diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.501336002076591,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.0374,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.603127269181422,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.0494,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.390476827128191,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.026,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.344372997834995,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.0232,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 5.955313716049971,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.0203,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.871092466781864,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.0172,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.544025456793008,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.9776,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.214362663677726,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.9739,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.706585034732052,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.9448,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.4788989383018474,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.9345,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.4448026860023186,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.9376,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.828966636115953,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.9233,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.214570529021246,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.9308,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.9876944787415796,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.9264,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.7071241518481544,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.9151,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.7526648527366113,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.875,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.512621240973615,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.8589,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.0342814546802015,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.8507,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.50533515109058,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8347,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.1288158504826755,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.8218,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.038025718651718,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8216,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.2046483803919945,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.8087,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.1060900153719064,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.8253,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.9080806649793854,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.7961,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.9329349605984499,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.7857,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.9570043965402277,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.7832,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.831879016527591,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.7744,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.8328620461891418,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.7467,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.7853244560622368,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.747,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.801226529369726,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.7576,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.7554283240011549,
+      "learning_rate": 1e-05,
+      "loss": 0.7513,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.6238770147399739,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.7506,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.649473953024947,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.7411,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.6818004619553264,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7357,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.5868446993770708,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7497,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.5529997856303241,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.725,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6075372125471553,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.721,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.5657687985185882,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.7256,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.5001278856720773,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.6949,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.5917192334577692,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.7051,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.5300108979610939,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.696,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.48256067922382906,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.6908,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.5769727254437385,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.6954,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.576942853426186,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.6993,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.5341431094164741,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.6939,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5694619536985795,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.6869,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.46860414181331367,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.691,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.4299622289893042,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.6948,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.5744194508087261,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.6905,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.4008021628224624,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.7102,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.48979194139824833,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.6905,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.6005622679234734,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.6735,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.5132887902339205,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.6843,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.5487932495992293,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.6858,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.5782293870968532,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.6768,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.5288512793357714,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.6812,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6865065704393304,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.6756,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.6730117435210778,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.6727,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.5964209430869492,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.6773,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.6375603944278386,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.6683,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.6473669575935326,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.6787,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.5766646400420069,
+      "learning_rate": 2e-05,
+      "loss": 0.6611,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.5645258916018057,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.673,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.5941218937446974,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.6729,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.6725271413010109,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.6636,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.6656944243790763,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.6489,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.7863321654028452,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.6613,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.8499358793536573,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.654,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.6552331803102749,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.6529,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.6123017836205994,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.672,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.8980513335604925,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.6593,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.0469522480355413,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.6666,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.9282365607152543,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.6499,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7642505838701882,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.6629,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.7473950275263416,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.6582,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.7266415570465198,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.6599,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.8931207060421503,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.6445,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.0653650519596147,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.6686,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.8093948717733653,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.6491,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.021193644061852,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.6476,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.393504401052146,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.6531,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.8657372861034439,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.6512,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.4078308451942245,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.6705,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 1.1511664026805637,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.6576,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.9920034658233652,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.6493,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.0513185074130087,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.6548,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 1.3768318263406591,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.6332,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.8511732604098855,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.6416,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.6959644785469117,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.6549,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.7477945078632898,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.6692,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.505831132014918,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.6385,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.7910391998729247,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.6238,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.5918353771636815,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6446,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.9472226502653913,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.6363,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.2928469170524732,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.6369,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.2836852557586935,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.6534,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.1187368424581636,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.6638,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.4058710746408583,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.6541,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.6882011013660615,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.6467,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.377793420448162,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.6349,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.7721740141397605,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.6354,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.241297521479032,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.6521,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.1245131993526902,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.6543,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.9271034911061992,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.6278,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.4297626039912292,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.6403,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.1693585563731375,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.6384,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.2468915779612153,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.6493,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.9481247727311128,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.6527,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.0300861371463794,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.6396,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.8434181308550983,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.6332,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.7943732572023293,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.6182,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.0346254024414947,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.6355,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.794706789659031,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.6516,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.9162589076526716,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.6436,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.2999187959554206,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.6535,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.5909379892372442,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.6299,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.8575735449305615,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.6396,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.469756383965448,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.6405,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.9547386723052693,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.641,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.3474828926229374,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.6345,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.1865690746173445,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.6292,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.3024417228570773,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.6277,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.9550193691941039,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.6428,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 1.2781738615134308,
+      "learning_rate": 4e-05,
+      "loss": 0.6394,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.9403656528851696,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.6418,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.2023623588268089,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.6283,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.7949538044738175,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.6351,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.3287424324238,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.6433,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.8368764277164075,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.6361,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.4907381117219332,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.6264,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.846289008486981,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.6344,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.221572967072379,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.6247,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.6744374425776652,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.6323,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.8741375665944096,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.6231,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.7629945092554689,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.6251,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 1.1029445268137403,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.6266,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.7312680098303111,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.617,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.7202002294256711,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.6444,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.8310247475687784,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.6427,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 1.1374946556663053,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.6218,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.9773983804635239,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.6079,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.1891758712199745,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.6274,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.9123711431296269,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.6156,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.7399330806738491,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.6226,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.7083628557811023,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.6219,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.7902444017288264,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.6254,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.8508104615217651,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.6312,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.7224565014152766,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.616,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.9079365123447389,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.609,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.9588435395850354,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.6353,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.6602866203957769,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.6168,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.6396718532430083,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.6512,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.7940693560708856,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.6284,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.7660735376671776,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.6145,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.6817483081407721,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.6241,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.7573658860170402,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.6258,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.8314071599807082,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.6166,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.8307786623963549,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.6109,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.8321110713255597,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.6186,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.8126735753640898,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.6219,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.7949347153804247,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.6097,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.5703554213966108,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.6088,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.5839624906867992,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.6071,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.6064829063108845,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.6216,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.7274541352492694,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.6098,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.9025353782305422,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.6143,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.8405215837199052,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.5968,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.6706953464321106,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.6238,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.4692676701282885,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.6308,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.6101528489101323,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.6203,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.7688049138285785,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.6344,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.9168439644595777,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.6118,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.6737998673576476,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.6174,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.5118308300646365,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.6182,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.7340542428492554,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.6063,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.862023274765519,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.6169,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.6218496742395199,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.6142,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6148551062704386,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.6243,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.6860580055473009,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.6213,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.6095316170617261,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.6205,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.6383662263090235,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.6079,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.5543644691042288,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.6042,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.6100346525040897,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.6263,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.7207419997797473,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.6153,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.7052285772262556,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.6256,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.5986959622725678,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.6159,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.5506479166138359,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.6191,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.5691261203342551,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.6112,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.6637721977156701,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.6155,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.8131976380154874,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.5956,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.6033537291945624,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.6105,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.7900197266198759,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.6123,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 1.0658014983920785,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.6046,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.6936650601979,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.6127,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.4825307931875446,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.6009,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.5636451958572147,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.6079,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.5592638968981687,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.6183,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.467545502144663,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.5959,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.4465810134139462,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.6144,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.5952111501200492,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.6148,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.6143582272143587,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.5978,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.49244889554899257,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.6059,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.4707966777642392,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.6134,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.5779042850975099,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.6166,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.5377628110002729,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.6083,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.4425197593734987,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.6032,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.5220456070665787,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.5945,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.6128790074006477,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.6157,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.673143350330518,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.6034,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.6445006430415718,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.5985,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.6457885318986645,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.6079,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.5935933045419325,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.5939,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.471866152473157,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.6021,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.4934849926473707,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.5985,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.7057354855927641,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.5998,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.7633761086814128,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.6124,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.7916000672525193,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.5929,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5750472798061009,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.6149,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.5709206939883914,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.6021,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.6316632686874118,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.6078,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.6281242686977094,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.6024,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.5462012519214096,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.6127,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.4210626513541926,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.6056,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.5299953077804459,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.6075,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.503794922290119,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.5933,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.6580339071800156,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.6142,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.6814187071942606,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.5991,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6957284098007691,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.6341,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.5271940679305018,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.5993,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.48424340135019533,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.5942,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.528559892284716,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.6072,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.486238409557306,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.5947,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5059601165307738,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.5901,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.5016160632395985,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.6043,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.44089762243458314,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.5949,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.486037823958588,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.5943,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.4312480969734151,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.6105,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.4887395250131878,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.5813,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.4344116620675808,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.595,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.4614126668266943,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.6051,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.44881477386667307,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.6004,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.6120189804187672,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.6037,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.5080534438398463,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.6065,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.5241427695754926,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.6017,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.4561166471955584,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.5943,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.3996995924734061,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.5934,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6546652217498222,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.6016,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.47338533848122066,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.5451,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.521967839245862,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.535,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.5175268180356786,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.5546,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.5225338626832894,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.5445,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.460730896835714,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.5542,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.5155685256492357,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.547,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.5894562396017856,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.5672,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.8756586497154124,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.5875,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.8352042566109957,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.5578,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.8004991720277312,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.5594,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.6379424715452059,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.5364,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.4444937846436943,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.5515,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.49077907235968354,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5529,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.4724161962461811,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.5609,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.586139352395267,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.5395,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.4989815330530784,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.5462,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.6295395954896866,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5609,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.48152487869893684,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.5681,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.4306788564546898,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.5493,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.43232105715448443,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.5521,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.44401921825812873,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5718,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.41212889722245777,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.5541,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.4561848537917732,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5554,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.418803863895386,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.5444,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.396023944570863,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.5517,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.35611938231005885,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.546,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.44301532851603753,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5542,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.392961127199416,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5524,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.38746556448928954,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.5375,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.45803983173862245,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.5354,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.6643074652811681,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5887,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.4182223358522735,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.5563,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.5024539637755465,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5569,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.5672927995862945,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.5643,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.5210090520363159,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.5427,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.5125595705791086,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.5429,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.4507242585218264,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.5388,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.3446212195352754,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5473,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.468492031837788,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.5582,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.47581066853126314,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.5581,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.4782128383503499,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5462,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.44597861410110173,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5731,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.40056511922566285,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5549,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.6065699106248879,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.56,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.5506606939004168,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.5595,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.4357774442279925,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5565,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5506546336208802,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.5538,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.6523346069129353,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5305,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.537235988247275,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.551,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.5472199787969219,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.554,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.6335632165474292,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.5491,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.6378935293711532,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.5462,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.7079054307921262,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5511,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.8212371957251346,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5413,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.658939306444265,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.5579,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.467004155198878,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.5981,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.8218219038532556,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.5574,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.8399001164284492,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.5521,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.6239188596941861,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.5622,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.42929022135486483,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.5542,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.48666998873989653,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.5525,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.5162086187265852,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.5553,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.45032598683571273,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.5564,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.4341353039722502,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5545,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.5497473305461197,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5475,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.5508573308849614,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.5359,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5448300871860186,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5492,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4745061967053439,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.5605,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.5440925240076461,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.5472,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.47527864710574896,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.5463,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.4520521630351715,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.54,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.5264382581396784,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5573,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5247472899948156,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.5509,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.4618058120991241,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5347,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.5142883462544702,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.5522,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.4500378355876767,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.5474,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.40747772831792284,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.5524,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.4990887230049701,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.553,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.39580666043354923,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.5408,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.4568633191595009,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5694,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.4997169928143083,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.5374,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.39419443459143566,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.5397,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.45557243273156744,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.5471,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.47957471676838886,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5624,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.40370159976251485,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.5396,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.5153630428198558,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.5475,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.42524882041430795,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.5494,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.33519893425310476,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.5482,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.4754831316840997,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.5478,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.5501590016590703,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.5506,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.48989536645833026,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5408,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.5005026246594406,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.5543,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.5131250425444224,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.5542,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.601684706685955,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.55,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.6786243026416414,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.5525,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.5040160265128553,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.549,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.5521764043535455,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.5477,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.6062956585159838,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.5435,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.5326912048118194,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5473,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.4857337249062357,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.5582,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.3979606816383933,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.5492,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.48411996177838995,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.5588,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.42214303098406125,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.551,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.4964519538018437,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.5492,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.4132793985815599,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5622,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.41171325333725,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5369,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.4446604606890363,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.543,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.34415724083875726,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.5395,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.42761047866358387,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.5522,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.44257076988490324,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5476,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.37499651231039105,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.5354,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.41621381062502366,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5466,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.39588365235638046,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.559,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.38449203348513933,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.5464,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.3899325871444137,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5433,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.3809301336806042,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5362,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.45051511581927023,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.553,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.4965784493176186,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.5519,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.5409330038568996,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.5455,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.34761463313599544,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5518,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.4649367233204076,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.5387,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.45036446287585097,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5504,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.37436334323071524,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5524,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.4157817923586303,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.5532,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.3449915467010953,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.5495,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.4370465402912321,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.548,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.44714234665160124,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.5341,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.3390088421568354,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.5475,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.5137941898011829,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.5491,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.3898265164757896,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5481,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.4302847825575368,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.5495,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.599124545959164,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.5512,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.4042714847391145,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.5439,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.4883065222836041,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.5478,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.5037086359697386,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5454,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.362150688430488,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.5548,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.4397624699733895,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5595,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.41618215505990847,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.5573,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.31087525027492613,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5471,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.4607388219382896,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5493,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.399055306470891,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.5414,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.39560262587393347,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.5405,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.4518765836318716,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.5506,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.44378327736878476,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.5393,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.3961926533473978,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5392,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.43337590813533633,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.5542,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.45859273201021183,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.5549,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.4482343388684534,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.549,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.3478560029462995,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.5553,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.4493663853781007,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.5421,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.4286063468766196,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.547,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.3961035712982849,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.538,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.38553303420130025,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.5417,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.4337339325892025,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.5615,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.46691552562135896,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.5488,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.40610426184957543,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.556,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.4203135495379824,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.5505,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.4202314249755647,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.5453,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.37771936629331543,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5506,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.3900164504175787,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.5555,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.3404341248272031,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.5347,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.3832055022177208,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.5412,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.4028751244788782,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.5417,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.3722422127959474,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.5363,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.36243249878558753,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.543,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.3529855269830736,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.5434,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.36135380098071856,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5378,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.4177458995568904,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.554,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.3808021859383369,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.5426,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.43435621731104057,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.5447,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.38334676175779053,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5521,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.3405837262283973,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.5487,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.4383480714379587,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.5558,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.38066530120752495,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.5364,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.3505657750540238,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.5434,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.4480319677487732,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.5536,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.4293103213852662,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.5327,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.3551547994680098,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.5473,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.36742759229111144,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.5365,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.5076184504858816,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.5519,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.3712499921790561,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.526,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.39280847688588905,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.5425,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.37867755382883017,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.5493,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.3577071160808504,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.553,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.3875833310413329,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.5484,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.4195555171855959,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.5401,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.3539479846539754,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.5366,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.38493883990851036,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.5338,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.42835660503242734,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.5423,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.3793038513987804,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.5446,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.3796766318708962,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5346,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.38960191584586845,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.546,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.45198643435474245,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.5423,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.4174073901846527,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.5332,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.35687499765455055,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.5624,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.4863380913486964,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.5396,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.3305009926015577,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.5373,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.4245063336434092,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.5567,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.44656205780360114,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.5596,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.4128007515350705,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.5432,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.3605803426957879,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.5382,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.3961998944153289,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5445,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.3321572400594232,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.531,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.36020684234442923,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.5441,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.3145312010541274,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.5508,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.3736538366140589,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.5447,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.3425166309768136,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.5311,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.3504965502152043,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.5428,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.41908698338437267,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.5453,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.3054715823973169,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.5494,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.3886113290092192,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.5479,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.34692187387797335,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.5332,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.3532459670336893,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.5494,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.36038862070952943,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.5295,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.3721063332338532,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.5317,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.38492040976500497,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5414,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.4020254384510381,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.546,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3615198951549212,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.5369,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.4365005288165114,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.5293,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.46455872210833254,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.5495,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.3736336979072726,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.5453,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.4777898559432648,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.5355,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.3643426787830856,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.5344,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.42382908751565523,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5365,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.38871688012129807,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5414,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.4111125375225109,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5397,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.3559953164678273,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.5356,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.29326220394195146,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.545,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.3995655972094481,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.5286,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.32656060913300683,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.5397,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.37879673167447725,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.5415,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.335076260865658,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5367,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.34152473216821483,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5429,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.30543414367865096,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5378,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.3449139297930993,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.5335,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.29870983514902155,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.5367,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.30032177125421466,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.5334,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.32371772575676894,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.5388,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.3090828899168418,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5586,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.31737877304410983,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.5432,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.34401180068628495,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.5375,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.3178066066847312,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.5481,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.2912063376589897,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.5416,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.31189621563355674,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.5487,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.3022551740685457,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.5443,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.32686012388575036,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.5348,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3218674765776564,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5347,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.44523705855399254,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.4869,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.3599689936498873,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.4759,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.5603284862428709,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.4862,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.5034941855562995,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.4675,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.5070967138131882,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.486,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.46582035107908065,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.4733,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.5929957133377151,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.4828,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.45058637715300565,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.4634,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.43254488546426806,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.4728,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.4409001803065222,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.4904,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.37291571782375404,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.481,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.43923706357549247,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.4761,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 1.29451336341998,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.477,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.4319451770682443,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.4835,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.5037597765392053,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.4636,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.49461865585645615,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.4919,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.3731371220806982,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.4828,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.4054573769760416,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.4833,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.41229477215371235,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.4968,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.39117942029379177,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.4753,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.3408766112597396,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.4857,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.43299231978898234,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.463,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.34213347340836037,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.4925,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.3598102585553423,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.4684,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3804486708481998,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.4836,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.29840395281604115,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.4726,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.43119767577278834,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.4842,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.3616213379674547,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.481,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.36185255607112077,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.4742,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.34853045205194,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.4829,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.32881393171975826,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.4825,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.33580444250238406,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.4843,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.3483477861352959,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.4846,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.3641102901109439,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.4735,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3029697370975726,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.4832,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.33786841112886795,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.4858,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.3331598324017113,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.4928,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.4518747128992889,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.4794,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3398452352763984,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.4852,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3206473690840172,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.478,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.368516150541396,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.4806,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.31406248425369093,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.479,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.3784052030282786,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.4776,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.27464202476949906,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.4838,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.37302127404354085,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.4814,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.2978719961190508,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.4804,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.30617316848754556,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.4771,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.2983777317632612,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.5037,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3149106496674001,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.4811,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.2934670174070717,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.4751,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.3194391653904097,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.4819,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.3265203092935218,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.4896,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.30377917215999417,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.477,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.3016086384351667,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.4791,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.2943963518749922,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.4708,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.29827142100569703,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.4871,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.2972320012276692,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.4898,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.30207524290542614,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.4689,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.32188307037351876,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.5045,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.3818420734650697,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.4887,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.28827469245831616,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.4865,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.3755384360689266,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.4769,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.3329208881631146,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.4858,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.3922352633777127,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.4832,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.3298515323755388,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.4787,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.295589324528678,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.4875,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.3235214348342221,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.4799,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.3074333873340973,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.4786,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.34903482816838327,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.4714,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.2710207571592716,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.4857,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.32904818268631597,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.4719,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.2850060128535107,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.4746,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.30788268852775524,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.4774,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.2916576723214955,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.4838,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.308306630403208,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.4838,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.33493189810003793,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.4815,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.31462853250631345,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.4804,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.3324907556414214,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.4835,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.38548710538302244,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.4806,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.29723416078065784,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.4942,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3161939307665217,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.4885,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.3941655292106356,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.4799,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.3766250240090942,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.4792,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.29879945418215303,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.4825,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.4160621490951331,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.4856,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.34220619071514013,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.4877,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.3480634819259696,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.4808,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.4040256810170426,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.4842,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.30571053271864146,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.4779,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3613618097862634,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.4891,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.39869098715970547,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.487,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.2608016412306887,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.4798,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.3816354908383628,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.4747,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.2757701942428882,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.4825,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.3444201427855753,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.4763,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.3694999314700325,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.4826,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.3124523056175555,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.4808,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.29870649798328686,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.471,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.29576436332726747,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.4807,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.3005563449005866,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.4919,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3107154829119454,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.4928,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.2875431056730711,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.4777,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.34278207339656513,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.5071,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.29707441035680704,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.486,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.34908576688849247,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.4829,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.33655016878426114,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.4684,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.31311312274529624,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.4748,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.40207148506389995,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.4848,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.3388598242009374,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.483,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.3221774730474711,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.487,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.36718875390641614,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.4856,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.3239433206743579,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.4845,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.3252536465739171,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.4834,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.2883791110876588,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.4731,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.30485012256057115,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.4882,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.32658954895157133,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.4808,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.3257606565774221,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.4932,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.2928302639772323,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.4599,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.3211873710284641,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.4893,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.2853316183744641,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.488,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.31927081774735905,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.4862,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.2764958528878502,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.4984,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.3183467548848915,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.4772,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.3223922294891695,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.4833,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.28935679258785524,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.481,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.310323982348667,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.4742,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.33924361174118417,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.472,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.3007260306530266,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.4776,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.34381335947752456,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.4844,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.29624252441132787,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.484,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.29333337750884536,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.4764,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.30873885482653,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.4716,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.2472736398798577,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.4829,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.28378130217310826,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.4813,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.2536900619091271,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.4794,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.3054712294431091,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.4858,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.3153630490242766,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.5038,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.32269657559996273,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.4846,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.31447730534796237,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.4796,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.29958565274099824,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.4901,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.35809255615860414,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.479,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.28984975632670923,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.4911,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.3422164106769867,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.4874,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.2769787915819864,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.4935,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.28311255967785437,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.4823,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.34792231111042776,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.4891,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.30550388693510283,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.4771,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.3416577806061086,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.4795,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.2939685164180667,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.483,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.306391504963146,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.4794,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3137265483805074,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.4851,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.295761598964308,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.4813,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.3177896588743363,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.4913,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.3545205206719438,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.4749,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.3099506991232152,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.4907,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.3003561764927816,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.4835,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.2855821182438034,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.4732,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.3126853343924098,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.4783,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.29465108670011303,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.4746,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.3209515284521271,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.4867,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.3196996450946175,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.4862,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.27776074408707857,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.4924,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.3118840238357117,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.493,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.24784656091956575,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.4767,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.2865727076690227,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.4944,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.26889698305646886,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.4856,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.3423374980262315,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.4853,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.3397759116199281,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.4723,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.3179281628392728,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.4855,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.37071793740898923,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.487,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.2974901185610592,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.4783,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.331925939352178,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.4762,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.2982258044880553,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.4812,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.37096452423822013,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.4822,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.2596277109148044,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.478,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.30576548064940556,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.4704,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.25013373711658893,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.4776,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.2888149635012136,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.4657,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.27128128016119735,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.4665,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.3106490375485201,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.4906,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.28316137237546674,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.4676,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.24910339076639976,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.4761,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.31752209012950466,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.4921,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.26727909098826763,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.4735,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.35011895741395294,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.4969,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.27774522586541966,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.4803,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.3479271121648441,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.4735,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.2909576236066905,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.4823,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.31200062044596505,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.4819,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.2916692043014418,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.4806,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.2813863461401582,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.4703,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.2776069943226141,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.493,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.27087656999076404,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.4805,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.2659110072576431,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.4885,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.2645064619634614,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.4672,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.2881898708686908,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.471,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.27536110316489376,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.4748,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.2707363391966337,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.4843,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.31416598409826185,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.4807,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.26091623733265273,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.4896,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.27685502474654894,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.4844,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.2950174075775018,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.4834,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.2804086174478964,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.4861,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.2835377216936194,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.4773,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.27222514828756356,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.481,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.27975713497075855,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.4736,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.25569308683247294,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.4659,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3076844795651064,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.4717,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.2621262676871629,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.4745,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.32438231424802305,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.4803,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.2741759527113326,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.4835,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.28823037378668404,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.4783,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.2814202348514938,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.4789,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.26088790548884777,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.4803,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.2979968887932747,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.4925,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.25270740033110756,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.4783,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.31574459044014863,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.4733,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.2641097525472468,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.4875,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.2942358356237042,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.4598,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.27781420449130523,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.4804,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.2979472490315894,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.4859,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.25974143266356015,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.4833,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.30845376036493233,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.4813,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.23891761240897105,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.4652,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.2844305900355645,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.4832,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.2614647669335602,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.4753,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.31688171711183194,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.4751,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.2770931637448464,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.4794,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.3646351821879984,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.4888,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.27279931152908027,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.4769,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.35945324609653234,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.4864,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.2562270640894489,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.4797,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.28364045149167716,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.4889,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.2492040670140214,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.4879,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.2778253786041453,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.4717,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.2438452565272541,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.4766,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.2740725149276371,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.4757,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.2490126796558031,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.4878,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.23730070954255017,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.4855,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.26137608617733643,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.4828,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.25283686963074936,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.4679,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.2611773733451272,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.493,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.2798346576855019,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.4823,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.29298930367359033,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.4818,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.26689920053053173,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.4761,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.2783424861703408,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.4874,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.26472557817598324,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.4753,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.42716546448880505,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.4283,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.3247676469977295,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.4335,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.574272577585603,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.4292,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.4139621799247635,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.4177,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.46453561850939057,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.4301,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.40206670697553254,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.4205,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.381933114368622,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.4245,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.41673867931658076,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.4382,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.40395213340232516,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.4302,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.38939967475554976,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.419,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.396333235214822,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.404,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.3872161056021772,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.424,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.3468275690949465,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.4079,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.3747363846006543,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.4311,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.30028059191685663,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.4064,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.3186483624204562,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.4083,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.34389066688591796,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.4264,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.298137107617706,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.4164,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.29647303780526535,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.4139,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.3008722301666454,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.4182,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.2814930272105444,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.4134,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.29130075759686286,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.4146,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.27540686992119096,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.4295,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.2645070151366103,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.4189,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.2854799700663522,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.43,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.28023162125074563,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.4139,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.2850399912993896,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.4161,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.2588870619232719,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.4269,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.2837664414150749,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.4337,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.285009526664894,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.4111,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.2603820094242142,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.4155,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.25648004642270217,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.4094,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.2633371900695218,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.4116,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.26864279541656416,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.4244,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.24662249108363465,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.4284,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.28273062859768017,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.4198,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.2735963850143561,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.4191,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.2379761987243131,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.4309,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.28984463043709213,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.4248,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.2521221156540418,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.4196,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.26052271617825523,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.4105,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.2856530472690246,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.4321,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.2405566000497225,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.4262,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.27677702354525613,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.4134,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.23363800364562248,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.4278,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.27122612515818034,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.4237,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.24989386141949352,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.4298,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.2639609439578666,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.4158,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.26555262762080417,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.4219,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.25533428912844647,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.4225,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.23601853943336226,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.4288,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.2832821760318479,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.4277,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.25129681670385756,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.4256,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.2580929021101108,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.4306,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.23122460201281825,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.4182,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.25341371582255623,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.4345,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.24207938557878375,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.4259,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.2456338159720481,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.4189,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.2580966296106912,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.4151,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.3424132993013671,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.437,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.2554756280638256,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.4286,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.24113703448082538,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.4278,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.2512520207170782,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.415,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.23554332597317054,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.4307,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.2372779369689775,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.4192,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2716692036525646,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.4245,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2499301916080145,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.4239,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.26497684175777486,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.4115,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.24975355423061574,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.419,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.2562093967422027,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.4318,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.24249390807524152,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.4216,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.25591640368617,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.4204,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.23704705404542153,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.4201,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.27896334059658323,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.4299,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.2505663633107243,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.4232,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.2746116742692593,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.4257,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.247328995577574,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.4281,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.254470751071893,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.4085,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.24974719068943727,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.4321,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.27395148194248736,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.4212,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.32390863221900623,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.4146,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.24421693849883422,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.4175,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.25797833349564303,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.4294,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.26095724268359755,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.4152,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.2734472497912189,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.4202,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2315737794113412,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.4208,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.2491152388561685,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.4181,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.25801876091551973,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.4091,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.23759375284481232,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.4254,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.26765830907928156,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.4202,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.22474066619008304,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.4191,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.2567919907310414,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.4208,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.23773731234696655,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.4168,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.23855394055467735,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.4212,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.24037158409724027,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.4257,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.23180780206579846,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.4239,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.24524438921464164,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.4279,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.2624884483493446,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.4184,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.2620010141451373,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.4196,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.24322436482964246,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.4205,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.2517150531716751,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.4129,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.253417809339605,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.4218,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.2489896962623098,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.4127,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.22003248349075505,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.4167,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.2703727065574688,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.4192,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.23520417265396715,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.4049,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.22746141836773628,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.4248,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.25439380333008177,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.4167,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.22839790625516493,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.4238,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.26825572399542774,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.4284,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.25213055027094633,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.4216,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.23026539551002762,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.4106,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.2667397923118754,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.422,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.2288248645783561,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.4299,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.24633415706831985,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.433,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.2621835744494584,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.4184,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.2272686539826848,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.4197,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.26888296361787795,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.4255,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.22319159272396555,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.4114,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.24786046871731648,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.4369,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.24248747943408508,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.4217,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.22424647199821302,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.4182,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.25800862761631976,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.4221,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.2219294368284842,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.43,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.24460864053611842,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.4185,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.23837236561034347,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.4291,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.23518932755476474,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.4232,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.23651001545921632,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.4314,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.23213118842218736,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.4166,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.24161176185046734,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.4175,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.2277412102184374,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.4157,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.24565268272396262,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.4118,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.22568344279606758,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.4268,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.2384661147460029,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.4229,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2355536989377484,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.429,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.24274058640072294,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.4049,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.23970966084963208,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.4256,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.23678231394762644,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.4213,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.24474601259027734,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.4156,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.23136019173971867,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.4182,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.2333032080611408,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.4281,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.22309756159901706,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.4056,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.22697767973374391,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.4073,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.2398001488953483,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.4297,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.2203587824908215,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.4212,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.23470797646656713,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.4097,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.2339476038368614,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.42,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.2335647944962018,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.4229,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.2181098033031804,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.4321,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2746328153681407,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.4121,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.2359513996536879,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.4042,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.22376799104824974,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.4131,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.23164216231612275,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.4232,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.23529455110232053,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.4219,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.23317879919552179,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.4205,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.22206106839914092,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.4249,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.24188859246112496,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.4165,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.2389768552319045,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.4202,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.21206298169814491,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.4294,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.23806992851109976,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.4205,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.23208322281577654,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.414,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.21238687229469866,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.4265,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.22834622478772995,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.4207,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.22780319136854724,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.4317,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.2132840607981698,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.4212,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.2291458251741146,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.4222,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.2258687304140681,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.4227,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.2392791285171366,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.4235,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.2318032939569747,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.41,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.24099464905103962,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.4143,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.23511381209890714,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.4153,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.22661513231358685,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.4221,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.22589544649719848,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.4092,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.24515332660513237,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.4127,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.23392152403487454,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.4071,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.2473066390794541,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.4023,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.2515377527913947,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.4322,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.2279846757729996,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.4101,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.24607743387117595,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.4089,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.23097127867842374,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.4234,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.22411295988567023,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.4254,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.2543150983782804,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.4088,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.22947593250213624,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.419,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.2296370704567956,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.4191,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.22677778241800042,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.4197,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.21426103646008898,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.4245,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.2133040377351012,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.4219,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.23625300057558643,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.4236,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.2023043236123225,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.4132,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.22664130966801968,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.402,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.21448584271027493,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.4214,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.22002472791691027,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.422,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.21717135413123542,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.4009,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.20752629612362603,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.4184,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2079391480699532,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.42,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.22484847354556586,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.4191,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.21007207342133366,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.4255,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.20822960534514295,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.4138,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.22419928457723898,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.4366,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.23257472786548125,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.414,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.2059831972137516,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.3973,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.21674733828271756,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.4215,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.20882900315691102,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.4217,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.22922844678327925,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.4221,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.22789871074734555,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.4162,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.21972778184239136,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.4347,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.23282857272525193,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.4269,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.22282925442146267,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.4166,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.20401047619987397,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.4111,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.2169863913085385,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.4341,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.2400954148161607,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.4141,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.21891440547321483,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.4143,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.2225289372190161,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.4119,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2365592334667968,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.4203,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.1973280981125247,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.4235,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.2117817118466791,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.4141,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.2065313135957648,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.4145,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.21295797036118982,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.401,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.21965832752958503,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.429,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.2080360302905839,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.4237,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.20745740123391665,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.4244,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.20636438624210054,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.4171,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.22150503090651155,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.4129,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.23042417649577907,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.4222,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.22501461048453256,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.4144,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.23507007653163037,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.4237,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.215031865524335,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.4079,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.2243939164626856,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.4151,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.21268397830120633,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.4153,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.20146991885523657,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.422,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.2166426535349822,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.4085,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.21460068577035885,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.4283,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.20182961000669478,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.4274,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.22296563275075476,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.4244,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.22023892403612924,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.4185,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.21018252705917093,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.4226,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2129681290712287,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.4114,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.21295384650415167,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.4232,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.21217008017508499,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.4166,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.21370861713617614,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.3977,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.20632714523109716,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.4171,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.20789570341231708,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.4371,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.22162285933707807,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.4264,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.2095356021358193,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.4203,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.2122486929606399,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.4172,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.22122458918808668,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.411,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.20158510576107652,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.4122,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.40329413081410664,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.36,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.2947115853914343,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.3798,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.2807413785619938,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.397,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.4711662780354046,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.3822,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.3384826908440774,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.3763,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.290882348880419,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.3677,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.37499662707586073,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.3703,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.3124499566334527,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.3751,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.269145857003835,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.3695,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.2644664620010784,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.3825,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.32021921236778955,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.3791,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.3067448840075858,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.3786,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.2561863389138064,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.3699,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.2508533033776017,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.3676,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.24700642442352336,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.381,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.25048533338323786,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.3814,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.2559300014065233,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.3726,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.23125710515395967,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.3971,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.23671837304490181,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.3825,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.24787808744889686,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.3806,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.23585574124788586,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.3824,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.23489778920630566,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.3819,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.22703996421839184,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.3754,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.21407877594644453,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.3714,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.22533849866214095,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.3676,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.2226388908343112,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.3883,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.2189219800427245,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.3737,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.21906806915105803,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.3792,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.23004352169801592,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.3696,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.21996542327484908,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.3837,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.2070180222551748,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.3741,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.22011412308496364,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.3836,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.20600108306797218,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.3888,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.21045916149828317,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.3815,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.2132900635527811,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.3812,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.20312135562805128,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.3849,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.21040885458549818,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.3651,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.20552693713787873,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.3736,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.20560001161413705,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.3841,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.2073760182871921,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.3679,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.2049389103343872,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.3846,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.2074296197464054,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.3768,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.20687560760899926,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.3609,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.20998196486763163,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.381,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.20790065350987227,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.3877,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.19697689960414286,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.3909,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.20668808285937953,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.3806,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.22063525840379433,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.3578,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.20196031630367464,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.3649,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.20352395842897159,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.3752,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.21509300496056127,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.3648,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.21110059533933972,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.3723,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.2225337596147317,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.3783,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.19865998007475955,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.3774,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.2089757627658412,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.3843,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.19803354842582335,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.3776,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.2098786549898044,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.3719,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.19697404372691804,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.3829,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.2112063731495641,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.3749,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.20450498927034402,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.3597,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.20659976176781047,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.3624,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.20624490372258633,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.3643,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.20837213342172123,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.387,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.21514039112894934,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.3762,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.19696325086903121,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.3698,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.20438740421379134,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.3719,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.2211087876842422,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.3768,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.21458699725289027,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.3604,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.21219819897616404,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.3693,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.20651192352056952,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.389,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.19682208731688006,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.3707,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.200799906944616,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.3807,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.2053813591756652,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.3708,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.1974995455913308,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.391,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.20706765690919182,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.3587,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.20053423671090823,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.3636,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.20302844686192795,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.3634,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.20350219379629628,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.3716,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.199837805664711,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.3717,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.20209330510006454,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.3827,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.19534895177346262,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.3793,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.19894006641306852,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.3821,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.20473800202690354,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.3554,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.19621506088630739,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.3856,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.19894092313812217,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.3845,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.1975617242611743,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.3759,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.1936042489147819,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.3705,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.20251409562405856,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.3686,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.21167550435951835,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.3731,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.18481315774794255,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.3885,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.19790764186467688,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.368,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.20794772340753456,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.3585,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.2135659545437907,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.3617,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.20408411834670662,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.3727,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.19944854752634789,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.3686,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.2003527464183118,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.3806,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.1942787411638653,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.3771,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.18820950263554106,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.3771,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.20307660013088805,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.3709,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.20717217338530572,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.3718,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.19646244979542343,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.3722,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.19126045485933155,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.4003,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.1931701433335075,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.3752,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 4.958052233753441,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.4217,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.19355924789010262,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.386,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.20797848676315273,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.3779,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.196496841351256,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.3722,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.19645882526473274,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.3726,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.19466546900467677,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.3705,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.1988501178759439,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.3689,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.20174816662503742,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.3646,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.2019968238357404,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.3773,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.2192265955500457,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.3751,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.18928758156426959,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.3776,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.2213766257067429,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.3911,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.19551719177729585,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.3806,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.20054273615060722,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.3755,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.19165680312496533,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.3759,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.19931729377317153,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.363,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.19979034547919178,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.3759,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.20015535842161775,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.3785,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.1911314429773178,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.3929,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.1981724064710611,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.3679,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.1974462037584655,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.3771,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.18775654710139492,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.3912,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.19467585455306208,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.3658,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.19622123014783938,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.3713,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.19735185671216382,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.3697,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.18995555834225453,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.3905,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.18973959868103263,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.3854,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.19214500538216947,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.3857,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.19784036939983574,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.3774,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.20131885967413707,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.3709,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2006452209234428,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.3778,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.19014964003548934,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.3768,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.19649410758050886,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.3741,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.1916845561111861,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.3822,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.19390518952387623,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.3914,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.18638742672471326,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.3939,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.18932496106622246,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.3675,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.20531868011299137,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.3904,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.19288305561959923,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.3876,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.19497998634310748,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.3827,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.19547412834291222,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.3785,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.19890773723429003,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.3729,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.2066933918341747,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.371,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.19221639105598579,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.3836,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.19027860499672167,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.376,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.19269073849141996,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.378,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.1912651106118381,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.3901,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.18942736772363475,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.3852,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.21031900978821275,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.3735,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.19114531242415436,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.3817,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.19280815308611224,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.3755,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.20028471340359252,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.3613,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.1925748631081432,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.3781,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.19533975525290087,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.376,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.20003972928960395,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.3758,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.1877199995212104,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.3698,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.19473089625981801,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.3784,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.1899238078757953,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.3722,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.19114274270392387,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.3903,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.18950205178422527,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.387,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.18709163910901214,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.3688,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.1939845893760318,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.3712,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.19223688148479107,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.3858,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.19615663645742212,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.3774,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.20614456165243383,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.3714,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.19828108716609413,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.3754,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.19751019105692552,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.367,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.18938281439358753,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.369,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.1951715875494076,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.3735,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.19390834247579927,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.3866,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.19533193901819854,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.3724,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.19988371185920903,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.3532,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.19243021770826688,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.3696,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.20105494990308467,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.3789,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.18406817723574032,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.3933,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.19052232485095572,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.3628,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.201599853926091,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.3798,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.19654233789580774,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.3693,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.18717432272525023,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.3834,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.19031499646515548,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.3616,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.18741584961697377,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.3875,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.1958743754142958,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.3566,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.19377793034769605,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.3841,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.1879483920318766,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.3761,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.1939836798849069,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.3773,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.19074704616227992,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.3952,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.18293069001026602,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.3852,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.19035828400827876,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.3776,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.1868857691064927,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.3712,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.1950715673634053,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.3727,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.18408051639081763,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.3898,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.18360096175536852,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.3874,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.1818420590398244,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.3778,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.1897304038217063,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.3858,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.19064756418857604,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.3735,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.18497334635146298,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.3855,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.18810944094482646,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.3824,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.19231605755963416,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.3761,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.1914493318496071,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.3826,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.19058578453976824,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.3616,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.1981521742247874,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.3667,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.2006670186499817,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.3796,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.19285763471357673,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.3665,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.19309608365012862,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.3715,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.18476394649354536,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.3705,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.19177650723767883,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.378,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.1875787091040574,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.3709,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.19069547869124892,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.3773,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.19144432128732752,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.376,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.19137354561832268,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.373,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.18979471472206996,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.3795,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.18799202155898165,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.374,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.18984691122857228,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.3845,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.19325619561294563,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.3694,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.18577168465679772,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.3812,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.1912999886276741,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.3808,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.1921398715322506,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.3761,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.1899986225767572,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.3754,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.20213226914416862,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.3772,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.18951671968579806,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.3767,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.19155384322096836,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.384,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.19489286426107724,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.3812,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.19449051485410473,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.3683,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.18736281467594412,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.3875,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.19102344647874553,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.3728,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.1886686131416818,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.3714,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.18474622903447435,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.3865,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.1922224162990348,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.3648,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.1936293567669899,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.3788,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.19161666615446105,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.3705,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.19481749848759483,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.3637,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.1901689915033234,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.3712,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.1867197499206148,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.3872,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.19237073957694936,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.3815,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.19815707431168472,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.3728,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.20411437437600582,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.3778,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.19105403115631492,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.3751,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.18658483953812924,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.379,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.22611122276230808,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.3656,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.18826255684912405,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.3866,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.21474695293805077,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.3758,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.19416722429393574,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.3762,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.18711458164169548,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.3729,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.19350093354897577,
+      "learning_rate": 0.0,
+      "loss": 0.3746,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 5.285962813456515e+18,
+      "train_loss": 0.07530430794244836,
+      "train_runtime": 6016.657,
+      "train_samples_per_second": 26.26,
+      "train_steps_per_second": 0.205
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.285962813456515e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}