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": 8.026525434503938,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.9532,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 7.882164434687356,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.9998,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 7.843464344635058,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.9928,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 7.8417389652780845,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 2.0056,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 7.488306953846181,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.9705,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 7.140361861754341,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.9495,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 6.0091420217465386,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.9684,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.34166081415712,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.9404,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.103511596453666,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.8856,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.6932726650653223,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.8164,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.6437232982129117,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 1.8727,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 4.126682317393127,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 1.8312,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.4340869617039465,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 1.8481,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 4.325238139805762,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 1.8145,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.770335271022918,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.7742,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.741664810906421,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 1.713,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.243295392443877,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 1.8069,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 1.9842127697371874,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 1.7623,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.6200885269091334,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 1.6979,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.247795026563355,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.6948,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.2726729437280964,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 1.7282,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.4621936801870372,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 1.6604,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.505189210214472,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 1.6955,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.1499972959258873,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 1.6882,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.168842740393128,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.6931,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.2500550211816315,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 1.6853,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.225258566356079,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 1.6658,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.1972428424512904,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 1.5756,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.1897188115499329,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 1.6044,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 1.0749366331851946,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.6971,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 1.2489319761873838,
+      "learning_rate": 1e-05,
+      "loss": 1.6776,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 1.1312114160287388,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 1.631,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 1.1838411214865872,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 1.6277,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.8719912984003391,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 1.583,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 1.1150109480452228,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 1.6242,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.9593927294720196,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 1.6405,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.8537367462058044,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 1.5771,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.9554387817719772,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 1.5968,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.998122803757252,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 1.6196,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.8609903555935631,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 1.6345,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.8236523281360677,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 1.5832,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.9863773097777224,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 1.5895,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.848171222569732,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 1.5904,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.947013965687503,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 1.598,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.9289851969062997,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 1.5995,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.9951666246980823,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 1.5359,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 1.1006414604644887,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 1.5971,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 1.1862327895085363,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 1.6031,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 1.065824486654051,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 1.5321,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.954297255279065,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 1.6258,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 1.1755908404630064,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 1.5481,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 1.1865057150169516,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 1.5646,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.8888921407171231,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 1.5446,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 1.496082280310215,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 1.6169,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 1.1047364213596524,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 1.5929,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 1.3869923237066393,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 1.5586,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 1.357079045926828,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 1.5535,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 1.0490458408083558,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 1.6196,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 1.3581608354121941,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 1.5285,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.895983496835496,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 1.6061,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 1.2564498494934928,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 1.5769,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 1.0817459303859907,
+      "learning_rate": 2e-05,
+      "loss": 1.5505,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.9400457776598536,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 1.5594,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 1.6378854250047004,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 1.5314,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 1.0077927736946168,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 1.5467,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 1.9184602511687527,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 1.5615,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 1.3423057700708185,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 1.5415,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 2.274376348013867,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 1.5329,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.6549611522870724,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 1.5572,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 1.9631930656266023,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 1.5635,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 1.848341843156435,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 1.5614,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.9768481429852454,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 1.6079,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.9125903696732804,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 1.5382,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 2.118773812556432,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 1.5566,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.9269077353769797,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 1.5953,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.754569482919269,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 1.5123,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.6258638049713179,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 1.5569,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.8941499636803927,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 1.5243,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.564065390964657,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 1.5829,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 2.0329346083646502,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 1.5401,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.7147394504209161,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 1.5347,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.9645959988868995,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 1.561,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.5336045082396508,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 1.545,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 2.346291670884258,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 1.5561,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 2.1387883310246862,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 1.5551,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 2.0323933791898967,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 1.6041,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 1.9751315507319607,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 1.5895,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.718697384325289,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 1.5138,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.6374714194867745,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 1.5495,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 2.4175446869999733,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 1.5189,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 2.186300126655659,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 1.5561,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 2.2295044986455927,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 1.5384,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 2.210398699346369,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.5593,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.3374616554322618,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 1.516,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.3511707006261275,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 1.5674,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.4283631612814345,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 1.5392,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.9767650704059205,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 1.5538,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.475677434717399,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 1.5797,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.1034181022177074,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 1.488,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.6252869073148664,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 1.5404,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.1362580732107121,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 1.5947,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 2.143478212861006,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 1.5459,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.3839205382757072,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 1.5333,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 2.1007940206814593,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 1.5349,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.6466013248951366,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 1.5495,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 2.109037061419322,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 1.6037,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.625503314129484,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 1.551,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.4821404612081739,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 1.5434,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.3765477716589132,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 1.5499,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.301115474375949,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 1.5392,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.329779072906729,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 1.5639,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.835203341318974,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 1.5584,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.6773951688478266,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 1.5614,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.3521685622879787,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 1.5554,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 2.295896592914549,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 1.5245,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.693162415118145,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 1.5848,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 2.1189575711139876,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 1.532,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.8626412404049402,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 1.5624,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.8520676050734755,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 1.4908,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.762879488797456,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 1.5732,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.3294718965832582,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 1.5047,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 2.7557514613611174,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 1.6103,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.7190976847522046,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 1.5123,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 3.182230961229658,
+      "learning_rate": 4e-05,
+      "loss": 1.5695,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 2.2996450241236612,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 1.5615,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 3.2253344965075694,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 1.5256,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 3.282999107197885,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 1.5392,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.510830089016821,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 1.5544,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 1.85993758832767,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 1.5299,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.9652303281675816,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 1.5603,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.5452715950228741,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 1.5404,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 2.7290116955979555,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 1.5256,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.9044879078845642,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 1.568,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 3.045786210269286,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 1.5119,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 2.698373601238988,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 1.5815,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 2.193863718780223,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 1.5348,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 2.000773294021231,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 1.4846,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 2.072763967445434,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 1.5696,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.6667327998501393,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 1.5533,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 2.560954487662822,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 1.5214,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 2.164555471839649,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 1.5256,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 2.435194027539477,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 1.5637,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 2.003826412419521,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 1.5633,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 2.4905083260886975,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 1.537,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 2.305398830425363,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 1.5478,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.7438089304577027,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 1.5421,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 1.667434997342051,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 1.5666,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.9659591710851059,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 1.5356,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 1.5473905000326356,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 1.5117,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 2.244080279991039,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 1.5596,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.585309948089412,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 1.5333,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 2.2824728746115164,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 1.5179,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.8284220319431301,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 1.5328,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 2.3556334590348253,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 1.5367,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 1.8878335030160982,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 1.5565,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 2.243943212591818,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 1.5429,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 2.0235357962448313,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 1.5581,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 1.5653436913901708,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 1.4896,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 1.9562040955784044,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 1.5623,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 1.7880740580266685,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 1.5465,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 1.5278508831093587,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 1.5451,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 1.6986711892939381,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 1.5424,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 1.3112541849091732,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 1.492,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 1.8374153020056638,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 1.524,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 1.4359954463916338,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 1.5649,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 2.145841922623473,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 1.5344,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 1.7781167693080226,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 1.56,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 1.695240420897763,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 1.5137,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 1.7241814721597757,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 1.5572,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 1.319087911304726,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 1.5893,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 1.6758695678158944,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 1.5438,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 1.6243258189102097,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 1.5875,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 1.2090418869266344,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 1.512,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 2.235841729362679,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 1.5474,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 1.8145544563498666,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 1.5267,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 1.8265639163747172,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 1.5524,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 1.7778511357758318,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 1.5061,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 1.3605661780975529,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 1.569,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 1.5443923287369783,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 1.5328,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 1.0409783847936858,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 1.5046,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 1.6773971482763264,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 1.558,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 1.7126105776044434,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 1.5296,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 1.170781748476002,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 1.5342,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 3.0326841045138813,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 1.5429,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 2.444607281648493,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 1.5786,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 2.568002059002384,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 1.5573,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 2.281741850412629,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 1.5133,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 2.212717632319648,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 1.5312,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 2.235216149644005,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 1.5525,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.844718724225604,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 1.5645,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 1.432950052596294,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 1.5306,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 2.5889971067864455,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 1.5197,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 1.7517186559527682,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 1.5431,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 2.7800855390190806,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 1.5587,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 2.686436041427755,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 1.5098,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 1.9830044627130645,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 1.5709,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 2.059606160243582,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 1.5478,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 1.5645920086611544,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 1.5453,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 1.4668724420821615,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 1.5337,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 2.181173193680312,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 1.5376,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 1.7134334553211652,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 1.5281,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 2.353620888401153,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 1.5629,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 2.1110829161447824,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 1.5295,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 2.3202654431139793,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 1.5009,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 1.9573289369709683,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 1.5391,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 2.198929858693133,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 1.5412,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 2.086679416265024,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 1.4945,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 1.803793360353434,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 1.5593,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 2.025556600428043,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 1.5308,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 2.0206467425645704,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 1.5373,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 1.7258946216874347,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 1.5159,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 2.1591858485673834,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 1.4897,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 1.9396860731347654,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 1.5346,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 2.1160690138798426,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 1.5311,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 1.8472161450995226,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 1.5187,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 2.2438573626071774,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 1.5062,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 1.9757023531934303,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 1.5105,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 2.2510791548909754,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 1.527,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 1.893908819594474,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 1.564,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 2.210235644384317,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 1.5133,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 1.812551623407467,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 1.511,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 2.3124169925735183,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 1.5485,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 2.1096582997860085,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 1.5129,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 2.022488513775969,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 1.5178,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 1.7074625473337215,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 1.539,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 1.829715747627828,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 1.5285,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 1.7878640234023824,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 1.5282,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 2.148751261096328,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 1.5065,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 1.6303795959243155,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 1.5411,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 2.520588822026832,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 1.5834,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 2.479680253117511,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 1.5183,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 1.7029940599344275,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 1.5878,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 1.8325299249288673,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 1.5684,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 1.6308916768200545,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 1.5248,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 1.2034658830215734,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 1.4804,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 2.3695138388687305,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 1.4933,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 1.7381705929328566,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 1.5321,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 2.3260296122645863,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 1.5172,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 1.9763094756040906,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 1.5165,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 1.9224002368295328,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 1.5181,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 1.695701858060466,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 1.5266,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 2.0512935733178637,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 1.5374,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 1.761449025379216,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 1.5545,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 2.0796232158101757,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 1.5427,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 2.042488061710048,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 1.4754,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 1.8633629521165969,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 1.5458,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.828627245282665,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 1.5405,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 1.9516526002991035,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 1.4919,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 1.6024118847043638,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 1.4922,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 1.778439932031488,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 1.4135,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 1.499079527661675,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 1.4411,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 2.257732692135664,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 1.4745,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 1.6567959554864715,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 1.4118,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 2.2281560519069554,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 1.473,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 2.0417861851290633,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 1.4733,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 1.7377778177661911,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 1.4573,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 1.7330300029483587,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 1.462,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 1.8425824263146662,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 1.4086,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 1.5114170960928361,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 1.479,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 2.043914168900245,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 1.458,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 1.8698583963811974,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 1.4475,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 1.906413577736406,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 1.4846,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 1.843951716044497,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 1.4722,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 1.826171623108835,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 1.4898,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 1.9691423022053642,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 1.4247,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 1.8856143872464275,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 1.4752,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 1.8073789678561616,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 1.4674,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 1.814468556440197,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 1.4702,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 1.5905663273199844,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 1.4798,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 1.9685651922159049,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 1.4414,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 1.457444023001711,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 1.4842,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 2.0904417208992565,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 1.4202,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 1.8299527539424458,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 1.4128,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 1.8957120001835466,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 1.3957,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 1.8027509504620824,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 1.4103,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 1.8235315063326347,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 1.4429,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 1.6313676559624086,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 1.492,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 1.945809243686613,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 1.437,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 1.8518486298194334,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 1.4271,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 1.7149935212272156,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 1.4614,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 1.788708987192345,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 1.4806,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 1.4936417276230842,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 1.4413,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 1.5465160300430871,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 1.4566,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 1.619921939940084,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 1.4838,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 1.1940675821357196,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 1.4707,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 2.084424924899279,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 1.4645,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 1.5494661278987931,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 1.4621,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 2.2950747026558895,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 1.4755,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 2.1865970540772275,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 1.4495,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 1.575273045070695,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 1.5086,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 1.594284861149251,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 1.4782,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 1.3996336380985948,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 1.4391,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.9034180595812057,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 1.4851,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 1.876290592953088,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 1.4536,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 1.204327552151393,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 1.4377,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 1.876510042789472,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 1.4668,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 1.5361574903983923,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 1.4532,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 1.6156788730104323,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 1.4287,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 1.3052414565592194,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 1.4767,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 1.6790359075813623,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 1.4847,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 1.3299387407261543,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 1.4549,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 1.4173990011980024,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 1.4417,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 1.3110041627087787,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 1.4776,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 1.4112123654302542,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 1.4718,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 1.3638621860404503,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 1.4526,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 1.3038258684013035,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 1.427,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 1.7410553337155938,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 1.4519,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 1.6900836186068007,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 1.4884,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 1.5314208929296924,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 1.4525,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 1.5148120100945952,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 1.4371,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 1.5098607359142575,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 1.4076,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 1.327864037710334,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 1.4686,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 1.592160299585359,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 1.4524,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 1.3004528427128519,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 1.4475,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 1.7103372368658185,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 1.472,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 1.2779572416706253,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 1.4829,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 1.5594886461289368,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 1.4815,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 1.3955816167919821,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 1.4702,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 1.596953333177511,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 1.4979,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 1.6498853222331544,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 1.4924,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 1.3173730929391683,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 1.4684,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 1.6151180938166807,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 1.4541,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 1.1300167717375267,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 1.4246,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 1.5349940245477935,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 1.4143,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 1.3833361153360522,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 1.4833,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 1.4802131326480736,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 1.4431,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 1.4238126988807094,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 1.4719,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 1.0382271901338347,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 1.3996,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 1.5775750188600648,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 1.4636,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 1.1943343700309819,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 1.4403,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 1.693526942070933,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 1.4728,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 1.5429147223834976,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 1.4693,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 1.49303734051667,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 1.4578,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 1.4734049942710026,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 1.4554,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 1.2760211948042013,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 1.425,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 1.443908742863055,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 1.449,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 1.3744414697641059,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 1.5007,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 1.3653420451907459,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 1.4777,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 1.1348061097271116,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 1.4518,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 1.5960709145904004,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 1.4347,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 1.1538540683196536,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 1.4715,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 1.5240884576430527,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 1.4737,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 1.2645416178710849,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 1.4977,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 1.3329288677625157,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 1.5057,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 1.104635087292853,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 1.4321,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 1.1184393443449332,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 1.4881,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 1.384186351158884,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 1.4384,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 1.0415888478237458,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 1.4595,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 1.3606808132732657,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 1.4493,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 1.289518061342375,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 1.4343,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 1.0860686692795323,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 1.4421,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 1.4920526834801162,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 1.4341,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 1.1449660005899456,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 1.4153,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 1.4304648576255001,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 1.4907,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 1.1421909269376853,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 1.4609,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 1.5931159149175058,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 1.4665,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 1.3505986026763352,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 1.4888,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 1.4487673394669582,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 1.5157,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 1.3121025571816487,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 1.4279,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 1.3687617192747497,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 1.5134,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 1.3178626899979822,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 1.4966,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 1.6757628265944693,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 1.5323,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 1.2721601685834123,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 1.4399,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 1.453479664944701,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 1.4606,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 1.2219586182494662,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 1.4741,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 1.2787707028900102,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 1.4576,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 1.274763453766783,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 1.5116,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.7944820130802454,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 1.4666,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.9164010409521497,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 1.5749,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 1.3006557111300547,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 1.4335,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 1.0682749890379142,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 1.4577,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 1.5145396256136825,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 1.4412,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 1.307643472905405,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 1.4585,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 1.7069708019431333,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 1.3979,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 1.2555715247322958,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 1.4754,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 1.4945449779131745,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 1.5124,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 1.3382373487903652,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 1.495,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 1.1385571689442475,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 1.4791,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 1.6424656556603883,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 1.4533,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 1.1513219579602938,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 1.4558,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 1.6240332417595487,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 1.4866,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 1.3214820059195738,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 1.5546,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 1.58408006105677,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 1.4668,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 1.3090896143432416,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 1.4893,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 1.3839759477559703,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 1.4748,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 1.4177761124186856,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 1.4669,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 1.0956139424043776,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 1.464,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 1.775122909433424,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 1.4684,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 1.6112279465461563,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 1.4986,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 1.6466391522830601,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 1.5006,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 1.3046799982703883,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 1.4703,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 1.8800321034607097,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 1.4563,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 1.2908360670338301,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 1.4427,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 1.9414668284807288,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 1.4814,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 1.8325369917671959,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 1.4273,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 1.4151141853358074,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 1.4862,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 1.655465102620202,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 1.4714,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 1.224694814217516,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 1.46,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.0828997245019631,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 1.4483,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 1.5889131887130408,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 1.4845,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 1.0360362948955826,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 1.4196,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 1.7745338982223129,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 1.4952,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 1.2958548255639457,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 1.5003,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 1.887028898589429,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 1.4611,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 1.5456361858321828,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 1.4859,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 1.78889515939473,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 1.4924,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 1.3763587319210693,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 1.4173,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 1.784114266463693,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 1.4818,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 1.2554082985273476,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 1.4326,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 1.6225995046399928,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 1.4301,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 1.384616688110807,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 1.5231,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 1.740130136889657,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 1.4327,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 1.2977801477508235,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 1.4472,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 1.5951011541875229,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 1.4319,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 1.4275586581468618,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 1.4533,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 1.6845346876562166,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 1.4783,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 1.3611228087939786,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 1.4815,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 1.626726432391903,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 1.4843,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 1.5291195043916948,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 1.4611,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 1.55845164614389,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 1.4251,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 1.4969100901305772,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 1.4397,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 1.1699796629308048,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 1.4529,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 1.4038147044185991,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 1.5146,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 1.2533957835446832,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 1.4755,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 1.5451404026736748,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 1.4489,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 1.3076136755403516,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 1.4864,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 1.5173757692500407,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 1.4555,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 1.4996955575122535,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 1.4682,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 1.5829260615422402,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 1.4736,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 1.2926456292896293,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 1.4515,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 1.7668991704416048,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 1.4367,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 1.4784214870320136,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 1.4678,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 1.8391857980586834,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 1.4608,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 1.5054132876638042,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 1.479,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 1.5642460260189335,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 1.4478,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 1.423216335538991,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 1.4822,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 1.5080923531956116,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 1.4887,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 1.4633430555431786,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 1.4887,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 1.6673835943697075,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 1.4763,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 1.4766417226779849,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 1.4387,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 1.4725130939828635,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 1.467,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 1.2552340420524868,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 1.3969,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 1.6285984102799758,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 1.4312,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 1.3601919224917105,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 1.4597,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 1.806815486367513,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 1.4609,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 1.5594243475356249,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 1.4696,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 1.4565663959813366,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 1.4361,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 1.3584553278399139,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 1.4721,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 1.4352682468277345,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 1.4803,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.9626769565434922,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 1.4614,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 2.050603673243594,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 1.4526,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 1.6642078265610445,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 1.5099,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 1.7260249668990029,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 1.4838,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 1.8087777837940449,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 1.4394,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 1.258681539553661,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 1.4322,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 1.4327919175910646,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 1.4756,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 1.3283800751028683,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 1.446,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 1.0851054346668365,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 1.4231,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 1.584818189242404,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 1.4735,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 1.1395192176375906,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 1.4293,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 1.6004724872786222,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 1.471,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 1.1279656176933905,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 1.4879,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 1.9397484504225193,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 1.4944,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 1.5411920414151226,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 1.4566,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 1.705879030002293,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 1.4562,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 1.7547504673655065,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 1.4841,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 1.453261459858435,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 1.4859,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 1.5001714116594078,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 1.4474,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 1.1235406948372713,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 1.484,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 1.1185994275581814,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 1.4553,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 1.423398096437211,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 1.4324,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.9565337349377421,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 1.4524,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 1.383411407220692,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 1.4268,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.9105890276982307,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 1.4596,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 1.3856429339053513,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 1.468,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.9893237872260133,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 1.4648,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.949300117840019,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 1.4798,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 1.20455677230502,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 1.4447,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.9846276935164346,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 1.438,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 1.5389068390973635,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 1.4855,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 1.0930603215233847,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 1.4493,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 1.5105327562204485,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 1.458,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 1.0922958176960458,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 1.4155,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 1.536156720566966,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 1.4442,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 1.1324967228341594,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 1.4935,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 1.5796656320860485,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 1.4954,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 1.474579580875223,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 1.4987,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 1.781220660073662,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 1.4465,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 1.6640453458396438,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 1.4414,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 1.319303029817513,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 1.4808,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 1.3074153314507415,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 1.4474,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 1.5725205497220727,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 1.4944,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 1.25547677222195,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 1.506,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6703048659319772,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 1.5072,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 1.5200338788323136,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 1.337,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 1.2171013716072796,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 1.3275,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 1.0623037125775403,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 1.3182,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 1.576256872205973,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 1.3549,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 1.1466555888148038,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 1.3493,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 1.5104794322664017,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 1.3268,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 1.4716926862562918,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 1.3545,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 1.6188708084028105,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 1.3309,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 1.3830240923416734,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 1.3347,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 1.4298485002672854,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 1.3265,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 1.2220778802871601,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 1.2927,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 1.3290847411940625,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 1.3294,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 1.1092189578819713,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 1.3135,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 1.2564736894944075,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 1.3189,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 1.209966319342894,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 1.3143,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 1.0660614273539868,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 1.3219,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 1.1095259126179362,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 1.3707,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 1.2357291908164043,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 1.34,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 1.1068803280542427,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 1.3112,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 1.0324139833573918,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 1.3029,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 1.171407958375176,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 1.3928,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.9468482222332095,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 1.3153,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 1.1760094151290164,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 1.3467,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.9482387305539002,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 1.3454,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 1.282073873784778,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 1.3712,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.9945304038449438,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 1.2721,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 1.1317132577378664,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 1.3215,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.9917489985745699,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 1.3288,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.9815714997486087,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 1.3364,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.9474480102286158,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 1.3461,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.9037297680380816,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 1.3243,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.9277686185377051,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 1.374,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 1.0406156293391742,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 1.3657,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 1.116343179345882,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 1.3347,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.8554310618464325,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 1.3573,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 1.1265907345330657,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 1.3813,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.8067529546166173,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 1.3705,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 1.1289006355431617,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 1.342,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.9691005972222443,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 1.3487,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.9624213501350912,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 1.3122,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 1.0858681977967308,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 1.2965,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.816583451646021,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 1.3574,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 1.0741501388358752,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 1.3037,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.8952962800698278,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 1.2786,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.90394820521206,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 1.3148,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 1.0410689085982527,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 1.3662,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.8613044344020855,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 1.3335,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.864921929993728,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 1.3646,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.969663515115518,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 1.3349,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.714900949737254,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 1.3444,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 1.0115862345459,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 1.3675,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.8158997645609584,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 1.3925,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.8393457812418104,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 1.3545,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.8029273834602888,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 1.3241,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.8872262268091543,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 1.3586,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.8059694785117653,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 1.3607,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.8394075053748343,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 1.3576,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.8854484865516419,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 1.3649,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.8713455433966444,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 1.3628,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.8326257187714247,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 1.3367,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.7605151394813732,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 1.3768,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.752264950494583,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 1.3064,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.8570835555383621,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 1.3461,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.8647738261767182,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 1.3317,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.8805214731302021,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 1.3211,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.8449987586447775,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 1.3369,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.8532493701081851,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 1.3494,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.9235700023154165,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 1.4024,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.93808682744413,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 1.3793,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.6763483158027912,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 1.3886,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.9324249277144254,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 1.4046,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.744151187525395,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 1.3401,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.7359921092860575,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 1.3591,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.8425580999564389,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 1.357,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.8554636499262447,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 1.36,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 1.1757118688151909,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 1.3293,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.8230193747857821,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 1.3474,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 1.0382175745869637,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 1.3485,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.8504234163144797,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 1.3413,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.7911038926574405,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 1.3467,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.8922502266869108,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 1.3663,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.8349851977186469,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 1.3848,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.8514592639926716,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 1.3945,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.9704758222337911,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 1.3522,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.8721090409958855,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 1.3845,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.9476217587767076,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 1.3505,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.9472805695344462,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 1.3034,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.7905387238053974,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 1.3657,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.6719148369367203,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 1.3449,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.9433340120635818,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 1.3193,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.7751337809059741,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 1.3591,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.7854434365539742,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 1.3764,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.6807731301107439,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 1.3752,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.8594945918012924,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 1.3435,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.7633396231283969,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 1.3698,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.6843494506331644,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 1.3772,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.8637428986116925,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 1.3675,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.6545080936539937,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 1.3617,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.8520710764535974,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 1.3573,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.8249948069094276,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 1.3962,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.6660734963088515,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 1.4046,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.7108108259403759,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 1.3575,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.6980518871440919,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 1.3559,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.8047283183611125,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 1.3794,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.8004411896745335,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 1.3514,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.7994355399253518,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 1.3535,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.837764533146181,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 1.4169,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.6453602280435115,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 1.3717,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.8268488363852792,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 1.3381,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.7991954467850577,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 1.3158,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.809504611777723,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 1.4,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 1.1330531626100468,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 1.3458,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.7633956888911831,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 1.3454,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 1.1860673502242804,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 1.3603,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.8199238578751732,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 1.3546,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 1.0536328507153672,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 1.3686,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.9691822344123713,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 1.3775,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.9862320402889717,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 1.3643,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.8932094633368629,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 1.3865,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.8727337613313854,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 1.3429,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.8185015694372462,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 1.3767,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.7507848356103802,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 1.363,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.8039146764505831,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 1.3023,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.8750271472155938,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 1.3437,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.7372135199295209,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 1.3373,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 1.0802341214086362,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 1.3707,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.7839169606521996,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 1.3561,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.7829876846527618,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 1.3332,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.7025680486323816,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 1.3363,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.7454160421478038,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 1.3683,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.747879603330156,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 1.3911,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.7899671242887376,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 1.3805,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.6675445044739393,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 1.3437,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.7367896324517105,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 1.3612,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.7390993402068606,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 1.3334,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.7332488032176564,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 1.3841,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.799734103529994,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 1.35,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.6888499237062715,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 1.352,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.6451463906726327,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 1.3677,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.7296476372395322,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 1.3708,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.7194814512885057,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 1.3593,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.6897436999631273,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 1.3767,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.706954993933064,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 1.406,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.7507839829543848,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 1.3816,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.7800087645327112,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 1.3709,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.750378634025515,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 1.3679,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.7794979398607556,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 1.3886,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.7335673695537557,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 1.3503,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.7642841629491732,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 1.3721,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.7228418128634028,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 1.3637,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.7370727916648733,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 1.3748,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.718888427068339,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 1.3285,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.7494495923823151,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 1.3258,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.6654270068233933,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 1.3796,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.8312163353990792,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 1.353,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.6963952751109488,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 1.3349,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.7513027298951658,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 1.3214,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.8745410670205829,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 1.3559,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.8598647108722274,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 1.371,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.7939051693494025,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 1.3647,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.7755667386953683,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 1.4011,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.8455617938079358,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 1.3431,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.7577729579286112,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 1.3354,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.8861926316007983,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 1.3866,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.8417657717206692,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 1.3715,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.7894647600082629,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 1.3167,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.8498469589276549,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 1.3529,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.7150660423726173,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 1.3185,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.9121071096117757,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 1.3558,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.818663341690981,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 1.3629,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.7864131375811231,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 1.3446,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.8037213471718253,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 1.3479,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.754175616960917,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 1.3033,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.9001606106107812,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 1.3394,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.6952765443163472,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 1.3519,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.8037412919163156,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 1.3383,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.6392085129562798,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 1.3394,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.8274963837996411,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 1.3098,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.7092541565704569,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 1.3755,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.7267291579882804,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 1.3613,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.7488114955878362,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 1.3627,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.7281773952370632,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 1.3729,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.7893754041249252,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 1.3955,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.8484711604750808,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 1.3707,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.7420392382623489,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 1.3622,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.642243107301126,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 1.3304,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.7758112736840123,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 1.3659,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.6509013515913762,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 1.368,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.7163547438854224,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 1.3978,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.6647117121172667,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 1.3287,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.712594962884902,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 1.4,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.7272026873710101,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 1.3799,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.6715363086183569,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 1.3558,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.8074388371853205,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 1.3347,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.7254382486834361,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 1.3773,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.9312700929931609,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 1.3563,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.6854458177176904,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 1.3236,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.9074864596180356,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 1.3503,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.6392330503156646,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 1.3146,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.75034230682138,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 1.3365,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.7405953050270885,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 1.3333,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.7238811530761463,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 1.3356,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.7076419341246591,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 1.372,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.884822829622381,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 1.315,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.6649308575588457,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 1.3847,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.8391575927656353,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 1.3191,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.7150525274580681,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 1.3675,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.7020170817183811,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 1.387,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.8056946793417017,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 1.3524,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.8668649627516799,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 1.3228,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.7725264687548425,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 1.3923,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.8054878029060134,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 1.3336,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.8518906480933459,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 1.3763,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.7901204475275555,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 1.3792,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.8131472707983223,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 1.3914,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.7663046092234481,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 1.3595,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.8276865419292814,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 1.4172,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.7340216726339628,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 1.3383,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.7686864996281839,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 1.3408,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.7759050991284056,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 1.3562,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.9565069038137378,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 1.3447,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.6883698691000869,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 1.3661,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.8887672449464507,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 1.3919,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.6604513343365663,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 1.3509,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.9085301543646872,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 1.2791,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.6768473156641006,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 1.3394,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.852427360253586,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 1.3626,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.62590253763281,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 1.3744,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.779495048758307,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 1.3649,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.7501240911385546,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 1.337,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.711433315656692,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 1.3794,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.7440222489775266,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 1.3149,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.8345574706040749,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 1.3757,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.6955090287193482,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 1.3645,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.7311939115085901,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 1.3797,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.7902670588817831,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 1.3725,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.7020825241576522,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 1.3316,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.6946946177746777,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 1.37,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.7021502166572839,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 1.3509,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.7551067834045687,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 1.3797,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.634739180490126,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 1.3575,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.6868062251913953,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 1.3335,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.7257059834141432,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 1.3726,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.7554815867454113,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 1.3898,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.6882873674303993,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 1.2973,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.7647292915986252,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 1.3729,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6647047844954986,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 1.3535,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 1.0310913053682185,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 1.1835,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.7938789512695357,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 1.1875,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 1.2290619653037034,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 1.1656,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.9488287158324726,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 1.1686,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 1.2821711170154533,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 1.1783,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.9397299250045513,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 1.1819,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 1.0407539809553394,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 1.1894,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 1.1435168193079024,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 1.1902,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.9288378239355154,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 1.1589,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 1.0914883159167925,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 1.1522,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 1.1849329010103224,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 1.1668,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.8643632805576965,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 1.1412,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.8366966457436835,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 1.1875,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 1.0601868992295942,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 1.2095,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.797130971322437,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 1.1713,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 1.17290829643875,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 1.1954,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.8922269505696444,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 1.1674,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.7640425079512063,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 1.1837,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.9393969272223371,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 1.1564,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.9572549020595033,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 1.1631,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.6850654728694074,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 1.1909,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.9680240080335087,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 1.1806,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.8862192378542401,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 1.2034,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.6658115187657312,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 1.169,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.8294914186407537,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 1.1868,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.7354859687122971,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 1.1808,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.6823943414308418,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 1.1642,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.8021079299040427,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 1.172,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.7380707577066659,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 1.1665,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.6653327426544046,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 1.1865,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.8420594580428824,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 1.2244,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.7736198827699564,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 1.1916,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.644993663368803,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 1.171,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.7789486808328437,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 1.1526,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.6631503814516665,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 1.1787,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.6975394342366994,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 1.1543,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.6161693021077388,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 1.1567,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.6288417948553625,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 1.1904,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.7128581485320354,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 1.1716,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.680473870820667,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 1.1937,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.6625254767807587,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 1.177,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.6539076042847667,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 1.1939,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.6594045092340314,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 1.1547,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.6476808580994446,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 1.1911,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.6988016959501989,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 1.1407,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.6413968085583593,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 1.1664,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.552684199105306,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 1.1721,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.583271990608553,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 1.1751,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.6511425564216744,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 1.2059,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.5988889400548736,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 1.184,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.6834383057047795,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 1.1852,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.6321874629562481,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 1.1999,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.5966804140796729,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 1.1618,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.6184167145826728,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 1.1415,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.6094869126972463,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 1.2087,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.5802111227287393,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 1.174,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.6581021616924407,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 1.1824,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.6341135561711874,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 1.1436,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.6964504197360859,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 1.1547,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.7270577974951784,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 1.1762,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.6525703608993937,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 1.234,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.7266570487542904,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 1.1694,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.6349547628914392,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 1.2256,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.5540138786296759,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 1.1724,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.5952548936430602,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 1.1463,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.6360767335141344,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 1.1891,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.5954637683460638,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 1.1731,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.5991336332920943,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 1.1879,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.7161622465620239,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 1.2055,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.6240269580522437,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 1.2364,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.7634047081081468,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 1.1521,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.5654858832704541,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 1.1607,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.7381213768533125,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 1.1456,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.6256540616515218,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 1.1982,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.6205160637008112,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 1.149,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.73741958185056,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 1.2273,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.6957809363077071,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 1.1546,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.652960376050605,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 1.1636,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.7956005072165365,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 1.1919,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.7408653543657551,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 1.1636,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.5987402314945357,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 1.1876,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.7753399368930705,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 1.1686,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.6576146237624735,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 1.1764,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.6651148486201507,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 1.2218,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.5544407673615969,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 1.1802,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.7878285019532115,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 1.1467,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.6349654463141001,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 1.2,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.5990428112401303,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 1.1982,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.6261309921660293,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 1.1809,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.7107542841316558,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 1.1645,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.7187738799934636,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 1.1584,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.6104273679481167,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 1.1575,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.6722057347228764,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 1.1643,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.7286104047190327,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 1.1523,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.6727161493971696,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 1.2123,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.6931504447497595,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 1.1962,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.6363479867565578,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 1.1815,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.6446022631003733,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 1.1704,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.6675781343954964,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 1.1394,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.7336376994190794,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 1.1865,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.6916195995964635,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 1.1948,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.5984307493500672,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 1.1768,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.5861835520824048,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 1.1658,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.6946547851010797,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 1.1608,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.59934946580021,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 1.1577,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.6512698608765569,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 1.1899,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.6522889919301373,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 1.1772,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.6714318478427695,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 1.1793,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.6768180621487315,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 1.1644,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.63790861009633,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 1.1616,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.6266565181802534,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 1.1771,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.6442273700548274,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 1.1515,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.6429367679360597,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 1.1361,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.6330436546271315,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 1.19,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.5984681257521319,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 1.1776,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.6772186704715986,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 1.1829,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.6985899165624111,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 1.1725,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.5839296865765894,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 1.2077,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.6138671610542094,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 1.1928,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.6345642935458115,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 1.2137,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.5525020495809264,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 1.1568,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.6373460952817885,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 1.1873,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.5717162855036374,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 1.1719,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.6080492830000146,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 1.1712,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.6216497050910281,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 1.1601,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.5473709527091764,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 1.1586,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.557196459882288,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 1.1776,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.5843292231610796,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 1.1741,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.6323745911864286,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 1.1909,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.5523809387863426,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 1.1647,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.5385573218481451,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 1.1567,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.6033355596576097,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 1.161,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.6039381860790051,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 1.2017,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.5402578675983593,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 1.1825,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.6000208876644434,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 1.21,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.6201282744454086,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 1.1508,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.5737866976704377,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 1.1558,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.6301008910068358,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 1.186,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.5503430287890065,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 1.2007,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.5686774051918606,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 1.1641,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.5811342547842181,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 1.1755,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.5496253569023846,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 1.2043,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.6201466136194412,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 1.1782,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.5689983404515472,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 1.1845,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.6248615801054077,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 1.1796,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.5384826381192532,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 1.1856,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.5742777250825883,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 1.1657,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.5381610780712001,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 1.1668,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.5556970219607167,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 1.1979,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.5397837744429611,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 1.179,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.5647955229006898,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 1.1967,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.5405252812681757,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 1.1617,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.599271938815631,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 1.176,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.5403418098253385,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 1.1442,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.5238903963538231,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 1.181,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.5944437783195748,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 1.1712,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.5693263690107824,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 1.1868,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.5227428530527674,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 1.1632,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.554572696888225,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 1.1684,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.5880677150717764,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 1.1952,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.572521419167408,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 1.2006,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.5981246389516333,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 1.1995,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.6171894987738646,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 1.187,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.5145835356992159,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 1.1411,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.64682682386901,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 1.1992,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.5810363775338714,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 1.1525,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.5021572427008222,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 1.1764,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.5863306726005857,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 1.2048,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.5561338644241655,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 1.201,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.525682740186182,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 1.1875,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.5436489276719904,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 1.1933,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.5289253391894925,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 1.1691,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.5310139262988243,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 1.1849,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.5409869828603879,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 1.1813,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.5363736769038229,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 1.1712,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.5621697868796991,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 1.1996,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.5245876770751328,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 1.1842,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.5278219001953175,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 1.1915,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.5446082773489994,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 1.171,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.5841017592346412,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 1.2087,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.49753853335138354,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 1.1697,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.5625823351531547,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 1.1527,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.516646640899215,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 1.2055,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.5261120580414831,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 1.183,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.5694567447475214,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 1.1829,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.5525057920354985,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 1.1963,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.5271491548854899,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 1.146,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.5224542070890492,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 1.1872,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.5217885077436958,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 1.1896,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.5288853966543048,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 1.1771,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.553623900666312,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 1.1363,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.5253715412272851,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 1.1601,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.5155946100203869,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 1.1468,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.5734588111255852,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 1.1926,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.5427124904609653,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 1.2026,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.533679752636476,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 1.2006,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.5314313756738094,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 1.1399,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.5866293857682436,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 1.1841,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.5355632993869306,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 1.1731,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.5231275398921714,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 1.1698,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.5258535560263184,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 1.1732,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.6284786957444448,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 1.2036,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.5371091293990691,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 1.1853,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.5597964368892996,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 1.1935,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.5522431501645808,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 1.2023,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.5941410232375488,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 1.1439,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.6312147405148821,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 1.1523,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.5970900931626714,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 1.1697,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.5404054747794634,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 1.1969,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.5887671043045948,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 1.18,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.5250229555753759,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 1.1805,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.5606990943439549,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 1.1682,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.5464186563614354,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 1.1983,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.5101356496712628,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 1.1672,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.5444504765271971,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 1.1378,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.5075890582197089,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 1.1934,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.5560297449580451,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 1.1792,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.5246196920184715,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 1.1643,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.5191674217635841,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 1.1655,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.5178236560527186,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 1.2287,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.555389305367719,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 1.1798,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.553484031865255,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 1.1783,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.5392997539537439,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 1.1752,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.5624614058718049,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 1.1811,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.5170985590992573,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 1.1479,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.55311545717254,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 1.1906,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.5007445162186746,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 1.1696,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.5185422430249235,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 1.1475,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.5562808764025864,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 1.1775,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.5023012217420441,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 1.1823,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.5287870149873268,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 1.2479,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.5258370814506844,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 1.1704,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.5361035921540985,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 1.1191,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.5053190141376568,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 1.1791,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.5288494229935701,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 1.2011,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.5429636692864086,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 1.1793,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.5147196183025035,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 1.1955,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.586510155873436,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 1.1752,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.5286917460693905,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 1.2,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.48380404987145326,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 1.1783,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.5087497948275741,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 1.1806,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.4924406512763891,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 1.1028,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.5001326207237416,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 1.1622,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.5132743879934816,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 1.1404,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.5296010925521708,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 1.1724,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.5379888930195299,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 1.1822,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.5128703617145899,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 1.1473,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.973073046807191,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 1.0313,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.8412645135630926,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 1.0347,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.634613171724193,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 1.0149,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.9478977687228641,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.9732,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 1.1609921524304192,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 1.0341,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.9680958710262052,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.9817,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.8223084582460958,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 1.0141,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.9503170166829432,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 1.0068,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.8850189013261334,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 1.0578,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.7450442693580758,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 1.0313,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.6804852767975672,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 1.0119,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.7179835711658554,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 1.0094,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.7917739141141629,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.9992,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.76068893455233,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.9988,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.6741663311283637,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 1.0219,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.6389841382279978,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 1.022,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.6303219179177103,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 1.006,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.6884273855470108,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 1.0128,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.7357530952600735,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 1.0211,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.6211318146874908,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 1.0227,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.6170895295761802,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 1.0336,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.5687569327112564,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 1.0237,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.6571008874693811,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 1.0074,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.6656159357586171,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 1.0047,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.619938629508229,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 1.0265,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.5140850041386252,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 1.0284,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.5991132754283238,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 1.0172,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.6537056220523807,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 1.0568,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.5998034279359259,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 1.0246,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.5861918791876246,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 1.0232,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.5462216432873963,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 1.0186,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.586157391605592,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 1.0252,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.5473614692907793,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 1.0074,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.5583174051545087,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 1.0076,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.5560565969178766,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.9862,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.557525141542047,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.9973,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.5464639970953783,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 1.0268,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.542316577745015,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 1.0028,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.5316396478376516,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 1.0181,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.5259380798262479,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 1.0055,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.509957292441247,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 1.0166,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.535582122105427,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 1.0372,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.5306439113645566,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.9968,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.5430680345781942,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 1.0303,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.5441969031379472,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 1.0365,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.5078471680877957,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 1.0314,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.534925844582882,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 1.0116,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.5158814420111696,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 1.0127,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.5167109159121469,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.9905,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.5200832465968186,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 1.0233,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.49972521297935335,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 1.0084,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.5218513198412142,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.9795,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.5432110876244947,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 1.0255,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.5185156087946599,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 1.0411,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.5129402963573609,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 1.0376,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.5188443129828767,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 1.0028,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.512862017321004,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 1.0115,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.5046749384616654,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 1.0189,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.5272736699571927,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 1.0093,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.5068495013970551,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 1.0078,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.512690747150822,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.9833,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.49056612232833074,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 1.0178,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.5129707249044791,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.9988,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.5319179026462345,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 1.0295,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.5016256077262993,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 1.0072,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.543074631887413,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 1.0381,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.5540281987378094,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 1.031,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.5339860488642917,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 1.0175,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.5185203762540869,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 1.0462,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.509278928170136,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 1.0236,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.5271452700089436,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.9961,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.540551353087377,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 1.016,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.5273325667309819,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 1.036,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.5231899088278212,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 1.0292,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.5231874807989146,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 1.0198,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.5193784323196139,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 1.0073,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.5218394617616687,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 1.0084,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.523009634016985,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 1.0007,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.5364207025388761,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 1.0127,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.5199023294676158,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 1.0009,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.5066284454168705,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 1.0332,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.5053366618153782,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 1.018,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.5051294547884078,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 1.042,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.4879936872848984,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.9898,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.5036548254273029,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 1.0014,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.507731920739191,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 1.0393,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.515076614536176,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 1.0099,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.501306711022661,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.9999,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.5068795944709831,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 1.0121,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.5033967503482065,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 1.0029,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.49704362175995154,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 1.0147,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.4921998788711315,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 1.0018,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.5138885248714826,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 1.0241,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.5065089897218314,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.9829,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.5057912207095364,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 1.0216,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.5317049734567968,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 1.0316,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.5263026828826555,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 1.0075,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.5139936742736465,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.978,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.5035536757209479,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 1.0494,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.5066592674127866,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 1.0093,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.509767902641016,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 1.0204,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.5094678028492994,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.9924,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.506994224522678,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 1.0178,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.536258585934558,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 1.0244,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.5440462828251401,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 1.0131,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.5107399152818294,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 1.0241,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.5073610123672274,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.9985,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.5388363347962158,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 1.0416,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.5147862353036384,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 1.0244,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.5214220063214451,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 1.0124,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.5094222771878422,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.9938,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.4984495591438386,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 1.0247,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.5261181374463476,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 1.0346,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.5000015684086486,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.992,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.4936952305178453,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 1.0305,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.5079051259088926,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 1.0391,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.5041113555765385,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 1.0213,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.49937388723701276,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 1.001,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.4989955983420819,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 1.0395,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.5108736233612111,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 1.0169,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.5588938412726985,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 1.0377,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.49273635906808877,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 1.0064,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.5043341606028461,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 1.0225,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.4946953137473859,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.9871,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.4881890549402293,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.9907,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.5201492937517667,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 1.0518,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.5332819831416252,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 1.0245,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.5187063717069921,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 1.0064,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.5137701094732642,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 1.0034,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.5127286575310469,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 1.0232,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.5309936825619415,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 1.0489,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.5100536702098136,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 1.0206,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.48553024368010483,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 1.0229,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.4973794237558555,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 1.0233,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.4954731765740399,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 1.0197,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.49964728710761436,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 1.0182,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.49188784833631766,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 1.0112,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.4928603166996521,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 1.033,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.4963936121245415,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 1.0103,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.49930368922719987,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 1.0194,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.4958003546305566,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 1.0269,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.4857999840570667,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 1.0152,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.4945920658510709,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 1.0075,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.5101601993761083,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 1.0064,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.522801296152262,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 1.0589,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.5016957468087143,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.996,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.5115625561092892,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 1.0492,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.4892861296656082,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 1.0176,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.49064776961883494,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 1.0214,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.5143189815362103,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 1.0103,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.5039555000563258,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.9913,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.4928163598108281,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 1.0204,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.4978815063732532,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 1.0154,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.5057939564641576,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 1.0337,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.491880334720136,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 1.0032,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.4932083455044759,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 1.0068,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.4851625740904872,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.9758,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.500545348381862,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 1.0166,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.4942998052998467,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 1.0129,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.5006596410963156,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 1.0153,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.5150221061182075,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 1.0414,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.4837484590732504,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 1.0346,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.4974876180795454,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.9967,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.4921449576878621,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 1.0076,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.4968034747164335,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.9863,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.5077785466527265,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 1.0199,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.5042529021093726,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 1.0327,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.502883811778466,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 1.0139,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.49341468377986025,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.9869,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.49574855217347,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.9998,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.5078825929495664,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 1.0367,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.48995182452813907,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.999,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.5051309459476098,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 1.0124,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.5017878889713839,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.9956,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.49461201950825023,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 1.0113,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.497281233890468,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 1.0348,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.4836313083956368,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 1.0095,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.5006282340926338,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 1.011,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.5187326129197486,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 1.0189,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.4951434024517875,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.998,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.5052536437182817,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 1.0364,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.4797415042999193,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.9977,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.48713718247330695,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 1.0114,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.5046354016193971,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 1.0065,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.49119172531215427,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 1.0413,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.47371740914429633,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.9785,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.49372718803416893,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 1.0175,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.5072195090280328,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 1.0135,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.4854708995329299,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 1.0041,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.5143277382923933,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 1.0074,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.49808789875293635,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 1.0291,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.4915655459675146,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 1.0194,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.4931884841564934,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 1.0195,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.5154405318968666,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 1.0163,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.4791982907838555,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 1.0339,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.491762879146296,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 1.0213,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.4938588278499789,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 1.0022,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.5008782188066053,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.9737,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.4967481607948311,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 1.0067,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.49505867269631,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.995,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.4953443062360086,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.9846,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.49790525643393985,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 1.0314,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.4922814074766445,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 1.0217,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.49905919772422347,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 1.0122,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.5089702627608678,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 1.0058,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.488671456781952,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 1.0102,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.5023466187969754,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 1.0467,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.4990832136482074,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 1.0323,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.499900688973547,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 1.0125,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.4854762906216797,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 1.0114,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.5007803916388764,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 1.0172,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.47732787820507067,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.9795,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.49578828194646685,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 1.0204,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.49839740589079135,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 1.0149,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.4772554631452556,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 1.0052,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.4865164225648338,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 1.0102,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.4887645798657354,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 1.0298,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.4841358736553759,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 1.0324,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.4843089794456919,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.9864,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.4966419261356894,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 1.0271,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.4911933143504285,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 1.0216,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.4799248263133449,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.9787,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.49261969330245564,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 1.0185,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.4922636117132134,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.9805,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.4911675829838036,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 1.0202,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.4792586775616364,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 1.0137,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.48548471515497355,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.985,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.5093873097961316,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 1.0291,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.4871176382246962,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 1.0215,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.4922876837259011,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 1.0045,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.4922080483788156,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 1.0042,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.5111775893933592,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 1.0308,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.49093480823021257,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 1.0011,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.4817305073545766,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 1.0458,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.49216446673111736,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 1.0147,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.4862800221761653,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.9786,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.49769535777059004,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 1.0225,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.48259574662931726,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 1.0252,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.49499694007668565,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.9895,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.4936531734447972,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.997,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.4920178807659698,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 1.005,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.5057628377410248,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 1.0388,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.4782983401306831,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 1.0253,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.48888799812266964,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 1.0143,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.4959897873661303,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.9775,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.49824354779671787,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.9992,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.5016009139481801,
+      "learning_rate": 0.0,
+      "loss": 0.9941,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 1.6351152977299374e+18,
+      "train_loss": 1.317481616392792,
+      "train_runtime": 13924.3936,
+      "train_samples_per_second": 11.347,
+      "train_steps_per_second": 0.089
+    }
+  ],
+  "logging_steps": 1.0,
+  "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": 1.6351152977299374e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}