "loss": 1.1477,
+      "step": 10393
+    },
+    {
+      "epoch": 0.6315085971201166,
+      "grad_norm": 0.14687050879001617,
+      "learning_rate": 3.00957893311003e-05,
+      "loss": 1.0985,
+      "step": 10394
+    },
+    {
+      "epoch": 0.6315693541527432,
+      "grad_norm": 0.6837555170059204,
+      "learning_rate": 3.0087008123375787e-05,
+      "loss": 1.0365,
+      "step": 10395
+    },
+    {
+      "epoch": 0.6316301111853697,
+      "grad_norm": 0.14196206629276276,
+      "learning_rate": 3.007822764556587e-05,
+      "loss": 0.999,
+      "step": 10396
+    },
+    {
+      "epoch": 0.6316908682179962,
+      "grad_norm": 0.16935744881629944,
+      "learning_rate": 3.0069447897992342e-05,
+      "loss": 1.1195,
+      "step": 10397
+    },
+    {
+      "epoch": 0.6317516252506228,
+      "grad_norm": 0.13387644290924072,
+      "learning_rate": 3.006066888097706e-05,
+      "loss": 1.0343,
+      "step": 10398
+    },
+    {
+      "epoch": 0.6318123822832493,
+      "grad_norm": 0.1921854168176651,
+      "learning_rate": 3.0051890594841815e-05,
+      "loss": 1.0586,
+      "step": 10399
+    },
+    {
+      "epoch": 0.6318731393158759,
+      "grad_norm": 0.36501345038414,
+      "learning_rate": 3.0043113039908367e-05,
+      "loss": 1.1813,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6319338963485024,
+      "grad_norm": 0.15174295008182526,
+      "learning_rate": 3.0034336216498477e-05,
+      "loss": 1.0601,
+      "step": 10401
+    },
+    {
+      "epoch": 0.6319946533811288,
+      "grad_norm": 0.19241754710674286,
+      "learning_rate": 3.0025560124933856e-05,
+      "loss": 1.0191,
+      "step": 10402
+    },
+    {
+      "epoch": 0.6320554104137553,
+      "grad_norm": 0.1643778681755066,
+      "learning_rate": 3.0016784765536194e-05,
+      "loss": 1.1427,
+      "step": 10403
+    },
+    {
+      "epoch": 0.6321161674463819,
+      "grad_norm": 0.23425880074501038,
+      "learning_rate": 3.0008010138627142e-05,
+      "loss": 1.1788,
+      "step": 10404
+    },
+    {
+      "epoch": 0.6321769244790084,
+      "grad_norm": 0.23254826664924622,
+      "learning_rate": 2.9999236244528333e-05,
+      "loss": 1.0011,
+      "step": 10405
+    },
+    {
+      "epoch": 0.632237681511635,
+      "grad_norm": 0.4401704668998718,
+      "learning_rate": 2.999046308356139e-05,
+      "loss": 1.1651,
+      "step": 10406
+    },
+    {
+      "epoch": 0.6322984385442615,
+      "grad_norm": 0.1550683230161667,
+      "learning_rate": 2.99816906560479e-05,
+      "loss": 1.0353,
+      "step": 10407
+    },
+    {
+      "epoch": 0.632359195576888,
+      "grad_norm": 0.20659492909908295,
+      "learning_rate": 2.9972918962309393e-05,
+      "loss": 1.0239,
+      "step": 10408
+    },
+    {
+      "epoch": 0.6324199526095146,
+      "grad_norm": 0.14924851059913635,
+      "learning_rate": 2.9964148002667437e-05,
+      "loss": 1.0378,
+      "step": 10409
+    },
+    {
+      "epoch": 0.6324807096421411,
+      "grad_norm": 0.15685848891735077,
+      "learning_rate": 2.9955377777443493e-05,
+      "loss": 1.1113,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6325414666747676,
+      "grad_norm": 0.16926518082618713,
+      "learning_rate": 2.9946608286959056e-05,
+      "loss": 1.0968,
+      "step": 10411
+    },
+    {
+      "epoch": 0.6326022237073942,
+      "grad_norm": 0.1526908278465271,
+      "learning_rate": 2.993783953153556e-05,
+      "loss": 1.0395,
+      "step": 10412
+    },
+    {
+      "epoch": 0.6326629807400207,
+      "grad_norm": 0.11706577986478806,
+      "learning_rate": 2.9929071511494433e-05,
+      "loss": 1.0492,
+      "step": 10413
+    },
+    {
+      "epoch": 0.6327237377726472,
+      "grad_norm": 0.13299399614334106,
+      "learning_rate": 2.9920304227157082e-05,
+      "loss": 1.0153,
+      "step": 10414
+    },
+    {
+      "epoch": 0.6327844948052737,
+      "grad_norm": 0.1327035278081894,
+      "learning_rate": 2.9911537678844846e-05,
+      "loss": 1.1181,
+      "step": 10415
+    },
+    {
+      "epoch": 0.6328452518379002,
+      "grad_norm": 0.12151546776294708,
+      "learning_rate": 2.9902771866879104e-05,
+      "loss": 1.1217,
+      "step": 10416
+    },
+    {
+      "epoch": 0.6329060088705267,
+      "grad_norm": 0.11267746984958649,
+      "learning_rate": 2.989400679158112e-05,
+      "loss": 1.0755,
+      "step": 10417
+    },
+    {
+      "epoch": 0.6329667659031533,
+      "grad_norm": 0.4333186745643616,
+      "learning_rate": 2.988524245327221e-05,
+      "loss": 1.0757,
+      "step": 10418
+    },
+    {
+      "epoch": 0.6330275229357798,
+      "grad_norm": 0.22246746718883514,
+      "learning_rate": 2.9876478852273626e-05,
+      "loss": 1.2279,
+      "step": 10419
+    },
+    {
+      "epoch": 0.6330882799684063,
+      "grad_norm": 0.1636882722377777,
+      "learning_rate": 2.9867715988906602e-05,
+      "loss": 1.1471,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6331490370010329,
+      "grad_norm": 0.16150227189064026,
+      "learning_rate": 2.9858953863492334e-05,
+      "loss": 1.1191,
+      "step": 10421
+    },
+    {
+      "epoch": 0.6332097940336594,
+      "grad_norm": 0.1300274282693863,
+      "learning_rate": 2.9850192476352006e-05,
+      "loss": 1.0367,
+      "step": 10422
+    },
+    {
+      "epoch": 0.6332705510662859,
+      "grad_norm": 0.16975072026252747,
+      "learning_rate": 2.9841431827806787e-05,
+      "loss": 1.0628,
+      "step": 10423
+    },
+    {
+      "epoch": 0.6333313080989125,
+      "grad_norm": 0.17236849665641785,
+      "learning_rate": 2.9832671918177762e-05,
+      "loss": 1.0494,
+      "step": 10424
+    },
+    {
+      "epoch": 0.633392065131539,
+      "grad_norm": 0.1864895075559616,
+      "learning_rate": 2.9823912747786043e-05,
+      "loss": 1.1578,
+      "step": 10425
+    },
+    {
+      "epoch": 0.6334528221641655,
+      "grad_norm": 0.1635923981666565,
+      "learning_rate": 2.9815154316952708e-05,
+      "loss": 1.0333,
+      "step": 10426
+    },
+    {
+      "epoch": 0.6335135791967921,
+      "grad_norm": 0.30224552750587463,
+      "learning_rate": 2.9806396625998794e-05,
+      "loss": 1.1293,
+      "step": 10427
+    },
+    {
+      "epoch": 0.6335743362294185,
+      "grad_norm": 0.15837016701698303,
+      "learning_rate": 2.9797639675245305e-05,
+      "loss": 1.0523,
+      "step": 10428
+    },
+    {
+      "epoch": 0.633635093262045,
+      "grad_norm": 0.2511090636253357,
+      "learning_rate": 2.9788883465013274e-05,
+      "loss": 1.0683,
+      "step": 10429
+    },
+    {
+      "epoch": 0.6336958502946716,
+      "grad_norm": 0.13588260114192963,
+      "learning_rate": 2.9780127995623587e-05,
+      "loss": 1.0483,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6337566073272981,
+      "grad_norm": 0.11286661028862,
+      "learning_rate": 2.977137326739723e-05,
+      "loss": 1.0583,
+      "step": 10431
+    },
+    {
+      "epoch": 0.6338173643599246,
+      "grad_norm": 0.18842312693595886,
+      "learning_rate": 2.97626192806551e-05,
+      "loss": 1.0946,
+      "step": 10432
+    },
+    {
+      "epoch": 0.6338781213925512,
+      "grad_norm": 0.45880672335624695,
+      "learning_rate": 2.9753866035718058e-05,
+      "loss": 1.2263,
+      "step": 10433
+    },
+    {
+      "epoch": 0.6339388784251777,
+      "grad_norm": 0.30782759189605713,
+      "learning_rate": 2.974511353290698e-05,
+      "loss": 1.1732,
+      "step": 10434
+    },
+    {
+      "epoch": 0.6339996354578042,
+      "grad_norm": 0.1627044677734375,
+      "learning_rate": 2.9736361772542687e-05,
+      "loss": 1.06,
+      "step": 10435
+    },
+    {
+      "epoch": 0.6340603924904308,
+      "grad_norm": 0.13880908489227295,
+      "learning_rate": 2.9727610754945972e-05,
+      "loss": 1.0873,
+      "step": 10436
+    },
+    {
+      "epoch": 0.6341211495230573,
+      "grad_norm": 0.1428736448287964,
+      "learning_rate": 2.9718860480437594e-05,
+      "loss": 1.077,
+      "step": 10437
+    },
+    {
+      "epoch": 0.6341819065556838,
+      "grad_norm": 0.10837585479021072,
+      "learning_rate": 2.9710110949338298e-05,
+      "loss": 0.9905,
+      "step": 10438
+    },
+    {
+      "epoch": 0.6342426635883104,
+      "grad_norm": 0.13783054053783417,
+      "learning_rate": 2.970136216196882e-05,
+      "loss": 1.0985,
+      "step": 10439
+    },
+    {
+      "epoch": 0.6343034206209369,
+      "grad_norm": 0.1523531973361969,
+      "learning_rate": 2.969261411864983e-05,
+      "loss": 1.0682,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6343641776535635,
+      "grad_norm": 0.19871826469898224,
+      "learning_rate": 2.968386681970199e-05,
+      "loss": 1.0593,
+      "step": 10441
+    },
+    {
+      "epoch": 0.6344249346861899,
+      "grad_norm": 0.13012000918388367,
+      "learning_rate": 2.9675120265445944e-05,
+      "loss": 1.069,
+      "step": 10442
+    },
+    {
+      "epoch": 0.6344856917188164,
+      "grad_norm": 0.4994335174560547,
+      "learning_rate": 2.9666374456202306e-05,
+      "loss": 1.0828,
+      "step": 10443
+    },
+    {
+      "epoch": 0.634546448751443,
+      "grad_norm": 0.19163455069065094,
+      "learning_rate": 2.9657629392291632e-05,
+      "loss": 1.0965,
+      "step": 10444
+    },
+    {
+      "epoch": 0.6346072057840695,
+      "grad_norm": 0.15011663734912872,
+      "learning_rate": 2.9648885074034483e-05,
+      "loss": 1.1319,
+      "step": 10445
+    },
+    {
+      "epoch": 0.634667962816696,
+      "grad_norm": 0.27260082960128784,
+      "learning_rate": 2.9640141501751384e-05,
+      "loss": 1.3209,
+      "step": 10446
+    },
+    {
+      "epoch": 0.6347287198493226,
+      "grad_norm": 0.12057328224182129,
+      "learning_rate": 2.9631398675762835e-05,
+      "loss": 1.0218,
+      "step": 10447
+    },
+    {
+      "epoch": 0.6347894768819491,
+      "grad_norm": 0.1451476365327835,
+      "learning_rate": 2.9622656596389308e-05,
+      "loss": 1.079,
+      "step": 10448
+    },
+    {
+      "epoch": 0.6348502339145756,
+      "grad_norm": 0.12040580809116364,
+      "learning_rate": 2.9613915263951247e-05,
+      "loss": 1.0153,
+      "step": 10449
+    },
+    {
+      "epoch": 0.6349109909472022,
+      "grad_norm": 0.22193381190299988,
+      "learning_rate": 2.9605174678769054e-05,
+      "loss": 1.1129,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6349717479798287,
+      "grad_norm": 0.3113779127597809,
+      "learning_rate": 2.9596434841163124e-05,
+      "loss": 1.1277,
+      "step": 10451
+    },
+    {
+      "epoch": 0.6350325050124552,
+      "grad_norm": 0.17773836851119995,
+      "learning_rate": 2.958769575145382e-05,
+      "loss": 1.1239,
+      "step": 10452
+    },
+    {
+      "epoch": 0.6350932620450818,
+      "grad_norm": 0.20376388728618622,
+      "learning_rate": 2.9578957409961478e-05,
+      "loss": 1.0878,
+      "step": 10453
+    },
+    {
+      "epoch": 0.6351540190777083,
+      "grad_norm": 0.1555800437927246,
+      "learning_rate": 2.9570219817006384e-05,
+      "loss": 1.2266,
+      "step": 10454
+    },
+    {
+      "epoch": 0.6352147761103347,
+      "grad_norm": 0.1502917855978012,
+      "learning_rate": 2.956148297290885e-05,
+      "loss": 1.1663,
+      "step": 10455
+    },
+    {
+      "epoch": 0.6352755331429613,
+      "grad_norm": 0.5676726698875427,
+      "learning_rate": 2.9552746877989113e-05,
+      "loss": 1.2282,
+      "step": 10456
+    },
+    {
+      "epoch": 0.6353362901755878,
+      "grad_norm": 0.2551223635673523,
+      "learning_rate": 2.9544011532567385e-05,
+      "loss": 1.1928,
+      "step": 10457
+    },
+    {
+      "epoch": 0.6353970472082143,
+      "grad_norm": 0.17729078233242035,
+      "learning_rate": 2.9535276936963864e-05,
+      "loss": 1.1115,
+      "step": 10458
+    },
+    {
+      "epoch": 0.6354578042408409,
+      "grad_norm": 0.1536313146352768,
+      "learning_rate": 2.9526543091498715e-05,
+      "loss": 1.1336,
+      "step": 10459
+    },
+    {
+      "epoch": 0.6355185612734674,
+      "grad_norm": 0.1466873288154602,
+      "learning_rate": 2.9517809996492096e-05,
+      "loss": 1.1072,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6355793183060939,
+      "grad_norm": 0.14344127476215363,
+      "learning_rate": 2.9509077652264106e-05,
+      "loss": 1.0029,
+      "step": 10461
+    },
+    {
+      "epoch": 0.6356400753387205,
+      "grad_norm": 0.6649088263511658,
+      "learning_rate": 2.950034605913483e-05,
+      "loss": 1.1936,
+      "step": 10462
+    },
+    {
+      "epoch": 0.635700832371347,
+      "grad_norm": 0.17644619941711426,
+      "learning_rate": 2.949161521742436e-05,
+      "loss": 1.0747,
+      "step": 10463
+    },
+    {
+      "epoch": 0.6357615894039735,
+      "grad_norm": 0.11065807938575745,
+      "learning_rate": 2.9482885127452676e-05,
+      "loss": 1.0062,
+      "step": 10464
+    },
+    {
+      "epoch": 0.6358223464366001,
+      "grad_norm": 0.13071666657924652,
+      "learning_rate": 2.9474155789539803e-05,
+      "loss": 1.0401,
+      "step": 10465
+    },
+    {
+      "epoch": 0.6358831034692266,
+      "grad_norm": 0.26256754994392395,
+      "learning_rate": 2.9465427204005725e-05,
+      "loss": 1.2293,
+      "step": 10466
+    },
+    {
+      "epoch": 0.6359438605018531,
+      "grad_norm": 0.20208019018173218,
+      "learning_rate": 2.9456699371170372e-05,
+      "loss": 1.0935,
+      "step": 10467
+    },
+    {
+      "epoch": 0.6360046175344796,
+      "grad_norm": 0.11011306941509247,
+      "learning_rate": 2.9447972291353676e-05,
+      "loss": 1.0297,
+      "step": 10468
+    },
+    {
+      "epoch": 0.6360653745671061,
+      "grad_norm": 0.6539217829704285,
+      "learning_rate": 2.943924596487554e-05,
+      "loss": 1.1171,
+      "step": 10469
+    },
+    {
+      "epoch": 0.6361261315997326,
+      "grad_norm": 0.12469131499528885,
+      "learning_rate": 2.9430520392055817e-05,
+      "loss": 1.0509,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6361868886323592,
+      "grad_norm": 0.176783487200737,
+      "learning_rate": 2.9421795573214337e-05,
+      "loss": 1.1422,
+      "step": 10471
+    },
+    {
+      "epoch": 0.6362476456649857,
+      "grad_norm": 0.15529657900333405,
+      "learning_rate": 2.9413071508670907e-05,
+      "loss": 1.1086,
+      "step": 10472
+    },
+    {
+      "epoch": 0.6363084026976122,
+      "grad_norm": 0.13071373105049133,
+      "learning_rate": 2.9404348198745334e-05,
+      "loss": 1.0379,
+      "step": 10473
+    },
+    {
+      "epoch": 0.6363691597302388,
+      "grad_norm": 0.1773880273103714,
+      "learning_rate": 2.9395625643757365e-05,
+      "loss": 1.0744,
+      "step": 10474
+    },
+    {
+      "epoch": 0.6364299167628653,
+      "grad_norm": 0.18066534399986267,
+      "learning_rate": 2.93869038440267e-05,
+      "loss": 1.1468,
+      "step": 10475
+    },
+    {
+      "epoch": 0.6364906737954918,
+      "grad_norm": 0.1577855795621872,
+      "learning_rate": 2.9378182799873094e-05,
+      "loss": 1.1089,
+      "step": 10476
+    },
+    {
+      "epoch": 0.6365514308281184,
+      "grad_norm": 0.2647530436515808,
+      "learning_rate": 2.9369462511616154e-05,
+      "loss": 1.1218,
+      "step": 10477
+    },
+    {
+      "epoch": 0.6366121878607449,
+      "grad_norm": 0.1956561654806137,
+      "learning_rate": 2.936074297957556e-05,
+      "loss": 1.1213,
+      "step": 10478
+    },
+    {
+      "epoch": 0.6366729448933715,
+      "grad_norm": 0.1589953899383545,
+      "learning_rate": 2.9352024204070927e-05,
+      "loss": 1.0411,
+      "step": 10479
+    },
+    {
+      "epoch": 0.636733701925998,
+      "grad_norm": 0.5645908117294312,
+      "learning_rate": 2.934330618542182e-05,
+      "loss": 1.2063,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6367944589586244,
+      "grad_norm": 0.1910964995622635,
+      "learning_rate": 2.9334588923947838e-05,
+      "loss": 1.016,
+      "step": 10481
+    },
+    {
+      "epoch": 0.636855215991251,
+      "grad_norm": 0.14719392359256744,
+      "learning_rate": 2.932587241996848e-05,
+      "loss": 1.0743,
+      "step": 10482
+    },
+    {
+      "epoch": 0.6369159730238775,
+      "grad_norm": 0.17193061113357544,
+      "learning_rate": 2.9317156673803282e-05,
+      "loss": 1.0588,
+      "step": 10483
+    },
+    {
+      "epoch": 0.636976730056504,
+      "grad_norm": 0.13495969772338867,
+      "learning_rate": 2.9308441685771692e-05,
+      "loss": 1.0239,
+      "step": 10484
+    },
+    {
+      "epoch": 0.6370374870891305,
+      "grad_norm": 0.18823975324630737,
+      "learning_rate": 2.9299727456193172e-05,
+      "loss": 1.1506,
+      "step": 10485
+    },
+    {
+      "epoch": 0.6370982441217571,
+      "grad_norm": 0.16708354651927948,
+      "learning_rate": 2.929101398538714e-05,
+      "loss": 1.0709,
+      "step": 10486
+    },
+    {
+      "epoch": 0.6371590011543836,
+      "grad_norm": 0.1167384460568428,
+      "learning_rate": 2.9282301273672996e-05,
+      "loss": 1.0237,
+      "step": 10487
+    },
+    {
+      "epoch": 0.6372197581870102,
+      "grad_norm": 0.18278738856315613,
+      "learning_rate": 2.9273589321370098e-05,
+      "loss": 1.0955,
+      "step": 10488
+    },
+    {
+      "epoch": 0.6372805152196367,
+      "grad_norm": 0.14713147282600403,
+      "learning_rate": 2.9264878128797802e-05,
+      "loss": 1.0901,
+      "step": 10489
+    },
+    {
+      "epoch": 0.6373412722522632,
+      "grad_norm": 0.21573582291603088,
+      "learning_rate": 2.925616769627541e-05,
+      "loss": 1.0395,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6374020292848898,
+      "grad_norm": 0.1737573891878128,
+      "learning_rate": 2.9247458024122193e-05,
+      "loss": 1.0336,
+      "step": 10491
+    },
+    {
+      "epoch": 0.6374627863175163,
+      "grad_norm": 0.38860005140304565,
+      "learning_rate": 2.9238749112657414e-05,
+      "loss": 1.0884,
+      "step": 10492
+    },
+    {
+      "epoch": 0.6375235433501428,
+      "grad_norm": 0.14887909591197968,
+      "learning_rate": 2.923004096220029e-05,
+      "loss": 1.0261,
+      "step": 10493
+    },
+    {
+      "epoch": 0.6375843003827693,
+      "grad_norm": 0.15561585128307343,
+      "learning_rate": 2.922133357307003e-05,
+      "loss": 1.0969,
+      "step": 10494
+    },
+    {
+      "epoch": 0.6376450574153958,
+      "grad_norm": 0.15924541652202606,
+      "learning_rate": 2.921262694558582e-05,
+      "loss": 1.0666,
+      "step": 10495
+    },
+    {
+      "epoch": 0.6377058144480223,
+      "grad_norm": 0.12230934947729111,
+      "learning_rate": 2.9203921080066786e-05,
+      "loss": 0.9906,
+      "step": 10496
+    },
+    {
+      "epoch": 0.6377665714806489,
+      "grad_norm": 0.21622680127620697,
+      "learning_rate": 2.9195215976832034e-05,
+      "loss": 1.1063,
+      "step": 10497
+    },
+    {
+      "epoch": 0.6378273285132754,
+      "grad_norm": 0.27397438883781433,
+      "learning_rate": 2.9186511636200666e-05,
+      "loss": 1.0661,
+      "step": 10498
+    },
+    {
+      "epoch": 0.6378880855459019,
+      "grad_norm": 8.239015579223633,
+      "learning_rate": 2.917780805849173e-05,
+      "loss": 1.1225,
+      "step": 10499
+    },
+    {
+      "epoch": 0.6379488425785285,
+      "grad_norm": 0.17071060836315155,
+      "learning_rate": 2.9169105244024275e-05,
+      "loss": 1.1244,
+      "step": 10500
+    },
+    {
+      "epoch": 0.638009599611155,
+      "grad_norm": 0.14276830852031708,
+      "learning_rate": 2.9160403193117292e-05,
+      "loss": 1.1034,
+      "step": 10501
+    },
+    {
+      "epoch": 0.6380703566437815,
+      "grad_norm": 0.30681511759757996,
+      "learning_rate": 2.9151701906089735e-05,
+      "loss": 1.1993,
+      "step": 10502
+    },
+    {
+      "epoch": 0.6381311136764081,
+      "grad_norm": 0.18345583975315094,
+      "learning_rate": 2.9143001383260614e-05,
+      "loss": 1.1962,
+      "step": 10503
+    },
+    {
+      "epoch": 0.6381918707090346,
+      "grad_norm": 0.11108536273241043,
+      "learning_rate": 2.913430162494879e-05,
+      "loss": 1.0579,
+      "step": 10504
+    },
+    {
+      "epoch": 0.6382526277416611,
+      "grad_norm": 0.14707481861114502,
+      "learning_rate": 2.9125602631473176e-05,
+      "loss": 1.0368,
+      "step": 10505
+    },
+    {
+      "epoch": 0.6383133847742877,
+      "grad_norm": 0.19864681363105774,
+      "learning_rate": 2.911690440315263e-05,
+      "loss": 1.1421,
+      "step": 10506
+    },
+    {
+      "epoch": 0.6383741418069141,
+      "grad_norm": 0.16419346630573273,
+      "learning_rate": 2.9108206940305972e-05,
+      "loss": 1.08,
+      "step": 10507
+    },
+    {
+      "epoch": 0.6384348988395406,
+      "grad_norm": 0.6969701647758484,
+      "learning_rate": 2.9099510243252042e-05,
+      "loss": 1.0362,
+      "step": 10508
+    },
+    {
+      "epoch": 0.6384956558721672,
+      "grad_norm": 0.32078859210014343,
+      "learning_rate": 2.909081431230961e-05,
+      "loss": 1.0556,
+      "step": 10509
+    },
+    {
+      "epoch": 0.6385564129047937,
+      "grad_norm": 0.14009247720241547,
+      "learning_rate": 2.9082119147797437e-05,
+      "loss": 1.0372,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6386171699374202,
+      "grad_norm": 0.9470230340957642,
+      "learning_rate": 2.907342475003421e-05,
+      "loss": 1.0585,
+      "step": 10511
+    },
+    {
+      "epoch": 0.6386779269700468,
+      "grad_norm": 0.14937447011470795,
+      "learning_rate": 2.906473111933863e-05,
+      "loss": 1.2305,
+      "step": 10512
+    },
+    {
+      "epoch": 0.6387386840026733,
+      "grad_norm": 0.17845764756202698,
+      "learning_rate": 2.9056038256029388e-05,
+      "loss": 1.1116,
+      "step": 10513
+    },
+    {
+      "epoch": 0.6387994410352998,
+      "grad_norm": 0.19338946044445038,
+      "learning_rate": 2.9047346160425114e-05,
+      "loss": 1.0547,
+      "step": 10514
+    },
+    {
+      "epoch": 0.6388601980679264,
+      "grad_norm": 0.7391157746315002,
+      "learning_rate": 2.9038654832844425e-05,
+      "loss": 1.0786,
+      "step": 10515
+    },
+    {
+      "epoch": 0.6389209551005529,
+      "grad_norm": 12.721567153930664,
+      "learning_rate": 2.9029964273605902e-05,
+      "loss": 1.0449,
+      "step": 10516
+    },
+    {
+      "epoch": 0.6389817121331794,
+      "grad_norm": 0.18733571469783783,
+      "learning_rate": 2.902127448302806e-05,
+      "loss": 1.0663,
+      "step": 10517
+    },
+    {
+      "epoch": 0.639042469165806,
+      "grad_norm": 0.1624296009540558,
+      "learning_rate": 2.9012585461429475e-05,
+      "loss": 1.036,
+      "step": 10518
+    },
+    {
+      "epoch": 0.6391032261984325,
+      "grad_norm": 0.1422363519668579,
+      "learning_rate": 2.9003897209128627e-05,
+      "loss": 1.02,
+      "step": 10519
+    },
+    {
+      "epoch": 0.6391639832310589,
+      "grad_norm": 0.15627799928188324,
+      "learning_rate": 2.8995209726443988e-05,
+      "loss": 1.0093,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6392247402636855,
+      "grad_norm": 0.43517640233039856,
+      "learning_rate": 2.8986523013694e-05,
+      "loss": 1.0599,
+      "step": 10521
+    },
+    {
+      "epoch": 0.639285497296312,
+      "grad_norm": 0.1844719499349594,
+      "learning_rate": 2.8977837071197067e-05,
+      "loss": 1.1382,
+      "step": 10522
+    },
+    {
+      "epoch": 0.6393462543289385,
+      "grad_norm": 1.516577959060669,
+      "learning_rate": 2.8969151899271573e-05,
+      "loss": 1.0643,
+      "step": 10523
+    },
+    {
+      "epoch": 0.6394070113615651,
+      "grad_norm": 0.1679154336452484,
+      "learning_rate": 2.8960467498235887e-05,
+      "loss": 1.0892,
+      "step": 10524
+    },
+    {
+      "epoch": 0.6394677683941916,
+      "grad_norm": 0.18547673523426056,
+      "learning_rate": 2.895178386840834e-05,
+      "loss": 1.0279,
+      "step": 10525
+    },
+    {
+      "epoch": 0.6395285254268182,
+      "grad_norm": 0.23361608386039734,
+      "learning_rate": 2.8943101010107214e-05,
+      "loss": 1.1299,
+      "step": 10526
+    },
+    {
+      "epoch": 0.6395892824594447,
+      "grad_norm": 0.19014549255371094,
+      "learning_rate": 2.8934418923650795e-05,
+      "loss": 1.032,
+      "step": 10527
+    },
+    {
+      "epoch": 0.6396500394920712,
+      "grad_norm": 0.18541701138019562,
+      "learning_rate": 2.8925737609357296e-05,
+      "loss": 1.0271,
+      "step": 10528
+    },
+    {
+      "epoch": 0.6397107965246978,
+      "grad_norm": 0.18043692409992218,
+      "learning_rate": 2.8917057067544983e-05,
+      "loss": 1.1251,
+      "step": 10529
+    },
+    {
+      "epoch": 0.6397715535573243,
+      "grad_norm": 0.14186227321624756,
+      "learning_rate": 2.8908377298532042e-05,
+      "loss": 1.0757,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6398323105899508,
+      "grad_norm": 0.19671863317489624,
+      "learning_rate": 2.8899698302636586e-05,
+      "loss": 1.1594,
+      "step": 10531
+    },
+    {
+      "epoch": 0.6398930676225774,
+      "grad_norm": 0.14527735114097595,
+      "learning_rate": 2.8891020080176767e-05,
+      "loss": 1.0831,
+      "step": 10532
+    },
+    {
+      "epoch": 0.6399538246552038,
+      "grad_norm": 1.8104153871536255,
+      "learning_rate": 2.888234263147066e-05,
+      "loss": 1.0576,
+      "step": 10533
+    },
+    {
+      "epoch": 0.6400145816878303,
+      "grad_norm": 0.1322147399187088,
+      "learning_rate": 2.887366595683639e-05,
+      "loss": 1.0787,
+      "step": 10534
+    },
+    {
+      "epoch": 0.6400753387204569,
+      "grad_norm": 0.3149656057357788,
+      "learning_rate": 2.8864990056591984e-05,
+      "loss": 1.1864,
+      "step": 10535
+    },
+    {
+      "epoch": 0.6401360957530834,
+      "grad_norm": 0.32331764698028564,
+      "learning_rate": 2.885631493105545e-05,
+      "loss": 1.0815,
+      "step": 10536
+    },
+    {
+      "epoch": 0.6401968527857099,
+      "grad_norm": 0.2264787256717682,
+      "learning_rate": 2.88476405805448e-05,
+      "loss": 1.262,
+      "step": 10537
+    },
+    {
+      "epoch": 0.6402576098183365,
+      "grad_norm": 0.16247673332691193,
+      "learning_rate": 2.8838967005377927e-05,
+      "loss": 1.1141,
+      "step": 10538
+    },
+    {
+      "epoch": 0.640318366850963,
+      "grad_norm": 0.26086339354515076,
+      "learning_rate": 2.8830294205872833e-05,
+      "loss": 1.1664,
+      "step": 10539
+    },
+    {
+      "epoch": 0.6403791238835895,
+      "grad_norm": 0.14740754663944244,
+      "learning_rate": 2.8821622182347395e-05,
+      "loss": 1.1476,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6404398809162161,
+      "grad_norm": 0.5690014362335205,
+      "learning_rate": 2.8812950935119498e-05,
+      "loss": 1.1018,
+      "step": 10541
+    },
+    {
+      "epoch": 0.6405006379488426,
+      "grad_norm": 0.18084542453289032,
+      "learning_rate": 2.8804280464506973e-05,
+      "loss": 1.1514,
+      "step": 10542
+    },
+    {
+      "epoch": 0.6405613949814691,
+      "grad_norm": 0.16567806899547577,
+      "learning_rate": 2.8795610770827643e-05,
+      "loss": 1.0586,
+      "step": 10543
+    },
+    {
+      "epoch": 0.6406221520140957,
+      "grad_norm": 1.2102230787277222,
+      "learning_rate": 2.8786941854399297e-05,
+      "loss": 1.0514,
+      "step": 10544
+    },
+    {
+      "epoch": 0.6406829090467222,
+      "grad_norm": 0.3500695824623108,
+      "learning_rate": 2.8778273715539706e-05,
+      "loss": 1.0911,
+      "step": 10545
+    },
+    {
+      "epoch": 0.6407436660793487,
+      "grad_norm": 0.5379045009613037,
+      "learning_rate": 2.8769606354566585e-05,
+      "loss": 1.0863,
+      "step": 10546
+    },
+    {
+      "epoch": 0.6408044231119752,
+      "grad_norm": 0.22588512301445007,
+      "learning_rate": 2.8760939771797656e-05,
+      "loss": 1.0849,
+      "step": 10547
+    },
+    {
+      "epoch": 0.6408651801446017,
+      "grad_norm": 0.133767768740654,
+      "learning_rate": 2.875227396755058e-05,
+      "loss": 1.0827,
+      "step": 10548
+    },
+    {
+      "epoch": 0.6409259371772282,
+      "grad_norm": 0.22321519255638123,
+      "learning_rate": 2.8743608942142986e-05,
+      "loss": 1.1239,
+      "step": 10549
+    },
+    {
+      "epoch": 0.6409866942098548,
+      "grad_norm": 0.23637591302394867,
+      "learning_rate": 2.8734944695892564e-05,
+      "loss": 1.177,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6410474512424813,
+      "grad_norm": 0.13198213279247284,
+      "learning_rate": 2.8726281229116826e-05,
+      "loss": 1.0856,
+      "step": 10551
+    },
+    {
+      "epoch": 0.6411082082751078,
+      "grad_norm": 0.16887839138507843,
+      "learning_rate": 2.8717618542133362e-05,
+      "loss": 1.0927,
+      "step": 10552
+    },
+    {
+      "epoch": 0.6411689653077344,
+      "grad_norm": 0.2495092898607254,
+      "learning_rate": 2.870895663525971e-05,
+      "loss": 1.0293,
+      "step": 10553
+    },
+    {
+      "epoch": 0.6412297223403609,
+      "grad_norm": 0.17197994887828827,
+      "learning_rate": 2.8700295508813342e-05,
+      "loss": 1.2436,
+      "step": 10554
+    },
+    {
+      "epoch": 0.6412904793729874,
+      "grad_norm": 0.21060949563980103,
+      "learning_rate": 2.869163516311178e-05,
+      "loss": 1.112,
+      "step": 10555
+    },
+    {
+      "epoch": 0.641351236405614,
+      "grad_norm": 0.1529807299375534,
+      "learning_rate": 2.868297559847245e-05,
+      "loss": 1.0931,
+      "step": 10556
+    },
+    {
+      "epoch": 0.6414119934382405,
+      "grad_norm": 0.16953857243061066,
+      "learning_rate": 2.8674316815212797e-05,
+      "loss": 1.0739,
+      "step": 10557
+    },
+    {
+      "epoch": 0.641472750470867,
+      "grad_norm": 0.35685837268829346,
+      "learning_rate": 2.8665658813650153e-05,
+      "loss": 1.3015,
+      "step": 10558
+    },
+    {
+      "epoch": 0.6415335075034936,
+      "grad_norm": 0.11360722035169601,
+      "learning_rate": 2.8657001594101895e-05,
+      "loss": 1.0546,
+      "step": 10559
+    },
+    {
+      "epoch": 0.64159426453612,
+      "grad_norm": 0.22124333679676056,
+      "learning_rate": 2.864834515688539e-05,
+      "loss": 1.0973,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6416550215687465,
+      "grad_norm": 0.14160633087158203,
+      "learning_rate": 2.8639689502317917e-05,
+      "loss": 1.0706,
+      "step": 10561
+    },
+    {
+      "epoch": 0.6417157786013731,
+      "grad_norm": 0.2581811845302582,
+      "learning_rate": 2.863103463071676e-05,
+      "loss": 1.0185,
+      "step": 10562
+    },
+    {
+      "epoch": 0.6417765356339996,
+      "grad_norm": 0.11835815757513046,
+      "learning_rate": 2.862238054239915e-05,
+      "loss": 1.0791,
+      "step": 10563
+    },
+    {
+      "epoch": 0.6418372926666261,
+      "grad_norm": 0.24685421586036682,
+      "learning_rate": 2.861372723768232e-05,
+      "loss": 1.0465,
+      "step": 10564
+    },
+    {
+      "epoch": 0.6418980496992527,
+      "grad_norm": 0.15230706334114075,
+      "learning_rate": 2.8605074716883452e-05,
+      "loss": 1.0551,
+      "step": 10565
+    },
+    {
+      "epoch": 0.6419588067318792,
+      "grad_norm": 0.6146996021270752,
+      "learning_rate": 2.8596422980319704e-05,
+      "loss": 1.0456,
+      "step": 10566
+    },
+    {
+      "epoch": 0.6420195637645058,
+      "grad_norm": 5.732402801513672,
+      "learning_rate": 2.8587772028308208e-05,
+      "loss": 1.0217,
+      "step": 10567
+    },
+    {
+      "epoch": 0.6420803207971323,
+      "grad_norm": 0.1213982030749321,
+      "learning_rate": 2.857912186116607e-05,
+      "loss": 1.0752,
+      "step": 10568
+    },
+    {
+      "epoch": 0.6421410778297588,
+      "grad_norm": 0.12367095053195953,
+      "learning_rate": 2.8570472479210354e-05,
+      "loss": 1.0006,
+      "step": 10569
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 0.14044378697872162,
+      "learning_rate": 2.856182388275812e-05,
+      "loss": 1.0052,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6422625918950119,
+      "grad_norm": 0.14465275406837463,
+      "learning_rate": 2.855317607212637e-05,
+      "loss": 1.04,
+      "step": 10571
+    },
+    {
+      "epoch": 0.6423233489276384,
+      "grad_norm": 0.14098306000232697,
+      "learning_rate": 2.8544529047632095e-05,
+      "loss": 1.143,
+      "step": 10572
+    },
+    {
+      "epoch": 0.6423841059602649,
+      "grad_norm": 0.10634411126375198,
+      "learning_rate": 2.853588280959226e-05,
+      "loss": 1.0059,
+      "step": 10573
+    },
+    {
+      "epoch": 0.6424448629928914,
+      "grad_norm": 0.13710854947566986,
+      "learning_rate": 2.8527237358323768e-05,
+      "loss": 1.1003,
+      "step": 10574
+    },
+    {
+      "epoch": 0.6425056200255179,
+      "grad_norm": 0.1620447039604187,
+      "learning_rate": 2.8518592694143564e-05,
+      "loss": 1.0908,
+      "step": 10575
+    },
+    {
+      "epoch": 0.6425663770581445,
+      "grad_norm": 0.2394874393939972,
+      "learning_rate": 2.8509948817368492e-05,
+      "loss": 1.0776,
+      "step": 10576
+    },
+    {
+      "epoch": 0.642627134090771,
+      "grad_norm": 0.1275450736284256,
+      "learning_rate": 2.850130572831542e-05,
+      "loss": 1.0457,
+      "step": 10577
+    },
+    {
+      "epoch": 0.6426878911233975,
+      "grad_norm": 0.10778269916772842,
+      "learning_rate": 2.8492663427301127e-05,
+      "loss": 1.0434,
+      "step": 10578
+    },
+    {
+      "epoch": 0.6427486481560241,
+      "grad_norm": 0.3755713403224945,
+      "learning_rate": 2.848402191464239e-05,
+      "loss": 1.1312,
+      "step": 10579
+    },
+    {
+      "epoch": 0.6428094051886506,
+      "grad_norm": 0.17444919049739838,
+      "learning_rate": 2.8475381190656013e-05,
+      "loss": 1.0979,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6428701622212771,
+      "grad_norm": 0.13567398488521576,
+      "learning_rate": 2.8466741255658693e-05,
+      "loss": 1.1309,
+      "step": 10581
+    },
+    {
+      "epoch": 0.6429309192539037,
+      "grad_norm": 0.5624414086341858,
+      "learning_rate": 2.8458102109967133e-05,
+      "loss": 1.2213,
+      "step": 10582
+    },
+    {
+      "epoch": 0.6429916762865302,
+      "grad_norm": 0.16844765841960907,
+      "learning_rate": 2.844946375389803e-05,
+      "loss": 1.1224,
+      "step": 10583
+    },
+    {
+      "epoch": 0.6430524333191567,
+      "grad_norm": 1.1051650047302246,
+      "learning_rate": 2.844082618776796e-05,
+      "loss": 1.0876,
+      "step": 10584
+    },
+    {
+      "epoch": 0.6431131903517833,
+      "grad_norm": 0.619744598865509,
+      "learning_rate": 2.843218941189359e-05,
+      "loss": 1.084,
+      "step": 10585
+    },
+    {
+      "epoch": 0.6431739473844097,
+      "grad_norm": 0.15182097256183624,
+      "learning_rate": 2.8423553426591486e-05,
+      "loss": 1.0927,
+      "step": 10586
+    },
+    {
+      "epoch": 0.6432347044170362,
+      "grad_norm": 0.14888085424900055,
+      "learning_rate": 2.84149182321782e-05,
+      "loss": 1.0007,
+      "step": 10587
+    },
+    {
+      "epoch": 0.6432954614496628,
+      "grad_norm": 0.5216175317764282,
+      "learning_rate": 2.840628382897026e-05,
+      "loss": 1.073,
+      "step": 10588
+    },
+    {
+      "epoch": 0.6433562184822893,
+      "grad_norm": 0.19235432147979736,
+      "learning_rate": 2.8397650217284165e-05,
+      "loss": 1.1685,
+      "step": 10589
+    },
+    {
+      "epoch": 0.6434169755149158,
+      "grad_norm": 0.12589845061302185,
+      "learning_rate": 2.838901739743637e-05,
+      "loss": 1.0045,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6434777325475424,
+      "grad_norm": 0.5008223652839661,
+      "learning_rate": 2.8380385369743325e-05,
+      "loss": 1.1294,
+      "step": 10591
+    },
+    {
+      "epoch": 0.6435384895801689,
+      "grad_norm": 0.15247169137001038,
+      "learning_rate": 2.837175413452143e-05,
+      "loss": 1.0924,
+      "step": 10592
+    },
+    {
+      "epoch": 0.6435992466127954,
+      "grad_norm": 0.15467919409275055,
+      "learning_rate": 2.836312369208707e-05,
+      "loss": 1.0603,
+      "step": 10593
+    },
+    {
+      "epoch": 0.643660003645422,
+      "grad_norm": 0.23309412598609924,
+      "learning_rate": 2.835449404275659e-05,
+      "loss": 1.0223,
+      "step": 10594
+    },
+    {
+      "epoch": 0.6437207606780485,
+      "grad_norm": 0.12963947653770447,
+      "learning_rate": 2.83458651868463e-05,
+      "loss": 1.0094,
+      "step": 10595
+    },
+    {
+      "epoch": 0.643781517710675,
+      "grad_norm": 0.1775350123643875,
+      "learning_rate": 2.833723712467252e-05,
+      "loss": 1.163,
+      "step": 10596
+    },
+    {
+      "epoch": 0.6438422747433016,
+      "grad_norm": 0.11030242592096329,
+      "learning_rate": 2.8328609856551524e-05,
+      "loss": 1.002,
+      "step": 10597
+    },
+    {
+      "epoch": 0.6439030317759281,
+      "grad_norm": 0.1264825314283371,
+      "learning_rate": 2.83199833827995e-05,
+      "loss": 1.0638,
+      "step": 10598
+    },
+    {
+      "epoch": 0.6439637888085545,
+      "grad_norm": 0.16903752088546753,
+      "learning_rate": 2.8311357703732683e-05,
+      "loss": 1.1039,
+      "step": 10599
+    },
+    {
+      "epoch": 0.6440245458411811,
+      "grad_norm": 0.14801651239395142,
+      "learning_rate": 2.830273281966722e-05,
+      "loss": 1.0716,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6440853028738076,
+      "grad_norm": 0.120904341340065,
+      "learning_rate": 2.8294108730919295e-05,
+      "loss": 1.0582,
+      "step": 10601
+    },
+    {
+      "epoch": 0.6441460599064341,
+      "grad_norm": 0.23208607733249664,
+      "learning_rate": 2.8285485437805016e-05,
+      "loss": 1.1607,
+      "step": 10602
+    },
+    {
+      "epoch": 0.6442068169390607,
+      "grad_norm": 0.1501462310552597,
+      "learning_rate": 2.827686294064048e-05,
+      "loss": 1.0478,
+      "step": 10603
+    },
+    {
+      "epoch": 0.6442675739716872,
+      "grad_norm": 0.18621088564395905,
+      "learning_rate": 2.8268241239741712e-05,
+      "loss": 1.1535,
+      "step": 10604
+    },
+    {
+      "epoch": 0.6443283310043137,
+      "grad_norm": 0.13974034786224365,
+      "learning_rate": 2.825962033542474e-05,
+      "loss": 1.0888,
+      "step": 10605
+    },
+    {
+      "epoch": 0.6443890880369403,
+      "grad_norm": 0.18857252597808838,
+      "learning_rate": 2.8251000228005614e-05,
+      "loss": 1.1022,
+      "step": 10606
+    },
+    {
+      "epoch": 0.6444498450695668,
+      "grad_norm": 0.12503036856651306,
+      "learning_rate": 2.8242380917800278e-05,
+      "loss": 1.0141,
+      "step": 10607
+    },
+    {
+      "epoch": 0.6445106021021934,
+      "grad_norm": 0.13200336694717407,
+      "learning_rate": 2.823376240512467e-05,
+      "loss": 1.0721,
+      "step": 10608
+    },
+    {
+      "epoch": 0.6445713591348199,
+      "grad_norm": 2.661329507827759,
+      "learning_rate": 2.8225144690294714e-05,
+      "loss": 1.0722,
+      "step": 10609
+    },
+    {
+      "epoch": 0.6446321161674464,
+      "grad_norm": 0.31004250049591064,
+      "learning_rate": 2.8216527773626277e-05,
+      "loss": 1.3008,
+      "step": 10610
+    },
+    {
+      "epoch": 0.644692873200073,
+      "grad_norm": 0.11052168905735016,
+      "learning_rate": 2.8207911655435238e-05,
+      "loss": 1.1473,
+      "step": 10611
+    },
+    {
+      "epoch": 0.6447536302326994,
+      "grad_norm": 0.13838598132133484,
+      "learning_rate": 2.8199296336037406e-05,
+      "loss": 1.0776,
+      "step": 10612
+    },
+    {
+      "epoch": 0.6448143872653259,
+      "grad_norm": 0.22109512984752655,
+      "learning_rate": 2.8190681815748577e-05,
+      "loss": 1.0097,
+      "step": 10613
+    },
+    {
+      "epoch": 0.6448751442979525,
+      "grad_norm": 0.5599778294563293,
+      "learning_rate": 2.8182068094884527e-05,
+      "loss": 1.1104,
+      "step": 10614
+    },
+    {
+      "epoch": 0.644935901330579,
+      "grad_norm": 0.15470832586288452,
+      "learning_rate": 2.8173455173760987e-05,
+      "loss": 1.1861,
+      "step": 10615
+    },
+    {
+      "epoch": 0.6449966583632055,
+      "grad_norm": 0.11315450072288513,
+      "learning_rate": 2.8164843052693645e-05,
+      "loss": 1.004,
+      "step": 10616
+    },
+    {
+      "epoch": 0.6450574153958321,
+      "grad_norm": 0.26550528407096863,
+      "learning_rate": 2.815623173199824e-05,
+      "loss": 1.1666,
+      "step": 10617
+    },
+    {
+      "epoch": 0.6451181724284586,
+      "grad_norm": 0.22693970799446106,
+      "learning_rate": 2.814762121199036e-05,
+      "loss": 1.0744,
+      "step": 10618
+    },
+    {
+      "epoch": 0.6451789294610851,
+      "grad_norm": 0.1459822803735733,
+      "learning_rate": 2.8139011492985657e-05,
+      "loss": 1.0092,
+      "step": 10619
+    },
+    {
+      "epoch": 0.6452396864937117,
+      "grad_norm": 0.21222250163555145,
+      "learning_rate": 2.813040257529971e-05,
+      "loss": 1.1666,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6453004435263382,
+      "grad_norm": 0.1804226189851761,
+      "learning_rate": 2.8121794459248063e-05,
+      "loss": 1.1138,
+      "step": 10621
+    },
+    {
+      "epoch": 0.6453612005589647,
+      "grad_norm": 0.1522487848997116,
+      "learning_rate": 2.811318714514629e-05,
+      "loss": 1.0307,
+      "step": 10622
+    },
+    {
+      "epoch": 0.6454219575915913,
+      "grad_norm": 0.13227388262748718,
+      "learning_rate": 2.8104580633309872e-05,
+      "loss": 1.0279,
+      "step": 10623
+    },
+    {
+      "epoch": 0.6454827146242178,
+      "grad_norm": 0.1361694186925888,
+      "learning_rate": 2.8095974924054303e-05,
+      "loss": 1.0827,
+      "step": 10624
+    },
+    {
+      "epoch": 0.6455434716568442,
+      "grad_norm": 0.12480373680591583,
+      "learning_rate": 2.8087370017694992e-05,
+      "loss": 1.0458,
+      "step": 10625
+    },
+    {
+      "epoch": 0.6456042286894708,
+      "grad_norm": 0.17040294408798218,
+      "learning_rate": 2.8078765914547345e-05,
+      "loss": 1.0888,
+      "step": 10626
+    },
+    {
+      "epoch": 0.6456649857220973,
+      "grad_norm": 0.17852675914764404,
+      "learning_rate": 2.8070162614926787e-05,
+      "loss": 1.1456,
+      "step": 10627
+    },
+    {
+      "epoch": 0.6457257427547238,
+      "grad_norm": 0.11578996479511261,
+      "learning_rate": 2.8061560119148662e-05,
+      "loss": 1.0609,
+      "step": 10628
+    },
+    {
+      "epoch": 0.6457864997873504,
+      "grad_norm": 0.11707695573568344,
+      "learning_rate": 2.8052958427528285e-05,
+      "loss": 0.9949,
+      "step": 10629
+    },
+    {
+      "epoch": 0.6458472568199769,
+      "grad_norm": 0.12889371812343597,
+      "learning_rate": 2.8044357540380984e-05,
+      "loss": 1.0478,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6459080138526034,
+      "grad_norm": 0.1696360856294632,
+      "learning_rate": 2.803575745802196e-05,
+      "loss": 1.08,
+      "step": 10631
+    },
+    {
+      "epoch": 0.64596877088523,
+      "grad_norm": 0.12604303658008575,
+      "learning_rate": 2.8027158180766518e-05,
+      "loss": 1.0659,
+      "step": 10632
+    },
+    {
+      "epoch": 0.6460295279178565,
+      "grad_norm": 0.16087479889392853,
+      "learning_rate": 2.8018559708929837e-05,
+      "loss": 1.048,
+      "step": 10633
+    },
+    {
+      "epoch": 0.646090284950483,
+      "grad_norm": 0.14207258820533752,
+      "learning_rate": 2.8009962042827098e-05,
+      "loss": 1.0718,
+      "step": 10634
+    },
+    {
+      "epoch": 0.6461510419831096,
+      "grad_norm": 0.161500945687294,
+      "learning_rate": 2.8001365182773447e-05,
+      "loss": 1.1255,
+      "step": 10635
+    },
+    {
+      "epoch": 0.6462117990157361,
+      "grad_norm": 0.17684707045555115,
+      "learning_rate": 2.7992769129084005e-05,
+      "loss": 1.1101,
+      "step": 10636
+    },
+    {
+      "epoch": 0.6462725560483626,
+      "grad_norm": 0.1507025808095932,
+      "learning_rate": 2.7984173882073872e-05,
+      "loss": 1.0229,
+      "step": 10637
+    },
+    {
+      "epoch": 0.6463333130809891,
+      "grad_norm": 0.1448822170495987,
+      "learning_rate": 2.7975579442058102e-05,
+      "loss": 1.0163,
+      "step": 10638
+    },
+    {
+      "epoch": 0.6463940701136156,
+      "grad_norm": 0.17672374844551086,
+      "learning_rate": 2.796698580935172e-05,
+      "loss": 1.0837,
+      "step": 10639
+    },
+    {
+      "epoch": 0.6464548271462421,
+      "grad_norm": 0.1632174402475357,
+      "learning_rate": 2.795839298426973e-05,
+      "loss": 1.124,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6465155841788687,
+      "grad_norm": 0.2601765990257263,
+      "learning_rate": 2.7949800967127116e-05,
+      "loss": 1.1536,
+      "step": 10641
+    },
+    {
+      "epoch": 0.6465763412114952,
+      "grad_norm": 0.37360918521881104,
+      "learning_rate": 2.7941209758238786e-05,
+      "loss": 1.2292,
+      "step": 10642
+    },
+    {
+      "epoch": 0.6466370982441217,
+      "grad_norm": 0.3798327147960663,
+      "learning_rate": 2.7932619357919687e-05,
+      "loss": 1.133,
+      "step": 10643
+    },
+    {
+      "epoch": 0.6466978552767483,
+      "grad_norm": 0.11715422570705414,
+      "learning_rate": 2.7924029766484717e-05,
+      "loss": 1.0242,
+      "step": 10644
+    },
+    {
+      "epoch": 0.6467586123093748,
+      "grad_norm": 0.11328744888305664,
+      "learning_rate": 2.7915440984248686e-05,
+      "loss": 1.0338,
+      "step": 10645
+    },
+    {
+      "epoch": 0.6468193693420013,
+      "grad_norm": 0.189795583486557,
+      "learning_rate": 2.7906853011526435e-05,
+      "loss": 1.0895,
+      "step": 10646
+    },
+    {
+      "epoch": 0.6468801263746279,
+      "grad_norm": 0.12398957461118698,
+      "learning_rate": 2.7898265848632733e-05,
+      "loss": 1.0871,
+      "step": 10647
+    },
+    {
+      "epoch": 0.6469408834072544,
+      "grad_norm": 0.12003659456968307,
+      "learning_rate": 2.7889679495882385e-05,
+      "loss": 1.0596,
+      "step": 10648
+    },
+    {
+      "epoch": 0.647001640439881,
+      "grad_norm": 0.19922269880771637,
+      "learning_rate": 2.788109395359012e-05,
+      "loss": 1.0853,
+      "step": 10649
+    },
+    {
+      "epoch": 0.6470623974725075,
+      "grad_norm": 0.12221385538578033,
+      "learning_rate": 2.7872509222070642e-05,
+      "loss": 1.0815,
+      "step": 10650
+    },
+    {
+      "epoch": 0.647123154505134,
+      "grad_norm": 0.21900205314159393,
+      "learning_rate": 2.7863925301638598e-05,
+      "loss": 1.0685,
+      "step": 10651
+    },
+    {
+      "epoch": 0.6471839115377604,
+      "grad_norm": 0.16819904744625092,
+      "learning_rate": 2.7855342192608637e-05,
+      "loss": 1.1614,
+      "step": 10652
+    },
+    {
+      "epoch": 0.647244668570387,
+      "grad_norm": 0.1284661889076233,
+      "learning_rate": 2.7846759895295397e-05,
+      "loss": 1.0313,
+      "step": 10653
+    },
+    {
+      "epoch": 0.6473054256030135,
+      "grad_norm": 0.1185019314289093,
+      "learning_rate": 2.783817841001346e-05,
+      "loss": 1.0814,
+      "step": 10654
+    },
+    {
+      "epoch": 0.64736618263564,
+      "grad_norm": 0.14398090541362762,
+      "learning_rate": 2.7829597737077385e-05,
+      "loss": 1.086,
+      "step": 10655
+    },
+    {
+      "epoch": 0.6474269396682666,
+      "grad_norm": 0.14481377601623535,
+      "learning_rate": 2.782101787680168e-05,
+      "loss": 1.0586,
+      "step": 10656
+    },
+    {
+      "epoch": 0.6474876967008931,
+      "grad_norm": 0.13460202515125275,
+      "learning_rate": 2.781243882950085e-05,
+      "loss": 1.063,
+      "step": 10657
+    },
+    {
+      "epoch": 0.6475484537335197,
+      "grad_norm": 0.3281172811985016,
+      "learning_rate": 2.7803860595489373e-05,
+      "loss": 1.1042,
+      "step": 10658
+    },
+    {
+      "epoch": 0.6476092107661462,
+      "grad_norm": 0.15195156633853912,
+      "learning_rate": 2.7795283175081678e-05,
+      "loss": 1.1059,
+      "step": 10659
+    },
+    {
+      "epoch": 0.6476699677987727,
+      "grad_norm": 0.3348652720451355,
+      "learning_rate": 2.7786706568592163e-05,
+      "loss": 1.1535,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6477307248313993,
+      "grad_norm": 0.20453037321567535,
+      "learning_rate": 2.7778130776335217e-05,
+      "loss": 1.0573,
+      "step": 10661
+    },
+    {
+      "epoch": 0.6477914818640258,
+      "grad_norm": 0.20219139754772186,
+      "learning_rate": 2.776955579862518e-05,
+      "loss": 1.051,
+      "step": 10662
+    },
+    {
+      "epoch": 0.6478522388966523,
+      "grad_norm": 0.1860876977443695,
+      "learning_rate": 2.776098163577635e-05,
+      "loss": 1.1814,
+      "step": 10663
+    },
+    {
+      "epoch": 0.6479129959292789,
+      "grad_norm": 0.7542808055877686,
+      "learning_rate": 2.7752408288103082e-05,
+      "loss": 1.096,
+      "step": 10664
+    },
+    {
+      "epoch": 0.6479737529619053,
+      "grad_norm": 0.17214106023311615,
+      "learning_rate": 2.774383575591956e-05,
+      "loss": 1.0859,
+      "step": 10665
+    },
+    {
+      "epoch": 0.6480345099945318,
+      "grad_norm": 0.17734161019325256,
+      "learning_rate": 2.773526403954005e-05,
+      "loss": 1.0953,
+      "step": 10666
+    },
+    {
+      "epoch": 0.6480952670271584,
+      "grad_norm": 0.1600418984889984,
+      "learning_rate": 2.7726693139278716e-05,
+      "loss": 1.0923,
+      "step": 10667
+    },
+    {
+      "epoch": 0.6481560240597849,
+      "grad_norm": 0.11942606419324875,
+      "learning_rate": 2.7718123055449773e-05,
+      "loss": 1.0367,
+      "step": 10668
+    },
+    {
+      "epoch": 0.6482167810924114,
+      "grad_norm": 0.1817196160554886,
+      "learning_rate": 2.770955378836733e-05,
+      "loss": 1.1031,
+      "step": 10669
+    },
+    {
+      "epoch": 0.648277538125038,
+      "grad_norm": 0.12810561060905457,
+      "learning_rate": 2.7700985338345525e-05,
+      "loss": 1.129,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6483382951576645,
+      "grad_norm": 0.15423093736171722,
+      "learning_rate": 2.769241770569839e-05,
+      "loss": 1.0173,
+      "step": 10671
+    },
+    {
+      "epoch": 0.648399052190291,
+      "grad_norm": 0.1347077190876007,
+      "learning_rate": 2.768385089073998e-05,
+      "loss": 1.0857,
+      "step": 10672
+    },
+    {
+      "epoch": 0.6484598092229176,
+      "grad_norm": 0.09607825428247452,
+      "learning_rate": 2.7675284893784353e-05,
+      "loss": 1.0281,
+      "step": 10673
+    },
+    {
+      "epoch": 0.6485205662555441,
+      "grad_norm": 0.11636609584093094,
+      "learning_rate": 2.766671971514547e-05,
+      "loss": 1.0655,
+      "step": 10674
+    },
+    {
+      "epoch": 0.6485813232881706,
+      "grad_norm": 4.822450160980225,
+      "learning_rate": 2.7658155355137295e-05,
+      "loss": 1.173,
+      "step": 10675
+    },
+    {
+      "epoch": 0.6486420803207972,
+      "grad_norm": 0.14269623160362244,
+      "learning_rate": 2.764959181407375e-05,
+      "loss": 1.0798,
+      "step": 10676
+    },
+    {
+      "epoch": 0.6487028373534237,
+      "grad_norm": 0.21184715628623962,
+      "learning_rate": 2.7641029092268733e-05,
+      "loss": 1.1282,
+      "step": 10677
+    },
+    {
+      "epoch": 0.6487635943860501,
+      "grad_norm": 0.19763189554214478,
+      "learning_rate": 2.7632467190036128e-05,
+      "loss": 1.1804,
+      "step": 10678
+    },
+    {
+      "epoch": 0.6488243514186767,
+      "grad_norm": 0.16328218579292297,
+      "learning_rate": 2.762390610768975e-05,
+      "loss": 1.134,
+      "step": 10679
+    },
+    {
+      "epoch": 0.6488851084513032,
+      "grad_norm": 0.1238720566034317,
+      "learning_rate": 2.761534584554343e-05,
+      "loss": 1.0484,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6489458654839297,
+      "grad_norm": 0.1482764333486557,
+      "learning_rate": 2.7606786403910923e-05,
+      "loss": 1.1044,
+      "step": 10681
+    },
+    {
+      "epoch": 0.6490066225165563,
+      "grad_norm": 0.12053503841161728,
+      "learning_rate": 2.7598227783105995e-05,
+      "loss": 1.0508,
+      "step": 10682
+    },
+    {
+      "epoch": 0.6490673795491828,
+      "grad_norm": 0.2628561556339264,
+      "learning_rate": 2.758966998344234e-05,
+      "loss": 1.2877,
+      "step": 10683
+    },
+    {
+      "epoch": 0.6491281365818093,
+      "grad_norm": 0.12281625717878342,
+      "learning_rate": 2.75811130052337e-05,
+      "loss": 0.9922,
+      "step": 10684
+    },
+    {
+      "epoch": 0.6491888936144359,
+      "grad_norm": 0.11099492013454437,
+      "learning_rate": 2.757255684879368e-05,
+      "loss": 1.0338,
+      "step": 10685
+    },
+    {
+      "epoch": 0.6492496506470624,
+      "grad_norm": 0.12935888767242432,
+      "learning_rate": 2.756400151443592e-05,
+      "loss": 1.2749,
+      "step": 10686
+    },
+    {
+      "epoch": 0.649310407679689,
+      "grad_norm": 0.11092004179954529,
+      "learning_rate": 2.7555447002474023e-05,
+      "loss": 1.0279,
+      "step": 10687
+    },
+    {
+      "epoch": 0.6493711647123155,
+      "grad_norm": 1.9871572256088257,
+      "learning_rate": 2.7546893313221535e-05,
+      "loss": 1.0454,
+      "step": 10688
+    },
+    {
+      "epoch": 0.649431921744942,
+      "grad_norm": 0.19476854801177979,
+      "learning_rate": 2.753834044699203e-05,
+      "loss": 1.2029,
+      "step": 10689
+    },
+    {
+      "epoch": 0.6494926787775686,
+      "grad_norm": 0.4752329885959625,
+      "learning_rate": 2.7529788404098998e-05,
+      "loss": 1.0481,
+      "step": 10690
+    },
+    {
+      "epoch": 0.649553435810195,
+      "grad_norm": 0.1420932561159134,
+      "learning_rate": 2.752123718485594e-05,
+      "loss": 1.0116,
+      "step": 10691
+    },
+    {
+      "epoch": 0.6496141928428215,
+      "grad_norm": 0.15726050734519958,
+      "learning_rate": 2.7512686789576253e-05,
+      "loss": 1.076,
+      "step": 10692
+    },
+    {
+      "epoch": 0.649674949875448,
+      "grad_norm": 0.1392820030450821,
+      "learning_rate": 2.750413721857336e-05,
+      "loss": 1.0952,
+      "step": 10693
+    },
+    {
+      "epoch": 0.6497357069080746,
+      "grad_norm": 0.1012481153011322,
+      "learning_rate": 2.7495588472160687e-05,
+      "loss": 1.0534,
+      "step": 10694
+    },
+    {
+      "epoch": 0.6497964639407011,
+      "grad_norm": 0.1261141300201416,
+      "learning_rate": 2.7487040550651564e-05,
+      "loss": 1.0513,
+      "step": 10695
+    },
+    {
+      "epoch": 0.6498572209733277,
+      "grad_norm": 0.1541125625371933,
+      "learning_rate": 2.7478493454359334e-05,
+      "loss": 1.1897,
+      "step": 10696
+    },
+    {
+      "epoch": 0.6499179780059542,
+      "grad_norm": 0.10333828628063202,
+      "learning_rate": 2.746994718359729e-05,
+      "loss": 1.069,
+      "step": 10697
+    },
+    {
+      "epoch": 0.6499787350385807,
+      "grad_norm": 0.2629009485244751,
+      "learning_rate": 2.7461401738678643e-05,
+      "loss": 1.0752,
+      "step": 10698
+    },
+    {
+      "epoch": 0.6500394920712073,
+      "grad_norm": 0.7989603281021118,
+      "learning_rate": 2.7452857119916696e-05,
+      "loss": 1.3133,
+      "step": 10699
+    },
+    {
+      "epoch": 0.6501002491038338,
+      "grad_norm": 0.3149626851081848,
+      "learning_rate": 2.7444313327624632e-05,
+      "loss": 1.2889,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6501610061364603,
+      "grad_norm": 0.11603736877441406,
+      "learning_rate": 2.743577036211562e-05,
+      "loss": 1.1079,
+      "step": 10701
+    },
+    {
+      "epoch": 0.6502217631690869,
+      "grad_norm": 0.13387934863567352,
+      "learning_rate": 2.7427228223702805e-05,
+      "loss": 1.0513,
+      "step": 10702
+    },
+    {
+      "epoch": 0.6502825202017134,
+      "grad_norm": 0.13568900525569916,
+      "learning_rate": 2.741868691269931e-05,
+      "loss": 1.0568,
+      "step": 10703
+    },
+    {
+      "epoch": 0.6503432772343398,
+      "grad_norm": 0.13343168795108795,
+      "learning_rate": 2.7410146429418206e-05,
+      "loss": 1.0544,
+      "step": 10704
+    },
+    {
+      "epoch": 0.6504040342669664,
+      "grad_norm": 0.12232920527458191,
+      "learning_rate": 2.740160677417255e-05,
+      "loss": 1.0325,
+      "step": 10705
+    },
+    {
+      "epoch": 0.6504647912995929,
+      "grad_norm": 0.7413262724876404,
+      "learning_rate": 2.739306794727537e-05,
+      "loss": 1.2225,
+      "step": 10706
+    },
+    {
+      "epoch": 0.6505255483322194,
+      "grad_norm": 0.15551714599132538,
+      "learning_rate": 2.738452994903965e-05,
+      "loss": 1.1337,
+      "step": 10707
+    },
+    {
+      "epoch": 0.650586305364846,
+      "grad_norm": 0.1549120545387268,
+      "learning_rate": 2.737599277977836e-05,
+      "loss": 1.0281,
+      "step": 10708
+    },
+    {
+      "epoch": 0.6506470623974725,
+      "grad_norm": 0.21087591350078583,
+      "learning_rate": 2.7367456439804407e-05,
+      "loss": 1.0843,
+      "step": 10709
+    },
+    {
+      "epoch": 0.650707819430099,
+      "grad_norm": 0.12115465104579926,
+      "learning_rate": 2.735892092943073e-05,
+      "loss": 1.0472,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6507685764627256,
+      "grad_norm": 0.19685101509094238,
+      "learning_rate": 2.7350386248970207e-05,
+      "loss": 1.1719,
+      "step": 10711
+    },
+    {
+      "epoch": 0.6508293334953521,
+      "grad_norm": 0.12675067782402039,
+      "learning_rate": 2.734185239873563e-05,
+      "loss": 1.0328,
+      "step": 10712
+    },
+    {
+      "epoch": 0.6508900905279786,
+      "grad_norm": 0.2035692185163498,
+      "learning_rate": 2.7333319379039834e-05,
+      "loss": 1.1672,
+      "step": 10713
+    },
+    {
+      "epoch": 0.6509508475606052,
+      "grad_norm": 0.26875996589660645,
+      "learning_rate": 2.7324787190195577e-05,
+      "loss": 1.1056,
+      "step": 10714
+    },
+    {
+      "epoch": 0.6510116045932317,
+      "grad_norm": 0.12651421129703522,
+      "learning_rate": 2.731625583251565e-05,
+      "loss": 1.0526,
+      "step": 10715
+    },
+    {
+      "epoch": 0.6510723616258582,
+      "grad_norm": 0.18824221193790436,
+      "learning_rate": 2.730772530631275e-05,
+      "loss": 1.0381,
+      "step": 10716
+    },
+    {
+      "epoch": 0.6511331186584847,
+      "grad_norm": 0.13760711252689362,
+      "learning_rate": 2.7299195611899587e-05,
+      "loss": 1.0657,
+      "step": 10717
+    },
+    {
+      "epoch": 0.6511938756911112,
+      "grad_norm": 0.14851705729961395,
+      "learning_rate": 2.729066674958878e-05,
+      "loss": 1.1141,
+      "step": 10718
+    },
+    {
+      "epoch": 0.6512546327237377,
+      "grad_norm": 0.1626603901386261,
+      "learning_rate": 2.728213871969294e-05,
+      "loss": 1.0949,
+      "step": 10719
+    },
+    {
+      "epoch": 0.6513153897563643,
+      "grad_norm": 0.14219172298908234,
+      "learning_rate": 2.7273611522524728e-05,
+      "loss": 1.0762,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6513761467889908,
+      "grad_norm": 0.16569769382476807,
+      "learning_rate": 2.7265085158396675e-05,
+      "loss": 1.0168,
+      "step": 10721
+    },
+    {
+      "epoch": 0.6514369038216173,
+      "grad_norm": 0.1216806098818779,
+      "learning_rate": 2.725655962762132e-05,
+      "loss": 1.0689,
+      "step": 10722
+    },
+    {
+      "epoch": 0.6514976608542439,
+      "grad_norm": 0.15485937893390656,
+      "learning_rate": 2.724803493051117e-05,
+      "loss": 1.0672,
+      "step": 10723
+    },
+    {
+      "epoch": 0.6515584178868704,
+      "grad_norm": 0.12191330641508102,
+      "learning_rate": 2.7239511067378692e-05,
+      "loss": 1.0433,
+      "step": 10724
+    },
+    {
+      "epoch": 0.651619174919497,
+      "grad_norm": 0.10191213339567184,
+      "learning_rate": 2.723098803853634e-05,
+      "loss": 1.0429,
+      "step": 10725
+    },
+    {
+      "epoch": 0.6516799319521235,
+      "grad_norm": 0.16689635813236237,
+      "learning_rate": 2.722246584429652e-05,
+      "loss": 1.1634,
+      "step": 10726
+    },
+    {
+      "epoch": 0.65174068898475,
+      "grad_norm": 0.13971641659736633,
+      "learning_rate": 2.7213944484971622e-05,
+      "loss": 1.0802,
+      "step": 10727
+    },
+    {
+      "epoch": 0.6518014460173766,
+      "grad_norm": 0.20681428909301758,
+      "learning_rate": 2.7205423960873987e-05,
+      "loss": 1.0837,
+      "step": 10728
+    },
+    {
+      "epoch": 0.6518622030500031,
+      "grad_norm": 0.4985653758049011,
+      "learning_rate": 2.7196904272315947e-05,
+      "loss": 1.1315,
+      "step": 10729
+    },
+    {
+      "epoch": 0.6519229600826295,
+      "grad_norm": 0.12481482326984406,
+      "learning_rate": 2.718838541960977e-05,
+      "loss": 1.1028,
+      "step": 10730
+    },
+    {
+      "epoch": 0.651983717115256,
+      "grad_norm": 0.1376858800649643,
+      "learning_rate": 2.717986740306777e-05,
+      "loss": 1.1033,
+      "step": 10731
+    },
+    {
+      "epoch": 0.6520444741478826,
+      "grad_norm": 0.3027942180633545,
+      "learning_rate": 2.7171350223002123e-05,
+      "loss": 1.1748,
+      "step": 10732
+    },
+    {
+      "epoch": 0.6521052311805091,
+      "grad_norm": 0.16355596482753754,
+      "learning_rate": 2.716283387972505e-05,
+      "loss": 1.0503,
+      "step": 10733
+    },
+    {
+      "epoch": 0.6521659882131357,
+      "grad_norm": 0.12199196964502335,
+      "learning_rate": 2.715431837354871e-05,
+      "loss": 1.0452,
+      "step": 10734
+    },
+    {
+      "epoch": 0.6522267452457622,
+      "grad_norm": 0.17801955342292786,
+      "learning_rate": 2.7145803704785234e-05,
+      "loss": 1.1548,
+      "step": 10735
+    },
+    {
+      "epoch": 0.6522875022783887,
+      "grad_norm": 0.15752169489860535,
+      "learning_rate": 2.7137289873746752e-05,
+      "loss": 1.1457,
+      "step": 10736
+    },
+    {
+      "epoch": 0.6523482593110153,
+      "grad_norm": 0.16104267537593842,
+      "learning_rate": 2.7128776880745354e-05,
+      "loss": 1.0534,
+      "step": 10737
+    },
+    {
+      "epoch": 0.6524090163436418,
+      "grad_norm": 0.16529831290245056,
+      "learning_rate": 2.7120264726093038e-05,
+      "loss": 1.1007,
+      "step": 10738
+    },
+    {
+      "epoch": 0.6524697733762683,
+      "grad_norm": 0.14979039132595062,
+      "learning_rate": 2.7111753410101838e-05,
+      "loss": 1.1359,
+      "step": 10739
+    },
+    {
+      "epoch": 0.6525305304088949,
+      "grad_norm": 0.129634290933609,
+      "learning_rate": 2.710324293308373e-05,
+      "loss": 1.0222,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6525912874415214,
+      "grad_norm": 0.10899734497070312,
+      "learning_rate": 2.709473329535069e-05,
+      "loss": 1.0715,
+      "step": 10741
+    },
+    {
+      "epoch": 0.6526520444741479,
+      "grad_norm": 0.227709099650383,
+      "learning_rate": 2.708622449721463e-05,
+      "loss": 1.0479,
+      "step": 10742
+    },
+    {
+      "epoch": 0.6527128015067744,
+      "grad_norm": 0.1568709909915924,
+      "learning_rate": 2.707771653898744e-05,
+      "loss": 1.0215,
+      "step": 10743
+    },
+    {
+      "epoch": 0.6527735585394009,
+      "grad_norm": 0.14843539893627167,
+      "learning_rate": 2.7069209420981002e-05,
+      "loss": 1.1602,
+      "step": 10744
+    },
+    {
+      "epoch": 0.6528343155720274,
+      "grad_norm": 0.2109106034040451,
+      "learning_rate": 2.7060703143507088e-05,
+      "loss": 1.0605,
+      "step": 10745
+    },
+    {
+      "epoch": 0.652895072604654,
+      "grad_norm": 0.12173515558242798,
+      "learning_rate": 2.7052197706877553e-05,
+      "loss": 1.0649,
+      "step": 10746
+    },
+    {
+      "epoch": 0.6529558296372805,
+      "grad_norm": 0.14209100604057312,
+      "learning_rate": 2.7043693111404146e-05,
+      "loss": 1.0505,
+      "step": 10747
+    },
+    {
+      "epoch": 0.653016586669907,
+      "grad_norm": 0.15922392904758453,
+      "learning_rate": 2.7035189357398605e-05,
+      "loss": 1.0151,
+      "step": 10748
+    },
+    {
+      "epoch": 0.6530773437025336,
+      "grad_norm": 0.19324637949466705,
+      "learning_rate": 2.7026686445172634e-05,
+      "loss": 1.0934,
+      "step": 10749
+    },
+    {
+      "epoch": 0.6531381007351601,
+      "grad_norm": 0.3077450096607208,
+      "learning_rate": 2.7018184375037913e-05,
+      "loss": 1.11,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6531988577677866,
+      "grad_norm": 0.30469441413879395,
+      "learning_rate": 2.700968314730609e-05,
+      "loss": 1.1193,
+      "step": 10751
+    },
+    {
+      "epoch": 0.6532596148004132,
+      "grad_norm": 0.14849016070365906,
+      "learning_rate": 2.7001182762288774e-05,
+      "loss": 1.0774,
+      "step": 10752
+    },
+    {
+      "epoch": 0.6533203718330397,
+      "grad_norm": 0.18265512585639954,
+      "learning_rate": 2.699268322029755e-05,
+      "loss": 1.0336,
+      "step": 10753
+    },
+    {
+      "epoch": 0.6533811288656662,
+      "grad_norm": 0.3886958062648773,
+      "learning_rate": 2.6984184521643968e-05,
+      "loss": 1.3925,
+      "step": 10754
+    },
+    {
+      "epoch": 0.6534418858982928,
+      "grad_norm": 0.12576435506343842,
+      "learning_rate": 2.6975686666639556e-05,
+      "loss": 1.0325,
+      "step": 10755
+    },
+    {
+      "epoch": 0.6535026429309193,
+      "grad_norm": 0.1552131474018097,
+      "learning_rate": 2.696718965559577e-05,
+      "loss": 1.1016,
+      "step": 10756
+    },
+    {
+      "epoch": 0.6535633999635457,
+      "grad_norm": 0.15450161695480347,
+      "learning_rate": 2.6958693488824127e-05,
+      "loss": 1.1005,
+      "step": 10757
+    },
+    {
+      "epoch": 0.6536241569961723,
+      "grad_norm": 0.12544409930706024,
+      "learning_rate": 2.6950198166636042e-05,
+      "loss": 1.024,
+      "step": 10758
+    },
+    {
+      "epoch": 0.6536849140287988,
+      "grad_norm": 0.21194180846214294,
+      "learning_rate": 2.6941703689342885e-05,
+      "loss": 1.1391,
+      "step": 10759
+    },
+    {
+      "epoch": 0.6537456710614253,
+      "grad_norm": 0.18956564366817474,
+      "learning_rate": 2.6933210057256014e-05,
+      "loss": 0.9917,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6538064280940519,
+      "grad_norm": 0.21792666614055634,
+      "learning_rate": 2.6924717270686807e-05,
+      "loss": 1.1133,
+      "step": 10761
+    },
+    {
+      "epoch": 0.6538671851266784,
+      "grad_norm": 0.12324512004852295,
+      "learning_rate": 2.691622532994655e-05,
+      "loss": 1.0287,
+      "step": 10762
+    },
+    {
+      "epoch": 0.6539279421593049,
+      "grad_norm": 0.21461355686187744,
+      "learning_rate": 2.6907734235346515e-05,
+      "loss": 1.1449,
+      "step": 10763
+    },
+    {
+      "epoch": 0.6539886991919315,
+      "grad_norm": 0.16204892098903656,
+      "learning_rate": 2.6899243987197963e-05,
+      "loss": 1.0116,
+      "step": 10764
+    },
+    {
+      "epoch": 0.654049456224558,
+      "grad_norm": 0.5267785787582397,
+      "learning_rate": 2.689075458581205e-05,
+      "loss": 1.2177,
+      "step": 10765
+    },
+    {
+      "epoch": 0.6541102132571845,
+      "grad_norm": 0.1675375998020172,
+      "learning_rate": 2.6882266031500016e-05,
+      "loss": 1.0919,
+      "step": 10766
+    },
+    {
+      "epoch": 0.6541709702898111,
+      "grad_norm": 0.27366146445274353,
+      "learning_rate": 2.6873778324572986e-05,
+      "loss": 1.0982,
+      "step": 10767
+    },
+    {
+      "epoch": 0.6542317273224376,
+      "grad_norm": 0.18844632804393768,
+      "learning_rate": 2.686529146534208e-05,
+      "loss": 1.0637,
+      "step": 10768
+    },
+    {
+      "epoch": 0.6542924843550642,
+      "grad_norm": 0.16073046624660492,
+      "learning_rate": 2.685680545411839e-05,
+      "loss": 1.0852,
+      "step": 10769
+    },
+    {
+      "epoch": 0.6543532413876906,
+      "grad_norm": 0.1605251580476761,
+      "learning_rate": 2.684832029121297e-05,
+      "loss": 1.103,
+      "step": 10770
+    },
+    {
+      "epoch": 0.6544139984203171,
+      "grad_norm": 0.12175038456916809,
+      "learning_rate": 2.6839835976936843e-05,
+      "loss": 1.0475,
+      "step": 10771
+    },
+    {
+      "epoch": 0.6544747554529436,
+      "grad_norm": 0.22805438935756683,
+      "learning_rate": 2.6831352511601006e-05,
+      "loss": 1.1575,
+      "step": 10772
+    },
+    {
+      "epoch": 0.6545355124855702,
+      "grad_norm": 0.16298699378967285,
+      "learning_rate": 2.6822869895516424e-05,
+      "loss": 1.0302,
+      "step": 10773
+    },
+    {
+      "epoch": 0.6545962695181967,
+      "grad_norm": 0.15765489637851715,
+      "learning_rate": 2.6814388128994023e-05,
+      "loss": 1.0899,
+      "step": 10774
+    },
+    {
+      "epoch": 0.6546570265508233,
+      "grad_norm": 0.17365390062332153,
+      "learning_rate": 2.6805907212344705e-05,
+      "loss": 1.1037,
+      "step": 10775
+    },
+    {
+      "epoch": 0.6547177835834498,
+      "grad_norm": 0.12098949402570724,
+      "learning_rate": 2.6797427145879327e-05,
+      "loss": 1.0278,
+      "step": 10776
+    },
+    {
+      "epoch": 0.6547785406160763,
+      "grad_norm": 0.13909979164600372,
+      "learning_rate": 2.6788947929908757e-05,
+      "loss": 1.0753,
+      "step": 10777
+    },
+    {
+      "epoch": 0.6548392976487029,
+      "grad_norm": 0.1650567203760147,
+      "learning_rate": 2.678046956474381e-05,
+      "loss": 1.0894,
+      "step": 10778
+    },
+    {
+      "epoch": 0.6549000546813294,
+      "grad_norm": 0.4109768271446228,
+      "learning_rate": 2.6771992050695215e-05,
+      "loss": 1.0635,
+      "step": 10779
+    },
+    {
+      "epoch": 0.6549608117139559,
+      "grad_norm": 0.12718398869037628,
+      "learning_rate": 2.6763515388073745e-05,
+      "loss": 1.0622,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6550215687465825,
+      "grad_norm": 2.839046001434326,
+      "learning_rate": 2.6755039577190087e-05,
+      "loss": 1.1341,
+      "step": 10781
+    },
+    {
+      "epoch": 0.655082325779209,
+      "grad_norm": 0.3101520538330078,
+      "learning_rate": 2.6746564618354974e-05,
+      "loss": 1.0789,
+      "step": 10782
+    },
+    {
+      "epoch": 0.6551430828118354,
+      "grad_norm": 0.159837543964386,
+      "learning_rate": 2.6738090511879022e-05,
+      "loss": 1.0418,
+      "step": 10783
+    },
+    {
+      "epoch": 0.655203839844462,
+      "grad_norm": 1.0854791402816772,
+      "learning_rate": 2.6729617258072883e-05,
+      "loss": 1.0776,
+      "step": 10784
+    },
+    {
+      "epoch": 0.6552645968770885,
+      "grad_norm": 0.6836337447166443,
+      "learning_rate": 2.67211448572471e-05,
+      "loss": 1.1007,
+      "step": 10785
+    },
+    {
+      "epoch": 0.655325353909715,
+      "grad_norm": 0.31533119082450867,
+      "learning_rate": 2.6712673309712232e-05,
+      "loss": 1.1158,
+      "step": 10786
+    },
+    {
+      "epoch": 0.6553861109423416,
+      "grad_norm": 0.17787127196788788,
+      "learning_rate": 2.6704202615778844e-05,
+      "loss": 1.1277,
+      "step": 10787
+    },
+    {
+      "epoch": 0.6554468679749681,
+      "grad_norm": 0.16356894373893738,
+      "learning_rate": 2.669573277575741e-05,
+      "loss": 1.1824,
+      "step": 10788
+    },
+    {
+      "epoch": 0.6555076250075946,
+      "grad_norm": 0.23183858394622803,
+      "learning_rate": 2.6687263789958407e-05,
+      "loss": 1.0299,
+      "step": 10789
+    },
+    {
+      "epoch": 0.6555683820402212,
+      "grad_norm": 0.11725512892007828,
+      "learning_rate": 2.6678795658692247e-05,
+      "loss": 1.0734,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6556291390728477,
+      "grad_norm": 0.4919125437736511,
+      "learning_rate": 2.6670328382269338e-05,
+      "loss": 0.9961,
+      "step": 10791
+    },
+    {
+      "epoch": 0.6556898961054742,
+      "grad_norm": 0.185536190867424,
+      "learning_rate": 2.6661861961000058e-05,
+      "loss": 1.1405,
+      "step": 10792
+    },
+    {
+      "epoch": 0.6557506531381008,
+      "grad_norm": 0.1605636030435562,
+      "learning_rate": 2.6653396395194728e-05,
+      "loss": 1.1675,
+      "step": 10793
+    },
+    {
+      "epoch": 0.6558114101707273,
+      "grad_norm": 0.1225455105304718,
+      "learning_rate": 2.6644931685163678e-05,
+      "loss": 1.0777,
+      "step": 10794
+    },
+    {
+      "epoch": 0.6558721672033538,
+      "grad_norm": 0.13150855898857117,
+      "learning_rate": 2.663646783121716e-05,
+      "loss": 1.0669,
+      "step": 10795
+    },
+    {
+      "epoch": 0.6559329242359803,
+      "grad_norm": 0.9739035964012146,
+      "learning_rate": 2.6628004833665443e-05,
+      "loss": 1.1015,
+      "step": 10796
+    },
+    {
+      "epoch": 0.6559936812686068,
+      "grad_norm": 0.10610534995794296,
+      "learning_rate": 2.66195426928187e-05,
+      "loss": 1.0493,
+      "step": 10797
+    },
+    {
+      "epoch": 0.6560544383012333,
+      "grad_norm": 0.20631539821624756,
+      "learning_rate": 2.6611081408987177e-05,
+      "loss": 1.2029,
+      "step": 10798
+    },
+    {
+      "epoch": 0.6561151953338599,
+      "grad_norm": 0.10467001795768738,
+      "learning_rate": 2.660262098248097e-05,
+      "loss": 1.0358,
+      "step": 10799
+    },
+    {
+      "epoch": 0.6561759523664864,
+      "grad_norm": 0.1426679641008377,
+      "learning_rate": 2.6594161413610207e-05,
+      "loss": 1.041,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6562367093991129,
+      "grad_norm": 0.16207429766654968,
+      "learning_rate": 2.6585702702684988e-05,
+      "loss": 1.0635,
+      "step": 10801
+    },
+    {
+      "epoch": 0.6562974664317395,
+      "grad_norm": 0.17047178745269775,
+      "learning_rate": 2.6577244850015337e-05,
+      "loss": 1.0086,
+      "step": 10802
+    },
+    {
+      "epoch": 0.656358223464366,
+      "grad_norm": 0.1677495539188385,
+      "learning_rate": 2.656878785591133e-05,
+      "loss": 1.0827,
+      "step": 10803
+    },
+    {
+      "epoch": 0.6564189804969925,
+      "grad_norm": 0.19676806032657623,
+      "learning_rate": 2.6560331720682952e-05,
+      "loss": 1.1453,
+      "step": 10804
+    },
+    {
+      "epoch": 0.6564797375296191,
+      "grad_norm": 0.12434057891368866,
+      "learning_rate": 2.6551876444640123e-05,
+      "loss": 1.0852,
+      "step": 10805
+    },
+    {
+      "epoch": 0.6565404945622456,
+      "grad_norm": 0.130045548081398,
+      "learning_rate": 2.6543422028092797e-05,
+      "loss": 1.0049,
+      "step": 10806
+    },
+    {
+      "epoch": 0.6566012515948721,
+      "grad_norm": 0.31346607208251953,
+      "learning_rate": 2.6534968471350852e-05,
+      "loss": 1.0867,
+      "step": 10807
+    },
+    {
+      "epoch": 0.6566620086274987,
+      "grad_norm": 0.15215517580509186,
+      "learning_rate": 2.6526515774724193e-05,
+      "loss": 1.0468,
+      "step": 10808
+    },
+    {
+      "epoch": 0.6567227656601251,
+      "grad_norm": 0.15309739112854004,
+      "learning_rate": 2.651806393852263e-05,
+      "loss": 1.0681,
+      "step": 10809
+    },
+    {
+      "epoch": 0.6567835226927516,
+      "grad_norm": 0.12594451010227203,
+      "learning_rate": 2.6509612963055974e-05,
+      "loss": 1.0111,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6568442797253782,
+      "grad_norm": 0.1614355444908142,
+      "learning_rate": 2.650116284863402e-05,
+      "loss": 1.0984,
+      "step": 10811
+    },
+    {
+      "epoch": 0.6569050367580047,
+      "grad_norm": 0.13417722284793854,
+      "learning_rate": 2.6492713595566443e-05,
+      "loss": 1.0524,
+      "step": 10812
+    },
+    {
+      "epoch": 0.6569657937906312,
+      "grad_norm": 0.13377425074577332,
+      "learning_rate": 2.6484265204163005e-05,
+      "loss": 1.0751,
+      "step": 10813
+    },
+    {
+      "epoch": 0.6570265508232578,
+      "grad_norm": 0.14764824509620667,
+      "learning_rate": 2.6475817674733383e-05,
+      "loss": 1.075,
+      "step": 10814
+    },
+    {
+      "epoch": 0.6570873078558843,
+      "grad_norm": 0.11233925819396973,
+      "learning_rate": 2.6467371007587203e-05,
+      "loss": 1.0509,
+      "step": 10815
+    },
+    {
+      "epoch": 0.6571480648885109,
+      "grad_norm": 0.13913224637508392,
+      "learning_rate": 2.645892520303409e-05,
+      "loss": 1.0815,
+      "step": 10816
+    },
+    {
+      "epoch": 0.6572088219211374,
+      "grad_norm": 0.6653662919998169,
+      "learning_rate": 2.6450480261383626e-05,
+      "loss": 1.2268,
+      "step": 10817
+    },
+    {
+      "epoch": 0.6572695789537639,
+      "grad_norm": 0.24871478974819183,
+      "learning_rate": 2.644203618294536e-05,
+      "loss": 1.1791,
+      "step": 10818
+    },
+    {
+      "epoch": 0.6573303359863905,
+      "grad_norm": 0.14004692435264587,
+      "learning_rate": 2.6433592968028807e-05,
+      "loss": 1.0971,
+      "step": 10819
+    },
+    {
+      "epoch": 0.657391093019017,
+      "grad_norm": 0.12995389103889465,
+      "learning_rate": 2.6425150616943463e-05,
+      "loss": 1.0689,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6574518500516435,
+      "grad_norm": 0.16400907933712006,
+      "learning_rate": 2.6416709129998783e-05,
+      "loss": 1.1193,
+      "step": 10821
+    },
+    {
+      "epoch": 0.65751260708427,
+      "grad_norm": 0.19335129857063293,
+      "learning_rate": 2.6408268507504186e-05,
+      "loss": 1.1282,
+      "step": 10822
+    },
+    {
+      "epoch": 0.6575733641168965,
+      "grad_norm": 0.14659562706947327,
+      "learning_rate": 2.639982874976904e-05,
+      "loss": 1.0798,
+      "step": 10823
+    },
+    {
+      "epoch": 0.657634121149523,
+      "grad_norm": 0.20350515842437744,
+      "learning_rate": 2.6391389857102762e-05,
+      "loss": 1.0526,
+      "step": 10824
+    },
+    {
+      "epoch": 0.6576948781821496,
+      "grad_norm": 0.14672085642814636,
+      "learning_rate": 2.638295182981467e-05,
+      "loss": 1.0872,
+      "step": 10825
+    },
+    {
+      "epoch": 0.6577556352147761,
+      "grad_norm": 0.19582423567771912,
+      "learning_rate": 2.6374514668214024e-05,
+      "loss": 1.0811,
+      "step": 10826
+    },
+    {
+      "epoch": 0.6578163922474026,
+      "grad_norm": 0.11572594195604324,
+      "learning_rate": 2.6366078372610103e-05,
+      "loss": 1.0072,
+      "step": 10827
+    },
+    {
+      "epoch": 0.6578771492800292,
+      "grad_norm": 0.2721400260925293,
+      "learning_rate": 2.6357642943312134e-05,
+      "loss": 1.0856,
+      "step": 10828
+    },
+    {
+      "epoch": 0.6579379063126557,
+      "grad_norm": 0.217032328248024,
+      "learning_rate": 2.6349208380629353e-05,
+      "loss": 1.1815,
+      "step": 10829
+    },
+    {
+      "epoch": 0.6579986633452822,
+      "grad_norm": 0.16566398739814758,
+      "learning_rate": 2.6340774684870904e-05,
+      "loss": 1.1236,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6580594203779088,
+      "grad_norm": 0.21326296031475067,
+      "learning_rate": 2.633234185634596e-05,
+      "loss": 1.0908,
+      "step": 10831
+    },
+    {
+      "epoch": 0.6581201774105353,
+      "grad_norm": 0.12842196226119995,
+      "learning_rate": 2.6323909895363575e-05,
+      "loss": 1.0734,
+      "step": 10832
+    },
+    {
+      "epoch": 0.6581809344431618,
+      "grad_norm": 0.18694041669368744,
+      "learning_rate": 2.6315478802232835e-05,
+      "loss": 1.1869,
+      "step": 10833
+    },
+    {
+      "epoch": 0.6582416914757884,
+      "grad_norm": 0.13833561539649963,
+      "learning_rate": 2.6307048577262815e-05,
+      "loss": 1.0638,
+      "step": 10834
+    },
+    {
+      "epoch": 0.6583024485084148,
+      "grad_norm": 0.15762682259082794,
+      "learning_rate": 2.629861922076251e-05,
+      "loss": 1.0498,
+      "step": 10835
+    },
+    {
+      "epoch": 0.6583632055410413,
+      "grad_norm": 0.1437039077281952,
+      "learning_rate": 2.6290190733040897e-05,
+      "loss": 1.0247,
+      "step": 10836
+    },
+    {
+      "epoch": 0.6584239625736679,
+      "grad_norm": 0.23634670674800873,
+      "learning_rate": 2.6281763114406923e-05,
+      "loss": 1.1219,
+      "step": 10837
+    },
+    {
+      "epoch": 0.6584847196062944,
+      "grad_norm": 0.13402417302131653,
+      "learning_rate": 2.627333636516951e-05,
+      "loss": 1.0591,
+      "step": 10838
+    },
+    {
+      "epoch": 0.6585454766389209,
+      "grad_norm": 0.15598180890083313,
+      "learning_rate": 2.626491048563754e-05,
+      "loss": 1.0986,
+      "step": 10839
+    },
+    {
+      "epoch": 0.6586062336715475,
+      "grad_norm": 0.12586364150047302,
+      "learning_rate": 2.6256485476119863e-05,
+      "loss": 1.0323,
+      "step": 10840
+    },
+    {
+      "epoch": 0.658666990704174,
+      "grad_norm": 0.9201759696006775,
+      "learning_rate": 2.6248061336925302e-05,
+      "loss": 1.1661,
+      "step": 10841
+    },
+    {
+      "epoch": 0.6587277477368005,
+      "grad_norm": 0.09927939623594284,
+      "learning_rate": 2.6239638068362636e-05,
+      "loss": 1.0428,
+      "step": 10842
+    },
+    {
+      "epoch": 0.6587885047694271,
+      "grad_norm": 0.16602684557437897,
+      "learning_rate": 2.623121567074064e-05,
+      "loss": 1.0052,
+      "step": 10843
+    },
+    {
+      "epoch": 0.6588492618020536,
+      "grad_norm": 0.22177504003047943,
+      "learning_rate": 2.6222794144368e-05,
+      "loss": 1.1179,
+      "step": 10844
+    },
+    {
+      "epoch": 0.6589100188346801,
+      "grad_norm": 0.13188055157661438,
+      "learning_rate": 2.621437348955348e-05,
+      "loss": 1.0621,
+      "step": 10845
+    },
+    {
+      "epoch": 0.6589707758673067,
+      "grad_norm": 0.12681160867214203,
+      "learning_rate": 2.620595370660568e-05,
+      "loss": 1.0535,
+      "step": 10846
+    },
+    {
+      "epoch": 0.6590315328999332,
+      "grad_norm": 0.162693053483963,
+      "learning_rate": 2.619753479583324e-05,
+      "loss": 1.0399,
+      "step": 10847
+    },
+    {
+      "epoch": 0.6590922899325596,
+      "grad_norm": 0.13260555267333984,
+      "learning_rate": 2.6189116757544764e-05,
+      "loss": 1.0678,
+      "step": 10848
+    },
+    {
+      "epoch": 0.6591530469651862,
+      "grad_norm": 0.16105875372886658,
+      "learning_rate": 2.6180699592048795e-05,
+      "loss": 1.0822,
+      "step": 10849
+    },
+    {
+      "epoch": 0.6592138039978127,
+      "grad_norm": 0.20426367223262787,
+      "learning_rate": 2.617228329965391e-05,
+      "loss": 1.1188,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6592745610304392,
+      "grad_norm": 0.14040818810462952,
+      "learning_rate": 2.6163867880668614e-05,
+      "loss": 1.0195,
+      "step": 10851
+    },
+    {
+      "epoch": 0.6593353180630658,
+      "grad_norm": 0.2034810185432434,
+      "learning_rate": 2.6155453335401315e-05,
+      "loss": 1.1623,
+      "step": 10852
+    },
+    {
+      "epoch": 0.6593960750956923,
+      "grad_norm": 0.14832821488380432,
+      "learning_rate": 2.6147039664160488e-05,
+      "loss": 1.0944,
+      "step": 10853
+    },
+    {
+      "epoch": 0.6594568321283188,
+      "grad_norm": 0.15600016713142395,
+      "learning_rate": 2.6138626867254512e-05,
+      "loss": 1.0418,
+      "step": 10854
+    },
+    {
+      "epoch": 0.6595175891609454,
+      "grad_norm": 0.11755774170160294,
+      "learning_rate": 2.6130214944991804e-05,
+      "loss": 1.0597,
+      "step": 10855
+    },
+    {
+      "epoch": 0.6595783461935719,
+      "grad_norm": 0.14025630056858063,
+      "learning_rate": 2.612180389768068e-05,
+      "loss": 1.0553,
+      "step": 10856
+    },
+    {
+      "epoch": 0.6596391032261985,
+      "grad_norm": 0.15888909995555878,
+      "learning_rate": 2.6113393725629447e-05,
+      "loss": 1.0819,
+      "step": 10857
+    },
+    {
+      "epoch": 0.659699860258825,
+      "grad_norm": 2.1332199573516846,
+      "learning_rate": 2.6104984429146384e-05,
+      "loss": 1.1181,
+      "step": 10858
+    },
+    {
+      "epoch": 0.6597606172914515,
+      "grad_norm": 6.67297887802124,
+      "learning_rate": 2.609657600853974e-05,
+      "loss": 1.1161,
+      "step": 10859
+    },
+    {
+      "epoch": 0.6598213743240781,
+      "grad_norm": 0.16308479011058807,
+      "learning_rate": 2.608816846411772e-05,
+      "loss": 1.1502,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6598821313567046,
+      "grad_norm": 0.12557394802570343,
+      "learning_rate": 2.6079761796188508e-05,
+      "loss": 1.0559,
+      "step": 10861
+    },
+    {
+      "epoch": 0.659942888389331,
+      "grad_norm": 0.4175530672073364,
+      "learning_rate": 2.6071356005060248e-05,
+      "loss": 1.0965,
+      "step": 10862
+    },
+    {
+      "epoch": 0.6600036454219576,
+      "grad_norm": 0.111270472407341,
+      "learning_rate": 2.606295109104106e-05,
+      "loss": 1.0279,
+      "step": 10863
+    },
+    {
+      "epoch": 0.6600644024545841,
+      "grad_norm": 0.11335918307304382,
+      "learning_rate": 2.6054547054439026e-05,
+      "loss": 1.0291,
+      "step": 10864
+    },
+    {
+      "epoch": 0.6601251594872106,
+      "grad_norm": 0.15028226375579834,
+      "learning_rate": 2.60461438955622e-05,
+      "loss": 1.0169,
+      "step": 10865
+    },
+    {
+      "epoch": 0.6601859165198372,
+      "grad_norm": 0.12362881749868393,
+      "learning_rate": 2.6037741614718602e-05,
+      "loss": 1.0731,
+      "step": 10866
+    },
+    {
+      "epoch": 0.6602466735524637,
+      "grad_norm": 0.1826833039522171,
+      "learning_rate": 2.602934021221621e-05,
+      "loss": 1.1245,
+      "step": 10867
+    },
+    {
+      "epoch": 0.6603074305850902,
+      "grad_norm": 0.15229396522045135,
+      "learning_rate": 2.6020939688362995e-05,
+      "loss": 1.1351,
+      "step": 10868
+    },
+    {
+      "epoch": 0.6603681876177168,
+      "grad_norm": 0.5797024965286255,
+      "learning_rate": 2.601254004346685e-05,
+      "loss": 1.4027,
+      "step": 10869
+    },
+    {
+      "epoch": 0.6604289446503433,
+      "grad_norm": 0.3920063078403473,
+      "learning_rate": 2.6004141277835703e-05,
+      "loss": 1.154,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6604897016829698,
+      "grad_norm": 0.12447533756494522,
+      "learning_rate": 2.599574339177741e-05,
+      "loss": 1.0448,
+      "step": 10871
+    },
+    {
+      "epoch": 0.6605504587155964,
+      "grad_norm": 0.16003093123435974,
+      "learning_rate": 2.598734638559977e-05,
+      "loss": 1.0382,
+      "step": 10872
+    },
+    {
+      "epoch": 0.6606112157482229,
+      "grad_norm": 0.17947323620319366,
+      "learning_rate": 2.597895025961059e-05,
+      "loss": 1.1228,
+      "step": 10873
+    },
+    {
+      "epoch": 0.6606719727808494,
+      "grad_norm": 0.22580967843532562,
+      "learning_rate": 2.5970555014117605e-05,
+      "loss": 1.2768,
+      "step": 10874
+    },
+    {
+      "epoch": 0.6607327298134759,
+      "grad_norm": 0.496307909488678,
+      "learning_rate": 2.59621606494286e-05,
+      "loss": 1.0932,
+      "step": 10875
+    },
+    {
+      "epoch": 0.6607934868461024,
+      "grad_norm": 0.1197834238409996,
+      "learning_rate": 2.5953767165851238e-05,
+      "loss": 1.0422,
+      "step": 10876
+    },
+    {
+      "epoch": 0.6608542438787289,
+      "grad_norm": 0.15819180011749268,
+      "learning_rate": 2.5945374563693185e-05,
+      "loss": 1.0388,
+      "step": 10877
+    },
+    {
+      "epoch": 0.6609150009113555,
+      "grad_norm": 0.19857493042945862,
+      "learning_rate": 2.5936982843262102e-05,
+      "loss": 1.0388,
+      "step": 10878
+    },
+    {
+      "epoch": 0.660975757943982,
+      "grad_norm": 0.11137481033802032,
+      "learning_rate": 2.5928592004865526e-05,
+      "loss": 1.0291,
+      "step": 10879
+    },
+    {
+      "epoch": 0.6610365149766085,
+      "grad_norm": 0.29282453656196594,
+      "learning_rate": 2.5920202048811075e-05,
+      "loss": 1.0689,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6610972720092351,
+      "grad_norm": 0.17970162630081177,
+      "learning_rate": 2.591181297540628e-05,
+      "loss": 1.1914,
+      "step": 10881
+    },
+    {
+      "epoch": 0.6611580290418616,
+      "grad_norm": 0.1269974559545517,
+      "learning_rate": 2.590342478495863e-05,
+      "loss": 1.053,
+      "step": 10882
+    },
+    {
+      "epoch": 0.6612187860744881,
+      "grad_norm": 0.13030850887298584,
+      "learning_rate": 2.589503747777561e-05,
+      "loss": 1.0744,
+      "step": 10883
+    },
+    {
+      "epoch": 0.6612795431071147,
+      "grad_norm": 0.13220517337322235,
+      "learning_rate": 2.5886651054164647e-05,
+      "loss": 1.0142,
+      "step": 10884
+    },
+    {
+      "epoch": 0.6613403001397412,
+      "grad_norm": 0.14280858635902405,
+      "learning_rate": 2.5878265514433148e-05,
+      "loss": 1.1011,
+      "step": 10885
+    },
+    {
+      "epoch": 0.6614010571723677,
+      "grad_norm": 0.15455102920532227,
+      "learning_rate": 2.586988085888849e-05,
+      "loss": 1.1168,
+      "step": 10886
+    },
+    {
+      "epoch": 0.6614618142049943,
+      "grad_norm": 0.13458912074565887,
+      "learning_rate": 2.5861497087838016e-05,
+      "loss": 0.9959,
+      "step": 10887
+    },
+    {
+      "epoch": 0.6615225712376207,
+      "grad_norm": 0.22777879238128662,
+      "learning_rate": 2.585311420158903e-05,
+      "loss": 1.0927,
+      "step": 10888
+    },
+    {
+      "epoch": 0.6615833282702472,
+      "grad_norm": 0.18483905494213104,
+      "learning_rate": 2.5844732200448813e-05,
+      "loss": 1.4205,
+      "step": 10889
+    },
+    {
+      "epoch": 0.6616440853028738,
+      "grad_norm": 0.23490746319293976,
+      "learning_rate": 2.5836351084724587e-05,
+      "loss": 1.056,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6617048423355003,
+      "grad_norm": 0.23946048319339752,
+      "learning_rate": 2.5827970854723615e-05,
+      "loss": 1.2313,
+      "step": 10891
+    },
+    {
+      "epoch": 0.6617655993681268,
+      "grad_norm": 0.21814477443695068,
+      "learning_rate": 2.581959151075303e-05,
+      "loss": 1.1219,
+      "step": 10892
+    },
+    {
+      "epoch": 0.6618263564007534,
+      "grad_norm": 0.15023307502269745,
+      "learning_rate": 2.581121305311999e-05,
+      "loss": 1.0767,
+      "step": 10893
+    },
+    {
+      "epoch": 0.6618871134333799,
+      "grad_norm": 0.4145217835903168,
+      "learning_rate": 2.5802835482131616e-05,
+      "loss": 1.0436,
+      "step": 10894
+    },
+    {
+      "epoch": 0.6619478704660064,
+      "grad_norm": 0.21603821218013763,
+      "learning_rate": 2.5794458798094955e-05,
+      "loss": 1.0727,
+      "step": 10895
+    },
+    {
+      "epoch": 0.662008627498633,
+      "grad_norm": 0.13193009793758392,
+      "learning_rate": 2.578608300131712e-05,
+      "loss": 1.0198,
+      "step": 10896
+    },
+    {
+      "epoch": 0.6620693845312595,
+      "grad_norm": 0.1871628314256668,
+      "learning_rate": 2.577770809210508e-05,
+      "loss": 1.1044,
+      "step": 10897
+    },
+    {
+      "epoch": 0.662130141563886,
+      "grad_norm": 0.1724691241979599,
+      "learning_rate": 2.5769334070765856e-05,
+      "loss": 1.0182,
+      "step": 10898
+    },
+    {
+      "epoch": 0.6621908985965126,
+      "grad_norm": 0.1359548419713974,
+      "learning_rate": 2.5760960937606353e-05,
+      "loss": 1.0623,
+      "step": 10899
+    },
+    {
+      "epoch": 0.6622516556291391,
+      "grad_norm": 0.20106449723243713,
+      "learning_rate": 2.57525886929335e-05,
+      "loss": 1.2013,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6623124126617655,
+      "grad_norm": 0.2684769928455353,
+      "learning_rate": 2.574421733705421e-05,
+      "loss": 1.0162,
+      "step": 10901
+    },
+    {
+      "epoch": 0.6623731696943921,
+      "grad_norm": 0.12225805222988129,
+      "learning_rate": 2.5735846870275322e-05,
+      "loss": 0.9762,
+      "step": 10902
+    },
+    {
+      "epoch": 0.6624339267270186,
+      "grad_norm": 0.1472126990556717,
+      "learning_rate": 2.5727477292903655e-05,
+      "loss": 1.1052,
+      "step": 10903
+    },
+    {
+      "epoch": 0.6624946837596452,
+      "grad_norm": 2.2278921604156494,
+      "learning_rate": 2.5719108605246e-05,
+      "loss": 1.0728,
+      "step": 10904
+    },
+    {
+      "epoch": 0.6625554407922717,
+      "grad_norm": 0.2266005277633667,
+      "learning_rate": 2.5710740807609117e-05,
+      "loss": 1.1619,
+      "step": 10905
+    },
+    {
+      "epoch": 0.6626161978248982,
+      "grad_norm": 0.998564600944519,
+      "learning_rate": 2.5702373900299725e-05,
+      "loss": 1.3294,
+      "step": 10906
+    },
+    {
+      "epoch": 0.6626769548575248,
+      "grad_norm": 0.15773192048072815,
+      "learning_rate": 2.569400788362451e-05,
+      "loss": 1.0401,
+      "step": 10907
+    },
+    {
+      "epoch": 0.6627377118901513,
+      "grad_norm": 0.11532074213027954,
+      "learning_rate": 2.5685642757890138e-05,
+      "loss": 0.9735,
+      "step": 10908
+    },
+    {
+      "epoch": 0.6627984689227778,
+      "grad_norm": 4.395709991455078,
+      "learning_rate": 2.5677278523403226e-05,
+      "loss": 1.1131,
+      "step": 10909
+    },
+    {
+      "epoch": 0.6628592259554044,
+      "grad_norm": 0.2470182329416275,
+      "learning_rate": 2.5668915180470376e-05,
+      "loss": 1.1702,
+      "step": 10910
+    },
+    {
+      "epoch": 0.6629199829880309,
+      "grad_norm": 0.17347879707813263,
+      "learning_rate": 2.5660552729398118e-05,
+      "loss": 1.1674,
+      "step": 10911
+    },
+    {
+      "epoch": 0.6629807400206574,
+      "grad_norm": 0.20448634028434753,
+      "learning_rate": 2.5652191170493045e-05,
+      "loss": 1.1145,
+      "step": 10912
+    },
+    {
+      "epoch": 0.663041497053284,
+      "grad_norm": 0.13824328780174255,
+      "learning_rate": 2.5643830504061594e-05,
+      "loss": 1.0981,
+      "step": 10913
+    },
+    {
+      "epoch": 0.6631022540859104,
+      "grad_norm": 0.2224777191877365,
+      "learning_rate": 2.5635470730410243e-05,
+      "loss": 1.0506,
+      "step": 10914
+    },
+    {
+      "epoch": 0.6631630111185369,
+      "grad_norm": 0.17914369702339172,
+      "learning_rate": 2.562711184984542e-05,
+      "loss": 1.0237,
+      "step": 10915
+    },
+    {
+      "epoch": 0.6632237681511635,
+      "grad_norm": 0.16801179945468903,
+      "learning_rate": 2.5618753862673506e-05,
+      "loss": 1.1143,
+      "step": 10916
+    },
+    {
+      "epoch": 0.66328452518379,
+      "grad_norm": 0.13636700809001923,
+      "learning_rate": 2.5610396769200906e-05,
+      "loss": 1.0748,
+      "step": 10917
+    },
+    {
+      "epoch": 0.6633452822164165,
+      "grad_norm": 0.17905090749263763,
+      "learning_rate": 2.5602040569733936e-05,
+      "loss": 1.1598,
+      "step": 10918
+    },
+    {
+      "epoch": 0.6634060392490431,
+      "grad_norm": 0.14596080780029297,
+      "learning_rate": 2.559368526457887e-05,
+      "loss": 1.0201,
+      "step": 10919
+    },
+    {
+      "epoch": 0.6634667962816696,
+      "grad_norm": 0.31742027401924133,
+      "learning_rate": 2.5585330854041988e-05,
+      "loss": 1.0801,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6635275533142961,
+      "grad_norm": 0.33845844864845276,
+      "learning_rate": 2.557697733842951e-05,
+      "loss": 1.1417,
+      "step": 10921
+    },
+    {
+      "epoch": 0.6635883103469227,
+      "grad_norm": 0.12948045134544373,
+      "learning_rate": 2.556862471804766e-05,
+      "loss": 1.0289,
+      "step": 10922
+    },
+    {
+      "epoch": 0.6636490673795492,
+      "grad_norm": 0.15239079296588898,
+      "learning_rate": 2.5560272993202594e-05,
+      "loss": 1.0488,
+      "step": 10923
+    },
+    {
+      "epoch": 0.6637098244121757,
+      "grad_norm": 0.3333180844783783,
+      "learning_rate": 2.555192216420045e-05,
+      "loss": 1.2639,
+      "step": 10924
+    },
+    {
+      "epoch": 0.6637705814448023,
+      "grad_norm": 0.13333414494991302,
+      "learning_rate": 2.5543572231347345e-05,
+      "loss": 1.066,
+      "step": 10925
+    },
+    {
+      "epoch": 0.6638313384774288,
+      "grad_norm": 0.2191387116909027,
+      "learning_rate": 2.553522319494928e-05,
+      "loss": 1.0223,
+      "step": 10926
+    },
+    {
+      "epoch": 0.6638920955100552,
+      "grad_norm": 0.10697758942842484,
+      "learning_rate": 2.552687505531236e-05,
+      "loss": 1.0463,
+      "step": 10927
+    },
+    {
+      "epoch": 0.6639528525426818,
+      "grad_norm": 0.12608550488948822,
+      "learning_rate": 2.551852781274257e-05,
+      "loss": 1.039,
+      "step": 10928
+    },
+    {
+      "epoch": 0.6640136095753083,
+      "grad_norm": 0.16342447698116302,
+      "learning_rate": 2.5510181467545873e-05,
+      "loss": 1.0519,
+      "step": 10929
+    },
+    {
+      "epoch": 0.6640743666079348,
+      "grad_norm": 0.16439864039421082,
+      "learning_rate": 2.5501836020028208e-05,
+      "loss": 1.0758,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6641351236405614,
+      "grad_norm": 0.1616610437631607,
+      "learning_rate": 2.549349147049548e-05,
+      "loss": 1.0354,
+      "step": 10931
+    },
+    {
+      "epoch": 0.6641958806731879,
+      "grad_norm": 0.17169110476970673,
+      "learning_rate": 2.5485147819253556e-05,
+      "loss": 1.0739,
+      "step": 10932
+    },
+    {
+      "epoch": 0.6642566377058144,
+      "grad_norm": 0.11216206848621368,
+      "learning_rate": 2.547680506660828e-05,
+      "loss": 1.0598,
+      "step": 10933
+    },
+    {
+      "epoch": 0.664317394738441,
+      "grad_norm": 0.265135258436203,
+      "learning_rate": 2.5468463212865455e-05,
+      "loss": 1.21,
+      "step": 10934
+    },
+    {
+      "epoch": 0.6643781517710675,
+      "grad_norm": 0.1278349906206131,
+      "learning_rate": 2.5460122258330853e-05,
+      "loss": 1.0443,
+      "step": 10935
+    },
+    {
+      "epoch": 0.664438908803694,
+      "grad_norm": 0.1615898460149765,
+      "learning_rate": 2.5451782203310208e-05,
+      "loss": 1.1094,
+      "step": 10936
+    },
+    {
+      "epoch": 0.6644996658363206,
+      "grad_norm": 0.1697765439748764,
+      "learning_rate": 2.5443443048109217e-05,
+      "loss": 1.0075,
+      "step": 10937
+    },
+    {
+      "epoch": 0.6645604228689471,
+      "grad_norm": 0.2195979505777359,
+      "learning_rate": 2.5435104793033605e-05,
+      "loss": 1.0721,
+      "step": 10938
+    },
+    {
+      "epoch": 0.6646211799015737,
+      "grad_norm": 0.25098177790641785,
+      "learning_rate": 2.5426767438388954e-05,
+      "loss": 1.1371,
+      "step": 10939
+    },
+    {
+      "epoch": 0.6646819369342001,
+      "grad_norm": 0.3639383316040039,
+      "learning_rate": 2.54184309844809e-05,
+      "loss": 1.1418,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6647426939668266,
+      "grad_norm": 0.18113672733306885,
+      "learning_rate": 2.5410095431615e-05,
+      "loss": 1.1016,
+      "step": 10941
+    },
+    {
+      "epoch": 0.6648034509994531,
+      "grad_norm": 0.1271345317363739,
+      "learning_rate": 2.5401760780096794e-05,
+      "loss": 1.036,
+      "step": 10942
+    },
+    {
+      "epoch": 0.6648642080320797,
+      "grad_norm": 0.19353890419006348,
+      "learning_rate": 2.5393427030231816e-05,
+      "loss": 1.1762,
+      "step": 10943
+    },
+    {
+      "epoch": 0.6649249650647062,
+      "grad_norm": 0.14769570529460907,
+      "learning_rate": 2.5385094182325525e-05,
+      "loss": 1.1017,
+      "step": 10944
+    },
+    {
+      "epoch": 0.6649857220973328,
+      "grad_norm": 0.13683219254016876,
+      "learning_rate": 2.537676223668338e-05,
+      "loss": 1.052,
+      "step": 10945
+    },
+    {
+      "epoch": 0.6650464791299593,
+      "grad_norm": 0.10853105783462524,
+      "learning_rate": 2.5368431193610763e-05,
+      "loss": 1.0403,
+      "step": 10946
+    },
+    {
+      "epoch": 0.6651072361625858,
+      "grad_norm": 0.18989138305187225,
+      "learning_rate": 2.5360101053413045e-05,
+      "loss": 1.078,
+      "step": 10947
+    },
+    {
+      "epoch": 0.6651679931952124,
+      "grad_norm": 0.12131781131029129,
+      "learning_rate": 2.5351771816395597e-05,
+      "loss": 1.0689,
+      "step": 10948
+    },
+    {
+      "epoch": 0.6652287502278389,
+      "grad_norm": 0.11907931417226791,
+      "learning_rate": 2.5343443482863728e-05,
+      "loss": 1.0483,
+      "step": 10949
+    },
+    {
+      "epoch": 0.6652895072604654,
+      "grad_norm": 0.10854718089103699,
+      "learning_rate": 2.53351160531227e-05,
+      "loss": 0.9932,
+      "step": 10950
+    },
+    {
+      "epoch": 0.665350264293092,
+      "grad_norm": 0.28004562854766846,
+      "learning_rate": 2.5326789527477756e-05,
+      "loss": 1.1395,
+      "step": 10951
+    },
+    {
+      "epoch": 0.6654110213257185,
+      "grad_norm": 0.20339399576187134,
+      "learning_rate": 2.5318463906234113e-05,
+      "loss": 1.1368,
+      "step": 10952
+    },
+    {
+      "epoch": 0.6654717783583449,
+      "grad_norm": 0.37540802359580994,
+      "learning_rate": 2.5310139189696945e-05,
+      "loss": 1.2658,
+      "step": 10953
+    },
+    {
+      "epoch": 0.6655325353909715,
+      "grad_norm": 0.13918866217136383,
+      "learning_rate": 2.53018153781714e-05,
+      "loss": 0.9907,
+      "step": 10954
+    },
+    {
+      "epoch": 0.665593292423598,
+      "grad_norm": 0.3053690195083618,
+      "learning_rate": 2.5293492471962587e-05,
+      "loss": 1.1532,
+      "step": 10955
+    },
+    {
+      "epoch": 0.6656540494562245,
+      "grad_norm": 0.24435748159885406,
+      "learning_rate": 2.528517047137558e-05,
+      "loss": 1.1165,
+      "step": 10956
+    },
+    {
+      "epoch": 0.6657148064888511,
+      "grad_norm": 0.2918207347393036,
+      "learning_rate": 2.527684937671543e-05,
+      "loss": 1.2466,
+      "step": 10957
+    },
+    {
+      "epoch": 0.6657755635214776,
+      "grad_norm": 0.10876744985580444,
+      "learning_rate": 2.5268529188287137e-05,
+      "loss": 1.0431,
+      "step": 10958
+    },
+    {
+      "epoch": 0.6658363205541041,
+      "grad_norm": 3.3927807807922363,
+      "learning_rate": 2.52602099063957e-05,
+      "loss": 1.0761,
+      "step": 10959
+    },
+    {
+      "epoch": 0.6658970775867307,
+      "grad_norm": 0.20972682535648346,
+      "learning_rate": 2.5251891531346038e-05,
+      "loss": 1.0863,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6659578346193572,
+      "grad_norm": 0.2238680124282837,
+      "learning_rate": 2.524357406344309e-05,
+      "loss": 1.0027,
+      "step": 10961
+    },
+    {
+      "epoch": 0.6660185916519837,
+      "grad_norm": 1.1363712549209595,
+      "learning_rate": 2.523525750299169e-05,
+      "loss": 1.1623,
+      "step": 10962
+    },
+    {
+      "epoch": 0.6660793486846103,
+      "grad_norm": 0.13125662505626678,
+      "learning_rate": 2.5226941850296737e-05,
+      "loss": 1.1134,
+      "step": 10963
+    },
+    {
+      "epoch": 0.6661401057172368,
+      "grad_norm": 0.10065761208534241,
+      "learning_rate": 2.5218627105663017e-05,
+      "loss": 1.0206,
+      "step": 10964
+    },
+    {
+      "epoch": 0.6662008627498633,
+      "grad_norm": 0.16558073461055756,
+      "learning_rate": 2.5210313269395326e-05,
+      "loss": 1.0784,
+      "step": 10965
+    },
+    {
+      "epoch": 0.6662616197824899,
+      "grad_norm": 0.16034385561943054,
+      "learning_rate": 2.520200034179838e-05,
+      "loss": 1.0932,
+      "step": 10966
+    },
+    {
+      "epoch": 0.6663223768151163,
+      "grad_norm": 0.13057871162891388,
+      "learning_rate": 2.5193688323176882e-05,
+      "loss": 1.0698,
+      "step": 10967
+    },
+    {
+      "epoch": 0.6663831338477428,
+      "grad_norm": 0.10659851133823395,
+      "learning_rate": 2.518537721383556e-05,
+      "loss": 1.0417,
+      "step": 10968
+    },
+    {
+      "epoch": 0.6664438908803694,
+      "grad_norm": 0.1049768477678299,
+      "learning_rate": 2.5177067014079026e-05,
+      "loss": 1.0249,
+      "step": 10969
+    },
+    {
+      "epoch": 0.6665046479129959,
+      "grad_norm": 0.1295790672302246,
+      "learning_rate": 2.51687577242119e-05,
+      "loss": 1.0233,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6665654049456224,
+      "grad_norm": 0.140318363904953,
+      "learning_rate": 2.516044934453876e-05,
+      "loss": 1.0856,
+      "step": 10971
+    },
+    {
+      "epoch": 0.666626161978249,
+      "grad_norm": 0.10695676505565643,
+      "learning_rate": 2.5152141875364148e-05,
+      "loss": 1.0294,
+      "step": 10972
+    },
+    {
+      "epoch": 0.6666869190108755,
+      "grad_norm": 0.14935553073883057,
+      "learning_rate": 2.514383531699258e-05,
+      "loss": 1.0153,
+      "step": 10973
+    },
+    {
+      "epoch": 0.666747676043502,
+      "grad_norm": 0.15685604512691498,
+      "learning_rate": 2.5135529669728537e-05,
+      "loss": 1.1119,
+      "step": 10974
+    },
+    {
+      "epoch": 0.6668084330761286,
+      "grad_norm": 0.13498510420322418,
+      "learning_rate": 2.512722493387646e-05,
+      "loss": 1.0648,
+      "step": 10975
+    },
+    {
+      "epoch": 0.6668691901087551,
+      "grad_norm": 0.1312147080898285,
+      "learning_rate": 2.5118921109740757e-05,
+      "loss": 1.0833,
+      "step": 10976
+    },
+    {
+      "epoch": 0.6669299471413817,
+      "grad_norm": 0.09532418102025986,
+      "learning_rate": 2.5110618197625806e-05,
+      "loss": 1.0211,
+      "step": 10977
+    },
+    {
+      "epoch": 0.6669907041740082,
+      "grad_norm": 0.126339852809906,
+      "learning_rate": 2.510231619783594e-05,
+      "loss": 1.0884,
+      "step": 10978
+    },
+    {
+      "epoch": 0.6670514612066347,
+      "grad_norm": 0.14915476739406586,
+      "learning_rate": 2.509401511067553e-05,
+      "loss": 1.0718,
+      "step": 10979
+    },
+    {
+      "epoch": 0.6671122182392611,
+      "grad_norm": 0.5402998924255371,
+      "learning_rate": 2.508571493644879e-05,
+      "loss": 1.0478,
+      "step": 10980
+    },
+    {
+      "epoch": 0.6671729752718877,
+      "grad_norm": 0.3424145579338074,
+      "learning_rate": 2.5077415675459982e-05,
+      "loss": 1.0312,
+      "step": 10981
+    },
+    {
+      "epoch": 0.6672337323045142,
+      "grad_norm": 0.3049122393131256,
+      "learning_rate": 2.506911732801332e-05,
+      "loss": 1.1784,
+      "step": 10982
+    },
+    {
+      "epoch": 0.6672944893371408,
+      "grad_norm": 0.25699368119239807,
+      "learning_rate": 2.5060819894412963e-05,
+      "loss": 1.1183,
+      "step": 10983
+    },
+    {
+      "epoch": 0.6673552463697673,
+      "grad_norm": 1.1047624349594116,
+      "learning_rate": 2.505252337496309e-05,
+      "loss": 1.0959,
+      "step": 10984
+    },
+    {
+      "epoch": 0.6674160034023938,
+      "grad_norm": 0.1271514594554901,
+      "learning_rate": 2.504422776996782e-05,
+      "loss": 1.053,
+      "step": 10985
+    },
+    {
+      "epoch": 0.6674767604350204,
+      "grad_norm": 0.17269587516784668,
+      "learning_rate": 2.5035933079731176e-05,
+      "loss": 1.0488,
+      "step": 10986
+    },
+    {
+      "epoch": 0.6675375174676469,
+      "grad_norm": 0.21040619909763336,
+      "learning_rate": 2.5027639304557236e-05,
+      "loss": 1.1364,
+      "step": 10987
+    },
+    {
+      "epoch": 0.6675982745002734,
+      "grad_norm": 0.1320820450782776,
+      "learning_rate": 2.5019346444749992e-05,
+      "loss": 1.0364,
+      "step": 10988
+    },
+    {
+      "epoch": 0.6676590315329,
+      "grad_norm": 0.11433776468038559,
+      "learning_rate": 2.5011054500613446e-05,
+      "loss": 1.0156,
+      "step": 10989
+    },
+    {
+      "epoch": 0.6677197885655265,
+      "grad_norm": 0.13606613874435425,
+      "learning_rate": 2.5002763472451528e-05,
+      "loss": 1.0408,
+      "step": 10990
+    },
+    {
+      "epoch": 0.667780545598153,
+      "grad_norm": 0.14474043250083923,
+      "learning_rate": 2.4994473360568145e-05,
+      "loss": 1.0466,
+      "step": 10991
+    },
+    {
+      "epoch": 0.6678413026307796,
+      "grad_norm": 0.20358166098594666,
+      "learning_rate": 2.4986184165267195e-05,
+      "loss": 1.0616,
+      "step": 10992
+    },
+    {
+      "epoch": 0.667902059663406,
+      "grad_norm": 0.11155235022306442,
+      "learning_rate": 2.497789588685246e-05,
+      "loss": 1.0487,
+      "step": 10993
+    },
+    {
+      "epoch": 0.6679628166960325,
+      "grad_norm": 0.18793247640132904,
+      "learning_rate": 2.4969608525627807e-05,
+      "loss": 1.0539,
+      "step": 10994
+    },
+    {
+      "epoch": 0.6680235737286591,
+      "grad_norm": 0.1075793206691742,
+      "learning_rate": 2.4961322081896994e-05,
+      "loss": 1.068,
+      "step": 10995
+    },
+    {
+      "epoch": 0.6680843307612856,
+      "grad_norm": 1.0784072875976562,
+      "learning_rate": 2.4953036555963754e-05,
+      "loss": 1.0165,
+      "step": 10996
+    },
+    {
+      "epoch": 0.6681450877939121,
+      "grad_norm": 0.17310203611850739,
+      "learning_rate": 2.4944751948131806e-05,
+      "loss": 1.0761,
+      "step": 10997
+    },
+    {
+      "epoch": 0.6682058448265387,
+      "grad_norm": 0.17463995516300201,
+      "learning_rate": 2.493646825870482e-05,
+      "loss": 1.0954,
+      "step": 10998
+    },
+    {
+      "epoch": 0.6682666018591652,
+      "grad_norm": 0.13798128068447113,
+      "learning_rate": 2.4928185487986433e-05,
+      "loss": 1.0186,
+      "step": 10999
+    },
+    {
+      "epoch": 0.6683273588917917,
+      "grad_norm": 0.12489543855190277,
+      "learning_rate": 2.491990363628025e-05,
+      "loss": 1.113,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6683881159244183,
+      "grad_norm": 0.30142515897750854,
+      "learning_rate": 2.4911622703889848e-05,
+      "loss": 1.1783,
+      "step": 11001
+    },
+    {
+      "epoch": 0.6684488729570448,
+      "grad_norm": 0.7820678353309631,
+      "learning_rate": 2.490334269111877e-05,
+      "loss": 1.1694,
+      "step": 11002
+    },
+    {
+      "epoch": 0.6685096299896713,
+      "grad_norm": 0.45104581117630005,
+      "learning_rate": 2.489506359827051e-05,
+      "loss": 1.1753,
+      "step": 11003
+    },
+    {
+      "epoch": 0.6685703870222979,
+      "grad_norm": 0.20650026202201843,
+      "learning_rate": 2.4886785425648533e-05,
+      "loss": 1.0191,
+      "step": 11004
+    },
+    {
+      "epoch": 0.6686311440549244,
+      "grad_norm": 0.18666957318782806,
+      "learning_rate": 2.4878508173556327e-05,
+      "loss": 1.0125,
+      "step": 11005
+    },
+    {
+      "epoch": 0.6686919010875508,
+      "grad_norm": 0.43002381920814514,
+      "learning_rate": 2.4870231842297236e-05,
+      "loss": 1.1658,
+      "step": 11006
+    },
+    {
+      "epoch": 0.6687526581201774,
+      "grad_norm": 0.16620266437530518,
+      "learning_rate": 2.486195643217466e-05,
+      "loss": 1.0975,
+      "step": 11007
+    },
+    {
+      "epoch": 0.6688134151528039,
+      "grad_norm": 0.21435941755771637,
+      "learning_rate": 2.485368194349193e-05,
+      "loss": 1.0637,
+      "step": 11008
+    },
+    {
+      "epoch": 0.6688741721854304,
+      "grad_norm": 0.18640761077404022,
+      "learning_rate": 2.4845408376552325e-05,
+      "loss": 1.0572,
+      "step": 11009
+    },
+    {
+      "epoch": 0.668934929218057,
+      "grad_norm": 0.1668463796377182,
+      "learning_rate": 2.4837135731659168e-05,
+      "loss": 1.2412,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6689956862506835,
+      "grad_norm": 0.14010213315486908,
+      "learning_rate": 2.4828864009115654e-05,
+      "loss": 1.058,
+      "step": 11011
+    },
+    {
+      "epoch": 0.66905644328331,
+      "grad_norm": 0.18175546824932098,
+      "learning_rate": 2.4820593209225017e-05,
+      "loss": 1.1308,
+      "step": 11012
+    },
+    {
+      "epoch": 0.6691172003159366,
+      "grad_norm": 0.2023133635520935,
+      "learning_rate": 2.4812323332290384e-05,
+      "loss": 1.0996,
+      "step": 11013
+    },
+    {
+      "epoch": 0.6691779573485631,
+      "grad_norm": 0.11518340557813644,
+      "learning_rate": 2.4804054378614887e-05,
+      "loss": 1.0337,
+      "step": 11014
+    },
+    {
+      "epoch": 0.6692387143811896,
+      "grad_norm": 0.10573253035545349,
+      "learning_rate": 2.479578634850167e-05,
+      "loss": 1.0204,
+      "step": 11015
+    },
+    {
+      "epoch": 0.6692994714138162,
+      "grad_norm": 0.12148343026638031,
+      "learning_rate": 2.4787519242253776e-05,
+      "loss": 1.0455,
+      "step": 11016
+    },
+    {
+      "epoch": 0.6693602284464427,
+      "grad_norm": 0.3998861312866211,
+      "learning_rate": 2.4779253060174233e-05,
+      "loss": 1.069,
+      "step": 11017
+    },
+    {
+      "epoch": 0.6694209854790693,
+      "grad_norm": 0.14318203926086426,
+      "learning_rate": 2.4770987802566054e-05,
+      "loss": 1.0613,
+      "step": 11018
+    },
+    {
+      "epoch": 0.6694817425116957,
+      "grad_norm": 0.15260010957717896,
+      "learning_rate": 2.4762723469732186e-05,
+      "loss": 1.0182,
+      "step": 11019
+    },
+    {
+      "epoch": 0.6695424995443222,
+      "grad_norm": 0.18030261993408203,
+      "learning_rate": 2.475446006197557e-05,
+      "loss": 1.0358,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6696032565769487,
+      "grad_norm": 0.18028858304023743,
+      "learning_rate": 2.4746197579599095e-05,
+      "loss": 1.1461,
+      "step": 11021
+    },
+    {
+      "epoch": 0.6696640136095753,
+      "grad_norm": 0.14516298472881317,
+      "learning_rate": 2.4737936022905632e-05,
+      "loss": 1.1109,
+      "step": 11022
+    },
+    {
+      "epoch": 0.6697247706422018,
+      "grad_norm": 0.15879233181476593,
+      "learning_rate": 2.472967539219801e-05,
+      "loss": 1.067,
+      "step": 11023
+    },
+    {
+      "epoch": 0.6697855276748284,
+      "grad_norm": 0.16960769891738892,
+      "learning_rate": 2.472141568777902e-05,
+      "loss": 1.0521,
+      "step": 11024
+    },
+    {
+      "epoch": 0.6698462847074549,
+      "grad_norm": 0.16727851331233978,
+      "learning_rate": 2.4713156909951424e-05,
+      "loss": 1.0142,
+      "step": 11025
+    },
+    {
+      "epoch": 0.6699070417400814,
+      "grad_norm": 0.20663772523403168,
+      "learning_rate": 2.470489905901795e-05,
+      "loss": 1.1121,
+      "step": 11026
+    },
+    {
+      "epoch": 0.669967798772708,
+      "grad_norm": 0.1411844938993454,
+      "learning_rate": 2.469664213528129e-05,
+      "loss": 1.053,
+      "step": 11027
+    },
+    {
+      "epoch": 0.6700285558053345,
+      "grad_norm": 0.13510943949222565,
+      "learning_rate": 2.46883861390441e-05,
+      "loss": 1.0884,
+      "step": 11028
+    },
+    {
+      "epoch": 0.670089312837961,
+      "grad_norm": 0.16399888694286346,
+      "learning_rate": 2.4680131070609015e-05,
+      "loss": 1.0639,
+      "step": 11029
+    },
+    {
+      "epoch": 0.6701500698705876,
+      "grad_norm": 0.22690443694591522,
+      "learning_rate": 2.4671876930278598e-05,
+      "loss": 1.1131,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6702108269032141,
+      "grad_norm": 0.12540382146835327,
+      "learning_rate": 2.4663623718355444e-05,
+      "loss": 1.1023,
+      "step": 11031
+    },
+    {
+      "epoch": 0.6702715839358405,
+      "grad_norm": 0.18903689086437225,
+      "learning_rate": 2.4655371435142084e-05,
+      "loss": 1.103,
+      "step": 11032
+    },
+    {
+      "epoch": 0.670332340968467,
+      "grad_norm": 0.11257229000329971,
+      "learning_rate": 2.4647120080940962e-05,
+      "loss": 1.0481,
+      "step": 11033
+    },
+    {
+      "epoch": 0.6703930980010936,
+      "grad_norm": 0.14155890047550201,
+      "learning_rate": 2.4638869656054554e-05,
+      "loss": 1.1018,
+      "step": 11034
+    },
+    {
+      "epoch": 0.6704538550337201,
+      "grad_norm": 0.2668917179107666,
+      "learning_rate": 2.463062016078526e-05,
+      "loss": 1.1144,
+      "step": 11035
+    },
+    {
+      "epoch": 0.6705146120663467,
+      "grad_norm": 0.10257581621408463,
+      "learning_rate": 2.4622371595435506e-05,
+      "loss": 1.0188,
+      "step": 11036
+    },
+    {
+      "epoch": 0.6705753690989732,
+      "grad_norm": 0.2013516128063202,
+      "learning_rate": 2.4614123960307634e-05,
+      "loss": 1.1496,
+      "step": 11037
+    },
+    {
+      "epoch": 0.6706361261315997,
+      "grad_norm": 1.106760859489441,
+      "learning_rate": 2.4605877255703947e-05,
+      "loss": 1.0553,
+      "step": 11038
+    },
+    {
+      "epoch": 0.6706968831642263,
+      "grad_norm": 0.2565065324306488,
+      "learning_rate": 2.459763148192676e-05,
+      "loss": 1.029,
+      "step": 11039
+    },
+    {
+      "epoch": 0.6707576401968528,
+      "grad_norm": 0.16055181622505188,
+      "learning_rate": 2.458938663927826e-05,
+      "loss": 0.9793,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6708183972294793,
+      "grad_norm": 0.2202807515859604,
+      "learning_rate": 2.458114272806073e-05,
+      "loss": 1.0342,
+      "step": 11041
+    },
+    {
+      "epoch": 0.6708791542621059,
+      "grad_norm": 0.13346879184246063,
+      "learning_rate": 2.457289974857632e-05,
+      "loss": 1.0884,
+      "step": 11042
+    },
+    {
+      "epoch": 0.6709399112947324,
+      "grad_norm": 0.1459265798330307,
+      "learning_rate": 2.4564657701127193e-05,
+      "loss": 1.0081,
+      "step": 11043
+    },
+    {
+      "epoch": 0.6710006683273589,
+      "grad_norm": 0.15224696695804596,
+      "learning_rate": 2.4556416586015446e-05,
+      "loss": 1.0997,
+      "step": 11044
+    },
+    {
+      "epoch": 0.6710614253599854,
+      "grad_norm": 0.22559818625450134,
+      "learning_rate": 2.454817640354317e-05,
+      "loss": 1.108,
+      "step": 11045
+    },
+    {
+      "epoch": 0.6711221823926119,
+      "grad_norm": 0.3387386202812195,
+      "learning_rate": 2.453993715401241e-05,
+      "loss": 1.1044,
+      "step": 11046
+    },
+    {
+      "epoch": 0.6711829394252384,
+      "grad_norm": 0.19596999883651733,
+      "learning_rate": 2.4531698837725175e-05,
+      "loss": 1.1141,
+      "step": 11047
+    },
+    {
+      "epoch": 0.671243696457865,
+      "grad_norm": 0.12199502438306808,
+      "learning_rate": 2.4523461454983436e-05,
+      "loss": 1.0643,
+      "step": 11048
+    },
+    {
+      "epoch": 0.6713044534904915,
+      "grad_norm": 0.13556912541389465,
+      "learning_rate": 2.4515225006089147e-05,
+      "loss": 1.0302,
+      "step": 11049
+    },
+    {
+      "epoch": 0.671365210523118,
+      "grad_norm": 0.19067586958408356,
+      "learning_rate": 2.4506989491344213e-05,
+      "loss": 1.1643,
+      "step": 11050
+    },
+    {
+      "epoch": 0.6714259675557446,
+      "grad_norm": 0.1274019181728363,
+      "learning_rate": 2.4498754911050485e-05,
+      "loss": 0.9875,
+      "step": 11051
+    },
+    {
+      "epoch": 0.6714867245883711,
+      "grad_norm": 0.1863957643508911,
+      "learning_rate": 2.4490521265509863e-05,
+      "loss": 1.1279,
+      "step": 11052
+    },
+    {
+      "epoch": 0.6715474816209976,
+      "grad_norm": 0.1458347737789154,
+      "learning_rate": 2.4482288555024097e-05,
+      "loss": 1.0337,
+      "step": 11053
+    },
+    {
+      "epoch": 0.6716082386536242,
+      "grad_norm": 0.2231624722480774,
+      "learning_rate": 2.4474056779894978e-05,
+      "loss": 1.1914,
+      "step": 11054
+    },
+    {
+      "epoch": 0.6716689956862507,
+      "grad_norm": 0.18422985076904297,
+      "learning_rate": 2.4465825940424236e-05,
+      "loss": 1.1243,
+      "step": 11055
+    },
+    {
+      "epoch": 0.6717297527188772,
+      "grad_norm": 0.178894504904747,
+      "learning_rate": 2.4457596036913567e-05,
+      "loss": 1.0835,
+      "step": 11056
+    },
+    {
+      "epoch": 0.6717905097515038,
+      "grad_norm": 0.15781378746032715,
+      "learning_rate": 2.444936706966467e-05,
+      "loss": 1.0731,
+      "step": 11057
+    },
+    {
+      "epoch": 0.6718512667841302,
+      "grad_norm": 0.15718761086463928,
+      "learning_rate": 2.444113903897916e-05,
+      "loss": 1.039,
+      "step": 11058
+    },
+    {
+      "epoch": 0.6719120238167567,
+      "grad_norm": 0.1350620836019516,
+      "learning_rate": 2.4432911945158653e-05,
+      "loss": 1.0101,
+      "step": 11059
+    },
+    {
+      "epoch": 0.6719727808493833,
+      "grad_norm": 0.355625718832016,
+      "learning_rate": 2.442468578850467e-05,
+      "loss": 1.179,
+      "step": 11060
+    },
+    {
+      "epoch": 0.6720335378820098,
+      "grad_norm": 0.1486140340566635,
+      "learning_rate": 2.4416460569318796e-05,
+      "loss": 1.0926,
+      "step": 11061
+    },
+    {
+      "epoch": 0.6720942949146363,
+      "grad_norm": 0.2064017951488495,
+      "learning_rate": 2.4408236287902497e-05,
+      "loss": 1.0995,
+      "step": 11062
+    },
+    {
+      "epoch": 0.6721550519472629,
+      "grad_norm": 0.19788654148578644,
+      "learning_rate": 2.4400012944557244e-05,
+      "loss": 1.1173,
+      "step": 11063
+    },
+    {
+      "epoch": 0.6722158089798894,
+      "grad_norm": 0.2470221221446991,
+      "learning_rate": 2.439179053958447e-05,
+      "loss": 1.0547,
+      "step": 11064
+    },
+    {
+      "epoch": 0.672276566012516,
+      "grad_norm": 0.7126647233963013,
+      "learning_rate": 2.438356907328556e-05,
+      "loss": 1.2172,
+      "step": 11065
+    },
+    {
+      "epoch": 0.6723373230451425,
+      "grad_norm": 0.25403162837028503,
+      "learning_rate": 2.4375348545961878e-05,
+      "loss": 1.1078,
+      "step": 11066
+    },
+    {
+      "epoch": 0.672398080077769,
+      "grad_norm": 0.16500109434127808,
+      "learning_rate": 2.4367128957914746e-05,
+      "loss": 1.0563,
+      "step": 11067
+    },
+    {
+      "epoch": 0.6724588371103956,
+      "grad_norm": 0.1750573068857193,
+      "learning_rate": 2.435891030944546e-05,
+      "loss": 1.1769,
+      "step": 11068
+    },
+    {
+      "epoch": 0.6725195941430221,
+      "grad_norm": 0.14198659360408783,
+      "learning_rate": 2.4350692600855274e-05,
+      "loss": 0.985,
+      "step": 11069
+    },
+    {
+      "epoch": 0.6725803511756486,
+      "grad_norm": 0.15442423522472382,
+      "learning_rate": 2.4342475832445405e-05,
+      "loss": 1.0744,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6726411082082752,
+      "grad_norm": 0.41567301750183105,
+      "learning_rate": 2.4334260004517022e-05,
+      "loss": 1.2301,
+      "step": 11071
+    },
+    {
+      "epoch": 0.6727018652409016,
+      "grad_norm": 0.1515897661447525,
+      "learning_rate": 2.4326045117371343e-05,
+      "loss": 1.1096,
+      "step": 11072
+    },
+    {
+      "epoch": 0.6727626222735281,
+      "grad_norm": 0.9282751679420471,
+      "learning_rate": 2.4317831171309423e-05,
+      "loss": 1.0783,
+      "step": 11073
+    },
+    {
+      "epoch": 0.6728233793061547,
+      "grad_norm": 0.21398164331912994,
+      "learning_rate": 2.430961816663237e-05,
+      "loss": 1.0437,
+      "step": 11074
+    },
+    {
+      "epoch": 0.6728841363387812,
+      "grad_norm": 0.15392647683620453,
+      "learning_rate": 2.430140610364122e-05,
+      "loss": 1.0981,
+      "step": 11075
+    },
+    {
+      "epoch": 0.6729448933714077,
+      "grad_norm": 0.23070895671844482,
+      "learning_rate": 2.4293194982636986e-05,
+      "loss": 1.1094,
+      "step": 11076
+    },
+    {
+      "epoch": 0.6730056504040343,
+      "grad_norm": 0.13760994374752045,
+      "learning_rate": 2.4284984803920673e-05,
+      "loss": 1.0755,
+      "step": 11077
+    },
+    {
+      "epoch": 0.6730664074366608,
+      "grad_norm": 0.22460895776748657,
+      "learning_rate": 2.4276775567793214e-05,
+      "loss": 1.0105,
+      "step": 11078
+    },
+    {
+      "epoch": 0.6731271644692873,
+      "grad_norm": 0.11733175069093704,
+      "learning_rate": 2.4268567274555538e-05,
+      "loss": 1.0348,
+      "step": 11079
+    },
+    {
+      "epoch": 0.6731879215019139,
+      "grad_norm": 0.11407577246427536,
+      "learning_rate": 2.426035992450848e-05,
+      "loss": 0.9997,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6732486785345404,
+      "grad_norm": 0.15263628959655762,
+      "learning_rate": 2.4252153517952884e-05,
+      "loss": 1.1412,
+      "step": 11081
+    },
+    {
+      "epoch": 0.6733094355671669,
+      "grad_norm": 0.19868876039981842,
+      "learning_rate": 2.4243948055189602e-05,
+      "loss": 1.1893,
+      "step": 11082
+    },
+    {
+      "epoch": 0.6733701925997935,
+      "grad_norm": 0.4263297915458679,
+      "learning_rate": 2.4235743536519383e-05,
+      "loss": 1.1724,
+      "step": 11083
+    },
+    {
+      "epoch": 0.67343094963242,
+      "grad_norm": 0.2823430001735687,
+      "learning_rate": 2.422753996224296e-05,
+      "loss": 1.0477,
+      "step": 11084
+    },
+    {
+      "epoch": 0.6734917066650464,
+      "grad_norm": 0.13645730912685394,
+      "learning_rate": 2.4219337332661045e-05,
+      "loss": 1.1096,
+      "step": 11085
+    },
+    {
+      "epoch": 0.673552463697673,
+      "grad_norm": 0.28503063321113586,
+      "learning_rate": 2.421113564807431e-05,
+      "loss": 1.1504,
+      "step": 11086
+    },
+    {
+      "epoch": 0.6736132207302995,
+      "grad_norm": 0.17133508622646332,
+      "learning_rate": 2.420293490878337e-05,
+      "loss": 1.1179,
+      "step": 11087
+    },
+    {
+      "epoch": 0.673673977762926,
+      "grad_norm": 0.24579021334648132,
+      "learning_rate": 2.4194735115088846e-05,
+      "loss": 1.1998,
+      "step": 11088
+    },
+    {
+      "epoch": 0.6737347347955526,
+      "grad_norm": 0.12109159678220749,
+      "learning_rate": 2.4186536267291287e-05,
+      "loss": 1.0868,
+      "step": 11089
+    },
+    {
+      "epoch": 0.6737954918281791,
+      "grad_norm": 0.12620693445205688,
+      "learning_rate": 2.4178338365691238e-05,
+      "loss": 1.0801,
+      "step": 11090
+    },
+    {
+      "epoch": 0.6738562488608056,
+      "grad_norm": 0.24572831392288208,
+      "learning_rate": 2.4170141410589185e-05,
+      "loss": 1.2019,
+      "step": 11091
+    },
+    {
+      "epoch": 0.6739170058934322,
+      "grad_norm": 0.11499699205160141,
+      "learning_rate": 2.416194540228559e-05,
+      "loss": 1.0401,
+      "step": 11092
+    },
+    {
+      "epoch": 0.6739777629260587,
+      "grad_norm": 0.2524833381175995,
+      "learning_rate": 2.415375034108088e-05,
+      "loss": 1.3011,
+      "step": 11093
+    },
+    {
+      "epoch": 0.6740385199586852,
+      "grad_norm": 0.14241334795951843,
+      "learning_rate": 2.414555622727545e-05,
+      "loss": 1.0803,
+      "step": 11094
+    },
+    {
+      "epoch": 0.6740992769913118,
+      "grad_norm": 0.21915780007839203,
+      "learning_rate": 2.4137363061169656e-05,
+      "loss": 1.1197,
+      "step": 11095
+    },
+    {
+      "epoch": 0.6741600340239383,
+      "grad_norm": 0.11939509958028793,
+      "learning_rate": 2.4129170843063816e-05,
+      "loss": 1.0567,
+      "step": 11096
+    },
+    {
+      "epoch": 0.6742207910565648,
+      "grad_norm": 0.19560158252716064,
+      "learning_rate": 2.4120979573258195e-05,
+      "loss": 1.1072,
+      "step": 11097
+    },
+    {
+      "epoch": 0.6742815480891913,
+      "grad_norm": 2.686720848083496,
+      "learning_rate": 2.4112789252053092e-05,
+      "loss": 1.1111,
+      "step": 11098
+    },
+    {
+      "epoch": 0.6743423051218178,
+      "grad_norm": 0.22586117684841156,
+      "learning_rate": 2.4104599879748724e-05,
+      "loss": 1.1085,
+      "step": 11099
+    },
+    {
+      "epoch": 0.6744030621544443,
+      "grad_norm": 0.1262160837650299,
+      "learning_rate": 2.4096411456645235e-05,
+      "loss": 1.0541,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6744638191870709,
+      "grad_norm": 2.3152763843536377,
+      "learning_rate": 2.408822398304279e-05,
+      "loss": 1.0395,
+      "step": 11101
+    },
+    {
+      "epoch": 0.6745245762196974,
+      "grad_norm": 0.15275660157203674,
+      "learning_rate": 2.4080037459241482e-05,
+      "loss": 1.0939,
+      "step": 11102
+    },
+    {
+      "epoch": 0.674585333252324,
+      "grad_norm": 0.5373700857162476,
+      "learning_rate": 2.4071851885541435e-05,
+      "loss": 1.0866,
+      "step": 11103
+    },
+    {
+      "epoch": 0.6746460902849505,
+      "grad_norm": 0.13188719749450684,
+      "learning_rate": 2.4063667262242662e-05,
+      "loss": 1.1003,
+      "step": 11104
+    },
+    {
+      "epoch": 0.674706847317577,
+      "grad_norm": 0.17168322205543518,
+      "learning_rate": 2.4055483589645185e-05,
+      "loss": 1.0888,
+      "step": 11105
+    },
+    {
+      "epoch": 0.6747676043502036,
+      "grad_norm": 0.17305390536785126,
+      "learning_rate": 2.404730086804899e-05,
+      "loss": 1.0685,
+      "step": 11106
+    },
+    {
+      "epoch": 0.6748283613828301,
+      "grad_norm": 0.3638569712638855,
+      "learning_rate": 2.403911909775396e-05,
+      "loss": 1.0102,
+      "step": 11107
+    },
+    {
+      "epoch": 0.6748891184154566,
+      "grad_norm": 0.47309207916259766,
+      "learning_rate": 2.4030938279060054e-05,
+      "loss": 1.1301,
+      "step": 11108
+    },
+    {
+      "epoch": 0.6749498754480832,
+      "grad_norm": 0.1250740885734558,
+      "learning_rate": 2.4022758412267126e-05,
+      "loss": 1.0399,
+      "step": 11109
+    },
+    {
+      "epoch": 0.6750106324807097,
+      "grad_norm": 0.1532147377729416,
+      "learning_rate": 2.401457949767501e-05,
+      "loss": 1.0876,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6750713895133361,
+      "grad_norm": 0.23548004031181335,
+      "learning_rate": 2.4006401535583507e-05,
+      "loss": 0.9906,
+      "step": 11111
+    },
+    {
+      "epoch": 0.6751321465459627,
+      "grad_norm": 0.21655581891536713,
+      "learning_rate": 2.3998224526292377e-05,
+      "loss": 1.1653,
+      "step": 11112
+    },
+    {
+      "epoch": 0.6751929035785892,
+      "grad_norm": 0.14888663589954376,
+      "learning_rate": 2.3990048470101355e-05,
+      "loss": 1.1008,
+      "step": 11113
+    },
+    {
+      "epoch": 0.6752536606112157,
+      "grad_norm": 0.18984763324260712,
+      "learning_rate": 2.3981873367310136e-05,
+      "loss": 1.1244,
+      "step": 11114
+    },
+    {
+      "epoch": 0.6753144176438423,
+      "grad_norm": 0.1585441529750824,
+      "learning_rate": 2.3973699218218375e-05,
+      "loss": 1.113,
+      "step": 11115
+    },
+    {
+      "epoch": 0.6753751746764688,
+      "grad_norm": 0.11531566083431244,
+      "learning_rate": 2.39655260231257e-05,
+      "loss": 1.05,
+      "step": 11116
+    },
+    {
+      "epoch": 0.6754359317090953,
+      "grad_norm": 0.17235319316387177,
+      "learning_rate": 2.3957353782331706e-05,
+      "loss": 1.1019,
+      "step": 11117
+    },
+    {
+      "epoch": 0.6754966887417219,
+      "grad_norm": 0.12548208236694336,
+      "learning_rate": 2.3949182496135913e-05,
+      "loss": 1.018,
+      "step": 11118
+    },
+    {
+      "epoch": 0.6755574457743484,
+      "grad_norm": 0.14066247642040253,
+      "learning_rate": 2.3941012164837917e-05,
+      "loss": 1.0494,
+      "step": 11119
+    },
+    {
+      "epoch": 0.6756182028069749,
+      "grad_norm": 0.16103151440620422,
+      "learning_rate": 2.3932842788737135e-05,
+      "loss": 1.1113,
+      "step": 11120
+    },
+    {
+      "epoch": 0.6756789598396015,
+      "grad_norm": 0.2735982835292816,
+      "learning_rate": 2.3924674368133042e-05,
+      "loss": 1.113,
+      "step": 11121
+    },
+    {
+      "epoch": 0.675739716872228,
+      "grad_norm": 0.1598770171403885,
+      "learning_rate": 2.3916506903325053e-05,
+      "loss": 1.0469,
+      "step": 11122
+    },
+    {
+      "epoch": 0.6758004739048545,
+      "grad_norm": 0.20828813314437866,
+      "learning_rate": 2.390834039461252e-05,
+      "loss": 1.1276,
+      "step": 11123
+    },
+    {
+      "epoch": 0.675861230937481,
+      "grad_norm": 0.1409241110086441,
+      "learning_rate": 2.3900174842294836e-05,
+      "loss": 1.0763,
+      "step": 11124
+    },
+    {
+      "epoch": 0.6759219879701075,
+      "grad_norm": 0.12823322415351868,
+      "learning_rate": 2.3892010246671287e-05,
+      "loss": 1.0786,
+      "step": 11125
+    },
+    {
+      "epoch": 0.675982745002734,
+      "grad_norm": 0.16855323314666748,
+      "learning_rate": 2.388384660804117e-05,
+      "loss": 1.1017,
+      "step": 11126
+    },
+    {
+      "epoch": 0.6760435020353606,
+      "grad_norm": 0.16470292210578918,
+      "learning_rate": 2.3875683926703683e-05,
+      "loss": 1.0998,
+      "step": 11127
+    },
+    {
+      "epoch": 0.6761042590679871,
+      "grad_norm": 0.13120052218437195,
+      "learning_rate": 2.386752220295803e-05,
+      "loss": 1.073,
+      "step": 11128
+    },
+    {
+      "epoch": 0.6761650161006136,
+      "grad_norm": 0.31621286273002625,
+      "learning_rate": 2.3859361437103428e-05,
+      "loss": 1.2871,
+      "step": 11129
+    },
+    {
+      "epoch": 0.6762257731332402,
+      "grad_norm": 0.4286743998527527,
+      "learning_rate": 2.385120162943898e-05,
+      "loss": 1.0333,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6762865301658667,
+      "grad_norm": 0.16835333406925201,
+      "learning_rate": 2.384304278026379e-05,
+      "loss": 1.1153,
+      "step": 11131
+    },
+    {
+      "epoch": 0.6763472871984932,
+      "grad_norm": 0.12320169061422348,
+      "learning_rate": 2.3834884889876925e-05,
+      "loss": 1.0807,
+      "step": 11132
+    },
+    {
+      "epoch": 0.6764080442311198,
+      "grad_norm": 0.19194099307060242,
+      "learning_rate": 2.3826727958577412e-05,
+      "loss": 1.0582,
+      "step": 11133
+    },
+    {
+      "epoch": 0.6764688012637463,
+      "grad_norm": 0.11436954885721207,
+      "learning_rate": 2.381857198666424e-05,
+      "loss": 1.0233,
+      "step": 11134
+    },
+    {
+      "epoch": 0.6765295582963728,
+      "grad_norm": 0.1854429394006729,
+      "learning_rate": 2.3810416974436378e-05,
+      "loss": 1.1164,
+      "step": 11135
+    },
+    {
+      "epoch": 0.6765903153289994,
+      "grad_norm": 0.10896814614534378,
+      "learning_rate": 2.3802262922192742e-05,
+      "loss": 1.0463,
+      "step": 11136
+    },
+    {
+      "epoch": 0.6766510723616258,
+      "grad_norm": 0.15312983095645905,
+      "learning_rate": 2.3794109830232226e-05,
+      "loss": 1.1177,
+      "step": 11137
+    },
+    {
+      "epoch": 0.6767118293942523,
+      "grad_norm": 0.1960471272468567,
+      "learning_rate": 2.378595769885368e-05,
+      "loss": 1.1229,
+      "step": 11138
+    },
+    {
+      "epoch": 0.6767725864268789,
+      "grad_norm": 0.21762306988239288,
+      "learning_rate": 2.3777806528355918e-05,
+      "loss": 1.0554,
+      "step": 11139
+    },
+    {
+      "epoch": 0.6768333434595054,
+      "grad_norm": 0.15447430312633514,
+      "learning_rate": 2.3769656319037724e-05,
+      "loss": 1.0138,
+      "step": 11140
+    },
+    {
+      "epoch": 0.676894100492132,
+      "grad_norm": 0.13208043575286865,
+      "learning_rate": 2.3761507071197854e-05,
+      "loss": 1.0548,
+      "step": 11141
+    },
+    {
+      "epoch": 0.6769548575247585,
+      "grad_norm": 0.12053914368152618,
+      "learning_rate": 2.3753358785135005e-05,
+      "loss": 1.064,
+      "step": 11142
+    },
+    {
+      "epoch": 0.677015614557385,
+      "grad_norm": 0.2518937587738037,
+      "learning_rate": 2.374521146114787e-05,
+      "loss": 1.1704,
+      "step": 11143
+    },
+    {
+      "epoch": 0.6770763715900115,
+      "grad_norm": 0.14782080054283142,
+      "learning_rate": 2.3737065099535066e-05,
+      "loss": 1.0677,
+      "step": 11144
+    },
+    {
+      "epoch": 0.6771371286226381,
+      "grad_norm": 0.1666611135005951,
+      "learning_rate": 2.3728919700595233e-05,
+      "loss": 1.1756,
+      "step": 11145
+    },
+    {
+      "epoch": 0.6771978856552646,
+      "grad_norm": 0.11371220648288727,
+      "learning_rate": 2.372077526462695e-05,
+      "loss": 1.0639,
+      "step": 11146
+    },
+    {
+      "epoch": 0.6772586426878912,
+      "grad_norm": 0.15326513350009918,
+      "learning_rate": 2.3712631791928703e-05,
+      "loss": 1.1054,
+      "step": 11147
+    },
+    {
+      "epoch": 0.6773193997205177,
+      "grad_norm": 0.14061607420444489,
+      "learning_rate": 2.3704489282799025e-05,
+      "loss": 1.0928,
+      "step": 11148
+    },
+    {
+      "epoch": 0.6773801567531442,
+      "grad_norm": 0.1306842565536499,
+      "learning_rate": 2.3696347737536352e-05,
+      "loss": 1.0713,
+      "step": 11149
+    },
+    {
+      "epoch": 0.6774409137857706,
+      "grad_norm": 0.17402581870555878,
+      "learning_rate": 2.3688207156439156e-05,
+      "loss": 1.158,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6775016708183972,
+      "grad_norm": 0.1504419595003128,
+      "learning_rate": 2.3680067539805817e-05,
+      "loss": 1.0521,
+      "step": 11151
+    },
+    {
+      "epoch": 0.6775624278510237,
+      "grad_norm": 0.15994706749916077,
+      "learning_rate": 2.3671928887934687e-05,
+      "loss": 0.995,
+      "step": 11152
+    },
+    {
+      "epoch": 0.6776231848836503,
+      "grad_norm": 0.2099013328552246,
+      "learning_rate": 2.366379120112409e-05,
+      "loss": 1.0981,
+      "step": 11153
+    },
+    {
+      "epoch": 0.6776839419162768,
+      "grad_norm": 0.32417577505111694,
+      "learning_rate": 2.3655654479672324e-05,
+      "loss": 1.1424,
+      "step": 11154
+    },
+    {
+      "epoch": 0.6777446989489033,
+      "grad_norm": 0.15805046260356903,
+      "learning_rate": 2.3647518723877632e-05,
+      "loss": 1.11,
+      "step": 11155
+    },
+    {
+      "epoch": 0.6778054559815299,
+      "grad_norm": 0.17728765308856964,
+      "learning_rate": 2.3639383934038233e-05,
+      "loss": 1.0659,
+      "step": 11156
+    },
+    {
+      "epoch": 0.6778662130141564,
+      "grad_norm": 0.2740498185157776,
+      "learning_rate": 2.3631250110452312e-05,
+      "loss": 1.057,
+      "step": 11157
+    },
+    {
+      "epoch": 0.6779269700467829,
+      "grad_norm": 0.1369834691286087,
+      "learning_rate": 2.362311725341802e-05,
+      "loss": 1.0916,
+      "step": 11158
+    },
+    {
+      "epoch": 0.6779877270794095,
+      "grad_norm": 0.13214300572872162,
+      "learning_rate": 2.361498536323346e-05,
+      "loss": 1.0806,
+      "step": 11159
+    },
+    {
+      "epoch": 0.678048484112036,
+      "grad_norm": 0.13550175726413727,
+      "learning_rate": 2.3606854440196714e-05,
+      "loss": 1.0746,
+      "step": 11160
+    },
+    {
+      "epoch": 0.6781092411446625,
+      "grad_norm": 0.1341206282377243,
+      "learning_rate": 2.3598724484605815e-05,
+      "loss": 1.0567,
+      "step": 11161
+    },
+    {
+      "epoch": 0.6781699981772891,
+      "grad_norm": 0.2170247882604599,
+      "learning_rate": 2.3590595496758776e-05,
+      "loss": 1.041,
+      "step": 11162
+    },
+    {
+      "epoch": 0.6782307552099155,
+      "grad_norm": 0.12049305438995361,
+      "learning_rate": 2.3582467476953563e-05,
+      "loss": 1.0234,
+      "step": 11163
+    },
+    {
+      "epoch": 0.678291512242542,
+      "grad_norm": 0.26586124300956726,
+      "learning_rate": 2.357434042548809e-05,
+      "loss": 0.9987,
+      "step": 11164
+    },
+    {
+      "epoch": 0.6783522692751686,
+      "grad_norm": 0.14460714161396027,
+      "learning_rate": 2.3566214342660304e-05,
+      "loss": 1.0291,
+      "step": 11165
+    },
+    {
+      "epoch": 0.6784130263077951,
+      "grad_norm": 0.2643889784812927,
+      "learning_rate": 2.3558089228768048e-05,
+      "loss": 1.2214,
+      "step": 11166
+    },
+    {
+      "epoch": 0.6784737833404216,
+      "grad_norm": 0.8038681149482727,
+      "learning_rate": 2.3549965084109132e-05,
+      "loss": 1.0938,
+      "step": 11167
+    },
+    {
+      "epoch": 0.6785345403730482,
+      "grad_norm": 0.3597032129764557,
+      "learning_rate": 2.3541841908981353e-05,
+      "loss": 1.1543,
+      "step": 11168
+    },
+    {
+      "epoch": 0.6785952974056747,
+      "grad_norm": 0.14750725030899048,
+      "learning_rate": 2.353371970368246e-05,
+      "loss": 1.097,
+      "step": 11169
+    },
+    {
+      "epoch": 0.6786560544383012,
+      "grad_norm": 0.1521051973104477,
+      "learning_rate": 2.3525598468510208e-05,
+      "loss": 1.0609,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6787168114709278,
+      "grad_norm": 0.14152897894382477,
+      "learning_rate": 2.3517478203762255e-05,
+      "loss": 1.0153,
+      "step": 11171
+    },
+    {
+      "epoch": 0.6787775685035543,
+      "grad_norm": 0.18961024284362793,
+      "learning_rate": 2.3509358909736256e-05,
+      "loss": 1.0387,
+      "step": 11172
+    },
+    {
+      "epoch": 0.6788383255361808,
+      "grad_norm": 0.3233139216899872,
+      "learning_rate": 2.350124058672985e-05,
+      "loss": 1.0516,
+      "step": 11173
+    },
+    {
+      "epoch": 0.6788990825688074,
+      "grad_norm": 0.13173529505729675,
+      "learning_rate": 2.3493123235040554e-05,
+      "loss": 1.0767,
+      "step": 11174
+    },
+    {
+      "epoch": 0.6789598396014339,
+      "grad_norm": 0.1663878709077835,
+      "learning_rate": 2.348500685496597e-05,
+      "loss": 1.0475,
+      "step": 11175
+    },
+    {
+      "epoch": 0.6790205966340604,
+      "grad_norm": 0.24867284297943115,
+      "learning_rate": 2.347689144680358e-05,
+      "loss": 1.2727,
+      "step": 11176
+    },
+    {
+      "epoch": 0.6790813536666869,
+      "grad_norm": 0.13534094393253326,
+      "learning_rate": 2.3468777010850863e-05,
+      "loss": 1.0833,
+      "step": 11177
+    },
+    {
+      "epoch": 0.6791421106993134,
+      "grad_norm": 0.1138947457075119,
+      "learning_rate": 2.3460663547405253e-05,
+      "loss": 1.018,
+      "step": 11178
+    },
+    {
+      "epoch": 0.6792028677319399,
+      "grad_norm": 0.15786230564117432,
+      "learning_rate": 2.3452551056764156e-05,
+      "loss": 1.1388,
+      "step": 11179
+    },
+    {
+      "epoch": 0.6792636247645665,
+      "grad_norm": 0.18799610435962677,
+      "learning_rate": 2.344443953922492e-05,
+      "loss": 1.0976,
+      "step": 11180
+    },
+    {
+      "epoch": 0.679324381797193,
+      "grad_norm": 0.13827615976333618,
+      "learning_rate": 2.3436328995084895e-05,
+      "loss": 1.0099,
+      "step": 11181
+    },
+    {
+      "epoch": 0.6793851388298195,
+      "grad_norm": 0.16614465415477753,
+      "learning_rate": 2.3428219424641363e-05,
+      "loss": 1.1709,
+      "step": 11182
+    },
+    {
+      "epoch": 0.6794458958624461,
+      "grad_norm": 0.12641862034797668,
+      "learning_rate": 2.3420110828191584e-05,
+      "loss": 1.0408,
+      "step": 11183
+    },
+    {
+      "epoch": 0.6795066528950726,
+      "grad_norm": 0.12962312996387482,
+      "learning_rate": 2.3412003206032786e-05,
+      "loss": 1.1191,
+      "step": 11184
+    },
+    {
+      "epoch": 0.6795674099276992,
+      "grad_norm": 0.24748635292053223,
+      "learning_rate": 2.3403896558462118e-05,
+      "loss": 1.1064,
+      "step": 11185
+    },
+    {
+      "epoch": 0.6796281669603257,
+      "grad_norm": 0.19643889367580414,
+      "learning_rate": 2.3395790885776807e-05,
+      "loss": 1.1248,
+      "step": 11186
+    },
+    {
+      "epoch": 0.6796889239929522,
+      "grad_norm": 4.767789363861084,
+      "learning_rate": 2.338768618827391e-05,
+      "loss": 1.0578,
+      "step": 11187
+    },
+    {
+      "epoch": 0.6797496810255788,
+      "grad_norm": 0.1541973203420639,
+      "learning_rate": 2.3379582466250515e-05,
+      "loss": 1.0691,
+      "step": 11188
+    },
+    {
+      "epoch": 0.6798104380582053,
+      "grad_norm": 0.13764290511608124,
+      "learning_rate": 2.337147972000367e-05,
+      "loss": 1.0353,
+      "step": 11189
+    },
+    {
+      "epoch": 0.6798711950908317,
+      "grad_norm": 0.1775774359703064,
+      "learning_rate": 2.3363377949830368e-05,
+      "loss": 1.0523,
+      "step": 11190
+    },
+    {
+      "epoch": 0.6799319521234582,
+      "grad_norm": 0.13941790163516998,
+      "learning_rate": 2.335527715602761e-05,
+      "loss": 1.0844,
+      "step": 11191
+    },
+    {
+      "epoch": 0.6799927091560848,
+      "grad_norm": 0.1713925004005432,
+      "learning_rate": 2.3347177338892318e-05,
+      "loss": 1.1128,
+      "step": 11192
+    },
+    {
+      "epoch": 0.6800534661887113,
+      "grad_norm": 0.5088496208190918,
+      "learning_rate": 2.3339078498721413e-05,
+      "loss": 1.1756,
+      "step": 11193
+    },
+    {
+      "epoch": 0.6801142232213379,
+      "grad_norm": 0.17205387353897095,
+      "learning_rate": 2.3330980635811722e-05,
+      "loss": 1.0744,
+      "step": 11194
+    },
+    {
+      "epoch": 0.6801749802539644,
+      "grad_norm": 0.15832848846912384,
+      "learning_rate": 2.3322883750460068e-05,
+      "loss": 1.0926,
+      "step": 11195
+    },
+    {
+      "epoch": 0.6802357372865909,
+      "grad_norm": 0.19478711485862732,
+      "learning_rate": 2.3314787842963286e-05,
+      "loss": 1.1616,
+      "step": 11196
+    },
+    {
+      "epoch": 0.6802964943192175,
+      "grad_norm": 0.11826007068157196,
+      "learning_rate": 2.3306692913618116e-05,
+      "loss": 1.0181,
+      "step": 11197
+    },
+    {
+      "epoch": 0.680357251351844,
+      "grad_norm": 0.11528857052326202,
+      "learning_rate": 2.329859896272128e-05,
+      "loss": 1.0055,
+      "step": 11198
+    },
+    {
+      "epoch": 0.6804180083844705,
+      "grad_norm": 0.14169877767562866,
+      "learning_rate": 2.3290505990569457e-05,
+      "loss": 1.1219,
+      "step": 11199
+    },
+    {
+      "epoch": 0.6804787654170971,
+      "grad_norm": 0.12776890397071838,
+      "learning_rate": 2.3282413997459295e-05,
+      "loss": 1.0739,
+      "step": 11200
+    },
+    {
+      "epoch": 0.6805395224497236,
+      "grad_norm": 0.1028270497918129,
+      "learning_rate": 2.3274322983687417e-05,
+      "loss": 1.0428,
+      "step": 11201
+    },
+    {
+      "epoch": 0.6806002794823501,
+      "grad_norm": 0.12106592953205109,
+      "learning_rate": 2.3266232949550397e-05,
+      "loss": 1.0452,
+      "step": 11202
+    },
+    {
+      "epoch": 0.6806610365149766,
+      "grad_norm": 0.14473384618759155,
+      "learning_rate": 2.3258143895344774e-05,
+      "loss": 1.0717,
+      "step": 11203
+    },
+    {
+      "epoch": 0.6807217935476031,
+      "grad_norm": 0.2115626037120819,
+      "learning_rate": 2.3250055821367056e-05,
+      "loss": 1.1266,
+      "step": 11204
+    },
+    {
+      "epoch": 0.6807825505802296,
+      "grad_norm": 0.159811332821846,
+      "learning_rate": 2.3241968727913716e-05,
+      "loss": 1.0065,
+      "step": 11205
+    },
+    {
+      "epoch": 0.6808433076128562,
+      "grad_norm": 0.11820317059755325,
+      "learning_rate": 2.3233882615281184e-05,
+      "loss": 1.0482,
+      "step": 11206
+    },
+    {
+      "epoch": 0.6809040646454827,
+      "grad_norm": 0.1523427814245224,
+      "learning_rate": 2.3225797483765855e-05,
+      "loss": 1.0416,
+      "step": 11207
+    },
+    {
+      "epoch": 0.6809648216781092,
+      "grad_norm": 0.15173062682151794,
+      "learning_rate": 2.321771333366409e-05,
+      "loss": 1.102,
+      "step": 11208
+    },
+    {
+      "epoch": 0.6810255787107358,
+      "grad_norm": 0.18065033853054047,
+      "learning_rate": 2.320963016527223e-05,
+      "loss": 1.0746,
+      "step": 11209
+    },
+    {
+      "epoch": 0.6810863357433623,
+      "grad_norm": 0.8758243918418884,
+      "learning_rate": 2.3201547978886544e-05,
+      "loss": 1.149,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6811470927759888,
+      "grad_norm": 0.23307965695858002,
+      "learning_rate": 2.3193466774803284e-05,
+      "loss": 1.0976,
+      "step": 11211
+    },
+    {
+      "epoch": 0.6812078498086154,
+      "grad_norm": 0.12007883936166763,
+      "learning_rate": 2.31853865533187e-05,
+      "loss": 1.0643,
+      "step": 11212
+    },
+    {
+      "epoch": 0.6812686068412419,
+      "grad_norm": 0.12658709287643433,
+      "learning_rate": 2.3177307314728974e-05,
+      "loss": 1.059,
+      "step": 11213
+    },
+    {
+      "epoch": 0.6813293638738684,
+      "grad_norm": 0.17198549211025238,
+      "learning_rate": 2.3169229059330217e-05,
+      "loss": 1.0674,
+      "step": 11214
+    },
+    {
+      "epoch": 0.681390120906495,
+      "grad_norm": 0.2031710147857666,
+      "learning_rate": 2.316115178741855e-05,
+      "loss": 1.1218,
+      "step": 11215
+    },
+    {
+      "epoch": 0.6814508779391214,
+      "grad_norm": 0.2512986361980438,
+      "learning_rate": 2.3153075499290034e-05,
+      "loss": 1.0482,
+      "step": 11216
+    },
+    {
+      "epoch": 0.6815116349717479,
+      "grad_norm": 0.1166347935795784,
+      "learning_rate": 2.3145000195240745e-05,
+      "loss": 1.0264,
+      "step": 11217
+    },
+    {
+      "epoch": 0.6815723920043745,
+      "grad_norm": 0.24622927606105804,
+      "learning_rate": 2.313692587556667e-05,
+      "loss": 1.3013,
+      "step": 11218
+    },
+    {
+      "epoch": 0.681633149037001,
+      "grad_norm": 0.12514033913612366,
+      "learning_rate": 2.3128852540563762e-05,
+      "loss": 1.0634,
+      "step": 11219
+    },
+    {
+      "epoch": 0.6816939060696275,
+      "grad_norm": 0.22220630943775177,
+      "learning_rate": 2.3120780190527975e-05,
+      "loss": 1.0936,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6817546631022541,
+      "grad_norm": 0.1998179405927658,
+      "learning_rate": 2.3112708825755153e-05,
+      "loss": 1.0548,
+      "step": 11221
+    },
+    {
+      "epoch": 0.6818154201348806,
+      "grad_norm": 0.4532238245010376,
+      "learning_rate": 2.3104638446541204e-05,
+      "loss": 1.1619,
+      "step": 11222
+    },
+    {
+      "epoch": 0.6818761771675071,
+      "grad_norm": 0.14723381400108337,
+      "learning_rate": 2.3096569053181933e-05,
+      "loss": 1.0154,
+      "step": 11223
+    },
+    {
+      "epoch": 0.6819369342001337,
+      "grad_norm": 0.12438743561506271,
+      "learning_rate": 2.308850064597312e-05,
+      "loss": 1.0344,
+      "step": 11224
+    },
+    {
+      "epoch": 0.6819976912327602,
+      "grad_norm": 0.11856713891029358,
+      "learning_rate": 2.3080433225210523e-05,
+      "loss": 1.0176,
+      "step": 11225
+    },
+    {
+      "epoch": 0.6820584482653868,
+      "grad_norm": 0.19358035922050476,
+      "learning_rate": 2.3072366791189847e-05,
+      "loss": 1.0851,
+      "step": 11226
+    },
+    {
+      "epoch": 0.6821192052980133,
+      "grad_norm": 0.1101171150803566,
+      "learning_rate": 2.3064301344206768e-05,
+      "loss": 1.0453,
+      "step": 11227
+    },
+    {
+      "epoch": 0.6821799623306398,
+      "grad_norm": 0.16870665550231934,
+      "learning_rate": 2.3056236884556937e-05,
+      "loss": 1.0954,
+      "step": 11228
+    },
+    {
+      "epoch": 0.6822407193632662,
+      "grad_norm": 1.1572208404541016,
+      "learning_rate": 2.3048173412535945e-05,
+      "loss": 1.0756,
+      "step": 11229
+    },
+    {
+      "epoch": 0.6823014763958928,
+      "grad_norm": 0.12642860412597656,
+      "learning_rate": 2.3040110928439367e-05,
+      "loss": 1.0724,
+      "step": 11230
+    },
+    {
+      "epoch": 0.6823622334285193,
+      "grad_norm": 0.2222655713558197,
+      "learning_rate": 2.3032049432562734e-05,
+      "loss": 1.0656,
+      "step": 11231
+    },
+    {
+      "epoch": 0.6824229904611459,
+      "grad_norm": 0.12913857400417328,
+      "learning_rate": 2.3023988925201527e-05,
+      "loss": 1.054,
+      "step": 11232
+    },
+    {
+      "epoch": 0.6824837474937724,
+      "grad_norm": 0.2229100614786148,
+      "learning_rate": 2.3015929406651255e-05,
+      "loss": 1.1127,
+      "step": 11233
+    },
+    {
+      "epoch": 0.6825445045263989,
+      "grad_norm": 0.13377363979816437,
+      "learning_rate": 2.3007870877207288e-05,
+      "loss": 1.0141,
+      "step": 11234
+    },
+    {
+      "epoch": 0.6826052615590255,
+      "grad_norm": 0.22718104720115662,
+      "learning_rate": 2.2999813337165034e-05,
+      "loss": 1.1603,
+      "step": 11235
+    },
+    {
+      "epoch": 0.682666018591652,
+      "grad_norm": 0.1535019725561142,
+      "learning_rate": 2.2991756786819833e-05,
+      "loss": 1.0866,
+      "step": 11236
+    },
+    {
+      "epoch": 0.6827267756242785,
+      "grad_norm": 0.14151044189929962,
+      "learning_rate": 2.2983701226467e-05,
+      "loss": 1.006,
+      "step": 11237
+    },
+    {
+      "epoch": 0.6827875326569051,
+      "grad_norm": 0.12746143341064453,
+      "learning_rate": 2.297564665640183e-05,
+      "loss": 1.0689,
+      "step": 11238
+    },
+    {
+      "epoch": 0.6828482896895316,
+      "grad_norm": 0.1492866575717926,
+      "learning_rate": 2.2967593076919558e-05,
+      "loss": 1.135,
+      "step": 11239
+    },
+    {
+      "epoch": 0.6829090467221581,
+      "grad_norm": 0.16964216530323029,
+      "learning_rate": 2.2959540488315408e-05,
+      "loss": 1.022,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6829698037547847,
+      "grad_norm": 0.1894938200712204,
+      "learning_rate": 2.295148889088451e-05,
+      "loss": 1.0979,
+      "step": 11241
+    },
+    {
+      "epoch": 0.6830305607874111,
+      "grad_norm": 0.17001105844974518,
+      "learning_rate": 2.2943438284921994e-05,
+      "loss": 1.1067,
+      "step": 11242
+    },
+    {
+      "epoch": 0.6830913178200376,
+      "grad_norm": 0.5521331429481506,
+      "learning_rate": 2.2935388670722996e-05,
+      "loss": 1.198,
+      "step": 11243
+    },
+    {
+      "epoch": 0.6831520748526642,
+      "grad_norm": 0.1320745050907135,
+      "learning_rate": 2.2927340048582563e-05,
+      "loss": 1.0598,
+      "step": 11244
+    },
+    {
+      "epoch": 0.6832128318852907,
+      "grad_norm": 0.15190023183822632,
+      "learning_rate": 2.2919292418795706e-05,
+      "loss": 1.1067,
+      "step": 11245
+    },
+    {
+      "epoch": 0.6832735889179172,
+      "grad_norm": 0.12350473552942276,
+      "learning_rate": 2.2911245781657425e-05,
+      "loss": 1.0186,
+      "step": 11246
+    },
+    {
+      "epoch": 0.6833343459505438,
+      "grad_norm": 0.2489331215620041,
+      "learning_rate": 2.2903200137462668e-05,
+      "loss": 1.1251,
+      "step": 11247
+    },
+    {
+      "epoch": 0.6833951029831703,
+      "grad_norm": 0.29204264283180237,
+      "learning_rate": 2.2895155486506337e-05,
+      "loss": 0.9742,
+      "step": 11248
+    },
+    {
+      "epoch": 0.6834558600157968,
+      "grad_norm": 0.36877569556236267,
+      "learning_rate": 2.288711182908333e-05,
+      "loss": 1.0788,
+      "step": 11249
+    },
+    {
+      "epoch": 0.6835166170484234,
+      "grad_norm": 0.2829798758029938,
+      "learning_rate": 2.2879069165488477e-05,
+      "loss": 1.176,
+      "step": 11250
+    },
+    {
+      "epoch": 0.6835773740810499,
+      "grad_norm": 0.8081707954406738,
+      "learning_rate": 2.287102749601659e-05,
+      "loss": 1.0698,
+      "step": 11251
+    },
+    {
+      "epoch": 0.6836381311136764,
+      "grad_norm": 0.18228988349437714,
+      "learning_rate": 2.2862986820962428e-05,
+      "loss": 1.0294,
+      "step": 11252
+    },
+    {
+      "epoch": 0.683698888146303,
+      "grad_norm": 0.11081290990114212,
+      "learning_rate": 2.285494714062073e-05,
+      "loss": 1.0573,
+      "step": 11253
+    },
+    {
+      "epoch": 0.6837596451789295,
+      "grad_norm": 0.16781076788902283,
+      "learning_rate": 2.2846908455286198e-05,
+      "loss": 1.1313,
+      "step": 11254
+    },
+    {
+      "epoch": 0.6838204022115559,
+      "grad_norm": 0.18613655865192413,
+      "learning_rate": 2.283887076525348e-05,
+      "loss": 1.1006,
+      "step": 11255
+    },
+    {
+      "epoch": 0.6838811592441825,
+      "grad_norm": 0.13374072313308716,
+      "learning_rate": 2.2830834070817204e-05,
+      "loss": 1.0478,
+      "step": 11256
+    },
+    {
+      "epoch": 0.683941916276809,
+      "grad_norm": 0.12921583652496338,
+      "learning_rate": 2.282279837227194e-05,
+      "loss": 1.0537,
+      "step": 11257
+    },
+    {
+      "epoch": 0.6840026733094355,
+      "grad_norm": 0.19202256202697754,
+      "learning_rate": 2.2814763669912276e-05,
+      "loss": 1.162,
+      "step": 11258
+    },
+    {
+      "epoch": 0.6840634303420621,
+      "grad_norm": 0.11374222487211227,
+      "learning_rate": 2.2806729964032696e-05,
+      "loss": 1.0306,
+      "step": 11259
+    },
+    {
+      "epoch": 0.6841241873746886,
+      "grad_norm": 0.12950368225574493,
+      "learning_rate": 2.2798697254927716e-05,
+      "loss": 1.0692,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6841849444073151,
+      "grad_norm": 0.17806057631969452,
+      "learning_rate": 2.2790665542891716e-05,
+      "loss": 1.0438,
+      "step": 11261
+    },
+    {
+      "epoch": 0.6842457014399417,
+      "grad_norm": 0.2502976655960083,
+      "learning_rate": 2.278263482821912e-05,
+      "loss": 1.1371,
+      "step": 11262
+    },
+    {
+      "epoch": 0.6843064584725682,
+      "grad_norm": 0.12404952943325043,
+      "learning_rate": 2.2774605111204317e-05,
+      "loss": 1.0468,
+      "step": 11263
+    },
+    {
+      "epoch": 0.6843672155051947,
+      "grad_norm": 0.4306713342666626,
+      "learning_rate": 2.2766576392141632e-05,
+      "loss": 1.1541,
+      "step": 11264
+    },
+    {
+      "epoch": 0.6844279725378213,
+      "grad_norm": 0.21711395680904388,
+      "learning_rate": 2.2758548671325348e-05,
+      "loss": 1.0316,
+      "step": 11265
+    },
+    {
+      "epoch": 0.6844887295704478,
+      "grad_norm": 0.23647603392601013,
+      "learning_rate": 2.275052194904973e-05,
+      "loss": 1.1839,
+      "step": 11266
+    },
+    {
+      "epoch": 0.6845494866030744,
+      "grad_norm": 0.15655535459518433,
+      "learning_rate": 2.2742496225609e-05,
+      "loss": 1.0443,
+      "step": 11267
+    },
+    {
+      "epoch": 0.6846102436357008,
+      "grad_norm": 0.1550855040550232,
+      "learning_rate": 2.273447150129734e-05,
+      "loss": 1.1046,
+      "step": 11268
+    },
+    {
+      "epoch": 0.6846710006683273,
+      "grad_norm": 0.22014521062374115,
+      "learning_rate": 2.2726447776408894e-05,
+      "loss": 1.0747,
+      "step": 11269
+    },
+    {
+      "epoch": 0.6847317577009538,
+      "grad_norm": 0.15115638077259064,
+      "learning_rate": 2.271842505123778e-05,
+      "loss": 1.0504,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6847925147335804,
+      "grad_norm": 0.20195476710796356,
+      "learning_rate": 2.2710403326078066e-05,
+      "loss": 1.0528,
+      "step": 11271
+    },
+    {
+      "epoch": 0.6848532717662069,
+      "grad_norm": 0.15342214703559875,
+      "learning_rate": 2.270238260122379e-05,
+      "loss": 1.0804,
+      "step": 11272
+    },
+    {
+      "epoch": 0.6849140287988335,
+      "grad_norm": 0.2931040823459625,
+      "learning_rate": 2.2694362876968965e-05,
+      "loss": 1.051,
+      "step": 11273
+    },
+    {
+      "epoch": 0.68497478583146,
+      "grad_norm": 0.23116204142570496,
+      "learning_rate": 2.268634415360754e-05,
+      "loss": 1.0184,
+      "step": 11274
+    },
+    {
+      "epoch": 0.6850355428640865,
+      "grad_norm": 0.11853302270174026,
+      "learning_rate": 2.2678326431433454e-05,
+      "loss": 1.0903,
+      "step": 11275
+    },
+    {
+      "epoch": 0.6850962998967131,
+      "grad_norm": 0.59605872631073,
+      "learning_rate": 2.267030971074059e-05,
+      "loss": 1.1906,
+      "step": 11276
+    },
+    {
+      "epoch": 0.6851570569293396,
+      "grad_norm": 0.14413706958293915,
+      "learning_rate": 2.2662293991822807e-05,
+      "loss": 1.0237,
+      "step": 11277
+    },
+    {
+      "epoch": 0.6852178139619661,
+      "grad_norm": 0.21427135169506073,
+      "learning_rate": 2.2654279274973905e-05,
+      "loss": 1.0428,
+      "step": 11278
+    },
+    {
+      "epoch": 0.6852785709945927,
+      "grad_norm": 0.21035341918468475,
+      "learning_rate": 2.26462655604877e-05,
+      "loss": 1.1416,
+      "step": 11279
+    },
+    {
+      "epoch": 0.6853393280272192,
+      "grad_norm": 0.1456451565027237,
+      "learning_rate": 2.2638252848657938e-05,
+      "loss": 1.0368,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6854000850598457,
+      "grad_norm": 0.1785242259502411,
+      "learning_rate": 2.2630241139778286e-05,
+      "loss": 1.0541,
+      "step": 11281
+    },
+    {
+      "epoch": 0.6854608420924722,
+      "grad_norm": 0.2017097920179367,
+      "learning_rate": 2.2622230434142438e-05,
+      "loss": 1.1664,
+      "step": 11282
+    },
+    {
+      "epoch": 0.6855215991250987,
+      "grad_norm": 0.2288617640733719,
+      "learning_rate": 2.2614220732044006e-05,
+      "loss": 1.1246,
+      "step": 11283
+    },
+    {
+      "epoch": 0.6855823561577252,
+      "grad_norm": 0.5452085137367249,
+      "learning_rate": 2.2606212033776625e-05,
+      "loss": 1.1486,
+      "step": 11284
+    },
+    {
+      "epoch": 0.6856431131903518,
+      "grad_norm": 0.24827709794044495,
+      "learning_rate": 2.2598204339633833e-05,
+      "loss": 1.1406,
+      "step": 11285
+    },
+    {
+      "epoch": 0.6857038702229783,
+      "grad_norm": 0.22711576521396637,
+      "learning_rate": 2.259019764990916e-05,
+      "loss": 1.0707,
+      "step": 11286
+    },
+    {
+      "epoch": 0.6857646272556048,
+      "grad_norm": 0.11822300404310226,
+      "learning_rate": 2.258219196489611e-05,
+      "loss": 1.0378,
+      "step": 11287
+    },
+    {
+      "epoch": 0.6858253842882314,
+      "grad_norm": 0.13896025717258453,
+      "learning_rate": 2.2574187284888076e-05,
+      "loss": 1.0406,
+      "step": 11288
+    },
+    {
+      "epoch": 0.6858861413208579,
+      "grad_norm": 0.11160395294427872,
+      "learning_rate": 2.2566183610178525e-05,
+      "loss": 1.0201,
+      "step": 11289
+    },
+    {
+      "epoch": 0.6859468983534844,
+      "grad_norm": 0.5209375619888306,
+      "learning_rate": 2.2558180941060823e-05,
+      "loss": 1.1955,
+      "step": 11290
+    },
+    {
+      "epoch": 0.686007655386111,
+      "grad_norm": 0.11260931938886642,
+      "learning_rate": 2.25501792778283e-05,
+      "loss": 1.0778,
+      "step": 11291
+    },
+    {
+      "epoch": 0.6860684124187375,
+      "grad_norm": 0.21111446619033813,
+      "learning_rate": 2.2542178620774257e-05,
+      "loss": 1.0234,
+      "step": 11292
+    },
+    {
+      "epoch": 0.686129169451364,
+      "grad_norm": 0.12418483197689056,
+      "learning_rate": 2.2534178970191966e-05,
+      "loss": 0.9883,
+      "step": 11293
+    },
+    {
+      "epoch": 0.6861899264839906,
+      "grad_norm": 1.1833758354187012,
+      "learning_rate": 2.2526180326374656e-05,
+      "loss": 1.0256,
+      "step": 11294
+    },
+    {
+      "epoch": 0.686250683516617,
+      "grad_norm": 0.13782691955566406,
+      "learning_rate": 2.2518182689615518e-05,
+      "loss": 1.0233,
+      "step": 11295
+    },
+    {
+      "epoch": 0.6863114405492435,
+      "grad_norm": 0.1393684297800064,
+      "learning_rate": 2.2510186060207707e-05,
+      "loss": 1.0351,
+      "step": 11296
+    },
+    {
+      "epoch": 0.6863721975818701,
+      "grad_norm": 0.11530917882919312,
+      "learning_rate": 2.250219043844433e-05,
+      "loss": 1.0398,
+      "step": 11297
+    },
+    {
+      "epoch": 0.6864329546144966,
+      "grad_norm": 0.13274326920509338,
+      "learning_rate": 2.2494195824618485e-05,
+      "loss": 1.048,
+      "step": 11298
+    },
+    {
+      "epoch": 0.6864937116471231,
+      "grad_norm": 0.1376384198665619,
+      "learning_rate": 2.2486202219023188e-05,
+      "loss": 1.0707,
+      "step": 11299
+    },
+    {
+      "epoch": 0.6865544686797497,
+      "grad_norm": 0.1071726530790329,
+      "learning_rate": 2.2478209621951508e-05,
+      "loss": 1.0238,
+      "step": 11300
+    },
+    {
+      "epoch": 0.6866152257123762,
+      "grad_norm": 0.17224492132663727,
+      "learning_rate": 2.247021803369635e-05,
+      "loss": 1.105,
+      "step": 11301
+    },
+    {
+      "epoch": 0.6866759827450027,
+      "grad_norm": 0.24265849590301514,
+      "learning_rate": 2.2462227454550677e-05,
+      "loss": 1.1862,
+      "step": 11302
+    },
+    {
+      "epoch": 0.6867367397776293,
+      "grad_norm": 0.14050258696079254,
+      "learning_rate": 2.245423788480737e-05,
+      "loss": 1.116,
+      "step": 11303
+    },
+    {
+      "epoch": 0.6867974968102558,
+      "grad_norm": 0.14598074555397034,
+      "learning_rate": 2.244624932475929e-05,
+      "loss": 1.0575,
+      "step": 11304
+    },
+    {
+      "epoch": 0.6868582538428823,
+      "grad_norm": 0.12525883316993713,
+      "learning_rate": 2.2438261774699283e-05,
+      "loss": 1.0121,
+      "step": 11305
+    },
+    {
+      "epoch": 0.6869190108755089,
+      "grad_norm": 0.43297070264816284,
+      "learning_rate": 2.2430275234920118e-05,
+      "loss": 1.1337,
+      "step": 11306
+    },
+    {
+      "epoch": 0.6869797679081354,
+      "grad_norm": 0.8391994833946228,
+      "learning_rate": 2.242228970571456e-05,
+      "loss": 1.0812,
+      "step": 11307
+    },
+    {
+      "epoch": 0.6870405249407618,
+      "grad_norm": 0.1682329922914505,
+      "learning_rate": 2.2414305187375286e-05,
+      "loss": 1.1872,
+      "step": 11308
+    },
+    {
+      "epoch": 0.6871012819733884,
+      "grad_norm": 0.15211036801338196,
+      "learning_rate": 2.2406321680194965e-05,
+      "loss": 1.1219,
+      "step": 11309
+    },
+    {
+      "epoch": 0.6871620390060149,
+      "grad_norm": 0.11000309139490128,
+      "learning_rate": 2.2398339184466278e-05,
+      "loss": 1.0118,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6872227960386414,
+      "grad_norm": 0.16255216300487518,
+      "learning_rate": 2.2390357700481808e-05,
+      "loss": 1.1365,
+      "step": 11311
+    },
+    {
+      "epoch": 0.687283553071268,
+      "grad_norm": 0.2930504381656647,
+      "learning_rate": 2.238237722853411e-05,
+      "loss": 1.1252,
+      "step": 11312
+    },
+    {
+      "epoch": 0.6873443101038945,
+      "grad_norm": 0.12112327665090561,
+      "learning_rate": 2.237439776891573e-05,
+      "loss": 1.027,
+      "step": 11313
+    },
+    {
+      "epoch": 0.687405067136521,
+      "grad_norm": 0.14984898269176483,
+      "learning_rate": 2.2366419321919102e-05,
+      "loss": 1.1272,
+      "step": 11314
+    },
+    {
+      "epoch": 0.6874658241691476,
+      "grad_norm": 0.5324952602386475,
+      "learning_rate": 2.2358441887836735e-05,
+      "loss": 1.2852,
+      "step": 11315
+    },
+    {
+      "epoch": 0.6875265812017741,
+      "grad_norm": 0.21704143285751343,
+      "learning_rate": 2.2350465466961024e-05,
+      "loss": 1.1833,
+      "step": 11316
+    },
+    {
+      "epoch": 0.6875873382344007,
+      "grad_norm": 0.28614747524261475,
+      "learning_rate": 2.2342490059584347e-05,
+      "loss": 1.0676,
+      "step": 11317
+    },
+    {
+      "epoch": 0.6876480952670272,
+      "grad_norm": 0.19358699023723602,
+      "learning_rate": 2.233451566599904e-05,
+      "loss": 1.0704,
+      "step": 11318
+    },
+    {
+      "epoch": 0.6877088522996537,
+      "grad_norm": 0.13374973833560944,
+      "learning_rate": 2.232654228649741e-05,
+      "loss": 1.0887,
+      "step": 11319
+    },
+    {
+      "epoch": 0.6877696093322803,
+      "grad_norm": 0.12314755469560623,
+      "learning_rate": 2.2318569921371725e-05,
+      "loss": 1.1698,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6878303663649067,
+      "grad_norm": 0.12458351999521255,
+      "learning_rate": 2.2310598570914204e-05,
+      "loss": 1.0192,
+      "step": 11321
+    },
+    {
+      "epoch": 0.6878911233975332,
+      "grad_norm": 0.16685117781162262,
+      "learning_rate": 2.230262823541705e-05,
+      "loss": 1.1628,
+      "step": 11322
+    },
+    {
+      "epoch": 0.6879518804301598,
+      "grad_norm": 0.14661340415477753,
+      "learning_rate": 2.229465891517241e-05,
+      "loss": 1.1166,
+      "step": 11323
+    },
+    {
+      "epoch": 0.6880126374627863,
+      "grad_norm": 0.22112315893173218,
+      "learning_rate": 2.2286690610472404e-05,
+      "loss": 1.2693,
+      "step": 11324
+    },
+    {
+      "epoch": 0.6880733944954128,
+      "grad_norm": 9.382180213928223,
+      "learning_rate": 2.227872332160909e-05,
+      "loss": 1.1014,
+      "step": 11325
+    },
+    {
+      "epoch": 0.6881341515280394,
+      "grad_norm": 0.2312825620174408,
+      "learning_rate": 2.2270757048874553e-05,
+      "loss": 1.185,
+      "step": 11326
+    },
+    {
+      "epoch": 0.6881949085606659,
+      "grad_norm": 0.17301885783672333,
+      "learning_rate": 2.226279179256079e-05,
+      "loss": 1.0081,
+      "step": 11327
+    },
+    {
+      "epoch": 0.6882556655932924,
+      "grad_norm": 0.12839139997959137,
+      "learning_rate": 2.225482755295974e-05,
+      "loss": 1.0546,
+      "step": 11328
+    },
+    {
+      "epoch": 0.688316422625919,
+      "grad_norm": 0.18592028319835663,
+      "learning_rate": 2.2246864330363348e-05,
+      "loss": 1.0263,
+      "step": 11329
+    },
+    {
+      "epoch": 0.6883771796585455,
+      "grad_norm": 0.1334301233291626,
+      "learning_rate": 2.2238902125063487e-05,
+      "loss": 1.0454,
+      "step": 11330
+    },
+    {
+      "epoch": 0.688437936691172,
+      "grad_norm": 0.12021974474191666,
+      "learning_rate": 2.2230940937352057e-05,
+      "loss": 1.0313,
+      "step": 11331
+    },
+    {
+      "epoch": 0.6884986937237986,
+      "grad_norm": 0.10884489864110947,
+      "learning_rate": 2.2222980767520856e-05,
+      "loss": 1.0278,
+      "step": 11332
+    },
+    {
+      "epoch": 0.6885594507564251,
+      "grad_norm": 0.1361217349767685,
+      "learning_rate": 2.2215021615861658e-05,
+      "loss": 1.017,
+      "step": 11333
+    },
+    {
+      "epoch": 0.6886202077890515,
+      "grad_norm": 0.17185622453689575,
+      "learning_rate": 2.220706348266624e-05,
+      "loss": 1.2327,
+      "step": 11334
+    },
+    {
+      "epoch": 0.6886809648216781,
+      "grad_norm": 0.3156260848045349,
+      "learning_rate": 2.219910636822624e-05,
+      "loss": 1.0775,
+      "step": 11335
+    },
+    {
+      "epoch": 0.6887417218543046,
+      "grad_norm": 0.4924779534339905,
+      "learning_rate": 2.219115027283339e-05,
+      "loss": 1.1998,
+      "step": 11336
+    },
+    {
+      "epoch": 0.6888024788869311,
+      "grad_norm": 0.167219877243042,
+      "learning_rate": 2.21831951967793e-05,
+      "loss": 1.0629,
+      "step": 11337
+    },
+    {
+      "epoch": 0.6888632359195577,
+      "grad_norm": 0.24007928371429443,
+      "learning_rate": 2.2175241140355568e-05,
+      "loss": 1.0389,
+      "step": 11338
+    },
+    {
+      "epoch": 0.6889239929521842,
+      "grad_norm": 0.18390145897865295,
+      "learning_rate": 2.216728810385375e-05,
+      "loss": 1.1026,
+      "step": 11339
+    },
+    {
+      "epoch": 0.6889847499848107,
+      "grad_norm": 0.12695859372615814,
+      "learning_rate": 2.215933608756537e-05,
+      "loss": 1.0668,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6890455070174373,
+      "grad_norm": 0.12433171272277832,
+      "learning_rate": 2.2151385091781907e-05,
+      "loss": 1.0158,
+      "step": 11341
+    },
+    {
+      "epoch": 0.6891062640500638,
+      "grad_norm": 0.4193268120288849,
+      "learning_rate": 2.2143435116794803e-05,
+      "loss": 1.073,
+      "step": 11342
+    },
+    {
+      "epoch": 0.6891670210826903,
+      "grad_norm": 0.19214269518852234,
+      "learning_rate": 2.2135486162895468e-05,
+      "loss": 1.0785,
+      "step": 11343
+    },
+    {
+      "epoch": 0.6892277781153169,
+      "grad_norm": 0.16940905153751373,
+      "learning_rate": 2.2127538230375277e-05,
+      "loss": 1.0851,
+      "step": 11344
+    },
+    {
+      "epoch": 0.6892885351479434,
+      "grad_norm": 0.15236569941043854,
+      "learning_rate": 2.2119591319525557e-05,
+      "loss": 1.0669,
+      "step": 11345
+    },
+    {
+      "epoch": 0.68934929218057,
+      "grad_norm": 0.28772369027137756,
+      "learning_rate": 2.211164543063759e-05,
+      "loss": 1.0773,
+      "step": 11346
+    },
+    {
+      "epoch": 0.6894100492131964,
+      "grad_norm": 0.2376510053873062,
+      "learning_rate": 2.2103700564002688e-05,
+      "loss": 1.0759,
+      "step": 11347
+    },
+    {
+      "epoch": 0.6894708062458229,
+      "grad_norm": 0.20857170224189758,
+      "learning_rate": 2.2095756719912014e-05,
+      "loss": 1.0781,
+      "step": 11348
+    },
+    {
+      "epoch": 0.6895315632784494,
+      "grad_norm": 0.139313742518425,
+      "learning_rate": 2.2087813898656774e-05,
+      "loss": 1.0252,
+      "step": 11349
+    },
+    {
+      "epoch": 0.689592320311076,
+      "grad_norm": 0.1044449508190155,
+      "learning_rate": 2.2079872100528108e-05,
+      "loss": 1.0784,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6896530773437025,
+      "grad_norm": 0.12288221716880798,
+      "learning_rate": 2.2071931325817114e-05,
+      "loss": 1.0203,
+      "step": 11351
+    },
+    {
+      "epoch": 0.689713834376329,
+      "grad_norm": 0.12695297598838806,
+      "learning_rate": 2.2063991574814884e-05,
+      "loss": 1.0102,
+      "step": 11352
+    },
+    {
+      "epoch": 0.6897745914089556,
+      "grad_norm": 0.18684028089046478,
+      "learning_rate": 2.2056052847812454e-05,
+      "loss": 1.0566,
+      "step": 11353
+    },
+    {
+      "epoch": 0.6898353484415821,
+      "grad_norm": 0.18013936281204224,
+      "learning_rate": 2.2048115145100817e-05,
+      "loss": 1.1283,
+      "step": 11354
+    },
+    {
+      "epoch": 0.6898961054742087,
+      "grad_norm": 0.14498640596866608,
+      "learning_rate": 2.204017846697089e-05,
+      "loss": 1.1544,
+      "step": 11355
+    },
+    {
+      "epoch": 0.6899568625068352,
+      "grad_norm": 0.15685342252254486,
+      "learning_rate": 2.2032242813713643e-05,
+      "loss": 1.0622,
+      "step": 11356
+    },
+    {
+      "epoch": 0.6900176195394617,
+      "grad_norm": 0.1755741387605667,
+      "learning_rate": 2.2024308185619946e-05,
+      "loss": 1.0581,
+      "step": 11357
+    },
+    {
+      "epoch": 0.6900783765720883,
+      "grad_norm": 1.366424560546875,
+      "learning_rate": 2.2016374582980638e-05,
+      "loss": 1.2208,
+      "step": 11358
+    },
+    {
+      "epoch": 0.6901391336047148,
+      "grad_norm": 0.12541456520557404,
+      "learning_rate": 2.200844200608652e-05,
+      "loss": 1.0127,
+      "step": 11359
+    },
+    {
+      "epoch": 0.6901998906373412,
+      "grad_norm": 0.15432308614253998,
+      "learning_rate": 2.200051045522838e-05,
+      "loss": 1.0694,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6902606476699678,
+      "grad_norm": 0.15922018885612488,
+      "learning_rate": 2.1992579930696945e-05,
+      "loss": 1.0379,
+      "step": 11361
+    },
+    {
+      "epoch": 0.6903214047025943,
+      "grad_norm": 0.11118892580270767,
+      "learning_rate": 2.19846504327829e-05,
+      "loss": 1.0874,
+      "step": 11362
+    },
+    {
+      "epoch": 0.6903821617352208,
+      "grad_norm": 0.18039391934871674,
+      "learning_rate": 2.197672196177691e-05,
+      "loss": 1.148,
+      "step": 11363
+    },
+    {
+      "epoch": 0.6904429187678474,
+      "grad_norm": 0.5095503926277161,
+      "learning_rate": 2.1968794517969588e-05,
+      "loss": 1.0956,
+      "step": 11364
+    },
+    {
+      "epoch": 0.6905036758004739,
+      "grad_norm": 0.19518274068832397,
+      "learning_rate": 2.1960868101651527e-05,
+      "loss": 1.0276,
+      "step": 11365
+    },
+    {
+      "epoch": 0.6905644328331004,
+      "grad_norm": 0.11616974323987961,
+      "learning_rate": 2.1952942713113246e-05,
+      "loss": 1.0895,
+      "step": 11366
+    },
+    {
+      "epoch": 0.690625189865727,
+      "grad_norm": 0.13836927711963654,
+      "learning_rate": 2.1945018352645308e-05,
+      "loss": 1.035,
+      "step": 11367
+    },
+    {
+      "epoch": 0.6906859468983535,
+      "grad_norm": 3.9429945945739746,
+      "learning_rate": 2.193709502053813e-05,
+      "loss": 1.1059,
+      "step": 11368
+    },
+    {
+      "epoch": 0.69074670393098,
+      "grad_norm": 0.16126501560211182,
+      "learning_rate": 2.1929172717082152e-05,
+      "loss": 1.1234,
+      "step": 11369
+    },
+    {
+      "epoch": 0.6908074609636066,
+      "grad_norm": 0.2413211315870285,
+      "learning_rate": 2.1921251442567782e-05,
+      "loss": 1.1332,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6908682179962331,
+      "grad_norm": 0.22435836493968964,
+      "learning_rate": 2.191333119728534e-05,
+      "loss": 1.0584,
+      "step": 11371
+    },
+    {
+      "epoch": 0.6909289750288596,
+      "grad_norm": 0.13752995431423187,
+      "learning_rate": 2.1905411981525197e-05,
+      "loss": 1.0454,
+      "step": 11372
+    },
+    {
+      "epoch": 0.6909897320614861,
+      "grad_norm": 0.32286930084228516,
+      "learning_rate": 2.1897493795577606e-05,
+      "loss": 1.0526,
+      "step": 11373
+    },
+    {
+      "epoch": 0.6910504890941126,
+      "grad_norm": 0.3062862157821655,
+      "learning_rate": 2.1889576639732834e-05,
+      "loss": 1.1951,
+      "step": 11374
+    },
+    {
+      "epoch": 0.6911112461267391,
+      "grad_norm": 0.24602198600769043,
+      "learning_rate": 2.1881660514281043e-05,
+      "loss": 1.2371,
+      "step": 11375
+    },
+    {
+      "epoch": 0.6911720031593657,
+      "grad_norm": 0.16420650482177734,
+      "learning_rate": 2.1873745419512403e-05,
+      "loss": 1.13,
+      "step": 11376
+    },
+    {
+      "epoch": 0.6912327601919922,
+      "grad_norm": 0.11032281070947647,
+      "learning_rate": 2.1865831355717075e-05,
+      "loss": 1.0498,
+      "step": 11377
+    },
+    {
+      "epoch": 0.6912935172246187,
+      "grad_norm": 0.2551252841949463,
+      "learning_rate": 2.1857918323185145e-05,
+      "loss": 1.1135,
+      "step": 11378
+    },
+    {
+      "epoch": 0.6913542742572453,
+      "grad_norm": 1.0878547430038452,
+      "learning_rate": 2.1850006322206656e-05,
+      "loss": 1.029,
+      "step": 11379
+    },
+    {
+      "epoch": 0.6914150312898718,
+      "grad_norm": 1.1755034923553467,
+      "learning_rate": 2.184209535307164e-05,
+      "loss": 1.287,
+      "step": 11380
+    },
+    {
+      "epoch": 0.6914757883224983,
+      "grad_norm": 0.17688456177711487,
+      "learning_rate": 2.183418541607003e-05,
+      "loss": 1.055,
+      "step": 11381
+    },
+    {
+      "epoch": 0.6915365453551249,
+      "grad_norm": 0.2281084954738617,
+      "learning_rate": 2.1826276511491815e-05,
+      "loss": 1.1393,
+      "step": 11382
+    },
+    {
+      "epoch": 0.6915973023877514,
+      "grad_norm": 0.20563869178295135,
+      "learning_rate": 2.181836863962688e-05,
+      "loss": 1.0698,
+      "step": 11383
+    },
+    {
+      "epoch": 0.691658059420378,
+      "grad_norm": 0.7641192674636841,
+      "learning_rate": 2.181046180076509e-05,
+      "loss": 1.2132,
+      "step": 11384
+    },
+    {
+      "epoch": 0.6917188164530045,
+      "grad_norm": 0.3450571894645691,
+      "learning_rate": 2.1802555995196273e-05,
+      "loss": 0.9568,
+      "step": 11385
+    },
+    {
+      "epoch": 0.691779573485631,
+      "grad_norm": 0.2267465442419052,
+      "learning_rate": 2.179465122321021e-05,
+      "loss": 1.1404,
+      "step": 11386
+    },
+    {
+      "epoch": 0.6918403305182574,
+      "grad_norm": 0.1363375186920166,
+      "learning_rate": 2.1786747485096665e-05,
+      "loss": 1.0125,
+      "step": 11387
+    },
+    {
+      "epoch": 0.691901087550884,
+      "grad_norm": 0.2029709368944168,
+      "learning_rate": 2.1778844781145347e-05,
+      "loss": 1.0236,
+      "step": 11388
+    },
+    {
+      "epoch": 0.6919618445835105,
+      "grad_norm": 0.15069545805454254,
+      "learning_rate": 2.1770943111645918e-05,
+      "loss": 1.0264,
+      "step": 11389
+    },
+    {
+      "epoch": 0.692022601616137,
+      "grad_norm": 0.163344606757164,
+      "learning_rate": 2.1763042476888036e-05,
+      "loss": 1.0488,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6920833586487636,
+      "grad_norm": 0.13086891174316406,
+      "learning_rate": 2.1755142877161282e-05,
+      "loss": 1.0315,
+      "step": 11391
+    },
+    {
+      "epoch": 0.6921441156813901,
+      "grad_norm": 0.16269154846668243,
+      "learning_rate": 2.1747244312755218e-05,
+      "loss": 1.0962,
+      "step": 11392
+    },
+    {
+      "epoch": 0.6922048727140167,
+      "grad_norm": 0.14945822954177856,
+      "learning_rate": 2.1739346783959387e-05,
+      "loss": 1.1243,
+      "step": 11393
+    },
+    {
+      "epoch": 0.6922656297466432,
+      "grad_norm": 0.1559775173664093,
+      "learning_rate": 2.1731450291063275e-05,
+      "loss": 1.106,
+      "step": 11394
+    },
+    {
+      "epoch": 0.6923263867792697,
+      "grad_norm": 0.16421636939048767,
+      "learning_rate": 2.1723554834356307e-05,
+      "loss": 1.1212,
+      "step": 11395
+    },
+    {
+      "epoch": 0.6923871438118963,
+      "grad_norm": 0.1422232836484909,
+      "learning_rate": 2.1715660414127896e-05,
+      "loss": 1.0992,
+      "step": 11396
+    },
+    {
+      "epoch": 0.6924479008445228,
+      "grad_norm": 0.2752242088317871,
+      "learning_rate": 2.1707767030667403e-05,
+      "loss": 1.1142,
+      "step": 11397
+    },
+    {
+      "epoch": 0.6925086578771493,
+      "grad_norm": 0.11977748572826385,
+      "learning_rate": 2.16998746842642e-05,
+      "loss": 1.0349,
+      "step": 11398
+    },
+    {
+      "epoch": 0.6925694149097759,
+      "grad_norm": 0.13460689783096313,
+      "learning_rate": 2.1691983375207557e-05,
+      "loss": 1.0567,
+      "step": 11399
+    },
+    {
+      "epoch": 0.6926301719424023,
+      "grad_norm": 0.09862243384122849,
+      "learning_rate": 2.168409310378674e-05,
+      "loss": 1.0159,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6926909289750288,
+      "grad_norm": 0.10983823984861374,
+      "learning_rate": 2.167620387029098e-05,
+      "loss": 1.0426,
+      "step": 11401
+    },
+    {
+      "epoch": 0.6927516860076554,
+      "grad_norm": 1.2687128782272339,
+      "learning_rate": 2.1668315675009404e-05,
+      "loss": 1.0385,
+      "step": 11402
+    },
+    {
+      "epoch": 0.6928124430402819,
+      "grad_norm": 0.8386787176132202,
+      "learning_rate": 2.1660428518231207e-05,
+      "loss": 1.0136,
+      "step": 11403
+    },
+    {
+      "epoch": 0.6928732000729084,
+      "grad_norm": 0.14497166872024536,
+      "learning_rate": 2.165254240024549e-05,
+      "loss": 1.0494,
+      "step": 11404
+    },
+    {
+      "epoch": 0.692933957105535,
+      "grad_norm": 0.4801327884197235,
+      "learning_rate": 2.1644657321341304e-05,
+      "loss": 1.2055,
+      "step": 11405
+    },
+    {
+      "epoch": 0.6929947141381615,
+      "grad_norm": 0.6916252374649048,
+      "learning_rate": 2.1636773281807687e-05,
+      "loss": 1.253,
+      "step": 11406
+    },
+    {
+      "epoch": 0.693055471170788,
+      "grad_norm": 0.10596638172864914,
+      "learning_rate": 2.1628890281933627e-05,
+      "loss": 1.048,
+      "step": 11407
+    },
+    {
+      "epoch": 0.6931162282034146,
+      "grad_norm": 0.35621994733810425,
+      "learning_rate": 2.162100832200808e-05,
+      "loss": 1.0984,
+      "step": 11408
+    },
+    {
+      "epoch": 0.6931769852360411,
+      "grad_norm": 0.23045781254768372,
+      "learning_rate": 2.1613127402319954e-05,
+      "loss": 1.0596,
+      "step": 11409
+    },
+    {
+      "epoch": 0.6932377422686676,
+      "grad_norm": 0.20667323470115662,
+      "learning_rate": 2.1605247523158133e-05,
+      "loss": 1.1198,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6932984993012942,
+      "grad_norm": 0.1826099157333374,
+      "learning_rate": 2.1597368684811454e-05,
+      "loss": 1.1868,
+      "step": 11411
+    },
+    {
+      "epoch": 0.6933592563339207,
+      "grad_norm": 0.15626446902751923,
+      "learning_rate": 2.1589490887568715e-05,
+      "loss": 1.0739,
+      "step": 11412
+    },
+    {
+      "epoch": 0.6934200133665471,
+      "grad_norm": 0.12921178340911865,
+      "learning_rate": 2.1581614131718663e-05,
+      "loss": 1.0857,
+      "step": 11413
+    },
+    {
+      "epoch": 0.6934807703991737,
+      "grad_norm": 0.2649064362049103,
+      "learning_rate": 2.1573738417550072e-05,
+      "loss": 1.1169,
+      "step": 11414
+    },
+    {
+      "epoch": 0.6935415274318002,
+      "grad_norm": 0.1484181433916092,
+      "learning_rate": 2.1565863745351577e-05,
+      "loss": 1.1053,
+      "step": 11415
+    },
+    {
+      "epoch": 0.6936022844644267,
+      "grad_norm": 0.4540680944919586,
+      "learning_rate": 2.1557990115411843e-05,
+      "loss": 1.1722,
+      "step": 11416
+    },
+    {
+      "epoch": 0.6936630414970533,
+      "grad_norm": 0.22185561060905457,
+      "learning_rate": 2.1550117528019474e-05,
+      "loss": 1.0228,
+      "step": 11417
+    },
+    {
+      "epoch": 0.6937237985296798,
+      "grad_norm": 0.13270683586597443,
+      "learning_rate": 2.1542245983463035e-05,
+      "loss": 1.0491,
+      "step": 11418
+    },
+    {
+      "epoch": 0.6937845555623063,
+      "grad_norm": 0.349508672952652,
+      "learning_rate": 2.153437548203108e-05,
+      "loss": 1.0694,
+      "step": 11419
+    },
+    {
+      "epoch": 0.6938453125949329,
+      "grad_norm": 0.2577899396419525,
+      "learning_rate": 2.15265060240121e-05,
+      "loss": 1.0473,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6939060696275594,
+      "grad_norm": 0.12111520767211914,
+      "learning_rate": 2.1518637609694565e-05,
+      "loss": 1.0436,
+      "step": 11421
+    },
+    {
+      "epoch": 0.6939668266601859,
+      "grad_norm": 0.1914292424917221,
+      "learning_rate": 2.151077023936685e-05,
+      "loss": 1.1916,
+      "step": 11422
+    },
+    {
+      "epoch": 0.6940275836928125,
+      "grad_norm": 0.1535150110721588,
+      "learning_rate": 2.1502903913317346e-05,
+      "loss": 1.0549,
+      "step": 11423
+    },
+    {
+      "epoch": 0.694088340725439,
+      "grad_norm": 0.30049237608909607,
+      "learning_rate": 2.1495038631834426e-05,
+      "loss": 1.0861,
+      "step": 11424
+    },
+    {
+      "epoch": 0.6941490977580655,
+      "grad_norm": 0.15553316473960876,
+      "learning_rate": 2.148717439520637e-05,
+      "loss": 1.0094,
+      "step": 11425
+    },
+    {
+      "epoch": 0.694209854790692,
+      "grad_norm": 0.1438228189945221,
+      "learning_rate": 2.1479311203721452e-05,
+      "loss": 0.9906,
+      "step": 11426
+    },
+    {
+      "epoch": 0.6942706118233185,
+      "grad_norm": 0.11101258546113968,
+      "learning_rate": 2.1471449057667907e-05,
+      "loss": 1.0369,
+      "step": 11427
+    },
+    {
+      "epoch": 0.694331368855945,
+      "grad_norm": 0.16801387071609497,
+      "learning_rate": 2.146358795733388e-05,
+      "loss": 1.189,
+      "step": 11428
+    },
+    {
+      "epoch": 0.6943921258885716,
+      "grad_norm": 0.15543152391910553,
+      "learning_rate": 2.145572790300757e-05,
+      "loss": 1.0902,
+      "step": 11429
+    },
+    {
+      "epoch": 0.6944528829211981,
+      "grad_norm": 0.36190688610076904,
+      "learning_rate": 2.1447868894977073e-05,
+      "loss": 1.2635,
+      "step": 11430
+    },
+    {
+      "epoch": 0.6945136399538246,
+      "grad_norm": 0.1915113925933838,
+      "learning_rate": 2.1440010933530456e-05,
+      "loss": 1.0793,
+      "step": 11431
+    },
+    {
+      "epoch": 0.6945743969864512,
+      "grad_norm": 0.13725712895393372,
+      "learning_rate": 2.1432154018955763e-05,
+      "loss": 1.0968,
+      "step": 11432
+    },
+    {
+      "epoch": 0.6946351540190777,
+      "grad_norm": 0.17632752656936646,
+      "learning_rate": 2.1424298151540983e-05,
+      "loss": 1.122,
+      "step": 11433
+    },
+    {
+      "epoch": 0.6946959110517043,
+      "grad_norm": 0.16672050952911377,
+      "learning_rate": 2.141644333157408e-05,
+      "loss": 1.0549,
+      "step": 11434
+    },
+    {
+      "epoch": 0.6947566680843308,
+      "grad_norm": 0.9295737743377686,
+      "learning_rate": 2.1408589559342974e-05,
+      "loss": 1.117,
+      "step": 11435
+    },
+    {
+      "epoch": 0.6948174251169573,
+      "grad_norm": 0.15703879296779633,
+      "learning_rate": 2.1400736835135545e-05,
+      "loss": 1.0562,
+      "step": 11436
+    },
+    {
+      "epoch": 0.6948781821495839,
+      "grad_norm": 0.15473277866840363,
+      "learning_rate": 2.1392885159239628e-05,
+      "loss": 1.0567,
+      "step": 11437
+    },
+    {
+      "epoch": 0.6949389391822104,
+      "grad_norm": 0.1298961490392685,
+      "learning_rate": 2.138503453194304e-05,
+      "loss": 1.0552,
+      "step": 11438
+    },
+    {
+      "epoch": 0.6949996962148368,
+      "grad_norm": 0.25411659479141235,
+      "learning_rate": 2.1377184953533526e-05,
+      "loss": 1.0615,
+      "step": 11439
+    },
+    {
+      "epoch": 0.6950604532474634,
+      "grad_norm": 0.12521715462207794,
+      "learning_rate": 2.136933642429884e-05,
+      "loss": 1.0257,
+      "step": 11440
+    },
+    {
+      "epoch": 0.6951212102800899,
+      "grad_norm": 0.11640459299087524,
+      "learning_rate": 2.1361488944526686e-05,
+      "loss": 1.0124,
+      "step": 11441
+    },
+    {
+      "epoch": 0.6951819673127164,
+      "grad_norm": 0.146500825881958,
+      "learning_rate": 2.1353642514504675e-05,
+      "loss": 1.0278,
+      "step": 11442
+    },
+    {
+      "epoch": 0.695242724345343,
+      "grad_norm": 0.12997595965862274,
+      "learning_rate": 2.134579713452043e-05,
+      "loss": 1.0052,
+      "step": 11443
+    },
+    {
+      "epoch": 0.6953034813779695,
+      "grad_norm": 0.22321529686450958,
+      "learning_rate": 2.1337952804861515e-05,
+      "loss": 1.2223,
+      "step": 11444
+    },
+    {
+      "epoch": 0.695364238410596,
+      "grad_norm": 0.9836867451667786,
+      "learning_rate": 2.1330109525815494e-05,
+      "loss": 1.1228,
+      "step": 11445
+    },
+    {
+      "epoch": 0.6954249954432226,
+      "grad_norm": 0.22070465981960297,
+      "learning_rate": 2.132226729766985e-05,
+      "loss": 1.1562,
+      "step": 11446
+    },
+    {
+      "epoch": 0.6954857524758491,
+      "grad_norm": 0.1521066129207611,
+      "learning_rate": 2.1314426120712066e-05,
+      "loss": 1.1043,
+      "step": 11447
+    },
+    {
+      "epoch": 0.6955465095084756,
+      "grad_norm": 0.18961137533187866,
+      "learning_rate": 2.13065859952295e-05,
+      "loss": 1.0512,
+      "step": 11448
+    },
+    {
+      "epoch": 0.6956072665411022,
+      "grad_norm": 0.13366352021694183,
+      "learning_rate": 2.1298746921509582e-05,
+      "loss": 1.1128,
+      "step": 11449
+    },
+    {
+      "epoch": 0.6956680235737287,
+      "grad_norm": 0.10963138937950134,
+      "learning_rate": 2.1290908899839653e-05,
+      "loss": 0.979,
+      "step": 11450
+    },
+    {
+      "epoch": 0.6957287806063552,
+      "grad_norm": 0.18463166058063507,
+      "learning_rate": 2.1283071930506998e-05,
+      "loss": 1.0449,
+      "step": 11451
+    },
+    {
+      "epoch": 0.6957895376389817,
+      "grad_norm": 0.3294398784637451,
+      "learning_rate": 2.12752360137989e-05,
+      "loss": 1.1423,
+      "step": 11452
+    },
+    {
+      "epoch": 0.6958502946716082,
+      "grad_norm": 0.1256033331155777,
+      "learning_rate": 2.126740115000257e-05,
+      "loss": 1.0647,
+      "step": 11453
+    },
+    {
+      "epoch": 0.6959110517042347,
+      "grad_norm": 0.09753338247537613,
+      "learning_rate": 2.1259567339405213e-05,
+      "loss": 1.0281,
+      "step": 11454
+    },
+    {
+      "epoch": 0.6959718087368613,
+      "grad_norm": 0.13088829815387726,
+      "learning_rate": 2.1251734582293963e-05,
+      "loss": 1.0716,
+      "step": 11455
+    },
+    {
+      "epoch": 0.6960325657694878,
+      "grad_norm": 0.15262579917907715,
+      "learning_rate": 2.1243902878955946e-05,
+      "loss": 1.0808,
+      "step": 11456
+    },
+    {
+      "epoch": 0.6960933228021143,
+      "grad_norm": 0.1111823245882988,
+      "learning_rate": 2.1236072229678223e-05,
+      "loss": 1.0569,
+      "step": 11457
+    },
+    {
+      "epoch": 0.6961540798347409,
+      "grad_norm": 0.18385662138462067,
+      "learning_rate": 2.1228242634747835e-05,
+      "loss": 1.0541,
+      "step": 11458
+    },
+    {
+      "epoch": 0.6962148368673674,
+      "grad_norm": 0.18422475457191467,
+      "learning_rate": 2.1220414094451752e-05,
+      "loss": 1.0532,
+      "step": 11459
+    },
+    {
+      "epoch": 0.6962755938999939,
+      "grad_norm": 0.1610908806324005,
+      "learning_rate": 2.121258660907698e-05,
+      "loss": 1.1046,
+      "step": 11460
+    },
+    {
+      "epoch": 0.6963363509326205,
+      "grad_norm": 0.11393465846776962,
+      "learning_rate": 2.1204760178910422e-05,
+      "loss": 1.0313,
+      "step": 11461
+    },
+    {
+      "epoch": 0.696397107965247,
+      "grad_norm": 0.16193285584449768,
+      "learning_rate": 2.119693480423893e-05,
+      "loss": 1.0794,
+      "step": 11462
+    },
+    {
+      "epoch": 0.6964578649978735,
+      "grad_norm": 0.212175652384758,
+      "learning_rate": 2.1189110485349354e-05,
+      "loss": 1.138,
+      "step": 11463
+    },
+    {
+      "epoch": 0.6965186220305001,
+      "grad_norm": 0.11896412819623947,
+      "learning_rate": 2.118128722252849e-05,
+      "loss": 1.0222,
+      "step": 11464
+    },
+    {
+      "epoch": 0.6965793790631265,
+      "grad_norm": 0.1666450798511505,
+      "learning_rate": 2.1173465016063134e-05,
+      "loss": 1.1224,
+      "step": 11465
+    },
+    {
+      "epoch": 0.696640136095753,
+      "grad_norm": 0.15710842609405518,
+      "learning_rate": 2.1165643866239982e-05,
+      "loss": 1.089,
+      "step": 11466
+    },
+    {
+      "epoch": 0.6967008931283796,
+      "grad_norm": 0.14782632887363434,
+      "learning_rate": 2.1157823773345752e-05,
+      "loss": 0.9961,
+      "step": 11467
+    },
+    {
+      "epoch": 0.6967616501610061,
+      "grad_norm": 0.14348682761192322,
+      "learning_rate": 2.1150004737667045e-05,
+      "loss": 1.0893,
+      "step": 11468
+    },
+    {
+      "epoch": 0.6968224071936326,
+      "grad_norm": 0.12337322533130646,
+      "learning_rate": 2.1142186759490473e-05,
+      "loss": 1.0436,
+      "step": 11469
+    },
+    {
+      "epoch": 0.6968831642262592,
+      "grad_norm": 0.1315898895263672,
+      "learning_rate": 2.1134369839102646e-05,
+      "loss": 1.0505,
+      "step": 11470
+    },
+    {
+      "epoch": 0.6969439212588857,
+      "grad_norm": 0.405222624540329,
+      "learning_rate": 2.112655397679007e-05,
+      "loss": 1.1287,
+      "step": 11471
+    },
+    {
+      "epoch": 0.6970046782915122,
+      "grad_norm": 0.2001056671142578,
+      "learning_rate": 2.1118739172839236e-05,
+      "loss": 1.1667,
+      "step": 11472
+    },
+    {
+      "epoch": 0.6970654353241388,
+      "grad_norm": 0.16974933445453644,
+      "learning_rate": 2.1110925427536604e-05,
+      "loss": 1.1377,
+      "step": 11473
+    },
+    {
+      "epoch": 0.6971261923567653,
+      "grad_norm": 0.12256669998168945,
+      "learning_rate": 2.1103112741168578e-05,
+      "loss": 1.025,
+      "step": 11474
+    },
+    {
+      "epoch": 0.6971869493893919,
+      "grad_norm": 0.1347525417804718,
+      "learning_rate": 2.109530111402155e-05,
+      "loss": 1.0915,
+      "step": 11475
+    },
+    {
+      "epoch": 0.6972477064220184,
+      "grad_norm": 0.11152520030736923,
+      "learning_rate": 2.108749054638184e-05,
+      "loss": 1.0437,
+      "step": 11476
+    },
+    {
+      "epoch": 0.6973084634546449,
+      "grad_norm": 0.40980181097984314,
+      "learning_rate": 2.1079681038535753e-05,
+      "loss": 1.2565,
+      "step": 11477
+    },
+    {
+      "epoch": 0.6973692204872713,
+      "grad_norm": 0.1527237445116043,
+      "learning_rate": 2.1071872590769554e-05,
+      "loss": 1.0618,
+      "step": 11478
+    },
+    {
+      "epoch": 0.6974299775198979,
+      "grad_norm": 0.2353675663471222,
+      "learning_rate": 2.1064065203369448e-05,
+      "loss": 1.1475,
+      "step": 11479
+    },
+    {
+      "epoch": 0.6974907345525244,
+      "grad_norm": 0.18684354424476624,
+      "learning_rate": 2.1056258876621616e-05,
+      "loss": 1.1217,
+      "step": 11480
+    },
+    {
+      "epoch": 0.697551491585151,
+      "grad_norm": 0.13205884397029877,
+      "learning_rate": 2.104845361081224e-05,
+      "loss": 1.0379,
+      "step": 11481
+    },
+    {
+      "epoch": 0.6976122486177775,
+      "grad_norm": 0.16660812497138977,
+      "learning_rate": 2.1040649406227376e-05,
+      "loss": 1.1084,
+      "step": 11482
+    },
+    {
+      "epoch": 0.697673005650404,
+      "grad_norm": 0.24975334107875824,
+      "learning_rate": 2.1032846263153105e-05,
+      "loss": 1.1076,
+      "step": 11483
+    },
+    {
+      "epoch": 0.6977337626830306,
+      "grad_norm": 0.20460328459739685,
+      "learning_rate": 2.1025044181875453e-05,
+      "loss": 1.0736,
+      "step": 11484
+    },
+    {
+      "epoch": 0.6977945197156571,
+      "grad_norm": 0.11432195454835892,
+      "learning_rate": 2.101724316268039e-05,
+      "loss": 1.0284,
+      "step": 11485
+    },
+    {
+      "epoch": 0.6978552767482836,
+      "grad_norm": 0.15732809901237488,
+      "learning_rate": 2.1009443205853902e-05,
+      "loss": 1.1179,
+      "step": 11486
+    },
+    {
+      "epoch": 0.6979160337809102,
+      "grad_norm": 0.12171676009893417,
+      "learning_rate": 2.1001644311681873e-05,
+      "loss": 1.0094,
+      "step": 11487
+    },
+    {
+      "epoch": 0.6979767908135367,
+      "grad_norm": 0.14438176155090332,
+      "learning_rate": 2.0993846480450198e-05,
+      "loss": 1.0545,
+      "step": 11488
+    },
+    {
+      "epoch": 0.6980375478461632,
+      "grad_norm": 0.1527024507522583,
+      "learning_rate": 2.0986049712444666e-05,
+      "loss": 1.0182,
+      "step": 11489
+    },
+    {
+      "epoch": 0.6980983048787898,
+      "grad_norm": 0.1665276437997818,
+      "learning_rate": 2.0978254007951077e-05,
+      "loss": 1.0243,
+      "step": 11490
+    },
+    {
+      "epoch": 0.6981590619114163,
+      "grad_norm": 0.11962895095348358,
+      "learning_rate": 2.0970459367255207e-05,
+      "loss": 1.1142,
+      "step": 11491
+    },
+    {
+      "epoch": 0.6982198189440427,
+      "grad_norm": 0.10681907087564468,
+      "learning_rate": 2.0962665790642765e-05,
+      "loss": 1.0157,
+      "step": 11492
+    },
+    {
+      "epoch": 0.6982805759766693,
+      "grad_norm": 0.31860029697418213,
+      "learning_rate": 2.0954873278399422e-05,
+      "loss": 1.0798,
+      "step": 11493
+    },
+    {
+      "epoch": 0.6983413330092958,
+      "grad_norm": 0.7815797328948975,
+      "learning_rate": 2.0947081830810833e-05,
+      "loss": 1.1652,
+      "step": 11494
+    },
+    {
+      "epoch": 0.6984020900419223,
+      "grad_norm": 0.2584242522716522,
+      "learning_rate": 2.0939291448162534e-05,
+      "loss": 1.1049,
+      "step": 11495
+    },
+    {
+      "epoch": 0.6984628470745489,
+      "grad_norm": 0.23324842751026154,
+      "learning_rate": 2.093150213074015e-05,
+      "loss": 1.0847,
+      "step": 11496
+    },
+    {
+      "epoch": 0.6985236041071754,
+      "grad_norm": 0.14944741129875183,
+      "learning_rate": 2.0923713878829165e-05,
+      "loss": 1.0428,
+      "step": 11497
+    },
+    {
+      "epoch": 0.6985843611398019,
+      "grad_norm": 0.260637104511261,
+      "learning_rate": 2.0915926692715073e-05,
+      "loss": 1.1506,
+      "step": 11498
+    },
+    {
+      "epoch": 0.6986451181724285,
+      "grad_norm": 0.13984835147857666,
+      "learning_rate": 2.090814057268331e-05,
+      "loss": 1.0728,
+      "step": 11499
+    },
+    {
+      "epoch": 0.698705875205055,
+      "grad_norm": 0.15960261225700378,
+      "learning_rate": 2.0900355519019277e-05,
+      "loss": 1.0212,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6987666322376815,
+      "grad_norm": 0.10564113408327103,
+      "learning_rate": 2.0892571532008338e-05,
+      "loss": 1.019,
+      "step": 11501
+    },
+    {
+      "epoch": 0.6988273892703081,
+      "grad_norm": 0.19508783519268036,
+      "learning_rate": 2.0884788611935813e-05,
+      "loss": 1.1809,
+      "step": 11502
+    },
+    {
+      "epoch": 0.6988881463029346,
+      "grad_norm": 0.10866780579090118,
+      "learning_rate": 2.0877006759086993e-05,
+      "loss": 1.0216,
+      "step": 11503
+    },
+    {
+      "epoch": 0.6989489033355611,
+      "grad_norm": 0.17517589032649994,
+      "learning_rate": 2.086922597374712e-05,
+      "loss": 1.0548,
+      "step": 11504
+    },
+    {
+      "epoch": 0.6990096603681876,
+      "grad_norm": 0.13531221449375153,
+      "learning_rate": 2.0861446256201394e-05,
+      "loss": 1.0234,
+      "step": 11505
+    },
+    {
+      "epoch": 0.6990704174008141,
+      "grad_norm": 0.14674626290798187,
+      "learning_rate": 2.085366760673497e-05,
+      "loss": 1.1378,
+      "step": 11506
+    },
+    {
+      "epoch": 0.6991311744334406,
+      "grad_norm": 0.63823401927948,
+      "learning_rate": 2.084589002563302e-05,
+      "loss": 1.0373,
+      "step": 11507
+    },
+    {
+      "epoch": 0.6991919314660672,
+      "grad_norm": 0.2077614665031433,
+      "learning_rate": 2.083811351318062e-05,
+      "loss": 1.008,
+      "step": 11508
+    },
+    {
+      "epoch": 0.6992526884986937,
+      "grad_norm": 0.14270423352718353,
+      "learning_rate": 2.0830338069662785e-05,
+      "loss": 1.1257,
+      "step": 11509
+    },
+    {
+      "epoch": 0.6993134455313202,
+      "grad_norm": 0.1445399820804596,
+      "learning_rate": 2.082256369536455e-05,
+      "loss": 1.0397,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6993742025639468,
+      "grad_norm": 0.17797310650348663,
+      "learning_rate": 2.0814790390570865e-05,
+      "loss": 1.0898,
+      "step": 11511
+    },
+    {
+      "epoch": 0.6994349595965733,
+      "grad_norm": 0.1764162927865982,
+      "learning_rate": 2.08070181555667e-05,
+      "loss": 1.078,
+      "step": 11512
+    },
+    {
+      "epoch": 0.6994957166291998,
+      "grad_norm": 0.19250036776065826,
+      "learning_rate": 2.0799246990636922e-05,
+      "loss": 1.0716,
+      "step": 11513
+    },
+    {
+      "epoch": 0.6995564736618264,
+      "grad_norm": 1.214611291885376,
+      "learning_rate": 2.0791476896066407e-05,
+      "loss": 1.0709,
+      "step": 11514
+    },
+    {
+      "epoch": 0.6996172306944529,
+      "grad_norm": 0.1410617232322693,
+      "learning_rate": 2.078370787213994e-05,
+      "loss": 1.0417,
+      "step": 11515
+    },
+    {
+      "epoch": 0.6996779877270795,
+      "grad_norm": 2.257185935974121,
+      "learning_rate": 2.0775939919142284e-05,
+      "loss": 1.1272,
+      "step": 11516
+    },
+    {
+      "epoch": 0.699738744759706,
+      "grad_norm": 0.19316335022449493,
+      "learning_rate": 2.0768173037358224e-05,
+      "loss": 1.0308,
+      "step": 11517
+    },
+    {
+      "epoch": 0.6997995017923324,
+      "grad_norm": 0.27087971568107605,
+      "learning_rate": 2.0760407227072424e-05,
+      "loss": 1.1541,
+      "step": 11518
+    },
+    {
+      "epoch": 0.699860258824959,
+      "grad_norm": 0.10225939005613327,
+      "learning_rate": 2.0752642488569556e-05,
+      "loss": 1.0157,
+      "step": 11519
+    },
+    {
+      "epoch": 0.6999210158575855,
+      "grad_norm": 0.10104874521493912,
+      "learning_rate": 2.074487882213423e-05,
+      "loss": 1.047,
+      "step": 11520
+    },
+    {
+      "epoch": 0.699981772890212,
+      "grad_norm": 0.4039386510848999,
+      "learning_rate": 2.0737116228051023e-05,
+      "loss": 1.0645,
+      "step": 11521
+    },
+    {
+      "epoch": 0.7000425299228386,
+      "grad_norm": 0.11678233742713928,
+      "learning_rate": 2.0729354706604476e-05,
+      "loss": 1.0396,
+      "step": 11522
+    },
+    {
+      "epoch": 0.7001032869554651,
+      "grad_norm": 11.37902545928955,
+      "learning_rate": 2.0721594258079093e-05,
+      "loss": 1.0509,
+      "step": 11523
+    },
+    {
+      "epoch": 0.7001640439880916,
+      "grad_norm": 0.1936226338148117,
+      "learning_rate": 2.071383488275933e-05,
+      "loss": 1.0874,
+      "step": 11524
+    },
+    {
+      "epoch": 0.7002248010207182,
+      "grad_norm": 0.11814224720001221,
+      "learning_rate": 2.070607658092961e-05,
+      "loss": 1.0227,
+      "step": 11525
+    },
+    {
+      "epoch": 0.7002855580533447,
+      "grad_norm": 0.14713111519813538,
+      "learning_rate": 2.069831935287432e-05,
+      "loss": 1.0856,
+      "step": 11526
+    },
+    {
+      "epoch": 0.7003463150859712,
+      "grad_norm": 3.862598180770874,
+      "learning_rate": 2.0690563198877776e-05,
+      "loss": 1.0789,
+      "step": 11527
+    },
+    {
+      "epoch": 0.7004070721185978,
+      "grad_norm": 0.1093360111117363,
+      "learning_rate": 2.068280811922434e-05,
+      "loss": 1.0277,
+      "step": 11528
+    },
+    {
+      "epoch": 0.7004678291512243,
+      "grad_norm": 0.1838693618774414,
+      "learning_rate": 2.067505411419822e-05,
+      "loss": 1.0592,
+      "step": 11529
+    },
+    {
+      "epoch": 0.7005285861838508,
+      "grad_norm": 0.1865403652191162,
+      "learning_rate": 2.066730118408366e-05,
+      "loss": 1.0853,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7005893432164773,
+      "grad_norm": 0.14745014905929565,
+      "learning_rate": 2.0659549329164845e-05,
+      "loss": 1.0585,
+      "step": 11531
+    },
+    {
+      "epoch": 0.7006501002491038,
+      "grad_norm": 0.1056385412812233,
+      "learning_rate": 2.0651798549725903e-05,
+      "loss": 1.0336,
+      "step": 11532
+    },
+    {
+      "epoch": 0.7007108572817303,
+      "grad_norm": 0.11104696989059448,
+      "learning_rate": 2.0644048846050977e-05,
+      "loss": 1.0837,
+      "step": 11533
+    },
+    {
+      "epoch": 0.7007716143143569,
+      "grad_norm": 0.21141260862350464,
+      "learning_rate": 2.0636300218424136e-05,
+      "loss": 1.1472,
+      "step": 11534
+    },
+    {
+      "epoch": 0.7008323713469834,
+      "grad_norm": 6.516228199005127,
+      "learning_rate": 2.0628552667129363e-05,
+      "loss": 1.0657,
+      "step": 11535
+    },
+    {
+      "epoch": 0.7008931283796099,
+      "grad_norm": 0.24817363917827606,
+      "learning_rate": 2.062080619245067e-05,
+      "loss": 1.0763,
+      "step": 11536
+    },
+    {
+      "epoch": 0.7009538854122365,
+      "grad_norm": 0.47558996081352234,
+      "learning_rate": 2.061306079467199e-05,
+      "loss": 1.2482,
+      "step": 11537
+    },
+    {
+      "epoch": 0.701014642444863,
+      "grad_norm": 0.11814045161008835,
+      "learning_rate": 2.0605316474077264e-05,
+      "loss": 1.0928,
+      "step": 11538
+    },
+    {
+      "epoch": 0.7010753994774895,
+      "grad_norm": 0.18713463842868805,
+      "learning_rate": 2.0597573230950347e-05,
+      "loss": 1.0827,
+      "step": 11539
+    },
+    {
+      "epoch": 0.7011361565101161,
+      "grad_norm": 0.2151116132736206,
+      "learning_rate": 2.058983106557506e-05,
+      "loss": 1.0757,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7011969135427426,
+      "grad_norm": 0.12864471971988678,
+      "learning_rate": 2.0582089978235204e-05,
+      "loss": 1.037,
+      "step": 11541
+    },
+    {
+      "epoch": 0.7012576705753691,
+      "grad_norm": 0.14003022015094757,
+      "learning_rate": 2.0574349969214524e-05,
+      "loss": 1.0369,
+      "step": 11542
+    },
+    {
+      "epoch": 0.7013184276079957,
+      "grad_norm": 0.19501356780529022,
+      "learning_rate": 2.0566611038796735e-05,
+      "loss": 1.068,
+      "step": 11543
+    },
+    {
+      "epoch": 0.7013791846406221,
+      "grad_norm": 0.11872688680887222,
+      "learning_rate": 2.055887318726551e-05,
+      "loss": 1.0429,
+      "step": 11544
+    },
+    {
+      "epoch": 0.7014399416732486,
+      "grad_norm": 0.12238573282957077,
+      "learning_rate": 2.0551136414904475e-05,
+      "loss": 0.9964,
+      "step": 11545
+    },
+    {
+      "epoch": 0.7015006987058752,
+      "grad_norm": 0.16445164382457733,
+      "learning_rate": 2.054340072199723e-05,
+      "loss": 1.0246,
+      "step": 11546
+    },
+    {
+      "epoch": 0.7015614557385017,
+      "grad_norm": 0.13491672277450562,
+      "learning_rate": 2.0535666108827323e-05,
+      "loss": 1.0629,
+      "step": 11547
+    },
+    {
+      "epoch": 0.7016222127711282,
+      "grad_norm": 0.20657508075237274,
+      "learning_rate": 2.052793257567827e-05,
+      "loss": 1.1592,
+      "step": 11548
+    },
+    {
+      "epoch": 0.7016829698037548,
+      "grad_norm": 0.1072314977645874,
+      "learning_rate": 2.052020012283355e-05,
+      "loss": 1.004,
+      "step": 11549
+    },
+    {
+      "epoch": 0.7017437268363813,
+      "grad_norm": 0.13779258728027344,
+      "learning_rate": 2.0512468750576592e-05,
+      "loss": 1.0574,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7018044838690078,
+      "grad_norm": 0.16147683560848236,
+      "learning_rate": 2.0504738459190785e-05,
+      "loss": 1.1305,
+      "step": 11551
+    },
+    {
+      "epoch": 0.7018652409016344,
+      "grad_norm": 0.1400379240512848,
+      "learning_rate": 2.04970092489595e-05,
+      "loss": 1.0113,
+      "step": 11552
+    },
+    {
+      "epoch": 0.7019259979342609,
+      "grad_norm": 0.38990890979766846,
+      "learning_rate": 2.048928112016602e-05,
+      "loss": 1.1475,
+      "step": 11553
+    },
+    {
+      "epoch": 0.7019867549668874,
+      "grad_norm": 0.12363315373659134,
+      "learning_rate": 2.0481554073093666e-05,
+      "loss": 1.0476,
+      "step": 11554
+    },
+    {
+      "epoch": 0.702047511999514,
+      "grad_norm": 0.17727401852607727,
+      "learning_rate": 2.0473828108025666e-05,
+      "loss": 1.124,
+      "step": 11555
+    },
+    {
+      "epoch": 0.7021082690321405,
+      "grad_norm": 0.11687791347503662,
+      "learning_rate": 2.0466103225245192e-05,
+      "loss": 1.0628,
+      "step": 11556
+    },
+    {
+      "epoch": 0.7021690260647669,
+      "grad_norm": 0.12113591283559799,
+      "learning_rate": 2.045837942503539e-05,
+      "loss": 1.069,
+      "step": 11557
+    },
+    {
+      "epoch": 0.7022297830973935,
+      "grad_norm": 0.20920348167419434,
+      "learning_rate": 2.0450656707679422e-05,
+      "loss": 1.303,
+      "step": 11558
+    },
+    {
+      "epoch": 0.70229054013002,
+      "grad_norm": 0.10549594461917877,
+      "learning_rate": 2.0442935073460345e-05,
+      "loss": 1.0291,
+      "step": 11559
+    },
+    {
+      "epoch": 0.7023512971626465,
+      "grad_norm": 0.22226223349571228,
+      "learning_rate": 2.0435214522661194e-05,
+      "loss": 1.0751,
+      "step": 11560
+    },
+    {
+      "epoch": 0.7024120541952731,
+      "grad_norm": 0.10856860131025314,
+      "learning_rate": 2.042749505556499e-05,
+      "loss": 1.0625,
+      "step": 11561
+    },
+    {
+      "epoch": 0.7024728112278996,
+      "grad_norm": 0.09366056323051453,
+      "learning_rate": 2.0419776672454634e-05,
+      "loss": 1.0127,
+      "step": 11562
+    },
+    {
+      "epoch": 0.7025335682605262,
+      "grad_norm": 0.15423716604709625,
+      "learning_rate": 2.0412059373613096e-05,
+      "loss": 1.1022,
+      "step": 11563
+    },
+    {
+      "epoch": 0.7025943252931527,
+      "grad_norm": 0.1703069806098938,
+      "learning_rate": 2.0404343159323243e-05,
+      "loss": 1.0425,
+      "step": 11564
+    },
+    {
+      "epoch": 0.7026550823257792,
+      "grad_norm": 0.1194060817360878,
+      "learning_rate": 2.039662802986791e-05,
+      "loss": 1.0708,
+      "step": 11565
+    },
+    {
+      "epoch": 0.7027158393584058,
+      "grad_norm": 0.18426650762557983,
+      "learning_rate": 2.03889139855299e-05,
+      "loss": 1.1813,
+      "step": 11566
+    },
+    {
+      "epoch": 0.7027765963910323,
+      "grad_norm": 0.17231197655200958,
+      "learning_rate": 2.038120102659198e-05,
+      "loss": 1.1226,
+      "step": 11567
+    },
+    {
+      "epoch": 0.7028373534236588,
+      "grad_norm": 2.869856595993042,
+      "learning_rate": 2.037348915333685e-05,
+      "loss": 1.1197,
+      "step": 11568
+    },
+    {
+      "epoch": 0.7028981104562854,
+      "grad_norm": 0.11700932681560516,
+      "learning_rate": 2.0365778366047206e-05,
+      "loss": 1.0465,
+      "step": 11569
+    },
+    {
+      "epoch": 0.7029588674889118,
+      "grad_norm": 0.29313933849334717,
+      "learning_rate": 2.0358068665005685e-05,
+      "loss": 1.0747,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7030196245215383,
+      "grad_norm": 0.12552626430988312,
+      "learning_rate": 2.0350360050494888e-05,
+      "loss": 1.0245,
+      "step": 11571
+    },
+    {
+      "epoch": 0.7030803815541649,
+      "grad_norm": 0.19134995341300964,
+      "learning_rate": 2.0342652522797374e-05,
+      "loss": 1.0625,
+      "step": 11572
+    },
+    {
+      "epoch": 0.7031411385867914,
+      "grad_norm": 0.11859611421823502,
+      "learning_rate": 2.0334946082195643e-05,
+      "loss": 1.0288,
+      "step": 11573
+    },
+    {
+      "epoch": 0.7032018956194179,
+      "grad_norm": 0.1901741623878479,
+      "learning_rate": 2.0327240728972215e-05,
+      "loss": 1.1176,
+      "step": 11574
+    },
+    {
+      "epoch": 0.7032626526520445,
+      "grad_norm": 0.22749671339988708,
+      "learning_rate": 2.0319536463409533e-05,
+      "loss": 1.0367,
+      "step": 11575
+    },
+    {
+      "epoch": 0.703323409684671,
+      "grad_norm": 0.16825591027736664,
+      "learning_rate": 2.0311833285789968e-05,
+      "loss": 0.9985,
+      "step": 11576
+    },
+    {
+      "epoch": 0.7033841667172975,
+      "grad_norm": 0.25161734223365784,
+      "learning_rate": 2.030413119639588e-05,
+      "loss": 1.0393,
+      "step": 11577
+    },
+    {
+      "epoch": 0.7034449237499241,
+      "grad_norm": 0.14071886241436005,
+      "learning_rate": 2.0296430195509597e-05,
+      "loss": 1.0845,
+      "step": 11578
+    },
+    {
+      "epoch": 0.7035056807825506,
+      "grad_norm": 0.35491394996643066,
+      "learning_rate": 2.0288730283413416e-05,
+      "loss": 1.153,
+      "step": 11579
+    },
+    {
+      "epoch": 0.7035664378151771,
+      "grad_norm": 2.2419357299804688,
+      "learning_rate": 2.0281031460389576e-05,
+      "loss": 0.9888,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7036271948478037,
+      "grad_norm": 0.13546213507652283,
+      "learning_rate": 2.0273333726720284e-05,
+      "loss": 1.0679,
+      "step": 11581
+    },
+    {
+      "epoch": 0.7036879518804302,
+      "grad_norm": 0.11944033205509186,
+      "learning_rate": 2.0265637082687677e-05,
+      "loss": 1.0736,
+      "step": 11582
+    },
+    {
+      "epoch": 0.7037487089130566,
+      "grad_norm": 0.2002701759338379,
+      "learning_rate": 2.0257941528573872e-05,
+      "loss": 1.2367,
+      "step": 11583
+    },
+    {
+      "epoch": 0.7038094659456832,
+      "grad_norm": 0.23972612619400024,
+      "learning_rate": 2.0250247064660987e-05,
+      "loss": 1.1442,
+      "step": 11584
+    },
+    {
+      "epoch": 0.7038702229783097,
+      "grad_norm": 0.19439685344696045,
+      "learning_rate": 2.024255369123104e-05,
+      "loss": 1.1399,
+      "step": 11585
+    },
+    {
+      "epoch": 0.7039309800109362,
+      "grad_norm": 0.1960258185863495,
+      "learning_rate": 2.0234861408566052e-05,
+      "loss": 1.13,
+      "step": 11586
+    },
+    {
+      "epoch": 0.7039917370435628,
+      "grad_norm": 2.3999454975128174,
+      "learning_rate": 2.0227170216947967e-05,
+      "loss": 1.1277,
+      "step": 11587
+    },
+    {
+      "epoch": 0.7040524940761893,
+      "grad_norm": 0.16559021174907684,
+      "learning_rate": 2.0219480116658707e-05,
+      "loss": 1.0738,
+      "step": 11588
+    },
+    {
+      "epoch": 0.7041132511088158,
+      "grad_norm": 0.17407269775867462,
+      "learning_rate": 2.0211791107980167e-05,
+      "loss": 0.9817,
+      "step": 11589
+    },
+    {
+      "epoch": 0.7041740081414424,
+      "grad_norm": 0.24084053933620453,
+      "learning_rate": 2.0204103191194178e-05,
+      "loss": 1.2063,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7042347651740689,
+      "grad_norm": 0.14700211584568024,
+      "learning_rate": 2.0196416366582543e-05,
+      "loss": 1.0464,
+      "step": 11591
+    },
+    {
+      "epoch": 0.7042955222066954,
+      "grad_norm": 0.13604646921157837,
+      "learning_rate": 2.0188730634427035e-05,
+      "loss": 1.0574,
+      "step": 11592
+    },
+    {
+      "epoch": 0.704356279239322,
+      "grad_norm": 0.18877767026424408,
+      "learning_rate": 2.018104599500936e-05,
+      "loss": 1.0423,
+      "step": 11593
+    },
+    {
+      "epoch": 0.7044170362719485,
+      "grad_norm": 0.12309647351503372,
+      "learning_rate": 2.0173362448611195e-05,
+      "loss": 1.0944,
+      "step": 11594
+    },
+    {
+      "epoch": 0.704477793304575,
+      "grad_norm": 0.5432168245315552,
+      "learning_rate": 2.016567999551423e-05,
+      "loss": 1.2447,
+      "step": 11595
+    },
+    {
+      "epoch": 0.7045385503372016,
+      "grad_norm": 0.16699868440628052,
+      "learning_rate": 2.0157998636000015e-05,
+      "loss": 1.0167,
+      "step": 11596
+    },
+    {
+      "epoch": 0.704599307369828,
+      "grad_norm": 0.1616109162569046,
+      "learning_rate": 2.0150318370350125e-05,
+      "loss": 1.11,
+      "step": 11597
+    },
+    {
+      "epoch": 0.7046600644024545,
+      "grad_norm": 0.12054099887609482,
+      "learning_rate": 2.0142639198846086e-05,
+      "loss": 1.0459,
+      "step": 11598
+    },
+    {
+      "epoch": 0.7047208214350811,
+      "grad_norm": 0.19042310118675232,
+      "learning_rate": 2.0134961121769364e-05,
+      "loss": 1.0666,
+      "step": 11599
+    },
+    {
+      "epoch": 0.7047815784677076,
+      "grad_norm": 0.11563307791948318,
+      "learning_rate": 2.0127284139401426e-05,
+      "loss": 1.003,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7048423355003341,
+      "grad_norm": 0.16380497813224792,
+      "learning_rate": 2.0119608252023688e-05,
+      "loss": 1.047,
+      "step": 11601
+    },
+    {
+      "epoch": 0.7049030925329607,
+      "grad_norm": 0.15207651257514954,
+      "learning_rate": 2.011193345991746e-05,
+      "loss": 1.0427,
+      "step": 11602
+    },
+    {
+      "epoch": 0.7049638495655872,
+      "grad_norm": 0.3474125266075134,
+      "learning_rate": 2.010425976336409e-05,
+      "loss": 1.0842,
+      "step": 11603
+    },
+    {
+      "epoch": 0.7050246065982138,
+      "grad_norm": 0.1500178575515747,
+      "learning_rate": 2.0096587162644842e-05,
+      "loss": 1.0639,
+      "step": 11604
+    },
+    {
+      "epoch": 0.7050853636308403,
+      "grad_norm": 0.11654957383871078,
+      "learning_rate": 2.0088915658040992e-05,
+      "loss": 1.0401,
+      "step": 11605
+    },
+    {
+      "epoch": 0.7051461206634668,
+      "grad_norm": 0.115399070084095,
+      "learning_rate": 2.0081245249833713e-05,
+      "loss": 1.035,
+      "step": 11606
+    },
+    {
+      "epoch": 0.7052068776960934,
+      "grad_norm": 0.17292408645153046,
+      "learning_rate": 2.0073575938304178e-05,
+      "loss": 1.0734,
+      "step": 11607
+    },
+    {
+      "epoch": 0.7052676347287199,
+      "grad_norm": 0.15123586356639862,
+      "learning_rate": 2.006590772373352e-05,
+      "loss": 1.0592,
+      "step": 11608
+    },
+    {
+      "epoch": 0.7053283917613464,
+      "grad_norm": 0.1322067379951477,
+      "learning_rate": 2.0058240606402755e-05,
+      "loss": 1.021,
+      "step": 11609
+    },
+    {
+      "epoch": 0.7053891487939729,
+      "grad_norm": 0.48658719658851624,
+      "learning_rate": 2.0050574586592996e-05,
+      "loss": 1.0216,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7054499058265994,
+      "grad_norm": 0.2290707677602768,
+      "learning_rate": 2.0042909664585214e-05,
+      "loss": 1.1641,
+      "step": 11611
+    },
+    {
+      "epoch": 0.7055106628592259,
+      "grad_norm": 0.16731087863445282,
+      "learning_rate": 2.0035245840660367e-05,
+      "loss": 1.0914,
+      "step": 11612
+    },
+    {
+      "epoch": 0.7055714198918525,
+      "grad_norm": 0.2058500200510025,
+      "learning_rate": 2.0027583115099375e-05,
+      "loss": 1.1184,
+      "step": 11613
+    },
+    {
+      "epoch": 0.705632176924479,
+      "grad_norm": 0.11693782359361649,
+      "learning_rate": 2.0019921488183118e-05,
+      "loss": 1.0551,
+      "step": 11614
+    },
+    {
+      "epoch": 0.7056929339571055,
+      "grad_norm": 0.12149243801832199,
+      "learning_rate": 2.001226096019243e-05,
+      "loss": 1.0696,
+      "step": 11615
+    },
+    {
+      "epoch": 0.7057536909897321,
+      "grad_norm": 0.1214408278465271,
+      "learning_rate": 2.000460153140812e-05,
+      "loss": 1.0698,
+      "step": 11616
+    },
+    {
+      "epoch": 0.7058144480223586,
+      "grad_norm": 0.10472622513771057,
+      "learning_rate": 1.9996943202110934e-05,
+      "loss": 1.0172,
+      "step": 11617
+    },
+    {
+      "epoch": 0.7058752050549851,
+      "grad_norm": 0.18020954728126526,
+      "learning_rate": 1.9989285972581595e-05,
+      "loss": 1.1391,
+      "step": 11618
+    },
+    {
+      "epoch": 0.7059359620876117,
+      "grad_norm": 0.1960398256778717,
+      "learning_rate": 1.9981629843100784e-05,
+      "loss": 1.1194,
+      "step": 11619
+    },
+    {
+      "epoch": 0.7059967191202382,
+      "grad_norm": 0.19699768722057343,
+      "learning_rate": 1.9973974813949115e-05,
+      "loss": 1.0165,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7060574761528647,
+      "grad_norm": 0.15521733462810516,
+      "learning_rate": 1.9966320885407213e-05,
+      "loss": 1.0824,
+      "step": 11621
+    },
+    {
+      "epoch": 0.7061182331854913,
+      "grad_norm": 0.12691210210323334,
+      "learning_rate": 1.995866805775565e-05,
+      "loss": 0.9915,
+      "step": 11622
+    },
+    {
+      "epoch": 0.7061789902181177,
+      "grad_norm": 0.15325871109962463,
+      "learning_rate": 1.9951016331274897e-05,
+      "loss": 1.0658,
+      "step": 11623
+    },
+    {
+      "epoch": 0.7062397472507442,
+      "grad_norm": 0.13043555617332458,
+      "learning_rate": 1.9943365706245447e-05,
+      "loss": 1.0327,
+      "step": 11624
+    },
+    {
+      "epoch": 0.7063005042833708,
+      "grad_norm": 0.12522801756858826,
+      "learning_rate": 1.9935716182947716e-05,
+      "loss": 1.1267,
+      "step": 11625
+    },
+    {
+      "epoch": 0.7063612613159973,
+      "grad_norm": 1.1309356689453125,
+      "learning_rate": 1.9928067761662146e-05,
+      "loss": 1.0665,
+      "step": 11626
+    },
+    {
+      "epoch": 0.7064220183486238,
+      "grad_norm": 0.3840314745903015,
+      "learning_rate": 1.992042044266906e-05,
+      "loss": 1.1726,
+      "step": 11627
+    },
+    {
+      "epoch": 0.7064827753812504,
+      "grad_norm": 0.11954229325056076,
+      "learning_rate": 1.991277422624879e-05,
+      "loss": 1.0611,
+      "step": 11628
+    },
+    {
+      "epoch": 0.7065435324138769,
+      "grad_norm": 0.24044634401798248,
+      "learning_rate": 1.9905129112681585e-05,
+      "loss": 1.0721,
+      "step": 11629
+    },
+    {
+      "epoch": 0.7066042894465034,
+      "grad_norm": 0.1350989192724228,
+      "learning_rate": 1.989748510224767e-05,
+      "loss": 1.1328,
+      "step": 11630
+    },
+    {
+      "epoch": 0.70666504647913,
+      "grad_norm": 0.11142914742231369,
+      "learning_rate": 1.9889842195227276e-05,
+      "loss": 1.0081,
+      "step": 11631
+    },
+    {
+      "epoch": 0.7067258035117565,
+      "grad_norm": 0.18580536544322968,
+      "learning_rate": 1.9882200391900536e-05,
+      "loss": 1.1613,
+      "step": 11632
+    },
+    {
+      "epoch": 0.706786560544383,
+      "grad_norm": 0.14073748886585236,
+      "learning_rate": 1.987455969254756e-05,
+      "loss": 1.0609,
+      "step": 11633
+    },
+    {
+      "epoch": 0.7068473175770096,
+      "grad_norm": 0.11089940369129181,
+      "learning_rate": 1.9866920097448426e-05,
+      "loss": 1.01,
+      "step": 11634
+    },
+    {
+      "epoch": 0.7069080746096361,
+      "grad_norm": 0.17491337656974792,
+      "learning_rate": 1.9859281606883158e-05,
+      "loss": 1.1184,
+      "step": 11635
+    },
+    {
+      "epoch": 0.7069688316422625,
+      "grad_norm": 0.1741686910390854,
+      "learning_rate": 1.985164422113175e-05,
+      "loss": 1.0952,
+      "step": 11636
+    },
+    {
+      "epoch": 0.7070295886748891,
+      "grad_norm": 1.6490581035614014,
+      "learning_rate": 1.9844007940474153e-05,
+      "loss": 1.0455,
+      "step": 11637
+    },
+    {
+      "epoch": 0.7070903457075156,
+      "grad_norm": 0.17692893743515015,
+      "learning_rate": 1.983637276519027e-05,
+      "loss": 1.1266,
+      "step": 11638
+    },
+    {
+      "epoch": 0.7071511027401421,
+      "grad_norm": 0.16384969651699066,
+      "learning_rate": 1.9828738695559972e-05,
+      "loss": 1.0853,
+      "step": 11639
+    },
+    {
+      "epoch": 0.7072118597727687,
+      "grad_norm": 0.647416889667511,
+      "learning_rate": 1.9821105731863093e-05,
+      "loss": 1.0369,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7072726168053952,
+      "grad_norm": 0.13874442875385284,
+      "learning_rate": 1.9813473874379395e-05,
+      "loss": 1.0949,
+      "step": 11641
+    },
+    {
+      "epoch": 0.7073333738380218,
+      "grad_norm": 0.11145878583192825,
+      "learning_rate": 1.9805843123388686e-05,
+      "loss": 1.0891,
+      "step": 11642
+    },
+    {
+      "epoch": 0.7073941308706483,
+      "grad_norm": 0.12096337229013443,
+      "learning_rate": 1.9798213479170613e-05,
+      "loss": 1.0212,
+      "step": 11643
+    },
+    {
+      "epoch": 0.7074548879032748,
+      "grad_norm": 0.1196274682879448,
+      "learning_rate": 1.9790584942004862e-05,
+      "loss": 1.0435,
+      "step": 11644
+    },
+    {
+      "epoch": 0.7075156449359014,
+      "grad_norm": 0.11231572180986404,
+      "learning_rate": 1.9782957512171068e-05,
+      "loss": 1.0428,
+      "step": 11645
+    },
+    {
+      "epoch": 0.7075764019685279,
+      "grad_norm": 0.755441427230835,
+      "learning_rate": 1.977533118994878e-05,
+      "loss": 1.0262,
+      "step": 11646
+    },
+    {
+      "epoch": 0.7076371590011544,
+      "grad_norm": 0.1362077295780182,
+      "learning_rate": 1.976770597561759e-05,
+      "loss": 1.0145,
+      "step": 11647
+    },
+    {
+      "epoch": 0.707697916033781,
+      "grad_norm": 0.11280715465545654,
+      "learning_rate": 1.9760081869457005e-05,
+      "loss": 1.0652,
+      "step": 11648
+    },
+    {
+      "epoch": 0.7077586730664074,
+      "grad_norm": 0.14085575938224792,
+      "learning_rate": 1.9752458871746437e-05,
+      "loss": 1.114,
+      "step": 11649
+    },
+    {
+      "epoch": 0.7078194300990339,
+      "grad_norm": 0.1524488478899002,
+      "learning_rate": 1.9744836982765325e-05,
+      "loss": 1.1108,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7078801871316605,
+      "grad_norm": 0.15377359092235565,
+      "learning_rate": 1.9737216202793073e-05,
+      "loss": 0.9882,
+      "step": 11651
+    },
+    {
+      "epoch": 0.707940944164287,
+      "grad_norm": 0.11559594422578812,
+      "learning_rate": 1.972959653210902e-05,
+      "loss": 1.0137,
+      "step": 11652
+    },
+    {
+      "epoch": 0.7080017011969135,
+      "grad_norm": 0.1677667200565338,
+      "learning_rate": 1.972197797099245e-05,
+      "loss": 1.1743,
+      "step": 11653
+    },
+    {
+      "epoch": 0.7080624582295401,
+      "grad_norm": 0.13336063921451569,
+      "learning_rate": 1.971436051972263e-05,
+      "loss": 1.0211,
+      "step": 11654
+    },
+    {
+      "epoch": 0.7081232152621666,
+      "grad_norm": 0.10918305814266205,
+      "learning_rate": 1.9706744178578783e-05,
+      "loss": 1.0419,
+      "step": 11655
+    },
+    {
+      "epoch": 0.7081839722947931,
+      "grad_norm": 0.19564831256866455,
+      "learning_rate": 1.969912894784008e-05,
+      "loss": 1.0255,
+      "step": 11656
+    },
+    {
+      "epoch": 0.7082447293274197,
+      "grad_norm": 0.21771742403507233,
+      "learning_rate": 1.9691514827785663e-05,
+      "loss": 1.107,
+      "step": 11657
+    },
+    {
+      "epoch": 0.7083054863600462,
+      "grad_norm": 0.24258442223072052,
+      "learning_rate": 1.9683901818694632e-05,
+      "loss": 1.0283,
+      "step": 11658
+    },
+    {
+      "epoch": 0.7083662433926727,
+      "grad_norm": 0.17653697729110718,
+      "learning_rate": 1.9676289920846038e-05,
+      "loss": 1.0946,
+      "step": 11659
+    },
+    {
+      "epoch": 0.7084270004252993,
+      "grad_norm": 0.12213408946990967,
+      "learning_rate": 1.9668679134518898e-05,
+      "loss": 1.0595,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7084877574579258,
+      "grad_norm": 0.1997223198413849,
+      "learning_rate": 1.966106945999217e-05,
+      "loss": 1.1843,
+      "step": 11661
+    },
+    {
+      "epoch": 0.7085485144905522,
+      "grad_norm": 0.16083401441574097,
+      "learning_rate": 1.965346089754484e-05,
+      "loss": 1.0753,
+      "step": 11662
+    },
+    {
+      "epoch": 0.7086092715231788,
+      "grad_norm": 0.17574000358581543,
+      "learning_rate": 1.9645853447455748e-05,
+      "loss": 1.1772,
+      "step": 11663
+    },
+    {
+      "epoch": 0.7086700285558053,
+      "grad_norm": 0.1727604866027832,
+      "learning_rate": 1.9638247110003765e-05,
+      "loss": 1.1427,
+      "step": 11664
+    },
+    {
+      "epoch": 0.7087307855884318,
+      "grad_norm": 0.10754947364330292,
+      "learning_rate": 1.9630641885467704e-05,
+      "loss": 1.0053,
+      "step": 11665
+    },
+    {
+      "epoch": 0.7087915426210584,
+      "grad_norm": 0.16104000806808472,
+      "learning_rate": 1.9623037774126317e-05,
+      "loss": 1.0288,
+      "step": 11666
+    },
+    {
+      "epoch": 0.7088522996536849,
+      "grad_norm": 0.1376451998949051,
+      "learning_rate": 1.9615434776258372e-05,
+      "loss": 1.0181,
+      "step": 11667
+    },
+    {
+      "epoch": 0.7089130566863114,
+      "grad_norm": 0.14112310111522675,
+      "learning_rate": 1.9607832892142554e-05,
+      "loss": 1.1272,
+      "step": 11668
+    },
+    {
+      "epoch": 0.708973813718938,
+      "grad_norm": 0.12976184487342834,
+      "learning_rate": 1.9600232122057478e-05,
+      "loss": 1.1176,
+      "step": 11669
+    },
+    {
+      "epoch": 0.7090345707515645,
+      "grad_norm": 0.1489325910806656,
+      "learning_rate": 1.9592632466281775e-05,
+      "loss": 1.0262,
+      "step": 11670
+    },
+    {
+      "epoch": 0.709095327784191,
+      "grad_norm": 0.16774222254753113,
+      "learning_rate": 1.9585033925093982e-05,
+      "loss": 1.1054,
+      "step": 11671
+    },
+    {
+      "epoch": 0.7091560848168176,
+      "grad_norm": 0.3156583309173584,
+      "learning_rate": 1.9577436498772666e-05,
+      "loss": 1.0832,
+      "step": 11672
+    },
+    {
+      "epoch": 0.7092168418494441,
+      "grad_norm": 0.13944421708583832,
+      "learning_rate": 1.95698401875963e-05,
+      "loss": 1.0354,
+      "step": 11673
+    },
+    {
+      "epoch": 0.7092775988820706,
+      "grad_norm": 0.22708818316459656,
+      "learning_rate": 1.9562244991843327e-05,
+      "loss": 1.0832,
+      "step": 11674
+    },
+    {
+      "epoch": 0.7093383559146971,
+      "grad_norm": 0.4300410747528076,
+      "learning_rate": 1.955465091179216e-05,
+      "loss": 1.2113,
+      "step": 11675
+    },
+    {
+      "epoch": 0.7093991129473236,
+      "grad_norm": 0.14897175133228302,
+      "learning_rate": 1.9547057947721116e-05,
+      "loss": 1.0562,
+      "step": 11676
+    },
+    {
+      "epoch": 0.7094598699799501,
+      "grad_norm": 0.13382786512374878,
+      "learning_rate": 1.9539466099908565e-05,
+      "loss": 1.0461,
+      "step": 11677
+    },
+    {
+      "epoch": 0.7095206270125767,
+      "grad_norm": 0.1356203854084015,
+      "learning_rate": 1.9531875368632774e-05,
+      "loss": 1.0654,
+      "step": 11678
+    },
+    {
+      "epoch": 0.7095813840452032,
+      "grad_norm": 0.145925372838974,
+      "learning_rate": 1.9524285754171978e-05,
+      "loss": 1.0768,
+      "step": 11679
+    },
+    {
+      "epoch": 0.7096421410778297,
+      "grad_norm": 0.17852981388568878,
+      "learning_rate": 1.9516697256804378e-05,
+      "loss": 1.0411,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7097028981104563,
+      "grad_norm": 0.12320247292518616,
+      "learning_rate": 1.9509109876808135e-05,
+      "loss": 1.0199,
+      "step": 11681
+    },
+    {
+      "epoch": 0.7097636551430828,
+      "grad_norm": 0.11359871923923492,
+      "learning_rate": 1.9501523614461365e-05,
+      "loss": 1.0602,
+      "step": 11682
+    },
+    {
+      "epoch": 0.7098244121757094,
+      "grad_norm": 0.15387961268424988,
+      "learning_rate": 1.949393847004214e-05,
+      "loss": 1.0988,
+      "step": 11683
+    },
+    {
+      "epoch": 0.7098851692083359,
+      "grad_norm": 0.15408551692962646,
+      "learning_rate": 1.9486354443828498e-05,
+      "loss": 1.0873,
+      "step": 11684
+    },
+    {
+      "epoch": 0.7099459262409624,
+      "grad_norm": 0.11309153586626053,
+      "learning_rate": 1.947877153609843e-05,
+      "loss": 1.0026,
+      "step": 11685
+    },
+    {
+      "epoch": 0.710006683273589,
+      "grad_norm": 0.15927071869373322,
+      "learning_rate": 1.947118974712989e-05,
+      "loss": 1.0389,
+      "step": 11686
+    },
+    {
+      "epoch": 0.7100674403062155,
+      "grad_norm": 0.1430070549249649,
+      "learning_rate": 1.9463609077200774e-05,
+      "loss": 1.1247,
+      "step": 11687
+    },
+    {
+      "epoch": 0.7101281973388419,
+      "grad_norm": 0.1528358906507492,
+      "learning_rate": 1.945602952658901e-05,
+      "loss": 0.9834,
+      "step": 11688
+    },
+    {
+      "epoch": 0.7101889543714685,
+      "grad_norm": 0.14306876063346863,
+      "learning_rate": 1.9448451095572362e-05,
+      "loss": 1.1184,
+      "step": 11689
+    },
+    {
+      "epoch": 0.710249711404095,
+      "grad_norm": 1.314124584197998,
+      "learning_rate": 1.944087378442865e-05,
+      "loss": 1.0671,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7103104684367215,
+      "grad_norm": 0.12583403289318085,
+      "learning_rate": 1.9433297593435617e-05,
+      "loss": 1.0415,
+      "step": 11691
+    },
+    {
+      "epoch": 0.710371225469348,
+      "grad_norm": 0.13705484569072723,
+      "learning_rate": 1.9425722522870948e-05,
+      "loss": 1.0628,
+      "step": 11692
+    },
+    {
+      "epoch": 0.7104319825019746,
+      "grad_norm": 0.22917865216732025,
+      "learning_rate": 1.9418148573012352e-05,
+      "loss": 1.1318,
+      "step": 11693
+    },
+    {
+      "epoch": 0.7104927395346011,
+      "grad_norm": 0.21376343071460724,
+      "learning_rate": 1.9410575744137426e-05,
+      "loss": 1.043,
+      "step": 11694
+    },
+    {
+      "epoch": 0.7105534965672277,
+      "grad_norm": 0.12615026533603668,
+      "learning_rate": 1.940300403652378e-05,
+      "loss": 1.0492,
+      "step": 11695
+    },
+    {
+      "epoch": 0.7106142535998542,
+      "grad_norm": 4.3653106689453125,
+      "learning_rate": 1.9395433450448914e-05,
+      "loss": 1.2242,
+      "step": 11696
+    },
+    {
+      "epoch": 0.7106750106324807,
+      "grad_norm": 0.1611001193523407,
+      "learning_rate": 1.9387863986190337e-05,
+      "loss": 1.0784,
+      "step": 11697
+    },
+    {
+      "epoch": 0.7107357676651073,
+      "grad_norm": 0.179573193192482,
+      "learning_rate": 1.9380295644025538e-05,
+      "loss": 1.0448,
+      "step": 11698
+    },
+    {
+      "epoch": 0.7107965246977338,
+      "grad_norm": 0.11063571274280548,
+      "learning_rate": 1.9372728424231923e-05,
+      "loss": 1.0488,
+      "step": 11699
+    },
+    {
+      "epoch": 0.7108572817303603,
+      "grad_norm": 0.16249185800552368,
+      "learning_rate": 1.9365162327086867e-05,
+      "loss": 1.0264,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7109180387629869,
+      "grad_norm": 0.28120604157447815,
+      "learning_rate": 1.9357597352867706e-05,
+      "loss": 1.0068,
+      "step": 11701
+    },
+    {
+      "epoch": 0.7109787957956133,
+      "grad_norm": 0.13428813219070435,
+      "learning_rate": 1.9350033501851737e-05,
+      "loss": 1.0645,
+      "step": 11702
+    },
+    {
+      "epoch": 0.7110395528282398,
+      "grad_norm": 0.12566058337688446,
+      "learning_rate": 1.9342470774316217e-05,
+      "loss": 1.0386,
+      "step": 11703
+    },
+    {
+      "epoch": 0.7111003098608664,
+      "grad_norm": 0.1819542646408081,
+      "learning_rate": 1.9334909170538356e-05,
+      "loss": 1.01,
+      "step": 11704
+    },
+    {
+      "epoch": 0.7111610668934929,
+      "grad_norm": 0.24413032829761505,
+      "learning_rate": 1.9327348690795326e-05,
+      "loss": 1.1124,
+      "step": 11705
+    },
+    {
+      "epoch": 0.7112218239261194,
+      "grad_norm": 0.15328948199748993,
+      "learning_rate": 1.9319789335364258e-05,
+      "loss": 1.0426,
+      "step": 11706
+    },
+    {
+      "epoch": 0.711282580958746,
+      "grad_norm": 0.17072217166423798,
+      "learning_rate": 1.9312231104522242e-05,
+      "loss": 1.0202,
+      "step": 11707
+    },
+    {
+      "epoch": 0.7113433379913725,
+      "grad_norm": 0.1225581243634224,
+      "learning_rate": 1.9304673998546312e-05,
+      "loss": 1.0133,
+      "step": 11708
+    },
+    {
+      "epoch": 0.711404095023999,
+      "grad_norm": 0.25380784273147583,
+      "learning_rate": 1.9297118017713524e-05,
+      "loss": 1.225,
+      "step": 11709
+    },
+    {
+      "epoch": 0.7114648520566256,
+      "grad_norm": 0.1189921498298645,
+      "learning_rate": 1.9289563162300787e-05,
+      "loss": 1.0839,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7115256090892521,
+      "grad_norm": 0.21586130559444427,
+      "learning_rate": 1.9282009432585056e-05,
+      "loss": 1.1584,
+      "step": 11711
+    },
+    {
+      "epoch": 0.7115863661218786,
+      "grad_norm": 0.21023230254650116,
+      "learning_rate": 1.9274456828843206e-05,
+      "loss": 1.0591,
+      "step": 11712
+    },
+    {
+      "epoch": 0.7116471231545052,
+      "grad_norm": 0.10978969186544418,
+      "learning_rate": 1.926690535135206e-05,
+      "loss": 0.9849,
+      "step": 11713
+    },
+    {
+      "epoch": 0.7117078801871317,
+      "grad_norm": 0.17131248116493225,
+      "learning_rate": 1.9259355000388457e-05,
+      "loss": 1.1152,
+      "step": 11714
+    },
+    {
+      "epoch": 0.7117686372197581,
+      "grad_norm": 0.2565372884273529,
+      "learning_rate": 1.9251805776229153e-05,
+      "loss": 1.1037,
+      "step": 11715
+    },
+    {
+      "epoch": 0.7118293942523847,
+      "grad_norm": 0.13198554515838623,
+      "learning_rate": 1.924425767915084e-05,
+      "loss": 1.0037,
+      "step": 11716
+    },
+    {
+      "epoch": 0.7118901512850112,
+      "grad_norm": 0.1634645015001297,
+      "learning_rate": 1.92367107094302e-05,
+      "loss": 1.0472,
+      "step": 11717
+    },
+    {
+      "epoch": 0.7119509083176377,
+      "grad_norm": 0.11387443542480469,
+      "learning_rate": 1.9229164867343863e-05,
+      "loss": 1.0396,
+      "step": 11718
+    },
+    {
+      "epoch": 0.7120116653502643,
+      "grad_norm": 0.11026987433433533,
+      "learning_rate": 1.922162015316845e-05,
+      "loss": 1.0756,
+      "step": 11719
+    },
+    {
+      "epoch": 0.7120724223828908,
+      "grad_norm": 0.10341232270002365,
+      "learning_rate": 1.9214076567180506e-05,
+      "loss": 0.9706,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7121331794155173,
+      "grad_norm": 0.1101595014333725,
+      "learning_rate": 1.9206534109656532e-05,
+      "loss": 0.9948,
+      "step": 11721
+    },
+    {
+      "epoch": 0.7121939364481439,
+      "grad_norm": 0.17904014885425568,
+      "learning_rate": 1.9198992780873015e-05,
+      "loss": 1.0907,
+      "step": 11722
+    },
+    {
+      "epoch": 0.7122546934807704,
+      "grad_norm": 0.12634815275669098,
+      "learning_rate": 1.9191452581106346e-05,
+      "loss": 1.0334,
+      "step": 11723
+    },
+    {
+      "epoch": 0.712315450513397,
+      "grad_norm": 0.19385063648223877,
+      "learning_rate": 1.9183913510632945e-05,
+      "loss": 1.0873,
+      "step": 11724
+    },
+    {
+      "epoch": 0.7123762075460235,
+      "grad_norm": 0.1293027549982071,
+      "learning_rate": 1.917637556972916e-05,
+      "loss": 1.022,
+      "step": 11725
+    },
+    {
+      "epoch": 0.71243696457865,
+      "grad_norm": 0.12360483407974243,
+      "learning_rate": 1.9168838758671282e-05,
+      "loss": 1.0981,
+      "step": 11726
+    },
+    {
+      "epoch": 0.7124977216112766,
+      "grad_norm": 0.17805059254169464,
+      "learning_rate": 1.9161303077735575e-05,
+      "loss": 1.1905,
+      "step": 11727
+    },
+    {
+      "epoch": 0.712558478643903,
+      "grad_norm": 0.12121447920799255,
+      "learning_rate": 1.9153768527198272e-05,
+      "loss": 1.0508,
+      "step": 11728
+    },
+    {
+      "epoch": 0.7126192356765295,
+      "grad_norm": 0.1323792040348053,
+      "learning_rate": 1.9146235107335543e-05,
+      "loss": 1.0333,
+      "step": 11729
+    },
+    {
+      "epoch": 0.712679992709156,
+      "grad_norm": 0.2845796048641205,
+      "learning_rate": 1.913870281842353e-05,
+      "loss": 1.1417,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7127407497417826,
+      "grad_norm": 0.11830468475818634,
+      "learning_rate": 1.9131171660738335e-05,
+      "loss": 1.0567,
+      "step": 11731
+    },
+    {
+      "epoch": 0.7128015067744091,
+      "grad_norm": 0.2766297459602356,
+      "learning_rate": 1.9123641634556e-05,
+      "loss": 1.1617,
+      "step": 11732
+    },
+    {
+      "epoch": 0.7128622638070357,
+      "grad_norm": 0.21680931746959686,
+      "learning_rate": 1.911611274015256e-05,
+      "loss": 1.1791,
+      "step": 11733
+    },
+    {
+      "epoch": 0.7129230208396622,
+      "grad_norm": 0.1302758753299713,
+      "learning_rate": 1.910858497780395e-05,
+      "loss": 1.0951,
+      "step": 11734
+    },
+    {
+      "epoch": 0.7129837778722887,
+      "grad_norm": 0.3375566601753235,
+      "learning_rate": 1.9101058347786173e-05,
+      "loss": 1.0107,
+      "step": 11735
+    },
+    {
+      "epoch": 0.7130445349049153,
+      "grad_norm": 0.1535346359014511,
+      "learning_rate": 1.909353285037505e-05,
+      "loss": 1.0656,
+      "step": 11736
+    },
+    {
+      "epoch": 0.7131052919375418,
+      "grad_norm": 0.13433580100536346,
+      "learning_rate": 1.908600848584646e-05,
+      "loss": 1.0545,
+      "step": 11737
+    },
+    {
+      "epoch": 0.7131660489701683,
+      "grad_norm": 0.12825262546539307,
+      "learning_rate": 1.9078485254476198e-05,
+      "loss": 1.0438,
+      "step": 11738
+    },
+    {
+      "epoch": 0.7132268060027949,
+      "grad_norm": 0.1130940392613411,
+      "learning_rate": 1.9070963156540023e-05,
+      "loss": 1.0711,
+      "step": 11739
+    },
+    {
+      "epoch": 0.7132875630354214,
+      "grad_norm": 0.17816604673862457,
+      "learning_rate": 1.9063442192313687e-05,
+      "loss": 1.0674,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7133483200680478,
+      "grad_norm": 0.16936558485031128,
+      "learning_rate": 1.9055922362072858e-05,
+      "loss": 1.1216,
+      "step": 11741
+    },
+    {
+      "epoch": 0.7134090771006744,
+      "grad_norm": 0.36161476373672485,
+      "learning_rate": 1.9048403666093195e-05,
+      "loss": 1.1203,
+      "step": 11742
+    },
+    {
+      "epoch": 0.7134698341333009,
+      "grad_norm": 0.24623535573482513,
+      "learning_rate": 1.904088610465024e-05,
+      "loss": 1.0673,
+      "step": 11743
+    },
+    {
+      "epoch": 0.7135305911659274,
+      "grad_norm": 0.16006368398666382,
+      "learning_rate": 1.9033369678019613e-05,
+      "loss": 1.0809,
+      "step": 11744
+    },
+    {
+      "epoch": 0.713591348198554,
+      "grad_norm": 0.27312740683555603,
+      "learning_rate": 1.9025854386476798e-05,
+      "loss": 1.1195,
+      "step": 11745
+    },
+    {
+      "epoch": 0.7136521052311805,
+      "grad_norm": 0.1588553935289383,
+      "learning_rate": 1.9018340230297288e-05,
+      "loss": 1.1176,
+      "step": 11746
+    },
+    {
+      "epoch": 0.713712862263807,
+      "grad_norm": 0.1724746823310852,
+      "learning_rate": 1.90108272097565e-05,
+      "loss": 1.0724,
+      "step": 11747
+    },
+    {
+      "epoch": 0.7137736192964336,
+      "grad_norm": 0.18205776810646057,
+      "learning_rate": 1.900331532512983e-05,
+      "loss": 1.1882,
+      "step": 11748
+    },
+    {
+      "epoch": 0.7138343763290601,
+      "grad_norm": 0.11100442707538605,
+      "learning_rate": 1.8995804576692627e-05,
+      "loss": 1.066,
+      "step": 11749
+    },
+    {
+      "epoch": 0.7138951333616866,
+      "grad_norm": 0.12582969665527344,
+      "learning_rate": 1.89882949647202e-05,
+      "loss": 1.0308,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7139558903943132,
+      "grad_norm": 0.366331547498703,
+      "learning_rate": 1.898078648948782e-05,
+      "loss": 1.049,
+      "step": 11751
+    },
+    {
+      "epoch": 0.7140166474269397,
+      "grad_norm": 0.16668683290481567,
+      "learning_rate": 1.8973279151270702e-05,
+      "loss": 0.9968,
+      "step": 11752
+    },
+    {
+      "epoch": 0.7140774044595662,
+      "grad_norm": 0.23617713153362274,
+      "learning_rate": 1.8965772950344042e-05,
+      "loss": 1.0512,
+      "step": 11753
+    },
+    {
+      "epoch": 0.7141381614921927,
+      "grad_norm": 0.12710700929164886,
+      "learning_rate": 1.8958267886982945e-05,
+      "loss": 1.0293,
+      "step": 11754
+    },
+    {
+      "epoch": 0.7141989185248192,
+      "grad_norm": 0.1962948441505432,
+      "learning_rate": 1.8950763961462582e-05,
+      "loss": 1.2104,
+      "step": 11755
+    },
+    {
+      "epoch": 0.7142596755574457,
+      "grad_norm": 0.19061042368412018,
+      "learning_rate": 1.8943261174057948e-05,
+      "loss": 1.0444,
+      "step": 11756
+    },
+    {
+      "epoch": 0.7143204325900723,
+      "grad_norm": 0.18270787596702576,
+      "learning_rate": 1.893575952504408e-05,
+      "loss": 1.0478,
+      "step": 11757
+    },
+    {
+      "epoch": 0.7143811896226988,
+      "grad_norm": 0.1116868332028389,
+      "learning_rate": 1.8928259014695948e-05,
+      "loss": 1.0138,
+      "step": 11758
+    },
+    {
+      "epoch": 0.7144419466553253,
+      "grad_norm": 0.1393412947654724,
+      "learning_rate": 1.892075964328847e-05,
+      "loss": 1.1346,
+      "step": 11759
+    },
+    {
+      "epoch": 0.7145027036879519,
+      "grad_norm": 0.4019707143306732,
+      "learning_rate": 1.8913261411096572e-05,
+      "loss": 1.0636,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7145634607205784,
+      "grad_norm": 0.1373920887708664,
+      "learning_rate": 1.8905764318395093e-05,
+      "loss": 1.0183,
+      "step": 11761
+    },
+    {
+      "epoch": 0.714624217753205,
+      "grad_norm": 0.15592533349990845,
+      "learning_rate": 1.8898268365458844e-05,
+      "loss": 1.3152,
+      "step": 11762
+    },
+    {
+      "epoch": 0.7146849747858315,
+      "grad_norm": 0.1144845113158226,
+      "learning_rate": 1.889077355256256e-05,
+      "loss": 1.0361,
+      "step": 11763
+    },
+    {
+      "epoch": 0.714745731818458,
+      "grad_norm": 0.12774446606636047,
+      "learning_rate": 1.8883279879980976e-05,
+      "loss": 1.037,
+      "step": 11764
+    },
+    {
+      "epoch": 0.7148064888510846,
+      "grad_norm": 0.14454355835914612,
+      "learning_rate": 1.8875787347988793e-05,
+      "loss": 1.0525,
+      "step": 11765
+    },
+    {
+      "epoch": 0.7148672458837111,
+      "grad_norm": 0.19021962583065033,
+      "learning_rate": 1.8868295956860642e-05,
+      "loss": 1.1999,
+      "step": 11766
+    },
+    {
+      "epoch": 0.7149280029163375,
+      "grad_norm": 0.1694939285516739,
+      "learning_rate": 1.886080570687112e-05,
+      "loss": 1.0812,
+      "step": 11767
+    },
+    {
+      "epoch": 0.714988759948964,
+      "grad_norm": 0.19839046895503998,
+      "learning_rate": 1.885331659829479e-05,
+      "loss": 1.1567,
+      "step": 11768
+    },
+    {
+      "epoch": 0.7150495169815906,
+      "grad_norm": 0.142764151096344,
+      "learning_rate": 1.8845828631406153e-05,
+      "loss": 1.0673,
+      "step": 11769
+    },
+    {
+      "epoch": 0.7151102740142171,
+      "grad_norm": 0.14924031496047974,
+      "learning_rate": 1.883834180647969e-05,
+      "loss": 1.0916,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7151710310468437,
+      "grad_norm": 0.18129679560661316,
+      "learning_rate": 1.8830856123789836e-05,
+      "loss": 1.0506,
+      "step": 11771
+    },
+    {
+      "epoch": 0.7152317880794702,
+      "grad_norm": 0.1394844800233841,
+      "learning_rate": 1.8823371583610977e-05,
+      "loss": 1.0792,
+      "step": 11772
+    },
+    {
+      "epoch": 0.7152925451120967,
+      "grad_norm": 0.1567200869321823,
+      "learning_rate": 1.8815888186217457e-05,
+      "loss": 1.0068,
+      "step": 11773
+    },
+    {
+      "epoch": 0.7153533021447233,
+      "grad_norm": 0.17005854845046997,
+      "learning_rate": 1.8808405931883583e-05,
+      "loss": 1.0275,
+      "step": 11774
+    },
+    {
+      "epoch": 0.7154140591773498,
+      "grad_norm": 0.1875138133764267,
+      "learning_rate": 1.8800924820883597e-05,
+      "loss": 1.1628,
+      "step": 11775
+    },
+    {
+      "epoch": 0.7154748162099763,
+      "grad_norm": 0.12211021035909653,
+      "learning_rate": 1.8793444853491783e-05,
+      "loss": 1.0867,
+      "step": 11776
+    },
+    {
+      "epoch": 0.7155355732426029,
+      "grad_norm": 0.15800920128822327,
+      "learning_rate": 1.878596602998226e-05,
+      "loss": 1.1003,
+      "step": 11777
+    },
+    {
+      "epoch": 0.7155963302752294,
+      "grad_norm": 0.15006932616233826,
+      "learning_rate": 1.8778488350629187e-05,
+      "loss": 1.0562,
+      "step": 11778
+    },
+    {
+      "epoch": 0.7156570873078559,
+      "grad_norm": 0.16724249720573425,
+      "learning_rate": 1.8771011815706657e-05,
+      "loss": 1.089,
+      "step": 11779
+    },
+    {
+      "epoch": 0.7157178443404824,
+      "grad_norm": 0.17960305511951447,
+      "learning_rate": 1.8763536425488704e-05,
+      "loss": 1.0965,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7157786013731089,
+      "grad_norm": 0.2078692466020584,
+      "learning_rate": 1.8756062180249385e-05,
+      "loss": 1.1419,
+      "step": 11781
+    },
+    {
+      "epoch": 0.7158393584057354,
+      "grad_norm": 0.16367106139659882,
+      "learning_rate": 1.8748589080262662e-05,
+      "loss": 1.0788,
+      "step": 11782
+    },
+    {
+      "epoch": 0.715900115438362,
+      "grad_norm": 0.4139288067817688,
+      "learning_rate": 1.8741117125802427e-05,
+      "loss": 1.1373,
+      "step": 11783
+    },
+    {
+      "epoch": 0.7159608724709885,
+      "grad_norm": 0.1478399634361267,
+      "learning_rate": 1.873364631714259e-05,
+      "loss": 1.095,
+      "step": 11784
+    },
+    {
+      "epoch": 0.716021629503615,
+      "grad_norm": 0.2829451858997345,
+      "learning_rate": 1.872617665455697e-05,
+      "loss": 1.0528,
+      "step": 11785
+    },
+    {
+      "epoch": 0.7160823865362416,
+      "grad_norm": 0.1564915031194687,
+      "learning_rate": 1.871870813831941e-05,
+      "loss": 1.0406,
+      "step": 11786
+    },
+    {
+      "epoch": 0.7161431435688681,
+      "grad_norm": 0.2062319666147232,
+      "learning_rate": 1.8711240768703647e-05,
+      "loss": 1.1469,
+      "step": 11787
+    },
+    {
+      "epoch": 0.7162039006014946,
+      "grad_norm": 0.1241144984960556,
+      "learning_rate": 1.8703774545983405e-05,
+      "loss": 1.036,
+      "step": 11788
+    },
+    {
+      "epoch": 0.7162646576341212,
+      "grad_norm": 0.10204458236694336,
+      "learning_rate": 1.8696309470432372e-05,
+      "loss": 1.0282,
+      "step": 11789
+    },
+    {
+      "epoch": 0.7163254146667477,
+      "grad_norm": 0.16363711655139923,
+      "learning_rate": 1.868884554232413e-05,
+      "loss": 1.1376,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7163861716993742,
+      "grad_norm": 0.12379411607980728,
+      "learning_rate": 1.8681382761932324e-05,
+      "loss": 1.0577,
+      "step": 11791
+    },
+    {
+      "epoch": 0.7164469287320008,
+      "grad_norm": 0.19269931316375732,
+      "learning_rate": 1.8673921129530492e-05,
+      "loss": 1.0754,
+      "step": 11792
+    },
+    {
+      "epoch": 0.7165076857646272,
+      "grad_norm": 0.14565134048461914,
+      "learning_rate": 1.866646064539213e-05,
+      "loss": 1.0789,
+      "step": 11793
+    },
+    {
+      "epoch": 0.7165684427972537,
+      "grad_norm": 0.1668594628572464,
+      "learning_rate": 1.8659001309790715e-05,
+      "loss": 1.2008,
+      "step": 11794
+    },
+    {
+      "epoch": 0.7166291998298803,
+      "grad_norm": 0.1097547635436058,
+      "learning_rate": 1.865154312299966e-05,
+      "loss": 1.0236,
+      "step": 11795
+    },
+    {
+      "epoch": 0.7166899568625068,
+      "grad_norm": 0.17278924584388733,
+      "learning_rate": 1.8644086085292357e-05,
+      "loss": 1.1165,
+      "step": 11796
+    },
+    {
+      "epoch": 0.7167507138951333,
+      "grad_norm": 0.20955996215343475,
+      "learning_rate": 1.8636630196942136e-05,
+      "loss": 1.094,
+      "step": 11797
+    },
+    {
+      "epoch": 0.7168114709277599,
+      "grad_norm": 0.15809226036071777,
+      "learning_rate": 1.8629175458222293e-05,
+      "loss": 1.0972,
+      "step": 11798
+    },
+    {
+      "epoch": 0.7168722279603864,
+      "grad_norm": 0.22863392531871796,
+      "learning_rate": 1.8621721869406095e-05,
+      "loss": 1.1475,
+      "step": 11799
+    },
+    {
+      "epoch": 0.716932984993013,
+      "grad_norm": 0.12562121450901031,
+      "learning_rate": 1.861426943076674e-05,
+      "loss": 1.0516,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7169937420256395,
+      "grad_norm": 0.1876962035894394,
+      "learning_rate": 1.8606818142577392e-05,
+      "loss": 1.1917,
+      "step": 11801
+    },
+    {
+      "epoch": 0.717054499058266,
+      "grad_norm": 0.27948957681655884,
+      "learning_rate": 1.8599368005111233e-05,
+      "loss": 1.2759,
+      "step": 11802
+    },
+    {
+      "epoch": 0.7171152560908925,
+      "grad_norm": 0.3056049644947052,
+      "learning_rate": 1.8591919018641284e-05,
+      "loss": 1.2254,
+      "step": 11803
+    },
+    {
+      "epoch": 0.7171760131235191,
+      "grad_norm": 1.7359598875045776,
+      "learning_rate": 1.8584471183440623e-05,
+      "loss": 1.1066,
+      "step": 11804
+    },
+    {
+      "epoch": 0.7172367701561456,
+      "grad_norm": 0.12048758566379547,
+      "learning_rate": 1.8577024499782237e-05,
+      "loss": 1.0475,
+      "step": 11805
+    },
+    {
+      "epoch": 0.7172975271887722,
+      "grad_norm": 0.13789333403110504,
+      "learning_rate": 1.8569578967939073e-05,
+      "loss": 1.056,
+      "step": 11806
+    },
+    {
+      "epoch": 0.7173582842213986,
+      "grad_norm": 0.15338440239429474,
+      "learning_rate": 1.856213458818409e-05,
+      "loss": 1.1763,
+      "step": 11807
+    },
+    {
+      "epoch": 0.7174190412540251,
+      "grad_norm": 0.13651174306869507,
+      "learning_rate": 1.855469136079014e-05,
+      "loss": 1.122,
+      "step": 11808
+    },
+    {
+      "epoch": 0.7174797982866516,
+      "grad_norm": 0.14389897882938385,
+      "learning_rate": 1.8547249286030065e-05,
+      "loss": 1.0688,
+      "step": 11809
+    },
+    {
+      "epoch": 0.7175405553192782,
+      "grad_norm": 0.10650651901960373,
+      "learning_rate": 1.8539808364176635e-05,
+      "loss": 1.0866,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7176013123519047,
+      "grad_norm": 0.11491388827562332,
+      "learning_rate": 1.85323685955026e-05,
+      "loss": 1.0276,
+      "step": 11811
+    },
+    {
+      "epoch": 0.7176620693845313,
+      "grad_norm": 0.14369851350784302,
+      "learning_rate": 1.8524929980280685e-05,
+      "loss": 1.0363,
+      "step": 11812
+    },
+    {
+      "epoch": 0.7177228264171578,
+      "grad_norm": 0.17646241188049316,
+      "learning_rate": 1.851749251878355e-05,
+      "loss": 1.1339,
+      "step": 11813
+    },
+    {
+      "epoch": 0.7177835834497843,
+      "grad_norm": 0.1842593103647232,
+      "learning_rate": 1.851005621128381e-05,
+      "loss": 1.114,
+      "step": 11814
+    },
+    {
+      "epoch": 0.7178443404824109,
+      "grad_norm": 0.18943586945533752,
+      "learning_rate": 1.8502621058054048e-05,
+      "loss": 1.0658,
+      "step": 11815
+    },
+    {
+      "epoch": 0.7179050975150374,
+      "grad_norm": 0.11217565834522247,
+      "learning_rate": 1.8495187059366804e-05,
+      "loss": 1.0307,
+      "step": 11816
+    },
+    {
+      "epoch": 0.7179658545476639,
+      "grad_norm": 0.11643880605697632,
+      "learning_rate": 1.8487754215494564e-05,
+      "loss": 1.0323,
+      "step": 11817
+    },
+    {
+      "epoch": 0.7180266115802905,
+      "grad_norm": 0.11993326991796494,
+      "learning_rate": 1.8480322526709783e-05,
+      "loss": 0.9903,
+      "step": 11818
+    },
+    {
+      "epoch": 0.718087368612917,
+      "grad_norm": 0.14355207979679108,
+      "learning_rate": 1.8472891993284867e-05,
+      "loss": 1.0061,
+      "step": 11819
+    },
+    {
+      "epoch": 0.7181481256455434,
+      "grad_norm": 0.23948566615581512,
+      "learning_rate": 1.8465462615492196e-05,
+      "loss": 1.187,
+      "step": 11820
+    },
+    {
+      "epoch": 0.71820888267817,
+      "grad_norm": 0.5445346236228943,
+      "learning_rate": 1.845803439360409e-05,
+      "loss": 1.0417,
+      "step": 11821
+    },
+    {
+      "epoch": 0.7182696397107965,
+      "grad_norm": 0.13110193610191345,
+      "learning_rate": 1.845060732789283e-05,
+      "loss": 1.0378,
+      "step": 11822
+    },
+    {
+      "epoch": 0.718330396743423,
+      "grad_norm": 0.12356985360383987,
+      "learning_rate": 1.844318141863065e-05,
+      "loss": 1.0175,
+      "step": 11823
+    },
+    {
+      "epoch": 0.7183911537760496,
+      "grad_norm": 0.18453015387058258,
+      "learning_rate": 1.843575666608976e-05,
+      "loss": 1.0694,
+      "step": 11824
+    },
+    {
+      "epoch": 0.7184519108086761,
+      "grad_norm": 0.17998893558979034,
+      "learning_rate": 1.8428333070542308e-05,
+      "loss": 1.0021,
+      "step": 11825
+    },
+    {
+      "epoch": 0.7185126678413026,
+      "grad_norm": 0.10383981466293335,
+      "learning_rate": 1.8420910632260414e-05,
+      "loss": 0.9852,
+      "step": 11826
+    },
+    {
+      "epoch": 0.7185734248739292,
+      "grad_norm": 0.17593352496623993,
+      "learning_rate": 1.8413489351516128e-05,
+      "loss": 1.0634,
+      "step": 11827
+    },
+    {
+      "epoch": 0.7186341819065557,
+      "grad_norm": 0.12185440212488174,
+      "learning_rate": 1.840606922858151e-05,
+      "loss": 1.0412,
+      "step": 11828
+    },
+    {
+      "epoch": 0.7186949389391822,
+      "grad_norm": 0.2973095774650574,
+      "learning_rate": 1.8398650263728547e-05,
+      "loss": 1.0959,
+      "step": 11829
+    },
+    {
+      "epoch": 0.7187556959718088,
+      "grad_norm": 1.0236891508102417,
+      "learning_rate": 1.8391232457229153e-05,
+      "loss": 1.062,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7188164530044353,
+      "grad_norm": 0.15837439894676208,
+      "learning_rate": 1.8383815809355237e-05,
+      "loss": 0.9918,
+      "step": 11831
+    },
+    {
+      "epoch": 0.7188772100370618,
+      "grad_norm": 0.1673840433359146,
+      "learning_rate": 1.8376400320378654e-05,
+      "loss": 1.1054,
+      "step": 11832
+    },
+    {
+      "epoch": 0.7189379670696883,
+      "grad_norm": 0.17833562195301056,
+      "learning_rate": 1.836898599057124e-05,
+      "loss": 1.118,
+      "step": 11833
+    },
+    {
+      "epoch": 0.7189987241023148,
+      "grad_norm": 0.5794760584831238,
+      "learning_rate": 1.8361572820204768e-05,
+      "loss": 0.9715,
+      "step": 11834
+    },
+    {
+      "epoch": 0.7190594811349413,
+      "grad_norm": 0.13030590116977692,
+      "learning_rate": 1.8354160809550952e-05,
+      "loss": 1.0605,
+      "step": 11835
+    },
+    {
+      "epoch": 0.7191202381675679,
+      "grad_norm": 28.77001190185547,
+      "learning_rate": 1.8346749958881492e-05,
+      "loss": 1.0651,
+      "step": 11836
+    },
+    {
+      "epoch": 0.7191809952001944,
+      "grad_norm": 29.64858055114746,
+      "learning_rate": 1.833934026846803e-05,
+      "loss": 1.0844,
+      "step": 11837
+    },
+    {
+      "epoch": 0.7192417522328209,
+      "grad_norm": 0.1516307145357132,
+      "learning_rate": 1.833193173858217e-05,
+      "loss": 1.0323,
+      "step": 11838
+    },
+    {
+      "epoch": 0.7193025092654475,
+      "grad_norm": 0.14092974364757538,
+      "learning_rate": 1.8324524369495476e-05,
+      "loss": 1.0861,
+      "step": 11839
+    },
+    {
+      "epoch": 0.719363266298074,
+      "grad_norm": 1.4835987091064453,
+      "learning_rate": 1.8317118161479463e-05,
+      "loss": 1.0908,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7194240233307005,
+      "grad_norm": 0.22931554913520813,
+      "learning_rate": 1.8309713114805603e-05,
+      "loss": 1.1115,
+      "step": 11841
+    },
+    {
+      "epoch": 0.7194847803633271,
+      "grad_norm": 0.11138559132814407,
+      "learning_rate": 1.830230922974534e-05,
+      "loss": 0.9981,
+      "step": 11842
+    },
+    {
+      "epoch": 0.7195455373959536,
+      "grad_norm": 0.15380050241947174,
+      "learning_rate": 1.829490650657006e-05,
+      "loss": 1.2166,
+      "step": 11843
+    },
+    {
+      "epoch": 0.7196062944285802,
+      "grad_norm": 0.14814504981040955,
+      "learning_rate": 1.8287504945551104e-05,
+      "loss": 1.0389,
+      "step": 11844
+    },
+    {
+      "epoch": 0.7196670514612067,
+      "grad_norm": 0.197509765625,
+      "learning_rate": 1.828010454695979e-05,
+      "loss": 1.017,
+      "step": 11845
+    },
+    {
+      "epoch": 0.7197278084938331,
+      "grad_norm": 0.1280500888824463,
+      "learning_rate": 1.8272705311067374e-05,
+      "loss": 1.0313,
+      "step": 11846
+    },
+    {
+      "epoch": 0.7197885655264596,
+      "grad_norm": 0.8988205194473267,
+      "learning_rate": 1.8265307238145074e-05,
+      "loss": 1.0784,
+      "step": 11847
+    },
+    {
+      "epoch": 0.7198493225590862,
+      "grad_norm": 0.3101656436920166,
+      "learning_rate": 1.8257910328464057e-05,
+      "loss": 1.137,
+      "step": 11848
+    },
+    {
+      "epoch": 0.7199100795917127,
+      "grad_norm": 0.24984928965568542,
+      "learning_rate": 1.825051458229551e-05,
+      "loss": 1.0211,
+      "step": 11849
+    },
+    {
+      "epoch": 0.7199708366243392,
+      "grad_norm": 0.3139202296733856,
+      "learning_rate": 1.824311999991046e-05,
+      "loss": 1.0751,
+      "step": 11850
+    },
+    {
+      "epoch": 0.7200315936569658,
+      "grad_norm": 0.1468210220336914,
+      "learning_rate": 1.8235726581579998e-05,
+      "loss": 0.9974,
+      "step": 11851
+    },
+    {
+      "epoch": 0.7200923506895923,
+      "grad_norm": 0.10285634547472,
+      "learning_rate": 1.8228334327575097e-05,
+      "loss": 0.9946,
+      "step": 11852
+    },
+    {
+      "epoch": 0.7201531077222189,
+      "grad_norm": 0.5532729625701904,
+      "learning_rate": 1.8220943238166754e-05,
+      "loss": 1.0921,
+      "step": 11853
+    },
+    {
+      "epoch": 0.7202138647548454,
+      "grad_norm": 0.12387371063232422,
+      "learning_rate": 1.8213553313625888e-05,
+      "loss": 1.0732,
+      "step": 11854
+    },
+    {
+      "epoch": 0.7202746217874719,
+      "grad_norm": 0.1855122447013855,
+      "learning_rate": 1.8206164554223358e-05,
+      "loss": 1.1238,
+      "step": 11855
+    },
+    {
+      "epoch": 0.7203353788200985,
+      "grad_norm": 0.10577202588319778,
+      "learning_rate": 1.819877696023004e-05,
+      "loss": 1.0391,
+      "step": 11856
+    },
+    {
+      "epoch": 0.720396135852725,
+      "grad_norm": 0.23974394798278809,
+      "learning_rate": 1.8191390531916663e-05,
+      "loss": 1.0801,
+      "step": 11857
+    },
+    {
+      "epoch": 0.7204568928853515,
+      "grad_norm": 0.22621716558933258,
+      "learning_rate": 1.8184005269554027e-05,
+      "loss": 1.1183,
+      "step": 11858
+    },
+    {
+      "epoch": 0.720517649917978,
+      "grad_norm": 0.2566772401332855,
+      "learning_rate": 1.8176621173412834e-05,
+      "loss": 1.1688,
+      "step": 11859
+    },
+    {
+      "epoch": 0.7205784069506045,
+      "grad_norm": 0.21181346476078033,
+      "learning_rate": 1.8169238243763735e-05,
+      "loss": 1.0364,
+      "step": 11860
+    },
+    {
+      "epoch": 0.720639163983231,
+      "grad_norm": 0.6864669919013977,
+      "learning_rate": 1.8161856480877365e-05,
+      "loss": 1.0518,
+      "step": 11861
+    },
+    {
+      "epoch": 0.7206999210158576,
+      "grad_norm": 0.14368070662021637,
+      "learning_rate": 1.81544758850243e-05,
+      "loss": 1.0893,
+      "step": 11862
+    },
+    {
+      "epoch": 0.7207606780484841,
+      "grad_norm": 0.19693714380264282,
+      "learning_rate": 1.814709645647508e-05,
+      "loss": 1.1898,
+      "step": 11863
+    },
+    {
+      "epoch": 0.7208214350811106,
+      "grad_norm": 0.23038814961910248,
+      "learning_rate": 1.8139718195500193e-05,
+      "loss": 1.1314,
+      "step": 11864
+    },
+    {
+      "epoch": 0.7208821921137372,
+      "grad_norm": 0.1323247104883194,
+      "learning_rate": 1.81323411023701e-05,
+      "loss": 1.0347,
+      "step": 11865
+    },
+    {
+      "epoch": 0.7209429491463637,
+      "grad_norm": 0.1980310082435608,
+      "learning_rate": 1.8124965177355195e-05,
+      "loss": 1.0538,
+      "step": 11866
+    },
+    {
+      "epoch": 0.7210037061789902,
+      "grad_norm": 0.12161320447921753,
+      "learning_rate": 1.8117590420725856e-05,
+      "loss": 1.0722,
+      "step": 11867
+    },
+    {
+      "epoch": 0.7210644632116168,
+      "grad_norm": 0.11527378112077713,
+      "learning_rate": 1.8110216832752387e-05,
+      "loss": 1.0053,
+      "step": 11868
+    },
+    {
+      "epoch": 0.7211252202442433,
+      "grad_norm": 0.2602807283401489,
+      "learning_rate": 1.8102844413705117e-05,
+      "loss": 1.0766,
+      "step": 11869
+    },
+    {
+      "epoch": 0.7211859772768698,
+      "grad_norm": 0.13532786071300507,
+      "learning_rate": 1.8095473163854237e-05,
+      "loss": 1.0631,
+      "step": 11870
+    },
+    {
+      "epoch": 0.7212467343094964,
+      "grad_norm": 0.10426671802997589,
+      "learning_rate": 1.808810308346995e-05,
+      "loss": 1.0461,
+      "step": 11871
+    },
+    {
+      "epoch": 0.7213074913421228,
+      "grad_norm": 0.1360999047756195,
+      "learning_rate": 1.8080734172822417e-05,
+      "loss": 1.0698,
+      "step": 11872
+    },
+    {
+      "epoch": 0.7213682483747493,
+      "grad_norm": 0.10016483813524246,
+      "learning_rate": 1.807336643218172e-05,
+      "loss": 1.0322,
+      "step": 11873
+    },
+    {
+      "epoch": 0.7214290054073759,
+      "grad_norm": 0.20410671830177307,
+      "learning_rate": 1.8065999861817972e-05,
+      "loss": 1.0454,
+      "step": 11874
+    },
+    {
+      "epoch": 0.7214897624400024,
+      "grad_norm": 0.10459274053573608,
+      "learning_rate": 1.8058634462001168e-05,
+      "loss": 1.0299,
+      "step": 11875
+    },
+    {
+      "epoch": 0.7215505194726289,
+      "grad_norm": 0.19205737113952637,
+      "learning_rate": 1.805127023300131e-05,
+      "loss": 1.0933,
+      "step": 11876
+    },
+    {
+      "epoch": 0.7216112765052555,
+      "grad_norm": 0.12911053001880646,
+      "learning_rate": 1.8043907175088298e-05,
+      "loss": 1.0144,
+      "step": 11877
+    },
+    {
+      "epoch": 0.721672033537882,
+      "grad_norm": 0.4698108434677124,
+      "learning_rate": 1.803654528853203e-05,
+      "loss": 1.2417,
+      "step": 11878
+    },
+    {
+      "epoch": 0.7217327905705085,
+      "grad_norm": 0.1186222955584526,
+      "learning_rate": 1.8029184573602394e-05,
+      "loss": 1.0676,
+      "step": 11879
+    },
+    {
+      "epoch": 0.7217935476031351,
+      "grad_norm": 0.11740090698003769,
+      "learning_rate": 1.8021825030569168e-05,
+      "loss": 0.9866,
+      "step": 11880
+    },
+    {
+      "epoch": 0.7218543046357616,
+      "grad_norm": 0.14112940430641174,
+      "learning_rate": 1.8014466659702133e-05,
+      "loss": 1.1117,
+      "step": 11881
+    },
+    {
+      "epoch": 0.7219150616683881,
+      "grad_norm": 0.2183370590209961,
+      "learning_rate": 1.8007109461271006e-05,
+      "loss": 1.0429,
+      "step": 11882
+    },
+    {
+      "epoch": 0.7219758187010147,
+      "grad_norm": 0.1846650242805481,
+      "learning_rate": 1.7999753435545464e-05,
+      "loss": 1.2141,
+      "step": 11883
+    },
+    {
+      "epoch": 0.7220365757336412,
+      "grad_norm": 0.11909636110067368,
+      "learning_rate": 1.7992398582795146e-05,
+      "loss": 1.0356,
+      "step": 11884
+    },
+    {
+      "epoch": 0.7220973327662676,
+      "grad_norm": 0.10280126333236694,
+      "learning_rate": 1.7985044903289645e-05,
+      "loss": 1.0218,
+      "step": 11885
+    },
+    {
+      "epoch": 0.7221580897988942,
+      "grad_norm": 0.3202227056026459,
+      "learning_rate": 1.7977692397298507e-05,
+      "loss": 1.0067,
+      "step": 11886
+    },
+    {
+      "epoch": 0.7222188468315207,
+      "grad_norm": 0.11542931944131851,
+      "learning_rate": 1.7970341065091245e-05,
+      "loss": 1.0532,
+      "step": 11887
+    },
+    {
+      "epoch": 0.7222796038641472,
+      "grad_norm": 0.2886582612991333,
+      "learning_rate": 1.7962990906937326e-05,
+      "loss": 1.1831,
+      "step": 11888
+    },
+    {
+      "epoch": 0.7223403608967738,
+      "grad_norm": 0.1688026338815689,
+      "learning_rate": 1.795564192310617e-05,
+      "loss": 1.0442,
+      "step": 11889
+    },
+    {
+      "epoch": 0.7224011179294003,
+      "grad_norm": 0.18989261984825134,
+      "learning_rate": 1.7948294113867146e-05,
+      "loss": 1.0394,
+      "step": 11890
+    },
+    {
+      "epoch": 0.7224618749620269,
+      "grad_norm": 0.19108445942401886,
+      "learning_rate": 1.7940947479489605e-05,
+      "loss": 1.1426,
+      "step": 11891
+    },
+    {
+      "epoch": 0.7225226319946534,
+      "grad_norm": 0.17166262865066528,
+      "learning_rate": 1.7933602020242823e-05,
+      "loss": 1.0184,
+      "step": 11892
+    },
+    {
+      "epoch": 0.7225833890272799,
+      "grad_norm": 0.19406643509864807,
+      "learning_rate": 1.7926257736396062e-05,
+      "loss": 1.1246,
+      "step": 11893
+    },
+    {
+      "epoch": 0.7226441460599065,
+      "grad_norm": 0.18000058829784393,
+      "learning_rate": 1.791891462821851e-05,
+      "loss": 1.1396,
+      "step": 11894
+    },
+    {
+      "epoch": 0.722704903092533,
+      "grad_norm": 0.6983065009117126,
+      "learning_rate": 1.791157269597935e-05,
+      "loss": 1.0904,
+      "step": 11895
+    },
+    {
+      "epoch": 0.7227656601251595,
+      "grad_norm": 0.11733144521713257,
+      "learning_rate": 1.7904231939947724e-05,
+      "loss": 1.0607,
+      "step": 11896
+    },
+    {
+      "epoch": 0.7228264171577861,
+      "grad_norm": 27.400362014770508,
+      "learning_rate": 1.7896892360392654e-05,
+      "loss": 1.0329,
+      "step": 11897
+    },
+    {
+      "epoch": 0.7228871741904125,
+      "grad_norm": 0.12632763385772705,
+      "learning_rate": 1.7889553957583205e-05,
+      "loss": 1.0182,
+      "step": 11898
+    },
+    {
+      "epoch": 0.722947931223039,
+      "grad_norm": 0.11717429012060165,
+      "learning_rate": 1.788221673178834e-05,
+      "loss": 1.0502,
+      "step": 11899
+    },
+    {
+      "epoch": 0.7230086882556656,
+      "grad_norm": 0.1840904802083969,
+      "learning_rate": 1.7874880683277053e-05,
+      "loss": 1.1037,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7230694452882921,
+      "grad_norm": 0.22967280447483063,
+      "learning_rate": 1.786754581231822e-05,
+      "loss": 1.148,
+      "step": 11901
+    },
+    {
+      "epoch": 0.7231302023209186,
+      "grad_norm": 0.17893098294734955,
+      "learning_rate": 1.786021211918071e-05,
+      "loss": 1.0412,
+      "step": 11902
+    },
+    {
+      "epoch": 0.7231909593535452,
+      "grad_norm": 0.18129698932170868,
+      "learning_rate": 1.785287960413335e-05,
+      "loss": 1.2564,
+      "step": 11903
+    },
+    {
+      "epoch": 0.7232517163861717,
+      "grad_norm": 0.12053747475147247,
+      "learning_rate": 1.784554826744487e-05,
+      "loss": 1.0459,
+      "step": 11904
+    },
+    {
+      "epoch": 0.7233124734187982,
+      "grad_norm": 0.2647194266319275,
+      "learning_rate": 1.7838218109384053e-05,
+      "loss": 1.0365,
+      "step": 11905
+    },
+    {
+      "epoch": 0.7233732304514248,
+      "grad_norm": 0.12505848705768585,
+      "learning_rate": 1.7830889130219564e-05,
+      "loss": 1.0399,
+      "step": 11906
+    },
+    {
+      "epoch": 0.7234339874840513,
+      "grad_norm": 0.10708886384963989,
+      "learning_rate": 1.782356133022005e-05,
+      "loss": 1.043,
+      "step": 11907
+    },
+    {
+      "epoch": 0.7234947445166778,
+      "grad_norm": 0.10663751512765884,
+      "learning_rate": 1.7816234709654124e-05,
+      "loss": 1.0387,
+      "step": 11908
+    },
+    {
+      "epoch": 0.7235555015493044,
+      "grad_norm": 0.10991241782903671,
+      "learning_rate": 1.7808909268790323e-05,
+      "loss": 1.0226,
+      "step": 11909
+    },
+    {
+      "epoch": 0.7236162585819309,
+      "grad_norm": 0.22962409257888794,
+      "learning_rate": 1.7801585007897188e-05,
+      "loss": 1.2356,
+      "step": 11910
+    },
+    {
+      "epoch": 0.7236770156145574,
+      "grad_norm": 0.14256666600704193,
+      "learning_rate": 1.7794261927243173e-05,
+      "loss": 1.0354,
+      "step": 11911
+    },
+    {
+      "epoch": 0.7237377726471839,
+      "grad_norm": 0.11486692726612091,
+      "learning_rate": 1.778694002709671e-05,
+      "loss": 1.0556,
+      "step": 11912
+    },
+    {
+      "epoch": 0.7237985296798104,
+      "grad_norm": 0.184869185090065,
+      "learning_rate": 1.7779619307726186e-05,
+      "loss": 1.0473,
+      "step": 11913
+    },
+    {
+      "epoch": 0.7238592867124369,
+      "grad_norm": 0.1681230366230011,
+      "learning_rate": 1.7772299769399948e-05,
+      "loss": 1.1425,
+      "step": 11914
+    },
+    {
+      "epoch": 0.7239200437450635,
+      "grad_norm": 0.20325717329978943,
+      "learning_rate": 1.7764981412386273e-05,
+      "loss": 1.0678,
+      "step": 11915
+    },
+    {
+      "epoch": 0.72398080077769,
+      "grad_norm": 0.11284222453832626,
+      "learning_rate": 1.7757664236953468e-05,
+      "loss": 1.083,
+      "step": 11916
+    },
+    {
+      "epoch": 0.7240415578103165,
+      "grad_norm": 0.14161108434200287,
+      "learning_rate": 1.7750348243369703e-05,
+      "loss": 1.0956,
+      "step": 11917
+    },
+    {
+      "epoch": 0.7241023148429431,
+      "grad_norm": 0.18631429970264435,
+      "learning_rate": 1.774303343190315e-05,
+      "loss": 1.1746,
+      "step": 11918
+    },
+    {
+      "epoch": 0.7241630718755696,
+      "grad_norm": 0.11615665256977081,
+      "learning_rate": 1.7735719802821943e-05,
+      "loss": 1.0379,
+      "step": 11919
+    },
+    {
+      "epoch": 0.7242238289081961,
+      "grad_norm": 0.11753161996603012,
+      "learning_rate": 1.7728407356394144e-05,
+      "loss": 1.0209,
+      "step": 11920
+    },
+    {
+      "epoch": 0.7242845859408227,
+      "grad_norm": 0.3985714316368103,
+      "learning_rate": 1.7721096092887833e-05,
+      "loss": 1.1739,
+      "step": 11921
+    },
+    {
+      "epoch": 0.7243453429734492,
+      "grad_norm": 0.30062320828437805,
+      "learning_rate": 1.7713786012570983e-05,
+      "loss": 1.0979,
+      "step": 11922
+    },
+    {
+      "epoch": 0.7244061000060757,
+      "grad_norm": 0.1258896291255951,
+      "learning_rate": 1.7706477115711563e-05,
+      "loss": 1.0349,
+      "step": 11923
+    },
+    {
+      "epoch": 0.7244668570387023,
+      "grad_norm": 0.14100538194179535,
+      "learning_rate": 1.7699169402577447e-05,
+      "loss": 1.0864,
+      "step": 11924
+    },
+    {
+      "epoch": 0.7245276140713287,
+      "grad_norm": 0.17764458060264587,
+      "learning_rate": 1.7691862873436504e-05,
+      "loss": 1.0911,
+      "step": 11925
+    },
+    {
+      "epoch": 0.7245883711039552,
+      "grad_norm": 0.138014018535614,
+      "learning_rate": 1.7684557528556595e-05,
+      "loss": 1.0563,
+      "step": 11926
+    },
+    {
+      "epoch": 0.7246491281365818,
+      "grad_norm": 0.12526796758174896,
+      "learning_rate": 1.7677253368205472e-05,
+      "loss": 1.0369,
+      "step": 11927
+    },
+    {
+      "epoch": 0.7247098851692083,
+      "grad_norm": 0.2356281876564026,
+      "learning_rate": 1.7669950392650876e-05,
+      "loss": 1.2178,
+      "step": 11928
+    },
+    {
+      "epoch": 0.7247706422018348,
+      "grad_norm": 0.30109789967536926,
+      "learning_rate": 1.7662648602160493e-05,
+      "loss": 1.1486,
+      "step": 11929
+    },
+    {
+      "epoch": 0.7248313992344614,
+      "grad_norm": 0.24375095963478088,
+      "learning_rate": 1.765534799700198e-05,
+      "loss": 1.2662,
+      "step": 11930
+    },
+    {
+      "epoch": 0.7248921562670879,
+      "grad_norm": 0.151702880859375,
+      "learning_rate": 1.764804857744294e-05,
+      "loss": 1.1219,
+      "step": 11931
+    },
+    {
+      "epoch": 0.7249529132997145,
+      "grad_norm": 0.15219072997570038,
+      "learning_rate": 1.764075034375093e-05,
+      "loss": 1.0062,
+      "step": 11932
+    },
+    {
+      "epoch": 0.725013670332341,
+      "grad_norm": 0.19406989216804504,
+      "learning_rate": 1.763345329619347e-05,
+      "loss": 1.0271,
+      "step": 11933
+    },
+    {
+      "epoch": 0.7250744273649675,
+      "grad_norm": 0.9423526525497437,
+      "learning_rate": 1.762615743503804e-05,
+      "loss": 1.1378,
+      "step": 11934
+    },
+    {
+      "epoch": 0.7251351843975941,
+      "grad_norm": 0.1921922266483307,
+      "learning_rate": 1.7618862760552063e-05,
+      "loss": 1.1375,
+      "step": 11935
+    },
+    {
+      "epoch": 0.7251959414302206,
+      "grad_norm": 0.24168115854263306,
+      "learning_rate": 1.761156927300293e-05,
+      "loss": 1.1208,
+      "step": 11936
+    },
+    {
+      "epoch": 0.7252566984628471,
+      "grad_norm": 0.13272972404956818,
+      "learning_rate": 1.760427697265799e-05,
+      "loss": 1.0707,
+      "step": 11937
+    },
+    {
+      "epoch": 0.7253174554954736,
+      "grad_norm": 0.11252856999635696,
+      "learning_rate": 1.7596985859784537e-05,
+      "loss": 1.0489,
+      "step": 11938
+    },
+    {
+      "epoch": 0.7253782125281001,
+      "grad_norm": 0.10038076341152191,
+      "learning_rate": 1.7589695934649837e-05,
+      "loss": 1.0496,
+      "step": 11939
+    },
+    {
+      "epoch": 0.7254389695607266,
+      "grad_norm": 0.13965252041816711,
+      "learning_rate": 1.758240719752109e-05,
+      "loss": 1.0691,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7254997265933532,
+      "grad_norm": 0.28648120164871216,
+      "learning_rate": 1.7575119648665457e-05,
+      "loss": 1.0717,
+      "step": 11941
+    },
+    {
+      "epoch": 0.7255604836259797,
+      "grad_norm": 0.12222371995449066,
+      "learning_rate": 1.75678332883501e-05,
+      "loss": 1.0412,
+      "step": 11942
+    },
+    {
+      "epoch": 0.7256212406586062,
+      "grad_norm": 0.12257443368434906,
+      "learning_rate": 1.7560548116842097e-05,
+      "loss": 1.0781,
+      "step": 11943
+    },
+    {
+      "epoch": 0.7256819976912328,
+      "grad_norm": 0.3077135980129242,
+      "learning_rate": 1.755326413440846e-05,
+      "loss": 1.1147,
+      "step": 11944
+    },
+    {
+      "epoch": 0.7257427547238593,
+      "grad_norm": 0.19839894771575928,
+      "learning_rate": 1.7545981341316174e-05,
+      "loss": 1.1566,
+      "step": 11945
+    },
+    {
+      "epoch": 0.7258035117564858,
+      "grad_norm": 0.137166365981102,
+      "learning_rate": 1.7538699737832236e-05,
+      "loss": 1.0498,
+      "step": 11946
+    },
+    {
+      "epoch": 0.7258642687891124,
+      "grad_norm": 0.12378855049610138,
+      "learning_rate": 1.7531419324223536e-05,
+      "loss": 1.0502,
+      "step": 11947
+    },
+    {
+      "epoch": 0.7259250258217389,
+      "grad_norm": 0.13504281640052795,
+      "learning_rate": 1.7524140100756937e-05,
+      "loss": 1.0169,
+      "step": 11948
+    },
+    {
+      "epoch": 0.7259857828543654,
+      "grad_norm": 0.5035527348518372,
+      "learning_rate": 1.751686206769926e-05,
+      "loss": 1.1884,
+      "step": 11949
+    },
+    {
+      "epoch": 0.726046539886992,
+      "grad_norm": 0.20623120665550232,
+      "learning_rate": 1.750958522531728e-05,
+      "loss": 1.1622,
+      "step": 11950
+    },
+    {
+      "epoch": 0.7261072969196184,
+      "grad_norm": 0.177435040473938,
+      "learning_rate": 1.7502309573877735e-05,
+      "loss": 1.0072,
+      "step": 11951
+    },
+    {
+      "epoch": 0.7261680539522449,
+      "grad_norm": 0.11040189117193222,
+      "learning_rate": 1.7495035113647312e-05,
+      "loss": 0.9951,
+      "step": 11952
+    },
+    {
+      "epoch": 0.7262288109848715,
+      "grad_norm": 0.15410874783992767,
+      "learning_rate": 1.7487761844892663e-05,
+      "loss": 1.0728,
+      "step": 11953
+    },
+    {
+      "epoch": 0.726289568017498,
+      "grad_norm": 0.17565593123435974,
+      "learning_rate": 1.7480489767880388e-05,
+      "loss": 1.1792,
+      "step": 11954
+    },
+    {
+      "epoch": 0.7263503250501245,
+      "grad_norm": 0.23485685884952545,
+      "learning_rate": 1.7473218882877047e-05,
+      "loss": 1.225,
+      "step": 11955
+    },
+    {
+      "epoch": 0.7264110820827511,
+      "grad_norm": 0.16741417348384857,
+      "learning_rate": 1.7465949190149155e-05,
+      "loss": 1.0828,
+      "step": 11956
+    },
+    {
+      "epoch": 0.7264718391153776,
+      "grad_norm": 0.4004530608654022,
+      "learning_rate": 1.745868068996318e-05,
+      "loss": 1.028,
+      "step": 11957
+    },
+    {
+      "epoch": 0.7265325961480041,
+      "grad_norm": 0.17415936291217804,
+      "learning_rate": 1.7451413382585557e-05,
+      "loss": 1.0247,
+      "step": 11958
+    },
+    {
+      "epoch": 0.7265933531806307,
+      "grad_norm": 0.1294311285018921,
+      "learning_rate": 1.7444147268282673e-05,
+      "loss": 1.083,
+      "step": 11959
+    },
+    {
+      "epoch": 0.7266541102132572,
+      "grad_norm": 0.12395626306533813,
+      "learning_rate": 1.7436882347320856e-05,
+      "loss": 1.0623,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7267148672458837,
+      "grad_norm": 0.1823682337999344,
+      "learning_rate": 1.74296186199664e-05,
+      "loss": 1.058,
+      "step": 11961
+    },
+    {
+      "epoch": 0.7267756242785103,
+      "grad_norm": 0.10366540402173996,
+      "learning_rate": 1.742235608648558e-05,
+      "loss": 1.0377,
+      "step": 11962
+    },
+    {
+      "epoch": 0.7268363813111368,
+      "grad_norm": 0.15267492830753326,
+      "learning_rate": 1.7415094747144617e-05,
+      "loss": 1.1065,
+      "step": 11963
+    },
+    {
+      "epoch": 0.7268971383437632,
+      "grad_norm": 0.14518509805202484,
+      "learning_rate": 1.7407834602209633e-05,
+      "loss": 1.0374,
+      "step": 11964
+    },
+    {
+      "epoch": 0.7269578953763898,
+      "grad_norm": 0.14894914627075195,
+      "learning_rate": 1.7400575651946778e-05,
+      "loss": 1.1389,
+      "step": 11965
+    },
+    {
+      "epoch": 0.7270186524090163,
+      "grad_norm": 0.21335726976394653,
+      "learning_rate": 1.7393317896622103e-05,
+      "loss": 1.1297,
+      "step": 11966
+    },
+    {
+      "epoch": 0.7270794094416428,
+      "grad_norm": 0.12031121551990509,
+      "learning_rate": 1.738606133650168e-05,
+      "loss": 1.0386,
+      "step": 11967
+    },
+    {
+      "epoch": 0.7271401664742694,
+      "grad_norm": 0.11088176816701889,
+      "learning_rate": 1.737880597185148e-05,
+      "loss": 1.0218,
+      "step": 11968
+    },
+    {
+      "epoch": 0.7272009235068959,
+      "grad_norm": 0.13445907831192017,
+      "learning_rate": 1.7371551802937456e-05,
+      "loss": 1.1113,
+      "step": 11969
+    },
+    {
+      "epoch": 0.7272616805395224,
+      "grad_norm": 0.1174154281616211,
+      "learning_rate": 1.736429883002553e-05,
+      "loss": 1.0994,
+      "step": 11970
+    },
+    {
+      "epoch": 0.727322437572149,
+      "grad_norm": 0.1192723736166954,
+      "learning_rate": 1.7357047053381492e-05,
+      "loss": 1.0526,
+      "step": 11971
+    },
+    {
+      "epoch": 0.7273831946047755,
+      "grad_norm": 0.35053887963294983,
+      "learning_rate": 1.734979647327123e-05,
+      "loss": 1.0611,
+      "step": 11972
+    },
+    {
+      "epoch": 0.727443951637402,
+      "grad_norm": 0.13890041410923004,
+      "learning_rate": 1.7342547089960487e-05,
+      "loss": 1.049,
+      "step": 11973
+    },
+    {
+      "epoch": 0.7275047086700286,
+      "grad_norm": 0.12907762825489044,
+      "learning_rate": 1.733529890371499e-05,
+      "loss": 1.1144,
+      "step": 11974
+    },
+    {
+      "epoch": 0.7275654657026551,
+      "grad_norm": 0.1540406048297882,
+      "learning_rate": 1.7328051914800432e-05,
+      "loss": 1.0802,
+      "step": 11975
+    },
+    {
+      "epoch": 0.7276262227352817,
+      "grad_norm": 0.15978215634822845,
+      "learning_rate": 1.7320806123482447e-05,
+      "loss": 1.023,
+      "step": 11976
+    },
+    {
+      "epoch": 0.7276869797679081,
+      "grad_norm": 0.2384788542985916,
+      "learning_rate": 1.7313561530026624e-05,
+      "loss": 1.1058,
+      "step": 11977
+    },
+    {
+      "epoch": 0.7277477368005346,
+      "grad_norm": 0.16895924508571625,
+      "learning_rate": 1.7306318134698523e-05,
+      "loss": 1.0798,
+      "step": 11978
+    },
+    {
+      "epoch": 0.7278084938331612,
+      "grad_norm": 4.890584468841553,
+      "learning_rate": 1.7299075937763654e-05,
+      "loss": 1.0808,
+      "step": 11979
+    },
+    {
+      "epoch": 0.7278692508657877,
+      "grad_norm": 0.13193364441394806,
+      "learning_rate": 1.7291834939487473e-05,
+      "loss": 1.1062,
+      "step": 11980
+    },
+    {
+      "epoch": 0.7279300078984142,
+      "grad_norm": 0.1554129421710968,
+      "learning_rate": 1.7284595140135408e-05,
+      "loss": 1.0449,
+      "step": 11981
+    },
+    {
+      "epoch": 0.7279907649310408,
+      "grad_norm": 0.2118116170167923,
+      "learning_rate": 1.7277356539972812e-05,
+      "loss": 1.0861,
+      "step": 11982
+    },
+    {
+      "epoch": 0.7280515219636673,
+      "grad_norm": 0.20252224802970886,
+      "learning_rate": 1.7270119139265072e-05,
+      "loss": 1.0703,
+      "step": 11983
+    },
+    {
+      "epoch": 0.7281122789962938,
+      "grad_norm": 0.11053638160228729,
+      "learning_rate": 1.7262882938277425e-05,
+      "loss": 1.0638,
+      "step": 11984
+    },
+    {
+      "epoch": 0.7281730360289204,
+      "grad_norm": 0.11425353586673737,
+      "learning_rate": 1.7255647937275133e-05,
+      "loss": 1.0067,
+      "step": 11985
+    },
+    {
+      "epoch": 0.7282337930615469,
+      "grad_norm": 0.13651826977729797,
+      "learning_rate": 1.7248414136523396e-05,
+      "loss": 1.0203,
+      "step": 11986
+    },
+    {
+      "epoch": 0.7282945500941734,
+      "grad_norm": 0.2547251582145691,
+      "learning_rate": 1.724118153628735e-05,
+      "loss": 1.0748,
+      "step": 11987
+    },
+    {
+      "epoch": 0.7283553071268,
+      "grad_norm": 0.15423952043056488,
+      "learning_rate": 1.7233950136832143e-05,
+      "loss": 1.0642,
+      "step": 11988
+    },
+    {
+      "epoch": 0.7284160641594265,
+      "grad_norm": 0.10580328851938248,
+      "learning_rate": 1.722671993842283e-05,
+      "loss": 1.0218,
+      "step": 11989
+    },
+    {
+      "epoch": 0.7284768211920529,
+      "grad_norm": 0.1980973482131958,
+      "learning_rate": 1.7219490941324444e-05,
+      "loss": 1.0959,
+      "step": 11990
+    },
+    {
+      "epoch": 0.7285375782246795,
+      "grad_norm": 0.1393929123878479,
+      "learning_rate": 1.7212263145801943e-05,
+      "loss": 1.0355,
+      "step": 11991
+    },
+    {
+      "epoch": 0.728598335257306,
+      "grad_norm": 0.5448645949363708,
+      "learning_rate": 1.7205036552120257e-05,
+      "loss": 1.059,
+      "step": 11992
+    },
+    {
+      "epoch": 0.7286590922899325,
+      "grad_norm": 0.21412117779254913,
+      "learning_rate": 1.7197811160544307e-05,
+      "loss": 1.1711,
+      "step": 11993
+    },
+    {
+      "epoch": 0.7287198493225591,
+      "grad_norm": 0.6389798521995544,
+      "learning_rate": 1.7190586971338935e-05,
+      "loss": 1.0625,
+      "step": 11994
+    },
+    {
+      "epoch": 0.7287806063551856,
+      "grad_norm": 0.16428568959236145,
+      "learning_rate": 1.7183363984768935e-05,
+      "loss": 1.1916,
+      "step": 11995
+    },
+    {
+      "epoch": 0.7288413633878121,
+      "grad_norm": 0.13203231990337372,
+      "learning_rate": 1.7176142201099077e-05,
+      "loss": 1.0246,
+      "step": 11996
+    },
+    {
+      "epoch": 0.7289021204204387,
+      "grad_norm": 4.570551872253418,
+      "learning_rate": 1.716892162059407e-05,
+      "loss": 1.0759,
+      "step": 11997
+    },
+    {
+      "epoch": 0.7289628774530652,
+      "grad_norm": 0.7408970594406128,
+      "learning_rate": 1.7161702243518585e-05,
+      "loss": 1.0312,
+      "step": 11998
+    },
+    {
+      "epoch": 0.7290236344856917,
+      "grad_norm": 0.18424023687839508,
+      "learning_rate": 1.715448407013725e-05,
+      "loss": 1.064,
+      "step": 11999
+    },
+    {
+      "epoch": 0.7290843915183183,
+      "grad_norm": 0.12323799729347229,
+      "learning_rate": 1.7147267100714654e-05,
+      "loss": 1.0692,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7291451485509448,
+      "grad_norm": 0.1933857798576355,
+      "learning_rate": 1.714005133551533e-05,
+      "loss": 1.1501,
+      "step": 12001
+    },
+    {
+      "epoch": 0.7292059055835713,
+      "grad_norm": 0.13889583945274353,
+      "learning_rate": 1.713283677480378e-05,
+      "loss": 1.0426,
+      "step": 12002
+    },
+    {
+      "epoch": 0.7292666626161978,
+      "grad_norm": 0.09956985712051392,
+      "learning_rate": 1.7125623418844444e-05,
+      "loss": 1.0175,
+      "step": 12003
+    },
+    {
+      "epoch": 0.7293274196488243,
+      "grad_norm": 0.18919014930725098,
+      "learning_rate": 1.711841126790174e-05,
+      "loss": 1.1045,
+      "step": 12004
+    },
+    {
+      "epoch": 0.7293881766814508,
+      "grad_norm": 0.14285099506378174,
+      "learning_rate": 1.7111200322240027e-05,
+      "loss": 1.1028,
+      "step": 12005
+    },
+    {
+      "epoch": 0.7294489337140774,
+      "grad_norm": 0.3515326976776123,
+      "learning_rate": 1.7103990582123624e-05,
+      "loss": 1.0577,
+      "step": 12006
+    },
+    {
+      "epoch": 0.7295096907467039,
+      "grad_norm": 0.14477676153182983,
+      "learning_rate": 1.7096782047816805e-05,
+      "loss": 1.1403,
+      "step": 12007
+    },
+    {
+      "epoch": 0.7295704477793304,
+      "grad_norm": 0.12925733625888824,
+      "learning_rate": 1.7089574719583785e-05,
+      "loss": 1.0734,
+      "step": 12008
+    },
+    {
+      "epoch": 0.729631204811957,
+      "grad_norm": 0.13059315085411072,
+      "learning_rate": 1.708236859768878e-05,
+      "loss": 1.0159,
+      "step": 12009
+    },
+    {
+      "epoch": 0.7296919618445835,
+      "grad_norm": 0.20773211121559143,
+      "learning_rate": 1.7075163682395933e-05,
+      "loss": 1.0933,
+      "step": 12010
+    },
+    {
+      "epoch": 0.72975271887721,
+      "grad_norm": 0.13940361142158508,
+      "learning_rate": 1.706795997396931e-05,
+      "loss": 1.0617,
+      "step": 12011
+    },
+    {
+      "epoch": 0.7298134759098366,
+      "grad_norm": 0.15254928171634674,
+      "learning_rate": 1.7060757472672977e-05,
+      "loss": 1.122,
+      "step": 12012
+    },
+    {
+      "epoch": 0.7298742329424631,
+      "grad_norm": 0.13257399201393127,
+      "learning_rate": 1.705355617877093e-05,
+      "loss": 1.0407,
+      "step": 12013
+    },
+    {
+      "epoch": 0.7299349899750897,
+      "grad_norm": 0.20717093348503113,
+      "learning_rate": 1.704635609252717e-05,
+      "loss": 1.0555,
+      "step": 12014
+    },
+    {
+      "epoch": 0.7299957470077162,
+      "grad_norm": 0.12712256610393524,
+      "learning_rate": 1.7039157214205593e-05,
+      "loss": 1.0881,
+      "step": 12015
+    },
+    {
+      "epoch": 0.7300565040403427,
+      "grad_norm": 0.12965604662895203,
+      "learning_rate": 1.7031959544070082e-05,
+      "loss": 1.0441,
+      "step": 12016
+    },
+    {
+      "epoch": 0.7301172610729691,
+      "grad_norm": 0.22230522334575653,
+      "learning_rate": 1.7024763082384483e-05,
+      "loss": 1.1574,
+      "step": 12017
+    },
+    {
+      "epoch": 0.7301780181055957,
+      "grad_norm": 0.5337817072868347,
+      "learning_rate": 1.7017567829412535e-05,
+      "loss": 1.0982,
+      "step": 12018
+    },
+    {
+      "epoch": 0.7302387751382222,
+      "grad_norm": 0.09890550374984741,
+      "learning_rate": 1.7010373785418028e-05,
+      "loss": 1.0779,
+      "step": 12019
+    },
+    {
+      "epoch": 0.7302995321708488,
+      "grad_norm": 0.20456703007221222,
+      "learning_rate": 1.7003180950664642e-05,
+      "loss": 1.0771,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7303602892034753,
+      "grad_norm": 0.1896372139453888,
+      "learning_rate": 1.6995989325416046e-05,
+      "loss": 0.9824,
+      "step": 12021
+    },
+    {
+      "epoch": 0.7304210462361018,
+      "grad_norm": 0.09826211631298065,
+      "learning_rate": 1.6988798909935833e-05,
+      "loss": 1.0475,
+      "step": 12022
+    },
+    {
+      "epoch": 0.7304818032687284,
+      "grad_norm": 0.12386680394411087,
+      "learning_rate": 1.6981609704487583e-05,
+      "loss": 1.051,
+      "step": 12023
+    },
+    {
+      "epoch": 0.7305425603013549,
+      "grad_norm": 0.1515650451183319,
+      "learning_rate": 1.6974421709334808e-05,
+      "loss": 1.0871,
+      "step": 12024
+    },
+    {
+      "epoch": 0.7306033173339814,
+      "grad_norm": 0.22790172696113586,
+      "learning_rate": 1.696723492474099e-05,
+      "loss": 1.0778,
+      "step": 12025
+    },
+    {
+      "epoch": 0.730664074366608,
+      "grad_norm": 0.5616945624351501,
+      "learning_rate": 1.696004935096956e-05,
+      "loss": 1.207,
+      "step": 12026
+    },
+    {
+      "epoch": 0.7307248313992345,
+      "grad_norm": 0.1577715426683426,
+      "learning_rate": 1.6952864988283907e-05,
+      "loss": 1.1128,
+      "step": 12027
+    },
+    {
+      "epoch": 0.730785588431861,
+      "grad_norm": 0.12401007115840912,
+      "learning_rate": 1.6945681836947374e-05,
+      "loss": 1.0281,
+      "step": 12028
+    },
+    {
+      "epoch": 0.7308463454644876,
+      "grad_norm": 0.1118738129734993,
+      "learning_rate": 1.6938499897223252e-05,
+      "loss": 1.0026,
+      "step": 12029
+    },
+    {
+      "epoch": 0.730907102497114,
+      "grad_norm": 0.16849000751972198,
+      "learning_rate": 1.6931319169374836e-05,
+      "loss": 1.0547,
+      "step": 12030
+    },
+    {
+      "epoch": 0.7309678595297405,
+      "grad_norm": 0.18020136654376984,
+      "learning_rate": 1.6924139653665293e-05,
+      "loss": 1.0322,
+      "step": 12031
+    },
+    {
+      "epoch": 0.7310286165623671,
+      "grad_norm": 0.1565113067626953,
+      "learning_rate": 1.6916961350357802e-05,
+      "loss": 0.9947,
+      "step": 12032
+    },
+    {
+      "epoch": 0.7310893735949936,
+      "grad_norm": 0.1534905582666397,
+      "learning_rate": 1.690978425971549e-05,
+      "loss": 1.0182,
+      "step": 12033
+    },
+    {
+      "epoch": 0.7311501306276201,
+      "grad_norm": 0.123008593916893,
+      "learning_rate": 1.6902608382001415e-05,
+      "loss": 1.0365,
+      "step": 12034
+    },
+    {
+      "epoch": 0.7312108876602467,
+      "grad_norm": 0.2115435153245926,
+      "learning_rate": 1.6895433717478643e-05,
+      "loss": 1.1278,
+      "step": 12035
+    },
+    {
+      "epoch": 0.7312716446928732,
+      "grad_norm": 0.4212009608745575,
+      "learning_rate": 1.688826026641015e-05,
+      "loss": 1.0335,
+      "step": 12036
+    },
+    {
+      "epoch": 0.7313324017254997,
+      "grad_norm": 5.170857906341553,
+      "learning_rate": 1.688108802905889e-05,
+      "loss": 1.0288,
+      "step": 12037
+    },
+    {
+      "epoch": 0.7313931587581263,
+      "grad_norm": 0.22664779424667358,
+      "learning_rate": 1.687391700568772e-05,
+      "loss": 1.1439,
+      "step": 12038
+    },
+    {
+      "epoch": 0.7314539157907528,
+      "grad_norm": 0.11372818052768707,
+      "learning_rate": 1.686674719655954e-05,
+      "loss": 1.0972,
+      "step": 12039
+    },
+    {
+      "epoch": 0.7315146728233793,
+      "grad_norm": 0.19588282704353333,
+      "learning_rate": 1.6859578601937144e-05,
+      "loss": 1.1204,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7315754298560059,
+      "grad_norm": 0.17658144235610962,
+      "learning_rate": 1.6852411222083302e-05,
+      "loss": 1.0986,
+      "step": 12041
+    },
+    {
+      "epoch": 0.7316361868886324,
+      "grad_norm": 0.12479786574840546,
+      "learning_rate": 1.6845245057260734e-05,
+      "loss": 1.0102,
+      "step": 12042
+    },
+    {
+      "epoch": 0.7316969439212588,
+      "grad_norm": 7.837426662445068,
+      "learning_rate": 1.6838080107732117e-05,
+      "loss": 1.0823,
+      "step": 12043
+    },
+    {
+      "epoch": 0.7317577009538854,
+      "grad_norm": 0.6720539331436157,
+      "learning_rate": 1.6830916373760087e-05,
+      "loss": 1.1298,
+      "step": 12044
+    },
+    {
+      "epoch": 0.7318184579865119,
+      "grad_norm": 0.3161039650440216,
+      "learning_rate": 1.6823753855607222e-05,
+      "loss": 1.2,
+      "step": 12045
+    },
+    {
+      "epoch": 0.7318792150191384,
+      "grad_norm": 0.14072582125663757,
+      "learning_rate": 1.6816592553536077e-05,
+      "loss": 1.0535,
+      "step": 12046
+    },
+    {
+      "epoch": 0.731939972051765,
+      "grad_norm": 0.22343140840530396,
+      "learning_rate": 1.6809432467809144e-05,
+      "loss": 1.0222,
+      "step": 12047
+    },
+    {
+      "epoch": 0.7320007290843915,
+      "grad_norm": 0.12016142904758453,
+      "learning_rate": 1.6802273598688884e-05,
+      "loss": 1.0236,
+      "step": 12048
+    },
+    {
+      "epoch": 0.732061486117018,
+      "grad_norm": 0.1599024087190628,
+      "learning_rate": 1.67951159464377e-05,
+      "loss": 1.11,
+      "step": 12049
+    },
+    {
+      "epoch": 0.7321222431496446,
+      "grad_norm": 0.1445721685886383,
+      "learning_rate": 1.6787959511317958e-05,
+      "loss": 1.0586,
+      "step": 12050
+    },
+    {
+      "epoch": 0.7321830001822711,
+      "grad_norm": 0.10825826227664948,
+      "learning_rate": 1.678080429359198e-05,
+      "loss": 1.0494,
+      "step": 12051
+    },
+    {
+      "epoch": 0.7322437572148977,
+      "grad_norm": 0.12876258790493011,
+      "learning_rate": 1.677365029352204e-05,
+      "loss": 1.0637,
+      "step": 12052
+    },
+    {
+      "epoch": 0.7323045142475242,
+      "grad_norm": 1.277681827545166,
+      "learning_rate": 1.6766497511370376e-05,
+      "loss": 1.1165,
+      "step": 12053
+    },
+    {
+      "epoch": 0.7323652712801507,
+      "grad_norm": 0.16351719200611115,
+      "learning_rate": 1.6759345947399145e-05,
+      "loss": 1.0655,
+      "step": 12054
+    },
+    {
+      "epoch": 0.7324260283127773,
+      "grad_norm": 0.1531379669904709,
+      "learning_rate": 1.675219560187054e-05,
+      "loss": 1.0843,
+      "step": 12055
+    },
+    {
+      "epoch": 0.7324867853454037,
+      "grad_norm": 0.14996446669101715,
+      "learning_rate": 1.6745046475046627e-05,
+      "loss": 1.0445,
+      "step": 12056
+    },
+    {
+      "epoch": 0.7325475423780302,
+      "grad_norm": 0.15850290656089783,
+      "learning_rate": 1.673789856718948e-05,
+      "loss": 1.0966,
+      "step": 12057
+    },
+    {
+      "epoch": 0.7326082994106567,
+      "grad_norm": 1.516983151435852,
+      "learning_rate": 1.6730751878561074e-05,
+      "loss": 1.0543,
+      "step": 12058
+    },
+    {
+      "epoch": 0.7326690564432833,
+      "grad_norm": 0.17107877135276794,
+      "learning_rate": 1.672360640942337e-05,
+      "loss": 1.1013,
+      "step": 12059
+    },
+    {
+      "epoch": 0.7327298134759098,
+      "grad_norm": 0.48979371786117554,
+      "learning_rate": 1.6716462160038325e-05,
+      "loss": 1.1381,
+      "step": 12060
+    },
+    {
+      "epoch": 0.7327905705085364,
+      "grad_norm": 0.13007274270057678,
+      "learning_rate": 1.6709319130667793e-05,
+      "loss": 1.0532,
+      "step": 12061
+    },
+    {
+      "epoch": 0.7328513275411629,
+      "grad_norm": 0.11528556793928146,
+      "learning_rate": 1.6702177321573603e-05,
+      "loss": 1.031,
+      "step": 12062
+    },
+    {
+      "epoch": 0.7329120845737894,
+      "grad_norm": 0.10124536603689194,
+      "learning_rate": 1.669503673301754e-05,
+      "loss": 1.0465,
+      "step": 12063
+    },
+    {
+      "epoch": 0.732972841606416,
+      "grad_norm": 0.18992623686790466,
+      "learning_rate": 1.668789736526134e-05,
+      "loss": 1.1641,
+      "step": 12064
+    },
+    {
+      "epoch": 0.7330335986390425,
+      "grad_norm": 0.1198720708489418,
+      "learning_rate": 1.6680759218566704e-05,
+      "loss": 1.0275,
+      "step": 12065
+    },
+    {
+      "epoch": 0.733094355671669,
+      "grad_norm": 0.1708758920431137,
+      "learning_rate": 1.667362229319528e-05,
+      "loss": 1.0434,
+      "step": 12066
+    },
+    {
+      "epoch": 0.7331551127042956,
+      "grad_norm": 0.1209886223077774,
+      "learning_rate": 1.666648658940867e-05,
+      "loss": 1.0277,
+      "step": 12067
+    },
+    {
+      "epoch": 0.7332158697369221,
+      "grad_norm": 0.15256762504577637,
+      "learning_rate": 1.665935210746844e-05,
+      "loss": 1.1144,
+      "step": 12068
+    },
+    {
+      "epoch": 0.7332766267695485,
+      "grad_norm": 0.11685581505298615,
+      "learning_rate": 1.6652218847636098e-05,
+      "loss": 0.9854,
+      "step": 12069
+    },
+    {
+      "epoch": 0.7333373838021751,
+      "grad_norm": 0.12188849598169327,
+      "learning_rate": 1.664508681017313e-05,
+      "loss": 1.0406,
+      "step": 12070
+    },
+    {
+      "epoch": 0.7333981408348016,
+      "grad_norm": 0.14406493306159973,
+      "learning_rate": 1.663795599534094e-05,
+      "loss": 1.0887,
+      "step": 12071
+    },
+    {
+      "epoch": 0.7334588978674281,
+      "grad_norm": 0.13072574138641357,
+      "learning_rate": 1.663082640340093e-05,
+      "loss": 1.0692,
+      "step": 12072
+    },
+    {
+      "epoch": 0.7335196549000547,
+      "grad_norm": 0.2514955699443817,
+      "learning_rate": 1.6623698034614428e-05,
+      "loss": 1.1562,
+      "step": 12073
+    },
+    {
+      "epoch": 0.7335804119326812,
+      "grad_norm": 0.21190212666988373,
+      "learning_rate": 1.661657088924272e-05,
+      "loss": 1.0614,
+      "step": 12074
+    },
+    {
+      "epoch": 0.7336411689653077,
+      "grad_norm": 0.1246088445186615,
+      "learning_rate": 1.660944496754705e-05,
+      "loss": 1.0525,
+      "step": 12075
+    },
+    {
+      "epoch": 0.7337019259979343,
+      "grad_norm": 0.633964478969574,
+      "learning_rate": 1.660232026978864e-05,
+      "loss": 1.041,
+      "step": 12076
+    },
+    {
+      "epoch": 0.7337626830305608,
+      "grad_norm": 0.18055935204029083,
+      "learning_rate": 1.6595196796228655e-05,
+      "loss": 1.0343,
+      "step": 12077
+    },
+    {
+      "epoch": 0.7338234400631873,
+      "grad_norm": 0.10113296657800674,
+      "learning_rate": 1.6588074547128173e-05,
+      "loss": 0.9919,
+      "step": 12078
+    },
+    {
+      "epoch": 0.7338841970958139,
+      "grad_norm": 0.14399178326129913,
+      "learning_rate": 1.658095352274827e-05,
+      "loss": 1.0757,
+      "step": 12079
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 0.18908679485321045,
+      "learning_rate": 1.657383372334997e-05,
+      "loss": 1.1575,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7340057111610669,
+      "grad_norm": 0.14220589399337769,
+      "learning_rate": 1.656671514919427e-05,
+      "loss": 1.0921,
+      "step": 12081
+    },
+    {
+      "epoch": 0.7340664681936934,
+      "grad_norm": 0.380215585231781,
+      "learning_rate": 1.6559597800542082e-05,
+      "loss": 1.1456,
+      "step": 12082
+    },
+    {
+      "epoch": 0.7341272252263199,
+      "grad_norm": 0.5016540288925171,
+      "learning_rate": 1.6552481677654302e-05,
+      "loss": 1.1984,
+      "step": 12083
+    },
+    {
+      "epoch": 0.7341879822589464,
+      "grad_norm": 0.26983851194381714,
+      "learning_rate": 1.6545366780791792e-05,
+      "loss": 1.1842,
+      "step": 12084
+    },
+    {
+      "epoch": 0.734248739291573,
+      "grad_norm": 0.24236908555030823,
+      "learning_rate": 1.6538253110215292e-05,
+      "loss": 1.1977,
+      "step": 12085
+    },
+    {
+      "epoch": 0.7343094963241995,
+      "grad_norm": 0.14403243362903595,
+      "learning_rate": 1.6531140666185614e-05,
+      "loss": 1.0771,
+      "step": 12086
+    },
+    {
+      "epoch": 0.734370253356826,
+      "grad_norm": 0.18638917803764343,
+      "learning_rate": 1.652402944896344e-05,
+      "loss": 1.0281,
+      "step": 12087
+    },
+    {
+      "epoch": 0.7344310103894526,
+      "grad_norm": 0.1815180778503418,
+      "learning_rate": 1.651691945880944e-05,
+      "loss": 1.087,
+      "step": 12088
+    },
+    {
+      "epoch": 0.7344917674220791,
+      "grad_norm": 0.1412448137998581,
+      "learning_rate": 1.6509810695984225e-05,
+      "loss": 0.9898,
+      "step": 12089
+    },
+    {
+      "epoch": 0.7345525244547056,
+      "grad_norm": 0.13540153205394745,
+      "learning_rate": 1.650270316074838e-05,
+      "loss": 1.0673,
+      "step": 12090
+    },
+    {
+      "epoch": 0.7346132814873322,
+      "grad_norm": 35.782196044921875,
+      "learning_rate": 1.6495596853362415e-05,
+      "loss": 1.1115,
+      "step": 12091
+    },
+    {
+      "epoch": 0.7346740385199587,
+      "grad_norm": 0.1554182916879654,
+      "learning_rate": 1.6488491774086834e-05,
+      "loss": 1.0441,
+      "step": 12092
+    },
+    {
+      "epoch": 0.7347347955525853,
+      "grad_norm": 0.2042216658592224,
+      "learning_rate": 1.6481387923182055e-05,
+      "loss": 1.1018,
+      "step": 12093
+    },
+    {
+      "epoch": 0.7347955525852118,
+      "grad_norm": 0.1271388679742813,
+      "learning_rate": 1.6474285300908486e-05,
+      "loss": 1.0561,
+      "step": 12094
+    },
+    {
+      "epoch": 0.7348563096178382,
+      "grad_norm": 0.1373508721590042,
+      "learning_rate": 1.6467183907526472e-05,
+      "loss": 1.0392,
+      "step": 12095
+    },
+    {
+      "epoch": 0.7349170666504647,
+      "grad_norm": 0.1256594955921173,
+      "learning_rate": 1.6460083743296296e-05,
+      "loss": 1.0061,
+      "step": 12096
+    },
+    {
+      "epoch": 0.7349778236830913,
+      "grad_norm": 0.23491358757019043,
+      "learning_rate": 1.6452984808478267e-05,
+      "loss": 1.1557,
+      "step": 12097
+    },
+    {
+      "epoch": 0.7350385807157178,
+      "grad_norm": 0.13289356231689453,
+      "learning_rate": 1.644588710333254e-05,
+      "loss": 1.0213,
+      "step": 12098
+    },
+    {
+      "epoch": 0.7350993377483444,
+      "grad_norm": 0.21962270140647888,
+      "learning_rate": 1.6438790628119317e-05,
+      "loss": 1.1838,
+      "step": 12099
+    },
+    {
+      "epoch": 0.7351600947809709,
+      "grad_norm": 0.16216178238391876,
+      "learning_rate": 1.6431695383098704e-05,
+      "loss": 1.091,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7352208518135974,
+      "grad_norm": 0.3127570152282715,
+      "learning_rate": 1.6424601368530772e-05,
+      "loss": 1.1891,
+      "step": 12101
+    },
+    {
+      "epoch": 0.735281608846224,
+      "grad_norm": 0.22137241065502167,
+      "learning_rate": 1.6417508584675583e-05,
+      "loss": 1.1461,
+      "step": 12102
+    },
+    {
+      "epoch": 0.7353423658788505,
+      "grad_norm": 4.210534572601318,
+      "learning_rate": 1.6410417031793112e-05,
+      "loss": 1.0345,
+      "step": 12103
+    },
+    {
+      "epoch": 0.735403122911477,
+      "grad_norm": 0.23011243343353271,
+      "learning_rate": 1.6403326710143306e-05,
+      "loss": 1.0741,
+      "step": 12104
+    },
+    {
+      "epoch": 0.7354638799441036,
+      "grad_norm": 0.25292232632637024,
+      "learning_rate": 1.6396237619986042e-05,
+      "loss": 1.0699,
+      "step": 12105
+    },
+    {
+      "epoch": 0.7355246369767301,
+      "grad_norm": 0.13957048952579498,
+      "learning_rate": 1.638914976158117e-05,
+      "loss": 1.0243,
+      "step": 12106
+    },
+    {
+      "epoch": 0.7355853940093566,
+      "grad_norm": 0.12817224860191345,
+      "learning_rate": 1.6382063135188517e-05,
+      "loss": 1.0472,
+      "step": 12107
+    },
+    {
+      "epoch": 0.735646151041983,
+      "grad_norm": 0.12771575152873993,
+      "learning_rate": 1.6374977741067847e-05,
+      "loss": 1.0779,
+      "step": 12108
+    },
+    {
+      "epoch": 0.7357069080746096,
+      "grad_norm": 0.1966123878955841,
+      "learning_rate": 1.6367893579478865e-05,
+      "loss": 1.0523,
+      "step": 12109
+    },
+    {
+      "epoch": 0.7357676651072361,
+      "grad_norm": 0.16574272513389587,
+      "learning_rate": 1.6360810650681262e-05,
+      "loss": 1.1028,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7358284221398627,
+      "grad_norm": 0.12013058364391327,
+      "learning_rate": 1.635372895493461e-05,
+      "loss": 1.029,
+      "step": 12111
+    },
+    {
+      "epoch": 0.7358891791724892,
+      "grad_norm": 0.15217263996601105,
+      "learning_rate": 1.634664849249855e-05,
+      "loss": 1.1132,
+      "step": 12112
+    },
+    {
+      "epoch": 0.7359499362051157,
+      "grad_norm": 2.272892475128174,
+      "learning_rate": 1.6339569263632588e-05,
+      "loss": 1.0675,
+      "step": 12113
+    },
+    {
+      "epoch": 0.7360106932377423,
+      "grad_norm": 0.1886737197637558,
+      "learning_rate": 1.6332491268596223e-05,
+      "loss": 1.0251,
+      "step": 12114
+    },
+    {
+      "epoch": 0.7360714502703688,
+      "grad_norm": 0.15079881250858307,
+      "learning_rate": 1.63254145076489e-05,
+      "loss": 1.1271,
+      "step": 12115
+    },
+    {
+      "epoch": 0.7361322073029953,
+      "grad_norm": 0.16714252531528473,
+      "learning_rate": 1.631833898105002e-05,
+      "loss": 1.079,
+      "step": 12116
+    },
+    {
+      "epoch": 0.7361929643356219,
+      "grad_norm": 0.10738638043403625,
+      "learning_rate": 1.631126468905893e-05,
+      "loss": 1.0957,
+      "step": 12117
+    },
+    {
+      "epoch": 0.7362537213682484,
+      "grad_norm": 0.12305158376693726,
+      "learning_rate": 1.6304191631934945e-05,
+      "loss": 1.0236,
+      "step": 12118
+    },
+    {
+      "epoch": 0.7363144784008749,
+      "grad_norm": 0.18824654817581177,
+      "learning_rate": 1.6297119809937328e-05,
+      "loss": 1.1123,
+      "step": 12119
+    },
+    {
+      "epoch": 0.7363752354335015,
+      "grad_norm": 0.11218249052762985,
+      "learning_rate": 1.6290049223325304e-05,
+      "loss": 1.0697,
+      "step": 12120
+    },
+    {
+      "epoch": 0.736435992466128,
+      "grad_norm": 0.8790396451950073,
+      "learning_rate": 1.628297987235804e-05,
+      "loss": 1.0434,
+      "step": 12121
+    },
+    {
+      "epoch": 0.7364967494987544,
+      "grad_norm": 0.15753483772277832,
+      "learning_rate": 1.627591175729465e-05,
+      "loss": 1.141,
+      "step": 12122
+    },
+    {
+      "epoch": 0.736557506531381,
+      "grad_norm": 6.803406238555908,
+      "learning_rate": 1.6268844878394247e-05,
+      "loss": 0.9891,
+      "step": 12123
+    },
+    {
+      "epoch": 0.7366182635640075,
+      "grad_norm": 0.854327917098999,
+      "learning_rate": 1.626177923591587e-05,
+      "loss": 1.1061,
+      "step": 12124
+    },
+    {
+      "epoch": 0.736679020596634,
+      "grad_norm": 0.13866448402404785,
+      "learning_rate": 1.6254714830118482e-05,
+      "loss": 1.0706,
+      "step": 12125
+    },
+    {
+      "epoch": 0.7367397776292606,
+      "grad_norm": 0.2640054523944855,
+      "learning_rate": 1.6247651661261047e-05,
+      "loss": 1.1143,
+      "step": 12126
+    },
+    {
+      "epoch": 0.7368005346618871,
+      "grad_norm": 0.12271498143672943,
+      "learning_rate": 1.6240589729602446e-05,
+      "loss": 1.0639,
+      "step": 12127
+    },
+    {
+      "epoch": 0.7368612916945136,
+      "grad_norm": 0.18867793679237366,
+      "learning_rate": 1.6233529035401575e-05,
+      "loss": 1.1719,
+      "step": 12128
+    },
+    {
+      "epoch": 0.7369220487271402,
+      "grad_norm": 0.2110293060541153,
+      "learning_rate": 1.622646957891722e-05,
+      "loss": 1.0572,
+      "step": 12129
+    },
+    {
+      "epoch": 0.7369828057597667,
+      "grad_norm": 0.118031807243824,
+      "learning_rate": 1.6219411360408148e-05,
+      "loss": 1.0803,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7370435627923932,
+      "grad_norm": 0.12642765045166016,
+      "learning_rate": 1.6212354380133088e-05,
+      "loss": 1.0499,
+      "step": 12131
+    },
+    {
+      "epoch": 0.7371043198250198,
+      "grad_norm": 0.23367492854595184,
+      "learning_rate": 1.6205298638350704e-05,
+      "loss": 1.0253,
+      "step": 12132
+    },
+    {
+      "epoch": 0.7371650768576463,
+      "grad_norm": 0.28645071387290955,
+      "learning_rate": 1.6198244135319636e-05,
+      "loss": 1.1369,
+      "step": 12133
+    },
+    {
+      "epoch": 0.7372258338902729,
+      "grad_norm": 0.14862053096294403,
+      "learning_rate": 1.6191190871298457e-05,
+      "loss": 1.0935,
+      "step": 12134
+    },
+    {
+      "epoch": 0.7372865909228993,
+      "grad_norm": 0.14767231047153473,
+      "learning_rate": 1.6184138846545714e-05,
+      "loss": 1.0481,
+      "step": 12135
+    },
+    {
+      "epoch": 0.7373473479555258,
+      "grad_norm": 0.11814218759536743,
+      "learning_rate": 1.6177088061319902e-05,
+      "loss": 1.0082,
+      "step": 12136
+    },
+    {
+      "epoch": 0.7374081049881523,
+      "grad_norm": 0.5006747245788574,
+      "learning_rate": 1.617003851587946e-05,
+      "loss": 1.0453,
+      "step": 12137
+    },
+    {
+      "epoch": 0.7374688620207789,
+      "grad_norm": 0.14699113368988037,
+      "learning_rate": 1.61629902104828e-05,
+      "loss": 1.0858,
+      "step": 12138
+    },
+    {
+      "epoch": 0.7375296190534054,
+      "grad_norm": 0.114358089864254,
+      "learning_rate": 1.615594314538827e-05,
+      "loss": 1.0256,
+      "step": 12139
+    },
+    {
+      "epoch": 0.737590376086032,
+      "grad_norm": 0.26095932722091675,
+      "learning_rate": 1.614889732085419e-05,
+      "loss": 1.1842,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7376511331186585,
+      "grad_norm": 0.12283433228731155,
+      "learning_rate": 1.614185273713882e-05,
+      "loss": 1.0087,
+      "step": 12141
+    },
+    {
+      "epoch": 0.737711890151285,
+      "grad_norm": 0.11278723925352097,
+      "learning_rate": 1.613480939450038e-05,
+      "loss": 1.076,
+      "step": 12142
+    },
+    {
+      "epoch": 0.7377726471839116,
+      "grad_norm": 0.1318920999765396,
+      "learning_rate": 1.6127767293197034e-05,
+      "loss": 1.0812,
+      "step": 12143
+    },
+    {
+      "epoch": 0.7378334042165381,
+      "grad_norm": 0.10175958275794983,
+      "learning_rate": 1.6120726433486953e-05,
+      "loss": 0.9952,
+      "step": 12144
+    },
+    {
+      "epoch": 0.7378941612491646,
+      "grad_norm": 0.10769624263048172,
+      "learning_rate": 1.6113686815628187e-05,
+      "loss": 1.0417,
+      "step": 12145
+    },
+    {
+      "epoch": 0.7379549182817912,
+      "grad_norm": 0.1390557587146759,
+      "learning_rate": 1.6106648439878776e-05,
+      "loss": 1.109,
+      "step": 12146
+    },
+    {
+      "epoch": 0.7380156753144177,
+      "grad_norm": 0.451018363237381,
+      "learning_rate": 1.6099611306496698e-05,
+      "loss": 1.0775,
+      "step": 12147
+    },
+    {
+      "epoch": 0.7380764323470441,
+      "grad_norm": 0.14769063889980316,
+      "learning_rate": 1.6092575415739943e-05,
+      "loss": 1.0381,
+      "step": 12148
+    },
+    {
+      "epoch": 0.7381371893796707,
+      "grad_norm": 0.15738914906978607,
+      "learning_rate": 1.608554076786638e-05,
+      "loss": 1.0091,
+      "step": 12149
+    },
+    {
+      "epoch": 0.7381979464122972,
+      "grad_norm": 0.1770702749490738,
+      "learning_rate": 1.6078507363133888e-05,
+      "loss": 1.1069,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7382587034449237,
+      "grad_norm": 0.10331787168979645,
+      "learning_rate": 1.6071475201800272e-05,
+      "loss": 1.0161,
+      "step": 12151
+    },
+    {
+      "epoch": 0.7383194604775503,
+      "grad_norm": 0.11065951734781265,
+      "learning_rate": 1.6064444284123265e-05,
+      "loss": 1.0884,
+      "step": 12152
+    },
+    {
+      "epoch": 0.7383802175101768,
+      "grad_norm": 0.13166925311088562,
+      "learning_rate": 1.6057414610360627e-05,
+      "loss": 1.032,
+      "step": 12153
+    },
+    {
+      "epoch": 0.7384409745428033,
+      "grad_norm": 0.13021034002304077,
+      "learning_rate": 1.6050386180770017e-05,
+      "loss": 1.0543,
+      "step": 12154
+    },
+    {
+      "epoch": 0.7385017315754299,
+      "grad_norm": 0.16593828797340393,
+      "learning_rate": 1.604335899560906e-05,
+      "loss": 1.0737,
+      "step": 12155
+    },
+    {
+      "epoch": 0.7385624886080564,
+      "grad_norm": 0.13188204169273376,
+      "learning_rate": 1.6036333055135344e-05,
+      "loss": 1.0397,
+      "step": 12156
+    },
+    {
+      "epoch": 0.7386232456406829,
+      "grad_norm": 0.09796518087387085,
+      "learning_rate": 1.602930835960641e-05,
+      "loss": 0.9523,
+      "step": 12157
+    },
+    {
+      "epoch": 0.7386840026733095,
+      "grad_norm": 0.1713603436946869,
+      "learning_rate": 1.602228490927974e-05,
+      "loss": 1.0674,
+      "step": 12158
+    },
+    {
+      "epoch": 0.738744759705936,
+      "grad_norm": 0.17747315764427185,
+      "learning_rate": 1.6015262704412787e-05,
+      "loss": 1.1334,
+      "step": 12159
+    },
+    {
+      "epoch": 0.7388055167385625,
+      "grad_norm": 0.15091343224048615,
+      "learning_rate": 1.6008241745262947e-05,
+      "loss": 1.1572,
+      "step": 12160
+    },
+    {
+      "epoch": 0.738866273771189,
+      "grad_norm": 0.14343756437301636,
+      "learning_rate": 1.6001222032087575e-05,
+      "loss": 1.074,
+      "step": 12161
+    },
+    {
+      "epoch": 0.7389270308038155,
+      "grad_norm": 0.1796892285346985,
+      "learning_rate": 1.5994203565143982e-05,
+      "loss": 1.0948,
+      "step": 12162
+    },
+    {
+      "epoch": 0.738987787836442,
+      "grad_norm": 0.12104852497577667,
+      "learning_rate": 1.598718634468942e-05,
+      "loss": 1.061,
+      "step": 12163
+    },
+    {
+      "epoch": 0.7390485448690686,
+      "grad_norm": 0.4887903332710266,
+      "learning_rate": 1.598017037098114e-05,
+      "loss": 1.0691,
+      "step": 12164
+    },
+    {
+      "epoch": 0.7391093019016951,
+      "grad_norm": 0.15365274250507355,
+      "learning_rate": 1.5973155644276284e-05,
+      "loss": 1.0465,
+      "step": 12165
+    },
+    {
+      "epoch": 0.7391700589343216,
+      "grad_norm": 0.17284902930259705,
+      "learning_rate": 1.5966142164831976e-05,
+      "loss": 1.0845,
+      "step": 12166
+    },
+    {
+      "epoch": 0.7392308159669482,
+      "grad_norm": 0.1603914052248001,
+      "learning_rate": 1.5959129932905314e-05,
+      "loss": 1.116,
+      "step": 12167
+    },
+    {
+      "epoch": 0.7392915729995747,
+      "grad_norm": 0.13822023570537567,
+      "learning_rate": 1.5952118948753302e-05,
+      "loss": 1.0972,
+      "step": 12168
+    },
+    {
+      "epoch": 0.7393523300322012,
+      "grad_norm": 0.13838990032672882,
+      "learning_rate": 1.594510921263297e-05,
+      "loss": 1.0701,
+      "step": 12169
+    },
+    {
+      "epoch": 0.7394130870648278,
+      "grad_norm": 0.1430082470178604,
+      "learning_rate": 1.593810072480124e-05,
+      "loss": 1.0517,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7394738440974543,
+      "grad_norm": 0.11661330610513687,
+      "learning_rate": 1.593109348551503e-05,
+      "loss": 1.0801,
+      "step": 12171
+    },
+    {
+      "epoch": 0.7395346011300808,
+      "grad_norm": 0.11382436752319336,
+      "learning_rate": 1.592408749503116e-05,
+      "loss": 1.02,
+      "step": 12172
+    },
+    {
+      "epoch": 0.7395953581627074,
+      "grad_norm": 0.11320145428180695,
+      "learning_rate": 1.5917082753606434e-05,
+      "loss": 1.0246,
+      "step": 12173
+    },
+    {
+      "epoch": 0.7396561151953338,
+      "grad_norm": 0.11681144684553146,
+      "learning_rate": 1.591007926149764e-05,
+      "loss": 1.0359,
+      "step": 12174
+    },
+    {
+      "epoch": 0.7397168722279603,
+      "grad_norm": 0.10601933300495148,
+      "learning_rate": 1.5903077018961477e-05,
+      "loss": 1.0226,
+      "step": 12175
+    },
+    {
+      "epoch": 0.7397776292605869,
+      "grad_norm": 0.10508060455322266,
+      "learning_rate": 1.5896076026254626e-05,
+      "loss": 1.0079,
+      "step": 12176
+    },
+    {
+      "epoch": 0.7398383862932134,
+      "grad_norm": 0.2315458506345749,
+      "learning_rate": 1.5889076283633707e-05,
+      "loss": 1.1132,
+      "step": 12177
+    },
+    {
+      "epoch": 0.73989914332584,
+      "grad_norm": 0.2041950225830078,
+      "learning_rate": 1.5882077791355265e-05,
+      "loss": 1.1536,
+      "step": 12178
+    },
+    {
+      "epoch": 0.7399599003584665,
+      "grad_norm": 0.1524074524641037,
+      "learning_rate": 1.587508054967587e-05,
+      "loss": 1.0637,
+      "step": 12179
+    },
+    {
+      "epoch": 0.740020657391093,
+      "grad_norm": 0.4127258062362671,
+      "learning_rate": 1.5868084558851994e-05,
+      "loss": 1.1612,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7400814144237196,
+      "grad_norm": 3.113245964050293,
+      "learning_rate": 1.5861089819140074e-05,
+      "loss": 1.0365,
+      "step": 12181
+    },
+    {
+      "epoch": 0.7401421714563461,
+      "grad_norm": 0.19555634260177612,
+      "learning_rate": 1.5854096330796507e-05,
+      "loss": 1.087,
+      "step": 12182
+    },
+    {
+      "epoch": 0.7402029284889726,
+      "grad_norm": 0.15480531752109528,
+      "learning_rate": 1.5847104094077636e-05,
+      "loss": 1.0608,
+      "step": 12183
+    },
+    {
+      "epoch": 0.7402636855215992,
+      "grad_norm": 0.20994332432746887,
+      "learning_rate": 1.5840113109239772e-05,
+      "loss": 1.1236,
+      "step": 12184
+    },
+    {
+      "epoch": 0.7403244425542257,
+      "grad_norm": 0.1111488938331604,
+      "learning_rate": 1.583312337653916e-05,
+      "loss": 1.01,
+      "step": 12185
+    },
+    {
+      "epoch": 0.7403851995868522,
+      "grad_norm": 0.17714236676692963,
+      "learning_rate": 1.582613489623202e-05,
+      "loss": 1.0383,
+      "step": 12186
+    },
+    {
+      "epoch": 0.7404459566194787,
+      "grad_norm": 0.27466830611228943,
+      "learning_rate": 1.5819147668574507e-05,
+      "loss": 1.1285,
+      "step": 12187
+    },
+    {
+      "epoch": 0.7405067136521052,
+      "grad_norm": 0.15673334896564484,
+      "learning_rate": 1.5812161693822747e-05,
+      "loss": 1.0882,
+      "step": 12188
+    },
+    {
+      "epoch": 0.7405674706847317,
+      "grad_norm": 12.478750228881836,
+      "learning_rate": 1.580517697223279e-05,
+      "loss": 1.1318,
+      "step": 12189
+    },
+    {
+      "epoch": 0.7406282277173583,
+      "grad_norm": 0.197187140583992,
+      "learning_rate": 1.5798193504060692e-05,
+      "loss": 1.0544,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7406889847499848,
+      "grad_norm": 0.17191274464130402,
+      "learning_rate": 1.5791211289562447e-05,
+      "loss": 1.0347,
+      "step": 12191
+    },
+    {
+      "epoch": 0.7407497417826113,
+      "grad_norm": 0.14423011243343353,
+      "learning_rate": 1.5784230328993942e-05,
+      "loss": 1.1482,
+      "step": 12192
+    },
+    {
+      "epoch": 0.7408104988152379,
+      "grad_norm": 0.11650660634040833,
+      "learning_rate": 1.5777250622611096e-05,
+      "loss": 1.0523,
+      "step": 12193
+    },
+    {
+      "epoch": 0.7408712558478644,
+      "grad_norm": 0.1786743849515915,
+      "learning_rate": 1.577027217066972e-05,
+      "loss": 1.1743,
+      "step": 12194
+    },
+    {
+      "epoch": 0.7409320128804909,
+      "grad_norm": 0.15046550333499908,
+      "learning_rate": 1.576329497342565e-05,
+      "loss": 1.0375,
+      "step": 12195
+    },
+    {
+      "epoch": 0.7409927699131175,
+      "grad_norm": 0.11667760461568832,
+      "learning_rate": 1.5756319031134626e-05,
+      "loss": 1.0528,
+      "step": 12196
+    },
+    {
+      "epoch": 0.741053526945744,
+      "grad_norm": 0.13398396968841553,
+      "learning_rate": 1.574934434405234e-05,
+      "loss": 1.0074,
+      "step": 12197
+    },
+    {
+      "epoch": 0.7411142839783705,
+      "grad_norm": 0.4830251634120941,
+      "learning_rate": 1.574237091243448e-05,
+      "loss": 1.2222,
+      "step": 12198
+    },
+    {
+      "epoch": 0.7411750410109971,
+      "grad_norm": 0.12647002935409546,
+      "learning_rate": 1.57353987365366e-05,
+      "loss": 1.0748,
+      "step": 12199
+    },
+    {
+      "epoch": 0.7412357980436235,
+      "grad_norm": 0.21349944174289703,
+      "learning_rate": 1.572842781661432e-05,
+      "loss": 1.1692,
+      "step": 12200
+    },
+    {
+      "epoch": 0.74129655507625,
+      "grad_norm": 7.244073867797852,
+      "learning_rate": 1.5721458152923142e-05,
+      "loss": 1.0532,
+      "step": 12201
+    },
+    {
+      "epoch": 0.7413573121088766,
+      "grad_norm": 0.11956740915775299,
+      "learning_rate": 1.571448974571854e-05,
+      "loss": 1.0122,
+      "step": 12202
+    },
+    {
+      "epoch": 0.7414180691415031,
+      "grad_norm": 0.13429811596870422,
+      "learning_rate": 1.5707522595255953e-05,
+      "loss": 1.0962,
+      "step": 12203
+    },
+    {
+      "epoch": 0.7414788261741296,
+      "grad_norm": 0.16790010035037994,
+      "learning_rate": 1.570055670179074e-05,
+      "loss": 1.1043,
+      "step": 12204
+    },
+    {
+      "epoch": 0.7415395832067562,
+      "grad_norm": 0.17315034568309784,
+      "learning_rate": 1.569359206557826e-05,
+      "loss": 1.1732,
+      "step": 12205
+    },
+    {
+      "epoch": 0.7416003402393827,
+      "grad_norm": 0.221226766705513,
+      "learning_rate": 1.5686628686873796e-05,
+      "loss": 1.0745,
+      "step": 12206
+    },
+    {
+      "epoch": 0.7416610972720092,
+      "grad_norm": 0.11447381228208542,
+      "learning_rate": 1.5679666565932584e-05,
+      "loss": 1.0063,
+      "step": 12207
+    },
+    {
+      "epoch": 0.7417218543046358,
+      "grad_norm": 0.2961326837539673,
+      "learning_rate": 1.5672705703009828e-05,
+      "loss": 1.0339,
+      "step": 12208
+    },
+    {
+      "epoch": 0.7417826113372623,
+      "grad_norm": 0.572446882724762,
+      "learning_rate": 1.5665746098360685e-05,
+      "loss": 1.1093,
+      "step": 12209
+    },
+    {
+      "epoch": 0.7418433683698888,
+      "grad_norm": 0.7450178861618042,
+      "learning_rate": 1.5658787752240234e-05,
+      "loss": 1.1228,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7419041254025154,
+      "grad_norm": 0.14105729758739471,
+      "learning_rate": 1.5651830664903588e-05,
+      "loss": 1.0317,
+      "step": 12211
+    },
+    {
+      "epoch": 0.7419648824351419,
+      "grad_norm": 0.17359383404254913,
+      "learning_rate": 1.5644874836605716e-05,
+      "loss": 1.079,
+      "step": 12212
+    },
+    {
+      "epoch": 0.7420256394677683,
+      "grad_norm": 0.4730372428894043,
+      "learning_rate": 1.5637920267601597e-05,
+      "loss": 1.1325,
+      "step": 12213
+    },
+    {
+      "epoch": 0.7420863965003949,
+      "grad_norm": 0.1052330955862999,
+      "learning_rate": 1.5630966958146154e-05,
+      "loss": 1.0287,
+      "step": 12214
+    },
+    {
+      "epoch": 0.7421471535330214,
+      "grad_norm": 0.11401963233947754,
+      "learning_rate": 1.562401490849424e-05,
+      "loss": 1.0561,
+      "step": 12215
+    },
+    {
+      "epoch": 0.7422079105656479,
+      "grad_norm": 0.16880464553833008,
+      "learning_rate": 1.561706411890072e-05,
+      "loss": 1.1105,
+      "step": 12216
+    },
+    {
+      "epoch": 0.7422686675982745,
+      "grad_norm": 0.15431378781795502,
+      "learning_rate": 1.5610114589620366e-05,
+      "loss": 1.0083,
+      "step": 12217
+    },
+    {
+      "epoch": 0.742329424630901,
+      "grad_norm": 0.2946162223815918,
+      "learning_rate": 1.560316632090792e-05,
+      "loss": 1.2708,
+      "step": 12218
+    },
+    {
+      "epoch": 0.7423901816635275,
+      "grad_norm": 0.4436761140823364,
+      "learning_rate": 1.559621931301805e-05,
+      "loss": 1.2363,
+      "step": 12219
+    },
+    {
+      "epoch": 0.7424509386961541,
+      "grad_norm": 0.16144710779190063,
+      "learning_rate": 1.5589273566205394e-05,
+      "loss": 1.1268,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7425116957287806,
+      "grad_norm": 0.20668721199035645,
+      "learning_rate": 1.558232908072458e-05,
+      "loss": 1.0069,
+      "step": 12221
+    },
+    {
+      "epoch": 0.7425724527614072,
+      "grad_norm": 0.31884416937828064,
+      "learning_rate": 1.557538585683015e-05,
+      "loss": 1.0605,
+      "step": 12222
+    },
+    {
+      "epoch": 0.7426332097940337,
+      "grad_norm": 0.5737851858139038,
+      "learning_rate": 1.5568443894776602e-05,
+      "loss": 1.028,
+      "step": 12223
+    },
+    {
+      "epoch": 0.7426939668266602,
+      "grad_norm": 0.11045891046524048,
+      "learning_rate": 1.55615031948184e-05,
+      "loss": 1.0736,
+      "step": 12224
+    },
+    {
+      "epoch": 0.7427547238592868,
+      "grad_norm": 0.11405198276042938,
+      "learning_rate": 1.5554563757209956e-05,
+      "loss": 1.063,
+      "step": 12225
+    },
+    {
+      "epoch": 0.7428154808919133,
+      "grad_norm": 0.16240975260734558,
+      "learning_rate": 1.554762558220563e-05,
+      "loss": 1.073,
+      "step": 12226
+    },
+    {
+      "epoch": 0.7428762379245397,
+      "grad_norm": 0.3138273358345032,
+      "learning_rate": 1.5540688670059754e-05,
+      "loss": 1.0593,
+      "step": 12227
+    },
+    {
+      "epoch": 0.7429369949571663,
+      "grad_norm": 0.10800588876008987,
+      "learning_rate": 1.55337530210266e-05,
+      "loss": 1.0841,
+      "step": 12228
+    },
+    {
+      "epoch": 0.7429977519897928,
+      "grad_norm": 4.588499069213867,
+      "learning_rate": 1.552681863536038e-05,
+      "loss": 1.0594,
+      "step": 12229
+    },
+    {
+      "epoch": 0.7430585090224193,
+      "grad_norm": 0.15338660776615143,
+      "learning_rate": 1.551988551331529e-05,
+      "loss": 1.0901,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7431192660550459,
+      "grad_norm": 0.1269010752439499,
+      "learning_rate": 1.5512953655145467e-05,
+      "loss": 1.047,
+      "step": 12231
+    },
+    {
+      "epoch": 0.7431800230876724,
+      "grad_norm": 0.19623033702373505,
+      "learning_rate": 1.550602306110499e-05,
+      "loss": 1.0537,
+      "step": 12232
+    },
+    {
+      "epoch": 0.7432407801202989,
+      "grad_norm": 0.10276135802268982,
+      "learning_rate": 1.5499093731447905e-05,
+      "loss": 1.0172,
+      "step": 12233
+    },
+    {
+      "epoch": 0.7433015371529255,
+      "grad_norm": 0.16195812821388245,
+      "learning_rate": 1.549216566642821e-05,
+      "loss": 1.0106,
+      "step": 12234
+    },
+    {
+      "epoch": 0.743362294185552,
+      "grad_norm": 0.1606951355934143,
+      "learning_rate": 1.548523886629985e-05,
+      "loss": 1.1256,
+      "step": 12235
+    },
+    {
+      "epoch": 0.7434230512181785,
+      "grad_norm": 0.20359116792678833,
+      "learning_rate": 1.547831333131672e-05,
+      "loss": 1.0881,
+      "step": 12236
+    },
+    {
+      "epoch": 0.7434838082508051,
+      "grad_norm": 0.1122780293226242,
+      "learning_rate": 1.54713890617327e-05,
+      "loss": 1.0172,
+      "step": 12237
+    },
+    {
+      "epoch": 0.7435445652834316,
+      "grad_norm": 0.346225768327713,
+      "learning_rate": 1.54644660578016e-05,
+      "loss": 1.0837,
+      "step": 12238
+    },
+    {
+      "epoch": 0.7436053223160581,
+      "grad_norm": 0.2300349622964859,
+      "learning_rate": 1.5457544319777163e-05,
+      "loss": 1.15,
+      "step": 12239
+    },
+    {
+      "epoch": 0.7436660793486846,
+      "grad_norm": 0.11195831745862961,
+      "learning_rate": 1.54506238479131e-05,
+      "loss": 1.0275,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7437268363813111,
+      "grad_norm": 0.17067274451255798,
+      "learning_rate": 1.544370464246311e-05,
+      "loss": 1.0225,
+      "step": 12241
+    },
+    {
+      "epoch": 0.7437875934139376,
+      "grad_norm": 0.15167181193828583,
+      "learning_rate": 1.5436786703680806e-05,
+      "loss": 1.0461,
+      "step": 12242
+    },
+    {
+      "epoch": 0.7438483504465642,
+      "grad_norm": 0.17596226930618286,
+      "learning_rate": 1.5429870031819766e-05,
+      "loss": 1.0996,
+      "step": 12243
+    },
+    {
+      "epoch": 0.7439091074791907,
+      "grad_norm": 0.15617777407169342,
+      "learning_rate": 1.542295462713354e-05,
+      "loss": 1.0769,
+      "step": 12244
+    },
+    {
+      "epoch": 0.7439698645118172,
+      "grad_norm": 0.7419642210006714,
+      "learning_rate": 1.5416040489875565e-05,
+      "loss": 1.0329,
+      "step": 12245
+    },
+    {
+      "epoch": 0.7440306215444438,
+      "grad_norm": 0.15686364471912384,
+      "learning_rate": 1.5409127620299326e-05,
+      "loss": 0.9753,
+      "step": 12246
+    },
+    {
+      "epoch": 0.7440913785770703,
+      "grad_norm": 0.4805592894554138,
+      "learning_rate": 1.54022160186582e-05,
+      "loss": 1.0545,
+      "step": 12247
+    },
+    {
+      "epoch": 0.7441521356096968,
+      "grad_norm": 0.14140735566616058,
+      "learning_rate": 1.539530568520553e-05,
+      "loss": 1.0401,
+      "step": 12248
+    },
+    {
+      "epoch": 0.7442128926423234,
+      "grad_norm": 0.9698283076286316,
+      "learning_rate": 1.5388396620194623e-05,
+      "loss": 1.1128,
+      "step": 12249
+    },
+    {
+      "epoch": 0.7442736496749499,
+      "grad_norm": 0.10790181159973145,
+      "learning_rate": 1.5381488823878725e-05,
+      "loss": 1.0545,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7443344067075764,
+      "grad_norm": 0.21961985528469086,
+      "learning_rate": 1.5374582296511053e-05,
+      "loss": 1.1129,
+      "step": 12251
+    },
+    {
+      "epoch": 0.744395163740203,
+      "grad_norm": 0.11807654798030853,
+      "learning_rate": 1.5367677038344754e-05,
+      "loss": 1.016,
+      "step": 12252
+    },
+    {
+      "epoch": 0.7444559207728294,
+      "grad_norm": 0.20182126760482788,
+      "learning_rate": 1.536077304963295e-05,
+      "loss": 1.0361,
+      "step": 12253
+    },
+    {
+      "epoch": 0.7445166778054559,
+      "grad_norm": 0.12720921635627747,
+      "learning_rate": 1.53538703306287e-05,
+      "loss": 0.9927,
+      "step": 12254
+    },
+    {
+      "epoch": 0.7445774348380825,
+      "grad_norm": 0.18298062682151794,
+      "learning_rate": 1.5346968881585034e-05,
+      "loss": 1.212,
+      "step": 12255
+    },
+    {
+      "epoch": 0.744638191870709,
+      "grad_norm": 0.18976110219955444,
+      "learning_rate": 1.5340068702754906e-05,
+      "loss": 1.0713,
+      "step": 12256
+    },
+    {
+      "epoch": 0.7446989489033355,
+      "grad_norm": 0.16418978571891785,
+      "learning_rate": 1.533316979439128e-05,
+      "loss": 1.0432,
+      "step": 12257
+    },
+    {
+      "epoch": 0.7447597059359621,
+      "grad_norm": 0.10490068793296814,
+      "learning_rate": 1.5326272156747024e-05,
+      "loss": 1.0197,
+      "step": 12258
+    },
+    {
+      "epoch": 0.7448204629685886,
+      "grad_norm": 0.10313988476991653,
+      "learning_rate": 1.531937579007495e-05,
+      "loss": 1.0386,
+      "step": 12259
+    },
+    {
+      "epoch": 0.7448812200012151,
+      "grad_norm": 0.11124314367771149,
+      "learning_rate": 1.5312480694627862e-05,
+      "loss": 1.0259,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7449419770338417,
+      "grad_norm": 0.12163428962230682,
+      "learning_rate": 1.5305586870658485e-05,
+      "loss": 1.098,
+      "step": 12261
+    },
+    {
+      "epoch": 0.7450027340664682,
+      "grad_norm": 5.009799003601074,
+      "learning_rate": 1.5298694318419536e-05,
+      "loss": 1.0474,
+      "step": 12262
+    },
+    {
+      "epoch": 0.7450634910990948,
+      "grad_norm": 0.19056597352027893,
+      "learning_rate": 1.5291803038163665e-05,
+      "loss": 1.0466,
+      "step": 12263
+    },
+    {
+      "epoch": 0.7451242481317213,
+      "grad_norm": 0.11156143248081207,
+      "learning_rate": 1.5284913030143472e-05,
+      "loss": 1.0277,
+      "step": 12264
+    },
+    {
+      "epoch": 0.7451850051643478,
+      "grad_norm": 0.15466099977493286,
+      "learning_rate": 1.5278024294611487e-05,
+      "loss": 1.0494,
+      "step": 12265
+    },
+    {
+      "epoch": 0.7452457621969742,
+      "grad_norm": 0.11629234254360199,
+      "learning_rate": 1.5271136831820214e-05,
+      "loss": 1.0384,
+      "step": 12266
+    },
+    {
+      "epoch": 0.7453065192296008,
+      "grad_norm": 0.14503991603851318,
+      "learning_rate": 1.5264250642022153e-05,
+      "loss": 1.0128,
+      "step": 12267
+    },
+    {
+      "epoch": 0.7453672762622273,
+      "grad_norm": 0.1715267300605774,
+      "learning_rate": 1.5257365725469703e-05,
+      "loss": 1.159,
+      "step": 12268
+    },
+    {
+      "epoch": 0.7454280332948539,
+      "grad_norm": 0.18305064737796783,
+      "learning_rate": 1.5250482082415223e-05,
+      "loss": 1.2106,
+      "step": 12269
+    },
+    {
+      "epoch": 0.7454887903274804,
+      "grad_norm": 0.8006494641304016,
+      "learning_rate": 1.524359971311104e-05,
+      "loss": 1.1424,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7455495473601069,
+      "grad_norm": 0.25118541717529297,
+      "learning_rate": 1.5236718617809425e-05,
+      "loss": 1.143,
+      "step": 12271
+    },
+    {
+      "epoch": 0.7456103043927335,
+      "grad_norm": 0.12858976423740387,
+      "learning_rate": 1.5229838796762613e-05,
+      "loss": 1.1199,
+      "step": 12272
+    },
+    {
+      "epoch": 0.74567106142536,
+      "grad_norm": 0.114949531853199,
+      "learning_rate": 1.522296025022278e-05,
+      "loss": 1.0263,
+      "step": 12273
+    },
+    {
+      "epoch": 0.7457318184579865,
+      "grad_norm": 0.3934478759765625,
+      "learning_rate": 1.521608297844206e-05,
+      "loss": 1.0977,
+      "step": 12274
+    },
+    {
+      "epoch": 0.7457925754906131,
+      "grad_norm": 0.10857676714658737,
+      "learning_rate": 1.520920698167254e-05,
+      "loss": 1.0782,
+      "step": 12275
+    },
+    {
+      "epoch": 0.7458533325232396,
+      "grad_norm": 0.22125573456287384,
+      "learning_rate": 1.5202332260166269e-05,
+      "loss": 1.1118,
+      "step": 12276
+    },
+    {
+      "epoch": 0.7459140895558661,
+      "grad_norm": 0.16291722655296326,
+      "learning_rate": 1.5195458814175217e-05,
+      "loss": 1.0884,
+      "step": 12277
+    },
+    {
+      "epoch": 0.7459748465884927,
+      "grad_norm": 0.17648471891880035,
+      "learning_rate": 1.518858664395138e-05,
+      "loss": 1.0698,
+      "step": 12278
+    },
+    {
+      "epoch": 0.7460356036211191,
+      "grad_norm": 0.12250689417123795,
+      "learning_rate": 1.518171574974661e-05,
+      "loss": 1.014,
+      "step": 12279
+    },
+    {
+      "epoch": 0.7460963606537456,
+      "grad_norm": 0.27059710025787354,
+      "learning_rate": 1.5174846131812787e-05,
+      "loss": 1.078,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7461571176863722,
+      "grad_norm": 0.09177951514720917,
+      "learning_rate": 1.5167977790401705e-05,
+      "loss": 0.978,
+      "step": 12281
+    },
+    {
+      "epoch": 0.7462178747189987,
+      "grad_norm": 0.12109772115945816,
+      "learning_rate": 1.5161110725765116e-05,
+      "loss": 1.0156,
+      "step": 12282
+    },
+    {
+      "epoch": 0.7462786317516252,
+      "grad_norm": 0.11416082084178925,
+      "learning_rate": 1.515424493815476e-05,
+      "loss": 1.0002,
+      "step": 12283
+    },
+    {
+      "epoch": 0.7463393887842518,
+      "grad_norm": 0.12184463441371918,
+      "learning_rate": 1.5147380427822295e-05,
+      "loss": 1.0733,
+      "step": 12284
+    },
+    {
+      "epoch": 0.7464001458168783,
+      "grad_norm": 0.1866767257452011,
+      "learning_rate": 1.514051719501935e-05,
+      "loss": 1.0614,
+      "step": 12285
+    },
+    {
+      "epoch": 0.7464609028495048,
+      "grad_norm": 0.6591934561729431,
+      "learning_rate": 1.5133655239997474e-05,
+      "loss": 1.21,
+      "step": 12286
+    },
+    {
+      "epoch": 0.7465216598821314,
+      "grad_norm": 9.733088493347168,
+      "learning_rate": 1.5126794563008184e-05,
+      "loss": 1.1255,
+      "step": 12287
+    },
+    {
+      "epoch": 0.7465824169147579,
+      "grad_norm": 0.14364604651927948,
+      "learning_rate": 1.5119935164303e-05,
+      "loss": 1.0558,
+      "step": 12288
+    },
+    {
+      "epoch": 0.7466431739473844,
+      "grad_norm": 0.12493626773357391,
+      "learning_rate": 1.5113077044133334e-05,
+      "loss": 1.0376,
+      "step": 12289
+    },
+    {
+      "epoch": 0.746703930980011,
+      "grad_norm": 0.18352729082107544,
+      "learning_rate": 1.510622020275057e-05,
+      "loss": 1.1661,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7467646880126375,
+      "grad_norm": 0.16050124168395996,
+      "learning_rate": 1.5099364640406066e-05,
+      "loss": 1.059,
+      "step": 12291
+    },
+    {
+      "epoch": 0.7468254450452639,
+      "grad_norm": 0.5644650459289551,
+      "learning_rate": 1.509251035735107e-05,
+      "loss": 1.2231,
+      "step": 12292
+    },
+    {
+      "epoch": 0.7468862020778905,
+      "grad_norm": 0.15874862670898438,
+      "learning_rate": 1.5085657353836863e-05,
+      "loss": 1.0652,
+      "step": 12293
+    },
+    {
+      "epoch": 0.746946959110517,
+      "grad_norm": 0.1588876098394394,
+      "learning_rate": 1.5078805630114639e-05,
+      "loss": 1.1025,
+      "step": 12294
+    },
+    {
+      "epoch": 0.7470077161431435,
+      "grad_norm": 0.10968033969402313,
+      "learning_rate": 1.5071955186435544e-05,
+      "loss": 1.0647,
+      "step": 12295
+    },
+    {
+      "epoch": 0.7470684731757701,
+      "grad_norm": 0.19421853125095367,
+      "learning_rate": 1.506510602305068e-05,
+      "loss": 1.0627,
+      "step": 12296
+    },
+    {
+      "epoch": 0.7471292302083966,
+      "grad_norm": 0.1826222538948059,
+      "learning_rate": 1.5058258140211112e-05,
+      "loss": 1.1209,
+      "step": 12297
+    },
+    {
+      "epoch": 0.7471899872410231,
+      "grad_norm": 0.20129136741161346,
+      "learning_rate": 1.5051411538167843e-05,
+      "loss": 1.1843,
+      "step": 12298
+    },
+    {
+      "epoch": 0.7472507442736497,
+      "grad_norm": 0.20420442521572113,
+      "learning_rate": 1.504456621717184e-05,
+      "loss": 1.0034,
+      "step": 12299
+    },
+    {
+      "epoch": 0.7473115013062762,
+      "grad_norm": 0.23053786158561707,
+      "learning_rate": 1.5037722177474023e-05,
+      "loss": 1.0401,
+      "step": 12300
+    },
+    {
+      "epoch": 0.7473722583389028,
+      "grad_norm": 0.16696183383464813,
+      "learning_rate": 1.503087941932525e-05,
+      "loss": 1.034,
+      "step": 12301
+    },
+    {
+      "epoch": 0.7474330153715293,
+      "grad_norm": 0.171365886926651,
+      "learning_rate": 1.502403794297636e-05,
+      "loss": 1.0115,
+      "step": 12302
+    },
+    {
+      "epoch": 0.7474937724041558,
+      "grad_norm": 0.20434030890464783,
+      "learning_rate": 1.5017197748678096e-05,
+      "loss": 1.0599,
+      "step": 12303
+    },
+    {
+      "epoch": 0.7475545294367824,
+      "grad_norm": 0.24190174043178558,
+      "learning_rate": 1.5010358836681232e-05,
+      "loss": 1.0744,
+      "step": 12304
+    },
+    {
+      "epoch": 0.7476152864694088,
+      "grad_norm": 0.13057444989681244,
+      "learning_rate": 1.5003521207236443e-05,
+      "loss": 1.0539,
+      "step": 12305
+    },
+    {
+      "epoch": 0.7476760435020353,
+      "grad_norm": 0.14971034228801727,
+      "learning_rate": 1.4996684860594329e-05,
+      "loss": 1.073,
+      "step": 12306
+    },
+    {
+      "epoch": 0.7477368005346618,
+      "grad_norm": 0.19243723154067993,
+      "learning_rate": 1.49898497970055e-05,
+      "loss": 1.0818,
+      "step": 12307
+    },
+    {
+      "epoch": 0.7477975575672884,
+      "grad_norm": 0.14464719593524933,
+      "learning_rate": 1.498301601672048e-05,
+      "loss": 1.0696,
+      "step": 12308
+    },
+    {
+      "epoch": 0.7478583145999149,
+      "grad_norm": 0.12280137836933136,
+      "learning_rate": 1.497618351998979e-05,
+      "loss": 1.0549,
+      "step": 12309
+    },
+    {
+      "epoch": 0.7479190716325415,
+      "grad_norm": 0.12875477969646454,
+      "learning_rate": 1.4969352307063866e-05,
+      "loss": 1.0963,
+      "step": 12310
+    },
+    {
+      "epoch": 0.747979828665168,
+      "grad_norm": 0.8757300972938538,
+      "learning_rate": 1.4962522378193118e-05,
+      "loss": 1.1342,
+      "step": 12311
+    },
+    {
+      "epoch": 0.7480405856977945,
+      "grad_norm": 0.11182583123445511,
+      "learning_rate": 1.4955693733627867e-05,
+      "loss": 1.0095,
+      "step": 12312
+    },
+    {
+      "epoch": 0.7481013427304211,
+      "grad_norm": 0.9983725547790527,
+      "learning_rate": 1.4948866373618426e-05,
+      "loss": 1.0637,
+      "step": 12313
+    },
+    {
+      "epoch": 0.7481620997630476,
+      "grad_norm": 0.1568031758069992,
+      "learning_rate": 1.4942040298415072e-05,
+      "loss": 0.9961,
+      "step": 12314
+    },
+    {
+      "epoch": 0.7482228567956741,
+      "grad_norm": 0.266242653131485,
+      "learning_rate": 1.4935215508268014e-05,
+      "loss": 1.1979,
+      "step": 12315
+    },
+    {
+      "epoch": 0.7482836138283007,
+      "grad_norm": 0.16601009666919708,
+      "learning_rate": 1.4928392003427404e-05,
+      "loss": 1.0398,
+      "step": 12316
+    },
+    {
+      "epoch": 0.7483443708609272,
+      "grad_norm": 2.6673924922943115,
+      "learning_rate": 1.4921569784143368e-05,
+      "loss": 1.037,
+      "step": 12317
+    },
+    {
+      "epoch": 0.7484051278935536,
+      "grad_norm": 0.12594975531101227,
+      "learning_rate": 1.491474885066597e-05,
+      "loss": 1.0705,
+      "step": 12318
+    },
+    {
+      "epoch": 0.7484658849261802,
+      "grad_norm": 0.17537593841552734,
+      "learning_rate": 1.4907929203245235e-05,
+      "loss": 1.1044,
+      "step": 12319
+    },
+    {
+      "epoch": 0.7485266419588067,
+      "grad_norm": 0.12230952829122543,
+      "learning_rate": 1.4901110842131138e-05,
+      "loss": 1.02,
+      "step": 12320
+    },
+    {
+      "epoch": 0.7485873989914332,
+      "grad_norm": 0.18852370977401733,
+      "learning_rate": 1.4894293767573613e-05,
+      "loss": 1.082,
+      "step": 12321
+    },
+    {
+      "epoch": 0.7486481560240598,
+      "grad_norm": 0.11276037991046906,
+      "learning_rate": 1.488747797982253e-05,
+      "loss": 1.093,
+      "step": 12322
+    },
+    {
+      "epoch": 0.7487089130566863,
+      "grad_norm": 0.11569953709840775,
+      "learning_rate": 1.4880663479127727e-05,
+      "loss": 1.0358,
+      "step": 12323
+    },
+    {
+      "epoch": 0.7487696700893128,
+      "grad_norm": 0.622245728969574,
+      "learning_rate": 1.487385026573898e-05,
+      "loss": 1.1218,
+      "step": 12324
+    },
+    {
+      "epoch": 0.7488304271219394,
+      "grad_norm": 0.21036945283412933,
+      "learning_rate": 1.4867038339906076e-05,
+      "loss": 1.1005,
+      "step": 12325
+    },
+    {
+      "epoch": 0.7488911841545659,
+      "grad_norm": 0.11691872030496597,
+      "learning_rate": 1.4860227701878654e-05,
+      "loss": 1.0575,
+      "step": 12326
+    },
+    {
+      "epoch": 0.7489519411871924,
+      "grad_norm": 0.11479263007640839,
+      "learning_rate": 1.4853418351906374e-05,
+      "loss": 1.0512,
+      "step": 12327
+    },
+    {
+      "epoch": 0.749012698219819,
+      "grad_norm": 0.11864512413740158,
+      "learning_rate": 1.4846610290238843e-05,
+      "loss": 1.0454,
+      "step": 12328
+    },
+    {
+      "epoch": 0.7490734552524455,
+      "grad_norm": 0.18619774281978607,
+      "learning_rate": 1.483980351712559e-05,
+      "loss": 1.0688,
+      "step": 12329
+    },
+    {
+      "epoch": 0.749134212285072,
+      "grad_norm": 0.23992513120174408,
+      "learning_rate": 1.4832998032816148e-05,
+      "loss": 1.1711,
+      "step": 12330
+    },
+    {
+      "epoch": 0.7491949693176986,
+      "grad_norm": 0.24409279227256775,
+      "learning_rate": 1.4826193837559981e-05,
+      "loss": 1.2871,
+      "step": 12331
+    },
+    {
+      "epoch": 0.749255726350325,
+      "grad_norm": 0.14463911950588226,
+      "learning_rate": 1.4819390931606458e-05,
+      "loss": 1.0566,
+      "step": 12332
+    },
+    {
+      "epoch": 0.7493164833829515,
+      "grad_norm": 0.17773549258708954,
+      "learning_rate": 1.481258931520495e-05,
+      "loss": 1.1099,
+      "step": 12333
+    },
+    {
+      "epoch": 0.7493772404155781,
+      "grad_norm": 0.13054946064949036,
+      "learning_rate": 1.4805788988604791e-05,
+      "loss": 1.0594,
+      "step": 12334
+    },
+    {
+      "epoch": 0.7494379974482046,
+      "grad_norm": 0.19657039642333984,
+      "learning_rate": 1.4798989952055248e-05,
+      "loss": 1.0853,
+      "step": 12335
+    },
+    {
+      "epoch": 0.7494987544808311,
+      "grad_norm": 0.11151206493377686,
+      "learning_rate": 1.4792192205805527e-05,
+      "loss": 0.9878,
+      "step": 12336
+    },
+    {
+      "epoch": 0.7495595115134577,
+      "grad_norm": 0.1223205104470253,
+      "learning_rate": 1.4785395750104808e-05,
+      "loss": 1.0022,
+      "step": 12337
+    },
+    {
+      "epoch": 0.7496202685460842,
+      "grad_norm": 1.025681972503662,
+      "learning_rate": 1.4778600585202213e-05,
+      "loss": 1.018,
+      "step": 12338
+    },
+    {
+      "epoch": 0.7496810255787107,
+      "grad_norm": 0.920658528804779,
+      "learning_rate": 1.477180671134682e-05,
+      "loss": 1.067,
+      "step": 12339
+    },
+    {
+      "epoch": 0.7497417826113373,
+      "grad_norm": 0.4540456235408783,
+      "learning_rate": 1.4765014128787663e-05,
+      "loss": 1.1926,
+      "step": 12340
+    },
+    {
+      "epoch": 0.7498025396439638,
+      "grad_norm": 0.19764259457588196,
+      "learning_rate": 1.4758222837773727e-05,
+      "loss": 1.0697,
+      "step": 12341
+    },
+    {
+      "epoch": 0.7498632966765904,
+      "grad_norm": 0.940133810043335,
+      "learning_rate": 1.475143283855394e-05,
+      "loss": 1.1281,
+      "step": 12342
+    },
+    {
+      "epoch": 0.7499240537092169,
+      "grad_norm": 0.1195477619767189,
+      "learning_rate": 1.47446441313772e-05,
+      "loss": 1.0252,
+      "step": 12343
+    },
+    {
+      "epoch": 0.7499848107418434,
+      "grad_norm": 0.19660130143165588,
+      "learning_rate": 1.4737856716492344e-05,
+      "loss": 1.0857,
+      "step": 12344
+    },
+    {
+      "epoch": 0.7500455677744698,
+      "grad_norm": 0.16826272010803223,
+      "learning_rate": 1.4731070594148166e-05,
+      "loss": 1.0879,
+      "step": 12345
+    },
+    {
+      "epoch": 0.7501063248070964,
+      "grad_norm": 0.2910539507865906,
+      "learning_rate": 1.472428576459341e-05,
+      "loss": 1.1754,
+      "step": 12346
+    },
+    {
+      "epoch": 0.7501670818397229,
+      "grad_norm": 0.21095766127109528,
+      "learning_rate": 1.4717502228076779e-05,
+      "loss": 1.106,
+      "step": 12347
+    },
+    {
+      "epoch": 0.7502278388723495,
+      "grad_norm": 0.21968616545200348,
+      "learning_rate": 1.4710719984846927e-05,
+      "loss": 1.1062,
+      "step": 12348
+    },
+    {
+      "epoch": 0.750288595904976,
+      "grad_norm": 0.15479840338230133,
+      "learning_rate": 1.4703939035152436e-05,
+      "loss": 1.0531,
+      "step": 12349
+    },
+    {
+      "epoch": 0.7503493529376025,
+      "grad_norm": 0.14020900428295135,
+      "learning_rate": 1.4697159379241899e-05,
+      "loss": 1.05,
+      "step": 12350
+    },
+    {
+      "epoch": 0.7504101099702291,
+      "grad_norm": 0.15454229712486267,
+      "learning_rate": 1.4690381017363808e-05,
+      "loss": 1.0851,
+      "step": 12351
+    },
+    {
+      "epoch": 0.7504708670028556,
+      "grad_norm": 0.30164051055908203,
+      "learning_rate": 1.4683603949766639e-05,
+      "loss": 1.0625,
+      "step": 12352
+    },
+    {
+      "epoch": 0.7505316240354821,
+      "grad_norm": 0.17074117064476013,
+      "learning_rate": 1.4676828176698782e-05,
+      "loss": 1.065,
+      "step": 12353
+    },
+    {
+      "epoch": 0.7505923810681087,
+      "grad_norm": 0.12957727909088135,
+      "learning_rate": 1.4670053698408598e-05,
+      "loss": 1.0886,
+      "step": 12354
+    },
+    {
+      "epoch": 0.7506531381007352,
+      "grad_norm": 0.21802391111850739,
+      "learning_rate": 1.4663280515144445e-05,
+      "loss": 1.2373,
+      "step": 12355
+    },
+    {
+      "epoch": 0.7507138951333617,
+      "grad_norm": 0.11143873631954193,
+      "learning_rate": 1.465650862715457e-05,
+      "loss": 1.028,
+      "step": 12356
+    },
+    {
+      "epoch": 0.7507746521659883,
+      "grad_norm": 0.22323362529277802,
+      "learning_rate": 1.4649738034687206e-05,
+      "loss": 1.1476,
+      "step": 12357
+    },
+    {
+      "epoch": 0.7508354091986147,
+      "grad_norm": 0.16581158339977264,
+      "learning_rate": 1.4642968737990543e-05,
+      "loss": 1.111,
+      "step": 12358
+    },
+    {
+      "epoch": 0.7508961662312412,
+      "grad_norm": 0.25218355655670166,
+      "learning_rate": 1.463620073731266e-05,
+      "loss": 1.0233,
+      "step": 12359
+    },
+    {
+      "epoch": 0.7509569232638678,
+      "grad_norm": 0.15309728682041168,
+      "learning_rate": 1.4629434032901696e-05,
+      "loss": 1.1349,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7510176802964943,
+      "grad_norm": 0.16351532936096191,
+      "learning_rate": 1.462266862500566e-05,
+      "loss": 1.0063,
+      "step": 12361
+    },
+    {
+      "epoch": 0.7510784373291208,
+      "grad_norm": 0.12894141674041748,
+      "learning_rate": 1.4615904513872547e-05,
+      "loss": 1.1062,
+      "step": 12362
+    },
+    {
+      "epoch": 0.7511391943617474,
+      "grad_norm": 0.16792377829551697,
+      "learning_rate": 1.4609141699750295e-05,
+      "loss": 1.0515,
+      "step": 12363
+    },
+    {
+      "epoch": 0.7511999513943739,
+      "grad_norm": 0.11803977936506271,
+      "learning_rate": 1.4602380182886788e-05,
+      "loss": 1.0415,
+      "step": 12364
+    },
+    {
+      "epoch": 0.7512607084270004,
+      "grad_norm": 0.1203421950340271,
+      "learning_rate": 1.4595619963529883e-05,
+      "loss": 1.024,
+      "step": 12365
+    },
+    {
+      "epoch": 0.751321465459627,
+      "grad_norm": 0.1662909984588623,
+      "learning_rate": 1.458886104192737e-05,
+      "loss": 1.0982,
+      "step": 12366
+    },
+    {
+      "epoch": 0.7513822224922535,
+      "grad_norm": 0.19591106474399567,
+      "learning_rate": 1.4582103418327003e-05,
+      "loss": 1.0386,
+      "step": 12367
+    },
+    {
+      "epoch": 0.75144297952488,
+      "grad_norm": 0.16303405165672302,
+      "learning_rate": 1.4575347092976477e-05,
+      "loss": 1.0542,
+      "step": 12368
+    },
+    {
+      "epoch": 0.7515037365575066,
+      "grad_norm": 0.8739423751831055,
+      "learning_rate": 1.4568592066123454e-05,
+      "loss": 1.202,
+      "step": 12369
+    },
+    {
+      "epoch": 0.7515644935901331,
+      "grad_norm": 0.13482946157455444,
+      "learning_rate": 1.456183833801552e-05,
+      "loss": 1.0478,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7516252506227595,
+      "grad_norm": 0.21555595099925995,
+      "learning_rate": 1.4555085908900267e-05,
+      "loss": 1.1375,
+      "step": 12371
+    },
+    {
+      "epoch": 0.7516860076553861,
+      "grad_norm": 0.15941737592220306,
+      "learning_rate": 1.4548334779025213e-05,
+      "loss": 1.097,
+      "step": 12372
+    },
+    {
+      "epoch": 0.7517467646880126,
+      "grad_norm": 0.11273514479398727,
+      "learning_rate": 1.4541584948637776e-05,
+      "loss": 1.0163,
+      "step": 12373
+    },
+    {
+      "epoch": 0.7518075217206391,
+      "grad_norm": 0.18924646079540253,
+      "learning_rate": 1.4534836417985403e-05,
+      "loss": 1.0429,
+      "step": 12374
+    },
+    {
+      "epoch": 0.7518682787532657,
+      "grad_norm": 0.1637611985206604,
+      "learning_rate": 1.4528089187315436e-05,
+      "loss": 1.1046,
+      "step": 12375
+    },
+    {
+      "epoch": 0.7519290357858922,
+      "grad_norm": 0.1842072606086731,
+      "learning_rate": 1.4521343256875235e-05,
+      "loss": 1.087,
+      "step": 12376
+    },
+    {
+      "epoch": 0.7519897928185187,
+      "grad_norm": 0.38574734330177307,
+      "learning_rate": 1.4514598626912057e-05,
+      "loss": 1.1483,
+      "step": 12377
+    },
+    {
+      "epoch": 0.7520505498511453,
+      "grad_norm": 0.30958425998687744,
+      "learning_rate": 1.450785529767314e-05,
+      "loss": 1.1951,
+      "step": 12378
+    },
+    {
+      "epoch": 0.7521113068837718,
+      "grad_norm": 0.14373669028282166,
+      "learning_rate": 1.450111326940563e-05,
+      "loss": 1.0685,
+      "step": 12379
+    },
+    {
+      "epoch": 0.7521720639163983,
+      "grad_norm": 0.742682158946991,
+      "learning_rate": 1.449437254235666e-05,
+      "loss": 1.0204,
+      "step": 12380
+    },
+    {
+      "epoch": 0.7522328209490249,
+      "grad_norm": 0.21696442365646362,
+      "learning_rate": 1.4487633116773341e-05,
+      "loss": 1.0672,
+      "step": 12381
+    },
+    {
+      "epoch": 0.7522935779816514,
+      "grad_norm": 0.6758202314376831,
+      "learning_rate": 1.4480894992902694e-05,
+      "loss": 1.239,
+      "step": 12382
+    },
+    {
+      "epoch": 0.752354335014278,
+      "grad_norm": 0.16507484018802643,
+      "learning_rate": 1.4474158170991708e-05,
+      "loss": 1.007,
+      "step": 12383
+    },
+    {
+      "epoch": 0.7524150920469044,
+      "grad_norm": 0.2582539916038513,
+      "learning_rate": 1.4467422651287326e-05,
+      "loss": 1.1495,
+      "step": 12384
+    },
+    {
+      "epoch": 0.7524758490795309,
+      "grad_norm": 0.1980871707201004,
+      "learning_rate": 1.4460688434036429e-05,
+      "loss": 0.9937,
+      "step": 12385
+    },
+    {
+      "epoch": 0.7525366061121574,
+      "grad_norm": 0.23131780326366425,
+      "learning_rate": 1.4453955519485868e-05,
+      "loss": 1.0921,
+      "step": 12386
+    },
+    {
+      "epoch": 0.752597363144784,
+      "grad_norm": 0.14088793098926544,
+      "learning_rate": 1.444722390788244e-05,
+      "loss": 1.0398,
+      "step": 12387
+    },
+    {
+      "epoch": 0.7526581201774105,
+      "grad_norm": 0.12949642539024353,
+      "learning_rate": 1.4440493599472894e-05,
+      "loss": 1.0913,
+      "step": 12388
+    },
+    {
+      "epoch": 0.752718877210037,
+      "grad_norm": 0.15121890604496002,
+      "learning_rate": 1.4433764594503923e-05,
+      "loss": 1.1299,
+      "step": 12389
+    },
+    {
+      "epoch": 0.7527796342426636,
+      "grad_norm": 0.5240110158920288,
+      "learning_rate": 1.442703689322219e-05,
+      "loss": 1.2855,
+      "step": 12390
+    },
+    {
+      "epoch": 0.7528403912752901,
+      "grad_norm": 0.17915916442871094,
+      "learning_rate": 1.4420310495874279e-05,
+      "loss": 1.1821,
+      "step": 12391
+    },
+    {
+      "epoch": 0.7529011483079167,
+      "grad_norm": 0.157648965716362,
+      "learning_rate": 1.4413585402706798e-05,
+      "loss": 1.0633,
+      "step": 12392
+    },
+    {
+      "epoch": 0.7529619053405432,
+      "grad_norm": 0.14002954959869385,
+      "learning_rate": 1.4406861613966205e-05,
+      "loss": 1.0221,
+      "step": 12393
+    },
+    {
+      "epoch": 0.7530226623731697,
+      "grad_norm": 0.2535688877105713,
+      "learning_rate": 1.4400139129898977e-05,
+      "loss": 1.1571,
+      "step": 12394
+    },
+    {
+      "epoch": 0.7530834194057963,
+      "grad_norm": 0.11355717480182648,
+      "learning_rate": 1.439341795075153e-05,
+      "loss": 1.0427,
+      "step": 12395
+    },
+    {
+      "epoch": 0.7531441764384228,
+      "grad_norm": 0.20710282027721405,
+      "learning_rate": 1.4386698076770216e-05,
+      "loss": 1.0965,
+      "step": 12396
+    },
+    {
+      "epoch": 0.7532049334710492,
+      "grad_norm": 0.23512859642505646,
+      "learning_rate": 1.4379979508201385e-05,
+      "loss": 1.1267,
+      "step": 12397
+    },
+    {
+      "epoch": 0.7532656905036758,
+      "grad_norm": 0.19728349149227142,
+      "learning_rate": 1.4373262245291308e-05,
+      "loss": 1.1702,
+      "step": 12398
+    },
+    {
+      "epoch": 0.7533264475363023,
+      "grad_norm": 0.10688597708940506,
+      "learning_rate": 1.4366546288286176e-05,
+      "loss": 0.9936,
+      "step": 12399
+    },
+    {
+      "epoch": 0.7533872045689288,
+      "grad_norm": 0.15643784403800964,
+      "learning_rate": 1.435983163743217e-05,
+      "loss": 1.1451,
+      "step": 12400
+    },
+    {
+      "epoch": 0.7534479616015554,
+      "grad_norm": 0.914742648601532,
+      "learning_rate": 1.4353118292975414e-05,
+      "loss": 1.1157,
+      "step": 12401
+    },
+    {
+      "epoch": 0.7535087186341819,
+      "grad_norm": 0.17047327756881714,
+      "learning_rate": 1.434640625516201e-05,
+      "loss": 1.0628,
+      "step": 12402
+    },
+    {
+      "epoch": 0.7535694756668084,
+      "grad_norm": 0.1906994879245758,
+      "learning_rate": 1.4339695524237979e-05,
+      "loss": 1.1583,
+      "step": 12403
+    },
+    {
+      "epoch": 0.753630232699435,
+      "grad_norm": 0.20968379080295563,
+      "learning_rate": 1.4332986100449303e-05,
+      "loss": 1.099,
+      "step": 12404
+    },
+    {
+      "epoch": 0.7536909897320615,
+      "grad_norm": 0.15112251043319702,
+      "learning_rate": 1.4326277984041925e-05,
+      "loss": 1.0748,
+      "step": 12405
+    },
+    {
+      "epoch": 0.753751746764688,
+      "grad_norm": 0.18442794680595398,
+      "learning_rate": 1.4319571175261698e-05,
+      "loss": 1.0924,
+      "step": 12406
+    },
+    {
+      "epoch": 0.7538125037973146,
+      "grad_norm": 0.11638113856315613,
+      "learning_rate": 1.4312865674354497e-05,
+      "loss": 1.0333,
+      "step": 12407
+    },
+    {
+      "epoch": 0.7538732608299411,
+      "grad_norm": 0.21298284828662872,
+      "learning_rate": 1.4306161481566105e-05,
+      "loss": 1.0891,
+      "step": 12408
+    },
+    {
+      "epoch": 0.7539340178625676,
+      "grad_norm": 0.10447728633880615,
+      "learning_rate": 1.4299458597142267e-05,
+      "loss": 1.0405,
+      "step": 12409
+    },
+    {
+      "epoch": 0.7539947748951941,
+      "grad_norm": 0.168282613158226,
+      "learning_rate": 1.4292757021328673e-05,
+      "loss": 1.0638,
+      "step": 12410
+    },
+    {
+      "epoch": 0.7540555319278206,
+      "grad_norm": 0.16812369227409363,
+      "learning_rate": 1.4286056754370968e-05,
+      "loss": 1.083,
+      "step": 12411
+    },
+    {
+      "epoch": 0.7541162889604471,
+      "grad_norm": 0.20447130501270294,
+      "learning_rate": 1.4279357796514759e-05,
+      "loss": 1.0735,
+      "step": 12412
+    },
+    {
+      "epoch": 0.7541770459930737,
+      "grad_norm": 0.12645472586154938,
+      "learning_rate": 1.4272660148005596e-05,
+      "loss": 1.0476,
+      "step": 12413
+    },
+    {
+      "epoch": 0.7542378030257002,
+      "grad_norm": 0.09878091514110565,
+      "learning_rate": 1.4265963809088977e-05,
+      "loss": 1.0258,
+      "step": 12414
+    },
+    {
+      "epoch": 0.7542985600583267,
+      "grad_norm": 0.21717269718647003,
+      "learning_rate": 1.4259268780010371e-05,
+      "loss": 1.0834,
+      "step": 12415
+    },
+    {
+      "epoch": 0.7543593170909533,
+      "grad_norm": 0.09347713738679886,
+      "learning_rate": 1.4252575061015166e-05,
+      "loss": 1.0204,
+      "step": 12416
+    },
+    {
+      "epoch": 0.7544200741235798,
+      "grad_norm": 0.21152333915233612,
+      "learning_rate": 1.4245882652348724e-05,
+      "loss": 1.1361,
+      "step": 12417
+    },
+    {
+      "epoch": 0.7544808311562063,
+      "grad_norm": 0.18669576942920685,
+      "learning_rate": 1.4239191554256376e-05,
+      "loss": 1.0662,
+      "step": 12418
+    },
+    {
+      "epoch": 0.7545415881888329,
+      "grad_norm": 0.11657535284757614,
+      "learning_rate": 1.4232501766983391e-05,
+      "loss": 1.0741,
+      "step": 12419
+    },
+    {
+      "epoch": 0.7546023452214594,
+      "grad_norm": 0.10764314979314804,
+      "learning_rate": 1.4225813290774948e-05,
+      "loss": 1.0436,
+      "step": 12420
+    },
+    {
+      "epoch": 0.754663102254086,
+      "grad_norm": 0.21123827993869781,
+      "learning_rate": 1.4219126125876237e-05,
+      "loss": 1.1184,
+      "step": 12421
+    },
+    {
+      "epoch": 0.7547238592867125,
+      "grad_norm": 0.19128882884979248,
+      "learning_rate": 1.4212440272532362e-05,
+      "loss": 1.0029,
+      "step": 12422
+    },
+    {
+      "epoch": 0.754784616319339,
+      "grad_norm": 0.1671556532382965,
+      "learning_rate": 1.4205755730988418e-05,
+      "loss": 1.0467,
+      "step": 12423
+    },
+    {
+      "epoch": 0.7548453733519654,
+      "grad_norm": 0.22247761487960815,
+      "learning_rate": 1.4199072501489418e-05,
+      "loss": 1.176,
+      "step": 12424
+    },
+    {
+      "epoch": 0.754906130384592,
+      "grad_norm": 0.12096160650253296,
+      "learning_rate": 1.4192390584280346e-05,
+      "loss": 1.0593,
+      "step": 12425
+    },
+    {
+      "epoch": 0.7549668874172185,
+      "grad_norm": 0.17339357733726501,
+      "learning_rate": 1.4185709979606094e-05,
+      "loss": 0.9622,
+      "step": 12426
+    },
+    {
+      "epoch": 0.755027644449845,
+      "grad_norm": 0.1513710767030716,
+      "learning_rate": 1.417903068771158e-05,
+      "loss": 1.0831,
+      "step": 12427
+    },
+    {
+      "epoch": 0.7550884014824716,
+      "grad_norm": 0.3548330068588257,
+      "learning_rate": 1.4172352708841618e-05,
+      "loss": 1.1623,
+      "step": 12428
+    },
+    {
+      "epoch": 0.7551491585150981,
+      "grad_norm": 0.1272754818201065,
+      "learning_rate": 1.4165676043240995e-05,
+      "loss": 1.0407,
+      "step": 12429
+    },
+    {
+      "epoch": 0.7552099155477247,
+      "grad_norm": 0.1449151635169983,
+      "learning_rate": 1.4159000691154439e-05,
+      "loss": 1.0723,
+      "step": 12430
+    },
+    {
+      "epoch": 0.7552706725803512,
+      "grad_norm": 0.11652117967605591,
+      "learning_rate": 1.4152326652826647e-05,
+      "loss": 1.0292,
+      "step": 12431
+    },
+    {
+      "epoch": 0.7553314296129777,
+      "grad_norm": 0.14926502108573914,
+      "learning_rate": 1.4145653928502245e-05,
+      "loss": 1.0288,
+      "step": 12432
+    },
+    {
+      "epoch": 0.7553921866456043,
+      "grad_norm": 0.11315730214118958,
+      "learning_rate": 1.4138982518425831e-05,
+      "loss": 1.0105,
+      "step": 12433
+    },
+    {
+      "epoch": 0.7554529436782308,
+      "grad_norm": 0.1694616824388504,
+      "learning_rate": 1.4132312422841947e-05,
+      "loss": 1.1487,
+      "step": 12434
+    },
+    {
+      "epoch": 0.7555137007108573,
+      "grad_norm": 0.14117653667926788,
+      "learning_rate": 1.4125643641995078e-05,
+      "loss": 1.1222,
+      "step": 12435
+    },
+    {
+      "epoch": 0.7555744577434839,
+      "grad_norm": 0.11897718161344528,
+      "learning_rate": 1.4118976176129684e-05,
+      "loss": 1.0341,
+      "step": 12436
+    },
+    {
+      "epoch": 0.7556352147761103,
+      "grad_norm": 0.12595760822296143,
+      "learning_rate": 1.411231002549015e-05,
+      "loss": 1.0318,
+      "step": 12437
+    },
+    {
+      "epoch": 0.7556959718087368,
+      "grad_norm": 0.11892735213041306,
+      "learning_rate": 1.410564519032081e-05,
+      "loss": 1.0856,
+      "step": 12438
+    },
+    {
+      "epoch": 0.7557567288413634,
+      "grad_norm": 0.12020894885063171,
+      "learning_rate": 1.4098981670866018e-05,
+      "loss": 1.0343,
+      "step": 12439
+    },
+    {
+      "epoch": 0.7558174858739899,
+      "grad_norm": 0.1720348298549652,
+      "learning_rate": 1.4092319467369974e-05,
+      "loss": 1.1239,
+      "step": 12440
+    },
+    {
+      "epoch": 0.7558782429066164,
+      "grad_norm": 0.21336239576339722,
+      "learning_rate": 1.4085658580076905e-05,
+      "loss": 1.1394,
+      "step": 12441
+    },
+    {
+      "epoch": 0.755938999939243,
+      "grad_norm": 0.28349900245666504,
+      "learning_rate": 1.4078999009230947e-05,
+      "loss": 1.0413,
+      "step": 12442
+    },
+    {
+      "epoch": 0.7559997569718695,
+      "grad_norm": 0.15616722404956818,
+      "learning_rate": 1.4072340755076235e-05,
+      "loss": 1.0589,
+      "step": 12443
+    },
+    {
+      "epoch": 0.756060514004496,
+      "grad_norm": 0.15842755138874054,
+      "learning_rate": 1.4065683817856823e-05,
+      "loss": 1.1563,
+      "step": 12444
+    },
+    {
+      "epoch": 0.7561212710371226,
+      "grad_norm": 0.14412358403205872,
+      "learning_rate": 1.4059028197816725e-05,
+      "loss": 0.993,
+      "step": 12445
+    },
+    {
+      "epoch": 0.7561820280697491,
+      "grad_norm": 1.4671894311904907,
+      "learning_rate": 1.4052373895199888e-05,
+      "loss": 1.2388,
+      "step": 12446
+    },
+    {
+      "epoch": 0.7562427851023756,
+      "grad_norm": 0.11971438676118851,
+      "learning_rate": 1.4045720910250216e-05,
+      "loss": 1.0773,
+      "step": 12447
+    },
+    {
+      "epoch": 0.7563035421350022,
+      "grad_norm": 0.14378733932971954,
+      "learning_rate": 1.4039069243211605e-05,
+      "loss": 1.0301,
+      "step": 12448
+    },
+    {
+      "epoch": 0.7563642991676287,
+      "grad_norm": 0.3817189335823059,
+      "learning_rate": 1.403241889432787e-05,
+      "loss": 1.1437,
+      "step": 12449
+    },
+    {
+      "epoch": 0.7564250562002551,
+      "grad_norm": 3.130676507949829,
+      "learning_rate": 1.4025769863842764e-05,
+      "loss": 1.0221,
+      "step": 12450
+    },
+    {
+      "epoch": 0.7564858132328817,
+      "grad_norm": 0.19876748323440552,
+      "learning_rate": 1.4019122152000025e-05,
+      "loss": 1.0384,
+      "step": 12451
+    },
+    {
+      "epoch": 0.7565465702655082,
+      "grad_norm": 0.20972973108291626,
+      "learning_rate": 1.4012475759043314e-05,
+      "loss": 1.139,
+      "step": 12452
+    },
+    {
+      "epoch": 0.7566073272981347,
+      "grad_norm": 0.11403432488441467,
+      "learning_rate": 1.4005830685216259e-05,
+      "loss": 1.0368,
+      "step": 12453
+    },
+    {
+      "epoch": 0.7566680843307613,
+      "grad_norm": 0.13528011739253998,
+      "learning_rate": 1.3999186930762442e-05,
+      "loss": 1.0917,
+      "step": 12454
+    },
+    {
+      "epoch": 0.7567288413633878,
+      "grad_norm": 0.15780939161777496,
+      "learning_rate": 1.3992544495925386e-05,
+      "loss": 1.043,
+      "step": 12455
+    },
+    {
+      "epoch": 0.7567895983960143,
+      "grad_norm": 0.12039822340011597,
+      "learning_rate": 1.3985903380948573e-05,
+      "loss": 1.05,
+      "step": 12456
+    },
+    {
+      "epoch": 0.7568503554286409,
+      "grad_norm": 0.10794486850500107,
+      "learning_rate": 1.3979263586075426e-05,
+      "loss": 1.0666,
+      "step": 12457
+    },
+    {
+      "epoch": 0.7569111124612674,
+      "grad_norm": 0.1303398609161377,
+      "learning_rate": 1.3972625111549325e-05,
+      "loss": 1.0089,
+      "step": 12458
+    },
+    {
+      "epoch": 0.756971869493894,
+      "grad_norm": 0.12871064245700836,
+      "learning_rate": 1.3965987957613647e-05,
+      "loss": 1.0701,
+      "step": 12459
+    },
+    {
+      "epoch": 0.7570326265265205,
+      "grad_norm": 0.17288193106651306,
+      "learning_rate": 1.3959352124511627e-05,
+      "loss": 1.096,
+      "step": 12460
+    },
+    {
+      "epoch": 0.757093383559147,
+      "grad_norm": 0.13613943755626678,
+      "learning_rate": 1.3952717612486522e-05,
+      "loss": 1.0485,
+      "step": 12461
+    },
+    {
+      "epoch": 0.7571541405917736,
+      "grad_norm": 0.15245550870895386,
+      "learning_rate": 1.394608442178152e-05,
+      "loss": 1.0997,
+      "step": 12462
+    },
+    {
+      "epoch": 0.7572148976244,
+      "grad_norm": 0.1725936233997345,
+      "learning_rate": 1.3939452552639748e-05,
+      "loss": 1.0967,
+      "step": 12463
+    },
+    {
+      "epoch": 0.7572756546570265,
+      "grad_norm": 0.1804647296667099,
+      "learning_rate": 1.3932822005304325e-05,
+      "loss": 1.0522,
+      "step": 12464
+    },
+    {
+      "epoch": 0.757336411689653,
+      "grad_norm": 3.0747997760772705,
+      "learning_rate": 1.3926192780018305e-05,
+      "loss": 1.1279,
+      "step": 12465
+    },
+    {
+      "epoch": 0.7573971687222796,
+      "grad_norm": 0.13486158847808838,
+      "learning_rate": 1.3919564877024644e-05,
+      "loss": 1.0703,
+      "step": 12466
+    },
+    {
+      "epoch": 0.7574579257549061,
+      "grad_norm": 0.15360982716083527,
+      "learning_rate": 1.3912938296566302e-05,
+      "loss": 1.0675,
+      "step": 12467
+    },
+    {
+      "epoch": 0.7575186827875326,
+      "grad_norm": 0.17876802384853363,
+      "learning_rate": 1.3906313038886171e-05,
+      "loss": 1.1884,
+      "step": 12468
+    },
+    {
+      "epoch": 0.7575794398201592,
+      "grad_norm": 0.16275517642498016,
+      "learning_rate": 1.3899689104227126e-05,
+      "loss": 1.0359,
+      "step": 12469
+    },
+    {
+      "epoch": 0.7576401968527857,
+      "grad_norm": 0.120885469019413,
+      "learning_rate": 1.3893066492831957e-05,
+      "loss": 1.0314,
+      "step": 12470
+    },
+    {
+      "epoch": 0.7577009538854123,
+      "grad_norm": 0.27184921503067017,
+      "learning_rate": 1.3886445204943416e-05,
+      "loss": 1.0179,
+      "step": 12471
+    },
+    {
+      "epoch": 0.7577617109180388,
+      "grad_norm": 0.141347736120224,
+      "learning_rate": 1.3879825240804217e-05,
+      "loss": 1.0039,
+      "step": 12472
+    },
+    {
+      "epoch": 0.7578224679506653,
+      "grad_norm": 0.13252796232700348,
+      "learning_rate": 1.387320660065698e-05,
+      "loss": 1.0423,
+      "step": 12473
+    },
+    {
+      "epoch": 0.7578832249832919,
+      "grad_norm": 0.18378327786922455,
+      "learning_rate": 1.3866589284744352e-05,
+      "loss": 1.0412,
+      "step": 12474
+    },
+    {
+      "epoch": 0.7579439820159184,
+      "grad_norm": 0.10154604911804199,
+      "learning_rate": 1.3859973293308876e-05,
+      "loss": 1.0551,
+      "step": 12475
+    },
+    {
+      "epoch": 0.7580047390485448,
+      "grad_norm": 0.28289902210235596,
+      "learning_rate": 1.3853358626593065e-05,
+      "loss": 1.0512,
+      "step": 12476
+    },
+    {
+      "epoch": 0.7580654960811714,
+      "grad_norm": 0.9105175137519836,
+      "learning_rate": 1.3846745284839385e-05,
+      "loss": 1.0816,
+      "step": 12477
+    },
+    {
+      "epoch": 0.7581262531137979,
+      "grad_norm": 0.10451336205005646,
+      "learning_rate": 1.3840133268290239e-05,
+      "loss": 0.99,
+      "step": 12478
+    },
+    {
+      "epoch": 0.7581870101464244,
+      "grad_norm": 0.2284788191318512,
+      "learning_rate": 1.3833522577188002e-05,
+      "loss": 1.2458,
+      "step": 12479
+    },
+    {
+      "epoch": 0.758247767179051,
+      "grad_norm": 0.15661358833312988,
+      "learning_rate": 1.3826913211774983e-05,
+      "loss": 1.1283,
+      "step": 12480
+    },
+    {
+      "epoch": 0.7583085242116775,
+      "grad_norm": 0.15281632542610168,
+      "learning_rate": 1.3820305172293451e-05,
+      "loss": 1.0337,
+      "step": 12481
+    },
+    {
+      "epoch": 0.758369281244304,
+      "grad_norm": 0.13122248649597168,
+      "learning_rate": 1.381369845898563e-05,
+      "loss": 1.074,
+      "step": 12482
+    },
+    {
+      "epoch": 0.7584300382769306,
+      "grad_norm": 0.10453058034181595,
+      "learning_rate": 1.3807093072093686e-05,
+      "loss": 1.0441,
+      "step": 12483
+    },
+    {
+      "epoch": 0.7584907953095571,
+      "grad_norm": 0.1655968874692917,
+      "learning_rate": 1.3800489011859724e-05,
+      "loss": 1.1208,
+      "step": 12484
+    },
+    {
+      "epoch": 0.7585515523421836,
+      "grad_norm": 0.1780315637588501,
+      "learning_rate": 1.3793886278525853e-05,
+      "loss": 1.0966,
+      "step": 12485
+    },
+    {
+      "epoch": 0.7586123093748102,
+      "grad_norm": 0.37785041332244873,
+      "learning_rate": 1.3787284872334095e-05,
+      "loss": 1.0914,
+      "step": 12486
+    },
+    {
+      "epoch": 0.7586730664074367,
+      "grad_norm": 0.15251842141151428,
+      "learning_rate": 1.3780684793526389e-05,
+      "loss": 1.1238,
+      "step": 12487
+    },
+    {
+      "epoch": 0.7587338234400632,
+      "grad_norm": 0.2933756113052368,
+      "learning_rate": 1.3774086042344686e-05,
+      "loss": 1.1127,
+      "step": 12488
+    },
+    {
+      "epoch": 0.7587945804726897,
+      "grad_norm": 0.11405391246080399,
+      "learning_rate": 1.3767488619030843e-05,
+      "loss": 1.0085,
+      "step": 12489
+    },
+    {
+      "epoch": 0.7588553375053162,
+      "grad_norm": 0.37702250480651855,
+      "learning_rate": 1.3760892523826729e-05,
+      "loss": 1.2371,
+      "step": 12490
+    },
+    {
+      "epoch": 0.7589160945379427,
+      "grad_norm": 0.14275868237018585,
+      "learning_rate": 1.3754297756974099e-05,
+      "loss": 1.1094,
+      "step": 12491
+    },
+    {
+      "epoch": 0.7589768515705693,
+      "grad_norm": 0.15083137154579163,
+      "learning_rate": 1.374770431871471e-05,
+      "loss": 1.1236,
+      "step": 12492
+    },
+    {
+      "epoch": 0.7590376086031958,
+      "grad_norm": 0.23004627227783203,
+      "learning_rate": 1.3741112209290214e-05,
+      "loss": 1.0212,
+      "step": 12493
+    },
+    {
+      "epoch": 0.7590983656358223,
+      "grad_norm": 0.1742396503686905,
+      "learning_rate": 1.3734521428942238e-05,
+      "loss": 1.0842,
+      "step": 12494
+    },
+    {
+      "epoch": 0.7591591226684489,
+      "grad_norm": 0.10751919448375702,
+      "learning_rate": 1.3727931977912407e-05,
+      "loss": 1.0743,
+      "step": 12495
+    },
+    {
+      "epoch": 0.7592198797010754,
+      "grad_norm": 0.15435214340686798,
+      "learning_rate": 1.3721343856442238e-05,
+      "loss": 1.084,
+      "step": 12496
+    },
+    {
+      "epoch": 0.7592806367337019,
+      "grad_norm": 0.20320147275924683,
+      "learning_rate": 1.3714757064773226e-05,
+      "loss": 1.0652,
+      "step": 12497
+    },
+    {
+      "epoch": 0.7593413937663285,
+      "grad_norm": 0.3919805586338043,
+      "learning_rate": 1.3708171603146803e-05,
+      "loss": 1.1089,
+      "step": 12498
+    },
+    {
+      "epoch": 0.759402150798955,
+      "grad_norm": 0.2511536478996277,
+      "learning_rate": 1.3701587471804367e-05,
+      "loss": 0.9924,
+      "step": 12499
+    },
+    {
+      "epoch": 0.7594629078315815,
+      "grad_norm": 0.1596825122833252,
+      "learning_rate": 1.369500467098726e-05,
+      "loss": 1.0995,
+      "step": 12500
+    },
+    {
+      "epoch": 0.7595236648642081,
+      "grad_norm": 0.42139509320259094,
+      "learning_rate": 1.3688423200936773e-05,
+      "loss": 1.1318,
+      "step": 12501
+    },
+    {
+      "epoch": 0.7595844218968345,
+      "grad_norm": 0.24840416014194489,
+      "learning_rate": 1.368184306189415e-05,
+      "loss": 1.083,
+      "step": 12502
+    },
+    {
+      "epoch": 0.759645178929461,
+      "grad_norm": 0.22676849365234375,
+      "learning_rate": 1.3675264254100595e-05,
+      "loss": 1.1176,
+      "step": 12503
+    },
+    {
+      "epoch": 0.7597059359620876,
+      "grad_norm": 0.1771671622991562,
+      "learning_rate": 1.3668686777797247e-05,
+      "loss": 1.0393,
+      "step": 12504
+    },
+    {
+      "epoch": 0.7597666929947141,
+      "grad_norm": 0.1207723468542099,
+      "learning_rate": 1.3662110633225196e-05,
+      "loss": 1.0225,
+      "step": 12505
+    },
+    {
+      "epoch": 0.7598274500273406,
+      "grad_norm": 0.1388079673051834,
+      "learning_rate": 1.3655535820625532e-05,
+      "loss": 1.0324,
+      "step": 12506
+    },
+    {
+      "epoch": 0.7598882070599672,
+      "grad_norm": 0.3035030961036682,
+      "learning_rate": 1.3648962340239208e-05,
+      "loss": 1.0867,
+      "step": 12507
+    },
+    {
+      "epoch": 0.7599489640925937,
+      "grad_norm": 0.1515635848045349,
+      "learning_rate": 1.3642390192307204e-05,
+      "loss": 1.1514,
+      "step": 12508
+    },
+    {
+      "epoch": 0.7600097211252203,
+      "grad_norm": 0.11775967478752136,
+      "learning_rate": 1.3635819377070407e-05,
+      "loss": 1.011,
+      "step": 12509
+    },
+    {
+      "epoch": 0.7600704781578468,
+      "grad_norm": 0.14208143949508667,
+      "learning_rate": 1.3629249894769668e-05,
+      "loss": 1.0877,
+      "step": 12510
+    },
+    {
+      "epoch": 0.7601312351904733,
+      "grad_norm": 0.14568772912025452,
+      "learning_rate": 1.3622681745645821e-05,
+      "loss": 1.0621,
+      "step": 12511
+    },
+    {
+      "epoch": 0.7601919922230999,
+      "grad_norm": 0.1941111981868744,
+      "learning_rate": 1.3616114929939622e-05,
+      "loss": 1.0849,
+      "step": 12512
+    },
+    {
+      "epoch": 0.7602527492557264,
+      "grad_norm": 0.16677498817443848,
+      "learning_rate": 1.3609549447891745e-05,
+      "loss": 1.1111,
+      "step": 12513
+    },
+    {
+      "epoch": 0.7603135062883529,
+      "grad_norm": 0.1665506213903427,
+      "learning_rate": 1.360298529974287e-05,
+      "loss": 1.0657,
+      "step": 12514
+    },
+    {
+      "epoch": 0.7603742633209793,
+      "grad_norm": 0.1916867196559906,
+      "learning_rate": 1.3596422485733589e-05,
+      "loss": 1.1055,
+      "step": 12515
+    },
+    {
+      "epoch": 0.7604350203536059,
+      "grad_norm": 0.1459992378950119,
+      "learning_rate": 1.3589861006104494e-05,
+      "loss": 1.0527,
+      "step": 12516
+    },
+    {
+      "epoch": 0.7604957773862324,
+      "grad_norm": 0.12731795012950897,
+      "learning_rate": 1.3583300861096077e-05,
+      "loss": 1.0137,
+      "step": 12517
+    },
+    {
+      "epoch": 0.760556534418859,
+      "grad_norm": 0.19834287464618683,
+      "learning_rate": 1.357674205094881e-05,
+      "loss": 1.1117,
+      "step": 12518
+    },
+    {
+      "epoch": 0.7606172914514855,
+      "grad_norm": 0.21660210192203522,
+      "learning_rate": 1.3570184575903105e-05,
+      "loss": 1.0666,
+      "step": 12519
+    },
+    {
+      "epoch": 0.760678048484112,
+      "grad_norm": 0.17140260338783264,
+      "learning_rate": 1.3563628436199322e-05,
+      "loss": 1.1116,
+      "step": 12520
+    },
+    {
+      "epoch": 0.7607388055167386,
+      "grad_norm": 0.5122325420379639,
+      "learning_rate": 1.3557073632077783e-05,
+      "loss": 1.0389,
+      "step": 12521
+    },
+    {
+      "epoch": 0.7607995625493651,
+      "grad_norm": 0.1227804496884346,
+      "learning_rate": 1.3550520163778757e-05,
+      "loss": 0.9998,
+      "step": 12522
+    },
+    {
+      "epoch": 0.7608603195819916,
+      "grad_norm": 0.12985700368881226,
+      "learning_rate": 1.3543968031542459e-05,
+      "loss": 1.0405,
+      "step": 12523
+    },
+    {
+      "epoch": 0.7609210766146182,
+      "grad_norm": 0.1181931421160698,
+      "learning_rate": 1.3537417235609057e-05,
+      "loss": 1.0142,
+      "step": 12524
+    },
+    {
+      "epoch": 0.7609818336472447,
+      "grad_norm": 0.12929067015647888,
+      "learning_rate": 1.3530867776218675e-05,
+      "loss": 1.0285,
+      "step": 12525
+    },
+    {
+      "epoch": 0.7610425906798712,
+      "grad_norm": 0.14773963391780853,
+      "learning_rate": 1.3524319653611384e-05,
+      "loss": 1.09,
+      "step": 12526
+    },
+    {
+      "epoch": 0.7611033477124978,
+      "grad_norm": 0.11514797061681747,
+      "learning_rate": 1.3517772868027206e-05,
+      "loss": 1.0791,
+      "step": 12527
+    },
+    {
+      "epoch": 0.7611641047451243,
+      "grad_norm": 0.10016696900129318,
+      "learning_rate": 1.3511227419706119e-05,
+      "loss": 1.0048,
+      "step": 12528
+    },
+    {
+      "epoch": 0.7612248617777507,
+      "grad_norm": 0.1402485966682434,
+      "learning_rate": 1.3504683308888043e-05,
+      "loss": 1.0774,
+      "step": 12529
+    },
+    {
+      "epoch": 0.7612856188103773,
+      "grad_norm": 0.14354290068149567,
+      "learning_rate": 1.3498140535812852e-05,
+      "loss": 1.0617,
+      "step": 12530
+    },
+    {
+      "epoch": 0.7613463758430038,
+      "grad_norm": 0.25909605622291565,
+      "learning_rate": 1.3491599100720365e-05,
+      "loss": 1.14,
+      "step": 12531
+    },
+    {
+      "epoch": 0.7614071328756303,
+      "grad_norm": 3.9069385528564453,
+      "learning_rate": 1.3485059003850392e-05,
+      "loss": 1.0566,
+      "step": 12532
+    },
+    {
+      "epoch": 0.7614678899082569,
+      "grad_norm": 0.22198474407196045,
+      "learning_rate": 1.3478520245442633e-05,
+      "loss": 1.0977,
+      "step": 12533
+    },
+    {
+      "epoch": 0.7615286469408834,
+      "grad_norm": 0.09995206445455551,
+      "learning_rate": 1.3471982825736767e-05,
+      "loss": 1.0127,
+      "step": 12534
+    },
+    {
+      "epoch": 0.7615894039735099,
+      "grad_norm": 0.2163786143064499,
+      "learning_rate": 1.346544674497242e-05,
+      "loss": 1.128,
+      "step": 12535
+    },
+    {
+      "epoch": 0.7616501610061365,
+      "grad_norm": 0.16006778180599213,
+      "learning_rate": 1.3458912003389201e-05,
+      "loss": 1.0914,
+      "step": 12536
+    },
+    {
+      "epoch": 0.761710918038763,
+      "grad_norm": 0.4079490005970001,
+      "learning_rate": 1.3452378601226624e-05,
+      "loss": 1.116,
+      "step": 12537
+    },
+    {
+      "epoch": 0.7617716750713895,
+      "grad_norm": 0.2563630938529968,
+      "learning_rate": 1.3445846538724171e-05,
+      "loss": 1.1518,
+      "step": 12538
+    },
+    {
+      "epoch": 0.7618324321040161,
+      "grad_norm": 0.11792192608118057,
+      "learning_rate": 1.3439315816121294e-05,
+      "loss": 1.0224,
+      "step": 12539
+    },
+    {
+      "epoch": 0.7618931891366426,
+      "grad_norm": 0.2553277611732483,
+      "learning_rate": 1.3432786433657335e-05,
+      "loss": 1.2811,
+      "step": 12540
+    },
+    {
+      "epoch": 0.7619539461692691,
+      "grad_norm": 0.14214277267456055,
+      "learning_rate": 1.3426258391571677e-05,
+      "loss": 1.0724,
+      "step": 12541
+    },
+    {
+      "epoch": 0.7620147032018956,
+      "grad_norm": 0.33141443133354187,
+      "learning_rate": 1.3419731690103581e-05,
+      "loss": 1.0426,
+      "step": 12542
+    },
+    {
+      "epoch": 0.7620754602345221,
+      "grad_norm": 0.38680481910705566,
+      "learning_rate": 1.3413206329492295e-05,
+      "loss": 1.1436,
+      "step": 12543
+    },
+    {
+      "epoch": 0.7621362172671486,
+      "grad_norm": 0.2522801160812378,
+      "learning_rate": 1.3406682309976998e-05,
+      "loss": 1.0829,
+      "step": 12544
+    },
+    {
+      "epoch": 0.7621969742997752,
+      "grad_norm": 0.16162461042404175,
+      "learning_rate": 1.3400159631796838e-05,
+      "loss": 1.0985,
+      "step": 12545
+    },
+    {
+      "epoch": 0.7622577313324017,
+      "grad_norm": 0.13372021913528442,
+      "learning_rate": 1.3393638295190896e-05,
+      "loss": 1.0622,
+      "step": 12546
+    },
+    {
+      "epoch": 0.7623184883650282,
+      "grad_norm": 0.1485844850540161,
+      "learning_rate": 1.3387118300398227e-05,
+      "loss": 1.1156,
+      "step": 12547
+    },
+    {
+      "epoch": 0.7623792453976548,
+      "grad_norm": 0.1145109161734581,
+      "learning_rate": 1.338059964765781e-05,
+      "loss": 1.0838,
+      "step": 12548
+    },
+    {
+      "epoch": 0.7624400024302813,
+      "grad_norm": 0.1353783905506134,
+      "learning_rate": 1.3374082337208587e-05,
+      "loss": 1.008,
+      "step": 12549
+    },
+    {
+      "epoch": 0.7625007594629079,
+      "grad_norm": 0.17304907739162445,
+      "learning_rate": 1.3367566369289457e-05,
+      "loss": 1.1866,
+      "step": 12550
+    },
+    {
+      "epoch": 0.7625615164955344,
+      "grad_norm": 0.13886840641498566,
+      "learning_rate": 1.3361051744139253e-05,
+      "loss": 1.0019,
+      "step": 12551
+    },
+    {
+      "epoch": 0.7626222735281609,
+      "grad_norm": 0.18054184317588806,
+      "learning_rate": 1.3354538461996807e-05,
+      "loss": 1.131,
+      "step": 12552
+    },
+    {
+      "epoch": 0.7626830305607875,
+      "grad_norm": 1.101049542427063,
+      "learning_rate": 1.334802652310082e-05,
+      "loss": 1.0338,
+      "step": 12553
+    },
+    {
+      "epoch": 0.762743787593414,
+      "grad_norm": 0.1377827525138855,
+      "learning_rate": 1.3341515927690007e-05,
+      "loss": 1.0876,
+      "step": 12554
+    },
+    {
+      "epoch": 0.7628045446260404,
+      "grad_norm": 0.1813245266675949,
+      "learning_rate": 1.3335006676003014e-05,
+      "loss": 1.1161,
+      "step": 12555
+    },
+    {
+      "epoch": 0.762865301658667,
+      "grad_norm": 0.14815883338451385,
+      "learning_rate": 1.332849876827842e-05,
+      "loss": 1.0937,
+      "step": 12556
+    },
+    {
+      "epoch": 0.7629260586912935,
+      "grad_norm": 0.5480836033821106,
+      "learning_rate": 1.3321992204754813e-05,
+      "loss": 1.151,
+      "step": 12557
+    },
+    {
+      "epoch": 0.76298681572392,
+      "grad_norm": 0.3353072702884674,
+      "learning_rate": 1.3315486985670667e-05,
+      "loss": 1.0173,
+      "step": 12558
+    },
+    {
+      "epoch": 0.7630475727565466,
+      "grad_norm": 0.21494805812835693,
+      "learning_rate": 1.330898311126445e-05,
+      "loss": 1.1367,
+      "step": 12559
+    },
+    {
+      "epoch": 0.7631083297891731,
+      "grad_norm": 0.10747019946575165,
+      "learning_rate": 1.330248058177454e-05,
+      "loss": 1.0232,
+      "step": 12560
+    },
+    {
+      "epoch": 0.7631690868217996,
+      "grad_norm": 0.1298971027135849,
+      "learning_rate": 1.3295979397439279e-05,
+      "loss": 1.0208,
+      "step": 12561
+    },
+    {
+      "epoch": 0.7632298438544262,
+      "grad_norm": 0.14917881786823273,
+      "learning_rate": 1.3289479558497003e-05,
+      "loss": 1.0154,
+      "step": 12562
+    },
+    {
+      "epoch": 0.7632906008870527,
+      "grad_norm": 0.14374469220638275,
+      "learning_rate": 1.328298106518595e-05,
+      "loss": 1.0469,
+      "step": 12563
+    },
+    {
+      "epoch": 0.7633513579196792,
+      "grad_norm": 0.1334616094827652,
+      "learning_rate": 1.3276483917744326e-05,
+      "loss": 1.0582,
+      "step": 12564
+    },
+    {
+      "epoch": 0.7634121149523058,
+      "grad_norm": 0.15402697026729584,
+      "learning_rate": 1.3269988116410276e-05,
+      "loss": 1.0611,
+      "step": 12565
+    },
+    {
+      "epoch": 0.7634728719849323,
+      "grad_norm": 0.8438519835472107,
+      "learning_rate": 1.3263493661421916e-05,
+      "loss": 1.0733,
+      "step": 12566
+    },
+    {
+      "epoch": 0.7635336290175588,
+      "grad_norm": 0.12258189916610718,
+      "learning_rate": 1.3257000553017296e-05,
+      "loss": 1.0386,
+      "step": 12567
+    },
+    {
+      "epoch": 0.7635943860501853,
+      "grad_norm": 0.13384394347667694,
+      "learning_rate": 1.3250508791434424e-05,
+      "loss": 1.0002,
+      "step": 12568
+    },
+    {
+      "epoch": 0.7636551430828118,
+      "grad_norm": 0.11661837995052338,
+      "learning_rate": 1.3244018376911255e-05,
+      "loss": 1.0488,
+      "step": 12569
+    },
+    {
+      "epoch": 0.7637159001154383,
+      "grad_norm": 0.10678543895483017,
+      "learning_rate": 1.3237529309685697e-05,
+      "loss": 1.0065,
+      "step": 12570
+    },
+    {
+      "epoch": 0.7637766571480649,
+      "grad_norm": 0.10978089272975922,
+      "learning_rate": 1.3231041589995607e-05,
+      "loss": 1.0077,
+      "step": 12571
+    },
+    {
+      "epoch": 0.7638374141806914,
+      "grad_norm": 0.14187034964561462,
+      "learning_rate": 1.3224555218078782e-05,
+      "loss": 1.0413,
+      "step": 12572
+    },
+    {
+      "epoch": 0.7638981712133179,
+      "grad_norm": 0.12638205289840698,
+      "learning_rate": 1.3218070194173027e-05,
+      "loss": 1.0565,
+      "step": 12573
+    },
+    {
+      "epoch": 0.7639589282459445,
+      "grad_norm": 0.1691051572561264,
+      "learning_rate": 1.3211586518516e-05,
+      "loss": 1.1334,
+      "step": 12574
+    },
+    {
+      "epoch": 0.764019685278571,
+      "grad_norm": 0.15638159215450287,
+      "learning_rate": 1.320510419134538e-05,
+      "loss": 1.1161,
+      "step": 12575
+    },
+    {
+      "epoch": 0.7640804423111975,
+      "grad_norm": 0.15488126873970032,
+      "learning_rate": 1.3198623212898776e-05,
+      "loss": 1.0244,
+      "step": 12576
+    },
+    {
+      "epoch": 0.7641411993438241,
+      "grad_norm": 0.17011305689811707,
+      "learning_rate": 1.319214358341374e-05,
+      "loss": 1.1553,
+      "step": 12577
+    },
+    {
+      "epoch": 0.7642019563764506,
+      "grad_norm": 0.5920494794845581,
+      "learning_rate": 1.3185665303127809e-05,
+      "loss": 1.0001,
+      "step": 12578
+    },
+    {
+      "epoch": 0.7642627134090771,
+      "grad_norm": 0.16712260246276855,
+      "learning_rate": 1.317918837227845e-05,
+      "loss": 1.1527,
+      "step": 12579
+    },
+    {
+      "epoch": 0.7643234704417037,
+      "grad_norm": 0.19923590123653412,
+      "learning_rate": 1.3172712791103043e-05,
+      "loss": 1.1439,
+      "step": 12580
+    },
+    {
+      "epoch": 0.7643842274743301,
+      "grad_norm": 0.13136544823646545,
+      "learning_rate": 1.316623855983896e-05,
+      "loss": 1.0158,
+      "step": 12581
+    },
+    {
+      "epoch": 0.7644449845069566,
+      "grad_norm": 0.13120952248573303,
+      "learning_rate": 1.3159765678723518e-05,
+      "loss": 1.0376,
+      "step": 12582
+    },
+    {
+      "epoch": 0.7645057415395832,
+      "grad_norm": 0.20183120667934418,
+      "learning_rate": 1.3153294147993995e-05,
+      "loss": 1.1735,
+      "step": 12583
+    },
+    {
+      "epoch": 0.7645664985722097,
+      "grad_norm": 0.5586725473403931,
+      "learning_rate": 1.31468239678876e-05,
+      "loss": 1.121,
+      "step": 12584
+    },
+    {
+      "epoch": 0.7646272556048362,
+      "grad_norm": 0.12262866646051407,
+      "learning_rate": 1.3140355138641496e-05,
+      "loss": 1.1113,
+      "step": 12585
+    },
+    {
+      "epoch": 0.7646880126374628,
+      "grad_norm": 0.25252658128738403,
+      "learning_rate": 1.313388766049281e-05,
+      "loss": 1.2143,
+      "step": 12586
+    },
+    {
+      "epoch": 0.7647487696700893,
+      "grad_norm": 0.29822754859924316,
+      "learning_rate": 1.3127421533678575e-05,
+      "loss": 1.2904,
+      "step": 12587
+    },
+    {
+      "epoch": 0.7648095267027158,
+      "grad_norm": 0.11995142698287964,
+      "learning_rate": 1.3120956758435837e-05,
+      "loss": 1.0784,
+      "step": 12588
+    },
+    {
+      "epoch": 0.7648702837353424,
+      "grad_norm": 0.15583375096321106,
+      "learning_rate": 1.3114493335001565e-05,
+      "loss": 1.0207,
+      "step": 12589
+    },
+    {
+      "epoch": 0.7649310407679689,
+      "grad_norm": 0.17384937405586243,
+      "learning_rate": 1.310803126361267e-05,
+      "loss": 1.0822,
+      "step": 12590
+    },
+    {
+      "epoch": 0.7649917978005955,
+      "grad_norm": 0.20129141211509705,
+      "learning_rate": 1.3101570544506014e-05,
+      "loss": 1.1052,
+      "step": 12591
+    },
+    {
+      "epoch": 0.765052554833222,
+      "grad_norm": 0.15278537571430206,
+      "learning_rate": 1.3095111177918423e-05,
+      "loss": 1.0464,
+      "step": 12592
+    },
+    {
+      "epoch": 0.7651133118658485,
+      "grad_norm": 0.1675327718257904,
+      "learning_rate": 1.3088653164086667e-05,
+      "loss": 1.1144,
+      "step": 12593
+    },
+    {
+      "epoch": 0.765174068898475,
+      "grad_norm": 0.56601482629776,
+      "learning_rate": 1.3082196503247469e-05,
+      "loss": 1.0663,
+      "step": 12594
+    },
+    {
+      "epoch": 0.7652348259311015,
+      "grad_norm": 0.14412938058376312,
+      "learning_rate": 1.3075741195637487e-05,
+      "loss": 1.0697,
+      "step": 12595
+    },
+    {
+      "epoch": 0.765295582963728,
+      "grad_norm": 0.1973801851272583,
+      "learning_rate": 1.306928724149335e-05,
+      "loss": 1.1855,
+      "step": 12596
+    },
+    {
+      "epoch": 0.7653563399963546,
+      "grad_norm": 0.1966913342475891,
+      "learning_rate": 1.306283464105163e-05,
+      "loss": 1.1904,
+      "step": 12597
+    },
+    {
+      "epoch": 0.7654170970289811,
+      "grad_norm": 0.1815931648015976,
+      "learning_rate": 1.3056383394548832e-05,
+      "loss": 1.0903,
+      "step": 12598
+    },
+    {
+      "epoch": 0.7654778540616076,
+      "grad_norm": 0.16487367451190948,
+      "learning_rate": 1.3049933502221468e-05,
+      "loss": 1.1026,
+      "step": 12599
+    },
+    {
+      "epoch": 0.7655386110942342,
+      "grad_norm": 0.15805833041667938,
+      "learning_rate": 1.3043484964305924e-05,
+      "loss": 1.0216,
+      "step": 12600
+    },
+    {
+      "epoch": 0.7655993681268607,
+      "grad_norm": 0.10602639615535736,
+      "learning_rate": 1.303703778103858e-05,
+      "loss": 1.0302,
+      "step": 12601
+    },
+    {
+      "epoch": 0.7656601251594872,
+      "grad_norm": 0.1442016214132309,
+      "learning_rate": 1.3030591952655757e-05,
+      "loss": 1.0711,
+      "step": 12602
+    },
+    {
+      "epoch": 0.7657208821921138,
+      "grad_norm": 0.133018359541893,
+      "learning_rate": 1.3024147479393722e-05,
+      "loss": 1.0445,
+      "step": 12603
+    },
+    {
+      "epoch": 0.7657816392247403,
+      "grad_norm": 0.18071715533733368,
+      "learning_rate": 1.301770436148872e-05,
+      "loss": 1.0399,
+      "step": 12604
+    },
+    {
+      "epoch": 0.7658423962573668,
+      "grad_norm": 0.22149837017059326,
+      "learning_rate": 1.3011262599176915e-05,
+      "loss": 1.0678,
+      "step": 12605
+    },
+    {
+      "epoch": 0.7659031532899934,
+      "grad_norm": 0.18498311936855316,
+      "learning_rate": 1.3004822192694444e-05,
+      "loss": 1.2245,
+      "step": 12606
+    },
+    {
+      "epoch": 0.7659639103226198,
+      "grad_norm": 0.12543554604053497,
+      "learning_rate": 1.2998383142277348e-05,
+      "loss": 1.0037,
+      "step": 12607
+    },
+    {
+      "epoch": 0.7660246673552463,
+      "grad_norm": 0.20280525088310242,
+      "learning_rate": 1.2991945448161658e-05,
+      "loss": 1.1559,
+      "step": 12608
+    },
+    {
+      "epoch": 0.7660854243878729,
+      "grad_norm": 0.2181260585784912,
+      "learning_rate": 1.2985509110583378e-05,
+      "loss": 1.139,
+      "step": 12609
+    },
+    {
+      "epoch": 0.7661461814204994,
+      "grad_norm": 0.1568908542394638,
+      "learning_rate": 1.2979074129778419e-05,
+      "loss": 1.1372,
+      "step": 12610
+    },
+    {
+      "epoch": 0.7662069384531259,
+      "grad_norm": 0.23146596550941467,
+      "learning_rate": 1.2972640505982647e-05,
+      "loss": 1.1664,
+      "step": 12611
+    },
+    {
+      "epoch": 0.7662676954857525,
+      "grad_norm": 0.18140651285648346,
+      "learning_rate": 1.2966208239431899e-05,
+      "loss": 0.9685,
+      "step": 12612
+    },
+    {
+      "epoch": 0.766328452518379,
+      "grad_norm": 0.23390540480613708,
+      "learning_rate": 1.2959777330361938e-05,
+      "loss": 1.0568,
+      "step": 12613
+    },
+    {
+      "epoch": 0.7663892095510055,
+      "grad_norm": 0.12980833649635315,
+      "learning_rate": 1.2953347779008508e-05,
+      "loss": 1.0603,
+      "step": 12614
+    },
+    {
+      "epoch": 0.7664499665836321,
+      "grad_norm": 0.17330950498580933,
+      "learning_rate": 1.2946919585607275e-05,
+      "loss": 1.0121,
+      "step": 12615
+    },
+    {
+      "epoch": 0.7665107236162586,
+      "grad_norm": 0.1783524602651596,
+      "learning_rate": 1.2940492750393863e-05,
+      "loss": 1.1702,
+      "step": 12616
+    },
+    {
+      "epoch": 0.7665714806488851,
+      "grad_norm": 0.2138451486825943,
+      "learning_rate": 1.2934067273603856e-05,
+      "loss": 1.0531,
+      "step": 12617
+    },
+    {
+      "epoch": 0.7666322376815117,
+      "grad_norm": 0.11507545411586761,
+      "learning_rate": 1.2927643155472774e-05,
+      "loss": 1.0518,
+      "step": 12618
+    },
+    {
+      "epoch": 0.7666929947141382,
+      "grad_norm": 0.3343908190727234,
+      "learning_rate": 1.29212203962361e-05,
+      "loss": 1.237,
+      "step": 12619
+    },
+    {
+      "epoch": 0.7667537517467646,
+      "grad_norm": 0.14791886508464813,
+      "learning_rate": 1.2914798996129262e-05,
+      "loss": 1.026,
+      "step": 12620
+    },
+    {
+      "epoch": 0.7668145087793912,
+      "grad_norm": 0.178230881690979,
+      "learning_rate": 1.2908378955387634e-05,
+      "loss": 1.1533,
+      "step": 12621
+    },
+    {
+      "epoch": 0.7668752658120177,
+      "grad_norm": 0.13301712274551392,
+      "learning_rate": 1.2901960274246544e-05,
+      "loss": 1.0041,
+      "step": 12622
+    },
+    {
+      "epoch": 0.7669360228446442,
+      "grad_norm": 0.20597591996192932,
+      "learning_rate": 1.289554295294127e-05,
+      "loss": 1.1047,
+      "step": 12623
+    },
+    {
+      "epoch": 0.7669967798772708,
+      "grad_norm": 0.13181617856025696,
+      "learning_rate": 1.288912699170703e-05,
+      "loss": 1.0647,
+      "step": 12624
+    },
+    {
+      "epoch": 0.7670575369098973,
+      "grad_norm": 0.1167430430650711,
+      "learning_rate": 1.2882712390779022e-05,
+      "loss": 1.047,
+      "step": 12625
+    },
+    {
+      "epoch": 0.7671182939425238,
+      "grad_norm": 0.1887039691209793,
+      "learning_rate": 1.287629915039239e-05,
+      "loss": 1.0784,
+      "step": 12626
+    },
+    {
+      "epoch": 0.7671790509751504,
+      "grad_norm": 0.18720541894435883,
+      "learning_rate": 1.2869887270782165e-05,
+      "loss": 1.15,
+      "step": 12627
+    },
+    {
+      "epoch": 0.7672398080077769,
+      "grad_norm": 0.13905468583106995,
+      "learning_rate": 1.286347675218339e-05,
+      "loss": 1.0477,
+      "step": 12628
+    },
+    {
+      "epoch": 0.7673005650404034,
+      "grad_norm": 0.1416422426700592,
+      "learning_rate": 1.2857067594831063e-05,
+      "loss": 0.9789,
+      "step": 12629
+    },
+    {
+      "epoch": 0.76736132207303,
+      "grad_norm": 0.20307201147079468,
+      "learning_rate": 1.2850659798960107e-05,
+      "loss": 1.1788,
+      "step": 12630
+    },
+    {
+      "epoch": 0.7674220791056565,
+      "grad_norm": 0.1398327350616455,
+      "learning_rate": 1.2844253364805398e-05,
+      "loss": 1.0557,
+      "step": 12631
+    },
+    {
+      "epoch": 0.767482836138283,
+      "grad_norm": 0.1993551403284073,
+      "learning_rate": 1.2837848292601755e-05,
+      "loss": 1.1008,
+      "step": 12632
+    },
+    {
+      "epoch": 0.7675435931709096,
+      "grad_norm": 0.13359631597995758,
+      "learning_rate": 1.2831444582583973e-05,
+      "loss": 1.039,
+      "step": 12633
+    },
+    {
+      "epoch": 0.767604350203536,
+      "grad_norm": 0.16576634347438812,
+      "learning_rate": 1.2825042234986779e-05,
+      "loss": 1.067,
+      "step": 12634
+    },
+    {
+      "epoch": 0.7676651072361625,
+      "grad_norm": 0.23622706532478333,
+      "learning_rate": 1.2818641250044838e-05,
+      "loss": 1.1062,
+      "step": 12635
+    },
+    {
+      "epoch": 0.7677258642687891,
+      "grad_norm": 0.18494898080825806,
+      "learning_rate": 1.2812241627992793e-05,
+      "loss": 1.1224,
+      "step": 12636
+    },
+    {
+      "epoch": 0.7677866213014156,
+      "grad_norm": 0.1257193684577942,
+      "learning_rate": 1.2805843369065223e-05,
+      "loss": 1.036,
+      "step": 12637
+    },
+    {
+      "epoch": 0.7678473783340422,
+      "grad_norm": 0.1562509834766388,
+      "learning_rate": 1.2799446473496651e-05,
+      "loss": 1.0862,
+      "step": 12638
+    },
+    {
+      "epoch": 0.7679081353666687,
+      "grad_norm": 0.2027399241924286,
+      "learning_rate": 1.2793050941521556e-05,
+      "loss": 1.1339,
+      "step": 12639
+    },
+    {
+      "epoch": 0.7679688923992952,
+      "grad_norm": 0.15130436420440674,
+      "learning_rate": 1.2786656773374373e-05,
+      "loss": 1.2908,
+      "step": 12640
+    },
+    {
+      "epoch": 0.7680296494319218,
+      "grad_norm": 0.1184222549200058,
+      "learning_rate": 1.278026396928948e-05,
+      "loss": 1.0829,
+      "step": 12641
+    },
+    {
+      "epoch": 0.7680904064645483,
+      "grad_norm": 0.15989989042282104,
+      "learning_rate": 1.2773872529501207e-05,
+      "loss": 1.1005,
+      "step": 12642
+    },
+    {
+      "epoch": 0.7681511634971748,
+      "grad_norm": 0.14799381792545319,
+      "learning_rate": 1.2767482454243829e-05,
+      "loss": 1.0421,
+      "step": 12643
+    },
+    {
+      "epoch": 0.7682119205298014,
+      "grad_norm": 0.10433269292116165,
+      "learning_rate": 1.2761093743751568e-05,
+      "loss": 1.014,
+      "step": 12644
+    },
+    {
+      "epoch": 0.7682726775624279,
+      "grad_norm": 0.19913047552108765,
+      "learning_rate": 1.2754706398258625e-05,
+      "loss": 1.112,
+      "step": 12645
+    },
+    {
+      "epoch": 0.7683334345950544,
+      "grad_norm": 0.11359617114067078,
+      "learning_rate": 1.2748320417999132e-05,
+      "loss": 1.012,
+      "step": 12646
+    },
+    {
+      "epoch": 0.7683941916276809,
+      "grad_norm": 0.1482570320367813,
+      "learning_rate": 1.2741935803207144e-05,
+      "loss": 1.1103,
+      "step": 12647
+    },
+    {
+      "epoch": 0.7684549486603074,
+      "grad_norm": 0.1119273379445076,
+      "learning_rate": 1.2735552554116698e-05,
+      "loss": 1.0069,
+      "step": 12648
+    },
+    {
+      "epoch": 0.7685157056929339,
+      "grad_norm": 0.22167113423347473,
+      "learning_rate": 1.272917067096176e-05,
+      "loss": 1.1505,
+      "step": 12649
+    },
+    {
+      "epoch": 0.7685764627255605,
+      "grad_norm": 0.49580806493759155,
+      "learning_rate": 1.2722790153976294e-05,
+      "loss": 1.0523,
+      "step": 12650
+    },
+    {
+      "epoch": 0.768637219758187,
+      "grad_norm": 0.2841155230998993,
+      "learning_rate": 1.2716411003394157e-05,
+      "loss": 1.1503,
+      "step": 12651
+    },
+    {
+      "epoch": 0.7686979767908135,
+      "grad_norm": 1.1812560558319092,
+      "learning_rate": 1.2710033219449185e-05,
+      "loss": 1.0168,
+      "step": 12652
+    },
+    {
+      "epoch": 0.7687587338234401,
+      "grad_norm": 0.13246190547943115,
+      "learning_rate": 1.2703656802375164e-05,
+      "loss": 1.0813,
+      "step": 12653
+    },
+    {
+      "epoch": 0.7688194908560666,
+      "grad_norm": 0.10599101334810257,
+      "learning_rate": 1.2697281752405787e-05,
+      "loss": 1.0439,
+      "step": 12654
+    },
+    {
+      "epoch": 0.7688802478886931,
+      "grad_norm": 0.2182929664850235,
+      "learning_rate": 1.269090806977477e-05,
+      "loss": 1.1778,
+      "step": 12655
+    },
+    {
+      "epoch": 0.7689410049213197,
+      "grad_norm": 0.184401273727417,
+      "learning_rate": 1.2684535754715732e-05,
+      "loss": 0.9632,
+      "step": 12656
+    },
+    {
+      "epoch": 0.7690017619539462,
+      "grad_norm": 0.12318959832191467,
+      "learning_rate": 1.2678164807462244e-05,
+      "loss": 1.022,
+      "step": 12657
+    },
+    {
+      "epoch": 0.7690625189865727,
+      "grad_norm": 0.206369087100029,
+      "learning_rate": 1.267179522824784e-05,
+      "loss": 1.0669,
+      "step": 12658
+    },
+    {
+      "epoch": 0.7691232760191993,
+      "grad_norm": 0.11718953400850296,
+      "learning_rate": 1.2665427017306002e-05,
+      "loss": 1.0868,
+      "step": 12659
+    },
+    {
+      "epoch": 0.7691840330518257,
+      "grad_norm": 0.3162151277065277,
+      "learning_rate": 1.2659060174870146e-05,
+      "loss": 1.197,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7692447900844522,
+      "grad_norm": 4.24622917175293,
+      "learning_rate": 1.2652694701173662e-05,
+      "loss": 1.1749,
+      "step": 12661
+    },
+    {
+      "epoch": 0.7693055471170788,
+      "grad_norm": 0.15454228222370148,
+      "learning_rate": 1.2646330596449873e-05,
+      "loss": 1.0722,
+      "step": 12662
+    },
+    {
+      "epoch": 0.7693663041497053,
+      "grad_norm": 0.1315198838710785,
+      "learning_rate": 1.2639967860932056e-05,
+      "loss": 1.0242,
+      "step": 12663
+    },
+    {
+      "epoch": 0.7694270611823318,
+      "grad_norm": 0.16658826172351837,
+      "learning_rate": 1.263360649485344e-05,
+      "loss": 1.1173,
+      "step": 12664
+    },
+    {
+      "epoch": 0.7694878182149584,
+      "grad_norm": 0.19234028458595276,
+      "learning_rate": 1.2627246498447182e-05,
+      "loss": 1.1459,
+      "step": 12665
+    },
+    {
+      "epoch": 0.7695485752475849,
+      "grad_norm": 0.18884725868701935,
+      "learning_rate": 1.2620887871946462e-05,
+      "loss": 1.0918,
+      "step": 12666
+    },
+    {
+      "epoch": 0.7696093322802114,
+      "grad_norm": 0.1349470168352127,
+      "learning_rate": 1.2614530615584303e-05,
+      "loss": 1.0923,
+      "step": 12667
+    },
+    {
+      "epoch": 0.769670089312838,
+      "grad_norm": 2.6494998931884766,
+      "learning_rate": 1.2608174729593753e-05,
+      "loss": 1.0735,
+      "step": 12668
+    },
+    {
+      "epoch": 0.7697308463454645,
+      "grad_norm": 0.159627303481102,
+      "learning_rate": 1.2601820214207783e-05,
+      "loss": 1.1251,
+      "step": 12669
+    },
+    {
+      "epoch": 0.769791603378091,
+      "grad_norm": 0.1379135102033615,
+      "learning_rate": 1.2595467069659312e-05,
+      "loss": 1.0477,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7698523604107176,
+      "grad_norm": 0.39774608612060547,
+      "learning_rate": 1.2589115296181237e-05,
+      "loss": 1.1864,
+      "step": 12671
+    },
+    {
+      "epoch": 0.7699131174433441,
+      "grad_norm": 0.15946069359779358,
+      "learning_rate": 1.2582764894006371e-05,
+      "loss": 1.1818,
+      "step": 12672
+    },
+    {
+      "epoch": 0.7699738744759705,
+      "grad_norm": 0.13292591273784637,
+      "learning_rate": 1.2576415863367508e-05,
+      "loss": 1.0215,
+      "step": 12673
+    },
+    {
+      "epoch": 0.7700346315085971,
+      "grad_norm": 0.1330001950263977,
+      "learning_rate": 1.2570068204497343e-05,
+      "loss": 1.0559,
+      "step": 12674
+    },
+    {
+      "epoch": 0.7700953885412236,
+      "grad_norm": 0.29064276814460754,
+      "learning_rate": 1.2563721917628546e-05,
+      "loss": 1.0435,
+      "step": 12675
+    },
+    {
+      "epoch": 0.7701561455738501,
+      "grad_norm": 0.1436695158481598,
+      "learning_rate": 1.255737700299377e-05,
+      "loss": 1.0638,
+      "step": 12676
+    },
+    {
+      "epoch": 0.7702169026064767,
+      "grad_norm": 0.0998484417796135,
+      "learning_rate": 1.2551033460825579e-05,
+      "loss": 1.0161,
+      "step": 12677
+    },
+    {
+      "epoch": 0.7702776596391032,
+      "grad_norm": 0.13796480000019073,
+      "learning_rate": 1.2544691291356497e-05,
+      "loss": 1.0462,
+      "step": 12678
+    },
+    {
+      "epoch": 0.7703384166717298,
+      "grad_norm": 0.18114705383777618,
+      "learning_rate": 1.2538350494818996e-05,
+      "loss": 1.094,
+      "step": 12679
+    },
+    {
+      "epoch": 0.7703991737043563,
+      "grad_norm": 0.1539202630519867,
+      "learning_rate": 1.25320110714455e-05,
+      "loss": 1.0909,
+      "step": 12680
+    },
+    {
+      "epoch": 0.7704599307369828,
+      "grad_norm": 0.10257013142108917,
+      "learning_rate": 1.2525673021468376e-05,
+      "loss": 0.9988,
+      "step": 12681
+    },
+    {
+      "epoch": 0.7705206877696094,
+      "grad_norm": 0.22239956259727478,
+      "learning_rate": 1.2519336345119953e-05,
+      "loss": 1.0645,
+      "step": 12682
+    },
+    {
+      "epoch": 0.7705814448022359,
+      "grad_norm": 0.1461649090051651,
+      "learning_rate": 1.25130010426325e-05,
+      "loss": 1.028,
+      "step": 12683
+    },
+    {
+      "epoch": 0.7706422018348624,
+      "grad_norm": 0.20986881852149963,
+      "learning_rate": 1.250666711423824e-05,
+      "loss": 1.1951,
+      "step": 12684
+    },
+    {
+      "epoch": 0.770702958867489,
+      "grad_norm": 0.1416889876127243,
+      "learning_rate": 1.2500334560169347e-05,
+      "loss": 1.0959,
+      "step": 12685
+    },
+    {
+      "epoch": 0.7707637159001154,
+      "grad_norm": 2.351872205734253,
+      "learning_rate": 1.2494003380657937e-05,
+      "loss": 1.0869,
+      "step": 12686
+    },
+    {
+      "epoch": 0.7708244729327419,
+      "grad_norm": 0.1239066869020462,
+      "learning_rate": 1.2487673575936082e-05,
+      "loss": 1.0489,
+      "step": 12687
+    },
+    {
+      "epoch": 0.7708852299653685,
+      "grad_norm": 0.2513124644756317,
+      "learning_rate": 1.2481345146235807e-05,
+      "loss": 1.2966,
+      "step": 12688
+    },
+    {
+      "epoch": 0.770945986997995,
+      "grad_norm": 0.15611149370670319,
+      "learning_rate": 1.2475018091789075e-05,
+      "loss": 1.1034,
+      "step": 12689
+    },
+    {
+      "epoch": 0.7710067440306215,
+      "grad_norm": 0.1298285275697708,
+      "learning_rate": 1.246869241282781e-05,
+      "loss": 1.067,
+      "step": 12690
+    },
+    {
+      "epoch": 0.7710675010632481,
+      "grad_norm": 1.1889944076538086,
+      "learning_rate": 1.2462368109583861e-05,
+      "loss": 1.1691,
+      "step": 12691
+    },
+    {
+      "epoch": 0.7711282580958746,
+      "grad_norm": 0.14899390935897827,
+      "learning_rate": 1.2456045182289083e-05,
+      "loss": 1.1229,
+      "step": 12692
+    },
+    {
+      "epoch": 0.7711890151285011,
+      "grad_norm": 0.14661161601543427,
+      "learning_rate": 1.2449723631175237e-05,
+      "loss": 1.2312,
+      "step": 12693
+    },
+    {
+      "epoch": 0.7712497721611277,
+      "grad_norm": 0.15550890564918518,
+      "learning_rate": 1.2443403456474017e-05,
+      "loss": 1.0361,
+      "step": 12694
+    },
+    {
+      "epoch": 0.7713105291937542,
+      "grad_norm": 0.1178649514913559,
+      "learning_rate": 1.2437084658417104e-05,
+      "loss": 1.0193,
+      "step": 12695
+    },
+    {
+      "epoch": 0.7713712862263807,
+      "grad_norm": 0.34149467945098877,
+      "learning_rate": 1.2430767237236096e-05,
+      "loss": 1.0803,
+      "step": 12696
+    },
+    {
+      "epoch": 0.7714320432590073,
+      "grad_norm": 0.20721812546253204,
+      "learning_rate": 1.2424451193162594e-05,
+      "loss": 1.0325,
+      "step": 12697
+    },
+    {
+      "epoch": 0.7714928002916338,
+      "grad_norm": 0.3458195924758911,
+      "learning_rate": 1.2418136526428098e-05,
+      "loss": 1.1719,
+      "step": 12698
+    },
+    {
+      "epoch": 0.7715535573242602,
+      "grad_norm": 0.1345955729484558,
+      "learning_rate": 1.2411823237264069e-05,
+      "loss": 1.1226,
+      "step": 12699
+    },
+    {
+      "epoch": 0.7716143143568868,
+      "grad_norm": 0.15379756689071655,
+      "learning_rate": 1.2405511325901942e-05,
+      "loss": 1.2648,
+      "step": 12700
+    },
+    {
+      "epoch": 0.7716750713895133,
+      "grad_norm": 0.10465983301401138,
+      "learning_rate": 1.239920079257303e-05,
+      "loss": 0.9921,
+      "step": 12701
+    },
+    {
+      "epoch": 0.7717358284221398,
+      "grad_norm": 0.3000970780849457,
+      "learning_rate": 1.2392891637508702e-05,
+      "loss": 1.1169,
+      "step": 12702
+    },
+    {
+      "epoch": 0.7717965854547664,
+      "grad_norm": 0.14947441220283508,
+      "learning_rate": 1.2386583860940198e-05,
+      "loss": 1.0486,
+      "step": 12703
+    },
+    {
+      "epoch": 0.7718573424873929,
+      "grad_norm": 0.18780438601970673,
+      "learning_rate": 1.2380277463098733e-05,
+      "loss": 1.0876,
+      "step": 12704
+    },
+    {
+      "epoch": 0.7719180995200194,
+      "grad_norm": 0.8837225437164307,
+      "learning_rate": 1.2373972444215464e-05,
+      "loss": 1.1414,
+      "step": 12705
+    },
+    {
+      "epoch": 0.771978856552646,
+      "grad_norm": 0.30864495038986206,
+      "learning_rate": 1.2367668804521515e-05,
+      "loss": 1.2305,
+      "step": 12706
+    },
+    {
+      "epoch": 0.7720396135852725,
+      "grad_norm": 0.17433524131774902,
+      "learning_rate": 1.2361366544247937e-05,
+      "loss": 1.1584,
+      "step": 12707
+    },
+    {
+      "epoch": 0.772100370617899,
+      "grad_norm": 0.14750021696090698,
+      "learning_rate": 1.2355065663625747e-05,
+      "loss": 1.073,
+      "step": 12708
+    },
+    {
+      "epoch": 0.7721611276505256,
+      "grad_norm": 0.1649530678987503,
+      "learning_rate": 1.23487661628859e-05,
+      "loss": 1.09,
+      "step": 12709
+    },
+    {
+      "epoch": 0.7722218846831521,
+      "grad_norm": 0.18619085848331451,
+      "learning_rate": 1.2342468042259309e-05,
+      "loss": 1.1082,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7722826417157787,
+      "grad_norm": 0.13991715013980865,
+      "learning_rate": 1.2336171301976824e-05,
+      "loss": 1.0397,
+      "step": 12711
+    },
+    {
+      "epoch": 0.7723433987484051,
+      "grad_norm": 0.18284957110881805,
+      "learning_rate": 1.2329875942269253e-05,
+      "loss": 1.0869,
+      "step": 12712
+    },
+    {
+      "epoch": 0.7724041557810316,
+      "grad_norm": 0.1092042550444603,
+      "learning_rate": 1.2323581963367386e-05,
+      "loss": 1.0545,
+      "step": 12713
+    },
+    {
+      "epoch": 0.7724649128136581,
+      "grad_norm": 0.16548630595207214,
+      "learning_rate": 1.2317289365501894e-05,
+      "loss": 1.0854,
+      "step": 12714
+    },
+    {
+      "epoch": 0.7725256698462847,
+      "grad_norm": 0.11044737696647644,
+      "learning_rate": 1.2310998148903441e-05,
+      "loss": 1.0844,
+      "step": 12715
+    },
+    {
+      "epoch": 0.7725864268789112,
+      "grad_norm": 0.13364773988723755,
+      "learning_rate": 1.2304708313802632e-05,
+      "loss": 1.0503,
+      "step": 12716
+    },
+    {
+      "epoch": 0.7726471839115377,
+      "grad_norm": 0.1477249413728714,
+      "learning_rate": 1.2298419860430016e-05,
+      "loss": 0.9986,
+      "step": 12717
+    },
+    {
+      "epoch": 0.7727079409441643,
+      "grad_norm": 0.12467773258686066,
+      "learning_rate": 1.2292132789016125e-05,
+      "loss": 1.0257,
+      "step": 12718
+    },
+    {
+      "epoch": 0.7727686979767908,
+      "grad_norm": 0.13956567645072937,
+      "learning_rate": 1.2285847099791391e-05,
+      "loss": 1.1357,
+      "step": 12719
+    },
+    {
+      "epoch": 0.7728294550094174,
+      "grad_norm": 0.2354406863451004,
+      "learning_rate": 1.2279562792986238e-05,
+      "loss": 1.0079,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7728902120420439,
+      "grad_norm": 0.39486417174339294,
+      "learning_rate": 1.2273279868830973e-05,
+      "loss": 1.1166,
+      "step": 12721
+    },
+    {
+      "epoch": 0.7729509690746704,
+      "grad_norm": 0.18616119027137756,
+      "learning_rate": 1.2266998327555939e-05,
+      "loss": 1.0442,
+      "step": 12722
+    },
+    {
+      "epoch": 0.773011726107297,
+      "grad_norm": 0.2511233389377594,
+      "learning_rate": 1.2260718169391377e-05,
+      "loss": 1.1556,
+      "step": 12723
+    },
+    {
+      "epoch": 0.7730724831399235,
+      "grad_norm": 0.13469520211219788,
+      "learning_rate": 1.225443939456749e-05,
+      "loss": 1.2609,
+      "step": 12724
+    },
+    {
+      "epoch": 0.7731332401725499,
+      "grad_norm": 0.1431630700826645,
+      "learning_rate": 1.2248162003314418e-05,
+      "loss": 1.041,
+      "step": 12725
+    },
+    {
+      "epoch": 0.7731939972051765,
+      "grad_norm": 0.25470998883247375,
+      "learning_rate": 1.2241885995862262e-05,
+      "loss": 1.1215,
+      "step": 12726
+    },
+    {
+      "epoch": 0.773254754237803,
+      "grad_norm": 0.6969568133354187,
+      "learning_rate": 1.223561137244108e-05,
+      "loss": 1.1334,
+      "step": 12727
+    },
+    {
+      "epoch": 0.7733155112704295,
+      "grad_norm": 0.11331703513860703,
+      "learning_rate": 1.222933813328086e-05,
+      "loss": 0.9843,
+      "step": 12728
+    },
+    {
+      "epoch": 0.7733762683030561,
+      "grad_norm": 0.1726769357919693,
+      "learning_rate": 1.222306627861155e-05,
+      "loss": 1.0938,
+      "step": 12729
+    },
+    {
+      "epoch": 0.7734370253356826,
+      "grad_norm": 0.12890057265758514,
+      "learning_rate": 1.221679580866305e-05,
+      "loss": 1.0258,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7734977823683091,
+      "grad_norm": 0.144577294588089,
+      "learning_rate": 1.22105267236652e-05,
+      "loss": 1.1019,
+      "step": 12731
+    },
+    {
+      "epoch": 0.7735585394009357,
+      "grad_norm": 0.1436421275138855,
+      "learning_rate": 1.2204259023847796e-05,
+      "loss": 1.102,
+      "step": 12732
+    },
+    {
+      "epoch": 0.7736192964335622,
+      "grad_norm": 0.21898682415485382,
+      "learning_rate": 1.2197992709440592e-05,
+      "loss": 1.124,
+      "step": 12733
+    },
+    {
+      "epoch": 0.7736800534661887,
+      "grad_norm": 0.28340819478034973,
+      "learning_rate": 1.2191727780673262e-05,
+      "loss": 1.1258,
+      "step": 12734
+    },
+    {
+      "epoch": 0.7737408104988153,
+      "grad_norm": 0.1154094710946083,
+      "learning_rate": 1.2185464237775467e-05,
+      "loss": 1.054,
+      "step": 12735
+    },
+    {
+      "epoch": 0.7738015675314418,
+      "grad_norm": 0.12246297299861908,
+      "learning_rate": 1.217920208097678e-05,
+      "loss": 1.0494,
+      "step": 12736
+    },
+    {
+      "epoch": 0.7738623245640683,
+      "grad_norm": 0.629007875919342,
+      "learning_rate": 1.2172941310506748e-05,
+      "loss": 1.0815,
+      "step": 12737
+    },
+    {
+      "epoch": 0.7739230815966949,
+      "grad_norm": 0.22257009148597717,
+      "learning_rate": 1.216668192659487e-05,
+      "loss": 1.0272,
+      "step": 12738
+    },
+    {
+      "epoch": 0.7739838386293213,
+      "grad_norm": 0.13889539241790771,
+      "learning_rate": 1.2160423929470582e-05,
+      "loss": 1.079,
+      "step": 12739
+    },
+    {
+      "epoch": 0.7740445956619478,
+      "grad_norm": 0.12575124204158783,
+      "learning_rate": 1.2154167319363286e-05,
+      "loss": 0.982,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7741053526945744,
+      "grad_norm": 0.12299622595310211,
+      "learning_rate": 1.2147912096502285e-05,
+      "loss": 1.0463,
+      "step": 12741
+    },
+    {
+      "epoch": 0.7741661097272009,
+      "grad_norm": 0.18291832506656647,
+      "learning_rate": 1.2141658261116867e-05,
+      "loss": 1.0606,
+      "step": 12742
+    },
+    {
+      "epoch": 0.7742268667598274,
+      "grad_norm": 0.16416339576244354,
+      "learning_rate": 1.2135405813436301e-05,
+      "loss": 1.0713,
+      "step": 12743
+    },
+    {
+      "epoch": 0.774287623792454,
+      "grad_norm": 0.1198337972164154,
+      "learning_rate": 1.2129154753689754e-05,
+      "loss": 1.1134,
+      "step": 12744
+    },
+    {
+      "epoch": 0.7743483808250805,
+      "grad_norm": 0.11347414553165436,
+      "learning_rate": 1.2122905082106357e-05,
+      "loss": 1.0261,
+      "step": 12745
+    },
+    {
+      "epoch": 0.774409137857707,
+      "grad_norm": 0.16280025243759155,
+      "learning_rate": 1.21166567989152e-05,
+      "loss": 1.2671,
+      "step": 12746
+    },
+    {
+      "epoch": 0.7744698948903336,
+      "grad_norm": 0.17514172196388245,
+      "learning_rate": 1.2110409904345305e-05,
+      "loss": 1.1239,
+      "step": 12747
+    },
+    {
+      "epoch": 0.7745306519229601,
+      "grad_norm": 0.15606075525283813,
+      "learning_rate": 1.2104164398625662e-05,
+      "loss": 1.131,
+      "step": 12748
+    },
+    {
+      "epoch": 0.7745914089555866,
+      "grad_norm": 0.12669014930725098,
+      "learning_rate": 1.2097920281985198e-05,
+      "loss": 1.0605,
+      "step": 12749
+    },
+    {
+      "epoch": 0.7746521659882132,
+      "grad_norm": 0.7310323119163513,
+      "learning_rate": 1.2091677554652792e-05,
+      "loss": 1.119,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7747129230208397,
+      "grad_norm": 0.10180962085723877,
+      "learning_rate": 1.2085436216857277e-05,
+      "loss": 1.0049,
+      "step": 12751
+    },
+    {
+      "epoch": 0.7747736800534661,
+      "grad_norm": 0.10264743119478226,
+      "learning_rate": 1.2079196268827419e-05,
+      "loss": 1.0467,
+      "step": 12752
+    },
+    {
+      "epoch": 0.7748344370860927,
+      "grad_norm": 0.11282286047935486,
+      "learning_rate": 1.2072957710791955e-05,
+      "loss": 1.0158,
+      "step": 12753
+    },
+    {
+      "epoch": 0.7748951941187192,
+      "grad_norm": 0.18218880891799927,
+      "learning_rate": 1.206672054297956e-05,
+      "loss": 1.1653,
+      "step": 12754
+    },
+    {
+      "epoch": 0.7749559511513457,
+      "grad_norm": 0.27505213022232056,
+      "learning_rate": 1.2060484765618851e-05,
+      "loss": 1.2276,
+      "step": 12755
+    },
+    {
+      "epoch": 0.7750167081839723,
+      "grad_norm": 0.17598998546600342,
+      "learning_rate": 1.205425037893841e-05,
+      "loss": 1.0465,
+      "step": 12756
+    },
+    {
+      "epoch": 0.7750774652165988,
+      "grad_norm": 0.11984530836343765,
+      "learning_rate": 1.2048017383166753e-05,
+      "loss": 0.9833,
+      "step": 12757
+    },
+    {
+      "epoch": 0.7751382222492254,
+      "grad_norm": 0.1310300976037979,
+      "learning_rate": 1.2041785778532339e-05,
+      "loss": 1.0269,
+      "step": 12758
+    },
+    {
+      "epoch": 0.7751989792818519,
+      "grad_norm": 0.23538441956043243,
+      "learning_rate": 1.2035555565263623e-05,
+      "loss": 1.1002,
+      "step": 12759
+    },
+    {
+      "epoch": 0.7752597363144784,
+      "grad_norm": 0.12445208430290222,
+      "learning_rate": 1.202932674358897e-05,
+      "loss": 1.0521,
+      "step": 12760
+    },
+    {
+      "epoch": 0.775320493347105,
+      "grad_norm": 0.1923869252204895,
+      "learning_rate": 1.202309931373667e-05,
+      "loss": 1.0318,
+      "step": 12761
+    },
+    {
+      "epoch": 0.7753812503797315,
+      "grad_norm": 0.3935041129589081,
+      "learning_rate": 1.2016873275935004e-05,
+      "loss": 1.0251,
+      "step": 12762
+    },
+    {
+      "epoch": 0.775442007412358,
+      "grad_norm": 0.13856910169124603,
+      "learning_rate": 1.2010648630412175e-05,
+      "loss": 1.0557,
+      "step": 12763
+    },
+    {
+      "epoch": 0.7755027644449846,
+      "grad_norm": 0.19792990386486053,
+      "learning_rate": 1.2004425377396383e-05,
+      "loss": 1.1648,
+      "step": 12764
+    },
+    {
+      "epoch": 0.775563521477611,
+      "grad_norm": 0.14647440612316132,
+      "learning_rate": 1.199820351711572e-05,
+      "loss": 1.0814,
+      "step": 12765
+    },
+    {
+      "epoch": 0.7756242785102375,
+      "grad_norm": 0.13246290385723114,
+      "learning_rate": 1.199198304979825e-05,
+      "loss": 1.0764,
+      "step": 12766
+    },
+    {
+      "epoch": 0.775685035542864,
+      "grad_norm": 1.2874914407730103,
+      "learning_rate": 1.1985763975672004e-05,
+      "loss": 1.003,
+      "step": 12767
+    },
+    {
+      "epoch": 0.7757457925754906,
+      "grad_norm": 0.24590452015399933,
+      "learning_rate": 1.1979546294964899e-05,
+      "loss": 1.2237,
+      "step": 12768
+    },
+    {
+      "epoch": 0.7758065496081171,
+      "grad_norm": 0.4111698567867279,
+      "learning_rate": 1.1973330007904892e-05,
+      "loss": 1.1349,
+      "step": 12769
+    },
+    {
+      "epoch": 0.7758673066407437,
+      "grad_norm": 0.10011672973632812,
+      "learning_rate": 1.1967115114719818e-05,
+      "loss": 1.0022,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7759280636733702,
+      "grad_norm": 0.26296618580818176,
+      "learning_rate": 1.1960901615637494e-05,
+      "loss": 1.1378,
+      "step": 12771
+    },
+    {
+      "epoch": 0.7759888207059967,
+      "grad_norm": 0.14681220054626465,
+      "learning_rate": 1.1954689510885675e-05,
+      "loss": 1.0925,
+      "step": 12772
+    },
+    {
+      "epoch": 0.7760495777386233,
+      "grad_norm": 0.154121994972229,
+      "learning_rate": 1.1948478800692065e-05,
+      "loss": 1.1075,
+      "step": 12773
+    },
+    {
+      "epoch": 0.7761103347712498,
+      "grad_norm": 0.10315140336751938,
+      "learning_rate": 1.194226948528432e-05,
+      "loss": 1.0448,
+      "step": 12774
+    },
+    {
+      "epoch": 0.7761710918038763,
+      "grad_norm": 0.116829514503479,
+      "learning_rate": 1.1936061564890044e-05,
+      "loss": 1.0315,
+      "step": 12775
+    },
+    {
+      "epoch": 0.7762318488365029,
+      "grad_norm": 0.13943418860435486,
+      "learning_rate": 1.1929855039736786e-05,
+      "loss": 1.0536,
+      "step": 12776
+    },
+    {
+      "epoch": 0.7762926058691294,
+      "grad_norm": 0.18070149421691895,
+      "learning_rate": 1.1923649910052054e-05,
+      "loss": 1.1006,
+      "step": 12777
+    },
+    {
+      "epoch": 0.7763533629017558,
+      "grad_norm": 0.13364963233470917,
+      "learning_rate": 1.191744617606329e-05,
+      "loss": 1.0934,
+      "step": 12778
+    },
+    {
+      "epoch": 0.7764141199343824,
+      "grad_norm": 0.1769493967294693,
+      "learning_rate": 1.1911243837997887e-05,
+      "loss": 1.095,
+      "step": 12779
+    },
+    {
+      "epoch": 0.7764748769670089,
+      "grad_norm": 0.12219464778900146,
+      "learning_rate": 1.1905042896083235e-05,
+      "loss": 1.0864,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7765356339996354,
+      "grad_norm": 0.23633837699890137,
+      "learning_rate": 1.1898843350546579e-05,
+      "loss": 1.1657,
+      "step": 12781
+    },
+    {
+      "epoch": 0.776596391032262,
+      "grad_norm": 0.16362082958221436,
+      "learning_rate": 1.1892645201615188e-05,
+      "loss": 1.0872,
+      "step": 12782
+    },
+    {
+      "epoch": 0.7766571480648885,
+      "grad_norm": 0.1066916361451149,
+      "learning_rate": 1.1886448449516251e-05,
+      "loss": 1.0152,
+      "step": 12783
+    },
+    {
+      "epoch": 0.776717905097515,
+      "grad_norm": 0.11957157403230667,
+      "learning_rate": 1.1880253094476901e-05,
+      "loss": 1.0419,
+      "step": 12784
+    },
+    {
+      "epoch": 0.7767786621301416,
+      "grad_norm": 0.12654800713062286,
+      "learning_rate": 1.1874059136724253e-05,
+      "loss": 1.0735,
+      "step": 12785
+    },
+    {
+      "epoch": 0.7768394191627681,
+      "grad_norm": 0.27567052841186523,
+      "learning_rate": 1.1867866576485342e-05,
+      "loss": 1.2144,
+      "step": 12786
+    },
+    {
+      "epoch": 0.7769001761953946,
+      "grad_norm": 0.14142964780330658,
+      "learning_rate": 1.1861675413987162e-05,
+      "loss": 1.0765,
+      "step": 12787
+    },
+    {
+      "epoch": 0.7769609332280212,
+      "grad_norm": 0.178693950176239,
+      "learning_rate": 1.1855485649456633e-05,
+      "loss": 1.1344,
+      "step": 12788
+    },
+    {
+      "epoch": 0.7770216902606477,
+      "grad_norm": 0.6197378635406494,
+      "learning_rate": 1.184929728312063e-05,
+      "loss": 1.0859,
+      "step": 12789
+    },
+    {
+      "epoch": 0.7770824472932742,
+      "grad_norm": 0.1488838493824005,
+      "learning_rate": 1.184311031520603e-05,
+      "loss": 1.0169,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7771432043259007,
+      "grad_norm": 0.1694076657295227,
+      "learning_rate": 1.1836924745939593e-05,
+      "loss": 1.1357,
+      "step": 12791
+    },
+    {
+      "epoch": 0.7772039613585272,
+      "grad_norm": 0.13343407213687897,
+      "learning_rate": 1.1830740575548055e-05,
+      "loss": 1.1007,
+      "step": 12792
+    },
+    {
+      "epoch": 0.7772647183911537,
+      "grad_norm": 0.21537846326828003,
+      "learning_rate": 1.1824557804258101e-05,
+      "loss": 1.1133,
+      "step": 12793
+    },
+    {
+      "epoch": 0.7773254754237803,
+      "grad_norm": 0.17056311666965485,
+      "learning_rate": 1.1818376432296358e-05,
+      "loss": 1.0999,
+      "step": 12794
+    },
+    {
+      "epoch": 0.7773862324564068,
+      "grad_norm": 0.19689084589481354,
+      "learning_rate": 1.1812196459889413e-05,
+      "loss": 1.2193,
+      "step": 12795
+    },
+    {
+      "epoch": 0.7774469894890333,
+      "grad_norm": 0.22491297125816345,
+      "learning_rate": 1.1806017887263782e-05,
+      "loss": 1.0621,
+      "step": 12796
+    },
+    {
+      "epoch": 0.7775077465216599,
+      "grad_norm": 0.2003372311592102,
+      "learning_rate": 1.179984071464595e-05,
+      "loss": 1.0682,
+      "step": 12797
+    },
+    {
+      "epoch": 0.7775685035542864,
+      "grad_norm": 0.17473378777503967,
+      "learning_rate": 1.179366494226234e-05,
+      "loss": 1.0865,
+      "step": 12798
+    },
+    {
+      "epoch": 0.777629260586913,
+      "grad_norm": 0.3054414391517639,
+      "learning_rate": 1.1787490570339326e-05,
+      "loss": 1.196,
+      "step": 12799
+    },
+    {
+      "epoch": 0.7776900176195395,
+      "grad_norm": 0.10129894316196442,
+      "learning_rate": 1.1781317599103236e-05,
+      "loss": 1.0289,
+      "step": 12800
+    },
+    {
+      "epoch": 0.777750774652166,
+      "grad_norm": 0.311766117811203,
+      "learning_rate": 1.1775146028780332e-05,
+      "loss": 1.1187,
+      "step": 12801
+    },
+    {
+      "epoch": 0.7778115316847926,
+      "grad_norm": 0.13467176258563995,
+      "learning_rate": 1.1768975859596843e-05,
+      "loss": 1.0423,
+      "step": 12802
+    },
+    {
+      "epoch": 0.7778722887174191,
+      "grad_norm": 0.18228508532047272,
+      "learning_rate": 1.1762807091778933e-05,
+      "loss": 1.0727,
+      "step": 12803
+    },
+    {
+      "epoch": 0.7779330457500455,
+      "grad_norm": 0.12224791198968887,
+      "learning_rate": 1.175663972555272e-05,
+      "loss": 1.1033,
+      "step": 12804
+    },
+    {
+      "epoch": 0.777993802782672,
+      "grad_norm": 0.14524850249290466,
+      "learning_rate": 1.1750473761144254e-05,
+      "loss": 1.046,
+      "step": 12805
+    },
+    {
+      "epoch": 0.7780545598152986,
+      "grad_norm": 0.1831561028957367,
+      "learning_rate": 1.174430919877958e-05,
+      "loss": 1.0119,
+      "step": 12806
+    },
+    {
+      "epoch": 0.7781153168479251,
+      "grad_norm": 0.3943927586078644,
+      "learning_rate": 1.1738146038684662e-05,
+      "loss": 1.2763,
+      "step": 12807
+    },
+    {
+      "epoch": 0.7781760738805517,
+      "grad_norm": 0.6518760323524475,
+      "learning_rate": 1.1731984281085385e-05,
+      "loss": 1.0511,
+      "step": 12808
+    },
+    {
+      "epoch": 0.7782368309131782,
+      "grad_norm": 0.12650218605995178,
+      "learning_rate": 1.1725823926207612e-05,
+      "loss": 1.0181,
+      "step": 12809
+    },
+    {
+      "epoch": 0.7782975879458047,
+      "grad_norm": 0.11816632747650146,
+      "learning_rate": 1.1719664974277156e-05,
+      "loss": 1.0329,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7783583449784313,
+      "grad_norm": 0.16062131524085999,
+      "learning_rate": 1.1713507425519793e-05,
+      "loss": 1.1158,
+      "step": 12811
+    },
+    {
+      "epoch": 0.7784191020110578,
+      "grad_norm": 0.17883680760860443,
+      "learning_rate": 1.1707351280161216e-05,
+      "loss": 1.1943,
+      "step": 12812
+    },
+    {
+      "epoch": 0.7784798590436843,
+      "grad_norm": 0.15480759739875793,
+      "learning_rate": 1.1701196538427079e-05,
+      "loss": 1.1563,
+      "step": 12813
+    },
+    {
+      "epoch": 0.7785406160763109,
+      "grad_norm": 0.1633390337228775,
+      "learning_rate": 1.1695043200542982e-05,
+      "loss": 1.0152,
+      "step": 12814
+    },
+    {
+      "epoch": 0.7786013731089374,
+      "grad_norm": 0.16703739762306213,
+      "learning_rate": 1.1688891266734486e-05,
+      "loss": 1.047,
+      "step": 12815
+    },
+    {
+      "epoch": 0.7786621301415639,
+      "grad_norm": 0.15224634110927582,
+      "learning_rate": 1.1682740737227083e-05,
+      "loss": 1.1063,
+      "step": 12816
+    },
+    {
+      "epoch": 0.7787228871741904,
+      "grad_norm": 0.13909094035625458,
+      "learning_rate": 1.1676591612246224e-05,
+      "loss": 1.0591,
+      "step": 12817
+    },
+    {
+      "epoch": 0.7787836442068169,
+      "grad_norm": 0.14900583028793335,
+      "learning_rate": 1.1670443892017314e-05,
+      "loss": 1.1542,
+      "step": 12818
+    },
+    {
+      "epoch": 0.7788444012394434,
+      "grad_norm": 0.22665028274059296,
+      "learning_rate": 1.1664297576765687e-05,
+      "loss": 1.1414,
+      "step": 12819
+    },
+    {
+      "epoch": 0.77890515827207,
+      "grad_norm": 0.15524421632289886,
+      "learning_rate": 1.1658152666716638e-05,
+      "loss": 1.0208,
+      "step": 12820
+    },
+    {
+      "epoch": 0.7789659153046965,
+      "grad_norm": 0.14574138820171356,
+      "learning_rate": 1.1652009162095424e-05,
+      "loss": 1.1487,
+      "step": 12821
+    },
+    {
+      "epoch": 0.779026672337323,
+      "grad_norm": 0.3664519786834717,
+      "learning_rate": 1.1645867063127219e-05,
+      "loss": 1.0267,
+      "step": 12822
+    },
+    {
+      "epoch": 0.7790874293699496,
+      "grad_norm": 0.20768223702907562,
+      "learning_rate": 1.1639726370037174e-05,
+      "loss": 1.022,
+      "step": 12823
+    },
+    {
+      "epoch": 0.7791481864025761,
+      "grad_norm": 0.17619727551937103,
+      "learning_rate": 1.1633587083050374e-05,
+      "loss": 1.1451,
+      "step": 12824
+    },
+    {
+      "epoch": 0.7792089434352026,
+      "grad_norm": 0.13255421817302704,
+      "learning_rate": 1.1627449202391849e-05,
+      "loss": 1.0363,
+      "step": 12825
+    },
+    {
+      "epoch": 0.7792697004678292,
+      "grad_norm": 0.12598589062690735,
+      "learning_rate": 1.1621312728286581e-05,
+      "loss": 1.0438,
+      "step": 12826
+    },
+    {
+      "epoch": 0.7793304575004557,
+      "grad_norm": 0.6745483875274658,
+      "learning_rate": 1.1615177660959547e-05,
+      "loss": 1.1095,
+      "step": 12827
+    },
+    {
+      "epoch": 0.7793912145330822,
+      "grad_norm": 0.18534834682941437,
+      "learning_rate": 1.1609044000635577e-05,
+      "loss": 1.0394,
+      "step": 12828
+    },
+    {
+      "epoch": 0.7794519715657088,
+      "grad_norm": 0.13179436326026917,
+      "learning_rate": 1.160291174753952e-05,
+      "loss": 1.0576,
+      "step": 12829
+    },
+    {
+      "epoch": 0.7795127285983352,
+      "grad_norm": 0.17568884789943695,
+      "learning_rate": 1.1596780901896138e-05,
+      "loss": 1.0925,
+      "step": 12830
+    },
+    {
+      "epoch": 0.7795734856309617,
+      "grad_norm": 0.2002159059047699,
+      "learning_rate": 1.1590651463930191e-05,
+      "loss": 1.0955,
+      "step": 12831
+    },
+    {
+      "epoch": 0.7796342426635883,
+      "grad_norm": 0.1371026635169983,
+      "learning_rate": 1.158452343386634e-05,
+      "loss": 0.9927,
+      "step": 12832
+    },
+    {
+      "epoch": 0.7796949996962148,
+      "grad_norm": 0.6081156730651855,
+      "learning_rate": 1.157839681192921e-05,
+      "loss": 1.2386,
+      "step": 12833
+    },
+    {
+      "epoch": 0.7797557567288413,
+      "grad_norm": 0.1321309506893158,
+      "learning_rate": 1.1572271598343381e-05,
+      "loss": 1.0653,
+      "step": 12834
+    },
+    {
+      "epoch": 0.7798165137614679,
+      "grad_norm": 0.16023296117782593,
+      "learning_rate": 1.1566147793333343e-05,
+      "loss": 1.0827,
+      "step": 12835
+    },
+    {
+      "epoch": 0.7798772707940944,
+      "grad_norm": 0.1974269151687622,
+      "learning_rate": 1.1560025397123592e-05,
+      "loss": 1.1422,
+      "step": 12836
+    },
+    {
+      "epoch": 0.779938027826721,
+      "grad_norm": 0.18258316814899445,
+      "learning_rate": 1.1553904409938548e-05,
+      "loss": 1.0663,
+      "step": 12837
+    },
+    {
+      "epoch": 0.7799987848593475,
+      "grad_norm": 0.15497559309005737,
+      "learning_rate": 1.1547784832002567e-05,
+      "loss": 1.0619,
+      "step": 12838
+    },
+    {
+      "epoch": 0.780059541891974,
+      "grad_norm": 0.09782491624355316,
+      "learning_rate": 1.1541666663539963e-05,
+      "loss": 0.9979,
+      "step": 12839
+    },
+    {
+      "epoch": 0.7801202989246006,
+      "grad_norm": 0.7276683449745178,
+      "learning_rate": 1.1535549904775006e-05,
+      "loss": 1.1006,
+      "step": 12840
+    },
+    {
+      "epoch": 0.7801810559572271,
+      "grad_norm": 7.402686595916748,
+      "learning_rate": 1.15294345559319e-05,
+      "loss": 1.0741,
+      "step": 12841
+    },
+    {
+      "epoch": 0.7802418129898536,
+      "grad_norm": 0.1841738522052765,
+      "learning_rate": 1.1523320617234806e-05,
+      "loss": 1.1202,
+      "step": 12842
+    },
+    {
+      "epoch": 0.7803025700224802,
+      "grad_norm": 0.14523489773273468,
+      "learning_rate": 1.1517208088907833e-05,
+      "loss": 1.0366,
+      "step": 12843
+    },
+    {
+      "epoch": 0.7803633270551066,
+      "grad_norm": 0.12169469147920609,
+      "learning_rate": 1.1511096971175034e-05,
+      "loss": 1.0584,
+      "step": 12844
+    },
+    {
+      "epoch": 0.7804240840877331,
+      "grad_norm": 0.16683024168014526,
+      "learning_rate": 1.1504987264260414e-05,
+      "loss": 1.0587,
+      "step": 12845
+    },
+    {
+      "epoch": 0.7804848411203597,
+      "grad_norm": 0.17324349284172058,
+      "learning_rate": 1.1498878968387916e-05,
+      "loss": 1.0863,
+      "step": 12846
+    },
+    {
+      "epoch": 0.7805455981529862,
+      "grad_norm": 0.27542775869369507,
+      "learning_rate": 1.1492772083781478e-05,
+      "loss": 1.022,
+      "step": 12847
+    },
+    {
+      "epoch": 0.7806063551856127,
+      "grad_norm": 0.13815228641033173,
+      "learning_rate": 1.1486666610664908e-05,
+      "loss": 1.1454,
+      "step": 12848
+    },
+    {
+      "epoch": 0.7806671122182393,
+      "grad_norm": 0.12550212442874908,
+      "learning_rate": 1.1480562549262019e-05,
+      "loss": 1.0226,
+      "step": 12849
+    },
+    {
+      "epoch": 0.7807278692508658,
+      "grad_norm": 0.12101548165082932,
+      "learning_rate": 1.1474459899796559e-05,
+      "loss": 1.0132,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7807886262834923,
+      "grad_norm": 0.1512622833251953,
+      "learning_rate": 1.1468358662492202e-05,
+      "loss": 1.0244,
+      "step": 12851
+    },
+    {
+      "epoch": 0.7808493833161189,
+      "grad_norm": 0.3264085054397583,
+      "learning_rate": 1.1462258837572615e-05,
+      "loss": 1.047,
+      "step": 12852
+    },
+    {
+      "epoch": 0.7809101403487454,
+      "grad_norm": 0.15057966113090515,
+      "learning_rate": 1.1456160425261386e-05,
+      "loss": 1.0285,
+      "step": 12853
+    },
+    {
+      "epoch": 0.7809708973813719,
+      "grad_norm": 0.11836346238851547,
+      "learning_rate": 1.145006342578206e-05,
+      "loss": 1.0454,
+      "step": 12854
+    },
+    {
+      "epoch": 0.7810316544139985,
+      "grad_norm": 0.19074952602386475,
+      "learning_rate": 1.1443967839358094e-05,
+      "loss": 1.1281,
+      "step": 12855
+    },
+    {
+      "epoch": 0.781092411446625,
+      "grad_norm": 0.2930346429347992,
+      "learning_rate": 1.1437873666212928e-05,
+      "loss": 1.0946,
+      "step": 12856
+    },
+    {
+      "epoch": 0.7811531684792514,
+      "grad_norm": 0.13799655437469482,
+      "learning_rate": 1.1431780906569966e-05,
+      "loss": 1.0709,
+      "step": 12857
+    },
+    {
+      "epoch": 0.781213925511878,
+      "grad_norm": 0.12468823045492172,
+      "learning_rate": 1.1425689560652536e-05,
+      "loss": 1.0806,
+      "step": 12858
+    },
+    {
+      "epoch": 0.7812746825445045,
+      "grad_norm": 0.1512451022863388,
+      "learning_rate": 1.1419599628683909e-05,
+      "loss": 1.0849,
+      "step": 12859
+    },
+    {
+      "epoch": 0.781335439577131,
+      "grad_norm": 0.2822954058647156,
+      "learning_rate": 1.1413511110887314e-05,
+      "loss": 1.0825,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7813961966097576,
+      "grad_norm": 0.10462968796491623,
+      "learning_rate": 1.1407424007485929e-05,
+      "loss": 0.9911,
+      "step": 12861
+    },
+    {
+      "epoch": 0.7814569536423841,
+      "grad_norm": 0.12267274409532547,
+      "learning_rate": 1.1401338318702876e-05,
+      "loss": 1.0221,
+      "step": 12862
+    },
+    {
+      "epoch": 0.7815177106750106,
+      "grad_norm": 0.14118769764900208,
+      "learning_rate": 1.1395254044761227e-05,
+      "loss": 1.1061,
+      "step": 12863
+    },
+    {
+      "epoch": 0.7815784677076372,
+      "grad_norm": 0.157334104180336,
+      "learning_rate": 1.1389171185884006e-05,
+      "loss": 1.1094,
+      "step": 12864
+    },
+    {
+      "epoch": 0.7816392247402637,
+      "grad_norm": 0.2385435551404953,
+      "learning_rate": 1.1383089742294179e-05,
+      "loss": 1.1388,
+      "step": 12865
+    },
+    {
+      "epoch": 0.7816999817728902,
+      "grad_norm": 0.2596820890903473,
+      "learning_rate": 1.1377009714214658e-05,
+      "loss": 1.11,
+      "step": 12866
+    },
+    {
+      "epoch": 0.7817607388055168,
+      "grad_norm": 0.2865890562534332,
+      "learning_rate": 1.1370931101868315e-05,
+      "loss": 1.0757,
+      "step": 12867
+    },
+    {
+      "epoch": 0.7818214958381433,
+      "grad_norm": 0.104372538626194,
+      "learning_rate": 1.1364853905477956e-05,
+      "loss": 1.0245,
+      "step": 12868
+    },
+    {
+      "epoch": 0.7818822528707698,
+      "grad_norm": 0.16807779669761658,
+      "learning_rate": 1.1358778125266344e-05,
+      "loss": 1.1249,
+      "step": 12869
+    },
+    {
+      "epoch": 0.7819430099033963,
+      "grad_norm": 0.25906169414520264,
+      "learning_rate": 1.1352703761456196e-05,
+      "loss": 1.1488,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7820037669360228,
+      "grad_norm": 0.3172304928302765,
+      "learning_rate": 1.1346630814270154e-05,
+      "loss": 1.0882,
+      "step": 12871
+    },
+    {
+      "epoch": 0.7820645239686493,
+      "grad_norm": 0.1424168199300766,
+      "learning_rate": 1.1340559283930823e-05,
+      "loss": 1.0628,
+      "step": 12872
+    },
+    {
+      "epoch": 0.7821252810012759,
+      "grad_norm": 0.13839729130268097,
+      "learning_rate": 1.1334489170660779e-05,
+      "loss": 1.0628,
+      "step": 12873
+    },
+    {
+      "epoch": 0.7821860380339024,
+      "grad_norm": 0.26125919818878174,
+      "learning_rate": 1.1328420474682516e-05,
+      "loss": 1.1765,
+      "step": 12874
+    },
+    {
+      "epoch": 0.7822467950665289,
+      "grad_norm": 0.1128741130232811,
+      "learning_rate": 1.1322353196218465e-05,
+      "loss": 1.0465,
+      "step": 12875
+    },
+    {
+      "epoch": 0.7823075520991555,
+      "grad_norm": 0.1515772044658661,
+      "learning_rate": 1.1316287335491032e-05,
+      "loss": 1.0886,
+      "step": 12876
+    },
+    {
+      "epoch": 0.782368309131782,
+      "grad_norm": 0.14717566967010498,
+      "learning_rate": 1.1310222892722556e-05,
+      "loss": 1.1337,
+      "step": 12877
+    },
+    {
+      "epoch": 0.7824290661644085,
+      "grad_norm": 0.19552476704120636,
+      "learning_rate": 1.1304159868135344e-05,
+      "loss": 1.1012,
+      "step": 12878
+    },
+    {
+      "epoch": 0.7824898231970351,
+      "grad_norm": 0.2012212723493576,
+      "learning_rate": 1.129809826195164e-05,
+      "loss": 1.1345,
+      "step": 12879
+    },
+    {
+      "epoch": 0.7825505802296616,
+      "grad_norm": 0.2269507199525833,
+      "learning_rate": 1.1292038074393618e-05,
+      "loss": 1.2227,
+      "step": 12880
+    },
+    {
+      "epoch": 0.7826113372622882,
+      "grad_norm": 0.1639718860387802,
+      "learning_rate": 1.1285979305683441e-05,
+      "loss": 1.0904,
+      "step": 12881
+    },
+    {
+      "epoch": 0.7826720942949147,
+      "grad_norm": 0.9077989459037781,
+      "learning_rate": 1.1279921956043144e-05,
+      "loss": 1.0791,
+      "step": 12882
+    },
+    {
+      "epoch": 0.7827328513275411,
+      "grad_norm": 0.14997003972530365,
+      "learning_rate": 1.1273866025694813e-05,
+      "loss": 1.0671,
+      "step": 12883
+    },
+    {
+      "epoch": 0.7827936083601676,
+      "grad_norm": 0.14225393533706665,
+      "learning_rate": 1.1267811514860399e-05,
+      "loss": 1.109,
+      "step": 12884
+    },
+    {
+      "epoch": 0.7828543653927942,
+      "grad_norm": 0.19047455489635468,
+      "learning_rate": 1.1261758423761848e-05,
+      "loss": 1.1537,
+      "step": 12885
+    },
+    {
+      "epoch": 0.7829151224254207,
+      "grad_norm": 0.15400217473506927,
+      "learning_rate": 1.1255706752621026e-05,
+      "loss": 1.0874,
+      "step": 12886
+    },
+    {
+      "epoch": 0.7829758794580473,
+      "grad_norm": 0.19490458071231842,
+      "learning_rate": 1.1249656501659767e-05,
+      "loss": 1.087,
+      "step": 12887
+    },
+    {
+      "epoch": 0.7830366364906738,
+      "grad_norm": 0.128581240773201,
+      "learning_rate": 1.1243607671099831e-05,
+      "loss": 1.0141,
+      "step": 12888
+    },
+    {
+      "epoch": 0.7830973935233003,
+      "grad_norm": 0.11357045918703079,
+      "learning_rate": 1.1237560261162954e-05,
+      "loss": 1.0539,
+      "step": 12889
+    },
+    {
+      "epoch": 0.7831581505559269,
+      "grad_norm": 0.12881328165531158,
+      "learning_rate": 1.1231514272070798e-05,
+      "loss": 1.0236,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7832189075885534,
+      "grad_norm": 0.1487404853105545,
+      "learning_rate": 1.1225469704044978e-05,
+      "loss": 1.0815,
+      "step": 12891
+    },
+    {
+      "epoch": 0.7832796646211799,
+      "grad_norm": 0.15983174741268158,
+      "learning_rate": 1.1219426557307061e-05,
+      "loss": 1.0992,
+      "step": 12892
+    },
+    {
+      "epoch": 0.7833404216538065,
+      "grad_norm": 0.10644016414880753,
+      "learning_rate": 1.1213384832078545e-05,
+      "loss": 1.0345,
+      "step": 12893
+    },
+    {
+      "epoch": 0.783401178686433,
+      "grad_norm": 0.13520319759845734,
+      "learning_rate": 1.1207344528580937e-05,
+      "loss": 1.2728,
+      "step": 12894
+    },
+    {
+      "epoch": 0.7834619357190595,
+      "grad_norm": 0.11113191395998001,
+      "learning_rate": 1.1201305647035598e-05,
+      "loss": 1.0074,
+      "step": 12895
+    },
+    {
+      "epoch": 0.783522692751686,
+      "grad_norm": 0.4640451967716217,
+      "learning_rate": 1.1195268187663899e-05,
+      "loss": 1.0263,
+      "step": 12896
+    },
+    {
+      "epoch": 0.7835834497843125,
+      "grad_norm": 0.13769207894802094,
+      "learning_rate": 1.1189232150687145e-05,
+      "loss": 1.0469,
+      "step": 12897
+    },
+    {
+      "epoch": 0.783644206816939,
+      "grad_norm": 0.3489821255207062,
+      "learning_rate": 1.1183197536326573e-05,
+      "loss": 1.1215,
+      "step": 12898
+    },
+    {
+      "epoch": 0.7837049638495656,
+      "grad_norm": 0.10755273699760437,
+      "learning_rate": 1.1177164344803415e-05,
+      "loss": 1.0593,
+      "step": 12899
+    },
+    {
+      "epoch": 0.7837657208821921,
+      "grad_norm": 0.13133902847766876,
+      "learning_rate": 1.11711325763388e-05,
+      "loss": 1.0788,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7838264779148186,
+      "grad_norm": 0.1868750900030136,
+      "learning_rate": 1.1165102231153846e-05,
+      "loss": 1.0827,
+      "step": 12901
+    },
+    {
+      "epoch": 0.7838872349474452,
+      "grad_norm": 0.2032591551542282,
+      "learning_rate": 1.1159073309469553e-05,
+      "loss": 1.1533,
+      "step": 12902
+    },
+    {
+      "epoch": 0.7839479919800717,
+      "grad_norm": 0.16748055815696716,
+      "learning_rate": 1.115304581150693e-05,
+      "loss": 1.1537,
+      "step": 12903
+    },
+    {
+      "epoch": 0.7840087490126982,
+      "grad_norm": 0.12188611924648285,
+      "learning_rate": 1.114701973748693e-05,
+      "loss": 1.0402,
+      "step": 12904
+    },
+    {
+      "epoch": 0.7840695060453248,
+      "grad_norm": 0.13133476674556732,
+      "learning_rate": 1.1140995087630434e-05,
+      "loss": 1.0876,
+      "step": 12905
+    },
+    {
+      "epoch": 0.7841302630779513,
+      "grad_norm": 0.13564378023147583,
+      "learning_rate": 1.1134971862158277e-05,
+      "loss": 1.0694,
+      "step": 12906
+    },
+    {
+      "epoch": 0.7841910201105778,
+      "grad_norm": 0.22386950254440308,
+      "learning_rate": 1.112895006129125e-05,
+      "loss": 1.1008,
+      "step": 12907
+    },
+    {
+      "epoch": 0.7842517771432044,
+      "grad_norm": 0.11448849737644196,
+      "learning_rate": 1.1122929685250045e-05,
+      "loss": 1.0933,
+      "step": 12908
+    },
+    {
+      "epoch": 0.7843125341758308,
+      "grad_norm": 0.16264112293720245,
+      "learning_rate": 1.1116910734255375e-05,
+      "loss": 1.1221,
+      "step": 12909
+    },
+    {
+      "epoch": 0.7843732912084573,
+      "grad_norm": 0.11998185515403748,
+      "learning_rate": 1.1110893208527866e-05,
+      "loss": 1.0744,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7844340482410839,
+      "grad_norm": 0.1568371057510376,
+      "learning_rate": 1.1104877108288082e-05,
+      "loss": 1.0677,
+      "step": 12911
+    },
+    {
+      "epoch": 0.7844948052737104,
+      "grad_norm": 0.13577401638031006,
+      "learning_rate": 1.1098862433756545e-05,
+      "loss": 1.0711,
+      "step": 12912
+    },
+    {
+      "epoch": 0.7845555623063369,
+      "grad_norm": 0.11962083727121353,
+      "learning_rate": 1.1092849185153725e-05,
+      "loss": 1.0296,
+      "step": 12913
+    },
+    {
+      "epoch": 0.7846163193389635,
+      "grad_norm": 0.13312940299510956,
+      "learning_rate": 1.1086837362700036e-05,
+      "loss": 1.0341,
+      "step": 12914
+    },
+    {
+      "epoch": 0.78467707637159,
+      "grad_norm": 0.11990492790937424,
+      "learning_rate": 1.1080826966615848e-05,
+      "loss": 1.0306,
+      "step": 12915
+    },
+    {
+      "epoch": 0.7847378334042165,
+      "grad_norm": 0.13836820423603058,
+      "learning_rate": 1.1074817997121472e-05,
+      "loss": 1.0391,
+      "step": 12916
+    },
+    {
+      "epoch": 0.7847985904368431,
+      "grad_norm": 0.1741345226764679,
+      "learning_rate": 1.1068810454437161e-05,
+      "loss": 1.1001,
+      "step": 12917
+    },
+    {
+      "epoch": 0.7848593474694696,
+      "grad_norm": 0.20143209397792816,
+      "learning_rate": 1.1062804338783134e-05,
+      "loss": 1.1875,
+      "step": 12918
+    },
+    {
+      "epoch": 0.7849201045020961,
+      "grad_norm": 0.14273950457572937,
+      "learning_rate": 1.1056799650379523e-05,
+      "loss": 1.0625,
+      "step": 12919
+    },
+    {
+      "epoch": 0.7849808615347227,
+      "grad_norm": 2.2902894020080566,
+      "learning_rate": 1.1050796389446466e-05,
+      "loss": 1.0467,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7850416185673492,
+      "grad_norm": 0.3513869345188141,
+      "learning_rate": 1.1044794556204009e-05,
+      "loss": 1.1476,
+      "step": 12921
+    },
+    {
+      "epoch": 0.7851023755999756,
+      "grad_norm": 0.1931939572095871,
+      "learning_rate": 1.1038794150872116e-05,
+      "loss": 1.1242,
+      "step": 12922
+    },
+    {
+      "epoch": 0.7851631326326022,
+      "grad_norm": 0.8756475448608398,
+      "learning_rate": 1.1032795173670747e-05,
+      "loss": 1.0923,
+      "step": 12923
+    },
+    {
+      "epoch": 0.7852238896652287,
+      "grad_norm": 0.36362606287002563,
+      "learning_rate": 1.1026797624819818e-05,
+      "loss": 1.1572,
+      "step": 12924
+    },
+    {
+      "epoch": 0.7852846466978552,
+      "grad_norm": 0.131103515625,
+      "learning_rate": 1.1020801504539152e-05,
+      "loss": 1.086,
+      "step": 12925
+    },
+    {
+      "epoch": 0.7853454037304818,
+      "grad_norm": 0.12699051201343536,
+      "learning_rate": 1.1014806813048545e-05,
+      "loss": 1.0769,
+      "step": 12926
+    },
+    {
+      "epoch": 0.7854061607631083,
+      "grad_norm": 0.1372574120759964,
+      "learning_rate": 1.1008813550567725e-05,
+      "loss": 1.0063,
+      "step": 12927
+    },
+    {
+      "epoch": 0.7854669177957349,
+      "grad_norm": 0.6051851511001587,
+      "learning_rate": 1.1002821717316387e-05,
+      "loss": 1.0447,
+      "step": 12928
+    },
+    {
+      "epoch": 0.7855276748283614,
+      "grad_norm": 5.440812110900879,
+      "learning_rate": 1.099683131351415e-05,
+      "loss": 1.071,
+      "step": 12929
+    },
+    {
+      "epoch": 0.7855884318609879,
+      "grad_norm": 0.1737460494041443,
+      "learning_rate": 1.0990842339380608e-05,
+      "loss": 1.1249,
+      "step": 12930
+    },
+    {
+      "epoch": 0.7856491888936145,
+      "grad_norm": 0.31205490231513977,
+      "learning_rate": 1.0984854795135274e-05,
+      "loss": 1.0677,
+      "step": 12931
+    },
+    {
+      "epoch": 0.785709945926241,
+      "grad_norm": 0.18849575519561768,
+      "learning_rate": 1.0978868680997634e-05,
+      "loss": 1.1083,
+      "step": 12932
+    },
+    {
+      "epoch": 0.7857707029588675,
+      "grad_norm": 0.18966080248355865,
+      "learning_rate": 1.09728839971871e-05,
+      "loss": 1.0447,
+      "step": 12933
+    },
+    {
+      "epoch": 0.7858314599914941,
+      "grad_norm": 0.12457823008298874,
+      "learning_rate": 1.0966900743923052e-05,
+      "loss": 1.0785,
+      "step": 12934
+    },
+    {
+      "epoch": 0.7858922170241205,
+      "grad_norm": 0.34969809651374817,
+      "learning_rate": 1.0960918921424801e-05,
+      "loss": 1.0148,
+      "step": 12935
+    },
+    {
+      "epoch": 0.785952974056747,
+      "grad_norm": 0.29265135526657104,
+      "learning_rate": 1.0954938529911608e-05,
+      "loss": 1.069,
+      "step": 12936
+    },
+    {
+      "epoch": 0.7860137310893736,
+      "grad_norm": 0.12106429040431976,
+      "learning_rate": 1.0948959569602695e-05,
+      "loss": 1.0792,
+      "step": 12937
+    },
+    {
+      "epoch": 0.7860744881220001,
+      "grad_norm": 0.241487056016922,
+      "learning_rate": 1.094298204071722e-05,
+      "loss": 1.1267,
+      "step": 12938
+    },
+    {
+      "epoch": 0.7861352451546266,
+      "grad_norm": 0.13090172410011292,
+      "learning_rate": 1.0937005943474266e-05,
+      "loss": 0.9939,
+      "step": 12939
+    },
+    {
+      "epoch": 0.7861960021872532,
+      "grad_norm": 0.10063038021326065,
+      "learning_rate": 1.093103127809293e-05,
+      "loss": 1.0189,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7862567592198797,
+      "grad_norm": 0.10935657471418381,
+      "learning_rate": 1.0925058044792208e-05,
+      "loss": 1.0644,
+      "step": 12941
+    },
+    {
+      "epoch": 0.7863175162525062,
+      "grad_norm": 0.2448442280292511,
+      "learning_rate": 1.091908624379102e-05,
+      "loss": 1.1166,
+      "step": 12942
+    },
+    {
+      "epoch": 0.7863782732851328,
+      "grad_norm": 0.7896729707717896,
+      "learning_rate": 1.0913115875308283e-05,
+      "loss": 1.1464,
+      "step": 12943
+    },
+    {
+      "epoch": 0.7864390303177593,
+      "grad_norm": 0.14838793873786926,
+      "learning_rate": 1.0907146939562818e-05,
+      "loss": 1.0086,
+      "step": 12944
+    },
+    {
+      "epoch": 0.7864997873503858,
+      "grad_norm": 0.10829726606607437,
+      "learning_rate": 1.0901179436773462e-05,
+      "loss": 1.0336,
+      "step": 12945
+    },
+    {
+      "epoch": 0.7865605443830124,
+      "grad_norm": 0.3901865482330322,
+      "learning_rate": 1.0895213367158924e-05,
+      "loss": 1.0696,
+      "step": 12946
+    },
+    {
+      "epoch": 0.7866213014156389,
+      "grad_norm": 0.16181686520576477,
+      "learning_rate": 1.08892487309379e-05,
+      "loss": 1.0718,
+      "step": 12947
+    },
+    {
+      "epoch": 0.7866820584482654,
+      "grad_norm": 0.16711625456809998,
+      "learning_rate": 1.088328552832904e-05,
+      "loss": 1.0335,
+      "step": 12948
+    },
+    {
+      "epoch": 0.7867428154808919,
+      "grad_norm": 0.18678444623947144,
+      "learning_rate": 1.0877323759550884e-05,
+      "loss": 1.0285,
+      "step": 12949
+    },
+    {
+      "epoch": 0.7868035725135184,
+      "grad_norm": 0.1567946821451187,
+      "learning_rate": 1.0871363424821995e-05,
+      "loss": 0.9879,
+      "step": 12950
+    },
+    {
+      "epoch": 0.7868643295461449,
+      "grad_norm": 0.17752927541732788,
+      "learning_rate": 1.086540452436085e-05,
+      "loss": 1.0294,
+      "step": 12951
+    },
+    {
+      "epoch": 0.7869250865787715,
+      "grad_norm": 0.42185690999031067,
+      "learning_rate": 1.0859447058385862e-05,
+      "loss": 1.1973,
+      "step": 12952
+    },
+    {
+      "epoch": 0.786985843611398,
+      "grad_norm": 0.13146072626113892,
+      "learning_rate": 1.085349102711541e-05,
+      "loss": 1.0571,
+      "step": 12953
+    },
+    {
+      "epoch": 0.7870466006440245,
+      "grad_norm": 0.10630039870738983,
+      "learning_rate": 1.0847536430767813e-05,
+      "loss": 0.9831,
+      "step": 12954
+    },
+    {
+      "epoch": 0.7871073576766511,
+      "grad_norm": 0.1909441202878952,
+      "learning_rate": 1.0841583269561334e-05,
+      "loss": 1.0351,
+      "step": 12955
+    },
+    {
+      "epoch": 0.7871681147092776,
+      "grad_norm": 0.1109812930226326,
+      "learning_rate": 1.0835631543714186e-05,
+      "loss": 0.9941,
+      "step": 12956
+    },
+    {
+      "epoch": 0.7872288717419041,
+      "grad_norm": 0.1555769443511963,
+      "learning_rate": 1.082968125344454e-05,
+      "loss": 1.0479,
+      "step": 12957
+    },
+    {
+      "epoch": 0.7872896287745307,
+      "grad_norm": 0.16438855230808258,
+      "learning_rate": 1.0823732398970492e-05,
+      "loss": 1.0779,
+      "step": 12958
+    },
+    {
+      "epoch": 0.7873503858071572,
+      "grad_norm": 0.11183220893144608,
+      "learning_rate": 1.0817784980510115e-05,
+      "loss": 0.9849,
+      "step": 12959
+    },
+    {
+      "epoch": 0.7874111428397838,
+      "grad_norm": 0.13389062881469727,
+      "learning_rate": 1.0811838998281382e-05,
+      "loss": 1.0351,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7874718998724103,
+      "grad_norm": 0.30436673760414124,
+      "learning_rate": 1.0805894452502296e-05,
+      "loss": 1.2856,
+      "step": 12961
+    },
+    {
+      "epoch": 0.7875326569050367,
+      "grad_norm": 0.10991024225950241,
+      "learning_rate": 1.0799951343390714e-05,
+      "loss": 1.0655,
+      "step": 12962
+    },
+    {
+      "epoch": 0.7875934139376632,
+      "grad_norm": 0.15333764255046844,
+      "learning_rate": 1.0794009671164484e-05,
+      "loss": 1.1108,
+      "step": 12963
+    },
+    {
+      "epoch": 0.7876541709702898,
+      "grad_norm": 1.4359647035598755,
+      "learning_rate": 1.0788069436041415e-05,
+      "loss": 1.0512,
+      "step": 12964
+    },
+    {
+      "epoch": 0.7877149280029163,
+      "grad_norm": 0.11067744344472885,
+      "learning_rate": 1.0782130638239224e-05,
+      "loss": 1.0046,
+      "step": 12965
+    },
+    {
+      "epoch": 0.7877756850355428,
+      "grad_norm": 0.23311765491962433,
+      "learning_rate": 1.0776193277975633e-05,
+      "loss": 1.0825,
+      "step": 12966
+    },
+    {
+      "epoch": 0.7878364420681694,
+      "grad_norm": 0.12318284064531326,
+      "learning_rate": 1.0770257355468256e-05,
+      "loss": 1.028,
+      "step": 12967
+    },
+    {
+      "epoch": 0.7878971991007959,
+      "grad_norm": 0.16160030663013458,
+      "learning_rate": 1.076432287093469e-05,
+      "loss": 1.179,
+      "step": 12968
+    },
+    {
+      "epoch": 0.7879579561334225,
+      "grad_norm": 0.3871358633041382,
+      "learning_rate": 1.075838982459244e-05,
+      "loss": 1.1621,
+      "step": 12969
+    },
+    {
+      "epoch": 0.788018713166049,
+      "grad_norm": 0.14440901577472687,
+      "learning_rate": 1.0752458216658984e-05,
+      "loss": 1.0397,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7880794701986755,
+      "grad_norm": 0.11363501846790314,
+      "learning_rate": 1.0746528047351772e-05,
+      "loss": 0.9929,
+      "step": 12971
+    },
+    {
+      "epoch": 0.7881402272313021,
+      "grad_norm": 0.09387975931167603,
+      "learning_rate": 1.0740599316888167e-05,
+      "loss": 1.074,
+      "step": 12972
+    },
+    {
+      "epoch": 0.7882009842639286,
+      "grad_norm": 0.4731582999229431,
+      "learning_rate": 1.0734672025485476e-05,
+      "loss": 1.0937,
+      "step": 12973
+    },
+    {
+      "epoch": 0.7882617412965551,
+      "grad_norm": 0.16211146116256714,
+      "learning_rate": 1.0728746173360987e-05,
+      "loss": 1.1173,
+      "step": 12974
+    },
+    {
+      "epoch": 0.7883224983291816,
+      "grad_norm": 0.15314048528671265,
+      "learning_rate": 1.0722821760731871e-05,
+      "loss": 1.0428,
+      "step": 12975
+    },
+    {
+      "epoch": 0.7883832553618081,
+      "grad_norm": 0.1534699648618698,
+      "learning_rate": 1.0716898787815327e-05,
+      "loss": 1.0606,
+      "step": 12976
+    },
+    {
+      "epoch": 0.7884440123944346,
+      "grad_norm": 0.12805882096290588,
+      "learning_rate": 1.0710977254828458e-05,
+      "loss": 1.0784,
+      "step": 12977
+    },
+    {
+      "epoch": 0.7885047694270612,
+      "grad_norm": 0.11373342573642731,
+      "learning_rate": 1.0705057161988308e-05,
+      "loss": 1.0482,
+      "step": 12978
+    },
+    {
+      "epoch": 0.7885655264596877,
+      "grad_norm": 0.10901042819023132,
+      "learning_rate": 1.069913850951188e-05,
+      "loss": 1.0653,
+      "step": 12979
+    },
+    {
+      "epoch": 0.7886262834923142,
+      "grad_norm": 0.18511372804641724,
+      "learning_rate": 1.069322129761613e-05,
+      "loss": 1.1228,
+      "step": 12980
+    },
+    {
+      "epoch": 0.7886870405249408,
+      "grad_norm": 0.20320042967796326,
+      "learning_rate": 1.0687305526517949e-05,
+      "loss": 1.1332,
+      "step": 12981
+    },
+    {
+      "epoch": 0.7887477975575673,
+      "grad_norm": 0.16992968320846558,
+      "learning_rate": 1.0681391196434187e-05,
+      "loss": 1.1155,
+      "step": 12982
+    },
+    {
+      "epoch": 0.7888085545901938,
+      "grad_norm": 0.21405334770679474,
+      "learning_rate": 1.0675478307581627e-05,
+      "loss": 1.0134,
+      "step": 12983
+    },
+    {
+      "epoch": 0.7888693116228204,
+      "grad_norm": 0.09224823117256165,
+      "learning_rate": 1.066956686017701e-05,
+      "loss": 0.9915,
+      "step": 12984
+    },
+    {
+      "epoch": 0.7889300686554469,
+      "grad_norm": 0.18220220506191254,
+      "learning_rate": 1.0663656854437027e-05,
+      "loss": 1.1035,
+      "step": 12985
+    },
+    {
+      "epoch": 0.7889908256880734,
+      "grad_norm": 0.10819485783576965,
+      "learning_rate": 1.065774829057829e-05,
+      "loss": 1.0231,
+      "step": 12986
+    },
+    {
+      "epoch": 0.7890515827207,
+      "grad_norm": 0.12159349024295807,
+      "learning_rate": 1.0651841168817406e-05,
+      "loss": 1.0249,
+      "step": 12987
+    },
+    {
+      "epoch": 0.7891123397533264,
+      "grad_norm": 0.21518905460834503,
+      "learning_rate": 1.06459354893709e-05,
+      "loss": 1.1489,
+      "step": 12988
+    },
+    {
+      "epoch": 0.7891730967859529,
+      "grad_norm": 0.13885504007339478,
+      "learning_rate": 1.0640031252455229e-05,
+      "loss": 1.0772,
+      "step": 12989
+    },
+    {
+      "epoch": 0.7892338538185795,
+      "grad_norm": 0.11387944966554642,
+      "learning_rate": 1.0634128458286819e-05,
+      "loss": 0.9961,
+      "step": 12990
+    },
+    {
+      "epoch": 0.789294610851206,
+      "grad_norm": 0.7089393138885498,
+      "learning_rate": 1.0628227107082022e-05,
+      "loss": 1.1075,
+      "step": 12991
+    },
+    {
+      "epoch": 0.7893553678838325,
+      "grad_norm": 0.13204385340213776,
+      "learning_rate": 1.0622327199057192e-05,
+      "loss": 1.0834,
+      "step": 12992
+    },
+    {
+      "epoch": 0.7894161249164591,
+      "grad_norm": 0.12831248342990875,
+      "learning_rate": 1.0616428734428562e-05,
+      "loss": 1.0608,
+      "step": 12993
+    },
+    {
+      "epoch": 0.7894768819490856,
+      "grad_norm": 0.25789281725883484,
+      "learning_rate": 1.0610531713412352e-05,
+      "loss": 1.0449,
+      "step": 12994
+    },
+    {
+      "epoch": 0.7895376389817121,
+      "grad_norm": 0.10960721224546432,
+      "learning_rate": 1.0604636136224733e-05,
+      "loss": 1.0431,
+      "step": 12995
+    },
+    {
+      "epoch": 0.7895983960143387,
+      "grad_norm": 0.1594148576259613,
+      "learning_rate": 1.0598742003081762e-05,
+      "loss": 1.1183,
+      "step": 12996
+    },
+    {
+      "epoch": 0.7896591530469652,
+      "grad_norm": 5.922727584838867,
+      "learning_rate": 1.0592849314199532e-05,
+      "loss": 1.0603,
+      "step": 12997
+    },
+    {
+      "epoch": 0.7897199100795917,
+      "grad_norm": 0.1328015923500061,
+      "learning_rate": 1.0586958069794028e-05,
+      "loss": 1.0169,
+      "step": 12998
+    },
+    {
+      "epoch": 0.7897806671122183,
+      "grad_norm": 0.16861291229724884,
+      "learning_rate": 1.058106827008119e-05,
+      "loss": 1.0451,
+      "step": 12999
+    },
+    {
+      "epoch": 0.7898414241448448,
+      "grad_norm": 6.649844169616699,
+      "learning_rate": 1.0575179915276917e-05,
+      "loss": 1.0618,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7899021811774712,
+      "grad_norm": 0.15430712699890137,
+      "learning_rate": 1.0569293005597048e-05,
+      "loss": 1.1052,
+      "step": 13001
+    },
+    {
+      "epoch": 0.7899629382100978,
+      "grad_norm": 0.2725692093372345,
+      "learning_rate": 1.0563407541257364e-05,
+      "loss": 1.1127,
+      "step": 13002
+    },
+    {
+      "epoch": 0.7900236952427243,
+      "grad_norm": 0.19224128127098083,
+      "learning_rate": 1.05575235224736e-05,
+      "loss": 1.0674,
+      "step": 13003
+    },
+    {
+      "epoch": 0.7900844522753508,
+      "grad_norm": 0.13214683532714844,
+      "learning_rate": 1.0551640949461433e-05,
+      "loss": 1.0691,
+      "step": 13004
+    },
+    {
+      "epoch": 0.7901452093079774,
+      "grad_norm": 0.15809671580791473,
+      "learning_rate": 1.0545759822436496e-05,
+      "loss": 1.1539,
+      "step": 13005
+    },
+    {
+      "epoch": 0.7902059663406039,
+      "grad_norm": 0.1760806292295456,
+      "learning_rate": 1.0539880141614355e-05,
+      "loss": 1.0844,
+      "step": 13006
+    },
+    {
+      "epoch": 0.7902667233732305,
+      "grad_norm": 0.15000326931476593,
+      "learning_rate": 1.0534001907210522e-05,
+      "loss": 1.1025,
+      "step": 13007
+    },
+    {
+      "epoch": 0.790327480405857,
+      "grad_norm": 0.11090702563524246,
+      "learning_rate": 1.0528125119440507e-05,
+      "loss": 1.0588,
+      "step": 13008
+    },
+    {
+      "epoch": 0.7903882374384835,
+      "grad_norm": 0.14829276502132416,
+      "learning_rate": 1.0522249778519677e-05,
+      "loss": 1.092,
+      "step": 13009
+    },
+    {
+      "epoch": 0.7904489944711101,
+      "grad_norm": 0.10501495748758316,
+      "learning_rate": 1.051637588466342e-05,
+      "loss": 1.0122,
+      "step": 13010
+    },
+    {
+      "epoch": 0.7905097515037366,
+      "grad_norm": 0.11800161004066467,
+      "learning_rate": 1.0510503438087033e-05,
+      "loss": 1.0314,
+      "step": 13011
+    },
+    {
+      "epoch": 0.7905705085363631,
+      "grad_norm": 0.21349094808101654,
+      "learning_rate": 1.0504632439005762e-05,
+      "loss": 1.0456,
+      "step": 13012
+    },
+    {
+      "epoch": 0.7906312655689897,
+      "grad_norm": 0.35080158710479736,
+      "learning_rate": 1.0498762887634832e-05,
+      "loss": 1.1781,
+      "step": 13013
+    },
+    {
+      "epoch": 0.7906920226016161,
+      "grad_norm": 0.14761504530906677,
+      "learning_rate": 1.0492894784189389e-05,
+      "loss": 0.9978,
+      "step": 13014
+    },
+    {
+      "epoch": 0.7907527796342426,
+      "grad_norm": 0.1098819300532341,
+      "learning_rate": 1.0487028128884541e-05,
+      "loss": 1.0609,
+      "step": 13015
+    },
+    {
+      "epoch": 0.7908135366668692,
+      "grad_norm": 1.724346399307251,
+      "learning_rate": 1.0481162921935283e-05,
+      "loss": 1.0793,
+      "step": 13016
+    },
+    {
+      "epoch": 0.7908742936994957,
+      "grad_norm": 0.11826807260513306,
+      "learning_rate": 1.0475299163556651e-05,
+      "loss": 1.0929,
+      "step": 13017
+    },
+    {
+      "epoch": 0.7909350507321222,
+      "grad_norm": 0.12821528315544128,
+      "learning_rate": 1.0469436853963566e-05,
+      "loss": 1.0892,
+      "step": 13018
+    },
+    {
+      "epoch": 0.7909958077647488,
+      "grad_norm": 0.2582976222038269,
+      "learning_rate": 1.0463575993370917e-05,
+      "loss": 1.1234,
+      "step": 13019
+    },
+    {
+      "epoch": 0.7910565647973753,
+      "grad_norm": 0.1912316530942917,
+      "learning_rate": 1.0457716581993531e-05,
+      "loss": 1.1253,
+      "step": 13020
+    },
+    {
+      "epoch": 0.7911173218300018,
+      "grad_norm": 0.10567454993724823,
+      "learning_rate": 1.045185862004619e-05,
+      "loss": 1.0588,
+      "step": 13021
+    },
+    {
+      "epoch": 0.7911780788626284,
+      "grad_norm": 0.17749686539173126,
+      "learning_rate": 1.0446002107743612e-05,
+      "loss": 1.0633,
+      "step": 13022
+    },
+    {
+      "epoch": 0.7912388358952549,
+      "grad_norm": 0.16335220634937286,
+      "learning_rate": 1.0440147045300474e-05,
+      "loss": 1.055,
+      "step": 13023
+    },
+    {
+      "epoch": 0.7912995929278814,
+      "grad_norm": 0.14854291081428528,
+      "learning_rate": 1.0434293432931398e-05,
+      "loss": 1.0141,
+      "step": 13024
+    },
+    {
+      "epoch": 0.791360349960508,
+      "grad_norm": 0.21400520205497742,
+      "learning_rate": 1.0428441270850935e-05,
+      "loss": 1.1369,
+      "step": 13025
+    },
+    {
+      "epoch": 0.7914211069931345,
+      "grad_norm": 0.09609528630971909,
+      "learning_rate": 1.0422590559273609e-05,
+      "loss": 1.0386,
+      "step": 13026
+    },
+    {
+      "epoch": 0.7914818640257609,
+      "grad_norm": 1.0455647706985474,
+      "learning_rate": 1.0416741298413878e-05,
+      "loss": 1.1828,
+      "step": 13027
+    },
+    {
+      "epoch": 0.7915426210583875,
+      "grad_norm": 0.19983692467212677,
+      "learning_rate": 1.0410893488486145e-05,
+      "loss": 1.1282,
+      "step": 13028
+    },
+    {
+      "epoch": 0.791603378091014,
+      "grad_norm": 4.148086071014404,
+      "learning_rate": 1.0405047129704764e-05,
+      "loss": 1.2251,
+      "step": 13029
+    },
+    {
+      "epoch": 0.7916641351236405,
+      "grad_norm": 0.18747232854366302,
+      "learning_rate": 1.0399202222284033e-05,
+      "loss": 1.0936,
+      "step": 13030
+    },
+    {
+      "epoch": 0.7917248921562671,
+      "grad_norm": 0.1455703228712082,
+      "learning_rate": 1.0393358766438199e-05,
+      "loss": 1.0376,
+      "step": 13031
+    },
+    {
+      "epoch": 0.7917856491888936,
+      "grad_norm": 0.14367344975471497,
+      "learning_rate": 1.0387516762381444e-05,
+      "loss": 1.0534,
+      "step": 13032
+    },
+    {
+      "epoch": 0.7918464062215201,
+      "grad_norm": 0.13840726017951965,
+      "learning_rate": 1.0381676210327934e-05,
+      "loss": 1.0164,
+      "step": 13033
+    },
+    {
+      "epoch": 0.7919071632541467,
+      "grad_norm": 0.11749805510044098,
+      "learning_rate": 1.0375837110491737e-05,
+      "loss": 1.0384,
+      "step": 13034
+    },
+    {
+      "epoch": 0.7919679202867732,
+      "grad_norm": 0.1618070900440216,
+      "learning_rate": 1.0369999463086905e-05,
+      "loss": 1.0941,
+      "step": 13035
+    },
+    {
+      "epoch": 0.7920286773193997,
+      "grad_norm": 0.21264484524726868,
+      "learning_rate": 1.0364163268327387e-05,
+      "loss": 1.1066,
+      "step": 13036
+    },
+    {
+      "epoch": 0.7920894343520263,
+      "grad_norm": 0.1712161749601364,
+      "learning_rate": 1.0358328526427113e-05,
+      "loss": 1.0919,
+      "step": 13037
+    },
+    {
+      "epoch": 0.7921501913846528,
+      "grad_norm": 0.10082321614027023,
+      "learning_rate": 1.0352495237599985e-05,
+      "loss": 1.0525,
+      "step": 13038
+    },
+    {
+      "epoch": 0.7922109484172793,
+      "grad_norm": 0.27976882457733154,
+      "learning_rate": 1.0346663402059809e-05,
+      "loss": 1.2963,
+      "step": 13039
+    },
+    {
+      "epoch": 0.7922717054499058,
+      "grad_norm": 0.12554937601089478,
+      "learning_rate": 1.0340833020020347e-05,
+      "loss": 1.0778,
+      "step": 13040
+    },
+    {
+      "epoch": 0.7923324624825323,
+      "grad_norm": 0.2319071739912033,
+      "learning_rate": 1.0335004091695333e-05,
+      "loss": 1.0929,
+      "step": 13041
+    },
+    {
+      "epoch": 0.7923932195151588,
+      "grad_norm": 0.09979699552059174,
+      "learning_rate": 1.0329176617298381e-05,
+      "loss": 1.0091,
+      "step": 13042
+    },
+    {
+      "epoch": 0.7924539765477854,
+      "grad_norm": 0.20330984890460968,
+      "learning_rate": 1.0323350597043141e-05,
+      "loss": 1.143,
+      "step": 13043
+    },
+    {
+      "epoch": 0.7925147335804119,
+      "grad_norm": 0.6742188334465027,
+      "learning_rate": 1.031752603114316e-05,
+      "loss": 1.0548,
+      "step": 13044
+    },
+    {
+      "epoch": 0.7925754906130384,
+      "grad_norm": 1.9805086851119995,
+      "learning_rate": 1.0311702919811928e-05,
+      "loss": 1.1302,
+      "step": 13045
+    },
+    {
+      "epoch": 0.792636247645665,
+      "grad_norm": 0.131484255194664,
+      "learning_rate": 1.03058812632629e-05,
+      "loss": 1.0748,
+      "step": 13046
+    },
+    {
+      "epoch": 0.7926970046782915,
+      "grad_norm": 0.16047367453575134,
+      "learning_rate": 1.0300061061709465e-05,
+      "loss": 1.1203,
+      "step": 13047
+    },
+    {
+      "epoch": 0.792757761710918,
+      "grad_norm": 0.14155474305152893,
+      "learning_rate": 1.0294242315364966e-05,
+      "loss": 1.0415,
+      "step": 13048
+    },
+    {
+      "epoch": 0.7928185187435446,
+      "grad_norm": 0.29167142510414124,
+      "learning_rate": 1.0288425024442689e-05,
+      "loss": 0.9946,
+      "step": 13049
+    },
+    {
+      "epoch": 0.7928792757761711,
+      "grad_norm": 0.10048561543226242,
+      "learning_rate": 1.0282609189155867e-05,
+      "loss": 1.0097,
+      "step": 13050
+    },
+    {
+      "epoch": 0.7929400328087977,
+      "grad_norm": 0.23381289839744568,
+      "learning_rate": 1.0276794809717688e-05,
+      "loss": 1.1833,
+      "step": 13051
+    },
+    {
+      "epoch": 0.7930007898414242,
+      "grad_norm": 0.16975292563438416,
+      "learning_rate": 1.0270981886341268e-05,
+      "loss": 1.0282,
+      "step": 13052
+    },
+    {
+      "epoch": 0.7930615468740507,
+      "grad_norm": 0.12489912658929825,
+      "learning_rate": 1.0265170419239678e-05,
+      "loss": 1.0749,
+      "step": 13053
+    },
+    {
+      "epoch": 0.7931223039066772,
+      "grad_norm": 0.17183059453964233,
+      "learning_rate": 1.0259360408625957e-05,
+      "loss": 1.0156,
+      "step": 13054
+    },
+    {
+      "epoch": 0.7931830609393037,
+      "grad_norm": 0.1626148819923401,
+      "learning_rate": 1.0253551854713079e-05,
+      "loss": 1.0509,
+      "step": 13055
+    },
+    {
+      "epoch": 0.7932438179719302,
+      "grad_norm": 0.1180085688829422,
+      "learning_rate": 1.0247744757713923e-05,
+      "loss": 1.0849,
+      "step": 13056
+    },
+    {
+      "epoch": 0.7933045750045568,
+      "grad_norm": 0.15574371814727783,
+      "learning_rate": 1.0241939117841371e-05,
+      "loss": 1.0382,
+      "step": 13057
+    },
+    {
+      "epoch": 0.7933653320371833,
+      "grad_norm": 0.12374011427164078,
+      "learning_rate": 1.0236134935308205e-05,
+      "loss": 1.0393,
+      "step": 13058
+    },
+    {
+      "epoch": 0.7934260890698098,
+      "grad_norm": 0.15752482414245605,
+      "learning_rate": 1.023033221032722e-05,
+      "loss": 1.1205,
+      "step": 13059
+    },
+    {
+      "epoch": 0.7934868461024364,
+      "grad_norm": 0.12764738500118256,
+      "learning_rate": 1.0224530943111094e-05,
+      "loss": 1.0722,
+      "step": 13060
+    },
+    {
+      "epoch": 0.7935476031350629,
+      "grad_norm": 0.10181684792041779,
+      "learning_rate": 1.0218731133872484e-05,
+      "loss": 1.045,
+      "step": 13061
+    },
+    {
+      "epoch": 0.7936083601676894,
+      "grad_norm": 0.10471001267433167,
+      "learning_rate": 1.0212932782823965e-05,
+      "loss": 1.0592,
+      "step": 13062
+    },
+    {
+      "epoch": 0.793669117200316,
+      "grad_norm": 0.38642391562461853,
+      "learning_rate": 1.0207135890178066e-05,
+      "loss": 1.0435,
+      "step": 13063
+    },
+    {
+      "epoch": 0.7937298742329425,
+      "grad_norm": 0.12425728142261505,
+      "learning_rate": 1.0201340456147313e-05,
+      "loss": 1.0719,
+      "step": 13064
+    },
+    {
+      "epoch": 0.793790631265569,
+      "grad_norm": 0.13001494109630585,
+      "learning_rate": 1.0195546480944113e-05,
+      "loss": 1.0305,
+      "step": 13065
+    },
+    {
+      "epoch": 0.7938513882981956,
+      "grad_norm": 0.12355656176805496,
+      "learning_rate": 1.0189753964780857e-05,
+      "loss": 1.0618,
+      "step": 13066
+    },
+    {
+      "epoch": 0.793912145330822,
+      "grad_norm": 0.12111663073301315,
+      "learning_rate": 1.018396290786986e-05,
+      "loss": 1.0489,
+      "step": 13067
+    },
+    {
+      "epoch": 0.7939729023634485,
+      "grad_norm": 0.11895778030157089,
+      "learning_rate": 1.01781733104234e-05,
+      "loss": 1.0285,
+      "step": 13068
+    },
+    {
+      "epoch": 0.7940336593960751,
+      "grad_norm": 0.09712397307157516,
+      "learning_rate": 1.0172385172653692e-05,
+      "loss": 1.0307,
+      "step": 13069
+    },
+    {
+      "epoch": 0.7940944164287016,
+      "grad_norm": 0.14626674354076385,
+      "learning_rate": 1.0166598494772905e-05,
+      "loss": 1.0442,
+      "step": 13070
+    },
+    {
+      "epoch": 0.7941551734613281,
+      "grad_norm": 0.12935839593410492,
+      "learning_rate": 1.0160813276993153e-05,
+      "loss": 1.0446,
+      "step": 13071
+    },
+    {
+      "epoch": 0.7942159304939547,
+      "grad_norm": 0.181484192609787,
+      "learning_rate": 1.0155029519526487e-05,
+      "loss": 1.1431,
+      "step": 13072
+    },
+    {
+      "epoch": 0.7942766875265812,
+      "grad_norm": 0.13355356454849243,
+      "learning_rate": 1.0149247222584923e-05,
+      "loss": 1.0258,
+      "step": 13073
+    },
+    {
+      "epoch": 0.7943374445592077,
+      "grad_norm": 0.13636019825935364,
+      "learning_rate": 1.0143466386380385e-05,
+      "loss": 1.0626,
+      "step": 13074
+    },
+    {
+      "epoch": 0.7943982015918343,
+      "grad_norm": 0.10556565225124359,
+      "learning_rate": 1.013768701112482e-05,
+      "loss": 1.0374,
+      "step": 13075
+    },
+    {
+      "epoch": 0.7944589586244608,
+      "grad_norm": 0.31504783034324646,
+      "learning_rate": 1.013190909703003e-05,
+      "loss": 1.0132,
+      "step": 13076
+    },
+    {
+      "epoch": 0.7945197156570873,
+      "grad_norm": 0.10009024292230606,
+      "learning_rate": 1.0126132644307817e-05,
+      "loss": 1.0124,
+      "step": 13077
+    },
+    {
+      "epoch": 0.7945804726897139,
+      "grad_norm": 0.1064009815454483,
+      "learning_rate": 1.0120357653169921e-05,
+      "loss": 1.0205,
+      "step": 13078
+    },
+    {
+      "epoch": 0.7946412297223404,
+      "grad_norm": 0.11540800333023071,
+      "learning_rate": 1.0114584123828008e-05,
+      "loss": 1.023,
+      "step": 13079
+    },
+    {
+      "epoch": 0.7947019867549668,
+      "grad_norm": 0.15264244377613068,
+      "learning_rate": 1.0108812056493738e-05,
+      "loss": 1.0702,
+      "step": 13080
+    },
+    {
+      "epoch": 0.7947627437875934,
+      "grad_norm": 0.16647522151470184,
+      "learning_rate": 1.010304145137867e-05,
+      "loss": 1.1173,
+      "step": 13081
+    },
+    {
+      "epoch": 0.7948235008202199,
+      "grad_norm": 0.3319341838359833,
+      "learning_rate": 1.0097272308694345e-05,
+      "loss": 1.273,
+      "step": 13082
+    },
+    {
+      "epoch": 0.7948842578528464,
+      "grad_norm": 0.12956970930099487,
+      "learning_rate": 1.0091504628652204e-05,
+      "loss": 1.0449,
+      "step": 13083
+    },
+    {
+      "epoch": 0.794945014885473,
+      "grad_norm": 0.5735976099967957,
+      "learning_rate": 1.0085738411463664e-05,
+      "loss": 0.9946,
+      "step": 13084
+    },
+    {
+      "epoch": 0.7950057719180995,
+      "grad_norm": 0.10754632204771042,
+      "learning_rate": 1.0079973657340108e-05,
+      "loss": 1.0732,
+      "step": 13085
+    },
+    {
+      "epoch": 0.795066528950726,
+      "grad_norm": 0.31312495470046997,
+      "learning_rate": 1.0074210366492831e-05,
+      "loss": 1.1918,
+      "step": 13086
+    },
+    {
+      "epoch": 0.7951272859833526,
+      "grad_norm": 0.12384498864412308,
+      "learning_rate": 1.0068448539133096e-05,
+      "loss": 1.0675,
+      "step": 13087
+    },
+    {
+      "epoch": 0.7951880430159791,
+      "grad_norm": 0.2554831802845001,
+      "learning_rate": 1.0062688175472113e-05,
+      "loss": 1.1105,
+      "step": 13088
+    },
+    {
+      "epoch": 0.7952488000486057,
+      "grad_norm": 0.15807794034481049,
+      "learning_rate": 1.0056929275720982e-05,
+      "loss": 1.0198,
+      "step": 13089
+    },
+    {
+      "epoch": 0.7953095570812322,
+      "grad_norm": 0.1424776017665863,
+      "learning_rate": 1.0051171840090845e-05,
+      "loss": 1.0663,
+      "step": 13090
+    },
+    {
+      "epoch": 0.7953703141138587,
+      "grad_norm": 0.12099411338567734,
+      "learning_rate": 1.0045415868792729e-05,
+      "loss": 1.0547,
+      "step": 13091
+    },
+    {
+      "epoch": 0.7954310711464853,
+      "grad_norm": 0.1377709060907364,
+      "learning_rate": 1.003966136203761e-05,
+      "loss": 1.0447,
+      "step": 13092
+    },
+    {
+      "epoch": 0.7954918281791117,
+      "grad_norm": 0.1375075876712799,
+      "learning_rate": 1.0033908320036433e-05,
+      "loss": 1.0632,
+      "step": 13093
+    },
+    {
+      "epoch": 0.7955525852117382,
+      "grad_norm": 0.1367979496717453,
+      "learning_rate": 1.0028156743000072e-05,
+      "loss": 1.0598,
+      "step": 13094
+    },
+    {
+      "epoch": 0.7956133422443648,
+      "grad_norm": 0.13092826306819916,
+      "learning_rate": 1.0022406631139352e-05,
+      "loss": 1.0403,
+      "step": 13095
+    },
+    {
+      "epoch": 0.7956740992769913,
+      "grad_norm": 0.15822453796863556,
+      "learning_rate": 1.0016657984665045e-05,
+      "loss": 1.112,
+      "step": 13096
+    },
+    {
+      "epoch": 0.7957348563096178,
+      "grad_norm": 0.17074939608573914,
+      "learning_rate": 1.0010910803787865e-05,
+      "loss": 1.1617,
+      "step": 13097
+    },
+    {
+      "epoch": 0.7957956133422444,
+      "grad_norm": 0.17768992483615875,
+      "learning_rate": 1.0005165088718483e-05,
+      "loss": 1.0631,
+      "step": 13098
+    },
+    {
+      "epoch": 0.7958563703748709,
+      "grad_norm": 0.13596971333026886,
+      "learning_rate": 9.999420839667505e-06,
+      "loss": 1.0526,
+      "step": 13099
+    },
+    {
+      "epoch": 0.7959171274074974,
+      "grad_norm": 0.1868571788072586,
+      "learning_rate": 9.993678056845474e-06,
+      "loss": 1.1884,
+      "step": 13100
+    },
+    {
+      "epoch": 0.795977884440124,
+      "grad_norm": 0.13141293823719025,
+      "learning_rate": 9.987936740462923e-06,
+      "loss": 1.0669,
+      "step": 13101
+    },
+    {
+      "epoch": 0.7960386414727505,
+      "grad_norm": 0.15654785931110382,
+      "learning_rate": 9.982196890730294e-06,
+      "loss": 1.1035,
+      "step": 13102
+    },
+    {
+      "epoch": 0.796099398505377,
+      "grad_norm": 0.15711770951747894,
+      "learning_rate": 9.97645850785796e-06,
+      "loss": 1.0229,
+      "step": 13103
+    },
+    {
+      "epoch": 0.7961601555380036,
+      "grad_norm": 0.3458845913410187,
+      "learning_rate": 9.970721592056276e-06,
+      "loss": 1.1981,
+      "step": 13104
+    },
+    {
+      "epoch": 0.7962209125706301,
+      "grad_norm": 0.2297106832265854,
+      "learning_rate": 9.964986143535515e-06,
+      "loss": 1.0576,
+      "step": 13105
+    },
+    {
+      "epoch": 0.7962816696032565,
+      "grad_norm": 0.14223851263523102,
+      "learning_rate": 9.959252162505944e-06,
+      "loss": 1.1083,
+      "step": 13106
+    },
+    {
+      "epoch": 0.7963424266358831,
+      "grad_norm": 0.2293814718723297,
+      "learning_rate": 9.953519649177717e-06,
+      "loss": 1.0875,
+      "step": 13107
+    },
+    {
+      "epoch": 0.7964031836685096,
+      "grad_norm": 0.13752363622188568,
+      "learning_rate": 9.947788603760987e-06,
+      "loss": 1.0701,
+      "step": 13108
+    },
+    {
+      "epoch": 0.7964639407011361,
+      "grad_norm": 0.12873654067516327,
+      "learning_rate": 9.942059026465794e-06,
+      "loss": 1.0256,
+      "step": 13109
+    },
+    {
+      "epoch": 0.7965246977337627,
+      "grad_norm": 0.13620485365390778,
+      "learning_rate": 9.936330917502157e-06,
+      "loss": 1.0144,
+      "step": 13110
+    },
+    {
+      "epoch": 0.7965854547663892,
+      "grad_norm": 0.1412217766046524,
+      "learning_rate": 9.93060427708007e-06,
+      "loss": 1.057,
+      "step": 13111
+    },
+    {
+      "epoch": 0.7966462117990157,
+      "grad_norm": 0.21335838735103607,
+      "learning_rate": 9.924879105409423e-06,
+      "loss": 1.0967,
+      "step": 13112
+    },
+    {
+      "epoch": 0.7967069688316423,
+      "grad_norm": 0.20370370149612427,
+      "learning_rate": 9.919155402700082e-06,
+      "loss": 1.087,
+      "step": 13113
+    },
+    {
+      "epoch": 0.7967677258642688,
+      "grad_norm": 0.1506512314081192,
+      "learning_rate": 9.913433169161845e-06,
+      "loss": 1.0187,
+      "step": 13114
+    },
+    {
+      "epoch": 0.7968284828968953,
+      "grad_norm": 0.19056439399719238,
+      "learning_rate": 9.907712405004466e-06,
+      "loss": 1.1814,
+      "step": 13115
+    },
+    {
+      "epoch": 0.7968892399295219,
+      "grad_norm": 0.12057289481163025,
+      "learning_rate": 9.901993110437635e-06,
+      "loss": 1.0754,
+      "step": 13116
+    },
+    {
+      "epoch": 0.7969499969621484,
+      "grad_norm": 0.12495239824056625,
+      "learning_rate": 9.896275285671003e-06,
+      "loss": 1.0486,
+      "step": 13117
+    },
+    {
+      "epoch": 0.797010753994775,
+      "grad_norm": 0.1455259621143341,
+      "learning_rate": 9.890558930914146e-06,
+      "loss": 1.0627,
+      "step": 13118
+    },
+    {
+      "epoch": 0.7970715110274014,
+      "grad_norm": 0.10583961755037308,
+      "learning_rate": 9.884844046376607e-06,
+      "loss": 1.0394,
+      "step": 13119
+    },
+    {
+      "epoch": 0.7971322680600279,
+      "grad_norm": 0.12401382625102997,
+      "learning_rate": 9.879130632267864e-06,
+      "loss": 1.0273,
+      "step": 13120
+    },
+    {
+      "epoch": 0.7971930250926544,
+      "grad_norm": 0.30011144280433655,
+      "learning_rate": 9.873418688797332e-06,
+      "loss": 1.0685,
+      "step": 13121
+    },
+    {
+      "epoch": 0.797253782125281,
+      "grad_norm": 0.4915805459022522,
+      "learning_rate": 9.86770821617442e-06,
+      "loss": 1.1651,
+      "step": 13122
+    },
+    {
+      "epoch": 0.7973145391579075,
+      "grad_norm": 0.17420677840709686,
+      "learning_rate": 9.861999214608403e-06,
+      "loss": 1.0862,
+      "step": 13123
+    },
+    {
+      "epoch": 0.797375296190534,
+      "grad_norm": 0.12418276816606522,
+      "learning_rate": 9.856291684308566e-06,
+      "loss": 1.0232,
+      "step": 13124
+    },
+    {
+      "epoch": 0.7974360532231606,
+      "grad_norm": 0.4045256972312927,
+      "learning_rate": 9.850585625484105e-06,
+      "loss": 1.1388,
+      "step": 13125
+    },
+    {
+      "epoch": 0.7974968102557871,
+      "grad_norm": 0.12089885771274567,
+      "learning_rate": 9.8448810383442e-06,
+      "loss": 1.0539,
+      "step": 13126
+    },
+    {
+      "epoch": 0.7975575672884136,
+      "grad_norm": 0.1419459581375122,
+      "learning_rate": 9.839177923097942e-06,
+      "loss": 1.1091,
+      "step": 13127
+    },
+    {
+      "epoch": 0.7976183243210402,
+      "grad_norm": 0.11342260241508484,
+      "learning_rate": 9.833476279954395e-06,
+      "loss": 1.0748,
+      "step": 13128
+    },
+    {
+      "epoch": 0.7976790813536667,
+      "grad_norm": 0.24692946672439575,
+      "learning_rate": 9.827776109122526e-06,
+      "loss": 1.1235,
+      "step": 13129
+    },
+    {
+      "epoch": 0.7977398383862933,
+      "grad_norm": 0.13997912406921387,
+      "learning_rate": 9.822077410811275e-06,
+      "loss": 1.0214,
+      "step": 13130
+    },
+    {
+      "epoch": 0.7978005954189198,
+      "grad_norm": 0.2458542138338089,
+      "learning_rate": 9.816380185229556e-06,
+      "loss": 1.2319,
+      "step": 13131
+    },
+    {
+      "epoch": 0.7978613524515462,
+      "grad_norm": 0.18516439199447632,
+      "learning_rate": 9.810684432586188e-06,
+      "loss": 1.115,
+      "step": 13132
+    },
+    {
+      "epoch": 0.7979221094841727,
+      "grad_norm": 0.18086761236190796,
+      "learning_rate": 9.80499015308995e-06,
+      "loss": 1.0498,
+      "step": 13133
+    },
+    {
+      "epoch": 0.7979828665167993,
+      "grad_norm": 0.16141141951084137,
+      "learning_rate": 9.79929734694957e-06,
+      "loss": 1.117,
+      "step": 13134
+    },
+    {
+      "epoch": 0.7980436235494258,
+      "grad_norm": 0.1359003484249115,
+      "learning_rate": 9.793606014373712e-06,
+      "loss": 1.0516,
+      "step": 13135
+    },
+    {
+      "epoch": 0.7981043805820524,
+      "grad_norm": 0.12042230367660522,
+      "learning_rate": 9.787916155571003e-06,
+      "loss": 1.0276,
+      "step": 13136
+    },
+    {
+      "epoch": 0.7981651376146789,
+      "grad_norm": 0.8082221746444702,
+      "learning_rate": 9.782227770749997e-06,
+      "loss": 1.2311,
+      "step": 13137
+    },
+    {
+      "epoch": 0.7982258946473054,
+      "grad_norm": 0.11634038388729095,
+      "learning_rate": 9.776540860119205e-06,
+      "loss": 1.0583,
+      "step": 13138
+    },
+    {
+      "epoch": 0.798286651679932,
+      "grad_norm": 0.12310948222875595,
+      "learning_rate": 9.770855423887088e-06,
+      "loss": 1.0488,
+      "step": 13139
+    },
+    {
+      "epoch": 0.7983474087125585,
+      "grad_norm": 0.13864214718341827,
+      "learning_rate": 9.765171462262036e-06,
+      "loss": 1.0353,
+      "step": 13140
+    },
+    {
+      "epoch": 0.798408165745185,
+      "grad_norm": 5.499795436859131,
+      "learning_rate": 9.759488975452386e-06,
+      "loss": 1.063,
+      "step": 13141
+    },
+    {
+      "epoch": 0.7984689227778116,
+      "grad_norm": 0.11382286250591278,
+      "learning_rate": 9.753807963666478e-06,
+      "loss": 1.0494,
+      "step": 13142
+    },
+    {
+      "epoch": 0.7985296798104381,
+      "grad_norm": 0.1148020401597023,
+      "learning_rate": 9.748128427112502e-06,
+      "loss": 1.0672,
+      "step": 13143
+    },
+    {
+      "epoch": 0.7985904368430646,
+      "grad_norm": 0.12468332052230835,
+      "learning_rate": 9.742450365998663e-06,
+      "loss": 1.0281,
+      "step": 13144
+    },
+    {
+      "epoch": 0.7986511938756911,
+      "grad_norm": 0.1946880966424942,
+      "learning_rate": 9.736773780533081e-06,
+      "loss": 1.1204,
+      "step": 13145
+    },
+    {
+      "epoch": 0.7987119509083176,
+      "grad_norm": 0.27400967478752136,
+      "learning_rate": 9.731098670923832e-06,
+      "loss": 1.1401,
+      "step": 13146
+    },
+    {
+      "epoch": 0.7987727079409441,
+      "grad_norm": 0.2063048630952835,
+      "learning_rate": 9.725425037378954e-06,
+      "loss": 1.0165,
+      "step": 13147
+    },
+    {
+      "epoch": 0.7988334649735707,
+      "grad_norm": 0.16999012231826782,
+      "learning_rate": 9.71975288010641e-06,
+      "loss": 1.1012,
+      "step": 13148
+    },
+    {
+      "epoch": 0.7988942220061972,
+      "grad_norm": 0.13342589139938354,
+      "learning_rate": 9.714082199314118e-06,
+      "loss": 1.0646,
+      "step": 13149
+    },
+    {
+      "epoch": 0.7989549790388237,
+      "grad_norm": 0.11692260950803757,
+      "learning_rate": 9.708412995209921e-06,
+      "loss": 1.051,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7990157360714503,
+      "grad_norm": 0.3315751552581787,
+      "learning_rate": 9.702745268001617e-06,
+      "loss": 1.2028,
+      "step": 13151
+    },
+    {
+      "epoch": 0.7990764931040768,
+      "grad_norm": 0.1433667093515396,
+      "learning_rate": 9.697079017896993e-06,
+      "loss": 1.0102,
+      "step": 13152
+    },
+    {
+      "epoch": 0.7991372501367033,
+      "grad_norm": 0.1572551429271698,
+      "learning_rate": 9.691414245103725e-06,
+      "loss": 1.1013,
+      "step": 13153
+    },
+    {
+      "epoch": 0.7991980071693299,
+      "grad_norm": 0.340774267911911,
+      "learning_rate": 9.685750949829458e-06,
+      "loss": 1.1231,
+      "step": 13154
+    },
+    {
+      "epoch": 0.7992587642019564,
+      "grad_norm": 0.1548185497522354,
+      "learning_rate": 9.6800891322818e-06,
+      "loss": 1.0659,
+      "step": 13155
+    },
+    {
+      "epoch": 0.7993195212345829,
+      "grad_norm": 0.15976443886756897,
+      "learning_rate": 9.674428792668238e-06,
+      "loss": 1.0074,
+      "step": 13156
+    },
+    {
+      "epoch": 0.7993802782672095,
+      "grad_norm": 0.13568075001239777,
+      "learning_rate": 9.668769931196297e-06,
+      "loss": 1.0429,
+      "step": 13157
+    },
+    {
+      "epoch": 0.799441035299836,
+      "grad_norm": 0.14354725182056427,
+      "learning_rate": 9.663112548073394e-06,
+      "loss": 1.0901,
+      "step": 13158
+    },
+    {
+      "epoch": 0.7995017923324624,
+      "grad_norm": 0.1283273547887802,
+      "learning_rate": 9.65745664350689e-06,
+      "loss": 1.0648,
+      "step": 13159
+    },
+    {
+      "epoch": 0.799562549365089,
+      "grad_norm": 0.2266400009393692,
+      "learning_rate": 9.65180221770412e-06,
+      "loss": 1.1014,
+      "step": 13160
+    },
+    {
+      "epoch": 0.7996233063977155,
+      "grad_norm": 0.11634045839309692,
+      "learning_rate": 9.646149270872334e-06,
+      "loss": 1.0607,
+      "step": 13161
+    },
+    {
+      "epoch": 0.799684063430342,
+      "grad_norm": 0.2146165668964386,
+      "learning_rate": 9.640497803218745e-06,
+      "loss": 1.0938,
+      "step": 13162
+    },
+    {
+      "epoch": 0.7997448204629686,
+      "grad_norm": 0.19099931418895721,
+      "learning_rate": 9.634847814950509e-06,
+      "loss": 1.152,
+      "step": 13163
+    },
+    {
+      "epoch": 0.7998055774955951,
+      "grad_norm": 0.11193753778934479,
+      "learning_rate": 9.629199306274733e-06,
+      "loss": 1.0193,
+      "step": 13164
+    },
+    {
+      "epoch": 0.7998663345282216,
+      "grad_norm": 0.16398470103740692,
+      "learning_rate": 9.623552277398461e-06,
+      "loss": 1.0807,
+      "step": 13165
+    },
+    {
+      "epoch": 0.7999270915608482,
+      "grad_norm": 15.153077125549316,
+      "learning_rate": 9.61790672852868e-06,
+      "loss": 1.0093,
+      "step": 13166
+    },
+    {
+      "epoch": 0.7999878485934747,
+      "grad_norm": 0.11747323721647263,
+      "learning_rate": 9.612262659872322e-06,
+      "loss": 1.0322,
+      "step": 13167
+    },
+    {
+      "epoch": 0.8000486056261013,
+      "grad_norm": 0.1298796683549881,
+      "learning_rate": 9.606620071636302e-06,
+      "loss": 1.0672,
+      "step": 13168
+    },
+    {
+      "epoch": 0.8001093626587278,
+      "grad_norm": 2.933661937713623,
+      "learning_rate": 9.600978964027435e-06,
+      "loss": 0.9854,
+      "step": 13169
+    },
+    {
+      "epoch": 0.8001701196913543,
+      "grad_norm": 5.170802116394043,
+      "learning_rate": 9.595339337252485e-06,
+      "loss": 1.0516,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8002308767239809,
+      "grad_norm": 0.5858650803565979,
+      "learning_rate": 9.589701191518186e-06,
+      "loss": 1.4054,
+      "step": 13171
+    },
+    {
+      "epoch": 0.8002916337566073,
+      "grad_norm": 0.18030177056789398,
+      "learning_rate": 9.584064527031178e-06,
+      "loss": 1.0384,
+      "step": 13172
+    },
+    {
+      "epoch": 0.8003523907892338,
+      "grad_norm": 0.555079996585846,
+      "learning_rate": 9.578429343998119e-06,
+      "loss": 1.4273,
+      "step": 13173
+    },
+    {
+      "epoch": 0.8004131478218603,
+      "grad_norm": 0.1230132132768631,
+      "learning_rate": 9.572795642625537e-06,
+      "loss": 1.0499,
+      "step": 13174
+    },
+    {
+      "epoch": 0.8004739048544869,
+      "grad_norm": 0.1732676774263382,
+      "learning_rate": 9.567163423119968e-06,
+      "loss": 1.1516,
+      "step": 13175
+    },
+    {
+      "epoch": 0.8005346618871134,
+      "grad_norm": 0.2401278018951416,
+      "learning_rate": 9.561532685687818e-06,
+      "loss": 1.0172,
+      "step": 13176
+    },
+    {
+      "epoch": 0.80059541891974,
+      "grad_norm": 0.9617950320243835,
+      "learning_rate": 9.555903430535496e-06,
+      "loss": 1.1345,
+      "step": 13177
+    },
+    {
+      "epoch": 0.8006561759523665,
+      "grad_norm": 0.14101830124855042,
+      "learning_rate": 9.550275657869362e-06,
+      "loss": 1.0538,
+      "step": 13178
+    },
+    {
+      "epoch": 0.800716932984993,
+      "grad_norm": 0.12207241356372833,
+      "learning_rate": 9.544649367895692e-06,
+      "loss": 1.0581,
+      "step": 13179
+    },
+    {
+      "epoch": 0.8007776900176196,
+      "grad_norm": 0.19040468335151672,
+      "learning_rate": 9.53902456082072e-06,
+      "loss": 1.0202,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8008384470502461,
+      "grad_norm": 1.8374100923538208,
+      "learning_rate": 9.533401236850626e-06,
+      "loss": 1.0007,
+      "step": 13181
+    },
+    {
+      "epoch": 0.8008992040828726,
+      "grad_norm": 0.1839904934167862,
+      "learning_rate": 9.527779396191528e-06,
+      "loss": 0.9872,
+      "step": 13182
+    },
+    {
+      "epoch": 0.8009599611154992,
+      "grad_norm": 0.20559729635715485,
+      "learning_rate": 9.5221590390495e-06,
+      "loss": 1.1505,
+      "step": 13183
+    },
+    {
+      "epoch": 0.8010207181481257,
+      "grad_norm": 0.15975527465343475,
+      "learning_rate": 9.516540165630556e-06,
+      "loss": 1.0522,
+      "step": 13184
+    },
+    {
+      "epoch": 0.8010814751807521,
+      "grad_norm": 1.3011716604232788,
+      "learning_rate": 9.510922776140663e-06,
+      "loss": 1.0767,
+      "step": 13185
+    },
+    {
+      "epoch": 0.8011422322133787,
+      "grad_norm": 0.29181694984436035,
+      "learning_rate": 9.505306870785713e-06,
+      "loss": 1.1833,
+      "step": 13186
+    },
+    {
+      "epoch": 0.8012029892460052,
+      "grad_norm": 0.13684216141700745,
+      "learning_rate": 9.499692449771574e-06,
+      "loss": 1.0307,
+      "step": 13187
+    },
+    {
+      "epoch": 0.8012637462786317,
+      "grad_norm": 1.0310965776443481,
+      "learning_rate": 9.494079513304021e-06,
+      "loss": 1.1975,
+      "step": 13188
+    },
+    {
+      "epoch": 0.8013245033112583,
+      "grad_norm": 0.11666496843099594,
+      "learning_rate": 9.48846806158884e-06,
+      "loss": 1.0423,
+      "step": 13189
+    },
+    {
+      "epoch": 0.8013852603438848,
+      "grad_norm": 0.10867491364479065,
+      "learning_rate": 9.482858094831681e-06,
+      "loss": 1.0596,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8014460173765113,
+      "grad_norm": 0.24226537346839905,
+      "learning_rate": 9.477249613238186e-06,
+      "loss": 1.0954,
+      "step": 13191
+    },
+    {
+      "epoch": 0.8015067744091379,
+      "grad_norm": 0.32712113857269287,
+      "learning_rate": 9.471642617013948e-06,
+      "loss": 1.2661,
+      "step": 13192
+    },
+    {
+      "epoch": 0.8015675314417644,
+      "grad_norm": 0.14403632283210754,
+      "learning_rate": 9.466037106364462e-06,
+      "loss": 1.134,
+      "step": 13193
+    },
+    {
+      "epoch": 0.8016282884743909,
+      "grad_norm": 0.160531684756279,
+      "learning_rate": 9.46043308149524e-06,
+      "loss": 1.0509,
+      "step": 13194
+    },
+    {
+      "epoch": 0.8016890455070175,
+      "grad_norm": 0.15243718028068542,
+      "learning_rate": 9.454830542611686e-06,
+      "loss": 1.0402,
+      "step": 13195
+    },
+    {
+      "epoch": 0.801749802539644,
+      "grad_norm": 0.17986595630645752,
+      "learning_rate": 9.449229489919147e-06,
+      "loss": 1.0973,
+      "step": 13196
+    },
+    {
+      "epoch": 0.8018105595722705,
+      "grad_norm": 0.22678661346435547,
+      "learning_rate": 9.443629923622937e-06,
+      "loss": 1.0546,
+      "step": 13197
+    },
+    {
+      "epoch": 0.801871316604897,
+      "grad_norm": 0.22991134226322174,
+      "learning_rate": 9.438031843928302e-06,
+      "loss": 1.2182,
+      "step": 13198
+    },
+    {
+      "epoch": 0.8019320736375235,
+      "grad_norm": 0.1474834829568863,
+      "learning_rate": 9.43243525104046e-06,
+      "loss": 1.0439,
+      "step": 13199
+    },
+    {
+      "epoch": 0.80199283067015,
+      "grad_norm": 0.20641137659549713,
+      "learning_rate": 9.426840145164551e-06,
+      "loss": 1.1445,
+      "step": 13200
+    },
+    {
+      "epoch": 0.8020535877027766,
+      "grad_norm": 0.14994990825653076,
+      "learning_rate": 9.421246526505656e-06,
+      "loss": 1.0733,
+      "step": 13201
+    },
+    {
+      "epoch": 0.8021143447354031,
+      "grad_norm": 0.09064609557390213,
+      "learning_rate": 9.415654395268826e-06,
+      "loss": 1.0211,
+      "step": 13202
+    },
+    {
+      "epoch": 0.8021751017680296,
+      "grad_norm": 0.12877589464187622,
+      "learning_rate": 9.410063751658998e-06,
+      "loss": 1.0603,
+      "step": 13203
+    },
+    {
+      "epoch": 0.8022358588006562,
+      "grad_norm": 0.1695820689201355,
+      "learning_rate": 9.404474595881146e-06,
+      "loss": 1.0732,
+      "step": 13204
+    },
+    {
+      "epoch": 0.8022966158332827,
+      "grad_norm": 0.10930179059505463,
+      "learning_rate": 9.398886928140127e-06,
+      "loss": 1.0257,
+      "step": 13205
+    },
+    {
+      "epoch": 0.8023573728659092,
+      "grad_norm": 0.2766391634941101,
+      "learning_rate": 9.393300748640755e-06,
+      "loss": 1.0819,
+      "step": 13206
+    },
+    {
+      "epoch": 0.8024181298985358,
+      "grad_norm": 0.4591292440891266,
+      "learning_rate": 9.387716057587792e-06,
+      "loss": 1.2201,
+      "step": 13207
+    },
+    {
+      "epoch": 0.8024788869311623,
+      "grad_norm": 0.4646061062812805,
+      "learning_rate": 9.382132855185943e-06,
+      "loss": 1.245,
+      "step": 13208
+    },
+    {
+      "epoch": 0.8025396439637889,
+      "grad_norm": 0.13472218811511993,
+      "learning_rate": 9.376551141639872e-06,
+      "loss": 1.0377,
+      "step": 13209
+    },
+    {
+      "epoch": 0.8026004009964154,
+      "grad_norm": 0.11996764689683914,
+      "learning_rate": 9.370970917154165e-06,
+      "loss": 1.0544,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8026611580290418,
+      "grad_norm": 0.17723794281482697,
+      "learning_rate": 9.36539218193338e-06,
+      "loss": 1.0522,
+      "step": 13211
+    },
+    {
+      "epoch": 0.8027219150616683,
+      "grad_norm": 0.18431003391742706,
+      "learning_rate": 9.359814936181992e-06,
+      "loss": 1.2172,
+      "step": 13212
+    },
+    {
+      "epoch": 0.8027826720942949,
+      "grad_norm": 0.188750758767128,
+      "learning_rate": 9.354239180104445e-06,
+      "loss": 1.0233,
+      "step": 13213
+    },
+    {
+      "epoch": 0.8028434291269214,
+      "grad_norm": 0.12524496018886566,
+      "learning_rate": 9.348664913905108e-06,
+      "loss": 1.0335,
+      "step": 13214
+    },
+    {
+      "epoch": 0.802904186159548,
+      "grad_norm": 0.12777726352214813,
+      "learning_rate": 9.343092137788329e-06,
+      "loss": 1.0841,
+      "step": 13215
+    },
+    {
+      "epoch": 0.8029649431921745,
+      "grad_norm": 0.09639745205640793,
+      "learning_rate": 9.337520851958376e-06,
+      "loss": 1.0334,
+      "step": 13216
+    },
+    {
+      "epoch": 0.803025700224801,
+      "grad_norm": 0.2233753651380539,
+      "learning_rate": 9.331951056619442e-06,
+      "loss": 1.1472,
+      "step": 13217
+    },
+    {
+      "epoch": 0.8030864572574276,
+      "grad_norm": 0.16895906627178192,
+      "learning_rate": 9.326382751975693e-06,
+      "loss": 1.0756,
+      "step": 13218
+    },
+    {
+      "epoch": 0.8031472142900541,
+      "grad_norm": 0.16784971952438354,
+      "learning_rate": 9.320815938231258e-06,
+      "loss": 1.076,
+      "step": 13219
+    },
+    {
+      "epoch": 0.8032079713226806,
+      "grad_norm": 4.107651710510254,
+      "learning_rate": 9.315250615590182e-06,
+      "loss": 1.1153,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8032687283553072,
+      "grad_norm": 0.10829100012779236,
+      "learning_rate": 9.309686784256456e-06,
+      "loss": 1.049,
+      "step": 13221
+    },
+    {
+      "epoch": 0.8033294853879337,
+      "grad_norm": 0.10990607738494873,
+      "learning_rate": 9.304124444434037e-06,
+      "loss": 1.0091,
+      "step": 13222
+    },
+    {
+      "epoch": 0.8033902424205602,
+      "grad_norm": 0.1258353441953659,
+      "learning_rate": 9.298563596326777e-06,
+      "loss": 1.0488,
+      "step": 13223
+    },
+    {
+      "epoch": 0.8034509994531867,
+      "grad_norm": 0.1442100703716278,
+      "learning_rate": 9.293004240138548e-06,
+      "loss": 1.082,
+      "step": 13224
+    },
+    {
+      "epoch": 0.8035117564858132,
+      "grad_norm": 0.160639688372612,
+      "learning_rate": 9.287446376073123e-06,
+      "loss": 1.1263,
+      "step": 13225
+    },
+    {
+      "epoch": 0.8035725135184397,
+      "grad_norm": 0.32805612683296204,
+      "learning_rate": 9.281890004334214e-06,
+      "loss": 1.1557,
+      "step": 13226
+    },
+    {
+      "epoch": 0.8036332705510663,
+      "grad_norm": 0.1491694450378418,
+      "learning_rate": 9.276335125125501e-06,
+      "loss": 1.123,
+      "step": 13227
+    },
+    {
+      "epoch": 0.8036940275836928,
+      "grad_norm": 0.11665769666433334,
+      "learning_rate": 9.270781738650597e-06,
+      "loss": 1.0358,
+      "step": 13228
+    },
+    {
+      "epoch": 0.8037547846163193,
+      "grad_norm": 0.1171853318810463,
+      "learning_rate": 9.26522984511306e-06,
+      "loss": 1.0452,
+      "step": 13229
+    },
+    {
+      "epoch": 0.8038155416489459,
+      "grad_norm": 0.15451809763908386,
+      "learning_rate": 9.259679444716397e-06,
+      "loss": 1.0003,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8038762986815724,
+      "grad_norm": 0.10360805690288544,
+      "learning_rate": 9.254130537664057e-06,
+      "loss": 1.0297,
+      "step": 13231
+    },
+    {
+      "epoch": 0.8039370557141989,
+      "grad_norm": 0.15688200294971466,
+      "learning_rate": 9.248583124159438e-06,
+      "loss": 1.0976,
+      "step": 13232
+    },
+    {
+      "epoch": 0.8039978127468255,
+      "grad_norm": 0.6644073128700256,
+      "learning_rate": 9.24303720440588e-06,
+      "loss": 1.0816,
+      "step": 13233
+    },
+    {
+      "epoch": 0.804058569779452,
+      "grad_norm": 0.10312240570783615,
+      "learning_rate": 9.237492778606665e-06,
+      "loss": 1.0075,
+      "step": 13234
+    },
+    {
+      "epoch": 0.8041193268120785,
+      "grad_norm": 0.12448840588331223,
+      "learning_rate": 9.231949846965038e-06,
+      "loss": 1.0182,
+      "step": 13235
+    },
+    {
+      "epoch": 0.8041800838447051,
+      "grad_norm": 0.12514646351337433,
+      "learning_rate": 9.226408409684189e-06,
+      "loss": 1.0028,
+      "step": 13236
+    },
+    {
+      "epoch": 0.8042408408773315,
+      "grad_norm": 18.818635940551758,
+      "learning_rate": 9.220868466967203e-06,
+      "loss": 1.1145,
+      "step": 13237
+    },
+    {
+      "epoch": 0.804301597909958,
+      "grad_norm": 0.1440465897321701,
+      "learning_rate": 9.215330019017165e-06,
+      "loss": 1.0281,
+      "step": 13238
+    },
+    {
+      "epoch": 0.8043623549425846,
+      "grad_norm": 0.4489496648311615,
+      "learning_rate": 9.209793066037081e-06,
+      "loss": 1.2157,
+      "step": 13239
+    },
+    {
+      "epoch": 0.8044231119752111,
+      "grad_norm": 0.17046141624450684,
+      "learning_rate": 9.204257608229928e-06,
+      "loss": 1.0976,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8044838690078376,
+      "grad_norm": 2.879633665084839,
+      "learning_rate": 9.198723645798595e-06,
+      "loss": 1.0512,
+      "step": 13241
+    },
+    {
+      "epoch": 0.8045446260404642,
+      "grad_norm": 0.11650007218122482,
+      "learning_rate": 9.193191178945948e-06,
+      "loss": 0.9942,
+      "step": 13242
+    },
+    {
+      "epoch": 0.8046053830730907,
+      "grad_norm": 0.14756543934345245,
+      "learning_rate": 9.187660207874754e-06,
+      "loss": 1.1114,
+      "step": 13243
+    },
+    {
+      "epoch": 0.8046661401057172,
+      "grad_norm": 0.19879499077796936,
+      "learning_rate": 9.182130732787747e-06,
+      "loss": 0.9994,
+      "step": 13244
+    },
+    {
+      "epoch": 0.8047268971383438,
+      "grad_norm": 6.629814147949219,
+      "learning_rate": 9.176602753887648e-06,
+      "loss": 1.0806,
+      "step": 13245
+    },
+    {
+      "epoch": 0.8047876541709703,
+      "grad_norm": 0.2768288254737854,
+      "learning_rate": 9.171076271377061e-06,
+      "loss": 1.1867,
+      "step": 13246
+    },
+    {
+      "epoch": 0.8048484112035968,
+      "grad_norm": 0.1433205008506775,
+      "learning_rate": 9.165551285458562e-06,
+      "loss": 1.0882,
+      "step": 13247
+    },
+    {
+      "epoch": 0.8049091682362234,
+      "grad_norm": 0.133346289396286,
+      "learning_rate": 9.16002779633468e-06,
+      "loss": 1.0438,
+      "step": 13248
+    },
+    {
+      "epoch": 0.8049699252688499,
+      "grad_norm": 0.1338362991809845,
+      "learning_rate": 9.154505804207875e-06,
+      "loss": 1.0284,
+      "step": 13249
+    },
+    {
+      "epoch": 0.8050306823014763,
+      "grad_norm": 0.15715482831001282,
+      "learning_rate": 9.14898530928055e-06,
+      "loss": 1.0422,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8050914393341029,
+      "grad_norm": 0.13264602422714233,
+      "learning_rate": 9.143466311755067e-06,
+      "loss": 0.984,
+      "step": 13251
+    },
+    {
+      "epoch": 0.8051521963667294,
+      "grad_norm": 0.1752057522535324,
+      "learning_rate": 9.137948811833718e-06,
+      "loss": 1.1425,
+      "step": 13252
+    },
+    {
+      "epoch": 0.805212953399356,
+      "grad_norm": 0.14663147926330566,
+      "learning_rate": 9.13243280971876e-06,
+      "loss": 1.0541,
+      "step": 13253
+    },
+    {
+      "epoch": 0.8052737104319825,
+      "grad_norm": 0.1315969079732895,
+      "learning_rate": 9.126918305612376e-06,
+      "loss": 1.0325,
+      "step": 13254
+    },
+    {
+      "epoch": 0.805334467464609,
+      "grad_norm": 0.18285177648067474,
+      "learning_rate": 9.12140529971669e-06,
+      "loss": 1.1071,
+      "step": 13255
+    },
+    {
+      "epoch": 0.8053952244972356,
+      "grad_norm": 0.13508103787899017,
+      "learning_rate": 9.11589379223382e-06,
+      "loss": 1.0691,
+      "step": 13256
+    },
+    {
+      "epoch": 0.8054559815298621,
+      "grad_norm": 0.10481157153844833,
+      "learning_rate": 9.11038378336575e-06,
+      "loss": 1.0188,
+      "step": 13257
+    },
+    {
+      "epoch": 0.8055167385624886,
+      "grad_norm": 0.15566766262054443,
+      "learning_rate": 9.104875273314472e-06,
+      "loss": 1.0761,
+      "step": 13258
+    },
+    {
+      "epoch": 0.8055774955951152,
+      "grad_norm": 0.1830064058303833,
+      "learning_rate": 9.099368262281887e-06,
+      "loss": 1.0461,
+      "step": 13259
+    },
+    {
+      "epoch": 0.8056382526277417,
+      "grad_norm": 0.21882294118404388,
+      "learning_rate": 9.093862750469862e-06,
+      "loss": 1.0922,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8056990096603682,
+      "grad_norm": 0.14830783009529114,
+      "learning_rate": 9.08835873808021e-06,
+      "loss": 1.052,
+      "step": 13261
+    },
+    {
+      "epoch": 0.8057597666929948,
+      "grad_norm": 0.09598112851381302,
+      "learning_rate": 9.08285622531469e-06,
+      "loss": 1.0082,
+      "step": 13262
+    },
+    {
+      "epoch": 0.8058205237256213,
+      "grad_norm": 0.14668449759483337,
+      "learning_rate": 9.077355212374972e-06,
+      "loss": 1.1288,
+      "step": 13263
+    },
+    {
+      "epoch": 0.8058812807582477,
+      "grad_norm": 0.18051615357398987,
+      "learning_rate": 9.071855699462712e-06,
+      "loss": 1.0731,
+      "step": 13264
+    },
+    {
+      "epoch": 0.8059420377908743,
+      "grad_norm": 0.1578369289636612,
+      "learning_rate": 9.066357686779475e-06,
+      "loss": 1.1266,
+      "step": 13265
+    },
+    {
+      "epoch": 0.8060027948235008,
+      "grad_norm": 0.13977807760238647,
+      "learning_rate": 9.060861174526824e-06,
+      "loss": 1.0327,
+      "step": 13266
+    },
+    {
+      "epoch": 0.8060635518561273,
+      "grad_norm": 0.21295364201068878,
+      "learning_rate": 9.05536616290622e-06,
+      "loss": 1.1553,
+      "step": 13267
+    },
+    {
+      "epoch": 0.8061243088887539,
+      "grad_norm": 0.13446098566055298,
+      "learning_rate": 9.049872652119085e-06,
+      "loss": 1.0259,
+      "step": 13268
+    },
+    {
+      "epoch": 0.8061850659213804,
+      "grad_norm": 0.41465988755226135,
+      "learning_rate": 9.044380642366795e-06,
+      "loss": 1.139,
+      "step": 13269
+    },
+    {
+      "epoch": 0.8062458229540069,
+      "grad_norm": 0.16579413414001465,
+      "learning_rate": 9.038890133850625e-06,
+      "loss": 1.0413,
+      "step": 13270
+    },
+    {
+      "epoch": 0.8063065799866335,
+      "grad_norm": 0.20012642443180084,
+      "learning_rate": 9.033401126771867e-06,
+      "loss": 1.0729,
+      "step": 13271
+    },
+    {
+      "epoch": 0.80636733701926,
+      "grad_norm": 0.1496468335390091,
+      "learning_rate": 9.027913621331707e-06,
+      "loss": 1.1078,
+      "step": 13272
+    },
+    {
+      "epoch": 0.8064280940518865,
+      "grad_norm": 0.13976065814495087,
+      "learning_rate": 9.022427617731293e-06,
+      "loss": 1.0792,
+      "step": 13273
+    },
+    {
+      "epoch": 0.8064888510845131,
+      "grad_norm": 0.1137533038854599,
+      "learning_rate": 9.016943116171722e-06,
+      "loss": 1.0368,
+      "step": 13274
+    },
+    {
+      "epoch": 0.8065496081171396,
+      "grad_norm": 0.11362741887569427,
+      "learning_rate": 9.011460116854015e-06,
+      "loss": 1.0346,
+      "step": 13275
+    },
+    {
+      "epoch": 0.8066103651497661,
+      "grad_norm": 0.14698021113872528,
+      "learning_rate": 9.005978619979167e-06,
+      "loss": 1.0043,
+      "step": 13276
+    },
+    {
+      "epoch": 0.8066711221823926,
+      "grad_norm": 0.1776665300130844,
+      "learning_rate": 9.000498625748088e-06,
+      "loss": 1.0662,
+      "step": 13277
+    },
+    {
+      "epoch": 0.8067318792150191,
+      "grad_norm": 0.17688356339931488,
+      "learning_rate": 8.995020134361664e-06,
+      "loss": 1.1296,
+      "step": 13278
+    },
+    {
+      "epoch": 0.8067926362476456,
+      "grad_norm": 0.14069749414920807,
+      "learning_rate": 8.989543146020701e-06,
+      "loss": 1.0873,
+      "step": 13279
+    },
+    {
+      "epoch": 0.8068533932802722,
+      "grad_norm": 0.38233307003974915,
+      "learning_rate": 8.98406766092596e-06,
+      "loss": 1.2014,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8069141503128987,
+      "grad_norm": 0.13473199307918549,
+      "learning_rate": 8.978593679278136e-06,
+      "loss": 1.0326,
+      "step": 13281
+    },
+    {
+      "epoch": 0.8069749073455252,
+      "grad_norm": 0.1711856573820114,
+      "learning_rate": 8.973121201277901e-06,
+      "loss": 1.1841,
+      "step": 13282
+    },
+    {
+      "epoch": 0.8070356643781518,
+      "grad_norm": 0.09955224394798279,
+      "learning_rate": 8.967650227125856e-06,
+      "loss": 1.0142,
+      "step": 13283
+    },
+    {
+      "epoch": 0.8070964214107783,
+      "grad_norm": 0.2441820651292801,
+      "learning_rate": 8.962180757022504e-06,
+      "loss": 1.1502,
+      "step": 13284
+    },
+    {
+      "epoch": 0.8071571784434048,
+      "grad_norm": 0.13988153636455536,
+      "learning_rate": 8.956712791168354e-06,
+      "loss": 1.0682,
+      "step": 13285
+    },
+    {
+      "epoch": 0.8072179354760314,
+      "grad_norm": 0.11172053217887878,
+      "learning_rate": 8.951246329763813e-06,
+      "loss": 1.0601,
+      "step": 13286
+    },
+    {
+      "epoch": 0.8072786925086579,
+      "grad_norm": 0.12162996828556061,
+      "learning_rate": 8.945781373009288e-06,
+      "loss": 1.0532,
+      "step": 13287
+    },
+    {
+      "epoch": 0.8073394495412844,
+      "grad_norm": 0.1264953315258026,
+      "learning_rate": 8.940317921105085e-06,
+      "loss": 1.1064,
+      "step": 13288
+    },
+    {
+      "epoch": 0.807400206573911,
+      "grad_norm": 0.21767602860927582,
+      "learning_rate": 8.934855974251472e-06,
+      "loss": 1.0879,
+      "step": 13289
+    },
+    {
+      "epoch": 0.8074609636065374,
+      "grad_norm": 0.1758887618780136,
+      "learning_rate": 8.929395532648643e-06,
+      "loss": 1.1053,
+      "step": 13290
+    },
+    {
+      "epoch": 0.8075217206391639,
+      "grad_norm": 0.13397012650966644,
+      "learning_rate": 8.923936596496745e-06,
+      "loss": 1.0486,
+      "step": 13291
+    },
+    {
+      "epoch": 0.8075824776717905,
+      "grad_norm": 0.21539229154586792,
+      "learning_rate": 8.918479165995901e-06,
+      "loss": 1.0518,
+      "step": 13292
+    },
+    {
+      "epoch": 0.807643234704417,
+      "grad_norm": 0.13243921101093292,
+      "learning_rate": 8.913023241346147e-06,
+      "loss": 1.155,
+      "step": 13293
+    },
+    {
+      "epoch": 0.8077039917370435,
+      "grad_norm": 0.18805481493473053,
+      "learning_rate": 8.907568822747469e-06,
+      "loss": 1.0687,
+      "step": 13294
+    },
+    {
+      "epoch": 0.8077647487696701,
+      "grad_norm": 0.1260208636522293,
+      "learning_rate": 8.902115910399794e-06,
+      "loss": 1.1178,
+      "step": 13295
+    },
+    {
+      "epoch": 0.8078255058022966,
+      "grad_norm": 0.1716604381799698,
+      "learning_rate": 8.89666450450301e-06,
+      "loss": 1.0875,
+      "step": 13296
+    },
+    {
+      "epoch": 0.8078862628349232,
+      "grad_norm": 0.1804407685995102,
+      "learning_rate": 8.89121460525693e-06,
+      "loss": 1.0911,
+      "step": 13297
+    },
+    {
+      "epoch": 0.8079470198675497,
+      "grad_norm": 0.18404801189899445,
+      "learning_rate": 8.885766212861329e-06,
+      "loss": 1.0872,
+      "step": 13298
+    },
+    {
+      "epoch": 0.8080077769001762,
+      "grad_norm": 0.1195506826043129,
+      "learning_rate": 8.880319327515913e-06,
+      "loss": 1.0159,
+      "step": 13299
+    },
+    {
+      "epoch": 0.8080685339328028,
+      "grad_norm": 0.13083086907863617,
+      "learning_rate": 8.874873949420336e-06,
+      "loss": 1.0683,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8081292909654293,
+      "grad_norm": 0.45422470569610596,
+      "learning_rate": 8.869430078774205e-06,
+      "loss": 1.1042,
+      "step": 13301
+    },
+    {
+      "epoch": 0.8081900479980558,
+      "grad_norm": 0.0996997058391571,
+      "learning_rate": 8.863987715777055e-06,
+      "loss": 1.0408,
+      "step": 13302
+    },
+    {
+      "epoch": 0.8082508050306823,
+      "grad_norm": 0.1602700650691986,
+      "learning_rate": 8.858546860628415e-06,
+      "loss": 1.0789,
+      "step": 13303
+    },
+    {
+      "epoch": 0.8083115620633088,
+      "grad_norm": 0.7398463487625122,
+      "learning_rate": 8.853107513527676e-06,
+      "loss": 1.0957,
+      "step": 13304
+    },
+    {
+      "epoch": 0.8083723190959353,
+      "grad_norm": 0.18652676045894623,
+      "learning_rate": 8.84766967467423e-06,
+      "loss": 1.073,
+      "step": 13305
+    },
+    {
+      "epoch": 0.8084330761285619,
+      "grad_norm": 0.13182881474494934,
+      "learning_rate": 8.842233344267414e-06,
+      "loss": 1.092,
+      "step": 13306
+    },
+    {
+      "epoch": 0.8084938331611884,
+      "grad_norm": 0.09149140119552612,
+      "learning_rate": 8.836798522506473e-06,
+      "loss": 1.035,
+      "step": 13307
+    },
+    {
+      "epoch": 0.8085545901938149,
+      "grad_norm": 0.1158941239118576,
+      "learning_rate": 8.831365209590653e-06,
+      "loss": 1.019,
+      "step": 13308
+    },
+    {
+      "epoch": 0.8086153472264415,
+      "grad_norm": 0.16232800483703613,
+      "learning_rate": 8.825933405719112e-06,
+      "loss": 1.0799,
+      "step": 13309
+    },
+    {
+      "epoch": 0.808676104259068,
+      "grad_norm": 0.3569292426109314,
+      "learning_rate": 8.820503111090927e-06,
+      "loss": 1.0566,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8087368612916945,
+      "grad_norm": 0.2154925912618637,
+      "learning_rate": 8.815074325905147e-06,
+      "loss": 1.034,
+      "step": 13311
+    },
+    {
+      "epoch": 0.8087976183243211,
+      "grad_norm": 0.12039048224687576,
+      "learning_rate": 8.809647050360793e-06,
+      "loss": 1.0018,
+      "step": 13312
+    },
+    {
+      "epoch": 0.8088583753569476,
+      "grad_norm": 0.1591581404209137,
+      "learning_rate": 8.804221284656788e-06,
+      "loss": 1.0799,
+      "step": 13313
+    },
+    {
+      "epoch": 0.8089191323895741,
+      "grad_norm": 0.13909871876239777,
+      "learning_rate": 8.798797028992018e-06,
+      "loss": 1.0332,
+      "step": 13314
+    },
+    {
+      "epoch": 0.8089798894222007,
+      "grad_norm": 0.13049043715000153,
+      "learning_rate": 8.793374283565303e-06,
+      "loss": 1.0218,
+      "step": 13315
+    },
+    {
+      "epoch": 0.8090406464548271,
+      "grad_norm": 0.20883135497570038,
+      "learning_rate": 8.787953048575414e-06,
+      "loss": 1.1019,
+      "step": 13316
+    },
+    {
+      "epoch": 0.8091014034874536,
+      "grad_norm": 0.18288029730319977,
+      "learning_rate": 8.782533324221081e-06,
+      "loss": 1.2101,
+      "step": 13317
+    },
+    {
+      "epoch": 0.8091621605200802,
+      "grad_norm": 0.1941082775592804,
+      "learning_rate": 8.77711511070095e-06,
+      "loss": 1.1775,
+      "step": 13318
+    },
+    {
+      "epoch": 0.8092229175527067,
+      "grad_norm": 0.11101412773132324,
+      "learning_rate": 8.771698408213635e-06,
+      "loss": 1.0599,
+      "step": 13319
+    },
+    {
+      "epoch": 0.8092836745853332,
+      "grad_norm": 0.11275576055049896,
+      "learning_rate": 8.766283216957689e-06,
+      "loss": 1.0599,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8093444316179598,
+      "grad_norm": 0.18482142686843872,
+      "learning_rate": 8.760869537131594e-06,
+      "loss": 1.033,
+      "step": 13321
+    },
+    {
+      "epoch": 0.8094051886505863,
+      "grad_norm": 0.2128552347421646,
+      "learning_rate": 8.755457368933801e-06,
+      "loss": 1.1556,
+      "step": 13322
+    },
+    {
+      "epoch": 0.8094659456832128,
+      "grad_norm": 0.15571439266204834,
+      "learning_rate": 8.750046712562693e-06,
+      "loss": 1.1257,
+      "step": 13323
+    },
+    {
+      "epoch": 0.8095267027158394,
+      "grad_norm": 0.10723083466291428,
+      "learning_rate": 8.74463756821659e-06,
+      "loss": 1.0042,
+      "step": 13324
+    },
+    {
+      "epoch": 0.8095874597484659,
+      "grad_norm": 0.2939596176147461,
+      "learning_rate": 8.73922993609378e-06,
+      "loss": 1.1041,
+      "step": 13325
+    },
+    {
+      "epoch": 0.8096482167810924,
+      "grad_norm": 0.14527249336242676,
+      "learning_rate": 8.733823816392466e-06,
+      "loss": 1.0108,
+      "step": 13326
+    },
+    {
+      "epoch": 0.809708973813719,
+      "grad_norm": 0.09700893610715866,
+      "learning_rate": 8.7284192093108e-06,
+      "loss": 1.0229,
+      "step": 13327
+    },
+    {
+      "epoch": 0.8097697308463455,
+      "grad_norm": 0.10777320712804794,
+      "learning_rate": 8.723016115046928e-06,
+      "loss": 1.025,
+      "step": 13328
+    },
+    {
+      "epoch": 0.8098304878789719,
+      "grad_norm": 0.15057148039340973,
+      "learning_rate": 8.717614533798885e-06,
+      "loss": 1.1323,
+      "step": 13329
+    },
+    {
+      "epoch": 0.8098912449115985,
+      "grad_norm": 0.09742692112922668,
+      "learning_rate": 8.71221446576465e-06,
+      "loss": 1.0654,
+      "step": 13330
+    },
+    {
+      "epoch": 0.809952001944225,
+      "grad_norm": 0.16451206803321838,
+      "learning_rate": 8.706815911142175e-06,
+      "loss": 1.0946,
+      "step": 13331
+    },
+    {
+      "epoch": 0.8100127589768515,
+      "grad_norm": 0.14277468621730804,
+      "learning_rate": 8.70141887012933e-06,
+      "loss": 1.0022,
+      "step": 13332
+    },
+    {
+      "epoch": 0.8100735160094781,
+      "grad_norm": 0.11534839868545532,
+      "learning_rate": 8.696023342923976e-06,
+      "loss": 0.9816,
+      "step": 13333
+    },
+    {
+      "epoch": 0.8101342730421046,
+      "grad_norm": 0.13074979186058044,
+      "learning_rate": 8.690629329723865e-06,
+      "loss": 1.073,
+      "step": 13334
+    },
+    {
+      "epoch": 0.8101950300747311,
+      "grad_norm": 0.1668655276298523,
+      "learning_rate": 8.685236830726729e-06,
+      "loss": 1.1067,
+      "step": 13335
+    },
+    {
+      "epoch": 0.8102557871073577,
+      "grad_norm": 0.27570295333862305,
+      "learning_rate": 8.679845846130225e-06,
+      "loss": 1.0884,
+      "step": 13336
+    },
+    {
+      "epoch": 0.8103165441399842,
+      "grad_norm": 0.1755102276802063,
+      "learning_rate": 8.67445637613194e-06,
+      "loss": 1.1081,
+      "step": 13337
+    },
+    {
+      "epoch": 0.8103773011726108,
+      "grad_norm": 0.14132289588451385,
+      "learning_rate": 8.669068420929454e-06,
+      "loss": 1.087,
+      "step": 13338
+    },
+    {
+      "epoch": 0.8104380582052373,
+      "grad_norm": 0.1467886120080948,
+      "learning_rate": 8.663681980720257e-06,
+      "loss": 1.0114,
+      "step": 13339
+    },
+    {
+      "epoch": 0.8104988152378638,
+      "grad_norm": 0.10594131797552109,
+      "learning_rate": 8.658297055701786e-06,
+      "loss": 1.0128,
+      "step": 13340
+    },
+    {
+      "epoch": 0.8105595722704904,
+      "grad_norm": 0.11226150393486023,
+      "learning_rate": 8.652913646071425e-06,
+      "loss": 1.0077,
+      "step": 13341
+    },
+    {
+      "epoch": 0.8106203293031168,
+      "grad_norm": 0.17683349549770355,
+      "learning_rate": 8.647531752026511e-06,
+      "loss": 1.1241,
+      "step": 13342
+    },
+    {
+      "epoch": 0.8106810863357433,
+      "grad_norm": 0.1497465819120407,
+      "learning_rate": 8.642151373764312e-06,
+      "loss": 1.1421,
+      "step": 13343
+    },
+    {
+      "epoch": 0.8107418433683699,
+      "grad_norm": 0.11562196165323257,
+      "learning_rate": 8.63677251148205e-06,
+      "loss": 1.0636,
+      "step": 13344
+    },
+    {
+      "epoch": 0.8108026004009964,
+      "grad_norm": 0.1849144995212555,
+      "learning_rate": 8.631395165376888e-06,
+      "loss": 1.1321,
+      "step": 13345
+    },
+    {
+      "epoch": 0.8108633574336229,
+      "grad_norm": 0.21299001574516296,
+      "learning_rate": 8.62601933564593e-06,
+      "loss": 1.2577,
+      "step": 13346
+    },
+    {
+      "epoch": 0.8109241144662495,
+      "grad_norm": 0.17349374294281006,
+      "learning_rate": 8.62064502248623e-06,
+      "loss": 1.1222,
+      "step": 13347
+    },
+    {
+      "epoch": 0.810984871498876,
+      "grad_norm": 0.565649688243866,
+      "learning_rate": 8.615272226094773e-06,
+      "loss": 1.0232,
+      "step": 13348
+    },
+    {
+      "epoch": 0.8110456285315025,
+      "grad_norm": 0.5099070072174072,
+      "learning_rate": 8.609900946668537e-06,
+      "loss": 1.1207,
+      "step": 13349
+    },
+    {
+      "epoch": 0.8111063855641291,
+      "grad_norm": 0.11480657011270523,
+      "learning_rate": 8.604531184404368e-06,
+      "loss": 1.0016,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8111671425967556,
+      "grad_norm": 0.14679767191410065,
+      "learning_rate": 8.599162939499111e-06,
+      "loss": 1.002,
+      "step": 13351
+    },
+    {
+      "epoch": 0.8112278996293821,
+      "grad_norm": 0.16588066518306732,
+      "learning_rate": 8.59379621214954e-06,
+      "loss": 1.0356,
+      "step": 13352
+    },
+    {
+      "epoch": 0.8112886566620087,
+      "grad_norm": 0.1046774610877037,
+      "learning_rate": 8.588431002552361e-06,
+      "loss": 1.0527,
+      "step": 13353
+    },
+    {
+      "epoch": 0.8113494136946352,
+      "grad_norm": 0.1544579565525055,
+      "learning_rate": 8.583067310904264e-06,
+      "loss": 1.1128,
+      "step": 13354
+    },
+    {
+      "epoch": 0.8114101707272616,
+      "grad_norm": 0.10491519421339035,
+      "learning_rate": 8.577705137401837e-06,
+      "loss": 1.0483,
+      "step": 13355
+    },
+    {
+      "epoch": 0.8114709277598882,
+      "grad_norm": 0.1678817868232727,
+      "learning_rate": 8.572344482241645e-06,
+      "loss": 1.0977,
+      "step": 13356
+    },
+    {
+      "epoch": 0.8115316847925147,
+      "grad_norm": 0.13530850410461426,
+      "learning_rate": 8.56698534562016e-06,
+      "loss": 1.0349,
+      "step": 13357
+    },
+    {
+      "epoch": 0.8115924418251412,
+      "grad_norm": 0.4289415776729584,
+      "learning_rate": 8.561627727733828e-06,
+      "loss": 1.1378,
+      "step": 13358
+    },
+    {
+      "epoch": 0.8116531988577678,
+      "grad_norm": 0.19829007983207703,
+      "learning_rate": 8.556271628779056e-06,
+      "loss": 1.0593,
+      "step": 13359
+    },
+    {
+      "epoch": 0.8117139558903943,
+      "grad_norm": 0.23053620755672455,
+      "learning_rate": 8.550917048952151e-06,
+      "loss": 1.202,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8117747129230208,
+      "grad_norm": 0.2767316401004791,
+      "learning_rate": 8.545563988449401e-06,
+      "loss": 1.1134,
+      "step": 13361
+    },
+    {
+      "epoch": 0.8118354699556474,
+      "grad_norm": 0.14083430171012878,
+      "learning_rate": 8.54021244746701e-06,
+      "loss": 1.2361,
+      "step": 13362
+    },
+    {
+      "epoch": 0.8118962269882739,
+      "grad_norm": 0.14965178072452545,
+      "learning_rate": 8.534862426201151e-06,
+      "loss": 1.0433,
+      "step": 13363
+    },
+    {
+      "epoch": 0.8119569840209004,
+      "grad_norm": 0.12796726822853088,
+      "learning_rate": 8.52951392484792e-06,
+      "loss": 1.0932,
+      "step": 13364
+    },
+    {
+      "epoch": 0.812017741053527,
+      "grad_norm": 0.11587242782115936,
+      "learning_rate": 8.524166943603373e-06,
+      "loss": 1.046,
+      "step": 13365
+    },
+    {
+      "epoch": 0.8120784980861535,
+      "grad_norm": 0.1596328616142273,
+      "learning_rate": 8.518821482663508e-06,
+      "loss": 1.0971,
+      "step": 13366
+    },
+    {
+      "epoch": 0.81213925511878,
+      "grad_norm": 0.11747420579195023,
+      "learning_rate": 8.513477542224258e-06,
+      "loss": 0.9845,
+      "step": 13367
+    },
+    {
+      "epoch": 0.8122000121514066,
+      "grad_norm": 0.2287248969078064,
+      "learning_rate": 8.508135122481503e-06,
+      "loss": 1.073,
+      "step": 13368
+    },
+    {
+      "epoch": 0.812260769184033,
+      "grad_norm": 0.20100581645965576,
+      "learning_rate": 8.502794223631067e-06,
+      "loss": 1.0521,
+      "step": 13369
+    },
+    {
+      "epoch": 0.8123215262166595,
+      "grad_norm": 0.13511161506175995,
+      "learning_rate": 8.497454845868752e-06,
+      "loss": 1.0252,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8123822832492861,
+      "grad_norm": 0.1565893143415451,
+      "learning_rate": 8.492116989390243e-06,
+      "loss": 1.042,
+      "step": 13371
+    },
+    {
+      "epoch": 0.8124430402819126,
+      "grad_norm": 11.986188888549805,
+      "learning_rate": 8.486780654391207e-06,
+      "loss": 1.0436,
+      "step": 13372
+    },
+    {
+      "epoch": 0.8125037973145391,
+      "grad_norm": 0.310553640127182,
+      "learning_rate": 8.481445841067253e-06,
+      "loss": 1.0624,
+      "step": 13373
+    },
+    {
+      "epoch": 0.8125645543471657,
+      "grad_norm": 0.15974737703800201,
+      "learning_rate": 8.476112549613912e-06,
+      "loss": 1.0767,
+      "step": 13374
+    },
+    {
+      "epoch": 0.8126253113797922,
+      "grad_norm": 0.11792033910751343,
+      "learning_rate": 8.470780780226705e-06,
+      "loss": 1.0487,
+      "step": 13375
+    },
+    {
+      "epoch": 0.8126860684124187,
+      "grad_norm": 0.114162877202034,
+      "learning_rate": 8.465450533101072e-06,
+      "loss": 1.1031,
+      "step": 13376
+    },
+    {
+      "epoch": 0.8127468254450453,
+      "grad_norm": 0.5604024529457092,
+      "learning_rate": 8.460121808432358e-06,
+      "loss": 1.0376,
+      "step": 13377
+    },
+    {
+      "epoch": 0.8128075824776718,
+      "grad_norm": 2.486450433731079,
+      "learning_rate": 8.454794606415911e-06,
+      "loss": 1.0815,
+      "step": 13378
+    },
+    {
+      "epoch": 0.8128683395102984,
+      "grad_norm": 0.48364776372909546,
+      "learning_rate": 8.44946892724699e-06,
+      "loss": 1.0377,
+      "step": 13379
+    },
+    {
+      "epoch": 0.8129290965429249,
+      "grad_norm": 0.177617147564888,
+      "learning_rate": 8.444144771120826e-06,
+      "loss": 1.1836,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8129898535755514,
+      "grad_norm": 0.2059275209903717,
+      "learning_rate": 8.438822138232572e-06,
+      "loss": 1.1152,
+      "step": 13381
+    },
+    {
+      "epoch": 0.8130506106081778,
+      "grad_norm": 0.11329291760921478,
+      "learning_rate": 8.433501028777318e-06,
+      "loss": 1.0223,
+      "step": 13382
+    },
+    {
+      "epoch": 0.8131113676408044,
+      "grad_norm": 0.11364471912384033,
+      "learning_rate": 8.42818144295014e-06,
+      "loss": 1.0489,
+      "step": 13383
+    },
+    {
+      "epoch": 0.8131721246734309,
+      "grad_norm": 0.10933076590299606,
+      "learning_rate": 8.422863380945977e-06,
+      "loss": 1.0082,
+      "step": 13384
+    },
+    {
+      "epoch": 0.8132328817060575,
+      "grad_norm": 0.20043650269508362,
+      "learning_rate": 8.417546842959806e-06,
+      "loss": 1.1137,
+      "step": 13385
+    },
+    {
+      "epoch": 0.813293638738684,
+      "grad_norm": 0.2203894853591919,
+      "learning_rate": 8.412231829186495e-06,
+      "loss": 1.1604,
+      "step": 13386
+    },
+    {
+      "epoch": 0.8133543957713105,
+      "grad_norm": 0.11832256615161896,
+      "learning_rate": 8.406918339820864e-06,
+      "loss": 1.0116,
+      "step": 13387
+    },
+    {
+      "epoch": 0.8134151528039371,
+      "grad_norm": 0.1803184449672699,
+      "learning_rate": 8.40160637505768e-06,
+      "loss": 1.1081,
+      "step": 13388
+    },
+    {
+      "epoch": 0.8134759098365636,
+      "grad_norm": 0.5034940838813782,
+      "learning_rate": 8.396295935091659e-06,
+      "loss": 0.9652,
+      "step": 13389
+    },
+    {
+      "epoch": 0.8135366668691901,
+      "grad_norm": 0.12402615696191788,
+      "learning_rate": 8.390987020117447e-06,
+      "loss": 1.0652,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8135974239018167,
+      "grad_norm": 0.1698220670223236,
+      "learning_rate": 8.385679630329652e-06,
+      "loss": 0.9991,
+      "step": 13391
+    },
+    {
+      "epoch": 0.8136581809344432,
+      "grad_norm": 0.09587369859218597,
+      "learning_rate": 8.380373765922812e-06,
+      "loss": 1.0545,
+      "step": 13392
+    },
+    {
+      "epoch": 0.8137189379670697,
+      "grad_norm": 0.11561412364244461,
+      "learning_rate": 8.37506942709142e-06,
+      "loss": 1.0178,
+      "step": 13393
+    },
+    {
+      "epoch": 0.8137796949996963,
+      "grad_norm": 0.36636975407600403,
+      "learning_rate": 8.3697666140299e-06,
+      "loss": 1.1586,
+      "step": 13394
+    },
+    {
+      "epoch": 0.8138404520323227,
+      "grad_norm": 0.2472551465034485,
+      "learning_rate": 8.364465326932619e-06,
+      "loss": 1.0352,
+      "step": 13395
+    },
+    {
+      "epoch": 0.8139012090649492,
+      "grad_norm": 0.11036510765552521,
+      "learning_rate": 8.359165565993938e-06,
+      "loss": 1.0487,
+      "step": 13396
+    },
+    {
+      "epoch": 0.8139619660975758,
+      "grad_norm": 0.19025571644306183,
+      "learning_rate": 8.353867331408078e-06,
+      "loss": 1.0521,
+      "step": 13397
+    },
+    {
+      "epoch": 0.8140227231302023,
+      "grad_norm": 0.2817869484424591,
+      "learning_rate": 8.348570623369262e-06,
+      "loss": 1.1052,
+      "step": 13398
+    },
+    {
+      "epoch": 0.8140834801628288,
+      "grad_norm": 0.17201749980449677,
+      "learning_rate": 8.34327544207164e-06,
+      "loss": 1.072,
+      "step": 13399
+    },
+    {
+      "epoch": 0.8141442371954554,
+      "grad_norm": 0.12416765838861465,
+      "learning_rate": 8.337981787709297e-06,
+      "loss": 1.0651,
+      "step": 13400
+    },
+    {
+      "epoch": 0.8142049942280819,
+      "grad_norm": 0.1194167211651802,
+      "learning_rate": 8.332689660476294e-06,
+      "loss": 1.0725,
+      "step": 13401
+    },
+    {
+      "epoch": 0.8142657512607084,
+      "grad_norm": 0.16337068378925323,
+      "learning_rate": 8.32739906056661e-06,
+      "loss": 1.144,
+      "step": 13402
+    },
+    {
+      "epoch": 0.814326508293335,
+      "grad_norm": 0.14371144771575928,
+      "learning_rate": 8.322109988174177e-06,
+      "loss": 1.0232,
+      "step": 13403
+    },
+    {
+      "epoch": 0.8143872653259615,
+      "grad_norm": 0.14053042232990265,
+      "learning_rate": 8.31682244349285e-06,
+      "loss": 0.9842,
+      "step": 13404
+    },
+    {
+      "epoch": 0.814448022358588,
+      "grad_norm": 0.2046688050031662,
+      "learning_rate": 8.311536426716437e-06,
+      "loss": 1.0372,
+      "step": 13405
+    },
+    {
+      "epoch": 0.8145087793912146,
+      "grad_norm": 0.1138419359922409,
+      "learning_rate": 8.306251938038728e-06,
+      "loss": 1.1145,
+      "step": 13406
+    },
+    {
+      "epoch": 0.8145695364238411,
+      "grad_norm": 0.12448260933160782,
+      "learning_rate": 8.300968977653417e-06,
+      "loss": 1.0261,
+      "step": 13407
+    },
+    {
+      "epoch": 0.8146302934564675,
+      "grad_norm": 0.10281818360090256,
+      "learning_rate": 8.295687545754143e-06,
+      "loss": 1.0405,
+      "step": 13408
+    },
+    {
+      "epoch": 0.8146910504890941,
+      "grad_norm": 0.13087216019630432,
+      "learning_rate": 8.29040764253451e-06,
+      "loss": 1.0052,
+      "step": 13409
+    },
+    {
+      "epoch": 0.8147518075217206,
+      "grad_norm": 0.18908251821994781,
+      "learning_rate": 8.285129268188042e-06,
+      "loss": 1.1077,
+      "step": 13410
+    },
+    {
+      "epoch": 0.8148125645543471,
+      "grad_norm": 0.24766699969768524,
+      "learning_rate": 8.279852422908225e-06,
+      "loss": 1.0524,
+      "step": 13411
+    },
+    {
+      "epoch": 0.8148733215869737,
+      "grad_norm": 0.32727837562561035,
+      "learning_rate": 8.274577106888481e-06,
+      "loss": 1.2578,
+      "step": 13412
+    },
+    {
+      "epoch": 0.8149340786196002,
+      "grad_norm": 0.11442811787128448,
+      "learning_rate": 8.26930332032218e-06,
+      "loss": 1.0316,
+      "step": 13413
+    },
+    {
+      "epoch": 0.8149948356522267,
+      "grad_norm": 0.20276670157909393,
+      "learning_rate": 8.264031063402627e-06,
+      "loss": 1.0174,
+      "step": 13414
+    },
+    {
+      "epoch": 0.8150555926848533,
+      "grad_norm": 0.15397687256336212,
+      "learning_rate": 8.25876033632309e-06,
+      "loss": 1.0995,
+      "step": 13415
+    },
+    {
+      "epoch": 0.8151163497174798,
+      "grad_norm": 0.1613105684518814,
+      "learning_rate": 8.253491139276759e-06,
+      "loss": 1.0569,
+      "step": 13416
+    },
+    {
+      "epoch": 0.8151771067501064,
+      "grad_norm": 0.09520558267831802,
+      "learning_rate": 8.248223472456778e-06,
+      "loss": 1.0463,
+      "step": 13417
+    },
+    {
+      "epoch": 0.8152378637827329,
+      "grad_norm": 0.3520737290382385,
+      "learning_rate": 8.24295733605624e-06,
+      "loss": 1.2388,
+      "step": 13418
+    },
+    {
+      "epoch": 0.8152986208153594,
+      "grad_norm": 0.2062932848930359,
+      "learning_rate": 8.23769273026817e-06,
+      "loss": 1.2183,
+      "step": 13419
+    },
+    {
+      "epoch": 0.815359377847986,
+      "grad_norm": 0.1090056300163269,
+      "learning_rate": 8.232429655285529e-06,
+      "loss": 1.0164,
+      "step": 13420
+    },
+    {
+      "epoch": 0.8154201348806124,
+      "grad_norm": 0.11583462357521057,
+      "learning_rate": 8.22716811130127e-06,
+      "loss": 1.0591,
+      "step": 13421
+    },
+    {
+      "epoch": 0.8154808919132389,
+      "grad_norm": 0.31722715497016907,
+      "learning_rate": 8.221908098508235e-06,
+      "loss": 1.0835,
+      "step": 13422
+    },
+    {
+      "epoch": 0.8155416489458654,
+      "grad_norm": 0.15766093134880066,
+      "learning_rate": 8.216649617099248e-06,
+      "loss": 1.0525,
+      "step": 13423
+    },
+    {
+      "epoch": 0.815602405978492,
+      "grad_norm": 0.137874573469162,
+      "learning_rate": 8.211392667267037e-06,
+      "loss": 1.0473,
+      "step": 13424
+    },
+    {
+      "epoch": 0.8156631630111185,
+      "grad_norm": 0.11918137222528458,
+      "learning_rate": 8.206137249204288e-06,
+      "loss": 1.065,
+      "step": 13425
+    },
+    {
+      "epoch": 0.815723920043745,
+      "grad_norm": 0.20145457983016968,
+      "learning_rate": 8.200883363103673e-06,
+      "loss": 1.0876,
+      "step": 13426
+    },
+    {
+      "epoch": 0.8157846770763716,
+      "grad_norm": 1.0978052616119385,
+      "learning_rate": 8.195631009157756e-06,
+      "loss": 1.0407,
+      "step": 13427
+    },
+    {
+      "epoch": 0.8158454341089981,
+      "grad_norm": 0.1237892359495163,
+      "learning_rate": 8.190380187559065e-06,
+      "loss": 1.0636,
+      "step": 13428
+    },
+    {
+      "epoch": 0.8159061911416247,
+      "grad_norm": 0.12328382581472397,
+      "learning_rate": 8.185130898500071e-06,
+      "loss": 1.0438,
+      "step": 13429
+    },
+    {
+      "epoch": 0.8159669481742512,
+      "grad_norm": 0.14856725931167603,
+      "learning_rate": 8.179883142173184e-06,
+      "loss": 1.2112,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8160277052068777,
+      "grad_norm": 0.1367659866809845,
+      "learning_rate": 8.174636918770768e-06,
+      "loss": 0.9977,
+      "step": 13431
+    },
+    {
+      "epoch": 0.8160884622395043,
+      "grad_norm": 0.25983452796936035,
+      "learning_rate": 8.169392228485111e-06,
+      "loss": 1.2778,
+      "step": 13432
+    },
+    {
+      "epoch": 0.8161492192721308,
+      "grad_norm": 0.18509969115257263,
+      "learning_rate": 8.164149071508476e-06,
+      "loss": 1.1337,
+      "step": 13433
+    },
+    {
+      "epoch": 0.8162099763047572,
+      "grad_norm": 0.2002178430557251,
+      "learning_rate": 8.158907448033043e-06,
+      "loss": 1.2526,
+      "step": 13434
+    },
+    {
+      "epoch": 0.8162707333373838,
+      "grad_norm": 0.15770648419857025,
+      "learning_rate": 8.153667358250944e-06,
+      "loss": 1.082,
+      "step": 13435
+    },
+    {
+      "epoch": 0.8163314903700103,
+      "grad_norm": 0.11276135593652725,
+      "learning_rate": 8.148428802354246e-06,
+      "loss": 1.0172,
+      "step": 13436
+    },
+    {
+      "epoch": 0.8163922474026368,
+      "grad_norm": 0.1375294029712677,
+      "learning_rate": 8.143191780535003e-06,
+      "loss": 1.0377,
+      "step": 13437
+    },
+    {
+      "epoch": 0.8164530044352634,
+      "grad_norm": 0.13543906807899475,
+      "learning_rate": 8.137956292985144e-06,
+      "loss": 1.0538,
+      "step": 13438
+    },
+    {
+      "epoch": 0.8165137614678899,
+      "grad_norm": 0.09280017763376236,
+      "learning_rate": 8.132722339896588e-06,
+      "loss": 1.0167,
+      "step": 13439
+    },
+    {
+      "epoch": 0.8165745185005164,
+      "grad_norm": 0.15531599521636963,
+      "learning_rate": 8.127489921461185e-06,
+      "loss": 1.0726,
+      "step": 13440
+    },
+    {
+      "epoch": 0.816635275533143,
+      "grad_norm": 0.19414614140987396,
+      "learning_rate": 8.122259037870728e-06,
+      "loss": 1.1199,
+      "step": 13441
+    },
+    {
+      "epoch": 0.8166960325657695,
+      "grad_norm": 0.15643206238746643,
+      "learning_rate": 8.117029689316962e-06,
+      "loss": 1.0448,
+      "step": 13442
+    },
+    {
+      "epoch": 0.816756789598396,
+      "grad_norm": 0.14813822507858276,
+      "learning_rate": 8.111801875991593e-06,
+      "loss": 1.0673,
+      "step": 13443
+    },
+    {
+      "epoch": 0.8168175466310226,
+      "grad_norm": 0.1387462615966797,
+      "learning_rate": 8.106575598086203e-06,
+      "loss": 1.0505,
+      "step": 13444
+    },
+    {
+      "epoch": 0.8168783036636491,
+      "grad_norm": 0.21125616133213043,
+      "learning_rate": 8.101350855792383e-06,
+      "loss": 1.1712,
+      "step": 13445
+    },
+    {
+      "epoch": 0.8169390606962756,
+      "grad_norm": 2.0209147930145264,
+      "learning_rate": 8.096127649301638e-06,
+      "loss": 1.1521,
+      "step": 13446
+    },
+    {
+      "epoch": 0.8169998177289021,
+      "grad_norm": 0.14098483324050903,
+      "learning_rate": 8.09090597880544e-06,
+      "loss": 1.0115,
+      "step": 13447
+    },
+    {
+      "epoch": 0.8170605747615286,
+      "grad_norm": 0.1621711254119873,
+      "learning_rate": 8.08568584449519e-06,
+      "loss": 1.1544,
+      "step": 13448
+    },
+    {
+      "epoch": 0.8171213317941551,
+      "grad_norm": 0.14340583980083466,
+      "learning_rate": 8.080467246562224e-06,
+      "loss": 1.0983,
+      "step": 13449
+    },
+    {
+      "epoch": 0.8171820888267817,
+      "grad_norm": 0.15204134583473206,
+      "learning_rate": 8.075250185197847e-06,
+      "loss": 1.0353,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8172428458594082,
+      "grad_norm": 3.1373207569122314,
+      "learning_rate": 8.07003466059325e-06,
+      "loss": 1.0434,
+      "step": 13451
+    },
+    {
+      "epoch": 0.8173036028920347,
+      "grad_norm": 0.3946131765842438,
+      "learning_rate": 8.064820672939654e-06,
+      "loss": 1.1012,
+      "step": 13452
+    },
+    {
+      "epoch": 0.8173643599246613,
+      "grad_norm": 0.19499781727790833,
+      "learning_rate": 8.059608222428162e-06,
+      "loss": 1.1419,
+      "step": 13453
+    },
+    {
+      "epoch": 0.8174251169572878,
+      "grad_norm": 0.33928677439689636,
+      "learning_rate": 8.054397309249839e-06,
+      "loss": 1.177,
+      "step": 13454
+    },
+    {
+      "epoch": 0.8174858739899143,
+      "grad_norm": 0.11727161705493927,
+      "learning_rate": 8.049187933595686e-06,
+      "loss": 1.0469,
+      "step": 13455
+    },
+    {
+      "epoch": 0.8175466310225409,
+      "grad_norm": 0.27368858456611633,
+      "learning_rate": 8.043980095656662e-06,
+      "loss": 1.1589,
+      "step": 13456
+    },
+    {
+      "epoch": 0.8176073880551674,
+      "grad_norm": 0.17845216393470764,
+      "learning_rate": 8.038773795623656e-06,
+      "loss": 1.0977,
+      "step": 13457
+    },
+    {
+      "epoch": 0.817668145087794,
+      "grad_norm": 0.19985872507095337,
+      "learning_rate": 8.03356903368751e-06,
+      "loss": 1.1126,
+      "step": 13458
+    },
+    {
+      "epoch": 0.8177289021204205,
+      "grad_norm": 0.1213662400841713,
+      "learning_rate": 8.028365810039002e-06,
+      "loss": 1.0387,
+      "step": 13459
+    },
+    {
+      "epoch": 0.8177896591530469,
+      "grad_norm": 0.16757020354270935,
+      "learning_rate": 8.023164124868855e-06,
+      "loss": 1.1035,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8178504161856734,
+      "grad_norm": 0.12357483059167862,
+      "learning_rate": 8.017963978367742e-06,
+      "loss": 1.0775,
+      "step": 13461
+    },
+    {
+      "epoch": 0.8179111732183,
+      "grad_norm": 0.1485072374343872,
+      "learning_rate": 8.012765370726261e-06,
+      "loss": 1.0058,
+      "step": 13462
+    },
+    {
+      "epoch": 0.8179719302509265,
+      "grad_norm": 0.17385391891002655,
+      "learning_rate": 8.007568302135011e-06,
+      "loss": 1.0988,
+      "step": 13463
+    },
+    {
+      "epoch": 0.818032687283553,
+      "grad_norm": 0.13299430906772614,
+      "learning_rate": 8.002372772784445e-06,
+      "loss": 1.0398,
+      "step": 13464
+    },
+    {
+      "epoch": 0.8180934443161796,
+      "grad_norm": 0.4909656345844269,
+      "learning_rate": 7.997178782865023e-06,
+      "loss": 1.1298,
+      "step": 13465
+    },
+    {
+      "epoch": 0.8181542013488061,
+      "grad_norm": 0.14939147233963013,
+      "learning_rate": 7.99198633256713e-06,
+      "loss": 1.105,
+      "step": 13466
+    },
+    {
+      "epoch": 0.8182149583814327,
+      "grad_norm": 0.16109956800937653,
+      "learning_rate": 7.986795422081089e-06,
+      "loss": 1.129,
+      "step": 13467
+    },
+    {
+      "epoch": 0.8182757154140592,
+      "grad_norm": 0.10842107981443405,
+      "learning_rate": 7.981606051597196e-06,
+      "loss": 0.9968,
+      "step": 13468
+    },
+    {
+      "epoch": 0.8183364724466857,
+      "grad_norm": 0.5431641340255737,
+      "learning_rate": 7.976418221305653e-06,
+      "loss": 1.152,
+      "step": 13469
+    },
+    {
+      "epoch": 0.8183972294793123,
+      "grad_norm": 0.16564422845840454,
+      "learning_rate": 7.971231931396638e-06,
+      "loss": 1.027,
+      "step": 13470
+    },
+    {
+      "epoch": 0.8184579865119388,
+      "grad_norm": 0.18083353340625763,
+      "learning_rate": 7.966047182060226e-06,
+      "loss": 1.0913,
+      "step": 13471
+    },
+    {
+      "epoch": 0.8185187435445653,
+      "grad_norm": 0.13449324667453766,
+      "learning_rate": 7.960863973486466e-06,
+      "loss": 1.0507,
+      "step": 13472
+    },
+    {
+      "epoch": 0.8185795005771919,
+      "grad_norm": 0.1310717612504959,
+      "learning_rate": 7.955682305865381e-06,
+      "loss": 1.0563,
+      "step": 13473
+    },
+    {
+      "epoch": 0.8186402576098183,
+      "grad_norm": 0.20050659775733948,
+      "learning_rate": 7.950502179386882e-06,
+      "loss": 1.1739,
+      "step": 13474
+    },
+    {
+      "epoch": 0.8187010146424448,
+      "grad_norm": 0.16730952262878418,
+      "learning_rate": 7.945323594240856e-06,
+      "loss": 1.0201,
+      "step": 13475
+    },
+    {
+      "epoch": 0.8187617716750714,
+      "grad_norm": 0.10560943186283112,
+      "learning_rate": 7.940146550617123e-06,
+      "loss": 1.0624,
+      "step": 13476
+    },
+    {
+      "epoch": 0.8188225287076979,
+      "grad_norm": 0.1055358275771141,
+      "learning_rate": 7.934971048705447e-06,
+      "loss": 1.0492,
+      "step": 13477
+    },
+    {
+      "epoch": 0.8188832857403244,
+      "grad_norm": 0.11885208636522293,
+      "learning_rate": 7.929797088695539e-06,
+      "loss": 1.0096,
+      "step": 13478
+    },
+    {
+      "epoch": 0.818944042772951,
+      "grad_norm": 0.12103947252035141,
+      "learning_rate": 7.924624670777047e-06,
+      "loss": 1.0342,
+      "step": 13479
+    },
+    {
+      "epoch": 0.8190047998055775,
+      "grad_norm": 0.14237910509109497,
+      "learning_rate": 7.919453795139576e-06,
+      "loss": 1.0502,
+      "step": 13480
+    },
+    {
+      "epoch": 0.819065556838204,
+      "grad_norm": 0.09899695962667465,
+      "learning_rate": 7.91428446197266e-06,
+      "loss": 0.9941,
+      "step": 13481
+    },
+    {
+      "epoch": 0.8191263138708306,
+      "grad_norm": 0.3681263327598572,
+      "learning_rate": 7.909116671465778e-06,
+      "loss": 1.093,
+      "step": 13482
+    },
+    {
+      "epoch": 0.8191870709034571,
+      "grad_norm": 0.3283785581588745,
+      "learning_rate": 7.903950423808365e-06,
+      "loss": 1.1385,
+      "step": 13483
+    },
+    {
+      "epoch": 0.8192478279360836,
+      "grad_norm": 0.1327342689037323,
+      "learning_rate": 7.898785719189784e-06,
+      "loss": 1.0461,
+      "step": 13484
+    },
+    {
+      "epoch": 0.8193085849687102,
+      "grad_norm": 0.12868739664554596,
+      "learning_rate": 7.893622557799346e-06,
+      "loss": 1.0096,
+      "step": 13485
+    },
+    {
+      "epoch": 0.8193693420013367,
+      "grad_norm": 0.3868829905986786,
+      "learning_rate": 7.88846093982632e-06,
+      "loss": 1.0969,
+      "step": 13486
+    },
+    {
+      "epoch": 0.8194300990339631,
+      "grad_norm": 6.048282623291016,
+      "learning_rate": 7.883300865459891e-06,
+      "loss": 1.2119,
+      "step": 13487
+    },
+    {
+      "epoch": 0.8194908560665897,
+      "grad_norm": 0.23110385239124298,
+      "learning_rate": 7.878142334889205e-06,
+      "loss": 1.1701,
+      "step": 13488
+    },
+    {
+      "epoch": 0.8195516130992162,
+      "grad_norm": 0.228868767619133,
+      "learning_rate": 7.872985348303358e-06,
+      "loss": 1.1009,
+      "step": 13489
+    },
+    {
+      "epoch": 0.8196123701318427,
+      "grad_norm": 0.11409421265125275,
+      "learning_rate": 7.867829905891394e-06,
+      "loss": 1.0083,
+      "step": 13490
+    },
+    {
+      "epoch": 0.8196731271644693,
+      "grad_norm": 0.11872085183858871,
+      "learning_rate": 7.862676007842257e-06,
+      "loss": 1.0471,
+      "step": 13491
+    },
+    {
+      "epoch": 0.8197338841970958,
+      "grad_norm": 0.8929527997970581,
+      "learning_rate": 7.857523654344873e-06,
+      "loss": 1.0941,
+      "step": 13492
+    },
+    {
+      "epoch": 0.8197946412297223,
+      "grad_norm": 0.22200007736682892,
+      "learning_rate": 7.85237284558809e-06,
+      "loss": 1.0776,
+      "step": 13493
+    },
+    {
+      "epoch": 0.8198553982623489,
+      "grad_norm": 0.11394982039928436,
+      "learning_rate": 7.847223581760744e-06,
+      "loss": 1.0413,
+      "step": 13494
+    },
+    {
+      "epoch": 0.8199161552949754,
+      "grad_norm": 0.09477784484624863,
+      "learning_rate": 7.842075863051567e-06,
+      "loss": 1.036,
+      "step": 13495
+    },
+    {
+      "epoch": 0.819976912327602,
+      "grad_norm": 0.17983530461788177,
+      "learning_rate": 7.836929689649247e-06,
+      "loss": 1.112,
+      "step": 13496
+    },
+    {
+      "epoch": 0.8200376693602285,
+      "grad_norm": 0.10790320485830307,
+      "learning_rate": 7.831785061742431e-06,
+      "loss": 1.0104,
+      "step": 13497
+    },
+    {
+      "epoch": 0.820098426392855,
+      "grad_norm": 0.6867464184761047,
+      "learning_rate": 7.826641979519655e-06,
+      "loss": 1.0825,
+      "step": 13498
+    },
+    {
+      "epoch": 0.8201591834254816,
+      "grad_norm": 0.19029100239276886,
+      "learning_rate": 7.821500443169487e-06,
+      "loss": 0.9832,
+      "step": 13499
+    },
+    {
+      "epoch": 0.820219940458108,
+      "grad_norm": 0.6250835657119751,
+      "learning_rate": 7.816360452880373e-06,
+      "loss": 1.0707,
+      "step": 13500
+    },
+    {
+      "epoch": 0.8202806974907345,
+      "grad_norm": 0.12136763334274292,
+      "learning_rate": 7.81122200884072e-06,
+      "loss": 1.1161,
+      "step": 13501
+    },
+    {
+      "epoch": 0.820341454523361,
+      "grad_norm": 0.33576223254203796,
+      "learning_rate": 7.80608511123888e-06,
+      "loss": 1.2448,
+      "step": 13502
+    },
+    {
+      "epoch": 0.8204022115559876,
+      "grad_norm": 0.7326078414916992,
+      "learning_rate": 7.800949760263144e-06,
+      "loss": 1.2602,
+      "step": 13503
+    },
+    {
+      "epoch": 0.8204629685886141,
+      "grad_norm": 0.16755549609661102,
+      "learning_rate": 7.795815956101755e-06,
+      "loss": 1.1486,
+      "step": 13504
+    },
+    {
+      "epoch": 0.8205237256212407,
+      "grad_norm": 0.10186265408992767,
+      "learning_rate": 7.790683698942885e-06,
+      "loss": 1.0032,
+      "step": 13505
+    },
+    {
+      "epoch": 0.8205844826538672,
+      "grad_norm": 0.14835381507873535,
+      "learning_rate": 7.785552988974665e-06,
+      "loss": 1.1042,
+      "step": 13506
+    },
+    {
+      "epoch": 0.8206452396864937,
+      "grad_norm": 0.1081235483288765,
+      "learning_rate": 7.780423826385158e-06,
+      "loss": 1.0296,
+      "step": 13507
+    },
+    {
+      "epoch": 0.8207059967191203,
+      "grad_norm": 0.2186422199010849,
+      "learning_rate": 7.775296211362382e-06,
+      "loss": 1.0677,
+      "step": 13508
+    },
+    {
+      "epoch": 0.8207667537517468,
+      "grad_norm": 0.1254095733165741,
+      "learning_rate": 7.770170144094269e-06,
+      "loss": 1.0395,
+      "step": 13509
+    },
+    {
+      "epoch": 0.8208275107843733,
+      "grad_norm": 0.36395975947380066,
+      "learning_rate": 7.765045624768758e-06,
+      "loss": 1.0946,
+      "step": 13510
+    },
+    {
+      "epoch": 0.8208882678169999,
+      "grad_norm": 0.15172605216503143,
+      "learning_rate": 7.759922653573648e-06,
+      "loss": 1.0255,
+      "step": 13511
+    },
+    {
+      "epoch": 0.8209490248496264,
+      "grad_norm": 0.14489217102527618,
+      "learning_rate": 7.754801230696745e-06,
+      "loss": 1.0556,
+      "step": 13512
+    },
+    {
+      "epoch": 0.8210097818822528,
+      "grad_norm": 0.20732025802135468,
+      "learning_rate": 7.74968135632575e-06,
+      "loss": 1.1075,
+      "step": 13513
+    },
+    {
+      "epoch": 0.8210705389148794,
+      "grad_norm": 0.2300223410129547,
+      "learning_rate": 7.744563030648367e-06,
+      "loss": 1.2089,
+      "step": 13514
+    },
+    {
+      "epoch": 0.8211312959475059,
+      "grad_norm": 0.15074793994426727,
+      "learning_rate": 7.739446253852195e-06,
+      "loss": 1.0471,
+      "step": 13515
+    },
+    {
+      "epoch": 0.8211920529801324,
+      "grad_norm": 0.1597336381673813,
+      "learning_rate": 7.734331026124786e-06,
+      "loss": 1.0606,
+      "step": 13516
+    },
+    {
+      "epoch": 0.821252810012759,
+      "grad_norm": 0.14109216630458832,
+      "learning_rate": 7.729217347653655e-06,
+      "loss": 0.9996,
+      "step": 13517
+    },
+    {
+      "epoch": 0.8213135670453855,
+      "grad_norm": 0.10818999260663986,
+      "learning_rate": 7.72410521862621e-06,
+      "loss": 1.0427,
+      "step": 13518
+    },
+    {
+      "epoch": 0.821374324078012,
+      "grad_norm": 0.4802270531654358,
+      "learning_rate": 7.71899463922987e-06,
+      "loss": 1.0507,
+      "step": 13519
+    },
+    {
+      "epoch": 0.8214350811106386,
+      "grad_norm": 0.10484717041254044,
+      "learning_rate": 7.71388560965196e-06,
+      "loss": 0.9752,
+      "step": 13520
+    },
+    {
+      "epoch": 0.8214958381432651,
+      "grad_norm": 0.1392170637845993,
+      "learning_rate": 7.708778130079742e-06,
+      "loss": 1.1028,
+      "step": 13521
+    },
+    {
+      "epoch": 0.8215565951758916,
+      "grad_norm": 0.12246867269277573,
+      "learning_rate": 7.70367220070044e-06,
+      "loss": 1.0444,
+      "step": 13522
+    },
+    {
+      "epoch": 0.8216173522085182,
+      "grad_norm": 0.2541763484477997,
+      "learning_rate": 7.698567821701203e-06,
+      "loss": 1.1154,
+      "step": 13523
+    },
+    {
+      "epoch": 0.8216781092411447,
+      "grad_norm": 0.17111891508102417,
+      "learning_rate": 7.693464993269145e-06,
+      "loss": 1.0929,
+      "step": 13524
+    },
+    {
+      "epoch": 0.8217388662737712,
+      "grad_norm": 0.11510283499956131,
+      "learning_rate": 7.6883637155913e-06,
+      "loss": 1.0316,
+      "step": 13525
+    },
+    {
+      "epoch": 0.8217996233063977,
+      "grad_norm": 0.1308744251728058,
+      "learning_rate": 7.683263988854667e-06,
+      "loss": 1.0596,
+      "step": 13526
+    },
+    {
+      "epoch": 0.8218603803390242,
+      "grad_norm": 0.10676344484090805,
+      "learning_rate": 7.678165813246168e-06,
+      "loss": 1.0318,
+      "step": 13527
+    },
+    {
+      "epoch": 0.8219211373716507,
+      "grad_norm": 0.12430138885974884,
+      "learning_rate": 7.673069188952686e-06,
+      "loss": 1.0768,
+      "step": 13528
+    },
+    {
+      "epoch": 0.8219818944042773,
+      "grad_norm": 0.1866612583398819,
+      "learning_rate": 7.66797411616102e-06,
+      "loss": 1.0705,
+      "step": 13529
+    },
+    {
+      "epoch": 0.8220426514369038,
+      "grad_norm": 0.1155172660946846,
+      "learning_rate": 7.66288059505797e-06,
+      "loss": 1.0221,
+      "step": 13530
+    },
+    {
+      "epoch": 0.8221034084695303,
+      "grad_norm": 0.143506720662117,
+      "learning_rate": 7.657788625830204e-06,
+      "loss": 1.0939,
+      "step": 13531
+    },
+    {
+      "epoch": 0.8221641655021569,
+      "grad_norm": 0.45693644881248474,
+      "learning_rate": 7.652698208664378e-06,
+      "loss": 1.047,
+      "step": 13532
+    },
+    {
+      "epoch": 0.8222249225347834,
+      "grad_norm": 0.16410203278064728,
+      "learning_rate": 7.647609343747086e-06,
+      "loss": 1.1678,
+      "step": 13533
+    },
+    {
+      "epoch": 0.8222856795674099,
+      "grad_norm": 0.1787978559732437,
+      "learning_rate": 7.642522031264849e-06,
+      "loss": 1.0448,
+      "step": 13534
+    },
+    {
+      "epoch": 0.8223464366000365,
+      "grad_norm": 0.13374283909797668,
+      "learning_rate": 7.637436271404168e-06,
+      "loss": 1.0733,
+      "step": 13535
+    },
+    {
+      "epoch": 0.822407193632663,
+      "grad_norm": 0.10483134537935257,
+      "learning_rate": 7.632352064351445e-06,
+      "loss": 1.0516,
+      "step": 13536
+    },
+    {
+      "epoch": 0.8224679506652895,
+      "grad_norm": 0.10286300629377365,
+      "learning_rate": 7.627269410293059e-06,
+      "loss": 1.0518,
+      "step": 13537
+    },
+    {
+      "epoch": 0.8225287076979161,
+      "grad_norm": 0.18806321918964386,
+      "learning_rate": 7.6221883094152955e-06,
+      "loss": 1.216,
+      "step": 13538
+    },
+    {
+      "epoch": 0.8225894647305425,
+      "grad_norm": 0.9146344065666199,
+      "learning_rate": 7.617108761904396e-06,
+      "loss": 1.2358,
+      "step": 13539
+    },
+    {
+      "epoch": 0.822650221763169,
+      "grad_norm": 0.1602848917245865,
+      "learning_rate": 7.612030767946577e-06,
+      "loss": 1.0769,
+      "step": 13540
+    },
+    {
+      "epoch": 0.8227109787957956,
+      "grad_norm": 0.20304131507873535,
+      "learning_rate": 7.606954327727967e-06,
+      "loss": 1.0745,
+      "step": 13541
+    },
+    {
+      "epoch": 0.8227717358284221,
+      "grad_norm": 0.1013583168387413,
+      "learning_rate": 7.601879441434639e-06,
+      "loss": 1.0125,
+      "step": 13542
+    },
+    {
+      "epoch": 0.8228324928610486,
+      "grad_norm": 0.14118294417858124,
+      "learning_rate": 7.596806109252613e-06,
+      "loss": 1.0837,
+      "step": 13543
+    },
+    {
+      "epoch": 0.8228932498936752,
+      "grad_norm": 0.7267357707023621,
+      "learning_rate": 7.591734331367861e-06,
+      "loss": 1.2208,
+      "step": 13544
+    },
+    {
+      "epoch": 0.8229540069263017,
+      "grad_norm": 0.18252035975456238,
+      "learning_rate": 7.586664107966279e-06,
+      "loss": 1.1129,
+      "step": 13545
+    },
+    {
+      "epoch": 0.8230147639589283,
+      "grad_norm": 0.18453197181224823,
+      "learning_rate": 7.58159543923373e-06,
+      "loss": 1.12,
+      "step": 13546
+    },
+    {
+      "epoch": 0.8230755209915548,
+      "grad_norm": 0.24387039244174957,
+      "learning_rate": 7.576528325355991e-06,
+      "loss": 1.0128,
+      "step": 13547
+    },
+    {
+      "epoch": 0.8231362780241813,
+      "grad_norm": 0.3113418221473694,
+      "learning_rate": 7.571462766518811e-06,
+      "loss": 1.0681,
+      "step": 13548
+    },
+    {
+      "epoch": 0.8231970350568079,
+      "grad_norm": 0.4675925672054291,
+      "learning_rate": 7.566398762907861e-06,
+      "loss": 1.1145,
+      "step": 13549
+    },
+    {
+      "epoch": 0.8232577920894344,
+      "grad_norm": 0.16404248774051666,
+      "learning_rate": 7.561336314708773e-06,
+      "loss": 1.039,
+      "step": 13550
+    },
+    {
+      "epoch": 0.8233185491220609,
+      "grad_norm": 0.10805781930685043,
+      "learning_rate": 7.5562754221071005e-06,
+      "loss": 1.0403,
+      "step": 13551
+    },
+    {
+      "epoch": 0.8233793061546874,
+      "grad_norm": 0.257723331451416,
+      "learning_rate": 7.551216085288359e-06,
+      "loss": 1.145,
+      "step": 13552
+    },
+    {
+      "epoch": 0.8234400631873139,
+      "grad_norm": 0.17606984078884125,
+      "learning_rate": 7.546158304437995e-06,
+      "loss": 1.1894,
+      "step": 13553
+    },
+    {
+      "epoch": 0.8235008202199404,
+      "grad_norm": 0.12058721482753754,
+      "learning_rate": 7.541102079741402e-06,
+      "loss": 0.975,
+      "step": 13554
+    },
+    {
+      "epoch": 0.823561577252567,
+      "grad_norm": 0.09002473205327988,
+      "learning_rate": 7.536047411383912e-06,
+      "loss": 0.9875,
+      "step": 13555
+    },
+    {
+      "epoch": 0.8236223342851935,
+      "grad_norm": 0.12132207304239273,
+      "learning_rate": 7.530994299550826e-06,
+      "loss": 1.0886,
+      "step": 13556
+    },
+    {
+      "epoch": 0.82368309131782,
+      "grad_norm": 0.19114543497562408,
+      "learning_rate": 7.5259427444273614e-06,
+      "loss": 1.0621,
+      "step": 13557
+    },
+    {
+      "epoch": 0.8237438483504466,
+      "grad_norm": 1.147022008895874,
+      "learning_rate": 7.520892746198666e-06,
+      "loss": 1.0021,
+      "step": 13558
+    },
+    {
+      "epoch": 0.8238046053830731,
+      "grad_norm": 0.16448919475078583,
+      "learning_rate": 7.5158443050498515e-06,
+      "loss": 1.0586,
+      "step": 13559
+    },
+    {
+      "epoch": 0.8238653624156996,
+      "grad_norm": 0.14355164766311646,
+      "learning_rate": 7.510797421165972e-06,
+      "loss": 1.2218,
+      "step": 13560
+    },
+    {
+      "epoch": 0.8239261194483262,
+      "grad_norm": 0.11638607829809189,
+      "learning_rate": 7.505752094732032e-06,
+      "loss": 1.0586,
+      "step": 13561
+    },
+    {
+      "epoch": 0.8239868764809527,
+      "grad_norm": 0.18874981999397278,
+      "learning_rate": 7.500708325932965e-06,
+      "loss": 1.1447,
+      "step": 13562
+    },
+    {
+      "epoch": 0.8240476335135792,
+      "grad_norm": 0.16098664700984955,
+      "learning_rate": 7.495666114953653e-06,
+      "loss": 1.1337,
+      "step": 13563
+    },
+    {
+      "epoch": 0.8241083905462058,
+      "grad_norm": 0.1381879299879074,
+      "learning_rate": 7.490625461978923e-06,
+      "loss": 1.0181,
+      "step": 13564
+    },
+    {
+      "epoch": 0.8241691475788322,
+      "grad_norm": 0.12066946178674698,
+      "learning_rate": 7.485586367193509e-06,
+      "loss": 1.1,
+      "step": 13565
+    },
+    {
+      "epoch": 0.8242299046114587,
+      "grad_norm": 0.22286900877952576,
+      "learning_rate": 7.480548830782164e-06,
+      "loss": 1.1605,
+      "step": 13566
+    },
+    {
+      "epoch": 0.8242906616440853,
+      "grad_norm": 0.2938409149646759,
+      "learning_rate": 7.475512852929511e-06,
+      "loss": 1.0216,
+      "step": 13567
+    },
+    {
+      "epoch": 0.8243514186767118,
+      "grad_norm": 0.11672308295965195,
+      "learning_rate": 7.470478433820161e-06,
+      "loss": 1.0389,
+      "step": 13568
+    },
+    {
+      "epoch": 0.8244121757093383,
+      "grad_norm": 0.27402544021606445,
+      "learning_rate": 7.4654455736386406e-06,
+      "loss": 1.1318,
+      "step": 13569
+    },
+    {
+      "epoch": 0.8244729327419649,
+      "grad_norm": 21.82569122314453,
+      "learning_rate": 7.460414272569433e-06,
+      "loss": 1.0366,
+      "step": 13570
+    },
+    {
+      "epoch": 0.8245336897745914,
+      "grad_norm": 0.12488722056150436,
+      "learning_rate": 7.455384530796966e-06,
+      "loss": 1.0117,
+      "step": 13571
+    },
+    {
+      "epoch": 0.8245944468072179,
+      "grad_norm": 0.17637819051742554,
+      "learning_rate": 7.450356348505605e-06,
+      "loss": 1.0513,
+      "step": 13572
+    },
+    {
+      "epoch": 0.8246552038398445,
+      "grad_norm": 0.2478485256433487,
+      "learning_rate": 7.445329725879652e-06,
+      "loss": 1.1749,
+      "step": 13573
+    },
+    {
+      "epoch": 0.824715960872471,
+      "grad_norm": 0.1008739322423935,
+      "learning_rate": 7.4403046631033645e-06,
+      "loss": 1.0179,
+      "step": 13574
+    },
+    {
+      "epoch": 0.8247767179050975,
+      "grad_norm": 0.12400525063276291,
+      "learning_rate": 7.435281160360941e-06,
+      "loss": 1.0189,
+      "step": 13575
+    },
+    {
+      "epoch": 0.8248374749377241,
+      "grad_norm": 0.11206913739442825,
+      "learning_rate": 7.4302592178365e-06,
+      "loss": 1.0422,
+      "step": 13576
+    },
+    {
+      "epoch": 0.8248982319703506,
+      "grad_norm": 0.13070032000541687,
+      "learning_rate": 7.425238835714155e-06,
+      "loss": 1.0435,
+      "step": 13577
+    },
+    {
+      "epoch": 0.8249589890029771,
+      "grad_norm": 0.12553338706493378,
+      "learning_rate": 7.420220014177903e-06,
+      "loss": 1.0368,
+      "step": 13578
+    },
+    {
+      "epoch": 0.8250197460356036,
+      "grad_norm": 0.180640310049057,
+      "learning_rate": 7.4152027534117155e-06,
+      "loss": 1.1231,
+      "step": 13579
+    },
+    {
+      "epoch": 0.8250805030682301,
+      "grad_norm": 0.2928134799003601,
+      "learning_rate": 7.410187053599499e-06,
+      "loss": 1.1799,
+      "step": 13580
+    },
+    {
+      "epoch": 0.8251412601008566,
+      "grad_norm": 0.1468118131160736,
+      "learning_rate": 7.405172914925101e-06,
+      "loss": 1.0684,
+      "step": 13581
+    },
+    {
+      "epoch": 0.8252020171334832,
+      "grad_norm": 0.16805818676948547,
+      "learning_rate": 7.400160337572337e-06,
+      "loss": 1.0013,
+      "step": 13582
+    },
+    {
+      "epoch": 0.8252627741661097,
+      "grad_norm": 0.2454906702041626,
+      "learning_rate": 7.395149321724926e-06,
+      "loss": 1.1418,
+      "step": 13583
+    },
+    {
+      "epoch": 0.8253235311987362,
+      "grad_norm": 0.14615805447101593,
+      "learning_rate": 7.390139867566565e-06,
+      "loss": 1.026,
+      "step": 13584
+    },
+    {
+      "epoch": 0.8253842882313628,
+      "grad_norm": 0.10640106350183487,
+      "learning_rate": 7.385131975280851e-06,
+      "loss": 1.0046,
+      "step": 13585
+    },
+    {
+      "epoch": 0.8254450452639893,
+      "grad_norm": 1.4889495372772217,
+      "learning_rate": 7.38012564505135e-06,
+      "loss": 1.0464,
+      "step": 13586
+    },
+    {
+      "epoch": 0.8255058022966159,
+      "grad_norm": 1.630642056465149,
+      "learning_rate": 7.375120877061592e-06,
+      "loss": 1.0621,
+      "step": 13587
+    },
+    {
+      "epoch": 0.8255665593292424,
+      "grad_norm": 0.16819359362125397,
+      "learning_rate": 7.370117671495019e-06,
+      "loss": 1.1247,
+      "step": 13588
+    },
+    {
+      "epoch": 0.8256273163618689,
+      "grad_norm": 0.33381471037864685,
+      "learning_rate": 7.365116028535024e-06,
+      "loss": 1.07,
+      "step": 13589
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 0.12128517776727676,
+      "learning_rate": 7.360115948364943e-06,
+      "loss": 1.0396,
+      "step": 13590
+    },
+    {
+      "epoch": 0.825748830427122,
+      "grad_norm": 0.12619876861572266,
+      "learning_rate": 7.355117431168052e-06,
+      "loss": 1.0541,
+      "step": 13591
+    },
+    {
+      "epoch": 0.8258095874597484,
+      "grad_norm": 0.09805592894554138,
+      "learning_rate": 7.350120477127576e-06,
+      "loss": 1.0304,
+      "step": 13592
+    },
+    {
+      "epoch": 0.825870344492375,
+      "grad_norm": 0.10195305943489075,
+      "learning_rate": 7.345125086426674e-06,
+      "loss": 1.0454,
+      "step": 13593
+    },
+    {
+      "epoch": 0.8259311015250015,
+      "grad_norm": 0.10392196476459503,
+      "learning_rate": 7.340131259248461e-06,
+      "loss": 1.0472,
+      "step": 13594
+    },
+    {
+      "epoch": 0.825991858557628,
+      "grad_norm": 0.13752086460590363,
+      "learning_rate": 7.335138995775981e-06,
+      "loss": 1.1008,
+      "step": 13595
+    },
+    {
+      "epoch": 0.8260526155902546,
+      "grad_norm": 0.11507494747638702,
+      "learning_rate": 7.330148296192229e-06,
+      "loss": 1.004,
+      "step": 13596
+    },
+    {
+      "epoch": 0.8261133726228811,
+      "grad_norm": 0.15243516862392426,
+      "learning_rate": 7.325159160680139e-06,
+      "loss": 1.0588,
+      "step": 13597
+    },
+    {
+      "epoch": 0.8261741296555076,
+      "grad_norm": 0.11210864037275314,
+      "learning_rate": 7.320171589422592e-06,
+      "loss": 1.0571,
+      "step": 13598
+    },
+    {
+      "epoch": 0.8262348866881342,
+      "grad_norm": 0.1640331745147705,
+      "learning_rate": 7.315185582602402e-06,
+      "loss": 1.1551,
+      "step": 13599
+    },
+    {
+      "epoch": 0.8262956437207607,
+      "grad_norm": 0.1828596144914627,
+      "learning_rate": 7.310201140402334e-06,
+      "loss": 1.1405,
+      "step": 13600
+    },
+    {
+      "epoch": 0.8263564007533872,
+      "grad_norm": 0.15175187587738037,
+      "learning_rate": 7.305218263005098e-06,
+      "loss": 1.0813,
+      "step": 13601
+    },
+    {
+      "epoch": 0.8264171577860138,
+      "grad_norm": 0.953859269618988,
+      "learning_rate": 7.300236950593326e-06,
+      "loss": 1.019,
+      "step": 13602
+    },
+    {
+      "epoch": 0.8264779148186403,
+      "grad_norm": 0.14325015246868134,
+      "learning_rate": 7.295257203349632e-06,
+      "loss": 1.0881,
+      "step": 13603
+    },
+    {
+      "epoch": 0.8265386718512668,
+      "grad_norm": 0.15648460388183594,
+      "learning_rate": 7.290279021456553e-06,
+      "loss": 1.0504,
+      "step": 13604
+    },
+    {
+      "epoch": 0.8265994288838933,
+      "grad_norm": 0.19723069667816162,
+      "learning_rate": 7.285302405096539e-06,
+      "loss": 1.1383,
+      "step": 13605
+    },
+    {
+      "epoch": 0.8266601859165198,
+      "grad_norm": 0.13852082192897797,
+      "learning_rate": 7.280327354452027e-06,
+      "loss": 1.051,
+      "step": 13606
+    },
+    {
+      "epoch": 0.8267209429491463,
+      "grad_norm": 0.32211220264434814,
+      "learning_rate": 7.275353869705353e-06,
+      "loss": 1.2097,
+      "step": 13607
+    },
+    {
+      "epoch": 0.8267816999817729,
+      "grad_norm": 0.09656880795955658,
+      "learning_rate": 7.270381951038863e-06,
+      "loss": 1.0389,
+      "step": 13608
+    },
+    {
+      "epoch": 0.8268424570143994,
+      "grad_norm": 0.16907167434692383,
+      "learning_rate": 7.265411598634775e-06,
+      "loss": 1.1242,
+      "step": 13609
+    },
+    {
+      "epoch": 0.8269032140470259,
+      "grad_norm": 1.975132703781128,
+      "learning_rate": 7.260442812675288e-06,
+      "loss": 1.062,
+      "step": 13610
+    },
+    {
+      "epoch": 0.8269639710796525,
+      "grad_norm": 0.24325786530971527,
+      "learning_rate": 7.255475593342531e-06,
+      "loss": 1.2642,
+      "step": 13611
+    },
+    {
+      "epoch": 0.827024728112279,
+      "grad_norm": 0.19007137417793274,
+      "learning_rate": 7.250509940818578e-06,
+      "loss": 1.0715,
+      "step": 13612
+    },
+    {
+      "epoch": 0.8270854851449055,
+      "grad_norm": 0.1458977311849594,
+      "learning_rate": 7.245545855285446e-06,
+      "loss": 1.0858,
+      "step": 13613
+    },
+    {
+      "epoch": 0.8271462421775321,
+      "grad_norm": 0.15047360956668854,
+      "learning_rate": 7.240583336925094e-06,
+      "loss": 1.0298,
+      "step": 13614
+    },
+    {
+      "epoch": 0.8272069992101586,
+      "grad_norm": 0.12451022863388062,
+      "learning_rate": 7.235622385919427e-06,
+      "loss": 1.0188,
+      "step": 13615
+    },
+    {
+      "epoch": 0.8272677562427851,
+      "grad_norm": 0.2222929149866104,
+      "learning_rate": 7.230663002450289e-06,
+      "loss": 1.079,
+      "step": 13616
+    },
+    {
+      "epoch": 0.8273285132754117,
+      "grad_norm": 0.5067652463912964,
+      "learning_rate": 7.22570518669946e-06,
+      "loss": 1.0343,
+      "step": 13617
+    },
+    {
+      "epoch": 0.8273892703080381,
+      "grad_norm": 0.12848037481307983,
+      "learning_rate": 7.220748938848682e-06,
+      "loss": 1.0676,
+      "step": 13618
+    },
+    {
+      "epoch": 0.8274500273406646,
+      "grad_norm": 0.19044119119644165,
+      "learning_rate": 7.215794259079617e-06,
+      "loss": 1.136,
+      "step": 13619
+    },
+    {
+      "epoch": 0.8275107843732912,
+      "grad_norm": 0.1902448534965515,
+      "learning_rate": 7.2108411475738825e-06,
+      "loss": 1.0809,
+      "step": 13620
+    },
+    {
+      "epoch": 0.8275715414059177,
+      "grad_norm": 0.15893591940402985,
+      "learning_rate": 7.2058896045130384e-06,
+      "loss": 1.0257,
+      "step": 13621
+    },
+    {
+      "epoch": 0.8276322984385442,
+      "grad_norm": 0.10953414440155029,
+      "learning_rate": 7.200939630078573e-06,
+      "loss": 1.014,
+      "step": 13622
+    },
+    {
+      "epoch": 0.8276930554711708,
+      "grad_norm": 0.14103396236896515,
+      "learning_rate": 7.195991224451948e-06,
+      "loss": 1.0724,
+      "step": 13623
+    },
+    {
+      "epoch": 0.8277538125037973,
+      "grad_norm": 0.1850893795490265,
+      "learning_rate": 7.19104438781455e-06,
+      "loss": 1.019,
+      "step": 13624
+    },
+    {
+      "epoch": 0.8278145695364238,
+      "grad_norm": 0.21538183093070984,
+      "learning_rate": 7.186099120347683e-06,
+      "loss": 1.0432,
+      "step": 13625
+    },
+    {
+      "epoch": 0.8278753265690504,
+      "grad_norm": 0.2647867202758789,
+      "learning_rate": 7.181155422232633e-06,
+      "loss": 1.2063,
+      "step": 13626
+    },
+    {
+      "epoch": 0.8279360836016769,
+      "grad_norm": 0.11590036749839783,
+      "learning_rate": 7.1762132936505914e-06,
+      "loss": 1.0521,
+      "step": 13627
+    },
+    {
+      "epoch": 0.8279968406343035,
+      "grad_norm": 0.1335471123456955,
+      "learning_rate": 7.171272734782741e-06,
+      "loss": 1.066,
+      "step": 13628
+    },
+    {
+      "epoch": 0.82805759766693,
+      "grad_norm": 0.1113278940320015,
+      "learning_rate": 7.166333745810166e-06,
+      "loss": 1.0212,
+      "step": 13629
+    },
+    {
+      "epoch": 0.8281183546995565,
+      "grad_norm": 0.22371621429920197,
+      "learning_rate": 7.161396326913911e-06,
+      "loss": 1.1098,
+      "step": 13630
+    },
+    {
+      "epoch": 0.828179111732183,
+      "grad_norm": 0.19300615787506104,
+      "learning_rate": 7.156460478274962e-06,
+      "loss": 1.0784,
+      "step": 13631
+    },
+    {
+      "epoch": 0.8282398687648095,
+      "grad_norm": 0.2759637236595154,
+      "learning_rate": 7.1515262000742156e-06,
+      "loss": 1.1484,
+      "step": 13632
+    },
+    {
+      "epoch": 0.828300625797436,
+      "grad_norm": 0.12149547040462494,
+      "learning_rate": 7.146593492492565e-06,
+      "loss": 0.9509,
+      "step": 13633
+    },
+    {
+      "epoch": 0.8283613828300626,
+      "grad_norm": 0.15606677532196045,
+      "learning_rate": 7.141662355710821e-06,
+      "loss": 1.0416,
+      "step": 13634
+    },
+    {
+      "epoch": 0.8284221398626891,
+      "grad_norm": 0.19518235325813293,
+      "learning_rate": 7.136732789909723e-06,
+      "loss": 1.1857,
+      "step": 13635
+    },
+    {
+      "epoch": 0.8284828968953156,
+      "grad_norm": 0.1401486098766327,
+      "learning_rate": 7.1318047952699695e-06,
+      "loss": 1.0899,
+      "step": 13636
+    },
+    {
+      "epoch": 0.8285436539279422,
+      "grad_norm": 0.20847487449645996,
+      "learning_rate": 7.1268783719721955e-06,
+      "loss": 1.1556,
+      "step": 13637
+    },
+    {
+      "epoch": 0.8286044109605687,
+      "grad_norm": 0.6007376909255981,
+      "learning_rate": 7.121953520196989e-06,
+      "loss": 1.2849,
+      "step": 13638
+    },
+    {
+      "epoch": 0.8286651679931952,
+      "grad_norm": 0.18197812139987946,
+      "learning_rate": 7.117030240124855e-06,
+      "loss": 1.2203,
+      "step": 13639
+    },
+    {
+      "epoch": 0.8287259250258218,
+      "grad_norm": 0.13139967620372772,
+      "learning_rate": 7.1121085319362736e-06,
+      "loss": 1.0126,
+      "step": 13640
+    },
+    {
+      "epoch": 0.8287866820584483,
+      "grad_norm": 0.5589107275009155,
+      "learning_rate": 7.107188395811643e-06,
+      "loss": 1.0957,
+      "step": 13641
+    },
+    {
+      "epoch": 0.8288474390910748,
+      "grad_norm": 0.16172543168067932,
+      "learning_rate": 7.102269831931313e-06,
+      "loss": 1.1093,
+      "step": 13642
+    },
+    {
+      "epoch": 0.8289081961237014,
+      "grad_norm": 0.09160090237855911,
+      "learning_rate": 7.097352840475563e-06,
+      "loss": 1.0263,
+      "step": 13643
+    },
+    {
+      "epoch": 0.8289689531563278,
+      "grad_norm": 0.1511775255203247,
+      "learning_rate": 7.092437421624659e-06,
+      "loss": 1.1145,
+      "step": 13644
+    },
+    {
+      "epoch": 0.8290297101889543,
+      "grad_norm": 0.0995846539735794,
+      "learning_rate": 7.087523575558741e-06,
+      "loss": 1.0743,
+      "step": 13645
+    },
+    {
+      "epoch": 0.8290904672215809,
+      "grad_norm": 0.12825095653533936,
+      "learning_rate": 7.082611302457948e-06,
+      "loss": 0.9914,
+      "step": 13646
+    },
+    {
+      "epoch": 0.8291512242542074,
+      "grad_norm": 0.19835519790649414,
+      "learning_rate": 7.07770060250233e-06,
+      "loss": 1.1042,
+      "step": 13647
+    },
+    {
+      "epoch": 0.8292119812868339,
+      "grad_norm": 0.19569192826747894,
+      "learning_rate": 7.0727914758718815e-06,
+      "loss": 1.1402,
+      "step": 13648
+    },
+    {
+      "epoch": 0.8292727383194605,
+      "grad_norm": 0.19548919796943665,
+      "learning_rate": 7.067883922746565e-06,
+      "loss": 1.1665,
+      "step": 13649
+    },
+    {
+      "epoch": 0.829333495352087,
+      "grad_norm": 0.1786556839942932,
+      "learning_rate": 7.06297794330627e-06,
+      "loss": 1.2982,
+      "step": 13650
+    },
+    {
+      "epoch": 0.8293942523847135,
+      "grad_norm": 0.13037210702896118,
+      "learning_rate": 7.0580735377308245e-06,
+      "loss": 1.055,
+      "step": 13651
+    },
+    {
+      "epoch": 0.8294550094173401,
+      "grad_norm": 0.10718393325805664,
+      "learning_rate": 7.053170706199985e-06,
+      "loss": 1.039,
+      "step": 13652
+    },
+    {
+      "epoch": 0.8295157664499666,
+      "grad_norm": 0.1495639830827713,
+      "learning_rate": 7.048269448893463e-06,
+      "loss": 1.0966,
+      "step": 13653
+    },
+    {
+      "epoch": 0.8295765234825931,
+      "grad_norm": 0.15754443407058716,
+      "learning_rate": 7.043369765990943e-06,
+      "loss": 1.1105,
+      "step": 13654
+    },
+    {
+      "epoch": 0.8296372805152197,
+      "grad_norm": 0.12724776566028595,
+      "learning_rate": 7.0384716576720025e-06,
+      "loss": 1.0565,
+      "step": 13655
+    },
+    {
+      "epoch": 0.8296980375478462,
+      "grad_norm": 0.17613442242145538,
+      "learning_rate": 7.033575124116187e-06,
+      "loss": 1.0838,
+      "step": 13656
+    },
+    {
+      "epoch": 0.8297587945804726,
+      "grad_norm": 0.6000800132751465,
+      "learning_rate": 7.028680165502982e-06,
+      "loss": 1.0815,
+      "step": 13657
+    },
+    {
+      "epoch": 0.8298195516130992,
+      "grad_norm": 0.2269129902124405,
+      "learning_rate": 7.0237867820118094e-06,
+      "loss": 1.2047,
+      "step": 13658
+    },
+    {
+      "epoch": 0.8298803086457257,
+      "grad_norm": 0.11749386787414551,
+      "learning_rate": 7.018894973822043e-06,
+      "loss": 1.0573,
+      "step": 13659
+    },
+    {
+      "epoch": 0.8299410656783522,
+      "grad_norm": 0.12844781577587128,
+      "learning_rate": 7.014004741112984e-06,
+      "loss": 1.0787,
+      "step": 13660
+    },
+    {
+      "epoch": 0.8300018227109788,
+      "grad_norm": 0.30096015334129333,
+      "learning_rate": 7.009116084063893e-06,
+      "loss": 1.0929,
+      "step": 13661
+    },
+    {
+      "epoch": 0.8300625797436053,
+      "grad_norm": 1.499553918838501,
+      "learning_rate": 7.004229002853962e-06,
+      "loss": 1.0759,
+      "step": 13662
+    },
+    {
+      "epoch": 0.8301233367762318,
+      "grad_norm": 0.12127718329429626,
+      "learning_rate": 6.99934349766233e-06,
+      "loss": 1.0543,
+      "step": 13663
+    },
+    {
+      "epoch": 0.8301840938088584,
+      "grad_norm": 0.18528977036476135,
+      "learning_rate": 6.994459568668071e-06,
+      "loss": 1.0711,
+      "step": 13664
+    },
+    {
+      "epoch": 0.8302448508414849,
+      "grad_norm": 0.14939551055431366,
+      "learning_rate": 6.989577216050214e-06,
+      "loss": 1.0764,
+      "step": 13665
+    },
+    {
+      "epoch": 0.8303056078741115,
+      "grad_norm": 0.2261781245470047,
+      "learning_rate": 6.984696439987715e-06,
+      "loss": 1.1361,
+      "step": 13666
+    },
+    {
+      "epoch": 0.830366364906738,
+      "grad_norm": 0.12320876121520996,
+      "learning_rate": 6.979817240659481e-06,
+      "loss": 1.0552,
+      "step": 13667
+    },
+    {
+      "epoch": 0.8304271219393645,
+      "grad_norm": 3.8952367305755615,
+      "learning_rate": 6.974939618244364e-06,
+      "loss": 1.0586,
+      "step": 13668
+    },
+    {
+      "epoch": 0.8304878789719911,
+      "grad_norm": 0.10868556797504425,
+      "learning_rate": 6.970063572921137e-06,
+      "loss": 1.0734,
+      "step": 13669
+    },
+    {
+      "epoch": 0.8305486360046175,
+      "grad_norm": 0.11588618904352188,
+      "learning_rate": 6.965189104868564e-06,
+      "loss": 1.0109,
+      "step": 13670
+    },
+    {
+      "epoch": 0.830609393037244,
+      "grad_norm": 0.10129258036613464,
+      "learning_rate": 6.960316214265311e-06,
+      "loss": 1.0545,
+      "step": 13671
+    },
+    {
+      "epoch": 0.8306701500698705,
+      "grad_norm": 0.3364144563674927,
+      "learning_rate": 6.9554449012899755e-06,
+      "loss": 1.1406,
+      "step": 13672
+    },
+    {
+      "epoch": 0.8307309071024971,
+      "grad_norm": 0.19322070479393005,
+      "learning_rate": 6.950575166121126e-06,
+      "loss": 1.1751,
+      "step": 13673
+    },
+    {
+      "epoch": 0.8307916641351236,
+      "grad_norm": 0.10311407595872879,
+      "learning_rate": 6.9457070089372575e-06,
+      "loss": 1.0058,
+      "step": 13674
+    },
+    {
+      "epoch": 0.8308524211677502,
+      "grad_norm": 0.1127227172255516,
+      "learning_rate": 6.940840429916828e-06,
+      "loss": 0.9993,
+      "step": 13675
+    },
+    {
+      "epoch": 0.8309131782003767,
+      "grad_norm": 0.2369873970746994,
+      "learning_rate": 6.935975429238218e-06,
+      "loss": 1.129,
+      "step": 13676
+    },
+    {
+      "epoch": 0.8309739352330032,
+      "grad_norm": 0.13883095979690552,
+      "learning_rate": 6.931112007079749e-06,
+      "loss": 1.0715,
+      "step": 13677
+    },
+    {
+      "epoch": 0.8310346922656298,
+      "grad_norm": 0.22597213089466095,
+      "learning_rate": 6.926250163619713e-06,
+      "loss": 1.0718,
+      "step": 13678
+    },
+    {
+      "epoch": 0.8310954492982563,
+      "grad_norm": 0.1657181680202484,
+      "learning_rate": 6.9213898990362736e-06,
+      "loss": 1.0405,
+      "step": 13679
+    },
+    {
+      "epoch": 0.8311562063308828,
+      "grad_norm": 0.3855338990688324,
+      "learning_rate": 6.916531213507626e-06,
+      "loss": 1.1183,
+      "step": 13680
+    },
+    {
+      "epoch": 0.8312169633635094,
+      "grad_norm": 0.12398028373718262,
+      "learning_rate": 6.911674107211858e-06,
+      "loss": 1.0507,
+      "step": 13681
+    },
+    {
+      "epoch": 0.8312777203961359,
+      "grad_norm": 0.16409063339233398,
+      "learning_rate": 6.906818580327007e-06,
+      "loss": 1.0393,
+      "step": 13682
+    },
+    {
+      "epoch": 0.8313384774287624,
+      "grad_norm": 0.18331725895404816,
+      "learning_rate": 6.901964633031044e-06,
+      "loss": 1.02,
+      "step": 13683
+    },
+    {
+      "epoch": 0.8313992344613889,
+      "grad_norm": 0.15774329006671906,
+      "learning_rate": 6.897112265501904e-06,
+      "loss": 1.0643,
+      "step": 13684
+    },
+    {
+      "epoch": 0.8314599914940154,
+      "grad_norm": 0.9368786215782166,
+      "learning_rate": 6.892261477917444e-06,
+      "loss": 1.204,
+      "step": 13685
+    },
+    {
+      "epoch": 0.8315207485266419,
+      "grad_norm": 0.26125773787498474,
+      "learning_rate": 6.8874122704554676e-06,
+      "loss": 1.1126,
+      "step": 13686
+    },
+    {
+      "epoch": 0.8315815055592685,
+      "grad_norm": 0.4213768541812897,
+      "learning_rate": 6.882564643293732e-06,
+      "loss": 1.107,
+      "step": 13687
+    },
+    {
+      "epoch": 0.831642262591895,
+      "grad_norm": 0.1398756057024002,
+      "learning_rate": 6.8777185966099225e-06,
+      "loss": 1.0566,
+      "step": 13688
+    },
+    {
+      "epoch": 0.8317030196245215,
+      "grad_norm": 0.14308799803256989,
+      "learning_rate": 6.872874130581674e-06,
+      "loss": 1.101,
+      "step": 13689
+    },
+    {
+      "epoch": 0.8317637766571481,
+      "grad_norm": 0.21111813187599182,
+      "learning_rate": 6.868031245386547e-06,
+      "loss": 1.1003,
+      "step": 13690
+    },
+    {
+      "epoch": 0.8318245336897746,
+      "grad_norm": 0.2287217676639557,
+      "learning_rate": 6.8631899412020974e-06,
+      "loss": 1.0466,
+      "step": 13691
+    },
+    {
+      "epoch": 0.8318852907224011,
+      "grad_norm": 0.19451963901519775,
+      "learning_rate": 6.858350218205745e-06,
+      "loss": 1.0643,
+      "step": 13692
+    },
+    {
+      "epoch": 0.8319460477550277,
+      "grad_norm": 0.157549649477005,
+      "learning_rate": 6.853512076574908e-06,
+      "loss": 1.0906,
+      "step": 13693
+    },
+    {
+      "epoch": 0.8320068047876542,
+      "grad_norm": 0.28237295150756836,
+      "learning_rate": 6.848675516486924e-06,
+      "loss": 1.1226,
+      "step": 13694
+    },
+    {
+      "epoch": 0.8320675618202807,
+      "grad_norm": 0.12846878170967102,
+      "learning_rate": 6.843840538119067e-06,
+      "loss": 1.0677,
+      "step": 13695
+    },
+    {
+      "epoch": 0.8321283188529073,
+      "grad_norm": 0.09501422941684723,
+      "learning_rate": 6.8390071416485905e-06,
+      "loss": 1.0524,
+      "step": 13696
+    },
+    {
+      "epoch": 0.8321890758855337,
+      "grad_norm": 0.17965899407863617,
+      "learning_rate": 6.834175327252651e-06,
+      "loss": 1.0786,
+      "step": 13697
+    },
+    {
+      "epoch": 0.8322498329181602,
+      "grad_norm": 0.1301775574684143,
+      "learning_rate": 6.829345095108364e-06,
+      "loss": 1.0571,
+      "step": 13698
+    },
+    {
+      "epoch": 0.8323105899507868,
+      "grad_norm": 0.23039108514785767,
+      "learning_rate": 6.82451644539277e-06,
+      "loss": 1.1818,
+      "step": 13699
+    },
+    {
+      "epoch": 0.8323713469834133,
+      "grad_norm": 0.2390698790550232,
+      "learning_rate": 6.819689378282862e-06,
+      "loss": 1.0777,
+      "step": 13700
+    },
+    {
+      "epoch": 0.8324321040160398,
+      "grad_norm": 0.16512732207775116,
+      "learning_rate": 6.814863893955598e-06,
+      "loss": 1.0842,
+      "step": 13701
+    },
+    {
+      "epoch": 0.8324928610486664,
+      "grad_norm": 0.4279487729072571,
+      "learning_rate": 6.810039992587841e-06,
+      "loss": 1.0516,
+      "step": 13702
+    },
+    {
+      "epoch": 0.8325536180812929,
+      "grad_norm": 0.23650000989437103,
+      "learning_rate": 6.805217674356418e-06,
+      "loss": 1.1109,
+      "step": 13703
+    },
+    {
+      "epoch": 0.8326143751139194,
+      "grad_norm": 0.12996117770671844,
+      "learning_rate": 6.8003969394380915e-06,
+      "loss": 0.9953,
+      "step": 13704
+    },
+    {
+      "epoch": 0.832675132146546,
+      "grad_norm": 1.8573832511901855,
+      "learning_rate": 6.7955777880095665e-06,
+      "loss": 1.0391,
+      "step": 13705
+    },
+    {
+      "epoch": 0.8327358891791725,
+      "grad_norm": 0.10993792116641998,
+      "learning_rate": 6.790760220247488e-06,
+      "loss": 1.0667,
+      "step": 13706
+    },
+    {
+      "epoch": 0.832796646211799,
+      "grad_norm": 0.1834808737039566,
+      "learning_rate": 6.785944236328445e-06,
+      "loss": 1.1014,
+      "step": 13707
+    },
+    {
+      "epoch": 0.8328574032444256,
+      "grad_norm": 0.11124822497367859,
+      "learning_rate": 6.7811298364289675e-06,
+      "loss": 1.0246,
+      "step": 13708
+    },
+    {
+      "epoch": 0.8329181602770521,
+      "grad_norm": 0.16302169859409332,
+      "learning_rate": 6.776317020725531e-06,
+      "loss": 1.0169,
+      "step": 13709
+    },
+    {
+      "epoch": 0.8329789173096785,
+      "grad_norm": 0.6890310645103455,
+      "learning_rate": 6.77150578939455e-06,
+      "loss": 1.0296,
+      "step": 13710
+    },
+    {
+      "epoch": 0.8330396743423051,
+      "grad_norm": 0.11164145916700363,
+      "learning_rate": 6.766696142612383e-06,
+      "loss": 1.0556,
+      "step": 13711
+    },
+    {
+      "epoch": 0.8331004313749316,
+      "grad_norm": 0.12510815262794495,
+      "learning_rate": 6.761888080555323e-06,
+      "loss": 1.0287,
+      "step": 13712
+    },
+    {
+      "epoch": 0.8331611884075582,
+      "grad_norm": 0.31896454095840454,
+      "learning_rate": 6.757081603399612e-06,
+      "loss": 1.0798,
+      "step": 13713
+    },
+    {
+      "epoch": 0.8332219454401847,
+      "grad_norm": 0.2016814947128296,
+      "learning_rate": 6.752276711321432e-06,
+      "loss": 1.2079,
+      "step": 13714
+    },
+    {
+      "epoch": 0.8332827024728112,
+      "grad_norm": 0.2229379564523697,
+      "learning_rate": 6.7474734044969016e-06,
+      "loss": 1.1087,
+      "step": 13715
+    },
+    {
+      "epoch": 0.8333434595054378,
+      "grad_norm": 0.2151334434747696,
+      "learning_rate": 6.742671683102103e-06,
+      "loss": 1.1717,
+      "step": 13716
+    },
+    {
+      "epoch": 0.8334042165380643,
+      "grad_norm": 0.19906583428382874,
+      "learning_rate": 6.737871547313035e-06,
+      "loss": 1.0985,
+      "step": 13717
+    },
+    {
+      "epoch": 0.8334649735706908,
+      "grad_norm": 0.11293461918830872,
+      "learning_rate": 6.733072997305661e-06,
+      "loss": 0.9888,
+      "step": 13718
+    },
+    {
+      "epoch": 0.8335257306033174,
+      "grad_norm": 0.09920281916856766,
+      "learning_rate": 6.728276033255848e-06,
+      "loss": 1.0293,
+      "step": 13719
+    },
+    {
+      "epoch": 0.8335864876359439,
+      "grad_norm": 0.1459367573261261,
+      "learning_rate": 6.723480655339431e-06,
+      "loss": 1.0148,
+      "step": 13720
+    },
+    {
+      "epoch": 0.8336472446685704,
+      "grad_norm": 0.23297899961471558,
+      "learning_rate": 6.718686863732204e-06,
+      "loss": 1.128,
+      "step": 13721
+    },
+    {
+      "epoch": 0.833708001701197,
+      "grad_norm": 0.17642824351787567,
+      "learning_rate": 6.713894658609876e-06,
+      "loss": 1.0808,
+      "step": 13722
+    },
+    {
+      "epoch": 0.8337687587338234,
+      "grad_norm": 0.12494087219238281,
+      "learning_rate": 6.709104040148106e-06,
+      "loss": 0.9638,
+      "step": 13723
+    },
+    {
+      "epoch": 0.8338295157664499,
+      "grad_norm": 0.22081853449344635,
+      "learning_rate": 6.7043150085225e-06,
+      "loss": 1.1811,
+      "step": 13724
+    },
+    {
+      "epoch": 0.8338902727990765,
+      "grad_norm": 0.24359643459320068,
+      "learning_rate": 6.699527563908592e-06,
+      "loss": 1.2602,
+      "step": 13725
+    },
+    {
+      "epoch": 0.833951029831703,
+      "grad_norm": 0.10297096520662308,
+      "learning_rate": 6.694741706481872e-06,
+      "loss": 1.0032,
+      "step": 13726
+    },
+    {
+      "epoch": 0.8340117868643295,
+      "grad_norm": 0.0980038195848465,
+      "learning_rate": 6.689957436417766e-06,
+      "loss": 1.0307,
+      "step": 13727
+    },
+    {
+      "epoch": 0.8340725438969561,
+      "grad_norm": 0.14997591078281403,
+      "learning_rate": 6.685174753891643e-06,
+      "loss": 1.0007,
+      "step": 13728
+    },
+    {
+      "epoch": 0.8341333009295826,
+      "grad_norm": 0.20200970768928528,
+      "learning_rate": 6.68039365907881e-06,
+      "loss": 1.1032,
+      "step": 13729
+    },
+    {
+      "epoch": 0.8341940579622091,
+      "grad_norm": 0.2090281993150711,
+      "learning_rate": 6.675614152154519e-06,
+      "loss": 1.1025,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8342548149948357,
+      "grad_norm": 0.19217103719711304,
+      "learning_rate": 6.67083623329397e-06,
+      "loss": 1.0885,
+      "step": 13731
+    },
+    {
+      "epoch": 0.8343155720274622,
+      "grad_norm": 0.10333271324634552,
+      "learning_rate": 6.666059902672295e-06,
+      "loss": 1.0105,
+      "step": 13732
+    },
+    {
+      "epoch": 0.8343763290600887,
+      "grad_norm": 0.15675486624240875,
+      "learning_rate": 6.661285160464564e-06,
+      "loss": 1.0837,
+      "step": 13733
+    },
+    {
+      "epoch": 0.8344370860927153,
+      "grad_norm": 0.10371602326631546,
+      "learning_rate": 6.656512006845805e-06,
+      "loss": 1.0317,
+      "step": 13734
+    },
+    {
+      "epoch": 0.8344978431253418,
+      "grad_norm": 0.24126096069812775,
+      "learning_rate": 6.6517404419909835e-06,
+      "loss": 1.108,
+      "step": 13735
+    },
+    {
+      "epoch": 0.8345586001579682,
+      "grad_norm": 0.1097838431596756,
+      "learning_rate": 6.646970466074975e-06,
+      "loss": 0.9996,
+      "step": 13736
+    },
+    {
+      "epoch": 0.8346193571905948,
+      "grad_norm": 0.1136871725320816,
+      "learning_rate": 6.642202079272658e-06,
+      "loss": 1.0573,
+      "step": 13737
+    },
+    {
+      "epoch": 0.8346801142232213,
+      "grad_norm": 0.17664246261119843,
+      "learning_rate": 6.637435281758819e-06,
+      "loss": 1.1226,
+      "step": 13738
+    },
+    {
+      "epoch": 0.8347408712558478,
+      "grad_norm": 0.19927391409873962,
+      "learning_rate": 6.632670073708158e-06,
+      "loss": 1.0615,
+      "step": 13739
+    },
+    {
+      "epoch": 0.8348016282884744,
+      "grad_norm": 0.17190268635749817,
+      "learning_rate": 6.627906455295358e-06,
+      "loss": 1.0155,
+      "step": 13740
+    },
+    {
+      "epoch": 0.8348623853211009,
+      "grad_norm": 0.10104475915431976,
+      "learning_rate": 6.623144426695016e-06,
+      "loss": 1.001,
+      "step": 13741
+    },
+    {
+      "epoch": 0.8349231423537274,
+      "grad_norm": 0.11691311001777649,
+      "learning_rate": 6.618383988081717e-06,
+      "loss": 1.0219,
+      "step": 13742
+    },
+    {
+      "epoch": 0.834983899386354,
+      "grad_norm": 0.16260352730751038,
+      "learning_rate": 6.613625139629936e-06,
+      "loss": 1.0935,
+      "step": 13743
+    },
+    {
+      "epoch": 0.8350446564189805,
+      "grad_norm": 0.16674138605594635,
+      "learning_rate": 6.608867881514108e-06,
+      "loss": 1.0991,
+      "step": 13744
+    },
+    {
+      "epoch": 0.835105413451607,
+      "grad_norm": 0.10440758615732193,
+      "learning_rate": 6.604112213908631e-06,
+      "loss": 1.0249,
+      "step": 13745
+    },
+    {
+      "epoch": 0.8351661704842336,
+      "grad_norm": 0.11826813966035843,
+      "learning_rate": 6.5993581369877874e-06,
+      "loss": 0.995,
+      "step": 13746
+    },
+    {
+      "epoch": 0.8352269275168601,
+      "grad_norm": 0.18261979520320892,
+      "learning_rate": 6.594605650925867e-06,
+      "loss": 0.9989,
+      "step": 13747
+    },
+    {
+      "epoch": 0.8352876845494867,
+      "grad_norm": 0.12629005312919617,
+      "learning_rate": 6.589854755897068e-06,
+      "loss": 1.0658,
+      "step": 13748
+    },
+    {
+      "epoch": 0.8353484415821131,
+      "grad_norm": 0.10424727201461792,
+      "learning_rate": 6.585105452075535e-06,
+      "loss": 1.0041,
+      "step": 13749
+    },
+    {
+      "epoch": 0.8354091986147396,
+      "grad_norm": 0.22772297263145447,
+      "learning_rate": 6.580357739635357e-06,
+      "loss": 1.3067,
+      "step": 13750
+    },
+    {
+      "epoch": 0.8354699556473661,
+      "grad_norm": 0.1549176573753357,
+      "learning_rate": 6.575611618750555e-06,
+      "loss": 1.0807,
+      "step": 13751
+    },
+    {
+      "epoch": 0.8355307126799927,
+      "grad_norm": 0.11761268973350525,
+      "learning_rate": 6.5708670895951e-06,
+      "loss": 1.0924,
+      "step": 13752
+    },
+    {
+      "epoch": 0.8355914697126192,
+      "grad_norm": 0.10721135884523392,
+      "learning_rate": 6.566124152342912e-06,
+      "loss": 1.0559,
+      "step": 13753
+    },
+    {
+      "epoch": 0.8356522267452458,
+      "grad_norm": 0.12835678458213806,
+      "learning_rate": 6.5613828071678374e-06,
+      "loss": 1.0451,
+      "step": 13754
+    },
+    {
+      "epoch": 0.8357129837778723,
+      "grad_norm": 0.18677636981010437,
+      "learning_rate": 6.556643054243672e-06,
+      "loss": 1.0333,
+      "step": 13755
+    },
+    {
+      "epoch": 0.8357737408104988,
+      "grad_norm": 0.10661299526691437,
+      "learning_rate": 6.551904893744154e-06,
+      "loss": 1.0408,
+      "step": 13756
+    },
+    {
+      "epoch": 0.8358344978431254,
+      "grad_norm": 0.6153198480606079,
+      "learning_rate": 6.547168325842945e-06,
+      "loss": 1.0523,
+      "step": 13757
+    },
+    {
+      "epoch": 0.8358952548757519,
+      "grad_norm": 0.12669655680656433,
+      "learning_rate": 6.542433350713706e-06,
+      "loss": 1.026,
+      "step": 13758
+    },
+    {
+      "epoch": 0.8359560119083784,
+      "grad_norm": 0.14585493505001068,
+      "learning_rate": 6.537699968529965e-06,
+      "loss": 1.0532,
+      "step": 13759
+    },
+    {
+      "epoch": 0.836016768941005,
+      "grad_norm": 0.18196462094783783,
+      "learning_rate": 6.532968179465227e-06,
+      "loss": 1.1514,
+      "step": 13760
+    },
+    {
+      "epoch": 0.8360775259736315,
+      "grad_norm": 0.15091489255428314,
+      "learning_rate": 6.528237983692942e-06,
+      "loss": 1.0346,
+      "step": 13761
+    },
+    {
+      "epoch": 0.8361382830062579,
+      "grad_norm": 0.20028628408908844,
+      "learning_rate": 6.523509381386489e-06,
+      "loss": 1.0668,
+      "step": 13762
+    },
+    {
+      "epoch": 0.8361990400388845,
+      "grad_norm": 0.1909007877111435,
+      "learning_rate": 6.5187823727192125e-06,
+      "loss": 0.994,
+      "step": 13763
+    },
+    {
+      "epoch": 0.836259797071511,
+      "grad_norm": 9.37720012664795,
+      "learning_rate": 6.514056957864373e-06,
+      "loss": 1.314,
+      "step": 13764
+    },
+    {
+      "epoch": 0.8363205541041375,
+      "grad_norm": 0.13656282424926758,
+      "learning_rate": 6.509333136995194e-06,
+      "loss": 0.9967,
+      "step": 13765
+    },
+    {
+      "epoch": 0.8363813111367641,
+      "grad_norm": 0.13284167647361755,
+      "learning_rate": 6.504610910284803e-06,
+      "loss": 1.0325,
+      "step": 13766
+    },
+    {
+      "epoch": 0.8364420681693906,
+      "grad_norm": 0.13866139948368073,
+      "learning_rate": 6.499890277906295e-06,
+      "loss": 1.0621,
+      "step": 13767
+    },
+    {
+      "epoch": 0.8365028252020171,
+      "grad_norm": 0.10912806540727615,
+      "learning_rate": 6.495171240032733e-06,
+      "loss": 1.0505,
+      "step": 13768
+    },
+    {
+      "epoch": 0.8365635822346437,
+      "grad_norm": 0.14385144412517548,
+      "learning_rate": 6.490453796837071e-06,
+      "loss": 1.0154,
+      "step": 13769
+    },
+    {
+      "epoch": 0.8366243392672702,
+      "grad_norm": 0.19134403765201569,
+      "learning_rate": 6.4857379484922375e-06,
+      "loss": 1.1225,
+      "step": 13770
+    },
+    {
+      "epoch": 0.8366850962998967,
+      "grad_norm": 0.4064330458641052,
+      "learning_rate": 6.481023695171107e-06,
+      "loss": 1.0581,
+      "step": 13771
+    },
+    {
+      "epoch": 0.8367458533325233,
+      "grad_norm": 0.12208230793476105,
+      "learning_rate": 6.47631103704644e-06,
+      "loss": 1.0183,
+      "step": 13772
+    },
+    {
+      "epoch": 0.8368066103651498,
+      "grad_norm": 0.09965897351503372,
+      "learning_rate": 6.471599974291009e-06,
+      "loss": 1.0188,
+      "step": 13773
+    },
+    {
+      "epoch": 0.8368673673977763,
+      "grad_norm": 0.2437160164117813,
+      "learning_rate": 6.4668905070775055e-06,
+      "loss": 1.125,
+      "step": 13774
+    },
+    {
+      "epoch": 0.8369281244304028,
+      "grad_norm": 0.10612661391496658,
+      "learning_rate": 6.46218263557854e-06,
+      "loss": 1.022,
+      "step": 13775
+    },
+    {
+      "epoch": 0.8369888814630293,
+      "grad_norm": 0.12046092748641968,
+      "learning_rate": 6.4574763599666856e-06,
+      "loss": 1.0311,
+      "step": 13776
+    },
+    {
+      "epoch": 0.8370496384956558,
+      "grad_norm": 0.16663911938667297,
+      "learning_rate": 6.452771680414449e-06,
+      "loss": 1.1517,
+      "step": 13777
+    },
+    {
+      "epoch": 0.8371103955282824,
+      "grad_norm": 0.2208757847547531,
+      "learning_rate": 6.448068597094292e-06,
+      "loss": 1.1361,
+      "step": 13778
+    },
+    {
+      "epoch": 0.8371711525609089,
+      "grad_norm": 0.1213504746556282,
+      "learning_rate": 6.443367110178594e-06,
+      "loss": 1.064,
+      "step": 13779
+    },
+    {
+      "epoch": 0.8372319095935354,
+      "grad_norm": 0.16529811918735504,
+      "learning_rate": 6.438667219839689e-06,
+      "loss": 1.0732,
+      "step": 13780
+    },
+    {
+      "epoch": 0.837292666626162,
+      "grad_norm": 0.18973691761493683,
+      "learning_rate": 6.433968926249867e-06,
+      "loss": 1.0192,
+      "step": 13781
+    },
+    {
+      "epoch": 0.8373534236587885,
+      "grad_norm": 2.0302512645721436,
+      "learning_rate": 6.429272229581329e-06,
+      "loss": 1.091,
+      "step": 13782
+    },
+    {
+      "epoch": 0.837414180691415,
+      "grad_norm": 0.1460111439228058,
+      "learning_rate": 6.424577130006232e-06,
+      "loss": 1.0026,
+      "step": 13783
+    },
+    {
+      "epoch": 0.8374749377240416,
+      "grad_norm": 0.177422896027565,
+      "learning_rate": 6.419883627696688e-06,
+      "loss": 1.1193,
+      "step": 13784
+    },
+    {
+      "epoch": 0.8375356947566681,
+      "grad_norm": 0.11280520260334015,
+      "learning_rate": 6.415191722824749e-06,
+      "loss": 1.028,
+      "step": 13785
+    },
+    {
+      "epoch": 0.8375964517892946,
+      "grad_norm": 0.12574908137321472,
+      "learning_rate": 6.41050141556237e-06,
+      "loss": 1.0779,
+      "step": 13786
+    },
+    {
+      "epoch": 0.8376572088219212,
+      "grad_norm": 0.13504955172538757,
+      "learning_rate": 6.405812706081488e-06,
+      "loss": 1.0827,
+      "step": 13787
+    },
+    {
+      "epoch": 0.8377179658545477,
+      "grad_norm": 0.11524248868227005,
+      "learning_rate": 6.401125594553959e-06,
+      "loss": 1.0086,
+      "step": 13788
+    },
+    {
+      "epoch": 0.8377787228871741,
+      "grad_norm": 0.22904406487941742,
+      "learning_rate": 6.396440081151606e-06,
+      "loss": 1.161,
+      "step": 13789
+    },
+    {
+      "epoch": 0.8378394799198007,
+      "grad_norm": 0.12628456950187683,
+      "learning_rate": 6.391756166046176e-06,
+      "loss": 1.0432,
+      "step": 13790
+    },
+    {
+      "epoch": 0.8379002369524272,
+      "grad_norm": 0.17692239582538605,
+      "learning_rate": 6.387073849409347e-06,
+      "loss": 1.0568,
+      "step": 13791
+    },
+    {
+      "epoch": 0.8379609939850537,
+      "grad_norm": 0.13144269585609436,
+      "learning_rate": 6.3823931314127675e-06,
+      "loss": 1.0635,
+      "step": 13792
+    },
+    {
+      "epoch": 0.8380217510176803,
+      "grad_norm": 0.17531993985176086,
+      "learning_rate": 6.377714012227981e-06,
+      "loss": 1.0814,
+      "step": 13793
+    },
+    {
+      "epoch": 0.8380825080503068,
+      "grad_norm": 0.2876814901828766,
+      "learning_rate": 6.3730364920265295e-06,
+      "loss": 1.1806,
+      "step": 13794
+    },
+    {
+      "epoch": 0.8381432650829334,
+      "grad_norm": 0.953540563583374,
+      "learning_rate": 6.3683605709798534e-06,
+      "loss": 1.0599,
+      "step": 13795
+    },
+    {
+      "epoch": 0.8382040221155599,
+      "grad_norm": 0.14210133254528046,
+      "learning_rate": 6.36368624925936e-06,
+      "loss": 1.0592,
+      "step": 13796
+    },
+    {
+      "epoch": 0.8382647791481864,
+      "grad_norm": 0.10907606780529022,
+      "learning_rate": 6.3590135270363784e-06,
+      "loss": 1.0734,
+      "step": 13797
+    },
+    {
+      "epoch": 0.838325536180813,
+      "grad_norm": 0.15380249917507172,
+      "learning_rate": 6.354342404482194e-06,
+      "loss": 1.0379,
+      "step": 13798
+    },
+    {
+      "epoch": 0.8383862932134395,
+      "grad_norm": 0.490646630525589,
+      "learning_rate": 6.349672881768021e-06,
+      "loss": 1.1746,
+      "step": 13799
+    },
+    {
+      "epoch": 0.838447050246066,
+      "grad_norm": 0.24855506420135498,
+      "learning_rate": 6.345004959065032e-06,
+      "loss": 1.0777,
+      "step": 13800
+    },
+    {
+      "epoch": 0.8385078072786926,
+      "grad_norm": 0.14284101128578186,
+      "learning_rate": 6.340338636544319e-06,
+      "loss": 1.0026,
+      "step": 13801
+    },
+    {
+      "epoch": 0.838568564311319,
+      "grad_norm": 0.09635406732559204,
+      "learning_rate": 6.335673914376938e-06,
+      "loss": 1.0234,
+      "step": 13802
+    },
+    {
+      "epoch": 0.8386293213439455,
+      "grad_norm": 0.19218318164348602,
+      "learning_rate": 6.331010792733866e-06,
+      "loss": 1.1586,
+      "step": 13803
+    },
+    {
+      "epoch": 0.8386900783765721,
+      "grad_norm": 0.1273650974035263,
+      "learning_rate": 6.326349271786025e-06,
+      "loss": 1.0265,
+      "step": 13804
+    },
+    {
+      "epoch": 0.8387508354091986,
+      "grad_norm": 0.15145963430404663,
+      "learning_rate": 6.321689351704313e-06,
+      "loss": 1.017,
+      "step": 13805
+    },
+    {
+      "epoch": 0.8388115924418251,
+      "grad_norm": 0.19609273970127106,
+      "learning_rate": 6.31703103265951e-06,
+      "loss": 1.1041,
+      "step": 13806
+    },
+    {
+      "epoch": 0.8388723494744517,
+      "grad_norm": 0.21271000802516937,
+      "learning_rate": 6.312374314822378e-06,
+      "loss": 1.2021,
+      "step": 13807
+    },
+    {
+      "epoch": 0.8389331065070782,
+      "grad_norm": 0.11968686431646347,
+      "learning_rate": 6.307719198363615e-06,
+      "loss": 1.0439,
+      "step": 13808
+    },
+    {
+      "epoch": 0.8389938635397047,
+      "grad_norm": 0.12191746383905411,
+      "learning_rate": 6.303065683453835e-06,
+      "loss": 1.0517,
+      "step": 13809
+    },
+    {
+      "epoch": 0.8390546205723313,
+      "grad_norm": 0.1229800134897232,
+      "learning_rate": 6.298413770263639e-06,
+      "loss": 1.0184,
+      "step": 13810
+    },
+    {
+      "epoch": 0.8391153776049578,
+      "grad_norm": 0.15104541182518005,
+      "learning_rate": 6.293763458963536e-06,
+      "loss": 1.0986,
+      "step": 13811
+    },
+    {
+      "epoch": 0.8391761346375843,
+      "grad_norm": 0.11182752996683121,
+      "learning_rate": 6.28911474972399e-06,
+      "loss": 1.0656,
+      "step": 13812
+    },
+    {
+      "epoch": 0.8392368916702109,
+      "grad_norm": 0.11685499548912048,
+      "learning_rate": 6.284467642715375e-06,
+      "loss": 1.0433,
+      "step": 13813
+    },
+    {
+      "epoch": 0.8392976487028374,
+      "grad_norm": 0.1364884078502655,
+      "learning_rate": 6.279822138108054e-06,
+      "loss": 1.0881,
+      "step": 13814
+    },
+    {
+      "epoch": 0.8393584057354638,
+      "grad_norm": 0.17139606177806854,
+      "learning_rate": 6.27517823607231e-06,
+      "loss": 1.1082,
+      "step": 13815
+    },
+    {
+      "epoch": 0.8394191627680904,
+      "grad_norm": 0.19314952194690704,
+      "learning_rate": 6.270535936778354e-06,
+      "loss": 1.113,
+      "step": 13816
+    },
+    {
+      "epoch": 0.8394799198007169,
+      "grad_norm": 0.2756255269050598,
+      "learning_rate": 6.265895240396358e-06,
+      "loss": 1.1312,
+      "step": 13817
+    },
+    {
+      "epoch": 0.8395406768333434,
+      "grad_norm": 0.1790992170572281,
+      "learning_rate": 6.261256147096423e-06,
+      "loss": 1.0781,
+      "step": 13818
+    },
+    {
+      "epoch": 0.83960143386597,
+      "grad_norm": 0.12227267771959305,
+      "learning_rate": 6.256618657048602e-06,
+      "loss": 1.0565,
+      "step": 13819
+    },
+    {
+      "epoch": 0.8396621908985965,
+      "grad_norm": 0.14711852371692657,
+      "learning_rate": 6.2519827704228765e-06,
+      "loss": 1.0691,
+      "step": 13820
+    },
+    {
+      "epoch": 0.839722947931223,
+      "grad_norm": 1.5554476976394653,
+      "learning_rate": 6.247348487389182e-06,
+      "loss": 1.0783,
+      "step": 13821
+    },
+    {
+      "epoch": 0.8397837049638496,
+      "grad_norm": 0.19400683045387268,
+      "learning_rate": 6.242715808117383e-06,
+      "loss": 1.1859,
+      "step": 13822
+    },
+    {
+      "epoch": 0.8398444619964761,
+      "grad_norm": 0.17753420770168304,
+      "learning_rate": 6.238084732777294e-06,
+      "loss": 1.1051,
+      "step": 13823
+    },
+    {
+      "epoch": 0.8399052190291026,
+      "grad_norm": 0.17847858369350433,
+      "learning_rate": 6.233455261538662e-06,
+      "loss": 1.1659,
+      "step": 13824
+    },
+    {
+      "epoch": 0.8399659760617292,
+      "grad_norm": 0.10870621353387833,
+      "learning_rate": 6.228827394571201e-06,
+      "loss": 1.001,
+      "step": 13825
+    },
+    {
+      "epoch": 0.8400267330943557,
+      "grad_norm": 0.12984980642795563,
+      "learning_rate": 6.224201132044527e-06,
+      "loss": 1.0062,
+      "step": 13826
+    },
+    {
+      "epoch": 0.8400874901269823,
+      "grad_norm": 0.12406597286462784,
+      "learning_rate": 6.219576474128224e-06,
+      "loss": 1.0093,
+      "step": 13827
+    },
+    {
+      "epoch": 0.8401482471596087,
+      "grad_norm": 0.09741330146789551,
+      "learning_rate": 6.214953420991803e-06,
+      "loss": 1.0695,
+      "step": 13828
+    },
+    {
+      "epoch": 0.8402090041922352,
+      "grad_norm": 0.11878557503223419,
+      "learning_rate": 6.210331972804723e-06,
+      "loss": 1.1275,
+      "step": 13829
+    },
+    {
+      "epoch": 0.8402697612248617,
+      "grad_norm": 0.25833576917648315,
+      "learning_rate": 6.205712129736396e-06,
+      "loss": 1.3048,
+      "step": 13830
+    },
+    {
+      "epoch": 0.8403305182574883,
+      "grad_norm": 0.13255338370800018,
+      "learning_rate": 6.201093891956156e-06,
+      "loss": 1.012,
+      "step": 13831
+    },
+    {
+      "epoch": 0.8403912752901148,
+      "grad_norm": 0.13217753171920776,
+      "learning_rate": 6.196477259633299e-06,
+      "loss": 1.0539,
+      "step": 13832
+    },
+    {
+      "epoch": 0.8404520323227413,
+      "grad_norm": 0.7342213988304138,
+      "learning_rate": 6.191862232937018e-06,
+      "loss": 1.1737,
+      "step": 13833
+    },
+    {
+      "epoch": 0.8405127893553679,
+      "grad_norm": 0.39364326000213623,
+      "learning_rate": 6.187248812036489e-06,
+      "loss": 1.1532,
+      "step": 13834
+    },
+    {
+      "epoch": 0.8405735463879944,
+      "grad_norm": 0.09241849929094315,
+      "learning_rate": 6.182636997100833e-06,
+      "loss": 0.9985,
+      "step": 13835
+    },
+    {
+      "epoch": 0.840634303420621,
+      "grad_norm": 0.22844815254211426,
+      "learning_rate": 6.178026788299091e-06,
+      "loss": 1.0934,
+      "step": 13836
+    },
+    {
+      "epoch": 0.8406950604532475,
+      "grad_norm": 0.16626201570034027,
+      "learning_rate": 6.17341818580024e-06,
+      "loss": 1.0266,
+      "step": 13837
+    },
+    {
+      "epoch": 0.840755817485874,
+      "grad_norm": 0.10953666269779205,
+      "learning_rate": 6.168811189773233e-06,
+      "loss": 1.0366,
+      "step": 13838
+    },
+    {
+      "epoch": 0.8408165745185006,
+      "grad_norm": 0.12286390364170074,
+      "learning_rate": 6.164205800386902e-06,
+      "loss": 1.0569,
+      "step": 13839
+    },
+    {
+      "epoch": 0.8408773315511271,
+      "grad_norm": 0.14571483433246613,
+      "learning_rate": 6.15960201781009e-06,
+      "loss": 1.0828,
+      "step": 13840
+    },
+    {
+      "epoch": 0.8409380885837535,
+      "grad_norm": 0.12175819277763367,
+      "learning_rate": 6.154999842211545e-06,
+      "loss": 0.9593,
+      "step": 13841
+    },
+    {
+      "epoch": 0.84099884561638,
+      "grad_norm": 0.20199422538280487,
+      "learning_rate": 6.150399273759949e-06,
+      "loss": 1.068,
+      "step": 13842
+    },
+    {
+      "epoch": 0.8410596026490066,
+      "grad_norm": 0.10887544602155685,
+      "learning_rate": 6.145800312623951e-06,
+      "loss": 1.0704,
+      "step": 13843
+    },
+    {
+      "epoch": 0.8411203596816331,
+      "grad_norm": 0.179683119058609,
+      "learning_rate": 6.141202958972114e-06,
+      "loss": 0.9995,
+      "step": 13844
+    },
+    {
+      "epoch": 0.8411811167142597,
+      "grad_norm": 0.16286370158195496,
+      "learning_rate": 6.1366072129729615e-06,
+      "loss": 0.9997,
+      "step": 13845
+    },
+    {
+      "epoch": 0.8412418737468862,
+      "grad_norm": 0.3153090178966522,
+      "learning_rate": 6.132013074794951e-06,
+      "loss": 1.2551,
+      "step": 13846
+    },
+    {
+      "epoch": 0.8413026307795127,
+      "grad_norm": 0.3104003667831421,
+      "learning_rate": 6.127420544606477e-06,
+      "loss": 1.1072,
+      "step": 13847
+    },
+    {
+      "epoch": 0.8413633878121393,
+      "grad_norm": 0.2739587724208832,
+      "learning_rate": 6.122829622575888e-06,
+      "loss": 1.1687,
+      "step": 13848
+    },
+    {
+      "epoch": 0.8414241448447658,
+      "grad_norm": 0.6147429347038269,
+      "learning_rate": 6.118240308871459e-06,
+      "loss": 0.9993,
+      "step": 13849
+    },
+    {
+      "epoch": 0.8414849018773923,
+      "grad_norm": 0.15113450586795807,
+      "learning_rate": 6.113652603661402e-06,
+      "loss": 1.0322,
+      "step": 13850
+    },
+    {
+      "epoch": 0.8415456589100189,
+      "grad_norm": 0.23021289706230164,
+      "learning_rate": 6.109066507113903e-06,
+      "loss": 1.1207,
+      "step": 13851
+    },
+    {
+      "epoch": 0.8416064159426454,
+      "grad_norm": 0.6288475394248962,
+      "learning_rate": 6.104482019397068e-06,
+      "loss": 1.1348,
+      "step": 13852
+    },
+    {
+      "epoch": 0.8416671729752719,
+      "grad_norm": 0.480734258890152,
+      "learning_rate": 6.099899140678916e-06,
+      "loss": 1.1913,
+      "step": 13853
+    },
+    {
+      "epoch": 0.8417279300078984,
+      "grad_norm": 0.1646154522895813,
+      "learning_rate": 6.095317871127448e-06,
+      "loss": 1.1624,
+      "step": 13854
+    },
+    {
+      "epoch": 0.8417886870405249,
+      "grad_norm": 0.11706171184778214,
+      "learning_rate": 6.090738210910579e-06,
+      "loss": 1.0541,
+      "step": 13855
+    },
+    {
+      "epoch": 0.8418494440731514,
+      "grad_norm": 0.1765625923871994,
+      "learning_rate": 6.086160160196197e-06,
+      "loss": 1.0522,
+      "step": 13856
+    },
+    {
+      "epoch": 0.841910201105778,
+      "grad_norm": 0.19297125935554504,
+      "learning_rate": 6.081583719152106e-06,
+      "loss": 1.162,
+      "step": 13857
+    },
+    {
+      "epoch": 0.8419709581384045,
+      "grad_norm": 0.738193690776825,
+      "learning_rate": 6.07700888794605e-06,
+      "loss": 1.3687,
+      "step": 13858
+    },
+    {
+      "epoch": 0.842031715171031,
+      "grad_norm": 0.16662824153900146,
+      "learning_rate": 6.072435666745735e-06,
+      "loss": 1.1661,
+      "step": 13859
+    },
+    {
+      "epoch": 0.8420924722036576,
+      "grad_norm": 0.12282543629407883,
+      "learning_rate": 6.067864055718764e-06,
+      "loss": 1.0225,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8421532292362841,
+      "grad_norm": 0.11731891334056854,
+      "learning_rate": 6.063294055032731e-06,
+      "loss": 1.0572,
+      "step": 13861
+    },
+    {
+      "epoch": 0.8422139862689106,
+      "grad_norm": 0.15579040348529816,
+      "learning_rate": 6.058725664855153e-06,
+      "loss": 1.1617,
+      "step": 13862
+    },
+    {
+      "epoch": 0.8422747433015372,
+      "grad_norm": 0.11521507054567337,
+      "learning_rate": 6.054158885353478e-06,
+      "loss": 1.037,
+      "step": 13863
+    },
+    {
+      "epoch": 0.8423355003341637,
+      "grad_norm": 0.15756145119667053,
+      "learning_rate": 6.049593716695107e-06,
+      "loss": 1.096,
+      "step": 13864
+    },
+    {
+      "epoch": 0.8423962573667902,
+      "grad_norm": 0.355928510427475,
+      "learning_rate": 6.045030159047366e-06,
+      "loss": 1.019,
+      "step": 13865
+    },
+    {
+      "epoch": 0.8424570143994168,
+      "grad_norm": 0.19757333397865295,
+      "learning_rate": 6.040468212577549e-06,
+      "loss": 1.2141,
+      "step": 13866
+    },
+    {
+      "epoch": 0.8425177714320432,
+      "grad_norm": 0.15494762361049652,
+      "learning_rate": 6.0359078774528625e-06,
+      "loss": 1.0597,
+      "step": 13867
+    },
+    {
+      "epoch": 0.8425785284646697,
+      "grad_norm": 0.11002039909362793,
+      "learning_rate": 6.031349153840471e-06,
+      "loss": 1.0286,
+      "step": 13868
+    },
+    {
+      "epoch": 0.8426392854972963,
+      "grad_norm": 0.15614134073257446,
+      "learning_rate": 6.026792041907475e-06,
+      "loss": 0.9997,
+      "step": 13869
+    },
+    {
+      "epoch": 0.8427000425299228,
+      "grad_norm": 0.1367620974779129,
+      "learning_rate": 6.022236541820919e-06,
+      "loss": 1.0647,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8427607995625493,
+      "grad_norm": 0.1541340947151184,
+      "learning_rate": 6.01768265374777e-06,
+      "loss": 1.0453,
+      "step": 13871
+    },
+    {
+      "epoch": 0.8428215565951759,
+      "grad_norm": 0.10972689092159271,
+      "learning_rate": 6.013130377854986e-06,
+      "loss": 1.0448,
+      "step": 13872
+    },
+    {
+      "epoch": 0.8428823136278024,
+      "grad_norm": 0.15365009009838104,
+      "learning_rate": 6.008579714309398e-06,
+      "loss": 1.1324,
+      "step": 13873
+    },
+    {
+      "epoch": 0.842943070660429,
+      "grad_norm": 0.40709957480430603,
+      "learning_rate": 6.004030663277832e-06,
+      "loss": 1.4185,
+      "step": 13874
+    },
+    {
+      "epoch": 0.8430038276930555,
+      "grad_norm": 0.13938423991203308,
+      "learning_rate": 5.999483224927016e-06,
+      "loss": 1.0445,
+      "step": 13875
+    },
+    {
+      "epoch": 0.843064584725682,
+      "grad_norm": 0.16435745358467102,
+      "learning_rate": 5.994937399423645e-06,
+      "loss": 1.0489,
+      "step": 13876
+    },
+    {
+      "epoch": 0.8431253417583086,
+      "grad_norm": 0.15925735235214233,
+      "learning_rate": 5.990393186934357e-06,
+      "loss": 1.0384,
+      "step": 13877
+    },
+    {
+      "epoch": 0.8431860987909351,
+      "grad_norm": 0.11503589153289795,
+      "learning_rate": 5.9858505876257136e-06,
+      "loss": 1.0924,
+      "step": 13878
+    },
+    {
+      "epoch": 0.8432468558235616,
+      "grad_norm": 0.1875385046005249,
+      "learning_rate": 5.981309601664237e-06,
+      "loss": 1.1136,
+      "step": 13879
+    },
+    {
+      "epoch": 0.843307612856188,
+      "grad_norm": 0.14069220423698425,
+      "learning_rate": 5.97677022921635e-06,
+      "loss": 1.096,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8433683698888146,
+      "grad_norm": 0.10320793092250824,
+      "learning_rate": 5.972232470448452e-06,
+      "loss": 0.972,
+      "step": 13881
+    },
+    {
+      "epoch": 0.8434291269214411,
+      "grad_norm": 0.2010670155286789,
+      "learning_rate": 5.967696325526895e-06,
+      "loss": 0.9984,
+      "step": 13882
+    },
+    {
+      "epoch": 0.8434898839540677,
+      "grad_norm": 0.13203686475753784,
+      "learning_rate": 5.9631617946179385e-06,
+      "loss": 1.0238,
+      "step": 13883
+    },
+    {
+      "epoch": 0.8435506409866942,
+      "grad_norm": 1.4844870567321777,
+      "learning_rate": 5.958628877887801e-06,
+      "loss": 1.0436,
+      "step": 13884
+    },
+    {
+      "epoch": 0.8436113980193207,
+      "grad_norm": 0.16063804924488068,
+      "learning_rate": 5.9540975755026426e-06,
+      "loss": 1.1197,
+      "step": 13885
+    },
+    {
+      "epoch": 0.8436721550519473,
+      "grad_norm": 0.09038233011960983,
+      "learning_rate": 5.949567887628532e-06,
+      "loss": 1.0234,
+      "step": 13886
+    },
+    {
+      "epoch": 0.8437329120845738,
+      "grad_norm": 0.18728040158748627,
+      "learning_rate": 5.945039814431541e-06,
+      "loss": 1.1137,
+      "step": 13887
+    },
+    {
+      "epoch": 0.8437936691172003,
+      "grad_norm": 0.19329407811164856,
+      "learning_rate": 5.940513356077626e-06,
+      "loss": 1.0977,
+      "step": 13888
+    },
+    {
+      "epoch": 0.8438544261498269,
+      "grad_norm": 0.231111079454422,
+      "learning_rate": 5.93598851273271e-06,
+      "loss": 1.0863,
+      "step": 13889
+    },
+    {
+      "epoch": 0.8439151831824534,
+      "grad_norm": 0.25511226058006287,
+      "learning_rate": 5.931465284562648e-06,
+      "loss": 1.1529,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8439759402150799,
+      "grad_norm": 0.12302915006875992,
+      "learning_rate": 5.926943671733254e-06,
+      "loss": 1.0417,
+      "step": 13891
+    },
+    {
+      "epoch": 0.8440366972477065,
+      "grad_norm": 0.139537513256073,
+      "learning_rate": 5.922423674410255e-06,
+      "loss": 1.0403,
+      "step": 13892
+    },
+    {
+      "epoch": 0.844097454280333,
+      "grad_norm": 0.13094432651996613,
+      "learning_rate": 5.917905292759334e-06,
+      "loss": 1.0463,
+      "step": 13893
+    },
+    {
+      "epoch": 0.8441582113129594,
+      "grad_norm": 0.2075701355934143,
+      "learning_rate": 5.913388526946117e-06,
+      "loss": 1.0801,
+      "step": 13894
+    },
+    {
+      "epoch": 0.844218968345586,
+      "grad_norm": 0.18712620437145233,
+      "learning_rate": 5.9088733771361645e-06,
+      "loss": 1.1441,
+      "step": 13895
+    },
+    {
+      "epoch": 0.8442797253782125,
+      "grad_norm": 0.19830457866191864,
+      "learning_rate": 5.904359843494978e-06,
+      "loss": 1.1013,
+      "step": 13896
+    },
+    {
+      "epoch": 0.844340482410839,
+      "grad_norm": 1.6655857563018799,
+      "learning_rate": 5.899847926188001e-06,
+      "loss": 1.0438,
+      "step": 13897
+    },
+    {
+      "epoch": 0.8444012394434656,
+      "grad_norm": 0.11538496613502502,
+      "learning_rate": 5.895337625380632e-06,
+      "loss": 0.9785,
+      "step": 13898
+    },
+    {
+      "epoch": 0.8444619964760921,
+      "grad_norm": 0.1874082088470459,
+      "learning_rate": 5.890828941238191e-06,
+      "loss": 1.0697,
+      "step": 13899
+    },
+    {
+      "epoch": 0.8445227535087186,
+      "grad_norm": 0.18893639743328094,
+      "learning_rate": 5.886321873925937e-06,
+      "loss": 1.0719,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8445835105413452,
+      "grad_norm": 0.19268958270549774,
+      "learning_rate": 5.881816423609082e-06,
+      "loss": 1.1023,
+      "step": 13901
+    },
+    {
+      "epoch": 0.8446442675739717,
+      "grad_norm": 0.14129790663719177,
+      "learning_rate": 5.87731259045276e-06,
+      "loss": 1.0181,
+      "step": 13902
+    },
+    {
+      "epoch": 0.8447050246065982,
+      "grad_norm": 0.11569572240114212,
+      "learning_rate": 5.8728103746220894e-06,
+      "loss": 1.1015,
+      "step": 13903
+    },
+    {
+      "epoch": 0.8447657816392248,
+      "grad_norm": 0.2612714469432831,
+      "learning_rate": 5.868309776282077e-06,
+      "loss": 1.0468,
+      "step": 13904
+    },
+    {
+      "epoch": 0.8448265386718513,
+      "grad_norm": 0.16901415586471558,
+      "learning_rate": 5.863810795597718e-06,
+      "loss": 1.1631,
+      "step": 13905
+    },
+    {
+      "epoch": 0.8448872957044778,
+      "grad_norm": 0.20452101528644562,
+      "learning_rate": 5.859313432733882e-06,
+      "loss": 1.062,
+      "step": 13906
+    },
+    {
+      "epoch": 0.8449480527371043,
+      "grad_norm": 0.20321491360664368,
+      "learning_rate": 5.854817687855451e-06,
+      "loss": 1.059,
+      "step": 13907
+    },
+    {
+      "epoch": 0.8450088097697308,
+      "grad_norm": 0.1561172753572464,
+      "learning_rate": 5.850323561127213e-06,
+      "loss": 1.1638,
+      "step": 13908
+    },
+    {
+      "epoch": 0.8450695668023573,
+      "grad_norm": 0.14651024341583252,
+      "learning_rate": 5.845831052713902e-06,
+      "loss": 1.0839,
+      "step": 13909
+    },
+    {
+      "epoch": 0.8451303238349839,
+      "grad_norm": 0.11238010972738266,
+      "learning_rate": 5.841340162780184e-06,
+      "loss": 1.0364,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8451910808676104,
+      "grad_norm": 0.0998220443725586,
+      "learning_rate": 5.836850891490681e-06,
+      "loss": 0.9538,
+      "step": 13911
+    },
+    {
+      "epoch": 0.845251837900237,
+      "grad_norm": 3.9812381267547607,
+      "learning_rate": 5.832363239009947e-06,
+      "loss": 1.0316,
+      "step": 13912
+    },
+    {
+      "epoch": 0.8453125949328635,
+      "grad_norm": 0.1369580626487732,
+      "learning_rate": 5.827877205502469e-06,
+      "loss": 1.0209,
+      "step": 13913
+    },
+    {
+      "epoch": 0.84537335196549,
+      "grad_norm": 1.0016809701919556,
+      "learning_rate": 5.823392791132698e-06,
+      "loss": 1.1488,
+      "step": 13914
+    },
+    {
+      "epoch": 0.8454341089981166,
+      "grad_norm": 0.12096266448497772,
+      "learning_rate": 5.818909996064998e-06,
+      "loss": 1.0222,
+      "step": 13915
+    },
+    {
+      "epoch": 0.8454948660307431,
+      "grad_norm": 0.24976760149002075,
+      "learning_rate": 5.814428820463691e-06,
+      "loss": 1.1279,
+      "step": 13916
+    },
+    {
+      "epoch": 0.8455556230633696,
+      "grad_norm": 0.7143173217773438,
+      "learning_rate": 5.809949264493031e-06,
+      "loss": 1.0125,
+      "step": 13917
+    },
+    {
+      "epoch": 0.8456163800959962,
+      "grad_norm": 0.4793556332588196,
+      "learning_rate": 5.80547132831723e-06,
+      "loss": 1.0407,
+      "step": 13918
+    },
+    {
+      "epoch": 0.8456771371286227,
+      "grad_norm": 0.16486790776252747,
+      "learning_rate": 5.800995012100435e-06,
+      "loss": 1.1637,
+      "step": 13919
+    },
+    {
+      "epoch": 0.8457378941612491,
+      "grad_norm": 0.19586597383022308,
+      "learning_rate": 5.796520316006693e-06,
+      "loss": 1.1231,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8457986511938757,
+      "grad_norm": 0.21166226267814636,
+      "learning_rate": 5.792047240200049e-06,
+      "loss": 1.0789,
+      "step": 13921
+    },
+    {
+      "epoch": 0.8458594082265022,
+      "grad_norm": 0.11669517308473587,
+      "learning_rate": 5.787575784844446e-06,
+      "loss": 1.0335,
+      "step": 13922
+    },
+    {
+      "epoch": 0.8459201652591287,
+      "grad_norm": 0.20192952454090118,
+      "learning_rate": 5.783105950103812e-06,
+      "loss": 1.0376,
+      "step": 13923
+    },
+    {
+      "epoch": 0.8459809222917553,
+      "grad_norm": 0.10634389519691467,
+      "learning_rate": 5.778637736141973e-06,
+      "loss": 0.9854,
+      "step": 13924
+    },
+    {
+      "epoch": 0.8460416793243818,
+      "grad_norm": 1.1997287273406982,
+      "learning_rate": 5.7741711431227255e-06,
+      "loss": 1.1657,
+      "step": 13925
+    },
+    {
+      "epoch": 0.8461024363570083,
+      "grad_norm": 0.23737677931785583,
+      "learning_rate": 5.769706171209777e-06,
+      "loss": 1.1759,
+      "step": 13926
+    },
+    {
+      "epoch": 0.8461631933896349,
+      "grad_norm": 0.13504615426063538,
+      "learning_rate": 5.7652428205667855e-06,
+      "loss": 1.0699,
+      "step": 13927
+    },
+    {
+      "epoch": 0.8462239504222614,
+      "grad_norm": 0.14231258630752563,
+      "learning_rate": 5.760781091357376e-06,
+      "loss": 1.0612,
+      "step": 13928
+    },
+    {
+      "epoch": 0.8462847074548879,
+      "grad_norm": 0.1106368899345398,
+      "learning_rate": 5.756320983745089e-06,
+      "loss": 1.0621,
+      "step": 13929
+    },
+    {
+      "epoch": 0.8463454644875145,
+      "grad_norm": 0.21288204193115234,
+      "learning_rate": 5.751862497893412e-06,
+      "loss": 1.1791,
+      "step": 13930
+    },
+    {
+      "epoch": 0.846406221520141,
+      "grad_norm": 0.16346928477287292,
+      "learning_rate": 5.747405633965763e-06,
+      "loss": 1.0293,
+      "step": 13931
+    },
+    {
+      "epoch": 0.8464669785527675,
+      "grad_norm": 0.21459725499153137,
+      "learning_rate": 5.742950392125518e-06,
+      "loss": 0.9784,
+      "step": 13932
+    },
+    {
+      "epoch": 0.846527735585394,
+      "grad_norm": 0.30261167883872986,
+      "learning_rate": 5.738496772535984e-06,
+      "loss": 1.2409,
+      "step": 13933
+    },
+    {
+      "epoch": 0.8465884926180205,
+      "grad_norm": 0.14750361442565918,
+      "learning_rate": 5.734044775360397e-06,
+      "loss": 1.083,
+      "step": 13934
+    },
+    {
+      "epoch": 0.846649249650647,
+      "grad_norm": 0.15585088729858398,
+      "learning_rate": 5.729594400761961e-06,
+      "loss": 1.0613,
+      "step": 13935
+    },
+    {
+      "epoch": 0.8467100066832736,
+      "grad_norm": 0.12576434016227722,
+      "learning_rate": 5.725145648903801e-06,
+      "loss": 1.0309,
+      "step": 13936
+    },
+    {
+      "epoch": 0.8467707637159001,
+      "grad_norm": 0.13810418546199799,
+      "learning_rate": 5.720698519948986e-06,
+      "loss": 1.0646,
+      "step": 13937
+    },
+    {
+      "epoch": 0.8468315207485266,
+      "grad_norm": 0.21547050774097443,
+      "learning_rate": 5.716253014060513e-06,
+      "loss": 1.132,
+      "step": 13938
+    },
+    {
+      "epoch": 0.8468922777811532,
+      "grad_norm": 0.16368132829666138,
+      "learning_rate": 5.7118091314013635e-06,
+      "loss": 1.0583,
+      "step": 13939
+    },
+    {
+      "epoch": 0.8469530348137797,
+      "grad_norm": 0.15535663068294525,
+      "learning_rate": 5.707366872134401e-06,
+      "loss": 1.0338,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8470137918464062,
+      "grad_norm": 0.13681825995445251,
+      "learning_rate": 5.702926236422474e-06,
+      "loss": 0.9994,
+      "step": 13941
+    },
+    {
+      "epoch": 0.8470745488790328,
+      "grad_norm": 0.14484292268753052,
+      "learning_rate": 5.6984872244283395e-06,
+      "loss": 1.1008,
+      "step": 13942
+    },
+    {
+      "epoch": 0.8471353059116593,
+      "grad_norm": 0.2158062756061554,
+      "learning_rate": 5.694049836314713e-06,
+      "loss": 1.1981,
+      "step": 13943
+    },
+    {
+      "epoch": 0.8471960629442858,
+      "grad_norm": 0.1107899472117424,
+      "learning_rate": 5.689614072244265e-06,
+      "loss": 1.0519,
+      "step": 13944
+    },
+    {
+      "epoch": 0.8472568199769124,
+      "grad_norm": 0.12145136296749115,
+      "learning_rate": 5.685179932379581e-06,
+      "loss": 1.0633,
+      "step": 13945
+    },
+    {
+      "epoch": 0.8473175770095388,
+      "grad_norm": 0.11620216071605682,
+      "learning_rate": 5.680747416883198e-06,
+      "loss": 1.0644,
+      "step": 13946
+    },
+    {
+      "epoch": 0.8473783340421653,
+      "grad_norm": 0.1601884812116623,
+      "learning_rate": 5.676316525917575e-06,
+      "loss": 1.132,
+      "step": 13947
+    },
+    {
+      "epoch": 0.8474390910747919,
+      "grad_norm": 0.10534972697496414,
+      "learning_rate": 5.671887259645131e-06,
+      "loss": 0.9992,
+      "step": 13948
+    },
+    {
+      "epoch": 0.8474998481074184,
+      "grad_norm": 0.17490743100643158,
+      "learning_rate": 5.667459618228232e-06,
+      "loss": 1.2029,
+      "step": 13949
+    },
+    {
+      "epoch": 0.8475606051400449,
+      "grad_norm": 0.16057506203651428,
+      "learning_rate": 5.663033601829177e-06,
+      "loss": 1.0507,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8476213621726715,
+      "grad_norm": 0.3475726246833801,
+      "learning_rate": 5.658609210610189e-06,
+      "loss": 1.1973,
+      "step": 13951
+    },
+    {
+      "epoch": 0.847682119205298,
+      "grad_norm": 0.22239616513252258,
+      "learning_rate": 5.6541864447334665e-06,
+      "loss": 1.1561,
+      "step": 13952
+    },
+    {
+      "epoch": 0.8477428762379245,
+      "grad_norm": 0.10262919962406158,
+      "learning_rate": 5.649765304361088e-06,
+      "loss": 1.0142,
+      "step": 13953
+    },
+    {
+      "epoch": 0.8478036332705511,
+      "grad_norm": 0.19047772884368896,
+      "learning_rate": 5.645345789655149e-06,
+      "loss": 1.0935,
+      "step": 13954
+    },
+    {
+      "epoch": 0.8478643903031776,
+      "grad_norm": 0.37129777669906616,
+      "learning_rate": 5.640927900777626e-06,
+      "loss": 1.1662,
+      "step": 13955
+    },
+    {
+      "epoch": 0.8479251473358042,
+      "grad_norm": 0.12124883383512497,
+      "learning_rate": 5.636511637890474e-06,
+      "loss": 1.0394,
+      "step": 13956
+    },
+    {
+      "epoch": 0.8479859043684307,
+      "grad_norm": 0.15603722631931305,
+      "learning_rate": 5.632097001155556e-06,
+      "loss": 1.1198,
+      "step": 13957
+    },
+    {
+      "epoch": 0.8480466614010572,
+      "grad_norm": 0.1487198919057846,
+      "learning_rate": 5.6276839907347054e-06,
+      "loss": 1.1079,
+      "step": 13958
+    },
+    {
+      "epoch": 0.8481074184336836,
+      "grad_norm": 0.1818862110376358,
+      "learning_rate": 5.62327260678967e-06,
+      "loss": 1.1105,
+      "step": 13959
+    },
+    {
+      "epoch": 0.8481681754663102,
+      "grad_norm": 0.13143955171108246,
+      "learning_rate": 5.618862849482159e-06,
+      "loss": 1.0979,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8482289324989367,
+      "grad_norm": 0.11102840304374695,
+      "learning_rate": 5.6144547189738055e-06,
+      "loss": 1.0216,
+      "step": 13961
+    },
+    {
+      "epoch": 0.8482896895315633,
+      "grad_norm": 0.7428750991821289,
+      "learning_rate": 5.6100482154261945e-06,
+      "loss": 1.0789,
+      "step": 13962
+    },
+    {
+      "epoch": 0.8483504465641898,
+      "grad_norm": 0.185042142868042,
+      "learning_rate": 5.605643339000849e-06,
+      "loss": 1.1868,
+      "step": 13963
+    },
+    {
+      "epoch": 0.8484112035968163,
+      "grad_norm": 0.1131417453289032,
+      "learning_rate": 5.6012400898592166e-06,
+      "loss": 1.0566,
+      "step": 13964
+    },
+    {
+      "epoch": 0.8484719606294429,
+      "grad_norm": 0.12369358539581299,
+      "learning_rate": 5.596838468162718e-06,
+      "loss": 1.0471,
+      "step": 13965
+    },
+    {
+      "epoch": 0.8485327176620694,
+      "grad_norm": 0.15037371218204498,
+      "learning_rate": 5.592438474072703e-06,
+      "loss": 1.1091,
+      "step": 13966
+    },
+    {
+      "epoch": 0.8485934746946959,
+      "grad_norm": 0.11376750469207764,
+      "learning_rate": 5.588040107750425e-06,
+      "loss": 1.0585,
+      "step": 13967
+    },
+    {
+      "epoch": 0.8486542317273225,
+      "grad_norm": 0.13548539578914642,
+      "learning_rate": 5.583643369357122e-06,
+      "loss": 1.0781,
+      "step": 13968
+    },
+    {
+      "epoch": 0.848714988759949,
+      "grad_norm": 0.13752390444278717,
+      "learning_rate": 5.579248259053949e-06,
+      "loss": 1.0263,
+      "step": 13969
+    },
+    {
+      "epoch": 0.8487757457925755,
+      "grad_norm": 0.1105789989233017,
+      "learning_rate": 5.574854777002026e-06,
+      "loss": 1.0754,
+      "step": 13970
+    },
+    {
+      "epoch": 0.8488365028252021,
+      "grad_norm": 0.1519423872232437,
+      "learning_rate": 5.570462923362391e-06,
+      "loss": 1.1217,
+      "step": 13971
+    },
+    {
+      "epoch": 0.8488972598578285,
+      "grad_norm": 2.2291924953460693,
+      "learning_rate": 5.566072698296032e-06,
+      "loss": 1.1915,
+      "step": 13972
+    },
+    {
+      "epoch": 0.848958016890455,
+      "grad_norm": 0.09913677722215652,
+      "learning_rate": 5.56168410196386e-06,
+      "loss": 1.0571,
+      "step": 13973
+    },
+    {
+      "epoch": 0.8490187739230816,
+      "grad_norm": 0.10555507242679596,
+      "learning_rate": 5.557297134526734e-06,
+      "loss": 1.0546,
+      "step": 13974
+    },
+    {
+      "epoch": 0.8490795309557081,
+      "grad_norm": 0.20103463530540466,
+      "learning_rate": 5.552911796145488e-06,
+      "loss": 1.143,
+      "step": 13975
+    },
+    {
+      "epoch": 0.8491402879883346,
+      "grad_norm": 0.09908825159072876,
+      "learning_rate": 5.548528086980848e-06,
+      "loss": 1.0404,
+      "step": 13976
+    },
+    {
+      "epoch": 0.8492010450209612,
+      "grad_norm": 0.4844468832015991,
+      "learning_rate": 5.544146007193501e-06,
+      "loss": 1.2248,
+      "step": 13977
+    },
+    {
+      "epoch": 0.8492618020535877,
+      "grad_norm": 0.13323774933815002,
+      "learning_rate": 5.5397655569440765e-06,
+      "loss": 1.0883,
+      "step": 13978
+    },
+    {
+      "epoch": 0.8493225590862142,
+      "grad_norm": 0.1800452023744583,
+      "learning_rate": 5.535386736393139e-06,
+      "loss": 1.1608,
+      "step": 13979
+    },
+    {
+      "epoch": 0.8493833161188408,
+      "grad_norm": 0.18974387645721436,
+      "learning_rate": 5.531009545701199e-06,
+      "loss": 1.0439,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8494440731514673,
+      "grad_norm": 0.14465534687042236,
+      "learning_rate": 5.5266339850286955e-06,
+      "loss": 1.1212,
+      "step": 13981
+    },
+    {
+      "epoch": 0.8495048301840938,
+      "grad_norm": 0.19995470345020294,
+      "learning_rate": 5.522260054536022e-06,
+      "loss": 1.1382,
+      "step": 13982
+    },
+    {
+      "epoch": 0.8495655872167204,
+      "grad_norm": 0.11444170773029327,
+      "learning_rate": 5.517887754383505e-06,
+      "loss": 1.0484,
+      "step": 13983
+    },
+    {
+      "epoch": 0.8496263442493469,
+      "grad_norm": 0.20048007369041443,
+      "learning_rate": 5.5135170847314065e-06,
+      "loss": 1.0072,
+      "step": 13984
+    },
+    {
+      "epoch": 0.8496871012819733,
+      "grad_norm": 0.12816888093948364,
+      "learning_rate": 5.509148045739932e-06,
+      "loss": 1.0688,
+      "step": 13985
+    },
+    {
+      "epoch": 0.8497478583145999,
+      "grad_norm": 0.16064612567424774,
+      "learning_rate": 5.504780637569251e-06,
+      "loss": 1.067,
+      "step": 13986
+    },
+    {
+      "epoch": 0.8498086153472264,
+      "grad_norm": 0.15055491030216217,
+      "learning_rate": 5.500414860379433e-06,
+      "loss": 1.0457,
+      "step": 13987
+    },
+    {
+      "epoch": 0.8498693723798529,
+      "grad_norm": 0.17457495629787445,
+      "learning_rate": 5.496050714330508e-06,
+      "loss": 1.086,
+      "step": 13988
+    },
+    {
+      "epoch": 0.8499301294124795,
+      "grad_norm": 0.11255787312984467,
+      "learning_rate": 5.491688199582446e-06,
+      "loss": 1.0399,
+      "step": 13989
+    },
+    {
+      "epoch": 0.849990886445106,
+      "grad_norm": 0.2116011530160904,
+      "learning_rate": 5.487327316295143e-06,
+      "loss": 1.1662,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8500516434777325,
+      "grad_norm": 0.15451011061668396,
+      "learning_rate": 5.482968064628474e-06,
+      "loss": 1.0324,
+      "step": 13991
+    },
+    {
+      "epoch": 0.8501124005103591,
+      "grad_norm": 0.13074050843715668,
+      "learning_rate": 5.478610444742227e-06,
+      "loss": 0.9919,
+      "step": 13992
+    },
+    {
+      "epoch": 0.8501731575429856,
+      "grad_norm": 0.10223536938428879,
+      "learning_rate": 5.474254456796107e-06,
+      "loss": 1.0534,
+      "step": 13993
+    },
+    {
+      "epoch": 0.8502339145756121,
+      "grad_norm": 0.09384096413850784,
+      "learning_rate": 5.469900100949798e-06,
+      "loss": 1.0026,
+      "step": 13994
+    },
+    {
+      "epoch": 0.8502946716082387,
+      "grad_norm": 0.14617595076560974,
+      "learning_rate": 5.465547377362895e-06,
+      "loss": 1.3121,
+      "step": 13995
+    },
+    {
+      "epoch": 0.8503554286408652,
+      "grad_norm": 0.12430252134799957,
+      "learning_rate": 5.461196286194975e-06,
+      "loss": 1.0583,
+      "step": 13996
+    },
+    {
+      "epoch": 0.8504161856734918,
+      "grad_norm": 0.18041782081127167,
+      "learning_rate": 5.4568468276055165e-06,
+      "loss": 1.1298,
+      "step": 13997
+    },
+    {
+      "epoch": 0.8504769427061183,
+      "grad_norm": 0.13031385838985443,
+      "learning_rate": 5.4524990017539415e-06,
+      "loss": 1.075,
+      "step": 13998
+    },
+    {
+      "epoch": 0.8505376997387447,
+      "grad_norm": 0.20246997475624084,
+      "learning_rate": 5.4481528087996284e-06,
+      "loss": 1.1979,
+      "step": 13999
+    },
+    {
+      "epoch": 0.8505984567713712,
+      "grad_norm": 0.2741469740867615,
+      "learning_rate": 5.443808248901888e-06,
+      "loss": 1.0297,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8506592138039978,
+      "grad_norm": 0.12367981672286987,
+      "learning_rate": 5.439465322219966e-06,
+      "loss": 1.033,
+      "step": 14001
+    },
+    {
+      "epoch": 0.8507199708366243,
+      "grad_norm": 0.15394064784049988,
+      "learning_rate": 5.435124028913058e-06,
+      "loss": 1.0034,
+      "step": 14002
+    },
+    {
+      "epoch": 0.8507807278692509,
+      "grad_norm": 0.20726700127124786,
+      "learning_rate": 5.430784369140296e-06,
+      "loss": 1.199,
+      "step": 14003
+    },
+    {
+      "epoch": 0.8508414849018774,
+      "grad_norm": 0.13614840805530548,
+      "learning_rate": 5.426446343060743e-06,
+      "loss": 1.0427,
+      "step": 14004
+    },
+    {
+      "epoch": 0.8509022419345039,
+      "grad_norm": 0.10408087074756622,
+      "learning_rate": 5.422109950833415e-06,
+      "loss": 1.0179,
+      "step": 14005
+    },
+    {
+      "epoch": 0.8509629989671305,
+      "grad_norm": 0.0921889990568161,
+      "learning_rate": 5.417775192617269e-06,
+      "loss": 0.9978,
+      "step": 14006
+    },
+    {
+      "epoch": 0.851023755999757,
+      "grad_norm": 0.14945179224014282,
+      "learning_rate": 5.413442068571184e-06,
+      "loss": 1.0741,
+      "step": 14007
+    },
+    {
+      "epoch": 0.8510845130323835,
+      "grad_norm": 0.16615328192710876,
+      "learning_rate": 5.4091105788540045e-06,
+      "loss": 1.088,
+      "step": 14008
+    },
+    {
+      "epoch": 0.8511452700650101,
+      "grad_norm": 0.29710111021995544,
+      "learning_rate": 5.404780723624492e-06,
+      "loss": 1.0532,
+      "step": 14009
+    },
+    {
+      "epoch": 0.8512060270976366,
+      "grad_norm": 0.13664542138576508,
+      "learning_rate": 5.400452503041353e-06,
+      "loss": 1.0783,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8512667841302631,
+      "grad_norm": 0.12013693898916245,
+      "learning_rate": 5.396125917263256e-06,
+      "loss": 1.0257,
+      "step": 14011
+    },
+    {
+      "epoch": 0.8513275411628896,
+      "grad_norm": 0.1466490477323532,
+      "learning_rate": 5.39180096644879e-06,
+      "loss": 1.0319,
+      "step": 14012
+    },
+    {
+      "epoch": 0.8513882981955161,
+      "grad_norm": 0.1422511637210846,
+      "learning_rate": 5.387477650756489e-06,
+      "loss": 1.0595,
+      "step": 14013
+    },
+    {
+      "epoch": 0.8514490552281426,
+      "grad_norm": 0.10382051020860672,
+      "learning_rate": 5.383155970344806e-06,
+      "loss": 1.0395,
+      "step": 14014
+    },
+    {
+      "epoch": 0.8515098122607692,
+      "grad_norm": 0.28484755754470825,
+      "learning_rate": 5.378835925372161e-06,
+      "loss": 1.0452,
+      "step": 14015
+    },
+    {
+      "epoch": 0.8515705692933957,
+      "grad_norm": 0.12596751749515533,
+      "learning_rate": 5.374517515996919e-06,
+      "loss": 1.0188,
+      "step": 14016
+    },
+    {
+      "epoch": 0.8516313263260222,
+      "grad_norm": 0.19154946506023407,
+      "learning_rate": 5.370200742377357e-06,
+      "loss": 1.0316,
+      "step": 14017
+    },
+    {
+      "epoch": 0.8516920833586488,
+      "grad_norm": 0.3829948902130127,
+      "learning_rate": 5.365885604671722e-06,
+      "loss": 1.0672,
+      "step": 14018
+    },
+    {
+      "epoch": 0.8517528403912753,
+      "grad_norm": 0.15549182891845703,
+      "learning_rate": 5.3615721030381815e-06,
+      "loss": 1.0702,
+      "step": 14019
+    },
+    {
+      "epoch": 0.8518135974239018,
+      "grad_norm": 0.15565907955169678,
+      "learning_rate": 5.357260237634826e-06,
+      "loss": 1.0909,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8518743544565284,
+      "grad_norm": 0.4052349627017975,
+      "learning_rate": 5.3529500086197394e-06,
+      "loss": 1.0723,
+      "step": 14021
+    },
+    {
+      "epoch": 0.8519351114891549,
+      "grad_norm": 0.30731961131095886,
+      "learning_rate": 5.348641416150896e-06,
+      "loss": 1.0973,
+      "step": 14022
+    },
+    {
+      "epoch": 0.8519958685217814,
+      "grad_norm": 0.1267758160829544,
+      "learning_rate": 5.34433446038623e-06,
+      "loss": 1.0609,
+      "step": 14023
+    },
+    {
+      "epoch": 0.852056625554408,
+      "grad_norm": 0.16898858547210693,
+      "learning_rate": 5.340029141483621e-06,
+      "loss": 1.0255,
+      "step": 14024
+    },
+    {
+      "epoch": 0.8521173825870344,
+      "grad_norm": 0.3386203944683075,
+      "learning_rate": 5.335725459600876e-06,
+      "loss": 1.1653,
+      "step": 14025
+    },
+    {
+      "epoch": 0.8521781396196609,
+      "grad_norm": 0.24327914416790009,
+      "learning_rate": 5.331423414895747e-06,
+      "loss": 1.1276,
+      "step": 14026
+    },
+    {
+      "epoch": 0.8522388966522875,
+      "grad_norm": 0.4767051041126251,
+      "learning_rate": 5.327123007525925e-06,
+      "loss": 1.1859,
+      "step": 14027
+    },
+    {
+      "epoch": 0.852299653684914,
+      "grad_norm": 0.14266517758369446,
+      "learning_rate": 5.322824237649043e-06,
+      "loss": 1.0888,
+      "step": 14028
+    },
+    {
+      "epoch": 0.8523604107175405,
+      "grad_norm": 0.1028783991932869,
+      "learning_rate": 5.3185271054226825e-06,
+      "loss": 1.034,
+      "step": 14029
+    },
+    {
+      "epoch": 0.8524211677501671,
+      "grad_norm": 0.19003506004810333,
+      "learning_rate": 5.314231611004339e-06,
+      "loss": 1.0175,
+      "step": 14030
+    },
+    {
+      "epoch": 0.8524819247827936,
+      "grad_norm": 0.16263505816459656,
+      "learning_rate": 5.309937754551469e-06,
+      "loss": 1.0863,
+      "step": 14031
+    },
+    {
+      "epoch": 0.8525426818154201,
+      "grad_norm": 0.20552152395248413,
+      "learning_rate": 5.305645536221476e-06,
+      "loss": 1.1712,
+      "step": 14032
+    },
+    {
+      "epoch": 0.8526034388480467,
+      "grad_norm": 0.19740070402622223,
+      "learning_rate": 5.301354956171695e-06,
+      "loss": 1.1318,
+      "step": 14033
+    },
+    {
+      "epoch": 0.8526641958806732,
+      "grad_norm": 0.1874537169933319,
+      "learning_rate": 5.297066014559382e-06,
+      "loss": 1.0039,
+      "step": 14034
+    },
+    {
+      "epoch": 0.8527249529132997,
+      "grad_norm": 0.24981433153152466,
+      "learning_rate": 5.292778711541751e-06,
+      "loss": 1.0573,
+      "step": 14035
+    },
+    {
+      "epoch": 0.8527857099459263,
+      "grad_norm": 0.10621805489063263,
+      "learning_rate": 5.28849304727595e-06,
+      "loss": 1.0049,
+      "step": 14036
+    },
+    {
+      "epoch": 0.8528464669785528,
+      "grad_norm": 0.1408226042985916,
+      "learning_rate": 5.284209021919085e-06,
+      "loss": 1.0894,
+      "step": 14037
+    },
+    {
+      "epoch": 0.8529072240111792,
+      "grad_norm": 0.1325346976518631,
+      "learning_rate": 5.279926635628185e-06,
+      "loss": 1.0021,
+      "step": 14038
+    },
+    {
+      "epoch": 0.8529679810438058,
+      "grad_norm": 0.13675685226917267,
+      "learning_rate": 5.275645888560232e-06,
+      "loss": 1.1116,
+      "step": 14039
+    },
+    {
+      "epoch": 0.8530287380764323,
+      "grad_norm": 0.1163097470998764,
+      "learning_rate": 5.271366780872105e-06,
+      "loss": 1.0652,
+      "step": 14040
+    },
+    {
+      "epoch": 0.8530894951090588,
+      "grad_norm": 0.1079884022474289,
+      "learning_rate": 5.267089312720674e-06,
+      "loss": 1.0231,
+      "step": 14041
+    },
+    {
+      "epoch": 0.8531502521416854,
+      "grad_norm": 0.2544534206390381,
+      "learning_rate": 5.262813484262735e-06,
+      "loss": 1.1633,
+      "step": 14042
+    },
+    {
+      "epoch": 0.8532110091743119,
+      "grad_norm": 2.2221500873565674,
+      "learning_rate": 5.258539295655019e-06,
+      "loss": 1.0751,
+      "step": 14043
+    },
+    {
+      "epoch": 0.8532717662069385,
+      "grad_norm": 0.15971939265727997,
+      "learning_rate": 5.254266747054188e-06,
+      "loss": 1.0725,
+      "step": 14044
+    },
+    {
+      "epoch": 0.853332523239565,
+      "grad_norm": 0.14076870679855347,
+      "learning_rate": 5.249995838616861e-06,
+      "loss": 1.0241,
+      "step": 14045
+    },
+    {
+      "epoch": 0.8533932802721915,
+      "grad_norm": 0.22558552026748657,
+      "learning_rate": 5.245726570499587e-06,
+      "loss": 1.0576,
+      "step": 14046
+    },
+    {
+      "epoch": 0.8534540373048181,
+      "grad_norm": 0.10706805437803268,
+      "learning_rate": 5.241458942858857e-06,
+      "loss": 1.0184,
+      "step": 14047
+    },
+    {
+      "epoch": 0.8535147943374446,
+      "grad_norm": 0.1199878379702568,
+      "learning_rate": 5.2371929558511e-06,
+      "loss": 1.028,
+      "step": 14048
+    },
+    {
+      "epoch": 0.8535755513700711,
+      "grad_norm": 0.32045525312423706,
+      "learning_rate": 5.232928609632692e-06,
+      "loss": 1.1553,
+      "step": 14049
+    },
+    {
+      "epoch": 0.8536363084026977,
+      "grad_norm": 0.15111345052719116,
+      "learning_rate": 5.228665904359936e-06,
+      "loss": 1.1162,
+      "step": 14050
+    },
+    {
+      "epoch": 0.8536970654353241,
+      "grad_norm": 0.13601776957511902,
+      "learning_rate": 5.224404840189084e-06,
+      "loss": 1.0531,
+      "step": 14051
+    },
+    {
+      "epoch": 0.8537578224679506,
+      "grad_norm": 0.1124998927116394,
+      "learning_rate": 5.2201454172763175e-06,
+      "loss": 1.0374,
+      "step": 14052
+    },
+    {
+      "epoch": 0.8538185795005772,
+      "grad_norm": 0.14750927686691284,
+      "learning_rate": 5.2158876357777995e-06,
+      "loss": 1.0862,
+      "step": 14053
+    },
+    {
+      "epoch": 0.8538793365332037,
+      "grad_norm": 1.0044686794281006,
+      "learning_rate": 5.211631495849562e-06,
+      "loss": 1.112,
+      "step": 14054
+    },
+    {
+      "epoch": 0.8539400935658302,
+      "grad_norm": 0.1462719887495041,
+      "learning_rate": 5.2073769976476335e-06,
+      "loss": 1.0509,
+      "step": 14055
+    },
+    {
+      "epoch": 0.8540008505984568,
+      "grad_norm": 0.39019280672073364,
+      "learning_rate": 5.203124141327959e-06,
+      "loss": 1.1632,
+      "step": 14056
+    },
+    {
+      "epoch": 0.8540616076310833,
+      "grad_norm": 0.14390425384044647,
+      "learning_rate": 5.198872927046417e-06,
+      "loss": 1.0763,
+      "step": 14057
+    },
+    {
+      "epoch": 0.8541223646637098,
+      "grad_norm": 0.1774798184633255,
+      "learning_rate": 5.194623354958855e-06,
+      "loss": 1.0547,
+      "step": 14058
+    },
+    {
+      "epoch": 0.8541831216963364,
+      "grad_norm": 0.2587502896785736,
+      "learning_rate": 5.19037542522105e-06,
+      "loss": 1.1674,
+      "step": 14059
+    },
+    {
+      "epoch": 0.8542438787289629,
+      "grad_norm": 0.14740493893623352,
+      "learning_rate": 5.186129137988683e-06,
+      "loss": 1.0952,
+      "step": 14060
+    },
+    {
+      "epoch": 0.8543046357615894,
+      "grad_norm": 0.1673775166273117,
+      "learning_rate": 5.181884493417416e-06,
+      "loss": 1.0447,
+      "step": 14061
+    },
+    {
+      "epoch": 0.854365392794216,
+      "grad_norm": 0.4900735020637512,
+      "learning_rate": 5.177641491662821e-06,
+      "loss": 1.2021,
+      "step": 14062
+    },
+    {
+      "epoch": 0.8544261498268425,
+      "grad_norm": 0.15155020356178284,
+      "learning_rate": 5.173400132880457e-06,
+      "loss": 1.0754,
+      "step": 14063
+    },
+    {
+      "epoch": 0.8544869068594689,
+      "grad_norm": 0.10192351043224335,
+      "learning_rate": 5.169160417225771e-06,
+      "loss": 1.0477,
+      "step": 14064
+    },
+    {
+      "epoch": 0.8545476638920955,
+      "grad_norm": 0.1388743370771408,
+      "learning_rate": 5.1649223448541786e-06,
+      "loss": 1.053,
+      "step": 14065
+    },
+    {
+      "epoch": 0.854608420924722,
+      "grad_norm": 0.26271912455558777,
+      "learning_rate": 5.160685915921032e-06,
+      "loss": 1.2382,
+      "step": 14066
+    },
+    {
+      "epoch": 0.8546691779573485,
+      "grad_norm": 0.261271595954895,
+      "learning_rate": 5.156451130581597e-06,
+      "loss": 1.069,
+      "step": 14067
+    },
+    {
+      "epoch": 0.8547299349899751,
+      "grad_norm": 0.1277359127998352,
+      "learning_rate": 5.152217988991115e-06,
+      "loss": 1.0757,
+      "step": 14068
+    },
+    {
+      "epoch": 0.8547906920226016,
+      "grad_norm": 0.2207491248846054,
+      "learning_rate": 5.147986491304757e-06,
+      "loss": 1.0602,
+      "step": 14069
+    },
+    {
+      "epoch": 0.8548514490552281,
+      "grad_norm": 0.10567124933004379,
+      "learning_rate": 5.143756637677621e-06,
+      "loss": 1.072,
+      "step": 14070
+    },
+    {
+      "epoch": 0.8549122060878547,
+      "grad_norm": 6.4839253425598145,
+      "learning_rate": 5.139528428264756e-06,
+      "loss": 1.0505,
+      "step": 14071
+    },
+    {
+      "epoch": 0.8549729631204812,
+      "grad_norm": 0.1824016273021698,
+      "learning_rate": 5.135301863221154e-06,
+      "loss": 1.0083,
+      "step": 14072
+    },
+    {
+      "epoch": 0.8550337201531077,
+      "grad_norm": 0.12151354551315308,
+      "learning_rate": 5.131076942701729e-06,
+      "loss": 1.0784,
+      "step": 14073
+    },
+    {
+      "epoch": 0.8550944771857343,
+      "grad_norm": 0.1690412163734436,
+      "learning_rate": 5.126853666861347e-06,
+      "loss": 1.0106,
+      "step": 14074
+    },
+    {
+      "epoch": 0.8551552342183608,
+      "grad_norm": 0.2130904495716095,
+      "learning_rate": 5.122632035854824e-06,
+      "loss": 1.0766,
+      "step": 14075
+    },
+    {
+      "epoch": 0.8552159912509874,
+      "grad_norm": 0.12926946580410004,
+      "learning_rate": 5.118412049836896e-06,
+      "loss": 1.0916,
+      "step": 14076
+    },
+    {
+      "epoch": 0.8552767482836138,
+      "grad_norm": 0.1278305947780609,
+      "learning_rate": 5.11419370896225e-06,
+      "loss": 1.0319,
+      "step": 14077
+    },
+    {
+      "epoch": 0.8553375053162403,
+      "grad_norm": 0.48680371046066284,
+      "learning_rate": 5.109977013385503e-06,
+      "loss": 1.115,
+      "step": 14078
+    },
+    {
+      "epoch": 0.8553982623488668,
+      "grad_norm": 0.13173919916152954,
+      "learning_rate": 5.105761963261236e-06,
+      "loss": 1.0611,
+      "step": 14079
+    },
+    {
+      "epoch": 0.8554590193814934,
+      "grad_norm": 0.11212275177240372,
+      "learning_rate": 5.101548558743946e-06,
+      "loss": 1.0894,
+      "step": 14080
+    },
+    {
+      "epoch": 0.8555197764141199,
+      "grad_norm": 9.608139038085938,
+      "learning_rate": 5.097336799988067e-06,
+      "loss": 1.0463,
+      "step": 14081
+    },
+    {
+      "epoch": 0.8555805334467464,
+      "grad_norm": 0.11307689547538757,
+      "learning_rate": 5.093126687147986e-06,
+      "loss": 1.0792,
+      "step": 14082
+    },
+    {
+      "epoch": 0.855641290479373,
+      "grad_norm": 0.13400432467460632,
+      "learning_rate": 5.088918220378014e-06,
+      "loss": 1.0328,
+      "step": 14083
+    },
+    {
+      "epoch": 0.8557020475119995,
+      "grad_norm": 0.11246509850025177,
+      "learning_rate": 5.084711399832437e-06,
+      "loss": 0.9947,
+      "step": 14084
+    },
+    {
+      "epoch": 0.855762804544626,
+      "grad_norm": 0.18977652490139008,
+      "learning_rate": 5.080506225665449e-06,
+      "loss": 1.0919,
+      "step": 14085
+    },
+    {
+      "epoch": 0.8558235615772526,
+      "grad_norm": 0.1288769692182541,
+      "learning_rate": 5.076302698031193e-06,
+      "loss": 1.0324,
+      "step": 14086
+    },
+    {
+      "epoch": 0.8558843186098791,
+      "grad_norm": 0.15988439321517944,
+      "learning_rate": 5.072100817083741e-06,
+      "loss": 1.1172,
+      "step": 14087
+    },
+    {
+      "epoch": 0.8559450756425057,
+      "grad_norm": 0.2972189784049988,
+      "learning_rate": 5.067900582977103e-06,
+      "loss": 1.0873,
+      "step": 14088
+    },
+    {
+      "epoch": 0.8560058326751322,
+      "grad_norm": 0.17799590528011322,
+      "learning_rate": 5.063701995865266e-06,
+      "loss": 1.0474,
+      "step": 14089
+    },
+    {
+      "epoch": 0.8560665897077586,
+      "grad_norm": 0.12099135667085648,
+      "learning_rate": 5.0595050559021196e-06,
+      "loss": 1.0697,
+      "step": 14090
+    },
+    {
+      "epoch": 0.8561273467403852,
+      "grad_norm": 0.10432138293981552,
+      "learning_rate": 5.055309763241501e-06,
+      "loss": 0.9967,
+      "step": 14091
+    },
+    {
+      "epoch": 0.8561881037730117,
+      "grad_norm": 0.10693902522325516,
+      "learning_rate": 5.0511161180371916e-06,
+      "loss": 1.0283,
+      "step": 14092
+    },
+    {
+      "epoch": 0.8562488608056382,
+      "grad_norm": 0.1719645857810974,
+      "learning_rate": 5.046924120442903e-06,
+      "loss": 1.1292,
+      "step": 14093
+    },
+    {
+      "epoch": 0.8563096178382648,
+      "grad_norm": 0.2054222822189331,
+      "learning_rate": 5.042733770612307e-06,
+      "loss": 1.1198,
+      "step": 14094
+    },
+    {
+      "epoch": 0.8563703748708913,
+      "grad_norm": 0.1370614618062973,
+      "learning_rate": 5.038545068698991e-06,
+      "loss": 1.036,
+      "step": 14095
+    },
+    {
+      "epoch": 0.8564311319035178,
+      "grad_norm": 0.16334545612335205,
+      "learning_rate": 5.034358014856499e-06,
+      "loss": 1.1287,
+      "step": 14096
+    },
+    {
+      "epoch": 0.8564918889361444,
+      "grad_norm": 0.33314672112464905,
+      "learning_rate": 5.030172609238304e-06,
+      "loss": 1.1162,
+      "step": 14097
+    },
+    {
+      "epoch": 0.8565526459687709,
+      "grad_norm": 0.2050005942583084,
+      "learning_rate": 5.025988851997826e-06,
+      "loss": 1.0288,
+      "step": 14098
+    },
+    {
+      "epoch": 0.8566134030013974,
+      "grad_norm": 0.18852275609970093,
+      "learning_rate": 5.02180674328841e-06,
+      "loss": 1.0533,
+      "step": 14099
+    },
+    {
+      "epoch": 0.856674160034024,
+      "grad_norm": 0.2342219352722168,
+      "learning_rate": 5.017626283263382e-06,
+      "loss": 1.1181,
+      "step": 14100
+    },
+    {
+      "epoch": 0.8567349170666505,
+      "grad_norm": 0.1488930583000183,
+      "learning_rate": 5.01344747207595e-06,
+      "loss": 1.0883,
+      "step": 14101
+    },
+    {
+      "epoch": 0.856795674099277,
+      "grad_norm": 0.1642390340566635,
+      "learning_rate": 5.009270309879288e-06,
+      "loss": 1.025,
+      "step": 14102
+    },
+    {
+      "epoch": 0.8568564311319036,
+      "grad_norm": 0.14734713733196259,
+      "learning_rate": 5.005094796826526e-06,
+      "loss": 1.0894,
+      "step": 14103
+    },
+    {
+      "epoch": 0.85691718816453,
+      "grad_norm": 0.16555152833461761,
+      "learning_rate": 5.000920933070702e-06,
+      "loss": 1.093,
+      "step": 14104
+    },
+    {
+      "epoch": 0.8569779451971565,
+      "grad_norm": 0.30410251021385193,
+      "learning_rate": 4.9967487187648266e-06,
+      "loss": 1.263,
+      "step": 14105
+    },
+    {
+      "epoch": 0.8570387022297831,
+      "grad_norm": 0.09450112283229828,
+      "learning_rate": 4.992578154061839e-06,
+      "loss": 1.0099,
+      "step": 14106
+    },
+    {
+      "epoch": 0.8570994592624096,
+      "grad_norm": 0.11336357891559601,
+      "learning_rate": 4.988409239114583e-06,
+      "loss": 1.0278,
+      "step": 14107
+    },
+    {
+      "epoch": 0.8571602162950361,
+      "grad_norm": 0.25878921151161194,
+      "learning_rate": 4.984241974075882e-06,
+      "loss": 1.1545,
+      "step": 14108
+    },
+    {
+      "epoch": 0.8572209733276627,
+      "grad_norm": 0.16125600039958954,
+      "learning_rate": 4.980076359098501e-06,
+      "loss": 1.0565,
+      "step": 14109
+    },
+    {
+      "epoch": 0.8572817303602892,
+      "grad_norm": 0.12710823118686676,
+      "learning_rate": 4.975912394335125e-06,
+      "loss": 1.0849,
+      "step": 14110
+    },
+    {
+      "epoch": 0.8573424873929157,
+      "grad_norm": 0.1700657159090042,
+      "learning_rate": 4.971750079938381e-06,
+      "loss": 1.2611,
+      "step": 14111
+    },
+    {
+      "epoch": 0.8574032444255423,
+      "grad_norm": 0.12558159232139587,
+      "learning_rate": 4.967589416060842e-06,
+      "loss": 1.0416,
+      "step": 14112
+    },
+    {
+      "epoch": 0.8574640014581688,
+      "grad_norm": 0.1135503426194191,
+      "learning_rate": 4.9634304028550185e-06,
+      "loss": 1.0282,
+      "step": 14113
+    },
+    {
+      "epoch": 0.8575247584907953,
+      "grad_norm": 0.15996211767196655,
+      "learning_rate": 4.959273040473356e-06,
+      "loss": 1.1139,
+      "step": 14114
+    },
+    {
+      "epoch": 0.8575855155234219,
+      "grad_norm": 0.20730850100517273,
+      "learning_rate": 4.955117329068249e-06,
+      "loss": 1.1364,
+      "step": 14115
+    },
+    {
+      "epoch": 0.8576462725560484,
+      "grad_norm": 0.28644946217536926,
+      "learning_rate": 4.950963268792025e-06,
+      "loss": 1.1264,
+      "step": 14116
+    },
+    {
+      "epoch": 0.8577070295886748,
+      "grad_norm": 0.21113575994968414,
+      "learning_rate": 4.9468108597969465e-06,
+      "loss": 1.0931,
+      "step": 14117
+    },
+    {
+      "epoch": 0.8577677866213014,
+      "grad_norm": 0.9280003309249878,
+      "learning_rate": 4.942660102235225e-06,
+      "loss": 1.1053,
+      "step": 14118
+    },
+    {
+      "epoch": 0.8578285436539279,
+      "grad_norm": 0.2833453118801117,
+      "learning_rate": 4.938510996258999e-06,
+      "loss": 1.0873,
+      "step": 14119
+    },
+    {
+      "epoch": 0.8578893006865544,
+      "grad_norm": 0.1464274525642395,
+      "learning_rate": 4.934363542020381e-06,
+      "loss": 1.008,
+      "step": 14120
+    },
+    {
+      "epoch": 0.857950057719181,
+      "grad_norm": 0.174157977104187,
+      "learning_rate": 4.930217739671367e-06,
+      "loss": 1.1519,
+      "step": 14121
+    },
+    {
+      "epoch": 0.8580108147518075,
+      "grad_norm": 0.14351730048656464,
+      "learning_rate": 4.926073589363939e-06,
+      "loss": 1.0704,
+      "step": 14122
+    },
+    {
+      "epoch": 0.858071571784434,
+      "grad_norm": 0.21605785191059113,
+      "learning_rate": 4.921931091249993e-06,
+      "loss": 1.121,
+      "step": 14123
+    },
+    {
+      "epoch": 0.8581323288170606,
+      "grad_norm": 0.2234364002943039,
+      "learning_rate": 4.917790245481363e-06,
+      "loss": 1.0369,
+      "step": 14124
+    },
+    {
+      "epoch": 0.8581930858496871,
+      "grad_norm": 0.12953612208366394,
+      "learning_rate": 4.913651052209861e-06,
+      "loss": 1.1133,
+      "step": 14125
+    },
+    {
+      "epoch": 0.8582538428823137,
+      "grad_norm": 0.10417594760656357,
+      "learning_rate": 4.909513511587199e-06,
+      "loss": 1.0122,
+      "step": 14126
+    },
+    {
+      "epoch": 0.8583145999149402,
+      "grad_norm": 0.18923239409923553,
+      "learning_rate": 4.905377623765029e-06,
+      "loss": 1.1166,
+      "step": 14127
+    },
+    {
+      "epoch": 0.8583753569475667,
+      "grad_norm": 0.28426477313041687,
+      "learning_rate": 4.901243388894961e-06,
+      "loss": 1.0946,
+      "step": 14128
+    },
+    {
+      "epoch": 0.8584361139801933,
+      "grad_norm": 0.14293935894966125,
+      "learning_rate": 4.897110807128519e-06,
+      "loss": 1.0768,
+      "step": 14129
+    },
+    {
+      "epoch": 0.8584968710128197,
+      "grad_norm": 4.603726863861084,
+      "learning_rate": 4.89297987861721e-06,
+      "loss": 1.0716,
+      "step": 14130
+    },
+    {
+      "epoch": 0.8585576280454462,
+      "grad_norm": 0.15873724222183228,
+      "learning_rate": 4.888850603512446e-06,
+      "loss": 1.0715,
+      "step": 14131
+    },
+    {
+      "epoch": 0.8586183850780728,
+      "grad_norm": 0.16651402413845062,
+      "learning_rate": 4.884722981965578e-06,
+      "loss": 1.0712,
+      "step": 14132
+    },
+    {
+      "epoch": 0.8586791421106993,
+      "grad_norm": 0.1254180669784546,
+      "learning_rate": 4.880597014127925e-06,
+      "loss": 1.0405,
+      "step": 14133
+    },
+    {
+      "epoch": 0.8587398991433258,
+      "grad_norm": 4.309706211090088,
+      "learning_rate": 4.876472700150686e-06,
+      "loss": 1.0467,
+      "step": 14134
+    },
+    {
+      "epoch": 0.8588006561759524,
+      "grad_norm": 0.9488537311553955,
+      "learning_rate": 4.87235004018507e-06,
+      "loss": 1.0888,
+      "step": 14135
+    },
+    {
+      "epoch": 0.8588614132085789,
+      "grad_norm": 0.26401764154434204,
+      "learning_rate": 4.8682290343821875e-06,
+      "loss": 1.2698,
+      "step": 14136
+    },
+    {
+      "epoch": 0.8589221702412054,
+      "grad_norm": 0.1238195076584816,
+      "learning_rate": 4.864109682893098e-06,
+      "loss": 1.0337,
+      "step": 14137
+    },
+    {
+      "epoch": 0.858982927273832,
+      "grad_norm": 0.42624789476394653,
+      "learning_rate": 4.859991985868789e-06,
+      "loss": 1.1827,
+      "step": 14138
+    },
+    {
+      "epoch": 0.8590436843064585,
+      "grad_norm": 0.2221917361021042,
+      "learning_rate": 4.855875943460203e-06,
+      "loss": 1.1298,
+      "step": 14139
+    },
+    {
+      "epoch": 0.859104441339085,
+      "grad_norm": 0.25979480147361755,
+      "learning_rate": 4.851761555818207e-06,
+      "loss": 1.1176,
+      "step": 14140
+    },
+    {
+      "epoch": 0.8591651983717116,
+      "grad_norm": 0.20306643843650818,
+      "learning_rate": 4.847648823093615e-06,
+      "loss": 1.122,
+      "step": 14141
+    },
+    {
+      "epoch": 0.8592259554043381,
+      "grad_norm": 0.10130464285612106,
+      "learning_rate": 4.843537745437188e-06,
+      "loss": 0.9902,
+      "step": 14142
+    },
+    {
+      "epoch": 0.8592867124369645,
+      "grad_norm": 0.1563415229320526,
+      "learning_rate": 4.839428322999612e-06,
+      "loss": 1.1351,
+      "step": 14143
+    },
+    {
+      "epoch": 0.8593474694695911,
+      "grad_norm": 0.2717992067337036,
+      "learning_rate": 4.835320555931522e-06,
+      "loss": 1.043,
+      "step": 14144
+    },
+    {
+      "epoch": 0.8594082265022176,
+      "grad_norm": 0.12131844460964203,
+      "learning_rate": 4.831214444383475e-06,
+      "loss": 1.0318,
+      "step": 14145
+    },
+    {
+      "epoch": 0.8594689835348441,
+      "grad_norm": 0.30003729462623596,
+      "learning_rate": 4.8271099885060125e-06,
+      "loss": 1.0986,
+      "step": 14146
+    },
+    {
+      "epoch": 0.8595297405674707,
+      "grad_norm": 0.09912194311618805,
+      "learning_rate": 4.823007188449557e-06,
+      "loss": 1.0117,
+      "step": 14147
+    },
+    {
+      "epoch": 0.8595904976000972,
+      "grad_norm": 1.0891952514648438,
+      "learning_rate": 4.818906044364507e-06,
+      "loss": 1.2641,
+      "step": 14148
+    },
+    {
+      "epoch": 0.8596512546327237,
+      "grad_norm": 0.17120946943759918,
+      "learning_rate": 4.814806556401186e-06,
+      "loss": 1.1154,
+      "step": 14149
+    },
+    {
+      "epoch": 0.8597120116653503,
+      "grad_norm": 1.8278672695159912,
+      "learning_rate": 4.810708724709856e-06,
+      "loss": 1.0324,
+      "step": 14150
+    },
+    {
+      "epoch": 0.8597727686979768,
+      "grad_norm": 0.11879687756299973,
+      "learning_rate": 4.806612549440742e-06,
+      "loss": 1.0133,
+      "step": 14151
+    },
+    {
+      "epoch": 0.8598335257306033,
+      "grad_norm": 0.18011632561683655,
+      "learning_rate": 4.802518030743985e-06,
+      "loss": 1.0537,
+      "step": 14152
+    },
+    {
+      "epoch": 0.8598942827632299,
+      "grad_norm": 0.1315927356481552,
+      "learning_rate": 4.798425168769671e-06,
+      "loss": 1.0653,
+      "step": 14153
+    },
+    {
+      "epoch": 0.8599550397958564,
+      "grad_norm": 0.19348208606243134,
+      "learning_rate": 4.794333963667813e-06,
+      "loss": 1.1729,
+      "step": 14154
+    },
+    {
+      "epoch": 0.860015796828483,
+      "grad_norm": 0.24430322647094727,
+      "learning_rate": 4.790244415588369e-06,
+      "loss": 1.0538,
+      "step": 14155
+    },
+    {
+      "epoch": 0.8600765538611094,
+      "grad_norm": 0.27006277441978455,
+      "learning_rate": 4.786156524681268e-06,
+      "loss": 1.2019,
+      "step": 14156
+    },
+    {
+      "epoch": 0.8601373108937359,
+      "grad_norm": 0.22173461318016052,
+      "learning_rate": 4.782070291096336e-06,
+      "loss": 1.0858,
+      "step": 14157
+    },
+    {
+      "epoch": 0.8601980679263624,
+      "grad_norm": 0.10452301800251007,
+      "learning_rate": 4.777985714983357e-06,
+      "loss": 1.0287,
+      "step": 14158
+    },
+    {
+      "epoch": 0.860258824958989,
+      "grad_norm": 0.12470521032810211,
+      "learning_rate": 4.773902796492052e-06,
+      "loss": 1.0533,
+      "step": 14159
+    },
+    {
+      "epoch": 0.8603195819916155,
+      "grad_norm": 0.26975512504577637,
+      "learning_rate": 4.7698215357720895e-06,
+      "loss": 1.0738,
+      "step": 14160
+    },
+    {
+      "epoch": 0.860380339024242,
+      "grad_norm": 0.19559358060359955,
+      "learning_rate": 4.765741932973055e-06,
+      "loss": 1.1401,
+      "step": 14161
+    },
+    {
+      "epoch": 0.8604410960568686,
+      "grad_norm": 0.10784042626619339,
+      "learning_rate": 4.761663988244497e-06,
+      "loss": 1.0178,
+      "step": 14162
+    },
+    {
+      "epoch": 0.8605018530894951,
+      "grad_norm": 0.15287598967552185,
+      "learning_rate": 4.757587701735883e-06,
+      "loss": 1.0631,
+      "step": 14163
+    },
+    {
+      "epoch": 0.8605626101221217,
+      "grad_norm": 0.1741325855255127,
+      "learning_rate": 4.753513073596644e-06,
+      "loss": 1.1007,
+      "step": 14164
+    },
+    {
+      "epoch": 0.8606233671547482,
+      "grad_norm": 0.807770848274231,
+      "learning_rate": 4.749440103976127e-06,
+      "loss": 1.0559,
+      "step": 14165
+    },
+    {
+      "epoch": 0.8606841241873747,
+      "grad_norm": 0.352888822555542,
+      "learning_rate": 4.745368793023619e-06,
+      "loss": 1.2144,
+      "step": 14166
+    },
+    {
+      "epoch": 0.8607448812200013,
+      "grad_norm": 45.675228118896484,
+      "learning_rate": 4.741299140888389e-06,
+      "loss": 1.0253,
+      "step": 14167
+    },
+    {
+      "epoch": 0.8608056382526278,
+      "grad_norm": 15.36818790435791,
+      "learning_rate": 4.737231147719573e-06,
+      "loss": 1.0231,
+      "step": 14168
+    },
+    {
+      "epoch": 0.8608663952852542,
+      "grad_norm": 0.1279137283563614,
+      "learning_rate": 4.733164813666302e-06,
+      "loss": 1.052,
+      "step": 14169
+    },
+    {
+      "epoch": 0.8609271523178808,
+      "grad_norm": 0.16352872550487518,
+      "learning_rate": 4.729100138877623e-06,
+      "loss": 1.0183,
+      "step": 14170
+    },
+    {
+      "epoch": 0.8609879093505073,
+      "grad_norm": 0.14424246549606323,
+      "learning_rate": 4.725037123502518e-06,
+      "loss": 1.0729,
+      "step": 14171
+    },
+    {
+      "epoch": 0.8610486663831338,
+      "grad_norm": 0.12785501778125763,
+      "learning_rate": 4.720975767689939e-06,
+      "loss": 1.0509,
+      "step": 14172
+    },
+    {
+      "epoch": 0.8611094234157604,
+      "grad_norm": 0.1404189169406891,
+      "learning_rate": 4.7169160715887566e-06,
+      "loss": 1.0793,
+      "step": 14173
+    },
+    {
+      "epoch": 0.8611701804483869,
+      "grad_norm": 0.21100686490535736,
+      "learning_rate": 4.712858035347756e-06,
+      "loss": 1.0954,
+      "step": 14174
+    },
+    {
+      "epoch": 0.8612309374810134,
+      "grad_norm": 6.07974910736084,
+      "learning_rate": 4.708801659115702e-06,
+      "loss": 0.9877,
+      "step": 14175
+    },
+    {
+      "epoch": 0.86129169451364,
+      "grad_norm": 0.17326590418815613,
+      "learning_rate": 4.704746943041266e-06,
+      "loss": 1.0594,
+      "step": 14176
+    },
+    {
+      "epoch": 0.8613524515462665,
+      "grad_norm": 0.11902457475662231,
+      "learning_rate": 4.7006938872730945e-06,
+      "loss": 1.082,
+      "step": 14177
+    },
+    {
+      "epoch": 0.861413208578893,
+      "grad_norm": 0.09663273394107819,
+      "learning_rate": 4.696642491959746e-06,
+      "loss": 1.0214,
+      "step": 14178
+    },
+    {
+      "epoch": 0.8614739656115196,
+      "grad_norm": 0.1515512764453888,
+      "learning_rate": 4.692592757249725e-06,
+      "loss": 1.075,
+      "step": 14179
+    },
+    {
+      "epoch": 0.8615347226441461,
+      "grad_norm": 0.13148732483386993,
+      "learning_rate": 4.688544683291479e-06,
+      "loss": 1.109,
+      "step": 14180
+    },
+    {
+      "epoch": 0.8615954796767726,
+      "grad_norm": 0.17569436132907867,
+      "learning_rate": 4.6844982702333725e-06,
+      "loss": 1.0373,
+      "step": 14181
+    },
+    {
+      "epoch": 0.8616562367093991,
+      "grad_norm": 0.12792392075061798,
+      "learning_rate": 4.6804535182237485e-06,
+      "loss": 1.0156,
+      "step": 14182
+    },
+    {
+      "epoch": 0.8617169937420256,
+      "grad_norm": 0.16717316210269928,
+      "learning_rate": 4.67641042741086e-06,
+      "loss": 1.1186,
+      "step": 14183
+    },
+    {
+      "epoch": 0.8617777507746521,
+      "grad_norm": 0.15593482553958893,
+      "learning_rate": 4.672368997942911e-06,
+      "loss": 1.1887,
+      "step": 14184
+    },
+    {
+      "epoch": 0.8618385078072787,
+      "grad_norm": 0.25899389386177063,
+      "learning_rate": 4.668329229968038e-06,
+      "loss": 1.0453,
+      "step": 14185
+    },
+    {
+      "epoch": 0.8618992648399052,
+      "grad_norm": 0.15924355387687683,
+      "learning_rate": 4.664291123634318e-06,
+      "loss": 1.0993,
+      "step": 14186
+    },
+    {
+      "epoch": 0.8619600218725317,
+      "grad_norm": 0.12288684397935867,
+      "learning_rate": 4.660254679089771e-06,
+      "loss": 1.053,
+      "step": 14187
+    },
+    {
+      "epoch": 0.8620207789051583,
+      "grad_norm": 0.11554786562919617,
+      "learning_rate": 4.65621989648235e-06,
+      "loss": 1.0203,
+      "step": 14188
+    },
+    {
+      "epoch": 0.8620815359377848,
+      "grad_norm": 0.1808425486087799,
+      "learning_rate": 4.652186775959954e-06,
+      "loss": 1.1128,
+      "step": 14189
+    },
+    {
+      "epoch": 0.8621422929704113,
+      "grad_norm": 0.20926572382450104,
+      "learning_rate": 4.648155317670422e-06,
+      "loss": 1.1271,
+      "step": 14190
+    },
+    {
+      "epoch": 0.8622030500030379,
+      "grad_norm": 0.23795105516910553,
+      "learning_rate": 4.644125521761516e-06,
+      "loss": 1.0368,
+      "step": 14191
+    },
+    {
+      "epoch": 0.8622638070356644,
+      "grad_norm": 0.1222681999206543,
+      "learning_rate": 4.640097388380954e-06,
+      "loss": 0.9786,
+      "step": 14192
+    },
+    {
+      "epoch": 0.8623245640682909,
+      "grad_norm": 0.16433022916316986,
+      "learning_rate": 4.636070917676405e-06,
+      "loss": 1.0514,
+      "step": 14193
+    },
+    {
+      "epoch": 0.8623853211009175,
+      "grad_norm": 0.12489213049411774,
+      "learning_rate": 4.632046109795435e-06,
+      "loss": 1.014,
+      "step": 14194
+    },
+    {
+      "epoch": 0.8624460781335439,
+      "grad_norm": 0.2033950239419937,
+      "learning_rate": 4.628022964885587e-06,
+      "loss": 1.1627,
+      "step": 14195
+    },
+    {
+      "epoch": 0.8625068351661704,
+      "grad_norm": 0.22645170986652374,
+      "learning_rate": 4.62400148309432e-06,
+      "loss": 1.1643,
+      "step": 14196
+    },
+    {
+      "epoch": 0.862567592198797,
+      "grad_norm": 0.10457313805818558,
+      "learning_rate": 4.6199816645690505e-06,
+      "loss": 1.0445,
+      "step": 14197
+    },
+    {
+      "epoch": 0.8626283492314235,
+      "grad_norm": 0.16166149079799652,
+      "learning_rate": 4.615963509457128e-06,
+      "loss": 1.1237,
+      "step": 14198
+    },
+    {
+      "epoch": 0.86268910626405,
+      "grad_norm": 1.4195982217788696,
+      "learning_rate": 4.611947017905838e-06,
+      "loss": 1.0219,
+      "step": 14199
+    },
+    {
+      "epoch": 0.8627498632966766,
+      "grad_norm": 0.25008589029312134,
+      "learning_rate": 4.607932190062408e-06,
+      "loss": 1.181,
+      "step": 14200
+    },
+    {
+      "epoch": 0.8628106203293031,
+      "grad_norm": 0.19807790219783783,
+      "learning_rate": 4.603919026073989e-06,
+      "loss": 1.1015,
+      "step": 14201
+    },
+    {
+      "epoch": 0.8628713773619296,
+      "grad_norm": 0.21317186951637268,
+      "learning_rate": 4.5999075260876935e-06,
+      "loss": 1.0791,
+      "step": 14202
+    },
+    {
+      "epoch": 0.8629321343945562,
+      "grad_norm": 0.13090331852436066,
+      "learning_rate": 4.595897690250567e-06,
+      "loss": 1.0227,
+      "step": 14203
+    },
+    {
+      "epoch": 0.8629928914271827,
+      "grad_norm": 0.2109454721212387,
+      "learning_rate": 4.591889518709585e-06,
+      "loss": 1.2276,
+      "step": 14204
+    },
+    {
+      "epoch": 0.8630536484598093,
+      "grad_norm": 0.10929658263921738,
+      "learning_rate": 4.587883011611671e-06,
+      "loss": 1.0149,
+      "step": 14205
+    },
+    {
+      "epoch": 0.8631144054924358,
+      "grad_norm": 0.6618485450744629,
+      "learning_rate": 4.583878169103684e-06,
+      "loss": 1.0616,
+      "step": 14206
+    },
+    {
+      "epoch": 0.8631751625250623,
+      "grad_norm": 0.1818619817495346,
+      "learning_rate": 4.579874991332423e-06,
+      "loss": 1.1181,
+      "step": 14207
+    },
+    {
+      "epoch": 0.8632359195576889,
+      "grad_norm": 0.17064742743968964,
+      "learning_rate": 4.57587347844462e-06,
+      "loss": 1.0393,
+      "step": 14208
+    },
+    {
+      "epoch": 0.8632966765903153,
+      "grad_norm": 0.20591960847377777,
+      "learning_rate": 4.571873630586954e-06,
+      "loss": 1.1793,
+      "step": 14209
+    },
+    {
+      "epoch": 0.8633574336229418,
+      "grad_norm": 0.2499299794435501,
+      "learning_rate": 4.567875447906045e-06,
+      "loss": 1.0405,
+      "step": 14210
+    },
+    {
+      "epoch": 0.8634181906555684,
+      "grad_norm": 0.14137180149555206,
+      "learning_rate": 4.563878930548443e-06,
+      "loss": 1.0978,
+      "step": 14211
+    },
+    {
+      "epoch": 0.8634789476881949,
+      "grad_norm": 0.14389704167842865,
+      "learning_rate": 4.559884078660626e-06,
+      "loss": 1.145,
+      "step": 14212
+    },
+    {
+      "epoch": 0.8635397047208214,
+      "grad_norm": 0.2118283361196518,
+      "learning_rate": 4.55589089238907e-06,
+      "loss": 1.0046,
+      "step": 14213
+    },
+    {
+      "epoch": 0.863600461753448,
+      "grad_norm": 1.3960285186767578,
+      "learning_rate": 4.551899371880103e-06,
+      "loss": 1.0388,
+      "step": 14214
+    },
+    {
+      "epoch": 0.8636612187860745,
+      "grad_norm": 0.1342649757862091,
+      "learning_rate": 4.547909517280047e-06,
+      "loss": 1.0483,
+      "step": 14215
+    },
+    {
+      "epoch": 0.863721975818701,
+      "grad_norm": 0.12921537458896637,
+      "learning_rate": 4.543921328735157e-06,
+      "loss": 1.0534,
+      "step": 14216
+    },
+    {
+      "epoch": 0.8637827328513276,
+      "grad_norm": 0.12537027895450592,
+      "learning_rate": 4.539934806391605e-06,
+      "loss": 1.0032,
+      "step": 14217
+    },
+    {
+      "epoch": 0.8638434898839541,
+      "grad_norm": 0.11934049427509308,
+      "learning_rate": 4.535949950395541e-06,
+      "loss": 1.055,
+      "step": 14218
+    },
+    {
+      "epoch": 0.8639042469165806,
+      "grad_norm": 0.13322697579860687,
+      "learning_rate": 4.53196676089302e-06,
+      "loss": 1.0497,
+      "step": 14219
+    },
+    {
+      "epoch": 0.8639650039492072,
+      "grad_norm": 0.18889231979846954,
+      "learning_rate": 4.5279852380300524e-06,
+      "loss": 1.0634,
+      "step": 14220
+    },
+    {
+      "epoch": 0.8640257609818337,
+      "grad_norm": 0.17164137959480286,
+      "learning_rate": 4.5240053819525726e-06,
+      "loss": 1.0604,
+      "step": 14221
+    },
+    {
+      "epoch": 0.8640865180144601,
+      "grad_norm": 0.14856575429439545,
+      "learning_rate": 4.520027192806453e-06,
+      "loss": 1.1066,
+      "step": 14222
+    },
+    {
+      "epoch": 0.8641472750470867,
+      "grad_norm": 0.38683027029037476,
+      "learning_rate": 4.516050670737537e-06,
+      "loss": 1.061,
+      "step": 14223
+    },
+    {
+      "epoch": 0.8642080320797132,
+      "grad_norm": 0.42218300700187683,
+      "learning_rate": 4.5120758158915755e-06,
+      "loss": 1.0734,
+      "step": 14224
+    },
+    {
+      "epoch": 0.8642687891123397,
+      "grad_norm": 0.23691798746585846,
+      "learning_rate": 4.5081026284142625e-06,
+      "loss": 1.062,
+      "step": 14225
+    },
+    {
+      "epoch": 0.8643295461449663,
+      "grad_norm": 0.10736768692731857,
+      "learning_rate": 4.504131108451248e-06,
+      "loss": 1.0433,
+      "step": 14226
+    },
+    {
+      "epoch": 0.8643903031775928,
+      "grad_norm": 0.1270807534456253,
+      "learning_rate": 4.5001612561481e-06,
+      "loss": 1.0299,
+      "step": 14227
+    },
+    {
+      "epoch": 0.8644510602102193,
+      "grad_norm": 0.13499632477760315,
+      "learning_rate": 4.496193071650334e-06,
+      "loss": 1.0845,
+      "step": 14228
+    },
+    {
+      "epoch": 0.8645118172428459,
+      "grad_norm": 0.15015171468257904,
+      "learning_rate": 4.492226555103407e-06,
+      "loss": 1.0697,
+      "step": 14229
+    },
+    {
+      "epoch": 0.8645725742754724,
+      "grad_norm": 0.1274319738149643,
+      "learning_rate": 4.488261706652713e-06,
+      "loss": 1.041,
+      "step": 14230
+    },
+    {
+      "epoch": 0.8646333313080989,
+      "grad_norm": 0.1719062626361847,
+      "learning_rate": 4.4842985264435864e-06,
+      "loss": 1.153,
+      "step": 14231
+    },
+    {
+      "epoch": 0.8646940883407255,
+      "grad_norm": 0.23702573776245117,
+      "learning_rate": 4.480337014621294e-06,
+      "loss": 1.1169,
+      "step": 14232
+    },
+    {
+      "epoch": 0.864754845373352,
+      "grad_norm": 0.16583359241485596,
+      "learning_rate": 4.476377171331037e-06,
+      "loss": 1.0569,
+      "step": 14233
+    },
+    {
+      "epoch": 0.8648156024059785,
+      "grad_norm": 0.15656907856464386,
+      "learning_rate": 4.472418996717992e-06,
+      "loss": 1.0844,
+      "step": 14234
+    },
+    {
+      "epoch": 0.864876359438605,
+      "grad_norm": 0.22169530391693115,
+      "learning_rate": 4.468462490927216e-06,
+      "loss": 1.0993,
+      "step": 14235
+    },
+    {
+      "epoch": 0.8649371164712315,
+      "grad_norm": 0.1542995125055313,
+      "learning_rate": 4.464507654103756e-06,
+      "loss": 1.0316,
+      "step": 14236
+    },
+    {
+      "epoch": 0.864997873503858,
+      "grad_norm": 0.10396695137023926,
+      "learning_rate": 4.4605544863925654e-06,
+      "loss": 1.0052,
+      "step": 14237
+    },
+    {
+      "epoch": 0.8650586305364846,
+      "grad_norm": 0.29740509390830994,
+      "learning_rate": 4.456602987938546e-06,
+      "loss": 1.1668,
+      "step": 14238
+    },
+    {
+      "epoch": 0.8651193875691111,
+      "grad_norm": 0.14284315705299377,
+      "learning_rate": 4.452653158886555e-06,
+      "loss": 1.083,
+      "step": 14239
+    },
+    {
+      "epoch": 0.8651801446017376,
+      "grad_norm": 0.11881954222917557,
+      "learning_rate": 4.448704999381376e-06,
+      "loss": 1.0358,
+      "step": 14240
+    },
+    {
+      "epoch": 0.8652409016343642,
+      "grad_norm": 0.11248598247766495,
+      "learning_rate": 4.444758509567709e-06,
+      "loss": 1.1083,
+      "step": 14241
+    },
+    {
+      "epoch": 0.8653016586669907,
+      "grad_norm": 0.18575093150138855,
+      "learning_rate": 4.4408136895902286e-06,
+      "loss": 1.0914,
+      "step": 14242
+    },
+    {
+      "epoch": 0.8653624156996172,
+      "grad_norm": 2.2854063510894775,
+      "learning_rate": 4.436870539593513e-06,
+      "loss": 1.0581,
+      "step": 14243
+    },
+    {
+      "epoch": 0.8654231727322438,
+      "grad_norm": 0.15236227214336395,
+      "learning_rate": 4.4329290597221295e-06,
+      "loss": 1.1288,
+      "step": 14244
+    },
+    {
+      "epoch": 0.8654839297648703,
+      "grad_norm": 0.5712029933929443,
+      "learning_rate": 4.428989250120541e-06,
+      "loss": 1.2311,
+      "step": 14245
+    },
+    {
+      "epoch": 0.8655446867974969,
+      "grad_norm": 0.14029696583747864,
+      "learning_rate": 4.4250511109331534e-06,
+      "loss": 1.0724,
+      "step": 14246
+    },
+    {
+      "epoch": 0.8656054438301234,
+      "grad_norm": 0.12488316744565964,
+      "learning_rate": 4.421114642304342e-06,
+      "loss": 1.0861,
+      "step": 14247
+    },
+    {
+      "epoch": 0.8656662008627498,
+      "grad_norm": 0.20525529980659485,
+      "learning_rate": 4.417179844378361e-06,
+      "loss": 1.1568,
+      "step": 14248
+    },
+    {
+      "epoch": 0.8657269578953763,
+      "grad_norm": 0.13475671410560608,
+      "learning_rate": 4.413246717299474e-06,
+      "loss": 1.0144,
+      "step": 14249
+    },
+    {
+      "epoch": 0.8657877149280029,
+      "grad_norm": 0.11134091019630432,
+      "learning_rate": 4.4093152612118434e-06,
+      "loss": 1.0229,
+      "step": 14250
+    },
+    {
+      "epoch": 0.8658484719606294,
+      "grad_norm": 0.1131577268242836,
+      "learning_rate": 4.405385476259571e-06,
+      "loss": 1.0198,
+      "step": 14251
+    },
+    {
+      "epoch": 0.865909228993256,
+      "grad_norm": 0.10383344441652298,
+      "learning_rate": 4.401457362586709e-06,
+      "loss": 1.0528,
+      "step": 14252
+    },
+    {
+      "epoch": 0.8659699860258825,
+      "grad_norm": 0.17939306795597076,
+      "learning_rate": 4.3975309203372396e-06,
+      "loss": 1.2157,
+      "step": 14253
+    },
+    {
+      "epoch": 0.866030743058509,
+      "grad_norm": 0.18818438053131104,
+      "learning_rate": 4.393606149655094e-06,
+      "loss": 0.9997,
+      "step": 14254
+    },
+    {
+      "epoch": 0.8660915000911356,
+      "grad_norm": 0.31205496191978455,
+      "learning_rate": 4.389683050684124e-06,
+      "loss": 1.0796,
+      "step": 14255
+    },
+    {
+      "epoch": 0.8661522571237621,
+      "grad_norm": 0.17409072816371918,
+      "learning_rate": 4.385761623568141e-06,
+      "loss": 1.2496,
+      "step": 14256
+    },
+    {
+      "epoch": 0.8662130141563886,
+      "grad_norm": 0.33971136808395386,
+      "learning_rate": 4.381841868450881e-06,
+      "loss": 1.0332,
+      "step": 14257
+    },
+    {
+      "epoch": 0.8662737711890152,
+      "grad_norm": 0.13563384115695953,
+      "learning_rate": 4.377923785476029e-06,
+      "loss": 1.0465,
+      "step": 14258
+    },
+    {
+      "epoch": 0.8663345282216417,
+      "grad_norm": 0.2389136403799057,
+      "learning_rate": 4.3740073747871865e-06,
+      "loss": 1.0715,
+      "step": 14259
+    },
+    {
+      "epoch": 0.8663952852542682,
+      "grad_norm": 0.6087927222251892,
+      "learning_rate": 4.3700926365279445e-06,
+      "loss": 1.1517,
+      "step": 14260
+    },
+    {
+      "epoch": 0.8664560422868947,
+      "grad_norm": 0.1144602820277214,
+      "learning_rate": 4.366179570841766e-06,
+      "loss": 0.9902,
+      "step": 14261
+    },
+    {
+      "epoch": 0.8665167993195212,
+      "grad_norm": 0.12300249189138412,
+      "learning_rate": 4.362268177872097e-06,
+      "loss": 1.024,
+      "step": 14262
+    },
+    {
+      "epoch": 0.8665775563521477,
+      "grad_norm": 0.14014755189418793,
+      "learning_rate": 4.358358457762307e-06,
+      "loss": 1.0035,
+      "step": 14263
+    },
+    {
+      "epoch": 0.8666383133847743,
+      "grad_norm": 0.2034769058227539,
+      "learning_rate": 4.3544504106557025e-06,
+      "loss": 1.2037,
+      "step": 14264
+    },
+    {
+      "epoch": 0.8666990704174008,
+      "grad_norm": 0.583436131477356,
+      "learning_rate": 4.350544036695553e-06,
+      "loss": 1.1972,
+      "step": 14265
+    },
+    {
+      "epoch": 0.8667598274500273,
+      "grad_norm": 0.16248038411140442,
+      "learning_rate": 4.346639336025033e-06,
+      "loss": 1.0567,
+      "step": 14266
+    },
+    {
+      "epoch": 0.8668205844826539,
+      "grad_norm": 2.0765669345855713,
+      "learning_rate": 4.3427363087872815e-06,
+      "loss": 1.1516,
+      "step": 14267
+    },
+    {
+      "epoch": 0.8668813415152804,
+      "grad_norm": 0.206801176071167,
+      "learning_rate": 4.338834955125343e-06,
+      "loss": 1.0645,
+      "step": 14268
+    },
+    {
+      "epoch": 0.8669420985479069,
+      "grad_norm": 0.18729735910892487,
+      "learning_rate": 4.334935275182234e-06,
+      "loss": 1.0654,
+      "step": 14269
+    },
+    {
+      "epoch": 0.8670028555805335,
+      "grad_norm": 0.6838131546974182,
+      "learning_rate": 4.331037269100901e-06,
+      "loss": 1.2174,
+      "step": 14270
+    },
+    {
+      "epoch": 0.86706361261316,
+      "grad_norm": 0.14661040902137756,
+      "learning_rate": 4.327140937024232e-06,
+      "loss": 1.0889,
+      "step": 14271
+    },
+    {
+      "epoch": 0.8671243696457865,
+      "grad_norm": 0.15410764515399933,
+      "learning_rate": 4.323246279095033e-06,
+      "loss": 1.0795,
+      "step": 14272
+    },
+    {
+      "epoch": 0.8671851266784131,
+      "grad_norm": 0.17606182396411896,
+      "learning_rate": 4.319353295456074e-06,
+      "loss": 1.1158,
+      "step": 14273
+    },
+    {
+      "epoch": 0.8672458837110395,
+      "grad_norm": 0.2107379138469696,
+      "learning_rate": 4.315461986250052e-06,
+      "loss": 1.0633,
+      "step": 14274
+    },
+    {
+      "epoch": 0.867306640743666,
+      "grad_norm": 0.19956208765506744,
+      "learning_rate": 4.311572351619603e-06,
+      "loss": 1.058,
+      "step": 14275
+    },
+    {
+      "epoch": 0.8673673977762926,
+      "grad_norm": 0.4064960181713104,
+      "learning_rate": 4.3076843917072955e-06,
+      "loss": 1.1979,
+      "step": 14276
+    },
+    {
+      "epoch": 0.8674281548089191,
+      "grad_norm": 0.21979095041751862,
+      "learning_rate": 4.303798106655649e-06,
+      "loss": 1.0984,
+      "step": 14277
+    },
+    {
+      "epoch": 0.8674889118415456,
+      "grad_norm": 0.14662480354309082,
+      "learning_rate": 4.299913496607122e-06,
+      "loss": 1.028,
+      "step": 14278
+    },
+    {
+      "epoch": 0.8675496688741722,
+      "grad_norm": 0.21457675099372864,
+      "learning_rate": 4.2960305617040955e-06,
+      "loss": 1.1901,
+      "step": 14279
+    },
+    {
+      "epoch": 0.8676104259067987,
+      "grad_norm": 0.10797221213579178,
+      "learning_rate": 4.292149302088899e-06,
+      "loss": 1.0355,
+      "step": 14280
+    },
+    {
+      "epoch": 0.8676711829394252,
+      "grad_norm": 0.14204855263233185,
+      "learning_rate": 4.288269717903809e-06,
+      "loss": 1.0063,
+      "step": 14281
+    },
+    {
+      "epoch": 0.8677319399720518,
+      "grad_norm": 0.11679262667894363,
+      "learning_rate": 4.284391809291027e-06,
+      "loss": 1.0634,
+      "step": 14282
+    },
+    {
+      "epoch": 0.8677926970046783,
+      "grad_norm": 0.1416119486093521,
+      "learning_rate": 4.280515576392702e-06,
+      "loss": 1.0696,
+      "step": 14283
+    },
+    {
+      "epoch": 0.8678534540373049,
+      "grad_norm": 0.5395480990409851,
+      "learning_rate": 4.2766410193509136e-06,
+      "loss": 1.1584,
+      "step": 14284
+    },
+    {
+      "epoch": 0.8679142110699314,
+      "grad_norm": 0.13430988788604736,
+      "learning_rate": 4.2727681383076816e-06,
+      "loss": 1.0678,
+      "step": 14285
+    },
+    {
+      "epoch": 0.8679749681025579,
+      "grad_norm": 4.478382587432861,
+      "learning_rate": 4.268896933404981e-06,
+      "loss": 1.1018,
+      "step": 14286
+    },
+    {
+      "epoch": 0.8680357251351843,
+      "grad_norm": 0.11822658032178879,
+      "learning_rate": 4.265027404784705e-06,
+      "loss": 1.0135,
+      "step": 14287
+    },
+    {
+      "epoch": 0.8680964821678109,
+      "grad_norm": 0.20759336650371552,
+      "learning_rate": 4.261159552588689e-06,
+      "loss": 1.0894,
+      "step": 14288
+    },
+    {
+      "epoch": 0.8681572392004374,
+      "grad_norm": 0.11896127462387085,
+      "learning_rate": 4.2572933769587045e-06,
+      "loss": 1.043,
+      "step": 14289
+    },
+    {
+      "epoch": 0.868217996233064,
+      "grad_norm": 0.10811272263526917,
+      "learning_rate": 4.25342887803647e-06,
+      "loss": 1.0165,
+      "step": 14290
+    },
+    {
+      "epoch": 0.8682787532656905,
+      "grad_norm": 0.1244196742773056,
+      "learning_rate": 4.24956605596365e-06,
+      "loss": 1.0522,
+      "step": 14291
+    },
+    {
+      "epoch": 0.868339510298317,
+      "grad_norm": 0.19802682101726532,
+      "learning_rate": 4.245704910881826e-06,
+      "loss": 1.0674,
+      "step": 14292
+    },
+    {
+      "epoch": 0.8684002673309436,
+      "grad_norm": 0.3971807062625885,
+      "learning_rate": 4.241845442932535e-06,
+      "loss": 1.1721,
+      "step": 14293
+    },
+    {
+      "epoch": 0.8684610243635701,
+      "grad_norm": 0.5288057327270508,
+      "learning_rate": 4.237987652257242e-06,
+      "loss": 1.1029,
+      "step": 14294
+    },
+    {
+      "epoch": 0.8685217813961966,
+      "grad_norm": 0.13204018771648407,
+      "learning_rate": 4.23413153899736e-06,
+      "loss": 1.0733,
+      "step": 14295
+    },
+    {
+      "epoch": 0.8685825384288232,
+      "grad_norm": 0.10832633078098297,
+      "learning_rate": 4.2302771032942334e-06,
+      "loss": 1.0083,
+      "step": 14296
+    },
+    {
+      "epoch": 0.8686432954614497,
+      "grad_norm": 0.10118073970079422,
+      "learning_rate": 4.226424345289143e-06,
+      "loss": 1.0782,
+      "step": 14297
+    },
+    {
+      "epoch": 0.8687040524940762,
+      "grad_norm": 0.19373410940170288,
+      "learning_rate": 4.2225732651233195e-06,
+      "loss": 1.0414,
+      "step": 14298
+    },
+    {
+      "epoch": 0.8687648095267028,
+      "grad_norm": 0.20592525601387024,
+      "learning_rate": 4.218723862937917e-06,
+      "loss": 1.1401,
+      "step": 14299
+    },
+    {
+      "epoch": 0.8688255665593292,
+      "grad_norm": 0.5029820799827576,
+      "learning_rate": 4.214876138874046e-06,
+      "loss": 1.1679,
+      "step": 14300
+    },
+    {
+      "epoch": 0.8688863235919557,
+      "grad_norm": 0.130966916680336,
+      "learning_rate": 4.211030093072738e-06,
+      "loss": 1.0334,
+      "step": 14301
+    },
+    {
+      "epoch": 0.8689470806245823,
+      "grad_norm": 0.13666383922100067,
+      "learning_rate": 4.2071857256749725e-06,
+      "loss": 1.0496,
+      "step": 14302
+    },
+    {
+      "epoch": 0.8690078376572088,
+      "grad_norm": 0.09910415858030319,
+      "learning_rate": 4.203343036821667e-06,
+      "loss": 1.0108,
+      "step": 14303
+    },
+    {
+      "epoch": 0.8690685946898353,
+      "grad_norm": 0.2526373267173767,
+      "learning_rate": 4.199502026653674e-06,
+      "loss": 1.1438,
+      "step": 14304
+    },
+    {
+      "epoch": 0.8691293517224619,
+      "grad_norm": 0.5141761898994446,
+      "learning_rate": 4.195662695311786e-06,
+      "loss": 1.223,
+      "step": 14305
+    },
+    {
+      "epoch": 0.8691901087550884,
+      "grad_norm": 0.16585469245910645,
+      "learning_rate": 4.1918250429367255e-06,
+      "loss": 1.1055,
+      "step": 14306
+    },
+    {
+      "epoch": 0.8692508657877149,
+      "grad_norm": 0.13301222026348114,
+      "learning_rate": 4.187989069669196e-06,
+      "loss": 1.065,
+      "step": 14307
+    },
+    {
+      "epoch": 0.8693116228203415,
+      "grad_norm": 0.18069688975811005,
+      "learning_rate": 4.184154775649768e-06,
+      "loss": 1.0944,
+      "step": 14308
+    },
+    {
+      "epoch": 0.869372379852968,
+      "grad_norm": 0.14474423229694366,
+      "learning_rate": 4.180322161019007e-06,
+      "loss": 1.0705,
+      "step": 14309
+    },
+    {
+      "epoch": 0.8694331368855945,
+      "grad_norm": 0.1404387652873993,
+      "learning_rate": 4.176491225917384e-06,
+      "loss": 1.0815,
+      "step": 14310
+    },
+    {
+      "epoch": 0.8694938939182211,
+      "grad_norm": 0.12162946909666061,
+      "learning_rate": 4.172661970485336e-06,
+      "loss": 1.0218,
+      "step": 14311
+    },
+    {
+      "epoch": 0.8695546509508476,
+      "grad_norm": 0.12406567484140396,
+      "learning_rate": 4.168834394863231e-06,
+      "loss": 1.0487,
+      "step": 14312
+    },
+    {
+      "epoch": 0.8696154079834741,
+      "grad_norm": 0.19263829290866852,
+      "learning_rate": 4.165008499191353e-06,
+      "loss": 1.1119,
+      "step": 14313
+    },
+    {
+      "epoch": 0.8696761650161006,
+      "grad_norm": 0.1953824758529663,
+      "learning_rate": 4.161184283609959e-06,
+      "loss": 1.0972,
+      "step": 14314
+    },
+    {
+      "epoch": 0.8697369220487271,
+      "grad_norm": 0.1113588958978653,
+      "learning_rate": 4.157361748259198e-06,
+      "loss": 1.0449,
+      "step": 14315
+    },
+    {
+      "epoch": 0.8697976790813536,
+      "grad_norm": 0.12395302951335907,
+      "learning_rate": 4.153540893279212e-06,
+      "loss": 1.0492,
+      "step": 14316
+    },
+    {
+      "epoch": 0.8698584361139802,
+      "grad_norm": 0.17302490770816803,
+      "learning_rate": 4.149721718810051e-06,
+      "loss": 1.0272,
+      "step": 14317
+    },
+    {
+      "epoch": 0.8699191931466067,
+      "grad_norm": 0.14489449560642242,
+      "learning_rate": 4.145904224991698e-06,
+      "loss": 1.0945,
+      "step": 14318
+    },
+    {
+      "epoch": 0.8699799501792332,
+      "grad_norm": 0.16393961012363434,
+      "learning_rate": 4.1420884119640905e-06,
+      "loss": 1.094,
+      "step": 14319
+    },
+    {
+      "epoch": 0.8700407072118598,
+      "grad_norm": 0.10775593668222427,
+      "learning_rate": 4.1382742798670935e-06,
+      "loss": 1.0414,
+      "step": 14320
+    },
+    {
+      "epoch": 0.8701014642444863,
+      "grad_norm": 0.23243103921413422,
+      "learning_rate": 4.134461828840519e-06,
+      "loss": 1.2255,
+      "step": 14321
+    },
+    {
+      "epoch": 0.8701622212771128,
+      "grad_norm": 0.15937195718288422,
+      "learning_rate": 4.130651059024116e-06,
+      "loss": 1.0756,
+      "step": 14322
+    },
+    {
+      "epoch": 0.8702229783097394,
+      "grad_norm": 3.0323517322540283,
+      "learning_rate": 4.12684197055756e-06,
+      "loss": 1.0131,
+      "step": 14323
+    },
+    {
+      "epoch": 0.8702837353423659,
+      "grad_norm": 0.19789010286331177,
+      "learning_rate": 4.1230345635804804e-06,
+      "loss": 1.1573,
+      "step": 14324
+    },
+    {
+      "epoch": 0.8703444923749925,
+      "grad_norm": 0.16213564574718475,
+      "learning_rate": 4.119228838232436e-06,
+      "loss": 1.1004,
+      "step": 14325
+    },
+    {
+      "epoch": 0.870405249407619,
+      "grad_norm": 0.22013337910175323,
+      "learning_rate": 4.115424794652917e-06,
+      "loss": 1.0789,
+      "step": 14326
+    },
+    {
+      "epoch": 0.8704660064402454,
+      "grad_norm": 0.11992393434047699,
+      "learning_rate": 4.111622432981393e-06,
+      "loss": 1.0526,
+      "step": 14327
+    },
+    {
+      "epoch": 0.870526763472872,
+      "grad_norm": 0.15205895900726318,
+      "learning_rate": 4.10782175335721e-06,
+      "loss": 1.1187,
+      "step": 14328
+    },
+    {
+      "epoch": 0.8705875205054985,
+      "grad_norm": 0.1337515264749527,
+      "learning_rate": 4.104022755919689e-06,
+      "loss": 0.983,
+      "step": 14329
+    },
+    {
+      "epoch": 0.870648277538125,
+      "grad_norm": 3.0589065551757812,
+      "learning_rate": 4.1002254408080865e-06,
+      "loss": 1.0804,
+      "step": 14330
+    },
+    {
+      "epoch": 0.8707090345707515,
+      "grad_norm": 0.0997358188033104,
+      "learning_rate": 4.096429808161578e-06,
+      "loss": 1.0412,
+      "step": 14331
+    },
+    {
+      "epoch": 0.8707697916033781,
+      "grad_norm": 0.1250249147415161,
+      "learning_rate": 4.092635858119326e-06,
+      "loss": 1.0676,
+      "step": 14332
+    },
+    {
+      "epoch": 0.8708305486360046,
+      "grad_norm": 0.23462393879890442,
+      "learning_rate": 4.088843590820374e-06,
+      "loss": 1.1005,
+      "step": 14333
+    },
+    {
+      "epoch": 0.8708913056686312,
+      "grad_norm": 0.2286270260810852,
+      "learning_rate": 4.085053006403744e-06,
+      "loss": 1.0357,
+      "step": 14334
+    },
+    {
+      "epoch": 0.8709520627012577,
+      "grad_norm": 0.19235652685165405,
+      "learning_rate": 4.081264105008364e-06,
+      "loss": 1.056,
+      "step": 14335
+    },
+    {
+      "epoch": 0.8710128197338842,
+      "grad_norm": 0.12831154465675354,
+      "learning_rate": 4.077476886773118e-06,
+      "loss": 1.0069,
+      "step": 14336
+    },
+    {
+      "epoch": 0.8710735767665108,
+      "grad_norm": 0.196796253323555,
+      "learning_rate": 4.073691351836839e-06,
+      "loss": 1.0868,
+      "step": 14337
+    },
+    {
+      "epoch": 0.8711343337991373,
+      "grad_norm": 0.37093472480773926,
+      "learning_rate": 4.069907500338277e-06,
+      "loss": 1.1283,
+      "step": 14338
+    },
+    {
+      "epoch": 0.8711950908317638,
+      "grad_norm": 0.20871317386627197,
+      "learning_rate": 4.066125332416138e-06,
+      "loss": 1.1847,
+      "step": 14339
+    },
+    {
+      "epoch": 0.8712558478643903,
+      "grad_norm": 1.4458787441253662,
+      "learning_rate": 4.062344848209055e-06,
+      "loss": 1.0755,
+      "step": 14340
+    },
+    {
+      "epoch": 0.8713166048970168,
+      "grad_norm": 0.11043219268321991,
+      "learning_rate": 4.0585660478556e-06,
+      "loss": 1.0243,
+      "step": 14341
+    },
+    {
+      "epoch": 0.8713773619296433,
+      "grad_norm": 0.11328072100877762,
+      "learning_rate": 4.0547889314942855e-06,
+      "loss": 0.9849,
+      "step": 14342
+    },
+    {
+      "epoch": 0.8714381189622699,
+      "grad_norm": 0.16323398053646088,
+      "learning_rate": 4.051013499263568e-06,
+      "loss": 1.1197,
+      "step": 14343
+    },
+    {
+      "epoch": 0.8714988759948964,
+      "grad_norm": 0.16723348200321198,
+      "learning_rate": 4.047239751301829e-06,
+      "loss": 1.0409,
+      "step": 14344
+    },
+    {
+      "epoch": 0.8715596330275229,
+      "grad_norm": 0.15389733016490936,
+      "learning_rate": 4.043467687747399e-06,
+      "loss": 1.1401,
+      "step": 14345
+    },
+    {
+      "epoch": 0.8716203900601495,
+      "grad_norm": 0.12652705609798431,
+      "learning_rate": 4.039697308738544e-06,
+      "loss": 1.0526,
+      "step": 14346
+    },
+    {
+      "epoch": 0.871681147092776,
+      "grad_norm": 1.9126840829849243,
+      "learning_rate": 4.03592861441347e-06,
+      "loss": 1.1255,
+      "step": 14347
+    },
+    {
+      "epoch": 0.8717419041254025,
+      "grad_norm": 0.1312454640865326,
+      "learning_rate": 4.0321616049103116e-06,
+      "loss": 1.0382,
+      "step": 14348
+    },
+    {
+      "epoch": 0.8718026611580291,
+      "grad_norm": 0.16049225628376007,
+      "learning_rate": 4.028396280367164e-06,
+      "loss": 1.1389,
+      "step": 14349
+    },
+    {
+      "epoch": 0.8718634181906556,
+      "grad_norm": 0.2220696061849594,
+      "learning_rate": 4.024632640922027e-06,
+      "loss": 1.0807,
+      "step": 14350
+    },
+    {
+      "epoch": 0.8719241752232821,
+      "grad_norm": 0.1569354385137558,
+      "learning_rate": 4.0208706867128744e-06,
+      "loss": 1.0815,
+      "step": 14351
+    },
+    {
+      "epoch": 0.8719849322559087,
+      "grad_norm": 0.12963464856147766,
+      "learning_rate": 4.017110417877585e-06,
+      "loss": 1.0417,
+      "step": 14352
+    },
+    {
+      "epoch": 0.8720456892885351,
+      "grad_norm": 0.10035611689090729,
+      "learning_rate": 4.013351834554008e-06,
+      "loss": 1.0455,
+      "step": 14353
+    },
+    {
+      "epoch": 0.8721064463211616,
+      "grad_norm": 0.11115585267543793,
+      "learning_rate": 4.009594936879918e-06,
+      "loss": 1.0292,
+      "step": 14354
+    },
+    {
+      "epoch": 0.8721672033537882,
+      "grad_norm": 0.10728547722101212,
+      "learning_rate": 4.00583972499301e-06,
+      "loss": 1.0573,
+      "step": 14355
+    },
+    {
+      "epoch": 0.8722279603864147,
+      "grad_norm": 5.861358165740967,
+      "learning_rate": 4.00208619903093e-06,
+      "loss": 1.02,
+      "step": 14356
+    },
+    {
+      "epoch": 0.8722887174190412,
+      "grad_norm": 0.1270265430212021,
+      "learning_rate": 3.998334359131267e-06,
+      "loss": 1.0581,
+      "step": 14357
+    },
+    {
+      "epoch": 0.8723494744516678,
+      "grad_norm": 0.08880908042192459,
+      "learning_rate": 3.994584205431562e-06,
+      "loss": 1.0108,
+      "step": 14358
+    },
+    {
+      "epoch": 0.8724102314842943,
+      "grad_norm": 0.4161021113395691,
+      "learning_rate": 3.99083573806926e-06,
+      "loss": 1.2731,
+      "step": 14359
+    },
+    {
+      "epoch": 0.8724709885169208,
+      "grad_norm": 0.1339392364025116,
+      "learning_rate": 3.987088957181772e-06,
+      "loss": 1.0502,
+      "step": 14360
+    },
+    {
+      "epoch": 0.8725317455495474,
+      "grad_norm": 0.6967560052871704,
+      "learning_rate": 3.983343862906436e-06,
+      "loss": 1.0221,
+      "step": 14361
+    },
+    {
+      "epoch": 0.8725925025821739,
+      "grad_norm": 0.15927958488464355,
+      "learning_rate": 3.979600455380511e-06,
+      "loss": 1.0628,
+      "step": 14362
+    },
+    {
+      "epoch": 0.8726532596148004,
+      "grad_norm": 0.11636935919523239,
+      "learning_rate": 3.975858734741239e-06,
+      "loss": 1.035,
+      "step": 14363
+    },
+    {
+      "epoch": 0.872714016647427,
+      "grad_norm": 0.25674161314964294,
+      "learning_rate": 3.972118701125754e-06,
+      "loss": 1.1213,
+      "step": 14364
+    },
+    {
+      "epoch": 0.8727747736800535,
+      "grad_norm": 0.14380165934562683,
+      "learning_rate": 3.968380354671164e-06,
+      "loss": 1.0911,
+      "step": 14365
+    },
+    {
+      "epoch": 0.8728355307126799,
+      "grad_norm": 0.09789226949214935,
+      "learning_rate": 3.964643695514481e-06,
+      "loss": 1.0164,
+      "step": 14366
+    },
+    {
+      "epoch": 0.8728962877453065,
+      "grad_norm": 0.10349620878696442,
+      "learning_rate": 3.960908723792689e-06,
+      "loss": 0.9742,
+      "step": 14367
+    },
+    {
+      "epoch": 0.872957044777933,
+      "grad_norm": 0.11412148177623749,
+      "learning_rate": 3.95717543964268e-06,
+      "loss": 1.0696,
+      "step": 14368
+    },
+    {
+      "epoch": 0.8730178018105595,
+      "grad_norm": 2.68796443939209,
+      "learning_rate": 3.95344384320131e-06,
+      "loss": 0.991,
+      "step": 14369
+    },
+    {
+      "epoch": 0.8730785588431861,
+      "grad_norm": 0.16747133433818817,
+      "learning_rate": 3.949713934605354e-06,
+      "loss": 1.0614,
+      "step": 14370
+    },
+    {
+      "epoch": 0.8731393158758126,
+      "grad_norm": 0.09575141966342926,
+      "learning_rate": 3.945985713991534e-06,
+      "loss": 1.0033,
+      "step": 14371
+    },
+    {
+      "epoch": 0.8732000729084392,
+      "grad_norm": 0.2550472021102905,
+      "learning_rate": 3.942259181496516e-06,
+      "loss": 1.1353,
+      "step": 14372
+    },
+    {
+      "epoch": 0.8732608299410657,
+      "grad_norm": 0.10565902292728424,
+      "learning_rate": 3.938534337256877e-06,
+      "loss": 1.0307,
+      "step": 14373
+    },
+    {
+      "epoch": 0.8733215869736922,
+      "grad_norm": 0.3183503746986389,
+      "learning_rate": 3.934811181409181e-06,
+      "loss": 1.1327,
+      "step": 14374
+    },
+    {
+      "epoch": 0.8733823440063188,
+      "grad_norm": 0.19271215796470642,
+      "learning_rate": 3.9310897140898805e-06,
+      "loss": 1.1201,
+      "step": 14375
+    },
+    {
+      "epoch": 0.8734431010389453,
+      "grad_norm": 0.24655552208423615,
+      "learning_rate": 3.927369935435387e-06,
+      "loss": 1.1384,
+      "step": 14376
+    },
+    {
+      "epoch": 0.8735038580715718,
+      "grad_norm": 0.16251324117183685,
+      "learning_rate": 3.923651845582061e-06,
+      "loss": 1.1044,
+      "step": 14377
+    },
+    {
+      "epoch": 0.8735646151041984,
+      "grad_norm": 0.09870925545692444,
+      "learning_rate": 3.9199354446661675e-06,
+      "loss": 1.0222,
+      "step": 14378
+    },
+    {
+      "epoch": 0.8736253721368248,
+      "grad_norm": 0.17102311551570892,
+      "learning_rate": 3.916220732823961e-06,
+      "loss": 1.078,
+      "step": 14379
+    },
+    {
+      "epoch": 0.8736861291694513,
+      "grad_norm": 0.111696757376194,
+      "learning_rate": 3.9125077101915875e-06,
+      "loss": 1.0412,
+      "step": 14380
+    },
+    {
+      "epoch": 0.8737468862020779,
+      "grad_norm": 0.1547119915485382,
+      "learning_rate": 3.908796376905166e-06,
+      "loss": 1.047,
+      "step": 14381
+    },
+    {
+      "epoch": 0.8738076432347044,
+      "grad_norm": 0.3426506519317627,
+      "learning_rate": 3.90508673310071e-06,
+      "loss": 1.1596,
+      "step": 14382
+    },
+    {
+      "epoch": 0.8738684002673309,
+      "grad_norm": 0.09747691452503204,
+      "learning_rate": 3.9013787789142e-06,
+      "loss": 1.0123,
+      "step": 14383
+    },
+    {
+      "epoch": 0.8739291572999575,
+      "grad_norm": 0.09974610805511475,
+      "learning_rate": 3.89767251448157e-06,
+      "loss": 1.0447,
+      "step": 14384
+    },
+    {
+      "epoch": 0.873989914332584,
+      "grad_norm": 0.10408081859350204,
+      "learning_rate": 3.893967939938669e-06,
+      "loss": 1.0935,
+      "step": 14385
+    },
+    {
+      "epoch": 0.8740506713652105,
+      "grad_norm": 0.16880090534687042,
+      "learning_rate": 3.890265055421283e-06,
+      "loss": 1.0666,
+      "step": 14386
+    },
+    {
+      "epoch": 0.8741114283978371,
+      "grad_norm": 0.19213543832302094,
+      "learning_rate": 3.886563861065146e-06,
+      "loss": 1.1783,
+      "step": 14387
+    },
+    {
+      "epoch": 0.8741721854304636,
+      "grad_norm": 0.172330841422081,
+      "learning_rate": 3.882864357005922e-06,
+      "loss": 1.0489,
+      "step": 14388
+    },
+    {
+      "epoch": 0.8742329424630901,
+      "grad_norm": 0.11979315429925919,
+      "learning_rate": 3.87916654337922e-06,
+      "loss": 1.0029,
+      "step": 14389
+    },
+    {
+      "epoch": 0.8742936994957167,
+      "grad_norm": 0.18948282301425934,
+      "learning_rate": 3.875470420320582e-06,
+      "loss": 1.1096,
+      "step": 14390
+    },
+    {
+      "epoch": 0.8743544565283432,
+      "grad_norm": 0.3669520914554596,
+      "learning_rate": 3.871775987965498e-06,
+      "loss": 1.1074,
+      "step": 14391
+    },
+    {
+      "epoch": 0.8744152135609696,
+      "grad_norm": 0.14308518171310425,
+      "learning_rate": 3.868083246449372e-06,
+      "loss": 1.1311,
+      "step": 14392
+    },
+    {
+      "epoch": 0.8744759705935962,
+      "grad_norm": 0.16750699281692505,
+      "learning_rate": 3.864392195907579e-06,
+      "loss": 1.0572,
+      "step": 14393
+    },
+    {
+      "epoch": 0.8745367276262227,
+      "grad_norm": 0.14526700973510742,
+      "learning_rate": 3.860702836475405e-06,
+      "loss": 1.096,
+      "step": 14394
+    },
+    {
+      "epoch": 0.8745974846588492,
+      "grad_norm": 0.26115116477012634,
+      "learning_rate": 3.8570151682880865e-06,
+      "loss": 1.0689,
+      "step": 14395
+    },
+    {
+      "epoch": 0.8746582416914758,
+      "grad_norm": 0.17449040710926056,
+      "learning_rate": 3.853329191480798e-06,
+      "loss": 1.1736,
+      "step": 14396
+    },
+    {
+      "epoch": 0.8747189987241023,
+      "grad_norm": 0.10472922772169113,
+      "learning_rate": 3.849644906188643e-06,
+      "loss": 0.9713,
+      "step": 14397
+    },
+    {
+      "epoch": 0.8747797557567288,
+      "grad_norm": 0.455191969871521,
+      "learning_rate": 3.845962312546681e-06,
+      "loss": 1.2272,
+      "step": 14398
+    },
+    {
+      "epoch": 0.8748405127893554,
+      "grad_norm": 0.16802021861076355,
+      "learning_rate": 3.842281410689879e-06,
+      "loss": 1.0846,
+      "step": 14399
+    },
+    {
+      "epoch": 0.8749012698219819,
+      "grad_norm": 0.1144322007894516,
+      "learning_rate": 3.838602200753183e-06,
+      "loss": 1.0231,
+      "step": 14400
+    },
+    {
+      "epoch": 0.8749620268546084,
+      "grad_norm": 0.1819830983877182,
+      "learning_rate": 3.83492468287146e-06,
+      "loss": 1.0766,
+      "step": 14401
+    },
+    {
+      "epoch": 0.875022783887235,
+      "grad_norm": 0.1960277408361435,
+      "learning_rate": 3.8312488571794816e-06,
+      "loss": 1.0489,
+      "step": 14402
+    },
+    {
+      "epoch": 0.8750835409198615,
+      "grad_norm": 0.2229665070772171,
+      "learning_rate": 3.827574723811994e-06,
+      "loss": 1.2662,
+      "step": 14403
+    },
+    {
+      "epoch": 0.875144297952488,
+      "grad_norm": 0.16860918700695038,
+      "learning_rate": 3.8239022829036905e-06,
+      "loss": 1.1115,
+      "step": 14404
+    },
+    {
+      "epoch": 0.8752050549851145,
+      "grad_norm": 0.13054722547531128,
+      "learning_rate": 3.8202315345891696e-06,
+      "loss": 1.0217,
+      "step": 14405
+    },
+    {
+      "epoch": 0.875265812017741,
+      "grad_norm": 0.14450567960739136,
+      "learning_rate": 3.816562479002994e-06,
+      "loss": 1.0582,
+      "step": 14406
+    },
+    {
+      "epoch": 0.8753265690503675,
+      "grad_norm": 0.10299817472696304,
+      "learning_rate": 3.812895116279641e-06,
+      "loss": 1.0167,
+      "step": 14407
+    },
+    {
+      "epoch": 0.8753873260829941,
+      "grad_norm": 0.14837372303009033,
+      "learning_rate": 3.8092294465535527e-06,
+      "loss": 1.0695,
+      "step": 14408
+    },
+    {
+      "epoch": 0.8754480831156206,
+      "grad_norm": 0.11886508762836456,
+      "learning_rate": 3.8055654699590827e-06,
+      "loss": 1.049,
+      "step": 14409
+    },
+    {
+      "epoch": 0.8755088401482471,
+      "grad_norm": 0.15067318081855774,
+      "learning_rate": 3.80190318663054e-06,
+      "loss": 1.1297,
+      "step": 14410
+    },
+    {
+      "epoch": 0.8755695971808737,
+      "grad_norm": 0.164082869887352,
+      "learning_rate": 3.7982425967021673e-06,
+      "loss": 1.112,
+      "step": 14411
+    },
+    {
+      "epoch": 0.8756303542135002,
+      "grad_norm": 0.2626175880432129,
+      "learning_rate": 3.7945837003081406e-06,
+      "loss": 1.0367,
+      "step": 14412
+    },
+    {
+      "epoch": 0.8756911112461268,
+      "grad_norm": 0.16389034688472748,
+      "learning_rate": 3.790926497582581e-06,
+      "loss": 1.135,
+      "step": 14413
+    },
+    {
+      "epoch": 0.8757518682787533,
+      "grad_norm": 0.7456648349761963,
+      "learning_rate": 3.7872709886595413e-06,
+      "loss": 1.0796,
+      "step": 14414
+    },
+    {
+      "epoch": 0.8758126253113798,
+      "grad_norm": 0.12602953612804413,
+      "learning_rate": 3.7836171736730206e-06,
+      "loss": 1.0384,
+      "step": 14415
+    },
+    {
+      "epoch": 0.8758733823440064,
+      "grad_norm": 0.11357144266366959,
+      "learning_rate": 3.7799650527569396e-06,
+      "loss": 1.0601,
+      "step": 14416
+    },
+    {
+      "epoch": 0.8759341393766329,
+      "grad_norm": 0.10738505423069,
+      "learning_rate": 3.77631462604518e-06,
+      "loss": 1.0487,
+      "step": 14417
+    },
+    {
+      "epoch": 0.8759948964092594,
+      "grad_norm": 2.844250440597534,
+      "learning_rate": 3.7726658936715397e-06,
+      "loss": 1.0078,
+      "step": 14418
+    },
+    {
+      "epoch": 0.8760556534418859,
+      "grad_norm": 0.12740908563137054,
+      "learning_rate": 3.769018855769757e-06,
+      "loss": 1.0331,
+      "step": 14419
+    },
+    {
+      "epoch": 0.8761164104745124,
+      "grad_norm": 0.10843446105718613,
+      "learning_rate": 3.7653735124735355e-06,
+      "loss": 1.087,
+      "step": 14420
+    },
+    {
+      "epoch": 0.8761771675071389,
+      "grad_norm": 0.11844402551651001,
+      "learning_rate": 3.7617298639164913e-06,
+      "loss": 1.0338,
+      "step": 14421
+    },
+    {
+      "epoch": 0.8762379245397655,
+      "grad_norm": 0.6083289980888367,
+      "learning_rate": 3.7580879102321663e-06,
+      "loss": 1.0459,
+      "step": 14422
+    },
+    {
+      "epoch": 0.876298681572392,
+      "grad_norm": 0.16263823211193085,
+      "learning_rate": 3.7544476515540715e-06,
+      "loss": 1.158,
+      "step": 14423
+    },
+    {
+      "epoch": 0.8763594386050185,
+      "grad_norm": 0.15546706318855286,
+      "learning_rate": 3.750809088015628e-06,
+      "loss": 1.0721,
+      "step": 14424
+    },
+    {
+      "epoch": 0.8764201956376451,
+      "grad_norm": 22.740657806396484,
+      "learning_rate": 3.747172219750222e-06,
+      "loss": 1.0248,
+      "step": 14425
+    },
+    {
+      "epoch": 0.8764809526702716,
+      "grad_norm": 0.298325777053833,
+      "learning_rate": 3.74353704689116e-06,
+      "loss": 1.0601,
+      "step": 14426
+    },
+    {
+      "epoch": 0.8765417097028981,
+      "grad_norm": 2.110259771347046,
+      "learning_rate": 3.73990356957169e-06,
+      "loss": 1.0454,
+      "step": 14427
+    },
+    {
+      "epoch": 0.8766024667355247,
+      "grad_norm": 0.1646614968776703,
+      "learning_rate": 3.7362717879250053e-06,
+      "loss": 1.1097,
+      "step": 14428
+    },
+    {
+      "epoch": 0.8766632237681512,
+      "grad_norm": 0.278971403837204,
+      "learning_rate": 3.732641702084205e-06,
+      "loss": 1.1391,
+      "step": 14429
+    },
+    {
+      "epoch": 0.8767239808007777,
+      "grad_norm": 0.1312994360923767,
+      "learning_rate": 3.7290133121823723e-06,
+      "loss": 1.011,
+      "step": 14430
+    },
+    {
+      "epoch": 0.8767847378334043,
+      "grad_norm": 0.2943628132343292,
+      "learning_rate": 3.7253866183525e-06,
+      "loss": 1.0538,
+      "step": 14431
+    },
+    {
+      "epoch": 0.8768454948660307,
+      "grad_norm": 0.15289996564388275,
+      "learning_rate": 3.721761620727532e-06,
+      "loss": 1.0203,
+      "step": 14432
+    },
+    {
+      "epoch": 0.8769062518986572,
+      "grad_norm": 0.18115881085395813,
+      "learning_rate": 3.7181383194403286e-06,
+      "loss": 1.0979,
+      "step": 14433
+    },
+    {
+      "epoch": 0.8769670089312838,
+      "grad_norm": 0.2012121081352234,
+      "learning_rate": 3.714516714623717e-06,
+      "loss": 1.1157,
+      "step": 14434
+    },
+    {
+      "epoch": 0.8770277659639103,
+      "grad_norm": 0.1832466423511505,
+      "learning_rate": 3.7108968064104412e-06,
+      "loss": 1.0592,
+      "step": 14435
+    },
+    {
+      "epoch": 0.8770885229965368,
+      "grad_norm": 0.13826370239257812,
+      "learning_rate": 3.707278594933189e-06,
+      "loss": 1.0607,
+      "step": 14436
+    },
+    {
+      "epoch": 0.8771492800291634,
+      "grad_norm": 0.190177783370018,
+      "learning_rate": 3.703662080324588e-06,
+      "loss": 1.1455,
+      "step": 14437
+    },
+    {
+      "epoch": 0.8772100370617899,
+      "grad_norm": 0.14002488553524017,
+      "learning_rate": 3.7000472627172047e-06,
+      "loss": 1.0244,
+      "step": 14438
+    },
+    {
+      "epoch": 0.8772707940944164,
+      "grad_norm": 0.17376451194286346,
+      "learning_rate": 3.696434142243538e-06,
+      "loss": 1.1515,
+      "step": 14439
+    },
+    {
+      "epoch": 0.877331551127043,
+      "grad_norm": 0.15191587805747986,
+      "learning_rate": 3.6928227190360154e-06,
+      "loss": 1.0563,
+      "step": 14440
+    },
+    {
+      "epoch": 0.8773923081596695,
+      "grad_norm": 0.5717953443527222,
+      "learning_rate": 3.6892129932270482e-06,
+      "loss": 1.062,
+      "step": 14441
+    },
+    {
+      "epoch": 0.877453065192296,
+      "grad_norm": 0.09690621495246887,
+      "learning_rate": 3.685604964948919e-06,
+      "loss": 1.0183,
+      "step": 14442
+    },
+    {
+      "epoch": 0.8775138222249226,
+      "grad_norm": 0.1264795958995819,
+      "learning_rate": 3.681998634333894e-06,
+      "loss": 1.0749,
+      "step": 14443
+    },
+    {
+      "epoch": 0.8775745792575491,
+      "grad_norm": 0.12863373756408691,
+      "learning_rate": 3.6783940015141625e-06,
+      "loss": 1.0854,
+      "step": 14444
+    },
+    {
+      "epoch": 0.8776353362901755,
+      "grad_norm": 1.2896924018859863,
+      "learning_rate": 3.6747910666218466e-06,
+      "loss": 1.06,
+      "step": 14445
+    },
+    {
+      "epoch": 0.8776960933228021,
+      "grad_norm": 0.12301401793956757,
+      "learning_rate": 3.6711898297890234e-06,
+      "loss": 1.0481,
+      "step": 14446
+    },
+    {
+      "epoch": 0.8777568503554286,
+      "grad_norm": 0.15882372856140137,
+      "learning_rate": 3.667590291147693e-06,
+      "loss": 1.0705,
+      "step": 14447
+    },
+    {
+      "epoch": 0.8778176073880551,
+      "grad_norm": 0.12979115545749664,
+      "learning_rate": 3.663992450829801e-06,
+      "loss": 1.0566,
+      "step": 14448
+    },
+    {
+      "epoch": 0.8778783644206817,
+      "grad_norm": 9.955571174621582,
+      "learning_rate": 3.660396308967218e-06,
+      "loss": 0.9969,
+      "step": 14449
+    },
+    {
+      "epoch": 0.8779391214533082,
+      "grad_norm": 0.11079701036214828,
+      "learning_rate": 3.6568018656917614e-06,
+      "loss": 1.0505,
+      "step": 14450
+    },
+    {
+      "epoch": 0.8779998784859347,
+      "grad_norm": 0.1300920695066452,
+      "learning_rate": 3.6532091211351928e-06,
+      "loss": 0.9929,
+      "step": 14451
+    },
+    {
+      "epoch": 0.8780606355185613,
+      "grad_norm": 0.18141959607601166,
+      "learning_rate": 3.6496180754292007e-06,
+      "loss": 1.1408,
+      "step": 14452
+    },
+    {
+      "epoch": 0.8781213925511878,
+      "grad_norm": 1.4117289781570435,
+      "learning_rate": 3.6460287287054196e-06,
+      "loss": 1.0695,
+      "step": 14453
+    },
+    {
+      "epoch": 0.8781821495838144,
+      "grad_norm": 0.1022375300526619,
+      "learning_rate": 3.642441081095421e-06,
+      "loss": 1.0178,
+      "step": 14454
+    },
+    {
+      "epoch": 0.8782429066164409,
+      "grad_norm": 0.17482461035251617,
+      "learning_rate": 3.638855132730701e-06,
+      "loss": 1.0692,
+      "step": 14455
+    },
+    {
+      "epoch": 0.8783036636490674,
+      "grad_norm": 0.1493391990661621,
+      "learning_rate": 3.635270883742703e-06,
+      "loss": 1.0183,
+      "step": 14456
+    },
+    {
+      "epoch": 0.878364420681694,
+      "grad_norm": 0.3150024712085724,
+      "learning_rate": 3.631688334262817e-06,
+      "loss": 1.0442,
+      "step": 14457
+    },
+    {
+      "epoch": 0.8784251777143204,
+      "grad_norm": 0.1289951056241989,
+      "learning_rate": 3.6281074844223608e-06,
+      "loss": 1.0338,
+      "step": 14458
+    },
+    {
+      "epoch": 0.8784859347469469,
+      "grad_norm": 0.40785345435142517,
+      "learning_rate": 3.6245283343525837e-06,
+      "loss": 1.1256,
+      "step": 14459
+    },
+    {
+      "epoch": 0.8785466917795735,
+      "grad_norm": 0.1404150277376175,
+      "learning_rate": 3.6209508841846873e-06,
+      "loss": 1.0624,
+      "step": 14460
+    },
+    {
+      "epoch": 0.8786074488122,
+      "grad_norm": 0.12771236896514893,
+      "learning_rate": 3.6173751340497997e-06,
+      "loss": 1.045,
+      "step": 14461
+    },
+    {
+      "epoch": 0.8786682058448265,
+      "grad_norm": 0.140315443277359,
+      "learning_rate": 3.6138010840789936e-06,
+      "loss": 0.9968,
+      "step": 14462
+    },
+    {
+      "epoch": 0.8787289628774531,
+      "grad_norm": 0.18975098431110382,
+      "learning_rate": 3.6102287344032703e-06,
+      "loss": 1.102,
+      "step": 14463
+    },
+    {
+      "epoch": 0.8787897199100796,
+      "grad_norm": 0.1953096240758896,
+      "learning_rate": 3.6066580851535857e-06,
+      "loss": 1.1883,
+      "step": 14464
+    },
+    {
+      "epoch": 0.8788504769427061,
+      "grad_norm": 0.1363508254289627,
+      "learning_rate": 3.6030891364608133e-06,
+      "loss": 1.0215,
+      "step": 14465
+    },
+    {
+      "epoch": 0.8789112339753327,
+      "grad_norm": 0.14448469877243042,
+      "learning_rate": 3.59952188845577e-06,
+      "loss": 1.0609,
+      "step": 14466
+    },
+    {
+      "epoch": 0.8789719910079592,
+      "grad_norm": 0.5873912572860718,
+      "learning_rate": 3.5959563412692297e-06,
+      "loss": 1.2934,
+      "step": 14467
+    },
+    {
+      "epoch": 0.8790327480405857,
+      "grad_norm": 0.2090008556842804,
+      "learning_rate": 3.5923924950318875e-06,
+      "loss": 1.1541,
+      "step": 14468
+    },
+    {
+      "epoch": 0.8790935050732123,
+      "grad_norm": 0.13231422007083893,
+      "learning_rate": 3.588830349874356e-06,
+      "loss": 1.0369,
+      "step": 14469
+    },
+    {
+      "epoch": 0.8791542621058388,
+      "grad_norm": 0.15206478536128998,
+      "learning_rate": 3.5852699059272243e-06,
+      "loss": 1.0779,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8792150191384652,
+      "grad_norm": 0.1617000550031662,
+      "learning_rate": 3.581711163320989e-06,
+      "loss": 1.0624,
+      "step": 14471
+    },
+    {
+      "epoch": 0.8792757761710918,
+      "grad_norm": 0.12221449613571167,
+      "learning_rate": 3.578154122186106e-06,
+      "loss": 1.0475,
+      "step": 14472
+    },
+    {
+      "epoch": 0.8793365332037183,
+      "grad_norm": 0.11573433130979538,
+      "learning_rate": 3.5745987826529604e-06,
+      "loss": 1.0414,
+      "step": 14473
+    },
+    {
+      "epoch": 0.8793972902363448,
+      "grad_norm": 0.2252984195947647,
+      "learning_rate": 3.57104514485187e-06,
+      "loss": 1.1763,
+      "step": 14474
+    },
+    {
+      "epoch": 0.8794580472689714,
+      "grad_norm": 0.1389123648405075,
+      "learning_rate": 3.567493208913103e-06,
+      "loss": 1.0547,
+      "step": 14475
+    },
+    {
+      "epoch": 0.8795188043015979,
+      "grad_norm": 0.09723228961229324,
+      "learning_rate": 3.5639429749668272e-06,
+      "loss": 0.9914,
+      "step": 14476
+    },
+    {
+      "epoch": 0.8795795613342244,
+      "grad_norm": 0.16194170713424683,
+      "learning_rate": 3.5603944431432114e-06,
+      "loss": 1.0777,
+      "step": 14477
+    },
+    {
+      "epoch": 0.879640318366851,
+      "grad_norm": 0.12059863656759262,
+      "learning_rate": 3.5568476135723116e-06,
+      "loss": 1.0466,
+      "step": 14478
+    },
+    {
+      "epoch": 0.8797010753994775,
+      "grad_norm": 0.1424207240343094,
+      "learning_rate": 3.553302486384141e-06,
+      "loss": 1.055,
+      "step": 14479
+    },
+    {
+      "epoch": 0.879761832432104,
+      "grad_norm": 0.13578937947750092,
+      "learning_rate": 3.549759061708641e-06,
+      "loss": 1.1214,
+      "step": 14480
+    },
+    {
+      "epoch": 0.8798225894647306,
+      "grad_norm": 0.13324271142482758,
+      "learning_rate": 3.546217339675706e-06,
+      "loss": 1.1089,
+      "step": 14481
+    },
+    {
+      "epoch": 0.8798833464973571,
+      "grad_norm": 0.10492102056741714,
+      "learning_rate": 3.5426773204151498e-06,
+      "loss": 1.0897,
+      "step": 14482
+    },
+    {
+      "epoch": 0.8799441035299836,
+      "grad_norm": 0.3962917625904083,
+      "learning_rate": 3.5391390040567408e-06,
+      "loss": 1.2425,
+      "step": 14483
+    },
+    {
+      "epoch": 0.8800048605626101,
+      "grad_norm": 0.10516023635864258,
+      "learning_rate": 3.5356023907301696e-06,
+      "loss": 1.0185,
+      "step": 14484
+    },
+    {
+      "epoch": 0.8800656175952366,
+      "grad_norm": 0.14894473552703857,
+      "learning_rate": 3.5320674805650767e-06,
+      "loss": 1.0935,
+      "step": 14485
+    },
+    {
+      "epoch": 0.8801263746278631,
+      "grad_norm": 0.16137200593948364,
+      "learning_rate": 3.528534273691031e-06,
+      "loss": 1.0546,
+      "step": 14486
+    },
+    {
+      "epoch": 0.8801871316604897,
+      "grad_norm": 0.10996224731206894,
+      "learning_rate": 3.5250027702375345e-06,
+      "loss": 1.0507,
+      "step": 14487
+    },
+    {
+      "epoch": 0.8802478886931162,
+      "grad_norm": 0.15636777877807617,
+      "learning_rate": 3.521472970334061e-06,
+      "loss": 1.1398,
+      "step": 14488
+    },
+    {
+      "epoch": 0.8803086457257427,
+      "grad_norm": 0.22474385797977448,
+      "learning_rate": 3.5179448741099743e-06,
+      "loss": 1.1368,
+      "step": 14489
+    },
+    {
+      "epoch": 0.8803694027583693,
+      "grad_norm": 0.1425997018814087,
+      "learning_rate": 3.5144184816946036e-06,
+      "loss": 0.9917,
+      "step": 14490
+    },
+    {
+      "epoch": 0.8804301597909958,
+      "grad_norm": 0.236920565366745,
+      "learning_rate": 3.510893793217207e-06,
+      "loss": 1.1124,
+      "step": 14491
+    },
+    {
+      "epoch": 0.8804909168236223,
+      "grad_norm": 0.11593442410230637,
+      "learning_rate": 3.5073708088069756e-06,
+      "loss": 1.0283,
+      "step": 14492
+    },
+    {
+      "epoch": 0.8805516738562489,
+      "grad_norm": 0.19685669243335724,
+      "learning_rate": 3.503849528593062e-06,
+      "loss": 1.0703,
+      "step": 14493
+    },
+    {
+      "epoch": 0.8806124308888754,
+      "grad_norm": 0.15274019539356232,
+      "learning_rate": 3.500329952704534e-06,
+      "loss": 1.0925,
+      "step": 14494
+    },
+    {
+      "epoch": 0.880673187921502,
+      "grad_norm": 0.1302100419998169,
+      "learning_rate": 3.496812081270406e-06,
+      "loss": 1.065,
+      "step": 14495
+    },
+    {
+      "epoch": 0.8807339449541285,
+      "grad_norm": 0.14908307790756226,
+      "learning_rate": 3.493295914419603e-06,
+      "loss": 1.0319,
+      "step": 14496
+    },
+    {
+      "epoch": 0.8807947019867549,
+      "grad_norm": 0.1167093962430954,
+      "learning_rate": 3.489781452281038e-06,
+      "loss": 1.0376,
+      "step": 14497
+    },
+    {
+      "epoch": 0.8808554590193814,
+      "grad_norm": 0.24106687307357788,
+      "learning_rate": 3.4862686949835244e-06,
+      "loss": 1.0563,
+      "step": 14498
+    },
+    {
+      "epoch": 0.880916216052008,
+      "grad_norm": 0.11067598313093185,
+      "learning_rate": 3.4827576426558206e-06,
+      "loss": 1.0273,
+      "step": 14499
+    },
+    {
+      "epoch": 0.8809769730846345,
+      "grad_norm": 0.23668546974658966,
+      "learning_rate": 3.479248295426624e-06,
+      "loss": 1.1028,
+      "step": 14500
+    },
+    {
+      "epoch": 0.881037730117261,
+      "grad_norm": 0.3210560977458954,
+      "learning_rate": 3.4757406534245808e-06,
+      "loss": 1.0954,
+      "step": 14501
+    },
+    {
+      "epoch": 0.8810984871498876,
+      "grad_norm": 0.39827099442481995,
+      "learning_rate": 3.4722347167782498e-06,
+      "loss": 1.0256,
+      "step": 14502
+    },
+    {
+      "epoch": 0.8811592441825141,
+      "grad_norm": 1.8480387926101685,
+      "learning_rate": 3.468730485616156e-06,
+      "loss": 1.1882,
+      "step": 14503
+    },
+    {
+      "epoch": 0.8812200012151407,
+      "grad_norm": 0.14183840155601501,
+      "learning_rate": 3.4652279600667357e-06,
+      "loss": 1.0937,
+      "step": 14504
+    },
+    {
+      "epoch": 0.8812807582477672,
+      "grad_norm": 0.2680705487728119,
+      "learning_rate": 3.4617271402583806e-06,
+      "loss": 1.1024,
+      "step": 14505
+    },
+    {
+      "epoch": 0.8813415152803937,
+      "grad_norm": 0.12620899081230164,
+      "learning_rate": 3.4582280263194157e-06,
+      "loss": 1.0671,
+      "step": 14506
+    },
+    {
+      "epoch": 0.8814022723130203,
+      "grad_norm": 0.14713971316814423,
+      "learning_rate": 3.4547306183780825e-06,
+      "loss": 1.0373,
+      "step": 14507
+    },
+    {
+      "epoch": 0.8814630293456468,
+      "grad_norm": 0.12060145288705826,
+      "learning_rate": 3.451234916562618e-06,
+      "loss": 1.0737,
+      "step": 14508
+    },
+    {
+      "epoch": 0.8815237863782733,
+      "grad_norm": 0.11234474182128906,
+      "learning_rate": 3.44774092100113e-06,
+      "loss": 1.0373,
+      "step": 14509
+    },
+    {
+      "epoch": 0.8815845434108998,
+      "grad_norm": 0.18484778702259064,
+      "learning_rate": 3.444248631821689e-06,
+      "loss": 1.1451,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8816453004435263,
+      "grad_norm": 0.1472494900226593,
+      "learning_rate": 3.4407580491523206e-06,
+      "loss": 1.0421,
+      "step": 14511
+    },
+    {
+      "epoch": 0.8817060574761528,
+      "grad_norm": 0.10259804129600525,
+      "learning_rate": 3.437269173120955e-06,
+      "loss": 0.9976,
+      "step": 14512
+    },
+    {
+      "epoch": 0.8817668145087794,
+      "grad_norm": 0.14788947999477386,
+      "learning_rate": 3.4337820038554958e-06,
+      "loss": 1.0323,
+      "step": 14513
+    },
+    {
+      "epoch": 0.8818275715414059,
+      "grad_norm": 0.14503687620162964,
+      "learning_rate": 3.4302965414837575e-06,
+      "loss": 1.0969,
+      "step": 14514
+    },
+    {
+      "epoch": 0.8818883285740324,
+      "grad_norm": 0.10066328942775726,
+      "learning_rate": 3.4268127861335097e-06,
+      "loss": 1.017,
+      "step": 14515
+    },
+    {
+      "epoch": 0.881949085606659,
+      "grad_norm": 0.15164169669151306,
+      "learning_rate": 3.423330737932434e-06,
+      "loss": 1.1951,
+      "step": 14516
+    },
+    {
+      "epoch": 0.8820098426392855,
+      "grad_norm": 1.1713110208511353,
+      "learning_rate": 3.4198503970081675e-06,
+      "loss": 1.0925,
+      "step": 14517
+    },
+    {
+      "epoch": 0.882070599671912,
+      "grad_norm": 0.1136820986866951,
+      "learning_rate": 3.416371763488291e-06,
+      "loss": 1.0381,
+      "step": 14518
+    },
+    {
+      "epoch": 0.8821313567045386,
+      "grad_norm": 0.15022777020931244,
+      "learning_rate": 3.412894837500319e-06,
+      "loss": 1.0667,
+      "step": 14519
+    },
+    {
+      "epoch": 0.8821921137371651,
+      "grad_norm": 0.3239344358444214,
+      "learning_rate": 3.409419619171683e-06,
+      "loss": 1.073,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8822528707697916,
+      "grad_norm": 0.11347474157810211,
+      "learning_rate": 3.4059461086297815e-06,
+      "loss": 1.0118,
+      "step": 14521
+    },
+    {
+      "epoch": 0.8823136278024182,
+      "grad_norm": 0.14050550758838654,
+      "learning_rate": 3.402474306001935e-06,
+      "loss": 1.0807,
+      "step": 14522
+    },
+    {
+      "epoch": 0.8823743848350447,
+      "grad_norm": 0.12560464441776276,
+      "learning_rate": 3.3990042114153965e-06,
+      "loss": 1.0569,
+      "step": 14523
+    },
+    {
+      "epoch": 0.8824351418676711,
+      "grad_norm": 0.09681057929992676,
+      "learning_rate": 3.3955358249973646e-06,
+      "loss": 1.0486,
+      "step": 14524
+    },
+    {
+      "epoch": 0.8824958989002977,
+      "grad_norm": 0.48717978596687317,
+      "learning_rate": 3.3920691468749768e-06,
+      "loss": 1.0973,
+      "step": 14525
+    },
+    {
+      "epoch": 0.8825566559329242,
+      "grad_norm": 0.11208033561706543,
+      "learning_rate": 3.3886041771753086e-06,
+      "loss": 1.0222,
+      "step": 14526
+    },
+    {
+      "epoch": 0.8826174129655507,
+      "grad_norm": 0.1361130326986313,
+      "learning_rate": 3.385140916025359e-06,
+      "loss": 1.1048,
+      "step": 14527
+    },
+    {
+      "epoch": 0.8826781699981773,
+      "grad_norm": 0.10553893446922302,
+      "learning_rate": 3.3816793635520816e-06,
+      "loss": 1.0176,
+      "step": 14528
+    },
+    {
+      "epoch": 0.8827389270308038,
+      "grad_norm": 0.1062655821442604,
+      "learning_rate": 3.378219519882353e-06,
+      "loss": 1.0626,
+      "step": 14529
+    },
+    {
+      "epoch": 0.8827996840634303,
+      "grad_norm": 0.522912323474884,
+      "learning_rate": 3.374761385143005e-06,
+      "loss": 1.18,
+      "step": 14530
+    },
+    {
+      "epoch": 0.8828604410960569,
+      "grad_norm": 0.148276224732399,
+      "learning_rate": 3.3713049594607915e-06,
+      "loss": 1.0676,
+      "step": 14531
+    },
+    {
+      "epoch": 0.8829211981286834,
+      "grad_norm": 0.23223872482776642,
+      "learning_rate": 3.3678502429624057e-06,
+      "loss": 1.1133,
+      "step": 14532
+    },
+    {
+      "epoch": 0.88298195516131,
+      "grad_norm": 0.11547403037548065,
+      "learning_rate": 3.3643972357744746e-06,
+      "loss": 1.0592,
+      "step": 14533
+    },
+    {
+      "epoch": 0.8830427121939365,
+      "grad_norm": 0.1985643059015274,
+      "learning_rate": 3.360945938023585e-06,
+      "loss": 1.2399,
+      "step": 14534
+    },
+    {
+      "epoch": 0.883103469226563,
+      "grad_norm": 0.386805921792984,
+      "learning_rate": 3.357496349836248e-06,
+      "loss": 1.2557,
+      "step": 14535
+    },
+    {
+      "epoch": 0.8831642262591896,
+      "grad_norm": 1.045297622680664,
+      "learning_rate": 3.3540484713388832e-06,
+      "loss": 1.1679,
+      "step": 14536
+    },
+    {
+      "epoch": 0.883224983291816,
+      "grad_norm": 0.21725356578826904,
+      "learning_rate": 3.350602302657896e-06,
+      "loss": 1.1921,
+      "step": 14537
+    },
+    {
+      "epoch": 0.8832857403244425,
+      "grad_norm": 0.10408329963684082,
+      "learning_rate": 3.347157843919585e-06,
+      "loss": 1.0146,
+      "step": 14538
+    },
+    {
+      "epoch": 0.883346497357069,
+      "grad_norm": 0.181331604719162,
+      "learning_rate": 3.3437150952502327e-06,
+      "loss": 1.1562,
+      "step": 14539
+    },
+    {
+      "epoch": 0.8834072543896956,
+      "grad_norm": 0.11471977829933167,
+      "learning_rate": 3.3402740567760214e-06,
+      "loss": 1.0624,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8834680114223221,
+      "grad_norm": 0.6379608511924744,
+      "learning_rate": 3.336834728623084e-06,
+      "loss": 1.123,
+      "step": 14541
+    },
+    {
+      "epoch": 0.8835287684549487,
+      "grad_norm": 0.1922662854194641,
+      "learning_rate": 3.333397110917497e-06,
+      "loss": 1.1686,
+      "step": 14542
+    },
+    {
+      "epoch": 0.8835895254875752,
+      "grad_norm": 0.14018617570400238,
+      "learning_rate": 3.329961203785242e-06,
+      "loss": 1.066,
+      "step": 14543
+    },
+    {
+      "epoch": 0.8836502825202017,
+      "grad_norm": 0.14638058841228485,
+      "learning_rate": 3.3265270073522925e-06,
+      "loss": 0.9903,
+      "step": 14544
+    },
+    {
+      "epoch": 0.8837110395528283,
+      "grad_norm": 0.14833728969097137,
+      "learning_rate": 3.323094521744513e-06,
+      "loss": 1.1354,
+      "step": 14545
+    },
+    {
+      "epoch": 0.8837717965854548,
+      "grad_norm": 0.4308561682701111,
+      "learning_rate": 3.3196637470877256e-06,
+      "loss": 1.2485,
+      "step": 14546
+    },
+    {
+      "epoch": 0.8838325536180813,
+      "grad_norm": 2.2455477714538574,
+      "learning_rate": 3.3162346835076906e-06,
+      "loss": 1.1876,
+      "step": 14547
+    },
+    {
+      "epoch": 0.8838933106507079,
+      "grad_norm": 0.23969553411006927,
+      "learning_rate": 3.312807331130091e-06,
+      "loss": 1.1379,
+      "step": 14548
+    },
+    {
+      "epoch": 0.8839540676833344,
+      "grad_norm": 0.7818536758422852,
+      "learning_rate": 3.309381690080571e-06,
+      "loss": 1.1782,
+      "step": 14549
+    },
+    {
+      "epoch": 0.8840148247159608,
+      "grad_norm": 0.14767079055309296,
+      "learning_rate": 3.305957760484685e-06,
+      "loss": 1.1263,
+      "step": 14550
+    },
+    {
+      "epoch": 0.8840755817485874,
+      "grad_norm": 1.5664955377578735,
+      "learning_rate": 3.302535542467944e-06,
+      "loss": 1.0775,
+      "step": 14551
+    },
+    {
+      "epoch": 0.8841363387812139,
+      "grad_norm": 0.23392058908939362,
+      "learning_rate": 3.2991150361557865e-06,
+      "loss": 1.0968,
+      "step": 14552
+    },
+    {
+      "epoch": 0.8841970958138404,
+      "grad_norm": 0.19733375310897827,
+      "learning_rate": 3.2956962416736015e-06,
+      "loss": 1.0642,
+      "step": 14553
+    },
+    {
+      "epoch": 0.884257852846467,
+      "grad_norm": 4.169165134429932,
+      "learning_rate": 3.292279159146683e-06,
+      "loss": 1.0619,
+      "step": 14554
+    },
+    {
+      "epoch": 0.8843186098790935,
+      "grad_norm": 0.1237090602517128,
+      "learning_rate": 3.288863788700319e-06,
+      "loss": 1.0435,
+      "step": 14555
+    },
+    {
+      "epoch": 0.88437936691172,
+      "grad_norm": 0.12797720730304718,
+      "learning_rate": 3.2854501304596774e-06,
+      "loss": 1.0792,
+      "step": 14556
+    },
+    {
+      "epoch": 0.8844401239443466,
+      "grad_norm": 0.164870023727417,
+      "learning_rate": 3.282038184549885e-06,
+      "loss": 1.0304,
+      "step": 14557
+    },
+    {
+      "epoch": 0.8845008809769731,
+      "grad_norm": 0.12417539209127426,
+      "learning_rate": 3.2786279510960193e-06,
+      "loss": 1.0287,
+      "step": 14558
+    },
+    {
+      "epoch": 0.8845616380095996,
+      "grad_norm": 0.11996292322874069,
+      "learning_rate": 3.2752194302230644e-06,
+      "loss": 1.0658,
+      "step": 14559
+    },
+    {
+      "epoch": 0.8846223950422262,
+      "grad_norm": 0.11469022184610367,
+      "learning_rate": 3.271812622055981e-06,
+      "loss": 1.0098,
+      "step": 14560
+    },
+    {
+      "epoch": 0.8846831520748527,
+      "grad_norm": 0.10035540908575058,
+      "learning_rate": 3.268407526719641e-06,
+      "loss": 1.0082,
+      "step": 14561
+    },
+    {
+      "epoch": 0.8847439091074792,
+      "grad_norm": 0.177669957280159,
+      "learning_rate": 3.265004144338862e-06,
+      "loss": 1.1048,
+      "step": 14562
+    },
+    {
+      "epoch": 0.8848046661401057,
+      "grad_norm": 0.16027499735355377,
+      "learning_rate": 3.2616024750383833e-06,
+      "loss": 1.0394,
+      "step": 14563
+    },
+    {
+      "epoch": 0.8848654231727322,
+      "grad_norm": 0.6399933695793152,
+      "learning_rate": 3.2582025189428934e-06,
+      "loss": 1.1645,
+      "step": 14564
+    },
+    {
+      "epoch": 0.8849261802053587,
+      "grad_norm": 0.11340584605932236,
+      "learning_rate": 3.2548042761770316e-06,
+      "loss": 1.0378,
+      "step": 14565
+    },
+    {
+      "epoch": 0.8849869372379853,
+      "grad_norm": 0.24245789647102356,
+      "learning_rate": 3.2514077468653537e-06,
+      "loss": 1.1092,
+      "step": 14566
+    },
+    {
+      "epoch": 0.8850476942706118,
+      "grad_norm": 0.17457427084445953,
+      "learning_rate": 3.248012931132366e-06,
+      "loss": 1.1266,
+      "step": 14567
+    },
+    {
+      "epoch": 0.8851084513032383,
+      "grad_norm": 0.1674877405166626,
+      "learning_rate": 3.244619829102502e-06,
+      "loss": 1.0469,
+      "step": 14568
+    },
+    {
+      "epoch": 0.8851692083358649,
+      "grad_norm": 0.09634416550397873,
+      "learning_rate": 3.2412284409001235e-06,
+      "loss": 1.0049,
+      "step": 14569
+    },
+    {
+      "epoch": 0.8852299653684914,
+      "grad_norm": 0.3651910722255707,
+      "learning_rate": 3.2378387666495645e-06,
+      "loss": 1.1461,
+      "step": 14570
+    },
+    {
+      "epoch": 0.885290722401118,
+      "grad_norm": 0.17827612161636353,
+      "learning_rate": 3.234450806475059e-06,
+      "loss": 1.1395,
+      "step": 14571
+    },
+    {
+      "epoch": 0.8853514794337445,
+      "grad_norm": 0.1108350157737732,
+      "learning_rate": 3.2310645605008015e-06,
+      "loss": 1.0429,
+      "step": 14572
+    },
+    {
+      "epoch": 0.885412236466371,
+      "grad_norm": 0.2824110984802246,
+      "learning_rate": 3.2276800288509157e-06,
+      "loss": 1.2115,
+      "step": 14573
+    },
+    {
+      "epoch": 0.8854729934989976,
+      "grad_norm": 0.17498032748699188,
+      "learning_rate": 3.2242972116494574e-06,
+      "loss": 1.0318,
+      "step": 14574
+    },
+    {
+      "epoch": 0.8855337505316241,
+      "grad_norm": 0.10410551726818085,
+      "learning_rate": 3.2209161090204277e-06,
+      "loss": 1.0404,
+      "step": 14575
+    },
+    {
+      "epoch": 0.8855945075642505,
+      "grad_norm": 0.1333116739988327,
+      "learning_rate": 3.217536721087755e-06,
+      "loss": 1.0476,
+      "step": 14576
+    },
+    {
+      "epoch": 0.885655264596877,
+      "grad_norm": 0.24690653383731842,
+      "learning_rate": 3.2141590479753236e-06,
+      "loss": 1.0213,
+      "step": 14577
+    },
+    {
+      "epoch": 0.8857160216295036,
+      "grad_norm": 0.14949177205562592,
+      "learning_rate": 3.210783089806929e-06,
+      "loss": 1.1178,
+      "step": 14578
+    },
+    {
+      "epoch": 0.8857767786621301,
+      "grad_norm": 0.22718439996242523,
+      "learning_rate": 3.2074088467063277e-06,
+      "loss": 1.0418,
+      "step": 14579
+    },
+    {
+      "epoch": 0.8858375356947567,
+      "grad_norm": 0.16216522455215454,
+      "learning_rate": 3.2040363187971935e-06,
+      "loss": 1.1333,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8858982927273832,
+      "grad_norm": 0.11710798740386963,
+      "learning_rate": 3.2006655062031597e-06,
+      "loss": 1.0534,
+      "step": 14581
+    },
+    {
+      "epoch": 0.8859590497600097,
+      "grad_norm": 0.14996717870235443,
+      "learning_rate": 3.197296409047784e-06,
+      "loss": 1.2264,
+      "step": 14582
+    },
+    {
+      "epoch": 0.8860198067926363,
+      "grad_norm": 0.14383438229560852,
+      "learning_rate": 3.193929027454551e-06,
+      "loss": 1.056,
+      "step": 14583
+    },
+    {
+      "epoch": 0.8860805638252628,
+      "grad_norm": 0.10204288363456726,
+      "learning_rate": 3.1905633615468944e-06,
+      "loss": 1.0455,
+      "step": 14584
+    },
+    {
+      "epoch": 0.8861413208578893,
+      "grad_norm": 0.16288653016090393,
+      "learning_rate": 3.1871994114481772e-06,
+      "loss": 1.1167,
+      "step": 14585
+    },
+    {
+      "epoch": 0.8862020778905159,
+      "grad_norm": 0.1325506567955017,
+      "learning_rate": 3.1838371772817286e-06,
+      "loss": 1.0647,
+      "step": 14586
+    },
+    {
+      "epoch": 0.8862628349231424,
+      "grad_norm": 0.14301139116287231,
+      "learning_rate": 3.1804766591707724e-06,
+      "loss": 1.0752,
+      "step": 14587
+    },
+    {
+      "epoch": 0.8863235919557689,
+      "grad_norm": 0.22541333734989166,
+      "learning_rate": 3.177117857238493e-06,
+      "loss": 1.1439,
+      "step": 14588
+    },
+    {
+      "epoch": 0.8863843489883954,
+      "grad_norm": 0.14628830552101135,
+      "learning_rate": 3.173760771608009e-06,
+      "loss": 1.1141,
+      "step": 14589
+    },
+    {
+      "epoch": 0.8864451060210219,
+      "grad_norm": 0.17259405553340912,
+      "learning_rate": 3.1704054024023767e-06,
+      "loss": 1.0724,
+      "step": 14590
+    },
+    {
+      "epoch": 0.8865058630536484,
+      "grad_norm": 0.1687692552804947,
+      "learning_rate": 3.167051749744587e-06,
+      "loss": 1.0829,
+      "step": 14591
+    },
+    {
+      "epoch": 0.886566620086275,
+      "grad_norm": 0.1520577371120453,
+      "learning_rate": 3.16369981375757e-06,
+      "loss": 1.0366,
+      "step": 14592
+    },
+    {
+      "epoch": 0.8866273771189015,
+      "grad_norm": 0.17253956198692322,
+      "learning_rate": 3.160349594564188e-06,
+      "loss": 1.0822,
+      "step": 14593
+    },
+    {
+      "epoch": 0.886688134151528,
+      "grad_norm": 8.534601211547852,
+      "learning_rate": 3.157001092287243e-06,
+      "loss": 1.1007,
+      "step": 14594
+    },
+    {
+      "epoch": 0.8867488911841546,
+      "grad_norm": 0.13734091818332672,
+      "learning_rate": 3.1536543070494806e-06,
+      "loss": 1.0672,
+      "step": 14595
+    },
+    {
+      "epoch": 0.8868096482167811,
+      "grad_norm": 0.2312757670879364,
+      "learning_rate": 3.1503092389735754e-06,
+      "loss": 1.051,
+      "step": 14596
+    },
+    {
+      "epoch": 0.8868704052494076,
+      "grad_norm": 1.7190401554107666,
+      "learning_rate": 3.146965888182135e-06,
+      "loss": 1.1829,
+      "step": 14597
+    },
+    {
+      "epoch": 0.8869311622820342,
+      "grad_norm": 0.31289583444595337,
+      "learning_rate": 3.1436242547977225e-06,
+      "loss": 1.2083,
+      "step": 14598
+    },
+    {
+      "epoch": 0.8869919193146607,
+      "grad_norm": 0.1963265836238861,
+      "learning_rate": 3.140284338942817e-06,
+      "loss": 1.1283,
+      "step": 14599
+    },
+    {
+      "epoch": 0.8870526763472872,
+      "grad_norm": 0.17142769694328308,
+      "learning_rate": 3.1369461407398495e-06,
+      "loss": 1.0933,
+      "step": 14600
+    },
+    {
+      "epoch": 0.8871134333799138,
+      "grad_norm": 0.10682990401983261,
+      "learning_rate": 3.133609660311171e-06,
+      "loss": 0.9715,
+      "step": 14601
+    },
+    {
+      "epoch": 0.8871741904125402,
+      "grad_norm": 0.10956233739852905,
+      "learning_rate": 3.1302748977791007e-06,
+      "loss": 1.0109,
+      "step": 14602
+    },
+    {
+      "epoch": 0.8872349474451667,
+      "grad_norm": 0.1335013508796692,
+      "learning_rate": 3.1269418532658635e-06,
+      "loss": 1.1286,
+      "step": 14603
+    },
+    {
+      "epoch": 0.8872957044777933,
+      "grad_norm": 0.1755015254020691,
+      "learning_rate": 3.1236105268936277e-06,
+      "loss": 1.1408,
+      "step": 14604
+    },
+    {
+      "epoch": 0.8873564615104198,
+      "grad_norm": 0.1585562825202942,
+      "learning_rate": 3.1202809187845016e-06,
+      "loss": 1.0053,
+      "step": 14605
+    },
+    {
+      "epoch": 0.8874172185430463,
+      "grad_norm": 0.22049635648727417,
+      "learning_rate": 3.1169530290605486e-06,
+      "loss": 1.2608,
+      "step": 14606
+    },
+    {
+      "epoch": 0.8874779755756729,
+      "grad_norm": 0.11492518335580826,
+      "learning_rate": 3.113626857843743e-06,
+      "loss": 1.023,
+      "step": 14607
+    },
+    {
+      "epoch": 0.8875387326082994,
+      "grad_norm": 0.19224266707897186,
+      "learning_rate": 3.1103024052560105e-06,
+      "loss": 1.1204,
+      "step": 14608
+    },
+    {
+      "epoch": 0.8875994896409259,
+      "grad_norm": 0.32298731803894043,
+      "learning_rate": 3.1069796714192132e-06,
+      "loss": 1.0594,
+      "step": 14609
+    },
+    {
+      "epoch": 0.8876602466735525,
+      "grad_norm": 0.11298403888940811,
+      "learning_rate": 3.1036586564551273e-06,
+      "loss": 0.9929,
+      "step": 14610
+    },
+    {
+      "epoch": 0.887721003706179,
+      "grad_norm": 0.1807352751493454,
+      "learning_rate": 3.1003393604855047e-06,
+      "loss": 1.0969,
+      "step": 14611
+    },
+    {
+      "epoch": 0.8877817607388055,
+      "grad_norm": 0.14948639273643494,
+      "learning_rate": 3.097021783632009e-06,
+      "loss": 1.029,
+      "step": 14612
+    },
+    {
+      "epoch": 0.8878425177714321,
+      "grad_norm": 0.14266057312488556,
+      "learning_rate": 3.0937059260162438e-06,
+      "loss": 1.0347,
+      "step": 14613
+    },
+    {
+      "epoch": 0.8879032748040586,
+      "grad_norm": 0.13356633484363556,
+      "learning_rate": 3.0903917877597556e-06,
+      "loss": 1.0597,
+      "step": 14614
+    },
+    {
+      "epoch": 0.887964031836685,
+      "grad_norm": 0.23884007334709167,
+      "learning_rate": 3.087079368984025e-06,
+      "loss": 1.118,
+      "step": 14615
+    },
+    {
+      "epoch": 0.8880247888693116,
+      "grad_norm": 0.2663938105106354,
+      "learning_rate": 3.0837686698104663e-06,
+      "loss": 1.1783,
+      "step": 14616
+    },
+    {
+      "epoch": 0.8880855459019381,
+      "grad_norm": 0.2507220506668091,
+      "learning_rate": 3.0804596903604376e-06,
+      "loss": 1.0454,
+      "step": 14617
+    },
+    {
+      "epoch": 0.8881463029345646,
+      "grad_norm": 0.3995007276535034,
+      "learning_rate": 3.0771524307552256e-06,
+      "loss": 1.0603,
+      "step": 14618
+    },
+    {
+      "epoch": 0.8882070599671912,
+      "grad_norm": 0.10384316742420197,
+      "learning_rate": 3.0738468911160614e-06,
+      "loss": 1.0473,
+      "step": 14619
+    },
+    {
+      "epoch": 0.8882678169998177,
+      "grad_norm": 0.11711938679218292,
+      "learning_rate": 3.0705430715641135e-06,
+      "loss": 1.0116,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8883285740324443,
+      "grad_norm": 0.12873300909996033,
+      "learning_rate": 3.06724097222047e-06,
+      "loss": 1.0806,
+      "step": 14621
+    },
+    {
+      "epoch": 0.8883893310650708,
+      "grad_norm": 0.129280686378479,
+      "learning_rate": 3.0639405932061994e-06,
+      "loss": 1.0045,
+      "step": 14622
+    },
+    {
+      "epoch": 0.8884500880976973,
+      "grad_norm": 0.12430091947317123,
+      "learning_rate": 3.0606419346422444e-06,
+      "loss": 1.0395,
+      "step": 14623
+    },
+    {
+      "epoch": 0.8885108451303239,
+      "grad_norm": 0.19456730782985687,
+      "learning_rate": 3.057344996649536e-06,
+      "loss": 1.1176,
+      "step": 14624
+    },
+    {
+      "epoch": 0.8885716021629504,
+      "grad_norm": 0.1140284612774849,
+      "learning_rate": 3.054049779348922e-06,
+      "loss": 1.03,
+      "step": 14625
+    },
+    {
+      "epoch": 0.8886323591955769,
+      "grad_norm": 0.1789393126964569,
+      "learning_rate": 3.050756282861178e-06,
+      "loss": 1.0705,
+      "step": 14626
+    },
+    {
+      "epoch": 0.8886931162282035,
+      "grad_norm": 0.25484660267829895,
+      "learning_rate": 3.0474645073070517e-06,
+      "loss": 1.1812,
+      "step": 14627
+    },
+    {
+      "epoch": 0.88875387326083,
+      "grad_norm": 0.139180988073349,
+      "learning_rate": 3.04417445280718e-06,
+      "loss": 1.0934,
+      "step": 14628
+    },
+    {
+      "epoch": 0.8888146302934564,
+      "grad_norm": 0.15103241801261902,
+      "learning_rate": 3.040886119482184e-06,
+      "loss": 0.9761,
+      "step": 14629
+    },
+    {
+      "epoch": 0.888875387326083,
+      "grad_norm": 0.17447814345359802,
+      "learning_rate": 3.0375995074525764e-06,
+      "loss": 1.0903,
+      "step": 14630
+    },
+    {
+      "epoch": 0.8889361443587095,
+      "grad_norm": 0.12317290902137756,
+      "learning_rate": 3.0343146168388235e-06,
+      "loss": 1.064,
+      "step": 14631
+    },
+    {
+      "epoch": 0.888996901391336,
+      "grad_norm": 0.13857927918434143,
+      "learning_rate": 3.031031447761362e-06,
+      "loss": 1.0435,
+      "step": 14632
+    },
+    {
+      "epoch": 0.8890576584239626,
+      "grad_norm": 0.19820040464401245,
+      "learning_rate": 3.0277500003405178e-06,
+      "loss": 1.1088,
+      "step": 14633
+    },
+    {
+      "epoch": 0.8891184154565891,
+      "grad_norm": 0.1622518002986908,
+      "learning_rate": 3.0244702746965723e-06,
+      "loss": 1.0283,
+      "step": 14634
+    },
+    {
+      "epoch": 0.8891791724892156,
+      "grad_norm": 0.16865570843219757,
+      "learning_rate": 3.021192270949763e-06,
+      "loss": 0.992,
+      "step": 14635
+    },
+    {
+      "epoch": 0.8892399295218422,
+      "grad_norm": 0.16449861228466034,
+      "learning_rate": 3.0179159892202157e-06,
+      "loss": 1.0331,
+      "step": 14636
+    },
+    {
+      "epoch": 0.8893006865544687,
+      "grad_norm": 0.20679821074008942,
+      "learning_rate": 3.0146414296280454e-06,
+      "loss": 1.1515,
+      "step": 14637
+    },
+    {
+      "epoch": 0.8893614435870952,
+      "grad_norm": 0.10066144168376923,
+      "learning_rate": 3.0113685922932734e-06,
+      "loss": 1.03,
+      "step": 14638
+    },
+    {
+      "epoch": 0.8894222006197218,
+      "grad_norm": 0.12101301550865173,
+      "learning_rate": 3.008097477335875e-06,
+      "loss": 1.0367,
+      "step": 14639
+    },
+    {
+      "epoch": 0.8894829576523483,
+      "grad_norm": 0.10414587706327438,
+      "learning_rate": 3.004828084875744e-06,
+      "loss": 1.0178,
+      "step": 14640
+    },
+    {
+      "epoch": 0.8895437146849748,
+      "grad_norm": 0.24158069491386414,
+      "learning_rate": 3.001560415032717e-06,
+      "loss": 1.1149,
+      "step": 14641
+    },
+    {
+      "epoch": 0.8896044717176013,
+      "grad_norm": 1.1494296789169312,
+      "learning_rate": 2.9982944679265877e-06,
+      "loss": 1.1292,
+      "step": 14642
+    },
+    {
+      "epoch": 0.8896652287502278,
+      "grad_norm": 0.2280639261007309,
+      "learning_rate": 2.995030243677055e-06,
+      "loss": 1.2318,
+      "step": 14643
+    },
+    {
+      "epoch": 0.8897259857828543,
+      "grad_norm": 0.167388454079628,
+      "learning_rate": 2.9917677424037725e-06,
+      "loss": 1.0778,
+      "step": 14644
+    },
+    {
+      "epoch": 0.8897867428154809,
+      "grad_norm": 0.11865873634815216,
+      "learning_rate": 2.988506964226334e-06,
+      "loss": 1.0208,
+      "step": 14645
+    },
+    {
+      "epoch": 0.8898474998481074,
+      "grad_norm": 0.1320948302745819,
+      "learning_rate": 2.985247909264255e-06,
+      "loss": 1.0266,
+      "step": 14646
+    },
+    {
+      "epoch": 0.8899082568807339,
+      "grad_norm": 0.2526165246963501,
+      "learning_rate": 2.9819905776370006e-06,
+      "loss": 1.1069,
+      "step": 14647
+    },
+    {
+      "epoch": 0.8899690139133605,
+      "grad_norm": 0.10250655561685562,
+      "learning_rate": 2.97873496946397e-06,
+      "loss": 0.9924,
+      "step": 14648
+    },
+    {
+      "epoch": 0.890029770945987,
+      "grad_norm": 0.1981375366449356,
+      "learning_rate": 2.975481084864512e-06,
+      "loss": 1.0848,
+      "step": 14649
+    },
+    {
+      "epoch": 0.8900905279786135,
+      "grad_norm": 0.1285632848739624,
+      "learning_rate": 2.972228923957876e-06,
+      "loss": 0.9948,
+      "step": 14650
+    },
+    {
+      "epoch": 0.8901512850112401,
+      "grad_norm": 0.2181164026260376,
+      "learning_rate": 2.9689784868632773e-06,
+      "loss": 1.0004,
+      "step": 14651
+    },
+    {
+      "epoch": 0.8902120420438666,
+      "grad_norm": 0.13612231612205505,
+      "learning_rate": 2.9657297736998548e-06,
+      "loss": 1.0731,
+      "step": 14652
+    },
+    {
+      "epoch": 0.8902727990764931,
+      "grad_norm": 0.1471223533153534,
+      "learning_rate": 2.9624827845867065e-06,
+      "loss": 1.1116,
+      "step": 14653
+    },
+    {
+      "epoch": 0.8903335561091197,
+      "grad_norm": 0.2809170186519623,
+      "learning_rate": 2.9592375196428436e-06,
+      "loss": 1.1078,
+      "step": 14654
+    },
+    {
+      "epoch": 0.8903943131417461,
+      "grad_norm": 0.12725882232189178,
+      "learning_rate": 2.955993978987226e-06,
+      "loss": 1.0368,
+      "step": 14655
+    },
+    {
+      "epoch": 0.8904550701743726,
+      "grad_norm": 0.21599172055721283,
+      "learning_rate": 2.9527521627387476e-06,
+      "loss": 1.1954,
+      "step": 14656
+    },
+    {
+      "epoch": 0.8905158272069992,
+      "grad_norm": 0.18019451200962067,
+      "learning_rate": 2.949512071016214e-06,
+      "loss": 1.1317,
+      "step": 14657
+    },
+    {
+      "epoch": 0.8905765842396257,
+      "grad_norm": 0.2100551873445511,
+      "learning_rate": 2.9462737039384236e-06,
+      "loss": 1.0517,
+      "step": 14658
+    },
+    {
+      "epoch": 0.8906373412722522,
+      "grad_norm": 0.14522022008895874,
+      "learning_rate": 2.9430370616240656e-06,
+      "loss": 1.0404,
+      "step": 14659
+    },
+    {
+      "epoch": 0.8906980983048788,
+      "grad_norm": 0.24829906225204468,
+      "learning_rate": 2.9398021441917724e-06,
+      "loss": 1.1041,
+      "step": 14660
+    },
+    {
+      "epoch": 0.8907588553375053,
+      "grad_norm": 0.1998528391122818,
+      "learning_rate": 2.936568951760138e-06,
+      "loss": 1.0044,
+      "step": 14661
+    },
+    {
+      "epoch": 0.8908196123701319,
+      "grad_norm": 0.17631661891937256,
+      "learning_rate": 2.933337484447657e-06,
+      "loss": 1.1724,
+      "step": 14662
+    },
+    {
+      "epoch": 0.8908803694027584,
+      "grad_norm": 0.09878667443990707,
+      "learning_rate": 2.9301077423727953e-06,
+      "loss": 1.0421,
+      "step": 14663
+    },
+    {
+      "epoch": 0.8909411264353849,
+      "grad_norm": 0.13521592319011688,
+      "learning_rate": 2.9268797256539305e-06,
+      "loss": 1.0611,
+      "step": 14664
+    },
+    {
+      "epoch": 0.8910018834680115,
+      "grad_norm": 0.14376071095466614,
+      "learning_rate": 2.9236534344093904e-06,
+      "loss": 1.0151,
+      "step": 14665
+    },
+    {
+      "epoch": 0.891062640500638,
+      "grad_norm": 0.10521792620420456,
+      "learning_rate": 2.92042886875743e-06,
+      "loss": 1.0532,
+      "step": 14666
+    },
+    {
+      "epoch": 0.8911233975332645,
+      "grad_norm": 0.20890606939792633,
+      "learning_rate": 2.9172060288162493e-06,
+      "loss": 1.1579,
+      "step": 14667
+    },
+    {
+      "epoch": 0.891184154565891,
+      "grad_norm": 0.2172417789697647,
+      "learning_rate": 2.913984914703977e-06,
+      "loss": 1.1392,
+      "step": 14668
+    },
+    {
+      "epoch": 0.8912449115985175,
+      "grad_norm": 0.468802273273468,
+      "learning_rate": 2.9107655265387e-06,
+      "loss": 1.1962,
+      "step": 14669
+    },
+    {
+      "epoch": 0.891305668631144,
+      "grad_norm": 0.15653954446315765,
+      "learning_rate": 2.9075478644384146e-06,
+      "loss": 1.1079,
+      "step": 14670
+    },
+    {
+      "epoch": 0.8913664256637706,
+      "grad_norm": 0.2025233507156372,
+      "learning_rate": 2.904331928521059e-06,
+      "loss": 1.0116,
+      "step": 14671
+    },
+    {
+      "epoch": 0.8914271826963971,
+      "grad_norm": 0.14733392000198364,
+      "learning_rate": 2.9011177189045225e-06,
+      "loss": 1.0793,
+      "step": 14672
+    },
+    {
+      "epoch": 0.8914879397290236,
+      "grad_norm": 0.11526741832494736,
+      "learning_rate": 2.8979052357066104e-06,
+      "loss": 1.0578,
+      "step": 14673
+    },
+    {
+      "epoch": 0.8915486967616502,
+      "grad_norm": 0.1686958372592926,
+      "learning_rate": 2.894694479045096e-06,
+      "loss": 1.0547,
+      "step": 14674
+    },
+    {
+      "epoch": 0.8916094537942767,
+      "grad_norm": 0.10652989149093628,
+      "learning_rate": 2.891485449037662e-06,
+      "loss": 1.0395,
+      "step": 14675
+    },
+    {
+      "epoch": 0.8916702108269032,
+      "grad_norm": 0.1593276858329773,
+      "learning_rate": 2.8882781458019426e-06,
+      "loss": 1.0542,
+      "step": 14676
+    },
+    {
+      "epoch": 0.8917309678595298,
+      "grad_norm": 0.10352182388305664,
+      "learning_rate": 2.8850725694554827e-06,
+      "loss": 1.0187,
+      "step": 14677
+    },
+    {
+      "epoch": 0.8917917248921563,
+      "grad_norm": 0.13450925052165985,
+      "learning_rate": 2.881868720115788e-06,
+      "loss": 1.1116,
+      "step": 14678
+    },
+    {
+      "epoch": 0.8918524819247828,
+      "grad_norm": 0.5086029767990112,
+      "learning_rate": 2.8786665979003146e-06,
+      "loss": 1.0086,
+      "step": 14679
+    },
+    {
+      "epoch": 0.8919132389574094,
+      "grad_norm": 0.109037846326828,
+      "learning_rate": 2.8754662029264248e-06,
+      "loss": 1.0607,
+      "step": 14680
+    },
+    {
+      "epoch": 0.8919739959900358,
+      "grad_norm": 0.2711449861526489,
+      "learning_rate": 2.8722675353114236e-06,
+      "loss": 1.1371,
+      "step": 14681
+    },
+    {
+      "epoch": 0.8920347530226623,
+      "grad_norm": 0.08838333189487457,
+      "learning_rate": 2.869070595172574e-06,
+      "loss": 0.9772,
+      "step": 14682
+    },
+    {
+      "epoch": 0.8920955100552889,
+      "grad_norm": 0.13673768937587738,
+      "learning_rate": 2.8658753826270477e-06,
+      "loss": 1.0905,
+      "step": 14683
+    },
+    {
+      "epoch": 0.8921562670879154,
+      "grad_norm": 0.15699419379234314,
+      "learning_rate": 2.862681897791969e-06,
+      "loss": 1.1289,
+      "step": 14684
+    },
+    {
+      "epoch": 0.8922170241205419,
+      "grad_norm": 0.6075234413146973,
+      "learning_rate": 2.859490140784399e-06,
+      "loss": 1.0319,
+      "step": 14685
+    },
+    {
+      "epoch": 0.8922777811531685,
+      "grad_norm": 0.12360430508852005,
+      "learning_rate": 2.856300111721327e-06,
+      "loss": 1.1443,
+      "step": 14686
+    },
+    {
+      "epoch": 0.892338538185795,
+      "grad_norm": 0.29364779591560364,
+      "learning_rate": 2.853111810719694e-06,
+      "loss": 1.0884,
+      "step": 14687
+    },
+    {
+      "epoch": 0.8923992952184215,
+      "grad_norm": 0.3638782799243927,
+      "learning_rate": 2.849925237896356e-06,
+      "loss": 1.0868,
+      "step": 14688
+    },
+    {
+      "epoch": 0.8924600522510481,
+      "grad_norm": 0.14094452559947968,
+      "learning_rate": 2.8467403933681193e-06,
+      "loss": 1.0028,
+      "step": 14689
+    },
+    {
+      "epoch": 0.8925208092836746,
+      "grad_norm": 0.09357637912034988,
+      "learning_rate": 2.843557277251735e-06,
+      "loss": 1.0289,
+      "step": 14690
+    },
+    {
+      "epoch": 0.8925815663163011,
+      "grad_norm": 0.1741989254951477,
+      "learning_rate": 2.840375889663871e-06,
+      "loss": 1.0886,
+      "step": 14691
+    },
+    {
+      "epoch": 0.8926423233489277,
+      "grad_norm": 0.18354468047618866,
+      "learning_rate": 2.8371962307211395e-06,
+      "loss": 1.0822,
+      "step": 14692
+    },
+    {
+      "epoch": 0.8927030803815542,
+      "grad_norm": 0.14537058770656586,
+      "learning_rate": 2.834018300540103e-06,
+      "loss": 1.0258,
+      "step": 14693
+    },
+    {
+      "epoch": 0.8927638374141806,
+      "grad_norm": 0.11739664524793625,
+      "learning_rate": 2.8308420992372298e-06,
+      "loss": 1.0519,
+      "step": 14694
+    },
+    {
+      "epoch": 0.8928245944468072,
+      "grad_norm": 0.17187419533729553,
+      "learning_rate": 2.8276676269289704e-06,
+      "loss": 1.0675,
+      "step": 14695
+    },
+    {
+      "epoch": 0.8928853514794337,
+      "grad_norm": 0.17267253994941711,
+      "learning_rate": 2.8244948837316764e-06,
+      "loss": 1.0788,
+      "step": 14696
+    },
+    {
+      "epoch": 0.8929461085120602,
+      "grad_norm": 0.16254602372646332,
+      "learning_rate": 2.821323869761633e-06,
+      "loss": 1.0853,
+      "step": 14697
+    },
+    {
+      "epoch": 0.8930068655446868,
+      "grad_norm": 0.1279696673154831,
+      "learning_rate": 2.81815458513508e-06,
+      "loss": 1.0458,
+      "step": 14698
+    },
+    {
+      "epoch": 0.8930676225773133,
+      "grad_norm": 0.16091276705265045,
+      "learning_rate": 2.8149870299681913e-06,
+      "loss": 1.1893,
+      "step": 14699
+    },
+    {
+      "epoch": 0.8931283796099398,
+      "grad_norm": 0.1802794635295868,
+      "learning_rate": 2.8118212043770798e-06,
+      "loss": 1.067,
+      "step": 14700
+    },
+    {
+      "epoch": 0.8931891366425664,
+      "grad_norm": 0.1672443151473999,
+      "learning_rate": 2.80865710847778e-06,
+      "loss": 1.105,
+      "step": 14701
+    },
+    {
+      "epoch": 0.8932498936751929,
+      "grad_norm": 0.48433154821395874,
+      "learning_rate": 2.8054947423862833e-06,
+      "loss": 1.0651,
+      "step": 14702
+    },
+    {
+      "epoch": 0.8933106507078195,
+      "grad_norm": 0.1625070869922638,
+      "learning_rate": 2.802334106218485e-06,
+      "loss": 1.09,
+      "step": 14703
+    },
+    {
+      "epoch": 0.893371407740446,
+      "grad_norm": 0.16016674041748047,
+      "learning_rate": 2.7991752000902595e-06,
+      "loss": 1.0444,
+      "step": 14704
+    },
+    {
+      "epoch": 0.8934321647730725,
+      "grad_norm": 0.16798751056194305,
+      "learning_rate": 2.7960180241173917e-06,
+      "loss": 1.0838,
+      "step": 14705
+    },
+    {
+      "epoch": 0.8934929218056991,
+      "grad_norm": 0.11957212537527084,
+      "learning_rate": 2.7928625784156115e-06,
+      "loss": 1.0446,
+      "step": 14706
+    },
+    {
+      "epoch": 0.8935536788383255,
+      "grad_norm": 0.1715918779373169,
+      "learning_rate": 2.789708863100571e-06,
+      "loss": 1.0617,
+      "step": 14707
+    },
+    {
+      "epoch": 0.893614435870952,
+      "grad_norm": 0.1891513615846634,
+      "learning_rate": 2.7865568782878826e-06,
+      "loss": 1.0494,
+      "step": 14708
+    },
+    {
+      "epoch": 0.8936751929035786,
+      "grad_norm": 0.14633367955684662,
+      "learning_rate": 2.783406624093077e-06,
+      "loss": 1.0881,
+      "step": 14709
+    },
+    {
+      "epoch": 0.8937359499362051,
+      "grad_norm": 0.319070428609848,
+      "learning_rate": 2.7802581006316273e-06,
+      "loss": 1.0443,
+      "step": 14710
+    },
+    {
+      "epoch": 0.8937967069688316,
+      "grad_norm": 0.1441330760717392,
+      "learning_rate": 2.777111308018948e-06,
+      "loss": 1.0995,
+      "step": 14711
+    },
+    {
+      "epoch": 0.8938574640014582,
+      "grad_norm": 0.20820242166519165,
+      "learning_rate": 2.7739662463703743e-06,
+      "loss": 1.1511,
+      "step": 14712
+    },
+    {
+      "epoch": 0.8939182210340847,
+      "grad_norm": 0.1259198635816574,
+      "learning_rate": 2.770822915801202e-06,
+      "loss": 1.0475,
+      "step": 14713
+    },
+    {
+      "epoch": 0.8939789780667112,
+      "grad_norm": 0.39245837926864624,
+      "learning_rate": 2.76768131642664e-06,
+      "loss": 1.0277,
+      "step": 14714
+    },
+    {
+      "epoch": 0.8940397350993378,
+      "grad_norm": 0.13661637902259827,
+      "learning_rate": 2.7645414483618516e-06,
+      "loss": 1.0478,
+      "step": 14715
+    },
+    {
+      "epoch": 0.8941004921319643,
+      "grad_norm": 0.16632981598377228,
+      "learning_rate": 2.7614033117219327e-06,
+      "loss": 1.0753,
+      "step": 14716
+    },
+    {
+      "epoch": 0.8941612491645908,
+      "grad_norm": 0.11878080666065216,
+      "learning_rate": 2.7582669066219034e-06,
+      "loss": 1.01,
+      "step": 14717
+    },
+    {
+      "epoch": 0.8942220061972174,
+      "grad_norm": 0.1428791731595993,
+      "learning_rate": 2.7551322331767326e-06,
+      "loss": 1.0535,
+      "step": 14718
+    },
+    {
+      "epoch": 0.8942827632298439,
+      "grad_norm": 0.2595476806163788,
+      "learning_rate": 2.7519992915013115e-06,
+      "loss": 0.9776,
+      "step": 14719
+    },
+    {
+      "epoch": 0.8943435202624703,
+      "grad_norm": 0.4360501170158386,
+      "learning_rate": 2.7488680817104982e-06,
+      "loss": 1.0733,
+      "step": 14720
+    },
+    {
+      "epoch": 0.8944042772950969,
+      "grad_norm": 0.20499996840953827,
+      "learning_rate": 2.7457386039190568e-06,
+      "loss": 1.0537,
+      "step": 14721
+    },
+    {
+      "epoch": 0.8944650343277234,
+      "grad_norm": 0.15135543048381805,
+      "learning_rate": 2.742610858241712e-06,
+      "loss": 1.019,
+      "step": 14722
+    },
+    {
+      "epoch": 0.8945257913603499,
+      "grad_norm": 0.13484498858451843,
+      "learning_rate": 2.7394848447930887e-06,
+      "loss": 1.211,
+      "step": 14723
+    },
+    {
+      "epoch": 0.8945865483929765,
+      "grad_norm": 0.19778922200202942,
+      "learning_rate": 2.736360563687779e-06,
+      "loss": 1.1677,
+      "step": 14724
+    },
+    {
+      "epoch": 0.894647305425603,
+      "grad_norm": 0.17957265675067902,
+      "learning_rate": 2.7332380150403182e-06,
+      "loss": 1.1081,
+      "step": 14725
+    },
+    {
+      "epoch": 0.8947080624582295,
+      "grad_norm": 8.465873718261719,
+      "learning_rate": 2.730117198965154e-06,
+      "loss": 1.0996,
+      "step": 14726
+    },
+    {
+      "epoch": 0.8947688194908561,
+      "grad_norm": 0.6284109950065613,
+      "learning_rate": 2.726998115576679e-06,
+      "loss": 1.2421,
+      "step": 14727
+    },
+    {
+      "epoch": 0.8948295765234826,
+      "grad_norm": 0.15040907263755798,
+      "learning_rate": 2.7238807649892226e-06,
+      "loss": 1.0695,
+      "step": 14728
+    },
+    {
+      "epoch": 0.8948903335561091,
+      "grad_norm": 0.16046516597270966,
+      "learning_rate": 2.7207651473170615e-06,
+      "loss": 1.077,
+      "step": 14729
+    },
+    {
+      "epoch": 0.8949510905887357,
+      "grad_norm": 0.41832560300827026,
+      "learning_rate": 2.7176512626743867e-06,
+      "loss": 1.0307,
+      "step": 14730
+    },
+    {
+      "epoch": 0.8950118476213622,
+      "grad_norm": 0.1138959676027298,
+      "learning_rate": 2.714539111175346e-06,
+      "loss": 1.036,
+      "step": 14731
+    },
+    {
+      "epoch": 0.8950726046539887,
+      "grad_norm": 0.16120284795761108,
+      "learning_rate": 2.7114286929340206e-06,
+      "loss": 1.1776,
+      "step": 14732
+    },
+    {
+      "epoch": 0.8951333616866153,
+      "grad_norm": 0.2626010775566101,
+      "learning_rate": 2.7083200080644077e-06,
+      "loss": 1.111,
+      "step": 14733
+    },
+    {
+      "epoch": 0.8951941187192417,
+      "grad_norm": 0.1747281551361084,
+      "learning_rate": 2.7052130566804723e-06,
+      "loss": 1.0399,
+      "step": 14734
+    },
+    {
+      "epoch": 0.8952548757518682,
+      "grad_norm": 0.12358351051807404,
+      "learning_rate": 2.702107838896084e-06,
+      "loss": 1.0774,
+      "step": 14735
+    },
+    {
+      "epoch": 0.8953156327844948,
+      "grad_norm": 0.1457117795944214,
+      "learning_rate": 2.699004354825091e-06,
+      "loss": 1.0574,
+      "step": 14736
+    },
+    {
+      "epoch": 0.8953763898171213,
+      "grad_norm": 0.1838354617357254,
+      "learning_rate": 2.69590260458123e-06,
+      "loss": 1.1001,
+      "step": 14737
+    },
+    {
+      "epoch": 0.8954371468497478,
+      "grad_norm": 0.16018275916576385,
+      "learning_rate": 2.6928025882782037e-06,
+      "loss": 1.0407,
+      "step": 14738
+    },
+    {
+      "epoch": 0.8954979038823744,
+      "grad_norm": 0.7069351673126221,
+      "learning_rate": 2.689704306029639e-06,
+      "loss": 1.2702,
+      "step": 14739
+    },
+    {
+      "epoch": 0.8955586609150009,
+      "grad_norm": 0.12320427596569061,
+      "learning_rate": 2.6866077579491054e-06,
+      "loss": 1.0563,
+      "step": 14740
+    },
+    {
+      "epoch": 0.8956194179476274,
+      "grad_norm": 0.1000717282295227,
+      "learning_rate": 2.6835129441501126e-06,
+      "loss": 1.0168,
+      "step": 14741
+    },
+    {
+      "epoch": 0.895680174980254,
+      "grad_norm": 0.6606243252754211,
+      "learning_rate": 2.6804198647461033e-06,
+      "loss": 1.0581,
+      "step": 14742
+    },
+    {
+      "epoch": 0.8957409320128805,
+      "grad_norm": 0.1815868616104126,
+      "learning_rate": 2.677328519850453e-06,
+      "loss": 1.1081,
+      "step": 14743
+    },
+    {
+      "epoch": 0.895801689045507,
+      "grad_norm": 0.09678613394498825,
+      "learning_rate": 2.674238909576471e-06,
+      "loss": 1.0217,
+      "step": 14744
+    },
+    {
+      "epoch": 0.8958624460781336,
+      "grad_norm": 0.12373851239681244,
+      "learning_rate": 2.671151034037395e-06,
+      "loss": 1.0852,
+      "step": 14745
+    },
+    {
+      "epoch": 0.8959232031107601,
+      "grad_norm": 0.3140968680381775,
+      "learning_rate": 2.668064893346439e-06,
+      "loss": 1.188,
+      "step": 14746
+    },
+    {
+      "epoch": 0.8959839601433865,
+      "grad_norm": 0.17102232575416565,
+      "learning_rate": 2.664980487616714e-06,
+      "loss": 1.1088,
+      "step": 14747
+    },
+    {
+      "epoch": 0.8960447171760131,
+      "grad_norm": 0.14022096991539001,
+      "learning_rate": 2.6618978169612786e-06,
+      "loss": 1.0149,
+      "step": 14748
+    },
+    {
+      "epoch": 0.8961054742086396,
+      "grad_norm": 0.10651004314422607,
+      "learning_rate": 2.658816881493131e-06,
+      "loss": 1.0699,
+      "step": 14749
+    },
+    {
+      "epoch": 0.8961662312412662,
+      "grad_norm": 0.10498160868883133,
+      "learning_rate": 2.6557376813251934e-06,
+      "loss": 1.0357,
+      "step": 14750
+    },
+    {
+      "epoch": 0.8962269882738927,
+      "grad_norm": 0.11480096727609634,
+      "learning_rate": 2.6526602165703464e-06,
+      "loss": 1.0319,
+      "step": 14751
+    },
+    {
+      "epoch": 0.8962877453065192,
+      "grad_norm": 0.12573085725307465,
+      "learning_rate": 2.6495844873413943e-06,
+      "loss": 1.0453,
+      "step": 14752
+    },
+    {
+      "epoch": 0.8963485023391458,
+      "grad_norm": 0.17667414247989655,
+      "learning_rate": 2.64651049375107e-06,
+      "loss": 1.1554,
+      "step": 14753
+    },
+    {
+      "epoch": 0.8964092593717723,
+      "grad_norm": 0.11018501967191696,
+      "learning_rate": 2.6434382359120604e-06,
+      "loss": 1.0749,
+      "step": 14754
+    },
+    {
+      "epoch": 0.8964700164043988,
+      "grad_norm": 0.1279340535402298,
+      "learning_rate": 2.6403677139369754e-06,
+      "loss": 1.0774,
+      "step": 14755
+    },
+    {
+      "epoch": 0.8965307734370254,
+      "grad_norm": 0.149647518992424,
+      "learning_rate": 2.63729892793837e-06,
+      "loss": 1.0736,
+      "step": 14756
+    },
+    {
+      "epoch": 0.8965915304696519,
+      "grad_norm": 0.16375941038131714,
+      "learning_rate": 2.6342318780287203e-06,
+      "loss": 1.1356,
+      "step": 14757
+    },
+    {
+      "epoch": 0.8966522875022784,
+      "grad_norm": 0.136698916554451,
+      "learning_rate": 2.6311665643204586e-06,
+      "loss": 1.0437,
+      "step": 14758
+    },
+    {
+      "epoch": 0.896713044534905,
+      "grad_norm": 0.18355245888233185,
+      "learning_rate": 2.6281029869259454e-06,
+      "loss": 1.0928,
+      "step": 14759
+    },
+    {
+      "epoch": 0.8967738015675314,
+      "grad_norm": 0.0999312549829483,
+      "learning_rate": 2.625041145957474e-06,
+      "loss": 1.0195,
+      "step": 14760
+    },
+    {
+      "epoch": 0.8968345586001579,
+      "grad_norm": 0.15324987471103668,
+      "learning_rate": 2.6219810415272714e-06,
+      "loss": 1.0849,
+      "step": 14761
+    },
+    {
+      "epoch": 0.8968953156327845,
+      "grad_norm": 0.10880114883184433,
+      "learning_rate": 2.6189226737475146e-06,
+      "loss": 1.0414,
+      "step": 14762
+    },
+    {
+      "epoch": 0.896956072665411,
+      "grad_norm": 0.11195948719978333,
+      "learning_rate": 2.615866042730314e-06,
+      "loss": 1.0144,
+      "step": 14763
+    },
+    {
+      "epoch": 0.8970168296980375,
+      "grad_norm": 0.635172963142395,
+      "learning_rate": 2.612811148587696e-06,
+      "loss": 1.065,
+      "step": 14764
+    },
+    {
+      "epoch": 0.8970775867306641,
+      "grad_norm": 0.09788442403078079,
+      "learning_rate": 2.609757991431644e-06,
+      "loss": 0.9867,
+      "step": 14765
+    },
+    {
+      "epoch": 0.8971383437632906,
+      "grad_norm": 0.1582690328359604,
+      "learning_rate": 2.6067065713740677e-06,
+      "loss": 1.0674,
+      "step": 14766
+    },
+    {
+      "epoch": 0.8971991007959171,
+      "grad_norm": 0.25265640020370483,
+      "learning_rate": 2.6036568885268287e-06,
+      "loss": 1.1109,
+      "step": 14767
+    },
+    {
+      "epoch": 0.8972598578285437,
+      "grad_norm": 0.15409798920154572,
+      "learning_rate": 2.600608943001709e-06,
+      "loss": 1.1001,
+      "step": 14768
+    },
+    {
+      "epoch": 0.8973206148611702,
+      "grad_norm": 0.11112014204263687,
+      "learning_rate": 2.597562734910436e-06,
+      "loss": 1.0156,
+      "step": 14769
+    },
+    {
+      "epoch": 0.8973813718937967,
+      "grad_norm": 0.23407863080501556,
+      "learning_rate": 2.594518264364659e-06,
+      "loss": 1.0893,
+      "step": 14770
+    },
+    {
+      "epoch": 0.8974421289264233,
+      "grad_norm": 0.27085840702056885,
+      "learning_rate": 2.591475531475973e-06,
+      "loss": 1.2096,
+      "step": 14771
+    },
+    {
+      "epoch": 0.8975028859590498,
+      "grad_norm": 0.4738995134830475,
+      "learning_rate": 2.588434536355916e-06,
+      "loss": 1.1933,
+      "step": 14772
+    },
+    {
+      "epoch": 0.8975636429916762,
+      "grad_norm": 0.18575914204120636,
+      "learning_rate": 2.58539527911596e-06,
+      "loss": 1.1266,
+      "step": 14773
+    },
+    {
+      "epoch": 0.8976244000243028,
+      "grad_norm": 0.29799142479896545,
+      "learning_rate": 2.5823577598675043e-06,
+      "loss": 1.2722,
+      "step": 14774
+    },
+    {
+      "epoch": 0.8976851570569293,
+      "grad_norm": 0.15743301808834076,
+      "learning_rate": 2.579321978721888e-06,
+      "loss": 1.1032,
+      "step": 14775
+    },
+    {
+      "epoch": 0.8977459140895558,
+      "grad_norm": 0.17843139171600342,
+      "learning_rate": 2.5762879357903946e-06,
+      "loss": 1.0774,
+      "step": 14776
+    },
+    {
+      "epoch": 0.8978066711221824,
+      "grad_norm": 0.10879513621330261,
+      "learning_rate": 2.573255631184235e-06,
+      "loss": 1.0216,
+      "step": 14777
+    },
+    {
+      "epoch": 0.8978674281548089,
+      "grad_norm": 0.09962224215269089,
+      "learning_rate": 2.570225065014553e-06,
+      "loss": 1.0091,
+      "step": 14778
+    },
+    {
+      "epoch": 0.8979281851874354,
+      "grad_norm": 0.3224875032901764,
+      "learning_rate": 2.5671962373924384e-06,
+      "loss": 1.1435,
+      "step": 14779
+    },
+    {
+      "epoch": 0.897988942220062,
+      "grad_norm": 0.2196751981973648,
+      "learning_rate": 2.5641691484289187e-06,
+      "loss": 1.0831,
+      "step": 14780
+    },
+    {
+      "epoch": 0.8980496992526885,
+      "grad_norm": 0.1339867115020752,
+      "learning_rate": 2.5611437982349493e-06,
+      "loss": 1.0189,
+      "step": 14781
+    },
+    {
+      "epoch": 0.898110456285315,
+      "grad_norm": 0.32741713523864746,
+      "learning_rate": 2.5581201869214087e-06,
+      "loss": 1.0599,
+      "step": 14782
+    },
+    {
+      "epoch": 0.8981712133179416,
+      "grad_norm": 0.11647716909646988,
+      "learning_rate": 2.5550983145991635e-06,
+      "loss": 1.0442,
+      "step": 14783
+    },
+    {
+      "epoch": 0.8982319703505681,
+      "grad_norm": 0.19201484322547913,
+      "learning_rate": 2.552078181378953e-06,
+      "loss": 1.0715,
+      "step": 14784
+    },
+    {
+      "epoch": 0.8982927273831947,
+      "grad_norm": 0.1456119567155838,
+      "learning_rate": 2.549059787371488e-06,
+      "loss": 1.0855,
+      "step": 14785
+    },
+    {
+      "epoch": 0.8983534844158211,
+      "grad_norm": 0.20046569406986237,
+      "learning_rate": 2.546043132687409e-06,
+      "loss": 1.006,
+      "step": 14786
+    },
+    {
+      "epoch": 0.8984142414484476,
+      "grad_norm": 0.09116701036691666,
+      "learning_rate": 2.5430282174372823e-06,
+      "loss": 1.0245,
+      "step": 14787
+    },
+    {
+      "epoch": 0.8984749984810741,
+      "grad_norm": 0.11472343653440475,
+      "learning_rate": 2.5400150417316358e-06,
+      "loss": 1.059,
+      "step": 14788
+    },
+    {
+      "epoch": 0.8985357555137007,
+      "grad_norm": 0.11024674028158188,
+      "learning_rate": 2.537003605680921e-06,
+      "loss": 1.0503,
+      "step": 14789
+    },
+    {
+      "epoch": 0.8985965125463272,
+      "grad_norm": 0.11182813346385956,
+      "learning_rate": 2.5339939093955047e-06,
+      "loss": 1.0394,
+      "step": 14790
+    },
+    {
+      "epoch": 0.8986572695789538,
+      "grad_norm": 0.1158798485994339,
+      "learning_rate": 2.5309859529857095e-06,
+      "loss": 1.0521,
+      "step": 14791
+    },
+    {
+      "epoch": 0.8987180266115803,
+      "grad_norm": 0.11270305514335632,
+      "learning_rate": 2.5279797365618085e-06,
+      "loss": 1.0086,
+      "step": 14792
+    },
+    {
+      "epoch": 0.8987787836442068,
+      "grad_norm": 0.2143092006444931,
+      "learning_rate": 2.524975260233986e-06,
+      "loss": 1.0624,
+      "step": 14793
+    },
+    {
+      "epoch": 0.8988395406768334,
+      "grad_norm": 0.14864306151866913,
+      "learning_rate": 2.521972524112376e-06,
+      "loss": 1.1003,
+      "step": 14794
+    },
+    {
+      "epoch": 0.8989002977094599,
+      "grad_norm": 0.26273003220558167,
+      "learning_rate": 2.5189715283070346e-06,
+      "loss": 1.1184,
+      "step": 14795
+    },
+    {
+      "epoch": 0.8989610547420864,
+      "grad_norm": 0.1566545069217682,
+      "learning_rate": 2.515972272927969e-06,
+      "loss": 1.0751,
+      "step": 14796
+    },
+    {
+      "epoch": 0.899021811774713,
+      "grad_norm": 0.19338296353816986,
+      "learning_rate": 2.512974758085124e-06,
+      "loss": 1.0438,
+      "step": 14797
+    },
+    {
+      "epoch": 0.8990825688073395,
+      "grad_norm": 0.10696569085121155,
+      "learning_rate": 2.509978983888367e-06,
+      "loss": 1.0386,
+      "step": 14798
+    },
+    {
+      "epoch": 0.8991433258399659,
+      "grad_norm": 0.1505601853132248,
+      "learning_rate": 2.506984950447505e-06,
+      "loss": 1.0853,
+      "step": 14799
+    },
+    {
+      "epoch": 0.8992040828725925,
+      "grad_norm": 0.15337173640727997,
+      "learning_rate": 2.503992657872295e-06,
+      "loss": 1.0246,
+      "step": 14800
+    },
+    {
+      "epoch": 0.899264839905219,
+      "grad_norm": 0.31898120045661926,
+      "learning_rate": 2.5010021062724097e-06,
+      "loss": 1.1141,
+      "step": 14801
+    },
+    {
+      "epoch": 0.8993255969378455,
+      "grad_norm": 0.21222637593746185,
+      "learning_rate": 2.498013295757479e-06,
+      "loss": 1.1675,
+      "step": 14802
+    },
+    {
+      "epoch": 0.8993863539704721,
+      "grad_norm": 0.2428138107061386,
+      "learning_rate": 2.4950262264370473e-06,
+      "loss": 1.2234,
+      "step": 14803
+    },
+    {
+      "epoch": 0.8994471110030986,
+      "grad_norm": 0.13729248940944672,
+      "learning_rate": 2.4920408984206166e-06,
+      "loss": 1.0823,
+      "step": 14804
+    },
+    {
+      "epoch": 0.8995078680357251,
+      "grad_norm": 0.14754195511341095,
+      "learning_rate": 2.4890573118176107e-06,
+      "loss": 1.1047,
+      "step": 14805
+    },
+    {
+      "epoch": 0.8995686250683517,
+      "grad_norm": 1.1289266347885132,
+      "learning_rate": 2.4860754667373866e-06,
+      "loss": 1.1209,
+      "step": 14806
+    },
+    {
+      "epoch": 0.8996293821009782,
+      "grad_norm": 0.31131842732429504,
+      "learning_rate": 2.4830953632892505e-06,
+      "loss": 1.288,
+      "step": 14807
+    },
+    {
+      "epoch": 0.8996901391336047,
+      "grad_norm": 0.10699814558029175,
+      "learning_rate": 2.4801170015824384e-06,
+      "loss": 1.0468,
+      "step": 14808
+    },
+    {
+      "epoch": 0.8997508961662313,
+      "grad_norm": 0.42613834142684937,
+      "learning_rate": 2.4771403817261285e-06,
+      "loss": 1.0187,
+      "step": 14809
+    },
+    {
+      "epoch": 0.8998116531988578,
+      "grad_norm": 0.13182315230369568,
+      "learning_rate": 2.4741655038294285e-06,
+      "loss": 1.0689,
+      "step": 14810
+    },
+    {
+      "epoch": 0.8998724102314843,
+      "grad_norm": 0.13865163922309875,
+      "learning_rate": 2.471192368001374e-06,
+      "loss": 1.0806,
+      "step": 14811
+    },
+    {
+      "epoch": 0.8999331672641108,
+      "grad_norm": 0.1953926384449005,
+      "learning_rate": 2.4682209743509376e-06,
+      "loss": 1.0799,
+      "step": 14812
+    },
+    {
+      "epoch": 0.8999939242967373,
+      "grad_norm": 2.2017624378204346,
+      "learning_rate": 2.4652513229870612e-06,
+      "loss": 1.0499,
+      "step": 14813
+    },
+    {
+      "epoch": 0.9000546813293638,
+      "grad_norm": 0.17436793446540833,
+      "learning_rate": 2.4622834140185846e-06,
+      "loss": 1.0822,
+      "step": 14814
+    },
+    {
+      "epoch": 0.9001154383619904,
+      "grad_norm": 0.3115310072898865,
+      "learning_rate": 2.459317247554299e-06,
+      "loss": 1.1325,
+      "step": 14815
+    },
+    {
+      "epoch": 0.9001761953946169,
+      "grad_norm": 0.13118942081928253,
+      "learning_rate": 2.4563528237029344e-06,
+      "loss": 1.0601,
+      "step": 14816
+    },
+    {
+      "epoch": 0.9002369524272434,
+      "grad_norm": 0.9630343317985535,
+      "learning_rate": 2.4533901425731365e-06,
+      "loss": 1.0786,
+      "step": 14817
+    },
+    {
+      "epoch": 0.90029770945987,
+      "grad_norm": 0.2966160476207733,
+      "learning_rate": 2.4504292042735187e-06,
+      "loss": 1.0761,
+      "step": 14818
+    },
+    {
+      "epoch": 0.9003584664924965,
+      "grad_norm": 0.12345825135707855,
+      "learning_rate": 2.4474700089126055e-06,
+      "loss": 1.0625,
+      "step": 14819
+    },
+    {
+      "epoch": 0.900419223525123,
+      "grad_norm": 0.239149272441864,
+      "learning_rate": 2.4445125565988767e-06,
+      "loss": 1.2008,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9004799805577496,
+      "grad_norm": 0.19035352766513824,
+      "learning_rate": 2.441556847440729e-06,
+      "loss": 1.095,
+      "step": 14821
+    },
+    {
+      "epoch": 0.9005407375903761,
+      "grad_norm": 0.4163861870765686,
+      "learning_rate": 2.438602881546509e-06,
+      "loss": 1.0479,
+      "step": 14822
+    },
+    {
+      "epoch": 0.9006014946230027,
+      "grad_norm": 0.09577146172523499,
+      "learning_rate": 2.4356506590244967e-06,
+      "loss": 1.0097,
+      "step": 14823
+    },
+    {
+      "epoch": 0.9006622516556292,
+      "grad_norm": 0.15774449706077576,
+      "learning_rate": 2.4327001799828995e-06,
+      "loss": 1.0381,
+      "step": 14824
+    },
+    {
+      "epoch": 0.9007230086882556,
+      "grad_norm": 0.22189322113990784,
+      "learning_rate": 2.429751444529871e-06,
+      "loss": 1.1019,
+      "step": 14825
+    },
+    {
+      "epoch": 0.9007837657208821,
+      "grad_norm": 0.1437055468559265,
+      "learning_rate": 2.426804452773501e-06,
+      "loss": 1.0409,
+      "step": 14826
+    },
+    {
+      "epoch": 0.9008445227535087,
+      "grad_norm": 0.1471676081418991,
+      "learning_rate": 2.42385920482181e-06,
+      "loss": 1.0899,
+      "step": 14827
+    },
+    {
+      "epoch": 0.9009052797861352,
+      "grad_norm": 0.2153516709804535,
+      "learning_rate": 2.420915700782744e-06,
+      "loss": 1.2429,
+      "step": 14828
+    },
+    {
+      "epoch": 0.9009660368187618,
+      "grad_norm": 0.17334984242916107,
+      "learning_rate": 2.417973940764223e-06,
+      "loss": 1.1008,
+      "step": 14829
+    },
+    {
+      "epoch": 0.9010267938513883,
+      "grad_norm": 0.18309518694877625,
+      "learning_rate": 2.415033924874066e-06,
+      "loss": 1.0947,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9010875508840148,
+      "grad_norm": 0.18143358826637268,
+      "learning_rate": 2.4120956532200367e-06,
+      "loss": 1.0275,
+      "step": 14831
+    },
+    {
+      "epoch": 0.9011483079166414,
+      "grad_norm": 0.1206001490354538,
+      "learning_rate": 2.409159125909838e-06,
+      "loss": 1.0806,
+      "step": 14832
+    },
+    {
+      "epoch": 0.9012090649492679,
+      "grad_norm": 1.0380929708480835,
+      "learning_rate": 2.4062243430511056e-06,
+      "loss": 1.0939,
+      "step": 14833
+    },
+    {
+      "epoch": 0.9012698219818944,
+      "grad_norm": 0.14141760766506195,
+      "learning_rate": 2.4032913047514204e-06,
+      "loss": 1.0364,
+      "step": 14834
+    },
+    {
+      "epoch": 0.901330579014521,
+      "grad_norm": 0.13730286061763763,
+      "learning_rate": 2.400360011118297e-06,
+      "loss": 1.0002,
+      "step": 14835
+    },
+    {
+      "epoch": 0.9013913360471475,
+      "grad_norm": 0.12602296471595764,
+      "learning_rate": 2.397430462259187e-06,
+      "loss": 1.0192,
+      "step": 14836
+    },
+    {
+      "epoch": 0.901452093079774,
+      "grad_norm": 0.2573122978210449,
+      "learning_rate": 2.3945026582814557e-06,
+      "loss": 1.1678,
+      "step": 14837
+    },
+    {
+      "epoch": 0.9015128501124006,
+      "grad_norm": 0.1748371571302414,
+      "learning_rate": 2.391576599292422e-06,
+      "loss": 1.1519,
+      "step": 14838
+    },
+    {
+      "epoch": 0.901573607145027,
+      "grad_norm": 0.18285953998565674,
+      "learning_rate": 2.388652285399362e-06,
+      "loss": 1.0914,
+      "step": 14839
+    },
+    {
+      "epoch": 0.9016343641776535,
+      "grad_norm": 0.15655595064163208,
+      "learning_rate": 2.3857297167094508e-06,
+      "loss": 1.0505,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9016951212102801,
+      "grad_norm": 0.18352477252483368,
+      "learning_rate": 2.3828088933298243e-06,
+      "loss": 1.049,
+      "step": 14841
+    },
+    {
+      "epoch": 0.9017558782429066,
+      "grad_norm": 0.22545355558395386,
+      "learning_rate": 2.379889815367542e-06,
+      "loss": 1.2236,
+      "step": 14842
+    },
+    {
+      "epoch": 0.9018166352755331,
+      "grad_norm": 0.11439885944128036,
+      "learning_rate": 2.3769724829296015e-06,
+      "loss": 1.0496,
+      "step": 14843
+    },
+    {
+      "epoch": 0.9018773923081597,
+      "grad_norm": 0.15851031243801117,
+      "learning_rate": 2.3740568961229394e-06,
+      "loss": 1.0856,
+      "step": 14844
+    },
+    {
+      "epoch": 0.9019381493407862,
+      "grad_norm": 0.14722847938537598,
+      "learning_rate": 2.3711430550544257e-06,
+      "loss": 1.0487,
+      "step": 14845
+    },
+    {
+      "epoch": 0.9019989063734127,
+      "grad_norm": 0.2087150514125824,
+      "learning_rate": 2.3682309598308747e-06,
+      "loss": 1.1448,
+      "step": 14846
+    },
+    {
+      "epoch": 0.9020596634060393,
+      "grad_norm": 0.17388825118541718,
+      "learning_rate": 2.365320610559024e-06,
+      "loss": 1.0812,
+      "step": 14847
+    },
+    {
+      "epoch": 0.9021204204386658,
+      "grad_norm": 0.10240533202886581,
+      "learning_rate": 2.3624120073455548e-06,
+      "loss": 1.0514,
+      "step": 14848
+    },
+    {
+      "epoch": 0.9021811774712923,
+      "grad_norm": 0.1334536075592041,
+      "learning_rate": 2.35950515029707e-06,
+      "loss": 1.0451,
+      "step": 14849
+    },
+    {
+      "epoch": 0.9022419345039189,
+      "grad_norm": 0.12250686436891556,
+      "learning_rate": 2.3566000395201513e-06,
+      "loss": 1.0514,
+      "step": 14850
+    },
+    {
+      "epoch": 0.9023026915365454,
+      "grad_norm": 0.19700130820274353,
+      "learning_rate": 2.3536966751212586e-06,
+      "loss": 1.0974,
+      "step": 14851
+    },
+    {
+      "epoch": 0.9023634485691718,
+      "grad_norm": 0.10672856867313385,
+      "learning_rate": 2.3507950572068228e-06,
+      "loss": 1.1092,
+      "step": 14852
+    },
+    {
+      "epoch": 0.9024242056017984,
+      "grad_norm": 0.28611481189727783,
+      "learning_rate": 2.347895185883203e-06,
+      "loss": 1.0191,
+      "step": 14853
+    },
+    {
+      "epoch": 0.9024849626344249,
+      "grad_norm": 0.18546275794506073,
+      "learning_rate": 2.344997061256693e-06,
+      "loss": 1.0589,
+      "step": 14854
+    },
+    {
+      "epoch": 0.9025457196670514,
+      "grad_norm": 0.14217063784599304,
+      "learning_rate": 2.3421006834335348e-06,
+      "loss": 1.0837,
+      "step": 14855
+    },
+    {
+      "epoch": 0.902606476699678,
+      "grad_norm": 0.18248297274112701,
+      "learning_rate": 2.3392060525198935e-06,
+      "loss": 1.0928,
+      "step": 14856
+    },
+    {
+      "epoch": 0.9026672337323045,
+      "grad_norm": 0.10684803873300552,
+      "learning_rate": 2.3363131686218563e-06,
+      "loss": 1.0422,
+      "step": 14857
+    },
+    {
+      "epoch": 0.902727990764931,
+      "grad_norm": 0.09788000583648682,
+      "learning_rate": 2.333422031845478e-06,
+      "loss": 0.9968,
+      "step": 14858
+    },
+    {
+      "epoch": 0.9027887477975576,
+      "grad_norm": 0.10721543431282043,
+      "learning_rate": 2.3305326422967178e-06,
+      "loss": 1.0339,
+      "step": 14859
+    },
+    {
+      "epoch": 0.9028495048301841,
+      "grad_norm": 0.1177595779299736,
+      "learning_rate": 2.3276450000815076e-06,
+      "loss": 1.025,
+      "step": 14860
+    },
+    {
+      "epoch": 0.9029102618628106,
+      "grad_norm": 0.27633893489837646,
+      "learning_rate": 2.324759105305685e-06,
+      "loss": 1.1643,
+      "step": 14861
+    },
+    {
+      "epoch": 0.9029710188954372,
+      "grad_norm": 0.10858333855867386,
+      "learning_rate": 2.321874958075032e-06,
+      "loss": 1.0431,
+      "step": 14862
+    },
+    {
+      "epoch": 0.9030317759280637,
+      "grad_norm": 0.15864817798137665,
+      "learning_rate": 2.318992558495281e-06,
+      "loss": 1.1289,
+      "step": 14863
+    },
+    {
+      "epoch": 0.9030925329606903,
+      "grad_norm": 0.09833846986293793,
+      "learning_rate": 2.316111906672053e-06,
+      "loss": 1.0525,
+      "step": 14864
+    },
+    {
+      "epoch": 0.9031532899933167,
+      "grad_norm": 0.14392895996570587,
+      "learning_rate": 2.3132330027109683e-06,
+      "loss": 1.0411,
+      "step": 14865
+    },
+    {
+      "epoch": 0.9032140470259432,
+      "grad_norm": 0.1342892348766327,
+      "learning_rate": 2.310355846717549e-06,
+      "loss": 1.0434,
+      "step": 14866
+    },
+    {
+      "epoch": 0.9032748040585697,
+      "grad_norm": 0.213627889752388,
+      "learning_rate": 2.3074804387972493e-06,
+      "loss": 1.0886,
+      "step": 14867
+    },
+    {
+      "epoch": 0.9033355610911963,
+      "grad_norm": 0.15513767302036285,
+      "learning_rate": 2.3046067790554794e-06,
+      "loss": 1.1355,
+      "step": 14868
+    },
+    {
+      "epoch": 0.9033963181238228,
+      "grad_norm": 0.30452588200569153,
+      "learning_rate": 2.3017348675975604e-06,
+      "loss": 1.2791,
+      "step": 14869
+    },
+    {
+      "epoch": 0.9034570751564494,
+      "grad_norm": 19.968107223510742,
+      "learning_rate": 2.2988647045287747e-06,
+      "loss": 1.105,
+      "step": 14870
+    },
+    {
+      "epoch": 0.9035178321890759,
+      "grad_norm": 0.13805511593818665,
+      "learning_rate": 2.295996289954322e-06,
+      "loss": 1.06,
+      "step": 14871
+    },
+    {
+      "epoch": 0.9035785892217024,
+      "grad_norm": 0.16649939119815826,
+      "learning_rate": 2.2931296239793455e-06,
+      "loss": 1.094,
+      "step": 14872
+    },
+    {
+      "epoch": 0.903639346254329,
+      "grad_norm": 0.10299520939588547,
+      "learning_rate": 2.290264706708922e-06,
+      "loss": 1.0535,
+      "step": 14873
+    },
+    {
+      "epoch": 0.9037001032869555,
+      "grad_norm": 0.1757773607969284,
+      "learning_rate": 2.287401538248074e-06,
+      "loss": 1.0288,
+      "step": 14874
+    },
+    {
+      "epoch": 0.903760860319582,
+      "grad_norm": 0.22568610310554504,
+      "learning_rate": 2.2845401187017333e-06,
+      "loss": 1.2472,
+      "step": 14875
+    },
+    {
+      "epoch": 0.9038216173522086,
+      "grad_norm": 0.2091546654701233,
+      "learning_rate": 2.2816804481748055e-06,
+      "loss": 1.1392,
+      "step": 14876
+    },
+    {
+      "epoch": 0.9038823743848351,
+      "grad_norm": 0.20075146853923798,
+      "learning_rate": 2.278822526772112e-06,
+      "loss": 1.0969,
+      "step": 14877
+    },
+    {
+      "epoch": 0.9039431314174615,
+      "grad_norm": 0.12191522121429443,
+      "learning_rate": 2.2759663545983966e-06,
+      "loss": 1.0571,
+      "step": 14878
+    },
+    {
+      "epoch": 0.9040038884500881,
+      "grad_norm": 0.09095070511102676,
+      "learning_rate": 2.2731119317583593e-06,
+      "loss": 0.9987,
+      "step": 14879
+    },
+    {
+      "epoch": 0.9040646454827146,
+      "grad_norm": 0.1410599797964096,
+      "learning_rate": 2.270259258356622e-06,
+      "loss": 1.0615,
+      "step": 14880
+    },
+    {
+      "epoch": 0.9041254025153411,
+      "grad_norm": 0.18395306169986725,
+      "learning_rate": 2.2674083344977615e-06,
+      "loss": 1.1135,
+      "step": 14881
+    },
+    {
+      "epoch": 0.9041861595479677,
+      "grad_norm": 0.5182232856750488,
+      "learning_rate": 2.2645591602862836e-06,
+      "loss": 1.0553,
+      "step": 14882
+    },
+    {
+      "epoch": 0.9042469165805942,
+      "grad_norm": 0.25021278858184814,
+      "learning_rate": 2.261711735826616e-06,
+      "loss": 1.1731,
+      "step": 14883
+    },
+    {
+      "epoch": 0.9043076736132207,
+      "grad_norm": 0.09032687544822693,
+      "learning_rate": 2.258866061223119e-06,
+      "loss": 1.0013,
+      "step": 14884
+    },
+    {
+      "epoch": 0.9043684306458473,
+      "grad_norm": 0.11903296411037445,
+      "learning_rate": 2.2560221365801273e-06,
+      "loss": 0.9914,
+      "step": 14885
+    },
+    {
+      "epoch": 0.9044291876784738,
+      "grad_norm": 0.13379590213298798,
+      "learning_rate": 2.2531799620018667e-06,
+      "loss": 1.0353,
+      "step": 14886
+    },
+    {
+      "epoch": 0.9044899447111003,
+      "grad_norm": 0.12906059622764587,
+      "learning_rate": 2.2503395375925275e-06,
+      "loss": 0.9949,
+      "step": 14887
+    },
+    {
+      "epoch": 0.9045507017437269,
+      "grad_norm": 0.18384777009487152,
+      "learning_rate": 2.24750086345622e-06,
+      "loss": 1.085,
+      "step": 14888
+    },
+    {
+      "epoch": 0.9046114587763534,
+      "grad_norm": 0.3484676480293274,
+      "learning_rate": 2.2446639396970006e-06,
+      "loss": 1.1427,
+      "step": 14889
+    },
+    {
+      "epoch": 0.9046722158089799,
+      "grad_norm": 0.17215284705162048,
+      "learning_rate": 2.241828766418852e-06,
+      "loss": 1.1103,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9047329728416064,
+      "grad_norm": 0.20290076732635498,
+      "learning_rate": 2.238995343725703e-06,
+      "loss": 1.0607,
+      "step": 14891
+    },
+    {
+      "epoch": 0.9047937298742329,
+      "grad_norm": 0.10693349689245224,
+      "learning_rate": 2.236163671721414e-06,
+      "loss": 1.068,
+      "step": 14892
+    },
+    {
+      "epoch": 0.9048544869068594,
+      "grad_norm": 0.1770641803741455,
+      "learning_rate": 2.233333750509775e-06,
+      "loss": 1.0573,
+      "step": 14893
+    },
+    {
+      "epoch": 0.904915243939486,
+      "grad_norm": 0.11270394921302795,
+      "learning_rate": 2.2305055801945253e-06,
+      "loss": 0.9978,
+      "step": 14894
+    },
+    {
+      "epoch": 0.9049760009721125,
+      "grad_norm": 0.17094773054122925,
+      "learning_rate": 2.2276791608793257e-06,
+      "loss": 1.1332,
+      "step": 14895
+    },
+    {
+      "epoch": 0.905036758004739,
+      "grad_norm": 0.16758529841899872,
+      "learning_rate": 2.2248544926677717e-06,
+      "loss": 1.0376,
+      "step": 14896
+    },
+    {
+      "epoch": 0.9050975150373656,
+      "grad_norm": 0.15046589076519012,
+      "learning_rate": 2.2220315756634245e-06,
+      "loss": 1.0281,
+      "step": 14897
+    },
+    {
+      "epoch": 0.9051582720699921,
+      "grad_norm": 0.1624196469783783,
+      "learning_rate": 2.2192104099697407e-06,
+      "loss": 1.0697,
+      "step": 14898
+    },
+    {
+      "epoch": 0.9052190291026186,
+      "grad_norm": 0.16985276341438293,
+      "learning_rate": 2.216390995690132e-06,
+      "loss": 1.0238,
+      "step": 14899
+    },
+    {
+      "epoch": 0.9052797861352452,
+      "grad_norm": 0.1171732023358345,
+      "learning_rate": 2.213573332927943e-06,
+      "loss": 1.0048,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9053405431678717,
+      "grad_norm": 0.17110712826251984,
+      "learning_rate": 2.2107574217864634e-06,
+      "loss": 1.0858,
+      "step": 14901
+    },
+    {
+      "epoch": 0.9054013002004982,
+      "grad_norm": 1.0572800636291504,
+      "learning_rate": 2.2079432623689113e-06,
+      "loss": 1.0112,
+      "step": 14902
+    },
+    {
+      "epoch": 0.9054620572331248,
+      "grad_norm": 1.7373898029327393,
+      "learning_rate": 2.205130854778442e-06,
+      "loss": 1.0129,
+      "step": 14903
+    },
+    {
+      "epoch": 0.9055228142657512,
+      "grad_norm": 0.08647758513689041,
+      "learning_rate": 2.2023201991181354e-06,
+      "loss": 1.0087,
+      "step": 14904
+    },
+    {
+      "epoch": 0.9055835712983777,
+      "grad_norm": 0.13178007304668427,
+      "learning_rate": 2.1995112954910134e-06,
+      "loss": 1.0393,
+      "step": 14905
+    },
+    {
+      "epoch": 0.9056443283310043,
+      "grad_norm": 0.15629468858242035,
+      "learning_rate": 2.1967041440000558e-06,
+      "loss": 1.0491,
+      "step": 14906
+    },
+    {
+      "epoch": 0.9057050853636308,
+      "grad_norm": 0.11591527611017227,
+      "learning_rate": 2.1938987447481405e-06,
+      "loss": 1.0753,
+      "step": 14907
+    },
+    {
+      "epoch": 0.9057658423962573,
+      "grad_norm": 0.27414581179618835,
+      "learning_rate": 2.1910950978381086e-06,
+      "loss": 1.1585,
+      "step": 14908
+    },
+    {
+      "epoch": 0.9058265994288839,
+      "grad_norm": 0.14085903763771057,
+      "learning_rate": 2.188293203372732e-06,
+      "loss": 1.1002,
+      "step": 14909
+    },
+    {
+      "epoch": 0.9058873564615104,
+      "grad_norm": 0.0992974266409874,
+      "learning_rate": 2.1854930614547074e-06,
+      "loss": 1.0517,
+      "step": 14910
+    },
+    {
+      "epoch": 0.905948113494137,
+      "grad_norm": 0.12857897579669952,
+      "learning_rate": 2.1826946721866747e-06,
+      "loss": 1.0681,
+      "step": 14911
+    },
+    {
+      "epoch": 0.9060088705267635,
+      "grad_norm": 0.17564108967781067,
+      "learning_rate": 2.179898035671213e-06,
+      "loss": 1.0624,
+      "step": 14912
+    },
+    {
+      "epoch": 0.90606962755939,
+      "grad_norm": 0.08895411342382431,
+      "learning_rate": 2.1771031520108343e-06,
+      "loss": 1.0147,
+      "step": 14913
+    },
+    {
+      "epoch": 0.9061303845920166,
+      "grad_norm": 0.20652857422828674,
+      "learning_rate": 2.1743100213079794e-06,
+      "loss": 1.1615,
+      "step": 14914
+    },
+    {
+      "epoch": 0.9061911416246431,
+      "grad_norm": 0.24078267812728882,
+      "learning_rate": 2.171518643665038e-06,
+      "loss": 1.0412,
+      "step": 14915
+    },
+    {
+      "epoch": 0.9062518986572696,
+      "grad_norm": 0.17702774703502655,
+      "learning_rate": 2.1687290191843178e-06,
+      "loss": 1.0171,
+      "step": 14916
+    },
+    {
+      "epoch": 0.906312655689896,
+      "grad_norm": 0.13435055315494537,
+      "learning_rate": 2.165941147968098e-06,
+      "loss": 1.0695,
+      "step": 14917
+    },
+    {
+      "epoch": 0.9063734127225226,
+      "grad_norm": 0.11238982528448105,
+      "learning_rate": 2.163155030118541e-06,
+      "loss": 1.016,
+      "step": 14918
+    },
+    {
+      "epoch": 0.9064341697551491,
+      "grad_norm": 0.11210320889949799,
+      "learning_rate": 2.160370665737782e-06,
+      "loss": 1.0352,
+      "step": 14919
+    },
+    {
+      "epoch": 0.9064949267877757,
+      "grad_norm": 0.13768525421619415,
+      "learning_rate": 2.157588054927884e-06,
+      "loss": 1.0964,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9065556838204022,
+      "grad_norm": 0.11701498925685883,
+      "learning_rate": 2.1548071977908423e-06,
+      "loss": 1.0417,
+      "step": 14921
+    },
+    {
+      "epoch": 0.9066164408530287,
+      "grad_norm": 0.10916713625192642,
+      "learning_rate": 2.152028094428593e-06,
+      "loss": 1.0241,
+      "step": 14922
+    },
+    {
+      "epoch": 0.9066771978856553,
+      "grad_norm": 1.759683609008789,
+      "learning_rate": 2.1492507449430098e-06,
+      "loss": 1.0213,
+      "step": 14923
+    },
+    {
+      "epoch": 0.9067379549182818,
+      "grad_norm": 0.19697317481040955,
+      "learning_rate": 2.1464751494358837e-06,
+      "loss": 1.1385,
+      "step": 14924
+    },
+    {
+      "epoch": 0.9067987119509083,
+      "grad_norm": 0.11534097790718079,
+      "learning_rate": 2.1437013080089608e-06,
+      "loss": 1.0203,
+      "step": 14925
+    },
+    {
+      "epoch": 0.9068594689835349,
+      "grad_norm": 0.12873831391334534,
+      "learning_rate": 2.1409292207639096e-06,
+      "loss": 1.0734,
+      "step": 14926
+    },
+    {
+      "epoch": 0.9069202260161614,
+      "grad_norm": 0.1048264354467392,
+      "learning_rate": 2.138158887802355e-06,
+      "loss": 1.0258,
+      "step": 14927
+    },
+    {
+      "epoch": 0.9069809830487879,
+      "grad_norm": 0.16935966908931732,
+      "learning_rate": 2.1353903092258374e-06,
+      "loss": 1.1621,
+      "step": 14928
+    },
+    {
+      "epoch": 0.9070417400814145,
+      "grad_norm": 0.13281705975532532,
+      "learning_rate": 2.132623485135843e-06,
+      "loss": 1.0636,
+      "step": 14929
+    },
+    {
+      "epoch": 0.9071024971140409,
+      "grad_norm": 0.15051071345806122,
+      "learning_rate": 2.1298584156337896e-06,
+      "loss": 1.0907,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9071632541466674,
+      "grad_norm": 0.16882653534412384,
+      "learning_rate": 2.1270951008210192e-06,
+      "loss": 1.1023,
+      "step": 14931
+    },
+    {
+      "epoch": 0.907224011179294,
+      "grad_norm": 0.1358702927827835,
+      "learning_rate": 2.1243335407988396e-06,
+      "loss": 1.038,
+      "step": 14932
+    },
+    {
+      "epoch": 0.9072847682119205,
+      "grad_norm": 0.13505752384662628,
+      "learning_rate": 2.121573735668464e-06,
+      "loss": 1.0085,
+      "step": 14933
+    },
+    {
+      "epoch": 0.907345525244547,
+      "grad_norm": 0.1285618245601654,
+      "learning_rate": 2.1188156855310613e-06,
+      "loss": 1.0184,
+      "step": 14934
+    },
+    {
+      "epoch": 0.9074062822771736,
+      "grad_norm": 0.15828286111354828,
+      "learning_rate": 2.1160593904877235e-06,
+      "loss": 1.1131,
+      "step": 14935
+    },
+    {
+      "epoch": 0.9074670393098001,
+      "grad_norm": 0.12114787101745605,
+      "learning_rate": 2.113304850639486e-06,
+      "loss": 1.0311,
+      "step": 14936
+    },
+    {
+      "epoch": 0.9075277963424266,
+      "grad_norm": 1.705637812614441,
+      "learning_rate": 2.1105520660873123e-06,
+      "loss": 1.0907,
+      "step": 14937
+    },
+    {
+      "epoch": 0.9075885533750532,
+      "grad_norm": 0.11159954220056534,
+      "learning_rate": 2.107801036932111e-06,
+      "loss": 1.0591,
+      "step": 14938
+    },
+    {
+      "epoch": 0.9076493104076797,
+      "grad_norm": 0.13896577060222626,
+      "learning_rate": 2.1050517632747178e-06,
+      "loss": 1.0834,
+      "step": 14939
+    },
+    {
+      "epoch": 0.9077100674403062,
+      "grad_norm": 0.16339844465255737,
+      "learning_rate": 2.1023042452159136e-06,
+      "loss": 1.1279,
+      "step": 14940
+    },
+    {
+      "epoch": 0.9077708244729328,
+      "grad_norm": 0.14502331614494324,
+      "learning_rate": 2.0995584828564064e-06,
+      "loss": 1.0893,
+      "step": 14941
+    },
+    {
+      "epoch": 0.9078315815055593,
+      "grad_norm": 0.11531216651201248,
+      "learning_rate": 2.096814476296832e-06,
+      "loss": 1.0076,
+      "step": 14942
+    },
+    {
+      "epoch": 0.9078923385381859,
+      "grad_norm": 0.17742054164409637,
+      "learning_rate": 2.094072225637794e-06,
+      "loss": 1.165,
+      "step": 14943
+    },
+    {
+      "epoch": 0.9079530955708123,
+      "grad_norm": 0.15709777176380157,
+      "learning_rate": 2.091331730979801e-06,
+      "loss": 1.0494,
+      "step": 14944
+    },
+    {
+      "epoch": 0.9080138526034388,
+      "grad_norm": 0.20858605206012726,
+      "learning_rate": 2.0885929924232994e-06,
+      "loss": 1.0875,
+      "step": 14945
+    },
+    {
+      "epoch": 0.9080746096360653,
+      "grad_norm": 0.14397738873958588,
+      "learning_rate": 2.0858560100686763e-06,
+      "loss": 1.067,
+      "step": 14946
+    },
+    {
+      "epoch": 0.9081353666686919,
+      "grad_norm": 0.12032930552959442,
+      "learning_rate": 2.0831207840162626e-06,
+      "loss": 1.0613,
+      "step": 14947
+    },
+    {
+      "epoch": 0.9081961237013184,
+      "grad_norm": 0.14941547811031342,
+      "learning_rate": 2.0803873143663277e-06,
+      "loss": 1.0364,
+      "step": 14948
+    },
+    {
+      "epoch": 0.908256880733945,
+      "grad_norm": 0.10667446255683899,
+      "learning_rate": 2.0776556012190527e-06,
+      "loss": 1.0121,
+      "step": 14949
+    },
+    {
+      "epoch": 0.9083176377665715,
+      "grad_norm": 0.3162420988082886,
+      "learning_rate": 2.07492564467458e-06,
+      "loss": 1.1724,
+      "step": 14950
+    },
+    {
+      "epoch": 0.908378394799198,
+      "grad_norm": 0.1320541799068451,
+      "learning_rate": 2.0721974448329683e-06,
+      "loss": 1.0384,
+      "step": 14951
+    },
+    {
+      "epoch": 0.9084391518318246,
+      "grad_norm": 0.27250608801841736,
+      "learning_rate": 2.069471001794221e-06,
+      "loss": 1.1995,
+      "step": 14952
+    },
+    {
+      "epoch": 0.9084999088644511,
+      "grad_norm": 0.09595417976379395,
+      "learning_rate": 2.06674631565828e-06,
+      "loss": 1.0296,
+      "step": 14953
+    },
+    {
+      "epoch": 0.9085606658970776,
+      "grad_norm": 0.1230664774775505,
+      "learning_rate": 2.0640233865250215e-06,
+      "loss": 1.044,
+      "step": 14954
+    },
+    {
+      "epoch": 0.9086214229297042,
+      "grad_norm": 0.10095830261707306,
+      "learning_rate": 2.0613022144942482e-06,
+      "loss": 1.0377,
+      "step": 14955
+    },
+    {
+      "epoch": 0.9086821799623307,
+      "grad_norm": 1.5195826292037964,
+      "learning_rate": 2.058582799665709e-06,
+      "loss": 1.0577,
+      "step": 14956
+    },
+    {
+      "epoch": 0.9087429369949571,
+      "grad_norm": 0.12075446546077728,
+      "learning_rate": 2.055865142139085e-06,
+      "loss": 1.0343,
+      "step": 14957
+    },
+    {
+      "epoch": 0.9088036940275837,
+      "grad_norm": 0.10185153037309647,
+      "learning_rate": 2.053149242013991e-06,
+      "loss": 0.9888,
+      "step": 14958
+    },
+    {
+      "epoch": 0.9088644510602102,
+      "grad_norm": 0.1275072693824768,
+      "learning_rate": 2.05043509938998e-06,
+      "loss": 1.0409,
+      "step": 14959
+    },
+    {
+      "epoch": 0.9089252080928367,
+      "grad_norm": 0.19518202543258667,
+      "learning_rate": 2.04772271436654e-06,
+      "loss": 1.0735,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9089859651254633,
+      "grad_norm": 0.10764231532812119,
+      "learning_rate": 2.045012087043091e-06,
+      "loss": 1.0423,
+      "step": 14961
+    },
+    {
+      "epoch": 0.9090467221580898,
+      "grad_norm": 0.2531062960624695,
+      "learning_rate": 2.0423032175189983e-06,
+      "loss": 1.1535,
+      "step": 14962
+    },
+    {
+      "epoch": 0.9091074791907163,
+      "grad_norm": 0.25562238693237305,
+      "learning_rate": 2.0395961058935377e-06,
+      "loss": 1.0257,
+      "step": 14963
+    },
+    {
+      "epoch": 0.9091682362233429,
+      "grad_norm": 0.12102354317903519,
+      "learning_rate": 2.0368907522659686e-06,
+      "loss": 1.0686,
+      "step": 14964
+    },
+    {
+      "epoch": 0.9092289932559694,
+      "grad_norm": 0.13641558587551117,
+      "learning_rate": 2.034187156735434e-06,
+      "loss": 1.0498,
+      "step": 14965
+    },
+    {
+      "epoch": 0.9092897502885959,
+      "grad_norm": 0.1522539108991623,
+      "learning_rate": 2.0314853194010385e-06,
+      "loss": 1.1173,
+      "step": 14966
+    },
+    {
+      "epoch": 0.9093505073212225,
+      "grad_norm": 0.14811113476753235,
+      "learning_rate": 2.0287852403618247e-06,
+      "loss": 1.0094,
+      "step": 14967
+    },
+    {
+      "epoch": 0.909411264353849,
+      "grad_norm": 0.1169791966676712,
+      "learning_rate": 2.026086919716752e-06,
+      "loss": 1.0588,
+      "step": 14968
+    },
+    {
+      "epoch": 0.9094720213864755,
+      "grad_norm": 0.11501649022102356,
+      "learning_rate": 2.023390357564742e-06,
+      "loss": 1.0412,
+      "step": 14969
+    },
+    {
+      "epoch": 0.909532778419102,
+      "grad_norm": 0.12850110232830048,
+      "learning_rate": 2.020695554004637e-06,
+      "loss": 1.0739,
+      "step": 14970
+    },
+    {
+      "epoch": 0.9095935354517285,
+      "grad_norm": 0.1282275766134262,
+      "learning_rate": 2.0180025091352086e-06,
+      "loss": 1.0616,
+      "step": 14971
+    },
+    {
+      "epoch": 0.909654292484355,
+      "grad_norm": 0.28371360898017883,
+      "learning_rate": 2.0153112230551717e-06,
+      "loss": 1.1005,
+      "step": 14972
+    },
+    {
+      "epoch": 0.9097150495169816,
+      "grad_norm": 0.14681534469127655,
+      "learning_rate": 2.0126216958631707e-06,
+      "loss": 1.0195,
+      "step": 14973
+    },
+    {
+      "epoch": 0.9097758065496081,
+      "grad_norm": 0.13920894265174866,
+      "learning_rate": 2.009933927657803e-06,
+      "loss": 1.0926,
+      "step": 14974
+    },
+    {
+      "epoch": 0.9098365635822346,
+      "grad_norm": 0.09868458658456802,
+      "learning_rate": 2.0072479185375805e-06,
+      "loss": 1.0385,
+      "step": 14975
+    },
+    {
+      "epoch": 0.9098973206148612,
+      "grad_norm": 0.13641227781772614,
+      "learning_rate": 2.004563668600967e-06,
+      "loss": 1.0784,
+      "step": 14976
+    },
+    {
+      "epoch": 0.9099580776474877,
+      "grad_norm": 0.19605377316474915,
+      "learning_rate": 2.0018811779463464e-06,
+      "loss": 1.0987,
+      "step": 14977
+    },
+    {
+      "epoch": 0.9100188346801142,
+      "grad_norm": 0.10820509493350983,
+      "learning_rate": 1.9992004466720503e-06,
+      "loss": 1.0334,
+      "step": 14978
+    },
+    {
+      "epoch": 0.9100795917127408,
+      "grad_norm": 0.09274052083492279,
+      "learning_rate": 1.9965214748763395e-06,
+      "loss": 1.0015,
+      "step": 14979
+    },
+    {
+      "epoch": 0.9101403487453673,
+      "grad_norm": 0.22717702388763428,
+      "learning_rate": 1.993844262657413e-06,
+      "loss": 1.0885,
+      "step": 14980
+    },
+    {
+      "epoch": 0.9102011057779938,
+      "grad_norm": 0.13725969195365906,
+      "learning_rate": 1.991168810113403e-06,
+      "loss": 1.038,
+      "step": 14981
+    },
+    {
+      "epoch": 0.9102618628106204,
+      "grad_norm": 0.13086746633052826,
+      "learning_rate": 1.9884951173423814e-06,
+      "loss": 1.0778,
+      "step": 14982
+    },
+    {
+      "epoch": 0.9103226198432468,
+      "grad_norm": 0.09827271103858948,
+      "learning_rate": 1.9858231844423536e-06,
+      "loss": 1.0303,
+      "step": 14983
+    },
+    {
+      "epoch": 0.9103833768758733,
+      "grad_norm": 0.10377005487680435,
+      "learning_rate": 1.9831530115112575e-06,
+      "loss": 1.0548,
+      "step": 14984
+    },
+    {
+      "epoch": 0.9104441339084999,
+      "grad_norm": 0.1241227388381958,
+      "learning_rate": 1.98048459864697e-06,
+      "loss": 1.0995,
+      "step": 14985
+    },
+    {
+      "epoch": 0.9105048909411264,
+      "grad_norm": 0.11876312643289566,
+      "learning_rate": 1.9778179459472968e-06,
+      "loss": 1.0295,
+      "step": 14986
+    },
+    {
+      "epoch": 0.910565647973753,
+      "grad_norm": 0.17811179161071777,
+      "learning_rate": 1.9751530535099926e-06,
+      "loss": 1.1226,
+      "step": 14987
+    },
+    {
+      "epoch": 0.9106264050063795,
+      "grad_norm": 0.11316623538732529,
+      "learning_rate": 1.972489921432735e-06,
+      "loss": 1.0308,
+      "step": 14988
+    },
+    {
+      "epoch": 0.910687162039006,
+      "grad_norm": 0.24598747491836548,
+      "learning_rate": 1.969828549813141e-06,
+      "loss": 1.0775,
+      "step": 14989
+    },
+    {
+      "epoch": 0.9107479190716326,
+      "grad_norm": 0.6849556565284729,
+      "learning_rate": 1.9671689387487756e-06,
+      "loss": 1.1127,
+      "step": 14990
+    },
+    {
+      "epoch": 0.9108086761042591,
+      "grad_norm": 2.2096712589263916,
+      "learning_rate": 1.9645110883371064e-06,
+      "loss": 1.0574,
+      "step": 14991
+    },
+    {
+      "epoch": 0.9108694331368856,
+      "grad_norm": 0.14541193842887878,
+      "learning_rate": 1.9618549986755717e-06,
+      "loss": 1.0415,
+      "step": 14992
+    },
+    {
+      "epoch": 0.9109301901695122,
+      "grad_norm": 0.1036265566945076,
+      "learning_rate": 1.959200669861522e-06,
+      "loss": 1.0633,
+      "step": 14993
+    },
+    {
+      "epoch": 0.9109909472021387,
+      "grad_norm": 0.17369700968265533,
+      "learning_rate": 1.956548101992267e-06,
+      "loss": 1.0696,
+      "step": 14994
+    },
+    {
+      "epoch": 0.9110517042347652,
+      "grad_norm": 0.13502176105976105,
+      "learning_rate": 1.9538972951650245e-06,
+      "loss": 1.0682,
+      "step": 14995
+    },
+    {
+      "epoch": 0.9111124612673916,
+      "grad_norm": 0.1685733199119568,
+      "learning_rate": 1.9512482494769613e-06,
+      "loss": 1.0532,
+      "step": 14996
+    },
+    {
+      "epoch": 0.9111732183000182,
+      "grad_norm": 0.501579999923706,
+      "learning_rate": 1.948600965025188e-06,
+      "loss": 1.0255,
+      "step": 14997
+    },
+    {
+      "epoch": 0.9112339753326447,
+      "grad_norm": 0.20074892044067383,
+      "learning_rate": 1.9459554419067273e-06,
+      "loss": 1.1113,
+      "step": 14998
+    },
+    {
+      "epoch": 0.9112947323652713,
+      "grad_norm": 0.13688123226165771,
+      "learning_rate": 1.9433116802185626e-06,
+      "loss": 1.0556,
+      "step": 14999
+    },
+    {
+      "epoch": 0.9113554893978978,
+      "grad_norm": 0.21024824678897858,
+      "learning_rate": 1.940669680057594e-06,
+      "loss": 1.0449,
+      "step": 15000
+    },
+    {
+      "epoch": 0.9114162464305243,
+      "grad_norm": 0.13234519958496094,
+      "learning_rate": 1.9380294415206723e-06,
+      "loss": 1.0707,
+      "step": 15001
+    },
+    {
+      "epoch": 0.9114770034631509,
+      "grad_norm": 0.10887365788221359,
+      "learning_rate": 1.935390964704564e-06,
+      "loss": 1.0696,
+      "step": 15002
+    },
+    {
+      "epoch": 0.9115377604957774,
+      "grad_norm": 0.1296595185995102,
+      "learning_rate": 1.932754249705998e-06,
+      "loss": 1.0096,
+      "step": 15003
+    },
+    {
+      "epoch": 0.9115985175284039,
+      "grad_norm": 0.12822416424751282,
+      "learning_rate": 1.930119296621613e-06,
+      "loss": 1.0501,
+      "step": 15004
+    },
+    {
+      "epoch": 0.9116592745610305,
+      "grad_norm": 0.27920305728912354,
+      "learning_rate": 1.9274861055479986e-06,
+      "loss": 1.1158,
+      "step": 15005
+    },
+    {
+      "epoch": 0.911720031593657,
+      "grad_norm": 0.1370418816804886,
+      "learning_rate": 1.9248546765816722e-06,
+      "loss": 1.0468,
+      "step": 15006
+    },
+    {
+      "epoch": 0.9117807886262835,
+      "grad_norm": 0.13695718348026276,
+      "learning_rate": 1.9222250098190897e-06,
+      "loss": 1.0816,
+      "step": 15007
+    },
+    {
+      "epoch": 0.9118415456589101,
+      "grad_norm": 0.1371576189994812,
+      "learning_rate": 1.9195971053566464e-06,
+      "loss": 1.115,
+      "step": 15008
+    },
+    {
+      "epoch": 0.9119023026915365,
+      "grad_norm": 0.1741715967655182,
+      "learning_rate": 1.916970963290654e-06,
+      "loss": 1.0969,
+      "step": 15009
+    },
+    {
+      "epoch": 0.911963059724163,
+      "grad_norm": 0.13772989809513092,
+      "learning_rate": 1.9143465837174023e-06,
+      "loss": 1.0289,
+      "step": 15010
+    },
+    {
+      "epoch": 0.9120238167567896,
+      "grad_norm": 0.1589927226305008,
+      "learning_rate": 1.911723966733059e-06,
+      "loss": 1.0327,
+      "step": 15011
+    },
+    {
+      "epoch": 0.9120845737894161,
+      "grad_norm": 0.0933261290192604,
+      "learning_rate": 1.9091031124337744e-06,
+      "loss": 1.0767,
+      "step": 15012
+    },
+    {
+      "epoch": 0.9121453308220426,
+      "grad_norm": 0.19839192926883698,
+      "learning_rate": 1.906484020915611e-06,
+      "loss": 0.9879,
+      "step": 15013
+    },
+    {
+      "epoch": 0.9122060878546692,
+      "grad_norm": 0.1780577450990677,
+      "learning_rate": 1.903866692274564e-06,
+      "loss": 1.1104,
+      "step": 15014
+    },
+    {
+      "epoch": 0.9122668448872957,
+      "grad_norm": 0.1848917007446289,
+      "learning_rate": 1.9012511266065847e-06,
+      "loss": 1.0603,
+      "step": 15015
+    },
+    {
+      "epoch": 0.9123276019199222,
+      "grad_norm": 0.1053280457854271,
+      "learning_rate": 1.8986373240075462e-06,
+      "loss": 1.041,
+      "step": 15016
+    },
+    {
+      "epoch": 0.9123883589525488,
+      "grad_norm": 0.10170558094978333,
+      "learning_rate": 1.8960252845732551e-06,
+      "loss": 1.0641,
+      "step": 15017
+    },
+    {
+      "epoch": 0.9124491159851753,
+      "grad_norm": 0.1257053166627884,
+      "learning_rate": 1.8934150083994518e-06,
+      "loss": 1.0007,
+      "step": 15018
+    },
+    {
+      "epoch": 0.9125098730178018,
+      "grad_norm": 0.22070614993572235,
+      "learning_rate": 1.890806495581815e-06,
+      "loss": 1.0581,
+      "step": 15019
+    },
+    {
+      "epoch": 0.9125706300504284,
+      "grad_norm": 0.1336812525987625,
+      "learning_rate": 1.888199746215974e-06,
+      "loss": 1.0686,
+      "step": 15020
+    },
+    {
+      "epoch": 0.9126313870830549,
+      "grad_norm": 0.1422586888074875,
+      "learning_rate": 1.8855947603974688e-06,
+      "loss": 1.0763,
+      "step": 15021
+    },
+    {
+      "epoch": 0.9126921441156813,
+      "grad_norm": 0.12930350005626678,
+      "learning_rate": 1.8829915382217899e-06,
+      "loss": 1.0757,
+      "step": 15022
+    },
+    {
+      "epoch": 0.9127529011483079,
+      "grad_norm": 0.21522289514541626,
+      "learning_rate": 1.88039007978435e-06,
+      "loss": 1.1454,
+      "step": 15023
+    },
+    {
+      "epoch": 0.9128136581809344,
+      "grad_norm": 0.1586683988571167,
+      "learning_rate": 1.8777903851805222e-06,
+      "loss": 1.1496,
+      "step": 15024
+    },
+    {
+      "epoch": 0.9128744152135609,
+      "grad_norm": 0.16037622094154358,
+      "learning_rate": 1.875192454505581e-06,
+      "loss": 1.0363,
+      "step": 15025
+    },
+    {
+      "epoch": 0.9129351722461875,
+      "grad_norm": 0.17235533893108368,
+      "learning_rate": 1.8725962878547721e-06,
+      "loss": 1.0429,
+      "step": 15026
+    },
+    {
+      "epoch": 0.912995929278814,
+      "grad_norm": 0.2739880084991455,
+      "learning_rate": 1.8700018853232416e-06,
+      "loss": 1.1995,
+      "step": 15027
+    },
+    {
+      "epoch": 0.9130566863114405,
+      "grad_norm": 0.10850818455219269,
+      "learning_rate": 1.8674092470060966e-06,
+      "loss": 1.0333,
+      "step": 15028
+    },
+    {
+      "epoch": 0.9131174433440671,
+      "grad_norm": 0.13918934762477875,
+      "learning_rate": 1.8648183729983726e-06,
+      "loss": 1.0514,
+      "step": 15029
+    },
+    {
+      "epoch": 0.9131782003766936,
+      "grad_norm": 0.20980829000473022,
+      "learning_rate": 1.8622292633950266e-06,
+      "loss": 1.1437,
+      "step": 15030
+    },
+    {
+      "epoch": 0.9132389574093202,
+      "grad_norm": 0.15558724105358124,
+      "learning_rate": 1.859641918290983e-06,
+      "loss": 1.08,
+      "step": 15031
+    },
+    {
+      "epoch": 0.9132997144419467,
+      "grad_norm": 0.13440930843353271,
+      "learning_rate": 1.8570563377810657e-06,
+      "loss": 1.073,
+      "step": 15032
+    },
+    {
+      "epoch": 0.9133604714745732,
+      "grad_norm": 0.12515634298324585,
+      "learning_rate": 1.8544725219600544e-06,
+      "loss": 1.0554,
+      "step": 15033
+    },
+    {
+      "epoch": 0.9134212285071998,
+      "grad_norm": 0.13739170134067535,
+      "learning_rate": 1.8518904709226626e-06,
+      "loss": 1.1265,
+      "step": 15034
+    },
+    {
+      "epoch": 0.9134819855398262,
+      "grad_norm": 0.11433909088373184,
+      "learning_rate": 1.8493101847635196e-06,
+      "loss": 1.0616,
+      "step": 15035
+    },
+    {
+      "epoch": 0.9135427425724527,
+      "grad_norm": 0.12163405865430832,
+      "learning_rate": 1.846731663577228e-06,
+      "loss": 1.1028,
+      "step": 15036
+    },
+    {
+      "epoch": 0.9136034996050793,
+      "grad_norm": 0.1588326394557953,
+      "learning_rate": 1.8441549074583008e-06,
+      "loss": 1.0424,
+      "step": 15037
+    },
+    {
+      "epoch": 0.9136642566377058,
+      "grad_norm": 0.10011840611696243,
+      "learning_rate": 1.8415799165011793e-06,
+      "loss": 1.0136,
+      "step": 15038
+    },
+    {
+      "epoch": 0.9137250136703323,
+      "grad_norm": 0.1284933090209961,
+      "learning_rate": 1.8390066908002545e-06,
+      "loss": 1.0365,
+      "step": 15039
+    },
+    {
+      "epoch": 0.9137857707029589,
+      "grad_norm": 0.3652207851409912,
+      "learning_rate": 1.83643523044984e-06,
+      "loss": 1.1992,
+      "step": 15040
+    },
+    {
+      "epoch": 0.9138465277355854,
+      "grad_norm": 0.11399808526039124,
+      "learning_rate": 1.8338655355442104e-06,
+      "loss": 1.0214,
+      "step": 15041
+    },
+    {
+      "epoch": 0.9139072847682119,
+      "grad_norm": 0.1427631676197052,
+      "learning_rate": 1.8312976061775512e-06,
+      "loss": 1.1073,
+      "step": 15042
+    },
+    {
+      "epoch": 0.9139680418008385,
+      "grad_norm": 0.18603037297725677,
+      "learning_rate": 1.8287314424439872e-06,
+      "loss": 1.1403,
+      "step": 15043
+    },
+    {
+      "epoch": 0.914028798833465,
+      "grad_norm": 0.2763344645500183,
+      "learning_rate": 1.8261670444375934e-06,
+      "loss": 1.2671,
+      "step": 15044
+    },
+    {
+      "epoch": 0.9140895558660915,
+      "grad_norm": 0.18840569257736206,
+      "learning_rate": 1.8236044122523389e-06,
+      "loss": 1.0861,
+      "step": 15045
+    },
+    {
+      "epoch": 0.9141503128987181,
+      "grad_norm": 0.19211970269680023,
+      "learning_rate": 1.8210435459821873e-06,
+      "loss": 1.0524,
+      "step": 15046
+    },
+    {
+      "epoch": 0.9142110699313446,
+      "grad_norm": 0.11699900031089783,
+      "learning_rate": 1.818484445720997e-06,
+      "loss": 1.0632,
+      "step": 15047
+    },
+    {
+      "epoch": 0.9142718269639711,
+      "grad_norm": 0.17466813325881958,
+      "learning_rate": 1.815927111562571e-06,
+      "loss": 1.1245,
+      "step": 15048
+    },
+    {
+      "epoch": 0.9143325839965976,
+      "grad_norm": 0.11343840509653091,
+      "learning_rate": 1.813371543600656e-06,
+      "loss": 1.0551,
+      "step": 15049
+    },
+    {
+      "epoch": 0.9143933410292241,
+      "grad_norm": 0.10502791404724121,
+      "learning_rate": 1.8108177419289162e-06,
+      "loss": 1.0623,
+      "step": 15050
+    },
+    {
+      "epoch": 0.9144540980618506,
+      "grad_norm": 0.2084721028804779,
+      "learning_rate": 1.808265706640966e-06,
+      "loss": 1.1051,
+      "step": 15051
+    },
+    {
+      "epoch": 0.9145148550944772,
+      "grad_norm": 0.1667078286409378,
+      "learning_rate": 1.805715437830352e-06,
+      "loss": 1.0962,
+      "step": 15052
+    },
+    {
+      "epoch": 0.9145756121271037,
+      "grad_norm": 0.1975841522216797,
+      "learning_rate": 1.8031669355905556e-06,
+      "loss": 1.0915,
+      "step": 15053
+    },
+    {
+      "epoch": 0.9146363691597302,
+      "grad_norm": 0.18476633727550507,
+      "learning_rate": 1.8006202000149963e-06,
+      "loss": 1.0782,
+      "step": 15054
+    },
+    {
+      "epoch": 0.9146971261923568,
+      "grad_norm": 0.5259315371513367,
+      "learning_rate": 1.7980752311970106e-06,
+      "loss": 1.0697,
+      "step": 15055
+    },
+    {
+      "epoch": 0.9147578832249833,
+      "grad_norm": 0.8986549377441406,
+      "learning_rate": 1.795532029229896e-06,
+      "loss": 1.1258,
+      "step": 15056
+    },
+    {
+      "epoch": 0.9148186402576098,
+      "grad_norm": 0.11824187636375427,
+      "learning_rate": 1.7929905942068836e-06,
+      "loss": 1.0232,
+      "step": 15057
+    },
+    {
+      "epoch": 0.9148793972902364,
+      "grad_norm": 0.14921750128269196,
+      "learning_rate": 1.79045092622111e-06,
+      "loss": 1.0841,
+      "step": 15058
+    },
+    {
+      "epoch": 0.9149401543228629,
+      "grad_norm": 0.31009456515312195,
+      "learning_rate": 1.7879130253656783e-06,
+      "loss": 1.2014,
+      "step": 15059
+    },
+    {
+      "epoch": 0.9150009113554894,
+      "grad_norm": 0.09869231283664703,
+      "learning_rate": 1.7853768917336144e-06,
+      "loss": 0.998,
+      "step": 15060
+    },
+    {
+      "epoch": 0.915061668388116,
+      "grad_norm": 0.17841672897338867,
+      "learning_rate": 1.7828425254178715e-06,
+      "loss": 1.159,
+      "step": 15061
+    },
+    {
+      "epoch": 0.9151224254207424,
+      "grad_norm": 0.20998667180538177,
+      "learning_rate": 1.7803099265113643e-06,
+      "loss": 1.0843,
+      "step": 15062
+    },
+    {
+      "epoch": 0.9151831824533689,
+      "grad_norm": 0.10971344262361526,
+      "learning_rate": 1.7777790951069185e-06,
+      "loss": 1.0163,
+      "step": 15063
+    },
+    {
+      "epoch": 0.9152439394859955,
+      "grad_norm": 0.4894367754459381,
+      "learning_rate": 1.7752500312973042e-06,
+      "loss": 1.1286,
+      "step": 15064
+    },
+    {
+      "epoch": 0.915304696518622,
+      "grad_norm": 11.082429885864258,
+      "learning_rate": 1.7727227351752195e-06,
+      "loss": 1.0615,
+      "step": 15065
+    },
+    {
+      "epoch": 0.9153654535512485,
+      "grad_norm": 0.2113294005393982,
+      "learning_rate": 1.7701972068332962e-06,
+      "loss": 1.0273,
+      "step": 15066
+    },
+    {
+      "epoch": 0.9154262105838751,
+      "grad_norm": 0.21624793112277985,
+      "learning_rate": 1.7676734463641209e-06,
+      "loss": 1.1947,
+      "step": 15067
+    },
+    {
+      "epoch": 0.9154869676165016,
+      "grad_norm": 0.12739229202270508,
+      "learning_rate": 1.7651514538602033e-06,
+      "loss": 1.0036,
+      "step": 15068
+    },
+    {
+      "epoch": 0.9155477246491281,
+      "grad_norm": 0.47212859988212585,
+      "learning_rate": 1.7626312294139801e-06,
+      "loss": 1.293,
+      "step": 15069
+    },
+    {
+      "epoch": 0.9156084816817547,
+      "grad_norm": 0.14549385011196136,
+      "learning_rate": 1.7601127731178335e-06,
+      "loss": 1.068,
+      "step": 15070
+    },
+    {
+      "epoch": 0.9156692387143812,
+      "grad_norm": 0.09311727434396744,
+      "learning_rate": 1.7575960850640783e-06,
+      "loss": 1.0015,
+      "step": 15071
+    },
+    {
+      "epoch": 0.9157299957470078,
+      "grad_norm": 0.12194705754518509,
+      "learning_rate": 1.7550811653449628e-06,
+      "loss": 1.0511,
+      "step": 15072
+    },
+    {
+      "epoch": 0.9157907527796343,
+      "grad_norm": 1.0025709867477417,
+      "learning_rate": 1.7525680140526746e-06,
+      "loss": 1.0395,
+      "step": 15073
+    },
+    {
+      "epoch": 0.9158515098122608,
+      "grad_norm": 0.13267354667186737,
+      "learning_rate": 1.7500566312793287e-06,
+      "loss": 1.1499,
+      "step": 15074
+    },
+    {
+      "epoch": 0.9159122668448872,
+      "grad_norm": 0.30322837829589844,
+      "learning_rate": 1.747547017116985e-06,
+      "loss": 1.1591,
+      "step": 15075
+    },
+    {
+      "epoch": 0.9159730238775138,
+      "grad_norm": 0.15233691036701202,
+      "learning_rate": 1.7450391716576308e-06,
+      "loss": 1.0302,
+      "step": 15076
+    },
+    {
+      "epoch": 0.9160337809101403,
+      "grad_norm": 0.8492075204849243,
+      "learning_rate": 1.7425330949931985e-06,
+      "loss": 1.0863,
+      "step": 15077
+    },
+    {
+      "epoch": 0.9160945379427669,
+      "grad_norm": 0.4539567828178406,
+      "learning_rate": 1.7400287872155362e-06,
+      "loss": 1.1518,
+      "step": 15078
+    },
+    {
+      "epoch": 0.9161552949753934,
+      "grad_norm": 0.16213977336883545,
+      "learning_rate": 1.7375262484164546e-06,
+      "loss": 1.1028,
+      "step": 15079
+    },
+    {
+      "epoch": 0.9162160520080199,
+      "grad_norm": 0.9916358590126038,
+      "learning_rate": 1.735025478687674e-06,
+      "loss": 1.0673,
+      "step": 15080
+    },
+    {
+      "epoch": 0.9162768090406465,
+      "grad_norm": 0.1475456804037094,
+      "learning_rate": 1.7325264781208606e-06,
+      "loss": 1.0185,
+      "step": 15081
+    },
+    {
+      "epoch": 0.916337566073273,
+      "grad_norm": 0.2631661593914032,
+      "learning_rate": 1.7300292468076186e-06,
+      "loss": 1.0192,
+      "step": 15082
+    },
+    {
+      "epoch": 0.9163983231058995,
+      "grad_norm": 0.8500600457191467,
+      "learning_rate": 1.7275337848394856e-06,
+      "loss": 1.0965,
+      "step": 15083
+    },
+    {
+      "epoch": 0.9164590801385261,
+      "grad_norm": 0.17283283174037933,
+      "learning_rate": 1.7250400923079447e-06,
+      "loss": 1.1046,
+      "step": 15084
+    },
+    {
+      "epoch": 0.9165198371711526,
+      "grad_norm": 0.2416338324546814,
+      "learning_rate": 1.7225481693043776e-06,
+      "loss": 1.0715,
+      "step": 15085
+    },
+    {
+      "epoch": 0.9165805942037791,
+      "grad_norm": 0.1734868586063385,
+      "learning_rate": 1.7200580159201396e-06,
+      "loss": 1.0368,
+      "step": 15086
+    },
+    {
+      "epoch": 0.9166413512364057,
+      "grad_norm": 0.1387193202972412,
+      "learning_rate": 1.7175696322465129e-06,
+      "loss": 1.1174,
+      "step": 15087
+    },
+    {
+      "epoch": 0.9167021082690321,
+      "grad_norm": 0.22944627702236176,
+      "learning_rate": 1.7150830183746968e-06,
+      "loss": 1.1043,
+      "step": 15088
+    },
+    {
+      "epoch": 0.9167628653016586,
+      "grad_norm": 0.25378623604774475,
+      "learning_rate": 1.7125981743958518e-06,
+      "loss": 1.0573,
+      "step": 15089
+    },
+    {
+      "epoch": 0.9168236223342852,
+      "grad_norm": 0.226588174700737,
+      "learning_rate": 1.7101151004010551e-06,
+      "loss": 1.0743,
+      "step": 15090
+    },
+    {
+      "epoch": 0.9168843793669117,
+      "grad_norm": 0.11758585274219513,
+      "learning_rate": 1.7076337964813226e-06,
+      "loss": 1.0468,
+      "step": 15091
+    },
+    {
+      "epoch": 0.9169451363995382,
+      "grad_norm": 0.1250493824481964,
+      "learning_rate": 1.7051542627276097e-06,
+      "loss": 1.1005,
+      "step": 15092
+    },
+    {
+      "epoch": 0.9170058934321648,
+      "grad_norm": 0.10729379206895828,
+      "learning_rate": 1.7026764992307987e-06,
+      "loss": 1.0467,
+      "step": 15093
+    },
+    {
+      "epoch": 0.9170666504647913,
+      "grad_norm": 0.1321740299463272,
+      "learning_rate": 1.7002005060817173e-06,
+      "loss": 1.0143,
+      "step": 15094
+    },
+    {
+      "epoch": 0.9171274074974178,
+      "grad_norm": 0.13206522166728973,
+      "learning_rate": 1.6977262833711205e-06,
+      "loss": 1.1253,
+      "step": 15095
+    },
+    {
+      "epoch": 0.9171881645300444,
+      "grad_norm": 0.11639541387557983,
+      "learning_rate": 1.695253831189708e-06,
+      "loss": 1.0193,
+      "step": 15096
+    },
+    {
+      "epoch": 0.9172489215626709,
+      "grad_norm": 0.20463216304779053,
+      "learning_rate": 1.6927831496281021e-06,
+      "loss": 1.2974,
+      "step": 15097
+    },
+    {
+      "epoch": 0.9173096785952974,
+      "grad_norm": 0.14341233670711517,
+      "learning_rate": 1.6903142387768634e-06,
+      "loss": 1.0719,
+      "step": 15098
+    },
+    {
+      "epoch": 0.917370435627924,
+      "grad_norm": 0.11572027951478958,
+      "learning_rate": 1.6878470987264971e-06,
+      "loss": 1.0303,
+      "step": 15099
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 0.1278015822172165,
+      "learning_rate": 1.6853817295674368e-06,
+      "loss": 1.066,
+      "step": 15100
+    },
+    {
+      "epoch": 0.9174919496931769,
+      "grad_norm": 0.1656336933374405,
+      "learning_rate": 1.6829181313900488e-06,
+      "loss": 1.167,
+      "step": 15101
+    },
+    {
+      "epoch": 0.9175527067258035,
+      "grad_norm": 0.13989870250225067,
+      "learning_rate": 1.6804563042846223e-06,
+      "loss": 1.1451,
+      "step": 15102
+    },
+    {
+      "epoch": 0.91761346375843,
+      "grad_norm": 0.09838593751192093,
+      "learning_rate": 1.6779962483414235e-06,
+      "loss": 1.0545,
+      "step": 15103
+    },
+    {
+      "epoch": 0.9176742207910565,
+      "grad_norm": 0.2986197769641876,
+      "learning_rate": 1.6755379636506141e-06,
+      "loss": 1.122,
+      "step": 15104
+    },
+    {
+      "epoch": 0.9177349778236831,
+      "grad_norm": 0.12212865054607391,
+      "learning_rate": 1.6730814503022997e-06,
+      "loss": 1.0263,
+      "step": 15105
+    },
+    {
+      "epoch": 0.9177957348563096,
+      "grad_norm": 0.17864882946014404,
+      "learning_rate": 1.6706267083865245e-06,
+      "loss": 1.1397,
+      "step": 15106
+    },
+    {
+      "epoch": 0.9178564918889361,
+      "grad_norm": 0.15472503006458282,
+      "learning_rate": 1.6681737379932616e-06,
+      "loss": 1.0232,
+      "step": 15107
+    },
+    {
+      "epoch": 0.9179172489215627,
+      "grad_norm": 0.11831034719944,
+      "learning_rate": 1.6657225392124387e-06,
+      "loss": 1.0462,
+      "step": 15108
+    },
+    {
+      "epoch": 0.9179780059541892,
+      "grad_norm": 0.16926069557666779,
+      "learning_rate": 1.6632731121339008e-06,
+      "loss": 1.1335,
+      "step": 15109
+    },
+    {
+      "epoch": 0.9180387629868157,
+      "grad_norm": 0.11622993648052216,
+      "learning_rate": 1.6608254568474313e-06,
+      "loss": 1.0223,
+      "step": 15110
+    },
+    {
+      "epoch": 0.9180995200194423,
+      "grad_norm": 0.1877569556236267,
+      "learning_rate": 1.6583795734427477e-06,
+      "loss": 1.0648,
+      "step": 15111
+    },
+    {
+      "epoch": 0.9181602770520688,
+      "grad_norm": 0.1548006385564804,
+      "learning_rate": 1.6559354620095003e-06,
+      "loss": 1.0551,
+      "step": 15112
+    },
+    {
+      "epoch": 0.9182210340846954,
+      "grad_norm": 5.163941383361816,
+      "learning_rate": 1.6534931226372896e-06,
+      "loss": 1.0314,
+      "step": 15113
+    },
+    {
+      "epoch": 0.9182817911173218,
+      "grad_norm": 13.354634284973145,
+      "learning_rate": 1.6510525554156275e-06,
+      "loss": 1.0812,
+      "step": 15114
+    },
+    {
+      "epoch": 0.9183425481499483,
+      "grad_norm": 0.21274954080581665,
+      "learning_rate": 1.6486137604339813e-06,
+      "loss": 1.101,
+      "step": 15115
+    },
+    {
+      "epoch": 0.9184033051825748,
+      "grad_norm": 0.14278192818164825,
+      "learning_rate": 1.646176737781746e-06,
+      "loss": 1.1057,
+      "step": 15116
+    },
+    {
+      "epoch": 0.9184640622152014,
+      "grad_norm": 0.1299707293510437,
+      "learning_rate": 1.6437414875482504e-06,
+      "loss": 1.0963,
+      "step": 15117
+    },
+    {
+      "epoch": 0.9185248192478279,
+      "grad_norm": 0.1664724498987198,
+      "learning_rate": 1.6413080098227561e-06,
+      "loss": 1.0759,
+      "step": 15118
+    },
+    {
+      "epoch": 0.9185855762804545,
+      "grad_norm": 0.17971952259540558,
+      "learning_rate": 1.6388763046944644e-06,
+      "loss": 1.2105,
+      "step": 15119
+    },
+    {
+      "epoch": 0.918646333313081,
+      "grad_norm": 0.13144651055335999,
+      "learning_rate": 1.6364463722525092e-06,
+      "loss": 1.0894,
+      "step": 15120
+    },
+    {
+      "epoch": 0.9187070903457075,
+      "grad_norm": 0.1569184511899948,
+      "learning_rate": 1.6340182125859583e-06,
+      "loss": 1.111,
+      "step": 15121
+    },
+    {
+      "epoch": 0.9187678473783341,
+      "grad_norm": 0.11527897417545319,
+      "learning_rate": 1.6315918257838236e-06,
+      "loss": 1.0228,
+      "step": 15122
+    },
+    {
+      "epoch": 0.9188286044109606,
+      "grad_norm": 7.612890243530273,
+      "learning_rate": 1.6291672119350288e-06,
+      "loss": 1.0654,
+      "step": 15123
+    },
+    {
+      "epoch": 0.9188893614435871,
+      "grad_norm": 0.11632709950208664,
+      "learning_rate": 1.6267443711284747e-06,
+      "loss": 1.0385,
+      "step": 15124
+    },
+    {
+      "epoch": 0.9189501184762137,
+      "grad_norm": 0.18938694894313812,
+      "learning_rate": 1.6243233034529515e-06,
+      "loss": 1.2582,
+      "step": 15125
+    },
+    {
+      "epoch": 0.9190108755088402,
+      "grad_norm": 0.332832008600235,
+      "learning_rate": 1.621904008997205e-06,
+      "loss": 1.1757,
+      "step": 15126
+    },
+    {
+      "epoch": 0.9190716325414666,
+      "grad_norm": 0.18409976363182068,
+      "learning_rate": 1.6194864878499193e-06,
+      "loss": 1.0493,
+      "step": 15127
+    },
+    {
+      "epoch": 0.9191323895740932,
+      "grad_norm": 0.13636156916618347,
+      "learning_rate": 1.617070740099702e-06,
+      "loss": 0.9828,
+      "step": 15128
+    },
+    {
+      "epoch": 0.9191931466067197,
+      "grad_norm": 0.2805340588092804,
+      "learning_rate": 1.6146567658351154e-06,
+      "loss": 1.154,
+      "step": 15129
+    },
+    {
+      "epoch": 0.9192539036393462,
+      "grad_norm": 0.11665313690900803,
+      "learning_rate": 1.6122445651446383e-06,
+      "loss": 1.0406,
+      "step": 15130
+    },
+    {
+      "epoch": 0.9193146606719728,
+      "grad_norm": 0.14811690151691437,
+      "learning_rate": 1.6098341381166949e-06,
+      "loss": 1.0485,
+      "step": 15131
+    },
+    {
+      "epoch": 0.9193754177045993,
+      "grad_norm": 0.08609217405319214,
+      "learning_rate": 1.607425484839631e-06,
+      "loss": 1.0399,
+      "step": 15132
+    },
+    {
+      "epoch": 0.9194361747372258,
+      "grad_norm": 0.2471848428249359,
+      "learning_rate": 1.6050186054017313e-06,
+      "loss": 1.1401,
+      "step": 15133
+    },
+    {
+      "epoch": 0.9194969317698524,
+      "grad_norm": 0.13908128440380096,
+      "learning_rate": 1.6026134998912367e-06,
+      "loss": 1.0849,
+      "step": 15134
+    },
+    {
+      "epoch": 0.9195576888024789,
+      "grad_norm": 0.1085067093372345,
+      "learning_rate": 1.600210168396299e-06,
+      "loss": 1.0475,
+      "step": 15135
+    },
+    {
+      "epoch": 0.9196184458351054,
+      "grad_norm": 0.19212636351585388,
+      "learning_rate": 1.597808611005014e-06,
+      "loss": 1.1038,
+      "step": 15136
+    },
+    {
+      "epoch": 0.919679202867732,
+      "grad_norm": 0.11599438637495041,
+      "learning_rate": 1.5954088278054114e-06,
+      "loss": 1.0308,
+      "step": 15137
+    },
+    {
+      "epoch": 0.9197399599003585,
+      "grad_norm": 0.10855875164270401,
+      "learning_rate": 1.5930108188854487e-06,
+      "loss": 1.0726,
+      "step": 15138
+    },
+    {
+      "epoch": 0.919800716932985,
+      "grad_norm": 0.1338091939687729,
+      "learning_rate": 1.590614584333039e-06,
+      "loss": 1.0135,
+      "step": 15139
+    },
+    {
+      "epoch": 0.9198614739656115,
+      "grad_norm": 0.27087607979774475,
+      "learning_rate": 1.5882201242360006e-06,
+      "loss": 1.1385,
+      "step": 15140
+    },
+    {
+      "epoch": 0.919922230998238,
+      "grad_norm": 0.15763407945632935,
+      "learning_rate": 1.5858274386821192e-06,
+      "loss": 1.1272,
+      "step": 15141
+    },
+    {
+      "epoch": 0.9199829880308645,
+      "grad_norm": 0.265626460313797,
+      "learning_rate": 1.5834365277590857e-06,
+      "loss": 1.0558,
+      "step": 15142
+    },
+    {
+      "epoch": 0.9200437450634911,
+      "grad_norm": 0.15396176278591156,
+      "learning_rate": 1.5810473915545466e-06,
+      "loss": 1.0715,
+      "step": 15143
+    },
+    {
+      "epoch": 0.9201045020961176,
+      "grad_norm": 0.1626933515071869,
+      "learning_rate": 1.578660030156076e-06,
+      "loss": 1.1258,
+      "step": 15144
+    },
+    {
+      "epoch": 0.9201652591287441,
+      "grad_norm": 0.16092416644096375,
+      "learning_rate": 1.5762744436511822e-06,
+      "loss": 1.0809,
+      "step": 15145
+    },
+    {
+      "epoch": 0.9202260161613707,
+      "grad_norm": 0.3362584412097931,
+      "learning_rate": 1.573890632127306e-06,
+      "loss": 1.2333,
+      "step": 15146
+    },
+    {
+      "epoch": 0.9202867731939972,
+      "grad_norm": 0.13535462319850922,
+      "learning_rate": 1.5715085956718278e-06,
+      "loss": 1.0837,
+      "step": 15147
+    },
+    {
+      "epoch": 0.9203475302266237,
+      "grad_norm": 0.11573890596628189,
+      "learning_rate": 1.5691283343720664e-06,
+      "loss": 1.0326,
+      "step": 15148
+    },
+    {
+      "epoch": 0.9204082872592503,
+      "grad_norm": 0.2805766463279724,
+      "learning_rate": 1.566749848315263e-06,
+      "loss": 1.0142,
+      "step": 15149
+    },
+    {
+      "epoch": 0.9204690442918768,
+      "grad_norm": 0.12610600888729095,
+      "learning_rate": 1.5643731375886094e-06,
+      "loss": 1.0439,
+      "step": 15150
+    },
+    {
+      "epoch": 0.9205298013245033,
+      "grad_norm": 0.1797090768814087,
+      "learning_rate": 1.5619982022792245e-06,
+      "loss": 1.1453,
+      "step": 15151
+    },
+    {
+      "epoch": 0.9205905583571299,
+      "grad_norm": 0.15865083038806915,
+      "learning_rate": 1.5596250424741553e-06,
+      "loss": 1.1232,
+      "step": 15152
+    },
+    {
+      "epoch": 0.9206513153897564,
+      "grad_norm": 0.12154293060302734,
+      "learning_rate": 1.5572536582603935e-06,
+      "loss": 1.047,
+      "step": 15153
+    },
+    {
+      "epoch": 0.9207120724223828,
+      "grad_norm": 0.25497084856033325,
+      "learning_rate": 1.5548840497248584e-06,
+      "loss": 1.1329,
+      "step": 15154
+    },
+    {
+      "epoch": 0.9207728294550094,
+      "grad_norm": 0.10914847999811172,
+      "learning_rate": 1.5525162169544138e-06,
+      "loss": 1.0099,
+      "step": 15155
+    },
+    {
+      "epoch": 0.9208335864876359,
+      "grad_norm": 0.12534913420677185,
+      "learning_rate": 1.550150160035857e-06,
+      "loss": 1.0252,
+      "step": 15156
+    },
+    {
+      "epoch": 0.9208943435202624,
+      "grad_norm": 0.12034306675195694,
+      "learning_rate": 1.5477858790559075e-06,
+      "loss": 1.0439,
+      "step": 15157
+    },
+    {
+      "epoch": 0.920955100552889,
+      "grad_norm": 0.13736814260482788,
+      "learning_rate": 1.5454233741012403e-06,
+      "loss": 1.059,
+      "step": 15158
+    },
+    {
+      "epoch": 0.9210158575855155,
+      "grad_norm": 0.6928398013114929,
+      "learning_rate": 1.5430626452584306e-06,
+      "loss": 1.0643,
+      "step": 15159
+    },
+    {
+      "epoch": 0.921076614618142,
+      "grad_norm": 0.12043861299753189,
+      "learning_rate": 1.5407036926140372e-06,
+      "loss": 0.984,
+      "step": 15160
+    },
+    {
+      "epoch": 0.9211373716507686,
+      "grad_norm": 0.15041232109069824,
+      "learning_rate": 1.538346516254513e-06,
+      "loss": 1.0959,
+      "step": 15161
+    },
+    {
+      "epoch": 0.9211981286833951,
+      "grad_norm": 26.66671371459961,
+      "learning_rate": 1.5359911162662667e-06,
+      "loss": 1.0281,
+      "step": 15162
+    },
+    {
+      "epoch": 0.9212588857160217,
+      "grad_norm": 0.11382798850536346,
+      "learning_rate": 1.5336374927356344e-06,
+      "loss": 1.0151,
+      "step": 15163
+    },
+    {
+      "epoch": 0.9213196427486482,
+      "grad_norm": 0.247624009847641,
+      "learning_rate": 1.5312856457488866e-06,
+      "loss": 1.149,
+      "step": 15164
+    },
+    {
+      "epoch": 0.9213803997812747,
+      "grad_norm": 0.4614062011241913,
+      "learning_rate": 1.5289355753922318e-06,
+      "loss": 1.2893,
+      "step": 15165
+    },
+    {
+      "epoch": 0.9214411568139013,
+      "grad_norm": 0.1196788027882576,
+      "learning_rate": 1.5265872817518123e-06,
+      "loss": 0.992,
+      "step": 15166
+    },
+    {
+      "epoch": 0.9215019138465277,
+      "grad_norm": 0.2365286499261856,
+      "learning_rate": 1.5242407649137036e-06,
+      "loss": 1.0595,
+      "step": 15167
+    },
+    {
+      "epoch": 0.9215626708791542,
+      "grad_norm": 0.1954217255115509,
+      "learning_rate": 1.5218960249639258e-06,
+      "loss": 1.0842,
+      "step": 15168
+    },
+    {
+      "epoch": 0.9216234279117808,
+      "grad_norm": 0.16624826192855835,
+      "learning_rate": 1.5195530619884158e-06,
+      "loss": 1.0679,
+      "step": 15169
+    },
+    {
+      "epoch": 0.9216841849444073,
+      "grad_norm": 0.11108061671257019,
+      "learning_rate": 1.5172118760730546e-06,
+      "loss": 1.0124,
+      "step": 15170
+    },
+    {
+      "epoch": 0.9217449419770338,
+      "grad_norm": 0.13399088382720947,
+      "learning_rate": 1.514872467303674e-06,
+      "loss": 1.0822,
+      "step": 15171
+    },
+    {
+      "epoch": 0.9218056990096604,
+      "grad_norm": 0.1440088152885437,
+      "learning_rate": 1.5125348357660108e-06,
+      "loss": 1.0829,
+      "step": 15172
+    },
+    {
+      "epoch": 0.9218664560422869,
+      "grad_norm": 0.17107906937599182,
+      "learning_rate": 1.5101989815457573e-06,
+      "loss": 1.0695,
+      "step": 15173
+    },
+    {
+      "epoch": 0.9219272130749134,
+      "grad_norm": 0.3680818974971771,
+      "learning_rate": 1.5078649047285288e-06,
+      "loss": 1.0326,
+      "step": 15174
+    },
+    {
+      "epoch": 0.92198797010754,
+      "grad_norm": 0.14400503039360046,
+      "learning_rate": 1.505532605399884e-06,
+      "loss": 1.1473,
+      "step": 15175
+    },
+    {
+      "epoch": 0.9220487271401665,
+      "grad_norm": 0.11141887307167053,
+      "learning_rate": 1.5032020836453164e-06,
+      "loss": 1.0741,
+      "step": 15176
+    },
+    {
+      "epoch": 0.922109484172793,
+      "grad_norm": 0.15904878079891205,
+      "learning_rate": 1.5008733395502517e-06,
+      "loss": 1.0601,
+      "step": 15177
+    },
+    {
+      "epoch": 0.9221702412054196,
+      "grad_norm": 0.11423798650503159,
+      "learning_rate": 1.4985463732000604e-06,
+      "loss": 1.056,
+      "step": 15178
+    },
+    {
+      "epoch": 0.9222309982380461,
+      "grad_norm": 0.11866994947195053,
+      "learning_rate": 1.4962211846800077e-06,
+      "loss": 1.0049,
+      "step": 15179
+    },
+    {
+      "epoch": 0.9222917552706725,
+      "grad_norm": 0.29792261123657227,
+      "learning_rate": 1.4938977740753535e-06,
+      "loss": 1.0376,
+      "step": 15180
+    },
+    {
+      "epoch": 0.9223525123032991,
+      "grad_norm": 0.12448982894420624,
+      "learning_rate": 1.4915761414712514e-06,
+      "loss": 0.9684,
+      "step": 15181
+    },
+    {
+      "epoch": 0.9224132693359256,
+      "grad_norm": 0.24090053141117096,
+      "learning_rate": 1.4892562869528059e-06,
+      "loss": 1.1001,
+      "step": 15182
+    },
+    {
+      "epoch": 0.9224740263685521,
+      "grad_norm": 0.1544858068227768,
+      "learning_rate": 1.4869382106050433e-06,
+      "loss": 1.0554,
+      "step": 15183
+    },
+    {
+      "epoch": 0.9225347834011787,
+      "grad_norm": 0.12713445723056793,
+      "learning_rate": 1.48462191251294e-06,
+      "loss": 1.0475,
+      "step": 15184
+    },
+    {
+      "epoch": 0.9225955404338052,
+      "grad_norm": 0.7496867179870605,
+      "learning_rate": 1.4823073927614006e-06,
+      "loss": 1.1893,
+      "step": 15185
+    },
+    {
+      "epoch": 0.9226562974664317,
+      "grad_norm": 0.1444765031337738,
+      "learning_rate": 1.479994651435257e-06,
+      "loss": 1.1527,
+      "step": 15186
+    },
+    {
+      "epoch": 0.9227170544990583,
+      "grad_norm": 0.1176825538277626,
+      "learning_rate": 1.4776836886192914e-06,
+      "loss": 1.0687,
+      "step": 15187
+    },
+    {
+      "epoch": 0.9227778115316848,
+      "grad_norm": 0.11566071212291718,
+      "learning_rate": 1.4753745043982081e-06,
+      "loss": 1.0357,
+      "step": 15188
+    },
+    {
+      "epoch": 0.9228385685643113,
+      "grad_norm": 0.14826469123363495,
+      "learning_rate": 1.4730670988566508e-06,
+      "loss": 1.0522,
+      "step": 15189
+    },
+    {
+      "epoch": 0.9228993255969379,
+      "grad_norm": 0.13102050125598907,
+      "learning_rate": 1.4707614720792018e-06,
+      "loss": 1.0462,
+      "step": 15190
+    },
+    {
+      "epoch": 0.9229600826295644,
+      "grad_norm": 0.3773258328437805,
+      "learning_rate": 1.4684576241503712e-06,
+      "loss": 1.1564,
+      "step": 15191
+    },
+    {
+      "epoch": 0.923020839662191,
+      "grad_norm": 0.16633334755897522,
+      "learning_rate": 1.4661555551546024e-06,
+      "loss": 1.0712,
+      "step": 15192
+    },
+    {
+      "epoch": 0.9230815966948174,
+      "grad_norm": 0.1298397034406662,
+      "learning_rate": 1.4638552651762893e-06,
+      "loss": 1.0648,
+      "step": 15193
+    },
+    {
+      "epoch": 0.9231423537274439,
+      "grad_norm": 0.11218418926000595,
+      "learning_rate": 1.4615567542997422e-06,
+      "loss": 1.0078,
+      "step": 15194
+    },
+    {
+      "epoch": 0.9232031107600704,
+      "grad_norm": 0.2105770856142044,
+      "learning_rate": 1.4592600226092045e-06,
+      "loss": 1.0128,
+      "step": 15195
+    },
+    {
+      "epoch": 0.923263867792697,
+      "grad_norm": 2.036931276321411,
+      "learning_rate": 1.4569650701888815e-06,
+      "loss": 1.0732,
+      "step": 15196
+    },
+    {
+      "epoch": 0.9233246248253235,
+      "grad_norm": 0.11545518040657043,
+      "learning_rate": 1.454671897122889e-06,
+      "loss": 1.0226,
+      "step": 15197
+    },
+    {
+      "epoch": 0.92338538185795,
+      "grad_norm": 0.5267545580863953,
+      "learning_rate": 1.4523805034952874e-06,
+      "loss": 1.4146,
+      "step": 15198
+    },
+    {
+      "epoch": 0.9234461388905766,
+      "grad_norm": 0.10423264652490616,
+      "learning_rate": 1.4500908893900544e-06,
+      "loss": 1.0388,
+      "step": 15199
+    },
+    {
+      "epoch": 0.9235068959232031,
+      "grad_norm": 0.1641158014535904,
+      "learning_rate": 1.447803054891117e-06,
+      "loss": 1.0857,
+      "step": 15200
+    },
+    {
+      "epoch": 0.9235676529558297,
+      "grad_norm": 0.17542433738708496,
+      "learning_rate": 1.4455170000823526e-06,
+      "loss": 1.0079,
+      "step": 15201
+    },
+    {
+      "epoch": 0.9236284099884562,
+      "grad_norm": 0.1571839153766632,
+      "learning_rate": 1.4432327250475497e-06,
+      "loss": 1.0978,
+      "step": 15202
+    },
+    {
+      "epoch": 0.9236891670210827,
+      "grad_norm": 0.15233772993087769,
+      "learning_rate": 1.4409502298704359e-06,
+      "loss": 1.0884,
+      "step": 15203
+    },
+    {
+      "epoch": 0.9237499240537093,
+      "grad_norm": 0.1004517450928688,
+      "learning_rate": 1.4386695146346774e-06,
+      "loss": 0.9797,
+      "step": 15204
+    },
+    {
+      "epoch": 0.9238106810863358,
+      "grad_norm": 0.1002735048532486,
+      "learning_rate": 1.4363905794238686e-06,
+      "loss": 1.0276,
+      "step": 15205
+    },
+    {
+      "epoch": 0.9238714381189622,
+      "grad_norm": 0.1188020408153534,
+      "learning_rate": 1.4341134243215593e-06,
+      "loss": 1.0357,
+      "step": 15206
+    },
+    {
+      "epoch": 0.9239321951515888,
+      "grad_norm": 0.11223286390304565,
+      "learning_rate": 1.431838049411205e-06,
+      "loss": 1.0715,
+      "step": 15207
+    },
+    {
+      "epoch": 0.9239929521842153,
+      "grad_norm": 7.940769195556641,
+      "learning_rate": 1.4295644547762166e-06,
+      "loss": 1.0322,
+      "step": 15208
+    },
+    {
+      "epoch": 0.9240537092168418,
+      "grad_norm": 0.2006407082080841,
+      "learning_rate": 1.4272926404999276e-06,
+      "loss": 1.0644,
+      "step": 15209
+    },
+    {
+      "epoch": 0.9241144662494684,
+      "grad_norm": 0.18006841838359833,
+      "learning_rate": 1.4250226066656213e-06,
+      "loss": 0.9936,
+      "step": 15210
+    },
+    {
+      "epoch": 0.9241752232820949,
+      "grad_norm": 0.13213543593883514,
+      "learning_rate": 1.422754353356498e-06,
+      "loss": 1.0732,
+      "step": 15211
+    },
+    {
+      "epoch": 0.9242359803147214,
+      "grad_norm": 0.10727523267269135,
+      "learning_rate": 1.420487880655702e-06,
+      "loss": 1.0815,
+      "step": 15212
+    },
+    {
+      "epoch": 0.924296737347348,
+      "grad_norm": 0.13576774299144745,
+      "learning_rate": 1.4182231886463171e-06,
+      "loss": 1.0189,
+      "step": 15213
+    },
+    {
+      "epoch": 0.9243574943799745,
+      "grad_norm": 0.6041025519371033,
+      "learning_rate": 1.415960277411349e-06,
+      "loss": 1.129,
+      "step": 15214
+    },
+    {
+      "epoch": 0.924418251412601,
+      "grad_norm": 0.1084340512752533,
+      "learning_rate": 1.413699147033748e-06,
+      "loss": 1.0276,
+      "step": 15215
+    },
+    {
+      "epoch": 0.9244790084452276,
+      "grad_norm": 0.5081831216812134,
+      "learning_rate": 1.4114397975963923e-06,
+      "loss": 1.0542,
+      "step": 15216
+    },
+    {
+      "epoch": 0.9245397654778541,
+      "grad_norm": 0.19882328808307648,
+      "learning_rate": 1.40918222918211e-06,
+      "loss": 1.0765,
+      "step": 15217
+    },
+    {
+      "epoch": 0.9246005225104806,
+      "grad_norm": 0.0956098809838295,
+      "learning_rate": 1.406926441873646e-06,
+      "loss": 1.0108,
+      "step": 15218
+    },
+    {
+      "epoch": 0.9246612795431071,
+      "grad_norm": 0.32133370637893677,
+      "learning_rate": 1.4046724357536846e-06,
+      "loss": 1.1457,
+      "step": 15219
+    },
+    {
+      "epoch": 0.9247220365757336,
+      "grad_norm": 0.11757557094097137,
+      "learning_rate": 1.402420210904848e-06,
+      "loss": 1.0127,
+      "step": 15220
+    },
+    {
+      "epoch": 0.9247827936083601,
+      "grad_norm": 0.1253536343574524,
+      "learning_rate": 1.400169767409687e-06,
+      "loss": 1.0741,
+      "step": 15221
+    },
+    {
+      "epoch": 0.9248435506409867,
+      "grad_norm": 0.10463772714138031,
+      "learning_rate": 1.3979211053506968e-06,
+      "loss": 0.9783,
+      "step": 15222
+    },
+    {
+      "epoch": 0.9249043076736132,
+      "grad_norm": 0.13048705458641052,
+      "learning_rate": 1.3956742248103117e-06,
+      "loss": 1.0796,
+      "step": 15223
+    },
+    {
+      "epoch": 0.9249650647062397,
+      "grad_norm": 2.9244537353515625,
+      "learning_rate": 1.3934291258708765e-06,
+      "loss": 1.2361,
+      "step": 15224
+    },
+    {
+      "epoch": 0.9250258217388663,
+      "grad_norm": 0.15533210337162018,
+      "learning_rate": 1.3911858086147033e-06,
+      "loss": 1.1036,
+      "step": 15225
+    },
+    {
+      "epoch": 0.9250865787714928,
+      "grad_norm": 0.2609361708164215,
+      "learning_rate": 1.3889442731239933e-06,
+      "loss": 1.1349,
+      "step": 15226
+    },
+    {
+      "epoch": 0.9251473358041193,
+      "grad_norm": 0.19171033799648285,
+      "learning_rate": 1.3867045194809358e-06,
+      "loss": 1.0601,
+      "step": 15227
+    },
+    {
+      "epoch": 0.9252080928367459,
+      "grad_norm": 0.1614377349615097,
+      "learning_rate": 1.384466547767621e-06,
+      "loss": 1.0036,
+      "step": 15228
+    },
+    {
+      "epoch": 0.9252688498693724,
+      "grad_norm": 0.12945087254047394,
+      "learning_rate": 1.3822303580660833e-06,
+      "loss": 1.0721,
+      "step": 15229
+    },
+    {
+      "epoch": 0.925329606901999,
+      "grad_norm": 0.13309133052825928,
+      "learning_rate": 1.379995950458285e-06,
+      "loss": 1.1972,
+      "step": 15230
+    },
+    {
+      "epoch": 0.9253903639346255,
+      "grad_norm": 0.1279413104057312,
+      "learning_rate": 1.377763325026138e-06,
+      "loss": 1.0477,
+      "step": 15231
+    },
+    {
+      "epoch": 0.9254511209672519,
+      "grad_norm": 0.1321583092212677,
+      "learning_rate": 1.3755324818514714e-06,
+      "loss": 1.1255,
+      "step": 15232
+    },
+    {
+      "epoch": 0.9255118779998784,
+      "grad_norm": 0.19626502692699432,
+      "learning_rate": 1.373303421016059e-06,
+      "loss": 1.0662,
+      "step": 15233
+    },
+    {
+      "epoch": 0.925572635032505,
+      "grad_norm": 0.13864190876483917,
+      "learning_rate": 1.3710761426016127e-06,
+      "loss": 1.0627,
+      "step": 15234
+    },
+    {
+      "epoch": 0.9256333920651315,
+      "grad_norm": 0.20498597621917725,
+      "learning_rate": 1.3688506466897733e-06,
+      "loss": 1.0984,
+      "step": 15235
+    },
+    {
+      "epoch": 0.925694149097758,
+      "grad_norm": 0.17506982386112213,
+      "learning_rate": 1.3666269333621084e-06,
+      "loss": 1.1093,
+      "step": 15236
+    },
+    {
+      "epoch": 0.9257549061303846,
+      "grad_norm": 0.12914767861366272,
+      "learning_rate": 1.364405002700131e-06,
+      "loss": 1.033,
+      "step": 15237
+    },
+    {
+      "epoch": 0.9258156631630111,
+      "grad_norm": 0.11264797300100327,
+      "learning_rate": 1.3621848547852978e-06,
+      "loss": 1.0989,
+      "step": 15238
+    },
+    {
+      "epoch": 0.9258764201956377,
+      "grad_norm": 0.20568512380123138,
+      "learning_rate": 1.3599664896989773e-06,
+      "loss": 1.0369,
+      "step": 15239
+    },
+    {
+      "epoch": 0.9259371772282642,
+      "grad_norm": 0.11404545605182648,
+      "learning_rate": 1.357749907522482e-06,
+      "loss": 1.0098,
+      "step": 15240
+    },
+    {
+      "epoch": 0.9259979342608907,
+      "grad_norm": 0.46397948265075684,
+      "learning_rate": 1.3555351083370694e-06,
+      "loss": 1.0878,
+      "step": 15241
+    },
+    {
+      "epoch": 0.9260586912935173,
+      "grad_norm": 0.3841884136199951,
+      "learning_rate": 1.3533220922239188e-06,
+      "loss": 1.0557,
+      "step": 15242
+    },
+    {
+      "epoch": 0.9261194483261438,
+      "grad_norm": 0.2260499745607376,
+      "learning_rate": 1.3511108592641487e-06,
+      "loss": 1.2119,
+      "step": 15243
+    },
+    {
+      "epoch": 0.9261802053587703,
+      "grad_norm": 17.858728408813477,
+      "learning_rate": 1.3489014095388163e-06,
+      "loss": 1.0558,
+      "step": 15244
+    },
+    {
+      "epoch": 0.9262409623913967,
+      "grad_norm": 0.3694033920764923,
+      "learning_rate": 1.346693743128913e-06,
+      "loss": 1.0222,
+      "step": 15245
+    },
+    {
+      "epoch": 0.9263017194240233,
+      "grad_norm": 1.4111883640289307,
+      "learning_rate": 1.344487860115351e-06,
+      "loss": 1.0564,
+      "step": 15246
+    },
+    {
+      "epoch": 0.9263624764566498,
+      "grad_norm": 0.1914466917514801,
+      "learning_rate": 1.3422837605789828e-06,
+      "loss": 1.1455,
+      "step": 15247
+    },
+    {
+      "epoch": 0.9264232334892764,
+      "grad_norm": 0.20472809672355652,
+      "learning_rate": 1.340081444600616e-06,
+      "loss": 1.0565,
+      "step": 15248
+    },
+    {
+      "epoch": 0.9264839905219029,
+      "grad_norm": 0.17095518112182617,
+      "learning_rate": 1.3378809122609693e-06,
+      "loss": 1.077,
+      "step": 15249
+    },
+    {
+      "epoch": 0.9265447475545294,
+      "grad_norm": 0.5156824588775635,
+      "learning_rate": 1.3356821636407059e-06,
+      "loss": 1.0825,
+      "step": 15250
+    },
+    {
+      "epoch": 0.926605504587156,
+      "grad_norm": 0.25496575236320496,
+      "learning_rate": 1.3334851988204167e-06,
+      "loss": 1.142,
+      "step": 15251
+    },
+    {
+      "epoch": 0.9266662616197825,
+      "grad_norm": 0.13751541078090668,
+      "learning_rate": 1.331290017880632e-06,
+      "loss": 1.0444,
+      "step": 15252
+    },
+    {
+      "epoch": 0.926727018652409,
+      "grad_norm": 0.11381416767835617,
+      "learning_rate": 1.3290966209018264e-06,
+      "loss": 1.0318,
+      "step": 15253
+    },
+    {
+      "epoch": 0.9267877756850356,
+      "grad_norm": 0.13319475948810577,
+      "learning_rate": 1.3269050079643851e-06,
+      "loss": 1.0916,
+      "step": 15254
+    },
+    {
+      "epoch": 0.9268485327176621,
+      "grad_norm": 0.11884255707263947,
+      "learning_rate": 1.3247151791486557e-06,
+      "loss": 1.0822,
+      "step": 15255
+    },
+    {
+      "epoch": 0.9269092897502886,
+      "grad_norm": 0.21788297593593597,
+      "learning_rate": 1.3225271345348956e-06,
+      "loss": 1.1315,
+      "step": 15256
+    },
+    {
+      "epoch": 0.9269700467829152,
+      "grad_norm": 0.10422495752573013,
+      "learning_rate": 1.320340874203313e-06,
+      "loss": 1.007,
+      "step": 15257
+    },
+    {
+      "epoch": 0.9270308038155417,
+      "grad_norm": 0.14656057953834534,
+      "learning_rate": 1.3181563982340438e-06,
+      "loss": 1.1066,
+      "step": 15258
+    },
+    {
+      "epoch": 0.9270915608481681,
+      "grad_norm": 0.22052647173404694,
+      "learning_rate": 1.3159737067071631e-06,
+      "loss": 1.0927,
+      "step": 15259
+    },
+    {
+      "epoch": 0.9271523178807947,
+      "grad_norm": 0.1543327420949936,
+      "learning_rate": 1.313792799702679e-06,
+      "loss": 1.0855,
+      "step": 15260
+    },
+    {
+      "epoch": 0.9272130749134212,
+      "grad_norm": 0.2059401124715805,
+      "learning_rate": 1.311613677300527e-06,
+      "loss": 1.1389,
+      "step": 15261
+    },
+    {
+      "epoch": 0.9272738319460477,
+      "grad_norm": 0.18106365203857422,
+      "learning_rate": 1.3094363395805887e-06,
+      "loss": 1.0509,
+      "step": 15262
+    },
+    {
+      "epoch": 0.9273345889786743,
+      "grad_norm": 0.11952624469995499,
+      "learning_rate": 1.307260786622666e-06,
+      "loss": 1.0256,
+      "step": 15263
+    },
+    {
+      "epoch": 0.9273953460113008,
+      "grad_norm": 0.1100388839840889,
+      "learning_rate": 1.3050870185065178e-06,
+      "loss": 1.0106,
+      "step": 15264
+    },
+    {
+      "epoch": 0.9274561030439273,
+      "grad_norm": 0.17127452790737152,
+      "learning_rate": 1.3029150353118247e-06,
+      "loss": 1.0707,
+      "step": 15265
+    },
+    {
+      "epoch": 0.9275168600765539,
+      "grad_norm": 0.12455254793167114,
+      "learning_rate": 1.3007448371181897e-06,
+      "loss": 1.0833,
+      "step": 15266
+    },
+    {
+      "epoch": 0.9275776171091804,
+      "grad_norm": 0.11031661182641983,
+      "learning_rate": 1.29857642400516e-06,
+      "loss": 1.0643,
+      "step": 15267
+    },
+    {
+      "epoch": 0.9276383741418069,
+      "grad_norm": 0.13668157160282135,
+      "learning_rate": 1.2964097960522225e-06,
+      "loss": 1.0638,
+      "step": 15268
+    },
+    {
+      "epoch": 0.9276991311744335,
+      "grad_norm": 0.1880815476179123,
+      "learning_rate": 1.2942449533388078e-06,
+      "loss": 1.0867,
+      "step": 15269
+    },
+    {
+      "epoch": 0.92775988820706,
+      "grad_norm": 0.13995423913002014,
+      "learning_rate": 1.2920818959442526e-06,
+      "loss": 1.056,
+      "step": 15270
+    },
+    {
+      "epoch": 0.9278206452396865,
+      "grad_norm": 0.17679435014724731,
+      "learning_rate": 1.2899206239478545e-06,
+      "loss": 1.0798,
+      "step": 15271
+    },
+    {
+      "epoch": 0.927881402272313,
+      "grad_norm": 0.11338049918413162,
+      "learning_rate": 1.287761137428828e-06,
+      "loss": 1.031,
+      "step": 15272
+    },
+    {
+      "epoch": 0.9279421593049395,
+      "grad_norm": 0.13410165905952454,
+      "learning_rate": 1.2856034364663372e-06,
+      "loss": 1.0715,
+      "step": 15273
+    },
+    {
+      "epoch": 0.928002916337566,
+      "grad_norm": 0.1318570226430893,
+      "learning_rate": 1.2834475211394692e-06,
+      "loss": 1.0474,
+      "step": 15274
+    },
+    {
+      "epoch": 0.9280636733701926,
+      "grad_norm": 0.11561030149459839,
+      "learning_rate": 1.281293391527244e-06,
+      "loss": 1.0798,
+      "step": 15275
+    },
+    {
+      "epoch": 0.9281244304028191,
+      "grad_norm": 0.14250938594341278,
+      "learning_rate": 1.2791410477086319e-06,
+      "loss": 1.0462,
+      "step": 15276
+    },
+    {
+      "epoch": 0.9281851874354456,
+      "grad_norm": 0.18573889136314392,
+      "learning_rate": 1.276990489762514e-06,
+      "loss": 1.0046,
+      "step": 15277
+    },
+    {
+      "epoch": 0.9282459444680722,
+      "grad_norm": 1.8279820680618286,
+      "learning_rate": 1.274841717767733e-06,
+      "loss": 1.0845,
+      "step": 15278
+    },
+    {
+      "epoch": 0.9283067015006987,
+      "grad_norm": 0.8750011920928955,
+      "learning_rate": 1.272694731803048e-06,
+      "loss": 1.1081,
+      "step": 15279
+    },
+    {
+      "epoch": 0.9283674585333253,
+      "grad_norm": 0.2108527570962906,
+      "learning_rate": 1.2705495319471573e-06,
+      "loss": 0.9953,
+      "step": 15280
+    },
+    {
+      "epoch": 0.9284282155659518,
+      "grad_norm": 0.2060713768005371,
+      "learning_rate": 1.2684061182786922e-06,
+      "loss": 1.0865,
+      "step": 15281
+    },
+    {
+      "epoch": 0.9284889725985783,
+      "grad_norm": 0.1269395649433136,
+      "learning_rate": 1.2662644908762179e-06,
+      "loss": 1.069,
+      "step": 15282
+    },
+    {
+      "epoch": 0.9285497296312049,
+      "grad_norm": 0.14588280022144318,
+      "learning_rate": 1.264124649818238e-06,
+      "loss": 1.0895,
+      "step": 15283
+    },
+    {
+      "epoch": 0.9286104866638314,
+      "grad_norm": 0.1509881317615509,
+      "learning_rate": 1.26198659518319e-06,
+      "loss": 1.1111,
+      "step": 15284
+    },
+    {
+      "epoch": 0.9286712436964578,
+      "grad_norm": 0.11926669627428055,
+      "learning_rate": 1.25985032704945e-06,
+      "loss": 0.9683,
+      "step": 15285
+    },
+    {
+      "epoch": 0.9287320007290844,
+      "grad_norm": 0.49089598655700684,
+      "learning_rate": 1.2577158454953108e-06,
+      "loss": 1.0998,
+      "step": 15286
+    },
+    {
+      "epoch": 0.9287927577617109,
+      "grad_norm": 0.1705758273601532,
+      "learning_rate": 1.2555831505990267e-06,
+      "loss": 1.0665,
+      "step": 15287
+    },
+    {
+      "epoch": 0.9288535147943374,
+      "grad_norm": 0.13350285589694977,
+      "learning_rate": 1.2534522424387518e-06,
+      "loss": 1.0435,
+      "step": 15288
+    },
+    {
+      "epoch": 0.928914271826964,
+      "grad_norm": 0.12272652238607407,
+      "learning_rate": 1.2513231210926124e-06,
+      "loss": 1.0346,
+      "step": 15289
+    },
+    {
+      "epoch": 0.9289750288595905,
+      "grad_norm": 0.22617629170417786,
+      "learning_rate": 1.2491957866386516e-06,
+      "loss": 1.1836,
+      "step": 15290
+    },
+    {
+      "epoch": 0.929035785892217,
+      "grad_norm": 0.14095793664455414,
+      "learning_rate": 1.2470702391548406e-06,
+      "loss": 1.1044,
+      "step": 15291
+    },
+    {
+      "epoch": 0.9290965429248436,
+      "grad_norm": 0.18492470681667328,
+      "learning_rate": 1.2449464787191e-06,
+      "loss": 1.2188,
+      "step": 15292
+    },
+    {
+      "epoch": 0.9291572999574701,
+      "grad_norm": 0.2547471225261688,
+      "learning_rate": 1.242824505409257e-06,
+      "loss": 1.2218,
+      "step": 15293
+    },
+    {
+      "epoch": 0.9292180569900966,
+      "grad_norm": 0.22714532911777496,
+      "learning_rate": 1.2407043193031098e-06,
+      "loss": 1.0555,
+      "step": 15294
+    },
+    {
+      "epoch": 0.9292788140227232,
+      "grad_norm": 0.3288191556930542,
+      "learning_rate": 1.2385859204783746e-06,
+      "loss": 1.0663,
+      "step": 15295
+    },
+    {
+      "epoch": 0.9293395710553497,
+      "grad_norm": 0.13069987297058105,
+      "learning_rate": 1.2364693090126999e-06,
+      "loss": 1.0489,
+      "step": 15296
+    },
+    {
+      "epoch": 0.9294003280879762,
+      "grad_norm": 0.10086257755756378,
+      "learning_rate": 1.234354484983663e-06,
+      "loss": 1.047,
+      "step": 15297
+    },
+    {
+      "epoch": 0.9294610851206027,
+      "grad_norm": 0.0976250097155571,
+      "learning_rate": 1.2322414484687904e-06,
+      "loss": 1.0175,
+      "step": 15298
+    },
+    {
+      "epoch": 0.9295218421532292,
+      "grad_norm": 0.13630108535289764,
+      "learning_rate": 1.2301301995455371e-06,
+      "loss": 1.0772,
+      "step": 15299
+    },
+    {
+      "epoch": 0.9295825991858557,
+      "grad_norm": 0.22394976019859314,
+      "learning_rate": 1.2280207382912855e-06,
+      "loss": 1.0533,
+      "step": 15300
+    },
+    {
+      "epoch": 0.9296433562184823,
+      "grad_norm": 0.1796313226222992,
+      "learning_rate": 1.2259130647833627e-06,
+      "loss": 1.1555,
+      "step": 15301
+    },
+    {
+      "epoch": 0.9297041132511088,
+      "grad_norm": 0.13305111229419708,
+      "learning_rate": 1.2238071790990236e-06,
+      "loss": 1.0099,
+      "step": 15302
+    },
+    {
+      "epoch": 0.9297648702837353,
+      "grad_norm": 0.15687556564807892,
+      "learning_rate": 1.2217030813154617e-06,
+      "loss": 1.1247,
+      "step": 15303
+    },
+    {
+      "epoch": 0.9298256273163619,
+      "grad_norm": 0.12067729979753494,
+      "learning_rate": 1.2196007715097934e-06,
+      "loss": 1.0598,
+      "step": 15304
+    },
+    {
+      "epoch": 0.9298863843489884,
+      "grad_norm": 0.12093468010425568,
+      "learning_rate": 1.2175002497591016e-06,
+      "loss": 1.0716,
+      "step": 15305
+    },
+    {
+      "epoch": 0.9299471413816149,
+      "grad_norm": 1.464052677154541,
+      "learning_rate": 1.2154015161403631e-06,
+      "loss": 0.9718,
+      "step": 15306
+    },
+    {
+      "epoch": 0.9300078984142415,
+      "grad_norm": 0.20744918286800385,
+      "learning_rate": 1.2133045707305112e-06,
+      "loss": 1.1986,
+      "step": 15307
+    },
+    {
+      "epoch": 0.930068655446868,
+      "grad_norm": 0.13889753818511963,
+      "learning_rate": 1.2112094136064067e-06,
+      "loss": 1.1535,
+      "step": 15308
+    },
+    {
+      "epoch": 0.9301294124794945,
+      "grad_norm": 0.2360927164554596,
+      "learning_rate": 1.209116044844849e-06,
+      "loss": 1.0099,
+      "step": 15309
+    },
+    {
+      "epoch": 0.9301901695121211,
+      "grad_norm": 0.45913657546043396,
+      "learning_rate": 1.2070244645225825e-06,
+      "loss": 1.2555,
+      "step": 15310
+    },
+    {
+      "epoch": 0.9302509265447475,
+      "grad_norm": 0.09064749628305435,
+      "learning_rate": 1.2049346727162626e-06,
+      "loss": 0.9981,
+      "step": 15311
+    },
+    {
+      "epoch": 0.930311683577374,
+      "grad_norm": 0.17532262206077576,
+      "learning_rate": 1.2028466695024942e-06,
+      "loss": 1.1039,
+      "step": 15312
+    },
+    {
+      "epoch": 0.9303724406100006,
+      "grad_norm": 0.1494452804327011,
+      "learning_rate": 1.2007604549578167e-06,
+      "loss": 1.076,
+      "step": 15313
+    },
+    {
+      "epoch": 0.9304331976426271,
+      "grad_norm": 0.21924225986003876,
+      "learning_rate": 1.1986760291586908e-06,
+      "loss": 1.142,
+      "step": 15314
+    },
+    {
+      "epoch": 0.9304939546752536,
+      "grad_norm": 0.09638996422290802,
+      "learning_rate": 1.1965933921815275e-06,
+      "loss": 1.0379,
+      "step": 15315
+    },
+    {
+      "epoch": 0.9305547117078802,
+      "grad_norm": 0.11173415929079056,
+      "learning_rate": 1.1945125441026718e-06,
+      "loss": 1.0251,
+      "step": 15316
+    },
+    {
+      "epoch": 0.9306154687405067,
+      "grad_norm": 0.12360332906246185,
+      "learning_rate": 1.1924334849983898e-06,
+      "loss": 1.0203,
+      "step": 15317
+    },
+    {
+      "epoch": 0.9306762257731332,
+      "grad_norm": 0.10610894858837128,
+      "learning_rate": 1.1903562149448988e-06,
+      "loss": 1.0322,
+      "step": 15318
+    },
+    {
+      "epoch": 0.9307369828057598,
+      "grad_norm": 0.10047803074121475,
+      "learning_rate": 1.1882807340183265e-06,
+      "loss": 0.9895,
+      "step": 15319
+    },
+    {
+      "epoch": 0.9307977398383863,
+      "grad_norm": 0.11373595893383026,
+      "learning_rate": 1.186207042294768e-06,
+      "loss": 1.0652,
+      "step": 15320
+    },
+    {
+      "epoch": 0.9308584968710129,
+      "grad_norm": 0.1170571967959404,
+      "learning_rate": 1.1841351398502177e-06,
+      "loss": 1.0474,
+      "step": 15321
+    },
+    {
+      "epoch": 0.9309192539036394,
+      "grad_norm": 0.16566219925880432,
+      "learning_rate": 1.182065026760637e-06,
+      "loss": 1.0283,
+      "step": 15322
+    },
+    {
+      "epoch": 0.9309800109362659,
+      "grad_norm": 0.9753913879394531,
+      "learning_rate": 1.1799967031018933e-06,
+      "loss": 1.2524,
+      "step": 15323
+    },
+    {
+      "epoch": 0.9310407679688923,
+      "grad_norm": 0.19697614014148712,
+      "learning_rate": 1.1779301689498091e-06,
+      "loss": 1.1466,
+      "step": 15324
+    },
+    {
+      "epoch": 0.9311015250015189,
+      "grad_norm": 0.09586262702941895,
+      "learning_rate": 1.1758654243801348e-06,
+      "loss": 1.0151,
+      "step": 15325
+    },
+    {
+      "epoch": 0.9311622820341454,
+      "grad_norm": 0.10872571170330048,
+      "learning_rate": 1.1738024694685435e-06,
+      "loss": 1.0008,
+      "step": 15326
+    },
+    {
+      "epoch": 0.931223039066772,
+      "grad_norm": 0.8605412840843201,
+      "learning_rate": 1.171741304290669e-06,
+      "loss": 1.1964,
+      "step": 15327
+    },
+    {
+      "epoch": 0.9312837960993985,
+      "grad_norm": 0.10982364416122437,
+      "learning_rate": 1.1696819289220507e-06,
+      "loss": 1.0272,
+      "step": 15328
+    },
+    {
+      "epoch": 0.931344553132025,
+      "grad_norm": 0.14758697152137756,
+      "learning_rate": 1.1676243434381783e-06,
+      "loss": 1.0833,
+      "step": 15329
+    },
+    {
+      "epoch": 0.9314053101646516,
+      "grad_norm": 0.1348530799150467,
+      "learning_rate": 1.165568547914475e-06,
+      "loss": 1.0688,
+      "step": 15330
+    },
+    {
+      "epoch": 0.9314660671972781,
+      "grad_norm": 0.16634732484817505,
+      "learning_rate": 1.1635145424262973e-06,
+      "loss": 1.1161,
+      "step": 15331
+    },
+    {
+      "epoch": 0.9315268242299046,
+      "grad_norm": 0.30785343050956726,
+      "learning_rate": 1.16146232704894e-06,
+      "loss": 1.1312,
+      "step": 15332
+    },
+    {
+      "epoch": 0.9315875812625312,
+      "grad_norm": 0.15048499405384064,
+      "learning_rate": 1.1594119018576155e-06,
+      "loss": 1.0915,
+      "step": 15333
+    },
+    {
+      "epoch": 0.9316483382951577,
+      "grad_norm": 0.7665331363677979,
+      "learning_rate": 1.157363266927486e-06,
+      "loss": 1.0658,
+      "step": 15334
+    },
+    {
+      "epoch": 0.9317090953277842,
+      "grad_norm": 0.10684623569250107,
+      "learning_rate": 1.1553164223336465e-06,
+      "loss": 1.0454,
+      "step": 15335
+    },
+    {
+      "epoch": 0.9317698523604108,
+      "grad_norm": 0.10057730972766876,
+      "learning_rate": 1.153271368151132e-06,
+      "loss": 1.028,
+      "step": 15336
+    },
+    {
+      "epoch": 0.9318306093930372,
+      "grad_norm": 0.21029414236545563,
+      "learning_rate": 1.1512281044548933e-06,
+      "loss": 1.1419,
+      "step": 15337
+    },
+    {
+      "epoch": 0.9318913664256637,
+      "grad_norm": 0.19378623366355896,
+      "learning_rate": 1.1491866313198318e-06,
+      "loss": 1.0342,
+      "step": 15338
+    },
+    {
+      "epoch": 0.9319521234582903,
+      "grad_norm": 0.19587653875350952,
+      "learning_rate": 1.1471469488207765e-06,
+      "loss": 1.0876,
+      "step": 15339
+    },
+    {
+      "epoch": 0.9320128804909168,
+      "grad_norm": 0.2780442535877228,
+      "learning_rate": 1.1451090570324896e-06,
+      "loss": 1.0296,
+      "step": 15340
+    },
+    {
+      "epoch": 0.9320736375235433,
+      "grad_norm": 0.33275866508483887,
+      "learning_rate": 1.1430729560296782e-06,
+      "loss": 1.1214,
+      "step": 15341
+    },
+    {
+      "epoch": 0.9321343945561699,
+      "grad_norm": 0.1182708591222763,
+      "learning_rate": 1.1410386458869715e-06,
+      "loss": 1.0326,
+      "step": 15342
+    },
+    {
+      "epoch": 0.9321951515887964,
+      "grad_norm": 0.16430284082889557,
+      "learning_rate": 1.139006126678932e-06,
+      "loss": 1.0288,
+      "step": 15343
+    },
+    {
+      "epoch": 0.9322559086214229,
+      "grad_norm": 2.208350896835327,
+      "learning_rate": 1.1369753984800724e-06,
+      "loss": 1.1012,
+      "step": 15344
+    },
+    {
+      "epoch": 0.9323166656540495,
+      "grad_norm": 0.17178237438201904,
+      "learning_rate": 1.1349464613648275e-06,
+      "loss": 1.0325,
+      "step": 15345
+    },
+    {
+      "epoch": 0.932377422686676,
+      "grad_norm": 0.12913119792938232,
+      "learning_rate": 1.1329193154075601e-06,
+      "loss": 1.0632,
+      "step": 15346
+    },
+    {
+      "epoch": 0.9324381797193025,
+      "grad_norm": 0.14151978492736816,
+      "learning_rate": 1.130893960682583e-06,
+      "loss": 1.0634,
+      "step": 15347
+    },
+    {
+      "epoch": 0.9324989367519291,
+      "grad_norm": 0.10698746889829636,
+      "learning_rate": 1.1288703972641313e-06,
+      "loss": 1.037,
+      "step": 15348
+    },
+    {
+      "epoch": 0.9325596937845556,
+      "grad_norm": 0.12261773645877838,
+      "learning_rate": 1.1268486252263844e-06,
+      "loss": 0.9918,
+      "step": 15349
+    },
+    {
+      "epoch": 0.932620450817182,
+      "grad_norm": 0.19454126060009003,
+      "learning_rate": 1.1248286446434498e-06,
+      "loss": 1.058,
+      "step": 15350
+    },
+    {
+      "epoch": 0.9326812078498086,
+      "grad_norm": 2.1978588104248047,
+      "learning_rate": 1.1228104555893625e-06,
+      "loss": 1.1101,
+      "step": 15351
+    },
+    {
+      "epoch": 0.9327419648824351,
+      "grad_norm": 0.16989226639270782,
+      "learning_rate": 1.1207940581381138e-06,
+      "loss": 1.1179,
+      "step": 15352
+    },
+    {
+      "epoch": 0.9328027219150616,
+      "grad_norm": 0.10813550651073456,
+      "learning_rate": 1.1187794523635997e-06,
+      "loss": 1.0238,
+      "step": 15353
+    },
+    {
+      "epoch": 0.9328634789476882,
+      "grad_norm": 0.10100190341472626,
+      "learning_rate": 1.116766638339678e-06,
+      "loss": 1.0579,
+      "step": 15354
+    },
+    {
+      "epoch": 0.9329242359803147,
+      "grad_norm": 0.10340183973312378,
+      "learning_rate": 1.114755616140123e-06,
+      "loss": 1.0037,
+      "step": 15355
+    },
+    {
+      "epoch": 0.9329849930129412,
+      "grad_norm": 0.13751733303070068,
+      "learning_rate": 1.1127463858386477e-06,
+      "loss": 1.0692,
+      "step": 15356
+    },
+    {
+      "epoch": 0.9330457500455678,
+      "grad_norm": 0.151703879237175,
+      "learning_rate": 1.1107389475089048e-06,
+      "loss": 1.0709,
+      "step": 15357
+    },
+    {
+      "epoch": 0.9331065070781943,
+      "grad_norm": 0.1182345524430275,
+      "learning_rate": 1.1087333012244737e-06,
+      "loss": 1.0159,
+      "step": 15358
+    },
+    {
+      "epoch": 0.9331672641108208,
+      "grad_norm": 0.1722113937139511,
+      "learning_rate": 1.1067294470588852e-06,
+      "loss": 1.0687,
+      "step": 15359
+    },
+    {
+      "epoch": 0.9332280211434474,
+      "grad_norm": 0.13528689742088318,
+      "learning_rate": 1.1047273850855689e-06,
+      "loss": 1.0531,
+      "step": 15360
+    },
+    {
+      "epoch": 0.9332887781760739,
+      "grad_norm": 0.10084126889705658,
+      "learning_rate": 1.1027271153779218e-06,
+      "loss": 0.9678,
+      "step": 15361
+    },
+    {
+      "epoch": 0.9333495352087005,
+      "grad_norm": 0.09768477827310562,
+      "learning_rate": 1.100728638009263e-06,
+      "loss": 1.0103,
+      "step": 15362
+    },
+    {
+      "epoch": 0.933410292241327,
+      "grad_norm": 0.11217644810676575,
+      "learning_rate": 1.0987319530528506e-06,
+      "loss": 1.0143,
+      "step": 15363
+    },
+    {
+      "epoch": 0.9334710492739534,
+      "grad_norm": 0.38295334577560425,
+      "learning_rate": 1.0967370605818705e-06,
+      "loss": 1.0166,
+      "step": 15364
+    },
+    {
+      "epoch": 0.93353180630658,
+      "grad_norm": 0.18921901285648346,
+      "learning_rate": 1.0947439606694532e-06,
+      "loss": 1.1728,
+      "step": 15365
+    },
+    {
+      "epoch": 0.9335925633392065,
+      "grad_norm": 1.356097936630249,
+      "learning_rate": 1.0927526533886401e-06,
+      "loss": 1.1471,
+      "step": 15366
+    },
+    {
+      "epoch": 0.933653320371833,
+      "grad_norm": 0.10401421785354614,
+      "learning_rate": 1.0907631388124339e-06,
+      "loss": 1.0296,
+      "step": 15367
+    },
+    {
+      "epoch": 0.9337140774044596,
+      "grad_norm": 0.1892024278640747,
+      "learning_rate": 1.0887754170137598e-06,
+      "loss": 1.1636,
+      "step": 15368
+    },
+    {
+      "epoch": 0.9337748344370861,
+      "grad_norm": 0.42177772521972656,
+      "learning_rate": 1.0867894880654762e-06,
+      "loss": 1.0814,
+      "step": 15369
+    },
+    {
+      "epoch": 0.9338355914697126,
+      "grad_norm": 0.15873512625694275,
+      "learning_rate": 1.08480535204038e-06,
+      "loss": 1.1007,
+      "step": 15370
+    },
+    {
+      "epoch": 0.9338963485023392,
+      "grad_norm": 1.713375210762024,
+      "learning_rate": 1.0828230090111969e-06,
+      "loss": 1.0644,
+      "step": 15371
+    },
+    {
+      "epoch": 0.9339571055349657,
+      "grad_norm": 0.14453130960464478,
+      "learning_rate": 1.080842459050596e-06,
+      "loss": 1.0471,
+      "step": 15372
+    },
+    {
+      "epoch": 0.9340178625675922,
+      "grad_norm": 0.32694312930107117,
+      "learning_rate": 1.0788637022311642e-06,
+      "loss": 1.1044,
+      "step": 15373
+    },
+    {
+      "epoch": 0.9340786196002188,
+      "grad_norm": 0.11188525706529617,
+      "learning_rate": 1.0768867386254434e-06,
+      "loss": 1.0735,
+      "step": 15374
+    },
+    {
+      "epoch": 0.9341393766328453,
+      "grad_norm": 0.3183089792728424,
+      "learning_rate": 1.0749115683058974e-06,
+      "loss": 1.2225,
+      "step": 15375
+    },
+    {
+      "epoch": 0.9342001336654718,
+      "grad_norm": 0.15144848823547363,
+      "learning_rate": 1.0729381913449244e-06,
+      "loss": 0.9922,
+      "step": 15376
+    },
+    {
+      "epoch": 0.9342608906980983,
+      "grad_norm": 0.21125288307666779,
+      "learning_rate": 1.0709666078148551e-06,
+      "loss": 1.0736,
+      "step": 15377
+    },
+    {
+      "epoch": 0.9343216477307248,
+      "grad_norm": 0.22160610556602478,
+      "learning_rate": 1.068996817787965e-06,
+      "loss": 1.2534,
+      "step": 15378
+    },
+    {
+      "epoch": 0.9343824047633513,
+      "grad_norm": 0.10727578401565552,
+      "learning_rate": 1.0670288213364577e-06,
+      "loss": 1.0171,
+      "step": 15379
+    },
+    {
+      "epoch": 0.9344431617959779,
+      "grad_norm": 0.1001729741692543,
+      "learning_rate": 1.065062618532464e-06,
+      "loss": 0.9914,
+      "step": 15380
+    },
+    {
+      "epoch": 0.9345039188286044,
+      "grad_norm": 0.09848897904157639,
+      "learning_rate": 1.06309820944806e-06,
+      "loss": 1.0295,
+      "step": 15381
+    },
+    {
+      "epoch": 0.9345646758612309,
+      "grad_norm": 0.1327124387025833,
+      "learning_rate": 1.0611355941552491e-06,
+      "loss": 1.0136,
+      "step": 15382
+    },
+    {
+      "epoch": 0.9346254328938575,
+      "grad_norm": 0.15134766697883606,
+      "learning_rate": 1.0591747727259739e-06,
+      "loss": 1.0681,
+      "step": 15383
+    },
+    {
+      "epoch": 0.934686189926484,
+      "grad_norm": 1.4509658813476562,
+      "learning_rate": 1.0572157452321097e-06,
+      "loss": 1.0549,
+      "step": 15384
+    },
+    {
+      "epoch": 0.9347469469591105,
+      "grad_norm": 0.11305056512355804,
+      "learning_rate": 1.0552585117454606e-06,
+      "loss": 1.0669,
+      "step": 15385
+    },
+    {
+      "epoch": 0.9348077039917371,
+      "grad_norm": 0.13011938333511353,
+      "learning_rate": 1.053303072337769e-06,
+      "loss": 1.0947,
+      "step": 15386
+    },
+    {
+      "epoch": 0.9348684610243636,
+      "grad_norm": 0.17125792801380157,
+      "learning_rate": 1.0513494270807168e-06,
+      "loss": 1.1389,
+      "step": 15387
+    },
+    {
+      "epoch": 0.9349292180569901,
+      "grad_norm": 0.11153468489646912,
+      "learning_rate": 1.049397576045913e-06,
+      "loss": 0.9935,
+      "step": 15388
+    },
+    {
+      "epoch": 0.9349899750896167,
+      "grad_norm": 0.16348513960838318,
+      "learning_rate": 1.047447519304906e-06,
+      "loss": 1.1284,
+      "step": 15389
+    },
+    {
+      "epoch": 0.9350507321222431,
+      "grad_norm": 0.29993122816085815,
+      "learning_rate": 1.0454992569291666e-06,
+      "loss": 1.1022,
+      "step": 15390
+    },
+    {
+      "epoch": 0.9351114891548696,
+      "grad_norm": 0.12200228124856949,
+      "learning_rate": 1.0435527889901208e-06,
+      "loss": 1.0583,
+      "step": 15391
+    },
+    {
+      "epoch": 0.9351722461874962,
+      "grad_norm": 0.21584726870059967,
+      "learning_rate": 1.0416081155591061e-06,
+      "loss": 1.1769,
+      "step": 15392
+    },
+    {
+      "epoch": 0.9352330032201227,
+      "grad_norm": 0.18006779253482819,
+      "learning_rate": 1.03966523670741e-06,
+      "loss": 1.1844,
+      "step": 15393
+    },
+    {
+      "epoch": 0.9352937602527492,
+      "grad_norm": 0.13245415687561035,
+      "learning_rate": 1.0377241525062531e-06,
+      "loss": 1.0496,
+      "step": 15394
+    },
+    {
+      "epoch": 0.9353545172853758,
+      "grad_norm": 0.25103774666786194,
+      "learning_rate": 1.035784863026773e-06,
+      "loss": 1.1934,
+      "step": 15395
+    },
+    {
+      "epoch": 0.9354152743180023,
+      "grad_norm": 0.16308242082595825,
+      "learning_rate": 1.0338473683400685e-06,
+      "loss": 1.1009,
+      "step": 15396
+    },
+    {
+      "epoch": 0.9354760313506288,
+      "grad_norm": 0.19510650634765625,
+      "learning_rate": 1.0319116685171492e-06,
+      "loss": 1.108,
+      "step": 15397
+    },
+    {
+      "epoch": 0.9355367883832554,
+      "grad_norm": 0.20332162082195282,
+      "learning_rate": 1.0299777636289808e-06,
+      "loss": 1.1157,
+      "step": 15398
+    },
+    {
+      "epoch": 0.9355975454158819,
+      "grad_norm": 0.14856691658496857,
+      "learning_rate": 1.02804565374644e-06,
+      "loss": 1.0897,
+      "step": 15399
+    },
+    {
+      "epoch": 0.9356583024485084,
+      "grad_norm": 0.09579699486494064,
+      "learning_rate": 1.0261153389403533e-06,
+      "loss": 0.9748,
+      "step": 15400
+    },
+    {
+      "epoch": 0.935719059481135,
+      "grad_norm": 0.15310057997703552,
+      "learning_rate": 1.0241868192814752e-06,
+      "loss": 1.2507,
+      "step": 15401
+    },
+    {
+      "epoch": 0.9357798165137615,
+      "grad_norm": 0.11506348848342896,
+      "learning_rate": 1.0222600948404882e-06,
+      "loss": 1.0618,
+      "step": 15402
+    },
+    {
+      "epoch": 0.9358405735463879,
+      "grad_norm": 0.0961810052394867,
+      "learning_rate": 1.02033516568803e-06,
+      "loss": 1.086,
+      "step": 15403
+    },
+    {
+      "epoch": 0.9359013305790145,
+      "grad_norm": 0.10804151743650436,
+      "learning_rate": 1.0184120318946554e-06,
+      "loss": 1.0123,
+      "step": 15404
+    },
+    {
+      "epoch": 0.935962087611641,
+      "grad_norm": 0.31057628989219666,
+      "learning_rate": 1.0164906935308526e-06,
+      "loss": 1.0942,
+      "step": 15405
+    },
+    {
+      "epoch": 0.9360228446442675,
+      "grad_norm": 0.12351847440004349,
+      "learning_rate": 1.0145711506670596e-06,
+      "loss": 1.003,
+      "step": 15406
+    },
+    {
+      "epoch": 0.9360836016768941,
+      "grad_norm": 0.1282394677400589,
+      "learning_rate": 1.0126534033736202e-06,
+      "loss": 1.1411,
+      "step": 15407
+    },
+    {
+      "epoch": 0.9361443587095206,
+      "grad_norm": 0.19097188115119934,
+      "learning_rate": 1.0107374517208447e-06,
+      "loss": 1.0953,
+      "step": 15408
+    },
+    {
+      "epoch": 0.9362051157421472,
+      "grad_norm": 0.20312348008155823,
+      "learning_rate": 1.0088232957789545e-06,
+      "loss": 1.0992,
+      "step": 15409
+    },
+    {
+      "epoch": 0.9362658727747737,
+      "grad_norm": 0.13771170377731323,
+      "learning_rate": 1.0069109356181162e-06,
+      "loss": 1.1084,
+      "step": 15410
+    },
+    {
+      "epoch": 0.9363266298074002,
+      "grad_norm": 0.10973066091537476,
+      "learning_rate": 1.005000371308429e-06,
+      "loss": 1.0339,
+      "step": 15411
+    },
+    {
+      "epoch": 0.9363873868400268,
+      "grad_norm": 0.12341954559087753,
+      "learning_rate": 1.0030916029199256e-06,
+      "loss": 1.0248,
+      "step": 15412
+    },
+    {
+      "epoch": 0.9364481438726533,
+      "grad_norm": 0.11936056613922119,
+      "learning_rate": 1.001184630522567e-06,
+      "loss": 1.0348,
+      "step": 15413
+    },
+    {
+      "epoch": 0.9365089009052798,
+      "grad_norm": 0.17877939343452454,
+      "learning_rate": 9.992794541862582e-07,
+      "loss": 1.0628,
+      "step": 15414
+    },
+    {
+      "epoch": 0.9365696579379064,
+      "grad_norm": 0.16074557602405548,
+      "learning_rate": 9.973760739808325e-07,
+      "loss": 1.1226,
+      "step": 15415
+    },
+    {
+      "epoch": 0.9366304149705328,
+      "grad_norm": 0.29493623971939087,
+      "learning_rate": 9.954744899760616e-07,
+      "loss": 1.2371,
+      "step": 15416
+    },
+    {
+      "epoch": 0.9366911720031593,
+      "grad_norm": 1.5070557594299316,
+      "learning_rate": 9.935747022416397e-07,
+      "loss": 1.0696,
+      "step": 15417
+    },
+    {
+      "epoch": 0.9367519290357859,
+      "grad_norm": 0.12263424694538116,
+      "learning_rate": 9.916767108472114e-07,
+      "loss": 1.0877,
+      "step": 15418
+    },
+    {
+      "epoch": 0.9368126860684124,
+      "grad_norm": 0.11993416398763657,
+      "learning_rate": 9.89780515862354e-07,
+      "loss": 1.0615,
+      "step": 15419
+    },
+    {
+      "epoch": 0.9368734431010389,
+      "grad_norm": 0.12146300077438354,
+      "learning_rate": 9.878861173565569e-07,
+      "loss": 1.0379,
+      "step": 15420
+    },
+    {
+      "epoch": 0.9369342001336655,
+      "grad_norm": 0.11891285330057144,
+      "learning_rate": 9.859935153992694e-07,
+      "loss": 1.0215,
+      "step": 15421
+    },
+    {
+      "epoch": 0.936994957166292,
+      "grad_norm": 0.14614197611808777,
+      "learning_rate": 9.841027100598699e-07,
+      "loss": 1.1099,
+      "step": 15422
+    },
+    {
+      "epoch": 0.9370557141989185,
+      "grad_norm": 0.20706407725811005,
+      "learning_rate": 9.82213701407647e-07,
+      "loss": 1.1478,
+      "step": 15423
+    },
+    {
+      "epoch": 0.9371164712315451,
+      "grad_norm": 0.19839854538440704,
+      "learning_rate": 9.80326489511868e-07,
+      "loss": 1.2426,
+      "step": 15424
+    },
+    {
+      "epoch": 0.9371772282641716,
+      "grad_norm": 0.158394455909729,
+      "learning_rate": 9.784410744416882e-07,
+      "loss": 1.1307,
+      "step": 15425
+    },
+    {
+      "epoch": 0.9372379852967981,
+      "grad_norm": 0.468840092420578,
+      "learning_rate": 9.765574562662362e-07,
+      "loss": 1.0582,
+      "step": 15426
+    },
+    {
+      "epoch": 0.9372987423294247,
+      "grad_norm": 0.25128600001335144,
+      "learning_rate": 9.746756350545449e-07,
+      "loss": 1.1505,
+      "step": 15427
+    },
+    {
+      "epoch": 0.9373594993620512,
+      "grad_norm": 0.10685040801763535,
+      "learning_rate": 9.727956108755875e-07,
+      "loss": 1.0198,
+      "step": 15428
+    },
+    {
+      "epoch": 0.9374202563946776,
+      "grad_norm": 0.09054891765117645,
+      "learning_rate": 9.709173837982866e-07,
+      "loss": 1.039,
+      "step": 15429
+    },
+    {
+      "epoch": 0.9374810134273042,
+      "grad_norm": 0.12794844806194305,
+      "learning_rate": 9.690409538914869e-07,
+      "loss": 1.0365,
+      "step": 15430
+    },
+    {
+      "epoch": 0.9375417704599307,
+      "grad_norm": 0.2145228534936905,
+      "learning_rate": 9.671663212239667e-07,
+      "loss": 1.0805,
+      "step": 15431
+    },
+    {
+      "epoch": 0.9376025274925572,
+      "grad_norm": 0.13872674107551575,
+      "learning_rate": 9.65293485864449e-07,
+      "loss": 1.093,
+      "step": 15432
+    },
+    {
+      "epoch": 0.9376632845251838,
+      "grad_norm": 0.09915322065353394,
+      "learning_rate": 9.63422447881568e-07,
+      "loss": 1.0321,
+      "step": 15433
+    },
+    {
+      "epoch": 0.9377240415578103,
+      "grad_norm": 0.19035974144935608,
+      "learning_rate": 9.615532073439238e-07,
+      "loss": 1.1623,
+      "step": 15434
+    },
+    {
+      "epoch": 0.9377847985904368,
+      "grad_norm": 0.10336155444383621,
+      "learning_rate": 9.59685764320023e-07,
+      "loss": 1.0581,
+      "step": 15435
+    },
+    {
+      "epoch": 0.9378455556230634,
+      "grad_norm": 3.234569549560547,
+      "learning_rate": 9.578201188783165e-07,
+      "loss": 1.1341,
+      "step": 15436
+    },
+    {
+      "epoch": 0.9379063126556899,
+      "grad_norm": 3.544027805328369,
+      "learning_rate": 9.55956271087194e-07,
+      "loss": 1.0675,
+      "step": 15437
+    },
+    {
+      "epoch": 0.9379670696883164,
+      "grad_norm": 0.14697876572608948,
+      "learning_rate": 9.540942210149783e-07,
+      "loss": 1.0665,
+      "step": 15438
+    },
+    {
+      "epoch": 0.938027826720943,
+      "grad_norm": 0.3609788119792938,
+      "learning_rate": 9.522339687299153e-07,
+      "loss": 1.1431,
+      "step": 15439
+    },
+    {
+      "epoch": 0.9380885837535695,
+      "grad_norm": 0.11817152053117752,
+      "learning_rate": 9.503755143002002e-07,
+      "loss": 1.0406,
+      "step": 15440
+    },
+    {
+      "epoch": 0.938149340786196,
+      "grad_norm": 0.16668759286403656,
+      "learning_rate": 9.485188577939508e-07,
+      "loss": 1.1371,
+      "step": 15441
+    },
+    {
+      "epoch": 0.9382100978188225,
+      "grad_norm": 2.793811321258545,
+      "learning_rate": 9.466639992792237e-07,
+      "loss": 1.0374,
+      "step": 15442
+    },
+    {
+      "epoch": 0.938270854851449,
+      "grad_norm": 0.1596498191356659,
+      "learning_rate": 9.448109388240089e-07,
+      "loss": 1.0866,
+      "step": 15443
+    },
+    {
+      "epoch": 0.9383316118840755,
+      "grad_norm": 0.12353494763374329,
+      "learning_rate": 9.429596764962245e-07,
+      "loss": 1.0382,
+      "step": 15444
+    },
+    {
+      "epoch": 0.9383923689167021,
+      "grad_norm": 0.11589280515909195,
+      "learning_rate": 9.411102123637439e-07,
+      "loss": 1.0551,
+      "step": 15445
+    },
+    {
+      "epoch": 0.9384531259493286,
+      "grad_norm": 42.043296813964844,
+      "learning_rate": 9.392625464943572e-07,
+      "loss": 1.0201,
+      "step": 15446
+    },
+    {
+      "epoch": 0.9385138829819551,
+      "grad_norm": 0.0957220047712326,
+      "learning_rate": 9.37416678955777e-07,
+      "loss": 1.1029,
+      "step": 15447
+    },
+    {
+      "epoch": 0.9385746400145817,
+      "grad_norm": 0.253751665353775,
+      "learning_rate": 9.355726098156714e-07,
+      "loss": 1.3224,
+      "step": 15448
+    },
+    {
+      "epoch": 0.9386353970472082,
+      "grad_norm": 0.11063911765813828,
+      "learning_rate": 9.337303391416363e-07,
+      "loss": 1.0602,
+      "step": 15449
+    },
+    {
+      "epoch": 0.9386961540798348,
+      "grad_norm": 0.1451398879289627,
+      "learning_rate": 9.318898670012011e-07,
+      "loss": 1.0168,
+      "step": 15450
+    },
+    {
+      "epoch": 0.9387569111124613,
+      "grad_norm": 0.09249144047498703,
+      "learning_rate": 9.300511934618284e-07,
+      "loss": 1.0056,
+      "step": 15451
+    },
+    {
+      "epoch": 0.9388176681450878,
+      "grad_norm": 0.17374804615974426,
+      "learning_rate": 9.282143185909087e-07,
+      "loss": 1.1407,
+      "step": 15452
+    },
+    {
+      "epoch": 0.9388784251777144,
+      "grad_norm": 0.19422796368598938,
+      "learning_rate": 9.263792424557882e-07,
+      "loss": 1.1103,
+      "step": 15453
+    },
+    {
+      "epoch": 0.9389391822103409,
+      "grad_norm": 0.09566016495227814,
+      "learning_rate": 9.245459651237132e-07,
+      "loss": 1.0108,
+      "step": 15454
+    },
+    {
+      "epoch": 0.9389999392429673,
+      "grad_norm": 0.10348668694496155,
+      "learning_rate": 9.227144866618964e-07,
+      "loss": 1.0474,
+      "step": 15455
+    },
+    {
+      "epoch": 0.9390606962755939,
+      "grad_norm": 0.16968800127506256,
+      "learning_rate": 9.208848071374676e-07,
+      "loss": 1.0663,
+      "step": 15456
+    },
+    {
+      "epoch": 0.9391214533082204,
+      "grad_norm": 0.1156466007232666,
+      "learning_rate": 9.190569266174898e-07,
+      "loss": 1.0593,
+      "step": 15457
+    },
+    {
+      "epoch": 0.9391822103408469,
+      "grad_norm": 0.1115623265504837,
+      "learning_rate": 9.172308451689704e-07,
+      "loss": 1.021,
+      "step": 15458
+    },
+    {
+      "epoch": 0.9392429673734735,
+      "grad_norm": 0.09910662472248077,
+      "learning_rate": 9.154065628588392e-07,
+      "loss": 1.0201,
+      "step": 15459
+    },
+    {
+      "epoch": 0.9393037244061,
+      "grad_norm": 0.13908688724040985,
+      "learning_rate": 9.135840797539708e-07,
+      "loss": 1.0597,
+      "step": 15460
+    },
+    {
+      "epoch": 0.9393644814387265,
+      "grad_norm": 0.10848038643598557,
+      "learning_rate": 9.117633959211668e-07,
+      "loss": 1.0248,
+      "step": 15461
+    },
+    {
+      "epoch": 0.9394252384713531,
+      "grad_norm": 0.12842628359794617,
+      "learning_rate": 9.099445114271632e-07,
+      "loss": 1.053,
+      "step": 15462
+    },
+    {
+      "epoch": 0.9394859955039796,
+      "grad_norm": 0.20164434611797333,
+      "learning_rate": 9.081274263386285e-07,
+      "loss": 1.1054,
+      "step": 15463
+    },
+    {
+      "epoch": 0.9395467525366061,
+      "grad_norm": 0.1131063848733902,
+      "learning_rate": 9.063121407221764e-07,
+      "loss": 1.0463,
+      "step": 15464
+    },
+    {
+      "epoch": 0.9396075095692327,
+      "grad_norm": 0.14073964953422546,
+      "learning_rate": 9.044986546443368e-07,
+      "loss": 1.0738,
+      "step": 15465
+    },
+    {
+      "epoch": 0.9396682666018592,
+      "grad_norm": 0.14815548062324524,
+      "learning_rate": 9.026869681715954e-07,
+      "loss": 1.0619,
+      "step": 15466
+    },
+    {
+      "epoch": 0.9397290236344857,
+      "grad_norm": 0.13954861462116241,
+      "learning_rate": 9.008770813703549e-07,
+      "loss": 1.0497,
+      "step": 15467
+    },
+    {
+      "epoch": 0.9397897806671123,
+      "grad_norm": 0.14102184772491455,
+      "learning_rate": 8.990689943069508e-07,
+      "loss": 1.1041,
+      "step": 15468
+    },
+    {
+      "epoch": 0.9398505376997387,
+      "grad_norm": 0.23908531665802002,
+      "learning_rate": 8.972627070476636e-07,
+      "loss": 1.1724,
+      "step": 15469
+    },
+    {
+      "epoch": 0.9399112947323652,
+      "grad_norm": 0.15082105994224548,
+      "learning_rate": 8.954582196587013e-07,
+      "loss": 1.0835,
+      "step": 15470
+    },
+    {
+      "epoch": 0.9399720517649918,
+      "grad_norm": 0.09962394833564758,
+      "learning_rate": 8.936555322062168e-07,
+      "loss": 1.0273,
+      "step": 15471
+    },
+    {
+      "epoch": 0.9400328087976183,
+      "grad_norm": 0.32642045617103577,
+      "learning_rate": 8.918546447562737e-07,
+      "loss": 1.1048,
+      "step": 15472
+    },
+    {
+      "epoch": 0.9400935658302448,
+      "grad_norm": 0.13679812848567963,
+      "learning_rate": 8.90055557374897e-07,
+      "loss": 1.0399,
+      "step": 15473
+    },
+    {
+      "epoch": 0.9401543228628714,
+      "grad_norm": 0.12505611777305603,
+      "learning_rate": 8.882582701280228e-07,
+      "loss": 1.036,
+      "step": 15474
+    },
+    {
+      "epoch": 0.9402150798954979,
+      "grad_norm": 0.14543262124061584,
+      "learning_rate": 8.864627830815375e-07,
+      "loss": 1.0111,
+      "step": 15475
+    },
+    {
+      "epoch": 0.9402758369281244,
+      "grad_norm": 0.10226338356733322,
+      "learning_rate": 8.846690963012494e-07,
+      "loss": 1.0437,
+      "step": 15476
+    },
+    {
+      "epoch": 0.940336593960751,
+      "grad_norm": 0.11086393892765045,
+      "learning_rate": 8.828772098529114e-07,
+      "loss": 1.0519,
+      "step": 15477
+    },
+    {
+      "epoch": 0.9403973509933775,
+      "grad_norm": 0.1930517852306366,
+      "learning_rate": 8.81087123802199e-07,
+      "loss": 1.1278,
+      "step": 15478
+    },
+    {
+      "epoch": 0.940458108026004,
+      "grad_norm": 0.1946292221546173,
+      "learning_rate": 8.792988382147372e-07,
+      "loss": 1.0244,
+      "step": 15479
+    },
+    {
+      "epoch": 0.9405188650586306,
+      "grad_norm": 0.11169382929801941,
+      "learning_rate": 8.775123531560736e-07,
+      "loss": 1.0799,
+      "step": 15480
+    },
+    {
+      "epoch": 0.9405796220912571,
+      "grad_norm": 0.12485785037279129,
+      "learning_rate": 8.757276686916837e-07,
+      "loss": 1.0749,
+      "step": 15481
+    },
+    {
+      "epoch": 0.9406403791238835,
+      "grad_norm": 0.119688481092453,
+      "learning_rate": 8.739447848869987e-07,
+      "loss": 1.0125,
+      "step": 15482
+    },
+    {
+      "epoch": 0.9407011361565101,
+      "grad_norm": 0.13595597445964813,
+      "learning_rate": 8.721637018073548e-07,
+      "loss": 1.0779,
+      "step": 15483
+    },
+    {
+      "epoch": 0.9407618931891366,
+      "grad_norm": 0.2306956648826599,
+      "learning_rate": 8.703844195180555e-07,
+      "loss": 1.0528,
+      "step": 15484
+    },
+    {
+      "epoch": 0.9408226502217631,
+      "grad_norm": 0.27294921875,
+      "learning_rate": 8.686069380843042e-07,
+      "loss": 1.1304,
+      "step": 15485
+    },
+    {
+      "epoch": 0.9408834072543897,
+      "grad_norm": 0.26540976762771606,
+      "learning_rate": 8.668312575712656e-07,
+      "loss": 1.0042,
+      "step": 15486
+    },
+    {
+      "epoch": 0.9409441642870162,
+      "grad_norm": 0.18844637274742126,
+      "learning_rate": 8.650573780440263e-07,
+      "loss": 1.0722,
+      "step": 15487
+    },
+    {
+      "epoch": 0.9410049213196428,
+      "grad_norm": 0.16316348314285278,
+      "learning_rate": 8.632852995676066e-07,
+      "loss": 1.0507,
+      "step": 15488
+    },
+    {
+      "epoch": 0.9410656783522693,
+      "grad_norm": 0.15009582042694092,
+      "learning_rate": 8.615150222069601e-07,
+      "loss": 1.1169,
+      "step": 15489
+    },
+    {
+      "epoch": 0.9411264353848958,
+      "grad_norm": 0.1508311778306961,
+      "learning_rate": 8.597465460269738e-07,
+      "loss": 1.0106,
+      "step": 15490
+    },
+    {
+      "epoch": 0.9411871924175224,
+      "grad_norm": 0.21480034291744232,
+      "learning_rate": 8.579798710924791e-07,
+      "loss": 1.1273,
+      "step": 15491
+    },
+    {
+      "epoch": 0.9412479494501489,
+      "grad_norm": 0.11603716760873795,
+      "learning_rate": 8.562149974682354e-07,
+      "loss": 1.081,
+      "step": 15492
+    },
+    {
+      "epoch": 0.9413087064827754,
+      "grad_norm": 0.1249760165810585,
+      "learning_rate": 8.544519252189354e-07,
+      "loss": 1.0765,
+      "step": 15493
+    },
+    {
+      "epoch": 0.941369463515402,
+      "grad_norm": 0.15420284867286682,
+      "learning_rate": 8.526906544091884e-07,
+      "loss": 1.1285,
+      "step": 15494
+    },
+    {
+      "epoch": 0.9414302205480284,
+      "grad_norm": 0.24884967505931854,
+      "learning_rate": 8.50931185103565e-07,
+      "loss": 1.0758,
+      "step": 15495
+    },
+    {
+      "epoch": 0.9414909775806549,
+      "grad_norm": 0.20638661086559296,
+      "learning_rate": 8.491735173665693e-07,
+      "loss": 1.1614,
+      "step": 15496
+    },
+    {
+      "epoch": 0.9415517346132815,
+      "grad_norm": 9.35626220703125,
+      "learning_rate": 8.474176512626108e-07,
+      "loss": 1.0402,
+      "step": 15497
+    },
+    {
+      "epoch": 0.941612491645908,
+      "grad_norm": 0.13855469226837158,
+      "learning_rate": 8.456635868560603e-07,
+      "loss": 1.1637,
+      "step": 15498
+    },
+    {
+      "epoch": 0.9416732486785345,
+      "grad_norm": 0.1717992126941681,
+      "learning_rate": 8.439113242112217e-07,
+      "loss": 1.116,
+      "step": 15499
+    },
+    {
+      "epoch": 0.9417340057111611,
+      "grad_norm": 0.26355358958244324,
+      "learning_rate": 8.421608633923051e-07,
+      "loss": 1.0636,
+      "step": 15500
+    },
+    {
+      "epoch": 0.9417947627437876,
+      "grad_norm": 0.12528984248638153,
+      "learning_rate": 8.404122044634865e-07,
+      "loss": 0.9809,
+      "step": 15501
+    },
+    {
+      "epoch": 0.9418555197764141,
+      "grad_norm": 0.184946209192276,
+      "learning_rate": 8.386653474888595e-07,
+      "loss": 1.1069,
+      "step": 15502
+    },
+    {
+      "epoch": 0.9419162768090407,
+      "grad_norm": 0.12299303710460663,
+      "learning_rate": 8.369202925324559e-07,
+      "loss": 1.0393,
+      "step": 15503
+    },
+    {
+      "epoch": 0.9419770338416672,
+      "grad_norm": 0.14128755033016205,
+      "learning_rate": 8.35177039658247e-07,
+      "loss": 1.001,
+      "step": 15504
+    },
+    {
+      "epoch": 0.9420377908742937,
+      "grad_norm": 0.35349786281585693,
+      "learning_rate": 8.33435588930126e-07,
+      "loss": 1.0806,
+      "step": 15505
+    },
+    {
+      "epoch": 0.9420985479069203,
+      "grad_norm": 0.2799130380153656,
+      "learning_rate": 8.316959404119252e-07,
+      "loss": 1.2533,
+      "step": 15506
+    },
+    {
+      "epoch": 0.9421593049395468,
+      "grad_norm": 0.254946231842041,
+      "learning_rate": 8.299580941674157e-07,
+      "loss": 1.0968,
+      "step": 15507
+    },
+    {
+      "epoch": 0.9422200619721732,
+      "grad_norm": 0.10920147597789764,
+      "learning_rate": 8.282220502602911e-07,
+      "loss": 1.051,
+      "step": 15508
+    },
+    {
+      "epoch": 0.9422808190047998,
+      "grad_norm": 1.0087721347808838,
+      "learning_rate": 8.264878087542005e-07,
+      "loss": 1.0609,
+      "step": 15509
+    },
+    {
+      "epoch": 0.9423415760374263,
+      "grad_norm": 0.15644152462482452,
+      "learning_rate": 8.247553697126986e-07,
+      "loss": 1.1413,
+      "step": 15510
+    },
+    {
+      "epoch": 0.9424023330700528,
+      "grad_norm": 0.08863753825426102,
+      "learning_rate": 8.230247331992902e-07,
+      "loss": 1.0224,
+      "step": 15511
+    },
+    {
+      "epoch": 0.9424630901026794,
+      "grad_norm": 0.17508651316165924,
+      "learning_rate": 8.212958992774189e-07,
+      "loss": 1.0603,
+      "step": 15512
+    },
+    {
+      "epoch": 0.9425238471353059,
+      "grad_norm": 0.17421424388885498,
+      "learning_rate": 8.195688680104618e-07,
+      "loss": 1.1268,
+      "step": 15513
+    },
+    {
+      "epoch": 0.9425846041679324,
+      "grad_norm": 0.3248521685600281,
+      "learning_rate": 8.178436394617073e-07,
+      "loss": 1.1632,
+      "step": 15514
+    },
+    {
+      "epoch": 0.942645361200559,
+      "grad_norm": 0.14074921607971191,
+      "learning_rate": 8.16120213694399e-07,
+      "loss": 1.0927,
+      "step": 15515
+    },
+    {
+      "epoch": 0.9427061182331855,
+      "grad_norm": 0.12300248444080353,
+      "learning_rate": 8.143985907717089e-07,
+      "loss": 1.0098,
+      "step": 15516
+    },
+    {
+      "epoch": 0.942766875265812,
+      "grad_norm": 0.6366365551948547,
+      "learning_rate": 8.126787707567473e-07,
+      "loss": 1.0995,
+      "step": 15517
+    },
+    {
+      "epoch": 0.9428276322984386,
+      "grad_norm": 0.17644186317920685,
+      "learning_rate": 8.109607537125585e-07,
+      "loss": 1.1225,
+      "step": 15518
+    },
+    {
+      "epoch": 0.9428883893310651,
+      "grad_norm": 0.14260059595108032,
+      "learning_rate": 8.092445397021142e-07,
+      "loss": 1.0811,
+      "step": 15519
+    },
+    {
+      "epoch": 0.9429491463636916,
+      "grad_norm": 0.14571037888526917,
+      "learning_rate": 8.075301287883141e-07,
+      "loss": 1.0701,
+      "step": 15520
+    },
+    {
+      "epoch": 0.9430099033963181,
+      "grad_norm": 0.14896683394908905,
+      "learning_rate": 8.058175210339968e-07,
+      "loss": 1.0904,
+      "step": 15521
+    },
+    {
+      "epoch": 0.9430706604289446,
+      "grad_norm": 0.09953810274600983,
+      "learning_rate": 8.041067165019567e-07,
+      "loss": 0.9973,
+      "step": 15522
+    },
+    {
+      "epoch": 0.9431314174615711,
+      "grad_norm": 0.10686226934194565,
+      "learning_rate": 8.023977152548934e-07,
+      "loss": 1.0325,
+      "step": 15523
+    },
+    {
+      "epoch": 0.9431921744941977,
+      "grad_norm": 0.1334177553653717,
+      "learning_rate": 8.006905173554513e-07,
+      "loss": 0.9956,
+      "step": 15524
+    },
+    {
+      "epoch": 0.9432529315268242,
+      "grad_norm": 0.15446703135967255,
+      "learning_rate": 7.989851228662082e-07,
+      "loss": 1.0661,
+      "step": 15525
+    },
+    {
+      "epoch": 0.9433136885594507,
+      "grad_norm": 0.13371652364730835,
+      "learning_rate": 7.972815318496807e-07,
+      "loss": 1.0973,
+      "step": 15526
+    },
+    {
+      "epoch": 0.9433744455920773,
+      "grad_norm": 0.12622101604938507,
+      "learning_rate": 7.955797443683022e-07,
+      "loss": 1.0525,
+      "step": 15527
+    },
+    {
+      "epoch": 0.9434352026247038,
+      "grad_norm": 0.08999328315258026,
+      "learning_rate": 7.938797604844616e-07,
+      "loss": 1.0099,
+      "step": 15528
+    },
+    {
+      "epoch": 0.9434959596573304,
+      "grad_norm": 0.12138425558805466,
+      "learning_rate": 7.921815802604704e-07,
+      "loss": 1.0302,
+      "step": 15529
+    },
+    {
+      "epoch": 0.9435567166899569,
+      "grad_norm": 0.11151651293039322,
+      "learning_rate": 7.904852037585731e-07,
+      "loss": 1.065,
+      "step": 15530
+    },
+    {
+      "epoch": 0.9436174737225834,
+      "grad_norm": 0.12263260781764984,
+      "learning_rate": 7.887906310409588e-07,
+      "loss": 1.0348,
+      "step": 15531
+    },
+    {
+      "epoch": 0.94367823075521,
+      "grad_norm": 0.11933889985084534,
+      "learning_rate": 7.870978621697222e-07,
+      "loss": 1.0526,
+      "step": 15532
+    },
+    {
+      "epoch": 0.9437389877878365,
+      "grad_norm": 0.1343223750591278,
+      "learning_rate": 7.854068972069418e-07,
+      "loss": 1.2137,
+      "step": 15533
+    },
+    {
+      "epoch": 0.9437997448204629,
+      "grad_norm": 0.25458455085754395,
+      "learning_rate": 7.837177362145786e-07,
+      "loss": 1.0393,
+      "step": 15534
+    },
+    {
+      "epoch": 0.9438605018530895,
+      "grad_norm": 0.11736898869276047,
+      "learning_rate": 7.820303792545558e-07,
+      "loss": 1.0824,
+      "step": 15535
+    },
+    {
+      "epoch": 0.943921258885716,
+      "grad_norm": 0.09872827678918839,
+      "learning_rate": 7.803448263887237e-07,
+      "loss": 1.0255,
+      "step": 15536
+    },
+    {
+      "epoch": 0.9439820159183425,
+      "grad_norm": 0.15799172222614288,
+      "learning_rate": 7.78661077678855e-07,
+      "loss": 1.1307,
+      "step": 15537
+    },
+    {
+      "epoch": 0.9440427729509691,
+      "grad_norm": 0.24274155497550964,
+      "learning_rate": 7.769791331866894e-07,
+      "loss": 1.1223,
+      "step": 15538
+    },
+    {
+      "epoch": 0.9441035299835956,
+      "grad_norm": 0.13727737963199615,
+      "learning_rate": 7.752989929738663e-07,
+      "loss": 1.0714,
+      "step": 15539
+    },
+    {
+      "epoch": 0.9441642870162221,
+      "grad_norm": 0.1504889577627182,
+      "learning_rate": 7.73620657101981e-07,
+      "loss": 1.0696,
+      "step": 15540
+    },
+    {
+      "epoch": 0.9442250440488487,
+      "grad_norm": 0.2271743267774582,
+      "learning_rate": 7.719441256325399e-07,
+      "loss": 1.1468,
+      "step": 15541
+    },
+    {
+      "epoch": 0.9442858010814752,
+      "grad_norm": 0.16605541110038757,
+      "learning_rate": 7.702693986269938e-07,
+      "loss": 1.1537,
+      "step": 15542
+    },
+    {
+      "epoch": 0.9443465581141017,
+      "grad_norm": 0.18032360076904297,
+      "learning_rate": 7.685964761467491e-07,
+      "loss": 1.0823,
+      "step": 15543
+    },
+    {
+      "epoch": 0.9444073151467283,
+      "grad_norm": 0.1634989231824875,
+      "learning_rate": 7.669253582531122e-07,
+      "loss": 1.0584,
+      "step": 15544
+    },
+    {
+      "epoch": 0.9444680721793548,
+      "grad_norm": 0.2044805884361267,
+      "learning_rate": 7.652560450073454e-07,
+      "loss": 1.1238,
+      "step": 15545
+    },
+    {
+      "epoch": 0.9445288292119813,
+      "grad_norm": 0.11974194645881653,
+      "learning_rate": 7.635885364706385e-07,
+      "loss": 1.0715,
+      "step": 15546
+    },
+    {
+      "epoch": 0.9445895862446078,
+      "grad_norm": 0.17538370192050934,
+      "learning_rate": 7.61922832704104e-07,
+      "loss": 1.0661,
+      "step": 15547
+    },
+    {
+      "epoch": 0.9446503432772343,
+      "grad_norm": 0.10671769827604294,
+      "learning_rate": 7.602589337688093e-07,
+      "loss": 1.0213,
+      "step": 15548
+    },
+    {
+      "epoch": 0.9447111003098608,
+      "grad_norm": 0.1608729511499405,
+      "learning_rate": 7.585968397257392e-07,
+      "loss": 1.1058,
+      "step": 15549
+    },
+    {
+      "epoch": 0.9447718573424874,
+      "grad_norm": 0.39861640334129333,
+      "learning_rate": 7.569365506358228e-07,
+      "loss": 1.0483,
+      "step": 15550
+    },
+    {
+      "epoch": 0.9448326143751139,
+      "grad_norm": 0.18744775652885437,
+      "learning_rate": 7.552780665599113e-07,
+      "loss": 0.9952,
+      "step": 15551
+    },
+    {
+      "epoch": 0.9448933714077404,
+      "grad_norm": 0.19755451381206512,
+      "learning_rate": 7.536213875588061e-07,
+      "loss": 1.0674,
+      "step": 15552
+    },
+    {
+      "epoch": 0.944954128440367,
+      "grad_norm": 0.140263631939888,
+      "learning_rate": 7.519665136932252e-07,
+      "loss": 1.0613,
+      "step": 15553
+    },
+    {
+      "epoch": 0.9450148854729935,
+      "grad_norm": 0.28901341557502747,
+      "learning_rate": 7.503134450238314e-07,
+      "loss": 1.3098,
+      "step": 15554
+    },
+    {
+      "epoch": 0.94507564250562,
+      "grad_norm": 0.13824376463890076,
+      "learning_rate": 7.486621816112149e-07,
+      "loss": 1.0958,
+      "step": 15555
+    },
+    {
+      "epoch": 0.9451363995382466,
+      "grad_norm": 0.16486625373363495,
+      "learning_rate": 7.470127235159108e-07,
+      "loss": 1.1009,
+      "step": 15556
+    },
+    {
+      "epoch": 0.9451971565708731,
+      "grad_norm": 0.1943669468164444,
+      "learning_rate": 7.453650707983761e-07,
+      "loss": 1.0475,
+      "step": 15557
+    },
+    {
+      "epoch": 0.9452579136034996,
+      "grad_norm": 0.17605499923229218,
+      "learning_rate": 7.437192235190016e-07,
+      "loss": 1.0456,
+      "step": 15558
+    },
+    {
+      "epoch": 0.9453186706361262,
+      "grad_norm": 0.19593696296215057,
+      "learning_rate": 7.420751817381222e-07,
+      "loss": 1.1297,
+      "step": 15559
+    },
+    {
+      "epoch": 0.9453794276687526,
+      "grad_norm": 0.19700486958026886,
+      "learning_rate": 7.404329455160008e-07,
+      "loss": 1.0603,
+      "step": 15560
+    },
+    {
+      "epoch": 0.9454401847013791,
+      "grad_norm": 0.14606669545173645,
+      "learning_rate": 7.387925149128283e-07,
+      "loss": 1.0968,
+      "step": 15561
+    },
+    {
+      "epoch": 0.9455009417340057,
+      "grad_norm": 0.17591744661331177,
+      "learning_rate": 7.371538899887398e-07,
+      "loss": 1.1222,
+      "step": 15562
+    },
+    {
+      "epoch": 0.9455616987666322,
+      "grad_norm": 0.20410175621509552,
+      "learning_rate": 7.355170708037928e-07,
+      "loss": 1.1199,
+      "step": 15563
+    },
+    {
+      "epoch": 0.9456224557992587,
+      "grad_norm": 0.18681199848651886,
+      "learning_rate": 7.33882057417995e-07,
+      "loss": 1.2142,
+      "step": 15564
+    },
+    {
+      "epoch": 0.9456832128318853,
+      "grad_norm": 0.1307850480079651,
+      "learning_rate": 7.322488498912705e-07,
+      "loss": 1.0561,
+      "step": 15565
+    },
+    {
+      "epoch": 0.9457439698645118,
+      "grad_norm": 0.1756134331226349,
+      "learning_rate": 7.306174482834938e-07,
+      "loss": 1.1029,
+      "step": 15566
+    },
+    {
+      "epoch": 0.9458047268971383,
+      "grad_norm": 0.21926729381084442,
+      "learning_rate": 7.289878526544503e-07,
+      "loss": 1.215,
+      "step": 15567
+    },
+    {
+      "epoch": 0.9458654839297649,
+      "grad_norm": 0.18988537788391113,
+      "learning_rate": 7.273600630638866e-07,
+      "loss": 1.3811,
+      "step": 15568
+    },
+    {
+      "epoch": 0.9459262409623914,
+      "grad_norm": 0.18372716009616852,
+      "learning_rate": 7.257340795714607e-07,
+      "loss": 1.0801,
+      "step": 15569
+    },
+    {
+      "epoch": 0.945986997995018,
+      "grad_norm": 0.13982757925987244,
+      "learning_rate": 7.241099022367803e-07,
+      "loss": 1.1232,
+      "step": 15570
+    },
+    {
+      "epoch": 0.9460477550276445,
+      "grad_norm": 0.19893155992031097,
+      "learning_rate": 7.224875311193757e-07,
+      "loss": 1.0736,
+      "step": 15571
+    },
+    {
+      "epoch": 0.946108512060271,
+      "grad_norm": 0.12165235728025436,
+      "learning_rate": 7.208669662787104e-07,
+      "loss": 1.0941,
+      "step": 15572
+    },
+    {
+      "epoch": 0.9461692690928976,
+      "grad_norm": 0.14123670756816864,
+      "learning_rate": 7.19248207774198e-07,
+      "loss": 1.046,
+      "step": 15573
+    },
+    {
+      "epoch": 0.946230026125524,
+      "grad_norm": 0.1697043925523758,
+      "learning_rate": 7.176312556651632e-07,
+      "loss": 1.0971,
+      "step": 15574
+    },
+    {
+      "epoch": 0.9462907831581505,
+      "grad_norm": 0.12591105699539185,
+      "learning_rate": 7.160161100108864e-07,
+      "loss": 1.1016,
+      "step": 15575
+    },
+    {
+      "epoch": 0.946351540190777,
+      "grad_norm": 0.14257919788360596,
+      "learning_rate": 7.144027708705592e-07,
+      "loss": 1.0616,
+      "step": 15576
+    },
+    {
+      "epoch": 0.9464122972234036,
+      "grad_norm": 0.10719078779220581,
+      "learning_rate": 7.127912383033286e-07,
+      "loss": 1.0762,
+      "step": 15577
+    },
+    {
+      "epoch": 0.9464730542560301,
+      "grad_norm": 0.1505705863237381,
+      "learning_rate": 7.111815123682641e-07,
+      "loss": 1.0204,
+      "step": 15578
+    },
+    {
+      "epoch": 0.9465338112886567,
+      "grad_norm": 0.10594255477190018,
+      "learning_rate": 7.09573593124363e-07,
+      "loss": 1.0437,
+      "step": 15579
+    },
+    {
+      "epoch": 0.9465945683212832,
+      "grad_norm": 0.16246521472930908,
+      "learning_rate": 7.07967480630578e-07,
+      "loss": 1.0342,
+      "step": 15580
+    },
+    {
+      "epoch": 0.9466553253539097,
+      "grad_norm": 0.11460207402706146,
+      "learning_rate": 7.063631749457677e-07,
+      "loss": 1.0481,
+      "step": 15581
+    },
+    {
+      "epoch": 0.9467160823865363,
+      "grad_norm": 0.1805431842803955,
+      "learning_rate": 7.047606761287406e-07,
+      "loss": 1.1141,
+      "step": 15582
+    },
+    {
+      "epoch": 0.9467768394191628,
+      "grad_norm": 0.26799359917640686,
+      "learning_rate": 7.031599842382386e-07,
+      "loss": 1.0467,
+      "step": 15583
+    },
+    {
+      "epoch": 0.9468375964517893,
+      "grad_norm": 0.32107555866241455,
+      "learning_rate": 7.015610993329424e-07,
+      "loss": 1.122,
+      "step": 15584
+    },
+    {
+      "epoch": 0.9468983534844159,
+      "grad_norm": 0.11507256329059601,
+      "learning_rate": 6.999640214714498e-07,
+      "loss": 1.0745,
+      "step": 15585
+    },
+    {
+      "epoch": 0.9469591105170424,
+      "grad_norm": 0.38523489236831665,
+      "learning_rate": 6.983687507123138e-07,
+      "loss": 1.2181,
+      "step": 15586
+    },
+    {
+      "epoch": 0.9470198675496688,
+      "grad_norm": 0.14577315747737885,
+      "learning_rate": 6.967752871139932e-07,
+      "loss": 1.0399,
+      "step": 15587
+    },
+    {
+      "epoch": 0.9470806245822954,
+      "grad_norm": 0.12074267119169235,
+      "learning_rate": 6.951836307349025e-07,
+      "loss": 1.0143,
+      "step": 15588
+    },
+    {
+      "epoch": 0.9471413816149219,
+      "grad_norm": 0.11242710798978806,
+      "learning_rate": 6.935937816333837e-07,
+      "loss": 1.0237,
+      "step": 15589
+    },
+    {
+      "epoch": 0.9472021386475484,
+      "grad_norm": 4.724917411804199,
+      "learning_rate": 6.920057398677238e-07,
+      "loss": 1.1851,
+      "step": 15590
+    },
+    {
+      "epoch": 0.947262895680175,
+      "grad_norm": 0.20853085815906525,
+      "learning_rate": 6.904195054961204e-07,
+      "loss": 1.1195,
+      "step": 15591
+    },
+    {
+      "epoch": 0.9473236527128015,
+      "grad_norm": 0.23358775675296783,
+      "learning_rate": 6.888350785767217e-07,
+      "loss": 1.1955,
+      "step": 15592
+    },
+    {
+      "epoch": 0.947384409745428,
+      "grad_norm": 0.13040676712989807,
+      "learning_rate": 6.872524591676089e-07,
+      "loss": 1.0172,
+      "step": 15593
+    },
+    {
+      "epoch": 0.9474451667780546,
+      "grad_norm": 0.15266183018684387,
+      "learning_rate": 6.856716473267854e-07,
+      "loss": 1.0712,
+      "step": 15594
+    },
+    {
+      "epoch": 0.9475059238106811,
+      "grad_norm": 0.3191109597682953,
+      "learning_rate": 6.840926431121997e-07,
+      "loss": 1.1593,
+      "step": 15595
+    },
+    {
+      "epoch": 0.9475666808433076,
+      "grad_norm": 0.1220596432685852,
+      "learning_rate": 6.82515446581733e-07,
+      "loss": 1.0495,
+      "step": 15596
+    },
+    {
+      "epoch": 0.9476274378759342,
+      "grad_norm": 0.7531089186668396,
+      "learning_rate": 6.809400577931946e-07,
+      "loss": 1.1295,
+      "step": 15597
+    },
+    {
+      "epoch": 0.9476881949085607,
+      "grad_norm": 0.44466760754585266,
+      "learning_rate": 6.793664768043273e-07,
+      "loss": 1.0922,
+      "step": 15598
+    },
+    {
+      "epoch": 0.9477489519411872,
+      "grad_norm": 0.5898516178131104,
+      "learning_rate": 6.777947036728182e-07,
+      "loss": 1.0571,
+      "step": 15599
+    },
+    {
+      "epoch": 0.9478097089738137,
+      "grad_norm": 0.11698181927204132,
+      "learning_rate": 6.762247384562825e-07,
+      "loss": 1.0672,
+      "step": 15600
+    },
+    {
+      "epoch": 0.9478704660064402,
+      "grad_norm": 0.9985892176628113,
+      "learning_rate": 6.746565812122574e-07,
+      "loss": 1.1468,
+      "step": 15601
+    },
+    {
+      "epoch": 0.9479312230390667,
+      "grad_norm": 1.2169983386993408,
+      "learning_rate": 6.730902319982302e-07,
+      "loss": 1.1957,
+      "step": 15602
+    },
+    {
+      "epoch": 0.9479919800716933,
+      "grad_norm": 0.6467977166175842,
+      "learning_rate": 6.715256908716161e-07,
+      "loss": 1.0532,
+      "step": 15603
+    },
+    {
+      "epoch": 0.9480527371043198,
+      "grad_norm": 1.040322184562683,
+      "learning_rate": 6.699629578897637e-07,
+      "loss": 0.9915,
+      "step": 15604
+    },
+    {
+      "epoch": 0.9481134941369463,
+      "grad_norm": 0.3417030870914459,
+      "learning_rate": 6.684020331099494e-07,
+      "loss": 1.205,
+      "step": 15605
+    },
+    {
+      "epoch": 0.9481742511695729,
+      "grad_norm": 0.1402425467967987,
+      "learning_rate": 6.668429165893997e-07,
+      "loss": 1.0995,
+      "step": 15606
+    },
+    {
+      "epoch": 0.9482350082021994,
+      "grad_norm": 0.17089398205280304,
+      "learning_rate": 6.652856083852632e-07,
+      "loss": 1.0777,
+      "step": 15607
+    },
+    {
+      "epoch": 0.948295765234826,
+      "grad_norm": 0.10627352446317673,
+      "learning_rate": 6.637301085546111e-07,
+      "loss": 1.0384,
+      "step": 15608
+    },
+    {
+      "epoch": 0.9483565222674525,
+      "grad_norm": 0.34206822514533997,
+      "learning_rate": 6.621764171544698e-07,
+      "loss": 1.0198,
+      "step": 15609
+    },
+    {
+      "epoch": 0.948417279300079,
+      "grad_norm": 0.10728035867214203,
+      "learning_rate": 6.606245342417882e-07,
+      "loss": 1.0321,
+      "step": 15610
+    },
+    {
+      "epoch": 0.9484780363327056,
+      "grad_norm": 0.10981914401054382,
+      "learning_rate": 6.590744598734544e-07,
+      "loss": 1.0015,
+      "step": 15611
+    },
+    {
+      "epoch": 0.9485387933653321,
+      "grad_norm": 0.7554893493652344,
+      "learning_rate": 6.575261941062838e-07,
+      "loss": 1.1793,
+      "step": 15612
+    },
+    {
+      "epoch": 0.9485995503979585,
+      "grad_norm": 0.1202106773853302,
+      "learning_rate": 6.559797369970256e-07,
+      "loss": 0.971,
+      "step": 15613
+    },
+    {
+      "epoch": 0.948660307430585,
+      "grad_norm": 0.1026490181684494,
+      "learning_rate": 6.544350886023675e-07,
+      "loss": 1.0654,
+      "step": 15614
+    },
+    {
+      "epoch": 0.9487210644632116,
+      "grad_norm": 0.19423659145832062,
+      "learning_rate": 6.528922489789313e-07,
+      "loss": 1.1146,
+      "step": 15615
+    },
+    {
+      "epoch": 0.9487818214958381,
+      "grad_norm": 0.16467399895191193,
+      "learning_rate": 6.513512181832659e-07,
+      "loss": 1.1322,
+      "step": 15616
+    },
+    {
+      "epoch": 0.9488425785284647,
+      "grad_norm": 0.13733038306236267,
+      "learning_rate": 6.498119962718652e-07,
+      "loss": 1.054,
+      "step": 15617
+    },
+    {
+      "epoch": 0.9489033355610912,
+      "grad_norm": 25.943130493164062,
+      "learning_rate": 6.482745833011394e-07,
+      "loss": 1.0367,
+      "step": 15618
+    },
+    {
+      "epoch": 0.9489640925937177,
+      "grad_norm": 0.12427359819412231,
+      "learning_rate": 6.467389793274547e-07,
+      "loss": 1.0664,
+      "step": 15619
+    },
+    {
+      "epoch": 0.9490248496263443,
+      "grad_norm": 0.1817326694726944,
+      "learning_rate": 6.452051844070883e-07,
+      "loss": 1.0023,
+      "step": 15620
+    },
+    {
+      "epoch": 0.9490856066589708,
+      "grad_norm": 0.14965835213661194,
+      "learning_rate": 6.436731985962674e-07,
+      "loss": 1.0867,
+      "step": 15621
+    },
+    {
+      "epoch": 0.9491463636915973,
+      "grad_norm": 0.21469591557979584,
+      "learning_rate": 6.421430219511415e-07,
+      "loss": 1.0936,
+      "step": 15622
+    },
+    {
+      "epoch": 0.9492071207242239,
+      "grad_norm": 0.20590290427207947,
+      "learning_rate": 6.406146545278102e-07,
+      "loss": 1.0878,
+      "step": 15623
+    },
+    {
+      "epoch": 0.9492678777568504,
+      "grad_norm": 0.21590803563594818,
+      "learning_rate": 6.390880963822842e-07,
+      "loss": 1.1184,
+      "step": 15624
+    },
+    {
+      "epoch": 0.9493286347894769,
+      "grad_norm": 0.09580983221530914,
+      "learning_rate": 6.375633475705245e-07,
+      "loss": 1.0025,
+      "step": 15625
+    },
+    {
+      "epoch": 0.9493893918221034,
+      "grad_norm": 0.19283616542816162,
+      "learning_rate": 6.36040408148425e-07,
+      "loss": 1.1458,
+      "step": 15626
+    },
+    {
+      "epoch": 0.9494501488547299,
+      "grad_norm": 0.14228251576423645,
+      "learning_rate": 6.345192781718134e-07,
+      "loss": 1.0288,
+      "step": 15627
+    },
+    {
+      "epoch": 0.9495109058873564,
+      "grad_norm": 0.10137403756380081,
+      "learning_rate": 6.329999576964341e-07,
+      "loss": 1.0268,
+      "step": 15628
+    },
+    {
+      "epoch": 0.949571662919983,
+      "grad_norm": 0.11061901599168777,
+      "learning_rate": 6.314824467779812e-07,
+      "loss": 1.0246,
+      "step": 15629
+    },
+    {
+      "epoch": 0.9496324199526095,
+      "grad_norm": 0.1095118448138237,
+      "learning_rate": 6.29966745472077e-07,
+      "loss": 1.0235,
+      "step": 15630
+    },
+    {
+      "epoch": 0.949693176985236,
+      "grad_norm": 0.11759944260120392,
+      "learning_rate": 6.284528538342937e-07,
+      "loss": 1.0645,
+      "step": 15631
+    },
+    {
+      "epoch": 0.9497539340178626,
+      "grad_norm": 0.4917198419570923,
+      "learning_rate": 6.269407719201092e-07,
+      "loss": 1.0552,
+      "step": 15632
+    },
+    {
+      "epoch": 0.9498146910504891,
+      "grad_norm": 0.11569838225841522,
+      "learning_rate": 6.254304997849624e-07,
+      "loss": 1.0497,
+      "step": 15633
+    },
+    {
+      "epoch": 0.9498754480831156,
+      "grad_norm": 0.17323362827301025,
+      "learning_rate": 6.239220374841981e-07,
+      "loss": 1.1177,
+      "step": 15634
+    },
+    {
+      "epoch": 0.9499362051157422,
+      "grad_norm": 0.10534697026014328,
+      "learning_rate": 6.224153850731107e-07,
+      "loss": 1.0024,
+      "step": 15635
+    },
+    {
+      "epoch": 0.9499969621483687,
+      "grad_norm": 0.3628701865673065,
+      "learning_rate": 6.209105426069395e-07,
+      "loss": 1.2407,
+      "step": 15636
+    },
+    {
+      "epoch": 0.9500577191809952,
+      "grad_norm": 0.12392721325159073,
+      "learning_rate": 6.194075101408348e-07,
+      "loss": 1.0336,
+      "step": 15637
+    },
+    {
+      "epoch": 0.9501184762136218,
+      "grad_norm": 0.12260594218969345,
+      "learning_rate": 6.179062877298969e-07,
+      "loss": 1.0222,
+      "step": 15638
+    },
+    {
+      "epoch": 0.9501792332462482,
+      "grad_norm": 0.17222723364830017,
+      "learning_rate": 6.16406875429143e-07,
+      "loss": 1.0944,
+      "step": 15639
+    },
+    {
+      "epoch": 0.9502399902788747,
+      "grad_norm": 0.13815884292125702,
+      "learning_rate": 6.149092732935457e-07,
+      "loss": 1.0336,
+      "step": 15640
+    },
+    {
+      "epoch": 0.9503007473115013,
+      "grad_norm": 0.22213789820671082,
+      "learning_rate": 6.13413481377989e-07,
+      "loss": 1.0869,
+      "step": 15641
+    },
+    {
+      "epoch": 0.9503615043441278,
+      "grad_norm": 0.14781610667705536,
+      "learning_rate": 6.119194997373123e-07,
+      "loss": 1.0634,
+      "step": 15642
+    },
+    {
+      "epoch": 0.9504222613767543,
+      "grad_norm": 0.11637907475233078,
+      "learning_rate": 6.104273284262718e-07,
+      "loss": 1.0912,
+      "step": 15643
+    },
+    {
+      "epoch": 0.9504830184093809,
+      "grad_norm": 0.11060317605733871,
+      "learning_rate": 6.089369674995626e-07,
+      "loss": 1.0476,
+      "step": 15644
+    },
+    {
+      "epoch": 0.9505437754420074,
+      "grad_norm": 0.14210262894630432,
+      "learning_rate": 6.074484170118188e-07,
+      "loss": 1.0488,
+      "step": 15645
+    },
+    {
+      "epoch": 0.950604532474634,
+      "grad_norm": 0.1596514880657196,
+      "learning_rate": 6.059616770175969e-07,
+      "loss": 1.0141,
+      "step": 15646
+    },
+    {
+      "epoch": 0.9506652895072605,
+      "grad_norm": 0.13412600755691528,
+      "learning_rate": 6.044767475714031e-07,
+      "loss": 1.1651,
+      "step": 15647
+    },
+    {
+      "epoch": 0.950726046539887,
+      "grad_norm": 0.14134304225444794,
+      "learning_rate": 6.029936287276605e-07,
+      "loss": 0.9804,
+      "step": 15648
+    },
+    {
+      "epoch": 0.9507868035725136,
+      "grad_norm": 0.1634296029806137,
+      "learning_rate": 6.015123205407313e-07,
+      "loss": 1.0712,
+      "step": 15649
+    },
+    {
+      "epoch": 0.9508475606051401,
+      "grad_norm": 0.10759787261486053,
+      "learning_rate": 6.000328230649166e-07,
+      "loss": 1.0393,
+      "step": 15650
+    },
+    {
+      "epoch": 0.9509083176377666,
+      "grad_norm": 0.09372863173484802,
+      "learning_rate": 5.98555136354445e-07,
+      "loss": 1.0099,
+      "step": 15651
+    },
+    {
+      "epoch": 0.950969074670393,
+      "grad_norm": 0.13221114873886108,
+      "learning_rate": 5.970792604634901e-07,
+      "loss": 1.0567,
+      "step": 15652
+    },
+    {
+      "epoch": 0.9510298317030196,
+      "grad_norm": 0.10997379571199417,
+      "learning_rate": 5.956051954461472e-07,
+      "loss": 1.057,
+      "step": 15653
+    },
+    {
+      "epoch": 0.9510905887356461,
+      "grad_norm": 0.15794794261455536,
+      "learning_rate": 5.941329413564401e-07,
+      "loss": 1.1131,
+      "step": 15654
+    },
+    {
+      "epoch": 0.9511513457682726,
+      "grad_norm": 0.11565219610929489,
+      "learning_rate": 5.926624982483364e-07,
+      "loss": 1.04,
+      "step": 15655
+    },
+    {
+      "epoch": 0.9512121028008992,
+      "grad_norm": 0.1022125706076622,
+      "learning_rate": 5.911938661757433e-07,
+      "loss": 0.9985,
+      "step": 15656
+    },
+    {
+      "epoch": 0.9512728598335257,
+      "grad_norm": 0.11516375094652176,
+      "learning_rate": 5.897270451924897e-07,
+      "loss": 1.0435,
+      "step": 15657
+    },
+    {
+      "epoch": 0.9513336168661523,
+      "grad_norm": 0.10048330575227737,
+      "learning_rate": 5.882620353523383e-07,
+      "loss": 0.977,
+      "step": 15658
+    },
+    {
+      "epoch": 0.9513943738987788,
+      "grad_norm": 0.143539696931839,
+      "learning_rate": 5.867988367089961e-07,
+      "loss": 1.0442,
+      "step": 15659
+    },
+    {
+      "epoch": 0.9514551309314053,
+      "grad_norm": 0.12086950242519379,
+      "learning_rate": 5.853374493160979e-07,
+      "loss": 1.0932,
+      "step": 15660
+    },
+    {
+      "epoch": 0.9515158879640319,
+      "grad_norm": 0.2632455825805664,
+      "learning_rate": 5.83877873227201e-07,
+      "loss": 1.0284,
+      "step": 15661
+    },
+    {
+      "epoch": 0.9515766449966584,
+      "grad_norm": 0.3406867980957031,
+      "learning_rate": 5.824201084958181e-07,
+      "loss": 1.0898,
+      "step": 15662
+    },
+    {
+      "epoch": 0.9516374020292849,
+      "grad_norm": 0.11601199954748154,
+      "learning_rate": 5.809641551753731e-07,
+      "loss": 1.0777,
+      "step": 15663
+    },
+    {
+      "epoch": 0.9516981590619115,
+      "grad_norm": 0.11150176078081131,
+      "learning_rate": 5.7951001331924e-07,
+      "loss": 1.0282,
+      "step": 15664
+    },
+    {
+      "epoch": 0.9517589160945379,
+      "grad_norm": 0.15512247383594513,
+      "learning_rate": 5.78057682980726e-07,
+      "loss": 1.1055,
+      "step": 15665
+    },
+    {
+      "epoch": 0.9518196731271644,
+      "grad_norm": 0.11280364543199539,
+      "learning_rate": 5.766071642130555e-07,
+      "loss": 0.9957,
+      "step": 15666
+    },
+    {
+      "epoch": 0.951880430159791,
+      "grad_norm": 0.11917781084775925,
+      "learning_rate": 5.751584570694024e-07,
+      "loss": 1.0478,
+      "step": 15667
+    },
+    {
+      "epoch": 0.9519411871924175,
+      "grad_norm": 0.10429950058460236,
+      "learning_rate": 5.737115616028687e-07,
+      "loss": 1.0165,
+      "step": 15668
+    },
+    {
+      "epoch": 0.952001944225044,
+      "grad_norm": 0.14690320193767548,
+      "learning_rate": 5.722664778664955e-07,
+      "loss": 1.0667,
+      "step": 15669
+    },
+    {
+      "epoch": 0.9520627012576706,
+      "grad_norm": 0.11742889136075974,
+      "learning_rate": 5.708232059132456e-07,
+      "loss": 1.0384,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9521234582902971,
+      "grad_norm": 0.1140822321176529,
+      "learning_rate": 5.693817457960271e-07,
+      "loss": 1.0401,
+      "step": 15671
+    },
+    {
+      "epoch": 0.9521842153229236,
+      "grad_norm": 0.09306786954402924,
+      "learning_rate": 5.679420975676753e-07,
+      "loss": 1.031,
+      "step": 15672
+    },
+    {
+      "epoch": 0.9522449723555502,
+      "grad_norm": 0.1260322630405426,
+      "learning_rate": 5.66504261280959e-07,
+      "loss": 1.0305,
+      "step": 15673
+    },
+    {
+      "epoch": 0.9523057293881767,
+      "grad_norm": 0.273558109998703,
+      "learning_rate": 5.650682369885918e-07,
+      "loss": 1.1694,
+      "step": 15674
+    },
+    {
+      "epoch": 0.9523664864208032,
+      "grad_norm": 1.9666259288787842,
+      "learning_rate": 5.636340247431981e-07,
+      "loss": 1.0251,
+      "step": 15675
+    },
+    {
+      "epoch": 0.9524272434534298,
+      "grad_norm": 0.17853352427482605,
+      "learning_rate": 5.622016245973527e-07,
+      "loss": 1.1639,
+      "step": 15676
+    },
+    {
+      "epoch": 0.9524880004860563,
+      "grad_norm": 0.16596582531929016,
+      "learning_rate": 5.607710366035634e-07,
+      "loss": 1.1504,
+      "step": 15677
+    },
+    {
+      "epoch": 0.9525487575186828,
+      "grad_norm": 0.20105420053005219,
+      "learning_rate": 5.593422608142718e-07,
+      "loss": 1.1802,
+      "step": 15678
+    },
+    {
+      "epoch": 0.9526095145513093,
+      "grad_norm": 0.15446464717388153,
+      "learning_rate": 5.579152972818469e-07,
+      "loss": 1.0664,
+      "step": 15679
+    },
+    {
+      "epoch": 0.9526702715839358,
+      "grad_norm": 0.17659232020378113,
+      "learning_rate": 5.564901460585914e-07,
+      "loss": 1.1251,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9527310286165623,
+      "grad_norm": 0.1474592089653015,
+      "learning_rate": 5.550668071967469e-07,
+      "loss": 1.0833,
+      "step": 15681
+    },
+    {
+      "epoch": 0.9527917856491889,
+      "grad_norm": 0.17089638113975525,
+      "learning_rate": 5.536452807484827e-07,
+      "loss": 1.0889,
+      "step": 15682
+    },
+    {
+      "epoch": 0.9528525426818154,
+      "grad_norm": 0.13901163637638092,
+      "learning_rate": 5.522255667659127e-07,
+      "loss": 1.018,
+      "step": 15683
+    },
+    {
+      "epoch": 0.9529132997144419,
+      "grad_norm": 0.19612663984298706,
+      "learning_rate": 5.508076653010729e-07,
+      "loss": 1.1671,
+      "step": 15684
+    },
+    {
+      "epoch": 0.9529740567470685,
+      "grad_norm": 0.14554022252559662,
+      "learning_rate": 5.493915764059332e-07,
+      "loss": 1.0425,
+      "step": 15685
+    },
+    {
+      "epoch": 0.953034813779695,
+      "grad_norm": 0.18441662192344666,
+      "learning_rate": 5.479773001324074e-07,
+      "loss": 1.2528,
+      "step": 15686
+    },
+    {
+      "epoch": 0.9530955708123215,
+      "grad_norm": 0.08343720436096191,
+      "learning_rate": 5.465648365323317e-07,
+      "loss": 0.9894,
+      "step": 15687
+    },
+    {
+      "epoch": 0.9531563278449481,
+      "grad_norm": 0.2077893167734146,
+      "learning_rate": 5.451541856574815e-07,
+      "loss": 1.1351,
+      "step": 15688
+    },
+    {
+      "epoch": 0.9532170848775746,
+      "grad_norm": 0.18338990211486816,
+      "learning_rate": 5.4374534755956e-07,
+      "loss": 1.1574,
+      "step": 15689
+    },
+    {
+      "epoch": 0.9532778419102012,
+      "grad_norm": 5.325088977813721,
+      "learning_rate": 5.4233832229022e-07,
+      "loss": 1.2211,
+      "step": 15690
+    },
+    {
+      "epoch": 0.9533385989428277,
+      "grad_norm": 0.15941472351551056,
+      "learning_rate": 5.409331099010206e-07,
+      "loss": 1.1056,
+      "step": 15691
+    },
+    {
+      "epoch": 0.9533993559754541,
+      "grad_norm": 0.21715497970581055,
+      "learning_rate": 5.395297104434815e-07,
+      "loss": 1.2403,
+      "step": 15692
+    },
+    {
+      "epoch": 0.9534601130080806,
+      "grad_norm": 0.19840890169143677,
+      "learning_rate": 5.381281239690394e-07,
+      "loss": 1.1974,
+      "step": 15693
+    },
+    {
+      "epoch": 0.9535208700407072,
+      "grad_norm": 0.16425780951976776,
+      "learning_rate": 5.36728350529081e-07,
+      "loss": 1.0497,
+      "step": 15694
+    },
+    {
+      "epoch": 0.9535816270733337,
+      "grad_norm": 0.1002884954214096,
+      "learning_rate": 5.353303901748985e-07,
+      "loss": 0.9873,
+      "step": 15695
+    },
+    {
+      "epoch": 0.9536423841059603,
+      "grad_norm": 0.15883687138557434,
+      "learning_rate": 5.339342429577398e-07,
+      "loss": 1.084,
+      "step": 15696
+    },
+    {
+      "epoch": 0.9537031411385868,
+      "grad_norm": 0.15514416992664337,
+      "learning_rate": 5.325399089287809e-07,
+      "loss": 1.084,
+      "step": 15697
+    },
+    {
+      "epoch": 0.9537638981712133,
+      "grad_norm": 0.1579744517803192,
+      "learning_rate": 5.311473881391415e-07,
+      "loss": 1.1398,
+      "step": 15698
+    },
+    {
+      "epoch": 0.9538246552038399,
+      "grad_norm": 0.10681905597448349,
+      "learning_rate": 5.297566806398535e-07,
+      "loss": 1.0223,
+      "step": 15699
+    },
+    {
+      "epoch": 0.9538854122364664,
+      "grad_norm": 0.9505147933959961,
+      "learning_rate": 5.283677864818981e-07,
+      "loss": 1.0676,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9539461692690929,
+      "grad_norm": 0.12085028737783432,
+      "learning_rate": 5.269807057161847e-07,
+      "loss": 1.0438,
+      "step": 15701
+    },
+    {
+      "epoch": 0.9540069263017195,
+      "grad_norm": 0.15130169689655304,
+      "learning_rate": 5.255954383935502e-07,
+      "loss": 1.0802,
+      "step": 15702
+    },
+    {
+      "epoch": 0.954067683334346,
+      "grad_norm": 0.28984054923057556,
+      "learning_rate": 5.242119845647819e-07,
+      "loss": 1.0393,
+      "step": 15703
+    },
+    {
+      "epoch": 0.9541284403669725,
+      "grad_norm": 0.12057067453861237,
+      "learning_rate": 5.228303442805893e-07,
+      "loss": 1.0825,
+      "step": 15704
+    },
+    {
+      "epoch": 0.954189197399599,
+      "grad_norm": 0.1457018107175827,
+      "learning_rate": 5.214505175916096e-07,
+      "loss": 1.081,
+      "step": 15705
+    },
+    {
+      "epoch": 0.9542499544322255,
+      "grad_norm": 0.747618556022644,
+      "learning_rate": 5.2007250454843e-07,
+      "loss": 1.0604,
+      "step": 15706
+    },
+    {
+      "epoch": 0.954310711464852,
+      "grad_norm": 0.09438946098089218,
+      "learning_rate": 5.186963052015547e-07,
+      "loss": 1.0137,
+      "step": 15707
+    },
+    {
+      "epoch": 0.9543714684974786,
+      "grad_norm": 0.15344533324241638,
+      "learning_rate": 5.173219196014267e-07,
+      "loss": 1.0576,
+      "step": 15708
+    },
+    {
+      "epoch": 0.9544322255301051,
+      "grad_norm": 0.17992642521858215,
+      "learning_rate": 5.159493477984335e-07,
+      "loss": 1.0509,
+      "step": 15709
+    },
+    {
+      "epoch": 0.9544929825627316,
+      "grad_norm": 0.9288654327392578,
+      "learning_rate": 5.145785898428846e-07,
+      "loss": 1.1897,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9545537395953582,
+      "grad_norm": 0.2120366394519806,
+      "learning_rate": 5.132096457850177e-07,
+      "loss": 1.0823,
+      "step": 15711
+    },
+    {
+      "epoch": 0.9546144966279847,
+      "grad_norm": 0.13646164536476135,
+      "learning_rate": 5.118425156750206e-07,
+      "loss": 1.0284,
+      "step": 15712
+    },
+    {
+      "epoch": 0.9546752536606112,
+      "grad_norm": 0.10246030241250992,
+      "learning_rate": 5.10477199562992e-07,
+      "loss": 1.0369,
+      "step": 15713
+    },
+    {
+      "epoch": 0.9547360106932378,
+      "grad_norm": 0.18350543081760406,
+      "learning_rate": 5.091136974990029e-07,
+      "loss": 1.1862,
+      "step": 15714
+    },
+    {
+      "epoch": 0.9547967677258643,
+      "grad_norm": 0.7761063575744629,
+      "learning_rate": 5.07752009533008e-07,
+      "loss": 1.0339,
+      "step": 15715
+    },
+    {
+      "epoch": 0.9548575247584908,
+      "grad_norm": 0.10499266535043716,
+      "learning_rate": 5.063921357149337e-07,
+      "loss": 1.0231,
+      "step": 15716
+    },
+    {
+      "epoch": 0.9549182817911174,
+      "grad_norm": 0.11632519960403442,
+      "learning_rate": 5.050340760946182e-07,
+      "loss": 1.0214,
+      "step": 15717
+    },
+    {
+      "epoch": 0.9549790388237438,
+      "grad_norm": 0.10571163892745972,
+      "learning_rate": 5.036778307218493e-07,
+      "loss": 1.009,
+      "step": 15718
+    },
+    {
+      "epoch": 0.9550397958563703,
+      "grad_norm": 0.11057620495557785,
+      "learning_rate": 5.023233996463317e-07,
+      "loss": 1.0243,
+      "step": 15719
+    },
+    {
+      "epoch": 0.9551005528889969,
+      "grad_norm": 0.19193731248378754,
+      "learning_rate": 5.009707829177257e-07,
+      "loss": 0.9708,
+      "step": 15720
+    },
+    {
+      "epoch": 0.9551613099216234,
+      "grad_norm": 0.18102526664733887,
+      "learning_rate": 4.996199805856028e-07,
+      "loss": 1.0966,
+      "step": 15721
+    },
+    {
+      "epoch": 0.9552220669542499,
+      "grad_norm": 0.15442179143428802,
+      "learning_rate": 4.982709926994733e-07,
+      "loss": 1.0796,
+      "step": 15722
+    },
+    {
+      "epoch": 0.9552828239868765,
+      "grad_norm": 0.2329256236553192,
+      "learning_rate": 4.969238193087867e-07,
+      "loss": 1.1888,
+      "step": 15723
+    },
+    {
+      "epoch": 0.955343581019503,
+      "grad_norm": 0.2275102138519287,
+      "learning_rate": 4.95578460462931e-07,
+      "loss": 1.1093,
+      "step": 15724
+    },
+    {
+      "epoch": 0.9554043380521295,
+      "grad_norm": 0.36220794916152954,
+      "learning_rate": 4.94234916211217e-07,
+      "loss": 1.034,
+      "step": 15725
+    },
+    {
+      "epoch": 0.9554650950847561,
+      "grad_norm": 0.10626080632209778,
+      "learning_rate": 4.928931866028885e-07,
+      "loss": 1.04,
+      "step": 15726
+    },
+    {
+      "epoch": 0.9555258521173826,
+      "grad_norm": 0.18155361711978912,
+      "learning_rate": 4.915532716871341e-07,
+      "loss": 1.1162,
+      "step": 15727
+    },
+    {
+      "epoch": 0.9555866091500091,
+      "grad_norm": 0.20560337603092194,
+      "learning_rate": 4.902151715130587e-07,
+      "loss": 1.133,
+      "step": 15728
+    },
+    {
+      "epoch": 0.9556473661826357,
+      "grad_norm": 0.15575452148914337,
+      "learning_rate": 4.888788861297178e-07,
+      "loss": 1.0354,
+      "step": 15729
+    },
+    {
+      "epoch": 0.9557081232152622,
+      "grad_norm": 0.12168926745653152,
+      "learning_rate": 4.875444155860886e-07,
+      "loss": 0.9853,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9557688802478886,
+      "grad_norm": 0.1620974838733673,
+      "learning_rate": 4.862117599310933e-07,
+      "loss": 1.0908,
+      "step": 15731
+    },
+    {
+      "epoch": 0.9558296372805152,
+      "grad_norm": 0.10455036163330078,
+      "learning_rate": 4.848809192135761e-07,
+      "loss": 1.0511,
+      "step": 15732
+    },
+    {
+      "epoch": 0.9558903943131417,
+      "grad_norm": 0.16169939935207367,
+      "learning_rate": 4.835518934823203e-07,
+      "loss": 1.0938,
+      "step": 15733
+    },
+    {
+      "epoch": 0.9559511513457682,
+      "grad_norm": 0.12686842679977417,
+      "learning_rate": 4.822246827860422e-07,
+      "loss": 0.9902,
+      "step": 15734
+    },
+    {
+      "epoch": 0.9560119083783948,
+      "grad_norm": 1.6132800579071045,
+      "learning_rate": 4.80899287173392e-07,
+      "loss": 1.0305,
+      "step": 15735
+    },
+    {
+      "epoch": 0.9560726654110213,
+      "grad_norm": 0.44043534994125366,
+      "learning_rate": 4.795757066929419e-07,
+      "loss": 1.3478,
+      "step": 15736
+    },
+    {
+      "epoch": 0.9561334224436479,
+      "grad_norm": 0.36503586173057556,
+      "learning_rate": 4.782539413932252e-07,
+      "loss": 1.129,
+      "step": 15737
+    },
+    {
+      "epoch": 0.9561941794762744,
+      "grad_norm": 0.15852870047092438,
+      "learning_rate": 4.769339913226756e-07,
+      "loss": 1.0866,
+      "step": 15738
+    },
+    {
+      "epoch": 0.9562549365089009,
+      "grad_norm": 4.603458404541016,
+      "learning_rate": 4.756158565296875e-07,
+      "loss": 1.0553,
+      "step": 15739
+    },
+    {
+      "epoch": 0.9563156935415275,
+      "grad_norm": 0.9752864837646484,
+      "learning_rate": 4.7429953706257246e-07,
+      "loss": 1.0211,
+      "step": 15740
+    },
+    {
+      "epoch": 0.956376450574154,
+      "grad_norm": 0.12347756326198578,
+      "learning_rate": 4.7298503296958065e-07,
+      "loss": 1.0389,
+      "step": 15741
+    },
+    {
+      "epoch": 0.9564372076067805,
+      "grad_norm": 0.1566256731748581,
+      "learning_rate": 4.716723442988957e-07,
+      "loss": 1.1258,
+      "step": 15742
+    },
+    {
+      "epoch": 0.9564979646394071,
+      "grad_norm": 0.42548832297325134,
+      "learning_rate": 4.7036147109863484e-07,
+      "loss": 1.0828,
+      "step": 15743
+    },
+    {
+      "epoch": 0.9565587216720335,
+      "grad_norm": 0.4719665050506592,
+      "learning_rate": 4.690524134168428e-07,
+      "loss": 1.1058,
+      "step": 15744
+    },
+    {
+      "epoch": 0.95661947870466,
+      "grad_norm": 0.09681335836648941,
+      "learning_rate": 4.677451713015146e-07,
+      "loss": 1.0263,
+      "step": 15745
+    },
+    {
+      "epoch": 0.9566802357372866,
+      "grad_norm": 0.12808607518672943,
+      "learning_rate": 4.66439744800562e-07,
+      "loss": 1.0149,
+      "step": 15746
+    },
+    {
+      "epoch": 0.9567409927699131,
+      "grad_norm": 0.13159635663032532,
+      "learning_rate": 4.651361339618354e-07,
+      "loss": 1.0513,
+      "step": 15747
+    },
+    {
+      "epoch": 0.9568017498025396,
+      "grad_norm": 0.48745205998420715,
+      "learning_rate": 4.63834338833119e-07,
+      "loss": 1.0448,
+      "step": 15748
+    },
+    {
+      "epoch": 0.9568625068351662,
+      "grad_norm": 0.1470884084701538,
+      "learning_rate": 4.6253435946212454e-07,
+      "loss": 1.1253,
+      "step": 15749
+    },
+    {
+      "epoch": 0.9569232638677927,
+      "grad_norm": 1.147767186164856,
+      "learning_rate": 4.612361958965139e-07,
+      "loss": 1.0394,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9569840209004192,
+      "grad_norm": 0.1831953525543213,
+      "learning_rate": 4.5993984818386015e-07,
+      "loss": 1.078,
+      "step": 15751
+    },
+    {
+      "epoch": 0.9570447779330458,
+      "grad_norm": 0.24117903411388397,
+      "learning_rate": 4.586453163716919e-07,
+      "loss": 1.2209,
+      "step": 15752
+    },
+    {
+      "epoch": 0.9571055349656723,
+      "grad_norm": 0.19036448001861572,
+      "learning_rate": 4.5735260050745464e-07,
+      "loss": 1.018,
+      "step": 15753
+    },
+    {
+      "epoch": 0.9571662919982988,
+      "grad_norm": 0.09763475507497787,
+      "learning_rate": 4.5606170063853816e-07,
+      "loss": 0.9756,
+      "step": 15754
+    },
+    {
+      "epoch": 0.9572270490309254,
+      "grad_norm": 0.2130412757396698,
+      "learning_rate": 4.54772616812249e-07,
+      "loss": 1.143,
+      "step": 15755
+    },
+    {
+      "epoch": 0.9572878060635519,
+      "grad_norm": 0.11960317939519882,
+      "learning_rate": 4.5348534907585507e-07,
+      "loss": 1.0223,
+      "step": 15756
+    },
+    {
+      "epoch": 0.9573485630961783,
+      "grad_norm": 0.47025609016418457,
+      "learning_rate": 4.5219989747652404e-07,
+      "loss": 1.0692,
+      "step": 15757
+    },
+    {
+      "epoch": 0.9574093201288049,
+      "grad_norm": 0.1826414167881012,
+      "learning_rate": 4.5091626206139047e-07,
+      "loss": 1.1241,
+      "step": 15758
+    },
+    {
+      "epoch": 0.9574700771614314,
+      "grad_norm": 0.1296650618314743,
+      "learning_rate": 4.496344428774946e-07,
+      "loss": 1.0579,
+      "step": 15759
+    },
+    {
+      "epoch": 0.9575308341940579,
+      "grad_norm": 0.35346952080726624,
+      "learning_rate": 4.483544399718265e-07,
+      "loss": 1.0649,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9575915912266845,
+      "grad_norm": 0.08745156228542328,
+      "learning_rate": 4.4707625339130976e-07,
+      "loss": 1.06,
+      "step": 15761
+    },
+    {
+      "epoch": 0.957652348259311,
+      "grad_norm": 0.10423334687948227,
+      "learning_rate": 4.4579988318278477e-07,
+      "loss": 1.0251,
+      "step": 15762
+    },
+    {
+      "epoch": 0.9577131052919375,
+      "grad_norm": 0.12803082168102264,
+      "learning_rate": 4.4452532939304734e-07,
+      "loss": 1.0227,
+      "step": 15763
+    },
+    {
+      "epoch": 0.9577738623245641,
+      "grad_norm": 0.1084895208477974,
+      "learning_rate": 4.432525920688102e-07,
+      "loss": 1.0396,
+      "step": 15764
+    },
+    {
+      "epoch": 0.9578346193571906,
+      "grad_norm": 0.19573159515857697,
+      "learning_rate": 4.4198167125672485e-07,
+      "loss": 1.0564,
+      "step": 15765
+    },
+    {
+      "epoch": 0.9578953763898171,
+      "grad_norm": 1.038779854774475,
+      "learning_rate": 4.4071256700338737e-07,
+      "loss": 1.087,
+      "step": 15766
+    },
+    {
+      "epoch": 0.9579561334224437,
+      "grad_norm": 0.0839330330491066,
+      "learning_rate": 4.3944527935531053e-07,
+      "loss": 1.0192,
+      "step": 15767
+    },
+    {
+      "epoch": 0.9580168904550702,
+      "grad_norm": 0.2146690934896469,
+      "learning_rate": 4.381798083589461e-07,
+      "loss": 1.1401,
+      "step": 15768
+    },
+    {
+      "epoch": 0.9580776474876967,
+      "grad_norm": 0.14495867490768433,
+      "learning_rate": 4.3691615406067365e-07,
+      "loss": 1.0922,
+      "step": 15769
+    },
+    {
+      "epoch": 0.9581384045203232,
+      "grad_norm": 0.13484595715999603,
+      "learning_rate": 4.356543165068283e-07,
+      "loss": 1.0681,
+      "step": 15770
+    },
+    {
+      "epoch": 0.9581991615529497,
+      "grad_norm": 0.11026013642549515,
+      "learning_rate": 4.3439429574365086e-07,
+      "loss": 1.0064,
+      "step": 15771
+    },
+    {
+      "epoch": 0.9582599185855762,
+      "grad_norm": 0.1724824160337448,
+      "learning_rate": 4.331360918173322e-07,
+      "loss": 1.1284,
+      "step": 15772
+    },
+    {
+      "epoch": 0.9583206756182028,
+      "grad_norm": 0.10396581888198853,
+      "learning_rate": 4.3187970477399087e-07,
+      "loss": 1.0039,
+      "step": 15773
+    },
+    {
+      "epoch": 0.9583814326508293,
+      "grad_norm": 0.11033745855093002,
+      "learning_rate": 4.306251346596846e-07,
+      "loss": 1.0926,
+      "step": 15774
+    },
+    {
+      "epoch": 0.9584421896834558,
+      "grad_norm": 0.09618641436100006,
+      "learning_rate": 4.293723815203876e-07,
+      "loss": 1.0383,
+      "step": 15775
+    },
+    {
+      "epoch": 0.9585029467160824,
+      "grad_norm": 0.10459344834089279,
+      "learning_rate": 4.2812144540203545e-07,
+      "loss": 1.0,
+      "step": 15776
+    },
+    {
+      "epoch": 0.9585637037487089,
+      "grad_norm": 0.13060784339904785,
+      "learning_rate": 4.2687232635046923e-07,
+      "loss": 1.0797,
+      "step": 15777
+    },
+    {
+      "epoch": 0.9586244607813355,
+      "grad_norm": 0.16173116862773895,
+      "learning_rate": 4.256250244114801e-07,
+      "loss": 1.1165,
+      "step": 15778
+    },
+    {
+      "epoch": 0.958685217813962,
+      "grad_norm": 0.2411125749349594,
+      "learning_rate": 4.2437953963078704e-07,
+      "loss": 1.1622,
+      "step": 15779
+    },
+    {
+      "epoch": 0.9587459748465885,
+      "grad_norm": 0.11841461062431335,
+      "learning_rate": 4.231358720540479e-07,
+      "loss": 1.0184,
+      "step": 15780
+    },
+    {
+      "epoch": 0.9588067318792151,
+      "grad_norm": 0.2478577345609665,
+      "learning_rate": 4.2189402172684856e-07,
+      "loss": 1.3158,
+      "step": 15781
+    },
+    {
+      "epoch": 0.9588674889118416,
+      "grad_norm": 0.24367450177669525,
+      "learning_rate": 4.206539886947025e-07,
+      "loss": 1.0433,
+      "step": 15782
+    },
+    {
+      "epoch": 0.9589282459444681,
+      "grad_norm": 0.110954150557518,
+      "learning_rate": 4.194157730030679e-07,
+      "loss": 1.0466,
+      "step": 15783
+    },
+    {
+      "epoch": 0.9589890029770946,
+      "grad_norm": 0.10849586874246597,
+      "learning_rate": 4.1817937469733613e-07,
+      "loss": 1.0486,
+      "step": 15784
+    },
+    {
+      "epoch": 0.9590497600097211,
+      "grad_norm": 0.13907769322395325,
+      "learning_rate": 4.169447938228155e-07,
+      "loss": 1.0129,
+      "step": 15785
+    },
+    {
+      "epoch": 0.9591105170423476,
+      "grad_norm": 0.10487202554941177,
+      "learning_rate": 4.1571203042477523e-07,
+      "loss": 1.0686,
+      "step": 15786
+    },
+    {
+      "epoch": 0.9591712740749742,
+      "grad_norm": 0.15854674577713013,
+      "learning_rate": 4.144810845483904e-07,
+      "loss": 1.0877,
+      "step": 15787
+    },
+    {
+      "epoch": 0.9592320311076007,
+      "grad_norm": 0.11448860168457031,
+      "learning_rate": 4.1325195623879153e-07,
+      "loss": 1.0534,
+      "step": 15788
+    },
+    {
+      "epoch": 0.9592927881402272,
+      "grad_norm": 0.1065724790096283,
+      "learning_rate": 4.120246455410204e-07,
+      "loss": 1.0341,
+      "step": 15789
+    },
+    {
+      "epoch": 0.9593535451728538,
+      "grad_norm": 0.14623962342739105,
+      "learning_rate": 4.1079915250007427e-07,
+      "loss": 1.0399,
+      "step": 15790
+    },
+    {
+      "epoch": 0.9594143022054803,
+      "grad_norm": 0.12341058254241943,
+      "learning_rate": 4.0957547716086733e-07,
+      "loss": 1.0453,
+      "step": 15791
+    },
+    {
+      "epoch": 0.9594750592381068,
+      "grad_norm": 0.1551622897386551,
+      "learning_rate": 4.083536195682636e-07,
+      "loss": 1.0949,
+      "step": 15792
+    },
+    {
+      "epoch": 0.9595358162707334,
+      "grad_norm": 0.24149124324321747,
+      "learning_rate": 4.0713357976703834e-07,
+      "loss": 1.067,
+      "step": 15793
+    },
+    {
+      "epoch": 0.9595965733033599,
+      "grad_norm": 0.1642986536026001,
+      "learning_rate": 4.059153578019226e-07,
+      "loss": 1.098,
+      "step": 15794
+    },
+    {
+      "epoch": 0.9596573303359864,
+      "grad_norm": 0.09819808602333069,
+      "learning_rate": 4.0469895371756383e-07,
+      "loss": 0.9867,
+      "step": 15795
+    },
+    {
+      "epoch": 0.959718087368613,
+      "grad_norm": 0.17577776312828064,
+      "learning_rate": 4.034843675585487e-07,
+      "loss": 1.0885,
+      "step": 15796
+    },
+    {
+      "epoch": 0.9597788444012394,
+      "grad_norm": 0.13872236013412476,
+      "learning_rate": 4.022715993694026e-07,
+      "loss": 1.0661,
+      "step": 15797
+    },
+    {
+      "epoch": 0.9598396014338659,
+      "grad_norm": 0.1984054297208786,
+      "learning_rate": 4.0106064919457897e-07,
+      "loss": 1.0258,
+      "step": 15798
+    },
+    {
+      "epoch": 0.9599003584664925,
+      "grad_norm": 0.09982796758413315,
+      "learning_rate": 3.9985151707846444e-07,
+      "loss": 0.9987,
+      "step": 15799
+    },
+    {
+      "epoch": 0.959961115499119,
+      "grad_norm": 0.17996053397655487,
+      "learning_rate": 3.986442030653792e-07,
+      "loss": 1.1298,
+      "step": 15800
+    },
+    {
+      "epoch": 0.9600218725317455,
+      "grad_norm": 0.10261844843626022,
+      "learning_rate": 3.9743870719957664e-07,
+      "loss": 1.0492,
+      "step": 15801
+    },
+    {
+      "epoch": 0.9600826295643721,
+      "grad_norm": 0.11543924361467361,
+      "learning_rate": 3.962350295252493e-07,
+      "loss": 1.0396,
+      "step": 15802
+    },
+    {
+      "epoch": 0.9601433865969986,
+      "grad_norm": 0.22956831753253937,
+      "learning_rate": 3.9503317008651173e-07,
+      "loss": 1.1858,
+      "step": 15803
+    },
+    {
+      "epoch": 0.9602041436296251,
+      "grad_norm": 0.18623453378677368,
+      "learning_rate": 3.938331289274233e-07,
+      "loss": 1.112,
+      "step": 15804
+    },
+    {
+      "epoch": 0.9602649006622517,
+      "grad_norm": 0.17458507418632507,
+      "learning_rate": 3.9263490609197094e-07,
+      "loss": 1.1401,
+      "step": 15805
+    },
+    {
+      "epoch": 0.9603256576948782,
+      "grad_norm": 0.3694145679473877,
+      "learning_rate": 3.914385016240751e-07,
+      "loss": 1.0779,
+      "step": 15806
+    },
+    {
+      "epoch": 0.9603864147275047,
+      "grad_norm": 0.13405080139636993,
+      "learning_rate": 3.902439155675897e-07,
+      "loss": 1.0505,
+      "step": 15807
+    },
+    {
+      "epoch": 0.9604471717601313,
+      "grad_norm": 0.23359820246696472,
+      "learning_rate": 3.8905114796630724e-07,
+      "loss": 1.0955,
+      "step": 15808
+    },
+    {
+      "epoch": 0.9605079287927578,
+      "grad_norm": 0.10692767798900604,
+      "learning_rate": 3.878601988639374e-07,
+      "loss": 1.0553,
+      "step": 15809
+    },
+    {
+      "epoch": 0.9605686858253842,
+      "grad_norm": 0.47428059577941895,
+      "learning_rate": 3.8667106830414525e-07,
+      "loss": 1.2993,
+      "step": 15810
+    },
+    {
+      "epoch": 0.9606294428580108,
+      "grad_norm": 0.25792521238327026,
+      "learning_rate": 3.8548375633051247e-07,
+      "loss": 1.1253,
+      "step": 15811
+    },
+    {
+      "epoch": 0.9606901998906373,
+      "grad_norm": 0.1352255940437317,
+      "learning_rate": 3.8429826298656544e-07,
+      "loss": 1.0429,
+      "step": 15812
+    },
+    {
+      "epoch": 0.9607509569232638,
+      "grad_norm": 0.09801478683948517,
+      "learning_rate": 3.831145883157528e-07,
+      "loss": 1.0848,
+      "step": 15813
+    },
+    {
+      "epoch": 0.9608117139558904,
+      "grad_norm": 0.11662235110998154,
+      "learning_rate": 3.81932732361473e-07,
+      "loss": 1.0356,
+      "step": 15814
+    },
+    {
+      "epoch": 0.9608724709885169,
+      "grad_norm": 0.13122665882110596,
+      "learning_rate": 3.8075269516703596e-07,
+      "loss": 1.0,
+      "step": 15815
+    },
+    {
+      "epoch": 0.9609332280211434,
+      "grad_norm": 0.11306820809841156,
+      "learning_rate": 3.795744767756959e-07,
+      "loss": 1.0648,
+      "step": 15816
+    },
+    {
+      "epoch": 0.96099398505377,
+      "grad_norm": 0.22960704565048218,
+      "learning_rate": 3.783980772306517e-07,
+      "loss": 1.1373,
+      "step": 15817
+    },
+    {
+      "epoch": 0.9610547420863965,
+      "grad_norm": 0.1593761444091797,
+      "learning_rate": 3.772234965750132e-07,
+      "loss": 1.1185,
+      "step": 15818
+    },
+    {
+      "epoch": 0.961115499119023,
+      "grad_norm": 0.09501218795776367,
+      "learning_rate": 3.760507348518405e-07,
+      "loss": 1.0277,
+      "step": 15819
+    },
+    {
+      "epoch": 0.9611762561516496,
+      "grad_norm": 0.13905061781406403,
+      "learning_rate": 3.748797921041214e-07,
+      "loss": 1.0464,
+      "step": 15820
+    },
+    {
+      "epoch": 0.9612370131842761,
+      "grad_norm": 1.2092803716659546,
+      "learning_rate": 3.7371066837477154e-07,
+      "loss": 1.0348,
+      "step": 15821
+    },
+    {
+      "epoch": 0.9612977702169027,
+      "grad_norm": 0.21236439049243927,
+      "learning_rate": 3.7254336370665666e-07,
+      "loss": 1.0205,
+      "step": 15822
+    },
+    {
+      "epoch": 0.9613585272495291,
+      "grad_norm": 0.11326601356267929,
+      "learning_rate": 3.7137787814255365e-07,
+      "loss": 1.0609,
+      "step": 15823
+    },
+    {
+      "epoch": 0.9614192842821556,
+      "grad_norm": 0.1466970294713974,
+      "learning_rate": 3.702142117251839e-07,
+      "loss": 1.0726,
+      "step": 15824
+    },
+    {
+      "epoch": 0.9614800413147822,
+      "grad_norm": 0.18312968313694,
+      "learning_rate": 3.690523644972077e-07,
+      "loss": 1.1707,
+      "step": 15825
+    },
+    {
+      "epoch": 0.9615407983474087,
+      "grad_norm": 0.10370973497629166,
+      "learning_rate": 3.678923365012132e-07,
+      "loss": 0.987,
+      "step": 15826
+    },
+    {
+      "epoch": 0.9616015553800352,
+      "grad_norm": 0.12054158747196198,
+      "learning_rate": 3.667341277797165e-07,
+      "loss": 1.0349,
+      "step": 15827
+    },
+    {
+      "epoch": 0.9616623124126618,
+      "grad_norm": 0.09524430334568024,
+      "learning_rate": 3.6557773837517797e-07,
+      "loss": 1.0522,
+      "step": 15828
+    },
+    {
+      "epoch": 0.9617230694452883,
+      "grad_norm": 0.12918563187122345,
+      "learning_rate": 3.6442316832998035e-07,
+      "loss": 1.0663,
+      "step": 15829
+    },
+    {
+      "epoch": 0.9617838264779148,
+      "grad_norm": 0.13953079283237457,
+      "learning_rate": 3.6327041768643987e-07,
+      "loss": 1.0688,
+      "step": 15830
+    },
+    {
+      "epoch": 0.9618445835105414,
+      "grad_norm": 0.1647307574748993,
+      "learning_rate": 3.621194864868227e-07,
+      "loss": 1.0776,
+      "step": 15831
+    },
+    {
+      "epoch": 0.9619053405431679,
+      "grad_norm": 0.11573868989944458,
+      "learning_rate": 3.609703747733062e-07,
+      "loss": 1.0396,
+      "step": 15832
+    },
+    {
+      "epoch": 0.9619660975757944,
+      "grad_norm": 0.22192254662513733,
+      "learning_rate": 3.598230825880178e-07,
+      "loss": 1.0959,
+      "step": 15833
+    },
+    {
+      "epoch": 0.962026854608421,
+      "grad_norm": 0.10462997108697891,
+      "learning_rate": 3.586776099730127e-07,
+      "loss": 0.9694,
+      "step": 15834
+    },
+    {
+      "epoch": 0.9620876116410475,
+      "grad_norm": 0.13647164404392242,
+      "learning_rate": 3.57533956970274e-07,
+      "loss": 1.0748,
+      "step": 15835
+    },
+    {
+      "epoch": 0.9621483686736739,
+      "grad_norm": 0.10433114320039749,
+      "learning_rate": 3.5639212362172383e-07,
+      "loss": 1.0215,
+      "step": 15836
+    },
+    {
+      "epoch": 0.9622091257063005,
+      "grad_norm": 0.2004692107439041,
+      "learning_rate": 3.5525210996921187e-07,
+      "loss": 1.1962,
+      "step": 15837
+    },
+    {
+      "epoch": 0.962269882738927,
+      "grad_norm": 1.9959053993225098,
+      "learning_rate": 3.5411391605453257e-07,
+      "loss": 1.101,
+      "step": 15838
+    },
+    {
+      "epoch": 0.9623306397715535,
+      "grad_norm": 0.23181526362895966,
+      "learning_rate": 3.529775419194081e-07,
+      "loss": 1.0963,
+      "step": 15839
+    },
+    {
+      "epoch": 0.9623913968041801,
+      "grad_norm": 0.1325407177209854,
+      "learning_rate": 3.5184298760548296e-07,
+      "loss": 1.0725,
+      "step": 15840
+    },
+    {
+      "epoch": 0.9624521538368066,
+      "grad_norm": 0.19533170759677887,
+      "learning_rate": 3.507102531543516e-07,
+      "loss": 1.1787,
+      "step": 15841
+    },
+    {
+      "epoch": 0.9625129108694331,
+      "grad_norm": 31.75570297241211,
+      "learning_rate": 3.495793386075308e-07,
+      "loss": 1.091,
+      "step": 15842
+    },
+    {
+      "epoch": 0.9625736679020597,
+      "grad_norm": 0.13796232640743256,
+      "learning_rate": 3.484502440064818e-07,
+      "loss": 0.9974,
+      "step": 15843
+    },
+    {
+      "epoch": 0.9626344249346862,
+      "grad_norm": 0.08848493546247482,
+      "learning_rate": 3.473229693925828e-07,
+      "loss": 1.0555,
+      "step": 15844
+    },
+    {
+      "epoch": 0.9626951819673127,
+      "grad_norm": 0.09830200672149658,
+      "learning_rate": 3.461975148071561e-07,
+      "loss": 0.99,
+      "step": 15845
+    },
+    {
+      "epoch": 0.9627559389999393,
+      "grad_norm": 0.16677135229110718,
+      "learning_rate": 3.4507388029146324e-07,
+      "loss": 1.0793,
+      "step": 15846
+    },
+    {
+      "epoch": 0.9628166960325658,
+      "grad_norm": 0.10981838405132294,
+      "learning_rate": 3.439520658866824e-07,
+      "loss": 1.0254,
+      "step": 15847
+    },
+    {
+      "epoch": 0.9628774530651923,
+      "grad_norm": 0.14670252799987793,
+      "learning_rate": 3.4283207163393614e-07,
+      "loss": 1.0497,
+      "step": 15848
+    },
+    {
+      "epoch": 0.9629382100978188,
+      "grad_norm": 0.1654321849346161,
+      "learning_rate": 3.417138975742806e-07,
+      "loss": 1.0519,
+      "step": 15849
+    },
+    {
+      "epoch": 0.9629989671304453,
+      "grad_norm": 0.1254972219467163,
+      "learning_rate": 3.405975437486997e-07,
+      "loss": 0.9463,
+      "step": 15850
+    },
+    {
+      "epoch": 0.9630597241630718,
+      "grad_norm": 0.1553478091955185,
+      "learning_rate": 3.3948301019811056e-07,
+      "loss": 1.1441,
+      "step": 15851
+    },
+    {
+      "epoch": 0.9631204811956984,
+      "grad_norm": 0.14452053606510162,
+      "learning_rate": 3.383702969633751e-07,
+      "loss": 1.0105,
+      "step": 15852
+    },
+    {
+      "epoch": 0.9631812382283249,
+      "grad_norm": 0.17748470604419708,
+      "learning_rate": 3.372594040852772e-07,
+      "loss": 1.0594,
+      "step": 15853
+    },
+    {
+      "epoch": 0.9632419952609514,
+      "grad_norm": 0.19587081670761108,
+      "learning_rate": 3.3615033160453444e-07,
+      "loss": 1.2383,
+      "step": 15854
+    },
+    {
+      "epoch": 0.963302752293578,
+      "grad_norm": 0.09949034452438354,
+      "learning_rate": 3.350430795617976e-07,
+      "loss": 1.0072,
+      "step": 15855
+    },
+    {
+      "epoch": 0.9633635093262045,
+      "grad_norm": 0.1357887089252472,
+      "learning_rate": 3.339376479976619e-07,
+      "loss": 1.0904,
+      "step": 15856
+    },
+    {
+      "epoch": 0.963424266358831,
+      "grad_norm": 0.2644502520561218,
+      "learning_rate": 3.328340369526395e-07,
+      "loss": 1.1511,
+      "step": 15857
+    },
+    {
+      "epoch": 0.9634850233914576,
+      "grad_norm": 0.5562261343002319,
+      "learning_rate": 3.317322464671868e-07,
+      "loss": 1.1867,
+      "step": 15858
+    },
+    {
+      "epoch": 0.9635457804240841,
+      "grad_norm": 0.16169697046279907,
+      "learning_rate": 3.3063227658168827e-07,
+      "loss": 1.1308,
+      "step": 15859
+    },
+    {
+      "epoch": 0.9636065374567107,
+      "grad_norm": 0.11124769598245621,
+      "learning_rate": 3.295341273364616e-07,
+      "loss": 1.0593,
+      "step": 15860
+    },
+    {
+      "epoch": 0.9636672944893372,
+      "grad_norm": 0.16454099118709564,
+      "learning_rate": 3.2843779877176904e-07,
+      "loss": 1.0574,
+      "step": 15861
+    },
+    {
+      "epoch": 0.9637280515219636,
+      "grad_norm": 0.6228294968605042,
+      "learning_rate": 3.273432909277896e-07,
+      "loss": 1.0999,
+      "step": 15862
+    },
+    {
+      "epoch": 0.9637888085545901,
+      "grad_norm": 0.17999854683876038,
+      "learning_rate": 3.262506038446356e-07,
+      "loss": 1.1772,
+      "step": 15863
+    },
+    {
+      "epoch": 0.9638495655872167,
+      "grad_norm": 0.20635171234607697,
+      "learning_rate": 3.2515973756237496e-07,
+      "loss": 1.0948,
+      "step": 15864
+    },
+    {
+      "epoch": 0.9639103226198432,
+      "grad_norm": 0.12803366780281067,
+      "learning_rate": 3.240706921209813e-07,
+      "loss": 1.0711,
+      "step": 15865
+    },
+    {
+      "epoch": 0.9639710796524698,
+      "grad_norm": 0.12278302013874054,
+      "learning_rate": 3.229834675603838e-07,
+      "loss": 1.0304,
+      "step": 15866
+    },
+    {
+      "epoch": 0.9640318366850963,
+      "grad_norm": 0.09925037622451782,
+      "learning_rate": 3.2189806392042275e-07,
+      "loss": 1.0281,
+      "step": 15867
+    },
+    {
+      "epoch": 0.9640925937177228,
+      "grad_norm": 0.12721189856529236,
+      "learning_rate": 3.208144812408942e-07,
+      "loss": 1.0925,
+      "step": 15868
+    },
+    {
+      "epoch": 0.9641533507503494,
+      "grad_norm": 0.12747856974601746,
+      "learning_rate": 3.197327195615163e-07,
+      "loss": 1.0564,
+      "step": 15869
+    },
+    {
+      "epoch": 0.9642141077829759,
+      "grad_norm": 0.19266590476036072,
+      "learning_rate": 3.1865277892193515e-07,
+      "loss": 1.1516,
+      "step": 15870
+    },
+    {
+      "epoch": 0.9642748648156024,
+      "grad_norm": 0.127415731549263,
+      "learning_rate": 3.1757465936174127e-07,
+      "loss": 1.0182,
+      "step": 15871
+    },
+    {
+      "epoch": 0.964335621848229,
+      "grad_norm": 0.15527105331420898,
+      "learning_rate": 3.1649836092045306e-07,
+      "loss": 1.0965,
+      "step": 15872
+    },
+    {
+      "epoch": 0.9643963788808555,
+      "grad_norm": 0.16879485547542572,
+      "learning_rate": 3.1542388363751676e-07,
+      "loss": 1.0524,
+      "step": 15873
+    },
+    {
+      "epoch": 0.964457135913482,
+      "grad_norm": 0.18120986223220825,
+      "learning_rate": 3.143512275523286e-07,
+      "loss": 1.1234,
+      "step": 15874
+    },
+    {
+      "epoch": 0.9645178929461085,
+      "grad_norm": 0.16349101066589355,
+      "learning_rate": 3.132803927041961e-07,
+      "loss": 1.0923,
+      "step": 15875
+    },
+    {
+      "epoch": 0.964578649978735,
+      "grad_norm": 0.22919566929340363,
+      "learning_rate": 3.1221137913237107e-07,
+      "loss": 1.0207,
+      "step": 15876
+    },
+    {
+      "epoch": 0.9646394070113615,
+      "grad_norm": 0.1649797111749649,
+      "learning_rate": 3.111441868760445e-07,
+      "loss": 1.0874,
+      "step": 15877
+    },
+    {
+      "epoch": 0.9647001640439881,
+      "grad_norm": 0.22419624030590057,
+      "learning_rate": 3.100788159743351e-07,
+      "loss": 1.1291,
+      "step": 15878
+    },
+    {
+      "epoch": 0.9647609210766146,
+      "grad_norm": 0.1662607342004776,
+      "learning_rate": 3.090152664662893e-07,
+      "loss": 1.2389,
+      "step": 15879
+    },
+    {
+      "epoch": 0.9648216781092411,
+      "grad_norm": 0.10185400396585464,
+      "learning_rate": 3.0795353839089826e-07,
+      "loss": 1.0005,
+      "step": 15880
+    },
+    {
+      "epoch": 0.9648824351418677,
+      "grad_norm": 0.11787091195583344,
+      "learning_rate": 3.0689363178706964e-07,
+      "loss": 1.0436,
+      "step": 15881
+    },
+    {
+      "epoch": 0.9649431921744942,
+      "grad_norm": 0.10902456194162369,
+      "learning_rate": 3.0583554669366687e-07,
+      "loss": 1.0051,
+      "step": 15882
+    },
+    {
+      "epoch": 0.9650039492071207,
+      "grad_norm": 0.2078927606344223,
+      "learning_rate": 3.047792831494589e-07,
+      "loss": 1.0755,
+      "step": 15883
+    },
+    {
+      "epoch": 0.9650647062397473,
+      "grad_norm": 0.15197834372520447,
+      "learning_rate": 3.0372484119318143e-07,
+      "loss": 1.0498,
+      "step": 15884
+    },
+    {
+      "epoch": 0.9651254632723738,
+      "grad_norm": 0.11288712918758392,
+      "learning_rate": 3.0267222086347026e-07,
+      "loss": 1.0557,
+      "step": 15885
+    },
+    {
+      "epoch": 0.9651862203050003,
+      "grad_norm": 1.0391411781311035,
+      "learning_rate": 3.0162142219891663e-07,
+      "loss": 1.0546,
+      "step": 15886
+    },
+    {
+      "epoch": 0.9652469773376269,
+      "grad_norm": 0.23355908691883087,
+      "learning_rate": 3.005724452380343e-07,
+      "loss": 1.1244,
+      "step": 15887
+    },
+    {
+      "epoch": 0.9653077343702534,
+      "grad_norm": 0.21837641298770905,
+      "learning_rate": 2.995252900192813e-07,
+      "loss": 1.1245,
+      "step": 15888
+    },
+    {
+      "epoch": 0.9653684914028798,
+      "grad_norm": 0.33931583166122437,
+      "learning_rate": 2.984799565810326e-07,
+      "loss": 1.1074,
+      "step": 15889
+    },
+    {
+      "epoch": 0.9654292484355064,
+      "grad_norm": 0.5034958124160767,
+      "learning_rate": 2.974364449616074e-07,
+      "loss": 1.0286,
+      "step": 15890
+    },
+    {
+      "epoch": 0.9654900054681329,
+      "grad_norm": 0.11125493794679642,
+      "learning_rate": 2.9639475519925297e-07,
+      "loss": 1.0886,
+      "step": 15891
+    },
+    {
+      "epoch": 0.9655507625007594,
+      "grad_norm": 0.1079137921333313,
+      "learning_rate": 2.95354887332161e-07,
+      "loss": 1.0344,
+      "step": 15892
+    },
+    {
+      "epoch": 0.965611519533386,
+      "grad_norm": 2.671691656112671,
+      "learning_rate": 2.943168413984454e-07,
+      "loss": 1.2258,
+      "step": 15893
+    },
+    {
+      "epoch": 0.9656722765660125,
+      "grad_norm": 0.10957320034503937,
+      "learning_rate": 2.9328061743614796e-07,
+      "loss": 1.0921,
+      "step": 15894
+    },
+    {
+      "epoch": 0.965733033598639,
+      "grad_norm": 0.4058312177658081,
+      "learning_rate": 2.922462154832606e-07,
+      "loss": 1.2526,
+      "step": 15895
+    },
+    {
+      "epoch": 0.9657937906312656,
+      "grad_norm": 0.13628971576690674,
+      "learning_rate": 2.912136355776973e-07,
+      "loss": 1.0909,
+      "step": 15896
+    },
+    {
+      "epoch": 0.9658545476638921,
+      "grad_norm": 0.11478254199028015,
+      "learning_rate": 2.9018287775730566e-07,
+      "loss": 1.002,
+      "step": 15897
+    },
+    {
+      "epoch": 0.9659153046965187,
+      "grad_norm": 0.2863271236419678,
+      "learning_rate": 2.891539420598721e-07,
+      "loss": 1.0041,
+      "step": 15898
+    },
+    {
+      "epoch": 0.9659760617291452,
+      "grad_norm": 0.10050107538700104,
+      "learning_rate": 2.881268285231054e-07,
+      "loss": 1.0308,
+      "step": 15899
+    },
+    {
+      "epoch": 0.9660368187617717,
+      "grad_norm": 0.10459680110216141,
+      "learning_rate": 2.871015371846586e-07,
+      "loss": 1.05,
+      "step": 15900
+    },
+    {
+      "epoch": 0.9660975757943983,
+      "grad_norm": 0.12734566628932953,
+      "learning_rate": 2.860780680821185e-07,
+      "loss": 1.004,
+      "step": 15901
+    },
+    {
+      "epoch": 0.9661583328270247,
+      "grad_norm": 0.11034282296895981,
+      "learning_rate": 2.850564212529938e-07,
+      "loss": 1.038,
+      "step": 15902
+    },
+    {
+      "epoch": 0.9662190898596512,
+      "grad_norm": 0.1617112010717392,
+      "learning_rate": 2.84036596734738e-07,
+      "loss": 1.0622,
+      "step": 15903
+    },
+    {
+      "epoch": 0.9662798468922777,
+      "grad_norm": 0.21670573949813843,
+      "learning_rate": 2.830185945647268e-07,
+      "loss": 1.1696,
+      "step": 15904
+    },
+    {
+      "epoch": 0.9663406039249043,
+      "grad_norm": 0.10613535344600677,
+      "learning_rate": 2.820024147802802e-07,
+      "loss": 1.0534,
+      "step": 15905
+    },
+    {
+      "epoch": 0.9664013609575308,
+      "grad_norm": 0.12204107642173767,
+      "learning_rate": 2.809880574186463e-07,
+      "loss": 1.0461,
+      "step": 15906
+    },
+    {
+      "epoch": 0.9664621179901574,
+      "grad_norm": 1.8109042644500732,
+      "learning_rate": 2.799755225170064e-07,
+      "loss": 1.1054,
+      "step": 15907
+    },
+    {
+      "epoch": 0.9665228750227839,
+      "grad_norm": 0.1148768812417984,
+      "learning_rate": 2.789648101124753e-07,
+      "loss": 1.0529,
+      "step": 15908
+    },
+    {
+      "epoch": 0.9665836320554104,
+      "grad_norm": 0.12113587558269501,
+      "learning_rate": 2.7795592024209563e-07,
+      "loss": 1.079,
+      "step": 15909
+    },
+    {
+      "epoch": 0.966644389088037,
+      "grad_norm": 0.14573897421360016,
+      "learning_rate": 2.769488529428543e-07,
+      "loss": 1.0579,
+      "step": 15910
+    },
+    {
+      "epoch": 0.9667051461206635,
+      "grad_norm": 0.20230604708194733,
+      "learning_rate": 2.7594360825166644e-07,
+      "loss": 1.1347,
+      "step": 15911
+    },
+    {
+      "epoch": 0.96676590315329,
+      "grad_norm": 0.12635920941829681,
+      "learning_rate": 2.7494018620537466e-07,
+      "loss": 1.0587,
+      "step": 15912
+    },
+    {
+      "epoch": 0.9668266601859166,
+      "grad_norm": 0.10484524816274643,
+      "learning_rate": 2.739385868407607e-07,
+      "loss": 0.9967,
+      "step": 15913
+    },
+    {
+      "epoch": 0.9668874172185431,
+      "grad_norm": 0.24736304581165314,
+      "learning_rate": 2.7293881019454513e-07,
+      "loss": 1.1609,
+      "step": 15914
+    },
+    {
+      "epoch": 0.9669481742511695,
+      "grad_norm": 0.12311619520187378,
+      "learning_rate": 2.719408563033654e-07,
+      "loss": 1.0413,
+      "step": 15915
+    },
+    {
+      "epoch": 0.9670089312837961,
+      "grad_norm": 0.1055658683180809,
+      "learning_rate": 2.709447252038089e-07,
+      "loss": 1.0791,
+      "step": 15916
+    },
+    {
+      "epoch": 0.9670696883164226,
+      "grad_norm": 0.15936604142189026,
+      "learning_rate": 2.6995041693237965e-07,
+      "loss": 1.0079,
+      "step": 15917
+    },
+    {
+      "epoch": 0.9671304453490491,
+      "grad_norm": 0.560377836227417,
+      "learning_rate": 2.689579315255375e-07,
+      "loss": 1.3305,
+      "step": 15918
+    },
+    {
+      "epoch": 0.9671912023816757,
+      "grad_norm": 0.17118674516677856,
+      "learning_rate": 2.679672690196533e-07,
+      "loss": 1.1252,
+      "step": 15919
+    },
+    {
+      "epoch": 0.9672519594143022,
+      "grad_norm": 0.3924512565135956,
+      "learning_rate": 2.669784294510369e-07,
+      "loss": 1.1713,
+      "step": 15920
+    },
+    {
+      "epoch": 0.9673127164469287,
+      "grad_norm": 0.1629311889410019,
+      "learning_rate": 2.659914128559426e-07,
+      "loss": 1.0323,
+      "step": 15921
+    },
+    {
+      "epoch": 0.9673734734795553,
+      "grad_norm": 0.1835370510816574,
+      "learning_rate": 2.6500621927054715e-07,
+      "loss": 1.1132,
+      "step": 15922
+    },
+    {
+      "epoch": 0.9674342305121818,
+      "grad_norm": 6.135840892791748,
+      "learning_rate": 2.6402284873096594e-07,
+      "loss": 1.038,
+      "step": 15923
+    },
+    {
+      "epoch": 0.9674949875448083,
+      "grad_norm": 0.21831414103507996,
+      "learning_rate": 2.630413012732369e-07,
+      "loss": 1.1087,
+      "step": 15924
+    },
+    {
+      "epoch": 0.9675557445774349,
+      "grad_norm": 0.11834456771612167,
+      "learning_rate": 2.6206157693333677e-07,
+      "loss": 1.0419,
+      "step": 15925
+    },
+    {
+      "epoch": 0.9676165016100614,
+      "grad_norm": 36.64017868041992,
+      "learning_rate": 2.610836757471924e-07,
+      "loss": 1.1122,
+      "step": 15926
+    },
+    {
+      "epoch": 0.9676772586426879,
+      "grad_norm": 0.6722599864006042,
+      "learning_rate": 2.601075977506362e-07,
+      "loss": 1.0483,
+      "step": 15927
+    },
+    {
+      "epoch": 0.9677380156753144,
+      "grad_norm": 0.3618413507938385,
+      "learning_rate": 2.5913334297945624e-07,
+      "loss": 1.1299,
+      "step": 15928
+    },
+    {
+      "epoch": 0.9677987727079409,
+      "grad_norm": 0.13960571587085724,
+      "learning_rate": 2.5816091146935176e-07,
+      "loss": 1.0873,
+      "step": 15929
+    },
+    {
+      "epoch": 0.9678595297405674,
+      "grad_norm": 0.15450404584407806,
+      "learning_rate": 2.5719030325597746e-07,
+      "loss": 1.0672,
+      "step": 15930
+    },
+    {
+      "epoch": 0.967920286773194,
+      "grad_norm": 0.10349221527576447,
+      "learning_rate": 2.5622151837490504e-07,
+      "loss": 1.0249,
+      "step": 15931
+    },
+    {
+      "epoch": 0.9679810438058205,
+      "grad_norm": 0.22325465083122253,
+      "learning_rate": 2.5525455686165043e-07,
+      "loss": 1.1754,
+      "step": 15932
+    },
+    {
+      "epoch": 0.968041800838447,
+      "grad_norm": 0.09836241602897644,
+      "learning_rate": 2.54289418751652e-07,
+      "loss": 1.0064,
+      "step": 15933
+    },
+    {
+      "epoch": 0.9681025578710736,
+      "grad_norm": 0.2870478928089142,
+      "learning_rate": 2.533261040802981e-07,
+      "loss": 1.3061,
+      "step": 15934
+    },
+    {
+      "epoch": 0.9681633149037001,
+      "grad_norm": 0.11183391511440277,
+      "learning_rate": 2.5236461288288274e-07,
+      "loss": 1.0314,
+      "step": 15935
+    },
+    {
+      "epoch": 0.9682240719363266,
+      "grad_norm": 0.10338148474693298,
+      "learning_rate": 2.514049451946665e-07,
+      "loss": 1.0182,
+      "step": 15936
+    },
+    {
+      "epoch": 0.9682848289689532,
+      "grad_norm": 0.2402750700712204,
+      "learning_rate": 2.5044710105081027e-07,
+      "loss": 1.1383,
+      "step": 15937
+    },
+    {
+      "epoch": 0.9683455860015797,
+      "grad_norm": 0.12980733811855316,
+      "learning_rate": 2.494910804864359e-07,
+      "loss": 1.0342,
+      "step": 15938
+    },
+    {
+      "epoch": 0.9684063430342063,
+      "grad_norm": 0.12378325313329697,
+      "learning_rate": 2.4853688353658753e-07,
+      "loss": 1.0294,
+      "step": 15939
+    },
+    {
+      "epoch": 0.9684671000668328,
+      "grad_norm": 0.26979950070381165,
+      "learning_rate": 2.475845102362262e-07,
+      "loss": 1.0667,
+      "step": 15940
+    },
+    {
+      "epoch": 0.9685278570994592,
+      "grad_norm": 0.2891567647457123,
+      "learning_rate": 2.466339606202794e-07,
+      "loss": 1.0716,
+      "step": 15941
+    },
+    {
+      "epoch": 0.9685886141320857,
+      "grad_norm": 0.12859764695167542,
+      "learning_rate": 2.4568523472358054e-07,
+      "loss": 1.0487,
+      "step": 15942
+    },
+    {
+      "epoch": 0.9686493711647123,
+      "grad_norm": 0.17575380206108093,
+      "learning_rate": 2.4473833258090715e-07,
+      "loss": 1.0588,
+      "step": 15943
+    },
+    {
+      "epoch": 0.9687101281973388,
+      "grad_norm": 0.1738256961107254,
+      "learning_rate": 2.4379325422696495e-07,
+      "loss": 1.1585,
+      "step": 15944
+    },
+    {
+      "epoch": 0.9687708852299654,
+      "grad_norm": 0.10076726227998734,
+      "learning_rate": 2.4284999969639844e-07,
+      "loss": 0.9953,
+      "step": 15945
+    },
+    {
+      "epoch": 0.9688316422625919,
+      "grad_norm": 0.13657309114933014,
+      "learning_rate": 2.419085690237799e-07,
+      "loss": 1.0956,
+      "step": 15946
+    },
+    {
+      "epoch": 0.9688923992952184,
+      "grad_norm": 0.12712503969669342,
+      "learning_rate": 2.4096896224362617e-07,
+      "loss": 0.9993,
+      "step": 15947
+    },
+    {
+      "epoch": 0.968953156327845,
+      "grad_norm": 0.16392002999782562,
+      "learning_rate": 2.4003117939037647e-07,
+      "loss": 1.1506,
+      "step": 15948
+    },
+    {
+      "epoch": 0.9690139133604715,
+      "grad_norm": 0.1706627756357193,
+      "learning_rate": 2.390952204983976e-07,
+      "loss": 1.1572,
+      "step": 15949
+    },
+    {
+      "epoch": 0.969074670393098,
+      "grad_norm": 0.15807820856571198,
+      "learning_rate": 2.3816108560200111e-07,
+      "loss": 1.0901,
+      "step": 15950
+    },
+    {
+      "epoch": 0.9691354274257246,
+      "grad_norm": 0.5251103639602661,
+      "learning_rate": 2.3722877473543182e-07,
+      "loss": 1.1488,
+      "step": 15951
+    },
+    {
+      "epoch": 0.9691961844583511,
+      "grad_norm": 0.3150418996810913,
+      "learning_rate": 2.362982879328568e-07,
+      "loss": 0.9907,
+      "step": 15952
+    },
+    {
+      "epoch": 0.9692569414909776,
+      "grad_norm": 5.522952079772949,
+      "learning_rate": 2.3536962522838768e-07,
+      "loss": 1.0944,
+      "step": 15953
+    },
+    {
+      "epoch": 0.969317698523604,
+      "grad_norm": 0.24826404452323914,
+      "learning_rate": 2.3444278665606944e-07,
+      "loss": 1.008,
+      "step": 15954
+    },
+    {
+      "epoch": 0.9693784555562306,
+      "grad_norm": 0.16055329144001007,
+      "learning_rate": 2.3351777224986938e-07,
+      "loss": 1.1287,
+      "step": 15955
+    },
+    {
+      "epoch": 0.9694392125888571,
+      "grad_norm": 0.12546609342098236,
+      "learning_rate": 2.3259458204368812e-07,
+      "loss": 1.0814,
+      "step": 15956
+    },
+    {
+      "epoch": 0.9694999696214837,
+      "grad_norm": 0.1838313192129135,
+      "learning_rate": 2.3167321607137637e-07,
+      "loss": 1.1996,
+      "step": 15957
+    },
+    {
+      "epoch": 0.9695607266541102,
+      "grad_norm": 0.11834122985601425,
+      "learning_rate": 2.3075367436670715e-07,
+      "loss": 1.0623,
+      "step": 15958
+    },
+    {
+      "epoch": 0.9696214836867367,
+      "grad_norm": 0.11417987942695618,
+      "learning_rate": 2.2983595696337567e-07,
+      "loss": 1.09,
+      "step": 15959
+    },
+    {
+      "epoch": 0.9696822407193633,
+      "grad_norm": 0.15419122576713562,
+      "learning_rate": 2.2892006389503285e-07,
+      "loss": 1.1104,
+      "step": 15960
+    },
+    {
+      "epoch": 0.9697429977519898,
+      "grad_norm": 0.10465330630540848,
+      "learning_rate": 2.280059951952407e-07,
+      "loss": 0.9956,
+      "step": 15961
+    },
+    {
+      "epoch": 0.9698037547846163,
+      "grad_norm": 0.21363714337348938,
+      "learning_rate": 2.2709375089751127e-07,
+      "loss": 1.1505,
+      "step": 15962
+    },
+    {
+      "epoch": 0.9698645118172429,
+      "grad_norm": 0.16582389175891876,
+      "learning_rate": 2.2618333103528455e-07,
+      "loss": 1.115,
+      "step": 15963
+    },
+    {
+      "epoch": 0.9699252688498694,
+      "grad_norm": 0.14295408129692078,
+      "learning_rate": 2.252747356419227e-07,
+      "loss": 1.0617,
+      "step": 15964
+    },
+    {
+      "epoch": 0.9699860258824959,
+      "grad_norm": 0.10888700187206268,
+      "learning_rate": 2.243679647507435e-07,
+      "loss": 1.0311,
+      "step": 15965
+    },
+    {
+      "epoch": 0.9700467829151225,
+      "grad_norm": 0.23514166474342346,
+      "learning_rate": 2.234630183949704e-07,
+      "loss": 1.098,
+      "step": 15966
+    },
+    {
+      "epoch": 0.9701075399477489,
+      "grad_norm": 0.10871703922748566,
+      "learning_rate": 2.2255989660778798e-07,
+      "loss": 1.0273,
+      "step": 15967
+    },
+    {
+      "epoch": 0.9701682969803754,
+      "grad_norm": 0.10922279953956604,
+      "learning_rate": 2.2165859942229194e-07,
+      "loss": 1.0241,
+      "step": 15968
+    },
+    {
+      "epoch": 0.970229054013002,
+      "grad_norm": 0.10100605338811874,
+      "learning_rate": 2.2075912687152255e-07,
+      "loss": 1.0537,
+      "step": 15969
+    },
+    {
+      "epoch": 0.9702898110456285,
+      "grad_norm": 0.10844356566667557,
+      "learning_rate": 2.1986147898844788e-07,
+      "loss": 1.0019,
+      "step": 15970
+    },
+    {
+      "epoch": 0.970350568078255,
+      "grad_norm": 0.15895801782608032,
+      "learning_rate": 2.1896565580597495e-07,
+      "loss": 1.0303,
+      "step": 15971
+    },
+    {
+      "epoch": 0.9704113251108816,
+      "grad_norm": 0.10335024446249008,
+      "learning_rate": 2.1807165735693857e-07,
+      "loss": 0.9887,
+      "step": 15972
+    },
+    {
+      "epoch": 0.9704720821435081,
+      "grad_norm": 0.11725698411464691,
+      "learning_rate": 2.17179483674107e-07,
+      "loss": 1.0343,
+      "step": 15973
+    },
+    {
+      "epoch": 0.9705328391761346,
+      "grad_norm": 1.839383602142334,
+      "learning_rate": 2.1628913479018186e-07,
+      "loss": 1.094,
+      "step": 15974
+    },
+    {
+      "epoch": 0.9705935962087612,
+      "grad_norm": 6.549253463745117,
+      "learning_rate": 2.1540061073780927e-07,
+      "loss": 1.0673,
+      "step": 15975
+    },
+    {
+      "epoch": 0.9706543532413877,
+      "grad_norm": 0.10236018896102905,
+      "learning_rate": 2.1451391154954093e-07,
+      "loss": 1.0173,
+      "step": 15976
+    },
+    {
+      "epoch": 0.9707151102740142,
+      "grad_norm": 0.1534181833267212,
+      "learning_rate": 2.136290372578953e-07,
+      "loss": 1.1281,
+      "step": 15977
+    },
+    {
+      "epoch": 0.9707758673066408,
+      "grad_norm": 0.20116965472698212,
+      "learning_rate": 2.1274598789529643e-07,
+      "loss": 1.1008,
+      "step": 15978
+    },
+    {
+      "epoch": 0.9708366243392673,
+      "grad_norm": 0.16923700273036957,
+      "learning_rate": 2.1186476349411843e-07,
+      "loss": 1.0648,
+      "step": 15979
+    },
+    {
+      "epoch": 0.9708973813718937,
+      "grad_norm": 0.10010872781276703,
+      "learning_rate": 2.1098536408666325e-07,
+      "loss": 1.0502,
+      "step": 15980
+    },
+    {
+      "epoch": 0.9709581384045203,
+      "grad_norm": 0.10427713394165039,
+      "learning_rate": 2.1010778970516064e-07,
+      "loss": 1.0487,
+      "step": 15981
+    },
+    {
+      "epoch": 0.9710188954371468,
+      "grad_norm": 0.09616519510746002,
+      "learning_rate": 2.092320403817849e-07,
+      "loss": 1.0517,
+      "step": 15982
+    },
+    {
+      "epoch": 0.9710796524697733,
+      "grad_norm": 0.18378973007202148,
+      "learning_rate": 2.0835811614863254e-07,
+      "loss": 1.0333,
+      "step": 15983
+    },
+    {
+      "epoch": 0.9711404095023999,
+      "grad_norm": 0.18472549319267273,
+      "learning_rate": 2.074860170377335e-07,
+      "loss": 1.0902,
+      "step": 15984
+    },
+    {
+      "epoch": 0.9712011665350264,
+      "grad_norm": 0.09981843829154968,
+      "learning_rate": 2.0661574308106224e-07,
+      "loss": 1.0282,
+      "step": 15985
+    },
+    {
+      "epoch": 0.971261923567653,
+      "grad_norm": 0.1477714329957962,
+      "learning_rate": 2.0574729431051543e-07,
+      "loss": 1.1393,
+      "step": 15986
+    },
+    {
+      "epoch": 0.9713226806002795,
+      "grad_norm": 0.2482462078332901,
+      "learning_rate": 2.0488067075792872e-07,
+      "loss": 1.1187,
+      "step": 15987
+    },
+    {
+      "epoch": 0.971383437632906,
+      "grad_norm": 0.10056059062480927,
+      "learning_rate": 2.040158724550656e-07,
+      "loss": 0.997,
+      "step": 15988
+    },
+    {
+      "epoch": 0.9714441946655326,
+      "grad_norm": 0.13931965827941895,
+      "learning_rate": 2.0315289943362292e-07,
+      "loss": 1.0828,
+      "step": 15989
+    },
+    {
+      "epoch": 0.9715049516981591,
+      "grad_norm": 0.13172860443592072,
+      "learning_rate": 2.022917517252365e-07,
+      "loss": 1.0808,
+      "step": 15990
+    },
+    {
+      "epoch": 0.9715657087307856,
+      "grad_norm": 0.16991360485553741,
+      "learning_rate": 2.0143242936147554e-07,
+      "loss": 1.1247,
+      "step": 15991
+    },
+    {
+      "epoch": 0.9716264657634122,
+      "grad_norm": 0.38997524976730347,
+      "learning_rate": 2.005749323738315e-07,
+      "loss": 1.1852,
+      "step": 15992
+    },
+    {
+      "epoch": 0.9716872227960387,
+      "grad_norm": 0.19578388333320618,
+      "learning_rate": 1.9971926079374037e-07,
+      "loss": 1.1603,
+      "step": 15993
+    },
+    {
+      "epoch": 0.9717479798286651,
+      "grad_norm": 0.2540781795978546,
+      "learning_rate": 1.9886541465256593e-07,
+      "loss": 1.0591,
+      "step": 15994
+    },
+    {
+      "epoch": 0.9718087368612917,
+      "grad_norm": 0.33513855934143066,
+      "learning_rate": 1.9801339398160534e-07,
+      "loss": 1.071,
+      "step": 15995
+    },
+    {
+      "epoch": 0.9718694938939182,
+      "grad_norm": 0.12066254019737244,
+      "learning_rate": 1.9716319881208922e-07,
+      "loss": 1.0337,
+      "step": 15996
+    },
+    {
+      "epoch": 0.9719302509265447,
+      "grad_norm": 0.11083171516656876,
+      "learning_rate": 1.9631482917518707e-07,
+      "loss": 1.0525,
+      "step": 15997
+    },
+    {
+      "epoch": 0.9719910079591713,
+      "grad_norm": 0.13840889930725098,
+      "learning_rate": 1.954682851019851e-07,
+      "loss": 1.0658,
+      "step": 15998
+    },
+    {
+      "epoch": 0.9720517649917978,
+      "grad_norm": 0.1476316601037979,
+      "learning_rate": 1.9462356662352522e-07,
+      "loss": 1.0947,
+      "step": 15999
+    },
+    {
+      "epoch": 0.9721125220244243,
+      "grad_norm": 1.4096200466156006,
+      "learning_rate": 1.9378067377076593e-07,
+      "loss": 1.2981,
+      "step": 16000
+    },
+    {
+      "epoch": 0.9721732790570509,
+      "grad_norm": 0.12837892770767212,
+      "learning_rate": 1.929396065745992e-07,
+      "loss": 1.0293,
+      "step": 16001
+    },
+    {
+      "epoch": 0.9722340360896774,
+      "grad_norm": 0.16172531247138977,
+      "learning_rate": 1.9210036506586148e-07,
+      "loss": 1.0312,
+      "step": 16002
+    },
+    {
+      "epoch": 0.9722947931223039,
+      "grad_norm": 0.15705092251300812,
+      "learning_rate": 1.9126294927531153e-07,
+      "loss": 1.0737,
+      "step": 16003
+    },
+    {
+      "epoch": 0.9723555501549305,
+      "grad_norm": 0.13786739110946655,
+      "learning_rate": 1.9042735923364696e-07,
+      "loss": 1.1181,
+      "step": 16004
+    },
+    {
+      "epoch": 0.972416307187557,
+      "grad_norm": 0.17052096128463745,
+      "learning_rate": 1.8959359497149886e-07,
+      "loss": 1.116,
+      "step": 16005
+    },
+    {
+      "epoch": 0.9724770642201835,
+      "grad_norm": 0.12219720333814621,
+      "learning_rate": 1.8876165651942056e-07,
+      "loss": 1.0122,
+      "step": 16006
+    },
+    {
+      "epoch": 0.97253782125281,
+      "grad_norm": 1.060339093208313,
+      "learning_rate": 1.8793154390791545e-07,
+      "loss": 1.0741,
+      "step": 16007
+    },
+    {
+      "epoch": 0.9725985782854365,
+      "grad_norm": 0.11641202121973038,
+      "learning_rate": 1.8710325716740363e-07,
+      "loss": 1.0312,
+      "step": 16008
+    },
+    {
+      "epoch": 0.972659335318063,
+      "grad_norm": 0.13792043924331665,
+      "learning_rate": 1.8627679632825523e-07,
+      "loss": 1.1228,
+      "step": 16009
+    },
+    {
+      "epoch": 0.9727200923506896,
+      "grad_norm": 0.1705060452222824,
+      "learning_rate": 1.8545216142075717e-07,
+      "loss": 1.1256,
+      "step": 16010
+    },
+    {
+      "epoch": 0.9727808493833161,
+      "grad_norm": 0.1110953837633133,
+      "learning_rate": 1.846293524751408e-07,
+      "loss": 1.0149,
+      "step": 16011
+    },
+    {
+      "epoch": 0.9728416064159426,
+      "grad_norm": 0.13366954028606415,
+      "learning_rate": 1.8380836952155979e-07,
+      "loss": 1.1379,
+      "step": 16012
+    },
+    {
+      "epoch": 0.9729023634485692,
+      "grad_norm": 0.22752730548381805,
+      "learning_rate": 1.8298921259011782e-07,
+      "loss": 1.1169,
+      "step": 16013
+    },
+    {
+      "epoch": 0.9729631204811957,
+      "grad_norm": 0.27061083912849426,
+      "learning_rate": 1.8217188171082977e-07,
+      "loss": 1.1635,
+      "step": 16014
+    },
+    {
+      "epoch": 0.9730238775138222,
+      "grad_norm": 0.7638154029846191,
+      "learning_rate": 1.8135637691366613e-07,
+      "loss": 1.073,
+      "step": 16015
+    },
+    {
+      "epoch": 0.9730846345464488,
+      "grad_norm": 0.1051676943898201,
+      "learning_rate": 1.805426982285141e-07,
+      "loss": 1.0303,
+      "step": 16016
+    },
+    {
+      "epoch": 0.9731453915790753,
+      "grad_norm": 1.0993504524230957,
+      "learning_rate": 1.7973084568519982e-07,
+      "loss": 1.0117,
+      "step": 16017
+    },
+    {
+      "epoch": 0.9732061486117018,
+      "grad_norm": 0.1313251554965973,
+      "learning_rate": 1.7892081931347726e-07,
+      "loss": 1.0436,
+      "step": 16018
+    },
+    {
+      "epoch": 0.9732669056443284,
+      "grad_norm": 1.02340829372406,
+      "learning_rate": 1.7811261914304488e-07,
+      "loss": 1.0913,
+      "step": 16019
+    },
+    {
+      "epoch": 0.9733276626769548,
+      "grad_norm": 0.12242753058671951,
+      "learning_rate": 1.7730624520352346e-07,
+      "loss": 1.0217,
+      "step": 16020
+    },
+    {
+      "epoch": 0.9733884197095813,
+      "grad_norm": 0.09510639309883118,
+      "learning_rate": 1.7650169752447265e-07,
+      "loss": 0.992,
+      "step": 16021
+    },
+    {
+      "epoch": 0.9734491767422079,
+      "grad_norm": 0.12729017436504364,
+      "learning_rate": 1.7569897613538555e-07,
+      "loss": 1.0467,
+      "step": 16022
+    },
+    {
+      "epoch": 0.9735099337748344,
+      "grad_norm": 0.16152387857437134,
+      "learning_rate": 1.7489808106568305e-07,
+      "loss": 1.0752,
+      "step": 16023
+    },
+    {
+      "epoch": 0.973570690807461,
+      "grad_norm": 0.10730086266994476,
+      "learning_rate": 1.7409901234471948e-07,
+      "loss": 1.0293,
+      "step": 16024
+    },
+    {
+      "epoch": 0.9736314478400875,
+      "grad_norm": 1.1608229875564575,
+      "learning_rate": 1.7330177000178805e-07,
+      "loss": 1.1188,
+      "step": 16025
+    },
+    {
+      "epoch": 0.973692204872714,
+      "grad_norm": 0.12683729827404022,
+      "learning_rate": 1.7250635406611538e-07,
+      "loss": 0.9897,
+      "step": 16026
+    },
+    {
+      "epoch": 0.9737529619053406,
+      "grad_norm": 0.13331612944602966,
+      "learning_rate": 1.7171276456685037e-07,
+      "loss": 1.0664,
+      "step": 16027
+    },
+    {
+      "epoch": 0.9738137189379671,
+      "grad_norm": 3.2299716472625732,
+      "learning_rate": 1.7092100153308643e-07,
+      "loss": 1.044,
+      "step": 16028
+    },
+    {
+      "epoch": 0.9738744759705936,
+      "grad_norm": 0.11321816593408585,
+      "learning_rate": 1.7013106499383924e-07,
+      "loss": 1.0706,
+      "step": 16029
+    },
+    {
+      "epoch": 0.9739352330032202,
+      "grad_norm": 0.11880458146333694,
+      "learning_rate": 1.6934295497808006e-07,
+      "loss": 1.0139,
+      "step": 16030
+    },
+    {
+      "epoch": 0.9739959900358467,
+      "grad_norm": 0.2528494596481323,
+      "learning_rate": 1.6855667151468024e-07,
+      "loss": 1.1883,
+      "step": 16031
+    },
+    {
+      "epoch": 0.9740567470684732,
+      "grad_norm": 0.10102248936891556,
+      "learning_rate": 1.6777221463246673e-07,
+      "loss": 0.9729,
+      "step": 16032
+    },
+    {
+      "epoch": 0.9741175041010997,
+      "grad_norm": 0.1477328985929489,
+      "learning_rate": 1.6698958436019985e-07,
+      "loss": 1.1111,
+      "step": 16033
+    },
+    {
+      "epoch": 0.9741782611337262,
+      "grad_norm": 0.15187117457389832,
+      "learning_rate": 1.6620878072655664e-07,
+      "loss": 1.0719,
+      "step": 16034
+    },
+    {
+      "epoch": 0.9742390181663527,
+      "grad_norm": 0.25257760286331177,
+      "learning_rate": 1.6542980376016982e-07,
+      "loss": 1.0492,
+      "step": 16035
+    },
+    {
+      "epoch": 0.9742997751989793,
+      "grad_norm": 0.14116472005844116,
+      "learning_rate": 1.6465265348958314e-07,
+      "loss": 1.0791,
+      "step": 16036
+    },
+    {
+      "epoch": 0.9743605322316058,
+      "grad_norm": 0.22963206470012665,
+      "learning_rate": 1.6387732994328496e-07,
+      "loss": 1.0766,
+      "step": 16037
+    },
+    {
+      "epoch": 0.9744212892642323,
+      "grad_norm": 0.15372848510742188,
+      "learning_rate": 1.63103833149697e-07,
+      "loss": 1.0821,
+      "step": 16038
+    },
+    {
+      "epoch": 0.9744820462968589,
+      "grad_norm": 0.1959514319896698,
+      "learning_rate": 1.6233216313717437e-07,
+      "loss": 1.0473,
+      "step": 16039
+    },
+    {
+      "epoch": 0.9745428033294854,
+      "grad_norm": 0.15264278650283813,
+      "learning_rate": 1.615623199339944e-07,
+      "loss": 1.1159,
+      "step": 16040
+    },
+    {
+      "epoch": 0.9746035603621119,
+      "grad_norm": 0.09602281451225281,
+      "learning_rate": 1.6079430356838454e-07,
+      "loss": 1.0424,
+      "step": 16041
+    },
+    {
+      "epoch": 0.9746643173947385,
+      "grad_norm": 0.0902513787150383,
+      "learning_rate": 1.600281140684945e-07,
+      "loss": 1.0112,
+      "step": 16042
+    },
+    {
+      "epoch": 0.974725074427365,
+      "grad_norm": 0.1508355736732483,
+      "learning_rate": 1.5926375146240734e-07,
+      "loss": 1.1134,
+      "step": 16043
+    },
+    {
+      "epoch": 0.9747858314599915,
+      "grad_norm": 0.13014133274555206,
+      "learning_rate": 1.5850121577813959e-07,
+      "loss": 1.0306,
+      "step": 16044
+    },
+    {
+      "epoch": 0.9748465884926181,
+      "grad_norm": 0.12807750701904297,
+      "learning_rate": 1.5774050704364106e-07,
+      "loss": 1.0856,
+      "step": 16045
+    },
+    {
+      "epoch": 0.9749073455252445,
+      "grad_norm": 0.17278870940208435,
+      "learning_rate": 1.5698162528680616e-07,
+      "loss": 1.1414,
+      "step": 16046
+    },
+    {
+      "epoch": 0.974968102557871,
+      "grad_norm": 0.18267321586608887,
+      "learning_rate": 1.5622457053544038e-07,
+      "loss": 1.0351,
+      "step": 16047
+    },
+    {
+      "epoch": 0.9750288595904976,
+      "grad_norm": 0.20184537768363953,
+      "learning_rate": 1.5546934281729375e-07,
+      "loss": 1.2628,
+      "step": 16048
+    },
+    {
+      "epoch": 0.9750896166231241,
+      "grad_norm": 0.34459781646728516,
+      "learning_rate": 1.547159421600608e-07,
+      "loss": 1.2446,
+      "step": 16049
+    },
+    {
+      "epoch": 0.9751503736557506,
+      "grad_norm": 0.27078482508659363,
+      "learning_rate": 1.5396436859134168e-07,
+      "loss": 1.0954,
+      "step": 16050
+    },
+    {
+      "epoch": 0.9752111306883772,
+      "grad_norm": 0.11293277144432068,
+      "learning_rate": 1.5321462213869764e-07,
+      "loss": 1.0631,
+      "step": 16051
+    },
+    {
+      "epoch": 0.9752718877210037,
+      "grad_norm": 0.2270057201385498,
+      "learning_rate": 1.5246670282960674e-07,
+      "loss": 1.1245,
+      "step": 16052
+    },
+    {
+      "epoch": 0.9753326447536302,
+      "grad_norm": 0.20857922732830048,
+      "learning_rate": 1.517206106914859e-07,
+      "loss": 1.1282,
+      "step": 16053
+    },
+    {
+      "epoch": 0.9753934017862568,
+      "grad_norm": 0.3938729166984558,
+      "learning_rate": 1.5097634575167996e-07,
+      "loss": 1.196,
+      "step": 16054
+    },
+    {
+      "epoch": 0.9754541588188833,
+      "grad_norm": 0.12796056270599365,
+      "learning_rate": 1.502339080374726e-07,
+      "loss": 1.032,
+      "step": 16055
+    },
+    {
+      "epoch": 0.9755149158515098,
+      "grad_norm": 0.1373816430568695,
+      "learning_rate": 1.4949329757607544e-07,
+      "loss": 1.0655,
+      "step": 16056
+    },
+    {
+      "epoch": 0.9755756728841364,
+      "grad_norm": 0.3361382484436035,
+      "learning_rate": 1.4875451439463894e-07,
+      "loss": 1.1071,
+      "step": 16057
+    },
+    {
+      "epoch": 0.9756364299167629,
+      "grad_norm": 0.14629597961902618,
+      "learning_rate": 1.4801755852024145e-07,
+      "loss": 1.102,
+      "step": 16058
+    },
+    {
+      "epoch": 0.9756971869493893,
+      "grad_norm": 0.11022738367319107,
+      "learning_rate": 1.4728242997989473e-07,
+      "loss": 1.1013,
+      "step": 16059
+    },
+    {
+      "epoch": 0.9757579439820159,
+      "grad_norm": 0.43559399247169495,
+      "learning_rate": 1.4654912880054382e-07,
+      "loss": 1.1323,
+      "step": 16060
+    },
+    {
+      "epoch": 0.9758187010146424,
+      "grad_norm": 0.12420898675918579,
+      "learning_rate": 1.458176550090784e-07,
+      "loss": 1.0982,
+      "step": 16061
+    },
+    {
+      "epoch": 0.9758794580472689,
+      "grad_norm": 0.340138703584671,
+      "learning_rate": 1.450880086322992e-07,
+      "loss": 1.1505,
+      "step": 16062
+    },
+    {
+      "epoch": 0.9759402150798955,
+      "grad_norm": 0.19947384297847748,
+      "learning_rate": 1.4436018969695708e-07,
+      "loss": 1.0682,
+      "step": 16063
+    },
+    {
+      "epoch": 0.976000972112522,
+      "grad_norm": 0.17929403483867645,
+      "learning_rate": 1.4363419822973067e-07,
+      "loss": 1.0751,
+      "step": 16064
+    },
+    {
+      "epoch": 0.9760617291451485,
+      "grad_norm": 0.1395479142665863,
+      "learning_rate": 1.4291003425722648e-07,
+      "loss": 1.1095,
+      "step": 16065
+    },
+    {
+      "epoch": 0.9761224861777751,
+      "grad_norm": 0.14332261681556702,
+      "learning_rate": 1.4218769780598995e-07,
+      "loss": 1.1459,
+      "step": 16066
+    },
+    {
+      "epoch": 0.9761832432104016,
+      "grad_norm": 0.10867652297019958,
+      "learning_rate": 1.4146718890250543e-07,
+      "loss": 1.0158,
+      "step": 16067
+    },
+    {
+      "epoch": 0.9762440002430282,
+      "grad_norm": 0.1907731145620346,
+      "learning_rate": 1.407485075731796e-07,
+      "loss": 1.2309,
+      "step": 16068
+    },
+    {
+      "epoch": 0.9763047572756547,
+      "grad_norm": 0.143650621175766,
+      "learning_rate": 1.4003165384435246e-07,
+      "loss": 1.008,
+      "step": 16069
+    },
+    {
+      "epoch": 0.9763655143082812,
+      "grad_norm": 0.14735780656337738,
+      "learning_rate": 1.3931662774230304e-07,
+      "loss": 1.0731,
+      "step": 16070
+    },
+    {
+      "epoch": 0.9764262713409078,
+      "grad_norm": 0.1586189568042755,
+      "learning_rate": 1.386034292932381e-07,
+      "loss": 1.1754,
+      "step": 16071
+    },
+    {
+      "epoch": 0.9764870283735342,
+      "grad_norm": 0.13290809094905853,
+      "learning_rate": 1.37892058523309e-07,
+      "loss": 1.0146,
+      "step": 16072
+    },
+    {
+      "epoch": 0.9765477854061607,
+      "grad_norm": 0.47854822874069214,
+      "learning_rate": 1.371825154585782e-07,
+      "loss": 1.1807,
+      "step": 16073
+    },
+    {
+      "epoch": 0.9766085424387873,
+      "grad_norm": 0.11864987015724182,
+      "learning_rate": 1.3647480012506375e-07,
+      "loss": 1.0684,
+      "step": 16074
+    },
+    {
+      "epoch": 0.9766692994714138,
+      "grad_norm": 0.15853135287761688,
+      "learning_rate": 1.357689125487005e-07,
+      "loss": 1.069,
+      "step": 16075
+    },
+    {
+      "epoch": 0.9767300565040403,
+      "grad_norm": 2.330995798110962,
+      "learning_rate": 1.3506485275536772e-07,
+      "loss": 1.0224,
+      "step": 16076
+    },
+    {
+      "epoch": 0.9767908135366669,
+      "grad_norm": 0.14817878603935242,
+      "learning_rate": 1.3436262077087258e-07,
+      "loss": 1.2449,
+      "step": 16077
+    },
+    {
+      "epoch": 0.9768515705692934,
+      "grad_norm": 0.14410121738910675,
+      "learning_rate": 1.3366221662095558e-07,
+      "loss": 1.0676,
+      "step": 16078
+    },
+    {
+      "epoch": 0.9769123276019199,
+      "grad_norm": 0.15412510931491852,
+      "learning_rate": 1.3296364033128506e-07,
+      "loss": 1.1063,
+      "step": 16079
+    },
+    {
+      "epoch": 0.9769730846345465,
+      "grad_norm": 0.2766765356063843,
+      "learning_rate": 1.3226689192747387e-07,
+      "loss": 1.1065,
+      "step": 16080
+    },
+    {
+      "epoch": 0.977033841667173,
+      "grad_norm": 0.19024202227592468,
+      "learning_rate": 1.3157197143506273e-07,
+      "loss": 1.0934,
+      "step": 16081
+    },
+    {
+      "epoch": 0.9770945986997995,
+      "grad_norm": 0.10379128903150558,
+      "learning_rate": 1.3087887887951457e-07,
+      "loss": 1.0493,
+      "step": 16082
+    },
+    {
+      "epoch": 0.9771553557324261,
+      "grad_norm": 0.4700990319252014,
+      "learning_rate": 1.3018761428624237e-07,
+      "loss": 0.9965,
+      "step": 16083
+    },
+    {
+      "epoch": 0.9772161127650526,
+      "grad_norm": 0.1554998755455017,
+      "learning_rate": 1.294981776805815e-07,
+      "loss": 1.0842,
+      "step": 16084
+    },
+    {
+      "epoch": 0.977276869797679,
+      "grad_norm": 0.10045026987791061,
+      "learning_rate": 1.2881056908780608e-07,
+      "loss": 1.0269,
+      "step": 16085
+    },
+    {
+      "epoch": 0.9773376268303056,
+      "grad_norm": 0.302529901266098,
+      "learning_rate": 1.2812478853311826e-07,
+      "loss": 1.1391,
+      "step": 16086
+    },
+    {
+      "epoch": 0.9773983838629321,
+      "grad_norm": 0.24632851779460907,
+      "learning_rate": 1.27440836041659e-07,
+      "loss": 1.0796,
+      "step": 16087
+    },
+    {
+      "epoch": 0.9774591408955586,
+      "grad_norm": 0.10734419524669647,
+      "learning_rate": 1.2675871163849718e-07,
+      "loss": 1.0819,
+      "step": 16088
+    },
+    {
+      "epoch": 0.9775198979281852,
+      "grad_norm": 0.31728655099868774,
+      "learning_rate": 1.260784153486294e-07,
+      "loss": 1.2579,
+      "step": 16089
+    },
+    {
+      "epoch": 0.9775806549608117,
+      "grad_norm": 0.48442772030830383,
+      "learning_rate": 1.2539994719700244e-07,
+      "loss": 1.0493,
+      "step": 16090
+    },
+    {
+      "epoch": 0.9776414119934382,
+      "grad_norm": 0.1007699966430664,
+      "learning_rate": 1.247233072084797e-07,
+      "loss": 0.987,
+      "step": 16091
+    },
+    {
+      "epoch": 0.9777021690260648,
+      "grad_norm": 0.29470014572143555,
+      "learning_rate": 1.2404849540786357e-07,
+      "loss": 1.087,
+      "step": 16092
+    },
+    {
+      "epoch": 0.9777629260586913,
+      "grad_norm": 0.12711603939533234,
+      "learning_rate": 1.2337551181988983e-07,
+      "loss": 1.0659,
+      "step": 16093
+    },
+    {
+      "epoch": 0.9778236830913178,
+      "grad_norm": 0.27971151471138,
+      "learning_rate": 1.2270435646922762e-07,
+      "loss": 1.2239,
+      "step": 16094
+    },
+    {
+      "epoch": 0.9778844401239444,
+      "grad_norm": 0.11354578286409378,
+      "learning_rate": 1.2203502938047952e-07,
+      "loss": 1.0161,
+      "step": 16095
+    },
+    {
+      "epoch": 0.9779451971565709,
+      "grad_norm": 0.2748790383338928,
+      "learning_rate": 1.213675305781814e-07,
+      "loss": 1.0479,
+      "step": 16096
+    },
+    {
+      "epoch": 0.9780059541891974,
+      "grad_norm": 0.09628867357969284,
+      "learning_rate": 1.2070186008679707e-07,
+      "loss": 1.0198,
+      "step": 16097
+    },
+    {
+      "epoch": 0.978066711221824,
+      "grad_norm": 0.16216102242469788,
+      "learning_rate": 1.2003801793072366e-07,
+      "loss": 1.095,
+      "step": 16098
+    },
+    {
+      "epoch": 0.9781274682544504,
+      "grad_norm": 0.19075436890125275,
+      "learning_rate": 1.1937600413430283e-07,
+      "loss": 1.0632,
+      "step": 16099
+    },
+    {
+      "epoch": 0.9781882252870769,
+      "grad_norm": 0.22025249898433685,
+      "learning_rate": 1.1871581872179294e-07,
+      "loss": 1.0657,
+      "step": 16100
+    },
+    {
+      "epoch": 0.9782489823197035,
+      "grad_norm": 0.28478437662124634,
+      "learning_rate": 1.1805746171739685e-07,
+      "loss": 1.0728,
+      "step": 16101
+    },
+    {
+      "epoch": 0.97830973935233,
+      "grad_norm": 0.09931562840938568,
+      "learning_rate": 1.1740093314524525e-07,
+      "loss": 1.063,
+      "step": 16102
+    },
+    {
+      "epoch": 0.9783704963849565,
+      "grad_norm": 0.14065153896808624,
+      "learning_rate": 1.167462330294078e-07,
+      "loss": 1.092,
+      "step": 16103
+    },
+    {
+      "epoch": 0.9784312534175831,
+      "grad_norm": 0.24711716175079346,
+      "learning_rate": 1.1609336139387639e-07,
+      "loss": 1.1335,
+      "step": 16104
+    },
+    {
+      "epoch": 0.9784920104502096,
+      "grad_norm": 0.1082901656627655,
+      "learning_rate": 1.1544231826258745e-07,
+      "loss": 1.0165,
+      "step": 16105
+    },
+    {
+      "epoch": 0.9785527674828361,
+      "grad_norm": 0.16817279160022736,
+      "learning_rate": 1.1479310365939966e-07,
+      "loss": 1.1308,
+      "step": 16106
+    },
+    {
+      "epoch": 0.9786135245154627,
+      "grad_norm": 0.13043826818466187,
+      "learning_rate": 1.1414571760811621e-07,
+      "loss": 1.096,
+      "step": 16107
+    },
+    {
+      "epoch": 0.9786742815480892,
+      "grad_norm": 0.14298072457313538,
+      "learning_rate": 1.1350016013246256e-07,
+      "loss": 1.0753,
+      "step": 16108
+    },
+    {
+      "epoch": 0.9787350385807158,
+      "grad_norm": 0.2213154435157776,
+      "learning_rate": 1.1285643125610867e-07,
+      "loss": 1.1939,
+      "step": 16109
+    },
+    {
+      "epoch": 0.9787957956133423,
+      "grad_norm": 0.1052158921957016,
+      "learning_rate": 1.1221453100264123e-07,
+      "loss": 1.0351,
+      "step": 16110
+    },
+    {
+      "epoch": 0.9788565526459688,
+      "grad_norm": 0.1351722925901413,
+      "learning_rate": 1.1157445939559141e-07,
+      "loss": 1.1005,
+      "step": 16111
+    },
+    {
+      "epoch": 0.9789173096785952,
+      "grad_norm": 0.12105369567871094,
+      "learning_rate": 1.1093621645842378e-07,
+      "loss": 1.0058,
+      "step": 16112
+    },
+    {
+      "epoch": 0.9789780667112218,
+      "grad_norm": 0.11711379885673523,
+      "learning_rate": 1.1029980221453073e-07,
+      "loss": 1.0504,
+      "step": 16113
+    },
+    {
+      "epoch": 0.9790388237438483,
+      "grad_norm": 0.12315700203180313,
+      "learning_rate": 1.0966521668724361e-07,
+      "loss": 1.0017,
+      "step": 16114
+    },
+    {
+      "epoch": 0.9790995807764749,
+      "grad_norm": 0.10018137097358704,
+      "learning_rate": 1.0903245989982714e-07,
+      "loss": 1.024,
+      "step": 16115
+    },
+    {
+      "epoch": 0.9791603378091014,
+      "grad_norm": 0.3580493628978729,
+      "learning_rate": 1.0840153187546276e-07,
+      "loss": 1.1111,
+      "step": 16116
+    },
+    {
+      "epoch": 0.9792210948417279,
+      "grad_norm": 0.11206480860710144,
+      "learning_rate": 1.0777243263728754e-07,
+      "loss": 1.0214,
+      "step": 16117
+    },
+    {
+      "epoch": 0.9792818518743545,
+      "grad_norm": 0.137099027633667,
+      "learning_rate": 1.0714516220835525e-07,
+      "loss": 1.0701,
+      "step": 16118
+    },
+    {
+      "epoch": 0.979342608906981,
+      "grad_norm": 0.187575101852417,
+      "learning_rate": 1.0651972061166415e-07,
+      "loss": 1.1005,
+      "step": 16119
+    },
+    {
+      "epoch": 0.9794033659396075,
+      "grad_norm": 0.16693246364593506,
+      "learning_rate": 1.0589610787013482e-07,
+      "loss": 1.1508,
+      "step": 16120
+    },
+    {
+      "epoch": 0.9794641229722341,
+      "grad_norm": 0.14592605829238892,
+      "learning_rate": 1.0527432400662674e-07,
+      "loss": 1.0543,
+      "step": 16121
+    },
+    {
+      "epoch": 0.9795248800048606,
+      "grad_norm": 0.1340041160583496,
+      "learning_rate": 1.0465436904393277e-07,
+      "loss": 1.0305,
+      "step": 16122
+    },
+    {
+      "epoch": 0.9795856370374871,
+      "grad_norm": 0.25268298387527466,
+      "learning_rate": 1.0403624300477921e-07,
+      "loss": 1.1417,
+      "step": 16123
+    },
+    {
+      "epoch": 0.9796463940701137,
+      "grad_norm": 0.16122040152549744,
+      "learning_rate": 1.0341994591182014e-07,
+      "loss": 1.0411,
+      "step": 16124
+    },
+    {
+      "epoch": 0.9797071511027401,
+      "grad_norm": 0.2182396501302719,
+      "learning_rate": 1.028054777876486e-07,
+      "loss": 1.0628,
+      "step": 16125
+    },
+    {
+      "epoch": 0.9797679081353666,
+      "grad_norm": 0.09938058257102966,
+      "learning_rate": 1.0219283865479101e-07,
+      "loss": 1.055,
+      "step": 16126
+    },
+    {
+      "epoch": 0.9798286651679932,
+      "grad_norm": 0.0970664769411087,
+      "learning_rate": 1.015820285356961e-07,
+      "loss": 1.0117,
+      "step": 16127
+    },
+    {
+      "epoch": 0.9798894222006197,
+      "grad_norm": 0.24902427196502686,
+      "learning_rate": 1.0097304745275704e-07,
+      "loss": 1.1219,
+      "step": 16128
+    },
+    {
+      "epoch": 0.9799501792332462,
+      "grad_norm": 0.1190854012966156,
+      "learning_rate": 1.0036589542829488e-07,
+      "loss": 1.0786,
+      "step": 16129
+    },
+    {
+      "epoch": 0.9800109362658728,
+      "grad_norm": 0.09834282100200653,
+      "learning_rate": 9.97605724845696e-08,
+      "loss": 1.0342,
+      "step": 16130
+    },
+    {
+      "epoch": 0.9800716932984993,
+      "grad_norm": 0.15337973833084106,
+      "learning_rate": 9.915707864376345e-08,
+      "loss": 1.1267,
+      "step": 16131
+    },
+    {
+      "epoch": 0.9801324503311258,
+      "grad_norm": 0.10228081047534943,
+      "learning_rate": 9.855541392800316e-08,
+      "loss": 1.0664,
+      "step": 16132
+    },
+    {
+      "epoch": 0.9801932073637524,
+      "grad_norm": 0.11617690324783325,
+      "learning_rate": 9.795557835933778e-08,
+      "loss": 1.0676,
+      "step": 16133
+    },
+    {
+      "epoch": 0.9802539643963789,
+      "grad_norm": 0.10063482820987701,
+      "learning_rate": 9.735757195975525e-08,
+      "loss": 1.0431,
+      "step": 16134
+    },
+    {
+      "epoch": 0.9803147214290054,
+      "grad_norm": 0.10332181304693222,
+      "learning_rate": 9.676139475117696e-08,
+      "loss": 1.0398,
+      "step": 16135
+    },
+    {
+      "epoch": 0.980375478461632,
+      "grad_norm": 0.3813704550266266,
+      "learning_rate": 9.616704675545762e-08,
+      "loss": 1.0076,
+      "step": 16136
+    },
+    {
+      "epoch": 0.9804362354942585,
+      "grad_norm": 0.19592706859111786,
+      "learning_rate": 9.557452799437982e-08,
+      "loss": 1.113,
+      "step": 16137
+    },
+    {
+      "epoch": 0.9804969925268849,
+      "grad_norm": 0.11222069710493088,
+      "learning_rate": 9.498383848966508e-08,
+      "loss": 1.013,
+      "step": 16138
+    },
+    {
+      "epoch": 0.9805577495595115,
+      "grad_norm": 0.22417768836021423,
+      "learning_rate": 9.439497826296828e-08,
+      "loss": 1.0343,
+      "step": 16139
+    },
+    {
+      "epoch": 0.980618506592138,
+      "grad_norm": 0.12241493910551071,
+      "learning_rate": 9.380794733586107e-08,
+      "loss": 1.0776,
+      "step": 16140
+    },
+    {
+      "epoch": 0.9806792636247645,
+      "grad_norm": 0.14164398610591888,
+      "learning_rate": 9.322274572987621e-08,
+      "loss": 1.0949,
+      "step": 16141
+    },
+    {
+      "epoch": 0.9807400206573911,
+      "grad_norm": 0.11311923712491989,
+      "learning_rate": 9.263937346645769e-08,
+      "loss": 1.0523,
+      "step": 16142
+    },
+    {
+      "epoch": 0.9808007776900176,
+      "grad_norm": 0.15364493429660797,
+      "learning_rate": 9.205783056698835e-08,
+      "loss": 1.0641,
+      "step": 16143
+    },
+    {
+      "epoch": 0.9808615347226441,
+      "grad_norm": 0.17510327696800232,
+      "learning_rate": 9.147811705279008e-08,
+      "loss": 1.1533,
+      "step": 16144
+    },
+    {
+      "epoch": 0.9809222917552707,
+      "grad_norm": 0.17928443849086761,
+      "learning_rate": 9.090023294510697e-08,
+      "loss": 1.129,
+      "step": 16145
+    },
+    {
+      "epoch": 0.9809830487878972,
+      "grad_norm": 0.13306649029254913,
+      "learning_rate": 9.032417826512207e-08,
+      "loss": 1.0595,
+      "step": 16146
+    },
+    {
+      "epoch": 0.9810438058205238,
+      "grad_norm": 0.09092667698860168,
+      "learning_rate": 8.974995303395184e-08,
+      "loss": 1.0479,
+      "step": 16147
+    },
+    {
+      "epoch": 0.9811045628531503,
+      "grad_norm": 0.11474576592445374,
+      "learning_rate": 8.917755727264054e-08,
+      "loss": 1.0573,
+      "step": 16148
+    },
+    {
+      "epoch": 0.9811653198857768,
+      "grad_norm": 0.12943261861801147,
+      "learning_rate": 8.860699100217696e-08,
+      "loss": 1.028,
+      "step": 16149
+    },
+    {
+      "epoch": 0.9812260769184034,
+      "grad_norm": 0.10790281742811203,
+      "learning_rate": 8.803825424347212e-08,
+      "loss": 0.9781,
+      "step": 16150
+    },
+    {
+      "epoch": 0.9812868339510298,
+      "grad_norm": 0.1185181587934494,
+      "learning_rate": 8.747134701737602e-08,
+      "loss": 1.0507,
+      "step": 16151
+    },
+    {
+      "epoch": 0.9813475909836563,
+      "grad_norm": 0.10440313816070557,
+      "learning_rate": 8.690626934466096e-08,
+      "loss": 1.0813,
+      "step": 16152
+    },
+    {
+      "epoch": 0.9814083480162828,
+      "grad_norm": 0.11123882979154587,
+      "learning_rate": 8.634302124604366e-08,
+      "loss": 1.0761,
+      "step": 16153
+    },
+    {
+      "epoch": 0.9814691050489094,
+      "grad_norm": 1.1919429302215576,
+      "learning_rate": 8.578160274217428e-08,
+      "loss": 1.0604,
+      "step": 16154
+    },
+    {
+      "epoch": 0.9815298620815359,
+      "grad_norm": 0.1616741269826889,
+      "learning_rate": 8.522201385362527e-08,
+      "loss": 1.1023,
+      "step": 16155
+    },
+    {
+      "epoch": 0.9815906191141625,
+      "grad_norm": 0.2502225339412689,
+      "learning_rate": 8.466425460091909e-08,
+      "loss": 1.0837,
+      "step": 16156
+    },
+    {
+      "epoch": 0.981651376146789,
+      "grad_norm": 0.10444097965955734,
+      "learning_rate": 8.410832500448385e-08,
+      "loss": 1.004,
+      "step": 16157
+    },
+    {
+      "epoch": 0.9817121331794155,
+      "grad_norm": 0.15894034504890442,
+      "learning_rate": 8.355422508470878e-08,
+      "loss": 1.0874,
+      "step": 16158
+    },
+    {
+      "epoch": 0.9817728902120421,
+      "grad_norm": 0.14622347056865692,
+      "learning_rate": 8.300195486190543e-08,
+      "loss": 1.0506,
+      "step": 16159
+    },
+    {
+      "epoch": 0.9818336472446686,
+      "grad_norm": 0.34307006001472473,
+      "learning_rate": 8.245151435631316e-08,
+      "loss": 1.0173,
+      "step": 16160
+    },
+    {
+      "epoch": 0.9818944042772951,
+      "grad_norm": 0.10111016035079956,
+      "learning_rate": 8.190290358811026e-08,
+      "loss": 0.9942,
+      "step": 16161
+    },
+    {
+      "epoch": 0.9819551613099217,
+      "grad_norm": 0.194380521774292,
+      "learning_rate": 8.135612257740289e-08,
+      "loss": 1.0556,
+      "step": 16162
+    },
+    {
+      "epoch": 0.9820159183425482,
+      "grad_norm": 0.5943927764892578,
+      "learning_rate": 8.08111713442361e-08,
+      "loss": 1.0523,
+      "step": 16163
+    },
+    {
+      "epoch": 0.9820766753751746,
+      "grad_norm": 0.11505883932113647,
+      "learning_rate": 8.026804990858283e-08,
+      "loss": 1.0105,
+      "step": 16164
+    },
+    {
+      "epoch": 0.9821374324078012,
+      "grad_norm": 0.08244690299034119,
+      "learning_rate": 7.972675829036046e-08,
+      "loss": 0.9533,
+      "step": 16165
+    },
+    {
+      "epoch": 0.9821981894404277,
+      "grad_norm": 0.1624736487865448,
+      "learning_rate": 7.91872965093976e-08,
+      "loss": 1.1099,
+      "step": 16166
+    },
+    {
+      "epoch": 0.9822589464730542,
+      "grad_norm": 0.1681128591299057,
+      "learning_rate": 7.86496645854784e-08,
+      "loss": 1.0665,
+      "step": 16167
+    },
+    {
+      "epoch": 0.9823197035056808,
+      "grad_norm": 0.13631050288677216,
+      "learning_rate": 7.811386253829822e-08,
+      "loss": 1.0573,
+      "step": 16168
+    },
+    {
+      "epoch": 0.9823804605383073,
+      "grad_norm": 1.784241795539856,
+      "learning_rate": 7.757989038750801e-08,
+      "loss": 1.1413,
+      "step": 16169
+    },
+    {
+      "epoch": 0.9824412175709338,
+      "grad_norm": 0.14992661774158478,
+      "learning_rate": 7.704774815268101e-08,
+      "loss": 1.1459,
+      "step": 16170
+    },
+    {
+      "epoch": 0.9825019746035604,
+      "grad_norm": 0.20285074412822723,
+      "learning_rate": 7.651743585331273e-08,
+      "loss": 1.0521,
+      "step": 16171
+    },
+    {
+      "epoch": 0.9825627316361869,
+      "grad_norm": 0.19562220573425293,
+      "learning_rate": 7.598895350884871e-08,
+      "loss": 1.0629,
+      "step": 16172
+    },
+    {
+      "epoch": 0.9826234886688134,
+      "grad_norm": 4.75029993057251,
+      "learning_rate": 7.546230113866237e-08,
+      "loss": 1.1577,
+      "step": 16173
+    },
+    {
+      "epoch": 0.98268424570144,
+      "grad_norm": 0.15583325922489166,
+      "learning_rate": 7.493747876204937e-08,
+      "loss": 1.1043,
+      "step": 16174
+    },
+    {
+      "epoch": 0.9827450027340665,
+      "grad_norm": 0.1735037863254547,
+      "learning_rate": 7.441448639826099e-08,
+      "loss": 1.1572,
+      "step": 16175
+    },
+    {
+      "epoch": 0.982805759766693,
+      "grad_norm": 2.5373318195343018,
+      "learning_rate": 7.389332406645966e-08,
+      "loss": 1.0495,
+      "step": 16176
+    },
+    {
+      "epoch": 0.9828665167993195,
+      "grad_norm": 0.147075355052948,
+      "learning_rate": 7.337399178574678e-08,
+      "loss": 1.0395,
+      "step": 16177
+    },
+    {
+      "epoch": 0.982927273831946,
+      "grad_norm": 0.10163099318742752,
+      "learning_rate": 7.285648957515712e-08,
+      "loss": 1.0222,
+      "step": 16178
+    },
+    {
+      "epoch": 0.9829880308645725,
+      "grad_norm": 0.13326889276504517,
+      "learning_rate": 7.234081745366438e-08,
+      "loss": 1.0021,
+      "step": 16179
+    },
+    {
+      "epoch": 0.9830487878971991,
+      "grad_norm": 0.22920936346054077,
+      "learning_rate": 7.182697544017569e-08,
+      "loss": 1.0977,
+      "step": 16180
+    },
+    {
+      "epoch": 0.9831095449298256,
+      "grad_norm": 0.13143767416477203,
+      "learning_rate": 7.131496355351485e-08,
+      "loss": 1.232,
+      "step": 16181
+    },
+    {
+      "epoch": 0.9831703019624521,
+      "grad_norm": 0.11583894491195679,
+      "learning_rate": 7.080478181245576e-08,
+      "loss": 1.0703,
+      "step": 16182
+    },
+    {
+      "epoch": 0.9832310589950787,
+      "grad_norm": 0.15936227142810822,
+      "learning_rate": 7.02964302357001e-08,
+      "loss": 1.0567,
+      "step": 16183
+    },
+    {
+      "epoch": 0.9832918160277052,
+      "grad_norm": 0.20767661929130554,
+      "learning_rate": 6.978990884187742e-08,
+      "loss": 1.0982,
+      "step": 16184
+    },
+    {
+      "epoch": 0.9833525730603317,
+      "grad_norm": 0.1629217118024826,
+      "learning_rate": 6.928521764956176e-08,
+      "loss": 1.0403,
+      "step": 16185
+    },
+    {
+      "epoch": 0.9834133300929583,
+      "grad_norm": 0.11175232380628586,
+      "learning_rate": 6.878235667724386e-08,
+      "loss": 1.0727,
+      "step": 16186
+    },
+    {
+      "epoch": 0.9834740871255848,
+      "grad_norm": 0.21810823678970337,
+      "learning_rate": 6.828132594336456e-08,
+      "loss": 1.2005,
+      "step": 16187
+    },
+    {
+      "epoch": 0.9835348441582114,
+      "grad_norm": 0.14823606610298157,
+      "learning_rate": 6.778212546628138e-08,
+      "loss": 1.0498,
+      "step": 16188
+    },
+    {
+      "epoch": 0.9835956011908379,
+      "grad_norm": 0.11760478466749191,
+      "learning_rate": 6.728475526429634e-08,
+      "loss": 1.0549,
+      "step": 16189
+    },
+    {
+      "epoch": 0.9836563582234643,
+      "grad_norm": 0.1506522297859192,
+      "learning_rate": 6.67892153556393e-08,
+      "loss": 1.1545,
+      "step": 16190
+    },
+    {
+      "epoch": 0.9837171152560908,
+      "grad_norm": 0.6151314377784729,
+      "learning_rate": 6.629550575847354e-08,
+      "loss": 1.0127,
+      "step": 16191
+    },
+    {
+      "epoch": 0.9837778722887174,
+      "grad_norm": 0.10732372850179672,
+      "learning_rate": 6.580362649090122e-08,
+      "loss": 1.0032,
+      "step": 16192
+    },
+    {
+      "epoch": 0.9838386293213439,
+      "grad_norm": 0.17160214483737946,
+      "learning_rate": 6.531357757095235e-08,
+      "loss": 1.1107,
+      "step": 16193
+    },
+    {
+      "epoch": 0.9838993863539705,
+      "grad_norm": 0.15070758759975433,
+      "learning_rate": 6.482535901657927e-08,
+      "loss": 1.0009,
+      "step": 16194
+    },
+    {
+      "epoch": 0.983960143386597,
+      "grad_norm": 0.15262217819690704,
+      "learning_rate": 6.433897084568985e-08,
+      "loss": 1.0278,
+      "step": 16195
+    },
+    {
+      "epoch": 0.9840209004192235,
+      "grad_norm": 0.18083415925502777,
+      "learning_rate": 6.385441307610873e-08,
+      "loss": 1.0858,
+      "step": 16196
+    },
+    {
+      "epoch": 0.9840816574518501,
+      "grad_norm": 0.1496964991092682,
+      "learning_rate": 6.337168572559393e-08,
+      "loss": 1.1024,
+      "step": 16197
+    },
+    {
+      "epoch": 0.9841424144844766,
+      "grad_norm": 0.16877521574497223,
+      "learning_rate": 6.289078881184241e-08,
+      "loss": 1.1185,
+      "step": 16198
+    },
+    {
+      "epoch": 0.9842031715171031,
+      "grad_norm": 0.17392559349536896,
+      "learning_rate": 6.241172235247894e-08,
+      "loss": 1.0851,
+      "step": 16199
+    },
+    {
+      "epoch": 0.9842639285497297,
+      "grad_norm": 0.8473336696624756,
+      "learning_rate": 6.193448636506726e-08,
+      "loss": 1.0939,
+      "step": 16200
+    },
+    {
+      "epoch": 0.9843246855823562,
+      "grad_norm": 0.17369872331619263,
+      "learning_rate": 6.14590808671045e-08,
+      "loss": 1.0155,
+      "step": 16201
+    },
+    {
+      "epoch": 0.9843854426149827,
+      "grad_norm": 0.2819485068321228,
+      "learning_rate": 6.098550587601004e-08,
+      "loss": 1.1304,
+      "step": 16202
+    },
+    {
+      "epoch": 0.9844461996476093,
+      "grad_norm": 0.17543594539165497,
+      "learning_rate": 6.051376140914778e-08,
+      "loss": 1.076,
+      "step": 16203
+    },
+    {
+      "epoch": 0.9845069566802357,
+      "grad_norm": 0.12441477179527283,
+      "learning_rate": 6.004384748379832e-08,
+      "loss": 1.028,
+      "step": 16204
+    },
+    {
+      "epoch": 0.9845677137128622,
+      "grad_norm": 0.13013726472854614,
+      "learning_rate": 5.957576411720345e-08,
+      "loss": 1.0308,
+      "step": 16205
+    },
+    {
+      "epoch": 0.9846284707454888,
+      "grad_norm": 0.24910157918930054,
+      "learning_rate": 5.9109511326504994e-08,
+      "loss": 1.0836,
+      "step": 16206
+    },
+    {
+      "epoch": 0.9846892277781153,
+      "grad_norm": 0.1251731514930725,
+      "learning_rate": 5.864508912880595e-08,
+      "loss": 1.0699,
+      "step": 16207
+    },
+    {
+      "epoch": 0.9847499848107418,
+      "grad_norm": 0.269575834274292,
+      "learning_rate": 5.818249754112603e-08,
+      "loss": 1.0653,
+      "step": 16208
+    },
+    {
+      "epoch": 0.9848107418433684,
+      "grad_norm": 0.24665388464927673,
+      "learning_rate": 5.77217365804128e-08,
+      "loss": 1.0616,
+      "step": 16209
+    },
+    {
+      "epoch": 0.9848714988759949,
+      "grad_norm": 0.1910027265548706,
+      "learning_rate": 5.726280626356939e-08,
+      "loss": 1.1048,
+      "step": 16210
+    },
+    {
+      "epoch": 0.9849322559086214,
+      "grad_norm": 0.1944764405488968,
+      "learning_rate": 5.680570660741014e-08,
+      "loss": 1.2086,
+      "step": 16211
+    },
+    {
+      "epoch": 0.984993012941248,
+      "grad_norm": 0.2127663940191269,
+      "learning_rate": 5.6350437628688303e-08,
+      "loss": 1.161,
+      "step": 16212
+    },
+    {
+      "epoch": 0.9850537699738745,
+      "grad_norm": 0.12328021228313446,
+      "learning_rate": 5.589699934410164e-08,
+      "loss": 0.9736,
+      "step": 16213
+    },
+    {
+      "epoch": 0.985114527006501,
+      "grad_norm": 0.4507182836532593,
+      "learning_rate": 5.544539177025909e-08,
+      "loss": 1.1743,
+      "step": 16214
+    },
+    {
+      "epoch": 0.9851752840391276,
+      "grad_norm": 0.1443803310394287,
+      "learning_rate": 5.499561492371963e-08,
+      "loss": 1.1052,
+      "step": 16215
+    },
+    {
+      "epoch": 0.9852360410717541,
+      "grad_norm": 0.09980019181966782,
+      "learning_rate": 5.454766882097007e-08,
+      "loss": 1.0177,
+      "step": 16216
+    },
+    {
+      "epoch": 0.9852967981043805,
+      "grad_norm": 0.1433623880147934,
+      "learning_rate": 5.410155347843615e-08,
+      "loss": 1.1324,
+      "step": 16217
+    },
+    {
+      "epoch": 0.9853575551370071,
+      "grad_norm": 0.15100646018981934,
+      "learning_rate": 5.3657268912454815e-08,
+      "loss": 1.0952,
+      "step": 16218
+    },
+    {
+      "epoch": 0.9854183121696336,
+      "grad_norm": 0.10645327717065811,
+      "learning_rate": 5.321481513932969e-08,
+      "loss": 0.9999,
+      "step": 16219
+    },
+    {
+      "epoch": 0.9854790692022601,
+      "grad_norm": 0.2934175729751587,
+      "learning_rate": 5.277419217526447e-08,
+      "loss": 1.1087,
+      "step": 16220
+    },
+    {
+      "epoch": 0.9855398262348867,
+      "grad_norm": 0.4645816683769226,
+      "learning_rate": 5.233540003641291e-08,
+      "loss": 1.0195,
+      "step": 16221
+    },
+    {
+      "epoch": 0.9856005832675132,
+      "grad_norm": 0.1823212206363678,
+      "learning_rate": 5.189843873886768e-08,
+      "loss": 1.1439,
+      "step": 16222
+    },
+    {
+      "epoch": 0.9856613403001397,
+      "grad_norm": 1.103392243385315,
+      "learning_rate": 5.1463308298638216e-08,
+      "loss": 1.0375,
+      "step": 16223
+    },
+    {
+      "epoch": 0.9857220973327663,
+      "grad_norm": 0.12636372447013855,
+      "learning_rate": 5.1030008731678406e-08,
+      "loss": 1.0799,
+      "step": 16224
+    },
+    {
+      "epoch": 0.9857828543653928,
+      "grad_norm": 0.15905901789665222,
+      "learning_rate": 5.059854005386444e-08,
+      "loss": 1.0932,
+      "step": 16225
+    },
+    {
+      "epoch": 0.9858436113980193,
+      "grad_norm": 0.39884480834007263,
+      "learning_rate": 5.0168902281017003e-08,
+      "loss": 1.0032,
+      "step": 16226
+    },
+    {
+      "epoch": 0.9859043684306459,
+      "grad_norm": 0.17216506600379944,
+      "learning_rate": 4.974109542888461e-08,
+      "loss": 1.0447,
+      "step": 16227
+    },
+    {
+      "epoch": 0.9859651254632724,
+      "grad_norm": 0.1548287570476532,
+      "learning_rate": 4.931511951314915e-08,
+      "loss": 1.1883,
+      "step": 16228
+    },
+    {
+      "epoch": 0.986025882495899,
+      "grad_norm": 0.08968793600797653,
+      "learning_rate": 4.889097454942037e-08,
+      "loss": 0.9937,
+      "step": 16229
+    },
+    {
+      "epoch": 0.9860866395285254,
+      "grad_norm": 0.1462932825088501,
+      "learning_rate": 4.846866055324695e-08,
+      "loss": 1.0091,
+      "step": 16230
+    },
+    {
+      "epoch": 0.9861473965611519,
+      "grad_norm": 0.11029744893312454,
+      "learning_rate": 4.8048177540116476e-08,
+      "loss": 1.0655,
+      "step": 16231
+    },
+    {
+      "epoch": 0.9862081535937784,
+      "grad_norm": 0.2198856770992279,
+      "learning_rate": 4.7629525525427767e-08,
+      "loss": 1.1506,
+      "step": 16232
+    },
+    {
+      "epoch": 0.986268910626405,
+      "grad_norm": 0.1832173764705658,
+      "learning_rate": 4.721270452454074e-08,
+      "loss": 1.1574,
+      "step": 16233
+    },
+    {
+      "epoch": 0.9863296676590315,
+      "grad_norm": 0.09299392253160477,
+      "learning_rate": 4.679771455272653e-08,
+      "loss": 1.0169,
+      "step": 16234
+    },
+    {
+      "epoch": 0.986390424691658,
+      "grad_norm": 0.14054886996746063,
+      "learning_rate": 4.638455562519517e-08,
+      "loss": 1.0209,
+      "step": 16235
+    },
+    {
+      "epoch": 0.9864511817242846,
+      "grad_norm": 0.15262296795845032,
+      "learning_rate": 4.597322775709567e-08,
+      "loss": 1.0365,
+      "step": 16236
+    },
+    {
+      "epoch": 0.9865119387569111,
+      "grad_norm": 0.22205904126167297,
+      "learning_rate": 4.5563730963499306e-08,
+      "loss": 1.0497,
+      "step": 16237
+    },
+    {
+      "epoch": 0.9865726957895377,
+      "grad_norm": 0.09753791987895966,
+      "learning_rate": 4.515606525942739e-08,
+      "loss": 1.0243,
+      "step": 16238
+    },
+    {
+      "epoch": 0.9866334528221642,
+      "grad_norm": 0.3189934194087982,
+      "learning_rate": 4.475023065981243e-08,
+      "loss": 1.0419,
+      "step": 16239
+    },
+    {
+      "epoch": 0.9866942098547907,
+      "grad_norm": 0.09614891558885574,
+      "learning_rate": 4.434622717953141e-08,
+      "loss": 1.0347,
+      "step": 16240
+    },
+    {
+      "epoch": 0.9867549668874173,
+      "grad_norm": 0.5230992436408997,
+      "learning_rate": 4.394405483340025e-08,
+      "loss": 1.2699,
+      "step": 16241
+    },
+    {
+      "epoch": 0.9868157239200438,
+      "grad_norm": 0.18224404752254486,
+      "learning_rate": 4.3543713636151615e-08,
+      "loss": 1.1505,
+      "step": 16242
+    },
+    {
+      "epoch": 0.9868764809526702,
+      "grad_norm": 0.09472563117742538,
+      "learning_rate": 4.3145203602468207e-08,
+      "loss": 1.0057,
+      "step": 16243
+    },
+    {
+      "epoch": 0.9869372379852968,
+      "grad_norm": 0.15544719994068146,
+      "learning_rate": 4.274852474694946e-08,
+      "loss": 1.1309,
+      "step": 16244
+    },
+    {
+      "epoch": 0.9869979950179233,
+      "grad_norm": 0.2506932318210602,
+      "learning_rate": 4.235367708414484e-08,
+      "loss": 1.2998,
+      "step": 16245
+    },
+    {
+      "epoch": 0.9870587520505498,
+      "grad_norm": 6.877426624298096,
+      "learning_rate": 4.1960660628520554e-08,
+      "loss": 1.1083,
+      "step": 16246
+    },
+    {
+      "epoch": 0.9871195090831764,
+      "grad_norm": 0.6385260224342346,
+      "learning_rate": 4.15694753944873e-08,
+      "loss": 1.1019,
+      "step": 16247
+    },
+    {
+      "epoch": 0.9871802661158029,
+      "grad_norm": 0.10798175632953644,
+      "learning_rate": 4.118012139637806e-08,
+      "loss": 1.0023,
+      "step": 16248
+    },
+    {
+      "epoch": 0.9872410231484294,
+      "grad_norm": 0.1646491289138794,
+      "learning_rate": 4.079259864847029e-08,
+      "loss": 1.0992,
+      "step": 16249
+    },
+    {
+      "epoch": 0.987301780181056,
+      "grad_norm": 0.1321839541196823,
+      "learning_rate": 4.040690716496376e-08,
+      "loss": 1.0144,
+      "step": 16250
+    },
+    {
+      "epoch": 0.9873625372136825,
+      "grad_norm": 0.14921942353248596,
+      "learning_rate": 4.00230469600027e-08,
+      "loss": 1.0754,
+      "step": 16251
+    },
+    {
+      "epoch": 0.987423294246309,
+      "grad_norm": 0.16771191358566284,
+      "learning_rate": 3.9641018047653633e-08,
+      "loss": 1.0244,
+      "step": 16252
+    },
+    {
+      "epoch": 0.9874840512789356,
+      "grad_norm": 0.10679901391267776,
+      "learning_rate": 3.9260820441916477e-08,
+      "loss": 1.0182,
+      "step": 16253
+    },
+    {
+      "epoch": 0.9875448083115621,
+      "grad_norm": 0.16618993878364563,
+      "learning_rate": 3.8882454156735634e-08,
+      "loss": 1.0964,
+      "step": 16254
+    },
+    {
+      "epoch": 0.9876055653441886,
+      "grad_norm": 3.1540703773498535,
+      "learning_rate": 3.850591920597224e-08,
+      "loss": 1.0817,
+      "step": 16255
+    },
+    {
+      "epoch": 0.9876663223768151,
+      "grad_norm": 0.1641586869955063,
+      "learning_rate": 3.8131215603437467e-08,
+      "loss": 1.0985,
+      "step": 16256
+    },
+    {
+      "epoch": 0.9877270794094416,
+      "grad_norm": 0.1605738401412964,
+      "learning_rate": 3.775834336285367e-08,
+      "loss": 1.072,
+      "step": 16257
+    },
+    {
+      "epoch": 0.9877878364420681,
+      "grad_norm": 0.14778538048267365,
+      "learning_rate": 3.738730249789879e-08,
+      "loss": 1.1756,
+      "step": 16258
+    },
+    {
+      "epoch": 0.9878485934746947,
+      "grad_norm": 0.16662347316741943,
+      "learning_rate": 3.701809302216752e-08,
+      "loss": 1.1221,
+      "step": 16259
+    },
+    {
+      "epoch": 0.9879093505073212,
+      "grad_norm": 0.11811026930809021,
+      "learning_rate": 3.665071494919903e-08,
+      "loss": 1.04,
+      "step": 16260
+    },
+    {
+      "epoch": 0.9879701075399477,
+      "grad_norm": 0.14614927768707275,
+      "learning_rate": 3.6285168292454766e-08,
+      "loss": 1.0863,
+      "step": 16261
+    },
+    {
+      "epoch": 0.9880308645725743,
+      "grad_norm": 0.10567299276590347,
+      "learning_rate": 3.592145306533512e-08,
+      "loss": 1.0259,
+      "step": 16262
+    },
+    {
+      "epoch": 0.9880916216052008,
+      "grad_norm": 0.11595337837934494,
+      "learning_rate": 3.555956928117388e-08,
+      "loss": 0.975,
+      "step": 16263
+    },
+    {
+      "epoch": 0.9881523786378273,
+      "grad_norm": 0.106806680560112,
+      "learning_rate": 3.519951695323265e-08,
+      "loss": 1.0369,
+      "step": 16264
+    },
+    {
+      "epoch": 0.9882131356704539,
+      "grad_norm": 0.19005513191223145,
+      "learning_rate": 3.4841296094711984e-08,
+      "loss": 1.0304,
+      "step": 16265
+    },
+    {
+      "epoch": 0.9882738927030804,
+      "grad_norm": 0.236617311835289,
+      "learning_rate": 3.4484906718740276e-08,
+      "loss": 1.0201,
+      "step": 16266
+    },
+    {
+      "epoch": 0.988334649735707,
+      "grad_norm": 0.17361678183078766,
+      "learning_rate": 3.413034883838484e-08,
+      "loss": 1.1204,
+      "step": 16267
+    },
+    {
+      "epoch": 0.9883954067683335,
+      "grad_norm": 0.16652122139930725,
+      "learning_rate": 3.377762246663529e-08,
+      "loss": 1.1523,
+      "step": 16268
+    },
+    {
+      "epoch": 0.9884561638009599,
+      "grad_norm": 0.16175909340381622,
+      "learning_rate": 3.3426727616431286e-08,
+      "loss": 1.0317,
+      "step": 16269
+    },
+    {
+      "epoch": 0.9885169208335864,
+      "grad_norm": 0.09417491406202316,
+      "learning_rate": 3.3077664300623645e-08,
+      "loss": 1.0093,
+      "step": 16270
+    },
+    {
+      "epoch": 0.988577677866213,
+      "grad_norm": 0.12074661254882812,
+      "learning_rate": 3.273043253201324e-08,
+      "loss": 0.9837,
+      "step": 16271
+    },
+    {
+      "epoch": 0.9886384348988395,
+      "grad_norm": 0.12248991429805756,
+      "learning_rate": 3.238503232333434e-08,
+      "loss": 1.0563,
+      "step": 16272
+    },
+    {
+      "epoch": 0.988699191931466,
+      "grad_norm": 0.1273757815361023,
+      "learning_rate": 3.2041463687232374e-08,
+      "loss": 0.9905,
+      "step": 16273
+    },
+    {
+      "epoch": 0.9887599489640926,
+      "grad_norm": 0.2450380176305771,
+      "learning_rate": 3.169972663631393e-08,
+      "loss": 1.2046,
+      "step": 16274
+    },
+    {
+      "epoch": 0.9888207059967191,
+      "grad_norm": 0.3407484292984009,
+      "learning_rate": 3.135982118309677e-08,
+      "loss": 1.1031,
+      "step": 16275
+    },
+    {
+      "epoch": 0.9888814630293457,
+      "grad_norm": 0.11378134042024612,
+      "learning_rate": 3.10217473400487e-08,
+      "loss": 1.0558,
+      "step": 16276
+    },
+    {
+      "epoch": 0.9889422200619722,
+      "grad_norm": 0.11455726623535156,
+      "learning_rate": 3.068550511955426e-08,
+      "loss": 1.0424,
+      "step": 16277
+    },
+    {
+      "epoch": 0.9890029770945987,
+      "grad_norm": 0.10960632562637329,
+      "learning_rate": 3.035109453393692e-08,
+      "loss": 1.0592,
+      "step": 16278
+    },
+    {
+      "epoch": 0.9890637341272253,
+      "grad_norm": 0.10926475375890732,
+      "learning_rate": 3.0018515595464645e-08,
+      "loss": 1.0339,
+      "step": 16279
+    },
+    {
+      "epoch": 0.9891244911598518,
+      "grad_norm": 0.14063212275505066,
+      "learning_rate": 2.968776831632214e-08,
+      "loss": 1.0296,
+      "step": 16280
+    },
+    {
+      "epoch": 0.9891852481924783,
+      "grad_norm": 0.13234879076480865,
+      "learning_rate": 2.9358852708633032e-08,
+      "loss": 1.0527,
+      "step": 16281
+    },
+    {
+      "epoch": 0.9892460052251048,
+      "grad_norm": 0.1297687143087387,
+      "learning_rate": 2.903176878445435e-08,
+      "loss": 1.0492,
+      "step": 16282
+    },
+    {
+      "epoch": 0.9893067622577313,
+      "grad_norm": 0.301035612821579,
+      "learning_rate": 2.8706516555776497e-08,
+      "loss": 1.0035,
+      "step": 16283
+    },
+    {
+      "epoch": 0.9893675192903578,
+      "grad_norm": 0.11623239517211914,
+      "learning_rate": 2.838309603452327e-08,
+      "loss": 1.0135,
+      "step": 16284
+    },
+    {
+      "epoch": 0.9894282763229844,
+      "grad_norm": 0.5636290311813354,
+      "learning_rate": 2.806150723254075e-08,
+      "loss": 1.1364,
+      "step": 16285
+    },
+    {
+      "epoch": 0.9894890333556109,
+      "grad_norm": 0.14663462340831757,
+      "learning_rate": 2.7741750161625057e-08,
+      "loss": 1.0931,
+      "step": 16286
+    },
+    {
+      "epoch": 0.9895497903882374,
+      "grad_norm": 0.24398909509181976,
+      "learning_rate": 2.7423824833494593e-08,
+      "loss": 1.0783,
+      "step": 16287
+    },
+    {
+      "epoch": 0.989610547420864,
+      "grad_norm": 0.12663604319095612,
+      "learning_rate": 2.71077312598067e-08,
+      "loss": 1.0267,
+      "step": 16288
+    },
+    {
+      "epoch": 0.9896713044534905,
+      "grad_norm": 0.19128142297267914,
+      "learning_rate": 2.6793469452141007e-08,
+      "loss": 1.0423,
+      "step": 16289
+    },
+    {
+      "epoch": 0.989732061486117,
+      "grad_norm": 0.18244168162345886,
+      "learning_rate": 2.6481039422021626e-08,
+      "loss": 1.0873,
+      "step": 16290
+    },
+    {
+      "epoch": 0.9897928185187436,
+      "grad_norm": 0.18155498802661896,
+      "learning_rate": 2.6170441180900506e-08,
+      "loss": 1.1018,
+      "step": 16291
+    },
+    {
+      "epoch": 0.9898535755513701,
+      "grad_norm": 0.11443190276622772,
+      "learning_rate": 2.586167474015744e-08,
+      "loss": 1.0669,
+      "step": 16292
+    },
+    {
+      "epoch": 0.9899143325839966,
+      "grad_norm": 0.1291109025478363,
+      "learning_rate": 2.5554740111116693e-08,
+      "loss": 1.0626,
+      "step": 16293
+    },
+    {
+      "epoch": 0.9899750896166232,
+      "grad_norm": 0.11472375690937042,
+      "learning_rate": 2.5249637305030383e-08,
+      "loss": 1.0227,
+      "step": 16294
+    },
+    {
+      "epoch": 0.9900358466492496,
+      "grad_norm": 0.15321263670921326,
+      "learning_rate": 2.49463663330729e-08,
+      "loss": 1.0819,
+      "step": 16295
+    },
+    {
+      "epoch": 0.9900966036818761,
+      "grad_norm": 0.3910882771015167,
+      "learning_rate": 2.4644927206368685e-08,
+      "loss": 1.0588,
+      "step": 16296
+    },
+    {
+      "epoch": 0.9901573607145027,
+      "grad_norm": 0.20071378350257874,
+      "learning_rate": 2.434531993596445e-08,
+      "loss": 1.1416,
+      "step": 16297
+    },
+    {
+      "epoch": 0.9902181177471292,
+      "grad_norm": 0.7159701585769653,
+      "learning_rate": 2.4047544532845857e-08,
+      "loss": 1.1266,
+      "step": 16298
+    },
+    {
+      "epoch": 0.9902788747797557,
+      "grad_norm": 0.19287821650505066,
+      "learning_rate": 2.3751601007920842e-08,
+      "loss": 1.0753,
+      "step": 16299
+    },
+    {
+      "epoch": 0.9903396318123823,
+      "grad_norm": 0.12025300413370132,
+      "learning_rate": 2.345748937204184e-08,
+      "loss": 1.0383,
+      "step": 16300
+    },
+    {
+      "epoch": 0.9904003888450088,
+      "grad_norm": 0.19872963428497314,
+      "learning_rate": 2.3165209635989116e-08,
+      "loss": 1.0726,
+      "step": 16301
+    },
+    {
+      "epoch": 0.9904611458776353,
+      "grad_norm": 0.11132209002971649,
+      "learning_rate": 2.287476181047632e-08,
+      "loss": 1.0227,
+      "step": 16302
+    },
+    {
+      "epoch": 0.9905219029102619,
+      "grad_norm": 0.10098928958177567,
+      "learning_rate": 2.258614590615049e-08,
+      "loss": 1.0289,
+      "step": 16303
+    },
+    {
+      "epoch": 0.9905826599428884,
+      "grad_norm": 0.17251595854759216,
+      "learning_rate": 2.2299361933586505e-08,
+      "loss": 1.0801,
+      "step": 16304
+    },
+    {
+      "epoch": 0.990643416975515,
+      "grad_norm": 0.221372589468956,
+      "learning_rate": 2.2014409903303724e-08,
+      "loss": 1.1968,
+      "step": 16305
+    },
+    {
+      "epoch": 0.9907041740081415,
+      "grad_norm": 0.15848013758659363,
+      "learning_rate": 2.1731289825738245e-08,
+      "loss": 1.0452,
+      "step": 16306
+    },
+    {
+      "epoch": 0.990764931040768,
+      "grad_norm": 0.5078273415565491,
+      "learning_rate": 2.14500017112762e-08,
+      "loss": 1.0953,
+      "step": 16307
+    },
+    {
+      "epoch": 0.9908256880733946,
+      "grad_norm": 0.09761633723974228,
+      "learning_rate": 2.117054557022602e-08,
+      "loss": 1.0434,
+      "step": 16308
+    },
+    {
+      "epoch": 0.990886445106021,
+      "grad_norm": 0.1328747570514679,
+      "learning_rate": 2.0892921412829504e-08,
+      "loss": 1.0404,
+      "step": 16309
+    },
+    {
+      "epoch": 0.9909472021386475,
+      "grad_norm": 0.1075906828045845,
+      "learning_rate": 2.0617129249261845e-08,
+      "loss": 1.0375,
+      "step": 16310
+    },
+    {
+      "epoch": 0.991007959171274,
+      "grad_norm": 0.13317234814167023,
+      "learning_rate": 2.0343169089637182e-08,
+      "loss": 1.0308,
+      "step": 16311
+    },
+    {
+      "epoch": 0.9910687162039006,
+      "grad_norm": 0.09540808945894241,
+      "learning_rate": 2.0071040943991926e-08,
+      "loss": 1.0466,
+      "step": 16312
+    },
+    {
+      "epoch": 0.9911294732365271,
+      "grad_norm": 0.09854982048273087,
+      "learning_rate": 1.980074482230698e-08,
+      "loss": 1.0191,
+      "step": 16313
+    },
+    {
+      "epoch": 0.9911902302691536,
+      "grad_norm": 0.2078581005334854,
+      "learning_rate": 1.953228073447999e-08,
+      "loss": 1.063,
+      "step": 16314
+    },
+    {
+      "epoch": 0.9912509873017802,
+      "grad_norm": 0.11335760354995728,
+      "learning_rate": 1.926564869036418e-08,
+      "loss": 1.0425,
+      "step": 16315
+    },
+    {
+      "epoch": 0.9913117443344067,
+      "grad_norm": 0.42150208353996277,
+      "learning_rate": 1.9000848699723962e-08,
+      "loss": 1.1199,
+      "step": 16316
+    },
+    {
+      "epoch": 0.9913725013670333,
+      "grad_norm": 0.12567904591560364,
+      "learning_rate": 1.8737880772273786e-08,
+      "loss": 1.0938,
+      "step": 16317
+    },
+    {
+      "epoch": 0.9914332583996598,
+      "grad_norm": 0.17655733227729797,
+      "learning_rate": 1.847674491763929e-08,
+      "loss": 1.1089,
+      "step": 16318
+    },
+    {
+      "epoch": 0.9914940154322863,
+      "grad_norm": 0.19671276211738586,
+      "learning_rate": 1.8217441145407243e-08,
+      "loss": 1.1662,
+      "step": 16319
+    },
+    {
+      "epoch": 0.9915547724649129,
+      "grad_norm": 0.24333301186561584,
+      "learning_rate": 1.795996946507561e-08,
+      "loss": 1.0904,
+      "step": 16320
+    },
+    {
+      "epoch": 0.9916155294975394,
+      "grad_norm": 0.15849147737026215,
+      "learning_rate": 1.7704329886081282e-08,
+      "loss": 1.0899,
+      "step": 16321
+    },
+    {
+      "epoch": 0.9916762865301658,
+      "grad_norm": 0.10567186027765274,
+      "learning_rate": 1.7450522417794546e-08,
+      "loss": 1.0363,
+      "step": 16322
+    },
+    {
+      "epoch": 0.9917370435627924,
+      "grad_norm": 0.23922015726566315,
+      "learning_rate": 1.719854706952462e-08,
+      "loss": 1.1539,
+      "step": 16323
+    },
+    {
+      "epoch": 0.9917978005954189,
+      "grad_norm": 0.9320358037948608,
+      "learning_rate": 1.6948403850503003e-08,
+      "loss": 1.0789,
+      "step": 16324
+    },
+    {
+      "epoch": 0.9918585576280454,
+      "grad_norm": 0.24150465428829193,
+      "learning_rate": 1.6700092769894593e-08,
+      "loss": 1.0223,
+      "step": 16325
+    },
+    {
+      "epoch": 0.991919314660672,
+      "grad_norm": 0.13452938199043274,
+      "learning_rate": 1.645361383680877e-08,
+      "loss": 0.9891,
+      "step": 16326
+    },
+    {
+      "epoch": 0.9919800716932985,
+      "grad_norm": 0.12692850828170776,
+      "learning_rate": 1.6208967060277193e-08,
+      "loss": 1.043,
+      "step": 16327
+    },
+    {
+      "epoch": 0.992040828725925,
+      "grad_norm": 0.1624302864074707,
+      "learning_rate": 1.5966152449264916e-08,
+      "loss": 1.1034,
+      "step": 16328
+    },
+    {
+      "epoch": 0.9921015857585516,
+      "grad_norm": 0.1466129571199417,
+      "learning_rate": 1.5725170012681478e-08,
+      "loss": 1.313,
+      "step": 16329
+    },
+    {
+      "epoch": 0.9921623427911781,
+      "grad_norm": 0.20591171085834503,
+      "learning_rate": 1.5486019759347603e-08,
+      "loss": 1.1587,
+      "step": 16330
+    },
+    {
+      "epoch": 0.9922230998238046,
+      "grad_norm": 0.10770957916975021,
+      "learning_rate": 1.5248701698039604e-08,
+      "loss": 1.0059,
+      "step": 16331
+    },
+    {
+      "epoch": 0.9922838568564312,
+      "grad_norm": 0.16154597699642181,
+      "learning_rate": 1.5013215837450522e-08,
+      "loss": 1.0685,
+      "step": 16332
+    },
+    {
+      "epoch": 0.9923446138890577,
+      "grad_norm": 0.18607288599014282,
+      "learning_rate": 1.4779562186212348e-08,
+      "loss": 1.1741,
+      "step": 16333
+    },
+    {
+      "epoch": 0.9924053709216842,
+      "grad_norm": 0.133961021900177,
+      "learning_rate": 1.4547740752890449e-08,
+      "loss": 1.0493,
+      "step": 16334
+    },
+    {
+      "epoch": 0.9924661279543107,
+      "grad_norm": 0.1740068644285202,
+      "learning_rate": 1.4317751545983582e-08,
+      "loss": 1.0907,
+      "step": 16335
+    },
+    {
+      "epoch": 0.9925268849869372,
+      "grad_norm": 0.11344313621520996,
+      "learning_rate": 1.4089594573923893e-08,
+      "loss": 1.0242,
+      "step": 16336
+    },
+    {
+      "epoch": 0.9925876420195637,
+      "grad_norm": 0.14299818873405457,
+      "learning_rate": 1.3863269845071359e-08,
+      "loss": 1.1523,
+      "step": 16337
+    },
+    {
+      "epoch": 0.9926483990521903,
+      "grad_norm": 0.12203660607337952,
+      "learning_rate": 1.3638777367724897e-08,
+      "loss": 1.1053,
+      "step": 16338
+    },
+    {
+      "epoch": 0.9927091560848168,
+      "grad_norm": 0.39186519384384155,
+      "learning_rate": 1.3416117150105711e-08,
+      "loss": 1.1477,
+      "step": 16339
+    },
+    {
+      "epoch": 0.9927699131174433,
+      "grad_norm": 0.1688614934682846,
+      "learning_rate": 1.3195289200385042e-08,
+      "loss": 1.1292,
+      "step": 16340
+    },
+    {
+      "epoch": 0.9928306701500699,
+      "grad_norm": 0.1663534939289093,
+      "learning_rate": 1.2976293526656414e-08,
+      "loss": 1.1446,
+      "step": 16341
+    },
+    {
+      "epoch": 0.9928914271826964,
+      "grad_norm": 0.1755792796611786,
+      "learning_rate": 1.2759130136935638e-08,
+      "loss": 1.0967,
+      "step": 16342
+    },
+    {
+      "epoch": 0.9929521842153229,
+      "grad_norm": 0.10801321268081665,
+      "learning_rate": 1.2543799039199666e-08,
+      "loss": 1.0146,
+      "step": 16343
+    },
+    {
+      "epoch": 0.9930129412479495,
+      "grad_norm": 0.339424729347229,
+      "learning_rate": 1.2330300241325531e-08,
+      "loss": 1.08,
+      "step": 16344
+    },
+    {
+      "epoch": 0.993073698280576,
+      "grad_norm": 0.11286205053329468,
+      "learning_rate": 1.211863375115141e-08,
+      "loss": 1.0761,
+      "step": 16345
+    },
+    {
+      "epoch": 0.9931344553132025,
+      "grad_norm": 0.1963888704776764,
+      "learning_rate": 1.1908799576432207e-08,
+      "loss": 1.0592,
+      "step": 16346
+    },
+    {
+      "epoch": 0.9931952123458291,
+      "grad_norm": 0.17381581664085388,
+      "learning_rate": 1.1700797724856217e-08,
+      "loss": 1.1608,
+      "step": 16347
+    },
+    {
+      "epoch": 0.9932559693784555,
+      "grad_norm": 0.16749228537082672,
+      "learning_rate": 1.1494628204050672e-08,
+      "loss": 1.0566,
+      "step": 16348
+    },
+    {
+      "epoch": 0.993316726411082,
+      "grad_norm": 0.10370411723852158,
+      "learning_rate": 1.1290291021570643e-08,
+      "loss": 0.9783,
+      "step": 16349
+    },
+    {
+      "epoch": 0.9933774834437086,
+      "grad_norm": 0.15005536377429962,
+      "learning_rate": 1.1087786184910132e-08,
+      "loss": 1.0797,
+      "step": 16350
+    },
+    {
+      "epoch": 0.9934382404763351,
+      "grad_norm": 0.4708842933177948,
+      "learning_rate": 1.088711370148543e-08,
+      "loss": 1.1091,
+      "step": 16351
+    },
+    {
+      "epoch": 0.9934989975089616,
+      "grad_norm": 0.09188425540924072,
+      "learning_rate": 1.0688273578657315e-08,
+      "loss": 1.0199,
+      "step": 16352
+    },
+    {
+      "epoch": 0.9935597545415882,
+      "grad_norm": 0.18060752749443054,
+      "learning_rate": 1.0491265823714403e-08,
+      "loss": 1.0648,
+      "step": 16353
+    },
+    {
+      "epoch": 0.9936205115742147,
+      "grad_norm": 0.18418973684310913,
+      "learning_rate": 1.0296090443878691e-08,
+      "loss": 1.0846,
+      "step": 16354
+    },
+    {
+      "epoch": 0.9936812686068413,
+      "grad_norm": 0.13063186407089233,
+      "learning_rate": 1.010274744630002e-08,
+      "loss": 1.0257,
+      "step": 16355
+    },
+    {
+      "epoch": 0.9937420256394678,
+      "grad_norm": 0.14973171055316925,
+      "learning_rate": 9.911236838067162e-09,
+      "loss": 1.1194,
+      "step": 16356
+    },
+    {
+      "epoch": 0.9938027826720943,
+      "grad_norm": 0.10031884908676147,
+      "learning_rate": 9.721558626207827e-09,
+      "loss": 1.0268,
+      "step": 16357
+    },
+    {
+      "epoch": 0.9938635397047209,
+      "grad_norm": 1.4674253463745117,
+      "learning_rate": 9.533712817660912e-09,
+      "loss": 1.2043,
+      "step": 16358
+    },
+    {
+      "epoch": 0.9939242967373474,
+      "grad_norm": 0.1417066901922226,
+      "learning_rate": 9.347699419320898e-09,
+      "loss": 1.0582,
+      "step": 16359
+    },
+    {
+      "epoch": 0.9939850537699739,
+      "grad_norm": 0.1254492700099945,
+      "learning_rate": 9.163518437999009e-09,
+      "loss": 1.0635,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9940458108026003,
+      "grad_norm": 0.1302124261856079,
+      "learning_rate": 8.981169880456497e-09,
+      "loss": 1.0724,
+      "step": 16361
+    },
+    {
+      "epoch": 0.9941065678352269,
+      "grad_norm": 0.15401579439640045,
+      "learning_rate": 8.800653753371357e-09,
+      "loss": 1.0717,
+      "step": 16362
+    },
+    {
+      "epoch": 0.9941673248678534,
+      "grad_norm": 0.2836470901966095,
+      "learning_rate": 8.621970063360518e-09,
+      "loss": 1.091,
+      "step": 16363
+    },
+    {
+      "epoch": 0.99422808190048,
+      "grad_norm": 0.17619681358337402,
+      "learning_rate": 8.445118816979846e-09,
+      "loss": 1.0607,
+      "step": 16364
+    },
+    {
+      "epoch": 0.9942888389331065,
+      "grad_norm": 0.1280762106180191,
+      "learning_rate": 8.270100020701943e-09,
+      "loss": 1.0516,
+      "step": 16365
+    },
+    {
+      "epoch": 0.994349595965733,
+      "grad_norm": 0.1501254439353943,
+      "learning_rate": 8.096913680943896e-09,
+      "loss": 1.0818,
+      "step": 16366
+    },
+    {
+      "epoch": 0.9944103529983596,
+      "grad_norm": 0.10833926498889923,
+      "learning_rate": 7.925559804061733e-09,
+      "loss": 1.0294,
+      "step": 16367
+    },
+    {
+      "epoch": 0.9944711100309861,
+      "grad_norm": 0.14364440739154816,
+      "learning_rate": 7.756038396328214e-09,
+      "loss": 1.0696,
+      "step": 16368
+    },
+    {
+      "epoch": 0.9945318670636126,
+      "grad_norm": 0.1562419831752777,
+      "learning_rate": 7.588349463960587e-09,
+      "loss": 1.1501,
+      "step": 16369
+    },
+    {
+      "epoch": 0.9945926240962392,
+      "grad_norm": 0.16588133573532104,
+      "learning_rate": 7.422493013103937e-09,
+      "loss": 1.0214,
+      "step": 16370
+    },
+    {
+      "epoch": 0.9946533811288657,
+      "grad_norm": 0.11225719749927521,
+      "learning_rate": 7.258469049842287e-09,
+      "loss": 1.0363,
+      "step": 16371
+    },
+    {
+      "epoch": 0.9947141381614922,
+      "grad_norm": 0.21635957062244415,
+      "learning_rate": 7.0962775801874935e-09,
+      "loss": 1.1032,
+      "step": 16372
+    },
+    {
+      "epoch": 0.9947748951941188,
+      "grad_norm": 0.30810874700546265,
+      "learning_rate": 6.9359186100737e-09,
+      "loss": 1.0764,
+      "step": 16373
+    },
+    {
+      "epoch": 0.9948356522267452,
+      "grad_norm": 0.09591729193925858,
+      "learning_rate": 6.777392145396188e-09,
+      "loss": 1.0214,
+      "step": 16374
+    },
+    {
+      "epoch": 0.9948964092593717,
+      "grad_norm": 0.11895853281021118,
+      "learning_rate": 6.620698191950325e-09,
+      "loss": 0.9957,
+      "step": 16375
+    },
+    {
+      "epoch": 0.9949571662919983,
+      "grad_norm": 0.1277838796377182,
+      "learning_rate": 6.4658367554870644e-09,
+      "loss": 1.0188,
+      "step": 16376
+    },
+    {
+      "epoch": 0.9950179233246248,
+      "grad_norm": 1.3013290166854858,
+      "learning_rate": 6.312807841685198e-09,
+      "loss": 1.1323,
+      "step": 16377
+    },
+    {
+      "epoch": 0.9950786803572513,
+      "grad_norm": 12.565542221069336,
+      "learning_rate": 6.16161145615135e-09,
+      "loss": 1.053,
+      "step": 16378
+    },
+    {
+      "epoch": 0.9951394373898779,
+      "grad_norm": 0.12348482757806778,
+      "learning_rate": 6.012247604425536e-09,
+      "loss": 1.0495,
+      "step": 16379
+    },
+    {
+      "epoch": 0.9952001944225044,
+      "grad_norm": 0.16675661504268646,
+      "learning_rate": 5.864716291981154e-09,
+      "loss": 1.1276,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9952609514551309,
+      "grad_norm": 0.1771889328956604,
+      "learning_rate": 5.719017524236092e-09,
+      "loss": 1.1399,
+      "step": 16381
+    },
+    {
+      "epoch": 0.9953217084877575,
+      "grad_norm": 0.1173153892159462,
+      "learning_rate": 5.575151306519421e-09,
+      "loss": 1.0114,
+      "step": 16382
+    },
+    {
+      "epoch": 0.995382465520384,
+      "grad_norm": 3.2511708736419678,
+      "learning_rate": 5.433117644110252e-09,
+      "loss": 1.0745,
+      "step": 16383
+    },
+    {
+      "epoch": 0.9954432225530105,
+      "grad_norm": 0.10906144231557846,
+      "learning_rate": 5.29291654220998e-09,
+      "loss": 1.0691,
+      "step": 16384
+    },
+    {
+      "epoch": 0.9955039795856371,
+      "grad_norm": 0.2170865833759308,
+      "learning_rate": 5.154548005964488e-09,
+      "loss": 1.0627,
+      "step": 16385
+    },
+    {
+      "epoch": 0.9955647366182636,
+      "grad_norm": 0.18898631632328033,
+      "learning_rate": 5.018012040447495e-09,
+      "loss": 1.0766,
+      "step": 16386
+    },
+    {
+      "epoch": 0.99562549365089,
+      "grad_norm": 0.12484834343194962,
+      "learning_rate": 4.883308650655005e-09,
+      "loss": 1.0526,
+      "step": 16387
+    },
+    {
+      "epoch": 0.9956862506835166,
+      "grad_norm": 0.1729438453912735,
+      "learning_rate": 4.750437841527511e-09,
+      "loss": 1.0582,
+      "step": 16388
+    },
+    {
+      "epoch": 0.9957470077161431,
+      "grad_norm": 0.16722114384174347,
+      "learning_rate": 4.619399617938891e-09,
+      "loss": 1.0758,
+      "step": 16389
+    },
+    {
+      "epoch": 0.9958077647487696,
+      "grad_norm": 0.768099844455719,
+      "learning_rate": 4.49019398469086e-09,
+      "loss": 1.2121,
+      "step": 16390
+    },
+    {
+      "epoch": 0.9958685217813962,
+      "grad_norm": 0.1186041310429573,
+      "learning_rate": 4.362820946512969e-09,
+      "loss": 1.0011,
+      "step": 16391
+    },
+    {
+      "epoch": 0.9959292788140227,
+      "grad_norm": 0.18163001537322998,
+      "learning_rate": 4.2372805080848065e-09,
+      "loss": 1.0594,
+      "step": 16392
+    },
+    {
+      "epoch": 0.9959900358466492,
+      "grad_norm": 0.15095829963684082,
+      "learning_rate": 4.113572674002697e-09,
+      "loss": 1.0616,
+      "step": 16393
+    },
+    {
+      "epoch": 0.9960507928792758,
+      "grad_norm": 0.4118235409259796,
+      "learning_rate": 3.991697448801901e-09,
+      "loss": 1.0738,
+      "step": 16394
+    },
+    {
+      "epoch": 0.9961115499119023,
+      "grad_norm": 0.3481995761394501,
+      "learning_rate": 3.871654836951066e-09,
+      "loss": 1.0326,
+      "step": 16395
+    },
+    {
+      "epoch": 0.9961723069445289,
+      "grad_norm": 0.09429150074720383,
+      "learning_rate": 3.753444842841125e-09,
+      "loss": 0.9885,
+      "step": 16396
+    },
+    {
+      "epoch": 0.9962330639771554,
+      "grad_norm": 0.09320554882287979,
+      "learning_rate": 3.6370674708186005e-09,
+      "loss": 1.0453,
+      "step": 16397
+    },
+    {
+      "epoch": 0.9962938210097819,
+      "grad_norm": 0.1572645604610443,
+      "learning_rate": 3.5225227251467484e-09,
+      "loss": 1.0304,
+      "step": 16398
+    },
+    {
+      "epoch": 0.9963545780424085,
+      "grad_norm": 0.0966515839099884,
+      "learning_rate": 3.4098106100166618e-09,
+      "loss": 1.0741,
+      "step": 16399
+    },
+    {
+      "epoch": 0.9964153350750349,
+      "grad_norm": 0.2818453311920166,
+      "learning_rate": 3.29893112956392e-09,
+      "loss": 1.0589,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9964760921076614,
+      "grad_norm": 0.09154333174228668,
+      "learning_rate": 3.189884287851941e-09,
+      "loss": 0.9976,
+      "step": 16401
+    },
+    {
+      "epoch": 0.996536849140288,
+      "grad_norm": 0.09504634886980057,
+      "learning_rate": 3.0826700888830773e-09,
+      "loss": 1.0373,
+      "step": 16402
+    },
+    {
+      "epoch": 0.9965976061729145,
+      "grad_norm": 0.09509482234716415,
+      "learning_rate": 2.9772885365764168e-09,
+      "loss": 1.0066,
+      "step": 16403
+    },
+    {
+      "epoch": 0.996658363205541,
+      "grad_norm": 0.2784222960472107,
+      "learning_rate": 2.873739634806638e-09,
+      "loss": 1.1793,
+      "step": 16404
+    },
+    {
+      "epoch": 0.9967191202381676,
+      "grad_norm": 0.1504630595445633,
+      "learning_rate": 2.7720233873651523e-09,
+      "loss": 1.1545,
+      "step": 16405
+    },
+    {
+      "epoch": 0.9967798772707941,
+      "grad_norm": 0.17525725066661835,
+      "learning_rate": 2.672139797976758e-09,
+      "loss": 1.0201,
+      "step": 16406
+    },
+    {
+      "epoch": 0.9968406343034206,
+      "grad_norm": 0.13617132604122162,
+      "learning_rate": 2.5740888703051915e-09,
+      "loss": 1.0824,
+      "step": 16407
+    },
+    {
+      "epoch": 0.9969013913360472,
+      "grad_norm": 0.08649192005395889,
+      "learning_rate": 2.4778706079475745e-09,
+      "loss": 0.9728,
+      "step": 16408
+    },
+    {
+      "epoch": 0.9969621483686737,
+      "grad_norm": 0.15265050530433655,
+      "learning_rate": 2.3834850144233143e-09,
+      "loss": 1.0927,
+      "step": 16409
+    },
+    {
+      "epoch": 0.9970229054013002,
+      "grad_norm": 0.12428773194551468,
+      "learning_rate": 2.2909320932018586e-09,
+      "loss": 1.0532,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9970836624339268,
+      "grad_norm": 0.1048312559723854,
+      "learning_rate": 2.2002118476638355e-09,
+      "loss": 1.0563,
+      "step": 16411
+    },
+    {
+      "epoch": 0.9971444194665533,
+      "grad_norm": 0.14031675457954407,
+      "learning_rate": 2.111324281145466e-09,
+      "loss": 1.0772,
+      "step": 16412
+    },
+    {
+      "epoch": 0.9972051764991798,
+      "grad_norm": 0.1768234372138977,
+      "learning_rate": 2.024269396899703e-09,
+      "loss": 1.0697,
+      "step": 16413
+    },
+    {
+      "epoch": 0.9972659335318063,
+      "grad_norm": 0.16252635419368744,
+      "learning_rate": 1.939047198123989e-09,
+      "loss": 1.1371,
+      "step": 16414
+    },
+    {
+      "epoch": 0.9973266905644328,
+      "grad_norm": 0.22938312590122223,
+      "learning_rate": 1.8556576879324993e-09,
+      "loss": 1.109,
+      "step": 16415
+    },
+    {
+      "epoch": 0.9973874475970593,
+      "grad_norm": 0.1136816069483757,
+      "learning_rate": 1.774100869383899e-09,
+      "loss": 0.9748,
+      "step": 16416
+    },
+    {
+      "epoch": 0.9974482046296859,
+      "grad_norm": 0.09893430024385452,
+      "learning_rate": 1.6943767454757897e-09,
+      "loss": 1.0262,
+      "step": 16417
+    },
+    {
+      "epoch": 0.9975089616623124,
+      "grad_norm": 0.12094129621982574,
+      "learning_rate": 1.616485319122507e-09,
+      "loss": 1.0346,
+      "step": 16418
+    },
+    {
+      "epoch": 0.9975697186949389,
+      "grad_norm": 0.12108869105577469,
+      "learning_rate": 1.5404265931773243e-09,
+      "loss": 1.0142,
+      "step": 16419
+    },
+    {
+      "epoch": 0.9976304757275655,
+      "grad_norm": 0.10918483138084412,
+      "learning_rate": 1.4662005704380033e-09,
+      "loss": 1.0577,
+      "step": 16420
+    },
+    {
+      "epoch": 0.997691232760192,
+      "grad_norm": 0.1524430811405182,
+      "learning_rate": 1.3938072536134884e-09,
+      "loss": 1.0922,
+      "step": 16421
+    },
+    {
+      "epoch": 0.9977519897928185,
+      "grad_norm": 0.1589829921722412,
+      "learning_rate": 1.3232466453683146e-09,
+      "loss": 1.1393,
+      "step": 16422
+    },
+    {
+      "epoch": 0.9978127468254451,
+      "grad_norm": 0.16543225944042206,
+      "learning_rate": 1.2545187482837505e-09,
+      "loss": 1.095,
+      "step": 16423
+    },
+    {
+      "epoch": 0.9978735038580716,
+      "grad_norm": 0.21794280409812927,
+      "learning_rate": 1.1876235648800027e-09,
+      "loss": 1.1503,
+      "step": 16424
+    },
+    {
+      "epoch": 0.9979342608906981,
+      "grad_norm": 0.12918835878372192,
+      "learning_rate": 1.1225610976051126e-09,
+      "loss": 1.05,
+      "step": 16425
+    },
+    {
+      "epoch": 0.9979950179233247,
+      "grad_norm": 0.1320475935935974,
+      "learning_rate": 1.059331348851611e-09,
+      "loss": 1.0702,
+      "step": 16426
+    },
+    {
+      "epoch": 0.9980557749559511,
+      "grad_norm": 0.10265633463859558,
+      "learning_rate": 9.979343209287618e-10,
+      "loss": 1.0249,
+      "step": 16427
+    },
+    {
+      "epoch": 0.9981165319885776,
+      "grad_norm": 0.1497434675693512,
+      "learning_rate": 9.383700160958686e-10,
+      "loss": 1.1647,
+      "step": 16428
+    },
+    {
+      "epoch": 0.9981772890212042,
+      "grad_norm": 0.14027445018291473,
+      "learning_rate": 8.806384365289688e-10,
+      "loss": 1.0638,
+      "step": 16429
+    },
+    {
+      "epoch": 0.9982380460538307,
+      "grad_norm": 0.20440469682216644,
+      "learning_rate": 8.247395843430372e-10,
+      "loss": 1.0517,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9982988030864572,
+      "grad_norm": 0.13694840669631958,
+      "learning_rate": 7.706734615975375e-10,
+      "loss": 1.0594,
+      "step": 16431
+    },
+    {
+      "epoch": 0.9983595601190838,
+      "grad_norm": 0.1576005518436432,
+      "learning_rate": 7.184400702631156e-10,
+      "loss": 1.0848,
+      "step": 16432
+    },
+    {
+      "epoch": 0.9984203171517103,
+      "grad_norm": 1.2473230361938477,
+      "learning_rate": 6.680394122604572e-10,
+      "loss": 1.1231,
+      "step": 16433
+    },
+    {
+      "epoch": 0.9984810741843368,
+      "grad_norm": 0.09906170517206192,
+      "learning_rate": 6.194714894380838e-10,
+      "loss": 0.9965,
+      "step": 16434
+    },
+    {
+      "epoch": 0.9985418312169634,
+      "grad_norm": 0.2739318311214447,
+      "learning_rate": 5.727363035668009e-10,
+      "loss": 1.1276,
+      "step": 16435
+    },
+    {
+      "epoch": 0.9986025882495899,
+      "grad_norm": 0.3435087203979492,
+      "learning_rate": 5.278338563730056e-10,
+      "loss": 1.1132,
+      "step": 16436
+    },
+    {
+      "epoch": 0.9986633452822165,
+      "grad_norm": 0.0959831178188324,
+      "learning_rate": 4.847641494942767e-10,
+      "loss": 1.0131,
+      "step": 16437
+    },
+    {
+      "epoch": 0.998724102314843,
+      "grad_norm": 1.0217487812042236,
+      "learning_rate": 4.4352718451268204e-10,
+      "loss": 1.0946,
+      "step": 16438
+    },
+    {
+      "epoch": 0.9987848593474695,
+      "grad_norm": 0.15110178291797638,
+      "learning_rate": 4.041229629325738e-10,
+      "loss": 1.1102,
+      "step": 16439
+    },
+    {
+      "epoch": 0.998845616380096,
+      "grad_norm": 0.10901308059692383,
+      "learning_rate": 3.6655148620834413e-10,
+      "loss": 1.0399,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9989063734127225,
+      "grad_norm": 0.1671479493379593,
+      "learning_rate": 3.308127557111185e-10,
+      "loss": 1.0423,
+      "step": 16441
+    },
+    {
+      "epoch": 0.998967130445349,
+      "grad_norm": 0.23347000777721405,
+      "learning_rate": 2.9690677275651114e-10,
+      "loss": 1.1833,
+      "step": 16442
+    },
+    {
+      "epoch": 0.9990278874779756,
+      "grad_norm": 0.2965988218784332,
+      "learning_rate": 2.648335385824208e-10,
+      "loss": 1.1272,
+      "step": 16443
+    },
+    {
+      "epoch": 0.9990886445106021,
+      "grad_norm": 0.18708714842796326,
+      "learning_rate": 2.345930543601327e-10,
+      "loss": 1.1634,
+      "step": 16444
+    },
+    {
+      "epoch": 0.9991494015432286,
+      "grad_norm": 0.12446684390306473,
+      "learning_rate": 2.0618532121097213e-10,
+      "loss": 1.0628,
+      "step": 16445
+    },
+    {
+      "epoch": 0.9992101585758552,
+      "grad_norm": 0.17894360423088074,
+      "learning_rate": 1.7961034016189538e-10,
+      "loss": 1.203,
+      "step": 16446
+    },
+    {
+      "epoch": 0.9992709156084817,
+      "grad_norm": 0.24358119070529938,
+      "learning_rate": 1.5486811220100096e-10,
+      "loss": 1.102,
+      "step": 16447
+    },
+    {
+      "epoch": 0.9993316726411082,
+      "grad_norm": 0.0997956171631813,
+      "learning_rate": 1.3195863822201838e-10,
+      "loss": 1.061,
+      "step": 16448
+    },
+    {
+      "epoch": 0.9993924296737348,
+      "grad_norm": 0.09750929474830627,
+      "learning_rate": 1.1088191907426825e-10,
+      "loss": 1.0389,
+      "step": 16449
+    },
+    {
+      "epoch": 0.9994531867063613,
+      "grad_norm": 0.1518574357032776,
+      "learning_rate": 9.163795552935561e-11,
+      "loss": 1.1509,
+      "step": 16450
+    },
+    {
+      "epoch": 0.9995139437389878,
+      "grad_norm": 0.108207106590271,
+      "learning_rate": 7.422674828672093e-11,
+      "loss": 1.0395,
+      "step": 16451
+    },
+    {
+      "epoch": 0.9995747007716144,
+      "grad_norm": 0.1269640177488327,
+      "learning_rate": 5.864829798474247e-11,
+      "loss": 1.0621,
+      "step": 16452
+    },
+    {
+      "epoch": 0.9996354578042408,
+      "grad_norm": 0.10491638630628586,
+      "learning_rate": 4.490260520628731e-11,
+      "loss": 1.0369,
+      "step": 16453
+    },
+    {
+      "epoch": 0.9996962148368673,
+      "grad_norm": 0.11752914637327194,
+      "learning_rate": 3.298967043985357e-11,
+      "loss": 1.0706,
+      "step": 16454
+    },
+    {
+      "epoch": 0.9997569718694939,
+      "grad_norm": 0.14386291801929474,
+      "learning_rate": 2.2909494129530474e-11,
+      "loss": 1.1535,
+      "step": 16455
+    },
+    {
+      "epoch": 0.9998177289021204,
+      "grad_norm": 0.10731829702854156,
+      "learning_rate": 1.466207664724273e-11,
+      "loss": 1.0542,
+      "step": 16456
+    },
+    {
+      "epoch": 0.9998784859347469,
+      "grad_norm": 0.24370910227298737,
+      "learning_rate": 8.247418292750553e-12,
+      "loss": 1.0762,
+      "step": 16457
+    },
+    {
+      "epoch": 0.9999392429673735,
+      "grad_norm": 0.14225801825523376,
+      "learning_rate": 3.665519293649666e-12,
+      "loss": 1.0693,
+      "step": 16458
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.16426825523376465,
+      "learning_rate": 9.16379833126868e-13,
+      "loss": 1.1252,
+      "step": 16459
+    },
+    {
+      "epoch": 1.0,
+      "step": 16459,
+      "total_flos": 1.4485486633586629e+20,
+      "train_loss": 1.1091928682934797,
+      "train_runtime": 49071.5718,
+      "train_samples_per_second": 171.723,
+      "train_steps_per_second": 0.335
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 16459,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4485486633586629e+20,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}