diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9597 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.99835255354201,
+  "eval_steps": 500,
+  "global_step": 1365,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002196595277320154,
+      "grad_norm": 0.29563626646995544,
+      "learning_rate": 2.18978102189781e-06,
+      "loss": 1.3568,
+      "step": 1
+    },
+    {
+      "epoch": 0.004393190554640308,
+      "grad_norm": 0.2691873013973236,
+      "learning_rate": 4.37956204379562e-06,
+      "loss": 1.5243,
+      "step": 2
+    },
+    {
+      "epoch": 0.006589785831960461,
+      "grad_norm": 0.236099511384964,
+      "learning_rate": 6.56934306569343e-06,
+      "loss": 1.2765,
+      "step": 3
+    },
+    {
+      "epoch": 0.008786381109280615,
+      "grad_norm": 0.2405385822057724,
+      "learning_rate": 8.75912408759124e-06,
+      "loss": 1.1744,
+      "step": 4
+    },
+    {
+      "epoch": 0.010982976386600769,
+      "grad_norm": 0.2426607310771942,
+      "learning_rate": 1.0948905109489049e-05,
+      "loss": 1.2871,
+      "step": 5
+    },
+    {
+      "epoch": 0.013179571663920923,
+      "grad_norm": 0.26438114047050476,
+      "learning_rate": 1.313868613138686e-05,
+      "loss": 1.1193,
+      "step": 6
+    },
+    {
+      "epoch": 0.015376166941241077,
+      "grad_norm": 0.2810760736465454,
+      "learning_rate": 1.532846715328467e-05,
+      "loss": 1.098,
+      "step": 7
+    },
+    {
+      "epoch": 0.01757276221856123,
+      "grad_norm": 0.29785090684890747,
+      "learning_rate": 1.751824817518248e-05,
+      "loss": 1.1888,
+      "step": 8
+    },
+    {
+      "epoch": 0.019769357495881382,
+      "grad_norm": 0.2746748626232147,
+      "learning_rate": 1.9708029197080292e-05,
+      "loss": 1.4592,
+      "step": 9
+    },
+    {
+      "epoch": 0.021965952773201538,
+      "grad_norm": 0.27117863297462463,
+      "learning_rate": 2.1897810218978098e-05,
+      "loss": 1.1677,
+      "step": 10
+    },
+    {
+      "epoch": 0.02416254805052169,
+      "grad_norm": 0.21555709838867188,
+      "learning_rate": 2.408759124087591e-05,
+      "loss": 1.3955,
+      "step": 11
+    },
+    {
+      "epoch": 0.026359143327841845,
+      "grad_norm": 0.2435390055179596,
+      "learning_rate": 2.627737226277372e-05,
+      "loss": 1.2526,
+      "step": 12
+    },
+    {
+      "epoch": 0.028555738605161998,
+      "grad_norm": 0.2634373903274536,
+      "learning_rate": 2.846715328467153e-05,
+      "loss": 1.2278,
+      "step": 13
+    },
+    {
+      "epoch": 0.030752333882482153,
+      "grad_norm": 0.24729566276073456,
+      "learning_rate": 3.065693430656934e-05,
+      "loss": 1.25,
+      "step": 14
+    },
+    {
+      "epoch": 0.032948929159802305,
+      "grad_norm": 0.23541638255119324,
+      "learning_rate": 3.284671532846715e-05,
+      "loss": 1.3238,
+      "step": 15
+    },
+    {
+      "epoch": 0.03514552443712246,
+      "grad_norm": 0.2321559190750122,
+      "learning_rate": 3.503649635036496e-05,
+      "loss": 1.0762,
+      "step": 16
+    },
+    {
+      "epoch": 0.037342119714442616,
+      "grad_norm": 0.24050624668598175,
+      "learning_rate": 3.722627737226277e-05,
+      "loss": 1.2027,
+      "step": 17
+    },
+    {
+      "epoch": 0.039538714991762765,
+      "grad_norm": 0.2811952233314514,
+      "learning_rate": 3.9416058394160584e-05,
+      "loss": 1.4484,
+      "step": 18
+    },
+    {
+      "epoch": 0.04173531026908292,
+      "grad_norm": 0.21632897853851318,
+      "learning_rate": 4.160583941605839e-05,
+      "loss": 1.3258,
+      "step": 19
+    },
+    {
+      "epoch": 0.043931905546403076,
+      "grad_norm": 0.3140592873096466,
+      "learning_rate": 4.3795620437956196e-05,
+      "loss": 1.2775,
+      "step": 20
+    },
+    {
+      "epoch": 0.04612850082372323,
+      "grad_norm": 0.2542482018470764,
+      "learning_rate": 4.5985401459854016e-05,
+      "loss": 1.2865,
+      "step": 21
+    },
+    {
+      "epoch": 0.04832509610104338,
+      "grad_norm": 0.23981837928295135,
+      "learning_rate": 4.817518248175182e-05,
+      "loss": 1.2595,
+      "step": 22
+    },
+    {
+      "epoch": 0.050521691378363535,
+      "grad_norm": 0.3068813681602478,
+      "learning_rate": 5.036496350364963e-05,
+      "loss": 1.5723,
+      "step": 23
+    },
+    {
+      "epoch": 0.05271828665568369,
+      "grad_norm": 0.22749613225460052,
+      "learning_rate": 5.255474452554744e-05,
+      "loss": 1.3017,
+      "step": 24
+    },
+    {
+      "epoch": 0.054914881933003847,
+      "grad_norm": 0.24904906749725342,
+      "learning_rate": 5.4744525547445253e-05,
+      "loss": 1.1908,
+      "step": 25
+    },
+    {
+      "epoch": 0.057111477210323995,
+      "grad_norm": 0.26143336296081543,
+      "learning_rate": 5.693430656934306e-05,
+      "loss": 1.2672,
+      "step": 26
+    },
+    {
+      "epoch": 0.05930807248764415,
+      "grad_norm": 0.26132732629776,
+      "learning_rate": 5.912408759124087e-05,
+      "loss": 1.1928,
+      "step": 27
+    },
+    {
+      "epoch": 0.061504667764964306,
+      "grad_norm": 0.22761230170726776,
+      "learning_rate": 6.131386861313868e-05,
+      "loss": 1.0721,
+      "step": 28
+    },
+    {
+      "epoch": 0.06370126304228446,
+      "grad_norm": 0.2821810841560364,
+      "learning_rate": 6.350364963503648e-05,
+      "loss": 1.212,
+      "step": 29
+    },
+    {
+      "epoch": 0.06589785831960461,
+      "grad_norm": 0.25160491466522217,
+      "learning_rate": 6.56934306569343e-05,
+      "loss": 1.1603,
+      "step": 30
+    },
+    {
+      "epoch": 0.06809445359692477,
+      "grad_norm": 0.24630850553512573,
+      "learning_rate": 6.788321167883211e-05,
+      "loss": 1.2326,
+      "step": 31
+    },
+    {
+      "epoch": 0.07029104887424492,
+      "grad_norm": 0.2598794996738434,
+      "learning_rate": 7.007299270072992e-05,
+      "loss": 1.4148,
+      "step": 32
+    },
+    {
+      "epoch": 0.07248764415156507,
+      "grad_norm": 0.22428004443645477,
+      "learning_rate": 7.226277372262774e-05,
+      "loss": 1.654,
+      "step": 33
+    },
+    {
+      "epoch": 0.07468423942888523,
+      "grad_norm": 0.2537097930908203,
+      "learning_rate": 7.445255474452554e-05,
+      "loss": 1.1965,
+      "step": 34
+    },
+    {
+      "epoch": 0.07688083470620538,
+      "grad_norm": 0.28156256675720215,
+      "learning_rate": 7.664233576642336e-05,
+      "loss": 1.3038,
+      "step": 35
+    },
+    {
+      "epoch": 0.07907742998352553,
+      "grad_norm": 0.29285484552383423,
+      "learning_rate": 7.883211678832117e-05,
+      "loss": 1.3168,
+      "step": 36
+    },
+    {
+      "epoch": 0.08127402526084569,
+      "grad_norm": 0.22810137271881104,
+      "learning_rate": 8.102189781021897e-05,
+      "loss": 1.2865,
+      "step": 37
+    },
+    {
+      "epoch": 0.08347062053816584,
+      "grad_norm": 0.26377180218696594,
+      "learning_rate": 8.321167883211678e-05,
+      "loss": 1.3118,
+      "step": 38
+    },
+    {
+      "epoch": 0.085667215815486,
+      "grad_norm": 0.2753414809703827,
+      "learning_rate": 8.540145985401459e-05,
+      "loss": 1.4673,
+      "step": 39
+    },
+    {
+      "epoch": 0.08786381109280615,
+      "grad_norm": 0.2623012363910675,
+      "learning_rate": 8.759124087591239e-05,
+      "loss": 1.1949,
+      "step": 40
+    },
+    {
+      "epoch": 0.0900604063701263,
+      "grad_norm": 0.23522089421749115,
+      "learning_rate": 8.978102189781021e-05,
+      "loss": 1.3,
+      "step": 41
+    },
+    {
+      "epoch": 0.09225700164744646,
+      "grad_norm": 0.3461747169494629,
+      "learning_rate": 9.197080291970803e-05,
+      "loss": 1.5409,
+      "step": 42
+    },
+    {
+      "epoch": 0.09445359692476661,
+      "grad_norm": 0.2799449563026428,
+      "learning_rate": 9.416058394160584e-05,
+      "loss": 1.2624,
+      "step": 43
+    },
+    {
+      "epoch": 0.09665019220208676,
+      "grad_norm": 0.2374962866306305,
+      "learning_rate": 9.635036496350364e-05,
+      "loss": 1.3386,
+      "step": 44
+    },
+    {
+      "epoch": 0.09884678747940692,
+      "grad_norm": 0.2388330101966858,
+      "learning_rate": 9.854014598540145e-05,
+      "loss": 1.3671,
+      "step": 45
+    },
+    {
+      "epoch": 0.10104338275672707,
+      "grad_norm": 0.23381935060024261,
+      "learning_rate": 0.00010072992700729926,
+      "loss": 1.2581,
+      "step": 46
+    },
+    {
+      "epoch": 0.10323997803404723,
+      "grad_norm": 0.2766358256340027,
+      "learning_rate": 0.00010291970802919708,
+      "loss": 1.3108,
+      "step": 47
+    },
+    {
+      "epoch": 0.10543657331136738,
+      "grad_norm": 0.23038944602012634,
+      "learning_rate": 0.00010510948905109488,
+      "loss": 1.3206,
+      "step": 48
+    },
+    {
+      "epoch": 0.10763316858868753,
+      "grad_norm": 0.24806389212608337,
+      "learning_rate": 0.00010729927007299269,
+      "loss": 1.5171,
+      "step": 49
+    },
+    {
+      "epoch": 0.10982976386600769,
+      "grad_norm": 0.25186586380004883,
+      "learning_rate": 0.00010948905109489051,
+      "loss": 1.3482,
+      "step": 50
+    },
+    {
+      "epoch": 0.11202635914332784,
+      "grad_norm": 0.2518808841705322,
+      "learning_rate": 0.00011167883211678831,
+      "loss": 1.3216,
+      "step": 51
+    },
+    {
+      "epoch": 0.11422295442064799,
+      "grad_norm": 0.22794225811958313,
+      "learning_rate": 0.00011386861313868612,
+      "loss": 1.1407,
+      "step": 52
+    },
+    {
+      "epoch": 0.11641954969796815,
+      "grad_norm": 0.229314923286438,
+      "learning_rate": 0.00011605839416058394,
+      "loss": 1.3288,
+      "step": 53
+    },
+    {
+      "epoch": 0.1186161449752883,
+      "grad_norm": 0.25637155771255493,
+      "learning_rate": 0.00011824817518248174,
+      "loss": 1.0873,
+      "step": 54
+    },
+    {
+      "epoch": 0.12081274025260846,
+      "grad_norm": 0.21607248485088348,
+      "learning_rate": 0.00012043795620437955,
+      "loss": 1.2132,
+      "step": 55
+    },
+    {
+      "epoch": 0.12300933552992861,
+      "grad_norm": 0.30300867557525635,
+      "learning_rate": 0.00012262773722627736,
+      "loss": 1.4721,
+      "step": 56
+    },
+    {
+      "epoch": 0.12520593080724876,
+      "grad_norm": 0.22752255201339722,
+      "learning_rate": 0.00012481751824817516,
+      "loss": 1.1253,
+      "step": 57
+    },
+    {
+      "epoch": 0.12740252608456892,
+      "grad_norm": 0.22841185331344604,
+      "learning_rate": 0.00012700729927007297,
+      "loss": 1.3185,
+      "step": 58
+    },
+    {
+      "epoch": 0.12959912136188906,
+      "grad_norm": 0.21192136406898499,
+      "learning_rate": 0.00012919708029197077,
+      "loss": 1.1176,
+      "step": 59
+    },
+    {
+      "epoch": 0.13179571663920922,
+      "grad_norm": 0.19719888269901276,
+      "learning_rate": 0.0001313868613138686,
+      "loss": 1.1665,
+      "step": 60
+    },
+    {
+      "epoch": 0.13399231191652938,
+      "grad_norm": 0.20132644474506378,
+      "learning_rate": 0.00013357664233576641,
+      "loss": 1.2747,
+      "step": 61
+    },
+    {
+      "epoch": 0.13618890719384955,
+      "grad_norm": 0.22799596190452576,
+      "learning_rate": 0.00013576642335766422,
+      "loss": 1.1192,
+      "step": 62
+    },
+    {
+      "epoch": 0.13838550247116968,
+      "grad_norm": 0.20367981493473053,
+      "learning_rate": 0.00013795620437956203,
+      "loss": 1.0725,
+      "step": 63
+    },
+    {
+      "epoch": 0.14058209774848984,
+      "grad_norm": 0.25830328464508057,
+      "learning_rate": 0.00014014598540145983,
+      "loss": 1.3098,
+      "step": 64
+    },
+    {
+      "epoch": 0.14277869302581,
+      "grad_norm": 0.18191899359226227,
+      "learning_rate": 0.00014233576642335764,
+      "loss": 1.1321,
+      "step": 65
+    },
+    {
+      "epoch": 0.14497528830313014,
+      "grad_norm": 0.22679124772548676,
+      "learning_rate": 0.00014452554744525547,
+      "loss": 0.9727,
+      "step": 66
+    },
+    {
+      "epoch": 0.1471718835804503,
+      "grad_norm": 0.2176261991262436,
+      "learning_rate": 0.00014671532846715328,
+      "loss": 1.2184,
+      "step": 67
+    },
+    {
+      "epoch": 0.14936847885777046,
+      "grad_norm": 0.21768714487552643,
+      "learning_rate": 0.00014890510948905108,
+      "loss": 1.0603,
+      "step": 68
+    },
+    {
+      "epoch": 0.1515650741350906,
+      "grad_norm": 0.26472052931785583,
+      "learning_rate": 0.00015109489051094892,
+      "loss": 1.3187,
+      "step": 69
+    },
+    {
+      "epoch": 0.15376166941241076,
+      "grad_norm": 0.25393205881118774,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 1.155,
+      "step": 70
+    },
+    {
+      "epoch": 0.15595826468973092,
+      "grad_norm": 0.19356046617031097,
+      "learning_rate": 0.00015547445255474453,
+      "loss": 1.2658,
+      "step": 71
+    },
+    {
+      "epoch": 0.15815485996705106,
+      "grad_norm": 0.19435086846351624,
+      "learning_rate": 0.00015766423357664234,
+      "loss": 1.1707,
+      "step": 72
+    },
+    {
+      "epoch": 0.16035145524437122,
+      "grad_norm": 0.20181354880332947,
+      "learning_rate": 0.00015985401459854014,
+      "loss": 1.361,
+      "step": 73
+    },
+    {
+      "epoch": 0.16254805052169138,
+      "grad_norm": 0.16649247705936432,
+      "learning_rate": 0.00016204379562043795,
+      "loss": 1.2512,
+      "step": 74
+    },
+    {
+      "epoch": 0.16474464579901152,
+      "grad_norm": 0.23832470178604126,
+      "learning_rate": 0.00016423357664233575,
+      "loss": 1.1152,
+      "step": 75
+    },
+    {
+      "epoch": 0.16694124107633168,
+      "grad_norm": 0.19680383801460266,
+      "learning_rate": 0.00016642335766423356,
+      "loss": 1.1779,
+      "step": 76
+    },
+    {
+      "epoch": 0.16913783635365184,
+      "grad_norm": 0.17287391424179077,
+      "learning_rate": 0.00016861313868613137,
+      "loss": 1.0942,
+      "step": 77
+    },
+    {
+      "epoch": 0.171334431630972,
+      "grad_norm": 0.18865327537059784,
+      "learning_rate": 0.00017080291970802917,
+      "loss": 0.8755,
+      "step": 78
+    },
+    {
+      "epoch": 0.17353102690829214,
+      "grad_norm": 0.1676090806722641,
+      "learning_rate": 0.00017299270072992698,
+      "loss": 1.1024,
+      "step": 79
+    },
+    {
+      "epoch": 0.1757276221856123,
+      "grad_norm": 0.18176300823688507,
+      "learning_rate": 0.00017518248175182478,
+      "loss": 1.2364,
+      "step": 80
+    },
+    {
+      "epoch": 0.17792421746293247,
+      "grad_norm": 0.20020973682403564,
+      "learning_rate": 0.00017737226277372262,
+      "loss": 1.132,
+      "step": 81
+    },
+    {
+      "epoch": 0.1801208127402526,
+      "grad_norm": 0.18806861340999603,
+      "learning_rate": 0.00017956204379562042,
+      "loss": 1.1437,
+      "step": 82
+    },
+    {
+      "epoch": 0.18231740801757276,
+      "grad_norm": 0.16758619248867035,
+      "learning_rate": 0.00018175182481751826,
+      "loss": 1.0396,
+      "step": 83
+    },
+    {
+      "epoch": 0.18451400329489293,
+      "grad_norm": 0.19573086500167847,
+      "learning_rate": 0.00018394160583941606,
+      "loss": 1.2229,
+      "step": 84
+    },
+    {
+      "epoch": 0.18671059857221306,
+      "grad_norm": 0.4174942672252655,
+      "learning_rate": 0.00018613138686131387,
+      "loss": 1.3434,
+      "step": 85
+    },
+    {
+      "epoch": 0.18890719384953322,
+      "grad_norm": 0.19190342724323273,
+      "learning_rate": 0.00018832116788321167,
+      "loss": 1.469,
+      "step": 86
+    },
+    {
+      "epoch": 0.19110378912685339,
+      "grad_norm": 0.17672614753246307,
+      "learning_rate": 0.00019051094890510948,
+      "loss": 1.2067,
+      "step": 87
+    },
+    {
+      "epoch": 0.19330038440417352,
+      "grad_norm": 0.17822261154651642,
+      "learning_rate": 0.0001927007299270073,
+      "loss": 1.2395,
+      "step": 88
+    },
+    {
+      "epoch": 0.19549697968149368,
+      "grad_norm": 0.19186727702617645,
+      "learning_rate": 0.0001948905109489051,
+      "loss": 0.9998,
+      "step": 89
+    },
+    {
+      "epoch": 0.19769357495881384,
+      "grad_norm": 0.19049587845802307,
+      "learning_rate": 0.0001970802919708029,
+      "loss": 1.3747,
+      "step": 90
+    },
+    {
+      "epoch": 0.19989017023613398,
+      "grad_norm": 0.1485290229320526,
+      "learning_rate": 0.0001992700729927007,
+      "loss": 1.086,
+      "step": 91
+    },
+    {
+      "epoch": 0.20208676551345414,
+      "grad_norm": 0.15687604248523712,
+      "learning_rate": 0.0002014598540145985,
+      "loss": 1.0398,
+      "step": 92
+    },
+    {
+      "epoch": 0.2042833607907743,
+      "grad_norm": 0.198198139667511,
+      "learning_rate": 0.00020364963503649632,
+      "loss": 1.1161,
+      "step": 93
+    },
+    {
+      "epoch": 0.20647995606809447,
+      "grad_norm": 0.19159619510173798,
+      "learning_rate": 0.00020583941605839415,
+      "loss": 1.3116,
+      "step": 94
+    },
+    {
+      "epoch": 0.2086765513454146,
+      "grad_norm": 0.18750174343585968,
+      "learning_rate": 0.00020802919708029196,
+      "loss": 1.0596,
+      "step": 95
+    },
+    {
+      "epoch": 0.21087314662273476,
+      "grad_norm": 0.18936151266098022,
+      "learning_rate": 0.00021021897810218976,
+      "loss": 1.2446,
+      "step": 96
+    },
+    {
+      "epoch": 0.21306974190005493,
+      "grad_norm": 0.2053549438714981,
+      "learning_rate": 0.00021240875912408757,
+      "loss": 1.2587,
+      "step": 97
+    },
+    {
+      "epoch": 0.21526633717737506,
+      "grad_norm": 0.19101539254188538,
+      "learning_rate": 0.00021459854014598537,
+      "loss": 1.1655,
+      "step": 98
+    },
+    {
+      "epoch": 0.21746293245469522,
+      "grad_norm": 0.1533312350511551,
+      "learning_rate": 0.0002167883211678832,
+      "loss": 1.1611,
+      "step": 99
+    },
+    {
+      "epoch": 0.21965952773201539,
+      "grad_norm": 0.22746139764785767,
+      "learning_rate": 0.00021897810218978101,
+      "loss": 1.239,
+      "step": 100
+    },
+    {
+      "epoch": 0.22185612300933552,
+      "grad_norm": 0.22471238672733307,
+      "learning_rate": 0.00022116788321167882,
+      "loss": 1.3873,
+      "step": 101
+    },
+    {
+      "epoch": 0.22405271828665568,
+      "grad_norm": 0.2001270055770874,
+      "learning_rate": 0.00022335766423357663,
+      "loss": 1.2151,
+      "step": 102
+    },
+    {
+      "epoch": 0.22624931356397585,
+      "grad_norm": 0.14863215386867523,
+      "learning_rate": 0.00022554744525547443,
+      "loss": 1.1842,
+      "step": 103
+    },
+    {
+      "epoch": 0.22844590884129598,
+      "grad_norm": 0.17552022635936737,
+      "learning_rate": 0.00022773722627737224,
+      "loss": 1.0759,
+      "step": 104
+    },
+    {
+      "epoch": 0.23064250411861614,
+      "grad_norm": 0.18896892666816711,
+      "learning_rate": 0.00022992700729927004,
+      "loss": 1.1398,
+      "step": 105
+    },
+    {
+      "epoch": 0.2328390993959363,
+      "grad_norm": 0.1679116189479828,
+      "learning_rate": 0.00023211678832116788,
+      "loss": 1.2333,
+      "step": 106
+    },
+    {
+      "epoch": 0.23503569467325644,
+      "grad_norm": 0.22668372094631195,
+      "learning_rate": 0.00023430656934306568,
+      "loss": 1.3287,
+      "step": 107
+    },
+    {
+      "epoch": 0.2372322899505766,
+      "grad_norm": 0.17289626598358154,
+      "learning_rate": 0.0002364963503649635,
+      "loss": 1.0318,
+      "step": 108
+    },
+    {
+      "epoch": 0.23942888522789676,
+      "grad_norm": 0.21098165214061737,
+      "learning_rate": 0.0002386861313868613,
+      "loss": 1.1363,
+      "step": 109
+    },
+    {
+      "epoch": 0.24162548050521693,
+      "grad_norm": 0.16004520654678345,
+      "learning_rate": 0.0002408759124087591,
+      "loss": 1.0748,
+      "step": 110
+    },
+    {
+      "epoch": 0.24382207578253706,
+      "grad_norm": 0.1771387904882431,
+      "learning_rate": 0.0002430656934306569,
+      "loss": 1.1566,
+      "step": 111
+    },
+    {
+      "epoch": 0.24601867105985722,
+      "grad_norm": 0.19743089377880096,
+      "learning_rate": 0.0002452554744525547,
+      "loss": 1.1383,
+      "step": 112
+    },
+    {
+      "epoch": 0.2482152663371774,
+      "grad_norm": 0.1913851499557495,
+      "learning_rate": 0.0002474452554744525,
+      "loss": 1.1035,
+      "step": 113
+    },
+    {
+      "epoch": 0.2504118616144975,
+      "grad_norm": 0.20614024996757507,
+      "learning_rate": 0.0002496350364963503,
+      "loss": 1.0944,
+      "step": 114
+    },
+    {
+      "epoch": 0.25260845689181766,
+      "grad_norm": 0.20137476921081543,
+      "learning_rate": 0.00025182481751824813,
+      "loss": 1.1802,
+      "step": 115
+    },
+    {
+      "epoch": 0.25480505216913785,
+      "grad_norm": 0.17677545547485352,
+      "learning_rate": 0.00025401459854014594,
+      "loss": 1.2076,
+      "step": 116
+    },
+    {
+      "epoch": 0.257001647446458,
+      "grad_norm": 0.19805948436260223,
+      "learning_rate": 0.00025620437956204374,
+      "loss": 1.1395,
+      "step": 117
+    },
+    {
+      "epoch": 0.2591982427237781,
+      "grad_norm": 0.14304672181606293,
+      "learning_rate": 0.00025839416058394155,
+      "loss": 1.04,
+      "step": 118
+    },
+    {
+      "epoch": 0.2613948380010983,
+      "grad_norm": 0.16515874862670898,
+      "learning_rate": 0.0002605839416058394,
+      "loss": 1.0161,
+      "step": 119
+    },
+    {
+      "epoch": 0.26359143327841844,
+      "grad_norm": 0.19261552393436432,
+      "learning_rate": 0.0002627737226277372,
+      "loss": 0.9795,
+      "step": 120
+    },
+    {
+      "epoch": 0.26578802855573863,
+      "grad_norm": 0.24200361967086792,
+      "learning_rate": 0.000264963503649635,
+      "loss": 1.0562,
+      "step": 121
+    },
+    {
+      "epoch": 0.26798462383305877,
+      "grad_norm": 0.14347536861896515,
+      "learning_rate": 0.00026715328467153283,
+      "loss": 1.0031,
+      "step": 122
+    },
+    {
+      "epoch": 0.2701812191103789,
+      "grad_norm": 0.19266776740550995,
+      "learning_rate": 0.00026934306569343063,
+      "loss": 1.0814,
+      "step": 123
+    },
+    {
+      "epoch": 0.2723778143876991,
+      "grad_norm": 0.17406077682971954,
+      "learning_rate": 0.00027153284671532844,
+      "loss": 1.0681,
+      "step": 124
+    },
+    {
+      "epoch": 0.2745744096650192,
+      "grad_norm": 0.15546488761901855,
+      "learning_rate": 0.00027372262773722625,
+      "loss": 1.1663,
+      "step": 125
+    },
+    {
+      "epoch": 0.27677100494233936,
+      "grad_norm": 0.17485150694847107,
+      "learning_rate": 0.00027591240875912405,
+      "loss": 0.8728,
+      "step": 126
+    },
+    {
+      "epoch": 0.27896760021965955,
+      "grad_norm": 0.18409255146980286,
+      "learning_rate": 0.00027810218978102186,
+      "loss": 1.0806,
+      "step": 127
+    },
+    {
+      "epoch": 0.2811641954969797,
+      "grad_norm": 0.15314918756484985,
+      "learning_rate": 0.00028029197080291966,
+      "loss": 1.375,
+      "step": 128
+    },
+    {
+      "epoch": 0.2833607907742998,
+      "grad_norm": 0.25007015466690063,
+      "learning_rate": 0.00028248175182481747,
+      "loss": 1.2101,
+      "step": 129
+    },
+    {
+      "epoch": 0.28555738605162,
+      "grad_norm": 0.4458443224430084,
+      "learning_rate": 0.0002846715328467153,
+      "loss": 1.1376,
+      "step": 130
+    },
+    {
+      "epoch": 0.28775398132894014,
+      "grad_norm": 0.17013037204742432,
+      "learning_rate": 0.00028686131386861314,
+      "loss": 1.0517,
+      "step": 131
+    },
+    {
+      "epoch": 0.2899505766062603,
+      "grad_norm": 0.16901728510856628,
+      "learning_rate": 0.00028905109489051094,
+      "loss": 0.9294,
+      "step": 132
+    },
+    {
+      "epoch": 0.29214717188358047,
+      "grad_norm": 0.1529814898967743,
+      "learning_rate": 0.00029124087591240875,
+      "loss": 0.8575,
+      "step": 133
+    },
+    {
+      "epoch": 0.2943437671609006,
+      "grad_norm": 0.16957421600818634,
+      "learning_rate": 0.00029343065693430656,
+      "loss": 1.007,
+      "step": 134
+    },
+    {
+      "epoch": 0.29654036243822074,
+      "grad_norm": 0.14046508073806763,
+      "learning_rate": 0.00029562043795620436,
+      "loss": 1.0019,
+      "step": 135
+    },
+    {
+      "epoch": 0.29873695771554093,
+      "grad_norm": 0.18718869984149933,
+      "learning_rate": 0.00029781021897810217,
+      "loss": 1.2505,
+      "step": 136
+    },
+    {
+      "epoch": 0.30093355299286106,
+      "grad_norm": 0.19814763963222504,
+      "learning_rate": 0.0003,
+      "loss": 1.1649,
+      "step": 137
+    },
+    {
+      "epoch": 0.3031301482701812,
+      "grad_norm": 0.15006978809833527,
+      "learning_rate": 0.00029975570032573286,
+      "loss": 1.1154,
+      "step": 138
+    },
+    {
+      "epoch": 0.3053267435475014,
+      "grad_norm": 0.14791452884674072,
+      "learning_rate": 0.00029951140065146575,
+      "loss": 1.0372,
+      "step": 139
+    },
+    {
+      "epoch": 0.3075233388248215,
+      "grad_norm": 0.15797799825668335,
+      "learning_rate": 0.0002992671009771987,
+      "loss": 1.1038,
+      "step": 140
+    },
+    {
+      "epoch": 0.30971993410214166,
+      "grad_norm": 0.24310439825057983,
+      "learning_rate": 0.0002990228013029316,
+      "loss": 1.1571,
+      "step": 141
+    },
+    {
+      "epoch": 0.31191652937946185,
+      "grad_norm": 0.18441380560398102,
+      "learning_rate": 0.0002987785016286645,
+      "loss": 1.2487,
+      "step": 142
+    },
+    {
+      "epoch": 0.314113124656782,
+      "grad_norm": 0.1581963747739792,
+      "learning_rate": 0.00029853420195439736,
+      "loss": 1.1036,
+      "step": 143
+    },
+    {
+      "epoch": 0.3163097199341021,
+      "grad_norm": 0.17542096972465515,
+      "learning_rate": 0.00029828990228013025,
+      "loss": 1.3265,
+      "step": 144
+    },
+    {
+      "epoch": 0.3185063152114223,
+      "grad_norm": 0.20331527292728424,
+      "learning_rate": 0.00029804560260586314,
+      "loss": 1.1321,
+      "step": 145
+    },
+    {
+      "epoch": 0.32070291048874244,
+      "grad_norm": 0.17392964661121368,
+      "learning_rate": 0.0002978013029315961,
+      "loss": 1.0601,
+      "step": 146
+    },
+    {
+      "epoch": 0.3228995057660626,
+      "grad_norm": 0.20815367996692657,
+      "learning_rate": 0.000297557003257329,
+      "loss": 1.0112,
+      "step": 147
+    },
+    {
+      "epoch": 0.32509610104338277,
+      "grad_norm": 1.4079077243804932,
+      "learning_rate": 0.00029731270358306186,
+      "loss": 0.9275,
+      "step": 148
+    },
+    {
+      "epoch": 0.3272926963207029,
+      "grad_norm": 0.16588929295539856,
+      "learning_rate": 0.00029706840390879475,
+      "loss": 1.0738,
+      "step": 149
+    },
+    {
+      "epoch": 0.32948929159802304,
+      "grad_norm": 0.12843909859657288,
+      "learning_rate": 0.00029682410423452764,
+      "loss": 1.3013,
+      "step": 150
+    },
+    {
+      "epoch": 0.3316858868753432,
+      "grad_norm": 0.14631293714046478,
+      "learning_rate": 0.0002965798045602606,
+      "loss": 1.1527,
+      "step": 151
+    },
+    {
+      "epoch": 0.33388248215266336,
+      "grad_norm": 0.2243501842021942,
+      "learning_rate": 0.0002963355048859935,
+      "loss": 1.2525,
+      "step": 152
+    },
+    {
+      "epoch": 0.33607907742998355,
+      "grad_norm": 0.18019184470176697,
+      "learning_rate": 0.00029609120521172636,
+      "loss": 0.9711,
+      "step": 153
+    },
+    {
+      "epoch": 0.3382756727073037,
+      "grad_norm": 0.16834694147109985,
+      "learning_rate": 0.00029584690553745925,
+      "loss": 1.0335,
+      "step": 154
+    },
+    {
+      "epoch": 0.3404722679846238,
+      "grad_norm": 0.18543079495429993,
+      "learning_rate": 0.00029560260586319214,
+      "loss": 1.0006,
+      "step": 155
+    },
+    {
+      "epoch": 0.342668863261944,
+      "grad_norm": 0.22068579494953156,
+      "learning_rate": 0.00029535830618892503,
+      "loss": 1.0242,
+      "step": 156
+    },
+    {
+      "epoch": 0.34486545853926415,
+      "grad_norm": 0.21913889050483704,
+      "learning_rate": 0.000295114006514658,
+      "loss": 1.0852,
+      "step": 157
+    },
+    {
+      "epoch": 0.3470620538165843,
+      "grad_norm": 0.1590992957353592,
+      "learning_rate": 0.00029486970684039086,
+      "loss": 1.0078,
+      "step": 158
+    },
+    {
+      "epoch": 0.34925864909390447,
+      "grad_norm": 0.16164982318878174,
+      "learning_rate": 0.00029462540716612375,
+      "loss": 1.1735,
+      "step": 159
+    },
+    {
+      "epoch": 0.3514552443712246,
+      "grad_norm": 0.15709275007247925,
+      "learning_rate": 0.00029438110749185664,
+      "loss": 1.0023,
+      "step": 160
+    },
+    {
+      "epoch": 0.35365183964854474,
+      "grad_norm": 0.1895107626914978,
+      "learning_rate": 0.00029413680781758953,
+      "loss": 1.0987,
+      "step": 161
+    },
+    {
+      "epoch": 0.35584843492586493,
+      "grad_norm": 0.14722630381584167,
+      "learning_rate": 0.0002938925081433224,
+      "loss": 0.9213,
+      "step": 162
+    },
+    {
+      "epoch": 0.35804503020318507,
+      "grad_norm": 0.28739166259765625,
+      "learning_rate": 0.00029364820846905536,
+      "loss": 0.732,
+      "step": 163
+    },
+    {
+      "epoch": 0.3602416254805052,
+      "grad_norm": 0.16817232966423035,
+      "learning_rate": 0.00029340390879478825,
+      "loss": 0.983,
+      "step": 164
+    },
+    {
+      "epoch": 0.3624382207578254,
+      "grad_norm": 0.16354228556156158,
+      "learning_rate": 0.00029315960912052114,
+      "loss": 1.1507,
+      "step": 165
+    },
+    {
+      "epoch": 0.3646348160351455,
+      "grad_norm": 0.24635201692581177,
+      "learning_rate": 0.00029291530944625403,
+      "loss": 1.0987,
+      "step": 166
+    },
+    {
+      "epoch": 0.36683141131246566,
+      "grad_norm": 0.139459028840065,
+      "learning_rate": 0.0002926710097719869,
+      "loss": 0.9402,
+      "step": 167
+    },
+    {
+      "epoch": 0.36902800658978585,
+      "grad_norm": 0.23190955817699432,
+      "learning_rate": 0.00029242671009771986,
+      "loss": 1.0623,
+      "step": 168
+    },
+    {
+      "epoch": 0.371224601867106,
+      "grad_norm": 0.2316746711730957,
+      "learning_rate": 0.00029218241042345275,
+      "loss": 1.1154,
+      "step": 169
+    },
+    {
+      "epoch": 0.3734211971444261,
+      "grad_norm": 0.16537421941757202,
+      "learning_rate": 0.00029193811074918564,
+      "loss": 1.1401,
+      "step": 170
+    },
+    {
+      "epoch": 0.3756177924217463,
+      "grad_norm": 0.19183523952960968,
+      "learning_rate": 0.00029169381107491853,
+      "loss": 0.9293,
+      "step": 171
+    },
+    {
+      "epoch": 0.37781438769906645,
+      "grad_norm": 0.15889643132686615,
+      "learning_rate": 0.0002914495114006514,
+      "loss": 0.8786,
+      "step": 172
+    },
+    {
+      "epoch": 0.3800109829763866,
+      "grad_norm": 0.14010745286941528,
+      "learning_rate": 0.0002912052117263843,
+      "loss": 1.0453,
+      "step": 173
+    },
+    {
+      "epoch": 0.38220757825370677,
+      "grad_norm": 0.2037530541419983,
+      "learning_rate": 0.00029096091205211725,
+      "loss": 1.3083,
+      "step": 174
+    },
+    {
+      "epoch": 0.3844041735310269,
+      "grad_norm": 0.12781217694282532,
+      "learning_rate": 0.00029071661237785014,
+      "loss": 0.8071,
+      "step": 175
+    },
+    {
+      "epoch": 0.38660076880834704,
+      "grad_norm": 0.12923461198806763,
+      "learning_rate": 0.00029047231270358303,
+      "loss": 1.1784,
+      "step": 176
+    },
+    {
+      "epoch": 0.38879736408566723,
+      "grad_norm": 0.16285191476345062,
+      "learning_rate": 0.0002902280130293159,
+      "loss": 0.743,
+      "step": 177
+    },
+    {
+      "epoch": 0.39099395936298736,
+      "grad_norm": 0.1604568362236023,
+      "learning_rate": 0.0002899837133550488,
+      "loss": 0.9997,
+      "step": 178
+    },
+    {
+      "epoch": 0.3931905546403075,
+      "grad_norm": 0.20771288871765137,
+      "learning_rate": 0.00028973941368078175,
+      "loss": 1.0923,
+      "step": 179
+    },
+    {
+      "epoch": 0.3953871499176277,
+      "grad_norm": 0.1615612506866455,
+      "learning_rate": 0.00028949511400651464,
+      "loss": 0.962,
+      "step": 180
+    },
+    {
+      "epoch": 0.3975837451949478,
+      "grad_norm": 0.1963646560907364,
+      "learning_rate": 0.00028925081433224753,
+      "loss": 0.9493,
+      "step": 181
+    },
+    {
+      "epoch": 0.39978034047226796,
+      "grad_norm": 0.2578367292881012,
+      "learning_rate": 0.0002890065146579804,
+      "loss": 0.8423,
+      "step": 182
+    },
+    {
+      "epoch": 0.40197693574958815,
+      "grad_norm": 0.152908056974411,
+      "learning_rate": 0.0002887622149837133,
+      "loss": 1.0492,
+      "step": 183
+    },
+    {
+      "epoch": 0.4041735310269083,
+      "grad_norm": 0.14469197392463684,
+      "learning_rate": 0.0002885179153094462,
+      "loss": 1.3497,
+      "step": 184
+    },
+    {
+      "epoch": 0.4063701263042284,
+      "grad_norm": 0.14025677740573883,
+      "learning_rate": 0.00028827361563517914,
+      "loss": 1.0967,
+      "step": 185
+    },
+    {
+      "epoch": 0.4085667215815486,
+      "grad_norm": 0.15028126537799835,
+      "learning_rate": 0.00028802931596091203,
+      "loss": 1.0097,
+      "step": 186
+    },
+    {
+      "epoch": 0.41076331685886874,
+      "grad_norm": 0.17144006490707397,
+      "learning_rate": 0.0002877850162866449,
+      "loss": 0.8305,
+      "step": 187
+    },
+    {
+      "epoch": 0.41295991213618893,
+      "grad_norm": 0.1907787322998047,
+      "learning_rate": 0.0002875407166123778,
+      "loss": 1.0635,
+      "step": 188
+    },
+    {
+      "epoch": 0.41515650741350907,
+      "grad_norm": 0.18187835812568665,
+      "learning_rate": 0.0002872964169381107,
+      "loss": 1.1003,
+      "step": 189
+    },
+    {
+      "epoch": 0.4173531026908292,
+      "grad_norm": 0.12759219110012054,
+      "learning_rate": 0.00028705211726384364,
+      "loss": 0.9768,
+      "step": 190
+    },
+    {
+      "epoch": 0.4195496979681494,
+      "grad_norm": 0.18710525333881378,
+      "learning_rate": 0.00028680781758957653,
+      "loss": 0.9866,
+      "step": 191
+    },
+    {
+      "epoch": 0.42174629324546953,
+      "grad_norm": 0.13361667096614838,
+      "learning_rate": 0.0002865635179153094,
+      "loss": 1.0859,
+      "step": 192
+    },
+    {
+      "epoch": 0.42394288852278966,
+      "grad_norm": 0.16906079649925232,
+      "learning_rate": 0.0002863192182410423,
+      "loss": 1.0412,
+      "step": 193
+    },
+    {
+      "epoch": 0.42613948380010985,
+      "grad_norm": 0.14500804245471954,
+      "learning_rate": 0.0002860749185667752,
+      "loss": 1.1073,
+      "step": 194
+    },
+    {
+      "epoch": 0.42833607907743,
+      "grad_norm": 0.15757642686367035,
+      "learning_rate": 0.0002858306188925081,
+      "loss": 1.2531,
+      "step": 195
+    },
+    {
+      "epoch": 0.4305326743547501,
+      "grad_norm": 0.1713370531797409,
+      "learning_rate": 0.00028558631921824103,
+      "loss": 1.4524,
+      "step": 196
+    },
+    {
+      "epoch": 0.4327292696320703,
+      "grad_norm": 0.13056722283363342,
+      "learning_rate": 0.0002853420195439739,
+      "loss": 0.884,
+      "step": 197
+    },
+    {
+      "epoch": 0.43492586490939045,
+      "grad_norm": 0.1722119301557541,
+      "learning_rate": 0.0002850977198697068,
+      "loss": 0.8845,
+      "step": 198
+    },
+    {
+      "epoch": 0.4371224601867106,
+      "grad_norm": 0.20874305069446564,
+      "learning_rate": 0.0002848534201954397,
+      "loss": 0.9916,
+      "step": 199
+    },
+    {
+      "epoch": 0.43931905546403077,
+      "grad_norm": 0.11862941086292267,
+      "learning_rate": 0.0002846091205211726,
+      "loss": 1.0668,
+      "step": 200
+    },
+    {
+      "epoch": 0.4415156507413509,
+      "grad_norm": 0.2141619473695755,
+      "learning_rate": 0.0002843648208469055,
+      "loss": 0.889,
+      "step": 201
+    },
+    {
+      "epoch": 0.44371224601867104,
+      "grad_norm": 0.12272541224956512,
+      "learning_rate": 0.0002841205211726384,
+      "loss": 0.8919,
+      "step": 202
+    },
+    {
+      "epoch": 0.44590884129599123,
+      "grad_norm": 0.12215017527341843,
+      "learning_rate": 0.0002838762214983713,
+      "loss": 1.0237,
+      "step": 203
+    },
+    {
+      "epoch": 0.44810543657331137,
+      "grad_norm": 0.19673074781894684,
+      "learning_rate": 0.0002836319218241042,
+      "loss": 1.1888,
+      "step": 204
+    },
+    {
+      "epoch": 0.4503020318506315,
+      "grad_norm": 0.2969737648963928,
+      "learning_rate": 0.0002833876221498371,
+      "loss": 1.1792,
+      "step": 205
+    },
+    {
+      "epoch": 0.4524986271279517,
+      "grad_norm": 0.17545737326145172,
+      "learning_rate": 0.00028314332247557,
+      "loss": 1.223,
+      "step": 206
+    },
+    {
+      "epoch": 0.4546952224052718,
+      "grad_norm": 0.1426302045583725,
+      "learning_rate": 0.0002828990228013029,
+      "loss": 1.0211,
+      "step": 207
+    },
+    {
+      "epoch": 0.45689181768259196,
+      "grad_norm": 0.12227180600166321,
+      "learning_rate": 0.0002826547231270358,
+      "loss": 0.9242,
+      "step": 208
+    },
+    {
+      "epoch": 0.45908841295991215,
+      "grad_norm": 0.19062446057796478,
+      "learning_rate": 0.0002824104234527687,
+      "loss": 1.0456,
+      "step": 209
+    },
+    {
+      "epoch": 0.4612850082372323,
+      "grad_norm": 0.27791446447372437,
+      "learning_rate": 0.0002821661237785016,
+      "loss": 1.091,
+      "step": 210
+    },
+    {
+      "epoch": 0.4634816035145524,
+      "grad_norm": 0.1330525130033493,
+      "learning_rate": 0.0002819218241042345,
+      "loss": 1.1993,
+      "step": 211
+    },
+    {
+      "epoch": 0.4656781987918726,
+      "grad_norm": 0.3401545286178589,
+      "learning_rate": 0.00028167752442996737,
+      "loss": 1.0701,
+      "step": 212
+    },
+    {
+      "epoch": 0.46787479406919275,
+      "grad_norm": 0.13736934959888458,
+      "learning_rate": 0.0002814332247557003,
+      "loss": 0.8018,
+      "step": 213
+    },
+    {
+      "epoch": 0.4700713893465129,
+      "grad_norm": 0.14482451975345612,
+      "learning_rate": 0.0002811889250814332,
+      "loss": 0.9095,
+      "step": 214
+    },
+    {
+      "epoch": 0.47226798462383307,
+      "grad_norm": 0.12449502944946289,
+      "learning_rate": 0.0002809446254071661,
+      "loss": 1.1254,
+      "step": 215
+    },
+    {
+      "epoch": 0.4744645799011532,
+      "grad_norm": 0.173649862408638,
+      "learning_rate": 0.000280700325732899,
+      "loss": 1.0482,
+      "step": 216
+    },
+    {
+      "epoch": 0.47666117517847334,
+      "grad_norm": 0.20561538636684418,
+      "learning_rate": 0.00028045602605863187,
+      "loss": 0.9055,
+      "step": 217
+    },
+    {
+      "epoch": 0.47885777045579353,
+      "grad_norm": 0.11849188804626465,
+      "learning_rate": 0.0002802117263843648,
+      "loss": 1.0627,
+      "step": 218
+    },
+    {
+      "epoch": 0.48105436573311366,
+      "grad_norm": 0.1373571902513504,
+      "learning_rate": 0.0002799674267100977,
+      "loss": 0.9428,
+      "step": 219
+    },
+    {
+      "epoch": 0.48325096101043385,
+      "grad_norm": 0.13120803236961365,
+      "learning_rate": 0.0002797231270358306,
+      "loss": 0.9975,
+      "step": 220
+    },
+    {
+      "epoch": 0.485447556287754,
+      "grad_norm": 0.15404781699180603,
+      "learning_rate": 0.0002794788273615635,
+      "loss": 0.9749,
+      "step": 221
+    },
+    {
+      "epoch": 0.4876441515650741,
+      "grad_norm": 0.17716452479362488,
+      "learning_rate": 0.00027923452768729637,
+      "loss": 0.858,
+      "step": 222
+    },
+    {
+      "epoch": 0.4898407468423943,
+      "grad_norm": 0.14362047612667084,
+      "learning_rate": 0.00027899022801302926,
+      "loss": 1.033,
+      "step": 223
+    },
+    {
+      "epoch": 0.49203734211971445,
+      "grad_norm": 0.17385074496269226,
+      "learning_rate": 0.0002787459283387622,
+      "loss": 1.0792,
+      "step": 224
+    },
+    {
+      "epoch": 0.4942339373970346,
+      "grad_norm": 0.15892551839351654,
+      "learning_rate": 0.0002785016286644951,
+      "loss": 0.8183,
+      "step": 225
+    },
+    {
+      "epoch": 0.4964305326743548,
+      "grad_norm": 0.15761619806289673,
+      "learning_rate": 0.000278257328990228,
+      "loss": 1.2309,
+      "step": 226
+    },
+    {
+      "epoch": 0.4986271279516749,
+      "grad_norm": 0.12778125703334808,
+      "learning_rate": 0.00027801302931596087,
+      "loss": 0.9014,
+      "step": 227
+    },
+    {
+      "epoch": 0.500823723228995,
+      "grad_norm": 0.13027290999889374,
+      "learning_rate": 0.00027776872964169376,
+      "loss": 0.8739,
+      "step": 228
+    },
+    {
+      "epoch": 0.5030203185063152,
+      "grad_norm": 0.12950505316257477,
+      "learning_rate": 0.0002775244299674267,
+      "loss": 0.9932,
+      "step": 229
+    },
+    {
+      "epoch": 0.5052169137836353,
+      "grad_norm": 0.19206170737743378,
+      "learning_rate": 0.0002772801302931596,
+      "loss": 1.1505,
+      "step": 230
+    },
+    {
+      "epoch": 0.5074135090609555,
+      "grad_norm": 0.18855592608451843,
+      "learning_rate": 0.0002770358306188925,
+      "loss": 1.0686,
+      "step": 231
+    },
+    {
+      "epoch": 0.5096101043382757,
+      "grad_norm": 0.17141026258468628,
+      "learning_rate": 0.00027679153094462537,
+      "loss": 0.9628,
+      "step": 232
+    },
+    {
+      "epoch": 0.5118066996155958,
+      "grad_norm": 0.133894145488739,
+      "learning_rate": 0.00027654723127035826,
+      "loss": 1.0597,
+      "step": 233
+    },
+    {
+      "epoch": 0.514003294892916,
+      "grad_norm": 0.15527842938899994,
+      "learning_rate": 0.00027630293159609115,
+      "loss": 1.0251,
+      "step": 234
+    },
+    {
+      "epoch": 0.5161998901702362,
+      "grad_norm": 0.15913671255111694,
+      "learning_rate": 0.0002760586319218241,
+      "loss": 1.0539,
+      "step": 235
+    },
+    {
+      "epoch": 0.5183964854475562,
+      "grad_norm": 0.15577663481235504,
+      "learning_rate": 0.000275814332247557,
+      "loss": 0.8986,
+      "step": 236
+    },
+    {
+      "epoch": 0.5205930807248764,
+      "grad_norm": 0.17092250287532806,
+      "learning_rate": 0.00027557003257328987,
+      "loss": 1.2322,
+      "step": 237
+    },
+    {
+      "epoch": 0.5227896760021966,
+      "grad_norm": 0.10807476937770844,
+      "learning_rate": 0.00027532573289902276,
+      "loss": 1.0364,
+      "step": 238
+    },
+    {
+      "epoch": 0.5249862712795168,
+      "grad_norm": 0.16352656483650208,
+      "learning_rate": 0.00027508143322475565,
+      "loss": 0.9099,
+      "step": 239
+    },
+    {
+      "epoch": 0.5271828665568369,
+      "grad_norm": 0.14007623493671417,
+      "learning_rate": 0.00027483713355048854,
+      "loss": 1.0071,
+      "step": 240
+    },
+    {
+      "epoch": 0.5293794618341571,
+      "grad_norm": 0.1446109116077423,
+      "learning_rate": 0.0002745928338762215,
+      "loss": 1.0081,
+      "step": 241
+    },
+    {
+      "epoch": 0.5315760571114773,
+      "grad_norm": 0.15928012132644653,
+      "learning_rate": 0.00027434853420195437,
+      "loss": 0.9658,
+      "step": 242
+    },
+    {
+      "epoch": 0.5337726523887973,
+      "grad_norm": 0.10745849460363388,
+      "learning_rate": 0.00027410423452768726,
+      "loss": 0.8841,
+      "step": 243
+    },
+    {
+      "epoch": 0.5359692476661175,
+      "grad_norm": 0.2036733329296112,
+      "learning_rate": 0.00027385993485342015,
+      "loss": 1.0754,
+      "step": 244
+    },
+    {
+      "epoch": 0.5381658429434377,
+      "grad_norm": 0.12989583611488342,
+      "learning_rate": 0.00027361563517915304,
+      "loss": 1.0079,
+      "step": 245
+    },
+    {
+      "epoch": 0.5403624382207578,
+      "grad_norm": 0.1270613670349121,
+      "learning_rate": 0.000273371335504886,
+      "loss": 0.9672,
+      "step": 246
+    },
+    {
+      "epoch": 0.542559033498078,
+      "grad_norm": 0.13745439052581787,
+      "learning_rate": 0.00027312703583061887,
+      "loss": 0.8429,
+      "step": 247
+    },
+    {
+      "epoch": 0.5447556287753982,
+      "grad_norm": 0.15008491277694702,
+      "learning_rate": 0.00027288273615635176,
+      "loss": 0.9877,
+      "step": 248
+    },
+    {
+      "epoch": 0.5469522240527183,
+      "grad_norm": 0.16772015392780304,
+      "learning_rate": 0.00027263843648208465,
+      "loss": 0.8542,
+      "step": 249
+    },
+    {
+      "epoch": 0.5491488193300385,
+      "grad_norm": 0.14769555628299713,
+      "learning_rate": 0.00027239413680781754,
+      "loss": 0.9732,
+      "step": 250
+    },
+    {
+      "epoch": 0.5513454146073586,
+      "grad_norm": 0.15418405830860138,
+      "learning_rate": 0.0002721498371335504,
+      "loss": 0.9765,
+      "step": 251
+    },
+    {
+      "epoch": 0.5535420098846787,
+      "grad_norm": 0.15011285245418549,
+      "learning_rate": 0.00027190553745928337,
+      "loss": 0.9914,
+      "step": 252
+    },
+    {
+      "epoch": 0.5557386051619989,
+      "grad_norm": 0.18463046848773956,
+      "learning_rate": 0.00027166123778501626,
+      "loss": 1.0225,
+      "step": 253
+    },
+    {
+      "epoch": 0.5579352004393191,
+      "grad_norm": 0.14542873203754425,
+      "learning_rate": 0.00027141693811074915,
+      "loss": 0.8873,
+      "step": 254
+    },
+    {
+      "epoch": 0.5601317957166392,
+      "grad_norm": 0.13668955862522125,
+      "learning_rate": 0.00027117263843648204,
+      "loss": 1.0166,
+      "step": 255
+    },
+    {
+      "epoch": 0.5623283909939594,
+      "grad_norm": 0.1389240324497223,
+      "learning_rate": 0.0002709283387622149,
+      "loss": 1.0224,
+      "step": 256
+    },
+    {
+      "epoch": 0.5645249862712796,
+      "grad_norm": 0.2767558991909027,
+      "learning_rate": 0.00027068403908794787,
+      "loss": 0.8512,
+      "step": 257
+    },
+    {
+      "epoch": 0.5667215815485996,
+      "grad_norm": 0.1494889259338379,
+      "learning_rate": 0.00027043973941368076,
+      "loss": 1.0538,
+      "step": 258
+    },
+    {
+      "epoch": 0.5689181768259198,
+      "grad_norm": 0.15071460604667664,
+      "learning_rate": 0.00027019543973941365,
+      "loss": 0.8536,
+      "step": 259
+    },
+    {
+      "epoch": 0.57111477210324,
+      "grad_norm": 0.1345864236354828,
+      "learning_rate": 0.00026995114006514654,
+      "loss": 0.9729,
+      "step": 260
+    },
+    {
+      "epoch": 0.5733113673805601,
+      "grad_norm": 0.11756473034620285,
+      "learning_rate": 0.00026970684039087943,
+      "loss": 1.0119,
+      "step": 261
+    },
+    {
+      "epoch": 0.5755079626578803,
+      "grad_norm": 0.16454483568668365,
+      "learning_rate": 0.0002694625407166123,
+      "loss": 0.9963,
+      "step": 262
+    },
+    {
+      "epoch": 0.5777045579352005,
+      "grad_norm": 0.11827659606933594,
+      "learning_rate": 0.00026921824104234526,
+      "loss": 0.8272,
+      "step": 263
+    },
+    {
+      "epoch": 0.5799011532125206,
+      "grad_norm": 0.21315135061740875,
+      "learning_rate": 0.00026897394136807815,
+      "loss": 0.9962,
+      "step": 264
+    },
+    {
+      "epoch": 0.5820977484898407,
+      "grad_norm": 0.1703679859638214,
+      "learning_rate": 0.00026872964169381104,
+      "loss": 1.3526,
+      "step": 265
+    },
+    {
+      "epoch": 0.5842943437671609,
+      "grad_norm": 0.11550629884004593,
+      "learning_rate": 0.00026848534201954393,
+      "loss": 1.0609,
+      "step": 266
+    },
+    {
+      "epoch": 0.586490939044481,
+      "grad_norm": 0.11357277631759644,
+      "learning_rate": 0.0002682410423452768,
+      "loss": 0.9941,
+      "step": 267
+    },
+    {
+      "epoch": 0.5886875343218012,
+      "grad_norm": 0.19084390997886658,
+      "learning_rate": 0.00026799674267100976,
+      "loss": 1.1029,
+      "step": 268
+    },
+    {
+      "epoch": 0.5908841295991214,
+      "grad_norm": 0.1855383664369583,
+      "learning_rate": 0.00026775244299674265,
+      "loss": 1.1651,
+      "step": 269
+    },
+    {
+      "epoch": 0.5930807248764415,
+      "grad_norm": 0.1874328851699829,
+      "learning_rate": 0.00026750814332247554,
+      "loss": 1.1025,
+      "step": 270
+    },
+    {
+      "epoch": 0.5952773201537617,
+      "grad_norm": 0.11996940523386002,
+      "learning_rate": 0.00026726384364820843,
+      "loss": 0.9033,
+      "step": 271
+    },
+    {
+      "epoch": 0.5974739154310819,
+      "grad_norm": 0.1481139212846756,
+      "learning_rate": 0.0002670195439739413,
+      "loss": 0.9462,
+      "step": 272
+    },
+    {
+      "epoch": 0.5996705107084019,
+      "grad_norm": 0.18385189771652222,
+      "learning_rate": 0.0002667752442996742,
+      "loss": 0.9308,
+      "step": 273
+    },
+    {
+      "epoch": 0.6018671059857221,
+      "grad_norm": 0.31651073694229126,
+      "learning_rate": 0.00026653094462540715,
+      "loss": 1.0155,
+      "step": 274
+    },
+    {
+      "epoch": 0.6040637012630423,
+      "grad_norm": 0.18079692125320435,
+      "learning_rate": 0.00026628664495114004,
+      "loss": 1.0197,
+      "step": 275
+    },
+    {
+      "epoch": 0.6062602965403624,
+      "grad_norm": 0.17467527091503143,
+      "learning_rate": 0.00026604234527687293,
+      "loss": 0.7781,
+      "step": 276
+    },
+    {
+      "epoch": 0.6084568918176826,
+      "grad_norm": 0.12584443390369415,
+      "learning_rate": 0.0002657980456026058,
+      "loss": 0.9929,
+      "step": 277
+    },
+    {
+      "epoch": 0.6106534870950028,
+      "grad_norm": 0.17152723670005798,
+      "learning_rate": 0.0002655537459283387,
+      "loss": 1.1381,
+      "step": 278
+    },
+    {
+      "epoch": 0.6128500823723229,
+      "grad_norm": 0.1319683939218521,
+      "learning_rate": 0.00026530944625407165,
+      "loss": 0.946,
+      "step": 279
+    },
+    {
+      "epoch": 0.615046677649643,
+      "grad_norm": 0.13232830166816711,
+      "learning_rate": 0.00026506514657980454,
+      "loss": 0.8509,
+      "step": 280
+    },
+    {
+      "epoch": 0.6172432729269632,
+      "grad_norm": 0.11810342967510223,
+      "learning_rate": 0.00026482084690553743,
+      "loss": 0.8146,
+      "step": 281
+    },
+    {
+      "epoch": 0.6194398682042833,
+      "grad_norm": 0.11837928742170334,
+      "learning_rate": 0.0002645765472312703,
+      "loss": 0.9077,
+      "step": 282
+    },
+    {
+      "epoch": 0.6216364634816035,
+      "grad_norm": 0.14008352160453796,
+      "learning_rate": 0.0002643322475570032,
+      "loss": 1.056,
+      "step": 283
+    },
+    {
+      "epoch": 0.6238330587589237,
+      "grad_norm": 0.1293041706085205,
+      "learning_rate": 0.0002640879478827361,
+      "loss": 0.8668,
+      "step": 284
+    },
+    {
+      "epoch": 0.6260296540362438,
+      "grad_norm": 0.10690130293369293,
+      "learning_rate": 0.00026384364820846904,
+      "loss": 0.8966,
+      "step": 285
+    },
+    {
+      "epoch": 0.628226249313564,
+      "grad_norm": 0.1987561136484146,
+      "learning_rate": 0.00026359934853420193,
+      "loss": 0.862,
+      "step": 286
+    },
+    {
+      "epoch": 0.6304228445908842,
+      "grad_norm": 0.11143385618925095,
+      "learning_rate": 0.0002633550488599348,
+      "loss": 0.9939,
+      "step": 287
+    },
+    {
+      "epoch": 0.6326194398682042,
+      "grad_norm": 0.13517679274082184,
+      "learning_rate": 0.0002631107491856677,
+      "loss": 0.8909,
+      "step": 288
+    },
+    {
+      "epoch": 0.6348160351455244,
+      "grad_norm": 0.18095937371253967,
+      "learning_rate": 0.0002628664495114006,
+      "loss": 0.8146,
+      "step": 289
+    },
+    {
+      "epoch": 0.6370126304228446,
+      "grad_norm": 0.13336767256259918,
+      "learning_rate": 0.0002626221498371335,
+      "loss": 0.8443,
+      "step": 290
+    },
+    {
+      "epoch": 0.6392092257001647,
+      "grad_norm": 0.14703990519046783,
+      "learning_rate": 0.00026237785016286643,
+      "loss": 1.0456,
+      "step": 291
+    },
+    {
+      "epoch": 0.6414058209774849,
+      "grad_norm": 0.1435512751340866,
+      "learning_rate": 0.0002621335504885993,
+      "loss": 1.0309,
+      "step": 292
+    },
+    {
+      "epoch": 0.6436024162548051,
+      "grad_norm": 0.15228092670440674,
+      "learning_rate": 0.00026188925081433226,
+      "loss": 1.0162,
+      "step": 293
+    },
+    {
+      "epoch": 0.6457990115321252,
+      "grad_norm": 0.1642136424779892,
+      "learning_rate": 0.00026164495114006515,
+      "loss": 1.0079,
+      "step": 294
+    },
+    {
+      "epoch": 0.6479956068094453,
+      "grad_norm": 0.14121542870998383,
+      "learning_rate": 0.00026140065146579804,
+      "loss": 0.9609,
+      "step": 295
+    },
+    {
+      "epoch": 0.6501922020867655,
+      "grad_norm": 0.13058587908744812,
+      "learning_rate": 0.00026115635179153093,
+      "loss": 0.9614,
+      "step": 296
+    },
+    {
+      "epoch": 0.6523887973640856,
+      "grad_norm": 0.12643840909004211,
+      "learning_rate": 0.0002609120521172638,
+      "loss": 0.9067,
+      "step": 297
+    },
+    {
+      "epoch": 0.6545853926414058,
+      "grad_norm": 0.20523543655872345,
+      "learning_rate": 0.00026066775244299676,
+      "loss": 0.9624,
+      "step": 298
+    },
+    {
+      "epoch": 0.656781987918726,
+      "grad_norm": 0.11762203276157379,
+      "learning_rate": 0.00026042345276872965,
+      "loss": 0.934,
+      "step": 299
+    },
+    {
+      "epoch": 0.6589785831960461,
+      "grad_norm": 0.15599507093429565,
+      "learning_rate": 0.00026017915309446254,
+      "loss": 0.9329,
+      "step": 300
+    },
+    {
+      "epoch": 0.6611751784733663,
+      "grad_norm": 0.22324621677398682,
+      "learning_rate": 0.00025993485342019543,
+      "loss": 1.1429,
+      "step": 301
+    },
+    {
+      "epoch": 0.6633717737506865,
+      "grad_norm": 0.1364641785621643,
+      "learning_rate": 0.0002596905537459283,
+      "loss": 0.9514,
+      "step": 302
+    },
+    {
+      "epoch": 0.6655683690280065,
+      "grad_norm": 0.12749195098876953,
+      "learning_rate": 0.0002594462540716612,
+      "loss": 0.8445,
+      "step": 303
+    },
+    {
+      "epoch": 0.6677649643053267,
+      "grad_norm": 0.1623351126909256,
+      "learning_rate": 0.00025920195439739415,
+      "loss": 1.0894,
+      "step": 304
+    },
+    {
+      "epoch": 0.6699615595826469,
+      "grad_norm": 0.1912679374217987,
+      "learning_rate": 0.00025895765472312704,
+      "loss": 1.0661,
+      "step": 305
+    },
+    {
+      "epoch": 0.6721581548599671,
+      "grad_norm": 0.14835670590400696,
+      "learning_rate": 0.00025871335504885993,
+      "loss": 0.9108,
+      "step": 306
+    },
+    {
+      "epoch": 0.6743547501372872,
+      "grad_norm": 0.12244424223899841,
+      "learning_rate": 0.0002584690553745928,
+      "loss": 0.9987,
+      "step": 307
+    },
+    {
+      "epoch": 0.6765513454146074,
+      "grad_norm": 0.2500230669975281,
+      "learning_rate": 0.0002582247557003257,
+      "loss": 1.0257,
+      "step": 308
+    },
+    {
+      "epoch": 0.6787479406919276,
+      "grad_norm": 0.12365983426570892,
+      "learning_rate": 0.00025798045602605865,
+      "loss": 1.0206,
+      "step": 309
+    },
+    {
+      "epoch": 0.6809445359692476,
+      "grad_norm": 0.20770707726478577,
+      "learning_rate": 0.00025773615635179154,
+      "loss": 1.073,
+      "step": 310
+    },
+    {
+      "epoch": 0.6831411312465678,
+      "grad_norm": 0.16401170194149017,
+      "learning_rate": 0.00025749185667752443,
+      "loss": 0.9209,
+      "step": 311
+    },
+    {
+      "epoch": 0.685337726523888,
+      "grad_norm": 0.14924022555351257,
+      "learning_rate": 0.0002572475570032573,
+      "loss": 1.0164,
+      "step": 312
+    },
+    {
+      "epoch": 0.6875343218012081,
+      "grad_norm": 0.1671312302350998,
+      "learning_rate": 0.0002570032573289902,
+      "loss": 0.9301,
+      "step": 313
+    },
+    {
+      "epoch": 0.6897309170785283,
+      "grad_norm": 0.1174243912100792,
+      "learning_rate": 0.0002567589576547231,
+      "loss": 0.6693,
+      "step": 314
+    },
+    {
+      "epoch": 0.6919275123558485,
+      "grad_norm": 0.144245907664299,
+      "learning_rate": 0.00025651465798045604,
+      "loss": 0.7654,
+      "step": 315
+    },
+    {
+      "epoch": 0.6941241076331686,
+      "grad_norm": 0.1435171663761139,
+      "learning_rate": 0.00025627035830618893,
+      "loss": 1.026,
+      "step": 316
+    },
+    {
+      "epoch": 0.6963207029104888,
+      "grad_norm": 0.13194742798805237,
+      "learning_rate": 0.0002560260586319218,
+      "loss": 1.1549,
+      "step": 317
+    },
+    {
+      "epoch": 0.6985172981878089,
+      "grad_norm": 0.154351145029068,
+      "learning_rate": 0.0002557817589576547,
+      "loss": 1.0758,
+      "step": 318
+    },
+    {
+      "epoch": 0.700713893465129,
+      "grad_norm": 0.12992306053638458,
+      "learning_rate": 0.0002555374592833876,
+      "loss": 0.8639,
+      "step": 319
+    },
+    {
+      "epoch": 0.7029104887424492,
+      "grad_norm": 0.11188488453626633,
+      "learning_rate": 0.0002552931596091205,
+      "loss": 0.9073,
+      "step": 320
+    },
+    {
+      "epoch": 0.7051070840197694,
+      "grad_norm": 0.10336731374263763,
+      "learning_rate": 0.00025504885993485343,
+      "loss": 0.9417,
+      "step": 321
+    },
+    {
+      "epoch": 0.7073036792970895,
+      "grad_norm": 0.11573520302772522,
+      "learning_rate": 0.0002548045602605863,
+      "loss": 0.9759,
+      "step": 322
+    },
+    {
+      "epoch": 0.7095002745744097,
+      "grad_norm": 0.13856199383735657,
+      "learning_rate": 0.0002545602605863192,
+      "loss": 0.8407,
+      "step": 323
+    },
+    {
+      "epoch": 0.7116968698517299,
+      "grad_norm": 0.15449021756649017,
+      "learning_rate": 0.0002543159609120521,
+      "loss": 1.0627,
+      "step": 324
+    },
+    {
+      "epoch": 0.7138934651290499,
+      "grad_norm": 0.15693782269954681,
+      "learning_rate": 0.000254071661237785,
+      "loss": 0.9362,
+      "step": 325
+    },
+    {
+      "epoch": 0.7160900604063701,
+      "grad_norm": 0.15206071734428406,
+      "learning_rate": 0.00025382736156351793,
+      "loss": 0.9224,
+      "step": 326
+    },
+    {
+      "epoch": 0.7182866556836903,
+      "grad_norm": 0.12770970165729523,
+      "learning_rate": 0.0002535830618892508,
+      "loss": 0.9816,
+      "step": 327
+    },
+    {
+      "epoch": 0.7204832509610104,
+      "grad_norm": 0.16101676225662231,
+      "learning_rate": 0.0002533387622149837,
+      "loss": 1.1218,
+      "step": 328
+    },
+    {
+      "epoch": 0.7226798462383306,
+      "grad_norm": 0.11391662061214447,
+      "learning_rate": 0.0002530944625407166,
+      "loss": 0.9282,
+      "step": 329
+    },
+    {
+      "epoch": 0.7248764415156508,
+      "grad_norm": 0.13816086947917938,
+      "learning_rate": 0.0002528501628664495,
+      "loss": 0.9214,
+      "step": 330
+    },
+    {
+      "epoch": 0.7270730367929709,
+      "grad_norm": 0.1510637253522873,
+      "learning_rate": 0.0002526058631921824,
+      "loss": 1.0897,
+      "step": 331
+    },
+    {
+      "epoch": 0.729269632070291,
+      "grad_norm": 0.4288111925125122,
+      "learning_rate": 0.0002523615635179153,
+      "loss": 0.8802,
+      "step": 332
+    },
+    {
+      "epoch": 0.7314662273476112,
+      "grad_norm": 0.1300736963748932,
+      "learning_rate": 0.0002521172638436482,
+      "loss": 1.0093,
+      "step": 333
+    },
+    {
+      "epoch": 0.7336628226249313,
+      "grad_norm": 0.1599997729063034,
+      "learning_rate": 0.0002518729641693811,
+      "loss": 1.1069,
+      "step": 334
+    },
+    {
+      "epoch": 0.7358594179022515,
+      "grad_norm": 0.1353105902671814,
+      "learning_rate": 0.000251628664495114,
+      "loss": 0.9674,
+      "step": 335
+    },
+    {
+      "epoch": 0.7380560131795717,
+      "grad_norm": 0.1413733959197998,
+      "learning_rate": 0.0002513843648208469,
+      "loss": 0.9184,
+      "step": 336
+    },
+    {
+      "epoch": 0.7402526084568918,
+      "grad_norm": 0.1279912292957306,
+      "learning_rate": 0.0002511400651465798,
+      "loss": 1.1273,
+      "step": 337
+    },
+    {
+      "epoch": 0.742449203734212,
+      "grad_norm": 0.17359627783298492,
+      "learning_rate": 0.0002508957654723127,
+      "loss": 0.909,
+      "step": 338
+    },
+    {
+      "epoch": 0.7446457990115322,
+      "grad_norm": 0.11573012173175812,
+      "learning_rate": 0.0002506514657980456,
+      "loss": 0.9949,
+      "step": 339
+    },
+    {
+      "epoch": 0.7468423942888522,
+      "grad_norm": 0.15121477842330933,
+      "learning_rate": 0.0002504071661237785,
+      "loss": 1.0173,
+      "step": 340
+    },
+    {
+      "epoch": 0.7490389895661724,
+      "grad_norm": 0.13637810945510864,
+      "learning_rate": 0.0002501628664495114,
+      "loss": 0.886,
+      "step": 341
+    },
+    {
+      "epoch": 0.7512355848434926,
+      "grad_norm": 0.13389886915683746,
+      "learning_rate": 0.00024991856677524427,
+      "loss": 0.8421,
+      "step": 342
+    },
+    {
+      "epoch": 0.7534321801208127,
+      "grad_norm": 0.11682885140180588,
+      "learning_rate": 0.0002496742671009772,
+      "loss": 0.8713,
+      "step": 343
+    },
+    {
+      "epoch": 0.7556287753981329,
+      "grad_norm": 0.1256597936153412,
+      "learning_rate": 0.0002494299674267101,
+      "loss": 0.9076,
+      "step": 344
+    },
+    {
+      "epoch": 0.7578253706754531,
+      "grad_norm": 0.11844318360090256,
+      "learning_rate": 0.000249185667752443,
+      "loss": 0.7949,
+      "step": 345
+    },
+    {
+      "epoch": 0.7600219659527732,
+      "grad_norm": 0.13575026392936707,
+      "learning_rate": 0.0002489413680781759,
+      "loss": 0.9276,
+      "step": 346
+    },
+    {
+      "epoch": 0.7622185612300933,
+      "grad_norm": 0.1453033983707428,
+      "learning_rate": 0.00024869706840390877,
+      "loss": 0.9773,
+      "step": 347
+    },
+    {
+      "epoch": 0.7644151565074135,
+      "grad_norm": 0.2682100534439087,
+      "learning_rate": 0.0002484527687296417,
+      "loss": 1.1746,
+      "step": 348
+    },
+    {
+      "epoch": 0.7666117517847336,
+      "grad_norm": 0.13317736983299255,
+      "learning_rate": 0.0002482084690553746,
+      "loss": 0.72,
+      "step": 349
+    },
+    {
+      "epoch": 0.7688083470620538,
+      "grad_norm": 0.14967526495456696,
+      "learning_rate": 0.0002479641693811075,
+      "loss": 1.1455,
+      "step": 350
+    },
+    {
+      "epoch": 0.771004942339374,
+      "grad_norm": 0.12718360126018524,
+      "learning_rate": 0.0002477198697068404,
+      "loss": 1.1075,
+      "step": 351
+    },
+    {
+      "epoch": 0.7732015376166941,
+      "grad_norm": 0.16455650329589844,
+      "learning_rate": 0.00024747557003257327,
+      "loss": 1.0254,
+      "step": 352
+    },
+    {
+      "epoch": 0.7753981328940143,
+      "grad_norm": 0.6730981469154358,
+      "learning_rate": 0.00024723127035830616,
+      "loss": 0.7783,
+      "step": 353
+    },
+    {
+      "epoch": 0.7775947281713345,
+      "grad_norm": 0.13041049242019653,
+      "learning_rate": 0.0002469869706840391,
+      "loss": 0.9921,
+      "step": 354
+    },
+    {
+      "epoch": 0.7797913234486545,
+      "grad_norm": 0.14412690699100494,
+      "learning_rate": 0.000246742671009772,
+      "loss": 0.7214,
+      "step": 355
+    },
+    {
+      "epoch": 0.7819879187259747,
+      "grad_norm": 0.14545854926109314,
+      "learning_rate": 0.0002464983713355049,
+      "loss": 0.8058,
+      "step": 356
+    },
+    {
+      "epoch": 0.7841845140032949,
+      "grad_norm": 0.14278969168663025,
+      "learning_rate": 0.00024625407166123777,
+      "loss": 0.928,
+      "step": 357
+    },
+    {
+      "epoch": 0.786381109280615,
+      "grad_norm": 0.10582005977630615,
+      "learning_rate": 0.00024600977198697066,
+      "loss": 0.8861,
+      "step": 358
+    },
+    {
+      "epoch": 0.7885777045579352,
+      "grad_norm": 0.14361511170864105,
+      "learning_rate": 0.00024576547231270354,
+      "loss": 0.8635,
+      "step": 359
+    },
+    {
+      "epoch": 0.7907742998352554,
+      "grad_norm": 0.10195020586252213,
+      "learning_rate": 0.0002455211726384365,
+      "loss": 0.77,
+      "step": 360
+    },
+    {
+      "epoch": 0.7929708951125755,
+      "grad_norm": 0.13805824518203735,
+      "learning_rate": 0.0002452768729641694,
+      "loss": 1.0706,
+      "step": 361
+    },
+    {
+      "epoch": 0.7951674903898956,
+      "grad_norm": 0.1413760930299759,
+      "learning_rate": 0.00024503257328990227,
+      "loss": 0.9664,
+      "step": 362
+    },
+    {
+      "epoch": 0.7973640856672158,
+      "grad_norm": 0.17732904851436615,
+      "learning_rate": 0.00024478827361563516,
+      "loss": 0.8565,
+      "step": 363
+    },
+    {
+      "epoch": 0.7995606809445359,
+      "grad_norm": 0.1548486202955246,
+      "learning_rate": 0.00024454397394136804,
+      "loss": 0.9019,
+      "step": 364
+    },
+    {
+      "epoch": 0.8017572762218561,
+      "grad_norm": 0.1472049206495285,
+      "learning_rate": 0.000244299674267101,
+      "loss": 0.7398,
+      "step": 365
+    },
+    {
+      "epoch": 0.8039538714991763,
+      "grad_norm": 0.10263271629810333,
+      "learning_rate": 0.00024405537459283385,
+      "loss": 0.9662,
+      "step": 366
+    },
+    {
+      "epoch": 0.8061504667764964,
+      "grad_norm": 0.1480238437652588,
+      "learning_rate": 0.00024381107491856677,
+      "loss": 0.9077,
+      "step": 367
+    },
+    {
+      "epoch": 0.8083470620538166,
+      "grad_norm": 0.26850226521492004,
+      "learning_rate": 0.00024356677524429966,
+      "loss": 0.8464,
+      "step": 368
+    },
+    {
+      "epoch": 0.8105436573311368,
+      "grad_norm": 0.12278976291418076,
+      "learning_rate": 0.00024332247557003257,
+      "loss": 0.8219,
+      "step": 369
+    },
+    {
+      "epoch": 0.8127402526084568,
+      "grad_norm": 0.11495000123977661,
+      "learning_rate": 0.00024307817589576546,
+      "loss": 0.7323,
+      "step": 370
+    },
+    {
+      "epoch": 0.814936847885777,
+      "grad_norm": 0.11105601489543915,
+      "learning_rate": 0.00024283387622149835,
+      "loss": 0.9939,
+      "step": 371
+    },
+    {
+      "epoch": 0.8171334431630972,
+      "grad_norm": 0.15675702691078186,
+      "learning_rate": 0.00024258957654723127,
+      "loss": 0.9948,
+      "step": 372
+    },
+    {
+      "epoch": 0.8193300384404174,
+      "grad_norm": 0.09821725636720657,
+      "learning_rate": 0.00024234527687296416,
+      "loss": 0.8889,
+      "step": 373
+    },
+    {
+      "epoch": 0.8215266337177375,
+      "grad_norm": 0.18814116716384888,
+      "learning_rate": 0.00024210097719869705,
+      "loss": 0.9287,
+      "step": 374
+    },
+    {
+      "epoch": 0.8237232289950577,
+      "grad_norm": 0.12179295718669891,
+      "learning_rate": 0.00024185667752442996,
+      "loss": 1.0035,
+      "step": 375
+    },
+    {
+      "epoch": 0.8259198242723779,
+      "grad_norm": 0.12768927216529846,
+      "learning_rate": 0.00024161237785016285,
+      "loss": 0.9297,
+      "step": 376
+    },
+    {
+      "epoch": 0.828116419549698,
+      "grad_norm": 0.1317479908466339,
+      "learning_rate": 0.00024136807817589574,
+      "loss": 0.9831,
+      "step": 377
+    },
+    {
+      "epoch": 0.8303130148270181,
+      "grad_norm": 0.1062789335846901,
+      "learning_rate": 0.00024112377850162866,
+      "loss": 0.8954,
+      "step": 378
+    },
+    {
+      "epoch": 0.8325096101043383,
+      "grad_norm": 0.11569977551698685,
+      "learning_rate": 0.00024087947882736155,
+      "loss": 0.7818,
+      "step": 379
+    },
+    {
+      "epoch": 0.8347062053816584,
+      "grad_norm": 0.13721084594726562,
+      "learning_rate": 0.00024063517915309446,
+      "loss": 1.0023,
+      "step": 380
+    },
+    {
+      "epoch": 0.8369028006589786,
+      "grad_norm": 0.12809288501739502,
+      "learning_rate": 0.00024039087947882735,
+      "loss": 0.9533,
+      "step": 381
+    },
+    {
+      "epoch": 0.8390993959362988,
+      "grad_norm": 0.15030856430530548,
+      "learning_rate": 0.00024014657980456024,
+      "loss": 0.9799,
+      "step": 382
+    },
+    {
+      "epoch": 0.8412959912136189,
+      "grad_norm": 0.11091528087854385,
+      "learning_rate": 0.00023990228013029316,
+      "loss": 0.9974,
+      "step": 383
+    },
+    {
+      "epoch": 0.8434925864909391,
+      "grad_norm": 0.16910214722156525,
+      "learning_rate": 0.00023965798045602605,
+      "loss": 1.0005,
+      "step": 384
+    },
+    {
+      "epoch": 0.8456891817682592,
+      "grad_norm": 0.21585018932819366,
+      "learning_rate": 0.00023941368078175893,
+      "loss": 1.0896,
+      "step": 385
+    },
+    {
+      "epoch": 0.8478857770455793,
+      "grad_norm": 0.16444918513298035,
+      "learning_rate": 0.00023916938110749185,
+      "loss": 1.0137,
+      "step": 386
+    },
+    {
+      "epoch": 0.8500823723228995,
+      "grad_norm": 0.1044774129986763,
+      "learning_rate": 0.00023892508143322474,
+      "loss": 0.8621,
+      "step": 387
+    },
+    {
+      "epoch": 0.8522789676002197,
+      "grad_norm": 0.13315646350383759,
+      "learning_rate": 0.00023868078175895763,
+      "loss": 0.9951,
+      "step": 388
+    },
+    {
+      "epoch": 0.8544755628775398,
+      "grad_norm": 0.12963467836380005,
+      "learning_rate": 0.00023843648208469055,
+      "loss": 0.7219,
+      "step": 389
+    },
+    {
+      "epoch": 0.85667215815486,
+      "grad_norm": 0.1121048629283905,
+      "learning_rate": 0.00023819218241042343,
+      "loss": 0.8462,
+      "step": 390
+    },
+    {
+      "epoch": 0.8588687534321802,
+      "grad_norm": 0.1195456013083458,
+      "learning_rate": 0.00023794788273615632,
+      "loss": 0.8953,
+      "step": 391
+    },
+    {
+      "epoch": 0.8610653487095002,
+      "grad_norm": 0.1233413964509964,
+      "learning_rate": 0.00023770358306188924,
+      "loss": 1.149,
+      "step": 392
+    },
+    {
+      "epoch": 0.8632619439868204,
+      "grad_norm": 0.1587863266468048,
+      "learning_rate": 0.00023745928338762213,
+      "loss": 1.261,
+      "step": 393
+    },
+    {
+      "epoch": 0.8654585392641406,
+      "grad_norm": 0.2218884378671646,
+      "learning_rate": 0.00023721498371335505,
+      "loss": 0.9636,
+      "step": 394
+    },
+    {
+      "epoch": 0.8676551345414607,
+      "grad_norm": 0.13585165143013,
+      "learning_rate": 0.00023697068403908794,
+      "loss": 0.8642,
+      "step": 395
+    },
+    {
+      "epoch": 0.8698517298187809,
+      "grad_norm": 0.11273190379142761,
+      "learning_rate": 0.00023672638436482082,
+      "loss": 0.8761,
+      "step": 396
+    },
+    {
+      "epoch": 0.8720483250961011,
+      "grad_norm": 0.12780971825122833,
+      "learning_rate": 0.00023648208469055374,
+      "loss": 0.9501,
+      "step": 397
+    },
+    {
+      "epoch": 0.8742449203734212,
+      "grad_norm": 0.12432140111923218,
+      "learning_rate": 0.00023623778501628663,
+      "loss": 0.8263,
+      "step": 398
+    },
+    {
+      "epoch": 0.8764415156507414,
+      "grad_norm": 0.14229018986225128,
+      "learning_rate": 0.00023599348534201952,
+      "loss": 0.9283,
+      "step": 399
+    },
+    {
+      "epoch": 0.8786381109280615,
+      "grad_norm": 0.15906508266925812,
+      "learning_rate": 0.00023574918566775244,
+      "loss": 1.0313,
+      "step": 400
+    },
+    {
+      "epoch": 0.8808347062053816,
+      "grad_norm": 0.14006449282169342,
+      "learning_rate": 0.00023550488599348532,
+      "loss": 0.7532,
+      "step": 401
+    },
+    {
+      "epoch": 0.8830313014827018,
+      "grad_norm": 0.11119817942380905,
+      "learning_rate": 0.00023526058631921821,
+      "loss": 0.9078,
+      "step": 402
+    },
+    {
+      "epoch": 0.885227896760022,
+      "grad_norm": 0.11699362099170685,
+      "learning_rate": 0.00023501628664495113,
+      "loss": 1.0563,
+      "step": 403
+    },
+    {
+      "epoch": 0.8874244920373421,
+      "grad_norm": 0.09198285639286041,
+      "learning_rate": 0.00023477198697068402,
+      "loss": 1.0481,
+      "step": 404
+    },
+    {
+      "epoch": 0.8896210873146623,
+      "grad_norm": 0.12001250684261322,
+      "learning_rate": 0.0002345276872964169,
+      "loss": 0.7775,
+      "step": 405
+    },
+    {
+      "epoch": 0.8918176825919825,
+      "grad_norm": 0.10852393507957458,
+      "learning_rate": 0.00023428338762214982,
+      "loss": 0.8179,
+      "step": 406
+    },
+    {
+      "epoch": 0.8940142778693025,
+      "grad_norm": 0.16269919276237488,
+      "learning_rate": 0.00023403908794788271,
+      "loss": 0.8072,
+      "step": 407
+    },
+    {
+      "epoch": 0.8962108731466227,
+      "grad_norm": 0.1735255867242813,
+      "learning_rate": 0.00023379478827361563,
+      "loss": 0.8179,
+      "step": 408
+    },
+    {
+      "epoch": 0.8984074684239429,
+      "grad_norm": 0.10125740617513657,
+      "learning_rate": 0.00023355048859934852,
+      "loss": 0.8397,
+      "step": 409
+    },
+    {
+      "epoch": 0.900604063701263,
+      "grad_norm": 0.1552150547504425,
+      "learning_rate": 0.0002333061889250814,
+      "loss": 1.0605,
+      "step": 410
+    },
+    {
+      "epoch": 0.9028006589785832,
+      "grad_norm": 0.11182007193565369,
+      "learning_rate": 0.00023306188925081432,
+      "loss": 0.7748,
+      "step": 411
+    },
+    {
+      "epoch": 0.9049972542559034,
+      "grad_norm": 0.16088099777698517,
+      "learning_rate": 0.00023281758957654721,
+      "loss": 0.8609,
+      "step": 412
+    },
+    {
+      "epoch": 0.9071938495332235,
+      "grad_norm": 0.10549604892730713,
+      "learning_rate": 0.0002325732899022801,
+      "loss": 0.8732,
+      "step": 413
+    },
+    {
+      "epoch": 0.9093904448105437,
+      "grad_norm": 0.185531347990036,
+      "learning_rate": 0.00023232899022801302,
+      "loss": 1.0346,
+      "step": 414
+    },
+    {
+      "epoch": 0.9115870400878638,
+      "grad_norm": 0.12503954768180847,
+      "learning_rate": 0.0002320846905537459,
+      "loss": 0.8717,
+      "step": 415
+    },
+    {
+      "epoch": 0.9137836353651839,
+      "grad_norm": 0.13538984954357147,
+      "learning_rate": 0.0002318403908794788,
+      "loss": 1.0148,
+      "step": 416
+    },
+    {
+      "epoch": 0.9159802306425041,
+      "grad_norm": 0.11026758700609207,
+      "learning_rate": 0.00023159609120521171,
+      "loss": 0.9259,
+      "step": 417
+    },
+    {
+      "epoch": 0.9181768259198243,
+      "grad_norm": 0.12433885037899017,
+      "learning_rate": 0.0002313517915309446,
+      "loss": 0.6691,
+      "step": 418
+    },
+    {
+      "epoch": 0.9203734211971444,
+      "grad_norm": 0.12900269031524658,
+      "learning_rate": 0.00023110749185667752,
+      "loss": 0.7597,
+      "step": 419
+    },
+    {
+      "epoch": 0.9225700164744646,
+      "grad_norm": 0.1040680930018425,
+      "learning_rate": 0.0002308631921824104,
+      "loss": 0.9268,
+      "step": 420
+    },
+    {
+      "epoch": 0.9247666117517848,
+      "grad_norm": 0.12501965463161469,
+      "learning_rate": 0.0002306188925081433,
+      "loss": 0.8766,
+      "step": 421
+    },
+    {
+      "epoch": 0.9269632070291048,
+      "grad_norm": 0.13488435745239258,
+      "learning_rate": 0.00023037459283387621,
+      "loss": 0.8919,
+      "step": 422
+    },
+    {
+      "epoch": 0.929159802306425,
+      "grad_norm": 0.19442269206047058,
+      "learning_rate": 0.0002301302931596091,
+      "loss": 1.05,
+      "step": 423
+    },
+    {
+      "epoch": 0.9313563975837452,
+      "grad_norm": 0.14239557087421417,
+      "learning_rate": 0.000229885993485342,
+      "loss": 0.9127,
+      "step": 424
+    },
+    {
+      "epoch": 0.9335529928610653,
+      "grad_norm": 0.13452358543872833,
+      "learning_rate": 0.0002296416938110749,
+      "loss": 0.8358,
+      "step": 425
+    },
+    {
+      "epoch": 0.9357495881383855,
+      "grad_norm": 0.12392102926969528,
+      "learning_rate": 0.0002293973941368078,
+      "loss": 0.8161,
+      "step": 426
+    },
+    {
+      "epoch": 0.9379461834157057,
+      "grad_norm": 0.09947165846824646,
+      "learning_rate": 0.0002291530944625407,
+      "loss": 0.8546,
+      "step": 427
+    },
+    {
+      "epoch": 0.9401427786930258,
+      "grad_norm": 0.11522234976291656,
+      "learning_rate": 0.0002289087947882736,
+      "loss": 0.787,
+      "step": 428
+    },
+    {
+      "epoch": 0.942339373970346,
+      "grad_norm": 0.21582691371440887,
+      "learning_rate": 0.0002286644951140065,
+      "loss": 0.7897,
+      "step": 429
+    },
+    {
+      "epoch": 0.9445359692476661,
+      "grad_norm": 0.12030959874391556,
+      "learning_rate": 0.00022842019543973938,
+      "loss": 0.9382,
+      "step": 430
+    },
+    {
+      "epoch": 0.9467325645249862,
+      "grad_norm": 0.15318769216537476,
+      "learning_rate": 0.0002281758957654723,
+      "loss": 0.77,
+      "step": 431
+    },
+    {
+      "epoch": 0.9489291598023064,
+      "grad_norm": 0.12884899973869324,
+      "learning_rate": 0.0002279315960912052,
+      "loss": 1.0212,
+      "step": 432
+    },
+    {
+      "epoch": 0.9511257550796266,
+      "grad_norm": 0.20430032908916473,
+      "learning_rate": 0.0002276872964169381,
+      "loss": 0.9684,
+      "step": 433
+    },
+    {
+      "epoch": 0.9533223503569467,
+      "grad_norm": 0.13589294254779816,
+      "learning_rate": 0.000227442996742671,
+      "loss": 0.8307,
+      "step": 434
+    },
+    {
+      "epoch": 0.9555189456342669,
+      "grad_norm": 0.11975546181201935,
+      "learning_rate": 0.00022719869706840388,
+      "loss": 0.8346,
+      "step": 435
+    },
+    {
+      "epoch": 0.9577155409115871,
+      "grad_norm": 0.10656123608350754,
+      "learning_rate": 0.0002269543973941368,
+      "loss": 0.7334,
+      "step": 436
+    },
+    {
+      "epoch": 0.9599121361889071,
+      "grad_norm": 0.13536810874938965,
+      "learning_rate": 0.0002267100977198697,
+      "loss": 0.912,
+      "step": 437
+    },
+    {
+      "epoch": 0.9621087314662273,
+      "grad_norm": 0.1288425326347351,
+      "learning_rate": 0.00022646579804560258,
+      "loss": 0.7726,
+      "step": 438
+    },
+    {
+      "epoch": 0.9643053267435475,
+      "grad_norm": 0.28297072649002075,
+      "learning_rate": 0.0002262214983713355,
+      "loss": 1.127,
+      "step": 439
+    },
+    {
+      "epoch": 0.9665019220208677,
+      "grad_norm": 0.12722466886043549,
+      "learning_rate": 0.00022597719869706838,
+      "loss": 1.0057,
+      "step": 440
+    },
+    {
+      "epoch": 0.9686985172981878,
+      "grad_norm": 0.17841112613677979,
+      "learning_rate": 0.00022573289902280127,
+      "loss": 0.8428,
+      "step": 441
+    },
+    {
+      "epoch": 0.970895112575508,
+      "grad_norm": 0.12090222537517548,
+      "learning_rate": 0.0002254885993485342,
+      "loss": 0.8305,
+      "step": 442
+    },
+    {
+      "epoch": 0.9730917078528282,
+      "grad_norm": 0.12015727162361145,
+      "learning_rate": 0.00022524429967426708,
+      "loss": 0.9071,
+      "step": 443
+    },
+    {
+      "epoch": 0.9752883031301482,
+      "grad_norm": 0.14614488184452057,
+      "learning_rate": 0.000225,
+      "loss": 1.0436,
+      "step": 444
+    },
+    {
+      "epoch": 0.9774848984074684,
+      "grad_norm": 0.1627441942691803,
+      "learning_rate": 0.00022475570032573288,
+      "loss": 1.0326,
+      "step": 445
+    },
+    {
+      "epoch": 0.9796814936847886,
+      "grad_norm": 0.14776629209518433,
+      "learning_rate": 0.00022451140065146577,
+      "loss": 0.9897,
+      "step": 446
+    },
+    {
+      "epoch": 0.9818780889621087,
+      "grad_norm": 0.23793768882751465,
+      "learning_rate": 0.0002242671009771987,
+      "loss": 1.1024,
+      "step": 447
+    },
+    {
+      "epoch": 0.9840746842394289,
+      "grad_norm": 0.2322225719690323,
+      "learning_rate": 0.00022402280130293158,
+      "loss": 0.8824,
+      "step": 448
+    },
+    {
+      "epoch": 0.9862712795167491,
+      "grad_norm": 0.09890755265951157,
+      "learning_rate": 0.00022377850162866447,
+      "loss": 0.9695,
+      "step": 449
+    },
+    {
+      "epoch": 0.9884678747940692,
+      "grad_norm": 0.1557415872812271,
+      "learning_rate": 0.00022353420195439738,
+      "loss": 0.9163,
+      "step": 450
+    },
+    {
+      "epoch": 0.9906644700713894,
+      "grad_norm": 0.13875216245651245,
+      "learning_rate": 0.00022328990228013027,
+      "loss": 0.9186,
+      "step": 451
+    },
+    {
+      "epoch": 0.9928610653487095,
+      "grad_norm": 0.12742303311824799,
+      "learning_rate": 0.00022304560260586316,
+      "loss": 0.9313,
+      "step": 452
+    },
+    {
+      "epoch": 0.9950576606260296,
+      "grad_norm": 0.10524967312812805,
+      "learning_rate": 0.00022280130293159608,
+      "loss": 0.9857,
+      "step": 453
+    },
+    {
+      "epoch": 0.9972542559033498,
+      "grad_norm": 0.11670508980751038,
+      "learning_rate": 0.00022255700325732897,
+      "loss": 0.7964,
+      "step": 454
+    },
+    {
+      "epoch": 0.99945085118067,
+      "grad_norm": 0.1263129562139511,
+      "learning_rate": 0.00022231270358306186,
+      "loss": 0.8646,
+      "step": 455
+    },
+    {
+      "epoch": 1.00164744645799,
+      "grad_norm": 0.12352462857961655,
+      "learning_rate": 0.00022206840390879477,
+      "loss": 1.0347,
+      "step": 456
+    },
+    {
+      "epoch": 1.0038440417353103,
+      "grad_norm": 0.11953890323638916,
+      "learning_rate": 0.00022182410423452766,
+      "loss": 0.9569,
+      "step": 457
+    },
+    {
+      "epoch": 1.0060406370126305,
+      "grad_norm": 0.09742620587348938,
+      "learning_rate": 0.00022157980456026058,
+      "loss": 0.8796,
+      "step": 458
+    },
+    {
+      "epoch": 1.0082372322899507,
+      "grad_norm": 0.15897619724273682,
+      "learning_rate": 0.00022133550488599347,
+      "loss": 0.8549,
+      "step": 459
+    },
+    {
+      "epoch": 1.0104338275672706,
+      "grad_norm": 0.11113815009593964,
+      "learning_rate": 0.00022109120521172636,
+      "loss": 0.8123,
+      "step": 460
+    },
+    {
+      "epoch": 1.0126304228445908,
+      "grad_norm": 0.12016556411981583,
+      "learning_rate": 0.00022084690553745927,
+      "loss": 0.737,
+      "step": 461
+    },
+    {
+      "epoch": 1.014827018121911,
+      "grad_norm": 0.12541921436786652,
+      "learning_rate": 0.00022060260586319216,
+      "loss": 1.1899,
+      "step": 462
+    },
+    {
+      "epoch": 1.0170236133992312,
+      "grad_norm": 0.10914123058319092,
+      "learning_rate": 0.00022035830618892505,
+      "loss": 1.0704,
+      "step": 463
+    },
+    {
+      "epoch": 1.0192202086765514,
+      "grad_norm": 0.1266413927078247,
+      "learning_rate": 0.00022011400651465797,
+      "loss": 0.8706,
+      "step": 464
+    },
+    {
+      "epoch": 1.0214168039538716,
+      "grad_norm": 0.1230873391032219,
+      "learning_rate": 0.00021986970684039086,
+      "loss": 1.0777,
+      "step": 465
+    },
+    {
+      "epoch": 1.0236133992311915,
+      "grad_norm": 0.13624320924282074,
+      "learning_rate": 0.00021962540716612375,
+      "loss": 1.1265,
+      "step": 466
+    },
+    {
+      "epoch": 1.0258099945085117,
+      "grad_norm": 0.1346532702445984,
+      "learning_rate": 0.00021938110749185666,
+      "loss": 0.9405,
+      "step": 467
+    },
+    {
+      "epoch": 1.028006589785832,
+      "grad_norm": 0.229434996843338,
+      "learning_rate": 0.00021913680781758955,
+      "loss": 0.8704,
+      "step": 468
+    },
+    {
+      "epoch": 1.0302031850631521,
+      "grad_norm": 0.1134609580039978,
+      "learning_rate": 0.00021889250814332244,
+      "loss": 0.9814,
+      "step": 469
+    },
+    {
+      "epoch": 1.0323997803404723,
+      "grad_norm": 0.12965336441993713,
+      "learning_rate": 0.00021864820846905536,
+      "loss": 1.0623,
+      "step": 470
+    },
+    {
+      "epoch": 1.0345963756177925,
+      "grad_norm": 0.16243180632591248,
+      "learning_rate": 0.00021840390879478825,
+      "loss": 0.8355,
+      "step": 471
+    },
+    {
+      "epoch": 1.0367929708951125,
+      "grad_norm": 0.15026062726974487,
+      "learning_rate": 0.00021815960912052116,
+      "loss": 0.9013,
+      "step": 472
+    },
+    {
+      "epoch": 1.0389895661724327,
+      "grad_norm": 0.14699122309684753,
+      "learning_rate": 0.00021791530944625405,
+      "loss": 1.1404,
+      "step": 473
+    },
+    {
+      "epoch": 1.0411861614497528,
+      "grad_norm": 0.11705996096134186,
+      "learning_rate": 0.00021767100977198694,
+      "loss": 0.8812,
+      "step": 474
+    },
+    {
+      "epoch": 1.043382756727073,
+      "grad_norm": 0.10791568458080292,
+      "learning_rate": 0.00021742671009771986,
+      "loss": 0.8937,
+      "step": 475
+    },
+    {
+      "epoch": 1.0455793520043932,
+      "grad_norm": 0.10326795279979706,
+      "learning_rate": 0.00021718241042345275,
+      "loss": 1.1436,
+      "step": 476
+    },
+    {
+      "epoch": 1.0477759472817134,
+      "grad_norm": 0.2032373547554016,
+      "learning_rate": 0.00021693811074918564,
+      "loss": 0.8422,
+      "step": 477
+    },
+    {
+      "epoch": 1.0499725425590336,
+      "grad_norm": 0.10944756865501404,
+      "learning_rate": 0.00021669381107491855,
+      "loss": 0.9308,
+      "step": 478
+    },
+    {
+      "epoch": 1.0521691378363536,
+      "grad_norm": 0.11999942362308502,
+      "learning_rate": 0.00021644951140065144,
+      "loss": 0.9699,
+      "step": 479
+    },
+    {
+      "epoch": 1.0543657331136738,
+      "grad_norm": 0.12406887859106064,
+      "learning_rate": 0.00021620521172638433,
+      "loss": 0.7794,
+      "step": 480
+    },
+    {
+      "epoch": 1.056562328390994,
+      "grad_norm": 0.10555816441774368,
+      "learning_rate": 0.00021596091205211725,
+      "loss": 0.8277,
+      "step": 481
+    },
+    {
+      "epoch": 1.0587589236683141,
+      "grad_norm": 0.1672298163175583,
+      "learning_rate": 0.00021571661237785014,
+      "loss": 0.8624,
+      "step": 482
+    },
+    {
+      "epoch": 1.0609555189456343,
+      "grad_norm": 0.14567258954048157,
+      "learning_rate": 0.00021547231270358305,
+      "loss": 0.9291,
+      "step": 483
+    },
+    {
+      "epoch": 1.0631521142229543,
+      "grad_norm": 0.10755012929439545,
+      "learning_rate": 0.00021522801302931594,
+      "loss": 0.9582,
+      "step": 484
+    },
+    {
+      "epoch": 1.0653487095002745,
+      "grad_norm": 0.09260177612304688,
+      "learning_rate": 0.00021498371335504883,
+      "loss": 0.8898,
+      "step": 485
+    },
+    {
+      "epoch": 1.0675453047775947,
+      "grad_norm": 0.1625719666481018,
+      "learning_rate": 0.00021473941368078175,
+      "loss": 0.8288,
+      "step": 486
+    },
+    {
+      "epoch": 1.0697419000549149,
+      "grad_norm": 0.13732042908668518,
+      "learning_rate": 0.00021449511400651464,
+      "loss": 0.8504,
+      "step": 487
+    },
+    {
+      "epoch": 1.071938495332235,
+      "grad_norm": 0.12692949175834656,
+      "learning_rate": 0.00021425081433224753,
+      "loss": 1.0163,
+      "step": 488
+    },
+    {
+      "epoch": 1.0741350906095553,
+      "grad_norm": 0.12060361355543137,
+      "learning_rate": 0.00021400651465798044,
+      "loss": 0.8124,
+      "step": 489
+    },
+    {
+      "epoch": 1.0763316858868754,
+      "grad_norm": 0.1100107729434967,
+      "learning_rate": 0.00021376221498371333,
+      "loss": 0.7702,
+      "step": 490
+    },
+    {
+      "epoch": 1.0785282811641954,
+      "grad_norm": 0.15467298030853271,
+      "learning_rate": 0.00021351791530944622,
+      "loss": 1.076,
+      "step": 491
+    },
+    {
+      "epoch": 1.0807248764415156,
+      "grad_norm": 0.16019757091999054,
+      "learning_rate": 0.00021327361563517914,
+      "loss": 0.8392,
+      "step": 492
+    },
+    {
+      "epoch": 1.0829214717188358,
+      "grad_norm": 0.13633836805820465,
+      "learning_rate": 0.00021302931596091203,
+      "loss": 1.0028,
+      "step": 493
+    },
+    {
+      "epoch": 1.085118066996156,
+      "grad_norm": 0.11205071955919266,
+      "learning_rate": 0.00021278501628664491,
+      "loss": 0.7715,
+      "step": 494
+    },
+    {
+      "epoch": 1.0873146622734762,
+      "grad_norm": 0.1546461582183838,
+      "learning_rate": 0.00021254071661237783,
+      "loss": 0.886,
+      "step": 495
+    },
+    {
+      "epoch": 1.0895112575507964,
+      "grad_norm": 0.11212033778429031,
+      "learning_rate": 0.00021229641693811072,
+      "loss": 0.7734,
+      "step": 496
+    },
+    {
+      "epoch": 1.0917078528281163,
+      "grad_norm": 0.07747768610715866,
+      "learning_rate": 0.00021205211726384364,
+      "loss": 0.7633,
+      "step": 497
+    },
+    {
+      "epoch": 1.0939044481054365,
+      "grad_norm": 0.14608106017112732,
+      "learning_rate": 0.00021180781758957653,
+      "loss": 0.8873,
+      "step": 498
+    },
+    {
+      "epoch": 1.0961010433827567,
+      "grad_norm": 0.1442476361989975,
+      "learning_rate": 0.00021156351791530941,
+      "loss": 0.9558,
+      "step": 499
+    },
+    {
+      "epoch": 1.098297638660077,
+      "grad_norm": 0.15996138751506805,
+      "learning_rate": 0.00021131921824104233,
+      "loss": 0.9507,
+      "step": 500
+    },
+    {
+      "epoch": 1.100494233937397,
+      "grad_norm": 0.11033248901367188,
+      "learning_rate": 0.00021107491856677522,
+      "loss": 1.0364,
+      "step": 501
+    },
+    {
+      "epoch": 1.1026908292147173,
+      "grad_norm": 0.18023528158664703,
+      "learning_rate": 0.0002108306188925081,
+      "loss": 0.7825,
+      "step": 502
+    },
+    {
+      "epoch": 1.1048874244920373,
+      "grad_norm": 0.13797014951705933,
+      "learning_rate": 0.00021058631921824103,
+      "loss": 0.9486,
+      "step": 503
+    },
+    {
+      "epoch": 1.1070840197693574,
+      "grad_norm": 0.09844983369112015,
+      "learning_rate": 0.00021034201954397392,
+      "loss": 0.7711,
+      "step": 504
+    },
+    {
+      "epoch": 1.1092806150466776,
+      "grad_norm": 0.14193366467952728,
+      "learning_rate": 0.0002100977198697068,
+      "loss": 1.1424,
+      "step": 505
+    },
+    {
+      "epoch": 1.1114772103239978,
+      "grad_norm": 0.11808117479085922,
+      "learning_rate": 0.00020985342019543972,
+      "loss": 0.9048,
+      "step": 506
+    },
+    {
+      "epoch": 1.113673805601318,
+      "grad_norm": 0.122902050614357,
+      "learning_rate": 0.0002096091205211726,
+      "loss": 0.9875,
+      "step": 507
+    },
+    {
+      "epoch": 1.1158704008786382,
+      "grad_norm": 0.1283336728811264,
+      "learning_rate": 0.0002093648208469055,
+      "loss": 1.0239,
+      "step": 508
+    },
+    {
+      "epoch": 1.1180669961559582,
+      "grad_norm": 0.0826176330447197,
+      "learning_rate": 0.00020912052117263842,
+      "loss": 0.8802,
+      "step": 509
+    },
+    {
+      "epoch": 1.1202635914332784,
+      "grad_norm": 0.14805635809898376,
+      "learning_rate": 0.0002088762214983713,
+      "loss": 0.9966,
+      "step": 510
+    },
+    {
+      "epoch": 1.1224601867105986,
+      "grad_norm": 0.1393139511346817,
+      "learning_rate": 0.00020863192182410422,
+      "loss": 0.9014,
+      "step": 511
+    },
+    {
+      "epoch": 1.1246567819879187,
+      "grad_norm": 0.12868785858154297,
+      "learning_rate": 0.0002083876221498371,
+      "loss": 0.8943,
+      "step": 512
+    },
+    {
+      "epoch": 1.126853377265239,
+      "grad_norm": 0.16476362943649292,
+      "learning_rate": 0.00020814332247557,
+      "loss": 1.0424,
+      "step": 513
+    },
+    {
+      "epoch": 1.1290499725425591,
+      "grad_norm": 0.10821756720542908,
+      "learning_rate": 0.00020789902280130292,
+      "loss": 0.8516,
+      "step": 514
+    },
+    {
+      "epoch": 1.131246567819879,
+      "grad_norm": 0.11177735030651093,
+      "learning_rate": 0.0002076547231270358,
+      "loss": 0.8613,
+      "step": 515
+    },
+    {
+      "epoch": 1.1334431630971993,
+      "grad_norm": 0.1270567625761032,
+      "learning_rate": 0.0002074104234527687,
+      "loss": 0.9054,
+      "step": 516
+    },
+    {
+      "epoch": 1.1356397583745195,
+      "grad_norm": 0.15233540534973145,
+      "learning_rate": 0.0002071661237785016,
+      "loss": 0.9845,
+      "step": 517
+    },
+    {
+      "epoch": 1.1378363536518397,
+      "grad_norm": 0.1232580691576004,
+      "learning_rate": 0.0002069218241042345,
+      "loss": 0.7256,
+      "step": 518
+    },
+    {
+      "epoch": 1.1400329489291599,
+      "grad_norm": 0.12508700788021088,
+      "learning_rate": 0.0002066775244299674,
+      "loss": 0.691,
+      "step": 519
+    },
+    {
+      "epoch": 1.14222954420648,
+      "grad_norm": 0.14727623760700226,
+      "learning_rate": 0.0002064332247557003,
+      "loss": 0.9317,
+      "step": 520
+    },
+    {
+      "epoch": 1.1444261394838002,
+      "grad_norm": 0.13053067028522491,
+      "learning_rate": 0.0002061889250814332,
+      "loss": 0.786,
+      "step": 521
+    },
+    {
+      "epoch": 1.1466227347611202,
+      "grad_norm": 0.12321554869413376,
+      "learning_rate": 0.0002059446254071661,
+      "loss": 0.8582,
+      "step": 522
+    },
+    {
+      "epoch": 1.1488193300384404,
+      "grad_norm": 0.12414780259132385,
+      "learning_rate": 0.000205700325732899,
+      "loss": 0.6982,
+      "step": 523
+    },
+    {
+      "epoch": 1.1510159253157606,
+      "grad_norm": 0.1262628436088562,
+      "learning_rate": 0.0002054560260586319,
+      "loss": 1.0142,
+      "step": 524
+    },
+    {
+      "epoch": 1.1532125205930808,
+      "grad_norm": 0.10715745389461517,
+      "learning_rate": 0.0002052117263843648,
+      "loss": 0.8689,
+      "step": 525
+    },
+    {
+      "epoch": 1.155409115870401,
+      "grad_norm": 0.14395667612552643,
+      "learning_rate": 0.0002049674267100977,
+      "loss": 1.161,
+      "step": 526
+    },
+    {
+      "epoch": 1.157605711147721,
+      "grad_norm": 0.13535287976264954,
+      "learning_rate": 0.00020472312703583058,
+      "loss": 0.8077,
+      "step": 527
+    },
+    {
+      "epoch": 1.1598023064250411,
+      "grad_norm": 0.13736827671527863,
+      "learning_rate": 0.0002044788273615635,
+      "loss": 1.0011,
+      "step": 528
+    },
+    {
+      "epoch": 1.1619989017023613,
+      "grad_norm": 0.12646494805812836,
+      "learning_rate": 0.0002042345276872964,
+      "loss": 0.6827,
+      "step": 529
+    },
+    {
+      "epoch": 1.1641954969796815,
+      "grad_norm": 0.1230311393737793,
+      "learning_rate": 0.00020399022801302928,
+      "loss": 0.6759,
+      "step": 530
+    },
+    {
+      "epoch": 1.1663920922570017,
+      "grad_norm": 0.1283644437789917,
+      "learning_rate": 0.0002037459283387622,
+      "loss": 0.8476,
+      "step": 531
+    },
+    {
+      "epoch": 1.1685886875343219,
+      "grad_norm": 0.13082940876483917,
+      "learning_rate": 0.00020350162866449508,
+      "loss": 0.9902,
+      "step": 532
+    },
+    {
+      "epoch": 1.170785282811642,
+      "grad_norm": 0.11400187760591507,
+      "learning_rate": 0.00020325732899022797,
+      "loss": 0.978,
+      "step": 533
+    },
+    {
+      "epoch": 1.172981878088962,
+      "grad_norm": 0.1399601399898529,
+      "learning_rate": 0.0002030130293159609,
+      "loss": 1.0107,
+      "step": 534
+    },
+    {
+      "epoch": 1.1751784733662822,
+      "grad_norm": 0.11771988868713379,
+      "learning_rate": 0.00020276872964169378,
+      "loss": 0.887,
+      "step": 535
+    },
+    {
+      "epoch": 1.1773750686436024,
+      "grad_norm": 0.09674248844385147,
+      "learning_rate": 0.0002025244299674267,
+      "loss": 0.8012,
+      "step": 536
+    },
+    {
+      "epoch": 1.1795716639209226,
+      "grad_norm": 0.11043241620063782,
+      "learning_rate": 0.00020228013029315958,
+      "loss": 0.8373,
+      "step": 537
+    },
+    {
+      "epoch": 1.1817682591982428,
+      "grad_norm": 0.13775348663330078,
+      "learning_rate": 0.00020203583061889247,
+      "loss": 0.8214,
+      "step": 538
+    },
+    {
+      "epoch": 1.1839648544755628,
+      "grad_norm": 0.13636445999145508,
+      "learning_rate": 0.0002017915309446254,
+      "loss": 1.1635,
+      "step": 539
+    },
+    {
+      "epoch": 1.186161449752883,
+      "grad_norm": 0.11465809494256973,
+      "learning_rate": 0.00020154723127035828,
+      "loss": 0.7822,
+      "step": 540
+    },
+    {
+      "epoch": 1.1883580450302031,
+      "grad_norm": 0.10634935647249222,
+      "learning_rate": 0.00020130293159609117,
+      "loss": 0.897,
+      "step": 541
+    },
+    {
+      "epoch": 1.1905546403075233,
+      "grad_norm": 0.14827103912830353,
+      "learning_rate": 0.00020105863192182408,
+      "loss": 0.772,
+      "step": 542
+    },
+    {
+      "epoch": 1.1927512355848435,
+      "grad_norm": 0.11914302408695221,
+      "learning_rate": 0.00020081433224755697,
+      "loss": 0.9464,
+      "step": 543
+    },
+    {
+      "epoch": 1.1949478308621637,
+      "grad_norm": 0.10644625127315521,
+      "learning_rate": 0.00020057003257328986,
+      "loss": 0.8442,
+      "step": 544
+    },
+    {
+      "epoch": 1.197144426139484,
+      "grad_norm": 0.15938997268676758,
+      "learning_rate": 0.00020032573289902278,
+      "loss": 1.0133,
+      "step": 545
+    },
+    {
+      "epoch": 1.1993410214168039,
+      "grad_norm": 0.10189756006002426,
+      "learning_rate": 0.00020008143322475567,
+      "loss": 0.9064,
+      "step": 546
+    },
+    {
+      "epoch": 1.201537616694124,
+      "grad_norm": 0.14229203760623932,
+      "learning_rate": 0.00019983713355048856,
+      "loss": 1.2163,
+      "step": 547
+    },
+    {
+      "epoch": 1.2037342119714443,
+      "grad_norm": 0.12794840335845947,
+      "learning_rate": 0.00019959283387622147,
+      "loss": 0.8754,
+      "step": 548
+    },
+    {
+      "epoch": 1.2059308072487644,
+      "grad_norm": 0.10476420819759369,
+      "learning_rate": 0.00019934853420195436,
+      "loss": 1.093,
+      "step": 549
+    },
+    {
+      "epoch": 1.2081274025260846,
+      "grad_norm": 0.17878666520118713,
+      "learning_rate": 0.00019910423452768728,
+      "loss": 1.0325,
+      "step": 550
+    },
+    {
+      "epoch": 1.2103239978034046,
+      "grad_norm": 0.12591899931430817,
+      "learning_rate": 0.00019885993485342017,
+      "loss": 0.8528,
+      "step": 551
+    },
+    {
+      "epoch": 1.2125205930807248,
+      "grad_norm": 0.1547940969467163,
+      "learning_rate": 0.00019861563517915306,
+      "loss": 1.0173,
+      "step": 552
+    },
+    {
+      "epoch": 1.214717188358045,
+      "grad_norm": 0.1096530333161354,
+      "learning_rate": 0.00019837133550488597,
+      "loss": 0.9622,
+      "step": 553
+    },
+    {
+      "epoch": 1.2169137836353652,
+      "grad_norm": 0.1093997210264206,
+      "learning_rate": 0.00019812703583061886,
+      "loss": 0.9198,
+      "step": 554
+    },
+    {
+      "epoch": 1.2191103789126854,
+      "grad_norm": 0.1510469913482666,
+      "learning_rate": 0.00019788273615635175,
+      "loss": 0.9788,
+      "step": 555
+    },
+    {
+      "epoch": 1.2213069741900056,
+      "grad_norm": 0.14497597515583038,
+      "learning_rate": 0.00019763843648208467,
+      "loss": 0.7993,
+      "step": 556
+    },
+    {
+      "epoch": 1.2235035694673257,
+      "grad_norm": 0.13534414768218994,
+      "learning_rate": 0.00019739413680781756,
+      "loss": 0.943,
+      "step": 557
+    },
+    {
+      "epoch": 1.2257001647446457,
+      "grad_norm": 0.16626757383346558,
+      "learning_rate": 0.00019714983713355045,
+      "loss": 0.8564,
+      "step": 558
+    },
+    {
+      "epoch": 1.227896760021966,
+      "grad_norm": 0.18214108049869537,
+      "learning_rate": 0.00019690553745928336,
+      "loss": 1.1927,
+      "step": 559
+    },
+    {
+      "epoch": 1.230093355299286,
+      "grad_norm": 0.11493250727653503,
+      "learning_rate": 0.00019666123778501625,
+      "loss": 0.9667,
+      "step": 560
+    },
+    {
+      "epoch": 1.2322899505766063,
+      "grad_norm": 0.13930507004261017,
+      "learning_rate": 0.00019641693811074917,
+      "loss": 0.8856,
+      "step": 561
+    },
+    {
+      "epoch": 1.2344865458539265,
+      "grad_norm": 0.17520129680633545,
+      "learning_rate": 0.00019617263843648206,
+      "loss": 0.9295,
+      "step": 562
+    },
+    {
+      "epoch": 1.2366831411312464,
+      "grad_norm": 0.13444574177265167,
+      "learning_rate": 0.00019592833876221495,
+      "loss": 1.0619,
+      "step": 563
+    },
+    {
+      "epoch": 1.2388797364085666,
+      "grad_norm": 0.14415059983730316,
+      "learning_rate": 0.00019568403908794786,
+      "loss": 0.8161,
+      "step": 564
+    },
+    {
+      "epoch": 1.2410763316858868,
+      "grad_norm": 0.19687232375144958,
+      "learning_rate": 0.00019543973941368075,
+      "loss": 0.9891,
+      "step": 565
+    },
+    {
+      "epoch": 1.243272926963207,
+      "grad_norm": 0.15264511108398438,
+      "learning_rate": 0.00019519543973941364,
+      "loss": 0.902,
+      "step": 566
+    },
+    {
+      "epoch": 1.2454695222405272,
+      "grad_norm": 0.09933913499116898,
+      "learning_rate": 0.00019495114006514656,
+      "loss": 0.847,
+      "step": 567
+    },
+    {
+      "epoch": 1.2476661175178474,
+      "grad_norm": 0.14997830986976624,
+      "learning_rate": 0.00019470684039087945,
+      "loss": 0.9773,
+      "step": 568
+    },
+    {
+      "epoch": 1.2498627127951676,
+      "grad_norm": 0.13064095377922058,
+      "learning_rate": 0.00019446254071661234,
+      "loss": 0.7979,
+      "step": 569
+    },
+    {
+      "epoch": 1.2520593080724876,
+      "grad_norm": 0.12269759923219681,
+      "learning_rate": 0.00019421824104234525,
+      "loss": 0.6343,
+      "step": 570
+    },
+    {
+      "epoch": 1.2542559033498077,
+      "grad_norm": 0.1915825605392456,
+      "learning_rate": 0.00019397394136807814,
+      "loss": 1.062,
+      "step": 571
+    },
+    {
+      "epoch": 1.256452498627128,
+      "grad_norm": 0.15152528882026672,
+      "learning_rate": 0.00019372964169381103,
+      "loss": 0.9421,
+      "step": 572
+    },
+    {
+      "epoch": 1.2586490939044481,
+      "grad_norm": 0.13969701528549194,
+      "learning_rate": 0.00019348534201954395,
+      "loss": 0.8899,
+      "step": 573
+    },
+    {
+      "epoch": 1.2608456891817683,
+      "grad_norm": 0.11450473964214325,
+      "learning_rate": 0.00019324104234527684,
+      "loss": 0.9817,
+      "step": 574
+    },
+    {
+      "epoch": 1.2630422844590883,
+      "grad_norm": 0.11630494147539139,
+      "learning_rate": 0.00019299674267100975,
+      "loss": 0.9982,
+      "step": 575
+    },
+    {
+      "epoch": 1.2652388797364087,
+      "grad_norm": 0.11381109803915024,
+      "learning_rate": 0.00019275244299674264,
+      "loss": 0.8481,
+      "step": 576
+    },
+    {
+      "epoch": 1.2674354750137287,
+      "grad_norm": 0.12480124086141586,
+      "learning_rate": 0.00019250814332247553,
+      "loss": 1.1028,
+      "step": 577
+    },
+    {
+      "epoch": 1.2696320702910489,
+      "grad_norm": 0.34552475810050964,
+      "learning_rate": 0.00019226384364820845,
+      "loss": 0.658,
+      "step": 578
+    },
+    {
+      "epoch": 1.271828665568369,
+      "grad_norm": 0.16247911751270294,
+      "learning_rate": 0.00019201954397394134,
+      "loss": 0.6989,
+      "step": 579
+    },
+    {
+      "epoch": 1.2740252608456892,
+      "grad_norm": 0.10835195332765579,
+      "learning_rate": 0.00019177524429967423,
+      "loss": 0.7314,
+      "step": 580
+    },
+    {
+      "epoch": 1.2762218561230094,
+      "grad_norm": 0.12320885062217712,
+      "learning_rate": 0.00019153094462540714,
+      "loss": 0.9698,
+      "step": 581
+    },
+    {
+      "epoch": 1.2784184514003294,
+      "grad_norm": 0.16261093318462372,
+      "learning_rate": 0.00019128664495114003,
+      "loss": 0.854,
+      "step": 582
+    },
+    {
+      "epoch": 1.2806150466776496,
+      "grad_norm": 0.08921562135219574,
+      "learning_rate": 0.00019104234527687292,
+      "loss": 0.8191,
+      "step": 583
+    },
+    {
+      "epoch": 1.2828116419549698,
+      "grad_norm": 0.23219412565231323,
+      "learning_rate": 0.00019079804560260584,
+      "loss": 0.8434,
+      "step": 584
+    },
+    {
+      "epoch": 1.28500823723229,
+      "grad_norm": 0.11112730950117111,
+      "learning_rate": 0.00019055374592833873,
+      "loss": 0.9127,
+      "step": 585
+    },
+    {
+      "epoch": 1.2872048325096102,
+      "grad_norm": 0.09262687712907791,
+      "learning_rate": 0.00019030944625407164,
+      "loss": 0.8598,
+      "step": 586
+    },
+    {
+      "epoch": 1.2894014277869301,
+      "grad_norm": 0.12439727038145065,
+      "learning_rate": 0.00019006514657980453,
+      "loss": 0.8828,
+      "step": 587
+    },
+    {
+      "epoch": 1.2915980230642505,
+      "grad_norm": 0.11980634927749634,
+      "learning_rate": 0.00018982084690553742,
+      "loss": 0.9491,
+      "step": 588
+    },
+    {
+      "epoch": 1.2937946183415705,
+      "grad_norm": 0.1995294839143753,
+      "learning_rate": 0.00018957654723127034,
+      "loss": 0.8797,
+      "step": 589
+    },
+    {
+      "epoch": 1.2959912136188907,
+      "grad_norm": 0.11529744416475296,
+      "learning_rate": 0.00018933224755700323,
+      "loss": 0.9009,
+      "step": 590
+    },
+    {
+      "epoch": 1.2981878088962109,
+      "grad_norm": 0.08105169236660004,
+      "learning_rate": 0.00018908794788273612,
+      "loss": 0.6717,
+      "step": 591
+    },
+    {
+      "epoch": 1.300384404173531,
+      "grad_norm": 0.10464441776275635,
+      "learning_rate": 0.00018884364820846903,
+      "loss": 0.9301,
+      "step": 592
+    },
+    {
+      "epoch": 1.3025809994508513,
+      "grad_norm": 0.09852095693349838,
+      "learning_rate": 0.00018859934853420192,
+      "loss": 0.7886,
+      "step": 593
+    },
+    {
+      "epoch": 1.3047775947281712,
+      "grad_norm": 0.11155886203050613,
+      "learning_rate": 0.0001883550488599348,
+      "loss": 0.9232,
+      "step": 594
+    },
+    {
+      "epoch": 1.3069741900054914,
+      "grad_norm": 0.15545082092285156,
+      "learning_rate": 0.00018811074918566773,
+      "loss": 1.076,
+      "step": 595
+    },
+    {
+      "epoch": 1.3091707852828116,
+      "grad_norm": 0.11825481057167053,
+      "learning_rate": 0.00018786644951140062,
+      "loss": 0.8115,
+      "step": 596
+    },
+    {
+      "epoch": 1.3113673805601318,
+      "grad_norm": 0.11334706842899323,
+      "learning_rate": 0.0001876221498371335,
+      "loss": 0.8273,
+      "step": 597
+    },
+    {
+      "epoch": 1.313563975837452,
+      "grad_norm": 0.14000928401947021,
+      "learning_rate": 0.00018737785016286645,
+      "loss": 0.7728,
+      "step": 598
+    },
+    {
+      "epoch": 1.3157605711147722,
+      "grad_norm": 0.1172076016664505,
+      "learning_rate": 0.00018713355048859934,
+      "loss": 0.8903,
+      "step": 599
+    },
+    {
+      "epoch": 1.3179571663920924,
+      "grad_norm": 0.10422193259000778,
+      "learning_rate": 0.00018688925081433225,
+      "loss": 1.0481,
+      "step": 600
+    },
+    {
+      "epoch": 1.3201537616694123,
+      "grad_norm": 0.13573327660560608,
+      "learning_rate": 0.00018664495114006514,
+      "loss": 0.9122,
+      "step": 601
+    },
+    {
+      "epoch": 1.3223503569467325,
+      "grad_norm": 0.08036984503269196,
+      "learning_rate": 0.00018640065146579803,
+      "loss": 0.815,
+      "step": 602
+    },
+    {
+      "epoch": 1.3245469522240527,
+      "grad_norm": 0.20787370204925537,
+      "learning_rate": 0.00018615635179153095,
+      "loss": 0.9997,
+      "step": 603
+    },
+    {
+      "epoch": 1.326743547501373,
+      "grad_norm": 0.14923392236232758,
+      "learning_rate": 0.00018591205211726384,
+      "loss": 0.9391,
+      "step": 604
+    },
+    {
+      "epoch": 1.328940142778693,
+      "grad_norm": 0.1163204088807106,
+      "learning_rate": 0.00018566775244299675,
+      "loss": 1.0032,
+      "step": 605
+    },
+    {
+      "epoch": 1.331136738056013,
+      "grad_norm": 0.1008303090929985,
+      "learning_rate": 0.00018542345276872964,
+      "loss": 0.7659,
+      "step": 606
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.1375746726989746,
+      "learning_rate": 0.00018517915309446253,
+      "loss": 0.6844,
+      "step": 607
+    },
+    {
+      "epoch": 1.3355299286106534,
+      "grad_norm": 0.11015655100345612,
+      "learning_rate": 0.00018493485342019545,
+      "loss": 0.8953,
+      "step": 608
+    },
+    {
+      "epoch": 1.3377265238879736,
+      "grad_norm": 0.15311221778392792,
+      "learning_rate": 0.00018469055374592834,
+      "loss": 0.9244,
+      "step": 609
+    },
+    {
+      "epoch": 1.3399231191652938,
+      "grad_norm": 0.11998546868562698,
+      "learning_rate": 0.00018444625407166123,
+      "loss": 0.941,
+      "step": 610
+    },
+    {
+      "epoch": 1.342119714442614,
+      "grad_norm": 0.11638596653938293,
+      "learning_rate": 0.00018420195439739414,
+      "loss": 0.9288,
+      "step": 611
+    },
+    {
+      "epoch": 1.3443163097199342,
+      "grad_norm": 0.11609652638435364,
+      "learning_rate": 0.00018395765472312703,
+      "loss": 0.7958,
+      "step": 612
+    },
+    {
+      "epoch": 1.3465129049972542,
+      "grad_norm": 0.14922203123569489,
+      "learning_rate": 0.00018371335504885992,
+      "loss": 0.9488,
+      "step": 613
+    },
+    {
+      "epoch": 1.3487095002745744,
+      "grad_norm": 0.11704708635807037,
+      "learning_rate": 0.00018346905537459284,
+      "loss": 0.8325,
+      "step": 614
+    },
+    {
+      "epoch": 1.3509060955518946,
+      "grad_norm": 0.11375498026609421,
+      "learning_rate": 0.00018322475570032573,
+      "loss": 1.0293,
+      "step": 615
+    },
+    {
+      "epoch": 1.3531026908292147,
+      "grad_norm": 0.15464071929454803,
+      "learning_rate": 0.00018298045602605864,
+      "loss": 0.9836,
+      "step": 616
+    },
+    {
+      "epoch": 1.355299286106535,
+      "grad_norm": 0.120681531727314,
+      "learning_rate": 0.00018273615635179153,
+      "loss": 0.9044,
+      "step": 617
+    },
+    {
+      "epoch": 1.357495881383855,
+      "grad_norm": 0.14832864701747894,
+      "learning_rate": 0.00018249185667752442,
+      "loss": 1.1233,
+      "step": 618
+    },
+    {
+      "epoch": 1.359692476661175,
+      "grad_norm": 0.13344135880470276,
+      "learning_rate": 0.00018224755700325734,
+      "loss": 0.7321,
+      "step": 619
+    },
+    {
+      "epoch": 1.3618890719384953,
+      "grad_norm": 0.12168313562870026,
+      "learning_rate": 0.00018200325732899023,
+      "loss": 0.823,
+      "step": 620
+    },
+    {
+      "epoch": 1.3640856672158155,
+      "grad_norm": 0.14547699689865112,
+      "learning_rate": 0.00018175895765472312,
+      "loss": 1.1099,
+      "step": 621
+    },
+    {
+      "epoch": 1.3662822624931357,
+      "grad_norm": 0.09874589741230011,
+      "learning_rate": 0.00018151465798045603,
+      "loss": 0.8498,
+      "step": 622
+    },
+    {
+      "epoch": 1.3684788577704559,
+      "grad_norm": 0.11414291709661484,
+      "learning_rate": 0.00018127035830618892,
+      "loss": 0.7547,
+      "step": 623
+    },
+    {
+      "epoch": 1.370675453047776,
+      "grad_norm": 0.13768859207630157,
+      "learning_rate": 0.0001810260586319218,
+      "loss": 0.9257,
+      "step": 624
+    },
+    {
+      "epoch": 1.372872048325096,
+      "grad_norm": 0.1421048492193222,
+      "learning_rate": 0.00018078175895765473,
+      "loss": 0.9813,
+      "step": 625
+    },
+    {
+      "epoch": 1.3750686436024162,
+      "grad_norm": 0.08750153332948685,
+      "learning_rate": 0.00018053745928338762,
+      "loss": 0.7925,
+      "step": 626
+    },
+    {
+      "epoch": 1.3772652388797364,
+      "grad_norm": 0.08550368249416351,
+      "learning_rate": 0.0001802931596091205,
+      "loss": 0.815,
+      "step": 627
+    },
+    {
+      "epoch": 1.3794618341570566,
+      "grad_norm": 0.12753528356552124,
+      "learning_rate": 0.00018004885993485342,
+      "loss": 0.6825,
+      "step": 628
+    },
+    {
+      "epoch": 1.3816584294343768,
+      "grad_norm": 0.14189237356185913,
+      "learning_rate": 0.0001798045602605863,
+      "loss": 0.8547,
+      "step": 629
+    },
+    {
+      "epoch": 1.3838550247116967,
+      "grad_norm": 0.14211580157279968,
+      "learning_rate": 0.00017956026058631923,
+      "loss": 0.8385,
+      "step": 630
+    },
+    {
+      "epoch": 1.3860516199890172,
+      "grad_norm": 0.10314823687076569,
+      "learning_rate": 0.00017931596091205212,
+      "loss": 0.8367,
+      "step": 631
+    },
+    {
+      "epoch": 1.3882482152663371,
+      "grad_norm": 0.11636725813150406,
+      "learning_rate": 0.000179071661237785,
+      "loss": 1.0586,
+      "step": 632
+    },
+    {
+      "epoch": 1.3904448105436573,
+      "grad_norm": 0.10327992588281631,
+      "learning_rate": 0.00017882736156351792,
+      "loss": 0.8606,
+      "step": 633
+    },
+    {
+      "epoch": 1.3926414058209775,
+      "grad_norm": 0.09629005938768387,
+      "learning_rate": 0.0001785830618892508,
+      "loss": 0.8157,
+      "step": 634
+    },
+    {
+      "epoch": 1.3948380010982977,
+      "grad_norm": 0.109748475253582,
+      "learning_rate": 0.0001783387622149837,
+      "loss": 0.8558,
+      "step": 635
+    },
+    {
+      "epoch": 1.3970345963756179,
+      "grad_norm": 0.15101934969425201,
+      "learning_rate": 0.00017809446254071662,
+      "loss": 1.0308,
+      "step": 636
+    },
+    {
+      "epoch": 1.3992311916529379,
+      "grad_norm": 0.13724303245544434,
+      "learning_rate": 0.0001778501628664495,
+      "loss": 0.8514,
+      "step": 637
+    },
+    {
+      "epoch": 1.401427786930258,
+      "grad_norm": 0.11386612057685852,
+      "learning_rate": 0.0001776058631921824,
+      "loss": 0.8192,
+      "step": 638
+    },
+    {
+      "epoch": 1.4036243822075782,
+      "grad_norm": 0.16043232381343842,
+      "learning_rate": 0.0001773615635179153,
+      "loss": 0.8499,
+      "step": 639
+    },
+    {
+      "epoch": 1.4058209774848984,
+      "grad_norm": 0.10591732710599899,
+      "learning_rate": 0.0001771172638436482,
+      "loss": 1.1152,
+      "step": 640
+    },
+    {
+      "epoch": 1.4080175727622186,
+      "grad_norm": 0.1334570199251175,
+      "learning_rate": 0.0001768729641693811,
+      "loss": 0.8322,
+      "step": 641
+    },
+    {
+      "epoch": 1.4102141680395386,
+      "grad_norm": 0.13207300007343292,
+      "learning_rate": 0.000176628664495114,
+      "loss": 1.0413,
+      "step": 642
+    },
+    {
+      "epoch": 1.412410763316859,
+      "grad_norm": 0.28668177127838135,
+      "learning_rate": 0.0001763843648208469,
+      "loss": 0.8736,
+      "step": 643
+    },
+    {
+      "epoch": 1.414607358594179,
+      "grad_norm": 0.09645982086658478,
+      "learning_rate": 0.0001761400651465798,
+      "loss": 0.742,
+      "step": 644
+    },
+    {
+      "epoch": 1.4168039538714992,
+      "grad_norm": 0.2059766948223114,
+      "learning_rate": 0.0001758957654723127,
+      "loss": 0.8617,
+      "step": 645
+    },
+    {
+      "epoch": 1.4190005491488193,
+      "grad_norm": 0.15119031071662903,
+      "learning_rate": 0.0001756514657980456,
+      "loss": 0.8509,
+      "step": 646
+    },
+    {
+      "epoch": 1.4211971444261395,
+      "grad_norm": 0.1710873544216156,
+      "learning_rate": 0.0001754071661237785,
+      "loss": 0.8479,
+      "step": 647
+    },
+    {
+      "epoch": 1.4233937397034597,
+      "grad_norm": 0.21338628232479095,
+      "learning_rate": 0.0001751628664495114,
+      "loss": 0.8165,
+      "step": 648
+    },
+    {
+      "epoch": 1.4255903349807797,
+      "grad_norm": 0.15700800716876984,
+      "learning_rate": 0.00017491856677524429,
+      "loss": 0.9531,
+      "step": 649
+    },
+    {
+      "epoch": 1.4277869302580999,
+      "grad_norm": 0.12311442941427231,
+      "learning_rate": 0.0001746742671009772,
+      "loss": 0.7548,
+      "step": 650
+    },
+    {
+      "epoch": 1.42998352553542,
+      "grad_norm": 0.12723010778427124,
+      "learning_rate": 0.0001744299674267101,
+      "loss": 0.8149,
+      "step": 651
+    },
+    {
+      "epoch": 1.4321801208127403,
+      "grad_norm": 0.10200917720794678,
+      "learning_rate": 0.00017418566775244298,
+      "loss": 0.8794,
+      "step": 652
+    },
+    {
+      "epoch": 1.4343767160900605,
+      "grad_norm": 0.1130010336637497,
+      "learning_rate": 0.0001739413680781759,
+      "loss": 0.8996,
+      "step": 653
+    },
+    {
+      "epoch": 1.4365733113673804,
+      "grad_norm": 0.13708995282649994,
+      "learning_rate": 0.00017369706840390879,
+      "loss": 0.8804,
+      "step": 654
+    },
+    {
+      "epoch": 1.4387699066447008,
+      "grad_norm": 0.1232062503695488,
+      "learning_rate": 0.0001734527687296417,
+      "loss": 0.8756,
+      "step": 655
+    },
+    {
+      "epoch": 1.4409665019220208,
+      "grad_norm": 0.12290848791599274,
+      "learning_rate": 0.0001732084690553746,
+      "loss": 0.914,
+      "step": 656
+    },
+    {
+      "epoch": 1.443163097199341,
+      "grad_norm": 0.11608665436506271,
+      "learning_rate": 0.00017296416938110748,
+      "loss": 1.0692,
+      "step": 657
+    },
+    {
+      "epoch": 1.4453596924766612,
+      "grad_norm": 0.10378572344779968,
+      "learning_rate": 0.0001727198697068404,
+      "loss": 0.974,
+      "step": 658
+    },
+    {
+      "epoch": 1.4475562877539814,
+      "grad_norm": 0.11187213659286499,
+      "learning_rate": 0.00017247557003257329,
+      "loss": 1.0405,
+      "step": 659
+    },
+    {
+      "epoch": 1.4497528830313016,
+      "grad_norm": 0.08394558727741241,
+      "learning_rate": 0.00017223127035830618,
+      "loss": 0.7802,
+      "step": 660
+    },
+    {
+      "epoch": 1.4519494783086215,
+      "grad_norm": 0.10943625122308731,
+      "learning_rate": 0.0001719869706840391,
+      "loss": 0.7626,
+      "step": 661
+    },
+    {
+      "epoch": 1.4541460735859417,
+      "grad_norm": 0.1991468071937561,
+      "learning_rate": 0.00017174267100977198,
+      "loss": 0.9317,
+      "step": 662
+    },
+    {
+      "epoch": 1.456342668863262,
+      "grad_norm": 0.14255008101463318,
+      "learning_rate": 0.00017149837133550487,
+      "loss": 0.9079,
+      "step": 663
+    },
+    {
+      "epoch": 1.458539264140582,
+      "grad_norm": 0.08645475655794144,
+      "learning_rate": 0.00017125407166123779,
+      "loss": 0.7694,
+      "step": 664
+    },
+    {
+      "epoch": 1.4607358594179023,
+      "grad_norm": 0.1551288515329361,
+      "learning_rate": 0.00017100977198697068,
+      "loss": 1.0687,
+      "step": 665
+    },
+    {
+      "epoch": 1.4629324546952225,
+      "grad_norm": 0.135439932346344,
+      "learning_rate": 0.00017076547231270356,
+      "loss": 1.0414,
+      "step": 666
+    },
+    {
+      "epoch": 1.4651290499725427,
+      "grad_norm": 0.11133774369955063,
+      "learning_rate": 0.00017052117263843648,
+      "loss": 0.8298,
+      "step": 667
+    },
+    {
+      "epoch": 1.4673256452498626,
+      "grad_norm": 0.10870947688817978,
+      "learning_rate": 0.00017027687296416937,
+      "loss": 0.9939,
+      "step": 668
+    },
+    {
+      "epoch": 1.4695222405271828,
+      "grad_norm": 0.12509098649024963,
+      "learning_rate": 0.00017003257328990229,
+      "loss": 0.7844,
+      "step": 669
+    },
+    {
+      "epoch": 1.471718835804503,
+      "grad_norm": 0.1253507435321808,
+      "learning_rate": 0.00016978827361563518,
+      "loss": 0.8973,
+      "step": 670
+    },
+    {
+      "epoch": 1.4739154310818232,
+      "grad_norm": 0.21508567035198212,
+      "learning_rate": 0.00016954397394136806,
+      "loss": 0.9953,
+      "step": 671
+    },
+    {
+      "epoch": 1.4761120263591434,
+      "grad_norm": 0.1221606582403183,
+      "learning_rate": 0.00016929967426710098,
+      "loss": 0.933,
+      "step": 672
+    },
+    {
+      "epoch": 1.4783086216364634,
+      "grad_norm": 0.12343398481607437,
+      "learning_rate": 0.00016905537459283387,
+      "loss": 0.818,
+      "step": 673
+    },
+    {
+      "epoch": 1.4805052169137836,
+      "grad_norm": 0.10601703077554703,
+      "learning_rate": 0.00016881107491856676,
+      "loss": 0.9526,
+      "step": 674
+    },
+    {
+      "epoch": 1.4827018121911038,
+      "grad_norm": 0.11659565567970276,
+      "learning_rate": 0.00016856677524429968,
+      "loss": 0.7822,
+      "step": 675
+    },
+    {
+      "epoch": 1.484898407468424,
+      "grad_norm": 0.1104501485824585,
+      "learning_rate": 0.00016832247557003257,
+      "loss": 0.8474,
+      "step": 676
+    },
+    {
+      "epoch": 1.4870950027457441,
+      "grad_norm": 0.10485420376062393,
+      "learning_rate": 0.00016807817589576545,
+      "loss": 0.8576,
+      "step": 677
+    },
+    {
+      "epoch": 1.4892915980230643,
+      "grad_norm": 0.14655143022537231,
+      "learning_rate": 0.00016783387622149837,
+      "loss": 0.9752,
+      "step": 678
+    },
+    {
+      "epoch": 1.4914881933003845,
+      "grad_norm": 0.09909292310476303,
+      "learning_rate": 0.00016758957654723126,
+      "loss": 0.9234,
+      "step": 679
+    },
+    {
+      "epoch": 1.4936847885777045,
+      "grad_norm": 0.17876508831977844,
+      "learning_rate": 0.00016734527687296415,
+      "loss": 1.0725,
+      "step": 680
+    },
+    {
+      "epoch": 1.4958813838550247,
+      "grad_norm": 0.12898045778274536,
+      "learning_rate": 0.00016710097719869707,
+      "loss": 0.9119,
+      "step": 681
+    },
+    {
+      "epoch": 1.4980779791323449,
+      "grad_norm": 0.12947189807891846,
+      "learning_rate": 0.00016685667752442995,
+      "loss": 1.066,
+      "step": 682
+    },
+    {
+      "epoch": 1.500274574409665,
+      "grad_norm": 0.10877087712287903,
+      "learning_rate": 0.00016661237785016287,
+      "loss": 0.6399,
+      "step": 683
+    },
+    {
+      "epoch": 1.5024711696869852,
+      "grad_norm": 0.1593281775712967,
+      "learning_rate": 0.00016636807817589576,
+      "loss": 0.7755,
+      "step": 684
+    },
+    {
+      "epoch": 1.5046677649643052,
+      "grad_norm": 0.12966813147068024,
+      "learning_rate": 0.00016612377850162865,
+      "loss": 0.9667,
+      "step": 685
+    },
+    {
+      "epoch": 1.5068643602416256,
+      "grad_norm": 0.18489818274974823,
+      "learning_rate": 0.00016587947882736157,
+      "loss": 1.1204,
+      "step": 686
+    },
+    {
+      "epoch": 1.5090609555189456,
+      "grad_norm": 0.13272960484027863,
+      "learning_rate": 0.00016563517915309445,
+      "loss": 0.9426,
+      "step": 687
+    },
+    {
+      "epoch": 1.5112575507962658,
+      "grad_norm": 0.1426813006401062,
+      "learning_rate": 0.00016539087947882734,
+      "loss": 1.0307,
+      "step": 688
+    },
+    {
+      "epoch": 1.513454146073586,
+      "grad_norm": 0.1931968629360199,
+      "learning_rate": 0.00016514657980456026,
+      "loss": 0.8254,
+      "step": 689
+    },
+    {
+      "epoch": 1.515650741350906,
+      "grad_norm": 0.10640034824609756,
+      "learning_rate": 0.00016490228013029315,
+      "loss": 0.8013,
+      "step": 690
+    },
+    {
+      "epoch": 1.5178473366282264,
+      "grad_norm": 0.1861136108636856,
+      "learning_rate": 0.00016465798045602604,
+      "loss": 0.9776,
+      "step": 691
+    },
+    {
+      "epoch": 1.5200439319055463,
+      "grad_norm": 0.12281373888254166,
+      "learning_rate": 0.00016441368078175895,
+      "loss": 0.8831,
+      "step": 692
+    },
+    {
+      "epoch": 1.5222405271828665,
+      "grad_norm": 0.12226267158985138,
+      "learning_rate": 0.00016416938110749184,
+      "loss": 0.902,
+      "step": 693
+    },
+    {
+      "epoch": 1.5244371224601867,
+      "grad_norm": 0.11832962930202484,
+      "learning_rate": 0.00016392508143322476,
+      "loss": 0.8099,
+      "step": 694
+    },
+    {
+      "epoch": 1.526633717737507,
+      "grad_norm": 0.09061731398105621,
+      "learning_rate": 0.00016368078175895765,
+      "loss": 0.8485,
+      "step": 695
+    },
+    {
+      "epoch": 1.528830313014827,
+      "grad_norm": 0.0978364497423172,
+      "learning_rate": 0.00016343648208469054,
+      "loss": 0.7945,
+      "step": 696
+    },
+    {
+      "epoch": 1.531026908292147,
+      "grad_norm": 0.11006689071655273,
+      "learning_rate": 0.00016319218241042346,
+      "loss": 0.8817,
+      "step": 697
+    },
+    {
+      "epoch": 1.5332235035694675,
+      "grad_norm": 0.09524192661046982,
+      "learning_rate": 0.00016294788273615634,
+      "loss": 0.8417,
+      "step": 698
+    },
+    {
+      "epoch": 1.5354200988467874,
+      "grad_norm": 0.16973896324634552,
+      "learning_rate": 0.00016270358306188923,
+      "loss": 0.9918,
+      "step": 699
+    },
+    {
+      "epoch": 1.5376166941241076,
+      "grad_norm": 0.09180596470832825,
+      "learning_rate": 0.00016245928338762215,
+      "loss": 1.0115,
+      "step": 700
+    },
+    {
+      "epoch": 1.5398132894014278,
+      "grad_norm": 0.11738311499357224,
+      "learning_rate": 0.00016221498371335504,
+      "loss": 0.8146,
+      "step": 701
+    },
+    {
+      "epoch": 1.5420098846787478,
+      "grad_norm": 0.12505404651165009,
+      "learning_rate": 0.00016197068403908793,
+      "loss": 0.8858,
+      "step": 702
+    },
+    {
+      "epoch": 1.5442064799560682,
+      "grad_norm": 0.3418639302253723,
+      "learning_rate": 0.00016172638436482084,
+      "loss": 0.7981,
+      "step": 703
+    },
+    {
+      "epoch": 1.5464030752333882,
+      "grad_norm": 0.19170954823493958,
+      "learning_rate": 0.00016148208469055373,
+      "loss": 0.7313,
+      "step": 704
+    },
+    {
+      "epoch": 1.5485996705107083,
+      "grad_norm": 0.1381804347038269,
+      "learning_rate": 0.00016123778501628662,
+      "loss": 0.9084,
+      "step": 705
+    },
+    {
+      "epoch": 1.5507962657880285,
+      "grad_norm": 0.13961224257946014,
+      "learning_rate": 0.00016099348534201954,
+      "loss": 0.8186,
+      "step": 706
+    },
+    {
+      "epoch": 1.5529928610653487,
+      "grad_norm": 0.13189095258712769,
+      "learning_rate": 0.00016074918566775243,
+      "loss": 0.921,
+      "step": 707
+    },
+    {
+      "epoch": 1.555189456342669,
+      "grad_norm": 0.17217697203159332,
+      "learning_rate": 0.00016050488599348534,
+      "loss": 1.0244,
+      "step": 708
+    },
+    {
+      "epoch": 1.5573860516199889,
+      "grad_norm": 0.14158421754837036,
+      "learning_rate": 0.00016026058631921823,
+      "loss": 1.0419,
+      "step": 709
+    },
+    {
+      "epoch": 1.5595826468973093,
+      "grad_norm": 0.1306632161140442,
+      "learning_rate": 0.00016001628664495112,
+      "loss": 0.6946,
+      "step": 710
+    },
+    {
+      "epoch": 1.5617792421746293,
+      "grad_norm": 0.13527843356132507,
+      "learning_rate": 0.00015977198697068404,
+      "loss": 1.0026,
+      "step": 711
+    },
+    {
+      "epoch": 1.5639758374519495,
+      "grad_norm": 0.10221362113952637,
+      "learning_rate": 0.00015952768729641693,
+      "loss": 0.6937,
+      "step": 712
+    },
+    {
+      "epoch": 1.5661724327292696,
+      "grad_norm": 0.1362054944038391,
+      "learning_rate": 0.00015928338762214982,
+      "loss": 0.8205,
+      "step": 713
+    },
+    {
+      "epoch": 1.5683690280065898,
+      "grad_norm": 0.09425926208496094,
+      "learning_rate": 0.00015903908794788273,
+      "loss": 0.7527,
+      "step": 714
+    },
+    {
+      "epoch": 1.57056562328391,
+      "grad_norm": 0.1400139033794403,
+      "learning_rate": 0.00015879478827361562,
+      "loss": 0.9037,
+      "step": 715
+    },
+    {
+      "epoch": 1.57276221856123,
+      "grad_norm": 0.12442310899496078,
+      "learning_rate": 0.0001585504885993485,
+      "loss": 1.0432,
+      "step": 716
+    },
+    {
+      "epoch": 1.5749588138385504,
+      "grad_norm": 0.1035342589020729,
+      "learning_rate": 0.00015830618892508143,
+      "loss": 0.8774,
+      "step": 717
+    },
+    {
+      "epoch": 1.5771554091158704,
+      "grad_norm": 0.15819863975048065,
+      "learning_rate": 0.00015806188925081432,
+      "loss": 0.8483,
+      "step": 718
+    },
+    {
+      "epoch": 1.5793520043931906,
+      "grad_norm": 0.1131882518529892,
+      "learning_rate": 0.00015781758957654723,
+      "loss": 0.8302,
+      "step": 719
+    },
+    {
+      "epoch": 1.5815485996705108,
+      "grad_norm": 0.15275083482265472,
+      "learning_rate": 0.00015757328990228012,
+      "loss": 0.8143,
+      "step": 720
+    },
+    {
+      "epoch": 1.5837451949478307,
+      "grad_norm": 0.14359870553016663,
+      "learning_rate": 0.000157328990228013,
+      "loss": 0.9724,
+      "step": 721
+    },
+    {
+      "epoch": 1.5859417902251511,
+      "grad_norm": 0.1490335613489151,
+      "learning_rate": 0.00015708469055374593,
+      "loss": 0.9982,
+      "step": 722
+    },
+    {
+      "epoch": 1.588138385502471,
+      "grad_norm": 0.0880749374628067,
+      "learning_rate": 0.00015684039087947882,
+      "loss": 0.8095,
+      "step": 723
+    },
+    {
+      "epoch": 1.5903349807797913,
+      "grad_norm": 0.11849076300859451,
+      "learning_rate": 0.0001565960912052117,
+      "loss": 0.7337,
+      "step": 724
+    },
+    {
+      "epoch": 1.5925315760571115,
+      "grad_norm": 0.10837572067975998,
+      "learning_rate": 0.00015635179153094462,
+      "loss": 0.8949,
+      "step": 725
+    },
+    {
+      "epoch": 1.5947281713344317,
+      "grad_norm": 0.11553801596164703,
+      "learning_rate": 0.0001561074918566775,
+      "loss": 0.8587,
+      "step": 726
+    },
+    {
+      "epoch": 1.5969247666117519,
+      "grad_norm": 0.3137112855911255,
+      "learning_rate": 0.0001558631921824104,
+      "loss": 0.9208,
+      "step": 727
+    },
+    {
+      "epoch": 1.5991213618890718,
+      "grad_norm": 0.09689981490373611,
+      "learning_rate": 0.00015561889250814332,
+      "loss": 0.9363,
+      "step": 728
+    },
+    {
+      "epoch": 1.6013179571663922,
+      "grad_norm": 0.17914997041225433,
+      "learning_rate": 0.0001553745928338762,
+      "loss": 0.9,
+      "step": 729
+    },
+    {
+      "epoch": 1.6035145524437122,
+      "grad_norm": 0.1740686595439911,
+      "learning_rate": 0.0001551302931596091,
+      "loss": 0.9292,
+      "step": 730
+    },
+    {
+      "epoch": 1.6057111477210324,
+      "grad_norm": 0.10363954305648804,
+      "learning_rate": 0.000154885993485342,
+      "loss": 0.6316,
+      "step": 731
+    },
+    {
+      "epoch": 1.6079077429983526,
+      "grad_norm": 0.11538717150688171,
+      "learning_rate": 0.0001546416938110749,
+      "loss": 0.8754,
+      "step": 732
+    },
+    {
+      "epoch": 1.6101043382756726,
+      "grad_norm": 0.12158719450235367,
+      "learning_rate": 0.00015439739413680782,
+      "loss": 1.029,
+      "step": 733
+    },
+    {
+      "epoch": 1.612300933552993,
+      "grad_norm": 0.12808012962341309,
+      "learning_rate": 0.0001541530944625407,
+      "loss": 0.9371,
+      "step": 734
+    },
+    {
+      "epoch": 1.614497528830313,
+      "grad_norm": 0.15139949321746826,
+      "learning_rate": 0.0001539087947882736,
+      "loss": 0.9559,
+      "step": 735
+    },
+    {
+      "epoch": 1.6166941241076331,
+      "grad_norm": 0.13471727073192596,
+      "learning_rate": 0.0001536644951140065,
+      "loss": 0.9801,
+      "step": 736
+    },
+    {
+      "epoch": 1.6188907193849533,
+      "grad_norm": 0.10330980271100998,
+      "learning_rate": 0.0001534201954397394,
+      "loss": 0.9047,
+      "step": 737
+    },
+    {
+      "epoch": 1.6210873146622735,
+      "grad_norm": 0.15653370320796967,
+      "learning_rate": 0.0001531758957654723,
+      "loss": 0.9197,
+      "step": 738
+    },
+    {
+      "epoch": 1.6232839099395937,
+      "grad_norm": 0.10454390943050385,
+      "learning_rate": 0.0001529315960912052,
+      "loss": 0.8084,
+      "step": 739
+    },
+    {
+      "epoch": 1.6254805052169137,
+      "grad_norm": 0.11925213783979416,
+      "learning_rate": 0.0001526872964169381,
+      "loss": 0.9183,
+      "step": 740
+    },
+    {
+      "epoch": 1.627677100494234,
+      "grad_norm": 0.0876498818397522,
+      "learning_rate": 0.00015244299674267099,
+      "loss": 0.9165,
+      "step": 741
+    },
+    {
+      "epoch": 1.629873695771554,
+      "grad_norm": 0.12860970199108124,
+      "learning_rate": 0.0001521986970684039,
+      "loss": 0.8579,
+      "step": 742
+    },
+    {
+      "epoch": 1.6320702910488742,
+      "grad_norm": 1.338373064994812,
+      "learning_rate": 0.0001519543973941368,
+      "loss": 0.8082,
+      "step": 743
+    },
+    {
+      "epoch": 1.6342668863261944,
+      "grad_norm": 0.1339615136384964,
+      "learning_rate": 0.00015171009771986968,
+      "loss": 0.7575,
+      "step": 744
+    },
+    {
+      "epoch": 1.6364634816035144,
+      "grad_norm": 0.14447470009326935,
+      "learning_rate": 0.0001514657980456026,
+      "loss": 0.7587,
+      "step": 745
+    },
+    {
+      "epoch": 1.6386600768808348,
+      "grad_norm": 0.21855181455612183,
+      "learning_rate": 0.0001512214983713355,
+      "loss": 0.6512,
+      "step": 746
+    },
+    {
+      "epoch": 1.6408566721581548,
+      "grad_norm": 0.14507411420345306,
+      "learning_rate": 0.0001509771986970684,
+      "loss": 0.8963,
+      "step": 747
+    },
+    {
+      "epoch": 1.643053267435475,
+      "grad_norm": 0.10199563205242157,
+      "learning_rate": 0.0001507328990228013,
+      "loss": 0.7399,
+      "step": 748
+    },
+    {
+      "epoch": 1.6452498627127952,
+      "grad_norm": 0.11459880322217941,
+      "learning_rate": 0.00015048859934853418,
+      "loss": 0.7371,
+      "step": 749
+    },
+    {
+      "epoch": 1.6474464579901154,
+      "grad_norm": 0.11854815483093262,
+      "learning_rate": 0.0001502442996742671,
+      "loss": 0.7704,
+      "step": 750
+    },
+    {
+      "epoch": 1.6496430532674355,
+      "grad_norm": 0.1053410992026329,
+      "learning_rate": 0.00015,
+      "loss": 0.8886,
+      "step": 751
+    },
+    {
+      "epoch": 1.6518396485447555,
+      "grad_norm": 0.13653919100761414,
+      "learning_rate": 0.00014975570032573288,
+      "loss": 0.9625,
+      "step": 752
+    },
+    {
+      "epoch": 1.654036243822076,
+      "grad_norm": 0.18638905882835388,
+      "learning_rate": 0.0001495114006514658,
+      "loss": 1.1029,
+      "step": 753
+    },
+    {
+      "epoch": 1.656232839099396,
+      "grad_norm": 0.11540637165307999,
+      "learning_rate": 0.00014926710097719868,
+      "loss": 0.707,
+      "step": 754
+    },
+    {
+      "epoch": 1.658429434376716,
+      "grad_norm": 0.10499729961156845,
+      "learning_rate": 0.00014902280130293157,
+      "loss": 0.806,
+      "step": 755
+    },
+    {
+      "epoch": 1.6606260296540363,
+      "grad_norm": 0.09875135868787766,
+      "learning_rate": 0.0001487785016286645,
+      "loss": 0.8331,
+      "step": 756
+    },
+    {
+      "epoch": 1.6628226249313562,
+      "grad_norm": 0.10569437593221664,
+      "learning_rate": 0.00014853420195439738,
+      "loss": 0.9295,
+      "step": 757
+    },
+    {
+      "epoch": 1.6650192202086767,
+      "grad_norm": 0.14653964340686798,
+      "learning_rate": 0.0001482899022801303,
+      "loss": 0.9011,
+      "step": 758
+    },
+    {
+      "epoch": 1.6672158154859966,
+      "grad_norm": 0.12184444814920425,
+      "learning_rate": 0.00014804560260586318,
+      "loss": 0.9801,
+      "step": 759
+    },
+    {
+      "epoch": 1.6694124107633168,
+      "grad_norm": 0.20912539958953857,
+      "learning_rate": 0.00014780130293159607,
+      "loss": 0.9838,
+      "step": 760
+    },
+    {
+      "epoch": 1.671609006040637,
+      "grad_norm": 0.09944535791873932,
+      "learning_rate": 0.000147557003257329,
+      "loss": 0.8665,
+      "step": 761
+    },
+    {
+      "epoch": 1.6738056013179572,
+      "grad_norm": 0.1299128234386444,
+      "learning_rate": 0.00014731270358306188,
+      "loss": 0.9021,
+      "step": 762
+    },
+    {
+      "epoch": 1.6760021965952774,
+      "grad_norm": 0.1312880963087082,
+      "learning_rate": 0.00014706840390879477,
+      "loss": 0.8935,
+      "step": 763
+    },
+    {
+      "epoch": 1.6781987918725974,
+      "grad_norm": 0.10421334952116013,
+      "learning_rate": 0.00014682410423452768,
+      "loss": 0.9539,
+      "step": 764
+    },
+    {
+      "epoch": 1.6803953871499178,
+      "grad_norm": 0.1472439467906952,
+      "learning_rate": 0.00014657980456026057,
+      "loss": 0.96,
+      "step": 765
+    },
+    {
+      "epoch": 1.6825919824272377,
+      "grad_norm": 0.08844450116157532,
+      "learning_rate": 0.00014633550488599346,
+      "loss": 0.7252,
+      "step": 766
+    },
+    {
+      "epoch": 1.684788577704558,
+      "grad_norm": 0.1259230375289917,
+      "learning_rate": 0.00014609120521172638,
+      "loss": 0.8747,
+      "step": 767
+    },
+    {
+      "epoch": 1.6869851729818781,
+      "grad_norm": 0.12990254163742065,
+      "learning_rate": 0.00014584690553745927,
+      "loss": 0.843,
+      "step": 768
+    },
+    {
+      "epoch": 1.689181768259198,
+      "grad_norm": 0.19402039051055908,
+      "learning_rate": 0.00014560260586319216,
+      "loss": 0.8468,
+      "step": 769
+    },
+    {
+      "epoch": 1.6913783635365185,
+      "grad_norm": 0.11491717398166656,
+      "learning_rate": 0.00014535830618892507,
+      "loss": 0.9671,
+      "step": 770
+    },
+    {
+      "epoch": 1.6935749588138385,
+      "grad_norm": 0.12389478087425232,
+      "learning_rate": 0.00014511400651465796,
+      "loss": 1.0344,
+      "step": 771
+    },
+    {
+      "epoch": 1.6957715540911587,
+      "grad_norm": 0.6181966662406921,
+      "learning_rate": 0.00014486970684039088,
+      "loss": 0.9723,
+      "step": 772
+    },
+    {
+      "epoch": 1.6979681493684788,
+      "grad_norm": 0.08625784516334534,
+      "learning_rate": 0.00014462540716612377,
+      "loss": 0.77,
+      "step": 773
+    },
+    {
+      "epoch": 1.700164744645799,
+      "grad_norm": 0.10580785572528839,
+      "learning_rate": 0.00014438110749185666,
+      "loss": 0.7802,
+      "step": 774
+    },
+    {
+      "epoch": 1.7023613399231192,
+      "grad_norm": 0.13621123135089874,
+      "learning_rate": 0.00014413680781758957,
+      "loss": 0.911,
+      "step": 775
+    },
+    {
+      "epoch": 1.7045579352004392,
+      "grad_norm": 0.11263362318277359,
+      "learning_rate": 0.00014389250814332246,
+      "loss": 0.7195,
+      "step": 776
+    },
+    {
+      "epoch": 1.7067545304777596,
+      "grad_norm": 0.10706797987222672,
+      "learning_rate": 0.00014364820846905535,
+      "loss": 0.9413,
+      "step": 777
+    },
+    {
+      "epoch": 1.7089511257550796,
+      "grad_norm": 0.11347454786300659,
+      "learning_rate": 0.00014340390879478827,
+      "loss": 0.9875,
+      "step": 778
+    },
+    {
+      "epoch": 1.7111477210323998,
+      "grad_norm": 0.15007586777210236,
+      "learning_rate": 0.00014315960912052116,
+      "loss": 0.8893,
+      "step": 779
+    },
+    {
+      "epoch": 1.71334431630972,
+      "grad_norm": 0.11199898272752762,
+      "learning_rate": 0.00014291530944625404,
+      "loss": 0.9329,
+      "step": 780
+    },
+    {
+      "epoch": 1.7155409115870401,
+      "grad_norm": 0.12283331900835037,
+      "learning_rate": 0.00014267100977198696,
+      "loss": 0.8806,
+      "step": 781
+    },
+    {
+      "epoch": 1.7177375068643603,
+      "grad_norm": 0.1893986016511917,
+      "learning_rate": 0.00014242671009771985,
+      "loss": 0.8324,
+      "step": 782
+    },
+    {
+      "epoch": 1.7199341021416803,
+      "grad_norm": 0.11999013274908066,
+      "learning_rate": 0.00014218241042345274,
+      "loss": 0.9289,
+      "step": 783
+    },
+    {
+      "epoch": 1.7221306974190007,
+      "grad_norm": 0.1124853864312172,
+      "learning_rate": 0.00014193811074918566,
+      "loss": 0.8432,
+      "step": 784
+    },
+    {
+      "epoch": 1.7243272926963207,
+      "grad_norm": 0.10903999954462051,
+      "learning_rate": 0.00014169381107491854,
+      "loss": 1.1278,
+      "step": 785
+    },
+    {
+      "epoch": 1.7265238879736409,
+      "grad_norm": 0.0857887715101242,
+      "learning_rate": 0.00014144951140065146,
+      "loss": 0.7642,
+      "step": 786
+    },
+    {
+      "epoch": 1.728720483250961,
+      "grad_norm": 0.16403093934059143,
+      "learning_rate": 0.00014120521172638435,
+      "loss": 0.7997,
+      "step": 787
+    },
+    {
+      "epoch": 1.730917078528281,
+      "grad_norm": 0.09752880781888962,
+      "learning_rate": 0.00014096091205211724,
+      "loss": 0.8113,
+      "step": 788
+    },
+    {
+      "epoch": 1.7331136738056014,
+      "grad_norm": 0.11372333765029907,
+      "learning_rate": 0.00014071661237785016,
+      "loss": 0.9008,
+      "step": 789
+    },
+    {
+      "epoch": 1.7353102690829214,
+      "grad_norm": 0.10009942203760147,
+      "learning_rate": 0.00014047231270358305,
+      "loss": 1.0619,
+      "step": 790
+    },
+    {
+      "epoch": 1.7375068643602416,
+      "grad_norm": 0.18353626132011414,
+      "learning_rate": 0.00014022801302931593,
+      "loss": 0.873,
+      "step": 791
+    },
+    {
+      "epoch": 1.7397034596375618,
+      "grad_norm": 0.1049690917134285,
+      "learning_rate": 0.00013998371335504885,
+      "loss": 0.8022,
+      "step": 792
+    },
+    {
+      "epoch": 1.741900054914882,
+      "grad_norm": 0.15703780949115753,
+      "learning_rate": 0.00013973941368078174,
+      "loss": 0.8547,
+      "step": 793
+    },
+    {
+      "epoch": 1.7440966501922022,
+      "grad_norm": 0.11201336979866028,
+      "learning_rate": 0.00013949511400651463,
+      "loss": 0.9946,
+      "step": 794
+    },
+    {
+      "epoch": 1.7462932454695221,
+      "grad_norm": 0.08562810719013214,
+      "learning_rate": 0.00013925081433224755,
+      "loss": 0.8022,
+      "step": 795
+    },
+    {
+      "epoch": 1.7484898407468425,
+      "grad_norm": 0.12546247243881226,
+      "learning_rate": 0.00013900651465798043,
+      "loss": 0.7971,
+      "step": 796
+    },
+    {
+      "epoch": 1.7506864360241625,
+      "grad_norm": 0.11172143369913101,
+      "learning_rate": 0.00013876221498371335,
+      "loss": 1.0148,
+      "step": 797
+    },
+    {
+      "epoch": 1.7528830313014827,
+      "grad_norm": 0.09192407876253128,
+      "learning_rate": 0.00013851791530944624,
+      "loss": 0.8432,
+      "step": 798
+    },
+    {
+      "epoch": 1.755079626578803,
+      "grad_norm": 0.14140523970127106,
+      "learning_rate": 0.00013827361563517913,
+      "loss": 1.0197,
+      "step": 799
+    },
+    {
+      "epoch": 1.7572762218561229,
+      "grad_norm": 0.12856413424015045,
+      "learning_rate": 0.00013802931596091205,
+      "loss": 1.0823,
+      "step": 800
+    },
+    {
+      "epoch": 1.7594728171334433,
+      "grad_norm": 0.11290059238672256,
+      "learning_rate": 0.00013778501628664493,
+      "loss": 0.9726,
+      "step": 801
+    },
+    {
+      "epoch": 1.7616694124107632,
+      "grad_norm": 0.12149149924516678,
+      "learning_rate": 0.00013754071661237782,
+      "loss": 1.0984,
+      "step": 802
+    },
+    {
+      "epoch": 1.7638660076880834,
+      "grad_norm": 0.19626955687999725,
+      "learning_rate": 0.00013729641693811074,
+      "loss": 1.3272,
+      "step": 803
+    },
+    {
+      "epoch": 1.7660626029654036,
+      "grad_norm": 0.17344090342521667,
+      "learning_rate": 0.00013705211726384363,
+      "loss": 1.1145,
+      "step": 804
+    },
+    {
+      "epoch": 1.7682591982427238,
+      "grad_norm": 0.10894408822059631,
+      "learning_rate": 0.00013680781758957652,
+      "loss": 0.8778,
+      "step": 805
+    },
+    {
+      "epoch": 1.770455793520044,
+      "grad_norm": 0.18599945306777954,
+      "learning_rate": 0.00013656351791530943,
+      "loss": 0.8431,
+      "step": 806
+    },
+    {
+      "epoch": 1.772652388797364,
+      "grad_norm": 0.12184733897447586,
+      "learning_rate": 0.00013631921824104232,
+      "loss": 0.931,
+      "step": 807
+    },
+    {
+      "epoch": 1.7748489840746844,
+      "grad_norm": 0.2208235114812851,
+      "learning_rate": 0.0001360749185667752,
+      "loss": 0.984,
+      "step": 808
+    },
+    {
+      "epoch": 1.7770455793520044,
+      "grad_norm": 0.1123606339097023,
+      "learning_rate": 0.00013583061889250813,
+      "loss": 0.9812,
+      "step": 809
+    },
+    {
+      "epoch": 1.7792421746293245,
+      "grad_norm": 0.10484428703784943,
+      "learning_rate": 0.00013558631921824102,
+      "loss": 0.6917,
+      "step": 810
+    },
+    {
+      "epoch": 1.7814387699066447,
+      "grad_norm": 0.12408167868852615,
+      "learning_rate": 0.00013534201954397394,
+      "loss": 0.788,
+      "step": 811
+    },
+    {
+      "epoch": 1.7836353651839647,
+      "grad_norm": 0.8263797760009766,
+      "learning_rate": 0.00013509771986970682,
+      "loss": 0.8711,
+      "step": 812
+    },
+    {
+      "epoch": 1.7858319604612851,
+      "grad_norm": 0.1291903704404831,
+      "learning_rate": 0.00013485342019543971,
+      "loss": 1.0006,
+      "step": 813
+    },
+    {
+      "epoch": 1.788028555738605,
+      "grad_norm": 0.33283305168151855,
+      "learning_rate": 0.00013460912052117263,
+      "loss": 0.9395,
+      "step": 814
+    },
+    {
+      "epoch": 1.7902251510159253,
+      "grad_norm": 0.13870319724082947,
+      "learning_rate": 0.00013436482084690552,
+      "loss": 0.9515,
+      "step": 815
+    },
+    {
+      "epoch": 1.7924217462932455,
+      "grad_norm": 0.10988850891590118,
+      "learning_rate": 0.0001341205211726384,
+      "loss": 0.8901,
+      "step": 816
+    },
+    {
+      "epoch": 1.7946183415705657,
+      "grad_norm": 0.1376354992389679,
+      "learning_rate": 0.00013387622149837132,
+      "loss": 0.8075,
+      "step": 817
+    },
+    {
+      "epoch": 1.7968149368478858,
+      "grad_norm": 0.11696015298366547,
+      "learning_rate": 0.00013363192182410421,
+      "loss": 0.9678,
+      "step": 818
+    },
+    {
+      "epoch": 1.7990115321252058,
+      "grad_norm": 0.28118568658828735,
+      "learning_rate": 0.0001333876221498371,
+      "loss": 0.8912,
+      "step": 819
+    },
+    {
+      "epoch": 1.8012081274025262,
+      "grad_norm": 0.11387281864881516,
+      "learning_rate": 0.00013314332247557002,
+      "loss": 0.9296,
+      "step": 820
+    },
+    {
+      "epoch": 1.8034047226798462,
+      "grad_norm": 0.10126536339521408,
+      "learning_rate": 0.0001328990228013029,
+      "loss": 0.9168,
+      "step": 821
+    },
+    {
+      "epoch": 1.8056013179571664,
+      "grad_norm": 0.14740599691867828,
+      "learning_rate": 0.00013265472312703582,
+      "loss": 1.0371,
+      "step": 822
+    },
+    {
+      "epoch": 1.8077979132344866,
+      "grad_norm": 0.11603671312332153,
+      "learning_rate": 0.00013241042345276871,
+      "loss": 0.9052,
+      "step": 823
+    },
+    {
+      "epoch": 1.8099945085118065,
+      "grad_norm": 0.09960336238145828,
+      "learning_rate": 0.0001321661237785016,
+      "loss": 0.8304,
+      "step": 824
+    },
+    {
+      "epoch": 1.812191103789127,
+      "grad_norm": 0.10368466377258301,
+      "learning_rate": 0.00013192182410423452,
+      "loss": 0.7777,
+      "step": 825
+    },
+    {
+      "epoch": 1.814387699066447,
+      "grad_norm": 0.11901511996984482,
+      "learning_rate": 0.0001316775244299674,
+      "loss": 0.7101,
+      "step": 826
+    },
+    {
+      "epoch": 1.8165842943437671,
+      "grad_norm": 0.12246643006801605,
+      "learning_rate": 0.0001314332247557003,
+      "loss": 0.9504,
+      "step": 827
+    },
+    {
+      "epoch": 1.8187808896210873,
+      "grad_norm": 0.13885262608528137,
+      "learning_rate": 0.00013118892508143321,
+      "loss": 0.8097,
+      "step": 828
+    },
+    {
+      "epoch": 1.8209774848984075,
+      "grad_norm": 0.12531216442584991,
+      "learning_rate": 0.00013094462540716613,
+      "loss": 0.7951,
+      "step": 829
+    },
+    {
+      "epoch": 1.8231740801757277,
+      "grad_norm": 0.12174840271472931,
+      "learning_rate": 0.00013070032573289902,
+      "loss": 0.7011,
+      "step": 830
+    },
+    {
+      "epoch": 1.8253706754530477,
+      "grad_norm": 0.1265886276960373,
+      "learning_rate": 0.0001304560260586319,
+      "loss": 0.6207,
+      "step": 831
+    },
+    {
+      "epoch": 1.827567270730368,
+      "grad_norm": 0.1295892745256424,
+      "learning_rate": 0.00013021172638436483,
+      "loss": 0.9675,
+      "step": 832
+    },
+    {
+      "epoch": 1.829763866007688,
+      "grad_norm": 0.11082609742879868,
+      "learning_rate": 0.00012996742671009771,
+      "loss": 0.9515,
+      "step": 833
+    },
+    {
+      "epoch": 1.8319604612850082,
+      "grad_norm": 0.10516710579395294,
+      "learning_rate": 0.0001297231270358306,
+      "loss": 1.052,
+      "step": 834
+    },
+    {
+      "epoch": 1.8341570565623284,
+      "grad_norm": 0.11737772077322006,
+      "learning_rate": 0.00012947882736156352,
+      "loss": 0.932,
+      "step": 835
+    },
+    {
+      "epoch": 1.8363536518396484,
+      "grad_norm": 0.11845795065164566,
+      "learning_rate": 0.0001292345276872964,
+      "loss": 0.7439,
+      "step": 836
+    },
+    {
+      "epoch": 1.8385502471169688,
+      "grad_norm": 0.09692834317684174,
+      "learning_rate": 0.00012899022801302933,
+      "loss": 0.8162,
+      "step": 837
+    },
+    {
+      "epoch": 1.8407468423942888,
+      "grad_norm": 0.13455092906951904,
+      "learning_rate": 0.00012874592833876221,
+      "loss": 0.9222,
+      "step": 838
+    },
+    {
+      "epoch": 1.842943437671609,
+      "grad_norm": 0.15182065963745117,
+      "learning_rate": 0.0001285016286644951,
+      "loss": 0.83,
+      "step": 839
+    },
+    {
+      "epoch": 1.8451400329489291,
+      "grad_norm": 0.14715076982975006,
+      "learning_rate": 0.00012825732899022802,
+      "loss": 0.9608,
+      "step": 840
+    },
+    {
+      "epoch": 1.8473366282262493,
+      "grad_norm": 0.0961465984582901,
+      "learning_rate": 0.0001280130293159609,
+      "loss": 0.8116,
+      "step": 841
+    },
+    {
+      "epoch": 1.8495332235035695,
+      "grad_norm": 0.12844809889793396,
+      "learning_rate": 0.0001277687296416938,
+      "loss": 0.9128,
+      "step": 842
+    },
+    {
+      "epoch": 1.8517298187808895,
+      "grad_norm": 0.10709239542484283,
+      "learning_rate": 0.00012752442996742671,
+      "loss": 1.073,
+      "step": 843
+    },
+    {
+      "epoch": 1.85392641405821,
+      "grad_norm": 0.13335658609867096,
+      "learning_rate": 0.0001272801302931596,
+      "loss": 0.9715,
+      "step": 844
+    },
+    {
+      "epoch": 1.8561230093355299,
+      "grad_norm": 0.11525517702102661,
+      "learning_rate": 0.0001270358306188925,
+      "loss": 0.8337,
+      "step": 845
+    },
+    {
+      "epoch": 1.85831960461285,
+      "grad_norm": 0.12004940956830978,
+      "learning_rate": 0.0001267915309446254,
+      "loss": 0.8188,
+      "step": 846
+    },
+    {
+      "epoch": 1.8605161998901703,
+      "grad_norm": 0.11972694844007492,
+      "learning_rate": 0.0001265472312703583,
+      "loss": 1.0254,
+      "step": 847
+    },
+    {
+      "epoch": 1.8627127951674904,
+      "grad_norm": 0.09509237110614777,
+      "learning_rate": 0.0001263029315960912,
+      "loss": 0.7737,
+      "step": 848
+    },
+    {
+      "epoch": 1.8649093904448106,
+      "grad_norm": 0.10498124361038208,
+      "learning_rate": 0.0001260586319218241,
+      "loss": 0.7782,
+      "step": 849
+    },
+    {
+      "epoch": 1.8671059857221306,
+      "grad_norm": 0.09272133558988571,
+      "learning_rate": 0.000125814332247557,
+      "loss": 0.7865,
+      "step": 850
+    },
+    {
+      "epoch": 1.869302580999451,
+      "grad_norm": 0.10824044793844223,
+      "learning_rate": 0.0001255700325732899,
+      "loss": 0.7151,
+      "step": 851
+    },
+    {
+      "epoch": 1.871499176276771,
+      "grad_norm": 0.12407927215099335,
+      "learning_rate": 0.0001253257328990228,
+      "loss": 0.9417,
+      "step": 852
+    },
+    {
+      "epoch": 1.8736957715540912,
+      "grad_norm": 0.11268027871847153,
+      "learning_rate": 0.0001250814332247557,
+      "loss": 1.0951,
+      "step": 853
+    },
+    {
+      "epoch": 1.8758923668314114,
+      "grad_norm": 0.16494135558605194,
+      "learning_rate": 0.0001248371335504886,
+      "loss": 1.0313,
+      "step": 854
+    },
+    {
+      "epoch": 1.8780889621087313,
+      "grad_norm": 0.09016561508178711,
+      "learning_rate": 0.0001245928338762215,
+      "loss": 0.7813,
+      "step": 855
+    },
+    {
+      "epoch": 1.8802855573860517,
+      "grad_norm": 0.15998411178588867,
+      "learning_rate": 0.00012434853420195438,
+      "loss": 1.0554,
+      "step": 856
+    },
+    {
+      "epoch": 1.8824821526633717,
+      "grad_norm": 0.12199036031961441,
+      "learning_rate": 0.0001241042345276873,
+      "loss": 0.8729,
+      "step": 857
+    },
+    {
+      "epoch": 1.884678747940692,
+      "grad_norm": 0.09834381937980652,
+      "learning_rate": 0.0001238599348534202,
+      "loss": 0.9086,
+      "step": 858
+    },
+    {
+      "epoch": 1.886875343218012,
+      "grad_norm": 0.13520832359790802,
+      "learning_rate": 0.00012361563517915308,
+      "loss": 0.7906,
+      "step": 859
+    },
+    {
+      "epoch": 1.8890719384953323,
+      "grad_norm": 0.09126309305429459,
+      "learning_rate": 0.000123371335504886,
+      "loss": 0.7824,
+      "step": 860
+    },
+    {
+      "epoch": 1.8912685337726525,
+      "grad_norm": 0.11749809235334396,
+      "learning_rate": 0.00012312703583061888,
+      "loss": 1.0263,
+      "step": 861
+    },
+    {
+      "epoch": 1.8934651290499724,
+      "grad_norm": 0.11738768219947815,
+      "learning_rate": 0.00012288273615635177,
+      "loss": 0.9186,
+      "step": 862
+    },
+    {
+      "epoch": 1.8956617243272929,
+      "grad_norm": 0.13642632961273193,
+      "learning_rate": 0.0001226384364820847,
+      "loss": 1.1592,
+      "step": 863
+    },
+    {
+      "epoch": 1.8978583196046128,
+      "grad_norm": 0.10935261100530624,
+      "learning_rate": 0.00012239413680781758,
+      "loss": 0.7785,
+      "step": 864
+    },
+    {
+      "epoch": 1.900054914881933,
+      "grad_norm": 0.12026354670524597,
+      "learning_rate": 0.0001221498371335505,
+      "loss": 0.9485,
+      "step": 865
+    },
+    {
+      "epoch": 1.9022515101592532,
+      "grad_norm": 0.09153635799884796,
+      "learning_rate": 0.00012190553745928338,
+      "loss": 0.7911,
+      "step": 866
+    },
+    {
+      "epoch": 1.9044481054365732,
+      "grad_norm": 0.11394686996936798,
+      "learning_rate": 0.00012166123778501629,
+      "loss": 0.8442,
+      "step": 867
+    },
+    {
+      "epoch": 1.9066447007138936,
+      "grad_norm": 0.14081139862537384,
+      "learning_rate": 0.00012141693811074918,
+      "loss": 1.0958,
+      "step": 868
+    },
+    {
+      "epoch": 1.9088412959912135,
+      "grad_norm": 0.14040987193584442,
+      "learning_rate": 0.00012117263843648208,
+      "loss": 1.0783,
+      "step": 869
+    },
+    {
+      "epoch": 1.9110378912685337,
+      "grad_norm": 0.10303716361522675,
+      "learning_rate": 0.00012092833876221498,
+      "loss": 0.8313,
+      "step": 870
+    },
+    {
+      "epoch": 1.913234486545854,
+      "grad_norm": 0.16119170188903809,
+      "learning_rate": 0.00012068403908794787,
+      "loss": 0.862,
+      "step": 871
+    },
+    {
+      "epoch": 1.9154310818231741,
+      "grad_norm": 0.09430427849292755,
+      "learning_rate": 0.00012043973941368077,
+      "loss": 0.9486,
+      "step": 872
+    },
+    {
+      "epoch": 1.9176276771004943,
+      "grad_norm": 0.13481692969799042,
+      "learning_rate": 0.00012019543973941368,
+      "loss": 0.9941,
+      "step": 873
+    },
+    {
+      "epoch": 1.9198242723778143,
+      "grad_norm": 0.11439478397369385,
+      "learning_rate": 0.00011995114006514658,
+      "loss": 1.1153,
+      "step": 874
+    },
+    {
+      "epoch": 1.9220208676551347,
+      "grad_norm": 0.11953114718198776,
+      "learning_rate": 0.00011970684039087947,
+      "loss": 1.0579,
+      "step": 875
+    },
+    {
+      "epoch": 1.9242174629324547,
+      "grad_norm": 0.14599184691905975,
+      "learning_rate": 0.00011946254071661237,
+      "loss": 1.0078,
+      "step": 876
+    },
+    {
+      "epoch": 1.9264140582097748,
+      "grad_norm": 0.11028395593166351,
+      "learning_rate": 0.00011921824104234527,
+      "loss": 0.7588,
+      "step": 877
+    },
+    {
+      "epoch": 1.928610653487095,
+      "grad_norm": 0.13960738480091095,
+      "learning_rate": 0.00011897394136807816,
+      "loss": 0.9111,
+      "step": 878
+    },
+    {
+      "epoch": 1.930807248764415,
+      "grad_norm": 0.09937304258346558,
+      "learning_rate": 0.00011872964169381106,
+      "loss": 0.8471,
+      "step": 879
+    },
+    {
+      "epoch": 1.9330038440417354,
+      "grad_norm": 0.11825273931026459,
+      "learning_rate": 0.00011848534201954397,
+      "loss": 0.8056,
+      "step": 880
+    },
+    {
+      "epoch": 1.9352004393190554,
+      "grad_norm": 0.16002795100212097,
+      "learning_rate": 0.00011824104234527687,
+      "loss": 0.9288,
+      "step": 881
+    },
+    {
+      "epoch": 1.9373970345963756,
+      "grad_norm": 0.09280494600534439,
+      "learning_rate": 0.00011799674267100976,
+      "loss": 0.6833,
+      "step": 882
+    },
+    {
+      "epoch": 1.9395936298736958,
+      "grad_norm": 0.12539535760879517,
+      "learning_rate": 0.00011775244299674266,
+      "loss": 0.9863,
+      "step": 883
+    },
+    {
+      "epoch": 1.941790225151016,
+      "grad_norm": 0.14245621860027313,
+      "learning_rate": 0.00011750814332247557,
+      "loss": 0.9779,
+      "step": 884
+    },
+    {
+      "epoch": 1.9439868204283361,
+      "grad_norm": 0.11080895364284515,
+      "learning_rate": 0.00011726384364820845,
+      "loss": 1.3187,
+      "step": 885
+    },
+    {
+      "epoch": 1.9461834157056561,
+      "grad_norm": 0.13062021136283875,
+      "learning_rate": 0.00011701954397394136,
+      "loss": 0.9064,
+      "step": 886
+    },
+    {
+      "epoch": 1.9483800109829765,
+      "grad_norm": 0.09109672904014587,
+      "learning_rate": 0.00011677524429967426,
+      "loss": 0.7718,
+      "step": 887
+    },
+    {
+      "epoch": 1.9505766062602965,
+      "grad_norm": 0.1112729161977768,
+      "learning_rate": 0.00011653094462540716,
+      "loss": 0.7683,
+      "step": 888
+    },
+    {
+      "epoch": 1.9527732015376167,
+      "grad_norm": 0.20752094686031342,
+      "learning_rate": 0.00011628664495114005,
+      "loss": 0.8981,
+      "step": 889
+    },
+    {
+      "epoch": 1.9549697968149369,
+      "grad_norm": 0.09921902418136597,
+      "learning_rate": 0.00011604234527687295,
+      "loss": 0.8923,
+      "step": 890
+    },
+    {
+      "epoch": 1.9571663920922568,
+      "grad_norm": 0.10895263403654099,
+      "learning_rate": 0.00011579804560260586,
+      "loss": 1.0144,
+      "step": 891
+    },
+    {
+      "epoch": 1.9593629873695773,
+      "grad_norm": 0.10391975194215775,
+      "learning_rate": 0.00011555374592833876,
+      "loss": 0.7779,
+      "step": 892
+    },
+    {
+      "epoch": 1.9615595826468972,
+      "grad_norm": 0.11912518739700317,
+      "learning_rate": 0.00011530944625407165,
+      "loss": 0.9034,
+      "step": 893
+    },
+    {
+      "epoch": 1.9637561779242174,
+      "grad_norm": 0.16799083352088928,
+      "learning_rate": 0.00011506514657980455,
+      "loss": 0.9564,
+      "step": 894
+    },
+    {
+      "epoch": 1.9659527732015376,
+      "grad_norm": 0.12434320151805878,
+      "learning_rate": 0.00011482084690553745,
+      "loss": 0.7682,
+      "step": 895
+    },
+    {
+      "epoch": 1.9681493684788578,
+      "grad_norm": 0.10150429606437683,
+      "learning_rate": 0.00011457654723127034,
+      "loss": 1.0215,
+      "step": 896
+    },
+    {
+      "epoch": 1.970345963756178,
+      "grad_norm": 0.09960570186376572,
+      "learning_rate": 0.00011433224755700325,
+      "loss": 0.7719,
+      "step": 897
+    },
+    {
+      "epoch": 1.972542559033498,
+      "grad_norm": 0.1367155760526657,
+      "learning_rate": 0.00011408794788273615,
+      "loss": 1.0384,
+      "step": 898
+    },
+    {
+      "epoch": 1.9747391543108184,
+      "grad_norm": 0.1492743194103241,
+      "learning_rate": 0.00011384364820846905,
+      "loss": 1.0917,
+      "step": 899
+    },
+    {
+      "epoch": 1.9769357495881383,
+      "grad_norm": 0.10855671018362045,
+      "learning_rate": 0.00011359934853420194,
+      "loss": 0.8038,
+      "step": 900
+    },
+    {
+      "epoch": 1.9791323448654585,
+      "grad_norm": 0.11136850714683533,
+      "learning_rate": 0.00011335504885993484,
+      "loss": 1.0149,
+      "step": 901
+    },
+    {
+      "epoch": 1.9813289401427787,
+      "grad_norm": 0.12785561382770538,
+      "learning_rate": 0.00011311074918566775,
+      "loss": 0.9795,
+      "step": 902
+    },
+    {
+      "epoch": 1.9835255354200987,
+      "grad_norm": 0.18793602287769318,
+      "learning_rate": 0.00011286644951140064,
+      "loss": 0.8667,
+      "step": 903
+    },
+    {
+      "epoch": 1.985722130697419,
+      "grad_norm": 0.12246491760015488,
+      "learning_rate": 0.00011262214983713354,
+      "loss": 0.7488,
+      "step": 904
+    },
+    {
+      "epoch": 1.987918725974739,
+      "grad_norm": 0.13461388647556305,
+      "learning_rate": 0.00011237785016286644,
+      "loss": 0.7434,
+      "step": 905
+    },
+    {
+      "epoch": 1.9901153212520593,
+      "grad_norm": 0.09910102933645248,
+      "learning_rate": 0.00011213355048859934,
+      "loss": 0.9827,
+      "step": 906
+    },
+    {
+      "epoch": 1.9923119165293794,
+      "grad_norm": 0.11667338013648987,
+      "learning_rate": 0.00011188925081433223,
+      "loss": 0.7422,
+      "step": 907
+    },
+    {
+      "epoch": 1.9945085118066996,
+      "grad_norm": 0.09112700074911118,
+      "learning_rate": 0.00011164495114006514,
+      "loss": 0.7409,
+      "step": 908
+    },
+    {
+      "epoch": 1.9967051070840198,
+      "grad_norm": 0.11038555949926376,
+      "learning_rate": 0.00011140065146579804,
+      "loss": 0.6907,
+      "step": 909
+    },
+    {
+      "epoch": 1.9989017023613398,
+      "grad_norm": 0.1778726726770401,
+      "learning_rate": 0.00011115635179153093,
+      "loss": 0.9287,
+      "step": 910
+    },
+    {
+      "epoch": 2.00109829763866,
+      "grad_norm": 0.10834711790084839,
+      "learning_rate": 0.00011091205211726383,
+      "loss": 1.0208,
+      "step": 911
+    },
+    {
+      "epoch": 2.00329489291598,
+      "grad_norm": 0.1098528653383255,
+      "learning_rate": 0.00011066775244299673,
+      "loss": 0.8143,
+      "step": 912
+    },
+    {
+      "epoch": 2.0054914881933006,
+      "grad_norm": 0.10828452557325363,
+      "learning_rate": 0.00011042345276872964,
+      "loss": 0.6948,
+      "step": 913
+    },
+    {
+      "epoch": 2.0076880834706206,
+      "grad_norm": 0.11567585170269012,
+      "learning_rate": 0.00011017915309446253,
+      "loss": 0.8821,
+      "step": 914
+    },
+    {
+      "epoch": 2.0098846787479405,
+      "grad_norm": 0.1728159487247467,
+      "learning_rate": 0.00010993485342019543,
+      "loss": 0.8682,
+      "step": 915
+    },
+    {
+      "epoch": 2.012081274025261,
+      "grad_norm": 0.15759354829788208,
+      "learning_rate": 0.00010969055374592833,
+      "loss": 1.0556,
+      "step": 916
+    },
+    {
+      "epoch": 2.014277869302581,
+      "grad_norm": 0.12905576825141907,
+      "learning_rate": 0.00010944625407166122,
+      "loss": 1.0246,
+      "step": 917
+    },
+    {
+      "epoch": 2.0164744645799013,
+      "grad_norm": 0.1394015997648239,
+      "learning_rate": 0.00010920195439739412,
+      "loss": 1.0676,
+      "step": 918
+    },
+    {
+      "epoch": 2.0186710598572213,
+      "grad_norm": 0.10404639691114426,
+      "learning_rate": 0.00010895765472312703,
+      "loss": 0.8143,
+      "step": 919
+    },
+    {
+      "epoch": 2.0208676551345413,
+      "grad_norm": 0.1602998971939087,
+      "learning_rate": 0.00010871335504885993,
+      "loss": 0.8015,
+      "step": 920
+    },
+    {
+      "epoch": 2.0230642504118617,
+      "grad_norm": 0.11231439560651779,
+      "learning_rate": 0.00010846905537459282,
+      "loss": 0.8063,
+      "step": 921
+    },
+    {
+      "epoch": 2.0252608456891816,
+      "grad_norm": 0.10337056964635849,
+      "learning_rate": 0.00010822475570032572,
+      "loss": 0.907,
+      "step": 922
+    },
+    {
+      "epoch": 2.027457440966502,
+      "grad_norm": 0.13463544845581055,
+      "learning_rate": 0.00010798045602605862,
+      "loss": 0.9741,
+      "step": 923
+    },
+    {
+      "epoch": 2.029654036243822,
+      "grad_norm": 0.12715274095535278,
+      "learning_rate": 0.00010773615635179153,
+      "loss": 1.0225,
+      "step": 924
+    },
+    {
+      "epoch": 2.0318506315211424,
+      "grad_norm": 0.09976208209991455,
+      "learning_rate": 0.00010749185667752442,
+      "loss": 0.8043,
+      "step": 925
+    },
+    {
+      "epoch": 2.0340472267984624,
+      "grad_norm": 0.14953263103961945,
+      "learning_rate": 0.00010724755700325732,
+      "loss": 0.9199,
+      "step": 926
+    },
+    {
+      "epoch": 2.0362438220757824,
+      "grad_norm": 0.11714182049036026,
+      "learning_rate": 0.00010700325732899022,
+      "loss": 1.0389,
+      "step": 927
+    },
+    {
+      "epoch": 2.0384404173531028,
+      "grad_norm": 0.10790516436100006,
+      "learning_rate": 0.00010675895765472311,
+      "loss": 0.8493,
+      "step": 928
+    },
+    {
+      "epoch": 2.0406370126304227,
+      "grad_norm": 0.0973561555147171,
+      "learning_rate": 0.00010651465798045601,
+      "loss": 0.7817,
+      "step": 929
+    },
+    {
+      "epoch": 2.042833607907743,
+      "grad_norm": 0.09318231046199799,
+      "learning_rate": 0.00010627035830618892,
+      "loss": 0.7389,
+      "step": 930
+    },
+    {
+      "epoch": 2.045030203185063,
+      "grad_norm": 0.09634759277105331,
+      "learning_rate": 0.00010602605863192182,
+      "loss": 0.8012,
+      "step": 931
+    },
+    {
+      "epoch": 2.047226798462383,
+      "grad_norm": 0.1438138335943222,
+      "learning_rate": 0.00010578175895765471,
+      "loss": 0.9138,
+      "step": 932
+    },
+    {
+      "epoch": 2.0494233937397035,
+      "grad_norm": 0.13972003757953644,
+      "learning_rate": 0.00010553745928338761,
+      "loss": 0.8922,
+      "step": 933
+    },
+    {
+      "epoch": 2.0516199890170235,
+      "grad_norm": 0.13632741570472717,
+      "learning_rate": 0.00010529315960912051,
+      "loss": 0.8391,
+      "step": 934
+    },
+    {
+      "epoch": 2.053816584294344,
+      "grad_norm": 0.09593293070793152,
+      "learning_rate": 0.0001050488599348534,
+      "loss": 0.8419,
+      "step": 935
+    },
+    {
+      "epoch": 2.056013179571664,
+      "grad_norm": 0.10019386559724808,
+      "learning_rate": 0.0001048045602605863,
+      "loss": 0.9998,
+      "step": 936
+    },
+    {
+      "epoch": 2.0582097748489843,
+      "grad_norm": 0.11537342518568039,
+      "learning_rate": 0.00010456026058631921,
+      "loss": 0.9397,
+      "step": 937
+    },
+    {
+      "epoch": 2.0604063701263042,
+      "grad_norm": 0.16264887154102325,
+      "learning_rate": 0.00010431596091205211,
+      "loss": 0.8103,
+      "step": 938
+    },
+    {
+      "epoch": 2.062602965403624,
+      "grad_norm": 0.13394120335578918,
+      "learning_rate": 0.000104071661237785,
+      "loss": 1.0848,
+      "step": 939
+    },
+    {
+      "epoch": 2.0647995606809446,
+      "grad_norm": 0.10955598205327988,
+      "learning_rate": 0.0001038273615635179,
+      "loss": 0.8217,
+      "step": 940
+    },
+    {
+      "epoch": 2.0669961559582646,
+      "grad_norm": 0.33553674817085266,
+      "learning_rate": 0.0001035830618892508,
+      "loss": 0.9567,
+      "step": 941
+    },
+    {
+      "epoch": 2.069192751235585,
+      "grad_norm": 0.11299476772546768,
+      "learning_rate": 0.0001033387622149837,
+      "loss": 0.706,
+      "step": 942
+    },
+    {
+      "epoch": 2.071389346512905,
+      "grad_norm": 0.11936173588037491,
+      "learning_rate": 0.0001030944625407166,
+      "loss": 0.8733,
+      "step": 943
+    },
+    {
+      "epoch": 2.073585941790225,
+      "grad_norm": 0.13407137989997864,
+      "learning_rate": 0.0001028501628664495,
+      "loss": 0.8092,
+      "step": 944
+    },
+    {
+      "epoch": 2.0757825370675453,
+      "grad_norm": 0.13549263775348663,
+      "learning_rate": 0.0001026058631921824,
+      "loss": 0.9345,
+      "step": 945
+    },
+    {
+      "epoch": 2.0779791323448653,
+      "grad_norm": 0.10706178098917007,
+      "learning_rate": 0.00010236156351791529,
+      "loss": 0.8371,
+      "step": 946
+    },
+    {
+      "epoch": 2.0801757276221857,
+      "grad_norm": 0.13867148756980896,
+      "learning_rate": 0.0001021172638436482,
+      "loss": 0.6567,
+      "step": 947
+    },
+    {
+      "epoch": 2.0823723228995057,
+      "grad_norm": 0.10158029943704605,
+      "learning_rate": 0.0001018729641693811,
+      "loss": 0.7288,
+      "step": 948
+    },
+    {
+      "epoch": 2.084568918176826,
+      "grad_norm": 0.15534426271915436,
+      "learning_rate": 0.00010162866449511399,
+      "loss": 1.0471,
+      "step": 949
+    },
+    {
+      "epoch": 2.086765513454146,
+      "grad_norm": 0.08808978646993637,
+      "learning_rate": 0.00010138436482084689,
+      "loss": 0.8555,
+      "step": 950
+    },
+    {
+      "epoch": 2.088962108731466,
+      "grad_norm": 0.5641061067581177,
+      "learning_rate": 0.00010114006514657979,
+      "loss": 0.9366,
+      "step": 951
+    },
+    {
+      "epoch": 2.0911587040087865,
+      "grad_norm": 0.1367720067501068,
+      "learning_rate": 0.0001008957654723127,
+      "loss": 0.9281,
+      "step": 952
+    },
+    {
+      "epoch": 2.0933552992861064,
+      "grad_norm": 0.12688738107681274,
+      "learning_rate": 0.00010065146579804558,
+      "loss": 1.0111,
+      "step": 953
+    },
+    {
+      "epoch": 2.095551894563427,
+      "grad_norm": 0.09540694206953049,
+      "learning_rate": 0.00010040716612377849,
+      "loss": 0.771,
+      "step": 954
+    },
+    {
+      "epoch": 2.097748489840747,
+      "grad_norm": 0.09818293154239655,
+      "learning_rate": 0.00010016286644951139,
+      "loss": 0.9361,
+      "step": 955
+    },
+    {
+      "epoch": 2.099945085118067,
+      "grad_norm": 0.14012299478054047,
+      "learning_rate": 9.991856677524428e-05,
+      "loss": 0.8626,
+      "step": 956
+    },
+    {
+      "epoch": 2.102141680395387,
+      "grad_norm": 0.18497447669506073,
+      "learning_rate": 9.967426710097718e-05,
+      "loss": 0.9116,
+      "step": 957
+    },
+    {
+      "epoch": 2.104338275672707,
+      "grad_norm": 0.09814685583114624,
+      "learning_rate": 9.942996742671008e-05,
+      "loss": 0.798,
+      "step": 958
+    },
+    {
+      "epoch": 2.1065348709500276,
+      "grad_norm": 0.14369000494480133,
+      "learning_rate": 9.918566775244299e-05,
+      "loss": 0.8412,
+      "step": 959
+    },
+    {
+      "epoch": 2.1087314662273475,
+      "grad_norm": 0.08997687697410583,
+      "learning_rate": 9.894136807817588e-05,
+      "loss": 0.8715,
+      "step": 960
+    },
+    {
+      "epoch": 2.110928061504668,
+      "grad_norm": 0.13769656419754028,
+      "learning_rate": 9.869706840390878e-05,
+      "loss": 1.0398,
+      "step": 961
+    },
+    {
+      "epoch": 2.113124656781988,
+      "grad_norm": 0.14535696804523468,
+      "learning_rate": 9.845276872964168e-05,
+      "loss": 0.8822,
+      "step": 962
+    },
+    {
+      "epoch": 2.115321252059308,
+      "grad_norm": 0.13286636769771576,
+      "learning_rate": 9.820846905537458e-05,
+      "loss": 0.7634,
+      "step": 963
+    },
+    {
+      "epoch": 2.1175178473366283,
+      "grad_norm": 0.1073593944311142,
+      "learning_rate": 9.796416938110747e-05,
+      "loss": 0.8491,
+      "step": 964
+    },
+    {
+      "epoch": 2.1197144426139483,
+      "grad_norm": 0.08950421959161758,
+      "learning_rate": 9.771986970684038e-05,
+      "loss": 0.7973,
+      "step": 965
+    },
+    {
+      "epoch": 2.1219110378912687,
+      "grad_norm": 0.12546946108341217,
+      "learning_rate": 9.747557003257328e-05,
+      "loss": 0.7642,
+      "step": 966
+    },
+    {
+      "epoch": 2.1241076331685886,
+      "grad_norm": 0.6640294194221497,
+      "learning_rate": 9.723127035830617e-05,
+      "loss": 0.7884,
+      "step": 967
+    },
+    {
+      "epoch": 2.1263042284459086,
+      "grad_norm": 0.11225269734859467,
+      "learning_rate": 9.698697068403907e-05,
+      "loss": 0.8343,
+      "step": 968
+    },
+    {
+      "epoch": 2.128500823723229,
+      "grad_norm": 0.11752647906541824,
+      "learning_rate": 9.674267100977197e-05,
+      "loss": 0.7734,
+      "step": 969
+    },
+    {
+      "epoch": 2.130697419000549,
+      "grad_norm": 0.10536935925483704,
+      "learning_rate": 9.649837133550488e-05,
+      "loss": 0.9609,
+      "step": 970
+    },
+    {
+      "epoch": 2.1328940142778694,
+      "grad_norm": 0.19671271741390228,
+      "learning_rate": 9.625407166123777e-05,
+      "loss": 0.8989,
+      "step": 971
+    },
+    {
+      "epoch": 2.1350906095551894,
+      "grad_norm": 0.13126900792121887,
+      "learning_rate": 9.600977198697067e-05,
+      "loss": 1.0036,
+      "step": 972
+    },
+    {
+      "epoch": 2.13728720483251,
+      "grad_norm": 0.11103689670562744,
+      "learning_rate": 9.576547231270357e-05,
+      "loss": 0.8695,
+      "step": 973
+    },
+    {
+      "epoch": 2.1394838001098297,
+      "grad_norm": 0.11287889629602432,
+      "learning_rate": 9.552117263843646e-05,
+      "loss": 0.9219,
+      "step": 974
+    },
+    {
+      "epoch": 2.1416803953871497,
+      "grad_norm": 0.11684291809797287,
+      "learning_rate": 9.527687296416936e-05,
+      "loss": 0.9166,
+      "step": 975
+    },
+    {
+      "epoch": 2.14387699066447,
+      "grad_norm": 0.10588674247264862,
+      "learning_rate": 9.503257328990227e-05,
+      "loss": 0.9172,
+      "step": 976
+    },
+    {
+      "epoch": 2.14607358594179,
+      "grad_norm": 0.09471606463193893,
+      "learning_rate": 9.478827361563517e-05,
+      "loss": 0.8476,
+      "step": 977
+    },
+    {
+      "epoch": 2.1482701812191105,
+      "grad_norm": 0.09641019999980927,
+      "learning_rate": 9.454397394136806e-05,
+      "loss": 0.7556,
+      "step": 978
+    },
+    {
+      "epoch": 2.1504667764964305,
+      "grad_norm": 0.12420414388179779,
+      "learning_rate": 9.429967426710096e-05,
+      "loss": 0.8655,
+      "step": 979
+    },
+    {
+      "epoch": 2.152663371773751,
+      "grad_norm": 0.12752383947372437,
+      "learning_rate": 9.405537459283386e-05,
+      "loss": 0.991,
+      "step": 980
+    },
+    {
+      "epoch": 2.154859967051071,
+      "grad_norm": 0.14335161447525024,
+      "learning_rate": 9.381107491856675e-05,
+      "loss": 0.7357,
+      "step": 981
+    },
+    {
+      "epoch": 2.157056562328391,
+      "grad_norm": 0.14505070447921753,
+      "learning_rate": 9.356677524429967e-05,
+      "loss": 0.7465,
+      "step": 982
+    },
+    {
+      "epoch": 2.1592531576057112,
+      "grad_norm": 0.14414353668689728,
+      "learning_rate": 9.332247557003257e-05,
+      "loss": 1.04,
+      "step": 983
+    },
+    {
+      "epoch": 2.161449752883031,
+      "grad_norm": 0.17704938352108002,
+      "learning_rate": 9.307817589576547e-05,
+      "loss": 0.7817,
+      "step": 984
+    },
+    {
+      "epoch": 2.1636463481603516,
+      "grad_norm": 0.10942701250314713,
+      "learning_rate": 9.283387622149838e-05,
+      "loss": 0.7848,
+      "step": 985
+    },
+    {
+      "epoch": 2.1658429434376716,
+      "grad_norm": 0.14481937885284424,
+      "learning_rate": 9.258957654723127e-05,
+      "loss": 0.8194,
+      "step": 986
+    },
+    {
+      "epoch": 2.168039538714992,
+      "grad_norm": 0.11127373576164246,
+      "learning_rate": 9.234527687296417e-05,
+      "loss": 0.9727,
+      "step": 987
+    },
+    {
+      "epoch": 2.170236133992312,
+      "grad_norm": 0.1370663195848465,
+      "learning_rate": 9.210097719869707e-05,
+      "loss": 0.7761,
+      "step": 988
+    },
+    {
+      "epoch": 2.172432729269632,
+      "grad_norm": 0.11923271417617798,
+      "learning_rate": 9.185667752442996e-05,
+      "loss": 0.8624,
+      "step": 989
+    },
+    {
+      "epoch": 2.1746293245469523,
+      "grad_norm": 0.12829798460006714,
+      "learning_rate": 9.161237785016286e-05,
+      "loss": 1.0648,
+      "step": 990
+    },
+    {
+      "epoch": 2.1768259198242723,
+      "grad_norm": 0.3094155192375183,
+      "learning_rate": 9.136807817589577e-05,
+      "loss": 0.8541,
+      "step": 991
+    },
+    {
+      "epoch": 2.1790225151015927,
+      "grad_norm": 0.08770016580820084,
+      "learning_rate": 9.112377850162867e-05,
+      "loss": 0.7612,
+      "step": 992
+    },
+    {
+      "epoch": 2.1812191103789127,
+      "grad_norm": 0.0878802165389061,
+      "learning_rate": 9.087947882736156e-05,
+      "loss": 0.9159,
+      "step": 993
+    },
+    {
+      "epoch": 2.1834157056562327,
+      "grad_norm": 0.12802590429782867,
+      "learning_rate": 9.063517915309446e-05,
+      "loss": 0.7851,
+      "step": 994
+    },
+    {
+      "epoch": 2.185612300933553,
+      "grad_norm": 0.10365637391805649,
+      "learning_rate": 9.039087947882736e-05,
+      "loss": 0.7921,
+      "step": 995
+    },
+    {
+      "epoch": 2.187808896210873,
+      "grad_norm": 0.16769397258758545,
+      "learning_rate": 9.014657980456025e-05,
+      "loss": 0.9867,
+      "step": 996
+    },
+    {
+      "epoch": 2.1900054914881935,
+      "grad_norm": 0.1343819797039032,
+      "learning_rate": 8.990228013029316e-05,
+      "loss": 0.9561,
+      "step": 997
+    },
+    {
+      "epoch": 2.1922020867655134,
+      "grad_norm": 0.12204426527023315,
+      "learning_rate": 8.965798045602606e-05,
+      "loss": 0.905,
+      "step": 998
+    },
+    {
+      "epoch": 2.1943986820428334,
+      "grad_norm": 0.11351475119590759,
+      "learning_rate": 8.941368078175896e-05,
+      "loss": 1.0283,
+      "step": 999
+    },
+    {
+      "epoch": 2.196595277320154,
+      "grad_norm": 0.11063317954540253,
+      "learning_rate": 8.916938110749185e-05,
+      "loss": 0.8638,
+      "step": 1000
+    },
+    {
+      "epoch": 2.1987918725974738,
+      "grad_norm": 0.1318429857492447,
+      "learning_rate": 8.892508143322475e-05,
+      "loss": 0.7013,
+      "step": 1001
+    },
+    {
+      "epoch": 2.200988467874794,
+      "grad_norm": 0.0957917645573616,
+      "learning_rate": 8.868078175895766e-05,
+      "loss": 0.773,
+      "step": 1002
+    },
+    {
+      "epoch": 2.203185063152114,
+      "grad_norm": 0.14474697411060333,
+      "learning_rate": 8.843648208469055e-05,
+      "loss": 0.9124,
+      "step": 1003
+    },
+    {
+      "epoch": 2.2053816584294346,
+      "grad_norm": 0.10862885415554047,
+      "learning_rate": 8.819218241042345e-05,
+      "loss": 0.7838,
+      "step": 1004
+    },
+    {
+      "epoch": 2.2075782537067545,
+      "grad_norm": 0.10028444230556488,
+      "learning_rate": 8.794788273615635e-05,
+      "loss": 0.7802,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2097748489840745,
+      "grad_norm": 0.11436627805233002,
+      "learning_rate": 8.770358306188925e-05,
+      "loss": 0.9259,
+      "step": 1006
+    },
+    {
+      "epoch": 2.211971444261395,
+      "grad_norm": 0.09756966680288315,
+      "learning_rate": 8.745928338762214e-05,
+      "loss": 0.7317,
+      "step": 1007
+    },
+    {
+      "epoch": 2.214168039538715,
+      "grad_norm": 0.12304557859897614,
+      "learning_rate": 8.721498371335505e-05,
+      "loss": 1.1036,
+      "step": 1008
+    },
+    {
+      "epoch": 2.2163646348160353,
+      "grad_norm": 0.11333401501178741,
+      "learning_rate": 8.697068403908795e-05,
+      "loss": 0.8628,
+      "step": 1009
+    },
+    {
+      "epoch": 2.2185612300933553,
+      "grad_norm": 0.36374685168266296,
+      "learning_rate": 8.672638436482085e-05,
+      "loss": 0.9347,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2207578253706757,
+      "grad_norm": 0.08426238596439362,
+      "learning_rate": 8.648208469055374e-05,
+      "loss": 0.8296,
+      "step": 1011
+    },
+    {
+      "epoch": 2.2229544206479956,
+      "grad_norm": 0.12020387500524521,
+      "learning_rate": 8.623778501628664e-05,
+      "loss": 1.118,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2251510159253156,
+      "grad_norm": 0.14909891784191132,
+      "learning_rate": 8.599348534201955e-05,
+      "loss": 0.9468,
+      "step": 1013
+    },
+    {
+      "epoch": 2.227347611202636,
+      "grad_norm": 0.1345577836036682,
+      "learning_rate": 8.574918566775243e-05,
+      "loss": 1.0365,
+      "step": 1014
+    },
+    {
+      "epoch": 2.229544206479956,
+      "grad_norm": 0.15218985080718994,
+      "learning_rate": 8.550488599348534e-05,
+      "loss": 1.0291,
+      "step": 1015
+    },
+    {
+      "epoch": 2.2317408017572764,
+      "grad_norm": 0.10053347051143646,
+      "learning_rate": 8.526058631921824e-05,
+      "loss": 0.8846,
+      "step": 1016
+    },
+    {
+      "epoch": 2.2339373970345964,
+      "grad_norm": 0.12174846231937408,
+      "learning_rate": 8.501628664495114e-05,
+      "loss": 1.0328,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2361339923119163,
+      "grad_norm": 0.12445506453514099,
+      "learning_rate": 8.477198697068403e-05,
+      "loss": 1.057,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2383305875892368,
+      "grad_norm": 0.10489024221897125,
+      "learning_rate": 8.452768729641694e-05,
+      "loss": 0.9636,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2405271828665567,
+      "grad_norm": 0.17431600391864777,
+      "learning_rate": 8.428338762214984e-05,
+      "loss": 0.8872,
+      "step": 1020
+    },
+    {
+      "epoch": 2.242723778143877,
+      "grad_norm": 0.09695300459861755,
+      "learning_rate": 8.403908794788273e-05,
+      "loss": 0.8359,
+      "step": 1021
+    },
+    {
+      "epoch": 2.244920373421197,
+      "grad_norm": 0.12773211300373077,
+      "learning_rate": 8.379478827361563e-05,
+      "loss": 0.7038,
+      "step": 1022
+    },
+    {
+      "epoch": 2.247116968698517,
+      "grad_norm": 0.12343446165323257,
+      "learning_rate": 8.355048859934853e-05,
+      "loss": 0.9726,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2493135639758375,
+      "grad_norm": 0.1308927983045578,
+      "learning_rate": 8.330618892508144e-05,
+      "loss": 0.8544,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2515101592531575,
+      "grad_norm": 0.1014384925365448,
+      "learning_rate": 8.306188925081432e-05,
+      "loss": 0.8096,
+      "step": 1025
+    },
+    {
+      "epoch": 2.253706754530478,
+      "grad_norm": 0.10430674254894257,
+      "learning_rate": 8.281758957654723e-05,
+      "loss": 0.8047,
+      "step": 1026
+    },
+    {
+      "epoch": 2.255903349807798,
+      "grad_norm": 0.15577596426010132,
+      "learning_rate": 8.257328990228013e-05,
+      "loss": 0.8886,
+      "step": 1027
+    },
+    {
+      "epoch": 2.2580999450851182,
+      "grad_norm": 0.08472185581922531,
+      "learning_rate": 8.232899022801302e-05,
+      "loss": 0.9404,
+      "step": 1028
+    },
+    {
+      "epoch": 2.260296540362438,
+      "grad_norm": 0.12852627038955688,
+      "learning_rate": 8.208469055374592e-05,
+      "loss": 0.9345,
+      "step": 1029
+    },
+    {
+      "epoch": 2.262493135639758,
+      "grad_norm": 0.09691162407398224,
+      "learning_rate": 8.184039087947882e-05,
+      "loss": 0.733,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2646897309170786,
+      "grad_norm": 0.1022610142827034,
+      "learning_rate": 8.159609120521173e-05,
+      "loss": 0.724,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2668863261943986,
+      "grad_norm": 0.12456182390451431,
+      "learning_rate": 8.135179153094462e-05,
+      "loss": 0.7492,
+      "step": 1032
+    },
+    {
+      "epoch": 2.269082921471719,
+      "grad_norm": 0.1188085600733757,
+      "learning_rate": 8.110749185667752e-05,
+      "loss": 1.0635,
+      "step": 1033
+    },
+    {
+      "epoch": 2.271279516749039,
+      "grad_norm": 0.12036298215389252,
+      "learning_rate": 8.086319218241042e-05,
+      "loss": 0.8237,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2734761120263594,
+      "grad_norm": 0.12051574140787125,
+      "learning_rate": 8.061889250814331e-05,
+      "loss": 0.7685,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2756727073036793,
+      "grad_norm": 0.10017012804746628,
+      "learning_rate": 8.037459283387621e-05,
+      "loss": 0.8531,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2778693025809993,
+      "grad_norm": 0.08991660177707672,
+      "learning_rate": 8.013029315960912e-05,
+      "loss": 0.7253,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2800658978583197,
+      "grad_norm": 0.15187709033489227,
+      "learning_rate": 7.988599348534202e-05,
+      "loss": 0.821,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2822624931356397,
+      "grad_norm": 0.11432093381881714,
+      "learning_rate": 7.964169381107491e-05,
+      "loss": 0.7913,
+      "step": 1039
+    },
+    {
+      "epoch": 2.28445908841296,
+      "grad_norm": 0.1613302379846573,
+      "learning_rate": 7.939739413680781e-05,
+      "loss": 0.9062,
+      "step": 1040
+    },
+    {
+      "epoch": 2.28665568369028,
+      "grad_norm": 0.2552662491798401,
+      "learning_rate": 7.915309446254071e-05,
+      "loss": 0.8257,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2888522789676005,
+      "grad_norm": 0.1103154718875885,
+      "learning_rate": 7.890879478827362e-05,
+      "loss": 0.9108,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2910488742449204,
+      "grad_norm": 0.07838074862957001,
+      "learning_rate": 7.86644951140065e-05,
+      "loss": 0.7037,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2932454695222404,
+      "grad_norm": 0.10078336298465729,
+      "learning_rate": 7.842019543973941e-05,
+      "loss": 0.812,
+      "step": 1044
+    },
+    {
+      "epoch": 2.295442064799561,
+      "grad_norm": 0.09943167120218277,
+      "learning_rate": 7.817589576547231e-05,
+      "loss": 0.9444,
+      "step": 1045
+    },
+    {
+      "epoch": 2.297638660076881,
+      "grad_norm": 0.10206114500761032,
+      "learning_rate": 7.79315960912052e-05,
+      "loss": 0.8009,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2998352553542007,
+      "grad_norm": 0.14142203330993652,
+      "learning_rate": 7.76872964169381e-05,
+      "loss": 0.7574,
+      "step": 1047
+    },
+    {
+      "epoch": 2.302031850631521,
+      "grad_norm": 0.10557323694229126,
+      "learning_rate": 7.7442996742671e-05,
+      "loss": 0.8428,
+      "step": 1048
+    },
+    {
+      "epoch": 2.304228445908841,
+      "grad_norm": 0.10821036249399185,
+      "learning_rate": 7.719869706840391e-05,
+      "loss": 0.8547,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3064250411861615,
+      "grad_norm": 0.11686493456363678,
+      "learning_rate": 7.69543973941368e-05,
+      "loss": 0.9028,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3086216364634815,
+      "grad_norm": 0.12370723485946655,
+      "learning_rate": 7.67100977198697e-05,
+      "loss": 0.6684,
+      "step": 1051
+    },
+    {
+      "epoch": 2.310818231740802,
+      "grad_norm": 0.1100446954369545,
+      "learning_rate": 7.64657980456026e-05,
+      "loss": 0.816,
+      "step": 1052
+    },
+    {
+      "epoch": 2.313014827018122,
+      "grad_norm": 0.13455165922641754,
+      "learning_rate": 7.622149837133549e-05,
+      "loss": 0.8847,
+      "step": 1053
+    },
+    {
+      "epoch": 2.315211422295442,
+      "grad_norm": 0.14979985356330872,
+      "learning_rate": 7.59771986970684e-05,
+      "loss": 0.8127,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3174080175727623,
+      "grad_norm": 0.0960378497838974,
+      "learning_rate": 7.57328990228013e-05,
+      "loss": 1.0673,
+      "step": 1055
+    },
+    {
+      "epoch": 2.3196046128500822,
+      "grad_norm": 0.10235123336315155,
+      "learning_rate": 7.54885993485342e-05,
+      "loss": 0.8586,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3218012081274026,
+      "grad_norm": 0.12242686003446579,
+      "learning_rate": 7.524429967426709e-05,
+      "loss": 1.0777,
+      "step": 1057
+    },
+    {
+      "epoch": 2.3239978034047226,
+      "grad_norm": 0.2502252161502838,
+      "learning_rate": 7.5e-05,
+      "loss": 0.6739,
+      "step": 1058
+    },
+    {
+      "epoch": 2.326194398682043,
+      "grad_norm": 0.11071707308292389,
+      "learning_rate": 7.47557003257329e-05,
+      "loss": 0.7769,
+      "step": 1059
+    },
+    {
+      "epoch": 2.328390993959363,
+      "grad_norm": 0.12622836232185364,
+      "learning_rate": 7.451140065146579e-05,
+      "loss": 0.8454,
+      "step": 1060
+    },
+    {
+      "epoch": 2.330587589236683,
+      "grad_norm": 0.09597466140985489,
+      "learning_rate": 7.426710097719869e-05,
+      "loss": 0.845,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3327841845140034,
+      "grad_norm": 0.08889559656381607,
+      "learning_rate": 7.402280130293159e-05,
+      "loss": 0.8408,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3349807797913233,
+      "grad_norm": 0.1203259825706482,
+      "learning_rate": 7.37785016286645e-05,
+      "loss": 0.9444,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3371773750686438,
+      "grad_norm": 0.10422413796186447,
+      "learning_rate": 7.353420195439738e-05,
+      "loss": 0.902,
+      "step": 1064
+    },
+    {
+      "epoch": 2.3393739703459637,
+      "grad_norm": 0.2195027619600296,
+      "learning_rate": 7.328990228013029e-05,
+      "loss": 0.9222,
+      "step": 1065
+    },
+    {
+      "epoch": 2.341570565623284,
+      "grad_norm": 0.101964570581913,
+      "learning_rate": 7.304560260586319e-05,
+      "loss": 0.9849,
+      "step": 1066
+    },
+    {
+      "epoch": 2.343767160900604,
+      "grad_norm": 0.10780394077301025,
+      "learning_rate": 7.280130293159608e-05,
+      "loss": 0.7606,
+      "step": 1067
+    },
+    {
+      "epoch": 2.345963756177924,
+      "grad_norm": 0.09408123046159744,
+      "learning_rate": 7.255700325732898e-05,
+      "loss": 0.9145,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3481603514552445,
+      "grad_norm": 0.12819473445415497,
+      "learning_rate": 7.231270358306188e-05,
+      "loss": 1.0558,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3503569467325645,
+      "grad_norm": 0.1219649538397789,
+      "learning_rate": 7.206840390879479e-05,
+      "loss": 0.7906,
+      "step": 1070
+    },
+    {
+      "epoch": 2.352553542009885,
+      "grad_norm": 0.11465225368738174,
+      "learning_rate": 7.182410423452768e-05,
+      "loss": 0.897,
+      "step": 1071
+    },
+    {
+      "epoch": 2.354750137287205,
+      "grad_norm": 0.12268295884132385,
+      "learning_rate": 7.157980456026058e-05,
+      "loss": 0.7881,
+      "step": 1072
+    },
+    {
+      "epoch": 2.356946732564525,
+      "grad_norm": 0.11631671339273453,
+      "learning_rate": 7.133550488599348e-05,
+      "loss": 0.7811,
+      "step": 1073
+    },
+    {
+      "epoch": 2.359143327841845,
+      "grad_norm": 0.09588739275932312,
+      "learning_rate": 7.109120521172637e-05,
+      "loss": 0.7807,
+      "step": 1074
+    },
+    {
+      "epoch": 2.361339923119165,
+      "grad_norm": 0.10745790600776672,
+      "learning_rate": 7.084690553745927e-05,
+      "loss": 0.9771,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3635365183964856,
+      "grad_norm": 0.1461924910545349,
+      "learning_rate": 7.060260586319218e-05,
+      "loss": 1.0123,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3657331136738056,
+      "grad_norm": 0.11303440481424332,
+      "learning_rate": 7.035830618892508e-05,
+      "loss": 0.8168,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3679297089511255,
+      "grad_norm": 0.12605060636997223,
+      "learning_rate": 7.011400651465797e-05,
+      "loss": 0.9039,
+      "step": 1078
+    },
+    {
+      "epoch": 2.370126304228446,
+      "grad_norm": 0.10458989441394806,
+      "learning_rate": 6.986970684039087e-05,
+      "loss": 0.678,
+      "step": 1079
+    },
+    {
+      "epoch": 2.372322899505766,
+      "grad_norm": 0.1569211781024933,
+      "learning_rate": 6.962540716612377e-05,
+      "loss": 0.8849,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3745194947830863,
+      "grad_norm": 0.1404038667678833,
+      "learning_rate": 6.938110749185668e-05,
+      "loss": 1.1972,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3767160900604063,
+      "grad_norm": 0.11211780458688736,
+      "learning_rate": 6.913680781758956e-05,
+      "loss": 0.7859,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3789126853377267,
+      "grad_norm": 0.11469191312789917,
+      "learning_rate": 6.889250814332247e-05,
+      "loss": 0.8281,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3811092806150467,
+      "grad_norm": 0.1525568664073944,
+      "learning_rate": 6.864820846905537e-05,
+      "loss": 1.0841,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3833058758923666,
+      "grad_norm": 0.1954464316368103,
+      "learning_rate": 6.840390879478826e-05,
+      "loss": 1.0662,
+      "step": 1085
+    },
+    {
+      "epoch": 2.385502471169687,
+      "grad_norm": 0.1499062180519104,
+      "learning_rate": 6.815960912052116e-05,
+      "loss": 0.7981,
+      "step": 1086
+    },
+    {
+      "epoch": 2.387699066447007,
+      "grad_norm": 0.14987434446811676,
+      "learning_rate": 6.791530944625406e-05,
+      "loss": 1.1608,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3898956617243274,
+      "grad_norm": 0.11380533128976822,
+      "learning_rate": 6.767100977198697e-05,
+      "loss": 0.7659,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3920922570016474,
+      "grad_norm": 0.13894803822040558,
+      "learning_rate": 6.742671009771986e-05,
+      "loss": 0.994,
+      "step": 1089
+    },
+    {
+      "epoch": 2.394288852278968,
+      "grad_norm": 0.18096709251403809,
+      "learning_rate": 6.718241042345276e-05,
+      "loss": 0.853,
+      "step": 1090
+    },
+    {
+      "epoch": 2.396485447556288,
+      "grad_norm": 0.10410935431718826,
+      "learning_rate": 6.693811074918566e-05,
+      "loss": 0.846,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3986820428336078,
+      "grad_norm": 0.10668093711137772,
+      "learning_rate": 6.669381107491855e-05,
+      "loss": 0.9061,
+      "step": 1092
+    },
+    {
+      "epoch": 2.400878638110928,
+      "grad_norm": 0.10716383159160614,
+      "learning_rate": 6.644951140065145e-05,
+      "loss": 0.9242,
+      "step": 1093
+    },
+    {
+      "epoch": 2.403075233388248,
+      "grad_norm": 0.09179557859897614,
+      "learning_rate": 6.620521172638436e-05,
+      "loss": 0.9276,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4052718286655685,
+      "grad_norm": 0.1189119815826416,
+      "learning_rate": 6.596091205211726e-05,
+      "loss": 0.8073,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4074684239428885,
+      "grad_norm": 0.11901132017374039,
+      "learning_rate": 6.571661237785015e-05,
+      "loss": 0.6989,
+      "step": 1096
+    },
+    {
+      "epoch": 2.409665019220209,
+      "grad_norm": 0.11676941066980362,
+      "learning_rate": 6.547231270358307e-05,
+      "loss": 0.9521,
+      "step": 1097
+    },
+    {
+      "epoch": 2.411861614497529,
+      "grad_norm": 0.08545584231615067,
+      "learning_rate": 6.522801302931595e-05,
+      "loss": 0.8783,
+      "step": 1098
+    },
+    {
+      "epoch": 2.414058209774849,
+      "grad_norm": 0.13409164547920227,
+      "learning_rate": 6.498371335504886e-05,
+      "loss": 0.965,
+      "step": 1099
+    },
+    {
+      "epoch": 2.4162548050521693,
+      "grad_norm": 0.12696473300457,
+      "learning_rate": 6.473941368078176e-05,
+      "loss": 0.8853,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4184514003294892,
+      "grad_norm": 0.12019042670726776,
+      "learning_rate": 6.449511400651466e-05,
+      "loss": 0.878,
+      "step": 1101
+    },
+    {
+      "epoch": 2.420647995606809,
+      "grad_norm": 0.11318603157997131,
+      "learning_rate": 6.425081433224755e-05,
+      "loss": 0.9264,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4228445908841296,
+      "grad_norm": 0.12048106640577316,
+      "learning_rate": 6.400651465798045e-05,
+      "loss": 0.8827,
+      "step": 1103
+    },
+    {
+      "epoch": 2.4250411861614496,
+      "grad_norm": 0.13073910772800446,
+      "learning_rate": 6.376221498371336e-05,
+      "loss": 0.9585,
+      "step": 1104
+    },
+    {
+      "epoch": 2.42723778143877,
+      "grad_norm": 0.1316949427127838,
+      "learning_rate": 6.351791530944625e-05,
+      "loss": 0.9124,
+      "step": 1105
+    },
+    {
+      "epoch": 2.42943437671609,
+      "grad_norm": 0.13904504477977753,
+      "learning_rate": 6.327361563517915e-05,
+      "loss": 0.859,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4316309719934104,
+      "grad_norm": 0.11975440382957458,
+      "learning_rate": 6.302931596091205e-05,
+      "loss": 0.9366,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4338275672707304,
+      "grad_norm": 0.1158430278301239,
+      "learning_rate": 6.278501628664495e-05,
+      "loss": 0.9697,
+      "step": 1108
+    },
+    {
+      "epoch": 2.4360241625480503,
+      "grad_norm": 0.10369361191987991,
+      "learning_rate": 6.254071661237784e-05,
+      "loss": 0.8445,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4382207578253707,
+      "grad_norm": 0.08451946079730988,
+      "learning_rate": 6.229641693811075e-05,
+      "loss": 0.9052,
+      "step": 1110
+    },
+    {
+      "epoch": 2.4404173531026907,
+      "grad_norm": 0.08018508553504944,
+      "learning_rate": 6.205211726384365e-05,
+      "loss": 0.8352,
+      "step": 1111
+    },
+    {
+      "epoch": 2.442613948380011,
+      "grad_norm": 0.19825762510299683,
+      "learning_rate": 6.180781758957654e-05,
+      "loss": 0.7406,
+      "step": 1112
+    },
+    {
+      "epoch": 2.444810543657331,
+      "grad_norm": 0.11737879365682602,
+      "learning_rate": 6.156351791530944e-05,
+      "loss": 1.0757,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4470071389346515,
+      "grad_norm": 0.1253841668367386,
+      "learning_rate": 6.131921824104234e-05,
+      "loss": 1.0581,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4492037342119715,
+      "grad_norm": 0.13745571672916412,
+      "learning_rate": 6.107491856677525e-05,
+      "loss": 0.8273,
+      "step": 1115
+    },
+    {
+      "epoch": 2.4514003294892914,
+      "grad_norm": 0.11050701886415482,
+      "learning_rate": 6.083061889250814e-05,
+      "loss": 0.7525,
+      "step": 1116
+    },
+    {
+      "epoch": 2.453596924766612,
+      "grad_norm": 0.14010542631149292,
+      "learning_rate": 6.058631921824104e-05,
+      "loss": 0.8932,
+      "step": 1117
+    },
+    {
+      "epoch": 2.455793520043932,
+      "grad_norm": 0.10861250013113022,
+      "learning_rate": 6.0342019543973935e-05,
+      "loss": 0.8726,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4579901153212522,
+      "grad_norm": 0.11014145612716675,
+      "learning_rate": 6.009771986970684e-05,
+      "loss": 0.9023,
+      "step": 1119
+    },
+    {
+      "epoch": 2.460186710598572,
+      "grad_norm": 0.10450796037912369,
+      "learning_rate": 5.9853420195439734e-05,
+      "loss": 0.8538,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4623833058758926,
+      "grad_norm": 0.11975032091140747,
+      "learning_rate": 5.9609120521172636e-05,
+      "loss": 0.8931,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4645799011532126,
+      "grad_norm": 0.15808100998401642,
+      "learning_rate": 5.936482084690553e-05,
+      "loss": 1.2687,
+      "step": 1122
+    },
+    {
+      "epoch": 2.4667764964305325,
+      "grad_norm": 0.12776662409305573,
+      "learning_rate": 5.9120521172638435e-05,
+      "loss": 0.901,
+      "step": 1123
+    },
+    {
+      "epoch": 2.468973091707853,
+      "grad_norm": 0.13078485429286957,
+      "learning_rate": 5.887622149837133e-05,
+      "loss": 0.9012,
+      "step": 1124
+    },
+    {
+      "epoch": 2.471169686985173,
+      "grad_norm": 0.15072587132453918,
+      "learning_rate": 5.863192182410423e-05,
+      "loss": 0.752,
+      "step": 1125
+    },
+    {
+      "epoch": 2.473366282262493,
+      "grad_norm": 0.16593004763126373,
+      "learning_rate": 5.838762214983713e-05,
+      "loss": 1.0136,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4755628775398133,
+      "grad_norm": 0.13337472081184387,
+      "learning_rate": 5.8143322475570026e-05,
+      "loss": 0.7877,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4777594728171333,
+      "grad_norm": 0.07173790037631989,
+      "learning_rate": 5.789902280130293e-05,
+      "loss": 0.8314,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4799560680944537,
+      "grad_norm": 0.12199293076992035,
+      "learning_rate": 5.7654723127035825e-05,
+      "loss": 1.1241,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4821526633717736,
+      "grad_norm": 0.09257610142230988,
+      "learning_rate": 5.741042345276873e-05,
+      "loss": 0.8616,
+      "step": 1130
+    },
+    {
+      "epoch": 2.484349258649094,
+      "grad_norm": 1.7121655941009521,
+      "learning_rate": 5.716612377850162e-05,
+      "loss": 0.817,
+      "step": 1131
+    },
+    {
+      "epoch": 2.486545853926414,
+      "grad_norm": 0.14271928369998932,
+      "learning_rate": 5.6921824104234526e-05,
+      "loss": 0.8963,
+      "step": 1132
+    },
+    {
+      "epoch": 2.488742449203734,
+      "grad_norm": 0.11492665857076645,
+      "learning_rate": 5.667752442996742e-05,
+      "loss": 0.7439,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4909390444810544,
+      "grad_norm": 0.11374574154615402,
+      "learning_rate": 5.643322475570032e-05,
+      "loss": 0.9042,
+      "step": 1134
+    },
+    {
+      "epoch": 2.4931356397583744,
+      "grad_norm": 0.45574748516082764,
+      "learning_rate": 5.618892508143322e-05,
+      "loss": 0.9122,
+      "step": 1135
+    },
+    {
+      "epoch": 2.495332235035695,
+      "grad_norm": 0.10924903303384781,
+      "learning_rate": 5.594462540716612e-05,
+      "loss": 0.7695,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4975288303130148,
+      "grad_norm": 0.09882228821516037,
+      "learning_rate": 5.570032573289902e-05,
+      "loss": 0.6651,
+      "step": 1137
+    },
+    {
+      "epoch": 2.499725425590335,
+      "grad_norm": 0.11616279184818268,
+      "learning_rate": 5.5456026058631915e-05,
+      "loss": 0.8962,
+      "step": 1138
+    },
+    {
+      "epoch": 2.501922020867655,
+      "grad_norm": 0.11904437094926834,
+      "learning_rate": 5.521172638436482e-05,
+      "loss": 1.0351,
+      "step": 1139
+    },
+    {
+      "epoch": 2.504118616144975,
+      "grad_norm": 0.16439133882522583,
+      "learning_rate": 5.4967426710097714e-05,
+      "loss": 0.8393,
+      "step": 1140
+    },
+    {
+      "epoch": 2.5063152114222955,
+      "grad_norm": 0.1203792616724968,
+      "learning_rate": 5.472312703583061e-05,
+      "loss": 0.8061,
+      "step": 1141
+    },
+    {
+      "epoch": 2.5085118066996155,
+      "grad_norm": 0.10818367451429367,
+      "learning_rate": 5.447882736156351e-05,
+      "loss": 0.9078,
+      "step": 1142
+    },
+    {
+      "epoch": 2.510708401976936,
+      "grad_norm": 0.12388180196285248,
+      "learning_rate": 5.423452768729641e-05,
+      "loss": 1.0576,
+      "step": 1143
+    },
+    {
+      "epoch": 2.512904997254256,
+      "grad_norm": 0.09219598770141602,
+      "learning_rate": 5.399022801302931e-05,
+      "loss": 0.8425,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5151015925315763,
+      "grad_norm": 0.11083900928497314,
+      "learning_rate": 5.374592833876221e-05,
+      "loss": 0.8014,
+      "step": 1145
+    },
+    {
+      "epoch": 2.5172981878088962,
+      "grad_norm": 0.14442667365074158,
+      "learning_rate": 5.350162866449511e-05,
+      "loss": 0.9311,
+      "step": 1146
+    },
+    {
+      "epoch": 2.519494783086216,
+      "grad_norm": 0.09543650597333908,
+      "learning_rate": 5.3257328990228006e-05,
+      "loss": 0.7896,
+      "step": 1147
+    },
+    {
+      "epoch": 2.5216913783635366,
+      "grad_norm": 0.11851215362548828,
+      "learning_rate": 5.301302931596091e-05,
+      "loss": 0.7517,
+      "step": 1148
+    },
+    {
+      "epoch": 2.5238879736408566,
+      "grad_norm": 0.09310846030712128,
+      "learning_rate": 5.2768729641693805e-05,
+      "loss": 0.9193,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5260845689181766,
+      "grad_norm": 0.10620040446519852,
+      "learning_rate": 5.25244299674267e-05,
+      "loss": 1.0673,
+      "step": 1150
+    },
+    {
+      "epoch": 2.528281164195497,
+      "grad_norm": 0.08555366098880768,
+      "learning_rate": 5.2280130293159604e-05,
+      "loss": 0.8518,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5304777594728174,
+      "grad_norm": 0.09683895111083984,
+      "learning_rate": 5.20358306188925e-05,
+      "loss": 0.8163,
+      "step": 1152
+    },
+    {
+      "epoch": 2.5326743547501374,
+      "grad_norm": 0.09580101817846298,
+      "learning_rate": 5.17915309446254e-05,
+      "loss": 0.8809,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5348709500274573,
+      "grad_norm": 0.10482166707515717,
+      "learning_rate": 5.15472312703583e-05,
+      "loss": 0.824,
+      "step": 1154
+    },
+    {
+      "epoch": 2.5370675453047777,
+      "grad_norm": 0.09214266389608383,
+      "learning_rate": 5.13029315960912e-05,
+      "loss": 0.8088,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5392641405820977,
+      "grad_norm": 0.16963458061218262,
+      "learning_rate": 5.10586319218241e-05,
+      "loss": 0.9941,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5414607358594177,
+      "grad_norm": 0.16200727224349976,
+      "learning_rate": 5.081433224755699e-05,
+      "loss": 1.0362,
+      "step": 1157
+    },
+    {
+      "epoch": 2.543657331136738,
+      "grad_norm": 0.11370562016963959,
+      "learning_rate": 5.0570032573289896e-05,
+      "loss": 0.786,
+      "step": 1158
+    },
+    {
+      "epoch": 2.545853926414058,
+      "grad_norm": 0.1073915883898735,
+      "learning_rate": 5.032573289902279e-05,
+      "loss": 0.7657,
+      "step": 1159
+    },
+    {
+      "epoch": 2.5480505216913785,
+      "grad_norm": 0.12035344541072845,
+      "learning_rate": 5.0081433224755695e-05,
+      "loss": 0.8303,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5502471169686984,
+      "grad_norm": 0.1374288648366928,
+      "learning_rate": 4.983713355048859e-05,
+      "loss": 0.7293,
+      "step": 1161
+    },
+    {
+      "epoch": 2.552443712246019,
+      "grad_norm": 0.0995606780052185,
+      "learning_rate": 4.9592833876221493e-05,
+      "loss": 0.8834,
+      "step": 1162
+    },
+    {
+      "epoch": 2.554640307523339,
+      "grad_norm": 0.31442973017692566,
+      "learning_rate": 4.934853420195439e-05,
+      "loss": 0.8309,
+      "step": 1163
+    },
+    {
+      "epoch": 2.556836902800659,
+      "grad_norm": 0.1037110835313797,
+      "learning_rate": 4.910423452768729e-05,
+      "loss": 0.9258,
+      "step": 1164
+    },
+    {
+      "epoch": 2.559033498077979,
+      "grad_norm": 0.09194368124008179,
+      "learning_rate": 4.885993485342019e-05,
+      "loss": 0.9448,
+      "step": 1165
+    },
+    {
+      "epoch": 2.561230093355299,
+      "grad_norm": 0.08944636583328247,
+      "learning_rate": 4.8615635179153084e-05,
+      "loss": 0.8009,
+      "step": 1166
+    },
+    {
+      "epoch": 2.5634266886326196,
+      "grad_norm": 0.10588474571704865,
+      "learning_rate": 4.837133550488599e-05,
+      "loss": 0.7626,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5656232839099395,
+      "grad_norm": 0.1165136769413948,
+      "learning_rate": 4.812703583061888e-05,
+      "loss": 0.886,
+      "step": 1168
+    },
+    {
+      "epoch": 2.56781987918726,
+      "grad_norm": 0.2114308625459671,
+      "learning_rate": 4.7882736156351786e-05,
+      "loss": 0.8946,
+      "step": 1169
+    },
+    {
+      "epoch": 2.57001647446458,
+      "grad_norm": 0.12104310095310211,
+      "learning_rate": 4.763843648208468e-05,
+      "loss": 0.7778,
+      "step": 1170
+    },
+    {
+      "epoch": 2.5722130697419,
+      "grad_norm": 0.10215938091278076,
+      "learning_rate": 4.7394136807817584e-05,
+      "loss": 0.7395,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5744096650192203,
+      "grad_norm": 0.16481240093708038,
+      "learning_rate": 4.714983713355048e-05,
+      "loss": 0.8767,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5766062602965403,
+      "grad_norm": 0.09567465633153915,
+      "learning_rate": 4.6905537459283376e-05,
+      "loss": 0.7702,
+      "step": 1173
+    },
+    {
+      "epoch": 2.5788028555738602,
+      "grad_norm": 0.17528395354747772,
+      "learning_rate": 4.6661237785016286e-05,
+      "loss": 0.9397,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5809994508511807,
+      "grad_norm": 0.13244639337062836,
+      "learning_rate": 4.641693811074919e-05,
+      "loss": 0.9044,
+      "step": 1175
+    },
+    {
+      "epoch": 2.583196046128501,
+      "grad_norm": 0.11964637041091919,
+      "learning_rate": 4.6172638436482085e-05,
+      "loss": 0.8062,
+      "step": 1176
+    },
+    {
+      "epoch": 2.585392641405821,
+      "grad_norm": 0.11471062153577805,
+      "learning_rate": 4.592833876221498e-05,
+      "loss": 1.0314,
+      "step": 1177
+    },
+    {
+      "epoch": 2.587589236683141,
+      "grad_norm": 0.17589324712753296,
+      "learning_rate": 4.568403908794788e-05,
+      "loss": 0.8285,
+      "step": 1178
+    },
+    {
+      "epoch": 2.5897858319604614,
+      "grad_norm": 0.17229029536247253,
+      "learning_rate": 4.543973941368078e-05,
+      "loss": 1.0243,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5919824272377814,
+      "grad_norm": 0.11634933948516846,
+      "learning_rate": 4.519543973941368e-05,
+      "loss": 0.797,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5941790225151014,
+      "grad_norm": 0.15863265097141266,
+      "learning_rate": 4.495114006514658e-05,
+      "loss": 1.0276,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5963756177924218,
+      "grad_norm": 0.11726535111665726,
+      "learning_rate": 4.470684039087948e-05,
+      "loss": 0.8456,
+      "step": 1182
+    },
+    {
+      "epoch": 2.598572213069742,
+      "grad_norm": 0.12106660008430481,
+      "learning_rate": 4.446254071661238e-05,
+      "loss": 0.9347,
+      "step": 1183
+    },
+    {
+      "epoch": 2.600768808347062,
+      "grad_norm": 0.1864086091518402,
+      "learning_rate": 4.421824104234527e-05,
+      "loss": 0.9487,
+      "step": 1184
+    },
+    {
+      "epoch": 2.602965403624382,
+      "grad_norm": 0.10796128213405609,
+      "learning_rate": 4.3973941368078175e-05,
+      "loss": 0.9071,
+      "step": 1185
+    },
+    {
+      "epoch": 2.6051619989017025,
+      "grad_norm": 0.11755198240280151,
+      "learning_rate": 4.372964169381107e-05,
+      "loss": 0.8967,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6073585941790225,
+      "grad_norm": 0.08717940002679825,
+      "learning_rate": 4.3485342019543974e-05,
+      "loss": 0.7675,
+      "step": 1187
+    },
+    {
+      "epoch": 2.6095551894563425,
+      "grad_norm": 0.0905216708779335,
+      "learning_rate": 4.324104234527687e-05,
+      "loss": 0.8314,
+      "step": 1188
+    },
+    {
+      "epoch": 2.611751784733663,
+      "grad_norm": 0.10230715572834015,
+      "learning_rate": 4.299674267100977e-05,
+      "loss": 0.81,
+      "step": 1189
+    },
+    {
+      "epoch": 2.613948380010983,
+      "grad_norm": 0.1062103882431984,
+      "learning_rate": 4.275244299674267e-05,
+      "loss": 0.8503,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6161449752883033,
+      "grad_norm": 0.09377019852399826,
+      "learning_rate": 4.250814332247557e-05,
+      "loss": 0.8295,
+      "step": 1191
+    },
+    {
+      "epoch": 2.618341570565623,
+      "grad_norm": 0.14294861257076263,
+      "learning_rate": 4.226384364820847e-05,
+      "loss": 0.683,
+      "step": 1192
+    },
+    {
+      "epoch": 2.6205381658429436,
+      "grad_norm": 0.14558130502700806,
+      "learning_rate": 4.2019543973941364e-05,
+      "loss": 1.3493,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6227347611202636,
+      "grad_norm": 0.15816231071949005,
+      "learning_rate": 4.1775244299674266e-05,
+      "loss": 1.0111,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6249313563975836,
+      "grad_norm": 0.09366089105606079,
+      "learning_rate": 4.153094462540716e-05,
+      "loss": 0.5664,
+      "step": 1195
+    },
+    {
+      "epoch": 2.627127951674904,
+      "grad_norm": 0.1251467764377594,
+      "learning_rate": 4.1286644951140065e-05,
+      "loss": 0.8534,
+      "step": 1196
+    },
+    {
+      "epoch": 2.629324546952224,
+      "grad_norm": 0.14125970005989075,
+      "learning_rate": 4.104234527687296e-05,
+      "loss": 0.8769,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6315211422295444,
+      "grad_norm": 0.10542935132980347,
+      "learning_rate": 4.0798045602605864e-05,
+      "loss": 0.9059,
+      "step": 1198
+    },
+    {
+      "epoch": 2.6337177375068643,
+      "grad_norm": 0.10802783071994781,
+      "learning_rate": 4.055374592833876e-05,
+      "loss": 0.6928,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6359143327841847,
+      "grad_norm": 0.10232524573802948,
+      "learning_rate": 4.0309446254071656e-05,
+      "loss": 0.8478,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6381109280615047,
+      "grad_norm": 0.09960202127695084,
+      "learning_rate": 4.006514657980456e-05,
+      "loss": 0.708,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6403075233388247,
+      "grad_norm": 0.12340527027845383,
+      "learning_rate": 3.9820846905537454e-05,
+      "loss": 0.7051,
+      "step": 1202
+    },
+    {
+      "epoch": 2.642504118616145,
+      "grad_norm": 0.19421131908893585,
+      "learning_rate": 3.957654723127036e-05,
+      "loss": 1.068,
+      "step": 1203
+    },
+    {
+      "epoch": 2.644700713893465,
+      "grad_norm": 0.12119914591312408,
+      "learning_rate": 3.933224755700325e-05,
+      "loss": 0.8252,
+      "step": 1204
+    },
+    {
+      "epoch": 2.646897309170785,
+      "grad_norm": 0.10593686997890472,
+      "learning_rate": 3.9087947882736156e-05,
+      "loss": 0.9415,
+      "step": 1205
+    },
+    {
+      "epoch": 2.6490939044481054,
+      "grad_norm": 0.13152405619621277,
+      "learning_rate": 3.884364820846905e-05,
+      "loss": 0.9931,
+      "step": 1206
+    },
+    {
+      "epoch": 2.651290499725426,
+      "grad_norm": 0.09771620482206345,
+      "learning_rate": 3.8599348534201955e-05,
+      "loss": 0.9239,
+      "step": 1207
+    },
+    {
+      "epoch": 2.653487095002746,
+      "grad_norm": 0.1133955642580986,
+      "learning_rate": 3.835504885993485e-05,
+      "loss": 0.8576,
+      "step": 1208
+    },
+    {
+      "epoch": 2.655683690280066,
+      "grad_norm": 0.14273396134376526,
+      "learning_rate": 3.8110749185667747e-05,
+      "loss": 0.7426,
+      "step": 1209
+    },
+    {
+      "epoch": 2.657880285557386,
+      "grad_norm": 0.1126021072268486,
+      "learning_rate": 3.786644951140065e-05,
+      "loss": 1.0656,
+      "step": 1210
+    },
+    {
+      "epoch": 2.660076880834706,
+      "grad_norm": 0.12580552697181702,
+      "learning_rate": 3.7622149837133545e-05,
+      "loss": 0.8081,
+      "step": 1211
+    },
+    {
+      "epoch": 2.662273476112026,
+      "grad_norm": 0.45561549067497253,
+      "learning_rate": 3.737785016286645e-05,
+      "loss": 0.812,
+      "step": 1212
+    },
+    {
+      "epoch": 2.6644700713893466,
+      "grad_norm": 0.11111915111541748,
+      "learning_rate": 3.7133550488599344e-05,
+      "loss": 0.9267,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.14537368714809418,
+      "learning_rate": 3.688925081433225e-05,
+      "loss": 0.8071,
+      "step": 1214
+    },
+    {
+      "epoch": 2.668863261943987,
+      "grad_norm": 0.1004401445388794,
+      "learning_rate": 3.664495114006514e-05,
+      "loss": 0.8361,
+      "step": 1215
+    },
+    {
+      "epoch": 2.671059857221307,
+      "grad_norm": 0.09084270894527435,
+      "learning_rate": 3.640065146579804e-05,
+      "loss": 0.8711,
+      "step": 1216
+    },
+    {
+      "epoch": 2.6732564524986273,
+      "grad_norm": 0.1070455014705658,
+      "learning_rate": 3.615635179153094e-05,
+      "loss": 0.9142,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6754530477759473,
+      "grad_norm": 0.11680350452661514,
+      "learning_rate": 3.591205211726384e-05,
+      "loss": 0.8034,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6776496430532672,
+      "grad_norm": 0.11271864175796509,
+      "learning_rate": 3.566775244299674e-05,
+      "loss": 0.8032,
+      "step": 1219
+    },
+    {
+      "epoch": 2.6798462383305877,
+      "grad_norm": 0.11815498024225235,
+      "learning_rate": 3.5423452768729636e-05,
+      "loss": 0.8866,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6820428336079076,
+      "grad_norm": 0.13534937798976898,
+      "learning_rate": 3.517915309446254e-05,
+      "loss": 1.085,
+      "step": 1221
+    },
+    {
+      "epoch": 2.684239428885228,
+      "grad_norm": 0.10078220069408417,
+      "learning_rate": 3.4934853420195435e-05,
+      "loss": 0.6457,
+      "step": 1222
+    },
+    {
+      "epoch": 2.686436024162548,
+      "grad_norm": 0.10122572630643845,
+      "learning_rate": 3.469055374592834e-05,
+      "loss": 0.9317,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6886326194398684,
+      "grad_norm": 0.13453632593154907,
+      "learning_rate": 3.4446254071661234e-05,
+      "loss": 0.8689,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6908292147171884,
+      "grad_norm": 0.10447756946086884,
+      "learning_rate": 3.420195439739413e-05,
+      "loss": 0.9565,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6930258099945084,
+      "grad_norm": 0.134364515542984,
+      "learning_rate": 3.395765472312703e-05,
+      "loss": 0.7717,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6952224052718288,
+      "grad_norm": 0.11273462325334549,
+      "learning_rate": 3.371335504885993e-05,
+      "loss": 0.9509,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6974190005491487,
+      "grad_norm": 0.09328631311655045,
+      "learning_rate": 3.346905537459283e-05,
+      "loss": 0.8198,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6996155958264687,
+      "grad_norm": 0.10468220710754395,
+      "learning_rate": 3.322475570032573e-05,
+      "loss": 0.9748,
+      "step": 1229
+    },
+    {
+      "epoch": 2.701812191103789,
+      "grad_norm": 0.11480540037155151,
+      "learning_rate": 3.298045602605863e-05,
+      "loss": 0.7446,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7040087863811095,
+      "grad_norm": 0.10628961771726608,
+      "learning_rate": 3.273615635179153e-05,
+      "loss": 0.7638,
+      "step": 1231
+    },
+    {
+      "epoch": 2.7062053816584295,
+      "grad_norm": 0.14755621552467346,
+      "learning_rate": 3.249185667752443e-05,
+      "loss": 0.9475,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7084019769357495,
+      "grad_norm": 0.1164088174700737,
+      "learning_rate": 3.224755700325733e-05,
+      "loss": 0.8001,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71059857221307,
+      "grad_norm": 0.129755437374115,
+      "learning_rate": 3.200325732899023e-05,
+      "loss": 0.8465,
+      "step": 1234
+    },
+    {
+      "epoch": 2.71279516749039,
+      "grad_norm": 0.08108958601951599,
+      "learning_rate": 3.175895765472312e-05,
+      "loss": 0.9232,
+      "step": 1235
+    },
+    {
+      "epoch": 2.71499176276771,
+      "grad_norm": 0.11728141456842422,
+      "learning_rate": 3.1514657980456026e-05,
+      "loss": 1.089,
+      "step": 1236
+    },
+    {
+      "epoch": 2.7171883580450302,
+      "grad_norm": 0.12024188041687012,
+      "learning_rate": 3.127035830618892e-05,
+      "loss": 0.8768,
+      "step": 1237
+    },
+    {
+      "epoch": 2.71938495332235,
+      "grad_norm": 0.17721788585186005,
+      "learning_rate": 3.1026058631921825e-05,
+      "loss": 0.9105,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7215815485996706,
+      "grad_norm": 0.14233282208442688,
+      "learning_rate": 3.078175895765472e-05,
+      "loss": 0.8457,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7237781438769906,
+      "grad_norm": 0.09806663542985916,
+      "learning_rate": 3.0537459283387624e-05,
+      "loss": 0.8724,
+      "step": 1240
+    },
+    {
+      "epoch": 2.725974739154311,
+      "grad_norm": 0.1209288164973259,
+      "learning_rate": 3.029315960912052e-05,
+      "loss": 0.6792,
+      "step": 1241
+    },
+    {
+      "epoch": 2.728171334431631,
+      "grad_norm": 0.1301451325416565,
+      "learning_rate": 3.004885993485342e-05,
+      "loss": 0.8894,
+      "step": 1242
+    },
+    {
+      "epoch": 2.730367929708951,
+      "grad_norm": 0.10529869794845581,
+      "learning_rate": 2.9804560260586318e-05,
+      "loss": 0.945,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7325645249862713,
+      "grad_norm": 0.10522414743900299,
+      "learning_rate": 2.9560260586319218e-05,
+      "loss": 0.899,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7347611202635913,
+      "grad_norm": 0.13185814023017883,
+      "learning_rate": 2.9315960912052114e-05,
+      "loss": 0.8959,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7369577155409117,
+      "grad_norm": 0.15432147681713104,
+      "learning_rate": 2.9071661237785013e-05,
+      "loss": 0.9493,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7391543108182317,
+      "grad_norm": 0.2684352993965149,
+      "learning_rate": 2.8827361563517912e-05,
+      "loss": 1.395,
+      "step": 1247
+    },
+    {
+      "epoch": 2.741350906095552,
+      "grad_norm": 0.09965763986110687,
+      "learning_rate": 2.858306188925081e-05,
+      "loss": 0.863,
+      "step": 1248
+    },
+    {
+      "epoch": 2.743547501372872,
+      "grad_norm": 0.1661844551563263,
+      "learning_rate": 2.833876221498371e-05,
+      "loss": 0.9552,
+      "step": 1249
+    },
+    {
+      "epoch": 2.745744096650192,
+      "grad_norm": 0.0882975161075592,
+      "learning_rate": 2.809446254071661e-05,
+      "loss": 0.872,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7479406919275124,
+      "grad_norm": 0.09738167375326157,
+      "learning_rate": 2.785016286644951e-05,
+      "loss": 0.996,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7501372872048324,
+      "grad_norm": 0.1423174887895584,
+      "learning_rate": 2.760586319218241e-05,
+      "loss": 0.9607,
+      "step": 1252
+    },
+    {
+      "epoch": 2.752333882482153,
+      "grad_norm": 0.12437476217746735,
+      "learning_rate": 2.7361563517915305e-05,
+      "loss": 0.764,
+      "step": 1253
+    },
+    {
+      "epoch": 2.754530477759473,
+      "grad_norm": 0.10257714986801147,
+      "learning_rate": 2.7117263843648204e-05,
+      "loss": 0.7773,
+      "step": 1254
+    },
+    {
+      "epoch": 2.756727073036793,
+      "grad_norm": 0.1244991198182106,
+      "learning_rate": 2.6872964169381104e-05,
+      "loss": 0.9961,
+      "step": 1255
+    },
+    {
+      "epoch": 2.758923668314113,
+      "grad_norm": 0.11062435805797577,
+      "learning_rate": 2.6628664495114003e-05,
+      "loss": 0.681,
+      "step": 1256
+    },
+    {
+      "epoch": 2.761120263591433,
+      "grad_norm": 0.1060105711221695,
+      "learning_rate": 2.6384364820846903e-05,
+      "loss": 0.8526,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7633168588687536,
+      "grad_norm": 0.09225158393383026,
+      "learning_rate": 2.6140065146579802e-05,
+      "loss": 0.7565,
+      "step": 1258
+    },
+    {
+      "epoch": 2.7655134541460735,
+      "grad_norm": 0.13103348016738892,
+      "learning_rate": 2.58957654723127e-05,
+      "loss": 1.015,
+      "step": 1259
+    },
+    {
+      "epoch": 2.7677100494233935,
+      "grad_norm": 0.12131506949663162,
+      "learning_rate": 2.56514657980456e-05,
+      "loss": 0.9368,
+      "step": 1260
+    },
+    {
+      "epoch": 2.769906644700714,
+      "grad_norm": 0.1271435022354126,
+      "learning_rate": 2.5407166123778497e-05,
+      "loss": 0.8401,
+      "step": 1261
+    },
+    {
+      "epoch": 2.7721032399780343,
+      "grad_norm": 0.12489993870258331,
+      "learning_rate": 2.5162866449511396e-05,
+      "loss": 0.8848,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7742998352553543,
+      "grad_norm": 0.19427073001861572,
+      "learning_rate": 2.4918566775244295e-05,
+      "loss": 0.9358,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7764964305326743,
+      "grad_norm": 0.08987414836883545,
+      "learning_rate": 2.4674267100977195e-05,
+      "loss": 1.0088,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7786930258099947,
+      "grad_norm": 0.08769197016954422,
+      "learning_rate": 2.4429967426710094e-05,
+      "loss": 0.9239,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7808896210873146,
+      "grad_norm": 0.08131453394889832,
+      "learning_rate": 2.4185667752442993e-05,
+      "loss": 0.7421,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7830862163646346,
+      "grad_norm": 0.09779416769742966,
+      "learning_rate": 2.3941368078175893e-05,
+      "loss": 0.7556,
+      "step": 1267
+    },
+    {
+      "epoch": 2.785282811641955,
+      "grad_norm": 0.11181338131427765,
+      "learning_rate": 2.3697068403908792e-05,
+      "loss": 0.8297,
+      "step": 1268
+    },
+    {
+      "epoch": 2.787479406919275,
+      "grad_norm": 0.11137125641107559,
+      "learning_rate": 2.3452768729641688e-05,
+      "loss": 0.7445,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7896760021965954,
+      "grad_norm": 0.10434068739414215,
+      "learning_rate": 2.3208469055374594e-05,
+      "loss": 0.819,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7918725974739154,
+      "grad_norm": 0.11664993315935135,
+      "learning_rate": 2.296416938110749e-05,
+      "loss": 0.8675,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7940691927512358,
+      "grad_norm": 0.1124923825263977,
+      "learning_rate": 2.271986970684039e-05,
+      "loss": 0.8802,
+      "step": 1272
+    },
+    {
+      "epoch": 2.7962657880285557,
+      "grad_norm": 0.08510121703147888,
+      "learning_rate": 2.247557003257329e-05,
+      "loss": 0.8887,
+      "step": 1273
+    },
+    {
+      "epoch": 2.7984623833058757,
+      "grad_norm": 0.11321862041950226,
+      "learning_rate": 2.223127035830619e-05,
+      "loss": 1.0771,
+      "step": 1274
+    },
+    {
+      "epoch": 2.800658978583196,
+      "grad_norm": 0.11527451127767563,
+      "learning_rate": 2.1986970684039088e-05,
+      "loss": 0.8285,
+      "step": 1275
+    },
+    {
+      "epoch": 2.802855573860516,
+      "grad_norm": 0.13034583628177643,
+      "learning_rate": 2.1742671009771987e-05,
+      "loss": 1.0537,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8050521691378365,
+      "grad_norm": 0.15568028390407562,
+      "learning_rate": 2.1498371335504886e-05,
+      "loss": 1.1101,
+      "step": 1277
+    },
+    {
+      "epoch": 2.8072487644151565,
+      "grad_norm": 0.10911327600479126,
+      "learning_rate": 2.1254071661237786e-05,
+      "loss": 0.921,
+      "step": 1278
+    },
+    {
+      "epoch": 2.809445359692477,
+      "grad_norm": 0.10432939231395721,
+      "learning_rate": 2.1009771986970682e-05,
+      "loss": 0.8153,
+      "step": 1279
+    },
+    {
+      "epoch": 2.811641954969797,
+      "grad_norm": 0.10309730470180511,
+      "learning_rate": 2.076547231270358e-05,
+      "loss": 0.8418,
+      "step": 1280
+    },
+    {
+      "epoch": 2.813838550247117,
+      "grad_norm": 0.1340543031692505,
+      "learning_rate": 2.052117263843648e-05,
+      "loss": 0.9567,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8160351455244372,
+      "grad_norm": 0.19127598404884338,
+      "learning_rate": 2.027687296416938e-05,
+      "loss": 0.9237,
+      "step": 1282
+    },
+    {
+      "epoch": 2.818231740801757,
+      "grad_norm": 0.11550787836313248,
+      "learning_rate": 2.003257328990228e-05,
+      "loss": 1.0204,
+      "step": 1283
+    },
+    {
+      "epoch": 2.820428336079077,
+      "grad_norm": 0.08870989084243774,
+      "learning_rate": 1.978827361563518e-05,
+      "loss": 0.8918,
+      "step": 1284
+    },
+    {
+      "epoch": 2.8226249313563976,
+      "grad_norm": 0.09681003540754318,
+      "learning_rate": 1.9543973941368078e-05,
+      "loss": 0.8972,
+      "step": 1285
+    },
+    {
+      "epoch": 2.824821526633718,
+      "grad_norm": 0.11896215379238129,
+      "learning_rate": 1.9299674267100977e-05,
+      "loss": 1.1292,
+      "step": 1286
+    },
+    {
+      "epoch": 2.827018121911038,
+      "grad_norm": 0.11330459266901016,
+      "learning_rate": 1.9055374592833873e-05,
+      "loss": 0.8224,
+      "step": 1287
+    },
+    {
+      "epoch": 2.829214717188358,
+      "grad_norm": 0.12380295991897583,
+      "learning_rate": 1.8811074918566773e-05,
+      "loss": 0.9391,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8314113124656783,
+      "grad_norm": 0.13449206948280334,
+      "learning_rate": 1.8566775244299672e-05,
+      "loss": 0.8121,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8336079077429983,
+      "grad_norm": 0.11702273041009903,
+      "learning_rate": 1.832247557003257e-05,
+      "loss": 0.9502,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8358045030203183,
+      "grad_norm": 0.4879542887210846,
+      "learning_rate": 1.807817589576547e-05,
+      "loss": 0.7549,
+      "step": 1291
+    },
+    {
+      "epoch": 2.8380010982976387,
+      "grad_norm": 0.11182669550180435,
+      "learning_rate": 1.783387622149837e-05,
+      "loss": 0.7784,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8401976935749587,
+      "grad_norm": 0.11299195140600204,
+      "learning_rate": 1.758957654723127e-05,
+      "loss": 0.6569,
+      "step": 1293
+    },
+    {
+      "epoch": 2.842394288852279,
+      "grad_norm": 0.125802680850029,
+      "learning_rate": 1.734527687296417e-05,
+      "loss": 0.9292,
+      "step": 1294
+    },
+    {
+      "epoch": 2.844590884129599,
+      "grad_norm": 0.14828677475452423,
+      "learning_rate": 1.7100977198697065e-05,
+      "loss": 0.9814,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8467874794069195,
+      "grad_norm": 0.10490228235721588,
+      "learning_rate": 1.6856677524429964e-05,
+      "loss": 1.009,
+      "step": 1296
+    },
+    {
+      "epoch": 2.8489840746842394,
+      "grad_norm": 0.1204921156167984,
+      "learning_rate": 1.6612377850162864e-05,
+      "loss": 0.7553,
+      "step": 1297
+    },
+    {
+      "epoch": 2.8511806699615594,
+      "grad_norm": 0.19527584314346313,
+      "learning_rate": 1.6368078175895766e-05,
+      "loss": 1.1167,
+      "step": 1298
+    },
+    {
+      "epoch": 2.85337726523888,
+      "grad_norm": 0.14950314164161682,
+      "learning_rate": 1.6123778501628666e-05,
+      "loss": 0.8865,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8555738605161998,
+      "grad_norm": 0.2148180454969406,
+      "learning_rate": 1.587947882736156e-05,
+      "loss": 1.0818,
+      "step": 1300
+    },
+    {
+      "epoch": 2.85777045579352,
+      "grad_norm": 0.13899223506450653,
+      "learning_rate": 1.563517915309446e-05,
+      "loss": 0.8979,
+      "step": 1301
+    },
+    {
+      "epoch": 2.85996705107084,
+      "grad_norm": 0.12174992263317108,
+      "learning_rate": 1.539087947882736e-05,
+      "loss": 0.7538,
+      "step": 1302
+    },
+    {
+      "epoch": 2.8621636463481606,
+      "grad_norm": 0.10811881721019745,
+      "learning_rate": 1.514657980456026e-05,
+      "loss": 1.1239,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8643602416254805,
+      "grad_norm": 0.1307932436466217,
+      "learning_rate": 1.4902280130293159e-05,
+      "loss": 0.784,
+      "step": 1304
+    },
+    {
+      "epoch": 2.8665568369028005,
+      "grad_norm": 0.09690524637699127,
+      "learning_rate": 1.4657980456026057e-05,
+      "loss": 0.8151,
+      "step": 1305
+    },
+    {
+      "epoch": 2.868753432180121,
+      "grad_norm": 0.11124424636363983,
+      "learning_rate": 1.4413680781758956e-05,
+      "loss": 0.8364,
+      "step": 1306
+    },
+    {
+      "epoch": 2.870950027457441,
+      "grad_norm": 0.0880993977189064,
+      "learning_rate": 1.4169381107491856e-05,
+      "loss": 0.9133,
+      "step": 1307
+    },
+    {
+      "epoch": 2.873146622734761,
+      "grad_norm": 0.09988746792078018,
+      "learning_rate": 1.3925081433224755e-05,
+      "loss": 0.7658,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8753432180120813,
+      "grad_norm": 0.10843254625797272,
+      "learning_rate": 1.3680781758957653e-05,
+      "loss": 0.8759,
+      "step": 1309
+    },
+    {
+      "epoch": 2.8775398132894017,
+      "grad_norm": 0.11732069402933121,
+      "learning_rate": 1.3436482084690552e-05,
+      "loss": 0.9421,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8797364085667216,
+      "grad_norm": 0.11630874872207642,
+      "learning_rate": 1.3192182410423451e-05,
+      "loss": 0.9172,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8819330038440416,
+      "grad_norm": 0.11069071292877197,
+      "learning_rate": 1.294788273615635e-05,
+      "loss": 0.8679,
+      "step": 1312
+    },
+    {
+      "epoch": 2.884129599121362,
+      "grad_norm": 0.10334814339876175,
+      "learning_rate": 1.2703583061889248e-05,
+      "loss": 0.8601,
+      "step": 1313
+    },
+    {
+      "epoch": 2.886326194398682,
+      "grad_norm": 0.12045039981603622,
+      "learning_rate": 1.2459283387622148e-05,
+      "loss": 0.8755,
+      "step": 1314
+    },
+    {
+      "epoch": 2.888522789676002,
+      "grad_norm": 0.11761267483234406,
+      "learning_rate": 1.2214983713355047e-05,
+      "loss": 0.8335,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8907193849533224,
+      "grad_norm": 0.3814225494861603,
+      "learning_rate": 1.1970684039087946e-05,
+      "loss": 0.8499,
+      "step": 1316
+    },
+    {
+      "epoch": 2.892915980230643,
+      "grad_norm": 0.13088801503181458,
+      "learning_rate": 1.1726384364820844e-05,
+      "loss": 0.8017,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8951125755079627,
+      "grad_norm": 0.1916990727186203,
+      "learning_rate": 1.1482084690553745e-05,
+      "loss": 0.7181,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8973091707852827,
+      "grad_norm": 0.10485924035310745,
+      "learning_rate": 1.1237785016286644e-05,
+      "loss": 0.7044,
+      "step": 1319
+    },
+    {
+      "epoch": 2.899505766062603,
+      "grad_norm": 0.1017177477478981,
+      "learning_rate": 1.0993485342019544e-05,
+      "loss": 0.936,
+      "step": 1320
+    },
+    {
+      "epoch": 2.901702361339923,
+      "grad_norm": 0.13602672517299652,
+      "learning_rate": 1.0749185667752443e-05,
+      "loss": 0.7147,
+      "step": 1321
+    },
+    {
+      "epoch": 2.903898956617243,
+      "grad_norm": 0.17867814004421234,
+      "learning_rate": 1.0504885993485341e-05,
+      "loss": 0.9345,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9060955518945635,
+      "grad_norm": 0.13777326047420502,
+      "learning_rate": 1.026058631921824e-05,
+      "loss": 0.8578,
+      "step": 1323
+    },
+    {
+      "epoch": 2.9082921471718834,
+      "grad_norm": 0.13840548694133759,
+      "learning_rate": 1.001628664495114e-05,
+      "loss": 0.8765,
+      "step": 1324
+    },
+    {
+      "epoch": 2.910488742449204,
+      "grad_norm": 0.10178028792142868,
+      "learning_rate": 9.771986970684039e-06,
+      "loss": 0.7602,
+      "step": 1325
+    },
+    {
+      "epoch": 2.912685337726524,
+      "grad_norm": 0.10320001095533371,
+      "learning_rate": 9.527687296416937e-06,
+      "loss": 0.8766,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9148819330038442,
+      "grad_norm": 0.18920600414276123,
+      "learning_rate": 9.283387622149836e-06,
+      "loss": 0.9352,
+      "step": 1327
+    },
+    {
+      "epoch": 2.917078528281164,
+      "grad_norm": 0.13869708776474,
+      "learning_rate": 9.039087947882735e-06,
+      "loss": 0.7025,
+      "step": 1328
+    },
+    {
+      "epoch": 2.919275123558484,
+      "grad_norm": 0.11738063395023346,
+      "learning_rate": 8.794788273615635e-06,
+      "loss": 0.6368,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9214717188358046,
+      "grad_norm": 0.1826699674129486,
+      "learning_rate": 8.550488599348532e-06,
+      "loss": 0.9034,
+      "step": 1330
+    },
+    {
+      "epoch": 2.9236683141131246,
+      "grad_norm": 0.10016272217035294,
+      "learning_rate": 8.306188925081432e-06,
+      "loss": 0.9104,
+      "step": 1331
+    },
+    {
+      "epoch": 2.925864909390445,
+      "grad_norm": 0.14078868925571442,
+      "learning_rate": 8.061889250814333e-06,
+      "loss": 0.5473,
+      "step": 1332
+    },
+    {
+      "epoch": 2.928061504667765,
+      "grad_norm": 0.1499549299478531,
+      "learning_rate": 7.81758957654723e-06,
+      "loss": 0.9098,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9302580999450853,
+      "grad_norm": 0.1144939512014389,
+      "learning_rate": 7.57328990228013e-06,
+      "loss": 0.7585,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9324546952224053,
+      "grad_norm": 0.12452825158834457,
+      "learning_rate": 7.328990228013028e-06,
+      "loss": 0.7524,
+      "step": 1335
+    },
+    {
+      "epoch": 2.9346512904997253,
+      "grad_norm": 0.14157415926456451,
+      "learning_rate": 7.084690553745928e-06,
+      "loss": 0.8206,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9368478857770457,
+      "grad_norm": 0.10405463725328445,
+      "learning_rate": 6.840390879478826e-06,
+      "loss": 0.7443,
+      "step": 1337
+    },
+    {
+      "epoch": 2.9390444810543657,
+      "grad_norm": 0.09900306165218353,
+      "learning_rate": 6.596091205211726e-06,
+      "loss": 0.8146,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9412410763316856,
+      "grad_norm": 0.10284154862165451,
+      "learning_rate": 6.351791530944624e-06,
+      "loss": 0.8396,
+      "step": 1339
+    },
+    {
+      "epoch": 2.943437671609006,
+      "grad_norm": 0.13008739054203033,
+      "learning_rate": 6.1074918566775235e-06,
+      "loss": 0.6294,
+      "step": 1340
+    },
+    {
+      "epoch": 2.9456342668863265,
+      "grad_norm": 0.11180314421653748,
+      "learning_rate": 5.863192182410422e-06,
+      "loss": 0.776,
+      "step": 1341
+    },
+    {
+      "epoch": 2.9478308621636464,
+      "grad_norm": 0.1007196307182312,
+      "learning_rate": 5.618892508143322e-06,
+      "loss": 0.8754,
+      "step": 1342
+    },
+    {
+      "epoch": 2.9500274574409664,
+      "grad_norm": 0.11838221549987793,
+      "learning_rate": 5.374592833876222e-06,
+      "loss": 0.9839,
+      "step": 1343
+    },
+    {
+      "epoch": 2.952224052718287,
+      "grad_norm": 0.13677622377872467,
+      "learning_rate": 5.13029315960912e-06,
+      "loss": 1.0426,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9544206479956068,
+      "grad_norm": 0.0994543507695198,
+      "learning_rate": 4.8859934853420195e-06,
+      "loss": 0.9163,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9566172432729267,
+      "grad_norm": 0.1238836720585823,
+      "learning_rate": 4.641693811074918e-06,
+      "loss": 0.7843,
+      "step": 1346
+    },
+    {
+      "epoch": 2.958813838550247,
+      "grad_norm": 0.09787926822900772,
+      "learning_rate": 4.397394136807817e-06,
+      "loss": 0.7792,
+      "step": 1347
+    },
+    {
+      "epoch": 2.961010433827567,
+      "grad_norm": 0.11010874062776566,
+      "learning_rate": 4.153094462540716e-06,
+      "loss": 0.7628,
+      "step": 1348
+    },
+    {
+      "epoch": 2.9632070291048875,
+      "grad_norm": 0.09954501688480377,
+      "learning_rate": 3.908794788273615e-06,
+      "loss": 1.0633,
+      "step": 1349
+    },
+    {
+      "epoch": 2.9654036243822075,
+      "grad_norm": 1.534223198890686,
+      "learning_rate": 3.664495114006514e-06,
+      "loss": 0.83,
+      "step": 1350
+    },
+    {
+      "epoch": 2.967600219659528,
+      "grad_norm": 0.23465970158576965,
+      "learning_rate": 3.420195439739413e-06,
+      "loss": 0.9729,
+      "step": 1351
+    },
+    {
+      "epoch": 2.969796814936848,
+      "grad_norm": 0.11462447047233582,
+      "learning_rate": 3.175895765472312e-06,
+      "loss": 0.8308,
+      "step": 1352
+    },
+    {
+      "epoch": 2.971993410214168,
+      "grad_norm": 0.10270516574382782,
+      "learning_rate": 2.931596091205211e-06,
+      "loss": 0.8947,
+      "step": 1353
+    },
+    {
+      "epoch": 2.9741900054914883,
+      "grad_norm": 0.11180620640516281,
+      "learning_rate": 2.687296416938111e-06,
+      "loss": 1.014,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9763866007688082,
+      "grad_norm": 0.11388438940048218,
+      "learning_rate": 2.4429967426710097e-06,
+      "loss": 0.7252,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9785831960461286,
+      "grad_norm": 0.17407070100307465,
+      "learning_rate": 2.1986970684039087e-06,
+      "loss": 0.9446,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9807797913234486,
+      "grad_norm": 0.20987039804458618,
+      "learning_rate": 1.9543973941368076e-06,
+      "loss": 0.9541,
+      "step": 1357
+    },
+    {
+      "epoch": 2.982976386600769,
+      "grad_norm": 0.1313517689704895,
+      "learning_rate": 1.7100977198697066e-06,
+      "loss": 0.9657,
+      "step": 1358
+    },
+    {
+      "epoch": 2.985172981878089,
+      "grad_norm": 0.1280854493379593,
+      "learning_rate": 1.4657980456026055e-06,
+      "loss": 1.0016,
+      "step": 1359
+    },
+    {
+      "epoch": 2.987369577155409,
+      "grad_norm": 0.11837625503540039,
+      "learning_rate": 1.2214983713355049e-06,
+      "loss": 0.7518,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9895661724327294,
+      "grad_norm": 0.08296261727809906,
+      "learning_rate": 9.771986970684038e-07,
+      "loss": 0.8744,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9917627677100493,
+      "grad_norm": 0.2100769281387329,
+      "learning_rate": 7.328990228013028e-07,
+      "loss": 1.1194,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9939593629873693,
+      "grad_norm": 0.12164104729890823,
+      "learning_rate": 4.885993485342019e-07,
+      "loss": 0.8977,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9961559582646897,
+      "grad_norm": 0.1288069188594818,
+      "learning_rate": 2.4429967426710095e-07,
+      "loss": 0.9247,
+      "step": 1364
+    },
+    {
+      "epoch": 2.99835255354201,
+      "grad_norm": 0.10124250501394272,
+      "learning_rate": 0.0,
+      "loss": 0.9201,
+      "step": 1365
+    },
+    {
+      "epoch": 2.99835255354201,
+      "step": 1365,
+      "total_flos": 4.5989900901679104e+17,
+      "train_loss": 0.9395189625439626,
+      "train_runtime": 3703.1427,
+      "train_samples_per_second": 5.9,
+      "train_steps_per_second": 0.369
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1365,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.5989900901679104e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}