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": 6.602364870852425,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.0587,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.536384165230049,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.045,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.4947516460688455,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.0476,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.390054407460938,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.0449,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.082337913880124,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.0378,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.959767522948545,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.028,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.5851068039771485,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.9852,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.387275864183347,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.0077,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.8302315025170257,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.9515,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.614856500182006,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.9552,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.4966337143151454,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.964,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.7834673381978856,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.9293,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.063157259774333,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.9393,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.9184055870708483,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.9379,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.545335050829733,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.9,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.7542311306091336,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.8659,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.423313239005812,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.8542,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.0752924287453274,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.8622,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.4887961931808453,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8325,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.1797077279110286,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.8102,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.0395577346551437,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8056,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.2225476260633292,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.768,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.1022384495169002,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.7829,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.003828637904622,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.7756,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.8651062889261786,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.7724,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.9576771029029284,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.7575,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.8903040236749123,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.7459,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.8083751245506642,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.7404,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.726506450942672,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.7494,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.741865355108755,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.7243,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.745687093902222,
+      "learning_rate": 1e-05,
+      "loss": 0.7445,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.6473512427817195,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.7205,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.6190965716747676,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.7166,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.643621684754558,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7053,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.6268347887396677,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7277,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.655293207507944,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.7069,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.536860773602323,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.7227,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.7917800711954374,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.7291,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.5889099805481052,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.7066,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.5360324647782755,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.7129,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.5673297575298154,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.7177,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.5592205426927226,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.6988,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.5620797966260039,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.6926,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.5786169280698802,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.7011,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.5670327984011351,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.6982,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5629073326780953,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.7085,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.6294965696587225,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.6828,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.6371771008868942,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.6877,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.5955750429248199,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.6819,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.496435155527878,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.6916,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.5944274559010211,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.6833,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.4850256657669469,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.6872,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.4926142468325687,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.7065,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 1.332956239523453,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.7247,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.6161221980583526,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.6802,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.6228077929794128,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.6812,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6029751925083219,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.6715,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.6845409649560567,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.6776,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.69895810858536,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.6658,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.5270582519789423,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.6828,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.6377468550017252,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.6783,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.6541082097077358,
+      "learning_rate": 2e-05,
+      "loss": 0.6644,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.6386042785304631,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.6655,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.6414465002829899,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.6554,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.7959255857346073,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.6532,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.5828705829364301,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.6638,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.5731536510909668,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.6666,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.5587551828179012,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.6558,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.748959445969651,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.672,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.6329139835738089,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.6536,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.7577160437753068,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.6472,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.7700034162728409,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.6631,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.6318088029820649,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.6517,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7957172850233919,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.6639,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.6702551014994682,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.6403,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.8256640436325909,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.6593,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.220881222743266,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.661,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.0203944469891377,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.6611,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.8732654735161509,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.6486,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.1104640078224268,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.6433,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.2182610751709637,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.6404,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.9366839942949482,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.6526,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.2669478551318762,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.6688,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.8514961273533114,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.6583,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.066588576812034,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.6682,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.645504524954639,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.6483,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.7848965096623411,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.648,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.7509731827016162,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.6584,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.8403826841800346,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.6513,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.812060300364977,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.6426,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.238456023828352,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.658,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.642476966469415,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.6554,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.6264945099552897,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6639,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.2261937300191834,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.645,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.3642097757027598,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.6411,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.2316935222576657,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.6471,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.2465449706356853,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.647,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.2569020743110573,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.6545,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.0398861392010297,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.6451,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.1914493725415987,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.633,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.0456809255725787,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.6538,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.3883142361116456,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.6395,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.8602930997211914,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.6471,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.1773283640202026,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.6449,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.8980003363821147,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.6498,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.8491908020397604,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.6386,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.039870056170339,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.6569,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.7355308068460247,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.6406,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.1549361732951913,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.6446,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.4165303957464856,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.6499,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.00950015403687,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.6339,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.9494369188945065,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.6456,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.2934142512262803,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.6326,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.0163751518078967,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.6411,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.2465800691169762,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.6413,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.1984430210975985,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.6525,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.9601150300846938,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.6345,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.2322691372640096,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.6464,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.1824824293897949,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.6294,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.9434092324311617,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.6361,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.9111331877383408,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.6328,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.577716715686919,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.6349,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.0628304435535572,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.6439,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 1.500198510277317,
+      "learning_rate": 4e-05,
+      "loss": 0.644,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.3501201837050818,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.6359,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.9884592721082537,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.6325,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 1.1253601217104636,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.6298,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.1526716503174121,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.6212,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.7813079637234989,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.6479,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.369614420765309,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.6451,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.715907928764769,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.6354,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.1666938745479,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.652,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.8394296070725411,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.627,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 1.0910856968663916,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.6322,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.1142011528395257,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.6317,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.6461732647756876,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.641,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.1405274002971155,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.6309,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.9544343606509988,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.6333,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.8608239813412496,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.6256,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.794619540998437,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.6333,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.6337295036737729,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.6326,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.8551155636850617,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.6407,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.8384416852105897,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.6394,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.8030286371295386,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.6246,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.9560845853893204,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.6144,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.9468994627333496,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.632,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 1.206041359435266,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.6167,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.9298504460257946,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.6325,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.716689211327568,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.6339,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.6782069537883003,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.6417,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.8781730891361822,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.6337,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.8102129579125712,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.6269,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.7123632942235609,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.6227,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.6622422530806926,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.6219,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.5767750559685224,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.6397,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.6453193969143344,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.6311,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.7697562230982612,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.6382,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.7333779567818925,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.6298,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.6010093192107744,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.6232,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.56638799267087,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.6388,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.5835379454587558,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.6304,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.5564091597904623,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.6174,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.6555421604752352,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.6245,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.6863957996837259,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.627,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.8829749833474497,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.6202,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.8028799825415299,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.6315,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.6390887668331995,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.6211,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.39087004934753794,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.6274,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.6852441677466784,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.629,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.8802054364687365,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.6282,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.9882663172662245,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.605,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.8617577613664832,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.6391,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.6425373802085277,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.6415,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.5484190969517374,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.6187,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.7917433064203937,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.6271,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.8485000870621956,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.6235,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.5822618546039361,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.6215,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6680323325472396,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.6336,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.9850930102021289,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.6157,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.9248522576401909,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.6162,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.6316485951300459,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.6203,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.7863257482979117,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.6306,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 1.0432060716664224,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.6165,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.671247810032658,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.6191,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.7369625888234924,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.6273,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.8617580650739206,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.6186,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.7455252043969549,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.6335,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.5306082155460724,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.6404,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.7200667106356771,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.6328,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.6346726832401974,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.6224,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.5852012322584065,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.627,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.7104466126645028,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.6235,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.5105198111371139,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.6282,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.5932713576440197,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.6271,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.5581245244828591,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.6239,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.546898506450062,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.6184,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.5461183651334127,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.6251,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.5006426756235877,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.6351,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.5501084928564289,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.6176,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.587441576331411,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.5977,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.6036918361476123,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.618,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.49562306882079393,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.6216,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.6773301956209086,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.6141,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.5611813302379542,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.627,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.6358921747589213,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.622,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.6638032013962079,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.6205,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.6128290108474879,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.601,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.7811721903469547,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.616,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.7322893722476004,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.6175,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.5834539075762069,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.6164,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.5795701465457385,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.6105,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.5664245236874806,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.6185,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.5210480350979448,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.6154,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.5798925725898019,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.6324,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.6020953913038037,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.6234,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.6160665653034046,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.6289,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.4828248102615635,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.6143,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5055856167456388,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.6121,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.4331174802126953,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.6135,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.6078252121166356,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.6064,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.5667380855490277,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.6088,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.5694843706722886,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.6255,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.5598083004066285,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.6091,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.656498956471545,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.6158,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.7153968021750797,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.596,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.5514114387505116,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.6326,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.690009919854715,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.6096,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6917526709156558,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.6233,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.7488083480880349,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.612,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.5514322208131387,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.6041,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.47009285688951,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.6018,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.5408322998883931,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.6017,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5638591179735352,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.5922,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.46690004350278147,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.6264,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.41398582400587486,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.6033,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.4529824659628971,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.6006,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.4189872114865156,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.6021,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.5610452777073683,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.6164,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.46522522108500675,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.6116,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.5182413962398081,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.6073,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.42992641081792726,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.6047,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.46204487538947075,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.6103,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.4649863182465308,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.6063,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.44155318161715396,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.6058,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.5163427339631887,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.6027,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.43456455056941456,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.6107,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49073483206107626,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.6011,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.5220758653912213,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.5679,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.5404840352958442,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.5409,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.5252873229183734,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.5646,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.6187384038667566,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.5489,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.5394830536890309,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.5685,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.48532121655469374,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.5443,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.5907850677779716,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.5501,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.5475104268927631,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.5419,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.565618175910083,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.5578,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.5562777128961544,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.5598,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.5066526639874164,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.5515,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.6494770559706674,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.5705,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.6144327769519504,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5263,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.5266875893784679,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.5576,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.5994876824717766,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.5388,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.5058840618658207,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.5553,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.5603626548545543,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5439,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.5841798579494322,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.5461,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.43768377420068205,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.543,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.6646730849155592,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.5395,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.5047877324834139,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5524,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.5520787025144108,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.5436,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.6928949004932072,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5619,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.5676660615089524,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.556,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.49713409801200964,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.5662,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.6933851587834108,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.5327,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.7000544932789022,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5589,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.5269922611945771,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5595,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.5657649514526565,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.5278,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.5760362591104088,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.543,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.6554806782797181,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5513,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.4495720074810593,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.5429,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.5660074933295147,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5516,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.5465137167457197,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.5499,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.5165007577623318,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.5557,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.5600064633495889,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.5678,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.5232726065852432,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.5608,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.5131367219194095,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5395,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.3855284111756315,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.5448,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.4922879445410444,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.5443,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.6526993855548081,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5483,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.5233727160864214,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5443,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.5377122142884363,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5513,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.614640637351339,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.5462,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 1.2279758968499703,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.5994,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.5585835543858766,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5552,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5356676987200341,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.5553,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.5946685887523425,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5599,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.4348113018870701,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.5598,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.4920678258590714,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.5714,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.44503125465153515,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.5336,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.48824684654043493,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.5523,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.4522766196030256,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5818,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.47552087978639446,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5464,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.4141021287001192,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.5644,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.42862912969838024,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.5425,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.40490942863311735,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.5533,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.45860337273620383,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.553,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.44399375727303775,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.5535,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.49856092233810523,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.5498,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.37150211834620606,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.5464,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.42572376704554127,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.5319,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.43241400599474034,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.5543,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.43974626494500446,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5533,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.486855880952978,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5484,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.4726814512051882,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.55,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.4652721427948761,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5395,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4265440951099493,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.5448,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.5997497862788008,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.5443,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.48902091803335934,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.5552,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.601614184036123,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.5265,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.42764679128308325,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5508,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5218935532527201,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.5406,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.55211338156659,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5649,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.6110009729071909,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.5644,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.5268660491571779,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.5512,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.47546258857212975,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.5432,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.7449377737376969,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.5766,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.6747232299341241,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.548,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.5282968890333397,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5541,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.46984121388330113,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.559,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.5743120184187319,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.5454,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.5411521886757791,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.5497,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.4919861157748932,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5645,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.5801014355900201,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.5437,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.4808886126711262,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.5516,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.5144450236265299,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.553,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.44846854257948104,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.5253,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.5389909975500007,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.5488,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.5076456664426007,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.5514,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.4397799557648089,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5543,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.4873771019144693,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.5385,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.45715145112520383,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.5501,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.43058012538341967,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.5348,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.41286198136086627,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.5601,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.4979666024646852,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.5472,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.5614032806528886,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.5597,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.49637747848254954,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.5499,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.5090573340461543,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5528,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.5164492790388148,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.5456,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.5980087233541166,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.5326,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.4635569093250687,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.547,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.5552096074941648,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.5691,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.5271632550175097,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.5492,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.4483064257346407,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5538,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.5037129919946178,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5597,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.41028230151782125,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.5213,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.5106382143391129,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.547,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.49033406537346796,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.5455,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.50740529612196,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5358,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.4301373444103061,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.5393,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.38421752777663876,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5357,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.3616919377829653,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.5417,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.42559720272464613,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.5555,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.4208787424973646,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5338,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.4870324513387533,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5528,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.43470230426292833,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.5466,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.4113935638342077,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.548,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.45504223116100273,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.5398,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.5608776108481133,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5462,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.4549613338750727,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.5531,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.4838504345132933,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5494,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.439520570625894,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5536,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.44059562975316696,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.5629,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.396970974546808,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.5597,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.46431993447618436,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.5379,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.4155238922873732,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.5376,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.4148213296760514,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.5401,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.3756846345749801,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.5472,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.4064662641213973,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5407,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.41409406592064507,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.5513,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.4345432454529761,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.5565,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.40423192126553936,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.5477,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.3963988069913631,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.5391,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.4384485645868657,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5458,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.5265961250032971,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.5367,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.45503129088574196,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5441,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.37516429683628494,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.5481,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.5152865509400678,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5458,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.41439128795427127,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5392,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.47683362939429735,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.5496,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.45598052767702935,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.5479,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.4573557197354449,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.552,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.47712580439964547,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.5412,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.4738218514675672,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5365,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.4666741475658191,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.5556,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.4085358273944987,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.5331,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.5586876221108085,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.5486,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.42770114056683156,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.5581,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.5190188172530634,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.5458,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.4600800473132533,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.541,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.6215995625442328,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.5382,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.45179611232417244,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.5437,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.5914541965583552,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.5405,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.5272695045599235,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.5423,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.4480890998480868,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.5524,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.551701512638712,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.5484,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.4483193713209524,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.5623,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.41867412290331923,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5452,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.5080450711607515,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.5324,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.38552141543044954,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.5237,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5223946562727823,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.5517,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.4518795611112794,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.5614,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.40210706631295917,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.5433,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.45983441198494635,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.5471,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.4466829709611504,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.5529,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.3943823128529213,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5416,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.47244011900165,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.5394,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.41806434535897063,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.5472,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.37471023440445206,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.5492,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.4010233793584313,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5464,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.41037038321967223,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.5617,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.3820091277945866,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.54,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.4158070631622863,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.5416,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.3795465136969974,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.549,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.3771115083381269,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.5349,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.45247381324617575,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.5206,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.41785850633546634,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.5446,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.47939374647792493,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.5531,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.5036716245009042,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.5323,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.372468814637545,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.5454,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.47100709187114076,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.5381,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.421752843601229,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.5359,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.5618047583672333,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.5507,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.45510354892565735,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.5445,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.40335323080400043,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.5513,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.4739847368748407,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.5429,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.44123279237603025,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.5388,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.47801537274411793,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.5463,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.4688924751864048,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.5388,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.42964551227497294,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5402,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.5652930899334484,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.5433,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.3995989550425198,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.5487,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.5117580877744605,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.5458,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.42087284616065146,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.5406,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.48032722719194654,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.5395,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.3636362366786614,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.5303,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.42257067879962007,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.547,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.3910077107863385,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.5534,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.4301244685506251,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.568,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.3924060288005666,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.5651,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.3746916988317762,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5397,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.40196774146788805,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.5164,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.39670034143976185,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.5485,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.462955856405395,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.552,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.4777721350626486,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.5587,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.4691132311079346,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.5571,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.43133550872303794,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.5557,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.4699276685215045,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.5467,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.3861532525863248,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.5354,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.43534340794760834,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.5517,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.4580032243581484,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.545,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.3851686731668547,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.5292,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.38785060041869973,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.5468,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.47404805203494155,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.5488,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.32552705268329335,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5352,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.4332761603604135,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.5504,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.5020831133239243,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.5393,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.37157744945921867,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.5411,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.4429503394733022,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.5359,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.38276963258428526,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.5487,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.37556025193393716,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.5372,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.4516041150852407,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.5366,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.4216250878208538,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5427,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.36361745230284276,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5299,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.47347225128443,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5376,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.5770105794800977,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.5347,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.34919713588369455,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.5524,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.5639796436930649,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.5377,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.4755391903096046,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.5532,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.526948264278401,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.5651,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.4708790328280815,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5389,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.5728814372954077,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5452,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.37436326696883554,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5406,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.6072795020475927,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.5203,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.3992747926480174,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.5507,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.5496205641006531,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.5496,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.40892550645319725,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.5358,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.4251516990635762,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5291,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.46483469191921845,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.5125,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.4627211237125352,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.5426,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.40889609292514245,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.5524,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.4358969871342766,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.5261,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.41120328177059273,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.5536,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.41861923898496595,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.5567,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.4423884475514054,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.5461,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3399491593160345,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5425,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.5806038964864362,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.4626,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 1.2050644087342965,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.482,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.6954169829122302,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.4728,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.5086479867267842,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.4562,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.5754953853247983,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.4712,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.4253261074202401,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.4625,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 4.375258799968719,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.4643,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.6700005815724194,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.4647,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.6951139531397786,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.475,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.5001288304177843,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.4511,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.5897164329834571,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.4914,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.5915736309243619,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.464,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.49529603725009164,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.4765,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.4986664498725745,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.4759,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.39507512704560976,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.4622,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.44686519319073387,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.4532,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.44941592836619537,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.4592,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.3843623499194085,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.4626,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.4086170871884704,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.4604,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.38325240167849633,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.4583,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.37828499865454157,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.4644,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.3873285458239388,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.468,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.369721179843046,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.4807,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.41480405230642114,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.4757,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.376634230481478,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.4745,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.4464083737322092,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.4665,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.368377846076216,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.457,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.4330372050027377,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.4574,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.38159980206012273,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.4758,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.4118020974025461,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.4676,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.39623631526541475,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.4827,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.39707426700887083,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.4719,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.42510117364784256,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.4441,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.3751511732455897,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.4492,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3704811619801165,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.4603,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.3764111130295978,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.4613,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.3583361331131131,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.4668,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.35014052431002096,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.4729,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3684403734332735,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.4656,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3620331387998109,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.4801,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.36285570594933375,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.4657,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.36066629116294446,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.4664,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.3316964126703964,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.4661,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.32205132301888617,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.4525,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.3420392808712483,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.4732,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.33335193709760413,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.4624,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3534388953572807,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.4546,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.3615945006516054,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.464,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3496313110147198,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.4617,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.38368545009135485,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.4623,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.325387281886099,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.472,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.37536100014932144,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.4607,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.39520144353287845,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.4608,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.3319095417614279,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.4525,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.3509350587835275,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.4687,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3284431221878698,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.4581,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.3950993050212088,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.4652,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.39718416267345746,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.4638,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.37315382923327667,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.4631,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.38781415143967074,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.4862,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.41234978453050425,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.4741,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.38438519032962376,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.462,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.41929975864957914,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.4844,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.3843606144406984,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.4451,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.4094940940624711,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.4548,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.47426671763897305,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.4762,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.3416933472884306,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.4492,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.4385977028007199,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.4766,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.31716969749012713,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.4591,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.40324155011639407,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.4676,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.32024685900752603,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.458,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.39792226170970474,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.4767,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.32914373581481476,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.4644,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.5449387039178665,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.4778,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.4697008995276977,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.4662,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.3383688263013948,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.4574,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.4169022595700285,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.4447,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.32972059037032275,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.4693,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.3305199925739852,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.4615,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.3454532480966081,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.4677,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3454494441577979,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.4673,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.3837917330017202,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.4744,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.3126389369462563,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.4809,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.5439784808505859,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.4795,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.32481514097458875,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.46,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.3360794023271947,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.4646,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.28947026294996736,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.4678,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.35631305272142993,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.4598,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.3342811589710453,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.4848,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3067293802231825,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.4502,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.40428418080981005,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.4569,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.3124827533533047,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.4765,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.31726727072327454,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.4615,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.3230774078613723,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.4599,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.3091285698870702,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.4477,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.2992165526046311,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.4593,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.3179670844627982,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.4747,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.2956216749277053,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.4456,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.32832672430447907,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.4554,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.32952945931851746,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.4604,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3141723549376203,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.4829,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.31876527484946526,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.4578,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.36318560411503575,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.4792,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.3179196362611832,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.4828,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.3355785750475508,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.4823,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.334874854109676,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.4585,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.33812953353340225,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.4715,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.37285795363304747,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.4869,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.32192466876849674,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.4788,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.36919129084838515,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.4818,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.3312417843035549,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.4869,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.3947153959455223,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.4701,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.3706047832265801,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.4729,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.8040383210143381,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.4957,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.319366115223068,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.4894,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.45832728261837274,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.4633,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.3769763338369308,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.4636,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.35946722785004154,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.4647,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.405818805321382,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.4745,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.3205237967059495,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.4566,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.38177410190965727,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.4723,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.5136658584683855,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.4656,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.38808974313171457,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.4631,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.3850216615342763,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.4698,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.393986621552104,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.478,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.40084403262318724,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.4639,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.459186504765098,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.4746,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.349754314364059,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.4641,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.8197006366764972,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.4501,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.3323147463879523,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.4712,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.42102304781519806,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.4631,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.38037749662296877,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.4871,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.353212580043208,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.4662,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.33693945297135686,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.4901,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3339763082219477,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.4759,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.3696006205735763,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.4604,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.3297246675984372,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.4717,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.33759115118048333,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.4701,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.322732089667909,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.4787,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.32402121433501696,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.4698,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.3263394903113246,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.4439,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.31486329743237174,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.4795,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.351876640330087,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.4583,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.3465348307289402,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.4763,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.3551126006046667,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.484,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.35899018389327125,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.4578,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.3343565061282508,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.4909,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.3819399314163558,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.4677,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.3873634888446953,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.4597,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.41518918871735533,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.4724,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3256013270892894,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.4667,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.3637322991192564,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.4864,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.41703938993019,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.4853,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.3511616088652933,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.4652,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.35820411876605296,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.4622,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.33979454123582054,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.4663,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.3266909239335476,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.4607,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.3408634018080721,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.4615,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.3802623309829871,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.4561,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.3321814583988797,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.4753,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.40130796669628954,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.4594,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.2878774256077493,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.456,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.33662423337868524,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.4523,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3156717633347705,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.4696,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.3227678613865946,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.4562,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.33221470920149704,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.4695,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.30525760417901837,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.4536,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.3106721408343425,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.4907,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.35154516178934997,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.4607,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.3006510992859843,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.4594,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.33959885959527986,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.4637,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.29236351870849236,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.4683,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.31754656718015706,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.4778,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.29368275361602103,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.4723,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.3089971986452745,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.4724,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.4174816820781015,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.4571,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.3315797335746648,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.4585,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.3023779417417779,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.4695,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.35284960342098737,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.4604,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.3291985542351881,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.4719,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.37052778651189755,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.4408,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.3338464657377094,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.4484,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.3453864624017306,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.4627,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.3043820872875355,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.4556,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.38247555459977284,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.4661,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.3492419323558293,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.4606,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.35471311163345054,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.466,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.4122446078708342,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.4643,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.34917100643544335,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.4927,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.3857108119653685,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.4478,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.33797385091212334,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.4735,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3698625809019641,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.4688,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.8435681195525462,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.4652,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.3347002072291885,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.4792,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.3035186002486259,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.4805,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.341152104211212,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.4885,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.347755530328265,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.4379,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.30802913569009105,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.4706,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.39226099737005327,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.4555,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.335998071936328,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.4581,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.30817913938011127,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.4689,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.35858606794343784,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.4477,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.5302222478384304,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.465,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3175918402248559,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.4541,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.2984607907892039,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.4838,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.38310077790629615,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.4638,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.3137968306437046,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.461,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3864850042333099,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.4539,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.3378520209829769,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.4675,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.34243229053389046,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.4474,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.32170392832341077,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.4694,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.3180232358092572,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.4432,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.3605520967644294,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.4813,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.293607027269886,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.4813,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.31722945569637356,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.4708,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.35451840175580573,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.4532,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.296613201093117,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.4717,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.3105722219137452,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.4411,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.33162232961807747,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.4634,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.30441139682377016,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.4625,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.31173296241203163,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.4569,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.3239694222254036,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.4648,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.30287255554987225,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.4428,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.315976739617753,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.4679,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.29947944989248443,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.4776,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.3121786917217868,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.4852,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.2974249375028055,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.4672,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.3038201923273246,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.4646,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.29864044517569616,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.4737,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.33114665850931524,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.4438,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.2752076393055526,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.4602,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.29706355760356606,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.4654,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.29909027052302906,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.4577,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.29652461911301936,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.4551,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.2878385375542494,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.455,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.3057599368820231,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.4595,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.2846123226732762,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.4622,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.28248655251533117,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.449,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.3075372630474363,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.4487,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.3054247225808893,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.4711,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.3655311153723275,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.4723,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.2809173969227828,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.4542,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.3260433247321309,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.471,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.3049648856839692,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.4523,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.29947519241118187,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.4663,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.3485739711118224,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.4482,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2825330097433293,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.4527,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.49792823106456885,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.3974,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.37197660858170384,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.3999,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.4996922656312685,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.4046,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.4052891080113843,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.3948,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.4635915906857298,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.3979,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.3188478144618427,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.4083,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.46607036324976037,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.3878,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.3421743114311729,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.4208,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.39570726074256946,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.3821,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.37649819704279885,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.4019,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.3150499350586564,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.3887,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.3706329884162076,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.3996,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.31936112353586443,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.3824,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.3048136554679152,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.3841,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.35122525971070734,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.3883,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.306714842699935,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.3967,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.3025171329352706,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.393,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.31759706263295545,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.3772,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.31812857135988015,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.4068,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.30082783497769783,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.3881,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.31401374227574125,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.3984,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.29903380491493786,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.3815,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 1.1664727090686968,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.3898,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.3235712442371602,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.3812,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.30274208992875534,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.3888,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.282875747593255,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.3993,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.27818148387125624,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.395,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.28013324547512247,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.401,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.2994934549687295,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.3866,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.2884379371492975,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.3872,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.27466916420180043,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.3887,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.2803014711534082,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.4007,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.2721777881035549,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.3972,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.32209779103213315,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.3943,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.27021716989590605,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.3897,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.31516430134368917,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.3738,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.27071522828959127,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.399,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.28051938886857253,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.3889,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.30630525568749994,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.3725,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.2749196032851082,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.3886,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.2815939027017723,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.3964,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.36858172639352377,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.4003,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.26874268221039344,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.381,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.2726002742905712,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.3998,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.27999363650339415,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.3831,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.272411346094521,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.3949,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.29505009393366094,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.3944,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.28679459023241877,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.404,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.2792041035035591,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.394,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.29879158221566376,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.3948,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.2617655916735681,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.4052,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.28288871804667276,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.4113,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.2747075275858017,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.3758,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.4152656155726165,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.3965,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.28368371922770064,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.3877,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.2833386558662039,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.3927,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.31809897356428607,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.4178,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.27710290387189024,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.4043,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.31178948899302056,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.4082,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.31096804737263994,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.4031,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.29014487030983405,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.4132,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.3001335162447315,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.4156,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.2845276668169966,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.3817,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.3060401190424785,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.3736,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.26262813442260985,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.398,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.30859406192693783,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.3808,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2903133489137776,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.3888,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.31029021791895856,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.3945,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.2634919971318218,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.3906,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.26395249722661734,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.4031,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.28455067026014874,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.415,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.2671541507171164,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.3959,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.2794318441246143,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.3888,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.26259353780478256,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.3853,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.26055173814829485,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.4073,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.2714761703014528,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.3926,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.2830101619429286,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.4056,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.286039239620889,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.3841,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.27914139463628784,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.4101,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.3150740906949134,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.3993,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.29597394128747984,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.3852,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.2760626917463679,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.4026,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.2917509378318829,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.3867,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.2711495814667351,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.3887,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.2970492333633172,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.3677,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2883464186216226,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.4065,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.2717740180683508,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.4116,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.2845650057895562,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.3884,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.2665177415214521,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.3955,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.26005988006320807,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.4171,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.2768940590108311,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.3955,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.27202107074847426,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.3786,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.2695914828236664,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.4067,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.2770241871792155,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.4084,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.26724786788673477,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.3853,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.2762244413926409,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.4018,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.2562634650958191,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.4085,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.26807194276540763,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.3949,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.26498955953820874,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.3943,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.2643294819476071,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.4078,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.24938643992717377,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.3928,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.28287950754277486,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.4028,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.27082576306375944,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.3771,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.28942268633759877,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.3913,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.2985744986925461,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.4222,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.2589216976939352,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.3877,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.2858648776654393,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.3923,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.27450970844493666,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.3754,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.2811689119112176,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.3964,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.2832118901300103,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.4116,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.26198796924686546,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.4024,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.25551810526693575,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.3825,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.2853140050773187,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.3933,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.2556640170872582,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.3703,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.27343304491883985,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.386,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.29256021563267737,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.3715,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.27329113835588104,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.3828,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.2724372201659281,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.4142,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.2988184445200956,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.407,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.2835434159484771,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.3672,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.2731123415611822,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.3689,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.29704601401201575,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.3973,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.28648869961190565,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.4028,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.28803045777106906,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.3797,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.2763887181959277,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.4092,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.2957295918716548,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.3986,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.275052084417325,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.4128,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.3010959440206444,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.3871,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.26652365831692604,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.4002,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.27749752524015586,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.3968,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.2791740818714642,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.3941,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.2855051707341536,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.3868,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.2697029430401036,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.389,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.27010461929897606,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.3894,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2773745457919873,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.3847,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.2552310564406112,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.4081,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.31289217196182606,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.3957,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.25603576626046437,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.3886,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.25797794683574965,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.3921,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.2505879231759512,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.4012,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.27130674723961223,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.3943,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.25183231163801256,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.3914,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.2560540778201356,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.3782,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.24621752571504055,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.4188,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.2690005536980214,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.3871,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.25047708769723276,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.4076,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.24919275831172746,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.4126,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.2741739154445522,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.3441,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.2812346334245182,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.3848,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2634122270651492,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.3809,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.2861800047401816,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.4029,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.23986038569640963,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.4016,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.2667140550028946,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.3873,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.264873354606196,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.3914,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.26522373111274267,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.397,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.28194461216241806,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.3647,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.2755963738045635,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.3659,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.26766684901504323,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.3731,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.2626585361476057,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.3681,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.27952201507700875,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.3741,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2679792874709975,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.3792,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.2534544166735033,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.3691,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.2729575703062057,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.4128,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.2830668482049247,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.3848,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.25244287163259876,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.4014,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.2581181138920526,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.3798,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.2726336646036804,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.4006,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.274303588051596,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.3836,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.2604614392882015,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.3999,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.25735331701120295,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.4177,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.2570369727530352,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.4001,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.27601137058508585,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.3805,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.27233869193680565,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.3933,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.25774169852193457,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.4144,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.28812147094977086,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.4,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.25472825656113113,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.392,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.28412431408470046,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.3955,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.26097876550381477,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.4062,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.25143064098973994,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.4075,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.27281538523170074,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.3895,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.2585654844865703,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.3911,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.2682502358443095,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.3727,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.2642528575188834,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.4041,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.29197307166866515,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.3974,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.25306449037079415,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.3979,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.2608887919245467,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.3816,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.2632728461822644,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.4172,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.2784252909426177,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.4016,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.24586382907915258,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.3794,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.2418115295296207,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.3773,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.2643479146011233,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.3759,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.26844689194908433,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.386,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.7209030892174727,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.4044,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.2540580452482532,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.4,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2699078060542038,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.3768,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.24854266842420653,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.3915,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.25506932229199913,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.398,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 2.8757886841706015,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.3858,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.27771403750529733,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.3682,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.2634881363820007,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.3664,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.25268540939337325,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.3904,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2712602558969762,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.4052,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.2683430802586374,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.3867,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.24777827992704007,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.396,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.24283992473927485,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.408,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.25772678721739034,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.3857,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.24206385124058769,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.3921,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.24134119095261672,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.3747,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.2528736262944833,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.3949,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.24605450798058517,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.3903,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.3158686421087087,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.4135,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.25017737204212903,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.3949,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.26097332521731736,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.3966,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2738261533938812,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.3672,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.23865724207907807,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.3954,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.24671449567035822,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.4058,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.25464211586435026,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.3935,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.24223119734040713,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.3749,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.2597433344000261,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.4048,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.2346767564513443,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.4227,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.2202656628882042,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.41,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.23567499685133173,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.4021,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.24251992380451107,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.3853,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.33487949026151864,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.4131,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.2434746689384033,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.4007,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.260419141633063,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.3768,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.2255130457948597,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.3993,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.23430997020481975,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.4077,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.2506000472146237,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.3883,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.25410174648113704,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.3674,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.23965869194164338,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.3835,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.25062058496488593,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.3892,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.25673711800787247,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.3921,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.23398640018861533,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.4001,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.250344718456366,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.3601,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.25009237901745385,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.3912,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.256648007364594,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.382,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.2496639293352834,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.4124,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.23326839576436526,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.3943,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.2279021337940373,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.3924,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.2430492648903128,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.3886,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.2402660403451493,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.4252,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.24764357996357889,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.3984,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.22636596935920383,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.4099,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.25890118201684925,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.37,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.25577907166661396,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.3908,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.23566777572153694,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.3887,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.4547774407746576,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.3231,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.2882776256452512,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.3357,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.30178083355730717,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.3709,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.4606584079134939,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.373,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.3881149803098579,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.363,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.27740543027532366,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.3368,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.37259944124789024,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.3393,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.36811189824599455,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.327,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.28444309643421206,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.3557,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.2581161767982335,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.3513,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.28837111211409344,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.3386,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.3477610684053666,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.3326,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.31351951342888756,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.3402,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.25774499265275336,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.3654,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.26561189373064453,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.3475,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.282461936597404,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.3626,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.2879200190745944,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.328,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.2657287394847022,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.3678,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.26480380747685306,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.3534,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.26347216326998385,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.3416,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.29147703608545794,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.3554,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.2769435592414976,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.3444,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.5373644676304785,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.3583,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.23295689129967168,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.3555,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.2746574618034218,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.3473,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.24706075813784834,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.3569,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.24730889098252973,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.3401,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.23764913871328358,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.3261,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.23801033509100192,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.3366,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.2451286041842507,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.3499,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.23490457283814856,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.3651,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.2565018912131711,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.3314,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.24516899887260066,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.3778,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.236242550110834,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.3579,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.2348822174778194,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.336,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.2563309868965885,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.34,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.22822577304539152,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.3552,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.23333109071105723,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.3525,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.25228673546087343,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.2945,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.2350858837509772,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.3278,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.23465057759988475,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.3392,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.22572920210381087,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.3439,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.2227944722987408,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.3478,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.22073166849607068,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.3508,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.2352644010158919,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.3399,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.22270996724081635,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.3721,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.2361200344335744,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.3246,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.21984003294446947,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.3411,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.2380960610578752,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.3658,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.23240292698506634,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.3364,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.2421956778808715,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.3404,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.22682706793800408,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.3496,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.2234805539318305,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.3493,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.21785309969978536,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.3531,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.24013805453501463,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.3298,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.2204969692181197,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.3361,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.22550755577208279,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.3166,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.23477978964748647,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.3388,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.2307372228741444,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.3265,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.24050669880738634,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.3166,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.2139349206745003,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.3476,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 3.568407213675263,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.3569,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.22122645335144253,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.3818,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.23344343951132304,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.345,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.23071832199605957,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.3366,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.2179416787426558,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.3514,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.2225604273733167,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.3343,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.22150658647702956,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.3719,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.22620217037057913,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.3456,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2322119921318404,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.3636,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.21912028470807238,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.3521,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.22162219322747218,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.3519,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.21659855036044764,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.3466,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.22283678220936637,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.3617,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.20833798929528066,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.3584,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.22943112963937792,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.3218,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.22630303506132235,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.3315,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.21938854782239756,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.3301,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.23297982620318097,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.3344,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.23012095342397382,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.3264,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.21084352112604443,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.3597,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.21506079804989695,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.3631,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.20954725193820045,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.3328,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.24635771270469806,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.3453,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.24406519724105988,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.3437,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.2129350685571337,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.3549,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.22306413412053688,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.3624,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.2297552080148167,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.334,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.22401530470256242,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.3186,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.23377130702131138,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.328,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.22431567283120354,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.3592,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.21067294081202145,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.3612,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.23364280934103066,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.3505,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.22674862863252343,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.3422,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.21484328353580523,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.3542,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.2090699565672095,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.3697,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.2161944502362044,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.3606,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.22945994789870933,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.3425,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.21717791617841448,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.3596,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.21465832420635217,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.3514,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.22008703709163838,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.3365,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.21079439086334723,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.3779,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.20481890254800963,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.3819,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.22213212087062234,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.3258,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.22177373403091835,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.3735,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.2332535177951933,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.3282,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.21653046313984475,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.3607,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.2161021374539964,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.3579,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.2136382501116449,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.3646,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.22134458155299766,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.3288,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.23549084536839227,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.3373,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.22638719937592056,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.3351,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.22381509879984435,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.3488,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.2177968709959963,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.3214,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.21609195127742858,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.3377,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.21709561473026306,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.3459,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.21461272086929298,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.3309,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.21446093463067528,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.3489,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.2075987330280051,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.3252,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.21082889174674965,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.326,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.21545484456106406,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.3583,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.2163203256919409,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.3502,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.20861631811274417,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.3607,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.21894601731115645,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.3453,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.2152107999920446,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.3483,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.25909495388915765,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.3425,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.20554913832896513,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.3533,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.21062358858909258,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.3595,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.22420160123984703,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.3421,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.2137925803478535,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.3228,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.2107561348318557,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.3619,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.21148558621087857,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.3474,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.22695557699728167,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.3266,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2226202385638799,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.3555,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.2232644716927722,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.3395,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.21151895357189923,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.3392,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.21614735965017312,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.3398,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.2232190769156918,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.3431,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.21098003044721383,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.3344,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.2428350160398916,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.325,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.2180532496794213,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.3293,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.22775548776892085,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.3369,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.2120599473118699,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.3557,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.21279498617551557,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.3355,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.21041059949587043,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.3698,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.21282711665793536,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.3618,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.21100798441874957,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.3636,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.22323520922173457,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.3523,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.2194769967720978,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.3196,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.21202626516398143,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.345,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.21105261088874244,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.3176,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.2336678987719303,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.329,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.22910086229390783,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.3344,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.22154546903414132,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.3432,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.2109257877187925,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.3365,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.21124929964410957,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.3608,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.21747688706619284,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.349,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.2144787508780946,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.3439,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.21617275733220015,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.3398,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.2199569222364298,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.3374,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.21851372515789547,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.3373,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.21893777957659835,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.3385,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.3117438311919384,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.367,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.2249233378599933,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.3448,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.21928294935688808,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.3274,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.2166608290793675,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.3433,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.21833731878594356,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.3378,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.20937329675354344,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.3264,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.20640094164912026,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.3526,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.21225757393560174,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.3383,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.20630214098807748,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.3564,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.2190645384057256,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.3253,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.2139919562043125,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.3495,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.218807365362262,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.3768,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.21346318248919138,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.3658,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.22469876207553452,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.3552,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.2146349052287394,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.3566,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.20991624957012917,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.3282,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.20810845913555664,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.3676,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.2220492820944153,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.343,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.21000773941153944,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.3294,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.21042451277375882,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.3268,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.23060844573354716,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.3308,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.2119942381264809,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.3585,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.22713721943412252,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.3531,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.21000977183863137,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.3572,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.20597103659520322,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.3494,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.20857445315336715,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.3379,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.21807496388129743,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.3309,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.21578353360227168,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.3387,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.22302017355369383,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.3322,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.2115317519910877,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.3611,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.21367300138271839,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.3403,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.2100064484368674,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.3358,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.21462023340524936,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.3368,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.22060605662380128,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.3498,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.21287922498639852,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.3447,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.21081794276077098,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.3553,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.20620298262931483,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.3618,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.19907385474401956,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.3672,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.2108014802296119,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.3374,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.21475898725637912,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.3429,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.22508233834677646,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.3421,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.3232213253119292,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.3509,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.21104568382307326,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.361,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.21956827669525383,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.3451,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.218766920660492,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.3632,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.21269681885438735,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.3765,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.21278540471074203,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.3249,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.2663903061965398,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.3413,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.19379662338514325,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.3507,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.21692413137299277,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.348,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.19929666988724626,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.377,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.21917659276597332,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.3557,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.2080889296672673,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.3612,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.20677457523593024,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.3513,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.2105368390461654,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.3295,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.21254144058394567,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.3524,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.21660256830646565,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.3476,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.2114312214765133,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.337,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2147418836057327,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.3384,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.21749551459856878,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.3175,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.2047114074412061,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.3579,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.22069240562475842,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.327,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.2228340728905051,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.3236,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.2144673313898419,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.339,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.22178536869136495,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.3514,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.2139737912823681,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.3563,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.21731903072501255,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.3502,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.20417471669125803,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.3533,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.21038110319896824,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.3539,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.2122568069506192,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.3479,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.2175283238589436,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.3494,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.20681561849245214,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.3493,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.21743865764502693,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.3686,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.2080119352874074,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.3633,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.2202017679106831,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.3419,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.20795143218481388,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.3531,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.2430799417575536,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.3436,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.20933786970577425,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.3428,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.2093104322574686,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.3467,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.22485765472188504,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.3224,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.21459779868048023,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.3409,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.9385349037069886,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.3319,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.20953810588231944,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.3272,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.22776432975513053,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.3339,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.20257053780754497,
+      "learning_rate": 0.0,
+      "loss": 0.364,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 4.3040093703891845e+18,
+      "train_loss": 0.4847247115030945,
+      "train_runtime": 29049.913,
+      "train_samples_per_second": 5.439,
+      "train_steps_per_second": 0.043
+    }
+  ],
+  "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": 4.3040093703891845e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}