diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,40406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 5768,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9437410831451416,
+      "learning_rate": 2.9994798890429956e-05,
+      "loss": 5.931,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.015024662017822,
+      "learning_rate": 2.9989597780859918e-05,
+      "loss": 5.7445,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.849590301513672,
+      "learning_rate": 2.9984396671289873e-05,
+      "loss": 5.5895,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.190009593963623,
+      "learning_rate": 2.9979195561719835e-05,
+      "loss": 5.4091,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.478919506072998,
+      "learning_rate": 2.997399445214979e-05,
+      "loss": 5.1776,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.940654754638672,
+      "learning_rate": 2.9968793342579752e-05,
+      "loss": 5.0447,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.351830959320068,
+      "learning_rate": 2.9963592233009708e-05,
+      "loss": 4.838,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.111618518829346,
+      "learning_rate": 2.9958391123439666e-05,
+      "loss": 4.9007,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.2174973487854,
+      "learning_rate": 2.9953190013869625e-05,
+      "loss": 4.5247,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.609569549560547,
+      "learning_rate": 2.9947988904299583e-05,
+      "loss": 4.5729,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.401117324829102,
+      "learning_rate": 2.9942787794729542e-05,
+      "loss": 4.2805,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.256618976593018,
+      "learning_rate": 2.99375866851595e-05,
+      "loss": 4.2401,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.445135593414307,
+      "learning_rate": 2.993238557558946e-05,
+      "loss": 4.2249,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.126568794250488,
+      "learning_rate": 2.9927184466019418e-05,
+      "loss": 4.0671,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.497378826141357,
+      "learning_rate": 2.9921983356449377e-05,
+      "loss": 3.9239,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.693580150604248,
+      "learning_rate": 2.9916782246879335e-05,
+      "loss": 3.8454,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.538710117340088,
+      "learning_rate": 2.991158113730929e-05,
+      "loss": 3.8575,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.401360511779785,
+      "learning_rate": 2.9906380027739252e-05,
+      "loss": 3.6229,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.2959623336792,
+      "learning_rate": 2.9901178918169208e-05,
+      "loss": 3.4782,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.22870635986328,
+      "learning_rate": 2.989597780859917e-05,
+      "loss": 3.3089,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.805909156799316,
+      "learning_rate": 2.9890776699029125e-05,
+      "loss": 3.2963,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.135955810546875,
+      "learning_rate": 2.9885575589459087e-05,
+      "loss": 2.9362,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.550091743469238,
+      "learning_rate": 2.9880374479889042e-05,
+      "loss": 2.9321,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.74002456665039,
+      "learning_rate": 2.9875173370319e-05,
+      "loss": 3.1441,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.57274055480957,
+      "learning_rate": 2.986997226074896e-05,
+      "loss": 2.9708,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.764720916748047,
+      "learning_rate": 2.9864771151178918e-05,
+      "loss": 2.86,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.409101486206055,
+      "learning_rate": 2.9859570041608877e-05,
+      "loss": 2.7167,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.179572105407715,
+      "learning_rate": 2.9854368932038835e-05,
+      "loss": 2.8979,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.99836254119873,
+      "learning_rate": 2.9849167822468794e-05,
+      "loss": 3.0337,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.729070663452148,
+      "learning_rate": 2.9843966712898752e-05,
+      "loss": 2.5007,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.038841247558594,
+      "learning_rate": 2.983876560332871e-05,
+      "loss": 2.5117,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.810430526733398,
+      "learning_rate": 2.983356449375867e-05,
+      "loss": 2.4796,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.403511047363281,
+      "learning_rate": 2.9828363384188625e-05,
+      "loss": 2.7448,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.946587562561035,
+      "learning_rate": 2.9823162274618587e-05,
+      "loss": 2.4083,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.916707038879395,
+      "learning_rate": 2.9817961165048542e-05,
+      "loss": 2.5852,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.283550262451172,
+      "learning_rate": 2.9812760055478504e-05,
+      "loss": 2.6747,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.464363098144531,
+      "learning_rate": 2.980755894590846e-05,
+      "loss": 2.4931,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.841303825378418,
+      "learning_rate": 2.980235783633842e-05,
+      "loss": 2.4501,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.056305885314941,
+      "learning_rate": 2.9797156726768376e-05,
+      "loss": 2.1492,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.749049186706543,
+      "learning_rate": 2.9791955617198335e-05,
+      "loss": 2.1279,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.045581817626953,
+      "learning_rate": 2.9786754507628294e-05,
+      "loss": 2.3315,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.27183723449707,
+      "learning_rate": 2.9781553398058252e-05,
+      "loss": 2.3391,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.684698104858398,
+      "learning_rate": 2.977635228848821e-05,
+      "loss": 2.4632,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.084192276000977,
+      "learning_rate": 2.977115117891817e-05,
+      "loss": 2.1287,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.7970552444458,
+      "learning_rate": 2.9765950069348128e-05,
+      "loss": 2.2731,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.639681816101074,
+      "learning_rate": 2.9760748959778087e-05,
+      "loss": 2.7399,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.61768627166748,
+      "learning_rate": 2.9755547850208042e-05,
+      "loss": 2.1801,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.624665260314941,
+      "learning_rate": 2.9750346740638004e-05,
+      "loss": 2.1888,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.417475700378418,
+      "learning_rate": 2.974514563106796e-05,
+      "loss": 2.236,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.710034370422363,
+      "learning_rate": 2.973994452149792e-05,
+      "loss": 2.2905,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.710832595825195,
+      "learning_rate": 2.9734743411927876e-05,
+      "loss": 2.3676,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.496001243591309,
+      "learning_rate": 2.972954230235784e-05,
+      "loss": 2.0832,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.159036636352539,
+      "learning_rate": 2.9724341192787794e-05,
+      "loss": 2.3043,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.032967567443848,
+      "learning_rate": 2.9719140083217756e-05,
+      "loss": 2.0554,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.112791061401367,
+      "learning_rate": 2.971393897364771e-05,
+      "loss": 2.0387,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.197330474853516,
+      "learning_rate": 2.970873786407767e-05,
+      "loss": 2.1835,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.417916297912598,
+      "learning_rate": 2.9703536754507628e-05,
+      "loss": 2.3736,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.876360893249512,
+      "learning_rate": 2.9698335644937587e-05,
+      "loss": 2.07,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.316340446472168,
+      "learning_rate": 2.9693134535367545e-05,
+      "loss": 1.9853,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.644152641296387,
+      "learning_rate": 2.9687933425797504e-05,
+      "loss": 2.2578,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.609979629516602,
+      "learning_rate": 2.9682732316227463e-05,
+      "loss": 1.9305,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.370908737182617,
+      "learning_rate": 2.967753120665742e-05,
+      "loss": 2.2532,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.825346946716309,
+      "learning_rate": 2.9672330097087376e-05,
+      "loss": 2.1042,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.844170570373535,
+      "learning_rate": 2.966712898751734e-05,
+      "loss": 1.9153,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.59283447265625,
+      "learning_rate": 2.9661927877947294e-05,
+      "loss": 1.9381,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.945977210998535,
+      "learning_rate": 2.9656726768377256e-05,
+      "loss": 2.1282,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.630114555358887,
+      "learning_rate": 2.965152565880721e-05,
+      "loss": 1.8607,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.14123249053955,
+      "learning_rate": 2.9646324549237173e-05,
+      "loss": 1.9686,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.771526336669922,
+      "learning_rate": 2.9641123439667128e-05,
+      "loss": 2.093,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.731348991394043,
+      "learning_rate": 2.9635922330097087e-05,
+      "loss": 2.0783,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.250776290893555,
+      "learning_rate": 2.9630721220527045e-05,
+      "loss": 2.0302,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.943378448486328,
+      "learning_rate": 2.9625520110957004e-05,
+      "loss": 1.9559,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.509943008422852,
+      "learning_rate": 2.9620319001386963e-05,
+      "loss": 2.0576,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.13676643371582,
+      "learning_rate": 2.961511789181692e-05,
+      "loss": 2.0763,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.131099700927734,
+      "learning_rate": 2.960991678224688e-05,
+      "loss": 1.9826,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.166464805603027,
+      "learning_rate": 2.960471567267684e-05,
+      "loss": 1.9439,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.928932189941406,
+      "learning_rate": 2.9599514563106797e-05,
+      "loss": 2.1786,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.96970272064209,
+      "learning_rate": 2.9594313453536756e-05,
+      "loss": 2.168,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.965185165405273,
+      "learning_rate": 2.958911234396671e-05,
+      "loss": 1.9261,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.88183307647705,
+      "learning_rate": 2.9583911234396673e-05,
+      "loss": 1.9665,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.76819133758545,
+      "learning_rate": 2.9578710124826628e-05,
+      "loss": 1.9668,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.642396926879883,
+      "learning_rate": 2.957350901525659e-05,
+      "loss": 2.0166,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.919540405273438,
+      "learning_rate": 2.9568307905686545e-05,
+      "loss": 2.0598,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.992358207702637,
+      "learning_rate": 2.9563106796116507e-05,
+      "loss": 2.1594,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.36939811706543,
+      "learning_rate": 2.9557905686546462e-05,
+      "loss": 1.6627,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.166382789611816,
+      "learning_rate": 2.955270457697642e-05,
+      "loss": 1.6491,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.406874656677246,
+      "learning_rate": 2.954750346740638e-05,
+      "loss": 1.9434,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.465152740478516,
+      "learning_rate": 2.954230235783634e-05,
+      "loss": 2.0523,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.742449760437012,
+      "learning_rate": 2.9537101248266297e-05,
+      "loss": 1.6452,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.21478271484375,
+      "learning_rate": 2.9531900138696256e-05,
+      "loss": 1.5676,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.817789077758789,
+      "learning_rate": 2.9526699029126214e-05,
+      "loss": 1.8384,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.952509880065918,
+      "learning_rate": 2.9521497919556173e-05,
+      "loss": 1.4751,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.394773483276367,
+      "learning_rate": 2.9516296809986128e-05,
+      "loss": 1.9604,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.468931198120117,
+      "learning_rate": 2.951109570041609e-05,
+      "loss": 1.7915,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.616106033325195,
+      "learning_rate": 2.9505894590846045e-05,
+      "loss": 1.8767,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.973332405090332,
+      "learning_rate": 2.9500693481276007e-05,
+      "loss": 1.9367,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.00455379486084,
+      "learning_rate": 2.9495492371705962e-05,
+      "loss": 1.8951,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.214734077453613,
+      "learning_rate": 2.9490291262135924e-05,
+      "loss": 1.8986,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.523670196533203,
+      "learning_rate": 2.948509015256588e-05,
+      "loss": 1.7126,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.66853141784668,
+      "learning_rate": 2.947988904299584e-05,
+      "loss": 1.8396,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.567336082458496,
+      "learning_rate": 2.9474687933425797e-05,
+      "loss": 1.9027,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.822750091552734,
+      "learning_rate": 2.9469486823855756e-05,
+      "loss": 1.5785,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.275505065917969,
+      "learning_rate": 2.9464285714285714e-05,
+      "loss": 1.6998,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.994288444519043,
+      "learning_rate": 2.9459084604715673e-05,
+      "loss": 1.762,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.087584495544434,
+      "learning_rate": 2.945388349514563e-05,
+      "loss": 1.8763,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.028717041015625,
+      "learning_rate": 2.944868238557559e-05,
+      "loss": 1.6841,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.4558687210083,
+      "learning_rate": 2.944348127600555e-05,
+      "loss": 1.8046,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.519815444946289,
+      "learning_rate": 2.9438280166435507e-05,
+      "loss": 1.6152,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.68604564666748,
+      "learning_rate": 2.9433079056865462e-05,
+      "loss": 1.5206,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.852556228637695,
+      "learning_rate": 2.9427877947295424e-05,
+      "loss": 1.7853,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.174117088317871,
+      "learning_rate": 2.942267683772538e-05,
+      "loss": 1.5928,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.336973190307617,
+      "learning_rate": 2.941747572815534e-05,
+      "loss": 2.0873,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.930852890014648,
+      "learning_rate": 2.9412274618585297e-05,
+      "loss": 1.9045,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.178422927856445,
+      "learning_rate": 2.940707350901526e-05,
+      "loss": 2.0958,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.622445106506348,
+      "learning_rate": 2.9401872399445214e-05,
+      "loss": 1.6986,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.305771827697754,
+      "learning_rate": 2.9396671289875173e-05,
+      "loss": 1.3866,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.963384628295898,
+      "learning_rate": 2.939147018030513e-05,
+      "loss": 1.9851,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.022110939025879,
+      "learning_rate": 2.938626907073509e-05,
+      "loss": 2.0555,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.478226661682129,
+      "learning_rate": 2.938106796116505e-05,
+      "loss": 2.2252,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.553197860717773,
+      "learning_rate": 2.9375866851595007e-05,
+      "loss": 1.9081,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.280830383300781,
+      "learning_rate": 2.9370665742024966e-05,
+      "loss": 1.987,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.980745315551758,
+      "learning_rate": 2.9365464632454924e-05,
+      "loss": 1.8298,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.949466705322266,
+      "learning_rate": 2.9360263522884883e-05,
+      "loss": 2.0186,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.642003059387207,
+      "learning_rate": 2.935506241331484e-05,
+      "loss": 1.6872,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.050252914428711,
+      "learning_rate": 2.9349861303744797e-05,
+      "loss": 1.842,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.293116569519043,
+      "learning_rate": 2.934466019417476e-05,
+      "loss": 1.4905,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.450900077819824,
+      "learning_rate": 2.9339459084604714e-05,
+      "loss": 1.5967,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.116268157958984,
+      "learning_rate": 2.9334257975034676e-05,
+      "loss": 1.6057,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.433635711669922,
+      "learning_rate": 2.932905686546463e-05,
+      "loss": 1.9678,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.165159225463867,
+      "learning_rate": 2.9323855755894593e-05,
+      "loss": 1.7544,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.805227279663086,
+      "learning_rate": 2.931865464632455e-05,
+      "loss": 1.5984,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.324423789978027,
+      "learning_rate": 2.9313453536754507e-05,
+      "loss": 1.5765,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.1013765335083,
+      "learning_rate": 2.9308252427184466e-05,
+      "loss": 2.0393,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.632279396057129,
+      "learning_rate": 2.9303051317614424e-05,
+      "loss": 1.7005,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.309398651123047,
+      "learning_rate": 2.9297850208044383e-05,
+      "loss": 2.0103,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.914822578430176,
+      "learning_rate": 2.929264909847434e-05,
+      "loss": 1.5841,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.877006530761719,
+      "learning_rate": 2.92874479889043e-05,
+      "loss": 1.6234,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.33832836151123,
+      "learning_rate": 2.928224687933426e-05,
+      "loss": 1.8756,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.796295166015625,
+      "learning_rate": 2.9277045769764217e-05,
+      "loss": 1.3099,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.88008975982666,
+      "learning_rate": 2.9271844660194176e-05,
+      "loss": 1.3502,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.271021842956543,
+      "learning_rate": 2.926664355062413e-05,
+      "loss": 1.7604,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.75383472442627,
+      "learning_rate": 2.9261442441054093e-05,
+      "loss": 1.5372,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.157800674438477,
+      "learning_rate": 2.925624133148405e-05,
+      "loss": 1.5213,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.073105812072754,
+      "learning_rate": 2.925104022191401e-05,
+      "loss": 1.8372,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.894213676452637,
+      "learning_rate": 2.9245839112343966e-05,
+      "loss": 1.4782,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.824172973632812,
+      "learning_rate": 2.9240638002773928e-05,
+      "loss": 1.3582,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.166221618652344,
+      "learning_rate": 2.9235436893203883e-05,
+      "loss": 1.638,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.869292259216309,
+      "learning_rate": 2.923023578363384e-05,
+      "loss": 1.5228,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.476653099060059,
+      "learning_rate": 2.92250346740638e-05,
+      "loss": 1.7702,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.233400344848633,
+      "learning_rate": 2.921983356449376e-05,
+      "loss": 1.4167,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.32200813293457,
+      "learning_rate": 2.9214632454923717e-05,
+      "loss": 1.9758,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.03254508972168,
+      "learning_rate": 2.9209431345353676e-05,
+      "loss": 1.6569,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.843968391418457,
+      "learning_rate": 2.9204230235783635e-05,
+      "loss": 1.916,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.503581047058105,
+      "learning_rate": 2.9199029126213593e-05,
+      "loss": 1.4348,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.225092887878418,
+      "learning_rate": 2.919382801664355e-05,
+      "loss": 1.4901,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.823345184326172,
+      "learning_rate": 2.918862690707351e-05,
+      "loss": 1.6924,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.971875190734863,
+      "learning_rate": 2.9183425797503466e-05,
+      "loss": 2.0068,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.116519927978516,
+      "learning_rate": 2.9178224687933428e-05,
+      "loss": 1.606,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.954903602600098,
+      "learning_rate": 2.9173023578363383e-05,
+      "loss": 2.0906,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.031601905822754,
+      "learning_rate": 2.9167822468793345e-05,
+      "loss": 1.6135,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.684679985046387,
+      "learning_rate": 2.91626213592233e-05,
+      "loss": 1.7958,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.21334171295166,
+      "learning_rate": 2.9157420249653262e-05,
+      "loss": 1.9522,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.80127239227295,
+      "learning_rate": 2.9152219140083217e-05,
+      "loss": 1.7372,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.802678108215332,
+      "learning_rate": 2.9147018030513176e-05,
+      "loss": 1.4523,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.90209674835205,
+      "learning_rate": 2.9141816920943135e-05,
+      "loss": 2.1396,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.637846946716309,
+      "learning_rate": 2.9136615811373093e-05,
+      "loss": 1.8181,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.279107093811035,
+      "learning_rate": 2.9131414701803052e-05,
+      "loss": 1.453,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.76880931854248,
+      "learning_rate": 2.912621359223301e-05,
+      "loss": 1.882,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.019669532775879,
+      "learning_rate": 2.912101248266297e-05,
+      "loss": 1.3744,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.197808265686035,
+      "learning_rate": 2.9115811373092928e-05,
+      "loss": 1.4871,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.867878913879395,
+      "learning_rate": 2.9110610263522883e-05,
+      "loss": 1.6954,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.549890518188477,
+      "learning_rate": 2.9105409153952845e-05,
+      "loss": 1.6982,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.21431827545166,
+      "learning_rate": 2.91002080443828e-05,
+      "loss": 1.8707,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.142809867858887,
+      "learning_rate": 2.9095006934812762e-05,
+      "loss": 1.7527,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.7861409187316895,
+      "learning_rate": 2.9089805825242717e-05,
+      "loss": 1.3259,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.310654640197754,
+      "learning_rate": 2.908460471567268e-05,
+      "loss": 1.6632,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.789751052856445,
+      "learning_rate": 2.9079403606102635e-05,
+      "loss": 1.7563,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.376636505126953,
+      "learning_rate": 2.9074202496532593e-05,
+      "loss": 1.4722,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.35213565826416,
+      "learning_rate": 2.9069001386962552e-05,
+      "loss": 1.5533,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.820728302001953,
+      "learning_rate": 2.906380027739251e-05,
+      "loss": 1.6021,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.684493064880371,
+      "learning_rate": 2.905859916782247e-05,
+      "loss": 1.6719,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.775588989257812,
+      "learning_rate": 2.9053398058252428e-05,
+      "loss": 1.9681,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.648788452148438,
+      "learning_rate": 2.9048196948682386e-05,
+      "loss": 1.6951,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.2158842086792,
+      "learning_rate": 2.9042995839112345e-05,
+      "loss": 1.55,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.66967248916626,
+      "learning_rate": 2.9037794729542303e-05,
+      "loss": 1.2228,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.79252815246582,
+      "learning_rate": 2.9032593619972262e-05,
+      "loss": 1.4182,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.891547203063965,
+      "learning_rate": 2.9027392510402217e-05,
+      "loss": 1.8917,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.30432415008545,
+      "learning_rate": 2.902219140083218e-05,
+      "loss": 1.3425,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.478687286376953,
+      "learning_rate": 2.9016990291262134e-05,
+      "loss": 1.675,
+      "step": 189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.538810729980469,
+      "learning_rate": 2.9011789181692096e-05,
+      "loss": 1.6318,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.746855735778809,
+      "learning_rate": 2.900658807212205e-05,
+      "loss": 1.4732,
+      "step": 191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.903489112854004,
+      "learning_rate": 2.9001386962552014e-05,
+      "loss": 1.7595,
+      "step": 192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.650206565856934,
+      "learning_rate": 2.899618585298197e-05,
+      "loss": 1.3362,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.096489906311035,
+      "learning_rate": 2.8990984743411928e-05,
+      "loss": 1.6629,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.80374526977539,
+      "learning_rate": 2.8985783633841886e-05,
+      "loss": 1.4659,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.025944709777832,
+      "learning_rate": 2.8980582524271845e-05,
+      "loss": 1.287,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.64016056060791,
+      "learning_rate": 2.8975381414701803e-05,
+      "loss": 1.5318,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.08736801147461,
+      "learning_rate": 2.8970180305131762e-05,
+      "loss": 1.3464,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.437697410583496,
+      "learning_rate": 2.896497919556172e-05,
+      "loss": 1.2239,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.105047225952148,
+      "learning_rate": 2.895977808599168e-05,
+      "loss": 1.4558,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.02853012084961,
+      "learning_rate": 2.8954576976421638e-05,
+      "loss": 1.6013,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.757847785949707,
+      "learning_rate": 2.8949375866851596e-05,
+      "loss": 1.4901,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.401432991027832,
+      "learning_rate": 2.894417475728155e-05,
+      "loss": 1.4704,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.462626457214355,
+      "learning_rate": 2.8938973647711514e-05,
+      "loss": 1.5824,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.464775085449219,
+      "learning_rate": 2.893377253814147e-05,
+      "loss": 1.7951,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.73655891418457,
+      "learning_rate": 2.892857142857143e-05,
+      "loss": 1.3574,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.306361198425293,
+      "learning_rate": 2.8923370319001386e-05,
+      "loss": 1.3476,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.368309020996094,
+      "learning_rate": 2.8918169209431348e-05,
+      "loss": 1.7274,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.37753963470459,
+      "learning_rate": 2.8912968099861303e-05,
+      "loss": 1.5802,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.50301456451416,
+      "learning_rate": 2.8907766990291262e-05,
+      "loss": 1.5111,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.222660064697266,
+      "learning_rate": 2.890256588072122e-05,
+      "loss": 1.5031,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.269126892089844,
+      "learning_rate": 2.889736477115118e-05,
+      "loss": 1.5726,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.064433097839355,
+      "learning_rate": 2.8892163661581138e-05,
+      "loss": 1.5452,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.877541542053223,
+      "learning_rate": 2.8886962552011096e-05,
+      "loss": 1.4323,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.784576416015625,
+      "learning_rate": 2.8881761442441055e-05,
+      "loss": 1.3438,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.718664169311523,
+      "learning_rate": 2.8876560332871014e-05,
+      "loss": 1.3949,
+      "step": 216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.591150283813477,
+      "learning_rate": 2.887135922330097e-05,
+      "loss": 1.4218,
+      "step": 217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.845672607421875,
+      "learning_rate": 2.886615811373093e-05,
+      "loss": 1.4808,
+      "step": 218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.82565689086914,
+      "learning_rate": 2.8860957004160886e-05,
+      "loss": 1.4606,
+      "step": 219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.769989967346191,
+      "learning_rate": 2.8855755894590848e-05,
+      "loss": 1.5558,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.301299095153809,
+      "learning_rate": 2.8850554785020803e-05,
+      "loss": 1.2951,
+      "step": 221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.620203971862793,
+      "learning_rate": 2.8845353675450765e-05,
+      "loss": 1.3628,
+      "step": 222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.985174179077148,
+      "learning_rate": 2.884015256588072e-05,
+      "loss": 1.5432,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.746687889099121,
+      "learning_rate": 2.8834951456310683e-05,
+      "loss": 1.5892,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.504813194274902,
+      "learning_rate": 2.8829750346740638e-05,
+      "loss": 1.9179,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.689440727233887,
+      "learning_rate": 2.8824549237170596e-05,
+      "loss": 1.5908,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.399290084838867,
+      "learning_rate": 2.8819348127600555e-05,
+      "loss": 1.6392,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.498589515686035,
+      "learning_rate": 2.8814147018030514e-05,
+      "loss": 1.5517,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.25301742553711,
+      "learning_rate": 2.8808945908460472e-05,
+      "loss": 1.4527,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.789177894592285,
+      "learning_rate": 2.880374479889043e-05,
+      "loss": 1.607,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.231470108032227,
+      "learning_rate": 2.879854368932039e-05,
+      "loss": 1.5014,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.351259231567383,
+      "learning_rate": 2.8793342579750348e-05,
+      "loss": 1.7941,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.7614006996154785,
+      "learning_rate": 2.8788141470180303e-05,
+      "loss": 1.243,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.26957893371582,
+      "learning_rate": 2.8782940360610265e-05,
+      "loss": 1.4235,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.312116622924805,
+      "learning_rate": 2.877773925104022e-05,
+      "loss": 1.7436,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.507518768310547,
+      "learning_rate": 2.8772538141470182e-05,
+      "loss": 1.258,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.010966300964355,
+      "learning_rate": 2.8767337031900138e-05,
+      "loss": 1.4923,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.918039321899414,
+      "learning_rate": 2.87621359223301e-05,
+      "loss": 1.6282,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.306294441223145,
+      "learning_rate": 2.8756934812760055e-05,
+      "loss": 1.5557,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.08404541015625,
+      "learning_rate": 2.8751733703190014e-05,
+      "loss": 1.3712,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.742134094238281,
+      "learning_rate": 2.8746532593619972e-05,
+      "loss": 1.4655,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.379899978637695,
+      "learning_rate": 2.874133148404993e-05,
+      "loss": 1.6538,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.22764778137207,
+      "learning_rate": 2.873613037447989e-05,
+      "loss": 1.6601,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.94730281829834,
+      "learning_rate": 2.8730929264909848e-05,
+      "loss": 1.5977,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.50390911102295,
+      "learning_rate": 2.8725728155339807e-05,
+      "loss": 1.7249,
+      "step": 245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.118917465209961,
+      "learning_rate": 2.8720527045769765e-05,
+      "loss": 1.4243,
+      "step": 246
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.047847747802734,
+      "learning_rate": 2.8715325936199724e-05,
+      "loss": 1.3193,
+      "step": 247
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.507089614868164,
+      "learning_rate": 2.8710124826629682e-05,
+      "loss": 1.3854,
+      "step": 248
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.858211517333984,
+      "learning_rate": 2.8704923717059638e-05,
+      "loss": 1.3367,
+      "step": 249
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.835949897766113,
+      "learning_rate": 2.86997226074896e-05,
+      "loss": 1.5726,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.892538070678711,
+      "learning_rate": 2.8694521497919555e-05,
+      "loss": 1.5588,
+      "step": 251
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.223431587219238,
+      "learning_rate": 2.8689320388349517e-05,
+      "loss": 1.1788,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.330002784729004,
+      "learning_rate": 2.8684119278779472e-05,
+      "loss": 1.4924,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.642496109008789,
+      "learning_rate": 2.8678918169209434e-05,
+      "loss": 1.3023,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.62267780303955,
+      "learning_rate": 2.867371705963939e-05,
+      "loss": 1.4322,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.397704124450684,
+      "learning_rate": 2.8668515950069348e-05,
+      "loss": 1.7847,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.231722831726074,
+      "learning_rate": 2.8663314840499307e-05,
+      "loss": 1.8086,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.09852123260498,
+      "learning_rate": 2.8658113730929265e-05,
+      "loss": 1.2227,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.120312690734863,
+      "learning_rate": 2.8652912621359224e-05,
+      "loss": 1.6057,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.849720001220703,
+      "learning_rate": 2.8647711511789182e-05,
+      "loss": 1.5508,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.763028144836426,
+      "learning_rate": 2.864251040221914e-05,
+      "loss": 1.4658,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.62955093383789,
+      "learning_rate": 2.86373092926491e-05,
+      "loss": 1.6115,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.502273559570312,
+      "learning_rate": 2.8632108183079055e-05,
+      "loss": 1.6352,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.01419448852539,
+      "learning_rate": 2.8626907073509017e-05,
+      "loss": 1.4374,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.824151992797852,
+      "learning_rate": 2.8621705963938972e-05,
+      "loss": 1.6391,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.038396835327148,
+      "learning_rate": 2.8616504854368934e-05,
+      "loss": 1.4313,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.165166854858398,
+      "learning_rate": 2.861130374479889e-05,
+      "loss": 1.6367,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.33757495880127,
+      "learning_rate": 2.860610263522885e-05,
+      "loss": 1.4531,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.644055366516113,
+      "learning_rate": 2.8600901525658807e-05,
+      "loss": 1.7592,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.630600929260254,
+      "learning_rate": 2.859570041608877e-05,
+      "loss": 1.4934,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.351295471191406,
+      "learning_rate": 2.8590499306518724e-05,
+      "loss": 1.5299,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.138675689697266,
+      "learning_rate": 2.8585298196948682e-05,
+      "loss": 1.4298,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.685686111450195,
+      "learning_rate": 2.858009708737864e-05,
+      "loss": 1.2862,
+      "step": 273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.845622062683105,
+      "learning_rate": 2.85748959778086e-05,
+      "loss": 1.9247,
+      "step": 274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.040325164794922,
+      "learning_rate": 2.8569694868238558e-05,
+      "loss": 1.3682,
+      "step": 275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.99345588684082,
+      "learning_rate": 2.8564493758668517e-05,
+      "loss": 1.3171,
+      "step": 276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.519624710083008,
+      "learning_rate": 2.8559292649098475e-05,
+      "loss": 1.4481,
+      "step": 277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.8758673667907715,
+      "learning_rate": 2.8554091539528434e-05,
+      "loss": 1.1836,
+      "step": 278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.770076751708984,
+      "learning_rate": 2.854889042995839e-05,
+      "loss": 1.3676,
+      "step": 279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.373793601989746,
+      "learning_rate": 2.854368932038835e-05,
+      "loss": 1.3345,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.5178861618042,
+      "learning_rate": 2.8538488210818306e-05,
+      "loss": 1.4434,
+      "step": 281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.02005672454834,
+      "learning_rate": 2.853328710124827e-05,
+      "loss": 1.7931,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.602951049804688,
+      "learning_rate": 2.8528085991678224e-05,
+      "loss": 1.2721,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.859600067138672,
+      "learning_rate": 2.8522884882108186e-05,
+      "loss": 1.6366,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.426437377929688,
+      "learning_rate": 2.851768377253814e-05,
+      "loss": 1.726,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.756291389465332,
+      "learning_rate": 2.8512482662968103e-05,
+      "loss": 1.4414,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.598320960998535,
+      "learning_rate": 2.8507281553398058e-05,
+      "loss": 1.4956,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.504106521606445,
+      "learning_rate": 2.8502080443828017e-05,
+      "loss": 1.9411,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.65526294708252,
+      "learning_rate": 2.8496879334257975e-05,
+      "loss": 1.708,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.396827697753906,
+      "learning_rate": 2.8491678224687934e-05,
+      "loss": 1.6304,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.522106170654297,
+      "learning_rate": 2.8486477115117893e-05,
+      "loss": 1.5055,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.690590858459473,
+      "learning_rate": 2.848127600554785e-05,
+      "loss": 1.5796,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.430390357971191,
+      "learning_rate": 2.847607489597781e-05,
+      "loss": 1.5833,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.3465657234191895,
+      "learning_rate": 2.847087378640777e-05,
+      "loss": 1.2282,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.93099308013916,
+      "learning_rate": 2.8465672676837724e-05,
+      "loss": 1.5277,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.543130874633789,
+      "learning_rate": 2.8460471567267686e-05,
+      "loss": 1.3845,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.719395637512207,
+      "learning_rate": 2.845527045769764e-05,
+      "loss": 1.3802,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.534639358520508,
+      "learning_rate": 2.8450069348127603e-05,
+      "loss": 1.3313,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.372000694274902,
+      "learning_rate": 2.8444868238557558e-05,
+      "loss": 1.8624,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.28885269165039,
+      "learning_rate": 2.843966712898752e-05,
+      "loss": 1.5539,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.065950393676758,
+      "learning_rate": 2.8434466019417475e-05,
+      "loss": 1.6695,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.871010780334473,
+      "learning_rate": 2.8429264909847434e-05,
+      "loss": 1.3748,
+      "step": 302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.788363456726074,
+      "learning_rate": 2.8424063800277393e-05,
+      "loss": 1.592,
+      "step": 303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.655359268188477,
+      "learning_rate": 2.841886269070735e-05,
+      "loss": 1.5082,
+      "step": 304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.832733154296875,
+      "learning_rate": 2.841366158113731e-05,
+      "loss": 1.4371,
+      "step": 305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.828662872314453,
+      "learning_rate": 2.840846047156727e-05,
+      "loss": 1.5956,
+      "step": 306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.4097261428833,
+      "learning_rate": 2.8403259361997227e-05,
+      "loss": 1.6857,
+      "step": 307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.562804222106934,
+      "learning_rate": 2.8398058252427186e-05,
+      "loss": 1.3949,
+      "step": 308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.861429691314697,
+      "learning_rate": 2.8392857142857144e-05,
+      "loss": 1.1801,
+      "step": 309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.197989463806152,
+      "learning_rate": 2.8387656033287103e-05,
+      "loss": 1.1885,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.123920440673828,
+      "learning_rate": 2.8382454923717058e-05,
+      "loss": 1.4692,
+      "step": 311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.70377254486084,
+      "learning_rate": 2.837725381414702e-05,
+      "loss": 1.2695,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.368098258972168,
+      "learning_rate": 2.8372052704576975e-05,
+      "loss": 1.5174,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.075979232788086,
+      "learning_rate": 2.8366851595006937e-05,
+      "loss": 1.3776,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.848196983337402,
+      "learning_rate": 2.8361650485436893e-05,
+      "loss": 1.3274,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.091851234436035,
+      "learning_rate": 2.8356449375866855e-05,
+      "loss": 1.0592,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.959482192993164,
+      "learning_rate": 2.835124826629681e-05,
+      "loss": 1.483,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.618368148803711,
+      "learning_rate": 2.834604715672677e-05,
+      "loss": 1.5094,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.955796241760254,
+      "learning_rate": 2.8340846047156727e-05,
+      "loss": 1.5875,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.122024536132812,
+      "learning_rate": 2.8335644937586686e-05,
+      "loss": 1.4752,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.10103702545166,
+      "learning_rate": 2.8330443828016644e-05,
+      "loss": 1.8409,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.0626859664917,
+      "learning_rate": 2.8325242718446603e-05,
+      "loss": 1.8296,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.014315605163574,
+      "learning_rate": 2.832004160887656e-05,
+      "loss": 1.5961,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.897581577301025,
+      "learning_rate": 2.831484049930652e-05,
+      "loss": 1.4184,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.454628944396973,
+      "learning_rate": 2.8309639389736475e-05,
+      "loss": 1.4505,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.656554222106934,
+      "learning_rate": 2.8304438280166437e-05,
+      "loss": 1.435,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.921539783477783,
+      "learning_rate": 2.8299237170596392e-05,
+      "loss": 1.4475,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.29591178894043,
+      "learning_rate": 2.8294036061026354e-05,
+      "loss": 1.2657,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.965700149536133,
+      "learning_rate": 2.828883495145631e-05,
+      "loss": 1.589,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.564988136291504,
+      "learning_rate": 2.8283633841886272e-05,
+      "loss": 1.3063,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.545437812805176,
+      "learning_rate": 2.8278432732316227e-05,
+      "loss": 1.6668,
+      "step": 331
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.251893043518066,
+      "learning_rate": 2.827323162274619e-05,
+      "loss": 1.147,
+      "step": 332
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.108394622802734,
+      "learning_rate": 2.8268030513176144e-05,
+      "loss": 1.3457,
+      "step": 333
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.6668119430542,
+      "learning_rate": 2.8262829403606103e-05,
+      "loss": 1.851,
+      "step": 334
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.64436149597168,
+      "learning_rate": 2.825762829403606e-05,
+      "loss": 1.5591,
+      "step": 335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.218013763427734,
+      "learning_rate": 2.825242718446602e-05,
+      "loss": 1.2768,
+      "step": 336
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.935311317443848,
+      "learning_rate": 2.824722607489598e-05,
+      "loss": 1.3413,
+      "step": 337
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.266963005065918,
+      "learning_rate": 2.8242024965325937e-05,
+      "loss": 1.4708,
+      "step": 338
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.789060592651367,
+      "learning_rate": 2.8236823855755896e-05,
+      "loss": 1.5046,
+      "step": 339
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.878279685974121,
+      "learning_rate": 2.8231622746185854e-05,
+      "loss": 1.368,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.320792198181152,
+      "learning_rate": 2.822642163661581e-05,
+      "loss": 1.1281,
+      "step": 341
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.46088981628418,
+      "learning_rate": 2.822122052704577e-05,
+      "loss": 1.449,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.5416898727417,
+      "learning_rate": 2.8216019417475727e-05,
+      "loss": 1.4967,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.307211875915527,
+      "learning_rate": 2.821081830790569e-05,
+      "loss": 1.2682,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.439652442932129,
+      "learning_rate": 2.8205617198335644e-05,
+      "loss": 1.5236,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.725629806518555,
+      "learning_rate": 2.8200416088765606e-05,
+      "loss": 1.4234,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.212013244628906,
+      "learning_rate": 2.819521497919556e-05,
+      "loss": 1.6286,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.123527526855469,
+      "learning_rate": 2.819001386962552e-05,
+      "loss": 1.2837,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.105483055114746,
+      "learning_rate": 2.818481276005548e-05,
+      "loss": 1.2424,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.891584396362305,
+      "learning_rate": 2.8179611650485437e-05,
+      "loss": 1.4202,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.434996604919434,
+      "learning_rate": 2.8174410540915396e-05,
+      "loss": 1.487,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.164681434631348,
+      "learning_rate": 2.8169209431345354e-05,
+      "loss": 1.3877,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.201833724975586,
+      "learning_rate": 2.8164008321775313e-05,
+      "loss": 1.6707,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.433565139770508,
+      "learning_rate": 2.815880721220527e-05,
+      "loss": 1.278,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.772685050964355,
+      "learning_rate": 2.815360610263523e-05,
+      "loss": 1.3358,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.333212852478027,
+      "learning_rate": 2.814840499306519e-05,
+      "loss": 1.4233,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.82320499420166,
+      "learning_rate": 2.8143203883495144e-05,
+      "loss": 1.3781,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.654339790344238,
+      "learning_rate": 2.8138002773925106e-05,
+      "loss": 1.5566,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.45020866394043,
+      "learning_rate": 2.813280166435506e-05,
+      "loss": 1.3737,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.881738662719727,
+      "learning_rate": 2.8127600554785023e-05,
+      "loss": 1.4644,
+      "step": 360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.509510040283203,
+      "learning_rate": 2.812239944521498e-05,
+      "loss": 1.4318,
+      "step": 361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.092405319213867,
+      "learning_rate": 2.811719833564494e-05,
+      "loss": 1.6218,
+      "step": 362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.22721004486084,
+      "learning_rate": 2.8111997226074896e-05,
+      "loss": 1.2614,
+      "step": 363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.05213737487793,
+      "learning_rate": 2.8106796116504854e-05,
+      "loss": 1.3824,
+      "step": 364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.391263008117676,
+      "learning_rate": 2.8101595006934813e-05,
+      "loss": 1.9592,
+      "step": 365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.291427612304688,
+      "learning_rate": 2.809639389736477e-05,
+      "loss": 1.2983,
+      "step": 366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.029029846191406,
+      "learning_rate": 2.809119278779473e-05,
+      "loss": 1.5213,
+      "step": 367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.243372917175293,
+      "learning_rate": 2.808599167822469e-05,
+      "loss": 1.0608,
+      "step": 368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.516368865966797,
+      "learning_rate": 2.8080790568654647e-05,
+      "loss": 1.6242,
+      "step": 369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.330872535705566,
+      "learning_rate": 2.8075589459084606e-05,
+      "loss": 1.2902,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.240985870361328,
+      "learning_rate": 2.8070388349514565e-05,
+      "loss": 1.5181,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.059120178222656,
+      "learning_rate": 2.8065187239944523e-05,
+      "loss": 1.7363,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.24335765838623,
+      "learning_rate": 2.805998613037448e-05,
+      "loss": 1.5289,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.84642505645752,
+      "learning_rate": 2.805478502080444e-05,
+      "loss": 1.3953,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.259860038757324,
+      "learning_rate": 2.8049583911234396e-05,
+      "loss": 1.2778,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.002691268920898,
+      "learning_rate": 2.8044382801664358e-05,
+      "loss": 1.2652,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.594935417175293,
+      "learning_rate": 2.8039181692094313e-05,
+      "loss": 1.6165,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.01926040649414,
+      "learning_rate": 2.8033980582524275e-05,
+      "loss": 1.3675,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.642001152038574,
+      "learning_rate": 2.802877947295423e-05,
+      "loss": 1.3396,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.012113571166992,
+      "learning_rate": 2.802357836338419e-05,
+      "loss": 1.6095,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.416895866394043,
+      "learning_rate": 2.8018377253814147e-05,
+      "loss": 1.4541,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.769719123840332,
+      "learning_rate": 2.8013176144244106e-05,
+      "loss": 1.4955,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.049176216125488,
+      "learning_rate": 2.8007975034674065e-05,
+      "loss": 1.3368,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.002891540527344,
+      "learning_rate": 2.8002773925104023e-05,
+      "loss": 1.6374,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.367919921875,
+      "learning_rate": 2.7997572815533982e-05,
+      "loss": 1.2031,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.265168190002441,
+      "learning_rate": 2.799237170596394e-05,
+      "loss": 1.1006,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.07726764678955,
+      "learning_rate": 2.7987170596393896e-05,
+      "loss": 1.2594,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.769450187683105,
+      "learning_rate": 2.7981969486823858e-05,
+      "loss": 1.3811,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.588899612426758,
+      "learning_rate": 2.7976768377253813e-05,
+      "loss": 1.5496,
+      "step": 389
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.344609260559082,
+      "learning_rate": 2.7971567267683775e-05,
+      "loss": 1.5366,
+      "step": 390
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.274745464324951,
+      "learning_rate": 2.796636615811373e-05,
+      "loss": 1.0506,
+      "step": 391
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.099116325378418,
+      "learning_rate": 2.7961165048543692e-05,
+      "loss": 1.494,
+      "step": 392
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.625064849853516,
+      "learning_rate": 2.7955963938973647e-05,
+      "loss": 1.4747,
+      "step": 393
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.768614768981934,
+      "learning_rate": 2.795076282940361e-05,
+      "loss": 1.6713,
+      "step": 394
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.399869918823242,
+      "learning_rate": 2.7945561719833565e-05,
+      "loss": 1.165,
+      "step": 395
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.504380226135254,
+      "learning_rate": 2.7940360610263523e-05,
+      "loss": 1.4972,
+      "step": 396
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.888386726379395,
+      "learning_rate": 2.7935159500693482e-05,
+      "loss": 1.539,
+      "step": 397
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.923397064208984,
+      "learning_rate": 2.792995839112344e-05,
+      "loss": 1.3147,
+      "step": 398
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.621848106384277,
+      "learning_rate": 2.79247572815534e-05,
+      "loss": 1.5109,
+      "step": 399
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.210451126098633,
+      "learning_rate": 2.7919556171983358e-05,
+      "loss": 1.64,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.4457426071167,
+      "learning_rate": 2.7914355062413316e-05,
+      "loss": 1.2854,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.9414701461792,
+      "learning_rate": 2.7909153952843275e-05,
+      "loss": 1.3474,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.315629005432129,
+      "learning_rate": 2.790395284327323e-05,
+      "loss": 1.1314,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.901406288146973,
+      "learning_rate": 2.7898751733703192e-05,
+      "loss": 1.3633,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.315669059753418,
+      "learning_rate": 2.7893550624133147e-05,
+      "loss": 1.4386,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.166473388671875,
+      "learning_rate": 2.788834951456311e-05,
+      "loss": 1.2566,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.371944427490234,
+      "learning_rate": 2.7883148404993065e-05,
+      "loss": 1.3418,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.535171508789062,
+      "learning_rate": 2.7877947295423027e-05,
+      "loss": 1.1223,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.04525089263916,
+      "learning_rate": 2.7872746185852982e-05,
+      "loss": 1.5583,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.082500457763672,
+      "learning_rate": 2.786754507628294e-05,
+      "loss": 0.9832,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.991056442260742,
+      "learning_rate": 2.78623439667129e-05,
+      "loss": 1.3849,
+      "step": 411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.799002647399902,
+      "learning_rate": 2.7857142857142858e-05,
+      "loss": 1.5725,
+      "step": 412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.829028129577637,
+      "learning_rate": 2.7851941747572816e-05,
+      "loss": 1.676,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.930892944335938,
+      "learning_rate": 2.7846740638002775e-05,
+      "loss": 1.3746,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.203141212463379,
+      "learning_rate": 2.7841539528432733e-05,
+      "loss": 1.4066,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.27408504486084,
+      "learning_rate": 2.7836338418862692e-05,
+      "loss": 1.415,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.122041702270508,
+      "learning_rate": 2.783113730929265e-05,
+      "loss": 1.3087,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.356685638427734,
+      "learning_rate": 2.782593619972261e-05,
+      "loss": 1.2503,
+      "step": 418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.67627239227295,
+      "learning_rate": 2.7820735090152565e-05,
+      "loss": 1.3564,
+      "step": 419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.199246406555176,
+      "learning_rate": 2.7815533980582527e-05,
+      "loss": 1.1028,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.373217582702637,
+      "learning_rate": 2.7810332871012482e-05,
+      "loss": 1.6462,
+      "step": 421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.345012664794922,
+      "learning_rate": 2.7805131761442444e-05,
+      "loss": 1.3382,
+      "step": 422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.964041709899902,
+      "learning_rate": 2.77999306518724e-05,
+      "loss": 1.419,
+      "step": 423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.264195442199707,
+      "learning_rate": 2.779472954230236e-05,
+      "loss": 1.5548,
+      "step": 424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.324397087097168,
+      "learning_rate": 2.7789528432732316e-05,
+      "loss": 1.3416,
+      "step": 425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.711651802062988,
+      "learning_rate": 2.7784327323162275e-05,
+      "loss": 1.2265,
+      "step": 426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.732048988342285,
+      "learning_rate": 2.7779126213592233e-05,
+      "loss": 1.4387,
+      "step": 427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.969171047210693,
+      "learning_rate": 2.7773925104022192e-05,
+      "loss": 1.3048,
+      "step": 428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.845348358154297,
+      "learning_rate": 2.776872399445215e-05,
+      "loss": 1.2735,
+      "step": 429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.570350646972656,
+      "learning_rate": 2.776352288488211e-05,
+      "loss": 1.298,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.428897857666016,
+      "learning_rate": 2.7758321775312068e-05,
+      "loss": 1.2991,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.16434097290039,
+      "learning_rate": 2.7753120665742026e-05,
+      "loss": 1.4366,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.72012996673584,
+      "learning_rate": 2.7747919556171985e-05,
+      "loss": 1.3194,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.145621299743652,
+      "learning_rate": 2.7742718446601944e-05,
+      "loss": 1.203,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.799650192260742,
+      "learning_rate": 2.77375173370319e-05,
+      "loss": 1.3634,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.241556167602539,
+      "learning_rate": 2.773231622746186e-05,
+      "loss": 1.3974,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.69413948059082,
+      "learning_rate": 2.7727115117891816e-05,
+      "loss": 1.2178,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.726613998413086,
+      "learning_rate": 2.7721914008321778e-05,
+      "loss": 1.1438,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.58704948425293,
+      "learning_rate": 2.7716712898751733e-05,
+      "loss": 1.2962,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.85359001159668,
+      "learning_rate": 2.7711511789181695e-05,
+      "loss": 1.5944,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.668953895568848,
+      "learning_rate": 2.770631067961165e-05,
+      "loss": 1.2699,
+      "step": 441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.348400115966797,
+      "learning_rate": 2.770110957004161e-05,
+      "loss": 1.2338,
+      "step": 442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.16502857208252,
+      "learning_rate": 2.7695908460471568e-05,
+      "loss": 1.2942,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.099385261535645,
+      "learning_rate": 2.7690707350901526e-05,
+      "loss": 1.4804,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.641454696655273,
+      "learning_rate": 2.7685506241331485e-05,
+      "loss": 1.5923,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.569912910461426,
+      "learning_rate": 2.7680305131761444e-05,
+      "loss": 1.3244,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.636981010437012,
+      "learning_rate": 2.7675104022191402e-05,
+      "loss": 1.2302,
+      "step": 447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.936580657958984,
+      "learning_rate": 2.766990291262136e-05,
+      "loss": 1.0607,
+      "step": 448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.924996376037598,
+      "learning_rate": 2.7664701803051316e-05,
+      "loss": 1.3845,
+      "step": 449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.043442726135254,
+      "learning_rate": 2.7659500693481278e-05,
+      "loss": 1.3618,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.751946449279785,
+      "learning_rate": 2.7654299583911233e-05,
+      "loss": 1.2871,
+      "step": 451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.493877410888672,
+      "learning_rate": 2.7649098474341195e-05,
+      "loss": 1.3445,
+      "step": 452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.293668746948242,
+      "learning_rate": 2.764389736477115e-05,
+      "loss": 1.5492,
+      "step": 453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.425002098083496,
+      "learning_rate": 2.7638696255201113e-05,
+      "loss": 1.5325,
+      "step": 454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.7320556640625,
+      "learning_rate": 2.7633495145631068e-05,
+      "loss": 1.2074,
+      "step": 455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.757223129272461,
+      "learning_rate": 2.762829403606103e-05,
+      "loss": 1.4239,
+      "step": 456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.895047187805176,
+      "learning_rate": 2.7623092926490985e-05,
+      "loss": 1.2738,
+      "step": 457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.04129695892334,
+      "learning_rate": 2.7617891816920944e-05,
+      "loss": 1.4542,
+      "step": 458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.399195671081543,
+      "learning_rate": 2.7612690707350902e-05,
+      "loss": 1.2419,
+      "step": 459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.854976654052734,
+      "learning_rate": 2.760748959778086e-05,
+      "loss": 1.1354,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.137279510498047,
+      "learning_rate": 2.760228848821082e-05,
+      "loss": 1.3733,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.829804420471191,
+      "learning_rate": 2.7597087378640778e-05,
+      "loss": 1.2878,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.23508071899414,
+      "learning_rate": 2.7591886269070737e-05,
+      "loss": 1.4659,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.140019416809082,
+      "learning_rate": 2.7586685159500695e-05,
+      "loss": 1.3147,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.00510311126709,
+      "learning_rate": 2.758148404993065e-05,
+      "loss": 1.4936,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.923553466796875,
+      "learning_rate": 2.7576282940360613e-05,
+      "loss": 1.2378,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.306476593017578,
+      "learning_rate": 2.7571081830790568e-05,
+      "loss": 1.2058,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.393396377563477,
+      "learning_rate": 2.756588072122053e-05,
+      "loss": 1.1767,
+      "step": 468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.396347999572754,
+      "learning_rate": 2.7560679611650485e-05,
+      "loss": 1.2193,
+      "step": 469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.10958480834961,
+      "learning_rate": 2.7555478502080447e-05,
+      "loss": 1.2058,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.978511810302734,
+      "learning_rate": 2.7550277392510402e-05,
+      "loss": 1.2157,
+      "step": 471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.409880638122559,
+      "learning_rate": 2.754507628294036e-05,
+      "loss": 1.336,
+      "step": 472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.793098449707031,
+      "learning_rate": 2.753987517337032e-05,
+      "loss": 1.5577,
+      "step": 473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.230104446411133,
+      "learning_rate": 2.7534674063800278e-05,
+      "loss": 1.1817,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.266423225402832,
+      "learning_rate": 2.7529472954230237e-05,
+      "loss": 1.0505,
+      "step": 475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.720837593078613,
+      "learning_rate": 2.7524271844660195e-05,
+      "loss": 1.3369,
+      "step": 476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.759696006774902,
+      "learning_rate": 2.7519070735090154e-05,
+      "loss": 1.1522,
+      "step": 477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.163063049316406,
+      "learning_rate": 2.7513869625520112e-05,
+      "loss": 0.9939,
+      "step": 478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.125991821289062,
+      "learning_rate": 2.750866851595007e-05,
+      "loss": 1.245,
+      "step": 479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.157820701599121,
+      "learning_rate": 2.750346740638003e-05,
+      "loss": 1.3198,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.371237754821777,
+      "learning_rate": 2.7498266296809985e-05,
+      "loss": 1.2423,
+      "step": 481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.840733528137207,
+      "learning_rate": 2.7493065187239947e-05,
+      "loss": 1.313,
+      "step": 482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.888921737670898,
+      "learning_rate": 2.7487864077669902e-05,
+      "loss": 1.3123,
+      "step": 483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.84160041809082,
+      "learning_rate": 2.7482662968099864e-05,
+      "loss": 1.452,
+      "step": 484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.27896785736084,
+      "learning_rate": 2.747746185852982e-05,
+      "loss": 1.4633,
+      "step": 485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.996109008789062,
+      "learning_rate": 2.747226074895978e-05,
+      "loss": 1.5629,
+      "step": 486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.334553718566895,
+      "learning_rate": 2.7467059639389737e-05,
+      "loss": 1.2917,
+      "step": 487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.613493919372559,
+      "learning_rate": 2.7461858529819695e-05,
+      "loss": 1.5451,
+      "step": 488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.705168724060059,
+      "learning_rate": 2.7456657420249654e-05,
+      "loss": 1.3911,
+      "step": 489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.480560779571533,
+      "learning_rate": 2.7451456310679612e-05,
+      "loss": 1.3897,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.389557838439941,
+      "learning_rate": 2.744625520110957e-05,
+      "loss": 1.4274,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.005056381225586,
+      "learning_rate": 2.744105409153953e-05,
+      "loss": 1.3592,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.160528182983398,
+      "learning_rate": 2.7435852981969488e-05,
+      "loss": 1.4221,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.53385066986084,
+      "learning_rate": 2.7430651872399447e-05,
+      "loss": 1.3141,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.193487167358398,
+      "learning_rate": 2.7425450762829402e-05,
+      "loss": 1.5571,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.185532569885254,
+      "learning_rate": 2.7420249653259364e-05,
+      "loss": 1.3829,
+      "step": 496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.980709552764893,
+      "learning_rate": 2.741504854368932e-05,
+      "loss": 1.4038,
+      "step": 497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.474987983703613,
+      "learning_rate": 2.740984743411928e-05,
+      "loss": 1.0118,
+      "step": 498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.607361793518066,
+      "learning_rate": 2.7404646324549237e-05,
+      "loss": 1.2332,
+      "step": 499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.5491943359375,
+      "learning_rate": 2.73994452149792e-05,
+      "loss": 1.2935,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.907444953918457,
+      "learning_rate": 2.7394244105409154e-05,
+      "loss": 1.541,
+      "step": 501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.223031997680664,
+      "learning_rate": 2.7389042995839116e-05,
+      "loss": 1.3943,
+      "step": 502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.930360794067383,
+      "learning_rate": 2.738384188626907e-05,
+      "loss": 1.6929,
+      "step": 503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.160101890563965,
+      "learning_rate": 2.737864077669903e-05,
+      "loss": 1.4027,
+      "step": 504
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.699889183044434,
+      "learning_rate": 2.7373439667128988e-05,
+      "loss": 1.4054,
+      "step": 505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.229127883911133,
+      "learning_rate": 2.7368238557558947e-05,
+      "loss": 1.285,
+      "step": 506
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.459737777709961,
+      "learning_rate": 2.7363037447988905e-05,
+      "loss": 1.3263,
+      "step": 507
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.814366340637207,
+      "learning_rate": 2.7357836338418864e-05,
+      "loss": 1.0304,
+      "step": 508
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.895322799682617,
+      "learning_rate": 2.7352635228848823e-05,
+      "loss": 1.2854,
+      "step": 509
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.840097427368164,
+      "learning_rate": 2.734743411927878e-05,
+      "loss": 1.2094,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.275677680969238,
+      "learning_rate": 2.7342233009708737e-05,
+      "loss": 1.2268,
+      "step": 511
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.84443473815918,
+      "learning_rate": 2.73370319001387e-05,
+      "loss": 1.0413,
+      "step": 512
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.269462585449219,
+      "learning_rate": 2.7331830790568654e-05,
+      "loss": 1.1883,
+      "step": 513
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.3368501663208,
+      "learning_rate": 2.7326629680998616e-05,
+      "loss": 1.3685,
+      "step": 514
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.128835678100586,
+      "learning_rate": 2.732142857142857e-05,
+      "loss": 1.181,
+      "step": 515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.23559284210205,
+      "learning_rate": 2.7316227461858533e-05,
+      "loss": 1.5048,
+      "step": 516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.153850555419922,
+      "learning_rate": 2.7311026352288488e-05,
+      "loss": 1.295,
+      "step": 517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.497044563293457,
+      "learning_rate": 2.7305825242718447e-05,
+      "loss": 1.2798,
+      "step": 518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.901271343231201,
+      "learning_rate": 2.7300624133148405e-05,
+      "loss": 1.0506,
+      "step": 519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.656876564025879,
+      "learning_rate": 2.7295423023578364e-05,
+      "loss": 1.3849,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.55753231048584,
+      "learning_rate": 2.7290221914008323e-05,
+      "loss": 0.9221,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.15007495880127,
+      "learning_rate": 2.728502080443828e-05,
+      "loss": 1.4006,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.69929313659668,
+      "learning_rate": 2.727981969486824e-05,
+      "loss": 1.4559,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.522904396057129,
+      "learning_rate": 2.72746185852982e-05,
+      "loss": 1.2029,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.01738166809082,
+      "learning_rate": 2.7269417475728157e-05,
+      "loss": 1.2296,
+      "step": 525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.683744430541992,
+      "learning_rate": 2.7264216366158116e-05,
+      "loss": 1.177,
+      "step": 526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.598075866699219,
+      "learning_rate": 2.725901525658807e-05,
+      "loss": 1.4221,
+      "step": 527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.345980644226074,
+      "learning_rate": 2.7253814147018033e-05,
+      "loss": 1.3842,
+      "step": 528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.505378723144531,
+      "learning_rate": 2.7248613037447988e-05,
+      "loss": 1.2198,
+      "step": 529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.662880897521973,
+      "learning_rate": 2.724341192787795e-05,
+      "loss": 1.4686,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.8166327476501465,
+      "learning_rate": 2.7238210818307905e-05,
+      "loss": 1.2902,
+      "step": 531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.421590805053711,
+      "learning_rate": 2.7233009708737867e-05,
+      "loss": 1.294,
+      "step": 532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.587312698364258,
+      "learning_rate": 2.7227808599167823e-05,
+      "loss": 1.203,
+      "step": 533
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.040580749511719,
+      "learning_rate": 2.722260748959778e-05,
+      "loss": 1.2855,
+      "step": 534
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.138426780700684,
+      "learning_rate": 2.721740638002774e-05,
+      "loss": 1.1934,
+      "step": 535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.222373008728027,
+      "learning_rate": 2.72122052704577e-05,
+      "loss": 1.2733,
+      "step": 536
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.592804908752441,
+      "learning_rate": 2.7207004160887657e-05,
+      "loss": 1.5383,
+      "step": 537
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.687430381774902,
+      "learning_rate": 2.7201803051317616e-05,
+      "loss": 1.0374,
+      "step": 538
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.278436660766602,
+      "learning_rate": 2.7196601941747574e-05,
+      "loss": 1.0368,
+      "step": 539
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.43661642074585,
+      "learning_rate": 2.7191400832177533e-05,
+      "loss": 1.0086,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.632880210876465,
+      "learning_rate": 2.718619972260749e-05,
+      "loss": 1.5453,
+      "step": 541
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.965327262878418,
+      "learning_rate": 2.718099861303745e-05,
+      "loss": 1.331,
+      "step": 542
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.598584175109863,
+      "learning_rate": 2.7175797503467405e-05,
+      "loss": 1.1553,
+      "step": 543
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.2766695022583,
+      "learning_rate": 2.7170596393897367e-05,
+      "loss": 1.1868,
+      "step": 544
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.193568229675293,
+      "learning_rate": 2.7165395284327323e-05,
+      "loss": 1.5067,
+      "step": 545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.209564208984375,
+      "learning_rate": 2.7160194174757285e-05,
+      "loss": 1.125,
+      "step": 546
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.241795539855957,
+      "learning_rate": 2.715499306518724e-05,
+      "loss": 1.247,
+      "step": 547
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.474015235900879,
+      "learning_rate": 2.7149791955617202e-05,
+      "loss": 1.0642,
+      "step": 548
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.994909286499023,
+      "learning_rate": 2.7144590846047157e-05,
+      "loss": 1.2889,
+      "step": 549
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.14266586303711,
+      "learning_rate": 2.7139389736477116e-05,
+      "loss": 1.2044,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.511383056640625,
+      "learning_rate": 2.7134188626907074e-05,
+      "loss": 1.393,
+      "step": 551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.348587036132812,
+      "learning_rate": 2.7128987517337033e-05,
+      "loss": 1.6398,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.827128410339355,
+      "learning_rate": 2.712378640776699e-05,
+      "loss": 1.0723,
+      "step": 553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.655224800109863,
+      "learning_rate": 2.711858529819695e-05,
+      "loss": 0.9112,
+      "step": 554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.04931926727295,
+      "learning_rate": 2.711338418862691e-05,
+      "loss": 1.4055,
+      "step": 555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.419055938720703,
+      "learning_rate": 2.7108183079056867e-05,
+      "loss": 0.899,
+      "step": 556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.574268341064453,
+      "learning_rate": 2.7102981969486823e-05,
+      "loss": 1.5408,
+      "step": 557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.961968421936035,
+      "learning_rate": 2.7097780859916785e-05,
+      "loss": 1.6198,
+      "step": 558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.2930269241333,
+      "learning_rate": 2.709257975034674e-05,
+      "loss": 1.1143,
+      "step": 559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.619170188903809,
+      "learning_rate": 2.7087378640776702e-05,
+      "loss": 1.3459,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.323966979980469,
+      "learning_rate": 2.7082177531206657e-05,
+      "loss": 1.1484,
+      "step": 561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.887866020202637,
+      "learning_rate": 2.707697642163662e-05,
+      "loss": 1.2725,
+      "step": 562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.413854598999023,
+      "learning_rate": 2.7071775312066574e-05,
+      "loss": 1.3608,
+      "step": 563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.635536193847656,
+      "learning_rate": 2.7066574202496536e-05,
+      "loss": 1.3567,
+      "step": 564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.059237480163574,
+      "learning_rate": 2.706137309292649e-05,
+      "loss": 1.2776,
+      "step": 565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.049764633178711,
+      "learning_rate": 2.705617198335645e-05,
+      "loss": 1.2873,
+      "step": 566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.216239929199219,
+      "learning_rate": 2.705097087378641e-05,
+      "loss": 1.3649,
+      "step": 567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.243322372436523,
+      "learning_rate": 2.7045769764216367e-05,
+      "loss": 1.1851,
+      "step": 568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.128588676452637,
+      "learning_rate": 2.7040568654646326e-05,
+      "loss": 1.3678,
+      "step": 569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.71036148071289,
+      "learning_rate": 2.7035367545076284e-05,
+      "loss": 1.2886,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.009879112243652,
+      "learning_rate": 2.7030166435506243e-05,
+      "loss": 1.481,
+      "step": 571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.493657112121582,
+      "learning_rate": 2.7024965325936202e-05,
+      "loss": 1.3071,
+      "step": 572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.145187377929688,
+      "learning_rate": 2.7019764216366157e-05,
+      "loss": 1.1801,
+      "step": 573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.753716468811035,
+      "learning_rate": 2.701456310679612e-05,
+      "loss": 1.4391,
+      "step": 574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.463996887207031,
+      "learning_rate": 2.7009361997226074e-05,
+      "loss": 1.2212,
+      "step": 575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.976245880126953,
+      "learning_rate": 2.7004160887656036e-05,
+      "loss": 1.3406,
+      "step": 576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.606514930725098,
+      "learning_rate": 2.699895977808599e-05,
+      "loss": 1.2241,
+      "step": 577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.8770833015441895,
+      "learning_rate": 2.6993758668515953e-05,
+      "loss": 1.2236,
+      "step": 578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.138198852539062,
+      "learning_rate": 2.698855755894591e-05,
+      "loss": 1.4478,
+      "step": 579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.556624412536621,
+      "learning_rate": 2.6983356449375867e-05,
+      "loss": 1.4338,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.021261215209961,
+      "learning_rate": 2.6978155339805826e-05,
+      "loss": 1.3438,
+      "step": 581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.09264087677002,
+      "learning_rate": 2.6972954230235784e-05,
+      "loss": 1.5377,
+      "step": 582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.825350761413574,
+      "learning_rate": 2.6967753120665743e-05,
+      "loss": 1.3647,
+      "step": 583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.765951156616211,
+      "learning_rate": 2.69625520110957e-05,
+      "loss": 1.1526,
+      "step": 584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.564873695373535,
+      "learning_rate": 2.695735090152566e-05,
+      "loss": 1.2825,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.675870895385742,
+      "learning_rate": 2.695214979195562e-05,
+      "loss": 1.307,
+      "step": 586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.908300399780273,
+      "learning_rate": 2.6946948682385578e-05,
+      "loss": 1.6178,
+      "step": 587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.084193229675293,
+      "learning_rate": 2.6941747572815536e-05,
+      "loss": 1.231,
+      "step": 588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.443767547607422,
+      "learning_rate": 2.693654646324549e-05,
+      "loss": 1.3709,
+      "step": 589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.63538932800293,
+      "learning_rate": 2.6931345353675453e-05,
+      "loss": 1.4566,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.425724506378174,
+      "learning_rate": 2.692614424410541e-05,
+      "loss": 1.1031,
+      "step": 591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.775505065917969,
+      "learning_rate": 2.692094313453537e-05,
+      "loss": 1.1987,
+      "step": 592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.558345794677734,
+      "learning_rate": 2.6915742024965326e-05,
+      "loss": 1.1807,
+      "step": 593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.071976661682129,
+      "learning_rate": 2.6910540915395288e-05,
+      "loss": 1.6044,
+      "step": 594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.29594612121582,
+      "learning_rate": 2.6905339805825243e-05,
+      "loss": 1.4283,
+      "step": 595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.983234405517578,
+      "learning_rate": 2.69001386962552e-05,
+      "loss": 1.2675,
+      "step": 596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.873207092285156,
+      "learning_rate": 2.689493758668516e-05,
+      "loss": 1.4138,
+      "step": 597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.375691413879395,
+      "learning_rate": 2.688973647711512e-05,
+      "loss": 1.2068,
+      "step": 598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.934732437133789,
+      "learning_rate": 2.6884535367545077e-05,
+      "loss": 1.2004,
+      "step": 599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.87802505493164,
+      "learning_rate": 2.6879334257975036e-05,
+      "loss": 1.0992,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.95258903503418,
+      "learning_rate": 2.6874133148404995e-05,
+      "loss": 1.3087,
+      "step": 601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.457464218139648,
+      "learning_rate": 2.6868932038834953e-05,
+      "loss": 1.0491,
+      "step": 602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.138510704040527,
+      "learning_rate": 2.6863730929264912e-05,
+      "loss": 1.3888,
+      "step": 603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.70171070098877,
+      "learning_rate": 2.685852981969487e-05,
+      "loss": 1.4489,
+      "step": 604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.099088668823242,
+      "learning_rate": 2.6853328710124826e-05,
+      "loss": 1.451,
+      "step": 605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.69382381439209,
+      "learning_rate": 2.6848127600554788e-05,
+      "loss": 1.4559,
+      "step": 606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.570367813110352,
+      "learning_rate": 2.6842926490984743e-05,
+      "loss": 1.4438,
+      "step": 607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.897492408752441,
+      "learning_rate": 2.6837725381414705e-05,
+      "loss": 1.2942,
+      "step": 608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.69155216217041,
+      "learning_rate": 2.683252427184466e-05,
+      "loss": 1.4265,
+      "step": 609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.02287483215332,
+      "learning_rate": 2.6827323162274622e-05,
+      "loss": 1.2375,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.613814353942871,
+      "learning_rate": 2.6822122052704577e-05,
+      "loss": 1.1142,
+      "step": 611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.46070671081543,
+      "learning_rate": 2.6816920943134536e-05,
+      "loss": 1.2436,
+      "step": 612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.500393867492676,
+      "learning_rate": 2.6811719833564495e-05,
+      "loss": 1.3554,
+      "step": 613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.409038543701172,
+      "learning_rate": 2.6806518723994453e-05,
+      "loss": 1.3417,
+      "step": 614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.760284423828125,
+      "learning_rate": 2.6801317614424412e-05,
+      "loss": 1.1129,
+      "step": 615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.49120044708252,
+      "learning_rate": 2.679611650485437e-05,
+      "loss": 1.2865,
+      "step": 616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.75191879272461,
+      "learning_rate": 2.679091539528433e-05,
+      "loss": 1.2648,
+      "step": 617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.837239265441895,
+      "learning_rate": 2.6785714285714288e-05,
+      "loss": 1.3845,
+      "step": 618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.587468147277832,
+      "learning_rate": 2.6780513176144243e-05,
+      "loss": 1.1051,
+      "step": 619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.658740997314453,
+      "learning_rate": 2.6775312066574205e-05,
+      "loss": 1.2333,
+      "step": 620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.361716270446777,
+      "learning_rate": 2.677011095700416e-05,
+      "loss": 1.0254,
+      "step": 621
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.653465270996094,
+      "learning_rate": 2.6764909847434122e-05,
+      "loss": 1.4137,
+      "step": 622
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.516178131103516,
+      "learning_rate": 2.6759708737864077e-05,
+      "loss": 1.1567,
+      "step": 623
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.31721305847168,
+      "learning_rate": 2.675450762829404e-05,
+      "loss": 1.373,
+      "step": 624
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.935283660888672,
+      "learning_rate": 2.6749306518723995e-05,
+      "loss": 1.3011,
+      "step": 625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.537062644958496,
+      "learning_rate": 2.6744105409153957e-05,
+      "loss": 1.2059,
+      "step": 626
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.03139877319336,
+      "learning_rate": 2.6738904299583912e-05,
+      "loss": 1.0379,
+      "step": 627
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.869452476501465,
+      "learning_rate": 2.673370319001387e-05,
+      "loss": 1.1409,
+      "step": 628
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.701508522033691,
+      "learning_rate": 2.672850208044383e-05,
+      "loss": 1.3377,
+      "step": 629
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.825364112854004,
+      "learning_rate": 2.6723300970873788e-05,
+      "loss": 1.3431,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.420677185058594,
+      "learning_rate": 2.6718099861303746e-05,
+      "loss": 1.1427,
+      "step": 631
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.031339645385742,
+      "learning_rate": 2.6712898751733705e-05,
+      "loss": 1.1762,
+      "step": 632
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.136801719665527,
+      "learning_rate": 2.6707697642163664e-05,
+      "loss": 1.2387,
+      "step": 633
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.06429672241211,
+      "learning_rate": 2.6702496532593622e-05,
+      "loss": 1.2794,
+      "step": 634
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.476702690124512,
+      "learning_rate": 2.6697295423023577e-05,
+      "loss": 1.2872,
+      "step": 635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.201651573181152,
+      "learning_rate": 2.669209431345354e-05,
+      "loss": 1.3713,
+      "step": 636
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.263508796691895,
+      "learning_rate": 2.6686893203883495e-05,
+      "loss": 0.9582,
+      "step": 637
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.963038444519043,
+      "learning_rate": 2.6681692094313457e-05,
+      "loss": 1.2708,
+      "step": 638
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.786746978759766,
+      "learning_rate": 2.6676490984743412e-05,
+      "loss": 0.9783,
+      "step": 639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.89470386505127,
+      "learning_rate": 2.6671289875173374e-05,
+      "loss": 1.214,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.329146385192871,
+      "learning_rate": 2.666608876560333e-05,
+      "loss": 1.3699,
+      "step": 641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.576723575592041,
+      "learning_rate": 2.6660887656033288e-05,
+      "loss": 0.9926,
+      "step": 642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.539240837097168,
+      "learning_rate": 2.6655686546463246e-05,
+      "loss": 1.3883,
+      "step": 643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.552998542785645,
+      "learning_rate": 2.6650485436893205e-05,
+      "loss": 1.2427,
+      "step": 644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.002632141113281,
+      "learning_rate": 2.6645284327323163e-05,
+      "loss": 1.2611,
+      "step": 645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.49915885925293,
+      "learning_rate": 2.6640083217753122e-05,
+      "loss": 1.4899,
+      "step": 646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.545672416687012,
+      "learning_rate": 2.663488210818308e-05,
+      "loss": 1.1518,
+      "step": 647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.826525688171387,
+      "learning_rate": 2.662968099861304e-05,
+      "loss": 0.9722,
+      "step": 648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.8317484855651855,
+      "learning_rate": 2.6624479889042998e-05,
+      "loss": 1.1108,
+      "step": 649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.123722076416016,
+      "learning_rate": 2.6619278779472957e-05,
+      "loss": 1.0138,
+      "step": 650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.588679313659668,
+      "learning_rate": 2.6614077669902912e-05,
+      "loss": 1.1157,
+      "step": 651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.59313678741455,
+      "learning_rate": 2.6608876560332874e-05,
+      "loss": 1.3583,
+      "step": 652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.904531478881836,
+      "learning_rate": 2.660367545076283e-05,
+      "loss": 1.5144,
+      "step": 653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.814980506896973,
+      "learning_rate": 2.659847434119279e-05,
+      "loss": 1.3732,
+      "step": 654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.91495132446289,
+      "learning_rate": 2.6593273231622746e-05,
+      "loss": 1.1658,
+      "step": 655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.983271598815918,
+      "learning_rate": 2.6588072122052708e-05,
+      "loss": 0.9849,
+      "step": 656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.348917961120605,
+      "learning_rate": 2.6582871012482663e-05,
+      "loss": 1.3562,
+      "step": 657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.976978302001953,
+      "learning_rate": 2.6577669902912622e-05,
+      "loss": 1.3029,
+      "step": 658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.533385276794434,
+      "learning_rate": 2.657246879334258e-05,
+      "loss": 1.0711,
+      "step": 659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.634657859802246,
+      "learning_rate": 2.656726768377254e-05,
+      "loss": 1.2808,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.234044075012207,
+      "learning_rate": 2.6562066574202498e-05,
+      "loss": 1.2323,
+      "step": 661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.877964973449707,
+      "learning_rate": 2.6556865464632457e-05,
+      "loss": 1.182,
+      "step": 662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.138318061828613,
+      "learning_rate": 2.6551664355062415e-05,
+      "loss": 1.1607,
+      "step": 663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.455748558044434,
+      "learning_rate": 2.6546463245492374e-05,
+      "loss": 1.1812,
+      "step": 664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.842740058898926,
+      "learning_rate": 2.654126213592233e-05,
+      "loss": 1.5691,
+      "step": 665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.516674041748047,
+      "learning_rate": 2.653606102635229e-05,
+      "loss": 1.1348,
+      "step": 666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.163476943969727,
+      "learning_rate": 2.6530859916782246e-05,
+      "loss": 1.2331,
+      "step": 667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.494277954101562,
+      "learning_rate": 2.6525658807212208e-05,
+      "loss": 1.2412,
+      "step": 668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.835494041442871,
+      "learning_rate": 2.6520457697642163e-05,
+      "loss": 1.3143,
+      "step": 669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.751224517822266,
+      "learning_rate": 2.6515256588072125e-05,
+      "loss": 1.3948,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.18032169342041,
+      "learning_rate": 2.651005547850208e-05,
+      "loss": 1.2613,
+      "step": 671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.477836608886719,
+      "learning_rate": 2.6504854368932043e-05,
+      "loss": 0.971,
+      "step": 672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.272271156311035,
+      "learning_rate": 2.6499653259361998e-05,
+      "loss": 1.0968,
+      "step": 673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.481215476989746,
+      "learning_rate": 2.6494452149791956e-05,
+      "loss": 1.1182,
+      "step": 674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.60176944732666,
+      "learning_rate": 2.6489251040221915e-05,
+      "loss": 1.3199,
+      "step": 675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.935422897338867,
+      "learning_rate": 2.6484049930651874e-05,
+      "loss": 1.1209,
+      "step": 676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.94033432006836,
+      "learning_rate": 2.6478848821081832e-05,
+      "loss": 1.2507,
+      "step": 677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.514902114868164,
+      "learning_rate": 2.647364771151179e-05,
+      "loss": 1.215,
+      "step": 678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.019514083862305,
+      "learning_rate": 2.646844660194175e-05,
+      "loss": 1.4208,
+      "step": 679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.107585906982422,
+      "learning_rate": 2.6463245492371708e-05,
+      "loss": 1.4244,
+      "step": 680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.146953582763672,
+      "learning_rate": 2.6458044382801663e-05,
+      "loss": 1.1731,
+      "step": 681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.342073440551758,
+      "learning_rate": 2.6452843273231625e-05,
+      "loss": 1.271,
+      "step": 682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.38988971710205,
+      "learning_rate": 2.644764216366158e-05,
+      "loss": 1.2265,
+      "step": 683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.83610725402832,
+      "learning_rate": 2.6442441054091543e-05,
+      "loss": 1.3186,
+      "step": 684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.806127548217773,
+      "learning_rate": 2.6437239944521498e-05,
+      "loss": 1.2249,
+      "step": 685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.570383548736572,
+      "learning_rate": 2.643203883495146e-05,
+      "loss": 1.1001,
+      "step": 686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.918679237365723,
+      "learning_rate": 2.6426837725381415e-05,
+      "loss": 1.4043,
+      "step": 687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.545655250549316,
+      "learning_rate": 2.6421636615811377e-05,
+      "loss": 1.4472,
+      "step": 688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.794044494628906,
+      "learning_rate": 2.6416435506241332e-05,
+      "loss": 1.178,
+      "step": 689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.89754581451416,
+      "learning_rate": 2.641123439667129e-05,
+      "loss": 1.1883,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.437743663787842,
+      "learning_rate": 2.640603328710125e-05,
+      "loss": 0.9985,
+      "step": 691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.914102554321289,
+      "learning_rate": 2.6400832177531208e-05,
+      "loss": 1.3658,
+      "step": 692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.124156951904297,
+      "learning_rate": 2.6395631067961167e-05,
+      "loss": 1.2106,
+      "step": 693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.209687232971191,
+      "learning_rate": 2.6390429958391125e-05,
+      "loss": 0.952,
+      "step": 694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.671976089477539,
+      "learning_rate": 2.6385228848821084e-05,
+      "loss": 1.0483,
+      "step": 695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.107882499694824,
+      "learning_rate": 2.6380027739251043e-05,
+      "loss": 1.1585,
+      "step": 696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.695256233215332,
+      "learning_rate": 2.6374826629680998e-05,
+      "loss": 1.1776,
+      "step": 697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.420937538146973,
+      "learning_rate": 2.636962552011096e-05,
+      "loss": 1.2612,
+      "step": 698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.463383674621582,
+      "learning_rate": 2.6364424410540915e-05,
+      "loss": 1.2096,
+      "step": 699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.406822204589844,
+      "learning_rate": 2.6359223300970877e-05,
+      "loss": 1.1055,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.851249694824219,
+      "learning_rate": 2.6354022191400832e-05,
+      "loss": 1.4411,
+      "step": 701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.304463386535645,
+      "learning_rate": 2.6348821081830794e-05,
+      "loss": 1.2038,
+      "step": 702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.872647285461426,
+      "learning_rate": 2.634361997226075e-05,
+      "loss": 1.3517,
+      "step": 703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.731727600097656,
+      "learning_rate": 2.6338418862690708e-05,
+      "loss": 1.0877,
+      "step": 704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.550158500671387,
+      "learning_rate": 2.6333217753120667e-05,
+      "loss": 1.2207,
+      "step": 705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.14528751373291,
+      "learning_rate": 2.6328016643550625e-05,
+      "loss": 1.3675,
+      "step": 706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.461470603942871,
+      "learning_rate": 2.6322815533980584e-05,
+      "loss": 1.3051,
+      "step": 707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.056265830993652,
+      "learning_rate": 2.6317614424410543e-05,
+      "loss": 1.1708,
+      "step": 708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.258934020996094,
+      "learning_rate": 2.63124133148405e-05,
+      "loss": 1.3828,
+      "step": 709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.749487400054932,
+      "learning_rate": 2.630721220527046e-05,
+      "loss": 1.0926,
+      "step": 710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.534918785095215,
+      "learning_rate": 2.630201109570042e-05,
+      "loss": 1.1352,
+      "step": 711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.562870979309082,
+      "learning_rate": 2.6296809986130377e-05,
+      "loss": 1.2397,
+      "step": 712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.155929565429688,
+      "learning_rate": 2.6291608876560332e-05,
+      "loss": 1.2855,
+      "step": 713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.94394302368164,
+      "learning_rate": 2.6286407766990294e-05,
+      "loss": 1.0019,
+      "step": 714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.855461120605469,
+      "learning_rate": 2.628120665742025e-05,
+      "loss": 1.0617,
+      "step": 715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.463730812072754,
+      "learning_rate": 2.627600554785021e-05,
+      "loss": 1.5647,
+      "step": 716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.241301536560059,
+      "learning_rate": 2.6270804438280167e-05,
+      "loss": 1.3845,
+      "step": 717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.891985893249512,
+      "learning_rate": 2.626560332871013e-05,
+      "loss": 1.2699,
+      "step": 718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.604130744934082,
+      "learning_rate": 2.6260402219140084e-05,
+      "loss": 1.2937,
+      "step": 719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.020197868347168,
+      "learning_rate": 2.6255201109570042e-05,
+      "loss": 1.0384,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.678190231323242,
+      "learning_rate": 2.625e-05,
+      "loss": 1.3818,
+      "step": 721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.847310543060303,
+      "learning_rate": 2.6244798890429956e-05,
+      "loss": 1.1503,
+      "step": 722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.211929321289062,
+      "learning_rate": 2.623959778085992e-05,
+      "loss": 1.5145,
+      "step": 723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.5321626663208,
+      "learning_rate": 2.6234396671289874e-05,
+      "loss": 1.2727,
+      "step": 724
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.7330961227417,
+      "learning_rate": 2.6229195561719836e-05,
+      "loss": 1.3171,
+      "step": 725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.253852844238281,
+      "learning_rate": 2.622399445214979e-05,
+      "loss": 1.4148,
+      "step": 726
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.089603424072266,
+      "learning_rate": 2.621879334257975e-05,
+      "loss": 1.0252,
+      "step": 727
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.47254467010498,
+      "learning_rate": 2.6213592233009708e-05,
+      "loss": 1.1915,
+      "step": 728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.198091506958008,
+      "learning_rate": 2.6208391123439667e-05,
+      "loss": 1.3681,
+      "step": 729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.844797134399414,
+      "learning_rate": 2.6203190013869625e-05,
+      "loss": 1.1388,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.806116104125977,
+      "learning_rate": 2.6197988904299584e-05,
+      "loss": 1.5571,
+      "step": 731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.732404708862305,
+      "learning_rate": 2.6192787794729542e-05,
+      "loss": 1.2289,
+      "step": 732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.778172492980957,
+      "learning_rate": 2.61875866851595e-05,
+      "loss": 1.4028,
+      "step": 733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.307780265808105,
+      "learning_rate": 2.618238557558946e-05,
+      "loss": 1.4206,
+      "step": 734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.113368034362793,
+      "learning_rate": 2.6177184466019418e-05,
+      "loss": 1.2585,
+      "step": 735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.215656280517578,
+      "learning_rate": 2.6171983356449373e-05,
+      "loss": 1.2637,
+      "step": 736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.9940767288208,
+      "learning_rate": 2.6166782246879335e-05,
+      "loss": 1.2459,
+      "step": 737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.71268081665039,
+      "learning_rate": 2.616158113730929e-05,
+      "loss": 1.3878,
+      "step": 738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.722239017486572,
+      "learning_rate": 2.6156380027739253e-05,
+      "loss": 0.9557,
+      "step": 739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.503205299377441,
+      "learning_rate": 2.6151178918169208e-05,
+      "loss": 1.2034,
+      "step": 740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.04439640045166,
+      "learning_rate": 2.614597780859917e-05,
+      "loss": 1.503,
+      "step": 741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.41606330871582,
+      "learning_rate": 2.6140776699029125e-05,
+      "loss": 1.182,
+      "step": 742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.777578353881836,
+      "learning_rate": 2.6135575589459084e-05,
+      "loss": 0.9642,
+      "step": 743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.453320503234863,
+      "learning_rate": 2.6130374479889042e-05,
+      "loss": 1.326,
+      "step": 744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.22025203704834,
+      "learning_rate": 2.6125173370319e-05,
+      "loss": 1.2529,
+      "step": 745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.686552047729492,
+      "learning_rate": 2.611997226074896e-05,
+      "loss": 1.0604,
+      "step": 746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.597603797912598,
+      "learning_rate": 2.6114771151178918e-05,
+      "loss": 1.2296,
+      "step": 747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.151700973510742,
+      "learning_rate": 2.6109570041608877e-05,
+      "loss": 1.1653,
+      "step": 748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.472678184509277,
+      "learning_rate": 2.6104368932038835e-05,
+      "loss": 1.1864,
+      "step": 749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.167405605316162,
+      "learning_rate": 2.6099167822468794e-05,
+      "loss": 0.8641,
+      "step": 750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.457036018371582,
+      "learning_rate": 2.6093966712898753e-05,
+      "loss": 0.8526,
+      "step": 751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.86585807800293,
+      "learning_rate": 2.6088765603328708e-05,
+      "loss": 1.2687,
+      "step": 752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.411893844604492,
+      "learning_rate": 2.608356449375867e-05,
+      "loss": 1.0087,
+      "step": 753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.614012718200684,
+      "learning_rate": 2.6078363384188625e-05,
+      "loss": 1.4972,
+      "step": 754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.547771453857422,
+      "learning_rate": 2.6073162274618587e-05,
+      "loss": 1.0555,
+      "step": 755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.479796409606934,
+      "learning_rate": 2.6067961165048542e-05,
+      "loss": 1.2343,
+      "step": 756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.810163497924805,
+      "learning_rate": 2.6062760055478504e-05,
+      "loss": 1.0418,
+      "step": 757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.572808265686035,
+      "learning_rate": 2.605755894590846e-05,
+      "loss": 1.2371,
+      "step": 758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.260298728942871,
+      "learning_rate": 2.6052357836338418e-05,
+      "loss": 1.2572,
+      "step": 759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.918755531311035,
+      "learning_rate": 2.6047156726768377e-05,
+      "loss": 1.0709,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.397783279418945,
+      "learning_rate": 2.6041955617198335e-05,
+      "loss": 0.8753,
+      "step": 761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.469996452331543,
+      "learning_rate": 2.6036754507628294e-05,
+      "loss": 1.0806,
+      "step": 762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.95573616027832,
+      "learning_rate": 2.6031553398058253e-05,
+      "loss": 1.3964,
+      "step": 763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.759084701538086,
+      "learning_rate": 2.602635228848821e-05,
+      "loss": 1.05,
+      "step": 764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.913603782653809,
+      "learning_rate": 2.602115117891817e-05,
+      "loss": 1.2691,
+      "step": 765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.396145820617676,
+      "learning_rate": 2.6015950069348125e-05,
+      "loss": 1.3313,
+      "step": 766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.28937816619873,
+      "learning_rate": 2.6010748959778087e-05,
+      "loss": 1.2919,
+      "step": 767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.175122261047363,
+      "learning_rate": 2.6005547850208042e-05,
+      "loss": 1.6025,
+      "step": 768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.259637832641602,
+      "learning_rate": 2.6000346740638004e-05,
+      "loss": 1.3895,
+      "step": 769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.588577270507812,
+      "learning_rate": 2.599514563106796e-05,
+      "loss": 1.0706,
+      "step": 770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.026396751403809,
+      "learning_rate": 2.598994452149792e-05,
+      "loss": 1.2088,
+      "step": 771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.164908409118652,
+      "learning_rate": 2.5984743411927877e-05,
+      "loss": 1.1384,
+      "step": 772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.302207946777344,
+      "learning_rate": 2.597954230235784e-05,
+      "loss": 1.1141,
+      "step": 773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.077268600463867,
+      "learning_rate": 2.5974341192787794e-05,
+      "loss": 1.1967,
+      "step": 774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.200212478637695,
+      "learning_rate": 2.5969140083217753e-05,
+      "loss": 0.8065,
+      "step": 775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.635682106018066,
+      "learning_rate": 2.596393897364771e-05,
+      "loss": 1.0078,
+      "step": 776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.121787071228027,
+      "learning_rate": 2.595873786407767e-05,
+      "loss": 1.1093,
+      "step": 777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.347357749938965,
+      "learning_rate": 2.595353675450763e-05,
+      "loss": 1.1084,
+      "step": 778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.573598861694336,
+      "learning_rate": 2.5948335644937587e-05,
+      "loss": 1.1512,
+      "step": 779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.280482292175293,
+      "learning_rate": 2.5943134535367546e-05,
+      "loss": 1.2386,
+      "step": 780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.696454048156738,
+      "learning_rate": 2.5937933425797504e-05,
+      "loss": 1.2979,
+      "step": 781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.360091209411621,
+      "learning_rate": 2.593273231622746e-05,
+      "loss": 1.0269,
+      "step": 782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.222893714904785,
+      "learning_rate": 2.592753120665742e-05,
+      "loss": 1.0173,
+      "step": 783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.139443397521973,
+      "learning_rate": 2.5922330097087377e-05,
+      "loss": 1.6817,
+      "step": 784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.55649471282959,
+      "learning_rate": 2.591712898751734e-05,
+      "loss": 1.3373,
+      "step": 785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.11771011352539,
+      "learning_rate": 2.5911927877947294e-05,
+      "loss": 1.3629,
+      "step": 786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.0601224899292,
+      "learning_rate": 2.5906726768377256e-05,
+      "loss": 1.2173,
+      "step": 787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.504056930541992,
+      "learning_rate": 2.590152565880721e-05,
+      "loss": 1.0658,
+      "step": 788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.688624382019043,
+      "learning_rate": 2.589632454923717e-05,
+      "loss": 1.2183,
+      "step": 789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.717049598693848,
+      "learning_rate": 2.589112343966713e-05,
+      "loss": 1.1386,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.444982528686523,
+      "learning_rate": 2.5885922330097087e-05,
+      "loss": 1.2706,
+      "step": 791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.381704330444336,
+      "learning_rate": 2.5880721220527046e-05,
+      "loss": 1.1309,
+      "step": 792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.749383926391602,
+      "learning_rate": 2.5875520110957004e-05,
+      "loss": 1.1089,
+      "step": 793
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.764632225036621,
+      "learning_rate": 2.5870319001386963e-05,
+      "loss": 1.2133,
+      "step": 794
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.557351112365723,
+      "learning_rate": 2.586511789181692e-05,
+      "loss": 1.1468,
+      "step": 795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.855218887329102,
+      "learning_rate": 2.585991678224688e-05,
+      "loss": 1.1151,
+      "step": 796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.569310188293457,
+      "learning_rate": 2.585471567267684e-05,
+      "loss": 1.009,
+      "step": 797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.135235786437988,
+      "learning_rate": 2.5849514563106794e-05,
+      "loss": 1.6472,
+      "step": 798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.711816787719727,
+      "learning_rate": 2.5844313453536756e-05,
+      "loss": 1.1054,
+      "step": 799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.59421443939209,
+      "learning_rate": 2.583911234396671e-05,
+      "loss": 1.3199,
+      "step": 800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.127828598022461,
+      "learning_rate": 2.5833911234396673e-05,
+      "loss": 0.9647,
+      "step": 801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.614999771118164,
+      "learning_rate": 2.582871012482663e-05,
+      "loss": 1.0966,
+      "step": 802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.848357200622559,
+      "learning_rate": 2.582350901525659e-05,
+      "loss": 1.0745,
+      "step": 803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.322065353393555,
+      "learning_rate": 2.5818307905686546e-05,
+      "loss": 1.3714,
+      "step": 804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.61287784576416,
+      "learning_rate": 2.5813106796116504e-05,
+      "loss": 1.4444,
+      "step": 805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.29521656036377,
+      "learning_rate": 2.5807905686546463e-05,
+      "loss": 1.173,
+      "step": 806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.3076753616333,
+      "learning_rate": 2.580270457697642e-05,
+      "loss": 1.534,
+      "step": 807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.124590873718262,
+      "learning_rate": 2.579750346740638e-05,
+      "loss": 1.2595,
+      "step": 808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.819221019744873,
+      "learning_rate": 2.579230235783634e-05,
+      "loss": 1.0259,
+      "step": 809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.594743728637695,
+      "learning_rate": 2.5787101248266297e-05,
+      "loss": 1.2043,
+      "step": 810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.951892852783203,
+      "learning_rate": 2.5781900138696256e-05,
+      "loss": 1.0434,
+      "step": 811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.361400604248047,
+      "learning_rate": 2.577669902912621e-05,
+      "loss": 1.3102,
+      "step": 812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.740880966186523,
+      "learning_rate": 2.5771497919556173e-05,
+      "loss": 1.391,
+      "step": 813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.772485733032227,
+      "learning_rate": 2.576629680998613e-05,
+      "loss": 1.2798,
+      "step": 814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.828903198242188,
+      "learning_rate": 2.576109570041609e-05,
+      "loss": 1.4228,
+      "step": 815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.499209880828857,
+      "learning_rate": 2.5755894590846046e-05,
+      "loss": 0.9988,
+      "step": 816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.469114303588867,
+      "learning_rate": 2.5750693481276008e-05,
+      "loss": 1.3631,
+      "step": 817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.403695106506348,
+      "learning_rate": 2.5745492371705963e-05,
+      "loss": 1.3496,
+      "step": 818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.451019287109375,
+      "learning_rate": 2.5740291262135925e-05,
+      "loss": 1.3658,
+      "step": 819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.174098014831543,
+      "learning_rate": 2.573509015256588e-05,
+      "loss": 1.3371,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.007797241210938,
+      "learning_rate": 2.572988904299584e-05,
+      "loss": 1.4078,
+      "step": 821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.366394996643066,
+      "learning_rate": 2.5724687933425797e-05,
+      "loss": 1.1704,
+      "step": 822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.588715553283691,
+      "learning_rate": 2.5719486823855756e-05,
+      "loss": 1.4759,
+      "step": 823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.157060623168945,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 1.1258,
+      "step": 824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.143661499023438,
+      "learning_rate": 2.5709084604715673e-05,
+      "loss": 1.0804,
+      "step": 825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.478647232055664,
+      "learning_rate": 2.570388349514563e-05,
+      "loss": 1.2246,
+      "step": 826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.546572685241699,
+      "learning_rate": 2.569868238557559e-05,
+      "loss": 1.1916,
+      "step": 827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.57593822479248,
+      "learning_rate": 2.5693481276005546e-05,
+      "loss": 1.2855,
+      "step": 828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.429585456848145,
+      "learning_rate": 2.5688280166435508e-05,
+      "loss": 1.3497,
+      "step": 829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.605025291442871,
+      "learning_rate": 2.5683079056865463e-05,
+      "loss": 1.2301,
+      "step": 830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.559940338134766,
+      "learning_rate": 2.5677877947295425e-05,
+      "loss": 1.2916,
+      "step": 831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.190064430236816,
+      "learning_rate": 2.567267683772538e-05,
+      "loss": 1.1674,
+      "step": 832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.020051002502441,
+      "learning_rate": 2.5667475728155342e-05,
+      "loss": 1.0053,
+      "step": 833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.703924179077148,
+      "learning_rate": 2.5662274618585297e-05,
+      "loss": 1.2267,
+      "step": 834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.728412628173828,
+      "learning_rate": 2.565707350901526e-05,
+      "loss": 1.1592,
+      "step": 835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.116409301757812,
+      "learning_rate": 2.5651872399445214e-05,
+      "loss": 1.1441,
+      "step": 836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.695294380187988,
+      "learning_rate": 2.5646671289875173e-05,
+      "loss": 1.1508,
+      "step": 837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.306589126586914,
+      "learning_rate": 2.564147018030513e-05,
+      "loss": 1.3467,
+      "step": 838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.978409767150879,
+      "learning_rate": 2.563626907073509e-05,
+      "loss": 1.4941,
+      "step": 839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.270303726196289,
+      "learning_rate": 2.563106796116505e-05,
+      "loss": 1.1325,
+      "step": 840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.547574043273926,
+      "learning_rate": 2.5625866851595007e-05,
+      "loss": 1.4015,
+      "step": 841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.216253280639648,
+      "learning_rate": 2.5620665742024966e-05,
+      "loss": 1.2381,
+      "step": 842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.689031600952148,
+      "learning_rate": 2.5615464632454925e-05,
+      "loss": 0.9999,
+      "step": 843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.411420822143555,
+      "learning_rate": 2.561026352288488e-05,
+      "loss": 1.1721,
+      "step": 844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.995872497558594,
+      "learning_rate": 2.5605062413314842e-05,
+      "loss": 1.2055,
+      "step": 845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.848401069641113,
+      "learning_rate": 2.5599861303744797e-05,
+      "loss": 1.0137,
+      "step": 846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.497431755065918,
+      "learning_rate": 2.559466019417476e-05,
+      "loss": 1.1014,
+      "step": 847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.560737609863281,
+      "learning_rate": 2.5589459084604714e-05,
+      "loss": 0.9234,
+      "step": 848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.658347129821777,
+      "learning_rate": 2.5584257975034676e-05,
+      "loss": 1.0307,
+      "step": 849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.714625358581543,
+      "learning_rate": 2.557905686546463e-05,
+      "loss": 1.4185,
+      "step": 850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.22028636932373,
+      "learning_rate": 2.557385575589459e-05,
+      "loss": 1.3661,
+      "step": 851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.131681442260742,
+      "learning_rate": 2.556865464632455e-05,
+      "loss": 0.843,
+      "step": 852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.221606254577637,
+      "learning_rate": 2.5563453536754507e-05,
+      "loss": 1.3294,
+      "step": 853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.481217384338379,
+      "learning_rate": 2.5558252427184466e-05,
+      "loss": 0.905,
+      "step": 854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.945030212402344,
+      "learning_rate": 2.5553051317614425e-05,
+      "loss": 1.6755,
+      "step": 855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.647332191467285,
+      "learning_rate": 2.5547850208044383e-05,
+      "loss": 1.0578,
+      "step": 856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.187901496887207,
+      "learning_rate": 2.5542649098474342e-05,
+      "loss": 1.2354,
+      "step": 857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.420130729675293,
+      "learning_rate": 2.55374479889043e-05,
+      "loss": 0.8506,
+      "step": 858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.184560775756836,
+      "learning_rate": 2.553224687933426e-05,
+      "loss": 1.2255,
+      "step": 859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.013655662536621,
+      "learning_rate": 2.5527045769764214e-05,
+      "loss": 1.0143,
+      "step": 860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.667200088500977,
+      "learning_rate": 2.5521844660194176e-05,
+      "loss": 1.3442,
+      "step": 861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.685121536254883,
+      "learning_rate": 2.551664355062413e-05,
+      "loss": 1.3532,
+      "step": 862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.339266777038574,
+      "learning_rate": 2.5511442441054094e-05,
+      "loss": 1.249,
+      "step": 863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.764921188354492,
+      "learning_rate": 2.550624133148405e-05,
+      "loss": 1.1926,
+      "step": 864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.76657772064209,
+      "learning_rate": 2.550104022191401e-05,
+      "loss": 1.2935,
+      "step": 865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.448904037475586,
+      "learning_rate": 2.5495839112343966e-05,
+      "loss": 1.0809,
+      "step": 866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.616414070129395,
+      "learning_rate": 2.5490638002773925e-05,
+      "loss": 0.9777,
+      "step": 867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.189936637878418,
+      "learning_rate": 2.5485436893203883e-05,
+      "loss": 0.9879,
+      "step": 868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.878035545349121,
+      "learning_rate": 2.5480235783633842e-05,
+      "loss": 0.9491,
+      "step": 869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.97451114654541,
+      "learning_rate": 2.54750346740638e-05,
+      "loss": 1.1131,
+      "step": 870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.322022438049316,
+      "learning_rate": 2.546983356449376e-05,
+      "loss": 1.1408,
+      "step": 871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.71009635925293,
+      "learning_rate": 2.5464632454923718e-05,
+      "loss": 1.43,
+      "step": 872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.999345779418945,
+      "learning_rate": 2.5459431345353676e-05,
+      "loss": 1.3047,
+      "step": 873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.789861679077148,
+      "learning_rate": 2.545423023578363e-05,
+      "loss": 1.1912,
+      "step": 874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.957731246948242,
+      "learning_rate": 2.5449029126213594e-05,
+      "loss": 1.155,
+      "step": 875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.823418617248535,
+      "learning_rate": 2.544382801664355e-05,
+      "loss": 0.8906,
+      "step": 876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.580862998962402,
+      "learning_rate": 2.543862690707351e-05,
+      "loss": 1.0805,
+      "step": 877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.127274513244629,
+      "learning_rate": 2.5433425797503466e-05,
+      "loss": 1.0458,
+      "step": 878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.295371055603027,
+      "learning_rate": 2.5428224687933428e-05,
+      "loss": 1.4238,
+      "step": 879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.931177139282227,
+      "learning_rate": 2.5423023578363383e-05,
+      "loss": 1.3701,
+      "step": 880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.384162902832031,
+      "learning_rate": 2.5417822468793345e-05,
+      "loss": 1.0701,
+      "step": 881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.992918014526367,
+      "learning_rate": 2.54126213592233e-05,
+      "loss": 1.156,
+      "step": 882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.223592758178711,
+      "learning_rate": 2.540742024965326e-05,
+      "loss": 0.9094,
+      "step": 883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.637628555297852,
+      "learning_rate": 2.5402219140083218e-05,
+      "loss": 1.3089,
+      "step": 884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.275469779968262,
+      "learning_rate": 2.5397018030513176e-05,
+      "loss": 0.9509,
+      "step": 885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.599825859069824,
+      "learning_rate": 2.5391816920943135e-05,
+      "loss": 1.4559,
+      "step": 886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.039251327514648,
+      "learning_rate": 2.5386615811373093e-05,
+      "loss": 1.0057,
+      "step": 887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.520696640014648,
+      "learning_rate": 2.5381414701803052e-05,
+      "loss": 1.2121,
+      "step": 888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.586304664611816,
+      "learning_rate": 2.537621359223301e-05,
+      "loss": 1.2331,
+      "step": 889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.068800926208496,
+      "learning_rate": 2.5371012482662966e-05,
+      "loss": 1.195,
+      "step": 890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.369816780090332,
+      "learning_rate": 2.5365811373092928e-05,
+      "loss": 1.206,
+      "step": 891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.612571716308594,
+      "learning_rate": 2.5360610263522883e-05,
+      "loss": 1.1441,
+      "step": 892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.516185760498047,
+      "learning_rate": 2.5355409153952845e-05,
+      "loss": 1.5696,
+      "step": 893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.411796569824219,
+      "learning_rate": 2.53502080443828e-05,
+      "loss": 0.98,
+      "step": 894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.329340934753418,
+      "learning_rate": 2.5345006934812762e-05,
+      "loss": 1.1452,
+      "step": 895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.6731367111206055,
+      "learning_rate": 2.5339805825242718e-05,
+      "loss": 1.1079,
+      "step": 896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.672621726989746,
+      "learning_rate": 2.5334604715672676e-05,
+      "loss": 1.1821,
+      "step": 897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.029902458190918,
+      "learning_rate": 2.5329403606102635e-05,
+      "loss": 1.4712,
+      "step": 898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.270554542541504,
+      "learning_rate": 2.5324202496532593e-05,
+      "loss": 1.2378,
+      "step": 899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.440147399902344,
+      "learning_rate": 2.5319001386962552e-05,
+      "loss": 1.035,
+      "step": 900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.348343849182129,
+      "learning_rate": 2.531380027739251e-05,
+      "loss": 1.0391,
+      "step": 901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.073831558227539,
+      "learning_rate": 2.530859916782247e-05,
+      "loss": 1.2136,
+      "step": 902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.846288681030273,
+      "learning_rate": 2.5303398058252428e-05,
+      "loss": 0.9843,
+      "step": 903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.772936820983887,
+      "learning_rate": 2.5298196948682387e-05,
+      "loss": 1.1549,
+      "step": 904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.2023286819458,
+      "learning_rate": 2.5292995839112345e-05,
+      "loss": 1.2877,
+      "step": 905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.033526420593262,
+      "learning_rate": 2.52877947295423e-05,
+      "loss": 0.9613,
+      "step": 906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.604496002197266,
+      "learning_rate": 2.5282593619972262e-05,
+      "loss": 1.6032,
+      "step": 907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.419151306152344,
+      "learning_rate": 2.5277392510402218e-05,
+      "loss": 1.0931,
+      "step": 908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.42763900756836,
+      "learning_rate": 2.527219140083218e-05,
+      "loss": 1.2777,
+      "step": 909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.89081859588623,
+      "learning_rate": 2.5266990291262135e-05,
+      "loss": 1.0294,
+      "step": 910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.147234916687012,
+      "learning_rate": 2.5261789181692097e-05,
+      "loss": 1.3203,
+      "step": 911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.739482879638672,
+      "learning_rate": 2.5256588072122052e-05,
+      "loss": 1.1486,
+      "step": 912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.371050834655762,
+      "learning_rate": 2.525138696255201e-05,
+      "loss": 1.2171,
+      "step": 913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.034295082092285,
+      "learning_rate": 2.524618585298197e-05,
+      "loss": 1.0927,
+      "step": 914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.903518676757812,
+      "learning_rate": 2.5240984743411928e-05,
+      "loss": 1.4597,
+      "step": 915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.512449264526367,
+      "learning_rate": 2.5235783633841886e-05,
+      "loss": 1.0094,
+      "step": 916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.698187828063965,
+      "learning_rate": 2.5230582524271845e-05,
+      "loss": 1.0736,
+      "step": 917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.61658000946045,
+      "learning_rate": 2.5225381414701804e-05,
+      "loss": 1.1264,
+      "step": 918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.534981727600098,
+      "learning_rate": 2.5220180305131762e-05,
+      "loss": 0.9651,
+      "step": 919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.417647361755371,
+      "learning_rate": 2.521497919556172e-05,
+      "loss": 1.1207,
+      "step": 920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.902402877807617,
+      "learning_rate": 2.520977808599168e-05,
+      "loss": 1.0134,
+      "step": 921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.409195899963379,
+      "learning_rate": 2.5204576976421635e-05,
+      "loss": 1.4207,
+      "step": 922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.239357948303223,
+      "learning_rate": 2.5199375866851597e-05,
+      "loss": 1.3169,
+      "step": 923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.165468215942383,
+      "learning_rate": 2.5194174757281552e-05,
+      "loss": 1.0104,
+      "step": 924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.005200386047363,
+      "learning_rate": 2.5188973647711514e-05,
+      "loss": 1.2666,
+      "step": 925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.400407314300537,
+      "learning_rate": 2.518377253814147e-05,
+      "loss": 1.1305,
+      "step": 926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.789124488830566,
+      "learning_rate": 2.517857142857143e-05,
+      "loss": 1.088,
+      "step": 927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.899168014526367,
+      "learning_rate": 2.5173370319001386e-05,
+      "loss": 0.9541,
+      "step": 928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.3808698654174805,
+      "learning_rate": 2.5168169209431345e-05,
+      "loss": 0.8101,
+      "step": 929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.771642684936523,
+      "learning_rate": 2.5162968099861304e-05,
+      "loss": 1.1734,
+      "step": 930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.127939224243164,
+      "learning_rate": 2.5157766990291262e-05,
+      "loss": 0.991,
+      "step": 931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.57542896270752,
+      "learning_rate": 2.515256588072122e-05,
+      "loss": 1.1924,
+      "step": 932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.621986389160156,
+      "learning_rate": 2.514736477115118e-05,
+      "loss": 1.128,
+      "step": 933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.981200218200684,
+      "learning_rate": 2.5142163661581138e-05,
+      "loss": 1.2711,
+      "step": 934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.787057876586914,
+      "learning_rate": 2.5136962552011097e-05,
+      "loss": 1.4416,
+      "step": 935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.694108009338379,
+      "learning_rate": 2.5131761442441052e-05,
+      "loss": 0.9907,
+      "step": 936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.345697402954102,
+      "learning_rate": 2.5126560332871014e-05,
+      "loss": 1.1176,
+      "step": 937
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.257140159606934,
+      "learning_rate": 2.512135922330097e-05,
+      "loss": 1.4549,
+      "step": 938
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.078888893127441,
+      "learning_rate": 2.511615811373093e-05,
+      "loss": 1.3047,
+      "step": 939
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.023653030395508,
+      "learning_rate": 2.5110957004160886e-05,
+      "loss": 1.0633,
+      "step": 940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.270830154418945,
+      "learning_rate": 2.510575589459085e-05,
+      "loss": 1.1547,
+      "step": 941
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.745702743530273,
+      "learning_rate": 2.5100554785020804e-05,
+      "loss": 1.4424,
+      "step": 942
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.310778617858887,
+      "learning_rate": 2.5095353675450766e-05,
+      "loss": 1.4323,
+      "step": 943
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.381902694702148,
+      "learning_rate": 2.509015256588072e-05,
+      "loss": 1.2702,
+      "step": 944
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.333796501159668,
+      "learning_rate": 2.508495145631068e-05,
+      "loss": 1.2697,
+      "step": 945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.717247009277344,
+      "learning_rate": 2.5079750346740638e-05,
+      "loss": 1.21,
+      "step": 946
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.057883262634277,
+      "learning_rate": 2.5074549237170597e-05,
+      "loss": 1.1431,
+      "step": 947
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.843471050262451,
+      "learning_rate": 2.5069348127600555e-05,
+      "loss": 0.9535,
+      "step": 948
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.294394493103027,
+      "learning_rate": 2.5064147018030514e-05,
+      "loss": 1.2971,
+      "step": 949
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.180764198303223,
+      "learning_rate": 2.5058945908460473e-05,
+      "loss": 1.2619,
+      "step": 950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.55435848236084,
+      "learning_rate": 2.505374479889043e-05,
+      "loss": 1.0289,
+      "step": 951
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.405740737915039,
+      "learning_rate": 2.5048543689320386e-05,
+      "loss": 1.2265,
+      "step": 952
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.818949699401855,
+      "learning_rate": 2.504334257975035e-05,
+      "loss": 1.1864,
+      "step": 953
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.069859504699707,
+      "learning_rate": 2.5038141470180304e-05,
+      "loss": 1.3724,
+      "step": 954
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.720675468444824,
+      "learning_rate": 2.5032940360610266e-05,
+      "loss": 1.1105,
+      "step": 955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.126970291137695,
+      "learning_rate": 2.502773925104022e-05,
+      "loss": 1.0775,
+      "step": 956
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.017008781433105,
+      "learning_rate": 2.5022538141470183e-05,
+      "loss": 0.9295,
+      "step": 957
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.708210945129395,
+      "learning_rate": 2.5017337031900138e-05,
+      "loss": 1.193,
+      "step": 958
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.278105735778809,
+      "learning_rate": 2.5012135922330097e-05,
+      "loss": 1.12,
+      "step": 959
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.904488563537598,
+      "learning_rate": 2.5006934812760055e-05,
+      "loss": 1.1381,
+      "step": 960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.970672607421875,
+      "learning_rate": 2.5001733703190014e-05,
+      "loss": 1.4081,
+      "step": 961
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.657395362854004,
+      "learning_rate": 2.4996532593619972e-05,
+      "loss": 1.4451,
+      "step": 962
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.288087844848633,
+      "learning_rate": 2.499133148404993e-05,
+      "loss": 0.9089,
+      "step": 963
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.267343521118164,
+      "learning_rate": 2.498613037447989e-05,
+      "loss": 1.196,
+      "step": 964
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.301567077636719,
+      "learning_rate": 2.498092926490985e-05,
+      "loss": 1.2113,
+      "step": 965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.985000610351562,
+      "learning_rate": 2.4975728155339807e-05,
+      "loss": 1.069,
+      "step": 966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.715841293334961,
+      "learning_rate": 2.4970527045769766e-05,
+      "loss": 0.9421,
+      "step": 967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.193259239196777,
+      "learning_rate": 2.496532593619972e-05,
+      "loss": 1.3781,
+      "step": 968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.626194953918457,
+      "learning_rate": 2.4960124826629683e-05,
+      "loss": 1.0547,
+      "step": 969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.996138572692871,
+      "learning_rate": 2.4954923717059638e-05,
+      "loss": 1.0488,
+      "step": 970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.520912170410156,
+      "learning_rate": 2.49497226074896e-05,
+      "loss": 1.0963,
+      "step": 971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.89794397354126,
+      "learning_rate": 2.4944521497919555e-05,
+      "loss": 1.0894,
+      "step": 972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.410442352294922,
+      "learning_rate": 2.4939320388349517e-05,
+      "loss": 1.2946,
+      "step": 973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.534004211425781,
+      "learning_rate": 2.4934119278779472e-05,
+      "loss": 0.9378,
+      "step": 974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.224431037902832,
+      "learning_rate": 2.492891816920943e-05,
+      "loss": 1.0569,
+      "step": 975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.730304718017578,
+      "learning_rate": 2.492371705963939e-05,
+      "loss": 1.1976,
+      "step": 976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.569767951965332,
+      "learning_rate": 2.4918515950069348e-05,
+      "loss": 1.0558,
+      "step": 977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.212870597839355,
+      "learning_rate": 2.4913314840499307e-05,
+      "loss": 0.9654,
+      "step": 978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.623671531677246,
+      "learning_rate": 2.4908113730929266e-05,
+      "loss": 1.278,
+      "step": 979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.362083435058594,
+      "learning_rate": 2.4902912621359224e-05,
+      "loss": 1.2954,
+      "step": 980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.127822875976562,
+      "learning_rate": 2.4897711511789183e-05,
+      "loss": 1.0067,
+      "step": 981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.847293853759766,
+      "learning_rate": 2.489251040221914e-05,
+      "loss": 1.031,
+      "step": 982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.626890182495117,
+      "learning_rate": 2.48873092926491e-05,
+      "loss": 1.0236,
+      "step": 983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.844161987304688,
+      "learning_rate": 2.4882108183079055e-05,
+      "loss": 1.2226,
+      "step": 984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.641225814819336,
+      "learning_rate": 2.4876907073509017e-05,
+      "loss": 1.1642,
+      "step": 985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.027875900268555,
+      "learning_rate": 2.4871705963938972e-05,
+      "loss": 1.1528,
+      "step": 986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.219902992248535,
+      "learning_rate": 2.4866504854368934e-05,
+      "loss": 1.0527,
+      "step": 987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.942973136901855,
+      "learning_rate": 2.486130374479889e-05,
+      "loss": 1.3433,
+      "step": 988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.952167510986328,
+      "learning_rate": 2.485610263522885e-05,
+      "loss": 1.2572,
+      "step": 989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.606302261352539,
+      "learning_rate": 2.4850901525658807e-05,
+      "loss": 1.0853,
+      "step": 990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.174311637878418,
+      "learning_rate": 2.4845700416088765e-05,
+      "loss": 1.3437,
+      "step": 991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.517129898071289,
+      "learning_rate": 2.4840499306518724e-05,
+      "loss": 1.3742,
+      "step": 992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.881694793701172,
+      "learning_rate": 2.4835298196948683e-05,
+      "loss": 1.0378,
+      "step": 993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.123956680297852,
+      "learning_rate": 2.483009708737864e-05,
+      "loss": 0.94,
+      "step": 994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.075549125671387,
+      "learning_rate": 2.48248959778086e-05,
+      "loss": 0.8657,
+      "step": 995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.832903861999512,
+      "learning_rate": 2.481969486823856e-05,
+      "loss": 0.8468,
+      "step": 996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.779098510742188,
+      "learning_rate": 2.4814493758668517e-05,
+      "loss": 1.1834,
+      "step": 997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.666390419006348,
+      "learning_rate": 2.4809292649098472e-05,
+      "loss": 0.9814,
+      "step": 998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.82064151763916,
+      "learning_rate": 2.4804091539528434e-05,
+      "loss": 1.3517,
+      "step": 999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.987336158752441,
+      "learning_rate": 2.479889042995839e-05,
+      "loss": 1.2796,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.123810768127441,
+      "learning_rate": 2.479368932038835e-05,
+      "loss": 1.0566,
+      "step": 1001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.618170738220215,
+      "learning_rate": 2.4788488210818307e-05,
+      "loss": 1.1385,
+      "step": 1002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.226157188415527,
+      "learning_rate": 2.478328710124827e-05,
+      "loss": 0.9243,
+      "step": 1003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.024703025817871,
+      "learning_rate": 2.4778085991678224e-05,
+      "loss": 1.175,
+      "step": 1004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.721633911132812,
+      "learning_rate": 2.4772884882108186e-05,
+      "loss": 1.0732,
+      "step": 1005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.2864408493042,
+      "learning_rate": 2.476768377253814e-05,
+      "loss": 1.2958,
+      "step": 1006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.634989738464355,
+      "learning_rate": 2.47624826629681e-05,
+      "loss": 1.2564,
+      "step": 1007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.418289184570312,
+      "learning_rate": 2.475728155339806e-05,
+      "loss": 1.2699,
+      "step": 1008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.35737419128418,
+      "learning_rate": 2.4752080443828017e-05,
+      "loss": 1.039,
+      "step": 1009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.1788969039917,
+      "learning_rate": 2.4746879334257976e-05,
+      "loss": 1.0507,
+      "step": 1010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.342625617980957,
+      "learning_rate": 2.4741678224687934e-05,
+      "loss": 0.9901,
+      "step": 1011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.026543617248535,
+      "learning_rate": 2.4736477115117893e-05,
+      "loss": 1.027,
+      "step": 1012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.265235900878906,
+      "learning_rate": 2.473127600554785e-05,
+      "loss": 1.0241,
+      "step": 1013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.276094436645508,
+      "learning_rate": 2.4726074895977807e-05,
+      "loss": 1.1927,
+      "step": 1014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.864497184753418,
+      "learning_rate": 2.472087378640777e-05,
+      "loss": 1.2668,
+      "step": 1015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.264768600463867,
+      "learning_rate": 2.4715672676837724e-05,
+      "loss": 1.3133,
+      "step": 1016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.0725736618042,
+      "learning_rate": 2.4710471567267686e-05,
+      "loss": 0.9061,
+      "step": 1017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.527504920959473,
+      "learning_rate": 2.470527045769764e-05,
+      "loss": 0.9165,
+      "step": 1018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.371817588806152,
+      "learning_rate": 2.4700069348127603e-05,
+      "loss": 1.0826,
+      "step": 1019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.555102348327637,
+      "learning_rate": 2.469486823855756e-05,
+      "loss": 1.257,
+      "step": 1020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.525373458862305,
+      "learning_rate": 2.4689667128987517e-05,
+      "loss": 0.9613,
+      "step": 1021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.181309700012207,
+      "learning_rate": 2.4684466019417476e-05,
+      "loss": 1.1744,
+      "step": 1022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.791714668273926,
+      "learning_rate": 2.4679264909847434e-05,
+      "loss": 1.0494,
+      "step": 1023
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.982948303222656,
+      "learning_rate": 2.4674063800277393e-05,
+      "loss": 1.2302,
+      "step": 1024
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.092002868652344,
+      "learning_rate": 2.466886269070735e-05,
+      "loss": 1.114,
+      "step": 1025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.838763236999512,
+      "learning_rate": 2.466366158113731e-05,
+      "loss": 1.4183,
+      "step": 1026
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.03321361541748,
+      "learning_rate": 2.465846047156727e-05,
+      "loss": 1.4375,
+      "step": 1027
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.820063591003418,
+      "learning_rate": 2.4653259361997227e-05,
+      "loss": 1.0735,
+      "step": 1028
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.756912231445312,
+      "learning_rate": 2.4648058252427186e-05,
+      "loss": 1.1506,
+      "step": 1029
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.341421127319336,
+      "learning_rate": 2.464285714285714e-05,
+      "loss": 1.0948,
+      "step": 1030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.801969528198242,
+      "learning_rate": 2.4637656033287103e-05,
+      "loss": 1.0057,
+      "step": 1031
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.50317907333374,
+      "learning_rate": 2.463245492371706e-05,
+      "loss": 1.0058,
+      "step": 1032
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.758399963378906,
+      "learning_rate": 2.462725381414702e-05,
+      "loss": 1.172,
+      "step": 1033
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.504880905151367,
+      "learning_rate": 2.4622052704576976e-05,
+      "loss": 1.0348,
+      "step": 1034
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.030147552490234,
+      "learning_rate": 2.4616851595006938e-05,
+      "loss": 1.5774,
+      "step": 1035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.725292205810547,
+      "learning_rate": 2.4611650485436893e-05,
+      "loss": 0.9994,
+      "step": 1036
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.23112964630127,
+      "learning_rate": 2.460644937586685e-05,
+      "loss": 0.9728,
+      "step": 1037
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.539372444152832,
+      "learning_rate": 2.460124826629681e-05,
+      "loss": 1.0321,
+      "step": 1038
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.27584457397461,
+      "learning_rate": 2.459604715672677e-05,
+      "loss": 1.2291,
+      "step": 1039
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.715142726898193,
+      "learning_rate": 2.4590846047156727e-05,
+      "loss": 0.9589,
+      "step": 1040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.366896629333496,
+      "learning_rate": 2.4585644937586686e-05,
+      "loss": 1.0161,
+      "step": 1041
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.83417272567749,
+      "learning_rate": 2.4580443828016645e-05,
+      "loss": 0.8507,
+      "step": 1042
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.914201736450195,
+      "learning_rate": 2.4575242718446603e-05,
+      "loss": 1.0624,
+      "step": 1043
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.985934257507324,
+      "learning_rate": 2.457004160887656e-05,
+      "loss": 1.2146,
+      "step": 1044
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.741414070129395,
+      "learning_rate": 2.456484049930652e-05,
+      "loss": 1.0748,
+      "step": 1045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.566314697265625,
+      "learning_rate": 2.4559639389736476e-05,
+      "loss": 1.1447,
+      "step": 1046
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.026985168457031,
+      "learning_rate": 2.4554438280166438e-05,
+      "loss": 1.0758,
+      "step": 1047
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.20632266998291,
+      "learning_rate": 2.4549237170596393e-05,
+      "loss": 1.1282,
+      "step": 1048
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.181867599487305,
+      "learning_rate": 2.4544036061026355e-05,
+      "loss": 0.973,
+      "step": 1049
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.265329360961914,
+      "learning_rate": 2.453883495145631e-05,
+      "loss": 1.2828,
+      "step": 1050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.085383415222168,
+      "learning_rate": 2.4533633841886272e-05,
+      "loss": 1.1528,
+      "step": 1051
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.86031436920166,
+      "learning_rate": 2.4528432732316227e-05,
+      "loss": 0.8715,
+      "step": 1052
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.629022598266602,
+      "learning_rate": 2.4523231622746186e-05,
+      "loss": 1.09,
+      "step": 1053
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.126294136047363,
+      "learning_rate": 2.4518030513176144e-05,
+      "loss": 0.9308,
+      "step": 1054
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.041877746582031,
+      "learning_rate": 2.4512829403606103e-05,
+      "loss": 0.8367,
+      "step": 1055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.501529693603516,
+      "learning_rate": 2.4507628294036062e-05,
+      "loss": 0.8986,
+      "step": 1056
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.171121597290039,
+      "learning_rate": 2.450242718446602e-05,
+      "loss": 1.0259,
+      "step": 1057
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.340670585632324,
+      "learning_rate": 2.449722607489598e-05,
+      "loss": 1.1911,
+      "step": 1058
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.52405834197998,
+      "learning_rate": 2.4492024965325938e-05,
+      "loss": 1.1545,
+      "step": 1059
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.735933303833008,
+      "learning_rate": 2.4486823855755893e-05,
+      "loss": 1.0609,
+      "step": 1060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.766894340515137,
+      "learning_rate": 2.4481622746185855e-05,
+      "loss": 1.1219,
+      "step": 1061
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.504926681518555,
+      "learning_rate": 2.447642163661581e-05,
+      "loss": 1.1772,
+      "step": 1062
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.092659950256348,
+      "learning_rate": 2.4471220527045772e-05,
+      "loss": 0.9835,
+      "step": 1063
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.04471206665039,
+      "learning_rate": 2.4466019417475727e-05,
+      "loss": 1.1203,
+      "step": 1064
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.268803596496582,
+      "learning_rate": 2.446081830790569e-05,
+      "loss": 1.1643,
+      "step": 1065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.073554992675781,
+      "learning_rate": 2.4455617198335644e-05,
+      "loss": 1.2833,
+      "step": 1066
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.308940887451172,
+      "learning_rate": 2.4450416088765603e-05,
+      "loss": 1.1885,
+      "step": 1067
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.012957572937012,
+      "learning_rate": 2.444521497919556e-05,
+      "loss": 1.3747,
+      "step": 1068
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.909781455993652,
+      "learning_rate": 2.444001386962552e-05,
+      "loss": 1.0999,
+      "step": 1069
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.929265022277832,
+      "learning_rate": 2.443481276005548e-05,
+      "loss": 1.1246,
+      "step": 1070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.58456802368164,
+      "learning_rate": 2.4429611650485438e-05,
+      "loss": 1.2431,
+      "step": 1071
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.826704025268555,
+      "learning_rate": 2.4424410540915396e-05,
+      "loss": 1.1727,
+      "step": 1072
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.28188705444336,
+      "learning_rate": 2.4419209431345355e-05,
+      "loss": 1.1979,
+      "step": 1073
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.940963745117188,
+      "learning_rate": 2.4414008321775313e-05,
+      "loss": 1.5762,
+      "step": 1074
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.420486450195312,
+      "learning_rate": 2.4408807212205272e-05,
+      "loss": 1.053,
+      "step": 1075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.389084815979004,
+      "learning_rate": 2.4403606102635227e-05,
+      "loss": 0.9294,
+      "step": 1076
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.52692985534668,
+      "learning_rate": 2.439840499306519e-05,
+      "loss": 1.244,
+      "step": 1077
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.708897113800049,
+      "learning_rate": 2.4393203883495144e-05,
+      "loss": 1.0711,
+      "step": 1078
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.215950965881348,
+      "learning_rate": 2.4388002773925106e-05,
+      "loss": 1.1457,
+      "step": 1079
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.291402816772461,
+      "learning_rate": 2.438280166435506e-05,
+      "loss": 1.0162,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.87237548828125,
+      "learning_rate": 2.4377600554785024e-05,
+      "loss": 1.1172,
+      "step": 1081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.507237434387207,
+      "learning_rate": 2.437239944521498e-05,
+      "loss": 0.913,
+      "step": 1082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.333834648132324,
+      "learning_rate": 2.4367198335644937e-05,
+      "loss": 1.1238,
+      "step": 1083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.693390846252441,
+      "learning_rate": 2.4361997226074896e-05,
+      "loss": 1.5313,
+      "step": 1084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.882821083068848,
+      "learning_rate": 2.4356796116504855e-05,
+      "loss": 1.4448,
+      "step": 1085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.85878849029541,
+      "learning_rate": 2.4351595006934813e-05,
+      "loss": 1.2379,
+      "step": 1086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.941121101379395,
+      "learning_rate": 2.4346393897364772e-05,
+      "loss": 1.1856,
+      "step": 1087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.795541763305664,
+      "learning_rate": 2.434119278779473e-05,
+      "loss": 0.9256,
+      "step": 1088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.0215425491333,
+      "learning_rate": 2.433599167822469e-05,
+      "loss": 1.1996,
+      "step": 1089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.123217582702637,
+      "learning_rate": 2.4330790568654648e-05,
+      "loss": 1.4386,
+      "step": 1090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.72570276260376,
+      "learning_rate": 2.4325589459084606e-05,
+      "loss": 0.9642,
+      "step": 1091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.576109886169434,
+      "learning_rate": 2.432038834951456e-05,
+      "loss": 0.9953,
+      "step": 1092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.179551124572754,
+      "learning_rate": 2.4315187239944524e-05,
+      "loss": 1.1983,
+      "step": 1093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.801851272583008,
+      "learning_rate": 2.430998613037448e-05,
+      "loss": 1.0367,
+      "step": 1094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.542741775512695,
+      "learning_rate": 2.430478502080444e-05,
+      "loss": 0.9967,
+      "step": 1095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.36408519744873,
+      "learning_rate": 2.4299583911234396e-05,
+      "loss": 1.1361,
+      "step": 1096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.644014358520508,
+      "learning_rate": 2.4294382801664358e-05,
+      "loss": 1.155,
+      "step": 1097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.076963424682617,
+      "learning_rate": 2.4289181692094313e-05,
+      "loss": 1.2028,
+      "step": 1098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.53864860534668,
+      "learning_rate": 2.4283980582524272e-05,
+      "loss": 1.3405,
+      "step": 1099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.777440071105957,
+      "learning_rate": 2.427877947295423e-05,
+      "loss": 1.1368,
+      "step": 1100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.756475448608398,
+      "learning_rate": 2.427357836338419e-05,
+      "loss": 1.1581,
+      "step": 1101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.373846054077148,
+      "learning_rate": 2.4268377253814148e-05,
+      "loss": 1.0969,
+      "step": 1102
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.722124099731445,
+      "learning_rate": 2.4263176144244106e-05,
+      "loss": 1.3536,
+      "step": 1103
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.702287197113037,
+      "learning_rate": 2.4257975034674065e-05,
+      "loss": 0.9448,
+      "step": 1104
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.817178726196289,
+      "learning_rate": 2.4252773925104024e-05,
+      "loss": 1.1317,
+      "step": 1105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.63927173614502,
+      "learning_rate": 2.424757281553398e-05,
+      "loss": 1.3797,
+      "step": 1106
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.519260883331299,
+      "learning_rate": 2.424237170596394e-05,
+      "loss": 0.8949,
+      "step": 1107
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.542991638183594,
+      "learning_rate": 2.4237170596393896e-05,
+      "loss": 1.2899,
+      "step": 1108
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.340275764465332,
+      "learning_rate": 2.4231969486823858e-05,
+      "loss": 1.3509,
+      "step": 1109
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.893980979919434,
+      "learning_rate": 2.4226768377253813e-05,
+      "loss": 1.2167,
+      "step": 1110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.78421688079834,
+      "learning_rate": 2.4221567267683775e-05,
+      "loss": 1.1153,
+      "step": 1111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.113122940063477,
+      "learning_rate": 2.421636615811373e-05,
+      "loss": 1.5261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.420022964477539,
+      "learning_rate": 2.4211165048543692e-05,
+      "loss": 1.2301,
+      "step": 1113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.237869262695312,
+      "learning_rate": 2.4205963938973648e-05,
+      "loss": 1.1088,
+      "step": 1114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.11514663696289,
+      "learning_rate": 2.4200762829403606e-05,
+      "loss": 1.274,
+      "step": 1115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.46660327911377,
+      "learning_rate": 2.4195561719833565e-05,
+      "loss": 1.2447,
+      "step": 1116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.6835551261901855,
+      "learning_rate": 2.4190360610263524e-05,
+      "loss": 1.02,
+      "step": 1117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.782994270324707,
+      "learning_rate": 2.4185159500693482e-05,
+      "loss": 1.1247,
+      "step": 1118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.751260757446289,
+      "learning_rate": 2.417995839112344e-05,
+      "loss": 1.007,
+      "step": 1119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.246830940246582,
+      "learning_rate": 2.41747572815534e-05,
+      "loss": 1.2054,
+      "step": 1120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.524815559387207,
+      "learning_rate": 2.4169556171983358e-05,
+      "loss": 1.1515,
+      "step": 1121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.020284652709961,
+      "learning_rate": 2.4164355062413313e-05,
+      "loss": 0.9278,
+      "step": 1122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.182473182678223,
+      "learning_rate": 2.4159153952843275e-05,
+      "loss": 1.0884,
+      "step": 1123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.089916229248047,
+      "learning_rate": 2.415395284327323e-05,
+      "loss": 1.153,
+      "step": 1124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.305865287780762,
+      "learning_rate": 2.4148751733703192e-05,
+      "loss": 1.0672,
+      "step": 1125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.837860107421875,
+      "learning_rate": 2.4143550624133148e-05,
+      "loss": 1.2601,
+      "step": 1126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.737589836120605,
+      "learning_rate": 2.413834951456311e-05,
+      "loss": 1.4116,
+      "step": 1127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.277080535888672,
+      "learning_rate": 2.4133148404993065e-05,
+      "loss": 1.2046,
+      "step": 1128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.174736022949219,
+      "learning_rate": 2.4127947295423023e-05,
+      "loss": 1.0667,
+      "step": 1129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.308316230773926,
+      "learning_rate": 2.4122746185852982e-05,
+      "loss": 0.9291,
+      "step": 1130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.660791397094727,
+      "learning_rate": 2.411754507628294e-05,
+      "loss": 1.2822,
+      "step": 1131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.834282875061035,
+      "learning_rate": 2.41123439667129e-05,
+      "loss": 0.9361,
+      "step": 1132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.625434875488281,
+      "learning_rate": 2.4107142857142858e-05,
+      "loss": 0.9491,
+      "step": 1133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.672274589538574,
+      "learning_rate": 2.4101941747572817e-05,
+      "loss": 1.06,
+      "step": 1134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.098197937011719,
+      "learning_rate": 2.4096740638002775e-05,
+      "loss": 1.1269,
+      "step": 1135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.68520450592041,
+      "learning_rate": 2.4091539528432734e-05,
+      "loss": 0.7465,
+      "step": 1136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.59404468536377,
+      "learning_rate": 2.4086338418862692e-05,
+      "loss": 0.8978,
+      "step": 1137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.088608741760254,
+      "learning_rate": 2.4081137309292648e-05,
+      "loss": 0.8554,
+      "step": 1138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.222411155700684,
+      "learning_rate": 2.407593619972261e-05,
+      "loss": 1.0137,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.211467742919922,
+      "learning_rate": 2.4070735090152565e-05,
+      "loss": 1.1854,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.382680892944336,
+      "learning_rate": 2.4065533980582527e-05,
+      "loss": 1.2403,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.017969131469727,
+      "learning_rate": 2.4060332871012482e-05,
+      "loss": 0.9388,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.538640022277832,
+      "learning_rate": 2.4055131761442444e-05,
+      "loss": 1.2075,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.590061664581299,
+      "learning_rate": 2.40499306518724e-05,
+      "loss": 0.9346,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.916756629943848,
+      "learning_rate": 2.4044729542302358e-05,
+      "loss": 1.0115,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.811272621154785,
+      "learning_rate": 2.4039528432732317e-05,
+      "loss": 1.1074,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.036528587341309,
+      "learning_rate": 2.4034327323162275e-05,
+      "loss": 0.7804,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.479144096374512,
+      "learning_rate": 2.4029126213592234e-05,
+      "loss": 1.0645,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.507318496704102,
+      "learning_rate": 2.4023925104022192e-05,
+      "loss": 1.129,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.612815856933594,
+      "learning_rate": 2.401872399445215e-05,
+      "loss": 1.0142,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.013586044311523,
+      "learning_rate": 2.401352288488211e-05,
+      "loss": 1.1626,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.889484882354736,
+      "learning_rate": 2.4008321775312068e-05,
+      "loss": 0.9979,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.065937042236328,
+      "learning_rate": 2.4003120665742027e-05,
+      "loss": 1.0879,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.342282295227051,
+      "learning_rate": 2.3997919556171982e-05,
+      "loss": 0.9227,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.012101173400879,
+      "learning_rate": 2.3992718446601944e-05,
+      "loss": 1.4276,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.782224655151367,
+      "learning_rate": 2.39875173370319e-05,
+      "loss": 1.0911,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.037576675415039,
+      "learning_rate": 2.398231622746186e-05,
+      "loss": 0.8669,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.523233413696289,
+      "learning_rate": 2.3977115117891816e-05,
+      "loss": 1.1692,
+      "step": 1158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.87097454071045,
+      "learning_rate": 2.397191400832178e-05,
+      "loss": 1.1726,
+      "step": 1159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.375226974487305,
+      "learning_rate": 2.3966712898751734e-05,
+      "loss": 1.2817,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.36783504486084,
+      "learning_rate": 2.3961511789181692e-05,
+      "loss": 1.034,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.328276634216309,
+      "learning_rate": 2.395631067961165e-05,
+      "loss": 1.2894,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.13671588897705,
+      "learning_rate": 2.395110957004161e-05,
+      "loss": 1.1535,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.682068824768066,
+      "learning_rate": 2.3945908460471568e-05,
+      "loss": 0.9866,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.748922348022461,
+      "learning_rate": 2.3940707350901527e-05,
+      "loss": 1.1479,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.274221420288086,
+      "learning_rate": 2.3935506241331485e-05,
+      "loss": 0.9959,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.725530624389648,
+      "learning_rate": 2.3930305131761444e-05,
+      "loss": 1.3972,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.407389640808105,
+      "learning_rate": 2.39251040221914e-05,
+      "loss": 0.9717,
+      "step": 1168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.43239974975586,
+      "learning_rate": 2.391990291262136e-05,
+      "loss": 1.3704,
+      "step": 1169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.855135917663574,
+      "learning_rate": 2.3914701803051316e-05,
+      "loss": 1.4062,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.510525703430176,
+      "learning_rate": 2.390950069348128e-05,
+      "loss": 1.4227,
+      "step": 1171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.531473159790039,
+      "learning_rate": 2.3904299583911234e-05,
+      "loss": 1.201,
+      "step": 1172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.590426445007324,
+      "learning_rate": 2.3899098474341196e-05,
+      "loss": 1.2005,
+      "step": 1173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.642756462097168,
+      "learning_rate": 2.389389736477115e-05,
+      "loss": 0.9738,
+      "step": 1174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.821189880371094,
+      "learning_rate": 2.3888696255201113e-05,
+      "loss": 1.2305,
+      "step": 1175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.640848159790039,
+      "learning_rate": 2.3883495145631068e-05,
+      "loss": 0.969,
+      "step": 1176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.572020053863525,
+      "learning_rate": 2.3878294036061027e-05,
+      "loss": 0.9315,
+      "step": 1177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.473845481872559,
+      "learning_rate": 2.3873092926490985e-05,
+      "loss": 0.9998,
+      "step": 1178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.2108793258667,
+      "learning_rate": 2.3867891816920944e-05,
+      "loss": 1.0261,
+      "step": 1179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.466670989990234,
+      "learning_rate": 2.3862690707350903e-05,
+      "loss": 1.1564,
+      "step": 1180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.330925941467285,
+      "learning_rate": 2.385748959778086e-05,
+      "loss": 1.1804,
+      "step": 1181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.133946418762207,
+      "learning_rate": 2.385228848821082e-05,
+      "loss": 0.9253,
+      "step": 1182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.068821907043457,
+      "learning_rate": 2.384708737864078e-05,
+      "loss": 1.2555,
+      "step": 1183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.222344398498535,
+      "learning_rate": 2.3841886269070734e-05,
+      "loss": 1.2266,
+      "step": 1184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.432536125183105,
+      "learning_rate": 2.3836685159500696e-05,
+      "loss": 0.9611,
+      "step": 1185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.089869499206543,
+      "learning_rate": 2.383148404993065e-05,
+      "loss": 1.1871,
+      "step": 1186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.460118293762207,
+      "learning_rate": 2.3826282940360613e-05,
+      "loss": 0.8947,
+      "step": 1187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.442312240600586,
+      "learning_rate": 2.3821081830790568e-05,
+      "loss": 1.0141,
+      "step": 1188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.029494285583496,
+      "learning_rate": 2.381588072122053e-05,
+      "loss": 1.0516,
+      "step": 1189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.022995948791504,
+      "learning_rate": 2.3810679611650485e-05,
+      "loss": 1.0564,
+      "step": 1190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.339932441711426,
+      "learning_rate": 2.3805478502080444e-05,
+      "loss": 1.0154,
+      "step": 1191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.115042686462402,
+      "learning_rate": 2.3800277392510403e-05,
+      "loss": 1.2647,
+      "step": 1192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.497125148773193,
+      "learning_rate": 2.379507628294036e-05,
+      "loss": 1.0132,
+      "step": 1193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.274333000183105,
+      "learning_rate": 2.378987517337032e-05,
+      "loss": 0.914,
+      "step": 1194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.603759765625,
+      "learning_rate": 2.378467406380028e-05,
+      "loss": 1.107,
+      "step": 1195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.571268081665039,
+      "learning_rate": 2.3779472954230237e-05,
+      "loss": 1.1238,
+      "step": 1196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.976668357849121,
+      "learning_rate": 2.3774271844660196e-05,
+      "loss": 1.1393,
+      "step": 1197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.990774154663086,
+      "learning_rate": 2.3769070735090154e-05,
+      "loss": 1.2335,
+      "step": 1198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.431981563568115,
+      "learning_rate": 2.3763869625520113e-05,
+      "loss": 0.9305,
+      "step": 1199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.552568435668945,
+      "learning_rate": 2.3758668515950068e-05,
+      "loss": 1.3516,
+      "step": 1200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.056077003479004,
+      "learning_rate": 2.375346740638003e-05,
+      "loss": 1.0237,
+      "step": 1201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.981109619140625,
+      "learning_rate": 2.3748266296809985e-05,
+      "loss": 1.5921,
+      "step": 1202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.443930625915527,
+      "learning_rate": 2.3743065187239947e-05,
+      "loss": 1.0152,
+      "step": 1203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.7150774002075195,
+      "learning_rate": 2.3737864077669902e-05,
+      "loss": 0.8389,
+      "step": 1204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.989673614501953,
+      "learning_rate": 2.3732662968099864e-05,
+      "loss": 1.5053,
+      "step": 1205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.148493766784668,
+      "learning_rate": 2.372746185852982e-05,
+      "loss": 1.2977,
+      "step": 1206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.171385765075684,
+      "learning_rate": 2.372226074895978e-05,
+      "loss": 0.8553,
+      "step": 1207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.461333274841309,
+      "learning_rate": 2.3717059639389737e-05,
+      "loss": 0.9436,
+      "step": 1208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.484066009521484,
+      "learning_rate": 2.3711858529819696e-05,
+      "loss": 1.0668,
+      "step": 1209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.399192810058594,
+      "learning_rate": 2.3706657420249654e-05,
+      "loss": 1.1381,
+      "step": 1210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.247984886169434,
+      "learning_rate": 2.3701456310679613e-05,
+      "loss": 1.1486,
+      "step": 1211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.47590446472168,
+      "learning_rate": 2.369625520110957e-05,
+      "loss": 1.0189,
+      "step": 1212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.393614768981934,
+      "learning_rate": 2.369105409153953e-05,
+      "loss": 1.1711,
+      "step": 1213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.59646987915039,
+      "learning_rate": 2.3685852981969485e-05,
+      "loss": 0.8645,
+      "step": 1214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.654733657836914,
+      "learning_rate": 2.3680651872399447e-05,
+      "loss": 1.1297,
+      "step": 1215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.508956909179688,
+      "learning_rate": 2.3675450762829402e-05,
+      "loss": 1.4101,
+      "step": 1216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.38708209991455,
+      "learning_rate": 2.3670249653259364e-05,
+      "loss": 1.1795,
+      "step": 1217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.311931610107422,
+      "learning_rate": 2.366504854368932e-05,
+      "loss": 1.1618,
+      "step": 1218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.355271339416504,
+      "learning_rate": 2.365984743411928e-05,
+      "loss": 1.0581,
+      "step": 1219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.80959415435791,
+      "learning_rate": 2.3654646324549237e-05,
+      "loss": 0.9306,
+      "step": 1220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.943846702575684,
+      "learning_rate": 2.36494452149792e-05,
+      "loss": 1.2703,
+      "step": 1221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.743581771850586,
+      "learning_rate": 2.3644244105409154e-05,
+      "loss": 1.159,
+      "step": 1222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.717692375183105,
+      "learning_rate": 2.3639042995839113e-05,
+      "loss": 0.9446,
+      "step": 1223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.457610130310059,
+      "learning_rate": 2.363384188626907e-05,
+      "loss": 1.0337,
+      "step": 1224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.194137573242188,
+      "learning_rate": 2.362864077669903e-05,
+      "loss": 0.7838,
+      "step": 1225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.252030372619629,
+      "learning_rate": 2.362343966712899e-05,
+      "loss": 0.8914,
+      "step": 1226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.27846622467041,
+      "learning_rate": 2.3618238557558947e-05,
+      "loss": 0.9535,
+      "step": 1227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.068190574645996,
+      "learning_rate": 2.3613037447988906e-05,
+      "loss": 1.1305,
+      "step": 1228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.952171325683594,
+      "learning_rate": 2.3607836338418864e-05,
+      "loss": 0.9433,
+      "step": 1229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.967394828796387,
+      "learning_rate": 2.360263522884882e-05,
+      "loss": 1.0745,
+      "step": 1230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.855907440185547,
+      "learning_rate": 2.359743411927878e-05,
+      "loss": 0.9864,
+      "step": 1231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.692862510681152,
+      "learning_rate": 2.3592233009708737e-05,
+      "loss": 1.2283,
+      "step": 1232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.340595245361328,
+      "learning_rate": 2.35870319001387e-05,
+      "loss": 1.107,
+      "step": 1233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.825403213500977,
+      "learning_rate": 2.3581830790568654e-05,
+      "loss": 1.1345,
+      "step": 1234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.333578109741211,
+      "learning_rate": 2.3576629680998616e-05,
+      "loss": 0.9995,
+      "step": 1235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.830286026000977,
+      "learning_rate": 2.357142857142857e-05,
+      "loss": 1.1312,
+      "step": 1236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.355093955993652,
+      "learning_rate": 2.3566227461858533e-05,
+      "loss": 1.2679,
+      "step": 1237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.924578666687012,
+      "learning_rate": 2.356102635228849e-05,
+      "loss": 1.2251,
+      "step": 1238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.760751724243164,
+      "learning_rate": 2.3555825242718447e-05,
+      "loss": 1.1256,
+      "step": 1239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.383611679077148,
+      "learning_rate": 2.3550624133148406e-05,
+      "loss": 0.9163,
+      "step": 1240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.290586471557617,
+      "learning_rate": 2.3545423023578364e-05,
+      "loss": 1.1117,
+      "step": 1241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.448060035705566,
+      "learning_rate": 2.3540221914008323e-05,
+      "loss": 1.1042,
+      "step": 1242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.17648983001709,
+      "learning_rate": 2.353502080443828e-05,
+      "loss": 0.9309,
+      "step": 1243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.363167762756348,
+      "learning_rate": 2.352981969486824e-05,
+      "loss": 0.9133,
+      "step": 1244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.177379608154297,
+      "learning_rate": 2.35246185852982e-05,
+      "loss": 1.3058,
+      "step": 1245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.346177101135254,
+      "learning_rate": 2.3519417475728154e-05,
+      "loss": 1.1128,
+      "step": 1246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.475929260253906,
+      "learning_rate": 2.3514216366158116e-05,
+      "loss": 1.1725,
+      "step": 1247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.509500026702881,
+      "learning_rate": 2.350901525658807e-05,
+      "loss": 0.8733,
+      "step": 1248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.9058308601379395,
+      "learning_rate": 2.3503814147018033e-05,
+      "loss": 0.9263,
+      "step": 1249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.062189102172852,
+      "learning_rate": 2.349861303744799e-05,
+      "loss": 0.9751,
+      "step": 1250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.07795238494873,
+      "learning_rate": 2.349341192787795e-05,
+      "loss": 1.0535,
+      "step": 1251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.404336929321289,
+      "learning_rate": 2.3488210818307906e-05,
+      "loss": 1.2915,
+      "step": 1252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.236377716064453,
+      "learning_rate": 2.3483009708737864e-05,
+      "loss": 1.1673,
+      "step": 1253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.207799911499023,
+      "learning_rate": 2.3477808599167823e-05,
+      "loss": 1.185,
+      "step": 1254
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.197190284729004,
+      "learning_rate": 2.347260748959778e-05,
+      "loss": 1.2629,
+      "step": 1255
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.299299240112305,
+      "learning_rate": 2.346740638002774e-05,
+      "loss": 1.0992,
+      "step": 1256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.60837459564209,
+      "learning_rate": 2.34622052704577e-05,
+      "loss": 1.1311,
+      "step": 1257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.604754447937012,
+      "learning_rate": 2.3457004160887657e-05,
+      "loss": 1.0449,
+      "step": 1258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.322650909423828,
+      "learning_rate": 2.3451803051317616e-05,
+      "loss": 1.0046,
+      "step": 1259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.921939849853516,
+      "learning_rate": 2.3446601941747575e-05,
+      "loss": 1.2151,
+      "step": 1260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.702530860900879,
+      "learning_rate": 2.3441400832177533e-05,
+      "loss": 1.1697,
+      "step": 1261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.461334228515625,
+      "learning_rate": 2.343619972260749e-05,
+      "loss": 0.9839,
+      "step": 1262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.252026557922363,
+      "learning_rate": 2.343099861303745e-05,
+      "loss": 1.0668,
+      "step": 1263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.6114091873168945,
+      "learning_rate": 2.3425797503467406e-05,
+      "loss": 0.9287,
+      "step": 1264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.058842658996582,
+      "learning_rate": 2.3420596393897368e-05,
+      "loss": 1.0005,
+      "step": 1265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.0571870803833,
+      "learning_rate": 2.3415395284327323e-05,
+      "loss": 1.0696,
+      "step": 1266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.875245094299316,
+      "learning_rate": 2.3410194174757285e-05,
+      "loss": 1.5693,
+      "step": 1267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.88283920288086,
+      "learning_rate": 2.340499306518724e-05,
+      "loss": 1.0236,
+      "step": 1268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.316110610961914,
+      "learning_rate": 2.33997919556172e-05,
+      "loss": 1.2925,
+      "step": 1269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.383142948150635,
+      "learning_rate": 2.3394590846047157e-05,
+      "loss": 0.8741,
+      "step": 1270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.273981094360352,
+      "learning_rate": 2.3389389736477116e-05,
+      "loss": 1.0541,
+      "step": 1271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.29300594329834,
+      "learning_rate": 2.3384188626907075e-05,
+      "loss": 0.7987,
+      "step": 1272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.059412956237793,
+      "learning_rate": 2.3378987517337033e-05,
+      "loss": 1.0688,
+      "step": 1273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.957596778869629,
+      "learning_rate": 2.3373786407766992e-05,
+      "loss": 1.2219,
+      "step": 1274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.236515998840332,
+      "learning_rate": 2.336858529819695e-05,
+      "loss": 0.9545,
+      "step": 1275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.394058227539062,
+      "learning_rate": 2.3363384188626906e-05,
+      "loss": 1.038,
+      "step": 1276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.381969451904297,
+      "learning_rate": 2.3358183079056868e-05,
+      "loss": 1.3885,
+      "step": 1277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.4012451171875,
+      "learning_rate": 2.3352981969486823e-05,
+      "loss": 1.3224,
+      "step": 1278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.917348861694336,
+      "learning_rate": 2.3347780859916785e-05,
+      "loss": 1.1962,
+      "step": 1279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.396660804748535,
+      "learning_rate": 2.334257975034674e-05,
+      "loss": 0.9996,
+      "step": 1280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.844945907592773,
+      "learning_rate": 2.3337378640776702e-05,
+      "loss": 1.1066,
+      "step": 1281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.999183654785156,
+      "learning_rate": 2.3332177531206657e-05,
+      "loss": 1.1658,
+      "step": 1282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.084219932556152,
+      "learning_rate": 2.332697642163662e-05,
+      "loss": 0.8741,
+      "step": 1283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.442358016967773,
+      "learning_rate": 2.3321775312066575e-05,
+      "loss": 1.2332,
+      "step": 1284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.607652187347412,
+      "learning_rate": 2.3316574202496533e-05,
+      "loss": 1.2351,
+      "step": 1285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.290276527404785,
+      "learning_rate": 2.3311373092926492e-05,
+      "loss": 1.1104,
+      "step": 1286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.6866044998168945,
+      "learning_rate": 2.330617198335645e-05,
+      "loss": 0.9723,
+      "step": 1287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.24308967590332,
+      "learning_rate": 2.330097087378641e-05,
+      "loss": 1.0913,
+      "step": 1288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.105758666992188,
+      "learning_rate": 2.3295769764216368e-05,
+      "loss": 1.2814,
+      "step": 1289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.306723594665527,
+      "learning_rate": 2.3290568654646326e-05,
+      "loss": 1.2134,
+      "step": 1290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.541769981384277,
+      "learning_rate": 2.3285367545076285e-05,
+      "loss": 1.1586,
+      "step": 1291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.296027183532715,
+      "learning_rate": 2.328016643550624e-05,
+      "loss": 1.0052,
+      "step": 1292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.786535263061523,
+      "learning_rate": 2.3274965325936202e-05,
+      "loss": 0.9456,
+      "step": 1293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.503477096557617,
+      "learning_rate": 2.3269764216366157e-05,
+      "loss": 0.9696,
+      "step": 1294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.591896057128906,
+      "learning_rate": 2.326456310679612e-05,
+      "loss": 1.3285,
+      "step": 1295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.798360824584961,
+      "learning_rate": 2.3259361997226074e-05,
+      "loss": 0.9284,
+      "step": 1296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.76888370513916,
+      "learning_rate": 2.3254160887656036e-05,
+      "loss": 1.0843,
+      "step": 1297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.628501892089844,
+      "learning_rate": 2.3248959778085992e-05,
+      "loss": 1.3514,
+      "step": 1298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.853160858154297,
+      "learning_rate": 2.324375866851595e-05,
+      "loss": 1.1638,
+      "step": 1299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.965124130249023,
+      "learning_rate": 2.323855755894591e-05,
+      "loss": 0.9772,
+      "step": 1300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.132728576660156,
+      "learning_rate": 2.3233356449375868e-05,
+      "loss": 1.1367,
+      "step": 1301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.044602394104004,
+      "learning_rate": 2.3228155339805826e-05,
+      "loss": 0.9947,
+      "step": 1302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.205873489379883,
+      "learning_rate": 2.3222954230235785e-05,
+      "loss": 0.9575,
+      "step": 1303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.13265609741211,
+      "learning_rate": 2.3217753120665743e-05,
+      "loss": 1.1565,
+      "step": 1304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.90551471710205,
+      "learning_rate": 2.3212552011095702e-05,
+      "loss": 0.9542,
+      "step": 1305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.722801208496094,
+      "learning_rate": 2.320735090152566e-05,
+      "loss": 1.168,
+      "step": 1306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.317292213439941,
+      "learning_rate": 2.320214979195562e-05,
+      "loss": 1.0746,
+      "step": 1307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.758176803588867,
+      "learning_rate": 2.3196948682385574e-05,
+      "loss": 1.1635,
+      "step": 1308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.692208290100098,
+      "learning_rate": 2.3191747572815536e-05,
+      "loss": 1.2896,
+      "step": 1309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.830174446105957,
+      "learning_rate": 2.318654646324549e-05,
+      "loss": 1.0774,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.671296119689941,
+      "learning_rate": 2.3181345353675454e-05,
+      "loss": 1.2536,
+      "step": 1311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.977215766906738,
+      "learning_rate": 2.317614424410541e-05,
+      "loss": 1.356,
+      "step": 1312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.817078590393066,
+      "learning_rate": 2.317094313453537e-05,
+      "loss": 0.8468,
+      "step": 1313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.786142349243164,
+      "learning_rate": 2.3165742024965326e-05,
+      "loss": 0.8657,
+      "step": 1314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.529461860656738,
+      "learning_rate": 2.3160540915395285e-05,
+      "loss": 0.9163,
+      "step": 1315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.992224216461182,
+      "learning_rate": 2.3155339805825243e-05,
+      "loss": 0.901,
+      "step": 1316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.781744003295898,
+      "learning_rate": 2.3150138696255202e-05,
+      "loss": 1.0633,
+      "step": 1317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.136956214904785,
+      "learning_rate": 2.314493758668516e-05,
+      "loss": 0.8602,
+      "step": 1318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.685998916625977,
+      "learning_rate": 2.313973647711512e-05,
+      "loss": 0.9918,
+      "step": 1319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.281638145446777,
+      "learning_rate": 2.3134535367545078e-05,
+      "loss": 1.2904,
+      "step": 1320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.022212028503418,
+      "learning_rate": 2.3129334257975036e-05,
+      "loss": 0.9157,
+      "step": 1321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.28804874420166,
+      "learning_rate": 2.3124133148404995e-05,
+      "loss": 1.0624,
+      "step": 1322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.589061260223389,
+      "learning_rate": 2.3118932038834954e-05,
+      "loss": 0.7463,
+      "step": 1323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.001674652099609,
+      "learning_rate": 2.311373092926491e-05,
+      "loss": 0.7009,
+      "step": 1324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.158838272094727,
+      "learning_rate": 2.310852981969487e-05,
+      "loss": 1.312,
+      "step": 1325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.729043960571289,
+      "learning_rate": 2.3103328710124826e-05,
+      "loss": 0.956,
+      "step": 1326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.35422420501709,
+      "learning_rate": 2.3098127600554788e-05,
+      "loss": 1.007,
+      "step": 1327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.660632610321045,
+      "learning_rate": 2.3092926490984743e-05,
+      "loss": 0.8923,
+      "step": 1328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.42774486541748,
+      "learning_rate": 2.3087725381414705e-05,
+      "loss": 0.9539,
+      "step": 1329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.63194465637207,
+      "learning_rate": 2.308252427184466e-05,
+      "loss": 0.9942,
+      "step": 1330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.046058654785156,
+      "learning_rate": 2.307732316227462e-05,
+      "loss": 0.8426,
+      "step": 1331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.714208602905273,
+      "learning_rate": 2.3072122052704578e-05,
+      "loss": 1.174,
+      "step": 1332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.150784492492676,
+      "learning_rate": 2.3066920943134536e-05,
+      "loss": 0.998,
+      "step": 1333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.920156478881836,
+      "learning_rate": 2.3061719833564495e-05,
+      "loss": 0.9871,
+      "step": 1334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.724345207214355,
+      "learning_rate": 2.3056518723994454e-05,
+      "loss": 1.1356,
+      "step": 1335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.275041580200195,
+      "learning_rate": 2.3051317614424412e-05,
+      "loss": 1.0688,
+      "step": 1336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.19969654083252,
+      "learning_rate": 2.304611650485437e-05,
+      "loss": 1.0304,
+      "step": 1337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.865304946899414,
+      "learning_rate": 2.3040915395284326e-05,
+      "loss": 1.0809,
+      "step": 1338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.30162525177002,
+      "learning_rate": 2.3035714285714288e-05,
+      "loss": 0.8631,
+      "step": 1339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.094459533691406,
+      "learning_rate": 2.3030513176144243e-05,
+      "loss": 1.2042,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.112343788146973,
+      "learning_rate": 2.3025312066574205e-05,
+      "loss": 0.8875,
+      "step": 1341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.832200527191162,
+      "learning_rate": 2.302011095700416e-05,
+      "loss": 0.8354,
+      "step": 1342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.404396057128906,
+      "learning_rate": 2.3014909847434123e-05,
+      "loss": 1.2688,
+      "step": 1343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.203967094421387,
+      "learning_rate": 2.3009708737864078e-05,
+      "loss": 0.9302,
+      "step": 1344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.179252624511719,
+      "learning_rate": 2.300450762829404e-05,
+      "loss": 1.2544,
+      "step": 1345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.832180976867676,
+      "learning_rate": 2.2999306518723995e-05,
+      "loss": 1.2695,
+      "step": 1346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.990243434906006,
+      "learning_rate": 2.2994105409153954e-05,
+      "loss": 1.0466,
+      "step": 1347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.516494274139404,
+      "learning_rate": 2.2988904299583912e-05,
+      "loss": 0.9457,
+      "step": 1348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.124815464019775,
+      "learning_rate": 2.298370319001387e-05,
+      "loss": 0.9788,
+      "step": 1349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.96698522567749,
+      "learning_rate": 2.297850208044383e-05,
+      "loss": 0.9831,
+      "step": 1350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.302984237670898,
+      "learning_rate": 2.2973300970873788e-05,
+      "loss": 1.1953,
+      "step": 1351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.600886344909668,
+      "learning_rate": 2.2968099861303747e-05,
+      "loss": 1.1145,
+      "step": 1352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.698504447937012,
+      "learning_rate": 2.2962898751733705e-05,
+      "loss": 0.9345,
+      "step": 1353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.916093826293945,
+      "learning_rate": 2.295769764216366e-05,
+      "loss": 0.8273,
+      "step": 1354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.775230407714844,
+      "learning_rate": 2.2952496532593622e-05,
+      "loss": 1.2258,
+      "step": 1355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.462348937988281,
+      "learning_rate": 2.2947295423023578e-05,
+      "loss": 1.0593,
+      "step": 1356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.030835151672363,
+      "learning_rate": 2.294209431345354e-05,
+      "loss": 0.9327,
+      "step": 1357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.478904724121094,
+      "learning_rate": 2.2936893203883495e-05,
+      "loss": 1.1901,
+      "step": 1358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.947803497314453,
+      "learning_rate": 2.2931692094313457e-05,
+      "loss": 1.1667,
+      "step": 1359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.647053718566895,
+      "learning_rate": 2.2926490984743412e-05,
+      "loss": 1.1701,
+      "step": 1360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.604714393615723,
+      "learning_rate": 2.292128987517337e-05,
+      "loss": 1.0226,
+      "step": 1361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.742690086364746,
+      "learning_rate": 2.291608876560333e-05,
+      "loss": 1.0379,
+      "step": 1362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.964884757995605,
+      "learning_rate": 2.2910887656033288e-05,
+      "loss": 0.9677,
+      "step": 1363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.102911949157715,
+      "learning_rate": 2.2905686546463247e-05,
+      "loss": 0.7463,
+      "step": 1364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.411565780639648,
+      "learning_rate": 2.2900485436893205e-05,
+      "loss": 0.9792,
+      "step": 1365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.508044242858887,
+      "learning_rate": 2.2895284327323164e-05,
+      "loss": 0.9859,
+      "step": 1366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.981679916381836,
+      "learning_rate": 2.2890083217753122e-05,
+      "loss": 1.2349,
+      "step": 1367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.517605781555176,
+      "learning_rate": 2.288488210818308e-05,
+      "loss": 0.8696,
+      "step": 1368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.76963996887207,
+      "learning_rate": 2.287968099861304e-05,
+      "loss": 1.1133,
+      "step": 1369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.391816139221191,
+      "learning_rate": 2.2874479889042995e-05,
+      "loss": 0.9627,
+      "step": 1370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.763904571533203,
+      "learning_rate": 2.2869278779472957e-05,
+      "loss": 1.1669,
+      "step": 1371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.367034912109375,
+      "learning_rate": 2.2864077669902912e-05,
+      "loss": 1.1123,
+      "step": 1372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.28658390045166,
+      "learning_rate": 2.2858876560332874e-05,
+      "loss": 1.1702,
+      "step": 1373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.158466339111328,
+      "learning_rate": 2.285367545076283e-05,
+      "loss": 1.0698,
+      "step": 1374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.469143867492676,
+      "learning_rate": 2.284847434119279e-05,
+      "loss": 1.0988,
+      "step": 1375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.338737487792969,
+      "learning_rate": 2.2843273231622747e-05,
+      "loss": 1.1977,
+      "step": 1376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.19676399230957,
+      "learning_rate": 2.2838072122052705e-05,
+      "loss": 1.1971,
+      "step": 1377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.603156089782715,
+      "learning_rate": 2.2832871012482664e-05,
+      "loss": 1.1384,
+      "step": 1378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.138150215148926,
+      "learning_rate": 2.2827669902912622e-05,
+      "loss": 0.9988,
+      "step": 1379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.479331970214844,
+      "learning_rate": 2.282246879334258e-05,
+      "loss": 1.187,
+      "step": 1380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.844976425170898,
+      "learning_rate": 2.281726768377254e-05,
+      "loss": 1.1752,
+      "step": 1381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.487810134887695,
+      "learning_rate": 2.2812066574202498e-05,
+      "loss": 1.2728,
+      "step": 1382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.476121425628662,
+      "learning_rate": 2.2806865464632457e-05,
+      "loss": 0.9389,
+      "step": 1383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.328630447387695,
+      "learning_rate": 2.2801664355062415e-05,
+      "loss": 0.8522,
+      "step": 1384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.808159351348877,
+      "learning_rate": 2.2796463245492374e-05,
+      "loss": 1.0046,
+      "step": 1385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.897589683532715,
+      "learning_rate": 2.279126213592233e-05,
+      "loss": 0.8851,
+      "step": 1386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.843727111816406,
+      "learning_rate": 2.278606102635229e-05,
+      "loss": 0.8608,
+      "step": 1387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.608739852905273,
+      "learning_rate": 2.2780859916782247e-05,
+      "loss": 0.8963,
+      "step": 1388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.096890449523926,
+      "learning_rate": 2.277565880721221e-05,
+      "loss": 1.1002,
+      "step": 1389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.005692481994629,
+      "learning_rate": 2.2770457697642164e-05,
+      "loss": 0.9287,
+      "step": 1390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.297920227050781,
+      "learning_rate": 2.2765256588072126e-05,
+      "loss": 0.9943,
+      "step": 1391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.686956405639648,
+      "learning_rate": 2.276005547850208e-05,
+      "loss": 0.824,
+      "step": 1392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.75847053527832,
+      "learning_rate": 2.275485436893204e-05,
+      "loss": 0.8796,
+      "step": 1393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.44327163696289,
+      "learning_rate": 2.2749653259361998e-05,
+      "loss": 1.055,
+      "step": 1394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.133024215698242,
+      "learning_rate": 2.2744452149791957e-05,
+      "loss": 0.817,
+      "step": 1395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.29399299621582,
+      "learning_rate": 2.2739251040221915e-05,
+      "loss": 1.0556,
+      "step": 1396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.315126419067383,
+      "learning_rate": 2.2734049930651874e-05,
+      "loss": 0.9373,
+      "step": 1397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.372446060180664,
+      "learning_rate": 2.2728848821081833e-05,
+      "loss": 0.8879,
+      "step": 1398
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.894015312194824,
+      "learning_rate": 2.272364771151179e-05,
+      "loss": 0.9891,
+      "step": 1399
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.150132179260254,
+      "learning_rate": 2.2718446601941746e-05,
+      "loss": 1.2077,
+      "step": 1400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.856317520141602,
+      "learning_rate": 2.271324549237171e-05,
+      "loss": 0.9945,
+      "step": 1401
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.501531600952148,
+      "learning_rate": 2.2708044382801664e-05,
+      "loss": 0.9938,
+      "step": 1402
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.236778259277344,
+      "learning_rate": 2.2702843273231626e-05,
+      "loss": 1.0559,
+      "step": 1403
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.717751502990723,
+      "learning_rate": 2.269764216366158e-05,
+      "loss": 1.0174,
+      "step": 1404
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.582865715026855,
+      "learning_rate": 2.2692441054091543e-05,
+      "loss": 1.1046,
+      "step": 1405
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.840753078460693,
+      "learning_rate": 2.2687239944521498e-05,
+      "loss": 0.9027,
+      "step": 1406
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.541910171508789,
+      "learning_rate": 2.268203883495146e-05,
+      "loss": 0.858,
+      "step": 1407
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.6116485595703125,
+      "learning_rate": 2.2676837725381415e-05,
+      "loss": 0.9182,
+      "step": 1408
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.536895751953125,
+      "learning_rate": 2.2671636615811374e-05,
+      "loss": 0.8825,
+      "step": 1409
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.147897243499756,
+      "learning_rate": 2.2666435506241333e-05,
+      "loss": 1.0518,
+      "step": 1410
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.131757736206055,
+      "learning_rate": 2.266123439667129e-05,
+      "loss": 0.9371,
+      "step": 1411
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.130814552307129,
+      "learning_rate": 2.265603328710125e-05,
+      "loss": 1.1732,
+      "step": 1412
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.699939727783203,
+      "learning_rate": 2.265083217753121e-05,
+      "loss": 1.0275,
+      "step": 1413
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.85390567779541,
+      "learning_rate": 2.2645631067961167e-05,
+      "loss": 1.0748,
+      "step": 1414
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.059505462646484,
+      "learning_rate": 2.2640429958391126e-05,
+      "loss": 1.016,
+      "step": 1415
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.568216323852539,
+      "learning_rate": 2.263522884882108e-05,
+      "loss": 0.9201,
+      "step": 1416
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.816947937011719,
+      "learning_rate": 2.2630027739251043e-05,
+      "loss": 1.0479,
+      "step": 1417
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.228033065795898,
+      "learning_rate": 2.2624826629680998e-05,
+      "loss": 1.0496,
+      "step": 1418
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.84034538269043,
+      "learning_rate": 2.261962552011096e-05,
+      "loss": 1.2668,
+      "step": 1419
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.270434379577637,
+      "learning_rate": 2.2614424410540915e-05,
+      "loss": 0.9574,
+      "step": 1420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.211235046386719,
+      "learning_rate": 2.2609223300970877e-05,
+      "loss": 1.1242,
+      "step": 1421
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.206026077270508,
+      "learning_rate": 2.2604022191400833e-05,
+      "loss": 0.9559,
+      "step": 1422
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.001397132873535,
+      "learning_rate": 2.259882108183079e-05,
+      "loss": 0.988,
+      "step": 1423
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.962864398956299,
+      "learning_rate": 2.259361997226075e-05,
+      "loss": 0.8649,
+      "step": 1424
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.528092384338379,
+      "learning_rate": 2.258841886269071e-05,
+      "loss": 1.1687,
+      "step": 1425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.0927734375,
+      "learning_rate": 2.2583217753120667e-05,
+      "loss": 1.1177,
+      "step": 1426
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.459978103637695,
+      "learning_rate": 2.2578016643550626e-05,
+      "loss": 0.9998,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.486920356750488,
+      "learning_rate": 2.2572815533980584e-05,
+      "loss": 1.1385,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.863340377807617,
+      "learning_rate": 2.2567614424410543e-05,
+      "loss": 1.1181,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.25797176361084,
+      "learning_rate": 2.25624133148405e-05,
+      "loss": 1.4947,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.920810699462891,
+      "learning_rate": 2.255721220527046e-05,
+      "loss": 1.1507,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.577263355255127,
+      "learning_rate": 2.2552011095700415e-05,
+      "loss": 1.2015,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.427802085876465,
+      "learning_rate": 2.2546809986130377e-05,
+      "loss": 1.0173,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.8864216804504395,
+      "learning_rate": 2.2541608876560333e-05,
+      "loss": 0.9637,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.268056869506836,
+      "learning_rate": 2.2536407766990295e-05,
+      "loss": 1.0138,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.228147506713867,
+      "learning_rate": 2.253120665742025e-05,
+      "loss": 1.0487,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.683311462402344,
+      "learning_rate": 2.2526005547850212e-05,
+      "loss": 1.1107,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.987105369567871,
+      "learning_rate": 2.2520804438280167e-05,
+      "loss": 1.0377,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.032308578491211,
+      "learning_rate": 2.2515603328710126e-05,
+      "loss": 1.0936,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.416810989379883,
+      "learning_rate": 2.2510402219140084e-05,
+      "loss": 1.1339,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.361400604248047,
+      "learning_rate": 2.2505201109570043e-05,
+      "loss": 1.0418,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.257600784301758,
+      "learning_rate": 2.25e-05,
+      "loss": 0.9715,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.659405708312988,
+      "learning_rate": 2.2494798890429957e-05,
+      "loss": 0.9982,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.017932891845703,
+      "learning_rate": 2.248959778085992e-05,
+      "loss": 0.921,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.5886125564575195,
+      "learning_rate": 2.2484396671289874e-05,
+      "loss": 0.6553,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.980467796325684,
+      "learning_rate": 2.2479195561719832e-05,
+      "loss": 0.7998,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.353504657745361,
+      "learning_rate": 2.247399445214979e-05,
+      "loss": 0.8026,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.6283674240112305,
+      "learning_rate": 2.246879334257975e-05,
+      "loss": 1.0834,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.931901931762695,
+      "learning_rate": 2.246359223300971e-05,
+      "loss": 0.8871,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.095620155334473,
+      "learning_rate": 2.2458391123439667e-05,
+      "loss": 1.2369,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.284668922424316,
+      "learning_rate": 2.2453190013869626e-05,
+      "loss": 0.8124,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.50693416595459,
+      "learning_rate": 2.2447988904299584e-05,
+      "loss": 1.0431,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.662418365478516,
+      "learning_rate": 2.2442787794729543e-05,
+      "loss": 0.7756,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.690422058105469,
+      "learning_rate": 2.24375866851595e-05,
+      "loss": 0.8169,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.109491348266602,
+      "learning_rate": 2.2432385575589457e-05,
+      "loss": 1.0202,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.763873100280762,
+      "learning_rate": 2.242718446601942e-05,
+      "loss": 0.7115,
+      "step": 1456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.554086685180664,
+      "learning_rate": 2.2421983356449374e-05,
+      "loss": 1.1135,
+      "step": 1457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.953583717346191,
+      "learning_rate": 2.2416782246879336e-05,
+      "loss": 1.1528,
+      "step": 1458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.443325996398926,
+      "learning_rate": 2.241158113730929e-05,
+      "loss": 1.5208,
+      "step": 1459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.562949180603027,
+      "learning_rate": 2.2406380027739253e-05,
+      "loss": 1.147,
+      "step": 1460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.276161193847656,
+      "learning_rate": 2.2401178918169208e-05,
+      "loss": 1.032,
+      "step": 1461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.173160552978516,
+      "learning_rate": 2.2395977808599167e-05,
+      "loss": 1.1618,
+      "step": 1462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.034249305725098,
+      "learning_rate": 2.2390776699029126e-05,
+      "loss": 0.8887,
+      "step": 1463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.635499000549316,
+      "learning_rate": 2.2385575589459084e-05,
+      "loss": 0.9459,
+      "step": 1464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.142465591430664,
+      "learning_rate": 2.2380374479889043e-05,
+      "loss": 1.2304,
+      "step": 1465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.973081588745117,
+      "learning_rate": 2.2375173370319e-05,
+      "loss": 1.0193,
+      "step": 1466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.091975212097168,
+      "learning_rate": 2.236997226074896e-05,
+      "loss": 0.9117,
+      "step": 1467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.6797685623168945,
+      "learning_rate": 2.236477115117892e-05,
+      "loss": 0.8438,
+      "step": 1468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.737372398376465,
+      "learning_rate": 2.2359570041608877e-05,
+      "loss": 1.0948,
+      "step": 1469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.149528503417969,
+      "learning_rate": 2.2354368932038836e-05,
+      "loss": 0.9689,
+      "step": 1470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.480108261108398,
+      "learning_rate": 2.234916782246879e-05,
+      "loss": 1.2653,
+      "step": 1471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.520005226135254,
+      "learning_rate": 2.2343966712898753e-05,
+      "loss": 0.9844,
+      "step": 1472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.171473503112793,
+      "learning_rate": 2.2338765603328708e-05,
+      "loss": 0.961,
+      "step": 1473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.034173011779785,
+      "learning_rate": 2.233356449375867e-05,
+      "loss": 0.8032,
+      "step": 1474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.520832061767578,
+      "learning_rate": 2.2328363384188625e-05,
+      "loss": 1.1094,
+      "step": 1475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.267046928405762,
+      "learning_rate": 2.2323162274618587e-05,
+      "loss": 1.0222,
+      "step": 1476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.487750053405762,
+      "learning_rate": 2.2317961165048543e-05,
+      "loss": 1.099,
+      "step": 1477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.80225658416748,
+      "learning_rate": 2.23127600554785e-05,
+      "loss": 0.9993,
+      "step": 1478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.422036170959473,
+      "learning_rate": 2.230755894590846e-05,
+      "loss": 1.0908,
+      "step": 1479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.777751922607422,
+      "learning_rate": 2.230235783633842e-05,
+      "loss": 1.0437,
+      "step": 1480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.511184692382812,
+      "learning_rate": 2.2297156726768377e-05,
+      "loss": 0.813,
+      "step": 1481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.171425819396973,
+      "learning_rate": 2.2291955617198336e-05,
+      "loss": 0.9494,
+      "step": 1482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.797185897827148,
+      "learning_rate": 2.2286754507628294e-05,
+      "loss": 1.0716,
+      "step": 1483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.69419002532959,
+      "learning_rate": 2.2281553398058253e-05,
+      "loss": 1.1345,
+      "step": 1484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.515954971313477,
+      "learning_rate": 2.2276352288488208e-05,
+      "loss": 1.1293,
+      "step": 1485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.527286529541016,
+      "learning_rate": 2.227115117891817e-05,
+      "loss": 0.7783,
+      "step": 1486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.144612312316895,
+      "learning_rate": 2.2265950069348125e-05,
+      "loss": 1.0086,
+      "step": 1487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.534097671508789,
+      "learning_rate": 2.2260748959778087e-05,
+      "loss": 1.1818,
+      "step": 1488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.501457214355469,
+      "learning_rate": 2.2255547850208043e-05,
+      "loss": 1.1665,
+      "step": 1489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.520267486572266,
+      "learning_rate": 2.2250346740638005e-05,
+      "loss": 1.0683,
+      "step": 1490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.602879524230957,
+      "learning_rate": 2.224514563106796e-05,
+      "loss": 1.4147,
+      "step": 1491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.314382553100586,
+      "learning_rate": 2.2239944521497922e-05,
+      "loss": 0.9975,
+      "step": 1492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.036949157714844,
+      "learning_rate": 2.2234743411927877e-05,
+      "loss": 1.176,
+      "step": 1493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.553271770477295,
+      "learning_rate": 2.2229542302357836e-05,
+      "loss": 0.8507,
+      "step": 1494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.7321882247924805,
+      "learning_rate": 2.2224341192787794e-05,
+      "loss": 1.0024,
+      "step": 1495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.204987525939941,
+      "learning_rate": 2.2219140083217753e-05,
+      "loss": 0.9279,
+      "step": 1496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.264081001281738,
+      "learning_rate": 2.221393897364771e-05,
+      "loss": 1.2349,
+      "step": 1497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.156037330627441,
+      "learning_rate": 2.220873786407767e-05,
+      "loss": 1.0557,
+      "step": 1498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.438545227050781,
+      "learning_rate": 2.220353675450763e-05,
+      "loss": 1.0319,
+      "step": 1499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.160483360290527,
+      "learning_rate": 2.2198335644937587e-05,
+      "loss": 1.0311,
+      "step": 1500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.576005935668945,
+      "learning_rate": 2.2193134535367543e-05,
+      "loss": 0.971,
+      "step": 1501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.509010314941406,
+      "learning_rate": 2.2187933425797505e-05,
+      "loss": 0.9782,
+      "step": 1502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.844980239868164,
+      "learning_rate": 2.218273231622746e-05,
+      "loss": 1.2092,
+      "step": 1503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.443933486938477,
+      "learning_rate": 2.2177531206657422e-05,
+      "loss": 0.9991,
+      "step": 1504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.989592552185059,
+      "learning_rate": 2.2172330097087377e-05,
+      "loss": 0.9856,
+      "step": 1505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.106234550476074,
+      "learning_rate": 2.216712898751734e-05,
+      "loss": 1.2354,
+      "step": 1506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.431119918823242,
+      "learning_rate": 2.2161927877947294e-05,
+      "loss": 0.8393,
+      "step": 1507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.707688331604004,
+      "learning_rate": 2.2156726768377253e-05,
+      "loss": 0.7764,
+      "step": 1508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.862491607666016,
+      "learning_rate": 2.215152565880721e-05,
+      "loss": 1.0189,
+      "step": 1509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.402446746826172,
+      "learning_rate": 2.214632454923717e-05,
+      "loss": 0.9292,
+      "step": 1510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.261404037475586,
+      "learning_rate": 2.214112343966713e-05,
+      "loss": 0.8126,
+      "step": 1511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.096583366394043,
+      "learning_rate": 2.2135922330097087e-05,
+      "loss": 0.8764,
+      "step": 1512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.44420051574707,
+      "learning_rate": 2.2130721220527046e-05,
+      "loss": 1.0473,
+      "step": 1513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.817628860473633,
+      "learning_rate": 2.2125520110957005e-05,
+      "loss": 1.0385,
+      "step": 1514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.099286079406738,
+      "learning_rate": 2.2120319001386963e-05,
+      "loss": 1.0449,
+      "step": 1515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.229570388793945,
+      "learning_rate": 2.2115117891816922e-05,
+      "loss": 1.1061,
+      "step": 1516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.953912734985352,
+      "learning_rate": 2.2109916782246877e-05,
+      "loss": 0.8533,
+      "step": 1517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.411382675170898,
+      "learning_rate": 2.210471567267684e-05,
+      "loss": 0.9692,
+      "step": 1518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.891387939453125,
+      "learning_rate": 2.2099514563106794e-05,
+      "loss": 0.9939,
+      "step": 1519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.954206466674805,
+      "learning_rate": 2.2094313453536756e-05,
+      "loss": 0.9918,
+      "step": 1520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.9855375289917,
+      "learning_rate": 2.208911234396671e-05,
+      "loss": 1.2752,
+      "step": 1521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.127771377563477,
+      "learning_rate": 2.2083911234396673e-05,
+      "loss": 1.0441,
+      "step": 1522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.93181037902832,
+      "learning_rate": 2.207871012482663e-05,
+      "loss": 1.0738,
+      "step": 1523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.426911354064941,
+      "learning_rate": 2.2073509015256587e-05,
+      "loss": 1.3534,
+      "step": 1524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.692221641540527,
+      "learning_rate": 2.2068307905686546e-05,
+      "loss": 0.8908,
+      "step": 1525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.48365592956543,
+      "learning_rate": 2.2063106796116505e-05,
+      "loss": 1.0935,
+      "step": 1526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.365015506744385,
+      "learning_rate": 2.2057905686546463e-05,
+      "loss": 0.8988,
+      "step": 1527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.990251541137695,
+      "learning_rate": 2.2052704576976422e-05,
+      "loss": 0.9955,
+      "step": 1528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.653663635253906,
+      "learning_rate": 2.204750346740638e-05,
+      "loss": 1.1738,
+      "step": 1529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.538841247558594,
+      "learning_rate": 2.204230235783634e-05,
+      "loss": 0.8388,
+      "step": 1530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.512665748596191,
+      "learning_rate": 2.2037101248266298e-05,
+      "loss": 1.2213,
+      "step": 1531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.83434772491455,
+      "learning_rate": 2.2031900138696256e-05,
+      "loss": 1.1859,
+      "step": 1532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.327853679656982,
+      "learning_rate": 2.202669902912621e-05,
+      "loss": 0.9314,
+      "step": 1533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.055566787719727,
+      "learning_rate": 2.2021497919556173e-05,
+      "loss": 1.1107,
+      "step": 1534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.323426246643066,
+      "learning_rate": 2.201629680998613e-05,
+      "loss": 1.1299,
+      "step": 1535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.520674705505371,
+      "learning_rate": 2.201109570041609e-05,
+      "loss": 1.1722,
+      "step": 1536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.748078346252441,
+      "learning_rate": 2.2005894590846046e-05,
+      "loss": 0.9746,
+      "step": 1537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.226656913757324,
+      "learning_rate": 2.2000693481276008e-05,
+      "loss": 1.071,
+      "step": 1538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.922107696533203,
+      "learning_rate": 2.1995492371705963e-05,
+      "loss": 1.0532,
+      "step": 1539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.011866569519043,
+      "learning_rate": 2.1990291262135922e-05,
+      "loss": 1.386,
+      "step": 1540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.168937683105469,
+      "learning_rate": 2.198509015256588e-05,
+      "loss": 0.966,
+      "step": 1541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.409341812133789,
+      "learning_rate": 2.197988904299584e-05,
+      "loss": 1.0327,
+      "step": 1542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.451764106750488,
+      "learning_rate": 2.1974687933425798e-05,
+      "loss": 1.0799,
+      "step": 1543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.158644676208496,
+      "learning_rate": 2.1969486823855756e-05,
+      "loss": 1.0766,
+      "step": 1544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.797407150268555,
+      "learning_rate": 2.1964285714285715e-05,
+      "loss": 1.0159,
+      "step": 1545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.949959754943848,
+      "learning_rate": 2.1959084604715673e-05,
+      "loss": 1.0661,
+      "step": 1546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.357747077941895,
+      "learning_rate": 2.195388349514563e-05,
+      "loss": 0.842,
+      "step": 1547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.632048606872559,
+      "learning_rate": 2.194868238557559e-05,
+      "loss": 0.9119,
+      "step": 1548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.59842586517334,
+      "learning_rate": 2.1943481276005546e-05,
+      "loss": 1.23,
+      "step": 1549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.422656059265137,
+      "learning_rate": 2.1938280166435508e-05,
+      "loss": 1.0267,
+      "step": 1550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.523341178894043,
+      "learning_rate": 2.1933079056865463e-05,
+      "loss": 0.78,
+      "step": 1551
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.048563003540039,
+      "learning_rate": 2.1927877947295425e-05,
+      "loss": 1.0274,
+      "step": 1552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.431428909301758,
+      "learning_rate": 2.192267683772538e-05,
+      "loss": 1.0079,
+      "step": 1553
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.648046493530273,
+      "learning_rate": 2.1917475728155342e-05,
+      "loss": 1.0895,
+      "step": 1554
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.49229621887207,
+      "learning_rate": 2.1912274618585298e-05,
+      "loss": 1.477,
+      "step": 1555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.820974349975586,
+      "learning_rate": 2.1907073509015256e-05,
+      "loss": 0.936,
+      "step": 1556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.980937004089355,
+      "learning_rate": 2.1901872399445215e-05,
+      "loss": 1.0907,
+      "step": 1557
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.645615577697754,
+      "learning_rate": 2.1896671289875173e-05,
+      "loss": 0.9362,
+      "step": 1558
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.582886695861816,
+      "learning_rate": 2.1891470180305132e-05,
+      "loss": 1.0923,
+      "step": 1559
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.8737154006958,
+      "learning_rate": 2.188626907073509e-05,
+      "loss": 1.4523,
+      "step": 1560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.57275104522705,
+      "learning_rate": 2.188106796116505e-05,
+      "loss": 0.8,
+      "step": 1561
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.305000305175781,
+      "learning_rate": 2.1875866851595008e-05,
+      "loss": 0.9912,
+      "step": 1562
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.768076419830322,
+      "learning_rate": 2.1870665742024963e-05,
+      "loss": 1.0351,
+      "step": 1563
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.738561153411865,
+      "learning_rate": 2.1865464632454925e-05,
+      "loss": 0.8003,
+      "step": 1564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.980818748474121,
+      "learning_rate": 2.186026352288488e-05,
+      "loss": 1.1299,
+      "step": 1565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.663748741149902,
+      "learning_rate": 2.1855062413314842e-05,
+      "loss": 0.9684,
+      "step": 1566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.401162147521973,
+      "learning_rate": 2.1849861303744797e-05,
+      "loss": 0.901,
+      "step": 1567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.05675983428955,
+      "learning_rate": 2.184466019417476e-05,
+      "loss": 1.2839,
+      "step": 1568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.898779392242432,
+      "learning_rate": 2.1839459084604715e-05,
+      "loss": 0.8175,
+      "step": 1569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.189223289489746,
+      "learning_rate": 2.1834257975034673e-05,
+      "loss": 1.1396,
+      "step": 1570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.153961181640625,
+      "learning_rate": 2.1829056865464632e-05,
+      "loss": 0.8548,
+      "step": 1571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.528942108154297,
+      "learning_rate": 2.182385575589459e-05,
+      "loss": 1.1299,
+      "step": 1572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.129383087158203,
+      "learning_rate": 2.181865464632455e-05,
+      "loss": 0.9284,
+      "step": 1573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.614042282104492,
+      "learning_rate": 2.1813453536754508e-05,
+      "loss": 1.0334,
+      "step": 1574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.856457710266113,
+      "learning_rate": 2.1808252427184466e-05,
+      "loss": 0.872,
+      "step": 1575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.59931468963623,
+      "learning_rate": 2.1803051317614425e-05,
+      "loss": 0.7515,
+      "step": 1576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.659500122070312,
+      "learning_rate": 2.1797850208044384e-05,
+      "loss": 1.2413,
+      "step": 1577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.69355583190918,
+      "learning_rate": 2.1792649098474342e-05,
+      "loss": 1.1086,
+      "step": 1578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.9990620613098145,
+      "learning_rate": 2.1787447988904297e-05,
+      "loss": 0.9258,
+      "step": 1579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.234999656677246,
+      "learning_rate": 2.178224687933426e-05,
+      "loss": 0.9191,
+      "step": 1580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.023730278015137,
+      "learning_rate": 2.1777045769764215e-05,
+      "loss": 0.7892,
+      "step": 1581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.296300888061523,
+      "learning_rate": 2.1771844660194177e-05,
+      "loss": 1.021,
+      "step": 1582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.660843849182129,
+      "learning_rate": 2.1766643550624132e-05,
+      "loss": 0.9192,
+      "step": 1583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.456892013549805,
+      "learning_rate": 2.1761442441054094e-05,
+      "loss": 0.8263,
+      "step": 1584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.629579544067383,
+      "learning_rate": 2.175624133148405e-05,
+      "loss": 0.9692,
+      "step": 1585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.33510684967041,
+      "learning_rate": 2.1751040221914008e-05,
+      "loss": 0.8235,
+      "step": 1586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.746157646179199,
+      "learning_rate": 2.1745839112343966e-05,
+      "loss": 0.8323,
+      "step": 1587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.982109546661377,
+      "learning_rate": 2.1740638002773925e-05,
+      "loss": 0.9284,
+      "step": 1588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.388298988342285,
+      "learning_rate": 2.1735436893203884e-05,
+      "loss": 1.0895,
+      "step": 1589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.567059516906738,
+      "learning_rate": 2.1730235783633842e-05,
+      "loss": 1.2042,
+      "step": 1590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.189192771911621,
+      "learning_rate": 2.17250346740638e-05,
+      "loss": 0.8364,
+      "step": 1591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.365422248840332,
+      "learning_rate": 2.171983356449376e-05,
+      "loss": 1.3045,
+      "step": 1592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.428990364074707,
+      "learning_rate": 2.1714632454923715e-05,
+      "loss": 1.1145,
+      "step": 1593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.830870628356934,
+      "learning_rate": 2.1709431345353677e-05,
+      "loss": 1.0767,
+      "step": 1594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.32997989654541,
+      "learning_rate": 2.1704230235783632e-05,
+      "loss": 1.1175,
+      "step": 1595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.383808135986328,
+      "learning_rate": 2.1699029126213594e-05,
+      "loss": 1.1011,
+      "step": 1596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.074614524841309,
+      "learning_rate": 2.169382801664355e-05,
+      "loss": 0.9203,
+      "step": 1597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.43778133392334,
+      "learning_rate": 2.168862690707351e-05,
+      "loss": 0.8644,
+      "step": 1598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.0752534866333,
+      "learning_rate": 2.1683425797503466e-05,
+      "loss": 0.8446,
+      "step": 1599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.0220947265625,
+      "learning_rate": 2.1678224687933428e-05,
+      "loss": 0.9458,
+      "step": 1600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.362180709838867,
+      "learning_rate": 2.1673023578363384e-05,
+      "loss": 0.7843,
+      "step": 1601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.262948989868164,
+      "learning_rate": 2.1667822468793342e-05,
+      "loss": 1.4149,
+      "step": 1602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.366640090942383,
+      "learning_rate": 2.16626213592233e-05,
+      "loss": 0.762,
+      "step": 1603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.15239429473877,
+      "learning_rate": 2.165742024965326e-05,
+      "loss": 1.0349,
+      "step": 1604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.624871253967285,
+      "learning_rate": 2.1652219140083218e-05,
+      "loss": 1.0013,
+      "step": 1605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.821836471557617,
+      "learning_rate": 2.1647018030513177e-05,
+      "loss": 0.8563,
+      "step": 1606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.177359580993652,
+      "learning_rate": 2.1641816920943135e-05,
+      "loss": 1.2133,
+      "step": 1607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.765803337097168,
+      "learning_rate": 2.1636615811373094e-05,
+      "loss": 1.1475,
+      "step": 1608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.836848258972168,
+      "learning_rate": 2.163141470180305e-05,
+      "loss": 0.9745,
+      "step": 1609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.660187721252441,
+      "learning_rate": 2.162621359223301e-05,
+      "loss": 0.978,
+      "step": 1610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.367073059082031,
+      "learning_rate": 2.1621012482662966e-05,
+      "loss": 0.9795,
+      "step": 1611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.526913166046143,
+      "learning_rate": 2.1615811373092928e-05,
+      "loss": 0.8306,
+      "step": 1612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.85236930847168,
+      "learning_rate": 2.1610610263522883e-05,
+      "loss": 1.1021,
+      "step": 1613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.483741760253906,
+      "learning_rate": 2.1605409153952845e-05,
+      "loss": 0.8563,
+      "step": 1614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.153908729553223,
+      "learning_rate": 2.16002080443828e-05,
+      "loss": 0.94,
+      "step": 1615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.052560806274414,
+      "learning_rate": 2.159500693481276e-05,
+      "loss": 1.0266,
+      "step": 1616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.532426834106445,
+      "learning_rate": 2.1589805825242718e-05,
+      "loss": 0.9681,
+      "step": 1617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.622851371765137,
+      "learning_rate": 2.1584604715672677e-05,
+      "loss": 1.1497,
+      "step": 1618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.737046241760254,
+      "learning_rate": 2.1579403606102635e-05,
+      "loss": 0.9209,
+      "step": 1619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.788216590881348,
+      "learning_rate": 2.1574202496532594e-05,
+      "loss": 0.8631,
+      "step": 1620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.907063484191895,
+      "learning_rate": 2.1569001386962552e-05,
+      "loss": 1.266,
+      "step": 1621
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.501041412353516,
+      "learning_rate": 2.156380027739251e-05,
+      "loss": 0.9295,
+      "step": 1622
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.46712589263916,
+      "learning_rate": 2.155859916782247e-05,
+      "loss": 0.9502,
+      "step": 1623
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.265917778015137,
+      "learning_rate": 2.1553398058252428e-05,
+      "loss": 1.0674,
+      "step": 1624
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.438003540039062,
+      "learning_rate": 2.1548196948682383e-05,
+      "loss": 1.1105,
+      "step": 1625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.949743270874023,
+      "learning_rate": 2.1542995839112345e-05,
+      "loss": 1.1875,
+      "step": 1626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.653990745544434,
+      "learning_rate": 2.15377947295423e-05,
+      "loss": 1.0083,
+      "step": 1627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.823734283447266,
+      "learning_rate": 2.1532593619972263e-05,
+      "loss": 0.9727,
+      "step": 1628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.251046180725098,
+      "learning_rate": 2.1527392510402218e-05,
+      "loss": 1.1671,
+      "step": 1629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.571043014526367,
+      "learning_rate": 2.152219140083218e-05,
+      "loss": 1.0371,
+      "step": 1630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.753584861755371,
+      "learning_rate": 2.1516990291262135e-05,
+      "loss": 0.9824,
+      "step": 1631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.445932388305664,
+      "learning_rate": 2.1511789181692094e-05,
+      "loss": 0.8462,
+      "step": 1632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.156420707702637,
+      "learning_rate": 2.1506588072122052e-05,
+      "loss": 0.8387,
+      "step": 1633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.6746244430542,
+      "learning_rate": 2.150138696255201e-05,
+      "loss": 0.9108,
+      "step": 1634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.927655220031738,
+      "learning_rate": 2.149618585298197e-05,
+      "loss": 1.2826,
+      "step": 1635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.081718444824219,
+      "learning_rate": 2.1490984743411928e-05,
+      "loss": 0.9389,
+      "step": 1636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.496515274047852,
+      "learning_rate": 2.1485783633841887e-05,
+      "loss": 1.1783,
+      "step": 1637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.20954418182373,
+      "learning_rate": 2.1480582524271845e-05,
+      "loss": 1.0799,
+      "step": 1638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.32349967956543,
+      "learning_rate": 2.1475381414701804e-05,
+      "loss": 0.9551,
+      "step": 1639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.724217414855957,
+      "learning_rate": 2.1470180305131763e-05,
+      "loss": 1.0046,
+      "step": 1640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.414569854736328,
+      "learning_rate": 2.1464979195561718e-05,
+      "loss": 1.022,
+      "step": 1641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.6725945472717285,
+      "learning_rate": 2.145977808599168e-05,
+      "loss": 0.9224,
+      "step": 1642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.927412986755371,
+      "learning_rate": 2.1454576976421635e-05,
+      "loss": 1.2199,
+      "step": 1643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.550128936767578,
+      "learning_rate": 2.1449375866851597e-05,
+      "loss": 0.847,
+      "step": 1644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.247906684875488,
+      "learning_rate": 2.1444174757281552e-05,
+      "loss": 0.9572,
+      "step": 1645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.220638275146484,
+      "learning_rate": 2.1438973647711514e-05,
+      "loss": 1.0399,
+      "step": 1646
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.04456615447998,
+      "learning_rate": 2.143377253814147e-05,
+      "loss": 0.7589,
+      "step": 1647
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.345590591430664,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 1.0668,
+      "step": 1648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.260354995727539,
+      "learning_rate": 2.1423370319001387e-05,
+      "loss": 1.0723,
+      "step": 1649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.36581802368164,
+      "learning_rate": 2.1418169209431345e-05,
+      "loss": 0.7731,
+      "step": 1650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.418099403381348,
+      "learning_rate": 2.1412968099861304e-05,
+      "loss": 1.1206,
+      "step": 1651
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.175243377685547,
+      "learning_rate": 2.1407766990291263e-05,
+      "loss": 1.1165,
+      "step": 1652
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.753148078918457,
+      "learning_rate": 2.140256588072122e-05,
+      "loss": 1.2074,
+      "step": 1653
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.045913696289062,
+      "learning_rate": 2.139736477115118e-05,
+      "loss": 0.8931,
+      "step": 1654
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.629289627075195,
+      "learning_rate": 2.1392163661581135e-05,
+      "loss": 1.1302,
+      "step": 1655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.870026588439941,
+      "learning_rate": 2.1386962552011097e-05,
+      "loss": 0.9737,
+      "step": 1656
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.948883056640625,
+      "learning_rate": 2.1381761442441052e-05,
+      "loss": 0.7934,
+      "step": 1657
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.169763565063477,
+      "learning_rate": 2.1376560332871014e-05,
+      "loss": 0.8089,
+      "step": 1658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.743636131286621,
+      "learning_rate": 2.137135922330097e-05,
+      "loss": 1.1734,
+      "step": 1659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.306285858154297,
+      "learning_rate": 2.136615811373093e-05,
+      "loss": 0.8468,
+      "step": 1660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.917943000793457,
+      "learning_rate": 2.1360957004160887e-05,
+      "loss": 0.9989,
+      "step": 1661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.121708869934082,
+      "learning_rate": 2.135575589459085e-05,
+      "loss": 1.3721,
+      "step": 1662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.0794677734375,
+      "learning_rate": 2.1350554785020804e-05,
+      "loss": 0.9388,
+      "step": 1663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.54685115814209,
+      "learning_rate": 2.1345353675450763e-05,
+      "loss": 1.0956,
+      "step": 1664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.469412326812744,
+      "learning_rate": 2.134015256588072e-05,
+      "loss": 0.7056,
+      "step": 1665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.555352687835693,
+      "learning_rate": 2.133495145631068e-05,
+      "loss": 0.7533,
+      "step": 1666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.837309837341309,
+      "learning_rate": 2.132975034674064e-05,
+      "loss": 0.8574,
+      "step": 1667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.03205394744873,
+      "learning_rate": 2.1324549237170597e-05,
+      "loss": 1.3477,
+      "step": 1668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.98725700378418,
+      "learning_rate": 2.1319348127600556e-05,
+      "loss": 0.8264,
+      "step": 1669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.989886283874512,
+      "learning_rate": 2.1314147018030514e-05,
+      "loss": 0.9308,
+      "step": 1670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.628488063812256,
+      "learning_rate": 2.130894590846047e-05,
+      "loss": 0.9402,
+      "step": 1671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.177940368652344,
+      "learning_rate": 2.130374479889043e-05,
+      "loss": 0.9051,
+      "step": 1672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.530177116394043,
+      "learning_rate": 2.1298543689320387e-05,
+      "loss": 0.9192,
+      "step": 1673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.620121002197266,
+      "learning_rate": 2.129334257975035e-05,
+      "loss": 0.9727,
+      "step": 1674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.336570739746094,
+      "learning_rate": 2.1288141470180304e-05,
+      "loss": 1.0059,
+      "step": 1675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.81037712097168,
+      "learning_rate": 2.1282940360610266e-05,
+      "loss": 0.9668,
+      "step": 1676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.010392189025879,
+      "learning_rate": 2.127773925104022e-05,
+      "loss": 0.7592,
+      "step": 1677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.468883514404297,
+      "learning_rate": 2.127253814147018e-05,
+      "loss": 1.0394,
+      "step": 1678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.315005302429199,
+      "learning_rate": 2.126733703190014e-05,
+      "loss": 0.7472,
+      "step": 1679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.25241470336914,
+      "learning_rate": 2.1262135922330097e-05,
+      "loss": 0.9165,
+      "step": 1680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.910595893859863,
+      "learning_rate": 2.1256934812760056e-05,
+      "loss": 1.3089,
+      "step": 1681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.70949935913086,
+      "learning_rate": 2.1251733703190014e-05,
+      "loss": 1.0669,
+      "step": 1682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.826355934143066,
+      "learning_rate": 2.1246532593619973e-05,
+      "loss": 0.8172,
+      "step": 1683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.849308967590332,
+      "learning_rate": 2.124133148404993e-05,
+      "loss": 0.8236,
+      "step": 1684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.871307373046875,
+      "learning_rate": 2.123613037447989e-05,
+      "loss": 1.1733,
+      "step": 1685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.873411178588867,
+      "learning_rate": 2.123092926490985e-05,
+      "loss": 1.024,
+      "step": 1686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.857860565185547,
+      "learning_rate": 2.1225728155339804e-05,
+      "loss": 0.8685,
+      "step": 1687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.926965713500977,
+      "learning_rate": 2.1220527045769766e-05,
+      "loss": 1.3536,
+      "step": 1688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.526103973388672,
+      "learning_rate": 2.121532593619972e-05,
+      "loss": 1.0057,
+      "step": 1689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.883389472961426,
+      "learning_rate": 2.1210124826629683e-05,
+      "loss": 0.937,
+      "step": 1690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.637239456176758,
+      "learning_rate": 2.120492371705964e-05,
+      "loss": 1.1863,
+      "step": 1691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.12449836730957,
+      "learning_rate": 2.11997226074896e-05,
+      "loss": 1.2121,
+      "step": 1692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.009220123291016,
+      "learning_rate": 2.1194521497919556e-05,
+      "loss": 0.9908,
+      "step": 1693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.494811058044434,
+      "learning_rate": 2.1189320388349514e-05,
+      "loss": 1.0037,
+      "step": 1694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.166579246520996,
+      "learning_rate": 2.1184119278779473e-05,
+      "loss": 0.9186,
+      "step": 1695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.355118751525879,
+      "learning_rate": 2.117891816920943e-05,
+      "loss": 1.1119,
+      "step": 1696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.580537796020508,
+      "learning_rate": 2.117371705963939e-05,
+      "loss": 1.0441,
+      "step": 1697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.429487228393555,
+      "learning_rate": 2.116851595006935e-05,
+      "loss": 0.9334,
+      "step": 1698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.64000415802002,
+      "learning_rate": 2.1163314840499307e-05,
+      "loss": 1.2891,
+      "step": 1699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.230514526367188,
+      "learning_rate": 2.1158113730929266e-05,
+      "loss": 0.8613,
+      "step": 1700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.051413536071777,
+      "learning_rate": 2.1152912621359224e-05,
+      "loss": 1.1433,
+      "step": 1701
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.793166637420654,
+      "learning_rate": 2.1147711511789183e-05,
+      "loss": 0.8766,
+      "step": 1702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.1639404296875,
+      "learning_rate": 2.1142510402219138e-05,
+      "loss": 0.8856,
+      "step": 1703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.963254928588867,
+      "learning_rate": 2.11373092926491e-05,
+      "loss": 0.6662,
+      "step": 1704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.71238899230957,
+      "learning_rate": 2.1132108183079056e-05,
+      "loss": 0.8402,
+      "step": 1705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.129924297332764,
+      "learning_rate": 2.1126907073509018e-05,
+      "loss": 0.9903,
+      "step": 1706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.466437339782715,
+      "learning_rate": 2.1121705963938973e-05,
+      "loss": 0.9505,
+      "step": 1707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.19483757019043,
+      "learning_rate": 2.1116504854368935e-05,
+      "loss": 1.1001,
+      "step": 1708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.989506721496582,
+      "learning_rate": 2.111130374479889e-05,
+      "loss": 1.0578,
+      "step": 1709
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.077580451965332,
+      "learning_rate": 2.110610263522885e-05,
+      "loss": 0.9551,
+      "step": 1710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.656124114990234,
+      "learning_rate": 2.1100901525658807e-05,
+      "loss": 1.292,
+      "step": 1711
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.727130889892578,
+      "learning_rate": 2.1095700416088766e-05,
+      "loss": 1.1123,
+      "step": 1712
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.74486780166626,
+      "learning_rate": 2.1090499306518724e-05,
+      "loss": 0.8495,
+      "step": 1713
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.348278045654297,
+      "learning_rate": 2.1085298196948683e-05,
+      "loss": 0.8393,
+      "step": 1714
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.923415184020996,
+      "learning_rate": 2.108009708737864e-05,
+      "loss": 1.0608,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.132942199707031,
+      "learning_rate": 2.10748959778086e-05,
+      "loss": 1.1176,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.8517842292785645,
+      "learning_rate": 2.1069694868238555e-05,
+      "loss": 0.9511,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.295762062072754,
+      "learning_rate": 2.1064493758668517e-05,
+      "loss": 1.0137,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.619576454162598,
+      "learning_rate": 2.1059292649098473e-05,
+      "loss": 1.0352,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.282169342041016,
+      "learning_rate": 2.1054091539528435e-05,
+      "loss": 0.9911,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.87074089050293,
+      "learning_rate": 2.104889042995839e-05,
+      "loss": 0.8844,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.744874954223633,
+      "learning_rate": 2.1043689320388352e-05,
+      "loss": 0.94,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.275975704193115,
+      "learning_rate": 2.1038488210818307e-05,
+      "loss": 0.7407,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.153521537780762,
+      "learning_rate": 2.103328710124827e-05,
+      "loss": 1.1153,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.648085594177246,
+      "learning_rate": 2.1028085991678224e-05,
+      "loss": 1.1463,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.292182922363281,
+      "learning_rate": 2.1022884882108183e-05,
+      "loss": 0.9604,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.345081329345703,
+      "learning_rate": 2.101768377253814e-05,
+      "loss": 1.2416,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.737083435058594,
+      "learning_rate": 2.10124826629681e-05,
+      "loss": 0.7615,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.526652336120605,
+      "learning_rate": 2.100728155339806e-05,
+      "loss": 1.3182,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.903183937072754,
+      "learning_rate": 2.1002080443828017e-05,
+      "loss": 1.0851,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.215039253234863,
+      "learning_rate": 2.0996879334257976e-05,
+      "loss": 0.8879,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.119863510131836,
+      "learning_rate": 2.0991678224687935e-05,
+      "loss": 1.0287,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.009673118591309,
+      "learning_rate": 2.098647711511789e-05,
+      "loss": 0.9087,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.262844085693359,
+      "learning_rate": 2.0981276005547852e-05,
+      "loss": 0.9149,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.75808048248291,
+      "learning_rate": 2.0976074895977807e-05,
+      "loss": 1.0588,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.085824012756348,
+      "learning_rate": 2.097087378640777e-05,
+      "loss": 0.6754,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.924566268920898,
+      "learning_rate": 2.0965672676837724e-05,
+      "loss": 1.0383,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.254755973815918,
+      "learning_rate": 2.0960471567267686e-05,
+      "loss": 0.805,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.93554401397705,
+      "learning_rate": 2.095527045769764e-05,
+      "loss": 0.9316,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.271368026733398,
+      "learning_rate": 2.09500693481276e-05,
+      "loss": 1.1695,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.455760955810547,
+      "learning_rate": 2.094486823855756e-05,
+      "loss": 0.7929,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.914581298828125,
+      "learning_rate": 2.0939667128987517e-05,
+      "loss": 1.0004,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.92013931274414,
+      "learning_rate": 2.0934466019417476e-05,
+      "loss": 1.186,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.798009872436523,
+      "learning_rate": 2.0929264909847435e-05,
+      "loss": 1.0501,
+      "step": 1744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.172391891479492,
+      "learning_rate": 2.0924063800277393e-05,
+      "loss": 0.7513,
+      "step": 1745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.1939849853515625,
+      "learning_rate": 2.0918862690707352e-05,
+      "loss": 0.6659,
+      "step": 1746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.16413402557373,
+      "learning_rate": 2.091366158113731e-05,
+      "loss": 1.0073,
+      "step": 1747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.985913276672363,
+      "learning_rate": 2.090846047156727e-05,
+      "loss": 0.8851,
+      "step": 1748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.731945037841797,
+      "learning_rate": 2.0903259361997224e-05,
+      "loss": 0.9554,
+      "step": 1749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.712846755981445,
+      "learning_rate": 2.0898058252427186e-05,
+      "loss": 1.2106,
+      "step": 1750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.07654857635498,
+      "learning_rate": 2.089285714285714e-05,
+      "loss": 0.8189,
+      "step": 1751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.146484851837158,
+      "learning_rate": 2.0887656033287104e-05,
+      "loss": 0.8135,
+      "step": 1752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.261504173278809,
+      "learning_rate": 2.088245492371706e-05,
+      "loss": 1.02,
+      "step": 1753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.727327346801758,
+      "learning_rate": 2.087725381414702e-05,
+      "loss": 0.9135,
+      "step": 1754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.599987983703613,
+      "learning_rate": 2.0872052704576976e-05,
+      "loss": 1.0359,
+      "step": 1755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.74177885055542,
+      "learning_rate": 2.0866851595006935e-05,
+      "loss": 0.9251,
+      "step": 1756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.007347106933594,
+      "learning_rate": 2.0861650485436893e-05,
+      "loss": 1.0688,
+      "step": 1757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.739408493041992,
+      "learning_rate": 2.0856449375866852e-05,
+      "loss": 1.0849,
+      "step": 1758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.04910945892334,
+      "learning_rate": 2.085124826629681e-05,
+      "loss": 0.8398,
+      "step": 1759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.735797882080078,
+      "learning_rate": 2.084604715672677e-05,
+      "loss": 0.8337,
+      "step": 1760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.785184383392334,
+      "learning_rate": 2.0840846047156728e-05,
+      "loss": 0.8296,
+      "step": 1761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.373104095458984,
+      "learning_rate": 2.0835644937586686e-05,
+      "loss": 1.0406,
+      "step": 1762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.556767463684082,
+      "learning_rate": 2.083044382801664e-05,
+      "loss": 0.8216,
+      "step": 1763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.50161361694336,
+      "learning_rate": 2.0825242718446603e-05,
+      "loss": 1.3081,
+      "step": 1764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.276495456695557,
+      "learning_rate": 2.082004160887656e-05,
+      "loss": 0.9097,
+      "step": 1765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.409287452697754,
+      "learning_rate": 2.081484049930652e-05,
+      "loss": 0.7794,
+      "step": 1766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.35347843170166,
+      "learning_rate": 2.0809639389736476e-05,
+      "loss": 0.8513,
+      "step": 1767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.337664604187012,
+      "learning_rate": 2.0804438280166438e-05,
+      "loss": 0.9523,
+      "step": 1768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.931621551513672,
+      "learning_rate": 2.0799237170596393e-05,
+      "loss": 0.693,
+      "step": 1769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.065896987915039,
+      "learning_rate": 2.0794036061026355e-05,
+      "loss": 0.859,
+      "step": 1770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.407849311828613,
+      "learning_rate": 2.078883495145631e-05,
+      "loss": 0.8172,
+      "step": 1771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.347042083740234,
+      "learning_rate": 2.078363384188627e-05,
+      "loss": 0.8553,
+      "step": 1772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.984518051147461,
+      "learning_rate": 2.0778432732316228e-05,
+      "loss": 0.8052,
+      "step": 1773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.469547271728516,
+      "learning_rate": 2.0773231622746186e-05,
+      "loss": 0.8248,
+      "step": 1774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.689716339111328,
+      "learning_rate": 2.0768030513176145e-05,
+      "loss": 0.9785,
+      "step": 1775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.626379013061523,
+      "learning_rate": 2.0762829403606103e-05,
+      "loss": 1.0119,
+      "step": 1776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.568920135498047,
+      "learning_rate": 2.0757628294036062e-05,
+      "loss": 1.114,
+      "step": 1777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.921881675720215,
+      "learning_rate": 2.075242718446602e-05,
+      "loss": 1.1523,
+      "step": 1778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.626132011413574,
+      "learning_rate": 2.0747226074895976e-05,
+      "loss": 1.0149,
+      "step": 1779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.758912086486816,
+      "learning_rate": 2.0742024965325938e-05,
+      "loss": 1.2241,
+      "step": 1780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.550247192382812,
+      "learning_rate": 2.0736823855755893e-05,
+      "loss": 1.1389,
+      "step": 1781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.472107887268066,
+      "learning_rate": 2.0731622746185855e-05,
+      "loss": 0.6814,
+      "step": 1782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.630594253540039,
+      "learning_rate": 2.072642163661581e-05,
+      "loss": 1.3379,
+      "step": 1783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.253286838531494,
+      "learning_rate": 2.0721220527045772e-05,
+      "loss": 0.6316,
+      "step": 1784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.315498352050781,
+      "learning_rate": 2.0716019417475728e-05,
+      "loss": 0.783,
+      "step": 1785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.76617431640625,
+      "learning_rate": 2.071081830790569e-05,
+      "loss": 0.8965,
+      "step": 1786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.727604866027832,
+      "learning_rate": 2.0705617198335645e-05,
+      "loss": 1.0326,
+      "step": 1787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.951493263244629,
+      "learning_rate": 2.0700416088765603e-05,
+      "loss": 1.1093,
+      "step": 1788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.183108329772949,
+      "learning_rate": 2.0695214979195562e-05,
+      "loss": 0.6822,
+      "step": 1789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.15321683883667,
+      "learning_rate": 2.069001386962552e-05,
+      "loss": 0.683,
+      "step": 1790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.896440505981445,
+      "learning_rate": 2.068481276005548e-05,
+      "loss": 1.0491,
+      "step": 1791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.498035430908203,
+      "learning_rate": 2.0679611650485438e-05,
+      "loss": 0.8614,
+      "step": 1792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.748805522918701,
+      "learning_rate": 2.0674410540915396e-05,
+      "loss": 0.7889,
+      "step": 1793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.501326560974121,
+      "learning_rate": 2.0669209431345355e-05,
+      "loss": 1.0563,
+      "step": 1794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.21589469909668,
+      "learning_rate": 2.066400832177531e-05,
+      "loss": 0.8673,
+      "step": 1795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.392748832702637,
+      "learning_rate": 2.0658807212205272e-05,
+      "loss": 0.8682,
+      "step": 1796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.426064491271973,
+      "learning_rate": 2.0653606102635228e-05,
+      "loss": 0.8775,
+      "step": 1797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.027569770812988,
+      "learning_rate": 2.064840499306519e-05,
+      "loss": 0.8761,
+      "step": 1798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.82761287689209,
+      "learning_rate": 2.0643203883495145e-05,
+      "loss": 0.9749,
+      "step": 1799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.067137718200684,
+      "learning_rate": 2.0638002773925107e-05,
+      "loss": 1.2451,
+      "step": 1800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.760767459869385,
+      "learning_rate": 2.0632801664355062e-05,
+      "loss": 0.7466,
+      "step": 1801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.869583129882812,
+      "learning_rate": 2.062760055478502e-05,
+      "loss": 1.1424,
+      "step": 1802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.646446228027344,
+      "learning_rate": 2.062239944521498e-05,
+      "loss": 0.979,
+      "step": 1803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.12800121307373,
+      "learning_rate": 2.0617198335644938e-05,
+      "loss": 1.0694,
+      "step": 1804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.228433609008789,
+      "learning_rate": 2.0611997226074896e-05,
+      "loss": 0.8537,
+      "step": 1805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.784979820251465,
+      "learning_rate": 2.0606796116504855e-05,
+      "loss": 1.0533,
+      "step": 1806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.853028297424316,
+      "learning_rate": 2.0601595006934814e-05,
+      "loss": 0.8897,
+      "step": 1807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.688620090484619,
+      "learning_rate": 2.0596393897364772e-05,
+      "loss": 0.839,
+      "step": 1808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.124772071838379,
+      "learning_rate": 2.059119278779473e-05,
+      "loss": 1.0203,
+      "step": 1809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.272348403930664,
+      "learning_rate": 2.058599167822469e-05,
+      "loss": 0.9748,
+      "step": 1810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.3377103805542,
+      "learning_rate": 2.0580790568654645e-05,
+      "loss": 1.029,
+      "step": 1811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.121694564819336,
+      "learning_rate": 2.0575589459084607e-05,
+      "loss": 0.9566,
+      "step": 1812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.211719512939453,
+      "learning_rate": 2.0570388349514562e-05,
+      "loss": 1.1143,
+      "step": 1813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.9921088218688965,
+      "learning_rate": 2.0565187239944524e-05,
+      "loss": 1.0478,
+      "step": 1814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.21432113647461,
+      "learning_rate": 2.055998613037448e-05,
+      "loss": 1.1374,
+      "step": 1815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.128217697143555,
+      "learning_rate": 2.055478502080444e-05,
+      "loss": 1.2524,
+      "step": 1816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.68552017211914,
+      "learning_rate": 2.0549583911234396e-05,
+      "loss": 0.9567,
+      "step": 1817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.655951976776123,
+      "learning_rate": 2.0544382801664355e-05,
+      "loss": 0.9436,
+      "step": 1818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.031831741333008,
+      "learning_rate": 2.0539181692094314e-05,
+      "loss": 1.0126,
+      "step": 1819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.143815040588379,
+      "learning_rate": 2.0533980582524272e-05,
+      "loss": 0.952,
+      "step": 1820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.34178352355957,
+      "learning_rate": 2.052877947295423e-05,
+      "loss": 1.2109,
+      "step": 1821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.36521053314209,
+      "learning_rate": 2.052357836338419e-05,
+      "loss": 0.7731,
+      "step": 1822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.812886238098145,
+      "learning_rate": 2.0518377253814148e-05,
+      "loss": 1.135,
+      "step": 1823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.14427661895752,
+      "learning_rate": 2.0513176144244107e-05,
+      "loss": 0.9146,
+      "step": 1824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.727337837219238,
+      "learning_rate": 2.0507975034674062e-05,
+      "loss": 1.1631,
+      "step": 1825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.827186584472656,
+      "learning_rate": 2.0502773925104024e-05,
+      "loss": 0.9367,
+      "step": 1826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.758828163146973,
+      "learning_rate": 2.049757281553398e-05,
+      "loss": 1.1972,
+      "step": 1827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.972197532653809,
+      "learning_rate": 2.049237170596394e-05,
+      "loss": 0.9267,
+      "step": 1828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.925103187561035,
+      "learning_rate": 2.0487170596393896e-05,
+      "loss": 1.0461,
+      "step": 1829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.668866157531738,
+      "learning_rate": 2.048196948682386e-05,
+      "loss": 0.7448,
+      "step": 1830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.152073860168457,
+      "learning_rate": 2.0476768377253814e-05,
+      "loss": 0.9772,
+      "step": 1831
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.280303478240967,
+      "learning_rate": 2.0471567267683776e-05,
+      "loss": 0.6909,
+      "step": 1832
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.157217025756836,
+      "learning_rate": 2.046636615811373e-05,
+      "loss": 1.1721,
+      "step": 1833
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.512612819671631,
+      "learning_rate": 2.046116504854369e-05,
+      "loss": 0.8809,
+      "step": 1834
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.897099494934082,
+      "learning_rate": 2.0455963938973648e-05,
+      "loss": 0.8384,
+      "step": 1835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.692654609680176,
+      "learning_rate": 2.0450762829403607e-05,
+      "loss": 0.8681,
+      "step": 1836
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.280620574951172,
+      "learning_rate": 2.0445561719833565e-05,
+      "loss": 1.0134,
+      "step": 1837
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.577531814575195,
+      "learning_rate": 2.0440360610263524e-05,
+      "loss": 0.7492,
+      "step": 1838
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.943808555603027,
+      "learning_rate": 2.0435159500693482e-05,
+      "loss": 0.8839,
+      "step": 1839
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.143321990966797,
+      "learning_rate": 2.042995839112344e-05,
+      "loss": 0.8749,
+      "step": 1840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.880051612854004,
+      "learning_rate": 2.0424757281553396e-05,
+      "loss": 1.1705,
+      "step": 1841
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.452839851379395,
+      "learning_rate": 2.0419556171983358e-05,
+      "loss": 0.9903,
+      "step": 1842
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.597046375274658,
+      "learning_rate": 2.0414355062413314e-05,
+      "loss": 0.7932,
+      "step": 1843
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.38146686553955,
+      "learning_rate": 2.0409153952843276e-05,
+      "loss": 1.0898,
+      "step": 1844
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.5260648727417,
+      "learning_rate": 2.040395284327323e-05,
+      "loss": 0.833,
+      "step": 1845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.469724178314209,
+      "learning_rate": 2.0398751733703193e-05,
+      "loss": 0.6147,
+      "step": 1846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.164047241210938,
+      "learning_rate": 2.0393550624133148e-05,
+      "loss": 0.8874,
+      "step": 1847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.305908203125,
+      "learning_rate": 2.0388349514563107e-05,
+      "loss": 1.1425,
+      "step": 1848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.798207759857178,
+      "learning_rate": 2.0383148404993065e-05,
+      "loss": 0.9972,
+      "step": 1849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.617688179016113,
+      "learning_rate": 2.0377947295423024e-05,
+      "loss": 0.7912,
+      "step": 1850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.213847160339355,
+      "learning_rate": 2.0372746185852982e-05,
+      "loss": 0.7733,
+      "step": 1851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.901324272155762,
+      "learning_rate": 2.036754507628294e-05,
+      "loss": 1.0761,
+      "step": 1852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.086406707763672,
+      "learning_rate": 2.03623439667129e-05,
+      "loss": 1.2476,
+      "step": 1853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.682315826416016,
+      "learning_rate": 2.0357142857142858e-05,
+      "loss": 1.0641,
+      "step": 1854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.457551002502441,
+      "learning_rate": 2.0351941747572817e-05,
+      "loss": 0.9083,
+      "step": 1855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.520429611206055,
+      "learning_rate": 2.0346740638002775e-05,
+      "loss": 1.0106,
+      "step": 1856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.64767074584961,
+      "learning_rate": 2.034153952843273e-05,
+      "loss": 0.9989,
+      "step": 1857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.595660209655762,
+      "learning_rate": 2.0336338418862693e-05,
+      "loss": 1.1712,
+      "step": 1858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.12858772277832,
+      "learning_rate": 2.0331137309292648e-05,
+      "loss": 0.9241,
+      "step": 1859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.823761940002441,
+      "learning_rate": 2.032593619972261e-05,
+      "loss": 0.903,
+      "step": 1860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.532835006713867,
+      "learning_rate": 2.0320735090152565e-05,
+      "loss": 0.9554,
+      "step": 1861
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.099420547485352,
+      "learning_rate": 2.0315533980582527e-05,
+      "loss": 0.8788,
+      "step": 1862
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.075092315673828,
+      "learning_rate": 2.0310332871012482e-05,
+      "loss": 1.0726,
+      "step": 1863
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.841543197631836,
+      "learning_rate": 2.030513176144244e-05,
+      "loss": 1.1965,
+      "step": 1864
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.167098045349121,
+      "learning_rate": 2.02999306518724e-05,
+      "loss": 1.2175,
+      "step": 1865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.781707763671875,
+      "learning_rate": 2.0294729542302358e-05,
+      "loss": 1.0769,
+      "step": 1866
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.558561325073242,
+      "learning_rate": 2.0289528432732317e-05,
+      "loss": 0.8989,
+      "step": 1867
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.53407096862793,
+      "learning_rate": 2.0284327323162275e-05,
+      "loss": 0.9089,
+      "step": 1868
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.919599533081055,
+      "learning_rate": 2.0279126213592234e-05,
+      "loss": 1.077,
+      "step": 1869
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.832406997680664,
+      "learning_rate": 2.0273925104022193e-05,
+      "loss": 1.0679,
+      "step": 1870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.640225887298584,
+      "learning_rate": 2.026872399445215e-05,
+      "loss": 0.9744,
+      "step": 1871
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.073670387268066,
+      "learning_rate": 2.026352288488211e-05,
+      "loss": 0.9386,
+      "step": 1872
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.000179290771484,
+      "learning_rate": 2.0258321775312065e-05,
+      "loss": 0.5979,
+      "step": 1873
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.370835304260254,
+      "learning_rate": 2.0253120665742027e-05,
+      "loss": 1.0725,
+      "step": 1874
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.661410331726074,
+      "learning_rate": 2.0247919556171982e-05,
+      "loss": 1.1436,
+      "step": 1875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.855201244354248,
+      "learning_rate": 2.0242718446601944e-05,
+      "loss": 0.8585,
+      "step": 1876
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.182552337646484,
+      "learning_rate": 2.02375173370319e-05,
+      "loss": 0.9837,
+      "step": 1877
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.174983024597168,
+      "learning_rate": 2.023231622746186e-05,
+      "loss": 0.9328,
+      "step": 1878
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.030082702636719,
+      "learning_rate": 2.0227115117891817e-05,
+      "loss": 1.0196,
+      "step": 1879
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.385581970214844,
+      "learning_rate": 2.0221914008321775e-05,
+      "loss": 1.1031,
+      "step": 1880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.314638137817383,
+      "learning_rate": 2.0216712898751734e-05,
+      "loss": 0.9081,
+      "step": 1881
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.271907806396484,
+      "learning_rate": 2.0211511789181693e-05,
+      "loss": 0.8553,
+      "step": 1882
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.60692024230957,
+      "learning_rate": 2.020631067961165e-05,
+      "loss": 0.9001,
+      "step": 1883
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.103350639343262,
+      "learning_rate": 2.020110957004161e-05,
+      "loss": 1.0298,
+      "step": 1884
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.624825477600098,
+      "learning_rate": 2.019590846047157e-05,
+      "loss": 1.1545,
+      "step": 1885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.381512641906738,
+      "learning_rate": 2.0190707350901527e-05,
+      "loss": 1.1658,
+      "step": 1886
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.114741325378418,
+      "learning_rate": 2.0185506241331482e-05,
+      "loss": 1.0475,
+      "step": 1887
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.848599433898926,
+      "learning_rate": 2.0180305131761444e-05,
+      "loss": 1.1993,
+      "step": 1888
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.5590181350708,
+      "learning_rate": 2.01751040221914e-05,
+      "loss": 1.3536,
+      "step": 1889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.70373249053955,
+      "learning_rate": 2.016990291262136e-05,
+      "loss": 0.9376,
+      "step": 1890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.016758441925049,
+      "learning_rate": 2.0164701803051317e-05,
+      "loss": 0.6931,
+      "step": 1891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.74405574798584,
+      "learning_rate": 2.015950069348128e-05,
+      "loss": 0.9366,
+      "step": 1892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.600968360900879,
+      "learning_rate": 2.0154299583911234e-05,
+      "loss": 0.9526,
+      "step": 1893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.830979824066162,
+      "learning_rate": 2.0149098474341196e-05,
+      "loss": 0.909,
+      "step": 1894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.545015335083008,
+      "learning_rate": 2.014389736477115e-05,
+      "loss": 1.13,
+      "step": 1895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.765753746032715,
+      "learning_rate": 2.013869625520111e-05,
+      "loss": 1.0516,
+      "step": 1896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.52077579498291,
+      "learning_rate": 2.013349514563107e-05,
+      "loss": 0.8004,
+      "step": 1897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.8366007804870605,
+      "learning_rate": 2.0128294036061027e-05,
+      "loss": 0.8787,
+      "step": 1898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.225493431091309,
+      "learning_rate": 2.0123092926490986e-05,
+      "loss": 0.7548,
+      "step": 1899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.855305194854736,
+      "learning_rate": 2.0117891816920944e-05,
+      "loss": 0.8534,
+      "step": 1900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.846075534820557,
+      "learning_rate": 2.0112690707350903e-05,
+      "loss": 0.7653,
+      "step": 1901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.08962631225586,
+      "learning_rate": 2.010748959778086e-05,
+      "loss": 1.1172,
+      "step": 1902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.12498664855957,
+      "learning_rate": 2.0102288488210817e-05,
+      "loss": 1.0684,
+      "step": 1903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.908328533172607,
+      "learning_rate": 2.009708737864078e-05,
+      "loss": 0.8131,
+      "step": 1904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.699455261230469,
+      "learning_rate": 2.0091886269070734e-05,
+      "loss": 0.8202,
+      "step": 1905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.407492637634277,
+      "learning_rate": 2.0086685159500696e-05,
+      "loss": 1.368,
+      "step": 1906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.775314331054688,
+      "learning_rate": 2.008148404993065e-05,
+      "loss": 0.8989,
+      "step": 1907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.947121620178223,
+      "learning_rate": 2.0076282940360613e-05,
+      "loss": 0.6972,
+      "step": 1908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.135316848754883,
+      "learning_rate": 2.007108183079057e-05,
+      "loss": 0.992,
+      "step": 1909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.010300636291504,
+      "learning_rate": 2.0065880721220527e-05,
+      "loss": 1.016,
+      "step": 1910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.803226470947266,
+      "learning_rate": 2.0060679611650486e-05,
+      "loss": 0.8972,
+      "step": 1911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.930852890014648,
+      "learning_rate": 2.0055478502080444e-05,
+      "loss": 0.7577,
+      "step": 1912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.240434646606445,
+      "learning_rate": 2.0050277392510403e-05,
+      "loss": 0.7245,
+      "step": 1913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.839241981506348,
+      "learning_rate": 2.004507628294036e-05,
+      "loss": 1.0298,
+      "step": 1914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.951171875,
+      "learning_rate": 2.003987517337032e-05,
+      "loss": 0.8549,
+      "step": 1915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.689400672912598,
+      "learning_rate": 2.003467406380028e-05,
+      "loss": 1.2937,
+      "step": 1916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.621824264526367,
+      "learning_rate": 2.0029472954230237e-05,
+      "loss": 1.2308,
+      "step": 1917
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.283914566040039,
+      "learning_rate": 2.0024271844660196e-05,
+      "loss": 0.8211,
+      "step": 1918
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.065135955810547,
+      "learning_rate": 2.001907073509015e-05,
+      "loss": 0.8206,
+      "step": 1919
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.092941284179688,
+      "learning_rate": 2.0013869625520113e-05,
+      "loss": 0.8047,
+      "step": 1920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.2218656539917,
+      "learning_rate": 2.000866851595007e-05,
+      "loss": 1.0216,
+      "step": 1921
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.780349731445312,
+      "learning_rate": 2.000346740638003e-05,
+      "loss": 0.893,
+      "step": 1922
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.80836296081543,
+      "learning_rate": 1.9998266296809986e-05,
+      "loss": 1.1336,
+      "step": 1923
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.763306617736816,
+      "learning_rate": 1.9993065187239948e-05,
+      "loss": 0.9787,
+      "step": 1924
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.962296485900879,
+      "learning_rate": 1.9987864077669903e-05,
+      "loss": 0.9264,
+      "step": 1925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.356396675109863,
+      "learning_rate": 1.998266296809986e-05,
+      "loss": 1.0764,
+      "step": 1926
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.952024936676025,
+      "learning_rate": 1.997746185852982e-05,
+      "loss": 0.9852,
+      "step": 1927
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.497747421264648,
+      "learning_rate": 1.997226074895978e-05,
+      "loss": 0.9103,
+      "step": 1928
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.837340354919434,
+      "learning_rate": 1.9967059639389737e-05,
+      "loss": 0.9342,
+      "step": 1929
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.383747100830078,
+      "learning_rate": 1.9961858529819696e-05,
+      "loss": 0.8522,
+      "step": 1930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.453402519226074,
+      "learning_rate": 1.9956657420249654e-05,
+      "loss": 0.863,
+      "step": 1931
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.432109832763672,
+      "learning_rate": 1.9951456310679613e-05,
+      "loss": 1.0155,
+      "step": 1932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.150177001953125,
+      "learning_rate": 1.9946255201109572e-05,
+      "loss": 1.0367,
+      "step": 1933
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.609443664550781,
+      "learning_rate": 1.994105409153953e-05,
+      "loss": 1.0919,
+      "step": 1934
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.569661140441895,
+      "learning_rate": 1.9935852981969486e-05,
+      "loss": 1.2232,
+      "step": 1935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.538910865783691,
+      "learning_rate": 1.9930651872399448e-05,
+      "loss": 1.0162,
+      "step": 1936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.364474296569824,
+      "learning_rate": 1.9925450762829403e-05,
+      "loss": 0.8026,
+      "step": 1937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.824756622314453,
+      "learning_rate": 1.9920249653259365e-05,
+      "loss": 0.716,
+      "step": 1938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.850571632385254,
+      "learning_rate": 1.991504854368932e-05,
+      "loss": 1.0227,
+      "step": 1939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.760404586791992,
+      "learning_rate": 1.9909847434119282e-05,
+      "loss": 1.2445,
+      "step": 1940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.572029113769531,
+      "learning_rate": 1.9904646324549237e-05,
+      "loss": 1.0212,
+      "step": 1941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.536205291748047,
+      "learning_rate": 1.9899445214979196e-05,
+      "loss": 1.0638,
+      "step": 1942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.509013175964355,
+      "learning_rate": 1.9894244105409154e-05,
+      "loss": 0.9248,
+      "step": 1943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.368556022644043,
+      "learning_rate": 1.9889042995839113e-05,
+      "loss": 1.0585,
+      "step": 1944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.215378761291504,
+      "learning_rate": 1.988384188626907e-05,
+      "loss": 0.9742,
+      "step": 1945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.661454200744629,
+      "learning_rate": 1.987864077669903e-05,
+      "loss": 1.0973,
+      "step": 1946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.293997764587402,
+      "learning_rate": 1.987343966712899e-05,
+      "loss": 1.1377,
+      "step": 1947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.668646812438965,
+      "learning_rate": 1.9868238557558948e-05,
+      "loss": 0.7215,
+      "step": 1948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.183780670166016,
+      "learning_rate": 1.9863037447988903e-05,
+      "loss": 1.0707,
+      "step": 1949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.956949234008789,
+      "learning_rate": 1.9857836338418865e-05,
+      "loss": 1.0052,
+      "step": 1950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.312695503234863,
+      "learning_rate": 1.985263522884882e-05,
+      "loss": 1.1984,
+      "step": 1951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.824272155761719,
+      "learning_rate": 1.9847434119278782e-05,
+      "loss": 0.9319,
+      "step": 1952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.198112964630127,
+      "learning_rate": 1.9842233009708737e-05,
+      "loss": 0.7657,
+      "step": 1953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.413636207580566,
+      "learning_rate": 1.98370319001387e-05,
+      "loss": 1.2627,
+      "step": 1954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.368317604064941,
+      "learning_rate": 1.9831830790568654e-05,
+      "loss": 1.0284,
+      "step": 1955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.67740535736084,
+      "learning_rate": 1.9826629680998616e-05,
+      "loss": 1.0415,
+      "step": 1956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.864145278930664,
+      "learning_rate": 1.982142857142857e-05,
+      "loss": 0.7497,
+      "step": 1957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.6328125,
+      "learning_rate": 1.981622746185853e-05,
+      "loss": 0.9803,
+      "step": 1958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.184209823608398,
+      "learning_rate": 1.981102635228849e-05,
+      "loss": 1.1211,
+      "step": 1959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.850543022155762,
+      "learning_rate": 1.9805825242718447e-05,
+      "loss": 1.1116,
+      "step": 1960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.11820125579834,
+      "learning_rate": 1.9800624133148406e-05,
+      "loss": 0.7132,
+      "step": 1961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.309416770935059,
+      "learning_rate": 1.9795423023578365e-05,
+      "loss": 1.1328,
+      "step": 1962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.136751174926758,
+      "learning_rate": 1.9790221914008323e-05,
+      "loss": 0.9406,
+      "step": 1963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.008854866027832,
+      "learning_rate": 1.9785020804438282e-05,
+      "loss": 1.0326,
+      "step": 1964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.989263534545898,
+      "learning_rate": 1.9779819694868237e-05,
+      "loss": 1.2186,
+      "step": 1965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.428730964660645,
+      "learning_rate": 1.97746185852982e-05,
+      "loss": 0.916,
+      "step": 1966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.886207580566406,
+      "learning_rate": 1.9769417475728154e-05,
+      "loss": 0.8013,
+      "step": 1967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.323176383972168,
+      "learning_rate": 1.9764216366158116e-05,
+      "loss": 0.7887,
+      "step": 1968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.727721214294434,
+      "learning_rate": 1.975901525658807e-05,
+      "loss": 1.0273,
+      "step": 1969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.624381065368652,
+      "learning_rate": 1.9753814147018034e-05,
+      "loss": 0.9516,
+      "step": 1970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.473042488098145,
+      "learning_rate": 1.974861303744799e-05,
+      "loss": 1.0007,
+      "step": 1971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.513883590698242,
+      "learning_rate": 1.9743411927877947e-05,
+      "loss": 0.8634,
+      "step": 1972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.367743492126465,
+      "learning_rate": 1.9738210818307906e-05,
+      "loss": 1.251,
+      "step": 1973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.775065898895264,
+      "learning_rate": 1.9733009708737865e-05,
+      "loss": 0.7074,
+      "step": 1974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.83116626739502,
+      "learning_rate": 1.9727808599167823e-05,
+      "loss": 1.1902,
+      "step": 1975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.30637264251709,
+      "learning_rate": 1.9722607489597782e-05,
+      "loss": 1.0356,
+      "step": 1976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.544965744018555,
+      "learning_rate": 1.971740638002774e-05,
+      "loss": 0.8709,
+      "step": 1977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.072301864624023,
+      "learning_rate": 1.97122052704577e-05,
+      "loss": 1.0345,
+      "step": 1978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.481698989868164,
+      "learning_rate": 1.9707004160887658e-05,
+      "loss": 1.0729,
+      "step": 1979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.442758560180664,
+      "learning_rate": 1.9701803051317616e-05,
+      "loss": 0.896,
+      "step": 1980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.822319030761719,
+      "learning_rate": 1.969660194174757e-05,
+      "loss": 0.9175,
+      "step": 1981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.676434516906738,
+      "learning_rate": 1.9691400832177534e-05,
+      "loss": 0.9214,
+      "step": 1982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.482654571533203,
+      "learning_rate": 1.968619972260749e-05,
+      "loss": 1.0187,
+      "step": 1983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.855206489562988,
+      "learning_rate": 1.968099861303745e-05,
+      "loss": 1.3318,
+      "step": 1984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.831679344177246,
+      "learning_rate": 1.9675797503467406e-05,
+      "loss": 0.8803,
+      "step": 1985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.750319004058838,
+      "learning_rate": 1.9670596393897368e-05,
+      "loss": 0.837,
+      "step": 1986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.4572529792785645,
+      "learning_rate": 1.9665395284327323e-05,
+      "loss": 0.672,
+      "step": 1987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.618048667907715,
+      "learning_rate": 1.9660194174757282e-05,
+      "loss": 0.871,
+      "step": 1988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.413116455078125,
+      "learning_rate": 1.965499306518724e-05,
+      "loss": 0.9491,
+      "step": 1989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.271251678466797,
+      "learning_rate": 1.96497919556172e-05,
+      "loss": 0.9297,
+      "step": 1990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.305636405944824,
+      "learning_rate": 1.9644590846047158e-05,
+      "loss": 1.2229,
+      "step": 1991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.503952980041504,
+      "learning_rate": 1.9639389736477116e-05,
+      "loss": 0.8154,
+      "step": 1992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.915284156799316,
+      "learning_rate": 1.9634188626907075e-05,
+      "loss": 0.9179,
+      "step": 1993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.02058219909668,
+      "learning_rate": 1.9628987517337034e-05,
+      "loss": 0.6974,
+      "step": 1994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.367935180664062,
+      "learning_rate": 1.962378640776699e-05,
+      "loss": 0.8111,
+      "step": 1995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.73412036895752,
+      "learning_rate": 1.961858529819695e-05,
+      "loss": 1.0132,
+      "step": 1996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.980670928955078,
+      "learning_rate": 1.9613384188626906e-05,
+      "loss": 0.744,
+      "step": 1997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.067268371582031,
+      "learning_rate": 1.9608183079056868e-05,
+      "loss": 0.7093,
+      "step": 1998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.334397315979004,
+      "learning_rate": 1.9602981969486823e-05,
+      "loss": 0.97,
+      "step": 1999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.687064170837402,
+      "learning_rate": 1.9597780859916785e-05,
+      "loss": 1.0755,
+      "step": 2000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.652106285095215,
+      "learning_rate": 1.959257975034674e-05,
+      "loss": 0.9779,
+      "step": 2001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.590825080871582,
+      "learning_rate": 1.9587378640776702e-05,
+      "loss": 1.07,
+      "step": 2002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.793490409851074,
+      "learning_rate": 1.9582177531206658e-05,
+      "loss": 0.9522,
+      "step": 2003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.043313026428223,
+      "learning_rate": 1.9576976421636616e-05,
+      "loss": 1.1544,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.957944869995117,
+      "learning_rate": 1.9571775312066575e-05,
+      "loss": 0.9598,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.078780174255371,
+      "learning_rate": 1.9566574202496533e-05,
+      "loss": 0.8192,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.922538757324219,
+      "learning_rate": 1.9561373092926492e-05,
+      "loss": 0.9312,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.726404190063477,
+      "learning_rate": 1.955617198335645e-05,
+      "loss": 0.9913,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.442599296569824,
+      "learning_rate": 1.955097087378641e-05,
+      "loss": 1.0741,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.096790313720703,
+      "learning_rate": 1.9545769764216368e-05,
+      "loss": 1.2073,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.42972469329834,
+      "learning_rate": 1.9540568654646323e-05,
+      "loss": 1.0211,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.446356773376465,
+      "learning_rate": 1.9535367545076285e-05,
+      "loss": 0.9855,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.08479118347168,
+      "learning_rate": 1.953016643550624e-05,
+      "loss": 1.0114,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.132721900939941,
+      "learning_rate": 1.9524965325936202e-05,
+      "loss": 0.922,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.556670188903809,
+      "learning_rate": 1.9519764216366158e-05,
+      "loss": 0.9425,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.472869873046875,
+      "learning_rate": 1.951456310679612e-05,
+      "loss": 0.7691,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.536955833435059,
+      "learning_rate": 1.9509361997226075e-05,
+      "loss": 0.903,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.88356876373291,
+      "learning_rate": 1.9504160887656033e-05,
+      "loss": 1.0953,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.646073341369629,
+      "learning_rate": 1.9498959778085992e-05,
+      "loss": 0.864,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.293815612792969,
+      "learning_rate": 1.949375866851595e-05,
+      "loss": 0.9991,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.753483772277832,
+      "learning_rate": 1.948855755894591e-05,
+      "loss": 0.9517,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.558955192565918,
+      "learning_rate": 1.9483356449375868e-05,
+      "loss": 0.8722,
+      "step": 2022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.219301223754883,
+      "learning_rate": 1.9478155339805826e-05,
+      "loss": 0.8832,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.743721961975098,
+      "learning_rate": 1.9472954230235785e-05,
+      "loss": 0.6626,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.82735538482666,
+      "learning_rate": 1.9467753120665744e-05,
+      "loss": 0.8755,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.090924263000488,
+      "learning_rate": 1.9462552011095702e-05,
+      "loss": 0.9198,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.881436347961426,
+      "learning_rate": 1.9457350901525658e-05,
+      "loss": 0.772,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.205047607421875,
+      "learning_rate": 1.945214979195562e-05,
+      "loss": 0.8558,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.38953685760498,
+      "learning_rate": 1.9446948682385575e-05,
+      "loss": 0.959,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.119656562805176,
+      "learning_rate": 1.9441747572815537e-05,
+      "loss": 0.8648,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.595319747924805,
+      "learning_rate": 1.9436546463245492e-05,
+      "loss": 0.7973,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.60218620300293,
+      "learning_rate": 1.9431345353675454e-05,
+      "loss": 1.0152,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.983809947967529,
+      "learning_rate": 1.942614424410541e-05,
+      "loss": 0.9678,
+      "step": 2033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.914435386657715,
+      "learning_rate": 1.9420943134535368e-05,
+      "loss": 0.9479,
+      "step": 2034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.48495101928711,
+      "learning_rate": 1.9415742024965326e-05,
+      "loss": 1.0644,
+      "step": 2035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.40905475616455,
+      "learning_rate": 1.9410540915395285e-05,
+      "loss": 0.7463,
+      "step": 2036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.827081680297852,
+      "learning_rate": 1.9405339805825244e-05,
+      "loss": 1.0272,
+      "step": 2037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.265745162963867,
+      "learning_rate": 1.9400138696255202e-05,
+      "loss": 0.6327,
+      "step": 2038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.547269821166992,
+      "learning_rate": 1.939493758668516e-05,
+      "loss": 1.0091,
+      "step": 2039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.374234199523926,
+      "learning_rate": 1.938973647711512e-05,
+      "loss": 1.0395,
+      "step": 2040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.79951000213623,
+      "learning_rate": 1.9384535367545078e-05,
+      "loss": 0.908,
+      "step": 2041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.179064750671387,
+      "learning_rate": 1.9379334257975037e-05,
+      "loss": 0.7608,
+      "step": 2042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.370891571044922,
+      "learning_rate": 1.9374133148404992e-05,
+      "loss": 0.8588,
+      "step": 2043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.158045768737793,
+      "learning_rate": 1.9368932038834954e-05,
+      "loss": 0.8422,
+      "step": 2044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.608132362365723,
+      "learning_rate": 1.936373092926491e-05,
+      "loss": 1.05,
+      "step": 2045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.803508758544922,
+      "learning_rate": 1.935852981969487e-05,
+      "loss": 0.9474,
+      "step": 2046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.3163628578186035,
+      "learning_rate": 1.9353328710124826e-05,
+      "loss": 0.6914,
+      "step": 2047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.768494606018066,
+      "learning_rate": 1.934812760055479e-05,
+      "loss": 1.089,
+      "step": 2048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.81220817565918,
+      "learning_rate": 1.9342926490984744e-05,
+      "loss": 0.6591,
+      "step": 2049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.743836402893066,
+      "learning_rate": 1.9337725381414702e-05,
+      "loss": 1.0297,
+      "step": 2050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.687013626098633,
+      "learning_rate": 1.933252427184466e-05,
+      "loss": 1.0166,
+      "step": 2051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.903140544891357,
+      "learning_rate": 1.932732316227462e-05,
+      "loss": 0.7113,
+      "step": 2052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.372147560119629,
+      "learning_rate": 1.9322122052704578e-05,
+      "loss": 1.2247,
+      "step": 2053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.53331470489502,
+      "learning_rate": 1.9316920943134537e-05,
+      "loss": 0.9602,
+      "step": 2054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.326908111572266,
+      "learning_rate": 1.9311719833564495e-05,
+      "loss": 0.7912,
+      "step": 2055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.111473083496094,
+      "learning_rate": 1.9306518723994454e-05,
+      "loss": 0.9509,
+      "step": 2056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.502007484436035,
+      "learning_rate": 1.930131761442441e-05,
+      "loss": 0.98,
+      "step": 2057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.454950332641602,
+      "learning_rate": 1.929611650485437e-05,
+      "loss": 0.6444,
+      "step": 2058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.469954490661621,
+      "learning_rate": 1.9290915395284326e-05,
+      "loss": 0.9352,
+      "step": 2059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.426725387573242,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 1.0196,
+      "step": 2060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.399835586547852,
+      "learning_rate": 1.9280513176144244e-05,
+      "loss": 0.9294,
+      "step": 2061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.630615234375,
+      "learning_rate": 1.9275312066574206e-05,
+      "loss": 1.114,
+      "step": 2062
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.40657901763916,
+      "learning_rate": 1.927011095700416e-05,
+      "loss": 1.0898,
+      "step": 2063
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.11788272857666,
+      "learning_rate": 1.9264909847434123e-05,
+      "loss": 1.0423,
+      "step": 2064
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.823526859283447,
+      "learning_rate": 1.9259708737864078e-05,
+      "loss": 0.823,
+      "step": 2065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.728775024414062,
+      "learning_rate": 1.9254507628294037e-05,
+      "loss": 0.8101,
+      "step": 2066
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.299310684204102,
+      "learning_rate": 1.9249306518723995e-05,
+      "loss": 1.2788,
+      "step": 2067
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.6565399169921875,
+      "learning_rate": 1.9244105409153954e-05,
+      "loss": 1.0632,
+      "step": 2068
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.727337837219238,
+      "learning_rate": 1.9238904299583913e-05,
+      "loss": 1.0,
+      "step": 2069
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.737884998321533,
+      "learning_rate": 1.923370319001387e-05,
+      "loss": 0.9133,
+      "step": 2070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.303391456604004,
+      "learning_rate": 1.922850208044383e-05,
+      "loss": 0.8154,
+      "step": 2071
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.779125213623047,
+      "learning_rate": 1.922330097087379e-05,
+      "loss": 1.0296,
+      "step": 2072
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.517130374908447,
+      "learning_rate": 1.9218099861303744e-05,
+      "loss": 0.9487,
+      "step": 2073
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.026442527770996,
+      "learning_rate": 1.9212898751733706e-05,
+      "loss": 1.0332,
+      "step": 2074
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.580415725708008,
+      "learning_rate": 1.920769764216366e-05,
+      "loss": 1.2613,
+      "step": 2075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.269636154174805,
+      "learning_rate": 1.9202496532593623e-05,
+      "loss": 1.0194,
+      "step": 2076
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.188154220581055,
+      "learning_rate": 1.9197295423023578e-05,
+      "loss": 0.9622,
+      "step": 2077
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.251532554626465,
+      "learning_rate": 1.919209431345354e-05,
+      "loss": 0.938,
+      "step": 2078
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.177571296691895,
+      "learning_rate": 1.9186893203883495e-05,
+      "loss": 0.8552,
+      "step": 2079
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.557371139526367,
+      "learning_rate": 1.9181692094313454e-05,
+      "loss": 0.7792,
+      "step": 2080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.38702917098999,
+      "learning_rate": 1.9176490984743412e-05,
+      "loss": 0.9306,
+      "step": 2081
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.370863914489746,
+      "learning_rate": 1.917128987517337e-05,
+      "loss": 0.9548,
+      "step": 2082
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.195610046386719,
+      "learning_rate": 1.916608876560333e-05,
+      "loss": 0.9193,
+      "step": 2083
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.917121887207031,
+      "learning_rate": 1.9160887656033288e-05,
+      "loss": 0.7968,
+      "step": 2084
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.3883695602417,
+      "learning_rate": 1.9155686546463247e-05,
+      "loss": 0.8685,
+      "step": 2085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.936119079589844,
+      "learning_rate": 1.9150485436893206e-05,
+      "loss": 0.7708,
+      "step": 2086
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.466489791870117,
+      "learning_rate": 1.9145284327323164e-05,
+      "loss": 0.9006,
+      "step": 2087
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.05833625793457,
+      "learning_rate": 1.9140083217753123e-05,
+      "loss": 0.8871,
+      "step": 2088
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.733646392822266,
+      "learning_rate": 1.9134882108183078e-05,
+      "loss": 0.9194,
+      "step": 2089
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.680283546447754,
+      "learning_rate": 1.912968099861304e-05,
+      "loss": 0.9495,
+      "step": 2090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.766976356506348,
+      "learning_rate": 1.9124479889042995e-05,
+      "loss": 0.9033,
+      "step": 2091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.9521613121032715,
+      "learning_rate": 1.9119278779472957e-05,
+      "loss": 0.9229,
+      "step": 2092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.641468048095703,
+      "learning_rate": 1.9114077669902912e-05,
+      "loss": 1.4355,
+      "step": 2093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.618013381958008,
+      "learning_rate": 1.9108876560332874e-05,
+      "loss": 0.9756,
+      "step": 2094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.770188331604004,
+      "learning_rate": 1.910367545076283e-05,
+      "loss": 0.9344,
+      "step": 2095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.623730659484863,
+      "learning_rate": 1.9098474341192788e-05,
+      "loss": 0.7913,
+      "step": 2096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.649590492248535,
+      "learning_rate": 1.9093273231622747e-05,
+      "loss": 1.1402,
+      "step": 2097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.936155319213867,
+      "learning_rate": 1.9088072122052705e-05,
+      "loss": 0.6855,
+      "step": 2098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.521100044250488,
+      "learning_rate": 1.9082871012482664e-05,
+      "loss": 1.0438,
+      "step": 2099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.807062149047852,
+      "learning_rate": 1.9077669902912623e-05,
+      "loss": 0.9439,
+      "step": 2100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.803032875061035,
+      "learning_rate": 1.907246879334258e-05,
+      "loss": 1.0596,
+      "step": 2101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.236330032348633,
+      "learning_rate": 1.906726768377254e-05,
+      "loss": 0.7866,
+      "step": 2102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.636486053466797,
+      "learning_rate": 1.90620665742025e-05,
+      "loss": 1.0252,
+      "step": 2103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.041923522949219,
+      "learning_rate": 1.9056865464632457e-05,
+      "loss": 0.7515,
+      "step": 2104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.264223098754883,
+      "learning_rate": 1.9051664355062412e-05,
+      "loss": 0.9622,
+      "step": 2105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.463504791259766,
+      "learning_rate": 1.9046463245492374e-05,
+      "loss": 1.2029,
+      "step": 2106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.661527633666992,
+      "learning_rate": 1.904126213592233e-05,
+      "loss": 0.8332,
+      "step": 2107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.251303195953369,
+      "learning_rate": 1.903606102635229e-05,
+      "loss": 0.849,
+      "step": 2108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.008356094360352,
+      "learning_rate": 1.9030859916782247e-05,
+      "loss": 1.0656,
+      "step": 2109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.574713706970215,
+      "learning_rate": 1.902565880721221e-05,
+      "loss": 1.2198,
+      "step": 2110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.594322681427002,
+      "learning_rate": 1.9020457697642164e-05,
+      "loss": 0.6383,
+      "step": 2111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.694341659545898,
+      "learning_rate": 1.9015256588072123e-05,
+      "loss": 0.9332,
+      "step": 2112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.3349027633667,
+      "learning_rate": 1.901005547850208e-05,
+      "loss": 1.1601,
+      "step": 2113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.435070037841797,
+      "learning_rate": 1.900485436893204e-05,
+      "loss": 0.9986,
+      "step": 2114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.733730792999268,
+      "learning_rate": 1.8999653259362e-05,
+      "loss": 0.7241,
+      "step": 2115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.48081636428833,
+      "learning_rate": 1.8994452149791957e-05,
+      "loss": 0.8055,
+      "step": 2116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.87580394744873,
+      "learning_rate": 1.8989251040221916e-05,
+      "loss": 1.1126,
+      "step": 2117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.852312088012695,
+      "learning_rate": 1.8984049930651874e-05,
+      "loss": 1.0604,
+      "step": 2118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.071954727172852,
+      "learning_rate": 1.897884882108183e-05,
+      "loss": 0.8878,
+      "step": 2119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.509331703186035,
+      "learning_rate": 1.897364771151179e-05,
+      "loss": 0.9331,
+      "step": 2120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.5162177085876465,
+      "learning_rate": 1.8968446601941747e-05,
+      "loss": 0.8107,
+      "step": 2121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.9848809242248535,
+      "learning_rate": 1.896324549237171e-05,
+      "loss": 0.9875,
+      "step": 2122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.150737762451172,
+      "learning_rate": 1.8958044382801664e-05,
+      "loss": 0.9652,
+      "step": 2123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.251764297485352,
+      "learning_rate": 1.8952843273231626e-05,
+      "loss": 0.7771,
+      "step": 2124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.258102416992188,
+      "learning_rate": 1.894764216366158e-05,
+      "loss": 0.7115,
+      "step": 2125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.184815406799316,
+      "learning_rate": 1.8942441054091543e-05,
+      "loss": 0.8252,
+      "step": 2126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.5755615234375,
+      "learning_rate": 1.89372399445215e-05,
+      "loss": 0.8764,
+      "step": 2127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.520837783813477,
+      "learning_rate": 1.8932038834951457e-05,
+      "loss": 0.7124,
+      "step": 2128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.503242492675781,
+      "learning_rate": 1.8926837725381416e-05,
+      "loss": 0.8605,
+      "step": 2129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.22625732421875,
+      "learning_rate": 1.8921636615811374e-05,
+      "loss": 0.8653,
+      "step": 2130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.072173118591309,
+      "learning_rate": 1.8916435506241333e-05,
+      "loss": 0.8651,
+      "step": 2131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.462240219116211,
+      "learning_rate": 1.891123439667129e-05,
+      "loss": 1.1205,
+      "step": 2132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.588064193725586,
+      "learning_rate": 1.890603328710125e-05,
+      "loss": 0.929,
+      "step": 2133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.94345474243164,
+      "learning_rate": 1.890083217753121e-05,
+      "loss": 0.9191,
+      "step": 2134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.450584411621094,
+      "learning_rate": 1.8895631067961164e-05,
+      "loss": 1.1645,
+      "step": 2135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.363498210906982,
+      "learning_rate": 1.8890429958391126e-05,
+      "loss": 0.8147,
+      "step": 2136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.624772071838379,
+      "learning_rate": 1.888522884882108e-05,
+      "loss": 0.8046,
+      "step": 2137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.755500793457031,
+      "learning_rate": 1.8880027739251043e-05,
+      "loss": 0.9705,
+      "step": 2138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.918044090270996,
+      "learning_rate": 1.8874826629681e-05,
+      "loss": 1.1202,
+      "step": 2139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.54166030883789,
+      "learning_rate": 1.886962552011096e-05,
+      "loss": 0.9204,
+      "step": 2140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.540922164916992,
+      "learning_rate": 1.8864424410540916e-05,
+      "loss": 1.012,
+      "step": 2141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.734829902648926,
+      "learning_rate": 1.8859223300970874e-05,
+      "loss": 0.8564,
+      "step": 2142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.56745719909668,
+      "learning_rate": 1.8854022191400833e-05,
+      "loss": 0.9095,
+      "step": 2143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.004289627075195,
+      "learning_rate": 1.884882108183079e-05,
+      "loss": 1.1517,
+      "step": 2144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.846745491027832,
+      "learning_rate": 1.884361997226075e-05,
+      "loss": 1.0248,
+      "step": 2145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.573404312133789,
+      "learning_rate": 1.883841886269071e-05,
+      "loss": 1.1074,
+      "step": 2146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.259777069091797,
+      "learning_rate": 1.8833217753120667e-05,
+      "loss": 1.0786,
+      "step": 2147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.032732963562012,
+      "learning_rate": 1.8828016643550626e-05,
+      "loss": 1.0075,
+      "step": 2148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.881424903869629,
+      "learning_rate": 1.8822815533980585e-05,
+      "loss": 0.9605,
+      "step": 2149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.001380920410156,
+      "learning_rate": 1.8817614424410543e-05,
+      "loss": 0.7307,
+      "step": 2150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.841034889221191,
+      "learning_rate": 1.88124133148405e-05,
+      "loss": 1.1413,
+      "step": 2151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.197729110717773,
+      "learning_rate": 1.880721220527046e-05,
+      "loss": 0.8443,
+      "step": 2152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.256385803222656,
+      "learning_rate": 1.8802011095700416e-05,
+      "loss": 1.1418,
+      "step": 2153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.092263221740723,
+      "learning_rate": 1.8796809986130378e-05,
+      "loss": 0.884,
+      "step": 2154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.154441833496094,
+      "learning_rate": 1.8791608876560333e-05,
+      "loss": 0.8684,
+      "step": 2155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.698145866394043,
+      "learning_rate": 1.8786407766990295e-05,
+      "loss": 0.9267,
+      "step": 2156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.537955284118652,
+      "learning_rate": 1.878120665742025e-05,
+      "loss": 0.9478,
+      "step": 2157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.632209777832031,
+      "learning_rate": 1.877600554785021e-05,
+      "loss": 1.1214,
+      "step": 2158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.722085952758789,
+      "learning_rate": 1.8770804438280167e-05,
+      "loss": 0.9345,
+      "step": 2159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.012622833251953,
+      "learning_rate": 1.8765603328710126e-05,
+      "loss": 0.9282,
+      "step": 2160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.907605171203613,
+      "learning_rate": 1.8760402219140085e-05,
+      "loss": 0.8733,
+      "step": 2161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.06229305267334,
+      "learning_rate": 1.8755201109570043e-05,
+      "loss": 1.2021,
+      "step": 2162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.526949882507324,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.9438,
+      "step": 2163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.348458290100098,
+      "learning_rate": 1.8744798890429957e-05,
+      "loss": 1.0672,
+      "step": 2164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.976863861083984,
+      "learning_rate": 1.8739597780859916e-05,
+      "loss": 1.0734,
+      "step": 2165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.545647621154785,
+      "learning_rate": 1.8734396671289874e-05,
+      "loss": 0.7463,
+      "step": 2166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.845981597900391,
+      "learning_rate": 1.8729195561719833e-05,
+      "loss": 0.934,
+      "step": 2167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.080180168151855,
+      "learning_rate": 1.872399445214979e-05,
+      "loss": 0.9144,
+      "step": 2168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.11190128326416,
+      "learning_rate": 1.871879334257975e-05,
+      "loss": 0.8645,
+      "step": 2169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.8780837059021,
+      "learning_rate": 1.871359223300971e-05,
+      "loss": 0.8808,
+      "step": 2170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.908954620361328,
+      "learning_rate": 1.8708391123439667e-05,
+      "loss": 0.934,
+      "step": 2171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.9501800537109375,
+      "learning_rate": 1.8703190013869626e-05,
+      "loss": 0.7466,
+      "step": 2172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.496336936950684,
+      "learning_rate": 1.8697988904299584e-05,
+      "loss": 1.0268,
+      "step": 2173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.970081329345703,
+      "learning_rate": 1.869278779472954e-05,
+      "loss": 0.8517,
+      "step": 2174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.94469165802002,
+      "learning_rate": 1.8687586685159502e-05,
+      "loss": 0.9174,
+      "step": 2175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.182271003723145,
+      "learning_rate": 1.8682385575589457e-05,
+      "loss": 0.8543,
+      "step": 2176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.8668794631958,
+      "learning_rate": 1.867718446601942e-05,
+      "loss": 0.9299,
+      "step": 2177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.724179744720459,
+      "learning_rate": 1.8671983356449374e-05,
+      "loss": 0.8318,
+      "step": 2178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.029430389404297,
+      "learning_rate": 1.8666782246879336e-05,
+      "loss": 0.6457,
+      "step": 2179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.864494323730469,
+      "learning_rate": 1.866158113730929e-05,
+      "loss": 0.8679,
+      "step": 2180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.018725395202637,
+      "learning_rate": 1.865638002773925e-05,
+      "loss": 0.9219,
+      "step": 2181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.799222946166992,
+      "learning_rate": 1.865117891816921e-05,
+      "loss": 0.7283,
+      "step": 2182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.699052810668945,
+      "learning_rate": 1.8645977808599167e-05,
+      "loss": 0.9226,
+      "step": 2183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.059530258178711,
+      "learning_rate": 1.8640776699029126e-05,
+      "loss": 0.8706,
+      "step": 2184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.743648529052734,
+      "learning_rate": 1.8635575589459084e-05,
+      "loss": 0.7915,
+      "step": 2185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.090658187866211,
+      "learning_rate": 1.8630374479889043e-05,
+      "loss": 0.7476,
+      "step": 2186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.281763076782227,
+      "learning_rate": 1.8625173370319e-05,
+      "loss": 1.326,
+      "step": 2187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.540687561035156,
+      "learning_rate": 1.861997226074896e-05,
+      "loss": 0.716,
+      "step": 2188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.704326629638672,
+      "learning_rate": 1.861477115117892e-05,
+      "loss": 1.0025,
+      "step": 2189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.876301765441895,
+      "learning_rate": 1.8609570041608874e-05,
+      "loss": 1.0494,
+      "step": 2190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.392189979553223,
+      "learning_rate": 1.8604368932038836e-05,
+      "loss": 0.9549,
+      "step": 2191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.822389602661133,
+      "learning_rate": 1.859916782246879e-05,
+      "loss": 0.9177,
+      "step": 2192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.604636192321777,
+      "learning_rate": 1.8593966712898753e-05,
+      "loss": 0.9384,
+      "step": 2193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.229218482971191,
+      "learning_rate": 1.858876560332871e-05,
+      "loss": 1.0134,
+      "step": 2194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.998059272766113,
+      "learning_rate": 1.858356449375867e-05,
+      "loss": 0.8782,
+      "step": 2195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.4927978515625,
+      "learning_rate": 1.8578363384188626e-05,
+      "loss": 1.172,
+      "step": 2196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.218864440917969,
+      "learning_rate": 1.8573162274618584e-05,
+      "loss": 0.8625,
+      "step": 2197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.244900703430176,
+      "learning_rate": 1.8567961165048543e-05,
+      "loss": 0.8598,
+      "step": 2198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.650553703308105,
+      "learning_rate": 1.85627600554785e-05,
+      "loss": 0.989,
+      "step": 2199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.119291305541992,
+      "learning_rate": 1.855755894590846e-05,
+      "loss": 1.0673,
+      "step": 2200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.532360076904297,
+      "learning_rate": 1.855235783633842e-05,
+      "loss": 0.7948,
+      "step": 2201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.159000396728516,
+      "learning_rate": 1.8547156726768377e-05,
+      "loss": 0.7696,
+      "step": 2202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.85336971282959,
+      "learning_rate": 1.8541955617198336e-05,
+      "loss": 1.0071,
+      "step": 2203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.341078758239746,
+      "learning_rate": 1.853675450762829e-05,
+      "loss": 1.0783,
+      "step": 2204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.988049030303955,
+      "learning_rate": 1.8531553398058253e-05,
+      "loss": 0.7235,
+      "step": 2205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.791369438171387,
+      "learning_rate": 1.852635228848821e-05,
+      "loss": 0.7335,
+      "step": 2206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.099921226501465,
+      "learning_rate": 1.852115117891817e-05,
+      "loss": 0.8204,
+      "step": 2207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.476777076721191,
+      "learning_rate": 1.8515950069348126e-05,
+      "loss": 0.9082,
+      "step": 2208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.101350784301758,
+      "learning_rate": 1.8510748959778088e-05,
+      "loss": 0.7992,
+      "step": 2209
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.631999969482422,
+      "learning_rate": 1.8505547850208043e-05,
+      "loss": 0.9138,
+      "step": 2210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.243071556091309,
+      "learning_rate": 1.8500346740638005e-05,
+      "loss": 0.8018,
+      "step": 2211
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.777984142303467,
+      "learning_rate": 1.849514563106796e-05,
+      "loss": 0.808,
+      "step": 2212
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.062716484069824,
+      "learning_rate": 1.848994452149792e-05,
+      "loss": 0.8887,
+      "step": 2213
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.70878791809082,
+      "learning_rate": 1.8484743411927877e-05,
+      "loss": 0.9179,
+      "step": 2214
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.725410461425781,
+      "learning_rate": 1.8479542302357836e-05,
+      "loss": 1.1297,
+      "step": 2215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.332430839538574,
+      "learning_rate": 1.8474341192787795e-05,
+      "loss": 1.2206,
+      "step": 2216
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.239778518676758,
+      "learning_rate": 1.8469140083217753e-05,
+      "loss": 0.7479,
+      "step": 2217
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.386820793151855,
+      "learning_rate": 1.8463938973647712e-05,
+      "loss": 1.1325,
+      "step": 2218
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.389747619628906,
+      "learning_rate": 1.845873786407767e-05,
+      "loss": 0.6831,
+      "step": 2219
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.92628288269043,
+      "learning_rate": 1.8453536754507626e-05,
+      "loss": 0.984,
+      "step": 2220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.7384033203125,
+      "learning_rate": 1.8448335644937588e-05,
+      "loss": 0.8993,
+      "step": 2221
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.0441312789917,
+      "learning_rate": 1.8443134535367543e-05,
+      "loss": 1.037,
+      "step": 2222
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.614768028259277,
+      "learning_rate": 1.8437933425797505e-05,
+      "loss": 1.0117,
+      "step": 2223
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.19947338104248,
+      "learning_rate": 1.843273231622746e-05,
+      "loss": 1.0617,
+      "step": 2224
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.319229125976562,
+      "learning_rate": 1.8427531206657422e-05,
+      "loss": 0.9729,
+      "step": 2225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.23625659942627,
+      "learning_rate": 1.8422330097087377e-05,
+      "loss": 1.0378,
+      "step": 2226
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.567047119140625,
+      "learning_rate": 1.8417128987517336e-05,
+      "loss": 0.7274,
+      "step": 2227
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.378421783447266,
+      "learning_rate": 1.8411927877947295e-05,
+      "loss": 0.9632,
+      "step": 2228
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.822378158569336,
+      "learning_rate": 1.8406726768377253e-05,
+      "loss": 0.681,
+      "step": 2229
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.502187728881836,
+      "learning_rate": 1.8401525658807212e-05,
+      "loss": 1.0021,
+      "step": 2230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.790970802307129,
+      "learning_rate": 1.839632454923717e-05,
+      "loss": 1.0371,
+      "step": 2231
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.35094928741455,
+      "learning_rate": 1.839112343966713e-05,
+      "loss": 1.2098,
+      "step": 2232
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.747018814086914,
+      "learning_rate": 1.8385922330097088e-05,
+      "loss": 0.9065,
+      "step": 2233
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.302157402038574,
+      "learning_rate": 1.8380721220527046e-05,
+      "loss": 0.9799,
+      "step": 2234
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.360464096069336,
+      "learning_rate": 1.8375520110957005e-05,
+      "loss": 0.8187,
+      "step": 2235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.154192924499512,
+      "learning_rate": 1.837031900138696e-05,
+      "loss": 1.0068,
+      "step": 2236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.185779571533203,
+      "learning_rate": 1.8365117891816922e-05,
+      "loss": 1.0812,
+      "step": 2237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.261066436767578,
+      "learning_rate": 1.8359916782246877e-05,
+      "loss": 0.9472,
+      "step": 2238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.425460338592529,
+      "learning_rate": 1.835471567267684e-05,
+      "loss": 0.6453,
+      "step": 2239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.561004638671875,
+      "learning_rate": 1.8349514563106795e-05,
+      "loss": 0.9135,
+      "step": 2240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.108255386352539,
+      "learning_rate": 1.8344313453536757e-05,
+      "loss": 1.0979,
+      "step": 2241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.77114486694336,
+      "learning_rate": 1.8339112343966712e-05,
+      "loss": 0.9154,
+      "step": 2242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.982231140136719,
+      "learning_rate": 1.833391123439667e-05,
+      "loss": 0.8853,
+      "step": 2243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.380264282226562,
+      "learning_rate": 1.832871012482663e-05,
+      "loss": 0.8926,
+      "step": 2244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.904301643371582,
+      "learning_rate": 1.8323509015256588e-05,
+      "loss": 1.4021,
+      "step": 2245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.555354118347168,
+      "learning_rate": 1.8318307905686546e-05,
+      "loss": 0.894,
+      "step": 2246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.472662925720215,
+      "learning_rate": 1.8313106796116505e-05,
+      "loss": 0.8173,
+      "step": 2247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.41767406463623,
+      "learning_rate": 1.8307905686546463e-05,
+      "loss": 0.891,
+      "step": 2248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.844581604003906,
+      "learning_rate": 1.8302704576976422e-05,
+      "loss": 0.8523,
+      "step": 2249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.160209655761719,
+      "learning_rate": 1.829750346740638e-05,
+      "loss": 0.8166,
+      "step": 2250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.458903312683105,
+      "learning_rate": 1.829230235783634e-05,
+      "loss": 1.2381,
+      "step": 2251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.618772506713867,
+      "learning_rate": 1.8287101248266295e-05,
+      "loss": 1.1545,
+      "step": 2252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.740551948547363,
+      "learning_rate": 1.8281900138696257e-05,
+      "loss": 1.1544,
+      "step": 2253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.552860260009766,
+      "learning_rate": 1.8276699029126212e-05,
+      "loss": 1.0461,
+      "step": 2254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.742663383483887,
+      "learning_rate": 1.8271497919556174e-05,
+      "loss": 0.7478,
+      "step": 2255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.2740302085876465,
+      "learning_rate": 1.826629680998613e-05,
+      "loss": 0.9961,
+      "step": 2256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.049434661865234,
+      "learning_rate": 1.826109570041609e-05,
+      "loss": 0.9251,
+      "step": 2257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.418859958648682,
+      "learning_rate": 1.8255894590846046e-05,
+      "loss": 0.9214,
+      "step": 2258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.603795051574707,
+      "learning_rate": 1.8250693481276005e-05,
+      "loss": 1.0878,
+      "step": 2259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.814173221588135,
+      "learning_rate": 1.8245492371705963e-05,
+      "loss": 0.9679,
+      "step": 2260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.489156723022461,
+      "learning_rate": 1.8240291262135922e-05,
+      "loss": 0.8131,
+      "step": 2261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.267471313476562,
+      "learning_rate": 1.823509015256588e-05,
+      "loss": 0.8191,
+      "step": 2262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.708309173583984,
+      "learning_rate": 1.822988904299584e-05,
+      "loss": 0.5344,
+      "step": 2263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.450493812561035,
+      "learning_rate": 1.8224687933425798e-05,
+      "loss": 0.9291,
+      "step": 2264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.621163368225098,
+      "learning_rate": 1.8219486823855756e-05,
+      "loss": 1.0115,
+      "step": 2265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.47472095489502,
+      "learning_rate": 1.8214285714285712e-05,
+      "loss": 0.9163,
+      "step": 2266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.177391052246094,
+      "learning_rate": 1.8209084604715674e-05,
+      "loss": 0.751,
+      "step": 2267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.412423133850098,
+      "learning_rate": 1.820388349514563e-05,
+      "loss": 0.8896,
+      "step": 2268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.271450996398926,
+      "learning_rate": 1.819868238557559e-05,
+      "loss": 1.0005,
+      "step": 2269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.328194618225098,
+      "learning_rate": 1.8193481276005546e-05,
+      "loss": 0.7072,
+      "step": 2270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.805480003356934,
+      "learning_rate": 1.8188280166435508e-05,
+      "loss": 0.8577,
+      "step": 2271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.031120300292969,
+      "learning_rate": 1.8183079056865463e-05,
+      "loss": 0.8562,
+      "step": 2272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.564994812011719,
+      "learning_rate": 1.8177877947295425e-05,
+      "loss": 0.8527,
+      "step": 2273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.943829536437988,
+      "learning_rate": 1.817267683772538e-05,
+      "loss": 0.9243,
+      "step": 2274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.689786911010742,
+      "learning_rate": 1.816747572815534e-05,
+      "loss": 0.9336,
+      "step": 2275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.556715965270996,
+      "learning_rate": 1.8162274618585298e-05,
+      "loss": 0.8656,
+      "step": 2276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.078814506530762,
+      "learning_rate": 1.8157073509015256e-05,
+      "loss": 1.0346,
+      "step": 2277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.934518814086914,
+      "learning_rate": 1.8151872399445215e-05,
+      "loss": 1.3483,
+      "step": 2278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.373231887817383,
+      "learning_rate": 1.8146671289875174e-05,
+      "loss": 0.723,
+      "step": 2279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.002766609191895,
+      "learning_rate": 1.8141470180305132e-05,
+      "loss": 0.8015,
+      "step": 2280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.259267807006836,
+      "learning_rate": 1.813626907073509e-05,
+      "loss": 0.9232,
+      "step": 2281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.15833854675293,
+      "learning_rate": 1.8131067961165046e-05,
+      "loss": 0.7547,
+      "step": 2282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.243614196777344,
+      "learning_rate": 1.8125866851595008e-05,
+      "loss": 0.9682,
+      "step": 2283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.676841735839844,
+      "learning_rate": 1.8120665742024963e-05,
+      "loss": 1.0853,
+      "step": 2284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.822103977203369,
+      "learning_rate": 1.8115464632454925e-05,
+      "loss": 0.9282,
+      "step": 2285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.27798080444336,
+      "learning_rate": 1.811026352288488e-05,
+      "loss": 0.7805,
+      "step": 2286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.019153594970703,
+      "learning_rate": 1.8105062413314843e-05,
+      "loss": 1.0803,
+      "step": 2287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.102422714233398,
+      "learning_rate": 1.8099861303744798e-05,
+      "loss": 1.2279,
+      "step": 2288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.919480323791504,
+      "learning_rate": 1.8094660194174756e-05,
+      "loss": 0.84,
+      "step": 2289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.81380844116211,
+      "learning_rate": 1.8089459084604715e-05,
+      "loss": 1.0674,
+      "step": 2290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.009980201721191,
+      "learning_rate": 1.8084257975034674e-05,
+      "loss": 0.5791,
+      "step": 2291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.17934799194336,
+      "learning_rate": 1.8079056865464632e-05,
+      "loss": 0.9096,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.44701623916626,
+      "learning_rate": 1.807385575589459e-05,
+      "loss": 0.7013,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.309456825256348,
+      "learning_rate": 1.806865464632455e-05,
+      "loss": 1.0013,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.704399108886719,
+      "learning_rate": 1.8063453536754508e-05,
+      "loss": 0.8194,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.357321739196777,
+      "learning_rate": 1.8058252427184467e-05,
+      "loss": 0.8676,
+      "step": 2296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.572833061218262,
+      "learning_rate": 1.8053051317614425e-05,
+      "loss": 0.8308,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.491618156433105,
+      "learning_rate": 1.804785020804438e-05,
+      "loss": 0.7864,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.273808479309082,
+      "learning_rate": 1.8042649098474343e-05,
+      "loss": 0.8442,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.177206039428711,
+      "learning_rate": 1.8037447988904298e-05,
+      "loss": 1.0329,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.943525791168213,
+      "learning_rate": 1.803224687933426e-05,
+      "loss": 0.8758,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.744579315185547,
+      "learning_rate": 1.8027045769764215e-05,
+      "loss": 1.0219,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.535143852233887,
+      "learning_rate": 1.8021844660194177e-05,
+      "loss": 0.8825,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.548876762390137,
+      "learning_rate": 1.8016643550624132e-05,
+      "loss": 0.7406,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.887797355651855,
+      "learning_rate": 1.801144244105409e-05,
+      "loss": 0.9716,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.860045433044434,
+      "learning_rate": 1.800624133148405e-05,
+      "loss": 0.9836,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.644034385681152,
+      "learning_rate": 1.8001040221914008e-05,
+      "loss": 1.0605,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.061759948730469,
+      "learning_rate": 1.7995839112343967e-05,
+      "loss": 0.6944,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.507060050964355,
+      "learning_rate": 1.7990638002773925e-05,
+      "loss": 1.0255,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.723271369934082,
+      "learning_rate": 1.7985436893203884e-05,
+      "loss": 1.0848,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.693058967590332,
+      "learning_rate": 1.7980235783633843e-05,
+      "loss": 0.8871,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.021513938903809,
+      "learning_rate": 1.7975034674063798e-05,
+      "loss": 0.8019,
+      "step": 2312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.312256813049316,
+      "learning_rate": 1.796983356449376e-05,
+      "loss": 0.5588,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.426703453063965,
+      "learning_rate": 1.7964632454923715e-05,
+      "loss": 1.1014,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.841800689697266,
+      "learning_rate": 1.7959431345353677e-05,
+      "loss": 0.9038,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.388137817382812,
+      "learning_rate": 1.7954230235783632e-05,
+      "loss": 0.7854,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.095231056213379,
+      "learning_rate": 1.7949029126213594e-05,
+      "loss": 1.072,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.084694862365723,
+      "learning_rate": 1.794382801664355e-05,
+      "loss": 0.942,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.855873107910156,
+      "learning_rate": 1.793862690707351e-05,
+      "loss": 1.0792,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.30201244354248,
+      "learning_rate": 1.7933425797503467e-05,
+      "loss": 0.918,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.305950164794922,
+      "learning_rate": 1.7928224687933425e-05,
+      "loss": 0.696,
+      "step": 2321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.654512405395508,
+      "learning_rate": 1.7923023578363384e-05,
+      "loss": 0.911,
+      "step": 2322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.394330024719238,
+      "learning_rate": 1.7917822468793342e-05,
+      "loss": 0.9838,
+      "step": 2323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.136809825897217,
+      "learning_rate": 1.79126213592233e-05,
+      "loss": 0.5377,
+      "step": 2324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.087574005126953,
+      "learning_rate": 1.790742024965326e-05,
+      "loss": 0.8624,
+      "step": 2325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.497315406799316,
+      "learning_rate": 1.7902219140083218e-05,
+      "loss": 0.9406,
+      "step": 2326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.024170875549316,
+      "learning_rate": 1.7897018030513177e-05,
+      "loss": 0.7597,
+      "step": 2327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.496444702148438,
+      "learning_rate": 1.7891816920943132e-05,
+      "loss": 0.9568,
+      "step": 2328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.309422492980957,
+      "learning_rate": 1.7886615811373094e-05,
+      "loss": 1.0179,
+      "step": 2329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.805585861206055,
+      "learning_rate": 1.788141470180305e-05,
+      "loss": 0.8447,
+      "step": 2330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.118191719055176,
+      "learning_rate": 1.787621359223301e-05,
+      "loss": 1.0072,
+      "step": 2331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.073431015014648,
+      "learning_rate": 1.7871012482662967e-05,
+      "loss": 0.8856,
+      "step": 2332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.128653526306152,
+      "learning_rate": 1.786581137309293e-05,
+      "loss": 0.844,
+      "step": 2333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.831501007080078,
+      "learning_rate": 1.7860610263522884e-05,
+      "loss": 0.9115,
+      "step": 2334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.639031410217285,
+      "learning_rate": 1.7855409153952846e-05,
+      "loss": 1.0755,
+      "step": 2335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.04238510131836,
+      "learning_rate": 1.78502080443828e-05,
+      "loss": 1.1489,
+      "step": 2336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.406225204467773,
+      "learning_rate": 1.784500693481276e-05,
+      "loss": 0.772,
+      "step": 2337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.882306098937988,
+      "learning_rate": 1.7839805825242718e-05,
+      "loss": 0.6553,
+      "step": 2338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.954991340637207,
+      "learning_rate": 1.7834604715672677e-05,
+      "loss": 0.9141,
+      "step": 2339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.760220527648926,
+      "learning_rate": 1.7829403606102635e-05,
+      "loss": 0.7574,
+      "step": 2340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.435726165771484,
+      "learning_rate": 1.7824202496532594e-05,
+      "loss": 1.0013,
+      "step": 2341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.783705711364746,
+      "learning_rate": 1.7819001386962553e-05,
+      "loss": 0.6511,
+      "step": 2342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.521049499511719,
+      "learning_rate": 1.781380027739251e-05,
+      "loss": 0.9809,
+      "step": 2343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.812851905822754,
+      "learning_rate": 1.7808599167822467e-05,
+      "loss": 0.7478,
+      "step": 2344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.54907512664795,
+      "learning_rate": 1.780339805825243e-05,
+      "loss": 0.8435,
+      "step": 2345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.744988441467285,
+      "learning_rate": 1.7798196948682384e-05,
+      "loss": 0.7734,
+      "step": 2346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.330835342407227,
+      "learning_rate": 1.7792995839112346e-05,
+      "loss": 0.8638,
+      "step": 2347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.717142105102539,
+      "learning_rate": 1.77877947295423e-05,
+      "loss": 0.9761,
+      "step": 2348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.584728717803955,
+      "learning_rate": 1.7782593619972263e-05,
+      "loss": 0.7205,
+      "step": 2349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.786359786987305,
+      "learning_rate": 1.7777392510402218e-05,
+      "loss": 0.7667,
+      "step": 2350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.683026313781738,
+      "learning_rate": 1.7772191400832177e-05,
+      "loss": 1.0213,
+      "step": 2351
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.391626358032227,
+      "learning_rate": 1.7766990291262135e-05,
+      "loss": 0.7852,
+      "step": 2352
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.887482643127441,
+      "learning_rate": 1.7761789181692094e-05,
+      "loss": 0.6744,
+      "step": 2353
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.953023910522461,
+      "learning_rate": 1.7756588072122053e-05,
+      "loss": 0.7991,
+      "step": 2354
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.051332473754883,
+      "learning_rate": 1.775138696255201e-05,
+      "loss": 0.8391,
+      "step": 2355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.919193267822266,
+      "learning_rate": 1.774618585298197e-05,
+      "loss": 0.7841,
+      "step": 2356
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.971954822540283,
+      "learning_rate": 1.774098474341193e-05,
+      "loss": 0.8846,
+      "step": 2357
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.400032997131348,
+      "learning_rate": 1.7735783633841887e-05,
+      "loss": 0.5575,
+      "step": 2358
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.656684875488281,
+      "learning_rate": 1.7730582524271846e-05,
+      "loss": 0.6832,
+      "step": 2359
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.167808532714844,
+      "learning_rate": 1.77253814147018e-05,
+      "loss": 0.8444,
+      "step": 2360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.860931396484375,
+      "learning_rate": 1.7720180305131763e-05,
+      "loss": 0.9354,
+      "step": 2361
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.741632461547852,
+      "learning_rate": 1.7714979195561718e-05,
+      "loss": 1.1878,
+      "step": 2362
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.147112846374512,
+      "learning_rate": 1.770977808599168e-05,
+      "loss": 0.9029,
+      "step": 2363
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.77216911315918,
+      "learning_rate": 1.7704576976421635e-05,
+      "loss": 0.9599,
+      "step": 2364
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.650917053222656,
+      "learning_rate": 1.7699375866851597e-05,
+      "loss": 0.9045,
+      "step": 2365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.98697566986084,
+      "learning_rate": 1.7694174757281553e-05,
+      "loss": 1.1834,
+      "step": 2366
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.696317672729492,
+      "learning_rate": 1.768897364771151e-05,
+      "loss": 0.8578,
+      "step": 2367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.21140193939209,
+      "learning_rate": 1.768377253814147e-05,
+      "loss": 0.8669,
+      "step": 2368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.710769653320312,
+      "learning_rate": 1.767857142857143e-05,
+      "loss": 1.0415,
+      "step": 2369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.567777633666992,
+      "learning_rate": 1.7673370319001387e-05,
+      "loss": 0.8103,
+      "step": 2370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.072882652282715,
+      "learning_rate": 1.7668169209431346e-05,
+      "loss": 0.9752,
+      "step": 2371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.471040725708008,
+      "learning_rate": 1.7662968099861304e-05,
+      "loss": 0.7433,
+      "step": 2372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.350770950317383,
+      "learning_rate": 1.7657766990291263e-05,
+      "loss": 0.9172,
+      "step": 2373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.344016075134277,
+      "learning_rate": 1.7652565880721218e-05,
+      "loss": 0.9608,
+      "step": 2374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.668880462646484,
+      "learning_rate": 1.764736477115118e-05,
+      "loss": 0.8833,
+      "step": 2375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.793963432312012,
+      "learning_rate": 1.7642163661581135e-05,
+      "loss": 0.933,
+      "step": 2376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.04218578338623,
+      "learning_rate": 1.7636962552011097e-05,
+      "loss": 0.7779,
+      "step": 2377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.825773239135742,
+      "learning_rate": 1.7631761442441053e-05,
+      "loss": 0.7552,
+      "step": 2378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.30412483215332,
+      "learning_rate": 1.7626560332871015e-05,
+      "loss": 0.7473,
+      "step": 2379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.877568244934082,
+      "learning_rate": 1.762135922330097e-05,
+      "loss": 0.9867,
+      "step": 2380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.957174777984619,
+      "learning_rate": 1.7616158113730932e-05,
+      "loss": 0.778,
+      "step": 2381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.511476516723633,
+      "learning_rate": 1.7610957004160887e-05,
+      "loss": 1.1605,
+      "step": 2382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.26945972442627,
+      "learning_rate": 1.7605755894590846e-05,
+      "loss": 0.8638,
+      "step": 2383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.366921424865723,
+      "learning_rate": 1.7600554785020804e-05,
+      "loss": 0.8811,
+      "step": 2384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.03262996673584,
+      "learning_rate": 1.7595353675450763e-05,
+      "loss": 0.785,
+      "step": 2385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.042497634887695,
+      "learning_rate": 1.759015256588072e-05,
+      "loss": 0.924,
+      "step": 2386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.961371421813965,
+      "learning_rate": 1.758495145631068e-05,
+      "loss": 1.0358,
+      "step": 2387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.913840293884277,
+      "learning_rate": 1.757975034674064e-05,
+      "loss": 0.8917,
+      "step": 2388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.438494682312012,
+      "learning_rate": 1.7574549237170597e-05,
+      "loss": 0.9991,
+      "step": 2389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.110587120056152,
+      "learning_rate": 1.7569348127600553e-05,
+      "loss": 0.7873,
+      "step": 2390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.91530990600586,
+      "learning_rate": 1.7564147018030515e-05,
+      "loss": 0.9665,
+      "step": 2391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.137900352478027,
+      "learning_rate": 1.755894590846047e-05,
+      "loss": 0.8899,
+      "step": 2392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.107134819030762,
+      "learning_rate": 1.7553744798890432e-05,
+      "loss": 0.7665,
+      "step": 2393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.467238426208496,
+      "learning_rate": 1.7548543689320387e-05,
+      "loss": 0.9722,
+      "step": 2394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.43931770324707,
+      "learning_rate": 1.754334257975035e-05,
+      "loss": 0.7236,
+      "step": 2395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.178537368774414,
+      "learning_rate": 1.7538141470180304e-05,
+      "loss": 0.5804,
+      "step": 2396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.658415794372559,
+      "learning_rate": 1.7532940360610263e-05,
+      "loss": 0.7485,
+      "step": 2397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.9293670654296875,
+      "learning_rate": 1.752773925104022e-05,
+      "loss": 0.8065,
+      "step": 2398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.793274879455566,
+      "learning_rate": 1.752253814147018e-05,
+      "loss": 1.0896,
+      "step": 2399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.740966796875,
+      "learning_rate": 1.751733703190014e-05,
+      "loss": 0.763,
+      "step": 2400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.857243537902832,
+      "learning_rate": 1.7512135922330097e-05,
+      "loss": 0.8411,
+      "step": 2401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.500646591186523,
+      "learning_rate": 1.7506934812760056e-05,
+      "loss": 0.7909,
+      "step": 2402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.384224891662598,
+      "learning_rate": 1.7501733703190015e-05,
+      "loss": 1.2372,
+      "step": 2403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.304952621459961,
+      "learning_rate": 1.7496532593619973e-05,
+      "loss": 0.7015,
+      "step": 2404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.753312110900879,
+      "learning_rate": 1.7491331484049932e-05,
+      "loss": 0.8772,
+      "step": 2405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.133476257324219,
+      "learning_rate": 1.7486130374479887e-05,
+      "loss": 0.7753,
+      "step": 2406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.844011306762695,
+      "learning_rate": 1.748092926490985e-05,
+      "loss": 0.6982,
+      "step": 2407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.317828178405762,
+      "learning_rate": 1.7475728155339804e-05,
+      "loss": 0.7575,
+      "step": 2408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.201967239379883,
+      "learning_rate": 1.7470527045769766e-05,
+      "loss": 0.9741,
+      "step": 2409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.227357864379883,
+      "learning_rate": 1.746532593619972e-05,
+      "loss": 1.0989,
+      "step": 2410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.313446044921875,
+      "learning_rate": 1.7460124826629683e-05,
+      "loss": 0.7395,
+      "step": 2411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.602742671966553,
+      "learning_rate": 1.745492371705964e-05,
+      "loss": 0.6014,
+      "step": 2412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.285748481750488,
+      "learning_rate": 1.7449722607489597e-05,
+      "loss": 0.9785,
+      "step": 2413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.758256912231445,
+      "learning_rate": 1.7444521497919556e-05,
+      "loss": 0.803,
+      "step": 2414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.255910873413086,
+      "learning_rate": 1.7439320388349514e-05,
+      "loss": 1.0908,
+      "step": 2415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.398155212402344,
+      "learning_rate": 1.7434119278779473e-05,
+      "loss": 0.8819,
+      "step": 2416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.095613479614258,
+      "learning_rate": 1.7428918169209432e-05,
+      "loss": 0.8399,
+      "step": 2417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.655876159667969,
+      "learning_rate": 1.742371705963939e-05,
+      "loss": 0.9854,
+      "step": 2418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.161911964416504,
+      "learning_rate": 1.741851595006935e-05,
+      "loss": 0.8689,
+      "step": 2419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.828033924102783,
+      "learning_rate": 1.7413314840499308e-05,
+      "loss": 0.8696,
+      "step": 2420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.90929889678955,
+      "learning_rate": 1.7408113730929266e-05,
+      "loss": 0.8821,
+      "step": 2421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.943587303161621,
+      "learning_rate": 1.740291262135922e-05,
+      "loss": 0.9394,
+      "step": 2422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.130565643310547,
+      "learning_rate": 1.7397711511789183e-05,
+      "loss": 0.7989,
+      "step": 2423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.091739654541016,
+      "learning_rate": 1.739251040221914e-05,
+      "loss": 0.937,
+      "step": 2424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.087077140808105,
+      "learning_rate": 1.73873092926491e-05,
+      "loss": 0.9408,
+      "step": 2425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.73260498046875,
+      "learning_rate": 1.7382108183079056e-05,
+      "loss": 0.9131,
+      "step": 2426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.79139232635498,
+      "learning_rate": 1.7376907073509018e-05,
+      "loss": 0.8761,
+      "step": 2427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.949048042297363,
+      "learning_rate": 1.7371705963938973e-05,
+      "loss": 1.1482,
+      "step": 2428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.010560989379883,
+      "learning_rate": 1.736650485436893e-05,
+      "loss": 0.9007,
+      "step": 2429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.050619125366211,
+      "learning_rate": 1.736130374479889e-05,
+      "loss": 0.9665,
+      "step": 2430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.110004425048828,
+      "learning_rate": 1.735610263522885e-05,
+      "loss": 0.8595,
+      "step": 2431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.303783416748047,
+      "learning_rate": 1.7350901525658808e-05,
+      "loss": 0.9507,
+      "step": 2432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.01194953918457,
+      "learning_rate": 1.7345700416088766e-05,
+      "loss": 0.6592,
+      "step": 2433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.967955112457275,
+      "learning_rate": 1.7340499306518725e-05,
+      "loss": 0.5877,
+      "step": 2434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.43539047241211,
+      "learning_rate": 1.7335298196948683e-05,
+      "loss": 0.7377,
+      "step": 2435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.450138092041016,
+      "learning_rate": 1.733009708737864e-05,
+      "loss": 1.0303,
+      "step": 2436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.505535125732422,
+      "learning_rate": 1.73248959778086e-05,
+      "loss": 0.7455,
+      "step": 2437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.08008861541748,
+      "learning_rate": 1.7319694868238556e-05,
+      "loss": 0.8179,
+      "step": 2438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.8351411819458,
+      "learning_rate": 1.7314493758668518e-05,
+      "loss": 0.8035,
+      "step": 2439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.4881591796875,
+      "learning_rate": 1.7309292649098473e-05,
+      "loss": 0.8068,
+      "step": 2440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.825074195861816,
+      "learning_rate": 1.7304091539528435e-05,
+      "loss": 0.8463,
+      "step": 2441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.412644386291504,
+      "learning_rate": 1.729889042995839e-05,
+      "loss": 1.2973,
+      "step": 2442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.676898002624512,
+      "learning_rate": 1.7293689320388352e-05,
+      "loss": 0.7902,
+      "step": 2443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.764787673950195,
+      "learning_rate": 1.7288488210818307e-05,
+      "loss": 0.7229,
+      "step": 2444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.119978427886963,
+      "learning_rate": 1.7283287101248266e-05,
+      "loss": 0.5837,
+      "step": 2445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.844925880432129,
+      "learning_rate": 1.7278085991678225e-05,
+      "loss": 0.9923,
+      "step": 2446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.811936378479004,
+      "learning_rate": 1.7272884882108183e-05,
+      "loss": 0.7396,
+      "step": 2447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.47705078125,
+      "learning_rate": 1.7267683772538142e-05,
+      "loss": 0.7603,
+      "step": 2448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.908052444458008,
+      "learning_rate": 1.72624826629681e-05,
+      "loss": 0.7184,
+      "step": 2449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.089224815368652,
+      "learning_rate": 1.725728155339806e-05,
+      "loss": 0.7389,
+      "step": 2450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.044531345367432,
+      "learning_rate": 1.7252080443828018e-05,
+      "loss": 0.7172,
+      "step": 2451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.51612663269043,
+      "learning_rate": 1.7246879334257973e-05,
+      "loss": 0.6591,
+      "step": 2452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.109050750732422,
+      "learning_rate": 1.7241678224687935e-05,
+      "loss": 0.9086,
+      "step": 2453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.914301872253418,
+      "learning_rate": 1.723647711511789e-05,
+      "loss": 0.9479,
+      "step": 2454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.524891376495361,
+      "learning_rate": 1.7231276005547852e-05,
+      "loss": 0.6447,
+      "step": 2455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.461121559143066,
+      "learning_rate": 1.7226074895977807e-05,
+      "loss": 0.7463,
+      "step": 2456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.200360298156738,
+      "learning_rate": 1.722087378640777e-05,
+      "loss": 0.8762,
+      "step": 2457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.930845260620117,
+      "learning_rate": 1.7215672676837725e-05,
+      "loss": 0.9794,
+      "step": 2458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.948577880859375,
+      "learning_rate": 1.7210471567267683e-05,
+      "loss": 0.8372,
+      "step": 2459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.146499633789062,
+      "learning_rate": 1.7205270457697642e-05,
+      "loss": 0.871,
+      "step": 2460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.771735191345215,
+      "learning_rate": 1.72000693481276e-05,
+      "loss": 1.0819,
+      "step": 2461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.840658187866211,
+      "learning_rate": 1.719486823855756e-05,
+      "loss": 0.9004,
+      "step": 2462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.275520324707031,
+      "learning_rate": 1.7189667128987518e-05,
+      "loss": 0.9342,
+      "step": 2463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.99124813079834,
+      "learning_rate": 1.7184466019417476e-05,
+      "loss": 0.8395,
+      "step": 2464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.867913246154785,
+      "learning_rate": 1.7179264909847435e-05,
+      "loss": 0.7121,
+      "step": 2465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.624526023864746,
+      "learning_rate": 1.7174063800277394e-05,
+      "loss": 0.9613,
+      "step": 2466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.419919013977051,
+      "learning_rate": 1.7168862690707352e-05,
+      "loss": 0.5952,
+      "step": 2467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.31462574005127,
+      "learning_rate": 1.7163661581137307e-05,
+      "loss": 0.77,
+      "step": 2468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.752237319946289,
+      "learning_rate": 1.715846047156727e-05,
+      "loss": 0.7791,
+      "step": 2469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.392569065093994,
+      "learning_rate": 1.7153259361997225e-05,
+      "loss": 0.5698,
+      "step": 2470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.568845748901367,
+      "learning_rate": 1.7148058252427187e-05,
+      "loss": 0.8982,
+      "step": 2471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.299123764038086,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.5982,
+      "step": 2472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.020793914794922,
+      "learning_rate": 1.7137656033287104e-05,
+      "loss": 0.6194,
+      "step": 2473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.09770393371582,
+      "learning_rate": 1.713245492371706e-05,
+      "loss": 0.7336,
+      "step": 2474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.859700202941895,
+      "learning_rate": 1.7127253814147018e-05,
+      "loss": 0.6841,
+      "step": 2475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.347691535949707,
+      "learning_rate": 1.7122052704576976e-05,
+      "loss": 0.848,
+      "step": 2476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.216526985168457,
+      "learning_rate": 1.7116851595006935e-05,
+      "loss": 1.0041,
+      "step": 2477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.711380958557129,
+      "learning_rate": 1.7111650485436894e-05,
+      "loss": 0.8952,
+      "step": 2478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.38117790222168,
+      "learning_rate": 1.7106449375866852e-05,
+      "loss": 0.9286,
+      "step": 2479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.421720504760742,
+      "learning_rate": 1.710124826629681e-05,
+      "loss": 0.7293,
+      "step": 2480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.691868305206299,
+      "learning_rate": 1.709604715672677e-05,
+      "loss": 0.6243,
+      "step": 2481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.48595905303955,
+      "learning_rate": 1.7090846047156728e-05,
+      "loss": 0.9194,
+      "step": 2482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.346985816955566,
+      "learning_rate": 1.7085644937586687e-05,
+      "loss": 0.7475,
+      "step": 2483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.760885238647461,
+      "learning_rate": 1.7080443828016642e-05,
+      "loss": 1.0393,
+      "step": 2484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.74984359741211,
+      "learning_rate": 1.7075242718446604e-05,
+      "loss": 0.8818,
+      "step": 2485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.497040748596191,
+      "learning_rate": 1.707004160887656e-05,
+      "loss": 0.6406,
+      "step": 2486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.185701370239258,
+      "learning_rate": 1.706484049930652e-05,
+      "loss": 0.6761,
+      "step": 2487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.780242919921875,
+      "learning_rate": 1.7059639389736476e-05,
+      "loss": 0.7758,
+      "step": 2488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.944270133972168,
+      "learning_rate": 1.7054438280166438e-05,
+      "loss": 0.9112,
+      "step": 2489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.39229965209961,
+      "learning_rate": 1.7049237170596393e-05,
+      "loss": 1.0375,
+      "step": 2490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.889898300170898,
+      "learning_rate": 1.7044036061026352e-05,
+      "loss": 0.7026,
+      "step": 2491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.671615600585938,
+      "learning_rate": 1.703883495145631e-05,
+      "loss": 0.8166,
+      "step": 2492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.967292785644531,
+      "learning_rate": 1.703363384188627e-05,
+      "loss": 1.0264,
+      "step": 2493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.948934078216553,
+      "learning_rate": 1.7028432732316228e-05,
+      "loss": 0.6997,
+      "step": 2494
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.33706283569336,
+      "learning_rate": 1.7023231622746187e-05,
+      "loss": 1.0927,
+      "step": 2495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.830198287963867,
+      "learning_rate": 1.7018030513176145e-05,
+      "loss": 0.9172,
+      "step": 2496
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.292044639587402,
+      "learning_rate": 1.7012829403606104e-05,
+      "loss": 0.9183,
+      "step": 2497
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.285040855407715,
+      "learning_rate": 1.700762829403606e-05,
+      "loss": 0.9779,
+      "step": 2498
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.579421997070312,
+      "learning_rate": 1.700242718446602e-05,
+      "loss": 0.661,
+      "step": 2499
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.195077896118164,
+      "learning_rate": 1.6997226074895976e-05,
+      "loss": 1.1502,
+      "step": 2500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.154961585998535,
+      "learning_rate": 1.6992024965325938e-05,
+      "loss": 1.1232,
+      "step": 2501
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.302929878234863,
+      "learning_rate": 1.6986823855755893e-05,
+      "loss": 0.8592,
+      "step": 2502
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.999261856079102,
+      "learning_rate": 1.6981622746185855e-05,
+      "loss": 1.1477,
+      "step": 2503
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.386415481567383,
+      "learning_rate": 1.697642163661581e-05,
+      "loss": 0.9504,
+      "step": 2504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.410075187683105,
+      "learning_rate": 1.6971220527045773e-05,
+      "loss": 0.8306,
+      "step": 2505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.828786849975586,
+      "learning_rate": 1.6966019417475728e-05,
+      "loss": 0.8891,
+      "step": 2506
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.442381858825684,
+      "learning_rate": 1.6960818307905686e-05,
+      "loss": 0.7647,
+      "step": 2507
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.126102447509766,
+      "learning_rate": 1.6955617198335645e-05,
+      "loss": 0.6803,
+      "step": 2508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.7232084274292,
+      "learning_rate": 1.6950416088765604e-05,
+      "loss": 1.245,
+      "step": 2509
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.542457580566406,
+      "learning_rate": 1.6945214979195562e-05,
+      "loss": 0.6954,
+      "step": 2510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.86044454574585,
+      "learning_rate": 1.694001386962552e-05,
+      "loss": 0.741,
+      "step": 2511
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.412029266357422,
+      "learning_rate": 1.693481276005548e-05,
+      "loss": 0.7828,
+      "step": 2512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.205294609069824,
+      "learning_rate": 1.6929611650485438e-05,
+      "loss": 0.6524,
+      "step": 2513
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.456207275390625,
+      "learning_rate": 1.6924410540915393e-05,
+      "loss": 0.9303,
+      "step": 2514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.1021089553833,
+      "learning_rate": 1.6919209431345355e-05,
+      "loss": 0.7705,
+      "step": 2515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.152643203735352,
+      "learning_rate": 1.691400832177531e-05,
+      "loss": 0.7265,
+      "step": 2516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.116086959838867,
+      "learning_rate": 1.6908807212205273e-05,
+      "loss": 0.9527,
+      "step": 2517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.554437637329102,
+      "learning_rate": 1.6903606102635228e-05,
+      "loss": 0.8834,
+      "step": 2518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.438138961791992,
+      "learning_rate": 1.689840499306519e-05,
+      "loss": 0.9939,
+      "step": 2519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.829917907714844,
+      "learning_rate": 1.6893203883495145e-05,
+      "loss": 0.9621,
+      "step": 2520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.805362701416016,
+      "learning_rate": 1.6888002773925104e-05,
+      "loss": 0.6978,
+      "step": 2521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.090920448303223,
+      "learning_rate": 1.6882801664355062e-05,
+      "loss": 1.0062,
+      "step": 2522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.062956809997559,
+      "learning_rate": 1.687760055478502e-05,
+      "loss": 0.9727,
+      "step": 2523
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.351447105407715,
+      "learning_rate": 1.687239944521498e-05,
+      "loss": 0.6701,
+      "step": 2524
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.651618003845215,
+      "learning_rate": 1.6867198335644938e-05,
+      "loss": 0.8567,
+      "step": 2525
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.612993240356445,
+      "learning_rate": 1.6861997226074897e-05,
+      "loss": 0.8122,
+      "step": 2526
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.483071327209473,
+      "learning_rate": 1.6856796116504855e-05,
+      "loss": 0.7624,
+      "step": 2527
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.562975883483887,
+      "learning_rate": 1.6851595006934814e-05,
+      "loss": 1.0593,
+      "step": 2528
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.557186126708984,
+      "learning_rate": 1.6846393897364773e-05,
+      "loss": 0.9346,
+      "step": 2529
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.598958969116211,
+      "learning_rate": 1.6841192787794728e-05,
+      "loss": 0.9642,
+      "step": 2530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.19286823272705,
+      "learning_rate": 1.683599167822469e-05,
+      "loss": 0.8314,
+      "step": 2531
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.810806274414062,
+      "learning_rate": 1.6830790568654645e-05,
+      "loss": 0.8908,
+      "step": 2532
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.63270378112793,
+      "learning_rate": 1.6825589459084607e-05,
+      "loss": 0.8329,
+      "step": 2533
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.54032564163208,
+      "learning_rate": 1.6820388349514562e-05,
+      "loss": 0.6287,
+      "step": 2534
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.717103004455566,
+      "learning_rate": 1.6815187239944524e-05,
+      "loss": 1.0587,
+      "step": 2535
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.456790924072266,
+      "learning_rate": 1.680998613037448e-05,
+      "loss": 0.8083,
+      "step": 2536
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.115432739257812,
+      "learning_rate": 1.6804785020804438e-05,
+      "loss": 0.6252,
+      "step": 2537
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.244967460632324,
+      "learning_rate": 1.6799583911234397e-05,
+      "loss": 0.9389,
+      "step": 2538
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.930976867675781,
+      "learning_rate": 1.6794382801664355e-05,
+      "loss": 0.7618,
+      "step": 2539
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.038630485534668,
+      "learning_rate": 1.6789181692094314e-05,
+      "loss": 0.7201,
+      "step": 2540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.477506637573242,
+      "learning_rate": 1.6783980582524273e-05,
+      "loss": 0.8921,
+      "step": 2541
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.920867443084717,
+      "learning_rate": 1.677877947295423e-05,
+      "loss": 0.7681,
+      "step": 2542
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.304810523986816,
+      "learning_rate": 1.677357836338419e-05,
+      "loss": 0.7343,
+      "step": 2543
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.622842788696289,
+      "learning_rate": 1.6768377253814145e-05,
+      "loss": 0.8758,
+      "step": 2544
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.683833122253418,
+      "learning_rate": 1.6763176144244107e-05,
+      "loss": 0.9119,
+      "step": 2545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.896069049835205,
+      "learning_rate": 1.6757975034674062e-05,
+      "loss": 0.7014,
+      "step": 2546
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.905839920043945,
+      "learning_rate": 1.6752773925104024e-05,
+      "loss": 0.7884,
+      "step": 2547
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.601353645324707,
+      "learning_rate": 1.674757281553398e-05,
+      "loss": 0.79,
+      "step": 2548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.496175765991211,
+      "learning_rate": 1.674237170596394e-05,
+      "loss": 1.1306,
+      "step": 2549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.17431354522705,
+      "learning_rate": 1.6737170596393897e-05,
+      "loss": 0.6654,
+      "step": 2550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.971198081970215,
+      "learning_rate": 1.673196948682386e-05,
+      "loss": 0.8139,
+      "step": 2551
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.989645957946777,
+      "learning_rate": 1.6726768377253814e-05,
+      "loss": 0.9589,
+      "step": 2552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.248270034790039,
+      "learning_rate": 1.6721567267683773e-05,
+      "loss": 0.8393,
+      "step": 2553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.279542922973633,
+      "learning_rate": 1.671636615811373e-05,
+      "loss": 1.1145,
+      "step": 2554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.996354103088379,
+      "learning_rate": 1.671116504854369e-05,
+      "loss": 0.7185,
+      "step": 2555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.842789649963379,
+      "learning_rate": 1.670596393897365e-05,
+      "loss": 0.8767,
+      "step": 2556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.75771427154541,
+      "learning_rate": 1.6700762829403607e-05,
+      "loss": 1.0715,
+      "step": 2557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.800548553466797,
+      "learning_rate": 1.6695561719833566e-05,
+      "loss": 1.0494,
+      "step": 2558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.054328918457031,
+      "learning_rate": 1.6690360610263524e-05,
+      "loss": 0.7683,
+      "step": 2559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.571981906890869,
+      "learning_rate": 1.668515950069348e-05,
+      "loss": 0.6542,
+      "step": 2560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.572015762329102,
+      "learning_rate": 1.667995839112344e-05,
+      "loss": 0.8284,
+      "step": 2561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.833404541015625,
+      "learning_rate": 1.6674757281553397e-05,
+      "loss": 0.842,
+      "step": 2562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.623393058776855,
+      "learning_rate": 1.666955617198336e-05,
+      "loss": 0.7869,
+      "step": 2563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.609304428100586,
+      "learning_rate": 1.6664355062413314e-05,
+      "loss": 0.8741,
+      "step": 2564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.867240905761719,
+      "learning_rate": 1.6659153952843276e-05,
+      "loss": 1.0073,
+      "step": 2565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.124246597290039,
+      "learning_rate": 1.665395284327323e-05,
+      "loss": 0.7034,
+      "step": 2566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.718267440795898,
+      "learning_rate": 1.664875173370319e-05,
+      "loss": 0.8898,
+      "step": 2567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.946948051452637,
+      "learning_rate": 1.6643550624133148e-05,
+      "loss": 0.8934,
+      "step": 2568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.544266700744629,
+      "learning_rate": 1.6638349514563107e-05,
+      "loss": 0.7021,
+      "step": 2569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.799774169921875,
+      "learning_rate": 1.6633148404993066e-05,
+      "loss": 0.7072,
+      "step": 2570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.382257461547852,
+      "learning_rate": 1.6627947295423024e-05,
+      "loss": 0.934,
+      "step": 2571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.109541893005371,
+      "learning_rate": 1.6622746185852983e-05,
+      "loss": 0.6898,
+      "step": 2572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.166972160339355,
+      "learning_rate": 1.661754507628294e-05,
+      "loss": 0.8258,
+      "step": 2573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.468461990356445,
+      "learning_rate": 1.66123439667129e-05,
+      "loss": 0.4508,
+      "step": 2574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.794669151306152,
+      "learning_rate": 1.660714285714286e-05,
+      "loss": 0.9579,
+      "step": 2575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.309988021850586,
+      "learning_rate": 1.6601941747572814e-05,
+      "loss": 0.9096,
+      "step": 2576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.887486457824707,
+      "learning_rate": 1.6596740638002776e-05,
+      "loss": 0.6987,
+      "step": 2577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.605618476867676,
+      "learning_rate": 1.659153952843273e-05,
+      "loss": 0.8531,
+      "step": 2578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.840025901794434,
+      "learning_rate": 1.6586338418862693e-05,
+      "loss": 0.8397,
+      "step": 2579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.52086353302002,
+      "learning_rate": 1.6581137309292648e-05,
+      "loss": 0.8333,
+      "step": 2580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.15660572052002,
+      "learning_rate": 1.657593619972261e-05,
+      "loss": 1.1245,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.230676651000977,
+      "learning_rate": 1.6570735090152565e-05,
+      "loss": 0.7665,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.708389282226562,
+      "learning_rate": 1.6565533980582524e-05,
+      "loss": 0.8237,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.242239952087402,
+      "learning_rate": 1.6560332871012483e-05,
+      "loss": 1.0153,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.276283264160156,
+      "learning_rate": 1.655513176144244e-05,
+      "loss": 0.8515,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.708937644958496,
+      "learning_rate": 1.65499306518724e-05,
+      "loss": 0.8673,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.583049774169922,
+      "learning_rate": 1.654472954230236e-05,
+      "loss": 0.8118,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.216681480407715,
+      "learning_rate": 1.6539528432732317e-05,
+      "loss": 0.8672,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.120989799499512,
+      "learning_rate": 1.6534327323162276e-05,
+      "loss": 0.9337,
+      "step": 2589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.040614604949951,
+      "learning_rate": 1.6529126213592234e-05,
+      "loss": 0.6085,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.534711837768555,
+      "learning_rate": 1.6523925104022193e-05,
+      "loss": 1.0516,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.224756240844727,
+      "learning_rate": 1.6518723994452148e-05,
+      "loss": 0.9007,
+      "step": 2592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.222489356994629,
+      "learning_rate": 1.651352288488211e-05,
+      "loss": 0.8684,
+      "step": 2593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.84293270111084,
+      "learning_rate": 1.6508321775312065e-05,
+      "loss": 0.817,
+      "step": 2594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.663381576538086,
+      "learning_rate": 1.6503120665742027e-05,
+      "loss": 0.8726,
+      "step": 2595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.93218994140625,
+      "learning_rate": 1.6497919556171983e-05,
+      "loss": 0.7987,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.264707565307617,
+      "learning_rate": 1.6492718446601945e-05,
+      "loss": 0.7798,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.941629409790039,
+      "learning_rate": 1.64875173370319e-05,
+      "loss": 0.8919,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.80352783203125,
+      "learning_rate": 1.648231622746186e-05,
+      "loss": 0.7566,
+      "step": 2599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.934093475341797,
+      "learning_rate": 1.6477115117891817e-05,
+      "loss": 0.5759,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.031767845153809,
+      "learning_rate": 1.6471914008321776e-05,
+      "loss": 0.995,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.236480712890625,
+      "learning_rate": 1.6466712898751734e-05,
+      "loss": 0.6286,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.051833152770996,
+      "learning_rate": 1.6461511789181693e-05,
+      "loss": 1.036,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.354971885681152,
+      "learning_rate": 1.645631067961165e-05,
+      "loss": 0.8339,
+      "step": 2604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.750553607940674,
+      "learning_rate": 1.645110957004161e-05,
+      "loss": 0.8342,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.685649871826172,
+      "learning_rate": 1.6445908460471565e-05,
+      "loss": 0.9277,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.626613616943359,
+      "learning_rate": 1.6440707350901527e-05,
+      "loss": 0.8496,
+      "step": 2607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.471017837524414,
+      "learning_rate": 1.6435506241331483e-05,
+      "loss": 0.7023,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.50967788696289,
+      "learning_rate": 1.6430305131761445e-05,
+      "loss": 0.8895,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.114291191101074,
+      "learning_rate": 1.64251040221914e-05,
+      "loss": 0.84,
+      "step": 2610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.273847579956055,
+      "learning_rate": 1.6419902912621362e-05,
+      "loss": 0.9929,
+      "step": 2611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.022987365722656,
+      "learning_rate": 1.6414701803051317e-05,
+      "loss": 0.7485,
+      "step": 2612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.339129447937012,
+      "learning_rate": 1.640950069348128e-05,
+      "loss": 1.0058,
+      "step": 2613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.417801856994629,
+      "learning_rate": 1.6404299583911234e-05,
+      "loss": 0.7583,
+      "step": 2614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.45248794555664,
+      "learning_rate": 1.6399098474341193e-05,
+      "loss": 0.7011,
+      "step": 2615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.127172470092773,
+      "learning_rate": 1.639389736477115e-05,
+      "loss": 0.8932,
+      "step": 2616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.642643928527832,
+      "learning_rate": 1.638869625520111e-05,
+      "loss": 0.7898,
+      "step": 2617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.13563060760498,
+      "learning_rate": 1.638349514563107e-05,
+      "loss": 0.9897,
+      "step": 2618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.836441516876221,
+      "learning_rate": 1.6378294036061027e-05,
+      "loss": 0.6971,
+      "step": 2619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.396356582641602,
+      "learning_rate": 1.6373092926490986e-05,
+      "loss": 0.5992,
+      "step": 2620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.763139724731445,
+      "learning_rate": 1.6367891816920945e-05,
+      "loss": 0.7869,
+      "step": 2621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.99718189239502,
+      "learning_rate": 1.63626907073509e-05,
+      "loss": 0.8111,
+      "step": 2622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.577115058898926,
+      "learning_rate": 1.6357489597780862e-05,
+      "loss": 0.8791,
+      "step": 2623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.145164489746094,
+      "learning_rate": 1.6352288488210817e-05,
+      "loss": 1.108,
+      "step": 2624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.246150970458984,
+      "learning_rate": 1.634708737864078e-05,
+      "loss": 0.8305,
+      "step": 2625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.842000961303711,
+      "learning_rate": 1.6341886269070734e-05,
+      "loss": 0.7215,
+      "step": 2626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.877351760864258,
+      "learning_rate": 1.6336685159500696e-05,
+      "loss": 0.9306,
+      "step": 2627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.722697257995605,
+      "learning_rate": 1.633148404993065e-05,
+      "loss": 0.7663,
+      "step": 2628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.962141036987305,
+      "learning_rate": 1.632628294036061e-05,
+      "loss": 0.7599,
+      "step": 2629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.453579902648926,
+      "learning_rate": 1.632108183079057e-05,
+      "loss": 0.9452,
+      "step": 2630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.481234550476074,
+      "learning_rate": 1.6315880721220527e-05,
+      "loss": 0.8767,
+      "step": 2631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.800584316253662,
+      "learning_rate": 1.6310679611650486e-05,
+      "loss": 0.871,
+      "step": 2632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.703936576843262,
+      "learning_rate": 1.6305478502080445e-05,
+      "loss": 0.8829,
+      "step": 2633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.616242408752441,
+      "learning_rate": 1.6300277392510403e-05,
+      "loss": 0.887,
+      "step": 2634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.493009567260742,
+      "learning_rate": 1.6295076282940362e-05,
+      "loss": 0.9376,
+      "step": 2635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.001906394958496,
+      "learning_rate": 1.628987517337032e-05,
+      "loss": 0.9616,
+      "step": 2636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.473123550415039,
+      "learning_rate": 1.628467406380028e-05,
+      "loss": 0.8614,
+      "step": 2637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.290766716003418,
+      "learning_rate": 1.6279472954230234e-05,
+      "loss": 0.9045,
+      "step": 2638
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.756927490234375,
+      "learning_rate": 1.6274271844660196e-05,
+      "loss": 0.8731,
+      "step": 2639
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.969849586486816,
+      "learning_rate": 1.626907073509015e-05,
+      "loss": 0.5766,
+      "step": 2640
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.447450160980225,
+      "learning_rate": 1.6263869625520113e-05,
+      "loss": 0.7123,
+      "step": 2641
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.92074203491211,
+      "learning_rate": 1.625866851595007e-05,
+      "loss": 0.8665,
+      "step": 2642
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.94494915008545,
+      "learning_rate": 1.625346740638003e-05,
+      "loss": 0.9903,
+      "step": 2643
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.734247207641602,
+      "learning_rate": 1.6248266296809986e-05,
+      "loss": 0.7357,
+      "step": 2644
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.129583358764648,
+      "learning_rate": 1.6243065187239945e-05,
+      "loss": 0.9716,
+      "step": 2645
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.066713333129883,
+      "learning_rate": 1.6237864077669903e-05,
+      "loss": 0.7938,
+      "step": 2646
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.284828186035156,
+      "learning_rate": 1.6232662968099862e-05,
+      "loss": 0.7645,
+      "step": 2647
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.484292030334473,
+      "learning_rate": 1.622746185852982e-05,
+      "loss": 0.9445,
+      "step": 2648
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.093836784362793,
+      "learning_rate": 1.622226074895978e-05,
+      "loss": 0.7491,
+      "step": 2649
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.289881706237793,
+      "learning_rate": 1.6217059639389738e-05,
+      "loss": 0.8581,
+      "step": 2650
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.494281768798828,
+      "learning_rate": 1.6211858529819696e-05,
+      "loss": 0.8538,
+      "step": 2651
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.814667701721191,
+      "learning_rate": 1.6206657420249655e-05,
+      "loss": 0.7379,
+      "step": 2652
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.264556884765625,
+      "learning_rate": 1.6201456310679613e-05,
+      "loss": 0.7088,
+      "step": 2653
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.852272033691406,
+      "learning_rate": 1.619625520110957e-05,
+      "loss": 0.6836,
+      "step": 2654
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.688441276550293,
+      "learning_rate": 1.619105409153953e-05,
+      "loss": 0.8211,
+      "step": 2655
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.646769046783447,
+      "learning_rate": 1.6185852981969486e-05,
+      "loss": 0.8015,
+      "step": 2656
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.994892597198486,
+      "learning_rate": 1.6180651872399448e-05,
+      "loss": 0.6728,
+      "step": 2657
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.00714111328125,
+      "learning_rate": 1.6175450762829403e-05,
+      "loss": 0.8038,
+      "step": 2658
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.38809585571289,
+      "learning_rate": 1.6170249653259365e-05,
+      "loss": 0.9475,
+      "step": 2659
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.100142478942871,
+      "learning_rate": 1.616504854368932e-05,
+      "loss": 0.5887,
+      "step": 2660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.103289604187012,
+      "learning_rate": 1.615984743411928e-05,
+      "loss": 0.8068,
+      "step": 2661
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.872264862060547,
+      "learning_rate": 1.6154646324549238e-05,
+      "loss": 0.9055,
+      "step": 2662
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.645809173583984,
+      "learning_rate": 1.6149445214979196e-05,
+      "loss": 0.9021,
+      "step": 2663
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.80757474899292,
+      "learning_rate": 1.6144244105409155e-05,
+      "loss": 0.5553,
+      "step": 2664
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.544624328613281,
+      "learning_rate": 1.6139042995839113e-05,
+      "loss": 0.9148,
+      "step": 2665
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.874537467956543,
+      "learning_rate": 1.6133841886269072e-05,
+      "loss": 0.8809,
+      "step": 2666
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.302000999450684,
+      "learning_rate": 1.612864077669903e-05,
+      "loss": 0.8719,
+      "step": 2667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.132102966308594,
+      "learning_rate": 1.6123439667128986e-05,
+      "loss": 0.955,
+      "step": 2668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.922053813934326,
+      "learning_rate": 1.6118238557558948e-05,
+      "loss": 0.7156,
+      "step": 2669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.246227264404297,
+      "learning_rate": 1.6113037447988903e-05,
+      "loss": 0.8391,
+      "step": 2670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.06718921661377,
+      "learning_rate": 1.6107836338418865e-05,
+      "loss": 0.8627,
+      "step": 2671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.306187629699707,
+      "learning_rate": 1.610263522884882e-05,
+      "loss": 0.8836,
+      "step": 2672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.726730346679688,
+      "learning_rate": 1.6097434119278782e-05,
+      "loss": 0.9284,
+      "step": 2673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.8587064743042,
+      "learning_rate": 1.6092233009708738e-05,
+      "loss": 0.744,
+      "step": 2674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.939717292785645,
+      "learning_rate": 1.60870319001387e-05,
+      "loss": 0.7142,
+      "step": 2675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.17357349395752,
+      "learning_rate": 1.6081830790568655e-05,
+      "loss": 0.6131,
+      "step": 2676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.26930046081543,
+      "learning_rate": 1.6076629680998613e-05,
+      "loss": 0.9129,
+      "step": 2677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.619955062866211,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 0.6853,
+      "step": 2678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.300370216369629,
+      "learning_rate": 1.606622746185853e-05,
+      "loss": 0.8831,
+      "step": 2679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.018220901489258,
+      "learning_rate": 1.606102635228849e-05,
+      "loss": 0.8153,
+      "step": 2680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.365309715270996,
+      "learning_rate": 1.6055825242718448e-05,
+      "loss": 0.8083,
+      "step": 2681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.034289360046387,
+      "learning_rate": 1.6050624133148406e-05,
+      "loss": 0.6757,
+      "step": 2682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.800156116485596,
+      "learning_rate": 1.6045423023578365e-05,
+      "loss": 0.7803,
+      "step": 2683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.604784965515137,
+      "learning_rate": 1.604022191400832e-05,
+      "loss": 0.9461,
+      "step": 2684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.897541999816895,
+      "learning_rate": 1.6035020804438282e-05,
+      "loss": 0.7582,
+      "step": 2685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.291213035583496,
+      "learning_rate": 1.6029819694868237e-05,
+      "loss": 1.0384,
+      "step": 2686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.733201026916504,
+      "learning_rate": 1.60246185852982e-05,
+      "loss": 0.8935,
+      "step": 2687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.66602611541748,
+      "learning_rate": 1.6019417475728155e-05,
+      "loss": 0.8684,
+      "step": 2688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.07101058959961,
+      "learning_rate": 1.6014216366158117e-05,
+      "loss": 0.6842,
+      "step": 2689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.33614444732666,
+      "learning_rate": 1.6009015256588072e-05,
+      "loss": 1.0049,
+      "step": 2690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.630635261535645,
+      "learning_rate": 1.600381414701803e-05,
+      "loss": 1.0027,
+      "step": 2691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.321612358093262,
+      "learning_rate": 1.599861303744799e-05,
+      "loss": 0.7713,
+      "step": 2692
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.09355640411377,
+      "learning_rate": 1.5993411927877948e-05,
+      "loss": 1.0829,
+      "step": 2693
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.01588249206543,
+      "learning_rate": 1.5988210818307906e-05,
+      "loss": 0.7123,
+      "step": 2694
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.683025360107422,
+      "learning_rate": 1.5983009708737865e-05,
+      "loss": 0.6601,
+      "step": 2695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.229838371276855,
+      "learning_rate": 1.5977808599167824e-05,
+      "loss": 1.0023,
+      "step": 2696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.834228515625,
+      "learning_rate": 1.5972607489597782e-05,
+      "loss": 0.6557,
+      "step": 2697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.497241020202637,
+      "learning_rate": 1.596740638002774e-05,
+      "loss": 0.9514,
+      "step": 2698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.045317649841309,
+      "learning_rate": 1.59622052704577e-05,
+      "loss": 0.8013,
+      "step": 2699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.65354061126709,
+      "learning_rate": 1.5957004160887655e-05,
+      "loss": 0.9131,
+      "step": 2700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.68481731414795,
+      "learning_rate": 1.5951803051317617e-05,
+      "loss": 0.8803,
+      "step": 2701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.939493179321289,
+      "learning_rate": 1.5946601941747572e-05,
+      "loss": 1.1382,
+      "step": 2702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.671445846557617,
+      "learning_rate": 1.5941400832177534e-05,
+      "loss": 0.8172,
+      "step": 2703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.500269889831543,
+      "learning_rate": 1.593619972260749e-05,
+      "loss": 0.6959,
+      "step": 2704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.744132995605469,
+      "learning_rate": 1.593099861303745e-05,
+      "loss": 0.9775,
+      "step": 2705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.186996459960938,
+      "learning_rate": 1.5925797503467406e-05,
+      "loss": 0.8097,
+      "step": 2706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.458073616027832,
+      "learning_rate": 1.5920596393897365e-05,
+      "loss": 0.691,
+      "step": 2707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.493504524230957,
+      "learning_rate": 1.5915395284327324e-05,
+      "loss": 0.8171,
+      "step": 2708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.44349479675293,
+      "learning_rate": 1.5910194174757282e-05,
+      "loss": 0.8527,
+      "step": 2709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.716134071350098,
+      "learning_rate": 1.590499306518724e-05,
+      "loss": 0.8394,
+      "step": 2710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.856500625610352,
+      "learning_rate": 1.58997919556172e-05,
+      "loss": 0.7968,
+      "step": 2711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.7447052001953125,
+      "learning_rate": 1.5894590846047158e-05,
+      "loss": 0.8695,
+      "step": 2712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.850167274475098,
+      "learning_rate": 1.5889389736477117e-05,
+      "loss": 1.0726,
+      "step": 2713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.88703727722168,
+      "learning_rate": 1.5884188626907072e-05,
+      "loss": 0.7605,
+      "step": 2714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.859131813049316,
+      "learning_rate": 1.5878987517337034e-05,
+      "loss": 0.7941,
+      "step": 2715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.549976348876953,
+      "learning_rate": 1.587378640776699e-05,
+      "loss": 0.9651,
+      "step": 2716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.094198226928711,
+      "learning_rate": 1.586858529819695e-05,
+      "loss": 0.9772,
+      "step": 2717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.321313858032227,
+      "learning_rate": 1.5863384188626906e-05,
+      "loss": 1.0067,
+      "step": 2718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.55509090423584,
+      "learning_rate": 1.5858183079056868e-05,
+      "loss": 0.9106,
+      "step": 2719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.154003143310547,
+      "learning_rate": 1.5852981969486824e-05,
+      "loss": 0.7489,
+      "step": 2720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.573179244995117,
+      "learning_rate": 1.5847780859916786e-05,
+      "loss": 0.6638,
+      "step": 2721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.511068344116211,
+      "learning_rate": 1.584257975034674e-05,
+      "loss": 0.6515,
+      "step": 2722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.486224174499512,
+      "learning_rate": 1.58373786407767e-05,
+      "loss": 1.0078,
+      "step": 2723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.8227219581604,
+      "learning_rate": 1.5832177531206658e-05,
+      "loss": 0.745,
+      "step": 2724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.142109870910645,
+      "learning_rate": 1.5826976421636617e-05,
+      "loss": 0.8015,
+      "step": 2725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.438566207885742,
+      "learning_rate": 1.5821775312066575e-05,
+      "loss": 1.1274,
+      "step": 2726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.971293926239014,
+      "learning_rate": 1.5816574202496534e-05,
+      "loss": 0.5685,
+      "step": 2727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.103302955627441,
+      "learning_rate": 1.5811373092926492e-05,
+      "loss": 0.7565,
+      "step": 2728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.285184383392334,
+      "learning_rate": 1.580617198335645e-05,
+      "loss": 0.7551,
+      "step": 2729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.508404731750488,
+      "learning_rate": 1.5800970873786406e-05,
+      "loss": 0.791,
+      "step": 2730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.229925155639648,
+      "learning_rate": 1.5795769764216368e-05,
+      "loss": 0.6538,
+      "step": 2731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.519681930541992,
+      "learning_rate": 1.5790568654646323e-05,
+      "loss": 0.8607,
+      "step": 2732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.970123291015625,
+      "learning_rate": 1.5785367545076285e-05,
+      "loss": 0.7041,
+      "step": 2733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.734323501586914,
+      "learning_rate": 1.578016643550624e-05,
+      "loss": 0.4817,
+      "step": 2734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.159523010253906,
+      "learning_rate": 1.5774965325936203e-05,
+      "loss": 0.8281,
+      "step": 2735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.204686164855957,
+      "learning_rate": 1.5769764216366158e-05,
+      "loss": 0.6481,
+      "step": 2736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.24722957611084,
+      "learning_rate": 1.5764563106796117e-05,
+      "loss": 1.0653,
+      "step": 2737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.194380760192871,
+      "learning_rate": 1.5759361997226075e-05,
+      "loss": 0.7497,
+      "step": 2738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.742879390716553,
+      "learning_rate": 1.5754160887656034e-05,
+      "loss": 0.7581,
+      "step": 2739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.092025756835938,
+      "learning_rate": 1.5748959778085992e-05,
+      "loss": 0.8585,
+      "step": 2740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.799080848693848,
+      "learning_rate": 1.574375866851595e-05,
+      "loss": 0.9431,
+      "step": 2741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.743739128112793,
+      "learning_rate": 1.573855755894591e-05,
+      "loss": 0.9264,
+      "step": 2742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.857199668884277,
+      "learning_rate": 1.5733356449375868e-05,
+      "loss": 0.8629,
+      "step": 2743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.325883865356445,
+      "learning_rate": 1.5728155339805827e-05,
+      "loss": 0.8016,
+      "step": 2744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.4783759117126465,
+      "learning_rate": 1.5722954230235785e-05,
+      "loss": 0.7256,
+      "step": 2745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.3095121383667,
+      "learning_rate": 1.571775312066574e-05,
+      "loss": 1.0435,
+      "step": 2746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.400176048278809,
+      "learning_rate": 1.5712552011095703e-05,
+      "loss": 0.645,
+      "step": 2747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.641757011413574,
+      "learning_rate": 1.5707350901525658e-05,
+      "loss": 0.7898,
+      "step": 2748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.999951362609863,
+      "learning_rate": 1.570214979195562e-05,
+      "loss": 0.9445,
+      "step": 2749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.229620933532715,
+      "learning_rate": 1.5696948682385575e-05,
+      "loss": 0.6356,
+      "step": 2750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.866415977478027,
+      "learning_rate": 1.5691747572815537e-05,
+      "loss": 0.9061,
+      "step": 2751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.695892333984375,
+      "learning_rate": 1.5686546463245492e-05,
+      "loss": 0.9337,
+      "step": 2752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.573240280151367,
+      "learning_rate": 1.568134535367545e-05,
+      "loss": 0.7226,
+      "step": 2753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.498461723327637,
+      "learning_rate": 1.567614424410541e-05,
+      "loss": 1.034,
+      "step": 2754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.051616668701172,
+      "learning_rate": 1.5670943134535368e-05,
+      "loss": 0.8391,
+      "step": 2755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.917240142822266,
+      "learning_rate": 1.5665742024965327e-05,
+      "loss": 0.8391,
+      "step": 2756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.570353507995605,
+      "learning_rate": 1.5660540915395285e-05,
+      "loss": 0.6924,
+      "step": 2757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.296908378601074,
+      "learning_rate": 1.5655339805825244e-05,
+      "loss": 0.5971,
+      "step": 2758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.444250106811523,
+      "learning_rate": 1.5650138696255203e-05,
+      "loss": 0.9077,
+      "step": 2759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.521012306213379,
+      "learning_rate": 1.564493758668516e-05,
+      "loss": 0.814,
+      "step": 2760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.404314994812012,
+      "learning_rate": 1.563973647711512e-05,
+      "loss": 0.8576,
+      "step": 2761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.548666954040527,
+      "learning_rate": 1.5634535367545075e-05,
+      "loss": 0.9781,
+      "step": 2762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.679899215698242,
+      "learning_rate": 1.5629334257975037e-05,
+      "loss": 0.7582,
+      "step": 2763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.389328002929688,
+      "learning_rate": 1.5624133148404992e-05,
+      "loss": 0.8144,
+      "step": 2764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.648536682128906,
+      "learning_rate": 1.5618932038834954e-05,
+      "loss": 0.8753,
+      "step": 2765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.204187393188477,
+      "learning_rate": 1.561373092926491e-05,
+      "loss": 0.7748,
+      "step": 2766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.056441307067871,
+      "learning_rate": 1.560852981969487e-05,
+      "loss": 0.8187,
+      "step": 2767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.9407377243042,
+      "learning_rate": 1.5603328710124827e-05,
+      "loss": 0.676,
+      "step": 2768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.89291000366211,
+      "learning_rate": 1.5598127600554785e-05,
+      "loss": 0.8257,
+      "step": 2769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.297140121459961,
+      "learning_rate": 1.5592926490984744e-05,
+      "loss": 0.7226,
+      "step": 2770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.355209350585938,
+      "learning_rate": 1.5587725381414703e-05,
+      "loss": 0.7509,
+      "step": 2771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.719225883483887,
+      "learning_rate": 1.558252427184466e-05,
+      "loss": 0.967,
+      "step": 2772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.206822395324707,
+      "learning_rate": 1.557732316227462e-05,
+      "loss": 0.9011,
+      "step": 2773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.215839385986328,
+      "learning_rate": 1.557212205270458e-05,
+      "loss": 0.8255,
+      "step": 2774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.222729682922363,
+      "learning_rate": 1.5566920943134537e-05,
+      "loss": 1.1003,
+      "step": 2775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.912050247192383,
+      "learning_rate": 1.5561719833564492e-05,
+      "loss": 0.8799,
+      "step": 2776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.180534362792969,
+      "learning_rate": 1.5556518723994454e-05,
+      "loss": 0.7134,
+      "step": 2777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.699484825134277,
+      "learning_rate": 1.555131761442441e-05,
+      "loss": 0.9529,
+      "step": 2778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.967399597167969,
+      "learning_rate": 1.554611650485437e-05,
+      "loss": 0.5688,
+      "step": 2779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.325196266174316,
+      "learning_rate": 1.5540915395284327e-05,
+      "loss": 0.8032,
+      "step": 2780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.936700820922852,
+      "learning_rate": 1.553571428571429e-05,
+      "loss": 0.8145,
+      "step": 2781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.552833557128906,
+      "learning_rate": 1.5530513176144244e-05,
+      "loss": 0.852,
+      "step": 2782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.914388656616211,
+      "learning_rate": 1.5525312066574206e-05,
+      "loss": 0.6022,
+      "step": 2783
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.78986930847168,
+      "learning_rate": 1.552011095700416e-05,
+      "loss": 0.689,
+      "step": 2784
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.121807098388672,
+      "learning_rate": 1.551490984743412e-05,
+      "loss": 0.7238,
+      "step": 2785
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.120682716369629,
+      "learning_rate": 1.550970873786408e-05,
+      "loss": 0.9493,
+      "step": 2786
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.852214813232422,
+      "learning_rate": 1.5504507628294037e-05,
+      "loss": 0.7346,
+      "step": 2787
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.964700698852539,
+      "learning_rate": 1.5499306518723996e-05,
+      "loss": 0.837,
+      "step": 2788
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.389676094055176,
+      "learning_rate": 1.5494105409153954e-05,
+      "loss": 0.7262,
+      "step": 2789
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.699017524719238,
+      "learning_rate": 1.5488904299583913e-05,
+      "loss": 0.6453,
+      "step": 2790
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.630216598510742,
+      "learning_rate": 1.548370319001387e-05,
+      "loss": 0.8196,
+      "step": 2791
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.08828067779541,
+      "learning_rate": 1.5478502080443827e-05,
+      "loss": 1.1384,
+      "step": 2792
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.585906982421875,
+      "learning_rate": 1.547330097087379e-05,
+      "loss": 0.7978,
+      "step": 2793
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.209080696105957,
+      "learning_rate": 1.5468099861303744e-05,
+      "loss": 0.7331,
+      "step": 2794
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.966225624084473,
+      "learning_rate": 1.5462898751733706e-05,
+      "loss": 0.7271,
+      "step": 2795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.752635478973389,
+      "learning_rate": 1.545769764216366e-05,
+      "loss": 0.5502,
+      "step": 2796
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.33692455291748,
+      "learning_rate": 1.5452496532593623e-05,
+      "loss": 0.6897,
+      "step": 2797
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.5021390914917,
+      "learning_rate": 1.544729542302358e-05,
+      "loss": 0.6569,
+      "step": 2798
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.656229019165039,
+      "learning_rate": 1.5442094313453537e-05,
+      "loss": 0.7575,
+      "step": 2799
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.655914306640625,
+      "learning_rate": 1.5436893203883496e-05,
+      "loss": 0.7894,
+      "step": 2800
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.49217414855957,
+      "learning_rate": 1.5431692094313454e-05,
+      "loss": 0.7045,
+      "step": 2801
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.39480209350586,
+      "learning_rate": 1.5426490984743413e-05,
+      "loss": 0.8626,
+      "step": 2802
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.8488130569458,
+      "learning_rate": 1.542128987517337e-05,
+      "loss": 0.6824,
+      "step": 2803
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.377074241638184,
+      "learning_rate": 1.541608876560333e-05,
+      "loss": 0.7999,
+      "step": 2804
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.25821304321289,
+      "learning_rate": 1.541088765603329e-05,
+      "loss": 0.6716,
+      "step": 2805
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.612323760986328,
+      "learning_rate": 1.5405686546463247e-05,
+      "loss": 1.101,
+      "step": 2806
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.930768013000488,
+      "learning_rate": 1.5400485436893206e-05,
+      "loss": 0.5392,
+      "step": 2807
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.386237144470215,
+      "learning_rate": 1.539528432732316e-05,
+      "loss": 0.8293,
+      "step": 2808
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.770174980163574,
+      "learning_rate": 1.5390083217753123e-05,
+      "loss": 0.8808,
+      "step": 2809
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.426881790161133,
+      "learning_rate": 1.5384882108183078e-05,
+      "loss": 0.8096,
+      "step": 2810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.554659843444824,
+      "learning_rate": 1.537968099861304e-05,
+      "loss": 0.8788,
+      "step": 2811
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.845392227172852,
+      "learning_rate": 1.5374479889042996e-05,
+      "loss": 0.8912,
+      "step": 2812
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.933225631713867,
+      "learning_rate": 1.5369278779472958e-05,
+      "loss": 0.804,
+      "step": 2813
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.933081150054932,
+      "learning_rate": 1.5364077669902913e-05,
+      "loss": 0.5709,
+      "step": 2814
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.562551498413086,
+      "learning_rate": 1.535887656033287e-05,
+      "loss": 0.8773,
+      "step": 2815
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.417220115661621,
+      "learning_rate": 1.535367545076283e-05,
+      "loss": 0.7495,
+      "step": 2816
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.193485736846924,
+      "learning_rate": 1.534847434119279e-05,
+      "loss": 0.6074,
+      "step": 2817
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.241901397705078,
+      "learning_rate": 1.5343273231622747e-05,
+      "loss": 1.0412,
+      "step": 2818
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.171204566955566,
+      "learning_rate": 1.5338072122052706e-05,
+      "loss": 0.7534,
+      "step": 2819
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.7554731369018555,
+      "learning_rate": 1.5332871012482664e-05,
+      "loss": 0.6085,
+      "step": 2820
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.644204139709473,
+      "learning_rate": 1.5327669902912623e-05,
+      "loss": 0.5946,
+      "step": 2821
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.639865875244141,
+      "learning_rate": 1.532246879334258e-05,
+      "loss": 0.7176,
+      "step": 2822
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.911556243896484,
+      "learning_rate": 1.531726768377254e-05,
+      "loss": 0.7895,
+      "step": 2823
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.712705612182617,
+      "learning_rate": 1.5312066574202495e-05,
+      "loss": 0.7533,
+      "step": 2824
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.221652030944824,
+      "learning_rate": 1.5306865464632457e-05,
+      "loss": 1.1135,
+      "step": 2825
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.696203231811523,
+      "learning_rate": 1.5301664355062413e-05,
+      "loss": 0.7372,
+      "step": 2826
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.604036331176758,
+      "learning_rate": 1.5296463245492375e-05,
+      "loss": 0.7577,
+      "step": 2827
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.920551300048828,
+      "learning_rate": 1.529126213592233e-05,
+      "loss": 0.8906,
+      "step": 2828
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.170001029968262,
+      "learning_rate": 1.5286061026352292e-05,
+      "loss": 0.7907,
+      "step": 2829
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.201112747192383,
+      "learning_rate": 1.5280859916782247e-05,
+      "loss": 0.9412,
+      "step": 2830
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.22433090209961,
+      "learning_rate": 1.5275658807212206e-05,
+      "loss": 0.9861,
+      "step": 2831
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.269522666931152,
+      "learning_rate": 1.5270457697642164e-05,
+      "loss": 0.8007,
+      "step": 2832
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.982280731201172,
+      "learning_rate": 1.5265256588072123e-05,
+      "loss": 0.6293,
+      "step": 2833
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.69751262664795,
+      "learning_rate": 1.526005547850208e-05,
+      "loss": 0.8613,
+      "step": 2834
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.559819221496582,
+      "learning_rate": 1.525485436893204e-05,
+      "loss": 0.832,
+      "step": 2835
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.080559730529785,
+      "learning_rate": 1.5249653259361997e-05,
+      "loss": 0.837,
+      "step": 2836
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.15544605255127,
+      "learning_rate": 1.5244452149791957e-05,
+      "loss": 0.5802,
+      "step": 2837
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.891303539276123,
+      "learning_rate": 1.5239251040221914e-05,
+      "loss": 0.7946,
+      "step": 2838
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.603619575500488,
+      "learning_rate": 1.5234049930651875e-05,
+      "loss": 0.7713,
+      "step": 2839
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.131147384643555,
+      "learning_rate": 1.5228848821081832e-05,
+      "loss": 0.7577,
+      "step": 2840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.367525100708008,
+      "learning_rate": 1.5223647711511792e-05,
+      "loss": 0.7504,
+      "step": 2841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.72475814819336,
+      "learning_rate": 1.5218446601941747e-05,
+      "loss": 0.8395,
+      "step": 2842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.811549186706543,
+      "learning_rate": 1.5213245492371707e-05,
+      "loss": 0.8078,
+      "step": 2843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.961498260498047,
+      "learning_rate": 1.5208044382801664e-05,
+      "loss": 0.6018,
+      "step": 2844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.471916198730469,
+      "learning_rate": 1.5202843273231625e-05,
+      "loss": 0.7703,
+      "step": 2845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.71827507019043,
+      "learning_rate": 1.5197642163661582e-05,
+      "loss": 0.7293,
+      "step": 2846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.617841720581055,
+      "learning_rate": 1.5192441054091542e-05,
+      "loss": 1.0498,
+      "step": 2847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.811444282531738,
+      "learning_rate": 1.5187239944521499e-05,
+      "loss": 0.5619,
+      "step": 2848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.619905471801758,
+      "learning_rate": 1.5182038834951459e-05,
+      "loss": 1.0238,
+      "step": 2849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.76431941986084,
+      "learning_rate": 1.5176837725381414e-05,
+      "loss": 0.8663,
+      "step": 2850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.397789001464844,
+      "learning_rate": 1.5171636615811375e-05,
+      "loss": 0.8491,
+      "step": 2851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.408298015594482,
+      "learning_rate": 1.5166435506241332e-05,
+      "loss": 0.5825,
+      "step": 2852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.531935691833496,
+      "learning_rate": 1.5161234396671292e-05,
+      "loss": 0.6986,
+      "step": 2853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.823439598083496,
+      "learning_rate": 1.5156033287101249e-05,
+      "loss": 0.7298,
+      "step": 2854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.596158981323242,
+      "learning_rate": 1.5150832177531209e-05,
+      "loss": 0.7639,
+      "step": 2855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.907723426818848,
+      "learning_rate": 1.5145631067961166e-05,
+      "loss": 1.0873,
+      "step": 2856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.978068828582764,
+      "learning_rate": 1.5140429958391125e-05,
+      "loss": 0.707,
+      "step": 2857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.259562492370605,
+      "learning_rate": 1.5135228848821082e-05,
+      "loss": 0.9389,
+      "step": 2858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.640412330627441,
+      "learning_rate": 1.5130027739251042e-05,
+      "loss": 0.9589,
+      "step": 2859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.88586139678955,
+      "learning_rate": 1.5124826629680999e-05,
+      "loss": 0.7275,
+      "step": 2860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.609187602996826,
+      "learning_rate": 1.5119625520110959e-05,
+      "loss": 0.5905,
+      "step": 2861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.925213813781738,
+      "learning_rate": 1.5114424410540916e-05,
+      "loss": 0.5891,
+      "step": 2862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.370050430297852,
+      "learning_rate": 1.5109223300970876e-05,
+      "loss": 0.6806,
+      "step": 2863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.442910194396973,
+      "learning_rate": 1.5104022191400833e-05,
+      "loss": 0.7918,
+      "step": 2864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.414563179016113,
+      "learning_rate": 1.5098821081830792e-05,
+      "loss": 0.739,
+      "step": 2865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.013062477111816,
+      "learning_rate": 1.5093619972260749e-05,
+      "loss": 1.0026,
+      "step": 2866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.054566383361816,
+      "learning_rate": 1.5088418862690709e-05,
+      "loss": 0.7561,
+      "step": 2867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.863681316375732,
+      "learning_rate": 1.5083217753120666e-05,
+      "loss": 0.8671,
+      "step": 2868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.060341835021973,
+      "learning_rate": 1.5078016643550626e-05,
+      "loss": 0.8724,
+      "step": 2869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.256401062011719,
+      "learning_rate": 1.5072815533980583e-05,
+      "loss": 0.6849,
+      "step": 2870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.666465759277344,
+      "learning_rate": 1.5067614424410544e-05,
+      "loss": 0.6511,
+      "step": 2871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.94410514831543,
+      "learning_rate": 1.50624133148405e-05,
+      "loss": 0.8416,
+      "step": 2872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.07627010345459,
+      "learning_rate": 1.5057212205270459e-05,
+      "loss": 0.7978,
+      "step": 2873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.325894355773926,
+      "learning_rate": 1.5052011095700416e-05,
+      "loss": 0.9741,
+      "step": 2874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.969085693359375,
+      "learning_rate": 1.5046809986130376e-05,
+      "loss": 0.7281,
+      "step": 2875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.575811386108398,
+      "learning_rate": 1.5041608876560333e-05,
+      "loss": 0.8218,
+      "step": 2876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.655391693115234,
+      "learning_rate": 1.5036407766990293e-05,
+      "loss": 1.3581,
+      "step": 2877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.593942165374756,
+      "learning_rate": 1.503120665742025e-05,
+      "loss": 0.6686,
+      "step": 2878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.832137107849121,
+      "learning_rate": 1.502600554785021e-05,
+      "loss": 0.702,
+      "step": 2879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.946168899536133,
+      "learning_rate": 1.5020804438280168e-05,
+      "loss": 0.9193,
+      "step": 2880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.695586204528809,
+      "learning_rate": 1.5015603328710126e-05,
+      "loss": 0.8515,
+      "step": 2881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.698012351989746,
+      "learning_rate": 1.5010402219140083e-05,
+      "loss": 0.8351,
+      "step": 2882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.690600395202637,
+      "learning_rate": 1.5005201109570043e-05,
+      "loss": 0.7162,
+      "step": 2883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.788239479064941,
+      "learning_rate": 1.5e-05,
+      "loss": 1.0762,
+      "step": 2884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.296771049499512,
+      "learning_rate": 1.4994798890429959e-05,
+      "loss": 0.5584,
+      "step": 2885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.880526065826416,
+      "learning_rate": 1.4989597780859918e-05,
+      "loss": 0.5338,
+      "step": 2886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.61114501953125,
+      "learning_rate": 1.4984396671289876e-05,
+      "loss": 0.5055,
+      "step": 2887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.791881561279297,
+      "learning_rate": 1.4979195561719833e-05,
+      "loss": 0.9085,
+      "step": 2888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.803398132324219,
+      "learning_rate": 1.4973994452149792e-05,
+      "loss": 0.9722,
+      "step": 2889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.344821929931641,
+      "learning_rate": 1.496879334257975e-05,
+      "loss": 0.6021,
+      "step": 2890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.132574081420898,
+      "learning_rate": 1.4963592233009709e-05,
+      "loss": 0.5836,
+      "step": 2891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.875996589660645,
+      "learning_rate": 1.4958391123439668e-05,
+      "loss": 0.9709,
+      "step": 2892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.95775032043457,
+      "learning_rate": 1.4953190013869626e-05,
+      "loss": 0.7477,
+      "step": 2893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.453639030456543,
+      "learning_rate": 1.4947988904299585e-05,
+      "loss": 0.596,
+      "step": 2894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.426093101501465,
+      "learning_rate": 1.4942787794729543e-05,
+      "loss": 0.5446,
+      "step": 2895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.665059089660645,
+      "learning_rate": 1.49375866851595e-05,
+      "loss": 0.8142,
+      "step": 2896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.162260055541992,
+      "learning_rate": 1.4932385575589459e-05,
+      "loss": 0.7518,
+      "step": 2897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.911294937133789,
+      "learning_rate": 1.4927184466019418e-05,
+      "loss": 0.4999,
+      "step": 2898
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.275508880615234,
+      "learning_rate": 1.4921983356449376e-05,
+      "loss": 0.763,
+      "step": 2899
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.829892158508301,
+      "learning_rate": 1.4916782246879335e-05,
+      "loss": 0.6067,
+      "step": 2900
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.743669509887695,
+      "learning_rate": 1.4911581137309293e-05,
+      "loss": 0.7344,
+      "step": 2901
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.011699676513672,
+      "learning_rate": 1.4906380027739252e-05,
+      "loss": 0.4819,
+      "step": 2902
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.467980861663818,
+      "learning_rate": 1.490117891816921e-05,
+      "loss": 0.679,
+      "step": 2903
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.17066478729248,
+      "learning_rate": 1.4895977808599168e-05,
+      "loss": 0.804,
+      "step": 2904
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.253417015075684,
+      "learning_rate": 1.4890776699029126e-05,
+      "loss": 0.5411,
+      "step": 2905
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.02830982208252,
+      "learning_rate": 1.4885575589459085e-05,
+      "loss": 0.7488,
+      "step": 2906
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.416342735290527,
+      "learning_rate": 1.4880374479889043e-05,
+      "loss": 0.6206,
+      "step": 2907
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.394293785095215,
+      "learning_rate": 1.4875173370319002e-05,
+      "loss": 0.5616,
+      "step": 2908
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.015763282775879,
+      "learning_rate": 1.486997226074896e-05,
+      "loss": 0.5737,
+      "step": 2909
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 12.703664779663086,
+      "learning_rate": 1.486477115117892e-05,
+      "loss": 0.7005,
+      "step": 2910
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.736967086791992,
+      "learning_rate": 1.4859570041608878e-05,
+      "loss": 0.745,
+      "step": 2911
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.646903038024902,
+      "learning_rate": 1.4854368932038835e-05,
+      "loss": 0.4536,
+      "step": 2912
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.828563213348389,
+      "learning_rate": 1.4849167822468793e-05,
+      "loss": 0.5246,
+      "step": 2913
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.979222297668457,
+      "learning_rate": 1.4843966712898752e-05,
+      "loss": 0.8207,
+      "step": 2914
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.122727394104004,
+      "learning_rate": 1.483876560332871e-05,
+      "loss": 0.7985,
+      "step": 2915
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.18850040435791,
+      "learning_rate": 1.483356449375867e-05,
+      "loss": 0.6621,
+      "step": 2916
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.915452003479004,
+      "learning_rate": 1.4828363384188628e-05,
+      "loss": 0.7176,
+      "step": 2917
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.788103103637695,
+      "learning_rate": 1.4823162274618586e-05,
+      "loss": 0.7398,
+      "step": 2918
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.424139976501465,
+      "learning_rate": 1.4817961165048543e-05,
+      "loss": 0.5139,
+      "step": 2919
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.855794906616211,
+      "learning_rate": 1.4812760055478502e-05,
+      "loss": 0.5614,
+      "step": 2920
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.075289726257324,
+      "learning_rate": 1.480755894590846e-05,
+      "loss": 0.7535,
+      "step": 2921
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.43271541595459,
+      "learning_rate": 1.480235783633842e-05,
+      "loss": 0.6099,
+      "step": 2922
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.21056079864502,
+      "learning_rate": 1.4797156726768378e-05,
+      "loss": 0.7552,
+      "step": 2923
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.864224433898926,
+      "learning_rate": 1.4791955617198336e-05,
+      "loss": 0.793,
+      "step": 2924
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.00338077545166,
+      "learning_rate": 1.4786754507628295e-05,
+      "loss": 0.6255,
+      "step": 2925
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.13648509979248,
+      "learning_rate": 1.4781553398058254e-05,
+      "loss": 0.7288,
+      "step": 2926
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.154802322387695,
+      "learning_rate": 1.477635228848821e-05,
+      "loss": 0.5707,
+      "step": 2927
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.031853199005127,
+      "learning_rate": 1.477115117891817e-05,
+      "loss": 0.4779,
+      "step": 2928
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.882038116455078,
+      "learning_rate": 1.4765950069348128e-05,
+      "loss": 0.5892,
+      "step": 2929
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.608929634094238,
+      "learning_rate": 1.4760748959778086e-05,
+      "loss": 0.8154,
+      "step": 2930
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.829019069671631,
+      "learning_rate": 1.4755547850208045e-05,
+      "loss": 0.4688,
+      "step": 2931
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.114839553833008,
+      "learning_rate": 1.4750346740638004e-05,
+      "loss": 0.7166,
+      "step": 2932
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.383705139160156,
+      "learning_rate": 1.4745145631067962e-05,
+      "loss": 0.6184,
+      "step": 2933
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.212657928466797,
+      "learning_rate": 1.473994452149792e-05,
+      "loss": 0.6354,
+      "step": 2934
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.5133562088012695,
+      "learning_rate": 1.4734743411927878e-05,
+      "loss": 0.3535,
+      "step": 2935
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.669210433959961,
+      "learning_rate": 1.4729542302357836e-05,
+      "loss": 0.5864,
+      "step": 2936
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.589670181274414,
+      "learning_rate": 1.4724341192787795e-05,
+      "loss": 0.727,
+      "step": 2937
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.909879684448242,
+      "learning_rate": 1.4719140083217754e-05,
+      "loss": 0.6842,
+      "step": 2938
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.818959712982178,
+      "learning_rate": 1.4713938973647712e-05,
+      "loss": 0.5381,
+      "step": 2939
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.44021987915039,
+      "learning_rate": 1.470873786407767e-05,
+      "loss": 0.7184,
+      "step": 2940
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.281930923461914,
+      "learning_rate": 1.470353675450763e-05,
+      "loss": 0.5144,
+      "step": 2941
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.562447547912598,
+      "learning_rate": 1.4698335644937586e-05,
+      "loss": 0.6964,
+      "step": 2942
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.162126541137695,
+      "learning_rate": 1.4693134535367545e-05,
+      "loss": 0.6943,
+      "step": 2943
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.883671760559082,
+      "learning_rate": 1.4687933425797504e-05,
+      "loss": 0.7755,
+      "step": 2944
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.838099479675293,
+      "learning_rate": 1.4682732316227462e-05,
+      "loss": 0.5395,
+      "step": 2945
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.091822624206543,
+      "learning_rate": 1.467753120665742e-05,
+      "loss": 0.7848,
+      "step": 2946
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.504087448120117,
+      "learning_rate": 1.467233009708738e-05,
+      "loss": 0.6012,
+      "step": 2947
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.135701179504395,
+      "learning_rate": 1.4667128987517338e-05,
+      "loss": 0.765,
+      "step": 2948
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.072742462158203,
+      "learning_rate": 1.4661927877947297e-05,
+      "loss": 0.7332,
+      "step": 2949
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.639333724975586,
+      "learning_rate": 1.4656726768377254e-05,
+      "loss": 0.7153,
+      "step": 2950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.423262596130371,
+      "learning_rate": 1.4651525658807212e-05,
+      "loss": 0.7063,
+      "step": 2951
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.710224151611328,
+      "learning_rate": 1.464632454923717e-05,
+      "loss": 0.6973,
+      "step": 2952
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.9091644287109375,
+      "learning_rate": 1.464112343966713e-05,
+      "loss": 0.6457,
+      "step": 2953
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.061420440673828,
+      "learning_rate": 1.4635922330097088e-05,
+      "loss": 0.7195,
+      "step": 2954
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.997738838195801,
+      "learning_rate": 1.4630721220527047e-05,
+      "loss": 0.61,
+      "step": 2955
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.495656967163086,
+      "learning_rate": 1.4625520110957005e-05,
+      "loss": 0.8504,
+      "step": 2956
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.097314834594727,
+      "learning_rate": 1.4620319001386964e-05,
+      "loss": 1.053,
+      "step": 2957
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.730443000793457,
+      "learning_rate": 1.461511789181692e-05,
+      "loss": 0.6342,
+      "step": 2958
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.7630205154418945,
+      "learning_rate": 1.460991678224688e-05,
+      "loss": 0.581,
+      "step": 2959
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.571893692016602,
+      "learning_rate": 1.4604715672676838e-05,
+      "loss": 0.9295,
+      "step": 2960
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.906530380249023,
+      "learning_rate": 1.4599514563106797e-05,
+      "loss": 0.8194,
+      "step": 2961
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.590948104858398,
+      "learning_rate": 1.4594313453536755e-05,
+      "loss": 0.6634,
+      "step": 2962
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.812472343444824,
+      "learning_rate": 1.4589112343966714e-05,
+      "loss": 0.7333,
+      "step": 2963
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.562383651733398,
+      "learning_rate": 1.4583911234396672e-05,
+      "loss": 0.7003,
+      "step": 2964
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.011970520019531,
+      "learning_rate": 1.4578710124826631e-05,
+      "loss": 0.6887,
+      "step": 2965
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.477331161499023,
+      "learning_rate": 1.4573509015256588e-05,
+      "loss": 0.8064,
+      "step": 2966
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.882014274597168,
+      "learning_rate": 1.4568307905686547e-05,
+      "loss": 0.6365,
+      "step": 2967
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.87427806854248,
+      "learning_rate": 1.4563106796116505e-05,
+      "loss": 0.8805,
+      "step": 2968
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.582829475402832,
+      "learning_rate": 1.4557905686546464e-05,
+      "loss": 0.6472,
+      "step": 2969
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.24707555770874,
+      "learning_rate": 1.4552704576976422e-05,
+      "loss": 0.4814,
+      "step": 2970
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.817434310913086,
+      "learning_rate": 1.4547503467406381e-05,
+      "loss": 0.9053,
+      "step": 2971
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.66775894165039,
+      "learning_rate": 1.454230235783634e-05,
+      "loss": 0.7781,
+      "step": 2972
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.775761604309082,
+      "learning_rate": 1.4537101248266297e-05,
+      "loss": 0.4567,
+      "step": 2973
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.641385078430176,
+      "learning_rate": 1.4531900138696255e-05,
+      "loss": 0.8476,
+      "step": 2974
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 10.27375602722168,
+      "learning_rate": 1.4526699029126214e-05,
+      "loss": 0.4869,
+      "step": 2975
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.533365249633789,
+      "learning_rate": 1.4521497919556172e-05,
+      "loss": 0.5322,
+      "step": 2976
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.908929824829102,
+      "learning_rate": 1.4516296809986131e-05,
+      "loss": 0.7134,
+      "step": 2977
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.531363487243652,
+      "learning_rate": 1.451109570041609e-05,
+      "loss": 0.7399,
+      "step": 2978
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.875345230102539,
+      "learning_rate": 1.4505894590846048e-05,
+      "loss": 0.4954,
+      "step": 2979
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.492164611816406,
+      "learning_rate": 1.4500693481276007e-05,
+      "loss": 0.6042,
+      "step": 2980
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.779877662658691,
+      "learning_rate": 1.4495492371705964e-05,
+      "loss": 0.485,
+      "step": 2981
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 13.125850677490234,
+      "learning_rate": 1.4490291262135922e-05,
+      "loss": 0.7077,
+      "step": 2982
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.895146369934082,
+      "learning_rate": 1.4485090152565881e-05,
+      "loss": 0.7053,
+      "step": 2983
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.739556789398193,
+      "learning_rate": 1.447988904299584e-05,
+      "loss": 0.4375,
+      "step": 2984
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.637362480163574,
+      "learning_rate": 1.4474687933425798e-05,
+      "loss": 0.7279,
+      "step": 2985
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.125722885131836,
+      "learning_rate": 1.4469486823855757e-05,
+      "loss": 0.6974,
+      "step": 2986
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.150125503540039,
+      "learning_rate": 1.4464285714285715e-05,
+      "loss": 0.7708,
+      "step": 2987
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.454068183898926,
+      "learning_rate": 1.4459084604715674e-05,
+      "loss": 0.5581,
+      "step": 2988
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.365803718566895,
+      "learning_rate": 1.4453883495145631e-05,
+      "loss": 0.6864,
+      "step": 2989
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.665801048278809,
+      "learning_rate": 1.444868238557559e-05,
+      "loss": 0.6367,
+      "step": 2990
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.856833457946777,
+      "learning_rate": 1.4443481276005548e-05,
+      "loss": 0.5993,
+      "step": 2991
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.591376304626465,
+      "learning_rate": 1.4438280166435507e-05,
+      "loss": 0.5865,
+      "step": 2992
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 12.617863655090332,
+      "learning_rate": 1.4433079056865465e-05,
+      "loss": 0.6744,
+      "step": 2993
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.354214668273926,
+      "learning_rate": 1.4427877947295424e-05,
+      "loss": 0.8505,
+      "step": 2994
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.892980575561523,
+      "learning_rate": 1.4422676837725383e-05,
+      "loss": 0.6683,
+      "step": 2995
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.853412628173828,
+      "learning_rate": 1.4417475728155341e-05,
+      "loss": 0.7431,
+      "step": 2996
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.582582473754883,
+      "learning_rate": 1.4412274618585298e-05,
+      "loss": 0.5658,
+      "step": 2997
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.623527526855469,
+      "learning_rate": 1.4407073509015257e-05,
+      "loss": 0.6643,
+      "step": 2998
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.61246395111084,
+      "learning_rate": 1.4401872399445215e-05,
+      "loss": 0.6372,
+      "step": 2999
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.683486938476562,
+      "learning_rate": 1.4396671289875174e-05,
+      "loss": 0.6981,
+      "step": 3000
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.080511093139648,
+      "learning_rate": 1.4391470180305133e-05,
+      "loss": 0.7633,
+      "step": 3001
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.806725978851318,
+      "learning_rate": 1.4386269070735091e-05,
+      "loss": 0.5768,
+      "step": 3002
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.041475296020508,
+      "learning_rate": 1.438106796116505e-05,
+      "loss": 0.6325,
+      "step": 3003
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.830497741699219,
+      "learning_rate": 1.4375866851595007e-05,
+      "loss": 0.6942,
+      "step": 3004
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.549383163452148,
+      "learning_rate": 1.4370665742024965e-05,
+      "loss": 0.5699,
+      "step": 3005
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.994110107421875,
+      "learning_rate": 1.4365464632454924e-05,
+      "loss": 0.5367,
+      "step": 3006
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.752497673034668,
+      "learning_rate": 1.4360263522884883e-05,
+      "loss": 0.5931,
+      "step": 3007
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.876712799072266,
+      "learning_rate": 1.4355062413314841e-05,
+      "loss": 0.6855,
+      "step": 3008
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.174016952514648,
+      "learning_rate": 1.43498613037448e-05,
+      "loss": 0.7325,
+      "step": 3009
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.930347442626953,
+      "learning_rate": 1.4344660194174758e-05,
+      "loss": 0.7633,
+      "step": 3010
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.097058296203613,
+      "learning_rate": 1.4339459084604717e-05,
+      "loss": 0.5558,
+      "step": 3011
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.351913452148438,
+      "learning_rate": 1.4334257975034674e-05,
+      "loss": 0.6062,
+      "step": 3012
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 7.742211818695068,
+      "learning_rate": 1.4329056865464633e-05,
+      "loss": 0.599,
+      "step": 3013
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.972253322601318,
+      "learning_rate": 1.4323855755894591e-05,
+      "loss": 0.5522,
+      "step": 3014
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.271120071411133,
+      "learning_rate": 1.431865464632455e-05,
+      "loss": 0.5933,
+      "step": 3015
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.537859916687012,
+      "learning_rate": 1.4313453536754508e-05,
+      "loss": 0.8168,
+      "step": 3016
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.072148323059082,
+      "learning_rate": 1.4308252427184467e-05,
+      "loss": 0.5698,
+      "step": 3017
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.69576358795166,
+      "learning_rate": 1.4303051317614426e-05,
+      "loss": 0.9067,
+      "step": 3018
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.900772094726562,
+      "learning_rate": 1.4297850208044384e-05,
+      "loss": 0.7807,
+      "step": 3019
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.132246494293213,
+      "learning_rate": 1.4292649098474341e-05,
+      "loss": 0.637,
+      "step": 3020
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.999605178833008,
+      "learning_rate": 1.42874479889043e-05,
+      "loss": 0.753,
+      "step": 3021
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.111726760864258,
+      "learning_rate": 1.4282246879334258e-05,
+      "loss": 0.5462,
+      "step": 3022
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.222118377685547,
+      "learning_rate": 1.4277045769764217e-05,
+      "loss": 0.4775,
+      "step": 3023
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.128279685974121,
+      "learning_rate": 1.4271844660194176e-05,
+      "loss": 0.5738,
+      "step": 3024
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.48021411895752,
+      "learning_rate": 1.4266643550624134e-05,
+      "loss": 0.6549,
+      "step": 3025
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.142247200012207,
+      "learning_rate": 1.4261442441054093e-05,
+      "loss": 0.6039,
+      "step": 3026
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.1122465133667,
+      "learning_rate": 1.4256241331484051e-05,
+      "loss": 0.7743,
+      "step": 3027
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.789977073669434,
+      "learning_rate": 1.4251040221914008e-05,
+      "loss": 0.772,
+      "step": 3028
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.591363906860352,
+      "learning_rate": 1.4245839112343967e-05,
+      "loss": 0.5895,
+      "step": 3029
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.25362777709961,
+      "learning_rate": 1.4240638002773926e-05,
+      "loss": 0.5915,
+      "step": 3030
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.759230613708496,
+      "learning_rate": 1.4235436893203884e-05,
+      "loss": 0.5305,
+      "step": 3031
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.457803726196289,
+      "learning_rate": 1.4230235783633843e-05,
+      "loss": 0.7034,
+      "step": 3032
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.757201194763184,
+      "learning_rate": 1.4225034674063801e-05,
+      "loss": 0.9803,
+      "step": 3033
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.975438117980957,
+      "learning_rate": 1.421983356449376e-05,
+      "loss": 0.6908,
+      "step": 3034
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.811638832092285,
+      "learning_rate": 1.4214632454923717e-05,
+      "loss": 0.5662,
+      "step": 3035
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.030835151672363,
+      "learning_rate": 1.4209431345353676e-05,
+      "loss": 0.7064,
+      "step": 3036
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.454142570495605,
+      "learning_rate": 1.4204230235783634e-05,
+      "loss": 0.5918,
+      "step": 3037
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.646995544433594,
+      "learning_rate": 1.4199029126213593e-05,
+      "loss": 0.7193,
+      "step": 3038
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.595490455627441,
+      "learning_rate": 1.4193828016643551e-05,
+      "loss": 0.6109,
+      "step": 3039
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.926987171173096,
+      "learning_rate": 1.418862690707351e-05,
+      "loss": 0.608,
+      "step": 3040
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.782769680023193,
+      "learning_rate": 1.4183425797503469e-05,
+      "loss": 0.5437,
+      "step": 3041
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 10.960162162780762,
+      "learning_rate": 1.4178224687933427e-05,
+      "loss": 0.6944,
+      "step": 3042
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.04822063446045,
+      "learning_rate": 1.4173023578363384e-05,
+      "loss": 0.7186,
+      "step": 3043
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.306471824645996,
+      "learning_rate": 1.4167822468793343e-05,
+      "loss": 0.8814,
+      "step": 3044
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.5531587600708,
+      "learning_rate": 1.4162621359223301e-05,
+      "loss": 0.6529,
+      "step": 3045
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.78360652923584,
+      "learning_rate": 1.415742024965326e-05,
+      "loss": 0.5894,
+      "step": 3046
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.288869857788086,
+      "learning_rate": 1.4152219140083219e-05,
+      "loss": 0.8349,
+      "step": 3047
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.004966735839844,
+      "learning_rate": 1.4147018030513177e-05,
+      "loss": 0.7166,
+      "step": 3048
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.292329788208008,
+      "learning_rate": 1.4141816920943136e-05,
+      "loss": 0.4188,
+      "step": 3049
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.87419605255127,
+      "learning_rate": 1.4136615811373094e-05,
+      "loss": 0.7878,
+      "step": 3050
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.182921409606934,
+      "learning_rate": 1.4131414701803051e-05,
+      "loss": 0.7652,
+      "step": 3051
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.460906982421875,
+      "learning_rate": 1.412621359223301e-05,
+      "loss": 0.7301,
+      "step": 3052
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.00776481628418,
+      "learning_rate": 1.4121012482662969e-05,
+      "loss": 0.8708,
+      "step": 3053
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.652389526367188,
+      "learning_rate": 1.4115811373092927e-05,
+      "loss": 0.6528,
+      "step": 3054
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.148279190063477,
+      "learning_rate": 1.4110610263522886e-05,
+      "loss": 0.5677,
+      "step": 3055
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.058034896850586,
+      "learning_rate": 1.4105409153952844e-05,
+      "loss": 0.715,
+      "step": 3056
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.713724136352539,
+      "learning_rate": 1.4100208044382803e-05,
+      "loss": 0.5859,
+      "step": 3057
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.186271667480469,
+      "learning_rate": 1.409500693481276e-05,
+      "loss": 0.7346,
+      "step": 3058
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.460636138916016,
+      "learning_rate": 1.4089805825242719e-05,
+      "loss": 0.7004,
+      "step": 3059
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.244153022766113,
+      "learning_rate": 1.4084604715672677e-05,
+      "loss": 0.6917,
+      "step": 3060
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.061360359191895,
+      "learning_rate": 1.4079403606102636e-05,
+      "loss": 0.6781,
+      "step": 3061
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.882063865661621,
+      "learning_rate": 1.4074202496532594e-05,
+      "loss": 0.8223,
+      "step": 3062
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.636629104614258,
+      "learning_rate": 1.4069001386962553e-05,
+      "loss": 0.8719,
+      "step": 3063
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.0866265296936035,
+      "learning_rate": 1.4063800277392512e-05,
+      "loss": 0.6504,
+      "step": 3064
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.153482437133789,
+      "learning_rate": 1.405859916782247e-05,
+      "loss": 0.6972,
+      "step": 3065
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.479498863220215,
+      "learning_rate": 1.4053398058252427e-05,
+      "loss": 0.7597,
+      "step": 3066
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.302968978881836,
+      "learning_rate": 1.4048196948682386e-05,
+      "loss": 0.7673,
+      "step": 3067
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.474740028381348,
+      "learning_rate": 1.4042995839112344e-05,
+      "loss": 0.7891,
+      "step": 3068
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.914313316345215,
+      "learning_rate": 1.4037794729542303e-05,
+      "loss": 0.6762,
+      "step": 3069
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.513034343719482,
+      "learning_rate": 1.4032593619972262e-05,
+      "loss": 0.5633,
+      "step": 3070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.935517311096191,
+      "learning_rate": 1.402739251040222e-05,
+      "loss": 0.6287,
+      "step": 3071
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.04997730255127,
+      "learning_rate": 1.4022191400832179e-05,
+      "loss": 0.8404,
+      "step": 3072
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.63071346282959,
+      "learning_rate": 1.4016990291262137e-05,
+      "loss": 0.721,
+      "step": 3073
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.222707748413086,
+      "learning_rate": 1.4011789181692094e-05,
+      "loss": 0.8229,
+      "step": 3074
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.635363578796387,
+      "learning_rate": 1.4006588072122053e-05,
+      "loss": 0.7761,
+      "step": 3075
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.812986373901367,
+      "learning_rate": 1.4001386962552012e-05,
+      "loss": 0.7019,
+      "step": 3076
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.50902795791626,
+      "learning_rate": 1.399618585298197e-05,
+      "loss": 0.6318,
+      "step": 3077
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.718852043151855,
+      "learning_rate": 1.3990984743411929e-05,
+      "loss": 0.5163,
+      "step": 3078
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.147627830505371,
+      "learning_rate": 1.3985783633841887e-05,
+      "loss": 0.7148,
+      "step": 3079
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.825345993041992,
+      "learning_rate": 1.3980582524271846e-05,
+      "loss": 0.8276,
+      "step": 3080
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.087700843811035,
+      "learning_rate": 1.3975381414701805e-05,
+      "loss": 0.5715,
+      "step": 3081
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.905198097229004,
+      "learning_rate": 1.3970180305131762e-05,
+      "loss": 0.6437,
+      "step": 3082
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.60615348815918,
+      "learning_rate": 1.396497919556172e-05,
+      "loss": 0.7156,
+      "step": 3083
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.642374992370605,
+      "learning_rate": 1.3959778085991679e-05,
+      "loss": 0.6027,
+      "step": 3084
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.211601734161377,
+      "learning_rate": 1.3954576976421637e-05,
+      "loss": 0.5075,
+      "step": 3085
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.48480749130249,
+      "learning_rate": 1.3949375866851596e-05,
+      "loss": 0.6077,
+      "step": 3086
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.834716796875,
+      "learning_rate": 1.3944174757281555e-05,
+      "loss": 0.6724,
+      "step": 3087
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.283005714416504,
+      "learning_rate": 1.3938973647711513e-05,
+      "loss": 0.7603,
+      "step": 3088
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.275066375732422,
+      "learning_rate": 1.393377253814147e-05,
+      "loss": 0.5594,
+      "step": 3089
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.036904335021973,
+      "learning_rate": 1.3928571428571429e-05,
+      "loss": 0.4803,
+      "step": 3090
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.866251945495605,
+      "learning_rate": 1.3923370319001387e-05,
+      "loss": 0.6427,
+      "step": 3091
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.316411018371582,
+      "learning_rate": 1.3918169209431346e-05,
+      "loss": 0.7155,
+      "step": 3092
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.2617669105529785,
+      "learning_rate": 1.3912968099861305e-05,
+      "loss": 0.3892,
+      "step": 3093
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.989748954772949,
+      "learning_rate": 1.3907766990291263e-05,
+      "loss": 0.5337,
+      "step": 3094
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.189265251159668,
+      "learning_rate": 1.3902565880721222e-05,
+      "loss": 0.6505,
+      "step": 3095
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.799735069274902,
+      "learning_rate": 1.389736477115118e-05,
+      "loss": 0.8246,
+      "step": 3096
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.646000862121582,
+      "learning_rate": 1.3892163661581137e-05,
+      "loss": 0.6625,
+      "step": 3097
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 10.361005783081055,
+      "learning_rate": 1.3886962552011096e-05,
+      "loss": 0.8168,
+      "step": 3098
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.837032318115234,
+      "learning_rate": 1.3881761442441055e-05,
+      "loss": 0.6356,
+      "step": 3099
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.574671745300293,
+      "learning_rate": 1.3876560332871013e-05,
+      "loss": 0.6473,
+      "step": 3100
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.949514389038086,
+      "learning_rate": 1.3871359223300972e-05,
+      "loss": 0.6357,
+      "step": 3101
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.902454376220703,
+      "learning_rate": 1.386615811373093e-05,
+      "loss": 0.712,
+      "step": 3102
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.581043243408203,
+      "learning_rate": 1.3860957004160889e-05,
+      "loss": 0.5943,
+      "step": 3103
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.05706787109375,
+      "learning_rate": 1.3855755894590848e-05,
+      "loss": 0.6317,
+      "step": 3104
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.301756858825684,
+      "learning_rate": 1.3850554785020805e-05,
+      "loss": 0.9087,
+      "step": 3105
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.385939598083496,
+      "learning_rate": 1.3845353675450763e-05,
+      "loss": 0.695,
+      "step": 3106
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.710298538208008,
+      "learning_rate": 1.3840152565880722e-05,
+      "loss": 0.5101,
+      "step": 3107
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.577681541442871,
+      "learning_rate": 1.383495145631068e-05,
+      "loss": 0.7025,
+      "step": 3108
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.318740844726562,
+      "learning_rate": 1.3829750346740639e-05,
+      "loss": 0.6069,
+      "step": 3109
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.339838981628418,
+      "learning_rate": 1.3824549237170598e-05,
+      "loss": 0.631,
+      "step": 3110
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.78890609741211,
+      "learning_rate": 1.3819348127600556e-05,
+      "loss": 0.5131,
+      "step": 3111
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.147135734558105,
+      "learning_rate": 1.3814147018030515e-05,
+      "loss": 0.6237,
+      "step": 3112
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.352085113525391,
+      "learning_rate": 1.3808945908460472e-05,
+      "loss": 0.5122,
+      "step": 3113
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.169585227966309,
+      "learning_rate": 1.380374479889043e-05,
+      "loss": 0.5331,
+      "step": 3114
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.105509757995605,
+      "learning_rate": 1.3798543689320389e-05,
+      "loss": 0.7164,
+      "step": 3115
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.797449111938477,
+      "learning_rate": 1.3793342579750348e-05,
+      "loss": 0.5868,
+      "step": 3116
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.017821311950684,
+      "learning_rate": 1.3788141470180306e-05,
+      "loss": 0.6235,
+      "step": 3117
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.426405906677246,
+      "learning_rate": 1.3782940360610265e-05,
+      "loss": 0.6324,
+      "step": 3118
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.32983112335205,
+      "learning_rate": 1.3777739251040223e-05,
+      "loss": 0.6227,
+      "step": 3119
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.697662353515625,
+      "learning_rate": 1.377253814147018e-05,
+      "loss": 0.5047,
+      "step": 3120
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.846199035644531,
+      "learning_rate": 1.3767337031900139e-05,
+      "loss": 0.6305,
+      "step": 3121
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.274434089660645,
+      "learning_rate": 1.3762135922330098e-05,
+      "loss": 0.6046,
+      "step": 3122
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.113862991333008,
+      "learning_rate": 1.3756934812760056e-05,
+      "loss": 0.7744,
+      "step": 3123
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.846778869628906,
+      "learning_rate": 1.3751733703190015e-05,
+      "loss": 0.6111,
+      "step": 3124
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.266072273254395,
+      "learning_rate": 1.3746532593619973e-05,
+      "loss": 0.8029,
+      "step": 3125
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.580093383789062,
+      "learning_rate": 1.3741331484049932e-05,
+      "loss": 0.7517,
+      "step": 3126
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 9.583841323852539,
+      "learning_rate": 1.373613037447989e-05,
+      "loss": 0.7639,
+      "step": 3127
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.653943061828613,
+      "learning_rate": 1.3730929264909848e-05,
+      "loss": 0.5757,
+      "step": 3128
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.179404258728027,
+      "learning_rate": 1.3725728155339806e-05,
+      "loss": 0.6449,
+      "step": 3129
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.89759635925293,
+      "learning_rate": 1.3720527045769765e-05,
+      "loss": 0.5745,
+      "step": 3130
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.3915205001831055,
+      "learning_rate": 1.3715325936199723e-05,
+      "loss": 0.5836,
+      "step": 3131
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.020711421966553,
+      "learning_rate": 1.3710124826629682e-05,
+      "loss": 0.4437,
+      "step": 3132
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.676058769226074,
+      "learning_rate": 1.370492371705964e-05,
+      "loss": 0.8392,
+      "step": 3133
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.910389423370361,
+      "learning_rate": 1.36997226074896e-05,
+      "loss": 0.7113,
+      "step": 3134
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.776093482971191,
+      "learning_rate": 1.3694521497919558e-05,
+      "loss": 0.6709,
+      "step": 3135
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.24519157409668,
+      "learning_rate": 1.3689320388349515e-05,
+      "loss": 0.5232,
+      "step": 3136
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.157966613769531,
+      "learning_rate": 1.3684119278779473e-05,
+      "loss": 0.8893,
+      "step": 3137
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.025310516357422,
+      "learning_rate": 1.3678918169209432e-05,
+      "loss": 0.825,
+      "step": 3138
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.52361011505127,
+      "learning_rate": 1.367371705963939e-05,
+      "loss": 0.6487,
+      "step": 3139
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.310822486877441,
+      "learning_rate": 1.366851595006935e-05,
+      "loss": 0.7936,
+      "step": 3140
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.864860534667969,
+      "learning_rate": 1.3663314840499308e-05,
+      "loss": 0.7542,
+      "step": 3141
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.326322555541992,
+      "learning_rate": 1.3658113730929266e-05,
+      "loss": 0.8431,
+      "step": 3142
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.193901062011719,
+      "learning_rate": 1.3652912621359223e-05,
+      "loss": 0.6452,
+      "step": 3143
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.51846981048584,
+      "learning_rate": 1.3647711511789182e-05,
+      "loss": 0.4525,
+      "step": 3144
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.184155464172363,
+      "learning_rate": 1.364251040221914e-05,
+      "loss": 0.5839,
+      "step": 3145
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.386479377746582,
+      "learning_rate": 1.36373092926491e-05,
+      "loss": 0.5519,
+      "step": 3146
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.047306060791016,
+      "learning_rate": 1.3632108183079058e-05,
+      "loss": 0.6554,
+      "step": 3147
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.247803688049316,
+      "learning_rate": 1.3626907073509016e-05,
+      "loss": 0.497,
+      "step": 3148
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.3960676193237305,
+      "learning_rate": 1.3621705963938975e-05,
+      "loss": 0.4902,
+      "step": 3149
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.214529991149902,
+      "learning_rate": 1.3616504854368934e-05,
+      "loss": 0.8096,
+      "step": 3150
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.934833526611328,
+      "learning_rate": 1.361130374479889e-05,
+      "loss": 0.661,
+      "step": 3151
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.45018196105957,
+      "learning_rate": 1.360610263522885e-05,
+      "loss": 0.7565,
+      "step": 3152
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.136610984802246,
+      "learning_rate": 1.3600901525658808e-05,
+      "loss": 0.8152,
+      "step": 3153
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.896270751953125,
+      "learning_rate": 1.3595700416088766e-05,
+      "loss": 0.5763,
+      "step": 3154
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.917746067047119,
+      "learning_rate": 1.3590499306518725e-05,
+      "loss": 0.5957,
+      "step": 3155
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.643387794494629,
+      "learning_rate": 1.3585298196948684e-05,
+      "loss": 0.6063,
+      "step": 3156
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.03211784362793,
+      "learning_rate": 1.3580097087378642e-05,
+      "loss": 0.7963,
+      "step": 3157
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.67426872253418,
+      "learning_rate": 1.3574895977808601e-05,
+      "loss": 0.7143,
+      "step": 3158
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.064115524291992,
+      "learning_rate": 1.3569694868238558e-05,
+      "loss": 0.7221,
+      "step": 3159
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.176915168762207,
+      "learning_rate": 1.3564493758668516e-05,
+      "loss": 0.689,
+      "step": 3160
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.35881233215332,
+      "learning_rate": 1.3559292649098475e-05,
+      "loss": 0.6485,
+      "step": 3161
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.977132797241211,
+      "learning_rate": 1.3554091539528434e-05,
+      "loss": 0.758,
+      "step": 3162
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.958552360534668,
+      "learning_rate": 1.3548890429958392e-05,
+      "loss": 0.5394,
+      "step": 3163
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.640252113342285,
+      "learning_rate": 1.3543689320388351e-05,
+      "loss": 0.7376,
+      "step": 3164
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.000346183776855,
+      "learning_rate": 1.353848821081831e-05,
+      "loss": 0.6707,
+      "step": 3165
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.459858894348145,
+      "learning_rate": 1.3533287101248268e-05,
+      "loss": 0.6238,
+      "step": 3166
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.655112266540527,
+      "learning_rate": 1.3528085991678225e-05,
+      "loss": 0.8418,
+      "step": 3167
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.524639129638672,
+      "learning_rate": 1.3522884882108184e-05,
+      "loss": 0.8966,
+      "step": 3168
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.536056518554688,
+      "learning_rate": 1.3517683772538142e-05,
+      "loss": 0.5724,
+      "step": 3169
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.569376468658447,
+      "learning_rate": 1.3512482662968101e-05,
+      "loss": 0.4851,
+      "step": 3170
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.668380737304688,
+      "learning_rate": 1.350728155339806e-05,
+      "loss": 0.7955,
+      "step": 3171
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.850291728973389,
+      "learning_rate": 1.3502080443828018e-05,
+      "loss": 0.5657,
+      "step": 3172
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.974855422973633,
+      "learning_rate": 1.3496879334257977e-05,
+      "loss": 0.4701,
+      "step": 3173
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.310314178466797,
+      "learning_rate": 1.3491678224687934e-05,
+      "loss": 0.6633,
+      "step": 3174
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.99842643737793,
+      "learning_rate": 1.3486477115117892e-05,
+      "loss": 0.5157,
+      "step": 3175
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.8536787033081055,
+      "learning_rate": 1.348127600554785e-05,
+      "loss": 0.6515,
+      "step": 3176
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.571218490600586,
+      "learning_rate": 1.347607489597781e-05,
+      "loss": 0.7062,
+      "step": 3177
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.816609382629395,
+      "learning_rate": 1.3470873786407768e-05,
+      "loss": 0.7803,
+      "step": 3178
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.771000862121582,
+      "learning_rate": 1.3465672676837727e-05,
+      "loss": 0.5145,
+      "step": 3179
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.20881462097168,
+      "learning_rate": 1.3460471567267685e-05,
+      "loss": 0.7769,
+      "step": 3180
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.335771560668945,
+      "learning_rate": 1.3455270457697644e-05,
+      "loss": 0.7654,
+      "step": 3181
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.264165878295898,
+      "learning_rate": 1.34500693481276e-05,
+      "loss": 0.4955,
+      "step": 3182
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.893646717071533,
+      "learning_rate": 1.344486823855756e-05,
+      "loss": 0.4287,
+      "step": 3183
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.084233283996582,
+      "learning_rate": 1.3439667128987518e-05,
+      "loss": 0.6696,
+      "step": 3184
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.354082107543945,
+      "learning_rate": 1.3434466019417477e-05,
+      "loss": 0.7054,
+      "step": 3185
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.452290534973145,
+      "learning_rate": 1.3429264909847435e-05,
+      "loss": 0.7776,
+      "step": 3186
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.879586219787598,
+      "learning_rate": 1.3424063800277394e-05,
+      "loss": 0.8935,
+      "step": 3187
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.810629844665527,
+      "learning_rate": 1.3418862690707352e-05,
+      "loss": 0.5345,
+      "step": 3188
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.777983665466309,
+      "learning_rate": 1.3413661581137311e-05,
+      "loss": 0.5225,
+      "step": 3189
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.801258087158203,
+      "learning_rate": 1.3408460471567268e-05,
+      "loss": 0.7283,
+      "step": 3190
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.98735523223877,
+      "learning_rate": 1.3403259361997227e-05,
+      "loss": 0.7185,
+      "step": 3191
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.128756523132324,
+      "learning_rate": 1.3398058252427185e-05,
+      "loss": 0.533,
+      "step": 3192
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.759527206420898,
+      "learning_rate": 1.3392857142857144e-05,
+      "loss": 0.7412,
+      "step": 3193
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.774765014648438,
+      "learning_rate": 1.3387656033287102e-05,
+      "loss": 0.5096,
+      "step": 3194
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.103494644165039,
+      "learning_rate": 1.3382454923717061e-05,
+      "loss": 0.7819,
+      "step": 3195
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.797850608825684,
+      "learning_rate": 1.337725381414702e-05,
+      "loss": 0.5953,
+      "step": 3196
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.022561073303223,
+      "learning_rate": 1.3372052704576978e-05,
+      "loss": 0.8236,
+      "step": 3197
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.96264362335205,
+      "learning_rate": 1.3366851595006935e-05,
+      "loss": 0.5832,
+      "step": 3198
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.87700843811035,
+      "learning_rate": 1.3361650485436894e-05,
+      "loss": 0.4889,
+      "step": 3199
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.97100830078125,
+      "learning_rate": 1.3356449375866852e-05,
+      "loss": 0.6398,
+      "step": 3200
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.052594184875488,
+      "learning_rate": 1.3351248266296811e-05,
+      "loss": 0.5751,
+      "step": 3201
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.391582489013672,
+      "learning_rate": 1.334604715672677e-05,
+      "loss": 0.5433,
+      "step": 3202
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.51190185546875,
+      "learning_rate": 1.3340846047156728e-05,
+      "loss": 0.6965,
+      "step": 3203
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.706512451171875,
+      "learning_rate": 1.3335644937586687e-05,
+      "loss": 0.6303,
+      "step": 3204
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.477424621582031,
+      "learning_rate": 1.3330443828016644e-05,
+      "loss": 0.6091,
+      "step": 3205
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.573928833007812,
+      "learning_rate": 1.3325242718446602e-05,
+      "loss": 0.7083,
+      "step": 3206
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.108811378479004,
+      "learning_rate": 1.3320041608876561e-05,
+      "loss": 0.8475,
+      "step": 3207
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.808083534240723,
+      "learning_rate": 1.331484049930652e-05,
+      "loss": 0.7819,
+      "step": 3208
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.24623966217041,
+      "learning_rate": 1.3309639389736478e-05,
+      "loss": 0.5752,
+      "step": 3209
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.245308876037598,
+      "learning_rate": 1.3304438280166437e-05,
+      "loss": 0.5954,
+      "step": 3210
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.578156471252441,
+      "learning_rate": 1.3299237170596395e-05,
+      "loss": 0.5806,
+      "step": 3211
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.326817512512207,
+      "learning_rate": 1.3294036061026354e-05,
+      "loss": 0.6902,
+      "step": 3212
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.642638206481934,
+      "learning_rate": 1.3288834951456311e-05,
+      "loss": 0.7629,
+      "step": 3213
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.23896312713623,
+      "learning_rate": 1.328363384188627e-05,
+      "loss": 0.6544,
+      "step": 3214
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.844769477844238,
+      "learning_rate": 1.3278432732316228e-05,
+      "loss": 0.6938,
+      "step": 3215
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.294123649597168,
+      "learning_rate": 1.3273231622746187e-05,
+      "loss": 0.5018,
+      "step": 3216
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.1334381103515625,
+      "learning_rate": 1.3268030513176145e-05,
+      "loss": 0.5015,
+      "step": 3217
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.979654788970947,
+      "learning_rate": 1.3262829403606104e-05,
+      "loss": 0.4749,
+      "step": 3218
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.270721435546875,
+      "learning_rate": 1.3257628294036063e-05,
+      "loss": 0.7834,
+      "step": 3219
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.549599647521973,
+      "learning_rate": 1.3252427184466021e-05,
+      "loss": 0.8575,
+      "step": 3220
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.1569342613220215,
+      "learning_rate": 1.3247226074895978e-05,
+      "loss": 0.4635,
+      "step": 3221
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.752510070800781,
+      "learning_rate": 1.3242024965325937e-05,
+      "loss": 0.6067,
+      "step": 3222
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.62195110321045,
+      "learning_rate": 1.3236823855755895e-05,
+      "loss": 0.5209,
+      "step": 3223
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.855537414550781,
+      "learning_rate": 1.3231622746185854e-05,
+      "loss": 0.681,
+      "step": 3224
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.61518383026123,
+      "learning_rate": 1.3226421636615813e-05,
+      "loss": 0.5536,
+      "step": 3225
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.82557487487793,
+      "learning_rate": 1.3221220527045771e-05,
+      "loss": 0.5777,
+      "step": 3226
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.304746627807617,
+      "learning_rate": 1.321601941747573e-05,
+      "loss": 0.8864,
+      "step": 3227
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.240342140197754,
+      "learning_rate": 1.3210818307905689e-05,
+      "loss": 0.8896,
+      "step": 3228
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.33933162689209,
+      "learning_rate": 1.3205617198335645e-05,
+      "loss": 0.6234,
+      "step": 3229
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.451299667358398,
+      "learning_rate": 1.3200416088765604e-05,
+      "loss": 0.4929,
+      "step": 3230
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.518176078796387,
+      "learning_rate": 1.3195214979195563e-05,
+      "loss": 0.8773,
+      "step": 3231
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.250840187072754,
+      "learning_rate": 1.3190013869625521e-05,
+      "loss": 0.9472,
+      "step": 3232
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.420426368713379,
+      "learning_rate": 1.318481276005548e-05,
+      "loss": 0.3831,
+      "step": 3233
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.71928596496582,
+      "learning_rate": 1.3179611650485439e-05,
+      "loss": 0.457,
+      "step": 3234
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.879851341247559,
+      "learning_rate": 1.3174410540915397e-05,
+      "loss": 0.6205,
+      "step": 3235
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.821012496948242,
+      "learning_rate": 1.3169209431345354e-05,
+      "loss": 0.8812,
+      "step": 3236
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.714361190795898,
+      "learning_rate": 1.3164008321775313e-05,
+      "loss": 0.58,
+      "step": 3237
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.787755966186523,
+      "learning_rate": 1.3158807212205271e-05,
+      "loss": 0.6848,
+      "step": 3238
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.218979835510254,
+      "learning_rate": 1.315360610263523e-05,
+      "loss": 0.5704,
+      "step": 3239
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.556578636169434,
+      "learning_rate": 1.3148404993065188e-05,
+      "loss": 0.7625,
+      "step": 3240
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.397248268127441,
+      "learning_rate": 1.3143203883495147e-05,
+      "loss": 1.1804,
+      "step": 3241
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.811513900756836,
+      "learning_rate": 1.3138002773925106e-05,
+      "loss": 0.7236,
+      "step": 3242
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.960922241210938,
+      "learning_rate": 1.3132801664355064e-05,
+      "loss": 0.7219,
+      "step": 3243
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.706040382385254,
+      "learning_rate": 1.3127600554785021e-05,
+      "loss": 0.4918,
+      "step": 3244
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.039717674255371,
+      "learning_rate": 1.3122399445214978e-05,
+      "loss": 0.6704,
+      "step": 3245
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.182718276977539,
+      "learning_rate": 1.3117198335644937e-05,
+      "loss": 0.5042,
+      "step": 3246
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.197659492492676,
+      "learning_rate": 1.3111997226074895e-05,
+      "loss": 0.5094,
+      "step": 3247
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.222517490386963,
+      "learning_rate": 1.3106796116504854e-05,
+      "loss": 0.6181,
+      "step": 3248
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.936732292175293,
+      "learning_rate": 1.3101595006934813e-05,
+      "loss": 0.7587,
+      "step": 3249
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.923095703125,
+      "learning_rate": 1.3096393897364771e-05,
+      "loss": 0.5106,
+      "step": 3250
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.653781890869141,
+      "learning_rate": 1.309119278779473e-05,
+      "loss": 0.6519,
+      "step": 3251
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.703441619873047,
+      "learning_rate": 1.3085991678224687e-05,
+      "loss": 0.6234,
+      "step": 3252
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.103178024291992,
+      "learning_rate": 1.3080790568654645e-05,
+      "loss": 0.6633,
+      "step": 3253
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.820837020874023,
+      "learning_rate": 1.3075589459084604e-05,
+      "loss": 0.7283,
+      "step": 3254
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.6021575927734375,
+      "learning_rate": 1.3070388349514563e-05,
+      "loss": 0.5479,
+      "step": 3255
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.364289283752441,
+      "learning_rate": 1.3065187239944521e-05,
+      "loss": 0.6148,
+      "step": 3256
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.978642463684082,
+      "learning_rate": 1.305998613037448e-05,
+      "loss": 0.5667,
+      "step": 3257
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.253656387329102,
+      "learning_rate": 1.3054785020804438e-05,
+      "loss": 0.3743,
+      "step": 3258
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.11502742767334,
+      "learning_rate": 1.3049583911234397e-05,
+      "loss": 0.5366,
+      "step": 3259
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.481905937194824,
+      "learning_rate": 1.3044382801664354e-05,
+      "loss": 0.6532,
+      "step": 3260
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.31740951538086,
+      "learning_rate": 1.3039181692094313e-05,
+      "loss": 0.6937,
+      "step": 3261
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.315475463867188,
+      "learning_rate": 1.3033980582524271e-05,
+      "loss": 0.5893,
+      "step": 3262
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.803540229797363,
+      "learning_rate": 1.302877947295423e-05,
+      "loss": 0.6722,
+      "step": 3263
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.481358528137207,
+      "learning_rate": 1.3023578363384188e-05,
+      "loss": 0.8302,
+      "step": 3264
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.009539604187012,
+      "learning_rate": 1.3018377253814147e-05,
+      "loss": 0.6885,
+      "step": 3265
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.66769027709961,
+      "learning_rate": 1.3013176144244106e-05,
+      "loss": 0.5675,
+      "step": 3266
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 13.474137306213379,
+      "learning_rate": 1.3007975034674063e-05,
+      "loss": 1.068,
+      "step": 3267
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.782398223876953,
+      "learning_rate": 1.3002773925104021e-05,
+      "loss": 0.5218,
+      "step": 3268
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.455309867858887,
+      "learning_rate": 1.299757281553398e-05,
+      "loss": 0.5776,
+      "step": 3269
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.48913860321045,
+      "learning_rate": 1.2992371705963938e-05,
+      "loss": 0.6031,
+      "step": 3270
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.330061912536621,
+      "learning_rate": 1.2987170596393897e-05,
+      "loss": 0.8823,
+      "step": 3271
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.518062591552734,
+      "learning_rate": 1.2981969486823856e-05,
+      "loss": 0.5532,
+      "step": 3272
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.971010208129883,
+      "learning_rate": 1.2976768377253814e-05,
+      "loss": 0.5383,
+      "step": 3273
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.17977237701416,
+      "learning_rate": 1.2971567267683773e-05,
+      "loss": 0.4313,
+      "step": 3274
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.222477912902832,
+      "learning_rate": 1.296636615811373e-05,
+      "loss": 0.6069,
+      "step": 3275
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.239995002746582,
+      "learning_rate": 1.2961165048543688e-05,
+      "loss": 0.6185,
+      "step": 3276
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.611314296722412,
+      "learning_rate": 1.2955963938973647e-05,
+      "loss": 0.4708,
+      "step": 3277
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.776871681213379,
+      "learning_rate": 1.2950762829403606e-05,
+      "loss": 0.6307,
+      "step": 3278
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.94337272644043,
+      "learning_rate": 1.2945561719833564e-05,
+      "loss": 0.6272,
+      "step": 3279
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.526337623596191,
+      "learning_rate": 1.2940360610263523e-05,
+      "loss": 0.5818,
+      "step": 3280
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.629937171936035,
+      "learning_rate": 1.2935159500693481e-05,
+      "loss": 0.8628,
+      "step": 3281
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.363411903381348,
+      "learning_rate": 1.292995839112344e-05,
+      "loss": 0.5897,
+      "step": 3282
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.285347938537598,
+      "learning_rate": 1.2924757281553397e-05,
+      "loss": 0.7358,
+      "step": 3283
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.940401077270508,
+      "learning_rate": 1.2919556171983356e-05,
+      "loss": 0.67,
+      "step": 3284
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.628547668457031,
+      "learning_rate": 1.2914355062413314e-05,
+      "loss": 0.6646,
+      "step": 3285
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.4731364250183105,
+      "learning_rate": 1.2909153952843273e-05,
+      "loss": 0.505,
+      "step": 3286
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.678828239440918,
+      "learning_rate": 1.2903952843273231e-05,
+      "loss": 0.5102,
+      "step": 3287
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.39566421508789,
+      "learning_rate": 1.289875173370319e-05,
+      "loss": 0.5126,
+      "step": 3288
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.233772277832031,
+      "learning_rate": 1.2893550624133149e-05,
+      "loss": 0.726,
+      "step": 3289
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.781078338623047,
+      "learning_rate": 1.2888349514563106e-05,
+      "loss": 0.6678,
+      "step": 3290
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.619208335876465,
+      "learning_rate": 1.2883148404993064e-05,
+      "loss": 0.5556,
+      "step": 3291
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.677033424377441,
+      "learning_rate": 1.2877947295423023e-05,
+      "loss": 0.6105,
+      "step": 3292
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.058754920959473,
+      "learning_rate": 1.2872746185852981e-05,
+      "loss": 0.7183,
+      "step": 3293
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.221075057983398,
+      "learning_rate": 1.286754507628294e-05,
+      "loss": 0.6105,
+      "step": 3294
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.350085258483887,
+      "learning_rate": 1.2862343966712899e-05,
+      "loss": 0.6241,
+      "step": 3295
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.64791488647461,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.6775,
+      "step": 3296
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.914546966552734,
+      "learning_rate": 1.2851941747572816e-05,
+      "loss": 0.6852,
+      "step": 3297
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.83916187286377,
+      "learning_rate": 1.2846740638002773e-05,
+      "loss": 0.6052,
+      "step": 3298
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.124293327331543,
+      "learning_rate": 1.2841539528432731e-05,
+      "loss": 0.5249,
+      "step": 3299
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.530363082885742,
+      "learning_rate": 1.283633841886269e-05,
+      "loss": 0.8334,
+      "step": 3300
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.027034759521484,
+      "learning_rate": 1.2831137309292649e-05,
+      "loss": 0.721,
+      "step": 3301
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.782017707824707,
+      "learning_rate": 1.2825936199722607e-05,
+      "loss": 0.6296,
+      "step": 3302
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.470296859741211,
+      "learning_rate": 1.2820735090152566e-05,
+      "loss": 0.6031,
+      "step": 3303
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.057510375976562,
+      "learning_rate": 1.2815533980582524e-05,
+      "loss": 0.5854,
+      "step": 3304
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.124804496765137,
+      "learning_rate": 1.2810332871012483e-05,
+      "loss": 0.777,
+      "step": 3305
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.020991325378418,
+      "learning_rate": 1.280513176144244e-05,
+      "loss": 0.4758,
+      "step": 3306
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.20478630065918,
+      "learning_rate": 1.2799930651872399e-05,
+      "loss": 0.7558,
+      "step": 3307
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.818323135375977,
+      "learning_rate": 1.2794729542302357e-05,
+      "loss": 0.9272,
+      "step": 3308
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.305425643920898,
+      "learning_rate": 1.2789528432732316e-05,
+      "loss": 0.8073,
+      "step": 3309
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.237946510314941,
+      "learning_rate": 1.2784327323162274e-05,
+      "loss": 0.6448,
+      "step": 3310
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.600101470947266,
+      "learning_rate": 1.2779126213592233e-05,
+      "loss": 0.5585,
+      "step": 3311
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.868979454040527,
+      "learning_rate": 1.2773925104022192e-05,
+      "loss": 0.6092,
+      "step": 3312
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.428675651550293,
+      "learning_rate": 1.276872399445215e-05,
+      "loss": 0.7173,
+      "step": 3313
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.9117350578308105,
+      "learning_rate": 1.2763522884882107e-05,
+      "loss": 0.4295,
+      "step": 3314
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.226484775543213,
+      "learning_rate": 1.2758321775312066e-05,
+      "loss": 0.6261,
+      "step": 3315
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.818290710449219,
+      "learning_rate": 1.2753120665742024e-05,
+      "loss": 0.7706,
+      "step": 3316
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.797225952148438,
+      "learning_rate": 1.2747919556171983e-05,
+      "loss": 0.4912,
+      "step": 3317
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.115281105041504,
+      "learning_rate": 1.2742718446601942e-05,
+      "loss": 0.7738,
+      "step": 3318
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.177691459655762,
+      "learning_rate": 1.27375173370319e-05,
+      "loss": 0.7364,
+      "step": 3319
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.015717506408691,
+      "learning_rate": 1.2732316227461859e-05,
+      "loss": 0.5079,
+      "step": 3320
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.374719619750977,
+      "learning_rate": 1.2727115117891816e-05,
+      "loss": 0.8427,
+      "step": 3321
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.752613067626953,
+      "learning_rate": 1.2721914008321774e-05,
+      "loss": 0.6962,
+      "step": 3322
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.641569137573242,
+      "learning_rate": 1.2716712898751733e-05,
+      "loss": 0.4684,
+      "step": 3323
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.728381156921387,
+      "learning_rate": 1.2711511789181692e-05,
+      "loss": 0.7192,
+      "step": 3324
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.93523120880127,
+      "learning_rate": 1.270631067961165e-05,
+      "loss": 0.527,
+      "step": 3325
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.022779941558838,
+      "learning_rate": 1.2701109570041609e-05,
+      "loss": 0.5377,
+      "step": 3326
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.60744857788086,
+      "learning_rate": 1.2695908460471567e-05,
+      "loss": 0.4875,
+      "step": 3327
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.518454551696777,
+      "learning_rate": 1.2690707350901526e-05,
+      "loss": 0.8863,
+      "step": 3328
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.117900848388672,
+      "learning_rate": 1.2685506241331483e-05,
+      "loss": 0.5592,
+      "step": 3329
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.28317642211914,
+      "learning_rate": 1.2680305131761442e-05,
+      "loss": 0.7833,
+      "step": 3330
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.295029640197754,
+      "learning_rate": 1.26751040221914e-05,
+      "loss": 0.606,
+      "step": 3331
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.220582008361816,
+      "learning_rate": 1.2669902912621359e-05,
+      "loss": 0.6392,
+      "step": 3332
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.813886642456055,
+      "learning_rate": 1.2664701803051317e-05,
+      "loss": 0.5092,
+      "step": 3333
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.230911254882812,
+      "learning_rate": 1.2659500693481276e-05,
+      "loss": 0.5808,
+      "step": 3334
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.152475357055664,
+      "learning_rate": 1.2654299583911235e-05,
+      "loss": 0.7679,
+      "step": 3335
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.826388359069824,
+      "learning_rate": 1.2649098474341193e-05,
+      "loss": 0.6005,
+      "step": 3336
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.25456428527832,
+      "learning_rate": 1.264389736477115e-05,
+      "loss": 0.7825,
+      "step": 3337
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.204198837280273,
+      "learning_rate": 1.2638696255201109e-05,
+      "loss": 0.5255,
+      "step": 3338
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.048077583312988,
+      "learning_rate": 1.2633495145631067e-05,
+      "loss": 0.7554,
+      "step": 3339
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.058822631835938,
+      "learning_rate": 1.2628294036061026e-05,
+      "loss": 0.826,
+      "step": 3340
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.6530961990356445,
+      "learning_rate": 1.2623092926490985e-05,
+      "loss": 0.6666,
+      "step": 3341
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.169814109802246,
+      "learning_rate": 1.2617891816920943e-05,
+      "loss": 0.6133,
+      "step": 3342
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.144367218017578,
+      "learning_rate": 1.2612690707350902e-05,
+      "loss": 0.5434,
+      "step": 3343
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.725295066833496,
+      "learning_rate": 1.260748959778086e-05,
+      "loss": 0.7541,
+      "step": 3344
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.431962490081787,
+      "learning_rate": 1.2602288488210817e-05,
+      "loss": 0.4696,
+      "step": 3345
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.880017280578613,
+      "learning_rate": 1.2597087378640776e-05,
+      "loss": 0.6362,
+      "step": 3346
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.642802238464355,
+      "learning_rate": 1.2591886269070735e-05,
+      "loss": 0.667,
+      "step": 3347
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.225004196166992,
+      "learning_rate": 1.2586685159500693e-05,
+      "loss": 0.547,
+      "step": 3348
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.70583724975586,
+      "learning_rate": 1.2581484049930652e-05,
+      "loss": 0.6633,
+      "step": 3349
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.09362506866455,
+      "learning_rate": 1.257628294036061e-05,
+      "loss": 0.4018,
+      "step": 3350
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.921455383300781,
+      "learning_rate": 1.2571081830790569e-05,
+      "loss": 0.7769,
+      "step": 3351
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.384669303894043,
+      "learning_rate": 1.2565880721220526e-05,
+      "loss": 0.5591,
+      "step": 3352
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.18743896484375,
+      "learning_rate": 1.2560679611650485e-05,
+      "loss": 0.8822,
+      "step": 3353
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.5005464553833,
+      "learning_rate": 1.2555478502080443e-05,
+      "loss": 0.5424,
+      "step": 3354
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.731698989868164,
+      "learning_rate": 1.2550277392510402e-05,
+      "loss": 0.8609,
+      "step": 3355
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.633663177490234,
+      "learning_rate": 1.254507628294036e-05,
+      "loss": 0.4743,
+      "step": 3356
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.037214279174805,
+      "learning_rate": 1.2539875173370319e-05,
+      "loss": 0.6608,
+      "step": 3357
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.657254219055176,
+      "learning_rate": 1.2534674063800278e-05,
+      "loss": 0.7582,
+      "step": 3358
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.576579093933105,
+      "learning_rate": 1.2529472954230236e-05,
+      "loss": 0.6336,
+      "step": 3359
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.000418663024902,
+      "learning_rate": 1.2524271844660193e-05,
+      "loss": 0.4758,
+      "step": 3360
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.400201797485352,
+      "learning_rate": 1.2519070735090152e-05,
+      "loss": 0.7136,
+      "step": 3361
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.290253639221191,
+      "learning_rate": 1.251386962552011e-05,
+      "loss": 0.5929,
+      "step": 3362
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.538841724395752,
+      "learning_rate": 1.2508668515950069e-05,
+      "loss": 0.6295,
+      "step": 3363
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.604605674743652,
+      "learning_rate": 1.2503467406380028e-05,
+      "loss": 0.7263,
+      "step": 3364
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.836129188537598,
+      "learning_rate": 1.2498266296809986e-05,
+      "loss": 0.9373,
+      "step": 3365
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.182531356811523,
+      "learning_rate": 1.2493065187239945e-05,
+      "loss": 0.663,
+      "step": 3366
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.684839248657227,
+      "learning_rate": 1.2487864077669903e-05,
+      "loss": 0.6762,
+      "step": 3367
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.806381225585938,
+      "learning_rate": 1.248266296809986e-05,
+      "loss": 0.6027,
+      "step": 3368
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.823851585388184,
+      "learning_rate": 1.2477461858529819e-05,
+      "loss": 0.8483,
+      "step": 3369
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.49300479888916,
+      "learning_rate": 1.2472260748959778e-05,
+      "loss": 0.6788,
+      "step": 3370
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.943394660949707,
+      "learning_rate": 1.2467059639389736e-05,
+      "loss": 0.6758,
+      "step": 3371
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.057083129882812,
+      "learning_rate": 1.2461858529819695e-05,
+      "loss": 0.7364,
+      "step": 3372
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.279767036437988,
+      "learning_rate": 1.2456657420249653e-05,
+      "loss": 0.5627,
+      "step": 3373
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.79541301727295,
+      "learning_rate": 1.2451456310679612e-05,
+      "loss": 0.6563,
+      "step": 3374
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.939632415771484,
+      "learning_rate": 1.244625520110957e-05,
+      "loss": 0.5935,
+      "step": 3375
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.298787117004395,
+      "learning_rate": 1.2441054091539528e-05,
+      "loss": 0.6929,
+      "step": 3376
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.814843654632568,
+      "learning_rate": 1.2435852981969486e-05,
+      "loss": 0.4488,
+      "step": 3377
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.773331642150879,
+      "learning_rate": 1.2430651872399445e-05,
+      "loss": 0.6422,
+      "step": 3378
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.483409881591797,
+      "learning_rate": 1.2425450762829403e-05,
+      "loss": 0.5109,
+      "step": 3379
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.157276153564453,
+      "learning_rate": 1.2420249653259362e-05,
+      "loss": 0.5134,
+      "step": 3380
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.490983009338379,
+      "learning_rate": 1.241504854368932e-05,
+      "loss": 0.7722,
+      "step": 3381
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.698285102844238,
+      "learning_rate": 1.240984743411928e-05,
+      "loss": 0.5914,
+      "step": 3382
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.54973030090332,
+      "learning_rate": 1.2404646324549236e-05,
+      "loss": 0.8772,
+      "step": 3383
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.269288063049316,
+      "learning_rate": 1.2399445214979195e-05,
+      "loss": 0.7119,
+      "step": 3384
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.533912658691406,
+      "learning_rate": 1.2394244105409153e-05,
+      "loss": 0.7636,
+      "step": 3385
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.014190673828125,
+      "learning_rate": 1.2389042995839112e-05,
+      "loss": 0.6086,
+      "step": 3386
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.346126556396484,
+      "learning_rate": 1.238384188626907e-05,
+      "loss": 0.5297,
+      "step": 3387
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.2058687210083,
+      "learning_rate": 1.237864077669903e-05,
+      "loss": 0.6363,
+      "step": 3388
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.494256973266602,
+      "learning_rate": 1.2373439667128988e-05,
+      "loss": 0.6523,
+      "step": 3389
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.047470092773438,
+      "learning_rate": 1.2368238557558946e-05,
+      "loss": 0.5961,
+      "step": 3390
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.113043785095215,
+      "learning_rate": 1.2363037447988903e-05,
+      "loss": 0.589,
+      "step": 3391
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.633227348327637,
+      "learning_rate": 1.2357836338418862e-05,
+      "loss": 0.3843,
+      "step": 3392
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.998676300048828,
+      "learning_rate": 1.235263522884882e-05,
+      "loss": 0.7914,
+      "step": 3393
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.908716201782227,
+      "learning_rate": 1.234743411927878e-05,
+      "loss": 0.8029,
+      "step": 3394
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.146324157714844,
+      "learning_rate": 1.2342233009708738e-05,
+      "loss": 0.5743,
+      "step": 3395
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.30378246307373,
+      "learning_rate": 1.2337031900138696e-05,
+      "loss": 0.829,
+      "step": 3396
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.828285217285156,
+      "learning_rate": 1.2331830790568655e-05,
+      "loss": 0.7887,
+      "step": 3397
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.819519996643066,
+      "learning_rate": 1.2326629680998614e-05,
+      "loss": 0.7625,
+      "step": 3398
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.316707611083984,
+      "learning_rate": 1.232142857142857e-05,
+      "loss": 0.5993,
+      "step": 3399
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.369057655334473,
+      "learning_rate": 1.231622746185853e-05,
+      "loss": 0.5939,
+      "step": 3400
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.899922370910645,
+      "learning_rate": 1.2311026352288488e-05,
+      "loss": 0.6673,
+      "step": 3401
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.653634071350098,
+      "learning_rate": 1.2305825242718446e-05,
+      "loss": 0.6638,
+      "step": 3402
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.815055847167969,
+      "learning_rate": 1.2300624133148405e-05,
+      "loss": 0.7741,
+      "step": 3403
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.783088684082031,
+      "learning_rate": 1.2295423023578364e-05,
+      "loss": 0.6617,
+      "step": 3404
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.725252151489258,
+      "learning_rate": 1.2290221914008322e-05,
+      "loss": 0.8027,
+      "step": 3405
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.819101333618164,
+      "learning_rate": 1.228502080443828e-05,
+      "loss": 0.766,
+      "step": 3406
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.205863952636719,
+      "learning_rate": 1.2279819694868238e-05,
+      "loss": 0.7261,
+      "step": 3407
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.84185791015625,
+      "learning_rate": 1.2274618585298196e-05,
+      "loss": 0.636,
+      "step": 3408
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.925177574157715,
+      "learning_rate": 1.2269417475728155e-05,
+      "loss": 0.7128,
+      "step": 3409
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.783062934875488,
+      "learning_rate": 1.2264216366158114e-05,
+      "loss": 0.6054,
+      "step": 3410
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.891144752502441,
+      "learning_rate": 1.2259015256588072e-05,
+      "loss": 0.8274,
+      "step": 3411
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.877708911895752,
+      "learning_rate": 1.2253814147018031e-05,
+      "loss": 0.5413,
+      "step": 3412
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.261727333068848,
+      "learning_rate": 1.224861303744799e-05,
+      "loss": 0.7579,
+      "step": 3413
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.06403923034668,
+      "learning_rate": 1.2243411927877946e-05,
+      "loss": 0.6609,
+      "step": 3414
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.297375679016113,
+      "learning_rate": 1.2238210818307905e-05,
+      "loss": 0.4867,
+      "step": 3415
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.9816179275512695,
+      "learning_rate": 1.2233009708737864e-05,
+      "loss": 0.6308,
+      "step": 3416
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.227006912231445,
+      "learning_rate": 1.2227808599167822e-05,
+      "loss": 0.8586,
+      "step": 3417
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.93570327758789,
+      "learning_rate": 1.222260748959778e-05,
+      "loss": 0.651,
+      "step": 3418
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.26896333694458,
+      "learning_rate": 1.221740638002774e-05,
+      "loss": 0.5372,
+      "step": 3419
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.537076950073242,
+      "learning_rate": 1.2212205270457698e-05,
+      "loss": 0.6494,
+      "step": 3420
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.881966590881348,
+      "learning_rate": 1.2207004160887657e-05,
+      "loss": 0.4655,
+      "step": 3421
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.236825942993164,
+      "learning_rate": 1.2201803051317614e-05,
+      "loss": 0.7525,
+      "step": 3422
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.505415439605713,
+      "learning_rate": 1.2196601941747572e-05,
+      "loss": 0.5862,
+      "step": 3423
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.314382553100586,
+      "learning_rate": 1.219140083217753e-05,
+      "loss": 0.6883,
+      "step": 3424
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.982744216918945,
+      "learning_rate": 1.218619972260749e-05,
+      "loss": 0.5451,
+      "step": 3425
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.66835880279541,
+      "learning_rate": 1.2180998613037448e-05,
+      "loss": 0.5071,
+      "step": 3426
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.133866310119629,
+      "learning_rate": 1.2175797503467407e-05,
+      "loss": 0.6327,
+      "step": 3427
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.707072257995605,
+      "learning_rate": 1.2170596393897365e-05,
+      "loss": 0.6608,
+      "step": 3428
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.292767524719238,
+      "learning_rate": 1.2165395284327324e-05,
+      "loss": 0.5286,
+      "step": 3429
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.527750015258789,
+      "learning_rate": 1.216019417475728e-05,
+      "loss": 0.6617,
+      "step": 3430
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.18224811553955,
+      "learning_rate": 1.215499306518724e-05,
+      "loss": 0.7161,
+      "step": 3431
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.421167373657227,
+      "learning_rate": 1.2149791955617198e-05,
+      "loss": 0.6214,
+      "step": 3432
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.574601173400879,
+      "learning_rate": 1.2144590846047157e-05,
+      "loss": 0.7118,
+      "step": 3433
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.407282829284668,
+      "learning_rate": 1.2139389736477115e-05,
+      "loss": 0.628,
+      "step": 3434
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.04910945892334,
+      "learning_rate": 1.2134188626907074e-05,
+      "loss": 0.5489,
+      "step": 3435
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.698755264282227,
+      "learning_rate": 1.2128987517337032e-05,
+      "loss": 0.693,
+      "step": 3436
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.336718559265137,
+      "learning_rate": 1.212378640776699e-05,
+      "loss": 0.7739,
+      "step": 3437
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.792193412780762,
+      "learning_rate": 1.2118585298196948e-05,
+      "loss": 0.7112,
+      "step": 3438
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.5117645263671875,
+      "learning_rate": 1.2113384188626907e-05,
+      "loss": 0.6584,
+      "step": 3439
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.126577377319336,
+      "learning_rate": 1.2108183079056865e-05,
+      "loss": 0.8542,
+      "step": 3440
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.368196487426758,
+      "learning_rate": 1.2102981969486824e-05,
+      "loss": 0.7738,
+      "step": 3441
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.827323913574219,
+      "learning_rate": 1.2097780859916782e-05,
+      "loss": 0.5136,
+      "step": 3442
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.90331745147705,
+      "learning_rate": 1.2092579750346741e-05,
+      "loss": 0.6026,
+      "step": 3443
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.287485122680664,
+      "learning_rate": 1.20873786407767e-05,
+      "loss": 0.688,
+      "step": 3444
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.875972747802734,
+      "learning_rate": 1.2082177531206657e-05,
+      "loss": 0.4893,
+      "step": 3445
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.019347190856934,
+      "learning_rate": 1.2076976421636615e-05,
+      "loss": 0.6897,
+      "step": 3446
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.782350540161133,
+      "learning_rate": 1.2071775312066574e-05,
+      "loss": 0.5911,
+      "step": 3447
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.03481388092041,
+      "learning_rate": 1.2066574202496532e-05,
+      "loss": 0.6997,
+      "step": 3448
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.124323844909668,
+      "learning_rate": 1.2061373092926491e-05,
+      "loss": 0.6317,
+      "step": 3449
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.264427185058594,
+      "learning_rate": 1.205617198335645e-05,
+      "loss": 0.5914,
+      "step": 3450
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.684024810791016,
+      "learning_rate": 1.2050970873786408e-05,
+      "loss": 0.6278,
+      "step": 3451
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.028885841369629,
+      "learning_rate": 1.2045769764216367e-05,
+      "loss": 0.7506,
+      "step": 3452
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.493610382080078,
+      "learning_rate": 1.2040568654646324e-05,
+      "loss": 0.6384,
+      "step": 3453
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.00268840789795,
+      "learning_rate": 1.2035367545076282e-05,
+      "loss": 0.661,
+      "step": 3454
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.41315746307373,
+      "learning_rate": 1.2030166435506241e-05,
+      "loss": 0.6229,
+      "step": 3455
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.451761245727539,
+      "learning_rate": 1.20249653259362e-05,
+      "loss": 0.7774,
+      "step": 3456
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.10692310333252,
+      "learning_rate": 1.2019764216366158e-05,
+      "loss": 0.577,
+      "step": 3457
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.20718765258789,
+      "learning_rate": 1.2014563106796117e-05,
+      "loss": 0.5403,
+      "step": 3458
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.779866218566895,
+      "learning_rate": 1.2009361997226075e-05,
+      "loss": 0.6762,
+      "step": 3459
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.717893600463867,
+      "learning_rate": 1.2004160887656034e-05,
+      "loss": 0.5356,
+      "step": 3460
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.684358596801758,
+      "learning_rate": 1.1998959778085991e-05,
+      "loss": 0.7044,
+      "step": 3461
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.072762489318848,
+      "learning_rate": 1.199375866851595e-05,
+      "loss": 0.6568,
+      "step": 3462
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.137213706970215,
+      "learning_rate": 1.1988557558945908e-05,
+      "loss": 0.675,
+      "step": 3463
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.729426383972168,
+      "learning_rate": 1.1983356449375867e-05,
+      "loss": 0.4798,
+      "step": 3464
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.54182243347168,
+      "learning_rate": 1.1978155339805825e-05,
+      "loss": 0.6188,
+      "step": 3465
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.265158653259277,
+      "learning_rate": 1.1972954230235784e-05,
+      "loss": 0.4722,
+      "step": 3466
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.444613456726074,
+      "learning_rate": 1.1967753120665743e-05,
+      "loss": 0.5981,
+      "step": 3467
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.550773620605469,
+      "learning_rate": 1.19625520110957e-05,
+      "loss": 0.7394,
+      "step": 3468
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.011622428894043,
+      "learning_rate": 1.1957350901525658e-05,
+      "loss": 0.654,
+      "step": 3469
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.600987434387207,
+      "learning_rate": 1.1952149791955617e-05,
+      "loss": 0.4681,
+      "step": 3470
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.751709461212158,
+      "learning_rate": 1.1946948682385575e-05,
+      "loss": 0.4352,
+      "step": 3471
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.890069007873535,
+      "learning_rate": 1.1941747572815534e-05,
+      "loss": 0.5832,
+      "step": 3472
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.769165992736816,
+      "learning_rate": 1.1936546463245493e-05,
+      "loss": 0.5785,
+      "step": 3473
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.495832443237305,
+      "learning_rate": 1.1931345353675451e-05,
+      "loss": 0.608,
+      "step": 3474
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.069620132446289,
+      "learning_rate": 1.192614424410541e-05,
+      "loss": 0.5838,
+      "step": 3475
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.266063690185547,
+      "learning_rate": 1.1920943134535367e-05,
+      "loss": 0.5597,
+      "step": 3476
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.009483337402344,
+      "learning_rate": 1.1915742024965325e-05,
+      "loss": 0.7955,
+      "step": 3477
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.434066772460938,
+      "learning_rate": 1.1910540915395284e-05,
+      "loss": 0.6952,
+      "step": 3478
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.868752479553223,
+      "learning_rate": 1.1905339805825243e-05,
+      "loss": 0.6376,
+      "step": 3479
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.30820369720459,
+      "learning_rate": 1.1900138696255201e-05,
+      "loss": 0.6164,
+      "step": 3480
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.177927017211914,
+      "learning_rate": 1.189493758668516e-05,
+      "loss": 0.5027,
+      "step": 3481
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.11686897277832,
+      "learning_rate": 1.1889736477115118e-05,
+      "loss": 0.6275,
+      "step": 3482
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.721661567687988,
+      "learning_rate": 1.1884535367545077e-05,
+      "loss": 0.7413,
+      "step": 3483
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.21695327758789,
+      "learning_rate": 1.1879334257975034e-05,
+      "loss": 0.4253,
+      "step": 3484
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.297143936157227,
+      "learning_rate": 1.1874133148404993e-05,
+      "loss": 0.6898,
+      "step": 3485
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.627429008483887,
+      "learning_rate": 1.1868932038834951e-05,
+      "loss": 0.5485,
+      "step": 3486
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.69112777709961,
+      "learning_rate": 1.186373092926491e-05,
+      "loss": 0.65,
+      "step": 3487
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.569755554199219,
+      "learning_rate": 1.1858529819694868e-05,
+      "loss": 0.9069,
+      "step": 3488
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.87176513671875,
+      "learning_rate": 1.1853328710124827e-05,
+      "loss": 0.3893,
+      "step": 3489
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.387989044189453,
+      "learning_rate": 1.1848127600554786e-05,
+      "loss": 0.7239,
+      "step": 3490
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.046439170837402,
+      "learning_rate": 1.1842926490984743e-05,
+      "loss": 0.5647,
+      "step": 3491
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.028524398803711,
+      "learning_rate": 1.1837725381414701e-05,
+      "loss": 0.6242,
+      "step": 3492
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.455018043518066,
+      "learning_rate": 1.183252427184466e-05,
+      "loss": 0.5894,
+      "step": 3493
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.086151123046875,
+      "learning_rate": 1.1827323162274618e-05,
+      "loss": 0.5768,
+      "step": 3494
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.730131149291992,
+      "learning_rate": 1.1822122052704577e-05,
+      "loss": 0.529,
+      "step": 3495
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.635631084442139,
+      "learning_rate": 1.1816920943134536e-05,
+      "loss": 0.4215,
+      "step": 3496
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.223538398742676,
+      "learning_rate": 1.1811719833564494e-05,
+      "loss": 0.6018,
+      "step": 3497
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.408818244934082,
+      "learning_rate": 1.1806518723994453e-05,
+      "loss": 0.5799,
+      "step": 3498
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.170156478881836,
+      "learning_rate": 1.180131761442441e-05,
+      "loss": 0.532,
+      "step": 3499
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.452635765075684,
+      "learning_rate": 1.1796116504854368e-05,
+      "loss": 0.6306,
+      "step": 3500
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.388873100280762,
+      "learning_rate": 1.1790915395284327e-05,
+      "loss": 0.659,
+      "step": 3501
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.51634407043457,
+      "learning_rate": 1.1785714285714286e-05,
+      "loss": 0.7335,
+      "step": 3502
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.370936393737793,
+      "learning_rate": 1.1780513176144244e-05,
+      "loss": 0.7392,
+      "step": 3503
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.5121431350708,
+      "learning_rate": 1.1775312066574203e-05,
+      "loss": 0.6442,
+      "step": 3504
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.838701248168945,
+      "learning_rate": 1.1770110957004161e-05,
+      "loss": 0.637,
+      "step": 3505
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.166336059570312,
+      "learning_rate": 1.176490984743412e-05,
+      "loss": 0.5935,
+      "step": 3506
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.335641860961914,
+      "learning_rate": 1.1759708737864077e-05,
+      "loss": 0.6772,
+      "step": 3507
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.741118431091309,
+      "learning_rate": 1.1754507628294036e-05,
+      "loss": 0.5371,
+      "step": 3508
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.308605194091797,
+      "learning_rate": 1.1749306518723994e-05,
+      "loss": 0.7059,
+      "step": 3509
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.188108444213867,
+      "learning_rate": 1.1744105409153953e-05,
+      "loss": 0.5005,
+      "step": 3510
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.655853271484375,
+      "learning_rate": 1.1738904299583911e-05,
+      "loss": 0.6562,
+      "step": 3511
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.267602920532227,
+      "learning_rate": 1.173370319001387e-05,
+      "loss": 0.5737,
+      "step": 3512
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.47374439239502,
+      "learning_rate": 1.1728502080443829e-05,
+      "loss": 0.5738,
+      "step": 3513
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.685359954833984,
+      "learning_rate": 1.1723300970873787e-05,
+      "loss": 0.7454,
+      "step": 3514
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.031201362609863,
+      "learning_rate": 1.1718099861303744e-05,
+      "loss": 0.6413,
+      "step": 3515
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.244865894317627,
+      "learning_rate": 1.1712898751733703e-05,
+      "loss": 0.5847,
+      "step": 3516
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.360899925231934,
+      "learning_rate": 1.1707697642163661e-05,
+      "loss": 0.7551,
+      "step": 3517
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.36923599243164,
+      "learning_rate": 1.170249653259362e-05,
+      "loss": 0.6244,
+      "step": 3518
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.072619438171387,
+      "learning_rate": 1.1697295423023579e-05,
+      "loss": 0.4559,
+      "step": 3519
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.594614028930664,
+      "learning_rate": 1.1692094313453537e-05,
+      "loss": 0.6579,
+      "step": 3520
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.841490745544434,
+      "learning_rate": 1.1686893203883496e-05,
+      "loss": 0.7536,
+      "step": 3521
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.399360656738281,
+      "learning_rate": 1.1681692094313453e-05,
+      "loss": 0.536,
+      "step": 3522
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.977686882019043,
+      "learning_rate": 1.1676490984743411e-05,
+      "loss": 0.6032,
+      "step": 3523
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.666833877563477,
+      "learning_rate": 1.167128987517337e-05,
+      "loss": 0.8936,
+      "step": 3524
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.861553192138672,
+      "learning_rate": 1.1666088765603329e-05,
+      "loss": 0.6878,
+      "step": 3525
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.891753196716309,
+      "learning_rate": 1.1660887656033287e-05,
+      "loss": 0.6448,
+      "step": 3526
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.3847551345825195,
+      "learning_rate": 1.1655686546463246e-05,
+      "loss": 0.491,
+      "step": 3527
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.977781295776367,
+      "learning_rate": 1.1650485436893204e-05,
+      "loss": 0.4924,
+      "step": 3528
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.65844440460205,
+      "learning_rate": 1.1645284327323163e-05,
+      "loss": 0.6633,
+      "step": 3529
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.103177070617676,
+      "learning_rate": 1.164008321775312e-05,
+      "loss": 0.5177,
+      "step": 3530
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.166363716125488,
+      "learning_rate": 1.1634882108183079e-05,
+      "loss": 0.6939,
+      "step": 3531
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.191306114196777,
+      "learning_rate": 1.1629680998613037e-05,
+      "loss": 0.6833,
+      "step": 3532
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.125207901000977,
+      "learning_rate": 1.1624479889042996e-05,
+      "loss": 0.5724,
+      "step": 3533
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.680253982543945,
+      "learning_rate": 1.1619278779472954e-05,
+      "loss": 0.6397,
+      "step": 3534
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.068096160888672,
+      "learning_rate": 1.1614077669902913e-05,
+      "loss": 0.7004,
+      "step": 3535
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.147455215454102,
+      "learning_rate": 1.1608876560332872e-05,
+      "loss": 0.7704,
+      "step": 3536
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.140295028686523,
+      "learning_rate": 1.160367545076283e-05,
+      "loss": 0.5299,
+      "step": 3537
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.924756050109863,
+      "learning_rate": 1.1598474341192787e-05,
+      "loss": 0.6143,
+      "step": 3538
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.255847930908203,
+      "learning_rate": 1.1593273231622746e-05,
+      "loss": 0.7724,
+      "step": 3539
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.994891166687012,
+      "learning_rate": 1.1588072122052704e-05,
+      "loss": 0.7712,
+      "step": 3540
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.042202949523926,
+      "learning_rate": 1.1582871012482663e-05,
+      "loss": 0.7086,
+      "step": 3541
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.918682098388672,
+      "learning_rate": 1.1577669902912622e-05,
+      "loss": 0.7919,
+      "step": 3542
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.006922721862793,
+      "learning_rate": 1.157246879334258e-05,
+      "loss": 0.5436,
+      "step": 3543
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.029532432556152,
+      "learning_rate": 1.1567267683772539e-05,
+      "loss": 0.5785,
+      "step": 3544
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.815452575683594,
+      "learning_rate": 1.1562066574202498e-05,
+      "loss": 0.6029,
+      "step": 3545
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.521928787231445,
+      "learning_rate": 1.1556865464632454e-05,
+      "loss": 0.4728,
+      "step": 3546
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.802687644958496,
+      "learning_rate": 1.1551664355062413e-05,
+      "loss": 0.5002,
+      "step": 3547
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.549832344055176,
+      "learning_rate": 1.1546463245492372e-05,
+      "loss": 0.4838,
+      "step": 3548
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.020637512207031,
+      "learning_rate": 1.154126213592233e-05,
+      "loss": 0.6789,
+      "step": 3549
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.153058052062988,
+      "learning_rate": 1.1536061026352289e-05,
+      "loss": 0.6309,
+      "step": 3550
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.691498756408691,
+      "learning_rate": 1.1530859916782247e-05,
+      "loss": 0.8944,
+      "step": 3551
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.685355186462402,
+      "learning_rate": 1.1525658807212206e-05,
+      "loss": 0.5224,
+      "step": 3552
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.103001594543457,
+      "learning_rate": 1.1520457697642163e-05,
+      "loss": 0.7689,
+      "step": 3553
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.591734886169434,
+      "learning_rate": 1.1515256588072122e-05,
+      "loss": 0.6965,
+      "step": 3554
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.260899543762207,
+      "learning_rate": 1.151005547850208e-05,
+      "loss": 0.8013,
+      "step": 3555
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.457902908325195,
+      "learning_rate": 1.1504854368932039e-05,
+      "loss": 0.8162,
+      "step": 3556
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.536871910095215,
+      "learning_rate": 1.1499653259361997e-05,
+      "loss": 0.5857,
+      "step": 3557
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.442811012268066,
+      "learning_rate": 1.1494452149791956e-05,
+      "loss": 0.6008,
+      "step": 3558
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.429636001586914,
+      "learning_rate": 1.1489251040221915e-05,
+      "loss": 0.4538,
+      "step": 3559
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.116147994995117,
+      "learning_rate": 1.1484049930651873e-05,
+      "loss": 0.7847,
+      "step": 3560
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.193581581115723,
+      "learning_rate": 1.147884882108183e-05,
+      "loss": 0.7698,
+      "step": 3561
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.240629196166992,
+      "learning_rate": 1.1473647711511789e-05,
+      "loss": 0.4896,
+      "step": 3562
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.261916160583496,
+      "learning_rate": 1.1468446601941747e-05,
+      "loss": 0.7631,
+      "step": 3563
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.094813346862793,
+      "learning_rate": 1.1463245492371706e-05,
+      "loss": 0.6772,
+      "step": 3564
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.36781120300293,
+      "learning_rate": 1.1458044382801665e-05,
+      "loss": 0.5301,
+      "step": 3565
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.723400115966797,
+      "learning_rate": 1.1452843273231623e-05,
+      "loss": 0.5341,
+      "step": 3566
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.431586265563965,
+      "learning_rate": 1.1447642163661582e-05,
+      "loss": 0.6168,
+      "step": 3567
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.995532989501953,
+      "learning_rate": 1.144244105409154e-05,
+      "loss": 0.64,
+      "step": 3568
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.50583267211914,
+      "learning_rate": 1.1437239944521497e-05,
+      "loss": 0.5616,
+      "step": 3569
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.404993057250977,
+      "learning_rate": 1.1432038834951456e-05,
+      "loss": 0.5625,
+      "step": 3570
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.248979568481445,
+      "learning_rate": 1.1426837725381415e-05,
+      "loss": 0.6388,
+      "step": 3571
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.594337463378906,
+      "learning_rate": 1.1421636615811373e-05,
+      "loss": 0.552,
+      "step": 3572
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.200927734375,
+      "learning_rate": 1.1416435506241332e-05,
+      "loss": 0.7927,
+      "step": 3573
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.512529373168945,
+      "learning_rate": 1.141123439667129e-05,
+      "loss": 0.852,
+      "step": 3574
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.243284225463867,
+      "learning_rate": 1.1406033287101249e-05,
+      "loss": 0.6302,
+      "step": 3575
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.984554290771484,
+      "learning_rate": 1.1400832177531208e-05,
+      "loss": 0.7656,
+      "step": 3576
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.20220947265625,
+      "learning_rate": 1.1395631067961165e-05,
+      "loss": 0.5538,
+      "step": 3577
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.826508522033691,
+      "learning_rate": 1.1390429958391123e-05,
+      "loss": 0.7804,
+      "step": 3578
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.686850547790527,
+      "learning_rate": 1.1385228848821082e-05,
+      "loss": 0.5799,
+      "step": 3579
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.29490852355957,
+      "learning_rate": 1.138002773925104e-05,
+      "loss": 0.6902,
+      "step": 3580
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.662853240966797,
+      "learning_rate": 1.1374826629680999e-05,
+      "loss": 0.7414,
+      "step": 3581
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.042679786682129,
+      "learning_rate": 1.1369625520110958e-05,
+      "loss": 0.842,
+      "step": 3582
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.75692081451416,
+      "learning_rate": 1.1364424410540916e-05,
+      "loss": 0.7177,
+      "step": 3583
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.88834285736084,
+      "learning_rate": 1.1359223300970873e-05,
+      "loss": 0.6396,
+      "step": 3584
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.446885108947754,
+      "learning_rate": 1.1354022191400832e-05,
+      "loss": 0.6433,
+      "step": 3585
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.543100833892822,
+      "learning_rate": 1.134882108183079e-05,
+      "loss": 0.5982,
+      "step": 3586
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.828396797180176,
+      "learning_rate": 1.1343619972260749e-05,
+      "loss": 0.8974,
+      "step": 3587
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.285717964172363,
+      "learning_rate": 1.1338418862690708e-05,
+      "loss": 0.5766,
+      "step": 3588
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.639152526855469,
+      "learning_rate": 1.1333217753120666e-05,
+      "loss": 0.4938,
+      "step": 3589
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.222545623779297,
+      "learning_rate": 1.1328016643550625e-05,
+      "loss": 0.5983,
+      "step": 3590
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.696405410766602,
+      "learning_rate": 1.1322815533980584e-05,
+      "loss": 0.768,
+      "step": 3591
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.821869850158691,
+      "learning_rate": 1.131761442441054e-05,
+      "loss": 0.7045,
+      "step": 3592
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.875047206878662,
+      "learning_rate": 1.1312413314840499e-05,
+      "loss": 0.45,
+      "step": 3593
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.569849014282227,
+      "learning_rate": 1.1307212205270458e-05,
+      "loss": 0.7144,
+      "step": 3594
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.270258903503418,
+      "learning_rate": 1.1302011095700416e-05,
+      "loss": 0.5832,
+      "step": 3595
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.983715057373047,
+      "learning_rate": 1.1296809986130375e-05,
+      "loss": 0.7102,
+      "step": 3596
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.028542518615723,
+      "learning_rate": 1.1291608876560334e-05,
+      "loss": 0.6728,
+      "step": 3597
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.697208404541016,
+      "learning_rate": 1.1286407766990292e-05,
+      "loss": 0.6511,
+      "step": 3598
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.470470428466797,
+      "learning_rate": 1.128120665742025e-05,
+      "loss": 0.6174,
+      "step": 3599
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.231191635131836,
+      "learning_rate": 1.1276005547850208e-05,
+      "loss": 0.6538,
+      "step": 3600
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.326043128967285,
+      "learning_rate": 1.1270804438280166e-05,
+      "loss": 0.5879,
+      "step": 3601
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.403915405273438,
+      "learning_rate": 1.1265603328710125e-05,
+      "loss": 0.6233,
+      "step": 3602
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.689613342285156,
+      "learning_rate": 1.1260402219140083e-05,
+      "loss": 0.6852,
+      "step": 3603
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.571495056152344,
+      "learning_rate": 1.1255201109570042e-05,
+      "loss": 0.7501,
+      "step": 3604
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.72501277923584,
+      "learning_rate": 1.125e-05,
+      "loss": 0.5094,
+      "step": 3605
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.710362434387207,
+      "learning_rate": 1.124479889042996e-05,
+      "loss": 0.6329,
+      "step": 3606
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.942561149597168,
+      "learning_rate": 1.1239597780859916e-05,
+      "loss": 0.7947,
+      "step": 3607
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.509535789489746,
+      "learning_rate": 1.1234396671289875e-05,
+      "loss": 0.7834,
+      "step": 3608
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.8983154296875,
+      "learning_rate": 1.1229195561719833e-05,
+      "loss": 0.4874,
+      "step": 3609
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.3330817222595215,
+      "learning_rate": 1.1223994452149792e-05,
+      "loss": 0.4646,
+      "step": 3610
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.778616905212402,
+      "learning_rate": 1.121879334257975e-05,
+      "loss": 0.6575,
+      "step": 3611
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.926790237426758,
+      "learning_rate": 1.121359223300971e-05,
+      "loss": 0.7332,
+      "step": 3612
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.327296257019043,
+      "learning_rate": 1.1208391123439668e-05,
+      "loss": 0.445,
+      "step": 3613
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.130929470062256,
+      "learning_rate": 1.1203190013869627e-05,
+      "loss": 0.5671,
+      "step": 3614
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.104205131530762,
+      "learning_rate": 1.1197988904299583e-05,
+      "loss": 0.6573,
+      "step": 3615
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.228527069091797,
+      "learning_rate": 1.1192787794729542e-05,
+      "loss": 0.6798,
+      "step": 3616
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.22072696685791,
+      "learning_rate": 1.11875866851595e-05,
+      "loss": 0.7694,
+      "step": 3617
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.88055419921875,
+      "learning_rate": 1.118238557558946e-05,
+      "loss": 0.7935,
+      "step": 3618
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.035874366760254,
+      "learning_rate": 1.1177184466019418e-05,
+      "loss": 0.6434,
+      "step": 3619
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.355406761169434,
+      "learning_rate": 1.1171983356449377e-05,
+      "loss": 0.6735,
+      "step": 3620
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.939910888671875,
+      "learning_rate": 1.1166782246879335e-05,
+      "loss": 0.5604,
+      "step": 3621
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.974050045013428,
+      "learning_rate": 1.1161581137309294e-05,
+      "loss": 0.4122,
+      "step": 3622
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.096267700195312,
+      "learning_rate": 1.115638002773925e-05,
+      "loss": 0.6066,
+      "step": 3623
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.868278503417969,
+      "learning_rate": 1.115117891816921e-05,
+      "loss": 0.4096,
+      "step": 3624
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.247589111328125,
+      "learning_rate": 1.1145977808599168e-05,
+      "loss": 0.5116,
+      "step": 3625
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.73166561126709,
+      "learning_rate": 1.1140776699029126e-05,
+      "loss": 0.7188,
+      "step": 3626
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.259160995483398,
+      "learning_rate": 1.1135575589459085e-05,
+      "loss": 0.7073,
+      "step": 3627
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.486572265625,
+      "learning_rate": 1.1130374479889044e-05,
+      "loss": 0.6473,
+      "step": 3628
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.768749237060547,
+      "learning_rate": 1.1125173370319002e-05,
+      "loss": 0.7211,
+      "step": 3629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.778606414794922,
+      "learning_rate": 1.1119972260748961e-05,
+      "loss": 0.6517,
+      "step": 3630
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.894742012023926,
+      "learning_rate": 1.1114771151178918e-05,
+      "loss": 0.5298,
+      "step": 3631
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.711938858032227,
+      "learning_rate": 1.1109570041608876e-05,
+      "loss": 0.7372,
+      "step": 3632
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.552787780761719,
+      "learning_rate": 1.1104368932038835e-05,
+      "loss": 0.5397,
+      "step": 3633
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.881089210510254,
+      "learning_rate": 1.1099167822468794e-05,
+      "loss": 0.7381,
+      "step": 3634
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.116681098937988,
+      "learning_rate": 1.1093966712898752e-05,
+      "loss": 0.5901,
+      "step": 3635
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.299844741821289,
+      "learning_rate": 1.1088765603328711e-05,
+      "loss": 0.5885,
+      "step": 3636
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.800332069396973,
+      "learning_rate": 1.108356449375867e-05,
+      "loss": 0.5568,
+      "step": 3637
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.643264293670654,
+      "learning_rate": 1.1078363384188626e-05,
+      "loss": 0.4908,
+      "step": 3638
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.182077407836914,
+      "learning_rate": 1.1073162274618585e-05,
+      "loss": 0.684,
+      "step": 3639
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.34622859954834,
+      "learning_rate": 1.1067961165048544e-05,
+      "loss": 0.4782,
+      "step": 3640
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.783387184143066,
+      "learning_rate": 1.1062760055478502e-05,
+      "loss": 0.6717,
+      "step": 3641
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.37402057647705,
+      "learning_rate": 1.1057558945908461e-05,
+      "loss": 0.6357,
+      "step": 3642
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.225423812866211,
+      "learning_rate": 1.105235783633842e-05,
+      "loss": 0.7882,
+      "step": 3643
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.563211441040039,
+      "learning_rate": 1.1047156726768378e-05,
+      "loss": 0.6414,
+      "step": 3644
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.393747329711914,
+      "learning_rate": 1.1041955617198337e-05,
+      "loss": 0.5813,
+      "step": 3645
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.752588272094727,
+      "learning_rate": 1.1036754507628294e-05,
+      "loss": 0.5208,
+      "step": 3646
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.95848560333252,
+      "learning_rate": 1.1031553398058252e-05,
+      "loss": 0.5139,
+      "step": 3647
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.983621597290039,
+      "learning_rate": 1.1026352288488211e-05,
+      "loss": 0.6646,
+      "step": 3648
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.908858299255371,
+      "learning_rate": 1.102115117891817e-05,
+      "loss": 0.6388,
+      "step": 3649
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.389294624328613,
+      "learning_rate": 1.1015950069348128e-05,
+      "loss": 0.6467,
+      "step": 3650
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.358412265777588,
+      "learning_rate": 1.1010748959778087e-05,
+      "loss": 0.654,
+      "step": 3651
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.030925750732422,
+      "learning_rate": 1.1005547850208045e-05,
+      "loss": 0.5294,
+      "step": 3652
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.3793253898620605,
+      "learning_rate": 1.1000346740638004e-05,
+      "loss": 0.5134,
+      "step": 3653
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.642773628234863,
+      "learning_rate": 1.0995145631067961e-05,
+      "loss": 0.5925,
+      "step": 3654
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.854351997375488,
+      "learning_rate": 1.098994452149792e-05,
+      "loss": 0.7133,
+      "step": 3655
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.291631698608398,
+      "learning_rate": 1.0984743411927878e-05,
+      "loss": 0.665,
+      "step": 3656
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.464147567749023,
+      "learning_rate": 1.0979542302357837e-05,
+      "loss": 0.5185,
+      "step": 3657
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.160975456237793,
+      "learning_rate": 1.0974341192787795e-05,
+      "loss": 0.5335,
+      "step": 3658
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.739657402038574,
+      "learning_rate": 1.0969140083217754e-05,
+      "loss": 0.4728,
+      "step": 3659
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.343751907348633,
+      "learning_rate": 1.0963938973647713e-05,
+      "loss": 0.527,
+      "step": 3660
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.232293128967285,
+      "learning_rate": 1.0958737864077671e-05,
+      "loss": 0.6638,
+      "step": 3661
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.891478538513184,
+      "learning_rate": 1.0953536754507628e-05,
+      "loss": 0.6332,
+      "step": 3662
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.056166648864746,
+      "learning_rate": 1.0948335644937587e-05,
+      "loss": 0.5513,
+      "step": 3663
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.036486625671387,
+      "learning_rate": 1.0943134535367545e-05,
+      "loss": 0.7471,
+      "step": 3664
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.198938369750977,
+      "learning_rate": 1.0937933425797504e-05,
+      "loss": 0.8004,
+      "step": 3665
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.020915031433105,
+      "learning_rate": 1.0932732316227463e-05,
+      "loss": 0.4542,
+      "step": 3666
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.08714485168457,
+      "learning_rate": 1.0927531206657421e-05,
+      "loss": 0.6942,
+      "step": 3667
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.208344459533691,
+      "learning_rate": 1.092233009708738e-05,
+      "loss": 0.6458,
+      "step": 3668
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.525196075439453,
+      "learning_rate": 1.0917128987517337e-05,
+      "loss": 0.4995,
+      "step": 3669
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.619588851928711,
+      "learning_rate": 1.0911927877947295e-05,
+      "loss": 0.7359,
+      "step": 3670
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.844137191772461,
+      "learning_rate": 1.0906726768377254e-05,
+      "loss": 0.5403,
+      "step": 3671
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.2015962600708,
+      "learning_rate": 1.0901525658807212e-05,
+      "loss": 0.7288,
+      "step": 3672
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.40717601776123,
+      "learning_rate": 1.0896324549237171e-05,
+      "loss": 0.4806,
+      "step": 3673
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.877468109130859,
+      "learning_rate": 1.089112343966713e-05,
+      "loss": 0.5884,
+      "step": 3674
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.342888832092285,
+      "learning_rate": 1.0885922330097088e-05,
+      "loss": 0.4884,
+      "step": 3675
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.268511772155762,
+      "learning_rate": 1.0880721220527047e-05,
+      "loss": 0.5634,
+      "step": 3676
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.146133422851562,
+      "learning_rate": 1.0875520110957004e-05,
+      "loss": 0.4833,
+      "step": 3677
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.588089942932129,
+      "learning_rate": 1.0870319001386962e-05,
+      "loss": 0.6398,
+      "step": 3678
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.915640354156494,
+      "learning_rate": 1.0865117891816921e-05,
+      "loss": 0.545,
+      "step": 3679
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.336564064025879,
+      "learning_rate": 1.085991678224688e-05,
+      "loss": 0.5768,
+      "step": 3680
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.044526100158691,
+      "learning_rate": 1.0854715672676838e-05,
+      "loss": 0.4638,
+      "step": 3681
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.849113464355469,
+      "learning_rate": 1.0849514563106797e-05,
+      "loss": 0.5709,
+      "step": 3682
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.113792419433594,
+      "learning_rate": 1.0844313453536756e-05,
+      "loss": 0.5061,
+      "step": 3683
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.69697380065918,
+      "learning_rate": 1.0839112343966714e-05,
+      "loss": 0.6462,
+      "step": 3684
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.150206565856934,
+      "learning_rate": 1.0833911234396671e-05,
+      "loss": 0.4995,
+      "step": 3685
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.727845191955566,
+      "learning_rate": 1.082871012482663e-05,
+      "loss": 0.6414,
+      "step": 3686
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.676175117492676,
+      "learning_rate": 1.0823509015256588e-05,
+      "loss": 0.4986,
+      "step": 3687
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.558026313781738,
+      "learning_rate": 1.0818307905686547e-05,
+      "loss": 0.6964,
+      "step": 3688
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.190624237060547,
+      "learning_rate": 1.0813106796116506e-05,
+      "loss": 0.5079,
+      "step": 3689
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.798722267150879,
+      "learning_rate": 1.0807905686546464e-05,
+      "loss": 0.8488,
+      "step": 3690
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.869168281555176,
+      "learning_rate": 1.0802704576976423e-05,
+      "loss": 0.6087,
+      "step": 3691
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.318862915039062,
+      "learning_rate": 1.079750346740638e-05,
+      "loss": 0.7555,
+      "step": 3692
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.117623329162598,
+      "learning_rate": 1.0792302357836338e-05,
+      "loss": 0.5136,
+      "step": 3693
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.413625717163086,
+      "learning_rate": 1.0787101248266297e-05,
+      "loss": 0.5788,
+      "step": 3694
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.791483879089355,
+      "learning_rate": 1.0781900138696255e-05,
+      "loss": 0.816,
+      "step": 3695
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.891353607177734,
+      "learning_rate": 1.0776699029126214e-05,
+      "loss": 0.6677,
+      "step": 3696
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.62878704071045,
+      "learning_rate": 1.0771497919556173e-05,
+      "loss": 0.6727,
+      "step": 3697
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.21375560760498,
+      "learning_rate": 1.0766296809986131e-05,
+      "loss": 0.6297,
+      "step": 3698
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.615203857421875,
+      "learning_rate": 1.076109570041609e-05,
+      "loss": 0.5331,
+      "step": 3699
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.448115348815918,
+      "learning_rate": 1.0755894590846047e-05,
+      "loss": 0.5401,
+      "step": 3700
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.984742164611816,
+      "learning_rate": 1.0750693481276005e-05,
+      "loss": 0.4288,
+      "step": 3701
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.584757804870605,
+      "learning_rate": 1.0745492371705964e-05,
+      "loss": 0.6701,
+      "step": 3702
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.599209785461426,
+      "learning_rate": 1.0740291262135923e-05,
+      "loss": 0.5942,
+      "step": 3703
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.69420051574707,
+      "learning_rate": 1.0735090152565881e-05,
+      "loss": 0.6914,
+      "step": 3704
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.482925415039062,
+      "learning_rate": 1.072988904299584e-05,
+      "loss": 0.8294,
+      "step": 3705
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.759831428527832,
+      "learning_rate": 1.0724687933425799e-05,
+      "loss": 0.8748,
+      "step": 3706
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.843296051025391,
+      "learning_rate": 1.0719486823855757e-05,
+      "loss": 0.482,
+      "step": 3707
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.958112716674805,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.591,
+      "step": 3708
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.663594245910645,
+      "learning_rate": 1.0709084604715673e-05,
+      "loss": 0.5092,
+      "step": 3709
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.100013732910156,
+      "learning_rate": 1.0703883495145631e-05,
+      "loss": 0.6935,
+      "step": 3710
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.836501121520996,
+      "learning_rate": 1.069868238557559e-05,
+      "loss": 0.5892,
+      "step": 3711
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.510650634765625,
+      "learning_rate": 1.0693481276005549e-05,
+      "loss": 0.756,
+      "step": 3712
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.632122993469238,
+      "learning_rate": 1.0688280166435507e-05,
+      "loss": 0.6604,
+      "step": 3713
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.79010534286499,
+      "learning_rate": 1.0683079056865466e-05,
+      "loss": 0.6148,
+      "step": 3714
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.928653717041016,
+      "learning_rate": 1.0677877947295424e-05,
+      "loss": 0.6427,
+      "step": 3715
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.870379447937012,
+      "learning_rate": 1.0672676837725381e-05,
+      "loss": 0.5397,
+      "step": 3716
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.885458946228027,
+      "learning_rate": 1.066747572815534e-05,
+      "loss": 0.7308,
+      "step": 3717
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.405346870422363,
+      "learning_rate": 1.0662274618585299e-05,
+      "loss": 0.6719,
+      "step": 3718
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.470855712890625,
+      "learning_rate": 1.0657073509015257e-05,
+      "loss": 0.8511,
+      "step": 3719
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.136871337890625,
+      "learning_rate": 1.0651872399445216e-05,
+      "loss": 0.7114,
+      "step": 3720
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.95911693572998,
+      "learning_rate": 1.0646671289875174e-05,
+      "loss": 0.7894,
+      "step": 3721
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.460307121276855,
+      "learning_rate": 1.0641470180305133e-05,
+      "loss": 0.5617,
+      "step": 3722
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.345236778259277,
+      "learning_rate": 1.063626907073509e-05,
+      "loss": 0.5745,
+      "step": 3723
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.408091068267822,
+      "learning_rate": 1.0631067961165048e-05,
+      "loss": 0.5207,
+      "step": 3724
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.530776977539062,
+      "learning_rate": 1.0625866851595007e-05,
+      "loss": 0.7548,
+      "step": 3725
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.741684913635254,
+      "learning_rate": 1.0620665742024966e-05,
+      "loss": 0.6866,
+      "step": 3726
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.9609904289245605,
+      "learning_rate": 1.0615464632454924e-05,
+      "loss": 0.5385,
+      "step": 3727
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.640748977661133,
+      "learning_rate": 1.0610263522884883e-05,
+      "loss": 0.6413,
+      "step": 3728
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.044856548309326,
+      "learning_rate": 1.0605062413314842e-05,
+      "loss": 0.4771,
+      "step": 3729
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.04239559173584,
+      "learning_rate": 1.05998613037448e-05,
+      "loss": 0.5731,
+      "step": 3730
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.213220596313477,
+      "learning_rate": 1.0594660194174757e-05,
+      "loss": 0.7696,
+      "step": 3731
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.55842399597168,
+      "learning_rate": 1.0589459084604716e-05,
+      "loss": 0.973,
+      "step": 3732
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.33584976196289,
+      "learning_rate": 1.0584257975034674e-05,
+      "loss": 0.5804,
+      "step": 3733
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.602910995483398,
+      "learning_rate": 1.0579056865464633e-05,
+      "loss": 0.4569,
+      "step": 3734
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.89824390411377,
+      "learning_rate": 1.0573855755894592e-05,
+      "loss": 0.663,
+      "step": 3735
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.318037986755371,
+      "learning_rate": 1.056865464632455e-05,
+      "loss": 0.7819,
+      "step": 3736
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.847009658813477,
+      "learning_rate": 1.0563453536754509e-05,
+      "loss": 0.4731,
+      "step": 3737
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.718804359436035,
+      "learning_rate": 1.0558252427184467e-05,
+      "loss": 0.6826,
+      "step": 3738
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.743633270263672,
+      "learning_rate": 1.0553051317614424e-05,
+      "loss": 0.8298,
+      "step": 3739
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.704736709594727,
+      "learning_rate": 1.0547850208044383e-05,
+      "loss": 0.5291,
+      "step": 3740
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.027371406555176,
+      "learning_rate": 1.0542649098474342e-05,
+      "loss": 0.8748,
+      "step": 3741
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.2656888961792,
+      "learning_rate": 1.05374479889043e-05,
+      "loss": 0.7256,
+      "step": 3742
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.274510383605957,
+      "learning_rate": 1.0532246879334259e-05,
+      "loss": 0.7476,
+      "step": 3743
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.005496978759766,
+      "learning_rate": 1.0527045769764217e-05,
+      "loss": 0.3673,
+      "step": 3744
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.409915924072266,
+      "learning_rate": 1.0521844660194176e-05,
+      "loss": 0.665,
+      "step": 3745
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.769361972808838,
+      "learning_rate": 1.0516643550624135e-05,
+      "loss": 0.4393,
+      "step": 3746
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.031036376953125,
+      "learning_rate": 1.0511442441054091e-05,
+      "loss": 0.5036,
+      "step": 3747
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.715139389038086,
+      "learning_rate": 1.050624133148405e-05,
+      "loss": 0.615,
+      "step": 3748
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.19223403930664,
+      "learning_rate": 1.0501040221914009e-05,
+      "loss": 0.6877,
+      "step": 3749
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.433135986328125,
+      "learning_rate": 1.0495839112343967e-05,
+      "loss": 0.77,
+      "step": 3750
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.748095512390137,
+      "learning_rate": 1.0490638002773926e-05,
+      "loss": 0.559,
+      "step": 3751
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.55932331085205,
+      "learning_rate": 1.0485436893203885e-05,
+      "loss": 0.6394,
+      "step": 3752
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.064731597900391,
+      "learning_rate": 1.0480235783633843e-05,
+      "loss": 0.4067,
+      "step": 3753
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.389376640319824,
+      "learning_rate": 1.04750346740638e-05,
+      "loss": 0.7174,
+      "step": 3754
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.762127876281738,
+      "learning_rate": 1.0469833564493759e-05,
+      "loss": 0.674,
+      "step": 3755
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.292263984680176,
+      "learning_rate": 1.0464632454923717e-05,
+      "loss": 0.6647,
+      "step": 3756
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.219281196594238,
+      "learning_rate": 1.0459431345353676e-05,
+      "loss": 0.6436,
+      "step": 3757
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.640595436096191,
+      "learning_rate": 1.0454230235783635e-05,
+      "loss": 0.7122,
+      "step": 3758
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.105461120605469,
+      "learning_rate": 1.0449029126213593e-05,
+      "loss": 0.624,
+      "step": 3759
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.169806480407715,
+      "learning_rate": 1.0443828016643552e-05,
+      "loss": 0.4663,
+      "step": 3760
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.900885581970215,
+      "learning_rate": 1.043862690707351e-05,
+      "loss": 0.5969,
+      "step": 3761
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.106913566589355,
+      "learning_rate": 1.0433425797503467e-05,
+      "loss": 0.5015,
+      "step": 3762
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.404703140258789,
+      "learning_rate": 1.0428224687933426e-05,
+      "loss": 0.7342,
+      "step": 3763
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.958085060119629,
+      "learning_rate": 1.0423023578363385e-05,
+      "loss": 0.6326,
+      "step": 3764
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.067142486572266,
+      "learning_rate": 1.0417822468793343e-05,
+      "loss": 0.6356,
+      "step": 3765
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.413936614990234,
+      "learning_rate": 1.0412621359223302e-05,
+      "loss": 0.7239,
+      "step": 3766
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.022473335266113,
+      "learning_rate": 1.040742024965326e-05,
+      "loss": 0.7344,
+      "step": 3767
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.348723411560059,
+      "learning_rate": 1.0402219140083219e-05,
+      "loss": 0.5924,
+      "step": 3768
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.624251842498779,
+      "learning_rate": 1.0397018030513178e-05,
+      "loss": 0.5081,
+      "step": 3769
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.908666133880615,
+      "learning_rate": 1.0391816920943134e-05,
+      "loss": 0.4802,
+      "step": 3770
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.758859634399414,
+      "learning_rate": 1.0386615811373093e-05,
+      "loss": 1.0064,
+      "step": 3771
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.225369453430176,
+      "learning_rate": 1.0381414701803052e-05,
+      "loss": 0.7325,
+      "step": 3772
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.708783149719238,
+      "learning_rate": 1.037621359223301e-05,
+      "loss": 0.8194,
+      "step": 3773
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.271992683410645,
+      "learning_rate": 1.0371012482662969e-05,
+      "loss": 0.5436,
+      "step": 3774
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.526802062988281,
+      "learning_rate": 1.0365811373092928e-05,
+      "loss": 0.7166,
+      "step": 3775
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.661243438720703,
+      "learning_rate": 1.0360610263522886e-05,
+      "loss": 0.4977,
+      "step": 3776
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.435197830200195,
+      "learning_rate": 1.0355409153952845e-05,
+      "loss": 0.5368,
+      "step": 3777
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.229312896728516,
+      "learning_rate": 1.0350208044382802e-05,
+      "loss": 0.5095,
+      "step": 3778
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.75640869140625,
+      "learning_rate": 1.034500693481276e-05,
+      "loss": 0.6574,
+      "step": 3779
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.211320877075195,
+      "learning_rate": 1.0339805825242719e-05,
+      "loss": 0.7689,
+      "step": 3780
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.575648307800293,
+      "learning_rate": 1.0334604715672678e-05,
+      "loss": 0.666,
+      "step": 3781
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.290717124938965,
+      "learning_rate": 1.0329403606102636e-05,
+      "loss": 0.5938,
+      "step": 3782
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.585853576660156,
+      "learning_rate": 1.0324202496532595e-05,
+      "loss": 0.5516,
+      "step": 3783
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.661524772644043,
+      "learning_rate": 1.0319001386962553e-05,
+      "loss": 0.6654,
+      "step": 3784
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.59035587310791,
+      "learning_rate": 1.031380027739251e-05,
+      "loss": 0.5843,
+      "step": 3785
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.972683429718018,
+      "learning_rate": 1.0308599167822469e-05,
+      "loss": 0.5542,
+      "step": 3786
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.60693073272705,
+      "learning_rate": 1.0303398058252428e-05,
+      "loss": 0.6036,
+      "step": 3787
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.217698097229004,
+      "learning_rate": 1.0298196948682386e-05,
+      "loss": 0.6799,
+      "step": 3788
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.603872299194336,
+      "learning_rate": 1.0292995839112345e-05,
+      "loss": 0.5258,
+      "step": 3789
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.252287864685059,
+      "learning_rate": 1.0287794729542303e-05,
+      "loss": 0.6554,
+      "step": 3790
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.290839195251465,
+      "learning_rate": 1.0282593619972262e-05,
+      "loss": 0.6299,
+      "step": 3791
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.930623054504395,
+      "learning_rate": 1.027739251040222e-05,
+      "loss": 0.5216,
+      "step": 3792
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.627785682678223,
+      "learning_rate": 1.0272191400832177e-05,
+      "loss": 0.5306,
+      "step": 3793
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.746150016784668,
+      "learning_rate": 1.0266990291262136e-05,
+      "loss": 0.7397,
+      "step": 3794
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.772236824035645,
+      "learning_rate": 1.0261789181692095e-05,
+      "loss": 0.7148,
+      "step": 3795
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.7358574867248535,
+      "learning_rate": 1.0256588072122053e-05,
+      "loss": 0.5583,
+      "step": 3796
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.95777416229248,
+      "learning_rate": 1.0251386962552012e-05,
+      "loss": 0.6133,
+      "step": 3797
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.890229225158691,
+      "learning_rate": 1.024618585298197e-05,
+      "loss": 0.6558,
+      "step": 3798
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.55774211883545,
+      "learning_rate": 1.024098474341193e-05,
+      "loss": 0.6385,
+      "step": 3799
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.218329429626465,
+      "learning_rate": 1.0235783633841888e-05,
+      "loss": 0.591,
+      "step": 3800
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.791455268859863,
+      "learning_rate": 1.0230582524271845e-05,
+      "loss": 0.4495,
+      "step": 3801
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.479294776916504,
+      "learning_rate": 1.0225381414701803e-05,
+      "loss": 0.5744,
+      "step": 3802
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.346686363220215,
+      "learning_rate": 1.0220180305131762e-05,
+      "loss": 0.8617,
+      "step": 3803
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.561738967895508,
+      "learning_rate": 1.021497919556172e-05,
+      "loss": 0.6714,
+      "step": 3804
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.492154121398926,
+      "learning_rate": 1.0209778085991679e-05,
+      "loss": 0.6339,
+      "step": 3805
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.612784385681152,
+      "learning_rate": 1.0204576976421638e-05,
+      "loss": 0.5525,
+      "step": 3806
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.790363311767578,
+      "learning_rate": 1.0199375866851596e-05,
+      "loss": 0.6043,
+      "step": 3807
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.476125717163086,
+      "learning_rate": 1.0194174757281553e-05,
+      "loss": 0.4484,
+      "step": 3808
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.046899795532227,
+      "learning_rate": 1.0188973647711512e-05,
+      "loss": 0.6605,
+      "step": 3809
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.054926872253418,
+      "learning_rate": 1.018377253814147e-05,
+      "loss": 0.6205,
+      "step": 3810
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.164474487304688,
+      "learning_rate": 1.0178571428571429e-05,
+      "loss": 0.4315,
+      "step": 3811
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.537627220153809,
+      "learning_rate": 1.0173370319001388e-05,
+      "loss": 0.561,
+      "step": 3812
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.928399085998535,
+      "learning_rate": 1.0168169209431346e-05,
+      "loss": 0.7649,
+      "step": 3813
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.27289867401123,
+      "learning_rate": 1.0162968099861305e-05,
+      "loss": 0.7146,
+      "step": 3814
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.742684841156006,
+      "learning_rate": 1.0157766990291264e-05,
+      "loss": 0.3662,
+      "step": 3815
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.036755561828613,
+      "learning_rate": 1.015256588072122e-05,
+      "loss": 0.4009,
+      "step": 3816
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.657604217529297,
+      "learning_rate": 1.0147364771151179e-05,
+      "loss": 0.5665,
+      "step": 3817
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.845881462097168,
+      "learning_rate": 1.0142163661581138e-05,
+      "loss": 0.603,
+      "step": 3818
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.628456115722656,
+      "learning_rate": 1.0136962552011096e-05,
+      "loss": 0.6826,
+      "step": 3819
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.19565486907959,
+      "learning_rate": 1.0131761442441055e-05,
+      "loss": 0.4351,
+      "step": 3820
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.329957008361816,
+      "learning_rate": 1.0126560332871014e-05,
+      "loss": 0.6847,
+      "step": 3821
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.708224296569824,
+      "learning_rate": 1.0121359223300972e-05,
+      "loss": 0.7536,
+      "step": 3822
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.138981819152832,
+      "learning_rate": 1.011615811373093e-05,
+      "loss": 0.8125,
+      "step": 3823
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.936006546020508,
+      "learning_rate": 1.0110957004160888e-05,
+      "loss": 0.6008,
+      "step": 3824
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.169323921203613,
+      "learning_rate": 1.0105755894590846e-05,
+      "loss": 0.6365,
+      "step": 3825
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.319169998168945,
+      "learning_rate": 1.0100554785020805e-05,
+      "loss": 0.5712,
+      "step": 3826
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.016650199890137,
+      "learning_rate": 1.0095353675450764e-05,
+      "loss": 0.6795,
+      "step": 3827
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.766692161560059,
+      "learning_rate": 1.0090152565880722e-05,
+      "loss": 0.5196,
+      "step": 3828
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.154924392700195,
+      "learning_rate": 1.008495145631068e-05,
+      "loss": 0.6859,
+      "step": 3829
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.076983451843262,
+      "learning_rate": 1.007975034674064e-05,
+      "loss": 0.5781,
+      "step": 3830
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.01843547821045,
+      "learning_rate": 1.0074549237170598e-05,
+      "loss": 0.7025,
+      "step": 3831
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.749052047729492,
+      "learning_rate": 1.0069348127600555e-05,
+      "loss": 0.588,
+      "step": 3832
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.127371788024902,
+      "learning_rate": 1.0064147018030514e-05,
+      "loss": 0.4954,
+      "step": 3833
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.212437629699707,
+      "learning_rate": 1.0058945908460472e-05,
+      "loss": 0.7786,
+      "step": 3834
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.40681266784668,
+      "learning_rate": 1.005374479889043e-05,
+      "loss": 0.6521,
+      "step": 3835
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.591726303100586,
+      "learning_rate": 1.004854368932039e-05,
+      "loss": 0.5681,
+      "step": 3836
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.678763389587402,
+      "learning_rate": 1.0043342579750348e-05,
+      "loss": 0.493,
+      "step": 3837
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.876646041870117,
+      "learning_rate": 1.0038141470180307e-05,
+      "loss": 0.6979,
+      "step": 3838
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.30510139465332,
+      "learning_rate": 1.0032940360610264e-05,
+      "loss": 0.4975,
+      "step": 3839
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.30159854888916,
+      "learning_rate": 1.0027739251040222e-05,
+      "loss": 0.6134,
+      "step": 3840
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.826953887939453,
+      "learning_rate": 1.002253814147018e-05,
+      "loss": 0.5057,
+      "step": 3841
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.066223621368408,
+      "learning_rate": 1.001733703190014e-05,
+      "loss": 0.4942,
+      "step": 3842
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.253846168518066,
+      "learning_rate": 1.0012135922330098e-05,
+      "loss": 0.663,
+      "step": 3843
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.306096076965332,
+      "learning_rate": 1.0006934812760057e-05,
+      "loss": 0.5623,
+      "step": 3844
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.658204078674316,
+      "learning_rate": 1.0001733703190015e-05,
+      "loss": 0.4769,
+      "step": 3845
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.005504608154297,
+      "learning_rate": 9.996532593619974e-06,
+      "loss": 0.5233,
+      "step": 3846
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.51896858215332,
+      "learning_rate": 9.99133148404993e-06,
+      "loss": 0.6273,
+      "step": 3847
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.959982872009277,
+      "learning_rate": 9.98613037447989e-06,
+      "loss": 0.6043,
+      "step": 3848
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.147887229919434,
+      "learning_rate": 9.980929264909848e-06,
+      "loss": 0.644,
+      "step": 3849
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.5405912399292,
+      "learning_rate": 9.975728155339807e-06,
+      "loss": 0.4971,
+      "step": 3850
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.598784446716309,
+      "learning_rate": 9.970527045769765e-06,
+      "loss": 0.4212,
+      "step": 3851
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.22982406616211,
+      "learning_rate": 9.965325936199724e-06,
+      "loss": 0.4829,
+      "step": 3852
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.149472713470459,
+      "learning_rate": 9.960124826629682e-06,
+      "loss": 0.456,
+      "step": 3853
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.093696594238281,
+      "learning_rate": 9.954923717059641e-06,
+      "loss": 0.5975,
+      "step": 3854
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.085118293762207,
+      "learning_rate": 9.949722607489598e-06,
+      "loss": 0.7179,
+      "step": 3855
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.900604248046875,
+      "learning_rate": 9.944521497919557e-06,
+      "loss": 0.775,
+      "step": 3856
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.482162475585938,
+      "learning_rate": 9.939320388349515e-06,
+      "loss": 0.6843,
+      "step": 3857
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.879345893859863,
+      "learning_rate": 9.934119278779474e-06,
+      "loss": 0.899,
+      "step": 3858
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.699666976928711,
+      "learning_rate": 9.928918169209432e-06,
+      "loss": 0.4956,
+      "step": 3859
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.208638191223145,
+      "learning_rate": 9.923717059639391e-06,
+      "loss": 0.4696,
+      "step": 3860
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.678022384643555,
+      "learning_rate": 9.91851595006935e-06,
+      "loss": 0.6784,
+      "step": 3861
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.46440315246582,
+      "learning_rate": 9.913314840499308e-06,
+      "loss": 0.4424,
+      "step": 3862
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.156517028808594,
+      "learning_rate": 9.908113730929265e-06,
+      "loss": 0.5818,
+      "step": 3863
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.908573150634766,
+      "learning_rate": 9.902912621359224e-06,
+      "loss": 0.4736,
+      "step": 3864
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.710371971130371,
+      "learning_rate": 9.897711511789182e-06,
+      "loss": 0.5954,
+      "step": 3865
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.852016925811768,
+      "learning_rate": 9.892510402219141e-06,
+      "loss": 0.6296,
+      "step": 3866
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.253012657165527,
+      "learning_rate": 9.8873092926491e-06,
+      "loss": 0.5246,
+      "step": 3867
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.992013931274414,
+      "learning_rate": 9.882108183079058e-06,
+      "loss": 0.7673,
+      "step": 3868
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.99361515045166,
+      "learning_rate": 9.876907073509017e-06,
+      "loss": 0.6685,
+      "step": 3869
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.311062812805176,
+      "learning_rate": 9.871705963938974e-06,
+      "loss": 0.887,
+      "step": 3870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.95138168334961,
+      "learning_rate": 9.866504854368932e-06,
+      "loss": 0.7881,
+      "step": 3871
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.49193000793457,
+      "learning_rate": 9.861303744798891e-06,
+      "loss": 0.6789,
+      "step": 3872
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.143805027008057,
+      "learning_rate": 9.85610263522885e-06,
+      "loss": 0.3395,
+      "step": 3873
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.873164653778076,
+      "learning_rate": 9.850901525658808e-06,
+      "loss": 0.4198,
+      "step": 3874
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.988961219787598,
+      "learning_rate": 9.845700416088767e-06,
+      "loss": 0.6284,
+      "step": 3875
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.002625465393066,
+      "learning_rate": 9.840499306518725e-06,
+      "loss": 0.6776,
+      "step": 3876
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.847956657409668,
+      "learning_rate": 9.835298196948684e-06,
+      "loss": 0.3871,
+      "step": 3877
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.774800300598145,
+      "learning_rate": 9.830097087378641e-06,
+      "loss": 0.7414,
+      "step": 3878
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.436928749084473,
+      "learning_rate": 9.8248959778086e-06,
+      "loss": 0.6903,
+      "step": 3879
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.175318717956543,
+      "learning_rate": 9.819694868238558e-06,
+      "loss": 0.4744,
+      "step": 3880
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.338264465332031,
+      "learning_rate": 9.814493758668517e-06,
+      "loss": 0.6609,
+      "step": 3881
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.019375801086426,
+      "learning_rate": 9.809292649098475e-06,
+      "loss": 0.6138,
+      "step": 3882
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.326991081237793,
+      "learning_rate": 9.804091539528434e-06,
+      "loss": 0.7161,
+      "step": 3883
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.768942832946777,
+      "learning_rate": 9.798890429958393e-06,
+      "loss": 0.6053,
+      "step": 3884
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.115130424499512,
+      "learning_rate": 9.793689320388351e-06,
+      "loss": 0.6684,
+      "step": 3885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.39709758758545,
+      "learning_rate": 9.788488210818308e-06,
+      "loss": 0.5602,
+      "step": 3886
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.34122371673584,
+      "learning_rate": 9.783287101248267e-06,
+      "loss": 0.8827,
+      "step": 3887
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.499935150146484,
+      "learning_rate": 9.778085991678225e-06,
+      "loss": 0.5824,
+      "step": 3888
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.745086669921875,
+      "learning_rate": 9.772884882108184e-06,
+      "loss": 0.4939,
+      "step": 3889
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.361988067626953,
+      "learning_rate": 9.767683772538143e-06,
+      "loss": 0.7093,
+      "step": 3890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.905316352844238,
+      "learning_rate": 9.762482662968101e-06,
+      "loss": 0.5959,
+      "step": 3891
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.968365669250488,
+      "learning_rate": 9.75728155339806e-06,
+      "loss": 0.7744,
+      "step": 3892
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.02449369430542,
+      "learning_rate": 9.752080443828017e-06,
+      "loss": 0.3579,
+      "step": 3893
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.883745193481445,
+      "learning_rate": 9.746879334257975e-06,
+      "loss": 0.6601,
+      "step": 3894
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.10117244720459,
+      "learning_rate": 9.741678224687934e-06,
+      "loss": 0.5062,
+      "step": 3895
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.180943489074707,
+      "learning_rate": 9.736477115117893e-06,
+      "loss": 0.612,
+      "step": 3896
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.975778579711914,
+      "learning_rate": 9.731276005547851e-06,
+      "loss": 0.5422,
+      "step": 3897
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.291251182556152,
+      "learning_rate": 9.72607489597781e-06,
+      "loss": 0.6075,
+      "step": 3898
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.062381744384766,
+      "learning_rate": 9.720873786407768e-06,
+      "loss": 0.6609,
+      "step": 3899
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.553919792175293,
+      "learning_rate": 9.715672676837727e-06,
+      "loss": 0.5115,
+      "step": 3900
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.35729694366455,
+      "learning_rate": 9.710471567267684e-06,
+      "loss": 0.6227,
+      "step": 3901
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.21863079071045,
+      "learning_rate": 9.705270457697643e-06,
+      "loss": 0.894,
+      "step": 3902
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.336612701416016,
+      "learning_rate": 9.700069348127601e-06,
+      "loss": 0.6471,
+      "step": 3903
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.187104225158691,
+      "learning_rate": 9.69486823855756e-06,
+      "loss": 0.6438,
+      "step": 3904
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.015570640563965,
+      "learning_rate": 9.689667128987518e-06,
+      "loss": 0.6675,
+      "step": 3905
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.77251148223877,
+      "learning_rate": 9.684466019417477e-06,
+      "loss": 0.6243,
+      "step": 3906
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.868675231933594,
+      "learning_rate": 9.679264909847436e-06,
+      "loss": 0.5957,
+      "step": 3907
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.361734390258789,
+      "learning_rate": 9.674063800277394e-06,
+      "loss": 0.5899,
+      "step": 3908
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.503995895385742,
+      "learning_rate": 9.668862690707351e-06,
+      "loss": 0.6255,
+      "step": 3909
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.476040840148926,
+      "learning_rate": 9.66366158113731e-06,
+      "loss": 0.5128,
+      "step": 3910
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.430875778198242,
+      "learning_rate": 9.658460471567268e-06,
+      "loss": 0.5344,
+      "step": 3911
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.807493209838867,
+      "learning_rate": 9.653259361997227e-06,
+      "loss": 0.5752,
+      "step": 3912
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.210729598999023,
+      "learning_rate": 9.648058252427186e-06,
+      "loss": 0.5714,
+      "step": 3913
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.432924270629883,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.7056,
+      "step": 3914
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.525184154510498,
+      "learning_rate": 9.637656033287103e-06,
+      "loss": 0.4636,
+      "step": 3915
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.021270751953125,
+      "learning_rate": 9.632454923717061e-06,
+      "loss": 0.7156,
+      "step": 3916
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.345635414123535,
+      "learning_rate": 9.627253814147018e-06,
+      "loss": 0.7232,
+      "step": 3917
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.87484359741211,
+      "learning_rate": 9.622052704576977e-06,
+      "loss": 0.5709,
+      "step": 3918
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.665295600891113,
+      "learning_rate": 9.616851595006936e-06,
+      "loss": 0.5435,
+      "step": 3919
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.096817970275879,
+      "learning_rate": 9.611650485436894e-06,
+      "loss": 0.547,
+      "step": 3920
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.888790607452393,
+      "learning_rate": 9.606449375866853e-06,
+      "loss": 0.4388,
+      "step": 3921
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.455520629882812,
+      "learning_rate": 9.601248266296811e-06,
+      "loss": 0.7217,
+      "step": 3922
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.429350852966309,
+      "learning_rate": 9.59604715672677e-06,
+      "loss": 0.5824,
+      "step": 3923
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.462335586547852,
+      "learning_rate": 9.590846047156727e-06,
+      "loss": 0.4629,
+      "step": 3924
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.63576602935791,
+      "learning_rate": 9.585644937586686e-06,
+      "loss": 0.4422,
+      "step": 3925
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.833889961242676,
+      "learning_rate": 9.580443828016644e-06,
+      "loss": 0.7287,
+      "step": 3926
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.13122844696045,
+      "learning_rate": 9.575242718446603e-06,
+      "loss": 0.3866,
+      "step": 3927
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.559880256652832,
+      "learning_rate": 9.570041608876561e-06,
+      "loss": 0.6179,
+      "step": 3928
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.636510848999023,
+      "learning_rate": 9.56484049930652e-06,
+      "loss": 0.5586,
+      "step": 3929
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.905501365661621,
+      "learning_rate": 9.559639389736479e-06,
+      "loss": 0.5741,
+      "step": 3930
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.041130065917969,
+      "learning_rate": 9.554438280166437e-06,
+      "loss": 0.6505,
+      "step": 3931
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.366535186767578,
+      "learning_rate": 9.549237170596394e-06,
+      "loss": 0.4729,
+      "step": 3932
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.981525421142578,
+      "learning_rate": 9.544036061026353e-06,
+      "loss": 0.7922,
+      "step": 3933
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.05492877960205,
+      "learning_rate": 9.538834951456311e-06,
+      "loss": 0.5604,
+      "step": 3934
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.363113403320312,
+      "learning_rate": 9.53363384188627e-06,
+      "loss": 0.5249,
+      "step": 3935
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.31744384765625,
+      "learning_rate": 9.528432732316229e-06,
+      "loss": 0.4783,
+      "step": 3936
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.005027770996094,
+      "learning_rate": 9.523231622746187e-06,
+      "loss": 0.6053,
+      "step": 3937
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.464433670043945,
+      "learning_rate": 9.518030513176146e-06,
+      "loss": 0.5561,
+      "step": 3938
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.705599784851074,
+      "learning_rate": 9.512829403606104e-06,
+      "loss": 0.6927,
+      "step": 3939
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.636458396911621,
+      "learning_rate": 9.507628294036061e-06,
+      "loss": 0.7235,
+      "step": 3940
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.677639961242676,
+      "learning_rate": 9.50242718446602e-06,
+      "loss": 0.5253,
+      "step": 3941
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.979580879211426,
+      "learning_rate": 9.497226074895979e-06,
+      "loss": 0.5638,
+      "step": 3942
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.313800811767578,
+      "learning_rate": 9.492024965325937e-06,
+      "loss": 0.6074,
+      "step": 3943
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.995197296142578,
+      "learning_rate": 9.486823855755896e-06,
+      "loss": 0.4743,
+      "step": 3944
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.639266967773438,
+      "learning_rate": 9.481622746185854e-06,
+      "loss": 0.6224,
+      "step": 3945
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.006468772888184,
+      "learning_rate": 9.476421636615813e-06,
+      "loss": 0.6149,
+      "step": 3946
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.965597152709961,
+      "learning_rate": 9.471220527045772e-06,
+      "loss": 0.6149,
+      "step": 3947
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.769302368164062,
+      "learning_rate": 9.466019417475729e-06,
+      "loss": 0.5224,
+      "step": 3948
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.580398559570312,
+      "learning_rate": 9.460818307905687e-06,
+      "loss": 0.7168,
+      "step": 3949
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.060805320739746,
+      "learning_rate": 9.455617198335646e-06,
+      "loss": 0.3802,
+      "step": 3950
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.89362621307373,
+      "learning_rate": 9.450416088765604e-06,
+      "loss": 0.7076,
+      "step": 3951
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.592148780822754,
+      "learning_rate": 9.445214979195563e-06,
+      "loss": 0.7378,
+      "step": 3952
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.761645317077637,
+      "learning_rate": 9.440013869625522e-06,
+      "loss": 0.7464,
+      "step": 3953
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.915506362915039,
+      "learning_rate": 9.43481276005548e-06,
+      "loss": 0.7414,
+      "step": 3954
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.695937156677246,
+      "learning_rate": 9.429611650485437e-06,
+      "loss": 0.7779,
+      "step": 3955
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.803906440734863,
+      "learning_rate": 9.424410540915396e-06,
+      "loss": 0.5459,
+      "step": 3956
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.288229942321777,
+      "learning_rate": 9.419209431345354e-06,
+      "loss": 0.5387,
+      "step": 3957
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.850482940673828,
+      "learning_rate": 9.414008321775313e-06,
+      "loss": 0.6187,
+      "step": 3958
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.090662002563477,
+      "learning_rate": 9.408807212205272e-06,
+      "loss": 0.5627,
+      "step": 3959
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.952340602874756,
+      "learning_rate": 9.40360610263523e-06,
+      "loss": 0.4703,
+      "step": 3960
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.194254875183105,
+      "learning_rate": 9.398404993065189e-06,
+      "loss": 0.5515,
+      "step": 3961
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.170499801635742,
+      "learning_rate": 9.393203883495147e-06,
+      "loss": 0.5723,
+      "step": 3962
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.303997993469238,
+      "learning_rate": 9.388002773925104e-06,
+      "loss": 0.3949,
+      "step": 3963
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.426458358764648,
+      "learning_rate": 9.382801664355063e-06,
+      "loss": 0.6375,
+      "step": 3964
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.935735702514648,
+      "learning_rate": 9.377600554785022e-06,
+      "loss": 0.7621,
+      "step": 3965
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.89598274230957,
+      "learning_rate": 9.372399445214978e-06,
+      "loss": 0.4387,
+      "step": 3966
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.986362457275391,
+      "learning_rate": 9.367198335644937e-06,
+      "loss": 0.5599,
+      "step": 3967
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.687256813049316,
+      "learning_rate": 9.361997226074896e-06,
+      "loss": 0.6241,
+      "step": 3968
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.004544258117676,
+      "learning_rate": 9.356796116504854e-06,
+      "loss": 0.8222,
+      "step": 3969
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.801164627075195,
+      "learning_rate": 9.351595006934813e-06,
+      "loss": 0.6082,
+      "step": 3970
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.311173439025879,
+      "learning_rate": 9.34639389736477e-06,
+      "loss": 0.5826,
+      "step": 3971
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.271430015563965,
+      "learning_rate": 9.341192787794728e-06,
+      "loss": 0.5891,
+      "step": 3972
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.223475456237793,
+      "learning_rate": 9.335991678224687e-06,
+      "loss": 0.6908,
+      "step": 3973
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.28635311126709,
+      "learning_rate": 9.330790568654646e-06,
+      "loss": 0.6853,
+      "step": 3974
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.262669563293457,
+      "learning_rate": 9.325589459084604e-06,
+      "loss": 0.9531,
+      "step": 3975
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.363923072814941,
+      "learning_rate": 9.320388349514563e-06,
+      "loss": 0.7144,
+      "step": 3976
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.309259414672852,
+      "learning_rate": 9.315187239944522e-06,
+      "loss": 0.4564,
+      "step": 3977
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.684765815734863,
+      "learning_rate": 9.30998613037448e-06,
+      "loss": 0.5123,
+      "step": 3978
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.175800323486328,
+      "learning_rate": 9.304785020804437e-06,
+      "loss": 0.3545,
+      "step": 3979
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.177339553833008,
+      "learning_rate": 9.299583911234396e-06,
+      "loss": 0.4766,
+      "step": 3980
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.557180404663086,
+      "learning_rate": 9.294382801664354e-06,
+      "loss": 0.5548,
+      "step": 3981
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.362018585205078,
+      "learning_rate": 9.289181692094313e-06,
+      "loss": 0.5355,
+      "step": 3982
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.076787948608398,
+      "learning_rate": 9.283980582524272e-06,
+      "loss": 0.6852,
+      "step": 3983
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.204864501953125,
+      "learning_rate": 9.27877947295423e-06,
+      "loss": 0.6527,
+      "step": 3984
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.979928970336914,
+      "learning_rate": 9.273578363384189e-06,
+      "loss": 0.6929,
+      "step": 3985
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.842995643615723,
+      "learning_rate": 9.268377253814146e-06,
+      "loss": 0.4866,
+      "step": 3986
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.858844757080078,
+      "learning_rate": 9.263176144244104e-06,
+      "loss": 0.5833,
+      "step": 3987
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.090740203857422,
+      "learning_rate": 9.257975034674063e-06,
+      "loss": 0.7141,
+      "step": 3988
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.092180252075195,
+      "learning_rate": 9.252773925104021e-06,
+      "loss": 0.3945,
+      "step": 3989
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.153999328613281,
+      "learning_rate": 9.24757281553398e-06,
+      "loss": 0.5336,
+      "step": 3990
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.9664435386657715,
+      "learning_rate": 9.242371705963939e-06,
+      "loss": 0.5008,
+      "step": 3991
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.706225395202637,
+      "learning_rate": 9.237170596393897e-06,
+      "loss": 0.6097,
+      "step": 3992
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.656311988830566,
+      "learning_rate": 9.231969486823856e-06,
+      "loss": 0.5414,
+      "step": 3993
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.129077911376953,
+      "learning_rate": 9.226768377253813e-06,
+      "loss": 0.4334,
+      "step": 3994
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.154330253601074,
+      "learning_rate": 9.221567267683771e-06,
+      "loss": 0.6636,
+      "step": 3995
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.524849891662598,
+      "learning_rate": 9.21636615811373e-06,
+      "loss": 0.9208,
+      "step": 3996
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.08932876586914,
+      "learning_rate": 9.211165048543689e-06,
+      "loss": 0.4183,
+      "step": 3997
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.20439338684082,
+      "learning_rate": 9.205963938973647e-06,
+      "loss": 0.6433,
+      "step": 3998
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.051145553588867,
+      "learning_rate": 9.200762829403606e-06,
+      "loss": 0.6261,
+      "step": 3999
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.316868782043457,
+      "learning_rate": 9.195561719833565e-06,
+      "loss": 0.7119,
+      "step": 4000
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.518169403076172,
+      "learning_rate": 9.190360610263523e-06,
+      "loss": 0.6989,
+      "step": 4001
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.748174667358398,
+      "learning_rate": 9.18515950069348e-06,
+      "loss": 0.5253,
+      "step": 4002
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.649458885192871,
+      "learning_rate": 9.179958391123439e-06,
+      "loss": 0.6381,
+      "step": 4003
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.529348373413086,
+      "learning_rate": 9.174757281553397e-06,
+      "loss": 0.4463,
+      "step": 4004
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.567290306091309,
+      "learning_rate": 9.169556171983356e-06,
+      "loss": 0.6905,
+      "step": 4005
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.410309791564941,
+      "learning_rate": 9.164355062413315e-06,
+      "loss": 0.5516,
+      "step": 4006
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.157402038574219,
+      "learning_rate": 9.159153952843273e-06,
+      "loss": 0.7797,
+      "step": 4007
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.833531856536865,
+      "learning_rate": 9.153952843273232e-06,
+      "loss": 0.5224,
+      "step": 4008
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.314779281616211,
+      "learning_rate": 9.14875173370319e-06,
+      "loss": 0.6569,
+      "step": 4009
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.548032760620117,
+      "learning_rate": 9.143550624133147e-06,
+      "loss": 0.647,
+      "step": 4010
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.714493751525879,
+      "learning_rate": 9.138349514563106e-06,
+      "loss": 0.4617,
+      "step": 4011
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.29658031463623,
+      "learning_rate": 9.133148404993064e-06,
+      "loss": 0.5471,
+      "step": 4012
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.495676040649414,
+      "learning_rate": 9.127947295423023e-06,
+      "loss": 0.6501,
+      "step": 4013
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.9902472496032715,
+      "learning_rate": 9.122746185852982e-06,
+      "loss": 0.6268,
+      "step": 4014
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.472746849060059,
+      "learning_rate": 9.11754507628294e-06,
+      "loss": 0.6418,
+      "step": 4015
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.643261909484863,
+      "learning_rate": 9.112343966712899e-06,
+      "loss": 0.6348,
+      "step": 4016
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.701826095581055,
+      "learning_rate": 9.107142857142856e-06,
+      "loss": 0.7532,
+      "step": 4017
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.086309432983398,
+      "learning_rate": 9.101941747572814e-06,
+      "loss": 0.6109,
+      "step": 4018
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.730475425720215,
+      "learning_rate": 9.096740638002773e-06,
+      "loss": 0.541,
+      "step": 4019
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.078007698059082,
+      "learning_rate": 9.091539528432732e-06,
+      "loss": 0.5549,
+      "step": 4020
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.938374519348145,
+      "learning_rate": 9.08633841886269e-06,
+      "loss": 0.6263,
+      "step": 4021
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.744109153747559,
+      "learning_rate": 9.081137309292649e-06,
+      "loss": 0.4254,
+      "step": 4022
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.604860305786133,
+      "learning_rate": 9.075936199722608e-06,
+      "loss": 0.6178,
+      "step": 4023
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.243579864501953,
+      "learning_rate": 9.070735090152566e-06,
+      "loss": 0.4514,
+      "step": 4024
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.828361511230469,
+      "learning_rate": 9.065533980582523e-06,
+      "loss": 0.476,
+      "step": 4025
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.489561080932617,
+      "learning_rate": 9.060332871012482e-06,
+      "loss": 0.7044,
+      "step": 4026
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.893275260925293,
+      "learning_rate": 9.05513176144244e-06,
+      "loss": 0.547,
+      "step": 4027
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.703782081604004,
+      "learning_rate": 9.049930651872399e-06,
+      "loss": 0.5462,
+      "step": 4028
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.014063835144043,
+      "learning_rate": 9.044729542302358e-06,
+      "loss": 0.7508,
+      "step": 4029
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.457059860229492,
+      "learning_rate": 9.039528432732316e-06,
+      "loss": 0.7048,
+      "step": 4030
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.868513107299805,
+      "learning_rate": 9.034327323162275e-06,
+      "loss": 0.594,
+      "step": 4031
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.472339630126953,
+      "learning_rate": 9.029126213592233e-06,
+      "loss": 0.6335,
+      "step": 4032
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.564633846282959,
+      "learning_rate": 9.02392510402219e-06,
+      "loss": 0.4016,
+      "step": 4033
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.511756896972656,
+      "learning_rate": 9.018723994452149e-06,
+      "loss": 0.6127,
+      "step": 4034
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.21804141998291,
+      "learning_rate": 9.013522884882107e-06,
+      "loss": 0.5462,
+      "step": 4035
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.767874717712402,
+      "learning_rate": 9.008321775312066e-06,
+      "loss": 0.6504,
+      "step": 4036
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.71042537689209,
+      "learning_rate": 9.003120665742025e-06,
+      "loss": 0.6468,
+      "step": 4037
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.777512550354004,
+      "learning_rate": 8.997919556171983e-06,
+      "loss": 0.6782,
+      "step": 4038
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.692131042480469,
+      "learning_rate": 8.992718446601942e-06,
+      "loss": 0.8006,
+      "step": 4039
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.406553268432617,
+      "learning_rate": 8.987517337031899e-06,
+      "loss": 0.4617,
+      "step": 4040
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.598189353942871,
+      "learning_rate": 8.982316227461857e-06,
+      "loss": 0.7759,
+      "step": 4041
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.143418312072754,
+      "learning_rate": 8.977115117891816e-06,
+      "loss": 0.583,
+      "step": 4042
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.970319747924805,
+      "learning_rate": 8.971914008321775e-06,
+      "loss": 0.601,
+      "step": 4043
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.748861312866211,
+      "learning_rate": 8.966712898751733e-06,
+      "loss": 0.5218,
+      "step": 4044
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.431510925292969,
+      "learning_rate": 8.961511789181692e-06,
+      "loss": 0.7011,
+      "step": 4045
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.7251615524292,
+      "learning_rate": 8.95631067961165e-06,
+      "loss": 0.5317,
+      "step": 4046
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.743082046508789,
+      "learning_rate": 8.951109570041609e-06,
+      "loss": 0.523,
+      "step": 4047
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.241541862487793,
+      "learning_rate": 8.945908460471566e-06,
+      "loss": 0.9379,
+      "step": 4048
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.906710624694824,
+      "learning_rate": 8.940707350901525e-06,
+      "loss": 0.6484,
+      "step": 4049
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.110555648803711,
+      "learning_rate": 8.935506241331483e-06,
+      "loss": 0.6925,
+      "step": 4050
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.475569725036621,
+      "learning_rate": 8.930305131761442e-06,
+      "loss": 0.5886,
+      "step": 4051
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.216607093811035,
+      "learning_rate": 8.9251040221914e-06,
+      "loss": 0.5686,
+      "step": 4052
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.518875122070312,
+      "learning_rate": 8.919902912621359e-06,
+      "loss": 0.8086,
+      "step": 4053
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.564156532287598,
+      "learning_rate": 8.914701803051318e-06,
+      "loss": 0.5504,
+      "step": 4054
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.248382568359375,
+      "learning_rate": 8.909500693481276e-06,
+      "loss": 0.6618,
+      "step": 4055
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.736210346221924,
+      "learning_rate": 8.904299583911233e-06,
+      "loss": 0.5564,
+      "step": 4056
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.78152084350586,
+      "learning_rate": 8.899098474341192e-06,
+      "loss": 0.5386,
+      "step": 4057
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.7623677253723145,
+      "learning_rate": 8.89389736477115e-06,
+      "loss": 0.5322,
+      "step": 4058
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.249896049499512,
+      "learning_rate": 8.888696255201109e-06,
+      "loss": 0.4869,
+      "step": 4059
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.190584182739258,
+      "learning_rate": 8.883495145631068e-06,
+      "loss": 0.6381,
+      "step": 4060
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.149601936340332,
+      "learning_rate": 8.878294036061026e-06,
+      "loss": 0.5465,
+      "step": 4061
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.451274871826172,
+      "learning_rate": 8.873092926490985e-06,
+      "loss": 0.7345,
+      "step": 4062
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.13867473602295,
+      "learning_rate": 8.867891816920944e-06,
+      "loss": 0.6338,
+      "step": 4063
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.233431816101074,
+      "learning_rate": 8.8626907073509e-06,
+      "loss": 0.5073,
+      "step": 4064
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.802664756774902,
+      "learning_rate": 8.857489597780859e-06,
+      "loss": 0.6268,
+      "step": 4065
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.31629753112793,
+      "learning_rate": 8.852288488210818e-06,
+      "loss": 0.5618,
+      "step": 4066
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.913722038269043,
+      "learning_rate": 8.847087378640776e-06,
+      "loss": 0.5816,
+      "step": 4067
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.453803062438965,
+      "learning_rate": 8.841886269070735e-06,
+      "loss": 0.4307,
+      "step": 4068
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.207311630249023,
+      "learning_rate": 8.836685159500694e-06,
+      "loss": 0.6779,
+      "step": 4069
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.781633377075195,
+      "learning_rate": 8.831484049930652e-06,
+      "loss": 0.7317,
+      "step": 4070
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.256462097167969,
+      "learning_rate": 8.826282940360609e-06,
+      "loss": 0.6943,
+      "step": 4071
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.30513334274292,
+      "learning_rate": 8.821081830790568e-06,
+      "loss": 0.4382,
+      "step": 4072
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.94314956665039,
+      "learning_rate": 8.815880721220526e-06,
+      "loss": 0.4324,
+      "step": 4073
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.678756713867188,
+      "learning_rate": 8.810679611650485e-06,
+      "loss": 0.6305,
+      "step": 4074
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.60449504852295,
+      "learning_rate": 8.805478502080444e-06,
+      "loss": 0.687,
+      "step": 4075
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.31718635559082,
+      "learning_rate": 8.800277392510402e-06,
+      "loss": 0.6167,
+      "step": 4076
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.407679557800293,
+      "learning_rate": 8.79507628294036e-06,
+      "loss": 0.639,
+      "step": 4077
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.106707572937012,
+      "learning_rate": 8.78987517337032e-06,
+      "loss": 0.3483,
+      "step": 4078
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.66930103302002,
+      "learning_rate": 8.784674063800276e-06,
+      "loss": 0.4578,
+      "step": 4079
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.265925407409668,
+      "learning_rate": 8.779472954230235e-06,
+      "loss": 0.596,
+      "step": 4080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.225571632385254,
+      "learning_rate": 8.774271844660194e-06,
+      "loss": 0.6296,
+      "step": 4081
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.265103340148926,
+      "learning_rate": 8.769070735090152e-06,
+      "loss": 0.631,
+      "step": 4082
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.087563514709473,
+      "learning_rate": 8.76386962552011e-06,
+      "loss": 0.6576,
+      "step": 4083
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.464720726013184,
+      "learning_rate": 8.75866851595007e-06,
+      "loss": 0.6548,
+      "step": 4084
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.453390121459961,
+      "learning_rate": 8.753467406380028e-06,
+      "loss": 0.573,
+      "step": 4085
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.286153793334961,
+      "learning_rate": 8.748266296809987e-06,
+      "loss": 0.5852,
+      "step": 4086
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.632871627807617,
+      "learning_rate": 8.743065187239943e-06,
+      "loss": 0.8391,
+      "step": 4087
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.9220147132873535,
+      "learning_rate": 8.737864077669902e-06,
+      "loss": 0.5266,
+      "step": 4088
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.175202369689941,
+      "learning_rate": 8.73266296809986e-06,
+      "loss": 0.4842,
+      "step": 4089
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.416153907775879,
+      "learning_rate": 8.72746185852982e-06,
+      "loss": 0.7275,
+      "step": 4090
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.664410591125488,
+      "learning_rate": 8.722260748959778e-06,
+      "loss": 0.5818,
+      "step": 4091
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.330435752868652,
+      "learning_rate": 8.717059639389737e-06,
+      "loss": 0.6986,
+      "step": 4092
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.058676719665527,
+      "learning_rate": 8.711858529819695e-06,
+      "loss": 0.6405,
+      "step": 4093
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.471726417541504,
+      "learning_rate": 8.706657420249654e-06,
+      "loss": 0.687,
+      "step": 4094
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.358205795288086,
+      "learning_rate": 8.70145631067961e-06,
+      "loss": 0.5006,
+      "step": 4095
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.522767066955566,
+      "learning_rate": 8.69625520110957e-06,
+      "loss": 0.8418,
+      "step": 4096
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.74323558807373,
+      "learning_rate": 8.691054091539528e-06,
+      "loss": 0.614,
+      "step": 4097
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.892989158630371,
+      "learning_rate": 8.685852981969487e-06,
+      "loss": 0.825,
+      "step": 4098
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.878817558288574,
+      "learning_rate": 8.680651872399445e-06,
+      "loss": 0.4729,
+      "step": 4099
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.31519889831543,
+      "learning_rate": 8.675450762829404e-06,
+      "loss": 0.6773,
+      "step": 4100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.638272762298584,
+      "learning_rate": 8.670249653259362e-06,
+      "loss": 0.5226,
+      "step": 4101
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.452232360839844,
+      "learning_rate": 8.66504854368932e-06,
+      "loss": 0.358,
+      "step": 4102
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.439286231994629,
+      "learning_rate": 8.659847434119278e-06,
+      "loss": 0.7011,
+      "step": 4103
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.544307231903076,
+      "learning_rate": 8.654646324549237e-06,
+      "loss": 0.4038,
+      "step": 4104
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.509504318237305,
+      "learning_rate": 8.649445214979195e-06,
+      "loss": 0.5808,
+      "step": 4105
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.752281188964844,
+      "learning_rate": 8.644244105409154e-06,
+      "loss": 0.6767,
+      "step": 4106
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.252452850341797,
+      "learning_rate": 8.639042995839112e-06,
+      "loss": 0.7246,
+      "step": 4107
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.565033912658691,
+      "learning_rate": 8.633841886269071e-06,
+      "loss": 0.5779,
+      "step": 4108
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.057456970214844,
+      "learning_rate": 8.62864077669903e-06,
+      "loss": 0.5973,
+      "step": 4109
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.389781951904297,
+      "learning_rate": 8.623439667128986e-06,
+      "loss": 0.4811,
+      "step": 4110
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.153032302856445,
+      "learning_rate": 8.618238557558945e-06,
+      "loss": 0.5512,
+      "step": 4111
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.65677261352539,
+      "learning_rate": 8.613037447988904e-06,
+      "loss": 0.5542,
+      "step": 4112
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.531868934631348,
+      "learning_rate": 8.607836338418862e-06,
+      "loss": 0.3992,
+      "step": 4113
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.977151393890381,
+      "learning_rate": 8.602635228848821e-06,
+      "loss": 0.6348,
+      "step": 4114
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.527854919433594,
+      "learning_rate": 8.59743411927878e-06,
+      "loss": 0.7857,
+      "step": 4115
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.593602180480957,
+      "learning_rate": 8.592233009708738e-06,
+      "loss": 0.5332,
+      "step": 4116
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.583084106445312,
+      "learning_rate": 8.587031900138697e-06,
+      "loss": 0.4846,
+      "step": 4117
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.632596969604492,
+      "learning_rate": 8.581830790568654e-06,
+      "loss": 0.4383,
+      "step": 4118
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.326327323913574,
+      "learning_rate": 8.576629680998612e-06,
+      "loss": 0.7644,
+      "step": 4119
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.119685173034668,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.503,
+      "step": 4120
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.573513984680176,
+      "learning_rate": 8.56622746185853e-06,
+      "loss": 0.6756,
+      "step": 4121
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.056134223937988,
+      "learning_rate": 8.561026352288488e-06,
+      "loss": 0.4853,
+      "step": 4122
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.918344497680664,
+      "learning_rate": 8.555825242718447e-06,
+      "loss": 0.5975,
+      "step": 4123
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.489954471588135,
+      "learning_rate": 8.550624133148405e-06,
+      "loss": 0.5043,
+      "step": 4124
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.384556770324707,
+      "learning_rate": 8.545423023578364e-06,
+      "loss": 0.5815,
+      "step": 4125
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.562905311584473,
+      "learning_rate": 8.540221914008321e-06,
+      "loss": 0.7072,
+      "step": 4126
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.60637378692627,
+      "learning_rate": 8.53502080443828e-06,
+      "loss": 0.7442,
+      "step": 4127
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.500212669372559,
+      "learning_rate": 8.529819694868238e-06,
+      "loss": 0.649,
+      "step": 4128
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.864381790161133,
+      "learning_rate": 8.524618585298197e-06,
+      "loss": 0.6704,
+      "step": 4129
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.950214862823486,
+      "learning_rate": 8.519417475728155e-06,
+      "loss": 0.406,
+      "step": 4130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.812705039978027,
+      "learning_rate": 8.514216366158114e-06,
+      "loss": 0.665,
+      "step": 4131
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.913187026977539,
+      "learning_rate": 8.509015256588073e-06,
+      "loss": 0.6361,
+      "step": 4132
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.151941299438477,
+      "learning_rate": 8.50381414701803e-06,
+      "loss": 0.5495,
+      "step": 4133
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.988035678863525,
+      "learning_rate": 8.498613037447988e-06,
+      "loss": 0.5976,
+      "step": 4134
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.692972183227539,
+      "learning_rate": 8.493411927877947e-06,
+      "loss": 0.5354,
+      "step": 4135
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.463018417358398,
+      "learning_rate": 8.488210818307905e-06,
+      "loss": 0.655,
+      "step": 4136
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.997400283813477,
+      "learning_rate": 8.483009708737864e-06,
+      "loss": 0.6122,
+      "step": 4137
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.222185134887695,
+      "learning_rate": 8.477808599167823e-06,
+      "loss": 0.4595,
+      "step": 4138
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.460000991821289,
+      "learning_rate": 8.472607489597781e-06,
+      "loss": 0.4696,
+      "step": 4139
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.542812347412109,
+      "learning_rate": 8.46740638002774e-06,
+      "loss": 0.4933,
+      "step": 4140
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.410696029663086,
+      "learning_rate": 8.462205270457697e-06,
+      "loss": 0.8858,
+      "step": 4141
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.888580322265625,
+      "learning_rate": 8.457004160887655e-06,
+      "loss": 0.4208,
+      "step": 4142
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.285781860351562,
+      "learning_rate": 8.451803051317614e-06,
+      "loss": 0.4526,
+      "step": 4143
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.837698936462402,
+      "learning_rate": 8.446601941747573e-06,
+      "loss": 0.6304,
+      "step": 4144
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.67211627960205,
+      "learning_rate": 8.441400832177531e-06,
+      "loss": 0.7947,
+      "step": 4145
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.829214096069336,
+      "learning_rate": 8.43619972260749e-06,
+      "loss": 0.7083,
+      "step": 4146
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.882656574249268,
+      "learning_rate": 8.430998613037448e-06,
+      "loss": 0.4646,
+      "step": 4147
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.243562698364258,
+      "learning_rate": 8.425797503467407e-06,
+      "loss": 0.5869,
+      "step": 4148
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.774547100067139,
+      "learning_rate": 8.420596393897364e-06,
+      "loss": 0.4616,
+      "step": 4149
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.992588996887207,
+      "learning_rate": 8.415395284327323e-06,
+      "loss": 0.5312,
+      "step": 4150
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.334717750549316,
+      "learning_rate": 8.410194174757281e-06,
+      "loss": 0.6648,
+      "step": 4151
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.001058578491211,
+      "learning_rate": 8.40499306518724e-06,
+      "loss": 0.4812,
+      "step": 4152
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.907970428466797,
+      "learning_rate": 8.399791955617198e-06,
+      "loss": 0.6458,
+      "step": 4153
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.245448112487793,
+      "learning_rate": 8.394590846047157e-06,
+      "loss": 0.3836,
+      "step": 4154
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.554336071014404,
+      "learning_rate": 8.389389736477116e-06,
+      "loss": 0.5906,
+      "step": 4155
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.601044654846191,
+      "learning_rate": 8.384188626907072e-06,
+      "loss": 0.663,
+      "step": 4156
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.154277801513672,
+      "learning_rate": 8.378987517337031e-06,
+      "loss": 0.7108,
+      "step": 4157
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.22092342376709,
+      "learning_rate": 8.37378640776699e-06,
+      "loss": 0.5305,
+      "step": 4158
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.64560604095459,
+      "learning_rate": 8.368585298196948e-06,
+      "loss": 0.6795,
+      "step": 4159
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.183959007263184,
+      "learning_rate": 8.363384188626907e-06,
+      "loss": 0.438,
+      "step": 4160
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.393022060394287,
+      "learning_rate": 8.358183079056866e-06,
+      "loss": 0.394,
+      "step": 4161
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.689142227172852,
+      "learning_rate": 8.352981969486824e-06,
+      "loss": 0.7846,
+      "step": 4162
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.87822437286377,
+      "learning_rate": 8.347780859916783e-06,
+      "loss": 0.6148,
+      "step": 4163
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.3933563232421875,
+      "learning_rate": 8.34257975034674e-06,
+      "loss": 0.3324,
+      "step": 4164
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.125384330749512,
+      "learning_rate": 8.337378640776698e-06,
+      "loss": 0.7019,
+      "step": 4165
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.272537231445312,
+      "learning_rate": 8.332177531206657e-06,
+      "loss": 0.4373,
+      "step": 4166
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.526453018188477,
+      "learning_rate": 8.326976421636616e-06,
+      "loss": 0.6455,
+      "step": 4167
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.810981750488281,
+      "learning_rate": 8.321775312066574e-06,
+      "loss": 0.5686,
+      "step": 4168
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.137959480285645,
+      "learning_rate": 8.316574202496533e-06,
+      "loss": 0.6625,
+      "step": 4169
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.717644691467285,
+      "learning_rate": 8.311373092926491e-06,
+      "loss": 0.5637,
+      "step": 4170
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.075149536132812,
+      "learning_rate": 8.30617198335645e-06,
+      "loss": 0.5775,
+      "step": 4171
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.180283546447754,
+      "learning_rate": 8.300970873786407e-06,
+      "loss": 0.4706,
+      "step": 4172
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.954649925231934,
+      "learning_rate": 8.295769764216366e-06,
+      "loss": 0.6894,
+      "step": 4173
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.631184101104736,
+      "learning_rate": 8.290568654646324e-06,
+      "loss": 0.2642,
+      "step": 4174
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.316287040710449,
+      "learning_rate": 8.285367545076283e-06,
+      "loss": 0.4262,
+      "step": 4175
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.786038398742676,
+      "learning_rate": 8.280166435506241e-06,
+      "loss": 0.6696,
+      "step": 4176
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.144767761230469,
+      "learning_rate": 8.2749653259362e-06,
+      "loss": 0.6631,
+      "step": 4177
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.797948360443115,
+      "learning_rate": 8.269764216366159e-06,
+      "loss": 0.3499,
+      "step": 4178
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.11242961883545,
+      "learning_rate": 8.264563106796117e-06,
+      "loss": 0.6446,
+      "step": 4179
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.720194339752197,
+      "learning_rate": 8.259361997226074e-06,
+      "loss": 0.3987,
+      "step": 4180
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.407652854919434,
+      "learning_rate": 8.254160887656033e-06,
+      "loss": 0.6114,
+      "step": 4181
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.690034866333008,
+      "learning_rate": 8.248959778085991e-06,
+      "loss": 0.6606,
+      "step": 4182
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.354161739349365,
+      "learning_rate": 8.24375866851595e-06,
+      "loss": 0.4108,
+      "step": 4183
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.432453155517578,
+      "learning_rate": 8.238557558945909e-06,
+      "loss": 0.5458,
+      "step": 4184
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.361937522888184,
+      "learning_rate": 8.233356449375867e-06,
+      "loss": 0.5598,
+      "step": 4185
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.418684959411621,
+      "learning_rate": 8.228155339805826e-06,
+      "loss": 0.5486,
+      "step": 4186
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.254600524902344,
+      "learning_rate": 8.222954230235783e-06,
+      "loss": 0.5018,
+      "step": 4187
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.48658561706543,
+      "learning_rate": 8.217753120665741e-06,
+      "loss": 0.6885,
+      "step": 4188
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.154390335083008,
+      "learning_rate": 8.2125520110957e-06,
+      "loss": 0.7845,
+      "step": 4189
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.572808265686035,
+      "learning_rate": 8.207350901525659e-06,
+      "loss": 0.6216,
+      "step": 4190
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.542267799377441,
+      "learning_rate": 8.202149791955617e-06,
+      "loss": 0.7855,
+      "step": 4191
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.745335578918457,
+      "learning_rate": 8.196948682385576e-06,
+      "loss": 0.6447,
+      "step": 4192
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.569515705108643,
+      "learning_rate": 8.191747572815534e-06,
+      "loss": 0.4476,
+      "step": 4193
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.545004844665527,
+      "learning_rate": 8.186546463245493e-06,
+      "loss": 0.477,
+      "step": 4194
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.688425064086914,
+      "learning_rate": 8.18134535367545e-06,
+      "loss": 0.4938,
+      "step": 4195
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.485764503479004,
+      "learning_rate": 8.176144244105409e-06,
+      "loss": 0.4644,
+      "step": 4196
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.267085075378418,
+      "learning_rate": 8.170943134535367e-06,
+      "loss": 0.4656,
+      "step": 4197
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.375263214111328,
+      "learning_rate": 8.165742024965326e-06,
+      "loss": 0.6188,
+      "step": 4198
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.074758529663086,
+      "learning_rate": 8.160540915395284e-06,
+      "loss": 0.7597,
+      "step": 4199
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.885358810424805,
+      "learning_rate": 8.155339805825243e-06,
+      "loss": 0.8247,
+      "step": 4200
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.012892723083496,
+      "learning_rate": 8.150138696255202e-06,
+      "loss": 0.6088,
+      "step": 4201
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.001459121704102,
+      "learning_rate": 8.14493758668516e-06,
+      "loss": 0.6151,
+      "step": 4202
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.38743782043457,
+      "learning_rate": 8.139736477115117e-06,
+      "loss": 0.6589,
+      "step": 4203
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.617903709411621,
+      "learning_rate": 8.134535367545076e-06,
+      "loss": 0.6585,
+      "step": 4204
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.352035522460938,
+      "learning_rate": 8.129334257975034e-06,
+      "loss": 0.6109,
+      "step": 4205
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.008535385131836,
+      "learning_rate": 8.124133148404993e-06,
+      "loss": 0.4885,
+      "step": 4206
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.695870399475098,
+      "learning_rate": 8.118932038834952e-06,
+      "loss": 0.7174,
+      "step": 4207
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.759750366210938,
+      "learning_rate": 8.11373092926491e-06,
+      "loss": 0.4558,
+      "step": 4208
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.075036525726318,
+      "learning_rate": 8.108529819694869e-06,
+      "loss": 0.3972,
+      "step": 4209
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.534637451171875,
+      "learning_rate": 8.103328710124827e-06,
+      "loss": 0.5836,
+      "step": 4210
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.297171592712402,
+      "learning_rate": 8.098127600554784e-06,
+      "loss": 0.6069,
+      "step": 4211
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.497100830078125,
+      "learning_rate": 8.092926490984743e-06,
+      "loss": 0.5252,
+      "step": 4212
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.9551420211792,
+      "learning_rate": 8.087725381414702e-06,
+      "loss": 0.811,
+      "step": 4213
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.333817005157471,
+      "learning_rate": 8.08252427184466e-06,
+      "loss": 0.2914,
+      "step": 4214
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.859624862670898,
+      "learning_rate": 8.077323162274619e-06,
+      "loss": 0.4955,
+      "step": 4215
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.63651180267334,
+      "learning_rate": 8.072122052704577e-06,
+      "loss": 0.7019,
+      "step": 4216
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.217057704925537,
+      "learning_rate": 8.066920943134536e-06,
+      "loss": 0.5321,
+      "step": 4217
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.898738861083984,
+      "learning_rate": 8.061719833564493e-06,
+      "loss": 0.6337,
+      "step": 4218
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.836544036865234,
+      "learning_rate": 8.056518723994452e-06,
+      "loss": 0.7399,
+      "step": 4219
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.220185279846191,
+      "learning_rate": 8.05131761442441e-06,
+      "loss": 0.5401,
+      "step": 4220
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.683444023132324,
+      "learning_rate": 8.046116504854369e-06,
+      "loss": 0.622,
+      "step": 4221
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.447877883911133,
+      "learning_rate": 8.040915395284327e-06,
+      "loss": 0.4633,
+      "step": 4222
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.923842430114746,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 0.4938,
+      "step": 4223
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.957904815673828,
+      "learning_rate": 8.030513176144245e-06,
+      "loss": 0.4601,
+      "step": 4224
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.874549865722656,
+      "learning_rate": 8.025312066574203e-06,
+      "loss": 0.6091,
+      "step": 4225
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.266454696655273,
+      "learning_rate": 8.02011095700416e-06,
+      "loss": 0.5242,
+      "step": 4226
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.043966293334961,
+      "learning_rate": 8.014909847434119e-06,
+      "loss": 0.6115,
+      "step": 4227
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.032555103302002,
+      "learning_rate": 8.009708737864077e-06,
+      "loss": 0.5056,
+      "step": 4228
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.58064079284668,
+      "learning_rate": 8.004507628294036e-06,
+      "loss": 0.4685,
+      "step": 4229
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.460078239440918,
+      "learning_rate": 7.999306518723995e-06,
+      "loss": 0.4969,
+      "step": 4230
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.311739921569824,
+      "learning_rate": 7.994105409153953e-06,
+      "loss": 0.5216,
+      "step": 4231
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.704087734222412,
+      "learning_rate": 7.988904299583912e-06,
+      "loss": 0.528,
+      "step": 4232
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.92158317565918,
+      "learning_rate": 7.98370319001387e-06,
+      "loss": 0.6106,
+      "step": 4233
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.909663200378418,
+      "learning_rate": 7.978502080443827e-06,
+      "loss": 0.5223,
+      "step": 4234
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.069239616394043,
+      "learning_rate": 7.973300970873786e-06,
+      "loss": 0.5404,
+      "step": 4235
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.999006271362305,
+      "learning_rate": 7.968099861303745e-06,
+      "loss": 0.4252,
+      "step": 4236
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.731756210327148,
+      "learning_rate": 7.962898751733703e-06,
+      "loss": 0.7384,
+      "step": 4237
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.565672874450684,
+      "learning_rate": 7.957697642163662e-06,
+      "loss": 0.4041,
+      "step": 4238
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.11467170715332,
+      "learning_rate": 7.95249653259362e-06,
+      "loss": 0.6798,
+      "step": 4239
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.365856170654297,
+      "learning_rate": 7.947295423023579e-06,
+      "loss": 0.455,
+      "step": 4240
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.259723663330078,
+      "learning_rate": 7.942094313453536e-06,
+      "loss": 0.6675,
+      "step": 4241
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.594016075134277,
+      "learning_rate": 7.936893203883495e-06,
+      "loss": 0.4983,
+      "step": 4242
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.081572532653809,
+      "learning_rate": 7.931692094313453e-06,
+      "loss": 0.5848,
+      "step": 4243
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.256620407104492,
+      "learning_rate": 7.926490984743412e-06,
+      "loss": 0.6263,
+      "step": 4244
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.013093948364258,
+      "learning_rate": 7.92128987517337e-06,
+      "loss": 0.4268,
+      "step": 4245
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.717179298400879,
+      "learning_rate": 7.916088765603329e-06,
+      "loss": 0.5684,
+      "step": 4246
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.643256187438965,
+      "learning_rate": 7.910887656033288e-06,
+      "loss": 0.7204,
+      "step": 4247
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.765620231628418,
+      "learning_rate": 7.905686546463246e-06,
+      "loss": 0.7858,
+      "step": 4248
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.790409088134766,
+      "learning_rate": 7.900485436893203e-06,
+      "loss": 0.6419,
+      "step": 4249
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.700429916381836,
+      "learning_rate": 7.895284327323162e-06,
+      "loss": 0.556,
+      "step": 4250
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.430482387542725,
+      "learning_rate": 7.89008321775312e-06,
+      "loss": 0.4401,
+      "step": 4251
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.089859008789062,
+      "learning_rate": 7.884882108183079e-06,
+      "loss": 0.4241,
+      "step": 4252
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.872696876525879,
+      "learning_rate": 7.879680998613038e-06,
+      "loss": 0.5251,
+      "step": 4253
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.393576622009277,
+      "learning_rate": 7.874479889042996e-06,
+      "loss": 0.6022,
+      "step": 4254
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.529916763305664,
+      "learning_rate": 7.869278779472955e-06,
+      "loss": 0.5409,
+      "step": 4255
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.638440132141113,
+      "learning_rate": 7.864077669902913e-06,
+      "loss": 0.5881,
+      "step": 4256
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.298266410827637,
+      "learning_rate": 7.85887656033287e-06,
+      "loss": 0.4504,
+      "step": 4257
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.067214965820312,
+      "learning_rate": 7.853675450762829e-06,
+      "loss": 0.5605,
+      "step": 4258
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.941815376281738,
+      "learning_rate": 7.848474341192788e-06,
+      "loss": 0.7303,
+      "step": 4259
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.214824676513672,
+      "learning_rate": 7.843273231622746e-06,
+      "loss": 0.5098,
+      "step": 4260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.805112361907959,
+      "learning_rate": 7.838072122052705e-06,
+      "loss": 0.3283,
+      "step": 4261
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.640807151794434,
+      "learning_rate": 7.832871012482663e-06,
+      "loss": 0.3529,
+      "step": 4262
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.576926231384277,
+      "learning_rate": 7.827669902912622e-06,
+      "loss": 0.5417,
+      "step": 4263
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.069106101989746,
+      "learning_rate": 7.82246879334258e-06,
+      "loss": 0.816,
+      "step": 4264
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.296953201293945,
+      "learning_rate": 7.817267683772538e-06,
+      "loss": 0.5627,
+      "step": 4265
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.217358589172363,
+      "learning_rate": 7.812066574202496e-06,
+      "loss": 0.4396,
+      "step": 4266
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.352422714233398,
+      "learning_rate": 7.806865464632455e-06,
+      "loss": 0.467,
+      "step": 4267
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.038369178771973,
+      "learning_rate": 7.801664355062413e-06,
+      "loss": 0.7209,
+      "step": 4268
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.441539764404297,
+      "learning_rate": 7.796463245492372e-06,
+      "loss": 0.7943,
+      "step": 4269
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.130648612976074,
+      "learning_rate": 7.79126213592233e-06,
+      "loss": 0.5577,
+      "step": 4270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.616244316101074,
+      "learning_rate": 7.78606102635229e-06,
+      "loss": 0.4225,
+      "step": 4271
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.805031776428223,
+      "learning_rate": 7.780859916782246e-06,
+      "loss": 0.6443,
+      "step": 4272
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.328771591186523,
+      "learning_rate": 7.775658807212205e-06,
+      "loss": 0.8338,
+      "step": 4273
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.804397106170654,
+      "learning_rate": 7.770457697642163e-06,
+      "loss": 0.5276,
+      "step": 4274
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.468535423278809,
+      "learning_rate": 7.765256588072122e-06,
+      "loss": 0.4909,
+      "step": 4275
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.982081890106201,
+      "learning_rate": 7.76005547850208e-06,
+      "loss": 0.5585,
+      "step": 4276
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.10932731628418,
+      "learning_rate": 7.75485436893204e-06,
+      "loss": 0.5543,
+      "step": 4277
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.218323707580566,
+      "learning_rate": 7.749653259361998e-06,
+      "loss": 0.5665,
+      "step": 4278
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.288492202758789,
+      "learning_rate": 7.744452149791956e-06,
+      "loss": 0.5533,
+      "step": 4279
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.647442817687988,
+      "learning_rate": 7.739251040221913e-06,
+      "loss": 0.6857,
+      "step": 4280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.648493766784668,
+      "learning_rate": 7.734049930651872e-06,
+      "loss": 0.3913,
+      "step": 4281
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.08032512664795,
+      "learning_rate": 7.72884882108183e-06,
+      "loss": 0.6183,
+      "step": 4282
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.675776481628418,
+      "learning_rate": 7.72364771151179e-06,
+      "loss": 0.8492,
+      "step": 4283
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.954825401306152,
+      "learning_rate": 7.718446601941748e-06,
+      "loss": 0.3769,
+      "step": 4284
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.183074951171875,
+      "learning_rate": 7.713245492371706e-06,
+      "loss": 0.6202,
+      "step": 4285
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.09035587310791,
+      "learning_rate": 7.708044382801665e-06,
+      "loss": 0.5231,
+      "step": 4286
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.045661926269531,
+      "learning_rate": 7.702843273231624e-06,
+      "loss": 0.6332,
+      "step": 4287
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.717236518859863,
+      "learning_rate": 7.69764216366158e-06,
+      "loss": 0.6077,
+      "step": 4288
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.841323852539062,
+      "learning_rate": 7.692441054091539e-06,
+      "loss": 0.4893,
+      "step": 4289
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.438873291015625,
+      "learning_rate": 7.687239944521498e-06,
+      "loss": 0.6367,
+      "step": 4290
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.015273094177246,
+      "learning_rate": 7.682038834951456e-06,
+      "loss": 0.7863,
+      "step": 4291
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.140739917755127,
+      "learning_rate": 7.676837725381415e-06,
+      "loss": 0.3429,
+      "step": 4292
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.693789482116699,
+      "learning_rate": 7.671636615811374e-06,
+      "loss": 0.5593,
+      "step": 4293
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.456700325012207,
+      "learning_rate": 7.666435506241332e-06,
+      "loss": 0.4708,
+      "step": 4294
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.857933044433594,
+      "learning_rate": 7.66123439667129e-06,
+      "loss": 0.5777,
+      "step": 4295
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.191983222961426,
+      "learning_rate": 7.656033287101248e-06,
+      "loss": 0.6524,
+      "step": 4296
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.548505783081055,
+      "learning_rate": 7.650832177531206e-06,
+      "loss": 0.5664,
+      "step": 4297
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.674901962280273,
+      "learning_rate": 7.645631067961165e-06,
+      "loss": 0.5602,
+      "step": 4298
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.963614463806152,
+      "learning_rate": 7.640429958391124e-06,
+      "loss": 0.3776,
+      "step": 4299
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.67429256439209,
+      "learning_rate": 7.635228848821082e-06,
+      "loss": 0.6215,
+      "step": 4300
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.14966869354248,
+      "learning_rate": 7.63002773925104e-06,
+      "loss": 0.5093,
+      "step": 4301
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.12026309967041,
+      "learning_rate": 7.6248266296809986e-06,
+      "loss": 0.6485,
+      "step": 4302
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.083515167236328,
+      "learning_rate": 7.619625520110957e-06,
+      "loss": 0.5357,
+      "step": 4303
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.37206745147705,
+      "learning_rate": 7.614424410540916e-06,
+      "loss": 0.6783,
+      "step": 4304
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.424660682678223,
+      "learning_rate": 7.6092233009708736e-06,
+      "loss": 0.4857,
+      "step": 4305
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.054511070251465,
+      "learning_rate": 7.604022191400832e-06,
+      "loss": 0.67,
+      "step": 4306
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.609232902526855,
+      "learning_rate": 7.598821081830791e-06,
+      "loss": 0.4666,
+      "step": 4307
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.386672973632812,
+      "learning_rate": 7.593619972260749e-06,
+      "loss": 0.5033,
+      "step": 4308
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.781952857971191,
+      "learning_rate": 7.588418862690707e-06,
+      "loss": 0.4958,
+      "step": 4309
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.638063430786133,
+      "learning_rate": 7.583217753120666e-06,
+      "loss": 0.4331,
+      "step": 4310
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.766578674316406,
+      "learning_rate": 7.578016643550624e-06,
+      "loss": 0.5939,
+      "step": 4311
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.418108940124512,
+      "learning_rate": 7.572815533980583e-06,
+      "loss": 0.6529,
+      "step": 4312
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.247451782226562,
+      "learning_rate": 7.567614424410541e-06,
+      "loss": 0.4743,
+      "step": 4313
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.566956520080566,
+      "learning_rate": 7.562413314840499e-06,
+      "loss": 0.5741,
+      "step": 4314
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.020586013793945,
+      "learning_rate": 7.557212205270458e-06,
+      "loss": 0.6158,
+      "step": 4315
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.60345458984375,
+      "learning_rate": 7.552011095700417e-06,
+      "loss": 0.7338,
+      "step": 4316
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.434239864349365,
+      "learning_rate": 7.546809986130374e-06,
+      "loss": 0.4486,
+      "step": 4317
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.682068347930908,
+      "learning_rate": 7.541608876560333e-06,
+      "loss": 0.3953,
+      "step": 4318
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.243685722351074,
+      "learning_rate": 7.536407766990292e-06,
+      "loss": 0.6296,
+      "step": 4319
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.143321990966797,
+      "learning_rate": 7.53120665742025e-06,
+      "loss": 0.6677,
+      "step": 4320
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.402073860168457,
+      "learning_rate": 7.526005547850208e-06,
+      "loss": 0.5972,
+      "step": 4321
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.47900104522705,
+      "learning_rate": 7.520804438280167e-06,
+      "loss": 0.5134,
+      "step": 4322
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.579303741455078,
+      "learning_rate": 7.515603328710125e-06,
+      "loss": 0.6755,
+      "step": 4323
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.476795196533203,
+      "learning_rate": 7.510402219140084e-06,
+      "loss": 0.4712,
+      "step": 4324
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.90173053741455,
+      "learning_rate": 7.505201109570042e-06,
+      "loss": 0.5161,
+      "step": 4325
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.640264511108398,
+      "learning_rate": 7.5e-06,
+      "loss": 0.5684,
+      "step": 4326
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.295740127563477,
+      "learning_rate": 7.494798890429959e-06,
+      "loss": 0.3574,
+      "step": 4327
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.7251458168029785,
+      "learning_rate": 7.4895977808599166e-06,
+      "loss": 0.3376,
+      "step": 4328
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.6414713859558105,
+      "learning_rate": 7.484396671289875e-06,
+      "loss": 0.2837,
+      "step": 4329
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.81088399887085,
+      "learning_rate": 7.479195561719834e-06,
+      "loss": 0.4702,
+      "step": 4330
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.749281883239746,
+      "learning_rate": 7.473994452149792e-06,
+      "loss": 0.651,
+      "step": 4331
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.477560997009277,
+      "learning_rate": 7.46879334257975e-06,
+      "loss": 0.5649,
+      "step": 4332
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.622367858886719,
+      "learning_rate": 7.463592233009709e-06,
+      "loss": 0.6658,
+      "step": 4333
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.993042945861816,
+      "learning_rate": 7.458391123439667e-06,
+      "loss": 0.4318,
+      "step": 4334
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.411239624023438,
+      "learning_rate": 7.453190013869626e-06,
+      "loss": 0.6509,
+      "step": 4335
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.33577823638916,
+      "learning_rate": 7.447988904299584e-06,
+      "loss": 0.6734,
+      "step": 4336
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.497056007385254,
+      "learning_rate": 7.442787794729542e-06,
+      "loss": 0.5663,
+      "step": 4337
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.181666374206543,
+      "learning_rate": 7.437586685159501e-06,
+      "loss": 0.8656,
+      "step": 4338
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.441461563110352,
+      "learning_rate": 7.43238557558946e-06,
+      "loss": 0.489,
+      "step": 4339
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.065693855285645,
+      "learning_rate": 7.427184466019417e-06,
+      "loss": 0.6714,
+      "step": 4340
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.192416667938232,
+      "learning_rate": 7.421983356449376e-06,
+      "loss": 0.5024,
+      "step": 4341
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.179305076599121,
+      "learning_rate": 7.416782246879335e-06,
+      "loss": 0.6154,
+      "step": 4342
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.138778209686279,
+      "learning_rate": 7.411581137309293e-06,
+      "loss": 0.3859,
+      "step": 4343
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.576511859893799,
+      "learning_rate": 7.406380027739251e-06,
+      "loss": 0.5121,
+      "step": 4344
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.265250205993652,
+      "learning_rate": 7.40117891816921e-06,
+      "loss": 0.6406,
+      "step": 4345
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.39477252960205,
+      "learning_rate": 7.395977808599168e-06,
+      "loss": 0.6695,
+      "step": 4346
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.421689987182617,
+      "learning_rate": 7.390776699029127e-06,
+      "loss": 0.4669,
+      "step": 4347
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.893508434295654,
+      "learning_rate": 7.385575589459085e-06,
+      "loss": 0.5533,
+      "step": 4348
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.162010192871094,
+      "learning_rate": 7.380374479889043e-06,
+      "loss": 0.5403,
+      "step": 4349
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.332808494567871,
+      "learning_rate": 7.375173370319002e-06,
+      "loss": 0.5082,
+      "step": 4350
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.95407772064209,
+      "learning_rate": 7.36997226074896e-06,
+      "loss": 0.5064,
+      "step": 4351
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.878086090087891,
+      "learning_rate": 7.364771151178918e-06,
+      "loss": 0.3474,
+      "step": 4352
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.826037406921387,
+      "learning_rate": 7.359570041608877e-06,
+      "loss": 0.6458,
+      "step": 4353
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.475729942321777,
+      "learning_rate": 7.354368932038835e-06,
+      "loss": 0.5023,
+      "step": 4354
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.646228790283203,
+      "learning_rate": 7.349167822468793e-06,
+      "loss": 0.7203,
+      "step": 4355
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.663756370544434,
+      "learning_rate": 7.343966712898752e-06,
+      "loss": 0.5612,
+      "step": 4356
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.432422637939453,
+      "learning_rate": 7.33876560332871e-06,
+      "loss": 0.5836,
+      "step": 4357
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.597227096557617,
+      "learning_rate": 7.333564493758669e-06,
+      "loss": 0.5755,
+      "step": 4358
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.239532470703125,
+      "learning_rate": 7.328363384188627e-06,
+      "loss": 0.9391,
+      "step": 4359
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.65991497039795,
+      "learning_rate": 7.323162274618585e-06,
+      "loss": 0.5839,
+      "step": 4360
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.248530387878418,
+      "learning_rate": 7.317961165048544e-06,
+      "loss": 0.5685,
+      "step": 4361
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.134427070617676,
+      "learning_rate": 7.312760055478503e-06,
+      "loss": 0.4315,
+      "step": 4362
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.760110855102539,
+      "learning_rate": 7.30755894590846e-06,
+      "loss": 0.7689,
+      "step": 4363
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.503153800964355,
+      "learning_rate": 7.302357836338419e-06,
+      "loss": 0.6562,
+      "step": 4364
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.33425235748291,
+      "learning_rate": 7.297156726768378e-06,
+      "loss": 0.4593,
+      "step": 4365
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.018160820007324,
+      "learning_rate": 7.291955617198336e-06,
+      "loss": 0.7237,
+      "step": 4366
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.20079231262207,
+      "learning_rate": 7.286754507628294e-06,
+      "loss": 0.718,
+      "step": 4367
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.262954235076904,
+      "learning_rate": 7.281553398058253e-06,
+      "loss": 0.2732,
+      "step": 4368
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.969773292541504,
+      "learning_rate": 7.276352288488211e-06,
+      "loss": 0.731,
+      "step": 4369
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.648700714111328,
+      "learning_rate": 7.27115117891817e-06,
+      "loss": 0.8298,
+      "step": 4370
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.20981216430664,
+      "learning_rate": 7.265950069348128e-06,
+      "loss": 0.624,
+      "step": 4371
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.294867515563965,
+      "learning_rate": 7.260748959778086e-06,
+      "loss": 0.5559,
+      "step": 4372
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.168396949768066,
+      "learning_rate": 7.255547850208045e-06,
+      "loss": 0.7218,
+      "step": 4373
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.793280601501465,
+      "learning_rate": 7.250346740638003e-06,
+      "loss": 0.5731,
+      "step": 4374
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.951629638671875,
+      "learning_rate": 7.245145631067961e-06,
+      "loss": 0.5636,
+      "step": 4375
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.70508861541748,
+      "learning_rate": 7.23994452149792e-06,
+      "loss": 0.581,
+      "step": 4376
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.057461738586426,
+      "learning_rate": 7.234743411927878e-06,
+      "loss": 0.6905,
+      "step": 4377
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.863632678985596,
+      "learning_rate": 7.229542302357837e-06,
+      "loss": 0.4943,
+      "step": 4378
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.559826850891113,
+      "learning_rate": 7.224341192787795e-06,
+      "loss": 0.6115,
+      "step": 4379
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.4469194412231445,
+      "learning_rate": 7.219140083217753e-06,
+      "loss": 0.4772,
+      "step": 4380
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.657947540283203,
+      "learning_rate": 7.213938973647712e-06,
+      "loss": 0.6634,
+      "step": 4381
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.880634307861328,
+      "learning_rate": 7.208737864077671e-06,
+      "loss": 0.5562,
+      "step": 4382
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.728498458862305,
+      "learning_rate": 7.203536754507628e-06,
+      "loss": 0.4411,
+      "step": 4383
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.558417320251465,
+      "learning_rate": 7.198335644937587e-06,
+      "loss": 0.4466,
+      "step": 4384
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.570333957672119,
+      "learning_rate": 7.193134535367546e-06,
+      "loss": 0.5757,
+      "step": 4385
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.362052917480469,
+      "learning_rate": 7.187933425797503e-06,
+      "loss": 0.4331,
+      "step": 4386
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.479609489440918,
+      "learning_rate": 7.182732316227462e-06,
+      "loss": 0.5064,
+      "step": 4387
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.962183952331543,
+      "learning_rate": 7.177531206657421e-06,
+      "loss": 0.5087,
+      "step": 4388
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.520251274108887,
+      "learning_rate": 7.172330097087379e-06,
+      "loss": 0.6821,
+      "step": 4389
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.301064491271973,
+      "learning_rate": 7.167128987517337e-06,
+      "loss": 0.4613,
+      "step": 4390
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.220833778381348,
+      "learning_rate": 7.161927877947296e-06,
+      "loss": 0.4038,
+      "step": 4391
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.604326248168945,
+      "learning_rate": 7.156726768377254e-06,
+      "loss": 0.6028,
+      "step": 4392
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.479988098144531,
+      "learning_rate": 7.151525658807213e-06,
+      "loss": 0.5304,
+      "step": 4393
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.298116683959961,
+      "learning_rate": 7.146324549237171e-06,
+      "loss": 0.7593,
+      "step": 4394
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.741998672485352,
+      "learning_rate": 7.141123439667129e-06,
+      "loss": 0.6752,
+      "step": 4395
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.229445457458496,
+      "learning_rate": 7.135922330097088e-06,
+      "loss": 0.5993,
+      "step": 4396
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.226974487304688,
+      "learning_rate": 7.1307212205270464e-06,
+      "loss": 0.7558,
+      "step": 4397
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.45487117767334,
+      "learning_rate": 7.125520110957004e-06,
+      "loss": 0.4015,
+      "step": 4398
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.824450492858887,
+      "learning_rate": 7.120319001386963e-06,
+      "loss": 0.4238,
+      "step": 4399
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.849211692810059,
+      "learning_rate": 7.115117891816921e-06,
+      "loss": 0.5558,
+      "step": 4400
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.239603042602539,
+      "learning_rate": 7.10991678224688e-06,
+      "loss": 0.6404,
+      "step": 4401
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.121785163879395,
+      "learning_rate": 7.104715672676838e-06,
+      "loss": 0.7871,
+      "step": 4402
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.636196136474609,
+      "learning_rate": 7.099514563106796e-06,
+      "loss": 0.4659,
+      "step": 4403
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.6942973136901855,
+      "learning_rate": 7.094313453536755e-06,
+      "loss": 0.3701,
+      "step": 4404
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.988786697387695,
+      "learning_rate": 7.089112343966714e-06,
+      "loss": 0.451,
+      "step": 4405
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.156188011169434,
+      "learning_rate": 7.083911234396671e-06,
+      "loss": 0.5068,
+      "step": 4406
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.416943550109863,
+      "learning_rate": 7.07871012482663e-06,
+      "loss": 0.4636,
+      "step": 4407
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.6533708572387695,
+      "learning_rate": 7.073509015256589e-06,
+      "loss": 0.5959,
+      "step": 4408
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.939712047576904,
+      "learning_rate": 7.068307905686547e-06,
+      "loss": 0.3399,
+      "step": 4409
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.941202163696289,
+      "learning_rate": 7.063106796116505e-06,
+      "loss": 0.4684,
+      "step": 4410
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.147394180297852,
+      "learning_rate": 7.057905686546464e-06,
+      "loss": 0.4449,
+      "step": 4411
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.369781494140625,
+      "learning_rate": 7.052704576976422e-06,
+      "loss": 0.6824,
+      "step": 4412
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.165634155273438,
+      "learning_rate": 7.04750346740638e-06,
+      "loss": 0.5587,
+      "step": 4413
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.812243461608887,
+      "learning_rate": 7.042302357836339e-06,
+      "loss": 0.5498,
+      "step": 4414
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.252660751342773,
+      "learning_rate": 7.037101248266297e-06,
+      "loss": 0.56,
+      "step": 4415
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.106646537780762,
+      "learning_rate": 7.031900138696256e-06,
+      "loss": 0.5908,
+      "step": 4416
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.74988842010498,
+      "learning_rate": 7.026699029126214e-06,
+      "loss": 0.5865,
+      "step": 4417
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.716312408447266,
+      "learning_rate": 7.021497919556172e-06,
+      "loss": 0.6733,
+      "step": 4418
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.273759841918945,
+      "learning_rate": 7.016296809986131e-06,
+      "loss": 0.5932,
+      "step": 4419
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.9290900230407715,
+      "learning_rate": 7.0110957004160894e-06,
+      "loss": 0.469,
+      "step": 4420
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.140643119812012,
+      "learning_rate": 7.005894590846047e-06,
+      "loss": 0.5472,
+      "step": 4421
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.967321395874023,
+      "learning_rate": 7.000693481276006e-06,
+      "loss": 0.4893,
+      "step": 4422
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.99087905883789,
+      "learning_rate": 6.995492371705964e-06,
+      "loss": 0.5588,
+      "step": 4423
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.682893753051758,
+      "learning_rate": 6.990291262135923e-06,
+      "loss": 0.541,
+      "step": 4424
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.070243835449219,
+      "learning_rate": 6.985090152565881e-06,
+      "loss": 0.5543,
+      "step": 4425
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.725927352905273,
+      "learning_rate": 6.979889042995839e-06,
+      "loss": 0.5718,
+      "step": 4426
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.693760871887207,
+      "learning_rate": 6.974687933425798e-06,
+      "loss": 0.6184,
+      "step": 4427
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.592596054077148,
+      "learning_rate": 6.969486823855757e-06,
+      "loss": 0.5574,
+      "step": 4428
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.673807144165039,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 0.5913,
+      "step": 4429
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.758624076843262,
+      "learning_rate": 6.959084604715673e-06,
+      "loss": 0.5826,
+      "step": 4430
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.930179119110107,
+      "learning_rate": 6.953883495145632e-06,
+      "loss": 0.4171,
+      "step": 4431
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.22437858581543,
+      "learning_rate": 6.94868238557559e-06,
+      "loss": 0.6349,
+      "step": 4432
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.040468215942383,
+      "learning_rate": 6.943481276005548e-06,
+      "loss": 0.5684,
+      "step": 4433
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.889357566833496,
+      "learning_rate": 6.938280166435507e-06,
+      "loss": 0.4039,
+      "step": 4434
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.8134765625,
+      "learning_rate": 6.933079056865465e-06,
+      "loss": 0.4883,
+      "step": 4435
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.289885520935059,
+      "learning_rate": 6.927877947295424e-06,
+      "loss": 0.7655,
+      "step": 4436
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.719887733459473,
+      "learning_rate": 6.922676837725382e-06,
+      "loss": 0.5648,
+      "step": 4437
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.434460639953613,
+      "learning_rate": 6.91747572815534e-06,
+      "loss": 0.4216,
+      "step": 4438
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.486008644104004,
+      "learning_rate": 6.912274618585299e-06,
+      "loss": 0.427,
+      "step": 4439
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.806597709655762,
+      "learning_rate": 6.9070735090152574e-06,
+      "loss": 0.6272,
+      "step": 4440
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.558038711547852,
+      "learning_rate": 6.901872399445215e-06,
+      "loss": 0.6051,
+      "step": 4441
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.224431991577148,
+      "learning_rate": 6.896671289875174e-06,
+      "loss": 0.7484,
+      "step": 4442
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.6970853805542,
+      "learning_rate": 6.8914701803051324e-06,
+      "loss": 0.4194,
+      "step": 4443
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.690580368041992,
+      "learning_rate": 6.88626907073509e-06,
+      "loss": 0.4825,
+      "step": 4444
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.449708938598633,
+      "learning_rate": 6.881067961165049e-06,
+      "loss": 0.8997,
+      "step": 4445
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.904961585998535,
+      "learning_rate": 6.8758668515950074e-06,
+      "loss": 0.7325,
+      "step": 4446
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.707748413085938,
+      "learning_rate": 6.870665742024966e-06,
+      "loss": 0.6934,
+      "step": 4447
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.434955596923828,
+      "learning_rate": 6.865464632454924e-06,
+      "loss": 0.4876,
+      "step": 4448
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.530936241149902,
+      "learning_rate": 6.860263522884882e-06,
+      "loss": 0.4194,
+      "step": 4449
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.412516593933105,
+      "learning_rate": 6.855062413314841e-06,
+      "loss": 0.5388,
+      "step": 4450
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.071410179138184,
+      "learning_rate": 6.8498613037448e-06,
+      "loss": 0.6128,
+      "step": 4451
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.549949645996094,
+      "learning_rate": 6.844660194174757e-06,
+      "loss": 0.6451,
+      "step": 4452
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.661032676696777,
+      "learning_rate": 6.839459084604716e-06,
+      "loss": 0.7227,
+      "step": 4453
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.13907241821289,
+      "learning_rate": 6.834257975034675e-06,
+      "loss": 0.3817,
+      "step": 4454
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.86040210723877,
+      "learning_rate": 6.829056865464633e-06,
+      "loss": 0.6459,
+      "step": 4455
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.246345520019531,
+      "learning_rate": 6.823855755894591e-06,
+      "loss": 0.6123,
+      "step": 4456
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.18960952758789,
+      "learning_rate": 6.81865464632455e-06,
+      "loss": 0.7054,
+      "step": 4457
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.510292053222656,
+      "learning_rate": 6.813453536754508e-06,
+      "loss": 0.474,
+      "step": 4458
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.142463684082031,
+      "learning_rate": 6.808252427184467e-06,
+      "loss": 0.3871,
+      "step": 4459
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.387166023254395,
+      "learning_rate": 6.803051317614425e-06,
+      "loss": 0.4871,
+      "step": 4460
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.050041198730469,
+      "learning_rate": 6.797850208044383e-06,
+      "loss": 0.4229,
+      "step": 4461
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.677462577819824,
+      "learning_rate": 6.792649098474342e-06,
+      "loss": 0.4192,
+      "step": 4462
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.539796829223633,
+      "learning_rate": 6.7874479889043004e-06,
+      "loss": 0.766,
+      "step": 4463
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.172012329101562,
+      "learning_rate": 6.782246879334258e-06,
+      "loss": 0.6833,
+      "step": 4464
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.846525192260742,
+      "learning_rate": 6.777045769764217e-06,
+      "loss": 0.5713,
+      "step": 4465
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.758662223815918,
+      "learning_rate": 6.7718446601941754e-06,
+      "loss": 0.6088,
+      "step": 4466
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.145308494567871,
+      "learning_rate": 6.766643550624134e-06,
+      "loss": 0.5522,
+      "step": 4467
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.576066970825195,
+      "learning_rate": 6.761442441054092e-06,
+      "loss": 0.6419,
+      "step": 4468
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.574542999267578,
+      "learning_rate": 6.7562413314840504e-06,
+      "loss": 0.5325,
+      "step": 4469
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.885310173034668,
+      "learning_rate": 6.751040221914009e-06,
+      "loss": 0.5315,
+      "step": 4470
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.407565116882324,
+      "learning_rate": 6.745839112343967e-06,
+      "loss": 0.5612,
+      "step": 4471
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.623705863952637,
+      "learning_rate": 6.740638002773925e-06,
+      "loss": 0.5955,
+      "step": 4472
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.04969310760498,
+      "learning_rate": 6.735436893203884e-06,
+      "loss": 0.5257,
+      "step": 4473
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.997536659240723,
+      "learning_rate": 6.730235783633843e-06,
+      "loss": 0.77,
+      "step": 4474
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.355415344238281,
+      "learning_rate": 6.7250346740638e-06,
+      "loss": 0.4885,
+      "step": 4475
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.179099082946777,
+      "learning_rate": 6.719833564493759e-06,
+      "loss": 0.4347,
+      "step": 4476
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.92045783996582,
+      "learning_rate": 6.714632454923718e-06,
+      "loss": 0.7477,
+      "step": 4477
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.073054313659668,
+      "learning_rate": 6.709431345353676e-06,
+      "loss": 0.4964,
+      "step": 4478
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.011826515197754,
+      "learning_rate": 6.704230235783634e-06,
+      "loss": 0.5605,
+      "step": 4479
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.45168685913086,
+      "learning_rate": 6.699029126213593e-06,
+      "loss": 0.8483,
+      "step": 4480
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.440661430358887,
+      "learning_rate": 6.693828016643551e-06,
+      "loss": 0.7024,
+      "step": 4481
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.831561088562012,
+      "learning_rate": 6.68862690707351e-06,
+      "loss": 0.677,
+      "step": 4482
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.124338150024414,
+      "learning_rate": 6.683425797503468e-06,
+      "loss": 0.6079,
+      "step": 4483
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.283355236053467,
+      "learning_rate": 6.678224687933426e-06,
+      "loss": 0.4746,
+      "step": 4484
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.086446762084961,
+      "learning_rate": 6.673023578363385e-06,
+      "loss": 0.5942,
+      "step": 4485
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.63311767578125,
+      "learning_rate": 6.6678224687933435e-06,
+      "loss": 0.6031,
+      "step": 4486
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.430500984191895,
+      "learning_rate": 6.662621359223301e-06,
+      "loss": 0.4856,
+      "step": 4487
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.953193664550781,
+      "learning_rate": 6.65742024965326e-06,
+      "loss": 0.4792,
+      "step": 4488
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.57575798034668,
+      "learning_rate": 6.6522191400832184e-06,
+      "loss": 0.7524,
+      "step": 4489
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.052308082580566,
+      "learning_rate": 6.647018030513177e-06,
+      "loss": 0.6172,
+      "step": 4490
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.29978084564209,
+      "learning_rate": 6.641816920943135e-06,
+      "loss": 0.6877,
+      "step": 4491
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.1370930671691895,
+      "learning_rate": 6.6366158113730934e-06,
+      "loss": 0.5243,
+      "step": 4492
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.911317348480225,
+      "learning_rate": 6.631414701803052e-06,
+      "loss": 0.4926,
+      "step": 4493
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.4662446975708,
+      "learning_rate": 6.626213592233011e-06,
+      "loss": 0.4863,
+      "step": 4494
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.402633666992188,
+      "learning_rate": 6.621012482662968e-06,
+      "loss": 0.9783,
+      "step": 4495
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.01633358001709,
+      "learning_rate": 6.615811373092927e-06,
+      "loss": 0.5542,
+      "step": 4496
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.498165130615234,
+      "learning_rate": 6.610610263522886e-06,
+      "loss": 0.5217,
+      "step": 4497
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.463785171508789,
+      "learning_rate": 6.605409153952844e-06,
+      "loss": 0.7177,
+      "step": 4498
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.894790649414062,
+      "learning_rate": 6.600208044382802e-06,
+      "loss": 0.8518,
+      "step": 4499
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.285511016845703,
+      "learning_rate": 6.595006934812761e-06,
+      "loss": 0.6555,
+      "step": 4500
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.179150581359863,
+      "learning_rate": 6.589805825242719e-06,
+      "loss": 0.4893,
+      "step": 4501
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.653487205505371,
+      "learning_rate": 6.584604715672677e-06,
+      "loss": 0.5316,
+      "step": 4502
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.710869789123535,
+      "learning_rate": 6.579403606102636e-06,
+      "loss": 0.684,
+      "step": 4503
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.22343635559082,
+      "learning_rate": 6.574202496532594e-06,
+      "loss": 0.4946,
+      "step": 4504
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.27607250213623,
+      "learning_rate": 6.569001386962553e-06,
+      "loss": 0.8848,
+      "step": 4505
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.971867561340332,
+      "learning_rate": 6.563800277392511e-06,
+      "loss": 0.5186,
+      "step": 4506
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.35292911529541,
+      "learning_rate": 6.558599167822468e-06,
+      "loss": 0.6035,
+      "step": 4507
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.926923751831055,
+      "learning_rate": 6.553398058252427e-06,
+      "loss": 0.6367,
+      "step": 4508
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.297164916992188,
+      "learning_rate": 6.548196948682386e-06,
+      "loss": 0.628,
+      "step": 4509
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.963873863220215,
+      "learning_rate": 6.542995839112343e-06,
+      "loss": 0.6407,
+      "step": 4510
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.813761711120605,
+      "learning_rate": 6.537794729542302e-06,
+      "loss": 0.5166,
+      "step": 4511
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.959879875183105,
+      "learning_rate": 6.532593619972261e-06,
+      "loss": 0.6102,
+      "step": 4512
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.85804271697998,
+      "learning_rate": 6.527392510402219e-06,
+      "loss": 0.4972,
+      "step": 4513
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.04681396484375,
+      "learning_rate": 6.522191400832177e-06,
+      "loss": 0.4687,
+      "step": 4514
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.48231029510498,
+      "learning_rate": 6.516990291262136e-06,
+      "loss": 0.6226,
+      "step": 4515
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.555724143981934,
+      "learning_rate": 6.511789181692094e-06,
+      "loss": 0.8331,
+      "step": 4516
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.663237571716309,
+      "learning_rate": 6.506588072122053e-06,
+      "loss": 0.527,
+      "step": 4517
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.576103210449219,
+      "learning_rate": 6.501386962552011e-06,
+      "loss": 0.6482,
+      "step": 4518
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.165956497192383,
+      "learning_rate": 6.496185852981969e-06,
+      "loss": 0.3585,
+      "step": 4519
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.671982765197754,
+      "learning_rate": 6.490984743411928e-06,
+      "loss": 0.5605,
+      "step": 4520
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.65974235534668,
+      "learning_rate": 6.485783633841886e-06,
+      "loss": 0.5106,
+      "step": 4521
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.536144256591797,
+      "learning_rate": 6.480582524271844e-06,
+      "loss": 0.5178,
+      "step": 4522
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.815608024597168,
+      "learning_rate": 6.475381414701803e-06,
+      "loss": 0.4689,
+      "step": 4523
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.665443420410156,
+      "learning_rate": 6.470180305131761e-06,
+      "loss": 0.5929,
+      "step": 4524
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.40139389038086,
+      "learning_rate": 6.46497919556172e-06,
+      "loss": 0.4919,
+      "step": 4525
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.3502197265625,
+      "learning_rate": 6.459778085991678e-06,
+      "loss": 0.6825,
+      "step": 4526
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.61600399017334,
+      "learning_rate": 6.454576976421636e-06,
+      "loss": 0.7504,
+      "step": 4527
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.970343112945557,
+      "learning_rate": 6.449375866851595e-06,
+      "loss": 0.4257,
+      "step": 4528
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.864808082580566,
+      "learning_rate": 6.444174757281553e-06,
+      "loss": 0.7141,
+      "step": 4529
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.176302909851074,
+      "learning_rate": 6.438973647711511e-06,
+      "loss": 0.4578,
+      "step": 4530
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.032179832458496,
+      "learning_rate": 6.43377253814147e-06,
+      "loss": 0.6713,
+      "step": 4531
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.468162536621094,
+      "learning_rate": 6.428571428571429e-06,
+      "loss": 0.691,
+      "step": 4532
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.331866264343262,
+      "learning_rate": 6.423370319001386e-06,
+      "loss": 0.627,
+      "step": 4533
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.430709838867188,
+      "learning_rate": 6.418169209431345e-06,
+      "loss": 0.5361,
+      "step": 4534
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.199605941772461,
+      "learning_rate": 6.412968099861304e-06,
+      "loss": 0.5214,
+      "step": 4535
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.830657005310059,
+      "learning_rate": 6.407766990291262e-06,
+      "loss": 0.5143,
+      "step": 4536
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.841787338256836,
+      "learning_rate": 6.40256588072122e-06,
+      "loss": 0.6246,
+      "step": 4537
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.6421122550964355,
+      "learning_rate": 6.397364771151179e-06,
+      "loss": 0.5663,
+      "step": 4538
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.312774658203125,
+      "learning_rate": 6.392163661581137e-06,
+      "loss": 0.7082,
+      "step": 4539
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.094574451446533,
+      "learning_rate": 6.386962552011096e-06,
+      "loss": 0.433,
+      "step": 4540
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.103632926940918,
+      "learning_rate": 6.381761442441054e-06,
+      "loss": 0.3746,
+      "step": 4541
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.94027328491211,
+      "learning_rate": 6.376560332871012e-06,
+      "loss": 0.6653,
+      "step": 4542
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.982824325561523,
+      "learning_rate": 6.371359223300971e-06,
+      "loss": 0.5428,
+      "step": 4543
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.473831176757812,
+      "learning_rate": 6.366158113730929e-06,
+      "loss": 0.5944,
+      "step": 4544
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.992941856384277,
+      "learning_rate": 6.360957004160887e-06,
+      "loss": 0.5653,
+      "step": 4545
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.311532974243164,
+      "learning_rate": 6.355755894590846e-06,
+      "loss": 0.6149,
+      "step": 4546
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.262786865234375,
+      "learning_rate": 6.350554785020804e-06,
+      "loss": 0.7343,
+      "step": 4547
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.517559051513672,
+      "learning_rate": 6.345353675450763e-06,
+      "loss": 0.6798,
+      "step": 4548
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.12297248840332,
+      "learning_rate": 6.340152565880721e-06,
+      "loss": 0.5044,
+      "step": 4549
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.601327896118164,
+      "learning_rate": 6.334951456310679e-06,
+      "loss": 0.5871,
+      "step": 4550
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.488725662231445,
+      "learning_rate": 6.329750346740638e-06,
+      "loss": 0.3711,
+      "step": 4551
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.98576545715332,
+      "learning_rate": 6.324549237170597e-06,
+      "loss": 0.6171,
+      "step": 4552
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.932962417602539,
+      "learning_rate": 6.319348127600554e-06,
+      "loss": 0.5268,
+      "step": 4553
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.685565948486328,
+      "learning_rate": 6.314147018030513e-06,
+      "loss": 0.5146,
+      "step": 4554
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.919050216674805,
+      "learning_rate": 6.308945908460472e-06,
+      "loss": 0.5299,
+      "step": 4555
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.189722061157227,
+      "learning_rate": 6.30374479889043e-06,
+      "loss": 0.7001,
+      "step": 4556
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.739821434020996,
+      "learning_rate": 6.298543689320388e-06,
+      "loss": 0.5382,
+      "step": 4557
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.862648963928223,
+      "learning_rate": 6.293342579750347e-06,
+      "loss": 0.299,
+      "step": 4558
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.414625644683838,
+      "learning_rate": 6.288141470180305e-06,
+      "loss": 0.3986,
+      "step": 4559
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.675583839416504,
+      "learning_rate": 6.282940360610263e-06,
+      "loss": 0.6135,
+      "step": 4560
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.534601211547852,
+      "learning_rate": 6.277739251040222e-06,
+      "loss": 0.5602,
+      "step": 4561
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.562333583831787,
+      "learning_rate": 6.27253814147018e-06,
+      "loss": 0.3032,
+      "step": 4562
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.365431785583496,
+      "learning_rate": 6.267337031900139e-06,
+      "loss": 0.4821,
+      "step": 4563
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.382303237915039,
+      "learning_rate": 6.262135922330097e-06,
+      "loss": 0.6031,
+      "step": 4564
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.555085182189941,
+      "learning_rate": 6.256934812760055e-06,
+      "loss": 0.5327,
+      "step": 4565
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.18924331665039,
+      "learning_rate": 6.251733703190014e-06,
+      "loss": 0.4522,
+      "step": 4566
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.133987426757812,
+      "learning_rate": 6.2465325936199724e-06,
+      "loss": 0.5936,
+      "step": 4567
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.584412574768066,
+      "learning_rate": 6.24133148404993e-06,
+      "loss": 0.6192,
+      "step": 4568
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.222260475158691,
+      "learning_rate": 6.236130374479889e-06,
+      "loss": 0.6625,
+      "step": 4569
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.422402381896973,
+      "learning_rate": 6.230929264909847e-06,
+      "loss": 0.6236,
+      "step": 4570
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.233536720275879,
+      "learning_rate": 6.225728155339806e-06,
+      "loss": 0.7199,
+      "step": 4571
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.142074584960938,
+      "learning_rate": 6.220527045769764e-06,
+      "loss": 0.4353,
+      "step": 4572
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.094156265258789,
+      "learning_rate": 6.215325936199722e-06,
+      "loss": 0.5545,
+      "step": 4573
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.945605278015137,
+      "learning_rate": 6.210124826629681e-06,
+      "loss": 0.6989,
+      "step": 4574
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.65359115600586,
+      "learning_rate": 6.20492371705964e-06,
+      "loss": 0.5839,
+      "step": 4575
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.10034465789795,
+      "learning_rate": 6.199722607489597e-06,
+      "loss": 0.612,
+      "step": 4576
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.373862266540527,
+      "learning_rate": 6.194521497919556e-06,
+      "loss": 0.577,
+      "step": 4577
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.763827323913574,
+      "learning_rate": 6.189320388349515e-06,
+      "loss": 0.4144,
+      "step": 4578
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.583927154541016,
+      "learning_rate": 6.184119278779473e-06,
+      "loss": 0.6278,
+      "step": 4579
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.352312088012695,
+      "learning_rate": 6.178918169209431e-06,
+      "loss": 0.4575,
+      "step": 4580
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.431353569030762,
+      "learning_rate": 6.17371705963939e-06,
+      "loss": 0.6137,
+      "step": 4581
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.074173927307129,
+      "learning_rate": 6.168515950069348e-06,
+      "loss": 0.4935,
+      "step": 4582
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.539366722106934,
+      "learning_rate": 6.163314840499307e-06,
+      "loss": 0.5461,
+      "step": 4583
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.34234619140625,
+      "learning_rate": 6.158113730929265e-06,
+      "loss": 0.4615,
+      "step": 4584
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.205765724182129,
+      "learning_rate": 6.152912621359223e-06,
+      "loss": 0.6207,
+      "step": 4585
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.45760440826416,
+      "learning_rate": 6.147711511789182e-06,
+      "loss": 0.5939,
+      "step": 4586
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.302465438842773,
+      "learning_rate": 6.14251040221914e-06,
+      "loss": 0.4923,
+      "step": 4587
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.905799865722656,
+      "learning_rate": 6.137309292649098e-06,
+      "loss": 0.6377,
+      "step": 4588
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.857898712158203,
+      "learning_rate": 6.132108183079057e-06,
+      "loss": 0.7527,
+      "step": 4589
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.1714448928833,
+      "learning_rate": 6.1269070735090154e-06,
+      "loss": 0.6692,
+      "step": 4590
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.891768455505371,
+      "learning_rate": 6.121705963938973e-06,
+      "loss": 0.3974,
+      "step": 4591
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.985422134399414,
+      "learning_rate": 6.116504854368932e-06,
+      "loss": 0.4449,
+      "step": 4592
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.910746574401855,
+      "learning_rate": 6.11130374479889e-06,
+      "loss": 0.6187,
+      "step": 4593
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.01278018951416,
+      "learning_rate": 6.106102635228849e-06,
+      "loss": 0.4063,
+      "step": 4594
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.295592308044434,
+      "learning_rate": 6.100901525658807e-06,
+      "loss": 0.6672,
+      "step": 4595
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.298104286193848,
+      "learning_rate": 6.095700416088765e-06,
+      "loss": 0.6188,
+      "step": 4596
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.946653366088867,
+      "learning_rate": 6.090499306518724e-06,
+      "loss": 0.6178,
+      "step": 4597
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.994405746459961,
+      "learning_rate": 6.085298196948683e-06,
+      "loss": 0.5983,
+      "step": 4598
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.542595863342285,
+      "learning_rate": 6.08009708737864e-06,
+      "loss": 0.6121,
+      "step": 4599
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.009209156036377,
+      "learning_rate": 6.074895977808599e-06,
+      "loss": 0.4075,
+      "step": 4600
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.3133320808410645,
+      "learning_rate": 6.069694868238558e-06,
+      "loss": 0.4564,
+      "step": 4601
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.347931861877441,
+      "learning_rate": 6.064493758668516e-06,
+      "loss": 0.4257,
+      "step": 4602
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.039680480957031,
+      "learning_rate": 6.059292649098474e-06,
+      "loss": 0.5576,
+      "step": 4603
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.166584968566895,
+      "learning_rate": 6.054091539528433e-06,
+      "loss": 0.6452,
+      "step": 4604
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.9139933586120605,
+      "learning_rate": 6.048890429958391e-06,
+      "loss": 0.4927,
+      "step": 4605
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.786900520324707,
+      "learning_rate": 6.04368932038835e-06,
+      "loss": 0.4006,
+      "step": 4606
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.418083190917969,
+      "learning_rate": 6.038488210818308e-06,
+      "loss": 0.493,
+      "step": 4607
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.27340316772461,
+      "learning_rate": 6.033287101248266e-06,
+      "loss": 0.6227,
+      "step": 4608
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.353404998779297,
+      "learning_rate": 6.028085991678225e-06,
+      "loss": 0.7556,
+      "step": 4609
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.997955322265625,
+      "learning_rate": 6.0228848821081834e-06,
+      "loss": 0.6237,
+      "step": 4610
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.979578971862793,
+      "learning_rate": 6.017683772538141e-06,
+      "loss": 0.7581,
+      "step": 4611
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.307106018066406,
+      "learning_rate": 6.0124826629681e-06,
+      "loss": 0.4711,
+      "step": 4612
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.932023048400879,
+      "learning_rate": 6.0072815533980584e-06,
+      "loss": 0.5408,
+      "step": 4613
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.715838432312012,
+      "learning_rate": 6.002080443828017e-06,
+      "loss": 0.4705,
+      "step": 4614
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.960479736328125,
+      "learning_rate": 5.996879334257975e-06,
+      "loss": 0.3701,
+      "step": 4615
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.47189998626709,
+      "learning_rate": 5.9916782246879334e-06,
+      "loss": 0.6459,
+      "step": 4616
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.412280082702637,
+      "learning_rate": 5.986477115117892e-06,
+      "loss": 0.5067,
+      "step": 4617
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.637069702148438,
+      "learning_rate": 5.98127600554785e-06,
+      "loss": 0.584,
+      "step": 4618
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.741731643676758,
+      "learning_rate": 5.976074895977808e-06,
+      "loss": 0.4302,
+      "step": 4619
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.752028465270996,
+      "learning_rate": 5.970873786407767e-06,
+      "loss": 0.5626,
+      "step": 4620
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.182929992675781,
+      "learning_rate": 5.965672676837726e-06,
+      "loss": 0.6016,
+      "step": 4621
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.07071590423584,
+      "learning_rate": 5.960471567267683e-06,
+      "loss": 0.6628,
+      "step": 4622
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.95657730102539,
+      "learning_rate": 5.955270457697642e-06,
+      "loss": 0.7281,
+      "step": 4623
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.644377708435059,
+      "learning_rate": 5.950069348127601e-06,
+      "loss": 0.4162,
+      "step": 4624
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.719254493713379,
+      "learning_rate": 5.944868238557559e-06,
+      "loss": 0.5797,
+      "step": 4625
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.57142448425293,
+      "learning_rate": 5.939667128987517e-06,
+      "loss": 0.6166,
+      "step": 4626
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.126080513000488,
+      "learning_rate": 5.934466019417476e-06,
+      "loss": 0.4149,
+      "step": 4627
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.517499923706055,
+      "learning_rate": 5.929264909847434e-06,
+      "loss": 0.5216,
+      "step": 4628
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.29642391204834,
+      "learning_rate": 5.924063800277393e-06,
+      "loss": 0.6801,
+      "step": 4629
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.94295072555542,
+      "learning_rate": 5.918862690707351e-06,
+      "loss": 0.5558,
+      "step": 4630
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.286441802978516,
+      "learning_rate": 5.913661581137309e-06,
+      "loss": 0.5421,
+      "step": 4631
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.493871688842773,
+      "learning_rate": 5.908460471567268e-06,
+      "loss": 0.5496,
+      "step": 4632
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.905692100524902,
+      "learning_rate": 5.9032593619972264e-06,
+      "loss": 0.6103,
+      "step": 4633
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.19190788269043,
+      "learning_rate": 5.898058252427184e-06,
+      "loss": 0.7019,
+      "step": 4634
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.9011969566345215,
+      "learning_rate": 5.892857142857143e-06,
+      "loss": 0.3486,
+      "step": 4635
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.015719413757324,
+      "learning_rate": 5.8876560332871014e-06,
+      "loss": 0.6571,
+      "step": 4636
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.557942390441895,
+      "learning_rate": 5.88245492371706e-06,
+      "loss": 0.4489,
+      "step": 4637
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.364808082580566,
+      "learning_rate": 5.877253814147018e-06,
+      "loss": 0.3175,
+      "step": 4638
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.263394355773926,
+      "learning_rate": 5.8720527045769764e-06,
+      "loss": 0.5402,
+      "step": 4639
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.654337406158447,
+      "learning_rate": 5.866851595006935e-06,
+      "loss": 0.4881,
+      "step": 4640
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.503586769104004,
+      "learning_rate": 5.861650485436894e-06,
+      "loss": 0.515,
+      "step": 4641
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.070056915283203,
+      "learning_rate": 5.856449375866851e-06,
+      "loss": 0.5294,
+      "step": 4642
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.950750350952148,
+      "learning_rate": 5.85124826629681e-06,
+      "loss": 0.3558,
+      "step": 4643
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.561210632324219,
+      "learning_rate": 5.846047156726769e-06,
+      "loss": 0.3787,
+      "step": 4644
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.275738716125488,
+      "learning_rate": 5.840846047156726e-06,
+      "loss": 0.3965,
+      "step": 4645
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.019781112670898,
+      "learning_rate": 5.835644937586685e-06,
+      "loss": 0.6739,
+      "step": 4646
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.6217679977417,
+      "learning_rate": 5.830443828016644e-06,
+      "loss": 0.6476,
+      "step": 4647
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.410270690917969,
+      "learning_rate": 5.825242718446602e-06,
+      "loss": 0.4663,
+      "step": 4648
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.365642547607422,
+      "learning_rate": 5.82004160887656e-06,
+      "loss": 0.4112,
+      "step": 4649
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.673617362976074,
+      "learning_rate": 5.814840499306519e-06,
+      "loss": 0.6146,
+      "step": 4650
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.130807876586914,
+      "learning_rate": 5.809639389736477e-06,
+      "loss": 0.4475,
+      "step": 4651
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.238598823547363,
+      "learning_rate": 5.804438280166436e-06,
+      "loss": 0.5639,
+      "step": 4652
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.528800010681152,
+      "learning_rate": 5.799237170596394e-06,
+      "loss": 0.6635,
+      "step": 4653
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.320645332336426,
+      "learning_rate": 5.794036061026352e-06,
+      "loss": 0.6264,
+      "step": 4654
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.657713890075684,
+      "learning_rate": 5.788834951456311e-06,
+      "loss": 0.7618,
+      "step": 4655
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.465720176696777,
+      "learning_rate": 5.7836338418862695e-06,
+      "loss": 0.5909,
+      "step": 4656
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.046468734741211,
+      "learning_rate": 5.778432732316227e-06,
+      "loss": 0.6291,
+      "step": 4657
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.130772590637207,
+      "learning_rate": 5.773231622746186e-06,
+      "loss": 0.4079,
+      "step": 4658
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.921156883239746,
+      "learning_rate": 5.7680305131761444e-06,
+      "loss": 0.4657,
+      "step": 4659
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.657792091369629,
+      "learning_rate": 5.762829403606103e-06,
+      "loss": 0.4541,
+      "step": 4660
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.475892066955566,
+      "learning_rate": 5.757628294036061e-06,
+      "loss": 0.5832,
+      "step": 4661
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.011570930480957,
+      "learning_rate": 5.7524271844660194e-06,
+      "loss": 0.3215,
+      "step": 4662
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.4091157913208,
+      "learning_rate": 5.747226074895978e-06,
+      "loss": 0.5817,
+      "step": 4663
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.53343391418457,
+      "learning_rate": 5.742024965325937e-06,
+      "loss": 0.6411,
+      "step": 4664
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.044541358947754,
+      "learning_rate": 5.736823855755894e-06,
+      "loss": 0.5345,
+      "step": 4665
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.4736909866333,
+      "learning_rate": 5.731622746185853e-06,
+      "loss": 0.5965,
+      "step": 4666
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.884674072265625,
+      "learning_rate": 5.726421636615812e-06,
+      "loss": 0.5523,
+      "step": 4667
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.143738746643066,
+      "learning_rate": 5.72122052704577e-06,
+      "loss": 0.5938,
+      "step": 4668
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.052058219909668,
+      "learning_rate": 5.716019417475728e-06,
+      "loss": 0.798,
+      "step": 4669
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.675034523010254,
+      "learning_rate": 5.710818307905687e-06,
+      "loss": 0.3789,
+      "step": 4670
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.425758361816406,
+      "learning_rate": 5.705617198335645e-06,
+      "loss": 0.7813,
+      "step": 4671
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.368963241577148,
+      "learning_rate": 5.700416088765604e-06,
+      "loss": 0.5316,
+      "step": 4672
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.56916332244873,
+      "learning_rate": 5.695214979195562e-06,
+      "loss": 0.5712,
+      "step": 4673
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.49671459197998,
+      "learning_rate": 5.69001386962552e-06,
+      "loss": 0.4336,
+      "step": 4674
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.713895797729492,
+      "learning_rate": 5.684812760055479e-06,
+      "loss": 0.3974,
+      "step": 4675
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.559703826904297,
+      "learning_rate": 5.679611650485437e-06,
+      "loss": 0.477,
+      "step": 4676
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.942390441894531,
+      "learning_rate": 5.674410540915395e-06,
+      "loss": 0.6508,
+      "step": 4677
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.353754043579102,
+      "learning_rate": 5.669209431345354e-06,
+      "loss": 0.4331,
+      "step": 4678
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.350062370300293,
+      "learning_rate": 5.6640083217753125e-06,
+      "loss": 0.5965,
+      "step": 4679
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.526127815246582,
+      "learning_rate": 5.65880721220527e-06,
+      "loss": 0.6245,
+      "step": 4680
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.306266784667969,
+      "learning_rate": 5.653606102635229e-06,
+      "loss": 0.4716,
+      "step": 4681
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.8848419189453125,
+      "learning_rate": 5.6484049930651874e-06,
+      "loss": 0.5785,
+      "step": 4682
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.913545608520508,
+      "learning_rate": 5.643203883495146e-06,
+      "loss": 0.4959,
+      "step": 4683
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.745185852050781,
+      "learning_rate": 5.638002773925104e-06,
+      "loss": 0.6695,
+      "step": 4684
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.824722290039062,
+      "learning_rate": 5.6328016643550624e-06,
+      "loss": 0.6988,
+      "step": 4685
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.387177467346191,
+      "learning_rate": 5.627600554785021e-06,
+      "loss": 0.5475,
+      "step": 4686
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.479121208190918,
+      "learning_rate": 5.62239944521498e-06,
+      "loss": 0.5904,
+      "step": 4687
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.485502243041992,
+      "learning_rate": 5.6171983356449374e-06,
+      "loss": 0.4731,
+      "step": 4688
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.742836952209473,
+      "learning_rate": 5.611997226074896e-06,
+      "loss": 0.6139,
+      "step": 4689
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.88058090209961,
+      "learning_rate": 5.606796116504855e-06,
+      "loss": 0.456,
+      "step": 4690
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.745894432067871,
+      "learning_rate": 5.601595006934813e-06,
+      "loss": 0.3685,
+      "step": 4691
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.775364398956299,
+      "learning_rate": 5.596393897364771e-06,
+      "loss": 0.4186,
+      "step": 4692
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.82680082321167,
+      "learning_rate": 5.59119278779473e-06,
+      "loss": 0.5695,
+      "step": 4693
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.991153717041016,
+      "learning_rate": 5.585991678224688e-06,
+      "loss": 0.5678,
+      "step": 4694
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.72563648223877,
+      "learning_rate": 5.580790568654647e-06,
+      "loss": 0.5198,
+      "step": 4695
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.077791690826416,
+      "learning_rate": 5.575589459084605e-06,
+      "loss": 0.4393,
+      "step": 4696
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.596837043762207,
+      "learning_rate": 5.570388349514563e-06,
+      "loss": 0.4896,
+      "step": 4697
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.61953353881836,
+      "learning_rate": 5.565187239944522e-06,
+      "loss": 0.5441,
+      "step": 4698
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.628092765808105,
+      "learning_rate": 5.5599861303744805e-06,
+      "loss": 0.6594,
+      "step": 4699
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.810408115386963,
+      "learning_rate": 5.554785020804438e-06,
+      "loss": 0.4682,
+      "step": 4700
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.556577682495117,
+      "learning_rate": 5.549583911234397e-06,
+      "loss": 0.6133,
+      "step": 4701
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.810951232910156,
+      "learning_rate": 5.5443828016643555e-06,
+      "loss": 0.4475,
+      "step": 4702
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.270831108093262,
+      "learning_rate": 5.539181692094313e-06,
+      "loss": 0.4991,
+      "step": 4703
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.410139083862305,
+      "learning_rate": 5.533980582524272e-06,
+      "loss": 0.5544,
+      "step": 4704
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.175043106079102,
+      "learning_rate": 5.5287794729542304e-06,
+      "loss": 0.4335,
+      "step": 4705
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.784987449645996,
+      "learning_rate": 5.523578363384189e-06,
+      "loss": 0.525,
+      "step": 4706
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.053572654724121,
+      "learning_rate": 5.518377253814147e-06,
+      "loss": 0.6556,
+      "step": 4707
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.732515335083008,
+      "learning_rate": 5.5131761442441054e-06,
+      "loss": 0.4645,
+      "step": 4708
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.38383960723877,
+      "learning_rate": 5.507975034674064e-06,
+      "loss": 0.4778,
+      "step": 4709
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.70075798034668,
+      "learning_rate": 5.502773925104023e-06,
+      "loss": 0.5184,
+      "step": 4710
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.29574203491211,
+      "learning_rate": 5.4975728155339804e-06,
+      "loss": 0.6068,
+      "step": 4711
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.11755657196045,
+      "learning_rate": 5.492371705963939e-06,
+      "loss": 0.524,
+      "step": 4712
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.591024398803711,
+      "learning_rate": 5.487170596393898e-06,
+      "loss": 0.5737,
+      "step": 4713
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.993274688720703,
+      "learning_rate": 5.481969486823856e-06,
+      "loss": 0.3369,
+      "step": 4714
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.691473007202148,
+      "learning_rate": 5.476768377253814e-06,
+      "loss": 0.4802,
+      "step": 4715
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.242079734802246,
+      "learning_rate": 5.471567267683773e-06,
+      "loss": 0.5566,
+      "step": 4716
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.794942855834961,
+      "learning_rate": 5.466366158113731e-06,
+      "loss": 0.794,
+      "step": 4717
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.388157844543457,
+      "learning_rate": 5.46116504854369e-06,
+      "loss": 0.717,
+      "step": 4718
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.066296577453613,
+      "learning_rate": 5.455963938973648e-06,
+      "loss": 0.4212,
+      "step": 4719
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.905529022216797,
+      "learning_rate": 5.450762829403606e-06,
+      "loss": 0.3903,
+      "step": 4720
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.534521102905273,
+      "learning_rate": 5.445561719833565e-06,
+      "loss": 0.7987,
+      "step": 4721
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.064787864685059,
+      "learning_rate": 5.4403606102635235e-06,
+      "loss": 0.6541,
+      "step": 4722
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.73471736907959,
+      "learning_rate": 5.435159500693481e-06,
+      "loss": 0.752,
+      "step": 4723
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.182475090026855,
+      "learning_rate": 5.42995839112344e-06,
+      "loss": 0.3872,
+      "step": 4724
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.404216766357422,
+      "learning_rate": 5.4247572815533985e-06,
+      "loss": 0.5948,
+      "step": 4725
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.488905906677246,
+      "learning_rate": 5.419556171983357e-06,
+      "loss": 0.6544,
+      "step": 4726
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.598878860473633,
+      "learning_rate": 5.414355062413315e-06,
+      "loss": 0.5592,
+      "step": 4727
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.527830123901367,
+      "learning_rate": 5.4091539528432735e-06,
+      "loss": 0.8274,
+      "step": 4728
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.561210632324219,
+      "learning_rate": 5.403952843273232e-06,
+      "loss": 0.5694,
+      "step": 4729
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.453078269958496,
+      "learning_rate": 5.39875173370319e-06,
+      "loss": 0.6627,
+      "step": 4730
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.080071449279785,
+      "learning_rate": 5.3935506241331484e-06,
+      "loss": 0.466,
+      "step": 4731
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.6302690505981445,
+      "learning_rate": 5.388349514563107e-06,
+      "loss": 0.4158,
+      "step": 4732
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.020702362060547,
+      "learning_rate": 5.383148404993066e-06,
+      "loss": 0.458,
+      "step": 4733
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.655231475830078,
+      "learning_rate": 5.3779472954230234e-06,
+      "loss": 0.6914,
+      "step": 4734
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.93510627746582,
+      "learning_rate": 5.372746185852982e-06,
+      "loss": 0.6122,
+      "step": 4735
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.599753379821777,
+      "learning_rate": 5.367545076282941e-06,
+      "loss": 0.5845,
+      "step": 4736
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.233242988586426,
+      "learning_rate": 5.362343966712899e-06,
+      "loss": 0.5682,
+      "step": 4737
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.524258613586426,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.5481,
+      "step": 4738
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.759553909301758,
+      "learning_rate": 5.351941747572816e-06,
+      "loss": 0.584,
+      "step": 4739
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.421677589416504,
+      "learning_rate": 5.346740638002774e-06,
+      "loss": 0.4068,
+      "step": 4740
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.38226318359375,
+      "learning_rate": 5.341539528432733e-06,
+      "loss": 0.694,
+      "step": 4741
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.39568042755127,
+      "learning_rate": 5.336338418862691e-06,
+      "loss": 0.3829,
+      "step": 4742
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.752991199493408,
+      "learning_rate": 5.331137309292649e-06,
+      "loss": 0.4545,
+      "step": 4743
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.38961410522461,
+      "learning_rate": 5.325936199722608e-06,
+      "loss": 0.5813,
+      "step": 4744
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.831655502319336,
+      "learning_rate": 5.3207350901525665e-06,
+      "loss": 0.5737,
+      "step": 4745
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.350545883178711,
+      "learning_rate": 5.315533980582524e-06,
+      "loss": 0.4588,
+      "step": 4746
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.970984935760498,
+      "learning_rate": 5.310332871012483e-06,
+      "loss": 0.5613,
+      "step": 4747
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.555718421936035,
+      "learning_rate": 5.3051317614424415e-06,
+      "loss": 0.5508,
+      "step": 4748
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.356681823730469,
+      "learning_rate": 5.2999306518724e-06,
+      "loss": 0.822,
+      "step": 4749
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.5854811668396,
+      "learning_rate": 5.294729542302358e-06,
+      "loss": 0.3921,
+      "step": 4750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.948698043823242,
+      "learning_rate": 5.2895284327323165e-06,
+      "loss": 0.5705,
+      "step": 4751
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.3611321449279785,
+      "learning_rate": 5.284327323162275e-06,
+      "loss": 0.5203,
+      "step": 4752
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.785578727722168,
+      "learning_rate": 5.279126213592234e-06,
+      "loss": 0.5167,
+      "step": 4753
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.921727657318115,
+      "learning_rate": 5.2739251040221914e-06,
+      "loss": 0.3402,
+      "step": 4754
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.027938842773438,
+      "learning_rate": 5.26872399445215e-06,
+      "loss": 0.564,
+      "step": 4755
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.96353816986084,
+      "learning_rate": 5.263522884882109e-06,
+      "loss": 0.7071,
+      "step": 4756
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.172010898590088,
+      "learning_rate": 5.258321775312067e-06,
+      "loss": 0.4691,
+      "step": 4757
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.45865249633789,
+      "learning_rate": 5.253120665742025e-06,
+      "loss": 0.4615,
+      "step": 4758
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.449492454528809,
+      "learning_rate": 5.247919556171984e-06,
+      "loss": 0.6589,
+      "step": 4759
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.55368709564209,
+      "learning_rate": 5.242718446601942e-06,
+      "loss": 0.4871,
+      "step": 4760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.585121154785156,
+      "learning_rate": 5.2375173370319e-06,
+      "loss": 0.5276,
+      "step": 4761
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.877589225769043,
+      "learning_rate": 5.232316227461859e-06,
+      "loss": 0.6187,
+      "step": 4762
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.961573600769043,
+      "learning_rate": 5.227115117891817e-06,
+      "loss": 0.688,
+      "step": 4763
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.164773941040039,
+      "learning_rate": 5.221914008321776e-06,
+      "loss": 0.5707,
+      "step": 4764
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.715240478515625,
+      "learning_rate": 5.216712898751734e-06,
+      "loss": 0.7904,
+      "step": 4765
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.679605484008789,
+      "learning_rate": 5.211511789181692e-06,
+      "loss": 0.4692,
+      "step": 4766
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.034242630004883,
+      "learning_rate": 5.206310679611651e-06,
+      "loss": 0.7598,
+      "step": 4767
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.673609733581543,
+      "learning_rate": 5.2011095700416095e-06,
+      "loss": 0.5646,
+      "step": 4768
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.50714111328125,
+      "learning_rate": 5.195908460471567e-06,
+      "loss": 0.5162,
+      "step": 4769
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.564245223999023,
+      "learning_rate": 5.190707350901526e-06,
+      "loss": 0.4705,
+      "step": 4770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.492937088012695,
+      "learning_rate": 5.1855062413314845e-06,
+      "loss": 0.6365,
+      "step": 4771
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.047977447509766,
+      "learning_rate": 5.180305131761443e-06,
+      "loss": 0.6903,
+      "step": 4772
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.5095853805542,
+      "learning_rate": 5.175104022191401e-06,
+      "loss": 0.5283,
+      "step": 4773
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.424959182739258,
+      "learning_rate": 5.1699029126213595e-06,
+      "loss": 0.6858,
+      "step": 4774
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.771688938140869,
+      "learning_rate": 5.164701803051318e-06,
+      "loss": 0.4234,
+      "step": 4775
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.200518608093262,
+      "learning_rate": 5.159500693481277e-06,
+      "loss": 0.7099,
+      "step": 4776
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.778925895690918,
+      "learning_rate": 5.1542995839112345e-06,
+      "loss": 0.53,
+      "step": 4777
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.96568775177002,
+      "learning_rate": 5.149098474341193e-06,
+      "loss": 0.5125,
+      "step": 4778
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.910531997680664,
+      "learning_rate": 5.143897364771152e-06,
+      "loss": 0.6156,
+      "step": 4779
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.39760971069336,
+      "learning_rate": 5.13869625520111e-06,
+      "loss": 0.4452,
+      "step": 4780
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.831864356994629,
+      "learning_rate": 5.133495145631068e-06,
+      "loss": 0.6141,
+      "step": 4781
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.359835624694824,
+      "learning_rate": 5.128294036061027e-06,
+      "loss": 0.4677,
+      "step": 4782
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.350821495056152,
+      "learning_rate": 5.123092926490985e-06,
+      "loss": 0.4662,
+      "step": 4783
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.681921005249023,
+      "learning_rate": 5.117891816920944e-06,
+      "loss": 0.56,
+      "step": 4784
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.614752769470215,
+      "learning_rate": 5.112690707350902e-06,
+      "loss": 0.5317,
+      "step": 4785
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.301433563232422,
+      "learning_rate": 5.10748959778086e-06,
+      "loss": 0.4238,
+      "step": 4786
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.63447380065918,
+      "learning_rate": 5.102288488210819e-06,
+      "loss": 0.5206,
+      "step": 4787
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.235017776489258,
+      "learning_rate": 5.097087378640777e-06,
+      "loss": 0.4154,
+      "step": 4788
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.44775676727295,
+      "learning_rate": 5.091886269070735e-06,
+      "loss": 0.5205,
+      "step": 4789
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.77042293548584,
+      "learning_rate": 5.086685159500694e-06,
+      "loss": 0.4504,
+      "step": 4790
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.584941864013672,
+      "learning_rate": 5.0814840499306525e-06,
+      "loss": 0.5897,
+      "step": 4791
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.163352966308594,
+      "learning_rate": 5.07628294036061e-06,
+      "loss": 0.5461,
+      "step": 4792
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.620095252990723,
+      "learning_rate": 5.071081830790569e-06,
+      "loss": 0.6549,
+      "step": 4793
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.035597801208496,
+      "learning_rate": 5.0658807212205275e-06,
+      "loss": 0.3462,
+      "step": 4794
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.600295066833496,
+      "learning_rate": 5.060679611650486e-06,
+      "loss": 0.5997,
+      "step": 4795
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.689885139465332,
+      "learning_rate": 5.055478502080444e-06,
+      "loss": 0.6204,
+      "step": 4796
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.588595390319824,
+      "learning_rate": 5.0502773925104025e-06,
+      "loss": 0.5376,
+      "step": 4797
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.987271785736084,
+      "learning_rate": 5.045076282940361e-06,
+      "loss": 0.4047,
+      "step": 4798
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.567432403564453,
+      "learning_rate": 5.03987517337032e-06,
+      "loss": 0.514,
+      "step": 4799
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.07776165008545,
+      "learning_rate": 5.0346740638002775e-06,
+      "loss": 0.624,
+      "step": 4800
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.842432022094727,
+      "learning_rate": 5.029472954230236e-06,
+      "loss": 0.5988,
+      "step": 4801
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.80256175994873,
+      "learning_rate": 5.024271844660195e-06,
+      "loss": 0.6029,
+      "step": 4802
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.67645263671875,
+      "learning_rate": 5.019070735090153e-06,
+      "loss": 0.5782,
+      "step": 4803
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.317459106445312,
+      "learning_rate": 5.013869625520111e-06,
+      "loss": 0.6667,
+      "step": 4804
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.021561622619629,
+      "learning_rate": 5.00866851595007e-06,
+      "loss": 0.5388,
+      "step": 4805
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.778569221496582,
+      "learning_rate": 5.003467406380028e-06,
+      "loss": 0.4733,
+      "step": 4806
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.417832374572754,
+      "learning_rate": 4.998266296809987e-06,
+      "loss": 0.4929,
+      "step": 4807
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.7926344871521,
+      "learning_rate": 4.993065187239945e-06,
+      "loss": 0.417,
+      "step": 4808
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.790334701538086,
+      "learning_rate": 4.987864077669903e-06,
+      "loss": 0.4676,
+      "step": 4809
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.09611988067627,
+      "learning_rate": 4.982662968099862e-06,
+      "loss": 0.5665,
+      "step": 4810
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.157032012939453,
+      "learning_rate": 4.9774618585298205e-06,
+      "loss": 0.3441,
+      "step": 4811
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.80765438079834,
+      "learning_rate": 4.972260748959778e-06,
+      "loss": 0.6365,
+      "step": 4812
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.459265232086182,
+      "learning_rate": 4.967059639389737e-06,
+      "loss": 0.3993,
+      "step": 4813
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.869949340820312,
+      "learning_rate": 4.9618585298196955e-06,
+      "loss": 0.6706,
+      "step": 4814
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.895922660827637,
+      "learning_rate": 4.956657420249654e-06,
+      "loss": 0.5438,
+      "step": 4815
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.757575035095215,
+      "learning_rate": 4.951456310679612e-06,
+      "loss": 0.6796,
+      "step": 4816
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.295160293579102,
+      "learning_rate": 4.9462552011095705e-06,
+      "loss": 0.6322,
+      "step": 4817
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.531583786010742,
+      "learning_rate": 4.941054091539529e-06,
+      "loss": 0.6684,
+      "step": 4818
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.631214141845703,
+      "learning_rate": 4.935852981969487e-06,
+      "loss": 0.6345,
+      "step": 4819
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.713045120239258,
+      "learning_rate": 4.9306518723994455e-06,
+      "loss": 0.3686,
+      "step": 4820
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.418646812438965,
+      "learning_rate": 4.925450762829404e-06,
+      "loss": 0.5396,
+      "step": 4821
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.766876697540283,
+      "learning_rate": 4.920249653259363e-06,
+      "loss": 0.4339,
+      "step": 4822
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.560930252075195,
+      "learning_rate": 4.9150485436893205e-06,
+      "loss": 0.804,
+      "step": 4823
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.593488693237305,
+      "learning_rate": 4.909847434119279e-06,
+      "loss": 0.4579,
+      "step": 4824
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.906489372253418,
+      "learning_rate": 4.904646324549238e-06,
+      "loss": 0.6204,
+      "step": 4825
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.75501823425293,
+      "learning_rate": 4.899445214979196e-06,
+      "loss": 0.4091,
+      "step": 4826
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.171737670898438,
+      "learning_rate": 4.894244105409154e-06,
+      "loss": 0.711,
+      "step": 4827
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.862494468688965,
+      "learning_rate": 4.889042995839113e-06,
+      "loss": 0.6959,
+      "step": 4828
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.759832382202148,
+      "learning_rate": 4.883841886269071e-06,
+      "loss": 0.819,
+      "step": 4829
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.064682006835938,
+      "learning_rate": 4.87864077669903e-06,
+      "loss": 0.6641,
+      "step": 4830
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.994148254394531,
+      "learning_rate": 4.873439667128988e-06,
+      "loss": 0.502,
+      "step": 4831
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.093571662902832,
+      "learning_rate": 4.868238557558946e-06,
+      "loss": 0.6276,
+      "step": 4832
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.681207656860352,
+      "learning_rate": 4.863037447988905e-06,
+      "loss": 0.6102,
+      "step": 4833
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.569230079650879,
+      "learning_rate": 4.8578363384188635e-06,
+      "loss": 0.5687,
+      "step": 4834
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.980050086975098,
+      "learning_rate": 4.852635228848821e-06,
+      "loss": 0.6016,
+      "step": 4835
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.384329795837402,
+      "learning_rate": 4.84743411927878e-06,
+      "loss": 0.7522,
+      "step": 4836
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.346293449401855,
+      "learning_rate": 4.8422330097087385e-06,
+      "loss": 0.4812,
+      "step": 4837
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.32577133178711,
+      "learning_rate": 4.837031900138697e-06,
+      "loss": 0.4982,
+      "step": 4838
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.932581901550293,
+      "learning_rate": 4.831830790568655e-06,
+      "loss": 0.4746,
+      "step": 4839
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.699601173400879,
+      "learning_rate": 4.8266296809986135e-06,
+      "loss": 0.5429,
+      "step": 4840
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.058354377746582,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 0.5573,
+      "step": 4841
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.789058685302734,
+      "learning_rate": 4.816227461858531e-06,
+      "loss": 0.6983,
+      "step": 4842
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.474084854125977,
+      "learning_rate": 4.8110263522884885e-06,
+      "loss": 0.7227,
+      "step": 4843
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.949374198913574,
+      "learning_rate": 4.805825242718447e-06,
+      "loss": 0.7397,
+      "step": 4844
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.027949333190918,
+      "learning_rate": 4.800624133148406e-06,
+      "loss": 0.578,
+      "step": 4845
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.910952091217041,
+      "learning_rate": 4.7954230235783635e-06,
+      "loss": 0.4357,
+      "step": 4846
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.168667793273926,
+      "learning_rate": 4.790221914008322e-06,
+      "loss": 0.5987,
+      "step": 4847
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.440926551818848,
+      "learning_rate": 4.785020804438281e-06,
+      "loss": 0.4746,
+      "step": 4848
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.469916343688965,
+      "learning_rate": 4.779819694868239e-06,
+      "loss": 0.468,
+      "step": 4849
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.806220054626465,
+      "learning_rate": 4.774618585298197e-06,
+      "loss": 0.6178,
+      "step": 4850
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.442523002624512,
+      "learning_rate": 4.769417475728156e-06,
+      "loss": 0.4434,
+      "step": 4851
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.141456604003906,
+      "learning_rate": 4.764216366158114e-06,
+      "loss": 0.5513,
+      "step": 4852
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.495190620422363,
+      "learning_rate": 4.759015256588073e-06,
+      "loss": 0.5182,
+      "step": 4853
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.7544527053833,
+      "learning_rate": 4.753814147018031e-06,
+      "loss": 0.6264,
+      "step": 4854
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.820914268493652,
+      "learning_rate": 4.748613037447989e-06,
+      "loss": 0.593,
+      "step": 4855
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.442625045776367,
+      "learning_rate": 4.743411927877948e-06,
+      "loss": 0.9139,
+      "step": 4856
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.91428279876709,
+      "learning_rate": 4.7382108183079065e-06,
+      "loss": 0.466,
+      "step": 4857
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.290508270263672,
+      "learning_rate": 4.733009708737864e-06,
+      "loss": 0.6959,
+      "step": 4858
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.165891647338867,
+      "learning_rate": 4.727808599167823e-06,
+      "loss": 0.6599,
+      "step": 4859
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.855887413024902,
+      "learning_rate": 4.7226074895977815e-06,
+      "loss": 0.4878,
+      "step": 4860
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.630184650421143,
+      "learning_rate": 4.71740638002774e-06,
+      "loss": 0.5363,
+      "step": 4861
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.6704683303833,
+      "learning_rate": 4.712205270457698e-06,
+      "loss": 0.6082,
+      "step": 4862
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.793848991394043,
+      "learning_rate": 4.7070041608876565e-06,
+      "loss": 0.5936,
+      "step": 4863
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.205232620239258,
+      "learning_rate": 4.701803051317615e-06,
+      "loss": 0.4907,
+      "step": 4864
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.486756324768066,
+      "learning_rate": 4.696601941747574e-06,
+      "loss": 0.3557,
+      "step": 4865
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.051569938659668,
+      "learning_rate": 4.6914008321775315e-06,
+      "loss": 0.4587,
+      "step": 4866
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.488103866577148,
+      "learning_rate": 4.686199722607489e-06,
+      "loss": 0.3502,
+      "step": 4867
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.69162654876709,
+      "learning_rate": 4.680998613037448e-06,
+      "loss": 0.5901,
+      "step": 4868
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.887885093688965,
+      "learning_rate": 4.6757975034674065e-06,
+      "loss": 0.5501,
+      "step": 4869
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.959454536437988,
+      "learning_rate": 4.670596393897364e-06,
+      "loss": 0.5023,
+      "step": 4870
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.80720853805542,
+      "learning_rate": 4.665395284327323e-06,
+      "loss": 0.4876,
+      "step": 4871
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.922835350036621,
+      "learning_rate": 4.6601941747572815e-06,
+      "loss": 0.5294,
+      "step": 4872
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.414713859558105,
+      "learning_rate": 4.65499306518724e-06,
+      "loss": 0.6196,
+      "step": 4873
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.759987831115723,
+      "learning_rate": 4.649791955617198e-06,
+      "loss": 0.5731,
+      "step": 4874
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.331056594848633,
+      "learning_rate": 4.6445908460471564e-06,
+      "loss": 0.538,
+      "step": 4875
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.234905242919922,
+      "learning_rate": 4.639389736477115e-06,
+      "loss": 0.4345,
+      "step": 4876
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.657195091247559,
+      "learning_rate": 4.634188626907073e-06,
+      "loss": 0.483,
+      "step": 4877
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.105978965759277,
+      "learning_rate": 4.6289875173370314e-06,
+      "loss": 0.698,
+      "step": 4878
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.6546173095703125,
+      "learning_rate": 4.62378640776699e-06,
+      "loss": 0.6106,
+      "step": 4879
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.658801078796387,
+      "learning_rate": 4.618585298196949e-06,
+      "loss": 0.6156,
+      "step": 4880
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.150995254516602,
+      "learning_rate": 4.6133841886269064e-06,
+      "loss": 0.5231,
+      "step": 4881
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.991193771362305,
+      "learning_rate": 4.608183079056865e-06,
+      "loss": 0.5021,
+      "step": 4882
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.943384170532227,
+      "learning_rate": 4.602981969486824e-06,
+      "loss": 0.6205,
+      "step": 4883
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.994600296020508,
+      "learning_rate": 4.597780859916782e-06,
+      "loss": 0.5638,
+      "step": 4884
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.176400184631348,
+      "learning_rate": 4.59257975034674e-06,
+      "loss": 0.5146,
+      "step": 4885
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.476892471313477,
+      "learning_rate": 4.587378640776699e-06,
+      "loss": 0.7961,
+      "step": 4886
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.044927597045898,
+      "learning_rate": 4.582177531206657e-06,
+      "loss": 0.5924,
+      "step": 4887
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.888240814208984,
+      "learning_rate": 4.576976421636616e-06,
+      "loss": 0.56,
+      "step": 4888
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.453103542327881,
+      "learning_rate": 4.571775312066574e-06,
+      "loss": 0.4029,
+      "step": 4889
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.545904159545898,
+      "learning_rate": 4.566574202496532e-06,
+      "loss": 0.5878,
+      "step": 4890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.470197677612305,
+      "learning_rate": 4.561373092926491e-06,
+      "loss": 0.491,
+      "step": 4891
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.497235298156738,
+      "learning_rate": 4.5561719833564495e-06,
+      "loss": 0.6299,
+      "step": 4892
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.628016471862793,
+      "learning_rate": 4.550970873786407e-06,
+      "loss": 0.4105,
+      "step": 4893
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.585450649261475,
+      "learning_rate": 4.545769764216366e-06,
+      "loss": 0.4929,
+      "step": 4894
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.23181438446045,
+      "learning_rate": 4.5405686546463245e-06,
+      "loss": 0.5486,
+      "step": 4895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.337618827819824,
+      "learning_rate": 4.535367545076283e-06,
+      "loss": 0.9825,
+      "step": 4896
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.610481262207031,
+      "learning_rate": 4.530166435506241e-06,
+      "loss": 0.5139,
+      "step": 4897
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.329362869262695,
+      "learning_rate": 4.5249653259361995e-06,
+      "loss": 0.3887,
+      "step": 4898
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.154709815979004,
+      "learning_rate": 4.519764216366158e-06,
+      "loss": 0.564,
+      "step": 4899
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.001935005187988,
+      "learning_rate": 4.514563106796117e-06,
+      "loss": 0.4252,
+      "step": 4900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.234025955200195,
+      "learning_rate": 4.5093619972260744e-06,
+      "loss": 0.4971,
+      "step": 4901
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.74514102935791,
+      "learning_rate": 4.504160887656033e-06,
+      "loss": 0.53,
+      "step": 4902
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.945734024047852,
+      "learning_rate": 4.498959778085992e-06,
+      "loss": 0.607,
+      "step": 4903
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.460039138793945,
+      "learning_rate": 4.4937586685159494e-06,
+      "loss": 0.5312,
+      "step": 4904
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.6507792472839355,
+      "learning_rate": 4.488557558945908e-06,
+      "loss": 0.5347,
+      "step": 4905
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.6899895668029785,
+      "learning_rate": 4.483356449375867e-06,
+      "loss": 0.4826,
+      "step": 4906
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.151714324951172,
+      "learning_rate": 4.478155339805825e-06,
+      "loss": 0.5053,
+      "step": 4907
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.950303077697754,
+      "learning_rate": 4.472954230235783e-06,
+      "loss": 0.675,
+      "step": 4908
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.29718017578125,
+      "learning_rate": 4.467753120665742e-06,
+      "loss": 0.8247,
+      "step": 4909
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.444043159484863,
+      "learning_rate": 4.4625520110957e-06,
+      "loss": 0.5598,
+      "step": 4910
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.246794700622559,
+      "learning_rate": 4.457350901525659e-06,
+      "loss": 0.5446,
+      "step": 4911
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.1778693199157715,
+      "learning_rate": 4.452149791955617e-06,
+      "loss": 0.3365,
+      "step": 4912
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.600996017456055,
+      "learning_rate": 4.446948682385575e-06,
+      "loss": 0.5809,
+      "step": 4913
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.867671966552734,
+      "learning_rate": 4.441747572815534e-06,
+      "loss": 0.6497,
+      "step": 4914
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.207243919372559,
+      "learning_rate": 4.4365464632454925e-06,
+      "loss": 0.483,
+      "step": 4915
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.614618301391602,
+      "learning_rate": 4.43134535367545e-06,
+      "loss": 0.2934,
+      "step": 4916
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.645122528076172,
+      "learning_rate": 4.426144244105409e-06,
+      "loss": 0.4267,
+      "step": 4917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.81100845336914,
+      "learning_rate": 4.4209431345353675e-06,
+      "loss": 0.5251,
+      "step": 4918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.40697193145752,
+      "learning_rate": 4.415742024965326e-06,
+      "loss": 0.577,
+      "step": 4919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.953437328338623,
+      "learning_rate": 4.410540915395284e-06,
+      "loss": 0.5502,
+      "step": 4920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.569183349609375,
+      "learning_rate": 4.4053398058252425e-06,
+      "loss": 0.6045,
+      "step": 4921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.912191390991211,
+      "learning_rate": 4.400138696255201e-06,
+      "loss": 0.4692,
+      "step": 4922
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.505646705627441,
+      "learning_rate": 4.39493758668516e-06,
+      "loss": 0.5616,
+      "step": 4923
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.276614665985107,
+      "learning_rate": 4.3897364771151174e-06,
+      "loss": 0.3107,
+      "step": 4924
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.339812278747559,
+      "learning_rate": 4.384535367545076e-06,
+      "loss": 0.3137,
+      "step": 4925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.777345657348633,
+      "learning_rate": 4.379334257975035e-06,
+      "loss": 0.4653,
+      "step": 4926
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.484979629516602,
+      "learning_rate": 4.374133148404993e-06,
+      "loss": 0.5293,
+      "step": 4927
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.387085914611816,
+      "learning_rate": 4.368932038834951e-06,
+      "loss": 0.6354,
+      "step": 4928
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.15458869934082,
+      "learning_rate": 4.36373092926491e-06,
+      "loss": 0.6595,
+      "step": 4929
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.315655708312988,
+      "learning_rate": 4.358529819694868e-06,
+      "loss": 0.6319,
+      "step": 4930
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.041146278381348,
+      "learning_rate": 4.353328710124827e-06,
+      "loss": 0.483,
+      "step": 4931
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.095645904541016,
+      "learning_rate": 4.348127600554785e-06,
+      "loss": 0.5381,
+      "step": 4932
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.553728103637695,
+      "learning_rate": 4.342926490984743e-06,
+      "loss": 0.4332,
+      "step": 4933
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.06398868560791,
+      "learning_rate": 4.337725381414702e-06,
+      "loss": 0.7249,
+      "step": 4934
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.731823921203613,
+      "learning_rate": 4.33252427184466e-06,
+      "loss": 0.5332,
+      "step": 4935
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.3746337890625,
+      "learning_rate": 4.327323162274618e-06,
+      "loss": 0.5245,
+      "step": 4936
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.122068405151367,
+      "learning_rate": 4.322122052704577e-06,
+      "loss": 0.4409,
+      "step": 4937
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.97861385345459,
+      "learning_rate": 4.3169209431345355e-06,
+      "loss": 0.5089,
+      "step": 4938
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.983196258544922,
+      "learning_rate": 4.311719833564493e-06,
+      "loss": 0.5671,
+      "step": 4939
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.612195014953613,
+      "learning_rate": 4.306518723994452e-06,
+      "loss": 0.5294,
+      "step": 4940
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.99580192565918,
+      "learning_rate": 4.3013176144244105e-06,
+      "loss": 0.5874,
+      "step": 4941
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.306250095367432,
+      "learning_rate": 4.296116504854369e-06,
+      "loss": 0.391,
+      "step": 4942
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.671103477478027,
+      "learning_rate": 4.290915395284327e-06,
+      "loss": 0.5304,
+      "step": 4943
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.937923431396484,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.5709,
+      "step": 4944
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.333708763122559,
+      "learning_rate": 4.280513176144244e-06,
+      "loss": 0.6687,
+      "step": 4945
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.15112018585205,
+      "learning_rate": 4.275312066574203e-06,
+      "loss": 0.5866,
+      "step": 4946
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.5026350021362305,
+      "learning_rate": 4.2701109570041605e-06,
+      "loss": 0.4207,
+      "step": 4947
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.838363647460938,
+      "learning_rate": 4.264909847434119e-06,
+      "loss": 0.619,
+      "step": 4948
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.845272064208984,
+      "learning_rate": 4.259708737864078e-06,
+      "loss": 0.3976,
+      "step": 4949
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.681865692138672,
+      "learning_rate": 4.254507628294036e-06,
+      "loss": 0.4749,
+      "step": 4950
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.957738876342773,
+      "learning_rate": 4.249306518723994e-06,
+      "loss": 0.5544,
+      "step": 4951
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.404147148132324,
+      "learning_rate": 4.244105409153953e-06,
+      "loss": 0.6389,
+      "step": 4952
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.386004447937012,
+      "learning_rate": 4.238904299583911e-06,
+      "loss": 0.4342,
+      "step": 4953
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.337724208831787,
+      "learning_rate": 4.23370319001387e-06,
+      "loss": 0.5146,
+      "step": 4954
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.598703384399414,
+      "learning_rate": 4.228502080443828e-06,
+      "loss": 0.5557,
+      "step": 4955
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.985430717468262,
+      "learning_rate": 4.223300970873786e-06,
+      "loss": 0.5541,
+      "step": 4956
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.595061302185059,
+      "learning_rate": 4.218099861303745e-06,
+      "loss": 0.6081,
+      "step": 4957
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.455885887145996,
+      "learning_rate": 4.2128987517337035e-06,
+      "loss": 0.5228,
+      "step": 4958
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.996994972229004,
+      "learning_rate": 4.207697642163661e-06,
+      "loss": 0.3967,
+      "step": 4959
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.238723754882812,
+      "learning_rate": 4.20249653259362e-06,
+      "loss": 0.4402,
+      "step": 4960
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.118901252746582,
+      "learning_rate": 4.1972954230235785e-06,
+      "loss": 0.7847,
+      "step": 4961
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.035292625427246,
+      "learning_rate": 4.192094313453536e-06,
+      "loss": 0.5637,
+      "step": 4962
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.109061241149902,
+      "learning_rate": 4.186893203883495e-06,
+      "loss": 0.6213,
+      "step": 4963
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.62488079071045,
+      "learning_rate": 4.1816920943134535e-06,
+      "loss": 0.506,
+      "step": 4964
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.33948040008545,
+      "learning_rate": 4.176490984743412e-06,
+      "loss": 0.4025,
+      "step": 4965
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.407170295715332,
+      "learning_rate": 4.17128987517337e-06,
+      "loss": 0.4417,
+      "step": 4966
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.37271785736084,
+      "learning_rate": 4.1660887656033285e-06,
+      "loss": 0.5204,
+      "step": 4967
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.935742378234863,
+      "learning_rate": 4.160887656033287e-06,
+      "loss": 0.5415,
+      "step": 4968
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.162992477416992,
+      "learning_rate": 4.155686546463246e-06,
+      "loss": 0.5609,
+      "step": 4969
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.570207595825195,
+      "learning_rate": 4.1504854368932035e-06,
+      "loss": 0.4482,
+      "step": 4970
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.60733699798584,
+      "learning_rate": 4.145284327323162e-06,
+      "loss": 0.5903,
+      "step": 4971
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.422455787658691,
+      "learning_rate": 4.140083217753121e-06,
+      "loss": 0.6722,
+      "step": 4972
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.750211715698242,
+      "learning_rate": 4.134882108183079e-06,
+      "loss": 0.5944,
+      "step": 4973
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.891785621643066,
+      "learning_rate": 4.129680998613037e-06,
+      "loss": 0.4587,
+      "step": 4974
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.68132209777832,
+      "learning_rate": 4.124479889042996e-06,
+      "loss": 0.8306,
+      "step": 4975
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.010201930999756,
+      "learning_rate": 4.119278779472954e-06,
+      "loss": 0.3805,
+      "step": 4976
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.67814302444458,
+      "learning_rate": 4.114077669902913e-06,
+      "loss": 0.3359,
+      "step": 4977
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.848470687866211,
+      "learning_rate": 4.108876560332871e-06,
+      "loss": 0.9756,
+      "step": 4978
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.61059856414795,
+      "learning_rate": 4.103675450762829e-06,
+      "loss": 0.608,
+      "step": 4979
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.426758766174316,
+      "learning_rate": 4.098474341192788e-06,
+      "loss": 0.5373,
+      "step": 4980
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.708855152130127,
+      "learning_rate": 4.0932732316227465e-06,
+      "loss": 0.4631,
+      "step": 4981
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.653117656707764,
+      "learning_rate": 4.088072122052704e-06,
+      "loss": 0.4134,
+      "step": 4982
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.89979887008667,
+      "learning_rate": 4.082871012482663e-06,
+      "loss": 0.4106,
+      "step": 4983
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.368355751037598,
+      "learning_rate": 4.0776699029126215e-06,
+      "loss": 0.4971,
+      "step": 4984
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.531725883483887,
+      "learning_rate": 4.07246879334258e-06,
+      "loss": 0.5547,
+      "step": 4985
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.074995994567871,
+      "learning_rate": 4.067267683772538e-06,
+      "loss": 0.4801,
+      "step": 4986
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.681770324707031,
+      "learning_rate": 4.0620665742024965e-06,
+      "loss": 0.4227,
+      "step": 4987
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.94908332824707,
+      "learning_rate": 4.056865464632455e-06,
+      "loss": 0.5363,
+      "step": 4988
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.671468734741211,
+      "learning_rate": 4.051664355062414e-06,
+      "loss": 0.5376,
+      "step": 4989
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.64040756225586,
+      "learning_rate": 4.0464632454923715e-06,
+      "loss": 0.5689,
+      "step": 4990
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.67824935913086,
+      "learning_rate": 4.04126213592233e-06,
+      "loss": 0.6223,
+      "step": 4991
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.98774528503418,
+      "learning_rate": 4.036061026352289e-06,
+      "loss": 0.6176,
+      "step": 4992
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.306077003479004,
+      "learning_rate": 4.0308599167822465e-06,
+      "loss": 0.7514,
+      "step": 4993
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.78920841217041,
+      "learning_rate": 4.025658807212205e-06,
+      "loss": 0.6121,
+      "step": 4994
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.273934364318848,
+      "learning_rate": 4.020457697642164e-06,
+      "loss": 0.4614,
+      "step": 4995
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.074139595031738,
+      "learning_rate": 4.015256588072122e-06,
+      "loss": 0.6183,
+      "step": 4996
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.734215259552002,
+      "learning_rate": 4.01005547850208e-06,
+      "loss": 0.555,
+      "step": 4997
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.903766632080078,
+      "learning_rate": 4.004854368932039e-06,
+      "loss": 0.6132,
+      "step": 4998
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.833539962768555,
+      "learning_rate": 3.999653259361997e-06,
+      "loss": 0.4395,
+      "step": 4999
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.995643615722656,
+      "learning_rate": 3.994452149791956e-06,
+      "loss": 0.5092,
+      "step": 5000
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.636122703552246,
+      "learning_rate": 3.989251040221914e-06,
+      "loss": 0.4484,
+      "step": 5001
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.80492877960205,
+      "learning_rate": 3.984049930651872e-06,
+      "loss": 0.4585,
+      "step": 5002
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.409432888031006,
+      "learning_rate": 3.978848821081831e-06,
+      "loss": 0.4653,
+      "step": 5003
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.014898300170898,
+      "learning_rate": 3.9736477115117895e-06,
+      "loss": 0.6102,
+      "step": 5004
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.950699806213379,
+      "learning_rate": 3.968446601941747e-06,
+      "loss": 0.4696,
+      "step": 5005
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.822463035583496,
+      "learning_rate": 3.963245492371706e-06,
+      "loss": 0.5656,
+      "step": 5006
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.995532035827637,
+      "learning_rate": 3.9580443828016645e-06,
+      "loss": 0.319,
+      "step": 5007
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.051712989807129,
+      "learning_rate": 3.952843273231623e-06,
+      "loss": 0.5818,
+      "step": 5008
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.535999298095703,
+      "learning_rate": 3.947642163661581e-06,
+      "loss": 0.7264,
+      "step": 5009
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.22530746459961,
+      "learning_rate": 3.9424410540915395e-06,
+      "loss": 0.4799,
+      "step": 5010
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.734002113342285,
+      "learning_rate": 3.937239944521498e-06,
+      "loss": 0.5545,
+      "step": 5011
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.25273323059082,
+      "learning_rate": 3.932038834951457e-06,
+      "loss": 0.5738,
+      "step": 5012
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.987253189086914,
+      "learning_rate": 3.9268377253814145e-06,
+      "loss": 0.5038,
+      "step": 5013
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.004745960235596,
+      "learning_rate": 3.921636615811373e-06,
+      "loss": 0.399,
+      "step": 5014
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.102212905883789,
+      "learning_rate": 3.916435506241332e-06,
+      "loss": 0.6887,
+      "step": 5015
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.797811508178711,
+      "learning_rate": 3.91123439667129e-06,
+      "loss": 0.6242,
+      "step": 5016
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.003517150878906,
+      "learning_rate": 3.906033287101248e-06,
+      "loss": 0.4838,
+      "step": 5017
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.417401313781738,
+      "learning_rate": 3.900832177531207e-06,
+      "loss": 0.4567,
+      "step": 5018
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.597020149230957,
+      "learning_rate": 3.895631067961165e-06,
+      "loss": 0.411,
+      "step": 5019
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.998377799987793,
+      "learning_rate": 3.890429958391123e-06,
+      "loss": 0.4479,
+      "step": 5020
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.701866149902344,
+      "learning_rate": 3.885228848821082e-06,
+      "loss": 0.4698,
+      "step": 5021
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.574202537536621,
+      "learning_rate": 3.88002773925104e-06,
+      "loss": 0.5533,
+      "step": 5022
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.91343879699707,
+      "learning_rate": 3.874826629680999e-06,
+      "loss": 0.6443,
+      "step": 5023
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.883694648742676,
+      "learning_rate": 3.869625520110957e-06,
+      "loss": 0.5193,
+      "step": 5024
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.71105670928955,
+      "learning_rate": 3.864424410540915e-06,
+      "loss": 0.6125,
+      "step": 5025
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.267354965209961,
+      "learning_rate": 3.859223300970874e-06,
+      "loss": 0.6872,
+      "step": 5026
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.462644577026367,
+      "learning_rate": 3.8540221914008325e-06,
+      "loss": 0.4625,
+      "step": 5027
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.893092155456543,
+      "learning_rate": 3.84882108183079e-06,
+      "loss": 0.4243,
+      "step": 5028
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.089622497558594,
+      "learning_rate": 3.843619972260749e-06,
+      "loss": 0.4282,
+      "step": 5029
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.173006057739258,
+      "learning_rate": 3.8384188626907075e-06,
+      "loss": 0.6457,
+      "step": 5030
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.720417976379395,
+      "learning_rate": 3.833217753120666e-06,
+      "loss": 0.6523,
+      "step": 5031
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.247106552124023,
+      "learning_rate": 3.828016643550624e-06,
+      "loss": 0.5517,
+      "step": 5032
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.573639869689941,
+      "learning_rate": 3.8228155339805825e-06,
+      "loss": 0.7993,
+      "step": 5033
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.140912055969238,
+      "learning_rate": 3.817614424410541e-06,
+      "loss": 0.6763,
+      "step": 5034
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.09041690826416,
+      "learning_rate": 3.8124133148404993e-06,
+      "loss": 0.4925,
+      "step": 5035
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.930782318115234,
+      "learning_rate": 3.807212205270458e-06,
+      "loss": 0.4455,
+      "step": 5036
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.404703140258789,
+      "learning_rate": 3.802011095700416e-06,
+      "loss": 0.4453,
+      "step": 5037
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.533443450927734,
+      "learning_rate": 3.7968099861303747e-06,
+      "loss": 0.5097,
+      "step": 5038
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.435882568359375,
+      "learning_rate": 3.791608876560333e-06,
+      "loss": 0.5653,
+      "step": 5039
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.221325874328613,
+      "learning_rate": 3.7864077669902915e-06,
+      "loss": 0.6005,
+      "step": 5040
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.41263198852539,
+      "learning_rate": 3.7812066574202497e-06,
+      "loss": 0.5517,
+      "step": 5041
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.290833473205566,
+      "learning_rate": 3.7760055478502083e-06,
+      "loss": 0.5427,
+      "step": 5042
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.631448745727539,
+      "learning_rate": 3.7708044382801665e-06,
+      "loss": 0.5324,
+      "step": 5043
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.314607620239258,
+      "learning_rate": 3.765603328710125e-06,
+      "loss": 0.6077,
+      "step": 5044
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.718101501464844,
+      "learning_rate": 3.7604022191400833e-06,
+      "loss": 0.6821,
+      "step": 5045
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.903497695922852,
+      "learning_rate": 3.755201109570042e-06,
+      "loss": 0.5896,
+      "step": 5046
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.467845916748047,
+      "learning_rate": 3.75e-06,
+      "loss": 0.6139,
+      "step": 5047
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.8733649253845215,
+      "learning_rate": 3.7447988904299583e-06,
+      "loss": 0.4466,
+      "step": 5048
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.175176620483398,
+      "learning_rate": 3.739597780859917e-06,
+      "loss": 0.5409,
+      "step": 5049
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.58254337310791,
+      "learning_rate": 3.734396671289875e-06,
+      "loss": 0.6315,
+      "step": 5050
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.74666690826416,
+      "learning_rate": 3.7291955617198337e-06,
+      "loss": 0.4486,
+      "step": 5051
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.981082916259766,
+      "learning_rate": 3.723994452149792e-06,
+      "loss": 0.4416,
+      "step": 5052
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.815807342529297,
+      "learning_rate": 3.7187933425797505e-06,
+      "loss": 0.4436,
+      "step": 5053
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.083806037902832,
+      "learning_rate": 3.7135922330097087e-06,
+      "loss": 0.4373,
+      "step": 5054
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.732730865478516,
+      "learning_rate": 3.7083911234396673e-06,
+      "loss": 0.5304,
+      "step": 5055
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.483092308044434,
+      "learning_rate": 3.7031900138696255e-06,
+      "loss": 0.589,
+      "step": 5056
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.427230834960938,
+      "learning_rate": 3.697988904299584e-06,
+      "loss": 0.5498,
+      "step": 5057
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.916574478149414,
+      "learning_rate": 3.6927877947295423e-06,
+      "loss": 0.6325,
+      "step": 5058
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.491131782531738,
+      "learning_rate": 3.687586685159501e-06,
+      "loss": 0.5697,
+      "step": 5059
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.239340782165527,
+      "learning_rate": 3.682385575589459e-06,
+      "loss": 0.3476,
+      "step": 5060
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.922371864318848,
+      "learning_rate": 3.6771844660194177e-06,
+      "loss": 0.4546,
+      "step": 5061
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.605368614196777,
+      "learning_rate": 3.671983356449376e-06,
+      "loss": 0.5838,
+      "step": 5062
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.837553024291992,
+      "learning_rate": 3.6667822468793345e-06,
+      "loss": 0.5717,
+      "step": 5063
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.404801368713379,
+      "learning_rate": 3.6615811373092927e-06,
+      "loss": 0.5233,
+      "step": 5064
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.065502166748047,
+      "learning_rate": 3.6563800277392513e-06,
+      "loss": 0.3489,
+      "step": 5065
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.114429473876953,
+      "learning_rate": 3.6511789181692095e-06,
+      "loss": 0.3649,
+      "step": 5066
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.182964324951172,
+      "learning_rate": 3.645977808599168e-06,
+      "loss": 0.5161,
+      "step": 5067
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.894523620605469,
+      "learning_rate": 3.6407766990291263e-06,
+      "loss": 0.5019,
+      "step": 5068
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.260491371154785,
+      "learning_rate": 3.635575589459085e-06,
+      "loss": 0.7404,
+      "step": 5069
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.5048828125,
+      "learning_rate": 3.630374479889043e-06,
+      "loss": 0.4572,
+      "step": 5070
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.326455116271973,
+      "learning_rate": 3.6251733703190017e-06,
+      "loss": 0.4192,
+      "step": 5071
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.430721282958984,
+      "learning_rate": 3.61997226074896e-06,
+      "loss": 0.5188,
+      "step": 5072
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.776074409484863,
+      "learning_rate": 3.6147711511789185e-06,
+      "loss": 0.4118,
+      "step": 5073
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.228639602661133,
+      "learning_rate": 3.6095700416088767e-06,
+      "loss": 0.4217,
+      "step": 5074
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.388314723968506,
+      "learning_rate": 3.6043689320388353e-06,
+      "loss": 0.3597,
+      "step": 5075
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.473511695861816,
+      "learning_rate": 3.5991678224687935e-06,
+      "loss": 0.3975,
+      "step": 5076
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.317442893981934,
+      "learning_rate": 3.5939667128987517e-06,
+      "loss": 0.5186,
+      "step": 5077
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.480192184448242,
+      "learning_rate": 3.5887656033287103e-06,
+      "loss": 0.6578,
+      "step": 5078
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.556422233581543,
+      "learning_rate": 3.5835644937586685e-06,
+      "loss": 0.5234,
+      "step": 5079
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.203537464141846,
+      "learning_rate": 3.578363384188627e-06,
+      "loss": 0.483,
+      "step": 5080
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.390336036682129,
+      "learning_rate": 3.5731622746185853e-06,
+      "loss": 0.4247,
+      "step": 5081
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.587024688720703,
+      "learning_rate": 3.567961165048544e-06,
+      "loss": 0.6673,
+      "step": 5082
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.273709297180176,
+      "learning_rate": 3.562760055478502e-06,
+      "loss": 0.4948,
+      "step": 5083
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.567320823669434,
+      "learning_rate": 3.5575589459084607e-06,
+      "loss": 0.5793,
+      "step": 5084
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.604662895202637,
+      "learning_rate": 3.552357836338419e-06,
+      "loss": 0.5708,
+      "step": 5085
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.450252532958984,
+      "learning_rate": 3.5471567267683775e-06,
+      "loss": 0.4079,
+      "step": 5086
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.701937675476074,
+      "learning_rate": 3.5419556171983357e-06,
+      "loss": 0.413,
+      "step": 5087
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.461211204528809,
+      "learning_rate": 3.5367545076282943e-06,
+      "loss": 0.4494,
+      "step": 5088
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.589776992797852,
+      "learning_rate": 3.5315533980582525e-06,
+      "loss": 0.8119,
+      "step": 5089
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.399984359741211,
+      "learning_rate": 3.526352288488211e-06,
+      "loss": 0.3875,
+      "step": 5090
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.69408893585205,
+      "learning_rate": 3.5211511789181693e-06,
+      "loss": 0.6168,
+      "step": 5091
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.891462326049805,
+      "learning_rate": 3.515950069348128e-06,
+      "loss": 0.5236,
+      "step": 5092
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.460165977478027,
+      "learning_rate": 3.510748959778086e-06,
+      "loss": 0.4798,
+      "step": 5093
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.056535720825195,
+      "learning_rate": 3.5055478502080447e-06,
+      "loss": 0.605,
+      "step": 5094
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.569316864013672,
+      "learning_rate": 3.500346740638003e-06,
+      "loss": 0.4705,
+      "step": 5095
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.7147440910339355,
+      "learning_rate": 3.4951456310679615e-06,
+      "loss": 0.3143,
+      "step": 5096
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.007050514221191,
+      "learning_rate": 3.4899445214979197e-06,
+      "loss": 0.6616,
+      "step": 5097
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.142555236816406,
+      "learning_rate": 3.4847434119278783e-06,
+      "loss": 0.5071,
+      "step": 5098
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.954296112060547,
+      "learning_rate": 3.4795423023578365e-06,
+      "loss": 0.6579,
+      "step": 5099
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.29862117767334,
+      "learning_rate": 3.474341192787795e-06,
+      "loss": 0.5434,
+      "step": 5100
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.657697677612305,
+      "learning_rate": 3.4691400832177533e-06,
+      "loss": 0.5599,
+      "step": 5101
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.100842475891113,
+      "learning_rate": 3.463938973647712e-06,
+      "loss": 0.4658,
+      "step": 5102
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.64625358581543,
+      "learning_rate": 3.45873786407767e-06,
+      "loss": 0.5754,
+      "step": 5103
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.744253158569336,
+      "learning_rate": 3.4535367545076287e-06,
+      "loss": 0.662,
+      "step": 5104
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.170718669891357,
+      "learning_rate": 3.448335644937587e-06,
+      "loss": 0.407,
+      "step": 5105
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.85448932647705,
+      "learning_rate": 3.443134535367545e-06,
+      "loss": 0.5894,
+      "step": 5106
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.644410133361816,
+      "learning_rate": 3.4379334257975037e-06,
+      "loss": 0.7162,
+      "step": 5107
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.399460315704346,
+      "learning_rate": 3.432732316227462e-06,
+      "loss": 0.3921,
+      "step": 5108
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.522823333740234,
+      "learning_rate": 3.4275312066574205e-06,
+      "loss": 0.5723,
+      "step": 5109
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.601675987243652,
+      "learning_rate": 3.4223300970873787e-06,
+      "loss": 0.4757,
+      "step": 5110
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.862188816070557,
+      "learning_rate": 3.4171289875173373e-06,
+      "loss": 0.4422,
+      "step": 5111
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.991525650024414,
+      "learning_rate": 3.4119278779472955e-06,
+      "loss": 0.6586,
+      "step": 5112
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.528097152709961,
+      "learning_rate": 3.406726768377254e-06,
+      "loss": 0.5263,
+      "step": 5113
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.179000854492188,
+      "learning_rate": 3.4015256588072123e-06,
+      "loss": 0.3875,
+      "step": 5114
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.81576919555664,
+      "learning_rate": 3.396324549237171e-06,
+      "loss": 0.5812,
+      "step": 5115
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.933558464050293,
+      "learning_rate": 3.391123439667129e-06,
+      "loss": 0.4574,
+      "step": 5116
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.656196594238281,
+      "learning_rate": 3.3859223300970877e-06,
+      "loss": 0.6226,
+      "step": 5117
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.358199119567871,
+      "learning_rate": 3.380721220527046e-06,
+      "loss": 0.4794,
+      "step": 5118
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.789251327514648,
+      "learning_rate": 3.3755201109570045e-06,
+      "loss": 0.5701,
+      "step": 5119
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.651115417480469,
+      "learning_rate": 3.3703190013869627e-06,
+      "loss": 0.5134,
+      "step": 5120
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.873651504516602,
+      "learning_rate": 3.3651178918169213e-06,
+      "loss": 0.4319,
+      "step": 5121
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.28253173828125,
+      "learning_rate": 3.3599167822468795e-06,
+      "loss": 0.5116,
+      "step": 5122
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.039361000061035,
+      "learning_rate": 3.354715672676838e-06,
+      "loss": 0.4407,
+      "step": 5123
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.253868103027344,
+      "learning_rate": 3.3495145631067963e-06,
+      "loss": 0.4525,
+      "step": 5124
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.620678901672363,
+      "learning_rate": 3.344313453536755e-06,
+      "loss": 0.536,
+      "step": 5125
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.398587226867676,
+      "learning_rate": 3.339112343966713e-06,
+      "loss": 0.3159,
+      "step": 5126
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.76865291595459,
+      "learning_rate": 3.3339112343966717e-06,
+      "loss": 0.521,
+      "step": 5127
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.530657768249512,
+      "learning_rate": 3.32871012482663e-06,
+      "loss": 0.4623,
+      "step": 5128
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.379667282104492,
+      "learning_rate": 3.3235090152565885e-06,
+      "loss": 0.4932,
+      "step": 5129
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.95638656616211,
+      "learning_rate": 3.3183079056865467e-06,
+      "loss": 0.5636,
+      "step": 5130
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.394503593444824,
+      "learning_rate": 3.3131067961165053e-06,
+      "loss": 0.4504,
+      "step": 5131
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.446746826171875,
+      "learning_rate": 3.3079056865464635e-06,
+      "loss": 0.4974,
+      "step": 5132
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.826835632324219,
+      "learning_rate": 3.302704576976422e-06,
+      "loss": 0.4926,
+      "step": 5133
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.267595291137695,
+      "learning_rate": 3.2975034674063803e-06,
+      "loss": 0.5498,
+      "step": 5134
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.812124252319336,
+      "learning_rate": 3.2923023578363385e-06,
+      "loss": 0.5157,
+      "step": 5135
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.93056583404541,
+      "learning_rate": 3.287101248266297e-06,
+      "loss": 0.4529,
+      "step": 5136
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.621736526489258,
+      "learning_rate": 3.2819001386962553e-06,
+      "loss": 0.6188,
+      "step": 5137
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.59837818145752,
+      "learning_rate": 3.2766990291262135e-06,
+      "loss": 0.5405,
+      "step": 5138
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.690279960632324,
+      "learning_rate": 3.2714979195561717e-06,
+      "loss": 0.5402,
+      "step": 5139
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.322529792785645,
+      "learning_rate": 3.2662968099861303e-06,
+      "loss": 0.668,
+      "step": 5140
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.626975059509277,
+      "learning_rate": 3.2610957004160885e-06,
+      "loss": 0.5092,
+      "step": 5141
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.521129131317139,
+      "learning_rate": 3.255894590846047e-06,
+      "loss": 0.438,
+      "step": 5142
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.497949600219727,
+      "learning_rate": 3.2506934812760053e-06,
+      "loss": 0.5666,
+      "step": 5143
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.912392616271973,
+      "learning_rate": 3.245492371705964e-06,
+      "loss": 0.5676,
+      "step": 5144
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.243478775024414,
+      "learning_rate": 3.240291262135922e-06,
+      "loss": 0.5331,
+      "step": 5145
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.305438041687012,
+      "learning_rate": 3.2350901525658807e-06,
+      "loss": 0.4277,
+      "step": 5146
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.113061904907227,
+      "learning_rate": 3.229889042995839e-06,
+      "loss": 0.5414,
+      "step": 5147
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.696189880371094,
+      "learning_rate": 3.2246879334257975e-06,
+      "loss": 0.6689,
+      "step": 5148
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.476223945617676,
+      "learning_rate": 3.2194868238557557e-06,
+      "loss": 0.5628,
+      "step": 5149
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.2350640296936035,
+      "learning_rate": 3.2142857142857143e-06,
+      "loss": 0.3421,
+      "step": 5150
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.876107215881348,
+      "learning_rate": 3.2090846047156725e-06,
+      "loss": 0.3202,
+      "step": 5151
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.153095245361328,
+      "learning_rate": 3.203883495145631e-06,
+      "loss": 0.7282,
+      "step": 5152
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.719401359558105,
+      "learning_rate": 3.1986823855755893e-06,
+      "loss": 0.5836,
+      "step": 5153
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.909844398498535,
+      "learning_rate": 3.193481276005548e-06,
+      "loss": 0.541,
+      "step": 5154
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.760603904724121,
+      "learning_rate": 3.188280166435506e-06,
+      "loss": 0.3924,
+      "step": 5155
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.751255512237549,
+      "learning_rate": 3.1830790568654647e-06,
+      "loss": 0.355,
+      "step": 5156
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.88051700592041,
+      "learning_rate": 3.177877947295423e-06,
+      "loss": 0.6016,
+      "step": 5157
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.878511428833008,
+      "learning_rate": 3.1726768377253815e-06,
+      "loss": 0.4771,
+      "step": 5158
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.28941535949707,
+      "learning_rate": 3.1674757281553397e-06,
+      "loss": 0.5018,
+      "step": 5159
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.609025955200195,
+      "learning_rate": 3.1622746185852983e-06,
+      "loss": 0.5823,
+      "step": 5160
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.898832321166992,
+      "learning_rate": 3.1570735090152565e-06,
+      "loss": 0.6523,
+      "step": 5161
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.355817794799805,
+      "learning_rate": 3.151872399445215e-06,
+      "loss": 0.6577,
+      "step": 5162
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.248767852783203,
+      "learning_rate": 3.1466712898751733e-06,
+      "loss": 0.572,
+      "step": 5163
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.17566204071045,
+      "learning_rate": 3.1414701803051315e-06,
+      "loss": 0.4225,
+      "step": 5164
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.516551971435547,
+      "learning_rate": 3.13626907073509e-06,
+      "loss": 0.6438,
+      "step": 5165
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.48836612701416,
+      "learning_rate": 3.1310679611650483e-06,
+      "loss": 0.4762,
+      "step": 5166
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.812480926513672,
+      "learning_rate": 3.125866851595007e-06,
+      "loss": 0.4074,
+      "step": 5167
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.504562377929688,
+      "learning_rate": 3.120665742024965e-06,
+      "loss": 0.4118,
+      "step": 5168
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.213545799255371,
+      "learning_rate": 3.1154646324549237e-06,
+      "loss": 0.5204,
+      "step": 5169
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.762345314025879,
+      "learning_rate": 3.110263522884882e-06,
+      "loss": 0.6366,
+      "step": 5170
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.883951187133789,
+      "learning_rate": 3.1050624133148405e-06,
+      "loss": 0.6018,
+      "step": 5171
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.050641059875488,
+      "learning_rate": 3.0998613037447987e-06,
+      "loss": 0.5525,
+      "step": 5172
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.763492107391357,
+      "learning_rate": 3.0946601941747573e-06,
+      "loss": 0.4688,
+      "step": 5173
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.20591926574707,
+      "learning_rate": 3.0894590846047155e-06,
+      "loss": 0.6111,
+      "step": 5174
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.419022560119629,
+      "learning_rate": 3.084257975034674e-06,
+      "loss": 0.5529,
+      "step": 5175
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.943574905395508,
+      "learning_rate": 3.0790568654646323e-06,
+      "loss": 0.4726,
+      "step": 5176
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.375443458557129,
+      "learning_rate": 3.073855755894591e-06,
+      "loss": 0.4544,
+      "step": 5177
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.67541790008545,
+      "learning_rate": 3.068654646324549e-06,
+      "loss": 0.4389,
+      "step": 5178
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.95535945892334,
+      "learning_rate": 3.0634535367545077e-06,
+      "loss": 0.419,
+      "step": 5179
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.732804298400879,
+      "learning_rate": 3.058252427184466e-06,
+      "loss": 0.617,
+      "step": 5180
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.88866138458252,
+      "learning_rate": 3.0530513176144245e-06,
+      "loss": 0.4791,
+      "step": 5181
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.839919090270996,
+      "learning_rate": 3.0478502080443827e-06,
+      "loss": 0.7243,
+      "step": 5182
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.665588855743408,
+      "learning_rate": 3.0426490984743413e-06,
+      "loss": 0.3258,
+      "step": 5183
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.950766563415527,
+      "learning_rate": 3.0374479889042995e-06,
+      "loss": 0.4978,
+      "step": 5184
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.863330841064453,
+      "learning_rate": 3.032246879334258e-06,
+      "loss": 0.5683,
+      "step": 5185
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.325726509094238,
+      "learning_rate": 3.0270457697642163e-06,
+      "loss": 0.6101,
+      "step": 5186
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.004590034484863,
+      "learning_rate": 3.021844660194175e-06,
+      "loss": 0.5501,
+      "step": 5187
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.310755252838135,
+      "learning_rate": 3.016643550624133e-06,
+      "loss": 0.3844,
+      "step": 5188
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.86838436126709,
+      "learning_rate": 3.0114424410540917e-06,
+      "loss": 0.6165,
+      "step": 5189
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.01700210571289,
+      "learning_rate": 3.00624133148405e-06,
+      "loss": 0.4659,
+      "step": 5190
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.266008377075195,
+      "learning_rate": 3.0010402219140085e-06,
+      "loss": 0.6373,
+      "step": 5191
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.68026876449585,
+      "learning_rate": 2.9958391123439667e-06,
+      "loss": 0.4758,
+      "step": 5192
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.09383773803711,
+      "learning_rate": 2.990638002773925e-06,
+      "loss": 0.4982,
+      "step": 5193
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.797573089599609,
+      "learning_rate": 2.9854368932038835e-06,
+      "loss": 0.4612,
+      "step": 5194
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.983869552612305,
+      "learning_rate": 2.9802357836338417e-06,
+      "loss": 0.5807,
+      "step": 5195
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.290656089782715,
+      "learning_rate": 2.9750346740638003e-06,
+      "loss": 0.4398,
+      "step": 5196
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.115141868591309,
+      "learning_rate": 2.9698335644937585e-06,
+      "loss": 0.718,
+      "step": 5197
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.781490325927734,
+      "learning_rate": 2.964632454923717e-06,
+      "loss": 0.3563,
+      "step": 5198
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.809248447418213,
+      "learning_rate": 2.9594313453536753e-06,
+      "loss": 0.3461,
+      "step": 5199
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.359139442443848,
+      "learning_rate": 2.954230235783634e-06,
+      "loss": 0.5305,
+      "step": 5200
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.278435707092285,
+      "learning_rate": 2.949029126213592e-06,
+      "loss": 0.7575,
+      "step": 5201
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.774267196655273,
+      "learning_rate": 2.9438280166435507e-06,
+      "loss": 0.6629,
+      "step": 5202
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.031609535217285,
+      "learning_rate": 2.938626907073509e-06,
+      "loss": 0.5223,
+      "step": 5203
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.484367847442627,
+      "learning_rate": 2.9334257975034675e-06,
+      "loss": 0.4362,
+      "step": 5204
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.879771709442139,
+      "learning_rate": 2.9282246879334257e-06,
+      "loss": 0.4095,
+      "step": 5205
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.406259536743164,
+      "learning_rate": 2.9230235783633843e-06,
+      "loss": 0.5681,
+      "step": 5206
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.694445610046387,
+      "learning_rate": 2.9178224687933425e-06,
+      "loss": 0.4359,
+      "step": 5207
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.090280532836914,
+      "learning_rate": 2.912621359223301e-06,
+      "loss": 0.6076,
+      "step": 5208
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.998620986938477,
+      "learning_rate": 2.9074202496532593e-06,
+      "loss": 0.5123,
+      "step": 5209
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.081467628479004,
+      "learning_rate": 2.902219140083218e-06,
+      "loss": 0.522,
+      "step": 5210
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.796858787536621,
+      "learning_rate": 2.897018030513176e-06,
+      "loss": 0.5434,
+      "step": 5211
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.902584075927734,
+      "learning_rate": 2.8918169209431347e-06,
+      "loss": 0.4601,
+      "step": 5212
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.346362590789795,
+      "learning_rate": 2.886615811373093e-06,
+      "loss": 0.5161,
+      "step": 5213
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.09678840637207,
+      "learning_rate": 2.8814147018030515e-06,
+      "loss": 0.6071,
+      "step": 5214
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.468668937683105,
+      "learning_rate": 2.8762135922330097e-06,
+      "loss": 0.6697,
+      "step": 5215
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.404544830322266,
+      "learning_rate": 2.8710124826629683e-06,
+      "loss": 0.4509,
+      "step": 5216
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.278600692749023,
+      "learning_rate": 2.8658113730929265e-06,
+      "loss": 0.516,
+      "step": 5217
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.174261093139648,
+      "learning_rate": 2.860610263522885e-06,
+      "loss": 0.5035,
+      "step": 5218
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.344681739807129,
+      "learning_rate": 2.8554091539528433e-06,
+      "loss": 0.4995,
+      "step": 5219
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.092296600341797,
+      "learning_rate": 2.850208044382802e-06,
+      "loss": 0.5929,
+      "step": 5220
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.519937515258789,
+      "learning_rate": 2.84500693481276e-06,
+      "loss": 0.422,
+      "step": 5221
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.79766845703125,
+      "learning_rate": 2.8398058252427183e-06,
+      "loss": 0.4581,
+      "step": 5222
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.690959930419922,
+      "learning_rate": 2.834604715672677e-06,
+      "loss": 0.7295,
+      "step": 5223
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.984220027923584,
+      "learning_rate": 2.829403606102635e-06,
+      "loss": 0.4652,
+      "step": 5224
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.503026008605957,
+      "learning_rate": 2.8242024965325937e-06,
+      "loss": 0.6358,
+      "step": 5225
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.675196647644043,
+      "learning_rate": 2.819001386962552e-06,
+      "loss": 0.6058,
+      "step": 5226
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.048329830169678,
+      "learning_rate": 2.8138002773925105e-06,
+      "loss": 0.4209,
+      "step": 5227
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.705938339233398,
+      "learning_rate": 2.8085991678224687e-06,
+      "loss": 0.532,
+      "step": 5228
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.166681289672852,
+      "learning_rate": 2.8033980582524273e-06,
+      "loss": 0.5039,
+      "step": 5229
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.74935245513916,
+      "learning_rate": 2.7981969486823855e-06,
+      "loss": 0.3808,
+      "step": 5230
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.42692756652832,
+      "learning_rate": 2.792995839112344e-06,
+      "loss": 0.5136,
+      "step": 5231
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.359210968017578,
+      "learning_rate": 2.7877947295423023e-06,
+      "loss": 0.424,
+      "step": 5232
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.956768989562988,
+      "learning_rate": 2.782593619972261e-06,
+      "loss": 0.5752,
+      "step": 5233
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.244669437408447,
+      "learning_rate": 2.777392510402219e-06,
+      "loss": 0.3862,
+      "step": 5234
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.538180351257324,
+      "learning_rate": 2.7721914008321777e-06,
+      "loss": 0.5076,
+      "step": 5235
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.4874114990234375,
+      "learning_rate": 2.766990291262136e-06,
+      "loss": 0.394,
+      "step": 5236
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.77630615234375,
+      "learning_rate": 2.7617891816920945e-06,
+      "loss": 0.5094,
+      "step": 5237
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.938010215759277,
+      "learning_rate": 2.7565880721220527e-06,
+      "loss": 0.7631,
+      "step": 5238
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.46627426147461,
+      "learning_rate": 2.7513869625520113e-06,
+      "loss": 0.4222,
+      "step": 5239
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.860413551330566,
+      "learning_rate": 2.7461858529819695e-06,
+      "loss": 0.5381,
+      "step": 5240
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.136298179626465,
+      "learning_rate": 2.740984743411928e-06,
+      "loss": 0.6334,
+      "step": 5241
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.265201568603516,
+      "learning_rate": 2.7357836338418863e-06,
+      "loss": 0.583,
+      "step": 5242
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.982163429260254,
+      "learning_rate": 2.730582524271845e-06,
+      "loss": 0.464,
+      "step": 5243
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.913843631744385,
+      "learning_rate": 2.725381414701803e-06,
+      "loss": 0.4098,
+      "step": 5244
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.160569190979004,
+      "learning_rate": 2.7201803051317617e-06,
+      "loss": 0.4543,
+      "step": 5245
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.121781826019287,
+      "learning_rate": 2.71497919556172e-06,
+      "loss": 0.321,
+      "step": 5246
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.15789794921875,
+      "learning_rate": 2.7097780859916785e-06,
+      "loss": 0.5464,
+      "step": 5247
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.849469184875488,
+      "learning_rate": 2.7045769764216367e-06,
+      "loss": 0.4457,
+      "step": 5248
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.328103065490723,
+      "learning_rate": 2.699375866851595e-06,
+      "loss": 0.555,
+      "step": 5249
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.87508773803711,
+      "learning_rate": 2.6941747572815535e-06,
+      "loss": 0.6775,
+      "step": 5250
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.241524696350098,
+      "learning_rate": 2.6889736477115117e-06,
+      "loss": 0.5521,
+      "step": 5251
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.993924140930176,
+      "learning_rate": 2.6837725381414703e-06,
+      "loss": 0.6343,
+      "step": 5252
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.316781044006348,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 0.6614,
+      "step": 5253
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.372742652893066,
+      "learning_rate": 2.673370319001387e-06,
+      "loss": 0.6833,
+      "step": 5254
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.608357429504395,
+      "learning_rate": 2.6681692094313453e-06,
+      "loss": 0.3472,
+      "step": 5255
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.530796527862549,
+      "learning_rate": 2.662968099861304e-06,
+      "loss": 0.3707,
+      "step": 5256
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.260453224182129,
+      "learning_rate": 2.657766990291262e-06,
+      "loss": 0.3999,
+      "step": 5257
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.115459442138672,
+      "learning_rate": 2.6525658807212207e-06,
+      "loss": 0.5979,
+      "step": 5258
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.591205596923828,
+      "learning_rate": 2.647364771151179e-06,
+      "loss": 0.5432,
+      "step": 5259
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.869034767150879,
+      "learning_rate": 2.6421636615811375e-06,
+      "loss": 0.4625,
+      "step": 5260
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.837038040161133,
+      "learning_rate": 2.6369625520110957e-06,
+      "loss": 0.4853,
+      "step": 5261
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.689262390136719,
+      "learning_rate": 2.6317614424410543e-06,
+      "loss": 0.4437,
+      "step": 5262
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.190145492553711,
+      "learning_rate": 2.6265603328710125e-06,
+      "loss": 0.4729,
+      "step": 5263
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.703058242797852,
+      "learning_rate": 2.621359223300971e-06,
+      "loss": 0.5345,
+      "step": 5264
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.674890518188477,
+      "learning_rate": 2.6161581137309293e-06,
+      "loss": 0.5921,
+      "step": 5265
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.038079261779785,
+      "learning_rate": 2.610957004160888e-06,
+      "loss": 0.5299,
+      "step": 5266
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.060151100158691,
+      "learning_rate": 2.605755894590846e-06,
+      "loss": 0.5282,
+      "step": 5267
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.185263633728027,
+      "learning_rate": 2.6005547850208047e-06,
+      "loss": 0.4106,
+      "step": 5268
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.756791114807129,
+      "learning_rate": 2.595353675450763e-06,
+      "loss": 0.557,
+      "step": 5269
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.644769668579102,
+      "learning_rate": 2.5901525658807215e-06,
+      "loss": 0.6169,
+      "step": 5270
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.885109901428223,
+      "learning_rate": 2.5849514563106797e-06,
+      "loss": 0.6202,
+      "step": 5271
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.204273223876953,
+      "learning_rate": 2.5797503467406383e-06,
+      "loss": 0.5136,
+      "step": 5272
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.340177059173584,
+      "learning_rate": 2.5745492371705965e-06,
+      "loss": 0.4421,
+      "step": 5273
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.935446262359619,
+      "learning_rate": 2.569348127600555e-06,
+      "loss": 0.5107,
+      "step": 5274
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.863086223602295,
+      "learning_rate": 2.5641470180305133e-06,
+      "loss": 0.607,
+      "step": 5275
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.92444896697998,
+      "learning_rate": 2.558945908460472e-06,
+      "loss": 0.4221,
+      "step": 5276
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.240103721618652,
+      "learning_rate": 2.55374479889043e-06,
+      "loss": 0.5972,
+      "step": 5277
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.008270263671875,
+      "learning_rate": 2.5485436893203883e-06,
+      "loss": 0.4839,
+      "step": 5278
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.223001480102539,
+      "learning_rate": 2.543342579750347e-06,
+      "loss": 0.5098,
+      "step": 5279
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.784637928009033,
+      "learning_rate": 2.538141470180305e-06,
+      "loss": 0.5256,
+      "step": 5280
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.747617721557617,
+      "learning_rate": 2.5329403606102637e-06,
+      "loss": 0.6416,
+      "step": 5281
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.727588653564453,
+      "learning_rate": 2.527739251040222e-06,
+      "loss": 0.4431,
+      "step": 5282
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.307159423828125,
+      "learning_rate": 2.5225381414701805e-06,
+      "loss": 0.6512,
+      "step": 5283
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.520806789398193,
+      "learning_rate": 2.5173370319001387e-06,
+      "loss": 0.3392,
+      "step": 5284
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.264139175415039,
+      "learning_rate": 2.5121359223300973e-06,
+      "loss": 0.5048,
+      "step": 5285
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.481681823730469,
+      "learning_rate": 2.5069348127600555e-06,
+      "loss": 0.572,
+      "step": 5286
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.879159450531006,
+      "learning_rate": 2.501733703190014e-06,
+      "loss": 0.3575,
+      "step": 5287
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.386326789855957,
+      "learning_rate": 2.4965325936199723e-06,
+      "loss": 0.6063,
+      "step": 5288
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.9546332359313965,
+      "learning_rate": 2.491331484049931e-06,
+      "loss": 0.5085,
+      "step": 5289
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.258877754211426,
+      "learning_rate": 2.486130374479889e-06,
+      "loss": 0.3482,
+      "step": 5290
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.44167709350586,
+      "learning_rate": 2.4809292649098477e-06,
+      "loss": 0.7855,
+      "step": 5291
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.956826210021973,
+      "learning_rate": 2.475728155339806e-06,
+      "loss": 0.4497,
+      "step": 5292
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.19681167602539,
+      "learning_rate": 2.4705270457697645e-06,
+      "loss": 0.5351,
+      "step": 5293
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.679778099060059,
+      "learning_rate": 2.4653259361997227e-06,
+      "loss": 0.5886,
+      "step": 5294
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.887847900390625,
+      "learning_rate": 2.4601248266296813e-06,
+      "loss": 0.5936,
+      "step": 5295
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.742365837097168,
+      "learning_rate": 2.4549237170596395e-06,
+      "loss": 0.358,
+      "step": 5296
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.748147964477539,
+      "learning_rate": 2.449722607489598e-06,
+      "loss": 0.3495,
+      "step": 5297
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.0347318649292,
+      "learning_rate": 2.4445214979195563e-06,
+      "loss": 0.4638,
+      "step": 5298
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.169685363769531,
+      "learning_rate": 2.439320388349515e-06,
+      "loss": 0.4882,
+      "step": 5299
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.598620414733887,
+      "learning_rate": 2.434119278779473e-06,
+      "loss": 0.6129,
+      "step": 5300
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.468547821044922,
+      "learning_rate": 2.4289181692094318e-06,
+      "loss": 0.5397,
+      "step": 5301
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.236961364746094,
+      "learning_rate": 2.42371705963939e-06,
+      "loss": 0.6679,
+      "step": 5302
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.051382064819336,
+      "learning_rate": 2.4185159500693486e-06,
+      "loss": 0.5623,
+      "step": 5303
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.312117576599121,
+      "learning_rate": 2.4133148404993067e-06,
+      "loss": 0.528,
+      "step": 5304
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.416163444519043,
+      "learning_rate": 2.4081137309292654e-06,
+      "loss": 0.5573,
+      "step": 5305
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.789859771728516,
+      "learning_rate": 2.4029126213592235e-06,
+      "loss": 0.5657,
+      "step": 5306
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.881768703460693,
+      "learning_rate": 2.3977115117891817e-06,
+      "loss": 0.528,
+      "step": 5307
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.989826202392578,
+      "learning_rate": 2.3925104022191403e-06,
+      "loss": 0.6351,
+      "step": 5308
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.130838394165039,
+      "learning_rate": 2.3873092926490985e-06,
+      "loss": 0.4499,
+      "step": 5309
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.118513107299805,
+      "learning_rate": 2.382108183079057e-06,
+      "loss": 0.5073,
+      "step": 5310
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.258551597595215,
+      "learning_rate": 2.3769070735090153e-06,
+      "loss": 0.7933,
+      "step": 5311
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.893186569213867,
+      "learning_rate": 2.371705963938974e-06,
+      "loss": 0.637,
+      "step": 5312
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.35175895690918,
+      "learning_rate": 2.366504854368932e-06,
+      "loss": 0.3987,
+      "step": 5313
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.651947021484375,
+      "learning_rate": 2.3613037447988907e-06,
+      "loss": 0.4067,
+      "step": 5314
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.844538688659668,
+      "learning_rate": 2.356102635228849e-06,
+      "loss": 0.4681,
+      "step": 5315
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.795241355895996,
+      "learning_rate": 2.3509015256588075e-06,
+      "loss": 0.4362,
+      "step": 5316
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.121227264404297,
+      "learning_rate": 2.3457004160887657e-06,
+      "loss": 0.5862,
+      "step": 5317
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.66531753540039,
+      "learning_rate": 2.340499306518724e-06,
+      "loss": 0.5211,
+      "step": 5318
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.236081600189209,
+      "learning_rate": 2.335298196948682e-06,
+      "loss": 0.4121,
+      "step": 5319
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.087831974029541,
+      "learning_rate": 2.3300970873786407e-06,
+      "loss": 0.4713,
+      "step": 5320
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.717195510864258,
+      "learning_rate": 2.324895977808599e-06,
+      "loss": 0.5742,
+      "step": 5321
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.623218536376953,
+      "learning_rate": 2.3196948682385575e-06,
+      "loss": 0.4791,
+      "step": 5322
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.326926231384277,
+      "learning_rate": 2.3144937586685157e-06,
+      "loss": 0.5411,
+      "step": 5323
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.117230415344238,
+      "learning_rate": 2.3092926490984743e-06,
+      "loss": 0.6223,
+      "step": 5324
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.496102333068848,
+      "learning_rate": 2.3040915395284325e-06,
+      "loss": 0.4509,
+      "step": 5325
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.239459991455078,
+      "learning_rate": 2.298890429958391e-06,
+      "loss": 0.4563,
+      "step": 5326
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.366392135620117,
+      "learning_rate": 2.2936893203883493e-06,
+      "loss": 0.6265,
+      "step": 5327
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.16166877746582,
+      "learning_rate": 2.288488210818308e-06,
+      "loss": 0.6355,
+      "step": 5328
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.446980476379395,
+      "learning_rate": 2.283287101248266e-06,
+      "loss": 0.5248,
+      "step": 5329
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.076930046081543,
+      "learning_rate": 2.2780859916782247e-06,
+      "loss": 0.6823,
+      "step": 5330
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.04551887512207,
+      "learning_rate": 2.272884882108183e-06,
+      "loss": 0.6891,
+      "step": 5331
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.048138618469238,
+      "learning_rate": 2.2676837725381415e-06,
+      "loss": 0.5723,
+      "step": 5332
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.243216514587402,
+      "learning_rate": 2.2624826629680997e-06,
+      "loss": 0.5489,
+      "step": 5333
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.852044105529785,
+      "learning_rate": 2.2572815533980583e-06,
+      "loss": 0.4487,
+      "step": 5334
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.690757751464844,
+      "learning_rate": 2.2520804438280165e-06,
+      "loss": 0.4775,
+      "step": 5335
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.312056541442871,
+      "learning_rate": 2.2468793342579747e-06,
+      "loss": 0.646,
+      "step": 5336
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.364581108093262,
+      "learning_rate": 2.2416782246879333e-06,
+      "loss": 0.4416,
+      "step": 5337
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.555436134338379,
+      "learning_rate": 2.2364771151178915e-06,
+      "loss": 0.4928,
+      "step": 5338
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.490155220031738,
+      "learning_rate": 2.23127600554785e-06,
+      "loss": 0.4041,
+      "step": 5339
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.041017532348633,
+      "learning_rate": 2.2260748959778083e-06,
+      "loss": 0.5915,
+      "step": 5340
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.60179615020752,
+      "learning_rate": 2.220873786407767e-06,
+      "loss": 0.4586,
+      "step": 5341
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.140592575073242,
+      "learning_rate": 2.215672676837725e-06,
+      "loss": 0.3746,
+      "step": 5342
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.835028648376465,
+      "learning_rate": 2.2104715672676837e-06,
+      "loss": 0.5142,
+      "step": 5343
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.578543663024902,
+      "learning_rate": 2.205270457697642e-06,
+      "loss": 0.5434,
+      "step": 5344
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.629638671875,
+      "learning_rate": 2.2000693481276005e-06,
+      "loss": 0.527,
+      "step": 5345
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.306648254394531,
+      "learning_rate": 2.1948682385575587e-06,
+      "loss": 0.4923,
+      "step": 5346
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.102993965148926,
+      "learning_rate": 2.1896671289875173e-06,
+      "loss": 0.3429,
+      "step": 5347
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.673915386199951,
+      "learning_rate": 2.1844660194174755e-06,
+      "loss": 0.4331,
+      "step": 5348
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.116801261901855,
+      "learning_rate": 2.179264909847434e-06,
+      "loss": 0.5227,
+      "step": 5349
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.23523235321045,
+      "learning_rate": 2.1740638002773923e-06,
+      "loss": 0.6114,
+      "step": 5350
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.317479133605957,
+      "learning_rate": 2.168862690707351e-06,
+      "loss": 0.4902,
+      "step": 5351
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.768195152282715,
+      "learning_rate": 2.163661581137309e-06,
+      "loss": 0.6388,
+      "step": 5352
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.972824096679688,
+      "learning_rate": 2.1584604715672677e-06,
+      "loss": 0.4744,
+      "step": 5353
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.882039070129395,
+      "learning_rate": 2.153259361997226e-06,
+      "loss": 0.6863,
+      "step": 5354
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.599577903747559,
+      "learning_rate": 2.1480582524271845e-06,
+      "loss": 0.5296,
+      "step": 5355
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.882028579711914,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.5386,
+      "step": 5356
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.97374439239502,
+      "learning_rate": 2.1376560332871013e-06,
+      "loss": 0.5889,
+      "step": 5357
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.996978759765625,
+      "learning_rate": 2.1324549237170595e-06,
+      "loss": 0.4299,
+      "step": 5358
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.675373077392578,
+      "learning_rate": 2.127253814147018e-06,
+      "loss": 0.5893,
+      "step": 5359
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.240413665771484,
+      "learning_rate": 2.1220527045769763e-06,
+      "loss": 0.4315,
+      "step": 5360
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.738235473632812,
+      "learning_rate": 2.116851595006935e-06,
+      "loss": 0.5736,
+      "step": 5361
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.012187004089355,
+      "learning_rate": 2.111650485436893e-06,
+      "loss": 0.532,
+      "step": 5362
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.261618614196777,
+      "learning_rate": 2.1064493758668517e-06,
+      "loss": 0.6727,
+      "step": 5363
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.339353561401367,
+      "learning_rate": 2.10124826629681e-06,
+      "loss": 0.6094,
+      "step": 5364
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.814212799072266,
+      "learning_rate": 2.096047156726768e-06,
+      "loss": 0.5725,
+      "step": 5365
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.956605911254883,
+      "learning_rate": 2.0908460471567267e-06,
+      "loss": 0.6143,
+      "step": 5366
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.191831588745117,
+      "learning_rate": 2.085644937586685e-06,
+      "loss": 0.5815,
+      "step": 5367
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.4771552085876465,
+      "learning_rate": 2.0804438280166435e-06,
+      "loss": 0.4082,
+      "step": 5368
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.205078125,
+      "learning_rate": 2.0752427184466017e-06,
+      "loss": 0.8029,
+      "step": 5369
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.498282432556152,
+      "learning_rate": 2.0700416088765603e-06,
+      "loss": 0.4939,
+      "step": 5370
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.823099613189697,
+      "learning_rate": 2.0648404993065185e-06,
+      "loss": 0.4639,
+      "step": 5371
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.29514217376709,
+      "learning_rate": 2.059639389736477e-06,
+      "loss": 0.5409,
+      "step": 5372
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.73094654083252,
+      "learning_rate": 2.0544382801664353e-06,
+      "loss": 0.6428,
+      "step": 5373
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.7875337600708,
+      "learning_rate": 2.049237170596394e-06,
+      "loss": 0.6755,
+      "step": 5374
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.925169944763184,
+      "learning_rate": 2.044036061026352e-06,
+      "loss": 0.5171,
+      "step": 5375
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.447174072265625,
+      "learning_rate": 2.0388349514563107e-06,
+      "loss": 0.6049,
+      "step": 5376
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.670456886291504,
+      "learning_rate": 2.033633841886269e-06,
+      "loss": 0.4817,
+      "step": 5377
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.747203826904297,
+      "learning_rate": 2.0284327323162275e-06,
+      "loss": 0.4458,
+      "step": 5378
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.75134563446045,
+      "learning_rate": 2.0232316227461857e-06,
+      "loss": 0.513,
+      "step": 5379
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.8532328605651855,
+      "learning_rate": 2.0180305131761443e-06,
+      "loss": 0.5558,
+      "step": 5380
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.561273574829102,
+      "learning_rate": 2.0128294036061025e-06,
+      "loss": 0.5024,
+      "step": 5381
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.286358833312988,
+      "learning_rate": 2.007628294036061e-06,
+      "loss": 0.5565,
+      "step": 5382
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.124828338623047,
+      "learning_rate": 2.0024271844660193e-06,
+      "loss": 0.5204,
+      "step": 5383
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.831273555755615,
+      "learning_rate": 1.997226074895978e-06,
+      "loss": 0.4164,
+      "step": 5384
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.941850662231445,
+      "learning_rate": 1.992024965325936e-06,
+      "loss": 0.4787,
+      "step": 5385
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.027563095092773,
+      "learning_rate": 1.9868238557558948e-06,
+      "loss": 0.549,
+      "step": 5386
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.545065879821777,
+      "learning_rate": 1.981622746185853e-06,
+      "loss": 0.5361,
+      "step": 5387
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.703097343444824,
+      "learning_rate": 1.9764216366158116e-06,
+      "loss": 0.4844,
+      "step": 5388
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.384164810180664,
+      "learning_rate": 1.9712205270457697e-06,
+      "loss": 0.5296,
+      "step": 5389
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.210725784301758,
+      "learning_rate": 1.9660194174757284e-06,
+      "loss": 0.4025,
+      "step": 5390
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.163189888000488,
+      "learning_rate": 1.9608183079056865e-06,
+      "loss": 0.6761,
+      "step": 5391
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.4579668045043945,
+      "learning_rate": 1.955617198335645e-06,
+      "loss": 0.4394,
+      "step": 5392
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.00284194946289,
+      "learning_rate": 1.9504160887656033e-06,
+      "loss": 0.5571,
+      "step": 5393
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.355062484741211,
+      "learning_rate": 1.9452149791955615e-06,
+      "loss": 0.5091,
+      "step": 5394
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.213737487792969,
+      "learning_rate": 1.94001386962552e-06,
+      "loss": 0.5164,
+      "step": 5395
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.67052173614502,
+      "learning_rate": 1.9348127600554783e-06,
+      "loss": 0.543,
+      "step": 5396
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.801335334777832,
+      "learning_rate": 1.929611650485437e-06,
+      "loss": 0.398,
+      "step": 5397
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.9794230461120605,
+      "learning_rate": 1.924410540915395e-06,
+      "loss": 0.512,
+      "step": 5398
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.415800094604492,
+      "learning_rate": 1.9192094313453537e-06,
+      "loss": 0.6752,
+      "step": 5399
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.890510559082031,
+      "learning_rate": 1.914008321775312e-06,
+      "loss": 0.5494,
+      "step": 5400
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.436013698577881,
+      "learning_rate": 1.9088072122052705e-06,
+      "loss": 0.3431,
+      "step": 5401
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.754990100860596,
+      "learning_rate": 1.903606102635229e-06,
+      "loss": 0.5204,
+      "step": 5402
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.385466575622559,
+      "learning_rate": 1.8984049930651874e-06,
+      "loss": 0.5945,
+      "step": 5403
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.527352333068848,
+      "learning_rate": 1.8932038834951458e-06,
+      "loss": 0.4973,
+      "step": 5404
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.268107414245605,
+      "learning_rate": 1.8880027739251042e-06,
+      "loss": 0.5482,
+      "step": 5405
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.46562671661377,
+      "learning_rate": 1.8828016643550626e-06,
+      "loss": 0.6225,
+      "step": 5406
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.252921104431152,
+      "learning_rate": 1.877600554785021e-06,
+      "loss": 0.5019,
+      "step": 5407
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.840207576751709,
+      "learning_rate": 1.8723994452149791e-06,
+      "loss": 0.4636,
+      "step": 5408
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.624959945678711,
+      "learning_rate": 1.8671983356449375e-06,
+      "loss": 0.5816,
+      "step": 5409
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.170961380004883,
+      "learning_rate": 1.861997226074896e-06,
+      "loss": 0.5611,
+      "step": 5410
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.581083297729492,
+      "learning_rate": 1.8567961165048543e-06,
+      "loss": 0.518,
+      "step": 5411
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.634629249572754,
+      "learning_rate": 1.8515950069348127e-06,
+      "loss": 0.3075,
+      "step": 5412
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.085752487182617,
+      "learning_rate": 1.8463938973647711e-06,
+      "loss": 0.6046,
+      "step": 5413
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.451269149780273,
+      "learning_rate": 1.8411927877947295e-06,
+      "loss": 0.5631,
+      "step": 5414
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.565841674804688,
+      "learning_rate": 1.835991678224688e-06,
+      "loss": 0.5644,
+      "step": 5415
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.718992233276367,
+      "learning_rate": 1.8307905686546463e-06,
+      "loss": 0.551,
+      "step": 5416
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.014156818389893,
+      "learning_rate": 1.8255894590846047e-06,
+      "loss": 0.3809,
+      "step": 5417
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.131178379058838,
+      "learning_rate": 1.8203883495145631e-06,
+      "loss": 0.3918,
+      "step": 5418
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.13056755065918,
+      "learning_rate": 1.8151872399445215e-06,
+      "loss": 0.5599,
+      "step": 5419
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.526254653930664,
+      "learning_rate": 1.80998613037448e-06,
+      "loss": 0.3622,
+      "step": 5420
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.74657917022705,
+      "learning_rate": 1.8047850208044384e-06,
+      "loss": 0.6992,
+      "step": 5421
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.88620376586914,
+      "learning_rate": 1.7995839112343968e-06,
+      "loss": 0.6705,
+      "step": 5422
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.42060375213623,
+      "learning_rate": 1.7943828016643552e-06,
+      "loss": 0.3095,
+      "step": 5423
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.153654098510742,
+      "learning_rate": 1.7891816920943136e-06,
+      "loss": 0.4034,
+      "step": 5424
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.137581825256348,
+      "learning_rate": 1.783980582524272e-06,
+      "loss": 0.4475,
+      "step": 5425
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.853110313415527,
+      "learning_rate": 1.7787794729542304e-06,
+      "loss": 0.6746,
+      "step": 5426
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.56912612915039,
+      "learning_rate": 1.7735783633841888e-06,
+      "loss": 0.6892,
+      "step": 5427
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.735822677612305,
+      "learning_rate": 1.7683772538141472e-06,
+      "loss": 0.4847,
+      "step": 5428
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.768712997436523,
+      "learning_rate": 1.7631761442441056e-06,
+      "loss": 0.4239,
+      "step": 5429
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.5568342208862305,
+      "learning_rate": 1.757975034674064e-06,
+      "loss": 0.3879,
+      "step": 5430
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.890365600585938,
+      "learning_rate": 1.7527739251040224e-06,
+      "loss": 0.4561,
+      "step": 5431
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.315093040466309,
+      "learning_rate": 1.7475728155339808e-06,
+      "loss": 0.6191,
+      "step": 5432
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.4700870513916,
+      "learning_rate": 1.7423717059639392e-06,
+      "loss": 0.5468,
+      "step": 5433
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.750593185424805,
+      "learning_rate": 1.7371705963938976e-06,
+      "loss": 0.6775,
+      "step": 5434
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.606192588806152,
+      "learning_rate": 1.731969486823856e-06,
+      "loss": 0.4814,
+      "step": 5435
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.48621654510498,
+      "learning_rate": 1.7267683772538144e-06,
+      "loss": 0.5194,
+      "step": 5436
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.744473934173584,
+      "learning_rate": 1.7215672676837726e-06,
+      "loss": 0.4819,
+      "step": 5437
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.09116268157959,
+      "learning_rate": 1.716366158113731e-06,
+      "loss": 0.7241,
+      "step": 5438
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.74068832397461,
+      "learning_rate": 1.7111650485436894e-06,
+      "loss": 0.6624,
+      "step": 5439
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.505138874053955,
+      "learning_rate": 1.7059639389736478e-06,
+      "loss": 0.3759,
+      "step": 5440
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.167184829711914,
+      "learning_rate": 1.7007628294036062e-06,
+      "loss": 0.6395,
+      "step": 5441
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.316577434539795,
+      "learning_rate": 1.6955617198335646e-06,
+      "loss": 0.5104,
+      "step": 5442
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.024138450622559,
+      "learning_rate": 1.690360610263523e-06,
+      "loss": 0.4572,
+      "step": 5443
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.616570472717285,
+      "learning_rate": 1.6851595006934814e-06,
+      "loss": 0.5762,
+      "step": 5444
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.436549186706543,
+      "learning_rate": 1.6799583911234398e-06,
+      "loss": 0.6288,
+      "step": 5445
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.12216567993164,
+      "learning_rate": 1.6747572815533982e-06,
+      "loss": 0.5152,
+      "step": 5446
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.223217010498047,
+      "learning_rate": 1.6695561719833566e-06,
+      "loss": 0.6232,
+      "step": 5447
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.40604305267334,
+      "learning_rate": 1.664355062413315e-06,
+      "loss": 0.5591,
+      "step": 5448
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.703156471252441,
+      "learning_rate": 1.6591539528432734e-06,
+      "loss": 0.7926,
+      "step": 5449
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.601611137390137,
+      "learning_rate": 1.6539528432732318e-06,
+      "loss": 0.462,
+      "step": 5450
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.49284553527832,
+      "learning_rate": 1.6487517337031902e-06,
+      "loss": 0.4541,
+      "step": 5451
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.794659614562988,
+      "learning_rate": 1.6435506241331486e-06,
+      "loss": 0.3857,
+      "step": 5452
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.860626220703125,
+      "learning_rate": 1.6383495145631067e-06,
+      "loss": 0.5035,
+      "step": 5453
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.20423412322998,
+      "learning_rate": 1.6331484049930651e-06,
+      "loss": 0.4159,
+      "step": 5454
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.443599700927734,
+      "learning_rate": 1.6279472954230236e-06,
+      "loss": 0.5221,
+      "step": 5455
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.240882873535156,
+      "learning_rate": 1.622746185852982e-06,
+      "loss": 0.3428,
+      "step": 5456
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.079323768615723,
+      "learning_rate": 1.6175450762829404e-06,
+      "loss": 0.538,
+      "step": 5457
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.526915550231934,
+      "learning_rate": 1.6123439667128988e-06,
+      "loss": 0.577,
+      "step": 5458
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.84830379486084,
+      "learning_rate": 1.6071428571428572e-06,
+      "loss": 0.5129,
+      "step": 5459
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.446946144104004,
+      "learning_rate": 1.6019417475728156e-06,
+      "loss": 0.5258,
+      "step": 5460
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.12124252319336,
+      "learning_rate": 1.596740638002774e-06,
+      "loss": 0.4299,
+      "step": 5461
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.65672779083252,
+      "learning_rate": 1.5915395284327324e-06,
+      "loss": 0.5997,
+      "step": 5462
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.282958984375,
+      "learning_rate": 1.5863384188626908e-06,
+      "loss": 0.4163,
+      "step": 5463
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.955150604248047,
+      "learning_rate": 1.5811373092926492e-06,
+      "loss": 0.6099,
+      "step": 5464
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.218280792236328,
+      "learning_rate": 1.5759361997226076e-06,
+      "loss": 0.4506,
+      "step": 5465
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.897521018981934,
+      "learning_rate": 1.5707350901525657e-06,
+      "loss": 0.5317,
+      "step": 5466
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.453750610351562,
+      "learning_rate": 1.5655339805825241e-06,
+      "loss": 0.545,
+      "step": 5467
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.906431674957275,
+      "learning_rate": 1.5603328710124825e-06,
+      "loss": 0.3131,
+      "step": 5468
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.856907844543457,
+      "learning_rate": 1.555131761442441e-06,
+      "loss": 0.2662,
+      "step": 5469
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.156163215637207,
+      "learning_rate": 1.5499306518723993e-06,
+      "loss": 0.6276,
+      "step": 5470
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.133134841918945,
+      "learning_rate": 1.5447295423023577e-06,
+      "loss": 0.4697,
+      "step": 5471
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.973803520202637,
+      "learning_rate": 1.5395284327323162e-06,
+      "loss": 0.3893,
+      "step": 5472
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.403199195861816,
+      "learning_rate": 1.5343273231622746e-06,
+      "loss": 0.4967,
+      "step": 5473
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.99625301361084,
+      "learning_rate": 1.529126213592233e-06,
+      "loss": 0.4888,
+      "step": 5474
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.768814563751221,
+      "learning_rate": 1.5239251040221914e-06,
+      "loss": 0.5473,
+      "step": 5475
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.248258590698242,
+      "learning_rate": 1.5187239944521498e-06,
+      "loss": 0.56,
+      "step": 5476
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.538490295410156,
+      "learning_rate": 1.5135228848821082e-06,
+      "loss": 0.6675,
+      "step": 5477
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.83627462387085,
+      "learning_rate": 1.5083217753120666e-06,
+      "loss": 0.5546,
+      "step": 5478
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.156693458557129,
+      "learning_rate": 1.503120665742025e-06,
+      "loss": 0.3054,
+      "step": 5479
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.498602867126465,
+      "learning_rate": 1.4979195561719834e-06,
+      "loss": 0.564,
+      "step": 5480
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.2786226272583,
+      "learning_rate": 1.4927184466019418e-06,
+      "loss": 0.573,
+      "step": 5481
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.660609245300293,
+      "learning_rate": 1.4875173370319002e-06,
+      "loss": 0.4347,
+      "step": 5482
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.62446928024292,
+      "learning_rate": 1.4823162274618586e-06,
+      "loss": 0.304,
+      "step": 5483
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.452156066894531,
+      "learning_rate": 1.477115117891817e-06,
+      "loss": 0.3851,
+      "step": 5484
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.545473098754883,
+      "learning_rate": 1.4719140083217754e-06,
+      "loss": 0.4154,
+      "step": 5485
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.249227523803711,
+      "learning_rate": 1.4667128987517338e-06,
+      "loss": 0.4768,
+      "step": 5486
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.482326507568359,
+      "learning_rate": 1.4615117891816922e-06,
+      "loss": 0.4066,
+      "step": 5487
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.896811485290527,
+      "learning_rate": 1.4563106796116506e-06,
+      "loss": 0.6511,
+      "step": 5488
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.48816967010498,
+      "learning_rate": 1.451109570041609e-06,
+      "loss": 0.7442,
+      "step": 5489
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.929283618927002,
+      "learning_rate": 1.4459084604715674e-06,
+      "loss": 0.5301,
+      "step": 5490
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.20522403717041,
+      "learning_rate": 1.4407073509015258e-06,
+      "loss": 0.5357,
+      "step": 5491
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.822327613830566,
+      "learning_rate": 1.4355062413314842e-06,
+      "loss": 0.5057,
+      "step": 5492
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.723872184753418,
+      "learning_rate": 1.4303051317614426e-06,
+      "loss": 0.3747,
+      "step": 5493
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.940017700195312,
+      "learning_rate": 1.425104022191401e-06,
+      "loss": 0.8074,
+      "step": 5494
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.203953742980957,
+      "learning_rate": 1.4199029126213592e-06,
+      "loss": 0.5534,
+      "step": 5495
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.275420188903809,
+      "learning_rate": 1.4147018030513176e-06,
+      "loss": 0.5334,
+      "step": 5496
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.0225830078125,
+      "learning_rate": 1.409500693481276e-06,
+      "loss": 0.4392,
+      "step": 5497
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.001708030700684,
+      "learning_rate": 1.4042995839112344e-06,
+      "loss": 0.6259,
+      "step": 5498
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.011641502380371,
+      "learning_rate": 1.3990984743411928e-06,
+      "loss": 0.4605,
+      "step": 5499
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.901179313659668,
+      "learning_rate": 1.3938973647711512e-06,
+      "loss": 0.5258,
+      "step": 5500
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.026860237121582,
+      "learning_rate": 1.3886962552011096e-06,
+      "loss": 0.3714,
+      "step": 5501
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.25665855407715,
+      "learning_rate": 1.383495145631068e-06,
+      "loss": 0.3793,
+      "step": 5502
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.650860786437988,
+      "learning_rate": 1.3782940360610264e-06,
+      "loss": 0.5261,
+      "step": 5503
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.719388008117676,
+      "learning_rate": 1.3730929264909848e-06,
+      "loss": 0.4981,
+      "step": 5504
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.54290771484375,
+      "learning_rate": 1.3678918169209432e-06,
+      "loss": 0.4972,
+      "step": 5505
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.76274299621582,
+      "learning_rate": 1.3626907073509016e-06,
+      "loss": 0.6609,
+      "step": 5506
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.364433288574219,
+      "learning_rate": 1.35748959778086e-06,
+      "loss": 0.439,
+      "step": 5507
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.834692001342773,
+      "learning_rate": 1.3522884882108184e-06,
+      "loss": 0.5054,
+      "step": 5508
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.207436561584473,
+      "learning_rate": 1.3470873786407768e-06,
+      "loss": 0.5282,
+      "step": 5509
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.9845452308654785,
+      "learning_rate": 1.3418862690707352e-06,
+      "loss": 0.4222,
+      "step": 5510
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.785455703735352,
+      "learning_rate": 1.3366851595006936e-06,
+      "loss": 0.577,
+      "step": 5511
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.50716495513916,
+      "learning_rate": 1.331484049930652e-06,
+      "loss": 0.4997,
+      "step": 5512
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.832232475280762,
+      "learning_rate": 1.3262829403606104e-06,
+      "loss": 0.5087,
+      "step": 5513
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.203140258789062,
+      "learning_rate": 1.3210818307905688e-06,
+      "loss": 0.6111,
+      "step": 5514
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.718155860900879,
+      "learning_rate": 1.3158807212205272e-06,
+      "loss": 0.8928,
+      "step": 5515
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.444329261779785,
+      "learning_rate": 1.3106796116504856e-06,
+      "loss": 0.4791,
+      "step": 5516
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.639923095703125,
+      "learning_rate": 1.305478502080444e-06,
+      "loss": 0.3287,
+      "step": 5517
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.016860008239746,
+      "learning_rate": 1.3002773925104024e-06,
+      "loss": 0.4609,
+      "step": 5518
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.928682327270508,
+      "learning_rate": 1.2950762829403608e-06,
+      "loss": 0.3919,
+      "step": 5519
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.042181968688965,
+      "learning_rate": 1.2898751733703192e-06,
+      "loss": 0.395,
+      "step": 5520
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.315461158752441,
+      "learning_rate": 1.2846740638002776e-06,
+      "loss": 0.7908,
+      "step": 5521
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.524872779846191,
+      "learning_rate": 1.279472954230236e-06,
+      "loss": 0.6428,
+      "step": 5522
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.66862964630127,
+      "learning_rate": 1.2742718446601942e-06,
+      "loss": 0.6371,
+      "step": 5523
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.545085906982422,
+      "learning_rate": 1.2690707350901526e-06,
+      "loss": 0.4525,
+      "step": 5524
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.386002063751221,
+      "learning_rate": 1.263869625520111e-06,
+      "loss": 0.3033,
+      "step": 5525
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.462306022644043,
+      "learning_rate": 1.2586685159500694e-06,
+      "loss": 0.3847,
+      "step": 5526
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.18625259399414,
+      "learning_rate": 1.2534674063800278e-06,
+      "loss": 0.4899,
+      "step": 5527
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.48562240600586,
+      "learning_rate": 1.2482662968099862e-06,
+      "loss": 0.4416,
+      "step": 5528
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.525089263916016,
+      "learning_rate": 1.2430651872399446e-06,
+      "loss": 0.5211,
+      "step": 5529
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.641313552856445,
+      "learning_rate": 1.237864077669903e-06,
+      "loss": 0.5918,
+      "step": 5530
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.363886833190918,
+      "learning_rate": 1.2326629680998614e-06,
+      "loss": 0.3217,
+      "step": 5531
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.022553443908691,
+      "learning_rate": 1.2274618585298198e-06,
+      "loss": 0.4499,
+      "step": 5532
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.949962615966797,
+      "learning_rate": 1.2222607489597782e-06,
+      "loss": 0.6063,
+      "step": 5533
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.251152992248535,
+      "learning_rate": 1.2170596393897366e-06,
+      "loss": 0.4956,
+      "step": 5534
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.011673927307129,
+      "learning_rate": 1.211858529819695e-06,
+      "loss": 0.5221,
+      "step": 5535
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.54979419708252,
+      "learning_rate": 1.2066574202496534e-06,
+      "loss": 0.673,
+      "step": 5536
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.997461318969727,
+      "learning_rate": 1.2014563106796118e-06,
+      "loss": 0.6152,
+      "step": 5537
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.408784866333008,
+      "learning_rate": 1.1962552011095702e-06,
+      "loss": 0.5184,
+      "step": 5538
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.96095085144043,
+      "learning_rate": 1.1910540915395286e-06,
+      "loss": 0.5041,
+      "step": 5539
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.913379669189453,
+      "learning_rate": 1.185852981969487e-06,
+      "loss": 0.6589,
+      "step": 5540
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.125801086425781,
+      "learning_rate": 1.1806518723994454e-06,
+      "loss": 0.6088,
+      "step": 5541
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.75349235534668,
+      "learning_rate": 1.1754507628294038e-06,
+      "loss": 0.4515,
+      "step": 5542
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.111106872558594,
+      "learning_rate": 1.170249653259362e-06,
+      "loss": 0.4205,
+      "step": 5543
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.613080978393555,
+      "learning_rate": 1.1650485436893204e-06,
+      "loss": 0.5378,
+      "step": 5544
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.606620788574219,
+      "learning_rate": 1.1598474341192788e-06,
+      "loss": 0.5236,
+      "step": 5545
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.3320136070251465,
+      "learning_rate": 1.1546463245492372e-06,
+      "loss": 0.3967,
+      "step": 5546
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.550835609436035,
+      "learning_rate": 1.1494452149791956e-06,
+      "loss": 0.587,
+      "step": 5547
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.7941255569458,
+      "learning_rate": 1.144244105409154e-06,
+      "loss": 0.4759,
+      "step": 5548
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.590585708618164,
+      "learning_rate": 1.1390429958391124e-06,
+      "loss": 0.4614,
+      "step": 5549
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.713522911071777,
+      "learning_rate": 1.1338418862690708e-06,
+      "loss": 0.7678,
+      "step": 5550
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.034919738769531,
+      "learning_rate": 1.1286407766990292e-06,
+      "loss": 0.4746,
+      "step": 5551
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.659708976745605,
+      "learning_rate": 1.1234396671289874e-06,
+      "loss": 0.5689,
+      "step": 5552
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.155282974243164,
+      "learning_rate": 1.1182385575589458e-06,
+      "loss": 0.5005,
+      "step": 5553
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.288532257080078,
+      "learning_rate": 1.1130374479889042e-06,
+      "loss": 0.5599,
+      "step": 5554
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 16.77480125427246,
+      "learning_rate": 1.1078363384188626e-06,
+      "loss": 0.5257,
+      "step": 5555
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.132994651794434,
+      "learning_rate": 1.102635228848821e-06,
+      "loss": 0.7015,
+      "step": 5556
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.276415824890137,
+      "learning_rate": 1.0974341192787794e-06,
+      "loss": 0.4448,
+      "step": 5557
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.760940551757812,
+      "learning_rate": 1.0922330097087378e-06,
+      "loss": 0.4668,
+      "step": 5558
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.513123512268066,
+      "learning_rate": 1.0870319001386962e-06,
+      "loss": 0.6046,
+      "step": 5559
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.877487182617188,
+      "learning_rate": 1.0818307905686546e-06,
+      "loss": 0.6387,
+      "step": 5560
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.777705192565918,
+      "learning_rate": 1.076629680998613e-06,
+      "loss": 0.4165,
+      "step": 5561
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.244582176208496,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 0.5777,
+      "step": 5562
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.034939765930176,
+      "learning_rate": 1.0662274618585298e-06,
+      "loss": 0.6108,
+      "step": 5563
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.756478786468506,
+      "learning_rate": 1.0610263522884882e-06,
+      "loss": 0.5166,
+      "step": 5564
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.013072967529297,
+      "learning_rate": 1.0558252427184466e-06,
+      "loss": 0.4359,
+      "step": 5565
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.908638954162598,
+      "learning_rate": 1.050624133148405e-06,
+      "loss": 0.3583,
+      "step": 5566
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.58902359008789,
+      "learning_rate": 1.0454230235783634e-06,
+      "loss": 0.6108,
+      "step": 5567
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.80733585357666,
+      "learning_rate": 1.0402219140083218e-06,
+      "loss": 0.4954,
+      "step": 5568
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.35794448852539,
+      "learning_rate": 1.0350208044382802e-06,
+      "loss": 0.5685,
+      "step": 5569
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.23429012298584,
+      "learning_rate": 1.0298196948682386e-06,
+      "loss": 0.5734,
+      "step": 5570
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.009252548217773,
+      "learning_rate": 1.024618585298197e-06,
+      "loss": 0.4205,
+      "step": 5571
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.637924194335938,
+      "learning_rate": 1.0194174757281554e-06,
+      "loss": 0.4076,
+      "step": 5572
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.729819297790527,
+      "learning_rate": 1.0142163661581138e-06,
+      "loss": 0.6051,
+      "step": 5573
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.341536521911621,
+      "learning_rate": 1.0090152565880722e-06,
+      "loss": 0.458,
+      "step": 5574
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.228802680969238,
+      "learning_rate": 1.0038141470180306e-06,
+      "loss": 0.5031,
+      "step": 5575
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.386637687683105,
+      "learning_rate": 9.98613037447989e-07,
+      "loss": 0.5359,
+      "step": 5576
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.616072654724121,
+      "learning_rate": 9.934119278779474e-07,
+      "loss": 0.4613,
+      "step": 5577
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.324079513549805,
+      "learning_rate": 9.882108183079058e-07,
+      "loss": 0.4264,
+      "step": 5578
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.688997268676758,
+      "learning_rate": 9.830097087378642e-07,
+      "loss": 0.4032,
+      "step": 5579
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.204578876495361,
+      "learning_rate": 9.778085991678226e-07,
+      "loss": 0.3356,
+      "step": 5580
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.072381973266602,
+      "learning_rate": 9.726074895977808e-07,
+      "loss": 0.7199,
+      "step": 5581
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.6514253616333,
+      "learning_rate": 9.674063800277392e-07,
+      "loss": 0.4167,
+      "step": 5582
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.144237518310547,
+      "learning_rate": 9.622052704576976e-07,
+      "loss": 0.4494,
+      "step": 5583
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.686187744140625,
+      "learning_rate": 9.57004160887656e-07,
+      "loss": 0.5843,
+      "step": 5584
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.992888450622559,
+      "learning_rate": 9.518030513176145e-07,
+      "loss": 0.6097,
+      "step": 5585
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.166951179504395,
+      "learning_rate": 9.466019417475729e-07,
+      "loss": 0.6787,
+      "step": 5586
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.92952823638916,
+      "learning_rate": 9.414008321775313e-07,
+      "loss": 0.637,
+      "step": 5587
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.889827728271484,
+      "learning_rate": 9.361997226074896e-07,
+      "loss": 0.6354,
+      "step": 5588
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.474685668945312,
+      "learning_rate": 9.30998613037448e-07,
+      "loss": 0.4813,
+      "step": 5589
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.801338195800781,
+      "learning_rate": 9.257975034674064e-07,
+      "loss": 0.4642,
+      "step": 5590
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.5969953536987305,
+      "learning_rate": 9.205963938973648e-07,
+      "loss": 0.4804,
+      "step": 5591
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.773614883422852,
+      "learning_rate": 9.153952843273232e-07,
+      "loss": 0.4893,
+      "step": 5592
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.695140838623047,
+      "learning_rate": 9.101941747572816e-07,
+      "loss": 0.5917,
+      "step": 5593
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.16102409362793,
+      "learning_rate": 9.0499306518724e-07,
+      "loss": 0.3906,
+      "step": 5594
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.148383140563965,
+      "learning_rate": 8.997919556171984e-07,
+      "loss": 0.581,
+      "step": 5595
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.057964324951172,
+      "learning_rate": 8.945908460471568e-07,
+      "loss": 0.5219,
+      "step": 5596
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.453672409057617,
+      "learning_rate": 8.893897364771152e-07,
+      "loss": 0.6452,
+      "step": 5597
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.834772109985352,
+      "learning_rate": 8.841886269070736e-07,
+      "loss": 0.6453,
+      "step": 5598
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.374686241149902,
+      "learning_rate": 8.78987517337032e-07,
+      "loss": 0.4173,
+      "step": 5599
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.548800468444824,
+      "learning_rate": 8.737864077669904e-07,
+      "loss": 0.5963,
+      "step": 5600
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.246313095092773,
+      "learning_rate": 8.685852981969488e-07,
+      "loss": 0.4767,
+      "step": 5601
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.919556617736816,
+      "learning_rate": 8.633841886269072e-07,
+      "loss": 0.6367,
+      "step": 5602
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.53325080871582,
+      "learning_rate": 8.581830790568655e-07,
+      "loss": 0.4979,
+      "step": 5603
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.755916595458984,
+      "learning_rate": 8.529819694868239e-07,
+      "loss": 0.541,
+      "step": 5604
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.111021995544434,
+      "learning_rate": 8.477808599167823e-07,
+      "loss": 0.4624,
+      "step": 5605
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.863731384277344,
+      "learning_rate": 8.425797503467407e-07,
+      "loss": 0.5083,
+      "step": 5606
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.953649520874023,
+      "learning_rate": 8.373786407766991e-07,
+      "loss": 0.461,
+      "step": 5607
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.685579299926758,
+      "learning_rate": 8.321775312066575e-07,
+      "loss": 0.7432,
+      "step": 5608
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.866486549377441,
+      "learning_rate": 8.269764216366159e-07,
+      "loss": 0.6409,
+      "step": 5609
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.629308700561523,
+      "learning_rate": 8.217753120665743e-07,
+      "loss": 0.483,
+      "step": 5610
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.344728469848633,
+      "learning_rate": 8.165742024965326e-07,
+      "loss": 0.4463,
+      "step": 5611
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.973027229309082,
+      "learning_rate": 8.11373092926491e-07,
+      "loss": 0.561,
+      "step": 5612
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.125194549560547,
+      "learning_rate": 8.061719833564494e-07,
+      "loss": 0.3739,
+      "step": 5613
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.540558815002441,
+      "learning_rate": 8.009708737864078e-07,
+      "loss": 0.4562,
+      "step": 5614
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.961179733276367,
+      "learning_rate": 7.957697642163662e-07,
+      "loss": 0.5548,
+      "step": 5615
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.724641799926758,
+      "learning_rate": 7.905686546463246e-07,
+      "loss": 0.3159,
+      "step": 5616
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.348427772521973,
+      "learning_rate": 7.853675450762829e-07,
+      "loss": 0.5735,
+      "step": 5617
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.630414962768555,
+      "learning_rate": 7.801664355062413e-07,
+      "loss": 0.6961,
+      "step": 5618
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.034157752990723,
+      "learning_rate": 7.749653259361997e-07,
+      "loss": 0.528,
+      "step": 5619
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.79368782043457,
+      "learning_rate": 7.697642163661581e-07,
+      "loss": 0.4369,
+      "step": 5620
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.048106670379639,
+      "learning_rate": 7.645631067961165e-07,
+      "loss": 0.4735,
+      "step": 5621
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.681997299194336,
+      "learning_rate": 7.593619972260749e-07,
+      "loss": 0.5336,
+      "step": 5622
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.878608703613281,
+      "learning_rate": 7.541608876560333e-07,
+      "loss": 0.5271,
+      "step": 5623
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.74239444732666,
+      "learning_rate": 7.489597780859917e-07,
+      "loss": 0.6309,
+      "step": 5624
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.905543327331543,
+      "learning_rate": 7.437586685159501e-07,
+      "loss": 0.4772,
+      "step": 5625
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.030902862548828,
+      "learning_rate": 7.385575589459085e-07,
+      "loss": 0.4588,
+      "step": 5626
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.67436408996582,
+      "learning_rate": 7.333564493758669e-07,
+      "loss": 0.5033,
+      "step": 5627
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.108711242675781,
+      "learning_rate": 7.281553398058253e-07,
+      "loss": 0.6034,
+      "step": 5628
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.056013107299805,
+      "learning_rate": 7.229542302357837e-07,
+      "loss": 0.6467,
+      "step": 5629
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.818899154663086,
+      "learning_rate": 7.177531206657421e-07,
+      "loss": 0.5473,
+      "step": 5630
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.366634368896484,
+      "learning_rate": 7.125520110957005e-07,
+      "loss": 0.3513,
+      "step": 5631
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.752056121826172,
+      "learning_rate": 7.073509015256588e-07,
+      "loss": 0.5911,
+      "step": 5632
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.267977714538574,
+      "learning_rate": 7.021497919556172e-07,
+      "loss": 0.6184,
+      "step": 5633
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.862519264221191,
+      "learning_rate": 6.969486823855756e-07,
+      "loss": 0.7123,
+      "step": 5634
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.289929389953613,
+      "learning_rate": 6.91747572815534e-07,
+      "loss": 0.4615,
+      "step": 5635
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.672406196594238,
+      "learning_rate": 6.865464632454924e-07,
+      "loss": 0.4313,
+      "step": 5636
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.213133811950684,
+      "learning_rate": 6.813453536754508e-07,
+      "loss": 0.724,
+      "step": 5637
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.158554077148438,
+      "learning_rate": 6.761442441054092e-07,
+      "loss": 0.6788,
+      "step": 5638
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.780182838439941,
+      "learning_rate": 6.709431345353676e-07,
+      "loss": 0.5515,
+      "step": 5639
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.613974571228027,
+      "learning_rate": 6.65742024965326e-07,
+      "loss": 0.3769,
+      "step": 5640
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.779675483703613,
+      "learning_rate": 6.605409153952844e-07,
+      "loss": 0.5289,
+      "step": 5641
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.533519744873047,
+      "learning_rate": 6.553398058252428e-07,
+      "loss": 0.4767,
+      "step": 5642
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.109447479248047,
+      "learning_rate": 6.501386962552012e-07,
+      "loss": 0.5836,
+      "step": 5643
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.126509666442871,
+      "learning_rate": 6.449375866851596e-07,
+      "loss": 0.3947,
+      "step": 5644
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.286943435668945,
+      "learning_rate": 6.39736477115118e-07,
+      "loss": 0.4212,
+      "step": 5645
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.421225547790527,
+      "learning_rate": 6.345353675450763e-07,
+      "loss": 0.5666,
+      "step": 5646
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.276981353759766,
+      "learning_rate": 6.293342579750347e-07,
+      "loss": 0.5702,
+      "step": 5647
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.017122268676758,
+      "learning_rate": 6.241331484049931e-07,
+      "loss": 0.56,
+      "step": 5648
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.8422698974609375,
+      "learning_rate": 6.189320388349515e-07,
+      "loss": 0.5816,
+      "step": 5649
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.58733081817627,
+      "learning_rate": 6.137309292649099e-07,
+      "loss": 0.4333,
+      "step": 5650
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.465547561645508,
+      "learning_rate": 6.085298196948683e-07,
+      "loss": 0.5604,
+      "step": 5651
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.633495330810547,
+      "learning_rate": 6.033287101248267e-07,
+      "loss": 0.4984,
+      "step": 5652
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.377093315124512,
+      "learning_rate": 5.981276005547851e-07,
+      "loss": 0.414,
+      "step": 5653
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.952862739562988,
+      "learning_rate": 5.929264909847435e-07,
+      "loss": 0.8115,
+      "step": 5654
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.213726043701172,
+      "learning_rate": 5.877253814147019e-07,
+      "loss": 0.5527,
+      "step": 5655
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.58105754852295,
+      "learning_rate": 5.825242718446602e-07,
+      "loss": 0.5525,
+      "step": 5656
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.233848571777344,
+      "learning_rate": 5.773231622746186e-07,
+      "loss": 0.5136,
+      "step": 5657
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.963421821594238,
+      "learning_rate": 5.72122052704577e-07,
+      "loss": 0.6301,
+      "step": 5658
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.292157173156738,
+      "learning_rate": 5.669209431345354e-07,
+      "loss": 0.4286,
+      "step": 5659
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.835273265838623,
+      "learning_rate": 5.617198335644937e-07,
+      "loss": 0.4256,
+      "step": 5660
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.338135719299316,
+      "learning_rate": 5.565187239944521e-07,
+      "loss": 0.6159,
+      "step": 5661
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.699509620666504,
+      "learning_rate": 5.513176144244105e-07,
+      "loss": 0.3777,
+      "step": 5662
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.880874156951904,
+      "learning_rate": 5.461165048543689e-07,
+      "loss": 0.3574,
+      "step": 5663
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.73765754699707,
+      "learning_rate": 5.409153952843273e-07,
+      "loss": 0.5709,
+      "step": 5664
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.43372917175293,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 0.5992,
+      "step": 5665
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.224180221557617,
+      "learning_rate": 5.305131761442441e-07,
+      "loss": 0.5377,
+      "step": 5666
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.165611267089844,
+      "learning_rate": 5.253120665742025e-07,
+      "loss": 0.3228,
+      "step": 5667
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.73080062866211,
+      "learning_rate": 5.201109570041609e-07,
+      "loss": 0.6311,
+      "step": 5668
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.552660942077637,
+      "learning_rate": 5.149098474341193e-07,
+      "loss": 0.4763,
+      "step": 5669
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.72230339050293,
+      "learning_rate": 5.097087378640777e-07,
+      "loss": 0.526,
+      "step": 5670
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.743865966796875,
+      "learning_rate": 5.045076282940361e-07,
+      "loss": 0.6436,
+      "step": 5671
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.154030799865723,
+      "learning_rate": 4.993065187239945e-07,
+      "loss": 0.4995,
+      "step": 5672
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.758808135986328,
+      "learning_rate": 4.941054091539529e-07,
+      "loss": 0.5083,
+      "step": 5673
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.685373306274414,
+      "learning_rate": 4.889042995839113e-07,
+      "loss": 0.4573,
+      "step": 5674
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.418045520782471,
+      "learning_rate": 4.837031900138696e-07,
+      "loss": 0.5379,
+      "step": 5675
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.198553085327148,
+      "learning_rate": 4.78502080443828e-07,
+      "loss": 0.4124,
+      "step": 5676
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.299234390258789,
+      "learning_rate": 4.7330097087378644e-07,
+      "loss": 0.4169,
+      "step": 5677
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.042285919189453,
+      "learning_rate": 4.680998613037448e-07,
+      "loss": 0.5099,
+      "step": 5678
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.933381080627441,
+      "learning_rate": 4.628987517337032e-07,
+      "loss": 0.5361,
+      "step": 5679
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.765474319458008,
+      "learning_rate": 4.576976421636616e-07,
+      "loss": 0.4769,
+      "step": 5680
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.4788179397583,
+      "learning_rate": 4.5249653259362e-07,
+      "loss": 0.5894,
+      "step": 5681
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.123950004577637,
+      "learning_rate": 4.472954230235784e-07,
+      "loss": 0.5514,
+      "step": 5682
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.62724494934082,
+      "learning_rate": 4.420943134535368e-07,
+      "loss": 0.6507,
+      "step": 5683
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.088509559631348,
+      "learning_rate": 4.368932038834952e-07,
+      "loss": 0.5527,
+      "step": 5684
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.34859848022461,
+      "learning_rate": 4.316920943134536e-07,
+      "loss": 0.4747,
+      "step": 5685
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.374835968017578,
+      "learning_rate": 4.2649098474341194e-07,
+      "loss": 0.529,
+      "step": 5686
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.51578140258789,
+      "learning_rate": 4.2128987517337034e-07,
+      "loss": 0.552,
+      "step": 5687
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.586883544921875,
+      "learning_rate": 4.1608876560332874e-07,
+      "loss": 0.4958,
+      "step": 5688
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.111226081848145,
+      "learning_rate": 4.1088765603328714e-07,
+      "loss": 0.6991,
+      "step": 5689
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.335448265075684,
+      "learning_rate": 4.056865464632455e-07,
+      "loss": 0.6418,
+      "step": 5690
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.923771858215332,
+      "learning_rate": 4.004854368932039e-07,
+      "loss": 0.5901,
+      "step": 5691
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.324898719787598,
+      "learning_rate": 3.952843273231623e-07,
+      "loss": 0.5674,
+      "step": 5692
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.700109481811523,
+      "learning_rate": 3.9008321775312064e-07,
+      "loss": 0.4604,
+      "step": 5693
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.656610488891602,
+      "learning_rate": 3.8488210818307904e-07,
+      "loss": 0.5309,
+      "step": 5694
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.859576225280762,
+      "learning_rate": 3.7968099861303744e-07,
+      "loss": 0.6521,
+      "step": 5695
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.083701133728027,
+      "learning_rate": 3.7447988904299584e-07,
+      "loss": 0.5679,
+      "step": 5696
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.518235206604004,
+      "learning_rate": 3.6927877947295424e-07,
+      "loss": 0.7255,
+      "step": 5697
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.332111358642578,
+      "learning_rate": 3.6407766990291264e-07,
+      "loss": 0.4369,
+      "step": 5698
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.89719295501709,
+      "learning_rate": 3.5887656033287104e-07,
+      "loss": 0.5765,
+      "step": 5699
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.386354446411133,
+      "learning_rate": 3.536754507628294e-07,
+      "loss": 0.6346,
+      "step": 5700
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.430556297302246,
+      "learning_rate": 3.484743411927878e-07,
+      "loss": 0.3636,
+      "step": 5701
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.712185859680176,
+      "learning_rate": 3.432732316227462e-07,
+      "loss": 0.5126,
+      "step": 5702
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.742746353149414,
+      "learning_rate": 3.380721220527046e-07,
+      "loss": 0.5879,
+      "step": 5703
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.612070083618164,
+      "learning_rate": 3.32871012482663e-07,
+      "loss": 0.5313,
+      "step": 5704
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.783296585083008,
+      "learning_rate": 3.276699029126214e-07,
+      "loss": 0.5506,
+      "step": 5705
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.386141777038574,
+      "learning_rate": 3.224687933425798e-07,
+      "loss": 0.4507,
+      "step": 5706
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.561898231506348,
+      "learning_rate": 3.1726768377253814e-07,
+      "loss": 0.3804,
+      "step": 5707
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.625502586364746,
+      "learning_rate": 3.1206657420249654e-07,
+      "loss": 0.4883,
+      "step": 5708
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.941847801208496,
+      "learning_rate": 3.0686546463245494e-07,
+      "loss": 0.3631,
+      "step": 5709
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.511348724365234,
+      "learning_rate": 3.0166435506241334e-07,
+      "loss": 0.5085,
+      "step": 5710
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.256333351135254,
+      "learning_rate": 2.9646324549237174e-07,
+      "loss": 0.4653,
+      "step": 5711
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.338321685791016,
+      "learning_rate": 2.912621359223301e-07,
+      "loss": 0.5919,
+      "step": 5712
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.443796157836914,
+      "learning_rate": 2.860610263522885e-07,
+      "loss": 0.651,
+      "step": 5713
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.370104789733887,
+      "learning_rate": 2.8085991678224684e-07,
+      "loss": 0.5801,
+      "step": 5714
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.64720344543457,
+      "learning_rate": 2.7565880721220524e-07,
+      "loss": 0.4186,
+      "step": 5715
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.748358726501465,
+      "learning_rate": 2.7045769764216364e-07,
+      "loss": 0.6314,
+      "step": 5716
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.342188835144043,
+      "learning_rate": 2.6525658807212204e-07,
+      "loss": 0.4741,
+      "step": 5717
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.483677864074707,
+      "learning_rate": 2.6005547850208044e-07,
+      "loss": 0.4449,
+      "step": 5718
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.704583168029785,
+      "learning_rate": 2.5485436893203884e-07,
+      "loss": 0.6135,
+      "step": 5719
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.352420806884766,
+      "learning_rate": 2.4965325936199724e-07,
+      "loss": 0.4549,
+      "step": 5720
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.678380489349365,
+      "learning_rate": 2.4445214979195564e-07,
+      "loss": 0.3504,
+      "step": 5721
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.719414710998535,
+      "learning_rate": 2.39251040221914e-07,
+      "loss": 0.5509,
+      "step": 5722
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.784076690673828,
+      "learning_rate": 2.340499306518724e-07,
+      "loss": 0.4031,
+      "step": 5723
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.03869342803955,
+      "learning_rate": 2.288488210818308e-07,
+      "loss": 0.4131,
+      "step": 5724
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.50780963897705,
+      "learning_rate": 2.236477115117892e-07,
+      "loss": 0.5007,
+      "step": 5725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.133684158325195,
+      "learning_rate": 2.184466019417476e-07,
+      "loss": 0.6299,
+      "step": 5726
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.1018452644348145,
+      "learning_rate": 2.1324549237170597e-07,
+      "loss": 0.3628,
+      "step": 5727
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.558178901672363,
+      "learning_rate": 2.0804438280166437e-07,
+      "loss": 0.5339,
+      "step": 5728
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.508993148803711,
+      "learning_rate": 2.0284327323162274e-07,
+      "loss": 0.6367,
+      "step": 5729
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.489725112915039,
+      "learning_rate": 1.9764216366158114e-07,
+      "loss": 0.4715,
+      "step": 5730
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.89338493347168,
+      "learning_rate": 1.9244105409153952e-07,
+      "loss": 0.6887,
+      "step": 5731
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.253499031066895,
+      "learning_rate": 1.8723994452149792e-07,
+      "loss": 0.5763,
+      "step": 5732
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.563721656799316,
+      "learning_rate": 1.8203883495145632e-07,
+      "loss": 0.3259,
+      "step": 5733
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.705044269561768,
+      "learning_rate": 1.768377253814147e-07,
+      "loss": 0.5327,
+      "step": 5734
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.891453742980957,
+      "learning_rate": 1.716366158113731e-07,
+      "loss": 0.727,
+      "step": 5735
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.37764835357666,
+      "learning_rate": 1.664355062413315e-07,
+      "loss": 0.4131,
+      "step": 5736
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.240230560302734,
+      "learning_rate": 1.612343966712899e-07,
+      "loss": 0.4463,
+      "step": 5737
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.740094184875488,
+      "learning_rate": 1.5603328710124827e-07,
+      "loss": 0.4653,
+      "step": 5738
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.351477146148682,
+      "learning_rate": 1.5083217753120667e-07,
+      "loss": 0.3624,
+      "step": 5739
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.157804489135742,
+      "learning_rate": 1.4563106796116505e-07,
+      "loss": 0.5236,
+      "step": 5740
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.707240104675293,
+      "learning_rate": 1.4042995839112342e-07,
+      "loss": 0.607,
+      "step": 5741
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.948281288146973,
+      "learning_rate": 1.3522884882108182e-07,
+      "loss": 0.3578,
+      "step": 5742
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.812734603881836,
+      "learning_rate": 1.3002773925104022e-07,
+      "loss": 0.3867,
+      "step": 5743
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.943075180053711,
+      "learning_rate": 1.2482662968099862e-07,
+      "loss": 0.6969,
+      "step": 5744
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.617497444152832,
+      "learning_rate": 1.19625520110957e-07,
+      "loss": 0.6343,
+      "step": 5745
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.172540664672852,
+      "learning_rate": 1.144244105409154e-07,
+      "loss": 0.555,
+      "step": 5746
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.448555946350098,
+      "learning_rate": 1.092233009708738e-07,
+      "loss": 0.4293,
+      "step": 5747
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.999434471130371,
+      "learning_rate": 1.0402219140083218e-07,
+      "loss": 0.4669,
+      "step": 5748
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.45499324798584,
+      "learning_rate": 9.882108183079057e-08,
+      "loss": 0.6874,
+      "step": 5749
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.219456672668457,
+      "learning_rate": 9.361997226074896e-08,
+      "loss": 0.3337,
+      "step": 5750
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.545730590820312,
+      "learning_rate": 8.841886269070735e-08,
+      "loss": 0.4373,
+      "step": 5751
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.210844039916992,
+      "learning_rate": 8.321775312066575e-08,
+      "loss": 0.5288,
+      "step": 5752
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.801623344421387,
+      "learning_rate": 7.801664355062414e-08,
+      "loss": 0.6583,
+      "step": 5753
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.976869583129883,
+      "learning_rate": 7.281553398058252e-08,
+      "loss": 0.46,
+      "step": 5754
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.374468803405762,
+      "learning_rate": 6.761442441054091e-08,
+      "loss": 0.4428,
+      "step": 5755
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.280649185180664,
+      "learning_rate": 6.241331484049931e-08,
+      "loss": 0.7606,
+      "step": 5756
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.488844871520996,
+      "learning_rate": 5.72122052704577e-08,
+      "loss": 0.5356,
+      "step": 5757
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.790217399597168,
+      "learning_rate": 5.201109570041609e-08,
+      "loss": 0.4178,
+      "step": 5758
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.859482288360596,
+      "learning_rate": 4.680998613037448e-08,
+      "loss": 0.509,
+      "step": 5759
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.070880889892578,
+      "learning_rate": 4.1608876560332874e-08,
+      "loss": 0.4296,
+      "step": 5760
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.405878067016602,
+      "learning_rate": 3.640776699029126e-08,
+      "loss": 0.7175,
+      "step": 5761
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.788263320922852,
+      "learning_rate": 3.1206657420249655e-08,
+      "loss": 0.458,
+      "step": 5762
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.697994232177734,
+      "learning_rate": 2.6005547850208046e-08,
+      "loss": 0.4598,
+      "step": 5763
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.880910396575928,
+      "learning_rate": 2.0804438280166437e-08,
+      "loss": 0.5437,
+      "step": 5764
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.364768981933594,
+      "learning_rate": 1.5603328710124828e-08,
+      "loss": 0.557,
+      "step": 5765
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.511802673339844,
+      "learning_rate": 1.0402219140083218e-08,
+      "loss": 0.5084,
+      "step": 5766
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.449084281921387,
+      "learning_rate": 5.201109570041609e-09,
+      "loss": 0.4083,
+      "step": 5767
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.23604965209961,
+      "learning_rate": 0.0,
+      "loss": 0.5297,
+      "step": 5768
+    },
+    {
+      "epoch": 2.0,
+      "step": 5768,
+      "total_flos": 1.0849198117732762e+17,
+      "train_loss": 0.8587615417055631,
+      "train_runtime": 1778.2981,
+      "train_samples_per_second": 311.313,
+      "train_steps_per_second": 3.244
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5768,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 1.0849198117732762e+17,
+  "train_batch_size": 96,
+  "trial_name": null,
+  "trial_params": null
+}