diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9315 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.0,
+  "eval_steps": 500,
+  "global_step": 1326,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004524886877828055,
+      "grad_norm": 3.2510592937469482,
+      "learning_rate": 4.524886877828055e-08,
+      "loss": 0.8401,
+      "step": 1
+    },
+    {
+      "epoch": 0.00904977375565611,
+      "grad_norm": 2.890446424484253,
+      "learning_rate": 9.04977375565611e-08,
+      "loss": 0.7699,
+      "step": 2
+    },
+    {
+      "epoch": 0.013574660633484163,
+      "grad_norm": 3.4535956382751465,
+      "learning_rate": 1.3574660633484163e-07,
+      "loss": 0.8325,
+      "step": 3
+    },
+    {
+      "epoch": 0.01809954751131222,
+      "grad_norm": 2.992128610610962,
+      "learning_rate": 1.809954751131222e-07,
+      "loss": 0.7607,
+      "step": 4
+    },
+    {
+      "epoch": 0.02262443438914027,
+      "grad_norm": 2.929943561553955,
+      "learning_rate": 2.2624434389140273e-07,
+      "loss": 0.7356,
+      "step": 5
+    },
+    {
+      "epoch": 0.027149321266968326,
+      "grad_norm": 3.091803550720215,
+      "learning_rate": 2.7149321266968326e-07,
+      "loss": 0.7924,
+      "step": 6
+    },
+    {
+      "epoch": 0.03167420814479638,
+      "grad_norm": 2.760162591934204,
+      "learning_rate": 3.167420814479638e-07,
+      "loss": 0.7152,
+      "step": 7
+    },
+    {
+      "epoch": 0.03619909502262444,
+      "grad_norm": 2.924128770828247,
+      "learning_rate": 3.619909502262444e-07,
+      "loss": 0.7361,
+      "step": 8
+    },
+    {
+      "epoch": 0.04072398190045249,
+      "grad_norm": 3.2693638801574707,
+      "learning_rate": 4.072398190045249e-07,
+      "loss": 0.8098,
+      "step": 9
+    },
+    {
+      "epoch": 0.04524886877828054,
+      "grad_norm": 2.807117462158203,
+      "learning_rate": 4.5248868778280546e-07,
+      "loss": 0.7414,
+      "step": 10
+    },
+    {
+      "epoch": 0.049773755656108594,
+      "grad_norm": 2.837460994720459,
+      "learning_rate": 4.977375565610859e-07,
+      "loss": 0.7736,
+      "step": 11
+    },
+    {
+      "epoch": 0.05429864253393665,
+      "grad_norm": 2.843400239944458,
+      "learning_rate": 5.429864253393665e-07,
+      "loss": 0.7254,
+      "step": 12
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 2.7627649307250977,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.7314,
+      "step": 13
+    },
+    {
+      "epoch": 0.06334841628959276,
+      "grad_norm": 2.5128955841064453,
+      "learning_rate": 6.334841628959276e-07,
+      "loss": 0.6816,
+      "step": 14
+    },
+    {
+      "epoch": 0.06787330316742081,
+      "grad_norm": 2.2178218364715576,
+      "learning_rate": 6.787330316742082e-07,
+      "loss": 0.687,
+      "step": 15
+    },
+    {
+      "epoch": 0.07239819004524888,
+      "grad_norm": 2.223222017288208,
+      "learning_rate": 7.239819004524888e-07,
+      "loss": 0.6834,
+      "step": 16
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 2.177546501159668,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.7176,
+      "step": 17
+    },
+    {
+      "epoch": 0.08144796380090498,
+      "grad_norm": 2.0788660049438477,
+      "learning_rate": 8.144796380090498e-07,
+      "loss": 0.6928,
+      "step": 18
+    },
+    {
+      "epoch": 0.08597285067873303,
+      "grad_norm": 2.013267993927002,
+      "learning_rate": 8.597285067873304e-07,
+      "loss": 0.7033,
+      "step": 19
+    },
+    {
+      "epoch": 0.09049773755656108,
+      "grad_norm": 1.640211820602417,
+      "learning_rate": 9.049773755656109e-07,
+      "loss": 0.7006,
+      "step": 20
+    },
+    {
+      "epoch": 0.09502262443438914,
+      "grad_norm": 1.2833340167999268,
+      "learning_rate": 9.502262443438914e-07,
+      "loss": 0.6464,
+      "step": 21
+    },
+    {
+      "epoch": 0.09954751131221719,
+      "grad_norm": 1.2619421482086182,
+      "learning_rate": 9.954751131221719e-07,
+      "loss": 0.6643,
+      "step": 22
+    },
+    {
+      "epoch": 0.10407239819004525,
+      "grad_norm": 1.2324565649032593,
+      "learning_rate": 1.0407239819004527e-06,
+      "loss": 0.6724,
+      "step": 23
+    },
+    {
+      "epoch": 0.1085972850678733,
+      "grad_norm": 1.1131088733673096,
+      "learning_rate": 1.085972850678733e-06,
+      "loss": 0.6059,
+      "step": 24
+    },
+    {
+      "epoch": 0.11312217194570136,
+      "grad_norm": 1.1473792791366577,
+      "learning_rate": 1.1312217194570136e-06,
+      "loss": 0.6494,
+      "step": 25
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 1.1130750179290771,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.6693,
+      "step": 26
+    },
+    {
+      "epoch": 0.12217194570135746,
+      "grad_norm": 1.1199308633804321,
+      "learning_rate": 1.2217194570135748e-06,
+      "loss": 0.7353,
+      "step": 27
+    },
+    {
+      "epoch": 0.12669683257918551,
+      "grad_norm": 1.094309687614441,
+      "learning_rate": 1.2669683257918552e-06,
+      "loss": 0.6729,
+      "step": 28
+    },
+    {
+      "epoch": 0.13122171945701358,
+      "grad_norm": 1.2093147039413452,
+      "learning_rate": 1.312217194570136e-06,
+      "loss": 0.6549,
+      "step": 29
+    },
+    {
+      "epoch": 0.13574660633484162,
+      "grad_norm": 1.2095019817352295,
+      "learning_rate": 1.3574660633484164e-06,
+      "loss": 0.6317,
+      "step": 30
+    },
+    {
+      "epoch": 0.14027149321266968,
+      "grad_norm": 1.43827223777771,
+      "learning_rate": 1.402714932126697e-06,
+      "loss": 0.6496,
+      "step": 31
+    },
+    {
+      "epoch": 0.14479638009049775,
+      "grad_norm": 1.4549624919891357,
+      "learning_rate": 1.4479638009049775e-06,
+      "loss": 0.6697,
+      "step": 32
+    },
+    {
+      "epoch": 0.1493212669683258,
+      "grad_norm": 1.165290117263794,
+      "learning_rate": 1.493212669683258e-06,
+      "loss": 0.6249,
+      "step": 33
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 1.257883906364441,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.6564,
+      "step": 34
+    },
+    {
+      "epoch": 0.1583710407239819,
+      "grad_norm": 1.027738094329834,
+      "learning_rate": 1.583710407239819e-06,
+      "loss": 0.5368,
+      "step": 35
+    },
+    {
+      "epoch": 0.16289592760180996,
+      "grad_norm": 0.9741559624671936,
+      "learning_rate": 1.6289592760180997e-06,
+      "loss": 0.6242,
+      "step": 36
+    },
+    {
+      "epoch": 0.167420814479638,
+      "grad_norm": 0.8538191318511963,
+      "learning_rate": 1.67420814479638e-06,
+      "loss": 0.6142,
+      "step": 37
+    },
+    {
+      "epoch": 0.17194570135746606,
+      "grad_norm": 0.6769871115684509,
+      "learning_rate": 1.7194570135746609e-06,
+      "loss": 0.5496,
+      "step": 38
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 0.6637622117996216,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.5751,
+      "step": 39
+    },
+    {
+      "epoch": 0.18099547511312217,
+      "grad_norm": 0.6483653783798218,
+      "learning_rate": 1.8099547511312218e-06,
+      "loss": 0.6127,
+      "step": 40
+    },
+    {
+      "epoch": 0.18552036199095023,
+      "grad_norm": 0.7997961044311523,
+      "learning_rate": 1.8552036199095024e-06,
+      "loss": 0.576,
+      "step": 41
+    },
+    {
+      "epoch": 0.19004524886877827,
+      "grad_norm": 0.7297148704528809,
+      "learning_rate": 1.9004524886877828e-06,
+      "loss": 0.6254,
+      "step": 42
+    },
+    {
+      "epoch": 0.19457013574660634,
+      "grad_norm": 0.6087789535522461,
+      "learning_rate": 1.9457013574660634e-06,
+      "loss": 0.5344,
+      "step": 43
+    },
+    {
+      "epoch": 0.19909502262443438,
+      "grad_norm": 0.7219811677932739,
+      "learning_rate": 1.9909502262443437e-06,
+      "loss": 0.5776,
+      "step": 44
+    },
+    {
+      "epoch": 0.20361990950226244,
+      "grad_norm": 0.7508623003959656,
+      "learning_rate": 2.0361990950226245e-06,
+      "loss": 0.5435,
+      "step": 45
+    },
+    {
+      "epoch": 0.2081447963800905,
+      "grad_norm": 0.6041704416275024,
+      "learning_rate": 2.0814479638009053e-06,
+      "loss": 0.5094,
+      "step": 46
+    },
+    {
+      "epoch": 0.21266968325791855,
+      "grad_norm": 0.6673488020896912,
+      "learning_rate": 2.1266968325791857e-06,
+      "loss": 0.556,
+      "step": 47
+    },
+    {
+      "epoch": 0.2171945701357466,
+      "grad_norm": 0.6235842704772949,
+      "learning_rate": 2.171945701357466e-06,
+      "loss": 0.5766,
+      "step": 48
+    },
+    {
+      "epoch": 0.22171945701357465,
+      "grad_norm": 0.5541844367980957,
+      "learning_rate": 2.2171945701357465e-06,
+      "loss": 0.5422,
+      "step": 49
+    },
+    {
+      "epoch": 0.22624434389140272,
+      "grad_norm": 0.6511553525924683,
+      "learning_rate": 2.2624434389140273e-06,
+      "loss": 0.6076,
+      "step": 50
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.5737389922142029,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.5804,
+      "step": 51
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 0.45331427454948425,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.4903,
+      "step": 52
+    },
+    {
+      "epoch": 0.2398190045248869,
+      "grad_norm": 0.5451053977012634,
+      "learning_rate": 2.3981900452488693e-06,
+      "loss": 0.5716,
+      "step": 53
+    },
+    {
+      "epoch": 0.24434389140271492,
+      "grad_norm": 0.5204625725746155,
+      "learning_rate": 2.4434389140271496e-06,
+      "loss": 0.5129,
+      "step": 54
+    },
+    {
+      "epoch": 0.248868778280543,
+      "grad_norm": 0.4971032440662384,
+      "learning_rate": 2.48868778280543e-06,
+      "loss": 0.5389,
+      "step": 55
+    },
+    {
+      "epoch": 0.25339366515837103,
+      "grad_norm": 0.5272262096405029,
+      "learning_rate": 2.5339366515837104e-06,
+      "loss": 0.5235,
+      "step": 56
+    },
+    {
+      "epoch": 0.2579185520361991,
+      "grad_norm": 0.5696743130683899,
+      "learning_rate": 2.5791855203619916e-06,
+      "loss": 0.508,
+      "step": 57
+    },
+    {
+      "epoch": 0.26244343891402716,
+      "grad_norm": 0.5246968269348145,
+      "learning_rate": 2.624434389140272e-06,
+      "loss": 0.5625,
+      "step": 58
+    },
+    {
+      "epoch": 0.2669683257918552,
+      "grad_norm": 0.5081693530082703,
+      "learning_rate": 2.6696832579185524e-06,
+      "loss": 0.501,
+      "step": 59
+    },
+    {
+      "epoch": 0.27149321266968324,
+      "grad_norm": 0.48727846145629883,
+      "learning_rate": 2.7149321266968327e-06,
+      "loss": 0.5259,
+      "step": 60
+    },
+    {
+      "epoch": 0.27601809954751133,
+      "grad_norm": 0.4420638084411621,
+      "learning_rate": 2.7601809954751135e-06,
+      "loss": 0.5165,
+      "step": 61
+    },
+    {
+      "epoch": 0.28054298642533937,
+      "grad_norm": 0.4388761818408966,
+      "learning_rate": 2.805429864253394e-06,
+      "loss": 0.4902,
+      "step": 62
+    },
+    {
+      "epoch": 0.2850678733031674,
+      "grad_norm": 0.42472442984580994,
+      "learning_rate": 2.8506787330316743e-06,
+      "loss": 0.4561,
+      "step": 63
+    },
+    {
+      "epoch": 0.2895927601809955,
+      "grad_norm": 0.4837457239627838,
+      "learning_rate": 2.895927601809955e-06,
+      "loss": 0.5295,
+      "step": 64
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 0.4987254738807678,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.515,
+      "step": 65
+    },
+    {
+      "epoch": 0.2986425339366516,
+      "grad_norm": 0.4454256594181061,
+      "learning_rate": 2.986425339366516e-06,
+      "loss": 0.5238,
+      "step": 66
+    },
+    {
+      "epoch": 0.3031674208144796,
+      "grad_norm": 0.5037821531295776,
+      "learning_rate": 3.0316742081447962e-06,
+      "loss": 0.5527,
+      "step": 67
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.4538140296936035,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.4905,
+      "step": 68
+    },
+    {
+      "epoch": 0.31221719457013575,
+      "grad_norm": 0.5122717022895813,
+      "learning_rate": 3.122171945701358e-06,
+      "loss": 0.5493,
+      "step": 69
+    },
+    {
+      "epoch": 0.3167420814479638,
+      "grad_norm": 0.44596371054649353,
+      "learning_rate": 3.167420814479638e-06,
+      "loss": 0.5484,
+      "step": 70
+    },
+    {
+      "epoch": 0.3212669683257919,
+      "grad_norm": 0.4473898410797119,
+      "learning_rate": 3.212669683257919e-06,
+      "loss": 0.5218,
+      "step": 71
+    },
+    {
+      "epoch": 0.3257918552036199,
+      "grad_norm": 0.3825899064540863,
+      "learning_rate": 3.2579185520361994e-06,
+      "loss": 0.4883,
+      "step": 72
+    },
+    {
+      "epoch": 0.33031674208144796,
+      "grad_norm": 0.40941205620765686,
+      "learning_rate": 3.3031674208144797e-06,
+      "loss": 0.4127,
+      "step": 73
+    },
+    {
+      "epoch": 0.334841628959276,
+      "grad_norm": 0.42915788292884827,
+      "learning_rate": 3.34841628959276e-06,
+      "loss": 0.5194,
+      "step": 74
+    },
+    {
+      "epoch": 0.3393665158371041,
+      "grad_norm": 0.44509854912757874,
+      "learning_rate": 3.3936651583710413e-06,
+      "loss": 0.4664,
+      "step": 75
+    },
+    {
+      "epoch": 0.3438914027149321,
+      "grad_norm": 0.41393935680389404,
+      "learning_rate": 3.4389140271493217e-06,
+      "loss": 0.5213,
+      "step": 76
+    },
+    {
+      "epoch": 0.34841628959276016,
+      "grad_norm": 0.4840753674507141,
+      "learning_rate": 3.484162895927602e-06,
+      "loss": 0.5456,
+      "step": 77
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.39024800062179565,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.446,
+      "step": 78
+    },
+    {
+      "epoch": 0.3574660633484163,
+      "grad_norm": 0.4665883481502533,
+      "learning_rate": 3.5746606334841633e-06,
+      "loss": 0.504,
+      "step": 79
+    },
+    {
+      "epoch": 0.36199095022624433,
+      "grad_norm": 0.38244664669036865,
+      "learning_rate": 3.6199095022624436e-06,
+      "loss": 0.4656,
+      "step": 80
+    },
+    {
+      "epoch": 0.3665158371040724,
+      "grad_norm": 0.39441755414009094,
+      "learning_rate": 3.665158371040724e-06,
+      "loss": 0.4861,
+      "step": 81
+    },
+    {
+      "epoch": 0.37104072398190047,
+      "grad_norm": 0.41697946190834045,
+      "learning_rate": 3.710407239819005e-06,
+      "loss": 0.4678,
+      "step": 82
+    },
+    {
+      "epoch": 0.3755656108597285,
+      "grad_norm": 0.4145350456237793,
+      "learning_rate": 3.755656108597285e-06,
+      "loss": 0.4501,
+      "step": 83
+    },
+    {
+      "epoch": 0.38009049773755654,
+      "grad_norm": 0.42326846718788147,
+      "learning_rate": 3.8009049773755656e-06,
+      "loss": 0.4617,
+      "step": 84
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.44864463806152344,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.5503,
+      "step": 85
+    },
+    {
+      "epoch": 0.3891402714932127,
+      "grad_norm": 0.4511263072490692,
+      "learning_rate": 3.891402714932127e-06,
+      "loss": 0.4581,
+      "step": 86
+    },
+    {
+      "epoch": 0.3936651583710407,
+      "grad_norm": 0.5121440291404724,
+      "learning_rate": 3.9366515837104075e-06,
+      "loss": 0.4622,
+      "step": 87
+    },
+    {
+      "epoch": 0.39819004524886875,
+      "grad_norm": 0.45248016715049744,
+      "learning_rate": 3.9819004524886875e-06,
+      "loss": 0.5035,
+      "step": 88
+    },
+    {
+      "epoch": 0.40271493212669685,
+      "grad_norm": 0.4109788239002228,
+      "learning_rate": 4.027149321266969e-06,
+      "loss": 0.4743,
+      "step": 89
+    },
+    {
+      "epoch": 0.4072398190045249,
+      "grad_norm": 0.4233633577823639,
+      "learning_rate": 4.072398190045249e-06,
+      "loss": 0.4829,
+      "step": 90
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 0.47834792733192444,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.5588,
+      "step": 91
+    },
+    {
+      "epoch": 0.416289592760181,
+      "grad_norm": 0.4849421977996826,
+      "learning_rate": 4.162895927601811e-06,
+      "loss": 0.4738,
+      "step": 92
+    },
+    {
+      "epoch": 0.42081447963800905,
+      "grad_norm": 0.4980224370956421,
+      "learning_rate": 4.208144796380091e-06,
+      "loss": 0.5288,
+      "step": 93
+    },
+    {
+      "epoch": 0.4253393665158371,
+      "grad_norm": 0.42526569962501526,
+      "learning_rate": 4.2533936651583714e-06,
+      "loss": 0.4128,
+      "step": 94
+    },
+    {
+      "epoch": 0.4298642533936652,
+      "grad_norm": 0.483390748500824,
+      "learning_rate": 4.298642533936652e-06,
+      "loss": 0.4445,
+      "step": 95
+    },
+    {
+      "epoch": 0.4343891402714932,
+      "grad_norm": 0.4766150116920471,
+      "learning_rate": 4.343891402714932e-06,
+      "loss": 0.5001,
+      "step": 96
+    },
+    {
+      "epoch": 0.43891402714932126,
+      "grad_norm": 0.5314707159996033,
+      "learning_rate": 4.389140271493213e-06,
+      "loss": 0.4891,
+      "step": 97
+    },
+    {
+      "epoch": 0.4434389140271493,
+      "grad_norm": 0.44213107228279114,
+      "learning_rate": 4.434389140271493e-06,
+      "loss": 0.5129,
+      "step": 98
+    },
+    {
+      "epoch": 0.4479638009049774,
+      "grad_norm": 0.43551620841026306,
+      "learning_rate": 4.479638009049775e-06,
+      "loss": 0.4557,
+      "step": 99
+    },
+    {
+      "epoch": 0.45248868778280543,
+      "grad_norm": 0.4333869516849518,
+      "learning_rate": 4.5248868778280546e-06,
+      "loss": 0.4746,
+      "step": 100
+    },
+    {
+      "epoch": 0.45701357466063347,
+      "grad_norm": 0.42583024501800537,
+      "learning_rate": 4.570135746606335e-06,
+      "loss": 0.45,
+      "step": 101
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.48476096987724304,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.5087,
+      "step": 102
+    },
+    {
+      "epoch": 0.4660633484162896,
+      "grad_norm": 0.4503658413887024,
+      "learning_rate": 4.660633484162896e-06,
+      "loss": 0.4574,
+      "step": 103
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.43352094292640686,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.4494,
+      "step": 104
+    },
+    {
+      "epoch": 0.4751131221719457,
+      "grad_norm": 0.4933190643787384,
+      "learning_rate": 4.751131221719457e-06,
+      "loss": 0.4441,
+      "step": 105
+    },
+    {
+      "epoch": 0.4796380090497738,
+      "grad_norm": 0.4639478325843811,
+      "learning_rate": 4.7963800904977385e-06,
+      "loss": 0.4721,
+      "step": 106
+    },
+    {
+      "epoch": 0.4841628959276018,
+      "grad_norm": 0.4035486876964569,
+      "learning_rate": 4.8416289592760185e-06,
+      "loss": 0.4574,
+      "step": 107
+    },
+    {
+      "epoch": 0.48868778280542985,
+      "grad_norm": 0.47403860092163086,
+      "learning_rate": 4.886877828054299e-06,
+      "loss": 0.4516,
+      "step": 108
+    },
+    {
+      "epoch": 0.49321266968325794,
+      "grad_norm": 0.48216933012008667,
+      "learning_rate": 4.93212669683258e-06,
+      "loss": 0.4521,
+      "step": 109
+    },
+    {
+      "epoch": 0.497737556561086,
+      "grad_norm": 0.43759751319885254,
+      "learning_rate": 4.97737556561086e-06,
+      "loss": 0.4409,
+      "step": 110
+    },
+    {
+      "epoch": 0.502262443438914,
+      "grad_norm": 0.4080127775669098,
+      "learning_rate": 5.022624434389141e-06,
+      "loss": 0.4762,
+      "step": 111
+    },
+    {
+      "epoch": 0.5067873303167421,
+      "grad_norm": 0.5794436931610107,
+      "learning_rate": 5.067873303167421e-06,
+      "loss": 0.4954,
+      "step": 112
+    },
+    {
+      "epoch": 0.5113122171945701,
+      "grad_norm": 0.5085976123809814,
+      "learning_rate": 5.1131221719457016e-06,
+      "loss": 0.4635,
+      "step": 113
+    },
+    {
+      "epoch": 0.5158371040723982,
+      "grad_norm": 0.41183438897132874,
+      "learning_rate": 5.158371040723983e-06,
+      "loss": 0.46,
+      "step": 114
+    },
+    {
+      "epoch": 0.5203619909502263,
+      "grad_norm": 0.4061497747898102,
+      "learning_rate": 5.203619909502263e-06,
+      "loss": 0.4559,
+      "step": 115
+    },
+    {
+      "epoch": 0.5248868778280543,
+      "grad_norm": 0.4753948748111725,
+      "learning_rate": 5.248868778280544e-06,
+      "loss": 0.4398,
+      "step": 116
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 0.47879111766815186,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.4709,
+      "step": 117
+    },
+    {
+      "epoch": 0.5339366515837104,
+      "grad_norm": 0.4345899820327759,
+      "learning_rate": 5.339366515837105e-06,
+      "loss": 0.4576,
+      "step": 118
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.4419268071651459,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.4489,
+      "step": 119
+    },
+    {
+      "epoch": 0.5429864253393665,
+      "grad_norm": 0.46015602350234985,
+      "learning_rate": 5.4298642533936655e-06,
+      "loss": 0.4463,
+      "step": 120
+    },
+    {
+      "epoch": 0.5475113122171946,
+      "grad_norm": 0.4941313862800598,
+      "learning_rate": 5.475113122171946e-06,
+      "loss": 0.5041,
+      "step": 121
+    },
+    {
+      "epoch": 0.5520361990950227,
+      "grad_norm": 0.4165358245372772,
+      "learning_rate": 5.520361990950227e-06,
+      "loss": 0.4749,
+      "step": 122
+    },
+    {
+      "epoch": 0.5565610859728507,
+      "grad_norm": 0.4693758189678192,
+      "learning_rate": 5.565610859728508e-06,
+      "loss": 0.4536,
+      "step": 123
+    },
+    {
+      "epoch": 0.5610859728506787,
+      "grad_norm": 0.4350532293319702,
+      "learning_rate": 5.610859728506788e-06,
+      "loss": 0.5089,
+      "step": 124
+    },
+    {
+      "epoch": 0.5656108597285068,
+      "grad_norm": 0.45343929529190063,
+      "learning_rate": 5.656108597285069e-06,
+      "loss": 0.4014,
+      "step": 125
+    },
+    {
+      "epoch": 0.5701357466063348,
+      "grad_norm": 0.507248044013977,
+      "learning_rate": 5.7013574660633486e-06,
+      "loss": 0.4932,
+      "step": 126
+    },
+    {
+      "epoch": 0.5746606334841629,
+      "grad_norm": 0.48399487137794495,
+      "learning_rate": 5.746606334841629e-06,
+      "loss": 0.5221,
+      "step": 127
+    },
+    {
+      "epoch": 0.579185520361991,
+      "grad_norm": 0.39801326394081116,
+      "learning_rate": 5.79185520361991e-06,
+      "loss": 0.4325,
+      "step": 128
+    },
+    {
+      "epoch": 0.583710407239819,
+      "grad_norm": 0.4183308780193329,
+      "learning_rate": 5.837104072398191e-06,
+      "loss": 0.4293,
+      "step": 129
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.4360859990119934,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.4135,
+      "step": 130
+    },
+    {
+      "epoch": 0.5927601809954751,
+      "grad_norm": 0.4908457398414612,
+      "learning_rate": 5.927601809954752e-06,
+      "loss": 0.4765,
+      "step": 131
+    },
+    {
+      "epoch": 0.5972850678733032,
+      "grad_norm": 0.4215766489505768,
+      "learning_rate": 5.972850678733032e-06,
+      "loss": 0.4622,
+      "step": 132
+    },
+    {
+      "epoch": 0.6018099547511312,
+      "grad_norm": 0.4052114188671112,
+      "learning_rate": 6.0180995475113125e-06,
+      "loss": 0.4317,
+      "step": 133
+    },
+    {
+      "epoch": 0.6063348416289592,
+      "grad_norm": 0.46360260248184204,
+      "learning_rate": 6.0633484162895924e-06,
+      "loss": 0.4827,
+      "step": 134
+    },
+    {
+      "epoch": 0.6108597285067874,
+      "grad_norm": 0.40575292706489563,
+      "learning_rate": 6.108597285067874e-06,
+      "loss": 0.4281,
+      "step": 135
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5005738139152527,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.5005,
+      "step": 136
+    },
+    {
+      "epoch": 0.6199095022624435,
+      "grad_norm": 0.451043963432312,
+      "learning_rate": 6.199095022624435e-06,
+      "loss": 0.4006,
+      "step": 137
+    },
+    {
+      "epoch": 0.6244343891402715,
+      "grad_norm": 0.4629691541194916,
+      "learning_rate": 6.244343891402716e-06,
+      "loss": 0.4941,
+      "step": 138
+    },
+    {
+      "epoch": 0.6289592760180995,
+      "grad_norm": 0.4519931375980377,
+      "learning_rate": 6.2895927601809956e-06,
+      "loss": 0.4438,
+      "step": 139
+    },
+    {
+      "epoch": 0.6334841628959276,
+      "grad_norm": 0.4885016679763794,
+      "learning_rate": 6.334841628959276e-06,
+      "loss": 0.4642,
+      "step": 140
+    },
+    {
+      "epoch": 0.6380090497737556,
+      "grad_norm": 0.4142250418663025,
+      "learning_rate": 6.380090497737556e-06,
+      "loss": 0.4164,
+      "step": 141
+    },
+    {
+      "epoch": 0.6425339366515838,
+      "grad_norm": 0.4803358316421509,
+      "learning_rate": 6.425339366515838e-06,
+      "loss": 0.4564,
+      "step": 142
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 0.47766202688217163,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.4106,
+      "step": 143
+    },
+    {
+      "epoch": 0.6515837104072398,
+      "grad_norm": 0.449381947517395,
+      "learning_rate": 6.515837104072399e-06,
+      "loss": 0.4301,
+      "step": 144
+    },
+    {
+      "epoch": 0.6561085972850679,
+      "grad_norm": 0.3938814103603363,
+      "learning_rate": 6.5610859728506795e-06,
+      "loss": 0.4916,
+      "step": 145
+    },
+    {
+      "epoch": 0.6606334841628959,
+      "grad_norm": 0.39036545157432556,
+      "learning_rate": 6.6063348416289595e-06,
+      "loss": 0.4538,
+      "step": 146
+    },
+    {
+      "epoch": 0.665158371040724,
+      "grad_norm": 0.4129018485546112,
+      "learning_rate": 6.65158371040724e-06,
+      "loss": 0.4414,
+      "step": 147
+    },
+    {
+      "epoch": 0.669683257918552,
+      "grad_norm": 0.4925534725189209,
+      "learning_rate": 6.69683257918552e-06,
+      "loss": 0.4393,
+      "step": 148
+    },
+    {
+      "epoch": 0.6742081447963801,
+      "grad_norm": 0.4131447374820709,
+      "learning_rate": 6.742081447963802e-06,
+      "loss": 0.4944,
+      "step": 149
+    },
+    {
+      "epoch": 0.6787330316742082,
+      "grad_norm": 0.4236401319503784,
+      "learning_rate": 6.787330316742083e-06,
+      "loss": 0.4603,
+      "step": 150
+    },
+    {
+      "epoch": 0.6832579185520362,
+      "grad_norm": 0.4191535413265228,
+      "learning_rate": 6.832579185520363e-06,
+      "loss": 0.4195,
+      "step": 151
+    },
+    {
+      "epoch": 0.6877828054298643,
+      "grad_norm": 0.36520934104919434,
+      "learning_rate": 6.8778280542986434e-06,
+      "loss": 0.4602,
+      "step": 152
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.5145202875137329,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.508,
+      "step": 153
+    },
+    {
+      "epoch": 0.6968325791855203,
+      "grad_norm": 0.5549051761627197,
+      "learning_rate": 6.968325791855204e-06,
+      "loss": 0.4974,
+      "step": 154
+    },
+    {
+      "epoch": 0.7013574660633484,
+      "grad_norm": 0.42621299624443054,
+      "learning_rate": 7.013574660633484e-06,
+      "loss": 0.4783,
+      "step": 155
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.562082052230835,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.519,
+      "step": 156
+    },
+    {
+      "epoch": 0.7104072398190046,
+      "grad_norm": 0.4220001697540283,
+      "learning_rate": 7.104072398190046e-06,
+      "loss": 0.4909,
+      "step": 157
+    },
+    {
+      "epoch": 0.7149321266968326,
+      "grad_norm": 0.450371652841568,
+      "learning_rate": 7.1493212669683265e-06,
+      "loss": 0.441,
+      "step": 158
+    },
+    {
+      "epoch": 0.7194570135746606,
+      "grad_norm": 0.4640924036502838,
+      "learning_rate": 7.1945701357466065e-06,
+      "loss": 0.4315,
+      "step": 159
+    },
+    {
+      "epoch": 0.7239819004524887,
+      "grad_norm": 0.417269229888916,
+      "learning_rate": 7.239819004524887e-06,
+      "loss": 0.4071,
+      "step": 160
+    },
+    {
+      "epoch": 0.7285067873303167,
+      "grad_norm": 0.5066007971763611,
+      "learning_rate": 7.285067873303168e-06,
+      "loss": 0.4618,
+      "step": 161
+    },
+    {
+      "epoch": 0.7330316742081447,
+      "grad_norm": 0.412210613489151,
+      "learning_rate": 7.330316742081448e-06,
+      "loss": 0.4871,
+      "step": 162
+    },
+    {
+      "epoch": 0.7375565610859729,
+      "grad_norm": 0.41878026723861694,
+      "learning_rate": 7.37556561085973e-06,
+      "loss": 0.437,
+      "step": 163
+    },
+    {
+      "epoch": 0.7420814479638009,
+      "grad_norm": 0.48414674401283264,
+      "learning_rate": 7.42081447963801e-06,
+      "loss": 0.4682,
+      "step": 164
+    },
+    {
+      "epoch": 0.746606334841629,
+      "grad_norm": 0.45116230845451355,
+      "learning_rate": 7.4660633484162904e-06,
+      "loss": 0.4137,
+      "step": 165
+    },
+    {
+      "epoch": 0.751131221719457,
+      "grad_norm": 0.4211224615573883,
+      "learning_rate": 7.51131221719457e-06,
+      "loss": 0.4063,
+      "step": 166
+    },
+    {
+      "epoch": 0.755656108597285,
+      "grad_norm": 0.46173834800720215,
+      "learning_rate": 7.556561085972851e-06,
+      "loss": 0.4924,
+      "step": 167
+    },
+    {
+      "epoch": 0.7601809954751131,
+      "grad_norm": 0.4191412925720215,
+      "learning_rate": 7.601809954751131e-06,
+      "loss": 0.4572,
+      "step": 168
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 0.4505658745765686,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.4481,
+      "step": 169
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.5168994069099426,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.4449,
+      "step": 170
+    },
+    {
+      "epoch": 0.7737556561085973,
+      "grad_norm": 0.5589344501495361,
+      "learning_rate": 7.737556561085974e-06,
+      "loss": 0.4761,
+      "step": 171
+    },
+    {
+      "epoch": 0.7782805429864253,
+      "grad_norm": 0.4831865429878235,
+      "learning_rate": 7.782805429864253e-06,
+      "loss": 0.4805,
+      "step": 172
+    },
+    {
+      "epoch": 0.7828054298642534,
+      "grad_norm": 0.46204084157943726,
+      "learning_rate": 7.828054298642534e-06,
+      "loss": 0.4535,
+      "step": 173
+    },
+    {
+      "epoch": 0.7873303167420814,
+      "grad_norm": 0.44106197357177734,
+      "learning_rate": 7.873303167420815e-06,
+      "loss": 0.4072,
+      "step": 174
+    },
+    {
+      "epoch": 0.7918552036199095,
+      "grad_norm": 0.5106877088546753,
+      "learning_rate": 7.918552036199096e-06,
+      "loss": 0.4125,
+      "step": 175
+    },
+    {
+      "epoch": 0.7963800904977375,
+      "grad_norm": 0.5074198842048645,
+      "learning_rate": 7.963800904977375e-06,
+      "loss": 0.434,
+      "step": 176
+    },
+    {
+      "epoch": 0.8009049773755657,
+      "grad_norm": 0.40606701374053955,
+      "learning_rate": 8.009049773755657e-06,
+      "loss": 0.4624,
+      "step": 177
+    },
+    {
+      "epoch": 0.8054298642533937,
+      "grad_norm": 0.5007057189941406,
+      "learning_rate": 8.054298642533938e-06,
+      "loss": 0.4476,
+      "step": 178
+    },
+    {
+      "epoch": 0.8099547511312217,
+      "grad_norm": 0.5165268778800964,
+      "learning_rate": 8.099547511312217e-06,
+      "loss": 0.4746,
+      "step": 179
+    },
+    {
+      "epoch": 0.8144796380090498,
+      "grad_norm": 0.46993181109428406,
+      "learning_rate": 8.144796380090498e-06,
+      "loss": 0.4432,
+      "step": 180
+    },
+    {
+      "epoch": 0.8190045248868778,
+      "grad_norm": 0.5677555799484253,
+      "learning_rate": 8.190045248868779e-06,
+      "loss": 0.4422,
+      "step": 181
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 0.5643273591995239,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.5143,
+      "step": 182
+    },
+    {
+      "epoch": 0.8280542986425339,
+      "grad_norm": 0.5022807717323303,
+      "learning_rate": 8.280542986425339e-06,
+      "loss": 0.4544,
+      "step": 183
+    },
+    {
+      "epoch": 0.832579185520362,
+      "grad_norm": 0.49226853251457214,
+      "learning_rate": 8.325791855203621e-06,
+      "loss": 0.4637,
+      "step": 184
+    },
+    {
+      "epoch": 0.8371040723981901,
+      "grad_norm": 0.48023420572280884,
+      "learning_rate": 8.371040723981902e-06,
+      "loss": 0.4264,
+      "step": 185
+    },
+    {
+      "epoch": 0.8416289592760181,
+      "grad_norm": 0.5742875337600708,
+      "learning_rate": 8.416289592760181e-06,
+      "loss": 0.4761,
+      "step": 186
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.5855184197425842,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.4668,
+      "step": 187
+    },
+    {
+      "epoch": 0.8506787330316742,
+      "grad_norm": 0.5004088282585144,
+      "learning_rate": 8.506787330316743e-06,
+      "loss": 0.4354,
+      "step": 188
+    },
+    {
+      "epoch": 0.8552036199095022,
+      "grad_norm": 0.5111898183822632,
+      "learning_rate": 8.552036199095024e-06,
+      "loss": 0.4494,
+      "step": 189
+    },
+    {
+      "epoch": 0.8597285067873304,
+      "grad_norm": 0.5159589052200317,
+      "learning_rate": 8.597285067873304e-06,
+      "loss": 0.493,
+      "step": 190
+    },
+    {
+      "epoch": 0.8642533936651584,
+      "grad_norm": 0.5572559833526611,
+      "learning_rate": 8.642533936651585e-06,
+      "loss": 0.4845,
+      "step": 191
+    },
+    {
+      "epoch": 0.8687782805429864,
+      "grad_norm": 0.47993189096450806,
+      "learning_rate": 8.687782805429864e-06,
+      "loss": 0.4709,
+      "step": 192
+    },
+    {
+      "epoch": 0.8733031674208145,
+      "grad_norm": 0.4479634463787079,
+      "learning_rate": 8.733031674208145e-06,
+      "loss": 0.4192,
+      "step": 193
+    },
+    {
+      "epoch": 0.8778280542986425,
+      "grad_norm": 0.48463767766952515,
+      "learning_rate": 8.778280542986426e-06,
+      "loss": 0.4468,
+      "step": 194
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 0.4300975203514099,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.4546,
+      "step": 195
+    },
+    {
+      "epoch": 0.8868778280542986,
+      "grad_norm": 0.43377694487571716,
+      "learning_rate": 8.868778280542986e-06,
+      "loss": 0.4517,
+      "step": 196
+    },
+    {
+      "epoch": 0.8914027149321267,
+      "grad_norm": 0.4832284152507782,
+      "learning_rate": 8.914027149321268e-06,
+      "loss": 0.4296,
+      "step": 197
+    },
+    {
+      "epoch": 0.8959276018099548,
+      "grad_norm": 0.47211429476737976,
+      "learning_rate": 8.95927601809955e-06,
+      "loss": 0.4487,
+      "step": 198
+    },
+    {
+      "epoch": 0.9004524886877828,
+      "grad_norm": 0.4101850688457489,
+      "learning_rate": 9.004524886877828e-06,
+      "loss": 0.4392,
+      "step": 199
+    },
+    {
+      "epoch": 0.9049773755656109,
+      "grad_norm": 0.5208500623703003,
+      "learning_rate": 9.049773755656109e-06,
+      "loss": 0.4276,
+      "step": 200
+    },
+    {
+      "epoch": 0.9095022624434389,
+      "grad_norm": 0.49378854036331177,
+      "learning_rate": 9.09502262443439e-06,
+      "loss": 0.454,
+      "step": 201
+    },
+    {
+      "epoch": 0.9140271493212669,
+      "grad_norm": 0.4764745831489563,
+      "learning_rate": 9.14027149321267e-06,
+      "loss": 0.4176,
+      "step": 202
+    },
+    {
+      "epoch": 0.918552036199095,
+      "grad_norm": 0.45192357897758484,
+      "learning_rate": 9.18552036199095e-06,
+      "loss": 0.4762,
+      "step": 203
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.4330678880214691,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.4158,
+      "step": 204
+    },
+    {
+      "epoch": 0.9276018099547512,
+      "grad_norm": 0.5406427383422852,
+      "learning_rate": 9.276018099547513e-06,
+      "loss": 0.4351,
+      "step": 205
+    },
+    {
+      "epoch": 0.9321266968325792,
+      "grad_norm": 0.4686480760574341,
+      "learning_rate": 9.321266968325792e-06,
+      "loss": 0.4683,
+      "step": 206
+    },
+    {
+      "epoch": 0.9366515837104072,
+      "grad_norm": 0.430961549282074,
+      "learning_rate": 9.366515837104073e-06,
+      "loss": 0.4077,
+      "step": 207
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.4596535563468933,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.4357,
+      "step": 208
+    },
+    {
+      "epoch": 0.9457013574660633,
+      "grad_norm": 0.49825915694236755,
+      "learning_rate": 9.457013574660635e-06,
+      "loss": 0.4438,
+      "step": 209
+    },
+    {
+      "epoch": 0.9502262443438914,
+      "grad_norm": 0.4968724846839905,
+      "learning_rate": 9.502262443438914e-06,
+      "loss": 0.4391,
+      "step": 210
+    },
+    {
+      "epoch": 0.9547511312217195,
+      "grad_norm": 0.5227912664413452,
+      "learning_rate": 9.547511312217196e-06,
+      "loss": 0.4682,
+      "step": 211
+    },
+    {
+      "epoch": 0.9592760180995475,
+      "grad_norm": 0.593227744102478,
+      "learning_rate": 9.592760180995477e-06,
+      "loss": 0.4375,
+      "step": 212
+    },
+    {
+      "epoch": 0.9638009049773756,
+      "grad_norm": 0.46191665530204773,
+      "learning_rate": 9.638009049773756e-06,
+      "loss": 0.4469,
+      "step": 213
+    },
+    {
+      "epoch": 0.9683257918552036,
+      "grad_norm": 0.49439120292663574,
+      "learning_rate": 9.683257918552037e-06,
+      "loss": 0.4167,
+      "step": 214
+    },
+    {
+      "epoch": 0.9728506787330317,
+      "grad_norm": 0.5323225855827332,
+      "learning_rate": 9.728506787330318e-06,
+      "loss": 0.462,
+      "step": 215
+    },
+    {
+      "epoch": 0.9773755656108597,
+      "grad_norm": 0.40613311529159546,
+      "learning_rate": 9.773755656108599e-06,
+      "loss": 0.3551,
+      "step": 216
+    },
+    {
+      "epoch": 0.9819004524886877,
+      "grad_norm": 0.523820161819458,
+      "learning_rate": 9.819004524886878e-06,
+      "loss": 0.4303,
+      "step": 217
+    },
+    {
+      "epoch": 0.9864253393665159,
+      "grad_norm": 0.4367105960845947,
+      "learning_rate": 9.86425339366516e-06,
+      "loss": 0.4617,
+      "step": 218
+    },
+    {
+      "epoch": 0.9909502262443439,
+      "grad_norm": 0.474623441696167,
+      "learning_rate": 9.90950226244344e-06,
+      "loss": 0.4098,
+      "step": 219
+    },
+    {
+      "epoch": 0.995475113122172,
+      "grad_norm": 0.5020592212677002,
+      "learning_rate": 9.95475113122172e-06,
+      "loss": 0.3816,
+      "step": 220
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4186856150627136,
+      "learning_rate": 1e-05,
+      "loss": 0.4303,
+      "step": 221
+    },
+    {
+      "epoch": 1.004524886877828,
+      "grad_norm": 0.5000538229942322,
+      "learning_rate": 9.999993763081091e-06,
+      "loss": 0.4021,
+      "step": 222
+    },
+    {
+      "epoch": 1.009049773755656,
+      "grad_norm": 0.4233403205871582,
+      "learning_rate": 9.999975052339923e-06,
+      "loss": 0.3917,
+      "step": 223
+    },
+    {
+      "epoch": 1.0135746606334841,
+      "grad_norm": 0.44955193996429443,
+      "learning_rate": 9.999943867823174e-06,
+      "loss": 0.4112,
+      "step": 224
+    },
+    {
+      "epoch": 1.0180995475113122,
+      "grad_norm": 0.5060259699821472,
+      "learning_rate": 9.999900209608642e-06,
+      "loss": 0.3585,
+      "step": 225
+    },
+    {
+      "epoch": 1.0226244343891402,
+      "grad_norm": 0.46285244822502136,
+      "learning_rate": 9.999844077805245e-06,
+      "loss": 0.3972,
+      "step": 226
+    },
+    {
+      "epoch": 1.0271493212669682,
+      "grad_norm": 0.519055962562561,
+      "learning_rate": 9.999775472553019e-06,
+      "loss": 0.3439,
+      "step": 227
+    },
+    {
+      "epoch": 1.0316742081447963,
+      "grad_norm": 0.49448850750923157,
+      "learning_rate": 9.999694394023119e-06,
+      "loss": 0.3777,
+      "step": 228
+    },
+    {
+      "epoch": 1.0361990950226245,
+      "grad_norm": 0.502755343914032,
+      "learning_rate": 9.999600842417815e-06,
+      "loss": 0.3685,
+      "step": 229
+    },
+    {
+      "epoch": 1.0407239819004526,
+      "grad_norm": 0.5343412756919861,
+      "learning_rate": 9.999494817970498e-06,
+      "loss": 0.3402,
+      "step": 230
+    },
+    {
+      "epoch": 1.0452488687782806,
+      "grad_norm": 0.5237849354743958,
+      "learning_rate": 9.999376320945673e-06,
+      "loss": 0.3453,
+      "step": 231
+    },
+    {
+      "epoch": 1.0497737556561086,
+      "grad_norm": 0.5596418380737305,
+      "learning_rate": 9.999245351638964e-06,
+      "loss": 0.3945,
+      "step": 232
+    },
+    {
+      "epoch": 1.0542986425339367,
+      "grad_norm": 0.3989311456680298,
+      "learning_rate": 9.999101910377107e-06,
+      "loss": 0.3383,
+      "step": 233
+    },
+    {
+      "epoch": 1.0588235294117647,
+      "grad_norm": 0.4172278642654419,
+      "learning_rate": 9.998945997517957e-06,
+      "loss": 0.392,
+      "step": 234
+    },
+    {
+      "epoch": 1.0633484162895928,
+      "grad_norm": 0.5389004349708557,
+      "learning_rate": 9.998777613450478e-06,
+      "loss": 0.3883,
+      "step": 235
+    },
+    {
+      "epoch": 1.0678733031674208,
+      "grad_norm": 0.5382748246192932,
+      "learning_rate": 9.998596758594752e-06,
+      "loss": 0.3582,
+      "step": 236
+    },
+    {
+      "epoch": 1.0723981900452488,
+      "grad_norm": 0.5459226965904236,
+      "learning_rate": 9.998403433401969e-06,
+      "loss": 0.4137,
+      "step": 237
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.48526811599731445,
+      "learning_rate": 9.998197638354428e-06,
+      "loss": 0.3618,
+      "step": 238
+    },
+    {
+      "epoch": 1.081447963800905,
+      "grad_norm": 0.47047311067581177,
+      "learning_rate": 9.997979373965542e-06,
+      "loss": 0.3097,
+      "step": 239
+    },
+    {
+      "epoch": 1.085972850678733,
+      "grad_norm": 0.5708703398704529,
+      "learning_rate": 9.997748640779829e-06,
+      "loss": 0.4371,
+      "step": 240
+    },
+    {
+      "epoch": 1.090497737556561,
+      "grad_norm": 0.5245316028594971,
+      "learning_rate": 9.997505439372914e-06,
+      "loss": 0.3988,
+      "step": 241
+    },
+    {
+      "epoch": 1.0950226244343892,
+      "grad_norm": 0.5281980633735657,
+      "learning_rate": 9.997249770351531e-06,
+      "loss": 0.414,
+      "step": 242
+    },
+    {
+      "epoch": 1.0995475113122173,
+      "grad_norm": 0.6034480333328247,
+      "learning_rate": 9.99698163435351e-06,
+      "loss": 0.4268,
+      "step": 243
+    },
+    {
+      "epoch": 1.1040723981900453,
+      "grad_norm": 0.479973703622818,
+      "learning_rate": 9.996701032047795e-06,
+      "loss": 0.4261,
+      "step": 244
+    },
+    {
+      "epoch": 1.1085972850678734,
+      "grad_norm": 0.4137464165687561,
+      "learning_rate": 9.996407964134416e-06,
+      "loss": 0.369,
+      "step": 245
+    },
+    {
+      "epoch": 1.1131221719457014,
+      "grad_norm": 0.7938151359558105,
+      "learning_rate": 9.996102431344514e-06,
+      "loss": 0.4152,
+      "step": 246
+    },
+    {
+      "epoch": 1.1176470588235294,
+      "grad_norm": 0.5222673416137695,
+      "learning_rate": 9.99578443444032e-06,
+      "loss": 0.3956,
+      "step": 247
+    },
+    {
+      "epoch": 1.1221719457013575,
+      "grad_norm": 0.41782158613204956,
+      "learning_rate": 9.995453974215164e-06,
+      "loss": 0.3423,
+      "step": 248
+    },
+    {
+      "epoch": 1.1266968325791855,
+      "grad_norm": 0.6398916244506836,
+      "learning_rate": 9.995111051493468e-06,
+      "loss": 0.3975,
+      "step": 249
+    },
+    {
+      "epoch": 1.1312217194570136,
+      "grad_norm": 0.5231488943099976,
+      "learning_rate": 9.99475566713074e-06,
+      "loss": 0.4051,
+      "step": 250
+    },
+    {
+      "epoch": 1.1357466063348416,
+      "grad_norm": 0.48015347123146057,
+      "learning_rate": 9.994387822013586e-06,
+      "loss": 0.4183,
+      "step": 251
+    },
+    {
+      "epoch": 1.1402714932126696,
+      "grad_norm": 0.44875526428222656,
+      "learning_rate": 9.994007517059693e-06,
+      "loss": 0.37,
+      "step": 252
+    },
+    {
+      "epoch": 1.1447963800904977,
+      "grad_norm": 0.6783897876739502,
+      "learning_rate": 9.993614753217833e-06,
+      "loss": 0.4063,
+      "step": 253
+    },
+    {
+      "epoch": 1.1493212669683257,
+      "grad_norm": 0.5309958457946777,
+      "learning_rate": 9.99320953146786e-06,
+      "loss": 0.3896,
+      "step": 254
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.5052987933158875,
+      "learning_rate": 9.992791852820709e-06,
+      "loss": 0.4243,
+      "step": 255
+    },
+    {
+      "epoch": 1.1583710407239818,
+      "grad_norm": 0.47427037358283997,
+      "learning_rate": 9.992361718318393e-06,
+      "loss": 0.3601,
+      "step": 256
+    },
+    {
+      "epoch": 1.16289592760181,
+      "grad_norm": 0.4682779312133789,
+      "learning_rate": 9.991919129033994e-06,
+      "loss": 0.3564,
+      "step": 257
+    },
+    {
+      "epoch": 1.167420814479638,
+      "grad_norm": 0.5049699544906616,
+      "learning_rate": 9.991464086071669e-06,
+      "loss": 0.3362,
+      "step": 258
+    },
+    {
+      "epoch": 1.1719457013574661,
+      "grad_norm": 0.43032440543174744,
+      "learning_rate": 9.990996590566648e-06,
+      "loss": 0.3884,
+      "step": 259
+    },
+    {
+      "epoch": 1.1764705882352942,
+      "grad_norm": 0.48647770285606384,
+      "learning_rate": 9.990516643685222e-06,
+      "loss": 0.3797,
+      "step": 260
+    },
+    {
+      "epoch": 1.1809954751131222,
+      "grad_norm": 0.6087546944618225,
+      "learning_rate": 9.990024246624745e-06,
+      "loss": 0.4037,
+      "step": 261
+    },
+    {
+      "epoch": 1.1855203619909502,
+      "grad_norm": 0.4716806709766388,
+      "learning_rate": 9.989519400613636e-06,
+      "loss": 0.4145,
+      "step": 262
+    },
+    {
+      "epoch": 1.1900452488687783,
+      "grad_norm": 0.47347491979599,
+      "learning_rate": 9.989002106911368e-06,
+      "loss": 0.3781,
+      "step": 263
+    },
+    {
+      "epoch": 1.1945701357466063,
+      "grad_norm": 0.6511465907096863,
+      "learning_rate": 9.988472366808468e-06,
+      "loss": 0.4168,
+      "step": 264
+    },
+    {
+      "epoch": 1.1990950226244343,
+      "grad_norm": 0.4355635941028595,
+      "learning_rate": 9.987930181626515e-06,
+      "loss": 0.3703,
+      "step": 265
+    },
+    {
+      "epoch": 1.2036199095022624,
+      "grad_norm": 0.4431900382041931,
+      "learning_rate": 9.987375552718133e-06,
+      "loss": 0.4355,
+      "step": 266
+    },
+    {
+      "epoch": 1.2081447963800904,
+      "grad_norm": 0.4137621223926544,
+      "learning_rate": 9.986808481466994e-06,
+      "loss": 0.3837,
+      "step": 267
+    },
+    {
+      "epoch": 1.2126696832579185,
+      "grad_norm": 0.49528124928474426,
+      "learning_rate": 9.98622896928781e-06,
+      "loss": 0.3732,
+      "step": 268
+    },
+    {
+      "epoch": 1.2171945701357467,
+      "grad_norm": 0.45777904987335205,
+      "learning_rate": 9.985637017626326e-06,
+      "loss": 0.4144,
+      "step": 269
+    },
+    {
+      "epoch": 1.2217194570135748,
+      "grad_norm": 0.5112358331680298,
+      "learning_rate": 9.985032627959325e-06,
+      "loss": 0.4616,
+      "step": 270
+    },
+    {
+      "epoch": 1.2262443438914028,
+      "grad_norm": 0.5006369948387146,
+      "learning_rate": 9.984415801794622e-06,
+      "loss": 0.4317,
+      "step": 271
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.48134738206863403,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.3928,
+      "step": 272
+    },
+    {
+      "epoch": 1.2352941176470589,
+      "grad_norm": 0.4563564956188202,
+      "learning_rate": 9.983144846158472e-06,
+      "loss": 0.3907,
+      "step": 273
+    },
+    {
+      "epoch": 1.239819004524887,
+      "grad_norm": 0.43982940912246704,
+      "learning_rate": 9.982490719857766e-06,
+      "loss": 0.41,
+      "step": 274
+    },
+    {
+      "epoch": 1.244343891402715,
+      "grad_norm": 0.4124949872493744,
+      "learning_rate": 9.981824163400827e-06,
+      "loss": 0.3775,
+      "step": 275
+    },
+    {
+      "epoch": 1.248868778280543,
+      "grad_norm": 0.44366535544395447,
+      "learning_rate": 9.981145178450555e-06,
+      "loss": 0.3958,
+      "step": 276
+    },
+    {
+      "epoch": 1.253393665158371,
+      "grad_norm": 0.4540594220161438,
+      "learning_rate": 9.980453766700861e-06,
+      "loss": 0.4276,
+      "step": 277
+    },
+    {
+      "epoch": 1.257918552036199,
+      "grad_norm": 0.4591408967971802,
+      "learning_rate": 9.979749929876658e-06,
+      "loss": 0.3899,
+      "step": 278
+    },
+    {
+      "epoch": 1.262443438914027,
+      "grad_norm": 0.4709743857383728,
+      "learning_rate": 9.979033669733855e-06,
+      "loss": 0.4155,
+      "step": 279
+    },
+    {
+      "epoch": 1.2669683257918551,
+      "grad_norm": 0.5153703689575195,
+      "learning_rate": 9.978304988059352e-06,
+      "loss": 0.355,
+      "step": 280
+    },
+    {
+      "epoch": 1.2714932126696832,
+      "grad_norm": 0.5748324990272522,
+      "learning_rate": 9.977563886671043e-06,
+      "loss": 0.3903,
+      "step": 281
+    },
+    {
+      "epoch": 1.2760180995475112,
+      "grad_norm": 0.42208483815193176,
+      "learning_rate": 9.976810367417801e-06,
+      "loss": 0.3711,
+      "step": 282
+    },
+    {
+      "epoch": 1.2805429864253393,
+      "grad_norm": 0.4175834059715271,
+      "learning_rate": 9.976044432179486e-06,
+      "loss": 0.3714,
+      "step": 283
+    },
+    {
+      "epoch": 1.2850678733031673,
+      "grad_norm": 0.48363685607910156,
+      "learning_rate": 9.975266082866923e-06,
+      "loss": 0.359,
+      "step": 284
+    },
+    {
+      "epoch": 1.2895927601809956,
+      "grad_norm": 0.4426751732826233,
+      "learning_rate": 9.974475321421918e-06,
+      "loss": 0.3497,
+      "step": 285
+    },
+    {
+      "epoch": 1.2941176470588236,
+      "grad_norm": 0.49176037311553955,
+      "learning_rate": 9.973672149817232e-06,
+      "loss": 0.4077,
+      "step": 286
+    },
+    {
+      "epoch": 1.2986425339366516,
+      "grad_norm": 0.47568652033805847,
+      "learning_rate": 9.972856570056594e-06,
+      "loss": 0.3892,
+      "step": 287
+    },
+    {
+      "epoch": 1.3031674208144797,
+      "grad_norm": 0.484466552734375,
+      "learning_rate": 9.972028584174687e-06,
+      "loss": 0.4417,
+      "step": 288
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.41258934140205383,
+      "learning_rate": 9.971188194237141e-06,
+      "loss": 0.3803,
+      "step": 289
+    },
+    {
+      "epoch": 1.3122171945701357,
+      "grad_norm": 0.463830828666687,
+      "learning_rate": 9.970335402340534e-06,
+      "loss": 0.4389,
+      "step": 290
+    },
+    {
+      "epoch": 1.3167420814479638,
+      "grad_norm": 0.425300270318985,
+      "learning_rate": 9.969470210612384e-06,
+      "loss": 0.4108,
+      "step": 291
+    },
+    {
+      "epoch": 1.3212669683257918,
+      "grad_norm": 0.5346444249153137,
+      "learning_rate": 9.968592621211146e-06,
+      "loss": 0.3969,
+      "step": 292
+    },
+    {
+      "epoch": 1.3257918552036199,
+      "grad_norm": 0.4658094644546509,
+      "learning_rate": 9.967702636326195e-06,
+      "loss": 0.4434,
+      "step": 293
+    },
+    {
+      "epoch": 1.330316742081448,
+      "grad_norm": 0.48898160457611084,
+      "learning_rate": 9.966800258177842e-06,
+      "loss": 0.4164,
+      "step": 294
+    },
+    {
+      "epoch": 1.334841628959276,
+      "grad_norm": 0.5616970062255859,
+      "learning_rate": 9.96588548901731e-06,
+      "loss": 0.4472,
+      "step": 295
+    },
+    {
+      "epoch": 1.3393665158371042,
+      "grad_norm": 0.43943265080451965,
+      "learning_rate": 9.964958331126735e-06,
+      "loss": 0.4284,
+      "step": 296
+    },
+    {
+      "epoch": 1.3438914027149322,
+      "grad_norm": 0.5051960945129395,
+      "learning_rate": 9.964018786819158e-06,
+      "loss": 0.3999,
+      "step": 297
+    },
+    {
+      "epoch": 1.3484162895927603,
+      "grad_norm": 0.5062891840934753,
+      "learning_rate": 9.963066858438525e-06,
+      "loss": 0.4063,
+      "step": 298
+    },
+    {
+      "epoch": 1.3529411764705883,
+      "grad_norm": 0.5116050243377686,
+      "learning_rate": 9.96210254835968e-06,
+      "loss": 0.3752,
+      "step": 299
+    },
+    {
+      "epoch": 1.3574660633484164,
+      "grad_norm": 0.4431830048561096,
+      "learning_rate": 9.961125858988348e-06,
+      "loss": 0.3763,
+      "step": 300
+    },
+    {
+      "epoch": 1.3619909502262444,
+      "grad_norm": 0.4586993455886841,
+      "learning_rate": 9.96013679276114e-06,
+      "loss": 0.3983,
+      "step": 301
+    },
+    {
+      "epoch": 1.3665158371040724,
+      "grad_norm": 0.4054921567440033,
+      "learning_rate": 9.959135352145552e-06,
+      "loss": 0.3679,
+      "step": 302
+    },
+    {
+      "epoch": 1.3710407239819005,
+      "grad_norm": 0.4635688066482544,
+      "learning_rate": 9.958121539639945e-06,
+      "loss": 0.3835,
+      "step": 303
+    },
+    {
+      "epoch": 1.3755656108597285,
+      "grad_norm": 0.4249536395072937,
+      "learning_rate": 9.95709535777354e-06,
+      "loss": 0.3947,
+      "step": 304
+    },
+    {
+      "epoch": 1.3800904977375565,
+      "grad_norm": 0.43373027443885803,
+      "learning_rate": 9.956056809106426e-06,
+      "loss": 0.3603,
+      "step": 305
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.45955413579940796,
+      "learning_rate": 9.955005896229543e-06,
+      "loss": 0.3909,
+      "step": 306
+    },
+    {
+      "epoch": 1.3891402714932126,
+      "grad_norm": 0.44556981325149536,
+      "learning_rate": 9.95394262176467e-06,
+      "loss": 0.3668,
+      "step": 307
+    },
+    {
+      "epoch": 1.3936651583710407,
+      "grad_norm": 0.4461539685726166,
+      "learning_rate": 9.952866988364431e-06,
+      "loss": 0.3902,
+      "step": 308
+    },
+    {
+      "epoch": 1.3981900452488687,
+      "grad_norm": 0.4445544481277466,
+      "learning_rate": 9.951778998712282e-06,
+      "loss": 0.407,
+      "step": 309
+    },
+    {
+      "epoch": 1.4027149321266967,
+      "grad_norm": 0.40608200430870056,
+      "learning_rate": 9.950678655522505e-06,
+      "loss": 0.3472,
+      "step": 310
+    },
+    {
+      "epoch": 1.4072398190045248,
+      "grad_norm": 0.4478733241558075,
+      "learning_rate": 9.9495659615402e-06,
+      "loss": 0.4088,
+      "step": 311
+    },
+    {
+      "epoch": 1.4117647058823528,
+      "grad_norm": 0.5241243243217468,
+      "learning_rate": 9.948440919541277e-06,
+      "loss": 0.3937,
+      "step": 312
+    },
+    {
+      "epoch": 1.416289592760181,
+      "grad_norm": 0.44417575001716614,
+      "learning_rate": 9.947303532332457e-06,
+      "loss": 0.3508,
+      "step": 313
+    },
+    {
+      "epoch": 1.420814479638009,
+      "grad_norm": 0.46424350142478943,
+      "learning_rate": 9.946153802751257e-06,
+      "loss": 0.4189,
+      "step": 314
+    },
+    {
+      "epoch": 1.4253393665158371,
+      "grad_norm": 0.41135942935943604,
+      "learning_rate": 9.944991733665986e-06,
+      "loss": 0.3924,
+      "step": 315
+    },
+    {
+      "epoch": 1.4298642533936652,
+      "grad_norm": 0.4415293037891388,
+      "learning_rate": 9.943817327975732e-06,
+      "loss": 0.3657,
+      "step": 316
+    },
+    {
+      "epoch": 1.4343891402714932,
+      "grad_norm": 0.40911784768104553,
+      "learning_rate": 9.942630588610368e-06,
+      "loss": 0.383,
+      "step": 317
+    },
+    {
+      "epoch": 1.4389140271493213,
+      "grad_norm": 0.43340831995010376,
+      "learning_rate": 9.94143151853053e-06,
+      "loss": 0.3657,
+      "step": 318
+    },
+    {
+      "epoch": 1.4434389140271493,
+      "grad_norm": 0.46907198429107666,
+      "learning_rate": 9.940220120727624e-06,
+      "loss": 0.4239,
+      "step": 319
+    },
+    {
+      "epoch": 1.4479638009049773,
+      "grad_norm": 0.4872211813926697,
+      "learning_rate": 9.938996398223802e-06,
+      "loss": 0.4008,
+      "step": 320
+    },
+    {
+      "epoch": 1.4524886877828054,
+      "grad_norm": 0.46110352873802185,
+      "learning_rate": 9.937760354071966e-06,
+      "loss": 0.3792,
+      "step": 321
+    },
+    {
+      "epoch": 1.4570135746606334,
+      "grad_norm": 0.46933513879776,
+      "learning_rate": 9.936511991355764e-06,
+      "loss": 0.3626,
+      "step": 322
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.47610121965408325,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.4133,
+      "step": 323
+    },
+    {
+      "epoch": 1.4660633484162897,
+      "grad_norm": 0.4467121660709381,
+      "learning_rate": 9.933978322718472e-06,
+      "loss": 0.4247,
+      "step": 324
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 0.4571016728878021,
+      "learning_rate": 9.932693023118299e-06,
+      "loss": 0.4107,
+      "step": 325
+    },
+    {
+      "epoch": 1.4751131221719458,
+      "grad_norm": 0.533100962638855,
+      "learning_rate": 9.931395417595568e-06,
+      "loss": 0.4091,
+      "step": 326
+    },
+    {
+      "epoch": 1.4796380090497738,
+      "grad_norm": 0.4366355538368225,
+      "learning_rate": 9.930085509387509e-06,
+      "loss": 0.3467,
+      "step": 327
+    },
+    {
+      "epoch": 1.4841628959276019,
+      "grad_norm": 0.41149038076400757,
+      "learning_rate": 9.92876330176203e-06,
+      "loss": 0.387,
+      "step": 328
+    },
+    {
+      "epoch": 1.48868778280543,
+      "grad_norm": 0.4341404438018799,
+      "learning_rate": 9.927428798017738e-06,
+      "loss": 0.4285,
+      "step": 329
+    },
+    {
+      "epoch": 1.493212669683258,
+      "grad_norm": 0.46218249201774597,
+      "learning_rate": 9.926082001483909e-06,
+      "loss": 0.4224,
+      "step": 330
+    },
+    {
+      "epoch": 1.497737556561086,
+      "grad_norm": 0.42733314633369446,
+      "learning_rate": 9.924722915520484e-06,
+      "loss": 0.375,
+      "step": 331
+    },
+    {
+      "epoch": 1.502262443438914,
+      "grad_norm": 0.4854833781719208,
+      "learning_rate": 9.92335154351807e-06,
+      "loss": 0.4287,
+      "step": 332
+    },
+    {
+      "epoch": 1.506787330316742,
+      "grad_norm": 0.4013662040233612,
+      "learning_rate": 9.92196788889792e-06,
+      "loss": 0.3262,
+      "step": 333
+    },
+    {
+      "epoch": 1.51131221719457,
+      "grad_norm": 0.4636831283569336,
+      "learning_rate": 9.92057195511193e-06,
+      "loss": 0.4486,
+      "step": 334
+    },
+    {
+      "epoch": 1.5158371040723981,
+      "grad_norm": 0.3883086144924164,
+      "learning_rate": 9.919163745642633e-06,
+      "loss": 0.3536,
+      "step": 335
+    },
+    {
+      "epoch": 1.5203619909502262,
+      "grad_norm": 0.47283634543418884,
+      "learning_rate": 9.91774326400318e-06,
+      "loss": 0.3475,
+      "step": 336
+    },
+    {
+      "epoch": 1.5248868778280542,
+      "grad_norm": 0.5383743047714233,
+      "learning_rate": 9.916310513737345e-06,
+      "loss": 0.5013,
+      "step": 337
+    },
+    {
+      "epoch": 1.5294117647058822,
+      "grad_norm": 0.4297215938568115,
+      "learning_rate": 9.91486549841951e-06,
+      "loss": 0.3726,
+      "step": 338
+    },
+    {
+      "epoch": 1.5339366515837103,
+      "grad_norm": 0.5066801905632019,
+      "learning_rate": 9.913408221654647e-06,
+      "loss": 0.3823,
+      "step": 339
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.4163861870765686,
+      "learning_rate": 9.911938687078324e-06,
+      "loss": 0.3728,
+      "step": 340
+    },
+    {
+      "epoch": 1.5429864253393664,
+      "grad_norm": 0.46085265278816223,
+      "learning_rate": 9.91045689835669e-06,
+      "loss": 0.3822,
+      "step": 341
+    },
+    {
+      "epoch": 1.5475113122171946,
+      "grad_norm": 0.4126754105091095,
+      "learning_rate": 9.908962859186465e-06,
+      "loss": 0.3634,
+      "step": 342
+    },
+    {
+      "epoch": 1.5520361990950227,
+      "grad_norm": 0.4993842542171478,
+      "learning_rate": 9.907456573294924e-06,
+      "loss": 0.3835,
+      "step": 343
+    },
+    {
+      "epoch": 1.5565610859728507,
+      "grad_norm": 0.4884220063686371,
+      "learning_rate": 9.905938044439904e-06,
+      "loss": 0.436,
+      "step": 344
+    },
+    {
+      "epoch": 1.5610859728506787,
+      "grad_norm": 0.3982529044151306,
+      "learning_rate": 9.90440727640978e-06,
+      "loss": 0.389,
+      "step": 345
+    },
+    {
+      "epoch": 1.5656108597285068,
+      "grad_norm": 0.4887678921222687,
+      "learning_rate": 9.902864273023465e-06,
+      "loss": 0.344,
+      "step": 346
+    },
+    {
+      "epoch": 1.5701357466063348,
+      "grad_norm": 0.4793776869773865,
+      "learning_rate": 9.901309038130392e-06,
+      "loss": 0.4078,
+      "step": 347
+    },
+    {
+      "epoch": 1.5746606334841629,
+      "grad_norm": 0.44910818338394165,
+      "learning_rate": 9.89974157561051e-06,
+      "loss": 0.3894,
+      "step": 348
+    },
+    {
+      "epoch": 1.5791855203619911,
+      "grad_norm": 0.45025259256362915,
+      "learning_rate": 9.898161889374272e-06,
+      "loss": 0.4182,
+      "step": 349
+    },
+    {
+      "epoch": 1.5837104072398192,
+      "grad_norm": 0.45233941078186035,
+      "learning_rate": 9.896569983362632e-06,
+      "loss": 0.3846,
+      "step": 350
+    },
+    {
+      "epoch": 1.5882352941176472,
+      "grad_norm": 0.4646618664264679,
+      "learning_rate": 9.894965861547023e-06,
+      "loss": 0.364,
+      "step": 351
+    },
+    {
+      "epoch": 1.5927601809954752,
+      "grad_norm": 0.44090190529823303,
+      "learning_rate": 9.893349527929357e-06,
+      "loss": 0.3949,
+      "step": 352
+    },
+    {
+      "epoch": 1.5972850678733033,
+      "grad_norm": 0.5286484360694885,
+      "learning_rate": 9.891720986542011e-06,
+      "loss": 0.4311,
+      "step": 353
+    },
+    {
+      "epoch": 1.6018099547511313,
+      "grad_norm": 0.44429534673690796,
+      "learning_rate": 9.890080241447816e-06,
+      "loss": 0.3906,
+      "step": 354
+    },
+    {
+      "epoch": 1.6063348416289593,
+      "grad_norm": 0.5068367123603821,
+      "learning_rate": 9.888427296740054e-06,
+      "loss": 0.4052,
+      "step": 355
+    },
+    {
+      "epoch": 1.6108597285067874,
+      "grad_norm": 0.507785975933075,
+      "learning_rate": 9.886762156542428e-06,
+      "loss": 0.4063,
+      "step": 356
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.49715733528137207,
+      "learning_rate": 9.885084825009085e-06,
+      "loss": 0.3965,
+      "step": 357
+    },
+    {
+      "epoch": 1.6199095022624435,
+      "grad_norm": 0.4482433497905731,
+      "learning_rate": 9.883395306324575e-06,
+      "loss": 0.4066,
+      "step": 358
+    },
+    {
+      "epoch": 1.6244343891402715,
+      "grad_norm": 0.41335636377334595,
+      "learning_rate": 9.881693604703853e-06,
+      "loss": 0.3674,
+      "step": 359
+    },
+    {
+      "epoch": 1.6289592760180995,
+      "grad_norm": 0.5023950934410095,
+      "learning_rate": 9.87997972439227e-06,
+      "loss": 0.3444,
+      "step": 360
+    },
+    {
+      "epoch": 1.6334841628959276,
+      "grad_norm": 0.5373045802116394,
+      "learning_rate": 9.878253669665557e-06,
+      "loss": 0.3853,
+      "step": 361
+    },
+    {
+      "epoch": 1.6380090497737556,
+      "grad_norm": 0.4574486315250397,
+      "learning_rate": 9.876515444829822e-06,
+      "loss": 0.4254,
+      "step": 362
+    },
+    {
+      "epoch": 1.6425339366515836,
+      "grad_norm": 0.5090571045875549,
+      "learning_rate": 9.874765054221532e-06,
+      "loss": 0.391,
+      "step": 363
+    },
+    {
+      "epoch": 1.6470588235294117,
+      "grad_norm": 0.4528499245643616,
+      "learning_rate": 9.873002502207502e-06,
+      "loss": 0.4292,
+      "step": 364
+    },
+    {
+      "epoch": 1.6515837104072397,
+      "grad_norm": 0.4229438304901123,
+      "learning_rate": 9.871227793184893e-06,
+      "loss": 0.3829,
+      "step": 365
+    },
+    {
+      "epoch": 1.6561085972850678,
+      "grad_norm": 0.4995672404766083,
+      "learning_rate": 9.869440931581186e-06,
+      "loss": 0.3534,
+      "step": 366
+    },
+    {
+      "epoch": 1.6606334841628958,
+      "grad_norm": 0.527503490447998,
+      "learning_rate": 9.867641921854192e-06,
+      "loss": 0.4057,
+      "step": 367
+    },
+    {
+      "epoch": 1.6651583710407238,
+      "grad_norm": 0.4658501446247101,
+      "learning_rate": 9.865830768492019e-06,
+      "loss": 0.4088,
+      "step": 368
+    },
+    {
+      "epoch": 1.6696832579185519,
+      "grad_norm": 0.5441451668739319,
+      "learning_rate": 9.864007476013075e-06,
+      "loss": 0.4149,
+      "step": 369
+    },
+    {
+      "epoch": 1.6742081447963801,
+      "grad_norm": 0.48692771792411804,
+      "learning_rate": 9.862172048966048e-06,
+      "loss": 0.3818,
+      "step": 370
+    },
+    {
+      "epoch": 1.6787330316742082,
+      "grad_norm": 0.40074610710144043,
+      "learning_rate": 9.860324491929905e-06,
+      "loss": 0.3657,
+      "step": 371
+    },
+    {
+      "epoch": 1.6832579185520362,
+      "grad_norm": 0.4921663999557495,
+      "learning_rate": 9.858464809513867e-06,
+      "loss": 0.4022,
+      "step": 372
+    },
+    {
+      "epoch": 1.6877828054298643,
+      "grad_norm": 0.46872252225875854,
+      "learning_rate": 9.856593006357415e-06,
+      "loss": 0.4099,
+      "step": 373
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.5141183137893677,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.3961,
+      "step": 374
+    },
+    {
+      "epoch": 1.6968325791855203,
+      "grad_norm": 0.46619221568107605,
+      "learning_rate": 9.852813056532345e-06,
+      "loss": 0.3172,
+      "step": 375
+    },
+    {
+      "epoch": 1.7013574660633484,
+      "grad_norm": 0.5484828352928162,
+      "learning_rate": 9.85090491929382e-06,
+      "loss": 0.4182,
+      "step": 376
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 0.5394891500473022,
+      "learning_rate": 9.848984680175049e-06,
+      "loss": 0.3898,
+      "step": 377
+    },
+    {
+      "epoch": 1.7104072398190047,
+      "grad_norm": 0.48792046308517456,
+      "learning_rate": 9.84705234396658e-06,
+      "loss": 0.3373,
+      "step": 378
+    },
+    {
+      "epoch": 1.7149321266968327,
+      "grad_norm": 0.3987761437892914,
+      "learning_rate": 9.845107915489143e-06,
+      "loss": 0.334,
+      "step": 379
+    },
+    {
+      "epoch": 1.7194570135746607,
+      "grad_norm": 0.525193452835083,
+      "learning_rate": 9.843151399593636e-06,
+      "loss": 0.3718,
+      "step": 380
+    },
+    {
+      "epoch": 1.7239819004524888,
+      "grad_norm": 0.48990345001220703,
+      "learning_rate": 9.84118280116111e-06,
+      "loss": 0.3763,
+      "step": 381
+    },
+    {
+      "epoch": 1.7285067873303168,
+      "grad_norm": 0.4889325797557831,
+      "learning_rate": 9.839202125102761e-06,
+      "loss": 0.3785,
+      "step": 382
+    },
+    {
+      "epoch": 1.7330316742081449,
+      "grad_norm": 0.5154342651367188,
+      "learning_rate": 9.837209376359918e-06,
+      "loss": 0.3883,
+      "step": 383
+    },
+    {
+      "epoch": 1.737556561085973,
+      "grad_norm": 0.5224171280860901,
+      "learning_rate": 9.835204559904021e-06,
+      "loss": 0.3666,
+      "step": 384
+    },
+    {
+      "epoch": 1.742081447963801,
+      "grad_norm": 0.5012906193733215,
+      "learning_rate": 9.833187680736624e-06,
+      "loss": 0.4093,
+      "step": 385
+    },
+    {
+      "epoch": 1.746606334841629,
+      "grad_norm": 0.5155039429664612,
+      "learning_rate": 9.831158743889373e-06,
+      "loss": 0.3768,
+      "step": 386
+    },
+    {
+      "epoch": 1.751131221719457,
+      "grad_norm": 0.44377192854881287,
+      "learning_rate": 9.829117754423991e-06,
+      "loss": 0.4136,
+      "step": 387
+    },
+    {
+      "epoch": 1.755656108597285,
+      "grad_norm": 0.6181144714355469,
+      "learning_rate": 9.827064717432272e-06,
+      "loss": 0.4698,
+      "step": 388
+    },
+    {
+      "epoch": 1.760180995475113,
+      "grad_norm": 0.5261528491973877,
+      "learning_rate": 9.82499963803607e-06,
+      "loss": 0.3924,
+      "step": 389
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 0.43123725056648254,
+      "learning_rate": 9.822922521387277e-06,
+      "loss": 0.4207,
+      "step": 390
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.4998953342437744,
+      "learning_rate": 9.820833372667813e-06,
+      "loss": 0.3933,
+      "step": 391
+    },
+    {
+      "epoch": 1.7737556561085972,
+      "grad_norm": 0.5654891133308411,
+      "learning_rate": 9.81873219708962e-06,
+      "loss": 0.3732,
+      "step": 392
+    },
+    {
+      "epoch": 1.7782805429864252,
+      "grad_norm": 0.4899405539035797,
+      "learning_rate": 9.816618999894645e-06,
+      "loss": 0.3752,
+      "step": 393
+    },
+    {
+      "epoch": 1.7828054298642533,
+      "grad_norm": 0.5240674614906311,
+      "learning_rate": 9.81449378635482e-06,
+      "loss": 0.4596,
+      "step": 394
+    },
+    {
+      "epoch": 1.7873303167420813,
+      "grad_norm": 0.46099403500556946,
+      "learning_rate": 9.81235656177206e-06,
+      "loss": 0.3748,
+      "step": 395
+    },
+    {
+      "epoch": 1.7918552036199094,
+      "grad_norm": 0.46495547890663147,
+      "learning_rate": 9.810207331478247e-06,
+      "loss": 0.4091,
+      "step": 396
+    },
+    {
+      "epoch": 1.7963800904977374,
+      "grad_norm": 0.5413225293159485,
+      "learning_rate": 9.808046100835207e-06,
+      "loss": 0.3903,
+      "step": 397
+    },
+    {
+      "epoch": 1.8009049773755657,
+      "grad_norm": 0.5968965888023376,
+      "learning_rate": 9.80587287523471e-06,
+      "loss": 0.3925,
+      "step": 398
+    },
+    {
+      "epoch": 1.8054298642533937,
+      "grad_norm": 0.46063658595085144,
+      "learning_rate": 9.803687660098446e-06,
+      "loss": 0.4055,
+      "step": 399
+    },
+    {
+      "epoch": 1.8099547511312217,
+      "grad_norm": 0.4856792390346527,
+      "learning_rate": 9.801490460878023e-06,
+      "loss": 0.3594,
+      "step": 400
+    },
+    {
+      "epoch": 1.8144796380090498,
+      "grad_norm": 0.4093053638935089,
+      "learning_rate": 9.79928128305494e-06,
+      "loss": 0.353,
+      "step": 401
+    },
+    {
+      "epoch": 1.8190045248868778,
+      "grad_norm": 0.5286359786987305,
+      "learning_rate": 9.797060132140584e-06,
+      "loss": 0.3798,
+      "step": 402
+    },
+    {
+      "epoch": 1.8235294117647058,
+      "grad_norm": 0.49250075221061707,
+      "learning_rate": 9.794827013676206e-06,
+      "loss": 0.412,
+      "step": 403
+    },
+    {
+      "epoch": 1.8280542986425339,
+      "grad_norm": 0.43814513087272644,
+      "learning_rate": 9.792581933232924e-06,
+      "loss": 0.386,
+      "step": 404
+    },
+    {
+      "epoch": 1.8325791855203621,
+      "grad_norm": 0.4019566774368286,
+      "learning_rate": 9.790324896411684e-06,
+      "loss": 0.3668,
+      "step": 405
+    },
+    {
+      "epoch": 1.8371040723981902,
+      "grad_norm": 0.44960689544677734,
+      "learning_rate": 9.788055908843275e-06,
+      "loss": 0.3761,
+      "step": 406
+    },
+    {
+      "epoch": 1.8416289592760182,
+      "grad_norm": 0.40887507796287537,
+      "learning_rate": 9.78577497618829e-06,
+      "loss": 0.3824,
+      "step": 407
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.5183314085006714,
+      "learning_rate": 9.783482104137127e-06,
+      "loss": 0.3466,
+      "step": 408
+    },
+    {
+      "epoch": 1.8506787330316743,
+      "grad_norm": 0.4178057312965393,
+      "learning_rate": 9.781177298409969e-06,
+      "loss": 0.3428,
+      "step": 409
+    },
+    {
+      "epoch": 1.8552036199095023,
+      "grad_norm": 0.45295459032058716,
+      "learning_rate": 9.778860564756769e-06,
+      "loss": 0.3886,
+      "step": 410
+    },
+    {
+      "epoch": 1.8597285067873304,
+      "grad_norm": 0.5001879334449768,
+      "learning_rate": 9.776531908957241e-06,
+      "loss": 0.4219,
+      "step": 411
+    },
+    {
+      "epoch": 1.8642533936651584,
+      "grad_norm": 0.47920331358909607,
+      "learning_rate": 9.77419133682084e-06,
+      "loss": 0.4203,
+      "step": 412
+    },
+    {
+      "epoch": 1.8687782805429864,
+      "grad_norm": 0.48827502131462097,
+      "learning_rate": 9.771838854186748e-06,
+      "loss": 0.4053,
+      "step": 413
+    },
+    {
+      "epoch": 1.8733031674208145,
+      "grad_norm": 0.42105311155319214,
+      "learning_rate": 9.769474466923864e-06,
+      "loss": 0.3749,
+      "step": 414
+    },
+    {
+      "epoch": 1.8778280542986425,
+      "grad_norm": 0.41811907291412354,
+      "learning_rate": 9.76709818093078e-06,
+      "loss": 0.3489,
+      "step": 415
+    },
+    {
+      "epoch": 1.8823529411764706,
+      "grad_norm": 0.42248275876045227,
+      "learning_rate": 9.764710002135784e-06,
+      "loss": 0.4341,
+      "step": 416
+    },
+    {
+      "epoch": 1.8868778280542986,
+      "grad_norm": 0.4599759578704834,
+      "learning_rate": 9.762309936496824e-06,
+      "loss": 0.3931,
+      "step": 417
+    },
+    {
+      "epoch": 1.8914027149321266,
+      "grad_norm": 0.47709447145462036,
+      "learning_rate": 9.759897990001504e-06,
+      "loss": 0.3697,
+      "step": 418
+    },
+    {
+      "epoch": 1.8959276018099547,
+      "grad_norm": 0.44299864768981934,
+      "learning_rate": 9.757474168667072e-06,
+      "loss": 0.4235,
+      "step": 419
+    },
+    {
+      "epoch": 1.9004524886877827,
+      "grad_norm": 0.45147445797920227,
+      "learning_rate": 9.7550384785404e-06,
+      "loss": 0.3915,
+      "step": 420
+    },
+    {
+      "epoch": 1.9049773755656108,
+      "grad_norm": 0.4743190109729767,
+      "learning_rate": 9.752590925697964e-06,
+      "loss": 0.4003,
+      "step": 421
+    },
+    {
+      "epoch": 1.9095022624434388,
+      "grad_norm": 0.515183687210083,
+      "learning_rate": 9.750131516245844e-06,
+      "loss": 0.4147,
+      "step": 422
+    },
+    {
+      "epoch": 1.9140271493212668,
+      "grad_norm": 0.4393065273761749,
+      "learning_rate": 9.747660256319695e-06,
+      "loss": 0.4065,
+      "step": 423
+    },
+    {
+      "epoch": 1.9185520361990949,
+      "grad_norm": 0.4030460715293884,
+      "learning_rate": 9.745177152084733e-06,
+      "loss": 0.3824,
+      "step": 424
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.47524887323379517,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.4089,
+      "step": 425
+    },
+    {
+      "epoch": 1.9276018099547512,
+      "grad_norm": 0.4686073660850525,
+      "learning_rate": 9.74017543549698e-06,
+      "loss": 0.3935,
+      "step": 426
+    },
+    {
+      "epoch": 1.9321266968325792,
+      "grad_norm": 0.3981013894081116,
+      "learning_rate": 9.737656835622312e-06,
+      "loss": 0.3755,
+      "step": 427
+    },
+    {
+      "epoch": 1.9366515837104072,
+      "grad_norm": 0.46402475237846375,
+      "learning_rate": 9.73512641639504e-06,
+      "loss": 0.3673,
+      "step": 428
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 0.502984881401062,
+      "learning_rate": 9.732584184127973e-06,
+      "loss": 0.3981,
+      "step": 429
+    },
+    {
+      "epoch": 1.9457013574660633,
+      "grad_norm": 0.42585644125938416,
+      "learning_rate": 9.730030145163392e-06,
+      "loss": 0.4047,
+      "step": 430
+    },
+    {
+      "epoch": 1.9502262443438914,
+      "grad_norm": 0.39187467098236084,
+      "learning_rate": 9.72746430587303e-06,
+      "loss": 0.391,
+      "step": 431
+    },
+    {
+      "epoch": 1.9547511312217196,
+      "grad_norm": 0.3843924105167389,
+      "learning_rate": 9.724886672658056e-06,
+      "loss": 0.4038,
+      "step": 432
+    },
+    {
+      "epoch": 1.9592760180995477,
+      "grad_norm": 0.47524020075798035,
+      "learning_rate": 9.72229725194907e-06,
+      "loss": 0.4075,
+      "step": 433
+    },
+    {
+      "epoch": 1.9638009049773757,
+      "grad_norm": 0.46401557326316833,
+      "learning_rate": 9.719696050206072e-06,
+      "loss": 0.3564,
+      "step": 434
+    },
+    {
+      "epoch": 1.9683257918552037,
+      "grad_norm": 0.4504929482936859,
+      "learning_rate": 9.717083073918459e-06,
+      "loss": 0.401,
+      "step": 435
+    },
+    {
+      "epoch": 1.9728506787330318,
+      "grad_norm": 0.4272751212120056,
+      "learning_rate": 9.714458329604997e-06,
+      "loss": 0.4096,
+      "step": 436
+    },
+    {
+      "epoch": 1.9773755656108598,
+      "grad_norm": 0.4029974341392517,
+      "learning_rate": 9.711821823813812e-06,
+      "loss": 0.4362,
+      "step": 437
+    },
+    {
+      "epoch": 1.9819004524886878,
+      "grad_norm": 0.4730750322341919,
+      "learning_rate": 9.709173563122377e-06,
+      "loss": 0.3892,
+      "step": 438
+    },
+    {
+      "epoch": 1.9864253393665159,
+      "grad_norm": 0.4282892644405365,
+      "learning_rate": 9.706513554137482e-06,
+      "loss": 0.4198,
+      "step": 439
+    },
+    {
+      "epoch": 1.990950226244344,
+      "grad_norm": 0.4743562340736389,
+      "learning_rate": 9.703841803495234e-06,
+      "loss": 0.4141,
+      "step": 440
+    },
+    {
+      "epoch": 1.995475113122172,
+      "grad_norm": 0.4112869203090668,
+      "learning_rate": 9.701158317861031e-06,
+      "loss": 0.3836,
+      "step": 441
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.41886061429977417,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3317,
+      "step": 442
+    },
+    {
+      "epoch": 2.004524886877828,
+      "grad_norm": 0.5465108752250671,
+      "learning_rate": 9.695756168424703e-06,
+      "loss": 0.3391,
+      "step": 443
+    },
+    {
+      "epoch": 2.009049773755656,
+      "grad_norm": 0.4423621892929077,
+      "learning_rate": 9.693037518099689e-06,
+      "loss": 0.2949,
+      "step": 444
+    },
+    {
+      "epoch": 2.013574660633484,
+      "grad_norm": 0.43876439332962036,
+      "learning_rate": 9.690307159736899e-06,
+      "loss": 0.3114,
+      "step": 445
+    },
+    {
+      "epoch": 2.018099547511312,
+      "grad_norm": 0.5939359664916992,
+      "learning_rate": 9.68756510014794e-06,
+      "loss": 0.3299,
+      "step": 446
+    },
+    {
+      "epoch": 2.02262443438914,
+      "grad_norm": 0.5051030516624451,
+      "learning_rate": 9.684811346173617e-06,
+      "loss": 0.2708,
+      "step": 447
+    },
+    {
+      "epoch": 2.0271493212669682,
+      "grad_norm": 0.46658316254615784,
+      "learning_rate": 9.682045904683905e-06,
+      "loss": 0.2798,
+      "step": 448
+    },
+    {
+      "epoch": 2.0316742081447963,
+      "grad_norm": 0.5176125168800354,
+      "learning_rate": 9.67926878257794e-06,
+      "loss": 0.3084,
+      "step": 449
+    },
+    {
+      "epoch": 2.0361990950226243,
+      "grad_norm": 0.5656089186668396,
+      "learning_rate": 9.676479986783992e-06,
+      "loss": 0.3079,
+      "step": 450
+    },
+    {
+      "epoch": 2.0407239819004523,
+      "grad_norm": 0.47683894634246826,
+      "learning_rate": 9.67367952425946e-06,
+      "loss": 0.2702,
+      "step": 451
+    },
+    {
+      "epoch": 2.0452488687782804,
+      "grad_norm": 0.4187275767326355,
+      "learning_rate": 9.67086740199085e-06,
+      "loss": 0.2688,
+      "step": 452
+    },
+    {
+      "epoch": 2.0497737556561084,
+      "grad_norm": 0.5175010561943054,
+      "learning_rate": 9.668043626993748e-06,
+      "loss": 0.3128,
+      "step": 453
+    },
+    {
+      "epoch": 2.0542986425339365,
+      "grad_norm": 0.43686679005622864,
+      "learning_rate": 9.66520820631282e-06,
+      "loss": 0.2665,
+      "step": 454
+    },
+    {
+      "epoch": 2.0588235294117645,
+      "grad_norm": 0.39973190426826477,
+      "learning_rate": 9.66236114702178e-06,
+      "loss": 0.2743,
+      "step": 455
+    },
+    {
+      "epoch": 2.0633484162895925,
+      "grad_norm": 0.3550076186656952,
+      "learning_rate": 9.65950245622338e-06,
+      "loss": 0.2471,
+      "step": 456
+    },
+    {
+      "epoch": 2.067873303167421,
+      "grad_norm": 0.408397376537323,
+      "learning_rate": 9.65663214104939e-06,
+      "loss": 0.2556,
+      "step": 457
+    },
+    {
+      "epoch": 2.072398190045249,
+      "grad_norm": 0.4247797131538391,
+      "learning_rate": 9.653750208660577e-06,
+      "loss": 0.2887,
+      "step": 458
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.4115331172943115,
+      "learning_rate": 9.650856666246693e-06,
+      "loss": 0.285,
+      "step": 459
+    },
+    {
+      "epoch": 2.081447963800905,
+      "grad_norm": 0.40403059124946594,
+      "learning_rate": 9.647951521026453e-06,
+      "loss": 0.2477,
+      "step": 460
+    },
+    {
+      "epoch": 2.085972850678733,
+      "grad_norm": 0.42889419198036194,
+      "learning_rate": 9.645034780247521e-06,
+      "loss": 0.2836,
+      "step": 461
+    },
+    {
+      "epoch": 2.090497737556561,
+      "grad_norm": 0.39554914832115173,
+      "learning_rate": 9.642106451186489e-06,
+      "loss": 0.2984,
+      "step": 462
+    },
+    {
+      "epoch": 2.0950226244343892,
+      "grad_norm": 0.4248838722705841,
+      "learning_rate": 9.639166541148855e-06,
+      "loss": 0.2922,
+      "step": 463
+    },
+    {
+      "epoch": 2.0995475113122173,
+      "grad_norm": 0.39132872223854065,
+      "learning_rate": 9.636215057469009e-06,
+      "loss": 0.3063,
+      "step": 464
+    },
+    {
+      "epoch": 2.1040723981900453,
+      "grad_norm": 0.40757620334625244,
+      "learning_rate": 9.63325200751022e-06,
+      "loss": 0.2808,
+      "step": 465
+    },
+    {
+      "epoch": 2.1085972850678734,
+      "grad_norm": 0.4235598146915436,
+      "learning_rate": 9.630277398664605e-06,
+      "loss": 0.2934,
+      "step": 466
+    },
+    {
+      "epoch": 2.1131221719457014,
+      "grad_norm": 0.38248857855796814,
+      "learning_rate": 9.627291238353127e-06,
+      "loss": 0.2453,
+      "step": 467
+    },
+    {
+      "epoch": 2.1176470588235294,
+      "grad_norm": 0.4244171679019928,
+      "learning_rate": 9.62429353402556e-06,
+      "loss": 0.2371,
+      "step": 468
+    },
+    {
+      "epoch": 2.1221719457013575,
+      "grad_norm": 0.41888427734375,
+      "learning_rate": 9.621284293160478e-06,
+      "loss": 0.3136,
+      "step": 469
+    },
+    {
+      "epoch": 2.1266968325791855,
+      "grad_norm": 0.3960363566875458,
+      "learning_rate": 9.618263523265238e-06,
+      "loss": 0.2585,
+      "step": 470
+    },
+    {
+      "epoch": 2.1312217194570136,
+      "grad_norm": 0.4441540241241455,
+      "learning_rate": 9.61523123187596e-06,
+      "loss": 0.2777,
+      "step": 471
+    },
+    {
+      "epoch": 2.1357466063348416,
+      "grad_norm": 0.4612972140312195,
+      "learning_rate": 9.612187426557505e-06,
+      "loss": 0.3011,
+      "step": 472
+    },
+    {
+      "epoch": 2.1402714932126696,
+      "grad_norm": 0.49506399035453796,
+      "learning_rate": 9.609132114903458e-06,
+      "loss": 0.289,
+      "step": 473
+    },
+    {
+      "epoch": 2.1447963800904977,
+      "grad_norm": 0.45658251643180847,
+      "learning_rate": 9.606065304536116e-06,
+      "loss": 0.3009,
+      "step": 474
+    },
+    {
+      "epoch": 2.1493212669683257,
+      "grad_norm": 0.47934991121292114,
+      "learning_rate": 9.602987003106455e-06,
+      "loss": 0.2878,
+      "step": 475
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.3920869529247284,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.2379,
+      "step": 476
+    },
+    {
+      "epoch": 2.158371040723982,
+      "grad_norm": 0.46996328234672546,
+      "learning_rate": 9.596795957807412e-06,
+      "loss": 0.287,
+      "step": 477
+    },
+    {
+      "epoch": 2.16289592760181,
+      "grad_norm": 0.4109254777431488,
+      "learning_rate": 9.593683229383249e-06,
+      "loss": 0.2708,
+      "step": 478
+    },
+    {
+      "epoch": 2.167420814479638,
+      "grad_norm": 0.37143900990486145,
+      "learning_rate": 9.590559040787168e-06,
+      "loss": 0.2497,
+      "step": 479
+    },
+    {
+      "epoch": 2.171945701357466,
+      "grad_norm": 0.42445284128189087,
+      "learning_rate": 9.587423399813292e-06,
+      "loss": 0.2806,
+      "step": 480
+    },
+    {
+      "epoch": 2.176470588235294,
+      "grad_norm": 0.3920595645904541,
+      "learning_rate": 9.584276314284316e-06,
+      "loss": 0.265,
+      "step": 481
+    },
+    {
+      "epoch": 2.180995475113122,
+      "grad_norm": 0.46008458733558655,
+      "learning_rate": 9.581117792051487e-06,
+      "loss": 0.3138,
+      "step": 482
+    },
+    {
+      "epoch": 2.1855203619909505,
+      "grad_norm": 0.39173129200935364,
+      "learning_rate": 9.577947840994585e-06,
+      "loss": 0.2933,
+      "step": 483
+    },
+    {
+      "epoch": 2.1900452488687785,
+      "grad_norm": 0.37566521763801575,
+      "learning_rate": 9.574766469021901e-06,
+      "loss": 0.2465,
+      "step": 484
+    },
+    {
+      "epoch": 2.1945701357466065,
+      "grad_norm": 0.43578964471817017,
+      "learning_rate": 9.57157368407022e-06,
+      "loss": 0.2658,
+      "step": 485
+    },
+    {
+      "epoch": 2.1990950226244346,
+      "grad_norm": 0.4081728160381317,
+      "learning_rate": 9.568369494104793e-06,
+      "loss": 0.2881,
+      "step": 486
+    },
+    {
+      "epoch": 2.2036199095022626,
+      "grad_norm": 0.3810422718524933,
+      "learning_rate": 9.565153907119336e-06,
+      "loss": 0.2875,
+      "step": 487
+    },
+    {
+      "epoch": 2.2081447963800906,
+      "grad_norm": 0.631112277507782,
+      "learning_rate": 9.561926931135985e-06,
+      "loss": 0.3056,
+      "step": 488
+    },
+    {
+      "epoch": 2.2126696832579187,
+      "grad_norm": 0.4002344310283661,
+      "learning_rate": 9.5586885742053e-06,
+      "loss": 0.2806,
+      "step": 489
+    },
+    {
+      "epoch": 2.2171945701357467,
+      "grad_norm": 0.46340012550354004,
+      "learning_rate": 9.555438844406226e-06,
+      "loss": 0.2862,
+      "step": 490
+    },
+    {
+      "epoch": 2.2217194570135748,
+      "grad_norm": 0.4309289753437042,
+      "learning_rate": 9.552177749846083e-06,
+      "loss": 0.2899,
+      "step": 491
+    },
+    {
+      "epoch": 2.226244343891403,
+      "grad_norm": 0.4052945375442505,
+      "learning_rate": 9.548905298660547e-06,
+      "loss": 0.257,
+      "step": 492
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.445239782333374,
+      "learning_rate": 9.54562149901362e-06,
+      "loss": 0.2582,
+      "step": 493
+    },
+    {
+      "epoch": 2.235294117647059,
+      "grad_norm": 0.4149002730846405,
+      "learning_rate": 9.542326359097619e-06,
+      "loss": 0.2558,
+      "step": 494
+    },
+    {
+      "epoch": 2.239819004524887,
+      "grad_norm": 0.5213009715080261,
+      "learning_rate": 9.539019887133154e-06,
+      "loss": 0.3235,
+      "step": 495
+    },
+    {
+      "epoch": 2.244343891402715,
+      "grad_norm": 0.4313045144081116,
+      "learning_rate": 9.535702091369103e-06,
+      "loss": 0.308,
+      "step": 496
+    },
+    {
+      "epoch": 2.248868778280543,
+      "grad_norm": 0.4291783273220062,
+      "learning_rate": 9.532372980082598e-06,
+      "loss": 0.273,
+      "step": 497
+    },
+    {
+      "epoch": 2.253393665158371,
+      "grad_norm": 0.4001004099845886,
+      "learning_rate": 9.529032561578992e-06,
+      "loss": 0.282,
+      "step": 498
+    },
+    {
+      "epoch": 2.257918552036199,
+      "grad_norm": 0.46362540125846863,
+      "learning_rate": 9.525680844191858e-06,
+      "loss": 0.3265,
+      "step": 499
+    },
+    {
+      "epoch": 2.262443438914027,
+      "grad_norm": 0.43766117095947266,
+      "learning_rate": 9.522317836282949e-06,
+      "loss": 0.3059,
+      "step": 500
+    },
+    {
+      "epoch": 2.266968325791855,
+      "grad_norm": 0.4003988206386566,
+      "learning_rate": 9.518943546242189e-06,
+      "loss": 0.3059,
+      "step": 501
+    },
+    {
+      "epoch": 2.271493212669683,
+      "grad_norm": 0.3825046420097351,
+      "learning_rate": 9.515557982487647e-06,
+      "loss": 0.2901,
+      "step": 502
+    },
+    {
+      "epoch": 2.276018099547511,
+      "grad_norm": 0.4281090199947357,
+      "learning_rate": 9.512161153465518e-06,
+      "loss": 0.3094,
+      "step": 503
+    },
+    {
+      "epoch": 2.2805429864253393,
+      "grad_norm": 0.4918365478515625,
+      "learning_rate": 9.5087530676501e-06,
+      "loss": 0.3387,
+      "step": 504
+    },
+    {
+      "epoch": 2.2850678733031673,
+      "grad_norm": 0.38389474153518677,
+      "learning_rate": 9.505333733543777e-06,
+      "loss": 0.2812,
+      "step": 505
+    },
+    {
+      "epoch": 2.2895927601809953,
+      "grad_norm": 0.5115017294883728,
+      "learning_rate": 9.501903159676993e-06,
+      "loss": 0.2665,
+      "step": 506
+    },
+    {
+      "epoch": 2.2941176470588234,
+      "grad_norm": 0.385798841714859,
+      "learning_rate": 9.498461354608228e-06,
+      "loss": 0.2978,
+      "step": 507
+    },
+    {
+      "epoch": 2.2986425339366514,
+      "grad_norm": 0.4701876938343048,
+      "learning_rate": 9.495008326923989e-06,
+      "loss": 0.3222,
+      "step": 508
+    },
+    {
+      "epoch": 2.3031674208144794,
+      "grad_norm": 0.37049078941345215,
+      "learning_rate": 9.491544085238778e-06,
+      "loss": 0.2672,
+      "step": 509
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.4914953410625458,
+      "learning_rate": 9.488068638195072e-06,
+      "loss": 0.3117,
+      "step": 510
+    },
+    {
+      "epoch": 2.3122171945701355,
+      "grad_norm": 0.3973662853240967,
+      "learning_rate": 9.484581994463303e-06,
+      "loss": 0.2915,
+      "step": 511
+    },
+    {
+      "epoch": 2.3167420814479636,
+      "grad_norm": 0.3646693825721741,
+      "learning_rate": 9.481084162741835e-06,
+      "loss": 0.2562,
+      "step": 512
+    },
+    {
+      "epoch": 2.321266968325792,
+      "grad_norm": 0.3842124938964844,
+      "learning_rate": 9.477575151756951e-06,
+      "loss": 0.263,
+      "step": 513
+    },
+    {
+      "epoch": 2.32579185520362,
+      "grad_norm": 0.4494537115097046,
+      "learning_rate": 9.474054970262811e-06,
+      "loss": 0.3218,
+      "step": 514
+    },
+    {
+      "epoch": 2.330316742081448,
+      "grad_norm": 0.46450093388557434,
+      "learning_rate": 9.470523627041452e-06,
+      "loss": 0.2917,
+      "step": 515
+    },
+    {
+      "epoch": 2.334841628959276,
+      "grad_norm": 0.4324088394641876,
+      "learning_rate": 9.466981130902758e-06,
+      "loss": 0.2871,
+      "step": 516
+    },
+    {
+      "epoch": 2.339366515837104,
+      "grad_norm": 0.4018588066101074,
+      "learning_rate": 9.46342749068443e-06,
+      "loss": 0.2783,
+      "step": 517
+    },
+    {
+      "epoch": 2.3438914027149322,
+      "grad_norm": 0.39708182215690613,
+      "learning_rate": 9.459862715251973e-06,
+      "loss": 0.2681,
+      "step": 518
+    },
+    {
+      "epoch": 2.3484162895927603,
+      "grad_norm": 0.403104305267334,
+      "learning_rate": 9.456286813498677e-06,
+      "loss": 0.276,
+      "step": 519
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 0.365902304649353,
+      "learning_rate": 9.452699794345583e-06,
+      "loss": 0.2397,
+      "step": 520
+    },
+    {
+      "epoch": 2.3574660633484164,
+      "grad_norm": 0.39722907543182373,
+      "learning_rate": 9.44910166674147e-06,
+      "loss": 0.2581,
+      "step": 521
+    },
+    {
+      "epoch": 2.3619909502262444,
+      "grad_norm": 0.3249555230140686,
+      "learning_rate": 9.445492439662832e-06,
+      "loss": 0.2592,
+      "step": 522
+    },
+    {
+      "epoch": 2.3665158371040724,
+      "grad_norm": 0.36296772956848145,
+      "learning_rate": 9.441872122113848e-06,
+      "loss": 0.3029,
+      "step": 523
+    },
+    {
+      "epoch": 2.3710407239819005,
+      "grad_norm": 0.44991156458854675,
+      "learning_rate": 9.438240723126376e-06,
+      "loss": 0.2628,
+      "step": 524
+    },
+    {
+      "epoch": 2.3755656108597285,
+      "grad_norm": 0.37070232629776,
+      "learning_rate": 9.434598251759904e-06,
+      "loss": 0.3199,
+      "step": 525
+    },
+    {
+      "epoch": 2.3800904977375565,
+      "grad_norm": 0.45003408193588257,
+      "learning_rate": 9.430944717101556e-06,
+      "loss": 0.2697,
+      "step": 526
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.49059706926345825,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.3067,
+      "step": 527
+    },
+    {
+      "epoch": 2.3891402714932126,
+      "grad_norm": 0.3652385175228119,
+      "learning_rate": 9.423604494395686e-06,
+      "loss": 0.336,
+      "step": 528
+    },
+    {
+      "epoch": 2.3936651583710407,
+      "grad_norm": 0.37585943937301636,
+      "learning_rate": 9.419917824660311e-06,
+      "loss": 0.297,
+      "step": 529
+    },
+    {
+      "epoch": 2.3981900452488687,
+      "grad_norm": 0.4154417812824249,
+      "learning_rate": 9.416220128257317e-06,
+      "loss": 0.3214,
+      "step": 530
+    },
+    {
+      "epoch": 2.4027149321266967,
+      "grad_norm": 0.3846287131309509,
+      "learning_rate": 9.41251141441159e-06,
+      "loss": 0.3087,
+      "step": 531
+    },
+    {
+      "epoch": 2.4072398190045248,
+      "grad_norm": 0.3685553967952728,
+      "learning_rate": 9.408791692375511e-06,
+      "loss": 0.2663,
+      "step": 532
+    },
+    {
+      "epoch": 2.411764705882353,
+      "grad_norm": 0.3650501072406769,
+      "learning_rate": 9.405060971428924e-06,
+      "loss": 0.2475,
+      "step": 533
+    },
+    {
+      "epoch": 2.416289592760181,
+      "grad_norm": 0.4251144230365753,
+      "learning_rate": 9.401319260879107e-06,
+      "loss": 0.346,
+      "step": 534
+    },
+    {
+      "epoch": 2.420814479638009,
+      "grad_norm": 0.3708731532096863,
+      "learning_rate": 9.397566570060761e-06,
+      "loss": 0.252,
+      "step": 535
+    },
+    {
+      "epoch": 2.425339366515837,
+      "grad_norm": 0.3827893137931824,
+      "learning_rate": 9.393802908335978e-06,
+      "loss": 0.284,
+      "step": 536
+    },
+    {
+      "epoch": 2.4298642533936654,
+      "grad_norm": 0.401019424200058,
+      "learning_rate": 9.390028285094215e-06,
+      "loss": 0.2954,
+      "step": 537
+    },
+    {
+      "epoch": 2.4343891402714934,
+      "grad_norm": 0.4581804871559143,
+      "learning_rate": 9.386242709752283e-06,
+      "loss": 0.3333,
+      "step": 538
+    },
+    {
+      "epoch": 2.4389140271493215,
+      "grad_norm": 0.3979438245296478,
+      "learning_rate": 9.382446191754313e-06,
+      "loss": 0.2901,
+      "step": 539
+    },
+    {
+      "epoch": 2.4434389140271495,
+      "grad_norm": 0.3686637580394745,
+      "learning_rate": 9.378638740571733e-06,
+      "loss": 0.275,
+      "step": 540
+    },
+    {
+      "epoch": 2.4479638009049776,
+      "grad_norm": 0.42701080441474915,
+      "learning_rate": 9.37482036570325e-06,
+      "loss": 0.3166,
+      "step": 541
+    },
+    {
+      "epoch": 2.4524886877828056,
+      "grad_norm": 0.39671456813812256,
+      "learning_rate": 9.370991076674821e-06,
+      "loss": 0.2416,
+      "step": 542
+    },
+    {
+      "epoch": 2.4570135746606336,
+      "grad_norm": 0.4166645109653473,
+      "learning_rate": 9.36715088303963e-06,
+      "loss": 0.2876,
+      "step": 543
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.4189673066139221,
+      "learning_rate": 9.363299794378072e-06,
+      "loss": 0.2872,
+      "step": 544
+    },
+    {
+      "epoch": 2.4660633484162897,
+      "grad_norm": 0.4147026836872101,
+      "learning_rate": 9.359437820297716e-06,
+      "loss": 0.2781,
+      "step": 545
+    },
+    {
+      "epoch": 2.4705882352941178,
+      "grad_norm": 0.40569254755973816,
+      "learning_rate": 9.355564970433288e-06,
+      "loss": 0.2911,
+      "step": 546
+    },
+    {
+      "epoch": 2.475113122171946,
+      "grad_norm": 0.40611332654953003,
+      "learning_rate": 9.351681254446651e-06,
+      "loss": 0.2691,
+      "step": 547
+    },
+    {
+      "epoch": 2.479638009049774,
+      "grad_norm": 0.3574923276901245,
+      "learning_rate": 9.347786682026774e-06,
+      "loss": 0.2476,
+      "step": 548
+    },
+    {
+      "epoch": 2.484162895927602,
+      "grad_norm": 0.44315120577812195,
+      "learning_rate": 9.343881262889706e-06,
+      "loss": 0.2976,
+      "step": 549
+    },
+    {
+      "epoch": 2.48868778280543,
+      "grad_norm": 0.45607924461364746,
+      "learning_rate": 9.339965006778564e-06,
+      "loss": 0.2874,
+      "step": 550
+    },
+    {
+      "epoch": 2.493212669683258,
+      "grad_norm": 0.4020167291164398,
+      "learning_rate": 9.336037923463494e-06,
+      "loss": 0.2975,
+      "step": 551
+    },
+    {
+      "epoch": 2.497737556561086,
+      "grad_norm": 0.3714078962802887,
+      "learning_rate": 9.332100022741657e-06,
+      "loss": 0.276,
+      "step": 552
+    },
+    {
+      "epoch": 2.502262443438914,
+      "grad_norm": 0.4652123749256134,
+      "learning_rate": 9.328151314437201e-06,
+      "loss": 0.3213,
+      "step": 553
+    },
+    {
+      "epoch": 2.506787330316742,
+      "grad_norm": 0.4106157124042511,
+      "learning_rate": 9.324191808401235e-06,
+      "loss": 0.3131,
+      "step": 554
+    },
+    {
+      "epoch": 2.51131221719457,
+      "grad_norm": 0.41118085384368896,
+      "learning_rate": 9.320221514511805e-06,
+      "loss": 0.3477,
+      "step": 555
+    },
+    {
+      "epoch": 2.515837104072398,
+      "grad_norm": 0.39603039622306824,
+      "learning_rate": 9.316240442673872e-06,
+      "loss": 0.3234,
+      "step": 556
+    },
+    {
+      "epoch": 2.520361990950226,
+      "grad_norm": 0.4552370309829712,
+      "learning_rate": 9.312248602819284e-06,
+      "loss": 0.3104,
+      "step": 557
+    },
+    {
+      "epoch": 2.524886877828054,
+      "grad_norm": 0.4341452717781067,
+      "learning_rate": 9.308246004906758e-06,
+      "loss": 0.3191,
+      "step": 558
+    },
+    {
+      "epoch": 2.5294117647058822,
+      "grad_norm": 0.33837419748306274,
+      "learning_rate": 9.30423265892184e-06,
+      "loss": 0.287,
+      "step": 559
+    },
+    {
+      "epoch": 2.5339366515837103,
+      "grad_norm": 0.4650215804576874,
+      "learning_rate": 9.300208574876897e-06,
+      "loss": 0.2909,
+      "step": 560
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.3926837146282196,
+      "learning_rate": 9.296173762811084e-06,
+      "loss": 0.2819,
+      "step": 561
+    },
+    {
+      "epoch": 2.5429864253393664,
+      "grad_norm": 0.4534710943698883,
+      "learning_rate": 9.292128232790321e-06,
+      "loss": 0.3077,
+      "step": 562
+    },
+    {
+      "epoch": 2.5475113122171944,
+      "grad_norm": 0.38600146770477295,
+      "learning_rate": 9.288071994907262e-06,
+      "loss": 0.3145,
+      "step": 563
+    },
+    {
+      "epoch": 2.5520361990950224,
+      "grad_norm": 0.4256378710269928,
+      "learning_rate": 9.284005059281278e-06,
+      "loss": 0.2503,
+      "step": 564
+    },
+    {
+      "epoch": 2.5565610859728505,
+      "grad_norm": 0.4025668799877167,
+      "learning_rate": 9.27992743605843e-06,
+      "loss": 0.3042,
+      "step": 565
+    },
+    {
+      "epoch": 2.5610859728506785,
+      "grad_norm": 0.43262800574302673,
+      "learning_rate": 9.275839135411439e-06,
+      "loss": 0.3031,
+      "step": 566
+    },
+    {
+      "epoch": 2.5656108597285066,
+      "grad_norm": 0.42043566703796387,
+      "learning_rate": 9.271740167539665e-06,
+      "loss": 0.2848,
+      "step": 567
+    },
+    {
+      "epoch": 2.5701357466063346,
+      "grad_norm": 0.43189898133277893,
+      "learning_rate": 9.26763054266908e-06,
+      "loss": 0.278,
+      "step": 568
+    },
+    {
+      "epoch": 2.5746606334841626,
+      "grad_norm": 0.40669795870780945,
+      "learning_rate": 9.263510271052243e-06,
+      "loss": 0.3034,
+      "step": 569
+    },
+    {
+      "epoch": 2.579185520361991,
+      "grad_norm": 0.3897744119167328,
+      "learning_rate": 9.259379362968276e-06,
+      "loss": 0.2455,
+      "step": 570
+    },
+    {
+      "epoch": 2.583710407239819,
+      "grad_norm": 0.4340456426143646,
+      "learning_rate": 9.255237828722828e-06,
+      "loss": 0.2672,
+      "step": 571
+    },
+    {
+      "epoch": 2.588235294117647,
+      "grad_norm": 0.4333140254020691,
+      "learning_rate": 9.251085678648072e-06,
+      "loss": 0.3283,
+      "step": 572
+    },
+    {
+      "epoch": 2.5927601809954752,
+      "grad_norm": 0.5001205205917358,
+      "learning_rate": 9.246922923102652e-06,
+      "loss": 0.3178,
+      "step": 573
+    },
+    {
+      "epoch": 2.5972850678733033,
+      "grad_norm": 0.42825013399124146,
+      "learning_rate": 9.242749572471679e-06,
+      "loss": 0.3468,
+      "step": 574
+    },
+    {
+      "epoch": 2.6018099547511313,
+      "grad_norm": 0.4305206835269928,
+      "learning_rate": 9.238565637166692e-06,
+      "loss": 0.2952,
+      "step": 575
+    },
+    {
+      "epoch": 2.6063348416289593,
+      "grad_norm": 0.4100704789161682,
+      "learning_rate": 9.234371127625634e-06,
+      "loss": 0.3348,
+      "step": 576
+    },
+    {
+      "epoch": 2.6108597285067874,
+      "grad_norm": 0.37542393803596497,
+      "learning_rate": 9.230166054312834e-06,
+      "loss": 0.3066,
+      "step": 577
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.38587042689323425,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 0.3205,
+      "step": 578
+    },
+    {
+      "epoch": 2.6199095022624435,
+      "grad_norm": 0.4810059368610382,
+      "learning_rate": 9.22172425836106e-06,
+      "loss": 0.3374,
+      "step": 579
+    },
+    {
+      "epoch": 2.6244343891402715,
+      "grad_norm": 0.43957462906837463,
+      "learning_rate": 9.217487556782402e-06,
+      "loss": 0.3029,
+      "step": 580
+    },
+    {
+      "epoch": 2.6289592760180995,
+      "grad_norm": 0.39742496609687805,
+      "learning_rate": 9.213240333552589e-06,
+      "loss": 0.3118,
+      "step": 581
+    },
+    {
+      "epoch": 2.6334841628959276,
+      "grad_norm": 0.4322555661201477,
+      "learning_rate": 9.208982599267454e-06,
+      "loss": 0.3043,
+      "step": 582
+    },
+    {
+      "epoch": 2.6380090497737556,
+      "grad_norm": 0.43412262201309204,
+      "learning_rate": 9.20471436454905e-06,
+      "loss": 0.3009,
+      "step": 583
+    },
+    {
+      "epoch": 2.6425339366515836,
+      "grad_norm": 0.35117828845977783,
+      "learning_rate": 9.200435640045637e-06,
+      "loss": 0.3099,
+      "step": 584
+    },
+    {
+      "epoch": 2.6470588235294117,
+      "grad_norm": 0.4399445354938507,
+      "learning_rate": 9.196146436431635e-06,
+      "loss": 0.3394,
+      "step": 585
+    },
+    {
+      "epoch": 2.6515837104072397,
+      "grad_norm": 0.4506850242614746,
+      "learning_rate": 9.191846764407609e-06,
+      "loss": 0.3124,
+      "step": 586
+    },
+    {
+      "epoch": 2.6561085972850678,
+      "grad_norm": 0.43107202649116516,
+      "learning_rate": 9.187536634700244e-06,
+      "loss": 0.2903,
+      "step": 587
+    },
+    {
+      "epoch": 2.660633484162896,
+      "grad_norm": 0.44276267290115356,
+      "learning_rate": 9.18321605806231e-06,
+      "loss": 0.2783,
+      "step": 588
+    },
+    {
+      "epoch": 2.665158371040724,
+      "grad_norm": 0.4580158293247223,
+      "learning_rate": 9.178885045272642e-06,
+      "loss": 0.3157,
+      "step": 589
+    },
+    {
+      "epoch": 2.669683257918552,
+      "grad_norm": 0.39571404457092285,
+      "learning_rate": 9.174543607136111e-06,
+      "loss": 0.3387,
+      "step": 590
+    },
+    {
+      "epoch": 2.6742081447963804,
+      "grad_norm": 0.4113115072250366,
+      "learning_rate": 9.170191754483594e-06,
+      "loss": 0.2723,
+      "step": 591
+    },
+    {
+      "epoch": 2.6787330316742084,
+      "grad_norm": 0.3653103709220886,
+      "learning_rate": 9.165829498171956e-06,
+      "loss": 0.2986,
+      "step": 592
+    },
+    {
+      "epoch": 2.6832579185520364,
+      "grad_norm": 0.381362646818161,
+      "learning_rate": 9.161456849084007e-06,
+      "loss": 0.2845,
+      "step": 593
+    },
+    {
+      "epoch": 2.6877828054298645,
+      "grad_norm": 0.38520559668540955,
+      "learning_rate": 9.157073818128495e-06,
+      "loss": 0.2455,
+      "step": 594
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.4085952043533325,
+      "learning_rate": 9.152680416240059e-06,
+      "loss": 0.3021,
+      "step": 595
+    },
+    {
+      "epoch": 2.6968325791855206,
+      "grad_norm": 0.3726223409175873,
+      "learning_rate": 9.14827665437922e-06,
+      "loss": 0.2836,
+      "step": 596
+    },
+    {
+      "epoch": 2.7013574660633486,
+      "grad_norm": 0.40436556935310364,
+      "learning_rate": 9.143862543532336e-06,
+      "loss": 0.3051,
+      "step": 597
+    },
+    {
+      "epoch": 2.7058823529411766,
+      "grad_norm": 0.46403786540031433,
+      "learning_rate": 9.13943809471159e-06,
+      "loss": 0.2604,
+      "step": 598
+    },
+    {
+      "epoch": 2.7104072398190047,
+      "grad_norm": 0.4343224763870239,
+      "learning_rate": 9.135003318954954e-06,
+      "loss": 0.3354,
+      "step": 599
+    },
+    {
+      "epoch": 2.7149321266968327,
+      "grad_norm": 0.463319331407547,
+      "learning_rate": 9.13055822732616e-06,
+      "loss": 0.3098,
+      "step": 600
+    },
+    {
+      "epoch": 2.7194570135746607,
+      "grad_norm": 0.48439162969589233,
+      "learning_rate": 9.126102830914682e-06,
+      "loss": 0.2805,
+      "step": 601
+    },
+    {
+      "epoch": 2.723981900452489,
+      "grad_norm": 0.46056264638900757,
+      "learning_rate": 9.121637140835696e-06,
+      "loss": 0.3309,
+      "step": 602
+    },
+    {
+      "epoch": 2.728506787330317,
+      "grad_norm": 0.44612205028533936,
+      "learning_rate": 9.11716116823006e-06,
+      "loss": 0.3131,
+      "step": 603
+    },
+    {
+      "epoch": 2.733031674208145,
+      "grad_norm": 0.45993444323539734,
+      "learning_rate": 9.112674924264287e-06,
+      "loss": 0.2702,
+      "step": 604
+    },
+    {
+      "epoch": 2.737556561085973,
+      "grad_norm": 0.41719764471054077,
+      "learning_rate": 9.108178420130514e-06,
+      "loss": 0.288,
+      "step": 605
+    },
+    {
+      "epoch": 2.742081447963801,
+      "grad_norm": 0.49494820833206177,
+      "learning_rate": 9.103671667046472e-06,
+      "loss": 0.3831,
+      "step": 606
+    },
+    {
+      "epoch": 2.746606334841629,
+      "grad_norm": 0.47351089119911194,
+      "learning_rate": 9.099154676255462e-06,
+      "loss": 0.3197,
+      "step": 607
+    },
+    {
+      "epoch": 2.751131221719457,
+      "grad_norm": 0.4318673312664032,
+      "learning_rate": 9.094627459026326e-06,
+      "loss": 0.278,
+      "step": 608
+    },
+    {
+      "epoch": 2.755656108597285,
+      "grad_norm": 0.4779452383518219,
+      "learning_rate": 9.09009002665342e-06,
+      "loss": 0.3151,
+      "step": 609
+    },
+    {
+      "epoch": 2.760180995475113,
+      "grad_norm": 0.4163115322589874,
+      "learning_rate": 9.085542390456582e-06,
+      "loss": 0.3073,
+      "step": 610
+    },
+    {
+      "epoch": 2.764705882352941,
+      "grad_norm": 0.39201343059539795,
+      "learning_rate": 9.08098456178111e-06,
+      "loss": 0.3137,
+      "step": 611
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.5050286650657654,
+      "learning_rate": 9.076416551997721e-06,
+      "loss": 0.3052,
+      "step": 612
+    },
+    {
+      "epoch": 2.773755656108597,
+      "grad_norm": 0.4312725365161896,
+      "learning_rate": 9.071838372502546e-06,
+      "loss": 0.266,
+      "step": 613
+    },
+    {
+      "epoch": 2.7782805429864252,
+      "grad_norm": 0.42680007219314575,
+      "learning_rate": 9.067250034717072e-06,
+      "loss": 0.2631,
+      "step": 614
+    },
+    {
+      "epoch": 2.7828054298642533,
+      "grad_norm": 0.40380024909973145,
+      "learning_rate": 9.062651550088137e-06,
+      "loss": 0.2268,
+      "step": 615
+    },
+    {
+      "epoch": 2.7873303167420813,
+      "grad_norm": 0.395770788192749,
+      "learning_rate": 9.058042930087894e-06,
+      "loss": 0.2925,
+      "step": 616
+    },
+    {
+      "epoch": 2.7918552036199094,
+      "grad_norm": 0.47229576110839844,
+      "learning_rate": 9.053424186213776e-06,
+      "loss": 0.3023,
+      "step": 617
+    },
+    {
+      "epoch": 2.7963800904977374,
+      "grad_norm": 0.4000561833381653,
+      "learning_rate": 9.048795329988473e-06,
+      "loss": 0.2983,
+      "step": 618
+    },
+    {
+      "epoch": 2.8009049773755654,
+      "grad_norm": 0.39670804142951965,
+      "learning_rate": 9.044156372959911e-06,
+      "loss": 0.2657,
+      "step": 619
+    },
+    {
+      "epoch": 2.8054298642533935,
+      "grad_norm": 0.38643792271614075,
+      "learning_rate": 9.039507326701207e-06,
+      "loss": 0.2793,
+      "step": 620
+    },
+    {
+      "epoch": 2.8099547511312215,
+      "grad_norm": 0.4341120421886444,
+      "learning_rate": 9.03484820281065e-06,
+      "loss": 0.3408,
+      "step": 621
+    },
+    {
+      "epoch": 2.8144796380090495,
+      "grad_norm": 0.3975633382797241,
+      "learning_rate": 9.030179012911672e-06,
+      "loss": 0.2507,
+      "step": 622
+    },
+    {
+      "epoch": 2.8190045248868776,
+      "grad_norm": 0.3742230236530304,
+      "learning_rate": 9.025499768652817e-06,
+      "loss": 0.2768,
+      "step": 623
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 0.40196242928504944,
+      "learning_rate": 9.020810481707709e-06,
+      "loss": 0.2508,
+      "step": 624
+    },
+    {
+      "epoch": 2.8280542986425337,
+      "grad_norm": 0.41893255710601807,
+      "learning_rate": 9.016111163775032e-06,
+      "loss": 0.3294,
+      "step": 625
+    },
+    {
+      "epoch": 2.832579185520362,
+      "grad_norm": 0.3998293876647949,
+      "learning_rate": 9.011401826578492e-06,
+      "loss": 0.2652,
+      "step": 626
+    },
+    {
+      "epoch": 2.83710407239819,
+      "grad_norm": 0.502552330493927,
+      "learning_rate": 9.00668248186679e-06,
+      "loss": 0.2816,
+      "step": 627
+    },
+    {
+      "epoch": 2.841628959276018,
+      "grad_norm": 0.4417928457260132,
+      "learning_rate": 9.001953141413593e-06,
+      "loss": 0.3273,
+      "step": 628
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.40362390875816345,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.3004,
+      "step": 629
+    },
+    {
+      "epoch": 2.8506787330316743,
+      "grad_norm": 0.39725273847579956,
+      "learning_rate": 8.992464520502045e-06,
+      "loss": 0.2706,
+      "step": 630
+    },
+    {
+      "epoch": 2.8552036199095023,
+      "grad_norm": 0.407359778881073,
+      "learning_rate": 8.987705263715598e-06,
+      "loss": 0.2857,
+      "step": 631
+    },
+    {
+      "epoch": 2.8597285067873304,
+      "grad_norm": 0.4246702790260315,
+      "learning_rate": 8.982936058531403e-06,
+      "loss": 0.2658,
+      "step": 632
+    },
+    {
+      "epoch": 2.8642533936651584,
+      "grad_norm": 0.3781566619873047,
+      "learning_rate": 8.978156916847523e-06,
+      "loss": 0.2845,
+      "step": 633
+    },
+    {
+      "epoch": 2.8687782805429864,
+      "grad_norm": 0.4160442650318146,
+      "learning_rate": 8.973367850586802e-06,
+      "loss": 0.3012,
+      "step": 634
+    },
+    {
+      "epoch": 2.8733031674208145,
+      "grad_norm": 0.44970279932022095,
+      "learning_rate": 8.968568871696847e-06,
+      "loss": 0.3212,
+      "step": 635
+    },
+    {
+      "epoch": 2.8778280542986425,
+      "grad_norm": 0.38865357637405396,
+      "learning_rate": 8.96375999215e-06,
+      "loss": 0.2997,
+      "step": 636
+    },
+    {
+      "epoch": 2.8823529411764706,
+      "grad_norm": 0.4079996943473816,
+      "learning_rate": 8.958941223943292e-06,
+      "loss": 0.3396,
+      "step": 637
+    },
+    {
+      "epoch": 2.8868778280542986,
+      "grad_norm": 0.40862202644348145,
+      "learning_rate": 8.95411257909843e-06,
+      "loss": 0.2484,
+      "step": 638
+    },
+    {
+      "epoch": 2.8914027149321266,
+      "grad_norm": 0.39957407116889954,
+      "learning_rate": 8.949274069661765e-06,
+      "loss": 0.2696,
+      "step": 639
+    },
+    {
+      "epoch": 2.8959276018099547,
+      "grad_norm": 0.412178635597229,
+      "learning_rate": 8.944425707704247e-06,
+      "loss": 0.2805,
+      "step": 640
+    },
+    {
+      "epoch": 2.9004524886877827,
+      "grad_norm": 0.40401560068130493,
+      "learning_rate": 8.939567505321418e-06,
+      "loss": 0.26,
+      "step": 641
+    },
+    {
+      "epoch": 2.9049773755656108,
+      "grad_norm": 0.42878180742263794,
+      "learning_rate": 8.934699474633362e-06,
+      "loss": 0.3284,
+      "step": 642
+    },
+    {
+      "epoch": 2.909502262443439,
+      "grad_norm": 0.4750427007675171,
+      "learning_rate": 8.929821627784682e-06,
+      "loss": 0.3018,
+      "step": 643
+    },
+    {
+      "epoch": 2.914027149321267,
+      "grad_norm": 0.42947152256965637,
+      "learning_rate": 8.924933976944474e-06,
+      "loss": 0.3251,
+      "step": 644
+    },
+    {
+      "epoch": 2.918552036199095,
+      "grad_norm": 0.36342573165893555,
+      "learning_rate": 8.92003653430629e-06,
+      "loss": 0.2878,
+      "step": 645
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.4305476248264313,
+      "learning_rate": 8.915129312088112e-06,
+      "loss": 0.2904,
+      "step": 646
+    },
+    {
+      "epoch": 2.9276018099547514,
+      "grad_norm": 0.4355884790420532,
+      "learning_rate": 8.910212322532317e-06,
+      "loss": 0.2952,
+      "step": 647
+    },
+    {
+      "epoch": 2.9321266968325794,
+      "grad_norm": 0.39653098583221436,
+      "learning_rate": 8.905285577905653e-06,
+      "loss": 0.3174,
+      "step": 648
+    },
+    {
+      "epoch": 2.9366515837104075,
+      "grad_norm": 0.38590750098228455,
+      "learning_rate": 8.900349090499203e-06,
+      "loss": 0.2551,
+      "step": 649
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 0.43237945437431335,
+      "learning_rate": 8.895402872628352e-06,
+      "loss": 0.3424,
+      "step": 650
+    },
+    {
+      "epoch": 2.9457013574660635,
+      "grad_norm": 0.4289698004722595,
+      "learning_rate": 8.890446936632768e-06,
+      "loss": 0.3434,
+      "step": 651
+    },
+    {
+      "epoch": 2.9502262443438916,
+      "grad_norm": 0.426191121339798,
+      "learning_rate": 8.88548129487636e-06,
+      "loss": 0.3057,
+      "step": 652
+    },
+    {
+      "epoch": 2.9547511312217196,
+      "grad_norm": 0.4147898852825165,
+      "learning_rate": 8.880505959747245e-06,
+      "loss": 0.262,
+      "step": 653
+    },
+    {
+      "epoch": 2.9592760180995477,
+      "grad_norm": 0.4329768717288971,
+      "learning_rate": 8.87552094365773e-06,
+      "loss": 0.2902,
+      "step": 654
+    },
+    {
+      "epoch": 2.9638009049773757,
+      "grad_norm": 0.4282679855823517,
+      "learning_rate": 8.870526259044275e-06,
+      "loss": 0.2878,
+      "step": 655
+    },
+    {
+      "epoch": 2.9683257918552037,
+      "grad_norm": 0.3876934349536896,
+      "learning_rate": 8.86552191836745e-06,
+      "loss": 0.3095,
+      "step": 656
+    },
+    {
+      "epoch": 2.9728506787330318,
+      "grad_norm": 0.44530197978019714,
+      "learning_rate": 8.860507934111929e-06,
+      "loss": 0.2807,
+      "step": 657
+    },
+    {
+      "epoch": 2.97737556561086,
+      "grad_norm": 0.4089668095111847,
+      "learning_rate": 8.855484318786433e-06,
+      "loss": 0.323,
+      "step": 658
+    },
+    {
+      "epoch": 2.981900452488688,
+      "grad_norm": 0.4198194146156311,
+      "learning_rate": 8.850451084923717e-06,
+      "loss": 0.3207,
+      "step": 659
+    },
+    {
+      "epoch": 2.986425339366516,
+      "grad_norm": 0.4462484121322632,
+      "learning_rate": 8.845408245080527e-06,
+      "loss": 0.3145,
+      "step": 660
+    },
+    {
+      "epoch": 2.990950226244344,
+      "grad_norm": 0.39047619700431824,
+      "learning_rate": 8.840355811837578e-06,
+      "loss": 0.3429,
+      "step": 661
+    },
+    {
+      "epoch": 2.995475113122172,
+      "grad_norm": 0.38414013385772705,
+      "learning_rate": 8.835293797799517e-06,
+      "loss": 0.3076,
+      "step": 662
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.41144996881484985,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.2834,
+      "step": 663
+    },
+    {
+      "epoch": 3.004524886877828,
+      "grad_norm": 0.4691222012042999,
+      "learning_rate": 8.82514107787612e-06,
+      "loss": 0.2059,
+      "step": 664
+    },
+    {
+      "epoch": 3.009049773755656,
+      "grad_norm": 0.40257298946380615,
+      "learning_rate": 8.82005039731946e-06,
+      "loss": 0.1973,
+      "step": 665
+    },
+    {
+      "epoch": 3.013574660633484,
+      "grad_norm": 0.40091562271118164,
+      "learning_rate": 8.814950186624978e-06,
+      "loss": 0.1897,
+      "step": 666
+    },
+    {
+      "epoch": 3.018099547511312,
+      "grad_norm": 0.5339998006820679,
+      "learning_rate": 8.809840458516511e-06,
+      "loss": 0.1607,
+      "step": 667
+    },
+    {
+      "epoch": 3.02262443438914,
+      "grad_norm": 0.5496724247932434,
+      "learning_rate": 8.804721225741646e-06,
+      "loss": 0.2262,
+      "step": 668
+    },
+    {
+      "epoch": 3.0271493212669682,
+      "grad_norm": 0.4552634060382843,
+      "learning_rate": 8.799592501071679e-06,
+      "loss": 0.2015,
+      "step": 669
+    },
+    {
+      "epoch": 3.0316742081447963,
+      "grad_norm": 0.4297693073749542,
+      "learning_rate": 8.794454297301583e-06,
+      "loss": 0.1849,
+      "step": 670
+    },
+    {
+      "epoch": 3.0361990950226243,
+      "grad_norm": 0.46762946248054504,
+      "learning_rate": 8.789306627249985e-06,
+      "loss": 0.2071,
+      "step": 671
+    },
+    {
+      "epoch": 3.0407239819004523,
+      "grad_norm": 0.43196043372154236,
+      "learning_rate": 8.784149503759124e-06,
+      "loss": 0.1807,
+      "step": 672
+    },
+    {
+      "epoch": 3.0452488687782804,
+      "grad_norm": 0.39062273502349854,
+      "learning_rate": 8.778982939694822e-06,
+      "loss": 0.2068,
+      "step": 673
+    },
+    {
+      "epoch": 3.0497737556561084,
+      "grad_norm": 0.4284631907939911,
+      "learning_rate": 8.77380694794646e-06,
+      "loss": 0.1752,
+      "step": 674
+    },
+    {
+      "epoch": 3.0542986425339365,
+      "grad_norm": 0.5311867594718933,
+      "learning_rate": 8.768621541426933e-06,
+      "loss": 0.2363,
+      "step": 675
+    },
+    {
+      "epoch": 3.0588235294117645,
+      "grad_norm": 0.4057539701461792,
+      "learning_rate": 8.763426733072624e-06,
+      "loss": 0.1695,
+      "step": 676
+    },
+    {
+      "epoch": 3.0633484162895925,
+      "grad_norm": 0.4523632526397705,
+      "learning_rate": 8.75822253584337e-06,
+      "loss": 0.1925,
+      "step": 677
+    },
+    {
+      "epoch": 3.067873303167421,
+      "grad_norm": 0.40796959400177,
+      "learning_rate": 8.75300896272244e-06,
+      "loss": 0.2148,
+      "step": 678
+    },
+    {
+      "epoch": 3.072398190045249,
+      "grad_norm": 0.39213889837265015,
+      "learning_rate": 8.747786026716481e-06,
+      "loss": 0.2128,
+      "step": 679
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.48423635959625244,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.2061,
+      "step": 680
+    },
+    {
+      "epoch": 3.081447963800905,
+      "grad_norm": 0.4347093403339386,
+      "learning_rate": 8.737312118192853e-06,
+      "loss": 0.2512,
+      "step": 681
+    },
+    {
+      "epoch": 3.085972850678733,
+      "grad_norm": 0.4026263356208801,
+      "learning_rate": 8.732061171805154e-06,
+      "loss": 0.1843,
+      "step": 682
+    },
+    {
+      "epoch": 3.090497737556561,
+      "grad_norm": 0.4047137200832367,
+      "learning_rate": 8.726800914792296e-06,
+      "loss": 0.1738,
+      "step": 683
+    },
+    {
+      "epoch": 3.0950226244343892,
+      "grad_norm": 0.4355809688568115,
+      "learning_rate": 8.721531360277398e-06,
+      "loss": 0.2185,
+      "step": 684
+    },
+    {
+      "epoch": 3.0995475113122173,
+      "grad_norm": 0.4358494281768799,
+      "learning_rate": 8.716252521406776e-06,
+      "loss": 0.2087,
+      "step": 685
+    },
+    {
+      "epoch": 3.1040723981900453,
+      "grad_norm": 0.4449985921382904,
+      "learning_rate": 8.710964411349902e-06,
+      "loss": 0.2285,
+      "step": 686
+    },
+    {
+      "epoch": 3.1085972850678734,
+      "grad_norm": 0.38388368487358093,
+      "learning_rate": 8.705667043299388e-06,
+      "loss": 0.1556,
+      "step": 687
+    },
+    {
+      "epoch": 3.1131221719457014,
+      "grad_norm": 0.4281388521194458,
+      "learning_rate": 8.70036043047093e-06,
+      "loss": 0.2041,
+      "step": 688
+    },
+    {
+      "epoch": 3.1176470588235294,
+      "grad_norm": 0.37215033173561096,
+      "learning_rate": 8.695044586103297e-06,
+      "loss": 0.2017,
+      "step": 689
+    },
+    {
+      "epoch": 3.1221719457013575,
+      "grad_norm": 0.4040149748325348,
+      "learning_rate": 8.689719523458282e-06,
+      "loss": 0.1838,
+      "step": 690
+    },
+    {
+      "epoch": 3.1266968325791855,
+      "grad_norm": 0.3721625804901123,
+      "learning_rate": 8.684385255820682e-06,
+      "loss": 0.1982,
+      "step": 691
+    },
+    {
+      "epoch": 3.1312217194570136,
+      "grad_norm": 0.37194955348968506,
+      "learning_rate": 8.679041796498253e-06,
+      "loss": 0.1832,
+      "step": 692
+    },
+    {
+      "epoch": 3.1357466063348416,
+      "grad_norm": 0.3825089633464813,
+      "learning_rate": 8.673689158821684e-06,
+      "loss": 0.1753,
+      "step": 693
+    },
+    {
+      "epoch": 3.1402714932126696,
+      "grad_norm": 0.41846585273742676,
+      "learning_rate": 8.668327356144562e-06,
+      "loss": 0.1979,
+      "step": 694
+    },
+    {
+      "epoch": 3.1447963800904977,
+      "grad_norm": 0.4502934515476227,
+      "learning_rate": 8.66295640184334e-06,
+      "loss": 0.2344,
+      "step": 695
+    },
+    {
+      "epoch": 3.1493212669683257,
+      "grad_norm": 0.41532212495803833,
+      "learning_rate": 8.657576309317297e-06,
+      "loss": 0.1949,
+      "step": 696
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.4130610525608063,
+      "learning_rate": 8.652187091988516e-06,
+      "loss": 0.2038,
+      "step": 697
+    },
+    {
+      "epoch": 3.158371040723982,
+      "grad_norm": 0.42346063256263733,
+      "learning_rate": 8.646788763301842e-06,
+      "loss": 0.2081,
+      "step": 698
+    },
+    {
+      "epoch": 3.16289592760181,
+      "grad_norm": 0.3540644645690918,
+      "learning_rate": 8.641381336724848e-06,
+      "loss": 0.1972,
+      "step": 699
+    },
+    {
+      "epoch": 3.167420814479638,
+      "grad_norm": 0.4263281226158142,
+      "learning_rate": 8.635964825747809e-06,
+      "loss": 0.1815,
+      "step": 700
+    },
+    {
+      "epoch": 3.171945701357466,
+      "grad_norm": 0.442663311958313,
+      "learning_rate": 8.630539243883659e-06,
+      "loss": 0.1829,
+      "step": 701
+    },
+    {
+      "epoch": 3.176470588235294,
+      "grad_norm": 0.4172739088535309,
+      "learning_rate": 8.625104604667965e-06,
+      "loss": 0.203,
+      "step": 702
+    },
+    {
+      "epoch": 3.180995475113122,
+      "grad_norm": 0.38168448209762573,
+      "learning_rate": 8.619660921658885e-06,
+      "loss": 0.2247,
+      "step": 703
+    },
+    {
+      "epoch": 3.1855203619909505,
+      "grad_norm": 0.3923109769821167,
+      "learning_rate": 8.61420820843715e-06,
+      "loss": 0.1861,
+      "step": 704
+    },
+    {
+      "epoch": 3.1900452488687785,
+      "grad_norm": 0.4268626272678375,
+      "learning_rate": 8.608746478606004e-06,
+      "loss": 0.1889,
+      "step": 705
+    },
+    {
+      "epoch": 3.1945701357466065,
+      "grad_norm": 0.567819356918335,
+      "learning_rate": 8.6032757457912e-06,
+      "loss": 0.2458,
+      "step": 706
+    },
+    {
+      "epoch": 3.1990950226244346,
+      "grad_norm": 0.4022853970527649,
+      "learning_rate": 8.59779602364094e-06,
+      "loss": 0.2101,
+      "step": 707
+    },
+    {
+      "epoch": 3.2036199095022626,
+      "grad_norm": 0.3837186098098755,
+      "learning_rate": 8.592307325825861e-06,
+      "loss": 0.1734,
+      "step": 708
+    },
+    {
+      "epoch": 3.2081447963800906,
+      "grad_norm": 0.3684345483779907,
+      "learning_rate": 8.586809666038986e-06,
+      "loss": 0.1862,
+      "step": 709
+    },
+    {
+      "epoch": 3.2126696832579187,
+      "grad_norm": 0.4099283814430237,
+      "learning_rate": 8.581303057995697e-06,
+      "loss": 0.1787,
+      "step": 710
+    },
+    {
+      "epoch": 3.2171945701357467,
+      "grad_norm": 0.4496215879917145,
+      "learning_rate": 8.575787515433705e-06,
+      "loss": 0.2169,
+      "step": 711
+    },
+    {
+      "epoch": 3.2217194570135748,
+      "grad_norm": 0.3973177969455719,
+      "learning_rate": 8.570263052113005e-06,
+      "loss": 0.2127,
+      "step": 712
+    },
+    {
+      "epoch": 3.226244343891403,
+      "grad_norm": 0.40007996559143066,
+      "learning_rate": 8.564729681815846e-06,
+      "loss": 0.1987,
+      "step": 713
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.37854504585266113,
+      "learning_rate": 8.559187418346703e-06,
+      "loss": 0.2099,
+      "step": 714
+    },
+    {
+      "epoch": 3.235294117647059,
+      "grad_norm": 0.4037870466709137,
+      "learning_rate": 8.553636275532236e-06,
+      "loss": 0.2049,
+      "step": 715
+    },
+    {
+      "epoch": 3.239819004524887,
+      "grad_norm": 0.4421459436416626,
+      "learning_rate": 8.548076267221258e-06,
+      "loss": 0.1905,
+      "step": 716
+    },
+    {
+      "epoch": 3.244343891402715,
+      "grad_norm": 0.9329970479011536,
+      "learning_rate": 8.54250740728469e-06,
+      "loss": 0.1844,
+      "step": 717
+    },
+    {
+      "epoch": 3.248868778280543,
+      "grad_norm": 0.44006776809692383,
+      "learning_rate": 8.536929709615552e-06,
+      "loss": 0.1995,
+      "step": 718
+    },
+    {
+      "epoch": 3.253393665158371,
+      "grad_norm": 0.3820533752441406,
+      "learning_rate": 8.531343188128896e-06,
+      "loss": 0.1971,
+      "step": 719
+    },
+    {
+      "epoch": 3.257918552036199,
+      "grad_norm": 0.3708641231060028,
+      "learning_rate": 8.525747856761799e-06,
+      "loss": 0.1673,
+      "step": 720
+    },
+    {
+      "epoch": 3.262443438914027,
+      "grad_norm": 0.37453895807266235,
+      "learning_rate": 8.520143729473312e-06,
+      "loss": 0.2085,
+      "step": 721
+    },
+    {
+      "epoch": 3.266968325791855,
+      "grad_norm": 0.44639456272125244,
+      "learning_rate": 8.514530820244427e-06,
+      "loss": 0.2165,
+      "step": 722
+    },
+    {
+      "epoch": 3.271493212669683,
+      "grad_norm": 0.3881756365299225,
+      "learning_rate": 8.50890914307805e-06,
+      "loss": 0.1891,
+      "step": 723
+    },
+    {
+      "epoch": 3.276018099547511,
+      "grad_norm": 0.42670780420303345,
+      "learning_rate": 8.503278711998958e-06,
+      "loss": 0.2126,
+      "step": 724
+    },
+    {
+      "epoch": 3.2805429864253393,
+      "grad_norm": 0.4481564164161682,
+      "learning_rate": 8.497639541053769e-06,
+      "loss": 0.2376,
+      "step": 725
+    },
+    {
+      "epoch": 3.2850678733031673,
+      "grad_norm": 0.4192044138908386,
+      "learning_rate": 8.491991644310903e-06,
+      "loss": 0.2133,
+      "step": 726
+    },
+    {
+      "epoch": 3.2895927601809953,
+      "grad_norm": 0.4070059657096863,
+      "learning_rate": 8.48633503586055e-06,
+      "loss": 0.1873,
+      "step": 727
+    },
+    {
+      "epoch": 3.2941176470588234,
+      "grad_norm": 0.3769683539867401,
+      "learning_rate": 8.480669729814635e-06,
+      "loss": 0.2125,
+      "step": 728
+    },
+    {
+      "epoch": 3.2986425339366514,
+      "grad_norm": 0.45319512486457825,
+      "learning_rate": 8.474995740306775e-06,
+      "loss": 0.2039,
+      "step": 729
+    },
+    {
+      "epoch": 3.3031674208144794,
+      "grad_norm": 0.38497766852378845,
+      "learning_rate": 8.469313081492258e-06,
+      "loss": 0.2126,
+      "step": 730
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.3879552483558655,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.2104,
+      "step": 731
+    },
+    {
+      "epoch": 3.3122171945701355,
+      "grad_norm": 0.4097852110862732,
+      "learning_rate": 8.457921812672498e-06,
+      "loss": 0.1755,
+      "step": 732
+    },
+    {
+      "epoch": 3.3167420814479636,
+      "grad_norm": 0.4118267595767975,
+      "learning_rate": 8.452213231085823e-06,
+      "loss": 0.1905,
+      "step": 733
+    },
+    {
+      "epoch": 3.321266968325792,
+      "grad_norm": 0.42877197265625,
+      "learning_rate": 8.446496037029555e-06,
+      "loss": 0.2186,
+      "step": 734
+    },
+    {
+      "epoch": 3.32579185520362,
+      "grad_norm": 0.4173864424228668,
+      "learning_rate": 8.440770244766764e-06,
+      "loss": 0.1567,
+      "step": 735
+    },
+    {
+      "epoch": 3.330316742081448,
+      "grad_norm": 0.4193741977214813,
+      "learning_rate": 8.435035868581974e-06,
+      "loss": 0.2107,
+      "step": 736
+    },
+    {
+      "epoch": 3.334841628959276,
+      "grad_norm": 0.44753381609916687,
+      "learning_rate": 8.429292922781115e-06,
+      "loss": 0.2011,
+      "step": 737
+    },
+    {
+      "epoch": 3.339366515837104,
+      "grad_norm": 0.3999078869819641,
+      "learning_rate": 8.42354142169151e-06,
+      "loss": 0.2168,
+      "step": 738
+    },
+    {
+      "epoch": 3.3438914027149322,
+      "grad_norm": 0.3881414532661438,
+      "learning_rate": 8.417781379661811e-06,
+      "loss": 0.2272,
+      "step": 739
+    },
+    {
+      "epoch": 3.3484162895927603,
+      "grad_norm": 0.3872552216053009,
+      "learning_rate": 8.412012811061985e-06,
+      "loss": 0.2018,
+      "step": 740
+    },
+    {
+      "epoch": 3.3529411764705883,
+      "grad_norm": 0.4065379798412323,
+      "learning_rate": 8.40623573028327e-06,
+      "loss": 0.1935,
+      "step": 741
+    },
+    {
+      "epoch": 3.3574660633484164,
+      "grad_norm": 0.3988831639289856,
+      "learning_rate": 8.400450151738143e-06,
+      "loss": 0.2169,
+      "step": 742
+    },
+    {
+      "epoch": 3.3619909502262444,
+      "grad_norm": 0.3781244456768036,
+      "learning_rate": 8.394656089860274e-06,
+      "loss": 0.1622,
+      "step": 743
+    },
+    {
+      "epoch": 3.3665158371040724,
+      "grad_norm": 0.4144964814186096,
+      "learning_rate": 8.388853559104502e-06,
+      "loss": 0.2296,
+      "step": 744
+    },
+    {
+      "epoch": 3.3710407239819005,
+      "grad_norm": 0.38190963864326477,
+      "learning_rate": 8.383042573946792e-06,
+      "loss": 0.209,
+      "step": 745
+    },
+    {
+      "epoch": 3.3755656108597285,
+      "grad_norm": 0.4152965247631073,
+      "learning_rate": 8.377223148884202e-06,
+      "loss": 0.211,
+      "step": 746
+    },
+    {
+      "epoch": 3.3800904977375565,
+      "grad_norm": 0.4132832884788513,
+      "learning_rate": 8.371395298434844e-06,
+      "loss": 0.2005,
+      "step": 747
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.37585464119911194,
+      "learning_rate": 8.36555903713785e-06,
+      "loss": 0.1745,
+      "step": 748
+    },
+    {
+      "epoch": 3.3891402714932126,
+      "grad_norm": 0.40911996364593506,
+      "learning_rate": 8.359714379553338e-06,
+      "loss": 0.2016,
+      "step": 749
+    },
+    {
+      "epoch": 3.3936651583710407,
+      "grad_norm": 0.5231017470359802,
+      "learning_rate": 8.353861340262367e-06,
+      "loss": 0.2217,
+      "step": 750
+    },
+    {
+      "epoch": 3.3981900452488687,
+      "grad_norm": 0.4193516671657562,
+      "learning_rate": 8.347999933866912e-06,
+      "loss": 0.2042,
+      "step": 751
+    },
+    {
+      "epoch": 3.4027149321266967,
+      "grad_norm": 0.40086203813552856,
+      "learning_rate": 8.342130174989819e-06,
+      "loss": 0.1837,
+      "step": 752
+    },
+    {
+      "epoch": 3.4072398190045248,
+      "grad_norm": 0.44360044598579407,
+      "learning_rate": 8.33625207827477e-06,
+      "loss": 0.1675,
+      "step": 753
+    },
+    {
+      "epoch": 3.411764705882353,
+      "grad_norm": 0.3777754604816437,
+      "learning_rate": 8.330365658386252e-06,
+      "loss": 0.1793,
+      "step": 754
+    },
+    {
+      "epoch": 3.416289592760181,
+      "grad_norm": 0.42608344554901123,
+      "learning_rate": 8.324470930009514e-06,
+      "loss": 0.2048,
+      "step": 755
+    },
+    {
+      "epoch": 3.420814479638009,
+      "grad_norm": 0.46820250153541565,
+      "learning_rate": 8.318567907850533e-06,
+      "loss": 0.1656,
+      "step": 756
+    },
+    {
+      "epoch": 3.425339366515837,
+      "grad_norm": 0.3514660596847534,
+      "learning_rate": 8.312656606635976e-06,
+      "loss": 0.2045,
+      "step": 757
+    },
+    {
+      "epoch": 3.4298642533936654,
+      "grad_norm": 0.38363519310951233,
+      "learning_rate": 8.306737041113169e-06,
+      "loss": 0.2201,
+      "step": 758
+    },
+    {
+      "epoch": 3.4343891402714934,
+      "grad_norm": 0.39812731742858887,
+      "learning_rate": 8.300809226050049e-06,
+      "loss": 0.1857,
+      "step": 759
+    },
+    {
+      "epoch": 3.4389140271493215,
+      "grad_norm": 0.4550320506095886,
+      "learning_rate": 8.294873176235137e-06,
+      "loss": 0.2109,
+      "step": 760
+    },
+    {
+      "epoch": 3.4434389140271495,
+      "grad_norm": 0.34271740913391113,
+      "learning_rate": 8.288928906477497e-06,
+      "loss": 0.1732,
+      "step": 761
+    },
+    {
+      "epoch": 3.4479638009049776,
+      "grad_norm": 0.41642773151397705,
+      "learning_rate": 8.282976431606703e-06,
+      "loss": 0.2115,
+      "step": 762
+    },
+    {
+      "epoch": 3.4524886877828056,
+      "grad_norm": 0.40923258662223816,
+      "learning_rate": 8.277015766472794e-06,
+      "loss": 0.1746,
+      "step": 763
+    },
+    {
+      "epoch": 3.4570135746606336,
+      "grad_norm": 0.40351900458335876,
+      "learning_rate": 8.271046925946247e-06,
+      "loss": 0.2355,
+      "step": 764
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.4274236261844635,
+      "learning_rate": 8.265069924917925e-06,
+      "loss": 0.1954,
+      "step": 765
+    },
+    {
+      "epoch": 3.4660633484162897,
+      "grad_norm": 0.4109535217285156,
+      "learning_rate": 8.259084778299064e-06,
+      "loss": 0.2156,
+      "step": 766
+    },
+    {
+      "epoch": 3.4705882352941178,
+      "grad_norm": 0.4322621524333954,
+      "learning_rate": 8.25309150102121e-06,
+      "loss": 0.2329,
+      "step": 767
+    },
+    {
+      "epoch": 3.475113122171946,
+      "grad_norm": 0.42734619975090027,
+      "learning_rate": 8.247090108036198e-06,
+      "loss": 0.2,
+      "step": 768
+    },
+    {
+      "epoch": 3.479638009049774,
+      "grad_norm": 0.41158321499824524,
+      "learning_rate": 8.241080614316108e-06,
+      "loss": 0.1893,
+      "step": 769
+    },
+    {
+      "epoch": 3.484162895927602,
+      "grad_norm": 0.39836597442626953,
+      "learning_rate": 8.235063034853228e-06,
+      "loss": 0.2268,
+      "step": 770
+    },
+    {
+      "epoch": 3.48868778280543,
+      "grad_norm": 0.4076845943927765,
+      "learning_rate": 8.229037384660023e-06,
+      "loss": 0.2043,
+      "step": 771
+    },
+    {
+      "epoch": 3.493212669683258,
+      "grad_norm": 0.40793871879577637,
+      "learning_rate": 8.223003678769089e-06,
+      "loss": 0.2147,
+      "step": 772
+    },
+    {
+      "epoch": 3.497737556561086,
+      "grad_norm": 0.4273108243942261,
+      "learning_rate": 8.216961932233118e-06,
+      "loss": 0.2211,
+      "step": 773
+    },
+    {
+      "epoch": 3.502262443438914,
+      "grad_norm": 0.4290085434913635,
+      "learning_rate": 8.210912160124866e-06,
+      "loss": 0.1792,
+      "step": 774
+    },
+    {
+      "epoch": 3.506787330316742,
+      "grad_norm": 0.4068942964076996,
+      "learning_rate": 8.204854377537105e-06,
+      "loss": 0.1925,
+      "step": 775
+    },
+    {
+      "epoch": 3.51131221719457,
+      "grad_norm": 0.4647829830646515,
+      "learning_rate": 8.198788599582596e-06,
+      "loss": 0.2074,
+      "step": 776
+    },
+    {
+      "epoch": 3.515837104072398,
+      "grad_norm": 0.4103906452655792,
+      "learning_rate": 8.192714841394048e-06,
+      "loss": 0.2492,
+      "step": 777
+    },
+    {
+      "epoch": 3.520361990950226,
+      "grad_norm": 0.3793966472148895,
+      "learning_rate": 8.18663311812407e-06,
+      "loss": 0.2051,
+      "step": 778
+    },
+    {
+      "epoch": 3.524886877828054,
+      "grad_norm": 0.37239229679107666,
+      "learning_rate": 8.180543444945154e-06,
+      "loss": 0.1596,
+      "step": 779
+    },
+    {
+      "epoch": 3.5294117647058822,
+      "grad_norm": 0.43737536668777466,
+      "learning_rate": 8.174445837049614e-06,
+      "loss": 0.1427,
+      "step": 780
+    },
+    {
+      "epoch": 3.5339366515837103,
+      "grad_norm": 0.41861864924430847,
+      "learning_rate": 8.168340309649569e-06,
+      "loss": 0.1594,
+      "step": 781
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.3884759247303009,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.2057,
+      "step": 782
+    },
+    {
+      "epoch": 3.5429864253393664,
+      "grad_norm": 0.38496264815330505,
+      "learning_rate": 8.156105557283161e-06,
+      "loss": 0.1997,
+      "step": 783
+    },
+    {
+      "epoch": 3.5475113122171944,
+      "grad_norm": 0.3633590638637543,
+      "learning_rate": 8.149976362839662e-06,
+      "loss": 0.2228,
+      "step": 784
+    },
+    {
+      "epoch": 3.5520361990950224,
+      "grad_norm": 0.3853006362915039,
+      "learning_rate": 8.143839309937307e-06,
+      "loss": 0.2287,
+      "step": 785
+    },
+    {
+      "epoch": 3.5565610859728505,
+      "grad_norm": 0.41965270042419434,
+      "learning_rate": 8.137694413886617e-06,
+      "loss": 0.21,
+      "step": 786
+    },
+    {
+      "epoch": 3.5610859728506785,
+      "grad_norm": 0.45740389823913574,
+      "learning_rate": 8.131541690017676e-06,
+      "loss": 0.1703,
+      "step": 787
+    },
+    {
+      "epoch": 3.5656108597285066,
+      "grad_norm": 0.38977181911468506,
+      "learning_rate": 8.125381153680103e-06,
+      "loss": 0.1868,
+      "step": 788
+    },
+    {
+      "epoch": 3.5701357466063346,
+      "grad_norm": 0.7932807207107544,
+      "learning_rate": 8.119212820243002e-06,
+      "loss": 0.2025,
+      "step": 789
+    },
+    {
+      "epoch": 3.5746606334841626,
+      "grad_norm": 0.40560823678970337,
+      "learning_rate": 8.113036705094935e-06,
+      "loss": 0.2107,
+      "step": 790
+    },
+    {
+      "epoch": 3.579185520361991,
+      "grad_norm": 0.3959384262561798,
+      "learning_rate": 8.10685282364387e-06,
+      "loss": 0.2358,
+      "step": 791
+    },
+    {
+      "epoch": 3.583710407239819,
+      "grad_norm": 0.4452955424785614,
+      "learning_rate": 8.100661191317153e-06,
+      "loss": 0.2314,
+      "step": 792
+    },
+    {
+      "epoch": 3.588235294117647,
+      "grad_norm": 0.3842819929122925,
+      "learning_rate": 8.094461823561473e-06,
+      "loss": 0.233,
+      "step": 793
+    },
+    {
+      "epoch": 3.5927601809954752,
+      "grad_norm": 0.45402851700782776,
+      "learning_rate": 8.088254735842808e-06,
+      "loss": 0.1525,
+      "step": 794
+    },
+    {
+      "epoch": 3.5972850678733033,
+      "grad_norm": 0.4239247739315033,
+      "learning_rate": 8.082039943646398e-06,
+      "loss": 0.1896,
+      "step": 795
+    },
+    {
+      "epoch": 3.6018099547511313,
+      "grad_norm": 0.42828717827796936,
+      "learning_rate": 8.075817462476706e-06,
+      "loss": 0.2204,
+      "step": 796
+    },
+    {
+      "epoch": 3.6063348416289593,
+      "grad_norm": 0.39242568612098694,
+      "learning_rate": 8.069587307857377e-06,
+      "loss": 0.2161,
+      "step": 797
+    },
+    {
+      "epoch": 3.6108597285067874,
+      "grad_norm": 0.4356323480606079,
+      "learning_rate": 8.063349495331197e-06,
+      "loss": 0.2312,
+      "step": 798
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.42608317732810974,
+      "learning_rate": 8.057104040460062e-06,
+      "loss": 0.1939,
+      "step": 799
+    },
+    {
+      "epoch": 3.6199095022624435,
+      "grad_norm": 0.3582107126712799,
+      "learning_rate": 8.050850958824926e-06,
+      "loss": 0.1914,
+      "step": 800
+    },
+    {
+      "epoch": 3.6244343891402715,
+      "grad_norm": 0.396219938993454,
+      "learning_rate": 8.044590266025776e-06,
+      "loss": 0.141,
+      "step": 801
+    },
+    {
+      "epoch": 3.6289592760180995,
+      "grad_norm": 0.38731884956359863,
+      "learning_rate": 8.038321977681586e-06,
+      "loss": 0.2041,
+      "step": 802
+    },
+    {
+      "epoch": 3.6334841628959276,
+      "grad_norm": 0.4682024121284485,
+      "learning_rate": 8.032046109430276e-06,
+      "loss": 0.2546,
+      "step": 803
+    },
+    {
+      "epoch": 3.6380090497737556,
+      "grad_norm": 0.44411900639533997,
+      "learning_rate": 8.025762676928682e-06,
+      "loss": 0.2448,
+      "step": 804
+    },
+    {
+      "epoch": 3.6425339366515836,
+      "grad_norm": 0.3928527534008026,
+      "learning_rate": 8.019471695852505e-06,
+      "loss": 0.2319,
+      "step": 805
+    },
+    {
+      "epoch": 3.6470588235294117,
+      "grad_norm": 0.41334059834480286,
+      "learning_rate": 8.013173181896283e-06,
+      "loss": 0.1934,
+      "step": 806
+    },
+    {
+      "epoch": 3.6515837104072397,
+      "grad_norm": 0.4208468198776245,
+      "learning_rate": 8.006867150773341e-06,
+      "loss": 0.2124,
+      "step": 807
+    },
+    {
+      "epoch": 3.6561085972850678,
+      "grad_norm": 0.38747453689575195,
+      "learning_rate": 8.000553618215762e-06,
+      "loss": 0.2031,
+      "step": 808
+    },
+    {
+      "epoch": 3.660633484162896,
+      "grad_norm": 0.3957184851169586,
+      "learning_rate": 7.994232599974346e-06,
+      "loss": 0.2107,
+      "step": 809
+    },
+    {
+      "epoch": 3.665158371040724,
+      "grad_norm": 0.37880805134773254,
+      "learning_rate": 7.987904111818559e-06,
+      "loss": 0.2105,
+      "step": 810
+    },
+    {
+      "epoch": 3.669683257918552,
+      "grad_norm": 0.3965386152267456,
+      "learning_rate": 7.981568169536512e-06,
+      "loss": 0.1898,
+      "step": 811
+    },
+    {
+      "epoch": 3.6742081447963804,
+      "grad_norm": 0.3958094120025635,
+      "learning_rate": 7.975224788934903e-06,
+      "loss": 0.206,
+      "step": 812
+    },
+    {
+      "epoch": 3.6787330316742084,
+      "grad_norm": 0.3693002164363861,
+      "learning_rate": 7.968873985838998e-06,
+      "loss": 0.1915,
+      "step": 813
+    },
+    {
+      "epoch": 3.6832579185520364,
+      "grad_norm": 0.4172878563404083,
+      "learning_rate": 7.96251577609257e-06,
+      "loss": 0.2276,
+      "step": 814
+    },
+    {
+      "epoch": 3.6877828054298645,
+      "grad_norm": 0.3927733898162842,
+      "learning_rate": 7.95615017555788e-06,
+      "loss": 0.1971,
+      "step": 815
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.417679101228714,
+      "learning_rate": 7.949777200115617e-06,
+      "loss": 0.2045,
+      "step": 816
+    },
+    {
+      "epoch": 3.6968325791855206,
+      "grad_norm": 0.394901841878891,
+      "learning_rate": 7.943396865664872e-06,
+      "loss": 0.2073,
+      "step": 817
+    },
+    {
+      "epoch": 3.7013574660633486,
+      "grad_norm": 0.41202467679977417,
+      "learning_rate": 7.937009188123102e-06,
+      "loss": 0.2414,
+      "step": 818
+    },
+    {
+      "epoch": 3.7058823529411766,
+      "grad_norm": 0.3696572184562683,
+      "learning_rate": 7.930614183426074e-06,
+      "loss": 0.2078,
+      "step": 819
+    },
+    {
+      "epoch": 3.7104072398190047,
+      "grad_norm": 0.4119456112384796,
+      "learning_rate": 7.92421186752784e-06,
+      "loss": 0.2133,
+      "step": 820
+    },
+    {
+      "epoch": 3.7149321266968327,
+      "grad_norm": 0.48625221848487854,
+      "learning_rate": 7.917802256400688e-06,
+      "loss": 0.2036,
+      "step": 821
+    },
+    {
+      "epoch": 3.7194570135746607,
+      "grad_norm": 0.43132078647613525,
+      "learning_rate": 7.911385366035111e-06,
+      "loss": 0.2037,
+      "step": 822
+    },
+    {
+      "epoch": 3.723981900452489,
+      "grad_norm": 0.44492626190185547,
+      "learning_rate": 7.904961212439756e-06,
+      "loss": 0.235,
+      "step": 823
+    },
+    {
+      "epoch": 3.728506787330317,
+      "grad_norm": 0.37732723355293274,
+      "learning_rate": 7.898529811641393e-06,
+      "loss": 0.2191,
+      "step": 824
+    },
+    {
+      "epoch": 3.733031674208145,
+      "grad_norm": 0.3524717688560486,
+      "learning_rate": 7.892091179684876e-06,
+      "loss": 0.215,
+      "step": 825
+    },
+    {
+      "epoch": 3.737556561085973,
+      "grad_norm": 0.4359301030635834,
+      "learning_rate": 7.88564533263309e-06,
+      "loss": 0.2027,
+      "step": 826
+    },
+    {
+      "epoch": 3.742081447963801,
+      "grad_norm": 0.36626332998275757,
+      "learning_rate": 7.879192286566929e-06,
+      "loss": 0.1884,
+      "step": 827
+    },
+    {
+      "epoch": 3.746606334841629,
+      "grad_norm": 0.36521342396736145,
+      "learning_rate": 7.87273205758524e-06,
+      "loss": 0.1919,
+      "step": 828
+    },
+    {
+      "epoch": 3.751131221719457,
+      "grad_norm": 0.38231372833251953,
+      "learning_rate": 7.866264661804795e-06,
+      "loss": 0.2165,
+      "step": 829
+    },
+    {
+      "epoch": 3.755656108597285,
+      "grad_norm": 0.5153652429580688,
+      "learning_rate": 7.859790115360243e-06,
+      "loss": 0.2336,
+      "step": 830
+    },
+    {
+      "epoch": 3.760180995475113,
+      "grad_norm": 0.3765285909175873,
+      "learning_rate": 7.853308434404071e-06,
+      "loss": 0.1948,
+      "step": 831
+    },
+    {
+      "epoch": 3.764705882352941,
+      "grad_norm": 0.40210476517677307,
+      "learning_rate": 7.846819635106569e-06,
+      "loss": 0.2053,
+      "step": 832
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.3615150451660156,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.2148,
+      "step": 833
+    },
+    {
+      "epoch": 3.773755656108597,
+      "grad_norm": 0.3671107590198517,
+      "learning_rate": 7.833820746257471e-06,
+      "loss": 0.1812,
+      "step": 834
+    },
+    {
+      "epoch": 3.7782805429864252,
+      "grad_norm": 0.48968544602394104,
+      "learning_rate": 7.82731068913508e-06,
+      "loss": 0.2532,
+      "step": 835
+    },
+    {
+      "epoch": 3.7828054298642533,
+      "grad_norm": 0.378062903881073,
+      "learning_rate": 7.82079357852969e-06,
+      "loss": 0.2263,
+      "step": 836
+    },
+    {
+      "epoch": 3.7873303167420813,
+      "grad_norm": 0.3558998107910156,
+      "learning_rate": 7.814269430699978e-06,
+      "loss": 0.2033,
+      "step": 837
+    },
+    {
+      "epoch": 3.7918552036199094,
+      "grad_norm": 0.4033679664134979,
+      "learning_rate": 7.807738261922173e-06,
+      "loss": 0.2005,
+      "step": 838
+    },
+    {
+      "epoch": 3.7963800904977374,
+      "grad_norm": 0.40575459599494934,
+      "learning_rate": 7.801200088490026e-06,
+      "loss": 0.2269,
+      "step": 839
+    },
+    {
+      "epoch": 3.8009049773755654,
+      "grad_norm": 0.42206406593322754,
+      "learning_rate": 7.794654926714755e-06,
+      "loss": 0.2176,
+      "step": 840
+    },
+    {
+      "epoch": 3.8054298642533935,
+      "grad_norm": 0.4195837080478668,
+      "learning_rate": 7.788102792925026e-06,
+      "loss": 0.1968,
+      "step": 841
+    },
+    {
+      "epoch": 3.8099547511312215,
+      "grad_norm": 0.3960534930229187,
+      "learning_rate": 7.781543703466881e-06,
+      "loss": 0.183,
+      "step": 842
+    },
+    {
+      "epoch": 3.8144796380090495,
+      "grad_norm": 0.3833674192428589,
+      "learning_rate": 7.774977674703728e-06,
+      "loss": 0.2106,
+      "step": 843
+    },
+    {
+      "epoch": 3.8190045248868776,
+      "grad_norm": 0.4212116301059723,
+      "learning_rate": 7.768404723016281e-06,
+      "loss": 0.1978,
+      "step": 844
+    },
+    {
+      "epoch": 3.8235294117647056,
+      "grad_norm": 0.38558632135391235,
+      "learning_rate": 7.76182486480253e-06,
+      "loss": 0.164,
+      "step": 845
+    },
+    {
+      "epoch": 3.8280542986425337,
+      "grad_norm": 0.4841013252735138,
+      "learning_rate": 7.755238116477688e-06,
+      "loss": 0.2152,
+      "step": 846
+    },
+    {
+      "epoch": 3.832579185520362,
+      "grad_norm": 0.41625547409057617,
+      "learning_rate": 7.748644494474163e-06,
+      "loss": 0.2322,
+      "step": 847
+    },
+    {
+      "epoch": 3.83710407239819,
+      "grad_norm": 0.3868163228034973,
+      "learning_rate": 7.742044015241508e-06,
+      "loss": 0.1888,
+      "step": 848
+    },
+    {
+      "epoch": 3.841628959276018,
+      "grad_norm": 0.3938911259174347,
+      "learning_rate": 7.735436695246385e-06,
+      "loss": 0.209,
+      "step": 849
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.4158743619918823,
+      "learning_rate": 7.728822550972523e-06,
+      "loss": 0.1966,
+      "step": 850
+    },
+    {
+      "epoch": 3.8506787330316743,
+      "grad_norm": 0.46179747581481934,
+      "learning_rate": 7.722201598920673e-06,
+      "loss": 0.2185,
+      "step": 851
+    },
+    {
+      "epoch": 3.8552036199095023,
+      "grad_norm": 0.3782907724380493,
+      "learning_rate": 7.715573855608574e-06,
+      "loss": 0.1929,
+      "step": 852
+    },
+    {
+      "epoch": 3.8597285067873304,
+      "grad_norm": 0.3639270067214966,
+      "learning_rate": 7.7089393375709e-06,
+      "loss": 0.2075,
+      "step": 853
+    },
+    {
+      "epoch": 3.8642533936651584,
+      "grad_norm": 0.3893643021583557,
+      "learning_rate": 7.702298061359236e-06,
+      "loss": 0.2073,
+      "step": 854
+    },
+    {
+      "epoch": 3.8687782805429864,
+      "grad_norm": 0.41035643219947815,
+      "learning_rate": 7.69565004354202e-06,
+      "loss": 0.2165,
+      "step": 855
+    },
+    {
+      "epoch": 3.8733031674208145,
+      "grad_norm": 0.3843505382537842,
+      "learning_rate": 7.688995300704511e-06,
+      "loss": 0.1806,
+      "step": 856
+    },
+    {
+      "epoch": 3.8778280542986425,
+      "grad_norm": 0.4357762038707733,
+      "learning_rate": 7.682333849448749e-06,
+      "loss": 0.2137,
+      "step": 857
+    },
+    {
+      "epoch": 3.8823529411764706,
+      "grad_norm": 0.424322247505188,
+      "learning_rate": 7.675665706393502e-06,
+      "loss": 0.2281,
+      "step": 858
+    },
+    {
+      "epoch": 3.8868778280542986,
+      "grad_norm": 0.46023404598236084,
+      "learning_rate": 7.668990888174242e-06,
+      "loss": 0.1564,
+      "step": 859
+    },
+    {
+      "epoch": 3.8914027149321266,
+      "grad_norm": 0.4409455358982086,
+      "learning_rate": 7.662309411443084e-06,
+      "loss": 0.2128,
+      "step": 860
+    },
+    {
+      "epoch": 3.8959276018099547,
+      "grad_norm": 0.4324515163898468,
+      "learning_rate": 7.65562129286876e-06,
+      "loss": 0.2044,
+      "step": 861
+    },
+    {
+      "epoch": 3.9004524886877827,
+      "grad_norm": 0.41723108291625977,
+      "learning_rate": 7.648926549136574e-06,
+      "loss": 0.1992,
+      "step": 862
+    },
+    {
+      "epoch": 3.9049773755656108,
+      "grad_norm": 0.40275150537490845,
+      "learning_rate": 7.642225196948357e-06,
+      "loss": 0.222,
+      "step": 863
+    },
+    {
+      "epoch": 3.909502262443439,
+      "grad_norm": 0.3896505832672119,
+      "learning_rate": 7.635517253022419e-06,
+      "loss": 0.2029,
+      "step": 864
+    },
+    {
+      "epoch": 3.914027149321267,
+      "grad_norm": 0.41698864102363586,
+      "learning_rate": 7.628802734093528e-06,
+      "loss": 0.1867,
+      "step": 865
+    },
+    {
+      "epoch": 3.918552036199095,
+      "grad_norm": 0.3895374536514282,
+      "learning_rate": 7.622081656912842e-06,
+      "loss": 0.2302,
+      "step": 866
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.42431965470314026,
+      "learning_rate": 7.615354038247889e-06,
+      "loss": 0.2197,
+      "step": 867
+    },
+    {
+      "epoch": 3.9276018099547514,
+      "grad_norm": 0.4233725965023041,
+      "learning_rate": 7.608619894882514e-06,
+      "loss": 0.1918,
+      "step": 868
+    },
+    {
+      "epoch": 3.9321266968325794,
+      "grad_norm": 0.39451637864112854,
+      "learning_rate": 7.601879243616838e-06,
+      "loss": 0.1807,
+      "step": 869
+    },
+    {
+      "epoch": 3.9366515837104075,
+      "grad_norm": 0.42328158020973206,
+      "learning_rate": 7.595132101267221e-06,
+      "loss": 0.179,
+      "step": 870
+    },
+    {
+      "epoch": 3.9411764705882355,
+      "grad_norm": 0.49399158358573914,
+      "learning_rate": 7.588378484666214e-06,
+      "loss": 0.2318,
+      "step": 871
+    },
+    {
+      "epoch": 3.9457013574660635,
+      "grad_norm": 0.4472160041332245,
+      "learning_rate": 7.581618410662519e-06,
+      "loss": 0.1861,
+      "step": 872
+    },
+    {
+      "epoch": 3.9502262443438916,
+      "grad_norm": 0.4840329587459564,
+      "learning_rate": 7.574851896120951e-06,
+      "loss": 0.2355,
+      "step": 873
+    },
+    {
+      "epoch": 3.9547511312217196,
+      "grad_norm": 0.40697556734085083,
+      "learning_rate": 7.568078957922394e-06,
+      "loss": 0.1865,
+      "step": 874
+    },
+    {
+      "epoch": 3.9592760180995477,
+      "grad_norm": 0.4209546744823456,
+      "learning_rate": 7.56129961296375e-06,
+      "loss": 0.231,
+      "step": 875
+    },
+    {
+      "epoch": 3.9638009049773757,
+      "grad_norm": 0.36913007497787476,
+      "learning_rate": 7.55451387815791e-06,
+      "loss": 0.1939,
+      "step": 876
+    },
+    {
+      "epoch": 3.9683257918552037,
+      "grad_norm": 0.45089685916900635,
+      "learning_rate": 7.547721770433706e-06,
+      "loss": 0.2428,
+      "step": 877
+    },
+    {
+      "epoch": 3.9728506787330318,
+      "grad_norm": 0.39802390336990356,
+      "learning_rate": 7.540923306735868e-06,
+      "loss": 0.1777,
+      "step": 878
+    },
+    {
+      "epoch": 3.97737556561086,
+      "grad_norm": 0.4216815233230591,
+      "learning_rate": 7.534118504024981e-06,
+      "loss": 0.2052,
+      "step": 879
+    },
+    {
+      "epoch": 3.981900452488688,
+      "grad_norm": 0.5024846792221069,
+      "learning_rate": 7.527307379277448e-06,
+      "loss": 0.2161,
+      "step": 880
+    },
+    {
+      "epoch": 3.986425339366516,
+      "grad_norm": 0.3847920000553131,
+      "learning_rate": 7.5204899494854415e-06,
+      "loss": 0.1682,
+      "step": 881
+    },
+    {
+      "epoch": 3.990950226244344,
+      "grad_norm": 0.41892680525779724,
+      "learning_rate": 7.5136662316568645e-06,
+      "loss": 0.2123,
+      "step": 882
+    },
+    {
+      "epoch": 3.995475113122172,
+      "grad_norm": 0.40700384974479675,
+      "learning_rate": 7.5068362428153055e-06,
+      "loss": 0.2115,
+      "step": 883
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.40372100472450256,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1805,
+      "step": 884
+    },
+    {
+      "epoch": 4.004524886877828,
+      "grad_norm": 0.4641391932964325,
+      "learning_rate": 7.493157520265787e-06,
+      "loss": 0.1209,
+      "step": 885
+    },
+    {
+      "epoch": 4.009049773755656,
+      "grad_norm": 0.47821393609046936,
+      "learning_rate": 7.48630882068306e-06,
+      "loss": 0.1517,
+      "step": 886
+    },
+    {
+      "epoch": 4.013574660633484,
+      "grad_norm": 0.3496209979057312,
+      "learning_rate": 7.479453918337733e-06,
+      "loss": 0.0967,
+      "step": 887
+    },
+    {
+      "epoch": 4.018099547511312,
+      "grad_norm": 0.5180306434631348,
+      "learning_rate": 7.472592830331196e-06,
+      "loss": 0.1436,
+      "step": 888
+    },
+    {
+      "epoch": 4.02262443438914,
+      "grad_norm": 0.6377127766609192,
+      "learning_rate": 7.465725573780267e-06,
+      "loss": 0.1437,
+      "step": 889
+    },
+    {
+      "epoch": 4.027149321266968,
+      "grad_norm": 0.4262060225009918,
+      "learning_rate": 7.458852165817153e-06,
+      "loss": 0.1164,
+      "step": 890
+    },
+    {
+      "epoch": 4.031674208144796,
+      "grad_norm": 0.46363940834999084,
+      "learning_rate": 7.451972623589414e-06,
+      "loss": 0.1251,
+      "step": 891
+    },
+    {
+      "epoch": 4.036199095022624,
+      "grad_norm": 0.38100776076316833,
+      "learning_rate": 7.445086964259905e-06,
+      "loss": 0.1282,
+      "step": 892
+    },
+    {
+      "epoch": 4.040723981900452,
+      "grad_norm": 0.37024420499801636,
+      "learning_rate": 7.438195205006749e-06,
+      "loss": 0.121,
+      "step": 893
+    },
+    {
+      "epoch": 4.04524886877828,
+      "grad_norm": 0.3664066791534424,
+      "learning_rate": 7.43129736302328e-06,
+      "loss": 0.0924,
+      "step": 894
+    },
+    {
+      "epoch": 4.049773755656108,
+      "grad_norm": 0.4302886426448822,
+      "learning_rate": 7.4243934555180116e-06,
+      "loss": 0.1134,
+      "step": 895
+    },
+    {
+      "epoch": 4.0542986425339365,
+      "grad_norm": 0.3795878291130066,
+      "learning_rate": 7.417483499714589e-06,
+      "loss": 0.1179,
+      "step": 896
+    },
+    {
+      "epoch": 4.0588235294117645,
+      "grad_norm": 0.3966153562068939,
+      "learning_rate": 7.4105675128517456e-06,
+      "loss": 0.1404,
+      "step": 897
+    },
+    {
+      "epoch": 4.0633484162895925,
+      "grad_norm": 0.4186801314353943,
+      "learning_rate": 7.4036455121832595e-06,
+      "loss": 0.1253,
+      "step": 898
+    },
+    {
+      "epoch": 4.067873303167421,
+      "grad_norm": 0.45986229181289673,
+      "learning_rate": 7.396717514977916e-06,
+      "loss": 0.1458,
+      "step": 899
+    },
+    {
+      "epoch": 4.072398190045249,
+      "grad_norm": 0.5305169820785522,
+      "learning_rate": 7.389783538519456e-06,
+      "loss": 0.1608,
+      "step": 900
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.42427879571914673,
+      "learning_rate": 7.382843600106539e-06,
+      "loss": 0.1207,
+      "step": 901
+    },
+    {
+      "epoch": 4.081447963800905,
+      "grad_norm": 0.3799537420272827,
+      "learning_rate": 7.3758977170527e-06,
+      "loss": 0.1235,
+      "step": 902
+    },
+    {
+      "epoch": 4.085972850678733,
+      "grad_norm": 0.3693038821220398,
+      "learning_rate": 7.368945906686303e-06,
+      "loss": 0.1144,
+      "step": 903
+    },
+    {
+      "epoch": 4.090497737556561,
+      "grad_norm": 0.3847089111804962,
+      "learning_rate": 7.361988186350495e-06,
+      "loss": 0.1265,
+      "step": 904
+    },
+    {
+      "epoch": 4.095022624434389,
+      "grad_norm": 0.448414146900177,
+      "learning_rate": 7.355024573403174e-06,
+      "loss": 0.1451,
+      "step": 905
+    },
+    {
+      "epoch": 4.099547511312217,
+      "grad_norm": 0.42488551139831543,
+      "learning_rate": 7.348055085216937e-06,
+      "loss": 0.0954,
+      "step": 906
+    },
+    {
+      "epoch": 4.104072398190045,
+      "grad_norm": 0.45263928174972534,
+      "learning_rate": 7.341079739179033e-06,
+      "loss": 0.143,
+      "step": 907
+    },
+    {
+      "epoch": 4.108597285067873,
+      "grad_norm": 0.40875062346458435,
+      "learning_rate": 7.3340985526913335e-06,
+      "loss": 0.1438,
+      "step": 908
+    },
+    {
+      "epoch": 4.113122171945701,
+      "grad_norm": 0.4271523356437683,
+      "learning_rate": 7.327111543170274e-06,
+      "loss": 0.1377,
+      "step": 909
+    },
+    {
+      "epoch": 4.117647058823529,
+      "grad_norm": 0.38401153683662415,
+      "learning_rate": 7.320118728046818e-06,
+      "loss": 0.1004,
+      "step": 910
+    },
+    {
+      "epoch": 4.122171945701357,
+      "grad_norm": 0.42597001791000366,
+      "learning_rate": 7.313120124766417e-06,
+      "loss": 0.1505,
+      "step": 911
+    },
+    {
+      "epoch": 4.126696832579185,
+      "grad_norm": 0.40540316700935364,
+      "learning_rate": 7.306115750788956e-06,
+      "loss": 0.1144,
+      "step": 912
+    },
+    {
+      "epoch": 4.131221719457014,
+      "grad_norm": 0.3972390592098236,
+      "learning_rate": 7.299105623588722e-06,
+      "loss": 0.1068,
+      "step": 913
+    },
+    {
+      "epoch": 4.135746606334842,
+      "grad_norm": 0.39872103929519653,
+      "learning_rate": 7.292089760654352e-06,
+      "loss": 0.1106,
+      "step": 914
+    },
+    {
+      "epoch": 4.14027149321267,
+      "grad_norm": 0.4083847403526306,
+      "learning_rate": 7.2850681794887946e-06,
+      "loss": 0.1402,
+      "step": 915
+    },
+    {
+      "epoch": 4.144796380090498,
+      "grad_norm": 0.4313944876194,
+      "learning_rate": 7.278040897609262e-06,
+      "loss": 0.1467,
+      "step": 916
+    },
+    {
+      "epoch": 4.149321266968326,
+      "grad_norm": 0.4161936640739441,
+      "learning_rate": 7.271007932547188e-06,
+      "loss": 0.1135,
+      "step": 917
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.385703444480896,
+      "learning_rate": 7.263969301848188e-06,
+      "loss": 0.1019,
+      "step": 918
+    },
+    {
+      "epoch": 4.158371040723982,
+      "grad_norm": 0.41860657930374146,
+      "learning_rate": 7.256925023072008e-06,
+      "loss": 0.129,
+      "step": 919
+    },
+    {
+      "epoch": 4.16289592760181,
+      "grad_norm": 0.4360624849796295,
+      "learning_rate": 7.249875113792485e-06,
+      "loss": 0.1349,
+      "step": 920
+    },
+    {
+      "epoch": 4.167420814479638,
+      "grad_norm": 0.39161208271980286,
+      "learning_rate": 7.242819591597507e-06,
+      "loss": 0.1048,
+      "step": 921
+    },
+    {
+      "epoch": 4.171945701357466,
+      "grad_norm": 0.36276566982269287,
+      "learning_rate": 7.235758474088961e-06,
+      "loss": 0.1253,
+      "step": 922
+    },
+    {
+      "epoch": 4.176470588235294,
+      "grad_norm": 0.414889931678772,
+      "learning_rate": 7.2286917788826926e-06,
+      "loss": 0.1252,
+      "step": 923
+    },
+    {
+      "epoch": 4.180995475113122,
+      "grad_norm": 0.38982105255126953,
+      "learning_rate": 7.221619523608464e-06,
+      "loss": 0.135,
+      "step": 924
+    },
+    {
+      "epoch": 4.1855203619909505,
+      "grad_norm": 0.3900318145751953,
+      "learning_rate": 7.21454172590991e-06,
+      "loss": 0.1232,
+      "step": 925
+    },
+    {
+      "epoch": 4.1900452488687785,
+      "grad_norm": 0.4147709608078003,
+      "learning_rate": 7.207458403444488e-06,
+      "loss": 0.1217,
+      "step": 926
+    },
+    {
+      "epoch": 4.1945701357466065,
+      "grad_norm": 0.43696555495262146,
+      "learning_rate": 7.200369573883444e-06,
+      "loss": 0.1292,
+      "step": 927
+    },
+    {
+      "epoch": 4.199095022624435,
+      "grad_norm": 0.4161735475063324,
+      "learning_rate": 7.193275254911758e-06,
+      "loss": 0.1456,
+      "step": 928
+    },
+    {
+      "epoch": 4.203619909502263,
+      "grad_norm": 0.4888911545276642,
+      "learning_rate": 7.186175464228109e-06,
+      "loss": 0.1266,
+      "step": 929
+    },
+    {
+      "epoch": 4.208144796380091,
+      "grad_norm": 0.39078477025032043,
+      "learning_rate": 7.179070219544822e-06,
+      "loss": 0.1055,
+      "step": 930
+    },
+    {
+      "epoch": 4.212669683257919,
+      "grad_norm": 0.4553499221801758,
+      "learning_rate": 7.171959538587831e-06,
+      "loss": 0.1048,
+      "step": 931
+    },
+    {
+      "epoch": 4.217194570135747,
+      "grad_norm": 0.4161282479763031,
+      "learning_rate": 7.1648434390966356e-06,
+      "loss": 0.1229,
+      "step": 932
+    },
+    {
+      "epoch": 4.221719457013575,
+      "grad_norm": 0.3798969089984894,
+      "learning_rate": 7.157721938824247e-06,
+      "loss": 0.1178,
+      "step": 933
+    },
+    {
+      "epoch": 4.226244343891403,
+      "grad_norm": 0.3456389009952545,
+      "learning_rate": 7.150595055537153e-06,
+      "loss": 0.1244,
+      "step": 934
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.4065212607383728,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.1242,
+      "step": 935
+    },
+    {
+      "epoch": 4.235294117647059,
+      "grad_norm": 0.43408218026161194,
+      "learning_rate": 7.136325211051905e-06,
+      "loss": 0.1324,
+      "step": 936
+    },
+    {
+      "epoch": 4.239819004524887,
+      "grad_norm": 0.4018978774547577,
+      "learning_rate": 7.129182285453697e-06,
+      "loss": 0.1435,
+      "step": 937
+    },
+    {
+      "epoch": 4.244343891402715,
+      "grad_norm": 0.38995468616485596,
+      "learning_rate": 7.122034048040586e-06,
+      "loss": 0.105,
+      "step": 938
+    },
+    {
+      "epoch": 4.248868778280543,
+      "grad_norm": 0.40599551796913147,
+      "learning_rate": 7.114880516645761e-06,
+      "loss": 0.122,
+      "step": 939
+    },
+    {
+      "epoch": 4.253393665158371,
+      "grad_norm": 0.40751129388809204,
+      "learning_rate": 7.107721709115624e-06,
+      "loss": 0.1338,
+      "step": 940
+    },
+    {
+      "epoch": 4.257918552036199,
+      "grad_norm": 0.38541626930236816,
+      "learning_rate": 7.100557643309732e-06,
+      "loss": 0.1121,
+      "step": 941
+    },
+    {
+      "epoch": 4.262443438914027,
+      "grad_norm": 0.38623547554016113,
+      "learning_rate": 7.0933883371007665e-06,
+      "loss": 0.1053,
+      "step": 942
+    },
+    {
+      "epoch": 4.266968325791855,
+      "grad_norm": 0.3839960992336273,
+      "learning_rate": 7.086213808374478e-06,
+      "loss": 0.1153,
+      "step": 943
+    },
+    {
+      "epoch": 4.271493212669683,
+      "grad_norm": 0.35866132378578186,
+      "learning_rate": 7.079034075029651e-06,
+      "loss": 0.1032,
+      "step": 944
+    },
+    {
+      "epoch": 4.276018099547511,
+      "grad_norm": 0.4211701452732086,
+      "learning_rate": 7.071849154978048e-06,
+      "loss": 0.1228,
+      "step": 945
+    },
+    {
+      "epoch": 4.280542986425339,
+      "grad_norm": 0.4204184114933014,
+      "learning_rate": 7.064659066144376e-06,
+      "loss": 0.1409,
+      "step": 946
+    },
+    {
+      "epoch": 4.285067873303167,
+      "grad_norm": 0.39663875102996826,
+      "learning_rate": 7.057463826466235e-06,
+      "loss": 0.1291,
+      "step": 947
+    },
+    {
+      "epoch": 4.289592760180995,
+      "grad_norm": 0.4302012622356415,
+      "learning_rate": 7.050263453894078e-06,
+      "loss": 0.1281,
+      "step": 948
+    },
+    {
+      "epoch": 4.294117647058823,
+      "grad_norm": 0.39377084374427795,
+      "learning_rate": 7.043057966391158e-06,
+      "loss": 0.1471,
+      "step": 949
+    },
+    {
+      "epoch": 4.298642533936651,
+      "grad_norm": 0.4178546369075775,
+      "learning_rate": 7.035847381933494e-06,
+      "loss": 0.1139,
+      "step": 950
+    },
+    {
+      "epoch": 4.3031674208144794,
+      "grad_norm": 0.4286903738975525,
+      "learning_rate": 7.028631718509816e-06,
+      "loss": 0.1389,
+      "step": 951
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.4679263234138489,
+      "learning_rate": 7.021410994121525e-06,
+      "loss": 0.141,
+      "step": 952
+    },
+    {
+      "epoch": 4.3122171945701355,
+      "grad_norm": 0.38693973422050476,
+      "learning_rate": 7.014185226782655e-06,
+      "loss": 0.1114,
+      "step": 953
+    },
+    {
+      "epoch": 4.316742081447964,
+      "grad_norm": 0.514302670955658,
+      "learning_rate": 7.006954434519814e-06,
+      "loss": 0.0925,
+      "step": 954
+    },
+    {
+      "epoch": 4.321266968325792,
+      "grad_norm": 0.36845308542251587,
+      "learning_rate": 6.999718635372148e-06,
+      "loss": 0.1151,
+      "step": 955
+    },
+    {
+      "epoch": 4.32579185520362,
+      "grad_norm": 0.4446454346179962,
+      "learning_rate": 6.992477847391292e-06,
+      "loss": 0.1057,
+      "step": 956
+    },
+    {
+      "epoch": 4.330316742081448,
+      "grad_norm": 0.39890649914741516,
+      "learning_rate": 6.985232088641333e-06,
+      "loss": 0.1156,
+      "step": 957
+    },
+    {
+      "epoch": 4.334841628959276,
+      "grad_norm": 0.4039772152900696,
+      "learning_rate": 6.9779813771987525e-06,
+      "loss": 0.1212,
+      "step": 958
+    },
+    {
+      "epoch": 4.339366515837104,
+      "grad_norm": 0.38077545166015625,
+      "learning_rate": 6.970725731152389e-06,
+      "loss": 0.0997,
+      "step": 959
+    },
+    {
+      "epoch": 4.343891402714932,
+      "grad_norm": 0.44557592272758484,
+      "learning_rate": 6.963465168603395e-06,
+      "loss": 0.1263,
+      "step": 960
+    },
+    {
+      "epoch": 4.34841628959276,
+      "grad_norm": 0.45519551634788513,
+      "learning_rate": 6.9561997076651854e-06,
+      "loss": 0.1457,
+      "step": 961
+    },
+    {
+      "epoch": 4.352941176470588,
+      "grad_norm": 0.37018322944641113,
+      "learning_rate": 6.948929366463397e-06,
+      "loss": 0.0994,
+      "step": 962
+    },
+    {
+      "epoch": 4.357466063348416,
+      "grad_norm": 0.40303707122802734,
+      "learning_rate": 6.941654163135841e-06,
+      "loss": 0.1476,
+      "step": 963
+    },
+    {
+      "epoch": 4.361990950226244,
+      "grad_norm": 0.39125677943229675,
+      "learning_rate": 6.934374115832459e-06,
+      "loss": 0.1516,
+      "step": 964
+    },
+    {
+      "epoch": 4.366515837104072,
+      "grad_norm": 0.3679739534854889,
+      "learning_rate": 6.927089242715277e-06,
+      "loss": 0.1074,
+      "step": 965
+    },
+    {
+      "epoch": 4.371040723981901,
+      "grad_norm": 0.38703691959381104,
+      "learning_rate": 6.919799561958359e-06,
+      "loss": 0.1351,
+      "step": 966
+    },
+    {
+      "epoch": 4.375565610859729,
+      "grad_norm": 0.45326265692710876,
+      "learning_rate": 6.912505091747765e-06,
+      "loss": 0.1347,
+      "step": 967
+    },
+    {
+      "epoch": 4.380090497737557,
+      "grad_norm": 0.39937224984169006,
+      "learning_rate": 6.905205850281502e-06,
+      "loss": 0.1658,
+      "step": 968
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.39014557003974915,
+      "learning_rate": 6.897901855769483e-06,
+      "loss": 0.1213,
+      "step": 969
+    },
+    {
+      "epoch": 4.389140271493213,
+      "grad_norm": 0.41682887077331543,
+      "learning_rate": 6.890593126433474e-06,
+      "loss": 0.1256,
+      "step": 970
+    },
+    {
+      "epoch": 4.393665158371041,
+      "grad_norm": 0.37030866742134094,
+      "learning_rate": 6.883279680507057e-06,
+      "loss": 0.1239,
+      "step": 971
+    },
+    {
+      "epoch": 4.398190045248869,
+      "grad_norm": 0.4141203463077545,
+      "learning_rate": 6.8759615362355795e-06,
+      "loss": 0.1365,
+      "step": 972
+    },
+    {
+      "epoch": 4.402714932126697,
+      "grad_norm": 0.5501455664634705,
+      "learning_rate": 6.868638711876109e-06,
+      "loss": 0.1241,
+      "step": 973
+    },
+    {
+      "epoch": 4.407239819004525,
+      "grad_norm": 0.42243292927742004,
+      "learning_rate": 6.861311225697392e-06,
+      "loss": 0.1347,
+      "step": 974
+    },
+    {
+      "epoch": 4.411764705882353,
+      "grad_norm": 0.43161842226982117,
+      "learning_rate": 6.8539790959798045e-06,
+      "loss": 0.1299,
+      "step": 975
+    },
+    {
+      "epoch": 4.416289592760181,
+      "grad_norm": 0.3907506465911865,
+      "learning_rate": 6.846642341015304e-06,
+      "loss": 0.1083,
+      "step": 976
+    },
+    {
+      "epoch": 4.420814479638009,
+      "grad_norm": 0.39436179399490356,
+      "learning_rate": 6.8393009791073895e-06,
+      "loss": 0.1177,
+      "step": 977
+    },
+    {
+      "epoch": 4.425339366515837,
+      "grad_norm": 0.3935137689113617,
+      "learning_rate": 6.831955028571051e-06,
+      "loss": 0.1141,
+      "step": 978
+    },
+    {
+      "epoch": 4.429864253393665,
+      "grad_norm": 0.3807559609413147,
+      "learning_rate": 6.82460450773273e-06,
+      "loss": 0.1033,
+      "step": 979
+    },
+    {
+      "epoch": 4.4343891402714934,
+      "grad_norm": 0.3852434456348419,
+      "learning_rate": 6.817249434930267e-06,
+      "loss": 0.134,
+      "step": 980
+    },
+    {
+      "epoch": 4.4389140271493215,
+      "grad_norm": 0.4558684825897217,
+      "learning_rate": 6.809889828512857e-06,
+      "loss": 0.1246,
+      "step": 981
+    },
+    {
+      "epoch": 4.4434389140271495,
+      "grad_norm": 0.3894815742969513,
+      "learning_rate": 6.80252570684101e-06,
+      "loss": 0.1376,
+      "step": 982
+    },
+    {
+      "epoch": 4.447963800904978,
+      "grad_norm": 0.43874591588974,
+      "learning_rate": 6.7951570882864944e-06,
+      "loss": 0.1458,
+      "step": 983
+    },
+    {
+      "epoch": 4.452488687782806,
+      "grad_norm": 0.39900916814804077,
+      "learning_rate": 6.787783991232305e-06,
+      "loss": 0.1398,
+      "step": 984
+    },
+    {
+      "epoch": 4.457013574660634,
+      "grad_norm": 0.4251669645309448,
+      "learning_rate": 6.780406434072603e-06,
+      "loss": 0.1445,
+      "step": 985
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.42457953095436096,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.1344,
+      "step": 986
+    },
+    {
+      "epoch": 4.46606334841629,
+      "grad_norm": 0.48800379037857056,
+      "learning_rate": 6.765638013068903e-06,
+      "loss": 0.1183,
+      "step": 987
+    },
+    {
+      "epoch": 4.470588235294118,
+      "grad_norm": 0.3914732336997986,
+      "learning_rate": 6.758247186068684e-06,
+      "loss": 0.1176,
+      "step": 988
+    },
+    {
+      "epoch": 4.475113122171946,
+      "grad_norm": 0.40365973114967346,
+      "learning_rate": 6.750851972650416e-06,
+      "loss": 0.1166,
+      "step": 989
+    },
+    {
+      "epoch": 4.479638009049774,
+      "grad_norm": 0.4439378082752228,
+      "learning_rate": 6.743452391263438e-06,
+      "loss": 0.1109,
+      "step": 990
+    },
+    {
+      "epoch": 4.484162895927602,
+      "grad_norm": 0.365218847990036,
+      "learning_rate": 6.736048460367983e-06,
+      "loss": 0.1292,
+      "step": 991
+    },
+    {
+      "epoch": 4.48868778280543,
+      "grad_norm": 0.37418243288993835,
+      "learning_rate": 6.728640198435143e-06,
+      "loss": 0.1404,
+      "step": 992
+    },
+    {
+      "epoch": 4.493212669683258,
+      "grad_norm": 0.3743608891963959,
+      "learning_rate": 6.721227623946804e-06,
+      "loss": 0.115,
+      "step": 993
+    },
+    {
+      "epoch": 4.497737556561086,
+      "grad_norm": 0.384598046541214,
+      "learning_rate": 6.71381075539562e-06,
+      "loss": 0.1036,
+      "step": 994
+    },
+    {
+      "epoch": 4.502262443438914,
+      "grad_norm": 0.4858543276786804,
+      "learning_rate": 6.706389611284953e-06,
+      "loss": 0.1347,
+      "step": 995
+    },
+    {
+      "epoch": 4.506787330316742,
+      "grad_norm": 0.3937552869319916,
+      "learning_rate": 6.698964210128832e-06,
+      "loss": 0.1274,
+      "step": 996
+    },
+    {
+      "epoch": 4.51131221719457,
+      "grad_norm": 0.4109607934951782,
+      "learning_rate": 6.691534570451908e-06,
+      "loss": 0.1214,
+      "step": 997
+    },
+    {
+      "epoch": 4.515837104072398,
+      "grad_norm": 0.37876155972480774,
+      "learning_rate": 6.684100710789405e-06,
+      "loss": 0.1353,
+      "step": 998
+    },
+    {
+      "epoch": 4.520361990950226,
+      "grad_norm": 0.4157622158527374,
+      "learning_rate": 6.676662649687074e-06,
+      "loss": 0.143,
+      "step": 999
+    },
+    {
+      "epoch": 4.524886877828054,
+      "grad_norm": 0.41489532589912415,
+      "learning_rate": 6.669220405701149e-06,
+      "loss": 0.1336,
+      "step": 1000
+    },
+    {
+      "epoch": 4.529411764705882,
+      "grad_norm": 0.3667726516723633,
+      "learning_rate": 6.6617739973982985e-06,
+      "loss": 0.1218,
+      "step": 1001
+    },
+    {
+      "epoch": 4.53393665158371,
+      "grad_norm": 0.4074230194091797,
+      "learning_rate": 6.6543234433555815e-06,
+      "loss": 0.1255,
+      "step": 1002
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.358986496925354,
+      "learning_rate": 6.646868762160399e-06,
+      "loss": 0.1037,
+      "step": 1003
+    },
+    {
+      "epoch": 4.542986425339366,
+      "grad_norm": 0.4198194742202759,
+      "learning_rate": 6.639409972410446e-06,
+      "loss": 0.1726,
+      "step": 1004
+    },
+    {
+      "epoch": 4.547511312217194,
+      "grad_norm": 0.3871258497238159,
+      "learning_rate": 6.6319470927136694e-06,
+      "loss": 0.104,
+      "step": 1005
+    },
+    {
+      "epoch": 4.552036199095022,
+      "grad_norm": 0.40250852704048157,
+      "learning_rate": 6.624480141688221e-06,
+      "loss": 0.1248,
+      "step": 1006
+    },
+    {
+      "epoch": 4.5565610859728505,
+      "grad_norm": 0.38853251934051514,
+      "learning_rate": 6.617009137962407e-06,
+      "loss": 0.1078,
+      "step": 1007
+    },
+    {
+      "epoch": 4.5610859728506785,
+      "grad_norm": 0.45916301012039185,
+      "learning_rate": 6.609534100174646e-06,
+      "loss": 0.1556,
+      "step": 1008
+    },
+    {
+      "epoch": 4.5656108597285066,
+      "grad_norm": 0.43607795238494873,
+      "learning_rate": 6.6020550469734175e-06,
+      "loss": 0.1119,
+      "step": 1009
+    },
+    {
+      "epoch": 4.570135746606335,
+      "grad_norm": 0.3727249503135681,
+      "learning_rate": 6.594571997017224e-06,
+      "loss": 0.1095,
+      "step": 1010
+    },
+    {
+      "epoch": 4.574660633484163,
+      "grad_norm": 0.4152163863182068,
+      "learning_rate": 6.587084968974533e-06,
+      "loss": 0.141,
+      "step": 1011
+    },
+    {
+      "epoch": 4.579185520361991,
+      "grad_norm": 0.3729671835899353,
+      "learning_rate": 6.5795939815237405e-06,
+      "loss": 0.1133,
+      "step": 1012
+    },
+    {
+      "epoch": 4.583710407239819,
+      "grad_norm": 0.3653905391693115,
+      "learning_rate": 6.57209905335312e-06,
+      "loss": 0.1186,
+      "step": 1013
+    },
+    {
+      "epoch": 4.588235294117647,
+      "grad_norm": 0.406921923160553,
+      "learning_rate": 6.5646002031607726e-06,
+      "loss": 0.1338,
+      "step": 1014
+    },
+    {
+      "epoch": 4.592760180995475,
+      "grad_norm": 0.46261557936668396,
+      "learning_rate": 6.557097449654588e-06,
+      "loss": 0.1515,
+      "step": 1015
+    },
+    {
+      "epoch": 4.597285067873303,
+      "grad_norm": 0.43040263652801514,
+      "learning_rate": 6.549590811552193e-06,
+      "loss": 0.1087,
+      "step": 1016
+    },
+    {
+      "epoch": 4.601809954751131,
+      "grad_norm": 0.4118295907974243,
+      "learning_rate": 6.542080307580905e-06,
+      "loss": 0.1161,
+      "step": 1017
+    },
+    {
+      "epoch": 4.606334841628959,
+      "grad_norm": 0.5026023387908936,
+      "learning_rate": 6.534565956477684e-06,
+      "loss": 0.1373,
+      "step": 1018
+    },
+    {
+      "epoch": 4.610859728506787,
+      "grad_norm": 0.37630221247673035,
+      "learning_rate": 6.5270477769890906e-06,
+      "loss": 0.1281,
+      "step": 1019
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.4093388617038727,
+      "learning_rate": 6.519525787871235e-06,
+      "loss": 0.1177,
+      "step": 1020
+    },
+    {
+      "epoch": 4.619909502262443,
+      "grad_norm": 0.38547810912132263,
+      "learning_rate": 6.512000007889731e-06,
+      "loss": 0.1268,
+      "step": 1021
+    },
+    {
+      "epoch": 4.624434389140271,
+      "grad_norm": 0.3508959412574768,
+      "learning_rate": 6.504470455819651e-06,
+      "loss": 0.1081,
+      "step": 1022
+    },
+    {
+      "epoch": 4.628959276018099,
+      "grad_norm": 0.40760180354118347,
+      "learning_rate": 6.496937150445478e-06,
+      "loss": 0.1152,
+      "step": 1023
+    },
+    {
+      "epoch": 4.633484162895927,
+      "grad_norm": 0.42812004685401917,
+      "learning_rate": 6.489400110561056e-06,
+      "loss": 0.1307,
+      "step": 1024
+    },
+    {
+      "epoch": 4.638009049773755,
+      "grad_norm": 0.3820403218269348,
+      "learning_rate": 6.481859354969549e-06,
+      "loss": 0.131,
+      "step": 1025
+    },
+    {
+      "epoch": 4.642533936651584,
+      "grad_norm": 0.374654620885849,
+      "learning_rate": 6.474314902483388e-06,
+      "loss": 0.1177,
+      "step": 1026
+    },
+    {
+      "epoch": 4.647058823529412,
+      "grad_norm": 0.3797123432159424,
+      "learning_rate": 6.466766771924231e-06,
+      "loss": 0.1328,
+      "step": 1027
+    },
+    {
+      "epoch": 4.65158371040724,
+      "grad_norm": 0.42189446091651917,
+      "learning_rate": 6.4592149821229064e-06,
+      "loss": 0.1309,
+      "step": 1028
+    },
+    {
+      "epoch": 4.656108597285068,
+      "grad_norm": 0.4253498911857605,
+      "learning_rate": 6.451659551919378e-06,
+      "loss": 0.1297,
+      "step": 1029
+    },
+    {
+      "epoch": 4.660633484162896,
+      "grad_norm": 0.4357294738292694,
+      "learning_rate": 6.444100500162684e-06,
+      "loss": 0.1669,
+      "step": 1030
+    },
+    {
+      "epoch": 4.665158371040724,
+      "grad_norm": 0.3664080798625946,
+      "learning_rate": 6.436537845710904e-06,
+      "loss": 0.1158,
+      "step": 1031
+    },
+    {
+      "epoch": 4.669683257918552,
+      "grad_norm": 0.4295387268066406,
+      "learning_rate": 6.428971607431102e-06,
+      "loss": 0.1322,
+      "step": 1032
+    },
+    {
+      "epoch": 4.67420814479638,
+      "grad_norm": 0.3990268409252167,
+      "learning_rate": 6.421401804199284e-06,
+      "loss": 0.1383,
+      "step": 1033
+    },
+    {
+      "epoch": 4.678733031674208,
+      "grad_norm": 0.4077833890914917,
+      "learning_rate": 6.413828454900351e-06,
+      "loss": 0.1474,
+      "step": 1034
+    },
+    {
+      "epoch": 4.683257918552036,
+      "grad_norm": 0.4128705859184265,
+      "learning_rate": 6.406251578428047e-06,
+      "loss": 0.1417,
+      "step": 1035
+    },
+    {
+      "epoch": 4.6877828054298645,
+      "grad_norm": 0.41966959834098816,
+      "learning_rate": 6.398671193684919e-06,
+      "loss": 0.0998,
+      "step": 1036
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.379185289144516,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.1333,
+      "step": 1037
+    },
+    {
+      "epoch": 4.6968325791855206,
+      "grad_norm": 0.38707613945007324,
+      "learning_rate": 6.383499975040086e-06,
+      "loss": 0.1288,
+      "step": 1038
+    },
+    {
+      "epoch": 4.701357466063349,
+      "grad_norm": 0.4196775257587433,
+      "learning_rate": 6.375909178987046e-06,
+      "loss": 0.1327,
+      "step": 1039
+    },
+    {
+      "epoch": 4.705882352941177,
+      "grad_norm": 0.36592400074005127,
+      "learning_rate": 6.368314950360416e-06,
+      "loss": 0.1233,
+      "step": 1040
+    },
+    {
+      "epoch": 4.710407239819005,
+      "grad_norm": 0.35522153973579407,
+      "learning_rate": 6.36071730810603e-06,
+      "loss": 0.1071,
+      "step": 1041
+    },
+    {
+      "epoch": 4.714932126696833,
+      "grad_norm": 0.3890746533870697,
+      "learning_rate": 6.35311627117824e-06,
+      "loss": 0.1321,
+      "step": 1042
+    },
+    {
+      "epoch": 4.719457013574661,
+      "grad_norm": 0.3768482506275177,
+      "learning_rate": 6.3455118585398676e-06,
+      "loss": 0.1092,
+      "step": 1043
+    },
+    {
+      "epoch": 4.723981900452489,
+      "grad_norm": 0.3709065318107605,
+      "learning_rate": 6.337904089162152e-06,
+      "loss": 0.1207,
+      "step": 1044
+    },
+    {
+      "epoch": 4.728506787330317,
+      "grad_norm": 0.4227345585823059,
+      "learning_rate": 6.3302929820247125e-06,
+      "loss": 0.1046,
+      "step": 1045
+    },
+    {
+      "epoch": 4.733031674208145,
+      "grad_norm": 0.44126981496810913,
+      "learning_rate": 6.3226785561154914e-06,
+      "loss": 0.1215,
+      "step": 1046
+    },
+    {
+      "epoch": 4.737556561085973,
+      "grad_norm": 0.3676055669784546,
+      "learning_rate": 6.315060830430711e-06,
+      "loss": 0.1379,
+      "step": 1047
+    },
+    {
+      "epoch": 4.742081447963801,
+      "grad_norm": 0.41453203558921814,
+      "learning_rate": 6.307439823974826e-06,
+      "loss": 0.1357,
+      "step": 1048
+    },
+    {
+      "epoch": 4.746606334841629,
+      "grad_norm": 0.43501806259155273,
+      "learning_rate": 6.299815555760478e-06,
+      "loss": 0.1295,
+      "step": 1049
+    },
+    {
+      "epoch": 4.751131221719457,
+      "grad_norm": 0.38064759969711304,
+      "learning_rate": 6.29218804480844e-06,
+      "loss": 0.132,
+      "step": 1050
+    },
+    {
+      "epoch": 4.755656108597285,
+      "grad_norm": 0.41559624671936035,
+      "learning_rate": 6.284557310147583e-06,
+      "loss": 0.1299,
+      "step": 1051
+    },
+    {
+      "epoch": 4.760180995475113,
+      "grad_norm": 0.40272775292396545,
+      "learning_rate": 6.276923370814815e-06,
+      "loss": 0.1378,
+      "step": 1052
+    },
+    {
+      "epoch": 4.764705882352941,
+      "grad_norm": 0.38200849294662476,
+      "learning_rate": 6.269286245855039e-06,
+      "loss": 0.121,
+      "step": 1053
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.42263320088386536,
+      "learning_rate": 6.261645954321109e-06,
+      "loss": 0.1377,
+      "step": 1054
+    },
+    {
+      "epoch": 4.773755656108597,
+      "grad_norm": 0.44799476861953735,
+      "learning_rate": 6.254002515273775e-06,
+      "loss": 0.1183,
+      "step": 1055
+    },
+    {
+      "epoch": 4.778280542986425,
+      "grad_norm": 0.39172661304473877,
+      "learning_rate": 6.246355947781641e-06,
+      "loss": 0.1239,
+      "step": 1056
+    },
+    {
+      "epoch": 4.782805429864253,
+      "grad_norm": 0.3744019567966461,
+      "learning_rate": 6.2387062709211155e-06,
+      "loss": 0.1221,
+      "step": 1057
+    },
+    {
+      "epoch": 4.787330316742081,
+      "grad_norm": 0.3859163820743561,
+      "learning_rate": 6.231053503776363e-06,
+      "loss": 0.1416,
+      "step": 1058
+    },
+    {
+      "epoch": 4.791855203619909,
+      "grad_norm": 0.39975130558013916,
+      "learning_rate": 6.223397665439261e-06,
+      "loss": 0.1323,
+      "step": 1059
+    },
+    {
+      "epoch": 4.796380090497737,
+      "grad_norm": 0.41266360878944397,
+      "learning_rate": 6.2157387750093455e-06,
+      "loss": 0.1067,
+      "step": 1060
+    },
+    {
+      "epoch": 4.800904977375565,
+      "grad_norm": 0.3963589668273926,
+      "learning_rate": 6.208076851593768e-06,
+      "loss": 0.1338,
+      "step": 1061
+    },
+    {
+      "epoch": 4.8054298642533935,
+      "grad_norm": 0.42241835594177246,
+      "learning_rate": 6.200411914307247e-06,
+      "loss": 0.123,
+      "step": 1062
+    },
+    {
+      "epoch": 4.8099547511312215,
+      "grad_norm": 0.350961834192276,
+      "learning_rate": 6.192743982272019e-06,
+      "loss": 0.1125,
+      "step": 1063
+    },
+    {
+      "epoch": 4.8144796380090495,
+      "grad_norm": 0.39947742223739624,
+      "learning_rate": 6.185073074617793e-06,
+      "loss": 0.1147,
+      "step": 1064
+    },
+    {
+      "epoch": 4.819004524886878,
+      "grad_norm": 0.36715415120124817,
+      "learning_rate": 6.177399210481698e-06,
+      "loss": 0.1121,
+      "step": 1065
+    },
+    {
+      "epoch": 4.823529411764706,
+      "grad_norm": 0.3727978467941284,
+      "learning_rate": 6.169722409008244e-06,
+      "loss": 0.1188,
+      "step": 1066
+    },
+    {
+      "epoch": 4.828054298642534,
+      "grad_norm": 0.39884573221206665,
+      "learning_rate": 6.1620426893492645e-06,
+      "loss": 0.1229,
+      "step": 1067
+    },
+    {
+      "epoch": 4.832579185520362,
+      "grad_norm": 0.38297924399375916,
+      "learning_rate": 6.1543600706638766e-06,
+      "loss": 0.135,
+      "step": 1068
+    },
+    {
+      "epoch": 4.83710407239819,
+      "grad_norm": 0.4219057559967041,
+      "learning_rate": 6.1466745721184275e-06,
+      "loss": 0.1362,
+      "step": 1069
+    },
+    {
+      "epoch": 4.841628959276018,
+      "grad_norm": 0.3759216070175171,
+      "learning_rate": 6.13898621288645e-06,
+      "loss": 0.1149,
+      "step": 1070
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.44968006014823914,
+      "learning_rate": 6.131295012148613e-06,
+      "loss": 0.1403,
+      "step": 1071
+    },
+    {
+      "epoch": 4.850678733031674,
+      "grad_norm": 0.3693656027317047,
+      "learning_rate": 6.123600989092673e-06,
+      "loss": 0.1185,
+      "step": 1072
+    },
+    {
+      "epoch": 4.855203619909502,
+      "grad_norm": 0.42584556341171265,
+      "learning_rate": 6.115904162913431e-06,
+      "loss": 0.1186,
+      "step": 1073
+    },
+    {
+      "epoch": 4.859728506787331,
+      "grad_norm": 0.4135635495185852,
+      "learning_rate": 6.10820455281268e-06,
+      "loss": 0.11,
+      "step": 1074
+    },
+    {
+      "epoch": 4.864253393665159,
+      "grad_norm": 0.40252724289894104,
+      "learning_rate": 6.100502177999156e-06,
+      "loss": 0.1194,
+      "step": 1075
+    },
+    {
+      "epoch": 4.868778280542987,
+      "grad_norm": 0.4541674852371216,
+      "learning_rate": 6.092797057688496e-06,
+      "loss": 0.127,
+      "step": 1076
+    },
+    {
+      "epoch": 4.873303167420815,
+      "grad_norm": 0.37728506326675415,
+      "learning_rate": 6.085089211103181e-06,
+      "loss": 0.128,
+      "step": 1077
+    },
+    {
+      "epoch": 4.877828054298643,
+      "grad_norm": 0.3978901505470276,
+      "learning_rate": 6.077378657472498e-06,
+      "loss": 0.1358,
+      "step": 1078
+    },
+    {
+      "epoch": 4.882352941176471,
+      "grad_norm": 0.3870754837989807,
+      "learning_rate": 6.0696654160324875e-06,
+      "loss": 0.1281,
+      "step": 1079
+    },
+    {
+      "epoch": 4.886877828054299,
+      "grad_norm": 0.40275850892066956,
+      "learning_rate": 6.061949506025892e-06,
+      "loss": 0.1183,
+      "step": 1080
+    },
+    {
+      "epoch": 4.891402714932127,
+      "grad_norm": 0.37450602650642395,
+      "learning_rate": 6.0542309467021156e-06,
+      "loss": 0.1533,
+      "step": 1081
+    },
+    {
+      "epoch": 4.895927601809955,
+      "grad_norm": 0.41943442821502686,
+      "learning_rate": 6.046509757317168e-06,
+      "loss": 0.1318,
+      "step": 1082
+    },
+    {
+      "epoch": 4.900452488687783,
+      "grad_norm": 0.3888932466506958,
+      "learning_rate": 6.038785957133624e-06,
+      "loss": 0.1224,
+      "step": 1083
+    },
+    {
+      "epoch": 4.904977375565611,
+      "grad_norm": 0.47233960032463074,
+      "learning_rate": 6.0310595654205674e-06,
+      "loss": 0.1557,
+      "step": 1084
+    },
+    {
+      "epoch": 4.909502262443439,
+      "grad_norm": 0.43002673983573914,
+      "learning_rate": 6.0233306014535505e-06,
+      "loss": 0.11,
+      "step": 1085
+    },
+    {
+      "epoch": 4.914027149321267,
+      "grad_norm": 0.39190760254859924,
+      "learning_rate": 6.015599084514542e-06,
+      "loss": 0.1115,
+      "step": 1086
+    },
+    {
+      "epoch": 4.918552036199095,
+      "grad_norm": 0.36419686675071716,
+      "learning_rate": 6.00786503389188e-06,
+      "loss": 0.1156,
+      "step": 1087
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.3950363099575043,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.1399,
+      "step": 1088
+    },
+    {
+      "epoch": 4.927601809954751,
+      "grad_norm": 0.44508564472198486,
+      "learning_rate": 5.992389408780501e-06,
+      "loss": 0.1134,
+      "step": 1089
+    },
+    {
+      "epoch": 4.932126696832579,
+      "grad_norm": 0.4318121373653412,
+      "learning_rate": 5.984647872899873e-06,
+      "loss": 0.1034,
+      "step": 1090
+    },
+    {
+      "epoch": 4.9366515837104075,
+      "grad_norm": 0.37849506735801697,
+      "learning_rate": 5.976903880551669e-06,
+      "loss": 0.11,
+      "step": 1091
+    },
+    {
+      "epoch": 4.9411764705882355,
+      "grad_norm": 0.44689616560935974,
+      "learning_rate": 5.9691574510553505e-06,
+      "loss": 0.1248,
+      "step": 1092
+    },
+    {
+      "epoch": 4.9457013574660635,
+      "grad_norm": 0.37326931953430176,
+      "learning_rate": 5.961408603736461e-06,
+      "loss": 0.1354,
+      "step": 1093
+    },
+    {
+      "epoch": 4.950226244343892,
+      "grad_norm": 0.3781431019306183,
+      "learning_rate": 5.953657357926569e-06,
+      "loss": 0.1296,
+      "step": 1094
+    },
+    {
+      "epoch": 4.95475113122172,
+      "grad_norm": 0.41875743865966797,
+      "learning_rate": 5.945903732963234e-06,
+      "loss": 0.1351,
+      "step": 1095
+    },
+    {
+      "epoch": 4.959276018099548,
+      "grad_norm": 0.3916160464286804,
+      "learning_rate": 5.938147748189947e-06,
+      "loss": 0.1383,
+      "step": 1096
+    },
+    {
+      "epoch": 4.963800904977376,
+      "grad_norm": 0.42897456884384155,
+      "learning_rate": 5.930389422956088e-06,
+      "loss": 0.1494,
+      "step": 1097
+    },
+    {
+      "epoch": 4.968325791855204,
+      "grad_norm": 0.4615175127983093,
+      "learning_rate": 5.922628776616876e-06,
+      "loss": 0.1263,
+      "step": 1098
+    },
+    {
+      "epoch": 4.972850678733032,
+      "grad_norm": 0.43058305978775024,
+      "learning_rate": 5.914865828533319e-06,
+      "loss": 0.1356,
+      "step": 1099
+    },
+    {
+      "epoch": 4.97737556561086,
+      "grad_norm": 0.42625221610069275,
+      "learning_rate": 5.907100598072166e-06,
+      "loss": 0.1388,
+      "step": 1100
+    },
+    {
+      "epoch": 4.981900452488688,
+      "grad_norm": 0.42407146096229553,
+      "learning_rate": 5.899333104605864e-06,
+      "loss": 0.1476,
+      "step": 1101
+    },
+    {
+      "epoch": 4.986425339366516,
+      "grad_norm": 0.3559970557689667,
+      "learning_rate": 5.891563367512505e-06,
+      "loss": 0.1148,
+      "step": 1102
+    },
+    {
+      "epoch": 4.990950226244344,
+      "grad_norm": 0.3517310917377472,
+      "learning_rate": 5.883791406175775e-06,
+      "loss": 0.11,
+      "step": 1103
+    },
+    {
+      "epoch": 4.995475113122172,
+      "grad_norm": 0.3917599320411682,
+      "learning_rate": 5.876017239984911e-06,
+      "loss": 0.1138,
+      "step": 1104
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.39552751183509827,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.1195,
+      "step": 1105
+    },
+    {
+      "epoch": 5.004524886877828,
+      "grad_norm": 0.3723715841770172,
+      "learning_rate": 5.860462370625189e-06,
+      "loss": 0.067,
+      "step": 1106
+    },
+    {
+      "epoch": 5.009049773755656,
+      "grad_norm": 0.4061602056026459,
+      "learning_rate": 5.85268170626211e-06,
+      "loss": 0.0793,
+      "step": 1107
+    },
+    {
+      "epoch": 5.013574660633484,
+      "grad_norm": 0.3349854052066803,
+      "learning_rate": 5.844898914656371e-06,
+      "loss": 0.0649,
+      "step": 1108
+    },
+    {
+      "epoch": 5.018099547511312,
+      "grad_norm": 0.383832186460495,
+      "learning_rate": 5.837114015224223e-06,
+      "loss": 0.0664,
+      "step": 1109
+    },
+    {
+      "epoch": 5.02262443438914,
+      "grad_norm": 0.4438365399837494,
+      "learning_rate": 5.829327027387184e-06,
+      "loss": 0.0767,
+      "step": 1110
+    },
+    {
+      "epoch": 5.027149321266968,
+      "grad_norm": 0.4521051049232483,
+      "learning_rate": 5.821537970571978e-06,
+      "loss": 0.0728,
+      "step": 1111
+    },
+    {
+      "epoch": 5.031674208144796,
+      "grad_norm": 0.411251038312912,
+      "learning_rate": 5.813746864210489e-06,
+      "loss": 0.0652,
+      "step": 1112
+    },
+    {
+      "epoch": 5.036199095022624,
+      "grad_norm": 0.42150846123695374,
+      "learning_rate": 5.805953727739719e-06,
+      "loss": 0.0615,
+      "step": 1113
+    },
+    {
+      "epoch": 5.040723981900452,
+      "grad_norm": 0.3648434281349182,
+      "learning_rate": 5.79815858060173e-06,
+      "loss": 0.0527,
+      "step": 1114
+    },
+    {
+      "epoch": 5.04524886877828,
+      "grad_norm": 0.37746983766555786,
+      "learning_rate": 5.790361442243605e-06,
+      "loss": 0.0689,
+      "step": 1115
+    },
+    {
+      "epoch": 5.049773755656108,
+      "grad_norm": 0.35608813166618347,
+      "learning_rate": 5.78256233211739e-06,
+      "loss": 0.0752,
+      "step": 1116
+    },
+    {
+      "epoch": 5.0542986425339365,
+      "grad_norm": 0.36897188425064087,
+      "learning_rate": 5.774761269680052e-06,
+      "loss": 0.0575,
+      "step": 1117
+    },
+    {
+      "epoch": 5.0588235294117645,
+      "grad_norm": 0.334196537733078,
+      "learning_rate": 5.766958274393428e-06,
+      "loss": 0.0578,
+      "step": 1118
+    },
+    {
+      "epoch": 5.0633484162895925,
+      "grad_norm": 0.3506363034248352,
+      "learning_rate": 5.759153365724178e-06,
+      "loss": 0.0642,
+      "step": 1119
+    },
+    {
+      "epoch": 5.067873303167421,
+      "grad_norm": 0.3522595763206482,
+      "learning_rate": 5.751346563143736e-06,
+      "loss": 0.0635,
+      "step": 1120
+    },
+    {
+      "epoch": 5.072398190045249,
+      "grad_norm": 0.6559270620346069,
+      "learning_rate": 5.743537886128258e-06,
+      "loss": 0.1065,
+      "step": 1121
+    },
+    {
+      "epoch": 5.076923076923077,
+      "grad_norm": 0.339643657207489,
+      "learning_rate": 5.735727354158581e-06,
+      "loss": 0.0674,
+      "step": 1122
+    },
+    {
+      "epoch": 5.081447963800905,
+      "grad_norm": 0.3840213418006897,
+      "learning_rate": 5.727914986720164e-06,
+      "loss": 0.0742,
+      "step": 1123
+    },
+    {
+      "epoch": 5.085972850678733,
+      "grad_norm": 0.3815678656101227,
+      "learning_rate": 5.72010080330305e-06,
+      "loss": 0.0602,
+      "step": 1124
+    },
+    {
+      "epoch": 5.090497737556561,
+      "grad_norm": 0.3994493782520294,
+      "learning_rate": 5.712284823401808e-06,
+      "loss": 0.0595,
+      "step": 1125
+    },
+    {
+      "epoch": 5.095022624434389,
+      "grad_norm": 0.40029487013816833,
+      "learning_rate": 5.704467066515492e-06,
+      "loss": 0.0728,
+      "step": 1126
+    },
+    {
+      "epoch": 5.099547511312217,
+      "grad_norm": 0.3628596067428589,
+      "learning_rate": 5.696647552147589e-06,
+      "loss": 0.0598,
+      "step": 1127
+    },
+    {
+      "epoch": 5.104072398190045,
+      "grad_norm": 0.36752235889434814,
+      "learning_rate": 5.688826299805971e-06,
+      "loss": 0.074,
+      "step": 1128
+    },
+    {
+      "epoch": 5.108597285067873,
+      "grad_norm": 0.3554799556732178,
+      "learning_rate": 5.681003329002842e-06,
+      "loss": 0.0614,
+      "step": 1129
+    },
+    {
+      "epoch": 5.113122171945701,
+      "grad_norm": 0.41899964213371277,
+      "learning_rate": 5.673178659254698e-06,
+      "loss": 0.0974,
+      "step": 1130
+    },
+    {
+      "epoch": 5.117647058823529,
+      "grad_norm": 0.31798073649406433,
+      "learning_rate": 5.66535231008227e-06,
+      "loss": 0.0629,
+      "step": 1131
+    },
+    {
+      "epoch": 5.122171945701357,
+      "grad_norm": 0.3718262016773224,
+      "learning_rate": 5.657524301010478e-06,
+      "loss": 0.0696,
+      "step": 1132
+    },
+    {
+      "epoch": 5.126696832579185,
+      "grad_norm": 0.3424396216869354,
+      "learning_rate": 5.64969465156839e-06,
+      "loss": 0.0553,
+      "step": 1133
+    },
+    {
+      "epoch": 5.131221719457014,
+      "grad_norm": 0.3795858323574066,
+      "learning_rate": 5.641863381289158e-06,
+      "loss": 0.0805,
+      "step": 1134
+    },
+    {
+      "epoch": 5.135746606334842,
+      "grad_norm": 0.40230146050453186,
+      "learning_rate": 5.634030509709982e-06,
+      "loss": 0.0814,
+      "step": 1135
+    },
+    {
+      "epoch": 5.14027149321267,
+      "grad_norm": 0.3625279366970062,
+      "learning_rate": 5.626196056372056e-06,
+      "loss": 0.0819,
+      "step": 1136
+    },
+    {
+      "epoch": 5.144796380090498,
+      "grad_norm": 0.37393632531166077,
+      "learning_rate": 5.618360040820521e-06,
+      "loss": 0.0704,
+      "step": 1137
+    },
+    {
+      "epoch": 5.149321266968326,
+      "grad_norm": 0.3487889766693115,
+      "learning_rate": 5.6105224826044105e-06,
+      "loss": 0.0496,
+      "step": 1138
+    },
+    {
+      "epoch": 5.153846153846154,
+      "grad_norm": 0.3692089319229126,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.0672,
+      "step": 1139
+    },
+    {
+      "epoch": 5.158371040723982,
+      "grad_norm": 0.4027572572231293,
+      "learning_rate": 5.594842816393819e-06,
+      "loss": 0.0616,
+      "step": 1140
+    },
+    {
+      "epoch": 5.16289592760181,
+      "grad_norm": 0.44265252351760864,
+      "learning_rate": 5.58700074751646e-06,
+      "loss": 0.0778,
+      "step": 1141
+    },
+    {
+      "epoch": 5.167420814479638,
+      "grad_norm": 0.36350980401039124,
+      "learning_rate": 5.579157214208675e-06,
+      "loss": 0.0637,
+      "step": 1142
+    },
+    {
+      "epoch": 5.171945701357466,
+      "grad_norm": 0.35687702894210815,
+      "learning_rate": 5.571312236038259e-06,
+      "loss": 0.0753,
+      "step": 1143
+    },
+    {
+      "epoch": 5.176470588235294,
+      "grad_norm": 0.35756856203079224,
+      "learning_rate": 5.5634658325766066e-06,
+      "loss": 0.0804,
+      "step": 1144
+    },
+    {
+      "epoch": 5.180995475113122,
+      "grad_norm": 0.35187193751335144,
+      "learning_rate": 5.555618023398671e-06,
+      "loss": 0.0776,
+      "step": 1145
+    },
+    {
+      "epoch": 5.1855203619909505,
+      "grad_norm": 0.361777126789093,
+      "learning_rate": 5.547768828082915e-06,
+      "loss": 0.0708,
+      "step": 1146
+    },
+    {
+      "epoch": 5.1900452488687785,
+      "grad_norm": 0.40918031334877014,
+      "learning_rate": 5.539918266211253e-06,
+      "loss": 0.0959,
+      "step": 1147
+    },
+    {
+      "epoch": 5.1945701357466065,
+      "grad_norm": 0.37231284379959106,
+      "learning_rate": 5.532066357369012e-06,
+      "loss": 0.0646,
+      "step": 1148
+    },
+    {
+      "epoch": 5.199095022624435,
+      "grad_norm": 0.3823784589767456,
+      "learning_rate": 5.524213121144883e-06,
+      "loss": 0.064,
+      "step": 1149
+    },
+    {
+      "epoch": 5.203619909502263,
+      "grad_norm": 0.3881456255912781,
+      "learning_rate": 5.516358577130863e-06,
+      "loss": 0.0799,
+      "step": 1150
+    },
+    {
+      "epoch": 5.208144796380091,
+      "grad_norm": 0.38889485597610474,
+      "learning_rate": 5.508502744922212e-06,
+      "loss": 0.0866,
+      "step": 1151
+    },
+    {
+      "epoch": 5.212669683257919,
+      "grad_norm": 0.34559571743011475,
+      "learning_rate": 5.500645644117406e-06,
+      "loss": 0.0659,
+      "step": 1152
+    },
+    {
+      "epoch": 5.217194570135747,
+      "grad_norm": 0.4092608392238617,
+      "learning_rate": 5.492787294318087e-06,
+      "loss": 0.0876,
+      "step": 1153
+    },
+    {
+      "epoch": 5.221719457013575,
+      "grad_norm": 0.3414454460144043,
+      "learning_rate": 5.484927715129011e-06,
+      "loss": 0.0681,
+      "step": 1154
+    },
+    {
+      "epoch": 5.226244343891403,
+      "grad_norm": 0.3446487486362457,
+      "learning_rate": 5.4770669261580014e-06,
+      "loss": 0.0532,
+      "step": 1155
+    },
+    {
+      "epoch": 5.230769230769231,
+      "grad_norm": 0.38337475061416626,
+      "learning_rate": 5.469204947015897e-06,
+      "loss": 0.0765,
+      "step": 1156
+    },
+    {
+      "epoch": 5.235294117647059,
+      "grad_norm": 0.41466450691223145,
+      "learning_rate": 5.46134179731651e-06,
+      "loss": 0.0788,
+      "step": 1157
+    },
+    {
+      "epoch": 5.239819004524887,
+      "grad_norm": 0.3642111122608185,
+      "learning_rate": 5.453477496676573e-06,
+      "loss": 0.0809,
+      "step": 1158
+    },
+    {
+      "epoch": 5.244343891402715,
+      "grad_norm": 0.3510957360267639,
+      "learning_rate": 5.445612064715684e-06,
+      "loss": 0.0683,
+      "step": 1159
+    },
+    {
+      "epoch": 5.248868778280543,
+      "grad_norm": 0.390898734331131,
+      "learning_rate": 5.437745521056272e-06,
+      "loss": 0.0803,
+      "step": 1160
+    },
+    {
+      "epoch": 5.253393665158371,
+      "grad_norm": 0.4196690618991852,
+      "learning_rate": 5.4298778853235315e-06,
+      "loss": 0.0717,
+      "step": 1161
+    },
+    {
+      "epoch": 5.257918552036199,
+      "grad_norm": 0.32665446400642395,
+      "learning_rate": 5.422009177145388e-06,
+      "loss": 0.0527,
+      "step": 1162
+    },
+    {
+      "epoch": 5.262443438914027,
+      "grad_norm": 0.35169556736946106,
+      "learning_rate": 5.414139416152435e-06,
+      "loss": 0.0507,
+      "step": 1163
+    },
+    {
+      "epoch": 5.266968325791855,
+      "grad_norm": 0.36220046877861023,
+      "learning_rate": 5.406268621977902e-06,
+      "loss": 0.0782,
+      "step": 1164
+    },
+    {
+      "epoch": 5.271493212669683,
+      "grad_norm": 0.3889884054660797,
+      "learning_rate": 5.398396814257588e-06,
+      "loss": 0.0694,
+      "step": 1165
+    },
+    {
+      "epoch": 5.276018099547511,
+      "grad_norm": 0.3402976095676422,
+      "learning_rate": 5.390524012629824e-06,
+      "loss": 0.0572,
+      "step": 1166
+    },
+    {
+      "epoch": 5.280542986425339,
+      "grad_norm": 0.3195909559726715,
+      "learning_rate": 5.382650236735421e-06,
+      "loss": 0.0651,
+      "step": 1167
+    },
+    {
+      "epoch": 5.285067873303167,
+      "grad_norm": 0.3863244354724884,
+      "learning_rate": 5.3747755062176185e-06,
+      "loss": 0.0598,
+      "step": 1168
+    },
+    {
+      "epoch": 5.289592760180995,
+      "grad_norm": 0.33079132437705994,
+      "learning_rate": 5.3668998407220385e-06,
+      "loss": 0.0635,
+      "step": 1169
+    },
+    {
+      "epoch": 5.294117647058823,
+      "grad_norm": 0.3895846903324127,
+      "learning_rate": 5.359023259896638e-06,
+      "loss": 0.065,
+      "step": 1170
+    },
+    {
+      "epoch": 5.298642533936651,
+      "grad_norm": 0.3789190948009491,
+      "learning_rate": 5.351145783391654e-06,
+      "loss": 0.0744,
+      "step": 1171
+    },
+    {
+      "epoch": 5.3031674208144794,
+      "grad_norm": 0.35495415329933167,
+      "learning_rate": 5.343267430859559e-06,
+      "loss": 0.0834,
+      "step": 1172
+    },
+    {
+      "epoch": 5.3076923076923075,
+      "grad_norm": 0.38719555735588074,
+      "learning_rate": 5.335388221955012e-06,
+      "loss": 0.0666,
+      "step": 1173
+    },
+    {
+      "epoch": 5.3122171945701355,
+      "grad_norm": 0.3902130126953125,
+      "learning_rate": 5.327508176334808e-06,
+      "loss": 0.0681,
+      "step": 1174
+    },
+    {
+      "epoch": 5.316742081447964,
+      "grad_norm": 0.391261488199234,
+      "learning_rate": 5.319627313657829e-06,
+      "loss": 0.0768,
+      "step": 1175
+    },
+    {
+      "epoch": 5.321266968325792,
+      "grad_norm": 0.3392927348613739,
+      "learning_rate": 5.311745653584995e-06,
+      "loss": 0.0679,
+      "step": 1176
+    },
+    {
+      "epoch": 5.32579185520362,
+      "grad_norm": 0.3632453978061676,
+      "learning_rate": 5.3038632157792155e-06,
+      "loss": 0.06,
+      "step": 1177
+    },
+    {
+      "epoch": 5.330316742081448,
+      "grad_norm": 0.3581770062446594,
+      "learning_rate": 5.295980019905342e-06,
+      "loss": 0.0695,
+      "step": 1178
+    },
+    {
+      "epoch": 5.334841628959276,
+      "grad_norm": 0.35193902254104614,
+      "learning_rate": 5.288096085630114e-06,
+      "loss": 0.0689,
+      "step": 1179
+    },
+    {
+      "epoch": 5.339366515837104,
+      "grad_norm": 0.356852650642395,
+      "learning_rate": 5.280211432622118e-06,
+      "loss": 0.0716,
+      "step": 1180
+    },
+    {
+      "epoch": 5.343891402714932,
+      "grad_norm": 0.389583021402359,
+      "learning_rate": 5.272326080551729e-06,
+      "loss": 0.0669,
+      "step": 1181
+    },
+    {
+      "epoch": 5.34841628959276,
+      "grad_norm": 0.3774275779724121,
+      "learning_rate": 5.264440049091068e-06,
+      "loss": 0.0414,
+      "step": 1182
+    },
+    {
+      "epoch": 5.352941176470588,
+      "grad_norm": 0.35961687564849854,
+      "learning_rate": 5.2565533579139484e-06,
+      "loss": 0.0772,
+      "step": 1183
+    },
+    {
+      "epoch": 5.357466063348416,
+      "grad_norm": 0.35463571548461914,
+      "learning_rate": 5.248666026695835e-06,
+      "loss": 0.07,
+      "step": 1184
+    },
+    {
+      "epoch": 5.361990950226244,
+      "grad_norm": 0.3400146961212158,
+      "learning_rate": 5.240778075113781e-06,
+      "loss": 0.0631,
+      "step": 1185
+    },
+    {
+      "epoch": 5.366515837104072,
+      "grad_norm": 0.40522345900535583,
+      "learning_rate": 5.232889522846398e-06,
+      "loss": 0.0726,
+      "step": 1186
+    },
+    {
+      "epoch": 5.371040723981901,
+      "grad_norm": 0.40270382165908813,
+      "learning_rate": 5.2250003895737865e-06,
+      "loss": 0.0655,
+      "step": 1187
+    },
+    {
+      "epoch": 5.375565610859729,
+      "grad_norm": 0.35652971267700195,
+      "learning_rate": 5.217110694977501e-06,
+      "loss": 0.0656,
+      "step": 1188
+    },
+    {
+      "epoch": 5.380090497737557,
+      "grad_norm": 0.36236080527305603,
+      "learning_rate": 5.209220458740496e-06,
+      "loss": 0.0645,
+      "step": 1189
+    },
+    {
+      "epoch": 5.384615384615385,
+      "grad_norm": 0.39016634225845337,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.0681,
+      "step": 1190
+    },
+    {
+      "epoch": 5.389140271493213,
+      "grad_norm": 0.3703252971172333,
+      "learning_rate": 5.193438440082851e-06,
+      "loss": 0.0579,
+      "step": 1191
+    },
+    {
+      "epoch": 5.393665158371041,
+      "grad_norm": 0.3513895571231842,
+      "learning_rate": 5.185546697034679e-06,
+      "loss": 0.0689,
+      "step": 1192
+    },
+    {
+      "epoch": 5.398190045248869,
+      "grad_norm": 0.3632599413394928,
+      "learning_rate": 5.177654491090627e-06,
+      "loss": 0.069,
+      "step": 1193
+    },
+    {
+      "epoch": 5.402714932126697,
+      "grad_norm": 0.46198567748069763,
+      "learning_rate": 5.169761841939912e-06,
+      "loss": 0.0813,
+      "step": 1194
+    },
+    {
+      "epoch": 5.407239819004525,
+      "grad_norm": 0.37140151858329773,
+      "learning_rate": 5.16186876927286e-06,
+      "loss": 0.0728,
+      "step": 1195
+    },
+    {
+      "epoch": 5.411764705882353,
+      "grad_norm": 0.3435964584350586,
+      "learning_rate": 5.153975292780852e-06,
+      "loss": 0.0605,
+      "step": 1196
+    },
+    {
+      "epoch": 5.416289592760181,
+      "grad_norm": 0.3748542368412018,
+      "learning_rate": 5.14608143215628e-06,
+      "loss": 0.06,
+      "step": 1197
+    },
+    {
+      "epoch": 5.420814479638009,
+      "grad_norm": 0.3335817754268646,
+      "learning_rate": 5.1381872070924884e-06,
+      "loss": 0.05,
+      "step": 1198
+    },
+    {
+      "epoch": 5.425339366515837,
+      "grad_norm": 0.4469960033893585,
+      "learning_rate": 5.130292637283735e-06,
+      "loss": 0.0653,
+      "step": 1199
+    },
+    {
+      "epoch": 5.429864253393665,
+      "grad_norm": 0.3295464813709259,
+      "learning_rate": 5.122397742425136e-06,
+      "loss": 0.0711,
+      "step": 1200
+    },
+    {
+      "epoch": 5.4343891402714934,
+      "grad_norm": 0.36631765961647034,
+      "learning_rate": 5.11450254221262e-06,
+      "loss": 0.0726,
+      "step": 1201
+    },
+    {
+      "epoch": 5.4389140271493215,
+      "grad_norm": 0.3538479208946228,
+      "learning_rate": 5.1066070563428736e-06,
+      "loss": 0.0586,
+      "step": 1202
+    },
+    {
+      "epoch": 5.4434389140271495,
+      "grad_norm": 0.3970729112625122,
+      "learning_rate": 5.098711304513303e-06,
+      "loss": 0.0701,
+      "step": 1203
+    },
+    {
+      "epoch": 5.447963800904978,
+      "grad_norm": 0.36112064123153687,
+      "learning_rate": 5.090815306421971e-06,
+      "loss": 0.065,
+      "step": 1204
+    },
+    {
+      "epoch": 5.452488687782806,
+      "grad_norm": 0.34354132413864136,
+      "learning_rate": 5.082919081767558e-06,
+      "loss": 0.0515,
+      "step": 1205
+    },
+    {
+      "epoch": 5.457013574660634,
+      "grad_norm": 0.3775116801261902,
+      "learning_rate": 5.075022650249309e-06,
+      "loss": 0.0684,
+      "step": 1206
+    },
+    {
+      "epoch": 5.461538461538462,
+      "grad_norm": 0.35393577814102173,
+      "learning_rate": 5.067126031566988e-06,
+      "loss": 0.0737,
+      "step": 1207
+    },
+    {
+      "epoch": 5.46606334841629,
+      "grad_norm": 0.36637890338897705,
+      "learning_rate": 5.059229245420819e-06,
+      "loss": 0.0708,
+      "step": 1208
+    },
+    {
+      "epoch": 5.470588235294118,
+      "grad_norm": 0.3605267405509949,
+      "learning_rate": 5.05133231151145e-06,
+      "loss": 0.0555,
+      "step": 1209
+    },
+    {
+      "epoch": 5.475113122171946,
+      "grad_norm": 0.3860551118850708,
+      "learning_rate": 5.0434352495398956e-06,
+      "loss": 0.0706,
+      "step": 1210
+    },
+    {
+      "epoch": 5.479638009049774,
+      "grad_norm": 0.36745893955230713,
+      "learning_rate": 5.035538079207488e-06,
+      "loss": 0.0734,
+      "step": 1211
+    },
+    {
+      "epoch": 5.484162895927602,
+      "grad_norm": 0.3685935437679291,
+      "learning_rate": 5.027640820215834e-06,
+      "loss": 0.0793,
+      "step": 1212
+    },
+    {
+      "epoch": 5.48868778280543,
+      "grad_norm": 0.3923707604408264,
+      "learning_rate": 5.01974349226676e-06,
+      "loss": 0.0827,
+      "step": 1213
+    },
+    {
+      "epoch": 5.493212669683258,
+      "grad_norm": 0.3473922312259674,
+      "learning_rate": 5.01184611506226e-06,
+      "loss": 0.0614,
+      "step": 1214
+    },
+    {
+      "epoch": 5.497737556561086,
+      "grad_norm": 0.3433387279510498,
+      "learning_rate": 5.003948708304457e-06,
+      "loss": 0.0716,
+      "step": 1215
+    },
+    {
+      "epoch": 5.502262443438914,
+      "grad_norm": 0.3532765805721283,
+      "learning_rate": 4.996051291695544e-06,
+      "loss": 0.0719,
+      "step": 1216
+    },
+    {
+      "epoch": 5.506787330316742,
+      "grad_norm": 0.3672707974910736,
+      "learning_rate": 4.988153884937742e-06,
+      "loss": 0.0691,
+      "step": 1217
+    },
+    {
+      "epoch": 5.51131221719457,
+      "grad_norm": 0.46894240379333496,
+      "learning_rate": 4.980256507733242e-06,
+      "loss": 0.0923,
+      "step": 1218
+    },
+    {
+      "epoch": 5.515837104072398,
+      "grad_norm": 0.390020489692688,
+      "learning_rate": 4.9723591797841665e-06,
+      "loss": 0.0738,
+      "step": 1219
+    },
+    {
+      "epoch": 5.520361990950226,
+      "grad_norm": 0.3452206254005432,
+      "learning_rate": 4.964461920792512e-06,
+      "loss": 0.0492,
+      "step": 1220
+    },
+    {
+      "epoch": 5.524886877828054,
+      "grad_norm": 0.3586297333240509,
+      "learning_rate": 4.956564750460107e-06,
+      "loss": 0.0609,
+      "step": 1221
+    },
+    {
+      "epoch": 5.529411764705882,
+      "grad_norm": 0.3771961033344269,
+      "learning_rate": 4.948667688488552e-06,
+      "loss": 0.0786,
+      "step": 1222
+    },
+    {
+      "epoch": 5.53393665158371,
+      "grad_norm": 0.37323540449142456,
+      "learning_rate": 4.940770754579183e-06,
+      "loss": 0.0733,
+      "step": 1223
+    },
+    {
+      "epoch": 5.538461538461538,
+      "grad_norm": 0.353660523891449,
+      "learning_rate": 4.932873968433014e-06,
+      "loss": 0.0729,
+      "step": 1224
+    },
+    {
+      "epoch": 5.542986425339366,
+      "grad_norm": 0.35802799463272095,
+      "learning_rate": 4.924977349750692e-06,
+      "loss": 0.0716,
+      "step": 1225
+    },
+    {
+      "epoch": 5.547511312217194,
+      "grad_norm": 0.4036112129688263,
+      "learning_rate": 4.917080918232444e-06,
+      "loss": 0.0836,
+      "step": 1226
+    },
+    {
+      "epoch": 5.552036199095022,
+      "grad_norm": 0.3766706585884094,
+      "learning_rate": 4.90918469357803e-06,
+      "loss": 0.0638,
+      "step": 1227
+    },
+    {
+      "epoch": 5.5565610859728505,
+      "grad_norm": 0.3591664135456085,
+      "learning_rate": 4.9012886954866985e-06,
+      "loss": 0.0677,
+      "step": 1228
+    },
+    {
+      "epoch": 5.5610859728506785,
+      "grad_norm": 0.37841859459877014,
+      "learning_rate": 4.893392943657127e-06,
+      "loss": 0.0624,
+      "step": 1229
+    },
+    {
+      "epoch": 5.5656108597285066,
+      "grad_norm": 0.3366260826587677,
+      "learning_rate": 4.885497457787383e-06,
+      "loss": 0.0855,
+      "step": 1230
+    },
+    {
+      "epoch": 5.570135746606335,
+      "grad_norm": 0.4003104269504547,
+      "learning_rate": 4.877602257574866e-06,
+      "loss": 0.0731,
+      "step": 1231
+    },
+    {
+      "epoch": 5.574660633484163,
+      "grad_norm": 0.3531114459037781,
+      "learning_rate": 4.8697073627162675e-06,
+      "loss": 0.0574,
+      "step": 1232
+    },
+    {
+      "epoch": 5.579185520361991,
+      "grad_norm": 0.3736576437950134,
+      "learning_rate": 4.861812792907514e-06,
+      "loss": 0.0599,
+      "step": 1233
+    },
+    {
+      "epoch": 5.583710407239819,
+      "grad_norm": 0.34564992785453796,
+      "learning_rate": 4.853918567843722e-06,
+      "loss": 0.0538,
+      "step": 1234
+    },
+    {
+      "epoch": 5.588235294117647,
+      "grad_norm": 0.32957449555397034,
+      "learning_rate": 4.846024707219149e-06,
+      "loss": 0.0594,
+      "step": 1235
+    },
+    {
+      "epoch": 5.592760180995475,
+      "grad_norm": 0.35310038924217224,
+      "learning_rate": 4.8381312307271425e-06,
+      "loss": 0.0728,
+      "step": 1236
+    },
+    {
+      "epoch": 5.597285067873303,
+      "grad_norm": 0.36492466926574707,
+      "learning_rate": 4.830238158060091e-06,
+      "loss": 0.0726,
+      "step": 1237
+    },
+    {
+      "epoch": 5.601809954751131,
+      "grad_norm": 0.35988250374794006,
+      "learning_rate": 4.822345508909376e-06,
+      "loss": 0.0552,
+      "step": 1238
+    },
+    {
+      "epoch": 5.606334841628959,
+      "grad_norm": 0.394717276096344,
+      "learning_rate": 4.814453302965323e-06,
+      "loss": 0.0817,
+      "step": 1239
+    },
+    {
+      "epoch": 5.610859728506787,
+      "grad_norm": 0.374766081571579,
+      "learning_rate": 4.806561559917152e-06,
+      "loss": 0.0695,
+      "step": 1240
+    },
+    {
+      "epoch": 5.615384615384615,
+      "grad_norm": 0.3558162450790405,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.077,
+      "step": 1241
+    },
+    {
+      "epoch": 5.619909502262443,
+      "grad_norm": 0.3893064558506012,
+      "learning_rate": 4.790779541259507e-06,
+      "loss": 0.077,
+      "step": 1242
+    },
+    {
+      "epoch": 5.624434389140271,
+      "grad_norm": 0.34312954545021057,
+      "learning_rate": 4.7828893050225015e-06,
+      "loss": 0.0788,
+      "step": 1243
+    },
+    {
+      "epoch": 5.628959276018099,
+      "grad_norm": 0.38066574931144714,
+      "learning_rate": 4.774999610426216e-06,
+      "loss": 0.0922,
+      "step": 1244
+    },
+    {
+      "epoch": 5.633484162895927,
+      "grad_norm": 0.3637709617614746,
+      "learning_rate": 4.767110477153605e-06,
+      "loss": 0.0565,
+      "step": 1245
+    },
+    {
+      "epoch": 5.638009049773755,
+      "grad_norm": 0.3352390229701996,
+      "learning_rate": 4.7592219248862205e-06,
+      "loss": 0.0703,
+      "step": 1246
+    },
+    {
+      "epoch": 5.642533936651584,
+      "grad_norm": 0.3382686376571655,
+      "learning_rate": 4.751333973304166e-06,
+      "loss": 0.0536,
+      "step": 1247
+    },
+    {
+      "epoch": 5.647058823529412,
+      "grad_norm": 0.3596508502960205,
+      "learning_rate": 4.7434466420860515e-06,
+      "loss": 0.0509,
+      "step": 1248
+    },
+    {
+      "epoch": 5.65158371040724,
+      "grad_norm": 0.3635534942150116,
+      "learning_rate": 4.735559950908933e-06,
+      "loss": 0.0577,
+      "step": 1249
+    },
+    {
+      "epoch": 5.656108597285068,
+      "grad_norm": 0.3797168731689453,
+      "learning_rate": 4.727673919448271e-06,
+      "loss": 0.0629,
+      "step": 1250
+    },
+    {
+      "epoch": 5.660633484162896,
+      "grad_norm": 0.3910801410675049,
+      "learning_rate": 4.7197885673778816e-06,
+      "loss": 0.0746,
+      "step": 1251
+    },
+    {
+      "epoch": 5.665158371040724,
+      "grad_norm": 0.35556650161743164,
+      "learning_rate": 4.711903914369885e-06,
+      "loss": 0.0784,
+      "step": 1252
+    },
+    {
+      "epoch": 5.669683257918552,
+      "grad_norm": 0.3415828347206116,
+      "learning_rate": 4.704019980094659e-06,
+      "loss": 0.0533,
+      "step": 1253
+    },
+    {
+      "epoch": 5.67420814479638,
+      "grad_norm": 0.34802836179733276,
+      "learning_rate": 4.696136784220785e-06,
+      "loss": 0.0734,
+      "step": 1254
+    },
+    {
+      "epoch": 5.678733031674208,
+      "grad_norm": 0.34953394532203674,
+      "learning_rate": 4.688254346415006e-06,
+      "loss": 0.0618,
+      "step": 1255
+    },
+    {
+      "epoch": 5.683257918552036,
+      "grad_norm": 0.3640146255493164,
+      "learning_rate": 4.680372686342173e-06,
+      "loss": 0.0708,
+      "step": 1256
+    },
+    {
+      "epoch": 5.6877828054298645,
+      "grad_norm": 0.5030450224876404,
+      "learning_rate": 4.672491823665193e-06,
+      "loss": 0.0661,
+      "step": 1257
+    },
+    {
+      "epoch": 5.6923076923076925,
+      "grad_norm": 0.3879057466983795,
+      "learning_rate": 4.664611778044988e-06,
+      "loss": 0.0769,
+      "step": 1258
+    },
+    {
+      "epoch": 5.6968325791855206,
+      "grad_norm": 0.353974848985672,
+      "learning_rate": 4.656732569140441e-06,
+      "loss": 0.0788,
+      "step": 1259
+    },
+    {
+      "epoch": 5.701357466063349,
+      "grad_norm": 0.3681681156158447,
+      "learning_rate": 4.648854216608346e-06,
+      "loss": 0.0546,
+      "step": 1260
+    },
+    {
+      "epoch": 5.705882352941177,
+      "grad_norm": 0.36380332708358765,
+      "learning_rate": 4.640976740103363e-06,
+      "loss": 0.0619,
+      "step": 1261
+    },
+    {
+      "epoch": 5.710407239819005,
+      "grad_norm": 0.3659432530403137,
+      "learning_rate": 4.6331001592779615e-06,
+      "loss": 0.0613,
+      "step": 1262
+    },
+    {
+      "epoch": 5.714932126696833,
+      "grad_norm": 0.3356708884239197,
+      "learning_rate": 4.625224493782382e-06,
+      "loss": 0.0357,
+      "step": 1263
+    },
+    {
+      "epoch": 5.719457013574661,
+      "grad_norm": 0.3507138192653656,
+      "learning_rate": 4.61734976326458e-06,
+      "loss": 0.0715,
+      "step": 1264
+    },
+    {
+      "epoch": 5.723981900452489,
+      "grad_norm": 0.3713688552379608,
+      "learning_rate": 4.609475987370177e-06,
+      "loss": 0.0763,
+      "step": 1265
+    },
+    {
+      "epoch": 5.728506787330317,
+      "grad_norm": 0.3401196599006653,
+      "learning_rate": 4.6016031857424135e-06,
+      "loss": 0.0674,
+      "step": 1266
+    },
+    {
+      "epoch": 5.733031674208145,
+      "grad_norm": 0.3747742772102356,
+      "learning_rate": 4.593731378022099e-06,
+      "loss": 0.0655,
+      "step": 1267
+    },
+    {
+      "epoch": 5.737556561085973,
+      "grad_norm": 0.3921152353286743,
+      "learning_rate": 4.585860583847566e-06,
+      "loss": 0.079,
+      "step": 1268
+    },
+    {
+      "epoch": 5.742081447963801,
+      "grad_norm": 0.4036964178085327,
+      "learning_rate": 4.577990822854615e-06,
+      "loss": 0.0728,
+      "step": 1269
+    },
+    {
+      "epoch": 5.746606334841629,
+      "grad_norm": 0.36047136783599854,
+      "learning_rate": 4.570122114676469e-06,
+      "loss": 0.0569,
+      "step": 1270
+    },
+    {
+      "epoch": 5.751131221719457,
+      "grad_norm": 0.3767181932926178,
+      "learning_rate": 4.562254478943729e-06,
+      "loss": 0.0691,
+      "step": 1271
+    },
+    {
+      "epoch": 5.755656108597285,
+      "grad_norm": 0.3421269953250885,
+      "learning_rate": 4.554387935284317e-06,
+      "loss": 0.0587,
+      "step": 1272
+    },
+    {
+      "epoch": 5.760180995475113,
+      "grad_norm": 0.3528580367565155,
+      "learning_rate": 4.546522503323429e-06,
+      "loss": 0.0531,
+      "step": 1273
+    },
+    {
+      "epoch": 5.764705882352941,
+      "grad_norm": 0.3879212737083435,
+      "learning_rate": 4.53865820268349e-06,
+      "loss": 0.0752,
+      "step": 1274
+    },
+    {
+      "epoch": 5.769230769230769,
+      "grad_norm": 0.34203389286994934,
+      "learning_rate": 4.530795052984104e-06,
+      "loss": 0.0619,
+      "step": 1275
+    },
+    {
+      "epoch": 5.773755656108597,
+      "grad_norm": 0.32171866297721863,
+      "learning_rate": 4.522933073842001e-06,
+      "loss": 0.0615,
+      "step": 1276
+    },
+    {
+      "epoch": 5.778280542986425,
+      "grad_norm": 0.409665584564209,
+      "learning_rate": 4.51507228487099e-06,
+      "loss": 0.0706,
+      "step": 1277
+    },
+    {
+      "epoch": 5.782805429864253,
+      "grad_norm": 0.341505765914917,
+      "learning_rate": 4.507212705681914e-06,
+      "loss": 0.0503,
+      "step": 1278
+    },
+    {
+      "epoch": 5.787330316742081,
+      "grad_norm": 0.34372374415397644,
+      "learning_rate": 4.499354355882595e-06,
+      "loss": 0.0685,
+      "step": 1279
+    },
+    {
+      "epoch": 5.791855203619909,
+      "grad_norm": 0.35839352011680603,
+      "learning_rate": 4.49149725507779e-06,
+      "loss": 0.0768,
+      "step": 1280
+    },
+    {
+      "epoch": 5.796380090497737,
+      "grad_norm": 0.331247478723526,
+      "learning_rate": 4.483641422869139e-06,
+      "loss": 0.0563,
+      "step": 1281
+    },
+    {
+      "epoch": 5.800904977375565,
+      "grad_norm": 0.3905595541000366,
+      "learning_rate": 4.475786878855118e-06,
+      "loss": 0.08,
+      "step": 1282
+    },
+    {
+      "epoch": 5.8054298642533935,
+      "grad_norm": 0.37202346324920654,
+      "learning_rate": 4.467933642630989e-06,
+      "loss": 0.0612,
+      "step": 1283
+    },
+    {
+      "epoch": 5.8099547511312215,
+      "grad_norm": 0.3457815945148468,
+      "learning_rate": 4.460081733788749e-06,
+      "loss": 0.064,
+      "step": 1284
+    },
+    {
+      "epoch": 5.8144796380090495,
+      "grad_norm": 0.3689752221107483,
+      "learning_rate": 4.452231171917087e-06,
+      "loss": 0.055,
+      "step": 1285
+    },
+    {
+      "epoch": 5.819004524886878,
+      "grad_norm": 0.3966318666934967,
+      "learning_rate": 4.44438197660133e-06,
+      "loss": 0.0654,
+      "step": 1286
+    },
+    {
+      "epoch": 5.823529411764706,
+      "grad_norm": 0.34534600377082825,
+      "learning_rate": 4.436534167423395e-06,
+      "loss": 0.0671,
+      "step": 1287
+    },
+    {
+      "epoch": 5.828054298642534,
+      "grad_norm": 0.3827734887599945,
+      "learning_rate": 4.4286877639617435e-06,
+      "loss": 0.0857,
+      "step": 1288
+    },
+    {
+      "epoch": 5.832579185520362,
+      "grad_norm": 0.3259190022945404,
+      "learning_rate": 4.420842785791326e-06,
+      "loss": 0.0556,
+      "step": 1289
+    },
+    {
+      "epoch": 5.83710407239819,
+      "grad_norm": 0.3990061581134796,
+      "learning_rate": 4.412999252483542e-06,
+      "loss": 0.0803,
+      "step": 1290
+    },
+    {
+      "epoch": 5.841628959276018,
+      "grad_norm": 0.37944766879081726,
+      "learning_rate": 4.405157183606182e-06,
+      "loss": 0.0729,
+      "step": 1291
+    },
+    {
+      "epoch": 5.846153846153846,
+      "grad_norm": 0.362619012594223,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.0605,
+      "step": 1292
+    },
+    {
+      "epoch": 5.850678733031674,
+      "grad_norm": 0.3391781747341156,
+      "learning_rate": 4.38947751739559e-06,
+      "loss": 0.0606,
+      "step": 1293
+    },
+    {
+      "epoch": 5.855203619909502,
+      "grad_norm": 0.33791670203208923,
+      "learning_rate": 4.381639959179482e-06,
+      "loss": 0.0655,
+      "step": 1294
+    },
+    {
+      "epoch": 5.859728506787331,
+      "grad_norm": 0.3713356852531433,
+      "learning_rate": 4.373803943627946e-06,
+      "loss": 0.0553,
+      "step": 1295
+    },
+    {
+      "epoch": 5.864253393665159,
+      "grad_norm": 0.33650147914886475,
+      "learning_rate": 4.36596949029002e-06,
+      "loss": 0.068,
+      "step": 1296
+    },
+    {
+      "epoch": 5.868778280542987,
+      "grad_norm": 0.39711567759513855,
+      "learning_rate": 4.358136618710844e-06,
+      "loss": 0.0789,
+      "step": 1297
+    },
+    {
+      "epoch": 5.873303167420815,
+      "grad_norm": 0.416520893573761,
+      "learning_rate": 4.350305348431612e-06,
+      "loss": 0.0667,
+      "step": 1298
+    },
+    {
+      "epoch": 5.877828054298643,
+      "grad_norm": 0.37820732593536377,
+      "learning_rate": 4.342475698989524e-06,
+      "loss": 0.0515,
+      "step": 1299
+    },
+    {
+      "epoch": 5.882352941176471,
+      "grad_norm": 0.388907253742218,
+      "learning_rate": 4.334647689917734e-06,
+      "loss": 0.0599,
+      "step": 1300
+    },
+    {
+      "epoch": 5.886877828054299,
+      "grad_norm": 0.3462461233139038,
+      "learning_rate": 4.326821340745304e-06,
+      "loss": 0.0792,
+      "step": 1301
+    },
+    {
+      "epoch": 5.891402714932127,
+      "grad_norm": 0.3591066896915436,
+      "learning_rate": 4.318996670997159e-06,
+      "loss": 0.0595,
+      "step": 1302
+    },
+    {
+      "epoch": 5.895927601809955,
+      "grad_norm": 0.38075944781303406,
+      "learning_rate": 4.31117370019403e-06,
+      "loss": 0.0688,
+      "step": 1303
+    },
+    {
+      "epoch": 5.900452488687783,
+      "grad_norm": 0.39448288083076477,
+      "learning_rate": 4.303352447852412e-06,
+      "loss": 0.0901,
+      "step": 1304
+    },
+    {
+      "epoch": 5.904977375565611,
+      "grad_norm": 0.3826497495174408,
+      "learning_rate": 4.295532933484509e-06,
+      "loss": 0.0745,
+      "step": 1305
+    },
+    {
+      "epoch": 5.909502262443439,
+      "grad_norm": 0.34782737493515015,
+      "learning_rate": 4.287715176598194e-06,
+      "loss": 0.0619,
+      "step": 1306
+    },
+    {
+      "epoch": 5.914027149321267,
+      "grad_norm": 0.3616624176502228,
+      "learning_rate": 4.279899196696953e-06,
+      "loss": 0.0722,
+      "step": 1307
+    },
+    {
+      "epoch": 5.918552036199095,
+      "grad_norm": 0.3477957546710968,
+      "learning_rate": 4.2720850132798375e-06,
+      "loss": 0.0705,
+      "step": 1308
+    },
+    {
+      "epoch": 5.923076923076923,
+      "grad_norm": 0.3332456350326538,
+      "learning_rate": 4.264272645841419e-06,
+      "loss": 0.0538,
+      "step": 1309
+    },
+    {
+      "epoch": 5.927601809954751,
+      "grad_norm": 0.34810853004455566,
+      "learning_rate": 4.256462113871741e-06,
+      "loss": 0.071,
+      "step": 1310
+    },
+    {
+      "epoch": 5.932126696832579,
+      "grad_norm": 0.39874809980392456,
+      "learning_rate": 4.2486534368562645e-06,
+      "loss": 0.0814,
+      "step": 1311
+    },
+    {
+      "epoch": 5.9366515837104075,
+      "grad_norm": 0.3651580214500427,
+      "learning_rate": 4.240846634275822e-06,
+      "loss": 0.0521,
+      "step": 1312
+    },
+    {
+      "epoch": 5.9411764705882355,
+      "grad_norm": 0.37812861800193787,
+      "learning_rate": 4.233041725606573e-06,
+      "loss": 0.0438,
+      "step": 1313
+    },
+    {
+      "epoch": 5.9457013574660635,
+      "grad_norm": 0.3588254153728485,
+      "learning_rate": 4.225238730319949e-06,
+      "loss": 0.078,
+      "step": 1314
+    },
+    {
+      "epoch": 5.950226244343892,
+      "grad_norm": 0.34432655572891235,
+      "learning_rate": 4.217437667882611e-06,
+      "loss": 0.0602,
+      "step": 1315
+    },
+    {
+      "epoch": 5.95475113122172,
+      "grad_norm": 0.33063921332359314,
+      "learning_rate": 4.209638557756396e-06,
+      "loss": 0.0564,
+      "step": 1316
+    },
+    {
+      "epoch": 5.959276018099548,
+      "grad_norm": 0.4011344909667969,
+      "learning_rate": 4.2018414193982695e-06,
+      "loss": 0.0671,
+      "step": 1317
+    },
+    {
+      "epoch": 5.963800904977376,
+      "grad_norm": 0.40701520442962646,
+      "learning_rate": 4.1940462722602815e-06,
+      "loss": 0.0851,
+      "step": 1318
+    },
+    {
+      "epoch": 5.968325791855204,
+      "grad_norm": 0.374459445476532,
+      "learning_rate": 4.186253135789511e-06,
+      "loss": 0.076,
+      "step": 1319
+    },
+    {
+      "epoch": 5.972850678733032,
+      "grad_norm": 0.3491608798503876,
+      "learning_rate": 4.178462029428023e-06,
+      "loss": 0.0718,
+      "step": 1320
+    },
+    {
+      "epoch": 5.97737556561086,
+      "grad_norm": 0.3967466354370117,
+      "learning_rate": 4.170672972612816e-06,
+      "loss": 0.0753,
+      "step": 1321
+    },
+    {
+      "epoch": 5.981900452488688,
+      "grad_norm": 0.3280808627605438,
+      "learning_rate": 4.162885984775777e-06,
+      "loss": 0.0594,
+      "step": 1322
+    },
+    {
+      "epoch": 5.986425339366516,
+      "grad_norm": 0.4042797088623047,
+      "learning_rate": 4.15510108534363e-06,
+      "loss": 0.0878,
+      "step": 1323
+    },
+    {
+      "epoch": 5.990950226244344,
+      "grad_norm": 0.36680734157562256,
+      "learning_rate": 4.1473182937378905e-06,
+      "loss": 0.0672,
+      "step": 1324
+    },
+    {
+      "epoch": 5.995475113122172,
+      "grad_norm": 0.3289303779602051,
+      "learning_rate": 4.139537629374814e-06,
+      "loss": 0.0632,
+      "step": 1325
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.3161666989326477,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.0474,
+      "step": 1326
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2210,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 442,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 272662612566016.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}