diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12727 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 18123,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016553550736633008,
+      "grad_norm": 7.984700679779053,
+      "learning_rate": 4.964147821290679e-08,
+      "loss": 1.0634,
+      "step": 10
+    },
+    {
+      "epoch": 0.0033107101473266016,
+      "grad_norm": 6.1845703125,
+      "learning_rate": 1.0479867622724768e-07,
+      "loss": 1.1529,
+      "step": 20
+    },
+    {
+      "epoch": 0.004966065220989902,
+      "grad_norm": 7.543320655822754,
+      "learning_rate": 1.5995587424158854e-07,
+      "loss": 1.2636,
+      "step": 30
+    },
+    {
+      "epoch": 0.006621420294653203,
+      "grad_norm": 6.893922328948975,
+      "learning_rate": 2.1511307225592942e-07,
+      "loss": 1.2297,
+      "step": 40
+    },
+    {
+      "epoch": 0.008276775368316504,
+      "grad_norm": 7.185118675231934,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 1.1841,
+      "step": 50
+    },
+    {
+      "epoch": 0.009932130441979804,
+      "grad_norm": 8.388279914855957,
+      "learning_rate": 3.2542746828461117e-07,
+      "loss": 1.1145,
+      "step": 60
+    },
+    {
+      "epoch": 0.011587485515643106,
+      "grad_norm": 6.8379292488098145,
+      "learning_rate": 3.805846662989521e-07,
+      "loss": 1.1385,
+      "step": 70
+    },
+    {
+      "epoch": 0.013242840589306406,
+      "grad_norm": 8.465346336364746,
+      "learning_rate": 4.357418643132929e-07,
+      "loss": 1.1756,
+      "step": 80
+    },
+    {
+      "epoch": 0.014898195662969706,
+      "grad_norm": 6.7175493240356445,
+      "learning_rate": 4.908990623276337e-07,
+      "loss": 1.029,
+      "step": 90
+    },
+    {
+      "epoch": 0.01655355073663301,
+      "grad_norm": 4.6793212890625,
+      "learning_rate": 5.460562603419746e-07,
+      "loss": 1.1213,
+      "step": 100
+    },
+    {
+      "epoch": 0.01820890581029631,
+      "grad_norm": 4.11850643157959,
+      "learning_rate": 6.012134583563156e-07,
+      "loss": 1.0167,
+      "step": 110
+    },
+    {
+      "epoch": 0.01986426088395961,
+      "grad_norm": 6.64889669418335,
+      "learning_rate": 6.563706563706564e-07,
+      "loss": 1.0059,
+      "step": 120
+    },
+    {
+      "epoch": 0.02151961595762291,
+      "grad_norm": 3.417647123336792,
+      "learning_rate": 7.115278543849972e-07,
+      "loss": 1.0292,
+      "step": 130
+    },
+    {
+      "epoch": 0.023174971031286212,
+      "grad_norm": 3.204617738723755,
+      "learning_rate": 7.666850523993381e-07,
+      "loss": 0.9241,
+      "step": 140
+    },
+    {
+      "epoch": 0.02483032610494951,
+      "grad_norm": 5.174026966094971,
+      "learning_rate": 8.21842250413679e-07,
+      "loss": 0.9826,
+      "step": 150
+    },
+    {
+      "epoch": 0.026485681178612813,
+      "grad_norm": 5.526364803314209,
+      "learning_rate": 8.769994484280199e-07,
+      "loss": 0.9245,
+      "step": 160
+    },
+    {
+      "epoch": 0.028141036252276114,
+      "grad_norm": 5.8991570472717285,
+      "learning_rate": 9.321566464423607e-07,
+      "loss": 0.9536,
+      "step": 170
+    },
+    {
+      "epoch": 0.029796391325939413,
+      "grad_norm": 4.1310248374938965,
+      "learning_rate": 9.873138444567016e-07,
+      "loss": 0.9059,
+      "step": 180
+    },
+    {
+      "epoch": 0.031451746399602715,
+      "grad_norm": 4.935130596160889,
+      "learning_rate": 1.0424710424710426e-06,
+      "loss": 0.9174,
+      "step": 190
+    },
+    {
+      "epoch": 0.03310710147326602,
+      "grad_norm": 3.474487066268921,
+      "learning_rate": 1.0976282404853834e-06,
+      "loss": 0.9133,
+      "step": 200
+    },
+    {
+      "epoch": 0.03476245654692932,
+      "grad_norm": 10.657366752624512,
+      "learning_rate": 1.1527854384997244e-06,
+      "loss": 0.7568,
+      "step": 210
+    },
+    {
+      "epoch": 0.03641781162059262,
+      "grad_norm": 3.7891452312469482,
+      "learning_rate": 1.2079426365140651e-06,
+      "loss": 0.8258,
+      "step": 220
+    },
+    {
+      "epoch": 0.038073166694255915,
+      "grad_norm": 5.292945861816406,
+      "learning_rate": 1.263099834528406e-06,
+      "loss": 0.784,
+      "step": 230
+    },
+    {
+      "epoch": 0.03972852176791922,
+      "grad_norm": 4.9213690757751465,
+      "learning_rate": 1.3182570325427469e-06,
+      "loss": 0.7955,
+      "step": 240
+    },
+    {
+      "epoch": 0.04138387684158252,
+      "grad_norm": 4.905694484710693,
+      "learning_rate": 1.3734142305570876e-06,
+      "loss": 0.877,
+      "step": 250
+    },
+    {
+      "epoch": 0.04303923191524582,
+      "grad_norm": 3.1677496433258057,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.8612,
+      "step": 260
+    },
+    {
+      "epoch": 0.04469458698890912,
+      "grad_norm": 3.0269603729248047,
+      "learning_rate": 1.4837286265857694e-06,
+      "loss": 0.821,
+      "step": 270
+    },
+    {
+      "epoch": 0.046349942062572425,
+      "grad_norm": 3.246804714202881,
+      "learning_rate": 1.5388858246001104e-06,
+      "loss": 0.8345,
+      "step": 280
+    },
+    {
+      "epoch": 0.04800529713623572,
+      "grad_norm": 3.3569605350494385,
+      "learning_rate": 1.5940430226144514e-06,
+      "loss": 0.7517,
+      "step": 290
+    },
+    {
+      "epoch": 0.04966065220989902,
+      "grad_norm": 3.597111701965332,
+      "learning_rate": 1.6492002206287921e-06,
+      "loss": 0.8384,
+      "step": 300
+    },
+    {
+      "epoch": 0.05131600728356232,
+      "grad_norm": 3.895134925842285,
+      "learning_rate": 1.7043574186431331e-06,
+      "loss": 0.8339,
+      "step": 310
+    },
+    {
+      "epoch": 0.052971362357225625,
+      "grad_norm": 4.427375316619873,
+      "learning_rate": 1.7595146166574739e-06,
+      "loss": 0.784,
+      "step": 320
+    },
+    {
+      "epoch": 0.05462671743088893,
+      "grad_norm": 3.222888708114624,
+      "learning_rate": 1.8146718146718149e-06,
+      "loss": 0.7882,
+      "step": 330
+    },
+    {
+      "epoch": 0.05628207250455223,
+      "grad_norm": 3.1282763481140137,
+      "learning_rate": 1.8698290126861556e-06,
+      "loss": 0.7937,
+      "step": 340
+    },
+    {
+      "epoch": 0.05793742757821553,
+      "grad_norm": 4.77147102355957,
+      "learning_rate": 1.9249862107004966e-06,
+      "loss": 0.7097,
+      "step": 350
+    },
+    {
+      "epoch": 0.059592782651878826,
+      "grad_norm": 4.3493571281433105,
+      "learning_rate": 1.9801434087148376e-06,
+      "loss": 0.6705,
+      "step": 360
+    },
+    {
+      "epoch": 0.06124813772554213,
+      "grad_norm": 3.3635661602020264,
+      "learning_rate": 2.035300606729178e-06,
+      "loss": 0.8396,
+      "step": 370
+    },
+    {
+      "epoch": 0.06290349279920543,
+      "grad_norm": 4.846203804016113,
+      "learning_rate": 2.090457804743519e-06,
+      "loss": 0.7266,
+      "step": 380
+    },
+    {
+      "epoch": 0.06455884787286872,
+      "grad_norm": 4.919671058654785,
+      "learning_rate": 2.14561500275786e-06,
+      "loss": 0.6616,
+      "step": 390
+    },
+    {
+      "epoch": 0.06621420294653203,
+      "grad_norm": 3.959589958190918,
+      "learning_rate": 2.200772200772201e-06,
+      "loss": 0.7379,
+      "step": 400
+    },
+    {
+      "epoch": 0.06786955802019533,
+      "grad_norm": 4.111237049102783,
+      "learning_rate": 2.255929398786542e-06,
+      "loss": 0.7201,
+      "step": 410
+    },
+    {
+      "epoch": 0.06952491309385864,
+      "grad_norm": 4.0857343673706055,
+      "learning_rate": 2.3110865968008826e-06,
+      "loss": 0.7806,
+      "step": 420
+    },
+    {
+      "epoch": 0.07118026816752193,
+      "grad_norm": 2.696610689163208,
+      "learning_rate": 2.3662437948152236e-06,
+      "loss": 0.7176,
+      "step": 430
+    },
+    {
+      "epoch": 0.07283562324118524,
+      "grad_norm": 3.1696786880493164,
+      "learning_rate": 2.4214009928295646e-06,
+      "loss": 0.7062,
+      "step": 440
+    },
+    {
+      "epoch": 0.07449097831484854,
+      "grad_norm": 4.226884365081787,
+      "learning_rate": 2.4765581908439056e-06,
+      "loss": 0.7227,
+      "step": 450
+    },
+    {
+      "epoch": 0.07614633338851183,
+      "grad_norm": 2.852306842803955,
+      "learning_rate": 2.531715388858246e-06,
+      "loss": 0.7354,
+      "step": 460
+    },
+    {
+      "epoch": 0.07780168846217514,
+      "grad_norm": 3.600233316421509,
+      "learning_rate": 2.5868725868725867e-06,
+      "loss": 0.7057,
+      "step": 470
+    },
+    {
+      "epoch": 0.07945704353583843,
+      "grad_norm": 6.20136833190918,
+      "learning_rate": 2.642029784886928e-06,
+      "loss": 0.713,
+      "step": 480
+    },
+    {
+      "epoch": 0.08111239860950174,
+      "grad_norm": 5.960463047027588,
+      "learning_rate": 2.6971869829012687e-06,
+      "loss": 0.7262,
+      "step": 490
+    },
+    {
+      "epoch": 0.08276775368316504,
+      "grad_norm": 2.7353596687316895,
+      "learning_rate": 2.7523441809156096e-06,
+      "loss": 0.7249,
+      "step": 500
+    },
+    {
+      "epoch": 0.08442310875682833,
+      "grad_norm": 2.9528822898864746,
+      "learning_rate": 2.8075013789299506e-06,
+      "loss": 0.7158,
+      "step": 510
+    },
+    {
+      "epoch": 0.08607846383049164,
+      "grad_norm": 3.323730230331421,
+      "learning_rate": 2.8626585769442916e-06,
+      "loss": 0.704,
+      "step": 520
+    },
+    {
+      "epoch": 0.08773381890415494,
+      "grad_norm": 3.2658655643463135,
+      "learning_rate": 2.917815774958632e-06,
+      "loss": 0.7194,
+      "step": 530
+    },
+    {
+      "epoch": 0.08938917397781825,
+      "grad_norm": 3.035658121109009,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.6506,
+      "step": 540
+    },
+    {
+      "epoch": 0.09104452905148154,
+      "grad_norm": 3.595717668533325,
+      "learning_rate": 3.028130170987314e-06,
+      "loss": 0.7421,
+      "step": 550
+    },
+    {
+      "epoch": 0.09269988412514485,
+      "grad_norm": 4.5476202964782715,
+      "learning_rate": 3.083287369001655e-06,
+      "loss": 0.704,
+      "step": 560
+    },
+    {
+      "epoch": 0.09435523919880814,
+      "grad_norm": 3.4544177055358887,
+      "learning_rate": 3.1384445670159957e-06,
+      "loss": 0.6521,
+      "step": 570
+    },
+    {
+      "epoch": 0.09601059427247144,
+      "grad_norm": 3.8915786743164062,
+      "learning_rate": 3.193601765030337e-06,
+      "loss": 0.6756,
+      "step": 580
+    },
+    {
+      "epoch": 0.09766594934613475,
+      "grad_norm": 4.146472930908203,
+      "learning_rate": 3.2487589630446776e-06,
+      "loss": 0.7626,
+      "step": 590
+    },
+    {
+      "epoch": 0.09932130441979804,
+      "grad_norm": 3.272199869155884,
+      "learning_rate": 3.3039161610590186e-06,
+      "loss": 0.7363,
+      "step": 600
+    },
+    {
+      "epoch": 0.10097665949346135,
+      "grad_norm": 2.5908915996551514,
+      "learning_rate": 3.359073359073359e-06,
+      "loss": 0.649,
+      "step": 610
+    },
+    {
+      "epoch": 0.10263201456712465,
+      "grad_norm": 3.507157325744629,
+      "learning_rate": 3.4142305570877006e-06,
+      "loss": 0.7292,
+      "step": 620
+    },
+    {
+      "epoch": 0.10428736964078796,
+      "grad_norm": 3.2821595668792725,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 0.7024,
+      "step": 630
+    },
+    {
+      "epoch": 0.10594272471445125,
+      "grad_norm": 3.930805206298828,
+      "learning_rate": 3.524544953116382e-06,
+      "loss": 0.7038,
+      "step": 640
+    },
+    {
+      "epoch": 0.10759807978811455,
+      "grad_norm": 2.831280469894409,
+      "learning_rate": 3.5797021511307227e-06,
+      "loss": 0.6795,
+      "step": 650
+    },
+    {
+      "epoch": 0.10925343486177785,
+      "grad_norm": 2.831242799758911,
+      "learning_rate": 3.634859349145064e-06,
+      "loss": 0.6852,
+      "step": 660
+    },
+    {
+      "epoch": 0.11090878993544115,
+      "grad_norm": 2.9769375324249268,
+      "learning_rate": 3.6900165471594046e-06,
+      "loss": 0.6547,
+      "step": 670
+    },
+    {
+      "epoch": 0.11256414500910446,
+      "grad_norm": 2.967987060546875,
+      "learning_rate": 3.745173745173745e-06,
+      "loss": 0.7238,
+      "step": 680
+    },
+    {
+      "epoch": 0.11421950008276775,
+      "grad_norm": 3.29638671875,
+      "learning_rate": 3.800330943188086e-06,
+      "loss": 0.6848,
+      "step": 690
+    },
+    {
+      "epoch": 0.11587485515643106,
+      "grad_norm": 4.40482759475708,
+      "learning_rate": 3.855488141202427e-06,
+      "loss": 0.673,
+      "step": 700
+    },
+    {
+      "epoch": 0.11753021023009436,
+      "grad_norm": 3.206632375717163,
+      "learning_rate": 3.910645339216768e-06,
+      "loss": 0.6148,
+      "step": 710
+    },
+    {
+      "epoch": 0.11918556530375765,
+      "grad_norm": 3.0711138248443604,
+      "learning_rate": 3.965802537231109e-06,
+      "loss": 0.6717,
+      "step": 720
+    },
+    {
+      "epoch": 0.12084092037742096,
+      "grad_norm": 3.105243444442749,
+      "learning_rate": 4.02095973524545e-06,
+      "loss": 0.5857,
+      "step": 730
+    },
+    {
+      "epoch": 0.12249627545108425,
+      "grad_norm": 4.079644680023193,
+      "learning_rate": 4.07611693325979e-06,
+      "loss": 0.6482,
+      "step": 740
+    },
+    {
+      "epoch": 0.12415163052474756,
+      "grad_norm": 3.890226364135742,
+      "learning_rate": 4.131274131274132e-06,
+      "loss": 0.5871,
+      "step": 750
+    },
+    {
+      "epoch": 0.12580698559841086,
+      "grad_norm": 2.5968997478485107,
+      "learning_rate": 4.186431329288472e-06,
+      "loss": 0.6882,
+      "step": 760
+    },
+    {
+      "epoch": 0.12746234067207415,
+      "grad_norm": 2.7869668006896973,
+      "learning_rate": 4.241588527302813e-06,
+      "loss": 0.6667,
+      "step": 770
+    },
+    {
+      "epoch": 0.12911769574573745,
+      "grad_norm": 3.2500760555267334,
+      "learning_rate": 4.296745725317154e-06,
+      "loss": 0.6381,
+      "step": 780
+    },
+    {
+      "epoch": 0.13077305081940077,
+      "grad_norm": 3.1198246479034424,
+      "learning_rate": 4.351902923331495e-06,
+      "loss": 0.6226,
+      "step": 790
+    },
+    {
+      "epoch": 0.13242840589306407,
+      "grad_norm": 3.8074584007263184,
+      "learning_rate": 4.407060121345836e-06,
+      "loss": 0.6221,
+      "step": 800
+    },
+    {
+      "epoch": 0.13408376096672736,
+      "grad_norm": 2.311058521270752,
+      "learning_rate": 4.462217319360177e-06,
+      "loss": 0.6789,
+      "step": 810
+    },
+    {
+      "epoch": 0.13573911604039066,
+      "grad_norm": 2.830004930496216,
+      "learning_rate": 4.517374517374517e-06,
+      "loss": 0.6512,
+      "step": 820
+    },
+    {
+      "epoch": 0.13739447111405395,
+      "grad_norm": 3.7663865089416504,
+      "learning_rate": 4.572531715388859e-06,
+      "loss": 0.7057,
+      "step": 830
+    },
+    {
+      "epoch": 0.13904982618771727,
+      "grad_norm": 2.8656136989593506,
+      "learning_rate": 4.627688913403199e-06,
+      "loss": 0.6411,
+      "step": 840
+    },
+    {
+      "epoch": 0.14070518126138057,
+      "grad_norm": 3.1035988330841064,
+      "learning_rate": 4.68284611141754e-06,
+      "loss": 0.6221,
+      "step": 850
+    },
+    {
+      "epoch": 0.14236053633504386,
+      "grad_norm": 3.384016990661621,
+      "learning_rate": 4.738003309431881e-06,
+      "loss": 0.6321,
+      "step": 860
+    },
+    {
+      "epoch": 0.14401589140870716,
+      "grad_norm": 3.3458902835845947,
+      "learning_rate": 4.793160507446222e-06,
+      "loss": 0.6163,
+      "step": 870
+    },
+    {
+      "epoch": 0.14567124648237048,
+      "grad_norm": 3.6326417922973633,
+      "learning_rate": 4.848317705460563e-06,
+      "loss": 0.6228,
+      "step": 880
+    },
+    {
+      "epoch": 0.14732660155603378,
+      "grad_norm": 3.407176971435547,
+      "learning_rate": 4.903474903474904e-06,
+      "loss": 0.675,
+      "step": 890
+    },
+    {
+      "epoch": 0.14898195662969707,
+      "grad_norm": 2.779155731201172,
+      "learning_rate": 4.958632101489245e-06,
+      "loss": 0.6416,
+      "step": 900
+    },
+    {
+      "epoch": 0.15063731170336037,
+      "grad_norm": 3.1110782623291016,
+      "learning_rate": 5.013789299503585e-06,
+      "loss": 0.5804,
+      "step": 910
+    },
+    {
+      "epoch": 0.15229266677702366,
+      "grad_norm": 3.0117592811584473,
+      "learning_rate": 5.068946497517927e-06,
+      "loss": 0.7233,
+      "step": 920
+    },
+    {
+      "epoch": 0.15394802185068698,
+      "grad_norm": 3.690901756286621,
+      "learning_rate": 5.124103695532267e-06,
+      "loss": 0.6569,
+      "step": 930
+    },
+    {
+      "epoch": 0.15560337692435028,
+      "grad_norm": 3.155195713043213,
+      "learning_rate": 5.179260893546608e-06,
+      "loss": 0.6537,
+      "step": 940
+    },
+    {
+      "epoch": 0.15725873199801357,
+      "grad_norm": 3.386868715286255,
+      "learning_rate": 5.234418091560949e-06,
+      "loss": 0.6306,
+      "step": 950
+    },
+    {
+      "epoch": 0.15891408707167687,
+      "grad_norm": 4.336483001708984,
+      "learning_rate": 5.28957528957529e-06,
+      "loss": 0.6967,
+      "step": 960
+    },
+    {
+      "epoch": 0.16056944214534016,
+      "grad_norm": 3.386204481124878,
+      "learning_rate": 5.344732487589631e-06,
+      "loss": 0.6151,
+      "step": 970
+    },
+    {
+      "epoch": 0.16222479721900349,
+      "grad_norm": 3.8770434856414795,
+      "learning_rate": 5.399889685603972e-06,
+      "loss": 0.6667,
+      "step": 980
+    },
+    {
+      "epoch": 0.16388015229266678,
+      "grad_norm": 3.62699294090271,
+      "learning_rate": 5.455046883618312e-06,
+      "loss": 0.6323,
+      "step": 990
+    },
+    {
+      "epoch": 0.16553550736633008,
+      "grad_norm": 2.923297882080078,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 0.6294,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16719086243999337,
+      "grad_norm": 3.1828577518463135,
+      "learning_rate": 5.565361279646995e-06,
+      "loss": 0.6644,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16884621751365667,
+      "grad_norm": 2.8371975421905518,
+      "learning_rate": 5.620518477661335e-06,
+      "loss": 0.6294,
+      "step": 1020
+    },
+    {
+      "epoch": 0.17050157258732,
+      "grad_norm": 2.7191975116729736,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.7093,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17215692766098328,
+      "grad_norm": 2.4802534580230713,
+      "learning_rate": 5.730832873690016e-06,
+      "loss": 0.5608,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17381228273464658,
+      "grad_norm": 5.264163017272949,
+      "learning_rate": 5.785990071704358e-06,
+      "loss": 0.6454,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17546763780830987,
+      "grad_norm": 2.8554723262786865,
+      "learning_rate": 5.841147269718699e-06,
+      "loss": 0.6278,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1771229928819732,
+      "grad_norm": 3.11765718460083,
+      "learning_rate": 5.896304467733039e-06,
+      "loss": 0.6203,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1787783479556365,
+      "grad_norm": 2.88675594329834,
+      "learning_rate": 5.95146166574738e-06,
+      "loss": 0.6984,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18043370302929979,
+      "grad_norm": 2.93239688873291,
+      "learning_rate": 6.006618863761722e-06,
+      "loss": 0.6054,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18208905810296308,
+      "grad_norm": 3.4710001945495605,
+      "learning_rate": 6.061776061776062e-06,
+      "loss": 0.6106,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18374441317662638,
+      "grad_norm": 2.3426156044006348,
+      "learning_rate": 6.116933259790403e-06,
+      "loss": 0.6256,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1853997682502897,
+      "grad_norm": 5.0461602210998535,
+      "learning_rate": 6.172090457804743e-06,
+      "loss": 0.6798,
+      "step": 1120
+    },
+    {
+      "epoch": 0.187055123323953,
+      "grad_norm": 2.646273374557495,
+      "learning_rate": 6.227247655819085e-06,
+      "loss": 0.5688,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1887104783976163,
+      "grad_norm": 3.764864444732666,
+      "learning_rate": 6.282404853833426e-06,
+      "loss": 0.5802,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19036583347127958,
+      "grad_norm": 2.7304487228393555,
+      "learning_rate": 6.337562051847766e-06,
+      "loss": 0.6034,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19202118854494288,
+      "grad_norm": 2.973935127258301,
+      "learning_rate": 6.392719249862107e-06,
+      "loss": 0.6278,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1936765436186062,
+      "grad_norm": 2.750199794769287,
+      "learning_rate": 6.447876447876449e-06,
+      "loss": 0.6995,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1953318986922695,
+      "grad_norm": 2.852870225906372,
+      "learning_rate": 6.503033645890789e-06,
+      "loss": 0.6436,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1969872537659328,
+      "grad_norm": 3.1429359912872314,
+      "learning_rate": 6.55819084390513e-06,
+      "loss": 0.6216,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19864260883959609,
+      "grad_norm": 2.8767406940460205,
+      "learning_rate": 6.61334804191947e-06,
+      "loss": 0.649,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20029796391325938,
+      "grad_norm": 4.3968281745910645,
+      "learning_rate": 6.668505239933812e-06,
+      "loss": 0.626,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2019533189869227,
+      "grad_norm": 2.600132465362549,
+      "learning_rate": 6.723662437948153e-06,
+      "loss": 0.6757,
+      "step": 1220
+    },
+    {
+      "epoch": 0.203608674060586,
+      "grad_norm": 3.5382864475250244,
+      "learning_rate": 6.778819635962493e-06,
+      "loss": 0.6118,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2052640291342493,
+      "grad_norm": 2.8111588954925537,
+      "learning_rate": 6.833976833976834e-06,
+      "loss": 0.58,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2069193842079126,
+      "grad_norm": 3.0125858783721924,
+      "learning_rate": 6.889134031991176e-06,
+      "loss": 0.6049,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2085747392815759,
+      "grad_norm": 3.686331272125244,
+      "learning_rate": 6.944291230005516e-06,
+      "loss": 0.5898,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2102300943552392,
+      "grad_norm": 4.72364616394043,
+      "learning_rate": 6.999448428019857e-06,
+      "loss": 0.5901,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2118854494289025,
+      "grad_norm": 2.531151294708252,
+      "learning_rate": 7.054605626034198e-06,
+      "loss": 0.6588,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2135408045025658,
+      "grad_norm": 3.2532570362091064,
+      "learning_rate": 7.109762824048539e-06,
+      "loss": 0.5635,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2151961595762291,
+      "grad_norm": 3.614664316177368,
+      "learning_rate": 7.16492002206288e-06,
+      "loss": 0.5676,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2168515146498924,
+      "grad_norm": 4.736164569854736,
+      "learning_rate": 7.22007722007722e-06,
+      "loss": 0.6029,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2185068697235557,
+      "grad_norm": 3.391282320022583,
+      "learning_rate": 7.275234418091561e-06,
+      "loss": 0.6341,
+      "step": 1320
+    },
+    {
+      "epoch": 0.220162224797219,
+      "grad_norm": 3.4282443523406982,
+      "learning_rate": 7.330391616105902e-06,
+      "loss": 0.6301,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2218175798708823,
+      "grad_norm": 2.516094923019409,
+      "learning_rate": 7.385548814120243e-06,
+      "loss": 0.616,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2234729349445456,
+      "grad_norm": 2.89660906791687,
+      "learning_rate": 7.440706012134584e-06,
+      "loss": 0.6081,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22512829001820892,
+      "grad_norm": 2.6692073345184326,
+      "learning_rate": 7.495863210148925e-06,
+      "loss": 0.5554,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2267836450918722,
+      "grad_norm": 2.7826545238494873,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 0.6513,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2284390001655355,
+      "grad_norm": 4.580424785614014,
+      "learning_rate": 7.606177606177607e-06,
+      "loss": 0.6532,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2300943552391988,
+      "grad_norm": 2.9366021156311035,
+      "learning_rate": 7.661334804191947e-06,
+      "loss": 0.6284,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23174971031286212,
+      "grad_norm": 2.872647762298584,
+      "learning_rate": 7.716492002206288e-06,
+      "loss": 0.65,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23340506538652542,
+      "grad_norm": 2.3038113117218018,
+      "learning_rate": 7.77164920022063e-06,
+      "loss": 0.6274,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2350604204601887,
+      "grad_norm": 2.807044744491577,
+      "learning_rate": 7.82680639823497e-06,
+      "loss": 0.5892,
+      "step": 1420
+    },
+    {
+      "epoch": 0.236715775533852,
+      "grad_norm": 2.3904168605804443,
+      "learning_rate": 7.881963596249311e-06,
+      "loss": 0.6505,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2383711306075153,
+      "grad_norm": 2.934279441833496,
+      "learning_rate": 7.937120794263652e-06,
+      "loss": 0.5834,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24002648568117863,
+      "grad_norm": 2.8970916271209717,
+      "learning_rate": 7.992277992277993e-06,
+      "loss": 0.6136,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24168184075484192,
+      "grad_norm": 2.5647435188293457,
+      "learning_rate": 8.047435190292334e-06,
+      "loss": 0.6831,
+      "step": 1460
+    },
+    {
+      "epoch": 0.24333719582850522,
+      "grad_norm": 3.708890438079834,
+      "learning_rate": 8.102592388306675e-06,
+      "loss": 0.6018,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2449925509021685,
+      "grad_norm": 3.2741858959198,
+      "learning_rate": 8.157749586321016e-06,
+      "loss": 0.5826,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2466479059758318,
+      "grad_norm": 2.7993056774139404,
+      "learning_rate": 8.212906784335355e-06,
+      "loss": 0.6209,
+      "step": 1490
+    },
+    {
+      "epoch": 0.24830326104949513,
+      "grad_norm": 2.5464768409729004,
+      "learning_rate": 8.268063982349698e-06,
+      "loss": 0.6195,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24995861612315842,
+      "grad_norm": 2.438197612762451,
+      "learning_rate": 8.323221180364039e-06,
+      "loss": 0.607,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2516139711968217,
+      "grad_norm": 3.0494797229766846,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.6154,
+      "step": 1520
+    },
+    {
+      "epoch": 0.25326932627048504,
+      "grad_norm": 3.6615447998046875,
+      "learning_rate": 8.43353557639272e-06,
+      "loss": 0.6569,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2549246813441483,
+      "grad_norm": 4.182124614715576,
+      "learning_rate": 8.488692774407062e-06,
+      "loss": 0.6069,
+      "step": 1540
+    },
+    {
+      "epoch": 0.25658003641781163,
+      "grad_norm": 3.966855525970459,
+      "learning_rate": 8.543849972421401e-06,
+      "loss": 0.6061,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2582353914914749,
+      "grad_norm": 3.1076576709747314,
+      "learning_rate": 8.599007170435742e-06,
+      "loss": 0.5949,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2598907465651382,
+      "grad_norm": 2.6115076541900635,
+      "learning_rate": 8.654164368450083e-06,
+      "loss": 0.5933,
+      "step": 1570
+    },
+    {
+      "epoch": 0.26154610163880154,
+      "grad_norm": 2.357980489730835,
+      "learning_rate": 8.709321566464424e-06,
+      "loss": 0.6573,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2632014567124648,
+      "grad_norm": 2.8000941276550293,
+      "learning_rate": 8.764478764478765e-06,
+      "loss": 0.6048,
+      "step": 1590
+    },
+    {
+      "epoch": 0.26485681178612813,
+      "grad_norm": 2.700944185256958,
+      "learning_rate": 8.819635962493106e-06,
+      "loss": 0.6463,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2665121668597914,
+      "grad_norm": 2.6585984230041504,
+      "learning_rate": 8.874793160507447e-06,
+      "loss": 0.6004,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2681675219334547,
+      "grad_norm": 2.9849841594696045,
+      "learning_rate": 8.929950358521788e-06,
+      "loss": 0.5947,
+      "step": 1620
+    },
+    {
+      "epoch": 0.26982287700711804,
+      "grad_norm": 2.6921303272247314,
+      "learning_rate": 8.985107556536129e-06,
+      "loss": 0.6284,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2714782320807813,
+      "grad_norm": 3.0850579738616943,
+      "learning_rate": 9.04026475455047e-06,
+      "loss": 0.5238,
+      "step": 1640
+    },
+    {
+      "epoch": 0.27313358715444463,
+      "grad_norm": 3.351339340209961,
+      "learning_rate": 9.09542195256481e-06,
+      "loss": 0.635,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2747889422281079,
+      "grad_norm": 2.2097935676574707,
+      "learning_rate": 9.15057915057915e-06,
+      "loss": 0.6214,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2764442973017712,
+      "grad_norm": 3.295842170715332,
+      "learning_rate": 9.205736348593493e-06,
+      "loss": 0.6673,
+      "step": 1670
+    },
+    {
+      "epoch": 0.27809965237543455,
+      "grad_norm": 3.469433069229126,
+      "learning_rate": 9.260893546607832e-06,
+      "loss": 0.6393,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2797550074490978,
+      "grad_norm": 2.8949124813079834,
+      "learning_rate": 9.316050744622173e-06,
+      "loss": 0.6166,
+      "step": 1690
+    },
+    {
+      "epoch": 0.28141036252276114,
+      "grad_norm": 3.850156545639038,
+      "learning_rate": 9.371207942636514e-06,
+      "loss": 0.5904,
+      "step": 1700
+    },
+    {
+      "epoch": 0.28306571759642446,
+      "grad_norm": 2.344604730606079,
+      "learning_rate": 9.426365140650855e-06,
+      "loss": 0.6567,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2847210726700877,
+      "grad_norm": 2.964304208755493,
+      "learning_rate": 9.481522338665196e-06,
+      "loss": 0.6023,
+      "step": 1720
+    },
+    {
+      "epoch": 0.28637642774375105,
+      "grad_norm": 3.1953861713409424,
+      "learning_rate": 9.536679536679537e-06,
+      "loss": 0.6116,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2880317828174143,
+      "grad_norm": 2.340214967727661,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 0.5774,
+      "step": 1740
+    },
+    {
+      "epoch": 0.28968713789107764,
+      "grad_norm": 2.7652084827423096,
+      "learning_rate": 9.64699393270822e-06,
+      "loss": 0.6332,
+      "step": 1750
+    },
+    {
+      "epoch": 0.29134249296474096,
+      "grad_norm": 2.380476713180542,
+      "learning_rate": 9.70215113072256e-06,
+      "loss": 0.5401,
+      "step": 1760
+    },
+    {
+      "epoch": 0.29299784803840423,
+      "grad_norm": 2.936915397644043,
+      "learning_rate": 9.757308328736901e-06,
+      "loss": 0.6103,
+      "step": 1770
+    },
+    {
+      "epoch": 0.29465320311206755,
+      "grad_norm": 2.580601215362549,
+      "learning_rate": 9.812465526751242e-06,
+      "loss": 0.6312,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2963085581857308,
+      "grad_norm": 2.9639976024627686,
+      "learning_rate": 9.867622724765583e-06,
+      "loss": 0.5844,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29796391325939414,
+      "grad_norm": 2.4568283557891846,
+      "learning_rate": 9.922779922779924e-06,
+      "loss": 0.6088,
+      "step": 1800
+    },
+    {
+      "epoch": 0.29961926833305746,
+      "grad_norm": 2.4476640224456787,
+      "learning_rate": 9.977937120794265e-06,
+      "loss": 0.5739,
+      "step": 1810
+    },
+    {
+      "epoch": 0.30127462340672073,
+      "grad_norm": 2.506065845489502,
+      "learning_rate": 9.999996660862644e-06,
+      "loss": 0.6742,
+      "step": 1820
+    },
+    {
+      "epoch": 0.30292997848038405,
+      "grad_norm": 2.4358372688293457,
+      "learning_rate": 9.999976255039388e-06,
+      "loss": 0.6403,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3045853335540473,
+      "grad_norm": 3.3008499145507812,
+      "learning_rate": 9.999937298544802e-06,
+      "loss": 0.6091,
+      "step": 1840
+    },
+    {
+      "epoch": 0.30624068862771064,
+      "grad_norm": 3.557581901550293,
+      "learning_rate": 9.99987979152342e-06,
+      "loss": 0.6313,
+      "step": 1850
+    },
+    {
+      "epoch": 0.30789604370137397,
+      "grad_norm": 2.6936228275299072,
+      "learning_rate": 9.999803734188604e-06,
+      "loss": 0.6611,
+      "step": 1860
+    },
+    {
+      "epoch": 0.30955139877503723,
+      "grad_norm": 3.0474772453308105,
+      "learning_rate": 9.999709126822536e-06,
+      "loss": 0.5883,
+      "step": 1870
+    },
+    {
+      "epoch": 0.31120675384870056,
+      "grad_norm": 2.5758018493652344,
+      "learning_rate": 9.999595969776225e-06,
+      "loss": 0.5915,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3128621089223638,
+      "grad_norm": 3.6697728633880615,
+      "learning_rate": 9.999464263469503e-06,
+      "loss": 0.5736,
+      "step": 1890
+    },
+    {
+      "epoch": 0.31451746399602715,
+      "grad_norm": 2.458699941635132,
+      "learning_rate": 9.999314008391014e-06,
+      "loss": 0.6448,
+      "step": 1900
+    },
+    {
+      "epoch": 0.31617281906969047,
+      "grad_norm": 2.6207146644592285,
+      "learning_rate": 9.999145205098234e-06,
+      "loss": 0.57,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31782817414335374,
+      "grad_norm": 1.6553720235824585,
+      "learning_rate": 9.998957854217444e-06,
+      "loss": 0.5705,
+      "step": 1920
+    },
+    {
+      "epoch": 0.31948352921701706,
+      "grad_norm": 2.402009963989258,
+      "learning_rate": 9.998751956443748e-06,
+      "loss": 0.6143,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3211388842906803,
+      "grad_norm": 2.5485851764678955,
+      "learning_rate": 9.998527512541056e-06,
+      "loss": 0.5782,
+      "step": 1940
+    },
+    {
+      "epoch": 0.32279423936434365,
+      "grad_norm": 2.4615025520324707,
+      "learning_rate": 9.998284523342088e-06,
+      "loss": 0.6105,
+      "step": 1950
+    },
+    {
+      "epoch": 0.32444959443800697,
+      "grad_norm": 2.0781445503234863,
+      "learning_rate": 9.998022989748371e-06,
+      "loss": 0.5435,
+      "step": 1960
+    },
+    {
+      "epoch": 0.32610494951167024,
+      "grad_norm": 2.8392488956451416,
+      "learning_rate": 9.997742912730238e-06,
+      "loss": 0.6502,
+      "step": 1970
+    },
+    {
+      "epoch": 0.32776030458533356,
+      "grad_norm": 3.554027557373047,
+      "learning_rate": 9.997444293326812e-06,
+      "loss": 0.6015,
+      "step": 1980
+    },
+    {
+      "epoch": 0.32941565965899683,
+      "grad_norm": 2.370697259902954,
+      "learning_rate": 9.99712713264602e-06,
+      "loss": 0.6348,
+      "step": 1990
+    },
+    {
+      "epoch": 0.33107101473266015,
+      "grad_norm": 2.2110366821289062,
+      "learning_rate": 9.996791431864577e-06,
+      "loss": 0.5641,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3327263698063235,
+      "grad_norm": 2.8644583225250244,
+      "learning_rate": 9.996437192227979e-06,
+      "loss": 0.5921,
+      "step": 2010
+    },
+    {
+      "epoch": 0.33438172487998674,
+      "grad_norm": 1.8451977968215942,
+      "learning_rate": 9.996064415050515e-06,
+      "loss": 0.6194,
+      "step": 2020
+    },
+    {
+      "epoch": 0.33603707995365006,
+      "grad_norm": 2.718644857406616,
+      "learning_rate": 9.99567310171524e-06,
+      "loss": 0.5981,
+      "step": 2030
+    },
+    {
+      "epoch": 0.33769243502731333,
+      "grad_norm": 2.4472815990448,
+      "learning_rate": 9.995263253673989e-06,
+      "loss": 0.6079,
+      "step": 2040
+    },
+    {
+      "epoch": 0.33934779010097665,
+      "grad_norm": 2.6010594367980957,
+      "learning_rate": 9.994834872447357e-06,
+      "loss": 0.608,
+      "step": 2050
+    },
+    {
+      "epoch": 0.34100314517464,
+      "grad_norm": 3.013094425201416,
+      "learning_rate": 9.994387959624707e-06,
+      "loss": 0.5139,
+      "step": 2060
+    },
+    {
+      "epoch": 0.34265850024830324,
+      "grad_norm": 2.3363730907440186,
+      "learning_rate": 9.993922516864154e-06,
+      "loss": 0.5739,
+      "step": 2070
+    },
+    {
+      "epoch": 0.34431385532196657,
+      "grad_norm": 2.283562421798706,
+      "learning_rate": 9.993438545892557e-06,
+      "loss": 0.5872,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3459692103956299,
+      "grad_norm": 2.672161817550659,
+      "learning_rate": 9.992936048505525e-06,
+      "loss": 0.592,
+      "step": 2090
+    },
+    {
+      "epoch": 0.34762456546929316,
+      "grad_norm": 2.5038950443267822,
+      "learning_rate": 9.9924150265674e-06,
+      "loss": 0.5921,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3492799205429565,
+      "grad_norm": 2.3302109241485596,
+      "learning_rate": 9.991875482011251e-06,
+      "loss": 0.6143,
+      "step": 2110
+    },
+    {
+      "epoch": 0.35093527561661975,
+      "grad_norm": 3.3396990299224854,
+      "learning_rate": 9.991317416838871e-06,
+      "loss": 0.5817,
+      "step": 2120
+    },
+    {
+      "epoch": 0.35259063069028307,
+      "grad_norm": 2.076221466064453,
+      "learning_rate": 9.990740833120765e-06,
+      "loss": 0.5941,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3542459857639464,
+      "grad_norm": 2.247899293899536,
+      "learning_rate": 9.990145732996149e-06,
+      "loss": 0.6015,
+      "step": 2140
+    },
+    {
+      "epoch": 0.35590134083760966,
+      "grad_norm": 2.4962527751922607,
+      "learning_rate": 9.989532118672933e-06,
+      "loss": 0.6052,
+      "step": 2150
+    },
+    {
+      "epoch": 0.357556695911273,
+      "grad_norm": 2.5895111560821533,
+      "learning_rate": 9.988899992427719e-06,
+      "loss": 0.6154,
+      "step": 2160
+    },
+    {
+      "epoch": 0.35921205098493625,
+      "grad_norm": 2.0279507637023926,
+      "learning_rate": 9.988249356605793e-06,
+      "loss": 0.5683,
+      "step": 2170
+    },
+    {
+      "epoch": 0.36086740605859957,
+      "grad_norm": 2.5845603942871094,
+      "learning_rate": 9.98758021362111e-06,
+      "loss": 0.6027,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3625227611322629,
+      "grad_norm": 2.4495720863342285,
+      "learning_rate": 9.986892565956296e-06,
+      "loss": 0.6118,
+      "step": 2190
+    },
+    {
+      "epoch": 0.36417811620592616,
+      "grad_norm": 2.7284724712371826,
+      "learning_rate": 9.986186416162624e-06,
+      "loss": 0.6262,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3658334712795895,
+      "grad_norm": 3.050304412841797,
+      "learning_rate": 9.985461766860021e-06,
+      "loss": 0.5922,
+      "step": 2210
+    },
+    {
+      "epoch": 0.36748882635325275,
+      "grad_norm": 2.5775296688079834,
+      "learning_rate": 9.984718620737044e-06,
+      "loss": 0.5898,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3691441814269161,
+      "grad_norm": 2.2817842960357666,
+      "learning_rate": 9.983956980550877e-06,
+      "loss": 0.5528,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3707995365005794,
+      "grad_norm": 2.512723445892334,
+      "learning_rate": 9.983176849127323e-06,
+      "loss": 0.6015,
+      "step": 2240
+    },
+    {
+      "epoch": 0.37245489157424266,
+      "grad_norm": 3.1209611892700195,
+      "learning_rate": 9.982378229360785e-06,
+      "loss": 0.5624,
+      "step": 2250
+    },
+    {
+      "epoch": 0.374110246647906,
+      "grad_norm": 2.4052345752716064,
+      "learning_rate": 9.981561124214267e-06,
+      "loss": 0.5279,
+      "step": 2260
+    },
+    {
+      "epoch": 0.37576560172156925,
+      "grad_norm": 3.7475814819335938,
+      "learning_rate": 9.980725536719352e-06,
+      "loss": 0.6319,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3774209567952326,
+      "grad_norm": 2.4055349826812744,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.5653,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3790763118688959,
+      "grad_norm": 2.1154866218566895,
+      "learning_rate": 9.978998927153516e-06,
+      "loss": 0.6503,
+      "step": 2290
+    },
+    {
+      "epoch": 0.38073166694255917,
+      "grad_norm": 2.19960618019104,
+      "learning_rate": 9.978107911488581e-06,
+      "loss": 0.5311,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3823870220162225,
+      "grad_norm": 2.6673924922943115,
+      "learning_rate": 9.97719842628719e-06,
+      "loss": 0.59,
+      "step": 2310
+    },
+    {
+      "epoch": 0.38404237708988576,
+      "grad_norm": 2.573767900466919,
+      "learning_rate": 9.976270474923675e-06,
+      "loss": 0.559,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3856977321635491,
+      "grad_norm": 2.6069254875183105,
+      "learning_rate": 9.975324060840874e-06,
+      "loss": 0.6021,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3873530872372124,
+      "grad_norm": 2.925135374069214,
+      "learning_rate": 9.97435918755013e-06,
+      "loss": 0.5488,
+      "step": 2340
+    },
+    {
+      "epoch": 0.38900844231087567,
+      "grad_norm": 2.631096601486206,
+      "learning_rate": 9.973375858631266e-06,
+      "loss": 0.5977,
+      "step": 2350
+    },
+    {
+      "epoch": 0.390663797384539,
+      "grad_norm": 2.7871811389923096,
+      "learning_rate": 9.972374077732585e-06,
+      "loss": 0.6449,
+      "step": 2360
+    },
+    {
+      "epoch": 0.39231915245820226,
+      "grad_norm": 1.6462632417678833,
+      "learning_rate": 9.971353848570845e-06,
+      "loss": 0.5528,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3939745075318656,
+      "grad_norm": 3.6753463745117188,
+      "learning_rate": 9.97031517493125e-06,
+      "loss": 0.562,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3956298626055289,
+      "grad_norm": 2.3143789768218994,
+      "learning_rate": 9.969258060667435e-06,
+      "loss": 0.59,
+      "step": 2390
+    },
+    {
+      "epoch": 0.39728521767919217,
+      "grad_norm": 1.9230400323867798,
+      "learning_rate": 9.96818250970146e-06,
+      "loss": 0.5805,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3989405727528555,
+      "grad_norm": 2.318477153778076,
+      "learning_rate": 9.96708852602378e-06,
+      "loss": 0.6163,
+      "step": 2410
+    },
+    {
+      "epoch": 0.40059592782651876,
+      "grad_norm": 2.983262300491333,
+      "learning_rate": 9.965976113693238e-06,
+      "loss": 0.5384,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4022512829001821,
+      "grad_norm": 2.5981662273406982,
+      "learning_rate": 9.964845276837057e-06,
+      "loss": 0.5324,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4039066379738454,
+      "grad_norm": 2.5291125774383545,
+      "learning_rate": 9.96369601965081e-06,
+      "loss": 0.5857,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4055619930475087,
+      "grad_norm": 2.275705575942993,
+      "learning_rate": 9.962528346398418e-06,
+      "loss": 0.5985,
+      "step": 2450
+    },
+    {
+      "epoch": 0.407217348121172,
+      "grad_norm": 3.31367826461792,
+      "learning_rate": 9.961342261412125e-06,
+      "loss": 0.5318,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4088727031948353,
+      "grad_norm": 2.098184585571289,
+      "learning_rate": 9.960137769092487e-06,
+      "loss": 0.5643,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4105280582684986,
+      "grad_norm": 2.2636635303497314,
+      "learning_rate": 9.958914873908353e-06,
+      "loss": 0.5759,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4121834133421619,
+      "grad_norm": 2.2176895141601562,
+      "learning_rate": 9.95767358039685e-06,
+      "loss": 0.5145,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4138387684158252,
+      "grad_norm": 2.798469305038452,
+      "learning_rate": 9.956413893163365e-06,
+      "loss": 0.6119,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4154941234894885,
+      "grad_norm": 2.1799635887145996,
+      "learning_rate": 9.95513581688153e-06,
+      "loss": 0.5912,
+      "step": 2510
+    },
+    {
+      "epoch": 0.4171494785631518,
+      "grad_norm": 3.491344690322876,
+      "learning_rate": 9.953839356293202e-06,
+      "loss": 0.5705,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4188048336368151,
+      "grad_norm": 3.2916722297668457,
+      "learning_rate": 9.952524516208447e-06,
+      "loss": 0.5886,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4204601887104784,
+      "grad_norm": 2.2452287673950195,
+      "learning_rate": 9.951191301505519e-06,
+      "loss": 0.4922,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4221155437841417,
+      "grad_norm": 2.5860681533813477,
+      "learning_rate": 9.949839717130849e-06,
+      "loss": 0.5827,
+      "step": 2550
+    },
+    {
+      "epoch": 0.423770898857805,
+      "grad_norm": 2.4785280227661133,
+      "learning_rate": 9.94846976809902e-06,
+      "loss": 0.5541,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4254262539314683,
+      "grad_norm": 1.9815374612808228,
+      "learning_rate": 9.947081459492751e-06,
+      "loss": 0.5863,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4270816090051316,
+      "grad_norm": 2.4965639114379883,
+      "learning_rate": 9.945674796462879e-06,
+      "loss": 0.5708,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4287369640787949,
+      "grad_norm": 3.2701520919799805,
+      "learning_rate": 9.944249784228335e-06,
+      "loss": 0.5784,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4303923191524582,
+      "grad_norm": 2.430939197540283,
+      "learning_rate": 9.942806428076132e-06,
+      "loss": 0.5823,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4320476742261215,
+      "grad_norm": 2.519144296646118,
+      "learning_rate": 9.941344733361344e-06,
+      "loss": 0.5767,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4337030292997848,
+      "grad_norm": 2.906513214111328,
+      "learning_rate": 9.939864705507073e-06,
+      "loss": 0.548,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4353583843734481,
+      "grad_norm": 2.306670665740967,
+      "learning_rate": 9.938366350004454e-06,
+      "loss": 0.5714,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4370137394471114,
+      "grad_norm": 2.199439525604248,
+      "learning_rate": 9.93684967241261e-06,
+      "loss": 0.5586,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4386690945207747,
+      "grad_norm": 2.4085533618927,
+      "learning_rate": 9.935314678358644e-06,
+      "loss": 0.5811,
+      "step": 2650
+    },
+    {
+      "epoch": 0.440324449594438,
+      "grad_norm": 2.153773307800293,
+      "learning_rate": 9.933761373537621e-06,
+      "loss": 0.5879,
+      "step": 2660
+    },
+    {
+      "epoch": 0.44197980466810133,
+      "grad_norm": 2.2745254039764404,
+      "learning_rate": 9.932189763712537e-06,
+      "loss": 0.5682,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4436351597417646,
+      "grad_norm": 2.499537706375122,
+      "learning_rate": 9.9305998547143e-06,
+      "loss": 0.5803,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4452905148154279,
+      "grad_norm": 2.2464380264282227,
+      "learning_rate": 9.928991652441717e-06,
+      "loss": 0.5679,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4469458698890912,
+      "grad_norm": 2.874124526977539,
+      "learning_rate": 9.92736516286146e-06,
+      "loss": 0.5925,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4486012249627545,
+      "grad_norm": 2.706362247467041,
+      "learning_rate": 9.925720392008056e-06,
+      "loss": 0.5674,
+      "step": 2710
+    },
+    {
+      "epoch": 0.45025658003641783,
+      "grad_norm": 2.4833574295043945,
+      "learning_rate": 9.924057345983851e-06,
+      "loss": 0.5762,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4519119351100811,
+      "grad_norm": 1.9858320951461792,
+      "learning_rate": 9.922376030959e-06,
+      "loss": 0.5669,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4535672901837444,
+      "grad_norm": 2.3540823459625244,
+      "learning_rate": 9.920676453171438e-06,
+      "loss": 0.5652,
+      "step": 2740
+    },
+    {
+      "epoch": 0.4552226452574077,
+      "grad_norm": 2.1306300163269043,
+      "learning_rate": 9.918958618926855e-06,
+      "loss": 0.5218,
+      "step": 2750
+    },
+    {
+      "epoch": 0.456878000331071,
+      "grad_norm": 2.7033262252807617,
+      "learning_rate": 9.917222534598679e-06,
+      "loss": 0.5731,
+      "step": 2760
+    },
+    {
+      "epoch": 0.45853335540473433,
+      "grad_norm": 1.9918255805969238,
+      "learning_rate": 9.915468206628046e-06,
+      "loss": 0.5535,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4601887104783976,
+      "grad_norm": 2.858841896057129,
+      "learning_rate": 9.913695641523777e-06,
+      "loss": 0.601,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4618440655520609,
+      "grad_norm": 2.050906181335449,
+      "learning_rate": 9.91190484586236e-06,
+      "loss": 0.5605,
+      "step": 2790
+    },
+    {
+      "epoch": 0.46349942062572425,
+      "grad_norm": 2.51007342338562,
+      "learning_rate": 9.910095826287918e-06,
+      "loss": 0.6002,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4651547756993875,
+      "grad_norm": 2.3648529052734375,
+      "learning_rate": 9.908268589512187e-06,
+      "loss": 0.6065,
+      "step": 2810
+    },
+    {
+      "epoch": 0.46681013077305084,
+      "grad_norm": 2.8302104473114014,
+      "learning_rate": 9.906423142314497e-06,
+      "loss": 0.6004,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4684654858467141,
+      "grad_norm": 2.7113876342773438,
+      "learning_rate": 9.904559491541735e-06,
+      "loss": 0.5727,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4701208409203774,
+      "grad_norm": 1.8766558170318604,
+      "learning_rate": 9.902677644108327e-06,
+      "loss": 0.4911,
+      "step": 2840
+    },
+    {
+      "epoch": 0.47177619599404075,
+      "grad_norm": 2.576292037963867,
+      "learning_rate": 9.900777606996213e-06,
+      "loss": 0.5602,
+      "step": 2850
+    },
+    {
+      "epoch": 0.473431551067704,
+      "grad_norm": 2.50095796585083,
+      "learning_rate": 9.898859387254823e-06,
+      "loss": 0.5659,
+      "step": 2860
+    },
+    {
+      "epoch": 0.47508690614136734,
+      "grad_norm": 2.3355982303619385,
+      "learning_rate": 9.89692299200104e-06,
+      "loss": 0.58,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4767422612150306,
+      "grad_norm": 2.1642611026763916,
+      "learning_rate": 9.894968428419187e-06,
+      "loss": 0.563,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4783976162886939,
+      "grad_norm": 2.645699977874756,
+      "learning_rate": 9.892995703760988e-06,
+      "loss": 0.5911,
+      "step": 2890
+    },
+    {
+      "epoch": 0.48005297136235725,
+      "grad_norm": 3.1719911098480225,
+      "learning_rate": 9.891004825345555e-06,
+      "loss": 0.5638,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4817083264360205,
+      "grad_norm": 2.728663444519043,
+      "learning_rate": 9.888995800559347e-06,
+      "loss": 0.5808,
+      "step": 2910
+    },
+    {
+      "epoch": 0.48336368150968384,
+      "grad_norm": 2.2633676528930664,
+      "learning_rate": 9.886968636856153e-06,
+      "loss": 0.4939,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4850190365833471,
+      "grad_norm": 1.8135114908218384,
+      "learning_rate": 9.884923341757056e-06,
+      "loss": 0.5163,
+      "step": 2930
+    },
+    {
+      "epoch": 0.48667439165701043,
+      "grad_norm": 2.147280216217041,
+      "learning_rate": 9.882859922850412e-06,
+      "loss": 0.5427,
+      "step": 2940
+    },
+    {
+      "epoch": 0.48832974673067375,
+      "grad_norm": 2.248060464859009,
+      "learning_rate": 9.88077838779182e-06,
+      "loss": 0.5357,
+      "step": 2950
+    },
+    {
+      "epoch": 0.489985101804337,
+      "grad_norm": 2.9170844554901123,
+      "learning_rate": 9.87867874430409e-06,
+      "loss": 0.5257,
+      "step": 2960
+    },
+    {
+      "epoch": 0.49164045687800034,
+      "grad_norm": 2.0134947299957275,
+      "learning_rate": 9.87656100017722e-06,
+      "loss": 0.559,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4932958119516636,
+      "grad_norm": 2.3166182041168213,
+      "learning_rate": 9.87442516326836e-06,
+      "loss": 0.5788,
+      "step": 2980
+    },
+    {
+      "epoch": 0.49495116702532693,
+      "grad_norm": 2.9038329124450684,
+      "learning_rate": 9.87227124150179e-06,
+      "loss": 0.6224,
+      "step": 2990
+    },
+    {
+      "epoch": 0.49660652209899026,
+      "grad_norm": 1.8098182678222656,
+      "learning_rate": 9.870099242868887e-06,
+      "loss": 0.5759,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4982618771726535,
+      "grad_norm": 2.722839593887329,
+      "learning_rate": 9.867909175428096e-06,
+      "loss": 0.4706,
+      "step": 3010
+    },
+    {
+      "epoch": 0.49991723224631684,
+      "grad_norm": 2.420764446258545,
+      "learning_rate": 9.8657010473049e-06,
+      "loss": 0.577,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5015725873199801,
+      "grad_norm": 2.254530906677246,
+      "learning_rate": 9.86347486669179e-06,
+      "loss": 0.5189,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5032279423936434,
+      "grad_norm": 2.428337335586548,
+      "learning_rate": 9.861230641848233e-06,
+      "loss": 0.6204,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5048832974673068,
+      "grad_norm": 2.543137550354004,
+      "learning_rate": 9.858968381100646e-06,
+      "loss": 0.5617,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5065386525409701,
+      "grad_norm": 2.627063751220703,
+      "learning_rate": 9.856688092842357e-06,
+      "loss": 0.5348,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5081940076146333,
+      "grad_norm": 2.149599075317383,
+      "learning_rate": 9.854389785533585e-06,
+      "loss": 0.5694,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5098493626882966,
+      "grad_norm": 2.343366861343384,
+      "learning_rate": 9.852073467701398e-06,
+      "loss": 0.5729,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5115047177619599,
+      "grad_norm": 2.3842127323150635,
+      "learning_rate": 9.849739147939685e-06,
+      "loss": 0.5948,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5131600728356233,
+      "grad_norm": 2.269623041152954,
+      "learning_rate": 9.84738683490913e-06,
+      "loss": 0.5486,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5148154279092866,
+      "grad_norm": 2.9736173152923584,
+      "learning_rate": 9.845016537337168e-06,
+      "loss": 0.5615,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5164707829829498,
+      "grad_norm": 2.21465802192688,
+      "learning_rate": 9.842628264017969e-06,
+      "loss": 0.5914,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5181261380566131,
+      "grad_norm": 2.490063428878784,
+      "learning_rate": 9.840222023812383e-06,
+      "loss": 0.5833,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5197814931302764,
+      "grad_norm": 2.3841171264648438,
+      "learning_rate": 9.83779782564793e-06,
+      "loss": 0.577,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5214368482039398,
+      "grad_norm": 1.9457083940505981,
+      "learning_rate": 9.835355678518754e-06,
+      "loss": 0.5872,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5230922032776031,
+      "grad_norm": 2.5294015407562256,
+      "learning_rate": 9.83289559148559e-06,
+      "loss": 0.504,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5247475583512663,
+      "grad_norm": 3.2622299194335938,
+      "learning_rate": 9.830417573675737e-06,
+      "loss": 0.5809,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5264029134249296,
+      "grad_norm": 2.026658773422241,
+      "learning_rate": 9.827921634283015e-06,
+      "loss": 0.5413,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5280582684985929,
+      "grad_norm": 2.9788129329681396,
+      "learning_rate": 9.825407782567738e-06,
+      "loss": 0.5578,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5297136235722563,
+      "grad_norm": 2.1878185272216797,
+      "learning_rate": 9.822876027856679e-06,
+      "loss": 0.5203,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5313689786459196,
+      "grad_norm": 2.129364013671875,
+      "learning_rate": 9.820326379543032e-06,
+      "loss": 0.5794,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5330243337195828,
+      "grad_norm": 2.1982228755950928,
+      "learning_rate": 9.817758847086381e-06,
+      "loss": 0.5827,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5346796887932461,
+      "grad_norm": 2.233294725418091,
+      "learning_rate": 9.815173440012657e-06,
+      "loss": 0.5868,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5363350438669094,
+      "grad_norm": 2.100982666015625,
+      "learning_rate": 9.81257016791412e-06,
+      "loss": 0.5339,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5379903989405728,
+      "grad_norm": 2.1711225509643555,
+      "learning_rate": 9.809949040449298e-06,
+      "loss": 0.5841,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5396457540142361,
+      "grad_norm": 2.9437408447265625,
+      "learning_rate": 9.807310067342976e-06,
+      "loss": 0.5521,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5413011090878993,
+      "grad_norm": 4.610905170440674,
+      "learning_rate": 9.804653258386145e-06,
+      "loss": 0.5947,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5429564641615626,
+      "grad_norm": 2.275070905685425,
+      "learning_rate": 9.801978623435967e-06,
+      "loss": 0.5725,
+      "step": 3280
+    },
+    {
+      "epoch": 0.544611819235226,
+      "grad_norm": 2.3280909061431885,
+      "learning_rate": 9.799286172415746e-06,
+      "loss": 0.5722,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5462671743088893,
+      "grad_norm": 2.2407679557800293,
+      "learning_rate": 9.796575915314884e-06,
+      "loss": 0.5641,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5479225293825526,
+      "grad_norm": 2.3367135524749756,
+      "learning_rate": 9.793847862188848e-06,
+      "loss": 0.5626,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5495778844562158,
+      "grad_norm": 1.977503776550293,
+      "learning_rate": 9.791102023159125e-06,
+      "loss": 0.5515,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5512332395298791,
+      "grad_norm": 2.1004018783569336,
+      "learning_rate": 9.7883384084132e-06,
+      "loss": 0.5421,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5528885946035424,
+      "grad_norm": 2.568941116333008,
+      "learning_rate": 9.7855570282045e-06,
+      "loss": 0.5618,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5545439496772058,
+      "grad_norm": 2.1921818256378174,
+      "learning_rate": 9.782757892852367e-06,
+      "loss": 0.5418,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5561993047508691,
+      "grad_norm": 2.2598955631256104,
+      "learning_rate": 9.779941012742025e-06,
+      "loss": 0.5643,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5578546598245324,
+      "grad_norm": 2.0559098720550537,
+      "learning_rate": 9.77710639832452e-06,
+      "loss": 0.6003,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5595100148981956,
+      "grad_norm": 2.366570234298706,
+      "learning_rate": 9.774254060116703e-06,
+      "loss": 0.5208,
+      "step": 3380
+    },
+    {
+      "epoch": 0.561165369971859,
+      "grad_norm": 2.092085123062134,
+      "learning_rate": 9.771384008701185e-06,
+      "loss": 0.5987,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5628207250455223,
+      "grad_norm": 2.402676820755005,
+      "learning_rate": 9.768496254726293e-06,
+      "loss": 0.5836,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5644760801191856,
+      "grad_norm": 2.451547861099243,
+      "learning_rate": 9.765590808906029e-06,
+      "loss": 0.5874,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5661314351928489,
+      "grad_norm": 1.9011714458465576,
+      "learning_rate": 9.76266768202004e-06,
+      "loss": 0.495,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5677867902665121,
+      "grad_norm": 2.3859992027282715,
+      "learning_rate": 9.759726884913572e-06,
+      "loss": 0.5934,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5694421453401755,
+      "grad_norm": 2.100632905960083,
+      "learning_rate": 9.756768428497427e-06,
+      "loss": 0.5215,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5710975004138388,
+      "grad_norm": 2.366837501525879,
+      "learning_rate": 9.753792323747928e-06,
+      "loss": 0.4966,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5727528554875021,
+      "grad_norm": 2.0204997062683105,
+      "learning_rate": 9.75079858170688e-06,
+      "loss": 0.5586,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5744082105611654,
+      "grad_norm": 2.957526683807373,
+      "learning_rate": 9.747787213481511e-06,
+      "loss": 0.5437,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5760635656348286,
+      "grad_norm": 2.0730838775634766,
+      "learning_rate": 9.744758230244465e-06,
+      "loss": 0.5447,
+      "step": 3480
+    },
+    {
+      "epoch": 0.577718920708492,
+      "grad_norm": 2.0039265155792236,
+      "learning_rate": 9.741711643233724e-06,
+      "loss": 0.5071,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5793742757821553,
+      "grad_norm": 1.7638771533966064,
+      "learning_rate": 9.73864746375259e-06,
+      "loss": 0.5415,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5810296308558186,
+      "grad_norm": 2.048380136489868,
+      "learning_rate": 9.735565703169634e-06,
+      "loss": 0.5371,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5826849859294819,
+      "grad_norm": 2.077610492706299,
+      "learning_rate": 9.732466372918656e-06,
+      "loss": 0.5548,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5843403410031451,
+      "grad_norm": 2.1865971088409424,
+      "learning_rate": 9.729349484498642e-06,
+      "loss": 0.5595,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5859956960768085,
+      "grad_norm": 2.1196391582489014,
+      "learning_rate": 9.726215049473722e-06,
+      "loss": 0.5333,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5876510511504718,
+      "grad_norm": 2.613009452819824,
+      "learning_rate": 9.723063079473124e-06,
+      "loss": 0.5031,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5893064062241351,
+      "grad_norm": 2.1916675567626953,
+      "learning_rate": 9.719893586191137e-06,
+      "loss": 0.4728,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5909617612977984,
+      "grad_norm": 1.9465389251708984,
+      "learning_rate": 9.716706581387065e-06,
+      "loss": 0.5766,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5926171163714616,
+      "grad_norm": 2.581099271774292,
+      "learning_rate": 9.713502076885174e-06,
+      "loss": 0.5806,
+      "step": 3580
+    },
+    {
+      "epoch": 0.594272471445125,
+      "grad_norm": 2.1889805793762207,
+      "learning_rate": 9.710280084574667e-06,
+      "loss": 0.5075,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5959278265187883,
+      "grad_norm": 1.5537147521972656,
+      "learning_rate": 9.707040616409623e-06,
+      "loss": 0.5388,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5975831815924516,
+      "grad_norm": 2.479175329208374,
+      "learning_rate": 9.703783684408961e-06,
+      "loss": 0.5215,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5992385366661149,
+      "grad_norm": 2.344923257827759,
+      "learning_rate": 9.700509300656395e-06,
+      "loss": 0.5267,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6008938917397781,
+      "grad_norm": 3.7248928546905518,
+      "learning_rate": 9.697217477300385e-06,
+      "loss": 0.5766,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6025492468134415,
+      "grad_norm": 2.325535297393799,
+      "learning_rate": 9.693908226554094e-06,
+      "loss": 0.5377,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6042046018871048,
+      "grad_norm": 2.5930399894714355,
+      "learning_rate": 9.690581560695346e-06,
+      "loss": 0.5351,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6058599569607681,
+      "grad_norm": 2.1282172203063965,
+      "learning_rate": 9.68723749206658e-06,
+      "loss": 0.5148,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6075153120344314,
+      "grad_norm": 2.241332530975342,
+      "learning_rate": 9.683876033074793e-06,
+      "loss": 0.5813,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6091706671080946,
+      "grad_norm": 1.9913647174835205,
+      "learning_rate": 9.680497196191511e-06,
+      "loss": 0.5826,
+      "step": 3680
+    },
+    {
+      "epoch": 0.610826022181758,
+      "grad_norm": 2.073176622390747,
+      "learning_rate": 9.677100993952732e-06,
+      "loss": 0.5039,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6124813772554213,
+      "grad_norm": 2.6423120498657227,
+      "learning_rate": 9.673687438958883e-06,
+      "loss": 0.4948,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6141367323290846,
+      "grad_norm": 2.061516761779785,
+      "learning_rate": 9.670256543874769e-06,
+      "loss": 0.5831,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6157920874027479,
+      "grad_norm": 2.189225196838379,
+      "learning_rate": 9.666808321429534e-06,
+      "loss": 0.5646,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6174474424764111,
+      "grad_norm": 2.219355344772339,
+      "learning_rate": 9.663342784416609e-06,
+      "loss": 0.553,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6191027975500745,
+      "grad_norm": 2.61342453956604,
+      "learning_rate": 9.659859945693658e-06,
+      "loss": 0.5279,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6207581526237378,
+      "grad_norm": 3.1807034015655518,
+      "learning_rate": 9.656359818182543e-06,
+      "loss": 0.5833,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6224135076974011,
+      "grad_norm": 2.533095121383667,
+      "learning_rate": 9.65284241486927e-06,
+      "loss": 0.5927,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6240688627710644,
+      "grad_norm": 2.798417329788208,
+      "learning_rate": 9.649307748803939e-06,
+      "loss": 0.5413,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6257242178447276,
+      "grad_norm": 2.3580000400543213,
+      "learning_rate": 9.645755833100699e-06,
+      "loss": 0.553,
+      "step": 3780
+    },
+    {
+      "epoch": 0.627379572918391,
+      "grad_norm": 1.9809789657592773,
+      "learning_rate": 9.642186680937695e-06,
+      "loss": 0.4882,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6290349279920543,
+      "grad_norm": 2.5609359741210938,
+      "learning_rate": 9.638600305557025e-06,
+      "loss": 0.5164,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6306902830657176,
+      "grad_norm": 2.4713363647460938,
+      "learning_rate": 9.634996720264684e-06,
+      "loss": 0.535,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6323456381393809,
+      "grad_norm": 2.4818222522735596,
+      "learning_rate": 9.631375938430525e-06,
+      "loss": 0.5605,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6340009932130442,
+      "grad_norm": 2.3091225624084473,
+      "learning_rate": 9.627737973488194e-06,
+      "loss": 0.5705,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6356563482867075,
+      "grad_norm": 2.624854803085327,
+      "learning_rate": 9.624082838935096e-06,
+      "loss": 0.4703,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6373117033603708,
+      "grad_norm": 1.9231913089752197,
+      "learning_rate": 9.620410548332336e-06,
+      "loss": 0.4887,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6389670584340341,
+      "grad_norm": 2.4588680267333984,
+      "learning_rate": 9.616721115304669e-06,
+      "loss": 0.4479,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6406224135076974,
+      "grad_norm": 2.4490933418273926,
+      "learning_rate": 9.61301455354045e-06,
+      "loss": 0.5036,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6422777685813607,
+      "grad_norm": 2.085653305053711,
+      "learning_rate": 9.609290876791589e-06,
+      "loss": 0.5332,
+      "step": 3880
+    },
+    {
+      "epoch": 0.643933123655024,
+      "grad_norm": 2.1635732650756836,
+      "learning_rate": 9.60555009887349e-06,
+      "loss": 0.5819,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6455884787286873,
+      "grad_norm": 1.8872817754745483,
+      "learning_rate": 9.601792233665007e-06,
+      "loss": 0.5124,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6472438338023506,
+      "grad_norm": 2.5642919540405273,
+      "learning_rate": 9.598017295108395e-06,
+      "loss": 0.5272,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6488991888760139,
+      "grad_norm": 2.362342357635498,
+      "learning_rate": 9.594225297209245e-06,
+      "loss": 0.512,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6505545439496772,
+      "grad_norm": 2.04681134223938,
+      "learning_rate": 9.590416254036447e-06,
+      "loss": 0.5667,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6522098990233405,
+      "grad_norm": 2.4778685569763184,
+      "learning_rate": 9.586590179722131e-06,
+      "loss": 0.5527,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6538652540970038,
+      "grad_norm": 2.4283761978149414,
+      "learning_rate": 9.58274708846161e-06,
+      "loss": 0.5567,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6555206091706671,
+      "grad_norm": 2.188366413116455,
+      "learning_rate": 9.57888699451334e-06,
+      "loss": 0.5653,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6571759642443304,
+      "grad_norm": 1.9464526176452637,
+      "learning_rate": 9.575009912198853e-06,
+      "loss": 0.5482,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6588313193179937,
+      "grad_norm": 2.024707555770874,
+      "learning_rate": 9.571115855902715e-06,
+      "loss": 0.5105,
+      "step": 3980
+    },
+    {
+      "epoch": 0.660486674391657,
+      "grad_norm": 2.27622127532959,
+      "learning_rate": 9.567204840072466e-06,
+      "loss": 0.5506,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6621420294653203,
+      "grad_norm": 1.9838156700134277,
+      "learning_rate": 9.563276879218568e-06,
+      "loss": 0.5619,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6637973845389836,
+      "grad_norm": 2.1720376014709473,
+      "learning_rate": 9.559331987914354e-06,
+      "loss": 0.549,
+      "step": 4010
+    },
+    {
+      "epoch": 0.665452739612647,
+      "grad_norm": 1.8095346689224243,
+      "learning_rate": 9.555370180795967e-06,
+      "loss": 0.5652,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6671080946863102,
+      "grad_norm": 2.1045024394989014,
+      "learning_rate": 9.551391472562316e-06,
+      "loss": 0.5002,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6687634497599735,
+      "grad_norm": 2.3684701919555664,
+      "learning_rate": 9.547395877975012e-06,
+      "loss": 0.5148,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6704188048336368,
+      "grad_norm": 2.073254108428955,
+      "learning_rate": 9.543383411858318e-06,
+      "loss": 0.5044,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6720741599073001,
+      "grad_norm": 2.361924171447754,
+      "learning_rate": 9.539354089099092e-06,
+      "loss": 0.5516,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6737295149809635,
+      "grad_norm": 2.4293785095214844,
+      "learning_rate": 9.535307924646735e-06,
+      "loss": 0.5263,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6753848700546267,
+      "grad_norm": 1.8119914531707764,
+      "learning_rate": 9.531244933513129e-06,
+      "loss": 0.5128,
+      "step": 4080
+    },
+    {
+      "epoch": 0.67704022512829,
+      "grad_norm": 1.9259525537490845,
+      "learning_rate": 9.52716513077259e-06,
+      "loss": 0.4829,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6786955802019533,
+      "grad_norm": 2.1588046550750732,
+      "learning_rate": 9.523068531561805e-06,
+      "loss": 0.5347,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6803509352756166,
+      "grad_norm": 2.8328020572662354,
+      "learning_rate": 9.518955151079781e-06,
+      "loss": 0.5013,
+      "step": 4110
+    },
+    {
+      "epoch": 0.68200629034928,
+      "grad_norm": 2.9788923263549805,
+      "learning_rate": 9.514825004587784e-06,
+      "loss": 0.515,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6836616454229433,
+      "grad_norm": 2.3657615184783936,
+      "learning_rate": 9.510678107409282e-06,
+      "loss": 0.5444,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6853170004966065,
+      "grad_norm": 2.1405718326568604,
+      "learning_rate": 9.506514474929896e-06,
+      "loss": 0.4501,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6869723555702698,
+      "grad_norm": 1.9537779092788696,
+      "learning_rate": 9.502334122597335e-06,
+      "loss": 0.4819,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6886277106439331,
+      "grad_norm": 2.405439615249634,
+      "learning_rate": 9.49813706592134e-06,
+      "loss": 0.4746,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6902830657175965,
+      "grad_norm": 2.6599020957946777,
+      "learning_rate": 9.493923320473628e-06,
+      "loss": 0.5271,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6919384207912598,
+      "grad_norm": 2.166412830352783,
+      "learning_rate": 9.489692901887837e-06,
+      "loss": 0.5563,
+      "step": 4180
+    },
+    {
+      "epoch": 0.693593775864923,
+      "grad_norm": 1.994268774986267,
+      "learning_rate": 9.48544582585946e-06,
+      "loss": 0.5235,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6952491309385863,
+      "grad_norm": 3.00091814994812,
+      "learning_rate": 9.481182108145798e-06,
+      "loss": 0.4902,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6969044860122496,
+      "grad_norm": 2.903369426727295,
+      "learning_rate": 9.476901764565887e-06,
+      "loss": 0.5175,
+      "step": 4210
+    },
+    {
+      "epoch": 0.698559841085913,
+      "grad_norm": 2.3588013648986816,
+      "learning_rate": 9.472604811000453e-06,
+      "loss": 0.5236,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7002151961595763,
+      "grad_norm": 2.4532833099365234,
+      "learning_rate": 9.468291263391847e-06,
+      "loss": 0.4895,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7018705512332395,
+      "grad_norm": 2.1600773334503174,
+      "learning_rate": 9.463961137743986e-06,
+      "loss": 0.5251,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7035259063069028,
+      "grad_norm": 1.243881344795227,
+      "learning_rate": 9.459614450122293e-06,
+      "loss": 0.5456,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7051812613805661,
+      "grad_norm": 2.715055465698242,
+      "learning_rate": 9.45525121665364e-06,
+      "loss": 0.5019,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7068366164542295,
+      "grad_norm": 2.0910329818725586,
+      "learning_rate": 9.450871453526285e-06,
+      "loss": 0.5024,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7084919715278928,
+      "grad_norm": 2.3491621017456055,
+      "learning_rate": 9.446475176989816e-06,
+      "loss": 0.5481,
+      "step": 4280
+    },
+    {
+      "epoch": 0.710147326601556,
+      "grad_norm": 2.1992132663726807,
+      "learning_rate": 9.442062403355085e-06,
+      "loss": 0.5166,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7118026816752193,
+      "grad_norm": 2.599609375,
+      "learning_rate": 9.437633148994154e-06,
+      "loss": 0.5459,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7134580367488826,
+      "grad_norm": 1.8802043199539185,
+      "learning_rate": 9.433187430340228e-06,
+      "loss": 0.557,
+      "step": 4310
+    },
+    {
+      "epoch": 0.715113391822546,
+      "grad_norm": 1.949325680732727,
+      "learning_rate": 9.428725263887599e-06,
+      "loss": 0.5385,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7167687468962093,
+      "grad_norm": 2.2506093978881836,
+      "learning_rate": 9.424246666191582e-06,
+      "loss": 0.5123,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7184241019698725,
+      "grad_norm": 2.1846063137054443,
+      "learning_rate": 9.419751653868456e-06,
+      "loss": 0.5491,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7200794570435358,
+      "grad_norm": 2.2196507453918457,
+      "learning_rate": 9.415240243595397e-06,
+      "loss": 0.4818,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7217348121171991,
+      "grad_norm": 1.7722293138504028,
+      "learning_rate": 9.410712452110425e-06,
+      "loss": 0.5207,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7233901671908625,
+      "grad_norm": 1.8259626626968384,
+      "learning_rate": 9.406168296212332e-06,
+      "loss": 0.5789,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7250455222645258,
+      "grad_norm": 3.340482711791992,
+      "learning_rate": 9.401607792760628e-06,
+      "loss": 0.5335,
+      "step": 4380
+    },
+    {
+      "epoch": 0.726700877338189,
+      "grad_norm": 2.3359949588775635,
+      "learning_rate": 9.397030958675473e-06,
+      "loss": 0.5716,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7283562324118523,
+      "grad_norm": 2.8024532794952393,
+      "learning_rate": 9.392437810937615e-06,
+      "loss": 0.5089,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7300115874855156,
+      "grad_norm": 2.381237506866455,
+      "learning_rate": 9.387828366588333e-06,
+      "loss": 0.5497,
+      "step": 4410
+    },
+    {
+      "epoch": 0.731666942559179,
+      "grad_norm": 2.007678270339966,
+      "learning_rate": 9.383202642729363e-06,
+      "loss": 0.4973,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7333222976328423,
+      "grad_norm": 2.0614326000213623,
+      "learning_rate": 9.378560656522845e-06,
+      "loss": 0.5366,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7349776527065055,
+      "grad_norm": 2.1177313327789307,
+      "learning_rate": 9.37390242519125e-06,
+      "loss": 0.5676,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7366330077801688,
+      "grad_norm": 1.7760319709777832,
+      "learning_rate": 9.369227966017326e-06,
+      "loss": 0.4562,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7382883628538321,
+      "grad_norm": 2.2758963108062744,
+      "learning_rate": 9.364537296344029e-06,
+      "loss": 0.5742,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7399437179274955,
+      "grad_norm": 1.834656834602356,
+      "learning_rate": 9.359830433574451e-06,
+      "loss": 0.5405,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7415990730011588,
+      "grad_norm": 1.730483889579773,
+      "learning_rate": 9.35510739517177e-06,
+      "loss": 0.5217,
+      "step": 4480
+    },
+    {
+      "epoch": 0.743254428074822,
+      "grad_norm": 2.00947642326355,
+      "learning_rate": 9.350368198659174e-06,
+      "loss": 0.5132,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7449097831484853,
+      "grad_norm": 2.0409324169158936,
+      "learning_rate": 9.345612861619805e-06,
+      "loss": 0.5914,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7465651382221486,
+      "grad_norm": 2.4126083850860596,
+      "learning_rate": 9.340841401696686e-06,
+      "loss": 0.5179,
+      "step": 4510
+    },
+    {
+      "epoch": 0.748220493295812,
+      "grad_norm": 1.9900450706481934,
+      "learning_rate": 9.336053836592653e-06,
+      "loss": 0.4732,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7498758483694753,
+      "grad_norm": 2.0462656021118164,
+      "learning_rate": 9.331250184070307e-06,
+      "loss": 0.4873,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7515312034431385,
+      "grad_norm": 2.1668014526367188,
+      "learning_rate": 9.326430461951922e-06,
+      "loss": 0.5338,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7531865585168018,
+      "grad_norm": 2.973837375640869,
+      "learning_rate": 9.3215946881194e-06,
+      "loss": 0.5087,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7548419135904652,
+      "grad_norm": 1.9659732580184937,
+      "learning_rate": 9.3167428805142e-06,
+      "loss": 0.5518,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7564972686641285,
+      "grad_norm": 1.702977180480957,
+      "learning_rate": 9.311875057137263e-06,
+      "loss": 0.5058,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7581526237377918,
+      "grad_norm": 1.8664321899414062,
+      "learning_rate": 9.306991236048953e-06,
+      "loss": 0.4974,
+      "step": 4580
+    },
+    {
+      "epoch": 0.759807978811455,
+      "grad_norm": 1.9654330015182495,
+      "learning_rate": 9.302091435368988e-06,
+      "loss": 0.5579,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7614633338851183,
+      "grad_norm": 2.2180731296539307,
+      "learning_rate": 9.297175673276372e-06,
+      "loss": 0.5048,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7631186889587817,
+      "grad_norm": 2.6190831661224365,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.5507,
+      "step": 4610
+    },
+    {
+      "epoch": 0.764774044032445,
+      "grad_norm": 2.6507656574249268,
+      "learning_rate": 9.28729633786524e-06,
+      "loss": 0.5322,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7664293991061083,
+      "grad_norm": 2.1486687660217285,
+      "learning_rate": 9.282332801200557e-06,
+      "loss": 0.5358,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7680847541797715,
+      "grad_norm": 2.3483917713165283,
+      "learning_rate": 9.277353376430758e-06,
+      "loss": 0.5372,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7697401092534348,
+      "grad_norm": 1.9544059038162231,
+      "learning_rate": 9.272358082030263e-06,
+      "loss": 0.5375,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7713954643270982,
+      "grad_norm": 2.1422975063323975,
+      "learning_rate": 9.267346936532377e-06,
+      "loss": 0.5314,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7730508194007615,
+      "grad_norm": 2.089794874191284,
+      "learning_rate": 9.26231995852921e-06,
+      "loss": 0.5177,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7747061744744248,
+      "grad_norm": 2.1427338123321533,
+      "learning_rate": 9.25727716667161e-06,
+      "loss": 0.5485,
+      "step": 4680
+    },
+    {
+      "epoch": 0.776361529548088,
+      "grad_norm": 1.6319947242736816,
+      "learning_rate": 9.252218579669105e-06,
+      "loss": 0.5343,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7780168846217513,
+      "grad_norm": 2.030916452407837,
+      "learning_rate": 9.247144216289821e-06,
+      "loss": 0.5499,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7796722396954147,
+      "grad_norm": 2.0921289920806885,
+      "learning_rate": 9.242054095360413e-06,
+      "loss": 0.4607,
+      "step": 4710
+    },
+    {
+      "epoch": 0.781327594769078,
+      "grad_norm": 1.9406028985977173,
+      "learning_rate": 9.236948235766004e-06,
+      "loss": 0.5819,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7829829498427413,
+      "grad_norm": 2.238834857940674,
+      "learning_rate": 9.231826656450112e-06,
+      "loss": 0.4634,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7846383049164045,
+      "grad_norm": 2.383314609527588,
+      "learning_rate": 9.226689376414571e-06,
+      "loss": 0.5393,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7862936599900678,
+      "grad_norm": 2.1563191413879395,
+      "learning_rate": 9.221536414719472e-06,
+      "loss": 0.5372,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7879490150637312,
+      "grad_norm": 2.5362823009490967,
+      "learning_rate": 9.216367790483085e-06,
+      "loss": 0.5739,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7896043701373945,
+      "grad_norm": 2.164787769317627,
+      "learning_rate": 9.211183522881788e-06,
+      "loss": 0.5437,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7912597252110578,
+      "grad_norm": 1.8392080068588257,
+      "learning_rate": 9.205983631150005e-06,
+      "loss": 0.5152,
+      "step": 4780
+    },
+    {
+      "epoch": 0.792915080284721,
+      "grad_norm": 2.01461124420166,
+      "learning_rate": 9.200768134580124e-06,
+      "loss": 0.5199,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7945704353583843,
+      "grad_norm": 1.9222743511199951,
+      "learning_rate": 9.195537052522428e-06,
+      "loss": 0.5363,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7962257904320477,
+      "grad_norm": 2.3261756896972656,
+      "learning_rate": 9.190290404385025e-06,
+      "loss": 0.5161,
+      "step": 4810
+    },
+    {
+      "epoch": 0.797881145505711,
+      "grad_norm": 1.8926310539245605,
+      "learning_rate": 9.18502820963378e-06,
+      "loss": 0.5086,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7995365005793743,
+      "grad_norm": 2.2464699745178223,
+      "learning_rate": 9.179750487792232e-06,
+      "loss": 0.5446,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8011918556530375,
+      "grad_norm": 2.444520950317383,
+      "learning_rate": 9.17445725844153e-06,
+      "loss": 0.5071,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8028472107267008,
+      "grad_norm": 2.0580263137817383,
+      "learning_rate": 9.169148541220361e-06,
+      "loss": 0.5257,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8045025658003642,
+      "grad_norm": 3.0564193725585938,
+      "learning_rate": 9.163824355824871e-06,
+      "loss": 0.5054,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8061579208740275,
+      "grad_norm": 2.9118645191192627,
+      "learning_rate": 9.158484722008596e-06,
+      "loss": 0.5212,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8078132759476908,
+      "grad_norm": 2.128197193145752,
+      "learning_rate": 9.15312965958239e-06,
+      "loss": 0.4898,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8094686310213541,
+      "grad_norm": 2.0659148693084717,
+      "learning_rate": 9.147759188414348e-06,
+      "loss": 0.5409,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8111239860950173,
+      "grad_norm": 2.312744140625,
+      "learning_rate": 9.142373328429733e-06,
+      "loss": 0.5638,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8127793411686807,
+      "grad_norm": 2.416036367416382,
+      "learning_rate": 9.136972099610901e-06,
+      "loss": 0.5098,
+      "step": 4910
+    },
+    {
+      "epoch": 0.814434696242344,
+      "grad_norm": 2.6420984268188477,
+      "learning_rate": 9.131555521997236e-06,
+      "loss": 0.5312,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8160900513160073,
+      "grad_norm": 2.3493874073028564,
+      "learning_rate": 9.126123615685061e-06,
+      "loss": 0.5222,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8177454063896706,
+      "grad_norm": 2.002620220184326,
+      "learning_rate": 9.120676400827575e-06,
+      "loss": 0.5188,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8194007614633338,
+      "grad_norm": 1.8725508451461792,
+      "learning_rate": 9.11521389763477e-06,
+      "loss": 0.5371,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8210561165369972,
+      "grad_norm": 2.774479389190674,
+      "learning_rate": 9.109736126373364e-06,
+      "loss": 0.4711,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8227114716106605,
+      "grad_norm": 1.985054850578308,
+      "learning_rate": 9.10424310736672e-06,
+      "loss": 0.5487,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8243668266843238,
+      "grad_norm": 2.2153406143188477,
+      "learning_rate": 9.098734860994774e-06,
+      "loss": 0.5096,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8260221817579871,
+      "grad_norm": 2.1796064376831055,
+      "learning_rate": 9.093211407693954e-06,
+      "loss": 0.5158,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8276775368316504,
+      "grad_norm": 2.429814338684082,
+      "learning_rate": 9.087672767957114e-06,
+      "loss": 0.4919,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8293328919053137,
+      "grad_norm": 2.0959694385528564,
+      "learning_rate": 9.082118962333445e-06,
+      "loss": 0.4874,
+      "step": 5010
+    },
+    {
+      "epoch": 0.830988246978977,
+      "grad_norm": 2.562439203262329,
+      "learning_rate": 9.076550011428415e-06,
+      "loss": 0.4963,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8326436020526403,
+      "grad_norm": 2.4915542602539062,
+      "learning_rate": 9.070965935903672e-06,
+      "loss": 0.5219,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8342989571263036,
+      "grad_norm": 2.3988428115844727,
+      "learning_rate": 9.065366756476987e-06,
+      "loss": 0.5804,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8359543121999669,
+      "grad_norm": 2.0617735385894775,
+      "learning_rate": 9.059752493922165e-06,
+      "loss": 0.5585,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8376096672736302,
+      "grad_norm": 2.474543571472168,
+      "learning_rate": 9.054123169068974e-06,
+      "loss": 0.5238,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8392650223472935,
+      "grad_norm": 1.58362877368927,
+      "learning_rate": 9.048478802803062e-06,
+      "loss": 0.5623,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8409203774209568,
+      "grad_norm": 1.7996039390563965,
+      "learning_rate": 9.042819416065888e-06,
+      "loss": 0.5107,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8425757324946201,
+      "grad_norm": 2.7114357948303223,
+      "learning_rate": 9.037145029854637e-06,
+      "loss": 0.4495,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8442310875682834,
+      "grad_norm": 2.9728622436523438,
+      "learning_rate": 9.03145566522214e-06,
+      "loss": 0.5361,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8458864426419467,
+      "grad_norm": 2.092686891555786,
+      "learning_rate": 9.025751343276806e-06,
+      "loss": 0.4656,
+      "step": 5110
+    },
+    {
+      "epoch": 0.84754179771561,
+      "grad_norm": 2.272784471511841,
+      "learning_rate": 9.020032085182535e-06,
+      "loss": 0.546,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8491971527892733,
+      "grad_norm": 2.0549049377441406,
+      "learning_rate": 9.014297912158645e-06,
+      "loss": 0.5034,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8508525078629366,
+      "grad_norm": 1.5659208297729492,
+      "learning_rate": 9.008548845479787e-06,
+      "loss": 0.57,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8525078629365999,
+      "grad_norm": 1.5937261581420898,
+      "learning_rate": 9.002784906475872e-06,
+      "loss": 0.4394,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8541632180102632,
+      "grad_norm": 1.9835437536239624,
+      "learning_rate": 8.99700611653199e-06,
+      "loss": 0.492,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8558185730839265,
+      "grad_norm": 2.044243574142456,
+      "learning_rate": 8.991212497088328e-06,
+      "loss": 0.5157,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8574739281575898,
+      "grad_norm": 2.410057544708252,
+      "learning_rate": 8.985404069640096e-06,
+      "loss": 0.557,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8591292832312531,
+      "grad_norm": 1.7656784057617188,
+      "learning_rate": 8.97958085573744e-06,
+      "loss": 0.5092,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8607846383049164,
+      "grad_norm": 2.221282958984375,
+      "learning_rate": 8.973742876985369e-06,
+      "loss": 0.5198,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8624399933785797,
+      "grad_norm": 1.847607135772705,
+      "learning_rate": 8.967890155043672e-06,
+      "loss": 0.516,
+      "step": 5210
+    },
+    {
+      "epoch": 0.864095348452243,
+      "grad_norm": 1.8600164651870728,
+      "learning_rate": 8.962022711626835e-06,
+      "loss": 0.4892,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8657507035259063,
+      "grad_norm": 1.9303275346755981,
+      "learning_rate": 8.956140568503964e-06,
+      "loss": 0.5309,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8674060585995697,
+      "grad_norm": 2.567194938659668,
+      "learning_rate": 8.950243747498704e-06,
+      "loss": 0.5249,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8690614136732329,
+      "grad_norm": 2.3800063133239746,
+      "learning_rate": 8.944332270489156e-06,
+      "loss": 0.4972,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8707167687468962,
+      "grad_norm": 1.9158269166946411,
+      "learning_rate": 8.938406159407798e-06,
+      "loss": 0.4903,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8723721238205595,
+      "grad_norm": 2.1870522499084473,
+      "learning_rate": 8.932465436241403e-06,
+      "loss": 0.5044,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8740274788942228,
+      "grad_norm": 2.8295249938964844,
+      "learning_rate": 8.926510123030955e-06,
+      "loss": 0.5025,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8756828339678862,
+      "grad_norm": 2.0139286518096924,
+      "learning_rate": 8.92054024187157e-06,
+      "loss": 0.5451,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8773381890415494,
+      "grad_norm": 1.9241939783096313,
+      "learning_rate": 8.914555814912416e-06,
+      "loss": 0.4705,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8789935441152127,
+      "grad_norm": 1.9577219486236572,
+      "learning_rate": 8.908556864356625e-06,
+      "loss": 0.5262,
+      "step": 5310
+    },
+    {
+      "epoch": 0.880648899188876,
+      "grad_norm": 2.2654361724853516,
+      "learning_rate": 8.902543412461214e-06,
+      "loss": 0.4939,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8823042542625393,
+      "grad_norm": 1.9551903009414673,
+      "learning_rate": 8.896515481537004e-06,
+      "loss": 0.5221,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8839596093362027,
+      "grad_norm": 2.27404522895813,
+      "learning_rate": 8.890473093948532e-06,
+      "loss": 0.5371,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8856149644098659,
+      "grad_norm": 2.3164634704589844,
+      "learning_rate": 8.884416272113977e-06,
+      "loss": 0.5328,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8872703194835292,
+      "grad_norm": 1.646775484085083,
+      "learning_rate": 8.878345038505067e-06,
+      "loss": 0.5262,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8889256745571925,
+      "grad_norm": 2.1168019771575928,
+      "learning_rate": 8.872259415646998e-06,
+      "loss": 0.5132,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8905810296308558,
+      "grad_norm": 1.5265601873397827,
+      "learning_rate": 8.86615942611836e-06,
+      "loss": 0.4874,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8922363847045192,
+      "grad_norm": 1.885336995124817,
+      "learning_rate": 8.860045092551034e-06,
+      "loss": 0.4704,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8938917397781824,
+      "grad_norm": 2.109318256378174,
+      "learning_rate": 8.853916437630135e-06,
+      "loss": 0.4941,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8955470948518457,
+      "grad_norm": 1.7149527072906494,
+      "learning_rate": 8.847773484093896e-06,
+      "loss": 0.5617,
+      "step": 5410
+    },
+    {
+      "epoch": 0.897202449925509,
+      "grad_norm": 2.4205265045166016,
+      "learning_rate": 8.84161625473361e-06,
+      "loss": 0.4991,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8988578049991723,
+      "grad_norm": 2.258826971054077,
+      "learning_rate": 8.835444772393534e-06,
+      "loss": 0.5105,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9005131600728357,
+      "grad_norm": 2.3182597160339355,
+      "learning_rate": 8.829259059970805e-06,
+      "loss": 0.5127,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9021685151464989,
+      "grad_norm": 1.9707034826278687,
+      "learning_rate": 8.823059140415355e-06,
+      "loss": 0.4753,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9038238702201622,
+      "grad_norm": 1.8292397260665894,
+      "learning_rate": 8.816845036729827e-06,
+      "loss": 0.4998,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9054792252938255,
+      "grad_norm": 1.737146258354187,
+      "learning_rate": 8.81061677196949e-06,
+      "loss": 0.4546,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9071345803674888,
+      "grad_norm": 1.6696151494979858,
+      "learning_rate": 8.80437436924215e-06,
+      "loss": 0.5443,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9087899354411522,
+      "grad_norm": 2.815847635269165,
+      "learning_rate": 8.798117851708072e-06,
+      "loss": 0.5335,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9104452905148154,
+      "grad_norm": 1.8941160440444946,
+      "learning_rate": 8.791847242579887e-06,
+      "loss": 0.5121,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9121006455884787,
+      "grad_norm": 2.581604242324829,
+      "learning_rate": 8.785562565122505e-06,
+      "loss": 0.5301,
+      "step": 5510
+    },
+    {
+      "epoch": 0.913756000662142,
+      "grad_norm": 1.7762013673782349,
+      "learning_rate": 8.779263842653034e-06,
+      "loss": 0.5159,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9154113557358053,
+      "grad_norm": 2.0996243953704834,
+      "learning_rate": 8.772951098540692e-06,
+      "loss": 0.5227,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9170667108094687,
+      "grad_norm": 2.195647716522217,
+      "learning_rate": 8.76662435620672e-06,
+      "loss": 0.4409,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9187220658831319,
+      "grad_norm": 3.163017749786377,
+      "learning_rate": 8.760283639124289e-06,
+      "loss": 0.5603,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9203774209567952,
+      "grad_norm": 2.1786811351776123,
+      "learning_rate": 8.753928970818426e-06,
+      "loss": 0.5127,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9220327760304585,
+      "grad_norm": 1.9021741151809692,
+      "learning_rate": 8.747560374865913e-06,
+      "loss": 0.5137,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9236881311041218,
+      "grad_norm": 2.3319544792175293,
+      "learning_rate": 8.74117787489521e-06,
+      "loss": 0.515,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9253434861777852,
+      "grad_norm": 2.084627389907837,
+      "learning_rate": 8.734781494586363e-06,
+      "loss": 0.5457,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9269988412514485,
+      "grad_norm": 2.1023311614990234,
+      "learning_rate": 8.72837125767091e-06,
+      "loss": 0.5222,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9286541963251117,
+      "grad_norm": 1.880613923072815,
+      "learning_rate": 8.721947187931807e-06,
+      "loss": 0.5315,
+      "step": 5610
+    },
+    {
+      "epoch": 0.930309551398775,
+      "grad_norm": 2.219151735305786,
+      "learning_rate": 8.715509309203327e-06,
+      "loss": 0.526,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9319649064724383,
+      "grad_norm": 2.4242360591888428,
+      "learning_rate": 8.709057645370977e-06,
+      "loss": 0.4777,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9336202615461017,
+      "grad_norm": 1.921246886253357,
+      "learning_rate": 8.702592220371413e-06,
+      "loss": 0.523,
+      "step": 5640
+    },
+    {
+      "epoch": 0.935275616619765,
+      "grad_norm": 2.206521987915039,
+      "learning_rate": 8.696113058192341e-06,
+      "loss": 0.5092,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9369309716934282,
+      "grad_norm": 1.9734609127044678,
+      "learning_rate": 8.689620182872435e-06,
+      "loss": 0.5065,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9385863267670915,
+      "grad_norm": 1.9264705181121826,
+      "learning_rate": 8.683113618501256e-06,
+      "loss": 0.5365,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9402416818407549,
+      "grad_norm": 2.4335641860961914,
+      "learning_rate": 8.676593389219137e-06,
+      "loss": 0.4719,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9418970369144182,
+      "grad_norm": 2.1883111000061035,
+      "learning_rate": 8.670059519217124e-06,
+      "loss": 0.4727,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9435523919880815,
+      "grad_norm": 2.0516345500946045,
+      "learning_rate": 8.663512032736868e-06,
+      "loss": 0.5501,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9452077470617447,
+      "grad_norm": 2.072754144668579,
+      "learning_rate": 8.656950954070536e-06,
+      "loss": 0.5249,
+      "step": 5710
+    },
+    {
+      "epoch": 0.946863102135408,
+      "grad_norm": 1.922422170639038,
+      "learning_rate": 8.650376307560726e-06,
+      "loss": 0.5304,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9485184572090714,
+      "grad_norm": 1.9621427059173584,
+      "learning_rate": 8.643788117600376e-06,
+      "loss": 0.5281,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9501738122827347,
+      "grad_norm": 2.2022109031677246,
+      "learning_rate": 8.637186408632673e-06,
+      "loss": 0.4689,
+      "step": 5740
+    },
+    {
+      "epoch": 0.951829167356398,
+      "grad_norm": 2.15291690826416,
+      "learning_rate": 8.630571205150957e-06,
+      "loss": 0.5009,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9534845224300612,
+      "grad_norm": 2.6725265979766846,
+      "learning_rate": 8.62394253169864e-06,
+      "loss": 0.4522,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9551398775037245,
+      "grad_norm": 2.1421377658843994,
+      "learning_rate": 8.617300412869105e-06,
+      "loss": 0.5002,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9567952325773879,
+      "grad_norm": 1.951465368270874,
+      "learning_rate": 8.610644873305625e-06,
+      "loss": 0.4692,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9584505876510512,
+      "grad_norm": 2.3892083168029785,
+      "learning_rate": 8.60397593770126e-06,
+      "loss": 0.485,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9601059427247145,
+      "grad_norm": 2.2727770805358887,
+      "learning_rate": 8.597293630798776e-06,
+      "loss": 0.5378,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9617612977983777,
+      "grad_norm": 1.7045542001724243,
+      "learning_rate": 8.590597977390542e-06,
+      "loss": 0.4663,
+      "step": 5810
+    },
+    {
+      "epoch": 0.963416652872041,
+      "grad_norm": 2.008007526397705,
+      "learning_rate": 8.583889002318455e-06,
+      "loss": 0.5119,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9650720079457044,
+      "grad_norm": 1.846248745918274,
+      "learning_rate": 8.57716673047383e-06,
+      "loss": 0.5105,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9667273630193677,
+      "grad_norm": 3.5718061923980713,
+      "learning_rate": 8.570431186797314e-06,
+      "loss": 0.4927,
+      "step": 5840
+    },
+    {
+      "epoch": 0.968382718093031,
+      "grad_norm": 2.1415624618530273,
+      "learning_rate": 8.563682396278799e-06,
+      "loss": 0.4978,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9700380731666942,
+      "grad_norm": 2.2250213623046875,
+      "learning_rate": 8.556920383957322e-06,
+      "loss": 0.4732,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9716934282403575,
+      "grad_norm": 2.6127257347106934,
+      "learning_rate": 8.550145174920977e-06,
+      "loss": 0.4866,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9733487833140209,
+      "grad_norm": 2.1522982120513916,
+      "learning_rate": 8.543356794306818e-06,
+      "loss": 0.5143,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9750041383876842,
+      "grad_norm": 2.036477565765381,
+      "learning_rate": 8.536555267300766e-06,
+      "loss": 0.5438,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9766594934613475,
+      "grad_norm": 2.8033342361450195,
+      "learning_rate": 8.529740619137523e-06,
+      "loss": 0.514,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9783148485350107,
+      "grad_norm": 2.0152909755706787,
+      "learning_rate": 8.522912875100467e-06,
+      "loss": 0.5071,
+      "step": 5910
+    },
+    {
+      "epoch": 0.979970203608674,
+      "grad_norm": 2.0931289196014404,
+      "learning_rate": 8.516072060521566e-06,
+      "loss": 0.5031,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9816255586823374,
+      "grad_norm": 1.396856427192688,
+      "learning_rate": 8.509218200781278e-06,
+      "loss": 0.5602,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9832809137560007,
+      "grad_norm": 2.251620054244995,
+      "learning_rate": 8.502351321308468e-06,
+      "loss": 0.5185,
+      "step": 5940
+    },
+    {
+      "epoch": 0.984936268829664,
+      "grad_norm": 2.3058903217315674,
+      "learning_rate": 8.4954714475803e-06,
+      "loss": 0.4617,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9865916239033272,
+      "grad_norm": 2.428722620010376,
+      "learning_rate": 8.488578605122149e-06,
+      "loss": 0.5123,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9882469789769905,
+      "grad_norm": 2.9448606967926025,
+      "learning_rate": 8.48167281950751e-06,
+      "loss": 0.4912,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9899023340506539,
+      "grad_norm": 1.9352291822433472,
+      "learning_rate": 8.474754116357895e-06,
+      "loss": 0.4906,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9915576891243172,
+      "grad_norm": 2.116455078125,
+      "learning_rate": 8.467822521342744e-06,
+      "loss": 0.4532,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9932130441979805,
+      "grad_norm": 1.96929931640625,
+      "learning_rate": 8.460878060179326e-06,
+      "loss": 0.4909,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9948683992716437,
+      "grad_norm": 2.103734254837036,
+      "learning_rate": 8.45392075863265e-06,
+      "loss": 0.4986,
+      "step": 6010
+    },
+    {
+      "epoch": 0.996523754345307,
+      "grad_norm": 1.8520708084106445,
+      "learning_rate": 8.446950642515359e-06,
+      "loss": 0.4959,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9981791094189704,
+      "grad_norm": 1.5398201942443848,
+      "learning_rate": 8.439967737687642e-06,
+      "loss": 0.5061,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9998344644926337,
+      "grad_norm": 2.7384068965911865,
+      "learning_rate": 8.432972070057137e-06,
+      "loss": 0.4164,
+      "step": 6040
+    },
+    {
+      "epoch": 1.001489819566297,
+      "grad_norm": 1.699214220046997,
+      "learning_rate": 8.425963665578833e-06,
+      "loss": 0.4717,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0031451746399602,
+      "grad_norm": 2.7517991065979004,
+      "learning_rate": 8.418942550254978e-06,
+      "loss": 0.4783,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0048005297136235,
+      "grad_norm": 2.293407440185547,
+      "learning_rate": 8.411908750134973e-06,
+      "loss": 0.428,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0064558847872869,
+      "grad_norm": 2.3699285984039307,
+      "learning_rate": 8.404862291315287e-06,
+      "loss": 0.4529,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0081112398609502,
+      "grad_norm": 2.27496337890625,
+      "learning_rate": 8.39780319993935e-06,
+      "loss": 0.4486,
+      "step": 6090
+    },
+    {
+      "epoch": 1.0097665949346135,
+      "grad_norm": 2.0127158164978027,
+      "learning_rate": 8.390731502197465e-06,
+      "loss": 0.4529,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0114219500082768,
+      "grad_norm": 2.6068265438079834,
+      "learning_rate": 8.383647224326704e-06,
+      "loss": 0.4397,
+      "step": 6110
+    },
+    {
+      "epoch": 1.0130773050819402,
+      "grad_norm": 1.9427989721298218,
+      "learning_rate": 8.376550392610813e-06,
+      "loss": 0.4236,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0147326601556035,
+      "grad_norm": 2.082646369934082,
+      "learning_rate": 8.369441033380119e-06,
+      "loss": 0.4352,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0163880152292666,
+      "grad_norm": 2.0163888931274414,
+      "learning_rate": 8.362319173011421e-06,
+      "loss": 0.4359,
+      "step": 6140
+    },
+    {
+      "epoch": 1.01804337030293,
+      "grad_norm": 1.7320841550827026,
+      "learning_rate": 8.355184837927906e-06,
+      "loss": 0.4454,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0196987253765932,
+      "grad_norm": 1.9709644317626953,
+      "learning_rate": 8.348038054599037e-06,
+      "loss": 0.4593,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0213540804502566,
+      "grad_norm": 2.382404327392578,
+      "learning_rate": 8.340878849540466e-06,
+      "loss": 0.3941,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0230094355239199,
+      "grad_norm": 2.566659450531006,
+      "learning_rate": 8.333707249313933e-06,
+      "loss": 0.4474,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0246647905975832,
+      "grad_norm": 1.6790319681167603,
+      "learning_rate": 8.326523280527165e-06,
+      "loss": 0.437,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0263201456712465,
+      "grad_norm": 1.9605789184570312,
+      "learning_rate": 8.319326969833776e-06,
+      "loss": 0.4346,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0279755007449098,
+      "grad_norm": 1.5337015390396118,
+      "learning_rate": 8.312118343933172e-06,
+      "loss": 0.4329,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0296308558185732,
+      "grad_norm": 2.0103812217712402,
+      "learning_rate": 8.304897429570448e-06,
+      "loss": 0.4199,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0312862108922365,
+      "grad_norm": 3.106372594833374,
+      "learning_rate": 8.297664253536296e-06,
+      "loss": 0.4429,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0329415659658996,
+      "grad_norm": 2.3438146114349365,
+      "learning_rate": 8.290418842666894e-06,
+      "loss": 0.4588,
+      "step": 6240
+    },
+    {
+      "epoch": 1.034596921039563,
+      "grad_norm": 1.6596814393997192,
+      "learning_rate": 8.28316122384382e-06,
+      "loss": 0.4314,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0362522761132262,
+      "grad_norm": 2.1829092502593994,
+      "learning_rate": 8.275891423993943e-06,
+      "loss": 0.4292,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0379076311868896,
+      "grad_norm": 1.525658130645752,
+      "learning_rate": 8.268609470089322e-06,
+      "loss": 0.4206,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0395629862605529,
+      "grad_norm": 1.846199870109558,
+      "learning_rate": 8.261315389147113e-06,
+      "loss": 0.3781,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0412183413342162,
+      "grad_norm": 2.2340259552001953,
+      "learning_rate": 8.254009208229464e-06,
+      "loss": 0.4107,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0428736964078795,
+      "grad_norm": 2.0799405574798584,
+      "learning_rate": 8.246690954443416e-06,
+      "loss": 0.3953,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0445290514815428,
+      "grad_norm": 2.073094367980957,
+      "learning_rate": 8.239360654940803e-06,
+      "loss": 0.4502,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0461844065552062,
+      "grad_norm": 2.092500925064087,
+      "learning_rate": 8.232018336918145e-06,
+      "loss": 0.4067,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0478397616288695,
+      "grad_norm": 2.2009036540985107,
+      "learning_rate": 8.224664027616565e-06,
+      "loss": 0.4153,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0494951167025326,
+      "grad_norm": 2.1179239749908447,
+      "learning_rate": 8.217297754321661e-06,
+      "loss": 0.3964,
+      "step": 6340
+    },
+    {
+      "epoch": 1.051150471776196,
+      "grad_norm": 1.9350873231887817,
+      "learning_rate": 8.209919544363428e-06,
+      "loss": 0.4776,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0528058268498592,
+      "grad_norm": 2.0609846115112305,
+      "learning_rate": 8.202529425116145e-06,
+      "loss": 0.4193,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0544611819235226,
+      "grad_norm": 2.755305528640747,
+      "learning_rate": 8.195127423998279e-06,
+      "loss": 0.4682,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0561165369971859,
+      "grad_norm": 1.6468199491500854,
+      "learning_rate": 8.187713568472375e-06,
+      "loss": 0.4609,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0577718920708492,
+      "grad_norm": 2.1090316772460938,
+      "learning_rate": 8.180287886044967e-06,
+      "loss": 0.4468,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0594272471445125,
+      "grad_norm": 3.3360707759857178,
+      "learning_rate": 8.172850404266462e-06,
+      "loss": 0.4611,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0610826022181759,
+      "grad_norm": 1.8891501426696777,
+      "learning_rate": 8.165401150731045e-06,
+      "loss": 0.42,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0627379572918392,
+      "grad_norm": 2.2766528129577637,
+      "learning_rate": 8.157940153076582e-06,
+      "loss": 0.3972,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0643933123655025,
+      "grad_norm": 2.1425061225891113,
+      "learning_rate": 8.150467438984507e-06,
+      "loss": 0.4378,
+      "step": 6430
+    },
+    {
+      "epoch": 1.0660486674391656,
+      "grad_norm": 2.0477399826049805,
+      "learning_rate": 8.142983036179723e-06,
+      "loss": 0.4627,
+      "step": 6440
+    },
+    {
+      "epoch": 1.067704022512829,
+      "grad_norm": 1.6100428104400635,
+      "learning_rate": 8.135486972430502e-06,
+      "loss": 0.415,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0693593775864922,
+      "grad_norm": 2.3228981494903564,
+      "learning_rate": 8.127979275548376e-06,
+      "loss": 0.4399,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0710147326601556,
+      "grad_norm": 2.075024127960205,
+      "learning_rate": 8.120459973388046e-06,
+      "loss": 0.4679,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0726700877338189,
+      "grad_norm": 2.06042218208313,
+      "learning_rate": 8.112929093847262e-06,
+      "loss": 0.3853,
+      "step": 6480
+    },
+    {
+      "epoch": 1.0743254428074822,
+      "grad_norm": 2.618030071258545,
+      "learning_rate": 8.105386664866732e-06,
+      "loss": 0.4163,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0759807978811455,
+      "grad_norm": 2.011997938156128,
+      "learning_rate": 8.09783271443001e-06,
+      "loss": 0.4632,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0776361529548089,
+      "grad_norm": 1.7011245489120483,
+      "learning_rate": 8.090267270563403e-06,
+      "loss": 0.422,
+      "step": 6510
+    },
+    {
+      "epoch": 1.0792915080284722,
+      "grad_norm": 2.1394636631011963,
+      "learning_rate": 8.082690361335857e-06,
+      "loss": 0.4856,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0809468631021355,
+      "grad_norm": 1.878475308418274,
+      "learning_rate": 8.075102014858854e-06,
+      "loss": 0.4819,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0826022181757988,
+      "grad_norm": 2.2121503353118896,
+      "learning_rate": 8.067502259286313e-06,
+      "loss": 0.4312,
+      "step": 6540
+    },
+    {
+      "epoch": 1.084257573249462,
+      "grad_norm": 2.6810719966888428,
+      "learning_rate": 8.059891122814481e-06,
+      "loss": 0.4271,
+      "step": 6550
+    },
+    {
+      "epoch": 1.0859129283231252,
+      "grad_norm": 2.5359325408935547,
+      "learning_rate": 8.05226863368183e-06,
+      "loss": 0.4325,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0875682833967886,
+      "grad_norm": 2.2482380867004395,
+      "learning_rate": 8.044634820168954e-06,
+      "loss": 0.4763,
+      "step": 6570
+    },
+    {
+      "epoch": 1.089223638470452,
+      "grad_norm": 1.8877545595169067,
+      "learning_rate": 8.036989710598458e-06,
+      "loss": 0.4688,
+      "step": 6580
+    },
+    {
+      "epoch": 1.0908789935441152,
+      "grad_norm": 2.3584036827087402,
+      "learning_rate": 8.029333333334863e-06,
+      "loss": 0.4688,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0925343486177785,
+      "grad_norm": 2.0854368209838867,
+      "learning_rate": 8.02166571678449e-06,
+      "loss": 0.4252,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0941897036914419,
+      "grad_norm": 1.9038151502609253,
+      "learning_rate": 8.01398688939536e-06,
+      "loss": 0.4127,
+      "step": 6610
+    },
+    {
+      "epoch": 1.0958450587651052,
+      "grad_norm": 1.8824822902679443,
+      "learning_rate": 8.00629687965709e-06,
+      "loss": 0.4672,
+      "step": 6620
+    },
+    {
+      "epoch": 1.0975004138387685,
+      "grad_norm": 2.430506467819214,
+      "learning_rate": 7.998595716100783e-06,
+      "loss": 0.4449,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0991557689124316,
+      "grad_norm": 2.1402385234832764,
+      "learning_rate": 7.990883427298927e-06,
+      "loss": 0.4782,
+      "step": 6640
+    },
+    {
+      "epoch": 1.100811123986095,
+      "grad_norm": 2.3543331623077393,
+      "learning_rate": 7.983160041865285e-06,
+      "loss": 0.4212,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1024664790597583,
+      "grad_norm": 2.142435312271118,
+      "learning_rate": 7.975425588454788e-06,
+      "loss": 0.4939,
+      "step": 6660
+    },
+    {
+      "epoch": 1.1041218341334216,
+      "grad_norm": 2.1134002208709717,
+      "learning_rate": 7.967680095763434e-06,
+      "loss": 0.4463,
+      "step": 6670
+    },
+    {
+      "epoch": 1.105777189207085,
+      "grad_norm": 2.716151237487793,
+      "learning_rate": 7.959923592528177e-06,
+      "loss": 0.4561,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1074325442807482,
+      "grad_norm": 1.9613409042358398,
+      "learning_rate": 7.952156107526826e-06,
+      "loss": 0.4141,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1090878993544115,
+      "grad_norm": 1.8834747076034546,
+      "learning_rate": 7.944377669577924e-06,
+      "loss": 0.4361,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1107432544280749,
+      "grad_norm": 1.4934910535812378,
+      "learning_rate": 7.93658830754066e-06,
+      "loss": 0.4413,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1123986095017382,
+      "grad_norm": 1.9365501403808594,
+      "learning_rate": 7.928788050314751e-06,
+      "loss": 0.4656,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1140539645754015,
+      "grad_norm": 1.9965449571609497,
+      "learning_rate": 7.920976926840334e-06,
+      "loss": 0.4671,
+      "step": 6730
+    },
+    {
+      "epoch": 1.1157093196490648,
+      "grad_norm": 1.9242488145828247,
+      "learning_rate": 7.913154966097865e-06,
+      "loss": 0.4345,
+      "step": 6740
+    },
+    {
+      "epoch": 1.117364674722728,
+      "grad_norm": 1.4173060655593872,
+      "learning_rate": 7.905322197108006e-06,
+      "loss": 0.4674,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1190200297963913,
+      "grad_norm": 1.9419013261795044,
+      "learning_rate": 7.897478648931521e-06,
+      "loss": 0.4157,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1206753848700546,
+      "grad_norm": 1.910142421722412,
+      "learning_rate": 7.889624350669162e-06,
+      "loss": 0.4764,
+      "step": 6770
+    },
+    {
+      "epoch": 1.122330739943718,
+      "grad_norm": 1.9535759687423706,
+      "learning_rate": 7.88175933146157e-06,
+      "loss": 0.4167,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1239860950173812,
+      "grad_norm": 2.105318784713745,
+      "learning_rate": 7.873883620489164e-06,
+      "loss": 0.4169,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1256414500910445,
+      "grad_norm": 2.383732318878174,
+      "learning_rate": 7.865997246972023e-06,
+      "loss": 0.4533,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1272968051647079,
+      "grad_norm": 1.8368053436279297,
+      "learning_rate": 7.858100240169792e-06,
+      "loss": 0.4594,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1289521602383712,
+      "grad_norm": 1.8236784934997559,
+      "learning_rate": 7.850192629381568e-06,
+      "loss": 0.4501,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1306075153120345,
+      "grad_norm": 1.8940480947494507,
+      "learning_rate": 7.842274443945785e-06,
+      "loss": 0.3969,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1322628703856976,
+      "grad_norm": 2.2269837856292725,
+      "learning_rate": 7.834345713240114e-06,
+      "loss": 0.4762,
+      "step": 6840
+    },
+    {
+      "epoch": 1.133918225459361,
+      "grad_norm": 2.336608409881592,
+      "learning_rate": 7.826406466681354e-06,
+      "loss": 0.4974,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1355735805330243,
+      "grad_norm": 2.0093603134155273,
+      "learning_rate": 7.81845673372531e-06,
+      "loss": 0.4577,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1372289356066876,
+      "grad_norm": 1.781472086906433,
+      "learning_rate": 7.810496543866704e-06,
+      "loss": 0.434,
+      "step": 6870
+    },
+    {
+      "epoch": 1.138884290680351,
+      "grad_norm": 3.0077342987060547,
+      "learning_rate": 7.802525926639045e-06,
+      "loss": 0.4651,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1405396457540142,
+      "grad_norm": 2.0327229499816895,
+      "learning_rate": 7.794544911614537e-06,
+      "loss": 0.4405,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1421950008276776,
+      "grad_norm": 1.7821011543273926,
+      "learning_rate": 7.786553528403954e-06,
+      "loss": 0.4293,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1438503559013409,
+      "grad_norm": 2.4968810081481934,
+      "learning_rate": 7.778551806656546e-06,
+      "loss": 0.4273,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1455057109750042,
+      "grad_norm": 2.041292905807495,
+      "learning_rate": 7.770539776059914e-06,
+      "loss": 0.4605,
+      "step": 6920
+    },
+    {
+      "epoch": 1.1471610660486675,
+      "grad_norm": 2.0464673042297363,
+      "learning_rate": 7.762517466339905e-06,
+      "loss": 0.4407,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1488164211223308,
+      "grad_norm": 1.8867651224136353,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.4575,
+      "step": 6940
+    },
+    {
+      "epoch": 1.150471776195994,
+      "grad_norm": 1.7853606939315796,
+      "learning_rate": 7.74644212862375e-06,
+      "loss": 0.4158,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1521271312696573,
+      "grad_norm": 1.7646677494049072,
+      "learning_rate": 7.738389160269542e-06,
+      "loss": 0.4148,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1537824863433206,
+      "grad_norm": 1.729858636856079,
+      "learning_rate": 7.73032603207563e-06,
+      "loss": 0.4153,
+      "step": 6970
+    },
+    {
+      "epoch": 1.155437841416984,
+      "grad_norm": 2.1216416358947754,
+      "learning_rate": 7.722252773957442e-06,
+      "loss": 0.4578,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1570931964906472,
+      "grad_norm": 2.449660062789917,
+      "learning_rate": 7.714169415867991e-06,
+      "loss": 0.4396,
+      "step": 6990
+    },
+    {
+      "epoch": 1.1587485515643106,
+      "grad_norm": 2.7764759063720703,
+      "learning_rate": 7.706075987797767e-06,
+      "loss": 0.4814,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1604039066379739,
+      "grad_norm": 1.6688860654830933,
+      "learning_rate": 7.697972519774612e-06,
+      "loss": 0.3879,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1620592617116372,
+      "grad_norm": 2.147960662841797,
+      "learning_rate": 7.689859041863628e-06,
+      "loss": 0.4821,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1637146167853005,
+      "grad_norm": 1.482928991317749,
+      "learning_rate": 7.681735584167048e-06,
+      "loss": 0.4695,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1653699718589636,
+      "grad_norm": 3.1044204235076904,
+      "learning_rate": 7.673602176824134e-06,
+      "loss": 0.4205,
+      "step": 7040
+    },
+    {
+      "epoch": 1.167025326932627,
+      "grad_norm": 1.5646145343780518,
+      "learning_rate": 7.665458850011062e-06,
+      "loss": 0.4767,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1686806820062903,
+      "grad_norm": 2.1758921146392822,
+      "learning_rate": 7.657305633940816e-06,
+      "loss": 0.4152,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1703360370799536,
+      "grad_norm": 2.1989285945892334,
+      "learning_rate": 7.649142558863056e-06,
+      "loss": 0.4307,
+      "step": 7070
+    },
+    {
+      "epoch": 1.171991392153617,
+      "grad_norm": 2.195272207260132,
+      "learning_rate": 7.640969655064042e-06,
+      "loss": 0.4449,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1736467472272802,
+      "grad_norm": 2.0945467948913574,
+      "learning_rate": 7.63278695286648e-06,
+      "loss": 0.4225,
+      "step": 7090
+    },
+    {
+      "epoch": 1.1753021023009436,
+      "grad_norm": 1.938080906867981,
+      "learning_rate": 7.624594482629442e-06,
+      "loss": 0.4001,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1769574573746069,
+      "grad_norm": 1.8821436166763306,
+      "learning_rate": 7.616392274748235e-06,
+      "loss": 0.3847,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1786128124482702,
+      "grad_norm": 2.344076633453369,
+      "learning_rate": 7.608180359654298e-06,
+      "loss": 0.4675,
+      "step": 7120
+    },
+    {
+      "epoch": 1.1802681675219335,
+      "grad_norm": 1.977381944656372,
+      "learning_rate": 7.599958767815081e-06,
+      "loss": 0.4387,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1819235225955969,
+      "grad_norm": 2.3042211532592773,
+      "learning_rate": 7.591727529733941e-06,
+      "loss": 0.459,
+      "step": 7140
+    },
+    {
+      "epoch": 1.1835788776692602,
+      "grad_norm": 1.93119478225708,
+      "learning_rate": 7.583486675950021e-06,
+      "loss": 0.4625,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1852342327429233,
+      "grad_norm": 1.6517666578292847,
+      "learning_rate": 7.575236237038136e-06,
+      "loss": 0.474,
+      "step": 7160
+    },
+    {
+      "epoch": 1.1868895878165866,
+      "grad_norm": 1.7987353801727295,
+      "learning_rate": 7.566976243608673e-06,
+      "loss": 0.4182,
+      "step": 7170
+    },
+    {
+      "epoch": 1.18854494289025,
+      "grad_norm": 2.2666845321655273,
+      "learning_rate": 7.558706726307459e-06,
+      "loss": 0.43,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1902002979639132,
+      "grad_norm": 1.540936827659607,
+      "learning_rate": 7.55042771581566e-06,
+      "loss": 0.4305,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1918556530375766,
+      "grad_norm": 1.6909763813018799,
+      "learning_rate": 7.542139242849664e-06,
+      "loss": 0.4059,
+      "step": 7200
+    },
+    {
+      "epoch": 1.19351100811124,
+      "grad_norm": 2.2971370220184326,
+      "learning_rate": 7.533841338160963e-06,
+      "loss": 0.4068,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1951663631849032,
+      "grad_norm": 1.584033727645874,
+      "learning_rate": 7.525534032536044e-06,
+      "loss": 0.4416,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1968217182585665,
+      "grad_norm": 1.8350242376327515,
+      "learning_rate": 7.517217356796272e-06,
+      "loss": 0.4115,
+      "step": 7230
+    },
+    {
+      "epoch": 1.1984770733322299,
+      "grad_norm": 1.7758424282073975,
+      "learning_rate": 7.508891341797777e-06,
+      "loss": 0.4243,
+      "step": 7240
+    },
+    {
+      "epoch": 1.200132428405893,
+      "grad_norm": 1.4368244409561157,
+      "learning_rate": 7.500556018431342e-06,
+      "loss": 0.4006,
+      "step": 7250
+    },
+    {
+      "epoch": 1.2017877834795563,
+      "grad_norm": 2.0176913738250732,
+      "learning_rate": 7.492211417622278e-06,
+      "loss": 0.4484,
+      "step": 7260
+    },
+    {
+      "epoch": 1.2034431385532196,
+      "grad_norm": 2.6586618423461914,
+      "learning_rate": 7.483857570330326e-06,
+      "loss": 0.4808,
+      "step": 7270
+    },
+    {
+      "epoch": 1.205098493626883,
+      "grad_norm": 2.3201072216033936,
+      "learning_rate": 7.475494507549526e-06,
+      "loss": 0.4636,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2067538487005462,
+      "grad_norm": 2.491241693496704,
+      "learning_rate": 7.4671222603081115e-06,
+      "loss": 0.454,
+      "step": 7290
+    },
+    {
+      "epoch": 1.2084092037742096,
+      "grad_norm": 2.3110857009887695,
+      "learning_rate": 7.458740859668391e-06,
+      "loss": 0.4583,
+      "step": 7300
+    },
+    {
+      "epoch": 1.210064558847873,
+      "grad_norm": 2.2038140296936035,
+      "learning_rate": 7.450350336726635e-06,
+      "loss": 0.4352,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2117199139215362,
+      "grad_norm": 1.6321648359298706,
+      "learning_rate": 7.441950722612957e-06,
+      "loss": 0.4219,
+      "step": 7320
+    },
+    {
+      "epoch": 1.2133752689951995,
+      "grad_norm": 1.9630392789840698,
+      "learning_rate": 7.433542048491201e-06,
+      "loss": 0.4495,
+      "step": 7330
+    },
+    {
+      "epoch": 1.2150306240688629,
+      "grad_norm": 2.2168731689453125,
+      "learning_rate": 7.4251243455588266e-06,
+      "loss": 0.4646,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2166859791425262,
+      "grad_norm": 2.295635938644409,
+      "learning_rate": 7.416697645046789e-06,
+      "loss": 0.4543,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2183413342161893,
+      "grad_norm": 2.0987188816070557,
+      "learning_rate": 7.408261978219426e-06,
+      "loss": 0.464,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2199966892898526,
+      "grad_norm": 2.4366235733032227,
+      "learning_rate": 7.399817376374346e-06,
+      "loss": 0.4494,
+      "step": 7370
+    },
+    {
+      "epoch": 1.221652044363516,
+      "grad_norm": 2.054619073867798,
+      "learning_rate": 7.391363870842299e-06,
+      "loss": 0.4304,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2233073994371793,
+      "grad_norm": 1.9227901697158813,
+      "learning_rate": 7.3829014929870805e-06,
+      "loss": 0.4561,
+      "step": 7390
+    },
+    {
+      "epoch": 1.2249627545108426,
+      "grad_norm": 1.8692800998687744,
+      "learning_rate": 7.374430274205395e-06,
+      "loss": 0.46,
+      "step": 7400
+    },
+    {
+      "epoch": 1.226618109584506,
+      "grad_norm": 2.2331535816192627,
+      "learning_rate": 7.3659502459267516e-06,
+      "loss": 0.3858,
+      "step": 7410
+    },
+    {
+      "epoch": 1.2282734646581692,
+      "grad_norm": 2.1220543384552,
+      "learning_rate": 7.357461439613341e-06,
+      "loss": 0.4511,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2299288197318325,
+      "grad_norm": 1.9789310693740845,
+      "learning_rate": 7.348963886759926e-06,
+      "loss": 0.4326,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2315841748054959,
+      "grad_norm": 1.7887349128723145,
+      "learning_rate": 7.340457618893717e-06,
+      "loss": 0.4375,
+      "step": 7440
+    },
+    {
+      "epoch": 1.233239529879159,
+      "grad_norm": 2.1581339836120605,
+      "learning_rate": 7.331942667574262e-06,
+      "loss": 0.4178,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2348948849528223,
+      "grad_norm": 2.1266531944274902,
+      "learning_rate": 7.323419064393321e-06,
+      "loss": 0.4299,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2365502400264856,
+      "grad_norm": 1.832108974456787,
+      "learning_rate": 7.3148868409747585e-06,
+      "loss": 0.429,
+      "step": 7470
+    },
+    {
+      "epoch": 1.238205595100149,
+      "grad_norm": 1.735487937927246,
+      "learning_rate": 7.306346028974418e-06,
+      "loss": 0.4511,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2398609501738123,
+      "grad_norm": 2.0371644496917725,
+      "learning_rate": 7.297796660080011e-06,
+      "loss": 0.471,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2415163052474756,
+      "grad_norm": 2.0136220455169678,
+      "learning_rate": 7.289238766010992e-06,
+      "loss": 0.507,
+      "step": 7500
+    },
+    {
+      "epoch": 1.243171660321139,
+      "grad_norm": 1.9500830173492432,
+      "learning_rate": 7.280672378518449e-06,
+      "loss": 0.461,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2448270153948022,
+      "grad_norm": 2.5668957233428955,
+      "learning_rate": 7.2720975293849824e-06,
+      "loss": 0.4572,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2464823704684656,
+      "grad_norm": 1.870226263999939,
+      "learning_rate": 7.263514250424582e-06,
+      "loss": 0.4433,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2481377255421289,
+      "grad_norm": 1.7757699489593506,
+      "learning_rate": 7.254922573482518e-06,
+      "loss": 0.4711,
+      "step": 7540
+    },
+    {
+      "epoch": 1.2497930806157922,
+      "grad_norm": 2.230046510696411,
+      "learning_rate": 7.246322530435217e-06,
+      "loss": 0.418,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2514484356894555,
+      "grad_norm": 1.9823356866836548,
+      "learning_rate": 7.237714153190143e-06,
+      "loss": 0.4646,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2531037907631186,
+      "grad_norm": 2.5432960987091064,
+      "learning_rate": 7.229097473685686e-06,
+      "loss": 0.457,
+      "step": 7570
+    },
+    {
+      "epoch": 1.254759145836782,
+      "grad_norm": 1.8689072132110596,
+      "learning_rate": 7.220472523891035e-06,
+      "loss": 0.4398,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2564145009104453,
+      "grad_norm": 2.0174477100372314,
+      "learning_rate": 7.211839335806061e-06,
+      "loss": 0.4594,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2580698559841086,
+      "grad_norm": 2.0547597408294678,
+      "learning_rate": 7.203197941461206e-06,
+      "loss": 0.3943,
+      "step": 7600
+    },
+    {
+      "epoch": 1.259725211057772,
+      "grad_norm": 1.7089883089065552,
+      "learning_rate": 7.194548372917356e-06,
+      "loss": 0.4635,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2613805661314352,
+      "grad_norm": 1.6787538528442383,
+      "learning_rate": 7.185890662265721e-06,
+      "loss": 0.3988,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2630359212050986,
+      "grad_norm": 2.023766279220581,
+      "learning_rate": 7.177224841627724e-06,
+      "loss": 0.4548,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2646912762787617,
+      "grad_norm": 1.8672133684158325,
+      "learning_rate": 7.168550943154877e-06,
+      "loss": 0.4203,
+      "step": 7640
+    },
+    {
+      "epoch": 1.266346631352425,
+      "grad_norm": 2.4731225967407227,
+      "learning_rate": 7.159868999028658e-06,
+      "loss": 0.4652,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2680019864260883,
+      "grad_norm": 1.876103401184082,
+      "learning_rate": 7.151179041460402e-06,
+      "loss": 0.4101,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2696573414997516,
+      "grad_norm": 1.7336941957473755,
+      "learning_rate": 7.142481102691167e-06,
+      "loss": 0.3927,
+      "step": 7670
+    },
+    {
+      "epoch": 1.271312696573415,
+      "grad_norm": 1.5859344005584717,
+      "learning_rate": 7.133775214991632e-06,
+      "loss": 0.4542,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2729680516470783,
+      "grad_norm": 1.7735705375671387,
+      "learning_rate": 7.125061410661959e-06,
+      "loss": 0.4417,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2746234067207416,
+      "grad_norm": 2.3329756259918213,
+      "learning_rate": 7.1163397220316865e-06,
+      "loss": 0.4277,
+      "step": 7700
+    },
+    {
+      "epoch": 1.276278761794405,
+      "grad_norm": 2.6831562519073486,
+      "learning_rate": 7.107610181459603e-06,
+      "loss": 0.3803,
+      "step": 7710
+    },
+    {
+      "epoch": 1.2779341168680682,
+      "grad_norm": 1.7790484428405762,
+      "learning_rate": 7.098872821333633e-06,
+      "loss": 0.4598,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2795894719417316,
+      "grad_norm": 1.8431192636489868,
+      "learning_rate": 7.090127674070707e-06,
+      "loss": 0.426,
+      "step": 7730
+    },
+    {
+      "epoch": 1.2812448270153949,
+      "grad_norm": 2.489999532699585,
+      "learning_rate": 7.081374772116652e-06,
+      "loss": 0.4192,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2829001820890582,
+      "grad_norm": 1.8376771211624146,
+      "learning_rate": 7.07261414794606e-06,
+      "loss": 0.4184,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2845555371627215,
+      "grad_norm": 1.8397380113601685,
+      "learning_rate": 7.063845834062178e-06,
+      "loss": 0.4131,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2862108922363846,
+      "grad_norm": 1.7695285081863403,
+      "learning_rate": 7.055069862996786e-06,
+      "loss": 0.424,
+      "step": 7770
+    },
+    {
+      "epoch": 1.287866247310048,
+      "grad_norm": 1.8908298015594482,
+      "learning_rate": 7.0462862673100675e-06,
+      "loss": 0.4538,
+      "step": 7780
+    },
+    {
+      "epoch": 1.2895216023837113,
+      "grad_norm": 2.012000799179077,
+      "learning_rate": 7.037495079590494e-06,
+      "loss": 0.4166,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2911769574573746,
+      "grad_norm": 2.016519784927368,
+      "learning_rate": 7.028696332454712e-06,
+      "loss": 0.4505,
+      "step": 7800
+    },
+    {
+      "epoch": 1.292832312531038,
+      "grad_norm": 1.9093087911605835,
+      "learning_rate": 7.0198900585474065e-06,
+      "loss": 0.4142,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2944876676047012,
+      "grad_norm": 1.9886523485183716,
+      "learning_rate": 7.01107629054119e-06,
+      "loss": 0.4588,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2961430226783646,
+      "grad_norm": 2.125847101211548,
+      "learning_rate": 7.0022550611364835e-06,
+      "loss": 0.4451,
+      "step": 7830
+    },
+    {
+      "epoch": 1.2977983777520279,
+      "grad_norm": 2.2600979804992676,
+      "learning_rate": 6.993426403061389e-06,
+      "loss": 0.4324,
+      "step": 7840
+    },
+    {
+      "epoch": 1.299453732825691,
+      "grad_norm": 2.3956971168518066,
+      "learning_rate": 6.984590349071564e-06,
+      "loss": 0.4473,
+      "step": 7850
+    },
+    {
+      "epoch": 1.3011090878993543,
+      "grad_norm": 2.047929286956787,
+      "learning_rate": 6.975746931950116e-06,
+      "loss": 0.4193,
+      "step": 7860
+    },
+    {
+      "epoch": 1.3027644429730176,
+      "grad_norm": 1.9703757762908936,
+      "learning_rate": 6.9668961845074615e-06,
+      "loss": 0.4446,
+      "step": 7870
+    },
+    {
+      "epoch": 1.304419798046681,
+      "grad_norm": 2.0574028491973877,
+      "learning_rate": 6.95803813958122e-06,
+      "loss": 0.4543,
+      "step": 7880
+    },
+    {
+      "epoch": 1.3060751531203443,
+      "grad_norm": 1.9040168523788452,
+      "learning_rate": 6.949172830036084e-06,
+      "loss": 0.3936,
+      "step": 7890
+    },
+    {
+      "epoch": 1.3077305081940076,
+      "grad_norm": 2.8889362812042236,
+      "learning_rate": 6.940300288763697e-06,
+      "loss": 0.4244,
+      "step": 7900
+    },
+    {
+      "epoch": 1.309385863267671,
+      "grad_norm": 2.0269064903259277,
+      "learning_rate": 6.931420548682535e-06,
+      "loss": 0.4036,
+      "step": 7910
+    },
+    {
+      "epoch": 1.3110412183413342,
+      "grad_norm": 3.0732882022857666,
+      "learning_rate": 6.9225336427377835e-06,
+      "loss": 0.459,
+      "step": 7920
+    },
+    {
+      "epoch": 1.3126965734149976,
+      "grad_norm": 2.3372762203216553,
+      "learning_rate": 6.9136396039012125e-06,
+      "loss": 0.4699,
+      "step": 7930
+    },
+    {
+      "epoch": 1.314351928488661,
+      "grad_norm": 1.708390474319458,
+      "learning_rate": 6.904738465171058e-06,
+      "loss": 0.412,
+      "step": 7940
+    },
+    {
+      "epoch": 1.3160072835623242,
+      "grad_norm": 1.5611287355422974,
+      "learning_rate": 6.895830259571894e-06,
+      "loss": 0.4658,
+      "step": 7950
+    },
+    {
+      "epoch": 1.3176626386359875,
+      "grad_norm": 2.084836721420288,
+      "learning_rate": 6.886915020154519e-06,
+      "loss": 0.4428,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3193179937096506,
+      "grad_norm": 4.934893608093262,
+      "learning_rate": 6.877992779995825e-06,
+      "loss": 0.4382,
+      "step": 7970
+    },
+    {
+      "epoch": 1.320973348783314,
+      "grad_norm": 1.7433263063430786,
+      "learning_rate": 6.869063572198678e-06,
+      "loss": 0.4303,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3226287038569773,
+      "grad_norm": 2.039020538330078,
+      "learning_rate": 6.860127429891792e-06,
+      "loss": 0.4334,
+      "step": 7990
+    },
+    {
+      "epoch": 1.3242840589306406,
+      "grad_norm": 2.0853657722473145,
+      "learning_rate": 6.851184386229617e-06,
+      "loss": 0.4279,
+      "step": 8000
+    },
+    {
+      "epoch": 1.325939414004304,
+      "grad_norm": 2.357513904571533,
+      "learning_rate": 6.842234474392201e-06,
+      "loss": 0.4589,
+      "step": 8010
+    },
+    {
+      "epoch": 1.3275947690779673,
+      "grad_norm": 1.799069881439209,
+      "learning_rate": 6.833277727585076e-06,
+      "loss": 0.3904,
+      "step": 8020
+    },
+    {
+      "epoch": 1.3292501241516306,
+      "grad_norm": 1.7437762022018433,
+      "learning_rate": 6.8243141790391345e-06,
+      "loss": 0.4397,
+      "step": 8030
+    },
+    {
+      "epoch": 1.330905479225294,
+      "grad_norm": 2.2342443466186523,
+      "learning_rate": 6.8153438620105005e-06,
+      "loss": 0.4762,
+      "step": 8040
+    },
+    {
+      "epoch": 1.332560834298957,
+      "grad_norm": 3.045323133468628,
+      "learning_rate": 6.806366809780415e-06,
+      "loss": 0.4217,
+      "step": 8050
+    },
+    {
+      "epoch": 1.3342161893726203,
+      "grad_norm": 1.7698055505752563,
+      "learning_rate": 6.797383055655105e-06,
+      "loss": 0.4276,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3358715444462836,
+      "grad_norm": 1.860013723373413,
+      "learning_rate": 6.788392632965661e-06,
+      "loss": 0.3845,
+      "step": 8070
+    },
+    {
+      "epoch": 1.337526899519947,
+      "grad_norm": 2.04567551612854,
+      "learning_rate": 6.779395575067919e-06,
+      "loss": 0.4531,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3391822545936103,
+      "grad_norm": 1.8434317111968994,
+      "learning_rate": 6.770391915342329e-06,
+      "loss": 0.4137,
+      "step": 8090
+    },
+    {
+      "epoch": 1.3408376096672736,
+      "grad_norm": 2.2670910358428955,
+      "learning_rate": 6.761381687193836e-06,
+      "loss": 0.4245,
+      "step": 8100
+    },
+    {
+      "epoch": 1.342492964740937,
+      "grad_norm": 2.1821208000183105,
+      "learning_rate": 6.752364924051757e-06,
+      "loss": 0.4651,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3441483198146003,
+      "grad_norm": 1.7987016439437866,
+      "learning_rate": 6.7433416593696485e-06,
+      "loss": 0.4426,
+      "step": 8120
+    },
+    {
+      "epoch": 1.3458036748882636,
+      "grad_norm": 2.464167833328247,
+      "learning_rate": 6.734311926625198e-06,
+      "loss": 0.4304,
+      "step": 8130
+    },
+    {
+      "epoch": 1.347459029961927,
+      "grad_norm": 2.0317628383636475,
+      "learning_rate": 6.725275759320082e-06,
+      "loss": 0.418,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3491143850355902,
+      "grad_norm": 1.8057821989059448,
+      "learning_rate": 6.716233190979855e-06,
+      "loss": 0.4229,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3507697401092535,
+      "grad_norm": 2.1078262329101562,
+      "learning_rate": 6.707184255153818e-06,
+      "loss": 0.4275,
+      "step": 8160
+    },
+    {
+      "epoch": 1.3524250951829169,
+      "grad_norm": 1.794978380203247,
+      "learning_rate": 6.698128985414899e-06,
+      "loss": 0.4377,
+      "step": 8170
+    },
+    {
+      "epoch": 1.35408045025658,
+      "grad_norm": 1.7811648845672607,
+      "learning_rate": 6.689067415359522e-06,
+      "loss": 0.4426,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3557358053302433,
+      "grad_norm": 2.394949197769165,
+      "learning_rate": 6.6799995786074916e-06,
+      "loss": 0.4604,
+      "step": 8190
+    },
+    {
+      "epoch": 1.3573911604039066,
+      "grad_norm": 2.0753695964813232,
+      "learning_rate": 6.6709255088018545e-06,
+      "loss": 0.4313,
+      "step": 8200
+    },
+    {
+      "epoch": 1.35904651547757,
+      "grad_norm": 2.000762939453125,
+      "learning_rate": 6.661845239608792e-06,
+      "loss": 0.3965,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3607018705512333,
+      "grad_norm": 2.1908514499664307,
+      "learning_rate": 6.652758804717479e-06,
+      "loss": 0.4589,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3623572256248966,
+      "grad_norm": 1.8629430532455444,
+      "learning_rate": 6.643666237839973e-06,
+      "loss": 0.4131,
+      "step": 8230
+    },
+    {
+      "epoch": 1.36401258069856,
+      "grad_norm": 1.983957290649414,
+      "learning_rate": 6.6345675727110745e-06,
+      "loss": 0.4443,
+      "step": 8240
+    },
+    {
+      "epoch": 1.365667935772223,
+      "grad_norm": 1.5461257696151733,
+      "learning_rate": 6.625462843088214e-06,
+      "loss": 0.3976,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3673232908458863,
+      "grad_norm": 2.0671067237854004,
+      "learning_rate": 6.616352082751322e-06,
+      "loss": 0.4613,
+      "step": 8260
+    },
+    {
+      "epoch": 1.3689786459195497,
+      "grad_norm": 1.9865403175354004,
+      "learning_rate": 6.607235325502703e-06,
+      "loss": 0.3965,
+      "step": 8270
+    },
+    {
+      "epoch": 1.370634000993213,
+      "grad_norm": 1.8592555522918701,
+      "learning_rate": 6.598112605166909e-06,
+      "loss": 0.4623,
+      "step": 8280
+    },
+    {
+      "epoch": 1.3722893560668763,
+      "grad_norm": 2.026170253753662,
+      "learning_rate": 6.588983955590622e-06,
+      "loss": 0.4354,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3739447111405396,
+      "grad_norm": 1.7887972593307495,
+      "learning_rate": 6.5798494106425155e-06,
+      "loss": 0.4408,
+      "step": 8300
+    },
+    {
+      "epoch": 1.375600066214203,
+      "grad_norm": 1.863389253616333,
+      "learning_rate": 6.570709004213139e-06,
+      "loss": 0.4266,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3772554212878663,
+      "grad_norm": 1.7692625522613525,
+      "learning_rate": 6.56156277021479e-06,
+      "loss": 0.444,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3789107763615296,
+      "grad_norm": 2.5655503273010254,
+      "learning_rate": 6.5524107425813834e-06,
+      "loss": 0.408,
+      "step": 8330
+    },
+    {
+      "epoch": 1.380566131435193,
+      "grad_norm": 2.5863282680511475,
+      "learning_rate": 6.543252955268335e-06,
+      "loss": 0.4313,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3822214865088562,
+      "grad_norm": 1.7978219985961914,
+      "learning_rate": 6.5340894422524246e-06,
+      "loss": 0.4399,
+      "step": 8350
+    },
+    {
+      "epoch": 1.3838768415825196,
+      "grad_norm": 1.9100663661956787,
+      "learning_rate": 6.524920237531678e-06,
+      "loss": 0.4277,
+      "step": 8360
+    },
+    {
+      "epoch": 1.3855321966561829,
+      "grad_norm": 1.9516162872314453,
+      "learning_rate": 6.515745375125236e-06,
+      "loss": 0.4612,
+      "step": 8370
+    },
+    {
+      "epoch": 1.387187551729846,
+      "grad_norm": 1.9329428672790527,
+      "learning_rate": 6.506564889073233e-06,
+      "loss": 0.4383,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3888429068035093,
+      "grad_norm": 1.867749810218811,
+      "learning_rate": 6.497378813436667e-06,
+      "loss": 0.4156,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3904982618771726,
+      "grad_norm": 2.531853437423706,
+      "learning_rate": 6.488187182297272e-06,
+      "loss": 0.4459,
+      "step": 8400
+    },
+    {
+      "epoch": 1.392153616950836,
+      "grad_norm": 1.8974616527557373,
+      "learning_rate": 6.4789900297573985e-06,
+      "loss": 0.4295,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3938089720244993,
+      "grad_norm": 2.2630221843719482,
+      "learning_rate": 6.4697873899398756e-06,
+      "loss": 0.4561,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3954643270981626,
+      "grad_norm": 2.2150561809539795,
+      "learning_rate": 6.460579296987899e-06,
+      "loss": 0.4594,
+      "step": 8430
+    },
+    {
+      "epoch": 1.397119682171826,
+      "grad_norm": 2.0618362426757812,
+      "learning_rate": 6.451365785064887e-06,
+      "loss": 0.4673,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3987750372454892,
+      "grad_norm": 1.7820813655853271,
+      "learning_rate": 6.442146888354373e-06,
+      "loss": 0.4342,
+      "step": 8450
+    },
+    {
+      "epoch": 1.4004303923191523,
+      "grad_norm": 1.9812285900115967,
+      "learning_rate": 6.4329226410598625e-06,
+      "loss": 0.4662,
+      "step": 8460
+    },
+    {
+      "epoch": 1.4020857473928157,
+      "grad_norm": 2.486772060394287,
+      "learning_rate": 6.423693077404713e-06,
+      "loss": 0.4428,
+      "step": 8470
+    },
+    {
+      "epoch": 1.403741102466479,
+      "grad_norm": 2.0029282569885254,
+      "learning_rate": 6.4144582316320085e-06,
+      "loss": 0.4074,
+      "step": 8480
+    },
+    {
+      "epoch": 1.4053964575401423,
+      "grad_norm": 1.755991816520691,
+      "learning_rate": 6.405218138004428e-06,
+      "loss": 0.4384,
+      "step": 8490
+    },
+    {
+      "epoch": 1.4070518126138056,
+      "grad_norm": 2.043971300125122,
+      "learning_rate": 6.395972830804125e-06,
+      "loss": 0.4143,
+      "step": 8500
+    },
+    {
+      "epoch": 1.408707167687469,
+      "grad_norm": 2.047865152359009,
+      "learning_rate": 6.386722344332591e-06,
+      "loss": 0.3758,
+      "step": 8510
+    },
+    {
+      "epoch": 1.4103625227611323,
+      "grad_norm": 1.7395375967025757,
+      "learning_rate": 6.3774667129105374e-06,
+      "loss": 0.4224,
+      "step": 8520
+    },
+    {
+      "epoch": 1.4120178778347956,
+      "grad_norm": 1.747536063194275,
+      "learning_rate": 6.36820597087776e-06,
+      "loss": 0.4524,
+      "step": 8530
+    },
+    {
+      "epoch": 1.413673232908459,
+      "grad_norm": 1.9488874673843384,
+      "learning_rate": 6.358940152593021e-06,
+      "loss": 0.4208,
+      "step": 8540
+    },
+    {
+      "epoch": 1.4153285879821222,
+      "grad_norm": 2.2660651206970215,
+      "learning_rate": 6.349669292433913e-06,
+      "loss": 0.426,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4169839430557856,
+      "grad_norm": 2.1231184005737305,
+      "learning_rate": 6.340393424796735e-06,
+      "loss": 0.432,
+      "step": 8560
+    },
+    {
+      "epoch": 1.418639298129449,
+      "grad_norm": 2.2942655086517334,
+      "learning_rate": 6.331112584096364e-06,
+      "loss": 0.4558,
+      "step": 8570
+    },
+    {
+      "epoch": 1.420294653203112,
+      "grad_norm": 2.481250762939453,
+      "learning_rate": 6.3218268047661316e-06,
+      "loss": 0.3894,
+      "step": 8580
+    },
+    {
+      "epoch": 1.4219500082767753,
+      "grad_norm": 1.7492702007293701,
+      "learning_rate": 6.312536121257685e-06,
+      "loss": 0.4185,
+      "step": 8590
+    },
+    {
+      "epoch": 1.4236053633504386,
+      "grad_norm": 1.907914400100708,
+      "learning_rate": 6.303240568040875e-06,
+      "loss": 0.4708,
+      "step": 8600
+    },
+    {
+      "epoch": 1.425260718424102,
+      "grad_norm": 2.4471542835235596,
+      "learning_rate": 6.293940179603614e-06,
+      "loss": 0.3845,
+      "step": 8610
+    },
+    {
+      "epoch": 1.4269160734977653,
+      "grad_norm": 2.4658238887786865,
+      "learning_rate": 6.284634990451755e-06,
+      "loss": 0.3813,
+      "step": 8620
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 1.6748656034469604,
+      "learning_rate": 6.275325035108966e-06,
+      "loss": 0.4004,
+      "step": 8630
+    },
+    {
+      "epoch": 1.430226783645092,
+      "grad_norm": 2.4648280143737793,
+      "learning_rate": 6.266010348116592e-06,
+      "loss": 0.4744,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4318821387187552,
+      "grad_norm": 1.831211805343628,
+      "learning_rate": 6.256690964033537e-06,
+      "loss": 0.4707,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4335374937924183,
+      "grad_norm": 1.9654415845870972,
+      "learning_rate": 6.247366917436135e-06,
+      "loss": 0.4265,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4351928488660817,
+      "grad_norm": 1.9499164819717407,
+      "learning_rate": 6.238038242918012e-06,
+      "loss": 0.4302,
+      "step": 8670
+    },
+    {
+      "epoch": 1.436848203939745,
+      "grad_norm": 1.837054967880249,
+      "learning_rate": 6.228704975089966e-06,
+      "loss": 0.4066,
+      "step": 8680
+    },
+    {
+      "epoch": 1.4385035590134083,
+      "grad_norm": 2.007383108139038,
+      "learning_rate": 6.2193671485798414e-06,
+      "loss": 0.4104,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4401589140870716,
+      "grad_norm": 1.9571911096572876,
+      "learning_rate": 6.2100247980323925e-06,
+      "loss": 0.4567,
+      "step": 8700
+    },
+    {
+      "epoch": 1.441814269160735,
+      "grad_norm": 1.5203781127929688,
+      "learning_rate": 6.200677958109156e-06,
+      "loss": 0.4417,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4434696242343983,
+      "grad_norm": 2.683063268661499,
+      "learning_rate": 6.191326663488331e-06,
+      "loss": 0.4542,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4451249793080616,
+      "grad_norm": 2.251321792602539,
+      "learning_rate": 6.181970948864637e-06,
+      "loss": 0.45,
+      "step": 8730
+    },
+    {
+      "epoch": 1.446780334381725,
+      "grad_norm": 1.943559169769287,
+      "learning_rate": 6.172610848949201e-06,
+      "loss": 0.42,
+      "step": 8740
+    },
+    {
+      "epoch": 1.4484356894553883,
+      "grad_norm": 1.8431415557861328,
+      "learning_rate": 6.163246398469413e-06,
+      "loss": 0.4268,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4500910445290516,
+      "grad_norm": 1.8626545667648315,
+      "learning_rate": 6.153877632168805e-06,
+      "loss": 0.4441,
+      "step": 8760
+    },
+    {
+      "epoch": 1.451746399602715,
+      "grad_norm": 1.4640800952911377,
+      "learning_rate": 6.144504584806924e-06,
+      "loss": 0.4061,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4534017546763782,
+      "grad_norm": 2.071951150894165,
+      "learning_rate": 6.135127291159201e-06,
+      "loss": 0.4492,
+      "step": 8780
+    },
+    {
+      "epoch": 1.4550571097500413,
+      "grad_norm": 2.0795013904571533,
+      "learning_rate": 6.125745786016818e-06,
+      "loss": 0.4285,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4567124648237046,
+      "grad_norm": 1.7705730199813843,
+      "learning_rate": 6.116360104186586e-06,
+      "loss": 0.3959,
+      "step": 8800
+    },
+    {
+      "epoch": 1.458367819897368,
+      "grad_norm": 2.478497266769409,
+      "learning_rate": 6.106970280490807e-06,
+      "loss": 0.4817,
+      "step": 8810
+    },
+    {
+      "epoch": 1.4600231749710313,
+      "grad_norm": 1.7252123355865479,
+      "learning_rate": 6.097576349767155e-06,
+      "loss": 0.4097,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4616785300446946,
+      "grad_norm": 1.9439191818237305,
+      "learning_rate": 6.08817834686854e-06,
+      "loss": 0.422,
+      "step": 8830
+    },
+    {
+      "epoch": 1.463333885118358,
+      "grad_norm": 1.86637282371521,
+      "learning_rate": 6.07877630666298e-06,
+      "loss": 0.4097,
+      "step": 8840
+    },
+    {
+      "epoch": 1.4649892401920213,
+      "grad_norm": 2.0580270290374756,
+      "learning_rate": 6.069370264033472e-06,
+      "loss": 0.4683,
+      "step": 8850
+    },
+    {
+      "epoch": 1.4666445952656844,
+      "grad_norm": 1.9036422967910767,
+      "learning_rate": 6.059960253877861e-06,
+      "loss": 0.4154,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4682999503393477,
+      "grad_norm": 2.2205872535705566,
+      "learning_rate": 6.050546311108718e-06,
+      "loss": 0.432,
+      "step": 8870
+    },
+    {
+      "epoch": 1.469955305413011,
+      "grad_norm": 1.7262464761734009,
+      "learning_rate": 6.041128470653197e-06,
+      "loss": 0.4073,
+      "step": 8880
+    },
+    {
+      "epoch": 1.4716106604866743,
+      "grad_norm": 1.961290955543518,
+      "learning_rate": 6.0317067674529186e-06,
+      "loss": 0.4177,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4732660155603376,
+      "grad_norm": 1.9213963747024536,
+      "learning_rate": 6.022281236463829e-06,
+      "loss": 0.4509,
+      "step": 8900
+    },
+    {
+      "epoch": 1.474921370634001,
+      "grad_norm": 2.405870199203491,
+      "learning_rate": 6.012851912656084e-06,
+      "loss": 0.4234,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4765767257076643,
+      "grad_norm": 1.8768267631530762,
+      "learning_rate": 6.003418831013908e-06,
+      "loss": 0.4133,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4782320807813276,
+      "grad_norm": 1.9255216121673584,
+      "learning_rate": 5.993982026535461e-06,
+      "loss": 0.4286,
+      "step": 8930
+    },
+    {
+      "epoch": 1.479887435854991,
+      "grad_norm": 1.9891632795333862,
+      "learning_rate": 5.984541534232725e-06,
+      "loss": 0.4363,
+      "step": 8940
+    },
+    {
+      "epoch": 1.4815427909286543,
+      "grad_norm": 2.4171807765960693,
+      "learning_rate": 5.97509738913136e-06,
+      "loss": 0.4239,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4831981460023176,
+      "grad_norm": 2.0096046924591064,
+      "learning_rate": 5.96564962627058e-06,
+      "loss": 0.4461,
+      "step": 8960
+    },
+    {
+      "epoch": 1.484853501075981,
+      "grad_norm": 2.0493736267089844,
+      "learning_rate": 5.956198280703016e-06,
+      "loss": 0.4759,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4865088561496442,
+      "grad_norm": 2.407358169555664,
+      "learning_rate": 5.946743387494598e-06,
+      "loss": 0.4372,
+      "step": 8980
+    },
+    {
+      "epoch": 1.4881642112233073,
+      "grad_norm": 2.0386157035827637,
+      "learning_rate": 5.937284981724416e-06,
+      "loss": 0.4227,
+      "step": 8990
+    },
+    {
+      "epoch": 1.4898195662969707,
+      "grad_norm": 1.7787805795669556,
+      "learning_rate": 5.9278230984845934e-06,
+      "loss": 0.384,
+      "step": 9000
+    },
+    {
+      "epoch": 1.491474921370634,
+      "grad_norm": 2.143181085586548,
+      "learning_rate": 5.9183577728801524e-06,
+      "loss": 0.4499,
+      "step": 9010
+    },
+    {
+      "epoch": 1.4931302764442973,
+      "grad_norm": 2.1203973293304443,
+      "learning_rate": 5.908889040028887e-06,
+      "loss": 0.4441,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4947856315179606,
+      "grad_norm": 1.7282978296279907,
+      "learning_rate": 5.899416935061237e-06,
+      "loss": 0.3813,
+      "step": 9030
+    },
+    {
+      "epoch": 1.496440986591624,
+      "grad_norm": 1.6689910888671875,
+      "learning_rate": 5.889941493120151e-06,
+      "loss": 0.4233,
+      "step": 9040
+    },
+    {
+      "epoch": 1.4980963416652873,
+      "grad_norm": 1.9057822227478027,
+      "learning_rate": 5.880462749360956e-06,
+      "loss": 0.4192,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4997516967389504,
+      "grad_norm": 2.1098954677581787,
+      "learning_rate": 5.8709807389512294e-06,
+      "loss": 0.431,
+      "step": 9060
+    },
+    {
+      "epoch": 1.5014070518126137,
+      "grad_norm": 1.6635946035385132,
+      "learning_rate": 5.861495497070675e-06,
+      "loss": 0.4421,
+      "step": 9070
+    },
+    {
+      "epoch": 1.503062406886277,
+      "grad_norm": 1.9180066585540771,
+      "learning_rate": 5.8520070589109755e-06,
+      "loss": 0.4599,
+      "step": 9080
+    },
+    {
+      "epoch": 1.5047177619599403,
+      "grad_norm": 1.7293899059295654,
+      "learning_rate": 5.842515459675681e-06,
+      "loss": 0.4043,
+      "step": 9090
+    },
+    {
+      "epoch": 1.5063731170336037,
+      "grad_norm": 2.8191874027252197,
+      "learning_rate": 5.833020734580065e-06,
+      "loss": 0.4037,
+      "step": 9100
+    },
+    {
+      "epoch": 1.508028472107267,
+      "grad_norm": 1.7623112201690674,
+      "learning_rate": 5.823522918851e-06,
+      "loss": 0.4251,
+      "step": 9110
+    },
+    {
+      "epoch": 1.5096838271809303,
+      "grad_norm": 2.5156078338623047,
+      "learning_rate": 5.814022047726826e-06,
+      "loss": 0.4173,
+      "step": 9120
+    },
+    {
+      "epoch": 1.5113391822545936,
+      "grad_norm": 2.159651279449463,
+      "learning_rate": 5.804518156457216e-06,
+      "loss": 0.4365,
+      "step": 9130
+    },
+    {
+      "epoch": 1.512994537328257,
+      "grad_norm": 1.5260064601898193,
+      "learning_rate": 5.7950112803030504e-06,
+      "loss": 0.4707,
+      "step": 9140
+    },
+    {
+      "epoch": 1.5146498924019203,
+      "grad_norm": 1.9621107578277588,
+      "learning_rate": 5.785501454536286e-06,
+      "loss": 0.4374,
+      "step": 9150
+    },
+    {
+      "epoch": 1.5163052474755836,
+      "grad_norm": 1.5603402853012085,
+      "learning_rate": 5.775988714439817e-06,
+      "loss": 0.457,
+      "step": 9160
+    },
+    {
+      "epoch": 1.517960602549247,
+      "grad_norm": 1.9429376125335693,
+      "learning_rate": 5.766473095307357e-06,
+      "loss": 0.4234,
+      "step": 9170
+    },
+    {
+      "epoch": 1.5196159576229102,
+      "grad_norm": 1.9809801578521729,
+      "learning_rate": 5.756954632443297e-06,
+      "loss": 0.4448,
+      "step": 9180
+    },
+    {
+      "epoch": 1.5212713126965736,
+      "grad_norm": 2.2237918376922607,
+      "learning_rate": 5.747433361162581e-06,
+      "loss": 0.3984,
+      "step": 9190
+    },
+    {
+      "epoch": 1.5229266677702367,
+      "grad_norm": 2.021394729614258,
+      "learning_rate": 5.737909316790571e-06,
+      "loss": 0.4425,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5245820228439,
+      "grad_norm": 2.0034067630767822,
+      "learning_rate": 5.728382534662917e-06,
+      "loss": 0.4359,
+      "step": 9210
+    },
+    {
+      "epoch": 1.5262373779175633,
+      "grad_norm": 1.3382004499435425,
+      "learning_rate": 5.718853050125429e-06,
+      "loss": 0.4135,
+      "step": 9220
+    },
+    {
+      "epoch": 1.5278927329912266,
+      "grad_norm": 2.4722602367401123,
+      "learning_rate": 5.709320898533942e-06,
+      "loss": 0.4485,
+      "step": 9230
+    },
+    {
+      "epoch": 1.52954808806489,
+      "grad_norm": 1.733469843864441,
+      "learning_rate": 5.699786115254187e-06,
+      "loss": 0.4223,
+      "step": 9240
+    },
+    {
+      "epoch": 1.531203443138553,
+      "grad_norm": 2.123274087905884,
+      "learning_rate": 5.690248735661655e-06,
+      "loss": 0.4305,
+      "step": 9250
+    },
+    {
+      "epoch": 1.5328587982122164,
+      "grad_norm": 1.6809306144714355,
+      "learning_rate": 5.680708795141478e-06,
+      "loss": 0.4472,
+      "step": 9260
+    },
+    {
+      "epoch": 1.5345141532858797,
+      "grad_norm": 1.8373470306396484,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.4031,
+      "step": 9270
+    },
+    {
+      "epoch": 1.536169508359543,
+      "grad_norm": 2.4719669818878174,
+      "learning_rate": 5.661621372906058e-06,
+      "loss": 0.3956,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5378248634332063,
+      "grad_norm": 2.367678642272949,
+      "learning_rate": 5.652073962008054e-06,
+      "loss": 0.4775,
+      "step": 9290
+    },
+    {
+      "epoch": 1.5394802185068697,
+      "grad_norm": 2.1993610858917236,
+      "learning_rate": 5.64252413181661e-06,
+      "loss": 0.4199,
+      "step": 9300
+    },
+    {
+      "epoch": 1.541135573580533,
+      "grad_norm": 2.5951197147369385,
+      "learning_rate": 5.632971917763047e-06,
+      "loss": 0.4151,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5427909286541963,
+      "grad_norm": 2.3838794231414795,
+      "learning_rate": 5.623417355287532e-06,
+      "loss": 0.4379,
+      "step": 9320
+    },
+    {
+      "epoch": 1.5444462837278596,
+      "grad_norm": 2.005969285964966,
+      "learning_rate": 5.61386047983894e-06,
+      "loss": 0.4107,
+      "step": 9330
+    },
+    {
+      "epoch": 1.546101638801523,
+      "grad_norm": 2.2176222801208496,
+      "learning_rate": 5.604301326874729e-06,
+      "loss": 0.451,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5477569938751863,
+      "grad_norm": 2.1976559162139893,
+      "learning_rate": 5.594739931860812e-06,
+      "loss": 0.4042,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5494123489488496,
+      "grad_norm": 2.2575743198394775,
+      "learning_rate": 5.585176330271417e-06,
+      "loss": 0.4359,
+      "step": 9360
+    },
+    {
+      "epoch": 1.551067704022513,
+      "grad_norm": 2.771317720413208,
+      "learning_rate": 5.575610557588955e-06,
+      "loss": 0.4128,
+      "step": 9370
+    },
+    {
+      "epoch": 1.5527230590961763,
+      "grad_norm": 1.7214008569717407,
+      "learning_rate": 5.566042649303899e-06,
+      "loss": 0.4136,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5543784141698396,
+      "grad_norm": 2.055534601211548,
+      "learning_rate": 5.556472640914639e-06,
+      "loss": 0.4742,
+      "step": 9390
+    },
+    {
+      "epoch": 1.556033769243503,
+      "grad_norm": 2.1422717571258545,
+      "learning_rate": 5.5469005679273616e-06,
+      "loss": 0.4336,
+      "step": 9400
+    },
+    {
+      "epoch": 1.557689124317166,
+      "grad_norm": 2.063735246658325,
+      "learning_rate": 5.537326465855911e-06,
+      "loss": 0.4168,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5593444793908293,
+      "grad_norm": 2.0593113899230957,
+      "learning_rate": 5.527750370221661e-06,
+      "loss": 0.4187,
+      "step": 9420
+    },
+    {
+      "epoch": 1.5609998344644926,
+      "grad_norm": 2.017012596130371,
+      "learning_rate": 5.518172316553378e-06,
+      "loss": 0.448,
+      "step": 9430
+    },
+    {
+      "epoch": 1.562655189538156,
+      "grad_norm": 1.9413402080535889,
+      "learning_rate": 5.5085923403871e-06,
+      "loss": 0.4058,
+      "step": 9440
+    },
+    {
+      "epoch": 1.5643105446118193,
+      "grad_norm": 1.7295361757278442,
+      "learning_rate": 5.4990104772659895e-06,
+      "loss": 0.4131,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5659658996854824,
+      "grad_norm": 2.06921124458313,
+      "learning_rate": 5.489426762740218e-06,
+      "loss": 0.4081,
+      "step": 9460
+    },
+    {
+      "epoch": 1.5676212547591457,
+      "grad_norm": 1.8260921239852905,
+      "learning_rate": 5.479841232366818e-06,
+      "loss": 0.3937,
+      "step": 9470
+    },
+    {
+      "epoch": 1.569276609832809,
+      "grad_norm": 2.1799001693725586,
+      "learning_rate": 5.470253921709565e-06,
+      "loss": 0.3873,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5709319649064724,
+      "grad_norm": 2.076521158218384,
+      "learning_rate": 5.46066486633884e-06,
+      "loss": 0.4254,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5725873199801357,
+      "grad_norm": 2.1282382011413574,
+      "learning_rate": 5.451074101831492e-06,
+      "loss": 0.3999,
+      "step": 9500
+    },
+    {
+      "epoch": 1.574242675053799,
+      "grad_norm": 2.0794641971588135,
+      "learning_rate": 5.4414816637707115e-06,
+      "loss": 0.3946,
+      "step": 9510
+    },
+    {
+      "epoch": 1.5758980301274623,
+      "grad_norm": 2.05676531791687,
+      "learning_rate": 5.431887587745906e-06,
+      "loss": 0.3978,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5775533852011256,
+      "grad_norm": 2.6630096435546875,
+      "learning_rate": 5.422291909352554e-06,
+      "loss": 0.4185,
+      "step": 9530
+    },
+    {
+      "epoch": 1.579208740274789,
+      "grad_norm": 2.0524003505706787,
+      "learning_rate": 5.4126946641920766e-06,
+      "loss": 0.4344,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5808640953484523,
+      "grad_norm": 2.0351216793060303,
+      "learning_rate": 5.403095887871712e-06,
+      "loss": 0.3743,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5825194504221156,
+      "grad_norm": 1.9727466106414795,
+      "learning_rate": 5.393495616004382e-06,
+      "loss": 0.4016,
+      "step": 9560
+    },
+    {
+      "epoch": 1.584174805495779,
+      "grad_norm": 2.116448402404785,
+      "learning_rate": 5.383893884208548e-06,
+      "loss": 0.4173,
+      "step": 9570
+    },
+    {
+      "epoch": 1.5858301605694423,
+      "grad_norm": 2.624253273010254,
+      "learning_rate": 5.3742907281080956e-06,
+      "loss": 0.4528,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5874855156431056,
+      "grad_norm": 1.7432576417922974,
+      "learning_rate": 5.364686183332194e-06,
+      "loss": 0.3978,
+      "step": 9590
+    },
+    {
+      "epoch": 1.589140870716769,
+      "grad_norm": 2.323892831802368,
+      "learning_rate": 5.35508028551516e-06,
+      "loss": 0.4039,
+      "step": 9600
+    },
+    {
+      "epoch": 1.590796225790432,
+      "grad_norm": 1.9422341585159302,
+      "learning_rate": 5.345473070296337e-06,
+      "loss": 0.42,
+      "step": 9610
+    },
+    {
+      "epoch": 1.5924515808640953,
+      "grad_norm": 1.8938024044036865,
+      "learning_rate": 5.335864573319951e-06,
+      "loss": 0.3762,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5941069359377587,
+      "grad_norm": 2.0808968544006348,
+      "learning_rate": 5.326254830234984e-06,
+      "loss": 0.4134,
+      "step": 9630
+    },
+    {
+      "epoch": 1.595762291011422,
+      "grad_norm": 1.8897898197174072,
+      "learning_rate": 5.316643876695043e-06,
+      "loss": 0.4059,
+      "step": 9640
+    },
+    {
+      "epoch": 1.5974176460850853,
+      "grad_norm": 1.8060656785964966,
+      "learning_rate": 5.307031748358227e-06,
+      "loss": 0.4234,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5990730011587484,
+      "grad_norm": 1.8971941471099854,
+      "learning_rate": 5.29741848088699e-06,
+      "loss": 0.4222,
+      "step": 9660
+    },
+    {
+      "epoch": 1.6007283562324117,
+      "grad_norm": 2.0121896266937256,
+      "learning_rate": 5.2878041099480145e-06,
+      "loss": 0.3959,
+      "step": 9670
+    },
+    {
+      "epoch": 1.602383711306075,
+      "grad_norm": 2.076582908630371,
+      "learning_rate": 5.278188671212079e-06,
+      "loss": 0.4592,
+      "step": 9680
+    },
+    {
+      "epoch": 1.6040390663797384,
+      "grad_norm": 2.6054913997650146,
+      "learning_rate": 5.2685722003539215e-06,
+      "loss": 0.4128,
+      "step": 9690
+    },
+    {
+      "epoch": 1.6056944214534017,
+      "grad_norm": 2.281392812728882,
+      "learning_rate": 5.258954733052109e-06,
+      "loss": 0.4024,
+      "step": 9700
+    },
+    {
+      "epoch": 1.607349776527065,
+      "grad_norm": 2.435349225997925,
+      "learning_rate": 5.249336304988904e-06,
+      "loss": 0.4275,
+      "step": 9710
+    },
+    {
+      "epoch": 1.6090051316007283,
+      "grad_norm": 1.3645190000534058,
+      "learning_rate": 5.239716951850136e-06,
+      "loss": 0.3877,
+      "step": 9720
+    },
+    {
+      "epoch": 1.6106604866743917,
+      "grad_norm": 2.2975997924804688,
+      "learning_rate": 5.230096709325069e-06,
+      "loss": 0.4439,
+      "step": 9730
+    },
+    {
+      "epoch": 1.612315841748055,
+      "grad_norm": 2.102614164352417,
+      "learning_rate": 5.220475613106261e-06,
+      "loss": 0.4095,
+      "step": 9740
+    },
+    {
+      "epoch": 1.6139711968217183,
+      "grad_norm": 1.812879204750061,
+      "learning_rate": 5.210853698889442e-06,
+      "loss": 0.3996,
+      "step": 9750
+    },
+    {
+      "epoch": 1.6156265518953816,
+      "grad_norm": 1.8167248964309692,
+      "learning_rate": 5.201231002373374e-06,
+      "loss": 0.3547,
+      "step": 9760
+    },
+    {
+      "epoch": 1.617281906969045,
+      "grad_norm": 1.8382624387741089,
+      "learning_rate": 5.191607559259723e-06,
+      "loss": 0.418,
+      "step": 9770
+    },
+    {
+      "epoch": 1.6189372620427083,
+      "grad_norm": 1.9886109828948975,
+      "learning_rate": 5.181983405252925e-06,
+      "loss": 0.4426,
+      "step": 9780
+    },
+    {
+      "epoch": 1.6205926171163716,
+      "grad_norm": 2.0963995456695557,
+      "learning_rate": 5.172358576060052e-06,
+      "loss": 0.438,
+      "step": 9790
+    },
+    {
+      "epoch": 1.622247972190035,
+      "grad_norm": 2.365187168121338,
+      "learning_rate": 5.162733107390684e-06,
+      "loss": 0.4224,
+      "step": 9800
+    },
+    {
+      "epoch": 1.623903327263698,
+      "grad_norm": 2.2666516304016113,
+      "learning_rate": 5.153107034956772e-06,
+      "loss": 0.4314,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6255586823373613,
+      "grad_norm": 2.123664379119873,
+      "learning_rate": 5.143480394472504e-06,
+      "loss": 0.409,
+      "step": 9820
+    },
+    {
+      "epoch": 1.6272140374110247,
+      "grad_norm": 2.0788767337799072,
+      "learning_rate": 5.13385322165418e-06,
+      "loss": 0.4135,
+      "step": 9830
+    },
+    {
+      "epoch": 1.628869392484688,
+      "grad_norm": 2.1448142528533936,
+      "learning_rate": 5.124225552220073e-06,
+      "loss": 0.3914,
+      "step": 9840
+    },
+    {
+      "epoch": 1.6305247475583513,
+      "grad_norm": 1.6150662899017334,
+      "learning_rate": 5.114597421890302e-06,
+      "loss": 0.3785,
+      "step": 9850
+    },
+    {
+      "epoch": 1.6321801026320144,
+      "grad_norm": 2.1484649181365967,
+      "learning_rate": 5.104968866386687e-06,
+      "loss": 0.4356,
+      "step": 9860
+    },
+    {
+      "epoch": 1.6338354577056777,
+      "grad_norm": 1.89021635055542,
+      "learning_rate": 5.095339921432636e-06,
+      "loss": 0.3862,
+      "step": 9870
+    },
+    {
+      "epoch": 1.635490812779341,
+      "grad_norm": 2.2360267639160156,
+      "learning_rate": 5.085710622752994e-06,
+      "loss": 0.4052,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6371461678530044,
+      "grad_norm": 1.9827935695648193,
+      "learning_rate": 5.076081006073925e-06,
+      "loss": 0.4089,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6388015229266677,
+      "grad_norm": 2.240103006362915,
+      "learning_rate": 5.0664511071227676e-06,
+      "loss": 0.3983,
+      "step": 9900
+    },
+    {
+      "epoch": 1.640456878000331,
+      "grad_norm": 2.3089191913604736,
+      "learning_rate": 5.0568209616279095e-06,
+      "loss": 0.4148,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6421122330739943,
+      "grad_norm": 2.1113412380218506,
+      "learning_rate": 5.047190605318652e-06,
+      "loss": 0.4184,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6437675881476577,
+      "grad_norm": 1.8843637704849243,
+      "learning_rate": 5.0375600739250855e-06,
+      "loss": 0.3846,
+      "step": 9930
+    },
+    {
+      "epoch": 1.645422943221321,
+      "grad_norm": 2.040696859359741,
+      "learning_rate": 5.027929403177936e-06,
+      "loss": 0.3744,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6470782982949843,
+      "grad_norm": 1.737169861793518,
+      "learning_rate": 5.01829862880846e-06,
+      "loss": 0.3958,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6487336533686476,
+      "grad_norm": 1.7232611179351807,
+      "learning_rate": 5.008667786548291e-06,
+      "loss": 0.4123,
+      "step": 9960
+    },
+    {
+      "epoch": 1.650389008442311,
+      "grad_norm": 2.533484935760498,
+      "learning_rate": 4.999036912129319e-06,
+      "loss": 0.4368,
+      "step": 9970
+    },
+    {
+      "epoch": 1.6520443635159743,
+      "grad_norm": 2.971538782119751,
+      "learning_rate": 4.989406041283549e-06,
+      "loss": 0.4178,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6536997185896376,
+      "grad_norm": 2.1657238006591797,
+      "learning_rate": 4.9797752097429744e-06,
+      "loss": 0.4145,
+      "step": 9990
+    },
+    {
+      "epoch": 1.655355073663301,
+      "grad_norm": 2.1859874725341797,
+      "learning_rate": 4.970144453239443e-06,
+      "loss": 0.4099,
+      "step": 10000
+    },
+    {
+      "epoch": 1.657010428736964,
+      "grad_norm": 1.8412467241287231,
+      "learning_rate": 4.960513807504523e-06,
+      "loss": 0.3941,
+      "step": 10010
+    },
+    {
+      "epoch": 1.6586657838106273,
+      "grad_norm": 1.6499814987182617,
+      "learning_rate": 4.950883308269378e-06,
+      "loss": 0.4175,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6603211388842907,
+      "grad_norm": 2.210827350616455,
+      "learning_rate": 4.941252991264619e-06,
+      "loss": 0.4301,
+      "step": 10030
+    },
+    {
+      "epoch": 1.661976493957954,
+      "grad_norm": 1.6489720344543457,
+      "learning_rate": 4.931622892220184e-06,
+      "loss": 0.4519,
+      "step": 10040
+    },
+    {
+      "epoch": 1.6636318490316173,
+      "grad_norm": 2.0074875354766846,
+      "learning_rate": 4.921993046865205e-06,
+      "loss": 0.4016,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6652872041052804,
+      "grad_norm": 1.9746689796447754,
+      "learning_rate": 4.9123634909278705e-06,
+      "loss": 0.4213,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6669425591789437,
+      "grad_norm": 1.6770844459533691,
+      "learning_rate": 4.902734260135293e-06,
+      "loss": 0.3922,
+      "step": 10070
+    },
+    {
+      "epoch": 1.668597914252607,
+      "grad_norm": 2.5734076499938965,
+      "learning_rate": 4.893105390213386e-06,
+      "loss": 0.4437,
+      "step": 10080
+    },
+    {
+      "epoch": 1.6702532693262704,
+      "grad_norm": 2.2573366165161133,
+      "learning_rate": 4.883476916886716e-06,
+      "loss": 0.4324,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6719086243999337,
+      "grad_norm": 1.611232042312622,
+      "learning_rate": 4.8738488758783835e-06,
+      "loss": 0.4148,
+      "step": 10100
+    },
+    {
+      "epoch": 1.673563979473597,
+      "grad_norm": 1.7315454483032227,
+      "learning_rate": 4.864221302909882e-06,
+      "loss": 0.4148,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6752193345472604,
+      "grad_norm": 1.6441781520843506,
+      "learning_rate": 4.854594233700969e-06,
+      "loss": 0.4179,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6768746896209237,
+      "grad_norm": 1.7164368629455566,
+      "learning_rate": 4.844967703969532e-06,
+      "loss": 0.399,
+      "step": 10130
+    },
+    {
+      "epoch": 1.678530044694587,
+      "grad_norm": 1.6734899282455444,
+      "learning_rate": 4.835341749431464e-06,
+      "loss": 0.421,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6801853997682503,
+      "grad_norm": 1.6350212097167969,
+      "learning_rate": 4.825716405800513e-06,
+      "loss": 0.4089,
+      "step": 10150
+    },
+    {
+      "epoch": 1.6818407548419136,
+      "grad_norm": 2.038209915161133,
+      "learning_rate": 4.816091708788168e-06,
+      "loss": 0.4091,
+      "step": 10160
+    },
+    {
+      "epoch": 1.683496109915577,
+      "grad_norm": 1.6218624114990234,
+      "learning_rate": 4.806467694103516e-06,
+      "loss": 0.411,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6851514649892403,
+      "grad_norm": 1.921259880065918,
+      "learning_rate": 4.796844397453115e-06,
+      "loss": 0.4145,
+      "step": 10180
+    },
+    {
+      "epoch": 1.6868068200629036,
+      "grad_norm": 2.2492053508758545,
+      "learning_rate": 4.787221854540853e-06,
+      "loss": 0.4506,
+      "step": 10190
+    },
+    {
+      "epoch": 1.688462175136567,
+      "grad_norm": 1.9784705638885498,
+      "learning_rate": 4.7776001010678305e-06,
+      "loss": 0.4293,
+      "step": 10200
+    },
+    {
+      "epoch": 1.6901175302102303,
+      "grad_norm": 1.732125997543335,
+      "learning_rate": 4.767979172732212e-06,
+      "loss": 0.412,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6917728852838934,
+      "grad_norm": 1.7823375463485718,
+      "learning_rate": 4.758359105229103e-06,
+      "loss": 0.4042,
+      "step": 10220
+    },
+    {
+      "epoch": 1.6934282403575567,
+      "grad_norm": 1.6599950790405273,
+      "learning_rate": 4.748739934250416e-06,
+      "loss": 0.3674,
+      "step": 10230
+    },
+    {
+      "epoch": 1.69508359543122,
+      "grad_norm": 1.7613921165466309,
+      "learning_rate": 4.739121695484734e-06,
+      "loss": 0.406,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6967389505048833,
+      "grad_norm": 1.9368239641189575,
+      "learning_rate": 4.7295044246171865e-06,
+      "loss": 0.4133,
+      "step": 10250
+    },
+    {
+      "epoch": 1.6983943055785466,
+      "grad_norm": 2.9349164962768555,
+      "learning_rate": 4.719888157329309e-06,
+      "loss": 0.4141,
+      "step": 10260
+    },
+    {
+      "epoch": 1.7000496606522097,
+      "grad_norm": 1.70059335231781,
+      "learning_rate": 4.710272929298912e-06,
+      "loss": 0.3912,
+      "step": 10270
+    },
+    {
+      "epoch": 1.701705015725873,
+      "grad_norm": 1.844007968902588,
+      "learning_rate": 4.700658776199952e-06,
+      "loss": 0.4032,
+      "step": 10280
+    },
+    {
+      "epoch": 1.7033603707995364,
+      "grad_norm": 1.9813511371612549,
+      "learning_rate": 4.691045733702398e-06,
+      "loss": 0.4067,
+      "step": 10290
+    },
+    {
+      "epoch": 1.7050157258731997,
+      "grad_norm": 2.3343167304992676,
+      "learning_rate": 4.681433837472097e-06,
+      "loss": 0.4264,
+      "step": 10300
+    },
+    {
+      "epoch": 1.706671080946863,
+      "grad_norm": 2.3405990600585938,
+      "learning_rate": 4.671823123170646e-06,
+      "loss": 0.3879,
+      "step": 10310
+    },
+    {
+      "epoch": 1.7083264360205264,
+      "grad_norm": 2.8165476322174072,
+      "learning_rate": 4.662213626455253e-06,
+      "loss": 0.4121,
+      "step": 10320
+    },
+    {
+      "epoch": 1.7099817910941897,
+      "grad_norm": 3.3409717082977295,
+      "learning_rate": 4.65260538297861e-06,
+      "loss": 0.4053,
+      "step": 10330
+    },
+    {
+      "epoch": 1.711637146167853,
+      "grad_norm": 2.1133646965026855,
+      "learning_rate": 4.642998428388761e-06,
+      "loss": 0.4122,
+      "step": 10340
+    },
+    {
+      "epoch": 1.7132925012415163,
+      "grad_norm": 2.1685373783111572,
+      "learning_rate": 4.633392798328966e-06,
+      "loss": 0.4506,
+      "step": 10350
+    },
+    {
+      "epoch": 1.7149478563151797,
+      "grad_norm": 2.0110175609588623,
+      "learning_rate": 4.623788528437571e-06,
+      "loss": 0.4,
+      "step": 10360
+    },
+    {
+      "epoch": 1.716603211388843,
+      "grad_norm": 2.60198712348938,
+      "learning_rate": 4.614185654347877e-06,
+      "loss": 0.4176,
+      "step": 10370
+    },
+    {
+      "epoch": 1.7182585664625063,
+      "grad_norm": 1.7562052011489868,
+      "learning_rate": 4.604584211688004e-06,
+      "loss": 0.3981,
+      "step": 10380
+    },
+    {
+      "epoch": 1.7199139215361696,
+      "grad_norm": 2.134413957595825,
+      "learning_rate": 4.594984236080765e-06,
+      "loss": 0.4336,
+      "step": 10390
+    },
+    {
+      "epoch": 1.721569276609833,
+      "grad_norm": 2.218580961227417,
+      "learning_rate": 4.585385763143526e-06,
+      "loss": 0.4235,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7232246316834963,
+      "grad_norm": 1.7758234739303589,
+      "learning_rate": 4.575788828488078e-06,
+      "loss": 0.4221,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7248799867571594,
+      "grad_norm": 1.8942153453826904,
+      "learning_rate": 4.566193467720506e-06,
+      "loss": 0.4035,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7265353418308227,
+      "grad_norm": 1.9919090270996094,
+      "learning_rate": 4.556599716441062e-06,
+      "loss": 0.4017,
+      "step": 10430
+    },
+    {
+      "epoch": 1.728190696904486,
+      "grad_norm": 2.020827293395996,
+      "learning_rate": 4.547007610244015e-06,
+      "loss": 0.4353,
+      "step": 10440
+    },
+    {
+      "epoch": 1.7298460519781493,
+      "grad_norm": 2.077829360961914,
+      "learning_rate": 4.537417184717536e-06,
+      "loss": 0.4721,
+      "step": 10450
+    },
+    {
+      "epoch": 1.7315014070518127,
+      "grad_norm": 3.0025055408477783,
+      "learning_rate": 4.527828475443562e-06,
+      "loss": 0.4413,
+      "step": 10460
+    },
+    {
+      "epoch": 1.7331567621254758,
+      "grad_norm": 2.1447229385375977,
+      "learning_rate": 4.518241517997657e-06,
+      "loss": 0.4025,
+      "step": 10470
+    },
+    {
+      "epoch": 1.734812117199139,
+      "grad_norm": 1.9372047185897827,
+      "learning_rate": 4.508656347948896e-06,
+      "loss": 0.455,
+      "step": 10480
+    },
+    {
+      "epoch": 1.7364674722728024,
+      "grad_norm": 2.317401647567749,
+      "learning_rate": 4.4990730008597115e-06,
+      "loss": 0.4008,
+      "step": 10490
+    },
+    {
+      "epoch": 1.7381228273464657,
+      "grad_norm": 1.678004264831543,
+      "learning_rate": 4.489491512285776e-06,
+      "loss": 0.4216,
+      "step": 10500
+    },
+    {
+      "epoch": 1.739778182420129,
+      "grad_norm": 1.9995408058166504,
+      "learning_rate": 4.47991191777587e-06,
+      "loss": 0.3926,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7414335374937924,
+      "grad_norm": 2.1435697078704834,
+      "learning_rate": 4.470334252871743e-06,
+      "loss": 0.398,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7430888925674557,
+      "grad_norm": 1.8928353786468506,
+      "learning_rate": 4.4607585531079844e-06,
+      "loss": 0.4211,
+      "step": 10530
+    },
+    {
+      "epoch": 1.744744247641119,
+      "grad_norm": 2.309457302093506,
+      "learning_rate": 4.451184854011898e-06,
+      "loss": 0.4261,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7463996027147823,
+      "grad_norm": 1.7776364088058472,
+      "learning_rate": 4.44161319110336e-06,
+      "loss": 0.4415,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7480549577884457,
+      "grad_norm": 1.907957911491394,
+      "learning_rate": 4.432043599894694e-06,
+      "loss": 0.4271,
+      "step": 10560
+    },
+    {
+      "epoch": 1.749710312862109,
+      "grad_norm": 2.8148648738861084,
+      "learning_rate": 4.422476115890537e-06,
+      "loss": 0.4588,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7513656679357723,
+      "grad_norm": 1.931691288948059,
+      "learning_rate": 4.412910774587708e-06,
+      "loss": 0.4122,
+      "step": 10580
+    },
+    {
+      "epoch": 1.7530210230094356,
+      "grad_norm": 2.7117698192596436,
+      "learning_rate": 4.403347611475073e-06,
+      "loss": 0.442,
+      "step": 10590
+    },
+    {
+      "epoch": 1.754676378083099,
+      "grad_norm": 1.596501111984253,
+      "learning_rate": 4.393786662033424e-06,
+      "loss": 0.4256,
+      "step": 10600
+    },
+    {
+      "epoch": 1.7563317331567623,
+      "grad_norm": 2.3910441398620605,
+      "learning_rate": 4.384227961735334e-06,
+      "loss": 0.458,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7579870882304254,
+      "grad_norm": 2.2331271171569824,
+      "learning_rate": 4.374671546045031e-06,
+      "loss": 0.3796,
+      "step": 10620
+    },
+    {
+      "epoch": 1.7596424433040887,
+      "grad_norm": 1.9223552942276,
+      "learning_rate": 4.365117450418274e-06,
+      "loss": 0.4167,
+      "step": 10630
+    },
+    {
+      "epoch": 1.761297798377752,
+      "grad_norm": 2.403165578842163,
+      "learning_rate": 4.355565710302203e-06,
+      "loss": 0.4277,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7629531534514153,
+      "grad_norm": 2.100135564804077,
+      "learning_rate": 4.346016361135225e-06,
+      "loss": 0.3965,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7646085085250787,
+      "grad_norm": 2.3119328022003174,
+      "learning_rate": 4.3364694383468805e-06,
+      "loss": 0.462,
+      "step": 10660
+    },
+    {
+      "epoch": 1.7662638635987418,
+      "grad_norm": 1.6545701026916504,
+      "learning_rate": 4.3269249773577e-06,
+      "loss": 0.3747,
+      "step": 10670
+    },
+    {
+      "epoch": 1.767919218672405,
+      "grad_norm": 2.4669320583343506,
+      "learning_rate": 4.317383013579084e-06,
+      "loss": 0.3581,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7695745737460684,
+      "grad_norm": 1.7150194644927979,
+      "learning_rate": 4.307843582413166e-06,
+      "loss": 0.38,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7712299288197317,
+      "grad_norm": 2.031128168106079,
+      "learning_rate": 4.298306719252686e-06,
+      "loss": 0.4136,
+      "step": 10700
+    },
+    {
+      "epoch": 1.772885283893395,
+      "grad_norm": 2.0807642936706543,
+      "learning_rate": 4.288772459480854e-06,
+      "loss": 0.4435,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7745406389670584,
+      "grad_norm": 2.0820484161376953,
+      "learning_rate": 4.2792408384712245e-06,
+      "loss": 0.4107,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7761959940407217,
+      "grad_norm": 2.2544493675231934,
+      "learning_rate": 4.269711891587556e-06,
+      "loss": 0.381,
+      "step": 10730
+    },
+    {
+      "epoch": 1.777851349114385,
+      "grad_norm": 1.900412917137146,
+      "learning_rate": 4.260185654183689e-06,
+      "loss": 0.4134,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7795067041880483,
+      "grad_norm": 1.8195431232452393,
+      "learning_rate": 4.250662161603414e-06,
+      "loss": 0.4334,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7811620592617117,
+      "grad_norm": 1.9061986207962036,
+      "learning_rate": 4.24114144918033e-06,
+      "loss": 0.4173,
+      "step": 10760
+    },
+    {
+      "epoch": 1.782817414335375,
+      "grad_norm": 2.0322930812835693,
+      "learning_rate": 4.231623552237731e-06,
+      "loss": 0.3825,
+      "step": 10770
+    },
+    {
+      "epoch": 1.7844727694090383,
+      "grad_norm": 1.757241129875183,
+      "learning_rate": 4.222108506088457e-06,
+      "loss": 0.3781,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7861281244827016,
+      "grad_norm": 2.6230099201202393,
+      "learning_rate": 4.212596346034778e-06,
+      "loss": 0.4582,
+      "step": 10790
+    },
+    {
+      "epoch": 1.787783479556365,
+      "grad_norm": 1.8682286739349365,
+      "learning_rate": 4.20308710736825e-06,
+      "loss": 0.416,
+      "step": 10800
+    },
+    {
+      "epoch": 1.7894388346300283,
+      "grad_norm": 1.6394546031951904,
+      "learning_rate": 4.193580825369591e-06,
+      "loss": 0.3774,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7910941897036916,
+      "grad_norm": 1.5209989547729492,
+      "learning_rate": 4.184077535308554e-06,
+      "loss": 0.4161,
+      "step": 10820
+    },
+    {
+      "epoch": 1.7927495447773547,
+      "grad_norm": 2.6634228229522705,
+      "learning_rate": 4.174577272443788e-06,
+      "loss": 0.3987,
+      "step": 10830
+    },
+    {
+      "epoch": 1.794404899851018,
+      "grad_norm": 2.2621631622314453,
+      "learning_rate": 4.165080072022711e-06,
+      "loss": 0.4146,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7960602549246814,
+      "grad_norm": 2.3683760166168213,
+      "learning_rate": 4.15558596928138e-06,
+      "loss": 0.4474,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7977156099983447,
+      "grad_norm": 2.4090867042541504,
+      "learning_rate": 4.146094999444355e-06,
+      "loss": 0.3767,
+      "step": 10860
+    },
+    {
+      "epoch": 1.799370965072008,
+      "grad_norm": 1.345140814781189,
+      "learning_rate": 4.13660719772458e-06,
+      "loss": 0.3677,
+      "step": 10870
+    },
+    {
+      "epoch": 1.801026320145671,
+      "grad_norm": 1.5577553510665894,
+      "learning_rate": 4.127122599323235e-06,
+      "loss": 0.4328,
+      "step": 10880
+    },
+    {
+      "epoch": 1.8026816752193344,
+      "grad_norm": 2.3623404502868652,
+      "learning_rate": 4.1176412394296265e-06,
+      "loss": 0.3946,
+      "step": 10890
+    },
+    {
+      "epoch": 1.8043370302929977,
+      "grad_norm": 1.6678826808929443,
+      "learning_rate": 4.108163153221036e-06,
+      "loss": 0.4202,
+      "step": 10900
+    },
+    {
+      "epoch": 1.805992385366661,
+      "grad_norm": 1.8554919958114624,
+      "learning_rate": 4.098688375862605e-06,
+      "loss": 0.4213,
+      "step": 10910
+    },
+    {
+      "epoch": 1.8076477404403244,
+      "grad_norm": 2.539498805999756,
+      "learning_rate": 4.089216942507196e-06,
+      "loss": 0.4458,
+      "step": 10920
+    },
+    {
+      "epoch": 1.8093030955139877,
+      "grad_norm": 2.0051186084747314,
+      "learning_rate": 4.0797488882952625e-06,
+      "loss": 0.4274,
+      "step": 10930
+    },
+    {
+      "epoch": 1.810958450587651,
+      "grad_norm": 2.1093010902404785,
+      "learning_rate": 4.070284248354728e-06,
+      "loss": 0.3565,
+      "step": 10940
+    },
+    {
+      "epoch": 1.8126138056613144,
+      "grad_norm": 2.032156229019165,
+      "learning_rate": 4.060823057800842e-06,
+      "loss": 0.4315,
+      "step": 10950
+    },
+    {
+      "epoch": 1.8142691607349777,
+      "grad_norm": 1.781123161315918,
+      "learning_rate": 4.05136535173606e-06,
+      "loss": 0.4531,
+      "step": 10960
+    },
+    {
+      "epoch": 1.815924515808641,
+      "grad_norm": 2.380314350128174,
+      "learning_rate": 4.041911165249905e-06,
+      "loss": 0.3931,
+      "step": 10970
+    },
+    {
+      "epoch": 1.8175798708823043,
+      "grad_norm": 1.8698779344558716,
+      "learning_rate": 4.0324605334188474e-06,
+      "loss": 0.3803,
+      "step": 10980
+    },
+    {
+      "epoch": 1.8192352259559676,
+      "grad_norm": 1.6024515628814697,
+      "learning_rate": 4.0230134913061645e-06,
+      "loss": 0.394,
+      "step": 10990
+    },
+    {
+      "epoch": 1.820890581029631,
+      "grad_norm": 2.1540608406066895,
+      "learning_rate": 4.0135700739618205e-06,
+      "loss": 0.401,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8225459361032943,
+      "grad_norm": 2.3779587745666504,
+      "learning_rate": 4.004130316422327e-06,
+      "loss": 0.4131,
+      "step": 11010
+    },
+    {
+      "epoch": 1.8242012911769576,
+      "grad_norm": 1.9901835918426514,
+      "learning_rate": 3.994694253710617e-06,
+      "loss": 0.4113,
+      "step": 11020
+    },
+    {
+      "epoch": 1.8258566462506207,
+      "grad_norm": 1.5083597898483276,
+      "learning_rate": 3.985261920835917e-06,
+      "loss": 0.3972,
+      "step": 11030
+    },
+    {
+      "epoch": 1.827512001324284,
+      "grad_norm": 2.1913838386535645,
+      "learning_rate": 3.975833352793615e-06,
+      "loss": 0.4391,
+      "step": 11040
+    },
+    {
+      "epoch": 1.8291673563979474,
+      "grad_norm": 2.0615127086639404,
+      "learning_rate": 3.966408584565127e-06,
+      "loss": 0.4167,
+      "step": 11050
+    },
+    {
+      "epoch": 1.8308227114716107,
+      "grad_norm": 1.6090701818466187,
+      "learning_rate": 3.9569876511177805e-06,
+      "loss": 0.4146,
+      "step": 11060
+    },
+    {
+      "epoch": 1.832478066545274,
+      "grad_norm": 2.14567494392395,
+      "learning_rate": 3.947570587404667e-06,
+      "loss": 0.3594,
+      "step": 11070
+    },
+    {
+      "epoch": 1.834133421618937,
+      "grad_norm": 1.3794012069702148,
+      "learning_rate": 3.9381574283645215e-06,
+      "loss": 0.4172,
+      "step": 11080
+    },
+    {
+      "epoch": 1.8357887766926004,
+      "grad_norm": 1.5953060388565063,
+      "learning_rate": 3.928748208921597e-06,
+      "loss": 0.356,
+      "step": 11090
+    },
+    {
+      "epoch": 1.8374441317662638,
+      "grad_norm": 2.281615972518921,
+      "learning_rate": 3.919342963985524e-06,
+      "loss": 0.3627,
+      "step": 11100
+    },
+    {
+      "epoch": 1.839099486839927,
+      "grad_norm": 1.8211688995361328,
+      "learning_rate": 3.909941728451188e-06,
+      "loss": 0.4197,
+      "step": 11110
+    },
+    {
+      "epoch": 1.8407548419135904,
+      "grad_norm": 2.0925374031066895,
+      "learning_rate": 3.900544537198607e-06,
+      "loss": 0.4497,
+      "step": 11120
+    },
+    {
+      "epoch": 1.8424101969872537,
+      "grad_norm": 2.1480743885040283,
+      "learning_rate": 3.891151425092783e-06,
+      "loss": 0.3789,
+      "step": 11130
+    },
+    {
+      "epoch": 1.844065552060917,
+      "grad_norm": 1.6905628442764282,
+      "learning_rate": 3.8817624269835904e-06,
+      "loss": 0.4086,
+      "step": 11140
+    },
+    {
+      "epoch": 1.8457209071345804,
+      "grad_norm": 1.8402177095413208,
+      "learning_rate": 3.872377577705637e-06,
+      "loss": 0.4026,
+      "step": 11150
+    },
+    {
+      "epoch": 1.8473762622082437,
+      "grad_norm": 1.9096100330352783,
+      "learning_rate": 3.862996912078138e-06,
+      "loss": 0.4017,
+      "step": 11160
+    },
+    {
+      "epoch": 1.849031617281907,
+      "grad_norm": 2.6157302856445312,
+      "learning_rate": 3.853620464904792e-06,
+      "loss": 0.3848,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8506869723555703,
+      "grad_norm": 3.2500317096710205,
+      "learning_rate": 3.844248270973639e-06,
+      "loss": 0.4319,
+      "step": 11180
+    },
+    {
+      "epoch": 1.8523423274292337,
+      "grad_norm": 2.082878351211548,
+      "learning_rate": 3.834880365056942e-06,
+      "loss": 0.4486,
+      "step": 11190
+    },
+    {
+      "epoch": 1.853997682502897,
+      "grad_norm": 2.111647605895996,
+      "learning_rate": 3.825516781911056e-06,
+      "loss": 0.4077,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8556530375765603,
+      "grad_norm": 1.6086077690124512,
+      "learning_rate": 3.816157556276295e-06,
+      "loss": 0.4098,
+      "step": 11210
+    },
+    {
+      "epoch": 1.8573083926502236,
+      "grad_norm": 2.252664566040039,
+      "learning_rate": 3.806802722876808e-06,
+      "loss": 0.4283,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8589637477238867,
+      "grad_norm": 2.4436564445495605,
+      "learning_rate": 3.79745231642045e-06,
+      "loss": 0.4302,
+      "step": 11230
+    },
+    {
+      "epoch": 1.86061910279755,
+      "grad_norm": 2.0436229705810547,
+      "learning_rate": 3.7881063715986466e-06,
+      "loss": 0.4131,
+      "step": 11240
+    },
+    {
+      "epoch": 1.8622744578712134,
+      "grad_norm": 1.9353731870651245,
+      "learning_rate": 3.7787649230862746e-06,
+      "loss": 0.3978,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8639298129448767,
+      "grad_norm": 1.7772618532180786,
+      "learning_rate": 3.769428005541525e-06,
+      "loss": 0.4193,
+      "step": 11260
+    },
+    {
+      "epoch": 1.86558516801854,
+      "grad_norm": 1.960239052772522,
+      "learning_rate": 3.76009565360578e-06,
+      "loss": 0.4456,
+      "step": 11270
+    },
+    {
+      "epoch": 1.8672405230922031,
+      "grad_norm": 2.3844010829925537,
+      "learning_rate": 3.7507679019034827e-06,
+      "loss": 0.4229,
+      "step": 11280
+    },
+    {
+      "epoch": 1.8688958781658664,
+      "grad_norm": 1.5547751188278198,
+      "learning_rate": 3.7414447850420116e-06,
+      "loss": 0.4223,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8705512332395298,
+      "grad_norm": 2.0283639430999756,
+      "learning_rate": 3.732126337611544e-06,
+      "loss": 0.4348,
+      "step": 11300
+    },
+    {
+      "epoch": 1.872206588313193,
+      "grad_norm": 2.0857224464416504,
+      "learning_rate": 3.7228125941849347e-06,
+      "loss": 0.4507,
+      "step": 11310
+    },
+    {
+      "epoch": 1.8738619433868564,
+      "grad_norm": 1.7453672885894775,
+      "learning_rate": 3.7135035893175873e-06,
+      "loss": 0.4225,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8755172984605197,
+      "grad_norm": 2.237485885620117,
+      "learning_rate": 3.7041993575473245e-06,
+      "loss": 0.4036,
+      "step": 11330
+    },
+    {
+      "epoch": 1.877172653534183,
+      "grad_norm": 1.7037478685379028,
+      "learning_rate": 3.6948999333942558e-06,
+      "loss": 0.391,
+      "step": 11340
+    },
+    {
+      "epoch": 1.8788280086078464,
+      "grad_norm": 2.2421579360961914,
+      "learning_rate": 3.6856053513606615e-06,
+      "loss": 0.3896,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8804833636815097,
+      "grad_norm": 1.633567452430725,
+      "learning_rate": 3.676315645930851e-06,
+      "loss": 0.4273,
+      "step": 11360
+    },
+    {
+      "epoch": 1.882138718755173,
+      "grad_norm": 1.8429170846939087,
+      "learning_rate": 3.667030851571043e-06,
+      "loss": 0.3861,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8837940738288363,
+      "grad_norm": 2.3093669414520264,
+      "learning_rate": 3.657751002729234e-06,
+      "loss": 0.3958,
+      "step": 11380
+    },
+    {
+      "epoch": 1.8854494289024997,
+      "grad_norm": 1.9372445344924927,
+      "learning_rate": 3.6484761338350703e-06,
+      "loss": 0.4537,
+      "step": 11390
+    },
+    {
+      "epoch": 1.887104783976163,
+      "grad_norm": 2.930211305618286,
+      "learning_rate": 3.6392062792997284e-06,
+      "loss": 0.3707,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8887601390498263,
+      "grad_norm": 1.6780815124511719,
+      "learning_rate": 3.6299414735157767e-06,
+      "loss": 0.3917,
+      "step": 11410
+    },
+    {
+      "epoch": 1.8904154941234896,
+      "grad_norm": 2.5266661643981934,
+      "learning_rate": 3.620681750857049e-06,
+      "loss": 0.4232,
+      "step": 11420
+    },
+    {
+      "epoch": 1.8920708491971527,
+      "grad_norm": 1.8783165216445923,
+      "learning_rate": 3.6114271456785237e-06,
+      "loss": 0.415,
+      "step": 11430
+    },
+    {
+      "epoch": 1.893726204270816,
+      "grad_norm": 1.8861968517303467,
+      "learning_rate": 3.6021776923161927e-06,
+      "loss": 0.3868,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8953815593444794,
+      "grad_norm": 1.9357882738113403,
+      "learning_rate": 3.5929334250869297e-06,
+      "loss": 0.4461,
+      "step": 11450
+    },
+    {
+      "epoch": 1.8970369144181427,
+      "grad_norm": 1.3567824363708496,
+      "learning_rate": 3.5836943782883747e-06,
+      "loss": 0.3844,
+      "step": 11460
+    },
+    {
+      "epoch": 1.898692269491806,
+      "grad_norm": 2.8328826427459717,
+      "learning_rate": 3.5744605861987925e-06,
+      "loss": 0.4203,
+      "step": 11470
+    },
+    {
+      "epoch": 1.9003476245654691,
+      "grad_norm": 2.007815361022949,
+      "learning_rate": 3.565232083076954e-06,
+      "loss": 0.3869,
+      "step": 11480
+    },
+    {
+      "epoch": 1.9020029796391325,
+      "grad_norm": 1.5280500650405884,
+      "learning_rate": 3.556008903162007e-06,
+      "loss": 0.4193,
+      "step": 11490
+    },
+    {
+      "epoch": 1.9036583347127958,
+      "grad_norm": 2.2297863960266113,
+      "learning_rate": 3.546791080673351e-06,
+      "loss": 0.408,
+      "step": 11500
+    },
+    {
+      "epoch": 1.905313689786459,
+      "grad_norm": 2.147474527359009,
+      "learning_rate": 3.537578649810505e-06,
+      "loss": 0.4071,
+      "step": 11510
+    },
+    {
+      "epoch": 1.9069690448601224,
+      "grad_norm": 2.1503307819366455,
+      "learning_rate": 3.52837164475299e-06,
+      "loss": 0.4037,
+      "step": 11520
+    },
+    {
+      "epoch": 1.9086243999337857,
+      "grad_norm": 2.1771047115325928,
+      "learning_rate": 3.519170099660192e-06,
+      "loss": 0.3948,
+      "step": 11530
+    },
+    {
+      "epoch": 1.910279755007449,
+      "grad_norm": 1.9087363481521606,
+      "learning_rate": 3.509974048671241e-06,
+      "loss": 0.4097,
+      "step": 11540
+    },
+    {
+      "epoch": 1.9119351100811124,
+      "grad_norm": 1.270500659942627,
+      "learning_rate": 3.5007835259048818e-06,
+      "loss": 0.395,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9135904651547757,
+      "grad_norm": 1.6682640314102173,
+      "learning_rate": 3.491598565459351e-06,
+      "loss": 0.3878,
+      "step": 11560
+    },
+    {
+      "epoch": 1.915245820228439,
+      "grad_norm": 1.9983856678009033,
+      "learning_rate": 3.482419201412246e-06,
+      "loss": 0.4232,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9169011753021024,
+      "grad_norm": 1.9866315126419067,
+      "learning_rate": 3.473245467820403e-06,
+      "loss": 0.4339,
+      "step": 11580
+    },
+    {
+      "epoch": 1.9185565303757657,
+      "grad_norm": 1.7992204427719116,
+      "learning_rate": 3.4640773987197662e-06,
+      "loss": 0.4178,
+      "step": 11590
+    },
+    {
+      "epoch": 1.920211885449429,
+      "grad_norm": 1.685878872871399,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.3865,
+      "step": 11600
+    },
+    {
+      "epoch": 1.9218672405230923,
+      "grad_norm": 1.3427140712738037,
+      "learning_rate": 3.4457583900306835e-06,
+      "loss": 0.4346,
+      "step": 11610
+    },
+    {
+      "epoch": 1.9235225955967556,
+      "grad_norm": 1.8562428951263428,
+      "learning_rate": 3.4366075184085403e-06,
+      "loss": 0.4133,
+      "step": 11620
+    },
+    {
+      "epoch": 1.925177950670419,
+      "grad_norm": 1.8822894096374512,
+      "learning_rate": 3.4274624472099626e-06,
+      "loss": 0.4147,
+      "step": 11630
+    },
+    {
+      "epoch": 1.926833305744082,
+      "grad_norm": 1.918009638786316,
+      "learning_rate": 3.4183232103645508e-06,
+      "loss": 0.392,
+      "step": 11640
+    },
+    {
+      "epoch": 1.9284886608177454,
+      "grad_norm": 1.8526767492294312,
+      "learning_rate": 3.409189841780263e-06,
+      "loss": 0.4166,
+      "step": 11650
+    },
+    {
+      "epoch": 1.9301440158914087,
+      "grad_norm": 2.163173198699951,
+      "learning_rate": 3.400062375343283e-06,
+      "loss": 0.4057,
+      "step": 11660
+    },
+    {
+      "epoch": 1.931799370965072,
+      "grad_norm": 2.5304856300354004,
+      "learning_rate": 3.390940844917897e-06,
+      "loss": 0.3884,
+      "step": 11670
+    },
+    {
+      "epoch": 1.9334547260387354,
+      "grad_norm": 2.086353302001953,
+      "learning_rate": 3.3818252843463676e-06,
+      "loss": 0.3999,
+      "step": 11680
+    },
+    {
+      "epoch": 1.9351100811123985,
+      "grad_norm": 1.9789005517959595,
+      "learning_rate": 3.3727157274488113e-06,
+      "loss": 0.4293,
+      "step": 11690
+    },
+    {
+      "epoch": 1.9367654361860618,
+      "grad_norm": 1.9313982725143433,
+      "learning_rate": 3.363612208023068e-06,
+      "loss": 0.3912,
+      "step": 11700
+    },
+    {
+      "epoch": 1.938420791259725,
+      "grad_norm": 1.6824219226837158,
+      "learning_rate": 3.354514759844576e-06,
+      "loss": 0.3633,
+      "step": 11710
+    },
+    {
+      "epoch": 1.9400761463333884,
+      "grad_norm": 2.064709424972534,
+      "learning_rate": 3.345423416666249e-06,
+      "loss": 0.3762,
+      "step": 11720
+    },
+    {
+      "epoch": 1.9417315014070518,
+      "grad_norm": 1.9601470232009888,
+      "learning_rate": 3.336338212218354e-06,
+      "loss": 0.3626,
+      "step": 11730
+    },
+    {
+      "epoch": 1.943386856480715,
+      "grad_norm": 1.799816608428955,
+      "learning_rate": 3.327259180208375e-06,
+      "loss": 0.3649,
+      "step": 11740
+    },
+    {
+      "epoch": 1.9450422115543784,
+      "grad_norm": 2.4325437545776367,
+      "learning_rate": 3.318186354320905e-06,
+      "loss": 0.3869,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9466975666280417,
+      "grad_norm": 1.9126996994018555,
+      "learning_rate": 3.3091197682175023e-06,
+      "loss": 0.4228,
+      "step": 11760
+    },
+    {
+      "epoch": 1.948352921701705,
+      "grad_norm": 1.600388526916504,
+      "learning_rate": 3.300059455536579e-06,
+      "loss": 0.3965,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9500082767753684,
+      "grad_norm": 2.1479532718658447,
+      "learning_rate": 3.291005449893273e-06,
+      "loss": 0.3816,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9516636318490317,
+      "grad_norm": 1.8163764476776123,
+      "learning_rate": 3.281957784879317e-06,
+      "loss": 0.4336,
+      "step": 11790
+    },
+    {
+      "epoch": 1.953318986922695,
+      "grad_norm": 2.1194796562194824,
+      "learning_rate": 3.2729164940629264e-06,
+      "loss": 0.4344,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9549743419963583,
+      "grad_norm": 1.6033542156219482,
+      "learning_rate": 3.2638816109886604e-06,
+      "loss": 0.3745,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9566296970700217,
+      "grad_norm": 2.0563807487487793,
+      "learning_rate": 3.254853169177311e-06,
+      "loss": 0.408,
+      "step": 11820
+    },
+    {
+      "epoch": 1.958285052143685,
+      "grad_norm": 2.165522813796997,
+      "learning_rate": 3.2458312021257656e-06,
+      "loss": 0.417,
+      "step": 11830
+    },
+    {
+      "epoch": 1.959940407217348,
+      "grad_norm": 1.5800554752349854,
+      "learning_rate": 3.2368157433068916e-06,
+      "loss": 0.4462,
+      "step": 11840
+    },
+    {
+      "epoch": 1.9615957622910114,
+      "grad_norm": 1.5213532447814941,
+      "learning_rate": 3.2278068261694106e-06,
+      "loss": 0.3837,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9632511173646747,
+      "grad_norm": 1.8690297603607178,
+      "learning_rate": 3.2188044841377773e-06,
+      "loss": 0.3819,
+      "step": 11860
+    },
+    {
+      "epoch": 1.964906472438338,
+      "grad_norm": 1.4503165483474731,
+      "learning_rate": 3.2098087506120456e-06,
+      "loss": 0.4349,
+      "step": 11870
+    },
+    {
+      "epoch": 1.9665618275120014,
+      "grad_norm": 1.9367406368255615,
+      "learning_rate": 3.2008196589677532e-06,
+      "loss": 0.465,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9682171825856645,
+      "grad_norm": 3.2155959606170654,
+      "learning_rate": 3.1918372425557932e-06,
+      "loss": 0.3958,
+      "step": 11890
+    },
+    {
+      "epoch": 1.9698725376593278,
+      "grad_norm": 2.281371593475342,
+      "learning_rate": 3.1828615347022984e-06,
+      "loss": 0.4041,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9715278927329911,
+      "grad_norm": 1.9571949243545532,
+      "learning_rate": 3.173892568708505e-06,
+      "loss": 0.3746,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9731832478066544,
+      "grad_norm": 1.9914932250976562,
+      "learning_rate": 3.1649303778506425e-06,
+      "loss": 0.4203,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9748386028803178,
+      "grad_norm": 1.7412394285202026,
+      "learning_rate": 3.1559749953797973e-06,
+      "loss": 0.3784,
+      "step": 11930
+    },
+    {
+      "epoch": 1.976493957953981,
+      "grad_norm": 1.6468333005905151,
+      "learning_rate": 3.147026454521801e-06,
+      "loss": 0.4356,
+      "step": 11940
+    },
+    {
+      "epoch": 1.9781493130276444,
+      "grad_norm": 1.9956331253051758,
+      "learning_rate": 3.138084788477098e-06,
+      "loss": 0.4486,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9798046681013077,
+      "grad_norm": 2.0449624061584473,
+      "learning_rate": 3.1291500304206262e-06,
+      "loss": 0.3309,
+      "step": 11960
+    },
+    {
+      "epoch": 1.981460023174971,
+      "grad_norm": 2.1764378547668457,
+      "learning_rate": 3.120222213501697e-06,
+      "loss": 0.4128,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9831153782486344,
+      "grad_norm": 1.891964077949524,
+      "learning_rate": 3.1113013708438653e-06,
+      "loss": 0.4075,
+      "step": 11980
+    },
+    {
+      "epoch": 1.9847707333222977,
+      "grad_norm": 2.1923632621765137,
+      "learning_rate": 3.1023875355448153e-06,
+      "loss": 0.4233,
+      "step": 11990
+    },
+    {
+      "epoch": 1.986426088395961,
+      "grad_norm": 2.2308120727539062,
+      "learning_rate": 3.093480740676228e-06,
+      "loss": 0.4074,
+      "step": 12000
+    },
+    {
+      "epoch": 1.9880814434696243,
+      "grad_norm": 2.1118459701538086,
+      "learning_rate": 3.0845810192836645e-06,
+      "loss": 0.423,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9897367985432877,
+      "grad_norm": 1.4029134511947632,
+      "learning_rate": 3.075688404386442e-06,
+      "loss": 0.3564,
+      "step": 12020
+    },
+    {
+      "epoch": 1.991392153616951,
+      "grad_norm": 1.9621480703353882,
+      "learning_rate": 3.0668029289775163e-06,
+      "loss": 0.3869,
+      "step": 12030
+    },
+    {
+      "epoch": 1.993047508690614,
+      "grad_norm": 1.8076422214508057,
+      "learning_rate": 3.0579246260233486e-06,
+      "loss": 0.4299,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9947028637642774,
+      "grad_norm": 2.1234450340270996,
+      "learning_rate": 3.04905352846379e-06,
+      "loss": 0.4006,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9963582188379407,
+      "grad_norm": 1.8213273286819458,
+      "learning_rate": 3.0401896692119626e-06,
+      "loss": 0.4036,
+      "step": 12060
+    },
+    {
+      "epoch": 1.998013573911604,
+      "grad_norm": 2.5499227046966553,
+      "learning_rate": 3.031333081154129e-06,
+      "loss": 0.4424,
+      "step": 12070
+    },
+    {
+      "epoch": 1.9996689289852674,
+      "grad_norm": 2.273587942123413,
+      "learning_rate": 3.0224837971495736e-06,
+      "loss": 0.4118,
+      "step": 12080
+    },
+    {
+      "epoch": 2.0013242840589305,
+      "grad_norm": 1.7035340070724487,
+      "learning_rate": 3.0136418500304888e-06,
+      "loss": 0.3556,
+      "step": 12090
+    },
+    {
+      "epoch": 2.002979639132594,
+      "grad_norm": 1.4942774772644043,
+      "learning_rate": 3.0048072726018386e-06,
+      "loss": 0.3315,
+      "step": 12100
+    },
+    {
+      "epoch": 2.004634994206257,
+      "grad_norm": 1.6278843879699707,
+      "learning_rate": 2.995980097641248e-06,
+      "loss": 0.3448,
+      "step": 12110
+    },
+    {
+      "epoch": 2.0062903492799204,
+      "grad_norm": 2.276960849761963,
+      "learning_rate": 2.987160357898877e-06,
+      "loss": 0.3363,
+      "step": 12120
+    },
+    {
+      "epoch": 2.0079457043535838,
+      "grad_norm": 1.993857502937317,
+      "learning_rate": 2.978348086097298e-06,
+      "loss": 0.3246,
+      "step": 12130
+    },
+    {
+      "epoch": 2.009601059427247,
+      "grad_norm": 1.7416425943374634,
+      "learning_rate": 2.9695433149313774e-06,
+      "loss": 0.3325,
+      "step": 12140
+    },
+    {
+      "epoch": 2.0112564145009104,
+      "grad_norm": 2.1714820861816406,
+      "learning_rate": 2.960746077068158e-06,
+      "loss": 0.3331,
+      "step": 12150
+    },
+    {
+      "epoch": 2.0129117695745737,
+      "grad_norm": 1.8540669679641724,
+      "learning_rate": 2.951956405146725e-06,
+      "loss": 0.3549,
+      "step": 12160
+    },
+    {
+      "epoch": 2.014567124648237,
+      "grad_norm": 1.6556892395019531,
+      "learning_rate": 2.9431743317780957e-06,
+      "loss": 0.3138,
+      "step": 12170
+    },
+    {
+      "epoch": 2.0162224797219004,
+      "grad_norm": 1.8258768320083618,
+      "learning_rate": 2.934399889545099e-06,
+      "loss": 0.3415,
+      "step": 12180
+    },
+    {
+      "epoch": 2.0178778347955637,
+      "grad_norm": 1.6194030046463013,
+      "learning_rate": 2.9256331110022463e-06,
+      "loss": 0.3436,
+      "step": 12190
+    },
+    {
+      "epoch": 2.019533189869227,
+      "grad_norm": 2.233779191970825,
+      "learning_rate": 2.9168740286756157e-06,
+      "loss": 0.3653,
+      "step": 12200
+    },
+    {
+      "epoch": 2.0211885449428904,
+      "grad_norm": 1.9311305284500122,
+      "learning_rate": 2.9081226750627367e-06,
+      "loss": 0.3901,
+      "step": 12210
+    },
+    {
+      "epoch": 2.0228439000165537,
+      "grad_norm": 2.0436785221099854,
+      "learning_rate": 2.89937908263246e-06,
+      "loss": 0.343,
+      "step": 12220
+    },
+    {
+      "epoch": 2.024499255090217,
+      "grad_norm": 1.8099346160888672,
+      "learning_rate": 2.890643283824837e-06,
+      "loss": 0.3493,
+      "step": 12230
+    },
+    {
+      "epoch": 2.0261546101638803,
+      "grad_norm": 2.596013307571411,
+      "learning_rate": 2.8819153110510147e-06,
+      "loss": 0.3721,
+      "step": 12240
+    },
+    {
+      "epoch": 2.0278099652375436,
+      "grad_norm": 2.1353633403778076,
+      "learning_rate": 2.8731951966930917e-06,
+      "loss": 0.361,
+      "step": 12250
+    },
+    {
+      "epoch": 2.029465320311207,
+      "grad_norm": 2.3608481884002686,
+      "learning_rate": 2.8644829731040214e-06,
+      "loss": 0.3404,
+      "step": 12260
+    },
+    {
+      "epoch": 2.03112067538487,
+      "grad_norm": 1.9163801670074463,
+      "learning_rate": 2.8557786726074755e-06,
+      "loss": 0.3844,
+      "step": 12270
+    },
+    {
+      "epoch": 2.032776030458533,
+      "grad_norm": 2.0741047859191895,
+      "learning_rate": 2.84708232749773e-06,
+      "loss": 0.3721,
+      "step": 12280
+    },
+    {
+      "epoch": 2.0344313855321965,
+      "grad_norm": 2.1986711025238037,
+      "learning_rate": 2.8383939700395456e-06,
+      "loss": 0.3018,
+      "step": 12290
+    },
+    {
+      "epoch": 2.03608674060586,
+      "grad_norm": 1.790903925895691,
+      "learning_rate": 2.8297136324680498e-06,
+      "loss": 0.3716,
+      "step": 12300
+    },
+    {
+      "epoch": 2.037742095679523,
+      "grad_norm": 1.8230623006820679,
+      "learning_rate": 2.8210413469886094e-06,
+      "loss": 0.3459,
+      "step": 12310
+    },
+    {
+      "epoch": 2.0393974507531865,
+      "grad_norm": 1.7864844799041748,
+      "learning_rate": 2.812377145776724e-06,
+      "loss": 0.3428,
+      "step": 12320
+    },
+    {
+      "epoch": 2.04105280582685,
+      "grad_norm": 1.8330215215682983,
+      "learning_rate": 2.8037210609778975e-06,
+      "loss": 0.3533,
+      "step": 12330
+    },
+    {
+      "epoch": 2.042708160900513,
+      "grad_norm": 2.0148110389709473,
+      "learning_rate": 2.795073124707518e-06,
+      "loss": 0.3093,
+      "step": 12340
+    },
+    {
+      "epoch": 2.0443635159741764,
+      "grad_norm": 1.9366060495376587,
+      "learning_rate": 2.786433369050742e-06,
+      "loss": 0.3289,
+      "step": 12350
+    },
+    {
+      "epoch": 2.0460188710478397,
+      "grad_norm": 3.5124638080596924,
+      "learning_rate": 2.777801826062375e-06,
+      "loss": 0.3476,
+      "step": 12360
+    },
+    {
+      "epoch": 2.047674226121503,
+      "grad_norm": 2.0207085609436035,
+      "learning_rate": 2.7691785277667506e-06,
+      "loss": 0.3595,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0493295811951664,
+      "grad_norm": 1.430230736732483,
+      "learning_rate": 2.7605635061576195e-06,
+      "loss": 0.3462,
+      "step": 12380
+    },
+    {
+      "epoch": 2.0509849362688297,
+      "grad_norm": 2.4573211669921875,
+      "learning_rate": 2.7519567931980185e-06,
+      "loss": 0.3782,
+      "step": 12390
+    },
+    {
+      "epoch": 2.052640291342493,
+      "grad_norm": 1.84562087059021,
+      "learning_rate": 2.7433584208201577e-06,
+      "loss": 0.3536,
+      "step": 12400
+    },
+    {
+      "epoch": 2.0542956464161564,
+      "grad_norm": 1.9053126573562622,
+      "learning_rate": 2.734768420925308e-06,
+      "loss": 0.3187,
+      "step": 12410
+    },
+    {
+      "epoch": 2.0559510014898197,
+      "grad_norm": 2.1721508502960205,
+      "learning_rate": 2.726186825383673e-06,
+      "loss": 0.3245,
+      "step": 12420
+    },
+    {
+      "epoch": 2.057606356563483,
+      "grad_norm": 2.5854411125183105,
+      "learning_rate": 2.717613666034272e-06,
+      "loss": 0.3681,
+      "step": 12430
+    },
+    {
+      "epoch": 2.0592617116371463,
+      "grad_norm": 1.7167317867279053,
+      "learning_rate": 2.7090489746848336e-06,
+      "loss": 0.3286,
+      "step": 12440
+    },
+    {
+      "epoch": 2.0609170667108097,
+      "grad_norm": 1.8206017017364502,
+      "learning_rate": 2.7004927831116614e-06,
+      "loss": 0.3294,
+      "step": 12450
+    },
+    {
+      "epoch": 2.062572421784473,
+      "grad_norm": 1.7760072946548462,
+      "learning_rate": 2.691945123059525e-06,
+      "loss": 0.3465,
+      "step": 12460
+    },
+    {
+      "epoch": 2.064227776858136,
+      "grad_norm": 2.4801549911499023,
+      "learning_rate": 2.6834060262415425e-06,
+      "loss": 0.3392,
+      "step": 12470
+    },
+    {
+      "epoch": 2.065883131931799,
+      "grad_norm": 2.1817026138305664,
+      "learning_rate": 2.674875524339057e-06,
+      "loss": 0.3275,
+      "step": 12480
+    },
+    {
+      "epoch": 2.0675384870054625,
+      "grad_norm": 1.8844431638717651,
+      "learning_rate": 2.6663536490015284e-06,
+      "loss": 0.3257,
+      "step": 12490
+    },
+    {
+      "epoch": 2.069193842079126,
+      "grad_norm": 2.3225162029266357,
+      "learning_rate": 2.6578404318464112e-06,
+      "loss": 0.3575,
+      "step": 12500
+    },
+    {
+      "epoch": 2.070849197152789,
+      "grad_norm": 1.740742802619934,
+      "learning_rate": 2.6493359044590317e-06,
+      "loss": 0.3333,
+      "step": 12510
+    },
+    {
+      "epoch": 2.0725045522264525,
+      "grad_norm": 2.0152180194854736,
+      "learning_rate": 2.640840098392478e-06,
+      "loss": 0.3601,
+      "step": 12520
+    },
+    {
+      "epoch": 2.074159907300116,
+      "grad_norm": 2.1492371559143066,
+      "learning_rate": 2.6323530451674815e-06,
+      "loss": 0.3131,
+      "step": 12530
+    },
+    {
+      "epoch": 2.075815262373779,
+      "grad_norm": 2.565166473388672,
+      "learning_rate": 2.623874776272296e-06,
+      "loss": 0.3644,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0774706174474424,
+      "grad_norm": 2.024240493774414,
+      "learning_rate": 2.615405323162592e-06,
+      "loss": 0.2927,
+      "step": 12550
+    },
+    {
+      "epoch": 2.0791259725211058,
+      "grad_norm": 1.9475926160812378,
+      "learning_rate": 2.606944717261323e-06,
+      "loss": 0.3036,
+      "step": 12560
+    },
+    {
+      "epoch": 2.080781327594769,
+      "grad_norm": 2.2539281845092773,
+      "learning_rate": 2.598492989958622e-06,
+      "loss": 0.3522,
+      "step": 12570
+    },
+    {
+      "epoch": 2.0824366826684324,
+      "grad_norm": 1.7754216194152832,
+      "learning_rate": 2.590050172611684e-06,
+      "loss": 0.3114,
+      "step": 12580
+    },
+    {
+      "epoch": 2.0840920377420957,
+      "grad_norm": 1.9726289510726929,
+      "learning_rate": 2.5816162965446416e-06,
+      "loss": 0.3297,
+      "step": 12590
+    },
+    {
+      "epoch": 2.085747392815759,
+      "grad_norm": 1.9303443431854248,
+      "learning_rate": 2.5731913930484543e-06,
+      "loss": 0.3168,
+      "step": 12600
+    },
+    {
+      "epoch": 2.0874027478894224,
+      "grad_norm": 3.2593839168548584,
+      "learning_rate": 2.564775493380798e-06,
+      "loss": 0.3679,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0890581029630857,
+      "grad_norm": 2.273982524871826,
+      "learning_rate": 2.556368628765935e-06,
+      "loss": 0.3396,
+      "step": 12620
+    },
+    {
+      "epoch": 2.090713458036749,
+      "grad_norm": 2.084463119506836,
+      "learning_rate": 2.5479708303946126e-06,
+      "loss": 0.3432,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0923688131104123,
+      "grad_norm": 1.9511936902999878,
+      "learning_rate": 2.539582129423934e-06,
+      "loss": 0.3304,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0940241681840757,
+      "grad_norm": 3.004499912261963,
+      "learning_rate": 2.531202556977259e-06,
+      "loss": 0.3302,
+      "step": 12650
+    },
+    {
+      "epoch": 2.095679523257739,
+      "grad_norm": 2.4408180713653564,
+      "learning_rate": 2.5228321441440705e-06,
+      "loss": 0.3696,
+      "step": 12660
+    },
+    {
+      "epoch": 2.0973348783314023,
+      "grad_norm": 1.5785390138626099,
+      "learning_rate": 2.514470921979874e-06,
+      "loss": 0.3248,
+      "step": 12670
+    },
+    {
+      "epoch": 2.098990233405065,
+      "grad_norm": 2.020374059677124,
+      "learning_rate": 2.5061189215060733e-06,
+      "loss": 0.3153,
+      "step": 12680
+    },
+    {
+      "epoch": 2.1006455884787285,
+      "grad_norm": 4.800167560577393,
+      "learning_rate": 2.497776173709857e-06,
+      "loss": 0.3579,
+      "step": 12690
+    },
+    {
+      "epoch": 2.102300943552392,
+      "grad_norm": 2.8129446506500244,
+      "learning_rate": 2.4894427095440883e-06,
+      "loss": 0.3541,
+      "step": 12700
+    },
+    {
+      "epoch": 2.103956298626055,
+      "grad_norm": 2.103689670562744,
+      "learning_rate": 2.4811185599271813e-06,
+      "loss": 0.3655,
+      "step": 12710
+    },
+    {
+      "epoch": 2.1056116536997185,
+      "grad_norm": 2.179983139038086,
+      "learning_rate": 2.4728037557430012e-06,
+      "loss": 0.381,
+      "step": 12720
+    },
+    {
+      "epoch": 2.107267008773382,
+      "grad_norm": 1.8805456161499023,
+      "learning_rate": 2.464498327840729e-06,
+      "loss": 0.3306,
+      "step": 12730
+    },
+    {
+      "epoch": 2.108922363847045,
+      "grad_norm": 1.8576078414916992,
+      "learning_rate": 2.4562023070347685e-06,
+      "loss": 0.3927,
+      "step": 12740
+    },
+    {
+      "epoch": 2.1105777189207084,
+      "grad_norm": 2.150397300720215,
+      "learning_rate": 2.4479157241046143e-06,
+      "loss": 0.3269,
+      "step": 12750
+    },
+    {
+      "epoch": 2.1122330739943718,
+      "grad_norm": 1.8863481283187866,
+      "learning_rate": 2.4396386097947482e-06,
+      "loss": 0.3511,
+      "step": 12760
+    },
+    {
+      "epoch": 2.113888429068035,
+      "grad_norm": 1.7663947343826294,
+      "learning_rate": 2.4313709948145193e-06,
+      "loss": 0.3388,
+      "step": 12770
+    },
+    {
+      "epoch": 2.1155437841416984,
+      "grad_norm": 2.267031669616699,
+      "learning_rate": 2.423112909838039e-06,
+      "loss": 0.3502,
+      "step": 12780
+    },
+    {
+      "epoch": 2.1171991392153617,
+      "grad_norm": 2.0435879230499268,
+      "learning_rate": 2.4148643855040543e-06,
+      "loss": 0.3621,
+      "step": 12790
+    },
+    {
+      "epoch": 2.118854494289025,
+      "grad_norm": 2.001474618911743,
+      "learning_rate": 2.406625452415844e-06,
+      "loss": 0.3321,
+      "step": 12800
+    },
+    {
+      "epoch": 2.1205098493626884,
+      "grad_norm": 1.835279941558838,
+      "learning_rate": 2.398396141141099e-06,
+      "loss": 0.369,
+      "step": 12810
+    },
+    {
+      "epoch": 2.1221652044363517,
+      "grad_norm": 1.7734911441802979,
+      "learning_rate": 2.390176482211818e-06,
+      "loss": 0.34,
+      "step": 12820
+    },
+    {
+      "epoch": 2.123820559510015,
+      "grad_norm": 2.4885096549987793,
+      "learning_rate": 2.3819665061241798e-06,
+      "loss": 0.3385,
+      "step": 12830
+    },
+    {
+      "epoch": 2.1254759145836783,
+      "grad_norm": 1.844212293624878,
+      "learning_rate": 2.3737662433384474e-06,
+      "loss": 0.3229,
+      "step": 12840
+    },
+    {
+      "epoch": 2.1271312696573417,
+      "grad_norm": 1.8578639030456543,
+      "learning_rate": 2.3655757242788385e-06,
+      "loss": 0.354,
+      "step": 12850
+    },
+    {
+      "epoch": 2.128786624731005,
+      "grad_norm": 2.1224701404571533,
+      "learning_rate": 2.357394979333423e-06,
+      "loss": 0.3277,
+      "step": 12860
+    },
+    {
+      "epoch": 2.130441979804668,
+      "grad_norm": 1.8834820985794067,
+      "learning_rate": 2.349224038854007e-06,
+      "loss": 0.3472,
+      "step": 12870
+    },
+    {
+      "epoch": 2.132097334878331,
+      "grad_norm": 1.910531759262085,
+      "learning_rate": 2.3410629331560197e-06,
+      "loss": 0.3553,
+      "step": 12880
+    },
+    {
+      "epoch": 2.1337526899519945,
+      "grad_norm": 2.2018206119537354,
+      "learning_rate": 2.332911692518406e-06,
+      "loss": 0.3408,
+      "step": 12890
+    },
+    {
+      "epoch": 2.135408045025658,
+      "grad_norm": 1.6149979829788208,
+      "learning_rate": 2.3247703471835028e-06,
+      "loss": 0.3158,
+      "step": 12900
+    },
+    {
+      "epoch": 2.137063400099321,
+      "grad_norm": 2.036621570587158,
+      "learning_rate": 2.3166389273569416e-06,
+      "loss": 0.3459,
+      "step": 12910
+    },
+    {
+      "epoch": 2.1387187551729845,
+      "grad_norm": 2.2086009979248047,
+      "learning_rate": 2.3085174632075234e-06,
+      "loss": 0.3457,
+      "step": 12920
+    },
+    {
+      "epoch": 2.140374110246648,
+      "grad_norm": 1.6859116554260254,
+      "learning_rate": 2.3004059848671133e-06,
+      "loss": 0.3355,
+      "step": 12930
+    },
+    {
+      "epoch": 2.142029465320311,
+      "grad_norm": 2.271038770675659,
+      "learning_rate": 2.2923045224305267e-06,
+      "loss": 0.3251,
+      "step": 12940
+    },
+    {
+      "epoch": 2.1436848203939745,
+      "grad_norm": 1.8713353872299194,
+      "learning_rate": 2.2842131059554228e-06,
+      "loss": 0.3506,
+      "step": 12950
+    },
+    {
+      "epoch": 2.1453401754676378,
+      "grad_norm": 2.219752550125122,
+      "learning_rate": 2.2761317654621853e-06,
+      "loss": 0.3681,
+      "step": 12960
+    },
+    {
+      "epoch": 2.146995530541301,
+      "grad_norm": 2.1416406631469727,
+      "learning_rate": 2.2680605309338103e-06,
+      "loss": 0.3645,
+      "step": 12970
+    },
+    {
+      "epoch": 2.1486508856149644,
+      "grad_norm": 1.8972928524017334,
+      "learning_rate": 2.2599994323158103e-06,
+      "loss": 0.3625,
+      "step": 12980
+    },
+    {
+      "epoch": 2.1503062406886277,
+      "grad_norm": 1.7961077690124512,
+      "learning_rate": 2.2519484995160818e-06,
+      "loss": 0.3336,
+      "step": 12990
+    },
+    {
+      "epoch": 2.151961595762291,
+      "grad_norm": 2.188765287399292,
+      "learning_rate": 2.2439077624048074e-06,
+      "loss": 0.3458,
+      "step": 13000
+    },
+    {
+      "epoch": 2.1536169508359544,
+      "grad_norm": 1.96406888961792,
+      "learning_rate": 2.2358772508143467e-06,
+      "loss": 0.3536,
+      "step": 13010
+    },
+    {
+      "epoch": 2.1552723059096177,
+      "grad_norm": 2.556922435760498,
+      "learning_rate": 2.227856994539115e-06,
+      "loss": 0.3177,
+      "step": 13020
+    },
+    {
+      "epoch": 2.156927660983281,
+      "grad_norm": 2.1325490474700928,
+      "learning_rate": 2.2198470233354817e-06,
+      "loss": 0.3601,
+      "step": 13030
+    },
+    {
+      "epoch": 2.1585830160569444,
+      "grad_norm": 2.0026392936706543,
+      "learning_rate": 2.2118473669216568e-06,
+      "loss": 0.3097,
+      "step": 13040
+    },
+    {
+      "epoch": 2.1602383711306077,
+      "grad_norm": 2.2289645671844482,
+      "learning_rate": 2.203858054977578e-06,
+      "loss": 0.4102,
+      "step": 13050
+    },
+    {
+      "epoch": 2.161893726204271,
+      "grad_norm": 1.3882899284362793,
+      "learning_rate": 2.1958791171448083e-06,
+      "loss": 0.3555,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1635490812779343,
+      "grad_norm": 1.9401888847351074,
+      "learning_rate": 2.1879105830264213e-06,
+      "loss": 0.3216,
+      "step": 13070
+    },
+    {
+      "epoch": 2.1652044363515976,
+      "grad_norm": 2.3579916954040527,
+      "learning_rate": 2.179952482186886e-06,
+      "loss": 0.3729,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1668597914252605,
+      "grad_norm": 1.7595404386520386,
+      "learning_rate": 2.1720048441519665e-06,
+      "loss": 0.3457,
+      "step": 13090
+    },
+    {
+      "epoch": 2.168515146498924,
+      "grad_norm": 1.9767661094665527,
+      "learning_rate": 2.164067698408606e-06,
+      "loss": 0.3442,
+      "step": 13100
+    },
+    {
+      "epoch": 2.170170501572587,
+      "grad_norm": 2.1314504146575928,
+      "learning_rate": 2.1561410744048213e-06,
+      "loss": 0.3674,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1718258566462505,
+      "grad_norm": 2.088491439819336,
+      "learning_rate": 2.148225001549589e-06,
+      "loss": 0.3049,
+      "step": 13120
+    },
+    {
+      "epoch": 2.173481211719914,
+      "grad_norm": 1.943539023399353,
+      "learning_rate": 2.140319509212746e-06,
+      "loss": 0.3338,
+      "step": 13130
+    },
+    {
+      "epoch": 2.175136566793577,
+      "grad_norm": 1.8165936470031738,
+      "learning_rate": 2.1324246267248657e-06,
+      "loss": 0.343,
+      "step": 13140
+    },
+    {
+      "epoch": 2.1767919218672405,
+      "grad_norm": 1.5986518859863281,
+      "learning_rate": 2.124540383377165e-06,
+      "loss": 0.3654,
+      "step": 13150
+    },
+    {
+      "epoch": 2.178447276940904,
+      "grad_norm": 2.2084920406341553,
+      "learning_rate": 2.1166668084213822e-06,
+      "loss": 0.3188,
+      "step": 13160
+    },
+    {
+      "epoch": 2.180102632014567,
+      "grad_norm": 1.8947356939315796,
+      "learning_rate": 2.1088039310696744e-06,
+      "loss": 0.3347,
+      "step": 13170
+    },
+    {
+      "epoch": 2.1817579870882304,
+      "grad_norm": 1.6848961114883423,
+      "learning_rate": 2.1009517804945146e-06,
+      "loss": 0.3666,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1834133421618938,
+      "grad_norm": 1.5556492805480957,
+      "learning_rate": 2.0931103858285725e-06,
+      "loss": 0.3552,
+      "step": 13190
+    },
+    {
+      "epoch": 2.185068697235557,
+      "grad_norm": 1.661304235458374,
+      "learning_rate": 2.0852797761646125e-06,
+      "loss": 0.3119,
+      "step": 13200
+    },
+    {
+      "epoch": 2.1867240523092204,
+      "grad_norm": 2.146554470062256,
+      "learning_rate": 2.0774599805553873e-06,
+      "loss": 0.3636,
+      "step": 13210
+    },
+    {
+      "epoch": 2.1883794073828837,
+      "grad_norm": 2.547508478164673,
+      "learning_rate": 2.069651028013523e-06,
+      "loss": 0.3619,
+      "step": 13220
+    },
+    {
+      "epoch": 2.190034762456547,
+      "grad_norm": 1.6659605503082275,
+      "learning_rate": 2.0618529475114218e-06,
+      "loss": 0.2849,
+      "step": 13230
+    },
+    {
+      "epoch": 2.1916901175302104,
+      "grad_norm": 2.031393051147461,
+      "learning_rate": 2.054065767981149e-06,
+      "loss": 0.3449,
+      "step": 13240
+    },
+    {
+      "epoch": 2.1933454726038737,
+      "grad_norm": 1.7782469987869263,
+      "learning_rate": 2.0462895183143217e-06,
+      "loss": 0.3301,
+      "step": 13250
+    },
+    {
+      "epoch": 2.195000827677537,
+      "grad_norm": 2.75504994392395,
+      "learning_rate": 2.038524227362007e-06,
+      "loss": 0.3668,
+      "step": 13260
+    },
+    {
+      "epoch": 2.1966561827512003,
+      "grad_norm": 2.0453379154205322,
+      "learning_rate": 2.0307699239346136e-06,
+      "loss": 0.366,
+      "step": 13270
+    },
+    {
+      "epoch": 2.198311537824863,
+      "grad_norm": 2.0684151649475098,
+      "learning_rate": 2.023026636801785e-06,
+      "loss": 0.361,
+      "step": 13280
+    },
+    {
+      "epoch": 2.1999668928985265,
+      "grad_norm": 2.427762746810913,
+      "learning_rate": 2.0152943946922904e-06,
+      "loss": 0.3451,
+      "step": 13290
+    },
+    {
+      "epoch": 2.20162224797219,
+      "grad_norm": 1.8866089582443237,
+      "learning_rate": 2.007573226293927e-06,
+      "loss": 0.3214,
+      "step": 13300
+    },
+    {
+      "epoch": 2.203277603045853,
+      "grad_norm": 2.268911600112915,
+      "learning_rate": 1.999863160253398e-06,
+      "loss": 0.3247,
+      "step": 13310
+    },
+    {
+      "epoch": 2.2049329581195165,
+      "grad_norm": 2.1680514812469482,
+      "learning_rate": 1.992164225176223e-06,
+      "loss": 0.3316,
+      "step": 13320
+    },
+    {
+      "epoch": 2.20658831319318,
+      "grad_norm": 1.8509600162506104,
+      "learning_rate": 1.9844764496266195e-06,
+      "loss": 0.3393,
+      "step": 13330
+    },
+    {
+      "epoch": 2.208243668266843,
+      "grad_norm": 1.7594846487045288,
+      "learning_rate": 1.9767998621274014e-06,
+      "loss": 0.3732,
+      "step": 13340
+    },
+    {
+      "epoch": 2.2098990233405065,
+      "grad_norm": 2.6831443309783936,
+      "learning_rate": 1.969134491159873e-06,
+      "loss": 0.3263,
+      "step": 13350
+    },
+    {
+      "epoch": 2.21155437841417,
+      "grad_norm": 1.7843753099441528,
+      "learning_rate": 1.961480365163729e-06,
+      "loss": 0.3278,
+      "step": 13360
+    },
+    {
+      "epoch": 2.213209733487833,
+      "grad_norm": 2.025347948074341,
+      "learning_rate": 1.953837512536936e-06,
+      "loss": 0.3137,
+      "step": 13370
+    },
+    {
+      "epoch": 2.2148650885614964,
+      "grad_norm": 2.1478426456451416,
+      "learning_rate": 1.9462059616356377e-06,
+      "loss": 0.3597,
+      "step": 13380
+    },
+    {
+      "epoch": 2.2165204436351598,
+      "grad_norm": 2.198599100112915,
+      "learning_rate": 1.9385857407740504e-06,
+      "loss": 0.3527,
+      "step": 13390
+    },
+    {
+      "epoch": 2.218175798708823,
+      "grad_norm": 2.009014844894409,
+      "learning_rate": 1.9309768782243473e-06,
+      "loss": 0.3131,
+      "step": 13400
+    },
+    {
+      "epoch": 2.2198311537824864,
+      "grad_norm": 2.736865997314453,
+      "learning_rate": 1.9233794022165674e-06,
+      "loss": 0.3361,
+      "step": 13410
+    },
+    {
+      "epoch": 2.2214865088561497,
+      "grad_norm": 1.7309480905532837,
+      "learning_rate": 1.9157933409384993e-06,
+      "loss": 0.3225,
+      "step": 13420
+    },
+    {
+      "epoch": 2.223141863929813,
+      "grad_norm": 2.736997127532959,
+      "learning_rate": 1.908218722535582e-06,
+      "loss": 0.3144,
+      "step": 13430
+    },
+    {
+      "epoch": 2.2247972190034764,
+      "grad_norm": 1.6784467697143555,
+      "learning_rate": 1.9006555751108001e-06,
+      "loss": 0.34,
+      "step": 13440
+    },
+    {
+      "epoch": 2.2264525740771397,
+      "grad_norm": 2.2146949768066406,
+      "learning_rate": 1.8931039267245803e-06,
+      "loss": 0.3752,
+      "step": 13450
+    },
+    {
+      "epoch": 2.228107929150803,
+      "grad_norm": 1.708247423171997,
+      "learning_rate": 1.8855638053946823e-06,
+      "loss": 0.3439,
+      "step": 13460
+    },
+    {
+      "epoch": 2.2297632842244663,
+      "grad_norm": 2.3849120140075684,
+      "learning_rate": 1.8780352390961042e-06,
+      "loss": 0.3661,
+      "step": 13470
+    },
+    {
+      "epoch": 2.2314186392981297,
+      "grad_norm": 2.215123176574707,
+      "learning_rate": 1.8705182557609714e-06,
+      "loss": 0.3341,
+      "step": 13480
+    },
+    {
+      "epoch": 2.2330739943717925,
+      "grad_norm": 3.159491777420044,
+      "learning_rate": 1.8630128832784323e-06,
+      "loss": 0.3365,
+      "step": 13490
+    },
+    {
+      "epoch": 2.234729349445456,
+      "grad_norm": 1.2012838125228882,
+      "learning_rate": 1.8555191494945586e-06,
+      "loss": 0.3193,
+      "step": 13500
+    },
+    {
+      "epoch": 2.236384704519119,
+      "grad_norm": 2.053283452987671,
+      "learning_rate": 1.8480370822122412e-06,
+      "loss": 0.3612,
+      "step": 13510
+    },
+    {
+      "epoch": 2.2380400595927825,
+      "grad_norm": 2.0240302085876465,
+      "learning_rate": 1.8405667091910845e-06,
+      "loss": 0.3702,
+      "step": 13520
+    },
+    {
+      "epoch": 2.239695414666446,
+      "grad_norm": 2.051102638244629,
+      "learning_rate": 1.83310805814731e-06,
+      "loss": 0.3525,
+      "step": 13530
+    },
+    {
+      "epoch": 2.241350769740109,
+      "grad_norm": 2.752277135848999,
+      "learning_rate": 1.8256611567536442e-06,
+      "loss": 0.3599,
+      "step": 13540
+    },
+    {
+      "epoch": 2.2430061248137725,
+      "grad_norm": 1.6095753908157349,
+      "learning_rate": 1.8182260326392208e-06,
+      "loss": 0.2926,
+      "step": 13550
+    },
+    {
+      "epoch": 2.244661479887436,
+      "grad_norm": 2.0322036743164062,
+      "learning_rate": 1.8108027133894828e-06,
+      "loss": 0.3475,
+      "step": 13560
+    },
+    {
+      "epoch": 2.246316834961099,
+      "grad_norm": 1.9392104148864746,
+      "learning_rate": 1.8033912265460695e-06,
+      "loss": 0.3489,
+      "step": 13570
+    },
+    {
+      "epoch": 2.2479721900347625,
+      "grad_norm": 1.6775577068328857,
+      "learning_rate": 1.7959915996067256e-06,
+      "loss": 0.3406,
+      "step": 13580
+    },
+    {
+      "epoch": 2.2496275451084258,
+      "grad_norm": 1.9670125246047974,
+      "learning_rate": 1.7886038600251888e-06,
+      "loss": 0.3347,
+      "step": 13590
+    },
+    {
+      "epoch": 2.251282900182089,
+      "grad_norm": 2.1814675331115723,
+      "learning_rate": 1.7812280352110956e-06,
+      "loss": 0.3371,
+      "step": 13600
+    },
+    {
+      "epoch": 2.2529382552557524,
+      "grad_norm": 1.8385521173477173,
+      "learning_rate": 1.7738641525298766e-06,
+      "loss": 0.3531,
+      "step": 13610
+    },
+    {
+      "epoch": 2.2545936103294157,
+      "grad_norm": 1.961987018585205,
+      "learning_rate": 1.7665122393026523e-06,
+      "loss": 0.3291,
+      "step": 13620
+    },
+    {
+      "epoch": 2.256248965403079,
+      "grad_norm": 2.28271746635437,
+      "learning_rate": 1.759172322806142e-06,
+      "loss": 0.3615,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2579043204767424,
+      "grad_norm": 1.908655047416687,
+      "learning_rate": 1.7518444302725467e-06,
+      "loss": 0.3458,
+      "step": 13640
+    },
+    {
+      "epoch": 2.2595596755504057,
+      "grad_norm": 2.255382776260376,
+      "learning_rate": 1.7445285888894641e-06,
+      "loss": 0.3442,
+      "step": 13650
+    },
+    {
+      "epoch": 2.261215030624069,
+      "grad_norm": 1.4033169746398926,
+      "learning_rate": 1.7372248257997753e-06,
+      "loss": 0.3455,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2628703856977324,
+      "grad_norm": 2.0262038707733154,
+      "learning_rate": 1.7299331681015508e-06,
+      "loss": 0.3265,
+      "step": 13670
+    },
+    {
+      "epoch": 2.2645257407713952,
+      "grad_norm": 2.379423141479492,
+      "learning_rate": 1.722653642847948e-06,
+      "loss": 0.3533,
+      "step": 13680
+    },
+    {
+      "epoch": 2.2661810958450586,
+      "grad_norm": 1.8121516704559326,
+      "learning_rate": 1.7153862770471096e-06,
+      "loss": 0.3274,
+      "step": 13690
+    },
+    {
+      "epoch": 2.267836450918722,
+      "grad_norm": 2.4560039043426514,
+      "learning_rate": 1.7081310976620696e-06,
+      "loss": 0.3399,
+      "step": 13700
+    },
+    {
+      "epoch": 2.269491805992385,
+      "grad_norm": 1.8034098148345947,
+      "learning_rate": 1.700888131610643e-06,
+      "loss": 0.3056,
+      "step": 13710
+    },
+    {
+      "epoch": 2.2711471610660485,
+      "grad_norm": 1.6858197450637817,
+      "learning_rate": 1.6936574057653366e-06,
+      "loss": 0.3334,
+      "step": 13720
+    },
+    {
+      "epoch": 2.272802516139712,
+      "grad_norm": 1.8124359846115112,
+      "learning_rate": 1.6864389469532393e-06,
+      "loss": 0.3583,
+      "step": 13730
+    },
+    {
+      "epoch": 2.274457871213375,
+      "grad_norm": 1.7046393156051636,
+      "learning_rate": 1.6792327819559313e-06,
+      "loss": 0.332,
+      "step": 13740
+    },
+    {
+      "epoch": 2.2761132262870385,
+      "grad_norm": 2.347473382949829,
+      "learning_rate": 1.672038937509376e-06,
+      "loss": 0.4029,
+      "step": 13750
+    },
+    {
+      "epoch": 2.277768581360702,
+      "grad_norm": 1.9951478242874146,
+      "learning_rate": 1.6648574403038325e-06,
+      "loss": 0.3345,
+      "step": 13760
+    },
+    {
+      "epoch": 2.279423936434365,
+      "grad_norm": 2.5333080291748047,
+      "learning_rate": 1.657688316983746e-06,
+      "loss": 0.3114,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2810792915080285,
+      "grad_norm": 1.9064871072769165,
+      "learning_rate": 1.6505315941476507e-06,
+      "loss": 0.3244,
+      "step": 13780
+    },
+    {
+      "epoch": 2.282734646581692,
+      "grad_norm": 1.6432058811187744,
+      "learning_rate": 1.6433872983480758e-06,
+      "loss": 0.3372,
+      "step": 13790
+    },
+    {
+      "epoch": 2.284390001655355,
+      "grad_norm": 1.8521987199783325,
+      "learning_rate": 1.636255456091444e-06,
+      "loss": 0.3509,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2860453567290184,
+      "grad_norm": 1.8956800699234009,
+      "learning_rate": 1.6291360938379752e-06,
+      "loss": 0.3816,
+      "step": 13810
+    },
+    {
+      "epoch": 2.2877007118026818,
+      "grad_norm": 2.190884828567505,
+      "learning_rate": 1.622029238001584e-06,
+      "loss": 0.3157,
+      "step": 13820
+    },
+    {
+      "epoch": 2.289356066876345,
+      "grad_norm": 2.1423823833465576,
+      "learning_rate": 1.6149349149497833e-06,
+      "loss": 0.2804,
+      "step": 13830
+    },
+    {
+      "epoch": 2.2910114219500084,
+      "grad_norm": 2.221036672592163,
+      "learning_rate": 1.607853151003591e-06,
+      "loss": 0.3681,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2926667770236717,
+      "grad_norm": 2.1888656616210938,
+      "learning_rate": 1.6007839724374253e-06,
+      "loss": 0.3536,
+      "step": 13850
+    },
+    {
+      "epoch": 2.294322132097335,
+      "grad_norm": 2.2025184631347656,
+      "learning_rate": 1.593727405479012e-06,
+      "loss": 0.3667,
+      "step": 13860
+    },
+    {
+      "epoch": 2.2959774871709984,
+      "grad_norm": 1.9448472261428833,
+      "learning_rate": 1.5866834763092885e-06,
+      "loss": 0.3342,
+      "step": 13870
+    },
+    {
+      "epoch": 2.2976328422446617,
+      "grad_norm": 2.2824110984802246,
+      "learning_rate": 1.5796522110623002e-06,
+      "loss": 0.3581,
+      "step": 13880
+    },
+    {
+      "epoch": 2.299288197318325,
+      "grad_norm": 2.112283706665039,
+      "learning_rate": 1.5726336358251104e-06,
+      "loss": 0.3352,
+      "step": 13890
+    },
+    {
+      "epoch": 2.300943552391988,
+      "grad_norm": 2.393695592880249,
+      "learning_rate": 1.5656277766376992e-06,
+      "loss": 0.3287,
+      "step": 13900
+    },
+    {
+      "epoch": 2.302598907465651,
+      "grad_norm": 2.341130018234253,
+      "learning_rate": 1.5586346594928675e-06,
+      "loss": 0.3384,
+      "step": 13910
+    },
+    {
+      "epoch": 2.3042542625393145,
+      "grad_norm": 2.0110530853271484,
+      "learning_rate": 1.5516543103361403e-06,
+      "loss": 0.3276,
+      "step": 13920
+    },
+    {
+      "epoch": 2.305909617612978,
+      "grad_norm": 3.0257723331451416,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.3544,
+      "step": 13930
+    },
+    {
+      "epoch": 2.307564972686641,
+      "grad_norm": 1.825762391090393,
+      "learning_rate": 1.5377320195321642e-06,
+      "loss": 0.3448,
+      "step": 13940
+    },
+    {
+      "epoch": 2.3092203277603045,
+      "grad_norm": 1.9047185182571411,
+      "learning_rate": 1.5307901295387268e-06,
+      "loss": 0.3239,
+      "step": 13950
+    },
+    {
+      "epoch": 2.310875682833968,
+      "grad_norm": 2.2745182514190674,
+      "learning_rate": 1.5238611108408292e-06,
+      "loss": 0.3447,
+      "step": 13960
+    },
+    {
+      "epoch": 2.312531037907631,
+      "grad_norm": 2.037247896194458,
+      "learning_rate": 1.5169449891461867e-06,
+      "loss": 0.3188,
+      "step": 13970
+    },
+    {
+      "epoch": 2.3141863929812945,
+      "grad_norm": 2.3445773124694824,
+      "learning_rate": 1.5100417901146585e-06,
+      "loss": 0.3335,
+      "step": 13980
+    },
+    {
+      "epoch": 2.315841748054958,
+      "grad_norm": 2.1629180908203125,
+      "learning_rate": 1.5031515393581642e-06,
+      "loss": 0.3525,
+      "step": 13990
+    },
+    {
+      "epoch": 2.317497103128621,
+      "grad_norm": 2.212402820587158,
+      "learning_rate": 1.496274262440579e-06,
+      "loss": 0.3469,
+      "step": 14000
+    },
+    {
+      "epoch": 2.3191524582022844,
+      "grad_norm": 2.41933012008667,
+      "learning_rate": 1.4894099848776444e-06,
+      "loss": 0.3755,
+      "step": 14010
+    },
+    {
+      "epoch": 2.3208078132759478,
+      "grad_norm": 1.9362802505493164,
+      "learning_rate": 1.482558732136874e-06,
+      "loss": 0.3416,
+      "step": 14020
+    },
+    {
+      "epoch": 2.322463168349611,
+      "grad_norm": 1.8047678470611572,
+      "learning_rate": 1.4757205296374532e-06,
+      "loss": 0.3323,
+      "step": 14030
+    },
+    {
+      "epoch": 2.3241185234232744,
+      "grad_norm": 1.8480398654937744,
+      "learning_rate": 1.4688954027501545e-06,
+      "loss": 0.3112,
+      "step": 14040
+    },
+    {
+      "epoch": 2.3257738784969377,
+      "grad_norm": 2.211897611618042,
+      "learning_rate": 1.4620833767972365e-06,
+      "loss": 0.3252,
+      "step": 14050
+    },
+    {
+      "epoch": 2.327429233570601,
+      "grad_norm": 2.3595709800720215,
+      "learning_rate": 1.4552844770523477e-06,
+      "loss": 0.336,
+      "step": 14060
+    },
+    {
+      "epoch": 2.3290845886442644,
+      "grad_norm": 3.3694913387298584,
+      "learning_rate": 1.4484987287404407e-06,
+      "loss": 0.3253,
+      "step": 14070
+    },
+    {
+      "epoch": 2.3307399437179273,
+      "grad_norm": 1.8503248691558838,
+      "learning_rate": 1.4417261570376701e-06,
+      "loss": 0.3293,
+      "step": 14080
+    },
+    {
+      "epoch": 2.3323952987915906,
+      "grad_norm": 2.116312265396118,
+      "learning_rate": 1.4349667870713057e-06,
+      "loss": 0.3472,
+      "step": 14090
+    },
+    {
+      "epoch": 2.334050653865254,
+      "grad_norm": 2.352787971496582,
+      "learning_rate": 1.4282206439196395e-06,
+      "loss": 0.3609,
+      "step": 14100
+    },
+    {
+      "epoch": 2.335706008938917,
+      "grad_norm": 1.6080447435379028,
+      "learning_rate": 1.4214877526118853e-06,
+      "loss": 0.3355,
+      "step": 14110
+    },
+    {
+      "epoch": 2.3373613640125805,
+      "grad_norm": 2.383387565612793,
+      "learning_rate": 1.4147681381280909e-06,
+      "loss": 0.3439,
+      "step": 14120
+    },
+    {
+      "epoch": 2.339016719086244,
+      "grad_norm": 2.3709840774536133,
+      "learning_rate": 1.4080618253990502e-06,
+      "loss": 0.3425,
+      "step": 14130
+    },
+    {
+      "epoch": 2.340672074159907,
+      "grad_norm": 2.1909902095794678,
+      "learning_rate": 1.4013688393062003e-06,
+      "loss": 0.3428,
+      "step": 14140
+    },
+    {
+      "epoch": 2.3423274292335705,
+      "grad_norm": 1.8654948472976685,
+      "learning_rate": 1.3946892046815341e-06,
+      "loss": 0.3237,
+      "step": 14150
+    },
+    {
+      "epoch": 2.343982784307234,
+      "grad_norm": 1.9633307456970215,
+      "learning_rate": 1.3880229463075146e-06,
+      "loss": 0.3227,
+      "step": 14160
+    },
+    {
+      "epoch": 2.345638139380897,
+      "grad_norm": 2.1636171340942383,
+      "learning_rate": 1.3813700889169707e-06,
+      "loss": 0.3633,
+      "step": 14170
+    },
+    {
+      "epoch": 2.3472934944545605,
+      "grad_norm": 2.097670555114746,
+      "learning_rate": 1.3747306571930141e-06,
+      "loss": 0.363,
+      "step": 14180
+    },
+    {
+      "epoch": 2.348948849528224,
+      "grad_norm": 2.1034417152404785,
+      "learning_rate": 1.3681046757689448e-06,
+      "loss": 0.3479,
+      "step": 14190
+    },
+    {
+      "epoch": 2.350604204601887,
+      "grad_norm": 1.7305580377578735,
+      "learning_rate": 1.3614921692281586e-06,
+      "loss": 0.3351,
+      "step": 14200
+    },
+    {
+      "epoch": 2.3522595596755504,
+      "grad_norm": 2.346315860748291,
+      "learning_rate": 1.3548931621040607e-06,
+      "loss": 0.382,
+      "step": 14210
+    },
+    {
+      "epoch": 2.3539149147492138,
+      "grad_norm": 2.112192392349243,
+      "learning_rate": 1.3483076788799715e-06,
+      "loss": 0.3498,
+      "step": 14220
+    },
+    {
+      "epoch": 2.355570269822877,
+      "grad_norm": 2.4844164848327637,
+      "learning_rate": 1.3417357439890323e-06,
+      "loss": 0.359,
+      "step": 14230
+    },
+    {
+      "epoch": 2.3572256248965404,
+      "grad_norm": 2.170315742492676,
+      "learning_rate": 1.33517738181412e-06,
+      "loss": 0.3237,
+      "step": 14240
+    },
+    {
+      "epoch": 2.3588809799702037,
+      "grad_norm": 1.9840915203094482,
+      "learning_rate": 1.328632616687754e-06,
+      "loss": 0.3087,
+      "step": 14250
+    },
+    {
+      "epoch": 2.360536335043867,
+      "grad_norm": 2.098491907119751,
+      "learning_rate": 1.3221014728920056e-06,
+      "loss": 0.3542,
+      "step": 14260
+    },
+    {
+      "epoch": 2.3621916901175304,
+      "grad_norm": 2.413897752761841,
+      "learning_rate": 1.3155839746584138e-06,
+      "loss": 0.3359,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3638470451911937,
+      "grad_norm": 2.1317267417907715,
+      "learning_rate": 1.3090801461678848e-06,
+      "loss": 0.3045,
+      "step": 14280
+    },
+    {
+      "epoch": 2.365502400264857,
+      "grad_norm": 1.6186710596084595,
+      "learning_rate": 1.3025900115506086e-06,
+      "loss": 0.347,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3671577553385204,
+      "grad_norm": 2.192939519882202,
+      "learning_rate": 1.2961135948859737e-06,
+      "loss": 0.4005,
+      "step": 14300
+    },
+    {
+      "epoch": 2.3688131104121832,
+      "grad_norm": 2.0540459156036377,
+      "learning_rate": 1.2896509202024682e-06,
+      "loss": 0.3477,
+      "step": 14310
+    },
+    {
+      "epoch": 2.3704684654858466,
+      "grad_norm": 2.3481903076171875,
+      "learning_rate": 1.2832020114775951e-06,
+      "loss": 0.3254,
+      "step": 14320
+    },
+    {
+      "epoch": 2.37212382055951,
+      "grad_norm": 1.9363404512405396,
+      "learning_rate": 1.2767668926377885e-06,
+      "loss": 0.3361,
+      "step": 14330
+    },
+    {
+      "epoch": 2.373779175633173,
+      "grad_norm": 2.1417112350463867,
+      "learning_rate": 1.2703455875583148e-06,
+      "loss": 0.3502,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3754345307068365,
+      "grad_norm": 2.0085740089416504,
+      "learning_rate": 1.263938120063191e-06,
+      "loss": 0.3358,
+      "step": 14350
+    },
+    {
+      "epoch": 2.3770898857805,
+      "grad_norm": 1.862365961074829,
+      "learning_rate": 1.2575445139250936e-06,
+      "loss": 0.3062,
+      "step": 14360
+    },
+    {
+      "epoch": 2.378745240854163,
+      "grad_norm": 1.9572957754135132,
+      "learning_rate": 1.2511647928652754e-06,
+      "loss": 0.3038,
+      "step": 14370
+    },
+    {
+      "epoch": 2.3804005959278265,
+      "grad_norm": 2.7064197063446045,
+      "learning_rate": 1.2447989805534677e-06,
+      "loss": 0.3101,
+      "step": 14380
+    },
+    {
+      "epoch": 2.38205595100149,
+      "grad_norm": 2.0222458839416504,
+      "learning_rate": 1.2384471006078036e-06,
+      "loss": 0.3894,
+      "step": 14390
+    },
+    {
+      "epoch": 2.383711306075153,
+      "grad_norm": 1.8793877363204956,
+      "learning_rate": 1.2321091765947214e-06,
+      "loss": 0.2965,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3853666611488165,
+      "grad_norm": 1.7460297346115112,
+      "learning_rate": 1.2257852320288815e-06,
+      "loss": 0.3337,
+      "step": 14410
+    },
+    {
+      "epoch": 2.38702201622248,
+      "grad_norm": 2.319016218185425,
+      "learning_rate": 1.219475290373079e-06,
+      "loss": 0.3176,
+      "step": 14420
+    },
+    {
+      "epoch": 2.388677371296143,
+      "grad_norm": 2.06717848777771,
+      "learning_rate": 1.2131793750381554e-06,
+      "loss": 0.3919,
+      "step": 14430
+    },
+    {
+      "epoch": 2.3903327263698064,
+      "grad_norm": 2.179219961166382,
+      "learning_rate": 1.2068975093829123e-06,
+      "loss": 0.3412,
+      "step": 14440
+    },
+    {
+      "epoch": 2.3919880814434697,
+      "grad_norm": 1.629726767539978,
+      "learning_rate": 1.2006297167140257e-06,
+      "loss": 0.307,
+      "step": 14450
+    },
+    {
+      "epoch": 2.393643436517133,
+      "grad_norm": 2.2001848220825195,
+      "learning_rate": 1.1943760202859606e-06,
+      "loss": 0.3739,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3952987915907964,
+      "grad_norm": 1.8462634086608887,
+      "learning_rate": 1.188136443300879e-06,
+      "loss": 0.3466,
+      "step": 14470
+    },
+    {
+      "epoch": 2.3969541466644597,
+      "grad_norm": 1.7766656875610352,
+      "learning_rate": 1.1819110089085595e-06,
+      "loss": 0.3227,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3986095017381226,
+      "grad_norm": 2.0982725620269775,
+      "learning_rate": 1.1756997402063069e-06,
+      "loss": 0.3111,
+      "step": 14490
+    },
+    {
+      "epoch": 2.400264856811786,
+      "grad_norm": 2.340324878692627,
+      "learning_rate": 1.1695026602388755e-06,
+      "loss": 0.3458,
+      "step": 14500
+    },
+    {
+      "epoch": 2.4019202118854492,
+      "grad_norm": 2.2285056114196777,
+      "learning_rate": 1.1633197919983707e-06,
+      "loss": 0.3496,
+      "step": 14510
+    },
+    {
+      "epoch": 2.4035755669591126,
+      "grad_norm": 1.9656524658203125,
+      "learning_rate": 1.1571511584241735e-06,
+      "loss": 0.3217,
+      "step": 14520
+    },
+    {
+      "epoch": 2.405230922032776,
+      "grad_norm": 2.775507688522339,
+      "learning_rate": 1.1509967824028496e-06,
+      "loss": 0.3409,
+      "step": 14530
+    },
+    {
+      "epoch": 2.406886277106439,
+      "grad_norm": 2.0390307903289795,
+      "learning_rate": 1.1448566867680715e-06,
+      "loss": 0.3302,
+      "step": 14540
+    },
+    {
+      "epoch": 2.4085416321801025,
+      "grad_norm": 2.3175621032714844,
+      "learning_rate": 1.1387308943005248e-06,
+      "loss": 0.3494,
+      "step": 14550
+    },
+    {
+      "epoch": 2.410196987253766,
+      "grad_norm": 2.031985282897949,
+      "learning_rate": 1.132619427727832e-06,
+      "loss": 0.3195,
+      "step": 14560
+    },
+    {
+      "epoch": 2.411852342327429,
+      "grad_norm": 1.8072073459625244,
+      "learning_rate": 1.1265223097244604e-06,
+      "loss": 0.3405,
+      "step": 14570
+    },
+    {
+      "epoch": 2.4135076974010925,
+      "grad_norm": 2.3952126502990723,
+      "learning_rate": 1.1204395629116445e-06,
+      "loss": 0.3388,
+      "step": 14580
+    },
+    {
+      "epoch": 2.415163052474756,
+      "grad_norm": 2.0482804775238037,
+      "learning_rate": 1.114371209857299e-06,
+      "loss": 0.337,
+      "step": 14590
+    },
+    {
+      "epoch": 2.416818407548419,
+      "grad_norm": 2.1736388206481934,
+      "learning_rate": 1.1083172730759351e-06,
+      "loss": 0.3629,
+      "step": 14600
+    },
+    {
+      "epoch": 2.4184737626220825,
+      "grad_norm": 2.3148396015167236,
+      "learning_rate": 1.1022777750285767e-06,
+      "loss": 0.3523,
+      "step": 14610
+    },
+    {
+      "epoch": 2.420129117695746,
+      "grad_norm": 2.455919027328491,
+      "learning_rate": 1.0962527381226795e-06,
+      "loss": 0.3413,
+      "step": 14620
+    },
+    {
+      "epoch": 2.421784472769409,
+      "grad_norm": 2.373774766921997,
+      "learning_rate": 1.0902421847120482e-06,
+      "loss": 0.3068,
+      "step": 14630
+    },
+    {
+      "epoch": 2.4234398278430724,
+      "grad_norm": 2.160219669342041,
+      "learning_rate": 1.084246137096746e-06,
+      "loss": 0.345,
+      "step": 14640
+    },
+    {
+      "epoch": 2.4250951829167358,
+      "grad_norm": 1.6808048486709595,
+      "learning_rate": 1.0782646175230217e-06,
+      "loss": 0.3279,
+      "step": 14650
+    },
+    {
+      "epoch": 2.426750537990399,
+      "grad_norm": 1.472030520439148,
+      "learning_rate": 1.0722976481832205e-06,
+      "loss": 0.3158,
+      "step": 14660
+    },
+    {
+      "epoch": 2.4284058930640624,
+      "grad_norm": 1.8282170295715332,
+      "learning_rate": 1.0663452512157035e-06,
+      "loss": 0.323,
+      "step": 14670
+    },
+    {
+      "epoch": 2.4300612481377257,
+      "grad_norm": 1.9233111143112183,
+      "learning_rate": 1.0604074487047705e-06,
+      "loss": 0.3741,
+      "step": 14680
+    },
+    {
+      "epoch": 2.431716603211389,
+      "grad_norm": 1.9156076908111572,
+      "learning_rate": 1.0544842626805684e-06,
+      "loss": 0.3296,
+      "step": 14690
+    },
+    {
+      "epoch": 2.4333719582850524,
+      "grad_norm": 2.5253238677978516,
+      "learning_rate": 1.0485757151190195e-06,
+      "loss": 0.3149,
+      "step": 14700
+    },
+    {
+      "epoch": 2.4350273133587153,
+      "grad_norm": 1.5516034364700317,
+      "learning_rate": 1.0426818279417306e-06,
+      "loss": 0.3297,
+      "step": 14710
+    },
+    {
+      "epoch": 2.4366826684323786,
+      "grad_norm": 1.9635651111602783,
+      "learning_rate": 1.036802623015918e-06,
+      "loss": 0.3331,
+      "step": 14720
+    },
+    {
+      "epoch": 2.438338023506042,
+      "grad_norm": 1.9763412475585938,
+      "learning_rate": 1.0309381221543268e-06,
+      "loss": 0.3308,
+      "step": 14730
+    },
+    {
+      "epoch": 2.439993378579705,
+      "grad_norm": 1.942713975906372,
+      "learning_rate": 1.0250883471151451e-06,
+      "loss": 0.3429,
+      "step": 14740
+    },
+    {
+      "epoch": 2.4416487336533685,
+      "grad_norm": 2.4745335578918457,
+      "learning_rate": 1.0192533196019267e-06,
+      "loss": 0.3287,
+      "step": 14750
+    },
+    {
+      "epoch": 2.443304088727032,
+      "grad_norm": 2.276266098022461,
+      "learning_rate": 1.0134330612635101e-06,
+      "loss": 0.3187,
+      "step": 14760
+    },
+    {
+      "epoch": 2.444959443800695,
+      "grad_norm": 1.9105377197265625,
+      "learning_rate": 1.0076275936939357e-06,
+      "loss": 0.2986,
+      "step": 14770
+    },
+    {
+      "epoch": 2.4466147988743585,
+      "grad_norm": 2.1453704833984375,
+      "learning_rate": 1.0018369384323722e-06,
+      "loss": 0.3588,
+      "step": 14780
+    },
+    {
+      "epoch": 2.448270153948022,
+      "grad_norm": 2.0307068824768066,
+      "learning_rate": 9.960611169630308e-07,
+      "loss": 0.3284,
+      "step": 14790
+    },
+    {
+      "epoch": 2.449925509021685,
+      "grad_norm": 2.2573883533477783,
+      "learning_rate": 9.90300150715085e-07,
+      "loss": 0.3408,
+      "step": 14800
+    },
+    {
+      "epoch": 2.4515808640953485,
+      "grad_norm": 2.224506378173828,
+      "learning_rate": 9.845540610625953e-07,
+      "loss": 0.338,
+      "step": 14810
+    },
+    {
+      "epoch": 2.453236219169012,
+      "grad_norm": 2.1470425128936768,
+      "learning_rate": 9.788228693244266e-07,
+      "loss": 0.3116,
+      "step": 14820
+    },
+    {
+      "epoch": 2.454891574242675,
+      "grad_norm": 1.8685119152069092,
+      "learning_rate": 9.731065967641712e-07,
+      "loss": 0.3046,
+      "step": 14830
+    },
+    {
+      "epoch": 2.4565469293163384,
+      "grad_norm": 2.515958070755005,
+      "learning_rate": 9.67405264590066e-07,
+      "loss": 0.3251,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4582022843900018,
+      "grad_norm": 2.4196338653564453,
+      "learning_rate": 9.617188939549232e-07,
+      "loss": 0.318,
+      "step": 14850
+    },
+    {
+      "epoch": 2.459857639463665,
+      "grad_norm": 2.0046846866607666,
+      "learning_rate": 9.560475059560388e-07,
+      "loss": 0.325,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4615129945373284,
+      "grad_norm": 2.182124376296997,
+      "learning_rate": 9.503911216351252e-07,
+      "loss": 0.3061,
+      "step": 14870
+    },
+    {
+      "epoch": 2.4631683496109917,
+      "grad_norm": 2.041778087615967,
+      "learning_rate": 9.447497619782269e-07,
+      "loss": 0.31,
+      "step": 14880
+    },
+    {
+      "epoch": 2.4648237046846546,
+      "grad_norm": 2.510472297668457,
+      "learning_rate": 9.391234479156452e-07,
+      "loss": 0.3278,
+      "step": 14890
+    },
+    {
+      "epoch": 2.466479059758318,
+      "grad_norm": 1.6733882427215576,
+      "learning_rate": 9.335122003218583e-07,
+      "loss": 0.3061,
+      "step": 14900
+    },
+    {
+      "epoch": 2.4681344148319813,
+      "grad_norm": 1.9581342935562134,
+      "learning_rate": 9.279160400154497e-07,
+      "loss": 0.3268,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4697897699056446,
+      "grad_norm": 2.368661642074585,
+      "learning_rate": 9.223349877590226e-07,
+      "loss": 0.3642,
+      "step": 14920
+    },
+    {
+      "epoch": 2.471445124979308,
+      "grad_norm": 1.5843191146850586,
+      "learning_rate": 9.167690642591287e-07,
+      "loss": 0.3566,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4731004800529712,
+      "grad_norm": 1.8079990148544312,
+      "learning_rate": 9.112182901661881e-07,
+      "loss": 0.3455,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4747558351266346,
+      "grad_norm": 2.184272289276123,
+      "learning_rate": 9.056826860744178e-07,
+      "loss": 0.3407,
+      "step": 14950
+    },
+    {
+      "epoch": 2.476411190200298,
+      "grad_norm": 1.532046914100647,
+      "learning_rate": 9.001622725217495e-07,
+      "loss": 0.31,
+      "step": 14960
+    },
+    {
+      "epoch": 2.478066545273961,
+      "grad_norm": 2.5433170795440674,
+      "learning_rate": 8.946570699897566e-07,
+      "loss": 0.3267,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4797219003476245,
+      "grad_norm": 2.535883903503418,
+      "learning_rate": 8.89167098903575e-07,
+      "loss": 0.3047,
+      "step": 14980
+    },
+    {
+      "epoch": 2.481377255421288,
+      "grad_norm": 1.7732888460159302,
+      "learning_rate": 8.836923796318325e-07,
+      "loss": 0.3173,
+      "step": 14990
+    },
+    {
+      "epoch": 2.483032610494951,
+      "grad_norm": 2.467151403427124,
+      "learning_rate": 8.78232932486568e-07,
+      "loss": 0.3411,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4846879655686145,
+      "grad_norm": 2.1632003784179688,
+      "learning_rate": 8.727887777231591e-07,
+      "loss": 0.3038,
+      "step": 15010
+    },
+    {
+      "epoch": 2.486343320642278,
+      "grad_norm": 2.1071043014526367,
+      "learning_rate": 8.673599355402496e-07,
+      "loss": 0.3386,
+      "step": 15020
+    },
+    {
+      "epoch": 2.487998675715941,
+      "grad_norm": 1.7851759195327759,
+      "learning_rate": 8.619464260796651e-07,
+      "loss": 0.3284,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4896540307896045,
+      "grad_norm": 2.402301788330078,
+      "learning_rate": 8.565482694263516e-07,
+      "loss": 0.3356,
+      "step": 15040
+    },
+    {
+      "epoch": 2.4913093858632678,
+      "grad_norm": 1.905591368675232,
+      "learning_rate": 8.51165485608288e-07,
+      "loss": 0.303,
+      "step": 15050
+    },
+    {
+      "epoch": 2.492964740936931,
+      "grad_norm": 2.015292167663574,
+      "learning_rate": 8.4579809459642e-07,
+      "loss": 0.3312,
+      "step": 15060
+    },
+    {
+      "epoch": 2.4946200960105944,
+      "grad_norm": 2.36039662361145,
+      "learning_rate": 8.404461163045829e-07,
+      "loss": 0.3066,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4962754510842577,
+      "grad_norm": 2.305978536605835,
+      "learning_rate": 8.351095705894308e-07,
+      "loss": 0.3414,
+      "step": 15080
+    },
+    {
+      "epoch": 2.497930806157921,
+      "grad_norm": 2.039421558380127,
+      "learning_rate": 8.297884772503578e-07,
+      "loss": 0.3564,
+      "step": 15090
+    },
+    {
+      "epoch": 2.4995861612315844,
+      "grad_norm": 2.0474462509155273,
+      "learning_rate": 8.244828560294272e-07,
+      "loss": 0.3178,
+      "step": 15100
+    },
+    {
+      "epoch": 2.5012415163052477,
+      "grad_norm": 2.152980089187622,
+      "learning_rate": 8.19192726611302e-07,
+      "loss": 0.3511,
+      "step": 15110
+    },
+    {
+      "epoch": 2.502896871378911,
+      "grad_norm": 2.019902467727661,
+      "learning_rate": 8.139181086231651e-07,
+      "loss": 0.3401,
+      "step": 15120
+    },
+    {
+      "epoch": 2.504552226452574,
+      "grad_norm": 2.122307777404785,
+      "learning_rate": 8.086590216346479e-07,
+      "loss": 0.3577,
+      "step": 15130
+    },
+    {
+      "epoch": 2.5062075815262372,
+      "grad_norm": 2.2243058681488037,
+      "learning_rate": 8.034154851577658e-07,
+      "loss": 0.361,
+      "step": 15140
+    },
+    {
+      "epoch": 2.5078629365999006,
+      "grad_norm": 2.078754425048828,
+      "learning_rate": 7.981875186468335e-07,
+      "loss": 0.3345,
+      "step": 15150
+    },
+    {
+      "epoch": 2.509518291673564,
+      "grad_norm": 2.2314717769622803,
+      "learning_rate": 7.929751414984011e-07,
+      "loss": 0.3215,
+      "step": 15160
+    },
+    {
+      "epoch": 2.511173646747227,
+      "grad_norm": 2.4789726734161377,
+      "learning_rate": 7.87778373051179e-07,
+      "loss": 0.3323,
+      "step": 15170
+    },
+    {
+      "epoch": 2.5128290018208905,
+      "grad_norm": 1.7479615211486816,
+      "learning_rate": 7.825972325859671e-07,
+      "loss": 0.3372,
+      "step": 15180
+    },
+    {
+      "epoch": 2.514484356894554,
+      "grad_norm": 2.23327374458313,
+      "learning_rate": 7.774317393255837e-07,
+      "loss": 0.3448,
+      "step": 15190
+    },
+    {
+      "epoch": 2.516139711968217,
+      "grad_norm": 1.7538516521453857,
+      "learning_rate": 7.722819124347952e-07,
+      "loss": 0.3118,
+      "step": 15200
+    },
+    {
+      "epoch": 2.5177950670418805,
+      "grad_norm": 2.249854803085327,
+      "learning_rate": 7.671477710202407e-07,
+      "loss": 0.3393,
+      "step": 15210
+    },
+    {
+      "epoch": 2.519450422115544,
+      "grad_norm": 1.923726201057434,
+      "learning_rate": 7.620293341303636e-07,
+      "loss": 0.3565,
+      "step": 15220
+    },
+    {
+      "epoch": 2.521105777189207,
+      "grad_norm": 2.8457915782928467,
+      "learning_rate": 7.569266207553427e-07,
+      "loss": 0.3383,
+      "step": 15230
+    },
+    {
+      "epoch": 2.5227611322628705,
+      "grad_norm": 2.412850856781006,
+      "learning_rate": 7.518396498270191e-07,
+      "loss": 0.3066,
+      "step": 15240
+    },
+    {
+      "epoch": 2.524416487336534,
+      "grad_norm": 2.3531320095062256,
+      "learning_rate": 7.467684402188291e-07,
+      "loss": 0.3237,
+      "step": 15250
+    },
+    {
+      "epoch": 2.526071842410197,
+      "grad_norm": 1.6720978021621704,
+      "learning_rate": 7.417130107457293e-07,
+      "loss": 0.3458,
+      "step": 15260
+    },
+    {
+      "epoch": 2.5277271974838604,
+      "grad_norm": 2.2779157161712646,
+      "learning_rate": 7.366733801641302e-07,
+      "loss": 0.3091,
+      "step": 15270
+    },
+    {
+      "epoch": 2.5293825525575233,
+      "grad_norm": 3.225557804107666,
+      "learning_rate": 7.316495671718293e-07,
+      "loss": 0.3412,
+      "step": 15280
+    },
+    {
+      "epoch": 2.5310379076311866,
+      "grad_norm": 2.389375925064087,
+      "learning_rate": 7.266415904079338e-07,
+      "loss": 0.3549,
+      "step": 15290
+    },
+    {
+      "epoch": 2.53269326270485,
+      "grad_norm": 2.0392401218414307,
+      "learning_rate": 7.216494684527975e-07,
+      "loss": 0.3286,
+      "step": 15300
+    },
+    {
+      "epoch": 2.5343486177785133,
+      "grad_norm": 1.929891586303711,
+      "learning_rate": 7.166732198279535e-07,
+      "loss": 0.3015,
+      "step": 15310
+    },
+    {
+      "epoch": 2.5360039728521766,
+      "grad_norm": 1.5978699922561646,
+      "learning_rate": 7.11712862996039e-07,
+      "loss": 0.3161,
+      "step": 15320
+    },
+    {
+      "epoch": 2.53765932792584,
+      "grad_norm": 2.1672823429107666,
+      "learning_rate": 7.067684163607308e-07,
+      "loss": 0.3129,
+      "step": 15330
+    },
+    {
+      "epoch": 2.5393146829995032,
+      "grad_norm": 1.8573296070098877,
+      "learning_rate": 7.018398982666757e-07,
+      "loss": 0.3553,
+      "step": 15340
+    },
+    {
+      "epoch": 2.5409700380731666,
+      "grad_norm": 1.951036810874939,
+      "learning_rate": 6.96927326999427e-07,
+      "loss": 0.3444,
+      "step": 15350
+    },
+    {
+      "epoch": 2.54262539314683,
+      "grad_norm": 2.200702428817749,
+      "learning_rate": 6.920307207853683e-07,
+      "loss": 0.3321,
+      "step": 15360
+    },
+    {
+      "epoch": 2.544280748220493,
+      "grad_norm": 1.9019625186920166,
+      "learning_rate": 6.87150097791654e-07,
+      "loss": 0.3775,
+      "step": 15370
+    },
+    {
+      "epoch": 2.5459361032941565,
+      "grad_norm": 1.8351606130599976,
+      "learning_rate": 6.822854761261355e-07,
+      "loss": 0.3357,
+      "step": 15380
+    },
+    {
+      "epoch": 2.54759145836782,
+      "grad_norm": 2.076394557952881,
+      "learning_rate": 6.774368738372988e-07,
+      "loss": 0.3164,
+      "step": 15390
+    },
+    {
+      "epoch": 2.549246813441483,
+      "grad_norm": 2.102830171585083,
+      "learning_rate": 6.726043089141943e-07,
+      "loss": 0.3377,
+      "step": 15400
+    },
+    {
+      "epoch": 2.5509021685151465,
+      "grad_norm": 2.1659622192382812,
+      "learning_rate": 6.677877992863707e-07,
+      "loss": 0.3157,
+      "step": 15410
+    },
+    {
+      "epoch": 2.55255752358881,
+      "grad_norm": 2.069488525390625,
+      "learning_rate": 6.629873628238126e-07,
+      "loss": 0.3388,
+      "step": 15420
+    },
+    {
+      "epoch": 2.554212878662473,
+      "grad_norm": 2.4191696643829346,
+      "learning_rate": 6.582030173368664e-07,
+      "loss": 0.3583,
+      "step": 15430
+    },
+    {
+      "epoch": 2.5558682337361365,
+      "grad_norm": 2.1078972816467285,
+      "learning_rate": 6.534347805761826e-07,
+      "loss": 0.3732,
+      "step": 15440
+    },
+    {
+      "epoch": 2.5575235888098,
+      "grad_norm": 1.9522989988327026,
+      "learning_rate": 6.486826702326426e-07,
+      "loss": 0.289,
+      "step": 15450
+    },
+    {
+      "epoch": 2.559178943883463,
+      "grad_norm": 1.9794495105743408,
+      "learning_rate": 6.439467039372971e-07,
+      "loss": 0.3604,
+      "step": 15460
+    },
+    {
+      "epoch": 2.5608342989571264,
+      "grad_norm": 2.4447927474975586,
+      "learning_rate": 6.392268992613005e-07,
+      "loss": 0.3306,
+      "step": 15470
+    },
+    {
+      "epoch": 2.5624896540307898,
+      "grad_norm": 2.0704166889190674,
+      "learning_rate": 6.345232737158458e-07,
+      "loss": 0.3003,
+      "step": 15480
+    },
+    {
+      "epoch": 2.564145009104453,
+      "grad_norm": 2.33072566986084,
+      "learning_rate": 6.298358447520985e-07,
+      "loss": 0.3249,
+      "step": 15490
+    },
+    {
+      "epoch": 2.5658003641781164,
+      "grad_norm": 2.0629100799560547,
+      "learning_rate": 6.251646297611308e-07,
+      "loss": 0.3381,
+      "step": 15500
+    },
+    {
+      "epoch": 2.5674557192517797,
+      "grad_norm": 3.117828607559204,
+      "learning_rate": 6.205096460738591e-07,
+      "loss": 0.3267,
+      "step": 15510
+    },
+    {
+      "epoch": 2.569111074325443,
+      "grad_norm": 1.9236739873886108,
+      "learning_rate": 6.158709109609828e-07,
+      "loss": 0.3281,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5707664293991064,
+      "grad_norm": 2.1610283851623535,
+      "learning_rate": 6.112484416329107e-07,
+      "loss": 0.3101,
+      "step": 15530
+    },
+    {
+      "epoch": 2.5724217844727693,
+      "grad_norm": 1.9747469425201416,
+      "learning_rate": 6.066422552397083e-07,
+      "loss": 0.3243,
+      "step": 15540
+    },
+    {
+      "epoch": 2.5740771395464326,
+      "grad_norm": 2.253056049346924,
+      "learning_rate": 6.020523688710256e-07,
+      "loss": 0.3359,
+      "step": 15550
+    },
+    {
+      "epoch": 2.575732494620096,
+      "grad_norm": 2.041837215423584,
+      "learning_rate": 5.974787995560389e-07,
+      "loss": 0.302,
+      "step": 15560
+    },
+    {
+      "epoch": 2.5773878496937592,
+      "grad_norm": 1.8018887042999268,
+      "learning_rate": 5.92921564263384e-07,
+      "loss": 0.36,
+      "step": 15570
+    },
+    {
+      "epoch": 2.5790432047674225,
+      "grad_norm": 2.423647880554199,
+      "learning_rate": 5.88380679901096e-07,
+      "loss": 0.3439,
+      "step": 15580
+    },
+    {
+      "epoch": 2.580698559841086,
+      "grad_norm": 2.334973096847534,
+      "learning_rate": 5.838561633165474e-07,
+      "loss": 0.2951,
+      "step": 15590
+    },
+    {
+      "epoch": 2.582353914914749,
+      "grad_norm": 1.6279757022857666,
+      "learning_rate": 5.793480312963789e-07,
+      "loss": 0.3056,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5840092699884125,
+      "grad_norm": 1.6940557956695557,
+      "learning_rate": 5.748563005664482e-07,
+      "loss": 0.3447,
+      "step": 15610
+    },
+    {
+      "epoch": 2.585664625062076,
+      "grad_norm": 2.257672071456909,
+      "learning_rate": 5.703809877917571e-07,
+      "loss": 0.3755,
+      "step": 15620
+    },
+    {
+      "epoch": 2.587319980135739,
+      "grad_norm": 1.9907878637313843,
+      "learning_rate": 5.659221095763955e-07,
+      "loss": 0.3425,
+      "step": 15630
+    },
+    {
+      "epoch": 2.5889753352094025,
+      "grad_norm": 1.69442617893219,
+      "learning_rate": 5.614796824634783e-07,
+      "loss": 0.332,
+      "step": 15640
+    },
+    {
+      "epoch": 2.590630690283066,
+      "grad_norm": 1.8623377084732056,
+      "learning_rate": 5.570537229350864e-07,
+      "loss": 0.3538,
+      "step": 15650
+    },
+    {
+      "epoch": 2.592286045356729,
+      "grad_norm": 2.767336845397949,
+      "learning_rate": 5.526442474122013e-07,
+      "loss": 0.3634,
+      "step": 15660
+    },
+    {
+      "epoch": 2.5939414004303925,
+      "grad_norm": 2.1302411556243896,
+      "learning_rate": 5.482512722546468e-07,
+      "loss": 0.3413,
+      "step": 15670
+    },
+    {
+      "epoch": 2.5955967555040558,
+      "grad_norm": 1.9488990306854248,
+      "learning_rate": 5.438748137610267e-07,
+      "loss": 0.3044,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5972521105777187,
+      "grad_norm": 1.6431246995925903,
+      "learning_rate": 5.395148881686685e-07,
+      "loss": 0.348,
+      "step": 15690
+    },
+    {
+      "epoch": 2.598907465651382,
+      "grad_norm": 2.266380548477173,
+      "learning_rate": 5.351715116535571e-07,
+      "loss": 0.3123,
+      "step": 15700
+    },
+    {
+      "epoch": 2.6005628207250453,
+      "grad_norm": 2.4466562271118164,
+      "learning_rate": 5.308447003302808e-07,
+      "loss": 0.3468,
+      "step": 15710
+    },
+    {
+      "epoch": 2.6022181757987086,
+      "grad_norm": 2.20597505569458,
+      "learning_rate": 5.265344702519654e-07,
+      "loss": 0.3658,
+      "step": 15720
+    },
+    {
+      "epoch": 2.603873530872372,
+      "grad_norm": 1.868127703666687,
+      "learning_rate": 5.2224083741022e-07,
+      "loss": 0.3256,
+      "step": 15730
+    },
+    {
+      "epoch": 2.6055288859460353,
+      "grad_norm": 2.128237724304199,
+      "learning_rate": 5.179638177350737e-07,
+      "loss": 0.3305,
+      "step": 15740
+    },
+    {
+      "epoch": 2.6071842410196986,
+      "grad_norm": 2.1606836318969727,
+      "learning_rate": 5.137034270949182e-07,
+      "loss": 0.3211,
+      "step": 15750
+    },
+    {
+      "epoch": 2.608839596093362,
+      "grad_norm": 2.227559804916382,
+      "learning_rate": 5.094596812964525e-07,
+      "loss": 0.3505,
+      "step": 15760
+    },
+    {
+      "epoch": 2.6104949511670252,
+      "grad_norm": 1.9581117630004883,
+      "learning_rate": 5.052325960846155e-07,
+      "loss": 0.3398,
+      "step": 15770
+    },
+    {
+      "epoch": 2.6121503062406886,
+      "grad_norm": 2.039578914642334,
+      "learning_rate": 5.010221871425375e-07,
+      "loss": 0.3093,
+      "step": 15780
+    },
+    {
+      "epoch": 2.613805661314352,
+      "grad_norm": 1.7403851747512817,
+      "learning_rate": 4.968284700914744e-07,
+      "loss": 0.3196,
+      "step": 15790
+    },
+    {
+      "epoch": 2.615461016388015,
+      "grad_norm": 1.892404317855835,
+      "learning_rate": 4.926514604907534e-07,
+      "loss": 0.3266,
+      "step": 15800
+    },
+    {
+      "epoch": 2.6171163714616785,
+      "grad_norm": 2.2020905017852783,
+      "learning_rate": 4.88491173837713e-07,
+      "loss": 0.3502,
+      "step": 15810
+    },
+    {
+      "epoch": 2.618771726535342,
+      "grad_norm": 1.3924425840377808,
+      "learning_rate": 4.843476255676516e-07,
+      "loss": 0.3446,
+      "step": 15820
+    },
+    {
+      "epoch": 2.620427081609005,
+      "grad_norm": 2.562934637069702,
+      "learning_rate": 4.802208310537604e-07,
+      "loss": 0.3755,
+      "step": 15830
+    },
+    {
+      "epoch": 2.6220824366826685,
+      "grad_norm": 1.5193594694137573,
+      "learning_rate": 4.7611080560707344e-07,
+      "loss": 0.3136,
+      "step": 15840
+    },
+    {
+      "epoch": 2.623737791756332,
+      "grad_norm": 2.2128007411956787,
+      "learning_rate": 4.720175644764119e-07,
+      "loss": 0.3413,
+      "step": 15850
+    },
+    {
+      "epoch": 2.625393146829995,
+      "grad_norm": 1.999158263206482,
+      "learning_rate": 4.6794112284831995e-07,
+      "loss": 0.3565,
+      "step": 15860
+    },
+    {
+      "epoch": 2.6270485019036585,
+      "grad_norm": 2.3605151176452637,
+      "learning_rate": 4.63881495847015e-07,
+      "loss": 0.317,
+      "step": 15870
+    },
+    {
+      "epoch": 2.628703856977322,
+      "grad_norm": 2.1883292198181152,
+      "learning_rate": 4.5983869853433174e-07,
+      "loss": 0.317,
+      "step": 15880
+    },
+    {
+      "epoch": 2.630359212050985,
+      "grad_norm": 2.075212240219116,
+      "learning_rate": 4.558127459096612e-07,
+      "loss": 0.308,
+      "step": 15890
+    },
+    {
+      "epoch": 2.6320145671246484,
+      "grad_norm": 1.9252678155899048,
+      "learning_rate": 4.51803652909899e-07,
+      "loss": 0.3231,
+      "step": 15900
+    },
+    {
+      "epoch": 2.6336699221983118,
+      "grad_norm": 2.443986415863037,
+      "learning_rate": 4.4781143440938923e-07,
+      "loss": 0.3437,
+      "step": 15910
+    },
+    {
+      "epoch": 2.635325277271975,
+      "grad_norm": 1.7434769868850708,
+      "learning_rate": 4.438361052198675e-07,
+      "loss": 0.3152,
+      "step": 15920
+    },
+    {
+      "epoch": 2.6369806323456384,
+      "grad_norm": 2.128065347671509,
+      "learning_rate": 4.3987768009041033e-07,
+      "loss": 0.3374,
+      "step": 15930
+    },
+    {
+      "epoch": 2.6386359874193013,
+      "grad_norm": 1.744676113128662,
+      "learning_rate": 4.3593617370737697e-07,
+      "loss": 0.31,
+      "step": 15940
+    },
+    {
+      "epoch": 2.6402913424929646,
+      "grad_norm": 1.8740226030349731,
+      "learning_rate": 4.3201160069435367e-07,
+      "loss": 0.3152,
+      "step": 15950
+    },
+    {
+      "epoch": 2.641946697566628,
+      "grad_norm": 3.100207567214966,
+      "learning_rate": 4.281039756121025e-07,
+      "loss": 0.3397,
+      "step": 15960
+    },
+    {
+      "epoch": 2.6436020526402912,
+      "grad_norm": 1.7563873529434204,
+      "learning_rate": 4.242133129585063e-07,
+      "loss": 0.3383,
+      "step": 15970
+    },
+    {
+      "epoch": 2.6452574077139546,
+      "grad_norm": 2.2582356929779053,
+      "learning_rate": 4.2033962716851396e-07,
+      "loss": 0.3247,
+      "step": 15980
+    },
+    {
+      "epoch": 2.646912762787618,
+      "grad_norm": 2.0863699913024902,
+      "learning_rate": 4.164829326140873e-07,
+      "loss": 0.335,
+      "step": 15990
+    },
+    {
+      "epoch": 2.648568117861281,
+      "grad_norm": 2.127856969833374,
+      "learning_rate": 4.126432436041511e-07,
+      "loss": 0.3716,
+      "step": 16000
+    },
+    {
+      "epoch": 2.6502234729349445,
+      "grad_norm": 1.7168405055999756,
+      "learning_rate": 4.0882057438453305e-07,
+      "loss": 0.3074,
+      "step": 16010
+    },
+    {
+      "epoch": 2.651878828008608,
+      "grad_norm": 2.110229969024658,
+      "learning_rate": 4.050149391379177e-07,
+      "loss": 0.3661,
+      "step": 16020
+    },
+    {
+      "epoch": 2.653534183082271,
+      "grad_norm": 1.760513186454773,
+      "learning_rate": 4.0122635198378943e-07,
+      "loss": 0.2615,
+      "step": 16030
+    },
+    {
+      "epoch": 2.6551895381559345,
+      "grad_norm": 1.8273522853851318,
+      "learning_rate": 3.974548269783807e-07,
+      "loss": 0.3055,
+      "step": 16040
+    },
+    {
+      "epoch": 2.656844893229598,
+      "grad_norm": 2.403547763824463,
+      "learning_rate": 3.9370037811462424e-07,
+      "loss": 0.341,
+      "step": 16050
+    },
+    {
+      "epoch": 2.658500248303261,
+      "grad_norm": 2.5805859565734863,
+      "learning_rate": 3.899630193220949e-07,
+      "loss": 0.3347,
+      "step": 16060
+    },
+    {
+      "epoch": 2.6601556033769245,
+      "grad_norm": 2.449556827545166,
+      "learning_rate": 3.8624276446696086e-07,
+      "loss": 0.3419,
+      "step": 16070
+    },
+    {
+      "epoch": 2.661810958450588,
+      "grad_norm": 1.9614695310592651,
+      "learning_rate": 3.825396273519322e-07,
+      "loss": 0.3109,
+      "step": 16080
+    },
+    {
+      "epoch": 2.663466313524251,
+      "grad_norm": 1.9693225622177124,
+      "learning_rate": 3.78853621716212e-07,
+      "loss": 0.3744,
+      "step": 16090
+    },
+    {
+      "epoch": 2.665121668597914,
+      "grad_norm": 2.334308385848999,
+      "learning_rate": 3.751847612354387e-07,
+      "loss": 0.3278,
+      "step": 16100
+    },
+    {
+      "epoch": 2.6667770236715773,
+      "grad_norm": 1.722469687461853,
+      "learning_rate": 3.715330595216443e-07,
+      "loss": 0.3293,
+      "step": 16110
+    },
+    {
+      "epoch": 2.6684323787452406,
+      "grad_norm": 2.3163626194000244,
+      "learning_rate": 3.678985301231952e-07,
+      "loss": 0.3562,
+      "step": 16120
+    },
+    {
+      "epoch": 2.670087733818904,
+      "grad_norm": 1.810492992401123,
+      "learning_rate": 3.6428118652474807e-07,
+      "loss": 0.3117,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6717430888925673,
+      "grad_norm": 1.643846869468689,
+      "learning_rate": 3.606810421471973e-07,
+      "loss": 0.3057,
+      "step": 16140
+    },
+    {
+      "epoch": 2.6733984439662306,
+      "grad_norm": 2.1098129749298096,
+      "learning_rate": 3.5709811034762456e-07,
+      "loss": 0.3583,
+      "step": 16150
+    },
+    {
+      "epoch": 2.675053799039894,
+      "grad_norm": 1.8769261837005615,
+      "learning_rate": 3.535324044192506e-07,
+      "loss": 0.3228,
+      "step": 16160
+    },
+    {
+      "epoch": 2.6767091541135573,
+      "grad_norm": 2.2012782096862793,
+      "learning_rate": 3.499839375913872e-07,
+      "loss": 0.3166,
+      "step": 16170
+    },
+    {
+      "epoch": 2.6783645091872206,
+      "grad_norm": 2.656381607055664,
+      "learning_rate": 3.464527230293852e-07,
+      "loss": 0.3546,
+      "step": 16180
+    },
+    {
+      "epoch": 2.680019864260884,
+      "grad_norm": 2.1426846981048584,
+      "learning_rate": 3.429387738345868e-07,
+      "loss": 0.3262,
+      "step": 16190
+    },
+    {
+      "epoch": 2.6816752193345472,
+      "grad_norm": 2.4014816284179688,
+      "learning_rate": 3.3944210304427736e-07,
+      "loss": 0.3479,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6833305744082105,
+      "grad_norm": 2.6231253147125244,
+      "learning_rate": 3.3596272363163594e-07,
+      "loss": 0.338,
+      "step": 16210
+    },
+    {
+      "epoch": 2.684985929481874,
+      "grad_norm": 2.0805563926696777,
+      "learning_rate": 3.325006485056881e-07,
+      "loss": 0.318,
+      "step": 16220
+    },
+    {
+      "epoch": 2.686641284555537,
+      "grad_norm": 1.7813467979431152,
+      "learning_rate": 3.2905589051126065e-07,
+      "loss": 0.336,
+      "step": 16230
+    },
+    {
+      "epoch": 2.6882966396292005,
+      "grad_norm": 2.336962938308716,
+      "learning_rate": 3.2562846242892744e-07,
+      "loss": 0.3539,
+      "step": 16240
+    },
+    {
+      "epoch": 2.689951994702864,
+      "grad_norm": 2.1561317443847656,
+      "learning_rate": 3.2221837697496597e-07,
+      "loss": 0.329,
+      "step": 16250
+    },
+    {
+      "epoch": 2.691607349776527,
+      "grad_norm": 2.0386962890625,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.3483,
+      "step": 16260
+    },
+    {
+      "epoch": 2.6932627048501905,
+      "grad_norm": 2.5452029705047607,
+      "learning_rate": 3.1545028449551375e-07,
+      "loss": 0.3314,
+      "step": 16270
+    },
+    {
+      "epoch": 2.694918059923854,
+      "grad_norm": 1.8535023927688599,
+      "learning_rate": 3.1209230258067324e-07,
+      "loss": 0.3055,
+      "step": 16280
+    },
+    {
+      "epoch": 2.696573414997517,
+      "grad_norm": 2.0997235774993896,
+      "learning_rate": 3.0875171351541497e-07,
+      "loss": 0.3775,
+      "step": 16290
+    },
+    {
+      "epoch": 2.6982287700711804,
+      "grad_norm": 2.2901790142059326,
+      "learning_rate": 3.0542852969383196e-07,
+      "loss": 0.3575,
+      "step": 16300
+    },
+    {
+      "epoch": 2.6998841251448438,
+      "grad_norm": 1.9664065837860107,
+      "learning_rate": 3.021227634454399e-07,
+      "loss": 0.2803,
+      "step": 16310
+    },
+    {
+      "epoch": 2.701539480218507,
+      "grad_norm": 2.0761005878448486,
+      "learning_rate": 2.988344270351351e-07,
+      "loss": 0.3469,
+      "step": 16320
+    },
+    {
+      "epoch": 2.7031948352921704,
+      "grad_norm": 1.9783636331558228,
+      "learning_rate": 2.955635326631434e-07,
+      "loss": 0.3665,
+      "step": 16330
+    },
+    {
+      "epoch": 2.7048501903658337,
+      "grad_norm": 2.280743360519409,
+      "learning_rate": 2.9231009246498077e-07,
+      "loss": 0.3547,
+      "step": 16340
+    },
+    {
+      "epoch": 2.7065055454394966,
+      "grad_norm": 3.1975841522216797,
+      "learning_rate": 2.8907411851140487e-07,
+      "loss": 0.3531,
+      "step": 16350
+    },
+    {
+      "epoch": 2.70816090051316,
+      "grad_norm": 2.105062961578369,
+      "learning_rate": 2.8585562280837033e-07,
+      "loss": 0.3093,
+      "step": 16360
+    },
+    {
+      "epoch": 2.7098162555868233,
+      "grad_norm": 2.0571696758270264,
+      "learning_rate": 2.826546172969852e-07,
+      "loss": 0.3689,
+      "step": 16370
+    },
+    {
+      "epoch": 2.7114716106604866,
+      "grad_norm": 2.0750467777252197,
+      "learning_rate": 2.794711138534656e-07,
+      "loss": 0.3626,
+      "step": 16380
+    },
+    {
+      "epoch": 2.71312696573415,
+      "grad_norm": 2.3040149211883545,
+      "learning_rate": 2.7630512428909183e-07,
+      "loss": 0.3472,
+      "step": 16390
+    },
+    {
+      "epoch": 2.7147823208078132,
+      "grad_norm": 2.162661552429199,
+      "learning_rate": 2.731566603501684e-07,
+      "loss": 0.305,
+      "step": 16400
+    },
+    {
+      "epoch": 2.7164376758814766,
+      "grad_norm": 2.7106826305389404,
+      "learning_rate": 2.7002573371797347e-07,
+      "loss": 0.3489,
+      "step": 16410
+    },
+    {
+      "epoch": 2.71809303095514,
+      "grad_norm": 2.0429866313934326,
+      "learning_rate": 2.6691235600872e-07,
+      "loss": 0.3004,
+      "step": 16420
+    },
+    {
+      "epoch": 2.719748386028803,
+      "grad_norm": 2.0341339111328125,
+      "learning_rate": 2.638165387735131e-07,
+      "loss": 0.3486,
+      "step": 16430
+    },
+    {
+      "epoch": 2.7214037411024665,
+      "grad_norm": 2.3097574710845947,
+      "learning_rate": 2.607382934983044e-07,
+      "loss": 0.3774,
+      "step": 16440
+    },
+    {
+      "epoch": 2.72305909617613,
+      "grad_norm": 2.3151650428771973,
+      "learning_rate": 2.5767763160385095e-07,
+      "loss": 0.3031,
+      "step": 16450
+    },
+    {
+      "epoch": 2.724714451249793,
+      "grad_norm": 2.07060170173645,
+      "learning_rate": 2.5463456444567436e-07,
+      "loss": 0.3331,
+      "step": 16460
+    },
+    {
+      "epoch": 2.7263698063234565,
+      "grad_norm": 1.5830529928207397,
+      "learning_rate": 2.51609103314015e-07,
+      "loss": 0.3127,
+      "step": 16470
+    },
+    {
+      "epoch": 2.72802516139712,
+      "grad_norm": 2.311042547225952,
+      "learning_rate": 2.486012594337939e-07,
+      "loss": 0.3813,
+      "step": 16480
+    },
+    {
+      "epoch": 2.729680516470783,
+      "grad_norm": 2.337822437286377,
+      "learning_rate": 2.4561104396456815e-07,
+      "loss": 0.3171,
+      "step": 16490
+    },
+    {
+      "epoch": 2.731335871544446,
+      "grad_norm": 2.3123905658721924,
+      "learning_rate": 2.426384680004917e-07,
+      "loss": 0.3243,
+      "step": 16500
+    },
+    {
+      "epoch": 2.7329912266181093,
+      "grad_norm": 2.2762632369995117,
+      "learning_rate": 2.3968354257027205e-07,
+      "loss": 0.3188,
+      "step": 16510
+    },
+    {
+      "epoch": 2.7346465816917727,
+      "grad_norm": 2.021897315979004,
+      "learning_rate": 2.3674627863713273e-07,
+      "loss": 0.3599,
+      "step": 16520
+    },
+    {
+      "epoch": 2.736301936765436,
+      "grad_norm": 1.663979172706604,
+      "learning_rate": 2.3382668709876878e-07,
+      "loss": 0.303,
+      "step": 16530
+    },
+    {
+      "epoch": 2.7379572918390993,
+      "grad_norm": 2.365117311477661,
+      "learning_rate": 2.3092477878730757e-07,
+      "loss": 0.3329,
+      "step": 16540
+    },
+    {
+      "epoch": 2.7396126469127626,
+      "grad_norm": 1.6717923879623413,
+      "learning_rate": 2.280405644692696e-07,
+      "loss": 0.3389,
+      "step": 16550
+    },
+    {
+      "epoch": 2.741268001986426,
+      "grad_norm": 1.7539387941360474,
+      "learning_rate": 2.2517405484552778e-07,
+      "loss": 0.3362,
+      "step": 16560
+    },
+    {
+      "epoch": 2.7429233570600893,
+      "grad_norm": 2.190721273422241,
+      "learning_rate": 2.2232526055126992e-07,
+      "loss": 0.3732,
+      "step": 16570
+    },
+    {
+      "epoch": 2.7445787121337526,
+      "grad_norm": 1.689123272895813,
+      "learning_rate": 2.1949419215595346e-07,
+      "loss": 0.3564,
+      "step": 16580
+    },
+    {
+      "epoch": 2.746234067207416,
+      "grad_norm": 2.2224695682525635,
+      "learning_rate": 2.1668086016327415e-07,
+      "loss": 0.3176,
+      "step": 16590
+    },
+    {
+      "epoch": 2.7478894222810792,
+      "grad_norm": 2.078242063522339,
+      "learning_rate": 2.1388527501111977e-07,
+      "loss": 0.3233,
+      "step": 16600
+    },
+    {
+      "epoch": 2.7495447773547426,
+      "grad_norm": 1.993311882019043,
+      "learning_rate": 2.1110744707153574e-07,
+      "loss": 0.3093,
+      "step": 16610
+    },
+    {
+      "epoch": 2.751200132428406,
+      "grad_norm": 2.2646372318267822,
+      "learning_rate": 2.0834738665068576e-07,
+      "loss": 0.3471,
+      "step": 16620
+    },
+    {
+      "epoch": 2.752855487502069,
+      "grad_norm": 1.9162275791168213,
+      "learning_rate": 2.056051039888135e-07,
+      "loss": 0.2779,
+      "step": 16630
+    },
+    {
+      "epoch": 2.7545108425757325,
+      "grad_norm": 2.898524045944214,
+      "learning_rate": 2.0288060926020425e-07,
+      "loss": 0.3635,
+      "step": 16640
+    },
+    {
+      "epoch": 2.756166197649396,
+      "grad_norm": 2.449143648147583,
+      "learning_rate": 2.0017391257314723e-07,
+      "loss": 0.3147,
+      "step": 16650
+    },
+    {
+      "epoch": 2.757821552723059,
+      "grad_norm": 2.447744131088257,
+      "learning_rate": 1.9748502396989722e-07,
+      "loss": 0.3296,
+      "step": 16660
+    },
+    {
+      "epoch": 2.7594769077967225,
+      "grad_norm": 1.8371167182922363,
+      "learning_rate": 1.9481395342664078e-07,
+      "loss": 0.3355,
+      "step": 16670
+    },
+    {
+      "epoch": 2.761132262870386,
+      "grad_norm": 1.8588885068893433,
+      "learning_rate": 1.92160710853454e-07,
+      "loss": 0.3565,
+      "step": 16680
+    },
+    {
+      "epoch": 2.762787617944049,
+      "grad_norm": 2.5657360553741455,
+      "learning_rate": 1.8952530609427145e-07,
+      "loss": 0.2915,
+      "step": 16690
+    },
+    {
+      "epoch": 2.7644429730177125,
+      "grad_norm": 2.608978033065796,
+      "learning_rate": 1.8690774892684395e-07,
+      "loss": 0.3885,
+      "step": 16700
+    },
+    {
+      "epoch": 2.766098328091376,
+      "grad_norm": 2.3337411880493164,
+      "learning_rate": 1.8430804906270638e-07,
+      "loss": 0.2748,
+      "step": 16710
+    },
+    {
+      "epoch": 2.767753683165039,
+      "grad_norm": 2.865518808364868,
+      "learning_rate": 1.8172621614714004e-07,
+      "loss": 0.369,
+      "step": 16720
+    },
+    {
+      "epoch": 2.7694090382387024,
+      "grad_norm": 2.097165822982788,
+      "learning_rate": 1.7916225975913693e-07,
+      "loss": 0.3264,
+      "step": 16730
+    },
+    {
+      "epoch": 2.7710643933123658,
+      "grad_norm": 2.0352957248687744,
+      "learning_rate": 1.766161894113655e-07,
+      "loss": 0.2894,
+      "step": 16740
+    },
+    {
+      "epoch": 2.7727197483860286,
+      "grad_norm": 1.8923014402389526,
+      "learning_rate": 1.7408801455013224e-07,
+      "loss": 0.2917,
+      "step": 16750
+    },
+    {
+      "epoch": 2.774375103459692,
+      "grad_norm": 2.412301540374756,
+      "learning_rate": 1.715777445553507e-07,
+      "loss": 0.3357,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7760304585333553,
+      "grad_norm": 2.138453722000122,
+      "learning_rate": 1.6908538874050417e-07,
+      "loss": 0.3632,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7776858136070186,
+      "grad_norm": 2.331045627593994,
+      "learning_rate": 1.6661095635261082e-07,
+      "loss": 0.3331,
+      "step": 16780
+    },
+    {
+      "epoch": 2.779341168680682,
+      "grad_norm": 1.8366329669952393,
+      "learning_rate": 1.6415445657218975e-07,
+      "loss": 0.2916,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7809965237543453,
+      "grad_norm": 1.9621044397354126,
+      "learning_rate": 1.6171589851322999e-07,
+      "loss": 0.3636,
+      "step": 16800
+    },
+    {
+      "epoch": 2.7826518788280086,
+      "grad_norm": 2.012941360473633,
+      "learning_rate": 1.59295291223151e-07,
+      "loss": 0.3652,
+      "step": 16810
+    },
+    {
+      "epoch": 2.784307233901672,
+      "grad_norm": 2.185295343399048,
+      "learning_rate": 1.568926436827739e-07,
+      "loss": 0.3466,
+      "step": 16820
+    },
+    {
+      "epoch": 2.785962588975335,
+      "grad_norm": 2.6037065982818604,
+      "learning_rate": 1.5450796480628637e-07,
+      "loss": 0.3798,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7876179440489985,
+      "grad_norm": 2.552006483078003,
+      "learning_rate": 1.5214126344120894e-07,
+      "loss": 0.3012,
+      "step": 16840
+    },
+    {
+      "epoch": 2.789273299122662,
+      "grad_norm": 1.1729012727737427,
+      "learning_rate": 1.4979254836836266e-07,
+      "loss": 0.3451,
+      "step": 16850
+    },
+    {
+      "epoch": 2.790928654196325,
+      "grad_norm": 2.044630765914917,
+      "learning_rate": 1.4746182830183809e-07,
+      "loss": 0.338,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7925840092699885,
+      "grad_norm": 1.5621954202651978,
+      "learning_rate": 1.4514911188895976e-07,
+      "loss": 0.3422,
+      "step": 16870
+    },
+    {
+      "epoch": 2.794239364343652,
+      "grad_norm": 2.243272066116333,
+      "learning_rate": 1.4285440771025784e-07,
+      "loss": 0.3366,
+      "step": 16880
+    },
+    {
+      "epoch": 2.795894719417315,
+      "grad_norm": 2.275195598602295,
+      "learning_rate": 1.405777242794326e-07,
+      "loss": 0.3844,
+      "step": 16890
+    },
+    {
+      "epoch": 2.7975500744909785,
+      "grad_norm": 2.221266746520996,
+      "learning_rate": 1.3831907004332512e-07,
+      "loss": 0.3765,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7992054295646414,
+      "grad_norm": 2.322007894515991,
+      "learning_rate": 1.3607845338188595e-07,
+      "loss": 0.3003,
+      "step": 16910
+    },
+    {
+      "epoch": 2.8008607846383047,
+      "grad_norm": 1.8471641540527344,
+      "learning_rate": 1.338558826081443e-07,
+      "loss": 0.3199,
+      "step": 16920
+    },
+    {
+      "epoch": 2.802516139711968,
+      "grad_norm": 1.6814085245132446,
+      "learning_rate": 1.3165136596817395e-07,
+      "loss": 0.332,
+      "step": 16930
+    },
+    {
+      "epoch": 2.8041714947856313,
+      "grad_norm": 2.1636056900024414,
+      "learning_rate": 1.2946491164106677e-07,
+      "loss": 0.3538,
+      "step": 16940
+    },
+    {
+      "epoch": 2.8058268498592946,
+      "grad_norm": 2.366356372833252,
+      "learning_rate": 1.27296527738901e-07,
+      "loss": 0.3084,
+      "step": 16950
+    },
+    {
+      "epoch": 2.807482204932958,
+      "grad_norm": 1.5104891061782837,
+      "learning_rate": 1.2514622230670958e-07,
+      "loss": 0.3326,
+      "step": 16960
+    },
+    {
+      "epoch": 2.8091375600066213,
+      "grad_norm": 2.3185157775878906,
+      "learning_rate": 1.230140033224525e-07,
+      "loss": 0.3529,
+      "step": 16970
+    },
+    {
+      "epoch": 2.8107929150802846,
+      "grad_norm": 2.288813352584839,
+      "learning_rate": 1.2089987869698615e-07,
+      "loss": 0.35,
+      "step": 16980
+    },
+    {
+      "epoch": 2.812448270153948,
+      "grad_norm": 1.8733060359954834,
+      "learning_rate": 1.1880385627403345e-07,
+      "loss": 0.3267,
+      "step": 16990
+    },
+    {
+      "epoch": 2.8141036252276113,
+      "grad_norm": 2.303619146347046,
+      "learning_rate": 1.1672594383015656e-07,
+      "loss": 0.3233,
+      "step": 17000
+    },
+    {
+      "epoch": 2.8157589803012746,
+      "grad_norm": 2.3300998210906982,
+      "learning_rate": 1.146661490747264e-07,
+      "loss": 0.3066,
+      "step": 17010
+    },
+    {
+      "epoch": 2.817414335374938,
+      "grad_norm": 2.013810634613037,
+      "learning_rate": 1.1262447964989376e-07,
+      "loss": 0.3384,
+      "step": 17020
+    },
+    {
+      "epoch": 2.8190696904486012,
+      "grad_norm": 2.131016254425049,
+      "learning_rate": 1.1060094313056268e-07,
+      "loss": 0.3419,
+      "step": 17030
+    },
+    {
+      "epoch": 2.8207250455222646,
+      "grad_norm": 2.358128070831299,
+      "learning_rate": 1.0859554702436104e-07,
+      "loss": 0.3166,
+      "step": 17040
+    },
+    {
+      "epoch": 2.822380400595928,
+      "grad_norm": 2.0983312129974365,
+      "learning_rate": 1.0660829877161327e-07,
+      "loss": 0.354,
+      "step": 17050
+    },
+    {
+      "epoch": 2.824035755669591,
+      "grad_norm": 1.8471043109893799,
+      "learning_rate": 1.0463920574531106e-07,
+      "loss": 0.3207,
+      "step": 17060
+    },
+    {
+      "epoch": 2.8256911107432545,
+      "grad_norm": 1.907423973083496,
+      "learning_rate": 1.0268827525108937e-07,
+      "loss": 0.3158,
+      "step": 17070
+    },
+    {
+      "epoch": 2.827346465816918,
+      "grad_norm": 1.6720637083053589,
+      "learning_rate": 1.0075551452719657e-07,
+      "loss": 0.3553,
+      "step": 17080
+    },
+    {
+      "epoch": 2.829001820890581,
+      "grad_norm": 1.6659080982208252,
+      "learning_rate": 9.884093074446932e-08,
+      "loss": 0.3111,
+      "step": 17090
+    },
+    {
+      "epoch": 2.8306571759642445,
+      "grad_norm": 1.5908949375152588,
+      "learning_rate": 9.694453100630275e-08,
+      "loss": 0.3329,
+      "step": 17100
+    },
+    {
+      "epoch": 2.832312531037908,
+      "grad_norm": 1.9934496879577637,
+      "learning_rate": 9.506632234862922e-08,
+      "loss": 0.3384,
+      "step": 17110
+    },
+    {
+      "epoch": 2.833967886111571,
+      "grad_norm": 1.8065050840377808,
+      "learning_rate": 9.320631173988737e-08,
+      "loss": 0.3315,
+      "step": 17120
+    },
+    {
+      "epoch": 2.8356232411852345,
+      "grad_norm": 2.1966464519500732,
+      "learning_rate": 9.136450608099978e-08,
+      "loss": 0.345,
+      "step": 17130
+    },
+    {
+      "epoch": 2.837278596258898,
+      "grad_norm": 1.9803756475448608,
+      "learning_rate": 8.954091220534366e-08,
+      "loss": 0.3242,
+      "step": 17140
+    },
+    {
+      "epoch": 2.838933951332561,
+      "grad_norm": 2.167667865753174,
+      "learning_rate": 8.773553687873082e-08,
+      "loss": 0.3268,
+      "step": 17150
+    },
+    {
+      "epoch": 2.840589306406224,
+      "grad_norm": 2.3544301986694336,
+      "learning_rate": 8.594838679937712e-08,
+      "loss": 0.3454,
+      "step": 17160
+    },
+    {
+      "epoch": 2.8422446614798873,
+      "grad_norm": 1.4057775735855103,
+      "learning_rate": 8.41794685978814e-08,
+      "loss": 0.3432,
+      "step": 17170
+    },
+    {
+      "epoch": 2.8439000165535506,
+      "grad_norm": 1.7976477146148682,
+      "learning_rate": 8.242878883719829e-08,
+      "loss": 0.3196,
+      "step": 17180
+    },
+    {
+      "epoch": 2.845555371627214,
+      "grad_norm": 2.1721675395965576,
+      "learning_rate": 8.0696354012616e-08,
+      "loss": 0.3821,
+      "step": 17190
+    },
+    {
+      "epoch": 2.8472107267008773,
+      "grad_norm": 2.5062599182128906,
+      "learning_rate": 7.898217055173075e-08,
+      "loss": 0.3178,
+      "step": 17200
+    },
+    {
+      "epoch": 2.8488660817745406,
+      "grad_norm": 1.7274291515350342,
+      "learning_rate": 7.728624481442348e-08,
+      "loss": 0.3445,
+      "step": 17210
+    },
+    {
+      "epoch": 2.850521436848204,
+      "grad_norm": 2.225858211517334,
+      "learning_rate": 7.560858309283658e-08,
+      "loss": 0.3166,
+      "step": 17220
+    },
+    {
+      "epoch": 2.8521767919218672,
+      "grad_norm": 1.7840882539749146,
+      "learning_rate": 7.394919161134884e-08,
+      "loss": 0.3327,
+      "step": 17230
+    },
+    {
+      "epoch": 2.8538321469955306,
+      "grad_norm": 2.3786237239837646,
+      "learning_rate": 7.230807652655603e-08,
+      "loss": 0.3784,
+      "step": 17240
+    },
+    {
+      "epoch": 2.855487502069194,
+      "grad_norm": 1.890674352645874,
+      "learning_rate": 7.068524392724319e-08,
+      "loss": 0.3401,
+      "step": 17250
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 2.209174156188965,
+      "learning_rate": 6.908069983436683e-08,
+      "loss": 0.3495,
+      "step": 17260
+    },
+    {
+      "epoch": 2.8587982122165205,
+      "grad_norm": 1.9331399202346802,
+      "learning_rate": 6.749445020102884e-08,
+      "loss": 0.3155,
+      "step": 17270
+    },
+    {
+      "epoch": 2.860453567290184,
+      "grad_norm": 1.9215776920318604,
+      "learning_rate": 6.5926500912456e-08,
+      "loss": 0.3355,
+      "step": 17280
+    },
+    {
+      "epoch": 2.862108922363847,
+      "grad_norm": 2.269278049468994,
+      "learning_rate": 6.437685778597824e-08,
+      "loss": 0.3326,
+      "step": 17290
+    },
+    {
+      "epoch": 2.8637642774375105,
+      "grad_norm": 2.3621773719787598,
+      "learning_rate": 6.284552657100761e-08,
+      "loss": 0.3111,
+      "step": 17300
+    },
+    {
+      "epoch": 2.8654196325111734,
+      "grad_norm": 2.254394769668579,
+      "learning_rate": 6.133251294901443e-08,
+      "loss": 0.3558,
+      "step": 17310
+    },
+    {
+      "epoch": 2.8670749875848367,
+      "grad_norm": 2.3596551418304443,
+      "learning_rate": 5.983782253350944e-08,
+      "loss": 0.3323,
+      "step": 17320
+    },
+    {
+      "epoch": 2.8687303426585,
+      "grad_norm": 2.3255999088287354,
+      "learning_rate": 5.836146087002226e-08,
+      "loss": 0.3472,
+      "step": 17330
+    },
+    {
+      "epoch": 2.8703856977321633,
+      "grad_norm": 2.21858549118042,
+      "learning_rate": 5.690343343607796e-08,
+      "loss": 0.3458,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8720410528058267,
+      "grad_norm": 2.3857204914093018,
+      "learning_rate": 5.54637456411794e-08,
+      "loss": 0.3269,
+      "step": 17350
+    },
+    {
+      "epoch": 2.87369640787949,
+      "grad_norm": 2.049962282180786,
+      "learning_rate": 5.4042402826787746e-08,
+      "loss": 0.3314,
+      "step": 17360
+    },
+    {
+      "epoch": 2.8753517629531533,
+      "grad_norm": 2.2532308101654053,
+      "learning_rate": 5.2639410266299705e-08,
+      "loss": 0.3339,
+      "step": 17370
+    },
+    {
+      "epoch": 2.8770071180268166,
+      "grad_norm": 1.2130613327026367,
+      "learning_rate": 5.1254773165032023e-08,
+      "loss": 0.3056,
+      "step": 17380
+    },
+    {
+      "epoch": 2.87866247310048,
+      "grad_norm": 2.2649145126342773,
+      "learning_rate": 4.988849666019757e-08,
+      "loss": 0.3218,
+      "step": 17390
+    },
+    {
+      "epoch": 2.8803178281741433,
+      "grad_norm": 1.4370695352554321,
+      "learning_rate": 4.854058582089038e-08,
+      "loss": 0.3054,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8819731832478066,
+      "grad_norm": 2.0458033084869385,
+      "learning_rate": 4.7211045648064004e-08,
+      "loss": 0.3226,
+      "step": 17410
+    },
+    {
+      "epoch": 2.88362853832147,
+      "grad_norm": 1.8102025985717773,
+      "learning_rate": 4.589988107451482e-08,
+      "loss": 0.3504,
+      "step": 17420
+    },
+    {
+      "epoch": 2.8852838933951332,
+      "grad_norm": 1.952150583267212,
+      "learning_rate": 4.4607096964862094e-08,
+      "loss": 0.3291,
+      "step": 17430
+    },
+    {
+      "epoch": 2.8869392484687966,
+      "grad_norm": 1.774665355682373,
+      "learning_rate": 4.333269811553187e-08,
+      "loss": 0.3339,
+      "step": 17440
+    },
+    {
+      "epoch": 2.88859460354246,
+      "grad_norm": 2.385225772857666,
+      "learning_rate": 4.20766892547364e-08,
+      "loss": 0.3291,
+      "step": 17450
+    },
+    {
+      "epoch": 2.890249958616123,
+      "grad_norm": 2.228212356567383,
+      "learning_rate": 4.0839075042460875e-08,
+      "loss": 0.3173,
+      "step": 17460
+    },
+    {
+      "epoch": 2.8919053136897865,
+      "grad_norm": 1.6550447940826416,
+      "learning_rate": 3.9619860070440056e-08,
+      "loss": 0.334,
+      "step": 17470
+    },
+    {
+      "epoch": 2.89356066876345,
+      "grad_norm": 2.184173822402954,
+      "learning_rate": 3.841904886214831e-08,
+      "loss": 0.3265,
+      "step": 17480
+    },
+    {
+      "epoch": 2.895216023837113,
+      "grad_norm": 2.3634161949157715,
+      "learning_rate": 3.723664587277631e-08,
+      "loss": 0.3615,
+      "step": 17490
+    },
+    {
+      "epoch": 2.8968713789107765,
+      "grad_norm": 2.1004719734191895,
+      "learning_rate": 3.60726554892199e-08,
+      "loss": 0.3142,
+      "step": 17500
+    },
+    {
+      "epoch": 2.89852673398444,
+      "grad_norm": 1.8737444877624512,
+      "learning_rate": 3.492708203006012e-08,
+      "loss": 0.3214,
+      "step": 17510
+    },
+    {
+      "epoch": 2.900182089058103,
+      "grad_norm": 1.9953153133392334,
+      "learning_rate": 3.3799929745547685e-08,
+      "loss": 0.3365,
+      "step": 17520
+    },
+    {
+      "epoch": 2.9018374441317665,
+      "grad_norm": 1.4854812622070312,
+      "learning_rate": 3.2691202817589086e-08,
+      "loss": 0.3054,
+      "step": 17530
+    },
+    {
+      "epoch": 2.90349279920543,
+      "grad_norm": 2.8899099826812744,
+      "learning_rate": 3.160090535972993e-08,
+      "loss": 0.3153,
+      "step": 17540
+    },
+    {
+      "epoch": 2.905148154279093,
+      "grad_norm": 2.137659788131714,
+      "learning_rate": 3.052904141713886e-08,
+      "loss": 0.3037,
+      "step": 17550
+    },
+    {
+      "epoch": 2.9068035093527564,
+      "grad_norm": 2.0721936225891113,
+      "learning_rate": 2.9475614966594233e-08,
+      "loss": 0.3211,
+      "step": 17560
+    },
+    {
+      "epoch": 2.9084588644264193,
+      "grad_norm": 2.115863561630249,
+      "learning_rate": 2.844062991646801e-08,
+      "loss": 0.2805,
+      "step": 17570
+    },
+    {
+      "epoch": 2.9101142195000826,
+      "grad_norm": 1.711255431175232,
+      "learning_rate": 2.742409010671243e-08,
+      "loss": 0.3288,
+      "step": 17580
+    },
+    {
+      "epoch": 2.911769574573746,
+      "grad_norm": 2.0368499755859375,
+      "learning_rate": 2.642599930884393e-08,
+      "loss": 0.3121,
+      "step": 17590
+    },
+    {
+      "epoch": 2.9134249296474093,
+      "grad_norm": 2.2370572090148926,
+      "learning_rate": 2.544636122593147e-08,
+      "loss": 0.3161,
+      "step": 17600
+    },
+    {
+      "epoch": 2.9150802847210726,
+      "grad_norm": 2.0604302883148193,
+      "learning_rate": 2.4485179492581e-08,
+      "loss": 0.3285,
+      "step": 17610
+    },
+    {
+      "epoch": 2.916735639794736,
+      "grad_norm": 2.018296003341675,
+      "learning_rate": 2.354245767492269e-08,
+      "loss": 0.3635,
+      "step": 17620
+    },
+    {
+      "epoch": 2.9183909948683993,
+      "grad_norm": 2.0412917137145996,
+      "learning_rate": 2.2618199270597607e-08,
+      "loss": 0.3264,
+      "step": 17630
+    },
+    {
+      "epoch": 2.9200463499420626,
+      "grad_norm": 1.991478681564331,
+      "learning_rate": 2.1712407708744386e-08,
+      "loss": 0.3597,
+      "step": 17640
+    },
+    {
+      "epoch": 2.921701705015726,
+      "grad_norm": 1.9765408039093018,
+      "learning_rate": 2.0825086349988145e-08,
+      "loss": 0.3418,
+      "step": 17650
+    },
+    {
+      "epoch": 2.9233570600893892,
+      "grad_norm": 2.1275086402893066,
+      "learning_rate": 1.995623848642547e-08,
+      "loss": 0.3345,
+      "step": 17660
+    },
+    {
+      "epoch": 2.9250124151630525,
+      "grad_norm": 1.7225489616394043,
+      "learning_rate": 1.9105867341613903e-08,
+      "loss": 0.3135,
+      "step": 17670
+    },
+    {
+      "epoch": 2.926667770236716,
+      "grad_norm": 1.9626611471176147,
+      "learning_rate": 1.8273976070559695e-08,
+      "loss": 0.3614,
+      "step": 17680
+    },
+    {
+      "epoch": 2.928323125310379,
+      "grad_norm": 2.476229667663574,
+      "learning_rate": 1.7460567759705615e-08,
+      "loss": 0.3594,
+      "step": 17690
+    },
+    {
+      "epoch": 2.9299784803840425,
+      "grad_norm": 1.9182066917419434,
+      "learning_rate": 1.6665645426920396e-08,
+      "loss": 0.3197,
+      "step": 17700
+    },
+    {
+      "epoch": 2.931633835457706,
+      "grad_norm": 2.101400852203369,
+      "learning_rate": 1.5889212021485966e-08,
+      "loss": 0.3522,
+      "step": 17710
+    },
+    {
+      "epoch": 2.9332891905313687,
+      "grad_norm": 1.9222100973129272,
+      "learning_rate": 1.5131270424088573e-08,
+      "loss": 0.3461,
+      "step": 17720
+    },
+    {
+      "epoch": 2.934944545605032,
+      "grad_norm": 2.66011905670166,
+      "learning_rate": 1.4391823446807117e-08,
+      "loss": 0.3557,
+      "step": 17730
+    },
+    {
+      "epoch": 2.9365999006786954,
+      "grad_norm": 2.2789878845214844,
+      "learning_rate": 1.3670873833101505e-08,
+      "loss": 0.3229,
+      "step": 17740
+    },
+    {
+      "epoch": 2.9382552557523587,
+      "grad_norm": 2.3933591842651367,
+      "learning_rate": 1.2968424257804313e-08,
+      "loss": 0.3315,
+      "step": 17750
+    },
+    {
+      "epoch": 2.939910610826022,
+      "grad_norm": 2.3596224784851074,
+      "learning_rate": 1.22844773271108e-08,
+      "loss": 0.3439,
+      "step": 17760
+    },
+    {
+      "epoch": 2.9415659658996853,
+      "grad_norm": 2.155430316925049,
+      "learning_rate": 1.161903557856725e-08,
+      "loss": 0.3253,
+      "step": 17770
+    },
+    {
+      "epoch": 2.9432213209733487,
+      "grad_norm": 2.187922954559326,
+      "learning_rate": 1.0972101481063758e-08,
+      "loss": 0.3758,
+      "step": 17780
+    },
+    {
+      "epoch": 2.944876676047012,
+      "grad_norm": 2.4473180770874023,
+      "learning_rate": 1.0343677434824783e-08,
+      "loss": 0.3408,
+      "step": 17790
+    },
+    {
+      "epoch": 2.9465320311206753,
+      "grad_norm": 2.050906181335449,
+      "learning_rate": 9.733765771398063e-09,
+      "loss": 0.3365,
+      "step": 17800
+    },
+    {
+      "epoch": 2.9481873861943386,
+      "grad_norm": 2.1611318588256836,
+      "learning_rate": 9.142368753649045e-09,
+      "loss": 0.3467,
+      "step": 17810
+    },
+    {
+      "epoch": 2.949842741268002,
+      "grad_norm": 2.38580322265625,
+      "learning_rate": 8.56948857575035e-09,
+      "loss": 0.314,
+      "step": 17820
+    },
+    {
+      "epoch": 2.9514980963416653,
+      "grad_norm": 1.6547008752822876,
+      "learning_rate": 8.015127363174558e-09,
+      "loss": 0.3313,
+      "step": 17830
+    },
+    {
+      "epoch": 2.9531534514153286,
+      "grad_norm": 2.329861640930176,
+      "learning_rate": 7.479287172685867e-09,
+      "loss": 0.3271,
+      "step": 17840
+    },
+    {
+      "epoch": 2.954808806488992,
+      "grad_norm": 1.9690601825714111,
+      "learning_rate": 6.96196999233234e-09,
+      "loss": 0.3351,
+      "step": 17850
+    },
+    {
+      "epoch": 2.9564641615626552,
+      "grad_norm": 2.030273199081421,
+      "learning_rate": 6.463177741440341e-09,
+      "loss": 0.3587,
+      "step": 17860
+    },
+    {
+      "epoch": 2.9581195166363186,
+      "grad_norm": 2.2005035877227783,
+      "learning_rate": 5.982912270603991e-09,
+      "loss": 0.3264,
+      "step": 17870
+    },
+    {
+      "epoch": 2.959774871709982,
+      "grad_norm": 2.0540130138397217,
+      "learning_rate": 5.521175361681286e-09,
+      "loss": 0.3387,
+      "step": 17880
+    },
+    {
+      "epoch": 2.961430226783645,
+      "grad_norm": 1.6892403364181519,
+      "learning_rate": 5.07796872778632e-09,
+      "loss": 0.2986,
+      "step": 17890
+    },
+    {
+      "epoch": 2.9630855818573085,
+      "grad_norm": 1.895839810371399,
+      "learning_rate": 4.653294013283183e-09,
+      "loss": 0.3291,
+      "step": 17900
+    },
+    {
+      "epoch": 2.964740936930972,
+      "grad_norm": 1.6137304306030273,
+      "learning_rate": 4.247152793779296e-09,
+      "loss": 0.3265,
+      "step": 17910
+    },
+    {
+      "epoch": 2.966396292004635,
+      "grad_norm": 1.9061734676361084,
+      "learning_rate": 3.859546576120421e-09,
+      "loss": 0.334,
+      "step": 17920
+    },
+    {
+      "epoch": 2.9680516470782985,
+      "grad_norm": 2.028872013092041,
+      "learning_rate": 3.4904767983851006e-09,
+      "loss": 0.3213,
+      "step": 17930
+    },
+    {
+      "epoch": 2.969707002151962,
+      "grad_norm": 1.7429839372634888,
+      "learning_rate": 3.1399448298774503e-09,
+      "loss": 0.3341,
+      "step": 17940
+    },
+    {
+      "epoch": 2.971362357225625,
+      "grad_norm": 2.5334107875823975,
+      "learning_rate": 2.8079519711243786e-09,
+      "loss": 0.3251,
+      "step": 17950
+    },
+    {
+      "epoch": 2.9730177122992885,
+      "grad_norm": 2.0650694370269775,
+      "learning_rate": 2.4944994538700363e-09,
+      "loss": 0.2856,
+      "step": 17960
+    },
+    {
+      "epoch": 2.9746730673729513,
+      "grad_norm": 1.9629024267196655,
+      "learning_rate": 2.1995884410708212e-09,
+      "loss": 0.2924,
+      "step": 17970
+    },
+    {
+      "epoch": 2.9763284224466147,
+      "grad_norm": 1.9057166576385498,
+      "learning_rate": 1.9232200268914923e-09,
+      "loss": 0.3179,
+      "step": 17980
+    },
+    {
+      "epoch": 2.977983777520278,
+      "grad_norm": 2.323155164718628,
+      "learning_rate": 1.6653952367007287e-09,
+      "loss": 0.3351,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9796391325939413,
+      "grad_norm": 1.7210339307785034,
+      "learning_rate": 1.426115027067243e-09,
+      "loss": 0.3269,
+      "step": 18000
+    },
+    {
+      "epoch": 2.9812944876676046,
+      "grad_norm": 1.83793044090271,
+      "learning_rate": 1.2053802857581176e-09,
+      "loss": 0.347,
+      "step": 18010
+    },
+    {
+      "epoch": 2.982949842741268,
+      "grad_norm": 1.847522258758545,
+      "learning_rate": 1.003191831731587e-09,
+      "loss": 0.2979,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9846051978149313,
+      "grad_norm": 1.777446985244751,
+      "learning_rate": 8.195504151381484e-10,
+      "loss": 0.3473,
+      "step": 18030
+    },
+    {
+      "epoch": 2.9862605528885946,
+      "grad_norm": 2.9953622817993164,
+      "learning_rate": 6.544567173150107e-10,
+      "loss": 0.3001,
+      "step": 18040
+    },
+    {
+      "epoch": 2.987915907962258,
+      "grad_norm": 1.8776600360870361,
+      "learning_rate": 5.079113507849842e-10,
+      "loss": 0.357,
+      "step": 18050
+    },
+    {
+      "epoch": 2.9895712630359212,
+      "grad_norm": 2.086153268814087,
+      "learning_rate": 3.7991485925370496e-10,
+      "loss": 0.3004,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9912266181095846,
+      "grad_norm": 2.1379270553588867,
+      "learning_rate": 2.704677176079695e-10,
+      "loss": 0.3721,
+      "step": 18070
+    },
+    {
+      "epoch": 2.992881973183248,
+      "grad_norm": 2.7157349586486816,
+      "learning_rate": 1.7957033191240425e-10,
+      "loss": 0.3555,
+      "step": 18080
+    },
+    {
+      "epoch": 2.994537328256911,
+      "grad_norm": 2.5094964504241943,
+      "learning_rate": 1.0722303941113065e-10,
+      "loss": 0.3435,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9961926833305745,
+      "grad_norm": 2.10398268699646,
+      "learning_rate": 5.3426108522769415e-11,
+      "loss": 0.3287,
+      "step": 18100
+    },
+    {
+      "epoch": 2.997848038404238,
+      "grad_norm": 3.0506503582000732,
+      "learning_rate": 1.8179738842660777e-11,
+      "loss": 0.3678,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9995033934779007,
+      "grad_norm": 1.7245928049087524,
+      "learning_rate": 1.484061139533921e-12,
+      "loss": 0.3259,
+      "step": 18120
+    },
+    {
+      "epoch": 3.0,
+      "step": 18123,
+      "total_flos": 1.2130803771160658e+18,
+      "train_loss": 0.45045166753521493,
+      "train_runtime": 7695.2204,
+      "train_samples_per_second": 37.681,
+      "train_steps_per_second": 2.355
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 18123,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2130803771160658e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}