diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.6770614396811565,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.047,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.5720165418179235,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.0514,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.623417775783283,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.0378,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.427853019579377,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.0375,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.203484742470254,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.0323,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.91914240519534,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.0231,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.653200315393375,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.9894,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.29945836770982,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.984,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.833396768179839,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.9535,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.5653346246384814,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.9386,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.555804042160347,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.9583,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.9456722501184114,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.9272,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.318871397306766,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.9272,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 4.07063327283913,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.9186,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.7913786244896133,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.9164,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.6834858505295207,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.8799,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.405936501443008,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.8397,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 1.9791497605986823,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.8395,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.4039873607036288,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8256,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.1465113214607765,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.8086,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.179662671326682,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8086,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.2490649562024685,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.7863,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.090364610141192,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.8013,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.8655878788053274,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.7784,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.9818340763815057,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.7725,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.0032551847647133,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.75,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.8660485798610846,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.7673,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.8356392767443123,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.7353,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.8696678972566079,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.7521,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.8453774103561974,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.7391,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.752659977503478,
+      "learning_rate": 1e-05,
+      "loss": 0.7438,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.7983519246846403,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.7384,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.6612182073778251,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.7406,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.6546929312971373,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7212,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.6678617838955332,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7181,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.6347631404559687,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.7138,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6213599715409495,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.7146,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.5635840519987737,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.6995,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.640185628762296,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.7134,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.5609027957462052,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.6875,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.624900349175753,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.6971,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.5536956825674297,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.6926,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.5533351182440647,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.6976,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.4808070978685459,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.6979,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.5988800700398359,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.6971,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5029675511401901,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.6955,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.5986721691041395,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.6977,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.5349508777666068,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.6893,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.682459244273546,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.6662,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.5350059495852396,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.691,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.786144221511296,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.6851,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.8833090098117402,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.6809,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.6083420604292281,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.6679,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.7202289619683798,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.6697,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.9065082262300278,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.6855,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.9748802976433879,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.6717,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.5910391847318918,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.6655,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.7664786076219275,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.6698,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.8007191484332802,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.6666,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.7723733124335609,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.6691,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.6148056842638159,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.6676,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.6199863435081587,
+      "learning_rate": 2e-05,
+      "loss": 0.6645,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.9558925482430334,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.6616,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.6186507195324296,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.6634,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.9250628851063303,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.6566,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 1.1084002371836599,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.6482,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.6788645216611452,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.6582,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 1.0202427504955032,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.6543,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.831469462756205,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.6516,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 1.1221892043411483,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.6558,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.7838115419945407,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.6483,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.5975064950038802,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.6509,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.8009994359233226,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.6397,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.9193397132252863,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.6487,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.109478661764541,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.6572,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.3618950636019105,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.6533,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.7022063260863192,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.6384,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.0649947995281628,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.6319,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.3747622260103451,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.6436,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.859615228785089,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.6478,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.289462354312135,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.6388,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.7343657196221312,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.6321,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.3974132386582228,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.6358,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.8656662671178927,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.6413,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.9191750840701232,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.6518,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.4218831628405586,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.642,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.9318998959127525,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.6312,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.5148232100554209,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.6499,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.0907863437066017,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.6421,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.3895913976100012,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.6345,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.097483878903688,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.6412,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.1340793321999059,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.6406,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 0.8354241477222023,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.64,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.0364434075302644,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.6411,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.3936482105582642,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.6452,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.0590673677139273,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.6395,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.2068108062490708,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.6391,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.3401017184548483,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.6347,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.7882216858436292,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.6374,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.4180246880411616,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.6135,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.0347558265866226,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.6415,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.9500475672785398,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.636,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.290604506020049,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.6358,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.3386386693438679,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.6263,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.2092541902732505,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.6181,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.9855862244339758,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.6372,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.3135634566540855,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.6227,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.1851631199784816,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.6308,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.1370140597817484,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.6185,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.6703819931536202,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.6306,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.9271502506345481,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.6245,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.4925905426651536,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.6359,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 0.891037961574886,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.6164,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.171292950392342,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.6279,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.122861198189587,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.6343,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.9962702045602401,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.6208,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.3155717259920585,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.6421,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.8751241372187302,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.6276,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.0600580349475215,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.6249,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.189068427732068,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.6288,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.947810113638025,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.6359,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.05070342568381,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.64,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.2126193825631015,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.6281,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.7632432070263748,
+      "learning_rate": 4e-05,
+      "loss": 0.6163,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.1370173602477764,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.6355,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.0403493494450824,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.6254,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.9358106089179096,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.625,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.1866519719314461,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.618,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 1.2258465000694299,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.6244,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6784075823638892,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.6143,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.7558431439952014,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.6196,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.8818731502049122,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.6236,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.6536024093973225,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.63,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 1.0139265960820971,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.6241,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.4755076124850197,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.6236,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.8360135721788237,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.628,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.234493298529743,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.6199,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 1.1811705104123817,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.6157,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.1128876496720932,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.623,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.8265956995646768,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.6138,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.894645430547532,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.6176,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.2605622496630187,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.6252,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.6998824156496295,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.6118,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.982641935279461,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.6163,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 1.395493329860821,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.6278,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.8338477912829805,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.6114,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 1.2375194493134454,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.6198,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.1181170459071303,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.6102,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.9630418648657157,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.6044,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.9416009618681032,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.6213,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.1827804822525256,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.6216,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5936739168043021,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.6168,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.2229484691079928,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.6354,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.6990594687364668,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.6208,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.9393377016544756,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.626,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.164746325911562,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.6032,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.6741054594967034,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.6185,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 1.206351287293731,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.6223,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.6519685810342625,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.6144,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.8698313015468718,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.6159,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.8399758461397203,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.6233,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.6613350347583516,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.6104,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.652383564690667,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.6115,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.7427581788428069,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.6051,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.7486191604993948,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.6136,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.7060826682807432,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.6298,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.837524347476137,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.6184,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.7484500244019386,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.6091,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.7263022181502578,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.6112,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.5648214295006123,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.6216,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.6672947156152292,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.6173,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.6194617320637231,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.6056,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.6740848286200452,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.5896,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.6311187334510855,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.5989,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.5546009415944356,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.5946,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.6502763853818285,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.6031,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.8241586708081766,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.6137,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6612517768998096,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.6089,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.8777814495325315,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.6071,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.9889218970797955,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.6217,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.6727486824098018,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.6128,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.5040573855680512,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.6146,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.5810058920223108,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.6017,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.6116811752318959,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.6034,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.4925074703211673,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.6142,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.5743080455690419,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.5943,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.46148576938164615,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.6127,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.5713727051550086,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.5969,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.5576359617403364,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.5956,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.5994054690261014,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.5951,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.603435761844662,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.6011,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.4854290731665034,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.611,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.5414928685159391,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.6083,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.5219215490701048,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.6087,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.531743761258157,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.59,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.5032231309082119,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.6038,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.6316592747527242,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.6067,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.7754824899425196,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.6094,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.6028945602447773,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.5905,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.6077258309649255,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.6123,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.81552034799745,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.5915,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.6556991955591158,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.595,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.5904811463363456,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.6015,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.7441298006191099,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.6016,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.8981193644372232,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.5899,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.858934208953566,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.5897,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.7586793956850301,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.5907,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.6409594140294211,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.5973,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.6479214315218988,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.5931,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.5040489295496302,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.6136,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.5520064886169511,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.5999,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.609204261135067,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.595,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.622484127480476,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.6005,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.5752683432187204,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.5981,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.47533761168088245,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.6004,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.5675955632573417,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.5855,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.715802175101121,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.603,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.693800393180471,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.5862,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.5567402283529728,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.5984,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.5111706058953025,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.5971,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.5520650627220177,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.5974,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.48494104212379735,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.5926,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.7192658084900968,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.6083,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.5635495192809477,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.6022,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.5007426423749876,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.5975,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.4658697943948959,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.5822,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.6199890970579145,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.6027,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6277692226726372,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.589,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.5083632539654307,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.5832,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.607295631574403,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.5916,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.603336797905677,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.5938,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.6292671845834785,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.5893,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5232779143186166,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.5935,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.6202351480862789,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.5816,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.5401030729340527,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.597,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.4830694888866663,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.5963,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.4602968858083967,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.5925,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.5007473650534557,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.5974,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.6157538288466411,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.6021,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.4852010949512677,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.5883,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.5501107267401318,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.571,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.5828315757558827,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.5938,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.5105128851454673,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.5952,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.5864574160328139,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.5804,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.5085484332897554,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.5705,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.4657772927960817,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.5945,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5453197825819873,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.6012,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.6121698418144877,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.556,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.6799666639312135,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.5479,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.7807377312660038,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.5345,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.6707684155147563,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.5354,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.5839615008617338,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.5398,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.6660743799459552,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.5452,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.6458287431054457,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.5395,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.42938225523527734,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.5404,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.6468281912905653,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.5484,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.48832940333614605,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.5362,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.5205253817622443,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.5342,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.5580324541366177,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.5548,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.529395322477276,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5409,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.541270163259141,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.5437,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.5571581707293383,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.5368,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.6178119160690728,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.5391,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.5835267400987517,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5428,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.6264689602609933,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.5329,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.5907939997612563,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.5476,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.5677759946827482,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.5453,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.5955905988581913,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5446,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.4436368624496444,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.5313,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.4346074948630049,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5472,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.49261286381289215,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.5429,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.5619200996092756,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.548,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.5859248416267161,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.5293,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.517244693958148,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5404,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.48763207789902113,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5385,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.4901680428612619,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.5197,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.5273572138777802,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.5223,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5661933692543476,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5534,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.575105551864902,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.5298,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.5409879332739779,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5301,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.5516945692379993,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.5454,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.41929663321827837,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.5448,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.5696596259716544,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.5476,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.5213402304344986,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.535,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.43440826705000524,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5404,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.5957542597502941,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.5345,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.5864665218664469,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.5383,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.5222436786907997,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5408,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.4474689267694553,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5376,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.5336739632341273,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5452,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.5449879529909202,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.5326,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.43230141061125504,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.5451,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.4513209470800544,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5365,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.564739184857527,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.5441,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.596497321143265,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5349,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.45739536273081904,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.5291,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.5721335665948157,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.5528,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.6322467766408432,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.535,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.444004566469624,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.5454,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.5389080540180462,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5431,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.46176721630084855,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5287,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.4975027507176509,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.5449,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.45097135876704797,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.526,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.4171173093406917,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.5323,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.43284819609335284,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.5271,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.43929488843011616,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.5452,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.47052835527532394,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.5524,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.5049202000356442,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.5399,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.4296808306061608,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.5327,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.5128513353696514,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.5342,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.4348543304015031,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5241,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.4318976730972049,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5229,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.5827583887946639,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.5403,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5262456950428089,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5468,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4607330805535689,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.5264,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.5816785952334897,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.5411,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.626120676518109,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.536,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.5468929513836228,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.5259,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.5526088641987981,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5496,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.41434936322310273,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.5312,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.5722115631442343,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5387,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.5600353878805963,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.5445,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.45910766786133705,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.519,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.4795634955246598,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.5328,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.4361396798658986,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.5349,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.5111486927989057,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.5323,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.46827069097324947,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5292,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.4684809901280864,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.5354,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.39782945337442455,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.5345,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.4518518724821306,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.5232,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.5300437621421001,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5402,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.43536681379799436,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.5424,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.39277923956436817,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.5386,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.5820402361595725,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.5315,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.4135749486222358,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.5396,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.4412311227566855,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.534,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.41509430557906624,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.541,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.43450937197371414,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5221,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.45034298521210836,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.5174,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.4140718766606287,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.5398,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.49051937531877077,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.5456,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.4008859175796377,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.549,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.3920025801321086,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.5283,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.4252076927360221,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.553,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.46263614656613256,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.54,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.4521681040950005,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5446,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.5399765524613661,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.5331,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.6045826380475248,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.543,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.5191830410118201,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.5318,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.37634231480339714,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.539,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.5269068189112107,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.5216,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.46619424389315495,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5373,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.3945991404261759,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5414,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.42293654669122416,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.5385,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.55884985416574,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.5343,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.5501465383842606,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.5397,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.47212786188426115,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5437,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.5499121872507033,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.5291,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.653879095743198,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5295,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.43377461992712724,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.5581,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.4247403842885718,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.5281,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.3909114788203863,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5413,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.4048629495032347,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5391,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.38822195884781235,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.5342,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.38095158731002626,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.5385,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.4725644264105119,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.5339,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.4095618189848485,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5357,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.44472584165416096,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.5347,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.4245471721043806,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5482,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.4561192559175261,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5254,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.42598563909070153,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.5376,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.6488979834449538,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.5406,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.49737984637953764,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.5398,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.5290581107009448,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.5305,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.6055441831869611,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.5406,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.6729419130950426,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.5331,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.5239262378734906,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5358,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.5133478212466595,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.5187,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.5955973339465626,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.5319,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.4798613333108687,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.519,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.4416950076276517,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.5225,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.5529043016083527,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5276,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.5357455265245704,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.5279,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.41054214237341463,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5297,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.470823954879595,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.5229,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.452402223098991,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5347,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.505273469575433,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5308,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.5005981340902681,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.532,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.49317098452331176,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.5321,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.5175994676866063,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.536,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.5188256179444247,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.515,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.5698805240042245,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5309,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.37417613767867497,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.5295,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.5105059527814495,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.536,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.3880400500777969,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.5349,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.5537589896837134,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.5445,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.39320316975030567,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.5364,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.4574616885358706,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.5362,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.48538358712062984,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.5311,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.4655160334631272,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.5426,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.5024530200567083,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.5478,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.4055633722982698,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.5338,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.43757820084404947,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.5382,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.37479635327069416,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.518,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.37088020467044475,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.5314,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.4381064596350898,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5315,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.40533202062499113,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.5167,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.45366471666952896,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.5308,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5069629775005734,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.5205,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.4669688481284078,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.5201,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.5132570044848103,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.5254,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.5349805956007716,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.523,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.5763810926448028,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.5185,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.43896810524628904,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5374,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.5359718934915019,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.5263,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.43284998454124574,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.5336,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.4284596134035596,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.5004,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.3738599047568368,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5339,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.43309627063846295,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.5401,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.38044463548171037,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.5297,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.40238730021866786,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.5353,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.42828839582783473,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.5253,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.44046544097684526,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.5325,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.42860814735709896,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.5335,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.39773814493322357,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.5433,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.4774559418526836,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.525,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.38235387673737453,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.5332,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.44301123817448823,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.539,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.42014544337844356,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.5378,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.43916970066266947,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.5362,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.424069701856985,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.5207,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.433342280737446,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.5338,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.5622850177948286,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.5285,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.45455691347537686,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.5286,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.4981515303265994,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.5233,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.5957958082906348,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.5347,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.49738183677406705,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.5207,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.5020337591217743,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5356,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.5285020578449169,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.5404,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.5355712158903253,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.5225,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.4916107627417785,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.5183,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.57091572097654,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.5332,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.5496125083207997,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.5362,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.5344046194313982,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.5226,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.5678058149165987,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.5337,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.5305732572425254,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.5125,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.6330321213517199,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.544,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.4442865834606439,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.5213,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.4843602090141509,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5234,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.4344637700055957,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.527,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.487323954846566,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.5252,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.4508879140742996,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.5339,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.44578185755864697,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.5237,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.42045049316018546,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.5405,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.4300764040972666,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.5374,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.4078694344229407,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.5299,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.6288043533055311,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.5255,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.4036666800460831,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.5321,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.4378147387686428,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.5268,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.3999399179373766,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.5269,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.42081684292342153,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.5217,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.4775390230825924,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.5215,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.349502756223713,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5181,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.5503754450196612,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.5207,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3301483907273916,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.5221,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.4250469406833546,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.5264,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.35232116354939236,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.524,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.38874191653606105,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.5162,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.38907763717560234,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.5257,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.37102597220576494,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.5126,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.3807323288418751,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5178,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.43025487327960665,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5385,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.371950829209977,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5387,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.4572100196576565,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.5356,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.44619186903565705,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.5292,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.4964747198270556,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.5215,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.41911062423162787,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.5474,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.43967956084266707,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.5317,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.4410398700259025,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5254,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.4753552505546128,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5411,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.4492733276419692,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5104,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.41694989234895474,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.5259,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.4827573884976489,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.5244,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.4252170425544368,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.5214,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.47646844293274276,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.5234,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.36392976995152004,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5202,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.43467242787706545,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.5205,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.3916508392831362,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.5195,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.42842999969010626,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.5257,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.44180591226545846,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.5301,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.4080515921128648,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.5191,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.43132658766186344,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.5383,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.38136174110338256,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.535,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3685220353135599,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5188,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.4982186979084491,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.4624,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.9861433212786074,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.458,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.6644536000818229,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.4566,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.42811787004274554,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.4636,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.4554101502649836,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.4489,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.46541319312797774,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.4494,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.4972177227045675,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.4556,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.4636267909173419,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.4517,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.5242994869595194,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.4584,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.5465802560188056,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.4568,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.45554866101629726,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.4553,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.466802165647535,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.4564,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.4499711787673512,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.4495,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.42812129896799483,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.4482,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.5054170114193537,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.4342,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.3992174697335095,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.4666,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.47408014215450117,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.4467,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.4055655183356069,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.4754,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.4283510245900734,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.4445,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.3948990952042367,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.4392,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.40299197343639015,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.4551,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.4073165107258351,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.4474,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.4062145727392078,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.452,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.3494366412180308,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.4573,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3719051094199161,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.4397,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.3724799755499589,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.4378,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.3587097769442861,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.4591,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.42477134360418983,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.4446,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.3088121959570487,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.4481,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.38374357957265826,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.4463,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.39253774059130625,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.4588,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.3347838734061824,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.4526,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.3739301120776672,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.4548,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.4083062364002318,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.4378,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.37158560531194756,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.4414,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.4245296839296661,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.4576,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.3578365892559112,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.4467,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.44087809320302573,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.44,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3639129835322985,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.4668,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3080685444779784,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.4569,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.3332793779320767,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.4502,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.32707237511184073,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.4564,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.3289372089432307,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.4398,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.36046141760772243,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.4311,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.30247002834906983,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.4428,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.3848384455626526,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.4478,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.32412180386241696,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.4584,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.30419981857646794,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.4567,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3461093634590385,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.4451,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.32536447967607446,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.4477,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.3599926458412535,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.4621,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.36246298389364084,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.4514,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.34876456214102153,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.4474,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.3726645676625901,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.4375,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.3226922764022285,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.4484,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3388598987803239,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.4459,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.31660145838145975,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.4551,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.37059132142942197,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.4495,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.29918162664482534,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.4334,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.37559433887150645,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.4592,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.31833801223046176,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.4525,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.37082768999468396,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.4489,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.35991354304829704,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.447,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.3200763651599732,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.4361,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.35153319454642346,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.4592,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.3434931180745406,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.4669,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.3737514789913785,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.4387,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.39347898282653476,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.4563,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.4079888491550999,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.4528,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.5248177857416259,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.448,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.43781716211396743,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.458,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.5542216489124976,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.4463,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.5890497245459078,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.4446,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.41393781488520837,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.4385,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.5373712629856561,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.4368,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.4046693960769037,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.4507,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.5029052450812233,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.4561,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.44211183566598083,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.4676,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.44178684438461924,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.4426,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.5045317850150866,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.4344,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3774628098424003,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.4595,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.5328271202576679,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.4524,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.4162419454479367,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.4638,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.4262495442315526,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.4588,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.4187837113607268,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.4392,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.4060670013465893,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.458,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.41280248618485743,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.4397,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.33290483169844465,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.446,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.3671788494520942,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.4492,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3686762986008626,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.4493,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.3933176744643983,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.4571,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.41034202988291096,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.4723,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.43946525010153104,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.4557,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.3868979201878358,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.4465,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.3953895245471286,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.4543,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.37024870188670855,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.4528,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.3822527615367403,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.4657,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.3590419805635205,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.4458,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.38851994918932564,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.4592,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.3819688802226215,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.4565,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.37514450897480955,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.4472,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.3144667815786388,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.4655,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.4036779240224185,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.4458,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.3277919073549885,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.4588,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.3413815509261373,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.4388,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.36382284174879403,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.4586,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.32995930826544007,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.4434,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.402716595110856,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.4485,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.30168936059219054,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.451,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.33778350364022297,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.4531,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.32755459937396353,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.455,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.3737878871669524,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.4395,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.3400903793354071,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.4641,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.35523907249568953,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.4399,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.3404351414704405,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.4488,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.3606151360514615,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.4442,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.3486083505109857,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.4533,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.3677574130486176,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.4409,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.3782508835579284,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.4526,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.3910853996477974,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.4498,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.3847718634459492,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.448,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.34705099992562605,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.4516,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.4207356254956071,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.4614,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.3818624104348472,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.4557,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.3879247006760846,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.4509,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.3731385073813017,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.4501,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.3632312711875936,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.4575,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.3861227008426653,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.4382,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.3845848783702631,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.4468,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.3784673931935262,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.4432,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.36598672388188486,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.4394,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.47230308656182024,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.4386,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.31478992757906665,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.442,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.3504079621143133,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.4424,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3199095485910556,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.4505,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.34818258931079393,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.4594,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.3275729578365557,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.4515,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.32783427691691175,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.4516,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.3693498408628378,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.4359,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.33494022271257184,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.4478,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.32923992937580177,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.4585,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.29844991169060286,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.4547,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.3017692772181284,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.4531,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.2842980264177825,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.4426,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.3296616027726146,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.4292,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.3362003826677216,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.4653,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.3320046217386775,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.4553,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.32058481581585474,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.4434,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.36169511292392265,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.4444,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.3114116604208412,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.4469,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.39469519510401446,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.4478,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.3590647205207763,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.4457,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.35784363066752195,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.4334,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.3744962799741945,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.4472,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.35068863193168953,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.4691,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.40632042988249095,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.4448,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.388533140037651,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.4474,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.3906469939457996,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.4537,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.3478315390361503,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.4613,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.39213606353846686,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.4415,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.32385799251484676,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.461,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.32336301888359936,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.434,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.3882604138277141,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.4494,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3464975755175381,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.4283,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.32910280449211654,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.4351,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.34895569057482456,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.4675,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.35819590465536233,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.4602,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.3472030759237662,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.4388,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.33050558393150414,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.4456,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.31758278979518584,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.4373,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.34188653885796144,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.4463,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.3002577366812194,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.4389,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.36995514025485715,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.4517,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.29430172486997375,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.4506,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.3682648953345056,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.4543,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.2778457957820066,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.4416,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.3019467645743983,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.4534,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.28410452155772653,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.4404,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.324574641653875,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.4543,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.2823564622737898,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.4665,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.322307942446871,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.4306,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.28958934714067175,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.4448,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.311499333030071,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.4397,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.3278733026386993,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.4488,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.3118842461780062,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.4662,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.30773603996361754,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.442,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.3108827689557035,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.4443,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.34810573929718247,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.4671,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.29533936406644345,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.4669,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.3016464647143477,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.4635,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.2937320630942315,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.4427,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3015208732371542,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.4472,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.29799963597694523,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.4279,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.2828228144842739,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.4531,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.27967328127821894,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.451,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.3210128997300531,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.4542,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.32318422532124286,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.4476,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.31831682193213795,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.4477,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.3082278369835042,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.4488,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.3177475461232894,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.4535,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.29659044351329256,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.4575,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.28846856793105335,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.4465,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.3015717507851192,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.4465,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3116142647304791,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.4589,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.2913389117416329,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.4572,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.3408146492618761,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.4334,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.2769633842192604,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.4373,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3479100165849521,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.4463,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.2976987685687192,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.4137,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.3128718413008926,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.4383,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.3198502946665914,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.455,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.3128862923948384,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.436,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.2969208602921343,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.4464,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.2717048752595772,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.455,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.3096871224996096,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.4474,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.29976007204515903,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.4421,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.27807375967530634,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.4428,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.31493220236857017,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.4478,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.308826110790317,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.4577,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.326571847521404,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.4488,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.3466672360363286,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.4536,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.31766928471726974,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.4582,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.286278597186735,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.4321,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.3577946547906509,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.4483,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.28687741326509975,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.4311,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.32578515850543155,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.4541,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.3183123442475607,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.4358,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.3317488725587882,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.4632,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.3149755066449571,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.4433,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.3289853375228605,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.416,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.3382073233101809,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.4451,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.32312370827253317,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.4432,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.328868392310754,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.4402,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.323617086558151,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.4339,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.3664517844625413,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.4431,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.3395919380784735,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.4296,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.3581669723193888,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.4283,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.3343923802574171,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.4439,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.36660728941651133,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.4275,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.29738989895346324,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.4372,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.36500377001650675,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.4468,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.29280531382720576,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.4419,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.35522445592449325,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.4452,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.32938804623180645,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.4443,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.31153846932700185,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.4569,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.3123380967458237,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.4345,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3146772894871681,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.4366,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.5600511296025902,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.3839,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.710335192338159,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.3651,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.42654978618165185,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.3715,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.4190661089888699,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.3578,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.39071690311045454,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.3626,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.3876080010089845,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.3816,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.34670635999786886,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.3633,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.37894314479343033,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.3621,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.3504457177644532,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.3596,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.39891232066728843,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.3846,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.30928947279649904,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.3584,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.36223742067447573,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.3693,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.30657684520769934,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.3755,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.3537644050738213,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.3578,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.3594484352392796,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.3666,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.32600942392902077,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.3574,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.3126599390451087,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.3603,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.311628980743899,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.353,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.3088107243539632,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.3726,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.3165678366151193,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.3706,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.3151511384244397,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.3689,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.29192016646461233,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.3545,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.31095926930412343,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.3635,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.33693312437438155,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.3767,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.3104005821883992,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.353,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.308826892751975,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.3731,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.308592217236878,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.368,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.28921252814912735,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.3645,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.33143970788460325,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.3703,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.30123422491657664,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.3591,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.3183875307985971,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.3603,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.2951499123622415,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.3603,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.3111449911792704,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.3714,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.2937302374890542,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.3543,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.32491362876549573,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.3517,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.296673818356201,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.3686,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.3320048072370048,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.3751,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.29538467059777734,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.3687,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.32256885535137936,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.3522,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.31909904319658494,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.3755,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.29447945829516314,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.3687,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.3235284229974414,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.362,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.28546452600525446,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.3697,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.32692885525590626,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.3286,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.2935148063954288,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.3665,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.31607695498180477,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.3682,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.31337903374007625,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.36,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.3204142897650025,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.3448,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.3065550644343511,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.3515,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.3063997185812491,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.3507,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.31797717224831573,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.3598,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.2926275130623967,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.3621,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.3197077340398651,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.3664,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.30329720365106533,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.357,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.2948978614876748,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.3549,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.30096978411827985,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.3621,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.29819556654505175,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.3517,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.31639394337126814,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.3701,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.3190771389239022,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.3623,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.33223839057071736,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.3645,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.33662233650508444,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.3548,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.3294422196579773,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.3694,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.32541450425109214,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.354,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.30098087133352636,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.3671,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.343581921385378,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.3522,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2699273121551218,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.3713,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2775489951309005,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.3708,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.27041456107886724,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.3716,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.2749238325046797,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.3617,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.29015175692639744,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.3588,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.30960843424526857,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.3778,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.29085906294363245,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.3624,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.306567075210477,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.3598,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.3087166990068477,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.3696,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.2902030739666147,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.3642,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.32581590890904194,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.3727,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.32127500334518516,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.374,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.3265878427776116,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.3522,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.3526783695108242,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.3371,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.3171362979461586,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.3667,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.3342603415755846,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.3731,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.3159178610686334,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.3695,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.30536679083084695,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.3833,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.3565286936706188,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.346,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.3184540276334274,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.3694,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.39600423303417415,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.3689,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.32485298303352494,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.3645,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.33342714599659445,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.3699,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.33237407017141196,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.3672,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.29400755883103624,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.3462,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.37857507506258353,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.3779,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.2902439687100761,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.3689,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.2969263010673951,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.3674,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.308636019724935,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.3638,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.3040383864918713,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.3625,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.3346014792660967,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.3817,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.2944236984358365,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.3601,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.3249473430719637,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.3672,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.3235783337523676,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.3738,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.2746304704131087,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.3659,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.32458086604937203,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.3627,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.31123870103723617,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.3673,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.29335966294128973,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.3701,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.3411036252647646,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.3758,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.29653226752615985,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.3465,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.3035252131637916,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.3624,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.3434369432390818,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.3656,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.292689640552362,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.3624,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.29121253680614057,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.3597,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.30790272443414857,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.3739,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.2948124286290516,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.3599,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.3050222676071881,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.3456,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.3505186661974429,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.3685,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.2817302116517245,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.3746,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.3092413238634837,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.3579,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.29120071424306393,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.3534,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.34474665186275016,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.3713,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.2749173338383228,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.3728,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.29699173737094064,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.3772,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.3013812829319007,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.3637,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.27068244944372444,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.3739,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2671412722027941,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.3703,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.2984772936028287,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.351,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.3255599392464346,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.3615,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.3008591776978645,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.3506,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.3168680622401215,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.3718,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.2696094775448585,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.3637,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.2921267899177677,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.3442,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.31372801168901193,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.3517,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.30527497374562723,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.3566,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.29647020092346615,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.3508,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.2816906753175448,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.3747,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.28279782715713164,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.3755,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.28775860875305553,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.367,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2865352954090003,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.3587,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 1.0080115746074838,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.3538,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.2691566708232681,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.3626,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.2752267230080385,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.3762,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.26814014125459745,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.3635,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.2763454191275235,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.3742,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.28022151114869454,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.3738,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.3035865285505278,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.3705,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.27820013429721585,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.3721,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.275282359742795,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.3739,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.27056547426192384,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.353,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.294921764492366,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.3493,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.3149860821423364,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.3526,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.2580076578503863,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.3532,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.33106538716348216,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.3544,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2856338774591864,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.3469,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.30876534748952433,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.3768,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.32771463280733903,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.3727,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.2674215865703401,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.3715,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.3045133464748243,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.3757,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.2984758902297245,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.3585,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.26438419262365914,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.3669,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.3301835043036692,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.3393,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.28796542601437264,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.366,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.28109967713487116,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.3418,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.27398746079010117,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.3574,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2684216727424452,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.3676,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.27636100673135344,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.3637,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.27433900281148693,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.3741,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.3156639536267058,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.3681,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.27916993794567674,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.3467,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.26814004499025146,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.3633,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.26403816747925923,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.3872,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.2900229662793025,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.3437,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.26018165619587613,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.3646,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.38924624632804117,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.3627,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.2791010991451726,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.3604,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.27449232590678596,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.3615,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.2664899667296481,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.3619,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.2597596514079537,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.3603,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.26258433803024894,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.3437,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.27298976084521764,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.3602,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.2739462199208159,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.3677,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.2630093220426507,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.3696,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.2663910543423427,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.3615,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.2816784338279642,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.3421,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.2545017435629262,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.3514,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.3010407235726683,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.3478,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.26347737209966754,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.3539,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.24933659058649213,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.3541,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.257181117565556,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.3623,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.2686740818040298,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.3613,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.2551318618288475,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.3459,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.2835026985365782,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.3697,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.2735288461239391,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.3317,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.26778554282267103,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.3492,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.2835024396599578,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.3748,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.26406269599748616,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.3646,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.2710120504704801,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.3426,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.32877521056795916,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.3398,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.28362709250377344,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.3496,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.26834931355954206,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.3434,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.30356958225102326,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.363,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.2915363597536052,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.3607,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.3253574831654109,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.3513,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.28087804477782075,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.3467,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.3106058654739635,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.372,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2722578862158183,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.3679,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.28155952747504037,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.3618,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.2756124728742785,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.3484,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.2743586758637543,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.3609,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.26458952414494874,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.3586,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.2720527152117719,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.3635,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.25825134658086507,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.3818,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.2735942462576192,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.352,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.2809752370045188,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.3732,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.25642029998157134,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.3662,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.2584345796352508,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.3673,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.2825315073245843,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.3576,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2590178441549067,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.3712,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.24225384082397503,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.3729,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.2756207084681472,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.3552,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.2639882502147775,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.3598,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.26186734959068453,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.3478,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.2828682368602177,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.3712,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.28031498599534893,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.3716,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.23475095227662815,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.346,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.2578851107103661,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.3662,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.277676976018707,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.3713,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.2636515366557519,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.3706,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.2537175891034692,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.3682,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.253840126070526,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.3292,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.2624477468123628,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.3565,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.2603112426203265,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.3531,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.28955075269244057,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.3435,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.24735980374034666,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.373,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.26077572284529915,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.3654,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.2719255085044422,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.3467,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.26357178557612526,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.3478,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.25875616981976196,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.3482,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.26934073160314914,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.3617,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.2722459488273823,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.3527,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2608320520229782,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.3254,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.2631718530154652,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.3604,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.2684937723945436,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.3666,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.28868892626871423,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.3923,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.26471830896817117,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.3578,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.2507807454992281,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.3403,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2618973928050619,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.3473,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.2541623182139187,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.3502,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.25701223954197033,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.3556,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.2741696299705811,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.3527,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2806694872140955,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.3454,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.47839328425562405,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.3162,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.3880545293969393,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.3099,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.3033651826440316,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.2951,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.5447002565514975,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.3008,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.3674302847309182,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.3009,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.3742943814447286,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.2948,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.4117501002081086,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.2955,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.3639733476037077,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.2746,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.3075504347426322,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.303,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.3475578508347755,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.3017,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.3196472434326312,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.2922,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.2859910803578143,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.2909,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.28655762566007575,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.3042,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.2836696146219331,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.3121,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.2660857150839966,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.3012,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.2606868208349283,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.3067,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.2765479657069638,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.2942,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.28036078627448097,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.291,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.27497756566015613,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.3072,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.29189955871692447,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.2967,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.2722036371006832,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.315,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.28038958648749895,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.313,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.3028352128816358,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.3177,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.2750508600564734,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.2895,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.2744007756293367,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.2988,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.3096237805907974,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.3232,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.270263143707075,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.2819,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.26583234421252216,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.3063,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.2640072970921033,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.2888,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.25728389313183936,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.3043,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.25843422906749425,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.3002,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.25599222262652604,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.3153,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.2576038023993918,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.304,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.2685028581314331,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.3039,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.24944421590053933,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.3143,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.2536007423366579,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.3055,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.25065534365613523,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.3213,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.24775813274021197,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.3057,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.24806142977112136,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.32,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.25533442536334555,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.2902,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.26057395143974715,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.3029,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.26565038918264855,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.2745,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.2546382411322198,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.3023,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.2517414923463102,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.2898,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.2571817636372127,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.3034,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.2520825083950354,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.2986,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.2519415156029847,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.3109,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.2610781551971523,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.2964,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.2515679448368982,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.288,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.25533271153094894,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.2958,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.260864266227643,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.2859,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.2857899375562561,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.2984,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.25933611888895863,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.3035,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.24849912050206996,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.3037,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.2615961912018796,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.2984,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.2554100336143482,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.2753,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.2535048165244602,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.2912,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.25353564772544984,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.3182,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.25275573398801754,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.2991,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.2513088616087457,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.2923,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.25474033265518287,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.3047,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.25180323402166066,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.3056,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.2535884659157945,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.299,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.2531430772859713,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.3006,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.25513400634091976,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.3114,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.25813372919050087,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.3031,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.2482883994611749,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.3185,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.24793974004071553,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.2924,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.26436665911153473,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.3034,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2407721936870262,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.3074,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.25280623051951256,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.295,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.2508260656844987,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.3056,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.2466119638692335,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.2955,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.23890402317529436,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.2917,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.2573537336512045,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.2922,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.25496939239183275,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.2813,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.2550349165292775,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.3065,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.2616220764478925,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.2865,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.2608858708290238,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.3012,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.25492307865671565,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.2964,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.2512703539095177,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.3032,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.24610208139182277,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.2855,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.24354003883415032,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.3028,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.27566299409664097,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.288,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.24808978011048666,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.3036,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.2559798360800705,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.3089,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.26379208021050876,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.281,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.23763959719927882,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.2984,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.24553760618950973,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.2994,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.27849227715519037,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.2895,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.2614213366566067,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.2842,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.24776455184100965,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.2839,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.26557955464135047,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.292,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.25758809537316296,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.3066,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.25860537718159654,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.2955,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.24252406201588442,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.295,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.2524423926042645,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.3131,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.25382745086249753,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.296,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.2678395128174614,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.2766,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.26177957354092907,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.3127,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.24893939637316917,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.3105,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.2535658989074071,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.3153,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.25188974871243885,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.2799,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.2520884608666502,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.2804,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.2431735995716966,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.3151,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.24847989714821409,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.3015,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.25040362888047746,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.3177,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.24266379568446328,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.3093,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.27186984759774635,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.2928,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.27059640688032827,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.2996,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.2559709791382835,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.3151,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.2524494971718647,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.2822,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.2541319334245632,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.304,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.250693370265044,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.3128,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.24576574775848511,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.2928,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.24757501337262966,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.334,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.2447981532675195,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.2865,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.2459491984810163,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.3016,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.26092106681087296,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.3159,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.23851931139369123,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.2899,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.262821873200895,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.3081,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.24399581012004817,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.2869,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.24099340519259257,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.2888,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.2373794799973832,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.3051,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.2538163195932624,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.3014,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.25859157571934277,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.2827,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.26659649463254265,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.292,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.22864352759753392,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.2899,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.252822400956289,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.294,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.25137973548860243,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.3091,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.24465651488890588,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.305,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.2564732421253968,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.2862,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.23520605365610234,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.2979,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2561221652625103,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.3065,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.25342935502817365,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.2888,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.2591711996505325,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.3085,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.24473935425707616,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.3057,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.23717107037468957,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.3116,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.24389037086299645,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.3003,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.26819320316197415,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.3196,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.24083416251782466,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.2969,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.2614929361488406,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.2832,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.26753785070081476,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.2974,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.2396720839619487,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.2976,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.24788840685113156,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.2944,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.23439443745326194,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.2778,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.2499644127934523,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.315,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.24355862469977693,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.2988,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.2387104951404753,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.2859,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.2535547006668319,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.2858,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.22815621185018464,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.306,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.24337934754709828,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.2846,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.27085189677239113,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.3033,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.2431194180629345,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.2812,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.23743499794484,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.2976,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.2519173561032551,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.2955,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.2435006237092135,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.2924,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.24901182137800212,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.2935,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.24236168459566848,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.2915,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.23431829105339577,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.3073,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.23892913678971228,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.2916,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.25699505635329134,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.2897,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.24118784784632216,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.3074,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.2340043899232646,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.2955,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.23144787151355198,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.3014,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.24464261421005326,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.2833,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.22912547868789665,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.3072,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.23705262168948185,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.3122,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.25616252880766877,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.3005,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.24884221110619392,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.2828,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.23779236290674013,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.298,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.25051589931873963,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.3207,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.24698233032812747,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.298,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.2427629571332196,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.3012,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.25586410193673126,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.2772,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.24913134500546097,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.2939,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.25387269387492384,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.3048,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.24841415594115615,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.2967,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.24608292125624254,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.2974,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.23184182253081334,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.2851,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.2534988271623732,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.287,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.2508920747367489,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.2988,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.24746034899015176,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.2825,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.23831343020979204,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.2787,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.24445010758066302,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.2963,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.23809149656636547,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.2991,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.22744460182277595,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.2876,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.24026537494653835,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.3119,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.23832098805858157,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.2956,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.24582714445002435,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.2902,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.2534709066359969,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.291,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.2452872195167858,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.3045,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.2497449213484799,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.3018,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.23759725702765996,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.3179,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.24199580522519737,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.2992,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.23735638326482628,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.2866,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.23865382119474166,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.3004,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.26057939434806604,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.2911,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.2366363547206248,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.313,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.25655595413557947,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.3076,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.23141492273863767,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.3087,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.22696530156779204,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.3009,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.23435880532386302,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.3056,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.23335022620230295,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.2971,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.2452895190393469,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.2859,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.24303548653516355,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.3062,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.252382768235967,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.3096,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.2335668425723605,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.3005,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.2422821400935202,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.3144,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.2356171129356005,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.3067,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.24659596810428785,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.2883,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.25617015018246053,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.288,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.24001563339988383,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.2716,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.24085196696635658,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.2891,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.2334702035132024,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.2795,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.24084261481919528,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.296,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.24762703141811054,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.2887,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.23198308509059382,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.2899,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.23409800851970095,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.3032,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.23867144498431223,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.2839,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2379448075221363,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.3046,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.26577754204157106,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.2724,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.24345992811667194,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.2948,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.24167512798273394,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.2757,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.24244116280025277,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.3019,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.23348674465914904,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.3134,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.22942736062801203,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.3132,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.24874351178779303,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.302,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.25374264344014974,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.2985,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.23026192150148087,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.2897,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.23680833632526638,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.2951,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.24785677608327464,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.2904,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.2421467360362013,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.2978,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.2499196011953196,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.288,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.23746407871161881,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.2888,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.23038902635306335,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.2925,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.22951482255730452,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.2975,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.23064635110834136,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.2962,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.25056686463332145,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.2954,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.23624467525946638,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.3105,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.24425894133377332,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.3045,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.23603057374936862,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.3079,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.2427924061113795,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.3087,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.24859364831517305,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.2906,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.24256469124693425,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.3084,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.24163179572884658,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.29,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.23119069006684073,
+      "learning_rate": 0.0,
+      "loss": 0.2788,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 3167726939537408.0,
+      "train_loss": 0.46034835494964227,
+      "train_runtime": 30235.0271,
+      "train_samples_per_second": 5.226,
+      "train_steps_per_second": 0.041
+    }
+  ],
+  "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": 3167726939537408.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}