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": 7.118504106173568,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.8132,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 7.437683771850795,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.8127,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 7.3698836493817375,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.803,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 7.6564056500006705,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.8152,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.933928627349438,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.8024,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 6.807868230017765,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.7892,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 5.299207232187285,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.7513,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.937063919901095,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.727,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.9779039354752315,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.6709,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.7176548211212643,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.6756,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.699552832101438,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.6974,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.8229288968436688,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.6633,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 3.6077652797484285,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.6191,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.648369460468369,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.6205,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.2736348179536368,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.5961,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.850691343692402,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.566,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.765869915821997,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.5437,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.3067804189327497,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.5459,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.992192763991348,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.5442,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.4682340885878082,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.5026,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.2873186474139044,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.4996,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.7034061525683986,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.4769,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.6843327947547262,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.509,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.1009732937872485,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.4916,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.8529767683892491,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.4692,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.0117563395914635,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.4612,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.8326015279207485,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.4813,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.8300376991324209,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.4538,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.7810718715726266,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.4457,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.6634811012949229,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.4401,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.7707205843359554,
+      "learning_rate": 1e-05,
+      "loss": 0.4536,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.7186670748695088,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.4295,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.6999183858921999,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.4468,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.6597331450644618,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.4212,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.6753001259094199,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.4135,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.5858039407129834,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.4191,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6331230832665196,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.4448,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.5874504121904155,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.4157,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.579186021280984,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.4064,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.6021923166709635,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.4222,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.533583019356421,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.4152,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.599143963947017,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.411,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.571013393995223,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.4063,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.6313600403447496,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.4096,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.5428780757678995,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.4071,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5958015024177655,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.4109,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.6464267505702169,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.3975,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.5461126640114172,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.3912,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.6915763653420839,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.4072,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.6581046818804854,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.4091,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.6405593535778749,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.3986,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.6215845948051093,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.3879,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.5584759145799774,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.4066,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.6838136779911529,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.3909,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.5797921063831135,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.3797,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.6641770206969924,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.388,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.48969945588166247,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.3927,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.6215341433504239,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.3847,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.585582596644805,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.398,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.5878874464678583,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.3915,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.6078087675099241,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.406,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.7027056694070521,
+      "learning_rate": 2e-05,
+      "loss": 0.3733,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.607865997602758,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.3872,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.9214397546779594,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.3772,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.6134932123317298,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.3891,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.8712631758660739,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.3796,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.6599348023730984,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.3793,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.7674385544980429,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.3955,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.623894194917959,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.3741,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.6550403032958642,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.3695,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.6873621221090948,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.3789,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.7400733693473771,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.3769,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.7399519275147962,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.3846,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7656264296356046,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.3628,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.9395682565672755,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.3843,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.6050007854793501,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.3653,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.8053390695737289,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.3844,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.6969477951311942,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.3711,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.6462773663254329,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.3801,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.7527456702975275,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.3702,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.6345608678130361,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.3663,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.5918841227511178,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.3694,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.6450356495832293,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.3619,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.5842880022130709,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.3697,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.6455247797148725,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.3705,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.6905906494585939,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.3741,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.6428945310044967,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.3701,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.7259797960786987,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.3898,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.5961927316060215,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.3966,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.6823565772257351,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.36,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.5737934640760237,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.3761,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.7572451209593847,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.3636,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 0.9765901179329265,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3707,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.7541065670751819,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.3777,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.68775543263562,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.3733,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 0.6790960545665073,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.3626,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.6941610556889463,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.3849,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.6511725868757499,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.3719,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.6934305049574052,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.361,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.7779872318693448,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.3585,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.7907047847741482,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.3731,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.5779424116039841,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.3822,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.6426807673797272,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.3703,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.5660426567230197,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.3546,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.6210509817911559,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.366,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.5955653099658536,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.3571,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 0.5449817769716765,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.373,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.5819299006225636,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.362,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.6143841142512009,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.3737,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.6738577140111235,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.3618,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.7745392599761887,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.3488,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.4711541057918743,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.3645,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 0.6738250246096954,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.359,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.7024068879314533,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.3527,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.9193386315536389,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.3608,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.7258932294314141,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.3754,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.7808892826288459,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.3753,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.5209577684710155,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.3708,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.7317227984394186,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.3493,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.7683347411111879,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.3774,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.7543571281634498,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.3652,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 0.5977533334986536,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.3594,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.773843650081236,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.3613,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.8748711941559487,
+      "learning_rate": 4e-05,
+      "loss": 0.3767,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.6857719478558509,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.3578,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.7944509390814164,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.3571,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 1.013642814704965,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.3663,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 0.7567319632453028,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.3664,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.6420174019654132,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.3757,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6388401215124393,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.3892,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.6351483038800267,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.3398,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.5670557896395965,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.3577,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.642322749123987,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.3601,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.6049476563658008,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.3686,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.8385148180725028,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.3676,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.6624198549347596,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.3612,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.5695159364284791,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.3552,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.6521845225537456,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.3745,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.517009727708426,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.3615,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.5733911369939108,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.3532,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.6023803177657967,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.3638,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.5832446086972279,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.3556,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.5067349431095973,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.3535,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.6831689718798493,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.3532,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.7898772161315158,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.3673,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.5394693613891067,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.36,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.6859049033861891,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.353,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.6612873785407317,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.3431,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.5247379860781273,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.356,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.65025675718642,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.3689,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.6323551300292811,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.3664,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5820028406549974,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.3665,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.6448592909986827,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.3375,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.5540832404937497,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.3557,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.6245489220291308,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.3454,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.5334685789760342,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.3523,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.7568952238189596,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.3538,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.525657481430182,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.341,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.6308579591847122,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.3601,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.5574897311248597,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.3645,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.4671554493590328,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.3575,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.5296103173063593,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.3388,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.4205925087735219,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.3412,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.4001381805991677,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.3232,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.4845915681572124,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.3434,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.4539393043503738,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.3412,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.4242615256011767,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.3645,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.4742882847838134,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.3342,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.47453802565060205,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.355,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.5360959490778419,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.3567,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.5526462066721366,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.3357,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.48272455205581916,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.3559,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.5004600352445369,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.3637,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.4600282291112551,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.3387,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.48321902215857504,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.3481,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.479318267028018,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.3646,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.5349155057244624,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.3559,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6311977755035523,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.3538,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.49907440591488766,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.3581,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.4963308140929735,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.3518,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.5012563392469723,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.3649,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.3886161090285058,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.3428,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.4850031944313548,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.3455,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.47252251208503765,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.3414,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.4299249005567526,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.3363,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.4636854450517645,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.345,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.3956285943647656,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.3703,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.4857592660542645,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.3275,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.420541912717,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.3278,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.588325897565343,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.3555,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.47107323803621304,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.3485,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.5502058064329407,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.3491,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.49672100506206945,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.3404,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.4705188817651836,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.3413,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.4739747027581809,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.3402,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.563217163252234,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.3573,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.44250622971692716,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.3353,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.6069501616387124,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.3496,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.46237359839553305,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.3451,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.5037121094291322,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.3447,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.4902784855929594,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.329,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.4040019352694586,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.3422,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.5062114584928085,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.3401,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.4657951722053156,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.3685,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.4686605866438157,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.3536,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.4286568740418365,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.3524,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.43569334213669314,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.3282,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.4124553897012794,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.3477,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.4306455305390766,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.347,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.39269935670444206,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.3377,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.4451854561015899,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.3677,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.39095243177664185,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.3371,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.42762784702854967,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.35,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.39766027387905717,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.3497,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.4483937791889805,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.3432,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.4401446486842396,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.3538,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.48214895563986415,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.3325,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.3954787793929229,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.352,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.3926384777725948,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.3183,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.4470001672942572,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.3366,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.39762246441400567,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.3342,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.525583485892809,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.3428,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.43864357982407903,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.3288,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.41720348347028685,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.3283,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.43554611093899026,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.3332,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.458461665051269,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.3427,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.45260173384094615,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.3379,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.45420446531928793,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.3403,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.47868812338318073,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.3411,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.4808772344329161,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.3485,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.462342837002592,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.3249,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.46213284410448435,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.3439,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.46764532900610817,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.3419,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.49654666839459016,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.3269,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.4351238411199033,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.3358,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.5072636078128997,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.3487,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.42499017389397853,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.3402,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.45076393612418514,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.341,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.41767894724500004,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.3368,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.39589741851895915,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.3345,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.417706994937218,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.3309,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.3734394613686048,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.3421,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.3553274033115188,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.3359,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.37224689139285483,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.3407,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.40459843209173485,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.3357,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.40209859672868786,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.3235,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3680711709038999,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.3501,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.45453137173789293,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.3056,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.3597686636520099,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.2922,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.4479302306812575,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.3029,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.40894192291807996,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.2899,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.46171514796455065,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.2814,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.4370107654886658,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.2714,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.46792556724792506,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.275,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.39228725628041017,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.2693,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.4864376687520969,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.2881,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.41737875001174185,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.2837,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.45106946855191077,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.2812,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.5615062303962325,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.294,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.44446504133523224,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.2903,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.4395040345359579,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.2885,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.41990036403056485,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.3068,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.4425763446744967,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.2857,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.45478682304087464,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.29,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.47820876128426276,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.2933,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.40784745707908865,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.2719,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.4277264489556251,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.2831,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.5205783483801626,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.2774,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.39947534569535725,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.2805,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.5248590696557114,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.3083,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.4613979460196519,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.2808,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.47558723995803276,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.2973,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.5517033822059885,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.2798,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.3765361598479448,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.2994,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.4740388493195104,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.2781,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.4006347356105626,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.2819,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.46348724469371394,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.2921,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.41850117457969227,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.276,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.5054895150058392,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.2958,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.3847725550948356,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.2894,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.46805657782164917,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.2812,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.39635714231678043,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.2705,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.44120573897898296,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.2867,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.4991019350971027,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.2808,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.42011956595389005,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.2681,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.45523022810599484,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.2866,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.4305025538394865,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.2959,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.3530264814260184,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.2945,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.4469402085130078,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.2877,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.3425411952701144,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.2634,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.4719102994046611,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.2887,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.3787483927794099,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.284,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.39159256201250237,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.2844,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.4432635256757022,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.2818,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.4264019711851413,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.2914,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.41626852653572227,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.288,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.4405589520952878,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.2719,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.365641624094588,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.2739,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.43501322036693396,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.282,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.47748339470417905,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.3087,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.3820438683055499,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.2794,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.41771674051684604,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.289,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.4035297512087731,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.2897,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.38543636397779346,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.286,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.378822006919542,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.2803,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.3829360503830937,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.2833,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.3710647720216621,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.2825,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.3559491748387755,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.2771,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.37784424426443647,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.2924,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.3624991185689394,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.2952,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.38028105113308164,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.2901,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.32144401421560137,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.2793,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.36992644256094537,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.2823,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.3844936335876559,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.2902,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.34482764705082225,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.3022,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.4413298671102936,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.2965,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.3420094055666824,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.2906,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.41745553381416295,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.2873,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.38590295958604826,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.295,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.37831512497432096,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.2929,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.4342483932333117,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.2707,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.3406045117971642,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.2802,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.38483029417273446,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.2889,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.3825673135173233,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.2883,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.37705797482948794,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.2943,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.4022249135043645,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.2845,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.4073214343731977,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.2825,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.40601002233261735,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.2739,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.37495419971174376,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.2845,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.45289064077094304,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.2938,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.40921014231031516,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.2836,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.5252306841741873,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.291,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.3833498855352915,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.2762,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.4506443610829949,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.285,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.4021181821034779,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.2864,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.3870979174986855,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.2949,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.39380861846872767,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.294,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.3577818980428499,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.278,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.4012081405623277,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.2828,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.35108465564224217,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.2813,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.3654985894008831,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.2823,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.3995379159031071,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.2973,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.39006430331208436,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.2842,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.3363212844582717,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.2715,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.39083022549341195,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.2976,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.3846921026146934,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.2959,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.3558568183200052,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.2708,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.36616259994632433,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.2816,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.3586112534203935,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.2907,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.32578910486637436,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.2897,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.33999737181575623,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.2884,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.3460773467528847,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.285,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.3450480974150443,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.2778,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.3495953698722658,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.2889,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.3520315779926534,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.2934,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.3843958344639909,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.289,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.4005040337249368,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.2826,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.3626519878753971,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.2797,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.37013442902177557,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.2816,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.34690733469083934,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.2977,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.364692795418529,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.2872,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.3381096480751062,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.2757,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.3696103809169015,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.2801,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.3728738282879922,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.3019,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.36859272925939335,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.2867,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.3439755306028168,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.2818,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.42240020057585304,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.2855,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.3692194977224418,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.2801,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.3924240108835827,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.2866,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.38227056062120207,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.2833,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.34741443906100117,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.2882,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.34007146330834387,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.2866,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.3870454268589245,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.2663,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.3649109155398038,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.2963,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.3687351720967959,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.2888,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.41590013074941085,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.2866,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.36178058252063344,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.2731,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.41584638562917337,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.2807,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.43552094979176426,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.2834,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.37914098056190254,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.2866,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.42998537361959416,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.2815,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.3338627462385257,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.2892,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.37929180491136455,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.2814,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.3236070737288094,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.2926,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.3445694018063381,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.2864,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.37019446747176155,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.2943,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.3275570337141377,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.2821,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.3970163817773542,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.2895,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.36349255566840216,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.2859,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.3972431148274603,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.2891,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.336992504284258,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.266,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.3618231128035915,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.2844,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.32332659283476556,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.2926,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.36166429592646276,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.2842,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.3687164731540504,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.2894,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.38673606053441334,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.3045,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.34395811044600894,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.2899,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.37744433467501115,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.2868,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.37994669238925194,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.2922,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.34667842415734135,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.2897,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.3788023793376194,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.2926,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.3385210889689392,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.2828,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.3039046401194759,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.2874,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.32242428851618543,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.2855,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.33561271634754675,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.2845,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.3492086247935101,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.2662,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.3318092610131119,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.2816,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.3528521287731289,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.2852,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.35567734115999783,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.2898,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.3810413204777462,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.2876,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.3560976105599745,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.3125,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.3155605773661707,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.2883,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.3480608737209863,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.282,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.3116247441880544,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.2794,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.3654953262529568,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.306,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.35840971484967077,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.3045,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.32893826649976227,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.2965,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.3964958214979197,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.296,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.3209581274294714,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.2808,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.35326022114413735,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.2916,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.36805756441913023,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.2924,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.41178323816629675,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.3002,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.33930772555916777,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.2912,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.3831753440629763,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.2842,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.40780261632508735,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.2869,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.4575739841611893,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.2945,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.41587905695399524,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.2825,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.45542432880850064,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.2889,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.39174658191727796,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.2891,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.42104923315673465,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.2899,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.3525310187754441,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.2746,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.4002337095769349,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.2847,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.34059097437951835,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.2802,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.3642006580319902,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.2923,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.4318547628908974,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.278,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.35464919766593483,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.2909,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.3860880319966583,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.2839,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.33563554237332116,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.2869,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.3456649511531796,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.2819,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.3431127901052577,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.289,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.31321739904489215,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.282,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.37387332150158137,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.2858,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.36371781383768154,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.2836,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.3959413476810763,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.2819,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.3700524502578912,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.2783,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.3346176072806406,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.2699,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.37220352539136936,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.2956,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.35291956486331066,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.2604,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.3484115399418516,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.2971,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.32286365182411725,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.2747,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.4196535718525144,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.2823,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.3255926194774518,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.2791,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.4359365437805825,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.2954,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.33608844158667645,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.2967,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.36799334413362117,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.2726,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.3618685935996532,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.2961,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.34342180427913654,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.296,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.36818249577833106,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.2895,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.36371228394361976,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.2668,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.33660317658754985,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.2821,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.36497218491337646,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.2766,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.3281943773147179,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.2778,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.3628932082364025,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.2969,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.3665502571130032,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.2846,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3340767570089708,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.2873,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.3755287011977607,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.2996,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.34295981947435816,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.2605,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.3616074236060979,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.2781,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.4198231150710399,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.2765,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.3229117664549257,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.2962,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.4255744938871955,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.2726,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.31721033540569354,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.2784,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.3617759291612736,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.2848,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.3687903757159512,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.2877,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.3312223053838889,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.2831,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.36338728568237266,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.2906,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.3341923415578516,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.2987,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.3216987499766274,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.2731,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.3677694018016833,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.2897,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.32433028220952875,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.288,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.36512848836271267,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.2984,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.32356904326807,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.2772,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.41275221556361524,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.2804,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.32265430069005413,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.2789,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.4094409616897546,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.2772,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.36754265119416585,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.28,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.3561567074898979,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.2907,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.38420131417965764,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.2724,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.37995967777613787,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.2764,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.3235346935376628,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.2762,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.47918325076539586,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.272,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.3582932582301638,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.2846,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.46286513574448734,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.3021,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.379211622363177,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.2785,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.435606566707197,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.2188,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.3492698298537602,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.2053,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.8821374954367378,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.2273,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.41564567416182807,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.2053,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.40933161136132573,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.2176,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.35004601529498247,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.2207,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.3498756413520669,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.1933,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.3543332479605268,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.2049,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.3700850757070293,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.2083,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.3576834966210862,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.2156,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.3290733703777842,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.2098,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.3130925496160468,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.1993,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.36045108106185275,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.2158,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.32249250370825955,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.2171,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.38219338130828867,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.2059,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.33289834738215335,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.2197,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.3552831733348515,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.2106,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.38995324345778654,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.2015,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.35880802127496875,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.217,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.3645842341655114,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.1924,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.32605189835144704,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.2159,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.3918108378398281,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.2223,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.32941912809268514,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.2122,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.35822968353885043,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.2072,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3609039347580025,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.2054,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.32819783449499135,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.2036,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.34164427241071793,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.2042,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.32224425344441626,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.2038,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.34744286543603764,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.2128,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.32309100089307397,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.1991,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.35038900465102496,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.2252,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.38339829664246466,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.2279,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.32204807423933396,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.2085,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.3923671486006484,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.217,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3370292680067782,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.2087,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.3030571862646256,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.2071,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.37469586468881155,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.2146,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.30804787000576045,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.2226,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.376804885674712,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.2181,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3123368989221429,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.2026,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.4322803011127289,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.2309,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.34948873941249314,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.2111,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.4088036515472493,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.2145,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.37800228716994266,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.2116,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.32092307030075023,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.2095,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.431025294825748,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.2258,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3001373817207075,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.2184,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.35052847622212774,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.2076,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.35785215066095344,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.2118,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.3285851367687868,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.2141,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.3754731068805666,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.2112,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.36025784699898467,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.2203,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.3279934950347539,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.2255,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.36152080390580194,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.2283,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.3732166931732956,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.209,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3488038348487151,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.2211,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.3161392010264208,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.2054,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.3209771352727502,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.2137,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.3267852198219319,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.2101,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.3242002017241673,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.221,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.29486024983431597,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.2232,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.34475241857620276,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.1917,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.3138443445847884,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.2099,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.3172320172000138,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.1902,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.3426318534893546,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.2406,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.31011986586493856,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.2112,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.367492075862912,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.2069,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.2865385114284549,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.2056,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.32258751992711154,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.2119,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.3179213493001169,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.2146,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.31334126173339044,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.2097,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.2916388148592191,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.2031,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.34641021925993326,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.2105,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.3121629708345099,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.2334,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.3147067566599831,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.219,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.3274967334161985,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.2089,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.3303106472808675,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.2266,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.29537211760106213,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.2118,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.3693423609108331,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.2072,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.33551709128132784,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.2101,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.33979823188749364,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.2124,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.3483588353022114,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.2101,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.33694022870450846,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.216,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.29604008132007936,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.2045,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.30794263051763254,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.2047,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.334282340871069,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.2351,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.31697193482138136,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.214,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.3515290533199975,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.2209,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.3368051198269745,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.2175,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3217169323686682,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.218,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.3243397236603551,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.2253,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.30702193315949544,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.2109,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.3231771534191789,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.2082,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.32795450948222776,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.2093,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.31217109373163676,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.209,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.33235747727971715,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.2221,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.3142093055130027,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.2146,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.3307549665012373,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.2075,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.34951231886770323,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.2144,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.3027081710699269,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.2113,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3886069096277919,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.232,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.29756930658361774,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.2191,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.29756218698877457,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.2171,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.30595638337579467,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.2073,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.298167981016101,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.2296,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.32552563624582426,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.2077,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.3291859330007517,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.2047,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.3013961383090731,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.2089,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.3442148359330752,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.2078,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.3133025194067528,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.208,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.34655088737643974,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.2237,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.3386711444365701,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.2189,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.3257768235935944,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.1978,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.3289215647975061,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.2043,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.3033272469443143,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.2116,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.33038137524866595,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.2109,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.3275952917320109,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.2307,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.38743035651840313,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.2119,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.30705321318129586,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.2068,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.3350128617864359,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.2186,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.34283699384017396,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.2194,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.29896430122046624,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.2147,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.312143060656905,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.2068,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.3187852085462923,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.2094,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.3007942186397909,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.2069,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.30997549126396534,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.2176,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.2828888748137361,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.2073,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.29539448428145665,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.2023,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.32663497848420353,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.2217,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.29227082376071495,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.2094,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.29781646122820793,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.2188,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.3147386741681833,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.2041,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.3006213113332684,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.2205,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.2980257384986708,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.2213,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.2996118531885569,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.2274,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.2939273456700929,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.1949,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.32685992599830954,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.1942,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.30399033158597,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.2047,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.28828024586632495,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.211,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.2977407058906097,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.2215,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.30178392772745105,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.2047,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.29549616986239646,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.2007,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.3151457015586685,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.2131,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.30680565469020377,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.2204,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.3302185710359948,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.2121,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.3059141319306752,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.2297,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.3010162893890906,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.2191,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.31690118557480634,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.2232,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.30243352677518465,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.2268,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.2887463689971157,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.2091,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.28892365958770017,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.2188,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.2932784749360128,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.2171,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.2928260096368639,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.2185,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.30808305020352933,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.2354,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.3095692870392122,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.2205,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.2810435593113388,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.2137,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.29293199788559315,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.2289,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.29759930470744533,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.2025,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.2824446140469468,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.2159,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.29113760776200115,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.2213,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.2982533115774585,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.2122,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.2802781362479399,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.2185,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.30203337369617,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.225,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3015296938789539,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.1924,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.311440082317942,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.2046,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.31886236352220054,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.2032,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.2963018764000533,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.1996,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.32918721640816534,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.2108,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.2938028261656638,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.2114,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.28271399715448325,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.2004,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.33289227037297536,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.2112,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.31750195887557464,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.2238,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.30641041955130693,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.2034,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.33736305490092616,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.2278,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.3655686202851491,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.2284,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.3105596596157722,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.2102,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.32403724843077497,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.2114,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.34238256473786527,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.2091,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.3374816671475373,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.2146,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.33646011944550125,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.2158,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.3616742339041378,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.2036,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.3138611667129198,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.2304,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.37537758959289846,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.2116,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.3034912810256056,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.2037,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.2961028425695122,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.1981,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.322145330380957,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.2197,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.3178696244466162,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.2139,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.3039265331424478,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.2258,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.30187668684637525,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.2244,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.30465854066715714,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.2153,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.3053968580169965,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.1993,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.2816900091712236,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.2165,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.3224611921952675,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.2248,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.3136148498621113,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.2074,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.31969490188724076,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.2295,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.3035370855476504,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.2064,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.2928123369640044,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.2011,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.3054994647649771,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.2154,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.2911834642046291,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.2224,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.27063362461086077,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.2101,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.29681132793895826,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.2025,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.2958406690555019,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.2113,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.30152024571644304,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.2226,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3014227863082137,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.2248,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.28017694764298984,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.2034,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.30214870335248867,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.2031,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.29683662341250105,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.2067,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3124941957727727,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.2153,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.29908187464337915,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.2099,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.27811619775745067,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.2128,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.31383229001211566,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.199,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.29333466427819277,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.1921,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.3048645683479711,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.2154,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.2905729870014576,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.2153,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.3326896546970244,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.2151,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.3215608430324406,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.2189,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.28714355469206765,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.2201,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.33748640069107466,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.2081,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.3090461862935928,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.2002,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.3102416994838391,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.2045,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.3216282361818051,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.2077,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.3185921547724932,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.2027,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.31142042193998093,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.2114,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.3291106986311232,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.2143,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.305664544115776,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.2081,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.29661789057403587,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.2103,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.2895962940728554,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.2092,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.29161098856454837,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.2284,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.2926161786927011,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.2117,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.2902736898669816,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.2062,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.293068517450829,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.2094,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.2920765205122855,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.213,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.30424113342140624,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.2054,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.2861388307634325,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.2149,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.30592440437360463,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.208,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.3082669596881048,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.2065,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.30898126083694744,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.2102,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.31387340390836915,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.2055,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.31189511202169645,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.2197,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.28865327263060964,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.2033,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.3304711531875617,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.2127,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.31579698008991786,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.2037,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.3075111737014022,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.2024,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.31571535073413604,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.2273,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.30337474348896604,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.2052,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.3104512870346111,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.1898,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.29041964518464214,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.2116,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.39166428844945256,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.1508,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.3556689046724947,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.1323,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.38047072807024374,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.1496,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.4997157989688932,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.1502,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.3910446278254283,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.1381,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.3415586279746324,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.1406,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.34496117559005085,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.1394,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.3488984923421123,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.1499,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.3144473630576686,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.1427,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.2963879685931031,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.1446,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.3177984346932708,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.1293,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.31960616493895544,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.1326,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.3042182379075833,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.1328,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.317143934099883,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.1359,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.29731383212206647,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.1511,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.3163842953853432,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.1451,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.30871471078974044,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.1445,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.29787163033973685,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.1411,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.29101936518890925,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.1394,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.3220387989050703,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.1545,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.3284098588075202,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.1364,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.29674423333494027,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.1354,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.3055669559329136,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.16,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.2949391940327327,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.1282,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.32223372692802116,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.1443,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.3321764901257534,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.1495,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.27938927867157093,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.1386,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.3023688822328664,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.1534,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.30733607153381937,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.1271,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.28351901323783507,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.1435,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.32383738447312366,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.1501,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.2981068068953314,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.1455,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.32121462300607795,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.1399,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.32277226419172994,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.1317,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.3021550368755597,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.1478,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.31417411381801247,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.1407,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.29043348958219384,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.1396,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.2932349553584771,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.1469,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.31023556079685494,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.1337,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.29100756429283076,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.1357,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.28732719916618876,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.1438,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.30546016490989075,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.1414,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.2940258380260032,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.148,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.2914958227814483,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.1311,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.3122300867938819,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.1372,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.2765287324365358,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.1415,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.28566849000259414,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.1508,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.30769450172720025,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.1538,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.2809611665128305,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.1391,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.28714037566022427,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.1521,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.28280246740839005,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.1394,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.2754983878250665,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.1347,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.28189718128053154,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.1445,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.27959019345941893,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.1458,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.29113315347817303,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.1417,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.2744952183414928,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.1422,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.27393169773347426,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.1334,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.2924305890204787,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.1404,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.28168598107448367,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.1478,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.27416055345543466,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.1355,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.27958909967728746,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.1399,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.28325797556796783,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.1407,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.2902090958130782,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.1489,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.2858091528909106,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.1519,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.27789746651413855,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.1535,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2893571561075367,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.1409,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2709112355469938,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.1475,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.2852850530347169,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.1359,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.2848171909317087,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.1512,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.28193851513068335,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.153,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.2814064122999566,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.1503,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.27580630312027354,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.1289,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.28340901834615273,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.1253,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.2670554404344175,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.1671,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.26899980976800153,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.1417,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.26924358013115746,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.1447,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.27705996462095867,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.149,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.2830893086874125,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.1389,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.2830890126271056,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.1338,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.27729554940942414,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.1523,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.27209463821335506,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.15,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.2799096858071698,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.1397,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.2714159908590545,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.1495,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.27643216698219253,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.1403,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.2751341543752728,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.1428,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2848980992221633,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.1358,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.26630604225905796,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.1479,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.27792591239219966,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.1611,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.2723135345698293,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.1536,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.26459219992165356,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.1473,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.2777237939859561,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.1406,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.2713377518267065,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.1513,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.2822248752124317,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.1437,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.27549431505715255,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.1479,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.27137922481972004,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.1532,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.27717003209374536,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.1452,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.27782541749622713,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.1512,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.28445558992306036,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.1406,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.267672499569631,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.1435,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.28138424214002294,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.1493,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.2827234534613304,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.1454,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.2714073476028501,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.1482,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.26965973631714507,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.1427,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.27517036418960666,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.1431,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.26041857190662504,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.1475,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.2684009046248813,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.1537,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.27860234405146594,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.1338,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.26451279646513814,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.1505,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.2825819665042925,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.13,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.2658401188580172,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.1442,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.27549413233065406,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.1528,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.27567567090007045,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.1275,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.275669038620812,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.144,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.28594868307818105,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.1405,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.2746168375055058,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.146,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.27393093075837693,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.1437,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.2819102876700014,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.159,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.29521306376844053,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.1294,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.2782171905686036,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.1457,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.2650299785012289,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.155,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.2743161211728509,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.1347,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2730653497332922,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.1472,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.2563948060015619,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.1405,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.2544282179384869,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.1498,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.2766853722628602,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.1291,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.2908588273150795,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.1475,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.27454479762403294,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.1406,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.26983421106654154,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.1431,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.2885511863036756,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.1396,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.289814239088395,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.1421,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.2881834487171524,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.1437,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.2819831635153483,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.1398,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.2562018503130822,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.1418,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.2742887045054092,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.1457,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.27753299062902337,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.1435,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.27431317894200563,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.1347,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.2704708325143352,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.1476,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.2680084617088411,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.1339,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.272957973863563,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.1545,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.2726202142484916,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.1339,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.2707649151114059,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.1428,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.2726306219829313,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.1375,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.2701390819338791,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.1365,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.2683931627259463,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.1353,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.261267341511818,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.1377,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.2849280625174732,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.1365,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.27609677122300447,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.141,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.26904341642473,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.1498,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.2735088148986847,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.139,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.26327400481586594,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.1425,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.2719568759077968,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.1497,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.2679866660024736,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.1385,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.29116250161746415,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.128,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.2825834128667403,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.1306,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.26326710930787095,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.147,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.2774165578049401,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.1531,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.26871941460067117,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.1411,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.274563042697694,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.1466,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.2743417924464114,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.14,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.2591693542823603,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.149,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2774385876173199,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.1404,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.2658838465913749,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.1429,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.26423465300707266,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.1422,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.27195447428845476,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.1402,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.26547376804883505,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.1608,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.261821476681177,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.1343,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.25933587339987535,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.1517,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.27087035833745665,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.1417,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.26961473215358345,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.1415,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.27576848110124785,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.1361,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.27851036780974775,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.1347,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.26383796744569443,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.1307,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.24868867707343986,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.149,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.26565696647916054,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.1394,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.27359455418722606,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.1488,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.26016469046643825,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.1562,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.27278171818119984,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.1359,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.2710020037383002,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.155,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.26517488804772277,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.1361,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.2699406548055054,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.1453,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.2585956594822022,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.1425,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.2516652729182362,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.155,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.2787081878120419,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.1352,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.2704742527720519,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.1445,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2636957734443454,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.1508,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.2666551170507827,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.1307,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.26085931802412987,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.1486,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.28968179494505414,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.1305,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.2597433438593562,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.1331,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.26126513485836617,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.1436,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.27035178559322115,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.1388,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.2676682183529048,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.1355,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.25507392507852805,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.1516,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.24968515801200206,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.1456,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2596250827617097,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.1416,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.25844970983702364,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.1467,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.2600542297610201,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.1454,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.2603207004277424,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.1454,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.2588498274715568,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.1337,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.28535050146030083,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.1383,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.27831388032830273,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.1323,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2677436043904479,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.1233,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.2790269167470897,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.1484,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.2811195400863537,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.1362,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.26947567340898,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.1523,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.2634951524274826,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.1329,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.2754450153329781,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.1354,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.26349804238874347,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.1462,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.26957195705333814,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.1385,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.2589650204679684,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.1425,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.26833711999413695,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.1368,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.2711362964514294,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.1519,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.25897460484083157,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.1391,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2616695935211343,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.1521,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.2589598751744456,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.1338,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.25946328408218444,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.1362,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.2616021544937748,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.1282,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.2682668613367297,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.1379,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.2642776016820741,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.128,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.261916802412176,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.1403,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.2643086600460106,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.1397,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.26237270587469325,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.1434,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.2661582644995895,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.1322,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.2672775964304117,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.1434,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.2666120305779957,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.1274,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.2698082769593131,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.1338,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.26048009170562225,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.1393,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.267466664850153,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.1371,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.274460242061545,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.1322,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.27947048798514634,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.1332,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.28134122191438293,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.1318,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.2657466662354418,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.1488,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.2703464777517752,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.1355,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.274319564514985,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.1258,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.27765328067917233,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.1438,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.27520560295086793,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.1417,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2542951819748229,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.1369,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.26071026831036276,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.1249,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.2681603041030777,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.1251,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.25779169625477627,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.1514,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.26401581594261186,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.1563,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.25584179290342174,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.1399,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2673204906507546,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.1468,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.258652998257817,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.1265,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.2699056184877001,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.1319,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.2598000244867781,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.1283,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2602641393928329,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.1413,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.3306811056004185,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.1045,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.3172633319051092,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.1093,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.270605607912256,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.091,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.22163054415433872,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.1052,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.2796234817965464,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.1029,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.3268946709812274,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.0999,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.3417728091802175,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.1046,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.3686280959919504,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.1002,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.2780540176862346,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.0897,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.26402007096848656,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.0894,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.2665353306051486,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.0963,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.29715480590580334,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.0931,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.26427674088488223,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.0954,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.25770597418827973,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.0873,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.2426846753089434,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.0916,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.2558514230815182,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.1023,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.2414222698950163,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.0907,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.2603379587294967,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.0908,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.26165239620238334,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.0992,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.26645262398402686,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.0916,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.24710991878062707,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.0842,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.23076755504134053,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.1064,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.24586586030075988,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.1067,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.2408786622461862,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.1051,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.24627739703639479,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.0883,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.23474026522897254,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.0987,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.2474220824288691,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.0932,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.23201451086997496,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.0877,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.23018734418351391,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.1035,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.2356279316610834,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.0883,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.2264626375203755,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.1049,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.24658556488834368,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.0951,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.24915182959141374,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.0892,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.265149949007998,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.0977,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.2505186392998259,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.1077,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.23104677347698574,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.1013,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.23659578633859182,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.0971,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.23612673110926374,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.098,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.23588045232244723,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.0845,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.23971592664369445,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.0865,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.2380577814185428,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.0927,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.24422771494460313,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.0971,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.23211936394714078,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.0921,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.2319861645041723,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.0859,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.24443432631312503,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.0936,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.24542601279230875,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.0935,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.23648225870972994,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.0845,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.23155872539896075,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.0942,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.23578182136129539,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.0995,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.24233695303256947,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.0909,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.24230957257340105,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.0975,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.2409016240124633,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.0781,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.22607647084231441,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.1,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.22589517547049004,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.0907,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.2369867573554531,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.0881,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.23104076300704635,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.0928,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.22667159204775683,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.0856,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.23349868491589562,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.0939,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.2282937989765456,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.1041,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.23167340494171088,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.0872,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.23321698906697796,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.0832,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.242387657162141,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.0817,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.22113676747091363,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.1038,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.24404995481941186,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.089,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.23259681057060078,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.0991,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.23761441292535493,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.1007,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.229925068466603,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.1092,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.23405265077197984,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.0886,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.2356040363124505,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.1005,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2329280596985319,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.1032,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.2285430559481105,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.0867,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.23466351559321003,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.0799,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.2425571330992205,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.0959,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.22874436097324133,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.0851,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.24075118228189296,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.086,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.23746261317919895,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.0854,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.2342868074120832,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.1034,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.23670021830124952,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.0922,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.23419453469966692,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.0808,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.2212749926155631,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.094,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.2364338550373761,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.0978,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.2329093074424003,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.0947,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.22849832705818468,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.0721,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.22828587651873303,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.0972,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.2466322658434479,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.08,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.23382816787549798,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.1084,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.22981701284482545,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.0938,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.22510448415623258,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.0872,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.23518303502401494,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.0927,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.23717228236566096,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.0923,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.2396836252148668,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.0941,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.23728540757225533,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.0863,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.24200502491406733,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.0848,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.23170598452975708,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.0965,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.22374322550258519,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.109,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.23751280176870218,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.1003,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.24634864192677583,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.0885,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.2342975836319459,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.098,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.22910130687669128,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.1079,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.22202280253123471,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.085,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.2401663146898724,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.0873,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.21592815032098814,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.0947,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.22567166326387936,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.0923,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.23929517124950409,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.0887,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.22447506515959634,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.0936,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.2234060081352694,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.0982,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.2269607118631367,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.1123,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.2357006991068411,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.0931,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.2274687227877024,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.1078,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.23557746800877288,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.0816,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.23503637513094455,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.1044,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.23533817706966817,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.0881,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.23184275472878327,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.1159,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.22537609876364426,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.0965,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.22993940221523199,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.0843,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.22765377975620524,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.095,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.23121431092137354,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.1021,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.22997094957789554,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.0946,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.22850544559136451,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.0853,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.2326467281718281,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.1077,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.23311363512909675,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.0986,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.23632499648511054,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.0991,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.23705002823634277,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.0859,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.22755159181743298,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.0995,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.23196504942048365,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.0939,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.22405008035783902,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.0883,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.22584291458903788,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.1057,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.22700035192715207,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.087,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.22794497945617354,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.0988,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.2246662257949943,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.094,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.23014412441809645,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.0878,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.22537559234150165,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.1052,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.23067301947292934,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.0929,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.23698928486075968,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.0959,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.22932429844776425,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.085,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.23693048670963843,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.1039,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.23428794350906001,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.0978,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.23778743049667264,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.096,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.2189703399883847,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.0994,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.23742584059709804,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.0989,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.22969622243015073,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.0994,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.23004855082405845,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.0981,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.23975851459759845,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.0862,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.23357475936269367,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.084,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.24747772460241574,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.0961,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.2354597138551483,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.0926,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.2385330892053356,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.0861,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.23665858474942675,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.0954,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.22246753734891428,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.0892,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.23371740545317238,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.1058,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.23178884601552097,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.0872,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.22139766477477155,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.0948,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.2244422580628035,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.0956,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.2369848166148219,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.1024,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.23065950213233416,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.0924,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.2235067519020081,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.0895,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.22247213954690445,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.0949,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.2274835523551316,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.0985,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.22780987573935,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.1026,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.23955754202763674,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.0978,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.22842897239150192,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.0941,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.22049414718860855,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.091,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.2274011551414241,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.1037,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.21832614222582092,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.1039,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.22747344534162472,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.0856,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.23346526718492613,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.1031,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.23329679143372223,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.0945,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.22665860987345013,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.1087,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.21589324331008433,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.095,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.23709801279163456,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.0856,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.2354192033957317,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.0881,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.23309333028578302,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.0944,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.22979733268280825,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.0821,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.22711521283310585,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.0933,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.23307794025404474,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.0972,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.23776343035870837,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.0911,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.23099221279106677,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.0962,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.2191111749238939,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.0931,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.2256510384902856,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.0862,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.22962121801785745,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.0888,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.2310462652625666,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.0974,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.230669307765609,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.0997,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.23107975942819073,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.1054,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.2299966424958166,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.0909,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.23054021625375945,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.0743,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.23093585885911036,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.1017,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.23109906269676883,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.0968,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.22966908108323128,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.0859,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.2319513029960635,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.0975,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.2408592794664067,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.0865,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.23201390571759076,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.0985,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.22075255581215825,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.0944,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.2214686915354993,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.0997,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.22234172907162988,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.1072,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.2267336462307183,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.096,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.2214043881759293,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.0976,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.2233531446246805,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.0861,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.2260401640540623,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.0974,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.22606676361601863,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.1123,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.23301053222201348,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.0968,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.2244422254293922,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.0922,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.2263891604621985,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.0922,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.22850228715753365,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.0897,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.21665260983038256,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.0874,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.2322663908727303,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.0903,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.22263701626810867,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.0989,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.23172476915683318,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.0912,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.23166336131761758,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.0866,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.23151361094627979,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.0958,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.2386096676010409,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.0974,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.2209785207432619,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.1059,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.23015836867822492,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.0921,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.22337339801460246,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.0951,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.21940572739771735,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.1053,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.22746335546900187,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.0899,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.22929537397389946,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.0991,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.23529992119544382,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.0989,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.23773183031701645,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.0961,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.22735212081336612,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.0978,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.21967481359963173,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.0859,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2495751020771636,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.087,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.22918053733342977,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.0919,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.22650121788339717,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.0871,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.2186204753072585,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.0962,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.24313186649446897,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.092,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.23914852650590881,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.0917,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.24876924124183108,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.0899,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.22283363775505127,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.0892,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.23303296023083442,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.0853,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.22589268012806046,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.1026,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.23080190448426904,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.0866,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.22864642639025923,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.104,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.23185918640921285,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.0949,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.22410784694983513,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.1115,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.22988084277681184,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.0932,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.23361938187186282,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.0851,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.22716373009472507,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.089,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.2307929013342504,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.0972,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.22962601522082246,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.1178,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.2389699301041591,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.092,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.23316333247230198,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.0853,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.2236849739663226,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.0956,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.2237378487496335,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.1047,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.21771032517528033,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.1067,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.22625628427361838,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.1056,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.23338272969811474,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.0919,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.21878313858032142,
+      "learning_rate": 0.0,
+      "loss": 0.0918,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 1.4008313732367647e+18,
+      "train_loss": 0.22568280391002957,
+      "train_runtime": 19446.634,
+      "train_samples_per_second": 8.125,
+      "train_steps_per_second": 0.064
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4008313732367647e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}