diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35490 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9991116375481197,
+  "eval_steps": 500,
+  "global_step": 5064,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005922416345869114,
+      "grad_norm": 18.717661840154292,
+      "learning_rate": 1.3157894736842107e-07,
+      "loss": 1.1434,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011844832691738228,
+      "grad_norm": 15.975211698366682,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 1.1722,
+      "step": 2
+    },
+    {
+      "epoch": 0.0017767249037607344,
+      "grad_norm": 13.436630281888167,
+      "learning_rate": 3.9473684210526315e-07,
+      "loss": 1.2216,
+      "step": 3
+    },
+    {
+      "epoch": 0.0023689665383476457,
+      "grad_norm": 14.141974273595771,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 1.2084,
+      "step": 4
+    },
+    {
+      "epoch": 0.002961208172934557,
+      "grad_norm": 12.817546549612011,
+      "learning_rate": 6.578947368421053e-07,
+      "loss": 1.1613,
+      "step": 5
+    },
+    {
+      "epoch": 0.0035534498075214687,
+      "grad_norm": 15.101738254046671,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 1.2291,
+      "step": 6
+    },
+    {
+      "epoch": 0.00414569144210838,
+      "grad_norm": 21.872300636479174,
+      "learning_rate": 9.210526315789474e-07,
+      "loss": 1.2358,
+      "step": 7
+    },
+    {
+      "epoch": 0.004737933076695291,
+      "grad_norm": 18.219829859695743,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 1.2111,
+      "step": 8
+    },
+    {
+      "epoch": 0.005330174711282203,
+      "grad_norm": 11.651800626605043,
+      "learning_rate": 1.1842105263157894e-06,
+      "loss": 1.1699,
+      "step": 9
+    },
+    {
+      "epoch": 0.005922416345869114,
+      "grad_norm": 14.428397731094817,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 1.2005,
+      "step": 10
+    },
+    {
+      "epoch": 0.006514657980456026,
+      "grad_norm": 11.505996928681398,
+      "learning_rate": 1.4473684210526317e-06,
+      "loss": 1.1428,
+      "step": 11
+    },
+    {
+      "epoch": 0.0071068996150429374,
+      "grad_norm": 9.258825562856831,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 1.1801,
+      "step": 12
+    },
+    {
+      "epoch": 0.007699141249629849,
+      "grad_norm": 11.003890134146193,
+      "learning_rate": 1.710526315789474e-06,
+      "loss": 1.1603,
+      "step": 13
+    },
+    {
+      "epoch": 0.00829138288421676,
+      "grad_norm": 8.53365801142655,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 1.1299,
+      "step": 14
+    },
+    {
+      "epoch": 0.008883624518803672,
+      "grad_norm": 9.898879357555042,
+      "learning_rate": 1.973684210526316e-06,
+      "loss": 1.1691,
+      "step": 15
+    },
+    {
+      "epoch": 0.009475866153390583,
+      "grad_norm": 10.549302142463828,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 1.1276,
+      "step": 16
+    },
+    {
+      "epoch": 0.010068107787977496,
+      "grad_norm": 13.735637524962215,
+      "learning_rate": 2.236842105263158e-06,
+      "loss": 1.0958,
+      "step": 17
+    },
+    {
+      "epoch": 0.010660349422564407,
+      "grad_norm": 9.075762665514619,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 1.1352,
+      "step": 18
+    },
+    {
+      "epoch": 0.011252591057151318,
+      "grad_norm": 8.125019295709604,
+      "learning_rate": 2.5e-06,
+      "loss": 1.0965,
+      "step": 19
+    },
+    {
+      "epoch": 0.011844832691738229,
+      "grad_norm": 6.28753990985354,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 1.0706,
+      "step": 20
+    },
+    {
+      "epoch": 0.01243707432632514,
+      "grad_norm": 9.479465602728467,
+      "learning_rate": 2.7631578947368424e-06,
+      "loss": 1.1351,
+      "step": 21
+    },
+    {
+      "epoch": 0.013029315960912053,
+      "grad_norm": 10.298865353341801,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 1.1403,
+      "step": 22
+    },
+    {
+      "epoch": 0.013621557595498964,
+      "grad_norm": 37.987752673055155,
+      "learning_rate": 3.0263157894736843e-06,
+      "loss": 1.0901,
+      "step": 23
+    },
+    {
+      "epoch": 0.014213799230085875,
+      "grad_norm": 9.457834823064632,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 1.1241,
+      "step": 24
+    },
+    {
+      "epoch": 0.014806040864672786,
+      "grad_norm": 6.66281436022819,
+      "learning_rate": 3.289473684210527e-06,
+      "loss": 1.0351,
+      "step": 25
+    },
+    {
+      "epoch": 0.015398282499259699,
+      "grad_norm": 6.610062092413164,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 1.0715,
+      "step": 26
+    },
+    {
+      "epoch": 0.01599052413384661,
+      "grad_norm": 6.998988378675071,
+      "learning_rate": 3.5526315789473687e-06,
+      "loss": 1.0664,
+      "step": 27
+    },
+    {
+      "epoch": 0.01658276576843352,
+      "grad_norm": 6.559109520573562,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 1.0371,
+      "step": 28
+    },
+    {
+      "epoch": 0.017175007403020432,
+      "grad_norm": 8.1722379764705,
+      "learning_rate": 3.815789473684211e-06,
+      "loss": 1.0594,
+      "step": 29
+    },
+    {
+      "epoch": 0.017767249037607343,
+      "grad_norm": 9.992648704415512,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 1.0802,
+      "step": 30
+    },
+    {
+      "epoch": 0.018359490672194254,
+      "grad_norm": 6.773662148667347,
+      "learning_rate": 4.078947368421053e-06,
+      "loss": 0.9767,
+      "step": 31
+    },
+    {
+      "epoch": 0.018951732306781165,
+      "grad_norm": 5.160580354167821,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 1.0125,
+      "step": 32
+    },
+    {
+      "epoch": 0.019543973941368076,
+      "grad_norm": 7.601356319162516,
+      "learning_rate": 4.342105263157895e-06,
+      "loss": 1.0109,
+      "step": 33
+    },
+    {
+      "epoch": 0.02013621557595499,
+      "grad_norm": 6.835933036484314,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 1.0355,
+      "step": 34
+    },
+    {
+      "epoch": 0.020728457210541902,
+      "grad_norm": 14.435112479372032,
+      "learning_rate": 4.605263157894737e-06,
+      "loss": 1.0154,
+      "step": 35
+    },
+    {
+      "epoch": 0.021320698845128813,
+      "grad_norm": 7.829860179549529,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 1.04,
+      "step": 36
+    },
+    {
+      "epoch": 0.021912940479715724,
+      "grad_norm": 7.169740079807402,
+      "learning_rate": 4.8684210526315795e-06,
+      "loss": 0.9694,
+      "step": 37
+    },
+    {
+      "epoch": 0.022505182114302635,
+      "grad_norm": 8.83069550799535,
+      "learning_rate": 5e-06,
+      "loss": 0.9532,
+      "step": 38
+    },
+    {
+      "epoch": 0.023097423748889546,
+      "grad_norm": 9.131263259747842,
+      "learning_rate": 5.131578947368422e-06,
+      "loss": 0.9806,
+      "step": 39
+    },
+    {
+      "epoch": 0.023689665383476458,
+      "grad_norm": 10.809549787669706,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.0129,
+      "step": 40
+    },
+    {
+      "epoch": 0.02428190701806337,
+      "grad_norm": 8.804492257877177,
+      "learning_rate": 5.394736842105264e-06,
+      "loss": 1.0173,
+      "step": 41
+    },
+    {
+      "epoch": 0.02487414865265028,
+      "grad_norm": 20.71245515215828,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 0.9932,
+      "step": 42
+    },
+    {
+      "epoch": 0.025466390287237194,
+      "grad_norm": 9.655846535170545,
+      "learning_rate": 5.657894736842106e-06,
+      "loss": 0.9879,
+      "step": 43
+    },
+    {
+      "epoch": 0.026058631921824105,
+      "grad_norm": 6.537124531460736,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 1.0188,
+      "step": 44
+    },
+    {
+      "epoch": 0.026650873556411016,
+      "grad_norm": 13.091149382863613,
+      "learning_rate": 5.921052631578948e-06,
+      "loss": 0.9964,
+      "step": 45
+    },
+    {
+      "epoch": 0.027243115190997928,
+      "grad_norm": 8.006244083133073,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 0.9726,
+      "step": 46
+    },
+    {
+      "epoch": 0.02783535682558484,
+      "grad_norm": 7.668734764475963,
+      "learning_rate": 6.18421052631579e-06,
+      "loss": 1.0153,
+      "step": 47
+    },
+    {
+      "epoch": 0.02842759846017175,
+      "grad_norm": 8.137447421123358,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 1.0091,
+      "step": 48
+    },
+    {
+      "epoch": 0.02901984009475866,
+      "grad_norm": 7.606771295715497,
+      "learning_rate": 6.447368421052632e-06,
+      "loss": 0.9991,
+      "step": 49
+    },
+    {
+      "epoch": 0.029612081729345572,
+      "grad_norm": 7.345858891479215,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 0.9509,
+      "step": 50
+    },
+    {
+      "epoch": 0.030204323363932483,
+      "grad_norm": 7.643466506554061,
+      "learning_rate": 6.710526315789474e-06,
+      "loss": 0.9572,
+      "step": 51
+    },
+    {
+      "epoch": 0.030796564998519398,
+      "grad_norm": 13.10135554836337,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.9797,
+      "step": 52
+    },
+    {
+      "epoch": 0.03138880663310631,
+      "grad_norm": 63.002320155093024,
+      "learning_rate": 6.973684210526316e-06,
+      "loss": 0.9633,
+      "step": 53
+    },
+    {
+      "epoch": 0.03198104826769322,
+      "grad_norm": 6.071766418833052,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 0.9146,
+      "step": 54
+    },
+    {
+      "epoch": 0.03257328990228013,
+      "grad_norm": 15.636734262160939,
+      "learning_rate": 7.236842105263158e-06,
+      "loss": 1.0414,
+      "step": 55
+    },
+    {
+      "epoch": 0.03316553153686704,
+      "grad_norm": 22.37892911472697,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.9515,
+      "step": 56
+    },
+    {
+      "epoch": 0.03375777317145395,
+      "grad_norm": 11.914023755244859,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.9538,
+      "step": 57
+    },
+    {
+      "epoch": 0.034350014806040864,
+      "grad_norm": 8.586686698837346,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 0.9835,
+      "step": 58
+    },
+    {
+      "epoch": 0.034942256440627775,
+      "grad_norm": 27.707875718601368,
+      "learning_rate": 7.763157894736843e-06,
+      "loss": 1.0073,
+      "step": 59
+    },
+    {
+      "epoch": 0.035534498075214686,
+      "grad_norm": 24.521633695176586,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.961,
+      "step": 60
+    },
+    {
+      "epoch": 0.0361267397098016,
+      "grad_norm": 100.54943552048488,
+      "learning_rate": 8.026315789473685e-06,
+      "loss": 1.0181,
+      "step": 61
+    },
+    {
+      "epoch": 0.03671898134438851,
+      "grad_norm": 9.987380890435652,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 0.9711,
+      "step": 62
+    },
+    {
+      "epoch": 0.03731122297897542,
+      "grad_norm": 9.222884295954952,
+      "learning_rate": 8.289473684210526e-06,
+      "loss": 0.9489,
+      "step": 63
+    },
+    {
+      "epoch": 0.03790346461356233,
+      "grad_norm": 19.82352747253693,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.9661,
+      "step": 64
+    },
+    {
+      "epoch": 0.03849570624814924,
+      "grad_norm": 8.36286785237061,
+      "learning_rate": 8.552631578947368e-06,
+      "loss": 0.9554,
+      "step": 65
+    },
+    {
+      "epoch": 0.03908794788273615,
+      "grad_norm": 14.58438374105295,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 0.9321,
+      "step": 66
+    },
+    {
+      "epoch": 0.03968018951732307,
+      "grad_norm": 12.929431130450986,
+      "learning_rate": 8.81578947368421e-06,
+      "loss": 0.9036,
+      "step": 67
+    },
+    {
+      "epoch": 0.04027243115190998,
+      "grad_norm": 16.089145436097088,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.9221,
+      "step": 68
+    },
+    {
+      "epoch": 0.04086467278649689,
+      "grad_norm": 9.617533163896221,
+      "learning_rate": 9.078947368421054e-06,
+      "loss": 0.9167,
+      "step": 69
+    },
+    {
+      "epoch": 0.041456914421083804,
+      "grad_norm": 7.694991032785801,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.9336,
+      "step": 70
+    },
+    {
+      "epoch": 0.042049156055670715,
+      "grad_norm": 14.08369539185404,
+      "learning_rate": 9.342105263157895e-06,
+      "loss": 0.9538,
+      "step": 71
+    },
+    {
+      "epoch": 0.042641397690257626,
+      "grad_norm": 7.803725255120761,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.9534,
+      "step": 72
+    },
+    {
+      "epoch": 0.04323363932484454,
+      "grad_norm": 10.048093652946386,
+      "learning_rate": 9.605263157894737e-06,
+      "loss": 1.0111,
+      "step": 73
+    },
+    {
+      "epoch": 0.04382588095943145,
+      "grad_norm": 8.06770139647209,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 1.0268,
+      "step": 74
+    },
+    {
+      "epoch": 0.04441812259401836,
+      "grad_norm": 26.383451642079425,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 0.9725,
+      "step": 75
+    },
+    {
+      "epoch": 0.04501036422860527,
+      "grad_norm": 10.64378030761482,
+      "learning_rate": 1e-05,
+      "loss": 0.9351,
+      "step": 76
+    },
+    {
+      "epoch": 0.04560260586319218,
+      "grad_norm": 17.542208054609976,
+      "learning_rate": 1.0131578947368421e-05,
+      "loss": 0.9737,
+      "step": 77
+    },
+    {
+      "epoch": 0.04619484749777909,
+      "grad_norm": 4.904257105145783,
+      "learning_rate": 1.0263157894736844e-05,
+      "loss": 0.9626,
+      "step": 78
+    },
+    {
+      "epoch": 0.046787089132366004,
+      "grad_norm": 11.772948651444537,
+      "learning_rate": 1.0394736842105264e-05,
+      "loss": 0.9473,
+      "step": 79
+    },
+    {
+      "epoch": 0.047379330766952915,
+      "grad_norm": 5.711100010911588,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.9313,
+      "step": 80
+    },
+    {
+      "epoch": 0.047971572401539826,
+      "grad_norm": 17.763080812645768,
+      "learning_rate": 1.0657894736842108e-05,
+      "loss": 0.9619,
+      "step": 81
+    },
+    {
+      "epoch": 0.04856381403612674,
+      "grad_norm": 48.7479319965714,
+      "learning_rate": 1.0789473684210528e-05,
+      "loss": 0.9389,
+      "step": 82
+    },
+    {
+      "epoch": 0.04915605567071365,
+      "grad_norm": 12.652307160279797,
+      "learning_rate": 1.0921052631578948e-05,
+      "loss": 0.9623,
+      "step": 83
+    },
+    {
+      "epoch": 0.04974829730530056,
+      "grad_norm": 5.463588308594221,
+      "learning_rate": 1.105263157894737e-05,
+      "loss": 0.9464,
+      "step": 84
+    },
+    {
+      "epoch": 0.05034053893988748,
+      "grad_norm": 24.19528693762655,
+      "learning_rate": 1.1184210526315792e-05,
+      "loss": 0.8953,
+      "step": 85
+    },
+    {
+      "epoch": 0.05093278057447439,
+      "grad_norm": 6.29193429817291,
+      "learning_rate": 1.1315789473684212e-05,
+      "loss": 0.9005,
+      "step": 86
+    },
+    {
+      "epoch": 0.0515250222090613,
+      "grad_norm": 13.36573014121912,
+      "learning_rate": 1.1447368421052632e-05,
+      "loss": 0.9075,
+      "step": 87
+    },
+    {
+      "epoch": 0.05211726384364821,
+      "grad_norm": 10.377086501174952,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 0.9572,
+      "step": 88
+    },
+    {
+      "epoch": 0.05270950547823512,
+      "grad_norm": 8.69662787671184,
+      "learning_rate": 1.1710526315789475e-05,
+      "loss": 0.934,
+      "step": 89
+    },
+    {
+      "epoch": 0.05330174711282203,
+      "grad_norm": 12.914962830795819,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 0.9381,
+      "step": 90
+    },
+    {
+      "epoch": 0.053893988747408944,
+      "grad_norm": 5.960162225544222,
+      "learning_rate": 1.1973684210526317e-05,
+      "loss": 0.8615,
+      "step": 91
+    },
+    {
+      "epoch": 0.054486230381995855,
+      "grad_norm": 14.528434970541078,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 0.9648,
+      "step": 92
+    },
+    {
+      "epoch": 0.055078472016582766,
+      "grad_norm": 10.384675365673601,
+      "learning_rate": 1.2236842105263159e-05,
+      "loss": 0.912,
+      "step": 93
+    },
+    {
+      "epoch": 0.05567071365116968,
+      "grad_norm": 7.33465251957887,
+      "learning_rate": 1.236842105263158e-05,
+      "loss": 0.9293,
+      "step": 94
+    },
+    {
+      "epoch": 0.05626295528575659,
+      "grad_norm": 4.727007339545679,
+      "learning_rate": 1.25e-05,
+      "loss": 0.9342,
+      "step": 95
+    },
+    {
+      "epoch": 0.0568551969203435,
+      "grad_norm": 5.324071485836242,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 0.9379,
+      "step": 96
+    },
+    {
+      "epoch": 0.05744743855493041,
+      "grad_norm": 7.490618326236765,
+      "learning_rate": 1.2763157894736844e-05,
+      "loss": 0.8976,
+      "step": 97
+    },
+    {
+      "epoch": 0.05803968018951732,
+      "grad_norm": 10.570705779798235,
+      "learning_rate": 1.2894736842105264e-05,
+      "loss": 0.9339,
+      "step": 98
+    },
+    {
+      "epoch": 0.05863192182410423,
+      "grad_norm": 6.638002466312831,
+      "learning_rate": 1.3026315789473684e-05,
+      "loss": 0.9004,
+      "step": 99
+    },
+    {
+      "epoch": 0.059224163458691144,
+      "grad_norm": 6.082259049790152,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 0.9741,
+      "step": 100
+    },
+    {
+      "epoch": 0.059816405093278055,
+      "grad_norm": 27.384068364009153,
+      "learning_rate": 1.3289473684210528e-05,
+      "loss": 0.9887,
+      "step": 101
+    },
+    {
+      "epoch": 0.060408646727864966,
+      "grad_norm": 7.414470248625567,
+      "learning_rate": 1.3421052631578948e-05,
+      "loss": 0.9548,
+      "step": 102
+    },
+    {
+      "epoch": 0.06100088836245188,
+      "grad_norm": 12.181747128869292,
+      "learning_rate": 1.3552631578947371e-05,
+      "loss": 0.9342,
+      "step": 103
+    },
+    {
+      "epoch": 0.061593129997038795,
+      "grad_norm": 18.814533980462837,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 0.8821,
+      "step": 104
+    },
+    {
+      "epoch": 0.062185371631625706,
+      "grad_norm": 11.37830438605211,
+      "learning_rate": 1.3815789473684211e-05,
+      "loss": 0.9453,
+      "step": 105
+    },
+    {
+      "epoch": 0.06277761326621262,
+      "grad_norm": 7.700584289573716,
+      "learning_rate": 1.3947368421052631e-05,
+      "loss": 0.9137,
+      "step": 106
+    },
+    {
+      "epoch": 0.06336985490079952,
+      "grad_norm": 24.239676785791914,
+      "learning_rate": 1.4078947368421055e-05,
+      "loss": 0.9604,
+      "step": 107
+    },
+    {
+      "epoch": 0.06396209653538644,
+      "grad_norm": 15.296887074877906,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 0.9241,
+      "step": 108
+    },
+    {
+      "epoch": 0.06455433816997334,
+      "grad_norm": 5.911563958223732,
+      "learning_rate": 1.4342105263157895e-05,
+      "loss": 0.8995,
+      "step": 109
+    },
+    {
+      "epoch": 0.06514657980456026,
+      "grad_norm": 6.472036308433493,
+      "learning_rate": 1.4473684210526317e-05,
+      "loss": 0.9123,
+      "step": 110
+    },
+    {
+      "epoch": 0.06573882143914717,
+      "grad_norm": 8.838225442941134,
+      "learning_rate": 1.4605263157894739e-05,
+      "loss": 0.866,
+      "step": 111
+    },
+    {
+      "epoch": 0.06633106307373408,
+      "grad_norm": 6.179553262811797,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 0.9363,
+      "step": 112
+    },
+    {
+      "epoch": 0.06692330470832099,
+      "grad_norm": 8.05988074037456,
+      "learning_rate": 1.486842105263158e-05,
+      "loss": 0.8737,
+      "step": 113
+    },
+    {
+      "epoch": 0.0675155463429079,
+      "grad_norm": 8.800587055109485,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.9535,
+      "step": 114
+    },
+    {
+      "epoch": 0.06810778797749482,
+      "grad_norm": 14.297791370718492,
+      "learning_rate": 1.5131578947368422e-05,
+      "loss": 0.9063,
+      "step": 115
+    },
+    {
+      "epoch": 0.06870002961208173,
+      "grad_norm": 12.532931475700789,
+      "learning_rate": 1.5263157894736846e-05,
+      "loss": 0.9646,
+      "step": 116
+    },
+    {
+      "epoch": 0.06929227124666865,
+      "grad_norm": 10.961935762322105,
+      "learning_rate": 1.5394736842105264e-05,
+      "loss": 0.9074,
+      "step": 117
+    },
+    {
+      "epoch": 0.06988451288125555,
+      "grad_norm": 16.197406867612408,
+      "learning_rate": 1.5526315789473686e-05,
+      "loss": 0.9213,
+      "step": 118
+    },
+    {
+      "epoch": 0.07047675451584247,
+      "grad_norm": 4.618325199175887,
+      "learning_rate": 1.5657894736842107e-05,
+      "loss": 0.9723,
+      "step": 119
+    },
+    {
+      "epoch": 0.07106899615042937,
+      "grad_norm": 12.33412742358314,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 0.9582,
+      "step": 120
+    },
+    {
+      "epoch": 0.07166123778501629,
+      "grad_norm": 8.051122431463254,
+      "learning_rate": 1.5921052631578948e-05,
+      "loss": 0.9051,
+      "step": 121
+    },
+    {
+      "epoch": 0.0722534794196032,
+      "grad_norm": 247.48175705261465,
+      "learning_rate": 1.605263157894737e-05,
+      "loss": 0.9365,
+      "step": 122
+    },
+    {
+      "epoch": 0.07284572105419011,
+      "grad_norm": 4.117202388731062,
+      "learning_rate": 1.618421052631579e-05,
+      "loss": 0.919,
+      "step": 123
+    },
+    {
+      "epoch": 0.07343796268877702,
+      "grad_norm": 4.4180789641519755,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.9452,
+      "step": 124
+    },
+    {
+      "epoch": 0.07403020432336394,
+      "grad_norm": 10.408859163405296,
+      "learning_rate": 1.644736842105263e-05,
+      "loss": 0.9171,
+      "step": 125
+    },
+    {
+      "epoch": 0.07462244595795084,
+      "grad_norm": 5.812635898712064,
+      "learning_rate": 1.6578947368421053e-05,
+      "loss": 0.9144,
+      "step": 126
+    },
+    {
+      "epoch": 0.07521468759253776,
+      "grad_norm": 16.681339227878798,
+      "learning_rate": 1.6710526315789475e-05,
+      "loss": 0.9183,
+      "step": 127
+    },
+    {
+      "epoch": 0.07580692922712466,
+      "grad_norm": 5.4477969668787445,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 0.9358,
+      "step": 128
+    },
+    {
+      "epoch": 0.07639917086171158,
+      "grad_norm": 6.20964288619415,
+      "learning_rate": 1.6973684210526318e-05,
+      "loss": 0.9687,
+      "step": 129
+    },
+    {
+      "epoch": 0.07699141249629848,
+      "grad_norm": 16.245213640484007,
+      "learning_rate": 1.7105263157894737e-05,
+      "loss": 0.9135,
+      "step": 130
+    },
+    {
+      "epoch": 0.0775836541308854,
+      "grad_norm": 11.924195905537667,
+      "learning_rate": 1.723684210526316e-05,
+      "loss": 0.9306,
+      "step": 131
+    },
+    {
+      "epoch": 0.0781758957654723,
+      "grad_norm": 234.49611616589422,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 0.95,
+      "step": 132
+    },
+    {
+      "epoch": 0.07876813740005922,
+      "grad_norm": 5.235845225399804,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.9042,
+      "step": 133
+    },
+    {
+      "epoch": 0.07936037903464614,
+      "grad_norm": 7.032132390324782,
+      "learning_rate": 1.763157894736842e-05,
+      "loss": 0.958,
+      "step": 134
+    },
+    {
+      "epoch": 0.07995262066923305,
+      "grad_norm": 9.064372001479308,
+      "learning_rate": 1.7763157894736845e-05,
+      "loss": 0.8924,
+      "step": 135
+    },
+    {
+      "epoch": 0.08054486230381996,
+      "grad_norm": 6.104808834461892,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 0.9176,
+      "step": 136
+    },
+    {
+      "epoch": 0.08113710393840687,
+      "grad_norm": 8.886467300119726,
+      "learning_rate": 1.8026315789473685e-05,
+      "loss": 0.8794,
+      "step": 137
+    },
+    {
+      "epoch": 0.08172934557299379,
+      "grad_norm": 7.344989153314532,
+      "learning_rate": 1.8157894736842107e-05,
+      "loss": 0.9546,
+      "step": 138
+    },
+    {
+      "epoch": 0.08232158720758069,
+      "grad_norm": 3.97659060586214,
+      "learning_rate": 1.828947368421053e-05,
+      "loss": 0.9321,
+      "step": 139
+    },
+    {
+      "epoch": 0.08291382884216761,
+      "grad_norm": 11.103481278827003,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.9627,
+      "step": 140
+    },
+    {
+      "epoch": 0.08350607047675451,
+      "grad_norm": 4.6592325848802565,
+      "learning_rate": 1.8552631578947373e-05,
+      "loss": 0.9245,
+      "step": 141
+    },
+    {
+      "epoch": 0.08409831211134143,
+      "grad_norm": 4.7813770731736245,
+      "learning_rate": 1.868421052631579e-05,
+      "loss": 0.9103,
+      "step": 142
+    },
+    {
+      "epoch": 0.08469055374592833,
+      "grad_norm": 3.923276266087163,
+      "learning_rate": 1.8815789473684213e-05,
+      "loss": 0.9362,
+      "step": 143
+    },
+    {
+      "epoch": 0.08528279538051525,
+      "grad_norm": 5.389802023966097,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 0.8964,
+      "step": 144
+    },
+    {
+      "epoch": 0.08587503701510216,
+      "grad_norm": 3.416809050980661,
+      "learning_rate": 1.9078947368421056e-05,
+      "loss": 0.9409,
+      "step": 145
+    },
+    {
+      "epoch": 0.08646727864968907,
+      "grad_norm": 7.72766710731213,
+      "learning_rate": 1.9210526315789474e-05,
+      "loss": 0.9519,
+      "step": 146
+    },
+    {
+      "epoch": 0.08705952028427598,
+      "grad_norm": 4.149089365478041,
+      "learning_rate": 1.9342105263157896e-05,
+      "loss": 0.8684,
+      "step": 147
+    },
+    {
+      "epoch": 0.0876517619188629,
+      "grad_norm": 3.6666693738461795,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 0.9186,
+      "step": 148
+    },
+    {
+      "epoch": 0.0882440035534498,
+      "grad_norm": 4.7322097295129195,
+      "learning_rate": 1.960526315789474e-05,
+      "loss": 0.942,
+      "step": 149
+    },
+    {
+      "epoch": 0.08883624518803672,
+      "grad_norm": 4.525436493311952,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 0.9201,
+      "step": 150
+    },
+    {
+      "epoch": 0.08942848682262364,
+      "grad_norm": 4.785414884022584,
+      "learning_rate": 1.986842105263158e-05,
+      "loss": 0.9087,
+      "step": 151
+    },
+    {
+      "epoch": 0.09002072845721054,
+      "grad_norm": 6.846600500963897,
+      "learning_rate": 2e-05,
+      "loss": 0.9397,
+      "step": 152
+    },
+    {
+      "epoch": 0.09061297009179746,
+      "grad_norm": 10.702916565838894,
+      "learning_rate": 1.9999997954718838e-05,
+      "loss": 0.9056,
+      "step": 153
+    },
+    {
+      "epoch": 0.09120521172638436,
+      "grad_norm": 5.25139552872273,
+      "learning_rate": 1.9999991818876183e-05,
+      "loss": 0.9303,
+      "step": 154
+    },
+    {
+      "epoch": 0.09179745336097128,
+      "grad_norm": 3.289991163103779,
+      "learning_rate": 1.999998159247455e-05,
+      "loss": 0.8763,
+      "step": 155
+    },
+    {
+      "epoch": 0.09238969499555819,
+      "grad_norm": 4.456341363473166,
+      "learning_rate": 1.9999967275518118e-05,
+      "loss": 0.8785,
+      "step": 156
+    },
+    {
+      "epoch": 0.0929819366301451,
+      "grad_norm": 3.149442922056867,
+      "learning_rate": 1.9999948868012743e-05,
+      "loss": 0.9021,
+      "step": 157
+    },
+    {
+      "epoch": 0.09357417826473201,
+      "grad_norm": 2.4588136993301704,
+      "learning_rate": 1.999992636996596e-05,
+      "loss": 0.8785,
+      "step": 158
+    },
+    {
+      "epoch": 0.09416641989931893,
+      "grad_norm": 4.5661756454164255,
+      "learning_rate": 1.9999899781386968e-05,
+      "loss": 0.8916,
+      "step": 159
+    },
+    {
+      "epoch": 0.09475866153390583,
+      "grad_norm": 2.30357391498451,
+      "learning_rate": 1.9999869102286638e-05,
+      "loss": 0.9262,
+      "step": 160
+    },
+    {
+      "epoch": 0.09535090316849275,
+      "grad_norm": 2.6065419794469467,
+      "learning_rate": 1.9999834332677534e-05,
+      "loss": 0.9292,
+      "step": 161
+    },
+    {
+      "epoch": 0.09594314480307965,
+      "grad_norm": 2.8765449701366164,
+      "learning_rate": 1.9999795472573865e-05,
+      "loss": 0.9565,
+      "step": 162
+    },
+    {
+      "epoch": 0.09653538643766657,
+      "grad_norm": 2.049749661177928,
+      "learning_rate": 1.9999752521991535e-05,
+      "loss": 0.934,
+      "step": 163
+    },
+    {
+      "epoch": 0.09712762807225347,
+      "grad_norm": 2.143243330636517,
+      "learning_rate": 1.9999705480948107e-05,
+      "loss": 0.9237,
+      "step": 164
+    },
+    {
+      "epoch": 0.09771986970684039,
+      "grad_norm": 4.902266517773235,
+      "learning_rate": 1.9999654349462828e-05,
+      "loss": 0.9111,
+      "step": 165
+    },
+    {
+      "epoch": 0.0983121113414273,
+      "grad_norm": 3.7313193755804397,
+      "learning_rate": 1.9999599127556614e-05,
+      "loss": 0.8766,
+      "step": 166
+    },
+    {
+      "epoch": 0.09890435297601421,
+      "grad_norm": 2.121600886528297,
+      "learning_rate": 1.9999539815252048e-05,
+      "loss": 0.8929,
+      "step": 167
+    },
+    {
+      "epoch": 0.09949659461060112,
+      "grad_norm": 3.3403412194087374,
+      "learning_rate": 1.99994764125734e-05,
+      "loss": 0.9495,
+      "step": 168
+    },
+    {
+      "epoch": 0.10008883624518804,
+      "grad_norm": 2.468640773279232,
+      "learning_rate": 1.99994089195466e-05,
+      "loss": 0.926,
+      "step": 169
+    },
+    {
+      "epoch": 0.10068107787977496,
+      "grad_norm": 6.834707950691288,
+      "learning_rate": 1.9999337336199257e-05,
+      "loss": 0.8343,
+      "step": 170
+    },
+    {
+      "epoch": 0.10127331951436186,
+      "grad_norm": 8.374475093505808,
+      "learning_rate": 1.9999261662560657e-05,
+      "loss": 0.8807,
+      "step": 171
+    },
+    {
+      "epoch": 0.10186556114894878,
+      "grad_norm": 3.7723195507143905,
+      "learning_rate": 1.999918189866175e-05,
+      "loss": 0.9351,
+      "step": 172
+    },
+    {
+      "epoch": 0.10245780278353568,
+      "grad_norm": 5.716189938070129,
+      "learning_rate": 1.9999098044535164e-05,
+      "loss": 0.9152,
+      "step": 173
+    },
+    {
+      "epoch": 0.1030500444181226,
+      "grad_norm": 9.612635045762634,
+      "learning_rate": 1.9999010100215202e-05,
+      "loss": 0.8843,
+      "step": 174
+    },
+    {
+      "epoch": 0.1036422860527095,
+      "grad_norm": 4.703911264627151,
+      "learning_rate": 1.999891806573784e-05,
+      "loss": 0.9066,
+      "step": 175
+    },
+    {
+      "epoch": 0.10423452768729642,
+      "grad_norm": 4.8130705468771335,
+      "learning_rate": 1.9998821941140716e-05,
+      "loss": 0.9073,
+      "step": 176
+    },
+    {
+      "epoch": 0.10482676932188333,
+      "grad_norm": 4.204520297054476,
+      "learning_rate": 1.999872172646316e-05,
+      "loss": 0.8713,
+      "step": 177
+    },
+    {
+      "epoch": 0.10541901095647024,
+      "grad_norm": 9.243056732431207,
+      "learning_rate": 1.9998617421746166e-05,
+      "loss": 0.899,
+      "step": 178
+    },
+    {
+      "epoch": 0.10601125259105715,
+      "grad_norm": 7.819090907308756,
+      "learning_rate": 1.9998509027032392e-05,
+      "loss": 0.9286,
+      "step": 179
+    },
+    {
+      "epoch": 0.10660349422564407,
+      "grad_norm": 3.3915991356633426,
+      "learning_rate": 1.9998396542366188e-05,
+      "loss": 0.9553,
+      "step": 180
+    },
+    {
+      "epoch": 0.10719573586023097,
+      "grad_norm": 6.875890351613609,
+      "learning_rate": 1.9998279967793558e-05,
+      "loss": 0.9107,
+      "step": 181
+    },
+    {
+      "epoch": 0.10778797749481789,
+      "grad_norm": 18.641411497663324,
+      "learning_rate": 1.9998159303362193e-05,
+      "loss": 0.9007,
+      "step": 182
+    },
+    {
+      "epoch": 0.10838021912940479,
+      "grad_norm": 2.8778263671228794,
+      "learning_rate": 1.9998034549121445e-05,
+      "loss": 0.9288,
+      "step": 183
+    },
+    {
+      "epoch": 0.10897246076399171,
+      "grad_norm": 7.460744753488814,
+      "learning_rate": 1.9997905705122352e-05,
+      "loss": 0.9221,
+      "step": 184
+    },
+    {
+      "epoch": 0.10956470239857861,
+      "grad_norm": 38.271685018077065,
+      "learning_rate": 1.9997772771417615e-05,
+      "loss": 0.8964,
+      "step": 185
+    },
+    {
+      "epoch": 0.11015694403316553,
+      "grad_norm": 5.046727664405057,
+      "learning_rate": 1.9997635748061615e-05,
+      "loss": 0.9444,
+      "step": 186
+    },
+    {
+      "epoch": 0.11074918566775244,
+      "grad_norm": 8.624881305642129,
+      "learning_rate": 1.99974946351104e-05,
+      "loss": 0.9131,
+      "step": 187
+    },
+    {
+      "epoch": 0.11134142730233935,
+      "grad_norm": 8.837284064529285,
+      "learning_rate": 1.999734943262169e-05,
+      "loss": 0.906,
+      "step": 188
+    },
+    {
+      "epoch": 0.11193366893692627,
+      "grad_norm": 10.041268687119777,
+      "learning_rate": 1.999720014065489e-05,
+      "loss": 0.901,
+      "step": 189
+    },
+    {
+      "epoch": 0.11252591057151318,
+      "grad_norm": 2.6502443902067765,
+      "learning_rate": 1.9997046759271055e-05,
+      "loss": 0.9126,
+      "step": 190
+    },
+    {
+      "epoch": 0.1131181522061001,
+      "grad_norm": 5.889543292461257,
+      "learning_rate": 1.999688928853294e-05,
+      "loss": 0.9007,
+      "step": 191
+    },
+    {
+      "epoch": 0.113710393840687,
+      "grad_norm": 3.5995531544601578,
+      "learning_rate": 1.999672772850495e-05,
+      "loss": 0.8766,
+      "step": 192
+    },
+    {
+      "epoch": 0.11430263547527392,
+      "grad_norm": 1.796533079320344,
+      "learning_rate": 1.9996562079253177e-05,
+      "loss": 0.9259,
+      "step": 193
+    },
+    {
+      "epoch": 0.11489487710986082,
+      "grad_norm": 3.417277613043803,
+      "learning_rate": 1.999639234084538e-05,
+      "loss": 0.8797,
+      "step": 194
+    },
+    {
+      "epoch": 0.11548711874444774,
+      "grad_norm": 4.546208064925828,
+      "learning_rate": 1.999621851335099e-05,
+      "loss": 0.9833,
+      "step": 195
+    },
+    {
+      "epoch": 0.11607936037903464,
+      "grad_norm": 1.8321385514505888,
+      "learning_rate": 1.9996040596841118e-05,
+      "loss": 0.8972,
+      "step": 196
+    },
+    {
+      "epoch": 0.11667160201362156,
+      "grad_norm": 4.278863948371289,
+      "learning_rate": 1.9995858591388532e-05,
+      "loss": 0.9177,
+      "step": 197
+    },
+    {
+      "epoch": 0.11726384364820847,
+      "grad_norm": 2.5684660110656488,
+      "learning_rate": 1.999567249706769e-05,
+      "loss": 0.9269,
+      "step": 198
+    },
+    {
+      "epoch": 0.11785608528279538,
+      "grad_norm": 2.6222054577372784,
+      "learning_rate": 1.9995482313954713e-05,
+      "loss": 0.9453,
+      "step": 199
+    },
+    {
+      "epoch": 0.11844832691738229,
+      "grad_norm": 2.4017263015983557,
+      "learning_rate": 1.9995288042127396e-05,
+      "loss": 0.9334,
+      "step": 200
+    },
+    {
+      "epoch": 0.1190405685519692,
+      "grad_norm": 3.4487920247999244,
+      "learning_rate": 1.999508968166521e-05,
+      "loss": 0.8751,
+      "step": 201
+    },
+    {
+      "epoch": 0.11963281018655611,
+      "grad_norm": 4.923653355915247,
+      "learning_rate": 1.999488723264929e-05,
+      "loss": 0.9157,
+      "step": 202
+    },
+    {
+      "epoch": 0.12022505182114303,
+      "grad_norm": 2.210102296979183,
+      "learning_rate": 1.9994680695162453e-05,
+      "loss": 0.9256,
+      "step": 203
+    },
+    {
+      "epoch": 0.12081729345572993,
+      "grad_norm": 3.8316888952990786,
+      "learning_rate": 1.999447006928918e-05,
+      "loss": 0.8524,
+      "step": 204
+    },
+    {
+      "epoch": 0.12140953509031685,
+      "grad_norm": 1.703827800390735,
+      "learning_rate": 1.999425535511564e-05,
+      "loss": 0.8918,
+      "step": 205
+    },
+    {
+      "epoch": 0.12200177672490375,
+      "grad_norm": 3.5473701576710623,
+      "learning_rate": 1.999403655272965e-05,
+      "loss": 0.8804,
+      "step": 206
+    },
+    {
+      "epoch": 0.12259401835949067,
+      "grad_norm": 2.2496153840955415,
+      "learning_rate": 1.999381366222072e-05,
+      "loss": 0.9551,
+      "step": 207
+    },
+    {
+      "epoch": 0.12318625999407759,
+      "grad_norm": 2.126988606794187,
+      "learning_rate": 1.999358668368002e-05,
+      "loss": 0.8835,
+      "step": 208
+    },
+    {
+      "epoch": 0.1237785016286645,
+      "grad_norm": 2.660694810075823,
+      "learning_rate": 1.9993355617200404e-05,
+      "loss": 0.8951,
+      "step": 209
+    },
+    {
+      "epoch": 0.12437074326325141,
+      "grad_norm": 4.602299174530589,
+      "learning_rate": 1.9993120462876385e-05,
+      "loss": 0.9475,
+      "step": 210
+    },
+    {
+      "epoch": 0.12496298489783832,
+      "grad_norm": 3.931184146620322,
+      "learning_rate": 1.9992881220804157e-05,
+      "loss": 0.8922,
+      "step": 211
+    },
+    {
+      "epoch": 0.12555522653242523,
+      "grad_norm": 2.6749433222616292,
+      "learning_rate": 1.9992637891081585e-05,
+      "loss": 0.9381,
+      "step": 212
+    },
+    {
+      "epoch": 0.12614746816701214,
+      "grad_norm": 3.4189957335214105,
+      "learning_rate": 1.9992390473808195e-05,
+      "loss": 0.8404,
+      "step": 213
+    },
+    {
+      "epoch": 0.12673970980159904,
+      "grad_norm": 2.1866401943466,
+      "learning_rate": 1.999213896908521e-05,
+      "loss": 0.8585,
+      "step": 214
+    },
+    {
+      "epoch": 0.12733195143618598,
+      "grad_norm": 1.6522576171627141,
+      "learning_rate": 1.9991883377015497e-05,
+      "loss": 0.8548,
+      "step": 215
+    },
+    {
+      "epoch": 0.12792419307077288,
+      "grad_norm": 2.2509610072884074,
+      "learning_rate": 1.9991623697703613e-05,
+      "loss": 0.9282,
+      "step": 216
+    },
+    {
+      "epoch": 0.12851643470535978,
+      "grad_norm": 4.028256239000191,
+      "learning_rate": 1.9991359931255782e-05,
+      "loss": 0.9042,
+      "step": 217
+    },
+    {
+      "epoch": 0.1291086763399467,
+      "grad_norm": 1.8685142977539413,
+      "learning_rate": 1.9991092077779895e-05,
+      "loss": 0.9028,
+      "step": 218
+    },
+    {
+      "epoch": 0.12970091797453362,
+      "grad_norm": 2.87573942157228,
+      "learning_rate": 1.9990820137385525e-05,
+      "loss": 0.8854,
+      "step": 219
+    },
+    {
+      "epoch": 0.13029315960912052,
+      "grad_norm": 2.4941325943289323,
+      "learning_rate": 1.9990544110183907e-05,
+      "loss": 0.9238,
+      "step": 220
+    },
+    {
+      "epoch": 0.13088540124370743,
+      "grad_norm": 3.4805013559739235,
+      "learning_rate": 1.999026399628795e-05,
+      "loss": 0.8405,
+      "step": 221
+    },
+    {
+      "epoch": 0.13147764287829433,
+      "grad_norm": 1.8670578876428667,
+      "learning_rate": 1.998997979581224e-05,
+      "loss": 0.8898,
+      "step": 222
+    },
+    {
+      "epoch": 0.13206988451288126,
+      "grad_norm": 2.2661788538769043,
+      "learning_rate": 1.9989691508873032e-05,
+      "loss": 0.9229,
+      "step": 223
+    },
+    {
+      "epoch": 0.13266212614746817,
+      "grad_norm": 2.499121551088041,
+      "learning_rate": 1.9989399135588246e-05,
+      "loss": 0.9035,
+      "step": 224
+    },
+    {
+      "epoch": 0.13325436778205507,
+      "grad_norm": 3.4358794596174334,
+      "learning_rate": 1.9989102676077484e-05,
+      "loss": 0.9143,
+      "step": 225
+    },
+    {
+      "epoch": 0.13384660941664198,
+      "grad_norm": 3.4214432168583353,
+      "learning_rate": 1.9988802130462017e-05,
+      "loss": 0.9384,
+      "step": 226
+    },
+    {
+      "epoch": 0.1344388510512289,
+      "grad_norm": 2.544662580599115,
+      "learning_rate": 1.9988497498864776e-05,
+      "loss": 0.8979,
+      "step": 227
+    },
+    {
+      "epoch": 0.1350310926858158,
+      "grad_norm": 2.087938927884032,
+      "learning_rate": 1.9988188781410377e-05,
+      "loss": 0.8918,
+      "step": 228
+    },
+    {
+      "epoch": 0.13562333432040272,
+      "grad_norm": 3.1166292965067925,
+      "learning_rate": 1.9987875978225107e-05,
+      "loss": 0.9169,
+      "step": 229
+    },
+    {
+      "epoch": 0.13621557595498965,
+      "grad_norm": 2.9358196092245383,
+      "learning_rate": 1.9987559089436917e-05,
+      "loss": 0.9254,
+      "step": 230
+    },
+    {
+      "epoch": 0.13680781758957655,
+      "grad_norm": 4.473321811652426,
+      "learning_rate": 1.9987238115175428e-05,
+      "loss": 0.9227,
+      "step": 231
+    },
+    {
+      "epoch": 0.13740005922416346,
+      "grad_norm": 3.988884973768667,
+      "learning_rate": 1.998691305557194e-05,
+      "loss": 0.8807,
+      "step": 232
+    },
+    {
+      "epoch": 0.13799230085875036,
+      "grad_norm": 3.5180020915988632,
+      "learning_rate": 1.9986583910759427e-05,
+      "loss": 0.9089,
+      "step": 233
+    },
+    {
+      "epoch": 0.1385845424933373,
+      "grad_norm": 4.311583066844405,
+      "learning_rate": 1.9986250680872515e-05,
+      "loss": 0.8793,
+      "step": 234
+    },
+    {
+      "epoch": 0.1391767841279242,
+      "grad_norm": 4.094646789970844,
+      "learning_rate": 1.9985913366047524e-05,
+      "loss": 0.8717,
+      "step": 235
+    },
+    {
+      "epoch": 0.1397690257625111,
+      "grad_norm": 5.37580338032932,
+      "learning_rate": 1.998557196642243e-05,
+      "loss": 0.9662,
+      "step": 236
+    },
+    {
+      "epoch": 0.140361267397098,
+      "grad_norm": 5.374652282579761,
+      "learning_rate": 1.9985226482136887e-05,
+      "loss": 0.9267,
+      "step": 237
+    },
+    {
+      "epoch": 0.14095350903168494,
+      "grad_norm": 8.012244025293068,
+      "learning_rate": 1.9984876913332215e-05,
+      "loss": 0.9403,
+      "step": 238
+    },
+    {
+      "epoch": 0.14154575066627184,
+      "grad_norm": 4.000814675128233,
+      "learning_rate": 1.998452326015141e-05,
+      "loss": 0.9192,
+      "step": 239
+    },
+    {
+      "epoch": 0.14213799230085875,
+      "grad_norm": 2.743822493310051,
+      "learning_rate": 1.9984165522739135e-05,
+      "loss": 0.876,
+      "step": 240
+    },
+    {
+      "epoch": 0.14273023393544565,
+      "grad_norm": 3.584538417944988,
+      "learning_rate": 1.9983803701241723e-05,
+      "loss": 0.9261,
+      "step": 241
+    },
+    {
+      "epoch": 0.14332247557003258,
+      "grad_norm": 7.9272167764769845,
+      "learning_rate": 1.998343779580718e-05,
+      "loss": 0.8728,
+      "step": 242
+    },
+    {
+      "epoch": 0.14391471720461949,
+      "grad_norm": 2.481390145790585,
+      "learning_rate": 1.9983067806585184e-05,
+      "loss": 0.8885,
+      "step": 243
+    },
+    {
+      "epoch": 0.1445069588392064,
+      "grad_norm": 6.220923433805533,
+      "learning_rate": 1.998269373372708e-05,
+      "loss": 0.8931,
+      "step": 244
+    },
+    {
+      "epoch": 0.1450992004737933,
+      "grad_norm": 4.045398395801386,
+      "learning_rate": 1.9982315577385885e-05,
+      "loss": 0.8867,
+      "step": 245
+    },
+    {
+      "epoch": 0.14569144210838023,
+      "grad_norm": 5.776802690732297,
+      "learning_rate": 1.9981933337716288e-05,
+      "loss": 0.8379,
+      "step": 246
+    },
+    {
+      "epoch": 0.14628368374296713,
+      "grad_norm": 14.36864250446338,
+      "learning_rate": 1.998154701487464e-05,
+      "loss": 0.9106,
+      "step": 247
+    },
+    {
+      "epoch": 0.14687592537755403,
+      "grad_norm": 3.967462127445808,
+      "learning_rate": 1.9981156609018977e-05,
+      "loss": 0.9349,
+      "step": 248
+    },
+    {
+      "epoch": 0.14746816701214097,
+      "grad_norm": 3.265424009929576,
+      "learning_rate": 1.998076212030899e-05,
+      "loss": 0.8946,
+      "step": 249
+    },
+    {
+      "epoch": 0.14806040864672787,
+      "grad_norm": 5.252463182631547,
+      "learning_rate": 1.9980363548906056e-05,
+      "loss": 0.9346,
+      "step": 250
+    },
+    {
+      "epoch": 0.14865265028131477,
+      "grad_norm": 3.913170557844579,
+      "learning_rate": 1.9979960894973202e-05,
+      "loss": 0.9643,
+      "step": 251
+    },
+    {
+      "epoch": 0.14924489191590168,
+      "grad_norm": 10.316912742558536,
+      "learning_rate": 1.9979554158675145e-05,
+      "loss": 0.8396,
+      "step": 252
+    },
+    {
+      "epoch": 0.1498371335504886,
+      "grad_norm": 3.808990535792799,
+      "learning_rate": 1.9979143340178258e-05,
+      "loss": 0.9148,
+      "step": 253
+    },
+    {
+      "epoch": 0.15042937518507551,
+      "grad_norm": 12.348960229120108,
+      "learning_rate": 1.997872843965059e-05,
+      "loss": 0.8809,
+      "step": 254
+    },
+    {
+      "epoch": 0.15102161681966242,
+      "grad_norm": 7.004456279640057,
+      "learning_rate": 1.997830945726186e-05,
+      "loss": 0.8835,
+      "step": 255
+    },
+    {
+      "epoch": 0.15161385845424932,
+      "grad_norm": 7.66590001684898,
+      "learning_rate": 1.9977886393183454e-05,
+      "loss": 0.8604,
+      "step": 256
+    },
+    {
+      "epoch": 0.15220610008883625,
+      "grad_norm": 3.272971996729421,
+      "learning_rate": 1.997745924758843e-05,
+      "loss": 0.9124,
+      "step": 257
+    },
+    {
+      "epoch": 0.15279834172342316,
+      "grad_norm": 3.163402751353713,
+      "learning_rate": 1.9977028020651516e-05,
+      "loss": 0.9089,
+      "step": 258
+    },
+    {
+      "epoch": 0.15339058335801006,
+      "grad_norm": 9.234252191086176,
+      "learning_rate": 1.9976592712549102e-05,
+      "loss": 0.9354,
+      "step": 259
+    },
+    {
+      "epoch": 0.15398282499259697,
+      "grad_norm": 3.471952759539684,
+      "learning_rate": 1.9976153323459262e-05,
+      "loss": 0.9084,
+      "step": 260
+    },
+    {
+      "epoch": 0.1545750666271839,
+      "grad_norm": 19.265326355686057,
+      "learning_rate": 1.9975709853561725e-05,
+      "loss": 0.8811,
+      "step": 261
+    },
+    {
+      "epoch": 0.1551673082617708,
+      "grad_norm": 25.404153390256994,
+      "learning_rate": 1.9975262303037896e-05,
+      "loss": 0.9463,
+      "step": 262
+    },
+    {
+      "epoch": 0.1557595498963577,
+      "grad_norm": 1.790796295391984,
+      "learning_rate": 1.997481067207085e-05,
+      "loss": 0.9076,
+      "step": 263
+    },
+    {
+      "epoch": 0.1563517915309446,
+      "grad_norm": 1.5429882740755128,
+      "learning_rate": 1.9974354960845326e-05,
+      "loss": 0.8787,
+      "step": 264
+    },
+    {
+      "epoch": 0.15694403316553154,
+      "grad_norm": 1.8706636553302713,
+      "learning_rate": 1.997389516954774e-05,
+      "loss": 0.9265,
+      "step": 265
+    },
+    {
+      "epoch": 0.15753627480011845,
+      "grad_norm": 2.730000619216694,
+      "learning_rate": 1.997343129836617e-05,
+      "loss": 0.9074,
+      "step": 266
+    },
+    {
+      "epoch": 0.15812851643470535,
+      "grad_norm": 6.628653976491072,
+      "learning_rate": 1.9972963347490366e-05,
+      "loss": 0.8885,
+      "step": 267
+    },
+    {
+      "epoch": 0.15872075806929228,
+      "grad_norm": 1.9178468971516744,
+      "learning_rate": 1.9972491317111745e-05,
+      "loss": 0.8867,
+      "step": 268
+    },
+    {
+      "epoch": 0.1593129997038792,
+      "grad_norm": 3.3558537963254267,
+      "learning_rate": 1.9972015207423396e-05,
+      "loss": 0.844,
+      "step": 269
+    },
+    {
+      "epoch": 0.1599052413384661,
+      "grad_norm": 4.932473991808671,
+      "learning_rate": 1.997153501862007e-05,
+      "loss": 0.9451,
+      "step": 270
+    },
+    {
+      "epoch": 0.160497482973053,
+      "grad_norm": 1.7319648821727869,
+      "learning_rate": 1.99710507508982e-05,
+      "loss": 0.8596,
+      "step": 271
+    },
+    {
+      "epoch": 0.16108972460763993,
+      "grad_norm": 3.1569769697720558,
+      "learning_rate": 1.9970562404455872e-05,
+      "loss": 0.89,
+      "step": 272
+    },
+    {
+      "epoch": 0.16168196624222683,
+      "grad_norm": 3.6061121751456753,
+      "learning_rate": 1.9970069979492846e-05,
+      "loss": 0.8865,
+      "step": 273
+    },
+    {
+      "epoch": 0.16227420787681374,
+      "grad_norm": 19.08626640686028,
+      "learning_rate": 1.9969573476210558e-05,
+      "loss": 0.9106,
+      "step": 274
+    },
+    {
+      "epoch": 0.16286644951140064,
+      "grad_norm": 1.5734768192780313,
+      "learning_rate": 1.99690728948121e-05,
+      "loss": 0.9331,
+      "step": 275
+    },
+    {
+      "epoch": 0.16345869114598757,
+      "grad_norm": 1.8545986500276248,
+      "learning_rate": 1.996856823550224e-05,
+      "loss": 0.8472,
+      "step": 276
+    },
+    {
+      "epoch": 0.16405093278057448,
+      "grad_norm": 2.3229100358592563,
+      "learning_rate": 1.9968059498487415e-05,
+      "loss": 0.8914,
+      "step": 277
+    },
+    {
+      "epoch": 0.16464317441516138,
+      "grad_norm": 1.736892427592765,
+      "learning_rate": 1.996754668397572e-05,
+      "loss": 0.9359,
+      "step": 278
+    },
+    {
+      "epoch": 0.16523541604974828,
+      "grad_norm": 3.6486468336727444,
+      "learning_rate": 1.9967029792176932e-05,
+      "loss": 0.8875,
+      "step": 279
+    },
+    {
+      "epoch": 0.16582765768433522,
+      "grad_norm": 1.7408683315370481,
+      "learning_rate": 1.9966508823302484e-05,
+      "loss": 0.877,
+      "step": 280
+    },
+    {
+      "epoch": 0.16641989931892212,
+      "grad_norm": 1.8687430980001674,
+      "learning_rate": 1.9965983777565483e-05,
+      "loss": 0.8905,
+      "step": 281
+    },
+    {
+      "epoch": 0.16701214095350903,
+      "grad_norm": 3.904642504972808,
+      "learning_rate": 1.9965454655180704e-05,
+      "loss": 0.852,
+      "step": 282
+    },
+    {
+      "epoch": 0.16760438258809596,
+      "grad_norm": 1.9130632597634436,
+      "learning_rate": 1.9964921456364584e-05,
+      "loss": 0.9199,
+      "step": 283
+    },
+    {
+      "epoch": 0.16819662422268286,
+      "grad_norm": 2.675290531374258,
+      "learning_rate": 1.9964384181335237e-05,
+      "loss": 0.9072,
+      "step": 284
+    },
+    {
+      "epoch": 0.16878886585726977,
+      "grad_norm": 1.5847471748990105,
+      "learning_rate": 1.9963842830312434e-05,
+      "loss": 0.9185,
+      "step": 285
+    },
+    {
+      "epoch": 0.16938110749185667,
+      "grad_norm": 10.321315145183203,
+      "learning_rate": 1.996329740351762e-05,
+      "loss": 0.8816,
+      "step": 286
+    },
+    {
+      "epoch": 0.1699733491264436,
+      "grad_norm": 10.276955241455104,
+      "learning_rate": 1.9962747901173904e-05,
+      "loss": 0.9097,
+      "step": 287
+    },
+    {
+      "epoch": 0.1705655907610305,
+      "grad_norm": 2.3615243550709155,
+      "learning_rate": 1.9962194323506064e-05,
+      "loss": 0.928,
+      "step": 288
+    },
+    {
+      "epoch": 0.1711578323956174,
+      "grad_norm": 1.1933250899884889,
+      "learning_rate": 1.9961636670740546e-05,
+      "loss": 0.9304,
+      "step": 289
+    },
+    {
+      "epoch": 0.1717500740302043,
+      "grad_norm": 3.5837635071883396,
+      "learning_rate": 1.9961074943105457e-05,
+      "loss": 0.9264,
+      "step": 290
+    },
+    {
+      "epoch": 0.17234231566479125,
+      "grad_norm": 1.4231322142951308,
+      "learning_rate": 1.996050914083058e-05,
+      "loss": 0.915,
+      "step": 291
+    },
+    {
+      "epoch": 0.17293455729937815,
+      "grad_norm": 2.3576679543712724,
+      "learning_rate": 1.9959939264147355e-05,
+      "loss": 0.9305,
+      "step": 292
+    },
+    {
+      "epoch": 0.17352679893396505,
+      "grad_norm": 2.292913955194318,
+      "learning_rate": 1.99593653132889e-05,
+      "loss": 0.9098,
+      "step": 293
+    },
+    {
+      "epoch": 0.17411904056855196,
+      "grad_norm": 1.9398072904004944,
+      "learning_rate": 1.9958787288489983e-05,
+      "loss": 0.9107,
+      "step": 294
+    },
+    {
+      "epoch": 0.1747112822031389,
+      "grad_norm": 11.08115964739409,
+      "learning_rate": 1.9958205189987066e-05,
+      "loss": 0.8475,
+      "step": 295
+    },
+    {
+      "epoch": 0.1753035238377258,
+      "grad_norm": 2.6304317819640852,
+      "learning_rate": 1.9957619018018243e-05,
+      "loss": 0.8905,
+      "step": 296
+    },
+    {
+      "epoch": 0.1758957654723127,
+      "grad_norm": 3.3173060476207477,
+      "learning_rate": 1.99570287728233e-05,
+      "loss": 0.8585,
+      "step": 297
+    },
+    {
+      "epoch": 0.1764880071068996,
+      "grad_norm": 1.7335753531753704,
+      "learning_rate": 1.9956434454643675e-05,
+      "loss": 0.8835,
+      "step": 298
+    },
+    {
+      "epoch": 0.17708024874148653,
+      "grad_norm": 1.5781518168435433,
+      "learning_rate": 1.995583606372248e-05,
+      "loss": 0.8636,
+      "step": 299
+    },
+    {
+      "epoch": 0.17767249037607344,
+      "grad_norm": 3.186713245410372,
+      "learning_rate": 1.9955233600304496e-05,
+      "loss": 0.8857,
+      "step": 300
+    },
+    {
+      "epoch": 0.17826473201066034,
+      "grad_norm": 2.4742503690715254,
+      "learning_rate": 1.9954627064636157e-05,
+      "loss": 0.9306,
+      "step": 301
+    },
+    {
+      "epoch": 0.17885697364524727,
+      "grad_norm": 2.1188838286209513,
+      "learning_rate": 1.995401645696557e-05,
+      "loss": 0.8874,
+      "step": 302
+    },
+    {
+      "epoch": 0.17944921527983418,
+      "grad_norm": 1.979483491889314,
+      "learning_rate": 1.9953401777542517e-05,
+      "loss": 0.9408,
+      "step": 303
+    },
+    {
+      "epoch": 0.18004145691442108,
+      "grad_norm": 1.285049652875338,
+      "learning_rate": 1.9952783026618424e-05,
+      "loss": 0.8872,
+      "step": 304
+    },
+    {
+      "epoch": 0.180633698549008,
+      "grad_norm": 1.9180889335059634,
+      "learning_rate": 1.9952160204446404e-05,
+      "loss": 0.8832,
+      "step": 305
+    },
+    {
+      "epoch": 0.18122594018359492,
+      "grad_norm": 2.556643811478596,
+      "learning_rate": 1.995153331128122e-05,
+      "loss": 0.8946,
+      "step": 306
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 2.4449018469180723,
+      "learning_rate": 1.9950902347379316e-05,
+      "loss": 0.9229,
+      "step": 307
+    },
+    {
+      "epoch": 0.18241042345276873,
+      "grad_norm": 2.436648980404225,
+      "learning_rate": 1.9950267312998783e-05,
+      "loss": 0.9155,
+      "step": 308
+    },
+    {
+      "epoch": 0.18300266508735563,
+      "grad_norm": 2.1396571184843793,
+      "learning_rate": 1.9949628208399388e-05,
+      "loss": 0.9097,
+      "step": 309
+    },
+    {
+      "epoch": 0.18359490672194256,
+      "grad_norm": 1.8397633591444138,
+      "learning_rate": 1.994898503384256e-05,
+      "loss": 0.9461,
+      "step": 310
+    },
+    {
+      "epoch": 0.18418714835652947,
+      "grad_norm": 5.288654703997796,
+      "learning_rate": 1.9948337789591396e-05,
+      "loss": 0.9367,
+      "step": 311
+    },
+    {
+      "epoch": 0.18477938999111637,
+      "grad_norm": 1.8505284885364202,
+      "learning_rate": 1.9947686475910656e-05,
+      "loss": 0.8994,
+      "step": 312
+    },
+    {
+      "epoch": 0.18537163162570328,
+      "grad_norm": 1.655781456697678,
+      "learning_rate": 1.9947031093066758e-05,
+      "loss": 0.8772,
+      "step": 313
+    },
+    {
+      "epoch": 0.1859638732602902,
+      "grad_norm": 2.0037040325972835,
+      "learning_rate": 1.9946371641327794e-05,
+      "loss": 0.9276,
+      "step": 314
+    },
+    {
+      "epoch": 0.1865561148948771,
+      "grad_norm": 4.111557030910506,
+      "learning_rate": 1.994570812096352e-05,
+      "loss": 0.9274,
+      "step": 315
+    },
+    {
+      "epoch": 0.18714835652946402,
+      "grad_norm": 1.6261247430270642,
+      "learning_rate": 1.9945040532245352e-05,
+      "loss": 0.9014,
+      "step": 316
+    },
+    {
+      "epoch": 0.18774059816405092,
+      "grad_norm": 1.6162415339185936,
+      "learning_rate": 1.9944368875446363e-05,
+      "loss": 0.9511,
+      "step": 317
+    },
+    {
+      "epoch": 0.18833283979863785,
+      "grad_norm": 1.5148471582795817,
+      "learning_rate": 1.9943693150841312e-05,
+      "loss": 0.8605,
+      "step": 318
+    },
+    {
+      "epoch": 0.18892508143322476,
+      "grad_norm": 1.540594634489607,
+      "learning_rate": 1.99430133587066e-05,
+      "loss": 0.8838,
+      "step": 319
+    },
+    {
+      "epoch": 0.18951732306781166,
+      "grad_norm": 2.049032648950319,
+      "learning_rate": 1.9942329499320298e-05,
+      "loss": 0.9253,
+      "step": 320
+    },
+    {
+      "epoch": 0.1901095647023986,
+      "grad_norm": 1.977962711697837,
+      "learning_rate": 1.994164157296215e-05,
+      "loss": 0.921,
+      "step": 321
+    },
+    {
+      "epoch": 0.1907018063369855,
+      "grad_norm": 5.4655707773851265,
+      "learning_rate": 1.994094957991355e-05,
+      "loss": 0.8246,
+      "step": 322
+    },
+    {
+      "epoch": 0.1912940479715724,
+      "grad_norm": 1.264635984959813,
+      "learning_rate": 1.994025352045757e-05,
+      "loss": 0.8664,
+      "step": 323
+    },
+    {
+      "epoch": 0.1918862896061593,
+      "grad_norm": 2.1323498253950275,
+      "learning_rate": 1.9939553394878926e-05,
+      "loss": 0.9058,
+      "step": 324
+    },
+    {
+      "epoch": 0.19247853124074624,
+      "grad_norm": 1.9659140774542014,
+      "learning_rate": 1.9938849203464023e-05,
+      "loss": 0.8865,
+      "step": 325
+    },
+    {
+      "epoch": 0.19307077287533314,
+      "grad_norm": 1.9108336737757217,
+      "learning_rate": 1.99381409465009e-05,
+      "loss": 0.9114,
+      "step": 326
+    },
+    {
+      "epoch": 0.19366301450992005,
+      "grad_norm": 1.9649788148181904,
+      "learning_rate": 1.9937428624279284e-05,
+      "loss": 0.8786,
+      "step": 327
+    },
+    {
+      "epoch": 0.19425525614450695,
+      "grad_norm": 1.999999763126164,
+      "learning_rate": 1.9936712237090554e-05,
+      "loss": 0.8931,
+      "step": 328
+    },
+    {
+      "epoch": 0.19484749777909388,
+      "grad_norm": 2.6307075206759976,
+      "learning_rate": 1.993599178522775e-05,
+      "loss": 0.8924,
+      "step": 329
+    },
+    {
+      "epoch": 0.19543973941368079,
+      "grad_norm": 1.6672297479747191,
+      "learning_rate": 1.9935267268985577e-05,
+      "loss": 0.8818,
+      "step": 330
+    },
+    {
+      "epoch": 0.1960319810482677,
+      "grad_norm": 1.9888771189937435,
+      "learning_rate": 1.9934538688660403e-05,
+      "loss": 0.9191,
+      "step": 331
+    },
+    {
+      "epoch": 0.1966242226828546,
+      "grad_norm": 2.656316010840173,
+      "learning_rate": 1.9933806044550262e-05,
+      "loss": 0.8806,
+      "step": 332
+    },
+    {
+      "epoch": 0.19721646431744153,
+      "grad_norm": 4.286512255009421,
+      "learning_rate": 1.9933069336954842e-05,
+      "loss": 0.8969,
+      "step": 333
+    },
+    {
+      "epoch": 0.19780870595202843,
+      "grad_norm": 3.055765430366512,
+      "learning_rate": 1.99323285661755e-05,
+      "loss": 0.9115,
+      "step": 334
+    },
+    {
+      "epoch": 0.19840094758661533,
+      "grad_norm": 1.7964416584438196,
+      "learning_rate": 1.9931583732515252e-05,
+      "loss": 0.8734,
+      "step": 335
+    },
+    {
+      "epoch": 0.19899318922120224,
+      "grad_norm": 1.8519391859954,
+      "learning_rate": 1.993083483627878e-05,
+      "loss": 0.9265,
+      "step": 336
+    },
+    {
+      "epoch": 0.19958543085578917,
+      "grad_norm": 1.9971914645128772,
+      "learning_rate": 1.993008187777242e-05,
+      "loss": 0.8784,
+      "step": 337
+    },
+    {
+      "epoch": 0.20017767249037607,
+      "grad_norm": 2.4232990241295522,
+      "learning_rate": 1.9929324857304175e-05,
+      "loss": 0.8837,
+      "step": 338
+    },
+    {
+      "epoch": 0.20076991412496298,
+      "grad_norm": 1.4178062062472774,
+      "learning_rate": 1.9928563775183713e-05,
+      "loss": 0.8635,
+      "step": 339
+    },
+    {
+      "epoch": 0.2013621557595499,
+      "grad_norm": 4.756076969204727,
+      "learning_rate": 1.9927798631722353e-05,
+      "loss": 0.9008,
+      "step": 340
+    },
+    {
+      "epoch": 0.20195439739413681,
+      "grad_norm": 5.072255729219359,
+      "learning_rate": 1.992702942723309e-05,
+      "loss": 0.8908,
+      "step": 341
+    },
+    {
+      "epoch": 0.20254663902872372,
+      "grad_norm": 3.724264153241877,
+      "learning_rate": 1.9926256162030564e-05,
+      "loss": 0.893,
+      "step": 342
+    },
+    {
+      "epoch": 0.20313888066331062,
+      "grad_norm": 2.284558332966103,
+      "learning_rate": 1.992547883643109e-05,
+      "loss": 0.8823,
+      "step": 343
+    },
+    {
+      "epoch": 0.20373112229789755,
+      "grad_norm": 7.51141744653435,
+      "learning_rate": 1.9924697450752636e-05,
+      "loss": 0.8618,
+      "step": 344
+    },
+    {
+      "epoch": 0.20432336393248446,
+      "grad_norm": 2.7293707676417487,
+      "learning_rate": 1.9923912005314827e-05,
+      "loss": 0.8983,
+      "step": 345
+    },
+    {
+      "epoch": 0.20491560556707136,
+      "grad_norm": 3.2471120301938776,
+      "learning_rate": 1.9923122500438964e-05,
+      "loss": 0.9105,
+      "step": 346
+    },
+    {
+      "epoch": 0.20550784720165827,
+      "grad_norm": 6.07549147250831,
+      "learning_rate": 1.9922328936447992e-05,
+      "loss": 0.8483,
+      "step": 347
+    },
+    {
+      "epoch": 0.2061000888362452,
+      "grad_norm": 2.791357973212501,
+      "learning_rate": 1.9921531313666526e-05,
+      "loss": 0.8484,
+      "step": 348
+    },
+    {
+      "epoch": 0.2066923304708321,
+      "grad_norm": 2.076877360733566,
+      "learning_rate": 1.992072963242084e-05,
+      "loss": 0.9362,
+      "step": 349
+    },
+    {
+      "epoch": 0.207284572105419,
+      "grad_norm": 2.7096944494505024,
+      "learning_rate": 1.9919923893038863e-05,
+      "loss": 0.8252,
+      "step": 350
+    },
+    {
+      "epoch": 0.2078768137400059,
+      "grad_norm": 2.355203227286866,
+      "learning_rate": 1.991911409585019e-05,
+      "loss": 0.92,
+      "step": 351
+    },
+    {
+      "epoch": 0.20846905537459284,
+      "grad_norm": 3.436005203104444,
+      "learning_rate": 1.991830024118607e-05,
+      "loss": 0.8602,
+      "step": 352
+    },
+    {
+      "epoch": 0.20906129700917975,
+      "grad_norm": 2.3276079656629856,
+      "learning_rate": 1.991748232937942e-05,
+      "loss": 0.8665,
+      "step": 353
+    },
+    {
+      "epoch": 0.20965353864376665,
+      "grad_norm": 3.406633969476107,
+      "learning_rate": 1.991666036076481e-05,
+      "loss": 0.8774,
+      "step": 354
+    },
+    {
+      "epoch": 0.21024578027835356,
+      "grad_norm": 3.2519752361578043,
+      "learning_rate": 1.9915834335678476e-05,
+      "loss": 0.938,
+      "step": 355
+    },
+    {
+      "epoch": 0.2108380219129405,
+      "grad_norm": 1.5571411069585908,
+      "learning_rate": 1.99150042544583e-05,
+      "loss": 0.8665,
+      "step": 356
+    },
+    {
+      "epoch": 0.2114302635475274,
+      "grad_norm": 1.797561656093368,
+      "learning_rate": 1.9914170117443837e-05,
+      "loss": 0.8467,
+      "step": 357
+    },
+    {
+      "epoch": 0.2120225051821143,
+      "grad_norm": 2.173677269835551,
+      "learning_rate": 1.9913331924976295e-05,
+      "loss": 0.8809,
+      "step": 358
+    },
+    {
+      "epoch": 0.21261474681670123,
+      "grad_norm": 1.4857064069360264,
+      "learning_rate": 1.9912489677398543e-05,
+      "loss": 0.8524,
+      "step": 359
+    },
+    {
+      "epoch": 0.21320698845128813,
+      "grad_norm": 1.5451577112951609,
+      "learning_rate": 1.991164337505511e-05,
+      "loss": 0.8625,
+      "step": 360
+    },
+    {
+      "epoch": 0.21379923008587504,
+      "grad_norm": 4.152036575903902,
+      "learning_rate": 1.9910793018292168e-05,
+      "loss": 0.891,
+      "step": 361
+    },
+    {
+      "epoch": 0.21439147172046194,
+      "grad_norm": 1.7176492006424662,
+      "learning_rate": 1.990993860745758e-05,
+      "loss": 0.8393,
+      "step": 362
+    },
+    {
+      "epoch": 0.21498371335504887,
+      "grad_norm": 16.044944435431123,
+      "learning_rate": 1.9909080142900835e-05,
+      "loss": 0.8994,
+      "step": 363
+    },
+    {
+      "epoch": 0.21557595498963578,
+      "grad_norm": 3.070726746879335,
+      "learning_rate": 1.99082176249731e-05,
+      "loss": 0.9275,
+      "step": 364
+    },
+    {
+      "epoch": 0.21616819662422268,
+      "grad_norm": 11.949735878042022,
+      "learning_rate": 1.9907351054027188e-05,
+      "loss": 0.8794,
+      "step": 365
+    },
+    {
+      "epoch": 0.21676043825880958,
+      "grad_norm": 4.743677400565406,
+      "learning_rate": 1.9906480430417575e-05,
+      "loss": 0.8665,
+      "step": 366
+    },
+    {
+      "epoch": 0.21735267989339652,
+      "grad_norm": 6.192596337832938,
+      "learning_rate": 1.99056057545004e-05,
+      "loss": 0.8838,
+      "step": 367
+    },
+    {
+      "epoch": 0.21794492152798342,
+      "grad_norm": 2.1298876915873426,
+      "learning_rate": 1.9904727026633453e-05,
+      "loss": 0.8655,
+      "step": 368
+    },
+    {
+      "epoch": 0.21853716316257032,
+      "grad_norm": 2.872112183326634,
+      "learning_rate": 1.9903844247176182e-05,
+      "loss": 0.8857,
+      "step": 369
+    },
+    {
+      "epoch": 0.21912940479715723,
+      "grad_norm": 2.116013503348104,
+      "learning_rate": 1.9902957416489693e-05,
+      "loss": 0.8675,
+      "step": 370
+    },
+    {
+      "epoch": 0.21972164643174416,
+      "grad_norm": 1.9650140489362669,
+      "learning_rate": 1.990206653493675e-05,
+      "loss": 0.8901,
+      "step": 371
+    },
+    {
+      "epoch": 0.22031388806633107,
+      "grad_norm": 2.1533395141766847,
+      "learning_rate": 1.9901171602881778e-05,
+      "loss": 0.8268,
+      "step": 372
+    },
+    {
+      "epoch": 0.22090612970091797,
+      "grad_norm": 1.8679750441425857,
+      "learning_rate": 1.990027262069085e-05,
+      "loss": 0.8956,
+      "step": 373
+    },
+    {
+      "epoch": 0.22149837133550487,
+      "grad_norm": 1.876927910343411,
+      "learning_rate": 1.9899369588731697e-05,
+      "loss": 0.915,
+      "step": 374
+    },
+    {
+      "epoch": 0.2220906129700918,
+      "grad_norm": 1.966063586945644,
+      "learning_rate": 1.9898462507373713e-05,
+      "loss": 0.85,
+      "step": 375
+    },
+    {
+      "epoch": 0.2226828546046787,
+      "grad_norm": 5.175814315545093,
+      "learning_rate": 1.9897551376987948e-05,
+      "loss": 0.8877,
+      "step": 376
+    },
+    {
+      "epoch": 0.2232750962392656,
+      "grad_norm": 3.3647238312551293,
+      "learning_rate": 1.9896636197947104e-05,
+      "loss": 0.9013,
+      "step": 377
+    },
+    {
+      "epoch": 0.22386733787385255,
+      "grad_norm": 3.0764967820975113,
+      "learning_rate": 1.9895716970625544e-05,
+      "loss": 0.9351,
+      "step": 378
+    },
+    {
+      "epoch": 0.22445957950843945,
+      "grad_norm": 2.601494620918347,
+      "learning_rate": 1.9894793695399276e-05,
+      "loss": 0.8853,
+      "step": 379
+    },
+    {
+      "epoch": 0.22505182114302635,
+      "grad_norm": 2.4828022845602735,
+      "learning_rate": 1.9893866372645975e-05,
+      "loss": 0.8749,
+      "step": 380
+    },
+    {
+      "epoch": 0.22564406277761326,
+      "grad_norm": 2.0904211550387846,
+      "learning_rate": 1.989293500274497e-05,
+      "loss": 0.8613,
+      "step": 381
+    },
+    {
+      "epoch": 0.2262363044122002,
+      "grad_norm": 2.594547419908662,
+      "learning_rate": 1.989199958607724e-05,
+      "loss": 0.8853,
+      "step": 382
+    },
+    {
+      "epoch": 0.2268285460467871,
+      "grad_norm": 3.3824116389291046,
+      "learning_rate": 1.9891060123025427e-05,
+      "loss": 0.8458,
+      "step": 383
+    },
+    {
+      "epoch": 0.227420787681374,
+      "grad_norm": 2.388543281852487,
+      "learning_rate": 1.9890116613973822e-05,
+      "loss": 0.8779,
+      "step": 384
+    },
+    {
+      "epoch": 0.2280130293159609,
+      "grad_norm": 2.3270199405136425,
+      "learning_rate": 1.9889169059308374e-05,
+      "loss": 0.9211,
+      "step": 385
+    },
+    {
+      "epoch": 0.22860527095054783,
+      "grad_norm": 3.4392018992585625,
+      "learning_rate": 1.9888217459416685e-05,
+      "loss": 0.8896,
+      "step": 386
+    },
+    {
+      "epoch": 0.22919751258513474,
+      "grad_norm": 2.9845110896198865,
+      "learning_rate": 1.9887261814688017e-05,
+      "loss": 0.8656,
+      "step": 387
+    },
+    {
+      "epoch": 0.22978975421972164,
+      "grad_norm": 1.7722551264363466,
+      "learning_rate": 1.9886302125513276e-05,
+      "loss": 0.9233,
+      "step": 388
+    },
+    {
+      "epoch": 0.23038199585430855,
+      "grad_norm": 2.3468863928571864,
+      "learning_rate": 1.9885338392285032e-05,
+      "loss": 0.9028,
+      "step": 389
+    },
+    {
+      "epoch": 0.23097423748889548,
+      "grad_norm": 2.2088822196102136,
+      "learning_rate": 1.9884370615397507e-05,
+      "loss": 0.876,
+      "step": 390
+    },
+    {
+      "epoch": 0.23156647912348238,
+      "grad_norm": 2.954411705632498,
+      "learning_rate": 1.9883398795246577e-05,
+      "loss": 0.9139,
+      "step": 391
+    },
+    {
+      "epoch": 0.2321587207580693,
+      "grad_norm": 2.2449696353833963,
+      "learning_rate": 1.9882422932229765e-05,
+      "loss": 0.8882,
+      "step": 392
+    },
+    {
+      "epoch": 0.2327509623926562,
+      "grad_norm": 1.911476833050929,
+      "learning_rate": 1.988144302674626e-05,
+      "loss": 0.8947,
+      "step": 393
+    },
+    {
+      "epoch": 0.23334320402724312,
+      "grad_norm": 2.0157438479335625,
+      "learning_rate": 1.9880459079196898e-05,
+      "loss": 0.8994,
+      "step": 394
+    },
+    {
+      "epoch": 0.23393544566183003,
+      "grad_norm": 2.2549219273884655,
+      "learning_rate": 1.9879471089984168e-05,
+      "loss": 0.8613,
+      "step": 395
+    },
+    {
+      "epoch": 0.23452768729641693,
+      "grad_norm": 2.369416456819026,
+      "learning_rate": 1.9878479059512212e-05,
+      "loss": 0.8987,
+      "step": 396
+    },
+    {
+      "epoch": 0.23511992893100386,
+      "grad_norm": 2.065708029987891,
+      "learning_rate": 1.9877482988186825e-05,
+      "loss": 0.8903,
+      "step": 397
+    },
+    {
+      "epoch": 0.23571217056559077,
+      "grad_norm": 1.7370389947259615,
+      "learning_rate": 1.987648287641546e-05,
+      "loss": 0.9043,
+      "step": 398
+    },
+    {
+      "epoch": 0.23630441220017767,
+      "grad_norm": 3.4678559229085546,
+      "learning_rate": 1.987547872460722e-05,
+      "loss": 0.867,
+      "step": 399
+    },
+    {
+      "epoch": 0.23689665383476458,
+      "grad_norm": 2.1054707516143694,
+      "learning_rate": 1.987447053317285e-05,
+      "loss": 0.863,
+      "step": 400
+    },
+    {
+      "epoch": 0.2374888954693515,
+      "grad_norm": 2.820738998283495,
+      "learning_rate": 1.9873458302524767e-05,
+      "loss": 0.9357,
+      "step": 401
+    },
+    {
+      "epoch": 0.2380811371039384,
+      "grad_norm": 4.024264238834477,
+      "learning_rate": 1.9872442033077027e-05,
+      "loss": 0.8398,
+      "step": 402
+    },
+    {
+      "epoch": 0.23867337873852532,
+      "grad_norm": 2.7872478316820755,
+      "learning_rate": 1.9871421725245342e-05,
+      "loss": 0.8621,
+      "step": 403
+    },
+    {
+      "epoch": 0.23926562037311222,
+      "grad_norm": 2.641334532671031,
+      "learning_rate": 1.9870397379447074e-05,
+      "loss": 0.9111,
+      "step": 404
+    },
+    {
+      "epoch": 0.23985786200769915,
+      "grad_norm": 11.371926019038037,
+      "learning_rate": 1.9869368996101238e-05,
+      "loss": 0.863,
+      "step": 405
+    },
+    {
+      "epoch": 0.24045010364228606,
+      "grad_norm": 2.7635300627133925,
+      "learning_rate": 1.98683365756285e-05,
+      "loss": 0.866,
+      "step": 406
+    },
+    {
+      "epoch": 0.24104234527687296,
+      "grad_norm": 2.7794994837778084,
+      "learning_rate": 1.986730011845118e-05,
+      "loss": 0.8632,
+      "step": 407
+    },
+    {
+      "epoch": 0.24163458691145986,
+      "grad_norm": 3.08645053943537,
+      "learning_rate": 1.9866259624993246e-05,
+      "loss": 0.9136,
+      "step": 408
+    },
+    {
+      "epoch": 0.2422268285460468,
+      "grad_norm": 2.6869600181928255,
+      "learning_rate": 1.9865215095680322e-05,
+      "loss": 0.8791,
+      "step": 409
+    },
+    {
+      "epoch": 0.2428190701806337,
+      "grad_norm": 2.772950886041293,
+      "learning_rate": 1.986416653093967e-05,
+      "loss": 0.9567,
+      "step": 410
+    },
+    {
+      "epoch": 0.2434113118152206,
+      "grad_norm": 4.210601184028655,
+      "learning_rate": 1.986311393120022e-05,
+      "loss": 0.8685,
+      "step": 411
+    },
+    {
+      "epoch": 0.2440035534498075,
+      "grad_norm": 9.174595783532942,
+      "learning_rate": 1.9862057296892546e-05,
+      "loss": 0.888,
+      "step": 412
+    },
+    {
+      "epoch": 0.24459579508439444,
+      "grad_norm": 4.247915627932793,
+      "learning_rate": 1.9860996628448866e-05,
+      "loss": 0.8967,
+      "step": 413
+    },
+    {
+      "epoch": 0.24518803671898134,
+      "grad_norm": 2.835218121637848,
+      "learning_rate": 1.985993192630305e-05,
+      "loss": 0.8579,
+      "step": 414
+    },
+    {
+      "epoch": 0.24578027835356825,
+      "grad_norm": 2.9565335495075638,
+      "learning_rate": 1.985886319089063e-05,
+      "loss": 0.9049,
+      "step": 415
+    },
+    {
+      "epoch": 0.24637251998815518,
+      "grad_norm": 3.778725444535079,
+      "learning_rate": 1.9857790422648774e-05,
+      "loss": 0.8764,
+      "step": 416
+    },
+    {
+      "epoch": 0.24696476162274208,
+      "grad_norm": 3.1572313052180685,
+      "learning_rate": 1.9856713622016305e-05,
+      "loss": 0.8345,
+      "step": 417
+    },
+    {
+      "epoch": 0.247557003257329,
+      "grad_norm": 2.24801735688883,
+      "learning_rate": 1.9855632789433695e-05,
+      "loss": 0.8645,
+      "step": 418
+    },
+    {
+      "epoch": 0.2481492448919159,
+      "grad_norm": 23.485871617876615,
+      "learning_rate": 1.985454792534306e-05,
+      "loss": 0.8602,
+      "step": 419
+    },
+    {
+      "epoch": 0.24874148652650283,
+      "grad_norm": 6.005824269857163,
+      "learning_rate": 1.9853459030188183e-05,
+      "loss": 0.8688,
+      "step": 420
+    },
+    {
+      "epoch": 0.24933372816108973,
+      "grad_norm": 2.1616583205115796,
+      "learning_rate": 1.985236610441447e-05,
+      "loss": 0.907,
+      "step": 421
+    },
+    {
+      "epoch": 0.24992596979567663,
+      "grad_norm": 2.7443618096985993,
+      "learning_rate": 1.9851269148468998e-05,
+      "loss": 0.8735,
+      "step": 422
+    },
+    {
+      "epoch": 0.25051821143026354,
+      "grad_norm": 1.9467808502303237,
+      "learning_rate": 1.9850168162800482e-05,
+      "loss": 0.88,
+      "step": 423
+    },
+    {
+      "epoch": 0.25111045306485047,
+      "grad_norm": 2.293568091632234,
+      "learning_rate": 1.9849063147859282e-05,
+      "loss": 0.8251,
+      "step": 424
+    },
+    {
+      "epoch": 0.25170269469943735,
+      "grad_norm": 6.45680121195122,
+      "learning_rate": 1.9847954104097416e-05,
+      "loss": 0.8671,
+      "step": 425
+    },
+    {
+      "epoch": 0.2522949363340243,
+      "grad_norm": 2.6574165173210362,
+      "learning_rate": 1.9846841031968545e-05,
+      "loss": 0.8692,
+      "step": 426
+    },
+    {
+      "epoch": 0.2528871779686112,
+      "grad_norm": 3.3705354945908645,
+      "learning_rate": 1.9845723931927975e-05,
+      "loss": 0.8723,
+      "step": 427
+    },
+    {
+      "epoch": 0.2534794196031981,
+      "grad_norm": 2.039139055549686,
+      "learning_rate": 1.9844602804432667e-05,
+      "loss": 0.8749,
+      "step": 428
+    },
+    {
+      "epoch": 0.254071661237785,
+      "grad_norm": 2.3388722575578482,
+      "learning_rate": 1.9843477649941223e-05,
+      "loss": 0.8881,
+      "step": 429
+    },
+    {
+      "epoch": 0.25466390287237195,
+      "grad_norm": 2.3766640178995257,
+      "learning_rate": 1.9842348468913895e-05,
+      "loss": 0.8715,
+      "step": 430
+    },
+    {
+      "epoch": 0.2552561445069588,
+      "grad_norm": 2.114898403529378,
+      "learning_rate": 1.9841215261812578e-05,
+      "loss": 0.8633,
+      "step": 431
+    },
+    {
+      "epoch": 0.25584838614154576,
+      "grad_norm": 3.219697021874039,
+      "learning_rate": 1.9840078029100826e-05,
+      "loss": 0.8636,
+      "step": 432
+    },
+    {
+      "epoch": 0.25644062777613263,
+      "grad_norm": 2.73094050276702,
+      "learning_rate": 1.9838936771243823e-05,
+      "loss": 0.8747,
+      "step": 433
+    },
+    {
+      "epoch": 0.25703286941071957,
+      "grad_norm": 3.2677036829236252,
+      "learning_rate": 1.983779148870841e-05,
+      "loss": 0.8409,
+      "step": 434
+    },
+    {
+      "epoch": 0.2576251110453065,
+      "grad_norm": 3.691432854617829,
+      "learning_rate": 1.9836642181963074e-05,
+      "loss": 0.9213,
+      "step": 435
+    },
+    {
+      "epoch": 0.2582173526798934,
+      "grad_norm": 1.913831109953419,
+      "learning_rate": 1.9835488851477943e-05,
+      "loss": 0.878,
+      "step": 436
+    },
+    {
+      "epoch": 0.2588095943144803,
+      "grad_norm": 1.362390320852231,
+      "learning_rate": 1.9834331497724795e-05,
+      "loss": 0.8654,
+      "step": 437
+    },
+    {
+      "epoch": 0.25940183594906724,
+      "grad_norm": 4.080901803953802,
+      "learning_rate": 1.983317012117705e-05,
+      "loss": 0.8479,
+      "step": 438
+    },
+    {
+      "epoch": 0.2599940775836541,
+      "grad_norm": 2.235992282079437,
+      "learning_rate": 1.983200472230979e-05,
+      "loss": 0.8802,
+      "step": 439
+    },
+    {
+      "epoch": 0.26058631921824105,
+      "grad_norm": 2.171461179516376,
+      "learning_rate": 1.983083530159971e-05,
+      "loss": 0.9045,
+      "step": 440
+    },
+    {
+      "epoch": 0.261178560852828,
+      "grad_norm": 2.7476770749878145,
+      "learning_rate": 1.9829661859525176e-05,
+      "loss": 0.9307,
+      "step": 441
+    },
+    {
+      "epoch": 0.26177080248741486,
+      "grad_norm": 4.118541834493762,
+      "learning_rate": 1.9828484396566197e-05,
+      "loss": 0.9237,
+      "step": 442
+    },
+    {
+      "epoch": 0.2623630441220018,
+      "grad_norm": 1.827944414998417,
+      "learning_rate": 1.982730291320442e-05,
+      "loss": 0.8942,
+      "step": 443
+    },
+    {
+      "epoch": 0.26295528575658866,
+      "grad_norm": 1.8056252338200751,
+      "learning_rate": 1.982611740992313e-05,
+      "loss": 0.8737,
+      "step": 444
+    },
+    {
+      "epoch": 0.2635475273911756,
+      "grad_norm": 2.1706398680744328,
+      "learning_rate": 1.982492788720727e-05,
+      "loss": 0.8847,
+      "step": 445
+    },
+    {
+      "epoch": 0.2641397690257625,
+      "grad_norm": 2.0469770330929933,
+      "learning_rate": 1.9823734345543422e-05,
+      "loss": 0.8588,
+      "step": 446
+    },
+    {
+      "epoch": 0.2647320106603494,
+      "grad_norm": 1.86522114555931,
+      "learning_rate": 1.9822536785419815e-05,
+      "loss": 0.8934,
+      "step": 447
+    },
+    {
+      "epoch": 0.26532425229493634,
+      "grad_norm": 1.5844917473664697,
+      "learning_rate": 1.982133520732631e-05,
+      "loss": 0.8978,
+      "step": 448
+    },
+    {
+      "epoch": 0.26591649392952327,
+      "grad_norm": 2.8109169891194394,
+      "learning_rate": 1.9820129611754428e-05,
+      "loss": 0.889,
+      "step": 449
+    },
+    {
+      "epoch": 0.26650873556411014,
+      "grad_norm": 2.176184810578095,
+      "learning_rate": 1.981891999919732e-05,
+      "loss": 0.8648,
+      "step": 450
+    },
+    {
+      "epoch": 0.2671009771986971,
+      "grad_norm": 2.343556853907917,
+      "learning_rate": 1.981770637014979e-05,
+      "loss": 0.8753,
+      "step": 451
+    },
+    {
+      "epoch": 0.26769321883328395,
+      "grad_norm": 2.961712486256478,
+      "learning_rate": 1.981648872510828e-05,
+      "loss": 0.8267,
+      "step": 452
+    },
+    {
+      "epoch": 0.2682854604678709,
+      "grad_norm": 1.4529926233743575,
+      "learning_rate": 1.981526706457087e-05,
+      "loss": 0.8387,
+      "step": 453
+    },
+    {
+      "epoch": 0.2688777021024578,
+      "grad_norm": 1.9505755733893613,
+      "learning_rate": 1.9814041389037292e-05,
+      "loss": 0.8853,
+      "step": 454
+    },
+    {
+      "epoch": 0.2694699437370447,
+      "grad_norm": 2.200797769974991,
+      "learning_rate": 1.981281169900892e-05,
+      "loss": 0.9057,
+      "step": 455
+    },
+    {
+      "epoch": 0.2700621853716316,
+      "grad_norm": 2.1870312653366137,
+      "learning_rate": 1.9811577994988755e-05,
+      "loss": 0.872,
+      "step": 456
+    },
+    {
+      "epoch": 0.27065442700621856,
+      "grad_norm": 1.2993955402450537,
+      "learning_rate": 1.9810340277481463e-05,
+      "loss": 0.8613,
+      "step": 457
+    },
+    {
+      "epoch": 0.27124666864080543,
+      "grad_norm": 2.425202483291421,
+      "learning_rate": 1.9809098546993333e-05,
+      "loss": 0.9174,
+      "step": 458
+    },
+    {
+      "epoch": 0.27183891027539236,
+      "grad_norm": 1.83353264512604,
+      "learning_rate": 1.9807852804032306e-05,
+      "loss": 0.8911,
+      "step": 459
+    },
+    {
+      "epoch": 0.2724311519099793,
+      "grad_norm": 1.8314326605476354,
+      "learning_rate": 1.980660304910796e-05,
+      "loss": 0.9341,
+      "step": 460
+    },
+    {
+      "epoch": 0.2730233935445662,
+      "grad_norm": 1.6973761426278493,
+      "learning_rate": 1.9805349282731513e-05,
+      "loss": 0.9077,
+      "step": 461
+    },
+    {
+      "epoch": 0.2736156351791531,
+      "grad_norm": 1.695076266085342,
+      "learning_rate": 1.9804091505415833e-05,
+      "loss": 0.8425,
+      "step": 462
+    },
+    {
+      "epoch": 0.27420787681374,
+      "grad_norm": 1.6422303345636073,
+      "learning_rate": 1.9802829717675413e-05,
+      "loss": 0.8885,
+      "step": 463
+    },
+    {
+      "epoch": 0.2748001184483269,
+      "grad_norm": 2.8627350079091456,
+      "learning_rate": 1.98015639200264e-05,
+      "loss": 0.8507,
+      "step": 464
+    },
+    {
+      "epoch": 0.27539236008291385,
+      "grad_norm": 1.3298065268134922,
+      "learning_rate": 1.980029411298657e-05,
+      "loss": 0.8834,
+      "step": 465
+    },
+    {
+      "epoch": 0.2759846017175007,
+      "grad_norm": 1.3923498378130503,
+      "learning_rate": 1.979902029707536e-05,
+      "loss": 0.8122,
+      "step": 466
+    },
+    {
+      "epoch": 0.27657684335208765,
+      "grad_norm": 2.558317417775508,
+      "learning_rate": 1.9797742472813815e-05,
+      "loss": 0.9,
+      "step": 467
+    },
+    {
+      "epoch": 0.2771690849866746,
+      "grad_norm": 2.184421600404926,
+      "learning_rate": 1.9796460640724646e-05,
+      "loss": 0.8803,
+      "step": 468
+    },
+    {
+      "epoch": 0.27776132662126146,
+      "grad_norm": 1.5965091803938036,
+      "learning_rate": 1.9795174801332195e-05,
+      "loss": 0.8966,
+      "step": 469
+    },
+    {
+      "epoch": 0.2783535682558484,
+      "grad_norm": 1.6260839104501157,
+      "learning_rate": 1.9793884955162442e-05,
+      "loss": 0.867,
+      "step": 470
+    },
+    {
+      "epoch": 0.27894580989043527,
+      "grad_norm": 1.7420396047568658,
+      "learning_rate": 1.9792591102743006e-05,
+      "loss": 0.8878,
+      "step": 471
+    },
+    {
+      "epoch": 0.2795380515250222,
+      "grad_norm": 1.306769898643035,
+      "learning_rate": 1.979129324460314e-05,
+      "loss": 0.8346,
+      "step": 472
+    },
+    {
+      "epoch": 0.28013029315960913,
+      "grad_norm": 2.108370805335531,
+      "learning_rate": 1.978999138127375e-05,
+      "loss": 0.8706,
+      "step": 473
+    },
+    {
+      "epoch": 0.280722534794196,
+      "grad_norm": 4.654961995545459,
+      "learning_rate": 1.9788685513287368e-05,
+      "loss": 0.8544,
+      "step": 474
+    },
+    {
+      "epoch": 0.28131477642878294,
+      "grad_norm": 5.160042089413392,
+      "learning_rate": 1.9787375641178162e-05,
+      "loss": 0.8469,
+      "step": 475
+    },
+    {
+      "epoch": 0.2819070180633699,
+      "grad_norm": 2.0855599172877337,
+      "learning_rate": 1.9786061765481954e-05,
+      "loss": 0.9022,
+      "step": 476
+    },
+    {
+      "epoch": 0.28249925969795675,
+      "grad_norm": 1.621103726662229,
+      "learning_rate": 1.9784743886736185e-05,
+      "loss": 0.8852,
+      "step": 477
+    },
+    {
+      "epoch": 0.2830915013325437,
+      "grad_norm": 1.959405094759062,
+      "learning_rate": 1.9783422005479942e-05,
+      "loss": 0.8392,
+      "step": 478
+    },
+    {
+      "epoch": 0.2836837429671306,
+      "grad_norm": 2.3454164384529523,
+      "learning_rate": 1.978209612225395e-05,
+      "loss": 0.8649,
+      "step": 479
+    },
+    {
+      "epoch": 0.2842759846017175,
+      "grad_norm": 2.3325577065069933,
+      "learning_rate": 1.9780766237600574e-05,
+      "loss": 0.8854,
+      "step": 480
+    },
+    {
+      "epoch": 0.2848682262363044,
+      "grad_norm": 2.937131824507951,
+      "learning_rate": 1.9779432352063806e-05,
+      "loss": 0.9174,
+      "step": 481
+    },
+    {
+      "epoch": 0.2854604678708913,
+      "grad_norm": 2.1825918189828744,
+      "learning_rate": 1.977809446618928e-05,
+      "loss": 0.8332,
+      "step": 482
+    },
+    {
+      "epoch": 0.28605270950547823,
+      "grad_norm": 2.6555766398144494,
+      "learning_rate": 1.9776752580524268e-05,
+      "loss": 0.8701,
+      "step": 483
+    },
+    {
+      "epoch": 0.28664495114006516,
+      "grad_norm": 2.1721573688476927,
+      "learning_rate": 1.9775406695617677e-05,
+      "loss": 0.9111,
+      "step": 484
+    },
+    {
+      "epoch": 0.28723719277465204,
+      "grad_norm": 2.605674968341765,
+      "learning_rate": 1.977405681202005e-05,
+      "loss": 0.9035,
+      "step": 485
+    },
+    {
+      "epoch": 0.28782943440923897,
+      "grad_norm": 1.8150442904058317,
+      "learning_rate": 1.977270293028357e-05,
+      "loss": 0.8809,
+      "step": 486
+    },
+    {
+      "epoch": 0.2884216760438259,
+      "grad_norm": 2.126220077493787,
+      "learning_rate": 1.977134505096204e-05,
+      "loss": 0.8752,
+      "step": 487
+    },
+    {
+      "epoch": 0.2890139176784128,
+      "grad_norm": 3.4772636452956593,
+      "learning_rate": 1.9769983174610918e-05,
+      "loss": 0.8583,
+      "step": 488
+    },
+    {
+      "epoch": 0.2896061593129997,
+      "grad_norm": 2.44750945614034,
+      "learning_rate": 1.9768617301787284e-05,
+      "loss": 0.9164,
+      "step": 489
+    },
+    {
+      "epoch": 0.2901984009475866,
+      "grad_norm": 2.0538731894956586,
+      "learning_rate": 1.9767247433049858e-05,
+      "loss": 0.8891,
+      "step": 490
+    },
+    {
+      "epoch": 0.2907906425821735,
+      "grad_norm": 2.315988950710287,
+      "learning_rate": 1.9765873568958996e-05,
+      "loss": 0.868,
+      "step": 491
+    },
+    {
+      "epoch": 0.29138288421676045,
+      "grad_norm": 2.6759070473500586,
+      "learning_rate": 1.9764495710076678e-05,
+      "loss": 0.8976,
+      "step": 492
+    },
+    {
+      "epoch": 0.29197512585134733,
+      "grad_norm": 2.8515729411143624,
+      "learning_rate": 1.9763113856966532e-05,
+      "loss": 0.8445,
+      "step": 493
+    },
+    {
+      "epoch": 0.29256736748593426,
+      "grad_norm": 5.706902531350699,
+      "learning_rate": 1.9761728010193812e-05,
+      "loss": 0.8302,
+      "step": 494
+    },
+    {
+      "epoch": 0.2931596091205212,
+      "grad_norm": 2.3398897793110813,
+      "learning_rate": 1.9760338170325405e-05,
+      "loss": 0.8355,
+      "step": 495
+    },
+    {
+      "epoch": 0.29375185075510807,
+      "grad_norm": 3.290716851896505,
+      "learning_rate": 1.975894433792984e-05,
+      "loss": 0.9306,
+      "step": 496
+    },
+    {
+      "epoch": 0.294344092389695,
+      "grad_norm": 2.140513045045442,
+      "learning_rate": 1.975754651357727e-05,
+      "loss": 0.8518,
+      "step": 497
+    },
+    {
+      "epoch": 0.29493633402428193,
+      "grad_norm": 2.1673411170485193,
+      "learning_rate": 1.9756144697839477e-05,
+      "loss": 0.8967,
+      "step": 498
+    },
+    {
+      "epoch": 0.2955285756588688,
+      "grad_norm": 3.750627708947347,
+      "learning_rate": 1.975473889128989e-05,
+      "loss": 0.8229,
+      "step": 499
+    },
+    {
+      "epoch": 0.29612081729345574,
+      "grad_norm": 2.3819533335097303,
+      "learning_rate": 1.9753329094503563e-05,
+      "loss": 0.8417,
+      "step": 500
+    },
+    {
+      "epoch": 0.2967130589280426,
+      "grad_norm": 2.1932086970836924,
+      "learning_rate": 1.9751915308057176e-05,
+      "loss": 0.8552,
+      "step": 501
+    },
+    {
+      "epoch": 0.29730530056262955,
+      "grad_norm": 3.717240525518049,
+      "learning_rate": 1.9750497532529053e-05,
+      "loss": 0.8384,
+      "step": 502
+    },
+    {
+      "epoch": 0.2978975421972165,
+      "grad_norm": 3.828126525639594,
+      "learning_rate": 1.9749075768499148e-05,
+      "loss": 0.8433,
+      "step": 503
+    },
+    {
+      "epoch": 0.29848978383180336,
+      "grad_norm": 2.6956236004584277,
+      "learning_rate": 1.974765001654903e-05,
+      "loss": 0.8441,
+      "step": 504
+    },
+    {
+      "epoch": 0.2990820254663903,
+      "grad_norm": 3.676685156782072,
+      "learning_rate": 1.974622027726192e-05,
+      "loss": 0.8943,
+      "step": 505
+    },
+    {
+      "epoch": 0.2996742671009772,
+      "grad_norm": 2.590931432467633,
+      "learning_rate": 1.9744786551222658e-05,
+      "loss": 0.8639,
+      "step": 506
+    },
+    {
+      "epoch": 0.3002665087355641,
+      "grad_norm": 3.9335673574606935,
+      "learning_rate": 1.9743348839017728e-05,
+      "loss": 0.8656,
+      "step": 507
+    },
+    {
+      "epoch": 0.30085875037015103,
+      "grad_norm": 2.2776114718789904,
+      "learning_rate": 1.974190714123522e-05,
+      "loss": 0.8451,
+      "step": 508
+    },
+    {
+      "epoch": 0.3014509920047379,
+      "grad_norm": 4.638920612322161,
+      "learning_rate": 1.974046145846488e-05,
+      "loss": 0.8153,
+      "step": 509
+    },
+    {
+      "epoch": 0.30204323363932484,
+      "grad_norm": 6.980939867299674,
+      "learning_rate": 1.9739011791298073e-05,
+      "loss": 0.8504,
+      "step": 510
+    },
+    {
+      "epoch": 0.30263547527391177,
+      "grad_norm": 4.408663359040869,
+      "learning_rate": 1.973755814032779e-05,
+      "loss": 0.8419,
+      "step": 511
+    },
+    {
+      "epoch": 0.30322771690849865,
+      "grad_norm": 3.0168512612679113,
+      "learning_rate": 1.9736100506148657e-05,
+      "loss": 0.8906,
+      "step": 512
+    },
+    {
+      "epoch": 0.3038199585430856,
+      "grad_norm": 3.541981364911518,
+      "learning_rate": 1.973463888935693e-05,
+      "loss": 0.8498,
+      "step": 513
+    },
+    {
+      "epoch": 0.3044122001776725,
+      "grad_norm": 2.628539633705203,
+      "learning_rate": 1.9733173290550494e-05,
+      "loss": 0.8722,
+      "step": 514
+    },
+    {
+      "epoch": 0.3050044418122594,
+      "grad_norm": 8.965610022562394,
+      "learning_rate": 1.973170371032886e-05,
+      "loss": 0.888,
+      "step": 515
+    },
+    {
+      "epoch": 0.3055966834468463,
+      "grad_norm": 3.339091960381498,
+      "learning_rate": 1.9730230149293167e-05,
+      "loss": 0.8096,
+      "step": 516
+    },
+    {
+      "epoch": 0.30618892508143325,
+      "grad_norm": 2.91895416655151,
+      "learning_rate": 1.9728752608046184e-05,
+      "loss": 0.8708,
+      "step": 517
+    },
+    {
+      "epoch": 0.3067811667160201,
+      "grad_norm": 2.8857357994218216,
+      "learning_rate": 1.9727271087192312e-05,
+      "loss": 0.8614,
+      "step": 518
+    },
+    {
+      "epoch": 0.30737340835060706,
+      "grad_norm": 3.8897082719903815,
+      "learning_rate": 1.9725785587337574e-05,
+      "loss": 0.8364,
+      "step": 519
+    },
+    {
+      "epoch": 0.30796564998519393,
+      "grad_norm": 3.0809821310404657,
+      "learning_rate": 1.9724296109089623e-05,
+      "loss": 0.8634,
+      "step": 520
+    },
+    {
+      "epoch": 0.30855789161978087,
+      "grad_norm": 4.31589497235874,
+      "learning_rate": 1.972280265305774e-05,
+      "loss": 0.8938,
+      "step": 521
+    },
+    {
+      "epoch": 0.3091501332543678,
+      "grad_norm": 3.4659891853357174,
+      "learning_rate": 1.9721305219852833e-05,
+      "loss": 0.8612,
+      "step": 522
+    },
+    {
+      "epoch": 0.3097423748889547,
+      "grad_norm": 3.1684707972088106,
+      "learning_rate": 1.9719803810087436e-05,
+      "loss": 0.8838,
+      "step": 523
+    },
+    {
+      "epoch": 0.3103346165235416,
+      "grad_norm": 5.701615286531985,
+      "learning_rate": 1.971829842437571e-05,
+      "loss": 0.8652,
+      "step": 524
+    },
+    {
+      "epoch": 0.31092685815812854,
+      "grad_norm": 3.4465521972170987,
+      "learning_rate": 1.971678906333344e-05,
+      "loss": 0.8845,
+      "step": 525
+    },
+    {
+      "epoch": 0.3115190997927154,
+      "grad_norm": 1.6217244261703954,
+      "learning_rate": 1.971527572757804e-05,
+      "loss": 0.8227,
+      "step": 526
+    },
+    {
+      "epoch": 0.31211134142730235,
+      "grad_norm": 2.3507099292364897,
+      "learning_rate": 1.9713758417728555e-05,
+      "loss": 0.863,
+      "step": 527
+    },
+    {
+      "epoch": 0.3127035830618892,
+      "grad_norm": 2.281502522601501,
+      "learning_rate": 1.971223713440564e-05,
+      "loss": 0.8567,
+      "step": 528
+    },
+    {
+      "epoch": 0.31329582469647616,
+      "grad_norm": 3.1146189997548976,
+      "learning_rate": 1.97107118782316e-05,
+      "loss": 0.8631,
+      "step": 529
+    },
+    {
+      "epoch": 0.3138880663310631,
+      "grad_norm": 1.9410780864006059,
+      "learning_rate": 1.970918264983034e-05,
+      "loss": 0.8465,
+      "step": 530
+    },
+    {
+      "epoch": 0.31448030796564996,
+      "grad_norm": 3.224337429475279,
+      "learning_rate": 1.97076494498274e-05,
+      "loss": 0.846,
+      "step": 531
+    },
+    {
+      "epoch": 0.3150725496002369,
+      "grad_norm": 6.086192190115328,
+      "learning_rate": 1.970611227884995e-05,
+      "loss": 0.8496,
+      "step": 532
+    },
+    {
+      "epoch": 0.3156647912348238,
+      "grad_norm": 2.644702902924804,
+      "learning_rate": 1.9704571137526775e-05,
+      "loss": 0.7961,
+      "step": 533
+    },
+    {
+      "epoch": 0.3162570328694107,
+      "grad_norm": 5.673305681652432,
+      "learning_rate": 1.9703026026488288e-05,
+      "loss": 0.8546,
+      "step": 534
+    },
+    {
+      "epoch": 0.31684927450399764,
+      "grad_norm": 2.5921802921198767,
+      "learning_rate": 1.9701476946366533e-05,
+      "loss": 0.8838,
+      "step": 535
+    },
+    {
+      "epoch": 0.31744151613858457,
+      "grad_norm": 2.008362256981687,
+      "learning_rate": 1.9699923897795165e-05,
+      "loss": 0.9067,
+      "step": 536
+    },
+    {
+      "epoch": 0.31803375777317144,
+      "grad_norm": 3.335473379056157,
+      "learning_rate": 1.969836688140947e-05,
+      "loss": 0.8315,
+      "step": 537
+    },
+    {
+      "epoch": 0.3186259994077584,
+      "grad_norm": 1.6713798386358392,
+      "learning_rate": 1.9696805897846353e-05,
+      "loss": 0.8996,
+      "step": 538
+    },
+    {
+      "epoch": 0.31921824104234525,
+      "grad_norm": 1.6247895343380683,
+      "learning_rate": 1.9695240947744345e-05,
+      "loss": 0.9091,
+      "step": 539
+    },
+    {
+      "epoch": 0.3198104826769322,
+      "grad_norm": 2.0514048927331685,
+      "learning_rate": 1.9693672031743604e-05,
+      "loss": 0.8807,
+      "step": 540
+    },
+    {
+      "epoch": 0.3204027243115191,
+      "grad_norm": 3.2346881141878923,
+      "learning_rate": 1.9692099150485897e-05,
+      "loss": 0.8744,
+      "step": 541
+    },
+    {
+      "epoch": 0.320994965946106,
+      "grad_norm": 1.993896753955855,
+      "learning_rate": 1.9690522304614624e-05,
+      "loss": 0.8935,
+      "step": 542
+    },
+    {
+      "epoch": 0.3215872075806929,
+      "grad_norm": 1.6718977284924557,
+      "learning_rate": 1.9688941494774807e-05,
+      "loss": 0.871,
+      "step": 543
+    },
+    {
+      "epoch": 0.32217944921527986,
+      "grad_norm": 1.635671027032527,
+      "learning_rate": 1.9687356721613084e-05,
+      "loss": 0.8915,
+      "step": 544
+    },
+    {
+      "epoch": 0.32277169084986673,
+      "grad_norm": 1.873327797284262,
+      "learning_rate": 1.968576798577771e-05,
+      "loss": 0.8994,
+      "step": 545
+    },
+    {
+      "epoch": 0.32336393248445366,
+      "grad_norm": 1.2751536161018124,
+      "learning_rate": 1.9684175287918576e-05,
+      "loss": 0.8471,
+      "step": 546
+    },
+    {
+      "epoch": 0.32395617411904054,
+      "grad_norm": 2.7974266868919426,
+      "learning_rate": 1.9682578628687183e-05,
+      "loss": 0.8911,
+      "step": 547
+    },
+    {
+      "epoch": 0.3245484157536275,
+      "grad_norm": 1.6428003910383486,
+      "learning_rate": 1.968097800873665e-05,
+      "loss": 0.877,
+      "step": 548
+    },
+    {
+      "epoch": 0.3251406573882144,
+      "grad_norm": 5.184321377796686,
+      "learning_rate": 1.9679373428721728e-05,
+      "loss": 0.8149,
+      "step": 549
+    },
+    {
+      "epoch": 0.3257328990228013,
+      "grad_norm": 3.034456419245674,
+      "learning_rate": 1.9677764889298775e-05,
+      "loss": 0.9059,
+      "step": 550
+    },
+    {
+      "epoch": 0.3263251406573882,
+      "grad_norm": 1.824534137714233,
+      "learning_rate": 1.9676152391125773e-05,
+      "loss": 0.8729,
+      "step": 551
+    },
+    {
+      "epoch": 0.32691738229197514,
+      "grad_norm": 2.2247918235652526,
+      "learning_rate": 1.9674535934862327e-05,
+      "loss": 0.899,
+      "step": 552
+    },
+    {
+      "epoch": 0.327509623926562,
+      "grad_norm": 2.0674047220831655,
+      "learning_rate": 1.967291552116966e-05,
+      "loss": 0.8418,
+      "step": 553
+    },
+    {
+      "epoch": 0.32810186556114895,
+      "grad_norm": 1.8961950219733992,
+      "learning_rate": 1.967129115071061e-05,
+      "loss": 0.8938,
+      "step": 554
+    },
+    {
+      "epoch": 0.3286941071957359,
+      "grad_norm": 1.5716138089131724,
+      "learning_rate": 1.9669662824149632e-05,
+      "loss": 0.8675,
+      "step": 555
+    },
+    {
+      "epoch": 0.32928634883032276,
+      "grad_norm": 2.505642914698843,
+      "learning_rate": 1.966803054215281e-05,
+      "loss": 0.8575,
+      "step": 556
+    },
+    {
+      "epoch": 0.3298785904649097,
+      "grad_norm": 3.1345316813906745,
+      "learning_rate": 1.966639430538784e-05,
+      "loss": 0.8786,
+      "step": 557
+    },
+    {
+      "epoch": 0.33047083209949657,
+      "grad_norm": 2.56591314502249,
+      "learning_rate": 1.966475411452403e-05,
+      "loss": 0.8637,
+      "step": 558
+    },
+    {
+      "epoch": 0.3310630737340835,
+      "grad_norm": 2.7664296698481032,
+      "learning_rate": 1.966310997023231e-05,
+      "loss": 0.9181,
+      "step": 559
+    },
+    {
+      "epoch": 0.33165531536867043,
+      "grad_norm": 1.5231814543200226,
+      "learning_rate": 1.966146187318523e-05,
+      "loss": 0.8748,
+      "step": 560
+    },
+    {
+      "epoch": 0.3322475570032573,
+      "grad_norm": 4.815465527837311,
+      "learning_rate": 1.9659809824056954e-05,
+      "loss": 0.8408,
+      "step": 561
+    },
+    {
+      "epoch": 0.33283979863784424,
+      "grad_norm": 1.6014447979733908,
+      "learning_rate": 1.9658153823523262e-05,
+      "loss": 0.8672,
+      "step": 562
+    },
+    {
+      "epoch": 0.3334320402724312,
+      "grad_norm": 2.2744955163709006,
+      "learning_rate": 1.9656493872261554e-05,
+      "loss": 0.9016,
+      "step": 563
+    },
+    {
+      "epoch": 0.33402428190701805,
+      "grad_norm": 1.781058566595245,
+      "learning_rate": 1.9654829970950838e-05,
+      "loss": 0.8764,
+      "step": 564
+    },
+    {
+      "epoch": 0.334616523541605,
+      "grad_norm": 1.7582008304848717,
+      "learning_rate": 1.9653162120271746e-05,
+      "loss": 0.877,
+      "step": 565
+    },
+    {
+      "epoch": 0.3352087651761919,
+      "grad_norm": 1.9319354513096203,
+      "learning_rate": 1.965149032090653e-05,
+      "loss": 0.8599,
+      "step": 566
+    },
+    {
+      "epoch": 0.3358010068107788,
+      "grad_norm": 1.6123157537046895,
+      "learning_rate": 1.9649814573539037e-05,
+      "loss": 0.8374,
+      "step": 567
+    },
+    {
+      "epoch": 0.3363932484453657,
+      "grad_norm": 2.605187020230856,
+      "learning_rate": 1.9648134878854747e-05,
+      "loss": 0.8699,
+      "step": 568
+    },
+    {
+      "epoch": 0.3369854900799526,
+      "grad_norm": 2.4840368350540016,
+      "learning_rate": 1.9646451237540756e-05,
+      "loss": 0.8727,
+      "step": 569
+    },
+    {
+      "epoch": 0.33757773171453953,
+      "grad_norm": 3.6970475679434465,
+      "learning_rate": 1.9644763650285758e-05,
+      "loss": 0.8354,
+      "step": 570
+    },
+    {
+      "epoch": 0.33816997334912646,
+      "grad_norm": 1.804745820267029,
+      "learning_rate": 1.964307211778008e-05,
+      "loss": 0.8231,
+      "step": 571
+    },
+    {
+      "epoch": 0.33876221498371334,
+      "grad_norm": 1.7814414820272138,
+      "learning_rate": 1.9641376640715646e-05,
+      "loss": 0.8829,
+      "step": 572
+    },
+    {
+      "epoch": 0.33935445661830027,
+      "grad_norm": 2.246104894860363,
+      "learning_rate": 1.963967721978601e-05,
+      "loss": 0.8279,
+      "step": 573
+    },
+    {
+      "epoch": 0.3399466982528872,
+      "grad_norm": 1.5276776993277508,
+      "learning_rate": 1.963797385568632e-05,
+      "loss": 0.8421,
+      "step": 574
+    },
+    {
+      "epoch": 0.3405389398874741,
+      "grad_norm": 3.984819253438696,
+      "learning_rate": 1.9636266549113358e-05,
+      "loss": 0.8715,
+      "step": 575
+    },
+    {
+      "epoch": 0.341131181522061,
+      "grad_norm": 1.6707727313470635,
+      "learning_rate": 1.96345553007655e-05,
+      "loss": 0.868,
+      "step": 576
+    },
+    {
+      "epoch": 0.3417234231566479,
+      "grad_norm": 1.8434045564594856,
+      "learning_rate": 1.9632840111342747e-05,
+      "loss": 0.8946,
+      "step": 577
+    },
+    {
+      "epoch": 0.3423156647912348,
+      "grad_norm": 2.5166196457952386,
+      "learning_rate": 1.9631120981546713e-05,
+      "loss": 0.8169,
+      "step": 578
+    },
+    {
+      "epoch": 0.34290790642582175,
+      "grad_norm": 1.7256723725893435,
+      "learning_rate": 1.962939791208061e-05,
+      "loss": 0.8983,
+      "step": 579
+    },
+    {
+      "epoch": 0.3435001480604086,
+      "grad_norm": 1.537497453982475,
+      "learning_rate": 1.9627670903649273e-05,
+      "loss": 0.8535,
+      "step": 580
+    },
+    {
+      "epoch": 0.34409238969499556,
+      "grad_norm": 2.0982710201843346,
+      "learning_rate": 1.9625939956959146e-05,
+      "loss": 0.8352,
+      "step": 581
+    },
+    {
+      "epoch": 0.3446846313295825,
+      "grad_norm": 1.6615651140303191,
+      "learning_rate": 1.9624205072718285e-05,
+      "loss": 0.8424,
+      "step": 582
+    },
+    {
+      "epoch": 0.34527687296416937,
+      "grad_norm": 2.2732967556610144,
+      "learning_rate": 1.9622466251636352e-05,
+      "loss": 0.8548,
+      "step": 583
+    },
+    {
+      "epoch": 0.3458691145987563,
+      "grad_norm": 1.9163123664346449,
+      "learning_rate": 1.9620723494424627e-05,
+      "loss": 0.8903,
+      "step": 584
+    },
+    {
+      "epoch": 0.34646135623334323,
+      "grad_norm": 1.8941032464371603,
+      "learning_rate": 1.961897680179599e-05,
+      "loss": 0.856,
+      "step": 585
+    },
+    {
+      "epoch": 0.3470535978679301,
+      "grad_norm": 1.851962175877676,
+      "learning_rate": 1.9617226174464945e-05,
+      "loss": 0.8798,
+      "step": 586
+    },
+    {
+      "epoch": 0.34764583950251704,
+      "grad_norm": 1.4058738019152999,
+      "learning_rate": 1.961547161314759e-05,
+      "loss": 0.862,
+      "step": 587
+    },
+    {
+      "epoch": 0.3482380811371039,
+      "grad_norm": 1.9819045004335991,
+      "learning_rate": 1.9613713118561638e-05,
+      "loss": 0.8779,
+      "step": 588
+    },
+    {
+      "epoch": 0.34883032277169085,
+      "grad_norm": 1.6060471271059031,
+      "learning_rate": 1.961195069142642e-05,
+      "loss": 0.843,
+      "step": 589
+    },
+    {
+      "epoch": 0.3494225644062778,
+      "grad_norm": 1.5640055985461452,
+      "learning_rate": 1.961018433246286e-05,
+      "loss": 0.8379,
+      "step": 590
+    },
+    {
+      "epoch": 0.35001480604086466,
+      "grad_norm": 3.0209204466667905,
+      "learning_rate": 1.9608414042393503e-05,
+      "loss": 0.825,
+      "step": 591
+    },
+    {
+      "epoch": 0.3506070476754516,
+      "grad_norm": 1.6509809673195002,
+      "learning_rate": 1.9606639821942496e-05,
+      "loss": 0.858,
+      "step": 592
+    },
+    {
+      "epoch": 0.3511992893100385,
+      "grad_norm": 1.8338700183391534,
+      "learning_rate": 1.9604861671835593e-05,
+      "loss": 0.9081,
+      "step": 593
+    },
+    {
+      "epoch": 0.3517915309446254,
+      "grad_norm": 1.6832166078423174,
+      "learning_rate": 1.9603079592800157e-05,
+      "loss": 0.8525,
+      "step": 594
+    },
+    {
+      "epoch": 0.35238377257921233,
+      "grad_norm": 3.6998428943163715,
+      "learning_rate": 1.960129358556516e-05,
+      "loss": 0.9052,
+      "step": 595
+    },
+    {
+      "epoch": 0.3529760142137992,
+      "grad_norm": 1.8174759599966348,
+      "learning_rate": 1.9599503650861183e-05,
+      "loss": 0.8153,
+      "step": 596
+    },
+    {
+      "epoch": 0.35356825584838614,
+      "grad_norm": 3.5284496969394277,
+      "learning_rate": 1.9597709789420404e-05,
+      "loss": 0.8488,
+      "step": 597
+    },
+    {
+      "epoch": 0.35416049748297307,
+      "grad_norm": 8.653779192107086,
+      "learning_rate": 1.959591200197662e-05,
+      "loss": 0.8625,
+      "step": 598
+    },
+    {
+      "epoch": 0.35475273911755995,
+      "grad_norm": 1.8759995754403318,
+      "learning_rate": 1.9594110289265218e-05,
+      "loss": 0.8618,
+      "step": 599
+    },
+    {
+      "epoch": 0.3553449807521469,
+      "grad_norm": 1.9485225987035864,
+      "learning_rate": 1.9592304652023208e-05,
+      "loss": 0.8226,
+      "step": 600
+    },
+    {
+      "epoch": 0.3559372223867338,
+      "grad_norm": 3.686184834700953,
+      "learning_rate": 1.959049509098919e-05,
+      "loss": 0.8781,
+      "step": 601
+    },
+    {
+      "epoch": 0.3565294640213207,
+      "grad_norm": 2.8230171313225783,
+      "learning_rate": 1.9588681606903385e-05,
+      "loss": 0.8445,
+      "step": 602
+    },
+    {
+      "epoch": 0.3571217056559076,
+      "grad_norm": 2.7255745530386193,
+      "learning_rate": 1.95868642005076e-05,
+      "loss": 0.8965,
+      "step": 603
+    },
+    {
+      "epoch": 0.35771394729049455,
+      "grad_norm": 3.1116205729926385,
+      "learning_rate": 1.9585042872545266e-05,
+      "loss": 0.8434,
+      "step": 604
+    },
+    {
+      "epoch": 0.3583061889250814,
+      "grad_norm": 3.8126207167526154,
+      "learning_rate": 1.9583217623761404e-05,
+      "loss": 0.8651,
+      "step": 605
+    },
+    {
+      "epoch": 0.35889843055966836,
+      "grad_norm": 1.658269096943048,
+      "learning_rate": 1.958138845490264e-05,
+      "loss": 0.8434,
+      "step": 606
+    },
+    {
+      "epoch": 0.35949067219425523,
+      "grad_norm": 2.040025083348231,
+      "learning_rate": 1.9579555366717214e-05,
+      "loss": 0.8795,
+      "step": 607
+    },
+    {
+      "epoch": 0.36008291382884217,
+      "grad_norm": 4.804650864705654,
+      "learning_rate": 1.9577718359954955e-05,
+      "loss": 0.8943,
+      "step": 608
+    },
+    {
+      "epoch": 0.3606751554634291,
+      "grad_norm": 1.7709023659733834,
+      "learning_rate": 1.957587743536731e-05,
+      "loss": 0.8878,
+      "step": 609
+    },
+    {
+      "epoch": 0.361267397098016,
+      "grad_norm": 2.25150632588813,
+      "learning_rate": 1.9574032593707314e-05,
+      "loss": 0.8498,
+      "step": 610
+    },
+    {
+      "epoch": 0.3618596387326029,
+      "grad_norm": 7.9605607465838695,
+      "learning_rate": 1.9572183835729613e-05,
+      "loss": 0.9201,
+      "step": 611
+    },
+    {
+      "epoch": 0.36245188036718984,
+      "grad_norm": 2.5393697091836627,
+      "learning_rate": 1.957033116219045e-05,
+      "loss": 0.8733,
+      "step": 612
+    },
+    {
+      "epoch": 0.3630441220017767,
+      "grad_norm": 1.6878302312671156,
+      "learning_rate": 1.956847457384768e-05,
+      "loss": 0.8936,
+      "step": 613
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 1.868290304790888,
+      "learning_rate": 1.956661407146075e-05,
+      "loss": 0.8469,
+      "step": 614
+    },
+    {
+      "epoch": 0.3642286052709505,
+      "grad_norm": 1.7524149989576479,
+      "learning_rate": 1.9564749655790706e-05,
+      "loss": 0.8834,
+      "step": 615
+    },
+    {
+      "epoch": 0.36482084690553745,
+      "grad_norm": 3.268017562205005,
+      "learning_rate": 1.9562881327600197e-05,
+      "loss": 0.8549,
+      "step": 616
+    },
+    {
+      "epoch": 0.3654130885401244,
+      "grad_norm": 1.976076053518677,
+      "learning_rate": 1.9561009087653483e-05,
+      "loss": 0.8568,
+      "step": 617
+    },
+    {
+      "epoch": 0.36600533017471126,
+      "grad_norm": 2.637521424974901,
+      "learning_rate": 1.955913293671641e-05,
+      "loss": 0.8918,
+      "step": 618
+    },
+    {
+      "epoch": 0.3665975718092982,
+      "grad_norm": 3.314444680794765,
+      "learning_rate": 1.9557252875556428e-05,
+      "loss": 0.8414,
+      "step": 619
+    },
+    {
+      "epoch": 0.3671898134438851,
+      "grad_norm": 2.3746791597113766,
+      "learning_rate": 1.9555368904942593e-05,
+      "loss": 0.8457,
+      "step": 620
+    },
+    {
+      "epoch": 0.367782055078472,
+      "grad_norm": 2.6011871499667523,
+      "learning_rate": 1.9553481025645545e-05,
+      "loss": 0.8873,
+      "step": 621
+    },
+    {
+      "epoch": 0.36837429671305894,
+      "grad_norm": 6.819452995677599,
+      "learning_rate": 1.9551589238437546e-05,
+      "loss": 0.8907,
+      "step": 622
+    },
+    {
+      "epoch": 0.36896653834764587,
+      "grad_norm": 2.048515035382417,
+      "learning_rate": 1.954969354409243e-05,
+      "loss": 0.8392,
+      "step": 623
+    },
+    {
+      "epoch": 0.36955877998223274,
+      "grad_norm": 2.3431036995490855,
+      "learning_rate": 1.954779394338566e-05,
+      "loss": 0.8916,
+      "step": 624
+    },
+    {
+      "epoch": 0.3701510216168197,
+      "grad_norm": 3.0084271040530073,
+      "learning_rate": 1.954589043709426e-05,
+      "loss": 0.904,
+      "step": 625
+    },
+    {
+      "epoch": 0.37074326325140655,
+      "grad_norm": 1.5517306350499644,
+      "learning_rate": 1.954398302599688e-05,
+      "loss": 0.8716,
+      "step": 626
+    },
+    {
+      "epoch": 0.3713355048859935,
+      "grad_norm": 4.467132411617738,
+      "learning_rate": 1.954207171087376e-05,
+      "loss": 0.8593,
+      "step": 627
+    },
+    {
+      "epoch": 0.3719277465205804,
+      "grad_norm": 2.292921810220871,
+      "learning_rate": 1.9540156492506734e-05,
+      "loss": 0.8766,
+      "step": 628
+    },
+    {
+      "epoch": 0.3725199881551673,
+      "grad_norm": 1.9559474188452068,
+      "learning_rate": 1.9538237371679233e-05,
+      "loss": 0.8705,
+      "step": 629
+    },
+    {
+      "epoch": 0.3731122297897542,
+      "grad_norm": 1.6930579045607708,
+      "learning_rate": 1.9536314349176288e-05,
+      "loss": 0.8455,
+      "step": 630
+    },
+    {
+      "epoch": 0.37370447142434116,
+      "grad_norm": 5.371918547587357,
+      "learning_rate": 1.9534387425784518e-05,
+      "loss": 0.8998,
+      "step": 631
+    },
+    {
+      "epoch": 0.37429671305892803,
+      "grad_norm": 1.7900417336112107,
+      "learning_rate": 1.9532456602292148e-05,
+      "loss": 0.8481,
+      "step": 632
+    },
+    {
+      "epoch": 0.37488895469351496,
+      "grad_norm": 2.2958736067272416,
+      "learning_rate": 1.9530521879488993e-05,
+      "loss": 0.9105,
+      "step": 633
+    },
+    {
+      "epoch": 0.37548119632810184,
+      "grad_norm": 2.638086764469702,
+      "learning_rate": 1.952858325816646e-05,
+      "loss": 0.8381,
+      "step": 634
+    },
+    {
+      "epoch": 0.3760734379626888,
+      "grad_norm": 2.439133220470199,
+      "learning_rate": 1.9526640739117555e-05,
+      "loss": 0.8751,
+      "step": 635
+    },
+    {
+      "epoch": 0.3766656795972757,
+      "grad_norm": 1.8675503716373216,
+      "learning_rate": 1.9524694323136883e-05,
+      "loss": 0.883,
+      "step": 636
+    },
+    {
+      "epoch": 0.3772579212318626,
+      "grad_norm": 4.742878698850084,
+      "learning_rate": 1.952274401102063e-05,
+      "loss": 0.888,
+      "step": 637
+    },
+    {
+      "epoch": 0.3778501628664495,
+      "grad_norm": 2.420961821182289,
+      "learning_rate": 1.952078980356659e-05,
+      "loss": 0.8812,
+      "step": 638
+    },
+    {
+      "epoch": 0.37844240450103644,
+      "grad_norm": 2.237864761118068,
+      "learning_rate": 1.9518831701574136e-05,
+      "loss": 0.8906,
+      "step": 639
+    },
+    {
+      "epoch": 0.3790346461356233,
+      "grad_norm": 3.5814294479841915,
+      "learning_rate": 1.951686970584425e-05,
+      "loss": 0.8762,
+      "step": 640
+    },
+    {
+      "epoch": 0.37962688777021025,
+      "grad_norm": 3.322882097775164,
+      "learning_rate": 1.951490381717949e-05,
+      "loss": 0.8855,
+      "step": 641
+    },
+    {
+      "epoch": 0.3802191294047972,
+      "grad_norm": 2.2781530112916104,
+      "learning_rate": 1.9512934036384026e-05,
+      "loss": 0.8861,
+      "step": 642
+    },
+    {
+      "epoch": 0.38081137103938406,
+      "grad_norm": 3.1293049656485166,
+      "learning_rate": 1.95109603642636e-05,
+      "loss": 0.8684,
+      "step": 643
+    },
+    {
+      "epoch": 0.381403612673971,
+      "grad_norm": 2.2366175666301626,
+      "learning_rate": 1.9508982801625557e-05,
+      "loss": 0.8209,
+      "step": 644
+    },
+    {
+      "epoch": 0.38199585430855787,
+      "grad_norm": 2.0870198727029776,
+      "learning_rate": 1.9507001349278834e-05,
+      "loss": 0.8726,
+      "step": 645
+    },
+    {
+      "epoch": 0.3825880959431448,
+      "grad_norm": 5.789578141166538,
+      "learning_rate": 1.9505016008033953e-05,
+      "loss": 0.8065,
+      "step": 646
+    },
+    {
+      "epoch": 0.38318033757773173,
+      "grad_norm": 3.0810399663606183,
+      "learning_rate": 1.9503026778703034e-05,
+      "loss": 0.8686,
+      "step": 647
+    },
+    {
+      "epoch": 0.3837725792123186,
+      "grad_norm": 3.6634728348627528,
+      "learning_rate": 1.950103366209978e-05,
+      "loss": 0.8719,
+      "step": 648
+    },
+    {
+      "epoch": 0.38436482084690554,
+      "grad_norm": 2.8736691268816874,
+      "learning_rate": 1.949903665903949e-05,
+      "loss": 0.8767,
+      "step": 649
+    },
+    {
+      "epoch": 0.3849570624814925,
+      "grad_norm": 5.882320046408276,
+      "learning_rate": 1.949703577033905e-05,
+      "loss": 0.8671,
+      "step": 650
+    },
+    {
+      "epoch": 0.38554930411607935,
+      "grad_norm": 2.9528476437318925,
+      "learning_rate": 1.9495030996816932e-05,
+      "loss": 0.9013,
+      "step": 651
+    },
+    {
+      "epoch": 0.3861415457506663,
+      "grad_norm": 2.6363739674663247,
+      "learning_rate": 1.9493022339293207e-05,
+      "loss": 0.8833,
+      "step": 652
+    },
+    {
+      "epoch": 0.38673378738525316,
+      "grad_norm": 2.17785246783246,
+      "learning_rate": 1.949100979858953e-05,
+      "loss": 0.8866,
+      "step": 653
+    },
+    {
+      "epoch": 0.3873260290198401,
+      "grad_norm": 5.128470877775802,
+      "learning_rate": 1.9488993375529137e-05,
+      "loss": 0.9099,
+      "step": 654
+    },
+    {
+      "epoch": 0.387918270654427,
+      "grad_norm": 2.804392512181631,
+      "learning_rate": 1.9486973070936862e-05,
+      "loss": 0.8498,
+      "step": 655
+    },
+    {
+      "epoch": 0.3885105122890139,
+      "grad_norm": 2.9434325093740905,
+      "learning_rate": 1.9484948885639122e-05,
+      "loss": 0.8449,
+      "step": 656
+    },
+    {
+      "epoch": 0.38910275392360083,
+      "grad_norm": 4.154023552286942,
+      "learning_rate": 1.9482920820463923e-05,
+      "loss": 0.8563,
+      "step": 657
+    },
+    {
+      "epoch": 0.38969499555818776,
+      "grad_norm": 3.148511348448834,
+      "learning_rate": 1.948088887624086e-05,
+      "loss": 0.8056,
+      "step": 658
+    },
+    {
+      "epoch": 0.39028723719277464,
+      "grad_norm": 3.926327591148058,
+      "learning_rate": 1.947885305380111e-05,
+      "loss": 0.8888,
+      "step": 659
+    },
+    {
+      "epoch": 0.39087947882736157,
+      "grad_norm": 5.980099519975163,
+      "learning_rate": 1.9476813353977442e-05,
+      "loss": 0.8454,
+      "step": 660
+    },
+    {
+      "epoch": 0.3914717204619485,
+      "grad_norm": 3.3367209292691,
+      "learning_rate": 1.9474769777604198e-05,
+      "loss": 0.8548,
+      "step": 661
+    },
+    {
+      "epoch": 0.3920639620965354,
+      "grad_norm": 2.968203146893538,
+      "learning_rate": 1.947272232551733e-05,
+      "loss": 0.84,
+      "step": 662
+    },
+    {
+      "epoch": 0.3926562037311223,
+      "grad_norm": 39.05789203370318,
+      "learning_rate": 1.9470670998554352e-05,
+      "loss": 0.8914,
+      "step": 663
+    },
+    {
+      "epoch": 0.3932484453657092,
+      "grad_norm": 3.4148463229844506,
+      "learning_rate": 1.9468615797554374e-05,
+      "loss": 0.8605,
+      "step": 664
+    },
+    {
+      "epoch": 0.3938406870002961,
+      "grad_norm": 6.8180037854052316,
+      "learning_rate": 1.9466556723358084e-05,
+      "loss": 0.9065,
+      "step": 665
+    },
+    {
+      "epoch": 0.39443292863488305,
+      "grad_norm": 2.1152979135979484,
+      "learning_rate": 1.946449377680777e-05,
+      "loss": 0.8731,
+      "step": 666
+    },
+    {
+      "epoch": 0.3950251702694699,
+      "grad_norm": 2.4870435911253033,
+      "learning_rate": 1.9462426958747285e-05,
+      "loss": 0.8652,
+      "step": 667
+    },
+    {
+      "epoch": 0.39561741190405686,
+      "grad_norm": 2.6453498019333277,
+      "learning_rate": 1.9460356270022073e-05,
+      "loss": 0.898,
+      "step": 668
+    },
+    {
+      "epoch": 0.3962096535386438,
+      "grad_norm": 3.0508069490190066,
+      "learning_rate": 1.945828171147917e-05,
+      "loss": 0.9055,
+      "step": 669
+    },
+    {
+      "epoch": 0.39680189517323067,
+      "grad_norm": 1.6241048949900765,
+      "learning_rate": 1.945620328396718e-05,
+      "loss": 0.9111,
+      "step": 670
+    },
+    {
+      "epoch": 0.3973941368078176,
+      "grad_norm": 1.2381656243162942,
+      "learning_rate": 1.9454120988336297e-05,
+      "loss": 0.9074,
+      "step": 671
+    },
+    {
+      "epoch": 0.3979863784424045,
+      "grad_norm": 1.8466695435602942,
+      "learning_rate": 1.9452034825438302e-05,
+      "loss": 0.9597,
+      "step": 672
+    },
+    {
+      "epoch": 0.3985786200769914,
+      "grad_norm": 1.2562124363063687,
+      "learning_rate": 1.9449944796126547e-05,
+      "loss": 0.9371,
+      "step": 673
+    },
+    {
+      "epoch": 0.39917086171157834,
+      "grad_norm": 1.7231260645325064,
+      "learning_rate": 1.9447850901255975e-05,
+      "loss": 0.8556,
+      "step": 674
+    },
+    {
+      "epoch": 0.3997631033461652,
+      "grad_norm": 1.6192363719486302,
+      "learning_rate": 1.9445753141683107e-05,
+      "loss": 0.8885,
+      "step": 675
+    },
+    {
+      "epoch": 0.40035534498075215,
+      "grad_norm": 1.675728205658867,
+      "learning_rate": 1.9443651518266044e-05,
+      "loss": 0.8692,
+      "step": 676
+    },
+    {
+      "epoch": 0.4009475866153391,
+      "grad_norm": 2.04303793153201,
+      "learning_rate": 1.9441546031864467e-05,
+      "loss": 0.8552,
+      "step": 677
+    },
+    {
+      "epoch": 0.40153982824992596,
+      "grad_norm": 1.9133034698626181,
+      "learning_rate": 1.943943668333964e-05,
+      "loss": 0.882,
+      "step": 678
+    },
+    {
+      "epoch": 0.4021320698845129,
+      "grad_norm": 2.1370541384543342,
+      "learning_rate": 1.9437323473554404e-05,
+      "loss": 0.8968,
+      "step": 679
+    },
+    {
+      "epoch": 0.4027243115190998,
+      "grad_norm": 1.5028317313676052,
+      "learning_rate": 1.943520640337318e-05,
+      "loss": 0.8582,
+      "step": 680
+    },
+    {
+      "epoch": 0.4033165531536867,
+      "grad_norm": 3.9594823865775868,
+      "learning_rate": 1.943308547366197e-05,
+      "loss": 0.8111,
+      "step": 681
+    },
+    {
+      "epoch": 0.40390879478827363,
+      "grad_norm": 3.417942000806396,
+      "learning_rate": 1.9430960685288355e-05,
+      "loss": 0.8385,
+      "step": 682
+    },
+    {
+      "epoch": 0.4045010364228605,
+      "grad_norm": 1.6652473512379486,
+      "learning_rate": 1.9428832039121487e-05,
+      "loss": 0.8686,
+      "step": 683
+    },
+    {
+      "epoch": 0.40509327805744744,
+      "grad_norm": 3.8546279443523663,
+      "learning_rate": 1.942669953603211e-05,
+      "loss": 0.8697,
+      "step": 684
+    },
+    {
+      "epoch": 0.40568551969203437,
+      "grad_norm": 1.9412501113108496,
+      "learning_rate": 1.9424563176892534e-05,
+      "loss": 0.8235,
+      "step": 685
+    },
+    {
+      "epoch": 0.40627776132662125,
+      "grad_norm": 1.8569982182714377,
+      "learning_rate": 1.9422422962576646e-05,
+      "loss": 0.9344,
+      "step": 686
+    },
+    {
+      "epoch": 0.4068700029612082,
+      "grad_norm": 2.1750669740257824,
+      "learning_rate": 1.9420278893959922e-05,
+      "loss": 0.868,
+      "step": 687
+    },
+    {
+      "epoch": 0.4074622445957951,
+      "grad_norm": 2.4649163202529607,
+      "learning_rate": 1.94181309719194e-05,
+      "loss": 0.8585,
+      "step": 688
+    },
+    {
+      "epoch": 0.408054486230382,
+      "grad_norm": 2.5516191944482443,
+      "learning_rate": 1.9415979197333704e-05,
+      "loss": 0.866,
+      "step": 689
+    },
+    {
+      "epoch": 0.4086467278649689,
+      "grad_norm": 2.140588212831023,
+      "learning_rate": 1.941382357108303e-05,
+      "loss": 0.8736,
+      "step": 690
+    },
+    {
+      "epoch": 0.4092389694995558,
+      "grad_norm": 2.6369171422119417,
+      "learning_rate": 1.941166409404915e-05,
+      "loss": 0.8362,
+      "step": 691
+    },
+    {
+      "epoch": 0.4098312111341427,
+      "grad_norm": 3.0607023188332074,
+      "learning_rate": 1.9409500767115414e-05,
+      "loss": 0.8399,
+      "step": 692
+    },
+    {
+      "epoch": 0.41042345276872966,
+      "grad_norm": 3.0627901913227924,
+      "learning_rate": 1.940733359116674e-05,
+      "loss": 0.879,
+      "step": 693
+    },
+    {
+      "epoch": 0.41101569440331653,
+      "grad_norm": 3.2361644739692292,
+      "learning_rate": 1.9405162567089627e-05,
+      "loss": 0.8678,
+      "step": 694
+    },
+    {
+      "epoch": 0.41160793603790347,
+      "grad_norm": 18.463939120851368,
+      "learning_rate": 1.940298769577215e-05,
+      "loss": 0.9256,
+      "step": 695
+    },
+    {
+      "epoch": 0.4122001776724904,
+      "grad_norm": 10.368145595990683,
+      "learning_rate": 1.9400808978103948e-05,
+      "loss": 0.8188,
+      "step": 696
+    },
+    {
+      "epoch": 0.4127924193070773,
+      "grad_norm": 2.4187494879479745,
+      "learning_rate": 1.939862641497624e-05,
+      "loss": 0.8745,
+      "step": 697
+    },
+    {
+      "epoch": 0.4133846609416642,
+      "grad_norm": 1.806561112554444,
+      "learning_rate": 1.939644000728182e-05,
+      "loss": 0.8622,
+      "step": 698
+    },
+    {
+      "epoch": 0.41397690257625114,
+      "grad_norm": 2.376096051071182,
+      "learning_rate": 1.9394249755915047e-05,
+      "loss": 0.8613,
+      "step": 699
+    },
+    {
+      "epoch": 0.414569144210838,
+      "grad_norm": 2.0606522501522724,
+      "learning_rate": 1.939205566177186e-05,
+      "loss": 0.8889,
+      "step": 700
+    },
+    {
+      "epoch": 0.41516138584542495,
+      "grad_norm": 1.6424519963083875,
+      "learning_rate": 1.9389857725749767e-05,
+      "loss": 0.924,
+      "step": 701
+    },
+    {
+      "epoch": 0.4157536274800118,
+      "grad_norm": 2.5660132236892608,
+      "learning_rate": 1.938765594874785e-05,
+      "loss": 0.8472,
+      "step": 702
+    },
+    {
+      "epoch": 0.41634586911459875,
+      "grad_norm": 2.8325159916694598,
+      "learning_rate": 1.9385450331666754e-05,
+      "loss": 0.8887,
+      "step": 703
+    },
+    {
+      "epoch": 0.4169381107491857,
+      "grad_norm": 1.7008840901930986,
+      "learning_rate": 1.93832408754087e-05,
+      "loss": 0.8748,
+      "step": 704
+    },
+    {
+      "epoch": 0.41753035238377256,
+      "grad_norm": 1.9216283469656232,
+      "learning_rate": 1.9381027580877486e-05,
+      "loss": 0.8415,
+      "step": 705
+    },
+    {
+      "epoch": 0.4181225940183595,
+      "grad_norm": 2.026996390561362,
+      "learning_rate": 1.937881044897847e-05,
+      "loss": 0.8416,
+      "step": 706
+    },
+    {
+      "epoch": 0.4187148356529464,
+      "grad_norm": 1.423893980632455,
+      "learning_rate": 1.9376589480618583e-05,
+      "loss": 0.8505,
+      "step": 707
+    },
+    {
+      "epoch": 0.4193070772875333,
+      "grad_norm": 1.8333091580377612,
+      "learning_rate": 1.937436467670633e-05,
+      "loss": 0.8715,
+      "step": 708
+    },
+    {
+      "epoch": 0.41989931892212023,
+      "grad_norm": 2.2298036589587658,
+      "learning_rate": 1.937213603815178e-05,
+      "loss": 0.8767,
+      "step": 709
+    },
+    {
+      "epoch": 0.4204915605567071,
+      "grad_norm": 3.440326466071599,
+      "learning_rate": 1.9369903565866565e-05,
+      "loss": 0.8322,
+      "step": 710
+    },
+    {
+      "epoch": 0.42108380219129404,
+      "grad_norm": 1.6529181371554102,
+      "learning_rate": 1.9367667260763898e-05,
+      "loss": 0.899,
+      "step": 711
+    },
+    {
+      "epoch": 0.421676043825881,
+      "grad_norm": 1.811748367956541,
+      "learning_rate": 1.936542712375855e-05,
+      "loss": 0.805,
+      "step": 712
+    },
+    {
+      "epoch": 0.42226828546046785,
+      "grad_norm": 1.4423225369050077,
+      "learning_rate": 1.9363183155766867e-05,
+      "loss": 0.9193,
+      "step": 713
+    },
+    {
+      "epoch": 0.4228605270950548,
+      "grad_norm": 3.711540769404225,
+      "learning_rate": 1.9360935357706756e-05,
+      "loss": 0.8143,
+      "step": 714
+    },
+    {
+      "epoch": 0.4234527687296417,
+      "grad_norm": 1.6013652544896453,
+      "learning_rate": 1.9358683730497695e-05,
+      "loss": 0.8883,
+      "step": 715
+    },
+    {
+      "epoch": 0.4240450103642286,
+      "grad_norm": 4.765502585221681,
+      "learning_rate": 1.9356428275060722e-05,
+      "loss": 0.845,
+      "step": 716
+    },
+    {
+      "epoch": 0.4246372519988155,
+      "grad_norm": 3.1849530858655983,
+      "learning_rate": 1.9354168992318448e-05,
+      "loss": 0.9136,
+      "step": 717
+    },
+    {
+      "epoch": 0.42522949363340246,
+      "grad_norm": 1.62933697491851,
+      "learning_rate": 1.9351905883195044e-05,
+      "loss": 0.8887,
+      "step": 718
+    },
+    {
+      "epoch": 0.42582173526798933,
+      "grad_norm": 1.8945696646562518,
+      "learning_rate": 1.9349638948616253e-05,
+      "loss": 0.8268,
+      "step": 719
+    },
+    {
+      "epoch": 0.42641397690257626,
+      "grad_norm": 2.4172838989795125,
+      "learning_rate": 1.934736818950937e-05,
+      "loss": 0.8683,
+      "step": 720
+    },
+    {
+      "epoch": 0.42700621853716314,
+      "grad_norm": 1.884065405882965,
+      "learning_rate": 1.9345093606803276e-05,
+      "loss": 0.8832,
+      "step": 721
+    },
+    {
+      "epoch": 0.42759846017175007,
+      "grad_norm": 1.77943938389625,
+      "learning_rate": 1.9342815201428394e-05,
+      "loss": 0.8842,
+      "step": 722
+    },
+    {
+      "epoch": 0.428190701806337,
+      "grad_norm": 1.4357452116165423,
+      "learning_rate": 1.9340532974316727e-05,
+      "loss": 0.8274,
+      "step": 723
+    },
+    {
+      "epoch": 0.4287829434409239,
+      "grad_norm": 1.4379221296115967,
+      "learning_rate": 1.9338246926401828e-05,
+      "loss": 0.8388,
+      "step": 724
+    },
+    {
+      "epoch": 0.4293751850755108,
+      "grad_norm": 2.07752636716361,
+      "learning_rate": 1.933595705861882e-05,
+      "loss": 0.833,
+      "step": 725
+    },
+    {
+      "epoch": 0.42996742671009774,
+      "grad_norm": 2.0105999576116322,
+      "learning_rate": 1.9333663371904388e-05,
+      "loss": 0.8473,
+      "step": 726
+    },
+    {
+      "epoch": 0.4305596683446846,
+      "grad_norm": 3.956301063546753,
+      "learning_rate": 1.933136586719678e-05,
+      "loss": 0.8496,
+      "step": 727
+    },
+    {
+      "epoch": 0.43115190997927155,
+      "grad_norm": 1.3549992353009284,
+      "learning_rate": 1.9329064545435803e-05,
+      "loss": 0.8747,
+      "step": 728
+    },
+    {
+      "epoch": 0.43174415161385843,
+      "grad_norm": 2.227577277579618,
+      "learning_rate": 1.932675940756283e-05,
+      "loss": 0.8362,
+      "step": 729
+    },
+    {
+      "epoch": 0.43233639324844536,
+      "grad_norm": 1.5147909801235255,
+      "learning_rate": 1.932445045452079e-05,
+      "loss": 0.8688,
+      "step": 730
+    },
+    {
+      "epoch": 0.4329286348830323,
+      "grad_norm": 1.920697239844712,
+      "learning_rate": 1.9322137687254175e-05,
+      "loss": 0.8608,
+      "step": 731
+    },
+    {
+      "epoch": 0.43352087651761917,
+      "grad_norm": 1.9550148078238179,
+      "learning_rate": 1.931982110670904e-05,
+      "loss": 0.8438,
+      "step": 732
+    },
+    {
+      "epoch": 0.4341131181522061,
+      "grad_norm": 1.9357461615720977,
+      "learning_rate": 1.9317500713832987e-05,
+      "loss": 0.8289,
+      "step": 733
+    },
+    {
+      "epoch": 0.43470535978679303,
+      "grad_norm": 2.436470304212335,
+      "learning_rate": 1.9315176509575196e-05,
+      "loss": 0.8505,
+      "step": 734
+    },
+    {
+      "epoch": 0.4352976014213799,
+      "grad_norm": 2.0394551856453633,
+      "learning_rate": 1.93128484948864e-05,
+      "loss": 0.8595,
+      "step": 735
+    },
+    {
+      "epoch": 0.43588984305596684,
+      "grad_norm": 16.717645135664075,
+      "learning_rate": 1.9310516670718877e-05,
+      "loss": 0.8358,
+      "step": 736
+    },
+    {
+      "epoch": 0.4364820846905538,
+      "grad_norm": 1.9423408377644282,
+      "learning_rate": 1.930818103802648e-05,
+      "loss": 0.861,
+      "step": 737
+    },
+    {
+      "epoch": 0.43707432632514065,
+      "grad_norm": 4.812386875770368,
+      "learning_rate": 1.9305841597764615e-05,
+      "loss": 0.8801,
+      "step": 738
+    },
+    {
+      "epoch": 0.4376665679597276,
+      "grad_norm": 1.4672305025896835,
+      "learning_rate": 1.9303498350890246e-05,
+      "loss": 0.8782,
+      "step": 739
+    },
+    {
+      "epoch": 0.43825880959431446,
+      "grad_norm": 2.2508593137348676,
+      "learning_rate": 1.9301151298361887e-05,
+      "loss": 0.877,
+      "step": 740
+    },
+    {
+      "epoch": 0.4388510512289014,
+      "grad_norm": 1.7135079210199697,
+      "learning_rate": 1.9298800441139623e-05,
+      "loss": 0.8972,
+      "step": 741
+    },
+    {
+      "epoch": 0.4394432928634883,
+      "grad_norm": 1.4490997536994275,
+      "learning_rate": 1.9296445780185077e-05,
+      "loss": 0.8224,
+      "step": 742
+    },
+    {
+      "epoch": 0.4400355344980752,
+      "grad_norm": 1.770609855482205,
+      "learning_rate": 1.9294087316461446e-05,
+      "loss": 0.8803,
+      "step": 743
+    },
+    {
+      "epoch": 0.44062777613266213,
+      "grad_norm": 4.232242865631687,
+      "learning_rate": 1.929172505093347e-05,
+      "loss": 0.8834,
+      "step": 744
+    },
+    {
+      "epoch": 0.44122001776724906,
+      "grad_norm": 2.0717939954326625,
+      "learning_rate": 1.9289358984567446e-05,
+      "loss": 0.8896,
+      "step": 745
+    },
+    {
+      "epoch": 0.44181225940183594,
+      "grad_norm": 4.304229428729412,
+      "learning_rate": 1.928698911833123e-05,
+      "loss": 0.8621,
+      "step": 746
+    },
+    {
+      "epoch": 0.44240450103642287,
+      "grad_norm": 2.200899929703116,
+      "learning_rate": 1.928461545319424e-05,
+      "loss": 0.8648,
+      "step": 747
+    },
+    {
+      "epoch": 0.44299674267100975,
+      "grad_norm": 1.6834184926932727,
+      "learning_rate": 1.9282237990127425e-05,
+      "loss": 0.8714,
+      "step": 748
+    },
+    {
+      "epoch": 0.4435889843055967,
+      "grad_norm": 2.104833385776733,
+      "learning_rate": 1.927985673010331e-05,
+      "loss": 0.8172,
+      "step": 749
+    },
+    {
+      "epoch": 0.4441812259401836,
+      "grad_norm": 3.6197381435137324,
+      "learning_rate": 1.9277471674095952e-05,
+      "loss": 0.8392,
+      "step": 750
+    },
+    {
+      "epoch": 0.4447734675747705,
+      "grad_norm": 2.11958620090702,
+      "learning_rate": 1.9275082823080992e-05,
+      "loss": 0.8735,
+      "step": 751
+    },
+    {
+      "epoch": 0.4453657092093574,
+      "grad_norm": 3.3384124014666874,
+      "learning_rate": 1.927269017803559e-05,
+      "loss": 0.9202,
+      "step": 752
+    },
+    {
+      "epoch": 0.44595795084394435,
+      "grad_norm": 2.705834935266405,
+      "learning_rate": 1.9270293739938477e-05,
+      "loss": 0.8619,
+      "step": 753
+    },
+    {
+      "epoch": 0.4465501924785312,
+      "grad_norm": 2.7094042650834527,
+      "learning_rate": 1.9267893509769927e-05,
+      "loss": 0.8918,
+      "step": 754
+    },
+    {
+      "epoch": 0.44714243411311816,
+      "grad_norm": 1.7854900824015136,
+      "learning_rate": 1.926548948851178e-05,
+      "loss": 0.8564,
+      "step": 755
+    },
+    {
+      "epoch": 0.4477346757477051,
+      "grad_norm": 1.6005502953829145,
+      "learning_rate": 1.92630816771474e-05,
+      "loss": 0.8412,
+      "step": 756
+    },
+    {
+      "epoch": 0.44832691738229197,
+      "grad_norm": 7.492700831605615,
+      "learning_rate": 1.9260670076661735e-05,
+      "loss": 0.8718,
+      "step": 757
+    },
+    {
+      "epoch": 0.4489191590168789,
+      "grad_norm": 5.427630233101737,
+      "learning_rate": 1.925825468804125e-05,
+      "loss": 0.8778,
+      "step": 758
+    },
+    {
+      "epoch": 0.4495114006514658,
+      "grad_norm": 1.8888166698278774,
+      "learning_rate": 1.9255835512273982e-05,
+      "loss": 0.8596,
+      "step": 759
+    },
+    {
+      "epoch": 0.4501036422860527,
+      "grad_norm": 2.1812538978009055,
+      "learning_rate": 1.9253412550349507e-05,
+      "loss": 0.8691,
+      "step": 760
+    },
+    {
+      "epoch": 0.45069588392063964,
+      "grad_norm": 2.924997761771959,
+      "learning_rate": 1.9250985803258957e-05,
+      "loss": 0.8579,
+      "step": 761
+    },
+    {
+      "epoch": 0.4512881255552265,
+      "grad_norm": 2.1898518842107424,
+      "learning_rate": 1.9248555271995006e-05,
+      "loss": 0.9125,
+      "step": 762
+    },
+    {
+      "epoch": 0.45188036718981345,
+      "grad_norm": 2.1171881776125234,
+      "learning_rate": 1.924612095755188e-05,
+      "loss": 0.818,
+      "step": 763
+    },
+    {
+      "epoch": 0.4524726088244004,
+      "grad_norm": 1.860211579546916,
+      "learning_rate": 1.924368286092534e-05,
+      "loss": 0.8827,
+      "step": 764
+    },
+    {
+      "epoch": 0.45306485045898726,
+      "grad_norm": 1.6407946441138859,
+      "learning_rate": 1.9241240983112718e-05,
+      "loss": 0.8979,
+      "step": 765
+    },
+    {
+      "epoch": 0.4536570920935742,
+      "grad_norm": 1.8118381222826243,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.8969,
+      "step": 766
+    },
+    {
+      "epoch": 0.45424933372816106,
+      "grad_norm": 1.5722837226227442,
+      "learning_rate": 1.9236345887926215e-05,
+      "loss": 0.8211,
+      "step": 767
+    },
+    {
+      "epoch": 0.454841575362748,
+      "grad_norm": 1.690859845921261,
+      "learning_rate": 1.92338926725547e-05,
+      "loss": 0.9023,
+      "step": 768
+    },
+    {
+      "epoch": 0.45543381699733493,
+      "grad_norm": 1.3414797354584787,
+      "learning_rate": 1.9231435680001844e-05,
+      "loss": 0.8433,
+      "step": 769
+    },
+    {
+      "epoch": 0.4560260586319218,
+      "grad_norm": 2.9314823394828884,
+      "learning_rate": 1.9228974911272682e-05,
+      "loss": 0.8694,
+      "step": 770
+    },
+    {
+      "epoch": 0.45661830026650874,
+      "grad_norm": 1.6345887387544766,
+      "learning_rate": 1.9226510367373812e-05,
+      "loss": 0.9751,
+      "step": 771
+    },
+    {
+      "epoch": 0.45721054190109567,
+      "grad_norm": 1.4992911219830736,
+      "learning_rate": 1.922404204931337e-05,
+      "loss": 0.8129,
+      "step": 772
+    },
+    {
+      "epoch": 0.45780278353568254,
+      "grad_norm": 1.4628930115387142,
+      "learning_rate": 1.9221569958101038e-05,
+      "loss": 0.8945,
+      "step": 773
+    },
+    {
+      "epoch": 0.4583950251702695,
+      "grad_norm": 2.723334540080781,
+      "learning_rate": 1.9219094094748035e-05,
+      "loss": 0.8383,
+      "step": 774
+    },
+    {
+      "epoch": 0.4589872668048564,
+      "grad_norm": 1.9492331431341308,
+      "learning_rate": 1.9216614460267132e-05,
+      "loss": 0.8664,
+      "step": 775
+    },
+    {
+      "epoch": 0.4595795084394433,
+      "grad_norm": 1.8901483017176506,
+      "learning_rate": 1.9214131055672648e-05,
+      "loss": 0.9114,
+      "step": 776
+    },
+    {
+      "epoch": 0.4601717500740302,
+      "grad_norm": 1.5947480549984965,
+      "learning_rate": 1.921164388198042e-05,
+      "loss": 0.8619,
+      "step": 777
+    },
+    {
+      "epoch": 0.4607639917086171,
+      "grad_norm": 2.944268928829697,
+      "learning_rate": 1.9209152940207846e-05,
+      "loss": 0.8841,
+      "step": 778
+    },
+    {
+      "epoch": 0.461356233343204,
+      "grad_norm": 1.9959401062119286,
+      "learning_rate": 1.920665823137387e-05,
+      "loss": 0.8942,
+      "step": 779
+    },
+    {
+      "epoch": 0.46194847497779096,
+      "grad_norm": 2.055054417491741,
+      "learning_rate": 1.920415975649896e-05,
+      "loss": 0.8985,
+      "step": 780
+    },
+    {
+      "epoch": 0.46254071661237783,
+      "grad_norm": 1.43509961761953,
+      "learning_rate": 1.9201657516605136e-05,
+      "loss": 0.8409,
+      "step": 781
+    },
+    {
+      "epoch": 0.46313295824696477,
+      "grad_norm": 2.0964064977562464,
+      "learning_rate": 1.9199151512715948e-05,
+      "loss": 0.8603,
+      "step": 782
+    },
+    {
+      "epoch": 0.4637251998815517,
+      "grad_norm": 2.0471728931299933,
+      "learning_rate": 1.9196641745856502e-05,
+      "loss": 0.8195,
+      "step": 783
+    },
+    {
+      "epoch": 0.4643174415161386,
+      "grad_norm": 1.9479876262651854,
+      "learning_rate": 1.919412821705343e-05,
+      "loss": 0.892,
+      "step": 784
+    },
+    {
+      "epoch": 0.4649096831507255,
+      "grad_norm": 1.8429301406660341,
+      "learning_rate": 1.9191610927334905e-05,
+      "loss": 0.8517,
+      "step": 785
+    },
+    {
+      "epoch": 0.4655019247853124,
+      "grad_norm": 1.6626463990171203,
+      "learning_rate": 1.918908987773064e-05,
+      "loss": 0.8422,
+      "step": 786
+    },
+    {
+      "epoch": 0.4660941664198993,
+      "grad_norm": 2.484399506113593,
+      "learning_rate": 1.918656506927189e-05,
+      "loss": 0.8349,
+      "step": 787
+    },
+    {
+      "epoch": 0.46668640805448625,
+      "grad_norm": 2.2394704693255325,
+      "learning_rate": 1.918403650299144e-05,
+      "loss": 0.8584,
+      "step": 788
+    },
+    {
+      "epoch": 0.4672786496890731,
+      "grad_norm": 2.7696631299861494,
+      "learning_rate": 1.9181504179923622e-05,
+      "loss": 0.8702,
+      "step": 789
+    },
+    {
+      "epoch": 0.46787089132366005,
+      "grad_norm": 2.3702451262189603,
+      "learning_rate": 1.9178968101104285e-05,
+      "loss": 0.8812,
+      "step": 790
+    },
+    {
+      "epoch": 0.468463132958247,
+      "grad_norm": 3.347228840796043,
+      "learning_rate": 1.9176428267570842e-05,
+      "loss": 0.8925,
+      "step": 791
+    },
+    {
+      "epoch": 0.46905537459283386,
+      "grad_norm": 1.8592907890766055,
+      "learning_rate": 1.917388468036222e-05,
+      "loss": 0.859,
+      "step": 792
+    },
+    {
+      "epoch": 0.4696476162274208,
+      "grad_norm": 1.8720866828747977,
+      "learning_rate": 1.917133734051889e-05,
+      "loss": 0.8683,
+      "step": 793
+    },
+    {
+      "epoch": 0.4702398578620077,
+      "grad_norm": 2.3125042184278684,
+      "learning_rate": 1.9168786249082862e-05,
+      "loss": 0.8256,
+      "step": 794
+    },
+    {
+      "epoch": 0.4708320994965946,
+      "grad_norm": 1.719445243382972,
+      "learning_rate": 1.916623140709767e-05,
+      "loss": 0.9095,
+      "step": 795
+    },
+    {
+      "epoch": 0.47142434113118153,
+      "grad_norm": 3.0132158639000695,
+      "learning_rate": 1.9163672815608392e-05,
+      "loss": 0.8053,
+      "step": 796
+    },
+    {
+      "epoch": 0.4720165827657684,
+      "grad_norm": 1.933417130240812,
+      "learning_rate": 1.9161110475661634e-05,
+      "loss": 0.8653,
+      "step": 797
+    },
+    {
+      "epoch": 0.47260882440035534,
+      "grad_norm": 2.001369020588165,
+      "learning_rate": 1.9158544388305534e-05,
+      "loss": 0.8821,
+      "step": 798
+    },
+    {
+      "epoch": 0.4732010660349423,
+      "grad_norm": 2.236004511554376,
+      "learning_rate": 1.9155974554589774e-05,
+      "loss": 0.8531,
+      "step": 799
+    },
+    {
+      "epoch": 0.47379330766952915,
+      "grad_norm": 3.7355229765335785,
+      "learning_rate": 1.915340097556555e-05,
+      "loss": 0.8913,
+      "step": 800
+    },
+    {
+      "epoch": 0.4743855493041161,
+      "grad_norm": 6.246619019749031,
+      "learning_rate": 1.915082365228561e-05,
+      "loss": 0.9032,
+      "step": 801
+    },
+    {
+      "epoch": 0.474977790938703,
+      "grad_norm": 1.8815129005316729,
+      "learning_rate": 1.914824258580422e-05,
+      "loss": 0.8619,
+      "step": 802
+    },
+    {
+      "epoch": 0.4755700325732899,
+      "grad_norm": 1.662450993675011,
+      "learning_rate": 1.9145657777177186e-05,
+      "loss": 0.8465,
+      "step": 803
+    },
+    {
+      "epoch": 0.4761622742078768,
+      "grad_norm": 1.8295549474984136,
+      "learning_rate": 1.914306922746183e-05,
+      "loss": 0.8569,
+      "step": 804
+    },
+    {
+      "epoch": 0.4767545158424637,
+      "grad_norm": 10.419046505396,
+      "learning_rate": 1.914047693771702e-05,
+      "loss": 0.8476,
+      "step": 805
+    },
+    {
+      "epoch": 0.47734675747705063,
+      "grad_norm": 4.545366161612398,
+      "learning_rate": 1.9137880909003155e-05,
+      "loss": 0.8815,
+      "step": 806
+    },
+    {
+      "epoch": 0.47793899911163756,
+      "grad_norm": 2.3239626037301266,
+      "learning_rate": 1.9135281142382147e-05,
+      "loss": 0.812,
+      "step": 807
+    },
+    {
+      "epoch": 0.47853124074622444,
+      "grad_norm": 3.2052660355799256,
+      "learning_rate": 1.913267763891745e-05,
+      "loss": 0.8444,
+      "step": 808
+    },
+    {
+      "epoch": 0.47912348238081137,
+      "grad_norm": 2.661447353848109,
+      "learning_rate": 1.913007039967404e-05,
+      "loss": 0.8676,
+      "step": 809
+    },
+    {
+      "epoch": 0.4797157240153983,
+      "grad_norm": 3.1975284655785043,
+      "learning_rate": 1.912745942571843e-05,
+      "loss": 0.8973,
+      "step": 810
+    },
+    {
+      "epoch": 0.4803079656499852,
+      "grad_norm": 8.817742135689066,
+      "learning_rate": 1.9124844718118657e-05,
+      "loss": 0.8338,
+      "step": 811
+    },
+    {
+      "epoch": 0.4809002072845721,
+      "grad_norm": 6.647384676362516,
+      "learning_rate": 1.9122226277944276e-05,
+      "loss": 0.8034,
+      "step": 812
+    },
+    {
+      "epoch": 0.48149244891915904,
+      "grad_norm": 3.0068701129220665,
+      "learning_rate": 1.911960410626638e-05,
+      "loss": 0.8558,
+      "step": 813
+    },
+    {
+      "epoch": 0.4820846905537459,
+      "grad_norm": 2.4470838917289055,
+      "learning_rate": 1.9116978204157583e-05,
+      "loss": 0.8781,
+      "step": 814
+    },
+    {
+      "epoch": 0.48267693218833285,
+      "grad_norm": 3.568749125351664,
+      "learning_rate": 1.911434857269203e-05,
+      "loss": 0.8366,
+      "step": 815
+    },
+    {
+      "epoch": 0.48326917382291973,
+      "grad_norm": 2.788154497136242,
+      "learning_rate": 1.9111715212945384e-05,
+      "loss": 0.8179,
+      "step": 816
+    },
+    {
+      "epoch": 0.48386141545750666,
+      "grad_norm": 2.561601471580433,
+      "learning_rate": 1.9109078125994843e-05,
+      "loss": 0.8388,
+      "step": 817
+    },
+    {
+      "epoch": 0.4844536570920936,
+      "grad_norm": 13.67055884933429,
+      "learning_rate": 1.9106437312919116e-05,
+      "loss": 0.851,
+      "step": 818
+    },
+    {
+      "epoch": 0.48504589872668047,
+      "grad_norm": 3.884383319733934,
+      "learning_rate": 1.910379277479845e-05,
+      "loss": 0.8999,
+      "step": 819
+    },
+    {
+      "epoch": 0.4856381403612674,
+      "grad_norm": 4.2385883179797625,
+      "learning_rate": 1.910114451271461e-05,
+      "loss": 0.8981,
+      "step": 820
+    },
+    {
+      "epoch": 0.48623038199585433,
+      "grad_norm": 2.8064585598908356,
+      "learning_rate": 1.909849252775088e-05,
+      "loss": 0.8633,
+      "step": 821
+    },
+    {
+      "epoch": 0.4868226236304412,
+      "grad_norm": 3.4012309556283227,
+      "learning_rate": 1.9095836820992074e-05,
+      "loss": 0.8425,
+      "step": 822
+    },
+    {
+      "epoch": 0.48741486526502814,
+      "grad_norm": 1.8123186335536858,
+      "learning_rate": 1.9093177393524524e-05,
+      "loss": 0.8615,
+      "step": 823
+    },
+    {
+      "epoch": 0.488007106899615,
+      "grad_norm": 3.468928077402017,
+      "learning_rate": 1.9090514246436085e-05,
+      "loss": 0.8995,
+      "step": 824
+    },
+    {
+      "epoch": 0.48859934853420195,
+      "grad_norm": 1.8021817875541082,
+      "learning_rate": 1.908784738081614e-05,
+      "loss": 0.8284,
+      "step": 825
+    },
+    {
+      "epoch": 0.4891915901687889,
+      "grad_norm": 2.4825340026394374,
+      "learning_rate": 1.9085176797755575e-05,
+      "loss": 0.8585,
+      "step": 826
+    },
+    {
+      "epoch": 0.48978383180337576,
+      "grad_norm": 2.47602922744696,
+      "learning_rate": 1.908250249834682e-05,
+      "loss": 0.8805,
+      "step": 827
+    },
+    {
+      "epoch": 0.4903760734379627,
+      "grad_norm": 1.9033817119185694,
+      "learning_rate": 1.9079824483683808e-05,
+      "loss": 0.9176,
+      "step": 828
+    },
+    {
+      "epoch": 0.4909683150725496,
+      "grad_norm": 1.5335390856862279,
+      "learning_rate": 1.9077142754862e-05,
+      "loss": 0.8394,
+      "step": 829
+    },
+    {
+      "epoch": 0.4915605567071365,
+      "grad_norm": 1.8269219296715522,
+      "learning_rate": 1.9074457312978373e-05,
+      "loss": 0.8565,
+      "step": 830
+    },
+    {
+      "epoch": 0.49215279834172343,
+      "grad_norm": 1.8745343150998126,
+      "learning_rate": 1.907176815913142e-05,
+      "loss": 0.8292,
+      "step": 831
+    },
+    {
+      "epoch": 0.49274503997631036,
+      "grad_norm": 3.137235022218658,
+      "learning_rate": 1.9069075294421163e-05,
+      "loss": 0.8564,
+      "step": 832
+    },
+    {
+      "epoch": 0.49333728161089724,
+      "grad_norm": 4.136469645798408,
+      "learning_rate": 1.906637871994913e-05,
+      "loss": 0.8611,
+      "step": 833
+    },
+    {
+      "epoch": 0.49392952324548417,
+      "grad_norm": 2.105315946203615,
+      "learning_rate": 1.9063678436818372e-05,
+      "loss": 0.8306,
+      "step": 834
+    },
+    {
+      "epoch": 0.49452176488007105,
+      "grad_norm": 1.6169111636378513,
+      "learning_rate": 1.906097444613346e-05,
+      "loss": 0.8065,
+      "step": 835
+    },
+    {
+      "epoch": 0.495114006514658,
+      "grad_norm": 2.1563275408886495,
+      "learning_rate": 1.905826674900048e-05,
+      "loss": 0.8578,
+      "step": 836
+    },
+    {
+      "epoch": 0.4957062481492449,
+      "grad_norm": 1.6035367865321255,
+      "learning_rate": 1.9055555346527024e-05,
+      "loss": 0.8172,
+      "step": 837
+    },
+    {
+      "epoch": 0.4962984897838318,
+      "grad_norm": 2.7262844772305868,
+      "learning_rate": 1.9052840239822218e-05,
+      "loss": 0.8535,
+      "step": 838
+    },
+    {
+      "epoch": 0.4968907314184187,
+      "grad_norm": 5.363131725816022,
+      "learning_rate": 1.9050121429996682e-05,
+      "loss": 0.8659,
+      "step": 839
+    },
+    {
+      "epoch": 0.49748297305300565,
+      "grad_norm": 8.968200408901346,
+      "learning_rate": 1.904739891816257e-05,
+      "loss": 0.8318,
+      "step": 840
+    },
+    {
+      "epoch": 0.4980752146875925,
+      "grad_norm": 1.7248247185116594,
+      "learning_rate": 1.904467270543354e-05,
+      "loss": 0.8957,
+      "step": 841
+    },
+    {
+      "epoch": 0.49866745632217946,
+      "grad_norm": 1.9562698832994654,
+      "learning_rate": 1.9041942792924767e-05,
+      "loss": 0.8345,
+      "step": 842
+    },
+    {
+      "epoch": 0.49925969795676634,
+      "grad_norm": 3.9947226743168156,
+      "learning_rate": 1.9039209181752942e-05,
+      "loss": 0.8652,
+      "step": 843
+    },
+    {
+      "epoch": 0.49985193959135327,
+      "grad_norm": 3.1571777018867677,
+      "learning_rate": 1.903647187303626e-05,
+      "loss": 0.8515,
+      "step": 844
+    },
+    {
+      "epoch": 0.5004441812259401,
+      "grad_norm": 2.1169391797594668,
+      "learning_rate": 1.9033730867894436e-05,
+      "loss": 0.8119,
+      "step": 845
+    },
+    {
+      "epoch": 0.5010364228605271,
+      "grad_norm": 2.7287026350418806,
+      "learning_rate": 1.9030986167448696e-05,
+      "loss": 0.9056,
+      "step": 846
+    },
+    {
+      "epoch": 0.501628664495114,
+      "grad_norm": 4.03173653799457,
+      "learning_rate": 1.9028237772821778e-05,
+      "loss": 0.835,
+      "step": 847
+    },
+    {
+      "epoch": 0.5022209061297009,
+      "grad_norm": 3.2102449210627246,
+      "learning_rate": 1.902548568513793e-05,
+      "loss": 0.869,
+      "step": 848
+    },
+    {
+      "epoch": 0.5028131477642879,
+      "grad_norm": 2.243822048790965,
+      "learning_rate": 1.9022729905522906e-05,
+      "loss": 0.8895,
+      "step": 849
+    },
+    {
+      "epoch": 0.5034053893988747,
+      "grad_norm": 2.4533940450724314,
+      "learning_rate": 1.9019970435103978e-05,
+      "loss": 0.8636,
+      "step": 850
+    },
+    {
+      "epoch": 0.5039976310334616,
+      "grad_norm": 2.848746378952194,
+      "learning_rate": 1.9017207275009925e-05,
+      "loss": 0.892,
+      "step": 851
+    },
+    {
+      "epoch": 0.5045898726680486,
+      "grad_norm": 2.3326834597031105,
+      "learning_rate": 1.9014440426371034e-05,
+      "loss": 0.7968,
+      "step": 852
+    },
+    {
+      "epoch": 0.5051821143026355,
+      "grad_norm": 2.745519544350419,
+      "learning_rate": 1.9011669890319104e-05,
+      "loss": 0.8378,
+      "step": 853
+    },
+    {
+      "epoch": 0.5057743559372224,
+      "grad_norm": 2.3271640191288183,
+      "learning_rate": 1.9008895667987434e-05,
+      "loss": 0.9074,
+      "step": 854
+    },
+    {
+      "epoch": 0.5063665975718094,
+      "grad_norm": 2.06754274248316,
+      "learning_rate": 1.9006117760510846e-05,
+      "loss": 0.8939,
+      "step": 855
+    },
+    {
+      "epoch": 0.5069588392063962,
+      "grad_norm": 4.335317472534966,
+      "learning_rate": 1.9003336169025655e-05,
+      "loss": 0.8571,
+      "step": 856
+    },
+    {
+      "epoch": 0.5075510808409831,
+      "grad_norm": 2.494123585371266,
+      "learning_rate": 1.9000550894669686e-05,
+      "loss": 0.8623,
+      "step": 857
+    },
+    {
+      "epoch": 0.50814332247557,
+      "grad_norm": 2.8756951470707577,
+      "learning_rate": 1.8997761938582277e-05,
+      "loss": 0.8292,
+      "step": 858
+    },
+    {
+      "epoch": 0.508735564110157,
+      "grad_norm": 2.761198259777182,
+      "learning_rate": 1.8994969301904266e-05,
+      "loss": 0.868,
+      "step": 859
+    },
+    {
+      "epoch": 0.5093278057447439,
+      "grad_norm": 6.139448459726971,
+      "learning_rate": 1.8992172985778002e-05,
+      "loss": 0.8189,
+      "step": 860
+    },
+    {
+      "epoch": 0.5099200473793307,
+      "grad_norm": 2.3982620378137,
+      "learning_rate": 1.898937299134733e-05,
+      "loss": 0.8413,
+      "step": 861
+    },
+    {
+      "epoch": 0.5105122890139177,
+      "grad_norm": 2.4360272672092758,
+      "learning_rate": 1.8986569319757605e-05,
+      "loss": 0.8573,
+      "step": 862
+    },
+    {
+      "epoch": 0.5111045306485046,
+      "grad_norm": 2.978466515723497,
+      "learning_rate": 1.898376197215569e-05,
+      "loss": 0.8423,
+      "step": 863
+    },
+    {
+      "epoch": 0.5116967722830915,
+      "grad_norm": 3.3896178499592464,
+      "learning_rate": 1.8980950949689952e-05,
+      "loss": 0.8559,
+      "step": 864
+    },
+    {
+      "epoch": 0.5122890139176784,
+      "grad_norm": 3.3035441398858993,
+      "learning_rate": 1.8978136253510248e-05,
+      "loss": 0.8613,
+      "step": 865
+    },
+    {
+      "epoch": 0.5128812555522653,
+      "grad_norm": 2.8070122645483035,
+      "learning_rate": 1.897531788476795e-05,
+      "loss": 0.8941,
+      "step": 866
+    },
+    {
+      "epoch": 0.5134734971868522,
+      "grad_norm": 3.414743000811395,
+      "learning_rate": 1.8972495844615933e-05,
+      "loss": 0.8576,
+      "step": 867
+    },
+    {
+      "epoch": 0.5140657388214391,
+      "grad_norm": 4.129209851440052,
+      "learning_rate": 1.896967013420857e-05,
+      "loss": 0.8392,
+      "step": 868
+    },
+    {
+      "epoch": 0.5146579804560261,
+      "grad_norm": 2.7614250927203696,
+      "learning_rate": 1.896684075470173e-05,
+      "loss": 0.8044,
+      "step": 869
+    },
+    {
+      "epoch": 0.515250222090613,
+      "grad_norm": 5.7190450126340275,
+      "learning_rate": 1.896400770725279e-05,
+      "loss": 0.8782,
+      "step": 870
+    },
+    {
+      "epoch": 0.5158424637251999,
+      "grad_norm": 3.264689428254472,
+      "learning_rate": 1.896117099302063e-05,
+      "loss": 0.8781,
+      "step": 871
+    },
+    {
+      "epoch": 0.5164347053597867,
+      "grad_norm": 2.3052957981729123,
+      "learning_rate": 1.8958330613165622e-05,
+      "loss": 0.7944,
+      "step": 872
+    },
+    {
+      "epoch": 0.5170269469943737,
+      "grad_norm": 3.6741300739615745,
+      "learning_rate": 1.895548656884964e-05,
+      "loss": 0.846,
+      "step": 873
+    },
+    {
+      "epoch": 0.5176191886289606,
+      "grad_norm": 2.633170158045045,
+      "learning_rate": 1.8952638861236066e-05,
+      "loss": 0.837,
+      "step": 874
+    },
+    {
+      "epoch": 0.5182114302635475,
+      "grad_norm": 3.933048892917864,
+      "learning_rate": 1.894978749148976e-05,
+      "loss": 0.8227,
+      "step": 875
+    },
+    {
+      "epoch": 0.5188036718981345,
+      "grad_norm": 3.415883244982847,
+      "learning_rate": 1.8946932460777105e-05,
+      "loss": 0.8591,
+      "step": 876
+    },
+    {
+      "epoch": 0.5193959135327213,
+      "grad_norm": 2.1556681217465905,
+      "learning_rate": 1.8944073770265958e-05,
+      "loss": 0.8476,
+      "step": 877
+    },
+    {
+      "epoch": 0.5199881551673082,
+      "grad_norm": 3.3058563739280307,
+      "learning_rate": 1.894121142112569e-05,
+      "loss": 0.8595,
+      "step": 878
+    },
+    {
+      "epoch": 0.5205803968018952,
+      "grad_norm": 1.6701206007058242,
+      "learning_rate": 1.8938345414527165e-05,
+      "loss": 0.8418,
+      "step": 879
+    },
+    {
+      "epoch": 0.5211726384364821,
+      "grad_norm": 2.8666163842304786,
+      "learning_rate": 1.8935475751642736e-05,
+      "loss": 0.8302,
+      "step": 880
+    },
+    {
+      "epoch": 0.521764880071069,
+      "grad_norm": 2.448842057434439,
+      "learning_rate": 1.893260243364626e-05,
+      "loss": 0.8475,
+      "step": 881
+    },
+    {
+      "epoch": 0.522357121705656,
+      "grad_norm": 2.8805816085789355,
+      "learning_rate": 1.8929725461713083e-05,
+      "loss": 0.8595,
+      "step": 882
+    },
+    {
+      "epoch": 0.5229493633402428,
+      "grad_norm": 2.190759226441507,
+      "learning_rate": 1.892684483702005e-05,
+      "loss": 0.8447,
+      "step": 883
+    },
+    {
+      "epoch": 0.5235416049748297,
+      "grad_norm": 2.7950857557738464,
+      "learning_rate": 1.8923960560745495e-05,
+      "loss": 0.8345,
+      "step": 884
+    },
+    {
+      "epoch": 0.5241338466094166,
+      "grad_norm": 2.2234121325584213,
+      "learning_rate": 1.8921072634069255e-05,
+      "loss": 0.8267,
+      "step": 885
+    },
+    {
+      "epoch": 0.5247260882440036,
+      "grad_norm": 1.457185905309736,
+      "learning_rate": 1.891818105817265e-05,
+      "loss": 0.8214,
+      "step": 886
+    },
+    {
+      "epoch": 0.5253183298785905,
+      "grad_norm": 3.4034755252325515,
+      "learning_rate": 1.8915285834238498e-05,
+      "loss": 0.8626,
+      "step": 887
+    },
+    {
+      "epoch": 0.5259105715131773,
+      "grad_norm": 3.715757488653005,
+      "learning_rate": 1.891238696345111e-05,
+      "loss": 0.8602,
+      "step": 888
+    },
+    {
+      "epoch": 0.5265028131477643,
+      "grad_norm": 1.8540026574196378,
+      "learning_rate": 1.890948444699629e-05,
+      "loss": 0.8251,
+      "step": 889
+    },
+    {
+      "epoch": 0.5270950547823512,
+      "grad_norm": 2.1775940247813343,
+      "learning_rate": 1.8906578286061325e-05,
+      "loss": 0.8593,
+      "step": 890
+    },
+    {
+      "epoch": 0.5276872964169381,
+      "grad_norm": 2.6307649588163047,
+      "learning_rate": 1.8903668481834996e-05,
+      "loss": 0.8612,
+      "step": 891
+    },
+    {
+      "epoch": 0.528279538051525,
+      "grad_norm": 3.989337506531906,
+      "learning_rate": 1.890075503550758e-05,
+      "loss": 0.8391,
+      "step": 892
+    },
+    {
+      "epoch": 0.528871779686112,
+      "grad_norm": 8.95506433868868,
+      "learning_rate": 1.889783794827085e-05,
+      "loss": 0.8594,
+      "step": 893
+    },
+    {
+      "epoch": 0.5294640213206988,
+      "grad_norm": 2.6218182036375026,
+      "learning_rate": 1.8894917221318038e-05,
+      "loss": 0.8828,
+      "step": 894
+    },
+    {
+      "epoch": 0.5300562629552857,
+      "grad_norm": 1.8329700035472636,
+      "learning_rate": 1.8891992855843902e-05,
+      "loss": 0.8307,
+      "step": 895
+    },
+    {
+      "epoch": 0.5306485045898727,
+      "grad_norm": 2.011555405209645,
+      "learning_rate": 1.888906485304467e-05,
+      "loss": 0.8204,
+      "step": 896
+    },
+    {
+      "epoch": 0.5312407462244596,
+      "grad_norm": 2.8138211953978933,
+      "learning_rate": 1.8886133214118053e-05,
+      "loss": 0.8544,
+      "step": 897
+    },
+    {
+      "epoch": 0.5318329878590465,
+      "grad_norm": 2.7779050915507666,
+      "learning_rate": 1.888319794026326e-05,
+      "loss": 0.8776,
+      "step": 898
+    },
+    {
+      "epoch": 0.5324252294936334,
+      "grad_norm": 2.867628838160467,
+      "learning_rate": 1.8880259032680985e-05,
+      "loss": 0.8714,
+      "step": 899
+    },
+    {
+      "epoch": 0.5330174711282203,
+      "grad_norm": 2.645795842168525,
+      "learning_rate": 1.88773164925734e-05,
+      "loss": 0.8673,
+      "step": 900
+    },
+    {
+      "epoch": 0.5336097127628072,
+      "grad_norm": 1.988822701440226,
+      "learning_rate": 1.887437032114418e-05,
+      "loss": 0.8058,
+      "step": 901
+    },
+    {
+      "epoch": 0.5342019543973942,
+      "grad_norm": 1.6378154166647196,
+      "learning_rate": 1.887142051959847e-05,
+      "loss": 0.8464,
+      "step": 902
+    },
+    {
+      "epoch": 0.5347941960319811,
+      "grad_norm": 2.2058743781314036,
+      "learning_rate": 1.8868467089142893e-05,
+      "loss": 0.8751,
+      "step": 903
+    },
+    {
+      "epoch": 0.5353864376665679,
+      "grad_norm": 2.5282653299151527,
+      "learning_rate": 1.8865510030985588e-05,
+      "loss": 0.8621,
+      "step": 904
+    },
+    {
+      "epoch": 0.5359786793011548,
+      "grad_norm": 27.31034341693317,
+      "learning_rate": 1.8862549346336144e-05,
+      "loss": 0.8408,
+      "step": 905
+    },
+    {
+      "epoch": 0.5365709209357418,
+      "grad_norm": 2.8532802195059985,
+      "learning_rate": 1.8859585036405653e-05,
+      "loss": 0.8242,
+      "step": 906
+    },
+    {
+      "epoch": 0.5371631625703287,
+      "grad_norm": 2.0534557906487887,
+      "learning_rate": 1.8856617102406685e-05,
+      "loss": 0.8428,
+      "step": 907
+    },
+    {
+      "epoch": 0.5377554042049156,
+      "grad_norm": 1.5482048628272183,
+      "learning_rate": 1.885364554555329e-05,
+      "loss": 0.8626,
+      "step": 908
+    },
+    {
+      "epoch": 0.5383476458395026,
+      "grad_norm": 2.1575688293112507,
+      "learning_rate": 1.8850670367061003e-05,
+      "loss": 0.8558,
+      "step": 909
+    },
+    {
+      "epoch": 0.5389398874740894,
+      "grad_norm": 2.102806685634922,
+      "learning_rate": 1.884769156814684e-05,
+      "loss": 0.8687,
+      "step": 910
+    },
+    {
+      "epoch": 0.5395321291086763,
+      "grad_norm": 1.4463393941880638,
+      "learning_rate": 1.884470915002929e-05,
+      "loss": 0.8818,
+      "step": 911
+    },
+    {
+      "epoch": 0.5401243707432632,
+      "grad_norm": 1.4609406531497033,
+      "learning_rate": 1.884172311392834e-05,
+      "loss": 0.8548,
+      "step": 912
+    },
+    {
+      "epoch": 0.5407166123778502,
+      "grad_norm": 1.8329647817569432,
+      "learning_rate": 1.883873346106544e-05,
+      "loss": 0.858,
+      "step": 913
+    },
+    {
+      "epoch": 0.5413088540124371,
+      "grad_norm": 3.9152641529549856,
+      "learning_rate": 1.883574019266353e-05,
+      "loss": 0.8783,
+      "step": 914
+    },
+    {
+      "epoch": 0.5419010956470239,
+      "grad_norm": 1.4892888981468162,
+      "learning_rate": 1.8832743309947026e-05,
+      "loss": 0.8504,
+      "step": 915
+    },
+    {
+      "epoch": 0.5424933372816109,
+      "grad_norm": 2.1974930679295963,
+      "learning_rate": 1.8829742814141813e-05,
+      "loss": 0.7888,
+      "step": 916
+    },
+    {
+      "epoch": 0.5430855789161978,
+      "grad_norm": 2.960377295236711,
+      "learning_rate": 1.8826738706475275e-05,
+      "loss": 0.9234,
+      "step": 917
+    },
+    {
+      "epoch": 0.5436778205507847,
+      "grad_norm": 2.413195078205313,
+      "learning_rate": 1.882373098817625e-05,
+      "loss": 0.8354,
+      "step": 918
+    },
+    {
+      "epoch": 0.5442700621853717,
+      "grad_norm": 2.29920506736285,
+      "learning_rate": 1.882071966047507e-05,
+      "loss": 0.8359,
+      "step": 919
+    },
+    {
+      "epoch": 0.5448623038199586,
+      "grad_norm": 2.2127645514863916,
+      "learning_rate": 1.8817704724603536e-05,
+      "loss": 0.8713,
+      "step": 920
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 2.5056883347384074,
+      "learning_rate": 1.8814686181794927e-05,
+      "loss": 0.8836,
+      "step": 921
+    },
+    {
+      "epoch": 0.5460467870891323,
+      "grad_norm": 2.319729335193821,
+      "learning_rate": 1.8811664033283993e-05,
+      "loss": 0.8758,
+      "step": 922
+    },
+    {
+      "epoch": 0.5466390287237193,
+      "grad_norm": 2.4646633874772603,
+      "learning_rate": 1.880863828030697e-05,
+      "loss": 0.8338,
+      "step": 923
+    },
+    {
+      "epoch": 0.5472312703583062,
+      "grad_norm": 3.5967734066777624,
+      "learning_rate": 1.880560892410155e-05,
+      "loss": 0.8142,
+      "step": 924
+    },
+    {
+      "epoch": 0.5478235119928931,
+      "grad_norm": 1.931083687931415,
+      "learning_rate": 1.8802575965906923e-05,
+      "loss": 0.8624,
+      "step": 925
+    },
+    {
+      "epoch": 0.54841575362748,
+      "grad_norm": 1.3658707686323024,
+      "learning_rate": 1.879953940696373e-05,
+      "loss": 0.82,
+      "step": 926
+    },
+    {
+      "epoch": 0.5490079952620669,
+      "grad_norm": 1.5399776635493545,
+      "learning_rate": 1.87964992485141e-05,
+      "loss": 0.8752,
+      "step": 927
+    },
+    {
+      "epoch": 0.5496002368966538,
+      "grad_norm": 1.2888985969973952,
+      "learning_rate": 1.8793455491801623e-05,
+      "loss": 0.8258,
+      "step": 928
+    },
+    {
+      "epoch": 0.5501924785312408,
+      "grad_norm": 1.2235740279377274,
+      "learning_rate": 1.8790408138071372e-05,
+      "loss": 0.7807,
+      "step": 929
+    },
+    {
+      "epoch": 0.5507847201658277,
+      "grad_norm": 1.3948987520636114,
+      "learning_rate": 1.878735718856988e-05,
+      "loss": 0.85,
+      "step": 930
+    },
+    {
+      "epoch": 0.5513769618004146,
+      "grad_norm": 1.1530555013033301,
+      "learning_rate": 1.8784302644545165e-05,
+      "loss": 0.8434,
+      "step": 931
+    },
+    {
+      "epoch": 0.5519692034350014,
+      "grad_norm": 1.1001632567365147,
+      "learning_rate": 1.8781244507246706e-05,
+      "loss": 0.8357,
+      "step": 932
+    },
+    {
+      "epoch": 0.5525614450695884,
+      "grad_norm": 1.1272615396567933,
+      "learning_rate": 1.8778182777925447e-05,
+      "loss": 0.8673,
+      "step": 933
+    },
+    {
+      "epoch": 0.5531536867041753,
+      "grad_norm": 1.5817714863387373,
+      "learning_rate": 1.877511745783381e-05,
+      "loss": 0.8628,
+      "step": 934
+    },
+    {
+      "epoch": 0.5537459283387622,
+      "grad_norm": 1.5329765777416222,
+      "learning_rate": 1.8772048548225684e-05,
+      "loss": 0.8412,
+      "step": 935
+    },
+    {
+      "epoch": 0.5543381699733492,
+      "grad_norm": 1.15668478337925,
+      "learning_rate": 1.8768976050356428e-05,
+      "loss": 0.8673,
+      "step": 936
+    },
+    {
+      "epoch": 0.554930411607936,
+      "grad_norm": 1.4034733165443383,
+      "learning_rate": 1.8765899965482858e-05,
+      "loss": 0.8063,
+      "step": 937
+    },
+    {
+      "epoch": 0.5555226532425229,
+      "grad_norm": 1.278219411060756,
+      "learning_rate": 1.876282029486328e-05,
+      "loss": 0.8432,
+      "step": 938
+    },
+    {
+      "epoch": 0.5561148948771099,
+      "grad_norm": 1.434098032436352,
+      "learning_rate": 1.8759737039757437e-05,
+      "loss": 0.8616,
+      "step": 939
+    },
+    {
+      "epoch": 0.5567071365116968,
+      "grad_norm": 1.3706128629162078,
+      "learning_rate": 1.8756650201426565e-05,
+      "loss": 0.8397,
+      "step": 940
+    },
+    {
+      "epoch": 0.5572993781462837,
+      "grad_norm": 1.4524914917856544,
+      "learning_rate": 1.875355978113335e-05,
+      "loss": 0.8423,
+      "step": 941
+    },
+    {
+      "epoch": 0.5578916197808705,
+      "grad_norm": 1.6212056588454582,
+      "learning_rate": 1.8750465780141946e-05,
+      "loss": 0.8259,
+      "step": 942
+    },
+    {
+      "epoch": 0.5584838614154575,
+      "grad_norm": 1.5764324461444237,
+      "learning_rate": 1.8747368199717977e-05,
+      "loss": 0.8723,
+      "step": 943
+    },
+    {
+      "epoch": 0.5590761030500444,
+      "grad_norm": 1.5399903048769072,
+      "learning_rate": 1.8744267041128528e-05,
+      "loss": 0.8563,
+      "step": 944
+    },
+    {
+      "epoch": 0.5596683446846313,
+      "grad_norm": 1.4178813586778414,
+      "learning_rate": 1.874116230564214e-05,
+      "loss": 0.8551,
+      "step": 945
+    },
+    {
+      "epoch": 0.5602605863192183,
+      "grad_norm": 1.2554205520730095,
+      "learning_rate": 1.8738053994528835e-05,
+      "loss": 0.84,
+      "step": 946
+    },
+    {
+      "epoch": 0.5608528279538052,
+      "grad_norm": 1.436381834061602,
+      "learning_rate": 1.873494210906008e-05,
+      "loss": 0.8633,
+      "step": 947
+    },
+    {
+      "epoch": 0.561445069588392,
+      "grad_norm": 1.572609783293438,
+      "learning_rate": 1.8731826650508812e-05,
+      "loss": 0.8575,
+      "step": 948
+    },
+    {
+      "epoch": 0.562037311222979,
+      "grad_norm": 1.9911513522446955,
+      "learning_rate": 1.872870762014943e-05,
+      "loss": 0.8699,
+      "step": 949
+    },
+    {
+      "epoch": 0.5626295528575659,
+      "grad_norm": 1.7221797946035173,
+      "learning_rate": 1.8725585019257794e-05,
+      "loss": 0.8288,
+      "step": 950
+    },
+    {
+      "epoch": 0.5632217944921528,
+      "grad_norm": 1.6316667513165497,
+      "learning_rate": 1.8722458849111218e-05,
+      "loss": 0.8476,
+      "step": 951
+    },
+    {
+      "epoch": 0.5638140361267397,
+      "grad_norm": 1.8817617861124867,
+      "learning_rate": 1.8719329110988487e-05,
+      "loss": 0.8812,
+      "step": 952
+    },
+    {
+      "epoch": 0.5644062777613266,
+      "grad_norm": 1.225383750075909,
+      "learning_rate": 1.871619580616984e-05,
+      "loss": 0.8646,
+      "step": 953
+    },
+    {
+      "epoch": 0.5649985193959135,
+      "grad_norm": 4.355740946012875,
+      "learning_rate": 1.871305893593697e-05,
+      "loss": 0.8008,
+      "step": 954
+    },
+    {
+      "epoch": 0.5655907610305004,
+      "grad_norm": 2.1927471549874564,
+      "learning_rate": 1.8709918501573038e-05,
+      "loss": 0.8762,
+      "step": 955
+    },
+    {
+      "epoch": 0.5661830026650874,
+      "grad_norm": 3.055980708697755,
+      "learning_rate": 1.8706774504362655e-05,
+      "loss": 0.8584,
+      "step": 956
+    },
+    {
+      "epoch": 0.5667752442996743,
+      "grad_norm": 1.5384769512510463,
+      "learning_rate": 1.8703626945591895e-05,
+      "loss": 0.874,
+      "step": 957
+    },
+    {
+      "epoch": 0.5673674859342612,
+      "grad_norm": 1.593507898135259,
+      "learning_rate": 1.8700475826548285e-05,
+      "loss": 0.8539,
+      "step": 958
+    },
+    {
+      "epoch": 0.567959727568848,
+      "grad_norm": 2.4658738412503185,
+      "learning_rate": 1.8697321148520812e-05,
+      "loss": 0.8464,
+      "step": 959
+    },
+    {
+      "epoch": 0.568551969203435,
+      "grad_norm": 2.3647390172875213,
+      "learning_rate": 1.8694162912799917e-05,
+      "loss": 0.8487,
+      "step": 960
+    },
+    {
+      "epoch": 0.5691442108380219,
+      "grad_norm": 1.8822227155497309,
+      "learning_rate": 1.869100112067749e-05,
+      "loss": 0.924,
+      "step": 961
+    },
+    {
+      "epoch": 0.5697364524726088,
+      "grad_norm": 3.1476849757648258,
+      "learning_rate": 1.868783577344689e-05,
+      "loss": 0.8641,
+      "step": 962
+    },
+    {
+      "epoch": 0.5703286941071958,
+      "grad_norm": 2.391640628039692,
+      "learning_rate": 1.8684666872402913e-05,
+      "loss": 0.8189,
+      "step": 963
+    },
+    {
+      "epoch": 0.5709209357417826,
+      "grad_norm": 1.9209124622082125,
+      "learning_rate": 1.8681494418841825e-05,
+      "loss": 0.9141,
+      "step": 964
+    },
+    {
+      "epoch": 0.5715131773763695,
+      "grad_norm": 1.883444300120634,
+      "learning_rate": 1.8678318414061336e-05,
+      "loss": 0.8244,
+      "step": 965
+    },
+    {
+      "epoch": 0.5721054190109565,
+      "grad_norm": 1.8683731975284492,
+      "learning_rate": 1.867513885936061e-05,
+      "loss": 0.8039,
+      "step": 966
+    },
+    {
+      "epoch": 0.5726976606455434,
+      "grad_norm": 2.186169687223295,
+      "learning_rate": 1.8671955756040264e-05,
+      "loss": 0.8481,
+      "step": 967
+    },
+    {
+      "epoch": 0.5732899022801303,
+      "grad_norm": 2.4126100375538746,
+      "learning_rate": 1.8668769105402366e-05,
+      "loss": 0.8283,
+      "step": 968
+    },
+    {
+      "epoch": 0.5738821439147173,
+      "grad_norm": 5.881790876957725,
+      "learning_rate": 1.8665578908750437e-05,
+      "loss": 0.8004,
+      "step": 969
+    },
+    {
+      "epoch": 0.5744743855493041,
+      "grad_norm": 2.1957974263362963,
+      "learning_rate": 1.8662385167389443e-05,
+      "loss": 0.8125,
+      "step": 970
+    },
+    {
+      "epoch": 0.575066627183891,
+      "grad_norm": 2.122100726977036,
+      "learning_rate": 1.8659187882625807e-05,
+      "loss": 0.8323,
+      "step": 971
+    },
+    {
+      "epoch": 0.5756588688184779,
+      "grad_norm": 2.0089182722346512,
+      "learning_rate": 1.8655987055767396e-05,
+      "loss": 0.8107,
+      "step": 972
+    },
+    {
+      "epoch": 0.5762511104530649,
+      "grad_norm": 3.4788064271268677,
+      "learning_rate": 1.8652782688123535e-05,
+      "loss": 0.8357,
+      "step": 973
+    },
+    {
+      "epoch": 0.5768433520876518,
+      "grad_norm": 1.9971681598642872,
+      "learning_rate": 1.864957478100498e-05,
+      "loss": 0.8487,
+      "step": 974
+    },
+    {
+      "epoch": 0.5774355937222386,
+      "grad_norm": 2.7963971479792096,
+      "learning_rate": 1.8646363335723952e-05,
+      "loss": 0.9236,
+      "step": 975
+    },
+    {
+      "epoch": 0.5780278353568256,
+      "grad_norm": 2.6501420919148706,
+      "learning_rate": 1.864314835359411e-05,
+      "loss": 0.8643,
+      "step": 976
+    },
+    {
+      "epoch": 0.5786200769914125,
+      "grad_norm": 2.5379734296943273,
+      "learning_rate": 1.863992983593056e-05,
+      "loss": 0.8564,
+      "step": 977
+    },
+    {
+      "epoch": 0.5792123186259994,
+      "grad_norm": 3.4686271540396936,
+      "learning_rate": 1.8636707784049867e-05,
+      "loss": 0.8665,
+      "step": 978
+    },
+    {
+      "epoch": 0.5798045602605864,
+      "grad_norm": 1.9639248298320788,
+      "learning_rate": 1.8633482199270025e-05,
+      "loss": 0.8785,
+      "step": 979
+    },
+    {
+      "epoch": 0.5803968018951732,
+      "grad_norm": 3.009335654283905,
+      "learning_rate": 1.8630253082910473e-05,
+      "loss": 0.8629,
+      "step": 980
+    },
+    {
+      "epoch": 0.5809890435297601,
+      "grad_norm": 2.5913527017658535,
+      "learning_rate": 1.862702043629211e-05,
+      "loss": 0.8483,
+      "step": 981
+    },
+    {
+      "epoch": 0.581581285164347,
+      "grad_norm": 2.3301173927122636,
+      "learning_rate": 1.8623784260737272e-05,
+      "loss": 0.8872,
+      "step": 982
+    },
+    {
+      "epoch": 0.582173526798934,
+      "grad_norm": 2.1728733813785954,
+      "learning_rate": 1.862054455756973e-05,
+      "loss": 0.8851,
+      "step": 983
+    },
+    {
+      "epoch": 0.5827657684335209,
+      "grad_norm": 2.6022654651479153,
+      "learning_rate": 1.8617301328114704e-05,
+      "loss": 0.853,
+      "step": 984
+    },
+    {
+      "epoch": 0.5833580100681078,
+      "grad_norm": 5.499406695169115,
+      "learning_rate": 1.8614054573698867e-05,
+      "loss": 0.8533,
+      "step": 985
+    },
+    {
+      "epoch": 0.5839502517026947,
+      "grad_norm": 7.358085625855621,
+      "learning_rate": 1.861080429565031e-05,
+      "loss": 0.8609,
+      "step": 986
+    },
+    {
+      "epoch": 0.5845424933372816,
+      "grad_norm": 2.5901934912509406,
+      "learning_rate": 1.8607550495298594e-05,
+      "loss": 0.8123,
+      "step": 987
+    },
+    {
+      "epoch": 0.5851347349718685,
+      "grad_norm": 3.577130582550033,
+      "learning_rate": 1.8604293173974694e-05,
+      "loss": 0.8424,
+      "step": 988
+    },
+    {
+      "epoch": 0.5857269766064555,
+      "grad_norm": 6.266496305800451,
+      "learning_rate": 1.8601032333011045e-05,
+      "loss": 0.8278,
+      "step": 989
+    },
+    {
+      "epoch": 0.5863192182410424,
+      "grad_norm": 3.688636492554437,
+      "learning_rate": 1.8597767973741514e-05,
+      "loss": 0.8391,
+      "step": 990
+    },
+    {
+      "epoch": 0.5869114598756292,
+      "grad_norm": 3.4278947346345556,
+      "learning_rate": 1.8594500097501403e-05,
+      "loss": 0.8925,
+      "step": 991
+    },
+    {
+      "epoch": 0.5875037015102161,
+      "grad_norm": 2.798245409340499,
+      "learning_rate": 1.8591228705627464e-05,
+      "loss": 0.8683,
+      "step": 992
+    },
+    {
+      "epoch": 0.5880959431448031,
+      "grad_norm": 2.6101217239588674,
+      "learning_rate": 1.858795379945787e-05,
+      "loss": 0.7957,
+      "step": 993
+    },
+    {
+      "epoch": 0.58868818477939,
+      "grad_norm": 1.9767054625050304,
+      "learning_rate": 1.8584675380332254e-05,
+      "loss": 0.8868,
+      "step": 994
+    },
+    {
+      "epoch": 0.5892804264139769,
+      "grad_norm": 3.561966477734429,
+      "learning_rate": 1.8581393449591667e-05,
+      "loss": 0.8462,
+      "step": 995
+    },
+    {
+      "epoch": 0.5898726680485639,
+      "grad_norm": 1.6872502667794813,
+      "learning_rate": 1.8578108008578603e-05,
+      "loss": 0.8334,
+      "step": 996
+    },
+    {
+      "epoch": 0.5904649096831507,
+      "grad_norm": 2.4287274023717615,
+      "learning_rate": 1.8574819058636993e-05,
+      "loss": 0.8921,
+      "step": 997
+    },
+    {
+      "epoch": 0.5910571513177376,
+      "grad_norm": 1.7111015618500356,
+      "learning_rate": 1.8571526601112202e-05,
+      "loss": 0.8314,
+      "step": 998
+    },
+    {
+      "epoch": 0.5916493929523245,
+      "grad_norm": 2.5502109328589233,
+      "learning_rate": 1.856823063735103e-05,
+      "loss": 0.8725,
+      "step": 999
+    },
+    {
+      "epoch": 0.5922416345869115,
+      "grad_norm": 2.040139228442696,
+      "learning_rate": 1.8564931168701713e-05,
+      "loss": 0.8347,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5928338762214984,
+      "grad_norm": 2.5055813610082303,
+      "learning_rate": 1.8561628196513923e-05,
+      "loss": 0.8585,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5934261178560852,
+      "grad_norm": 2.1524220648045427,
+      "learning_rate": 1.855832172213875e-05,
+      "loss": 0.8474,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5940183594906722,
+      "grad_norm": 1.619731922084106,
+      "learning_rate": 1.8555011746928738e-05,
+      "loss": 0.8179,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5946106011252591,
+      "grad_norm": 2.8662321460118805,
+      "learning_rate": 1.855169827223785e-05,
+      "loss": 0.8484,
+      "step": 1004
+    },
+    {
+      "epoch": 0.595202842759846,
+      "grad_norm": 2.449255266934718,
+      "learning_rate": 1.8548381299421486e-05,
+      "loss": 0.8856,
+      "step": 1005
+    },
+    {
+      "epoch": 0.595795084394433,
+      "grad_norm": 1.8813446878490585,
+      "learning_rate": 1.854506082983647e-05,
+      "loss": 0.8173,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5963873260290199,
+      "grad_norm": 2.5469931424576586,
+      "learning_rate": 1.8541736864841064e-05,
+      "loss": 0.8532,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5969795676636067,
+      "grad_norm": 2.641308461614614,
+      "learning_rate": 1.8538409405794952e-05,
+      "loss": 0.8217,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5975718092981936,
+      "grad_norm": 1.4950837736543847,
+      "learning_rate": 1.8535078454059256e-05,
+      "loss": 0.8381,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5981640509327806,
+      "grad_norm": 2.338715560487363,
+      "learning_rate": 1.8531744010996525e-05,
+      "loss": 0.8573,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5987562925673675,
+      "grad_norm": 1.6059242683751407,
+      "learning_rate": 1.8528406077970725e-05,
+      "loss": 0.8527,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5993485342019544,
+      "grad_norm": 12.361811261173756,
+      "learning_rate": 1.8525064656347265e-05,
+      "loss": 0.8298,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5999407758365413,
+      "grad_norm": 1.1511502155917168,
+      "learning_rate": 1.8521719747492974e-05,
+      "loss": 0.8212,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6005330174711282,
+      "grad_norm": 1.6632441653678172,
+      "learning_rate": 1.8518371352776107e-05,
+      "loss": 0.8415,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6011252591057151,
+      "grad_norm": 4.116651414311895,
+      "learning_rate": 1.8515019473566346e-05,
+      "loss": 0.8252,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6017175007403021,
+      "grad_norm": 1.8732938641491954,
+      "learning_rate": 1.85116641112348e-05,
+      "loss": 0.8283,
+      "step": 1016
+    },
+    {
+      "epoch": 0.602309742374889,
+      "grad_norm": 1.8390615077922212,
+      "learning_rate": 1.8508305267153992e-05,
+      "loss": 0.8132,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6029019840094758,
+      "grad_norm": 2.633892399081572,
+      "learning_rate": 1.850494294269789e-05,
+      "loss": 0.8758,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6034942256440627,
+      "grad_norm": 2.3622253842023335,
+      "learning_rate": 1.8501577139241865e-05,
+      "loss": 0.829,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6040864672786497,
+      "grad_norm": 2.037756142332905,
+      "learning_rate": 1.8498207858162724e-05,
+      "loss": 0.867,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6046787089132366,
+      "grad_norm": 4.528232825533869,
+      "learning_rate": 1.8494835100838693e-05,
+      "loss": 0.8648,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6052709505478235,
+      "grad_norm": 2.8907054221222,
+      "learning_rate": 1.8491458868649417e-05,
+      "loss": 0.8574,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6058631921824105,
+      "grad_norm": 2.182754237899523,
+      "learning_rate": 1.8488079162975965e-05,
+      "loss": 0.8716,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6064554338169973,
+      "grad_norm": 3.6488297829005316,
+      "learning_rate": 1.8484695985200832e-05,
+      "loss": 0.8888,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6070476754515842,
+      "grad_norm": 2.1360401343086375,
+      "learning_rate": 1.848130933670792e-05,
+      "loss": 0.853,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6076399170861712,
+      "grad_norm": 2.7634831832013074,
+      "learning_rate": 1.847791921888256e-05,
+      "loss": 0.8444,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6082321587207581,
+      "grad_norm": 2.548946934028294,
+      "learning_rate": 1.847452563311151e-05,
+      "loss": 0.8552,
+      "step": 1027
+    },
+    {
+      "epoch": 0.608824400355345,
+      "grad_norm": 2.00472769722451,
+      "learning_rate": 1.8471128580782923e-05,
+      "loss": 0.7907,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6094166419899318,
+      "grad_norm": 7.112058400031312,
+      "learning_rate": 1.8467728063286398e-05,
+      "loss": 0.811,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6100088836245188,
+      "grad_norm": 5.904134501716097,
+      "learning_rate": 1.8464324082012926e-05,
+      "loss": 0.8236,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6106011252591057,
+      "grad_norm": 2.677789919762913,
+      "learning_rate": 1.8460916638354934e-05,
+      "loss": 0.8855,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6111933668936926,
+      "grad_norm": 2.979499541128458,
+      "learning_rate": 1.845750573370626e-05,
+      "loss": 0.81,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6117856085282796,
+      "grad_norm": 4.395198347599419,
+      "learning_rate": 1.845409136946215e-05,
+      "loss": 0.8649,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6123778501628665,
+      "grad_norm": 1.9418555795640626,
+      "learning_rate": 1.8450673547019273e-05,
+      "loss": 0.8106,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6129700917974533,
+      "grad_norm": 2.50689095850713,
+      "learning_rate": 1.844725226777571e-05,
+      "loss": 0.8387,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6135623334320403,
+      "grad_norm": 2.812993440172385,
+      "learning_rate": 1.844382753313096e-05,
+      "loss": 0.8102,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6141545750666272,
+      "grad_norm": 2.6604257345575206,
+      "learning_rate": 1.844039934448593e-05,
+      "loss": 0.8277,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6147468167012141,
+      "grad_norm": 2.561145533436182,
+      "learning_rate": 1.8436967703242938e-05,
+      "loss": 0.8557,
+      "step": 1038
+    },
+    {
+      "epoch": 0.615339058335801,
+      "grad_norm": 3.2075888751958,
+      "learning_rate": 1.8433532610805724e-05,
+      "loss": 0.8399,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6159312999703879,
+      "grad_norm": 7.7990669355472475,
+      "learning_rate": 1.843009406857943e-05,
+      "loss": 0.8473,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6165235416049748,
+      "grad_norm": 3.132725554622429,
+      "learning_rate": 1.8426652077970616e-05,
+      "loss": 0.8058,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6171157832395617,
+      "grad_norm": 2.1854341940737187,
+      "learning_rate": 1.842320664038725e-05,
+      "loss": 0.8126,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6177080248741487,
+      "grad_norm": 2.9704427255656825,
+      "learning_rate": 1.841975775723871e-05,
+      "loss": 0.8319,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6183002665087356,
+      "grad_norm": 2.2725465194986367,
+      "learning_rate": 1.8416305429935776e-05,
+      "loss": 0.8729,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6188925081433225,
+      "grad_norm": 2.5307221134117674,
+      "learning_rate": 1.8412849659890652e-05,
+      "loss": 0.8412,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6194847497779093,
+      "grad_norm": 1.810981613810124,
+      "learning_rate": 1.840939044851694e-05,
+      "loss": 0.8673,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6200769914124963,
+      "grad_norm": 2.4443325681183907,
+      "learning_rate": 1.840592779722965e-05,
+      "loss": 0.7945,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6206692330470832,
+      "grad_norm": 2.8376072263141485,
+      "learning_rate": 1.8402461707445206e-05,
+      "loss": 0.8427,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6212614746816701,
+      "grad_norm": 3.6210685733471397,
+      "learning_rate": 1.8398992180581427e-05,
+      "loss": 0.826,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6218537163162571,
+      "grad_norm": 1.561394208799311,
+      "learning_rate": 1.839551921805755e-05,
+      "loss": 0.8572,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6224459579508439,
+      "grad_norm": 3.103802020893097,
+      "learning_rate": 1.839204282129421e-05,
+      "loss": 0.8646,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6230381995854308,
+      "grad_norm": 4.425061232798762,
+      "learning_rate": 1.8388562991713447e-05,
+      "loss": 0.8143,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6236304412200178,
+      "grad_norm": 2.191428810992078,
+      "learning_rate": 1.838507973073871e-05,
+      "loss": 0.8502,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6242226828546047,
+      "grad_norm": 4.95011030076338,
+      "learning_rate": 1.8381593039794846e-05,
+      "loss": 0.864,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6248149244891916,
+      "grad_norm": 2.7797880624241516,
+      "learning_rate": 1.837810292030811e-05,
+      "loss": 0.859,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6254071661237784,
+      "grad_norm": 4.869981957929177,
+      "learning_rate": 1.8374609373706156e-05,
+      "loss": 0.8444,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6259994077583654,
+      "grad_norm": 2.5746569296165522,
+      "learning_rate": 1.8371112401418042e-05,
+      "loss": 0.8364,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6265916493929523,
+      "grad_norm": 2.0625606665572653,
+      "learning_rate": 1.8367612004874224e-05,
+      "loss": 0.8906,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6271838910275392,
+      "grad_norm": 2.676271947186104,
+      "learning_rate": 1.8364108185506563e-05,
+      "loss": 0.8459,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6277761326621262,
+      "grad_norm": 1.5883481724433661,
+      "learning_rate": 1.8360600944748316e-05,
+      "loss": 0.7969,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6283683742967131,
+      "grad_norm": 1.695724970760026,
+      "learning_rate": 1.8357090284034145e-05,
+      "loss": 0.8488,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6289606159312999,
+      "grad_norm": 5.99104068302577,
+      "learning_rate": 1.8353576204800106e-05,
+      "loss": 0.8584,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6295528575658869,
+      "grad_norm": 1.4158669500557108,
+      "learning_rate": 1.8350058708483654e-05,
+      "loss": 0.8109,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6301450992004738,
+      "grad_norm": 4.619972925464011,
+      "learning_rate": 1.8346537796523643e-05,
+      "loss": 0.8387,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6307373408350607,
+      "grad_norm": 1.840623636179062,
+      "learning_rate": 1.834301347036033e-05,
+      "loss": 0.7773,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6313295824696477,
+      "grad_norm": 3.0573832313039073,
+      "learning_rate": 1.833948573143535e-05,
+      "loss": 0.8626,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6319218241042345,
+      "grad_norm": 2.518622510039811,
+      "learning_rate": 1.8335954581191758e-05,
+      "loss": 0.8783,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6325140657388214,
+      "grad_norm": 9.493928447056057,
+      "learning_rate": 1.8332420021073992e-05,
+      "loss": 0.8544,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6331063073734083,
+      "grad_norm": 1.720905288086446,
+      "learning_rate": 1.8328882052527878e-05,
+      "loss": 0.8282,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6336985490079953,
+      "grad_norm": 1.8020645303459293,
+      "learning_rate": 1.8325340677000646e-05,
+      "loss": 0.8923,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6342907906425822,
+      "grad_norm": 2.8749501095196517,
+      "learning_rate": 1.8321795895940925e-05,
+      "loss": 0.8544,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6348830322771691,
+      "grad_norm": 2.008670621988076,
+      "learning_rate": 1.8318247710798728e-05,
+      "loss": 0.8462,
+      "step": 1072
+    },
+    {
+      "epoch": 0.635475273911756,
+      "grad_norm": 1.8563596582266815,
+      "learning_rate": 1.8314696123025456e-05,
+      "loss": 0.8324,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6360675155463429,
+      "grad_norm": 2.228469119107762,
+      "learning_rate": 1.831114113407391e-05,
+      "loss": 0.8325,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6366597571809298,
+      "grad_norm": 2.5028199204197596,
+      "learning_rate": 1.8307582745398282e-05,
+      "loss": 0.7938,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6372519988155168,
+      "grad_norm": 2.563042207855269,
+      "learning_rate": 1.8304020958454156e-05,
+      "loss": 0.8539,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6378442404501037,
+      "grad_norm": 2.748425207975687,
+      "learning_rate": 1.83004557746985e-05,
+      "loss": 0.9026,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6384364820846905,
+      "grad_norm": 2.198078366256531,
+      "learning_rate": 1.8296887195589678e-05,
+      "loss": 0.836,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6390287237192774,
+      "grad_norm": 2.978264957639652,
+      "learning_rate": 1.829331522258743e-05,
+      "loss": 0.8374,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6396209653538644,
+      "grad_norm": 3.0902393566945694,
+      "learning_rate": 1.8289739857152903e-05,
+      "loss": 0.8522,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6402132069884513,
+      "grad_norm": 2.5510293746898793,
+      "learning_rate": 1.828616110074862e-05,
+      "loss": 0.8959,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6408054486230382,
+      "grad_norm": 2.4920369060953,
+      "learning_rate": 1.8282578954838493e-05,
+      "loss": 0.8635,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6413976902576252,
+      "grad_norm": 3.8584939608718174,
+      "learning_rate": 1.8278993420887822e-05,
+      "loss": 0.897,
+      "step": 1083
+    },
+    {
+      "epoch": 0.641989931892212,
+      "grad_norm": 1.6943535494324076,
+      "learning_rate": 1.8275404500363293e-05,
+      "loss": 0.8555,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6425821735267989,
+      "grad_norm": 3.625435056497657,
+      "learning_rate": 1.8271812194732972e-05,
+      "loss": 0.8555,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6431744151613858,
+      "grad_norm": 9.548040429248383,
+      "learning_rate": 1.8268216505466318e-05,
+      "loss": 0.9243,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6437666567959728,
+      "grad_norm": 6.204666207240246,
+      "learning_rate": 1.8264617434034168e-05,
+      "loss": 0.8679,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6443588984305597,
+      "grad_norm": 2.3034334204515767,
+      "learning_rate": 1.826101498190875e-05,
+      "loss": 0.8646,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6449511400651465,
+      "grad_norm": 4.514126342622965,
+      "learning_rate": 1.825740915056366e-05,
+      "loss": 0.8552,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6455433816997335,
+      "grad_norm": 2.4208151622387604,
+      "learning_rate": 1.8253799941473894e-05,
+      "loss": 0.8319,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6461356233343204,
+      "grad_norm": 3.3019190869678736,
+      "learning_rate": 1.8250187356115817e-05,
+      "loss": 0.8964,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6467278649689073,
+      "grad_norm": 6.537722710331606,
+      "learning_rate": 1.824657139596718e-05,
+      "loss": 0.8184,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6473201066034943,
+      "grad_norm": 5.245807404499727,
+      "learning_rate": 1.8242952062507115e-05,
+      "loss": 0.8338,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6479123482380811,
+      "grad_norm": 3.8290099537018167,
+      "learning_rate": 1.8239329357216135e-05,
+      "loss": 0.7772,
+      "step": 1094
+    },
+    {
+      "epoch": 0.648504589872668,
+      "grad_norm": 4.36239109924843,
+      "learning_rate": 1.8235703281576127e-05,
+      "loss": 0.8034,
+      "step": 1095
+    },
+    {
+      "epoch": 0.649096831507255,
+      "grad_norm": 3.266500591644286,
+      "learning_rate": 1.823207383707036e-05,
+      "loss": 0.8272,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6496890731418419,
+      "grad_norm": 1.9798959268550758,
+      "learning_rate": 1.822844102518348e-05,
+      "loss": 0.843,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6502813147764288,
+      "grad_norm": 2.4272575236741343,
+      "learning_rate": 1.8224804847401518e-05,
+      "loss": 0.8178,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6508735564110157,
+      "grad_norm": 2.643309325700988,
+      "learning_rate": 1.822116530521187e-05,
+      "loss": 0.8607,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6514657980456026,
+      "grad_norm": 2.7517207819440466,
+      "learning_rate": 1.821752240010331e-05,
+      "loss": 0.9101,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6520580396801895,
+      "grad_norm": 2.409184915147776,
+      "learning_rate": 1.8213876133565996e-05,
+      "loss": 0.894,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6526502813147764,
+      "grad_norm": 1.8447833000009395,
+      "learning_rate": 1.8210226507091454e-05,
+      "loss": 0.8607,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6532425229493634,
+      "grad_norm": 2.9713191095410174,
+      "learning_rate": 1.820657352217259e-05,
+      "loss": 0.834,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6538347645839503,
+      "grad_norm": 2.5876456677398956,
+      "learning_rate": 1.8202917180303673e-05,
+      "loss": 0.866,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6544270062185371,
+      "grad_norm": 2.3509085863918284,
+      "learning_rate": 1.8199257482980358e-05,
+      "loss": 0.8582,
+      "step": 1105
+    },
+    {
+      "epoch": 0.655019247853124,
+      "grad_norm": 1.5013962529594334,
+      "learning_rate": 1.819559443169967e-05,
+      "loss": 0.8376,
+      "step": 1106
+    },
+    {
+      "epoch": 0.655611489487711,
+      "grad_norm": 2.604248538807811,
+      "learning_rate": 1.8191928027959996e-05,
+      "loss": 0.8598,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6562037311222979,
+      "grad_norm": 1.8138214242680486,
+      "learning_rate": 1.8188258273261104e-05,
+      "loss": 0.8527,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6567959727568848,
+      "grad_norm": 2.716863323227044,
+      "learning_rate": 1.818458516910413e-05,
+      "loss": 0.8676,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6573882143914718,
+      "grad_norm": 1.9705395914912947,
+      "learning_rate": 1.818090871699158e-05,
+      "loss": 0.8832,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6579804560260586,
+      "grad_norm": 4.001918700211936,
+      "learning_rate": 1.817722891842733e-05,
+      "loss": 0.858,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6585726976606455,
+      "grad_norm": 1.7504239586578514,
+      "learning_rate": 1.8173545774916628e-05,
+      "loss": 0.8335,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6591649392952325,
+      "grad_norm": 2.103180849415543,
+      "learning_rate": 1.816985928796608e-05,
+      "loss": 0.8058,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6597571809298194,
+      "grad_norm": 2.042550989077482,
+      "learning_rate": 1.8166169459083673e-05,
+      "loss": 0.8622,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6603494225644063,
+      "grad_norm": 2.895471180154479,
+      "learning_rate": 1.8162476289778745e-05,
+      "loss": 0.8349,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6609416641989931,
+      "grad_norm": 2.1623953611204163,
+      "learning_rate": 1.8158779781562022e-05,
+      "loss": 0.8047,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6615339058335801,
+      "grad_norm": 4.82625679912337,
+      "learning_rate": 1.8155079935945577e-05,
+      "loss": 0.8058,
+      "step": 1117
+    },
+    {
+      "epoch": 0.662126147468167,
+      "grad_norm": 2.1209114049054163,
+      "learning_rate": 1.8151376754442856e-05,
+      "loss": 0.8345,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6627183891027539,
+      "grad_norm": 3.202247450296245,
+      "learning_rate": 1.8147670238568666e-05,
+      "loss": 0.8264,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6633106307373409,
+      "grad_norm": 4.683917620146975,
+      "learning_rate": 1.8143960389839184e-05,
+      "loss": 0.8997,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6639028723719278,
+      "grad_norm": 1.5146727248882648,
+      "learning_rate": 1.8140247209771946e-05,
+      "loss": 0.8525,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6644951140065146,
+      "grad_norm": 2.4296936822760746,
+      "learning_rate": 1.8136530699885852e-05,
+      "loss": 0.8435,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6650873556411016,
+      "grad_norm": 3.833524070202994,
+      "learning_rate": 1.813281086170116e-05,
+      "loss": 0.9018,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6656795972756885,
+      "grad_norm": 2.0040513437046648,
+      "learning_rate": 1.8129087696739497e-05,
+      "loss": 0.8821,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6662718389102754,
+      "grad_norm": 1.4186690415848557,
+      "learning_rate": 1.8125361206523845e-05,
+      "loss": 0.8447,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6668640805448623,
+      "grad_norm": 1.4088568380399054,
+      "learning_rate": 1.8121631392578545e-05,
+      "loss": 0.8548,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6674563221794492,
+      "grad_norm": 1.8279521308810265,
+      "learning_rate": 1.811789825642931e-05,
+      "loss": 0.8499,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6680485638140361,
+      "grad_norm": 1.7623710462745776,
+      "learning_rate": 1.8114161799603195e-05,
+      "loss": 0.8746,
+      "step": 1128
+    },
+    {
+      "epoch": 0.668640805448623,
+      "grad_norm": 1.6841484340107102,
+      "learning_rate": 1.8110422023628623e-05,
+      "loss": 0.8287,
+      "step": 1129
+    },
+    {
+      "epoch": 0.66923304708321,
+      "grad_norm": 1.5460477858044241,
+      "learning_rate": 1.810667893003537e-05,
+      "loss": 0.8403,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6698252887177969,
+      "grad_norm": 2.904407513681924,
+      "learning_rate": 1.8102932520354572e-05,
+      "loss": 0.8535,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6704175303523838,
+      "grad_norm": 2.3502598935192682,
+      "learning_rate": 1.8099182796118727e-05,
+      "loss": 0.9069,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6710097719869706,
+      "grad_norm": 2.267045869299457,
+      "learning_rate": 1.8095429758861682e-05,
+      "loss": 0.8095,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6716020136215576,
+      "grad_norm": 1.5438780262047374,
+      "learning_rate": 1.8091673410118633e-05,
+      "loss": 0.8264,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6721942552561445,
+      "grad_norm": 2.956645508277099,
+      "learning_rate": 1.8087913751426142e-05,
+      "loss": 0.8709,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6727864968907314,
+      "grad_norm": 2.2889022888507284,
+      "learning_rate": 1.8084150784322123e-05,
+      "loss": 0.836,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6733787385253184,
+      "grad_norm": 1.4850766527643833,
+      "learning_rate": 1.8080384510345838e-05,
+      "loss": 0.8603,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6739709801599052,
+      "grad_norm": 2.1026164044539315,
+      "learning_rate": 1.8076614931037908e-05,
+      "loss": 0.8917,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6745632217944921,
+      "grad_norm": 1.620826120946746,
+      "learning_rate": 1.80728420479403e-05,
+      "loss": 0.8327,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6751554634290791,
+      "grad_norm": 1.5100307488301057,
+      "learning_rate": 1.8069065862596338e-05,
+      "loss": 0.872,
+      "step": 1140
+    },
+    {
+      "epoch": 0.675747705063666,
+      "grad_norm": 1.3534401313900613,
+      "learning_rate": 1.8065286376550692e-05,
+      "loss": 0.8426,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6763399466982529,
+      "grad_norm": 12.092384661716544,
+      "learning_rate": 1.8061503591349386e-05,
+      "loss": 0.8416,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6769321883328397,
+      "grad_norm": 1.208766260196838,
+      "learning_rate": 1.8057717508539786e-05,
+      "loss": 0.8475,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6775244299674267,
+      "grad_norm": 1.2451969427779477,
+      "learning_rate": 1.8053928129670624e-05,
+      "loss": 0.8306,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6781166716020136,
+      "grad_norm": 1.799182555450414,
+      "learning_rate": 1.805013545629196e-05,
+      "loss": 0.8296,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6787089132366005,
+      "grad_norm": 1.3830942844154197,
+      "learning_rate": 1.8046339489955214e-05,
+      "loss": 0.8246,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6793011548711875,
+      "grad_norm": 1.6078573859008807,
+      "learning_rate": 1.804254023221315e-05,
+      "loss": 0.8071,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6798933965057744,
+      "grad_norm": 1.335961933557725,
+      "learning_rate": 1.8038737684619874e-05,
+      "loss": 0.8593,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6804856381403612,
+      "grad_norm": 1.546298180666286,
+      "learning_rate": 1.8034931848730846e-05,
+      "loss": 0.8249,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6810778797749482,
+      "grad_norm": 1.635069545524243,
+      "learning_rate": 1.8031122726102868e-05,
+      "loss": 0.877,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6816701214095351,
+      "grad_norm": 1.2691447098723208,
+      "learning_rate": 1.802731031829408e-05,
+      "loss": 0.8395,
+      "step": 1151
+    },
+    {
+      "epoch": 0.682262363044122,
+      "grad_norm": 1.869539168329112,
+      "learning_rate": 1.8023494626863976e-05,
+      "loss": 0.813,
+      "step": 1152
+    },
+    {
+      "epoch": 0.682854604678709,
+      "grad_norm": 4.282750886399974,
+      "learning_rate": 1.8019675653373387e-05,
+      "loss": 0.8337,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6834468463132958,
+      "grad_norm": 1.5109022276624124,
+      "learning_rate": 1.8015853399384488e-05,
+      "loss": 0.8654,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6840390879478827,
+      "grad_norm": 3.215342397054156,
+      "learning_rate": 1.8012027866460797e-05,
+      "loss": 0.9001,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6846313295824696,
+      "grad_norm": 1.311513238293177,
+      "learning_rate": 1.8008199056167167e-05,
+      "loss": 0.8889,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6852235712170566,
+      "grad_norm": 1.4742093712676303,
+      "learning_rate": 1.80043669700698e-05,
+      "loss": 0.8175,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6858158128516435,
+      "grad_norm": 1.2459495798186484,
+      "learning_rate": 1.8000531609736236e-05,
+      "loss": 0.8096,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6864080544862304,
+      "grad_norm": 1.6603825601369882,
+      "learning_rate": 1.799669297673535e-05,
+      "loss": 0.8859,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6870002961208173,
+      "grad_norm": 1.8179070682239424,
+      "learning_rate": 1.7992851072637366e-05,
+      "loss": 0.8541,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6875925377554042,
+      "grad_norm": 1.2406374896101122,
+      "learning_rate": 1.7989005899013828e-05,
+      "loss": 0.8413,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6881847793899911,
+      "grad_norm": 1.5721991668978201,
+      "learning_rate": 1.798515745743764e-05,
+      "loss": 0.8636,
+      "step": 1162
+    },
+    {
+      "epoch": 0.688777021024578,
+      "grad_norm": 1.8095072497988238,
+      "learning_rate": 1.798130574948302e-05,
+      "loss": 0.8892,
+      "step": 1163
+    },
+    {
+      "epoch": 0.689369262659165,
+      "grad_norm": 1.8048696574204277,
+      "learning_rate": 1.797745077672554e-05,
+      "loss": 0.8207,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6899615042937518,
+      "grad_norm": 1.9168478766964387,
+      "learning_rate": 1.7973592540742095e-05,
+      "loss": 0.8593,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6905537459283387,
+      "grad_norm": 1.7351792742264414,
+      "learning_rate": 1.7969731043110928e-05,
+      "loss": 0.845,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6911459875629257,
+      "grad_norm": 1.9689492610583388,
+      "learning_rate": 1.79658662854116e-05,
+      "loss": 0.8399,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6917382291975126,
+      "grad_norm": 2.378890204311825,
+      "learning_rate": 1.7961998269225024e-05,
+      "loss": 0.836,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6923304708320995,
+      "grad_norm": 1.7234996114622456,
+      "learning_rate": 1.7958126996133427e-05,
+      "loss": 0.8726,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6929227124666865,
+      "grad_norm": 12.35334136288384,
+      "learning_rate": 1.7954252467720386e-05,
+      "loss": 0.8219,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6935149541012733,
+      "grad_norm": 1.5256431304922908,
+      "learning_rate": 1.7950374685570794e-05,
+      "loss": 0.8227,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6941071957358602,
+      "grad_norm": 2.490658777946983,
+      "learning_rate": 1.7946493651270883e-05,
+      "loss": 0.8785,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6946994373704471,
+      "grad_norm": 2.143120458120011,
+      "learning_rate": 1.794260936640822e-05,
+      "loss": 0.7914,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6952916790050341,
+      "grad_norm": 1.961814139735007,
+      "learning_rate": 1.7938721832571688e-05,
+      "loss": 0.84,
+      "step": 1174
+    },
+    {
+      "epoch": 0.695883920639621,
+      "grad_norm": 1.8808018807507807,
+      "learning_rate": 1.7934831051351513e-05,
+      "loss": 0.8379,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6964761622742078,
+      "grad_norm": 3.8564691984179915,
+      "learning_rate": 1.793093702433924e-05,
+      "loss": 0.8931,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6970684039087948,
+      "grad_norm": 2.991169714160442,
+      "learning_rate": 1.792703975312774e-05,
+      "loss": 0.8591,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6976606455433817,
+      "grad_norm": 2.0564470226192775,
+      "learning_rate": 1.792313923931123e-05,
+      "loss": 0.8484,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6982528871779686,
+      "grad_norm": 2.7676305573686486,
+      "learning_rate": 1.791923548448523e-05,
+      "loss": 0.9096,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6988451288125556,
+      "grad_norm": 2.313908537479878,
+      "learning_rate": 1.7915328490246594e-05,
+      "loss": 0.8442,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6994373704471424,
+      "grad_norm": 2.6234620243010967,
+      "learning_rate": 1.7911418258193503e-05,
+      "loss": 0.8982,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7000296120817293,
+      "grad_norm": 1.8042865323547594,
+      "learning_rate": 1.7907504789925473e-05,
+      "loss": 0.8913,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7006218537163162,
+      "grad_norm": 2.7891631926045948,
+      "learning_rate": 1.7903588087043314e-05,
+      "loss": 0.833,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7012140953509032,
+      "grad_norm": 1.8602348121875936,
+      "learning_rate": 1.789966815114919e-05,
+      "loss": 0.8298,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7018063369854901,
+      "grad_norm": 2.8372707284352074,
+      "learning_rate": 1.7895744983846575e-05,
+      "loss": 0.8192,
+      "step": 1185
+    },
+    {
+      "epoch": 0.702398578620077,
+      "grad_norm": 1.8134109667963114,
+      "learning_rate": 1.789181858674026e-05,
+      "loss": 0.8336,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7029908202546639,
+      "grad_norm": 3.3699318930231392,
+      "learning_rate": 1.7887888961436367e-05,
+      "loss": 0.837,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7035830618892508,
+      "grad_norm": 3.5279273185363933,
+      "learning_rate": 1.788395610954233e-05,
+      "loss": 0.8584,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7041753035238377,
+      "grad_norm": 10.545341229489324,
+      "learning_rate": 1.7880020032666906e-05,
+      "loss": 0.8366,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7047675451584247,
+      "grad_norm": 1.8317791616806671,
+      "learning_rate": 1.7876080732420176e-05,
+      "loss": 0.8156,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7053597867930116,
+      "grad_norm": 2.8166796146116635,
+      "learning_rate": 1.7872138210413533e-05,
+      "loss": 0.8591,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7059520284275984,
+      "grad_norm": 2.3506890987289837,
+      "learning_rate": 1.7868192468259686e-05,
+      "loss": 0.8696,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7065442700621853,
+      "grad_norm": 1.9915077835053718,
+      "learning_rate": 1.7864243507572678e-05,
+      "loss": 0.8078,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7071365116967723,
+      "grad_norm": 2.4789943921714728,
+      "learning_rate": 1.7860291329967842e-05,
+      "loss": 0.8549,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7077287533313592,
+      "grad_norm": 3.5203512739721114,
+      "learning_rate": 1.7856335937061843e-05,
+      "loss": 0.8259,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7083209949659461,
+      "grad_norm": 2.7626514330217207,
+      "learning_rate": 1.7852377330472668e-05,
+      "loss": 0.8024,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7089132366005331,
+      "grad_norm": 2.9434175694356286,
+      "learning_rate": 1.7848415511819602e-05,
+      "loss": 0.8561,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7095054782351199,
+      "grad_norm": 2.938160794113487,
+      "learning_rate": 1.7844450482723258e-05,
+      "loss": 0.8348,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7100977198697068,
+      "grad_norm": 7.531717951600264,
+      "learning_rate": 1.7840482244805546e-05,
+      "loss": 0.8257,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7106899615042938,
+      "grad_norm": 2.5108516098902345,
+      "learning_rate": 1.783651079968971e-05,
+      "loss": 0.8359,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7112822031388807,
+      "grad_norm": 6.9259734842190515,
+      "learning_rate": 1.7832536149000283e-05,
+      "loss": 0.8493,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7118744447734676,
+      "grad_norm": 3.604438919377138,
+      "learning_rate": 1.782855829436313e-05,
+      "loss": 0.9005,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7124666864080544,
+      "grad_norm": 6.432790286739704,
+      "learning_rate": 1.782457723740541e-05,
+      "loss": 0.8502,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7130589280426414,
+      "grad_norm": 2.309337652505556,
+      "learning_rate": 1.7820592979755605e-05,
+      "loss": 0.8378,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7136511696772283,
+      "grad_norm": 4.86010637400942,
+      "learning_rate": 1.78166055230435e-05,
+      "loss": 0.8571,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7142434113118152,
+      "grad_norm": 2.4643862263555003,
+      "learning_rate": 1.7812614868900185e-05,
+      "loss": 0.8026,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7148356529464022,
+      "grad_norm": 4.488518458955996,
+      "learning_rate": 1.7808621018958063e-05,
+      "loss": 0.8474,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7154278945809891,
+      "grad_norm": 2.142014115136399,
+      "learning_rate": 1.7804623974850844e-05,
+      "loss": 0.8327,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7160201362155759,
+      "grad_norm": 3.059145127668178,
+      "learning_rate": 1.7800623738213544e-05,
+      "loss": 0.8184,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7166123778501629,
+      "grad_norm": 3.1694873538825927,
+      "learning_rate": 1.779662031068249e-05,
+      "loss": 0.8609,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7172046194847498,
+      "grad_norm": 1.8086628702754877,
+      "learning_rate": 1.7792613693895298e-05,
+      "loss": 0.8932,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7177968611193367,
+      "grad_norm": 3.4357386272756254,
+      "learning_rate": 1.7788603889490907e-05,
+      "loss": 0.825,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7183891027539236,
+      "grad_norm": 5.825010332449306,
+      "learning_rate": 1.7784590899109554e-05,
+      "loss": 0.8121,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7189813443885105,
+      "grad_norm": 4.868145531572263,
+      "learning_rate": 1.778057472439277e-05,
+      "loss": 0.8699,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7195735860230974,
+      "grad_norm": 2.4243315664795104,
+      "learning_rate": 1.7776555366983403e-05,
+      "loss": 0.8094,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7201658276576843,
+      "grad_norm": 2.2560480044043554,
+      "learning_rate": 1.7772532828525593e-05,
+      "loss": 0.8866,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7207580692922713,
+      "grad_norm": 2.7080185493295432,
+      "learning_rate": 1.7768507110664787e-05,
+      "loss": 0.8779,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7213503109268582,
+      "grad_norm": 2.1434635867346246,
+      "learning_rate": 1.7764478215047725e-05,
+      "loss": 0.8646,
+      "step": 1218
+    },
+    {
+      "epoch": 0.721942552561445,
+      "grad_norm": 3.651352165498139,
+      "learning_rate": 1.776044614332246e-05,
+      "loss": 0.8161,
+      "step": 1219
+    },
+    {
+      "epoch": 0.722534794196032,
+      "grad_norm": 3.1506750177909093,
+      "learning_rate": 1.7756410897138326e-05,
+      "loss": 0.8791,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7231270358306189,
+      "grad_norm": 3.4301185321356424,
+      "learning_rate": 1.7752372478145975e-05,
+      "loss": 0.8843,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7237192774652058,
+      "grad_norm": 2.985509123858241,
+      "learning_rate": 1.7748330887997344e-05,
+      "loss": 0.8521,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7243115190997927,
+      "grad_norm": 2.1395675990186964,
+      "learning_rate": 1.774428612834567e-05,
+      "loss": 0.8504,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7249037607343797,
+      "grad_norm": 1.8208520277656095,
+      "learning_rate": 1.7740238200845485e-05,
+      "loss": 0.8359,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7254960023689665,
+      "grad_norm": 2.5196284085738583,
+      "learning_rate": 1.773618710715262e-05,
+      "loss": 0.8758,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7260882440035534,
+      "grad_norm": 1.9151106256797734,
+      "learning_rate": 1.7732132848924206e-05,
+      "loss": 0.8402,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7266804856381404,
+      "grad_norm": 2.8093175656971927,
+      "learning_rate": 1.7728075427818658e-05,
+      "loss": 0.8215,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 3.506653025025842,
+      "learning_rate": 1.7724014845495684e-05,
+      "loss": 0.8823,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7278649689073142,
+      "grad_norm": 3.033233001673288,
+      "learning_rate": 1.77199511036163e-05,
+      "loss": 0.8542,
+      "step": 1229
+    },
+    {
+      "epoch": 0.728457210541901,
+      "grad_norm": 1.474917032531982,
+      "learning_rate": 1.77158842038428e-05,
+      "loss": 0.8106,
+      "step": 1230
+    },
+    {
+      "epoch": 0.729049452176488,
+      "grad_norm": 5.616912789611092,
+      "learning_rate": 1.7711814147838776e-05,
+      "loss": 0.8468,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7296416938110749,
+      "grad_norm": 2.2432187424799217,
+      "learning_rate": 1.7707740937269108e-05,
+      "loss": 0.7949,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7302339354456618,
+      "grad_norm": 3.5975642635267424,
+      "learning_rate": 1.770366457379997e-05,
+      "loss": 0.8752,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7308261770802488,
+      "grad_norm": 2.0011627880439096,
+      "learning_rate": 1.769958505909882e-05,
+      "loss": 0.8102,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7314184187148357,
+      "grad_norm": 2.8280112394549604,
+      "learning_rate": 1.7695502394834416e-05,
+      "loss": 0.8692,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7320106603494225,
+      "grad_norm": 5.669137853206307,
+      "learning_rate": 1.7691416582676792e-05,
+      "loss": 0.8747,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7326029019840095,
+      "grad_norm": 2.1063941357764935,
+      "learning_rate": 1.768732762429728e-05,
+      "loss": 0.7983,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7331951436185964,
+      "grad_norm": 2.4339492898585253,
+      "learning_rate": 1.7683235521368484e-05,
+      "loss": 0.8579,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7337873852531833,
+      "grad_norm": 3.432527823235171,
+      "learning_rate": 1.7679140275564315e-05,
+      "loss": 0.814,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7343796268877703,
+      "grad_norm": 2.4768439625101286,
+      "learning_rate": 1.7675041888559952e-05,
+      "loss": 0.8373,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7349718685223571,
+      "grad_norm": 3.360300015627485,
+      "learning_rate": 1.767094036203187e-05,
+      "loss": 0.8396,
+      "step": 1241
+    },
+    {
+      "epoch": 0.735564110156944,
+      "grad_norm": 3.705948689891925,
+      "learning_rate": 1.7666835697657824e-05,
+      "loss": 0.8773,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7361563517915309,
+      "grad_norm": 2.29301370449494,
+      "learning_rate": 1.7662727897116843e-05,
+      "loss": 0.8446,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7367485934261179,
+      "grad_norm": 5.029628549833236,
+      "learning_rate": 1.7658616962089262e-05,
+      "loss": 0.7834,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7373408350607048,
+      "grad_norm": 3.1057311683988056,
+      "learning_rate": 1.765450289425668e-05,
+      "loss": 0.832,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7379330766952917,
+      "grad_norm": 3.097561035378791,
+      "learning_rate": 1.765038569530198e-05,
+      "loss": 0.8543,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7385253183298786,
+      "grad_norm": 3.034406664943384,
+      "learning_rate": 1.7646265366909324e-05,
+      "loss": 0.8386,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7391175599644655,
+      "grad_norm": 2.6367359298667985,
+      "learning_rate": 1.7642141910764164e-05,
+      "loss": 0.8551,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7397098015990524,
+      "grad_norm": 2.3375256950022014,
+      "learning_rate": 1.763801532855323e-05,
+      "loss": 0.8636,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7403020432336394,
+      "grad_norm": 2.518186118551617,
+      "learning_rate": 1.7633885621964516e-05,
+      "loss": 0.8382,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7408942848682263,
+      "grad_norm": 2.8355859671570003,
+      "learning_rate": 1.762975279268731e-05,
+      "loss": 0.8467,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7414865265028131,
+      "grad_norm": 3.038664124903949,
+      "learning_rate": 1.7625616842412166e-05,
+      "loss": 0.8035,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7420787681374,
+      "grad_norm": 3.4648343011990033,
+      "learning_rate": 1.7621477772830927e-05,
+      "loss": 0.8167,
+      "step": 1253
+    },
+    {
+      "epoch": 0.742671009771987,
+      "grad_norm": 2.5985099036690764,
+      "learning_rate": 1.76173355856367e-05,
+      "loss": 0.852,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7432632514065739,
+      "grad_norm": 2.043804821692882,
+      "learning_rate": 1.761319028252388e-05,
+      "loss": 0.8029,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7438554930411608,
+      "grad_norm": 3.499822034990886,
+      "learning_rate": 1.7609041865188122e-05,
+      "loss": 0.8653,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7444477346757477,
+      "grad_norm": 2.351751327143647,
+      "learning_rate": 1.7604890335326362e-05,
+      "loss": 0.8544,
+      "step": 1257
+    },
+    {
+      "epoch": 0.7450399763103346,
+      "grad_norm": 1.989760304041313,
+      "learning_rate": 1.7600735694636814e-05,
+      "loss": 0.809,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7456322179449215,
+      "grad_norm": 3.2403502877273676,
+      "learning_rate": 1.7596577944818954e-05,
+      "loss": 0.8615,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7462244595795084,
+      "grad_norm": 1.5631166920770456,
+      "learning_rate": 1.759241708757354e-05,
+      "loss": 0.8369,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7468167012140954,
+      "grad_norm": 5.042322573986611,
+      "learning_rate": 1.7588253124602596e-05,
+      "loss": 0.7944,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7474089428486823,
+      "grad_norm": 4.732786628166864,
+      "learning_rate": 1.7584086057609413e-05,
+      "loss": 0.8202,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7480011844832691,
+      "grad_norm": 2.275400890981689,
+      "learning_rate": 1.757991588829856e-05,
+      "loss": 0.8174,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7485934261178561,
+      "grad_norm": 2.669086920815046,
+      "learning_rate": 1.757574261837587e-05,
+      "loss": 0.867,
+      "step": 1264
+    },
+    {
+      "epoch": 0.749185667752443,
+      "grad_norm": 2.6052774063562865,
+      "learning_rate": 1.7571566249548446e-05,
+      "loss": 0.8538,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7497779093870299,
+      "grad_norm": 2.405180495723794,
+      "learning_rate": 1.7567386783524655e-05,
+      "loss": 0.8009,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7503701510216169,
+      "grad_norm": 3.3676019565827886,
+      "learning_rate": 1.756320422201413e-05,
+      "loss": 0.8042,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7509623926562037,
+      "grad_norm": 2.966741525614473,
+      "learning_rate": 1.7559018566727788e-05,
+      "loss": 0.8026,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7515546342907906,
+      "grad_norm": 3.2543032515193993,
+      "learning_rate": 1.755482981937778e-05,
+      "loss": 0.8627,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7521468759253775,
+      "grad_norm": 2.0946507733084165,
+      "learning_rate": 1.755063798167755e-05,
+      "loss": 0.8687,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7527391175599645,
+      "grad_norm": 1.3831223141092206,
+      "learning_rate": 1.754644305534179e-05,
+      "loss": 0.8339,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7533313591945514,
+      "grad_norm": 2.0665293112108465,
+      "learning_rate": 1.754224504208647e-05,
+      "loss": 0.7964,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7539236008291383,
+      "grad_norm": 13.182638808399124,
+      "learning_rate": 1.7538043943628803e-05,
+      "loss": 0.851,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7545158424637252,
+      "grad_norm": 1.7213978689349443,
+      "learning_rate": 1.7533839761687278e-05,
+      "loss": 0.848,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7551080840983121,
+      "grad_norm": 4.2008316501367435,
+      "learning_rate": 1.7529632497981644e-05,
+      "loss": 0.7775,
+      "step": 1275
+    },
+    {
+      "epoch": 0.755700325732899,
+      "grad_norm": 1.7693985409902604,
+      "learning_rate": 1.7525422154232906e-05,
+      "loss": 0.8109,
+      "step": 1276
+    },
+    {
+      "epoch": 0.756292567367486,
+      "grad_norm": 2.2297535857353155,
+      "learning_rate": 1.752120873216333e-05,
+      "loss": 0.9018,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7568848090020729,
+      "grad_norm": 2.7453149628984215,
+      "learning_rate": 1.7516992233496443e-05,
+      "loss": 0.8978,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7574770506366597,
+      "grad_norm": 1.6904372827667524,
+      "learning_rate": 1.7512772659957037e-05,
+      "loss": 0.875,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7580692922712466,
+      "grad_norm": 2.943908693131239,
+      "learning_rate": 1.7508550013271146e-05,
+      "loss": 0.8365,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7586615339058336,
+      "grad_norm": 2.5437925239618684,
+      "learning_rate": 1.7504324295166073e-05,
+      "loss": 0.8528,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7592537755404205,
+      "grad_norm": 1.6253829546363479,
+      "learning_rate": 1.7500095507370376e-05,
+      "loss": 0.8533,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7598460171750074,
+      "grad_norm": 2.534930764133692,
+      "learning_rate": 1.7495863651613866e-05,
+      "loss": 0.8562,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7604382588095944,
+      "grad_norm": 1.6060087765012423,
+      "learning_rate": 1.749162872962761e-05,
+      "loss": 0.8215,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7610305004441812,
+      "grad_norm": 1.7368003536795598,
+      "learning_rate": 1.7487390743143927e-05,
+      "loss": 0.8376,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7616227420787681,
+      "grad_norm": 3.382650140379708,
+      "learning_rate": 1.7483149693896396e-05,
+      "loss": 0.8282,
+      "step": 1286
+    },
+    {
+      "epoch": 0.762214983713355,
+      "grad_norm": 3.2394560487893074,
+      "learning_rate": 1.747890558361984e-05,
+      "loss": 0.8334,
+      "step": 1287
+    },
+    {
+      "epoch": 0.762807225347942,
+      "grad_norm": 2.1963133524390295,
+      "learning_rate": 1.7474658414050344e-05,
+      "loss": 0.8512,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7633994669825289,
+      "grad_norm": 2.7546045395326986,
+      "learning_rate": 1.7470408186925233e-05,
+      "loss": 0.846,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7639917086171157,
+      "grad_norm": 2.4453730617259004,
+      "learning_rate": 1.7466154903983092e-05,
+      "loss": 0.8403,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7645839502517027,
+      "grad_norm": 3.391610937118782,
+      "learning_rate": 1.7461898566963754e-05,
+      "loss": 0.844,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7651761918862896,
+      "grad_norm": 2.9168088797780345,
+      "learning_rate": 1.74576391776083e-05,
+      "loss": 0.8823,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7657684335208765,
+      "grad_norm": 2.199990393304276,
+      "learning_rate": 1.745337673765906e-05,
+      "loss": 0.8407,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7663606751554635,
+      "grad_norm": 1.9287995776670166,
+      "learning_rate": 1.744911124885961e-05,
+      "loss": 0.8179,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7669529167900503,
+      "grad_norm": 2.2102693326836964,
+      "learning_rate": 1.7444842712954778e-05,
+      "loss": 0.7751,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7675451584246372,
+      "grad_norm": 2.388498039600017,
+      "learning_rate": 1.7440571131690626e-05,
+      "loss": 0.8686,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7681374000592242,
+      "grad_norm": 2.6036812285499322,
+      "learning_rate": 1.7436296506814483e-05,
+      "loss": 0.8603,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7687296416938111,
+      "grad_norm": 2.9525736440709425,
+      "learning_rate": 1.7432018840074905e-05,
+      "loss": 0.8569,
+      "step": 1298
+    },
+    {
+      "epoch": 0.769321883328398,
+      "grad_norm": 2.301471909050177,
+      "learning_rate": 1.7427738133221694e-05,
+      "loss": 0.8967,
+      "step": 1299
+    },
+    {
+      "epoch": 0.769914124962985,
+      "grad_norm": 2.399067499647489,
+      "learning_rate": 1.742345438800591e-05,
+      "loss": 0.886,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7705063665975718,
+      "grad_norm": 1.6895013555018883,
+      "learning_rate": 1.7419167606179837e-05,
+      "loss": 0.8087,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7710986082321587,
+      "grad_norm": 3.155550562789463,
+      "learning_rate": 1.7414877789497017e-05,
+      "loss": 0.8124,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7716908498667456,
+      "grad_norm": 4.009429752899135,
+      "learning_rate": 1.741058493971222e-05,
+      "loss": 0.835,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7722830915013326,
+      "grad_norm": 2.7343050132154483,
+      "learning_rate": 1.7406289058581466e-05,
+      "loss": 0.8687,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7728753331359195,
+      "grad_norm": 2.511148964673554,
+      "learning_rate": 1.7401990147862008e-05,
+      "loss": 0.8697,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7734675747705063,
+      "grad_norm": 1.6005516598974556,
+      "learning_rate": 1.739768820931235e-05,
+      "loss": 0.8558,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7740598164050932,
+      "grad_norm": 5.650942014927322,
+      "learning_rate": 1.7393383244692218e-05,
+      "loss": 0.8184,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7746520580396802,
+      "grad_norm": 3.5550381737938634,
+      "learning_rate": 1.7389075255762592e-05,
+      "loss": 0.8144,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7752442996742671,
+      "grad_norm": 2.764917013421342,
+      "learning_rate": 1.738476424428568e-05,
+      "loss": 0.8719,
+      "step": 1309
+    },
+    {
+      "epoch": 0.775836541308854,
+      "grad_norm": 3.587759059350319,
+      "learning_rate": 1.7380450212024924e-05,
+      "loss": 0.861,
+      "step": 1310
+    },
+    {
+      "epoch": 0.776428782943441,
+      "grad_norm": 2.253441060117673,
+      "learning_rate": 1.737613316074501e-05,
+      "loss": 0.8321,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7770210245780278,
+      "grad_norm": 1.4280085430434677,
+      "learning_rate": 1.737181309221185e-05,
+      "loss": 0.8575,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7776132662126147,
+      "grad_norm": 2.035397944647251,
+      "learning_rate": 1.73674900081926e-05,
+      "loss": 0.8998,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7782055078472017,
+      "grad_norm": 1.519998519535366,
+      "learning_rate": 1.7363163910455646e-05,
+      "loss": 0.7885,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7787977494817886,
+      "grad_norm": 3.650699183495285,
+      "learning_rate": 1.73588348007706e-05,
+      "loss": 0.8609,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7793899911163755,
+      "grad_norm": 5.6510846760274775,
+      "learning_rate": 1.735450268090831e-05,
+      "loss": 0.8084,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7799822327509623,
+      "grad_norm": 2.09775440341691,
+      "learning_rate": 1.735016755264086e-05,
+      "loss": 0.8738,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7805744743855493,
+      "grad_norm": 1.931648307541478,
+      "learning_rate": 1.7345829417741564e-05,
+      "loss": 0.815,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7811667160201362,
+      "grad_norm": 3.1720176685267183,
+      "learning_rate": 1.734148827798496e-05,
+      "loss": 0.8635,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7817589576547231,
+      "grad_norm": 3.202514276793904,
+      "learning_rate": 1.7337144135146818e-05,
+      "loss": 0.8523,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7823511992893101,
+      "grad_norm": 2.3467580376278767,
+      "learning_rate": 1.7332796991004137e-05,
+      "loss": 0.8929,
+      "step": 1321
+    },
+    {
+      "epoch": 0.782943440923897,
+      "grad_norm": 1.7193074019861174,
+      "learning_rate": 1.7328446847335142e-05,
+      "loss": 0.8243,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7835356825584838,
+      "grad_norm": 1.93789230603662,
+      "learning_rate": 1.7324093705919288e-05,
+      "loss": 0.8041,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7841279241930708,
+      "grad_norm": 2.813798090440131,
+      "learning_rate": 1.731973756853726e-05,
+      "loss": 0.8083,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7847201658276577,
+      "grad_norm": 5.447428960254784,
+      "learning_rate": 1.7315378436970952e-05,
+      "loss": 0.8272,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7853124074622446,
+      "grad_norm": 1.5252289352217188,
+      "learning_rate": 1.73110163130035e-05,
+      "loss": 0.9191,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7859046490968316,
+      "grad_norm": 4.098459400505394,
+      "learning_rate": 1.730665119841926e-05,
+      "loss": 0.8249,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7864968907314184,
+      "grad_norm": 1.3507732532666377,
+      "learning_rate": 1.7302283095003807e-05,
+      "loss": 0.8448,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7870891323660053,
+      "grad_norm": 1.8647038075327456,
+      "learning_rate": 1.729791200454394e-05,
+      "loss": 0.8271,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7876813740005922,
+      "grad_norm": 6.007731309930798,
+      "learning_rate": 1.729353792882768e-05,
+      "loss": 0.8679,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7882736156351792,
+      "grad_norm": 1.730951437129075,
+      "learning_rate": 1.7289160869644273e-05,
+      "loss": 0.8049,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7888658572697661,
+      "grad_norm": 2.5126410391090825,
+      "learning_rate": 1.728478082878418e-05,
+      "loss": 0.854,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7894580989043529,
+      "grad_norm": 17.97525643692378,
+      "learning_rate": 1.7280397808039087e-05,
+      "loss": 0.8243,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7900503405389399,
+      "grad_norm": 3.5691587966526512,
+      "learning_rate": 1.7276011809201896e-05,
+      "loss": 0.8118,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7906425821735268,
+      "grad_norm": 1.8559040171020436,
+      "learning_rate": 1.7271622834066722e-05,
+      "loss": 0.8886,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7912348238081137,
+      "grad_norm": 1.6994200843830343,
+      "learning_rate": 1.7267230884428905e-05,
+      "loss": 0.8215,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7918270654427007,
+      "grad_norm": 2.6015274500336343,
+      "learning_rate": 1.7262835962085e-05,
+      "loss": 0.7956,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7924193070772876,
+      "grad_norm": 1.8824140694819236,
+      "learning_rate": 1.725843806883278e-05,
+      "loss": 0.8642,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7930115487118744,
+      "grad_norm": 2.14510961510416,
+      "learning_rate": 1.7254037206471226e-05,
+      "loss": 0.7958,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7936037903464613,
+      "grad_norm": 1.978972467508621,
+      "learning_rate": 1.7249633376800542e-05,
+      "loss": 0.7886,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7941960319810483,
+      "grad_norm": 1.797453382477477,
+      "learning_rate": 1.724522658162214e-05,
+      "loss": 0.8117,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7947882736156352,
+      "grad_norm": 1.9495032921229507,
+      "learning_rate": 1.7240816822738646e-05,
+      "loss": 0.7986,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7953805152502221,
+      "grad_norm": 2.420632731466788,
+      "learning_rate": 1.72364041019539e-05,
+      "loss": 0.8428,
+      "step": 1343
+    },
+    {
+      "epoch": 0.795972756884809,
+      "grad_norm": 1.430427148223359,
+      "learning_rate": 1.7231988421072957e-05,
+      "loss": 0.8502,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7965649985193959,
+      "grad_norm": 2.0260574149321755,
+      "learning_rate": 1.7227569781902073e-05,
+      "loss": 0.819,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7971572401539828,
+      "grad_norm": 1.8930044657278047,
+      "learning_rate": 1.722314818624872e-05,
+      "loss": 0.8378,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7977494817885697,
+      "grad_norm": 1.368043792921863,
+      "learning_rate": 1.7218723635921587e-05,
+      "loss": 0.8363,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7983417234231567,
+      "grad_norm": 1.6583322562133502,
+      "learning_rate": 1.7214296132730555e-05,
+      "loss": 0.8587,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7989339650577436,
+      "grad_norm": 2.6952971468399842,
+      "learning_rate": 1.7209865678486727e-05,
+      "loss": 0.8388,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7995262066923304,
+      "grad_norm": 1.603441854661535,
+      "learning_rate": 1.7205432275002403e-05,
+      "loss": 0.8303,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8001184483269174,
+      "grad_norm": 1.3660414196014246,
+      "learning_rate": 1.7200995924091102e-05,
+      "loss": 0.8137,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8007106899615043,
+      "grad_norm": 1.3428573586009116,
+      "learning_rate": 1.719655662756753e-05,
+      "loss": 0.8324,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8013029315960912,
+      "grad_norm": 3.4551970616956185,
+      "learning_rate": 1.719211438724762e-05,
+      "loss": 0.8495,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8018951732306782,
+      "grad_norm": 3.699225184360079,
+      "learning_rate": 1.7187669204948495e-05,
+      "loss": 0.8617,
+      "step": 1354
+    },
+    {
+      "epoch": 0.802487414865265,
+      "grad_norm": 1.6762625181042976,
+      "learning_rate": 1.718322108248848e-05,
+      "loss": 0.8473,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8030796564998519,
+      "grad_norm": 1.3717529222149383,
+      "learning_rate": 1.7178770021687113e-05,
+      "loss": 0.834,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8036718981344388,
+      "grad_norm": 2.315814133619738,
+      "learning_rate": 1.7174316024365123e-05,
+      "loss": 0.8763,
+      "step": 1357
+    },
+    {
+      "epoch": 0.8042641397690258,
+      "grad_norm": 4.886422808842332,
+      "learning_rate": 1.7169859092344448e-05,
+      "loss": 0.8444,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8048563814036127,
+      "grad_norm": 1.7525565651645565,
+      "learning_rate": 1.7165399227448222e-05,
+      "loss": 0.8134,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8054486230381996,
+      "grad_norm": 2.806549546737031,
+      "learning_rate": 1.7160936431500785e-05,
+      "loss": 0.9172,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8060408646727865,
+      "grad_norm": 1.5747965592763717,
+      "learning_rate": 1.7156470706327665e-05,
+      "loss": 0.8433,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8066331063073734,
+      "grad_norm": 2.128730321422943,
+      "learning_rate": 1.7152002053755604e-05,
+      "loss": 0.8361,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8072253479419603,
+      "grad_norm": 2.327065092345198,
+      "learning_rate": 1.7147530475612524e-05,
+      "loss": 0.8628,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8078175895765473,
+      "grad_norm": 2.380546899275701,
+      "learning_rate": 1.714305597372755e-05,
+      "loss": 0.7691,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8084098312111342,
+      "grad_norm": 2.5637775941669063,
+      "learning_rate": 1.7138578549931013e-05,
+      "loss": 0.7957,
+      "step": 1365
+    },
+    {
+      "epoch": 0.809002072845721,
+      "grad_norm": 2.4241927150678175,
+      "learning_rate": 1.713409820605443e-05,
+      "loss": 0.8272,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8095943144803079,
+      "grad_norm": 1.8511672871361389,
+      "learning_rate": 1.7129614943930505e-05,
+      "loss": 0.8062,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8101865561148949,
+      "grad_norm": 5.52637450036081,
+      "learning_rate": 1.7125128765393157e-05,
+      "loss": 0.8065,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8107787977494818,
+      "grad_norm": 1.5937260830934799,
+      "learning_rate": 1.7120639672277474e-05,
+      "loss": 0.8085,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8113710393840687,
+      "grad_norm": 1.4639078667885639,
+      "learning_rate": 1.7116147666419755e-05,
+      "loss": 0.8631,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8119632810186556,
+      "grad_norm": 2.4168601194929518,
+      "learning_rate": 1.7111652749657473e-05,
+      "loss": 0.816,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8125555226532425,
+      "grad_norm": 7.289050587362146,
+      "learning_rate": 1.7107154923829317e-05,
+      "loss": 0.8525,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8131477642878294,
+      "grad_norm": 2.360028584777416,
+      "learning_rate": 1.710265419077514e-05,
+      "loss": 0.8313,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8137400059224164,
+      "grad_norm": 2.0945779721069457,
+      "learning_rate": 1.7098150552335997e-05,
+      "loss": 0.8624,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8143322475570033,
+      "grad_norm": 3.0412982148200283,
+      "learning_rate": 1.709364401035413e-05,
+      "loss": 0.8574,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8149244891915902,
+      "grad_norm": 1.5284655725565341,
+      "learning_rate": 1.708913456667297e-05,
+      "loss": 0.8844,
+      "step": 1376
+    },
+    {
+      "epoch": 0.815516730826177,
+      "grad_norm": 3.2308336643923368,
+      "learning_rate": 1.7084622223137128e-05,
+      "loss": 0.8091,
+      "step": 1377
+    },
+    {
+      "epoch": 0.816108972460764,
+      "grad_norm": 1.830650660023485,
+      "learning_rate": 1.7080106981592407e-05,
+      "loss": 0.8669,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8167012140953509,
+      "grad_norm": 2.814611843926401,
+      "learning_rate": 1.70755888438858e-05,
+      "loss": 0.8626,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8172934557299378,
+      "grad_norm": 2.183664883059378,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.8077,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8178856973645248,
+      "grad_norm": 3.4013315528606594,
+      "learning_rate": 1.7066543887380793e-05,
+      "loss": 0.8089,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8184779389991116,
+      "grad_norm": 2.299860689598534,
+      "learning_rate": 1.7062017072282285e-05,
+      "loss": 0.8607,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8190701806336985,
+      "grad_norm": 2.268163956608553,
+      "learning_rate": 1.7057487368421685e-05,
+      "loss": 0.8744,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8196624222682855,
+      "grad_norm": 3.0103061900949166,
+      "learning_rate": 1.705295477765188e-05,
+      "loss": 0.7831,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8202546639028724,
+      "grad_norm": 1.9846928716505232,
+      "learning_rate": 1.7048419301826973e-05,
+      "loss": 0.8019,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8208469055374593,
+      "grad_norm": 2.1436359754808434,
+      "learning_rate": 1.7043880942802212e-05,
+      "loss": 0.7726,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8214391471720462,
+      "grad_norm": 15.99919570552301,
+      "learning_rate": 1.7039339702434057e-05,
+      "loss": 0.8581,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8220313888066331,
+      "grad_norm": 1.9788780361165421,
+      "learning_rate": 1.7034795582580118e-05,
+      "loss": 0.8926,
+      "step": 1388
+    },
+    {
+      "epoch": 0.82262363044122,
+      "grad_norm": 1.9979870871622687,
+      "learning_rate": 1.70302485850992e-05,
+      "loss": 0.8445,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8232158720758069,
+      "grad_norm": 3.3526689648805106,
+      "learning_rate": 1.7025698711851283e-05,
+      "loss": 0.8779,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8238081137103939,
+      "grad_norm": 3.2663903535745553,
+      "learning_rate": 1.7021145964697518e-05,
+      "loss": 0.8397,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8244003553449808,
+      "grad_norm": 2.4623605879732278,
+      "learning_rate": 1.701659034550023e-05,
+      "loss": 0.8613,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8249925969795676,
+      "grad_norm": 4.055785244924466,
+      "learning_rate": 1.7012031856122936e-05,
+      "loss": 0.8329,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8255848386141545,
+      "grad_norm": 2.546760379245732,
+      "learning_rate": 1.70074704984303e-05,
+      "loss": 0.8129,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8261770802487415,
+      "grad_norm": 2.004660636750862,
+      "learning_rate": 1.7002906274288187e-05,
+      "loss": 0.8649,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8267693218833284,
+      "grad_norm": 3.63927143622832,
+      "learning_rate": 1.6998339185563614e-05,
+      "loss": 0.8651,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8273615635179153,
+      "grad_norm": 2.6389702902480354,
+      "learning_rate": 1.6993769234124777e-05,
+      "loss": 0.8868,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8279538051525023,
+      "grad_norm": 3.2789514048802326,
+      "learning_rate": 1.6989196421841045e-05,
+      "loss": 0.8329,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8285460467870891,
+      "grad_norm": 2.9698551323065248,
+      "learning_rate": 1.6984620750582957e-05,
+      "loss": 0.8506,
+      "step": 1399
+    },
+    {
+      "epoch": 0.829138288421676,
+      "grad_norm": 4.493208712376449,
+      "learning_rate": 1.6980042222222216e-05,
+      "loss": 0.8512,
+      "step": 1400
+    },
+    {
+      "epoch": 0.829730530056263,
+      "grad_norm": 3.2446475490162587,
+      "learning_rate": 1.69754608386317e-05,
+      "loss": 0.8283,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8303227716908499,
+      "grad_norm": 3.8139641340470707,
+      "learning_rate": 1.697087660168545e-05,
+      "loss": 0.8369,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8309150133254368,
+      "grad_norm": 2.537763008254762,
+      "learning_rate": 1.6966289513258683e-05,
+      "loss": 0.8677,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8315072549600236,
+      "grad_norm": 2.97503941123126,
+      "learning_rate": 1.6961699575227767e-05,
+      "loss": 0.8321,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8320994965946106,
+      "grad_norm": 5.054631443547838,
+      "learning_rate": 1.6957106789470252e-05,
+      "loss": 0.8819,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8326917382291975,
+      "grad_norm": 3.2685313483268086,
+      "learning_rate": 1.695251115786484e-05,
+      "loss": 0.8571,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8332839798637844,
+      "grad_norm": 2.6791989072294466,
+      "learning_rate": 1.6947912682291412e-05,
+      "loss": 0.82,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8338762214983714,
+      "grad_norm": 6.388461734354429,
+      "learning_rate": 1.694331136463099e-05,
+      "loss": 0.8654,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8344684631329582,
+      "grad_norm": 5.378746896229759,
+      "learning_rate": 1.6938707206765784e-05,
+      "loss": 0.8756,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8350607047675451,
+      "grad_norm": 2.382113562740471,
+      "learning_rate": 1.6934100210579144e-05,
+      "loss": 0.8521,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8356529464021321,
+      "grad_norm": 2.3110802302573936,
+      "learning_rate": 1.69294903779556e-05,
+      "loss": 0.8355,
+      "step": 1411
+    },
+    {
+      "epoch": 0.836245188036719,
+      "grad_norm": 2.217801291758983,
+      "learning_rate": 1.6924877710780818e-05,
+      "loss": 0.8798,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8368374296713059,
+      "grad_norm": 2.0560538956481036,
+      "learning_rate": 1.6920262210941657e-05,
+      "loss": 0.8128,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8374296713058929,
+      "grad_norm": 6.864186095594025,
+      "learning_rate": 1.69156438803261e-05,
+      "loss": 0.8533,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8380219129404797,
+      "grad_norm": 2.136582758478113,
+      "learning_rate": 1.6911022720823315e-05,
+      "loss": 0.8236,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8386141545750666,
+      "grad_norm": 2.499403080840201,
+      "learning_rate": 1.690639873432361e-05,
+      "loss": 0.8663,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8392063962096535,
+      "grad_norm": 6.035882530883013,
+      "learning_rate": 1.6901771922718453e-05,
+      "loss": 0.8783,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8397986378442405,
+      "grad_norm": 4.243343370936844,
+      "learning_rate": 1.6897142287900477e-05,
+      "loss": 0.8794,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8403908794788274,
+      "grad_norm": 6.601930358453118,
+      "learning_rate": 1.6892509831763467e-05,
+      "loss": 0.8406,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8409831211134142,
+      "grad_norm": 2.546863435034202,
+      "learning_rate": 1.6887874556202342e-05,
+      "loss": 0.8262,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8415753627480012,
+      "grad_norm": 1.626447021048131,
+      "learning_rate": 1.6883236463113207e-05,
+      "loss": 0.8047,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8421676043825881,
+      "grad_norm": 16.71050322079958,
+      "learning_rate": 1.687859555439329e-05,
+      "loss": 0.8308,
+      "step": 1422
+    },
+    {
+      "epoch": 0.842759846017175,
+      "grad_norm": 2.0493773260165353,
+      "learning_rate": 1.6873951831940993e-05,
+      "loss": 0.8338,
+      "step": 1423
+    },
+    {
+      "epoch": 0.843352087651762,
+      "grad_norm": 2.4162309237205988,
+      "learning_rate": 1.686930529765585e-05,
+      "loss": 0.8567,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8439443292863489,
+      "grad_norm": 2.4878273609774753,
+      "learning_rate": 1.6864655953438563e-05,
+      "loss": 0.8029,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8445365709209357,
+      "grad_norm": 3.2572445643608003,
+      "learning_rate": 1.6860003801190975e-05,
+      "loss": 0.8457,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8451288125555226,
+      "grad_norm": 5.916107483028511,
+      "learning_rate": 1.6855348842816074e-05,
+      "loss": 0.8797,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8457210541901096,
+      "grad_norm": 2.2527335241769166,
+      "learning_rate": 1.6850691080218e-05,
+      "loss": 0.8791,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8463132958246965,
+      "grad_norm": 5.083451401380974,
+      "learning_rate": 1.6846030515302044e-05,
+      "loss": 0.8085,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8469055374592834,
+      "grad_norm": 3.1992848879644495,
+      "learning_rate": 1.6841367149974638e-05,
+      "loss": 0.8713,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8474977790938703,
+      "grad_norm": 2.1404065000354775,
+      "learning_rate": 1.6836700986143354e-05,
+      "loss": 0.8767,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8480900207284572,
+      "grad_norm": 3.2502047486489327,
+      "learning_rate": 1.683203202571692e-05,
+      "loss": 0.825,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8486822623630441,
+      "grad_norm": 4.062783383833649,
+      "learning_rate": 1.682736027060521e-05,
+      "loss": 0.7975,
+      "step": 1433
+    },
+    {
+      "epoch": 0.849274503997631,
+      "grad_norm": 3.1346761905120077,
+      "learning_rate": 1.6822685722719224e-05,
+      "loss": 0.8598,
+      "step": 1434
+    },
+    {
+      "epoch": 0.849866745632218,
+      "grad_norm": 4.335065021022113,
+      "learning_rate": 1.681800838397112e-05,
+      "loss": 0.7865,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8504589872668049,
+      "grad_norm": 2.017703357099882,
+      "learning_rate": 1.681332825627419e-05,
+      "loss": 0.8488,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8510512289013917,
+      "grad_norm": 4.753567787732221,
+      "learning_rate": 1.680864534154287e-05,
+      "loss": 0.8702,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8516434705359787,
+      "grad_norm": 2.438375125937311,
+      "learning_rate": 1.680395964169274e-05,
+      "loss": 0.8571,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8522357121705656,
+      "grad_norm": 2.9251937289705916,
+      "learning_rate": 1.6799271158640517e-05,
+      "loss": 0.822,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8528279538051525,
+      "grad_norm": 2.2312157866256412,
+      "learning_rate": 1.6794579894304043e-05,
+      "loss": 0.8243,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8534201954397395,
+      "grad_norm": 7.074372608081394,
+      "learning_rate": 1.678988585060231e-05,
+      "loss": 0.8624,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8540124370743263,
+      "grad_norm": 2.7556705357562357,
+      "learning_rate": 1.6785189029455455e-05,
+      "loss": 0.8614,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8546046787089132,
+      "grad_norm": 1.9746989976797515,
+      "learning_rate": 1.6780489432784738e-05,
+      "loss": 0.8229,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8551969203435001,
+      "grad_norm": 1.9413368148573928,
+      "learning_rate": 1.6775787062512557e-05,
+      "loss": 0.8694,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8557891619780871,
+      "grad_norm": 2.2060934638723206,
+      "learning_rate": 1.6771081920562445e-05,
+      "loss": 0.8598,
+      "step": 1445
+    },
+    {
+      "epoch": 0.856381403612674,
+      "grad_norm": 3.5024309364669444,
+      "learning_rate": 1.676637400885907e-05,
+      "loss": 0.8485,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8569736452472608,
+      "grad_norm": 4.018519154408224,
+      "learning_rate": 1.676166332932824e-05,
+      "loss": 0.8383,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8575658868818478,
+      "grad_norm": 1.7615281242175937,
+      "learning_rate": 1.6756949883896874e-05,
+      "loss": 0.8035,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8581581285164347,
+      "grad_norm": 2.1966167616847403,
+      "learning_rate": 1.675223367449305e-05,
+      "loss": 0.831,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8587503701510216,
+      "grad_norm": 2.6550583205151685,
+      "learning_rate": 1.6747514703045952e-05,
+      "loss": 0.8216,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8593426117856086,
+      "grad_norm": 2.478511653719764,
+      "learning_rate": 1.6742792971485912e-05,
+      "loss": 0.8828,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8599348534201955,
+      "grad_norm": 5.962171814245029,
+      "learning_rate": 1.673806848174438e-05,
+      "loss": 0.8454,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8605270950547823,
+      "grad_norm": 2.083540364486756,
+      "learning_rate": 1.6733341235753938e-05,
+      "loss": 0.8842,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8611193366893692,
+      "grad_norm": 2.063306619099545,
+      "learning_rate": 1.67286112354483e-05,
+      "loss": 0.8155,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8617115783239562,
+      "grad_norm": 3.43994740810729,
+      "learning_rate": 1.6723878482762296e-05,
+      "loss": 0.8795,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8623038199585431,
+      "grad_norm": 3.1123310742355312,
+      "learning_rate": 1.671914297963189e-05,
+      "loss": 0.8512,
+      "step": 1456
+    },
+    {
+      "epoch": 0.86289606159313,
+      "grad_norm": 3.1237187679340455,
+      "learning_rate": 1.671440472799417e-05,
+      "loss": 0.8313,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8634883032277169,
+      "grad_norm": 2.450839707155073,
+      "learning_rate": 1.670966372978735e-05,
+      "loss": 0.8421,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8640805448623038,
+      "grad_norm": 10.847110912043622,
+      "learning_rate": 1.6704919986950757e-05,
+      "loss": 0.9009,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8646727864968907,
+      "grad_norm": 3.949302192305963,
+      "learning_rate": 1.670017350142486e-05,
+      "loss": 0.8754,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8652650281314777,
+      "grad_norm": 1.8975667299053336,
+      "learning_rate": 1.6695424275151228e-05,
+      "loss": 0.819,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8658572697660646,
+      "grad_norm": 3.3380137589905337,
+      "learning_rate": 1.669067231007256e-05,
+      "loss": 0.85,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8664495114006515,
+      "grad_norm": 2.2925444313034133,
+      "learning_rate": 1.668591760813269e-05,
+      "loss": 0.8283,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8670417530352383,
+      "grad_norm": 1.7424162303718678,
+      "learning_rate": 1.668116017127655e-05,
+      "loss": 0.8352,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8676339946698253,
+      "grad_norm": 2.3289766280717425,
+      "learning_rate": 1.66764000014502e-05,
+      "loss": 0.8662,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8682262363044122,
+      "grad_norm": 2.501537435831264,
+      "learning_rate": 1.667163710060082e-05,
+      "loss": 0.8126,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8688184779389991,
+      "grad_norm": 2.9512357732950405,
+      "learning_rate": 1.6666871470676692e-05,
+      "loss": 0.8318,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8694107195735861,
+      "grad_norm": 1.7783090866183635,
+      "learning_rate": 1.6662103113627246e-05,
+      "loss": 0.8499,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8700029612081729,
+      "grad_norm": 2.5732417344764875,
+      "learning_rate": 1.6657332031402992e-05,
+      "loss": 0.8511,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8705952028427598,
+      "grad_norm": 2.270239747001915,
+      "learning_rate": 1.6652558225955582e-05,
+      "loss": 0.8056,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8711874444773468,
+      "grad_norm": 2.0703557879671104,
+      "learning_rate": 1.6647781699237765e-05,
+      "loss": 0.8341,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8717796861119337,
+      "grad_norm": 3.563658846641916,
+      "learning_rate": 1.6643002453203405e-05,
+      "loss": 0.8619,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8723719277465206,
+      "grad_norm": 2.8049555007770453,
+      "learning_rate": 1.6638220489807495e-05,
+      "loss": 0.8093,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8729641693811075,
+      "grad_norm": 5.679515019474668,
+      "learning_rate": 1.6633435811006117e-05,
+      "loss": 0.8665,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8735564110156944,
+      "grad_norm": 5.038477719863848,
+      "learning_rate": 1.6628648418756474e-05,
+      "loss": 0.8382,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8741486526502813,
+      "grad_norm": 8.615254589030032,
+      "learning_rate": 1.662385831501688e-05,
+      "loss": 0.7959,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8747408942848682,
+      "grad_norm": 8.49226125726425,
+      "learning_rate": 1.6619065501746762e-05,
+      "loss": 0.8569,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8753331359194552,
+      "grad_norm": 2.216292664316629,
+      "learning_rate": 1.661426998090664e-05,
+      "loss": 0.8056,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8759253775540421,
+      "grad_norm": 3.1706560681895177,
+      "learning_rate": 1.6609471754458163e-05,
+      "loss": 0.8421,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8765176191886289,
+      "grad_norm": 2.596125359352298,
+      "learning_rate": 1.6604670824364067e-05,
+      "loss": 0.8084,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8771098608232158,
+      "grad_norm": 3.599816985420332,
+      "learning_rate": 1.6599867192588207e-05,
+      "loss": 0.8211,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8777021024578028,
+      "grad_norm": 5.1170671887832,
+      "learning_rate": 1.6595060861095534e-05,
+      "loss": 0.8538,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8782943440923897,
+      "grad_norm": 3.911881349947541,
+      "learning_rate": 1.6590251831852113e-05,
+      "loss": 0.8509,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8788865857269766,
+      "grad_norm": 3.6771517985702125,
+      "learning_rate": 1.6585440106825107e-05,
+      "loss": 0.7833,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8794788273615635,
+      "grad_norm": 2.5698713996771105,
+      "learning_rate": 1.6580625687982776e-05,
+      "loss": 0.814,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8800710689961504,
+      "grad_norm": 10.101577187451042,
+      "learning_rate": 1.6575808577294492e-05,
+      "loss": 0.819,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8806633106307373,
+      "grad_norm": 2.2073990472582476,
+      "learning_rate": 1.657098877673073e-05,
+      "loss": 0.885,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8812555522653243,
+      "grad_norm": 4.739398445699331,
+      "learning_rate": 1.6566166288263046e-05,
+      "loss": 0.8629,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8818477938999112,
+      "grad_norm": 3.1974524924806107,
+      "learning_rate": 1.656134111386412e-05,
+      "loss": 0.8511,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8824400355344981,
+      "grad_norm": 2.5663036406803608,
+      "learning_rate": 1.6556513255507714e-05,
+      "loss": 0.816,
+      "step": 1490
+    },
+    {
+      "epoch": 0.883032277169085,
+      "grad_norm": 1.9680923900191523,
+      "learning_rate": 1.65516827151687e-05,
+      "loss": 0.8843,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8836245188036719,
+      "grad_norm": 2.414157535179823,
+      "learning_rate": 1.6546849494823037e-05,
+      "loss": 0.8753,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8842167604382588,
+      "grad_norm": 2.4575703751091065,
+      "learning_rate": 1.654201359644778e-05,
+      "loss": 0.8654,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8848090020728457,
+      "grad_norm": 2.992592460545843,
+      "learning_rate": 1.653717502202109e-05,
+      "loss": 0.8567,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8854012437074327,
+      "grad_norm": 2.2703367582766525,
+      "learning_rate": 1.653233377352221e-05,
+      "loss": 0.8335,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8859934853420195,
+      "grad_norm": 20.27278103196212,
+      "learning_rate": 1.652748985293149e-05,
+      "loss": 0.883,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8865857269766064,
+      "grad_norm": 1.8516469130504767,
+      "learning_rate": 1.652264326223036e-05,
+      "loss": 0.7949,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8871779686111934,
+      "grad_norm": 4.524574044226397,
+      "learning_rate": 1.6517794003401345e-05,
+      "loss": 0.8369,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8877702102457803,
+      "grad_norm": 2.3066187010469865,
+      "learning_rate": 1.6512942078428072e-05,
+      "loss": 0.8298,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8883624518803672,
+      "grad_norm": 3.9255893035367437,
+      "learning_rate": 1.650808748929525e-05,
+      "loss": 0.8436,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8889546935149542,
+      "grad_norm": 2.475550455756812,
+      "learning_rate": 1.6503230237988676e-05,
+      "loss": 0.7807,
+      "step": 1501
+    },
+    {
+      "epoch": 0.889546935149541,
+      "grad_norm": 1.9064164754303783,
+      "learning_rate": 1.6498370326495242e-05,
+      "loss": 0.8097,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8901391767841279,
+      "grad_norm": 2.9818131068437967,
+      "learning_rate": 1.649350775680292e-05,
+      "loss": 0.8664,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8907314184187148,
+      "grad_norm": 1.928535242218003,
+      "learning_rate": 1.648864253090078e-05,
+      "loss": 0.8444,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8913236600533018,
+      "grad_norm": 2.4186278041533753,
+      "learning_rate": 1.6483774650778973e-05,
+      "loss": 0.8225,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8919159016878887,
+      "grad_norm": 6.329817463922851,
+      "learning_rate": 1.6478904118428735e-05,
+      "loss": 0.8687,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8925081433224755,
+      "grad_norm": 2.406656080352376,
+      "learning_rate": 1.647403093584238e-05,
+      "loss": 0.8689,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8931003849570625,
+      "grad_norm": 1.698392307205128,
+      "learning_rate": 1.6469155105013324e-05,
+      "loss": 0.8136,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8936926265916494,
+      "grad_norm": 1.7222065175556518,
+      "learning_rate": 1.646427662793605e-05,
+      "loss": 0.8621,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8942848682262363,
+      "grad_norm": 1.835429252270241,
+      "learning_rate": 1.6459395506606133e-05,
+      "loss": 0.8463,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8948771098608232,
+      "grad_norm": 2.02354089579679,
+      "learning_rate": 1.6454511743020222e-05,
+      "loss": 0.822,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8954693514954102,
+      "grad_norm": 2.178325990673068,
+      "learning_rate": 1.6449625339176056e-05,
+      "loss": 0.9125,
+      "step": 1512
+    },
+    {
+      "epoch": 0.896061593129997,
+      "grad_norm": 1.6566812304799738,
+      "learning_rate": 1.6444736297072446e-05,
+      "loss": 0.8098,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8966538347645839,
+      "grad_norm": 2.7595341729881344,
+      "learning_rate": 1.6439844618709285e-05,
+      "loss": 0.8959,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8972460763991709,
+      "grad_norm": 1.992051362483181,
+      "learning_rate": 1.6434950306087544e-05,
+      "loss": 0.8598,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8978383180337578,
+      "grad_norm": 1.4714721824680785,
+      "learning_rate": 1.6430053361209274e-05,
+      "loss": 0.8236,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8984305596683447,
+      "grad_norm": 1.6591161099805962,
+      "learning_rate": 1.6425153786077598e-05,
+      "loss": 0.8749,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8990228013029316,
+      "grad_norm": 1.2831071701329722,
+      "learning_rate": 1.642025158269672e-05,
+      "loss": 0.8128,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8996150429375185,
+      "grad_norm": 1.4872368851201596,
+      "learning_rate": 1.641534675307192e-05,
+      "loss": 0.7701,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9002072845721054,
+      "grad_norm": 1.756308879818358,
+      "learning_rate": 1.641043929920954e-05,
+      "loss": 0.8523,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9007995262066923,
+      "grad_norm": 2.161708847108366,
+      "learning_rate": 1.6405529223117013e-05,
+      "loss": 0.8145,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9013917678412793,
+      "grad_norm": 1.585827703784451,
+      "learning_rate": 1.6400616526802835e-05,
+      "loss": 0.8352,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9019840094758661,
+      "grad_norm": 1.823999072218306,
+      "learning_rate": 1.6395701212276573e-05,
+      "loss": 0.8103,
+      "step": 1523
+    },
+    {
+      "epoch": 0.902576251110453,
+      "grad_norm": 1.7892988299933283,
+      "learning_rate": 1.6390783281548865e-05,
+      "loss": 0.8468,
+      "step": 1524
+    },
+    {
+      "epoch": 0.90316849274504,
+      "grad_norm": 1.828224641878733,
+      "learning_rate": 1.638586273663143e-05,
+      "loss": 0.8189,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9037607343796269,
+      "grad_norm": 1.5516682705105311,
+      "learning_rate": 1.6380939579537033e-05,
+      "loss": 0.8347,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9043529760142138,
+      "grad_norm": 3.834168570081106,
+      "learning_rate": 1.6376013812279534e-05,
+      "loss": 0.8573,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9049452176488008,
+      "grad_norm": 2.1526452193356547,
+      "learning_rate": 1.6371085436873847e-05,
+      "loss": 0.8469,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9055374592833876,
+      "grad_norm": 1.525680512787262,
+      "learning_rate": 1.636615445533595e-05,
+      "loss": 0.8192,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9061297009179745,
+      "grad_norm": 2.288883878412789,
+      "learning_rate": 1.6361220869682896e-05,
+      "loss": 0.8153,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9067219425525614,
+      "grad_norm": 4.621322518818992,
+      "learning_rate": 1.63562846819328e-05,
+      "loss": 0.8144,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9073141841871484,
+      "grad_norm": 1.6212241602499209,
+      "learning_rate": 1.635134589410483e-05,
+      "loss": 0.8537,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9079064258217353,
+      "grad_norm": 1.750883625265076,
+      "learning_rate": 1.6346404508219244e-05,
+      "loss": 0.8252,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9084986674563221,
+      "grad_norm": 1.8359912026959841,
+      "learning_rate": 1.6341460526297335e-05,
+      "loss": 0.8425,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 1.4533966344737113,
+      "learning_rate": 1.6336513950361474e-05,
+      "loss": 0.8562,
+      "step": 1535
+    },
+    {
+      "epoch": 0.909683150725496,
+      "grad_norm": 2.2110168582468552,
+      "learning_rate": 1.6331564782435087e-05,
+      "loss": 0.7873,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9102753923600829,
+      "grad_norm": 2.4584163208420913,
+      "learning_rate": 1.6326613024542667e-05,
+      "loss": 0.8297,
+      "step": 1537
+    },
+    {
+      "epoch": 0.9108676339946699,
+      "grad_norm": 2.304101358316969,
+      "learning_rate": 1.6321658678709752e-05,
+      "loss": 0.8304,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9114598756292568,
+      "grad_norm": 1.5235885803236862,
+      "learning_rate": 1.6316701746962956e-05,
+      "loss": 0.8285,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9120521172638436,
+      "grad_norm": 2.0501459681351517,
+      "learning_rate": 1.6311742231329936e-05,
+      "loss": 0.8338,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9126443588984305,
+      "grad_norm": 1.7456834254936182,
+      "learning_rate": 1.630678013383942e-05,
+      "loss": 0.8232,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9132366005330175,
+      "grad_norm": 2.017939224936834,
+      "learning_rate": 1.6301815456521185e-05,
+      "loss": 0.8093,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9138288421676044,
+      "grad_norm": 1.7066296321756638,
+      "learning_rate": 1.629684820140606e-05,
+      "loss": 0.8545,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9144210838021913,
+      "grad_norm": 5.4294245310007,
+      "learning_rate": 1.6291878370525925e-05,
+      "loss": 0.828,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9150133254367782,
+      "grad_norm": 2.6780329423128446,
+      "learning_rate": 1.6286905965913732e-05,
+      "loss": 0.8682,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9156055670713651,
+      "grad_norm": 2.5467678711002364,
+      "learning_rate": 1.6281930989603466e-05,
+      "loss": 0.7832,
+      "step": 1546
+    },
+    {
+      "epoch": 0.916197808705952,
+      "grad_norm": 2.230883357213969,
+      "learning_rate": 1.627695344363018e-05,
+      "loss": 0.8445,
+      "step": 1547
+    },
+    {
+      "epoch": 0.916790050340539,
+      "grad_norm": 4.2585630086605315,
+      "learning_rate": 1.627197333002996e-05,
+      "loss": 0.7631,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9173822919751259,
+      "grad_norm": 2.018426386994712,
+      "learning_rate": 1.6266990650839965e-05,
+      "loss": 0.8065,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9179745336097128,
+      "grad_norm": 5.5918944687522405,
+      "learning_rate": 1.6262005408098378e-05,
+      "loss": 0.8405,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9185667752442996,
+      "grad_norm": 3.0051047010536234,
+      "learning_rate": 1.6257017603844452e-05,
+      "loss": 0.8329,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9191590168788866,
+      "grad_norm": 1.6536747992743932,
+      "learning_rate": 1.6252027240118472e-05,
+      "loss": 0.8,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9197512585134735,
+      "grad_norm": 2.3069408419926978,
+      "learning_rate": 1.6247034318961788e-05,
+      "loss": 0.7958,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9203435001480604,
+      "grad_norm": 2.5997053054107244,
+      "learning_rate": 1.624203884241678e-05,
+      "loss": 0.8594,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9209357417826474,
+      "grad_norm": 2.63780427340796,
+      "learning_rate": 1.6237040812526875e-05,
+      "loss": 0.84,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9215279834172342,
+      "grad_norm": 2.4212885909450206,
+      "learning_rate": 1.6232040231336556e-05,
+      "loss": 0.8491,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9221202250518211,
+      "grad_norm": 2.2044096639579815,
+      "learning_rate": 1.6227037100891335e-05,
+      "loss": 0.8598,
+      "step": 1557
+    },
+    {
+      "epoch": 0.922712466686408,
+      "grad_norm": 1.6328122663952063,
+      "learning_rate": 1.6222031423237776e-05,
+      "loss": 0.8353,
+      "step": 1558
+    },
+    {
+      "epoch": 0.923304708320995,
+      "grad_norm": 2.3168890361561263,
+      "learning_rate": 1.6217023200423483e-05,
+      "loss": 0.8548,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9238969499555819,
+      "grad_norm": 1.7424095619475186,
+      "learning_rate": 1.6212012434497103e-05,
+      "loss": 0.7988,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9244891915901687,
+      "grad_norm": 1.6126116008758955,
+      "learning_rate": 1.6206999127508318e-05,
+      "loss": 0.85,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9250814332247557,
+      "grad_norm": 1.4774563212178027,
+      "learning_rate": 1.620198328150785e-05,
+      "loss": 0.7911,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9256736748593426,
+      "grad_norm": 1.3520460937622927,
+      "learning_rate": 1.6196964898547474e-05,
+      "loss": 0.8204,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9262659164939295,
+      "grad_norm": 1.5690304650328533,
+      "learning_rate": 1.6191943980679975e-05,
+      "loss": 0.7729,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9268581581285165,
+      "grad_norm": 1.7240591652512482,
+      "learning_rate": 1.61869205299592e-05,
+      "loss": 0.8504,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9274503997631034,
+      "grad_norm": 1.629961765323752,
+      "learning_rate": 1.6181894548440022e-05,
+      "loss": 0.8158,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9280426413976902,
+      "grad_norm": 1.664765209050695,
+      "learning_rate": 1.6176866038178348e-05,
+      "loss": 0.8291,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9286348830322771,
+      "grad_norm": 1.3306049204265953,
+      "learning_rate": 1.617183500123112e-05,
+      "loss": 0.7634,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9292271246668641,
+      "grad_norm": 1.7500132992266746,
+      "learning_rate": 1.6166801439656322e-05,
+      "loss": 0.8743,
+      "step": 1569
+    },
+    {
+      "epoch": 0.929819366301451,
+      "grad_norm": 1.5049905349441384,
+      "learning_rate": 1.6161765355512958e-05,
+      "loss": 0.8252,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9304116079360379,
+      "grad_norm": 1.4564671298739658,
+      "learning_rate": 1.615672675086107e-05,
+      "loss": 0.8371,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9310038495706248,
+      "grad_norm": 1.3839434144832101,
+      "learning_rate": 1.615168562776173e-05,
+      "loss": 0.8293,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9315960912052117,
+      "grad_norm": 1.4540537733669165,
+      "learning_rate": 1.6146641988277044e-05,
+      "loss": 0.8373,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9321883328397986,
+      "grad_norm": 5.541810457275405,
+      "learning_rate": 1.6141595834470142e-05,
+      "loss": 0.8454,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9327805744743856,
+      "grad_norm": 1.5031210400984882,
+      "learning_rate": 1.6136547168405185e-05,
+      "loss": 0.8377,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9333728161089725,
+      "grad_norm": 2.397459980734325,
+      "learning_rate": 1.6131495992147363e-05,
+      "loss": 0.8208,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9339650577435594,
+      "grad_norm": 4.008252947997453,
+      "learning_rate": 1.6126442307762886e-05,
+      "loss": 0.8473,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9345572993781462,
+      "grad_norm": 2.148720748325416,
+      "learning_rate": 1.6121386117319e-05,
+      "loss": 0.8426,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9351495410127332,
+      "grad_norm": 4.17679478258533,
+      "learning_rate": 1.611632742288397e-05,
+      "loss": 0.7861,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9357417826473201,
+      "grad_norm": 1.458220294326226,
+      "learning_rate": 1.6111266226527086e-05,
+      "loss": 0.797,
+      "step": 1580
+    },
+    {
+      "epoch": 0.936334024281907,
+      "grad_norm": 2.424871404734393,
+      "learning_rate": 1.6106202530318662e-05,
+      "loss": 0.7973,
+      "step": 1581
+    },
+    {
+      "epoch": 0.936926265916494,
+      "grad_norm": 1.2771446682809802,
+      "learning_rate": 1.6101136336330037e-05,
+      "loss": 0.8365,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9375185075510808,
+      "grad_norm": 1.1198521871087879,
+      "learning_rate": 1.6096067646633568e-05,
+      "loss": 0.8309,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9381107491856677,
+      "grad_norm": 1.4872636584914338,
+      "learning_rate": 1.609099646330263e-05,
+      "loss": 0.8144,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9387029908202547,
+      "grad_norm": 1.2530678270199203,
+      "learning_rate": 1.6085922788411625e-05,
+      "loss": 0.8321,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9392952324548416,
+      "grad_norm": 3.1068270835989167,
+      "learning_rate": 1.6080846624035972e-05,
+      "loss": 0.8287,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9398874740894285,
+      "grad_norm": 1.4244796316413282,
+      "learning_rate": 1.6075767972252107e-05,
+      "loss": 0.827,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9404797157240155,
+      "grad_norm": 1.8678497643123235,
+      "learning_rate": 1.6070686835137484e-05,
+      "loss": 0.8164,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9410719573586023,
+      "grad_norm": 2.0794440551572744,
+      "learning_rate": 1.6065603214770576e-05,
+      "loss": 0.8368,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9416641989931892,
+      "grad_norm": 1.4310283817792016,
+      "learning_rate": 1.6060517113230866e-05,
+      "loss": 0.8414,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9422564406277761,
+      "grad_norm": 1.1400352112299277,
+      "learning_rate": 1.605542853259886e-05,
+      "loss": 0.8632,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9428486822623631,
+      "grad_norm": 2.2260369027177402,
+      "learning_rate": 1.605033747495607e-05,
+      "loss": 0.8193,
+      "step": 1592
+    },
+    {
+      "epoch": 0.94344092389695,
+      "grad_norm": 1.8613131439464332,
+      "learning_rate": 1.6045243942385026e-05,
+      "loss": 0.8796,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9440331655315368,
+      "grad_norm": 1.6025848171890213,
+      "learning_rate": 1.6040147936969263e-05,
+      "loss": 0.8037,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9446254071661238,
+      "grad_norm": 1.9429865966947775,
+      "learning_rate": 1.6035049460793346e-05,
+      "loss": 0.8611,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9452176488007107,
+      "grad_norm": 1.4125880162885809,
+      "learning_rate": 1.602994851594283e-05,
+      "loss": 0.8218,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9458098904352976,
+      "grad_norm": 1.4496484290895701,
+      "learning_rate": 1.6024845104504295e-05,
+      "loss": 0.831,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9464021320698845,
+      "grad_norm": 1.0965352833215942,
+      "learning_rate": 1.6019739228565314e-05,
+      "loss": 0.7957,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9469943737044714,
+      "grad_norm": 2.5680253220532525,
+      "learning_rate": 1.6014630890214483e-05,
+      "loss": 0.8338,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9475866153390583,
+      "grad_norm": 1.4680992447699333,
+      "learning_rate": 1.6009520091541403e-05,
+      "loss": 0.8043,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9481788569736452,
+      "grad_norm": 1.808937748229738,
+      "learning_rate": 1.600440683463667e-05,
+      "loss": 0.8217,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9487710986082322,
+      "grad_norm": 1.0990585324473146,
+      "learning_rate": 1.5999291121591894e-05,
+      "loss": 0.8004,
+      "step": 1602
+    },
+    {
+      "epoch": 0.9493633402428191,
+      "grad_norm": 1.3712584097732767,
+      "learning_rate": 1.59941729544997e-05,
+      "loss": 0.8557,
+      "step": 1603
+    },
+    {
+      "epoch": 0.949955581877406,
+      "grad_norm": 1.8683892068199166,
+      "learning_rate": 1.5989052335453695e-05,
+      "loss": 0.8331,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9505478235119929,
+      "grad_norm": 1.6272782502704557,
+      "learning_rate": 1.598392926654851e-05,
+      "loss": 0.8554,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9511400651465798,
+      "grad_norm": 6.533052707890747,
+      "learning_rate": 1.5978803749879754e-05,
+      "loss": 0.8254,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9517323067811667,
+      "grad_norm": 1.5123671105986136,
+      "learning_rate": 1.5973675787544062e-05,
+      "loss": 0.8491,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9523245484157536,
+      "grad_norm": 2.956300129012484,
+      "learning_rate": 1.596854538163906e-05,
+      "loss": 0.8472,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9529167900503406,
+      "grad_norm": 1.1789264958256336,
+      "learning_rate": 1.5963412534263368e-05,
+      "loss": 0.8399,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9535090316849274,
+      "grad_norm": 2.116364944004953,
+      "learning_rate": 1.595827724751661e-05,
+      "loss": 0.8087,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9541012733195143,
+      "grad_norm": 1.0294899321599131,
+      "learning_rate": 1.5953139523499407e-05,
+      "loss": 0.8103,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9546935149541013,
+      "grad_norm": 1.2277020777634202,
+      "learning_rate": 1.5947999364313378e-05,
+      "loss": 0.8414,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9552857565886882,
+      "grad_norm": 2.0861925635013807,
+      "learning_rate": 1.594285677206114e-05,
+      "loss": 0.8432,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9558779982232751,
+      "grad_norm": 1.6204140200740225,
+      "learning_rate": 1.5937711748846292e-05,
+      "loss": 0.8855,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9564702398578621,
+      "grad_norm": 1.5551835590922503,
+      "learning_rate": 1.5932564296773452e-05,
+      "loss": 0.8211,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9570624814924489,
+      "grad_norm": 1.338127305643877,
+      "learning_rate": 1.5927414417948205e-05,
+      "loss": 0.8845,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9576547231270358,
+      "grad_norm": 1.2387865170884367,
+      "learning_rate": 1.592226211447715e-05,
+      "loss": 0.7897,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9582469647616227,
+      "grad_norm": 1.226435037795598,
+      "learning_rate": 1.5917107388467866e-05,
+      "loss": 0.842,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9588392063962097,
+      "grad_norm": 1.2457195951382203,
+      "learning_rate": 1.5911950242028924e-05,
+      "loss": 0.804,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9594314480307966,
+      "grad_norm": 1.6797866491993712,
+      "learning_rate": 1.5906790677269887e-05,
+      "loss": 0.8548,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9600236896653834,
+      "grad_norm": 1.202377183866836,
+      "learning_rate": 1.590162869630131e-05,
+      "loss": 0.8212,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9606159312999704,
+      "grad_norm": 1.6157690809746326,
+      "learning_rate": 1.589646430123473e-05,
+      "loss": 0.8579,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9612081729345573,
+      "grad_norm": 1.0841272568029858,
+      "learning_rate": 1.5891297494182677e-05,
+      "loss": 0.8165,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9618004145691442,
+      "grad_norm": 1.0296129071614393,
+      "learning_rate": 1.5886128277258665e-05,
+      "loss": 0.8144,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9623926562037312,
+      "grad_norm": 8.702037235312739,
+      "learning_rate": 1.5880956652577194e-05,
+      "loss": 0.8223,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9629848978383181,
+      "grad_norm": 1.1958806153516213,
+      "learning_rate": 1.587578262225375e-05,
+      "loss": 0.8464,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9635771394729049,
+      "grad_norm": 3.794741977806126,
+      "learning_rate": 1.5870606188404803e-05,
+      "loss": 0.7858,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9641693811074918,
+      "grad_norm": 2.0476194711292743,
+      "learning_rate": 1.5865427353147805e-05,
+      "loss": 0.8624,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9647616227420788,
+      "grad_norm": 1.1041034856721486,
+      "learning_rate": 1.586024611860119e-05,
+      "loss": 0.827,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9653538643766657,
+      "grad_norm": 1.2670169849566177,
+      "learning_rate": 1.5855062486884377e-05,
+      "loss": 0.8591,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9659461060112526,
+      "grad_norm": 1.4173702686257748,
+      "learning_rate": 1.5849876460117756e-05,
+      "loss": 0.8619,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9665383476458395,
+      "grad_norm": 3.026025776762052,
+      "learning_rate": 1.5844688040422714e-05,
+      "loss": 0.8098,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9671305892804264,
+      "grad_norm": 1.4053556345848581,
+      "learning_rate": 1.5839497229921596e-05,
+      "loss": 0.807,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9677228309150133,
+      "grad_norm": 1.7031188178538146,
+      "learning_rate": 1.5834304030737744e-05,
+      "loss": 0.7775,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9683150725496003,
+      "grad_norm": 2.9127557245302187,
+      "learning_rate": 1.5829108444995463e-05,
+      "loss": 0.7833,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9689073141841872,
+      "grad_norm": 1.180567784206561,
+      "learning_rate": 1.582391047482004e-05,
+      "loss": 0.8379,
+      "step": 1636
+    },
+    {
+      "epoch": 0.969499555818774,
+      "grad_norm": 1.6229999823880097,
+      "learning_rate": 1.581871012233774e-05,
+      "loss": 0.8217,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9700917974533609,
+      "grad_norm": 1.208083563753721,
+      "learning_rate": 1.5813507389675796e-05,
+      "loss": 0.8457,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9706840390879479,
+      "grad_norm": 2.1554849675219407,
+      "learning_rate": 1.5808302278962425e-05,
+      "loss": 0.8622,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9712762807225348,
+      "grad_norm": 2.64337198259556,
+      "learning_rate": 1.58030947923268e-05,
+      "loss": 0.7997,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9718685223571217,
+      "grad_norm": 1.5576104708781473,
+      "learning_rate": 1.5797884931899085e-05,
+      "loss": 0.8174,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9724607639917087,
+      "grad_norm": 2.038533162623759,
+      "learning_rate": 1.57926726998104e-05,
+      "loss": 0.7796,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9730530056262955,
+      "grad_norm": 2.9263975793725403,
+      "learning_rate": 1.5787458098192846e-05,
+      "loss": 0.8364,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9736452472608824,
+      "grad_norm": 1.4490134642361368,
+      "learning_rate": 1.5782241129179482e-05,
+      "loss": 0.8124,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9742374888954694,
+      "grad_norm": 1.533779399794962,
+      "learning_rate": 1.5777021794904347e-05,
+      "loss": 0.8359,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9748297305300563,
+      "grad_norm": 1.3549406554518657,
+      "learning_rate": 1.5771800097502437e-05,
+      "loss": 0.7979,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9754219721646432,
+      "grad_norm": 1.5156946411040901,
+      "learning_rate": 1.5766576039109727e-05,
+      "loss": 0.8221,
+      "step": 1647
+    },
+    {
+      "epoch": 0.97601421379923,
+      "grad_norm": 1.0074291473613406,
+      "learning_rate": 1.5761349621863145e-05,
+      "loss": 0.7519,
+      "step": 1648
+    },
+    {
+      "epoch": 0.976606455433817,
+      "grad_norm": 1.749198076091115,
+      "learning_rate": 1.575612084790059e-05,
+      "loss": 0.8644,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9771986970684039,
+      "grad_norm": 1.576797332615058,
+      "learning_rate": 1.5750889719360927e-05,
+      "loss": 0.8243,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9777909387029908,
+      "grad_norm": 1.1947271437764446,
+      "learning_rate": 1.5745656238383983e-05,
+      "loss": 0.8444,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9783831803375778,
+      "grad_norm": 1.4690410500886613,
+      "learning_rate": 1.574042040711054e-05,
+      "loss": 0.8234,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9789754219721647,
+      "grad_norm": 1.3832210724891452,
+      "learning_rate": 1.5735182227682353e-05,
+      "loss": 0.8242,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9795676636067515,
+      "grad_norm": 1.0700735949850522,
+      "learning_rate": 1.572994170224213e-05,
+      "loss": 0.738,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9801599052413384,
+      "grad_norm": 1.048484464353663,
+      "learning_rate": 1.572469883293354e-05,
+      "loss": 0.8888,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9807521468759254,
+      "grad_norm": 1.2979362621765205,
+      "learning_rate": 1.571945362190121e-05,
+      "loss": 0.8947,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9813443885105123,
+      "grad_norm": 1.5547020809846688,
+      "learning_rate": 1.571420607129073e-05,
+      "loss": 0.8581,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9819366301450992,
+      "grad_norm": 2.0365430398639117,
+      "learning_rate": 1.5708956183248644e-05,
+      "loss": 0.856,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9825288717796861,
+      "grad_norm": 1.3434352080990184,
+      "learning_rate": 1.5703703959922444e-05,
+      "loss": 0.84,
+      "step": 1659
+    },
+    {
+      "epoch": 0.983121113414273,
+      "grad_norm": 1.5105070318212588,
+      "learning_rate": 1.5698449403460593e-05,
+      "loss": 0.8324,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9837133550488599,
+      "grad_norm": 1.2274682794404093,
+      "learning_rate": 1.5693192516012497e-05,
+      "loss": 0.7942,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9843055966834469,
+      "grad_norm": 1.5114527110798501,
+      "learning_rate": 1.5687933299728517e-05,
+      "loss": 0.9006,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9848978383180338,
+      "grad_norm": 1.118678025977753,
+      "learning_rate": 1.568267175675997e-05,
+      "loss": 0.8235,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9854900799526207,
+      "grad_norm": 0.9278952601780849,
+      "learning_rate": 1.567740788925912e-05,
+      "loss": 0.8125,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9860823215872075,
+      "grad_norm": 1.2284653816674624,
+      "learning_rate": 1.5672141699379195e-05,
+      "loss": 0.7883,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9866745632217945,
+      "grad_norm": 1.5199993795896831,
+      "learning_rate": 1.5666873189274344e-05,
+      "loss": 0.8336,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9872668048563814,
+      "grad_norm": 1.2261415724075908,
+      "learning_rate": 1.56616023610997e-05,
+      "loss": 0.8881,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9878590464909683,
+      "grad_norm": 1.0819795359579585,
+      "learning_rate": 1.5656329217011322e-05,
+      "loss": 0.782,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9884512881255553,
+      "grad_norm": 1.3268966177305566,
+      "learning_rate": 1.565105375916623e-05,
+      "loss": 0.8229,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9890435297601421,
+      "grad_norm": 1.2761359970445783,
+      "learning_rate": 1.5645775989722366e-05,
+      "loss": 0.8224,
+      "step": 1670
+    },
+    {
+      "epoch": 0.989635771394729,
+      "grad_norm": 1.194769993158141,
+      "learning_rate": 1.5640495910838652e-05,
+      "loss": 0.8327,
+      "step": 1671
+    },
+    {
+      "epoch": 0.990228013029316,
+      "grad_norm": 1.2775750866147921,
+      "learning_rate": 1.563521352467493e-05,
+      "loss": 0.8399,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9908202546639029,
+      "grad_norm": 1.0439693594751094,
+      "learning_rate": 1.562992883339199e-05,
+      "loss": 0.838,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9914124962984898,
+      "grad_norm": 1.4853858307458851,
+      "learning_rate": 1.562464183915157e-05,
+      "loss": 0.8309,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9920047379330766,
+      "grad_norm": 1.2338933063602229,
+      "learning_rate": 1.5619352544116354e-05,
+      "loss": 0.7848,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9925969795676636,
+      "grad_norm": 1.1283006710007517,
+      "learning_rate": 1.5614060950449948e-05,
+      "loss": 0.8699,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9931892212022505,
+      "grad_norm": 1.0265880293302552,
+      "learning_rate": 1.5608767060316927e-05,
+      "loss": 0.7997,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9937814628368374,
+      "grad_norm": 8.09364358410445,
+      "learning_rate": 1.560347087588278e-05,
+      "loss": 0.8407,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9943737044714244,
+      "grad_norm": 1.1900974681831218,
+      "learning_rate": 1.5598172399313946e-05,
+      "loss": 0.8168,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9949659461060113,
+      "grad_norm": 1.4239135447597135,
+      "learning_rate": 1.5592871632777798e-05,
+      "loss": 0.8455,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9955581877405981,
+      "grad_norm": 1.0018040213544295,
+      "learning_rate": 1.5587568578442654e-05,
+      "loss": 0.8209,
+      "step": 1681
+    },
+    {
+      "epoch": 0.996150429375185,
+      "grad_norm": 0.9312885357262206,
+      "learning_rate": 1.5582263238477753e-05,
+      "loss": 0.8201,
+      "step": 1682
+    },
+    {
+      "epoch": 0.996742671009772,
+      "grad_norm": 4.822722321751858,
+      "learning_rate": 1.5576955615053283e-05,
+      "loss": 0.826,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9973349126443589,
+      "grad_norm": 1.3705883900182685,
+      "learning_rate": 1.557164571034036e-05,
+      "loss": 0.8634,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9979271542789458,
+      "grad_norm": 1.2331718885455818,
+      "learning_rate": 1.5566333526511032e-05,
+      "loss": 0.836,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9985193959135327,
+      "grad_norm": 1.7535753725578784,
+      "learning_rate": 1.5561019065738282e-05,
+      "loss": 0.7917,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9991116375481196,
+      "grad_norm": 1.141962242390919,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 0.8084,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9997038791827065,
+      "grad_norm": 0.9912371715304618,
+      "learning_rate": 1.55503833220591e-05,
+      "loss": 0.8038,
+      "step": 1688
+    },
+    {
+      "epoch": 1.0002961208172934,
+      "grad_norm": 1.1500062970774998,
+      "learning_rate": 1.5545062043503284e-05,
+      "loss": 0.7195,
+      "step": 1689
+    },
+    {
+      "epoch": 1.0008883624518803,
+      "grad_norm": 1.162441466138965,
+      "learning_rate": 1.5539738496705277e-05,
+      "loss": 0.7466,
+      "step": 1690
+    },
+    {
+      "epoch": 1.0014806040864672,
+      "grad_norm": 2.452972866337579,
+      "learning_rate": 1.553441268384271e-05,
+      "loss": 0.7629,
+      "step": 1691
+    },
+    {
+      "epoch": 1.0020728457210542,
+      "grad_norm": 1.1676699232346432,
+      "learning_rate": 1.5529084607094144e-05,
+      "loss": 0.7211,
+      "step": 1692
+    },
+    {
+      "epoch": 1.002665087355641,
+      "grad_norm": 1.1249391475584591,
+      "learning_rate": 1.5523754268639053e-05,
+      "loss": 0.6833,
+      "step": 1693
+    },
+    {
+      "epoch": 1.003257328990228,
+      "grad_norm": 0.9431636171309491,
+      "learning_rate": 1.5518421670657856e-05,
+      "loss": 0.6873,
+      "step": 1694
+    },
+    {
+      "epoch": 1.003849570624815,
+      "grad_norm": 1.2138310832226853,
+      "learning_rate": 1.5513086815331876e-05,
+      "loss": 0.7482,
+      "step": 1695
+    },
+    {
+      "epoch": 1.0044418122594019,
+      "grad_norm": 1.3175202159803,
+      "learning_rate": 1.550774970484337e-05,
+      "loss": 0.7627,
+      "step": 1696
+    },
+    {
+      "epoch": 1.0050340538939888,
+      "grad_norm": 1.2998459350824845,
+      "learning_rate": 1.5502410341375525e-05,
+      "loss": 0.7568,
+      "step": 1697
+    },
+    {
+      "epoch": 1.0056262955285757,
+      "grad_norm": 1.3077350264029703,
+      "learning_rate": 1.5497068727112435e-05,
+      "loss": 0.7064,
+      "step": 1698
+    },
+    {
+      "epoch": 1.0062185371631627,
+      "grad_norm": 1.295716875999854,
+      "learning_rate": 1.5491724864239116e-05,
+      "loss": 0.7511,
+      "step": 1699
+    },
+    {
+      "epoch": 1.0068107787977494,
+      "grad_norm": 1.0973356675970865,
+      "learning_rate": 1.5486378754941514e-05,
+      "loss": 0.7318,
+      "step": 1700
+    },
+    {
+      "epoch": 1.0074030204323363,
+      "grad_norm": 1.064599007825042,
+      "learning_rate": 1.5481030401406486e-05,
+      "loss": 0.67,
+      "step": 1701
+    },
+    {
+      "epoch": 1.0079952620669232,
+      "grad_norm": 1.7373649804796532,
+      "learning_rate": 1.5475679805821814e-05,
+      "loss": 0.7182,
+      "step": 1702
+    },
+    {
+      "epoch": 1.0085875037015102,
+      "grad_norm": 1.1434410016960337,
+      "learning_rate": 1.547032697037619e-05,
+      "loss": 0.7471,
+      "step": 1703
+    },
+    {
+      "epoch": 1.0091797453360971,
+      "grad_norm": 1.027785100761493,
+      "learning_rate": 1.546497189725922e-05,
+      "loss": 0.7033,
+      "step": 1704
+    },
+    {
+      "epoch": 1.009771986970684,
+      "grad_norm": 1.4750262197277808,
+      "learning_rate": 1.5459614588661435e-05,
+      "loss": 0.6917,
+      "step": 1705
+    },
+    {
+      "epoch": 1.010364228605271,
+      "grad_norm": 1.5320626210105739,
+      "learning_rate": 1.5454255046774273e-05,
+      "loss": 0.7396,
+      "step": 1706
+    },
+    {
+      "epoch": 1.010956470239858,
+      "grad_norm": 1.2399324918591237,
+      "learning_rate": 1.5448893273790093e-05,
+      "loss": 0.7728,
+      "step": 1707
+    },
+    {
+      "epoch": 1.0115487118744448,
+      "grad_norm": 1.2465148539400586,
+      "learning_rate": 1.5443529271902155e-05,
+      "loss": 0.7421,
+      "step": 1708
+    },
+    {
+      "epoch": 1.0121409535090318,
+      "grad_norm": 1.0519427663128549,
+      "learning_rate": 1.543816304330464e-05,
+      "loss": 0.744,
+      "step": 1709
+    },
+    {
+      "epoch": 1.0127331951436187,
+      "grad_norm": 1.1516952366005058,
+      "learning_rate": 1.543279459019264e-05,
+      "loss": 0.7283,
+      "step": 1710
+    },
+    {
+      "epoch": 1.0133254367782054,
+      "grad_norm": 1.1650321254394644,
+      "learning_rate": 1.542742391476215e-05,
+      "loss": 0.7158,
+      "step": 1711
+    },
+    {
+      "epoch": 1.0139176784127923,
+      "grad_norm": 1.0202804013315474,
+      "learning_rate": 1.5422051019210082e-05,
+      "loss": 0.7283,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0145099200473793,
+      "grad_norm": 1.923691433363424,
+      "learning_rate": 1.5416675905734247e-05,
+      "loss": 0.7629,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0151021616819662,
+      "grad_norm": 1.5605996887729816,
+      "learning_rate": 1.5411298576533376e-05,
+      "loss": 0.7222,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0156944033165531,
+      "grad_norm": 1.181981096535394,
+      "learning_rate": 1.540591903380709e-05,
+      "loss": 0.7437,
+      "step": 1715
+    },
+    {
+      "epoch": 1.01628664495114,
+      "grad_norm": 1.2141712016450033,
+      "learning_rate": 1.5400537279755935e-05,
+      "loss": 0.7403,
+      "step": 1716
+    },
+    {
+      "epoch": 1.016878886585727,
+      "grad_norm": 1.4181664222010109,
+      "learning_rate": 1.539515331658134e-05,
+      "loss": 0.7366,
+      "step": 1717
+    },
+    {
+      "epoch": 1.017471128220314,
+      "grad_norm": 1.5554806786394424,
+      "learning_rate": 1.538976714648566e-05,
+      "loss": 0.7507,
+      "step": 1718
+    },
+    {
+      "epoch": 1.0180633698549009,
+      "grad_norm": 1.2281864029513376,
+      "learning_rate": 1.5384378771672132e-05,
+      "loss": 0.7295,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0186556114894878,
+      "grad_norm": 1.8180155521778687,
+      "learning_rate": 1.5378988194344913e-05,
+      "loss": 0.6907,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0192478531240745,
+      "grad_norm": 1.3836146904218154,
+      "learning_rate": 1.537359541670904e-05,
+      "loss": 0.7357,
+      "step": 1721
+    },
+    {
+      "epoch": 1.0198400947586614,
+      "grad_norm": 1.2844050388131933,
+      "learning_rate": 1.5368200440970478e-05,
+      "loss": 0.7001,
+      "step": 1722
+    },
+    {
+      "epoch": 1.0204323363932484,
+      "grad_norm": 1.0399087505464837,
+      "learning_rate": 1.5362803269336063e-05,
+      "loss": 0.707,
+      "step": 1723
+    },
+    {
+      "epoch": 1.0210245780278353,
+      "grad_norm": 1.2549348887269711,
+      "learning_rate": 1.5357403904013546e-05,
+      "loss": 0.7467,
+      "step": 1724
+    },
+    {
+      "epoch": 1.0216168196624222,
+      "grad_norm": 1.1036457909371953,
+      "learning_rate": 1.535200234721157e-05,
+      "loss": 0.7386,
+      "step": 1725
+    },
+    {
+      "epoch": 1.0222090612970092,
+      "grad_norm": 1.5796396509435675,
+      "learning_rate": 1.5346598601139677e-05,
+      "loss": 0.698,
+      "step": 1726
+    },
+    {
+      "epoch": 1.022801302931596,
+      "grad_norm": 1.172115897988437,
+      "learning_rate": 1.5341192668008305e-05,
+      "loss": 0.7232,
+      "step": 1727
+    },
+    {
+      "epoch": 1.023393544566183,
+      "grad_norm": 1.2380517696446258,
+      "learning_rate": 1.533578455002878e-05,
+      "loss": 0.7418,
+      "step": 1728
+    },
+    {
+      "epoch": 1.02398578620077,
+      "grad_norm": 1.0709459361086207,
+      "learning_rate": 1.5330374249413327e-05,
+      "loss": 0.7005,
+      "step": 1729
+    },
+    {
+      "epoch": 1.024578027835357,
+      "grad_norm": 1.7356966783518266,
+      "learning_rate": 1.5324961768375065e-05,
+      "loss": 0.7308,
+      "step": 1730
+    },
+    {
+      "epoch": 1.0251702694699438,
+      "grad_norm": 1.9959691714404597,
+      "learning_rate": 1.5319547109128e-05,
+      "loss": 0.6944,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0257625111045305,
+      "grad_norm": 1.1724773521675533,
+      "learning_rate": 1.531413027388704e-05,
+      "loss": 0.7128,
+      "step": 1732
+    },
+    {
+      "epoch": 1.0263547527391175,
+      "grad_norm": 1.3876472736084153,
+      "learning_rate": 1.5308711264867966e-05,
+      "loss": 0.7233,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0269469943737044,
+      "grad_norm": 0.9552747056186334,
+      "learning_rate": 1.5303290084287465e-05,
+      "loss": 0.7608,
+      "step": 1734
+    },
+    {
+      "epoch": 1.0275392360082913,
+      "grad_norm": 1.4703684447974952,
+      "learning_rate": 1.52978667343631e-05,
+      "loss": 0.7286,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0281314776428783,
+      "grad_norm": 1.295631529067337,
+      "learning_rate": 1.5292441217313324e-05,
+      "loss": 0.7348,
+      "step": 1736
+    },
+    {
+      "epoch": 1.0287237192774652,
+      "grad_norm": 1.7649342180752203,
+      "learning_rate": 1.5287013535357488e-05,
+      "loss": 0.7281,
+      "step": 1737
+    },
+    {
+      "epoch": 1.0293159609120521,
+      "grad_norm": 1.0923294902447087,
+      "learning_rate": 1.5281583690715805e-05,
+      "loss": 0.6939,
+      "step": 1738
+    },
+    {
+      "epoch": 1.029908202546639,
+      "grad_norm": 1.499563132831407,
+      "learning_rate": 1.52761516856094e-05,
+      "loss": 0.6817,
+      "step": 1739
+    },
+    {
+      "epoch": 1.030500444181226,
+      "grad_norm": 1.0171580664278979,
+      "learning_rate": 1.5270717522260264e-05,
+      "loss": 0.7234,
+      "step": 1740
+    },
+    {
+      "epoch": 1.031092685815813,
+      "grad_norm": 1.0354234657563626,
+      "learning_rate": 1.526528120289127e-05,
+      "loss": 0.7194,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0316849274503999,
+      "grad_norm": 1.2340045067110246,
+      "learning_rate": 1.5259842729726186e-05,
+      "loss": 0.7582,
+      "step": 1742
+    },
+    {
+      "epoch": 1.0322771690849866,
+      "grad_norm": 1.1262612916228454,
+      "learning_rate": 1.5254402104989652e-05,
+      "loss": 0.7168,
+      "step": 1743
+    },
+    {
+      "epoch": 1.0328694107195735,
+      "grad_norm": 1.2118168057016097,
+      "learning_rate": 1.5248959330907186e-05,
+      "loss": 0.743,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0334616523541604,
+      "grad_norm": 0.988865475204544,
+      "learning_rate": 1.5243514409705187e-05,
+      "loss": 0.7318,
+      "step": 1745
+    },
+    {
+      "epoch": 1.0340538939887474,
+      "grad_norm": 1.1771979349040396,
+      "learning_rate": 1.5238067343610943e-05,
+      "loss": 0.7278,
+      "step": 1746
+    },
+    {
+      "epoch": 1.0346461356233343,
+      "grad_norm": 1.3840039027011963,
+      "learning_rate": 1.5232618134852598e-05,
+      "loss": 0.7461,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0352383772579212,
+      "grad_norm": 6.602016761729066,
+      "learning_rate": 1.52271667856592e-05,
+      "loss": 0.7317,
+      "step": 1748
+    },
+    {
+      "epoch": 1.0358306188925082,
+      "grad_norm": 1.0065104505442493,
+      "learning_rate": 1.522171329826064e-05,
+      "loss": 0.7306,
+      "step": 1749
+    },
+    {
+      "epoch": 1.036422860527095,
+      "grad_norm": 1.0752528953543894,
+      "learning_rate": 1.5216257674887718e-05,
+      "loss": 0.683,
+      "step": 1750
+    },
+    {
+      "epoch": 1.037015102161682,
+      "grad_norm": 1.0970642144292997,
+      "learning_rate": 1.5210799917772076e-05,
+      "loss": 0.7274,
+      "step": 1751
+    },
+    {
+      "epoch": 1.037607343796269,
+      "grad_norm": 2.4805051543151233,
+      "learning_rate": 1.5205340029146256e-05,
+      "loss": 0.7554,
+      "step": 1752
+    },
+    {
+      "epoch": 1.0381995854308559,
+      "grad_norm": 1.7822116435903128,
+      "learning_rate": 1.5199878011243647e-05,
+      "loss": 0.7089,
+      "step": 1753
+    },
+    {
+      "epoch": 1.0387918270654426,
+      "grad_norm": 0.8928687106650833,
+      "learning_rate": 1.5194413866298536e-05,
+      "loss": 0.7456,
+      "step": 1754
+    },
+    {
+      "epoch": 1.0393840687000295,
+      "grad_norm": 1.19053453703619,
+      "learning_rate": 1.5188947596546053e-05,
+      "loss": 0.7291,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0399763103346165,
+      "grad_norm": 1.4988527163029588,
+      "learning_rate": 1.5183479204222216e-05,
+      "loss": 0.7221,
+      "step": 1756
+    },
+    {
+      "epoch": 1.0405685519692034,
+      "grad_norm": 1.7373245033776372,
+      "learning_rate": 1.5178008691563902e-05,
+      "loss": 0.7289,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0411607936037903,
+      "grad_norm": 1.838119156554147,
+      "learning_rate": 1.5172536060808857e-05,
+      "loss": 0.726,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0417530352383773,
+      "grad_norm": 1.3624164343579135,
+      "learning_rate": 1.5167061314195702e-05,
+      "loss": 0.7325,
+      "step": 1759
+    },
+    {
+      "epoch": 1.0423452768729642,
+      "grad_norm": 1.6612931900677586,
+      "learning_rate": 1.5161584453963908e-05,
+      "loss": 0.7035,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0429375185075511,
+      "grad_norm": 1.4616110065160015,
+      "learning_rate": 1.5156105482353827e-05,
+      "loss": 0.7442,
+      "step": 1761
+    },
+    {
+      "epoch": 1.043529760142138,
+      "grad_norm": 2.0905521686727737,
+      "learning_rate": 1.5150624401606658e-05,
+      "loss": 0.7269,
+      "step": 1762
+    },
+    {
+      "epoch": 1.044122001776725,
+      "grad_norm": 1.7611278415862754,
+      "learning_rate": 1.5145141213964479e-05,
+      "loss": 0.696,
+      "step": 1763
+    },
+    {
+      "epoch": 1.044714243411312,
+      "grad_norm": 2.1466849185074683,
+      "learning_rate": 1.5139655921670213e-05,
+      "loss": 0.6893,
+      "step": 1764
+    },
+    {
+      "epoch": 1.0453064850458986,
+      "grad_norm": 3.86853781945047,
+      "learning_rate": 1.5134168526967661e-05,
+      "loss": 0.7508,
+      "step": 1765
+    },
+    {
+      "epoch": 1.0458987266804856,
+      "grad_norm": 3.5972940752882137,
+      "learning_rate": 1.5128679032101472e-05,
+      "loss": 0.7366,
+      "step": 1766
+    },
+    {
+      "epoch": 1.0464909683150725,
+      "grad_norm": 3.6857544393967347,
+      "learning_rate": 1.5123187439317159e-05,
+      "loss": 0.7766,
+      "step": 1767
+    },
+    {
+      "epoch": 1.0470832099496594,
+      "grad_norm": 1.6678885841692537,
+      "learning_rate": 1.5117693750861096e-05,
+      "loss": 0.7361,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0476754515842464,
+      "grad_norm": 1.672518635910601,
+      "learning_rate": 1.5112197968980503e-05,
+      "loss": 0.6881,
+      "step": 1769
+    },
+    {
+      "epoch": 1.0482676932188333,
+      "grad_norm": 1.5591607166759955,
+      "learning_rate": 1.5106700095923471e-05,
+      "loss": 0.693,
+      "step": 1770
+    },
+    {
+      "epoch": 1.0488599348534202,
+      "grad_norm": 2.257758886616912,
+      "learning_rate": 1.5101200133938933e-05,
+      "loss": 0.6811,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0494521764880071,
+      "grad_norm": 2.0363297195079912,
+      "learning_rate": 1.5095698085276692e-05,
+      "loss": 0.7407,
+      "step": 1772
+    },
+    {
+      "epoch": 1.050044418122594,
+      "grad_norm": 2.830719217097853,
+      "learning_rate": 1.5090193952187382e-05,
+      "loss": 0.7523,
+      "step": 1773
+    },
+    {
+      "epoch": 1.050636659757181,
+      "grad_norm": 1.671928706296085,
+      "learning_rate": 1.5084687736922514e-05,
+      "loss": 0.7253,
+      "step": 1774
+    },
+    {
+      "epoch": 1.051228901391768,
+      "grad_norm": 2.790885707597884,
+      "learning_rate": 1.5079179441734435e-05,
+      "loss": 0.7218,
+      "step": 1775
+    },
+    {
+      "epoch": 1.0518211430263547,
+      "grad_norm": 1.7819284353063456,
+      "learning_rate": 1.5073669068876348e-05,
+      "loss": 0.73,
+      "step": 1776
+    },
+    {
+      "epoch": 1.0524133846609416,
+      "grad_norm": 2.7217998288857794,
+      "learning_rate": 1.5068156620602303e-05,
+      "loss": 0.7359,
+      "step": 1777
+    },
+    {
+      "epoch": 1.0530056262955285,
+      "grad_norm": 5.101430064202783,
+      "learning_rate": 1.5062642099167208e-05,
+      "loss": 0.7175,
+      "step": 1778
+    },
+    {
+      "epoch": 1.0535978679301155,
+      "grad_norm": 2.4798155357352902,
+      "learning_rate": 1.5057125506826806e-05,
+      "loss": 0.7212,
+      "step": 1779
+    },
+    {
+      "epoch": 1.0541901095647024,
+      "grad_norm": 1.4306493055601683,
+      "learning_rate": 1.5051606845837699e-05,
+      "loss": 0.7508,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0547823511992893,
+      "grad_norm": 1.9617350912249327,
+      "learning_rate": 1.5046086118457325e-05,
+      "loss": 0.7302,
+      "step": 1781
+    },
+    {
+      "epoch": 1.0553745928338762,
+      "grad_norm": 3.2294694912122766,
+      "learning_rate": 1.5040563326943974e-05,
+      "loss": 0.7164,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0559668344684632,
+      "grad_norm": 2.159249108038316,
+      "learning_rate": 1.5035038473556776e-05,
+      "loss": 0.7142,
+      "step": 1783
+    },
+    {
+      "epoch": 1.05655907610305,
+      "grad_norm": 1.810542509614688,
+      "learning_rate": 1.5029511560555707e-05,
+      "loss": 0.7212,
+      "step": 1784
+    },
+    {
+      "epoch": 1.057151317737637,
+      "grad_norm": 4.37550667109893,
+      "learning_rate": 1.5023982590201586e-05,
+      "loss": 0.6615,
+      "step": 1785
+    },
+    {
+      "epoch": 1.057743559372224,
+      "grad_norm": 1.4873887560427135,
+      "learning_rate": 1.5018451564756078e-05,
+      "loss": 0.6939,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0583358010068107,
+      "grad_norm": 3.2445272993387335,
+      "learning_rate": 1.5012918486481677e-05,
+      "loss": 0.7296,
+      "step": 1787
+    },
+    {
+      "epoch": 1.0589280426413976,
+      "grad_norm": 1.6958901351830118,
+      "learning_rate": 1.5007383357641723e-05,
+      "loss": 0.7386,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0595202842759845,
+      "grad_norm": 1.6693502713038142,
+      "learning_rate": 1.5001846180500399e-05,
+      "loss": 0.7317,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0601125259105715,
+      "grad_norm": 1.4628738488051636,
+      "learning_rate": 1.499630695732272e-05,
+      "loss": 0.7039,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0607047675451584,
+      "grad_norm": 1.4961438926110393,
+      "learning_rate": 1.4990765690374537e-05,
+      "loss": 0.7023,
+      "step": 1791
+    },
+    {
+      "epoch": 1.0612970091797453,
+      "grad_norm": 2.1544637844797125,
+      "learning_rate": 1.4985222381922543e-05,
+      "loss": 0.7214,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0618892508143323,
+      "grad_norm": 8.28217196999516,
+      "learning_rate": 1.4979677034234265e-05,
+      "loss": 0.717,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0624814924489192,
+      "grad_norm": 2.501132894894816,
+      "learning_rate": 1.4974129649578058e-05,
+      "loss": 0.7356,
+      "step": 1794
+    },
+    {
+      "epoch": 1.0630737340835061,
+      "grad_norm": 1.16138368637112,
+      "learning_rate": 1.4968580230223112e-05,
+      "loss": 0.7435,
+      "step": 1795
+    },
+    {
+      "epoch": 1.063665975718093,
+      "grad_norm": 1.5162128926090968,
+      "learning_rate": 1.496302877843946e-05,
+      "loss": 0.7392,
+      "step": 1796
+    },
+    {
+      "epoch": 1.06425821735268,
+      "grad_norm": 1.2373419739712945,
+      "learning_rate": 1.4957475296497953e-05,
+      "loss": 0.7,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0648504589872667,
+      "grad_norm": 1.9258973289098624,
+      "learning_rate": 1.4951919786670274e-05,
+      "loss": 0.7231,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0654427006218536,
+      "grad_norm": 1.2248509462971935,
+      "learning_rate": 1.4946362251228943e-05,
+      "loss": 0.7276,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0660349422564406,
+      "grad_norm": 1.5808585539599749,
+      "learning_rate": 1.4940802692447306e-05,
+      "loss": 0.7257,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0666271838910275,
+      "grad_norm": 1.3982040276801868,
+      "learning_rate": 1.493524111259953e-05,
+      "loss": 0.7563,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0672194255256144,
+      "grad_norm": 3.7095487179566713,
+      "learning_rate": 1.4929677513960621e-05,
+      "loss": 0.7217,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0678116671602014,
+      "grad_norm": 2.0014255531666936,
+      "learning_rate": 1.4924111898806395e-05,
+      "loss": 0.7425,
+      "step": 1803
+    },
+    {
+      "epoch": 1.0684039087947883,
+      "grad_norm": 1.4939878195406038,
+      "learning_rate": 1.4918544269413511e-05,
+      "loss": 0.727,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0689961504293752,
+      "grad_norm": 1.74927476414085,
+      "learning_rate": 1.4912974628059433e-05,
+      "loss": 0.7637,
+      "step": 1805
+    },
+    {
+      "epoch": 1.0695883920639622,
+      "grad_norm": 1.658292058667489,
+      "learning_rate": 1.4907402977022465e-05,
+      "loss": 0.7623,
+      "step": 1806
+    },
+    {
+      "epoch": 1.070180633698549,
+      "grad_norm": 1.4315694213602883,
+      "learning_rate": 1.4901829318581722e-05,
+      "loss": 0.7211,
+      "step": 1807
+    },
+    {
+      "epoch": 1.070772875333136,
+      "grad_norm": 1.704429216725604,
+      "learning_rate": 1.4896253655017146e-05,
+      "loss": 0.7316,
+      "step": 1808
+    },
+    {
+      "epoch": 1.0713651169677227,
+      "grad_norm": 1.3312378206473277,
+      "learning_rate": 1.4890675988609493e-05,
+      "loss": 0.7199,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0719573586023097,
+      "grad_norm": 3.97929527512917,
+      "learning_rate": 1.4885096321640346e-05,
+      "loss": 0.725,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0725496002368966,
+      "grad_norm": 1.50003812253246,
+      "learning_rate": 1.48795146563921e-05,
+      "loss": 0.7149,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0731418418714835,
+      "grad_norm": 1.67906861392692,
+      "learning_rate": 1.4873930995147971e-05,
+      "loss": 0.7448,
+      "step": 1812
+    },
+    {
+      "epoch": 1.0737340835060705,
+      "grad_norm": 1.6617287890649148,
+      "learning_rate": 1.4868345340191992e-05,
+      "loss": 0.7207,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0743263251406574,
+      "grad_norm": 1.3825842739615184,
+      "learning_rate": 1.4862757693809009e-05,
+      "loss": 0.7454,
+      "step": 1814
+    },
+    {
+      "epoch": 1.0749185667752443,
+      "grad_norm": 1.4144163977489808,
+      "learning_rate": 1.485716805828468e-05,
+      "loss": 0.7158,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0755108084098313,
+      "grad_norm": 1.8101168890390542,
+      "learning_rate": 1.4851576435905489e-05,
+      "loss": 0.7072,
+      "step": 1816
+    },
+    {
+      "epoch": 1.0761030500444182,
+      "grad_norm": 1.872316239101291,
+      "learning_rate": 1.4845982828958713e-05,
+      "loss": 0.7332,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0766952916790051,
+      "grad_norm": 1.3296117156310565,
+      "learning_rate": 1.484038723973246e-05,
+      "loss": 0.7099,
+      "step": 1818
+    },
+    {
+      "epoch": 1.077287533313592,
+      "grad_norm": 1.4154353624795581,
+      "learning_rate": 1.4834789670515637e-05,
+      "loss": 0.7188,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0778797749481788,
+      "grad_norm": 1.7457686235458405,
+      "learning_rate": 1.4829190123597965e-05,
+      "loss": 0.7769,
+      "step": 1820
+    },
+    {
+      "epoch": 1.0784720165827657,
+      "grad_norm": 1.7634196004341314,
+      "learning_rate": 1.4823588601269973e-05,
+      "loss": 0.7494,
+      "step": 1821
+    },
+    {
+      "epoch": 1.0790642582173526,
+      "grad_norm": 1.2440518586809848,
+      "learning_rate": 1.4817985105823003e-05,
+      "loss": 0.6979,
+      "step": 1822
+    },
+    {
+      "epoch": 1.0796564998519396,
+      "grad_norm": 2.2385183360083953,
+      "learning_rate": 1.4812379639549194e-05,
+      "loss": 0.7202,
+      "step": 1823
+    },
+    {
+      "epoch": 1.0802487414865265,
+      "grad_norm": 1.2215652923766105,
+      "learning_rate": 1.4806772204741503e-05,
+      "loss": 0.7091,
+      "step": 1824
+    },
+    {
+      "epoch": 1.0808409831211134,
+      "grad_norm": 1.4449642609762843,
+      "learning_rate": 1.4801162803693676e-05,
+      "loss": 0.7442,
+      "step": 1825
+    },
+    {
+      "epoch": 1.0814332247557004,
+      "grad_norm": 1.105763165926035,
+      "learning_rate": 1.4795551438700283e-05,
+      "loss": 0.7515,
+      "step": 1826
+    },
+    {
+      "epoch": 1.0820254663902873,
+      "grad_norm": 1.6473343255705488,
+      "learning_rate": 1.4789938112056683e-05,
+      "loss": 0.7527,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0826177080248742,
+      "grad_norm": 1.618781048885639,
+      "learning_rate": 1.4784322826059048e-05,
+      "loss": 0.7249,
+      "step": 1828
+    },
+    {
+      "epoch": 1.0832099496594612,
+      "grad_norm": 1.652228341540137,
+      "learning_rate": 1.4778705583004338e-05,
+      "loss": 0.7447,
+      "step": 1829
+    },
+    {
+      "epoch": 1.083802191294048,
+      "grad_norm": 2.4386960209217343,
+      "learning_rate": 1.4773086385190328e-05,
+      "loss": 0.7448,
+      "step": 1830
+    },
+    {
+      "epoch": 1.0843944329286348,
+      "grad_norm": 1.259552923914992,
+      "learning_rate": 1.4767465234915577e-05,
+      "loss": 0.7799,
+      "step": 1831
+    },
+    {
+      "epoch": 1.0849866745632217,
+      "grad_norm": 3.1470024337770446,
+      "learning_rate": 1.4761842134479463e-05,
+      "loss": 0.7429,
+      "step": 1832
+    },
+    {
+      "epoch": 1.0855789161978087,
+      "grad_norm": 1.2003867622349509,
+      "learning_rate": 1.4756217086182142e-05,
+      "loss": 0.7326,
+      "step": 1833
+    },
+    {
+      "epoch": 1.0861711578323956,
+      "grad_norm": 2.4507353417893314,
+      "learning_rate": 1.4750590092324579e-05,
+      "loss": 0.7183,
+      "step": 1834
+    },
+    {
+      "epoch": 1.0867633994669825,
+      "grad_norm": 1.1444638541124244,
+      "learning_rate": 1.474496115520853e-05,
+      "loss": 0.7436,
+      "step": 1835
+    },
+    {
+      "epoch": 1.0873556411015695,
+      "grad_norm": 1.9216426351330924,
+      "learning_rate": 1.4739330277136546e-05,
+      "loss": 0.7302,
+      "step": 1836
+    },
+    {
+      "epoch": 1.0879478827361564,
+      "grad_norm": 1.7272854817225367,
+      "learning_rate": 1.4733697460411973e-05,
+      "loss": 0.7041,
+      "step": 1837
+    },
+    {
+      "epoch": 1.0885401243707433,
+      "grad_norm": 1.1808892975599021,
+      "learning_rate": 1.4728062707338949e-05,
+      "loss": 0.7214,
+      "step": 1838
+    },
+    {
+      "epoch": 1.0891323660053303,
+      "grad_norm": 1.2060838830500324,
+      "learning_rate": 1.4722426020222406e-05,
+      "loss": 0.7387,
+      "step": 1839
+    },
+    {
+      "epoch": 1.0897246076399172,
+      "grad_norm": 1.2045178599251487,
+      "learning_rate": 1.4716787401368067e-05,
+      "loss": 0.708,
+      "step": 1840
+    },
+    {
+      "epoch": 1.090316849274504,
+      "grad_norm": 1.4796202285269815,
+      "learning_rate": 1.4711146853082443e-05,
+      "loss": 0.7453,
+      "step": 1841
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 1.2637405534511554,
+      "learning_rate": 1.4705504377672834e-05,
+      "loss": 0.7001,
+      "step": 1842
+    },
+    {
+      "epoch": 1.0915013325436778,
+      "grad_norm": 0.9993594395369135,
+      "learning_rate": 1.4699859977447335e-05,
+      "loss": 0.7461,
+      "step": 1843
+    },
+    {
+      "epoch": 1.0920935741782647,
+      "grad_norm": 1.3269419928105537,
+      "learning_rate": 1.4694213654714816e-05,
+      "loss": 0.7391,
+      "step": 1844
+    },
+    {
+      "epoch": 1.0926858158128516,
+      "grad_norm": 1.4669810778171086,
+      "learning_rate": 1.4688565411784943e-05,
+      "loss": 0.7218,
+      "step": 1845
+    },
+    {
+      "epoch": 1.0932780574474386,
+      "grad_norm": 3.385399768271858,
+      "learning_rate": 1.4682915250968169e-05,
+      "loss": 0.7459,
+      "step": 1846
+    },
+    {
+      "epoch": 1.0938702990820255,
+      "grad_norm": 1.6211356022303685,
+      "learning_rate": 1.4677263174575723e-05,
+      "loss": 0.7276,
+      "step": 1847
+    },
+    {
+      "epoch": 1.0944625407166124,
+      "grad_norm": 1.3469501603849017,
+      "learning_rate": 1.4671609184919622e-05,
+      "loss": 0.7409,
+      "step": 1848
+    },
+    {
+      "epoch": 1.0950547823511994,
+      "grad_norm": 1.3639447724050635,
+      "learning_rate": 1.4665953284312668e-05,
+      "loss": 0.7392,
+      "step": 1849
+    },
+    {
+      "epoch": 1.0956470239857863,
+      "grad_norm": 1.2226531191169623,
+      "learning_rate": 1.4660295475068443e-05,
+      "loss": 0.7384,
+      "step": 1850
+    },
+    {
+      "epoch": 1.0962392656203732,
+      "grad_norm": 1.5816714452127887,
+      "learning_rate": 1.4654635759501306e-05,
+      "loss": 0.7145,
+      "step": 1851
+    },
+    {
+      "epoch": 1.09683150725496,
+      "grad_norm": 1.3398064935081868,
+      "learning_rate": 1.4648974139926403e-05,
+      "loss": 0.7341,
+      "step": 1852
+    },
+    {
+      "epoch": 1.0974237488895469,
+      "grad_norm": 1.3669350828600808,
+      "learning_rate": 1.4643310618659646e-05,
+      "loss": 0.7229,
+      "step": 1853
+    },
+    {
+      "epoch": 1.0980159905241338,
+      "grad_norm": 1.7617904906850115,
+      "learning_rate": 1.4637645198017745e-05,
+      "loss": 0.7067,
+      "step": 1854
+    },
+    {
+      "epoch": 1.0986082321587207,
+      "grad_norm": 3.706221805008793,
+      "learning_rate": 1.463197788031817e-05,
+      "loss": 0.7075,
+      "step": 1855
+    },
+    {
+      "epoch": 1.0992004737933077,
+      "grad_norm": 1.6435750167047545,
+      "learning_rate": 1.4626308667879175e-05,
+      "loss": 0.758,
+      "step": 1856
+    },
+    {
+      "epoch": 1.0997927154278946,
+      "grad_norm": 1.7817244713218505,
+      "learning_rate": 1.462063756301978e-05,
+      "loss": 0.7666,
+      "step": 1857
+    },
+    {
+      "epoch": 1.1003849570624815,
+      "grad_norm": 3.183409165734301,
+      "learning_rate": 1.4614964568059795e-05,
+      "loss": 0.7431,
+      "step": 1858
+    },
+    {
+      "epoch": 1.1009771986970684,
+      "grad_norm": 1.5588746699566092,
+      "learning_rate": 1.4609289685319788e-05,
+      "loss": 0.6989,
+      "step": 1859
+    },
+    {
+      "epoch": 1.1015694403316554,
+      "grad_norm": 1.4715880184354002,
+      "learning_rate": 1.4603612917121107e-05,
+      "loss": 0.7476,
+      "step": 1860
+    },
+    {
+      "epoch": 1.1021616819662423,
+      "grad_norm": 1.5483999283147984,
+      "learning_rate": 1.4597934265785868e-05,
+      "loss": 0.7095,
+      "step": 1861
+    },
+    {
+      "epoch": 1.102753923600829,
+      "grad_norm": 1.8136131665101558,
+      "learning_rate": 1.4592253733636961e-05,
+      "loss": 0.7218,
+      "step": 1862
+    },
+    {
+      "epoch": 1.103346165235416,
+      "grad_norm": 1.898459387929092,
+      "learning_rate": 1.4586571322998041e-05,
+      "loss": 0.6928,
+      "step": 1863
+    },
+    {
+      "epoch": 1.1039384068700029,
+      "grad_norm": 1.1332820859860129,
+      "learning_rate": 1.4580887036193539e-05,
+      "loss": 0.7379,
+      "step": 1864
+    },
+    {
+      "epoch": 1.1045306485045898,
+      "grad_norm": 1.1910399671887877,
+      "learning_rate": 1.4575200875548639e-05,
+      "loss": 0.7093,
+      "step": 1865
+    },
+    {
+      "epoch": 1.1051228901391768,
+      "grad_norm": 1.7135675328136981,
+      "learning_rate": 1.4569512843389306e-05,
+      "loss": 0.7452,
+      "step": 1866
+    },
+    {
+      "epoch": 1.1057151317737637,
+      "grad_norm": 1.3131715172492713,
+      "learning_rate": 1.4563822942042264e-05,
+      "loss": 0.7512,
+      "step": 1867
+    },
+    {
+      "epoch": 1.1063073734083506,
+      "grad_norm": 1.301975931331709,
+      "learning_rate": 1.4558131173835002e-05,
+      "loss": 0.724,
+      "step": 1868
+    },
+    {
+      "epoch": 1.1068996150429375,
+      "grad_norm": 1.9451185830687892,
+      "learning_rate": 1.4552437541095774e-05,
+      "loss": 0.7658,
+      "step": 1869
+    },
+    {
+      "epoch": 1.1074918566775245,
+      "grad_norm": 2.5902094254473274,
+      "learning_rate": 1.4546742046153596e-05,
+      "loss": 0.7389,
+      "step": 1870
+    },
+    {
+      "epoch": 1.1080840983121114,
+      "grad_norm": 1.8005401821027935,
+      "learning_rate": 1.4541044691338246e-05,
+      "loss": 0.7113,
+      "step": 1871
+    },
+    {
+      "epoch": 1.1086763399466983,
+      "grad_norm": 1.3297390320707363,
+      "learning_rate": 1.453534547898026e-05,
+      "loss": 0.7192,
+      "step": 1872
+    },
+    {
+      "epoch": 1.109268581581285,
+      "grad_norm": 1.616999402923775,
+      "learning_rate": 1.452964441141094e-05,
+      "loss": 0.703,
+      "step": 1873
+    },
+    {
+      "epoch": 1.109860823215872,
+      "grad_norm": 3.5826588522082607,
+      "learning_rate": 1.4523941490962342e-05,
+      "loss": 0.7405,
+      "step": 1874
+    },
+    {
+      "epoch": 1.110453064850459,
+      "grad_norm": 1.5079201259207546,
+      "learning_rate": 1.451823671996728e-05,
+      "loss": 0.739,
+      "step": 1875
+    },
+    {
+      "epoch": 1.1110453064850458,
+      "grad_norm": 2.719011118153245,
+      "learning_rate": 1.451253010075933e-05,
+      "loss": 0.7482,
+      "step": 1876
+    },
+    {
+      "epoch": 1.1116375481196328,
+      "grad_norm": 2.508076865789526,
+      "learning_rate": 1.450682163567281e-05,
+      "loss": 0.7225,
+      "step": 1877
+    },
+    {
+      "epoch": 1.1122297897542197,
+      "grad_norm": 1.6745532655484892,
+      "learning_rate": 1.4501111327042817e-05,
+      "loss": 0.7662,
+      "step": 1878
+    },
+    {
+      "epoch": 1.1128220313888066,
+      "grad_norm": 1.617999952278557,
+      "learning_rate": 1.4495399177205177e-05,
+      "loss": 0.7055,
+      "step": 1879
+    },
+    {
+      "epoch": 1.1134142730233936,
+      "grad_norm": 1.783576483289883,
+      "learning_rate": 1.4489685188496488e-05,
+      "loss": 0.7175,
+      "step": 1880
+    },
+    {
+      "epoch": 1.1140065146579805,
+      "grad_norm": 1.8113648186864728,
+      "learning_rate": 1.4483969363254085e-05,
+      "loss": 0.7221,
+      "step": 1881
+    },
+    {
+      "epoch": 1.1145987562925674,
+      "grad_norm": 4.466398490357165,
+      "learning_rate": 1.447825170381607e-05,
+      "loss": 0.731,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1151909979271544,
+      "grad_norm": 2.3544006277235776,
+      "learning_rate": 1.4472532212521281e-05,
+      "loss": 0.7599,
+      "step": 1883
+    },
+    {
+      "epoch": 1.115783239561741,
+      "grad_norm": 1.5430309293543931,
+      "learning_rate": 1.446681089170932e-05,
+      "loss": 0.7418,
+      "step": 1884
+    },
+    {
+      "epoch": 1.116375481196328,
+      "grad_norm": 2.547169019555421,
+      "learning_rate": 1.4461087743720519e-05,
+      "loss": 0.7436,
+      "step": 1885
+    },
+    {
+      "epoch": 1.116967722830915,
+      "grad_norm": 2.4575093456645174,
+      "learning_rate": 1.4455362770895976e-05,
+      "loss": 0.7316,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1175599644655019,
+      "grad_norm": 1.5479754084135895,
+      "learning_rate": 1.444963597557752e-05,
+      "loss": 0.6865,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1181522061000888,
+      "grad_norm": 1.5331344869939385,
+      "learning_rate": 1.4443907360107734e-05,
+      "loss": 0.7007,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1187444477346757,
+      "grad_norm": 2.1333020071731896,
+      "learning_rate": 1.4438176926829944e-05,
+      "loss": 0.727,
+      "step": 1889
+    },
+    {
+      "epoch": 1.1193366893692627,
+      "grad_norm": 2.758567222418793,
+      "learning_rate": 1.4432444678088222e-05,
+      "loss": 0.7018,
+      "step": 1890
+    },
+    {
+      "epoch": 1.1199289310038496,
+      "grad_norm": 1.7694989327792088,
+      "learning_rate": 1.4426710616227377e-05,
+      "loss": 0.7148,
+      "step": 1891
+    },
+    {
+      "epoch": 1.1205211726384365,
+      "grad_norm": 1.674707947535516,
+      "learning_rate": 1.4420974743592964e-05,
+      "loss": 0.7497,
+      "step": 1892
+    },
+    {
+      "epoch": 1.1211134142730235,
+      "grad_norm": 1.7836099010979485,
+      "learning_rate": 1.4415237062531277e-05,
+      "loss": 0.7138,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1217056559076104,
+      "grad_norm": 1.2259620137684928,
+      "learning_rate": 1.4409497575389352e-05,
+      "loss": 0.6871,
+      "step": 1894
+    },
+    {
+      "epoch": 1.122297897542197,
+      "grad_norm": 3.7532717279318066,
+      "learning_rate": 1.4403756284514962e-05,
+      "loss": 0.7662,
+      "step": 1895
+    },
+    {
+      "epoch": 1.122890139176784,
+      "grad_norm": 1.6907860579891794,
+      "learning_rate": 1.4398013192256615e-05,
+      "loss": 0.7459,
+      "step": 1896
+    },
+    {
+      "epoch": 1.123482380811371,
+      "grad_norm": 2.1021110669948166,
+      "learning_rate": 1.439226830096356e-05,
+      "loss": 0.7327,
+      "step": 1897
+    },
+    {
+      "epoch": 1.124074622445958,
+      "grad_norm": 3.754781447508906,
+      "learning_rate": 1.438652161298578e-05,
+      "loss": 0.7541,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1246668640805448,
+      "grad_norm": 1.6070012623831906,
+      "learning_rate": 1.4380773130673994e-05,
+      "loss": 0.7569,
+      "step": 1899
+    },
+    {
+      "epoch": 1.1252591057151318,
+      "grad_norm": 7.979250377670888,
+      "learning_rate": 1.4375022856379657e-05,
+      "loss": 0.766,
+      "step": 1900
+    },
+    {
+      "epoch": 1.1258513473497187,
+      "grad_norm": 1.3313809617715828,
+      "learning_rate": 1.436927079245495e-05,
+      "loss": 0.7326,
+      "step": 1901
+    },
+    {
+      "epoch": 1.1264435889843056,
+      "grad_norm": 1.9341986555985717,
+      "learning_rate": 1.4363516941252795e-05,
+      "loss": 0.7408,
+      "step": 1902
+    },
+    {
+      "epoch": 1.1270358306188926,
+      "grad_norm": 1.6237139068552036,
+      "learning_rate": 1.4357761305126836e-05,
+      "loss": 0.7064,
+      "step": 1903
+    },
+    {
+      "epoch": 1.1276280722534795,
+      "grad_norm": 1.2456602536397703,
+      "learning_rate": 1.4352003886431459e-05,
+      "loss": 0.7131,
+      "step": 1904
+    },
+    {
+      "epoch": 1.1282203138880664,
+      "grad_norm": 1.8462015297534495,
+      "learning_rate": 1.4346244687521761e-05,
+      "loss": 0.719,
+      "step": 1905
+    },
+    {
+      "epoch": 1.1288125555226531,
+      "grad_norm": 2.308176089812671,
+      "learning_rate": 1.434048371075359e-05,
+      "loss": 0.6911,
+      "step": 1906
+    },
+    {
+      "epoch": 1.12940479715724,
+      "grad_norm": 13.436596947136401,
+      "learning_rate": 1.43347209584835e-05,
+      "loss": 0.7109,
+      "step": 1907
+    },
+    {
+      "epoch": 1.129997038791827,
+      "grad_norm": 2.4217564676011896,
+      "learning_rate": 1.4328956433068789e-05,
+      "loss": 0.7412,
+      "step": 1908
+    },
+    {
+      "epoch": 1.130589280426414,
+      "grad_norm": 1.917753337825582,
+      "learning_rate": 1.4323190136867464e-05,
+      "loss": 0.753,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1311815220610009,
+      "grad_norm": 1.411783160540838,
+      "learning_rate": 1.4317422072238271e-05,
+      "loss": 0.7423,
+      "step": 1910
+    },
+    {
+      "epoch": 1.1317737636955878,
+      "grad_norm": 1.6181997909355976,
+      "learning_rate": 1.4311652241540668e-05,
+      "loss": 0.7672,
+      "step": 1911
+    },
+    {
+      "epoch": 1.1323660053301747,
+      "grad_norm": 1.8668388500696302,
+      "learning_rate": 1.4305880647134847e-05,
+      "loss": 0.7447,
+      "step": 1912
+    },
+    {
+      "epoch": 1.1329582469647617,
+      "grad_norm": 1.6503141304547309,
+      "learning_rate": 1.4300107291381704e-05,
+      "loss": 0.7519,
+      "step": 1913
+    },
+    {
+      "epoch": 1.1335504885993486,
+      "grad_norm": 1.8711353472468055,
+      "learning_rate": 1.4294332176642875e-05,
+      "loss": 0.7117,
+      "step": 1914
+    },
+    {
+      "epoch": 1.1341427302339355,
+      "grad_norm": 1.5597160551862157,
+      "learning_rate": 1.4288555305280702e-05,
+      "loss": 0.7413,
+      "step": 1915
+    },
+    {
+      "epoch": 1.1347349718685225,
+      "grad_norm": 1.9438549888259047,
+      "learning_rate": 1.4282776679658255e-05,
+      "loss": 0.7131,
+      "step": 1916
+    },
+    {
+      "epoch": 1.1353272135031092,
+      "grad_norm": 3.576441284877924,
+      "learning_rate": 1.4276996302139312e-05,
+      "loss": 0.7173,
+      "step": 1917
+    },
+    {
+      "epoch": 1.135919455137696,
+      "grad_norm": 4.139553745005125,
+      "learning_rate": 1.4271214175088374e-05,
+      "loss": 0.7347,
+      "step": 1918
+    },
+    {
+      "epoch": 1.136511696772283,
+      "grad_norm": 2.094913208691617,
+      "learning_rate": 1.4265430300870656e-05,
+      "loss": 0.7003,
+      "step": 1919
+    },
+    {
+      "epoch": 1.13710393840687,
+      "grad_norm": 1.961827791646241,
+      "learning_rate": 1.425964468185209e-05,
+      "loss": 0.7393,
+      "step": 1920
+    },
+    {
+      "epoch": 1.137696180041457,
+      "grad_norm": 1.2595516677616476,
+      "learning_rate": 1.4253857320399316e-05,
+      "loss": 0.7199,
+      "step": 1921
+    },
+    {
+      "epoch": 1.1382884216760438,
+      "grad_norm": 1.76264901153827,
+      "learning_rate": 1.4248068218879691e-05,
+      "loss": 0.7213,
+      "step": 1922
+    },
+    {
+      "epoch": 1.1388806633106308,
+      "grad_norm": 1.8110656077220395,
+      "learning_rate": 1.4242277379661286e-05,
+      "loss": 0.7134,
+      "step": 1923
+    },
+    {
+      "epoch": 1.1394729049452177,
+      "grad_norm": 1.3189613973752465,
+      "learning_rate": 1.4236484805112878e-05,
+      "loss": 0.7287,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1400651465798046,
+      "grad_norm": 2.039277540030174,
+      "learning_rate": 1.4230690497603955e-05,
+      "loss": 0.716,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1406573882143916,
+      "grad_norm": 2.20069189081546,
+      "learning_rate": 1.4224894459504717e-05,
+      "loss": 0.7297,
+      "step": 1926
+    },
+    {
+      "epoch": 1.1412496298489785,
+      "grad_norm": 1.342282720151037,
+      "learning_rate": 1.4219096693186065e-05,
+      "loss": 0.7599,
+      "step": 1927
+    },
+    {
+      "epoch": 1.1418418714835652,
+      "grad_norm": 2.128575546767518,
+      "learning_rate": 1.4213297201019618e-05,
+      "loss": 0.727,
+      "step": 1928
+    },
+    {
+      "epoch": 1.1424341131181521,
+      "grad_norm": 5.679213605059689,
+      "learning_rate": 1.4207495985377687e-05,
+      "loss": 0.7075,
+      "step": 1929
+    },
+    {
+      "epoch": 1.143026354752739,
+      "grad_norm": 1.4477866874730978,
+      "learning_rate": 1.4201693048633302e-05,
+      "loss": 0.7705,
+      "step": 1930
+    },
+    {
+      "epoch": 1.143618596387326,
+      "grad_norm": 1.3689716380240244,
+      "learning_rate": 1.4195888393160183e-05,
+      "loss": 0.7145,
+      "step": 1931
+    },
+    {
+      "epoch": 1.144210838021913,
+      "grad_norm": 2.1302801407154153,
+      "learning_rate": 1.419008202133277e-05,
+      "loss": 0.7193,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1448030796564999,
+      "grad_norm": 1.300453821118966,
+      "learning_rate": 1.4184273935526184e-05,
+      "loss": 0.7635,
+      "step": 1933
+    },
+    {
+      "epoch": 1.1453953212910868,
+      "grad_norm": 1.9849938187340435,
+      "learning_rate": 1.4178464138116272e-05,
+      "loss": 0.7525,
+      "step": 1934
+    },
+    {
+      "epoch": 1.1459875629256737,
+      "grad_norm": 1.7532705143201168,
+      "learning_rate": 1.4172652631479558e-05,
+      "loss": 0.7159,
+      "step": 1935
+    },
+    {
+      "epoch": 1.1465798045602607,
+      "grad_norm": 1.993678172399548,
+      "learning_rate": 1.4166839417993281e-05,
+      "loss": 0.7287,
+      "step": 1936
+    },
+    {
+      "epoch": 1.1471720461948476,
+      "grad_norm": 1.6612921095486382,
+      "learning_rate": 1.4161024500035364e-05,
+      "loss": 0.7116,
+      "step": 1937
+    },
+    {
+      "epoch": 1.1477642878294345,
+      "grad_norm": 2.704022908452418,
+      "learning_rate": 1.4155207879984447e-05,
+      "loss": 0.7005,
+      "step": 1938
+    },
+    {
+      "epoch": 1.1483565294640212,
+      "grad_norm": 1.5244879998598664,
+      "learning_rate": 1.4149389560219846e-05,
+      "loss": 0.7588,
+      "step": 1939
+    },
+    {
+      "epoch": 1.1489487710986082,
+      "grad_norm": 3.6604754255800307,
+      "learning_rate": 1.414356954312158e-05,
+      "loss": 0.749,
+      "step": 1940
+    },
+    {
+      "epoch": 1.149541012733195,
+      "grad_norm": 1.9096638139679427,
+      "learning_rate": 1.4137747831070371e-05,
+      "loss": 0.7127,
+      "step": 1941
+    },
+    {
+      "epoch": 1.150133254367782,
+      "grad_norm": 1.5220531333094949,
+      "learning_rate": 1.4131924426447621e-05,
+      "loss": 0.7561,
+      "step": 1942
+    },
+    {
+      "epoch": 1.150725496002369,
+      "grad_norm": 2.3373596861886456,
+      "learning_rate": 1.412609933163543e-05,
+      "loss": 0.7053,
+      "step": 1943
+    },
+    {
+      "epoch": 1.1513177376369559,
+      "grad_norm": 1.4526913515699773,
+      "learning_rate": 1.4120272549016591e-05,
+      "loss": 0.7197,
+      "step": 1944
+    },
+    {
+      "epoch": 1.1519099792715428,
+      "grad_norm": 1.7959045608328434,
+      "learning_rate": 1.4114444080974585e-05,
+      "loss": 0.7105,
+      "step": 1945
+    },
+    {
+      "epoch": 1.1525022209061297,
+      "grad_norm": 2.2273046483347834,
+      "learning_rate": 1.4108613929893586e-05,
+      "loss": 0.7099,
+      "step": 1946
+    },
+    {
+      "epoch": 1.1530944625407167,
+      "grad_norm": 1.851272958075549,
+      "learning_rate": 1.410278209815845e-05,
+      "loss": 0.6989,
+      "step": 1947
+    },
+    {
+      "epoch": 1.1536867041753036,
+      "grad_norm": 1.4743710555973606,
+      "learning_rate": 1.4096948588154723e-05,
+      "loss": 0.7193,
+      "step": 1948
+    },
+    {
+      "epoch": 1.1542789458098905,
+      "grad_norm": 1.8040748031208247,
+      "learning_rate": 1.4091113402268644e-05,
+      "loss": 0.7769,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1548711874444773,
+      "grad_norm": 4.56346912920917,
+      "learning_rate": 1.4085276542887128e-05,
+      "loss": 0.7509,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1554634290790642,
+      "grad_norm": 1.83204117485103,
+      "learning_rate": 1.4079438012397777e-05,
+      "loss": 0.7835,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1560556707136511,
+      "grad_norm": 1.7404819759752297,
+      "learning_rate": 1.4073597813188884e-05,
+      "loss": 0.7317,
+      "step": 1952
+    },
+    {
+      "epoch": 1.156647912348238,
+      "grad_norm": 2.5984937357601803,
+      "learning_rate": 1.4067755947649416e-05,
+      "loss": 0.735,
+      "step": 1953
+    },
+    {
+      "epoch": 1.157240153982825,
+      "grad_norm": 1.6054589075238175,
+      "learning_rate": 1.4061912418169024e-05,
+      "loss": 0.7258,
+      "step": 1954
+    },
+    {
+      "epoch": 1.157832395617412,
+      "grad_norm": 1.9010263960704716,
+      "learning_rate": 1.405606722713804e-05,
+      "loss": 0.7318,
+      "step": 1955
+    },
+    {
+      "epoch": 1.1584246372519988,
+      "grad_norm": 1.859336257713254,
+      "learning_rate": 1.405022037694748e-05,
+      "loss": 0.721,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1590168788865858,
+      "grad_norm": 1.815312595979808,
+      "learning_rate": 1.4044371869989024e-05,
+      "loss": 0.7329,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1596091205211727,
+      "grad_norm": 2.126985969178438,
+      "learning_rate": 1.4038521708655054e-05,
+      "loss": 0.7508,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1602013621557596,
+      "grad_norm": 1.5244816275828186,
+      "learning_rate": 1.4032669895338602e-05,
+      "loss": 0.7254,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1607936037903466,
+      "grad_norm": 2.1945138390122834,
+      "learning_rate": 1.40268164324334e-05,
+      "loss": 0.7529,
+      "step": 1960
+    },
+    {
+      "epoch": 1.1613858454249333,
+      "grad_norm": 2.5793485490868235,
+      "learning_rate": 1.4020961322333833e-05,
+      "loss": 0.7489,
+      "step": 1961
+    },
+    {
+      "epoch": 1.1619780870595202,
+      "grad_norm": 1.6233997764262025,
+      "learning_rate": 1.4015104567434981e-05,
+      "loss": 0.7458,
+      "step": 1962
+    },
+    {
+      "epoch": 1.1625703286941071,
+      "grad_norm": 2.1892983877595875,
+      "learning_rate": 1.4009246170132575e-05,
+      "loss": 0.7234,
+      "step": 1963
+    },
+    {
+      "epoch": 1.163162570328694,
+      "grad_norm": 2.2188497921341654,
+      "learning_rate": 1.400338613282304e-05,
+      "loss": 0.7235,
+      "step": 1964
+    },
+    {
+      "epoch": 1.163754811963281,
+      "grad_norm": 2.6410164085265233,
+      "learning_rate": 1.3997524457903455e-05,
+      "loss": 0.7395,
+      "step": 1965
+    },
+    {
+      "epoch": 1.164347053597868,
+      "grad_norm": 2.8137504546585395,
+      "learning_rate": 1.3991661147771574e-05,
+      "loss": 0.725,
+      "step": 1966
+    },
+    {
+      "epoch": 1.1649392952324549,
+      "grad_norm": 4.3683730385339965,
+      "learning_rate": 1.398579620482582e-05,
+      "loss": 0.7569,
+      "step": 1967
+    },
+    {
+      "epoch": 1.1655315368670418,
+      "grad_norm": 3.000428917939229,
+      "learning_rate": 1.3979929631465286e-05,
+      "loss": 0.703,
+      "step": 1968
+    },
+    {
+      "epoch": 1.1661237785016287,
+      "grad_norm": 1.8828332802791297,
+      "learning_rate": 1.3974061430089731e-05,
+      "loss": 0.7377,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1667160201362157,
+      "grad_norm": 1.6402504327655474,
+      "learning_rate": 1.396819160309958e-05,
+      "loss": 0.7423,
+      "step": 1970
+    },
+    {
+      "epoch": 1.1673082617708026,
+      "grad_norm": 6.3377427586159625,
+      "learning_rate": 1.396232015289592e-05,
+      "loss": 0.7424,
+      "step": 1971
+    },
+    {
+      "epoch": 1.1679005034053893,
+      "grad_norm": 2.2909334335027407,
+      "learning_rate": 1.3956447081880506e-05,
+      "loss": 0.7274,
+      "step": 1972
+    },
+    {
+      "epoch": 1.1684927450399762,
+      "grad_norm": 1.7468340066260577,
+      "learning_rate": 1.3950572392455755e-05,
+      "loss": 0.7087,
+      "step": 1973
+    },
+    {
+      "epoch": 1.1690849866745632,
+      "grad_norm": 3.9317619791484644,
+      "learning_rate": 1.394469608702474e-05,
+      "loss": 0.7178,
+      "step": 1974
+    },
+    {
+      "epoch": 1.16967722830915,
+      "grad_norm": 1.8428851547057905,
+      "learning_rate": 1.3938818167991208e-05,
+      "loss": 0.7067,
+      "step": 1975
+    },
+    {
+      "epoch": 1.170269469943737,
+      "grad_norm": 3.147154685015463,
+      "learning_rate": 1.3932938637759555e-05,
+      "loss": 0.7499,
+      "step": 1976
+    },
+    {
+      "epoch": 1.170861711578324,
+      "grad_norm": 5.221103236332467,
+      "learning_rate": 1.3927057498734837e-05,
+      "loss": 0.7523,
+      "step": 1977
+    },
+    {
+      "epoch": 1.171453953212911,
+      "grad_norm": 2.367076528325581,
+      "learning_rate": 1.3921174753322775e-05,
+      "loss": 0.739,
+      "step": 1978
+    },
+    {
+      "epoch": 1.1720461948474978,
+      "grad_norm": 2.216413750813569,
+      "learning_rate": 1.3915290403929738e-05,
+      "loss": 0.732,
+      "step": 1979
+    },
+    {
+      "epoch": 1.1726384364820848,
+      "grad_norm": 2.7444238440546584,
+      "learning_rate": 1.390940445296276e-05,
+      "loss": 0.7295,
+      "step": 1980
+    },
+    {
+      "epoch": 1.1732306781166717,
+      "grad_norm": 3.5442659708991076,
+      "learning_rate": 1.3903516902829525e-05,
+      "loss": 0.7312,
+      "step": 1981
+    },
+    {
+      "epoch": 1.1738229197512586,
+      "grad_norm": 2.4148636767736478,
+      "learning_rate": 1.3897627755938372e-05,
+      "loss": 0.7149,
+      "step": 1982
+    },
+    {
+      "epoch": 1.1744151613858453,
+      "grad_norm": 1.686105294543067,
+      "learning_rate": 1.389173701469829e-05,
+      "loss": 0.7449,
+      "step": 1983
+    },
+    {
+      "epoch": 1.1750074030204323,
+      "grad_norm": 2.18226411659928,
+      "learning_rate": 1.388584468151893e-05,
+      "loss": 0.7074,
+      "step": 1984
+    },
+    {
+      "epoch": 1.1755996446550192,
+      "grad_norm": 1.6462875388280893,
+      "learning_rate": 1.3879950758810577e-05,
+      "loss": 0.7186,
+      "step": 1985
+    },
+    {
+      "epoch": 1.1761918862896061,
+      "grad_norm": 1.9514852968058827,
+      "learning_rate": 1.3874055248984191e-05,
+      "loss": 0.7708,
+      "step": 1986
+    },
+    {
+      "epoch": 1.176784127924193,
+      "grad_norm": 2.5287146246879058,
+      "learning_rate": 1.3868158154451354e-05,
+      "loss": 0.7552,
+      "step": 1987
+    },
+    {
+      "epoch": 1.17737636955878,
+      "grad_norm": 5.354876267561049,
+      "learning_rate": 1.3862259477624317e-05,
+      "loss": 0.7096,
+      "step": 1988
+    },
+    {
+      "epoch": 1.177968611193367,
+      "grad_norm": 2.080724856383468,
+      "learning_rate": 1.3856359220915967e-05,
+      "loss": 0.7306,
+      "step": 1989
+    },
+    {
+      "epoch": 1.1785608528279539,
+      "grad_norm": 1.9442702346562672,
+      "learning_rate": 1.3850457386739846e-05,
+      "loss": 0.7006,
+      "step": 1990
+    },
+    {
+      "epoch": 1.1791530944625408,
+      "grad_norm": 2.217412566401062,
+      "learning_rate": 1.3844553977510127e-05,
+      "loss": 0.7313,
+      "step": 1991
+    },
+    {
+      "epoch": 1.1797453360971275,
+      "grad_norm": 2.3005541144417094,
+      "learning_rate": 1.3838648995641645e-05,
+      "loss": 0.7372,
+      "step": 1992
+    },
+    {
+      "epoch": 1.1803375777317147,
+      "grad_norm": 1.5633590022176485,
+      "learning_rate": 1.3832742443549865e-05,
+      "loss": 0.7273,
+      "step": 1993
+    },
+    {
+      "epoch": 1.1809298193663014,
+      "grad_norm": 2.177628183648929,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.7433,
+      "step": 1994
+    },
+    {
+      "epoch": 1.1815220610008883,
+      "grad_norm": 2.1872512165965596,
+      "learning_rate": 1.3820924638361501e-05,
+      "loss": 0.7884,
+      "step": 1995
+    },
+    {
+      "epoch": 1.1821143026354752,
+      "grad_norm": 2.7491390741203183,
+      "learning_rate": 1.3815013390099068e-05,
+      "loss": 0.77,
+      "step": 1996
+    },
+    {
+      "epoch": 1.1827065442700622,
+      "grad_norm": 2.2399320892471595,
+      "learning_rate": 1.380910058128163e-05,
+      "loss": 0.7449,
+      "step": 1997
+    },
+    {
+      "epoch": 1.183298785904649,
+      "grad_norm": 1.8717444666929153,
+      "learning_rate": 1.3803186214327852e-05,
+      "loss": 0.7234,
+      "step": 1998
+    },
+    {
+      "epoch": 1.183891027539236,
+      "grad_norm": 2.3196057634912934,
+      "learning_rate": 1.3797270291657056e-05,
+      "loss": 0.7311,
+      "step": 1999
+    },
+    {
+      "epoch": 1.184483269173823,
+      "grad_norm": 1.5216917870082234,
+      "learning_rate": 1.3791352815689174e-05,
+      "loss": 0.7444,
+      "step": 2000
+    },
+    {
+      "epoch": 1.18507551080841,
+      "grad_norm": 4.333885469046637,
+      "learning_rate": 1.37854337888448e-05,
+      "loss": 0.7638,
+      "step": 2001
+    },
+    {
+      "epoch": 1.1856677524429968,
+      "grad_norm": 2.019243585845638,
+      "learning_rate": 1.3779513213545132e-05,
+      "loss": 0.7491,
+      "step": 2002
+    },
+    {
+      "epoch": 1.1862599940775835,
+      "grad_norm": 2.577233313193749,
+      "learning_rate": 1.3773591092212035e-05,
+      "loss": 0.7296,
+      "step": 2003
+    },
+    {
+      "epoch": 1.1868522357121707,
+      "grad_norm": 1.9298566377245834,
+      "learning_rate": 1.3767667427267976e-05,
+      "loss": 0.7447,
+      "step": 2004
+    },
+    {
+      "epoch": 1.1874444773467574,
+      "grad_norm": 2.427978631614327,
+      "learning_rate": 1.3761742221136078e-05,
+      "loss": 0.7494,
+      "step": 2005
+    },
+    {
+      "epoch": 1.1880367189813443,
+      "grad_norm": 5.679942493223697,
+      "learning_rate": 1.3755815476240076e-05,
+      "loss": 0.7124,
+      "step": 2006
+    },
+    {
+      "epoch": 1.1886289606159313,
+      "grad_norm": 2.087484384962119,
+      "learning_rate": 1.3749887195004345e-05,
+      "loss": 0.7371,
+      "step": 2007
+    },
+    {
+      "epoch": 1.1892212022505182,
+      "grad_norm": 2.329160941125,
+      "learning_rate": 1.3743957379853885e-05,
+      "loss": 0.7031,
+      "step": 2008
+    },
+    {
+      "epoch": 1.1898134438851051,
+      "grad_norm": 2.8641168542308226,
+      "learning_rate": 1.3738026033214323e-05,
+      "loss": 0.7713,
+      "step": 2009
+    },
+    {
+      "epoch": 1.190405685519692,
+      "grad_norm": 2.7900911750772703,
+      "learning_rate": 1.3732093157511914e-05,
+      "loss": 0.753,
+      "step": 2010
+    },
+    {
+      "epoch": 1.190997927154279,
+      "grad_norm": 1.8472081752565688,
+      "learning_rate": 1.372615875517354e-05,
+      "loss": 0.7305,
+      "step": 2011
+    },
+    {
+      "epoch": 1.191590168788866,
+      "grad_norm": 1.9016156353449258,
+      "learning_rate": 1.3720222828626699e-05,
+      "loss": 0.7529,
+      "step": 2012
+    },
+    {
+      "epoch": 1.1921824104234529,
+      "grad_norm": 1.9348033443818209,
+      "learning_rate": 1.3714285380299525e-05,
+      "loss": 0.7213,
+      "step": 2013
+    },
+    {
+      "epoch": 1.1927746520580396,
+      "grad_norm": 1.371298962671068,
+      "learning_rate": 1.3708346412620768e-05,
+      "loss": 0.7334,
+      "step": 2014
+    },
+    {
+      "epoch": 1.1933668936926265,
+      "grad_norm": 1.4004771260654063,
+      "learning_rate": 1.3702405928019795e-05,
+      "loss": 0.715,
+      "step": 2015
+    },
+    {
+      "epoch": 1.1939591353272134,
+      "grad_norm": 1.6163665195760184,
+      "learning_rate": 1.3696463928926602e-05,
+      "loss": 0.727,
+      "step": 2016
+    },
+    {
+      "epoch": 1.1945513769618004,
+      "grad_norm": 1.8911450014106126,
+      "learning_rate": 1.36905204177718e-05,
+      "loss": 0.7406,
+      "step": 2017
+    },
+    {
+      "epoch": 1.1951436185963873,
+      "grad_norm": 2.6466552053840413,
+      "learning_rate": 1.3684575396986622e-05,
+      "loss": 0.7064,
+      "step": 2018
+    },
+    {
+      "epoch": 1.1957358602309742,
+      "grad_norm": 1.7980339830263137,
+      "learning_rate": 1.3678628869002908e-05,
+      "loss": 0.6718,
+      "step": 2019
+    },
+    {
+      "epoch": 1.1963281018655612,
+      "grad_norm": 2.18736908567566,
+      "learning_rate": 1.3672680836253129e-05,
+      "loss": 0.7137,
+      "step": 2020
+    },
+    {
+      "epoch": 1.196920343500148,
+      "grad_norm": 6.371103342127131,
+      "learning_rate": 1.3666731301170364e-05,
+      "loss": 0.7216,
+      "step": 2021
+    },
+    {
+      "epoch": 1.197512585134735,
+      "grad_norm": 1.7740748237595747,
+      "learning_rate": 1.3660780266188306e-05,
+      "loss": 0.7277,
+      "step": 2022
+    },
+    {
+      "epoch": 1.198104826769322,
+      "grad_norm": 1.437742129128349,
+      "learning_rate": 1.3654827733741263e-05,
+      "loss": 0.7204,
+      "step": 2023
+    },
+    {
+      "epoch": 1.1986970684039089,
+      "grad_norm": 2.9002638880137597,
+      "learning_rate": 1.3648873706264159e-05,
+      "loss": 0.6936,
+      "step": 2024
+    },
+    {
+      "epoch": 1.1992893100384956,
+      "grad_norm": 1.6559528300674338,
+      "learning_rate": 1.3642918186192521e-05,
+      "loss": 0.7051,
+      "step": 2025
+    },
+    {
+      "epoch": 1.1998815516730825,
+      "grad_norm": 2.2195218190520465,
+      "learning_rate": 1.363696117596249e-05,
+      "loss": 0.7208,
+      "step": 2026
+    },
+    {
+      "epoch": 1.2004737933076695,
+      "grad_norm": 1.8165964870897855,
+      "learning_rate": 1.363100267801083e-05,
+      "loss": 0.7576,
+      "step": 2027
+    },
+    {
+      "epoch": 1.2010660349422564,
+      "grad_norm": 2.3943878965553655,
+      "learning_rate": 1.3625042694774886e-05,
+      "loss": 0.7357,
+      "step": 2028
+    },
+    {
+      "epoch": 1.2016582765768433,
+      "grad_norm": 1.7673693572256555,
+      "learning_rate": 1.3619081228692639e-05,
+      "loss": 0.7445,
+      "step": 2029
+    },
+    {
+      "epoch": 1.2022505182114303,
+      "grad_norm": 2.3818913908134585,
+      "learning_rate": 1.3613118282202653e-05,
+      "loss": 0.773,
+      "step": 2030
+    },
+    {
+      "epoch": 1.2028427598460172,
+      "grad_norm": 1.3134819673961666,
+      "learning_rate": 1.360715385774412e-05,
+      "loss": 0.7121,
+      "step": 2031
+    },
+    {
+      "epoch": 1.2034350014806041,
+      "grad_norm": 1.4209296259531388,
+      "learning_rate": 1.3601187957756814e-05,
+      "loss": 0.7476,
+      "step": 2032
+    },
+    {
+      "epoch": 1.204027243115191,
+      "grad_norm": 1.5555865610475328,
+      "learning_rate": 1.3595220584681132e-05,
+      "loss": 0.7267,
+      "step": 2033
+    },
+    {
+      "epoch": 1.204619484749778,
+      "grad_norm": 1.7478667815196771,
+      "learning_rate": 1.358925174095806e-05,
+      "loss": 0.736,
+      "step": 2034
+    },
+    {
+      "epoch": 1.205211726384365,
+      "grad_norm": 1.5866637854107661,
+      "learning_rate": 1.3583281429029197e-05,
+      "loss": 0.734,
+      "step": 2035
+    },
+    {
+      "epoch": 1.2058039680189516,
+      "grad_norm": 1.7725107135676927,
+      "learning_rate": 1.3577309651336728e-05,
+      "loss": 0.7305,
+      "step": 2036
+    },
+    {
+      "epoch": 1.2063962096535386,
+      "grad_norm": 1.623392832320982,
+      "learning_rate": 1.3571336410323448e-05,
+      "loss": 0.7349,
+      "step": 2037
+    },
+    {
+      "epoch": 1.2069884512881255,
+      "grad_norm": 1.6230978180494768,
+      "learning_rate": 1.3565361708432754e-05,
+      "loss": 0.7369,
+      "step": 2038
+    },
+    {
+      "epoch": 1.2075806929227124,
+      "grad_norm": 1.3893660474733283,
+      "learning_rate": 1.3559385548108628e-05,
+      "loss": 0.777,
+      "step": 2039
+    },
+    {
+      "epoch": 1.2081729345572993,
+      "grad_norm": 1.8002057697729996,
+      "learning_rate": 1.3553407931795662e-05,
+      "loss": 0.7434,
+      "step": 2040
+    },
+    {
+      "epoch": 1.2087651761918863,
+      "grad_norm": 2.208673349342069,
+      "learning_rate": 1.3547428861939031e-05,
+      "loss": 0.7277,
+      "step": 2041
+    },
+    {
+      "epoch": 1.2093574178264732,
+      "grad_norm": 2.130797939268362,
+      "learning_rate": 1.3541448340984516e-05,
+      "loss": 0.7517,
+      "step": 2042
+    },
+    {
+      "epoch": 1.2099496594610601,
+      "grad_norm": 1.6356617427047075,
+      "learning_rate": 1.3535466371378483e-05,
+      "loss": 0.7003,
+      "step": 2043
+    },
+    {
+      "epoch": 1.210541901095647,
+      "grad_norm": 1.41442682311907,
+      "learning_rate": 1.3529482955567896e-05,
+      "loss": 0.7375,
+      "step": 2044
+    },
+    {
+      "epoch": 1.211134142730234,
+      "grad_norm": 1.31478939013615,
+      "learning_rate": 1.352349809600031e-05,
+      "loss": 0.7268,
+      "step": 2045
+    },
+    {
+      "epoch": 1.211726384364821,
+      "grad_norm": 2.630006601001849,
+      "learning_rate": 1.3517511795123864e-05,
+      "loss": 0.7314,
+      "step": 2046
+    },
+    {
+      "epoch": 1.2123186259994077,
+      "grad_norm": 2.771345319898994,
+      "learning_rate": 1.3511524055387293e-05,
+      "loss": 0.738,
+      "step": 2047
+    },
+    {
+      "epoch": 1.2129108676339946,
+      "grad_norm": 1.4849713723280222,
+      "learning_rate": 1.3505534879239923e-05,
+      "loss": 0.7009,
+      "step": 2048
+    },
+    {
+      "epoch": 1.2135031092685815,
+      "grad_norm": 1.6458441198708356,
+      "learning_rate": 1.3499544269131662e-05,
+      "loss": 0.7104,
+      "step": 2049
+    },
+    {
+      "epoch": 1.2140953509031684,
+      "grad_norm": 1.7131561551779049,
+      "learning_rate": 1.3493552227513007e-05,
+      "loss": 0.7311,
+      "step": 2050
+    },
+    {
+      "epoch": 1.2146875925377554,
+      "grad_norm": 1.3615653142107633,
+      "learning_rate": 1.3487558756835037e-05,
+      "loss": 0.7221,
+      "step": 2051
+    },
+    {
+      "epoch": 1.2152798341723423,
+      "grad_norm": 1.2965127926167885,
+      "learning_rate": 1.348156385954942e-05,
+      "loss": 0.7176,
+      "step": 2052
+    },
+    {
+      "epoch": 1.2158720758069292,
+      "grad_norm": 1.2844748989522476,
+      "learning_rate": 1.347556753810841e-05,
+      "loss": 0.7511,
+      "step": 2053
+    },
+    {
+      "epoch": 1.2164643174415162,
+      "grad_norm": 1.530448235101446,
+      "learning_rate": 1.3469569794964832e-05,
+      "loss": 0.7388,
+      "step": 2054
+    },
+    {
+      "epoch": 1.217056559076103,
+      "grad_norm": 1.958572967417298,
+      "learning_rate": 1.346357063257211e-05,
+      "loss": 0.706,
+      "step": 2055
+    },
+    {
+      "epoch": 1.21764880071069,
+      "grad_norm": 2.5624134832322514,
+      "learning_rate": 1.3457570053384225e-05,
+      "loss": 0.7516,
+      "step": 2056
+    },
+    {
+      "epoch": 1.218241042345277,
+      "grad_norm": 1.4244459644738547,
+      "learning_rate": 1.3451568059855769e-05,
+      "loss": 0.7369,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2188332839798637,
+      "grad_norm": 1.3769283070535534,
+      "learning_rate": 1.3445564654441879e-05,
+      "loss": 0.7497,
+      "step": 2058
+    },
+    {
+      "epoch": 1.2194255256144506,
+      "grad_norm": 2.074590247262327,
+      "learning_rate": 1.3439559839598296e-05,
+      "loss": 0.7637,
+      "step": 2059
+    },
+    {
+      "epoch": 1.2200177672490375,
+      "grad_norm": 2.0719993078058923,
+      "learning_rate": 1.3433553617781318e-05,
+      "loss": 0.7669,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2206100088836245,
+      "grad_norm": 2.3062477000075248,
+      "learning_rate": 1.3427545991447838e-05,
+      "loss": 0.7561,
+      "step": 2061
+    },
+    {
+      "epoch": 1.2212022505182114,
+      "grad_norm": 1.45455392841872,
+      "learning_rate": 1.3421536963055304e-05,
+      "loss": 0.7921,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2217944921527983,
+      "grad_norm": 1.8743454895734402,
+      "learning_rate": 1.341552653506175e-05,
+      "loss": 0.7586,
+      "step": 2063
+    },
+    {
+      "epoch": 1.2223867337873853,
+      "grad_norm": 1.6979098890708813,
+      "learning_rate": 1.3409514709925777e-05,
+      "loss": 0.741,
+      "step": 2064
+    },
+    {
+      "epoch": 1.2229789754219722,
+      "grad_norm": 3.0439938556284045,
+      "learning_rate": 1.3403501490106562e-05,
+      "loss": 0.753,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2235712170565591,
+      "grad_norm": 1.4161385510159186,
+      "learning_rate": 1.3397486878063852e-05,
+      "loss": 0.7204,
+      "step": 2066
+    },
+    {
+      "epoch": 1.224163458691146,
+      "grad_norm": 1.4943178672108177,
+      "learning_rate": 1.3391470876257957e-05,
+      "loss": 0.7726,
+      "step": 2067
+    },
+    {
+      "epoch": 1.224755700325733,
+      "grad_norm": 9.496258794249528,
+      "learning_rate": 1.3385453487149765e-05,
+      "loss": 0.7337,
+      "step": 2068
+    },
+    {
+      "epoch": 1.2253479419603197,
+      "grad_norm": 2.31673978824299,
+      "learning_rate": 1.3379434713200719e-05,
+      "loss": 0.7277,
+      "step": 2069
+    },
+    {
+      "epoch": 1.2259401835949066,
+      "grad_norm": 4.688994620362354,
+      "learning_rate": 1.3373414556872844e-05,
+      "loss": 0.7385,
+      "step": 2070
+    },
+    {
+      "epoch": 1.2265324252294936,
+      "grad_norm": 2.7363030213596686,
+      "learning_rate": 1.3367393020628716e-05,
+      "loss": 0.7473,
+      "step": 2071
+    },
+    {
+      "epoch": 1.2271246668640805,
+      "grad_norm": 2.074298161390372,
+      "learning_rate": 1.3361370106931486e-05,
+      "loss": 0.7227,
+      "step": 2072
+    },
+    {
+      "epoch": 1.2277169084986674,
+      "grad_norm": 1.790196274702212,
+      "learning_rate": 1.3355345818244864e-05,
+      "loss": 0.7003,
+      "step": 2073
+    },
+    {
+      "epoch": 1.2283091501332544,
+      "grad_norm": 2.2647881453530636,
+      "learning_rate": 1.3349320157033121e-05,
+      "loss": 0.6884,
+      "step": 2074
+    },
+    {
+      "epoch": 1.2289013917678413,
+      "grad_norm": 2.009616238480604,
+      "learning_rate": 1.3343293125761095e-05,
+      "loss": 0.739,
+      "step": 2075
+    },
+    {
+      "epoch": 1.2294936334024282,
+      "grad_norm": 1.5327846716374411,
+      "learning_rate": 1.3337264726894175e-05,
+      "loss": 0.765,
+      "step": 2076
+    },
+    {
+      "epoch": 1.2300858750370152,
+      "grad_norm": 1.5570663444419837,
+      "learning_rate": 1.333123496289832e-05,
+      "loss": 0.8065,
+      "step": 2077
+    },
+    {
+      "epoch": 1.230678116671602,
+      "grad_norm": 9.590583102694426,
+      "learning_rate": 1.3325203836240039e-05,
+      "loss": 0.7578,
+      "step": 2078
+    },
+    {
+      "epoch": 1.231270358306189,
+      "grad_norm": 1.945591127870425,
+      "learning_rate": 1.3319171349386407e-05,
+      "loss": 0.7356,
+      "step": 2079
+    },
+    {
+      "epoch": 1.2318625999407757,
+      "grad_norm": 1.9607332954570817,
+      "learning_rate": 1.3313137504805042e-05,
+      "loss": 0.72,
+      "step": 2080
+    },
+    {
+      "epoch": 1.2324548415753627,
+      "grad_norm": 2.1351386740486387,
+      "learning_rate": 1.3307102304964137e-05,
+      "loss": 0.7345,
+      "step": 2081
+    },
+    {
+      "epoch": 1.2330470832099496,
+      "grad_norm": 2.0536248837102336,
+      "learning_rate": 1.3301065752332415e-05,
+      "loss": 0.7216,
+      "step": 2082
+    },
+    {
+      "epoch": 1.2336393248445365,
+      "grad_norm": 2.055956373479122,
+      "learning_rate": 1.329502784937918e-05,
+      "loss": 0.734,
+      "step": 2083
+    },
+    {
+      "epoch": 1.2342315664791235,
+      "grad_norm": 1.335091728441026,
+      "learning_rate": 1.328898859857426e-05,
+      "loss": 0.6955,
+      "step": 2084
+    },
+    {
+      "epoch": 1.2348238081137104,
+      "grad_norm": 3.0796986220464118,
+      "learning_rate": 1.328294800238806e-05,
+      "loss": 0.6923,
+      "step": 2085
+    },
+    {
+      "epoch": 1.2354160497482973,
+      "grad_norm": 1.4601736651102073,
+      "learning_rate": 1.3276906063291511e-05,
+      "loss": 0.7418,
+      "step": 2086
+    },
+    {
+      "epoch": 1.2360082913828843,
+      "grad_norm": 1.9846393822643258,
+      "learning_rate": 1.327086278375612e-05,
+      "loss": 0.7092,
+      "step": 2087
+    },
+    {
+      "epoch": 1.2366005330174712,
+      "grad_norm": 1.5795884904364481,
+      "learning_rate": 1.3264818166253917e-05,
+      "loss": 0.7119,
+      "step": 2088
+    },
+    {
+      "epoch": 1.2371927746520581,
+      "grad_norm": 3.1635752576713663,
+      "learning_rate": 1.3258772213257493e-05,
+      "loss": 0.724,
+      "step": 2089
+    },
+    {
+      "epoch": 1.237785016286645,
+      "grad_norm": 2.0561446054723573,
+      "learning_rate": 1.3252724927239986e-05,
+      "loss": 0.7008,
+      "step": 2090
+    },
+    {
+      "epoch": 1.2383772579212318,
+      "grad_norm": 1.977573696753446,
+      "learning_rate": 1.3246676310675076e-05,
+      "loss": 0.6925,
+      "step": 2091
+    },
+    {
+      "epoch": 1.2389694995558187,
+      "grad_norm": 1.459694866102705,
+      "learning_rate": 1.3240626366036982e-05,
+      "loss": 0.7235,
+      "step": 2092
+    },
+    {
+      "epoch": 1.2395617411904056,
+      "grad_norm": 1.7170760478119698,
+      "learning_rate": 1.3234575095800477e-05,
+      "loss": 0.672,
+      "step": 2093
+    },
+    {
+      "epoch": 1.2401539828249926,
+      "grad_norm": 2.4802114399082673,
+      "learning_rate": 1.3228522502440868e-05,
+      "loss": 0.7352,
+      "step": 2094
+    },
+    {
+      "epoch": 1.2407462244595795,
+      "grad_norm": 2.0363567883714713,
+      "learning_rate": 1.3222468588434007e-05,
+      "loss": 0.7406,
+      "step": 2095
+    },
+    {
+      "epoch": 1.2413384660941664,
+      "grad_norm": 1.6732345457040516,
+      "learning_rate": 1.3216413356256286e-05,
+      "loss": 0.7671,
+      "step": 2096
+    },
+    {
+      "epoch": 1.2419307077287534,
+      "grad_norm": 2.0841489770262744,
+      "learning_rate": 1.3210356808384634e-05,
+      "loss": 0.7337,
+      "step": 2097
+    },
+    {
+      "epoch": 1.2425229493633403,
+      "grad_norm": 2.033101647120512,
+      "learning_rate": 1.3204298947296521e-05,
+      "loss": 0.7288,
+      "step": 2098
+    },
+    {
+      "epoch": 1.2431151909979272,
+      "grad_norm": 7.475913929909522,
+      "learning_rate": 1.3198239775469953e-05,
+      "loss": 0.7209,
+      "step": 2099
+    },
+    {
+      "epoch": 1.2437074326325142,
+      "grad_norm": 2.093757868519759,
+      "learning_rate": 1.319217929538347e-05,
+      "loss": 0.7292,
+      "step": 2100
+    },
+    {
+      "epoch": 1.244299674267101,
+      "grad_norm": 1.9048288578338362,
+      "learning_rate": 1.318611750951615e-05,
+      "loss": 0.7345,
+      "step": 2101
+    },
+    {
+      "epoch": 1.2448919159016878,
+      "grad_norm": 2.054573126945815,
+      "learning_rate": 1.3180054420347603e-05,
+      "loss": 0.7389,
+      "step": 2102
+    },
+    {
+      "epoch": 1.2454841575362747,
+      "grad_norm": 1.8305207739909872,
+      "learning_rate": 1.317399003035798e-05,
+      "loss": 0.7297,
+      "step": 2103
+    },
+    {
+      "epoch": 1.2460763991708617,
+      "grad_norm": 1.9690959794278322,
+      "learning_rate": 1.3167924342027947e-05,
+      "loss": 0.7013,
+      "step": 2104
+    },
+    {
+      "epoch": 1.2466686408054486,
+      "grad_norm": 2.0772164992233715,
+      "learning_rate": 1.316185735783872e-05,
+      "loss": 0.7077,
+      "step": 2105
+    },
+    {
+      "epoch": 1.2472608824400355,
+      "grad_norm": 2.034723235828488,
+      "learning_rate": 1.315578908027203e-05,
+      "loss": 0.7378,
+      "step": 2106
+    },
+    {
+      "epoch": 1.2478531240746225,
+      "grad_norm": 1.5869955061233847,
+      "learning_rate": 1.3149719511810152e-05,
+      "loss": 0.7371,
+      "step": 2107
+    },
+    {
+      "epoch": 1.2484453657092094,
+      "grad_norm": 3.66716265133577,
+      "learning_rate": 1.3143648654935875e-05,
+      "loss": 0.7766,
+      "step": 2108
+    },
+    {
+      "epoch": 1.2490376073437963,
+      "grad_norm": 2.1574547752973707,
+      "learning_rate": 1.3137576512132524e-05,
+      "loss": 0.7387,
+      "step": 2109
+    },
+    {
+      "epoch": 1.2496298489783833,
+      "grad_norm": 2.0549597904621426,
+      "learning_rate": 1.313150308588394e-05,
+      "loss": 0.7119,
+      "step": 2110
+    },
+    {
+      "epoch": 1.25022209061297,
+      "grad_norm": 1.766968422618495,
+      "learning_rate": 1.3125428378674507e-05,
+      "loss": 0.7619,
+      "step": 2111
+    },
+    {
+      "epoch": 1.2508143322475571,
+      "grad_norm": 1.6504903814709435,
+      "learning_rate": 1.311935239298911e-05,
+      "loss": 0.7339,
+      "step": 2112
+    },
+    {
+      "epoch": 1.2514065738821438,
+      "grad_norm": 2.3755573371765846,
+      "learning_rate": 1.3113275131313179e-05,
+      "loss": 0.7706,
+      "step": 2113
+    },
+    {
+      "epoch": 1.2519988155167308,
+      "grad_norm": 6.419415940537187,
+      "learning_rate": 1.310719659613265e-05,
+      "loss": 0.7298,
+      "step": 2114
+    },
+    {
+      "epoch": 1.2525910571513177,
+      "grad_norm": 4.777728911657151,
+      "learning_rate": 1.3101116789933988e-05,
+      "loss": 0.7351,
+      "step": 2115
+    },
+    {
+      "epoch": 1.2531832987859046,
+      "grad_norm": 1.9830147062841872,
+      "learning_rate": 1.3095035715204171e-05,
+      "loss": 0.7352,
+      "step": 2116
+    },
+    {
+      "epoch": 1.2537755404204916,
+      "grad_norm": 3.8969606332634212,
+      "learning_rate": 1.3088953374430709e-05,
+      "loss": 0.7565,
+      "step": 2117
+    },
+    {
+      "epoch": 1.2543677820550785,
+      "grad_norm": 3.895878218256781,
+      "learning_rate": 1.3082869770101613e-05,
+      "loss": 0.737,
+      "step": 2118
+    },
+    {
+      "epoch": 1.2549600236896654,
+      "grad_norm": 3.0246769337442725,
+      "learning_rate": 1.3076784904705426e-05,
+      "loss": 0.7413,
+      "step": 2119
+    },
+    {
+      "epoch": 1.2555522653242523,
+      "grad_norm": 6.0066054184240105,
+      "learning_rate": 1.3070698780731194e-05,
+      "loss": 0.7304,
+      "step": 2120
+    },
+    {
+      "epoch": 1.2561445069588393,
+      "grad_norm": 3.00887439367028,
+      "learning_rate": 1.306461140066849e-05,
+      "loss": 0.7248,
+      "step": 2121
+    },
+    {
+      "epoch": 1.256736748593426,
+      "grad_norm": 2.3778171464319318,
+      "learning_rate": 1.305852276700739e-05,
+      "loss": 0.7199,
+      "step": 2122
+    },
+    {
+      "epoch": 1.2573289902280131,
+      "grad_norm": 3.933106640277816,
+      "learning_rate": 1.3052432882238487e-05,
+      "loss": 0.7539,
+      "step": 2123
+    },
+    {
+      "epoch": 1.2579212318625999,
+      "grad_norm": 12.537942854482242,
+      "learning_rate": 1.304634174885289e-05,
+      "loss": 0.7503,
+      "step": 2124
+    },
+    {
+      "epoch": 1.2585134734971868,
+      "grad_norm": 3.6650070710135174,
+      "learning_rate": 1.3040249369342215e-05,
+      "loss": 0.7451,
+      "step": 2125
+    },
+    {
+      "epoch": 1.2591057151317737,
+      "grad_norm": 3.320172986670575,
+      "learning_rate": 1.3034155746198588e-05,
+      "loss": 0.692,
+      "step": 2126
+    },
+    {
+      "epoch": 1.2596979567663606,
+      "grad_norm": 2.414366162856636,
+      "learning_rate": 1.3028060881914639e-05,
+      "loss": 0.7276,
+      "step": 2127
+    },
+    {
+      "epoch": 1.2602901984009476,
+      "grad_norm": 1.845695162557533,
+      "learning_rate": 1.3021964778983513e-05,
+      "loss": 0.7083,
+      "step": 2128
+    },
+    {
+      "epoch": 1.2608824400355345,
+      "grad_norm": 3.8714818295986957,
+      "learning_rate": 1.3015867439898862e-05,
+      "loss": 0.7238,
+      "step": 2129
+    },
+    {
+      "epoch": 1.2614746816701214,
+      "grad_norm": 3.0251438217693205,
+      "learning_rate": 1.3009768867154834e-05,
+      "loss": 0.7342,
+      "step": 2130
+    },
+    {
+      "epoch": 1.2620669233047084,
+      "grad_norm": 3.7078771253414717,
+      "learning_rate": 1.3003669063246096e-05,
+      "loss": 0.7253,
+      "step": 2131
+    },
+    {
+      "epoch": 1.2626591649392953,
+      "grad_norm": 1.6876561495268882,
+      "learning_rate": 1.2997568030667802e-05,
+      "loss": 0.7286,
+      "step": 2132
+    },
+    {
+      "epoch": 1.263251406573882,
+      "grad_norm": 2.3907642362721835,
+      "learning_rate": 1.2991465771915626e-05,
+      "loss": 0.7579,
+      "step": 2133
+    },
+    {
+      "epoch": 1.2638436482084692,
+      "grad_norm": 2.5589770747316867,
+      "learning_rate": 1.2985362289485728e-05,
+      "loss": 0.7411,
+      "step": 2134
+    },
+    {
+      "epoch": 1.2644358898430559,
+      "grad_norm": 1.670540837036476,
+      "learning_rate": 1.2979257585874782e-05,
+      "loss": 0.7457,
+      "step": 2135
+    },
+    {
+      "epoch": 1.2650281314776428,
+      "grad_norm": 2.1619277987058525,
+      "learning_rate": 1.2973151663579948e-05,
+      "loss": 0.7158,
+      "step": 2136
+    },
+    {
+      "epoch": 1.2656203731122297,
+      "grad_norm": 4.927979400324849,
+      "learning_rate": 1.2967044525098897e-05,
+      "loss": 0.7082,
+      "step": 2137
+    },
+    {
+      "epoch": 1.2662126147468167,
+      "grad_norm": 1.7102954049147032,
+      "learning_rate": 1.296093617292979e-05,
+      "loss": 0.731,
+      "step": 2138
+    },
+    {
+      "epoch": 1.2668048563814036,
+      "grad_norm": 4.087043152351807,
+      "learning_rate": 1.295482660957129e-05,
+      "loss": 0.7433,
+      "step": 2139
+    },
+    {
+      "epoch": 1.2673970980159905,
+      "grad_norm": 7.571384572693805,
+      "learning_rate": 1.2948715837522542e-05,
+      "loss": 0.7354,
+      "step": 2140
+    },
+    {
+      "epoch": 1.2679893396505775,
+      "grad_norm": 2.3629194134848928,
+      "learning_rate": 1.2942603859283207e-05,
+      "loss": 0.7327,
+      "step": 2141
+    },
+    {
+      "epoch": 1.2685815812851644,
+      "grad_norm": 2.8525406546001415,
+      "learning_rate": 1.2936490677353422e-05,
+      "loss": 0.7268,
+      "step": 2142
+    },
+    {
+      "epoch": 1.2691738229197513,
+      "grad_norm": 14.104534903708966,
+      "learning_rate": 1.2930376294233821e-05,
+      "loss": 0.723,
+      "step": 2143
+    },
+    {
+      "epoch": 1.269766064554338,
+      "grad_norm": 2.488060816626756,
+      "learning_rate": 1.2924260712425536e-05,
+      "loss": 0.7055,
+      "step": 2144
+    },
+    {
+      "epoch": 1.2703583061889252,
+      "grad_norm": 2.281444459835155,
+      "learning_rate": 1.2918143934430178e-05,
+      "loss": 0.7358,
+      "step": 2145
+    },
+    {
+      "epoch": 1.270950547823512,
+      "grad_norm": 2.863071025708633,
+      "learning_rate": 1.2912025962749856e-05,
+      "loss": 0.7511,
+      "step": 2146
+    },
+    {
+      "epoch": 1.2715427894580988,
+      "grad_norm": 3.4445132717535145,
+      "learning_rate": 1.2905906799887164e-05,
+      "loss": 0.7211,
+      "step": 2147
+    },
+    {
+      "epoch": 1.2721350310926858,
+      "grad_norm": 2.992321062019045,
+      "learning_rate": 1.2899786448345186e-05,
+      "loss": 0.7323,
+      "step": 2148
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 2.7754937679502767,
+      "learning_rate": 1.2893664910627486e-05,
+      "loss": 0.6995,
+      "step": 2149
+    },
+    {
+      "epoch": 1.2733195143618596,
+      "grad_norm": 4.043526562299961,
+      "learning_rate": 1.288754218923812e-05,
+      "loss": 0.7038,
+      "step": 2150
+    },
+    {
+      "epoch": 1.2739117559964466,
+      "grad_norm": 2.601133025890244,
+      "learning_rate": 1.2881418286681622e-05,
+      "loss": 0.7592,
+      "step": 2151
+    },
+    {
+      "epoch": 1.2745039976310335,
+      "grad_norm": 3.228479699181939,
+      "learning_rate": 1.2875293205463018e-05,
+      "loss": 0.7457,
+      "step": 2152
+    },
+    {
+      "epoch": 1.2750962392656204,
+      "grad_norm": 5.282129958437338,
+      "learning_rate": 1.2869166948087804e-05,
+      "loss": 0.7341,
+      "step": 2153
+    },
+    {
+      "epoch": 1.2756884809002074,
+      "grad_norm": 12.516944032609397,
+      "learning_rate": 1.2863039517061968e-05,
+      "loss": 0.7495,
+      "step": 2154
+    },
+    {
+      "epoch": 1.276280722534794,
+      "grad_norm": 4.462130212698797,
+      "learning_rate": 1.2856910914891973e-05,
+      "loss": 0.7348,
+      "step": 2155
+    },
+    {
+      "epoch": 1.2768729641693812,
+      "grad_norm": 4.499459533831295,
+      "learning_rate": 1.2850781144084763e-05,
+      "loss": 0.7377,
+      "step": 2156
+    },
+    {
+      "epoch": 1.277465205803968,
+      "grad_norm": 4.366118773756862,
+      "learning_rate": 1.284465020714776e-05,
+      "loss": 0.7213,
+      "step": 2157
+    },
+    {
+      "epoch": 1.2780574474385549,
+      "grad_norm": 5.796358264091171,
+      "learning_rate": 1.2838518106588856e-05,
+      "loss": 0.7296,
+      "step": 2158
+    },
+    {
+      "epoch": 1.2786496890731418,
+      "grad_norm": 3.051441847538537,
+      "learning_rate": 1.2832384844916433e-05,
+      "loss": 0.7283,
+      "step": 2159
+    },
+    {
+      "epoch": 1.2792419307077287,
+      "grad_norm": 2.3238935491908865,
+      "learning_rate": 1.2826250424639329e-05,
+      "loss": 0.7258,
+      "step": 2160
+    },
+    {
+      "epoch": 1.2798341723423157,
+      "grad_norm": 10.583266939609347,
+      "learning_rate": 1.282011484826688e-05,
+      "loss": 0.6906,
+      "step": 2161
+    },
+    {
+      "epoch": 1.2804264139769026,
+      "grad_norm": 1.9613859782854648,
+      "learning_rate": 1.2813978118308872e-05,
+      "loss": 0.7381,
+      "step": 2162
+    },
+    {
+      "epoch": 1.2810186556114895,
+      "grad_norm": 1.5522438888542716,
+      "learning_rate": 1.2807840237275578e-05,
+      "loss": 0.7759,
+      "step": 2163
+    },
+    {
+      "epoch": 1.2816108972460765,
+      "grad_norm": 2.0620688237603866,
+      "learning_rate": 1.2801701207677731e-05,
+      "loss": 0.7275,
+      "step": 2164
+    },
+    {
+      "epoch": 1.2822031388806634,
+      "grad_norm": 4.033929396881479,
+      "learning_rate": 1.2795561032026546e-05,
+      "loss": 0.7161,
+      "step": 2165
+    },
+    {
+      "epoch": 1.28279538051525,
+      "grad_norm": 2.2174066380945128,
+      "learning_rate": 1.2789419712833698e-05,
+      "loss": 0.7347,
+      "step": 2166
+    },
+    {
+      "epoch": 1.2833876221498373,
+      "grad_norm": 2.989193169042572,
+      "learning_rate": 1.2783277252611326e-05,
+      "loss": 0.744,
+      "step": 2167
+    },
+    {
+      "epoch": 1.283979863784424,
+      "grad_norm": 3.7142286905163817,
+      "learning_rate": 1.277713365387205e-05,
+      "loss": 0.711,
+      "step": 2168
+    },
+    {
+      "epoch": 1.284572105419011,
+      "grad_norm": 5.139644730862864,
+      "learning_rate": 1.2770988919128943e-05,
+      "loss": 0.7386,
+      "step": 2169
+    },
+    {
+      "epoch": 1.2851643470535978,
+      "grad_norm": 6.589804111626187,
+      "learning_rate": 1.2764843050895548e-05,
+      "loss": 0.7376,
+      "step": 2170
+    },
+    {
+      "epoch": 1.2857565886881848,
+      "grad_norm": 7.048157778907931,
+      "learning_rate": 1.275869605168587e-05,
+      "loss": 0.726,
+      "step": 2171
+    },
+    {
+      "epoch": 1.2863488303227717,
+      "grad_norm": 4.519497731902214,
+      "learning_rate": 1.2752547924014378e-05,
+      "loss": 0.7176,
+      "step": 2172
+    },
+    {
+      "epoch": 1.2869410719573586,
+      "grad_norm": 4.560564674061755,
+      "learning_rate": 1.2746398670396003e-05,
+      "loss": 0.7548,
+      "step": 2173
+    },
+    {
+      "epoch": 1.2875333135919456,
+      "grad_norm": 2.476033120505578,
+      "learning_rate": 1.2740248293346134e-05,
+      "loss": 0.6937,
+      "step": 2174
+    },
+    {
+      "epoch": 1.2881255552265325,
+      "grad_norm": 1.711083220222386,
+      "learning_rate": 1.2734096795380619e-05,
+      "loss": 0.7451,
+      "step": 2175
+    },
+    {
+      "epoch": 1.2887177968611194,
+      "grad_norm": 2.308003756437765,
+      "learning_rate": 1.2727944179015773e-05,
+      "loss": 0.7071,
+      "step": 2176
+    },
+    {
+      "epoch": 1.2893100384957061,
+      "grad_norm": 1.9018295125948241,
+      "learning_rate": 1.2721790446768355e-05,
+      "loss": 0.7064,
+      "step": 2177
+    },
+    {
+      "epoch": 1.2899022801302933,
+      "grad_norm": 3.3632041500633663,
+      "learning_rate": 1.271563560115559e-05,
+      "loss": 0.756,
+      "step": 2178
+    },
+    {
+      "epoch": 1.29049452176488,
+      "grad_norm": 9.199019371454177,
+      "learning_rate": 1.2709479644695157e-05,
+      "loss": 0.7183,
+      "step": 2179
+    },
+    {
+      "epoch": 1.291086763399467,
+      "grad_norm": 3.5389831036172223,
+      "learning_rate": 1.2703322579905191e-05,
+      "loss": 0.717,
+      "step": 2180
+    },
+    {
+      "epoch": 1.2916790050340539,
+      "grad_norm": 7.54002229069042,
+      "learning_rate": 1.2697164409304273e-05,
+      "loss": 0.7707,
+      "step": 2181
+    },
+    {
+      "epoch": 1.2922712466686408,
+      "grad_norm": 1.7866494430336994,
+      "learning_rate": 1.269100513541144e-05,
+      "loss": 0.7525,
+      "step": 2182
+    },
+    {
+      "epoch": 1.2928634883032277,
+      "grad_norm": 1.4683566571838764,
+      "learning_rate": 1.2684844760746188e-05,
+      "loss": 0.741,
+      "step": 2183
+    },
+    {
+      "epoch": 1.2934557299378147,
+      "grad_norm": 1.8765066274808877,
+      "learning_rate": 1.2678683287828451e-05,
+      "loss": 0.7657,
+      "step": 2184
+    },
+    {
+      "epoch": 1.2940479715724016,
+      "grad_norm": 3.013366849613377,
+      "learning_rate": 1.267252071917862e-05,
+      "loss": 0.7016,
+      "step": 2185
+    },
+    {
+      "epoch": 1.2946402132069885,
+      "grad_norm": 1.7177077320985006,
+      "learning_rate": 1.266635705731753e-05,
+      "loss": 0.692,
+      "step": 2186
+    },
+    {
+      "epoch": 1.2952324548415755,
+      "grad_norm": 3.9718676101648813,
+      "learning_rate": 1.266019230476647e-05,
+      "loss": 0.6845,
+      "step": 2187
+    },
+    {
+      "epoch": 1.2958246964761622,
+      "grad_norm": 7.981490810275518,
+      "learning_rate": 1.2654026464047165e-05,
+      "loss": 0.6967,
+      "step": 2188
+    },
+    {
+      "epoch": 1.2964169381107493,
+      "grad_norm": 1.6653196455250276,
+      "learning_rate": 1.2647859537681794e-05,
+      "loss": 0.7264,
+      "step": 2189
+    },
+    {
+      "epoch": 1.297009179745336,
+      "grad_norm": 3.229577009236302,
+      "learning_rate": 1.2641691528192976e-05,
+      "loss": 0.7175,
+      "step": 2190
+    },
+    {
+      "epoch": 1.297601421379923,
+      "grad_norm": 1.8793833759264493,
+      "learning_rate": 1.2635522438103775e-05,
+      "loss": 0.7383,
+      "step": 2191
+    },
+    {
+      "epoch": 1.29819366301451,
+      "grad_norm": 6.074008068005565,
+      "learning_rate": 1.262935226993769e-05,
+      "loss": 0.716,
+      "step": 2192
+    },
+    {
+      "epoch": 1.2987859046490968,
+      "grad_norm": 3.365277587578852,
+      "learning_rate": 1.2623181026218676e-05,
+      "loss": 0.7079,
+      "step": 2193
+    },
+    {
+      "epoch": 1.2993781462836838,
+      "grad_norm": 3.8730032760836712,
+      "learning_rate": 1.261700870947111e-05,
+      "loss": 0.7094,
+      "step": 2194
+    },
+    {
+      "epoch": 1.2999703879182707,
+      "grad_norm": 2.4954320743950205,
+      "learning_rate": 1.2610835322219822e-05,
+      "loss": 0.7249,
+      "step": 2195
+    },
+    {
+      "epoch": 1.3005626295528576,
+      "grad_norm": 2.328817696663847,
+      "learning_rate": 1.2604660866990072e-05,
+      "loss": 0.7027,
+      "step": 2196
+    },
+    {
+      "epoch": 1.3011548711874446,
+      "grad_norm": 1.7371517673037984,
+      "learning_rate": 1.259848534630756e-05,
+      "loss": 0.7134,
+      "step": 2197
+    },
+    {
+      "epoch": 1.3017471128220315,
+      "grad_norm": 2.4321062518581096,
+      "learning_rate": 1.2592308762698422e-05,
+      "loss": 0.7529,
+      "step": 2198
+    },
+    {
+      "epoch": 1.3023393544566182,
+      "grad_norm": 2.551030855888438,
+      "learning_rate": 1.2586131118689229e-05,
+      "loss": 0.7395,
+      "step": 2199
+    },
+    {
+      "epoch": 1.3029315960912053,
+      "grad_norm": 2.5192997738692253,
+      "learning_rate": 1.257995241680698e-05,
+      "loss": 0.767,
+      "step": 2200
+    },
+    {
+      "epoch": 1.303523837725792,
+      "grad_norm": 1.871844395912189,
+      "learning_rate": 1.2573772659579118e-05,
+      "loss": 0.7456,
+      "step": 2201
+    },
+    {
+      "epoch": 1.304116079360379,
+      "grad_norm": 2.1558955925829903,
+      "learning_rate": 1.2567591849533507e-05,
+      "loss": 0.7714,
+      "step": 2202
+    },
+    {
+      "epoch": 1.304708320994966,
+      "grad_norm": 1.5443144366947106,
+      "learning_rate": 1.2561409989198445e-05,
+      "loss": 0.7153,
+      "step": 2203
+    },
+    {
+      "epoch": 1.3053005626295529,
+      "grad_norm": 2.893987768096241,
+      "learning_rate": 1.2555227081102663e-05,
+      "loss": 0.7501,
+      "step": 2204
+    },
+    {
+      "epoch": 1.3058928042641398,
+      "grad_norm": 2.9864438149165293,
+      "learning_rate": 1.2549043127775317e-05,
+      "loss": 0.7392,
+      "step": 2205
+    },
+    {
+      "epoch": 1.3064850458987267,
+      "grad_norm": 2.2495043135725044,
+      "learning_rate": 1.2542858131745997e-05,
+      "loss": 0.7344,
+      "step": 2206
+    },
+    {
+      "epoch": 1.3070772875333136,
+      "grad_norm": 1.8073617958458001,
+      "learning_rate": 1.2536672095544705e-05,
+      "loss": 0.739,
+      "step": 2207
+    },
+    {
+      "epoch": 1.3076695291679006,
+      "grad_norm": 5.482475430635642,
+      "learning_rate": 1.253048502170188e-05,
+      "loss": 0.7081,
+      "step": 2208
+    },
+    {
+      "epoch": 1.3082617708024875,
+      "grad_norm": 1.6369830148297229,
+      "learning_rate": 1.2524296912748391e-05,
+      "loss": 0.743,
+      "step": 2209
+    },
+    {
+      "epoch": 1.3088540124370742,
+      "grad_norm": 2.3037478748956457,
+      "learning_rate": 1.2518107771215511e-05,
+      "loss": 0.7013,
+      "step": 2210
+    },
+    {
+      "epoch": 1.3094462540716614,
+      "grad_norm": 2.2512579980860727,
+      "learning_rate": 1.2511917599634957e-05,
+      "loss": 0.7202,
+      "step": 2211
+    },
+    {
+      "epoch": 1.310038495706248,
+      "grad_norm": 3.811836598474259,
+      "learning_rate": 1.2505726400538849e-05,
+      "loss": 0.735,
+      "step": 2212
+    },
+    {
+      "epoch": 1.310630737340835,
+      "grad_norm": 3.888905349606496,
+      "learning_rate": 1.2499534176459742e-05,
+      "loss": 0.7649,
+      "step": 2213
+    },
+    {
+      "epoch": 1.311222978975422,
+      "grad_norm": 2.6409628828714062,
+      "learning_rate": 1.24933409299306e-05,
+      "loss": 0.7387,
+      "step": 2214
+    },
+    {
+      "epoch": 1.3118152206100089,
+      "grad_norm": 1.1361250179762734,
+      "learning_rate": 1.2487146663484811e-05,
+      "loss": 0.7357,
+      "step": 2215
+    },
+    {
+      "epoch": 1.3124074622445958,
+      "grad_norm": 3.0048958781591812,
+      "learning_rate": 1.2480951379656175e-05,
+      "loss": 0.7016,
+      "step": 2216
+    },
+    {
+      "epoch": 1.3129997038791827,
+      "grad_norm": 3.618139046477133,
+      "learning_rate": 1.247475508097892e-05,
+      "loss": 0.7279,
+      "step": 2217
+    },
+    {
+      "epoch": 1.3135919455137697,
+      "grad_norm": 5.253731958808833,
+      "learning_rate": 1.2468557769987667e-05,
+      "loss": 0.749,
+      "step": 2218
+    },
+    {
+      "epoch": 1.3141841871483566,
+      "grad_norm": 1.6888707222430752,
+      "learning_rate": 1.2462359449217478e-05,
+      "loss": 0.7063,
+      "step": 2219
+    },
+    {
+      "epoch": 1.3147764287829435,
+      "grad_norm": 1.8167018428081732,
+      "learning_rate": 1.2456160121203808e-05,
+      "loss": 0.6972,
+      "step": 2220
+    },
+    {
+      "epoch": 1.3153686704175303,
+      "grad_norm": 3.091651135895501,
+      "learning_rate": 1.244995978848253e-05,
+      "loss": 0.7003,
+      "step": 2221
+    },
+    {
+      "epoch": 1.3159609120521172,
+      "grad_norm": 1.5525088892116699,
+      "learning_rate": 1.2443758453589934e-05,
+      "loss": 0.722,
+      "step": 2222
+    },
+    {
+      "epoch": 1.3165531536867041,
+      "grad_norm": 2.7538503380997894,
+      "learning_rate": 1.2437556119062707e-05,
+      "loss": 0.7218,
+      "step": 2223
+    },
+    {
+      "epoch": 1.317145395321291,
+      "grad_norm": 2.232065009961151,
+      "learning_rate": 1.243135278743796e-05,
+      "loss": 0.7338,
+      "step": 2224
+    },
+    {
+      "epoch": 1.317737636955878,
+      "grad_norm": 1.4109391466878773,
+      "learning_rate": 1.24251484612532e-05,
+      "loss": 0.7339,
+      "step": 2225
+    },
+    {
+      "epoch": 1.318329878590465,
+      "grad_norm": 5.387146562198588,
+      "learning_rate": 1.2418943143046346e-05,
+      "loss": 0.7163,
+      "step": 2226
+    },
+    {
+      "epoch": 1.3189221202250518,
+      "grad_norm": 7.336494868885874,
+      "learning_rate": 1.2412736835355725e-05,
+      "loss": 0.7206,
+      "step": 2227
+    },
+    {
+      "epoch": 1.3195143618596388,
+      "grad_norm": 5.47208025176641,
+      "learning_rate": 1.2406529540720063e-05,
+      "loss": 0.7154,
+      "step": 2228
+    },
+    {
+      "epoch": 1.3201066034942257,
+      "grad_norm": 2.6695992583732946,
+      "learning_rate": 1.2400321261678492e-05,
+      "loss": 0.7236,
+      "step": 2229
+    },
+    {
+      "epoch": 1.3206988451288126,
+      "grad_norm": 2.195132600070081,
+      "learning_rate": 1.239411200077055e-05,
+      "loss": 0.7574,
+      "step": 2230
+    },
+    {
+      "epoch": 1.3212910867633996,
+      "grad_norm": 11.165845514321358,
+      "learning_rate": 1.238790176053617e-05,
+      "loss": 0.7546,
+      "step": 2231
+    },
+    {
+      "epoch": 1.3218833283979863,
+      "grad_norm": 2.9312352374249864,
+      "learning_rate": 1.2381690543515692e-05,
+      "loss": 0.7301,
+      "step": 2232
+    },
+    {
+      "epoch": 1.3224755700325732,
+      "grad_norm": 1.6545166318992972,
+      "learning_rate": 1.2375478352249854e-05,
+      "loss": 0.7211,
+      "step": 2233
+    },
+    {
+      "epoch": 1.3230678116671601,
+      "grad_norm": 2.1861482000851518,
+      "learning_rate": 1.236926518927979e-05,
+      "loss": 0.6936,
+      "step": 2234
+    },
+    {
+      "epoch": 1.323660053301747,
+      "grad_norm": 1.8651361150526538,
+      "learning_rate": 1.2363051057147036e-05,
+      "loss": 0.7116,
+      "step": 2235
+    },
+    {
+      "epoch": 1.324252294936334,
+      "grad_norm": 2.1494204464514923,
+      "learning_rate": 1.2356835958393513e-05,
+      "loss": 0.746,
+      "step": 2236
+    },
+    {
+      "epoch": 1.324844536570921,
+      "grad_norm": 1.5645161950911375,
+      "learning_rate": 1.2350619895561557e-05,
+      "loss": 0.7028,
+      "step": 2237
+    },
+    {
+      "epoch": 1.3254367782055079,
+      "grad_norm": 1.7327580503628912,
+      "learning_rate": 1.2344402871193876e-05,
+      "loss": 0.7459,
+      "step": 2238
+    },
+    {
+      "epoch": 1.3260290198400948,
+      "grad_norm": 2.9366360287029085,
+      "learning_rate": 1.2338184887833595e-05,
+      "loss": 0.7929,
+      "step": 2239
+    },
+    {
+      "epoch": 1.3266212614746817,
+      "grad_norm": 2.234776359408874,
+      "learning_rate": 1.2331965948024209e-05,
+      "loss": 0.741,
+      "step": 2240
+    },
+    {
+      "epoch": 1.3272135031092684,
+      "grad_norm": 16.615421551346895,
+      "learning_rate": 1.232574605430962e-05,
+      "loss": 0.7348,
+      "step": 2241
+    },
+    {
+      "epoch": 1.3278057447438556,
+      "grad_norm": 1.836805999142044,
+      "learning_rate": 1.2319525209234109e-05,
+      "loss": 0.7431,
+      "step": 2242
+    },
+    {
+      "epoch": 1.3283979863784423,
+      "grad_norm": 3.6059103156218217,
+      "learning_rate": 1.2313303415342358e-05,
+      "loss": 0.702,
+      "step": 2243
+    },
+    {
+      "epoch": 1.3289902280130292,
+      "grad_norm": 2.494436178191213,
+      "learning_rate": 1.230708067517942e-05,
+      "loss": 0.7407,
+      "step": 2244
+    },
+    {
+      "epoch": 1.3295824696476162,
+      "grad_norm": 3.946731442600297,
+      "learning_rate": 1.230085699129076e-05,
+      "loss": 0.6811,
+      "step": 2245
+    },
+    {
+      "epoch": 1.330174711282203,
+      "grad_norm": 4.20508246282007,
+      "learning_rate": 1.2294632366222201e-05,
+      "loss": 0.7007,
+      "step": 2246
+    },
+    {
+      "epoch": 1.33076695291679,
+      "grad_norm": 6.980679537862269,
+      "learning_rate": 1.2288406802519974e-05,
+      "loss": 0.6982,
+      "step": 2247
+    },
+    {
+      "epoch": 1.331359194551377,
+      "grad_norm": 2.0063581908355195,
+      "learning_rate": 1.2282180302730683e-05,
+      "loss": 0.7171,
+      "step": 2248
+    },
+    {
+      "epoch": 1.331951436185964,
+      "grad_norm": 2.812205651172888,
+      "learning_rate": 1.2275952869401311e-05,
+      "loss": 0.7062,
+      "step": 2249
+    },
+    {
+      "epoch": 1.3325436778205508,
+      "grad_norm": 4.260513166725431,
+      "learning_rate": 1.2269724505079234e-05,
+      "loss": 0.7677,
+      "step": 2250
+    },
+    {
+      "epoch": 1.3331359194551378,
+      "grad_norm": 2.395786745837272,
+      "learning_rate": 1.2263495212312202e-05,
+      "loss": 0.7321,
+      "step": 2251
+    },
+    {
+      "epoch": 1.3337281610897245,
+      "grad_norm": 3.4523356684471933,
+      "learning_rate": 1.2257264993648345e-05,
+      "loss": 0.7213,
+      "step": 2252
+    },
+    {
+      "epoch": 1.3343204027243116,
+      "grad_norm": 1.6680835833538643,
+      "learning_rate": 1.2251033851636174e-05,
+      "loss": 0.7316,
+      "step": 2253
+    },
+    {
+      "epoch": 1.3349126443588983,
+      "grad_norm": 4.066851683497322,
+      "learning_rate": 1.2244801788824577e-05,
+      "loss": 0.7494,
+      "step": 2254
+    },
+    {
+      "epoch": 1.3355048859934853,
+      "grad_norm": 2.2024623170572575,
+      "learning_rate": 1.2238568807762813e-05,
+      "loss": 0.7078,
+      "step": 2255
+    },
+    {
+      "epoch": 1.3360971276280722,
+      "grad_norm": 4.096509077552267,
+      "learning_rate": 1.223233491100053e-05,
+      "loss": 0.6927,
+      "step": 2256
+    },
+    {
+      "epoch": 1.3366893692626591,
+      "grad_norm": 1.7054951183155915,
+      "learning_rate": 1.2226100101087737e-05,
+      "loss": 0.744,
+      "step": 2257
+    },
+    {
+      "epoch": 1.337281610897246,
+      "grad_norm": 2.256752220268974,
+      "learning_rate": 1.2219864380574822e-05,
+      "loss": 0.7134,
+      "step": 2258
+    },
+    {
+      "epoch": 1.337873852531833,
+      "grad_norm": 2.6221532584973217,
+      "learning_rate": 1.2213627752012547e-05,
+      "loss": 0.7147,
+      "step": 2259
+    },
+    {
+      "epoch": 1.33846609416642,
+      "grad_norm": 3.3038467185814953,
+      "learning_rate": 1.2207390217952044e-05,
+      "loss": 0.7271,
+      "step": 2260
+    },
+    {
+      "epoch": 1.3390583358010069,
+      "grad_norm": 1.9956526626820097,
+      "learning_rate": 1.2201151780944813e-05,
+      "loss": 0.7499,
+      "step": 2261
+    },
+    {
+      "epoch": 1.3396505774355938,
+      "grad_norm": 2.58029133331588,
+      "learning_rate": 1.2194912443542728e-05,
+      "loss": 0.7362,
+      "step": 2262
+    },
+    {
+      "epoch": 1.3402428190701805,
+      "grad_norm": 1.3093228436379185,
+      "learning_rate": 1.2188672208298028e-05,
+      "loss": 0.7464,
+      "step": 2263
+    },
+    {
+      "epoch": 1.3408350607047677,
+      "grad_norm": 2.2716810173689943,
+      "learning_rate": 1.2182431077763317e-05,
+      "loss": 0.7151,
+      "step": 2264
+    },
+    {
+      "epoch": 1.3414273023393544,
+      "grad_norm": 3.7636447060204112,
+      "learning_rate": 1.2176189054491576e-05,
+      "loss": 0.7064,
+      "step": 2265
+    },
+    {
+      "epoch": 1.3420195439739413,
+      "grad_norm": 1.6344445874345244,
+      "learning_rate": 1.2169946141036133e-05,
+      "loss": 0.7248,
+      "step": 2266
+    },
+    {
+      "epoch": 1.3426117856085282,
+      "grad_norm": 1.660994198119721,
+      "learning_rate": 1.2163702339950702e-05,
+      "loss": 0.7344,
+      "step": 2267
+    },
+    {
+      "epoch": 1.3432040272431152,
+      "grad_norm": 2.3481480865911912,
+      "learning_rate": 1.2157457653789337e-05,
+      "loss": 0.7358,
+      "step": 2268
+    },
+    {
+      "epoch": 1.343796268877702,
+      "grad_norm": 2.085597969173187,
+      "learning_rate": 1.2151212085106478e-05,
+      "loss": 0.7227,
+      "step": 2269
+    },
+    {
+      "epoch": 1.344388510512289,
+      "grad_norm": 1.6707958861915149,
+      "learning_rate": 1.2144965636456903e-05,
+      "loss": 0.7298,
+      "step": 2270
+    },
+    {
+      "epoch": 1.344980752146876,
+      "grad_norm": 2.7788857483382206,
+      "learning_rate": 1.213871831039577e-05,
+      "loss": 0.6869,
+      "step": 2271
+    },
+    {
+      "epoch": 1.345572993781463,
+      "grad_norm": 2.382424211273845,
+      "learning_rate": 1.2132470109478577e-05,
+      "loss": 0.6929,
+      "step": 2272
+    },
+    {
+      "epoch": 1.3461652354160498,
+      "grad_norm": 3.2307082039534047,
+      "learning_rate": 1.21262210362612e-05,
+      "loss": 0.716,
+      "step": 2273
+    },
+    {
+      "epoch": 1.3467574770506365,
+      "grad_norm": 2.304898917214821,
+      "learning_rate": 1.2119971093299852e-05,
+      "loss": 0.7246,
+      "step": 2274
+    },
+    {
+      "epoch": 1.3473497186852237,
+      "grad_norm": 4.297876675540321,
+      "learning_rate": 1.2113720283151115e-05,
+      "loss": 0.7579,
+      "step": 2275
+    },
+    {
+      "epoch": 1.3479419603198104,
+      "grad_norm": 2.19405196855775,
+      "learning_rate": 1.2107468608371924e-05,
+      "loss": 0.7604,
+      "step": 2276
+    },
+    {
+      "epoch": 1.3485342019543973,
+      "grad_norm": 2.236962795916548,
+      "learning_rate": 1.2101216071519561e-05,
+      "loss": 0.7458,
+      "step": 2277
+    },
+    {
+      "epoch": 1.3491264435889843,
+      "grad_norm": 3.9633695711684975,
+      "learning_rate": 1.209496267515167e-05,
+      "loss": 0.7194,
+      "step": 2278
+    },
+    {
+      "epoch": 1.3497186852235712,
+      "grad_norm": 2.4068817921340333,
+      "learning_rate": 1.2088708421826238e-05,
+      "loss": 0.7136,
+      "step": 2279
+    },
+    {
+      "epoch": 1.3503109268581581,
+      "grad_norm": 1.5247334800083856,
+      "learning_rate": 1.2082453314101607e-05,
+      "loss": 0.6947,
+      "step": 2280
+    },
+    {
+      "epoch": 1.350903168492745,
+      "grad_norm": 1.673058636797337,
+      "learning_rate": 1.2076197354536472e-05,
+      "loss": 0.7307,
+      "step": 2281
+    },
+    {
+      "epoch": 1.351495410127332,
+      "grad_norm": 2.7089039748271495,
+      "learning_rate": 1.2069940545689867e-05,
+      "loss": 0.7495,
+      "step": 2282
+    },
+    {
+      "epoch": 1.352087651761919,
+      "grad_norm": 1.4724209313434784,
+      "learning_rate": 1.2063682890121178e-05,
+      "loss": 0.7076,
+      "step": 2283
+    },
+    {
+      "epoch": 1.3526798933965059,
+      "grad_norm": 1.4347383550436956,
+      "learning_rate": 1.2057424390390141e-05,
+      "loss": 0.7122,
+      "step": 2284
+    },
+    {
+      "epoch": 1.3532721350310926,
+      "grad_norm": 1.6233887203509207,
+      "learning_rate": 1.2051165049056836e-05,
+      "loss": 0.7273,
+      "step": 2285
+    },
+    {
+      "epoch": 1.3538643766656797,
+      "grad_norm": 2.2293249045490504,
+      "learning_rate": 1.2044904868681684e-05,
+      "loss": 0.7501,
+      "step": 2286
+    },
+    {
+      "epoch": 1.3544566183002664,
+      "grad_norm": 2.3431634966622115,
+      "learning_rate": 1.2038643851825449e-05,
+      "loss": 0.7621,
+      "step": 2287
+    },
+    {
+      "epoch": 1.3550488599348534,
+      "grad_norm": 2.132037815874192,
+      "learning_rate": 1.203238200104924e-05,
+      "loss": 0.7268,
+      "step": 2288
+    },
+    {
+      "epoch": 1.3556411015694403,
+      "grad_norm": 2.230932140756579,
+      "learning_rate": 1.2026119318914507e-05,
+      "loss": 0.735,
+      "step": 2289
+    },
+    {
+      "epoch": 1.3562333432040272,
+      "grad_norm": 2.313542593223687,
+      "learning_rate": 1.2019855807983036e-05,
+      "loss": 0.7422,
+      "step": 2290
+    },
+    {
+      "epoch": 1.3568255848386142,
+      "grad_norm": 1.9178912189104353,
+      "learning_rate": 1.201359147081696e-05,
+      "loss": 0.73,
+      "step": 2291
+    },
+    {
+      "epoch": 1.357417826473201,
+      "grad_norm": 1.5513535945386705,
+      "learning_rate": 1.200732630997874e-05,
+      "loss": 0.7273,
+      "step": 2292
+    },
+    {
+      "epoch": 1.358010068107788,
+      "grad_norm": 2.3144421995406512,
+      "learning_rate": 1.2001060328031185e-05,
+      "loss": 0.7184,
+      "step": 2293
+    },
+    {
+      "epoch": 1.358602309742375,
+      "grad_norm": 1.8962622817640602,
+      "learning_rate": 1.1994793527537427e-05,
+      "loss": 0.7409,
+      "step": 2294
+    },
+    {
+      "epoch": 1.3591945513769619,
+      "grad_norm": 1.9723792893075207,
+      "learning_rate": 1.198852591106095e-05,
+      "loss": 0.7409,
+      "step": 2295
+    },
+    {
+      "epoch": 1.3597867930115486,
+      "grad_norm": 1.6685845910473676,
+      "learning_rate": 1.1982257481165547e-05,
+      "loss": 0.6975,
+      "step": 2296
+    },
+    {
+      "epoch": 1.3603790346461357,
+      "grad_norm": 1.8774532039410945,
+      "learning_rate": 1.1975988240415373e-05,
+      "loss": 0.7433,
+      "step": 2297
+    },
+    {
+      "epoch": 1.3609712762807225,
+      "grad_norm": 4.187478803679397,
+      "learning_rate": 1.1969718191374888e-05,
+      "loss": 0.7295,
+      "step": 2298
+    },
+    {
+      "epoch": 1.3615635179153094,
+      "grad_norm": 2.2037953249975315,
+      "learning_rate": 1.1963447336608906e-05,
+      "loss": 0.7087,
+      "step": 2299
+    },
+    {
+      "epoch": 1.3621557595498963,
+      "grad_norm": 3.372329419689456,
+      "learning_rate": 1.1957175678682548e-05,
+      "loss": 0.737,
+      "step": 2300
+    },
+    {
+      "epoch": 1.3627480011844832,
+      "grad_norm": 1.7609785926431862,
+      "learning_rate": 1.1950903220161286e-05,
+      "loss": 0.7049,
+      "step": 2301
+    },
+    {
+      "epoch": 1.3633402428190702,
+      "grad_norm": 10.575058480914846,
+      "learning_rate": 1.1944629963610897e-05,
+      "loss": 0.732,
+      "step": 2302
+    },
+    {
+      "epoch": 1.3639324844536571,
+      "grad_norm": 1.7154442069759137,
+      "learning_rate": 1.1938355911597503e-05,
+      "loss": 0.7155,
+      "step": 2303
+    },
+    {
+      "epoch": 1.364524726088244,
+      "grad_norm": 2.2620417290126946,
+      "learning_rate": 1.1932081066687544e-05,
+      "loss": 0.7689,
+      "step": 2304
+    },
+    {
+      "epoch": 1.365116967722831,
+      "grad_norm": 5.085468174057115,
+      "learning_rate": 1.1925805431447779e-05,
+      "loss": 0.7161,
+      "step": 2305
+    },
+    {
+      "epoch": 1.365709209357418,
+      "grad_norm": 2.4608886800316285,
+      "learning_rate": 1.1919529008445302e-05,
+      "loss": 0.7487,
+      "step": 2306
+    },
+    {
+      "epoch": 1.3663014509920046,
+      "grad_norm": 2.196016033444546,
+      "learning_rate": 1.1913251800247515e-05,
+      "loss": 0.7473,
+      "step": 2307
+    },
+    {
+      "epoch": 1.3668936926265918,
+      "grad_norm": 2.3351921735739136,
+      "learning_rate": 1.1906973809422163e-05,
+      "loss": 0.7114,
+      "step": 2308
+    },
+    {
+      "epoch": 1.3674859342611785,
+      "grad_norm": 3.712094735632108,
+      "learning_rate": 1.1900695038537283e-05,
+      "loss": 0.7126,
+      "step": 2309
+    },
+    {
+      "epoch": 1.3680781758957654,
+      "grad_norm": 3.0445067803167434,
+      "learning_rate": 1.1894415490161253e-05,
+      "loss": 0.7616,
+      "step": 2310
+    },
+    {
+      "epoch": 1.3686704175303523,
+      "grad_norm": 5.092443503760028,
+      "learning_rate": 1.1888135166862756e-05,
+      "loss": 0.7598,
+      "step": 2311
+    },
+    {
+      "epoch": 1.3692626591649393,
+      "grad_norm": 1.7515743364833825,
+      "learning_rate": 1.1881854071210805e-05,
+      "loss": 0.7519,
+      "step": 2312
+    },
+    {
+      "epoch": 1.3698549007995262,
+      "grad_norm": 1.4595659443345792,
+      "learning_rate": 1.1875572205774712e-05,
+      "loss": 0.6944,
+      "step": 2313
+    },
+    {
+      "epoch": 1.3704471424341131,
+      "grad_norm": 2.0276065761611566,
+      "learning_rate": 1.1869289573124125e-05,
+      "loss": 0.7117,
+      "step": 2314
+    },
+    {
+      "epoch": 1.3710393840687,
+      "grad_norm": 1.6385422571466461,
+      "learning_rate": 1.1863006175828984e-05,
+      "loss": 0.7348,
+      "step": 2315
+    },
+    {
+      "epoch": 1.371631625703287,
+      "grad_norm": 1.2986389722977396,
+      "learning_rate": 1.1856722016459554e-05,
+      "loss": 0.7225,
+      "step": 2316
+    },
+    {
+      "epoch": 1.372223867337874,
+      "grad_norm": 2.5907128138808284,
+      "learning_rate": 1.1850437097586412e-05,
+      "loss": 0.7053,
+      "step": 2317
+    },
+    {
+      "epoch": 1.3728161089724606,
+      "grad_norm": 1.72659341320089,
+      "learning_rate": 1.1844151421780442e-05,
+      "loss": 0.7747,
+      "step": 2318
+    },
+    {
+      "epoch": 1.3734083506070478,
+      "grad_norm": 2.3936636379326326,
+      "learning_rate": 1.1837864991612839e-05,
+      "loss": 0.746,
+      "step": 2319
+    },
+    {
+      "epoch": 1.3740005922416345,
+      "grad_norm": 3.268948302001786,
+      "learning_rate": 1.1831577809655105e-05,
+      "loss": 0.7287,
+      "step": 2320
+    },
+    {
+      "epoch": 1.3745928338762214,
+      "grad_norm": 1.8358669402385104,
+      "learning_rate": 1.1825289878479054e-05,
+      "loss": 0.7323,
+      "step": 2321
+    },
+    {
+      "epoch": 1.3751850755108084,
+      "grad_norm": 6.659136666043565,
+      "learning_rate": 1.18190012006568e-05,
+      "loss": 0.781,
+      "step": 2322
+    },
+    {
+      "epoch": 1.3757773171453953,
+      "grad_norm": 1.8103702241279684,
+      "learning_rate": 1.1812711778760768e-05,
+      "loss": 0.7562,
+      "step": 2323
+    },
+    {
+      "epoch": 1.3763695587799822,
+      "grad_norm": 3.4205148227314486,
+      "learning_rate": 1.1806421615363685e-05,
+      "loss": 0.7057,
+      "step": 2324
+    },
+    {
+      "epoch": 1.3769618004145692,
+      "grad_norm": 5.3299975994261874,
+      "learning_rate": 1.1800130713038582e-05,
+      "loss": 0.7554,
+      "step": 2325
+    },
+    {
+      "epoch": 1.377554042049156,
+      "grad_norm": 1.5067357582677707,
+      "learning_rate": 1.179383907435879e-05,
+      "loss": 0.728,
+      "step": 2326
+    },
+    {
+      "epoch": 1.378146283683743,
+      "grad_norm": 1.5937327086086364,
+      "learning_rate": 1.1787546701897947e-05,
+      "loss": 0.6923,
+      "step": 2327
+    },
+    {
+      "epoch": 1.37873852531833,
+      "grad_norm": 2.9084759352815137,
+      "learning_rate": 1.1781253598229982e-05,
+      "loss": 0.6867,
+      "step": 2328
+    },
+    {
+      "epoch": 1.3793307669529167,
+      "grad_norm": 1.8065003892691773,
+      "learning_rate": 1.177495976592913e-05,
+      "loss": 0.7264,
+      "step": 2329
+    },
+    {
+      "epoch": 1.3799230085875038,
+      "grad_norm": 1.374778675210592,
+      "learning_rate": 1.1768665207569922e-05,
+      "loss": 0.763,
+      "step": 2330
+    },
+    {
+      "epoch": 1.3805152502220905,
+      "grad_norm": 5.135815860167982,
+      "learning_rate": 1.176236992572719e-05,
+      "loss": 0.7287,
+      "step": 2331
+    },
+    {
+      "epoch": 1.3811074918566775,
+      "grad_norm": 1.9451779116322265,
+      "learning_rate": 1.1756073922976056e-05,
+      "loss": 0.7603,
+      "step": 2332
+    },
+    {
+      "epoch": 1.3816997334912644,
+      "grad_norm": 4.102609143725733,
+      "learning_rate": 1.1749777201891937e-05,
+      "loss": 0.6891,
+      "step": 2333
+    },
+    {
+      "epoch": 1.3822919751258513,
+      "grad_norm": 2.703163169605834,
+      "learning_rate": 1.1743479765050549e-05,
+      "loss": 0.7238,
+      "step": 2334
+    },
+    {
+      "epoch": 1.3828842167604383,
+      "grad_norm": 1.4161656206435274,
+      "learning_rate": 1.1737181615027894e-05,
+      "loss": 0.7072,
+      "step": 2335
+    },
+    {
+      "epoch": 1.3834764583950252,
+      "grad_norm": 2.111623863485621,
+      "learning_rate": 1.1730882754400274e-05,
+      "loss": 0.6903,
+      "step": 2336
+    },
+    {
+      "epoch": 1.3840687000296121,
+      "grad_norm": 1.9893098347865092,
+      "learning_rate": 1.172458318574427e-05,
+      "loss": 0.7353,
+      "step": 2337
+    },
+    {
+      "epoch": 1.384660941664199,
+      "grad_norm": 3.119953828491785,
+      "learning_rate": 1.1718282911636774e-05,
+      "loss": 0.7126,
+      "step": 2338
+    },
+    {
+      "epoch": 1.385253183298786,
+      "grad_norm": 2.0807212215419937,
+      "learning_rate": 1.1711981934654937e-05,
+      "loss": 0.7486,
+      "step": 2339
+    },
+    {
+      "epoch": 1.3858454249333727,
+      "grad_norm": 2.1409029627127025,
+      "learning_rate": 1.1705680257376224e-05,
+      "loss": 0.736,
+      "step": 2340
+    },
+    {
+      "epoch": 1.3864376665679599,
+      "grad_norm": 1.7678639437104227,
+      "learning_rate": 1.1699377882378367e-05,
+      "loss": 0.6811,
+      "step": 2341
+    },
+    {
+      "epoch": 1.3870299082025466,
+      "grad_norm": 1.1969173608840782,
+      "learning_rate": 1.1693074812239397e-05,
+      "loss": 0.7072,
+      "step": 2342
+    },
+    {
+      "epoch": 1.3876221498371335,
+      "grad_norm": 1.7602686350385015,
+      "learning_rate": 1.1686771049537621e-05,
+      "loss": 0.6984,
+      "step": 2343
+    },
+    {
+      "epoch": 1.3882143914717204,
+      "grad_norm": 1.4360466219925514,
+      "learning_rate": 1.1680466596851635e-05,
+      "loss": 0.7065,
+      "step": 2344
+    },
+    {
+      "epoch": 1.3888066331063074,
+      "grad_norm": 4.263733785806742,
+      "learning_rate": 1.1674161456760314e-05,
+      "loss": 0.7292,
+      "step": 2345
+    },
+    {
+      "epoch": 1.3893988747408943,
+      "grad_norm": 2.471335440136607,
+      "learning_rate": 1.1667855631842815e-05,
+      "loss": 0.7671,
+      "step": 2346
+    },
+    {
+      "epoch": 1.3899911163754812,
+      "grad_norm": 4.172593773758679,
+      "learning_rate": 1.1661549124678573e-05,
+      "loss": 0.751,
+      "step": 2347
+    },
+    {
+      "epoch": 1.3905833580100682,
+      "grad_norm": 2.5493693541099667,
+      "learning_rate": 1.1655241937847305e-05,
+      "loss": 0.7763,
+      "step": 2348
+    },
+    {
+      "epoch": 1.391175599644655,
+      "grad_norm": 12.734269954324368,
+      "learning_rate": 1.1648934073929008e-05,
+      "loss": 0.7433,
+      "step": 2349
+    },
+    {
+      "epoch": 1.391767841279242,
+      "grad_norm": 2.0172453728072015,
+      "learning_rate": 1.164262553550395e-05,
+      "loss": 0.7186,
+      "step": 2350
+    },
+    {
+      "epoch": 1.3923600829138287,
+      "grad_norm": 1.6347819754535515,
+      "learning_rate": 1.1636316325152678e-05,
+      "loss": 0.7313,
+      "step": 2351
+    },
+    {
+      "epoch": 1.3929523245484159,
+      "grad_norm": 1.6197215465398926,
+      "learning_rate": 1.1630006445456015e-05,
+      "loss": 0.7162,
+      "step": 2352
+    },
+    {
+      "epoch": 1.3935445661830026,
+      "grad_norm": 1.3419132909074647,
+      "learning_rate": 1.1623695898995057e-05,
+      "loss": 0.6812,
+      "step": 2353
+    },
+    {
+      "epoch": 1.3941368078175895,
+      "grad_norm": 1.999373727061114,
+      "learning_rate": 1.161738468835117e-05,
+      "loss": 0.6904,
+      "step": 2354
+    },
+    {
+      "epoch": 1.3947290494521765,
+      "grad_norm": 2.0145229640481106,
+      "learning_rate": 1.1611072816105995e-05,
+      "loss": 0.7258,
+      "step": 2355
+    },
+    {
+      "epoch": 1.3953212910867634,
+      "grad_norm": 1.6691310278707019,
+      "learning_rate": 1.1604760284841446e-05,
+      "loss": 0.7245,
+      "step": 2356
+    },
+    {
+      "epoch": 1.3959135327213503,
+      "grad_norm": 4.128799120066488,
+      "learning_rate": 1.1598447097139698e-05,
+      "loss": 0.7285,
+      "step": 2357
+    },
+    {
+      "epoch": 1.3965057743559373,
+      "grad_norm": 2.0773835779838663,
+      "learning_rate": 1.1592133255583204e-05,
+      "loss": 0.7516,
+      "step": 2358
+    },
+    {
+      "epoch": 1.3970980159905242,
+      "grad_norm": 1.3942800893003757,
+      "learning_rate": 1.1585818762754678e-05,
+      "loss": 0.731,
+      "step": 2359
+    },
+    {
+      "epoch": 1.3976902576251111,
+      "grad_norm": 1.8775327674483524,
+      "learning_rate": 1.1579503621237102e-05,
+      "loss": 0.7454,
+      "step": 2360
+    },
+    {
+      "epoch": 1.398282499259698,
+      "grad_norm": 3.027896060244714,
+      "learning_rate": 1.1573187833613723e-05,
+      "loss": 0.7275,
+      "step": 2361
+    },
+    {
+      "epoch": 1.3988747408942848,
+      "grad_norm": 1.4716030596428373,
+      "learning_rate": 1.156687140246806e-05,
+      "loss": 0.7367,
+      "step": 2362
+    },
+    {
+      "epoch": 1.399466982528872,
+      "grad_norm": 1.7820811493950572,
+      "learning_rate": 1.1560554330383881e-05,
+      "loss": 0.7146,
+      "step": 2363
+    },
+    {
+      "epoch": 1.4000592241634586,
+      "grad_norm": 2.7037191130979323,
+      "learning_rate": 1.1554236619945229e-05,
+      "loss": 0.7242,
+      "step": 2364
+    },
+    {
+      "epoch": 1.4006514657980456,
+      "grad_norm": 1.7688945832299963,
+      "learning_rate": 1.1547918273736397e-05,
+      "loss": 0.7575,
+      "step": 2365
+    },
+    {
+      "epoch": 1.4012437074326325,
+      "grad_norm": 1.7948443959599745,
+      "learning_rate": 1.1541599294341952e-05,
+      "loss": 0.7385,
+      "step": 2366
+    },
+    {
+      "epoch": 1.4018359490672194,
+      "grad_norm": 2.5494232660163414,
+      "learning_rate": 1.1535279684346702e-05,
+      "loss": 0.7321,
+      "step": 2367
+    },
+    {
+      "epoch": 1.4024281907018064,
+      "grad_norm": 1.131792892745796,
+      "learning_rate": 1.1528959446335735e-05,
+      "loss": 0.7251,
+      "step": 2368
+    },
+    {
+      "epoch": 1.4030204323363933,
+      "grad_norm": 1.7242196863607848,
+      "learning_rate": 1.1522638582894372e-05,
+      "loss": 0.7357,
+      "step": 2369
+    },
+    {
+      "epoch": 1.4036126739709802,
+      "grad_norm": 1.7494101461566682,
+      "learning_rate": 1.1516317096608207e-05,
+      "loss": 0.7277,
+      "step": 2370
+    },
+    {
+      "epoch": 1.4042049156055671,
+      "grad_norm": 2.45681555550249,
+      "learning_rate": 1.1509994990063085e-05,
+      "loss": 0.6967,
+      "step": 2371
+    },
+    {
+      "epoch": 1.404797157240154,
+      "grad_norm": 2.9094052043800223,
+      "learning_rate": 1.1503672265845098e-05,
+      "loss": 0.7258,
+      "step": 2372
+    },
+    {
+      "epoch": 1.4053893988747408,
+      "grad_norm": 2.4462949935085616,
+      "learning_rate": 1.1497348926540602e-05,
+      "loss": 0.7472,
+      "step": 2373
+    },
+    {
+      "epoch": 1.4059816405093277,
+      "grad_norm": 1.5009612756951904,
+      "learning_rate": 1.1491024974736191e-05,
+      "loss": 0.6824,
+      "step": 2374
+    },
+    {
+      "epoch": 1.4065738821439147,
+      "grad_norm": 3.095173697672372,
+      "learning_rate": 1.1484700413018724e-05,
+      "loss": 0.7173,
+      "step": 2375
+    },
+    {
+      "epoch": 1.4071661237785016,
+      "grad_norm": 2.4512057822789606,
+      "learning_rate": 1.1478375243975298e-05,
+      "loss": 0.7297,
+      "step": 2376
+    },
+    {
+      "epoch": 1.4077583654130885,
+      "grad_norm": 1.9421393264777882,
+      "learning_rate": 1.1472049470193263e-05,
+      "loss": 0.7504,
+      "step": 2377
+    },
+    {
+      "epoch": 1.4083506070476755,
+      "grad_norm": 2.4342391043313047,
+      "learning_rate": 1.1465723094260219e-05,
+      "loss": 0.7148,
+      "step": 2378
+    },
+    {
+      "epoch": 1.4089428486822624,
+      "grad_norm": 1.2678552133839875,
+      "learning_rate": 1.1459396118764007e-05,
+      "loss": 0.6937,
+      "step": 2379
+    },
+    {
+      "epoch": 1.4095350903168493,
+      "grad_norm": 1.6681069042211958,
+      "learning_rate": 1.1453068546292718e-05,
+      "loss": 0.7413,
+      "step": 2380
+    },
+    {
+      "epoch": 1.4101273319514362,
+      "grad_norm": 2.0274279135928706,
+      "learning_rate": 1.1446740379434681e-05,
+      "loss": 0.7516,
+      "step": 2381
+    },
+    {
+      "epoch": 1.4107195735860232,
+      "grad_norm": 1.4001394906264233,
+      "learning_rate": 1.1440411620778478e-05,
+      "loss": 0.7393,
+      "step": 2382
+    },
+    {
+      "epoch": 1.41131181522061,
+      "grad_norm": 3.0338701312366445,
+      "learning_rate": 1.1434082272912923e-05,
+      "loss": 0.7396,
+      "step": 2383
+    },
+    {
+      "epoch": 1.4119040568551968,
+      "grad_norm": 1.786481375473841,
+      "learning_rate": 1.1427752338427075e-05,
+      "loss": 0.7079,
+      "step": 2384
+    },
+    {
+      "epoch": 1.4124962984897838,
+      "grad_norm": 1.1712156976339714,
+      "learning_rate": 1.1421421819910235e-05,
+      "loss": 0.7258,
+      "step": 2385
+    },
+    {
+      "epoch": 1.4130885401243707,
+      "grad_norm": 2.996800157879306,
+      "learning_rate": 1.141509071995194e-05,
+      "loss": 0.7282,
+      "step": 2386
+    },
+    {
+      "epoch": 1.4136807817589576,
+      "grad_norm": 1.3892970824112951,
+      "learning_rate": 1.1408759041141963e-05,
+      "loss": 0.7463,
+      "step": 2387
+    },
+    {
+      "epoch": 1.4142730233935445,
+      "grad_norm": 1.6106966314025877,
+      "learning_rate": 1.1402426786070326e-05,
+      "loss": 0.7159,
+      "step": 2388
+    },
+    {
+      "epoch": 1.4148652650281315,
+      "grad_norm": 2.391973281082433,
+      "learning_rate": 1.1396093957327266e-05,
+      "loss": 0.6767,
+      "step": 2389
+    },
+    {
+      "epoch": 1.4154575066627184,
+      "grad_norm": 1.8064771701538858,
+      "learning_rate": 1.1389760557503275e-05,
+      "loss": 0.7618,
+      "step": 2390
+    },
+    {
+      "epoch": 1.4160497482973053,
+      "grad_norm": 1.9436742635325912,
+      "learning_rate": 1.1383426589189062e-05,
+      "loss": 0.7507,
+      "step": 2391
+    },
+    {
+      "epoch": 1.4166419899318923,
+      "grad_norm": 1.4722158504330858,
+      "learning_rate": 1.1377092054975586e-05,
+      "loss": 0.7029,
+      "step": 2392
+    },
+    {
+      "epoch": 1.417234231566479,
+      "grad_norm": 1.2331100621292894,
+      "learning_rate": 1.1370756957454015e-05,
+      "loss": 0.6877,
+      "step": 2393
+    },
+    {
+      "epoch": 1.4178264732010661,
+      "grad_norm": 1.8467647418967628,
+      "learning_rate": 1.1364421299215773e-05,
+      "loss": 0.734,
+      "step": 2394
+    },
+    {
+      "epoch": 1.4184187148356528,
+      "grad_norm": 2.0082819089866293,
+      "learning_rate": 1.135808508285249e-05,
+      "loss": 0.719,
+      "step": 2395
+    },
+    {
+      "epoch": 1.4190109564702398,
+      "grad_norm": 1.461710287319451,
+      "learning_rate": 1.135174831095604e-05,
+      "loss": 0.7138,
+      "step": 2396
+    },
+    {
+      "epoch": 1.4196031981048267,
+      "grad_norm": 1.519799568496123,
+      "learning_rate": 1.134541098611852e-05,
+      "loss": 0.7084,
+      "step": 2397
+    },
+    {
+      "epoch": 1.4201954397394136,
+      "grad_norm": 2.315400648418732,
+      "learning_rate": 1.133907311093225e-05,
+      "loss": 0.7295,
+      "step": 2398
+    },
+    {
+      "epoch": 1.4207876813740006,
+      "grad_norm": 1.8933519527297487,
+      "learning_rate": 1.133273468798978e-05,
+      "loss": 0.7078,
+      "step": 2399
+    },
+    {
+      "epoch": 1.4213799230085875,
+      "grad_norm": 1.91221104798189,
+      "learning_rate": 1.1326395719883876e-05,
+      "loss": 0.6819,
+      "step": 2400
+    },
+    {
+      "epoch": 1.4219721646431744,
+      "grad_norm": 1.979878107597815,
+      "learning_rate": 1.1320056209207538e-05,
+      "loss": 0.71,
+      "step": 2401
+    },
+    {
+      "epoch": 1.4225644062777614,
+      "grad_norm": 2.977882494720392,
+      "learning_rate": 1.1313716158553978e-05,
+      "loss": 0.7408,
+      "step": 2402
+    },
+    {
+      "epoch": 1.4231566479123483,
+      "grad_norm": 1.6024476014527083,
+      "learning_rate": 1.1307375570516637e-05,
+      "loss": 0.7167,
+      "step": 2403
+    },
+    {
+      "epoch": 1.423748889546935,
+      "grad_norm": 1.714605953833717,
+      "learning_rate": 1.130103444768917e-05,
+      "loss": 0.7197,
+      "step": 2404
+    },
+    {
+      "epoch": 1.4243411311815222,
+      "grad_norm": 2.5811940978711068,
+      "learning_rate": 1.1294692792665452e-05,
+      "loss": 0.6986,
+      "step": 2405
+    },
+    {
+      "epoch": 1.4249333728161089,
+      "grad_norm": 5.955582148467331,
+      "learning_rate": 1.1288350608039577e-05,
+      "loss": 0.7285,
+      "step": 2406
+    },
+    {
+      "epoch": 1.4255256144506958,
+      "grad_norm": 1.580417338453988,
+      "learning_rate": 1.1282007896405858e-05,
+      "loss": 0.7571,
+      "step": 2407
+    },
+    {
+      "epoch": 1.4261178560852827,
+      "grad_norm": 1.7271693274308317,
+      "learning_rate": 1.1275664660358818e-05,
+      "loss": 0.7756,
+      "step": 2408
+    },
+    {
+      "epoch": 1.4267100977198697,
+      "grad_norm": 1.5963788064488735,
+      "learning_rate": 1.1269320902493199e-05,
+      "loss": 0.7245,
+      "step": 2409
+    },
+    {
+      "epoch": 1.4273023393544566,
+      "grad_norm": 2.2232952704603828,
+      "learning_rate": 1.1262976625403954e-05,
+      "loss": 0.7156,
+      "step": 2410
+    },
+    {
+      "epoch": 1.4278945809890435,
+      "grad_norm": 4.047457885454327,
+      "learning_rate": 1.1256631831686245e-05,
+      "loss": 0.7397,
+      "step": 2411
+    },
+    {
+      "epoch": 1.4284868226236305,
+      "grad_norm": 2.079661525227277,
+      "learning_rate": 1.1250286523935456e-05,
+      "loss": 0.6906,
+      "step": 2412
+    },
+    {
+      "epoch": 1.4290790642582174,
+      "grad_norm": 2.3170134187542666,
+      "learning_rate": 1.1243940704747169e-05,
+      "loss": 0.7696,
+      "step": 2413
+    },
+    {
+      "epoch": 1.4296713058928043,
+      "grad_norm": 4.447126564702576,
+      "learning_rate": 1.1237594376717188e-05,
+      "loss": 0.7215,
+      "step": 2414
+    },
+    {
+      "epoch": 1.430263547527391,
+      "grad_norm": 1.487487996677641,
+      "learning_rate": 1.1231247542441507e-05,
+      "loss": 0.7387,
+      "step": 2415
+    },
+    {
+      "epoch": 1.4308557891619782,
+      "grad_norm": 2.7450360360522925,
+      "learning_rate": 1.122490020451635e-05,
+      "loss": 0.7184,
+      "step": 2416
+    },
+    {
+      "epoch": 1.431448030796565,
+      "grad_norm": 4.220383960594748,
+      "learning_rate": 1.1218552365538129e-05,
+      "loss": 0.7528,
+      "step": 2417
+    },
+    {
+      "epoch": 1.4320402724311518,
+      "grad_norm": 8.716959325200545,
+      "learning_rate": 1.1212204028103469e-05,
+      "loss": 0.7771,
+      "step": 2418
+    },
+    {
+      "epoch": 1.4326325140657388,
+      "grad_norm": 1.6083934310732948,
+      "learning_rate": 1.1205855194809191e-05,
+      "loss": 0.7349,
+      "step": 2419
+    },
+    {
+      "epoch": 1.4332247557003257,
+      "grad_norm": 2.098687005967919,
+      "learning_rate": 1.1199505868252336e-05,
+      "loss": 0.7355,
+      "step": 2420
+    },
+    {
+      "epoch": 1.4338169973349126,
+      "grad_norm": 1.6917431368313898,
+      "learning_rate": 1.1193156051030128e-05,
+      "loss": 0.6713,
+      "step": 2421
+    },
+    {
+      "epoch": 1.4344092389694996,
+      "grad_norm": 1.9121827759052907,
+      "learning_rate": 1.1186805745739999e-05,
+      "loss": 0.7065,
+      "step": 2422
+    },
+    {
+      "epoch": 1.4350014806040865,
+      "grad_norm": 1.7303129056373097,
+      "learning_rate": 1.1180454954979583e-05,
+      "loss": 0.6621,
+      "step": 2423
+    },
+    {
+      "epoch": 1.4355937222386734,
+      "grad_norm": 1.4653359584564212,
+      "learning_rate": 1.1174103681346711e-05,
+      "loss": 0.7288,
+      "step": 2424
+    },
+    {
+      "epoch": 1.4361859638732604,
+      "grad_norm": 1.6242879043550942,
+      "learning_rate": 1.1167751927439407e-05,
+      "loss": 0.7364,
+      "step": 2425
+    },
+    {
+      "epoch": 1.436778205507847,
+      "grad_norm": 1.879062950337473,
+      "learning_rate": 1.1161399695855903e-05,
+      "loss": 0.7296,
+      "step": 2426
+    },
+    {
+      "epoch": 1.4373704471424342,
+      "grad_norm": 2.2495686189430617,
+      "learning_rate": 1.1155046989194613e-05,
+      "loss": 0.775,
+      "step": 2427
+    },
+    {
+      "epoch": 1.437962688777021,
+      "grad_norm": 1.667428421117216,
+      "learning_rate": 1.1148693810054152e-05,
+      "loss": 0.7023,
+      "step": 2428
+    },
+    {
+      "epoch": 1.4385549304116079,
+      "grad_norm": 1.3747252199835087,
+      "learning_rate": 1.1142340161033331e-05,
+      "loss": 0.7057,
+      "step": 2429
+    },
+    {
+      "epoch": 1.4391471720461948,
+      "grad_norm": 2.463314444705448,
+      "learning_rate": 1.1135986044731144e-05,
+      "loss": 0.6567,
+      "step": 2430
+    },
+    {
+      "epoch": 1.4397394136807817,
+      "grad_norm": 1.5810081020557705,
+      "learning_rate": 1.1129631463746789e-05,
+      "loss": 0.6708,
+      "step": 2431
+    },
+    {
+      "epoch": 1.4403316553153687,
+      "grad_norm": 1.8170531880544196,
+      "learning_rate": 1.112327642067964e-05,
+      "loss": 0.7159,
+      "step": 2432
+    },
+    {
+      "epoch": 1.4409238969499556,
+      "grad_norm": 5.805209899239579,
+      "learning_rate": 1.1116920918129271e-05,
+      "loss": 0.7091,
+      "step": 2433
+    },
+    {
+      "epoch": 1.4415161385845425,
+      "grad_norm": 1.9375366699904837,
+      "learning_rate": 1.111056495869544e-05,
+      "loss": 0.7578,
+      "step": 2434
+    },
+    {
+      "epoch": 1.4421083802191295,
+      "grad_norm": 2.8102088648986525,
+      "learning_rate": 1.110420854497809e-05,
+      "loss": 0.7223,
+      "step": 2435
+    },
+    {
+      "epoch": 1.4427006218537164,
+      "grad_norm": 2.1131020428074456,
+      "learning_rate": 1.1097851679577351e-05,
+      "loss": 0.7278,
+      "step": 2436
+    },
+    {
+      "epoch": 1.443292863488303,
+      "grad_norm": 2.3630646661630004,
+      "learning_rate": 1.1091494365093542e-05,
+      "loss": 0.751,
+      "step": 2437
+    },
+    {
+      "epoch": 1.4438851051228903,
+      "grad_norm": 3.8773111167009624,
+      "learning_rate": 1.1085136604127161e-05,
+      "loss": 0.7495,
+      "step": 2438
+    },
+    {
+      "epoch": 1.444477346757477,
+      "grad_norm": 2.7572277967036456,
+      "learning_rate": 1.1078778399278885e-05,
+      "loss": 0.749,
+      "step": 2439
+    },
+    {
+      "epoch": 1.445069588392064,
+      "grad_norm": 1.469912383127116,
+      "learning_rate": 1.1072419753149585e-05,
+      "loss": 0.7198,
+      "step": 2440
+    },
+    {
+      "epoch": 1.4456618300266508,
+      "grad_norm": 2.777349956968284,
+      "learning_rate": 1.1066060668340298e-05,
+      "loss": 0.6924,
+      "step": 2441
+    },
+    {
+      "epoch": 1.4462540716612378,
+      "grad_norm": 1.4767013976366694,
+      "learning_rate": 1.105970114745225e-05,
+      "loss": 0.7943,
+      "step": 2442
+    },
+    {
+      "epoch": 1.4468463132958247,
+      "grad_norm": 3.0907295250096785,
+      "learning_rate": 1.1053341193086844e-05,
+      "loss": 0.6945,
+      "step": 2443
+    },
+    {
+      "epoch": 1.4474385549304116,
+      "grad_norm": 1.473657931484691,
+      "learning_rate": 1.104698080784566e-05,
+      "loss": 0.7246,
+      "step": 2444
+    },
+    {
+      "epoch": 1.4480307965649986,
+      "grad_norm": 1.5243657541187539,
+      "learning_rate": 1.1040619994330446e-05,
+      "loss": 0.7208,
+      "step": 2445
+    },
+    {
+      "epoch": 1.4486230381995855,
+      "grad_norm": 1.4396515407272992,
+      "learning_rate": 1.1034258755143141e-05,
+      "loss": 0.7082,
+      "step": 2446
+    },
+    {
+      "epoch": 1.4492152798341724,
+      "grad_norm": 4.67485917339473,
+      "learning_rate": 1.1027897092885846e-05,
+      "loss": 0.7418,
+      "step": 2447
+    },
+    {
+      "epoch": 1.4498075214687591,
+      "grad_norm": 1.8934539767167,
+      "learning_rate": 1.1021535010160838e-05,
+      "loss": 0.7048,
+      "step": 2448
+    },
+    {
+      "epoch": 1.4503997631033463,
+      "grad_norm": 1.7934976398624312,
+      "learning_rate": 1.1015172509570567e-05,
+      "loss": 0.7438,
+      "step": 2449
+    },
+    {
+      "epoch": 1.450992004737933,
+      "grad_norm": 3.1616972894112574,
+      "learning_rate": 1.1008809593717653e-05,
+      "loss": 0.7354,
+      "step": 2450
+    },
+    {
+      "epoch": 1.45158424637252,
+      "grad_norm": 2.5005464211548216,
+      "learning_rate": 1.1002446265204887e-05,
+      "loss": 0.7263,
+      "step": 2451
+    },
+    {
+      "epoch": 1.4521764880071069,
+      "grad_norm": 3.4868223181586875,
+      "learning_rate": 1.0996082526635227e-05,
+      "loss": 0.7188,
+      "step": 2452
+    },
+    {
+      "epoch": 1.4527687296416938,
+      "grad_norm": 1.6735246025817165,
+      "learning_rate": 1.0989718380611805e-05,
+      "loss": 0.6973,
+      "step": 2453
+    },
+    {
+      "epoch": 1.4533609712762807,
+      "grad_norm": 2.147911052347242,
+      "learning_rate": 1.0983353829737909e-05,
+      "loss": 0.6972,
+      "step": 2454
+    },
+    {
+      "epoch": 1.4539532129108677,
+      "grad_norm": 1.1261985891425348,
+      "learning_rate": 1.0976988876616998e-05,
+      "loss": 0.7311,
+      "step": 2455
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 1.3716117307279665,
+      "learning_rate": 1.0970623523852699e-05,
+      "loss": 0.7525,
+      "step": 2456
+    },
+    {
+      "epoch": 1.4551376961800415,
+      "grad_norm": 2.1195272638552165,
+      "learning_rate": 1.0964257774048799e-05,
+      "loss": 0.7471,
+      "step": 2457
+    },
+    {
+      "epoch": 1.4557299378146284,
+      "grad_norm": 1.9959343877258,
+      "learning_rate": 1.0957891629809248e-05,
+      "loss": 0.761,
+      "step": 2458
+    },
+    {
+      "epoch": 1.4563221794492152,
+      "grad_norm": 2.2291293933510095,
+      "learning_rate": 1.0951525093738151e-05,
+      "loss": 0.7153,
+      "step": 2459
+    },
+    {
+      "epoch": 1.4569144210838023,
+      "grad_norm": 1.5622339941082886,
+      "learning_rate": 1.0945158168439786e-05,
+      "loss": 0.7644,
+      "step": 2460
+    },
+    {
+      "epoch": 1.457506662718389,
+      "grad_norm": 2.3239762154569554,
+      "learning_rate": 1.0938790856518582e-05,
+      "loss": 0.7644,
+      "step": 2461
+    },
+    {
+      "epoch": 1.458098904352976,
+      "grad_norm": 2.650172654861937,
+      "learning_rate": 1.0932423160579126e-05,
+      "loss": 0.7249,
+      "step": 2462
+    },
+    {
+      "epoch": 1.4586911459875629,
+      "grad_norm": 1.717625601814134,
+      "learning_rate": 1.0926055083226166e-05,
+      "loss": 0.7325,
+      "step": 2463
+    },
+    {
+      "epoch": 1.4592833876221498,
+      "grad_norm": 3.9491611186011384,
+      "learning_rate": 1.0919686627064603e-05,
+      "loss": 0.7214,
+      "step": 2464
+    },
+    {
+      "epoch": 1.4598756292567368,
+      "grad_norm": 1.2001329473694484,
+      "learning_rate": 1.091331779469949e-05,
+      "loss": 0.7511,
+      "step": 2465
+    },
+    {
+      "epoch": 1.4604678708913237,
+      "grad_norm": 1.742997197381938,
+      "learning_rate": 1.0906948588736044e-05,
+      "loss": 0.7016,
+      "step": 2466
+    },
+    {
+      "epoch": 1.4610601125259106,
+      "grad_norm": 1.2448970057704796,
+      "learning_rate": 1.0900579011779622e-05,
+      "loss": 0.7431,
+      "step": 2467
+    },
+    {
+      "epoch": 1.4616523541604975,
+      "grad_norm": 1.3860151218112158,
+      "learning_rate": 1.0894209066435746e-05,
+      "loss": 0.715,
+      "step": 2468
+    },
+    {
+      "epoch": 1.4622445957950845,
+      "grad_norm": 16.17995815293019,
+      "learning_rate": 1.0887838755310072e-05,
+      "loss": 0.7018,
+      "step": 2469
+    },
+    {
+      "epoch": 1.4628368374296712,
+      "grad_norm": 1.7009020003527942,
+      "learning_rate": 1.0881468081008428e-05,
+      "loss": 0.7178,
+      "step": 2470
+    },
+    {
+      "epoch": 1.4634290790642583,
+      "grad_norm": 3.5437807058153696,
+      "learning_rate": 1.0875097046136764e-05,
+      "loss": 0.7219,
+      "step": 2471
+    },
+    {
+      "epoch": 1.464021320698845,
+      "grad_norm": 1.3944824507804543,
+      "learning_rate": 1.0868725653301206e-05,
+      "loss": 0.7535,
+      "step": 2472
+    },
+    {
+      "epoch": 1.464613562333432,
+      "grad_norm": 1.3126484653094563,
+      "learning_rate": 1.0862353905108002e-05,
+      "loss": 0.7229,
+      "step": 2473
+    },
+    {
+      "epoch": 1.465205803968019,
+      "grad_norm": 2.458706343995586,
+      "learning_rate": 1.085598180416356e-05,
+      "loss": 0.7186,
+      "step": 2474
+    },
+    {
+      "epoch": 1.4657980456026058,
+      "grad_norm": 4.852881691527256,
+      "learning_rate": 1.0849609353074423e-05,
+      "loss": 0.7252,
+      "step": 2475
+    },
+    {
+      "epoch": 1.4663902872371928,
+      "grad_norm": 2.036540439548759,
+      "learning_rate": 1.0843236554447288e-05,
+      "loss": 0.7141,
+      "step": 2476
+    },
+    {
+      "epoch": 1.4669825288717797,
+      "grad_norm": 2.066706696048189,
+      "learning_rate": 1.0836863410888983e-05,
+      "loss": 0.7087,
+      "step": 2477
+    },
+    {
+      "epoch": 1.4675747705063666,
+      "grad_norm": 1.7357600707145933,
+      "learning_rate": 1.0830489925006485e-05,
+      "loss": 0.7375,
+      "step": 2478
+    },
+    {
+      "epoch": 1.4681670121409536,
+      "grad_norm": 2.55498272013775,
+      "learning_rate": 1.0824116099406905e-05,
+      "loss": 0.7577,
+      "step": 2479
+    },
+    {
+      "epoch": 1.4687592537755405,
+      "grad_norm": 1.3261417160385482,
+      "learning_rate": 1.0817741936697499e-05,
+      "loss": 0.7181,
+      "step": 2480
+    },
+    {
+      "epoch": 1.4693514954101272,
+      "grad_norm": 2.4364379575015587,
+      "learning_rate": 1.0811367439485658e-05,
+      "loss": 0.7276,
+      "step": 2481
+    },
+    {
+      "epoch": 1.4699437370447144,
+      "grad_norm": 4.800149895264068,
+      "learning_rate": 1.0804992610378907e-05,
+      "loss": 0.7436,
+      "step": 2482
+    },
+    {
+      "epoch": 1.470535978679301,
+      "grad_norm": 5.046483999796505,
+      "learning_rate": 1.0798617451984912e-05,
+      "loss": 0.7433,
+      "step": 2483
+    },
+    {
+      "epoch": 1.471128220313888,
+      "grad_norm": 1.9692087823077262,
+      "learning_rate": 1.0792241966911472e-05,
+      "loss": 0.6953,
+      "step": 2484
+    },
+    {
+      "epoch": 1.471720461948475,
+      "grad_norm": 2.128256332160169,
+      "learning_rate": 1.0785866157766515e-05,
+      "loss": 0.7532,
+      "step": 2485
+    },
+    {
+      "epoch": 1.4723127035830619,
+      "grad_norm": 1.6170247391820483,
+      "learning_rate": 1.077949002715811e-05,
+      "loss": 0.747,
+      "step": 2486
+    },
+    {
+      "epoch": 1.4729049452176488,
+      "grad_norm": 2.938634854754934,
+      "learning_rate": 1.0773113577694452e-05,
+      "loss": 0.7153,
+      "step": 2487
+    },
+    {
+      "epoch": 1.4734971868522357,
+      "grad_norm": 4.000268288506294,
+      "learning_rate": 1.0766736811983864e-05,
+      "loss": 0.7287,
+      "step": 2488
+    },
+    {
+      "epoch": 1.4740894284868227,
+      "grad_norm": 3.613730312963113,
+      "learning_rate": 1.0760359732634806e-05,
+      "loss": 0.7625,
+      "step": 2489
+    },
+    {
+      "epoch": 1.4746816701214096,
+      "grad_norm": 5.433532028459535,
+      "learning_rate": 1.0753982342255863e-05,
+      "loss": 0.7634,
+      "step": 2490
+    },
+    {
+      "epoch": 1.4752739117559965,
+      "grad_norm": 3.32345033735752,
+      "learning_rate": 1.0747604643455735e-05,
+      "loss": 0.7312,
+      "step": 2491
+    },
+    {
+      "epoch": 1.4758661533905832,
+      "grad_norm": 1.5311040334516028,
+      "learning_rate": 1.0741226638843276e-05,
+      "loss": 0.7367,
+      "step": 2492
+    },
+    {
+      "epoch": 1.4764583950251704,
+      "grad_norm": 2.7319357660771355,
+      "learning_rate": 1.0734848331027437e-05,
+      "loss": 0.7359,
+      "step": 2493
+    },
+    {
+      "epoch": 1.477050636659757,
+      "grad_norm": 18.11777663466048,
+      "learning_rate": 1.072846972261731e-05,
+      "loss": 0.7704,
+      "step": 2494
+    },
+    {
+      "epoch": 1.477642878294344,
+      "grad_norm": 2.4993251519063895,
+      "learning_rate": 1.07220908162221e-05,
+      "loss": 0.7497,
+      "step": 2495
+    },
+    {
+      "epoch": 1.478235119928931,
+      "grad_norm": 1.8417261729994114,
+      "learning_rate": 1.0715711614451146e-05,
+      "loss": 0.7485,
+      "step": 2496
+    },
+    {
+      "epoch": 1.478827361563518,
+      "grad_norm": 2.7762420879853558,
+      "learning_rate": 1.0709332119913889e-05,
+      "loss": 0.7371,
+      "step": 2497
+    },
+    {
+      "epoch": 1.4794196031981048,
+      "grad_norm": 2.8156204757494394,
+      "learning_rate": 1.0702952335219912e-05,
+      "loss": 0.7295,
+      "step": 2498
+    },
+    {
+      "epoch": 1.4800118448326918,
+      "grad_norm": 3.4764320543388645,
+      "learning_rate": 1.0696572262978897e-05,
+      "loss": 0.7117,
+      "step": 2499
+    },
+    {
+      "epoch": 1.4806040864672787,
+      "grad_norm": 2.613928689321406,
+      "learning_rate": 1.0690191905800659e-05,
+      "loss": 0.7364,
+      "step": 2500
+    },
+    {
+      "epoch": 1.4811963281018656,
+      "grad_norm": 3.310002223294027,
+      "learning_rate": 1.0683811266295122e-05,
+      "loss": 0.7071,
+      "step": 2501
+    },
+    {
+      "epoch": 1.4817885697364526,
+      "grad_norm": 1.8031901176722984,
+      "learning_rate": 1.067743034707232e-05,
+      "loss": 0.7363,
+      "step": 2502
+    },
+    {
+      "epoch": 1.4823808113710393,
+      "grad_norm": 2.736654320785856,
+      "learning_rate": 1.0671049150742414e-05,
+      "loss": 0.7292,
+      "step": 2503
+    },
+    {
+      "epoch": 1.4829730530056264,
+      "grad_norm": 6.165916524526913,
+      "learning_rate": 1.066466767991567e-05,
+      "loss": 0.7259,
+      "step": 2504
+    },
+    {
+      "epoch": 1.4835652946402131,
+      "grad_norm": 3.2740591856838286,
+      "learning_rate": 1.065828593720247e-05,
+      "loss": 0.7333,
+      "step": 2505
+    },
+    {
+      "epoch": 1.4841575362748,
+      "grad_norm": 8.686085817094266,
+      "learning_rate": 1.0651903925213304e-05,
+      "loss": 0.694,
+      "step": 2506
+    },
+    {
+      "epoch": 1.484749777909387,
+      "grad_norm": 3.2639051263293744,
+      "learning_rate": 1.0645521646558774e-05,
+      "loss": 0.7347,
+      "step": 2507
+    },
+    {
+      "epoch": 1.485342019543974,
+      "grad_norm": 2.5554270332110334,
+      "learning_rate": 1.0639139103849591e-05,
+      "loss": 0.7331,
+      "step": 2508
+    },
+    {
+      "epoch": 1.4859342611785609,
+      "grad_norm": 2.926347424292201,
+      "learning_rate": 1.0632756299696576e-05,
+      "loss": 0.7181,
+      "step": 2509
+    },
+    {
+      "epoch": 1.4865265028131478,
+      "grad_norm": 2.3947137814149047,
+      "learning_rate": 1.062637323671065e-05,
+      "loss": 0.7079,
+      "step": 2510
+    },
+    {
+      "epoch": 1.4871187444477347,
+      "grad_norm": 2.6108737168898264,
+      "learning_rate": 1.061998991750285e-05,
+      "loss": 0.7863,
+      "step": 2511
+    },
+    {
+      "epoch": 1.4877109860823217,
+      "grad_norm": 2.7866821674655755,
+      "learning_rate": 1.0613606344684309e-05,
+      "loss": 0.7435,
+      "step": 2512
+    },
+    {
+      "epoch": 1.4883032277169086,
+      "grad_norm": 2.1621024720841935,
+      "learning_rate": 1.0607222520866268e-05,
+      "loss": 0.7383,
+      "step": 2513
+    },
+    {
+      "epoch": 1.4888954693514953,
+      "grad_norm": 1.4595498878836146,
+      "learning_rate": 1.060083844866007e-05,
+      "loss": 0.7215,
+      "step": 2514
+    },
+    {
+      "epoch": 1.4894877109860825,
+      "grad_norm": 2.4323235608077267,
+      "learning_rate": 1.0594454130677159e-05,
+      "loss": 0.7737,
+      "step": 2515
+    },
+    {
+      "epoch": 1.4900799526206692,
+      "grad_norm": 2.0951574345823167,
+      "learning_rate": 1.0588069569529085e-05,
+      "loss": 0.751,
+      "step": 2516
+    },
+    {
+      "epoch": 1.490672194255256,
+      "grad_norm": 2.7057235512306264,
+      "learning_rate": 1.0581684767827483e-05,
+      "loss": 0.7164,
+      "step": 2517
+    },
+    {
+      "epoch": 1.491264435889843,
+      "grad_norm": 2.3050182970407462,
+      "learning_rate": 1.0575299728184105e-05,
+      "loss": 0.7266,
+      "step": 2518
+    },
+    {
+      "epoch": 1.49185667752443,
+      "grad_norm": 2.658571871840395,
+      "learning_rate": 1.0568914453210784e-05,
+      "loss": 0.729,
+      "step": 2519
+    },
+    {
+      "epoch": 1.492448919159017,
+      "grad_norm": 2.6694118505347046,
+      "learning_rate": 1.0562528945519463e-05,
+      "loss": 0.7099,
+      "step": 2520
+    },
+    {
+      "epoch": 1.4930411607936038,
+      "grad_norm": 2.7419506378224257,
+      "learning_rate": 1.0556143207722167e-05,
+      "loss": 0.7001,
+      "step": 2521
+    },
+    {
+      "epoch": 1.4936334024281908,
+      "grad_norm": 1.5146904472068337,
+      "learning_rate": 1.0549757242431032e-05,
+      "loss": 0.7012,
+      "step": 2522
+    },
+    {
+      "epoch": 1.4942256440627777,
+      "grad_norm": 2.187917848241182,
+      "learning_rate": 1.0543371052258262e-05,
+      "loss": 0.7006,
+      "step": 2523
+    },
+    {
+      "epoch": 1.4948178856973646,
+      "grad_norm": 3.0173641599728307,
+      "learning_rate": 1.0536984639816183e-05,
+      "loss": 0.7009,
+      "step": 2524
+    },
+    {
+      "epoch": 1.4954101273319513,
+      "grad_norm": 4.342321069678089,
+      "learning_rate": 1.0530598007717188e-05,
+      "loss": 0.7325,
+      "step": 2525
+    },
+    {
+      "epoch": 1.4960023689665383,
+      "grad_norm": 2.168797900177454,
+      "learning_rate": 1.0524211158573772e-05,
+      "loss": 0.7077,
+      "step": 2526
+    },
+    {
+      "epoch": 1.4965946106011252,
+      "grad_norm": 2.459366128626486,
+      "learning_rate": 1.0517824094998514e-05,
+      "loss": 0.7149,
+      "step": 2527
+    },
+    {
+      "epoch": 1.4971868522357121,
+      "grad_norm": 3.5713298768798794,
+      "learning_rate": 1.0511436819604082e-05,
+      "loss": 0.734,
+      "step": 2528
+    },
+    {
+      "epoch": 1.497779093870299,
+      "grad_norm": 1.9883923998857207,
+      "learning_rate": 1.050504933500323e-05,
+      "loss": 0.6835,
+      "step": 2529
+    },
+    {
+      "epoch": 1.498371335504886,
+      "grad_norm": 2.939625720286527,
+      "learning_rate": 1.0498661643808801e-05,
+      "loss": 0.6833,
+      "step": 2530
+    },
+    {
+      "epoch": 1.498963577139473,
+      "grad_norm": 3.2278416983726586,
+      "learning_rate": 1.0492273748633718e-05,
+      "loss": 0.7615,
+      "step": 2531
+    },
+    {
+      "epoch": 1.4995558187740599,
+      "grad_norm": 2.134003837207207,
+      "learning_rate": 1.0485885652090992e-05,
+      "loss": 0.7285,
+      "step": 2532
+    },
+    {
+      "epoch": 1.5001480604086468,
+      "grad_norm": 1.4289709457674888,
+      "learning_rate": 1.0479497356793708e-05,
+      "loss": 0.7301,
+      "step": 2533
+    },
+    {
+      "epoch": 1.5007403020432335,
+      "grad_norm": 2.2718039122799936,
+      "learning_rate": 1.0473108865355046e-05,
+      "loss": 0.6829,
+      "step": 2534
+    },
+    {
+      "epoch": 1.5013325436778207,
+      "grad_norm": 2.1362613649289837,
+      "learning_rate": 1.046672018038825e-05,
+      "loss": 0.7585,
+      "step": 2535
+    },
+    {
+      "epoch": 1.5019247853124074,
+      "grad_norm": 4.9287065917580035,
+      "learning_rate": 1.0460331304506658e-05,
+      "loss": 0.6854,
+      "step": 2536
+    },
+    {
+      "epoch": 1.5025170269469945,
+      "grad_norm": 2.71738703423648,
+      "learning_rate": 1.0453942240323676e-05,
+      "loss": 0.6972,
+      "step": 2537
+    },
+    {
+      "epoch": 1.5031092685815812,
+      "grad_norm": 3.3571433608705967,
+      "learning_rate": 1.044755299045279e-05,
+      "loss": 0.7208,
+      "step": 2538
+    },
+    {
+      "epoch": 1.5037015102161682,
+      "grad_norm": 2.2306864648628104,
+      "learning_rate": 1.0441163557507565e-05,
+      "loss": 0.7,
+      "step": 2539
+    },
+    {
+      "epoch": 1.504293751850755,
+      "grad_norm": 2.127715558269432,
+      "learning_rate": 1.0434773944101637e-05,
+      "loss": 0.7028,
+      "step": 2540
+    },
+    {
+      "epoch": 1.504885993485342,
+      "grad_norm": 3.283305061966627,
+      "learning_rate": 1.0428384152848716e-05,
+      "loss": 0.7194,
+      "step": 2541
+    },
+    {
+      "epoch": 1.505478235119929,
+      "grad_norm": 1.7421560789370591,
+      "learning_rate": 1.0421994186362591e-05,
+      "loss": 0.7146,
+      "step": 2542
+    },
+    {
+      "epoch": 1.5060704767545159,
+      "grad_norm": 2.1998312820966035,
+      "learning_rate": 1.0415604047257108e-05,
+      "loss": 0.7055,
+      "step": 2543
+    },
+    {
+      "epoch": 1.5066627183891028,
+      "grad_norm": 2.8243735302280797,
+      "learning_rate": 1.0409213738146207e-05,
+      "loss": 0.7101,
+      "step": 2544
+    },
+    {
+      "epoch": 1.5072549600236895,
+      "grad_norm": 2.064730143131007,
+      "learning_rate": 1.0402823261643869e-05,
+      "loss": 0.7449,
+      "step": 2545
+    },
+    {
+      "epoch": 1.5078472016582767,
+      "grad_norm": 2.1534779272441016,
+      "learning_rate": 1.039643262036417e-05,
+      "loss": 0.7249,
+      "step": 2546
+    },
+    {
+      "epoch": 1.5084394432928634,
+      "grad_norm": 1.9313484117848871,
+      "learning_rate": 1.039004181692123e-05,
+      "loss": 0.7402,
+      "step": 2547
+    },
+    {
+      "epoch": 1.5090316849274505,
+      "grad_norm": 2.075887908145325,
+      "learning_rate": 1.0383650853929261e-05,
+      "loss": 0.7791,
+      "step": 2548
+    },
+    {
+      "epoch": 1.5096239265620373,
+      "grad_norm": 3.2492748129540985,
+      "learning_rate": 1.0377259734002514e-05,
+      "loss": 0.7529,
+      "step": 2549
+    },
+    {
+      "epoch": 1.5102161681966242,
+      "grad_norm": 1.900962739440968,
+      "learning_rate": 1.0370868459755325e-05,
+      "loss": 0.7034,
+      "step": 2550
+    },
+    {
+      "epoch": 1.5108084098312111,
+      "grad_norm": 1.477785833608308,
+      "learning_rate": 1.0364477033802079e-05,
+      "loss": 0.6937,
+      "step": 2551
+    },
+    {
+      "epoch": 1.511400651465798,
+      "grad_norm": 3.631747743457973,
+      "learning_rate": 1.0358085458757233e-05,
+      "loss": 0.7212,
+      "step": 2552
+    },
+    {
+      "epoch": 1.511992893100385,
+      "grad_norm": 3.5138447414359173,
+      "learning_rate": 1.0351693737235296e-05,
+      "loss": 0.7277,
+      "step": 2553
+    },
+    {
+      "epoch": 1.512585134734972,
+      "grad_norm": 1.3857492580523203,
+      "learning_rate": 1.0345301871850843e-05,
+      "loss": 0.7219,
+      "step": 2554
+    },
+    {
+      "epoch": 1.5131773763695588,
+      "grad_norm": 2.6281307473566673,
+      "learning_rate": 1.0338909865218509e-05,
+      "loss": 0.7223,
+      "step": 2555
+    },
+    {
+      "epoch": 1.5137696180041456,
+      "grad_norm": 11.099970381589284,
+      "learning_rate": 1.0332517719952982e-05,
+      "loss": 0.7396,
+      "step": 2556
+    },
+    {
+      "epoch": 1.5143618596387327,
+      "grad_norm": 5.440070964115479,
+      "learning_rate": 1.0326125438669008e-05,
+      "loss": 0.7263,
+      "step": 2557
+    },
+    {
+      "epoch": 1.5149541012733194,
+      "grad_norm": 3.966364535381863,
+      "learning_rate": 1.0319733023981392e-05,
+      "loss": 0.7212,
+      "step": 2558
+    },
+    {
+      "epoch": 1.5155463429079066,
+      "grad_norm": 9.420534733675245,
+      "learning_rate": 1.031334047850499e-05,
+      "loss": 0.7458,
+      "step": 2559
+    },
+    {
+      "epoch": 1.5161385845424933,
+      "grad_norm": 2.6027648620090997,
+      "learning_rate": 1.030694780485471e-05,
+      "loss": 0.7141,
+      "step": 2560
+    },
+    {
+      "epoch": 1.5167308261770802,
+      "grad_norm": 1.4387286301673126,
+      "learning_rate": 1.030055500564552e-05,
+      "loss": 0.7644,
+      "step": 2561
+    },
+    {
+      "epoch": 1.5173230678116671,
+      "grad_norm": 1.956401956469561,
+      "learning_rate": 1.0294162083492429e-05,
+      "loss": 0.7265,
+      "step": 2562
+    },
+    {
+      "epoch": 1.517915309446254,
+      "grad_norm": 1.9974664355897158,
+      "learning_rate": 1.0287769041010506e-05,
+      "loss": 0.752,
+      "step": 2563
+    },
+    {
+      "epoch": 1.518507551080841,
+      "grad_norm": 2.317095214652806,
+      "learning_rate": 1.0281375880814864e-05,
+      "loss": 0.7235,
+      "step": 2564
+    },
+    {
+      "epoch": 1.519099792715428,
+      "grad_norm": 2.301255206281844,
+      "learning_rate": 1.0274982605520662e-05,
+      "loss": 0.7452,
+      "step": 2565
+    },
+    {
+      "epoch": 1.5196920343500149,
+      "grad_norm": 3.1153435591558387,
+      "learning_rate": 1.0268589217743114e-05,
+      "loss": 0.7009,
+      "step": 2566
+    },
+    {
+      "epoch": 1.5202842759846016,
+      "grad_norm": 8.104588610609271,
+      "learning_rate": 1.0262195720097472e-05,
+      "loss": 0.7581,
+      "step": 2567
+    },
+    {
+      "epoch": 1.5208765176191887,
+      "grad_norm": 1.629901500067734,
+      "learning_rate": 1.0255802115199034e-05,
+      "loss": 0.746,
+      "step": 2568
+    },
+    {
+      "epoch": 1.5214687592537754,
+      "grad_norm": 2.5849788803020193,
+      "learning_rate": 1.0249408405663148e-05,
+      "loss": 0.7137,
+      "step": 2569
+    },
+    {
+      "epoch": 1.5220610008883626,
+      "grad_norm": 1.7706404764967858,
+      "learning_rate": 1.0243014594105201e-05,
+      "loss": 0.6867,
+      "step": 2570
+    },
+    {
+      "epoch": 1.5226532425229493,
+      "grad_norm": 3.0042524859021564,
+      "learning_rate": 1.0236620683140616e-05,
+      "loss": 0.6803,
+      "step": 2571
+    },
+    {
+      "epoch": 1.5232454841575362,
+      "grad_norm": 1.7908570112056044,
+      "learning_rate": 1.0230226675384868e-05,
+      "loss": 0.7065,
+      "step": 2572
+    },
+    {
+      "epoch": 1.5238377257921232,
+      "grad_norm": 2.8915396186908393,
+      "learning_rate": 1.0223832573453463e-05,
+      "loss": 0.7225,
+      "step": 2573
+    },
+    {
+      "epoch": 1.52442996742671,
+      "grad_norm": 10.14931172158288,
+      "learning_rate": 1.021743837996195e-05,
+      "loss": 0.7015,
+      "step": 2574
+    },
+    {
+      "epoch": 1.525022209061297,
+      "grad_norm": 2.3225645983111307,
+      "learning_rate": 1.0211044097525908e-05,
+      "loss": 0.7365,
+      "step": 2575
+    },
+    {
+      "epoch": 1.525614450695884,
+      "grad_norm": 1.8117439188164521,
+      "learning_rate": 1.0204649728760969e-05,
+      "loss": 0.7822,
+      "step": 2576
+    },
+    {
+      "epoch": 1.526206692330471,
+      "grad_norm": 2.4639010150094647,
+      "learning_rate": 1.0198255276282778e-05,
+      "loss": 0.7265,
+      "step": 2577
+    },
+    {
+      "epoch": 1.5267989339650576,
+      "grad_norm": 1.866652089335267,
+      "learning_rate": 1.0191860742707034e-05,
+      "loss": 0.7577,
+      "step": 2578
+    },
+    {
+      "epoch": 1.5273911755996448,
+      "grad_norm": 2.434779909956098,
+      "learning_rate": 1.0185466130649455e-05,
+      "loss": 0.7516,
+      "step": 2579
+    },
+    {
+      "epoch": 1.5279834172342315,
+      "grad_norm": 2.0991586945316416,
+      "learning_rate": 1.0179071442725801e-05,
+      "loss": 0.7407,
+      "step": 2580
+    },
+    {
+      "epoch": 1.5285756588688186,
+      "grad_norm": 2.5982167009442847,
+      "learning_rate": 1.0172676681551857e-05,
+      "loss": 0.7136,
+      "step": 2581
+    },
+    {
+      "epoch": 1.5291679005034053,
+      "grad_norm": 4.3978929309621835,
+      "learning_rate": 1.0166281849743438e-05,
+      "loss": 0.7124,
+      "step": 2582
+    },
+    {
+      "epoch": 1.5297601421379923,
+      "grad_norm": 1.4422002498656223,
+      "learning_rate": 1.0159886949916394e-05,
+      "loss": 0.7052,
+      "step": 2583
+    },
+    {
+      "epoch": 1.5303523837725792,
+      "grad_norm": 2.917865956798247,
+      "learning_rate": 1.0153491984686595e-05,
+      "loss": 0.7503,
+      "step": 2584
+    },
+    {
+      "epoch": 1.5309446254071661,
+      "grad_norm": 1.5687207674041863,
+      "learning_rate": 1.0147096956669945e-05,
+      "loss": 0.7344,
+      "step": 2585
+    },
+    {
+      "epoch": 1.531536867041753,
+      "grad_norm": 2.2546664576414828,
+      "learning_rate": 1.0140701868482365e-05,
+      "loss": 0.7531,
+      "step": 2586
+    },
+    {
+      "epoch": 1.53212910867634,
+      "grad_norm": 1.4308336411744214,
+      "learning_rate": 1.013430672273981e-05,
+      "loss": 0.7343,
+      "step": 2587
+    },
+    {
+      "epoch": 1.532721350310927,
+      "grad_norm": 2.396679670787782,
+      "learning_rate": 1.0127911522058256e-05,
+      "loss": 0.7603,
+      "step": 2588
+    },
+    {
+      "epoch": 1.5333135919455136,
+      "grad_norm": 2.149803137842192,
+      "learning_rate": 1.0121516269053693e-05,
+      "loss": 0.715,
+      "step": 2589
+    },
+    {
+      "epoch": 1.5339058335801008,
+      "grad_norm": 1.5562821807570975,
+      "learning_rate": 1.0115120966342145e-05,
+      "loss": 0.7023,
+      "step": 2590
+    },
+    {
+      "epoch": 1.5344980752146875,
+      "grad_norm": 3.920771426341168,
+      "learning_rate": 1.0108725616539648e-05,
+      "loss": 0.7465,
+      "step": 2591
+    },
+    {
+      "epoch": 1.5350903168492747,
+      "grad_norm": 2.6406023004555945,
+      "learning_rate": 1.0102330222262257e-05,
+      "loss": 0.6824,
+      "step": 2592
+    },
+    {
+      "epoch": 1.5356825584838614,
+      "grad_norm": 1.999159078713688,
+      "learning_rate": 1.0095934786126055e-05,
+      "loss": 0.7184,
+      "step": 2593
+    },
+    {
+      "epoch": 1.5362748001184483,
+      "grad_norm": 2.6618576577531012,
+      "learning_rate": 1.0089539310747127e-05,
+      "loss": 0.6993,
+      "step": 2594
+    },
+    {
+      "epoch": 1.5368670417530352,
+      "grad_norm": 2.088264527488616,
+      "learning_rate": 1.0083143798741587e-05,
+      "loss": 0.7093,
+      "step": 2595
+    },
+    {
+      "epoch": 1.5374592833876222,
+      "grad_norm": 1.2592517848012976,
+      "learning_rate": 1.007674825272556e-05,
+      "loss": 0.7187,
+      "step": 2596
+    },
+    {
+      "epoch": 1.538051525022209,
+      "grad_norm": 2.4043437880802703,
+      "learning_rate": 1.007035267531518e-05,
+      "loss": 0.7162,
+      "step": 2597
+    },
+    {
+      "epoch": 1.5386437666567958,
+      "grad_norm": 1.3595141476229315,
+      "learning_rate": 1.0063957069126602e-05,
+      "loss": 0.7122,
+      "step": 2598
+    },
+    {
+      "epoch": 1.539236008291383,
+      "grad_norm": 1.5459899337725218,
+      "learning_rate": 1.0057561436775982e-05,
+      "loss": 0.7631,
+      "step": 2599
+    },
+    {
+      "epoch": 1.5398282499259697,
+      "grad_norm": 1.6231837605639283,
+      "learning_rate": 1.0051165780879503e-05,
+      "loss": 0.7219,
+      "step": 2600
+    },
+    {
+      "epoch": 1.5404204915605568,
+      "grad_norm": 2.0281678385057265,
+      "learning_rate": 1.0044770104053336e-05,
+      "loss": 0.7281,
+      "step": 2601
+    },
+    {
+      "epoch": 1.5410127331951435,
+      "grad_norm": 1.761034640996413,
+      "learning_rate": 1.0038374408913684e-05,
+      "loss": 0.7147,
+      "step": 2602
+    },
+    {
+      "epoch": 1.5416049748297307,
+      "grad_norm": 2.4171957792447216,
+      "learning_rate": 1.0031978698076738e-05,
+      "loss": 0.7625,
+      "step": 2603
+    },
+    {
+      "epoch": 1.5421972164643174,
+      "grad_norm": 2.4766490230586955,
+      "learning_rate": 1.002558297415871e-05,
+      "loss": 0.7588,
+      "step": 2604
+    },
+    {
+      "epoch": 1.5427894580989043,
+      "grad_norm": 1.6895743443704907,
+      "learning_rate": 1.00191872397758e-05,
+      "loss": 0.7303,
+      "step": 2605
+    },
+    {
+      "epoch": 1.5433816997334913,
+      "grad_norm": 3.162304535278318,
+      "learning_rate": 1.0012791497544238e-05,
+      "loss": 0.7232,
+      "step": 2606
+    },
+    {
+      "epoch": 1.5439739413680782,
+      "grad_norm": 1.6202855977215762,
+      "learning_rate": 1.000639575008023e-05,
+      "loss": 0.7778,
+      "step": 2607
+    },
+    {
+      "epoch": 1.5445661830026651,
+      "grad_norm": 2.9346216243576317,
+      "learning_rate": 1e-05,
+      "loss": 0.7523,
+      "step": 2608
+    },
+    {
+      "epoch": 1.5451584246372518,
+      "grad_norm": 2.5997819364592805,
+      "learning_rate": 9.993604249919773e-06,
+      "loss": 0.7149,
+      "step": 2609
+    },
+    {
+      "epoch": 1.545750666271839,
+      "grad_norm": 2.2690927538275956,
+      "learning_rate": 9.987208502455767e-06,
+      "loss": 0.7492,
+      "step": 2610
+    },
+    {
+      "epoch": 1.5463429079064257,
+      "grad_norm": 2.0044261798649456,
+      "learning_rate": 9.980812760224202e-06,
+      "loss": 0.7161,
+      "step": 2611
+    },
+    {
+      "epoch": 1.5469351495410129,
+      "grad_norm": 1.7008339193116702,
+      "learning_rate": 9.974417025841293e-06,
+      "loss": 0.7075,
+      "step": 2612
+    },
+    {
+      "epoch": 1.5475273911755996,
+      "grad_norm": 2.802313549002366,
+      "learning_rate": 9.968021301923264e-06,
+      "loss": 0.7673,
+      "step": 2613
+    },
+    {
+      "epoch": 1.5481196328101867,
+      "grad_norm": 5.540492098493376,
+      "learning_rate": 9.961625591086321e-06,
+      "loss": 0.6843,
+      "step": 2614
+    },
+    {
+      "epoch": 1.5487118744447734,
+      "grad_norm": 1.7826155088059201,
+      "learning_rate": 9.955229895946666e-06,
+      "loss": 0.6854,
+      "step": 2615
+    },
+    {
+      "epoch": 1.5493041160793604,
+      "grad_norm": 3.214394133267478,
+      "learning_rate": 9.9488342191205e-06,
+      "loss": 0.7126,
+      "step": 2616
+    },
+    {
+      "epoch": 1.5498963577139473,
+      "grad_norm": 3.977691609819524,
+      "learning_rate": 9.942438563224018e-06,
+      "loss": 0.7541,
+      "step": 2617
+    },
+    {
+      "epoch": 1.5504885993485342,
+      "grad_norm": 2.7729164430172393,
+      "learning_rate": 9.936042930873403e-06,
+      "loss": 0.7343,
+      "step": 2618
+    },
+    {
+      "epoch": 1.5510808409831212,
+      "grad_norm": 3.2382422721234945,
+      "learning_rate": 9.929647324684823e-06,
+      "loss": 0.6899,
+      "step": 2619
+    },
+    {
+      "epoch": 1.5516730826177079,
+      "grad_norm": 1.712687804564103,
+      "learning_rate": 9.923251747274441e-06,
+      "loss": 0.7011,
+      "step": 2620
+    },
+    {
+      "epoch": 1.552265324252295,
+      "grad_norm": 2.1866276642869127,
+      "learning_rate": 9.916856201258413e-06,
+      "loss": 0.7067,
+      "step": 2621
+    },
+    {
+      "epoch": 1.5528575658868817,
+      "grad_norm": 1.4026718170049086,
+      "learning_rate": 9.910460689252876e-06,
+      "loss": 0.7055,
+      "step": 2622
+    },
+    {
+      "epoch": 1.5534498075214689,
+      "grad_norm": 1.3836603781583152,
+      "learning_rate": 9.904065213873949e-06,
+      "loss": 0.7652,
+      "step": 2623
+    },
+    {
+      "epoch": 1.5540420491560556,
+      "grad_norm": 2.193906980580802,
+      "learning_rate": 9.897669777737745e-06,
+      "loss": 0.7304,
+      "step": 2624
+    },
+    {
+      "epoch": 1.5546342907906427,
+      "grad_norm": 7.203396030176273,
+      "learning_rate": 9.891274383460354e-06,
+      "loss": 0.7812,
+      "step": 2625
+    },
+    {
+      "epoch": 1.5552265324252295,
+      "grad_norm": 1.8434299981876887,
+      "learning_rate": 9.884879033657859e-06,
+      "loss": 0.748,
+      "step": 2626
+    },
+    {
+      "epoch": 1.5558187740598164,
+      "grad_norm": 1.821123721515522,
+      "learning_rate": 9.878483730946308e-06,
+      "loss": 0.6896,
+      "step": 2627
+    },
+    {
+      "epoch": 1.5564110156944033,
+      "grad_norm": 1.7271066903414265,
+      "learning_rate": 9.872088477941748e-06,
+      "loss": 0.7252,
+      "step": 2628
+    },
+    {
+      "epoch": 1.5570032573289903,
+      "grad_norm": 1.925482681465068,
+      "learning_rate": 9.86569327726019e-06,
+      "loss": 0.7099,
+      "step": 2629
+    },
+    {
+      "epoch": 1.5575954989635772,
+      "grad_norm": 1.6134202565322096,
+      "learning_rate": 9.859298131517639e-06,
+      "loss": 0.7507,
+      "step": 2630
+    },
+    {
+      "epoch": 1.558187740598164,
+      "grad_norm": 3.513343627530898,
+      "learning_rate": 9.852903043330059e-06,
+      "loss": 0.7638,
+      "step": 2631
+    },
+    {
+      "epoch": 1.558779982232751,
+      "grad_norm": 1.442402404811214,
+      "learning_rate": 9.846508015313407e-06,
+      "loss": 0.7419,
+      "step": 2632
+    },
+    {
+      "epoch": 1.5593722238673378,
+      "grad_norm": 2.2944058513698544,
+      "learning_rate": 9.84011305008361e-06,
+      "loss": 0.7012,
+      "step": 2633
+    },
+    {
+      "epoch": 1.559964465501925,
+      "grad_norm": 4.277848662731923,
+      "learning_rate": 9.833718150256567e-06,
+      "loss": 0.7286,
+      "step": 2634
+    },
+    {
+      "epoch": 1.5605567071365116,
+      "grad_norm": 2.2320066876585987,
+      "learning_rate": 9.827323318448148e-06,
+      "loss": 0.7014,
+      "step": 2635
+    },
+    {
+      "epoch": 1.5611489487710986,
+      "grad_norm": 2.592881982866087,
+      "learning_rate": 9.820928557274202e-06,
+      "loss": 0.7711,
+      "step": 2636
+    },
+    {
+      "epoch": 1.5617411904056855,
+      "grad_norm": 2.3317654184305923,
+      "learning_rate": 9.814533869350547e-06,
+      "loss": 0.6949,
+      "step": 2637
+    },
+    {
+      "epoch": 1.5623334320402724,
+      "grad_norm": 3.3958407254332323,
+      "learning_rate": 9.808139257292971e-06,
+      "loss": 0.7517,
+      "step": 2638
+    },
+    {
+      "epoch": 1.5629256736748594,
+      "grad_norm": 1.8868331278047943,
+      "learning_rate": 9.801744723717225e-06,
+      "loss": 0.7492,
+      "step": 2639
+    },
+    {
+      "epoch": 1.5635179153094463,
+      "grad_norm": 1.899167075781751,
+      "learning_rate": 9.795350271239034e-06,
+      "loss": 0.7545,
+      "step": 2640
+    },
+    {
+      "epoch": 1.5641101569440332,
+      "grad_norm": 1.6299887369278279,
+      "learning_rate": 9.78895590247409e-06,
+      "loss": 0.7296,
+      "step": 2641
+    },
+    {
+      "epoch": 1.56470239857862,
+      "grad_norm": 86.37994019353313,
+      "learning_rate": 9.782561620038055e-06,
+      "loss": 0.7508,
+      "step": 2642
+    },
+    {
+      "epoch": 1.565294640213207,
+      "grad_norm": 2.0626173331144932,
+      "learning_rate": 9.77616742654654e-06,
+      "loss": 0.7579,
+      "step": 2643
+    },
+    {
+      "epoch": 1.5658868818477938,
+      "grad_norm": 3.3316806059009325,
+      "learning_rate": 9.769773324615133e-06,
+      "loss": 0.7028,
+      "step": 2644
+    },
+    {
+      "epoch": 1.566479123482381,
+      "grad_norm": 1.7459776374665734,
+      "learning_rate": 9.763379316859386e-06,
+      "loss": 0.711,
+      "step": 2645
+    },
+    {
+      "epoch": 1.5670713651169677,
+      "grad_norm": 4.838875211558403,
+      "learning_rate": 9.756985405894802e-06,
+      "loss": 0.7106,
+      "step": 2646
+    },
+    {
+      "epoch": 1.5676636067515546,
+      "grad_norm": 1.3863030191233248,
+      "learning_rate": 9.750591594336854e-06,
+      "loss": 0.7431,
+      "step": 2647
+    },
+    {
+      "epoch": 1.5682558483861415,
+      "grad_norm": 1.734491733463122,
+      "learning_rate": 9.744197884800968e-06,
+      "loss": 0.7221,
+      "step": 2648
+    },
+    {
+      "epoch": 1.5688480900207284,
+      "grad_norm": 1.4554333705299076,
+      "learning_rate": 9.73780427990253e-06,
+      "loss": 0.7297,
+      "step": 2649
+    },
+    {
+      "epoch": 1.5694403316553154,
+      "grad_norm": 3.2911736777923952,
+      "learning_rate": 9.731410782256889e-06,
+      "loss": 0.7361,
+      "step": 2650
+    },
+    {
+      "epoch": 1.5700325732899023,
+      "grad_norm": 1.3472702183611696,
+      "learning_rate": 9.72501739447934e-06,
+      "loss": 0.7268,
+      "step": 2651
+    },
+    {
+      "epoch": 1.5706248149244892,
+      "grad_norm": 1.7709537123753905,
+      "learning_rate": 9.718624119185138e-06,
+      "loss": 0.7226,
+      "step": 2652
+    },
+    {
+      "epoch": 1.571217056559076,
+      "grad_norm": 2.0914175608541434,
+      "learning_rate": 9.712230958989494e-06,
+      "loss": 0.7413,
+      "step": 2653
+    },
+    {
+      "epoch": 1.571809298193663,
+      "grad_norm": 1.1748662544040405,
+      "learning_rate": 9.705837916507575e-06,
+      "loss": 0.7256,
+      "step": 2654
+    },
+    {
+      "epoch": 1.5724015398282498,
+      "grad_norm": 1.9165526544435088,
+      "learning_rate": 9.699444994354483e-06,
+      "loss": 0.7225,
+      "step": 2655
+    },
+    {
+      "epoch": 1.572993781462837,
+      "grad_norm": 1.4887726373278507,
+      "learning_rate": 9.693052195145292e-06,
+      "loss": 0.7339,
+      "step": 2656
+    },
+    {
+      "epoch": 1.5735860230974237,
+      "grad_norm": 2.8971799022670233,
+      "learning_rate": 9.68665952149501e-06,
+      "loss": 0.7242,
+      "step": 2657
+    },
+    {
+      "epoch": 1.5741782647320106,
+      "grad_norm": 2.3325271582358895,
+      "learning_rate": 9.680266976018613e-06,
+      "loss": 0.7538,
+      "step": 2658
+    },
+    {
+      "epoch": 1.5747705063665975,
+      "grad_norm": 1.4580873725563652,
+      "learning_rate": 9.673874561330994e-06,
+      "loss": 0.7313,
+      "step": 2659
+    },
+    {
+      "epoch": 1.5753627480011845,
+      "grad_norm": 1.498906817025249,
+      "learning_rate": 9.66748228004702e-06,
+      "loss": 0.7552,
+      "step": 2660
+    },
+    {
+      "epoch": 1.5759549896357714,
+      "grad_norm": 1.0553079729268877,
+      "learning_rate": 9.661090134781493e-06,
+      "loss": 0.7426,
+      "step": 2661
+    },
+    {
+      "epoch": 1.5765472312703583,
+      "grad_norm": 1.3421402504246218,
+      "learning_rate": 9.654698128149162e-06,
+      "loss": 0.7538,
+      "step": 2662
+    },
+    {
+      "epoch": 1.5771394729049453,
+      "grad_norm": 2.3722370387111322,
+      "learning_rate": 9.648306262764708e-06,
+      "loss": 0.7284,
+      "step": 2663
+    },
+    {
+      "epoch": 1.577731714539532,
+      "grad_norm": 1.80136298232067,
+      "learning_rate": 9.64191454124277e-06,
+      "loss": 0.7216,
+      "step": 2664
+    },
+    {
+      "epoch": 1.5783239561741191,
+      "grad_norm": 2.4789404827141723,
+      "learning_rate": 9.635522966197923e-06,
+      "loss": 0.6985,
+      "step": 2665
+    },
+    {
+      "epoch": 1.5789161978087058,
+      "grad_norm": 1.6133875143981933,
+      "learning_rate": 9.62913154024468e-06,
+      "loss": 0.7065,
+      "step": 2666
+    },
+    {
+      "epoch": 1.579508439443293,
+      "grad_norm": 1.878564085742203,
+      "learning_rate": 9.622740265997488e-06,
+      "loss": 0.7201,
+      "step": 2667
+    },
+    {
+      "epoch": 1.5801006810778797,
+      "grad_norm": 1.561377075762865,
+      "learning_rate": 9.61634914607074e-06,
+      "loss": 0.7158,
+      "step": 2668
+    },
+    {
+      "epoch": 1.5806929227124666,
+      "grad_norm": 1.7864895574938302,
+      "learning_rate": 9.60995818307877e-06,
+      "loss": 0.729,
+      "step": 2669
+    },
+    {
+      "epoch": 1.5812851643470536,
+      "grad_norm": 1.661603731541404,
+      "learning_rate": 9.603567379635836e-06,
+      "loss": 0.7625,
+      "step": 2670
+    },
+    {
+      "epoch": 1.5818774059816405,
+      "grad_norm": 1.299944010246059,
+      "learning_rate": 9.597176738356134e-06,
+      "loss": 0.7081,
+      "step": 2671
+    },
+    {
+      "epoch": 1.5824696476162274,
+      "grad_norm": 2.1529097270049706,
+      "learning_rate": 9.590786261853798e-06,
+      "loss": 0.6694,
+      "step": 2672
+    },
+    {
+      "epoch": 1.5830618892508144,
+      "grad_norm": 1.4703201941101165,
+      "learning_rate": 9.584395952742892e-06,
+      "loss": 0.696,
+      "step": 2673
+    },
+    {
+      "epoch": 1.5836541308854013,
+      "grad_norm": 1.9214984260880197,
+      "learning_rate": 9.578005813637414e-06,
+      "loss": 0.7194,
+      "step": 2674
+    },
+    {
+      "epoch": 1.584246372519988,
+      "grad_norm": 1.3154793348516993,
+      "learning_rate": 9.571615847151287e-06,
+      "loss": 0.7264,
+      "step": 2675
+    },
+    {
+      "epoch": 1.5848386141545752,
+      "grad_norm": 2.0731383642895658,
+      "learning_rate": 9.565226055898366e-06,
+      "loss": 0.7398,
+      "step": 2676
+    },
+    {
+      "epoch": 1.5854308557891619,
+      "grad_norm": 1.269718722070663,
+      "learning_rate": 9.558836442492437e-06,
+      "loss": 0.7618,
+      "step": 2677
+    },
+    {
+      "epoch": 1.586023097423749,
+      "grad_norm": 2.326022738330298,
+      "learning_rate": 9.552447009547214e-06,
+      "loss": 0.7153,
+      "step": 2678
+    },
+    {
+      "epoch": 1.5866153390583357,
+      "grad_norm": 2.108051114609614,
+      "learning_rate": 9.546057759676328e-06,
+      "loss": 0.6837,
+      "step": 2679
+    },
+    {
+      "epoch": 1.5872075806929227,
+      "grad_norm": 1.4247839172166672,
+      "learning_rate": 9.539668695493344e-06,
+      "loss": 0.7276,
+      "step": 2680
+    },
+    {
+      "epoch": 1.5877998223275096,
+      "grad_norm": 1.0542131540553648,
+      "learning_rate": 9.53327981961175e-06,
+      "loss": 0.7446,
+      "step": 2681
+    },
+    {
+      "epoch": 1.5883920639620965,
+      "grad_norm": 1.8059660461442077,
+      "learning_rate": 9.52689113464496e-06,
+      "loss": 0.731,
+      "step": 2682
+    },
+    {
+      "epoch": 1.5889843055966835,
+      "grad_norm": 1.568683168663255,
+      "learning_rate": 9.520502643206293e-06,
+      "loss": 0.712,
+      "step": 2683
+    },
+    {
+      "epoch": 1.5895765472312704,
+      "grad_norm": 3.4501788436091254,
+      "learning_rate": 9.514114347909011e-06,
+      "loss": 0.7132,
+      "step": 2684
+    },
+    {
+      "epoch": 1.5901687888658573,
+      "grad_norm": 2.0059923334239707,
+      "learning_rate": 9.507726251366283e-06,
+      "loss": 0.7292,
+      "step": 2685
+    },
+    {
+      "epoch": 1.590761030500444,
+      "grad_norm": 3.8951776491133745,
+      "learning_rate": 9.501338356191204e-06,
+      "loss": 0.7396,
+      "step": 2686
+    },
+    {
+      "epoch": 1.5913532721350312,
+      "grad_norm": 2.4788086356042216,
+      "learning_rate": 9.494950664996771e-06,
+      "loss": 0.7391,
+      "step": 2687
+    },
+    {
+      "epoch": 1.591945513769618,
+      "grad_norm": 2.9671138222012954,
+      "learning_rate": 9.488563180395922e-06,
+      "loss": 0.7501,
+      "step": 2688
+    },
+    {
+      "epoch": 1.592537755404205,
+      "grad_norm": 1.7325289372388857,
+      "learning_rate": 9.482175905001489e-06,
+      "loss": 0.7592,
+      "step": 2689
+    },
+    {
+      "epoch": 1.5931299970387918,
+      "grad_norm": 2.5802186860418552,
+      "learning_rate": 9.475788841426232e-06,
+      "loss": 0.6955,
+      "step": 2690
+    },
+    {
+      "epoch": 1.5937222386733787,
+      "grad_norm": 2.804837641485897,
+      "learning_rate": 9.469401992282817e-06,
+      "loss": 0.7185,
+      "step": 2691
+    },
+    {
+      "epoch": 1.5943144803079656,
+      "grad_norm": 2.4503401391313893,
+      "learning_rate": 9.463015360183819e-06,
+      "loss": 0.7131,
+      "step": 2692
+    },
+    {
+      "epoch": 1.5949067219425526,
+      "grad_norm": 2.1268807947669606,
+      "learning_rate": 9.456628947741738e-06,
+      "loss": 0.7144,
+      "step": 2693
+    },
+    {
+      "epoch": 1.5954989635771395,
+      "grad_norm": 2.110094339237852,
+      "learning_rate": 9.450242757568975e-06,
+      "loss": 0.7282,
+      "step": 2694
+    },
+    {
+      "epoch": 1.5960912052117264,
+      "grad_norm": 1.7700346930524404,
+      "learning_rate": 9.443856792277836e-06,
+      "loss": 0.7554,
+      "step": 2695
+    },
+    {
+      "epoch": 1.5966834468463134,
+      "grad_norm": 1.0280242622620175,
+      "learning_rate": 9.43747105448054e-06,
+      "loss": 0.7444,
+      "step": 2696
+    },
+    {
+      "epoch": 1.5972756884809,
+      "grad_norm": 1.3872664321744306,
+      "learning_rate": 9.431085546789218e-06,
+      "loss": 0.7354,
+      "step": 2697
+    },
+    {
+      "epoch": 1.5978679301154872,
+      "grad_norm": 1.2479775404016484,
+      "learning_rate": 9.424700271815901e-06,
+      "loss": 0.7271,
+      "step": 2698
+    },
+    {
+      "epoch": 1.598460171750074,
+      "grad_norm": 1.6017123862586502,
+      "learning_rate": 9.41831523217252e-06,
+      "loss": 0.7117,
+      "step": 2699
+    },
+    {
+      "epoch": 1.599052413384661,
+      "grad_norm": 3.210802359158483,
+      "learning_rate": 9.41193043047092e-06,
+      "loss": 0.7564,
+      "step": 2700
+    },
+    {
+      "epoch": 1.5996446550192478,
+      "grad_norm": 3.6114823786582515,
+      "learning_rate": 9.405545869322843e-06,
+      "loss": 0.7498,
+      "step": 2701
+    },
+    {
+      "epoch": 1.6002368966538347,
+      "grad_norm": 1.6710996935533942,
+      "learning_rate": 9.399161551339933e-06,
+      "loss": 0.7275,
+      "step": 2702
+    },
+    {
+      "epoch": 1.6008291382884217,
+      "grad_norm": 1.6984346113201354,
+      "learning_rate": 9.392777479133736e-06,
+      "loss": 0.7232,
+      "step": 2703
+    },
+    {
+      "epoch": 1.6014213799230086,
+      "grad_norm": 1.9419725358546698,
+      "learning_rate": 9.386393655315696e-06,
+      "loss": 0.7397,
+      "step": 2704
+    },
+    {
+      "epoch": 1.6020136215575955,
+      "grad_norm": 3.9234250963518043,
+      "learning_rate": 9.380010082497152e-06,
+      "loss": 0.7185,
+      "step": 2705
+    },
+    {
+      "epoch": 1.6026058631921825,
+      "grad_norm": 1.6194229418389212,
+      "learning_rate": 9.373626763289352e-06,
+      "loss": 0.7287,
+      "step": 2706
+    },
+    {
+      "epoch": 1.6031981048267694,
+      "grad_norm": 1.3815447699004129,
+      "learning_rate": 9.367243700303427e-06,
+      "loss": 0.7354,
+      "step": 2707
+    },
+    {
+      "epoch": 1.603790346461356,
+      "grad_norm": 2.310394866629508,
+      "learning_rate": 9.36086089615041e-06,
+      "loss": 0.7429,
+      "step": 2708
+    },
+    {
+      "epoch": 1.6043825880959433,
+      "grad_norm": 24.261527338301693,
+      "learning_rate": 9.354478353441226e-06,
+      "loss": 0.7592,
+      "step": 2709
+    },
+    {
+      "epoch": 1.60497482973053,
+      "grad_norm": 2.170178314510858,
+      "learning_rate": 9.3480960747867e-06,
+      "loss": 0.7337,
+      "step": 2710
+    },
+    {
+      "epoch": 1.6055670713651171,
+      "grad_norm": 3.053942328323577,
+      "learning_rate": 9.341714062797533e-06,
+      "loss": 0.7174,
+      "step": 2711
+    },
+    {
+      "epoch": 1.6061593129997038,
+      "grad_norm": 1.3735266945797255,
+      "learning_rate": 9.335332320084331e-06,
+      "loss": 0.7538,
+      "step": 2712
+    },
+    {
+      "epoch": 1.6067515546342908,
+      "grad_norm": 1.8525060260331747,
+      "learning_rate": 9.32895084925759e-06,
+      "loss": 0.7206,
+      "step": 2713
+    },
+    {
+      "epoch": 1.6073437962688777,
+      "grad_norm": 1.6312829687512218,
+      "learning_rate": 9.322569652927685e-06,
+      "loss": 0.7047,
+      "step": 2714
+    },
+    {
+      "epoch": 1.6079360379034646,
+      "grad_norm": 1.5960390029649218,
+      "learning_rate": 9.316188733704883e-06,
+      "loss": 0.7373,
+      "step": 2715
+    },
+    {
+      "epoch": 1.6085282795380516,
+      "grad_norm": 1.6939174639156191,
+      "learning_rate": 9.309808094199343e-06,
+      "loss": 0.7343,
+      "step": 2716
+    },
+    {
+      "epoch": 1.6091205211726385,
+      "grad_norm": 5.647495987042836,
+      "learning_rate": 9.303427737021105e-06,
+      "loss": 0.7441,
+      "step": 2717
+    },
+    {
+      "epoch": 1.6097127628072254,
+      "grad_norm": 1.0834842055792864,
+      "learning_rate": 9.297047664780093e-06,
+      "loss": 0.7104,
+      "step": 2718
+    },
+    {
+      "epoch": 1.6103050044418121,
+      "grad_norm": 1.3533184427520821,
+      "learning_rate": 9.290667880086115e-06,
+      "loss": 0.7428,
+      "step": 2719
+    },
+    {
+      "epoch": 1.6108972460763993,
+      "grad_norm": 1.1679095552719452,
+      "learning_rate": 9.284288385548858e-06,
+      "loss": 0.7262,
+      "step": 2720
+    },
+    {
+      "epoch": 1.611489487710986,
+      "grad_norm": 8.140610596435375,
+      "learning_rate": 9.2779091837779e-06,
+      "loss": 0.7327,
+      "step": 2721
+    },
+    {
+      "epoch": 1.6120817293455731,
+      "grad_norm": 1.2293837136922272,
+      "learning_rate": 9.271530277382695e-06,
+      "loss": 0.7129,
+      "step": 2722
+    },
+    {
+      "epoch": 1.6126739709801599,
+      "grad_norm": 1.1470847973636114,
+      "learning_rate": 9.265151668972566e-06,
+      "loss": 0.6852,
+      "step": 2723
+    },
+    {
+      "epoch": 1.6132662126147468,
+      "grad_norm": 2.4405096525740038,
+      "learning_rate": 9.258773361156725e-06,
+      "loss": 0.725,
+      "step": 2724
+    },
+    {
+      "epoch": 1.6138584542493337,
+      "grad_norm": 1.336420955664339,
+      "learning_rate": 9.252395356544263e-06,
+      "loss": 0.6653,
+      "step": 2725
+    },
+    {
+      "epoch": 1.6144506958839207,
+      "grad_norm": 3.435337274510319,
+      "learning_rate": 9.246017657744142e-06,
+      "loss": 0.7157,
+      "step": 2726
+    },
+    {
+      "epoch": 1.6150429375185076,
+      "grad_norm": 1.7205900312208158,
+      "learning_rate": 9.239640267365197e-06,
+      "loss": 0.6767,
+      "step": 2727
+    },
+    {
+      "epoch": 1.6156351791530945,
+      "grad_norm": 0.9404428582680933,
+      "learning_rate": 9.233263188016138e-06,
+      "loss": 0.7388,
+      "step": 2728
+    },
+    {
+      "epoch": 1.6162274207876814,
+      "grad_norm": 1.643384773023064,
+      "learning_rate": 9.22688642230555e-06,
+      "loss": 0.7729,
+      "step": 2729
+    },
+    {
+      "epoch": 1.6168196624222682,
+      "grad_norm": 0.9226766736351931,
+      "learning_rate": 9.220509972841893e-06,
+      "loss": 0.7097,
+      "step": 2730
+    },
+    {
+      "epoch": 1.6174119040568553,
+      "grad_norm": 1.292558562901911,
+      "learning_rate": 9.214133842233486e-06,
+      "loss": 0.7638,
+      "step": 2731
+    },
+    {
+      "epoch": 1.618004145691442,
+      "grad_norm": 1.268236145418197,
+      "learning_rate": 9.207758033088533e-06,
+      "loss": 0.7443,
+      "step": 2732
+    },
+    {
+      "epoch": 1.6185963873260292,
+      "grad_norm": 1.152636990406385,
+      "learning_rate": 9.20138254801509e-06,
+      "loss": 0.7477,
+      "step": 2733
+    },
+    {
+      "epoch": 1.6191886289606159,
+      "grad_norm": 2.325992339134092,
+      "learning_rate": 9.195007389621098e-06,
+      "loss": 0.7597,
+      "step": 2734
+    },
+    {
+      "epoch": 1.6197808705952028,
+      "grad_norm": 1.3096504733315142,
+      "learning_rate": 9.188632560514345e-06,
+      "loss": 0.7028,
+      "step": 2735
+    },
+    {
+      "epoch": 1.6203731122297897,
+      "grad_norm": 1.0546949833315633,
+      "learning_rate": 9.182258063302504e-06,
+      "loss": 0.7131,
+      "step": 2736
+    },
+    {
+      "epoch": 1.6209653538643767,
+      "grad_norm": 0.9210797626479261,
+      "learning_rate": 9.175883900593095e-06,
+      "loss": 0.7453,
+      "step": 2737
+    },
+    {
+      "epoch": 1.6215575954989636,
+      "grad_norm": 1.5075058070737895,
+      "learning_rate": 9.16951007499352e-06,
+      "loss": 0.7248,
+      "step": 2738
+    },
+    {
+      "epoch": 1.6221498371335505,
+      "grad_norm": 1.169767212334899,
+      "learning_rate": 9.163136589111019e-06,
+      "loss": 0.732,
+      "step": 2739
+    },
+    {
+      "epoch": 1.6227420787681375,
+      "grad_norm": 1.615392214615148,
+      "learning_rate": 9.156763445552714e-06,
+      "loss": 0.7017,
+      "step": 2740
+    },
+    {
+      "epoch": 1.6233343204027242,
+      "grad_norm": 1.0292405348020153,
+      "learning_rate": 9.150390646925578e-06,
+      "loss": 0.69,
+      "step": 2741
+    },
+    {
+      "epoch": 1.6239265620373113,
+      "grad_norm": 1.8762128663128925,
+      "learning_rate": 9.144018195836445e-06,
+      "loss": 0.7546,
+      "step": 2742
+    },
+    {
+      "epoch": 1.624518803671898,
+      "grad_norm": 0.8711466459720059,
+      "learning_rate": 9.137646094892e-06,
+      "loss": 0.7039,
+      "step": 2743
+    },
+    {
+      "epoch": 1.6251110453064852,
+      "grad_norm": 0.9463455729109375,
+      "learning_rate": 9.131274346698797e-06,
+      "loss": 0.7569,
+      "step": 2744
+    },
+    {
+      "epoch": 1.625703286941072,
+      "grad_norm": 0.9011458492231906,
+      "learning_rate": 9.124902953863235e-06,
+      "loss": 0.7567,
+      "step": 2745
+    },
+    {
+      "epoch": 1.6262955285756588,
+      "grad_norm": 2.289091980573055,
+      "learning_rate": 9.118531918991578e-06,
+      "loss": 0.749,
+      "step": 2746
+    },
+    {
+      "epoch": 1.6268877702102458,
+      "grad_norm": 1.2146535737974766,
+      "learning_rate": 9.112161244689931e-06,
+      "loss": 0.7254,
+      "step": 2747
+    },
+    {
+      "epoch": 1.6274800118448327,
+      "grad_norm": 2.4627393306018646,
+      "learning_rate": 9.105790933564259e-06,
+      "loss": 0.7543,
+      "step": 2748
+    },
+    {
+      "epoch": 1.6280722534794196,
+      "grad_norm": 1.002159085261613,
+      "learning_rate": 9.09942098822038e-06,
+      "loss": 0.7394,
+      "step": 2749
+    },
+    {
+      "epoch": 1.6286644951140063,
+      "grad_norm": 1.639564743885887,
+      "learning_rate": 9.09305141126396e-06,
+      "loss": 0.7178,
+      "step": 2750
+    },
+    {
+      "epoch": 1.6292567367485935,
+      "grad_norm": 0.9813191864395472,
+      "learning_rate": 9.086682205300512e-06,
+      "loss": 0.7354,
+      "step": 2751
+    },
+    {
+      "epoch": 1.6298489783831802,
+      "grad_norm": 2.1747632589697155,
+      "learning_rate": 9.080313372935399e-06,
+      "loss": 0.7072,
+      "step": 2752
+    },
+    {
+      "epoch": 1.6304412200177674,
+      "grad_norm": 0.8902643816753388,
+      "learning_rate": 9.073944916773835e-06,
+      "loss": 0.7272,
+      "step": 2753
+    },
+    {
+      "epoch": 1.631033461652354,
+      "grad_norm": 0.883321085183472,
+      "learning_rate": 9.067576839420876e-06,
+      "loss": 0.7433,
+      "step": 2754
+    },
+    {
+      "epoch": 1.6316257032869412,
+      "grad_norm": 1.191509270778969,
+      "learning_rate": 9.06120914348142e-06,
+      "loss": 0.7682,
+      "step": 2755
+    },
+    {
+      "epoch": 1.632217944921528,
+      "grad_norm": 0.8931773470587896,
+      "learning_rate": 9.054841831560216e-06,
+      "loss": 0.7217,
+      "step": 2756
+    },
+    {
+      "epoch": 1.6328101865561149,
+      "grad_norm": 0.9206149509553022,
+      "learning_rate": 9.04847490626185e-06,
+      "loss": 0.7244,
+      "step": 2757
+    },
+    {
+      "epoch": 1.6334024281907018,
+      "grad_norm": 0.9602886650943193,
+      "learning_rate": 9.042108370190757e-06,
+      "loss": 0.7327,
+      "step": 2758
+    },
+    {
+      "epoch": 1.6339946698252887,
+      "grad_norm": 1.0731812685624111,
+      "learning_rate": 9.035742225951203e-06,
+      "loss": 0.7128,
+      "step": 2759
+    },
+    {
+      "epoch": 1.6345869114598757,
+      "grad_norm": 1.0001996418004009,
+      "learning_rate": 9.029376476147303e-06,
+      "loss": 0.7283,
+      "step": 2760
+    },
+    {
+      "epoch": 1.6351791530944624,
+      "grad_norm": 1.0101272999135227,
+      "learning_rate": 9.023011123383002e-06,
+      "loss": 0.7698,
+      "step": 2761
+    },
+    {
+      "epoch": 1.6357713947290495,
+      "grad_norm": 7.445183865475451,
+      "learning_rate": 9.016646170262096e-06,
+      "loss": 0.7248,
+      "step": 2762
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 1.0354333262038447,
+      "learning_rate": 9.010281619388198e-06,
+      "loss": 0.7165,
+      "step": 2763
+    },
+    {
+      "epoch": 1.6369558779982234,
+      "grad_norm": 1.0303790501337338,
+      "learning_rate": 9.003917473364774e-06,
+      "loss": 0.773,
+      "step": 2764
+    },
+    {
+      "epoch": 1.63754811963281,
+      "grad_norm": 1.0048176835409444,
+      "learning_rate": 8.997553734795115e-06,
+      "loss": 0.7446,
+      "step": 2765
+    },
+    {
+      "epoch": 1.6381403612673973,
+      "grad_norm": 0.8644074880461828,
+      "learning_rate": 8.991190406282352e-06,
+      "loss": 0.7565,
+      "step": 2766
+    },
+    {
+      "epoch": 1.638732602901984,
+      "grad_norm": 1.484278848736965,
+      "learning_rate": 8.984827490429437e-06,
+      "loss": 0.7242,
+      "step": 2767
+    },
+    {
+      "epoch": 1.639324844536571,
+      "grad_norm": 1.2307819145316636,
+      "learning_rate": 8.978464989839165e-06,
+      "loss": 0.7396,
+      "step": 2768
+    },
+    {
+      "epoch": 1.6399170861711578,
+      "grad_norm": 1.2176803902884394,
+      "learning_rate": 8.972102907114157e-06,
+      "loss": 0.745,
+      "step": 2769
+    },
+    {
+      "epoch": 1.6405093278057448,
+      "grad_norm": 1.0887898932988884,
+      "learning_rate": 8.965741244856864e-06,
+      "loss": 0.6943,
+      "step": 2770
+    },
+    {
+      "epoch": 1.6411015694403317,
+      "grad_norm": 0.988196915871078,
+      "learning_rate": 8.959380005669559e-06,
+      "loss": 0.7085,
+      "step": 2771
+    },
+    {
+      "epoch": 1.6416938110749184,
+      "grad_norm": 0.9375850317494164,
+      "learning_rate": 8.953019192154344e-06,
+      "loss": 0.7181,
+      "step": 2772
+    },
+    {
+      "epoch": 1.6422860527095056,
+      "grad_norm": 1.0916414927857014,
+      "learning_rate": 8.946658806913158e-06,
+      "loss": 0.7182,
+      "step": 2773
+    },
+    {
+      "epoch": 1.6428782943440923,
+      "grad_norm": 0.9559469293114459,
+      "learning_rate": 8.940298852547753e-06,
+      "loss": 0.6924,
+      "step": 2774
+    },
+    {
+      "epoch": 1.6434705359786794,
+      "grad_norm": 0.9122282528732937,
+      "learning_rate": 8.933939331659707e-06,
+      "loss": 0.6802,
+      "step": 2775
+    },
+    {
+      "epoch": 1.6440627776132661,
+      "grad_norm": 1.2827645783241866,
+      "learning_rate": 8.927580246850418e-06,
+      "loss": 0.7219,
+      "step": 2776
+    },
+    {
+      "epoch": 1.6446550192478533,
+      "grad_norm": 1.0856897372943664,
+      "learning_rate": 8.921221600721115e-06,
+      "loss": 0.7749,
+      "step": 2777
+    },
+    {
+      "epoch": 1.64524726088244,
+      "grad_norm": 0.9976872404160881,
+      "learning_rate": 8.914863395872844e-06,
+      "loss": 0.6998,
+      "step": 2778
+    },
+    {
+      "epoch": 1.645839502517027,
+      "grad_norm": 0.9187575917239228,
+      "learning_rate": 8.908505634906461e-06,
+      "loss": 0.6714,
+      "step": 2779
+    },
+    {
+      "epoch": 1.6464317441516139,
+      "grad_norm": 0.8887264946359307,
+      "learning_rate": 8.90214832042265e-06,
+      "loss": 0.7127,
+      "step": 2780
+    },
+    {
+      "epoch": 1.6470239857862008,
+      "grad_norm": 0.8836649493280779,
+      "learning_rate": 8.895791455021912e-06,
+      "loss": 0.716,
+      "step": 2781
+    },
+    {
+      "epoch": 1.6476162274207877,
+      "grad_norm": 0.9011624604992315,
+      "learning_rate": 8.889435041304565e-06,
+      "loss": 0.723,
+      "step": 2782
+    },
+    {
+      "epoch": 1.6482084690553744,
+      "grad_norm": 1.2551214513482096,
+      "learning_rate": 8.88307908187073e-06,
+      "loss": 0.7301,
+      "step": 2783
+    },
+    {
+      "epoch": 1.6488007106899616,
+      "grad_norm": 1.030561508128019,
+      "learning_rate": 8.876723579320363e-06,
+      "loss": 0.6991,
+      "step": 2784
+    },
+    {
+      "epoch": 1.6493929523245483,
+      "grad_norm": 0.9450181046757289,
+      "learning_rate": 8.870368536253213e-06,
+      "loss": 0.7391,
+      "step": 2785
+    },
+    {
+      "epoch": 1.6499851939591355,
+      "grad_norm": 1.0028902776005366,
+      "learning_rate": 8.86401395526886e-06,
+      "loss": 0.7461,
+      "step": 2786
+    },
+    {
+      "epoch": 1.6505774355937222,
+      "grad_norm": 1.0849709154792493,
+      "learning_rate": 8.857659838966672e-06,
+      "loss": 0.7124,
+      "step": 2787
+    },
+    {
+      "epoch": 1.651169677228309,
+      "grad_norm": 1.418277931700241,
+      "learning_rate": 8.85130618994585e-06,
+      "loss": 0.7166,
+      "step": 2788
+    },
+    {
+      "epoch": 1.651761918862896,
+      "grad_norm": 1.0799657812191223,
+      "learning_rate": 8.844953010805388e-06,
+      "loss": 0.6659,
+      "step": 2789
+    },
+    {
+      "epoch": 1.652354160497483,
+      "grad_norm": 0.9047344447663203,
+      "learning_rate": 8.838600304144102e-06,
+      "loss": 0.7237,
+      "step": 2790
+    },
+    {
+      "epoch": 1.65294640213207,
+      "grad_norm": 0.8954991208705074,
+      "learning_rate": 8.832248072560594e-06,
+      "loss": 0.7404,
+      "step": 2791
+    },
+    {
+      "epoch": 1.6535386437666568,
+      "grad_norm": 1.255225726733933,
+      "learning_rate": 8.825896318653294e-06,
+      "loss": 0.718,
+      "step": 2792
+    },
+    {
+      "epoch": 1.6541308854012438,
+      "grad_norm": 1.8935384231370476,
+      "learning_rate": 8.81954504502042e-06,
+      "loss": 0.741,
+      "step": 2793
+    },
+    {
+      "epoch": 1.6547231270358305,
+      "grad_norm": 0.8395781786009291,
+      "learning_rate": 8.813194254260006e-06,
+      "loss": 0.7077,
+      "step": 2794
+    },
+    {
+      "epoch": 1.6553153686704176,
+      "grad_norm": 1.1680365597009021,
+      "learning_rate": 8.806843948969875e-06,
+      "loss": 0.7082,
+      "step": 2795
+    },
+    {
+      "epoch": 1.6559076103050043,
+      "grad_norm": 0.912567977351368,
+      "learning_rate": 8.800494131747667e-06,
+      "loss": 0.7654,
+      "step": 2796
+    },
+    {
+      "epoch": 1.6564998519395915,
+      "grad_norm": 2.4309897712239166,
+      "learning_rate": 8.794144805190809e-06,
+      "loss": 0.7536,
+      "step": 2797
+    },
+    {
+      "epoch": 1.6570920935741782,
+      "grad_norm": 1.0403551057860285,
+      "learning_rate": 8.787795971896536e-06,
+      "loss": 0.7459,
+      "step": 2798
+    },
+    {
+      "epoch": 1.6576843352087651,
+      "grad_norm": 1.5383317536648582,
+      "learning_rate": 8.781447634461874e-06,
+      "loss": 0.722,
+      "step": 2799
+    },
+    {
+      "epoch": 1.658276576843352,
+      "grad_norm": 0.9088453416120332,
+      "learning_rate": 8.775099795483651e-06,
+      "loss": 0.7737,
+      "step": 2800
+    },
+    {
+      "epoch": 1.658868818477939,
+      "grad_norm": 1.042835871852608,
+      "learning_rate": 8.768752457558492e-06,
+      "loss": 0.7334,
+      "step": 2801
+    },
+    {
+      "epoch": 1.659461060112526,
+      "grad_norm": 1.0274279053461044,
+      "learning_rate": 8.762405623282817e-06,
+      "loss": 0.6733,
+      "step": 2802
+    },
+    {
+      "epoch": 1.6600533017471129,
+      "grad_norm": 0.9918727284685397,
+      "learning_rate": 8.756059295252833e-06,
+      "loss": 0.7416,
+      "step": 2803
+    },
+    {
+      "epoch": 1.6606455433816998,
+      "grad_norm": 1.306893253891884,
+      "learning_rate": 8.749713476064547e-06,
+      "loss": 0.7739,
+      "step": 2804
+    },
+    {
+      "epoch": 1.6612377850162865,
+      "grad_norm": 1.2615372668097604,
+      "learning_rate": 8.743368168313757e-06,
+      "loss": 0.7619,
+      "step": 2805
+    },
+    {
+      "epoch": 1.6618300266508736,
+      "grad_norm": 0.905731661272337,
+      "learning_rate": 8.737023374596051e-06,
+      "loss": 0.73,
+      "step": 2806
+    },
+    {
+      "epoch": 1.6624222682854604,
+      "grad_norm": 1.609306011429277,
+      "learning_rate": 8.730679097506804e-06,
+      "loss": 0.7174,
+      "step": 2807
+    },
+    {
+      "epoch": 1.6630145099200475,
+      "grad_norm": 1.294248981532454,
+      "learning_rate": 8.724335339641185e-06,
+      "loss": 0.7094,
+      "step": 2808
+    },
+    {
+      "epoch": 1.6636067515546342,
+      "grad_norm": 0.939745188957189,
+      "learning_rate": 8.717992103594142e-06,
+      "loss": 0.6777,
+      "step": 2809
+    },
+    {
+      "epoch": 1.6641989931892212,
+      "grad_norm": 1.3256782257737456,
+      "learning_rate": 8.711649391960424e-06,
+      "loss": 0.6926,
+      "step": 2810
+    },
+    {
+      "epoch": 1.664791234823808,
+      "grad_norm": 0.8735845886685419,
+      "learning_rate": 8.705307207334552e-06,
+      "loss": 0.7247,
+      "step": 2811
+    },
+    {
+      "epoch": 1.665383476458395,
+      "grad_norm": 1.0270651510259472,
+      "learning_rate": 8.698965552310834e-06,
+      "loss": 0.7213,
+      "step": 2812
+    },
+    {
+      "epoch": 1.665975718092982,
+      "grad_norm": 0.8887392623106478,
+      "learning_rate": 8.692624429483364e-06,
+      "loss": 0.7645,
+      "step": 2813
+    },
+    {
+      "epoch": 1.6665679597275689,
+      "grad_norm": 1.1018807997801454,
+      "learning_rate": 8.686283841446027e-06,
+      "loss": 0.7248,
+      "step": 2814
+    },
+    {
+      "epoch": 1.6671602013621558,
+      "grad_norm": 0.9580382702616205,
+      "learning_rate": 8.679943790792466e-06,
+      "loss": 0.704,
+      "step": 2815
+    },
+    {
+      "epoch": 1.6677524429967425,
+      "grad_norm": 1.775823863507453,
+      "learning_rate": 8.673604280116127e-06,
+      "loss": 0.7412,
+      "step": 2816
+    },
+    {
+      "epoch": 1.6683446846313297,
+      "grad_norm": 0.8788118750792545,
+      "learning_rate": 8.667265312010224e-06,
+      "loss": 0.7352,
+      "step": 2817
+    },
+    {
+      "epoch": 1.6689369262659164,
+      "grad_norm": 1.55508648257972,
+      "learning_rate": 8.660926889067753e-06,
+      "loss": 0.7357,
+      "step": 2818
+    },
+    {
+      "epoch": 1.6695291679005035,
+      "grad_norm": 1.062045887571016,
+      "learning_rate": 8.654589013881481e-06,
+      "loss": 0.7618,
+      "step": 2819
+    },
+    {
+      "epoch": 1.6701214095350903,
+      "grad_norm": 0.8983582875998769,
+      "learning_rate": 8.648251689043961e-06,
+      "loss": 0.7351,
+      "step": 2820
+    },
+    {
+      "epoch": 1.6707136511696772,
+      "grad_norm": 1.03505281944072,
+      "learning_rate": 8.641914917147512e-06,
+      "loss": 0.7224,
+      "step": 2821
+    },
+    {
+      "epoch": 1.6713058928042641,
+      "grad_norm": 0.9169880645190774,
+      "learning_rate": 8.635578700784232e-06,
+      "loss": 0.7079,
+      "step": 2822
+    },
+    {
+      "epoch": 1.671898134438851,
+      "grad_norm": 1.2057396000296163,
+      "learning_rate": 8.629243042545989e-06,
+      "loss": 0.7403,
+      "step": 2823
+    },
+    {
+      "epoch": 1.672490376073438,
+      "grad_norm": 1.0465851255608474,
+      "learning_rate": 8.622907945024418e-06,
+      "loss": 0.7507,
+      "step": 2824
+    },
+    {
+      "epoch": 1.673082617708025,
+      "grad_norm": 0.9218356861943021,
+      "learning_rate": 8.616573410810938e-06,
+      "loss": 0.6975,
+      "step": 2825
+    },
+    {
+      "epoch": 1.6736748593426118,
+      "grad_norm": 1.1722796718213668,
+      "learning_rate": 8.61023944249673e-06,
+      "loss": 0.6915,
+      "step": 2826
+    },
+    {
+      "epoch": 1.6742671009771986,
+      "grad_norm": 1.0126608768047232,
+      "learning_rate": 8.603906042672738e-06,
+      "loss": 0.7349,
+      "step": 2827
+    },
+    {
+      "epoch": 1.6748593426117857,
+      "grad_norm": 0.8864132486291652,
+      "learning_rate": 8.597573213929677e-06,
+      "loss": 0.7473,
+      "step": 2828
+    },
+    {
+      "epoch": 1.6754515842463724,
+      "grad_norm": 0.8297726828669127,
+      "learning_rate": 8.591240958858036e-06,
+      "loss": 0.6998,
+      "step": 2829
+    },
+    {
+      "epoch": 1.6760438258809596,
+      "grad_norm": 1.0399711698669063,
+      "learning_rate": 8.584909280048064e-06,
+      "loss": 0.6883,
+      "step": 2830
+    },
+    {
+      "epoch": 1.6766360675155463,
+      "grad_norm": 1.4494905018858395,
+      "learning_rate": 8.57857818008977e-06,
+      "loss": 0.7302,
+      "step": 2831
+    },
+    {
+      "epoch": 1.6772283091501332,
+      "grad_norm": 2.2622801471005656,
+      "learning_rate": 8.572247661572926e-06,
+      "loss": 0.7195,
+      "step": 2832
+    },
+    {
+      "epoch": 1.6778205507847201,
+      "grad_norm": 0.9090049659373163,
+      "learning_rate": 8.565917727087078e-06,
+      "loss": 0.6993,
+      "step": 2833
+    },
+    {
+      "epoch": 1.678412792419307,
+      "grad_norm": 1.0718271118863054,
+      "learning_rate": 8.559588379221525e-06,
+      "loss": 0.7332,
+      "step": 2834
+    },
+    {
+      "epoch": 1.679005034053894,
+      "grad_norm": 1.3513550671215728,
+      "learning_rate": 8.55325962056532e-06,
+      "loss": 0.7081,
+      "step": 2835
+    },
+    {
+      "epoch": 1.679597275688481,
+      "grad_norm": 1.2868537736190298,
+      "learning_rate": 8.546931453707285e-06,
+      "loss": 0.7261,
+      "step": 2836
+    },
+    {
+      "epoch": 1.6801895173230679,
+      "grad_norm": 0.8991809072064495,
+      "learning_rate": 8.540603881235993e-06,
+      "loss": 0.7581,
+      "step": 2837
+    },
+    {
+      "epoch": 1.6807817589576546,
+      "grad_norm": 0.9833047762425767,
+      "learning_rate": 8.534276905739783e-06,
+      "loss": 0.7371,
+      "step": 2838
+    },
+    {
+      "epoch": 1.6813740005922417,
+      "grad_norm": 1.3090020706900172,
+      "learning_rate": 8.527950529806739e-06,
+      "loss": 0.7371,
+      "step": 2839
+    },
+    {
+      "epoch": 1.6819662422268284,
+      "grad_norm": 1.3428401139365922,
+      "learning_rate": 8.521624756024706e-06,
+      "loss": 0.7124,
+      "step": 2840
+    },
+    {
+      "epoch": 1.6825584838614156,
+      "grad_norm": 0.9822516800617229,
+      "learning_rate": 8.515299586981278e-06,
+      "loss": 0.7122,
+      "step": 2841
+    },
+    {
+      "epoch": 1.6831507254960023,
+      "grad_norm": 1.5404692861197815,
+      "learning_rate": 8.508975025263814e-06,
+      "loss": 0.7303,
+      "step": 2842
+    },
+    {
+      "epoch": 1.6837429671305892,
+      "grad_norm": 0.946622538768465,
+      "learning_rate": 8.502651073459403e-06,
+      "loss": 0.7374,
+      "step": 2843
+    },
+    {
+      "epoch": 1.6843352087651762,
+      "grad_norm": 1.1084910982075946,
+      "learning_rate": 8.496327734154905e-06,
+      "loss": 0.6946,
+      "step": 2844
+    },
+    {
+      "epoch": 1.684927450399763,
+      "grad_norm": 0.8603924254623376,
+      "learning_rate": 8.490005009936918e-06,
+      "loss": 0.7159,
+      "step": 2845
+    },
+    {
+      "epoch": 1.68551969203435,
+      "grad_norm": 1.0846672515700364,
+      "learning_rate": 8.483682903391796e-06,
+      "loss": 0.7591,
+      "step": 2846
+    },
+    {
+      "epoch": 1.686111933668937,
+      "grad_norm": 1.1750404325244188,
+      "learning_rate": 8.477361417105631e-06,
+      "loss": 0.7344,
+      "step": 2847
+    },
+    {
+      "epoch": 1.686704175303524,
+      "grad_norm": 1.0501143321182151,
+      "learning_rate": 8.471040553664269e-06,
+      "loss": 0.7063,
+      "step": 2848
+    },
+    {
+      "epoch": 1.6872964169381106,
+      "grad_norm": 1.2763094400193624,
+      "learning_rate": 8.464720315653298e-06,
+      "loss": 0.6708,
+      "step": 2849
+    },
+    {
+      "epoch": 1.6878886585726978,
+      "grad_norm": 1.111124108113765,
+      "learning_rate": 8.458400705658051e-06,
+      "loss": 0.7201,
+      "step": 2850
+    },
+    {
+      "epoch": 1.6884809002072845,
+      "grad_norm": 0.9062869224487132,
+      "learning_rate": 8.452081726263604e-06,
+      "loss": 0.7048,
+      "step": 2851
+    },
+    {
+      "epoch": 1.6890731418418716,
+      "grad_norm": 1.2248497508498524,
+      "learning_rate": 8.445763380054773e-06,
+      "loss": 0.7171,
+      "step": 2852
+    },
+    {
+      "epoch": 1.6896653834764583,
+      "grad_norm": 1.6317101564085743,
+      "learning_rate": 8.43944566961612e-06,
+      "loss": 0.6921,
+      "step": 2853
+    },
+    {
+      "epoch": 1.6902576251110453,
+      "grad_norm": 0.8594789130404841,
+      "learning_rate": 8.433128597531943e-06,
+      "loss": 0.7166,
+      "step": 2854
+    },
+    {
+      "epoch": 1.6908498667456322,
+      "grad_norm": 5.253886253022391,
+      "learning_rate": 8.426812166386278e-06,
+      "loss": 0.6991,
+      "step": 2855
+    },
+    {
+      "epoch": 1.6914421083802191,
+      "grad_norm": 1.0371352611636877,
+      "learning_rate": 8.420496378762901e-06,
+      "loss": 0.7239,
+      "step": 2856
+    },
+    {
+      "epoch": 1.692034350014806,
+      "grad_norm": 0.8825325351256947,
+      "learning_rate": 8.414181237245324e-06,
+      "loss": 0.7318,
+      "step": 2857
+    },
+    {
+      "epoch": 1.692626591649393,
+      "grad_norm": 1.0387049681798395,
+      "learning_rate": 8.407866744416801e-06,
+      "loss": 0.7132,
+      "step": 2858
+    },
+    {
+      "epoch": 1.69321883328398,
+      "grad_norm": 1.0193831157896063,
+      "learning_rate": 8.401552902860306e-06,
+      "loss": 0.7193,
+      "step": 2859
+    },
+    {
+      "epoch": 1.6938110749185666,
+      "grad_norm": 1.2765032613192198,
+      "learning_rate": 8.395239715158558e-06,
+      "loss": 0.7425,
+      "step": 2860
+    },
+    {
+      "epoch": 1.6944033165531538,
+      "grad_norm": 1.0423155817988132,
+      "learning_rate": 8.388927183894005e-06,
+      "loss": 0.7557,
+      "step": 2861
+    },
+    {
+      "epoch": 1.6949955581877405,
+      "grad_norm": 1.2949288338272016,
+      "learning_rate": 8.382615311648833e-06,
+      "loss": 0.7196,
+      "step": 2862
+    },
+    {
+      "epoch": 1.6955877998223277,
+      "grad_norm": 1.203764230621225,
+      "learning_rate": 8.376304101004946e-06,
+      "loss": 0.7589,
+      "step": 2863
+    },
+    {
+      "epoch": 1.6961800414569144,
+      "grad_norm": 1.1439950575179505,
+      "learning_rate": 8.369993554543987e-06,
+      "loss": 0.7518,
+      "step": 2864
+    },
+    {
+      "epoch": 1.6967722830915013,
+      "grad_norm": 1.768457656980784,
+      "learning_rate": 8.363683674847323e-06,
+      "loss": 0.7382,
+      "step": 2865
+    },
+    {
+      "epoch": 1.6973645247260882,
+      "grad_norm": 1.0904038835740486,
+      "learning_rate": 8.357374464496056e-06,
+      "loss": 0.7482,
+      "step": 2866
+    },
+    {
+      "epoch": 1.6979567663606752,
+      "grad_norm": 1.4275207295309171,
+      "learning_rate": 8.351065926070994e-06,
+      "loss": 0.7192,
+      "step": 2867
+    },
+    {
+      "epoch": 1.698549007995262,
+      "grad_norm": 0.8975759750455716,
+      "learning_rate": 8.344758062152696e-06,
+      "loss": 0.7208,
+      "step": 2868
+    },
+    {
+      "epoch": 1.699141249629849,
+      "grad_norm": 0.9962272566707975,
+      "learning_rate": 8.338450875321428e-06,
+      "loss": 0.7174,
+      "step": 2869
+    },
+    {
+      "epoch": 1.699733491264436,
+      "grad_norm": 0.9458627764840206,
+      "learning_rate": 8.332144368157192e-06,
+      "loss": 0.7004,
+      "step": 2870
+    },
+    {
+      "epoch": 1.7003257328990227,
+      "grad_norm": 0.9243786772003465,
+      "learning_rate": 8.325838543239688e-06,
+      "loss": 0.7212,
+      "step": 2871
+    },
+    {
+      "epoch": 1.7009179745336098,
+      "grad_norm": 1.095152314018298,
+      "learning_rate": 8.319533403148368e-06,
+      "loss": 0.7413,
+      "step": 2872
+    },
+    {
+      "epoch": 1.7015102161681965,
+      "grad_norm": 1.179387764922565,
+      "learning_rate": 8.31322895046238e-06,
+      "loss": 0.7317,
+      "step": 2873
+    },
+    {
+      "epoch": 1.7021024578027837,
+      "grad_norm": 1.0825336416478861,
+      "learning_rate": 8.306925187760608e-06,
+      "loss": 0.7383,
+      "step": 2874
+    },
+    {
+      "epoch": 1.7026946994373704,
+      "grad_norm": 1.1435368543161897,
+      "learning_rate": 8.300622117621634e-06,
+      "loss": 0.7225,
+      "step": 2875
+    },
+    {
+      "epoch": 1.7032869410719573,
+      "grad_norm": 1.0508503526183521,
+      "learning_rate": 8.29431974262378e-06,
+      "loss": 0.7231,
+      "step": 2876
+    },
+    {
+      "epoch": 1.7038791827065443,
+      "grad_norm": 2.6026780734301576,
+      "learning_rate": 8.288018065345063e-06,
+      "loss": 0.7578,
+      "step": 2877
+    },
+    {
+      "epoch": 1.7044714243411312,
+      "grad_norm": 1.036737957257567,
+      "learning_rate": 8.28171708836323e-06,
+      "loss": 0.727,
+      "step": 2878
+    },
+    {
+      "epoch": 1.7050636659757181,
+      "grad_norm": 1.2395033785247513,
+      "learning_rate": 8.275416814255731e-06,
+      "loss": 0.7323,
+      "step": 2879
+    },
+    {
+      "epoch": 1.705655907610305,
+      "grad_norm": 5.3070838585377915,
+      "learning_rate": 8.269117245599729e-06,
+      "loss": 0.7276,
+      "step": 2880
+    },
+    {
+      "epoch": 1.706248149244892,
+      "grad_norm": 1.9790319279693616,
+      "learning_rate": 8.262818384972108e-06,
+      "loss": 0.7232,
+      "step": 2881
+    },
+    {
+      "epoch": 1.7068403908794787,
+      "grad_norm": 0.9027953973682145,
+      "learning_rate": 8.256520234949456e-06,
+      "loss": 0.7473,
+      "step": 2882
+    },
+    {
+      "epoch": 1.7074326325140659,
+      "grad_norm": 1.0026460097696173,
+      "learning_rate": 8.250222798108068e-06,
+      "loss": 0.6998,
+      "step": 2883
+    },
+    {
+      "epoch": 1.7080248741486526,
+      "grad_norm": 1.4657670953683113,
+      "learning_rate": 8.243926077023945e-06,
+      "loss": 0.663,
+      "step": 2884
+    },
+    {
+      "epoch": 1.7086171157832397,
+      "grad_norm": 1.0969379355734514,
+      "learning_rate": 8.23763007427281e-06,
+      "loss": 0.7078,
+      "step": 2885
+    },
+    {
+      "epoch": 1.7092093574178264,
+      "grad_norm": 0.967745297644781,
+      "learning_rate": 8.23133479243008e-06,
+      "loss": 0.7153,
+      "step": 2886
+    },
+    {
+      "epoch": 1.7098015990524134,
+      "grad_norm": 1.2060481495888675,
+      "learning_rate": 8.225040234070873e-06,
+      "loss": 0.7624,
+      "step": 2887
+    },
+    {
+      "epoch": 1.7103938406870003,
+      "grad_norm": 2.120123919682014,
+      "learning_rate": 8.218746401770021e-06,
+      "loss": 0.7574,
+      "step": 2888
+    },
+    {
+      "epoch": 1.7109860823215872,
+      "grad_norm": 1.1153560787494186,
+      "learning_rate": 8.212453298102054e-06,
+      "loss": 0.7589,
+      "step": 2889
+    },
+    {
+      "epoch": 1.7115783239561742,
+      "grad_norm": 1.5107438969118132,
+      "learning_rate": 8.206160925641211e-06,
+      "loss": 0.7536,
+      "step": 2890
+    },
+    {
+      "epoch": 1.712170565590761,
+      "grad_norm": 1.054181917285995,
+      "learning_rate": 8.19986928696142e-06,
+      "loss": 0.7439,
+      "step": 2891
+    },
+    {
+      "epoch": 1.712762807225348,
+      "grad_norm": 1.530532649227264,
+      "learning_rate": 8.193578384636317e-06,
+      "loss": 0.7304,
+      "step": 2892
+    },
+    {
+      "epoch": 1.7133550488599347,
+      "grad_norm": 1.4100039607836896,
+      "learning_rate": 8.187288221239232e-06,
+      "loss": 0.7115,
+      "step": 2893
+    },
+    {
+      "epoch": 1.7139472904945219,
+      "grad_norm": 1.3653865280730575,
+      "learning_rate": 8.180998799343203e-06,
+      "loss": 0.6963,
+      "step": 2894
+    },
+    {
+      "epoch": 1.7145395321291086,
+      "grad_norm": 1.3761845598375588,
+      "learning_rate": 8.17471012152095e-06,
+      "loss": 0.7766,
+      "step": 2895
+    },
+    {
+      "epoch": 1.7151317737636957,
+      "grad_norm": 0.9896735131841189,
+      "learning_rate": 8.168422190344896e-06,
+      "loss": 0.7104,
+      "step": 2896
+    },
+    {
+      "epoch": 1.7157240153982825,
+      "grad_norm": 1.8180792395206247,
+      "learning_rate": 8.162135008387164e-06,
+      "loss": 0.7355,
+      "step": 2897
+    },
+    {
+      "epoch": 1.7163162570328694,
+      "grad_norm": 1.2917397756929976,
+      "learning_rate": 8.155848578219563e-06,
+      "loss": 0.7736,
+      "step": 2898
+    },
+    {
+      "epoch": 1.7169084986674563,
+      "grad_norm": 1.2049090200411698,
+      "learning_rate": 8.14956290241359e-06,
+      "loss": 0.7332,
+      "step": 2899
+    },
+    {
+      "epoch": 1.7175007403020432,
+      "grad_norm": 2.413612490345052,
+      "learning_rate": 8.14327798354045e-06,
+      "loss": 0.7142,
+      "step": 2900
+    },
+    {
+      "epoch": 1.7180929819366302,
+      "grad_norm": 1.0571402249668937,
+      "learning_rate": 8.136993824171019e-06,
+      "loss": 0.7277,
+      "step": 2901
+    },
+    {
+      "epoch": 1.718685223571217,
+      "grad_norm": 1.2201789441380906,
+      "learning_rate": 8.130710426875881e-06,
+      "loss": 0.7176,
+      "step": 2902
+    },
+    {
+      "epoch": 1.719277465205804,
+      "grad_norm": 1.041688994211538,
+      "learning_rate": 8.12442779422529e-06,
+      "loss": 0.6969,
+      "step": 2903
+    },
+    {
+      "epoch": 1.7198697068403908,
+      "grad_norm": 1.1679307028654387,
+      "learning_rate": 8.118145928789198e-06,
+      "loss": 0.6998,
+      "step": 2904
+    },
+    {
+      "epoch": 1.720461948474978,
+      "grad_norm": 2.7733038021139875,
+      "learning_rate": 8.111864833137246e-06,
+      "loss": 0.6915,
+      "step": 2905
+    },
+    {
+      "epoch": 1.7210541901095646,
+      "grad_norm": 1.5759312229276083,
+      "learning_rate": 8.105584509838754e-06,
+      "loss": 0.7227,
+      "step": 2906
+    },
+    {
+      "epoch": 1.7216464317441518,
+      "grad_norm": 1.3336882101988723,
+      "learning_rate": 8.099304961462722e-06,
+      "loss": 0.741,
+      "step": 2907
+    },
+    {
+      "epoch": 1.7222386733787385,
+      "grad_norm": 1.2123410041999985,
+      "learning_rate": 8.093026190577839e-06,
+      "loss": 0.7105,
+      "step": 2908
+    },
+    {
+      "epoch": 1.7228309150133254,
+      "grad_norm": 1.0373687336456456,
+      "learning_rate": 8.086748199752483e-06,
+      "loss": 0.6718,
+      "step": 2909
+    },
+    {
+      "epoch": 1.7234231566479123,
+      "grad_norm": 2.863392250283223,
+      "learning_rate": 8.080470991554703e-06,
+      "loss": 0.7303,
+      "step": 2910
+    },
+    {
+      "epoch": 1.7240153982824993,
+      "grad_norm": 29.798561463464726,
+      "learning_rate": 8.074194568552224e-06,
+      "loss": 0.7341,
+      "step": 2911
+    },
+    {
+      "epoch": 1.7246076399170862,
+      "grad_norm": 1.3629065680833,
+      "learning_rate": 8.067918933312459e-06,
+      "loss": 0.7307,
+      "step": 2912
+    },
+    {
+      "epoch": 1.725199881551673,
+      "grad_norm": 1.0966158414516722,
+      "learning_rate": 8.061644088402499e-06,
+      "loss": 0.7015,
+      "step": 2913
+    },
+    {
+      "epoch": 1.72579212318626,
+      "grad_norm": 1.270462849175571,
+      "learning_rate": 8.055370036389105e-06,
+      "loss": 0.7317,
+      "step": 2914
+    },
+    {
+      "epoch": 1.7263843648208468,
+      "grad_norm": 1.1763045807935812,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 0.7091,
+      "step": 2915
+    },
+    {
+      "epoch": 1.726976606455434,
+      "grad_norm": 3.4592157510743173,
+      "learning_rate": 8.042824321317453e-06,
+      "loss": 0.7266,
+      "step": 2916
+    },
+    {
+      "epoch": 1.7275688480900206,
+      "grad_norm": 1.1717886947291323,
+      "learning_rate": 8.036552663391099e-06,
+      "loss": 0.7113,
+      "step": 2917
+    },
+    {
+      "epoch": 1.7281610897246078,
+      "grad_norm": 4.1475058496080965,
+      "learning_rate": 8.030281808625114e-06,
+      "loss": 0.7183,
+      "step": 2918
+    },
+    {
+      "epoch": 1.7287533313591945,
+      "grad_norm": 1.5657993451062597,
+      "learning_rate": 8.02401175958463e-06,
+      "loss": 0.7218,
+      "step": 2919
+    },
+    {
+      "epoch": 1.7293455729937814,
+      "grad_norm": 1.070325978428891,
+      "learning_rate": 8.017742518834454e-06,
+      "loss": 0.723,
+      "step": 2920
+    },
+    {
+      "epoch": 1.7299378146283684,
+      "grad_norm": 1.1119220102928689,
+      "learning_rate": 8.011474088939056e-06,
+      "loss": 0.6992,
+      "step": 2921
+    },
+    {
+      "epoch": 1.7305300562629553,
+      "grad_norm": 1.3076317154786299,
+      "learning_rate": 8.005206472462576e-06,
+      "loss": 0.7237,
+      "step": 2922
+    },
+    {
+      "epoch": 1.7311222978975422,
+      "grad_norm": 1.2441146503285951,
+      "learning_rate": 7.998939671968817e-06,
+      "loss": 0.7263,
+      "step": 2923
+    },
+    {
+      "epoch": 1.731714539532129,
+      "grad_norm": 1.3530915910623935,
+      "learning_rate": 7.99267369002126e-06,
+      "loss": 0.7264,
+      "step": 2924
+    },
+    {
+      "epoch": 1.732306781166716,
+      "grad_norm": 1.1749639635147615,
+      "learning_rate": 7.986408529183045e-06,
+      "loss": 0.7354,
+      "step": 2925
+    },
+    {
+      "epoch": 1.7328990228013028,
+      "grad_norm": 0.9443164977742067,
+      "learning_rate": 7.980144192016967e-06,
+      "loss": 0.7564,
+      "step": 2926
+    },
+    {
+      "epoch": 1.73349126443589,
+      "grad_norm": 1.2969219620095116,
+      "learning_rate": 7.973880681085495e-06,
+      "loss": 0.7822,
+      "step": 2927
+    },
+    {
+      "epoch": 1.7340835060704767,
+      "grad_norm": 1.3340663571052536,
+      "learning_rate": 7.967617998950762e-06,
+      "loss": 0.7495,
+      "step": 2928
+    },
+    {
+      "epoch": 1.7346757477050638,
+      "grad_norm": 1.2273379932145483,
+      "learning_rate": 7.961356148174554e-06,
+      "loss": 0.7072,
+      "step": 2929
+    },
+    {
+      "epoch": 1.7352679893396505,
+      "grad_norm": 1.3141531019330541,
+      "learning_rate": 7.955095131318319e-06,
+      "loss": 0.7258,
+      "step": 2930
+    },
+    {
+      "epoch": 1.7358602309742375,
+      "grad_norm": 1.203363496166313,
+      "learning_rate": 7.948834950943165e-06,
+      "loss": 0.721,
+      "step": 2931
+    },
+    {
+      "epoch": 1.7364524726088244,
+      "grad_norm": 1.1742529128764918,
+      "learning_rate": 7.942575609609857e-06,
+      "loss": 0.7185,
+      "step": 2932
+    },
+    {
+      "epoch": 1.7370447142434113,
+      "grad_norm": 1.428016405572957,
+      "learning_rate": 7.936317109878824e-06,
+      "loss": 0.7647,
+      "step": 2933
+    },
+    {
+      "epoch": 1.7376369558779983,
+      "grad_norm": 1.2790352849117015,
+      "learning_rate": 7.930059454310138e-06,
+      "loss": 0.7581,
+      "step": 2934
+    },
+    {
+      "epoch": 1.738229197512585,
+      "grad_norm": 0.9761646608961596,
+      "learning_rate": 7.923802645463532e-06,
+      "loss": 0.7217,
+      "step": 2935
+    },
+    {
+      "epoch": 1.7388214391471721,
+      "grad_norm": 1.5751448044306215,
+      "learning_rate": 7.917546685898393e-06,
+      "loss": 0.6937,
+      "step": 2936
+    },
+    {
+      "epoch": 1.7394136807817588,
+      "grad_norm": 1.1640978244444167,
+      "learning_rate": 7.911291578173767e-06,
+      "loss": 0.6968,
+      "step": 2937
+    },
+    {
+      "epoch": 1.740005922416346,
+      "grad_norm": 1.4304427410478942,
+      "learning_rate": 7.905037324848334e-06,
+      "loss": 0.7239,
+      "step": 2938
+    },
+    {
+      "epoch": 1.7405981640509327,
+      "grad_norm": 0.9786975945632999,
+      "learning_rate": 7.898783928480442e-06,
+      "loss": 0.744,
+      "step": 2939
+    },
+    {
+      "epoch": 1.7411904056855196,
+      "grad_norm": 1.206680722391421,
+      "learning_rate": 7.89253139162808e-06,
+      "loss": 0.732,
+      "step": 2940
+    },
+    {
+      "epoch": 1.7417826473201066,
+      "grad_norm": 1.9134032463262014,
+      "learning_rate": 7.88627971684889e-06,
+      "loss": 0.7539,
+      "step": 2941
+    },
+    {
+      "epoch": 1.7423748889546935,
+      "grad_norm": 1.7876317020070605,
+      "learning_rate": 7.880028906700153e-06,
+      "loss": 0.7261,
+      "step": 2942
+    },
+    {
+      "epoch": 1.7429671305892804,
+      "grad_norm": 1.2450546265182965,
+      "learning_rate": 7.873778963738806e-06,
+      "loss": 0.692,
+      "step": 2943
+    },
+    {
+      "epoch": 1.7435593722238674,
+      "grad_norm": 1.273671537217201,
+      "learning_rate": 7.867529890521424e-06,
+      "loss": 0.6972,
+      "step": 2944
+    },
+    {
+      "epoch": 1.7441516138584543,
+      "grad_norm": 0.9478991818498798,
+      "learning_rate": 7.861281689604237e-06,
+      "loss": 0.7267,
+      "step": 2945
+    },
+    {
+      "epoch": 1.744743855493041,
+      "grad_norm": 0.9418695622023932,
+      "learning_rate": 7.8550343635431e-06,
+      "loss": 0.7253,
+      "step": 2946
+    },
+    {
+      "epoch": 1.7453360971276282,
+      "grad_norm": 1.5972561604224107,
+      "learning_rate": 7.848787914893525e-06,
+      "loss": 0.719,
+      "step": 2947
+    },
+    {
+      "epoch": 1.7459283387622149,
+      "grad_norm": 1.6316651832184046,
+      "learning_rate": 7.842542346210663e-06,
+      "loss": 0.6981,
+      "step": 2948
+    },
+    {
+      "epoch": 1.746520580396802,
+      "grad_norm": 1.1436403440246294,
+      "learning_rate": 7.836297660049303e-06,
+      "loss": 0.7392,
+      "step": 2949
+    },
+    {
+      "epoch": 1.7471128220313887,
+      "grad_norm": 1.1760242256445133,
+      "learning_rate": 7.83005385896387e-06,
+      "loss": 0.7368,
+      "step": 2950
+    },
+    {
+      "epoch": 1.7477050636659757,
+      "grad_norm": 1.0420980744058013,
+      "learning_rate": 7.823810945508427e-06,
+      "loss": 0.7202,
+      "step": 2951
+    },
+    {
+      "epoch": 1.7482973053005626,
+      "grad_norm": 1.236600456703946,
+      "learning_rate": 7.817568922236683e-06,
+      "loss": 0.7232,
+      "step": 2952
+    },
+    {
+      "epoch": 1.7488895469351495,
+      "grad_norm": 1.1503525583737613,
+      "learning_rate": 7.811327791701977e-06,
+      "loss": 0.7134,
+      "step": 2953
+    },
+    {
+      "epoch": 1.7494817885697365,
+      "grad_norm": 2.7772710614120597,
+      "learning_rate": 7.805087556457275e-06,
+      "loss": 0.6838,
+      "step": 2954
+    },
+    {
+      "epoch": 1.7500740302043234,
+      "grad_norm": 1.5536873224508294,
+      "learning_rate": 7.798848219055189e-06,
+      "loss": 0.7342,
+      "step": 2955
+    },
+    {
+      "epoch": 1.7506662718389103,
+      "grad_norm": 1.3445549621095612,
+      "learning_rate": 7.792609782047958e-06,
+      "loss": 0.7372,
+      "step": 2956
+    },
+    {
+      "epoch": 1.751258513473497,
+      "grad_norm": 1.0834091497432776,
+      "learning_rate": 7.786372247987454e-06,
+      "loss": 0.6987,
+      "step": 2957
+    },
+    {
+      "epoch": 1.7518507551080842,
+      "grad_norm": 3.789898070049763,
+      "learning_rate": 7.78013561942518e-06,
+      "loss": 0.7085,
+      "step": 2958
+    },
+    {
+      "epoch": 1.752442996742671,
+      "grad_norm": 1.2154346445934634,
+      "learning_rate": 7.773899898912266e-06,
+      "loss": 0.7192,
+      "step": 2959
+    },
+    {
+      "epoch": 1.753035238377258,
+      "grad_norm": 1.3478249479015387,
+      "learning_rate": 7.76766508899947e-06,
+      "loss": 0.7343,
+      "step": 2960
+    },
+    {
+      "epoch": 1.7536274800118448,
+      "grad_norm": 1.4976847760098018,
+      "learning_rate": 7.761431192237192e-06,
+      "loss": 0.7646,
+      "step": 2961
+    },
+    {
+      "epoch": 1.7542197216464317,
+      "grad_norm": 1.81752309263868,
+      "learning_rate": 7.755198211175428e-06,
+      "loss": 0.7278,
+      "step": 2962
+    },
+    {
+      "epoch": 1.7548119632810186,
+      "grad_norm": 1.1597990895135137,
+      "learning_rate": 7.74896614836383e-06,
+      "loss": 0.7343,
+      "step": 2963
+    },
+    {
+      "epoch": 1.7554042049156056,
+      "grad_norm": 1.5225622446141434,
+      "learning_rate": 7.742735006351656e-06,
+      "loss": 0.7068,
+      "step": 2964
+    },
+    {
+      "epoch": 1.7559964465501925,
+      "grad_norm": 1.1131198166318628,
+      "learning_rate": 7.736504787687804e-06,
+      "loss": 0.6979,
+      "step": 2965
+    },
+    {
+      "epoch": 1.7565886881847794,
+      "grad_norm": 1.236323876399398,
+      "learning_rate": 7.73027549492077e-06,
+      "loss": 0.7224,
+      "step": 2966
+    },
+    {
+      "epoch": 1.7571809298193664,
+      "grad_norm": 1.2400368319010493,
+      "learning_rate": 7.724047130598692e-06,
+      "loss": 0.6906,
+      "step": 2967
+    },
+    {
+      "epoch": 1.757773171453953,
+      "grad_norm": 1.28829954217693,
+      "learning_rate": 7.717819697269322e-06,
+      "loss": 0.6975,
+      "step": 2968
+    },
+    {
+      "epoch": 1.7583654130885402,
+      "grad_norm": 0.9178925212400618,
+      "learning_rate": 7.711593197480031e-06,
+      "loss": 0.7548,
+      "step": 2969
+    },
+    {
+      "epoch": 1.758957654723127,
+      "grad_norm": 1.3227709395440883,
+      "learning_rate": 7.7053676337778e-06,
+      "loss": 0.7114,
+      "step": 2970
+    },
+    {
+      "epoch": 1.759549896357714,
+      "grad_norm": 0.9702755831253254,
+      "learning_rate": 7.699143008709245e-06,
+      "loss": 0.7199,
+      "step": 2971
+    },
+    {
+      "epoch": 1.7601421379923008,
+      "grad_norm": 0.9961523350834506,
+      "learning_rate": 7.69291932482058e-06,
+      "loss": 0.7949,
+      "step": 2972
+    },
+    {
+      "epoch": 1.7607343796268877,
+      "grad_norm": 1.2207456078535526,
+      "learning_rate": 7.686696584657649e-06,
+      "loss": 0.743,
+      "step": 2973
+    },
+    {
+      "epoch": 1.7613266212614747,
+      "grad_norm": 1.3380369368359777,
+      "learning_rate": 7.680474790765895e-06,
+      "loss": 0.6948,
+      "step": 2974
+    },
+    {
+      "epoch": 1.7619188628960616,
+      "grad_norm": 0.8789763724986712,
+      "learning_rate": 7.674253945690383e-06,
+      "loss": 0.6813,
+      "step": 2975
+    },
+    {
+      "epoch": 1.7625111045306485,
+      "grad_norm": 1.0544272650985724,
+      "learning_rate": 7.668034051975793e-06,
+      "loss": 0.6844,
+      "step": 2976
+    },
+    {
+      "epoch": 1.7631033461652355,
+      "grad_norm": 1.257614243686341,
+      "learning_rate": 7.661815112166408e-06,
+      "loss": 0.7071,
+      "step": 2977
+    },
+    {
+      "epoch": 1.7636955877998224,
+      "grad_norm": 1.0368071883700545,
+      "learning_rate": 7.655597128806125e-06,
+      "loss": 0.7018,
+      "step": 2978
+    },
+    {
+      "epoch": 1.764287829434409,
+      "grad_norm": 1.5446506502619204,
+      "learning_rate": 7.649380104438446e-06,
+      "loss": 0.7544,
+      "step": 2979
+    },
+    {
+      "epoch": 1.7648800710689962,
+      "grad_norm": 0.990885528622176,
+      "learning_rate": 7.643164041606489e-06,
+      "loss": 0.7179,
+      "step": 2980
+    },
+    {
+      "epoch": 1.765472312703583,
+      "grad_norm": 0.9953972053282248,
+      "learning_rate": 7.63694894285297e-06,
+      "loss": 0.7065,
+      "step": 2981
+    },
+    {
+      "epoch": 1.7660645543381701,
+      "grad_norm": 6.360092353723965,
+      "learning_rate": 7.630734810720212e-06,
+      "loss": 0.708,
+      "step": 2982
+    },
+    {
+      "epoch": 1.7666567959727568,
+      "grad_norm": 2.422095125571542,
+      "learning_rate": 7.624521647750149e-06,
+      "loss": 0.7443,
+      "step": 2983
+    },
+    {
+      "epoch": 1.7672490376073438,
+      "grad_norm": 1.0870002901322164,
+      "learning_rate": 7.618309456484309e-06,
+      "loss": 0.7297,
+      "step": 2984
+    },
+    {
+      "epoch": 1.7678412792419307,
+      "grad_norm": 2.3370356729061803,
+      "learning_rate": 7.612098239463833e-06,
+      "loss": 0.718,
+      "step": 2985
+    },
+    {
+      "epoch": 1.7684335208765176,
+      "grad_norm": 1.2080348447951503,
+      "learning_rate": 7.605887999229454e-06,
+      "loss": 0.72,
+      "step": 2986
+    },
+    {
+      "epoch": 1.7690257625111045,
+      "grad_norm": 1.0232344898785648,
+      "learning_rate": 7.599678738321512e-06,
+      "loss": 0.7236,
+      "step": 2987
+    },
+    {
+      "epoch": 1.7696180041456915,
+      "grad_norm": 1.347583694196258,
+      "learning_rate": 7.593470459279939e-06,
+      "loss": 0.7651,
+      "step": 2988
+    },
+    {
+      "epoch": 1.7702102457802784,
+      "grad_norm": 1.4355786367789864,
+      "learning_rate": 7.58726316464428e-06,
+      "loss": 0.7125,
+      "step": 2989
+    },
+    {
+      "epoch": 1.7708024874148651,
+      "grad_norm": 1.0074909388188276,
+      "learning_rate": 7.581056856953656e-06,
+      "loss": 0.7373,
+      "step": 2990
+    },
+    {
+      "epoch": 1.7713947290494523,
+      "grad_norm": 1.06347341429192,
+      "learning_rate": 7.574851538746802e-06,
+      "loss": 0.7273,
+      "step": 2991
+    },
+    {
+      "epoch": 1.771986970684039,
+      "grad_norm": 1.4128916933234186,
+      "learning_rate": 7.568647212562043e-06,
+      "loss": 0.7169,
+      "step": 2992
+    },
+    {
+      "epoch": 1.7725792123186261,
+      "grad_norm": 0.9450378936333973,
+      "learning_rate": 7.562443880937297e-06,
+      "loss": 0.7388,
+      "step": 2993
+    },
+    {
+      "epoch": 1.7731714539532129,
+      "grad_norm": 0.9163136638470174,
+      "learning_rate": 7.55624154641007e-06,
+      "loss": 0.7219,
+      "step": 2994
+    },
+    {
+      "epoch": 1.7737636955877998,
+      "grad_norm": 1.324578116509981,
+      "learning_rate": 7.550040211517472e-06,
+      "loss": 0.7602,
+      "step": 2995
+    },
+    {
+      "epoch": 1.7743559372223867,
+      "grad_norm": 1.1787072674151278,
+      "learning_rate": 7.543839878796195e-06,
+      "loss": 0.7183,
+      "step": 2996
+    },
+    {
+      "epoch": 1.7749481788569736,
+      "grad_norm": 3.8914286195953993,
+      "learning_rate": 7.537640550782527e-06,
+      "loss": 0.7204,
+      "step": 2997
+    },
+    {
+      "epoch": 1.7755404204915606,
+      "grad_norm": 1.0219901640955456,
+      "learning_rate": 7.531442230012336e-06,
+      "loss": 0.7438,
+      "step": 2998
+    },
+    {
+      "epoch": 1.7761326621261475,
+      "grad_norm": 1.2855857922477254,
+      "learning_rate": 7.525244919021084e-06,
+      "loss": 0.738,
+      "step": 2999
+    },
+    {
+      "epoch": 1.7767249037607344,
+      "grad_norm": 2.344182476915904,
+      "learning_rate": 7.519048620343825e-06,
+      "loss": 0.7248,
+      "step": 3000
+    },
+    {
+      "epoch": 1.7773171453953212,
+      "grad_norm": 1.1756563315618664,
+      "learning_rate": 7.512853336515193e-06,
+      "loss": 0.7267,
+      "step": 3001
+    },
+    {
+      "epoch": 1.7779093870299083,
+      "grad_norm": 0.9661265454991339,
+      "learning_rate": 7.506659070069404e-06,
+      "loss": 0.6421,
+      "step": 3002
+    },
+    {
+      "epoch": 1.778501628664495,
+      "grad_norm": 2.2188327107977934,
+      "learning_rate": 7.5004658235402594e-06,
+      "loss": 0.7092,
+      "step": 3003
+    },
+    {
+      "epoch": 1.7790938702990822,
+      "grad_norm": 2.792499182936962,
+      "learning_rate": 7.494273599461153e-06,
+      "loss": 0.7787,
+      "step": 3004
+    },
+    {
+      "epoch": 1.7796861119336689,
+      "grad_norm": 1.4377298749803848,
+      "learning_rate": 7.4880824003650475e-06,
+      "loss": 0.7323,
+      "step": 3005
+    },
+    {
+      "epoch": 1.7802783535682558,
+      "grad_norm": 1.1084690396732224,
+      "learning_rate": 7.481892228784491e-06,
+      "loss": 0.728,
+      "step": 3006
+    },
+    {
+      "epoch": 1.7808705952028427,
+      "grad_norm": 1.375063278971446,
+      "learning_rate": 7.475703087251611e-06,
+      "loss": 0.7102,
+      "step": 3007
+    },
+    {
+      "epoch": 1.7814628368374297,
+      "grad_norm": 1.251032911053362,
+      "learning_rate": 7.469514978298119e-06,
+      "loss": 0.7353,
+      "step": 3008
+    },
+    {
+      "epoch": 1.7820550784720166,
+      "grad_norm": 0.8237693730719295,
+      "learning_rate": 7.463327904455299e-06,
+      "loss": 0.6709,
+      "step": 3009
+    },
+    {
+      "epoch": 1.7826473201066035,
+      "grad_norm": 1.3285542970533006,
+      "learning_rate": 7.457141868254007e-06,
+      "loss": 0.7046,
+      "step": 3010
+    },
+    {
+      "epoch": 1.7832395617411905,
+      "grad_norm": 2.4110688518280003,
+      "learning_rate": 7.450956872224684e-06,
+      "loss": 0.6963,
+      "step": 3011
+    },
+    {
+      "epoch": 1.7838318033757772,
+      "grad_norm": 1.0311760681825095,
+      "learning_rate": 7.444772918897336e-06,
+      "loss": 0.7458,
+      "step": 3012
+    },
+    {
+      "epoch": 1.7844240450103643,
+      "grad_norm": 3.042245299524146,
+      "learning_rate": 7.438590010801558e-06,
+      "loss": 0.7007,
+      "step": 3013
+    },
+    {
+      "epoch": 1.785016286644951,
+      "grad_norm": 1.1638757581822774,
+      "learning_rate": 7.432408150466497e-06,
+      "loss": 0.727,
+      "step": 3014
+    },
+    {
+      "epoch": 1.7856085282795382,
+      "grad_norm": 1.8489857858233996,
+      "learning_rate": 7.426227340420886e-06,
+      "loss": 0.7187,
+      "step": 3015
+    },
+    {
+      "epoch": 1.786200769914125,
+      "grad_norm": 1.8469830538002985,
+      "learning_rate": 7.42004758319302e-06,
+      "loss": 0.7172,
+      "step": 3016
+    },
+    {
+      "epoch": 1.7867930115487118,
+      "grad_norm": 1.411130583258253,
+      "learning_rate": 7.413868881310778e-06,
+      "loss": 0.7044,
+      "step": 3017
+    },
+    {
+      "epoch": 1.7873852531832988,
+      "grad_norm": 1.2920931978075625,
+      "learning_rate": 7.40769123730158e-06,
+      "loss": 0.7222,
+      "step": 3018
+    },
+    {
+      "epoch": 1.7879774948178857,
+      "grad_norm": 1.3026695316857428,
+      "learning_rate": 7.401514653692442e-06,
+      "loss": 0.7283,
+      "step": 3019
+    },
+    {
+      "epoch": 1.7885697364524726,
+      "grad_norm": 1.002137271655342,
+      "learning_rate": 7.395339133009931e-06,
+      "loss": 0.7016,
+      "step": 3020
+    },
+    {
+      "epoch": 1.7891619780870596,
+      "grad_norm": 1.755038606150744,
+      "learning_rate": 7.3891646777801826e-06,
+      "loss": 0.7114,
+      "step": 3021
+    },
+    {
+      "epoch": 1.7897542197216465,
+      "grad_norm": 1.0069551996461956,
+      "learning_rate": 7.382991290528892e-06,
+      "loss": 0.7261,
+      "step": 3022
+    },
+    {
+      "epoch": 1.7903464613562332,
+      "grad_norm": 1.144265082942235,
+      "learning_rate": 7.376818973781328e-06,
+      "loss": 0.6922,
+      "step": 3023
+    },
+    {
+      "epoch": 1.7909387029908204,
+      "grad_norm": 0.9697054263146545,
+      "learning_rate": 7.370647730062311e-06,
+      "loss": 0.7872,
+      "step": 3024
+    },
+    {
+      "epoch": 1.791530944625407,
+      "grad_norm": 1.0689276075884193,
+      "learning_rate": 7.364477561896231e-06,
+      "loss": 0.6867,
+      "step": 3025
+    },
+    {
+      "epoch": 1.7921231862599942,
+      "grad_norm": 0.9895555328791956,
+      "learning_rate": 7.358308471807028e-06,
+      "loss": 0.7047,
+      "step": 3026
+    },
+    {
+      "epoch": 1.792715427894581,
+      "grad_norm": 1.426959406213989,
+      "learning_rate": 7.3521404623182065e-06,
+      "loss": 0.6768,
+      "step": 3027
+    },
+    {
+      "epoch": 1.7933076695291679,
+      "grad_norm": 1.6038922032905296,
+      "learning_rate": 7.3459735359528366e-06,
+      "loss": 0.7169,
+      "step": 3028
+    },
+    {
+      "epoch": 1.7938999111637548,
+      "grad_norm": 1.2100286028193288,
+      "learning_rate": 7.339807695233534e-06,
+      "loss": 0.732,
+      "step": 3029
+    },
+    {
+      "epoch": 1.7944921527983417,
+      "grad_norm": 2.091644826171607,
+      "learning_rate": 7.333642942682473e-06,
+      "loss": 0.7597,
+      "step": 3030
+    },
+    {
+      "epoch": 1.7950843944329287,
+      "grad_norm": 0.9734538843147006,
+      "learning_rate": 7.327479280821381e-06,
+      "loss": 0.6885,
+      "step": 3031
+    },
+    {
+      "epoch": 1.7956766360675156,
+      "grad_norm": 0.9935775103575984,
+      "learning_rate": 7.3213167121715514e-06,
+      "loss": 0.7143,
+      "step": 3032
+    },
+    {
+      "epoch": 1.7962688777021025,
+      "grad_norm": 1.2604409769783473,
+      "learning_rate": 7.315155239253815e-06,
+      "loss": 0.7302,
+      "step": 3033
+    },
+    {
+      "epoch": 1.7968611193366892,
+      "grad_norm": 1.1543012632612109,
+      "learning_rate": 7.308994864588562e-06,
+      "loss": 0.7388,
+      "step": 3034
+    },
+    {
+      "epoch": 1.7974533609712764,
+      "grad_norm": 1.242912338801917,
+      "learning_rate": 7.302835590695731e-06,
+      "loss": 0.732,
+      "step": 3035
+    },
+    {
+      "epoch": 1.798045602605863,
+      "grad_norm": 0.9047047765521495,
+      "learning_rate": 7.296677420094811e-06,
+      "loss": 0.6763,
+      "step": 3036
+    },
+    {
+      "epoch": 1.7986378442404503,
+      "grad_norm": 1.0900389528678076,
+      "learning_rate": 7.290520355304844e-06,
+      "loss": 0.7187,
+      "step": 3037
+    },
+    {
+      "epoch": 1.799230085875037,
+      "grad_norm": 1.1766868265659325,
+      "learning_rate": 7.284364398844412e-06,
+      "loss": 0.7759,
+      "step": 3038
+    },
+    {
+      "epoch": 1.799822327509624,
+      "grad_norm": 0.969905582098005,
+      "learning_rate": 7.2782095532316486e-06,
+      "loss": 0.6934,
+      "step": 3039
+    },
+    {
+      "epoch": 1.8004145691442108,
+      "grad_norm": 1.0483421180595986,
+      "learning_rate": 7.27205582098423e-06,
+      "loss": 0.6879,
+      "step": 3040
+    },
+    {
+      "epoch": 1.8010068107787978,
+      "grad_norm": 1.4279823666510263,
+      "learning_rate": 7.265903204619386e-06,
+      "loss": 0.6903,
+      "step": 3041
+    },
+    {
+      "epoch": 1.8015990524133847,
+      "grad_norm": 1.036318973747682,
+      "learning_rate": 7.25975170665387e-06,
+      "loss": 0.7208,
+      "step": 3042
+    },
+    {
+      "epoch": 1.8021912940479716,
+      "grad_norm": 1.1321513164832115,
+      "learning_rate": 7.253601329604001e-06,
+      "loss": 0.7162,
+      "step": 3043
+    },
+    {
+      "epoch": 1.8027835356825586,
+      "grad_norm": 7.227215013383004,
+      "learning_rate": 7.247452075985622e-06,
+      "loss": 0.7006,
+      "step": 3044
+    },
+    {
+      "epoch": 1.8033757773171453,
+      "grad_norm": 0.9690712025929449,
+      "learning_rate": 7.241303948314135e-06,
+      "loss": 0.7178,
+      "step": 3045
+    },
+    {
+      "epoch": 1.8039680189517324,
+      "grad_norm": 2.051690522200872,
+      "learning_rate": 7.235156949104455e-06,
+      "loss": 0.6948,
+      "step": 3046
+    },
+    {
+      "epoch": 1.8045602605863191,
+      "grad_norm": 0.975264804113868,
+      "learning_rate": 7.22901108087106e-06,
+      "loss": 0.7484,
+      "step": 3047
+    },
+    {
+      "epoch": 1.8051525022209063,
+      "grad_norm": 1.286263931105427,
+      "learning_rate": 7.222866346127952e-06,
+      "loss": 0.7124,
+      "step": 3048
+    },
+    {
+      "epoch": 1.805744743855493,
+      "grad_norm": 1.5309931798364385,
+      "learning_rate": 7.216722747388678e-06,
+      "loss": 0.7074,
+      "step": 3049
+    },
+    {
+      "epoch": 1.80633698549008,
+      "grad_norm": 1.7076934436600835,
+      "learning_rate": 7.210580287166307e-06,
+      "loss": 0.7149,
+      "step": 3050
+    },
+    {
+      "epoch": 1.8069292271246669,
+      "grad_norm": 1.1300533995630113,
+      "learning_rate": 7.2044389679734564e-06,
+      "loss": 0.7234,
+      "step": 3051
+    },
+    {
+      "epoch": 1.8075214687592538,
+      "grad_norm": 1.0540182120458212,
+      "learning_rate": 7.198298792322271e-06,
+      "loss": 0.7289,
+      "step": 3052
+    },
+    {
+      "epoch": 1.8081137103938407,
+      "grad_norm": 1.4987232839307902,
+      "learning_rate": 7.192159762724427e-06,
+      "loss": 0.7248,
+      "step": 3053
+    },
+    {
+      "epoch": 1.8087059520284274,
+      "grad_norm": 1.0497678988574621,
+      "learning_rate": 7.186021881691132e-06,
+      "loss": 0.6714,
+      "step": 3054
+    },
+    {
+      "epoch": 1.8092981936630146,
+      "grad_norm": 1.0903848686159592,
+      "learning_rate": 7.179885151733124e-06,
+      "loss": 0.6885,
+      "step": 3055
+    },
+    {
+      "epoch": 1.8098904352976013,
+      "grad_norm": 1.0133138934590773,
+      "learning_rate": 7.173749575360671e-06,
+      "loss": 0.7397,
+      "step": 3056
+    },
+    {
+      "epoch": 1.8104826769321885,
+      "grad_norm": 1.2805524494775062,
+      "learning_rate": 7.167615155083574e-06,
+      "loss": 0.6995,
+      "step": 3057
+    },
+    {
+      "epoch": 1.8110749185667752,
+      "grad_norm": 1.353348949222249,
+      "learning_rate": 7.1614818934111475e-06,
+      "loss": 0.704,
+      "step": 3058
+    },
+    {
+      "epoch": 1.8116671602013623,
+      "grad_norm": 1.7884004913807032,
+      "learning_rate": 7.155349792852242e-06,
+      "loss": 0.7066,
+      "step": 3059
+    },
+    {
+      "epoch": 1.812259401835949,
+      "grad_norm": 1.1372875659844779,
+      "learning_rate": 7.1492188559152364e-06,
+      "loss": 0.7382,
+      "step": 3060
+    },
+    {
+      "epoch": 1.812851643470536,
+      "grad_norm": 7.427167503376271,
+      "learning_rate": 7.143089085108028e-06,
+      "loss": 0.7224,
+      "step": 3061
+    },
+    {
+      "epoch": 1.8134438851051229,
+      "grad_norm": 1.602444693750489,
+      "learning_rate": 7.136960482938035e-06,
+      "loss": 0.7293,
+      "step": 3062
+    },
+    {
+      "epoch": 1.8140361267397098,
+      "grad_norm": 4.355339095528307,
+      "learning_rate": 7.130833051912198e-06,
+      "loss": 0.7455,
+      "step": 3063
+    },
+    {
+      "epoch": 1.8146283683742968,
+      "grad_norm": 1.2740183492816513,
+      "learning_rate": 7.124706794536984e-06,
+      "loss": 0.7043,
+      "step": 3064
+    },
+    {
+      "epoch": 1.8152206100088835,
+      "grad_norm": 1.0624879073095552,
+      "learning_rate": 7.11858171331838e-06,
+      "loss": 0.7437,
+      "step": 3065
+    },
+    {
+      "epoch": 1.8158128516434706,
+      "grad_norm": 1.1249819254474553,
+      "learning_rate": 7.112457810761883e-06,
+      "loss": 0.7151,
+      "step": 3066
+    },
+    {
+      "epoch": 1.8164050932780573,
+      "grad_norm": 2.4801786068161737,
+      "learning_rate": 7.106335089372517e-06,
+      "loss": 0.7494,
+      "step": 3067
+    },
+    {
+      "epoch": 1.8169973349126445,
+      "grad_norm": 1.7960579111312531,
+      "learning_rate": 7.100213551654816e-06,
+      "loss": 0.6956,
+      "step": 3068
+    },
+    {
+      "epoch": 1.8175895765472312,
+      "grad_norm": 1.206786649404624,
+      "learning_rate": 7.0940932001128395e-06,
+      "loss": 0.7599,
+      "step": 3069
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 1.3862079077433727,
+      "learning_rate": 7.087974037250146e-06,
+      "loss": 0.7371,
+      "step": 3070
+    },
+    {
+      "epoch": 1.818774059816405,
+      "grad_norm": 1.9789719604629894,
+      "learning_rate": 7.0818560655698246e-06,
+      "loss": 0.7472,
+      "step": 3071
+    },
+    {
+      "epoch": 1.819366301450992,
+      "grad_norm": 1.4121044579900446,
+      "learning_rate": 7.075739287574467e-06,
+      "loss": 0.7147,
+      "step": 3072
+    },
+    {
+      "epoch": 1.819958543085579,
+      "grad_norm": 2.1510778746002455,
+      "learning_rate": 7.069623705766182e-06,
+      "loss": 0.7472,
+      "step": 3073
+    },
+    {
+      "epoch": 1.8205507847201658,
+      "grad_norm": 1.0077829214521765,
+      "learning_rate": 7.063509322646581e-06,
+      "loss": 0.6905,
+      "step": 3074
+    },
+    {
+      "epoch": 1.8211430263547528,
+      "grad_norm": 1.3469407014130284,
+      "learning_rate": 7.057396140716796e-06,
+      "loss": 0.7024,
+      "step": 3075
+    },
+    {
+      "epoch": 1.8217352679893395,
+      "grad_norm": 1.876760028468839,
+      "learning_rate": 7.051284162477459e-06,
+      "loss": 0.7303,
+      "step": 3076
+    },
+    {
+      "epoch": 1.8223275096239266,
+      "grad_norm": 1.049747547083592,
+      "learning_rate": 7.0451733904287166e-06,
+      "loss": 0.7227,
+      "step": 3077
+    },
+    {
+      "epoch": 1.8229197512585134,
+      "grad_norm": 1.0058819206164897,
+      "learning_rate": 7.039063827070214e-06,
+      "loss": 0.713,
+      "step": 3078
+    },
+    {
+      "epoch": 1.8235119928931005,
+      "grad_norm": 3.6076196015946236,
+      "learning_rate": 7.0329554749011045e-06,
+      "loss": 0.7295,
+      "step": 3079
+    },
+    {
+      "epoch": 1.8241042345276872,
+      "grad_norm": 1.0057618249589486,
+      "learning_rate": 7.026848336420053e-06,
+      "loss": 0.7129,
+      "step": 3080
+    },
+    {
+      "epoch": 1.8246964761622744,
+      "grad_norm": 1.7386425971569412,
+      "learning_rate": 7.020742414125223e-06,
+      "loss": 0.6903,
+      "step": 3081
+    },
+    {
+      "epoch": 1.825288717796861,
+      "grad_norm": 2.473617947321098,
+      "learning_rate": 7.014637710514274e-06,
+      "loss": 0.7285,
+      "step": 3082
+    },
+    {
+      "epoch": 1.825880959431448,
+      "grad_norm": 0.9327751534341746,
+      "learning_rate": 7.008534228084376e-06,
+      "loss": 0.7183,
+      "step": 3083
+    },
+    {
+      "epoch": 1.826473201066035,
+      "grad_norm": 1.0584729282570624,
+      "learning_rate": 7.002431969332197e-06,
+      "loss": 0.7485,
+      "step": 3084
+    },
+    {
+      "epoch": 1.8270654427006219,
+      "grad_norm": 0.9797461090483691,
+      "learning_rate": 6.996330936753907e-06,
+      "loss": 0.7756,
+      "step": 3085
+    },
+    {
+      "epoch": 1.8276576843352088,
+      "grad_norm": 1.1221799073488412,
+      "learning_rate": 6.990231132845169e-06,
+      "loss": 0.7499,
+      "step": 3086
+    },
+    {
+      "epoch": 1.8282499259697955,
+      "grad_norm": 0.9306758935450702,
+      "learning_rate": 6.984132560101143e-06,
+      "loss": 0.7304,
+      "step": 3087
+    },
+    {
+      "epoch": 1.8288421676043827,
+      "grad_norm": 1.0887511525648792,
+      "learning_rate": 6.978035221016487e-06,
+      "loss": 0.7267,
+      "step": 3088
+    },
+    {
+      "epoch": 1.8294344092389694,
+      "grad_norm": 1.1273648529558058,
+      "learning_rate": 6.971939118085365e-06,
+      "loss": 0.6752,
+      "step": 3089
+    },
+    {
+      "epoch": 1.8300266508735565,
+      "grad_norm": 1.0281268202300238,
+      "learning_rate": 6.965844253801416e-06,
+      "loss": 0.723,
+      "step": 3090
+    },
+    {
+      "epoch": 1.8306188925081432,
+      "grad_norm": 1.0970082995953543,
+      "learning_rate": 6.959750630657787e-06,
+      "loss": 0.7901,
+      "step": 3091
+    },
+    {
+      "epoch": 1.8312111341427302,
+      "grad_norm": 1.0998649102940627,
+      "learning_rate": 6.953658251147109e-06,
+      "loss": 0.7464,
+      "step": 3092
+    },
+    {
+      "epoch": 1.8318033757773171,
+      "grad_norm": 1.6332086008816058,
+      "learning_rate": 6.947567117761517e-06,
+      "loss": 0.7037,
+      "step": 3093
+    },
+    {
+      "epoch": 1.832395617411904,
+      "grad_norm": 2.4583445925569114,
+      "learning_rate": 6.941477232992614e-06,
+      "loss": 0.698,
+      "step": 3094
+    },
+    {
+      "epoch": 1.832987859046491,
+      "grad_norm": 1.1349251674267553,
+      "learning_rate": 6.935388599331514e-06,
+      "loss": 0.6615,
+      "step": 3095
+    },
+    {
+      "epoch": 1.833580100681078,
+      "grad_norm": 1.0921356725030107,
+      "learning_rate": 6.929301219268806e-06,
+      "loss": 0.7015,
+      "step": 3096
+    },
+    {
+      "epoch": 1.8341723423156648,
+      "grad_norm": 1.534601523832624,
+      "learning_rate": 6.92321509529458e-06,
+      "loss": 0.6909,
+      "step": 3097
+    },
+    {
+      "epoch": 1.8347645839502515,
+      "grad_norm": 1.2237585847130814,
+      "learning_rate": 6.917130229898387e-06,
+      "loss": 0.6804,
+      "step": 3098
+    },
+    {
+      "epoch": 1.8353568255848387,
+      "grad_norm": 1.6141007678469457,
+      "learning_rate": 6.911046625569293e-06,
+      "loss": 0.7228,
+      "step": 3099
+    },
+    {
+      "epoch": 1.8359490672194254,
+      "grad_norm": 0.9997505945449734,
+      "learning_rate": 6.90496428479583e-06,
+      "loss": 0.7101,
+      "step": 3100
+    },
+    {
+      "epoch": 1.8365413088540126,
+      "grad_norm": 1.0339643235502096,
+      "learning_rate": 6.898883210066018e-06,
+      "loss": 0.7141,
+      "step": 3101
+    },
+    {
+      "epoch": 1.8371335504885993,
+      "grad_norm": 0.9214598557673855,
+      "learning_rate": 6.892803403867352e-06,
+      "loss": 0.7178,
+      "step": 3102
+    },
+    {
+      "epoch": 1.8377257921231862,
+      "grad_norm": 1.2899691433939153,
+      "learning_rate": 6.886724868686823e-06,
+      "loss": 0.7322,
+      "step": 3103
+    },
+    {
+      "epoch": 1.8383180337577731,
+      "grad_norm": 1.0878616869188478,
+      "learning_rate": 6.8806476070108905e-06,
+      "loss": 0.7229,
+      "step": 3104
+    },
+    {
+      "epoch": 1.83891027539236,
+      "grad_norm": 2.1890307846418455,
+      "learning_rate": 6.874571621325498e-06,
+      "loss": 0.7179,
+      "step": 3105
+    },
+    {
+      "epoch": 1.839502517026947,
+      "grad_norm": 1.2042122977243805,
+      "learning_rate": 6.868496914116063e-06,
+      "loss": 0.7208,
+      "step": 3106
+    },
+    {
+      "epoch": 1.840094758661534,
+      "grad_norm": 1.1067009260330778,
+      "learning_rate": 6.86242348786748e-06,
+      "loss": 0.7104,
+      "step": 3107
+    },
+    {
+      "epoch": 1.8406870002961209,
+      "grad_norm": 1.977612629506757,
+      "learning_rate": 6.856351345064127e-06,
+      "loss": 0.7083,
+      "step": 3108
+    },
+    {
+      "epoch": 1.8412792419307076,
+      "grad_norm": 1.183996644590801,
+      "learning_rate": 6.850280488189851e-06,
+      "loss": 0.7136,
+      "step": 3109
+    },
+    {
+      "epoch": 1.8418714835652947,
+      "grad_norm": 1.1336099389139767,
+      "learning_rate": 6.844210919727971e-06,
+      "loss": 0.7121,
+      "step": 3110
+    },
+    {
+      "epoch": 1.8424637251998814,
+      "grad_norm": 0.9915840493395013,
+      "learning_rate": 6.838142642161283e-06,
+      "loss": 0.7146,
+      "step": 3111
+    },
+    {
+      "epoch": 1.8430559668344686,
+      "grad_norm": 1.0672864776594517,
+      "learning_rate": 6.8320756579720545e-06,
+      "loss": 0.7185,
+      "step": 3112
+    },
+    {
+      "epoch": 1.8436482084690553,
+      "grad_norm": 1.2611537685575909,
+      "learning_rate": 6.826009969642027e-06,
+      "loss": 0.7399,
+      "step": 3113
+    },
+    {
+      "epoch": 1.8442404501036422,
+      "grad_norm": 3.728635542061314,
+      "learning_rate": 6.819945579652401e-06,
+      "loss": 0.7557,
+      "step": 3114
+    },
+    {
+      "epoch": 1.8448326917382292,
+      "grad_norm": 1.3284266412732086,
+      "learning_rate": 6.813882490483854e-06,
+      "loss": 0.6961,
+      "step": 3115
+    },
+    {
+      "epoch": 1.845424933372816,
+      "grad_norm": 19.468979938411216,
+      "learning_rate": 6.807820704616532e-06,
+      "loss": 0.7411,
+      "step": 3116
+    },
+    {
+      "epoch": 1.846017175007403,
+      "grad_norm": 0.8689154585001878,
+      "learning_rate": 6.801760224530052e-06,
+      "loss": 0.7244,
+      "step": 3117
+    },
+    {
+      "epoch": 1.84660941664199,
+      "grad_norm": 1.068439800318271,
+      "learning_rate": 6.795701052703482e-06,
+      "loss": 0.7105,
+      "step": 3118
+    },
+    {
+      "epoch": 1.847201658276577,
+      "grad_norm": 0.9526703958307156,
+      "learning_rate": 6.7896431916153684e-06,
+      "loss": 0.7437,
+      "step": 3119
+    },
+    {
+      "epoch": 1.8477938999111636,
+      "grad_norm": 1.0630490772363779,
+      "learning_rate": 6.783586643743714e-06,
+      "loss": 0.7221,
+      "step": 3120
+    },
+    {
+      "epoch": 1.8483861415457508,
+      "grad_norm": 3.5896361583850567,
+      "learning_rate": 6.777531411565996e-06,
+      "loss": 0.7202,
+      "step": 3121
+    },
+    {
+      "epoch": 1.8489783831803375,
+      "grad_norm": 1.7343409372636471,
+      "learning_rate": 6.7714774975591335e-06,
+      "loss": 0.7145,
+      "step": 3122
+    },
+    {
+      "epoch": 1.8495706248149246,
+      "grad_norm": 1.0396615169001493,
+      "learning_rate": 6.7654249041995256e-06,
+      "loss": 0.7005,
+      "step": 3123
+    },
+    {
+      "epoch": 1.8501628664495113,
+      "grad_norm": 1.642370057671462,
+      "learning_rate": 6.75937363396302e-06,
+      "loss": 0.7372,
+      "step": 3124
+    },
+    {
+      "epoch": 1.8507551080840983,
+      "grad_norm": 1.1074855634634668,
+      "learning_rate": 6.753323689324931e-06,
+      "loss": 0.7431,
+      "step": 3125
+    },
+    {
+      "epoch": 1.8513473497186852,
+      "grad_norm": 0.9324152655129814,
+      "learning_rate": 6.7472750727600155e-06,
+      "loss": 0.7135,
+      "step": 3126
+    },
+    {
+      "epoch": 1.8519395913532721,
+      "grad_norm": 0.8914241503209785,
+      "learning_rate": 6.741227786742509e-06,
+      "loss": 0.7181,
+      "step": 3127
+    },
+    {
+      "epoch": 1.852531832987859,
+      "grad_norm": 1.1891158801295643,
+      "learning_rate": 6.735181833746087e-06,
+      "loss": 0.6945,
+      "step": 3128
+    },
+    {
+      "epoch": 1.853124074622446,
+      "grad_norm": 1.0727481880906515,
+      "learning_rate": 6.729137216243886e-06,
+      "loss": 0.6876,
+      "step": 3129
+    },
+    {
+      "epoch": 1.853716316257033,
+      "grad_norm": 1.0797517345200798,
+      "learning_rate": 6.7230939367084915e-06,
+      "loss": 0.7165,
+      "step": 3130
+    },
+    {
+      "epoch": 1.8543085578916196,
+      "grad_norm": 1.4893641209790685,
+      "learning_rate": 6.717051997611944e-06,
+      "loss": 0.7433,
+      "step": 3131
+    },
+    {
+      "epoch": 1.8549007995262068,
+      "grad_norm": 1.0413984547576909,
+      "learning_rate": 6.711011401425741e-06,
+      "loss": 0.6766,
+      "step": 3132
+    },
+    {
+      "epoch": 1.8554930411607935,
+      "grad_norm": 1.1532788341257012,
+      "learning_rate": 6.704972150620825e-06,
+      "loss": 0.6584,
+      "step": 3133
+    },
+    {
+      "epoch": 1.8560852827953807,
+      "grad_norm": 1.680364023670173,
+      "learning_rate": 6.698934247667587e-06,
+      "loss": 0.7095,
+      "step": 3134
+    },
+    {
+      "epoch": 1.8566775244299674,
+      "grad_norm": 1.5274601112160506,
+      "learning_rate": 6.692897695035866e-06,
+      "loss": 0.6719,
+      "step": 3135
+    },
+    {
+      "epoch": 1.8572697660645543,
+      "grad_norm": 1.4344751525084347,
+      "learning_rate": 6.686862495194958e-06,
+      "loss": 0.7164,
+      "step": 3136
+    },
+    {
+      "epoch": 1.8578620076991412,
+      "grad_norm": 1.305338228972161,
+      "learning_rate": 6.6808286506135975e-06,
+      "loss": 0.7633,
+      "step": 3137
+    },
+    {
+      "epoch": 1.8584542493337282,
+      "grad_norm": 1.1939810781229667,
+      "learning_rate": 6.6747961637599645e-06,
+      "loss": 0.7098,
+      "step": 3138
+    },
+    {
+      "epoch": 1.859046490968315,
+      "grad_norm": 1.454501033733436,
+      "learning_rate": 6.668765037101682e-06,
+      "loss": 0.7129,
+      "step": 3139
+    },
+    {
+      "epoch": 1.859638732602902,
+      "grad_norm": 0.9270487154278511,
+      "learning_rate": 6.662735273105827e-06,
+      "loss": 0.7298,
+      "step": 3140
+    },
+    {
+      "epoch": 1.860230974237489,
+      "grad_norm": 0.887328217763018,
+      "learning_rate": 6.656706874238909e-06,
+      "loss": 0.6998,
+      "step": 3141
+    },
+    {
+      "epoch": 1.8608232158720757,
+      "grad_norm": 0.9240842979460729,
+      "learning_rate": 6.650679842966881e-06,
+      "loss": 0.7242,
+      "step": 3142
+    },
+    {
+      "epoch": 1.8614154575066628,
+      "grad_norm": 1.4560811731823002,
+      "learning_rate": 6.644654181755139e-06,
+      "loss": 0.7467,
+      "step": 3143
+    },
+    {
+      "epoch": 1.8620076991412495,
+      "grad_norm": 1.8642002876762704,
+      "learning_rate": 6.638629893068516e-06,
+      "loss": 0.7094,
+      "step": 3144
+    },
+    {
+      "epoch": 1.8625999407758367,
+      "grad_norm": 1.3637576327896082,
+      "learning_rate": 6.632606979371287e-06,
+      "loss": 0.7453,
+      "step": 3145
+    },
+    {
+      "epoch": 1.8631921824104234,
+      "grad_norm": 1.318364652725098,
+      "learning_rate": 6.62658544312716e-06,
+      "loss": 0.6683,
+      "step": 3146
+    },
+    {
+      "epoch": 1.8637844240450103,
+      "grad_norm": 1.457651252213059,
+      "learning_rate": 6.620565286799283e-06,
+      "loss": 0.7037,
+      "step": 3147
+    },
+    {
+      "epoch": 1.8643766656795973,
+      "grad_norm": 1.7273054218815638,
+      "learning_rate": 6.614546512850237e-06,
+      "loss": 0.7301,
+      "step": 3148
+    },
+    {
+      "epoch": 1.8649689073141842,
+      "grad_norm": 1.2049671686395615,
+      "learning_rate": 6.608529123742047e-06,
+      "loss": 0.728,
+      "step": 3149
+    },
+    {
+      "epoch": 1.8655611489487711,
+      "grad_norm": 1.17602916283294,
+      "learning_rate": 6.6025131219361505e-06,
+      "loss": 0.7274,
+      "step": 3150
+    },
+    {
+      "epoch": 1.866153390583358,
+      "grad_norm": 2.03746704987476,
+      "learning_rate": 6.596498509893438e-06,
+      "loss": 0.7063,
+      "step": 3151
+    },
+    {
+      "epoch": 1.866745632217945,
+      "grad_norm": 1.1900210213393627,
+      "learning_rate": 6.590485290074224e-06,
+      "loss": 0.721,
+      "step": 3152
+    },
+    {
+      "epoch": 1.8673378738525317,
+      "grad_norm": 2.1534721717971452,
+      "learning_rate": 6.584473464938257e-06,
+      "loss": 0.7324,
+      "step": 3153
+    },
+    {
+      "epoch": 1.8679301154871188,
+      "grad_norm": 2.9498552630407606,
+      "learning_rate": 6.5784630369447e-06,
+      "loss": 0.7615,
+      "step": 3154
+    },
+    {
+      "epoch": 1.8685223571217056,
+      "grad_norm": 1.2185800168618288,
+      "learning_rate": 6.572454008552166e-06,
+      "loss": 0.7466,
+      "step": 3155
+    },
+    {
+      "epoch": 1.8691145987562927,
+      "grad_norm": 1.1826866438599948,
+      "learning_rate": 6.566446382218683e-06,
+      "loss": 0.745,
+      "step": 3156
+    },
+    {
+      "epoch": 1.8697068403908794,
+      "grad_norm": 1.6999039368124431,
+      "learning_rate": 6.5604401604017095e-06,
+      "loss": 0.6985,
+      "step": 3157
+    },
+    {
+      "epoch": 1.8702990820254664,
+      "grad_norm": 1.244053260256039,
+      "learning_rate": 6.5544353455581245e-06,
+      "loss": 0.6739,
+      "step": 3158
+    },
+    {
+      "epoch": 1.8708913236600533,
+      "grad_norm": 1.718673210214095,
+      "learning_rate": 6.5484319401442346e-06,
+      "loss": 0.7268,
+      "step": 3159
+    },
+    {
+      "epoch": 1.8714835652946402,
+      "grad_norm": 1.9410211268230035,
+      "learning_rate": 6.542429946615774e-06,
+      "loss": 0.7235,
+      "step": 3160
+    },
+    {
+      "epoch": 1.8720758069292271,
+      "grad_norm": 1.1807820843110362,
+      "learning_rate": 6.536429367427896e-06,
+      "loss": 0.7566,
+      "step": 3161
+    },
+    {
+      "epoch": 1.872668048563814,
+      "grad_norm": 1.4651639003519918,
+      "learning_rate": 6.53043020503517e-06,
+      "loss": 0.7173,
+      "step": 3162
+    },
+    {
+      "epoch": 1.873260290198401,
+      "grad_norm": 2.002968615540737,
+      "learning_rate": 6.5244324618915925e-06,
+      "loss": 0.7393,
+      "step": 3163
+    },
+    {
+      "epoch": 1.8738525318329877,
+      "grad_norm": 1.1224074176105243,
+      "learning_rate": 6.5184361404505795e-06,
+      "loss": 0.7074,
+      "step": 3164
+    },
+    {
+      "epoch": 1.8744447734675749,
+      "grad_norm": 1.5396485670035198,
+      "learning_rate": 6.512441243164967e-06,
+      "loss": 0.7426,
+      "step": 3165
+    },
+    {
+      "epoch": 1.8750370151021616,
+      "grad_norm": 1.1447433665685038,
+      "learning_rate": 6.506447772486997e-06,
+      "loss": 0.7526,
+      "step": 3166
+    },
+    {
+      "epoch": 1.8756292567367487,
+      "grad_norm": 1.943211358510783,
+      "learning_rate": 6.50045573086834e-06,
+      "loss": 0.7328,
+      "step": 3167
+    },
+    {
+      "epoch": 1.8762214983713354,
+      "grad_norm": 1.8487347809196877,
+      "learning_rate": 6.4944651207600765e-06,
+      "loss": 0.7435,
+      "step": 3168
+    },
+    {
+      "epoch": 1.8768137400059224,
+      "grad_norm": 0.9270621419455923,
+      "learning_rate": 6.488475944612709e-06,
+      "loss": 0.7267,
+      "step": 3169
+    },
+    {
+      "epoch": 1.8774059816405093,
+      "grad_norm": 1.4497042883892517,
+      "learning_rate": 6.4824882048761406e-06,
+      "loss": 0.7384,
+      "step": 3170
+    },
+    {
+      "epoch": 1.8779982232750962,
+      "grad_norm": 1.7243801348273786,
+      "learning_rate": 6.476501903999695e-06,
+      "loss": 0.7276,
+      "step": 3171
+    },
+    {
+      "epoch": 1.8785904649096832,
+      "grad_norm": 2.212944195283541,
+      "learning_rate": 6.470517044432104e-06,
+      "loss": 0.7246,
+      "step": 3172
+    },
+    {
+      "epoch": 1.87918270654427,
+      "grad_norm": 2.9708159494514095,
+      "learning_rate": 6.464533628621521e-06,
+      "loss": 0.7134,
+      "step": 3173
+    },
+    {
+      "epoch": 1.879774948178857,
+      "grad_norm": 1.1598634102463192,
+      "learning_rate": 6.458551659015486e-06,
+      "loss": 0.6919,
+      "step": 3174
+    },
+    {
+      "epoch": 1.8803671898134438,
+      "grad_norm": 1.0582572327884274,
+      "learning_rate": 6.452571138060971e-06,
+      "loss": 0.7249,
+      "step": 3175
+    },
+    {
+      "epoch": 1.880959431448031,
+      "grad_norm": 1.2040138425826092,
+      "learning_rate": 6.446592068204341e-06,
+      "loss": 0.7295,
+      "step": 3176
+    },
+    {
+      "epoch": 1.8815516730826176,
+      "grad_norm": 2.157200215176374,
+      "learning_rate": 6.440614451891376e-06,
+      "loss": 0.7359,
+      "step": 3177
+    },
+    {
+      "epoch": 1.8821439147172048,
+      "grad_norm": 1.6977745304387328,
+      "learning_rate": 6.43463829156725e-06,
+      "loss": 0.7341,
+      "step": 3178
+    },
+    {
+      "epoch": 1.8827361563517915,
+      "grad_norm": 1.5684451007006281,
+      "learning_rate": 6.428663589676554e-06,
+      "loss": 0.72,
+      "step": 3179
+    },
+    {
+      "epoch": 1.8833283979863784,
+      "grad_norm": 1.1271225922581531,
+      "learning_rate": 6.422690348663276e-06,
+      "loss": 0.7196,
+      "step": 3180
+    },
+    {
+      "epoch": 1.8839206396209653,
+      "grad_norm": 1.8620701322408784,
+      "learning_rate": 6.41671857097081e-06,
+      "loss": 0.6995,
+      "step": 3181
+    },
+    {
+      "epoch": 1.8845128812555523,
+      "grad_norm": 1.7096834779094625,
+      "learning_rate": 6.410748259041941e-06,
+      "loss": 0.7097,
+      "step": 3182
+    },
+    {
+      "epoch": 1.8851051228901392,
+      "grad_norm": 1.4823000392501897,
+      "learning_rate": 6.40477941531887e-06,
+      "loss": 0.7037,
+      "step": 3183
+    },
+    {
+      "epoch": 1.8856973645247261,
+      "grad_norm": 1.226315830076985,
+      "learning_rate": 6.398812042243187e-06,
+      "loss": 0.728,
+      "step": 3184
+    },
+    {
+      "epoch": 1.886289606159313,
+      "grad_norm": 0.9904003037669937,
+      "learning_rate": 6.392846142255886e-06,
+      "loss": 0.6943,
+      "step": 3185
+    },
+    {
+      "epoch": 1.8868818477938998,
+      "grad_norm": 1.2611406463421255,
+      "learning_rate": 6.3868817177973505e-06,
+      "loss": 0.7045,
+      "step": 3186
+    },
+    {
+      "epoch": 1.887474089428487,
+      "grad_norm": 1.4663012894424847,
+      "learning_rate": 6.380918771307365e-06,
+      "loss": 0.6833,
+      "step": 3187
+    },
+    {
+      "epoch": 1.8880663310630736,
+      "grad_norm": 1.3127393848518931,
+      "learning_rate": 6.3749573052251155e-06,
+      "loss": 0.7176,
+      "step": 3188
+    },
+    {
+      "epoch": 1.8886585726976608,
+      "grad_norm": 1.4212019430587715,
+      "learning_rate": 6.368997321989176e-06,
+      "loss": 0.7226,
+      "step": 3189
+    },
+    {
+      "epoch": 1.8892508143322475,
+      "grad_norm": 1.2265922099963948,
+      "learning_rate": 6.363038824037511e-06,
+      "loss": 0.7691,
+      "step": 3190
+    },
+    {
+      "epoch": 1.8898430559668344,
+      "grad_norm": 2.4233051026500196,
+      "learning_rate": 6.357081813807482e-06,
+      "loss": 0.736,
+      "step": 3191
+    },
+    {
+      "epoch": 1.8904352976014214,
+      "grad_norm": 1.3155312645080448,
+      "learning_rate": 6.351126293735843e-06,
+      "loss": 0.7628,
+      "step": 3192
+    },
+    {
+      "epoch": 1.8910275392360083,
+      "grad_norm": 1.411204552132623,
+      "learning_rate": 6.345172266258739e-06,
+      "loss": 0.7293,
+      "step": 3193
+    },
+    {
+      "epoch": 1.8916197808705952,
+      "grad_norm": 1.5791438885255882,
+      "learning_rate": 6.339219733811697e-06,
+      "loss": 0.6975,
+      "step": 3194
+    },
+    {
+      "epoch": 1.8922120225051822,
+      "grad_norm": 1.4220630376908199,
+      "learning_rate": 6.333268698829639e-06,
+      "loss": 0.7482,
+      "step": 3195
+    },
+    {
+      "epoch": 1.892804264139769,
+      "grad_norm": 1.3250303025534889,
+      "learning_rate": 6.327319163746871e-06,
+      "loss": 0.7284,
+      "step": 3196
+    },
+    {
+      "epoch": 1.8933965057743558,
+      "grad_norm": 1.0527611253118723,
+      "learning_rate": 6.321371130997095e-06,
+      "loss": 0.6912,
+      "step": 3197
+    },
+    {
+      "epoch": 1.893988747408943,
+      "grad_norm": 1.1033965769900478,
+      "learning_rate": 6.315424603013382e-06,
+      "loss": 0.7213,
+      "step": 3198
+    },
+    {
+      "epoch": 1.8945809890435297,
+      "grad_norm": 1.2063474598713093,
+      "learning_rate": 6.309479582228202e-06,
+      "loss": 0.7157,
+      "step": 3199
+    },
+    {
+      "epoch": 1.8951732306781168,
+      "grad_norm": 1.269742008185876,
+      "learning_rate": 6.303536071073397e-06,
+      "loss": 0.7374,
+      "step": 3200
+    },
+    {
+      "epoch": 1.8957654723127035,
+      "grad_norm": 1.8184812717508338,
+      "learning_rate": 6.297594071980208e-06,
+      "loss": 0.6878,
+      "step": 3201
+    },
+    {
+      "epoch": 1.8963577139472905,
+      "grad_norm": 1.511960235444543,
+      "learning_rate": 6.291653587379236e-06,
+      "loss": 0.7505,
+      "step": 3202
+    },
+    {
+      "epoch": 1.8969499555818774,
+      "grad_norm": 1.2751220645522783,
+      "learning_rate": 6.2857146197004755e-06,
+      "loss": 0.8061,
+      "step": 3203
+    },
+    {
+      "epoch": 1.8975421972164643,
+      "grad_norm": 1.1058079431721495,
+      "learning_rate": 6.2797771713733025e-06,
+      "loss": 0.7353,
+      "step": 3204
+    },
+    {
+      "epoch": 1.8981344388510513,
+      "grad_norm": 0.9715354260971804,
+      "learning_rate": 6.273841244826466e-06,
+      "loss": 0.7067,
+      "step": 3205
+    },
+    {
+      "epoch": 1.898726680485638,
+      "grad_norm": 1.3513816230607942,
+      "learning_rate": 6.267906842488088e-06,
+      "loss": 0.7182,
+      "step": 3206
+    },
+    {
+      "epoch": 1.8993189221202251,
+      "grad_norm": 2.4611133660776714,
+      "learning_rate": 6.261973966785679e-06,
+      "loss": 0.7578,
+      "step": 3207
+    },
+    {
+      "epoch": 1.8999111637548118,
+      "grad_norm": 3.4200966306498968,
+      "learning_rate": 6.256042620146119e-06,
+      "loss": 0.7041,
+      "step": 3208
+    },
+    {
+      "epoch": 1.900503405389399,
+      "grad_norm": 1.0636075328664951,
+      "learning_rate": 6.250112804995661e-06,
+      "loss": 0.7919,
+      "step": 3209
+    },
+    {
+      "epoch": 1.9010956470239857,
+      "grad_norm": 1.1252808954511786,
+      "learning_rate": 6.2441845237599285e-06,
+      "loss": 0.7849,
+      "step": 3210
+    },
+    {
+      "epoch": 1.9016878886585729,
+      "grad_norm": 0.9806450437275593,
+      "learning_rate": 6.238257778863925e-06,
+      "loss": 0.7062,
+      "step": 3211
+    },
+    {
+      "epoch": 1.9022801302931596,
+      "grad_norm": 1.6392036224674011,
+      "learning_rate": 6.232332572732025e-06,
+      "loss": 0.7196,
+      "step": 3212
+    },
+    {
+      "epoch": 1.9028723719277465,
+      "grad_norm": 1.5302573409478557,
+      "learning_rate": 6.226408907787971e-06,
+      "loss": 0.7276,
+      "step": 3213
+    },
+    {
+      "epoch": 1.9034646135623334,
+      "grad_norm": 3.974467334315002,
+      "learning_rate": 6.22048678645487e-06,
+      "loss": 0.6877,
+      "step": 3214
+    },
+    {
+      "epoch": 1.9040568551969204,
+      "grad_norm": 1.2222534462798487,
+      "learning_rate": 6.2145662111552045e-06,
+      "loss": 0.7145,
+      "step": 3215
+    },
+    {
+      "epoch": 1.9046490968315073,
+      "grad_norm": 0.8412923141136364,
+      "learning_rate": 6.208647184310826e-06,
+      "loss": 0.7067,
+      "step": 3216
+    },
+    {
+      "epoch": 1.905241338466094,
+      "grad_norm": 1.8567585845983157,
+      "learning_rate": 6.202729708342948e-06,
+      "loss": 0.7041,
+      "step": 3217
+    },
+    {
+      "epoch": 1.9058335801006812,
+      "grad_norm": 1.3812265601103704,
+      "learning_rate": 6.196813785672149e-06,
+      "loss": 0.7327,
+      "step": 3218
+    },
+    {
+      "epoch": 1.9064258217352679,
+      "grad_norm": 1.490042310350291,
+      "learning_rate": 6.1908994187183726e-06,
+      "loss": 0.7277,
+      "step": 3219
+    },
+    {
+      "epoch": 1.907018063369855,
+      "grad_norm": 1.954642417359045,
+      "learning_rate": 6.184986609900934e-06,
+      "loss": 0.6789,
+      "step": 3220
+    },
+    {
+      "epoch": 1.9076103050044417,
+      "grad_norm": 1.837812310692102,
+      "learning_rate": 6.1790753616385e-06,
+      "loss": 0.7116,
+      "step": 3221
+    },
+    {
+      "epoch": 1.9082025466390289,
+      "grad_norm": 1.352234369694832,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.6998,
+      "step": 3222
+    },
+    {
+      "epoch": 1.9087947882736156,
+      "grad_norm": 1.353210285362207,
+      "learning_rate": 6.167257556450139e-06,
+      "loss": 0.7536,
+      "step": 3223
+    },
+    {
+      "epoch": 1.9093870299082025,
+      "grad_norm": 2.00212211868531,
+      "learning_rate": 6.16135100435836e-06,
+      "loss": 0.6977,
+      "step": 3224
+    },
+    {
+      "epoch": 1.9099792715427895,
+      "grad_norm": 1.5184121891582234,
+      "learning_rate": 6.155446022489877e-06,
+      "loss": 0.7228,
+      "step": 3225
+    },
+    {
+      "epoch": 1.9105715131773764,
+      "grad_norm": 1.7192454857712378,
+      "learning_rate": 6.149542613260157e-06,
+      "loss": 0.7029,
+      "step": 3226
+    },
+    {
+      "epoch": 1.9111637548119633,
+      "grad_norm": 0.948075216205065,
+      "learning_rate": 6.143640779084035e-06,
+      "loss": 0.6953,
+      "step": 3227
+    },
+    {
+      "epoch": 1.91175599644655,
+      "grad_norm": 1.6132188424013671,
+      "learning_rate": 6.137740522375687e-06,
+      "loss": 0.7213,
+      "step": 3228
+    },
+    {
+      "epoch": 1.9123482380811372,
+      "grad_norm": 3.072529086497967,
+      "learning_rate": 6.13184184554865e-06,
+      "loss": 0.7483,
+      "step": 3229
+    },
+    {
+      "epoch": 1.912940479715724,
+      "grad_norm": 2.1449484779319925,
+      "learning_rate": 6.1259447510158136e-06,
+      "loss": 0.7394,
+      "step": 3230
+    },
+    {
+      "epoch": 1.913532721350311,
+      "grad_norm": 1.2064970456317576,
+      "learning_rate": 6.120049241189423e-06,
+      "loss": 0.7599,
+      "step": 3231
+    },
+    {
+      "epoch": 1.9141249629848978,
+      "grad_norm": 1.5178757032575456,
+      "learning_rate": 6.114155318481076e-06,
+      "loss": 0.7253,
+      "step": 3232
+    },
+    {
+      "epoch": 1.914717204619485,
+      "grad_norm": 1.7616627788897157,
+      "learning_rate": 6.108262985301714e-06,
+      "loss": 0.713,
+      "step": 3233
+    },
+    {
+      "epoch": 1.9153094462540716,
+      "grad_norm": 1.5277812357867129,
+      "learning_rate": 6.102372244061631e-06,
+      "loss": 0.7322,
+      "step": 3234
+    },
+    {
+      "epoch": 1.9159016878886586,
+      "grad_norm": 1.241728693166342,
+      "learning_rate": 6.0964830971704755e-06,
+      "loss": 0.7061,
+      "step": 3235
+    },
+    {
+      "epoch": 1.9164939295232455,
+      "grad_norm": 8.016340143006273,
+      "learning_rate": 6.090595547037242e-06,
+      "loss": 0.7629,
+      "step": 3236
+    },
+    {
+      "epoch": 1.9170861711578324,
+      "grad_norm": 3.098312029179306,
+      "learning_rate": 6.084709596070264e-06,
+      "loss": 0.7099,
+      "step": 3237
+    },
+    {
+      "epoch": 1.9176784127924194,
+      "grad_norm": 1.570777816333163,
+      "learning_rate": 6.078825246677229e-06,
+      "loss": 0.6777,
+      "step": 3238
+    },
+    {
+      "epoch": 1.918270654427006,
+      "grad_norm": 1.1666760383543127,
+      "learning_rate": 6.072942501265164e-06,
+      "loss": 0.7347,
+      "step": 3239
+    },
+    {
+      "epoch": 1.9188628960615932,
+      "grad_norm": 1.1072653690074732,
+      "learning_rate": 6.06706136224045e-06,
+      "loss": 0.7289,
+      "step": 3240
+    },
+    {
+      "epoch": 1.91945513769618,
+      "grad_norm": 1.534029814565199,
+      "learning_rate": 6.061181832008795e-06,
+      "loss": 0.7361,
+      "step": 3241
+    },
+    {
+      "epoch": 1.920047379330767,
+      "grad_norm": 11.474053636259848,
+      "learning_rate": 6.055303912975261e-06,
+      "loss": 0.7246,
+      "step": 3242
+    },
+    {
+      "epoch": 1.9206396209653538,
+      "grad_norm": 1.1182633924025185,
+      "learning_rate": 6.049427607544247e-06,
+      "loss": 0.716,
+      "step": 3243
+    },
+    {
+      "epoch": 1.9212318625999407,
+      "grad_norm": 1.7977091796689737,
+      "learning_rate": 6.0435529181195e-06,
+      "loss": 0.691,
+      "step": 3244
+    },
+    {
+      "epoch": 1.9218241042345277,
+      "grad_norm": 1.2523074350306502,
+      "learning_rate": 6.0376798471040835e-06,
+      "loss": 0.7398,
+      "step": 3245
+    },
+    {
+      "epoch": 1.9224163458691146,
+      "grad_norm": 1.379470202336427,
+      "learning_rate": 6.031808396900422e-06,
+      "loss": 0.7037,
+      "step": 3246
+    },
+    {
+      "epoch": 1.9230085875037015,
+      "grad_norm": 1.1064953536693023,
+      "learning_rate": 6.025938569910271e-06,
+      "loss": 0.7371,
+      "step": 3247
+    },
+    {
+      "epoch": 1.9236008291382884,
+      "grad_norm": 1.3352312831811772,
+      "learning_rate": 6.020070368534719e-06,
+      "loss": 0.7153,
+      "step": 3248
+    },
+    {
+      "epoch": 1.9241930707728754,
+      "grad_norm": 6.84844144864244,
+      "learning_rate": 6.0142037951741824e-06,
+      "loss": 0.6729,
+      "step": 3249
+    },
+    {
+      "epoch": 1.924785312407462,
+      "grad_norm": 1.2959405118553768,
+      "learning_rate": 6.00833885222843e-06,
+      "loss": 0.6972,
+      "step": 3250
+    },
+    {
+      "epoch": 1.9253775540420492,
+      "grad_norm": 1.6921129797399264,
+      "learning_rate": 6.002475542096548e-06,
+      "loss": 0.7132,
+      "step": 3251
+    },
+    {
+      "epoch": 1.925969795676636,
+      "grad_norm": 1.899108534796794,
+      "learning_rate": 5.996613867176964e-06,
+      "loss": 0.7072,
+      "step": 3252
+    },
+    {
+      "epoch": 1.926562037311223,
+      "grad_norm": 1.6781426823646384,
+      "learning_rate": 5.9907538298674265e-06,
+      "loss": 0.7478,
+      "step": 3253
+    },
+    {
+      "epoch": 1.9271542789458098,
+      "grad_norm": 1.3313919012515292,
+      "learning_rate": 5.984895432565022e-06,
+      "loss": 0.7394,
+      "step": 3254
+    },
+    {
+      "epoch": 1.9277465205803967,
+      "grad_norm": 1.556101241994347,
+      "learning_rate": 5.979038677666167e-06,
+      "loss": 0.7304,
+      "step": 3255
+    },
+    {
+      "epoch": 1.9283387622149837,
+      "grad_norm": 1.434032373329083,
+      "learning_rate": 5.973183567566605e-06,
+      "loss": 0.7178,
+      "step": 3256
+    },
+    {
+      "epoch": 1.9289310038495706,
+      "grad_norm": 2.078218024535206,
+      "learning_rate": 5.967330104661402e-06,
+      "loss": 0.6953,
+      "step": 3257
+    },
+    {
+      "epoch": 1.9295232454841575,
+      "grad_norm": 1.9496636146502733,
+      "learning_rate": 5.96147829134495e-06,
+      "loss": 0.6895,
+      "step": 3258
+    },
+    {
+      "epoch": 1.9301154871187445,
+      "grad_norm": 1.7924248358414332,
+      "learning_rate": 5.955628130010977e-06,
+      "loss": 0.7637,
+      "step": 3259
+    },
+    {
+      "epoch": 1.9307077287533314,
+      "grad_norm": 1.6764468299545592,
+      "learning_rate": 5.949779623052526e-06,
+      "loss": 0.6733,
+      "step": 3260
+    },
+    {
+      "epoch": 1.9312999703879181,
+      "grad_norm": 1.3373116114928854,
+      "learning_rate": 5.9439327728619634e-06,
+      "loss": 0.6799,
+      "step": 3261
+    },
+    {
+      "epoch": 1.9318922120225053,
+      "grad_norm": 1.6119800507160023,
+      "learning_rate": 5.9380875818309805e-06,
+      "loss": 0.6576,
+      "step": 3262
+    },
+    {
+      "epoch": 1.932484453657092,
+      "grad_norm": 2.318030452874941,
+      "learning_rate": 5.932244052350585e-06,
+      "loss": 0.6927,
+      "step": 3263
+    },
+    {
+      "epoch": 1.9330766952916791,
+      "grad_norm": 1.498808583130641,
+      "learning_rate": 5.926402186811118e-06,
+      "loss": 0.6914,
+      "step": 3264
+    },
+    {
+      "epoch": 1.9336689369262658,
+      "grad_norm": 1.1247024199131759,
+      "learning_rate": 5.920561987602224e-06,
+      "loss": 0.721,
+      "step": 3265
+    },
+    {
+      "epoch": 1.9342611785608528,
+      "grad_norm": 2.2083931482923993,
+      "learning_rate": 5.914723457112877e-06,
+      "loss": 0.6962,
+      "step": 3266
+    },
+    {
+      "epoch": 1.9348534201954397,
+      "grad_norm": 2.703573235446084,
+      "learning_rate": 5.908886597731358e-06,
+      "loss": 0.6997,
+      "step": 3267
+    },
+    {
+      "epoch": 1.9354456618300266,
+      "grad_norm": 1.4305912748643144,
+      "learning_rate": 5.903051411845282e-06,
+      "loss": 0.7246,
+      "step": 3268
+    },
+    {
+      "epoch": 1.9360379034646136,
+      "grad_norm": 1.4530135063532168,
+      "learning_rate": 5.897217901841554e-06,
+      "loss": 0.6948,
+      "step": 3269
+    },
+    {
+      "epoch": 1.9366301450992005,
+      "grad_norm": 1.4638908118569964,
+      "learning_rate": 5.8913860701064175e-06,
+      "loss": 0.7293,
+      "step": 3270
+    },
+    {
+      "epoch": 1.9372223867337874,
+      "grad_norm": 1.3397680960176233,
+      "learning_rate": 5.885555919025414e-06,
+      "loss": 0.7114,
+      "step": 3271
+    },
+    {
+      "epoch": 1.9378146283683741,
+      "grad_norm": 1.4925051038532378,
+      "learning_rate": 5.879727450983412e-06,
+      "loss": 0.6961,
+      "step": 3272
+    },
+    {
+      "epoch": 1.9384068700029613,
+      "grad_norm": 1.2079008891186787,
+      "learning_rate": 5.873900668364572e-06,
+      "loss": 0.7337,
+      "step": 3273
+    },
+    {
+      "epoch": 1.938999111637548,
+      "grad_norm": 1.1990174795782194,
+      "learning_rate": 5.868075573552383e-06,
+      "loss": 0.7555,
+      "step": 3274
+    },
+    {
+      "epoch": 1.9395913532721352,
+      "grad_norm": 1.473905606977259,
+      "learning_rate": 5.862252168929632e-06,
+      "loss": 0.7142,
+      "step": 3275
+    },
+    {
+      "epoch": 1.9401835949067219,
+      "grad_norm": 2.1885900149709694,
+      "learning_rate": 5.856430456878424e-06,
+      "loss": 0.7396,
+      "step": 3276
+    },
+    {
+      "epoch": 1.9407758365413088,
+      "grad_norm": 1.9165088813619322,
+      "learning_rate": 5.850610439780158e-06,
+      "loss": 0.7055,
+      "step": 3277
+    },
+    {
+      "epoch": 1.9413680781758957,
+      "grad_norm": 1.1598467306680342,
+      "learning_rate": 5.844792120015556e-06,
+      "loss": 0.7183,
+      "step": 3278
+    },
+    {
+      "epoch": 1.9419603198104827,
+      "grad_norm": 1.4065104988302588,
+      "learning_rate": 5.838975499964636e-06,
+      "loss": 0.6838,
+      "step": 3279
+    },
+    {
+      "epoch": 1.9425525614450696,
+      "grad_norm": 1.3241034015935482,
+      "learning_rate": 5.833160582006722e-06,
+      "loss": 0.7029,
+      "step": 3280
+    },
+    {
+      "epoch": 1.9431448030796565,
+      "grad_norm": 1.5431286389019339,
+      "learning_rate": 5.827347368520444e-06,
+      "loss": 0.7218,
+      "step": 3281
+    },
+    {
+      "epoch": 1.9437370447142435,
+      "grad_norm": 1.4239447287632976,
+      "learning_rate": 5.821535861883729e-06,
+      "loss": 0.7142,
+      "step": 3282
+    },
+    {
+      "epoch": 1.9443292863488302,
+      "grad_norm": 1.4592705812164701,
+      "learning_rate": 5.815726064473812e-06,
+      "loss": 0.7701,
+      "step": 3283
+    },
+    {
+      "epoch": 1.9449215279834173,
+      "grad_norm": 1.449449394771617,
+      "learning_rate": 5.8099179786672365e-06,
+      "loss": 0.7341,
+      "step": 3284
+    },
+    {
+      "epoch": 1.945513769618004,
+      "grad_norm": 1.131051802200009,
+      "learning_rate": 5.80411160683982e-06,
+      "loss": 0.7324,
+      "step": 3285
+    },
+    {
+      "epoch": 1.9461060112525912,
+      "grad_norm": 1.2255859547243508,
+      "learning_rate": 5.798306951366701e-06,
+      "loss": 0.7226,
+      "step": 3286
+    },
+    {
+      "epoch": 1.946698252887178,
+      "grad_norm": 1.8162184207193348,
+      "learning_rate": 5.7925040146223155e-06,
+      "loss": 0.6954,
+      "step": 3287
+    },
+    {
+      "epoch": 1.9472904945217648,
+      "grad_norm": 1.2603648716863414,
+      "learning_rate": 5.786702798980388e-06,
+      "loss": 0.7035,
+      "step": 3288
+    },
+    {
+      "epoch": 1.9478827361563518,
+      "grad_norm": 2.0203946682183487,
+      "learning_rate": 5.780903306813937e-06,
+      "loss": 0.696,
+      "step": 3289
+    },
+    {
+      "epoch": 1.9484749777909387,
+      "grad_norm": 2.638924315398722,
+      "learning_rate": 5.775105540495284e-06,
+      "loss": 0.6752,
+      "step": 3290
+    },
+    {
+      "epoch": 1.9490672194255256,
+      "grad_norm": 4.143046528582363,
+      "learning_rate": 5.769309502396046e-06,
+      "loss": 0.7165,
+      "step": 3291
+    },
+    {
+      "epoch": 1.9496594610601126,
+      "grad_norm": 1.9906732930148696,
+      "learning_rate": 5.763515194887126e-06,
+      "loss": 0.7242,
+      "step": 3292
+    },
+    {
+      "epoch": 1.9502517026946995,
+      "grad_norm": 1.5769279032518237,
+      "learning_rate": 5.757722620338715e-06,
+      "loss": 0.7076,
+      "step": 3293
+    },
+    {
+      "epoch": 1.9508439443292862,
+      "grad_norm": 1.4341838760864,
+      "learning_rate": 5.751931781120308e-06,
+      "loss": 0.7051,
+      "step": 3294
+    },
+    {
+      "epoch": 1.9514361859638734,
+      "grad_norm": 1.4434054947134414,
+      "learning_rate": 5.746142679600687e-06,
+      "loss": 0.7102,
+      "step": 3295
+    },
+    {
+      "epoch": 1.95202842759846,
+      "grad_norm": 2.0084791627967924,
+      "learning_rate": 5.740355318147916e-06,
+      "loss": 0.7052,
+      "step": 3296
+    },
+    {
+      "epoch": 1.9526206692330472,
+      "grad_norm": 1.3474318765956996,
+      "learning_rate": 5.734569699129347e-06,
+      "loss": 0.7527,
+      "step": 3297
+    },
+    {
+      "epoch": 1.953212910867634,
+      "grad_norm": 1.9311183919130495,
+      "learning_rate": 5.728785824911627e-06,
+      "loss": 0.7138,
+      "step": 3298
+    },
+    {
+      "epoch": 1.9538051525022209,
+      "grad_norm": 1.8418123316642743,
+      "learning_rate": 5.723003697860692e-06,
+      "loss": 0.6989,
+      "step": 3299
+    },
+    {
+      "epoch": 1.9543973941368078,
+      "grad_norm": 1.8187532795120556,
+      "learning_rate": 5.717223320341751e-06,
+      "loss": 0.6974,
+      "step": 3300
+    },
+    {
+      "epoch": 1.9549896357713947,
+      "grad_norm": 1.355389837624177,
+      "learning_rate": 5.711444694719299e-06,
+      "loss": 0.7415,
+      "step": 3301
+    },
+    {
+      "epoch": 1.9555818774059817,
+      "grad_norm": 1.4523243460694153,
+      "learning_rate": 5.70566782335713e-06,
+      "loss": 0.7046,
+      "step": 3302
+    },
+    {
+      "epoch": 1.9561741190405686,
+      "grad_norm": 1.3091809995414065,
+      "learning_rate": 5.699892708618297e-06,
+      "loss": 0.7168,
+      "step": 3303
+    },
+    {
+      "epoch": 1.9567663606751555,
+      "grad_norm": 2.142139867997893,
+      "learning_rate": 5.69411935286516e-06,
+      "loss": 0.7549,
+      "step": 3304
+    },
+    {
+      "epoch": 1.9573586023097422,
+      "grad_norm": 2.5024110696824455,
+      "learning_rate": 5.6883477584593325e-06,
+      "loss": 0.7202,
+      "step": 3305
+    },
+    {
+      "epoch": 1.9579508439443294,
+      "grad_norm": 2.2513293693234813,
+      "learning_rate": 5.682577927761732e-06,
+      "loss": 0.7172,
+      "step": 3306
+    },
+    {
+      "epoch": 1.958543085578916,
+      "grad_norm": 1.2507660474168645,
+      "learning_rate": 5.676809863132537e-06,
+      "loss": 0.711,
+      "step": 3307
+    },
+    {
+      "epoch": 1.9591353272135033,
+      "grad_norm": 1.2539660203271004,
+      "learning_rate": 5.671043566931216e-06,
+      "loss": 0.7007,
+      "step": 3308
+    },
+    {
+      "epoch": 1.95972756884809,
+      "grad_norm": 1.9143243645347823,
+      "learning_rate": 5.665279041516501e-06,
+      "loss": 0.7193,
+      "step": 3309
+    },
+    {
+      "epoch": 1.960319810482677,
+      "grad_norm": 6.37124555621677,
+      "learning_rate": 5.659516289246414e-06,
+      "loss": 0.7501,
+      "step": 3310
+    },
+    {
+      "epoch": 1.9609120521172638,
+      "grad_norm": 1.4767774435352106,
+      "learning_rate": 5.6537553124782395e-06,
+      "loss": 0.7479,
+      "step": 3311
+    },
+    {
+      "epoch": 1.9615042937518508,
+      "grad_norm": 1.6017773415211547,
+      "learning_rate": 5.647996113568547e-06,
+      "loss": 0.7181,
+      "step": 3312
+    },
+    {
+      "epoch": 1.9620965353864377,
+      "grad_norm": 2.1133632644301747,
+      "learning_rate": 5.642238694873165e-06,
+      "loss": 0.7301,
+      "step": 3313
+    },
+    {
+      "epoch": 1.9626887770210246,
+      "grad_norm": 1.4351193282078467,
+      "learning_rate": 5.636483058747209e-06,
+      "loss": 0.6863,
+      "step": 3314
+    },
+    {
+      "epoch": 1.9632810186556116,
+      "grad_norm": 1.7463182715444514,
+      "learning_rate": 5.63072920754505e-06,
+      "loss": 0.6823,
+      "step": 3315
+    },
+    {
+      "epoch": 1.9638732602901983,
+      "grad_norm": 1.9641501791574498,
+      "learning_rate": 5.624977143620347e-06,
+      "loss": 0.7183,
+      "step": 3316
+    },
+    {
+      "epoch": 1.9644655019247854,
+      "grad_norm": 1.5984745729773937,
+      "learning_rate": 5.619226869326006e-06,
+      "loss": 0.7001,
+      "step": 3317
+    },
+    {
+      "epoch": 1.9650577435593721,
+      "grad_norm": 1.504200180863183,
+      "learning_rate": 5.613478387014223e-06,
+      "loss": 0.7457,
+      "step": 3318
+    },
+    {
+      "epoch": 1.9656499851939593,
+      "grad_norm": 1.165932682338627,
+      "learning_rate": 5.6077316990364415e-06,
+      "loss": 0.7008,
+      "step": 3319
+    },
+    {
+      "epoch": 1.966242226828546,
+      "grad_norm": 1.2741156158971194,
+      "learning_rate": 5.601986807743388e-06,
+      "loss": 0.724,
+      "step": 3320
+    },
+    {
+      "epoch": 1.966834468463133,
+      "grad_norm": 1.376216878070863,
+      "learning_rate": 5.5962437154850434e-06,
+      "loss": 0.738,
+      "step": 3321
+    },
+    {
+      "epoch": 1.9674267100977199,
+      "grad_norm": 2.0904006798570567,
+      "learning_rate": 5.5905024246106485e-06,
+      "loss": 0.7218,
+      "step": 3322
+    },
+    {
+      "epoch": 1.9680189517323068,
+      "grad_norm": 1.0737572438567344,
+      "learning_rate": 5.584762937468722e-06,
+      "loss": 0.7327,
+      "step": 3323
+    },
+    {
+      "epoch": 1.9686111933668937,
+      "grad_norm": 2.387906025872129,
+      "learning_rate": 5.579025256407038e-06,
+      "loss": 0.7739,
+      "step": 3324
+    },
+    {
+      "epoch": 1.9692034350014807,
+      "grad_norm": 4.887193226453433,
+      "learning_rate": 5.573289383772628e-06,
+      "loss": 0.7045,
+      "step": 3325
+    },
+    {
+      "epoch": 1.9697956766360676,
+      "grad_norm": 1.2124274832546234,
+      "learning_rate": 5.567555321911782e-06,
+      "loss": 0.6857,
+      "step": 3326
+    },
+    {
+      "epoch": 1.9703879182706543,
+      "grad_norm": 1.2868902005451346,
+      "learning_rate": 5.561823073170056e-06,
+      "loss": 0.7062,
+      "step": 3327
+    },
+    {
+      "epoch": 1.9709801599052414,
+      "grad_norm": 1.4155299558301717,
+      "learning_rate": 5.55609263989227e-06,
+      "loss": 0.6972,
+      "step": 3328
+    },
+    {
+      "epoch": 1.9715724015398282,
+      "grad_norm": 1.990767811294212,
+      "learning_rate": 5.550364024422486e-06,
+      "loss": 0.742,
+      "step": 3329
+    },
+    {
+      "epoch": 1.9721646431744153,
+      "grad_norm": 1.2244919516098602,
+      "learning_rate": 5.544637229104027e-06,
+      "loss": 0.7393,
+      "step": 3330
+    },
+    {
+      "epoch": 1.972756884809002,
+      "grad_norm": 1.5478990157497663,
+      "learning_rate": 5.538912256279479e-06,
+      "loss": 0.7008,
+      "step": 3331
+    },
+    {
+      "epoch": 1.973349126443589,
+      "grad_norm": 1.248386021207724,
+      "learning_rate": 5.533189108290682e-06,
+      "loss": 0.708,
+      "step": 3332
+    },
+    {
+      "epoch": 1.9739413680781759,
+      "grad_norm": 1.1494240153147253,
+      "learning_rate": 5.5274677874787195e-06,
+      "loss": 0.7127,
+      "step": 3333
+    },
+    {
+      "epoch": 1.9745336097127628,
+      "grad_norm": 2.0642350332580097,
+      "learning_rate": 5.5217482961839305e-06,
+      "loss": 0.7059,
+      "step": 3334
+    },
+    {
+      "epoch": 1.9751258513473497,
+      "grad_norm": 1.4089725977182308,
+      "learning_rate": 5.516030636745914e-06,
+      "loss": 0.6773,
+      "step": 3335
+    },
+    {
+      "epoch": 1.9757180929819367,
+      "grad_norm": 1.9085909255822204,
+      "learning_rate": 5.51031481150352e-06,
+      "loss": 0.7134,
+      "step": 3336
+    },
+    {
+      "epoch": 1.9763103346165236,
+      "grad_norm": 2.0113642440749877,
+      "learning_rate": 5.504600822794827e-06,
+      "loss": 0.7131,
+      "step": 3337
+    },
+    {
+      "epoch": 1.9769025762511103,
+      "grad_norm": 1.0613716617183546,
+      "learning_rate": 5.498888672957187e-06,
+      "loss": 0.7234,
+      "step": 3338
+    },
+    {
+      "epoch": 1.9774948178856975,
+      "grad_norm": 1.9231732798281154,
+      "learning_rate": 5.493178364327192e-06,
+      "loss": 0.7543,
+      "step": 3339
+    },
+    {
+      "epoch": 1.9780870595202842,
+      "grad_norm": 1.2516392211577612,
+      "learning_rate": 5.487469899240678e-06,
+      "loss": 0.6811,
+      "step": 3340
+    },
+    {
+      "epoch": 1.9786793011548713,
+      "grad_norm": 1.767242904928232,
+      "learning_rate": 5.481763280032722e-06,
+      "loss": 0.7349,
+      "step": 3341
+    },
+    {
+      "epoch": 1.979271542789458,
+      "grad_norm": 1.6839167459983475,
+      "learning_rate": 5.476058509037658e-06,
+      "loss": 0.7343,
+      "step": 3342
+    },
+    {
+      "epoch": 1.979863784424045,
+      "grad_norm": 2.2075665620689353,
+      "learning_rate": 5.470355588589063e-06,
+      "loss": 0.7268,
+      "step": 3343
+    },
+    {
+      "epoch": 1.980456026058632,
+      "grad_norm": 2.8495458199508636,
+      "learning_rate": 5.4646545210197435e-06,
+      "loss": 0.761,
+      "step": 3344
+    },
+    {
+      "epoch": 1.9810482676932188,
+      "grad_norm": 1.7526127261012925,
+      "learning_rate": 5.458955308661758e-06,
+      "loss": 0.7317,
+      "step": 3345
+    },
+    {
+      "epoch": 1.9816405093278058,
+      "grad_norm": 1.953204597527354,
+      "learning_rate": 5.453257953846405e-06,
+      "loss": 0.7254,
+      "step": 3346
+    },
+    {
+      "epoch": 1.9822327509623927,
+      "grad_norm": 2.6955617878160893,
+      "learning_rate": 5.447562458904227e-06,
+      "loss": 0.7285,
+      "step": 3347
+    },
+    {
+      "epoch": 1.9828249925969796,
+      "grad_norm": 1.6118188469147372,
+      "learning_rate": 5.441868826165002e-06,
+      "loss": 0.7359,
+      "step": 3348
+    },
+    {
+      "epoch": 1.9834172342315664,
+      "grad_norm": 1.9770522714718386,
+      "learning_rate": 5.436177057957739e-06,
+      "loss": 0.7308,
+      "step": 3349
+    },
+    {
+      "epoch": 1.9840094758661535,
+      "grad_norm": 1.629247494410455,
+      "learning_rate": 5.430487156610695e-06,
+      "loss": 0.7142,
+      "step": 3350
+    },
+    {
+      "epoch": 1.9846017175007402,
+      "grad_norm": 1.2839049398259734,
+      "learning_rate": 5.4247991244513635e-06,
+      "loss": 0.7203,
+      "step": 3351
+    },
+    {
+      "epoch": 1.9851939591353274,
+      "grad_norm": 1.864717196615187,
+      "learning_rate": 5.419112963806468e-06,
+      "loss": 0.6983,
+      "step": 3352
+    },
+    {
+      "epoch": 1.985786200769914,
+      "grad_norm": 1.4608471929423679,
+      "learning_rate": 5.4134286770019595e-06,
+      "loss": 0.6954,
+      "step": 3353
+    },
+    {
+      "epoch": 1.986378442404501,
+      "grad_norm": 2.368513304613487,
+      "learning_rate": 5.407746266363039e-06,
+      "loss": 0.7199,
+      "step": 3354
+    },
+    {
+      "epoch": 1.986970684039088,
+      "grad_norm": 2.3133080313204375,
+      "learning_rate": 5.402065734214135e-06,
+      "loss": 0.7527,
+      "step": 3355
+    },
+    {
+      "epoch": 1.9875629256736749,
+      "grad_norm": 1.236304503675414,
+      "learning_rate": 5.3963870828788975e-06,
+      "loss": 0.6784,
+      "step": 3356
+    },
+    {
+      "epoch": 1.9881551673082618,
+      "grad_norm": 1.6483424083328333,
+      "learning_rate": 5.390710314680214e-06,
+      "loss": 0.7044,
+      "step": 3357
+    },
+    {
+      "epoch": 1.9887474089428485,
+      "grad_norm": 6.5521490398349735,
+      "learning_rate": 5.3850354319402095e-06,
+      "loss": 0.6524,
+      "step": 3358
+    },
+    {
+      "epoch": 1.9893396505774357,
+      "grad_norm": 1.2554308783934172,
+      "learning_rate": 5.379362436980222e-06,
+      "loss": 0.7141,
+      "step": 3359
+    },
+    {
+      "epoch": 1.9899318922120224,
+      "grad_norm": 1.8973904162055975,
+      "learning_rate": 5.373691332120832e-06,
+      "loss": 0.71,
+      "step": 3360
+    },
+    {
+      "epoch": 1.9905241338466095,
+      "grad_norm": 1.3579617035026215,
+      "learning_rate": 5.368022119681833e-06,
+      "loss": 0.7219,
+      "step": 3361
+    },
+    {
+      "epoch": 1.9911163754811962,
+      "grad_norm": 3.758540120537383,
+      "learning_rate": 5.362354801982259e-06,
+      "loss": 0.7086,
+      "step": 3362
+    },
+    {
+      "epoch": 1.9917086171157834,
+      "grad_norm": 4.086844855136015,
+      "learning_rate": 5.356689381340354e-06,
+      "loss": 0.7156,
+      "step": 3363
+    },
+    {
+      "epoch": 1.99230085875037,
+      "grad_norm": 2.7000807650636185,
+      "learning_rate": 5.351025860073604e-06,
+      "loss": 0.7186,
+      "step": 3364
+    },
+    {
+      "epoch": 1.992893100384957,
+      "grad_norm": 1.302878544865362,
+      "learning_rate": 5.345364240498696e-06,
+      "loss": 0.7177,
+      "step": 3365
+    },
+    {
+      "epoch": 1.993485342019544,
+      "grad_norm": 1.341163287967209,
+      "learning_rate": 5.3397045249315615e-06,
+      "loss": 0.747,
+      "step": 3366
+    },
+    {
+      "epoch": 1.994077583654131,
+      "grad_norm": 1.8874536605468482,
+      "learning_rate": 5.334046715687334e-06,
+      "loss": 0.736,
+      "step": 3367
+    },
+    {
+      "epoch": 1.9946698252887178,
+      "grad_norm": 1.1757321216705952,
+      "learning_rate": 5.328390815080381e-06,
+      "loss": 0.707,
+      "step": 3368
+    },
+    {
+      "epoch": 1.9952620669233045,
+      "grad_norm": 2.40844448714098,
+      "learning_rate": 5.32273682542428e-06,
+      "loss": 0.7085,
+      "step": 3369
+    },
+    {
+      "epoch": 1.9958543085578917,
+      "grad_norm": 1.7338073370473768,
+      "learning_rate": 5.317084749031835e-06,
+      "loss": 0.6737,
+      "step": 3370
+    },
+    {
+      "epoch": 1.9964465501924784,
+      "grad_norm": 1.602486533397963,
+      "learning_rate": 5.311434588215057e-06,
+      "loss": 0.7334,
+      "step": 3371
+    },
+    {
+      "epoch": 1.9970387918270656,
+      "grad_norm": 1.8450498725558269,
+      "learning_rate": 5.3057863452851875e-06,
+      "loss": 0.7309,
+      "step": 3372
+    },
+    {
+      "epoch": 1.9976310334616523,
+      "grad_norm": 1.1497170574169295,
+      "learning_rate": 5.300140022552671e-06,
+      "loss": 0.7375,
+      "step": 3373
+    },
+    {
+      "epoch": 1.9982232750962394,
+      "grad_norm": 1.5753751548871004,
+      "learning_rate": 5.294495622327167e-06,
+      "loss": 0.7498,
+      "step": 3374
+    },
+    {
+      "epoch": 1.9988155167308261,
+      "grad_norm": 7.073970313653954,
+      "learning_rate": 5.288853146917557e-06,
+      "loss": 0.7538,
+      "step": 3375
+    },
+    {
+      "epoch": 1.999407758365413,
+      "grad_norm": 1.2888380404329518,
+      "learning_rate": 5.283212598631935e-06,
+      "loss": 0.7392,
+      "step": 3376
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1340654280485976,
+      "learning_rate": 5.277573979777597e-06,
+      "loss": 0.6702,
+      "step": 3377
+    },
+    {
+      "epoch": 2.0005922416345867,
+      "grad_norm": 1.1704057850103424,
+      "learning_rate": 5.271937292661054e-06,
+      "loss": 0.6327,
+      "step": 3378
+    },
+    {
+      "epoch": 2.001184483269174,
+      "grad_norm": 1.2928344272015833,
+      "learning_rate": 5.266302539588029e-06,
+      "loss": 0.651,
+      "step": 3379
+    },
+    {
+      "epoch": 2.0017767249037606,
+      "grad_norm": 1.8183856023739275,
+      "learning_rate": 5.260669722863457e-06,
+      "loss": 0.6276,
+      "step": 3380
+    },
+    {
+      "epoch": 2.0023689665383477,
+      "grad_norm": 1.232591682636013,
+      "learning_rate": 5.255038844791475e-06,
+      "loss": 0.655,
+      "step": 3381
+    },
+    {
+      "epoch": 2.0029612081729344,
+      "grad_norm": 1.439614040820058,
+      "learning_rate": 5.249409907675422e-06,
+      "loss": 0.6478,
+      "step": 3382
+    },
+    {
+      "epoch": 2.0035534498075216,
+      "grad_norm": 1.2378098354139504,
+      "learning_rate": 5.243782913817858e-06,
+      "loss": 0.5961,
+      "step": 3383
+    },
+    {
+      "epoch": 2.0041456914421083,
+      "grad_norm": 1.3329846166899064,
+      "learning_rate": 5.238157865520539e-06,
+      "loss": 0.6206,
+      "step": 3384
+    },
+    {
+      "epoch": 2.0047379330766955,
+      "grad_norm": 2.1513887798109037,
+      "learning_rate": 5.232534765084425e-06,
+      "loss": 0.6225,
+      "step": 3385
+    },
+    {
+      "epoch": 2.005330174711282,
+      "grad_norm": 1.4341250498776674,
+      "learning_rate": 5.226913614809677e-06,
+      "loss": 0.6466,
+      "step": 3386
+    },
+    {
+      "epoch": 2.0059224163458693,
+      "grad_norm": 2.65394252956428,
+      "learning_rate": 5.221294416995661e-06,
+      "loss": 0.6219,
+      "step": 3387
+    },
+    {
+      "epoch": 2.006514657980456,
+      "grad_norm": 1.939125905672404,
+      "learning_rate": 5.215677173940959e-06,
+      "loss": 0.6105,
+      "step": 3388
+    },
+    {
+      "epoch": 2.0071068996150427,
+      "grad_norm": 1.6103895073133687,
+      "learning_rate": 5.210061887943318e-06,
+      "loss": 0.5788,
+      "step": 3389
+    },
+    {
+      "epoch": 2.00769914124963,
+      "grad_norm": 5.940476923941293,
+      "learning_rate": 5.204448561299718e-06,
+      "loss": 0.608,
+      "step": 3390
+    },
+    {
+      "epoch": 2.0082913828842166,
+      "grad_norm": 3.4355880637707736,
+      "learning_rate": 5.1988371963063235e-06,
+      "loss": 0.6455,
+      "step": 3391
+    },
+    {
+      "epoch": 2.0088836245188038,
+      "grad_norm": 1.709718605173615,
+      "learning_rate": 5.193227795258505e-06,
+      "loss": 0.6509,
+      "step": 3392
+    },
+    {
+      "epoch": 2.0094758661533905,
+      "grad_norm": 1.521862412293069,
+      "learning_rate": 5.187620360450809e-06,
+      "loss": 0.6062,
+      "step": 3393
+    },
+    {
+      "epoch": 2.0100681077879776,
+      "grad_norm": 2.304842968308541,
+      "learning_rate": 5.182014894176999e-06,
+      "loss": 0.6268,
+      "step": 3394
+    },
+    {
+      "epoch": 2.0106603494225643,
+      "grad_norm": 1.6981834404747536,
+      "learning_rate": 5.176411398730028e-06,
+      "loss": 0.6083,
+      "step": 3395
+    },
+    {
+      "epoch": 2.0112525910571515,
+      "grad_norm": 1.5414486726862469,
+      "learning_rate": 5.170809876402039e-06,
+      "loss": 0.5742,
+      "step": 3396
+    },
+    {
+      "epoch": 2.011844832691738,
+      "grad_norm": 1.8324966309812314,
+      "learning_rate": 5.165210329484366e-06,
+      "loss": 0.6401,
+      "step": 3397
+    },
+    {
+      "epoch": 2.0124370743263253,
+      "grad_norm": 1.5280075849514587,
+      "learning_rate": 5.159612760267541e-06,
+      "loss": 0.612,
+      "step": 3398
+    },
+    {
+      "epoch": 2.013029315960912,
+      "grad_norm": 1.7053593517597887,
+      "learning_rate": 5.154017171041289e-06,
+      "loss": 0.6234,
+      "step": 3399
+    },
+    {
+      "epoch": 2.0136215575954988,
+      "grad_norm": 2.384672543167793,
+      "learning_rate": 5.148423564094517e-06,
+      "loss": 0.5944,
+      "step": 3400
+    },
+    {
+      "epoch": 2.014213799230086,
+      "grad_norm": 1.7192733690837867,
+      "learning_rate": 5.142831941715321e-06,
+      "loss": 0.6438,
+      "step": 3401
+    },
+    {
+      "epoch": 2.0148060408646726,
+      "grad_norm": 1.2298645317583594,
+      "learning_rate": 5.137242306190991e-06,
+      "loss": 0.6059,
+      "step": 3402
+    },
+    {
+      "epoch": 2.01539828249926,
+      "grad_norm": 1.421035277118102,
+      "learning_rate": 5.13165465980801e-06,
+      "loss": 0.6251,
+      "step": 3403
+    },
+    {
+      "epoch": 2.0159905241338465,
+      "grad_norm": 1.3393515437383583,
+      "learning_rate": 5.126069004852033e-06,
+      "loss": 0.6224,
+      "step": 3404
+    },
+    {
+      "epoch": 2.0165827657684336,
+      "grad_norm": 1.7027652434897638,
+      "learning_rate": 5.120485343607901e-06,
+      "loss": 0.6055,
+      "step": 3405
+    },
+    {
+      "epoch": 2.0171750074030204,
+      "grad_norm": 1.1354965871126625,
+      "learning_rate": 5.114903678359655e-06,
+      "loss": 0.6418,
+      "step": 3406
+    },
+    {
+      "epoch": 2.0177672490376075,
+      "grad_norm": 2.17653820273378,
+      "learning_rate": 5.10932401139051e-06,
+      "loss": 0.6267,
+      "step": 3407
+    },
+    {
+      "epoch": 2.0183594906721942,
+      "grad_norm": 1.1887528683987758,
+      "learning_rate": 5.103746344982859e-06,
+      "loss": 0.603,
+      "step": 3408
+    },
+    {
+      "epoch": 2.0189517323067814,
+      "grad_norm": 1.2166339620141267,
+      "learning_rate": 5.09817068141828e-06,
+      "loss": 0.6271,
+      "step": 3409
+    },
+    {
+      "epoch": 2.019543973941368,
+      "grad_norm": 1.4946678433457767,
+      "learning_rate": 5.092597022977539e-06,
+      "loss": 0.6135,
+      "step": 3410
+    },
+    {
+      "epoch": 2.020136215575955,
+      "grad_norm": 2.181670296560735,
+      "learning_rate": 5.087025371940568e-06,
+      "loss": 0.5295,
+      "step": 3411
+    },
+    {
+      "epoch": 2.020728457210542,
+      "grad_norm": 1.1109403860055909,
+      "learning_rate": 5.081455730586495e-06,
+      "loss": 0.6452,
+      "step": 3412
+    },
+    {
+      "epoch": 2.0213206988451287,
+      "grad_norm": 1.5294096669192148,
+      "learning_rate": 5.075888101193605e-06,
+      "loss": 0.6353,
+      "step": 3413
+    },
+    {
+      "epoch": 2.021912940479716,
+      "grad_norm": 1.2375562922087688,
+      "learning_rate": 5.070322486039383e-06,
+      "loss": 0.6426,
+      "step": 3414
+    },
+    {
+      "epoch": 2.0225051821143025,
+      "grad_norm": 1.1958259044194008,
+      "learning_rate": 5.06475888740047e-06,
+      "loss": 0.6161,
+      "step": 3415
+    },
+    {
+      "epoch": 2.0230974237488897,
+      "grad_norm": 1.3364125839027878,
+      "learning_rate": 5.059197307552698e-06,
+      "loss": 0.6045,
+      "step": 3416
+    },
+    {
+      "epoch": 2.0236896653834764,
+      "grad_norm": 1.471484617701428,
+      "learning_rate": 5.053637748771058e-06,
+      "loss": 0.6244,
+      "step": 3417
+    },
+    {
+      "epoch": 2.0242819070180635,
+      "grad_norm": 1.9697646130426174,
+      "learning_rate": 5.048080213329729e-06,
+      "loss": 0.6303,
+      "step": 3418
+    },
+    {
+      "epoch": 2.0248741486526503,
+      "grad_norm": 1.2533508619798654,
+      "learning_rate": 5.04252470350205e-06,
+      "loss": 0.5929,
+      "step": 3419
+    },
+    {
+      "epoch": 2.0254663902872374,
+      "grad_norm": 1.249684000829904,
+      "learning_rate": 5.036971221560543e-06,
+      "loss": 0.6563,
+      "step": 3420
+    },
+    {
+      "epoch": 2.026058631921824,
+      "grad_norm": 1.7695633483618758,
+      "learning_rate": 5.031419769776887e-06,
+      "loss": 0.6099,
+      "step": 3421
+    },
+    {
+      "epoch": 2.026650873556411,
+      "grad_norm": 1.5156337584508315,
+      "learning_rate": 5.025870350421945e-06,
+      "loss": 0.6385,
+      "step": 3422
+    },
+    {
+      "epoch": 2.027243115190998,
+      "grad_norm": 1.1787473479254456,
+      "learning_rate": 5.020322965765736e-06,
+      "loss": 0.6373,
+      "step": 3423
+    },
+    {
+      "epoch": 2.0278353568255847,
+      "grad_norm": 1.2307377200251604,
+      "learning_rate": 5.0147776180774575e-06,
+      "loss": 0.6469,
+      "step": 3424
+    },
+    {
+      "epoch": 2.028427598460172,
+      "grad_norm": 1.2005680095700468,
+      "learning_rate": 5.009234309625467e-06,
+      "loss": 0.5903,
+      "step": 3425
+    },
+    {
+      "epoch": 2.0290198400947586,
+      "grad_norm": 1.8848824259084624,
+      "learning_rate": 5.003693042677283e-06,
+      "loss": 0.5888,
+      "step": 3426
+    },
+    {
+      "epoch": 2.0296120817293457,
+      "grad_norm": 1.4702739853724482,
+      "learning_rate": 4.998153819499601e-06,
+      "loss": 0.5864,
+      "step": 3427
+    },
+    {
+      "epoch": 2.0302043233639324,
+      "grad_norm": 1.3620518239621033,
+      "learning_rate": 4.992616642358279e-06,
+      "loss": 0.617,
+      "step": 3428
+    },
+    {
+      "epoch": 2.0307965649985196,
+      "grad_norm": 1.1287181683000989,
+      "learning_rate": 4.9870815135183285e-06,
+      "loss": 0.605,
+      "step": 3429
+    },
+    {
+      "epoch": 2.0313888066331063,
+      "grad_norm": 1.8477536160925232,
+      "learning_rate": 4.9815484352439244e-06,
+      "loss": 0.6376,
+      "step": 3430
+    },
+    {
+      "epoch": 2.0319810482676934,
+      "grad_norm": 1.2689842522920158,
+      "learning_rate": 4.976017409798413e-06,
+      "loss": 0.5849,
+      "step": 3431
+    },
+    {
+      "epoch": 2.03257328990228,
+      "grad_norm": 2.2267139790954498,
+      "learning_rate": 4.970488439444296e-06,
+      "loss": 0.6276,
+      "step": 3432
+    },
+    {
+      "epoch": 2.033165531536867,
+      "grad_norm": 2.263989349996733,
+      "learning_rate": 4.964961526443231e-06,
+      "loss": 0.6187,
+      "step": 3433
+    },
+    {
+      "epoch": 2.033757773171454,
+      "grad_norm": 1.1002565279108922,
+      "learning_rate": 4.95943667305603e-06,
+      "loss": 0.6165,
+      "step": 3434
+    },
+    {
+      "epoch": 2.0343500148060407,
+      "grad_norm": 1.1437907896314625,
+      "learning_rate": 4.953913881542677e-06,
+      "loss": 0.6687,
+      "step": 3435
+    },
+    {
+      "epoch": 2.034942256440628,
+      "grad_norm": 1.4291661707165033,
+      "learning_rate": 4.948393154162303e-06,
+      "loss": 0.6324,
+      "step": 3436
+    },
+    {
+      "epoch": 2.0355344980752146,
+      "grad_norm": 1.2447703838850679,
+      "learning_rate": 4.9428744931731965e-06,
+      "loss": 0.5916,
+      "step": 3437
+    },
+    {
+      "epoch": 2.0361267397098017,
+      "grad_norm": 1.3396172149707037,
+      "learning_rate": 4.937357900832793e-06,
+      "loss": 0.6117,
+      "step": 3438
+    },
+    {
+      "epoch": 2.0367189813443884,
+      "grad_norm": 1.327487438241329,
+      "learning_rate": 4.931843379397695e-06,
+      "loss": 0.6275,
+      "step": 3439
+    },
+    {
+      "epoch": 2.0373112229789756,
+      "grad_norm": 5.634667309355351,
+      "learning_rate": 4.926330931123659e-06,
+      "loss": 0.5808,
+      "step": 3440
+    },
+    {
+      "epoch": 2.0379034646135623,
+      "grad_norm": 1.712596745761093,
+      "learning_rate": 4.920820558265569e-06,
+      "loss": 0.6298,
+      "step": 3441
+    },
+    {
+      "epoch": 2.038495706248149,
+      "grad_norm": 1.6350123351542276,
+      "learning_rate": 4.915312263077488e-06,
+      "loss": 0.6246,
+      "step": 3442
+    },
+    {
+      "epoch": 2.039087947882736,
+      "grad_norm": 1.4377006668550185,
+      "learning_rate": 4.909806047812617e-06,
+      "loss": 0.6212,
+      "step": 3443
+    },
+    {
+      "epoch": 2.039680189517323,
+      "grad_norm": 1.3626893907998683,
+      "learning_rate": 4.904301914723315e-06,
+      "loss": 0.5958,
+      "step": 3444
+    },
+    {
+      "epoch": 2.04027243115191,
+      "grad_norm": 1.881609654716602,
+      "learning_rate": 4.898799866061068e-06,
+      "loss": 0.6389,
+      "step": 3445
+    },
+    {
+      "epoch": 2.0408646727864967,
+      "grad_norm": 1.0881012049516798,
+      "learning_rate": 4.89329990407653e-06,
+      "loss": 0.6571,
+      "step": 3446
+    },
+    {
+      "epoch": 2.041456914421084,
+      "grad_norm": 1.4033919456100592,
+      "learning_rate": 4.887802031019498e-06,
+      "loss": 0.631,
+      "step": 3447
+    },
+    {
+      "epoch": 2.0420491560556706,
+      "grad_norm": 3.113208111035224,
+      "learning_rate": 4.882306249138909e-06,
+      "loss": 0.6045,
+      "step": 3448
+    },
+    {
+      "epoch": 2.0426413976902578,
+      "grad_norm": 1.2724517686667787,
+      "learning_rate": 4.876812560682842e-06,
+      "loss": 0.6156,
+      "step": 3449
+    },
+    {
+      "epoch": 2.0432336393248445,
+      "grad_norm": 1.472100349180877,
+      "learning_rate": 4.871320967898528e-06,
+      "loss": 0.6171,
+      "step": 3450
+    },
+    {
+      "epoch": 2.0438258809594316,
+      "grad_norm": 1.7554861202098981,
+      "learning_rate": 4.865831473032342e-06,
+      "loss": 0.6485,
+      "step": 3451
+    },
+    {
+      "epoch": 2.0444181225940183,
+      "grad_norm": 0.892915766556675,
+      "learning_rate": 4.860344078329791e-06,
+      "loss": 0.5951,
+      "step": 3452
+    },
+    {
+      "epoch": 2.045010364228605,
+      "grad_norm": 1.2477613762328605,
+      "learning_rate": 4.8548587860355255e-06,
+      "loss": 0.6422,
+      "step": 3453
+    },
+    {
+      "epoch": 2.045602605863192,
+      "grad_norm": 1.3476862158178415,
+      "learning_rate": 4.849375598393342e-06,
+      "loss": 0.6421,
+      "step": 3454
+    },
+    {
+      "epoch": 2.046194847497779,
+      "grad_norm": 1.3151784488318263,
+      "learning_rate": 4.843894517646176e-06,
+      "loss": 0.6417,
+      "step": 3455
+    },
+    {
+      "epoch": 2.046787089132366,
+      "grad_norm": 2.0867879393200934,
+      "learning_rate": 4.838415546036095e-06,
+      "loss": 0.6006,
+      "step": 3456
+    },
+    {
+      "epoch": 2.0473793307669528,
+      "grad_norm": 1.528214198655404,
+      "learning_rate": 4.8329386858043005e-06,
+      "loss": 0.6131,
+      "step": 3457
+    },
+    {
+      "epoch": 2.04797157240154,
+      "grad_norm": 1.6703970326441546,
+      "learning_rate": 4.827463939191141e-06,
+      "loss": 0.6009,
+      "step": 3458
+    },
+    {
+      "epoch": 2.0485638140361266,
+      "grad_norm": 1.084600846636688,
+      "learning_rate": 4.821991308436102e-06,
+      "loss": 0.5974,
+      "step": 3459
+    },
+    {
+      "epoch": 2.049156055670714,
+      "grad_norm": 1.4736165984007512,
+      "learning_rate": 4.816520795777789e-06,
+      "loss": 0.5891,
+      "step": 3460
+    },
+    {
+      "epoch": 2.0497482973053005,
+      "grad_norm": 1.4108193712127584,
+      "learning_rate": 4.811052403453949e-06,
+      "loss": 0.5992,
+      "step": 3461
+    },
+    {
+      "epoch": 2.0503405389398877,
+      "grad_norm": 1.2902478597441098,
+      "learning_rate": 4.805586133701468e-06,
+      "loss": 0.652,
+      "step": 3462
+    },
+    {
+      "epoch": 2.0509327805744744,
+      "grad_norm": 1.4732401085829676,
+      "learning_rate": 4.800121988756352e-06,
+      "loss": 0.6209,
+      "step": 3463
+    },
+    {
+      "epoch": 2.051525022209061,
+      "grad_norm": 1.7492854759727523,
+      "learning_rate": 4.7946599708537485e-06,
+      "loss": 0.5899,
+      "step": 3464
+    },
+    {
+      "epoch": 2.0521172638436482,
+      "grad_norm": 1.3793707141008569,
+      "learning_rate": 4.789200082227924e-06,
+      "loss": 0.6281,
+      "step": 3465
+    },
+    {
+      "epoch": 2.052709505478235,
+      "grad_norm": 1.5203850237041459,
+      "learning_rate": 4.783742325112286e-06,
+      "loss": 0.6517,
+      "step": 3466
+    },
+    {
+      "epoch": 2.053301747112822,
+      "grad_norm": 1.5248222664948767,
+      "learning_rate": 4.7782867017393585e-06,
+      "loss": 0.6476,
+      "step": 3467
+    },
+    {
+      "epoch": 2.053893988747409,
+      "grad_norm": 1.2625856377787512,
+      "learning_rate": 4.772833214340805e-06,
+      "loss": 0.6413,
+      "step": 3468
+    },
+    {
+      "epoch": 2.054486230381996,
+      "grad_norm": 1.2851858362812827,
+      "learning_rate": 4.7673818651474e-06,
+      "loss": 0.629,
+      "step": 3469
+    },
+    {
+      "epoch": 2.0550784720165827,
+      "grad_norm": 1.296157068335424,
+      "learning_rate": 4.761932656389061e-06,
+      "loss": 0.6189,
+      "step": 3470
+    },
+    {
+      "epoch": 2.05567071365117,
+      "grad_norm": 1.4886765308566954,
+      "learning_rate": 4.756485590294813e-06,
+      "loss": 0.6472,
+      "step": 3471
+    },
+    {
+      "epoch": 2.0562629552857565,
+      "grad_norm": 1.1373661307548197,
+      "learning_rate": 4.751040669092819e-06,
+      "loss": 0.5558,
+      "step": 3472
+    },
+    {
+      "epoch": 2.0568551969203437,
+      "grad_norm": 1.3633758311544337,
+      "learning_rate": 4.745597895010351e-06,
+      "loss": 0.604,
+      "step": 3473
+    },
+    {
+      "epoch": 2.0574474385549304,
+      "grad_norm": 1.1929108488372469,
+      "learning_rate": 4.740157270273816e-06,
+      "loss": 0.6514,
+      "step": 3474
+    },
+    {
+      "epoch": 2.058039680189517,
+      "grad_norm": 3.1263654625674975,
+      "learning_rate": 4.7347187971087294e-06,
+      "loss": 0.6156,
+      "step": 3475
+    },
+    {
+      "epoch": 2.0586319218241043,
+      "grad_norm": 2.327298864991031,
+      "learning_rate": 4.729282477739741e-06,
+      "loss": 0.6476,
+      "step": 3476
+    },
+    {
+      "epoch": 2.059224163458691,
+      "grad_norm": 1.5846789608099434,
+      "learning_rate": 4.723848314390604e-06,
+      "loss": 0.6529,
+      "step": 3477
+    },
+    {
+      "epoch": 2.059816405093278,
+      "grad_norm": 1.4670029209129003,
+      "learning_rate": 4.718416309284196e-06,
+      "loss": 0.6004,
+      "step": 3478
+    },
+    {
+      "epoch": 2.060408646727865,
+      "grad_norm": 4.684220098207152,
+      "learning_rate": 4.712986464642515e-06,
+      "loss": 0.5856,
+      "step": 3479
+    },
+    {
+      "epoch": 2.061000888362452,
+      "grad_norm": 1.0163949128750598,
+      "learning_rate": 4.707558782686677e-06,
+      "loss": 0.5652,
+      "step": 3480
+    },
+    {
+      "epoch": 2.0615931299970387,
+      "grad_norm": 1.9902642913838888,
+      "learning_rate": 4.702133265636905e-06,
+      "loss": 0.595,
+      "step": 3481
+    },
+    {
+      "epoch": 2.062185371631626,
+      "grad_norm": 0.976503070725885,
+      "learning_rate": 4.6967099157125384e-06,
+      "loss": 0.5816,
+      "step": 3482
+    },
+    {
+      "epoch": 2.0627776132662126,
+      "grad_norm": 1.1674854697438484,
+      "learning_rate": 4.6912887351320336e-06,
+      "loss": 0.6192,
+      "step": 3483
+    },
+    {
+      "epoch": 2.0633698549007997,
+      "grad_norm": 1.080121092579032,
+      "learning_rate": 4.685869726112963e-06,
+      "loss": 0.5851,
+      "step": 3484
+    },
+    {
+      "epoch": 2.0639620965353864,
+      "grad_norm": 1.1715927470368515,
+      "learning_rate": 4.680452890872003e-06,
+      "loss": 0.6208,
+      "step": 3485
+    },
+    {
+      "epoch": 2.064554338169973,
+      "grad_norm": 2.3721843450455093,
+      "learning_rate": 4.675038231624939e-06,
+      "loss": 0.6333,
+      "step": 3486
+    },
+    {
+      "epoch": 2.0651465798045603,
+      "grad_norm": 1.4695028389780564,
+      "learning_rate": 4.669625750586675e-06,
+      "loss": 0.5885,
+      "step": 3487
+    },
+    {
+      "epoch": 2.065738821439147,
+      "grad_norm": 6.3301735183432,
+      "learning_rate": 4.664215449971225e-06,
+      "loss": 0.6118,
+      "step": 3488
+    },
+    {
+      "epoch": 2.066331063073734,
+      "grad_norm": 0.9900749638912005,
+      "learning_rate": 4.658807331991702e-06,
+      "loss": 0.5987,
+      "step": 3489
+    },
+    {
+      "epoch": 2.066923304708321,
+      "grad_norm": 1.1137652898298571,
+      "learning_rate": 4.653401398860324e-06,
+      "loss": 0.6017,
+      "step": 3490
+    },
+    {
+      "epoch": 2.067515546342908,
+      "grad_norm": 1.2870621869899241,
+      "learning_rate": 4.64799765278843e-06,
+      "loss": 0.6256,
+      "step": 3491
+    },
+    {
+      "epoch": 2.0681077879774947,
+      "grad_norm": 1.8663809342938509,
+      "learning_rate": 4.6425960959864556e-06,
+      "loss": 0.6282,
+      "step": 3492
+    },
+    {
+      "epoch": 2.068700029612082,
+      "grad_norm": 1.1936625573372333,
+      "learning_rate": 4.637196730663941e-06,
+      "loss": 0.6056,
+      "step": 3493
+    },
+    {
+      "epoch": 2.0692922712466686,
+      "grad_norm": 1.1774196803926553,
+      "learning_rate": 4.631799559029524e-06,
+      "loss": 0.6547,
+      "step": 3494
+    },
+    {
+      "epoch": 2.0698845128812557,
+      "grad_norm": 1.0847314619442951,
+      "learning_rate": 4.626404583290956e-06,
+      "loss": 0.6047,
+      "step": 3495
+    },
+    {
+      "epoch": 2.0704767545158425,
+      "grad_norm": 1.8732417061022584,
+      "learning_rate": 4.621011805655093e-06,
+      "loss": 0.6184,
+      "step": 3496
+    },
+    {
+      "epoch": 2.071068996150429,
+      "grad_norm": 1.0566441622558125,
+      "learning_rate": 4.615621228327869e-06,
+      "loss": 0.6037,
+      "step": 3497
+    },
+    {
+      "epoch": 2.0716612377850163,
+      "grad_norm": 1.4098753153815882,
+      "learning_rate": 4.61023285351434e-06,
+      "loss": 0.6616,
+      "step": 3498
+    },
+    {
+      "epoch": 2.072253479419603,
+      "grad_norm": 1.0888925606822704,
+      "learning_rate": 4.60484668341866e-06,
+      "loss": 0.6057,
+      "step": 3499
+    },
+    {
+      "epoch": 2.07284572105419,
+      "grad_norm": 1.6235022191865385,
+      "learning_rate": 4.599462720244071e-06,
+      "loss": 0.615,
+      "step": 3500
+    },
+    {
+      "epoch": 2.073437962688777,
+      "grad_norm": 1.268826339976644,
+      "learning_rate": 4.594080966192912e-06,
+      "loss": 0.6819,
+      "step": 3501
+    },
+    {
+      "epoch": 2.074030204323364,
+      "grad_norm": 1.188326581207744,
+      "learning_rate": 4.5887014234666275e-06,
+      "loss": 0.5907,
+      "step": 3502
+    },
+    {
+      "epoch": 2.0746224459579508,
+      "grad_norm": 1.6688453348930554,
+      "learning_rate": 4.583324094265757e-06,
+      "loss": 0.6598,
+      "step": 3503
+    },
+    {
+      "epoch": 2.075214687592538,
+      "grad_norm": 1.3153998492810253,
+      "learning_rate": 4.577948980789924e-06,
+      "loss": 0.6265,
+      "step": 3504
+    },
+    {
+      "epoch": 2.0758069292271246,
+      "grad_norm": 1.2858231765513355,
+      "learning_rate": 4.572576085237853e-06,
+      "loss": 0.5932,
+      "step": 3505
+    },
+    {
+      "epoch": 2.0763991708617118,
+      "grad_norm": 0.9882491735932337,
+      "learning_rate": 4.567205409807362e-06,
+      "loss": 0.5705,
+      "step": 3506
+    },
+    {
+      "epoch": 2.0769914124962985,
+      "grad_norm": 1.731064255571183,
+      "learning_rate": 4.561836956695362e-06,
+      "loss": 0.6653,
+      "step": 3507
+    },
+    {
+      "epoch": 2.077583654130885,
+      "grad_norm": 2.457969135297415,
+      "learning_rate": 4.556470728097849e-06,
+      "loss": 0.6452,
+      "step": 3508
+    },
+    {
+      "epoch": 2.0781758957654723,
+      "grad_norm": 1.9842081906939593,
+      "learning_rate": 4.55110672620991e-06,
+      "loss": 0.6041,
+      "step": 3509
+    },
+    {
+      "epoch": 2.078768137400059,
+      "grad_norm": 1.9749587351865325,
+      "learning_rate": 4.545744953225726e-06,
+      "loss": 0.5994,
+      "step": 3510
+    },
+    {
+      "epoch": 2.079360379034646,
+      "grad_norm": 3.1505013963136412,
+      "learning_rate": 4.540385411338567e-06,
+      "loss": 0.6332,
+      "step": 3511
+    },
+    {
+      "epoch": 2.079952620669233,
+      "grad_norm": 1.275314172819857,
+      "learning_rate": 4.535028102740785e-06,
+      "loss": 0.6609,
+      "step": 3512
+    },
+    {
+      "epoch": 2.08054486230382,
+      "grad_norm": 1.4137589146456349,
+      "learning_rate": 4.529673029623815e-06,
+      "loss": 0.626,
+      "step": 3513
+    },
+    {
+      "epoch": 2.081137103938407,
+      "grad_norm": 3.083605720052062,
+      "learning_rate": 4.524320194178189e-06,
+      "loss": 0.607,
+      "step": 3514
+    },
+    {
+      "epoch": 2.081729345572994,
+      "grad_norm": 4.411371237327307,
+      "learning_rate": 4.518969598593515e-06,
+      "loss": 0.6471,
+      "step": 3515
+    },
+    {
+      "epoch": 2.0823215872075806,
+      "grad_norm": 1.3266195802580654,
+      "learning_rate": 4.5136212450584895e-06,
+      "loss": 0.6135,
+      "step": 3516
+    },
+    {
+      "epoch": 2.082913828842168,
+      "grad_norm": 1.4653968395659105,
+      "learning_rate": 4.508275135760887e-06,
+      "loss": 0.6175,
+      "step": 3517
+    },
+    {
+      "epoch": 2.0835060704767545,
+      "grad_norm": 0.990190635939963,
+      "learning_rate": 4.502931272887572e-06,
+      "loss": 0.6221,
+      "step": 3518
+    },
+    {
+      "epoch": 2.0840983121113412,
+      "grad_norm": 1.8306688688850197,
+      "learning_rate": 4.497589658624477e-06,
+      "loss": 0.5972,
+      "step": 3519
+    },
+    {
+      "epoch": 2.0846905537459284,
+      "grad_norm": 1.2748246161237164,
+      "learning_rate": 4.492250295156632e-06,
+      "loss": 0.6064,
+      "step": 3520
+    },
+    {
+      "epoch": 2.085282795380515,
+      "grad_norm": 3.1717060924349134,
+      "learning_rate": 4.486913184668127e-06,
+      "loss": 0.6155,
+      "step": 3521
+    },
+    {
+      "epoch": 2.0858750370151022,
+      "grad_norm": 1.6215550053410732,
+      "learning_rate": 4.481578329342149e-06,
+      "loss": 0.6155,
+      "step": 3522
+    },
+    {
+      "epoch": 2.086467278649689,
+      "grad_norm": 1.8412996687036585,
+      "learning_rate": 4.476245731360947e-06,
+      "loss": 0.6175,
+      "step": 3523
+    },
+    {
+      "epoch": 2.087059520284276,
+      "grad_norm": 1.6036947552013425,
+      "learning_rate": 4.470915392905862e-06,
+      "loss": 0.6359,
+      "step": 3524
+    },
+    {
+      "epoch": 2.087651761918863,
+      "grad_norm": 1.2185010704003587,
+      "learning_rate": 4.465587316157291e-06,
+      "loss": 0.648,
+      "step": 3525
+    },
+    {
+      "epoch": 2.08824400355345,
+      "grad_norm": 2.117675570470455,
+      "learning_rate": 4.460261503294726e-06,
+      "loss": 0.5918,
+      "step": 3526
+    },
+    {
+      "epoch": 2.0888362451880367,
+      "grad_norm": 1.26985997913265,
+      "learning_rate": 4.4549379564967174e-06,
+      "loss": 0.6935,
+      "step": 3527
+    },
+    {
+      "epoch": 2.089428486822624,
+      "grad_norm": 1.421007208158808,
+      "learning_rate": 4.449616677940904e-06,
+      "loss": 0.5882,
+      "step": 3528
+    },
+    {
+      "epoch": 2.0900207284572105,
+      "grad_norm": 2.5687635296741846,
+      "learning_rate": 4.444297669803981e-06,
+      "loss": 0.5908,
+      "step": 3529
+    },
+    {
+      "epoch": 2.0906129700917973,
+      "grad_norm": 1.6681381199853806,
+      "learning_rate": 4.4389809342617195e-06,
+      "loss": 0.6284,
+      "step": 3530
+    },
+    {
+      "epoch": 2.0912052117263844,
+      "grad_norm": 1.6690216078536573,
+      "learning_rate": 4.433666473488971e-06,
+      "loss": 0.6287,
+      "step": 3531
+    },
+    {
+      "epoch": 2.091797453360971,
+      "grad_norm": 1.231182818609419,
+      "learning_rate": 4.428354289659641e-06,
+      "loss": 0.5921,
+      "step": 3532
+    },
+    {
+      "epoch": 2.0923896949955583,
+      "grad_norm": 1.1422701274620846,
+      "learning_rate": 4.423044384946719e-06,
+      "loss": 0.5949,
+      "step": 3533
+    },
+    {
+      "epoch": 2.092981936630145,
+      "grad_norm": 1.5358988678809373,
+      "learning_rate": 4.417736761522249e-06,
+      "loss": 0.5869,
+      "step": 3534
+    },
+    {
+      "epoch": 2.093574178264732,
+      "grad_norm": 1.8994263747170306,
+      "learning_rate": 4.412431421557351e-06,
+      "loss": 0.6305,
+      "step": 3535
+    },
+    {
+      "epoch": 2.094166419899319,
+      "grad_norm": 1.0574749107241654,
+      "learning_rate": 4.407128367222203e-06,
+      "loss": 0.5842,
+      "step": 3536
+    },
+    {
+      "epoch": 2.094758661533906,
+      "grad_norm": 1.756684625648516,
+      "learning_rate": 4.401827600686059e-06,
+      "loss": 0.6044,
+      "step": 3537
+    },
+    {
+      "epoch": 2.0953509031684927,
+      "grad_norm": 1.1516953678691502,
+      "learning_rate": 4.396529124117223e-06,
+      "loss": 0.6213,
+      "step": 3538
+    },
+    {
+      "epoch": 2.09594314480308,
+      "grad_norm": 1.208360469786924,
+      "learning_rate": 4.391232939683077e-06,
+      "loss": 0.6195,
+      "step": 3539
+    },
+    {
+      "epoch": 2.0965353864376666,
+      "grad_norm": 1.7581006133683124,
+      "learning_rate": 4.38593904955005e-06,
+      "loss": 0.609,
+      "step": 3540
+    },
+    {
+      "epoch": 2.0971276280722533,
+      "grad_norm": 1.370855782157487,
+      "learning_rate": 4.380647455883651e-06,
+      "loss": 0.5826,
+      "step": 3541
+    },
+    {
+      "epoch": 2.0977198697068404,
+      "grad_norm": 2.388671548513241,
+      "learning_rate": 4.37535816084843e-06,
+      "loss": 0.6563,
+      "step": 3542
+    },
+    {
+      "epoch": 2.098312111341427,
+      "grad_norm": 1.6360347895846583,
+      "learning_rate": 4.3700711666080135e-06,
+      "loss": 0.6808,
+      "step": 3543
+    },
+    {
+      "epoch": 2.0989043529760143,
+      "grad_norm": 1.4588269440103798,
+      "learning_rate": 4.364786475325072e-06,
+      "loss": 0.6258,
+      "step": 3544
+    },
+    {
+      "epoch": 2.099496594610601,
+      "grad_norm": 1.4736283479225063,
+      "learning_rate": 4.359504089161351e-06,
+      "loss": 0.6395,
+      "step": 3545
+    },
+    {
+      "epoch": 2.100088836245188,
+      "grad_norm": 1.1782359902227395,
+      "learning_rate": 4.354224010277632e-06,
+      "loss": 0.6292,
+      "step": 3546
+    },
+    {
+      "epoch": 2.100681077879775,
+      "grad_norm": 1.1179803013332563,
+      "learning_rate": 4.348946240833774e-06,
+      "loss": 0.5898,
+      "step": 3547
+    },
+    {
+      "epoch": 2.101273319514362,
+      "grad_norm": 1.4815157595037098,
+      "learning_rate": 4.343670782988679e-06,
+      "loss": 0.6271,
+      "step": 3548
+    },
+    {
+      "epoch": 2.1018655611489487,
+      "grad_norm": 1.291616988218737,
+      "learning_rate": 4.338397638900301e-06,
+      "loss": 0.5947,
+      "step": 3549
+    },
+    {
+      "epoch": 2.102457802783536,
+      "grad_norm": 1.2765987298770318,
+      "learning_rate": 4.333126810725655e-06,
+      "loss": 0.5707,
+      "step": 3550
+    },
+    {
+      "epoch": 2.1030500444181226,
+      "grad_norm": 1.375312819341615,
+      "learning_rate": 4.32785830062081e-06,
+      "loss": 0.6335,
+      "step": 3551
+    },
+    {
+      "epoch": 2.1036422860527093,
+      "grad_norm": 1.6256621838008019,
+      "learning_rate": 4.322592110740882e-06,
+      "loss": 0.6424,
+      "step": 3552
+    },
+    {
+      "epoch": 2.1042345276872965,
+      "grad_norm": 1.289927765078695,
+      "learning_rate": 4.317328243240032e-06,
+      "loss": 0.5974,
+      "step": 3553
+    },
+    {
+      "epoch": 2.104826769321883,
+      "grad_norm": 1.2729987598808552,
+      "learning_rate": 4.312066700271483e-06,
+      "loss": 0.6065,
+      "step": 3554
+    },
+    {
+      "epoch": 2.1054190109564703,
+      "grad_norm": 1.5025651892089609,
+      "learning_rate": 4.306807483987505e-06,
+      "loss": 0.6541,
+      "step": 3555
+    },
+    {
+      "epoch": 2.106011252591057,
+      "grad_norm": 4.027247679133015,
+      "learning_rate": 4.30155059653941e-06,
+      "loss": 0.6414,
+      "step": 3556
+    },
+    {
+      "epoch": 2.106603494225644,
+      "grad_norm": 1.3103868133396175,
+      "learning_rate": 4.296296040077557e-06,
+      "loss": 0.6134,
+      "step": 3557
+    },
+    {
+      "epoch": 2.107195735860231,
+      "grad_norm": 1.6626367225184855,
+      "learning_rate": 4.291043816751357e-06,
+      "loss": 0.5832,
+      "step": 3558
+    },
+    {
+      "epoch": 2.107787977494818,
+      "grad_norm": 1.8842126011304945,
+      "learning_rate": 4.285793928709271e-06,
+      "loss": 0.6293,
+      "step": 3559
+    },
+    {
+      "epoch": 2.1083802191294048,
+      "grad_norm": 1.5670020537319092,
+      "learning_rate": 4.280546378098792e-06,
+      "loss": 0.6197,
+      "step": 3560
+    },
+    {
+      "epoch": 2.108972460763992,
+      "grad_norm": 1.4217647311568198,
+      "learning_rate": 4.2753011670664625e-06,
+      "loss": 0.6005,
+      "step": 3561
+    },
+    {
+      "epoch": 2.1095647023985786,
+      "grad_norm": 1.4382610743287825,
+      "learning_rate": 4.270058297757871e-06,
+      "loss": 0.6199,
+      "step": 3562
+    },
+    {
+      "epoch": 2.1101569440331653,
+      "grad_norm": 1.5200269742920969,
+      "learning_rate": 4.264817772317653e-06,
+      "loss": 0.6163,
+      "step": 3563
+    },
+    {
+      "epoch": 2.1107491856677525,
+      "grad_norm": 2.061967748457973,
+      "learning_rate": 4.259579592889464e-06,
+      "loss": 0.6036,
+      "step": 3564
+    },
+    {
+      "epoch": 2.111341427302339,
+      "grad_norm": 1.193306955598674,
+      "learning_rate": 4.25434376161602e-06,
+      "loss": 0.5957,
+      "step": 3565
+    },
+    {
+      "epoch": 2.1119336689369264,
+      "grad_norm": 1.15551634693182,
+      "learning_rate": 4.249110280639076e-06,
+      "loss": 0.5846,
+      "step": 3566
+    },
+    {
+      "epoch": 2.112525910571513,
+      "grad_norm": 1.9325467571821633,
+      "learning_rate": 4.243879152099415e-06,
+      "loss": 0.5902,
+      "step": 3567
+    },
+    {
+      "epoch": 2.1131181522061,
+      "grad_norm": 1.4819027125682924,
+      "learning_rate": 4.238650378136859e-06,
+      "loss": 0.5652,
+      "step": 3568
+    },
+    {
+      "epoch": 2.113710393840687,
+      "grad_norm": 1.221063734006737,
+      "learning_rate": 4.233423960890275e-06,
+      "loss": 0.6213,
+      "step": 3569
+    },
+    {
+      "epoch": 2.114302635475274,
+      "grad_norm": 1.2311749892527528,
+      "learning_rate": 4.228199902497565e-06,
+      "loss": 0.621,
+      "step": 3570
+    },
+    {
+      "epoch": 2.114894877109861,
+      "grad_norm": 1.6157414496724527,
+      "learning_rate": 4.222978205095659e-06,
+      "loss": 0.6328,
+      "step": 3571
+    },
+    {
+      "epoch": 2.115487118744448,
+      "grad_norm": 2.1238378827778654,
+      "learning_rate": 4.217758870820522e-06,
+      "loss": 0.6247,
+      "step": 3572
+    },
+    {
+      "epoch": 2.1160793603790347,
+      "grad_norm": 1.6576077793873196,
+      "learning_rate": 4.2125419018071565e-06,
+      "loss": 0.5893,
+      "step": 3573
+    },
+    {
+      "epoch": 2.1166716020136214,
+      "grad_norm": 7.289033310054799,
+      "learning_rate": 4.207327300189602e-06,
+      "loss": 0.6229,
+      "step": 3574
+    },
+    {
+      "epoch": 2.1172638436482085,
+      "grad_norm": 1.9096206920777807,
+      "learning_rate": 4.202115068100919e-06,
+      "loss": 0.6134,
+      "step": 3575
+    },
+    {
+      "epoch": 2.1178560852827952,
+      "grad_norm": 1.3294555847995595,
+      "learning_rate": 4.196905207673201e-06,
+      "loss": 0.6541,
+      "step": 3576
+    },
+    {
+      "epoch": 2.1184483269173824,
+      "grad_norm": 2.478570928826208,
+      "learning_rate": 4.191697721037577e-06,
+      "loss": 0.64,
+      "step": 3577
+    },
+    {
+      "epoch": 2.119040568551969,
+      "grad_norm": 1.3384108708761757,
+      "learning_rate": 4.186492610324204e-06,
+      "loss": 0.6313,
+      "step": 3578
+    },
+    {
+      "epoch": 2.1196328101865562,
+      "grad_norm": 2.059122205094186,
+      "learning_rate": 4.181289877662263e-06,
+      "loss": 0.6483,
+      "step": 3579
+    },
+    {
+      "epoch": 2.120225051821143,
+      "grad_norm": 2.210349418550527,
+      "learning_rate": 4.176089525179961e-06,
+      "loss": 0.6482,
+      "step": 3580
+    },
+    {
+      "epoch": 2.12081729345573,
+      "grad_norm": 1.2899545185125494,
+      "learning_rate": 4.1708915550045385e-06,
+      "loss": 0.6031,
+      "step": 3581
+    },
+    {
+      "epoch": 2.121409535090317,
+      "grad_norm": 1.0636519968386704,
+      "learning_rate": 4.165695969262259e-06,
+      "loss": 0.6238,
+      "step": 3582
+    },
+    {
+      "epoch": 2.122001776724904,
+      "grad_norm": 1.8256192982659298,
+      "learning_rate": 4.160502770078407e-06,
+      "loss": 0.6507,
+      "step": 3583
+    },
+    {
+      "epoch": 2.1225940183594907,
+      "grad_norm": 1.3887604621578677,
+      "learning_rate": 4.155311959577289e-06,
+      "loss": 0.5671,
+      "step": 3584
+    },
+    {
+      "epoch": 2.1231862599940774,
+      "grad_norm": 1.0622378454165742,
+      "learning_rate": 4.150123539882246e-06,
+      "loss": 0.5976,
+      "step": 3585
+    },
+    {
+      "epoch": 2.1237785016286646,
+      "grad_norm": 1.3574063696461265,
+      "learning_rate": 4.144937513115627e-06,
+      "loss": 0.6602,
+      "step": 3586
+    },
+    {
+      "epoch": 2.1243707432632513,
+      "grad_norm": 1.4538940676417582,
+      "learning_rate": 4.139753881398813e-06,
+      "loss": 0.5957,
+      "step": 3587
+    },
+    {
+      "epoch": 2.1249629848978384,
+      "grad_norm": 2.264238665767258,
+      "learning_rate": 4.134572646852196e-06,
+      "loss": 0.5999,
+      "step": 3588
+    },
+    {
+      "epoch": 2.125555226532425,
+      "grad_norm": 1.2966266388996097,
+      "learning_rate": 4.1293938115952e-06,
+      "loss": 0.6102,
+      "step": 3589
+    },
+    {
+      "epoch": 2.1261474681670123,
+      "grad_norm": 33.071460074035244,
+      "learning_rate": 4.124217377746251e-06,
+      "loss": 0.566,
+      "step": 3590
+    },
+    {
+      "epoch": 2.126739709801599,
+      "grad_norm": 1.1430460336126136,
+      "learning_rate": 4.11904334742281e-06,
+      "loss": 0.6487,
+      "step": 3591
+    },
+    {
+      "epoch": 2.127331951436186,
+      "grad_norm": 1.340637625601947,
+      "learning_rate": 4.113871722741337e-06,
+      "loss": 0.614,
+      "step": 3592
+    },
+    {
+      "epoch": 2.127924193070773,
+      "grad_norm": 4.321691288846663,
+      "learning_rate": 4.108702505817327e-06,
+      "loss": 0.6521,
+      "step": 3593
+    },
+    {
+      "epoch": 2.12851643470536,
+      "grad_norm": 1.3840804956759036,
+      "learning_rate": 4.103535698765272e-06,
+      "loss": 0.6095,
+      "step": 3594
+    },
+    {
+      "epoch": 2.1291086763399467,
+      "grad_norm": 1.5085339460848268,
+      "learning_rate": 4.098371303698694e-06,
+      "loss": 0.5992,
+      "step": 3595
+    },
+    {
+      "epoch": 2.1297009179745334,
+      "grad_norm": 2.0820647527427854,
+      "learning_rate": 4.093209322730114e-06,
+      "loss": 0.5808,
+      "step": 3596
+    },
+    {
+      "epoch": 2.1302931596091206,
+      "grad_norm": 1.2113506268447882,
+      "learning_rate": 4.08804975797108e-06,
+      "loss": 0.6079,
+      "step": 3597
+    },
+    {
+      "epoch": 2.1308854012437073,
+      "grad_norm": 1.450986037398985,
+      "learning_rate": 4.082892611532136e-06,
+      "loss": 0.634,
+      "step": 3598
+    },
+    {
+      "epoch": 2.1314776428782944,
+      "grad_norm": 1.8781499277404807,
+      "learning_rate": 4.077737885522852e-06,
+      "loss": 0.6374,
+      "step": 3599
+    },
+    {
+      "epoch": 2.132069884512881,
+      "grad_norm": 1.7864320630965398,
+      "learning_rate": 4.072585582051798e-06,
+      "loss": 0.627,
+      "step": 3600
+    },
+    {
+      "epoch": 2.1326621261474683,
+      "grad_norm": 1.4956135203030114,
+      "learning_rate": 4.067435703226552e-06,
+      "loss": 0.6108,
+      "step": 3601
+    },
+    {
+      "epoch": 2.133254367782055,
+      "grad_norm": 2.2180019781234,
+      "learning_rate": 4.0622882511537076e-06,
+      "loss": 0.5863,
+      "step": 3602
+    },
+    {
+      "epoch": 2.133846609416642,
+      "grad_norm": 4.119143056595805,
+      "learning_rate": 4.057143227938866e-06,
+      "loss": 0.6186,
+      "step": 3603
+    },
+    {
+      "epoch": 2.134438851051229,
+      "grad_norm": 1.5204561646279764,
+      "learning_rate": 4.052000635686627e-06,
+      "loss": 0.5892,
+      "step": 3604
+    },
+    {
+      "epoch": 2.135031092685816,
+      "grad_norm": 4.47672704923902,
+      "learning_rate": 4.046860476500596e-06,
+      "loss": 0.6405,
+      "step": 3605
+    },
+    {
+      "epoch": 2.1356233343204027,
+      "grad_norm": 1.7257594440244672,
+      "learning_rate": 4.0417227524833925e-06,
+      "loss": 0.6228,
+      "step": 3606
+    },
+    {
+      "epoch": 2.1362155759549895,
+      "grad_norm": 1.0415334302691686,
+      "learning_rate": 4.036587465736635e-06,
+      "loss": 0.6577,
+      "step": 3607
+    },
+    {
+      "epoch": 2.1368078175895766,
+      "grad_norm": 1.265219398940147,
+      "learning_rate": 4.031454618360945e-06,
+      "loss": 0.6531,
+      "step": 3608
+    },
+    {
+      "epoch": 2.1374000592241633,
+      "grad_norm": 1.2617611265687647,
+      "learning_rate": 4.02632421245594e-06,
+      "loss": 0.5512,
+      "step": 3609
+    },
+    {
+      "epoch": 2.1379923008587505,
+      "grad_norm": 1.241378144123858,
+      "learning_rate": 4.021196250120248e-06,
+      "loss": 0.6223,
+      "step": 3610
+    },
+    {
+      "epoch": 2.138584542493337,
+      "grad_norm": 1.8843558424320774,
+      "learning_rate": 4.016070733451496e-06,
+      "loss": 0.6307,
+      "step": 3611
+    },
+    {
+      "epoch": 2.1391767841279243,
+      "grad_norm": 1.152700464562051,
+      "learning_rate": 4.0109476645463076e-06,
+      "loss": 0.6149,
+      "step": 3612
+    },
+    {
+      "epoch": 2.139769025762511,
+      "grad_norm": 1.646886331579089,
+      "learning_rate": 4.005827045500301e-06,
+      "loss": 0.6022,
+      "step": 3613
+    },
+    {
+      "epoch": 2.140361267397098,
+      "grad_norm": 1.4285955128879986,
+      "learning_rate": 4.000708878408103e-06,
+      "loss": 0.5811,
+      "step": 3614
+    },
+    {
+      "epoch": 2.140953509031685,
+      "grad_norm": 1.8299679014912458,
+      "learning_rate": 3.9955931653633365e-06,
+      "loss": 0.6211,
+      "step": 3615
+    },
+    {
+      "epoch": 2.141545750666272,
+      "grad_norm": 1.4096002284217504,
+      "learning_rate": 3.990479908458602e-06,
+      "loss": 0.6149,
+      "step": 3616
+    },
+    {
+      "epoch": 2.1421379923008588,
+      "grad_norm": 2.69120669503384,
+      "learning_rate": 3.985369109785516e-06,
+      "loss": 0.6343,
+      "step": 3617
+    },
+    {
+      "epoch": 2.1427302339354455,
+      "grad_norm": 1.1754488546881272,
+      "learning_rate": 3.980260771434685e-06,
+      "loss": 0.6071,
+      "step": 3618
+    },
+    {
+      "epoch": 2.1433224755700326,
+      "grad_norm": 1.6113726977096838,
+      "learning_rate": 3.975154895495711e-06,
+      "loss": 0.5882,
+      "step": 3619
+    },
+    {
+      "epoch": 2.1439147172046193,
+      "grad_norm": 2.791107478639297,
+      "learning_rate": 3.970051484057171e-06,
+      "loss": 0.635,
+      "step": 3620
+    },
+    {
+      "epoch": 2.1445069588392065,
+      "grad_norm": 1.5913773207532422,
+      "learning_rate": 3.9649505392066544e-06,
+      "loss": 0.5979,
+      "step": 3621
+    },
+    {
+      "epoch": 2.145099200473793,
+      "grad_norm": 1.5408491987361748,
+      "learning_rate": 3.959852063030738e-06,
+      "loss": 0.6264,
+      "step": 3622
+    },
+    {
+      "epoch": 2.1456914421083804,
+      "grad_norm": 1.4228715868052781,
+      "learning_rate": 3.9547560576149815e-06,
+      "loss": 0.6112,
+      "step": 3623
+    },
+    {
+      "epoch": 2.146283683742967,
+      "grad_norm": 1.894084960567437,
+      "learning_rate": 3.949662525043935e-06,
+      "loss": 0.6981,
+      "step": 3624
+    },
+    {
+      "epoch": 2.1468759253775542,
+      "grad_norm": 29.083821486159245,
+      "learning_rate": 3.944571467401142e-06,
+      "loss": 0.5861,
+      "step": 3625
+    },
+    {
+      "epoch": 2.147468167012141,
+      "grad_norm": 1.778409125554096,
+      "learning_rate": 3.939482886769136e-06,
+      "loss": 0.6135,
+      "step": 3626
+    },
+    {
+      "epoch": 2.148060408646728,
+      "grad_norm": 1.7787742540030977,
+      "learning_rate": 3.934396785229429e-06,
+      "loss": 0.5599,
+      "step": 3627
+    },
+    {
+      "epoch": 2.148652650281315,
+      "grad_norm": 1.9679765606795838,
+      "learning_rate": 3.929313164862518e-06,
+      "loss": 0.637,
+      "step": 3628
+    },
+    {
+      "epoch": 2.1492448919159015,
+      "grad_norm": 1.5563281004545866,
+      "learning_rate": 3.924232027747894e-06,
+      "loss": 0.6389,
+      "step": 3629
+    },
+    {
+      "epoch": 2.1498371335504887,
+      "grad_norm": 1.2422466217603876,
+      "learning_rate": 3.919153375964032e-06,
+      "loss": 0.567,
+      "step": 3630
+    },
+    {
+      "epoch": 2.1504293751850754,
+      "grad_norm": 2.00976111403471,
+      "learning_rate": 3.91407721158838e-06,
+      "loss": 0.5458,
+      "step": 3631
+    },
+    {
+      "epoch": 2.1510216168196625,
+      "grad_norm": 1.6951508776864774,
+      "learning_rate": 3.909003536697374e-06,
+      "loss": 0.605,
+      "step": 3632
+    },
+    {
+      "epoch": 2.1516138584542492,
+      "grad_norm": 1.558169994124345,
+      "learning_rate": 3.903932353366435e-06,
+      "loss": 0.595,
+      "step": 3633
+    },
+    {
+      "epoch": 2.1522061000888364,
+      "grad_norm": 1.6276022870031455,
+      "learning_rate": 3.898863663669965e-06,
+      "loss": 0.6334,
+      "step": 3634
+    },
+    {
+      "epoch": 2.152798341723423,
+      "grad_norm": 1.6198572914267186,
+      "learning_rate": 3.8937974696813405e-06,
+      "loss": 0.5621,
+      "step": 3635
+    },
+    {
+      "epoch": 2.1533905833580103,
+      "grad_norm": 1.9472727059260369,
+      "learning_rate": 3.888733773472916e-06,
+      "loss": 0.5736,
+      "step": 3636
+    },
+    {
+      "epoch": 2.153982824992597,
+      "grad_norm": 2.552984173348855,
+      "learning_rate": 3.883672577116035e-06,
+      "loss": 0.6166,
+      "step": 3637
+    },
+    {
+      "epoch": 2.154575066627184,
+      "grad_norm": 1.9217704150949895,
+      "learning_rate": 3.878613882681002e-06,
+      "loss": 0.6168,
+      "step": 3638
+    },
+    {
+      "epoch": 2.155167308261771,
+      "grad_norm": 1.5327012862258527,
+      "learning_rate": 3.873557692237119e-06,
+      "loss": 0.5939,
+      "step": 3639
+    },
+    {
+      "epoch": 2.1557595498963575,
+      "grad_norm": 1.4830748099663695,
+      "learning_rate": 3.868504007852641e-06,
+      "loss": 0.636,
+      "step": 3640
+    },
+    {
+      "epoch": 2.1563517915309447,
+      "grad_norm": 1.4934300428699974,
+      "learning_rate": 3.86345283159482e-06,
+      "loss": 0.5907,
+      "step": 3641
+    },
+    {
+      "epoch": 2.1569440331655314,
+      "grad_norm": 1.8303353542056642,
+      "learning_rate": 3.8584041655298606e-06,
+      "loss": 0.6217,
+      "step": 3642
+    },
+    {
+      "epoch": 2.1575362748001186,
+      "grad_norm": 2.6188924474056203,
+      "learning_rate": 3.853358011722961e-06,
+      "loss": 0.6247,
+      "step": 3643
+    },
+    {
+      "epoch": 2.1581285164347053,
+      "grad_norm": 1.9829717288285205,
+      "learning_rate": 3.848314372238272e-06,
+      "loss": 0.5971,
+      "step": 3644
+    },
+    {
+      "epoch": 2.1587207580692924,
+      "grad_norm": 2.9821607547301348,
+      "learning_rate": 3.8432732491389345e-06,
+      "loss": 0.6207,
+      "step": 3645
+    },
+    {
+      "epoch": 2.159312999703879,
+      "grad_norm": 1.4864857735219268,
+      "learning_rate": 3.838234644487045e-06,
+      "loss": 0.6185,
+      "step": 3646
+    },
+    {
+      "epoch": 2.1599052413384663,
+      "grad_norm": 2.2631386629045287,
+      "learning_rate": 3.833198560343682e-06,
+      "loss": 0.5928,
+      "step": 3647
+    },
+    {
+      "epoch": 2.160497482973053,
+      "grad_norm": 1.2595848974397195,
+      "learning_rate": 3.828164998768879e-06,
+      "loss": 0.6202,
+      "step": 3648
+    },
+    {
+      "epoch": 2.16108972460764,
+      "grad_norm": 1.2855787943939143,
+      "learning_rate": 3.8231339618216556e-06,
+      "loss": 0.5912,
+      "step": 3649
+    },
+    {
+      "epoch": 2.161681966242227,
+      "grad_norm": 3.733946616282587,
+      "learning_rate": 3.8181054515599806e-06,
+      "loss": 0.642,
+      "step": 3650
+    },
+    {
+      "epoch": 2.1622742078768136,
+      "grad_norm": 1.9716595141173552,
+      "learning_rate": 3.8130794700408027e-06,
+      "loss": 0.576,
+      "step": 3651
+    },
+    {
+      "epoch": 2.1628664495114007,
+      "grad_norm": 10.436223336411803,
+      "learning_rate": 3.8080560193200288e-06,
+      "loss": 0.59,
+      "step": 3652
+    },
+    {
+      "epoch": 2.1634586911459874,
+      "grad_norm": 1.6433990265618101,
+      "learning_rate": 3.803035101452531e-06,
+      "loss": 0.6323,
+      "step": 3653
+    },
+    {
+      "epoch": 2.1640509327805746,
+      "grad_norm": 1.4190321545816333,
+      "learning_rate": 3.798016718492148e-06,
+      "loss": 0.6114,
+      "step": 3654
+    },
+    {
+      "epoch": 2.1646431744151613,
+      "grad_norm": 1.559115691980612,
+      "learning_rate": 3.7930008724916846e-06,
+      "loss": 0.583,
+      "step": 3655
+    },
+    {
+      "epoch": 2.1652354160497485,
+      "grad_norm": 2.1204592019182655,
+      "learning_rate": 3.7879875655029018e-06,
+      "loss": 0.6153,
+      "step": 3656
+    },
+    {
+      "epoch": 2.165827657684335,
+      "grad_norm": 1.9790725686824444,
+      "learning_rate": 3.782976799576519e-06,
+      "loss": 0.6118,
+      "step": 3657
+    },
+    {
+      "epoch": 2.1664198993189223,
+      "grad_norm": 1.7029886439459894,
+      "learning_rate": 3.7779685767622255e-06,
+      "loss": 0.6315,
+      "step": 3658
+    },
+    {
+      "epoch": 2.167012140953509,
+      "grad_norm": 1.3790840194704055,
+      "learning_rate": 3.7729628991086687e-06,
+      "loss": 0.6036,
+      "step": 3659
+    },
+    {
+      "epoch": 2.167604382588096,
+      "grad_norm": 1.7802518189464025,
+      "learning_rate": 3.7679597686634495e-06,
+      "loss": 0.5877,
+      "step": 3660
+    },
+    {
+      "epoch": 2.168196624222683,
+      "grad_norm": 1.746849203801897,
+      "learning_rate": 3.7629591874731264e-06,
+      "loss": 0.6504,
+      "step": 3661
+    },
+    {
+      "epoch": 2.1687888658572696,
+      "grad_norm": 1.6998904736750884,
+      "learning_rate": 3.757961157583221e-06,
+      "loss": 0.6304,
+      "step": 3662
+    },
+    {
+      "epoch": 2.1693811074918568,
+      "grad_norm": 1.3810219749838843,
+      "learning_rate": 3.7529656810382133e-06,
+      "loss": 0.6296,
+      "step": 3663
+    },
+    {
+      "epoch": 2.1699733491264435,
+      "grad_norm": 1.298383302796027,
+      "learning_rate": 3.7479727598815287e-06,
+      "loss": 0.6187,
+      "step": 3664
+    },
+    {
+      "epoch": 2.1705655907610306,
+      "grad_norm": 2.1800686642840987,
+      "learning_rate": 3.7429823961555513e-06,
+      "loss": 0.6048,
+      "step": 3665
+    },
+    {
+      "epoch": 2.1711578323956173,
+      "grad_norm": 1.138658124740238,
+      "learning_rate": 3.7379945919016225e-06,
+      "loss": 0.659,
+      "step": 3666
+    },
+    {
+      "epoch": 2.1717500740302045,
+      "grad_norm": 1.3434860594717237,
+      "learning_rate": 3.733009349160042e-06,
+      "loss": 0.6048,
+      "step": 3667
+    },
+    {
+      "epoch": 2.172342315664791,
+      "grad_norm": 1.764600534002413,
+      "learning_rate": 3.7280266699700406e-06,
+      "loss": 0.5861,
+      "step": 3668
+    },
+    {
+      "epoch": 2.1729345572993783,
+      "grad_norm": 1.1095526363681014,
+      "learning_rate": 3.7230465563698214e-06,
+      "loss": 0.6066,
+      "step": 3669
+    },
+    {
+      "epoch": 2.173526798933965,
+      "grad_norm": 3.064677575014838,
+      "learning_rate": 3.7180690103965313e-06,
+      "loss": 0.6762,
+      "step": 3670
+    },
+    {
+      "epoch": 2.1741190405685518,
+      "grad_norm": 1.0892290245854241,
+      "learning_rate": 3.713094034086273e-06,
+      "loss": 0.6142,
+      "step": 3671
+    },
+    {
+      "epoch": 2.174711282203139,
+      "grad_norm": 2.014740928840878,
+      "learning_rate": 3.7081216294740773e-06,
+      "loss": 0.6006,
+      "step": 3672
+    },
+    {
+      "epoch": 2.1753035238377256,
+      "grad_norm": 1.6840077237428435,
+      "learning_rate": 3.703151798593945e-06,
+      "loss": 0.6111,
+      "step": 3673
+    },
+    {
+      "epoch": 2.175895765472313,
+      "grad_norm": 3.147435725567642,
+      "learning_rate": 3.6981845434788188e-06,
+      "loss": 0.59,
+      "step": 3674
+    },
+    {
+      "epoch": 2.1764880071068995,
+      "grad_norm": 2.3668044579752547,
+      "learning_rate": 3.693219866160582e-06,
+      "loss": 0.6486,
+      "step": 3675
+    },
+    {
+      "epoch": 2.1770802487414866,
+      "grad_norm": 1.7822979150577114,
+      "learning_rate": 3.688257768670065e-06,
+      "loss": 0.617,
+      "step": 3676
+    },
+    {
+      "epoch": 2.1776724903760734,
+      "grad_norm": 1.5174818432951496,
+      "learning_rate": 3.6832982530370465e-06,
+      "loss": 0.62,
+      "step": 3677
+    },
+    {
+      "epoch": 2.1782647320106605,
+      "grad_norm": 1.4448335060389763,
+      "learning_rate": 3.678341321290252e-06,
+      "loss": 0.6469,
+      "step": 3678
+    },
+    {
+      "epoch": 2.178856973645247,
+      "grad_norm": 1.55193829165598,
+      "learning_rate": 3.6733869754573403e-06,
+      "loss": 0.602,
+      "step": 3679
+    },
+    {
+      "epoch": 2.1794492152798344,
+      "grad_norm": 1.1427395519868442,
+      "learning_rate": 3.668435217564915e-06,
+      "loss": 0.6552,
+      "step": 3680
+    },
+    {
+      "epoch": 2.180041456914421,
+      "grad_norm": 1.3343442976110556,
+      "learning_rate": 3.663486049638527e-06,
+      "loss": 0.6427,
+      "step": 3681
+    },
+    {
+      "epoch": 2.180633698549008,
+      "grad_norm": 1.6318631166682127,
+      "learning_rate": 3.658539473702667e-06,
+      "loss": 0.6476,
+      "step": 3682
+    },
+    {
+      "epoch": 2.181225940183595,
+      "grad_norm": 1.4447436327409222,
+      "learning_rate": 3.65359549178076e-06,
+      "loss": 0.595,
+      "step": 3683
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 1.1309723434982337,
+      "learning_rate": 3.6486541058951696e-06,
+      "loss": 0.6008,
+      "step": 3684
+    },
+    {
+      "epoch": 2.182410423452769,
+      "grad_norm": 1.7100247943717297,
+      "learning_rate": 3.6437153180672034e-06,
+      "loss": 0.6291,
+      "step": 3685
+    },
+    {
+      "epoch": 2.1830026650873555,
+      "grad_norm": 1.6181277098556224,
+      "learning_rate": 3.638779130317106e-06,
+      "loss": 0.6039,
+      "step": 3686
+    },
+    {
+      "epoch": 2.1835949067219427,
+      "grad_norm": 1.5754065682938263,
+      "learning_rate": 3.633845544664053e-06,
+      "loss": 0.6282,
+      "step": 3687
+    },
+    {
+      "epoch": 2.1841871483565294,
+      "grad_norm": 2.344647272098969,
+      "learning_rate": 3.628914563126156e-06,
+      "loss": 0.6411,
+      "step": 3688
+    },
+    {
+      "epoch": 2.1847793899911165,
+      "grad_norm": 6.915326366107702,
+      "learning_rate": 3.6239861877204684e-06,
+      "loss": 0.5722,
+      "step": 3689
+    },
+    {
+      "epoch": 2.1853716316257032,
+      "grad_norm": 1.3363840965456972,
+      "learning_rate": 3.6190604204629685e-06,
+      "loss": 0.6376,
+      "step": 3690
+    },
+    {
+      "epoch": 2.1859638732602904,
+      "grad_norm": 1.5541366844193774,
+      "learning_rate": 3.6141372633685767e-06,
+      "loss": 0.6474,
+      "step": 3691
+    },
+    {
+      "epoch": 2.186556114894877,
+      "grad_norm": 2.254316560958191,
+      "learning_rate": 3.6092167184511352e-06,
+      "loss": 0.6008,
+      "step": 3692
+    },
+    {
+      "epoch": 2.187148356529464,
+      "grad_norm": 2.255989122727738,
+      "learning_rate": 3.6042987877234304e-06,
+      "loss": 0.6062,
+      "step": 3693
+    },
+    {
+      "epoch": 2.187740598164051,
+      "grad_norm": 1.320086179706794,
+      "learning_rate": 3.5993834731971654e-06,
+      "loss": 0.5777,
+      "step": 3694
+    },
+    {
+      "epoch": 2.1883328397986377,
+      "grad_norm": 1.8280860055033141,
+      "learning_rate": 3.594470776882989e-06,
+      "loss": 0.5831,
+      "step": 3695
+    },
+    {
+      "epoch": 2.188925081433225,
+      "grad_norm": 1.9043201529453062,
+      "learning_rate": 3.5895607007904597e-06,
+      "loss": 0.5996,
+      "step": 3696
+    },
+    {
+      "epoch": 2.1895173230678115,
+      "grad_norm": 2.358121027410781,
+      "learning_rate": 3.584653246928085e-06,
+      "loss": 0.6427,
+      "step": 3697
+    },
+    {
+      "epoch": 2.1901095647023987,
+      "grad_norm": 1.3402210675480135,
+      "learning_rate": 3.5797484173032806e-06,
+      "loss": 0.6018,
+      "step": 3698
+    },
+    {
+      "epoch": 2.1907018063369854,
+      "grad_norm": 2.2835653872932706,
+      "learning_rate": 3.5748462139224048e-06,
+      "loss": 0.656,
+      "step": 3699
+    },
+    {
+      "epoch": 2.1912940479715726,
+      "grad_norm": 1.1925267592647002,
+      "learning_rate": 3.569946638790729e-06,
+      "loss": 0.5937,
+      "step": 3700
+    },
+    {
+      "epoch": 2.1918862896061593,
+      "grad_norm": 6.950862477600291,
+      "learning_rate": 3.5650496939124602e-06,
+      "loss": 0.5998,
+      "step": 3701
+    },
+    {
+      "epoch": 2.1924785312407464,
+      "grad_norm": 1.5302796583951903,
+      "learning_rate": 3.5601553812907174e-06,
+      "loss": 0.6095,
+      "step": 3702
+    },
+    {
+      "epoch": 2.193070772875333,
+      "grad_norm": 1.6634344074466962,
+      "learning_rate": 3.555263702927558e-06,
+      "loss": 0.6575,
+      "step": 3703
+    },
+    {
+      "epoch": 2.19366301450992,
+      "grad_norm": 1.3910377831755019,
+      "learning_rate": 3.5503746608239487e-06,
+      "loss": 0.6121,
+      "step": 3704
+    },
+    {
+      "epoch": 2.194255256144507,
+      "grad_norm": 1.6713226754844968,
+      "learning_rate": 3.54548825697978e-06,
+      "loss": 0.641,
+      "step": 3705
+    },
+    {
+      "epoch": 2.1948474977790937,
+      "grad_norm": 1.5747338103446598,
+      "learning_rate": 3.5406044933938688e-06,
+      "loss": 0.6064,
+      "step": 3706
+    },
+    {
+      "epoch": 2.195439739413681,
+      "grad_norm": 1.5768868641879534,
+      "learning_rate": 3.535723372063952e-06,
+      "loss": 0.6197,
+      "step": 3707
+    },
+    {
+      "epoch": 2.1960319810482676,
+      "grad_norm": 1.6152013346052505,
+      "learning_rate": 3.5308448949866805e-06,
+      "loss": 0.6304,
+      "step": 3708
+    },
+    {
+      "epoch": 2.1966242226828547,
+      "grad_norm": 1.3489870236047408,
+      "learning_rate": 3.5259690641576216e-06,
+      "loss": 0.6066,
+      "step": 3709
+    },
+    {
+      "epoch": 2.1972164643174414,
+      "grad_norm": 2.0855784992139825,
+      "learning_rate": 3.5210958815712672e-06,
+      "loss": 0.5858,
+      "step": 3710
+    },
+    {
+      "epoch": 2.1978087059520286,
+      "grad_norm": 2.1655140503921855,
+      "learning_rate": 3.5162253492210276e-06,
+      "loss": 0.6431,
+      "step": 3711
+    },
+    {
+      "epoch": 2.1984009475866153,
+      "grad_norm": 1.3963981694086718,
+      "learning_rate": 3.5113574690992203e-06,
+      "loss": 0.6041,
+      "step": 3712
+    },
+    {
+      "epoch": 2.198993189221202,
+      "grad_norm": 1.9394182753615543,
+      "learning_rate": 3.5064922431970793e-06,
+      "loss": 0.5795,
+      "step": 3713
+    },
+    {
+      "epoch": 2.199585430855789,
+      "grad_norm": 1.3857185510161663,
+      "learning_rate": 3.5016296735047584e-06,
+      "loss": 0.6244,
+      "step": 3714
+    },
+    {
+      "epoch": 2.200177672490376,
+      "grad_norm": 1.6061122144253255,
+      "learning_rate": 3.496769762011325e-06,
+      "loss": 0.6293,
+      "step": 3715
+    },
+    {
+      "epoch": 2.200769914124963,
+      "grad_norm": 5.40581536454228,
+      "learning_rate": 3.4919125107047537e-06,
+      "loss": 0.6259,
+      "step": 3716
+    },
+    {
+      "epoch": 2.2013621557595497,
+      "grad_norm": 1.2657815177175666,
+      "learning_rate": 3.487057921571929e-06,
+      "loss": 0.5988,
+      "step": 3717
+    },
+    {
+      "epoch": 2.201954397394137,
+      "grad_norm": 1.7259059897471152,
+      "learning_rate": 3.482205996598654e-06,
+      "loss": 0.6134,
+      "step": 3718
+    },
+    {
+      "epoch": 2.2025466390287236,
+      "grad_norm": 1.184129395946116,
+      "learning_rate": 3.477356737769645e-06,
+      "loss": 0.6274,
+      "step": 3719
+    },
+    {
+      "epoch": 2.2031388806633108,
+      "grad_norm": 1.7443680463119289,
+      "learning_rate": 3.472510147068515e-06,
+      "loss": 0.6375,
+      "step": 3720
+    },
+    {
+      "epoch": 2.2037311222978975,
+      "grad_norm": 1.4046681220312973,
+      "learning_rate": 3.4676662264777905e-06,
+      "loss": 0.6456,
+      "step": 3721
+    },
+    {
+      "epoch": 2.2043233639324846,
+      "grad_norm": 2.5357065102499994,
+      "learning_rate": 3.4628249779789105e-06,
+      "loss": 0.6201,
+      "step": 3722
+    },
+    {
+      "epoch": 2.2049156055670713,
+      "grad_norm": 2.357882593644408,
+      "learning_rate": 3.4579864035522236e-06,
+      "loss": 0.6352,
+      "step": 3723
+    },
+    {
+      "epoch": 2.205507847201658,
+      "grad_norm": 1.2137268888170112,
+      "learning_rate": 3.4531505051769665e-06,
+      "loss": 0.6398,
+      "step": 3724
+    },
+    {
+      "epoch": 2.206100088836245,
+      "grad_norm": 2.239315397553286,
+      "learning_rate": 3.4483172848312994e-06,
+      "loss": 0.6236,
+      "step": 3725
+    },
+    {
+      "epoch": 2.206692330470832,
+      "grad_norm": 1.557932862394711,
+      "learning_rate": 3.4434867444922857e-06,
+      "loss": 0.5774,
+      "step": 3726
+    },
+    {
+      "epoch": 2.207284572105419,
+      "grad_norm": 1.5777287070135115,
+      "learning_rate": 3.438658886135884e-06,
+      "loss": 0.5508,
+      "step": 3727
+    },
+    {
+      "epoch": 2.2078768137400058,
+      "grad_norm": 1.413920945713966,
+      "learning_rate": 3.433833711736957e-06,
+      "loss": 0.6097,
+      "step": 3728
+    },
+    {
+      "epoch": 2.208469055374593,
+      "grad_norm": 1.258492807065135,
+      "learning_rate": 3.429011223269274e-06,
+      "loss": 0.6256,
+      "step": 3729
+    },
+    {
+      "epoch": 2.2090612970091796,
+      "grad_norm": 3.75247152278292,
+      "learning_rate": 3.4241914227055096e-06,
+      "loss": 0.5718,
+      "step": 3730
+    },
+    {
+      "epoch": 2.209653538643767,
+      "grad_norm": 3.7102331792806478,
+      "learning_rate": 3.4193743120172297e-06,
+      "loss": 0.5975,
+      "step": 3731
+    },
+    {
+      "epoch": 2.2102457802783535,
+      "grad_norm": 1.7080087366642982,
+      "learning_rate": 3.414559893174898e-06,
+      "loss": 0.5949,
+      "step": 3732
+    },
+    {
+      "epoch": 2.2108380219129407,
+      "grad_norm": 2.0252022919779153,
+      "learning_rate": 3.4097481681478873e-06,
+      "loss": 0.6596,
+      "step": 3733
+    },
+    {
+      "epoch": 2.2114302635475274,
+      "grad_norm": 1.7291383197002848,
+      "learning_rate": 3.4049391389044674e-06,
+      "loss": 0.581,
+      "step": 3734
+    },
+    {
+      "epoch": 2.212022505182114,
+      "grad_norm": 2.2935670684623286,
+      "learning_rate": 3.4001328074117977e-06,
+      "loss": 0.6159,
+      "step": 3735
+    },
+    {
+      "epoch": 2.2126147468167012,
+      "grad_norm": 1.5768762657428204,
+      "learning_rate": 3.3953291756359354e-06,
+      "loss": 0.6298,
+      "step": 3736
+    },
+    {
+      "epoch": 2.213206988451288,
+      "grad_norm": 2.260996804956774,
+      "learning_rate": 3.3905282455418375e-06,
+      "loss": 0.6006,
+      "step": 3737
+    },
+    {
+      "epoch": 2.213799230085875,
+      "grad_norm": 1.4582667950449473,
+      "learning_rate": 3.3857300190933606e-06,
+      "loss": 0.6607,
+      "step": 3738
+    },
+    {
+      "epoch": 2.214391471720462,
+      "grad_norm": 1.7196961502523231,
+      "learning_rate": 3.3809344982532435e-06,
+      "loss": 0.6154,
+      "step": 3739
+    },
+    {
+      "epoch": 2.214983713355049,
+      "grad_norm": 1.4224301706484694,
+      "learning_rate": 3.376141684983121e-06,
+      "loss": 0.6257,
+      "step": 3740
+    },
+    {
+      "epoch": 2.2155759549896357,
+      "grad_norm": 1.5242446120056001,
+      "learning_rate": 3.3713515812435305e-06,
+      "loss": 0.6395,
+      "step": 3741
+    },
+    {
+      "epoch": 2.216168196624223,
+      "grad_norm": 3.082276038619601,
+      "learning_rate": 3.366564188993887e-06,
+      "loss": 0.607,
+      "step": 3742
+    },
+    {
+      "epoch": 2.2167604382588095,
+      "grad_norm": 1.1091952397891178,
+      "learning_rate": 3.36177951019251e-06,
+      "loss": 0.5673,
+      "step": 3743
+    },
+    {
+      "epoch": 2.2173526798933967,
+      "grad_norm": 1.0623328174773825,
+      "learning_rate": 3.3569975467965955e-06,
+      "loss": 0.6197,
+      "step": 3744
+    },
+    {
+      "epoch": 2.2179449215279834,
+      "grad_norm": 1.9139569358167408,
+      "learning_rate": 3.352218300762241e-06,
+      "loss": 0.6325,
+      "step": 3745
+    },
+    {
+      "epoch": 2.21853716316257,
+      "grad_norm": 1.754753069962319,
+      "learning_rate": 3.347441774044421e-06,
+      "loss": 0.6194,
+      "step": 3746
+    },
+    {
+      "epoch": 2.2191294047971573,
+      "grad_norm": 1.311850230899881,
+      "learning_rate": 3.3426679685970096e-06,
+      "loss": 0.6365,
+      "step": 3747
+    },
+    {
+      "epoch": 2.219721646431744,
+      "grad_norm": 2.3171585866726243,
+      "learning_rate": 3.337896886372757e-06,
+      "loss": 0.6015,
+      "step": 3748
+    },
+    {
+      "epoch": 2.220313888066331,
+      "grad_norm": 3.610142905808541,
+      "learning_rate": 3.3331285293233086e-06,
+      "loss": 0.5806,
+      "step": 3749
+    },
+    {
+      "epoch": 2.220906129700918,
+      "grad_norm": 1.4393742110166086,
+      "learning_rate": 3.3283628993991846e-06,
+      "loss": 0.5956,
+      "step": 3750
+    },
+    {
+      "epoch": 2.221498371335505,
+      "grad_norm": 1.5219021843139517,
+      "learning_rate": 3.3235999985498036e-06,
+      "loss": 0.6268,
+      "step": 3751
+    },
+    {
+      "epoch": 2.2220906129700917,
+      "grad_norm": 8.505541739782977,
+      "learning_rate": 3.3188398287234504e-06,
+      "loss": 0.602,
+      "step": 3752
+    },
+    {
+      "epoch": 2.222682854604679,
+      "grad_norm": 1.4777906472192541,
+      "learning_rate": 3.3140823918673117e-06,
+      "loss": 0.5723,
+      "step": 3753
+    },
+    {
+      "epoch": 2.2232750962392656,
+      "grad_norm": 3.427821946647888,
+      "learning_rate": 3.3093276899274373e-06,
+      "loss": 0.6404,
+      "step": 3754
+    },
+    {
+      "epoch": 2.2238673378738527,
+      "grad_norm": 1.5440256787529059,
+      "learning_rate": 3.3045757248487763e-06,
+      "loss": 0.5923,
+      "step": 3755
+    },
+    {
+      "epoch": 2.2244595795084394,
+      "grad_norm": 1.0121502394479982,
+      "learning_rate": 3.2998264985751425e-06,
+      "loss": 0.6091,
+      "step": 3756
+    },
+    {
+      "epoch": 2.225051821143026,
+      "grad_norm": 1.435124556629406,
+      "learning_rate": 3.2950800130492434e-06,
+      "loss": 0.5865,
+      "step": 3757
+    },
+    {
+      "epoch": 2.2256440627776133,
+      "grad_norm": 1.400676794978446,
+      "learning_rate": 3.2903362702126516e-06,
+      "loss": 0.5851,
+      "step": 3758
+    },
+    {
+      "epoch": 2.2262363044122,
+      "grad_norm": 2.776885503195463,
+      "learning_rate": 3.2855952720058303e-06,
+      "loss": 0.6104,
+      "step": 3759
+    },
+    {
+      "epoch": 2.226828546046787,
+      "grad_norm": 1.2754379174271564,
+      "learning_rate": 3.2808570203681135e-06,
+      "loss": 0.6243,
+      "step": 3760
+    },
+    {
+      "epoch": 2.227420787681374,
+      "grad_norm": 2.047090232279367,
+      "learning_rate": 3.2761215172377057e-06,
+      "loss": 0.6244,
+      "step": 3761
+    },
+    {
+      "epoch": 2.228013029315961,
+      "grad_norm": 1.9009466375062916,
+      "learning_rate": 3.271388764551702e-06,
+      "loss": 0.6162,
+      "step": 3762
+    },
+    {
+      "epoch": 2.2286052709505477,
+      "grad_norm": 1.7400174906099655,
+      "learning_rate": 3.266658764246062e-06,
+      "loss": 0.5828,
+      "step": 3763
+    },
+    {
+      "epoch": 2.229197512585135,
+      "grad_norm": 1.4745943692336556,
+      "learning_rate": 3.2619315182556234e-06,
+      "loss": 0.6394,
+      "step": 3764
+    },
+    {
+      "epoch": 2.2297897542197216,
+      "grad_norm": 2.7463760563137667,
+      "learning_rate": 3.25720702851409e-06,
+      "loss": 0.5372,
+      "step": 3765
+    },
+    {
+      "epoch": 2.2303819958543087,
+      "grad_norm": 1.291326363777691,
+      "learning_rate": 3.2524852969540477e-06,
+      "loss": 0.6244,
+      "step": 3766
+    },
+    {
+      "epoch": 2.2309742374888955,
+      "grad_norm": 3.0151646713135456,
+      "learning_rate": 3.2477663255069536e-06,
+      "loss": 0.5868,
+      "step": 3767
+    },
+    {
+      "epoch": 2.231566479123482,
+      "grad_norm": 1.3901935328359527,
+      "learning_rate": 3.243050116103128e-06,
+      "loss": 0.6149,
+      "step": 3768
+    },
+    {
+      "epoch": 2.2321587207580693,
+      "grad_norm": 2.8729426761941332,
+      "learning_rate": 3.2383366706717647e-06,
+      "loss": 0.6374,
+      "step": 3769
+    },
+    {
+      "epoch": 2.232750962392656,
+      "grad_norm": 1.3514288298661428,
+      "learning_rate": 3.2336259911409283e-06,
+      "loss": 0.6504,
+      "step": 3770
+    },
+    {
+      "epoch": 2.233343204027243,
+      "grad_norm": 1.2214092762638862,
+      "learning_rate": 3.228918079437556e-06,
+      "loss": 0.6331,
+      "step": 3771
+    },
+    {
+      "epoch": 2.23393544566183,
+      "grad_norm": 1.9604923888716306,
+      "learning_rate": 3.2242129374874478e-06,
+      "loss": 0.5907,
+      "step": 3772
+    },
+    {
+      "epoch": 2.234527687296417,
+      "grad_norm": 1.726601509131056,
+      "learning_rate": 3.219510567215264e-06,
+      "loss": 0.6266,
+      "step": 3773
+    },
+    {
+      "epoch": 2.2351199289310038,
+      "grad_norm": 1.518643995060012,
+      "learning_rate": 3.2148109705445442e-06,
+      "loss": 0.6445,
+      "step": 3774
+    },
+    {
+      "epoch": 2.235712170565591,
+      "grad_norm": 1.386321333561844,
+      "learning_rate": 3.2101141493976938e-06,
+      "loss": 0.6196,
+      "step": 3775
+    },
+    {
+      "epoch": 2.2363044122001776,
+      "grad_norm": 1.5786767296496107,
+      "learning_rate": 3.205420105695963e-06,
+      "loss": 0.6102,
+      "step": 3776
+    },
+    {
+      "epoch": 2.2368966538347648,
+      "grad_norm": 1.2013453929038438,
+      "learning_rate": 3.200728841359487e-06,
+      "loss": 0.6102,
+      "step": 3777
+    },
+    {
+      "epoch": 2.2374888954693515,
+      "grad_norm": 1.4238732131144223,
+      "learning_rate": 3.1960403583072596e-06,
+      "loss": 0.5799,
+      "step": 3778
+    },
+    {
+      "epoch": 2.238081137103938,
+      "grad_norm": 1.2261171846122096,
+      "learning_rate": 3.191354658457131e-06,
+      "loss": 0.6501,
+      "step": 3779
+    },
+    {
+      "epoch": 2.2386733787385253,
+      "grad_norm": 1.5582931846027446,
+      "learning_rate": 3.186671743725812e-06,
+      "loss": 0.6227,
+      "step": 3780
+    },
+    {
+      "epoch": 2.239265620373112,
+      "grad_norm": 1.4151694573036904,
+      "learning_rate": 3.181991616028882e-06,
+      "loss": 0.6248,
+      "step": 3781
+    },
+    {
+      "epoch": 2.239857862007699,
+      "grad_norm": 1.2588101337192434,
+      "learning_rate": 3.1773142772807796e-06,
+      "loss": 0.5951,
+      "step": 3782
+    },
+    {
+      "epoch": 2.240450103642286,
+      "grad_norm": 1.4993382133597462,
+      "learning_rate": 3.172639729394795e-06,
+      "loss": 0.6159,
+      "step": 3783
+    },
+    {
+      "epoch": 2.241042345276873,
+      "grad_norm": 1.8917475122461924,
+      "learning_rate": 3.1679679742830806e-06,
+      "loss": 0.5925,
+      "step": 3784
+    },
+    {
+      "epoch": 2.24163458691146,
+      "grad_norm": 2.418861532155358,
+      "learning_rate": 3.1632990138566467e-06,
+      "loss": 0.614,
+      "step": 3785
+    },
+    {
+      "epoch": 2.242226828546047,
+      "grad_norm": 1.705180026300204,
+      "learning_rate": 3.158632850025367e-06,
+      "loss": 0.5872,
+      "step": 3786
+    },
+    {
+      "epoch": 2.2428190701806336,
+      "grad_norm": 1.377966079629573,
+      "learning_rate": 3.1539694846979594e-06,
+      "loss": 0.6421,
+      "step": 3787
+    },
+    {
+      "epoch": 2.243411311815221,
+      "grad_norm": 1.58665211047525,
+      "learning_rate": 3.1493089197820015e-06,
+      "loss": 0.6665,
+      "step": 3788
+    },
+    {
+      "epoch": 2.2440035534498075,
+      "grad_norm": 1.338148955510755,
+      "learning_rate": 3.1446511571839267e-06,
+      "loss": 0.6043,
+      "step": 3789
+    },
+    {
+      "epoch": 2.244595795084394,
+      "grad_norm": 1.1705176333063276,
+      "learning_rate": 3.139996198809028e-06,
+      "loss": 0.6098,
+      "step": 3790
+    },
+    {
+      "epoch": 2.2451880367189814,
+      "grad_norm": 1.8169627762543634,
+      "learning_rate": 3.1353440465614403e-06,
+      "loss": 0.6236,
+      "step": 3791
+    },
+    {
+      "epoch": 2.245780278353568,
+      "grad_norm": 1.291671583977258,
+      "learning_rate": 3.1306947023441524e-06,
+      "loss": 0.616,
+      "step": 3792
+    },
+    {
+      "epoch": 2.2463725199881552,
+      "grad_norm": 3.0359361674296963,
+      "learning_rate": 3.1260481680590116e-06,
+      "loss": 0.6486,
+      "step": 3793
+    },
+    {
+      "epoch": 2.246964761622742,
+      "grad_norm": 1.4138943360346141,
+      "learning_rate": 3.121404445606714e-06,
+      "loss": 0.654,
+      "step": 3794
+    },
+    {
+      "epoch": 2.247557003257329,
+      "grad_norm": 1.5503123140519945,
+      "learning_rate": 3.1167635368867997e-06,
+      "loss": 0.6364,
+      "step": 3795
+    },
+    {
+      "epoch": 2.248149244891916,
+      "grad_norm": 1.4287633207856516,
+      "learning_rate": 3.112125443797659e-06,
+      "loss": 0.5875,
+      "step": 3796
+    },
+    {
+      "epoch": 2.248741486526503,
+      "grad_norm": 2.002912860104456,
+      "learning_rate": 3.107490168236539e-06,
+      "loss": 0.6282,
+      "step": 3797
+    },
+    {
+      "epoch": 2.2493337281610897,
+      "grad_norm": 1.4230050402558123,
+      "learning_rate": 3.1028577120995216e-06,
+      "loss": 0.6012,
+      "step": 3798
+    },
+    {
+      "epoch": 2.249925969795677,
+      "grad_norm": 1.877917813203428,
+      "learning_rate": 3.0982280772815478e-06,
+      "loss": 0.6339,
+      "step": 3799
+    },
+    {
+      "epoch": 2.2505182114302635,
+      "grad_norm": 1.6288711989281626,
+      "learning_rate": 3.0936012656763937e-06,
+      "loss": 0.5788,
+      "step": 3800
+    },
+    {
+      "epoch": 2.2511104530648502,
+      "grad_norm": 1.6985228890250441,
+      "learning_rate": 3.0889772791766892e-06,
+      "loss": 0.6384,
+      "step": 3801
+    },
+    {
+      "epoch": 2.2517026946994374,
+      "grad_norm": 1.2149735698196262,
+      "learning_rate": 3.0843561196739013e-06,
+      "loss": 0.6121,
+      "step": 3802
+    },
+    {
+      "epoch": 2.252294936334024,
+      "grad_norm": 1.8781077925643568,
+      "learning_rate": 3.079737789058348e-06,
+      "loss": 0.6069,
+      "step": 3803
+    },
+    {
+      "epoch": 2.2528871779686113,
+      "grad_norm": 2.685977214125362,
+      "learning_rate": 3.075122289219181e-06,
+      "loss": 0.5838,
+      "step": 3804
+    },
+    {
+      "epoch": 2.253479419603198,
+      "grad_norm": 1.951312531114018,
+      "learning_rate": 3.0705096220444066e-06,
+      "loss": 0.5911,
+      "step": 3805
+    },
+    {
+      "epoch": 2.254071661237785,
+      "grad_norm": 1.9964640855025602,
+      "learning_rate": 3.0658997894208573e-06,
+      "loss": 0.5825,
+      "step": 3806
+    },
+    {
+      "epoch": 2.254663902872372,
+      "grad_norm": 1.4301827317146103,
+      "learning_rate": 3.0612927932342205e-06,
+      "loss": 0.5751,
+      "step": 3807
+    },
+    {
+      "epoch": 2.255256144506959,
+      "grad_norm": 1.3725748768219552,
+      "learning_rate": 3.0566886353690106e-06,
+      "loss": 0.6418,
+      "step": 3808
+    },
+    {
+      "epoch": 2.2558483861415457,
+      "grad_norm": 1.7520734317776463,
+      "learning_rate": 3.052087317708593e-06,
+      "loss": 0.6551,
+      "step": 3809
+    },
+    {
+      "epoch": 2.256440627776133,
+      "grad_norm": 1.6914593574417534,
+      "learning_rate": 3.047488842135159e-06,
+      "loss": 0.6149,
+      "step": 3810
+    },
+    {
+      "epoch": 2.2570328694107196,
+      "grad_norm": 1.15940591732544,
+      "learning_rate": 3.0428932105297516e-06,
+      "loss": 0.5848,
+      "step": 3811
+    },
+    {
+      "epoch": 2.2576251110453063,
+      "grad_norm": 1.2240564347576872,
+      "learning_rate": 3.038300424772237e-06,
+      "loss": 0.6324,
+      "step": 3812
+    },
+    {
+      "epoch": 2.2582173526798934,
+      "grad_norm": 1.4337226917688144,
+      "learning_rate": 3.0337104867413215e-06,
+      "loss": 0.678,
+      "step": 3813
+    },
+    {
+      "epoch": 2.25880959431448,
+      "grad_norm": 1.691348892207366,
+      "learning_rate": 3.0291233983145494e-06,
+      "loss": 0.6343,
+      "step": 3814
+    },
+    {
+      "epoch": 2.2594018359490673,
+      "grad_norm": 1.5404862204108754,
+      "learning_rate": 3.0245391613683027e-06,
+      "loss": 0.5813,
+      "step": 3815
+    },
+    {
+      "epoch": 2.259994077583654,
+      "grad_norm": 3.706170880625637,
+      "learning_rate": 3.019957777777788e-06,
+      "loss": 0.5958,
+      "step": 3816
+    },
+    {
+      "epoch": 2.260586319218241,
+      "grad_norm": 1.4585279326793774,
+      "learning_rate": 3.015379249417045e-06,
+      "loss": 0.6097,
+      "step": 3817
+    },
+    {
+      "epoch": 2.261178560852828,
+      "grad_norm": 1.9982994091231006,
+      "learning_rate": 3.010803578158954e-06,
+      "loss": 0.6247,
+      "step": 3818
+    },
+    {
+      "epoch": 2.261770802487415,
+      "grad_norm": 17.925157107499732,
+      "learning_rate": 3.006230765875224e-06,
+      "loss": 0.6337,
+      "step": 3819
+    },
+    {
+      "epoch": 2.2623630441220017,
+      "grad_norm": 2.42901036591496,
+      "learning_rate": 3.00166081443639e-06,
+      "loss": 0.6131,
+      "step": 3820
+    },
+    {
+      "epoch": 2.262955285756589,
+      "grad_norm": 1.4874686031510487,
+      "learning_rate": 2.997093725711815e-06,
+      "loss": 0.6154,
+      "step": 3821
+    },
+    {
+      "epoch": 2.2635475273911756,
+      "grad_norm": 1.5630584463741388,
+      "learning_rate": 2.9925295015696978e-06,
+      "loss": 0.6295,
+      "step": 3822
+    },
+    {
+      "epoch": 2.2641397690257623,
+      "grad_norm": 2.901494892160737,
+      "learning_rate": 2.987968143877068e-06,
+      "loss": 0.6562,
+      "step": 3823
+    },
+    {
+      "epoch": 2.2647320106603495,
+      "grad_norm": 1.97840113224956,
+      "learning_rate": 2.9834096544997725e-06,
+      "loss": 0.633,
+      "step": 3824
+    },
+    {
+      "epoch": 2.265324252294936,
+      "grad_norm": 2.1415089601168034,
+      "learning_rate": 2.9788540353024863e-06,
+      "loss": 0.5627,
+      "step": 3825
+    },
+    {
+      "epoch": 2.2659164939295233,
+      "grad_norm": 1.8670402712788894,
+      "learning_rate": 2.9743012881487187e-06,
+      "loss": 0.579,
+      "step": 3826
+    },
+    {
+      "epoch": 2.26650873556411,
+      "grad_norm": 2.79632330229638,
+      "learning_rate": 2.9697514149008044e-06,
+      "loss": 0.6382,
+      "step": 3827
+    },
+    {
+      "epoch": 2.267100977198697,
+      "grad_norm": 2.1965219327895813,
+      "learning_rate": 2.965204417419886e-06,
+      "loss": 0.595,
+      "step": 3828
+    },
+    {
+      "epoch": 2.267693218833284,
+      "grad_norm": 1.456269981743204,
+      "learning_rate": 2.960660297565945e-06,
+      "loss": 0.6127,
+      "step": 3829
+    },
+    {
+      "epoch": 2.268285460467871,
+      "grad_norm": 1.5711432419219262,
+      "learning_rate": 2.956119057197785e-06,
+      "loss": 0.6269,
+      "step": 3830
+    },
+    {
+      "epoch": 2.2688777021024578,
+      "grad_norm": 2.2290008657723965,
+      "learning_rate": 2.9515806981730322e-06,
+      "loss": 0.6162,
+      "step": 3831
+    },
+    {
+      "epoch": 2.269469943737045,
+      "grad_norm": 1.8539455895230874,
+      "learning_rate": 2.9470452223481206e-06,
+      "loss": 0.6062,
+      "step": 3832
+    },
+    {
+      "epoch": 2.2700621853716316,
+      "grad_norm": 1.7914148986066236,
+      "learning_rate": 2.942512631578318e-06,
+      "loss": 0.6494,
+      "step": 3833
+    },
+    {
+      "epoch": 2.2706544270062183,
+      "grad_norm": 1.579902714082181,
+      "learning_rate": 2.9379829277177152e-06,
+      "loss": 0.6123,
+      "step": 3834
+    },
+    {
+      "epoch": 2.2712466686408055,
+      "grad_norm": 1.8966069617893526,
+      "learning_rate": 2.933456112619212e-06,
+      "loss": 0.6139,
+      "step": 3835
+    },
+    {
+      "epoch": 2.271838910275392,
+      "grad_norm": 3.372987103884868,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.6225,
+      "step": 3836
+    },
+    {
+      "epoch": 2.2724311519099794,
+      "grad_norm": 2.358810973987772,
+      "learning_rate": 2.9244111561141997e-06,
+      "loss": 0.6399,
+      "step": 3837
+    },
+    {
+      "epoch": 2.273023393544566,
+      "grad_norm": 1.660137040911655,
+      "learning_rate": 2.9198930184075944e-06,
+      "loss": 0.5812,
+      "step": 3838
+    },
+    {
+      "epoch": 2.273615635179153,
+      "grad_norm": 2.200855422546957,
+      "learning_rate": 2.915377776862878e-06,
+      "loss": 0.6089,
+      "step": 3839
+    },
+    {
+      "epoch": 2.27420787681374,
+      "grad_norm": 1.6572553582906615,
+      "learning_rate": 2.9108654333270346e-06,
+      "loss": 0.5949,
+      "step": 3840
+    },
+    {
+      "epoch": 2.274800118448327,
+      "grad_norm": 1.4498432049468442,
+      "learning_rate": 2.9063559896458704e-06,
+      "loss": 0.6329,
+      "step": 3841
+    },
+    {
+      "epoch": 2.275392360082914,
+      "grad_norm": 1.1737273223504179,
+      "learning_rate": 2.901849447664008e-06,
+      "loss": 0.6561,
+      "step": 3842
+    },
+    {
+      "epoch": 2.275984601717501,
+      "grad_norm": 2.66610971272177,
+      "learning_rate": 2.897345809224864e-06,
+      "loss": 0.5836,
+      "step": 3843
+    },
+    {
+      "epoch": 2.2765768433520877,
+      "grad_norm": 1.839143446701005,
+      "learning_rate": 2.892845076170685e-06,
+      "loss": 0.6346,
+      "step": 3844
+    },
+    {
+      "epoch": 2.2771690849866744,
+      "grad_norm": 1.60827795865389,
+      "learning_rate": 2.8883472503425236e-06,
+      "loss": 0.6318,
+      "step": 3845
+    },
+    {
+      "epoch": 2.2777613266212615,
+      "grad_norm": 1.5301557763688713,
+      "learning_rate": 2.8838523335802525e-06,
+      "loss": 0.6759,
+      "step": 3846
+    },
+    {
+      "epoch": 2.2783535682558482,
+      "grad_norm": 1.5562091375300113,
+      "learning_rate": 2.8793603277225302e-06,
+      "loss": 0.6203,
+      "step": 3847
+    },
+    {
+      "epoch": 2.2789458098904354,
+      "grad_norm": 1.588136398742716,
+      "learning_rate": 2.8748712346068464e-06,
+      "loss": 0.5912,
+      "step": 3848
+    },
+    {
+      "epoch": 2.279538051525022,
+      "grad_norm": 1.841240319682772,
+      "learning_rate": 2.8703850560694966e-06,
+      "loss": 0.6484,
+      "step": 3849
+    },
+    {
+      "epoch": 2.2801302931596092,
+      "grad_norm": 8.661704543502498,
+      "learning_rate": 2.865901793945576e-06,
+      "loss": 0.6019,
+      "step": 3850
+    },
+    {
+      "epoch": 2.280722534794196,
+      "grad_norm": 1.9390476098819909,
+      "learning_rate": 2.8614214500689886e-06,
+      "loss": 0.6344,
+      "step": 3851
+    },
+    {
+      "epoch": 2.281314776428783,
+      "grad_norm": 1.4608667959226727,
+      "learning_rate": 2.8569440262724502e-06,
+      "loss": 0.608,
+      "step": 3852
+    },
+    {
+      "epoch": 2.28190701806337,
+      "grad_norm": 2.253638236763826,
+      "learning_rate": 2.8524695243874814e-06,
+      "loss": 0.6115,
+      "step": 3853
+    },
+    {
+      "epoch": 2.282499259697957,
+      "grad_norm": 1.5747389470810782,
+      "learning_rate": 2.8479979462444017e-06,
+      "loss": 0.6291,
+      "step": 3854
+    },
+    {
+      "epoch": 2.2830915013325437,
+      "grad_norm": 1.216195294388743,
+      "learning_rate": 2.8435292936723356e-06,
+      "loss": 0.6106,
+      "step": 3855
+    },
+    {
+      "epoch": 2.2836837429671304,
+      "grad_norm": 2.7276242343011745,
+      "learning_rate": 2.8390635684992163e-06,
+      "loss": 0.5509,
+      "step": 3856
+    },
+    {
+      "epoch": 2.2842759846017175,
+      "grad_norm": 2.323606613377955,
+      "learning_rate": 2.83460077255178e-06,
+      "loss": 0.5926,
+      "step": 3857
+    },
+    {
+      "epoch": 2.2848682262363043,
+      "grad_norm": 2.38216532290336,
+      "learning_rate": 2.8301409076555574e-06,
+      "loss": 0.6453,
+      "step": 3858
+    },
+    {
+      "epoch": 2.2854604678708914,
+      "grad_norm": 1.3563075087072833,
+      "learning_rate": 2.8256839756348807e-06,
+      "loss": 0.6039,
+      "step": 3859
+    },
+    {
+      "epoch": 2.286052709505478,
+      "grad_norm": 1.2668051387431,
+      "learning_rate": 2.821229978312889e-06,
+      "loss": 0.5929,
+      "step": 3860
+    },
+    {
+      "epoch": 2.2866449511400653,
+      "grad_norm": 1.2956377312836582,
+      "learning_rate": 2.8167789175115223e-06,
+      "loss": 0.6503,
+      "step": 3861
+    },
+    {
+      "epoch": 2.287237192774652,
+      "grad_norm": 1.9302456323206396,
+      "learning_rate": 2.8123307950515087e-06,
+      "loss": 0.6626,
+      "step": 3862
+    },
+    {
+      "epoch": 2.287829434409239,
+      "grad_norm": 1.6044535773095645,
+      "learning_rate": 2.80788561275238e-06,
+      "loss": 0.5979,
+      "step": 3863
+    },
+    {
+      "epoch": 2.288421676043826,
+      "grad_norm": 2.2939252755635975,
+      "learning_rate": 2.8034433724324716e-06,
+      "loss": 0.6646,
+      "step": 3864
+    },
+    {
+      "epoch": 2.289013917678413,
+      "grad_norm": 2.471817823244044,
+      "learning_rate": 2.7990040759089022e-06,
+      "loss": 0.5829,
+      "step": 3865
+    },
+    {
+      "epoch": 2.2896061593129997,
+      "grad_norm": 3.3806573958288566,
+      "learning_rate": 2.7945677249976e-06,
+      "loss": 0.6553,
+      "step": 3866
+    },
+    {
+      "epoch": 2.2901984009475864,
+      "grad_norm": 2.737369171178317,
+      "learning_rate": 2.7901343215132758e-06,
+      "loss": 0.6349,
+      "step": 3867
+    },
+    {
+      "epoch": 2.2907906425821736,
+      "grad_norm": 2.298571285772045,
+      "learning_rate": 2.7857038672694492e-06,
+      "loss": 0.5691,
+      "step": 3868
+    },
+    {
+      "epoch": 2.2913828842167603,
+      "grad_norm": 1.5520258709926762,
+      "learning_rate": 2.7812763640784155e-06,
+      "loss": 0.6444,
+      "step": 3869
+    },
+    {
+      "epoch": 2.2919751258513474,
+      "grad_norm": 1.7364947241082105,
+      "learning_rate": 2.776851813751281e-06,
+      "loss": 0.6432,
+      "step": 3870
+    },
+    {
+      "epoch": 2.292567367485934,
+      "grad_norm": 1.5769815666091875,
+      "learning_rate": 2.77243021809793e-06,
+      "loss": 0.6317,
+      "step": 3871
+    },
+    {
+      "epoch": 2.2931596091205213,
+      "grad_norm": 2.137316840538586,
+      "learning_rate": 2.7680115789270478e-06,
+      "loss": 0.6133,
+      "step": 3872
+    },
+    {
+      "epoch": 2.293751850755108,
+      "grad_norm": 2.816072721070139,
+      "learning_rate": 2.763595898046101e-06,
+      "loss": 0.562,
+      "step": 3873
+    },
+    {
+      "epoch": 2.294344092389695,
+      "grad_norm": 1.3276083688656188,
+      "learning_rate": 2.7591831772613576e-06,
+      "loss": 0.5834,
+      "step": 3874
+    },
+    {
+      "epoch": 2.294936334024282,
+      "grad_norm": 1.2689499619456701,
+      "learning_rate": 2.754773418377863e-06,
+      "loss": 0.6007,
+      "step": 3875
+    },
+    {
+      "epoch": 2.295528575658869,
+      "grad_norm": 3.840896322229496,
+      "learning_rate": 2.750366623199462e-06,
+      "loss": 0.6271,
+      "step": 3876
+    },
+    {
+      "epoch": 2.2961208172934557,
+      "grad_norm": 1.1950201080078542,
+      "learning_rate": 2.745962793528775e-06,
+      "loss": 0.585,
+      "step": 3877
+    },
+    {
+      "epoch": 2.2967130589280425,
+      "grad_norm": 2.0462234251989484,
+      "learning_rate": 2.7415619311672236e-06,
+      "loss": 0.639,
+      "step": 3878
+    },
+    {
+      "epoch": 2.2973053005626296,
+      "grad_norm": 1.853413330196118,
+      "learning_rate": 2.7371640379150032e-06,
+      "loss": 0.6103,
+      "step": 3879
+    },
+    {
+      "epoch": 2.2978975421972163,
+      "grad_norm": 2.1374274208984647,
+      "learning_rate": 2.7327691155710978e-06,
+      "loss": 0.5992,
+      "step": 3880
+    },
+    {
+      "epoch": 2.2984897838318035,
+      "grad_norm": 1.9972945709906602,
+      "learning_rate": 2.7283771659332805e-06,
+      "loss": 0.62,
+      "step": 3881
+    },
+    {
+      "epoch": 2.29908202546639,
+      "grad_norm": 1.0982870260711513,
+      "learning_rate": 2.723988190798108e-06,
+      "loss": 0.6119,
+      "step": 3882
+    },
+    {
+      "epoch": 2.2996742671009773,
+      "grad_norm": 1.2876134638144012,
+      "learning_rate": 2.7196021919609163e-06,
+      "loss": 0.6366,
+      "step": 3883
+    },
+    {
+      "epoch": 2.300266508735564,
+      "grad_norm": 1.8016509667404628,
+      "learning_rate": 2.7152191712158207e-06,
+      "loss": 0.6041,
+      "step": 3884
+    },
+    {
+      "epoch": 2.300858750370151,
+      "grad_norm": 1.2850926199615083,
+      "learning_rate": 2.710839130355727e-06,
+      "loss": 0.6157,
+      "step": 3885
+    },
+    {
+      "epoch": 2.301450992004738,
+      "grad_norm": 1.4731799841992599,
+      "learning_rate": 2.706462071172322e-06,
+      "loss": 0.5742,
+      "step": 3886
+    },
+    {
+      "epoch": 2.302043233639325,
+      "grad_norm": 1.7403825798627295,
+      "learning_rate": 2.7020879954560642e-06,
+      "loss": 0.6451,
+      "step": 3887
+    },
+    {
+      "epoch": 2.3026354752739118,
+      "grad_norm": 3.5575301409939653,
+      "learning_rate": 2.697716904996196e-06,
+      "loss": 0.6125,
+      "step": 3888
+    },
+    {
+      "epoch": 2.3032277169084985,
+      "grad_norm": 1.283131812355814,
+      "learning_rate": 2.6933488015807406e-06,
+      "loss": 0.6057,
+      "step": 3889
+    },
+    {
+      "epoch": 2.3038199585430856,
+      "grad_norm": 3.1172645215689783,
+      "learning_rate": 2.6889836869965016e-06,
+      "loss": 0.6172,
+      "step": 3890
+    },
+    {
+      "epoch": 2.3044122001776723,
+      "grad_norm": 2.218448251870392,
+      "learning_rate": 2.6846215630290516e-06,
+      "loss": 0.6496,
+      "step": 3891
+    },
+    {
+      "epoch": 2.3050044418122595,
+      "grad_norm": 2.3254463534891796,
+      "learning_rate": 2.6802624314627436e-06,
+      "loss": 0.6223,
+      "step": 3892
+    },
+    {
+      "epoch": 2.305596683446846,
+      "grad_norm": 1.2275387722818925,
+      "learning_rate": 2.67590629408071e-06,
+      "loss": 0.5998,
+      "step": 3893
+    },
+    {
+      "epoch": 2.3061889250814334,
+      "grad_norm": 3.0555438618968584,
+      "learning_rate": 2.6715531526648585e-06,
+      "loss": 0.6273,
+      "step": 3894
+    },
+    {
+      "epoch": 2.30678116671602,
+      "grad_norm": 1.5752564674463383,
+      "learning_rate": 2.6672030089958668e-06,
+      "loss": 0.6596,
+      "step": 3895
+    },
+    {
+      "epoch": 2.3073734083506072,
+      "grad_norm": 1.7511275969987508,
+      "learning_rate": 2.6628558648531845e-06,
+      "loss": 0.5788,
+      "step": 3896
+    },
+    {
+      "epoch": 2.307965649985194,
+      "grad_norm": 0.9655204456900459,
+      "learning_rate": 2.6585117220150403e-06,
+      "loss": 0.639,
+      "step": 3897
+    },
+    {
+      "epoch": 2.308557891619781,
+      "grad_norm": 1.1990519232172137,
+      "learning_rate": 2.654170582258441e-06,
+      "loss": 0.5954,
+      "step": 3898
+    },
+    {
+      "epoch": 2.309150133254368,
+      "grad_norm": 3.208547550346025,
+      "learning_rate": 2.649832447359142e-06,
+      "loss": 0.6034,
+      "step": 3899
+    },
+    {
+      "epoch": 2.3097423748889545,
+      "grad_norm": 2.2300260612204577,
+      "learning_rate": 2.645497319091692e-06,
+      "loss": 0.5987,
+      "step": 3900
+    },
+    {
+      "epoch": 2.3103346165235417,
+      "grad_norm": 2.0946840308163543,
+      "learning_rate": 2.6411651992294065e-06,
+      "loss": 0.6319,
+      "step": 3901
+    },
+    {
+      "epoch": 2.3109268581581284,
+      "grad_norm": 2.079061501961239,
+      "learning_rate": 2.63683608954436e-06,
+      "loss": 0.636,
+      "step": 3902
+    },
+    {
+      "epoch": 2.3115190997927155,
+      "grad_norm": 1.2445711582719796,
+      "learning_rate": 2.6325099918074017e-06,
+      "loss": 0.6395,
+      "step": 3903
+    },
+    {
+      "epoch": 2.3121113414273022,
+      "grad_norm": 1.864032684312602,
+      "learning_rate": 2.6281869077881507e-06,
+      "loss": 0.5782,
+      "step": 3904
+    },
+    {
+      "epoch": 2.3127035830618894,
+      "grad_norm": 1.3316080049678458,
+      "learning_rate": 2.6238668392549947e-06,
+      "loss": 0.5901,
+      "step": 3905
+    },
+    {
+      "epoch": 2.313295824696476,
+      "grad_norm": 1.6276785587062519,
+      "learning_rate": 2.619549787975081e-06,
+      "loss": 0.6372,
+      "step": 3906
+    },
+    {
+      "epoch": 2.3138880663310633,
+      "grad_norm": 5.7918361147135045,
+      "learning_rate": 2.615235755714324e-06,
+      "loss": 0.6331,
+      "step": 3907
+    },
+    {
+      "epoch": 2.31448030796565,
+      "grad_norm": 2.2663927095141925,
+      "learning_rate": 2.6109247442374088e-06,
+      "loss": 0.6412,
+      "step": 3908
+    },
+    {
+      "epoch": 2.315072549600237,
+      "grad_norm": 7.124389289252508,
+      "learning_rate": 2.6066167553077826e-06,
+      "loss": 0.5689,
+      "step": 3909
+    },
+    {
+      "epoch": 2.315664791234824,
+      "grad_norm": 1.363067404396117,
+      "learning_rate": 2.602311790687655e-06,
+      "loss": 0.6523,
+      "step": 3910
+    },
+    {
+      "epoch": 2.3162570328694105,
+      "grad_norm": 1.5813779524915303,
+      "learning_rate": 2.5980098521379936e-06,
+      "loss": 0.6186,
+      "step": 3911
+    },
+    {
+      "epoch": 2.3168492745039977,
+      "grad_norm": 1.481882250550122,
+      "learning_rate": 2.593710941418537e-06,
+      "loss": 0.5903,
+      "step": 3912
+    },
+    {
+      "epoch": 2.3174415161385844,
+      "grad_norm": 1.9025600054836969,
+      "learning_rate": 2.5894150602877834e-06,
+      "loss": 0.6135,
+      "step": 3913
+    },
+    {
+      "epoch": 2.3180337577731716,
+      "grad_norm": 1.2895740270749483,
+      "learning_rate": 2.585122210502987e-06,
+      "loss": 0.5903,
+      "step": 3914
+    },
+    {
+      "epoch": 2.3186259994077583,
+      "grad_norm": 1.647388330853826,
+      "learning_rate": 2.5808323938201642e-06,
+      "loss": 0.5979,
+      "step": 3915
+    },
+    {
+      "epoch": 2.3192182410423454,
+      "grad_norm": 1.28519854345509,
+      "learning_rate": 2.5765456119940933e-06,
+      "loss": 0.6295,
+      "step": 3916
+    },
+    {
+      "epoch": 2.319810482676932,
+      "grad_norm": 1.2525366592486644,
+      "learning_rate": 2.5722618667783063e-06,
+      "loss": 0.5877,
+      "step": 3917
+    },
+    {
+      "epoch": 2.3204027243115193,
+      "grad_norm": 1.3317926678190848,
+      "learning_rate": 2.5679811599251003e-06,
+      "loss": 0.5926,
+      "step": 3918
+    },
+    {
+      "epoch": 2.320994965946106,
+      "grad_norm": 2.428793102384277,
+      "learning_rate": 2.5637034931855197e-06,
+      "loss": 0.5911,
+      "step": 3919
+    },
+    {
+      "epoch": 2.321587207580693,
+      "grad_norm": 1.4721974585666002,
+      "learning_rate": 2.559428868309377e-06,
+      "loss": 0.579,
+      "step": 3920
+    },
+    {
+      "epoch": 2.32217944921528,
+      "grad_norm": 3.8645920617049465,
+      "learning_rate": 2.5551572870452268e-06,
+      "loss": 0.6181,
+      "step": 3921
+    },
+    {
+      "epoch": 2.3227716908498666,
+      "grad_norm": 1.5226219139675943,
+      "learning_rate": 2.5508887511403936e-06,
+      "loss": 0.5791,
+      "step": 3922
+    },
+    {
+      "epoch": 2.3233639324844537,
+      "grad_norm": 2.2075587892085116,
+      "learning_rate": 2.5466232623409416e-06,
+      "loss": 0.634,
+      "step": 3923
+    },
+    {
+      "epoch": 2.3239561741190404,
+      "grad_norm": 1.3599172487421667,
+      "learning_rate": 2.542360822391702e-06,
+      "loss": 0.6161,
+      "step": 3924
+    },
+    {
+      "epoch": 2.3245484157536276,
+      "grad_norm": 2.3033687836929557,
+      "learning_rate": 2.538101433036246e-06,
+      "loss": 0.6239,
+      "step": 3925
+    },
+    {
+      "epoch": 2.3251406573882143,
+      "grad_norm": 1.9111808167290096,
+      "learning_rate": 2.5338450960169105e-06,
+      "loss": 0.5993,
+      "step": 3926
+    },
+    {
+      "epoch": 2.3257328990228014,
+      "grad_norm": 3.326963232812606,
+      "learning_rate": 2.52959181307477e-06,
+      "loss": 0.6112,
+      "step": 3927
+    },
+    {
+      "epoch": 2.326325140657388,
+      "grad_norm": 1.796700647763413,
+      "learning_rate": 2.525341585949662e-06,
+      "loss": 0.6028,
+      "step": 3928
+    },
+    {
+      "epoch": 2.3269173822919753,
+      "grad_norm": 2.312629837062101,
+      "learning_rate": 2.521094416380162e-06,
+      "loss": 0.6497,
+      "step": 3929
+    },
+    {
+      "epoch": 2.327509623926562,
+      "grad_norm": 1.2457487465053927,
+      "learning_rate": 2.5168503061036086e-06,
+      "loss": 0.6098,
+      "step": 3930
+    },
+    {
+      "epoch": 2.328101865561149,
+      "grad_norm": 3.791923080168204,
+      "learning_rate": 2.5126092568560754e-06,
+      "loss": 0.5886,
+      "step": 3931
+    },
+    {
+      "epoch": 2.328694107195736,
+      "grad_norm": 4.853398587684118,
+      "learning_rate": 2.5083712703723952e-06,
+      "loss": 0.6286,
+      "step": 3932
+    },
+    {
+      "epoch": 2.3292863488303226,
+      "grad_norm": 2.2593203003697915,
+      "learning_rate": 2.5041363483861357e-06,
+      "loss": 0.6453,
+      "step": 3933
+    },
+    {
+      "epoch": 2.3298785904649097,
+      "grad_norm": 1.6615315018284669,
+      "learning_rate": 2.499904492629627e-06,
+      "loss": 0.5813,
+      "step": 3934
+    },
+    {
+      "epoch": 2.3304708320994965,
+      "grad_norm": 1.6816787030820344,
+      "learning_rate": 2.4956757048339307e-06,
+      "loss": 0.5839,
+      "step": 3935
+    },
+    {
+      "epoch": 2.3310630737340836,
+      "grad_norm": 2.769364650023261,
+      "learning_rate": 2.4914499867288577e-06,
+      "loss": 0.6956,
+      "step": 3936
+    },
+    {
+      "epoch": 2.3316553153686703,
+      "grad_norm": 3.642860545017731,
+      "learning_rate": 2.487227340042966e-06,
+      "loss": 0.6321,
+      "step": 3937
+    },
+    {
+      "epoch": 2.3322475570032575,
+      "grad_norm": 1.3381965332512193,
+      "learning_rate": 2.483007766503558e-06,
+      "loss": 0.5996,
+      "step": 3938
+    },
+    {
+      "epoch": 2.332839798637844,
+      "grad_norm": 1.7008693234557766,
+      "learning_rate": 2.4787912678366755e-06,
+      "loss": 0.593,
+      "step": 3939
+    },
+    {
+      "epoch": 2.3334320402724313,
+      "grad_norm": 1.3507684647991987,
+      "learning_rate": 2.474577845767099e-06,
+      "loss": 0.6144,
+      "step": 3940
+    },
+    {
+      "epoch": 2.334024281907018,
+      "grad_norm": 2.0794417313783233,
+      "learning_rate": 2.4703675020183583e-06,
+      "loss": 0.5919,
+      "step": 3941
+    },
+    {
+      "epoch": 2.334616523541605,
+      "grad_norm": 2.4036165381923915,
+      "learning_rate": 2.4661602383127235e-06,
+      "loss": 0.6163,
+      "step": 3942
+    },
+    {
+      "epoch": 2.335208765176192,
+      "grad_norm": 2.006173079525012,
+      "learning_rate": 2.461956056371201e-06,
+      "loss": 0.6006,
+      "step": 3943
+    },
+    {
+      "epoch": 2.3358010068107786,
+      "grad_norm": 2.5021972165514677,
+      "learning_rate": 2.4577549579135318e-06,
+      "loss": 0.6524,
+      "step": 3944
+    },
+    {
+      "epoch": 2.3363932484453658,
+      "grad_norm": 2.2918385752916772,
+      "learning_rate": 2.453556944658206e-06,
+      "loss": 0.6192,
+      "step": 3945
+    },
+    {
+      "epoch": 2.3369854900799525,
+      "grad_norm": 2.0313012400244546,
+      "learning_rate": 2.449362018322451e-06,
+      "loss": 0.5882,
+      "step": 3946
+    },
+    {
+      "epoch": 2.3375777317145396,
+      "grad_norm": 2.1865753540293245,
+      "learning_rate": 2.445170180622223e-06,
+      "loss": 0.5776,
+      "step": 3947
+    },
+    {
+      "epoch": 2.3381699733491264,
+      "grad_norm": 1.6788725756694558,
+      "learning_rate": 2.440981433272216e-06,
+      "loss": 0.6286,
+      "step": 3948
+    },
+    {
+      "epoch": 2.3387622149837135,
+      "grad_norm": 1.3993996393149903,
+      "learning_rate": 2.4367957779858675e-06,
+      "loss": 0.6201,
+      "step": 3949
+    },
+    {
+      "epoch": 2.3393544566183,
+      "grad_norm": 2.5163778162713877,
+      "learning_rate": 2.43261321647535e-06,
+      "loss": 0.6333,
+      "step": 3950
+    },
+    {
+      "epoch": 2.3399466982528874,
+      "grad_norm": 2.1183056118895647,
+      "learning_rate": 2.4284337504515577e-06,
+      "loss": 0.6251,
+      "step": 3951
+    },
+    {
+      "epoch": 2.340538939887474,
+      "grad_norm": 2.9285242355007957,
+      "learning_rate": 2.42425738162413e-06,
+      "loss": 0.6026,
+      "step": 3952
+    },
+    {
+      "epoch": 2.3411311815220612,
+      "grad_norm": 2.1726586914478,
+      "learning_rate": 2.420084111701442e-06,
+      "loss": 0.6219,
+      "step": 3953
+    },
+    {
+      "epoch": 2.341723423156648,
+      "grad_norm": 1.7766065564148414,
+      "learning_rate": 2.4159139423905898e-06,
+      "loss": 0.6071,
+      "step": 3954
+    },
+    {
+      "epoch": 2.3423156647912347,
+      "grad_norm": 1.798438650040322,
+      "learning_rate": 2.411746875397407e-06,
+      "loss": 0.5952,
+      "step": 3955
+    },
+    {
+      "epoch": 2.342907906425822,
+      "grad_norm": 1.8115265686116828,
+      "learning_rate": 2.4075829124264606e-06,
+      "loss": 0.6037,
+      "step": 3956
+    },
+    {
+      "epoch": 2.3435001480604085,
+      "grad_norm": 1.2886622921652453,
+      "learning_rate": 2.4034220551810484e-06,
+      "loss": 0.6178,
+      "step": 3957
+    },
+    {
+      "epoch": 2.3440923896949957,
+      "grad_norm": 1.2555330420863857,
+      "learning_rate": 2.3992643053631904e-06,
+      "loss": 0.617,
+      "step": 3958
+    },
+    {
+      "epoch": 2.3446846313295824,
+      "grad_norm": 1.6387727645080599,
+      "learning_rate": 2.3951096646736403e-06,
+      "loss": 0.5982,
+      "step": 3959
+    },
+    {
+      "epoch": 2.3452768729641695,
+      "grad_norm": 1.2682217733893184,
+      "learning_rate": 2.3909581348118803e-06,
+      "loss": 0.6339,
+      "step": 3960
+    },
+    {
+      "epoch": 2.3458691145987562,
+      "grad_norm": 1.3586475278862904,
+      "learning_rate": 2.386809717476123e-06,
+      "loss": 0.6429,
+      "step": 3961
+    },
+    {
+      "epoch": 2.3464613562333434,
+      "grad_norm": 1.3204559918998942,
+      "learning_rate": 2.3826644143633017e-06,
+      "loss": 0.5899,
+      "step": 3962
+    },
+    {
+      "epoch": 2.34705359786793,
+      "grad_norm": 1.3928626985319164,
+      "learning_rate": 2.3785222271690754e-06,
+      "loss": 0.6113,
+      "step": 3963
+    },
+    {
+      "epoch": 2.3476458395025173,
+      "grad_norm": 9.312328145389811,
+      "learning_rate": 2.3743831575878352e-06,
+      "loss": 0.6527,
+      "step": 3964
+    },
+    {
+      "epoch": 2.348238081137104,
+      "grad_norm": 3.170276661797359,
+      "learning_rate": 2.370247207312695e-06,
+      "loss": 0.6355,
+      "step": 3965
+    },
+    {
+      "epoch": 2.3488303227716907,
+      "grad_norm": 2.153769559929862,
+      "learning_rate": 2.366114378035489e-06,
+      "loss": 0.5854,
+      "step": 3966
+    },
+    {
+      "epoch": 2.349422564406278,
+      "grad_norm": 2.010875700018389,
+      "learning_rate": 2.361984671446773e-06,
+      "loss": 0.6136,
+      "step": 3967
+    },
+    {
+      "epoch": 2.3500148060408645,
+      "grad_norm": 1.8883203647412998,
+      "learning_rate": 2.3578580892358337e-06,
+      "loss": 0.6464,
+      "step": 3968
+    },
+    {
+      "epoch": 2.3506070476754517,
+      "grad_norm": 1.3913502661830128,
+      "learning_rate": 2.3537346330906776e-06,
+      "loss": 0.6297,
+      "step": 3969
+    },
+    {
+      "epoch": 2.3511992893100384,
+      "grad_norm": 1.5332485123384902,
+      "learning_rate": 2.3496143046980256e-06,
+      "loss": 0.5928,
+      "step": 3970
+    },
+    {
+      "epoch": 2.3517915309446256,
+      "grad_norm": 1.3001291166457274,
+      "learning_rate": 2.345497105743323e-06,
+      "loss": 0.6043,
+      "step": 3971
+    },
+    {
+      "epoch": 2.3523837725792123,
+      "grad_norm": 1.4754086076781734,
+      "learning_rate": 2.3413830379107395e-06,
+      "loss": 0.6199,
+      "step": 3972
+    },
+    {
+      "epoch": 2.352976014213799,
+      "grad_norm": 2.409278659416967,
+      "learning_rate": 2.337272102883157e-06,
+      "loss": 0.6453,
+      "step": 3973
+    },
+    {
+      "epoch": 2.353568255848386,
+      "grad_norm": 2.3466875892630683,
+      "learning_rate": 2.3331643023421813e-06,
+      "loss": 0.6564,
+      "step": 3974
+    },
+    {
+      "epoch": 2.3541604974829733,
+      "grad_norm": 2.714570742335837,
+      "learning_rate": 2.329059637968132e-06,
+      "loss": 0.6404,
+      "step": 3975
+    },
+    {
+      "epoch": 2.35475273911756,
+      "grad_norm": 2.3929716670625356,
+      "learning_rate": 2.324958111440051e-06,
+      "loss": 0.6362,
+      "step": 3976
+    },
+    {
+      "epoch": 2.3553449807521467,
+      "grad_norm": 1.6228125008698313,
+      "learning_rate": 2.3208597244356867e-06,
+      "loss": 0.5842,
+      "step": 3977
+    },
+    {
+      "epoch": 2.355937222386734,
+      "grad_norm": 3.384644411190184,
+      "learning_rate": 2.316764478631518e-06,
+      "loss": 0.627,
+      "step": 3978
+    },
+    {
+      "epoch": 2.3565294640213206,
+      "grad_norm": 1.5260570773569566,
+      "learning_rate": 2.3126723757027245e-06,
+      "loss": 0.6295,
+      "step": 3979
+    },
+    {
+      "epoch": 2.3571217056559077,
+      "grad_norm": 1.8203121426278104,
+      "learning_rate": 2.30858341732321e-06,
+      "loss": 0.6369,
+      "step": 3980
+    },
+    {
+      "epoch": 2.3577139472904944,
+      "grad_norm": 1.1006202274680998,
+      "learning_rate": 2.3044976051655854e-06,
+      "loss": 0.6267,
+      "step": 3981
+    },
+    {
+      "epoch": 2.3583061889250816,
+      "grad_norm": 1.8643488805973478,
+      "learning_rate": 2.300414940901182e-06,
+      "loss": 0.6002,
+      "step": 3982
+    },
+    {
+      "epoch": 2.3588984305596683,
+      "grad_norm": 1.4796650275447034,
+      "learning_rate": 2.2963354262000335e-06,
+      "loss": 0.6101,
+      "step": 3983
+    },
+    {
+      "epoch": 2.359490672194255,
+      "grad_norm": 3.064100269638554,
+      "learning_rate": 2.292259062730897e-06,
+      "loss": 0.5567,
+      "step": 3984
+    },
+    {
+      "epoch": 2.360082913828842,
+      "grad_norm": 1.5811392863582345,
+      "learning_rate": 2.2881858521612275e-06,
+      "loss": 0.6186,
+      "step": 3985
+    },
+    {
+      "epoch": 2.3606751554634293,
+      "grad_norm": 2.1586453620244597,
+      "learning_rate": 2.2841157961572034e-06,
+      "loss": 0.6014,
+      "step": 3986
+    },
+    {
+      "epoch": 2.361267397098016,
+      "grad_norm": 2.641507002266892,
+      "learning_rate": 2.2800488963837043e-06,
+      "loss": 0.6133,
+      "step": 3987
+    },
+    {
+      "epoch": 2.3618596387326027,
+      "grad_norm": 1.1730513592343241,
+      "learning_rate": 2.2759851545043175e-06,
+      "loss": 0.6223,
+      "step": 3988
+    },
+    {
+      "epoch": 2.36245188036719,
+      "grad_norm": 1.4092715716805935,
+      "learning_rate": 2.2719245721813455e-06,
+      "loss": 0.6399,
+      "step": 3989
+    },
+    {
+      "epoch": 2.3630441220017766,
+      "grad_norm": 1.8498348981776163,
+      "learning_rate": 2.2678671510757953e-06,
+      "loss": 0.6188,
+      "step": 3990
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 2.143878596825647,
+      "learning_rate": 2.263812892847381e-06,
+      "loss": 0.6321,
+      "step": 3991
+    },
+    {
+      "epoch": 2.3642286052709505,
+      "grad_norm": 1.2933879029907467,
+      "learning_rate": 2.259761799154516e-06,
+      "loss": 0.6037,
+      "step": 3992
+    },
+    {
+      "epoch": 2.3648208469055376,
+      "grad_norm": 1.5882145664047522,
+      "learning_rate": 2.2557138716543316e-06,
+      "loss": 0.5955,
+      "step": 3993
+    },
+    {
+      "epoch": 2.3654130885401243,
+      "grad_norm": 1.686459234669479,
+      "learning_rate": 2.251669112002657e-06,
+      "loss": 0.6064,
+      "step": 3994
+    },
+    {
+      "epoch": 2.366005330174711,
+      "grad_norm": 1.7498942595889782,
+      "learning_rate": 2.2476275218540266e-06,
+      "loss": 0.5981,
+      "step": 3995
+    },
+    {
+      "epoch": 2.366597571809298,
+      "grad_norm": 2.8193121891643997,
+      "learning_rate": 2.243589102861673e-06,
+      "loss": 0.6234,
+      "step": 3996
+    },
+    {
+      "epoch": 2.3671898134438853,
+      "grad_norm": 1.4594782573094889,
+      "learning_rate": 2.239553856677541e-06,
+      "loss": 0.6007,
+      "step": 3997
+    },
+    {
+      "epoch": 2.367782055078472,
+      "grad_norm": 1.485059207419929,
+      "learning_rate": 2.235521784952275e-06,
+      "loss": 0.6011,
+      "step": 3998
+    },
+    {
+      "epoch": 2.3683742967130588,
+      "grad_norm": 1.3309379388804508,
+      "learning_rate": 2.231492889335217e-06,
+      "loss": 0.5943,
+      "step": 3999
+    },
+    {
+      "epoch": 2.368966538347646,
+      "grad_norm": 2.6267695132329454,
+      "learning_rate": 2.227467171474409e-06,
+      "loss": 0.6136,
+      "step": 4000
+    },
+    {
+      "epoch": 2.3695587799822326,
+      "grad_norm": 4.446529481550082,
+      "learning_rate": 2.223444633016597e-06,
+      "loss": 0.6042,
+      "step": 4001
+    },
+    {
+      "epoch": 2.37015102161682,
+      "grad_norm": 1.9069444415000925,
+      "learning_rate": 2.2194252756072343e-06,
+      "loss": 0.6064,
+      "step": 4002
+    },
+    {
+      "epoch": 2.3707432632514065,
+      "grad_norm": 1.8238334546399355,
+      "learning_rate": 2.2154091008904497e-06,
+      "loss": 0.6344,
+      "step": 4003
+    },
+    {
+      "epoch": 2.3713355048859937,
+      "grad_norm": 2.5613993978909693,
+      "learning_rate": 2.2113961105090933e-06,
+      "loss": 0.5878,
+      "step": 4004
+    },
+    {
+      "epoch": 2.3719277465205804,
+      "grad_norm": 1.2783779891008764,
+      "learning_rate": 2.207386306104701e-06,
+      "loss": 0.6141,
+      "step": 4005
+    },
+    {
+      "epoch": 2.372519988155167,
+      "grad_norm": 1.4416435791183768,
+      "learning_rate": 2.2033796893175152e-06,
+      "loss": 0.6002,
+      "step": 4006
+    },
+    {
+      "epoch": 2.3731122297897542,
+      "grad_norm": 1.6055773889827405,
+      "learning_rate": 2.1993762617864555e-06,
+      "loss": 0.6106,
+      "step": 4007
+    },
+    {
+      "epoch": 2.3737044714243414,
+      "grad_norm": 1.33887538582045,
+      "learning_rate": 2.195376025149156e-06,
+      "loss": 0.6547,
+      "step": 4008
+    },
+    {
+      "epoch": 2.374296713058928,
+      "grad_norm": 1.7979827304398968,
+      "learning_rate": 2.1913789810419393e-06,
+      "loss": 0.5853,
+      "step": 4009
+    },
+    {
+      "epoch": 2.374888954693515,
+      "grad_norm": 1.678514747907677,
+      "learning_rate": 2.1873851310998194e-06,
+      "loss": 0.6229,
+      "step": 4010
+    },
+    {
+      "epoch": 2.375481196328102,
+      "grad_norm": 12.535062449115955,
+      "learning_rate": 2.183394476956504e-06,
+      "loss": 0.636,
+      "step": 4011
+    },
+    {
+      "epoch": 2.3760734379626887,
+      "grad_norm": 3.6917293032690623,
+      "learning_rate": 2.179407020244395e-06,
+      "loss": 0.6194,
+      "step": 4012
+    },
+    {
+      "epoch": 2.376665679597276,
+      "grad_norm": 1.3674840173653329,
+      "learning_rate": 2.175422762594591e-06,
+      "loss": 0.6274,
+      "step": 4013
+    },
+    {
+      "epoch": 2.3772579212318625,
+      "grad_norm": 1.7064037619998782,
+      "learning_rate": 2.1714417056368752e-06,
+      "loss": 0.6281,
+      "step": 4014
+    },
+    {
+      "epoch": 2.3778501628664497,
+      "grad_norm": 4.498067534083709,
+      "learning_rate": 2.167463850999719e-06,
+      "loss": 0.6391,
+      "step": 4015
+    },
+    {
+      "epoch": 2.3784424045010364,
+      "grad_norm": 3.0010927370986615,
+      "learning_rate": 2.1634892003102935e-06,
+      "loss": 0.6198,
+      "step": 4016
+    },
+    {
+      "epoch": 2.379034646135623,
+      "grad_norm": 2.2759248099052205,
+      "learning_rate": 2.159517755194456e-06,
+      "loss": 0.6536,
+      "step": 4017
+    },
+    {
+      "epoch": 2.3796268877702103,
+      "grad_norm": 1.5564439174661013,
+      "learning_rate": 2.155549517276747e-06,
+      "loss": 0.6276,
+      "step": 4018
+    },
+    {
+      "epoch": 2.3802191294047974,
+      "grad_norm": 1.6225752200568844,
+      "learning_rate": 2.1515844881803993e-06,
+      "loss": 0.6166,
+      "step": 4019
+    },
+    {
+      "epoch": 2.380811371039384,
+      "grad_norm": 1.9066398001458083,
+      "learning_rate": 2.1476226695273326e-06,
+      "loss": 0.6092,
+      "step": 4020
+    },
+    {
+      "epoch": 2.381403612673971,
+      "grad_norm": 2.76829177620669,
+      "learning_rate": 2.143664062938158e-06,
+      "loss": 0.6401,
+      "step": 4021
+    },
+    {
+      "epoch": 2.381995854308558,
+      "grad_norm": 3.0015528560284026,
+      "learning_rate": 2.1397086700321635e-06,
+      "loss": 0.6389,
+      "step": 4022
+    },
+    {
+      "epoch": 2.3825880959431447,
+      "grad_norm": 1.4409882465741872,
+      "learning_rate": 2.1357564924273265e-06,
+      "loss": 0.6066,
+      "step": 4023
+    },
+    {
+      "epoch": 2.383180337577732,
+      "grad_norm": 1.6002150244136195,
+      "learning_rate": 2.1318075317403152e-06,
+      "loss": 0.6025,
+      "step": 4024
+    },
+    {
+      "epoch": 2.3837725792123186,
+      "grad_norm": 1.9985808815147001,
+      "learning_rate": 2.1278617895864706e-06,
+      "loss": 0.6046,
+      "step": 4025
+    },
+    {
+      "epoch": 2.3843648208469057,
+      "grad_norm": 1.694290847273592,
+      "learning_rate": 2.123919267579828e-06,
+      "loss": 0.6045,
+      "step": 4026
+    },
+    {
+      "epoch": 2.3849570624814924,
+      "grad_norm": 2.032855773037069,
+      "learning_rate": 2.1199799673330956e-06,
+      "loss": 0.6228,
+      "step": 4027
+    },
+    {
+      "epoch": 2.385549304116079,
+      "grad_norm": 1.4539445622297456,
+      "learning_rate": 2.1160438904576743e-06,
+      "loss": 0.6275,
+      "step": 4028
+    },
+    {
+      "epoch": 2.3861415457506663,
+      "grad_norm": 2.5084329606273528,
+      "learning_rate": 2.1121110385636357e-06,
+      "loss": 0.6437,
+      "step": 4029
+    },
+    {
+      "epoch": 2.386733787385253,
+      "grad_norm": 1.5794313508804994,
+      "learning_rate": 2.108181413259741e-06,
+      "loss": 0.5742,
+      "step": 4030
+    },
+    {
+      "epoch": 2.38732602901984,
+      "grad_norm": 1.3385924339428654,
+      "learning_rate": 2.104255016153426e-06,
+      "loss": 0.6265,
+      "step": 4031
+    },
+    {
+      "epoch": 2.387918270654427,
+      "grad_norm": 2.2669122640842505,
+      "learning_rate": 2.1003318488508107e-06,
+      "loss": 0.6635,
+      "step": 4032
+    },
+    {
+      "epoch": 2.388510512289014,
+      "grad_norm": 1.7123022659492335,
+      "learning_rate": 2.0964119129566864e-06,
+      "loss": 0.634,
+      "step": 4033
+    },
+    {
+      "epoch": 2.3891027539236007,
+      "grad_norm": 2.074583285155401,
+      "learning_rate": 2.092495210074532e-06,
+      "loss": 0.6422,
+      "step": 4034
+    },
+    {
+      "epoch": 2.389694995558188,
+      "grad_norm": 1.9435860623142942,
+      "learning_rate": 2.0885817418064947e-06,
+      "loss": 0.6352,
+      "step": 4035
+    },
+    {
+      "epoch": 2.3902872371927746,
+      "grad_norm": 1.7101502157361959,
+      "learning_rate": 2.0846715097534087e-06,
+      "loss": 0.6161,
+      "step": 4036
+    },
+    {
+      "epoch": 2.3908794788273617,
+      "grad_norm": 1.7335618347196917,
+      "learning_rate": 2.0807645155147726e-06,
+      "loss": 0.6222,
+      "step": 4037
+    },
+    {
+      "epoch": 2.3914717204619484,
+      "grad_norm": 1.5176551517398615,
+      "learning_rate": 2.0768607606887724e-06,
+      "loss": 0.6145,
+      "step": 4038
+    },
+    {
+      "epoch": 2.392063962096535,
+      "grad_norm": 1.091782921163788,
+      "learning_rate": 2.072960246872261e-06,
+      "loss": 0.5978,
+      "step": 4039
+    },
+    {
+      "epoch": 2.3926562037311223,
+      "grad_norm": 2.07357282739027,
+      "learning_rate": 2.069062975660765e-06,
+      "loss": 0.6421,
+      "step": 4040
+    },
+    {
+      "epoch": 2.393248445365709,
+      "grad_norm": 1.5424667085190742,
+      "learning_rate": 2.0651689486484894e-06,
+      "loss": 0.6532,
+      "step": 4041
+    },
+    {
+      "epoch": 2.393840687000296,
+      "grad_norm": 1.6880108409323265,
+      "learning_rate": 2.0612781674283142e-06,
+      "loss": 0.5966,
+      "step": 4042
+    },
+    {
+      "epoch": 2.394432928634883,
+      "grad_norm": 1.3807759801970425,
+      "learning_rate": 2.057390633591785e-06,
+      "loss": 0.5957,
+      "step": 4043
+    },
+    {
+      "epoch": 2.39502517026947,
+      "grad_norm": 1.3765906802348291,
+      "learning_rate": 2.0535063487291176e-06,
+      "loss": 0.636,
+      "step": 4044
+    },
+    {
+      "epoch": 2.3956174119040567,
+      "grad_norm": 3.0061313109171595,
+      "learning_rate": 2.049625314429207e-06,
+      "loss": 0.6586,
+      "step": 4045
+    },
+    {
+      "epoch": 2.396209653538644,
+      "grad_norm": 1.1890562111369984,
+      "learning_rate": 2.045747532279616e-06,
+      "loss": 0.5867,
+      "step": 4046
+    },
+    {
+      "epoch": 2.3968018951732306,
+      "grad_norm": 1.4910777330850993,
+      "learning_rate": 2.0418730038665747e-06,
+      "loss": 0.654,
+      "step": 4047
+    },
+    {
+      "epoch": 2.3973941368078178,
+      "grad_norm": 2.874685506178504,
+      "learning_rate": 2.038001730774978e-06,
+      "loss": 0.6087,
+      "step": 4048
+    },
+    {
+      "epoch": 2.3979863784424045,
+      "grad_norm": 1.3061726816651926,
+      "learning_rate": 2.034133714588399e-06,
+      "loss": 0.5734,
+      "step": 4049
+    },
+    {
+      "epoch": 2.398578620076991,
+      "grad_norm": 1.7785239459876332,
+      "learning_rate": 2.0302689568890753e-06,
+      "loss": 0.656,
+      "step": 4050
+    },
+    {
+      "epoch": 2.3991708617115783,
+      "grad_norm": 1.6332928430746765,
+      "learning_rate": 2.0264074592579087e-06,
+      "loss": 0.6034,
+      "step": 4051
+    },
+    {
+      "epoch": 2.399763103346165,
+      "grad_norm": 1.5176551764079678,
+      "learning_rate": 2.022549223274465e-06,
+      "loss": 0.608,
+      "step": 4052
+    },
+    {
+      "epoch": 2.400355344980752,
+      "grad_norm": 1.2607503040846897,
+      "learning_rate": 2.0186942505169827e-06,
+      "loss": 0.6439,
+      "step": 4053
+    },
+    {
+      "epoch": 2.400947586615339,
+      "grad_norm": 2.158511661493091,
+      "learning_rate": 2.0148425425623673e-06,
+      "loss": 0.6046,
+      "step": 4054
+    },
+    {
+      "epoch": 2.401539828249926,
+      "grad_norm": 2.465164565539923,
+      "learning_rate": 2.0109941009861743e-06,
+      "loss": 0.6046,
+      "step": 4055
+    },
+    {
+      "epoch": 2.4021320698845128,
+      "grad_norm": 1.3196892824682709,
+      "learning_rate": 2.0071489273626376e-06,
+      "loss": 0.6952,
+      "step": 4056
+    },
+    {
+      "epoch": 2.4027243115191,
+      "grad_norm": 1.6021832014855415,
+      "learning_rate": 2.0033070232646488e-06,
+      "loss": 0.6369,
+      "step": 4057
+    },
+    {
+      "epoch": 2.4033165531536866,
+      "grad_norm": 6.273666490794738,
+      "learning_rate": 1.999468390263769e-06,
+      "loss": 0.5949,
+      "step": 4058
+    },
+    {
+      "epoch": 2.403908794788274,
+      "grad_norm": 1.3248267529829056,
+      "learning_rate": 1.995633029930204e-06,
+      "loss": 0.6476,
+      "step": 4059
+    },
+    {
+      "epoch": 2.4045010364228605,
+      "grad_norm": 4.130727246499741,
+      "learning_rate": 1.9918009438328365e-06,
+      "loss": 0.6206,
+      "step": 4060
+    },
+    {
+      "epoch": 2.405093278057447,
+      "grad_norm": 1.8044158856252108,
+      "learning_rate": 1.9879721335392088e-06,
+      "loss": 0.6502,
+      "step": 4061
+    },
+    {
+      "epoch": 2.4056855196920344,
+      "grad_norm": 1.2394089890201714,
+      "learning_rate": 1.9841466006155162e-06,
+      "loss": 0.62,
+      "step": 4062
+    },
+    {
+      "epoch": 2.406277761326621,
+      "grad_norm": 1.3020178733946122,
+      "learning_rate": 1.9803243466266154e-06,
+      "loss": 0.604,
+      "step": 4063
+    },
+    {
+      "epoch": 2.4068700029612082,
+      "grad_norm": 2.3665692021469855,
+      "learning_rate": 1.976505373136025e-06,
+      "loss": 0.5555,
+      "step": 4064
+    },
+    {
+      "epoch": 2.407462244595795,
+      "grad_norm": 1.8076468168639461,
+      "learning_rate": 1.9726896817059214e-06,
+      "loss": 0.5771,
+      "step": 4065
+    },
+    {
+      "epoch": 2.408054486230382,
+      "grad_norm": 2.2441477724689425,
+      "learning_rate": 1.968877273897136e-06,
+      "loss": 0.6202,
+      "step": 4066
+    },
+    {
+      "epoch": 2.408646727864969,
+      "grad_norm": 1.2331684527553564,
+      "learning_rate": 1.965068151269156e-06,
+      "loss": 0.6105,
+      "step": 4067
+    },
+    {
+      "epoch": 2.409238969499556,
+      "grad_norm": 3.5419262263798377,
+      "learning_rate": 1.9612623153801267e-06,
+      "loss": 0.6276,
+      "step": 4068
+    },
+    {
+      "epoch": 2.4098312111341427,
+      "grad_norm": 1.2135255066194424,
+      "learning_rate": 1.9574597677868535e-06,
+      "loss": 0.5914,
+      "step": 4069
+    },
+    {
+      "epoch": 2.41042345276873,
+      "grad_norm": 1.1257223069045508,
+      "learning_rate": 1.953660510044789e-06,
+      "loss": 0.6071,
+      "step": 4070
+    },
+    {
+      "epoch": 2.4110156944033165,
+      "grad_norm": 1.7369991319720672,
+      "learning_rate": 1.949864543708042e-06,
+      "loss": 0.6332,
+      "step": 4071
+    },
+    {
+      "epoch": 2.4116079360379032,
+      "grad_norm": 1.42930483040224,
+      "learning_rate": 1.946071870329377e-06,
+      "loss": 0.6303,
+      "step": 4072
+    },
+    {
+      "epoch": 2.4122001776724904,
+      "grad_norm": 1.5717751665701418,
+      "learning_rate": 1.9422824914602135e-06,
+      "loss": 0.6262,
+      "step": 4073
+    },
+    {
+      "epoch": 2.412792419307077,
+      "grad_norm": 1.6832189402748274,
+      "learning_rate": 1.9384964086506185e-06,
+      "loss": 0.5809,
+      "step": 4074
+    },
+    {
+      "epoch": 2.4133846609416643,
+      "grad_norm": 1.7647206463793592,
+      "learning_rate": 1.9347136234493093e-06,
+      "loss": 0.6337,
+      "step": 4075
+    },
+    {
+      "epoch": 2.413976902576251,
+      "grad_norm": 1.9697614615701116,
+      "learning_rate": 1.930934137403665e-06,
+      "loss": 0.5884,
+      "step": 4076
+    },
+    {
+      "epoch": 2.414569144210838,
+      "grad_norm": 2.1805592052438523,
+      "learning_rate": 1.9271579520597005e-06,
+      "loss": 0.5893,
+      "step": 4077
+    },
+    {
+      "epoch": 2.415161385845425,
+      "grad_norm": 2.016248237439067,
+      "learning_rate": 1.923385068962095e-06,
+      "loss": 0.6267,
+      "step": 4078
+    },
+    {
+      "epoch": 2.415753627480012,
+      "grad_norm": 2.363174925154651,
+      "learning_rate": 1.919615489654163e-06,
+      "loss": 0.6469,
+      "step": 4079
+    },
+    {
+      "epoch": 2.4163458691145987,
+      "grad_norm": 1.3925989121360791,
+      "learning_rate": 1.9158492156778807e-06,
+      "loss": 0.6612,
+      "step": 4080
+    },
+    {
+      "epoch": 2.416938110749186,
+      "grad_norm": 1.273060654648486,
+      "learning_rate": 1.91208624857386e-06,
+      "loss": 0.6226,
+      "step": 4081
+    },
+    {
+      "epoch": 2.4175303523837726,
+      "grad_norm": 1.4601623756383526,
+      "learning_rate": 1.908326589881372e-06,
+      "loss": 0.6188,
+      "step": 4082
+    },
+    {
+      "epoch": 2.4181225940183593,
+      "grad_norm": 1.5467664483267456,
+      "learning_rate": 1.9045702411383227e-06,
+      "loss": 0.5895,
+      "step": 4083
+    },
+    {
+      "epoch": 2.4187148356529464,
+      "grad_norm": 2.82373984208031,
+      "learning_rate": 1.9008172038812744e-06,
+      "loss": 0.6258,
+      "step": 4084
+    },
+    {
+      "epoch": 2.419307077287533,
+      "grad_norm": 1.9417680433070386,
+      "learning_rate": 1.897067479645428e-06,
+      "loss": 0.6163,
+      "step": 4085
+    },
+    {
+      "epoch": 2.4198993189221203,
+      "grad_norm": 2.2916627310601374,
+      "learning_rate": 1.8933210699646342e-06,
+      "loss": 0.6046,
+      "step": 4086
+    },
+    {
+      "epoch": 2.420491560556707,
+      "grad_norm": 1.5024563307793963,
+      "learning_rate": 1.8895779763713806e-06,
+      "loss": 0.5865,
+      "step": 4087
+    },
+    {
+      "epoch": 2.421083802191294,
+      "grad_norm": 1.446643763862051,
+      "learning_rate": 1.885838200396808e-06,
+      "loss": 0.6188,
+      "step": 4088
+    },
+    {
+      "epoch": 2.421676043825881,
+      "grad_norm": 1.0593789622862213,
+      "learning_rate": 1.8821017435706912e-06,
+      "loss": 0.6205,
+      "step": 4089
+    },
+    {
+      "epoch": 2.422268285460468,
+      "grad_norm": 1.5753893128495062,
+      "learning_rate": 1.8783686074214546e-06,
+      "loss": 0.6676,
+      "step": 4090
+    },
+    {
+      "epoch": 2.4228605270950547,
+      "grad_norm": 1.5478395153422457,
+      "learning_rate": 1.874638793476159e-06,
+      "loss": 0.6609,
+      "step": 4091
+    },
+    {
+      "epoch": 2.423452768729642,
+      "grad_norm": 2.513139396256183,
+      "learning_rate": 1.8709123032605058e-06,
+      "loss": 0.6139,
+      "step": 4092
+    },
+    {
+      "epoch": 2.4240450103642286,
+      "grad_norm": 4.903771647987582,
+      "learning_rate": 1.8671891382988416e-06,
+      "loss": 0.6128,
+      "step": 4093
+    },
+    {
+      "epoch": 2.4246372519988153,
+      "grad_norm": 1.0479089231627525,
+      "learning_rate": 1.8634693001141513e-06,
+      "loss": 0.6405,
+      "step": 4094
+    },
+    {
+      "epoch": 2.4252294936334025,
+      "grad_norm": 4.228476276270857,
+      "learning_rate": 1.8597527902280577e-06,
+      "loss": 0.6362,
+      "step": 4095
+    },
+    {
+      "epoch": 2.425821735267989,
+      "grad_norm": 1.5814852714772147,
+      "learning_rate": 1.856039610160818e-06,
+      "loss": 0.6704,
+      "step": 4096
+    },
+    {
+      "epoch": 2.4264139769025763,
+      "grad_norm": 2.1012557519164776,
+      "learning_rate": 1.8523297614313351e-06,
+      "loss": 0.5896,
+      "step": 4097
+    },
+    {
+      "epoch": 2.427006218537163,
+      "grad_norm": 1.463430055540698,
+      "learning_rate": 1.8486232455571473e-06,
+      "loss": 0.5984,
+      "step": 4098
+    },
+    {
+      "epoch": 2.42759846017175,
+      "grad_norm": 1.6658819397730513,
+      "learning_rate": 1.8449200640544274e-06,
+      "loss": 0.5641,
+      "step": 4099
+    },
+    {
+      "epoch": 2.428190701806337,
+      "grad_norm": 2.26128039729551,
+      "learning_rate": 1.8412202184379801e-06,
+      "loss": 0.6307,
+      "step": 4100
+    },
+    {
+      "epoch": 2.428782943440924,
+      "grad_norm": 1.4559178783717623,
+      "learning_rate": 1.837523710221254e-06,
+      "loss": 0.6036,
+      "step": 4101
+    },
+    {
+      "epoch": 2.4293751850755108,
+      "grad_norm": 1.7791634905991207,
+      "learning_rate": 1.8338305409163314e-06,
+      "loss": 0.6759,
+      "step": 4102
+    },
+    {
+      "epoch": 2.429967426710098,
+      "grad_norm": 1.697277492617549,
+      "learning_rate": 1.8301407120339232e-06,
+      "loss": 0.6221,
+      "step": 4103
+    },
+    {
+      "epoch": 2.4305596683446846,
+      "grad_norm": 2.2136330956036696,
+      "learning_rate": 1.826454225083375e-06,
+      "loss": 0.6591,
+      "step": 4104
+    },
+    {
+      "epoch": 2.4311519099792713,
+      "grad_norm": 1.3508855440744538,
+      "learning_rate": 1.8227710815726686e-06,
+      "loss": 0.6392,
+      "step": 4105
+    },
+    {
+      "epoch": 2.4317441516138585,
+      "grad_norm": 1.8566181281306422,
+      "learning_rate": 1.8190912830084207e-06,
+      "loss": 0.5795,
+      "step": 4106
+    },
+    {
+      "epoch": 2.432336393248445,
+      "grad_norm": 3.144464503693746,
+      "learning_rate": 1.815414830895873e-06,
+      "loss": 0.593,
+      "step": 4107
+    },
+    {
+      "epoch": 2.4329286348830323,
+      "grad_norm": 2.7242904341679846,
+      "learning_rate": 1.811741726738898e-06,
+      "loss": 0.6229,
+      "step": 4108
+    },
+    {
+      "epoch": 2.433520876517619,
+      "grad_norm": 3.3833912508202517,
+      "learning_rate": 1.8080719720400052e-06,
+      "loss": 0.6076,
+      "step": 4109
+    },
+    {
+      "epoch": 2.434113118152206,
+      "grad_norm": 3.0225222759099992,
+      "learning_rate": 1.8044055683003358e-06,
+      "loss": 0.6193,
+      "step": 4110
+    },
+    {
+      "epoch": 2.434705359786793,
+      "grad_norm": 1.4370596639040112,
+      "learning_rate": 1.8007425170196435e-06,
+      "loss": 0.6325,
+      "step": 4111
+    },
+    {
+      "epoch": 2.43529760142138,
+      "grad_norm": 1.5211255655218439,
+      "learning_rate": 1.7970828196963286e-06,
+      "loss": 0.556,
+      "step": 4112
+    },
+    {
+      "epoch": 2.435889843055967,
+      "grad_norm": 1.3392790413616675,
+      "learning_rate": 1.7934264778274157e-06,
+      "loss": 0.5869,
+      "step": 4113
+    },
+    {
+      "epoch": 2.436482084690554,
+      "grad_norm": 2.2652578152092957,
+      "learning_rate": 1.7897734929085508e-06,
+      "loss": 0.6022,
+      "step": 4114
+    },
+    {
+      "epoch": 2.4370743263251406,
+      "grad_norm": 1.5991779329388756,
+      "learning_rate": 1.7861238664340075e-06,
+      "loss": 0.6318,
+      "step": 4115
+    },
+    {
+      "epoch": 2.4376665679597274,
+      "grad_norm": 1.6856573090275266,
+      "learning_rate": 1.7824775998966926e-06,
+      "loss": 0.6195,
+      "step": 4116
+    },
+    {
+      "epoch": 2.4382588095943145,
+      "grad_norm": 1.872131972586157,
+      "learning_rate": 1.7788346947881352e-06,
+      "loss": 0.6015,
+      "step": 4117
+    },
+    {
+      "epoch": 2.4388510512289012,
+      "grad_norm": 1.3474243564708543,
+      "learning_rate": 1.7751951525984857e-06,
+      "loss": 0.5991,
+      "step": 4118
+    },
+    {
+      "epoch": 2.4394432928634884,
+      "grad_norm": 1.44125755213156,
+      "learning_rate": 1.7715589748165196e-06,
+      "loss": 0.6336,
+      "step": 4119
+    },
+    {
+      "epoch": 2.440035534498075,
+      "grad_norm": 1.628282300330072,
+      "learning_rate": 1.7679261629296408e-06,
+      "loss": 0.628,
+      "step": 4120
+    },
+    {
+      "epoch": 2.4406277761326622,
+      "grad_norm": 1.5525192264857999,
+      "learning_rate": 1.7642967184238758e-06,
+      "loss": 0.5914,
+      "step": 4121
+    },
+    {
+      "epoch": 2.441220017767249,
+      "grad_norm": 1.3773547139505093,
+      "learning_rate": 1.7606706427838682e-06,
+      "loss": 0.6638,
+      "step": 4122
+    },
+    {
+      "epoch": 2.441812259401836,
+      "grad_norm": 2.496487937041218,
+      "learning_rate": 1.7570479374928862e-06,
+      "loss": 0.6213,
+      "step": 4123
+    },
+    {
+      "epoch": 2.442404501036423,
+      "grad_norm": 3.2817990638971257,
+      "learning_rate": 1.7534286040328208e-06,
+      "loss": 0.6255,
+      "step": 4124
+    },
+    {
+      "epoch": 2.44299674267101,
+      "grad_norm": 3.6057888078556712,
+      "learning_rate": 1.7498126438841857e-06,
+      "loss": 0.6445,
+      "step": 4125
+    },
+    {
+      "epoch": 2.4435889843055967,
+      "grad_norm": 1.5818140283182256,
+      "learning_rate": 1.7462000585261096e-06,
+      "loss": 0.6281,
+      "step": 4126
+    },
+    {
+      "epoch": 2.4441812259401834,
+      "grad_norm": 1.9261028578432795,
+      "learning_rate": 1.7425908494363408e-06,
+      "loss": 0.6367,
+      "step": 4127
+    },
+    {
+      "epoch": 2.4447734675747705,
+      "grad_norm": 2.984646605350875,
+      "learning_rate": 1.7389850180912537e-06,
+      "loss": 0.6526,
+      "step": 4128
+    },
+    {
+      "epoch": 2.4453657092093573,
+      "grad_norm": 1.361931780545937,
+      "learning_rate": 1.735382565965832e-06,
+      "loss": 0.5942,
+      "step": 4129
+    },
+    {
+      "epoch": 2.4459579508439444,
+      "grad_norm": 1.230243762561541,
+      "learning_rate": 1.7317834945336843e-06,
+      "loss": 0.6713,
+      "step": 4130
+    },
+    {
+      "epoch": 2.446550192478531,
+      "grad_norm": 2.9100521565655173,
+      "learning_rate": 1.7281878052670288e-06,
+      "loss": 0.641,
+      "step": 4131
+    },
+    {
+      "epoch": 2.4471424341131183,
+      "grad_norm": 1.3667964273082245,
+      "learning_rate": 1.724595499636711e-06,
+      "loss": 0.5686,
+      "step": 4132
+    },
+    {
+      "epoch": 2.447734675747705,
+      "grad_norm": 1.2327143660385451,
+      "learning_rate": 1.7210065791121789e-06,
+      "loss": 0.6167,
+      "step": 4133
+    },
+    {
+      "epoch": 2.448326917382292,
+      "grad_norm": 2.0388204744932894,
+      "learning_rate": 1.7174210451615091e-06,
+      "loss": 0.6042,
+      "step": 4134
+    },
+    {
+      "epoch": 2.448919159016879,
+      "grad_norm": 1.5150026413148958,
+      "learning_rate": 1.713838899251381e-06,
+      "loss": 0.627,
+      "step": 4135
+    },
+    {
+      "epoch": 2.449511400651466,
+      "grad_norm": 1.1466180408826923,
+      "learning_rate": 1.7102601428470988e-06,
+      "loss": 0.6109,
+      "step": 4136
+    },
+    {
+      "epoch": 2.4501036422860527,
+      "grad_norm": 2.0780281123685422,
+      "learning_rate": 1.7066847774125716e-06,
+      "loss": 0.6083,
+      "step": 4137
+    },
+    {
+      "epoch": 2.4506958839206394,
+      "grad_norm": 1.3954878819749057,
+      "learning_rate": 1.7031128044103272e-06,
+      "loss": 0.5951,
+      "step": 4138
+    },
+    {
+      "epoch": 2.4512881255552266,
+      "grad_norm": 1.2417256271708208,
+      "learning_rate": 1.6995442253015003e-06,
+      "loss": 0.5816,
+      "step": 4139
+    },
+    {
+      "epoch": 2.4518803671898133,
+      "grad_norm": 2.1291302266514998,
+      "learning_rate": 1.6959790415458454e-06,
+      "loss": 0.6318,
+      "step": 4140
+    },
+    {
+      "epoch": 2.4524726088244004,
+      "grad_norm": 1.294464705411972,
+      "learning_rate": 1.692417254601717e-06,
+      "loss": 0.6079,
+      "step": 4141
+    },
+    {
+      "epoch": 2.453064850458987,
+      "grad_norm": 1.723843313270676,
+      "learning_rate": 1.6888588659260929e-06,
+      "loss": 0.6003,
+      "step": 4142
+    },
+    {
+      "epoch": 2.4536570920935743,
+      "grad_norm": 2.01346115839394,
+      "learning_rate": 1.6853038769745466e-06,
+      "loss": 0.6387,
+      "step": 4143
+    },
+    {
+      "epoch": 2.454249333728161,
+      "grad_norm": 1.7697041782949663,
+      "learning_rate": 1.6817522892012762e-06,
+      "loss": 0.6412,
+      "step": 4144
+    },
+    {
+      "epoch": 2.454841575362748,
+      "grad_norm": 1.1993784853907747,
+      "learning_rate": 1.6782041040590769e-06,
+      "loss": 0.595,
+      "step": 4145
+    },
+    {
+      "epoch": 2.455433816997335,
+      "grad_norm": 1.5669145113783172,
+      "learning_rate": 1.6746593229993545e-06,
+      "loss": 0.5877,
+      "step": 4146
+    },
+    {
+      "epoch": 2.456026058631922,
+      "grad_norm": 1.1370169190841048,
+      "learning_rate": 1.6711179474721272e-06,
+      "loss": 0.6322,
+      "step": 4147
+    },
+    {
+      "epoch": 2.4566183002665087,
+      "grad_norm": 3.5535580287557127,
+      "learning_rate": 1.6675799789260128e-06,
+      "loss": 0.631,
+      "step": 4148
+    },
+    {
+      "epoch": 2.4572105419010954,
+      "grad_norm": 1.85149070130156,
+      "learning_rate": 1.6640454188082444e-06,
+      "loss": 0.6152,
+      "step": 4149
+    },
+    {
+      "epoch": 2.4578027835356826,
+      "grad_norm": 2.303568347956373,
+      "learning_rate": 1.6605142685646503e-06,
+      "loss": 0.6196,
+      "step": 4150
+    },
+    {
+      "epoch": 2.4583950251702693,
+      "grad_norm": 2.3689995504227372,
+      "learning_rate": 1.6569865296396748e-06,
+      "loss": 0.5813,
+      "step": 4151
+    },
+    {
+      "epoch": 2.4589872668048565,
+      "grad_norm": 1.894218094477947,
+      "learning_rate": 1.6534622034763558e-06,
+      "loss": 0.6069,
+      "step": 4152
+    },
+    {
+      "epoch": 2.459579508439443,
+      "grad_norm": 4.169690853566014,
+      "learning_rate": 1.6499412915163481e-06,
+      "loss": 0.6636,
+      "step": 4153
+    },
+    {
+      "epoch": 2.4601717500740303,
+      "grad_norm": 1.6611317565107804,
+      "learning_rate": 1.6464237951998952e-06,
+      "loss": 0.561,
+      "step": 4154
+    },
+    {
+      "epoch": 2.460763991708617,
+      "grad_norm": 2.215191682910294,
+      "learning_rate": 1.642909715965857e-06,
+      "loss": 0.63,
+      "step": 4155
+    },
+    {
+      "epoch": 2.461356233343204,
+      "grad_norm": 1.6918194298709026,
+      "learning_rate": 1.6393990552516848e-06,
+      "loss": 0.6134,
+      "step": 4156
+    },
+    {
+      "epoch": 2.461948474977791,
+      "grad_norm": 1.5262087510396043,
+      "learning_rate": 1.63589181449344e-06,
+      "loss": 0.6413,
+      "step": 4157
+    },
+    {
+      "epoch": 2.462540716612378,
+      "grad_norm": 1.2968558382847615,
+      "learning_rate": 1.6323879951257783e-06,
+      "loss": 0.6197,
+      "step": 4158
+    },
+    {
+      "epoch": 2.4631329582469648,
+      "grad_norm": 1.6679364569860284,
+      "learning_rate": 1.628887598581962e-06,
+      "loss": 0.5777,
+      "step": 4159
+    },
+    {
+      "epoch": 2.4637251998815515,
+      "grad_norm": 1.0369459266334156,
+      "learning_rate": 1.6253906262938457e-06,
+      "loss": 0.6209,
+      "step": 4160
+    },
+    {
+      "epoch": 2.4643174415161386,
+      "grad_norm": 1.5112263549928397,
+      "learning_rate": 1.6218970796918931e-06,
+      "loss": 0.6424,
+      "step": 4161
+    },
+    {
+      "epoch": 2.4649096831507253,
+      "grad_norm": 1.2179638572938345,
+      "learning_rate": 1.6184069602051578e-06,
+      "loss": 0.6032,
+      "step": 4162
+    },
+    {
+      "epoch": 2.4655019247853125,
+      "grad_norm": 1.2210989669346337,
+      "learning_rate": 1.614920269261293e-06,
+      "loss": 0.6041,
+      "step": 4163
+    },
+    {
+      "epoch": 2.466094166419899,
+      "grad_norm": 1.5638850592616937,
+      "learning_rate": 1.611437008286555e-06,
+      "loss": 0.6265,
+      "step": 4164
+    },
+    {
+      "epoch": 2.4666864080544864,
+      "grad_norm": 1.9348031966836825,
+      "learning_rate": 1.6079571787057946e-06,
+      "loss": 0.588,
+      "step": 4165
+    },
+    {
+      "epoch": 2.467278649689073,
+      "grad_norm": 1.4655585715236081,
+      "learning_rate": 1.6044807819424545e-06,
+      "loss": 0.6662,
+      "step": 4166
+    },
+    {
+      "epoch": 2.46787089132366,
+      "grad_norm": 1.35471687098539,
+      "learning_rate": 1.6010078194185752e-06,
+      "loss": 0.6332,
+      "step": 4167
+    },
+    {
+      "epoch": 2.468463132958247,
+      "grad_norm": 1.4021986342558357,
+      "learning_rate": 1.5975382925547966e-06,
+      "loss": 0.6344,
+      "step": 4168
+    },
+    {
+      "epoch": 2.469055374592834,
+      "grad_norm": 1.5104775229268375,
+      "learning_rate": 1.5940722027703516e-06,
+      "loss": 0.595,
+      "step": 4169
+    },
+    {
+      "epoch": 2.469647616227421,
+      "grad_norm": 1.5242057096864499,
+      "learning_rate": 1.5906095514830645e-06,
+      "loss": 0.6058,
+      "step": 4170
+    },
+    {
+      "epoch": 2.4702398578620075,
+      "grad_norm": 1.446713003214256,
+      "learning_rate": 1.5871503401093501e-06,
+      "loss": 0.6319,
+      "step": 4171
+    },
+    {
+      "epoch": 2.4708320994965947,
+      "grad_norm": 2.343073886207364,
+      "learning_rate": 1.5836945700642248e-06,
+      "loss": 0.6254,
+      "step": 4172
+    },
+    {
+      "epoch": 2.4714243411311814,
+      "grad_norm": 2.399540643140653,
+      "learning_rate": 1.580242242761295e-06,
+      "loss": 0.5846,
+      "step": 4173
+    },
+    {
+      "epoch": 2.4720165827657685,
+      "grad_norm": 1.928520767924082,
+      "learning_rate": 1.5767933596127528e-06,
+      "loss": 0.6639,
+      "step": 4174
+    },
+    {
+      "epoch": 2.4726088244003552,
+      "grad_norm": 1.1218273751756729,
+      "learning_rate": 1.5733479220293847e-06,
+      "loss": 0.5455,
+      "step": 4175
+    },
+    {
+      "epoch": 2.4732010660349424,
+      "grad_norm": 1.379515727004053,
+      "learning_rate": 1.56990593142057e-06,
+      "loss": 0.6029,
+      "step": 4176
+    },
+    {
+      "epoch": 2.473793307669529,
+      "grad_norm": 2.093979049163968,
+      "learning_rate": 1.5664673891942805e-06,
+      "loss": 0.5524,
+      "step": 4177
+    },
+    {
+      "epoch": 2.4743855493041162,
+      "grad_norm": 1.4867542033675507,
+      "learning_rate": 1.5630322967570655e-06,
+      "loss": 0.6112,
+      "step": 4178
+    },
+    {
+      "epoch": 2.474977790938703,
+      "grad_norm": 1.5435512842922596,
+      "learning_rate": 1.559600655514074e-06,
+      "loss": 0.609,
+      "step": 4179
+    },
+    {
+      "epoch": 2.47557003257329,
+      "grad_norm": 1.5223906435727905,
+      "learning_rate": 1.5561724668690436e-06,
+      "loss": 0.5927,
+      "step": 4180
+    },
+    {
+      "epoch": 2.476162274207877,
+      "grad_norm": 1.7939175275171766,
+      "learning_rate": 1.5527477322242934e-06,
+      "loss": 0.5754,
+      "step": 4181
+    },
+    {
+      "epoch": 2.4767545158424635,
+      "grad_norm": 2.322851003634435,
+      "learning_rate": 1.5493264529807305e-06,
+      "loss": 0.6463,
+      "step": 4182
+    },
+    {
+      "epoch": 2.4773467574770507,
+      "grad_norm": 1.4341109913608379,
+      "learning_rate": 1.5459086305378524e-06,
+      "loss": 0.6512,
+      "step": 4183
+    },
+    {
+      "epoch": 2.4779389991116374,
+      "grad_norm": 2.000472244942328,
+      "learning_rate": 1.5424942662937436e-06,
+      "loss": 0.6175,
+      "step": 4184
+    },
+    {
+      "epoch": 2.4785312407462246,
+      "grad_norm": 1.1364991284051775,
+      "learning_rate": 1.5390833616450684e-06,
+      "loss": 0.5778,
+      "step": 4185
+    },
+    {
+      "epoch": 2.4791234823808113,
+      "grad_norm": 2.5864290370042187,
+      "learning_rate": 1.5356759179870762e-06,
+      "loss": 0.6695,
+      "step": 4186
+    },
+    {
+      "epoch": 2.4797157240153984,
+      "grad_norm": 1.4569921907313288,
+      "learning_rate": 1.5322719367136064e-06,
+      "loss": 0.5809,
+      "step": 4187
+    },
+    {
+      "epoch": 2.480307965649985,
+      "grad_norm": 1.3478923100787537,
+      "learning_rate": 1.5288714192170796e-06,
+      "loss": 0.5834,
+      "step": 4188
+    },
+    {
+      "epoch": 2.4809002072845723,
+      "grad_norm": 3.810585927154581,
+      "learning_rate": 1.5254743668884963e-06,
+      "loss": 0.6608,
+      "step": 4189
+    },
+    {
+      "epoch": 2.481492448919159,
+      "grad_norm": 1.3909152535067673,
+      "learning_rate": 1.522080781117441e-06,
+      "loss": 0.6216,
+      "step": 4190
+    },
+    {
+      "epoch": 2.482084690553746,
+      "grad_norm": 1.7514258321845537,
+      "learning_rate": 1.5186906632920816e-06,
+      "loss": 0.6203,
+      "step": 4191
+    },
+    {
+      "epoch": 2.482676932188333,
+      "grad_norm": 1.731744905638932,
+      "learning_rate": 1.5153040147991716e-06,
+      "loss": 0.613,
+      "step": 4192
+    },
+    {
+      "epoch": 2.4832691738229196,
+      "grad_norm": 2.3774079021997765,
+      "learning_rate": 1.5119208370240369e-06,
+      "loss": 0.6039,
+      "step": 4193
+    },
+    {
+      "epoch": 2.4838614154575067,
+      "grad_norm": 1.9933956673478872,
+      "learning_rate": 1.5085411313505849e-06,
+      "loss": 0.6405,
+      "step": 4194
+    },
+    {
+      "epoch": 2.4844536570920934,
+      "grad_norm": 2.8685180286755187,
+      "learning_rate": 1.5051648991613077e-06,
+      "loss": 0.621,
+      "step": 4195
+    },
+    {
+      "epoch": 2.4850458987266806,
+      "grad_norm": 1.2386381396531745,
+      "learning_rate": 1.5017921418372772e-06,
+      "loss": 0.6757,
+      "step": 4196
+    },
+    {
+      "epoch": 2.4856381403612673,
+      "grad_norm": 1.2887006546671949,
+      "learning_rate": 1.4984228607581386e-06,
+      "loss": 0.599,
+      "step": 4197
+    },
+    {
+      "epoch": 2.4862303819958544,
+      "grad_norm": 1.3284199306461797,
+      "learning_rate": 1.4950570573021138e-06,
+      "loss": 0.5998,
+      "step": 4198
+    },
+    {
+      "epoch": 2.486822623630441,
+      "grad_norm": 1.505776114908257,
+      "learning_rate": 1.4916947328460108e-06,
+      "loss": 0.6197,
+      "step": 4199
+    },
+    {
+      "epoch": 2.4874148652650283,
+      "grad_norm": 1.4391666401175998,
+      "learning_rate": 1.4883358887652044e-06,
+      "loss": 0.5938,
+      "step": 4200
+    },
+    {
+      "epoch": 2.488007106899615,
+      "grad_norm": 1.7810197494911437,
+      "learning_rate": 1.484980526433657e-06,
+      "loss": 0.6247,
+      "step": 4201
+    },
+    {
+      "epoch": 2.488599348534202,
+      "grad_norm": 1.8202240648965637,
+      "learning_rate": 1.4816286472238939e-06,
+      "loss": 0.5803,
+      "step": 4202
+    },
+    {
+      "epoch": 2.489191590168789,
+      "grad_norm": 1.5697032510584463,
+      "learning_rate": 1.4782802525070282e-06,
+      "loss": 0.6114,
+      "step": 4203
+    },
+    {
+      "epoch": 2.4897838318033756,
+      "grad_norm": 1.3273698484858674,
+      "learning_rate": 1.474935343652736e-06,
+      "loss": 0.6103,
+      "step": 4204
+    },
+    {
+      "epoch": 2.4903760734379627,
+      "grad_norm": 1.825999416280271,
+      "learning_rate": 1.4715939220292775e-06,
+      "loss": 0.6109,
+      "step": 4205
+    },
+    {
+      "epoch": 2.4909683150725495,
+      "grad_norm": 6.697769541146492,
+      "learning_rate": 1.4682559890034787e-06,
+      "loss": 0.5785,
+      "step": 4206
+    },
+    {
+      "epoch": 2.4915605567071366,
+      "grad_norm": 4.951124117950075,
+      "learning_rate": 1.4649215459407462e-06,
+      "loss": 0.6181,
+      "step": 4207
+    },
+    {
+      "epoch": 2.4921527983417233,
+      "grad_norm": 2.2685218356339996,
+      "learning_rate": 1.461590594205049e-06,
+      "loss": 0.5906,
+      "step": 4208
+    },
+    {
+      "epoch": 2.4927450399763105,
+      "grad_norm": 1.187772811816334,
+      "learning_rate": 1.4582631351589405e-06,
+      "loss": 0.621,
+      "step": 4209
+    },
+    {
+      "epoch": 2.493337281610897,
+      "grad_norm": 1.568313648134276,
+      "learning_rate": 1.4549391701635308e-06,
+      "loss": 0.6407,
+      "step": 4210
+    },
+    {
+      "epoch": 2.4939295232454843,
+      "grad_norm": 2.056945291130729,
+      "learning_rate": 1.4516187005785153e-06,
+      "loss": 0.6526,
+      "step": 4211
+    },
+    {
+      "epoch": 2.494521764880071,
+      "grad_norm": 2.5897828120036372,
+      "learning_rate": 1.4483017277621482e-06,
+      "loss": 0.5857,
+      "step": 4212
+    },
+    {
+      "epoch": 2.495114006514658,
+      "grad_norm": 2.003851685407157,
+      "learning_rate": 1.4449882530712621e-06,
+      "loss": 0.5891,
+      "step": 4213
+    },
+    {
+      "epoch": 2.495706248149245,
+      "grad_norm": 2.031397305896863,
+      "learning_rate": 1.4416782778612514e-06,
+      "loss": 0.6374,
+      "step": 4214
+    },
+    {
+      "epoch": 2.4962984897838316,
+      "grad_norm": 2.460742739390165,
+      "learning_rate": 1.4383718034860806e-06,
+      "loss": 0.6282,
+      "step": 4215
+    },
+    {
+      "epoch": 2.4968907314184188,
+      "grad_norm": 1.8883652271349463,
+      "learning_rate": 1.4350688312982864e-06,
+      "loss": 0.607,
+      "step": 4216
+    },
+    {
+      "epoch": 2.4974829730530055,
+      "grad_norm": 1.6383575358198943,
+      "learning_rate": 1.4317693626489715e-06,
+      "loss": 0.6155,
+      "step": 4217
+    },
+    {
+      "epoch": 2.4980752146875926,
+      "grad_norm": 1.5789440462021105,
+      "learning_rate": 1.428473398887802e-06,
+      "loss": 0.6475,
+      "step": 4218
+    },
+    {
+      "epoch": 2.4986674563221793,
+      "grad_norm": 1.7872165908651498,
+      "learning_rate": 1.4251809413630103e-06,
+      "loss": 0.6318,
+      "step": 4219
+    },
+    {
+      "epoch": 2.4992596979567665,
+      "grad_norm": 1.2177122630184891,
+      "learning_rate": 1.421891991421399e-06,
+      "loss": 0.6186,
+      "step": 4220
+    },
+    {
+      "epoch": 2.499851939591353,
+      "grad_norm": 5.5110163664079765,
+      "learning_rate": 1.4186065504083356e-06,
+      "loss": 0.6125,
+      "step": 4221
+    },
+    {
+      "epoch": 2.50044418122594,
+      "grad_norm": 1.3204263434006236,
+      "learning_rate": 1.4153246196677483e-06,
+      "loss": 0.5692,
+      "step": 4222
+    },
+    {
+      "epoch": 2.501036422860527,
+      "grad_norm": 2.040538027329011,
+      "learning_rate": 1.4120462005421287e-06,
+      "loss": 0.6165,
+      "step": 4223
+    },
+    {
+      "epoch": 2.5016286644951142,
+      "grad_norm": 4.142369576433715,
+      "learning_rate": 1.4087712943725384e-06,
+      "loss": 0.6178,
+      "step": 4224
+    },
+    {
+      "epoch": 2.502220906129701,
+      "grad_norm": 1.1861793588300895,
+      "learning_rate": 1.405499902498597e-06,
+      "loss": 0.6187,
+      "step": 4225
+    },
+    {
+      "epoch": 2.5028131477642876,
+      "grad_norm": 1.2674089845867869,
+      "learning_rate": 1.40223202625849e-06,
+      "loss": 0.6052,
+      "step": 4226
+    },
+    {
+      "epoch": 2.503405389398875,
+      "grad_norm": 1.6386149471462386,
+      "learning_rate": 1.3989676669889562e-06,
+      "loss": 0.5914,
+      "step": 4227
+    },
+    {
+      "epoch": 2.5039976310334615,
+      "grad_norm": 1.5069585186135463,
+      "learning_rate": 1.395706826025306e-06,
+      "loss": 0.5885,
+      "step": 4228
+    },
+    {
+      "epoch": 2.5045898726680487,
+      "grad_norm": 1.5769533827442805,
+      "learning_rate": 1.392449504701412e-06,
+      "loss": 0.578,
+      "step": 4229
+    },
+    {
+      "epoch": 2.5051821143026354,
+      "grad_norm": 1.9577342067869337,
+      "learning_rate": 1.3891957043496917e-06,
+      "loss": 0.5767,
+      "step": 4230
+    },
+    {
+      "epoch": 2.5057743559372225,
+      "grad_norm": 1.820348565915433,
+      "learning_rate": 1.3859454263011373e-06,
+      "loss": 0.6608,
+      "step": 4231
+    },
+    {
+      "epoch": 2.5063665975718092,
+      "grad_norm": 1.4196495980962014,
+      "learning_rate": 1.3826986718852952e-06,
+      "loss": 0.6256,
+      "step": 4232
+    },
+    {
+      "epoch": 2.506958839206396,
+      "grad_norm": 2.614783295311572,
+      "learning_rate": 1.3794554424302752e-06,
+      "loss": 0.6216,
+      "step": 4233
+    },
+    {
+      "epoch": 2.507551080840983,
+      "grad_norm": 2.0862712740764726,
+      "learning_rate": 1.3762157392627317e-06,
+      "loss": 0.6103,
+      "step": 4234
+    },
+    {
+      "epoch": 2.5081433224755703,
+      "grad_norm": 1.4923395580122512,
+      "learning_rate": 1.372979563707889e-06,
+      "loss": 0.6247,
+      "step": 4235
+    },
+    {
+      "epoch": 2.508735564110157,
+      "grad_norm": 1.6903494282840945,
+      "learning_rate": 1.3697469170895282e-06,
+      "loss": 0.6138,
+      "step": 4236
+    },
+    {
+      "epoch": 2.5093278057447437,
+      "grad_norm": 1.7287805590209488,
+      "learning_rate": 1.3665178007299818e-06,
+      "loss": 0.5927,
+      "step": 4237
+    },
+    {
+      "epoch": 2.509920047379331,
+      "grad_norm": 2.0975391419269314,
+      "learning_rate": 1.363292215950135e-06,
+      "loss": 0.5833,
+      "step": 4238
+    },
+    {
+      "epoch": 2.5105122890139175,
+      "grad_norm": 1.7555964609395533,
+      "learning_rate": 1.3600701640694392e-06,
+      "loss": 0.6121,
+      "step": 4239
+    },
+    {
+      "epoch": 2.5111045306485047,
+      "grad_norm": 1.6961673222728129,
+      "learning_rate": 1.3568516464058946e-06,
+      "loss": 0.6195,
+      "step": 4240
+    },
+    {
+      "epoch": 2.5116967722830914,
+      "grad_norm": 2.079432102930987,
+      "learning_rate": 1.3536366642760534e-06,
+      "loss": 0.6057,
+      "step": 4241
+    },
+    {
+      "epoch": 2.5122890139176786,
+      "grad_norm": 1.1628789903318941,
+      "learning_rate": 1.350425218995024e-06,
+      "loss": 0.5979,
+      "step": 4242
+    },
+    {
+      "epoch": 2.5128812555522653,
+      "grad_norm": 1.454884994824475,
+      "learning_rate": 1.3472173118764686e-06,
+      "loss": 0.5959,
+      "step": 4243
+    },
+    {
+      "epoch": 2.513473497186852,
+      "grad_norm": 1.977230587423518,
+      "learning_rate": 1.3440129442326045e-06,
+      "loss": 0.604,
+      "step": 4244
+    },
+    {
+      "epoch": 2.514065738821439,
+      "grad_norm": 1.4424961332700483,
+      "learning_rate": 1.3408121173741962e-06,
+      "loss": 0.609,
+      "step": 4245
+    },
+    {
+      "epoch": 2.5146579804560263,
+      "grad_norm": 1.178276635818679,
+      "learning_rate": 1.3376148326105586e-06,
+      "loss": 0.6249,
+      "step": 4246
+    },
+    {
+      "epoch": 2.515250222090613,
+      "grad_norm": 1.5026051517578134,
+      "learning_rate": 1.3344210912495648e-06,
+      "loss": 0.642,
+      "step": 4247
+    },
+    {
+      "epoch": 2.5158424637251997,
+      "grad_norm": 1.8145268871556535,
+      "learning_rate": 1.3312308945976348e-06,
+      "loss": 0.6375,
+      "step": 4248
+    },
+    {
+      "epoch": 2.516434705359787,
+      "grad_norm": 1.7801182768357935,
+      "learning_rate": 1.3280442439597384e-06,
+      "loss": 0.6133,
+      "step": 4249
+    },
+    {
+      "epoch": 2.5170269469943736,
+      "grad_norm": 2.2296683859183224,
+      "learning_rate": 1.3248611406393918e-06,
+      "loss": 0.6702,
+      "step": 4250
+    },
+    {
+      "epoch": 2.5176191886289607,
+      "grad_norm": 1.8998316765515955,
+      "learning_rate": 1.3216815859386667e-06,
+      "loss": 0.5944,
+      "step": 4251
+    },
+    {
+      "epoch": 2.5182114302635474,
+      "grad_norm": 2.0543176892307833,
+      "learning_rate": 1.318505581158177e-06,
+      "loss": 0.6315,
+      "step": 4252
+    },
+    {
+      "epoch": 2.5188036718981346,
+      "grad_norm": 1.5059166609031631,
+      "learning_rate": 1.3153331275970904e-06,
+      "loss": 0.5729,
+      "step": 4253
+    },
+    {
+      "epoch": 2.5193959135327213,
+      "grad_norm": 1.6358415801788577,
+      "learning_rate": 1.3121642265531154e-06,
+      "loss": 0.6296,
+      "step": 4254
+    },
+    {
+      "epoch": 2.519988155167308,
+      "grad_norm": 1.36873712039331,
+      "learning_rate": 1.3089988793225139e-06,
+      "loss": 0.615,
+      "step": 4255
+    },
+    {
+      "epoch": 2.520580396801895,
+      "grad_norm": 1.2889128303477049,
+      "learning_rate": 1.305837087200087e-06,
+      "loss": 0.6231,
+      "step": 4256
+    },
+    {
+      "epoch": 2.5211726384364823,
+      "grad_norm": 4.424445990535661,
+      "learning_rate": 1.30267885147919e-06,
+      "loss": 0.6012,
+      "step": 4257
+    },
+    {
+      "epoch": 2.521764880071069,
+      "grad_norm": 1.5525798274801588,
+      "learning_rate": 1.299524173451715e-06,
+      "loss": 0.5881,
+      "step": 4258
+    },
+    {
+      "epoch": 2.5223571217056557,
+      "grad_norm": 2.316949095558541,
+      "learning_rate": 1.2963730544081065e-06,
+      "loss": 0.6059,
+      "step": 4259
+    },
+    {
+      "epoch": 2.522949363340243,
+      "grad_norm": 1.6494823394802665,
+      "learning_rate": 1.2932254956373457e-06,
+      "loss": 0.6213,
+      "step": 4260
+    },
+    {
+      "epoch": 2.5235416049748296,
+      "grad_norm": 1.5280751943361606,
+      "learning_rate": 1.290081498426965e-06,
+      "loss": 0.5645,
+      "step": 4261
+    },
+    {
+      "epoch": 2.5241338466094168,
+      "grad_norm": 1.9376208589060975,
+      "learning_rate": 1.286941064063031e-06,
+      "loss": 0.5926,
+      "step": 4262
+    },
+    {
+      "epoch": 2.5247260882440035,
+      "grad_norm": 1.429613590939648,
+      "learning_rate": 1.2838041938301638e-06,
+      "loss": 0.6164,
+      "step": 4263
+    },
+    {
+      "epoch": 2.5253183298785906,
+      "grad_norm": 2.5583596054226247,
+      "learning_rate": 1.2806708890115138e-06,
+      "loss": 0.5951,
+      "step": 4264
+    },
+    {
+      "epoch": 2.5259105715131773,
+      "grad_norm": 2.2738048703872513,
+      "learning_rate": 1.2775411508887837e-06,
+      "loss": 0.6259,
+      "step": 4265
+    },
+    {
+      "epoch": 2.526502813147764,
+      "grad_norm": 2.872265401198682,
+      "learning_rate": 1.2744149807422113e-06,
+      "loss": 0.5911,
+      "step": 4266
+    },
+    {
+      "epoch": 2.527095054782351,
+      "grad_norm": 1.6508438193817,
+      "learning_rate": 1.2712923798505727e-06,
+      "loss": 0.6179,
+      "step": 4267
+    },
+    {
+      "epoch": 2.5276872964169383,
+      "grad_norm": 2.2705730263320825,
+      "learning_rate": 1.2681733494911897e-06,
+      "loss": 0.6396,
+      "step": 4268
+    },
+    {
+      "epoch": 2.528279538051525,
+      "grad_norm": 1.7932950964539862,
+      "learning_rate": 1.2650578909399225e-06,
+      "loss": 0.5945,
+      "step": 4269
+    },
+    {
+      "epoch": 2.5288717796861118,
+      "grad_norm": 2.039463332472355,
+      "learning_rate": 1.2619460054711685e-06,
+      "loss": 0.598,
+      "step": 4270
+    },
+    {
+      "epoch": 2.529464021320699,
+      "grad_norm": 2.1940382419730446,
+      "learning_rate": 1.2588376943578594e-06,
+      "loss": 0.6173,
+      "step": 4271
+    },
+    {
+      "epoch": 2.5300562629552856,
+      "grad_norm": 1.334149394220382,
+      "learning_rate": 1.2557329588714739e-06,
+      "loss": 0.6212,
+      "step": 4272
+    },
+    {
+      "epoch": 2.530648504589873,
+      "grad_norm": 1.153175614903197,
+      "learning_rate": 1.2526318002820237e-06,
+      "loss": 0.5776,
+      "step": 4273
+    },
+    {
+      "epoch": 2.5312407462244595,
+      "grad_norm": 2.7701095890179848,
+      "learning_rate": 1.2495342198580562e-06,
+      "loss": 0.6507,
+      "step": 4274
+    },
+    {
+      "epoch": 2.5318329878590466,
+      "grad_norm": 1.9084746730843827,
+      "learning_rate": 1.2464402188666524e-06,
+      "loss": 0.5635,
+      "step": 4275
+    },
+    {
+      "epoch": 2.5324252294936334,
+      "grad_norm": 3.3261345903916824,
+      "learning_rate": 1.2433497985734356e-06,
+      "loss": 0.6515,
+      "step": 4276
+    },
+    {
+      "epoch": 2.53301747112822,
+      "grad_norm": 1.3772062277063388,
+      "learning_rate": 1.2402629602425643e-06,
+      "loss": 0.6683,
+      "step": 4277
+    },
+    {
+      "epoch": 2.533609712762807,
+      "grad_norm": 1.5163170294865773,
+      "learning_rate": 1.237179705136725e-06,
+      "loss": 0.6679,
+      "step": 4278
+    },
+    {
+      "epoch": 2.5342019543973944,
+      "grad_norm": 1.5235149346531982,
+      "learning_rate": 1.2341000345171417e-06,
+      "loss": 0.5914,
+      "step": 4279
+    },
+    {
+      "epoch": 2.534794196031981,
+      "grad_norm": 1.4537614359095834,
+      "learning_rate": 1.2310239496435749e-06,
+      "loss": 0.5615,
+      "step": 4280
+    },
+    {
+      "epoch": 2.535386437666568,
+      "grad_norm": 3.3433167384515285,
+      "learning_rate": 1.2279514517743208e-06,
+      "loss": 0.5961,
+      "step": 4281
+    },
+    {
+      "epoch": 2.535978679301155,
+      "grad_norm": 1.759299393387278,
+      "learning_rate": 1.2248825421661937e-06,
+      "loss": 0.5994,
+      "step": 4282
+    },
+    {
+      "epoch": 2.5365709209357417,
+      "grad_norm": 5.824067607433355,
+      "learning_rate": 1.2218172220745562e-06,
+      "loss": 0.6236,
+      "step": 4283
+    },
+    {
+      "epoch": 2.537163162570329,
+      "grad_norm": 4.849146537686694,
+      "learning_rate": 1.2187554927532963e-06,
+      "loss": 0.6548,
+      "step": 4284
+    },
+    {
+      "epoch": 2.5377554042049155,
+      "grad_norm": 1.8728352219215554,
+      "learning_rate": 1.2156973554548369e-06,
+      "loss": 0.6009,
+      "step": 4285
+    },
+    {
+      "epoch": 2.5383476458395027,
+      "grad_norm": 2.628941079406281,
+      "learning_rate": 1.2126428114301204e-06,
+      "loss": 0.5838,
+      "step": 4286
+    },
+    {
+      "epoch": 2.5389398874740894,
+      "grad_norm": 1.7988812066523692,
+      "learning_rate": 1.2095918619286306e-06,
+      "loss": 0.611,
+      "step": 4287
+    },
+    {
+      "epoch": 2.539532129108676,
+      "grad_norm": 5.152642550826023,
+      "learning_rate": 1.2065445081983795e-06,
+      "loss": 0.627,
+      "step": 4288
+    },
+    {
+      "epoch": 2.5401243707432632,
+      "grad_norm": 2.539557904372852,
+      "learning_rate": 1.2035007514859054e-06,
+      "loss": 0.5975,
+      "step": 4289
+    },
+    {
+      "epoch": 2.5407166123778504,
+      "grad_norm": 1.4457429738287353,
+      "learning_rate": 1.2004605930362724e-06,
+      "loss": 0.566,
+      "step": 4290
+    },
+    {
+      "epoch": 2.541308854012437,
+      "grad_norm": 3.4785094614740677,
+      "learning_rate": 1.19742403409308e-06,
+      "loss": 0.5893,
+      "step": 4291
+    },
+    {
+      "epoch": 2.541901095647024,
+      "grad_norm": 2.4252390321481703,
+      "learning_rate": 1.194391075898451e-06,
+      "loss": 0.5936,
+      "step": 4292
+    },
+    {
+      "epoch": 2.542493337281611,
+      "grad_norm": 1.483539871741351,
+      "learning_rate": 1.191361719693036e-06,
+      "loss": 0.6091,
+      "step": 4293
+    },
+    {
+      "epoch": 2.5430855789161977,
+      "grad_norm": 1.7394003556175763,
+      "learning_rate": 1.1883359667160087e-06,
+      "loss": 0.6362,
+      "step": 4294
+    },
+    {
+      "epoch": 2.543677820550785,
+      "grad_norm": 1.2312450282067198,
+      "learning_rate": 1.185313818205076e-06,
+      "loss": 0.6277,
+      "step": 4295
+    },
+    {
+      "epoch": 2.5442700621853716,
+      "grad_norm": 1.2235952845823084,
+      "learning_rate": 1.1822952753964667e-06,
+      "loss": 0.5776,
+      "step": 4296
+    },
+    {
+      "epoch": 2.5448623038199587,
+      "grad_norm": 1.3836979953577477,
+      "learning_rate": 1.179280339524933e-06,
+      "loss": 0.6458,
+      "step": 4297
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 1.550091891298104,
+      "learning_rate": 1.1762690118237518e-06,
+      "loss": 0.6219,
+      "step": 4298
+    },
+    {
+      "epoch": 2.546046787089132,
+      "grad_norm": 2.6309104401555286,
+      "learning_rate": 1.1732612935247267e-06,
+      "loss": 0.6267,
+      "step": 4299
+    },
+    {
+      "epoch": 2.5466390287237193,
+      "grad_norm": 1.7871342600964928,
+      "learning_rate": 1.1702571858581867e-06,
+      "loss": 0.6586,
+      "step": 4300
+    },
+    {
+      "epoch": 2.5472312703583064,
+      "grad_norm": 3.2118108366654257,
+      "learning_rate": 1.167256690052978e-06,
+      "loss": 0.6428,
+      "step": 4301
+    },
+    {
+      "epoch": 2.547823511992893,
+      "grad_norm": 1.8557141426320194,
+      "learning_rate": 1.1642598073364707e-06,
+      "loss": 0.5865,
+      "step": 4302
+    },
+    {
+      "epoch": 2.54841575362748,
+      "grad_norm": 1.7419326195322122,
+      "learning_rate": 1.1612665389345613e-06,
+      "loss": 0.586,
+      "step": 4303
+    },
+    {
+      "epoch": 2.549007995262067,
+      "grad_norm": 4.789685134855233,
+      "learning_rate": 1.158276886071662e-06,
+      "loss": 0.6206,
+      "step": 4304
+    },
+    {
+      "epoch": 2.5496002368966537,
+      "grad_norm": 1.681521269697521,
+      "learning_rate": 1.1552908499707115e-06,
+      "loss": 0.5966,
+      "step": 4305
+    },
+    {
+      "epoch": 2.550192478531241,
+      "grad_norm": 1.7353432303503473,
+      "learning_rate": 1.1523084318531641e-06,
+      "loss": 0.6347,
+      "step": 4306
+    },
+    {
+      "epoch": 2.5507847201658276,
+      "grad_norm": 1.4255690439020448,
+      "learning_rate": 1.1493296329390003e-06,
+      "loss": 0.6146,
+      "step": 4307
+    },
+    {
+      "epoch": 2.5513769618004147,
+      "grad_norm": 2.180356591475331,
+      "learning_rate": 1.1463544544467109e-06,
+      "loss": 0.6158,
+      "step": 4308
+    },
+    {
+      "epoch": 2.5519692034350014,
+      "grad_norm": 1.6710087648778493,
+      "learning_rate": 1.143382897593316e-06,
+      "loss": 0.6162,
+      "step": 4309
+    },
+    {
+      "epoch": 2.552561445069588,
+      "grad_norm": 1.5114876553797243,
+      "learning_rate": 1.1404149635943462e-06,
+      "loss": 0.5982,
+      "step": 4310
+    },
+    {
+      "epoch": 2.5531536867041753,
+      "grad_norm": 1.9274949790613265,
+      "learning_rate": 1.1374506536638574e-06,
+      "loss": 0.5799,
+      "step": 4311
+    },
+    {
+      "epoch": 2.5537459283387625,
+      "grad_norm": 1.1942475585564856,
+      "learning_rate": 1.134489969014414e-06,
+      "loss": 0.5993,
+      "step": 4312
+    },
+    {
+      "epoch": 2.554338169973349,
+      "grad_norm": 2.7949402099250795,
+      "learning_rate": 1.1315329108571072e-06,
+      "loss": 0.5959,
+      "step": 4313
+    },
+    {
+      "epoch": 2.554930411607936,
+      "grad_norm": 2.1002894683071345,
+      "learning_rate": 1.1285794804015349e-06,
+      "loss": 0.6032,
+      "step": 4314
+    },
+    {
+      "epoch": 2.555522653242523,
+      "grad_norm": 1.332872243171736,
+      "learning_rate": 1.125629678855822e-06,
+      "loss": 0.6241,
+      "step": 4315
+    },
+    {
+      "epoch": 2.5561148948771097,
+      "grad_norm": 2.3220966071152294,
+      "learning_rate": 1.1226835074265985e-06,
+      "loss": 0.6073,
+      "step": 4316
+    },
+    {
+      "epoch": 2.556707136511697,
+      "grad_norm": 1.166136350433321,
+      "learning_rate": 1.1197409673190186e-06,
+      "loss": 0.5955,
+      "step": 4317
+    },
+    {
+      "epoch": 2.5572993781462836,
+      "grad_norm": 1.4802954541867506,
+      "learning_rate": 1.1168020597367435e-06,
+      "loss": 0.6378,
+      "step": 4318
+    },
+    {
+      "epoch": 2.5578916197808708,
+      "grad_norm": 1.4086127534519697,
+      "learning_rate": 1.1138667858819497e-06,
+      "loss": 0.5804,
+      "step": 4319
+    },
+    {
+      "epoch": 2.5584838614154575,
+      "grad_norm": 1.6694041074811221,
+      "learning_rate": 1.1109351469553331e-06,
+      "loss": 0.6109,
+      "step": 4320
+    },
+    {
+      "epoch": 2.559076103050044,
+      "grad_norm": 1.81052273937072,
+      "learning_rate": 1.1080071441560992e-06,
+      "loss": 0.6108,
+      "step": 4321
+    },
+    {
+      "epoch": 2.5596683446846313,
+      "grad_norm": 2.0222094639871147,
+      "learning_rate": 1.105082778681964e-06,
+      "loss": 0.5742,
+      "step": 4322
+    },
+    {
+      "epoch": 2.5602605863192185,
+      "grad_norm": 2.3303481014699003,
+      "learning_rate": 1.1021620517291566e-06,
+      "loss": 0.6679,
+      "step": 4323
+    },
+    {
+      "epoch": 2.560852827953805,
+      "grad_norm": 1.2189662484050676,
+      "learning_rate": 1.0992449644924186e-06,
+      "loss": 0.5603,
+      "step": 4324
+    },
+    {
+      "epoch": 2.561445069588392,
+      "grad_norm": 1.6090711124893189,
+      "learning_rate": 1.0963315181650058e-06,
+      "loss": 0.5734,
+      "step": 4325
+    },
+    {
+      "epoch": 2.562037311222979,
+      "grad_norm": 1.4371331051564935,
+      "learning_rate": 1.0934217139386805e-06,
+      "loss": 0.6292,
+      "step": 4326
+    },
+    {
+      "epoch": 2.5626295528575658,
+      "grad_norm": 1.637674628017765,
+      "learning_rate": 1.0905155530037125e-06,
+      "loss": 0.6119,
+      "step": 4327
+    },
+    {
+      "epoch": 2.563221794492153,
+      "grad_norm": 2.25569504293535,
+      "learning_rate": 1.087613036548888e-06,
+      "loss": 0.5816,
+      "step": 4328
+    },
+    {
+      "epoch": 2.5638140361267396,
+      "grad_norm": 2.2331219302066136,
+      "learning_rate": 1.0847141657615023e-06,
+      "loss": 0.6207,
+      "step": 4329
+    },
+    {
+      "epoch": 2.564406277761327,
+      "grad_norm": 1.0824128878079005,
+      "learning_rate": 1.0818189418273527e-06,
+      "loss": 0.6287,
+      "step": 4330
+    },
+    {
+      "epoch": 2.5649985193959135,
+      "grad_norm": 1.7593659910084987,
+      "learning_rate": 1.0789273659307476e-06,
+      "loss": 0.6439,
+      "step": 4331
+    },
+    {
+      "epoch": 2.5655907610305,
+      "grad_norm": 1.761466082062497,
+      "learning_rate": 1.0760394392545058e-06,
+      "loss": 0.6039,
+      "step": 4332
+    },
+    {
+      "epoch": 2.5661830026650874,
+      "grad_norm": 1.7622676780184343,
+      "learning_rate": 1.0731551629799542e-06,
+      "loss": 0.5764,
+      "step": 4333
+    },
+    {
+      "epoch": 2.5667752442996745,
+      "grad_norm": 3.3213032917052576,
+      "learning_rate": 1.0702745382869207e-06,
+      "loss": 0.6205,
+      "step": 4334
+    },
+    {
+      "epoch": 2.5673674859342612,
+      "grad_norm": 1.14995546121375,
+      "learning_rate": 1.0673975663537428e-06,
+      "loss": 0.5977,
+      "step": 4335
+    },
+    {
+      "epoch": 2.567959727568848,
+      "grad_norm": 1.225755679706984,
+      "learning_rate": 1.064524248357265e-06,
+      "loss": 0.6445,
+      "step": 4336
+    },
+    {
+      "epoch": 2.568551969203435,
+      "grad_norm": 1.9102011927596232,
+      "learning_rate": 1.0616545854728388e-06,
+      "loss": 0.6223,
+      "step": 4337
+    },
+    {
+      "epoch": 2.569144210838022,
+      "grad_norm": 1.644287131836029,
+      "learning_rate": 1.0587885788743112e-06,
+      "loss": 0.6169,
+      "step": 4338
+    },
+    {
+      "epoch": 2.569736452472609,
+      "grad_norm": 2.229153860357448,
+      "learning_rate": 1.0559262297340433e-06,
+      "loss": 0.6087,
+      "step": 4339
+    },
+    {
+      "epoch": 2.5703286941071957,
+      "grad_norm": 1.8865674189894543,
+      "learning_rate": 1.0530675392228995e-06,
+      "loss": 0.6173,
+      "step": 4340
+    },
+    {
+      "epoch": 2.570920935741783,
+      "grad_norm": 6.002385683935582,
+      "learning_rate": 1.0502125085102422e-06,
+      "loss": 0.6208,
+      "step": 4341
+    },
+    {
+      "epoch": 2.5715131773763695,
+      "grad_norm": 2.24501027749272,
+      "learning_rate": 1.047361138763937e-06,
+      "loss": 0.6327,
+      "step": 4342
+    },
+    {
+      "epoch": 2.5721054190109562,
+      "grad_norm": 1.488334337447085,
+      "learning_rate": 1.0445134311503592e-06,
+      "loss": 0.6553,
+      "step": 4343
+    },
+    {
+      "epoch": 2.5726976606455434,
+      "grad_norm": 5.179170676032384,
+      "learning_rate": 1.0416693868343796e-06,
+      "loss": 0.5995,
+      "step": 4344
+    },
+    {
+      "epoch": 2.5732899022801305,
+      "grad_norm": 2.911488695193342,
+      "learning_rate": 1.0388290069793726e-06,
+      "loss": 0.6315,
+      "step": 4345
+    },
+    {
+      "epoch": 2.5738821439147173,
+      "grad_norm": 1.257777620049107,
+      "learning_rate": 1.035992292747211e-06,
+      "loss": 0.5844,
+      "step": 4346
+    },
+    {
+      "epoch": 2.574474385549304,
+      "grad_norm": 1.0731010623057406,
+      "learning_rate": 1.0331592452982718e-06,
+      "loss": 0.5987,
+      "step": 4347
+    },
+    {
+      "epoch": 2.575066627183891,
+      "grad_norm": 3.003901689507578,
+      "learning_rate": 1.030329865791434e-06,
+      "loss": 0.6051,
+      "step": 4348
+    },
+    {
+      "epoch": 2.575658868818478,
+      "grad_norm": 1.7372715719465501,
+      "learning_rate": 1.0275041553840691e-06,
+      "loss": 0.6226,
+      "step": 4349
+    },
+    {
+      "epoch": 2.576251110453065,
+      "grad_norm": 1.7371414977000834,
+      "learning_rate": 1.0246821152320507e-06,
+      "loss": 0.5935,
+      "step": 4350
+    },
+    {
+      "epoch": 2.5768433520876517,
+      "grad_norm": 1.408153611176849,
+      "learning_rate": 1.0218637464897541e-06,
+      "loss": 0.5643,
+      "step": 4351
+    },
+    {
+      "epoch": 2.577435593722239,
+      "grad_norm": 1.780798414186877,
+      "learning_rate": 1.0190490503100515e-06,
+      "loss": 0.6226,
+      "step": 4352
+    },
+    {
+      "epoch": 2.5780278353568256,
+      "grad_norm": 2.2724733020724424,
+      "learning_rate": 1.0162380278443107e-06,
+      "loss": 0.6212,
+      "step": 4353
+    },
+    {
+      "epoch": 2.5786200769914123,
+      "grad_norm": 2.3417587051881874,
+      "learning_rate": 1.0134306802423965e-06,
+      "loss": 0.5835,
+      "step": 4354
+    },
+    {
+      "epoch": 2.5792123186259994,
+      "grad_norm": 2.1212455863514625,
+      "learning_rate": 1.010627008652675e-06,
+      "loss": 0.6133,
+      "step": 4355
+    },
+    {
+      "epoch": 2.5798045602605866,
+      "grad_norm": 2.605778840709358,
+      "learning_rate": 1.0078270142220015e-06,
+      "loss": 0.6165,
+      "step": 4356
+    },
+    {
+      "epoch": 2.5803968018951733,
+      "grad_norm": 1.30684333675587,
+      "learning_rate": 1.0050306980957358e-06,
+      "loss": 0.6219,
+      "step": 4357
+    },
+    {
+      "epoch": 2.58098904352976,
+      "grad_norm": 1.3407378615869856,
+      "learning_rate": 1.0022380614177251e-06,
+      "loss": 0.6391,
+      "step": 4358
+    },
+    {
+      "epoch": 2.581581285164347,
+      "grad_norm": 1.673176460809278,
+      "learning_rate": 9.994491053303169e-07,
+      "loss": 0.6031,
+      "step": 4359
+    },
+    {
+      "epoch": 2.582173526798934,
+      "grad_norm": 1.3507011115002057,
+      "learning_rate": 9.966638309743481e-07,
+      "loss": 0.6097,
+      "step": 4360
+    },
+    {
+      "epoch": 2.582765768433521,
+      "grad_norm": 1.31675965356572,
+      "learning_rate": 9.938822394891568e-07,
+      "loss": 0.613,
+      "step": 4361
+    },
+    {
+      "epoch": 2.5833580100681077,
+      "grad_norm": 1.3590258887169568,
+      "learning_rate": 9.911043320125657e-07,
+      "loss": 0.6662,
+      "step": 4362
+    },
+    {
+      "epoch": 2.583950251702695,
+      "grad_norm": 1.8770135822453762,
+      "learning_rate": 9.883301096808995e-07,
+      "loss": 0.6422,
+      "step": 4363
+    },
+    {
+      "epoch": 2.5845424933372816,
+      "grad_norm": 2.121220810517652,
+      "learning_rate": 9.85559573628967e-07,
+      "loss": 0.6463,
+      "step": 4364
+    },
+    {
+      "epoch": 2.5851347349718683,
+      "grad_norm": 1.939410319473363,
+      "learning_rate": 9.827927249900782e-07,
+      "loss": 0.6276,
+      "step": 4365
+    },
+    {
+      "epoch": 2.5857269766064555,
+      "grad_norm": 1.5896698304385093,
+      "learning_rate": 9.800295648960245e-07,
+      "loss": 0.5878,
+      "step": 4366
+    },
+    {
+      "epoch": 2.5863192182410426,
+      "grad_norm": 2.349609733093395,
+      "learning_rate": 9.772700944770973e-07,
+      "loss": 0.639,
+      "step": 4367
+    },
+    {
+      "epoch": 2.5869114598756293,
+      "grad_norm": 1.502375187876737,
+      "learning_rate": 9.74514314862074e-07,
+      "loss": 0.623,
+      "step": 4368
+    },
+    {
+      "epoch": 2.587503701510216,
+      "grad_norm": 1.2592816904026312,
+      "learning_rate": 9.717622271782234e-07,
+      "loss": 0.6151,
+      "step": 4369
+    },
+    {
+      "epoch": 2.588095943144803,
+      "grad_norm": 1.5838796869432388,
+      "learning_rate": 9.690138325513043e-07,
+      "loss": 0.6271,
+      "step": 4370
+    },
+    {
+      "epoch": 2.58868818477939,
+      "grad_norm": 3.9815136759012213,
+      "learning_rate": 9.66269132105565e-07,
+      "loss": 0.6234,
+      "step": 4371
+    },
+    {
+      "epoch": 2.589280426413977,
+      "grad_norm": 1.332493171381054,
+      "learning_rate": 9.635281269637409e-07,
+      "loss": 0.6035,
+      "step": 4372
+    },
+    {
+      "epoch": 2.5898726680485638,
+      "grad_norm": 1.7683470516926967,
+      "learning_rate": 9.607908182470593e-07,
+      "loss": 0.6349,
+      "step": 4373
+    },
+    {
+      "epoch": 2.590464909683151,
+      "grad_norm": 2.0170932132048334,
+      "learning_rate": 9.580572070752335e-07,
+      "loss": 0.6105,
+      "step": 4374
+    },
+    {
+      "epoch": 2.5910571513177376,
+      "grad_norm": 2.740791397745886,
+      "learning_rate": 9.553272945664604e-07,
+      "loss": 0.5933,
+      "step": 4375
+    },
+    {
+      "epoch": 2.5916493929523243,
+      "grad_norm": 1.4563051854293856,
+      "learning_rate": 9.52601081837431e-07,
+      "loss": 0.6438,
+      "step": 4376
+    },
+    {
+      "epoch": 2.5922416345869115,
+      "grad_norm": 2.7013335054171974,
+      "learning_rate": 9.498785700033197e-07,
+      "loss": 0.6372,
+      "step": 4377
+    },
+    {
+      "epoch": 2.5928338762214986,
+      "grad_norm": 2.0736824122584423,
+      "learning_rate": 9.471597601777871e-07,
+      "loss": 0.6126,
+      "step": 4378
+    },
+    {
+      "epoch": 2.5934261178560853,
+      "grad_norm": 1.3687787675542844,
+      "learning_rate": 9.444446534729767e-07,
+      "loss": 0.6138,
+      "step": 4379
+    },
+    {
+      "epoch": 2.594018359490672,
+      "grad_norm": 1.4601120548746158,
+      "learning_rate": 9.417332509995203e-07,
+      "loss": 0.5814,
+      "step": 4380
+    },
+    {
+      "epoch": 2.594610601125259,
+      "grad_norm": 1.6919905995540607,
+      "learning_rate": 9.390255538665383e-07,
+      "loss": 0.6357,
+      "step": 4381
+    },
+    {
+      "epoch": 2.595202842759846,
+      "grad_norm": 1.5823616402702225,
+      "learning_rate": 9.36321563181628e-07,
+      "loss": 0.585,
+      "step": 4382
+    },
+    {
+      "epoch": 2.595795084394433,
+      "grad_norm": 1.960553817746592,
+      "learning_rate": 9.336212800508715e-07,
+      "loss": 0.6303,
+      "step": 4383
+    },
+    {
+      "epoch": 2.59638732602902,
+      "grad_norm": 17.328103997100726,
+      "learning_rate": 9.309247055788384e-07,
+      "loss": 0.6071,
+      "step": 4384
+    },
+    {
+      "epoch": 2.596979567663607,
+      "grad_norm": 2.096167704908563,
+      "learning_rate": 9.282318408685809e-07,
+      "loss": 0.6205,
+      "step": 4385
+    },
+    {
+      "epoch": 2.5975718092981936,
+      "grad_norm": 3.5082685509985216,
+      "learning_rate": 9.255426870216311e-07,
+      "loss": 0.6005,
+      "step": 4386
+    },
+    {
+      "epoch": 2.5981640509327804,
+      "grad_norm": 1.6098721047621571,
+      "learning_rate": 9.228572451380024e-07,
+      "loss": 0.6077,
+      "step": 4387
+    },
+    {
+      "epoch": 2.5987562925673675,
+      "grad_norm": 2.5927416189734593,
+      "learning_rate": 9.201755163161918e-07,
+      "loss": 0.597,
+      "step": 4388
+    },
+    {
+      "epoch": 2.5993485342019547,
+      "grad_norm": 1.9744478641135261,
+      "learning_rate": 9.174975016531828e-07,
+      "loss": 0.5956,
+      "step": 4389
+    },
+    {
+      "epoch": 2.5999407758365414,
+      "grad_norm": 1.587520907903784,
+      "learning_rate": 9.148232022444259e-07,
+      "loss": 0.5978,
+      "step": 4390
+    },
+    {
+      "epoch": 2.600533017471128,
+      "grad_norm": 1.7030721112666247,
+      "learning_rate": 9.121526191838626e-07,
+      "loss": 0.6412,
+      "step": 4391
+    },
+    {
+      "epoch": 2.6011252591057152,
+      "grad_norm": 1.6855774573137718,
+      "learning_rate": 9.094857535639157e-07,
+      "loss": 0.6673,
+      "step": 4392
+    },
+    {
+      "epoch": 2.601717500740302,
+      "grad_norm": 1.670526552319908,
+      "learning_rate": 9.068226064754781e-07,
+      "loss": 0.6057,
+      "step": 4393
+    },
+    {
+      "epoch": 2.602309742374889,
+      "grad_norm": 3.2516145639118346,
+      "learning_rate": 9.041631790079275e-07,
+      "loss": 0.607,
+      "step": 4394
+    },
+    {
+      "epoch": 2.602901984009476,
+      "grad_norm": 2.0502691399541777,
+      "learning_rate": 9.015074722491212e-07,
+      "loss": 0.6516,
+      "step": 4395
+    },
+    {
+      "epoch": 2.603494225644063,
+      "grad_norm": 1.9295772880960242,
+      "learning_rate": 8.988554872853927e-07,
+      "loss": 0.6131,
+      "step": 4396
+    },
+    {
+      "epoch": 2.6040864672786497,
+      "grad_norm": 2.8204499089521318,
+      "learning_rate": 8.962072252015519e-07,
+      "loss": 0.6764,
+      "step": 4397
+    },
+    {
+      "epoch": 2.6046787089132364,
+      "grad_norm": 1.449343492660855,
+      "learning_rate": 8.935626870808856e-07,
+      "loss": 0.5797,
+      "step": 4398
+    },
+    {
+      "epoch": 2.6052709505478235,
+      "grad_norm": 1.4023824555418185,
+      "learning_rate": 8.909218740051596e-07,
+      "loss": 0.6024,
+      "step": 4399
+    },
+    {
+      "epoch": 2.6058631921824107,
+      "grad_norm": 1.9603813068399214,
+      "learning_rate": 8.882847870546174e-07,
+      "loss": 0.5941,
+      "step": 4400
+    },
+    {
+      "epoch": 2.6064554338169974,
+      "grad_norm": 2.1424478675974252,
+      "learning_rate": 8.856514273079741e-07,
+      "loss": 0.6012,
+      "step": 4401
+    },
+    {
+      "epoch": 2.607047675451584,
+      "grad_norm": 5.123535144408223,
+      "learning_rate": 8.830217958424192e-07,
+      "loss": 0.5819,
+      "step": 4402
+    },
+    {
+      "epoch": 2.6076399170861713,
+      "grad_norm": 2.2597985178839117,
+      "learning_rate": 8.803958937336221e-07,
+      "loss": 0.6186,
+      "step": 4403
+    },
+    {
+      "epoch": 2.608232158720758,
+      "grad_norm": 2.218872286361537,
+      "learning_rate": 8.77773722055727e-07,
+      "loss": 0.6509,
+      "step": 4404
+    },
+    {
+      "epoch": 2.608824400355345,
+      "grad_norm": 1.4707962923229974,
+      "learning_rate": 8.751552818813469e-07,
+      "loss": 0.6754,
+      "step": 4405
+    },
+    {
+      "epoch": 2.609416641989932,
+      "grad_norm": 1.45480818288004,
+      "learning_rate": 8.725405742815695e-07,
+      "loss": 0.5915,
+      "step": 4406
+    },
+    {
+      "epoch": 2.610008883624519,
+      "grad_norm": 1.3986566797454838,
+      "learning_rate": 8.699296003259594e-07,
+      "loss": 0.5885,
+      "step": 4407
+    },
+    {
+      "epoch": 2.6106011252591057,
+      "grad_norm": 5.235464145421576,
+      "learning_rate": 8.673223610825532e-07,
+      "loss": 0.6181,
+      "step": 4408
+    },
+    {
+      "epoch": 2.6111933668936924,
+      "grad_norm": 1.2189308602666966,
+      "learning_rate": 8.647188576178567e-07,
+      "loss": 0.606,
+      "step": 4409
+    },
+    {
+      "epoch": 2.6117856085282796,
+      "grad_norm": 1.4608812222029703,
+      "learning_rate": 8.62119090996848e-07,
+      "loss": 0.6097,
+      "step": 4410
+    },
+    {
+      "epoch": 2.6123778501628667,
+      "grad_norm": 1.7812634754687648,
+      "learning_rate": 8.595230622829797e-07,
+      "loss": 0.6253,
+      "step": 4411
+    },
+    {
+      "epoch": 2.6129700917974534,
+      "grad_norm": 1.3064781743089184,
+      "learning_rate": 8.569307725381715e-07,
+      "loss": 0.6165,
+      "step": 4412
+    },
+    {
+      "epoch": 2.61356233343204,
+      "grad_norm": 1.335671673105394,
+      "learning_rate": 8.543422228228182e-07,
+      "loss": 0.6033,
+      "step": 4413
+    },
+    {
+      "epoch": 2.6141545750666273,
+      "grad_norm": 1.5391838884745948,
+      "learning_rate": 8.517574141957796e-07,
+      "loss": 0.6247,
+      "step": 4414
+    },
+    {
+      "epoch": 2.614746816701214,
+      "grad_norm": 1.620678307778965,
+      "learning_rate": 8.491763477143911e-07,
+      "loss": 0.5832,
+      "step": 4415
+    },
+    {
+      "epoch": 2.615339058335801,
+      "grad_norm": 1.9416245565474328,
+      "learning_rate": 8.46599024434449e-07,
+      "loss": 0.6067,
+      "step": 4416
+    },
+    {
+      "epoch": 2.615931299970388,
+      "grad_norm": 3.5183251486251983,
+      "learning_rate": 8.440254454102303e-07,
+      "loss": 0.5849,
+      "step": 4417
+    },
+    {
+      "epoch": 2.616523541604975,
+      "grad_norm": 1.735719411627244,
+      "learning_rate": 8.414556116944672e-07,
+      "loss": 0.637,
+      "step": 4418
+    },
+    {
+      "epoch": 2.6171157832395617,
+      "grad_norm": 1.2608882302510347,
+      "learning_rate": 8.388895243383699e-07,
+      "loss": 0.6155,
+      "step": 4419
+    },
+    {
+      "epoch": 2.6177080248741484,
+      "grad_norm": 2.0273357584791865,
+      "learning_rate": 8.363271843916099e-07,
+      "loss": 0.6128,
+      "step": 4420
+    },
+    {
+      "epoch": 2.6183002665087356,
+      "grad_norm": 1.125910415428338,
+      "learning_rate": 8.33768592902332e-07,
+      "loss": 0.6086,
+      "step": 4421
+    },
+    {
+      "epoch": 2.6188925081433228,
+      "grad_norm": 2.818588547251622,
+      "learning_rate": 8.312137509171392e-07,
+      "loss": 0.5505,
+      "step": 4422
+    },
+    {
+      "epoch": 2.6194847497779095,
+      "grad_norm": 2.4748929370800457,
+      "learning_rate": 8.286626594811098e-07,
+      "loss": 0.5675,
+      "step": 4423
+    },
+    {
+      "epoch": 2.620076991412496,
+      "grad_norm": 2.5020858763222456,
+      "learning_rate": 8.261153196377814e-07,
+      "loss": 0.5973,
+      "step": 4424
+    },
+    {
+      "epoch": 2.6206692330470833,
+      "grad_norm": 5.677134946889842,
+      "learning_rate": 8.235717324291604e-07,
+      "loss": 0.622,
+      "step": 4425
+    },
+    {
+      "epoch": 2.62126147468167,
+      "grad_norm": 3.5621389375037813,
+      "learning_rate": 8.210318988957166e-07,
+      "loss": 0.584,
+      "step": 4426
+    },
+    {
+      "epoch": 2.621853716316257,
+      "grad_norm": 1.8691129401457045,
+      "learning_rate": 8.184958200763826e-07,
+      "loss": 0.5916,
+      "step": 4427
+    },
+    {
+      "epoch": 2.622445957950844,
+      "grad_norm": 4.897419852333168,
+      "learning_rate": 8.159634970085595e-07,
+      "loss": 0.5899,
+      "step": 4428
+    },
+    {
+      "epoch": 2.623038199585431,
+      "grad_norm": 1.4035158403121284,
+      "learning_rate": 8.134349307281109e-07,
+      "loss": 0.5928,
+      "step": 4429
+    },
+    {
+      "epoch": 2.6236304412200178,
+      "grad_norm": 2.473039251325127,
+      "learning_rate": 8.109101222693616e-07,
+      "loss": 0.5929,
+      "step": 4430
+    },
+    {
+      "epoch": 2.6242226828546045,
+      "grad_norm": 2.0204745202248353,
+      "learning_rate": 8.083890726650978e-07,
+      "loss": 0.6055,
+      "step": 4431
+    },
+    {
+      "epoch": 2.6248149244891916,
+      "grad_norm": 1.4459314914665022,
+      "learning_rate": 8.058717829465723e-07,
+      "loss": 0.6025,
+      "step": 4432
+    },
+    {
+      "epoch": 2.6254071661237783,
+      "grad_norm": 2.1952381195541553,
+      "learning_rate": 8.033582541435003e-07,
+      "loss": 0.6302,
+      "step": 4433
+    },
+    {
+      "epoch": 2.6259994077583655,
+      "grad_norm": 1.7238053969572114,
+      "learning_rate": 8.008484872840538e-07,
+      "loss": 0.6178,
+      "step": 4434
+    },
+    {
+      "epoch": 2.626591649392952,
+      "grad_norm": 1.4421860814594145,
+      "learning_rate": 7.983424833948694e-07,
+      "loss": 0.6132,
+      "step": 4435
+    },
+    {
+      "epoch": 2.6271838910275394,
+      "grad_norm": 33.61594311760086,
+      "learning_rate": 7.958402435010415e-07,
+      "loss": 0.6302,
+      "step": 4436
+    },
+    {
+      "epoch": 2.627776132662126,
+      "grad_norm": 1.3890112121581686,
+      "learning_rate": 7.933417686261325e-07,
+      "loss": 0.6571,
+      "step": 4437
+    },
+    {
+      "epoch": 2.628368374296713,
+      "grad_norm": 1.5916247723609014,
+      "learning_rate": 7.908470597921547e-07,
+      "loss": 0.6012,
+      "step": 4438
+    },
+    {
+      "epoch": 2.6289606159313,
+      "grad_norm": 1.1892053559966738,
+      "learning_rate": 7.883561180195831e-07,
+      "loss": 0.6098,
+      "step": 4439
+    },
+    {
+      "epoch": 2.629552857565887,
+      "grad_norm": 3.0816929839131504,
+      "learning_rate": 7.858689443273548e-07,
+      "loss": 0.6013,
+      "step": 4440
+    },
+    {
+      "epoch": 2.630145099200474,
+      "grad_norm": 3.903155006740842,
+      "learning_rate": 7.833855397328682e-07,
+      "loss": 0.5941,
+      "step": 4441
+    },
+    {
+      "epoch": 2.6307373408350605,
+      "grad_norm": 2.618889450319004,
+      "learning_rate": 7.809059052519674e-07,
+      "loss": 0.644,
+      "step": 4442
+    },
+    {
+      "epoch": 2.6313295824696477,
+      "grad_norm": 1.993232060605033,
+      "learning_rate": 7.784300418989665e-07,
+      "loss": 0.591,
+      "step": 4443
+    },
+    {
+      "epoch": 2.6319218241042344,
+      "grad_norm": 1.2377163825617075,
+      "learning_rate": 7.759579506866311e-07,
+      "loss": 0.5866,
+      "step": 4444
+    },
+    {
+      "epoch": 2.6325140657388215,
+      "grad_norm": 1.843881731614733,
+      "learning_rate": 7.73489632626192e-07,
+      "loss": 0.642,
+      "step": 4445
+    },
+    {
+      "epoch": 2.6331063073734082,
+      "grad_norm": 2.4131522683099518,
+      "learning_rate": 7.710250887273196e-07,
+      "loss": 0.6004,
+      "step": 4446
+    },
+    {
+      "epoch": 2.6336985490079954,
+      "grad_norm": 1.1424501229679849,
+      "learning_rate": 7.685643199981574e-07,
+      "loss": 0.6154,
+      "step": 4447
+    },
+    {
+      "epoch": 2.634290790642582,
+      "grad_norm": 1.353394407111203,
+      "learning_rate": 7.66107327445299e-07,
+      "loss": 0.6123,
+      "step": 4448
+    },
+    {
+      "epoch": 2.6348830322771692,
+      "grad_norm": 1.4623696640133033,
+      "learning_rate": 7.636541120737906e-07,
+      "loss": 0.5979,
+      "step": 4449
+    },
+    {
+      "epoch": 2.635475273911756,
+      "grad_norm": 1.6100152183594005,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.6087,
+      "step": 4450
+    },
+    {
+      "epoch": 2.636067515546343,
+      "grad_norm": 1.5347034575841871,
+      "learning_rate": 7.58759016887286e-07,
+      "loss": 0.6232,
+      "step": 4451
+    },
+    {
+      "epoch": 2.63665975718093,
+      "grad_norm": 1.487392673062803,
+      "learning_rate": 7.563171390746627e-07,
+      "loss": 0.5846,
+      "step": 4452
+    },
+    {
+      "epoch": 2.6372519988155165,
+      "grad_norm": 1.9959820986132857,
+      "learning_rate": 7.53879042448128e-07,
+      "loss": 0.6068,
+      "step": 4453
+    },
+    {
+      "epoch": 2.6378442404501037,
+      "grad_norm": 1.3610216722187833,
+      "learning_rate": 7.514447280049964e-07,
+      "loss": 0.5942,
+      "step": 4454
+    },
+    {
+      "epoch": 2.6384364820846904,
+      "grad_norm": 2.2754665990675136,
+      "learning_rate": 7.490141967410436e-07,
+      "loss": 0.6712,
+      "step": 4455
+    },
+    {
+      "epoch": 2.6390287237192775,
+      "grad_norm": 1.3895620105466362,
+      "learning_rate": 7.465874496504944e-07,
+      "loss": 0.6119,
+      "step": 4456
+    },
+    {
+      "epoch": 2.6396209653538643,
+      "grad_norm": 1.1470130133503953,
+      "learning_rate": 7.441644877260212e-07,
+      "loss": 0.5964,
+      "step": 4457
+    },
+    {
+      "epoch": 2.6402132069884514,
+      "grad_norm": 2.521466762407184,
+      "learning_rate": 7.417453119587525e-07,
+      "loss": 0.5729,
+      "step": 4458
+    },
+    {
+      "epoch": 2.640805448623038,
+      "grad_norm": 2.4309719657662376,
+      "learning_rate": 7.393299233382678e-07,
+      "loss": 0.6652,
+      "step": 4459
+    },
+    {
+      "epoch": 2.6413976902576253,
+      "grad_norm": 3.395055166928025,
+      "learning_rate": 7.369183228526e-07,
+      "loss": 0.6361,
+      "step": 4460
+    },
+    {
+      "epoch": 2.641989931892212,
+      "grad_norm": 2.006109738193744,
+      "learning_rate": 7.345105114882245e-07,
+      "loss": 0.5852,
+      "step": 4461
+    },
+    {
+      "epoch": 2.642582173526799,
+      "grad_norm": 1.4964269607335803,
+      "learning_rate": 7.321064902300723e-07,
+      "loss": 0.6232,
+      "step": 4462
+    },
+    {
+      "epoch": 2.643174415161386,
+      "grad_norm": 3.2714212450603393,
+      "learning_rate": 7.297062600615268e-07,
+      "loss": 0.6469,
+      "step": 4463
+    },
+    {
+      "epoch": 2.6437666567959726,
+      "grad_norm": 1.8897563765286818,
+      "learning_rate": 7.273098219644137e-07,
+      "loss": 0.6354,
+      "step": 4464
+    },
+    {
+      "epoch": 2.6443588984305597,
+      "grad_norm": 1.528350989418592,
+      "learning_rate": 7.249171769190111e-07,
+      "loss": 0.6233,
+      "step": 4465
+    },
+    {
+      "epoch": 2.6449511400651464,
+      "grad_norm": 3.275762358290711,
+      "learning_rate": 7.225283259040472e-07,
+      "loss": 0.578,
+      "step": 4466
+    },
+    {
+      "epoch": 2.6455433816997336,
+      "grad_norm": 1.5673100456277456,
+      "learning_rate": 7.201432698966959e-07,
+      "loss": 0.6478,
+      "step": 4467
+    },
+    {
+      "epoch": 2.6461356233343203,
+      "grad_norm": 1.447196429106465,
+      "learning_rate": 7.17762009872579e-07,
+      "loss": 0.5887,
+      "step": 4468
+    },
+    {
+      "epoch": 2.6467278649689074,
+      "grad_norm": 1.4964135311485778,
+      "learning_rate": 7.15384546805764e-07,
+      "loss": 0.6,
+      "step": 4469
+    },
+    {
+      "epoch": 2.647320106603494,
+      "grad_norm": 1.9472910578091667,
+      "learning_rate": 7.130108816687687e-07,
+      "loss": 0.5853,
+      "step": 4470
+    },
+    {
+      "epoch": 2.647912348238081,
+      "grad_norm": 1.2610856530380568,
+      "learning_rate": 7.106410154325571e-07,
+      "loss": 0.6081,
+      "step": 4471
+    },
+    {
+      "epoch": 2.648504589872668,
+      "grad_norm": 1.3825561685529,
+      "learning_rate": 7.082749490665353e-07,
+      "loss": 0.6435,
+      "step": 4472
+    },
+    {
+      "epoch": 2.649096831507255,
+      "grad_norm": 1.5107349206863576,
+      "learning_rate": 7.059126835385577e-07,
+      "loss": 0.6626,
+      "step": 4473
+    },
+    {
+      "epoch": 2.649689073141842,
+      "grad_norm": 1.9273475424947197,
+      "learning_rate": 7.035542198149237e-07,
+      "loss": 0.5902,
+      "step": 4474
+    },
+    {
+      "epoch": 2.6502813147764286,
+      "grad_norm": 2.5130850136023075,
+      "learning_rate": 7.011995588603804e-07,
+      "loss": 0.6471,
+      "step": 4475
+    },
+    {
+      "epoch": 2.6508735564110157,
+      "grad_norm": 1.5585005044488192,
+      "learning_rate": 6.98848701638114e-07,
+      "loss": 0.6371,
+      "step": 4476
+    },
+    {
+      "epoch": 2.6514657980456025,
+      "grad_norm": 1.3238990292587902,
+      "learning_rate": 6.965016491097553e-07,
+      "loss": 0.6118,
+      "step": 4477
+    },
+    {
+      "epoch": 2.6520580396801896,
+      "grad_norm": 1.3046884838922062,
+      "learning_rate": 6.941584022353865e-07,
+      "loss": 0.6345,
+      "step": 4478
+    },
+    {
+      "epoch": 2.6526502813147763,
+      "grad_norm": 5.787734168290906,
+      "learning_rate": 6.918189619735205e-07,
+      "loss": 0.6323,
+      "step": 4479
+    },
+    {
+      "epoch": 2.6532425229493635,
+      "grad_norm": 1.9229458354471165,
+      "learning_rate": 6.894833292811265e-07,
+      "loss": 0.6234,
+      "step": 4480
+    },
+    {
+      "epoch": 2.65383476458395,
+      "grad_norm": 1.4760976120663631,
+      "learning_rate": 6.871515051136046e-07,
+      "loss": 0.6093,
+      "step": 4481
+    },
+    {
+      "epoch": 2.654427006218537,
+      "grad_norm": 1.6700900299285095,
+      "learning_rate": 6.848234904248041e-07,
+      "loss": 0.6118,
+      "step": 4482
+    },
+    {
+      "epoch": 2.655019247853124,
+      "grad_norm": 2.3039568980390683,
+      "learning_rate": 6.824992861670132e-07,
+      "loss": 0.654,
+      "step": 4483
+    },
+    {
+      "epoch": 2.655611489487711,
+      "grad_norm": 1.779803366967843,
+      "learning_rate": 6.801788932909648e-07,
+      "loss": 0.6177,
+      "step": 4484
+    },
+    {
+      "epoch": 2.656203731122298,
+      "grad_norm": 1.4560279559121052,
+      "learning_rate": 6.778623127458261e-07,
+      "loss": 0.6,
+      "step": 4485
+    },
+    {
+      "epoch": 2.6567959727568846,
+      "grad_norm": 1.6569231878529236,
+      "learning_rate": 6.755495454792116e-07,
+      "loss": 0.6316,
+      "step": 4486
+    },
+    {
+      "epoch": 2.6573882143914718,
+      "grad_norm": 1.4672246843225127,
+      "learning_rate": 6.73240592437171e-07,
+      "loss": 0.5814,
+      "step": 4487
+    },
+    {
+      "epoch": 2.6579804560260585,
+      "grad_norm": 1.929651108949212,
+      "learning_rate": 6.709354545641989e-07,
+      "loss": 0.6056,
+      "step": 4488
+    },
+    {
+      "epoch": 2.6585726976606456,
+      "grad_norm": 1.1762025778237375,
+      "learning_rate": 6.68634132803222e-07,
+      "loss": 0.5838,
+      "step": 4489
+    },
+    {
+      "epoch": 2.6591649392952323,
+      "grad_norm": 1.2260691704307056,
+      "learning_rate": 6.663366280956152e-07,
+      "loss": 0.5867,
+      "step": 4490
+    },
+    {
+      "epoch": 2.6597571809298195,
+      "grad_norm": 1.5821020306273104,
+      "learning_rate": 6.640429413811833e-07,
+      "loss": 0.6146,
+      "step": 4491
+    },
+    {
+      "epoch": 2.660349422564406,
+      "grad_norm": 3.7438552072655042,
+      "learning_rate": 6.617530735981758e-07,
+      "loss": 0.6057,
+      "step": 4492
+    },
+    {
+      "epoch": 2.660941664198993,
+      "grad_norm": 1.4861734342086816,
+      "learning_rate": 6.594670256832769e-07,
+      "loss": 0.5755,
+      "step": 4493
+    },
+    {
+      "epoch": 2.66153390583358,
+      "grad_norm": 1.3176811467599943,
+      "learning_rate": 6.571847985716063e-07,
+      "loss": 0.6114,
+      "step": 4494
+    },
+    {
+      "epoch": 2.6621261474681672,
+      "grad_norm": 2.5936137237565986,
+      "learning_rate": 6.549063931967247e-07,
+      "loss": 0.6174,
+      "step": 4495
+    },
+    {
+      "epoch": 2.662718389102754,
+      "grad_norm": 1.5641301693338778,
+      "learning_rate": 6.526318104906293e-07,
+      "loss": 0.6262,
+      "step": 4496
+    },
+    {
+      "epoch": 2.6633106307373406,
+      "grad_norm": 1.7900396459995547,
+      "learning_rate": 6.503610513837522e-07,
+      "loss": 0.557,
+      "step": 4497
+    },
+    {
+      "epoch": 2.663902872371928,
+      "grad_norm": 1.3448660046809024,
+      "learning_rate": 6.480941168049593e-07,
+      "loss": 0.5565,
+      "step": 4498
+    },
+    {
+      "epoch": 2.6644951140065145,
+      "grad_norm": 1.5352709594541674,
+      "learning_rate": 6.458310076815544e-07,
+      "loss": 0.6334,
+      "step": 4499
+    },
+    {
+      "epoch": 2.6650873556411017,
+      "grad_norm": 1.5829649420114509,
+      "learning_rate": 6.435717249392803e-07,
+      "loss": 0.6498,
+      "step": 4500
+    },
+    {
+      "epoch": 2.6656795972756884,
+      "grad_norm": 1.8146453565537977,
+      "learning_rate": 6.413162695023078e-07,
+      "loss": 0.6241,
+      "step": 4501
+    },
+    {
+      "epoch": 2.6662718389102755,
+      "grad_norm": 1.6561074771734265,
+      "learning_rate": 6.390646422932445e-07,
+      "loss": 0.6356,
+      "step": 4502
+    },
+    {
+      "epoch": 2.6668640805448622,
+      "grad_norm": 1.4096379062668936,
+      "learning_rate": 6.368168442331324e-07,
+      "loss": 0.6571,
+      "step": 4503
+    },
+    {
+      "epoch": 2.667456322179449,
+      "grad_norm": 1.568445220559726,
+      "learning_rate": 6.345728762414504e-07,
+      "loss": 0.5933,
+      "step": 4504
+    },
+    {
+      "epoch": 2.668048563814036,
+      "grad_norm": 1.2456099463597277,
+      "learning_rate": 6.323327392361056e-07,
+      "loss": 0.604,
+      "step": 4505
+    },
+    {
+      "epoch": 2.6686408054486233,
+      "grad_norm": 2.1193064430307804,
+      "learning_rate": 6.300964341334382e-07,
+      "loss": 0.6107,
+      "step": 4506
+    },
+    {
+      "epoch": 2.66923304708321,
+      "grad_norm": 1.981391634375128,
+      "learning_rate": 6.278639618482241e-07,
+      "loss": 0.6398,
+      "step": 4507
+    },
+    {
+      "epoch": 2.6698252887177967,
+      "grad_norm": 2.04267323424306,
+      "learning_rate": 6.256353232936718e-07,
+      "loss": 0.5972,
+      "step": 4508
+    },
+    {
+      "epoch": 2.670417530352384,
+      "grad_norm": 1.8887831971144367,
+      "learning_rate": 6.234105193814177e-07,
+      "loss": 0.6266,
+      "step": 4509
+    },
+    {
+      "epoch": 2.6710097719869705,
+      "grad_norm": 1.7381054901434658,
+      "learning_rate": 6.211895510215316e-07,
+      "loss": 0.6574,
+      "step": 4510
+    },
+    {
+      "epoch": 2.6716020136215577,
+      "grad_norm": 1.443595604409538,
+      "learning_rate": 6.189724191225155e-07,
+      "loss": 0.6292,
+      "step": 4511
+    },
+    {
+      "epoch": 2.6721942552561444,
+      "grad_norm": 1.6341383418243982,
+      "learning_rate": 6.167591245913029e-07,
+      "loss": 0.6265,
+      "step": 4512
+    },
+    {
+      "epoch": 2.6727864968907316,
+      "grad_norm": 1.468790273283455,
+      "learning_rate": 6.145496683332508e-07,
+      "loss": 0.6301,
+      "step": 4513
+    },
+    {
+      "epoch": 2.6733787385253183,
+      "grad_norm": 1.3435540901590721,
+      "learning_rate": 6.123440512521539e-07,
+      "loss": 0.6249,
+      "step": 4514
+    },
+    {
+      "epoch": 2.673970980159905,
+      "grad_norm": 2.159903078044722,
+      "learning_rate": 6.101422742502349e-07,
+      "loss": 0.6205,
+      "step": 4515
+    },
+    {
+      "epoch": 2.674563221794492,
+      "grad_norm": 1.5612885890524655,
+      "learning_rate": 6.079443382281424e-07,
+      "loss": 0.6187,
+      "step": 4516
+    },
+    {
+      "epoch": 2.6751554634290793,
+      "grad_norm": 1.9002762382063694,
+      "learning_rate": 6.05750244084956e-07,
+      "loss": 0.6108,
+      "step": 4517
+    },
+    {
+      "epoch": 2.675747705063666,
+      "grad_norm": 3.9695947033078127,
+      "learning_rate": 6.035599927181834e-07,
+      "loss": 0.6114,
+      "step": 4518
+    },
+    {
+      "epoch": 2.6763399466982527,
+      "grad_norm": 1.3099993291001881,
+      "learning_rate": 6.013735850237623e-07,
+      "loss": 0.603,
+      "step": 4519
+    },
+    {
+      "epoch": 2.67693218833284,
+      "grad_norm": 1.2559274936986125,
+      "learning_rate": 5.99191021896055e-07,
+      "loss": 0.591,
+      "step": 4520
+    },
+    {
+      "epoch": 2.6775244299674266,
+      "grad_norm": 1.494733461660258,
+      "learning_rate": 5.97012304227852e-07,
+      "loss": 0.5815,
+      "step": 4521
+    },
+    {
+      "epoch": 2.6781166716020137,
+      "grad_norm": 3.1926894336773475,
+      "learning_rate": 5.948374329103723e-07,
+      "loss": 0.5866,
+      "step": 4522
+    },
+    {
+      "epoch": 2.6787089132366004,
+      "grad_norm": 1.4852694724957944,
+      "learning_rate": 5.926664088332612e-07,
+      "loss": 0.5798,
+      "step": 4523
+    },
+    {
+      "epoch": 2.6793011548711876,
+      "grad_norm": 2.351294923764927,
+      "learning_rate": 5.904992328845893e-07,
+      "loss": 0.6286,
+      "step": 4524
+    },
+    {
+      "epoch": 2.6798933965057743,
+      "grad_norm": 1.243806142521426,
+      "learning_rate": 5.88335905950852e-07,
+      "loss": 0.6393,
+      "step": 4525
+    },
+    {
+      "epoch": 2.680485638140361,
+      "grad_norm": 1.4595958476913502,
+      "learning_rate": 5.861764289169713e-07,
+      "loss": 0.6362,
+      "step": 4526
+    },
+    {
+      "epoch": 2.681077879774948,
+      "grad_norm": 2.5372193151492066,
+      "learning_rate": 5.840208026662986e-07,
+      "loss": 0.6335,
+      "step": 4527
+    },
+    {
+      "epoch": 2.6816701214095353,
+      "grad_norm": 1.9022729218122443,
+      "learning_rate": 5.818690280806038e-07,
+      "loss": 0.6046,
+      "step": 4528
+    },
+    {
+      "epoch": 2.682262363044122,
+      "grad_norm": 1.300761914523257,
+      "learning_rate": 5.797211060400809e-07,
+      "loss": 0.6347,
+      "step": 4529
+    },
+    {
+      "epoch": 2.6828546046787087,
+      "grad_norm": 1.545954882819192,
+      "learning_rate": 5.775770374233558e-07,
+      "loss": 0.6014,
+      "step": 4530
+    },
+    {
+      "epoch": 2.683446846313296,
+      "grad_norm": 1.3134870233911644,
+      "learning_rate": 5.754368231074703e-07,
+      "loss": 0.624,
+      "step": 4531
+    },
+    {
+      "epoch": 2.6840390879478826,
+      "grad_norm": 1.257399217998619,
+      "learning_rate": 5.73300463967893e-07,
+      "loss": 0.6017,
+      "step": 4532
+    },
+    {
+      "epoch": 2.6846313295824697,
+      "grad_norm": 2.0429701939538267,
+      "learning_rate": 5.711679608785136e-07,
+      "loss": 0.6038,
+      "step": 4533
+    },
+    {
+      "epoch": 2.6852235712170565,
+      "grad_norm": 2.112515949275844,
+      "learning_rate": 5.690393147116491e-07,
+      "loss": 0.595,
+      "step": 4534
+    },
+    {
+      "epoch": 2.6858158128516436,
+      "grad_norm": 1.4181848302497644,
+      "learning_rate": 5.669145263380316e-07,
+      "loss": 0.618,
+      "step": 4535
+    },
+    {
+      "epoch": 2.6864080544862303,
+      "grad_norm": 1.817228663528393,
+      "learning_rate": 5.647935966268225e-07,
+      "loss": 0.6222,
+      "step": 4536
+    },
+    {
+      "epoch": 2.687000296120817,
+      "grad_norm": 2.6043085112867095,
+      "learning_rate": 5.626765264455992e-07,
+      "loss": 0.6255,
+      "step": 4537
+    },
+    {
+      "epoch": 2.687592537755404,
+      "grad_norm": 1.3403299480462842,
+      "learning_rate": 5.60563316660363e-07,
+      "loss": 0.6334,
+      "step": 4538
+    },
+    {
+      "epoch": 2.6881847793899913,
+      "grad_norm": 1.5039519498314093,
+      "learning_rate": 5.58453968135535e-07,
+      "loss": 0.5825,
+      "step": 4539
+    },
+    {
+      "epoch": 2.688777021024578,
+      "grad_norm": 1.2364951870402103,
+      "learning_rate": 5.563484817339581e-07,
+      "loss": 0.6074,
+      "step": 4540
+    },
+    {
+      "epoch": 2.6893692626591648,
+      "grad_norm": 2.509772691752077,
+      "learning_rate": 5.542468583168936e-07,
+      "loss": 0.6249,
+      "step": 4541
+    },
+    {
+      "epoch": 2.689961504293752,
+      "grad_norm": 1.1947349763864274,
+      "learning_rate": 5.521490987440259e-07,
+      "loss": 0.6369,
+      "step": 4542
+    },
+    {
+      "epoch": 2.6905537459283386,
+      "grad_norm": 1.541811330073855,
+      "learning_rate": 5.500552038734541e-07,
+      "loss": 0.5862,
+      "step": 4543
+    },
+    {
+      "epoch": 2.691145987562926,
+      "grad_norm": 1.287641189747786,
+      "learning_rate": 5.47965174561701e-07,
+      "loss": 0.5996,
+      "step": 4544
+    },
+    {
+      "epoch": 2.6917382291975125,
+      "grad_norm": 1.7408712569531013,
+      "learning_rate": 5.458790116637036e-07,
+      "loss": 0.639,
+      "step": 4545
+    },
+    {
+      "epoch": 2.6923304708320996,
+      "grad_norm": 1.2121113431420716,
+      "learning_rate": 5.437967160328228e-07,
+      "loss": 0.5543,
+      "step": 4546
+    },
+    {
+      "epoch": 2.6929227124666864,
+      "grad_norm": 3.4295681334615797,
+      "learning_rate": 5.417182885208317e-07,
+      "loss": 0.6128,
+      "step": 4547
+    },
+    {
+      "epoch": 2.693514954101273,
+      "grad_norm": 2.73332565881467,
+      "learning_rate": 5.396437299779278e-07,
+      "loss": 0.6197,
+      "step": 4548
+    },
+    {
+      "epoch": 2.69410719573586,
+      "grad_norm": 1.7659745103441442,
+      "learning_rate": 5.375730412527191e-07,
+      "loss": 0.6616,
+      "step": 4549
+    },
+    {
+      "epoch": 2.6946994373704474,
+      "grad_norm": 6.0658236906308955,
+      "learning_rate": 5.355062231922326e-07,
+      "loss": 0.6502,
+      "step": 4550
+    },
+    {
+      "epoch": 2.695291679005034,
+      "grad_norm": 1.6358319710857092,
+      "learning_rate": 5.334432766419162e-07,
+      "loss": 0.6126,
+      "step": 4551
+    },
+    {
+      "epoch": 2.695883920639621,
+      "grad_norm": 1.8693759376554604,
+      "learning_rate": 5.313842024456306e-07,
+      "loss": 0.639,
+      "step": 4552
+    },
+    {
+      "epoch": 2.696476162274208,
+      "grad_norm": 1.2740182256023065,
+      "learning_rate": 5.29329001445652e-07,
+      "loss": 0.648,
+      "step": 4553
+    },
+    {
+      "epoch": 2.6970684039087947,
+      "grad_norm": 2.009984038445754,
+      "learning_rate": 5.272776744826724e-07,
+      "loss": 0.6539,
+      "step": 4554
+    },
+    {
+      "epoch": 2.697660645543382,
+      "grad_norm": 1.7618233562419785,
+      "learning_rate": 5.252302223958006e-07,
+      "loss": 0.5866,
+      "step": 4555
+    },
+    {
+      "epoch": 2.6982528871779685,
+      "grad_norm": 2.567467853853205,
+      "learning_rate": 5.231866460225621e-07,
+      "loss": 0.603,
+      "step": 4556
+    },
+    {
+      "epoch": 2.6988451288125557,
+      "grad_norm": 1.9843404125639688,
+      "learning_rate": 5.211469461988916e-07,
+      "loss": 0.6404,
+      "step": 4557
+    },
+    {
+      "epoch": 2.6994373704471424,
+      "grad_norm": 1.87203615558653,
+      "learning_rate": 5.191111237591406e-07,
+      "loss": 0.6232,
+      "step": 4558
+    },
+    {
+      "epoch": 2.700029612081729,
+      "grad_norm": 1.8160459925973307,
+      "learning_rate": 5.170791795360752e-07,
+      "loss": 0.6328,
+      "step": 4559
+    },
+    {
+      "epoch": 2.7006218537163162,
+      "grad_norm": 1.776112360381722,
+      "learning_rate": 5.150511143608782e-07,
+      "loss": 0.5592,
+      "step": 4560
+    },
+    {
+      "epoch": 2.7012140953509034,
+      "grad_norm": 1.7788576070929014,
+      "learning_rate": 5.130269290631407e-07,
+      "loss": 0.6206,
+      "step": 4561
+    },
+    {
+      "epoch": 2.70180633698549,
+      "grad_norm": 1.3844584078213082,
+      "learning_rate": 5.110066244708645e-07,
+      "loss": 0.5974,
+      "step": 4562
+    },
+    {
+      "epoch": 2.702398578620077,
+      "grad_norm": 2.280195440016197,
+      "learning_rate": 5.08990201410472e-07,
+      "loss": 0.6314,
+      "step": 4563
+    },
+    {
+      "epoch": 2.702990820254664,
+      "grad_norm": 1.5039208783289522,
+      "learning_rate": 5.069776607067944e-07,
+      "loss": 0.5886,
+      "step": 4564
+    },
+    {
+      "epoch": 2.7035830618892507,
+      "grad_norm": 2.515470221955119,
+      "learning_rate": 5.04969003183069e-07,
+      "loss": 0.5987,
+      "step": 4565
+    },
+    {
+      "epoch": 2.704175303523838,
+      "grad_norm": 1.4253523242169355,
+      "learning_rate": 5.029642296609538e-07,
+      "loss": 0.63,
+      "step": 4566
+    },
+    {
+      "epoch": 2.7047675451584245,
+      "grad_norm": 1.0838622160262505,
+      "learning_rate": 5.009633409605131e-07,
+      "loss": 0.5878,
+      "step": 4567
+    },
+    {
+      "epoch": 2.7053597867930117,
+      "grad_norm": 2.5944428791188345,
+      "learning_rate": 4.98966337900224e-07,
+      "loss": 0.577,
+      "step": 4568
+    },
+    {
+      "epoch": 2.7059520284275984,
+      "grad_norm": 1.6365174307980475,
+      "learning_rate": 4.969732212969691e-07,
+      "loss": 0.6307,
+      "step": 4569
+    },
+    {
+      "epoch": 2.706544270062185,
+      "grad_norm": 1.2985396532324271,
+      "learning_rate": 4.949839919660481e-07,
+      "loss": 0.5999,
+      "step": 4570
+    },
+    {
+      "epoch": 2.7071365116967723,
+      "grad_norm": 1.6192926604384,
+      "learning_rate": 4.929986507211681e-07,
+      "loss": 0.6267,
+      "step": 4571
+    },
+    {
+      "epoch": 2.7077287533313594,
+      "grad_norm": 1.550515843078195,
+      "learning_rate": 4.91017198374445e-07,
+      "loss": 0.6368,
+      "step": 4572
+    },
+    {
+      "epoch": 2.708320994965946,
+      "grad_norm": 3.445551249960511,
+      "learning_rate": 4.890396357364025e-07,
+      "loss": 0.6394,
+      "step": 4573
+    },
+    {
+      "epoch": 2.708913236600533,
+      "grad_norm": 2.0889306837221677,
+      "learning_rate": 4.870659636159758e-07,
+      "loss": 0.6194,
+      "step": 4574
+    },
+    {
+      "epoch": 2.70950547823512,
+      "grad_norm": 1.512212913808209,
+      "learning_rate": 4.850961828205103e-07,
+      "loss": 0.64,
+      "step": 4575
+    },
+    {
+      "epoch": 2.7100977198697067,
+      "grad_norm": 1.4199048585388745,
+      "learning_rate": 4.831302941557537e-07,
+      "loss": 0.6473,
+      "step": 4576
+    },
+    {
+      "epoch": 2.710689961504294,
+      "grad_norm": 1.785317344909293,
+      "learning_rate": 4.811682984258659e-07,
+      "loss": 0.6042,
+      "step": 4577
+    },
+    {
+      "epoch": 2.7112822031388806,
+      "grad_norm": 1.3142205419516673,
+      "learning_rate": 4.79210196433414e-07,
+      "loss": 0.6127,
+      "step": 4578
+    },
+    {
+      "epoch": 2.7118744447734677,
+      "grad_norm": 2.677760144755771,
+      "learning_rate": 4.772559889793716e-07,
+      "loss": 0.571,
+      "step": 4579
+    },
+    {
+      "epoch": 2.7124666864080544,
+      "grad_norm": 1.6507834481358612,
+      "learning_rate": 4.7530567686312035e-07,
+      "loss": 0.65,
+      "step": 4580
+    },
+    {
+      "epoch": 2.713058928042641,
+      "grad_norm": 1.8447154534722792,
+      "learning_rate": 4.7335926088244556e-07,
+      "loss": 0.5999,
+      "step": 4581
+    },
+    {
+      "epoch": 2.7136511696772283,
+      "grad_norm": 1.488823816754773,
+      "learning_rate": 4.7141674183354247e-07,
+      "loss": 0.6284,
+      "step": 4582
+    },
+    {
+      "epoch": 2.7142434113118155,
+      "grad_norm": 1.7787433279339757,
+      "learning_rate": 4.6947812051100995e-07,
+      "loss": 0.5897,
+      "step": 4583
+    },
+    {
+      "epoch": 2.714835652946402,
+      "grad_norm": 1.2062389769136208,
+      "learning_rate": 4.6754339770785474e-07,
+      "loss": 0.5984,
+      "step": 4584
+    },
+    {
+      "epoch": 2.715427894580989,
+      "grad_norm": 1.48700055953898,
+      "learning_rate": 4.6561257421548377e-07,
+      "loss": 0.5989,
+      "step": 4585
+    },
+    {
+      "epoch": 2.716020136215576,
+      "grad_norm": 1.7381187253725767,
+      "learning_rate": 4.636856508237164e-07,
+      "loss": 0.5941,
+      "step": 4586
+    },
+    {
+      "epoch": 2.7166123778501627,
+      "grad_norm": 8.014229956967728,
+      "learning_rate": 4.617626283207688e-07,
+      "loss": 0.6453,
+      "step": 4587
+    },
+    {
+      "epoch": 2.71720461948475,
+      "grad_norm": 1.655651410898122,
+      "learning_rate": 4.5984350749326835e-07,
+      "loss": 0.6157,
+      "step": 4588
+    },
+    {
+      "epoch": 2.7177968611193366,
+      "grad_norm": 1.4064515132654336,
+      "learning_rate": 4.5792828912624154e-07,
+      "loss": 0.5953,
+      "step": 4589
+    },
+    {
+      "epoch": 2.7183891027539238,
+      "grad_norm": 2.5112416819185404,
+      "learning_rate": 4.5601697400312175e-07,
+      "loss": 0.609,
+      "step": 4590
+    },
+    {
+      "epoch": 2.7189813443885105,
+      "grad_norm": 1.484009351861582,
+      "learning_rate": 4.541095629057435e-07,
+      "loss": 0.6504,
+      "step": 4591
+    },
+    {
+      "epoch": 2.719573586023097,
+      "grad_norm": 1.4442625515083019,
+      "learning_rate": 4.5220605661434605e-07,
+      "loss": 0.5921,
+      "step": 4592
+    },
+    {
+      "epoch": 2.7201658276576843,
+      "grad_norm": 1.2531427351279538,
+      "learning_rate": 4.503064559075687e-07,
+      "loss": 0.622,
+      "step": 4593
+    },
+    {
+      "epoch": 2.7207580692922715,
+      "grad_norm": 1.772483151004884,
+      "learning_rate": 4.4841076156245665e-07,
+      "loss": 0.5807,
+      "step": 4594
+    },
+    {
+      "epoch": 2.721350310926858,
+      "grad_norm": 1.01520707328506,
+      "learning_rate": 4.465189743544551e-07,
+      "loss": 0.6132,
+      "step": 4595
+    },
+    {
+      "epoch": 2.721942552561445,
+      "grad_norm": 1.8850344959256942,
+      "learning_rate": 4.4463109505741177e-07,
+      "loss": 0.62,
+      "step": 4596
+    },
+    {
+      "epoch": 2.722534794196032,
+      "grad_norm": 1.742313192790708,
+      "learning_rate": 4.427471244435733e-07,
+      "loss": 0.6011,
+      "step": 4597
+    },
+    {
+      "epoch": 2.7231270358306188,
+      "grad_norm": 1.4656681253439428,
+      "learning_rate": 4.408670632835932e-07,
+      "loss": 0.5727,
+      "step": 4598
+    },
+    {
+      "epoch": 2.723719277465206,
+      "grad_norm": 1.7292315439029244,
+      "learning_rate": 4.389909123465186e-07,
+      "loss": 0.5966,
+      "step": 4599
+    },
+    {
+      "epoch": 2.7243115190997926,
+      "grad_norm": 1.6450950995513567,
+      "learning_rate": 4.3711867239980335e-07,
+      "loss": 0.6282,
+      "step": 4600
+    },
+    {
+      "epoch": 2.72490376073438,
+      "grad_norm": 1.5986011485085827,
+      "learning_rate": 4.3525034420929815e-07,
+      "loss": 0.5897,
+      "step": 4601
+    },
+    {
+      "epoch": 2.7254960023689665,
+      "grad_norm": 1.668306756549204,
+      "learning_rate": 4.3338592853925277e-07,
+      "loss": 0.6291,
+      "step": 4602
+    },
+    {
+      "epoch": 2.726088244003553,
+      "grad_norm": 1.3048391452095505,
+      "learning_rate": 4.315254261523194e-07,
+      "loss": 0.6368,
+      "step": 4603
+    },
+    {
+      "epoch": 2.7266804856381404,
+      "grad_norm": 1.4097561245163255,
+      "learning_rate": 4.296688378095493e-07,
+      "loss": 0.659,
+      "step": 4604
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 4.424354272711558,
+      "learning_rate": 4.278161642703904e-07,
+      "loss": 0.6217,
+      "step": 4605
+    },
+    {
+      "epoch": 2.7278649689073142,
+      "grad_norm": 1.5665435392205018,
+      "learning_rate": 4.2596740629268997e-07,
+      "loss": 0.6582,
+      "step": 4606
+    },
+    {
+      "epoch": 2.728457210541901,
+      "grad_norm": 1.2828718477664653,
+      "learning_rate": 4.2412256463269295e-07,
+      "loss": 0.6032,
+      "step": 4607
+    },
+    {
+      "epoch": 2.729049452176488,
+      "grad_norm": 1.6646289965329508,
+      "learning_rate": 4.222816400450458e-07,
+      "loss": 0.6442,
+      "step": 4608
+    },
+    {
+      "epoch": 2.729641693811075,
+      "grad_norm": 1.2239860350713117,
+      "learning_rate": 4.204446332827894e-07,
+      "loss": 0.5899,
+      "step": 4609
+    },
+    {
+      "epoch": 2.730233935445662,
+      "grad_norm": 1.1809420431718864,
+      "learning_rate": 4.186115450973616e-07,
+      "loss": 0.5778,
+      "step": 4610
+    },
+    {
+      "epoch": 2.7308261770802487,
+      "grad_norm": 3.218531272958351,
+      "learning_rate": 4.1678237623859917e-07,
+      "loss": 0.604,
+      "step": 4611
+    },
+    {
+      "epoch": 2.731418418714836,
+      "grad_norm": 4.9639832700100195,
+      "learning_rate": 4.1495712745473595e-07,
+      "loss": 0.5754,
+      "step": 4612
+    },
+    {
+      "epoch": 2.7320106603494225,
+      "grad_norm": 1.6291943122524715,
+      "learning_rate": 4.1313579949240123e-07,
+      "loss": 0.6341,
+      "step": 4613
+    },
+    {
+      "epoch": 2.7326029019840092,
+      "grad_norm": 1.2955725976386623,
+      "learning_rate": 4.1131839309661803e-07,
+      "loss": 0.6163,
+      "step": 4614
+    },
+    {
+      "epoch": 2.7331951436185964,
+      "grad_norm": 1.6474895593726941,
+      "learning_rate": 4.0950490901081055e-07,
+      "loss": 0.5947,
+      "step": 4615
+    },
+    {
+      "epoch": 2.7337873852531835,
+      "grad_norm": 1.1944008670704511,
+      "learning_rate": 4.076953479767964e-07,
+      "loss": 0.6119,
+      "step": 4616
+    },
+    {
+      "epoch": 2.7343796268877703,
+      "grad_norm": 1.8527741398776432,
+      "learning_rate": 4.0588971073478477e-07,
+      "loss": 0.597,
+      "step": 4617
+    },
+    {
+      "epoch": 2.734971868522357,
+      "grad_norm": 1.6717926231193392,
+      "learning_rate": 4.040879980233836e-07,
+      "loss": 0.5978,
+      "step": 4618
+    },
+    {
+      "epoch": 2.735564110156944,
+      "grad_norm": 1.351512519792787,
+      "learning_rate": 4.022902105795956e-07,
+      "loss": 0.5928,
+      "step": 4619
+    },
+    {
+      "epoch": 2.736156351791531,
+      "grad_norm": 2.021922048948864,
+      "learning_rate": 4.004963491388203e-07,
+      "loss": 0.5927,
+      "step": 4620
+    },
+    {
+      "epoch": 2.736748593426118,
+      "grad_norm": 2.259608972990153,
+      "learning_rate": 3.987064144348407e-07,
+      "loss": 0.5992,
+      "step": 4621
+    },
+    {
+      "epoch": 2.7373408350607047,
+      "grad_norm": 2.5222440416768124,
+      "learning_rate": 3.969204071998445e-07,
+      "loss": 0.6311,
+      "step": 4622
+    },
+    {
+      "epoch": 2.737933076695292,
+      "grad_norm": 2.0706914867727875,
+      "learning_rate": 3.951383281644106e-07,
+      "loss": 0.6518,
+      "step": 4623
+    },
+    {
+      "epoch": 2.7385253183298786,
+      "grad_norm": 1.0448933909161093,
+      "learning_rate": 3.93360178057508e-07,
+      "loss": 0.6289,
+      "step": 4624
+    },
+    {
+      "epoch": 2.7391175599644653,
+      "grad_norm": 1.7416780397549962,
+      "learning_rate": 3.9158595760649954e-07,
+      "loss": 0.6319,
+      "step": 4625
+    },
+    {
+      "epoch": 2.7397098015990524,
+      "grad_norm": 7.362023134660734,
+      "learning_rate": 3.8981566753714116e-07,
+      "loss": 0.6252,
+      "step": 4626
+    },
+    {
+      "epoch": 2.7403020432336396,
+      "grad_norm": 1.4619722411701648,
+      "learning_rate": 3.8804930857358256e-07,
+      "loss": 0.5815,
+      "step": 4627
+    },
+    {
+      "epoch": 2.7408942848682263,
+      "grad_norm": 2.5074140857959697,
+      "learning_rate": 3.8628688143836244e-07,
+      "loss": 0.656,
+      "step": 4628
+    },
+    {
+      "epoch": 2.741486526502813,
+      "grad_norm": 1.7710644494678003,
+      "learning_rate": 3.845283868524119e-07,
+      "loss": 0.6125,
+      "step": 4629
+    },
+    {
+      "epoch": 2.7420787681374,
+      "grad_norm": 1.302185304449621,
+      "learning_rate": 3.827738255350555e-07,
+      "loss": 0.6145,
+      "step": 4630
+    },
+    {
+      "epoch": 2.742671009771987,
+      "grad_norm": 1.488799773221385,
+      "learning_rate": 3.810231982040091e-07,
+      "loss": 0.6006,
+      "step": 4631
+    },
+    {
+      "epoch": 2.743263251406574,
+      "grad_norm": 1.1890826196658626,
+      "learning_rate": 3.792765055753755e-07,
+      "loss": 0.5658,
+      "step": 4632
+    },
+    {
+      "epoch": 2.7438554930411607,
+      "grad_norm": 1.6943845095513352,
+      "learning_rate": 3.775337483636488e-07,
+      "loss": 0.6312,
+      "step": 4633
+    },
+    {
+      "epoch": 2.744447734675748,
+      "grad_norm": 2.162351020574408,
+      "learning_rate": 3.757949272817174e-07,
+      "loss": 0.6157,
+      "step": 4634
+    },
+    {
+      "epoch": 2.7450399763103346,
+      "grad_norm": 1.8172507844780956,
+      "learning_rate": 3.7406004304085584e-07,
+      "loss": 0.5986,
+      "step": 4635
+    },
+    {
+      "epoch": 2.7456322179449213,
+      "grad_norm": 2.857384440493848,
+      "learning_rate": 3.723290963507309e-07,
+      "loss": 0.6173,
+      "step": 4636
+    },
+    {
+      "epoch": 2.7462244595795084,
+      "grad_norm": 2.336254189546982,
+      "learning_rate": 3.706020879193939e-07,
+      "loss": 0.6436,
+      "step": 4637
+    },
+    {
+      "epoch": 2.7468167012140956,
+      "grad_norm": 2.576256451190125,
+      "learning_rate": 3.688790184532909e-07,
+      "loss": 0.61,
+      "step": 4638
+    },
+    {
+      "epoch": 2.7474089428486823,
+      "grad_norm": 2.029285692911393,
+      "learning_rate": 3.671598886572525e-07,
+      "loss": 0.6187,
+      "step": 4639
+    },
+    {
+      "epoch": 2.748001184483269,
+      "grad_norm": 1.3899689164399947,
+      "learning_rate": 3.654446992345018e-07,
+      "loss": 0.6091,
+      "step": 4640
+    },
+    {
+      "epoch": 2.748593426117856,
+      "grad_norm": 1.3664074951874416,
+      "learning_rate": 3.6373345088664525e-07,
+      "loss": 0.5813,
+      "step": 4641
+    },
+    {
+      "epoch": 2.749185667752443,
+      "grad_norm": 2.891864884426199,
+      "learning_rate": 3.620261443136819e-07,
+      "loss": 0.5623,
+      "step": 4642
+    },
+    {
+      "epoch": 2.74977790938703,
+      "grad_norm": 2.0669641721163403,
+      "learning_rate": 3.6032278021399415e-07,
+      "loss": 0.5992,
+      "step": 4643
+    },
+    {
+      "epoch": 2.7503701510216167,
+      "grad_norm": 1.6857686046412492,
+      "learning_rate": 3.5862335928435465e-07,
+      "loss": 0.6651,
+      "step": 4644
+    },
+    {
+      "epoch": 2.750962392656204,
+      "grad_norm": 1.6062047172916332,
+      "learning_rate": 3.569278822199218e-07,
+      "loss": 0.5714,
+      "step": 4645
+    },
+    {
+      "epoch": 2.7515546342907906,
+      "grad_norm": 1.4375077948074273,
+      "learning_rate": 3.5523634971424194e-07,
+      "loss": 0.624,
+      "step": 4646
+    },
+    {
+      "epoch": 2.7521468759253773,
+      "grad_norm": 2.233044908101775,
+      "learning_rate": 3.5354876245924596e-07,
+      "loss": 0.5747,
+      "step": 4647
+    },
+    {
+      "epoch": 2.7527391175599645,
+      "grad_norm": 6.883080250715156,
+      "learning_rate": 3.5186512114525283e-07,
+      "loss": 0.5813,
+      "step": 4648
+    },
+    {
+      "epoch": 2.7533313591945516,
+      "grad_norm": 2.2999746530713394,
+      "learning_rate": 3.50185426460965e-07,
+      "loss": 0.6157,
+      "step": 4649
+    },
+    {
+      "epoch": 2.7539236008291383,
+      "grad_norm": 1.8968933961570913,
+      "learning_rate": 3.485096790934739e-07,
+      "loss": 0.6322,
+      "step": 4650
+    },
+    {
+      "epoch": 2.754515842463725,
+      "grad_norm": 1.4387652445210517,
+      "learning_rate": 3.4683787972825345e-07,
+      "loss": 0.6494,
+      "step": 4651
+    },
+    {
+      "epoch": 2.755108084098312,
+      "grad_norm": 1.7110296382242969,
+      "learning_rate": 3.451700290491633e-07,
+      "loss": 0.5789,
+      "step": 4652
+    },
+    {
+      "epoch": 2.755700325732899,
+      "grad_norm": 1.781421507136215,
+      "learning_rate": 3.4350612773844996e-07,
+      "loss": 0.5754,
+      "step": 4653
+    },
+    {
+      "epoch": 2.756292567367486,
+      "grad_norm": 1.6547622885370215,
+      "learning_rate": 3.4184617647674e-07,
+      "loss": 0.5839,
+      "step": 4654
+    },
+    {
+      "epoch": 2.756884809002073,
+      "grad_norm": 1.5141595029370973,
+      "learning_rate": 3.40190175943047e-07,
+      "loss": 0.6275,
+      "step": 4655
+    },
+    {
+      "epoch": 2.75747705063666,
+      "grad_norm": 1.4347172437913116,
+      "learning_rate": 3.3853812681477136e-07,
+      "loss": 0.6384,
+      "step": 4656
+    },
+    {
+      "epoch": 2.7580692922712466,
+      "grad_norm": 2.134280092349403,
+      "learning_rate": 3.368900297676925e-07,
+      "loss": 0.6151,
+      "step": 4657
+    },
+    {
+      "epoch": 2.7586615339058334,
+      "grad_norm": 1.6195000996671558,
+      "learning_rate": 3.3524588547597327e-07,
+      "loss": 0.5915,
+      "step": 4658
+    },
+    {
+      "epoch": 2.7592537755404205,
+      "grad_norm": 1.4226797326345686,
+      "learning_rate": 3.336056946121613e-07,
+      "loss": 0.6438,
+      "step": 4659
+    },
+    {
+      "epoch": 2.7598460171750077,
+      "grad_norm": 2.076976223041084,
+      "learning_rate": 3.3196945784718993e-07,
+      "loss": 0.6008,
+      "step": 4660
+    },
+    {
+      "epoch": 2.7604382588095944,
+      "grad_norm": 1.4596510371769844,
+      "learning_rate": 3.303371758503693e-07,
+      "loss": 0.6462,
+      "step": 4661
+    },
+    {
+      "epoch": 2.761030500444181,
+      "grad_norm": 1.596455551281381,
+      "learning_rate": 3.287088492893942e-07,
+      "loss": 0.5848,
+      "step": 4662
+    },
+    {
+      "epoch": 2.7616227420787682,
+      "grad_norm": 1.3584900013265102,
+      "learning_rate": 3.270844788303429e-07,
+      "loss": 0.6033,
+      "step": 4663
+    },
+    {
+      "epoch": 2.762214983713355,
+      "grad_norm": 1.4668378316726418,
+      "learning_rate": 3.25464065137675e-07,
+      "loss": 0.5542,
+      "step": 4664
+    },
+    {
+      "epoch": 2.762807225347942,
+      "grad_norm": 2.1567477871091936,
+      "learning_rate": 3.2384760887423036e-07,
+      "loss": 0.6195,
+      "step": 4665
+    },
+    {
+      "epoch": 2.763399466982529,
+      "grad_norm": 1.7423110133051698,
+      "learning_rate": 3.2223511070122893e-07,
+      "loss": 0.6229,
+      "step": 4666
+    },
+    {
+      "epoch": 2.763991708617116,
+      "grad_norm": 5.691959866524479,
+      "learning_rate": 3.2062657127827413e-07,
+      "loss": 0.6046,
+      "step": 4667
+    },
+    {
+      "epoch": 2.7645839502517027,
+      "grad_norm": 2.0605725048101124,
+      "learning_rate": 3.190219912633519e-07,
+      "loss": 0.5933,
+      "step": 4668
+    },
+    {
+      "epoch": 2.7651761918862894,
+      "grad_norm": 1.6855632316948044,
+      "learning_rate": 3.1742137131281937e-07,
+      "loss": 0.5716,
+      "step": 4669
+    },
+    {
+      "epoch": 2.7657684335208765,
+      "grad_norm": 1.4437261940020687,
+      "learning_rate": 3.158247120814251e-07,
+      "loss": 0.6427,
+      "step": 4670
+    },
+    {
+      "epoch": 2.7663606751554637,
+      "grad_norm": 1.9620940489950571,
+      "learning_rate": 3.142320142222899e-07,
+      "loss": 0.6286,
+      "step": 4671
+    },
+    {
+      "epoch": 2.7669529167900504,
+      "grad_norm": 1.5525721497968263,
+      "learning_rate": 3.1264327838692153e-07,
+      "loss": 0.6121,
+      "step": 4672
+    },
+    {
+      "epoch": 2.767545158424637,
+      "grad_norm": 1.864998629614895,
+      "learning_rate": 3.1105850522519574e-07,
+      "loss": 0.6336,
+      "step": 4673
+    },
+    {
+      "epoch": 2.7681374000592243,
+      "grad_norm": 2.834526941324686,
+      "learning_rate": 3.094776953853762e-07,
+      "loss": 0.5941,
+      "step": 4674
+    },
+    {
+      "epoch": 2.768729641693811,
+      "grad_norm": 2.1256789045491966,
+      "learning_rate": 3.079008495141056e-07,
+      "loss": 0.6248,
+      "step": 4675
+    },
+    {
+      "epoch": 2.769321883328398,
+      "grad_norm": 1.8764504253445144,
+      "learning_rate": 3.063279682564002e-07,
+      "loss": 0.6022,
+      "step": 4676
+    },
+    {
+      "epoch": 2.769914124962985,
+      "grad_norm": 2.287234147163577,
+      "learning_rate": 3.047590522556565e-07,
+      "loss": 0.6267,
+      "step": 4677
+    },
+    {
+      "epoch": 2.770506366597572,
+      "grad_norm": 2.975356179185165,
+      "learning_rate": 3.0319410215365e-07,
+      "loss": 0.6251,
+      "step": 4678
+    },
+    {
+      "epoch": 2.7710986082321587,
+      "grad_norm": 2.5144656587013507,
+      "learning_rate": 3.0163311859053524e-07,
+      "loss": 0.585,
+      "step": 4679
+    },
+    {
+      "epoch": 2.7716908498667454,
+      "grad_norm": 1.9765530263280324,
+      "learning_rate": 3.0007610220483927e-07,
+      "loss": 0.5958,
+      "step": 4680
+    },
+    {
+      "epoch": 2.7722830915013326,
+      "grad_norm": 1.8855944082384668,
+      "learning_rate": 2.9852305363347044e-07,
+      "loss": 0.6075,
+      "step": 4681
+    },
+    {
+      "epoch": 2.7728753331359197,
+      "grad_norm": 1.4481580120122306,
+      "learning_rate": 2.969739735117128e-07,
+      "loss": 0.631,
+      "step": 4682
+    },
+    {
+      "epoch": 2.7734675747705064,
+      "grad_norm": 1.4994595976642986,
+      "learning_rate": 2.954288624732293e-07,
+      "loss": 0.6311,
+      "step": 4683
+    },
+    {
+      "epoch": 2.774059816405093,
+      "grad_norm": 2.9375587558348415,
+      "learning_rate": 2.9388772115005457e-07,
+      "loss": 0.6206,
+      "step": 4684
+    },
+    {
+      "epoch": 2.7746520580396803,
+      "grad_norm": 1.3813361341030252,
+      "learning_rate": 2.9235055017260205e-07,
+      "loss": 0.6059,
+      "step": 4685
+    },
+    {
+      "epoch": 2.775244299674267,
+      "grad_norm": 1.2073925168982438,
+      "learning_rate": 2.9081735016966205e-07,
+      "loss": 0.6153,
+      "step": 4686
+    },
+    {
+      "epoch": 2.775836541308854,
+      "grad_norm": 1.6373735275001204,
+      "learning_rate": 2.892881217684007e-07,
+      "loss": 0.6376,
+      "step": 4687
+    },
+    {
+      "epoch": 2.776428782943441,
+      "grad_norm": 1.3737783710680564,
+      "learning_rate": 2.877628655943576e-07,
+      "loss": 0.5936,
+      "step": 4688
+    },
+    {
+      "epoch": 2.777021024578028,
+      "grad_norm": 1.4776319579183617,
+      "learning_rate": 2.8624158227144703e-07,
+      "loss": 0.6292,
+      "step": 4689
+    },
+    {
+      "epoch": 2.7776132662126147,
+      "grad_norm": 1.4187427316372916,
+      "learning_rate": 2.847242724219612e-07,
+      "loss": 0.631,
+      "step": 4690
+    },
+    {
+      "epoch": 2.7782055078472014,
+      "grad_norm": 1.5610769267747608,
+      "learning_rate": 2.8321093666656253e-07,
+      "loss": 0.6175,
+      "step": 4691
+    },
+    {
+      "epoch": 2.7787977494817886,
+      "grad_norm": 1.3150995164435961,
+      "learning_rate": 2.8170157562429466e-07,
+      "loss": 0.6882,
+      "step": 4692
+    },
+    {
+      "epoch": 2.7793899911163757,
+      "grad_norm": 1.3910028681163722,
+      "learning_rate": 2.801961899125671e-07,
+      "loss": 0.6123,
+      "step": 4693
+    },
+    {
+      "epoch": 2.7799822327509625,
+      "grad_norm": 1.8024828509396476,
+      "learning_rate": 2.7869478014716953e-07,
+      "loss": 0.6624,
+      "step": 4694
+    },
+    {
+      "epoch": 2.780574474385549,
+      "grad_norm": 1.2492833932312706,
+      "learning_rate": 2.7719734694226065e-07,
+      "loss": 0.6252,
+      "step": 4695
+    },
+    {
+      "epoch": 2.7811667160201363,
+      "grad_norm": 1.8896802955465037,
+      "learning_rate": 2.757038909103793e-07,
+      "loss": 0.5868,
+      "step": 4696
+    },
+    {
+      "epoch": 2.781758957654723,
+      "grad_norm": 2.317063717995178,
+      "learning_rate": 2.74214412662428e-07,
+      "loss": 0.6063,
+      "step": 4697
+    },
+    {
+      "epoch": 2.78235119928931,
+      "grad_norm": 1.2618887822595275,
+      "learning_rate": 2.7272891280769044e-07,
+      "loss": 0.6231,
+      "step": 4698
+    },
+    {
+      "epoch": 2.782943440923897,
+      "grad_norm": 2.519622445595849,
+      "learning_rate": 2.7124739195381724e-07,
+      "loss": 0.6046,
+      "step": 4699
+    },
+    {
+      "epoch": 2.783535682558484,
+      "grad_norm": 1.3397833913049007,
+      "learning_rate": 2.697698507068358e-07,
+      "loss": 0.6201,
+      "step": 4700
+    },
+    {
+      "epoch": 2.7841279241930708,
+      "grad_norm": 1.6268323399434212,
+      "learning_rate": 2.682962896711427e-07,
+      "loss": 0.5909,
+      "step": 4701
+    },
+    {
+      "epoch": 2.7847201658276575,
+      "grad_norm": 1.6990969743760185,
+      "learning_rate": 2.6682670944950804e-07,
+      "loss": 0.6215,
+      "step": 4702
+    },
+    {
+      "epoch": 2.7853124074622446,
+      "grad_norm": 1.8833596532734562,
+      "learning_rate": 2.653611106430698e-07,
+      "loss": 0.6729,
+      "step": 4703
+    },
+    {
+      "epoch": 2.7859046490968318,
+      "grad_norm": 1.6243542801393849,
+      "learning_rate": 2.638994938513451e-07,
+      "loss": 0.5851,
+      "step": 4704
+    },
+    {
+      "epoch": 2.7864968907314185,
+      "grad_norm": 1.2192739780041595,
+      "learning_rate": 2.624418596722134e-07,
+      "loss": 0.6167,
+      "step": 4705
+    },
+    {
+      "epoch": 2.787089132366005,
+      "grad_norm": 1.8467947724763827,
+      "learning_rate": 2.609882087019311e-07,
+      "loss": 0.5764,
+      "step": 4706
+    },
+    {
+      "epoch": 2.7876813740005923,
+      "grad_norm": 1.4594864341656142,
+      "learning_rate": 2.595385415351215e-07,
+      "loss": 0.612,
+      "step": 4707
+    },
+    {
+      "epoch": 2.788273615635179,
+      "grad_norm": 1.854617047861561,
+      "learning_rate": 2.580928587647824e-07,
+      "loss": 0.6462,
+      "step": 4708
+    },
+    {
+      "epoch": 2.788865857269766,
+      "grad_norm": 1.4487640953791494,
+      "learning_rate": 2.566511609822775e-07,
+      "loss": 0.6045,
+      "step": 4709
+    },
+    {
+      "epoch": 2.789458098904353,
+      "grad_norm": 5.36212138647543,
+      "learning_rate": 2.5521344877734165e-07,
+      "loss": 0.6486,
+      "step": 4710
+    },
+    {
+      "epoch": 2.79005034053894,
+      "grad_norm": 2.5695670114046285,
+      "learning_rate": 2.5377972273808115e-07,
+      "loss": 0.6292,
+      "step": 4711
+    },
+    {
+      "epoch": 2.790642582173527,
+      "grad_norm": 1.674042925609699,
+      "learning_rate": 2.523499834509724e-07,
+      "loss": 0.6021,
+      "step": 4712
+    },
+    {
+      "epoch": 2.7912348238081135,
+      "grad_norm": 1.7422315504699424,
+      "learning_rate": 2.5092423150085643e-07,
+      "loss": 0.6147,
+      "step": 4713
+    },
+    {
+      "epoch": 2.7918270654427007,
+      "grad_norm": 1.518935615818262,
+      "learning_rate": 2.495024674709468e-07,
+      "loss": 0.6175,
+      "step": 4714
+    },
+    {
+      "epoch": 2.792419307077288,
+      "grad_norm": 3.9189187326117443,
+      "learning_rate": 2.480846919428237e-07,
+      "loss": 0.6206,
+      "step": 4715
+    },
+    {
+      "epoch": 2.7930115487118745,
+      "grad_norm": 1.5020841924960988,
+      "learning_rate": 2.4667090549644e-07,
+      "loss": 0.6688,
+      "step": 4716
+    },
+    {
+      "epoch": 2.7936037903464612,
+      "grad_norm": 2.035594629956855,
+      "learning_rate": 2.45261108710112e-07,
+      "loss": 0.556,
+      "step": 4717
+    },
+    {
+      "epoch": 2.7941960319810484,
+      "grad_norm": 1.7405526485195462,
+      "learning_rate": 2.43855302160525e-07,
+      "loss": 0.6238,
+      "step": 4718
+    },
+    {
+      "epoch": 2.794788273615635,
+      "grad_norm": 4.375187374245842,
+      "learning_rate": 2.424534864227346e-07,
+      "loss": 0.6078,
+      "step": 4719
+    },
+    {
+      "epoch": 2.7953805152502222,
+      "grad_norm": 12.710028326671614,
+      "learning_rate": 2.4105566207016207e-07,
+      "loss": 0.6527,
+      "step": 4720
+    },
+    {
+      "epoch": 2.795972756884809,
+      "grad_norm": 1.3120633605538479,
+      "learning_rate": 2.396618296745956e-07,
+      "loss": 0.6098,
+      "step": 4721
+    },
+    {
+      "epoch": 2.796564998519396,
+      "grad_norm": 1.6308815559170025,
+      "learning_rate": 2.3827198980619025e-07,
+      "loss": 0.6458,
+      "step": 4722
+    },
+    {
+      "epoch": 2.797157240153983,
+      "grad_norm": 1.3307922640687166,
+      "learning_rate": 2.3688614303347012e-07,
+      "loss": 0.59,
+      "step": 4723
+    },
+    {
+      "epoch": 2.7977494817885695,
+      "grad_norm": 1.3953095149862986,
+      "learning_rate": 2.3550428992332508e-07,
+      "loss": 0.5971,
+      "step": 4724
+    },
+    {
+      "epoch": 2.7983417234231567,
+      "grad_norm": 1.241408618714826,
+      "learning_rate": 2.341264310410085e-07,
+      "loss": 0.6275,
+      "step": 4725
+    },
+    {
+      "epoch": 2.798933965057744,
+      "grad_norm": 2.7908340168986254,
+      "learning_rate": 2.327525669501418e-07,
+      "loss": 0.5977,
+      "step": 4726
+    },
+    {
+      "epoch": 2.7995262066923305,
+      "grad_norm": 1.46971406756966,
+      "learning_rate": 2.3138269821271654e-07,
+      "loss": 0.6049,
+      "step": 4727
+    },
+    {
+      "epoch": 2.8001184483269173,
+      "grad_norm": 1.23275893833151,
+      "learning_rate": 2.3001682538908333e-07,
+      "loss": 0.5917,
+      "step": 4728
+    },
+    {
+      "epoch": 2.8007106899615044,
+      "grad_norm": 2.1840457881629582,
+      "learning_rate": 2.286549490379597e-07,
+      "loss": 0.5872,
+      "step": 4729
+    },
+    {
+      "epoch": 2.801302931596091,
+      "grad_norm": 1.9787660250351953,
+      "learning_rate": 2.2729706971643117e-07,
+      "loss": 0.6328,
+      "step": 4730
+    },
+    {
+      "epoch": 2.8018951732306783,
+      "grad_norm": 1.6062910758904774,
+      "learning_rate": 2.2594318797994895e-07,
+      "loss": 0.5954,
+      "step": 4731
+    },
+    {
+      "epoch": 2.802487414865265,
+      "grad_norm": 1.7630976777586607,
+      "learning_rate": 2.245933043823234e-07,
+      "loss": 0.5963,
+      "step": 4732
+    },
+    {
+      "epoch": 2.803079656499852,
+      "grad_norm": 1.5256593573589792,
+      "learning_rate": 2.232474194757339e-07,
+      "loss": 0.6046,
+      "step": 4733
+    },
+    {
+      "epoch": 2.803671898134439,
+      "grad_norm": 1.811287510313242,
+      "learning_rate": 2.2190553381072234e-07,
+      "loss": 0.6255,
+      "step": 4734
+    },
+    {
+      "epoch": 2.8042641397690256,
+      "grad_norm": 1.6696578027302151,
+      "learning_rate": 2.2056764793619845e-07,
+      "loss": 0.6031,
+      "step": 4735
+    },
+    {
+      "epoch": 2.8048563814036127,
+      "grad_norm": 2.2390362799933348,
+      "learning_rate": 2.1923376239942895e-07,
+      "loss": 0.6192,
+      "step": 4736
+    },
+    {
+      "epoch": 2.8054486230382,
+      "grad_norm": 1.9676446261540672,
+      "learning_rate": 2.179038777460507e-07,
+      "loss": 0.5974,
+      "step": 4737
+    },
+    {
+      "epoch": 2.8060408646727866,
+      "grad_norm": 1.3850897301178713,
+      "learning_rate": 2.1657799452005856e-07,
+      "loss": 0.58,
+      "step": 4738
+    },
+    {
+      "epoch": 2.8066331063073733,
+      "grad_norm": 1.5591433252070857,
+      "learning_rate": 2.1525611326381756e-07,
+      "loss": 0.6151,
+      "step": 4739
+    },
+    {
+      "epoch": 2.8072253479419604,
+      "grad_norm": 1.190679482654428,
+      "learning_rate": 2.139382345180474e-07,
+      "loss": 0.6089,
+      "step": 4740
+    },
+    {
+      "epoch": 2.807817589576547,
+      "grad_norm": 1.4443287017918227,
+      "learning_rate": 2.1262435882183685e-07,
+      "loss": 0.649,
+      "step": 4741
+    },
+    {
+      "epoch": 2.8084098312111343,
+      "grad_norm": 2.138041941041724,
+      "learning_rate": 2.1131448671263378e-07,
+      "loss": 0.602,
+      "step": 4742
+    },
+    {
+      "epoch": 2.809002072845721,
+      "grad_norm": 2.021125849306582,
+      "learning_rate": 2.1000861872625066e-07,
+      "loss": 0.6106,
+      "step": 4743
+    },
+    {
+      "epoch": 2.809594314480308,
+      "grad_norm": 1.1718395287718038,
+      "learning_rate": 2.0870675539686024e-07,
+      "loss": 0.6226,
+      "step": 4744
+    },
+    {
+      "epoch": 2.810186556114895,
+      "grad_norm": 2.4236065030797453,
+      "learning_rate": 2.0740889725699654e-07,
+      "loss": 0.5738,
+      "step": 4745
+    },
+    {
+      "epoch": 2.8107787977494816,
+      "grad_norm": 2.134568247280834,
+      "learning_rate": 2.0611504483756038e-07,
+      "loss": 0.6524,
+      "step": 4746
+    },
+    {
+      "epoch": 2.8113710393840687,
+      "grad_norm": 1.2716048058881149,
+      "learning_rate": 2.0482519866780516e-07,
+      "loss": 0.5964,
+      "step": 4747
+    },
+    {
+      "epoch": 2.8119632810186554,
+      "grad_norm": 1.2026027684878298,
+      "learning_rate": 2.0353935927535428e-07,
+      "loss": 0.5937,
+      "step": 4748
+    },
+    {
+      "epoch": 2.8125555226532426,
+      "grad_norm": 1.7177800039620705,
+      "learning_rate": 2.0225752718618707e-07,
+      "loss": 0.6003,
+      "step": 4749
+    },
+    {
+      "epoch": 2.8131477642878293,
+      "grad_norm": 2.3836143414025974,
+      "learning_rate": 2.009797029246452e-07,
+      "loss": 0.568,
+      "step": 4750
+    },
+    {
+      "epoch": 2.8137400059224165,
+      "grad_norm": 1.5700728206598682,
+      "learning_rate": 1.997058870134294e-07,
+      "loss": 0.5908,
+      "step": 4751
+    },
+    {
+      "epoch": 2.814332247557003,
+      "grad_norm": 1.4091090434483953,
+      "learning_rate": 1.9843607997360403e-07,
+      "loss": 0.5847,
+      "step": 4752
+    },
+    {
+      "epoch": 2.8149244891915903,
+      "grad_norm": 3.983754518350977,
+      "learning_rate": 1.9717028232458912e-07,
+      "loss": 0.6242,
+      "step": 4753
+    },
+    {
+      "epoch": 2.815516730826177,
+      "grad_norm": 3.552480776764616,
+      "learning_rate": 1.959084945841705e-07,
+      "loss": 0.6299,
+      "step": 4754
+    },
+    {
+      "epoch": 2.816108972460764,
+      "grad_norm": 1.1920497610991834,
+      "learning_rate": 1.9465071726848638e-07,
+      "loss": 0.6314,
+      "step": 4755
+    },
+    {
+      "epoch": 2.816701214095351,
+      "grad_norm": 1.0342228137761196,
+      "learning_rate": 1.9339695089204192e-07,
+      "loss": 0.5991,
+      "step": 4756
+    },
+    {
+      "epoch": 2.8172934557299376,
+      "grad_norm": 1.5464771974705371,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 0.5717,
+      "step": 4757
+    },
+    {
+      "epoch": 2.8178856973645248,
+      "grad_norm": 9.400365344050027,
+      "learning_rate": 1.9090145300666885e-07,
+      "loss": 0.5665,
+      "step": 4758
+    },
+    {
+      "epoch": 2.8184779389991115,
+      "grad_norm": 1.451507907352805,
+      "learning_rate": 1.8965972251854038e-07,
+      "loss": 0.6476,
+      "step": 4759
+    },
+    {
+      "epoch": 2.8190701806336986,
+      "grad_norm": 1.467526846495732,
+      "learning_rate": 1.884220050112462e-07,
+      "loss": 0.5874,
+      "step": 4760
+    },
+    {
+      "epoch": 2.8196624222682853,
+      "grad_norm": 1.2322692339357366,
+      "learning_rate": 1.8718830099108464e-07,
+      "loss": 0.5639,
+      "step": 4761
+    },
+    {
+      "epoch": 2.8202546639028725,
+      "grad_norm": 1.9431671531415993,
+      "learning_rate": 1.8595861096270874e-07,
+      "loss": 0.6446,
+      "step": 4762
+    },
+    {
+      "epoch": 2.820846905537459,
+      "grad_norm": 1.6843027054345925,
+      "learning_rate": 1.8473293542913163e-07,
+      "loss": 0.6301,
+      "step": 4763
+    },
+    {
+      "epoch": 2.8214391471720464,
+      "grad_norm": 2.487244252300873,
+      "learning_rate": 1.8351127489172227e-07,
+      "loss": 0.646,
+      "step": 4764
+    },
+    {
+      "epoch": 2.822031388806633,
+      "grad_norm": 3.497230980934624,
+      "learning_rate": 1.8229362985021092e-07,
+      "loss": 0.5818,
+      "step": 4765
+    },
+    {
+      "epoch": 2.82262363044122,
+      "grad_norm": 2.308374655735379,
+      "learning_rate": 1.8108000080267918e-07,
+      "loss": 0.6312,
+      "step": 4766
+    },
+    {
+      "epoch": 2.823215872075807,
+      "grad_norm": 2.0176260134625874,
+      "learning_rate": 1.7987038824557323e-07,
+      "loss": 0.5929,
+      "step": 4767
+    },
+    {
+      "epoch": 2.8238081137103936,
+      "grad_norm": 1.7213851889188727,
+      "learning_rate": 1.7866479267369062e-07,
+      "loss": 0.6419,
+      "step": 4768
+    },
+    {
+      "epoch": 2.824400355344981,
+      "grad_norm": 1.510546311903195,
+      "learning_rate": 1.7746321458018802e-07,
+      "loss": 0.5843,
+      "step": 4769
+    },
+    {
+      "epoch": 2.8249925969795675,
+      "grad_norm": 1.9187963373200847,
+      "learning_rate": 1.7626565445657883e-07,
+      "loss": 0.5998,
+      "step": 4770
+    },
+    {
+      "epoch": 2.8255848386141547,
+      "grad_norm": 1.3445654001439231,
+      "learning_rate": 1.750721127927324e-07,
+      "loss": 0.6266,
+      "step": 4771
+    },
+    {
+      "epoch": 2.8261770802487414,
+      "grad_norm": 1.4215769707003076,
+      "learning_rate": 1.7388259007687368e-07,
+      "loss": 0.6382,
+      "step": 4772
+    },
+    {
+      "epoch": 2.8267693218833285,
+      "grad_norm": 2.129246036856138,
+      "learning_rate": 1.7269708679558572e-07,
+      "loss": 0.6008,
+      "step": 4773
+    },
+    {
+      "epoch": 2.8273615635179152,
+      "grad_norm": 1.4627320973002913,
+      "learning_rate": 1.715156034338039e-07,
+      "loss": 0.6084,
+      "step": 4774
+    },
+    {
+      "epoch": 2.8279538051525024,
+      "grad_norm": 2.8153006747845875,
+      "learning_rate": 1.7033814047482388e-07,
+      "loss": 0.6042,
+      "step": 4775
+    },
+    {
+      "epoch": 2.828546046787089,
+      "grad_norm": 1.507076721738337,
+      "learning_rate": 1.691646984002937e-07,
+      "loss": 0.6466,
+      "step": 4776
+    },
+    {
+      "epoch": 2.8291382884216763,
+      "grad_norm": 1.2884117944639297,
+      "learning_rate": 1.6799527769021495e-07,
+      "loss": 0.6195,
+      "step": 4777
+    },
+    {
+      "epoch": 2.829730530056263,
+      "grad_norm": 1.6946887003219235,
+      "learning_rate": 1.6682987882294722e-07,
+      "loss": 0.6082,
+      "step": 4778
+    },
+    {
+      "epoch": 2.8303227716908497,
+      "grad_norm": 1.5567973673661333,
+      "learning_rate": 1.6566850227520693e-07,
+      "loss": 0.5863,
+      "step": 4779
+    },
+    {
+      "epoch": 2.830915013325437,
+      "grad_norm": 1.7210984837021315,
+      "learning_rate": 1.6451114852206073e-07,
+      "loss": 0.633,
+      "step": 4780
+    },
+    {
+      "epoch": 2.8315072549600235,
+      "grad_norm": 1.5582674489351216,
+      "learning_rate": 1.6335781803692884e-07,
+      "loss": 0.5535,
+      "step": 4781
+    },
+    {
+      "epoch": 2.8320994965946107,
+      "grad_norm": 3.604937316955796,
+      "learning_rate": 1.6220851129159164e-07,
+      "loss": 0.6094,
+      "step": 4782
+    },
+    {
+      "epoch": 2.8326917382291974,
+      "grad_norm": 1.468440659774134,
+      "learning_rate": 1.6106322875617974e-07,
+      "loss": 0.6098,
+      "step": 4783
+    },
+    {
+      "epoch": 2.8332839798637846,
+      "grad_norm": 1.5367592010242979,
+      "learning_rate": 1.5992197089917727e-07,
+      "loss": 0.5844,
+      "step": 4784
+    },
+    {
+      "epoch": 2.8338762214983713,
+      "grad_norm": 1.7789542552715796,
+      "learning_rate": 1.587847381874219e-07,
+      "loss": 0.6287,
+      "step": 4785
+    },
+    {
+      "epoch": 2.834468463132958,
+      "grad_norm": 5.853973890713451,
+      "learning_rate": 1.576515310861071e-07,
+      "loss": 0.5924,
+      "step": 4786
+    },
+    {
+      "epoch": 2.835060704767545,
+      "grad_norm": 1.7798910211182182,
+      "learning_rate": 1.565223500587798e-07,
+      "loss": 0.6681,
+      "step": 4787
+    },
+    {
+      "epoch": 2.8356529464021323,
+      "grad_norm": 2.8553443898913935,
+      "learning_rate": 1.55397195567335e-07,
+      "loss": 0.6537,
+      "step": 4788
+    },
+    {
+      "epoch": 2.836245188036719,
+      "grad_norm": 1.1857098480480426,
+      "learning_rate": 1.5427606807202676e-07,
+      "loss": 0.6129,
+      "step": 4789
+    },
+    {
+      "epoch": 2.8368374296713057,
+      "grad_norm": 1.5704580950283231,
+      "learning_rate": 1.5315896803145824e-07,
+      "loss": 0.6026,
+      "step": 4790
+    },
+    {
+      "epoch": 2.837429671305893,
+      "grad_norm": 4.451172736833375,
+      "learning_rate": 1.5204589590258722e-07,
+      "loss": 0.6233,
+      "step": 4791
+    },
+    {
+      "epoch": 2.8380219129404796,
+      "grad_norm": 1.6353470217624213,
+      "learning_rate": 1.5093685214072173e-07,
+      "loss": 0.6129,
+      "step": 4792
+    },
+    {
+      "epoch": 2.8386141545750667,
+      "grad_norm": 1.660643984933811,
+      "learning_rate": 1.4983183719952222e-07,
+      "loss": 0.6326,
+      "step": 4793
+    },
+    {
+      "epoch": 2.8392063962096534,
+      "grad_norm": 1.456440719487532,
+      "learning_rate": 1.4873085153100485e-07,
+      "loss": 0.6038,
+      "step": 4794
+    },
+    {
+      "epoch": 2.8397986378442406,
+      "grad_norm": 1.2676177701134983,
+      "learning_rate": 1.4763389558553164e-07,
+      "loss": 0.5914,
+      "step": 4795
+    },
+    {
+      "epoch": 2.8403908794788273,
+      "grad_norm": 1.4386662121643279,
+      "learning_rate": 1.4654096981182031e-07,
+      "loss": 0.617,
+      "step": 4796
+    },
+    {
+      "epoch": 2.840983121113414,
+      "grad_norm": 1.3829547994752223,
+      "learning_rate": 1.4545207465693877e-07,
+      "loss": 0.6172,
+      "step": 4797
+    },
+    {
+      "epoch": 2.841575362748001,
+      "grad_norm": 2.9258003797550542,
+      "learning_rate": 1.4436721056630853e-07,
+      "loss": 0.6146,
+      "step": 4798
+    },
+    {
+      "epoch": 2.8421676043825883,
+      "grad_norm": 3.4493369680899795,
+      "learning_rate": 1.432863779836968e-07,
+      "loss": 0.6168,
+      "step": 4799
+    },
+    {
+      "epoch": 2.842759846017175,
+      "grad_norm": 1.7932986975776593,
+      "learning_rate": 1.4220957735122663e-07,
+      "loss": 0.624,
+      "step": 4800
+    },
+    {
+      "epoch": 2.8433520876517617,
+      "grad_norm": 2.6126448363470867,
+      "learning_rate": 1.41136809109369e-07,
+      "loss": 0.5994,
+      "step": 4801
+    },
+    {
+      "epoch": 2.843944329286349,
+      "grad_norm": 1.9157604337658336,
+      "learning_rate": 1.400680736969484e-07,
+      "loss": 0.6276,
+      "step": 4802
+    },
+    {
+      "epoch": 2.8445365709209356,
+      "grad_norm": 1.7072043505201628,
+      "learning_rate": 1.390033715511363e-07,
+      "loss": 0.6496,
+      "step": 4803
+    },
+    {
+      "epoch": 2.8451288125555227,
+      "grad_norm": 1.6570154960641135,
+      "learning_rate": 1.3794270310745538e-07,
+      "loss": 0.6308,
+      "step": 4804
+    },
+    {
+      "epoch": 2.8457210541901095,
+      "grad_norm": 1.344278584688019,
+      "learning_rate": 1.3688606879977863e-07,
+      "loss": 0.58,
+      "step": 4805
+    },
+    {
+      "epoch": 2.8463132958246966,
+      "grad_norm": 1.4565547659975684,
+      "learning_rate": 1.3583346906033024e-07,
+      "loss": 0.6234,
+      "step": 4806
+    },
+    {
+      "epoch": 2.8469055374592833,
+      "grad_norm": 1.5371160933434378,
+      "learning_rate": 1.347849043196814e-07,
+      "loss": 0.6329,
+      "step": 4807
+    },
+    {
+      "epoch": 2.84749777909387,
+      "grad_norm": 1.5015920663462485,
+      "learning_rate": 1.3374037500675452e-07,
+      "loss": 0.6034,
+      "step": 4808
+    },
+    {
+      "epoch": 2.848090020728457,
+      "grad_norm": 1.2589321162204534,
+      "learning_rate": 1.326998815488212e-07,
+      "loss": 0.5696,
+      "step": 4809
+    },
+    {
+      "epoch": 2.8486822623630443,
+      "grad_norm": 1.5267234806346195,
+      "learning_rate": 1.3166342437150204e-07,
+      "loss": 0.6538,
+      "step": 4810
+    },
+    {
+      "epoch": 2.849274503997631,
+      "grad_norm": 3.0702462243014934,
+      "learning_rate": 1.306310038987657e-07,
+      "loss": 0.63,
+      "step": 4811
+    },
+    {
+      "epoch": 2.8498667456322178,
+      "grad_norm": 1.6574653955539411,
+      "learning_rate": 1.2960262055292884e-07,
+      "loss": 0.6405,
+      "step": 4812
+    },
+    {
+      "epoch": 2.850458987266805,
+      "grad_norm": 1.55721639877075,
+      "learning_rate": 1.2857827475466045e-07,
+      "loss": 0.6251,
+      "step": 4813
+    },
+    {
+      "epoch": 2.8510512289013916,
+      "grad_norm": 1.5458939977128197,
+      "learning_rate": 1.275579669229743e-07,
+      "loss": 0.6109,
+      "step": 4814
+    },
+    {
+      "epoch": 2.8516434705359788,
+      "grad_norm": 2.5596201627651243,
+      "learning_rate": 1.2654169747523425e-07,
+      "loss": 0.632,
+      "step": 4815
+    },
+    {
+      "epoch": 2.8522357121705655,
+      "grad_norm": 1.4732233414875766,
+      "learning_rate": 1.2552946682715116e-07,
+      "loss": 0.6403,
+      "step": 4816
+    },
+    {
+      "epoch": 2.8528279538051526,
+      "grad_norm": 1.900721516109027,
+      "learning_rate": 1.2452127539278493e-07,
+      "loss": 0.5844,
+      "step": 4817
+    },
+    {
+      "epoch": 2.8534201954397393,
+      "grad_norm": 1.634171972189307,
+      "learning_rate": 1.2351712358454115e-07,
+      "loss": 0.6421,
+      "step": 4818
+    },
+    {
+      "epoch": 2.854012437074326,
+      "grad_norm": 1.4903972592743484,
+      "learning_rate": 1.2251701181317577e-07,
+      "loss": 0.5858,
+      "step": 4819
+    },
+    {
+      "epoch": 2.854604678708913,
+      "grad_norm": 5.834292688749914,
+      "learning_rate": 1.215209404877904e-07,
+      "loss": 0.6248,
+      "step": 4820
+    },
+    {
+      "epoch": 2.8551969203435004,
+      "grad_norm": 2.127865932703041,
+      "learning_rate": 1.2052891001583356e-07,
+      "loss": 0.5898,
+      "step": 4821
+    },
+    {
+      "epoch": 2.855789161978087,
+      "grad_norm": 1.2644572972552057,
+      "learning_rate": 1.1954092080310288e-07,
+      "loss": 0.6052,
+      "step": 4822
+    },
+    {
+      "epoch": 2.856381403612674,
+      "grad_norm": 3.002356796290768,
+      "learning_rate": 1.185569732537406e-07,
+      "loss": 0.6027,
+      "step": 4823
+    },
+    {
+      "epoch": 2.856973645247261,
+      "grad_norm": 2.3834117533925405,
+      "learning_rate": 1.1757706777023592e-07,
+      "loss": 0.6043,
+      "step": 4824
+    },
+    {
+      "epoch": 2.8575658868818476,
+      "grad_norm": 1.9852462249046476,
+      "learning_rate": 1.1660120475342707e-07,
+      "loss": 0.5817,
+      "step": 4825
+    },
+    {
+      "epoch": 2.858158128516435,
+      "grad_norm": 3.440130845092385,
+      "learning_rate": 1.1562938460249473e-07,
+      "loss": 0.647,
+      "step": 4826
+    },
+    {
+      "epoch": 2.8587503701510215,
+      "grad_norm": 1.5159101594522972,
+      "learning_rate": 1.1466160771496982e-07,
+      "loss": 0.6234,
+      "step": 4827
+    },
+    {
+      "epoch": 2.8593426117856087,
+      "grad_norm": 1.7623292425646084,
+      "learning_rate": 1.1369787448672675e-07,
+      "loss": 0.6387,
+      "step": 4828
+    },
+    {
+      "epoch": 2.8599348534201954,
+      "grad_norm": 1.1069616291653879,
+      "learning_rate": 1.1273818531198689e-07,
+      "loss": 0.651,
+      "step": 4829
+    },
+    {
+      "epoch": 2.860527095054782,
+      "grad_norm": 1.5908965577567722,
+      "learning_rate": 1.1178254058331616e-07,
+      "loss": 0.5953,
+      "step": 4830
+    },
+    {
+      "epoch": 2.8611193366893692,
+      "grad_norm": 1.1175423505897313,
+      "learning_rate": 1.1083094069162747e-07,
+      "loss": 0.5764,
+      "step": 4831
+    },
+    {
+      "epoch": 2.8617115783239564,
+      "grad_norm": 2.6373508177619223,
+      "learning_rate": 1.0988338602618053e-07,
+      "loss": 0.5694,
+      "step": 4832
+    },
+    {
+      "epoch": 2.862303819958543,
+      "grad_norm": 1.5234468544649382,
+      "learning_rate": 1.0893987697457531e-07,
+      "loss": 0.602,
+      "step": 4833
+    },
+    {
+      "epoch": 2.86289606159313,
+      "grad_norm": 2.2250099868509743,
+      "learning_rate": 1.0800041392276194e-07,
+      "loss": 0.6286,
+      "step": 4834
+    },
+    {
+      "epoch": 2.863488303227717,
+      "grad_norm": 2.59536596351284,
+      "learning_rate": 1.0706499725503306e-07,
+      "loss": 0.6136,
+      "step": 4835
+    },
+    {
+      "epoch": 2.8640805448623037,
+      "grad_norm": 1.4278623489444429,
+      "learning_rate": 1.061336273540281e-07,
+      "loss": 0.616,
+      "step": 4836
+    },
+    {
+      "epoch": 2.864672786496891,
+      "grad_norm": 4.753893258101949,
+      "learning_rate": 1.0520630460072789e-07,
+      "loss": 0.6092,
+      "step": 4837
+    },
+    {
+      "epoch": 2.8652650281314775,
+      "grad_norm": 5.378554417045982,
+      "learning_rate": 1.0428302937445899e-07,
+      "loss": 0.6194,
+      "step": 4838
+    },
+    {
+      "epoch": 2.8658572697660647,
+      "grad_norm": 1.8174164225079488,
+      "learning_rate": 1.0336380205289598e-07,
+      "loss": 0.6245,
+      "step": 4839
+    },
+    {
+      "epoch": 2.8664495114006514,
+      "grad_norm": 1.8785119619391066,
+      "learning_rate": 1.024486230120525e-07,
+      "loss": 0.6206,
+      "step": 4840
+    },
+    {
+      "epoch": 2.867041753035238,
+      "grad_norm": 1.4437126735611092,
+      "learning_rate": 1.0153749262628798e-07,
+      "loss": 0.604,
+      "step": 4841
+    },
+    {
+      "epoch": 2.8676339946698253,
+      "grad_norm": 2.6007194776549225,
+      "learning_rate": 1.0063041126830542e-07,
+      "loss": 0.6137,
+      "step": 4842
+    },
+    {
+      "epoch": 2.8682262363044124,
+      "grad_norm": 1.7004655793931742,
+      "learning_rate": 9.972737930915576e-08,
+      "loss": 0.6176,
+      "step": 4843
+    },
+    {
+      "epoch": 2.868818477938999,
+      "grad_norm": 1.7229767787502699,
+      "learning_rate": 9.882839711822468e-08,
+      "loss": 0.6027,
+      "step": 4844
+    },
+    {
+      "epoch": 2.869410719573586,
+      "grad_norm": 2.90523507649429,
+      "learning_rate": 9.793346506325019e-08,
+      "loss": 0.6238,
+      "step": 4845
+    },
+    {
+      "epoch": 2.870002961208173,
+      "grad_norm": 2.42243276703994,
+      "learning_rate": 9.704258351030838e-08,
+      "loss": 0.6451,
+      "step": 4846
+    },
+    {
+      "epoch": 2.8705952028427597,
+      "grad_norm": 1.7458977764762762,
+      "learning_rate": 9.615575282381995e-08,
+      "loss": 0.5998,
+      "step": 4847
+    },
+    {
+      "epoch": 2.871187444477347,
+      "grad_norm": 1.3809462087198652,
+      "learning_rate": 9.527297336654917e-08,
+      "loss": 0.6044,
+      "step": 4848
+    },
+    {
+      "epoch": 2.8717796861119336,
+      "grad_norm": 2.8002775238184037,
+      "learning_rate": 9.439424549960164e-08,
+      "loss": 0.5919,
+      "step": 4849
+    },
+    {
+      "epoch": 2.8723719277465207,
+      "grad_norm": 1.6723003456861165,
+      "learning_rate": 9.351956958242648e-08,
+      "loss": 0.6054,
+      "step": 4850
+    },
+    {
+      "epoch": 2.8729641693811074,
+      "grad_norm": 1.2925306658163453,
+      "learning_rate": 9.264894597281637e-08,
+      "loss": 0.5964,
+      "step": 4851
+    },
+    {
+      "epoch": 2.873556411015694,
+      "grad_norm": 1.5873219385850321,
+      "learning_rate": 9.178237502690423e-08,
+      "loss": 0.6179,
+      "step": 4852
+    },
+    {
+      "epoch": 2.8741486526502813,
+      "grad_norm": 1.839675287935247,
+      "learning_rate": 9.091985709916651e-08,
+      "loss": 0.6329,
+      "step": 4853
+    },
+    {
+      "epoch": 2.8747408942848685,
+      "grad_norm": 1.1428648528723437,
+      "learning_rate": 9.006139254242319e-08,
+      "loss": 0.5924,
+      "step": 4854
+    },
+    {
+      "epoch": 2.875333135919455,
+      "grad_norm": 1.5256355451305668,
+      "learning_rate": 8.920698170783226e-08,
+      "loss": 0.572,
+      "step": 4855
+    },
+    {
+      "epoch": 2.875925377554042,
+      "grad_norm": 1.4817600093717371,
+      "learning_rate": 8.835662494489638e-08,
+      "loss": 0.653,
+      "step": 4856
+    },
+    {
+      "epoch": 2.876517619188629,
+      "grad_norm": 1.9011431000162187,
+      "learning_rate": 8.751032260145841e-08,
+      "loss": 0.6457,
+      "step": 4857
+    },
+    {
+      "epoch": 2.8771098608232157,
+      "grad_norm": 1.4632309745360057,
+      "learning_rate": 8.6668075023707e-08,
+      "loss": 0.6112,
+      "step": 4858
+    },
+    {
+      "epoch": 2.877702102457803,
+      "grad_norm": 2.231165669864582,
+      "learning_rate": 8.582988255616542e-08,
+      "loss": 0.6457,
+      "step": 4859
+    },
+    {
+      "epoch": 2.8782943440923896,
+      "grad_norm": 1.6820428525262583,
+      "learning_rate": 8.499574554170276e-08,
+      "loss": 0.6196,
+      "step": 4860
+    },
+    {
+      "epoch": 2.8788865857269768,
+      "grad_norm": 2.545677098827465,
+      "learning_rate": 8.416566432152717e-08,
+      "loss": 0.6142,
+      "step": 4861
+    },
+    {
+      "epoch": 2.8794788273615635,
+      "grad_norm": 7.836633886663809,
+      "learning_rate": 8.333963923519039e-08,
+      "loss": 0.6608,
+      "step": 4862
+    },
+    {
+      "epoch": 2.88007106899615,
+      "grad_norm": 2.1612339210989977,
+      "learning_rate": 8.251767062058102e-08,
+      "loss": 0.5832,
+      "step": 4863
+    },
+    {
+      "epoch": 2.8806633106307373,
+      "grad_norm": 4.073221900832398,
+      "learning_rate": 8.169975881393122e-08,
+      "loss": 0.6748,
+      "step": 4864
+    },
+    {
+      "epoch": 2.8812555522653245,
+      "grad_norm": 1.332971531493207,
+      "learning_rate": 8.088590414981445e-08,
+      "loss": 0.6198,
+      "step": 4865
+    },
+    {
+      "epoch": 2.881847793899911,
+      "grad_norm": 1.322813916282647,
+      "learning_rate": 8.007610696114e-08,
+      "loss": 0.6119,
+      "step": 4866
+    },
+    {
+      "epoch": 2.882440035534498,
+      "grad_norm": 1.4855030858592297,
+      "learning_rate": 7.927036757916284e-08,
+      "loss": 0.556,
+      "step": 4867
+    },
+    {
+      "epoch": 2.883032277169085,
+      "grad_norm": 1.3953005756472212,
+      "learning_rate": 7.846868633347492e-08,
+      "loss": 0.5986,
+      "step": 4868
+    },
+    {
+      "epoch": 2.8836245188036718,
+      "grad_norm": 2.9292427514375365,
+      "learning_rate": 7.767106355200949e-08,
+      "loss": 0.637,
+      "step": 4869
+    },
+    {
+      "epoch": 2.884216760438259,
+      "grad_norm": 1.5163064833083102,
+      "learning_rate": 7.687749956103774e-08,
+      "loss": 0.6435,
+      "step": 4870
+    },
+    {
+      "epoch": 2.8848090020728456,
+      "grad_norm": 2.053869781792157,
+      "learning_rate": 7.608799468517336e-08,
+      "loss": 0.5933,
+      "step": 4871
+    },
+    {
+      "epoch": 2.885401243707433,
+      "grad_norm": 1.1888149397737269,
+      "learning_rate": 7.530254924736691e-08,
+      "loss": 0.6247,
+      "step": 4872
+    },
+    {
+      "epoch": 2.8859934853420195,
+      "grad_norm": 1.6691352976148741,
+      "learning_rate": 7.452116356891136e-08,
+      "loss": 0.5899,
+      "step": 4873
+    },
+    {
+      "epoch": 2.886585726976606,
+      "grad_norm": 1.9588081253912122,
+      "learning_rate": 7.374383796943663e-08,
+      "loss": 0.6164,
+      "step": 4874
+    },
+    {
+      "epoch": 2.8871779686111934,
+      "grad_norm": 1.983783309455853,
+      "learning_rate": 7.297057276691277e-08,
+      "loss": 0.5889,
+      "step": 4875
+    },
+    {
+      "epoch": 2.8877702102457805,
+      "grad_norm": 1.2950722515129418,
+      "learning_rate": 7.22013682776479e-08,
+      "loss": 0.5975,
+      "step": 4876
+    },
+    {
+      "epoch": 2.888362451880367,
+      "grad_norm": 1.469392578281443,
+      "learning_rate": 7.143622481629031e-08,
+      "loss": 0.5583,
+      "step": 4877
+    },
+    {
+      "epoch": 2.888954693514954,
+      "grad_norm": 1.7592018298109806,
+      "learning_rate": 7.067514269582743e-08,
+      "loss": 0.5739,
+      "step": 4878
+    },
+    {
+      "epoch": 2.889546935149541,
+      "grad_norm": 1.3423603999327254,
+      "learning_rate": 6.991812222758354e-08,
+      "loss": 0.5982,
+      "step": 4879
+    },
+    {
+      "epoch": 2.890139176784128,
+      "grad_norm": 1.6619904015701097,
+      "learning_rate": 6.916516372122429e-08,
+      "loss": 0.65,
+      "step": 4880
+    },
+    {
+      "epoch": 2.890731418418715,
+      "grad_norm": 1.287530941188681,
+      "learning_rate": 6.841626748474994e-08,
+      "loss": 0.6254,
+      "step": 4881
+    },
+    {
+      "epoch": 2.8913236600533017,
+      "grad_norm": 1.6059010609864288,
+      "learning_rate": 6.767143382450214e-08,
+      "loss": 0.609,
+      "step": 4882
+    },
+    {
+      "epoch": 2.891915901687889,
+      "grad_norm": 1.4171569214064548,
+      "learning_rate": 6.693066304516049e-08,
+      "loss": 0.6497,
+      "step": 4883
+    },
+    {
+      "epoch": 2.8925081433224755,
+      "grad_norm": 2.499973096967963,
+      "learning_rate": 6.619395544974039e-08,
+      "loss": 0.6378,
+      "step": 4884
+    },
+    {
+      "epoch": 2.8931003849570622,
+      "grad_norm": 2.2303635681374527,
+      "learning_rate": 6.546131133959743e-08,
+      "loss": 0.5796,
+      "step": 4885
+    },
+    {
+      "epoch": 2.8936926265916494,
+      "grad_norm": 2.4145279843725187,
+      "learning_rate": 6.473273101442412e-08,
+      "loss": 0.6072,
+      "step": 4886
+    },
+    {
+      "epoch": 2.8942848682262365,
+      "grad_norm": 1.9679250585944843,
+      "learning_rate": 6.400821477225206e-08,
+      "loss": 0.5963,
+      "step": 4887
+    },
+    {
+      "epoch": 2.8948771098608232,
+      "grad_norm": 2.3343239037159074,
+      "learning_rate": 6.32877629094475e-08,
+      "loss": 0.5846,
+      "step": 4888
+    },
+    {
+      "epoch": 2.89546935149541,
+      "grad_norm": 1.5461453160570697,
+      "learning_rate": 6.25713757207158e-08,
+      "loss": 0.5769,
+      "step": 4889
+    },
+    {
+      "epoch": 2.896061593129997,
+      "grad_norm": 1.0705482530363706,
+      "learning_rate": 6.185905349910038e-08,
+      "loss": 0.6512,
+      "step": 4890
+    },
+    {
+      "epoch": 2.896653834764584,
+      "grad_norm": 1.2649633299278897,
+      "learning_rate": 6.115079653598032e-08,
+      "loss": 0.6649,
+      "step": 4891
+    },
+    {
+      "epoch": 2.897246076399171,
+      "grad_norm": 1.8470616975035683,
+      "learning_rate": 6.044660512107392e-08,
+      "loss": 0.6089,
+      "step": 4892
+    },
+    {
+      "epoch": 2.8978383180337577,
+      "grad_norm": 2.1565586383593662,
+      "learning_rate": 5.974647954243295e-08,
+      "loss": 0.5952,
+      "step": 4893
+    },
+    {
+      "epoch": 2.898430559668345,
+      "grad_norm": 1.854936286527562,
+      "learning_rate": 5.905042008645057e-08,
+      "loss": 0.5861,
+      "step": 4894
+    },
+    {
+      "epoch": 2.8990228013029316,
+      "grad_norm": 1.935575054571336,
+      "learning_rate": 5.835842703785233e-08,
+      "loss": 0.6239,
+      "step": 4895
+    },
+    {
+      "epoch": 2.8996150429375183,
+      "grad_norm": 1.7416078769823777,
+      "learning_rate": 5.7670500679702925e-08,
+      "loss": 0.6335,
+      "step": 4896
+    },
+    {
+      "epoch": 2.9002072845721054,
+      "grad_norm": 1.1470576618099952,
+      "learning_rate": 5.698664129340281e-08,
+      "loss": 0.5932,
+      "step": 4897
+    },
+    {
+      "epoch": 2.9007995262066926,
+      "grad_norm": 1.2955327547356683,
+      "learning_rate": 5.630684915868934e-08,
+      "loss": 0.5998,
+      "step": 4898
+    },
+    {
+      "epoch": 2.9013917678412793,
+      "grad_norm": 3.332322229387898,
+      "learning_rate": 5.5631124553636726e-08,
+      "loss": 0.6397,
+      "step": 4899
+    },
+    {
+      "epoch": 2.901984009475866,
+      "grad_norm": 1.3915052794106848,
+      "learning_rate": 5.4959467754651665e-08,
+      "loss": 0.6373,
+      "step": 4900
+    },
+    {
+      "epoch": 2.902576251110453,
+      "grad_norm": 1.8217424896661099,
+      "learning_rate": 5.429187903647992e-08,
+      "loss": 0.6303,
+      "step": 4901
+    },
+    {
+      "epoch": 2.90316849274504,
+      "grad_norm": 1.5269216925197477,
+      "learning_rate": 5.3628358672205285e-08,
+      "loss": 0.6163,
+      "step": 4902
+    },
+    {
+      "epoch": 2.903760734379627,
+      "grad_norm": 2.2464586629579517,
+      "learning_rate": 5.2968906933243966e-08,
+      "loss": 0.6387,
+      "step": 4903
+    },
+    {
+      "epoch": 2.9043529760142137,
+      "grad_norm": 1.4028855974264578,
+      "learning_rate": 5.231352408934687e-08,
+      "loss": 0.6465,
+      "step": 4904
+    },
+    {
+      "epoch": 2.904945217648801,
+      "grad_norm": 2.456044814128531,
+      "learning_rate": 5.1662210408605084e-08,
+      "loss": 0.5744,
+      "step": 4905
+    },
+    {
+      "epoch": 2.9055374592833876,
+      "grad_norm": 1.8315857516887082,
+      "learning_rate": 5.101496615744106e-08,
+      "loss": 0.5918,
+      "step": 4906
+    },
+    {
+      "epoch": 2.9061297009179743,
+      "grad_norm": 1.3272661044838177,
+      "learning_rate": 5.0371791600614115e-08,
+      "loss": 0.6366,
+      "step": 4907
+    },
+    {
+      "epoch": 2.9067219425525614,
+      "grad_norm": 2.857393405458701,
+      "learning_rate": 4.973268700121936e-08,
+      "loss": 0.5693,
+      "step": 4908
+    },
+    {
+      "epoch": 2.9073141841871486,
+      "grad_norm": 2.217648509730232,
+      "learning_rate": 4.9097652620685444e-08,
+      "loss": 0.628,
+      "step": 4909
+    },
+    {
+      "epoch": 2.9079064258217353,
+      "grad_norm": 1.286864564902065,
+      "learning_rate": 4.846668871877902e-08,
+      "loss": 0.641,
+      "step": 4910
+    },
+    {
+      "epoch": 2.908498667456322,
+      "grad_norm": 1.297605169345718,
+      "learning_rate": 4.783979555359808e-08,
+      "loss": 0.6,
+      "step": 4911
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 1.2656130698182047,
+      "learning_rate": 4.721697338157749e-08,
+      "loss": 0.5856,
+      "step": 4912
+    },
+    {
+      "epoch": 2.909683150725496,
+      "grad_norm": 1.1801287041757442,
+      "learning_rate": 4.65982224574868e-08,
+      "loss": 0.5315,
+      "step": 4913
+    },
+    {
+      "epoch": 2.910275392360083,
+      "grad_norm": 1.7243061122338232,
+      "learning_rate": 4.59835430344302e-08,
+      "loss": 0.5509,
+      "step": 4914
+    },
+    {
+      "epoch": 2.9108676339946697,
+      "grad_norm": 3.045143418965923,
+      "learning_rate": 4.537293536384657e-08,
+      "loss": 0.5648,
+      "step": 4915
+    },
+    {
+      "epoch": 2.911459875629257,
+      "grad_norm": 1.4709546764859878,
+      "learning_rate": 4.476639969550722e-08,
+      "loss": 0.5994,
+      "step": 4916
+    },
+    {
+      "epoch": 2.9120521172638436,
+      "grad_norm": 2.2971830899861776,
+      "learning_rate": 4.416393627752147e-08,
+      "loss": 0.6151,
+      "step": 4917
+    },
+    {
+      "epoch": 2.9126443588984303,
+      "grad_norm": 1.832372143366767,
+      "learning_rate": 4.3565545356327734e-08,
+      "loss": 0.6323,
+      "step": 4918
+    },
+    {
+      "epoch": 2.9132366005330175,
+      "grad_norm": 2.17503203723051,
+      "learning_rate": 4.2971227176704656e-08,
+      "loss": 0.6108,
+      "step": 4919
+    },
+    {
+      "epoch": 2.9138288421676046,
+      "grad_norm": 2.3265306058858797,
+      "learning_rate": 4.2380981981759994e-08,
+      "loss": 0.613,
+      "step": 4920
+    },
+    {
+      "epoch": 2.9144210838021913,
+      "grad_norm": 1.367396635079341,
+      "learning_rate": 4.179481001293839e-08,
+      "loss": 0.5899,
+      "step": 4921
+    },
+    {
+      "epoch": 2.915013325436778,
+      "grad_norm": 2.8847118428481857,
+      "learning_rate": 4.1212711510015826e-08,
+      "loss": 0.6114,
+      "step": 4922
+    },
+    {
+      "epoch": 2.915605567071365,
+      "grad_norm": 1.3502886513621417,
+      "learning_rate": 4.0634686711104043e-08,
+      "loss": 0.6113,
+      "step": 4923
+    },
+    {
+      "epoch": 2.916197808705952,
+      "grad_norm": 1.2799009361253797,
+      "learning_rate": 4.006073585264725e-08,
+      "loss": 0.611,
+      "step": 4924
+    },
+    {
+      "epoch": 2.916790050340539,
+      "grad_norm": 42.704793158144476,
+      "learning_rate": 3.94908591694243e-08,
+      "loss": 0.6016,
+      "step": 4925
+    },
+    {
+      "epoch": 2.9173822919751258,
+      "grad_norm": 3.2359359053380894,
+      "learning_rate": 3.89250568945454e-08,
+      "loss": 0.6014,
+      "step": 4926
+    },
+    {
+      "epoch": 2.917974533609713,
+      "grad_norm": 1.5404185217083588,
+      "learning_rate": 3.836332925945874e-08,
+      "loss": 0.6352,
+      "step": 4927
+    },
+    {
+      "epoch": 2.9185667752442996,
+      "grad_norm": 1.193895601439141,
+      "learning_rate": 3.7805676493938294e-08,
+      "loss": 0.6499,
+      "step": 4928
+    },
+    {
+      "epoch": 2.9191590168788863,
+      "grad_norm": 1.7488780731222586,
+      "learning_rate": 3.7252098826098256e-08,
+      "loss": 0.6066,
+      "step": 4929
+    },
+    {
+      "epoch": 2.9197512585134735,
+      "grad_norm": 1.3649816816969873,
+      "learning_rate": 3.6702596482381946e-08,
+      "loss": 0.5781,
+      "step": 4930
+    },
+    {
+      "epoch": 2.9203435001480607,
+      "grad_norm": 1.4996743472643042,
+      "learning_rate": 3.615716968756733e-08,
+      "loss": 0.6191,
+      "step": 4931
+    },
+    {
+      "epoch": 2.9209357417826474,
+      "grad_norm": 2.174241162693474,
+      "learning_rate": 3.5615818664764825e-08,
+      "loss": 0.6638,
+      "step": 4932
+    },
+    {
+      "epoch": 2.921527983417234,
+      "grad_norm": 1.0378189780160854,
+      "learning_rate": 3.507854363541619e-08,
+      "loss": 0.6222,
+      "step": 4933
+    },
+    {
+      "epoch": 2.9221202250518212,
+      "grad_norm": 1.273845744800928,
+      "learning_rate": 3.454534481929783e-08,
+      "loss": 0.6489,
+      "step": 4934
+    },
+    {
+      "epoch": 2.922712466686408,
+      "grad_norm": 1.665710013846029,
+      "learning_rate": 3.4016222434518634e-08,
+      "loss": 0.5685,
+      "step": 4935
+    },
+    {
+      "epoch": 2.923304708320995,
+      "grad_norm": 1.59905029910758,
+      "learning_rate": 3.349117669751767e-08,
+      "loss": 0.581,
+      "step": 4936
+    },
+    {
+      "epoch": 2.923896949955582,
+      "grad_norm": 1.6375386411169495,
+      "learning_rate": 3.297020782307092e-08,
+      "loss": 0.648,
+      "step": 4937
+    },
+    {
+      "epoch": 2.924489191590169,
+      "grad_norm": 2.9794337991351627,
+      "learning_rate": 3.245331602428126e-08,
+      "loss": 0.6716,
+      "step": 4938
+    },
+    {
+      "epoch": 2.9250814332247557,
+      "grad_norm": 1.1832076972764827,
+      "learning_rate": 3.194050151258732e-08,
+      "loss": 0.6242,
+      "step": 4939
+    },
+    {
+      "epoch": 2.9256736748593424,
+      "grad_norm": 1.4696232417709514,
+      "learning_rate": 3.14317644977602e-08,
+      "loss": 0.6092,
+      "step": 4940
+    },
+    {
+      "epoch": 2.9262659164939295,
+      "grad_norm": 1.448707576808793,
+      "learning_rate": 3.09271051879001e-08,
+      "loss": 0.6279,
+      "step": 4941
+    },
+    {
+      "epoch": 2.9268581581285167,
+      "grad_norm": 1.6727775055688563,
+      "learning_rate": 3.0426523789442994e-08,
+      "loss": 0.6081,
+      "step": 4942
+    },
+    {
+      "epoch": 2.9274503997631034,
+      "grad_norm": 1.3971103869765082,
+      "learning_rate": 2.9930020507153986e-08,
+      "loss": 0.6083,
+      "step": 4943
+    },
+    {
+      "epoch": 2.92804264139769,
+      "grad_norm": 2.2833557926974057,
+      "learning_rate": 2.9437595544130615e-08,
+      "loss": 0.5705,
+      "step": 4944
+    },
+    {
+      "epoch": 2.9286348830322773,
+      "grad_norm": 3.0762266827362024,
+      "learning_rate": 2.8949249101801747e-08,
+      "loss": 0.575,
+      "step": 4945
+    },
+    {
+      "epoch": 2.929227124666864,
+      "grad_norm": 1.5877663565907878,
+      "learning_rate": 2.8464981379929814e-08,
+      "loss": 0.5788,
+      "step": 4946
+    },
+    {
+      "epoch": 2.929819366301451,
+      "grad_norm": 1.87679323242289,
+      "learning_rate": 2.7984792576606355e-08,
+      "loss": 0.6044,
+      "step": 4947
+    },
+    {
+      "epoch": 2.930411607936038,
+      "grad_norm": 1.579589442857244,
+      "learning_rate": 2.7508682888257587e-08,
+      "loss": 0.6532,
+      "step": 4948
+    },
+    {
+      "epoch": 2.931003849570625,
+      "grad_norm": 2.2731032093347125,
+      "learning_rate": 2.7036652509636607e-08,
+      "loss": 0.6169,
+      "step": 4949
+    },
+    {
+      "epoch": 2.9315960912052117,
+      "grad_norm": 1.6629613159148906,
+      "learning_rate": 2.6568701633832295e-08,
+      "loss": 0.6204,
+      "step": 4950
+    },
+    {
+      "epoch": 2.9321883328397984,
+      "grad_norm": 1.539079572245288,
+      "learning_rate": 2.610483045226264e-08,
+      "loss": 0.6334,
+      "step": 4951
+    },
+    {
+      "epoch": 2.9327805744743856,
+      "grad_norm": 3.602880436876509,
+      "learning_rate": 2.5645039154675867e-08,
+      "loss": 0.639,
+      "step": 4952
+    },
+    {
+      "epoch": 2.9333728161089727,
+      "grad_norm": 1.987049639752013,
+      "learning_rate": 2.518932792915263e-08,
+      "loss": 0.5984,
+      "step": 4953
+    },
+    {
+      "epoch": 2.9339650577435594,
+      "grad_norm": 1.4505610755352987,
+      "learning_rate": 2.4737696962106038e-08,
+      "loss": 0.5792,
+      "step": 4954
+    },
+    {
+      "epoch": 2.934557299378146,
+      "grad_norm": 1.5647319892731515,
+      "learning_rate": 2.4290146438277205e-08,
+      "loss": 0.6523,
+      "step": 4955
+    },
+    {
+      "epoch": 2.9351495410127333,
+      "grad_norm": 1.6632250660344898,
+      "learning_rate": 2.3846676540739687e-08,
+      "loss": 0.6055,
+      "step": 4956
+    },
+    {
+      "epoch": 2.93574178264732,
+      "grad_norm": 6.646406313636889,
+      "learning_rate": 2.3407287450897265e-08,
+      "loss": 0.553,
+      "step": 4957
+    },
+    {
+      "epoch": 2.936334024281907,
+      "grad_norm": 1.5348933781298375,
+      "learning_rate": 2.2971979348485053e-08,
+      "loss": 0.5773,
+      "step": 4958
+    },
+    {
+      "epoch": 2.936926265916494,
+      "grad_norm": 7.44752974245422,
+      "learning_rate": 2.25407524115695e-08,
+      "loss": 0.6163,
+      "step": 4959
+    },
+    {
+      "epoch": 2.937518507551081,
+      "grad_norm": 1.6171911575067186,
+      "learning_rate": 2.2113606816546172e-08,
+      "loss": 0.6417,
+      "step": 4960
+    },
+    {
+      "epoch": 2.9381107491856677,
+      "grad_norm": 3.9860985428266287,
+      "learning_rate": 2.169054273814086e-08,
+      "loss": 0.6266,
+      "step": 4961
+    },
+    {
+      "epoch": 2.9387029908202544,
+      "grad_norm": 2.1282877407533936,
+      "learning_rate": 2.127156034941069e-08,
+      "loss": 0.6488,
+      "step": 4962
+    },
+    {
+      "epoch": 2.9392952324548416,
+      "grad_norm": 1.54117884175081,
+      "learning_rate": 2.085665982174412e-08,
+      "loss": 0.5842,
+      "step": 4963
+    },
+    {
+      "epoch": 2.9398874740894287,
+      "grad_norm": 1.6319922889959433,
+      "learning_rate": 2.0445841324856497e-08,
+      "loss": 0.5998,
+      "step": 4964
+    },
+    {
+      "epoch": 2.9404797157240155,
+      "grad_norm": 1.9689793022752626,
+      "learning_rate": 2.0039105026798956e-08,
+      "loss": 0.5922,
+      "step": 4965
+    },
+    {
+      "epoch": 2.941071957358602,
+      "grad_norm": 1.9910821968525951,
+      "learning_rate": 1.9636451093947296e-08,
+      "loss": 0.6316,
+      "step": 4966
+    },
+    {
+      "epoch": 2.9416641989931893,
+      "grad_norm": 1.1742386957010071,
+      "learning_rate": 1.9237879691009764e-08,
+      "loss": 0.615,
+      "step": 4967
+    },
+    {
+      "epoch": 2.942256440627776,
+      "grad_norm": 1.9540025838105881,
+      "learning_rate": 1.8843390981024835e-08,
+      "loss": 0.6529,
+      "step": 4968
+    },
+    {
+      "epoch": 2.942848682262363,
+      "grad_norm": 1.426910256416881,
+      "learning_rate": 1.84529851253612e-08,
+      "loss": 0.6153,
+      "step": 4969
+    },
+    {
+      "epoch": 2.94344092389695,
+      "grad_norm": 1.3438358517937634,
+      "learning_rate": 1.8066662283715562e-08,
+      "loss": 0.628,
+      "step": 4970
+    },
+    {
+      "epoch": 2.944033165531537,
+      "grad_norm": 1.3977430066085448,
+      "learning_rate": 1.768442261411707e-08,
+      "loss": 0.5951,
+      "step": 4971
+    },
+    {
+      "epoch": 2.9446254071661238,
+      "grad_norm": 1.205901999673902,
+      "learning_rate": 1.7306266272921756e-08,
+      "loss": 0.63,
+      "step": 4972
+    },
+    {
+      "epoch": 2.9452176488007105,
+      "grad_norm": 1.4995147029787064,
+      "learning_rate": 1.6932193414817e-08,
+      "loss": 0.6355,
+      "step": 4973
+    },
+    {
+      "epoch": 2.9458098904352976,
+      "grad_norm": 1.2679556957348548,
+      "learning_rate": 1.6562204192821507e-08,
+      "loss": 0.6433,
+      "step": 4974
+    },
+    {
+      "epoch": 2.9464021320698848,
+      "grad_norm": 10.158096968146605,
+      "learning_rate": 1.619629875827977e-08,
+      "loss": 0.5938,
+      "step": 4975
+    },
+    {
+      "epoch": 2.9469943737044715,
+      "grad_norm": 2.776393491326393,
+      "learning_rate": 1.583447726086762e-08,
+      "loss": 0.6448,
+      "step": 4976
+    },
+    {
+      "epoch": 2.947586615339058,
+      "grad_norm": 2.679025395136911,
+      "learning_rate": 1.5476739848592216e-08,
+      "loss": 0.6748,
+      "step": 4977
+    },
+    {
+      "epoch": 2.9481788569736453,
+      "grad_norm": 1.4748116377281646,
+      "learning_rate": 1.5123086667786502e-08,
+      "loss": 0.598,
+      "step": 4978
+    },
+    {
+      "epoch": 2.948771098608232,
+      "grad_norm": 1.4962671971975132,
+      "learning_rate": 1.4773517863114761e-08,
+      "loss": 0.5961,
+      "step": 4979
+    },
+    {
+      "epoch": 2.949363340242819,
+      "grad_norm": 1.6298275886188416,
+      "learning_rate": 1.4428033577571498e-08,
+      "loss": 0.6241,
+      "step": 4980
+    },
+    {
+      "epoch": 2.949955581877406,
+      "grad_norm": 1.314024889434035,
+      "learning_rate": 1.4086633952478113e-08,
+      "loss": 0.6465,
+      "step": 4981
+    },
+    {
+      "epoch": 2.950547823511993,
+      "grad_norm": 1.0907588428254045,
+      "learning_rate": 1.3749319127486228e-08,
+      "loss": 0.5865,
+      "step": 4982
+    },
+    {
+      "epoch": 2.95114006514658,
+      "grad_norm": 1.3791495256977004,
+      "learning_rate": 1.341608924057658e-08,
+      "loss": 0.6294,
+      "step": 4983
+    },
+    {
+      "epoch": 2.9517323067811665,
+      "grad_norm": 1.42971017695065,
+      "learning_rate": 1.3086944428060132e-08,
+      "loss": 0.5849,
+      "step": 4984
+    },
+    {
+      "epoch": 2.9523245484157536,
+      "grad_norm": 1.9580880132072813,
+      "learning_rate": 1.2761884824573634e-08,
+      "loss": 0.6101,
+      "step": 4985
+    },
+    {
+      "epoch": 2.952916790050341,
+      "grad_norm": 1.678533797027778,
+      "learning_rate": 1.2440910563086273e-08,
+      "loss": 0.6036,
+      "step": 4986
+    },
+    {
+      "epoch": 2.9535090316849275,
+      "grad_norm": 1.8624668233784294,
+      "learning_rate": 1.2124021774894134e-08,
+      "loss": 0.625,
+      "step": 4987
+    },
+    {
+      "epoch": 2.954101273319514,
+      "grad_norm": 1.7622666153251665,
+      "learning_rate": 1.181121858962353e-08,
+      "loss": 0.5638,
+      "step": 4988
+    },
+    {
+      "epoch": 2.9546935149541014,
+      "grad_norm": 1.4944115685837704,
+      "learning_rate": 1.1502501135225442e-08,
+      "loss": 0.5821,
+      "step": 4989
+    },
+    {
+      "epoch": 2.955285756588688,
+      "grad_norm": 2.772196851809732,
+      "learning_rate": 1.1197869537986627e-08,
+      "loss": 0.6126,
+      "step": 4990
+    },
+    {
+      "epoch": 2.9558779982232752,
+      "grad_norm": 2.6820552266235147,
+      "learning_rate": 1.089732392251519e-08,
+      "loss": 0.5898,
+      "step": 4991
+    },
+    {
+      "epoch": 2.956470239857862,
+      "grad_norm": 1.26193680911539,
+      "learning_rate": 1.0600864411753897e-08,
+      "loss": 0.6345,
+      "step": 4992
+    },
+    {
+      "epoch": 2.957062481492449,
+      "grad_norm": 3.7109469716570436,
+      "learning_rate": 1.0308491126969077e-08,
+      "loss": 0.5613,
+      "step": 4993
+    },
+    {
+      "epoch": 2.957654723127036,
+      "grad_norm": 1.426244864019141,
+      "learning_rate": 1.0020204187759507e-08,
+      "loss": 0.625,
+      "step": 4994
+    },
+    {
+      "epoch": 2.9582469647616225,
+      "grad_norm": 1.3313942225304733,
+      "learning_rate": 9.736003712050857e-09,
+      "loss": 0.5922,
+      "step": 4995
+    },
+    {
+      "epoch": 2.9588392063962097,
+      "grad_norm": 3.724539629662786,
+      "learning_rate": 9.455889816095687e-09,
+      "loss": 0.585,
+      "step": 4996
+    },
+    {
+      "epoch": 2.959431448030797,
+      "grad_norm": 1.8116214951176344,
+      "learning_rate": 9.179862614476787e-09,
+      "loss": 0.646,
+      "step": 4997
+    },
+    {
+      "epoch": 2.9600236896653835,
+      "grad_norm": 1.4870564082539035,
+      "learning_rate": 8.907922220104947e-09,
+      "loss": 0.6339,
+      "step": 4998
+    },
+    {
+      "epoch": 2.9606159312999702,
+      "grad_norm": 2.5117803113313917,
+      "learning_rate": 8.640068744220077e-09,
+      "loss": 0.5936,
+      "step": 4999
+    },
+    {
+      "epoch": 2.9612081729345574,
+      "grad_norm": 2.643791847377505,
+      "learning_rate": 8.376302296387862e-09,
+      "loss": 0.6284,
+      "step": 5000
+    },
+    {
+      "epoch": 2.961800414569144,
+      "grad_norm": 1.5248955556340484,
+      "learning_rate": 8.116622984504219e-09,
+      "loss": 0.608,
+      "step": 5001
+    },
+    {
+      "epoch": 2.9623926562037313,
+      "grad_norm": 1.4206625140578077,
+      "learning_rate": 7.861030914791956e-09,
+      "loss": 0.6016,
+      "step": 5002
+    },
+    {
+      "epoch": 2.962984897838318,
+      "grad_norm": 2.7467705894682704,
+      "learning_rate": 7.609526191804107e-09,
+      "loss": 0.6142,
+      "step": 5003
+    },
+    {
+      "epoch": 2.963577139472905,
+      "grad_norm": 2.227943416338275,
+      "learning_rate": 7.362108918418376e-09,
+      "loss": 0.6039,
+      "step": 5004
+    },
+    {
+      "epoch": 2.964169381107492,
+      "grad_norm": 2.2376624114324635,
+      "learning_rate": 7.118779195843806e-09,
+      "loss": 0.6396,
+      "step": 5005
+    },
+    {
+      "epoch": 2.9647616227420786,
+      "grad_norm": 2.2974484985519372,
+      "learning_rate": 6.8795371236163315e-09,
+      "loss": 0.6084,
+      "step": 5006
+    },
+    {
+      "epoch": 2.9653538643766657,
+      "grad_norm": 2.445011281093722,
+      "learning_rate": 6.64438279959767e-09,
+      "loss": 0.6647,
+      "step": 5007
+    },
+    {
+      "epoch": 2.965946106011253,
+      "grad_norm": 2.5322699217659825,
+      "learning_rate": 6.413316319979768e-09,
+      "loss": 0.6158,
+      "step": 5008
+    },
+    {
+      "epoch": 2.9665383476458396,
+      "grad_norm": 2.264489636975484,
+      "learning_rate": 6.186337779282569e-09,
+      "loss": 0.6319,
+      "step": 5009
+    },
+    {
+      "epoch": 2.9671305892804263,
+      "grad_norm": 1.276152934848022,
+      "learning_rate": 5.9634472703518075e-09,
+      "loss": 0.5986,
+      "step": 5010
+    },
+    {
+      "epoch": 2.9677228309150134,
+      "grad_norm": 1.7722200702077937,
+      "learning_rate": 5.744644884363437e-09,
+      "loss": 0.6465,
+      "step": 5011
+    },
+    {
+      "epoch": 2.9683150725496,
+      "grad_norm": 2.8327566548651344,
+      "learning_rate": 5.529930710820308e-09,
+      "loss": 0.5873,
+      "step": 5012
+    },
+    {
+      "epoch": 2.9689073141841873,
+      "grad_norm": 15.271569350154476,
+      "learning_rate": 5.319304837549943e-09,
+      "loss": 0.5923,
+      "step": 5013
+    },
+    {
+      "epoch": 2.969499555818774,
+      "grad_norm": 3.7239511417804705,
+      "learning_rate": 5.112767350713421e-09,
+      "loss": 0.6025,
+      "step": 5014
+    },
+    {
+      "epoch": 2.970091797453361,
+      "grad_norm": 1.4648198398137937,
+      "learning_rate": 4.9103183347942725e-09,
+      "loss": 0.6499,
+      "step": 5015
+    },
+    {
+      "epoch": 2.970684039087948,
+      "grad_norm": 1.6998540857276536,
+      "learning_rate": 4.711957872606254e-09,
+      "loss": 0.6304,
+      "step": 5016
+    },
+    {
+      "epoch": 2.9712762807225346,
+      "grad_norm": 1.6467406630255363,
+      "learning_rate": 4.517686045288905e-09,
+      "loss": 0.6548,
+      "step": 5017
+    },
+    {
+      "epoch": 2.9718685223571217,
+      "grad_norm": 1.6033639446909023,
+      "learning_rate": 4.327502932311989e-09,
+      "loss": 0.5407,
+      "step": 5018
+    },
+    {
+      "epoch": 2.972460763991709,
+      "grad_norm": 2.013297951719648,
+      "learning_rate": 4.141408611469944e-09,
+      "loss": 0.6036,
+      "step": 5019
+    },
+    {
+      "epoch": 2.9730530056262956,
+      "grad_norm": 1.5184280052213182,
+      "learning_rate": 3.959403158885211e-09,
+      "loss": 0.655,
+      "step": 5020
+    },
+    {
+      "epoch": 2.9736452472608823,
+      "grad_norm": 1.2805590152228392,
+      "learning_rate": 3.781486649010458e-09,
+      "loss": 0.6361,
+      "step": 5021
+    },
+    {
+      "epoch": 2.9742374888954695,
+      "grad_norm": 1.4488055335719843,
+      "learning_rate": 3.607659154621912e-09,
+      "loss": 0.5857,
+      "step": 5022
+    },
+    {
+      "epoch": 2.974829730530056,
+      "grad_norm": 1.5067233321604163,
+      "learning_rate": 3.437920746824919e-09,
+      "loss": 0.6314,
+      "step": 5023
+    },
+    {
+      "epoch": 2.9754219721646433,
+      "grad_norm": 2.505217366867062,
+      "learning_rate": 3.2722714950517154e-09,
+      "loss": 0.593,
+      "step": 5024
+    },
+    {
+      "epoch": 2.97601421379923,
+      "grad_norm": 2.030933501052515,
+      "learning_rate": 3.110711467063654e-09,
+      "loss": 0.6655,
+      "step": 5025
+    },
+    {
+      "epoch": 2.976606455433817,
+      "grad_norm": 1.2552202263134016,
+      "learning_rate": 2.95324072894565e-09,
+      "loss": 0.6237,
+      "step": 5026
+    },
+    {
+      "epoch": 2.977198697068404,
+      "grad_norm": 1.3223714382480334,
+      "learning_rate": 2.7998593451139534e-09,
+      "loss": 0.5913,
+      "step": 5027
+    },
+    {
+      "epoch": 2.9777909387029906,
+      "grad_norm": 1.3538171647456099,
+      "learning_rate": 2.6505673783094875e-09,
+      "loss": 0.6199,
+      "step": 5028
+    },
+    {
+      "epoch": 2.9783831803375778,
+      "grad_norm": 1.3318176094069416,
+      "learning_rate": 2.5053648896011804e-09,
+      "loss": 0.5841,
+      "step": 5029
+    },
+    {
+      "epoch": 2.978975421972165,
+      "grad_norm": 1.589489686996947,
+      "learning_rate": 2.364251938384854e-09,
+      "loss": 0.6094,
+      "step": 5030
+    },
+    {
+      "epoch": 2.9795676636067516,
+      "grad_norm": 1.7400278108274327,
+      "learning_rate": 2.227228582384333e-09,
+      "loss": 0.6143,
+      "step": 5031
+    },
+    {
+      "epoch": 2.9801599052413383,
+      "grad_norm": 2.585378676689252,
+      "learning_rate": 2.0942948776481175e-09,
+      "loss": 0.5828,
+      "step": 5032
+    },
+    {
+      "epoch": 2.9807521468759255,
+      "grad_norm": 1.7280844207518962,
+      "learning_rate": 1.965450878556041e-09,
+      "loss": 0.6314,
+      "step": 5033
+    },
+    {
+      "epoch": 2.981344388510512,
+      "grad_norm": 5.265908624204134,
+      "learning_rate": 1.8406966378103909e-09,
+      "loss": 0.5925,
+      "step": 5034
+    },
+    {
+      "epoch": 2.9819366301450994,
+      "grad_norm": 1.4033183600587051,
+      "learning_rate": 1.720032206443678e-09,
+      "loss": 0.6348,
+      "step": 5035
+    },
+    {
+      "epoch": 2.982528871779686,
+      "grad_norm": 1.8992093800736356,
+      "learning_rate": 1.6034576338141982e-09,
+      "loss": 0.655,
+      "step": 5036
+    },
+    {
+      "epoch": 2.983121113414273,
+      "grad_norm": 1.6151253256582707,
+      "learning_rate": 1.4909729676071405e-09,
+      "loss": 0.5662,
+      "step": 5037
+    },
+    {
+      "epoch": 2.98371335504886,
+      "grad_norm": 1.2292587721661237,
+      "learning_rate": 1.3825782538368083e-09,
+      "loss": 0.6521,
+      "step": 5038
+    },
+    {
+      "epoch": 2.9843055966834466,
+      "grad_norm": 1.4529622161539926,
+      "learning_rate": 1.278273536839958e-09,
+      "loss": 0.6323,
+      "step": 5039
+    },
+    {
+      "epoch": 2.984897838318034,
+      "grad_norm": 1.498906553325035,
+      "learning_rate": 1.178058859285791e-09,
+      "loss": 0.5986,
+      "step": 5040
+    },
+    {
+      "epoch": 2.985490079952621,
+      "grad_norm": 1.2660500957601097,
+      "learning_rate": 1.0819342621648522e-09,
+      "loss": 0.5796,
+      "step": 5041
+    },
+    {
+      "epoch": 2.9860823215872077,
+      "grad_norm": 1.336021937389783,
+      "learning_rate": 9.898997848001302e-10,
+      "loss": 0.5771,
+      "step": 5042
+    },
+    {
+      "epoch": 2.9866745632217944,
+      "grad_norm": 2.936187746887494,
+      "learning_rate": 9.019554648381778e-10,
+      "loss": 0.5746,
+      "step": 5043
+    },
+    {
+      "epoch": 2.9872668048563815,
+      "grad_norm": 1.974248988367071,
+      "learning_rate": 8.181013382524416e-10,
+      "loss": 0.596,
+      "step": 5044
+    },
+    {
+      "epoch": 2.9878590464909682,
+      "grad_norm": 2.076337021390945,
+      "learning_rate": 7.383374393454823e-10,
+      "loss": 0.6021,
+      "step": 5045
+    },
+    {
+      "epoch": 2.9884512881255554,
+      "grad_norm": 1.2554188823022927,
+      "learning_rate": 6.626638007434239e-10,
+      "loss": 0.5964,
+      "step": 5046
+    },
+    {
+      "epoch": 2.989043529760142,
+      "grad_norm": 1.4522435596098013,
+      "learning_rate": 5.910804534015046e-10,
+      "loss": 0.6149,
+      "step": 5047
+    },
+    {
+      "epoch": 2.9896357713947292,
+      "grad_norm": 1.785638501758462,
+      "learning_rate": 5.23587426601857e-10,
+      "loss": 0.6229,
+      "step": 5048
+    },
+    {
+      "epoch": 2.990228013029316,
+      "grad_norm": 1.5511181558630578,
+      "learning_rate": 4.601847479523969e-10,
+      "loss": 0.5953,
+      "step": 5049
+    },
+    {
+      "epoch": 2.9908202546639027,
+      "grad_norm": 1.5809877291449523,
+      "learning_rate": 4.008724433890443e-10,
+      "loss": 0.6185,
+      "step": 5050
+    },
+    {
+      "epoch": 2.99141249629849,
+      "grad_norm": 1.25867971827954,
+      "learning_rate": 3.4565053717350303e-10,
+      "loss": 0.6286,
+      "step": 5051
+    },
+    {
+      "epoch": 2.9920047379330765,
+      "grad_norm": 1.6791689479616938,
+      "learning_rate": 2.94519051895481e-10,
+      "loss": 0.6264,
+      "step": 5052
+    },
+    {
+      "epoch": 2.9925969795676637,
+      "grad_norm": 2.2845001847932447,
+      "learning_rate": 2.474780084682493e-10,
+      "loss": 0.6321,
+      "step": 5053
+    },
+    {
+      "epoch": 2.9931892212022504,
+      "grad_norm": 2.632181608604146,
+      "learning_rate": 2.0452742613641386e-10,
+      "loss": 0.6344,
+      "step": 5054
+    },
+    {
+      "epoch": 2.9937814628368375,
+      "grad_norm": 1.8651242937205437,
+      "learning_rate": 1.6566732246925398e-10,
+      "loss": 0.6184,
+      "step": 5055
+    },
+    {
+      "epoch": 2.9943737044714243,
+      "grad_norm": 1.375082869113834,
+      "learning_rate": 1.3089771336072256e-10,
+      "loss": 0.6076,
+      "step": 5056
+    },
+    {
+      "epoch": 2.9949659461060114,
+      "grad_norm": 1.3906730573347121,
+      "learning_rate": 1.0021861303610714e-10,
+      "loss": 0.5459,
+      "step": 5057
+    },
+    {
+      "epoch": 2.995558187740598,
+      "grad_norm": 1.4138902354298646,
+      "learning_rate": 7.363003404314839e-11,
+      "loss": 0.5922,
+      "step": 5058
+    },
+    {
+      "epoch": 2.9961504293751853,
+      "grad_norm": 4.735285490670945,
+      "learning_rate": 5.113198725870128e-11,
+      "loss": 0.643,
+      "step": 5059
+    },
+    {
+      "epoch": 2.996742671009772,
+      "grad_norm": 4.0056402967283296,
+      "learning_rate": 3.272448188429422e-11,
+      "loss": 0.5758,
+      "step": 5060
+    },
+    {
+      "epoch": 2.9973349126443587,
+      "grad_norm": 1.6787822740596676,
+      "learning_rate": 1.8407525452790454e-11,
+      "loss": 0.5937,
+      "step": 5061
+    },
+    {
+      "epoch": 2.997927154278946,
+      "grad_norm": 1.5279759168116307,
+      "learning_rate": 8.181123817285752e-12,
+      "loss": 0.6155,
+      "step": 5062
+    },
+    {
+      "epoch": 2.9985193959135326,
+      "grad_norm": 1.977332811161579,
+      "learning_rate": 2.0452811633209224e-12,
+      "loss": 0.6343,
+      "step": 5063
+    },
+    {
+      "epoch": 2.9991116375481197,
+      "grad_norm": 1.904074374443596,
+      "learning_rate": 0.0,
+      "loss": 0.595,
+      "step": 5064
+    },
+    {
+      "epoch": 2.9991116375481197,
+      "step": 5064,
+      "total_flos": 3794121211510784.0,
+      "train_loss": 0.7367976606951505,
+      "train_runtime": 36668.0035,
+      "train_samples_per_second": 17.677,
+      "train_steps_per_second": 0.138
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5064,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 6000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3794121211510784.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}