diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 7.734643482805047,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.9488,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 7.3096977824430684,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.9214,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 7.461691284409732,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.9277,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 7.157892058831017,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.9162,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.96903092516795,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.926,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 6.820907145131145,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.9164,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 5.480552489820014,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.8688,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.020107130681183,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.8527,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.912113670172105,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.8012,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.427408630832678,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.8058,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.45500564607622,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.7969,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 4.32656848793759,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.7735,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.566260165536825,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.7634,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 4.6404954493744315,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.7686,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 4.548894132607465,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.7558,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 3.3928839292738284,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.7056,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 3.1443238710139547,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.691,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.7676844191206205,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.6974,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.8545749717137758,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.6871,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.4221487547155798,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.6695,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.6487912022264066,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.6564,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.9285691144877062,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.6446,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.502008869993801,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.6169,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.0146559841609954,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.6275,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.1259025576757071,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.6237,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.1596152562764515,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.6206,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.9909502685388257,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.5901,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.1120645470474066,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.5717,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.1946019086187338,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.5798,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.8204617865831552,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.5676,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.8170155650893812,
+      "learning_rate": 1e-05,
+      "loss": 0.5832,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 1.0786429582135955,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.5652,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.762374088679052,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.5588,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.7981841025342155,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.5838,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.7620285870401381,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.5528,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.7810574243217352,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.5547,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6914797249805275,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.5378,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.6707519846960208,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.5243,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.5963565565588691,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.5624,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.8798545386667745,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.536,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.5530901535008611,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.5298,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.704424513904442,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.5375,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.8485945160519994,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.5354,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.64317459328557,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.522,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.6727349449310095,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.5328,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.7308891854542383,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.5124,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.548024151224679,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.5117,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.7794801388741668,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.5054,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.6278877955265025,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.535,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.7601355087745416,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.5237,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.5914182233761857,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.515,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.7126630480725118,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.5208,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.7135592777239308,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.5141,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.7640945415613606,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.5027,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.6937232473798782,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.5266,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.7735455992383714,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.5125,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.8156536922091672,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.5146,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.8192512176595143,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.5058,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.8988868250279163,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.5024,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.7107530077488771,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.5103,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.7336396892995787,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.5108,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.8185073561534475,
+      "learning_rate": 2e-05,
+      "loss": 0.51,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.8600889161356826,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.5141,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.7160345006864813,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.5074,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.8285500941058619,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.5001,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.6135739931689834,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.5129,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.8050006028507454,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.5115,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.7685791158373413,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.5178,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.018104473443732,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.4949,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.8297702799020806,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.5008,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.6350803375322769,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.4879,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.654867122313852,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.4932,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.6309716319636256,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.4927,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7842745100619521,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.4827,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.9475282553402189,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.4739,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.0494660478721736,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.5026,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.8040309102172782,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.5093,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.9096235958366925,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.4865,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.8722589495547344,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.5019,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.71719831135619,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.4898,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.5885190444595908,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.4914,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.8646521434644713,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.5004,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.931073690610788,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.4968,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.9365061890826392,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.4985,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.767126084772357,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.484,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.0398828144646772,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.488,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 2.45989702092297,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.5008,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 2.5273160632712406,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.5122,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.2725551720411425,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.4921,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.6753425966757542,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.5114,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.4444172089129022,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.4923,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.276108649666813,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.4758,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.3560315100421565,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4745,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.0570083905625576,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.4854,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.0856470443888278,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.4921,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.1519740583991163,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.4834,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.8909279100970466,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.4908,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.9580383018388137,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.501,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.7485040055329627,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.4823,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.7866167735991243,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.4731,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.6332630032840428,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.4868,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.7574744200377741,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.4841,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.1064250709504093,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.4851,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.7497336271290229,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.4875,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.8452430031251453,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.489,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.8423552743822206,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.472,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 0.7661802787040434,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.4888,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.7361863706662598,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.492,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.6701439018920168,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.4756,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.6270137855721754,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.4788,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.7421264972479777,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.4703,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.732172580015961,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.491,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 0.6842938794710964,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.4864,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.7383925524196923,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.4892,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.9157573300260854,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.4887,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.0115514751554213,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.4845,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.1157241402226012,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.4859,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.9271795607126951,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.473,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.191768143335462,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.4716,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.7672578549689058,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.4861,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.150674389249796,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.4772,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 0.9128870105052341,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.4784,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.7437610944670909,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.4836,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.8654944000334199,
+      "learning_rate": 4e-05,
+      "loss": 0.4873,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.8564143637804627,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.4792,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.146909126708093,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.4708,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.6832197947243246,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.4754,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 0.7639353912783269,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.4865,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.7141546911912805,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.4678,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.6350617729765581,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.4915,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.6527337214670014,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.4689,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.6685005005785442,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.4723,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.7544004615783435,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.495,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.5473051849523892,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.4749,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.6952095186640709,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.4728,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.9333432634208799,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.4783,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.9622882912211412,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.4944,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.9474873918871494,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.4742,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.7939789214159919,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.4852,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.6492076378264587,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.4691,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.8822865206373757,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.4646,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.8247958431422373,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.4879,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.5277125456203351,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.4746,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.8550888705957267,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.478,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.6804364860861102,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.4816,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.5941538274594055,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.49,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.6280961103975754,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.4663,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.5993116648315235,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.471,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.5861071956376169,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.459,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.5151834483828006,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.4624,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.6434352384730035,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.4798,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5786109580433938,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.4953,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.6737422587964613,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.4701,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.6005221642547277,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.4666,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.6851524739150268,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.463,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.5621037241004109,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.4603,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.6168601230734387,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.4703,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.4840197288180902,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.4567,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.6553351208263217,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.4638,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.6763038694811396,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.4861,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.4776225063963911,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.4732,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.657323078817119,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.4574,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.5388858326154542,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.4588,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.5788939741406853,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.4828,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.6057276023247863,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.4628,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.5055215239242654,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.4562,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.7615103371967786,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.457,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.6148883869293625,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.4687,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.5666325622013614,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.46,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.6647134460344043,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.4665,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.531688668340499,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.4565,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.6208655967062184,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.4553,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.6246666474050012,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.4587,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.4427745013217288,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.4721,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.6057775438881163,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.4695,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.5144464163478175,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.473,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.5217898216837885,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.4615,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6370736466139368,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.4583,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.5241930762839094,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.4664,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.545754102519952,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.4668,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.6292708759528999,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.465,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.6562050271809652,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.4589,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.6298126360200873,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.4644,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.5692536990273296,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.4683,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.6272114809421838,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.4626,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.6456707882568733,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.4474,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.5275444621578714,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.468,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.8279548040712971,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.471,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.7415740643441256,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.4682,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.6766300674274127,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.4725,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.6823756090637796,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.4783,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.4573077466922982,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.4542,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.6034100235015191,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.4566,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.5435322830143426,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.4759,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.6367181950663509,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.4832,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.5015085207396889,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.469,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.5543357451832556,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.4548,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.6067998474109398,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.467,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.623239984941642,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.4578,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.6795410504735393,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.4769,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.5367294629677696,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.4576,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.4989189189529928,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.4652,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.5059215679302256,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.4661,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.4930286253120793,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.4495,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.5088350450095047,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.4557,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.46240995835342086,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.4503,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.49906687950294804,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.4642,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.5547390818135824,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.453,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.5400468572110573,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.4698,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.5213481375283806,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.4669,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.5618431115270125,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.4713,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.5461153796191732,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.4558,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.542366914387633,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.4596,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.5402276077330713,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.4545,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.5989317914242598,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.4541,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.4926166894950032,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.4559,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.504929079306462,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.4563,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5349251621850969,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.4651,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.5341752725241937,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.4611,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.5909362876275164,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.459,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.6512882167581038,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.4595,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.40326066823881174,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.4496,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.6632951477438782,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.4767,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.5515215370969041,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.4654,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.6475729508342792,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.4678,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.4905557106240392,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.4579,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.6048074572279282,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.4496,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6380302635412023,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.4562,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.4398664991391255,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.4463,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.525438161119976,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.4447,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.490200760120281,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.4497,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.4895284972140724,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.4545,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5848257881979536,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.4528,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.4985266413131312,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.4786,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.4949755754770271,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.4419,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.536804805670648,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.4627,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.48786260835782747,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.4592,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.5008313572017757,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.4674,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.5282345760048401,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.4594,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.5152991208768282,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.4468,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.45297186662033295,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.4417,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.4374491764095101,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.4628,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.5363894115221808,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.4501,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.44815265808401006,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.453,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.47524060230914333,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.4577,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.5904816289790633,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.4535,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.39775155684745195,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.4465,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.6682775967450688,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.406,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.4249998855823144,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.4252,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.5927701899468155,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.4088,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.5489414221946043,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.415,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.5407123395579995,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.4166,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.5927302650519181,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.4202,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.5427977732782798,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.4072,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.4239610499700297,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.4046,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.5611708765515259,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.4062,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.5038511113725981,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.4116,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.45957134435051183,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.4115,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.4662748786535415,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.4241,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.5937807818299485,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.3958,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.48085912246238455,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.4179,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.5074809885872571,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.4128,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.46987314708042033,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.3934,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.502195868033821,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.3912,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.5188287886043365,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.4151,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.4422477720981083,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.4152,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.529857285880172,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.4102,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.4748345674081377,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.3933,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.5731935358665898,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.4003,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.451288043397399,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.4022,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.4675787372357758,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.415,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.5410581354780803,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.411,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.5361239696886049,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.4055,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.6454002551622783,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.4137,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.6004787896714273,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.402,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.6838522460452854,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.3862,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.4368395534925611,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.3996,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5531786386024814,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.4081,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.48508722016054673,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.4109,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.5561179220068959,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.3975,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.42136421513460687,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.4011,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.5020469849644321,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.4099,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.4452798767993371,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.4101,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.5102653479408253,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.4102,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.47398901482619954,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.4128,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.46514763701790107,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.3955,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.46082046728552295,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.4048,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.44319599194712533,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.4115,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.5443787522489959,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.4142,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.43553690605999934,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.3982,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.4485483605885871,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.4038,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.5117643420404735,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.4038,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.45833222575652444,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.3998,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.4893198384991318,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.4044,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.4458892195432462,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.3809,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.49501443191861993,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.4086,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.5758217335181256,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.4172,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.5126736021925969,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.4048,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.5646488007492926,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.4181,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.5335675086694912,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.4022,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.4124779915694995,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.4116,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.4540804791307994,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.3941,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.4202418315115043,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.4116,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.555797393551368,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.4118,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.4874828594495045,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.4097,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.5023196220387411,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.3967,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.5137518069356997,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.4137,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.48943832071164806,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.4248,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.4359865163664032,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.4074,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.5150295294014279,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.4079,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.42316261750678485,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.382,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.49229270116077867,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.4103,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.4036350087971404,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.4063,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.4883716151952623,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.4092,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4174021807478955,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.393,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.3510450211295592,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.3908,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.4255857499687094,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.413,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.46073529931385676,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.3974,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.3806876107435705,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.4064,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5921226335642942,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.4084,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.46528336213161586,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.3888,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.5200006209182716,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.4136,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.49860762631375105,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.4092,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.42188208806130856,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.3986,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.3652076324151181,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.4064,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.44912525285498944,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.4044,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.42523316462449706,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.4056,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.47640503869793127,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.3957,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.4315666241433468,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.4001,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.3973527605392549,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.4143,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.4221229035148431,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.4073,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.43097802170168614,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.4107,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.5684974560534842,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.4174,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.4460477662536774,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.4324,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.6067474706246215,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.4027,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.5273067804015708,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.4019,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.5262203160830489,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.4166,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.4535383631930143,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.3944,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.6174815287116643,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.4059,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.4222162827489518,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.4025,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.7401281508422592,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.4198,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5296654916740015,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.4081,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.5945572100289696,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.4165,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.5777609340503994,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.4071,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.4955721907343552,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.4251,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.4571325009374186,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.398,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.5115291152860103,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.4156,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.48513507586649374,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.4105,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.4526201232775012,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.396,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.45510278573768403,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.4097,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.49062182938084087,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.3979,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.39300321114560854,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.3879,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.4354052503980858,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.3983,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.3645551187589373,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.4185,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.4902922669820506,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.401,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.3827271210500298,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.4129,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.4662860729839979,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.4085,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.5056681950001537,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.4005,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.5182326642284952,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.404,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.5011711855158377,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.3999,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.5090750707128616,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.4116,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.44813271306449043,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.3916,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.5219935067545384,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.4018,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.5120322458692358,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.396,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.4299774738887647,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.4108,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.4415918077858002,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.4105,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.42085528742882433,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.3906,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.5276199186208569,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.3996,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.4956540604713729,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.4179,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.5908564666340338,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.4152,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.46514729185581843,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.4354,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.5519220895929673,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.4175,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.388445538996307,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.3904,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.5307240563748337,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.4057,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.3801362123908875,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.407,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.4243712092459422,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.3894,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.4258917716541491,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.411,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.39911203251484656,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.3969,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.43551592338386214,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.4156,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.36326194689378083,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.4071,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.4398832095072527,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.3958,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.37481163605007334,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.4006,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.4298657276997694,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.4047,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.4011018967175088,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.3908,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.4679934328935953,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.4082,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.45569990031299673,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.4166,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.4990028599137004,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.3961,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.4420845517714287,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.4008,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.4153551882119935,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.4178,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.4486235898212959,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.4015,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.38026920850616186,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.39,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.44447648479165114,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.3926,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.4744695515475642,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.4235,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.37757948273219133,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.4001,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.4247226665752099,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.3987,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.41127042788984086,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.4109,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.3291750062581926,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.4063,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.4163567328087017,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.4041,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.3916523517332219,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.3859,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.39394166822956356,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.4033,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.44995005950666256,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.4051,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.42466761161224637,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.4139,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.49657189689384285,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.3929,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.39515959793661026,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.3957,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.40768736403915495,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.396,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.39996720691886595,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.4071,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.39894187115005086,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.4095,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.4081831078587505,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.4025,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.40554197100086503,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.3931,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.39038319398089943,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.4129,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.39423865160057475,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.391,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.37305619691240305,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.4124,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.4208788064194087,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.4093,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.40483700770815506,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.4035,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.4652663314774662,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.4096,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.40109926832412285,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.4192,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.44948379175270636,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.4054,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.4818324811906483,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.4154,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.42408202347621216,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.4109,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.4498238433998705,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.4057,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.4735223985054632,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.4014,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.38573271785090174,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.4003,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.4608040943787781,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.4077,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.4392752990610103,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.4026,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.43067880094994354,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.4011,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.4246806952181129,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.4114,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.3927736301719928,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.4024,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.4023754581338887,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.3997,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.3589001493438865,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.4031,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.45544435755133844,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.3979,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.34760632220905535,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.3968,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.4011115321463385,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.3924,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.3565301734645248,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.3929,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.35319067014658645,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.3959,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.40649540819011387,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.404,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.33854302289853855,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.4062,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.5040130454567064,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.3991,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.3964575967430866,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.4211,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.40254830977762185,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.412,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.4596193515010024,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.4119,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.308980559331853,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.3922,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.5380670589781682,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.407,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.4069283616464954,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.3994,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.41598343437778695,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.3869,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.4782545365993384,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.4033,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.36365538210707077,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.386,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.5427709556718253,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.4078,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.34942982548326246,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.4099,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.5063486772826034,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.4026,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.3877314337024519,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.3903,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.46357036378140576,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.3999,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.4731962795272894,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.4237,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.5057333069483468,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.4112,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.40482000378290134,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.4029,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.3885731616576007,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.4068,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.47047847503462437,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.4141,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.40143416326546205,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.4087,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.45837259024619664,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.3979,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.3864311057185557,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.3982,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.45032019565450937,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.4084,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.4039949369863002,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.4125,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.3829554588003787,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.3957,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.3799048712684953,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.4041,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.4114684296899068,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.4236,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.36465712940912526,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.4238,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.4654628401424767,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.3966,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.3968457581501677,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.4085,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.4073698121984442,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.4015,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.3968344220467487,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.411,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.3794528962715712,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.3977,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.36826263652691876,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.4001,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.39931712327022584,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.4085,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.3688570995977416,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.3908,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.4580885382554469,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.4046,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.3531732067253966,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.3992,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.4336722525694293,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.4,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.31733244103592373,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.4089,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.34598695432867,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.3935,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.39284792825731185,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.4117,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.360362706974779,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.4177,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.3537562495282027,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.4066,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.35137341320255566,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.3888,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.4540214143378079,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.3991,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.38009599138919603,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.4022,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.41777601694393135,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.4002,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.3791151401170273,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.4086,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.36887879177932187,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.4197,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.35145597364389,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.3848,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.4164740192940669,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.4102,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.3908646478250952,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.3931,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.3884885218221347,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.3827,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.3407944206975365,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.4089,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.41605632237467705,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.3976,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3666181924262721,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.3891,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.4741847068090917,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.3255,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.5058757386061294,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.3462,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.4419208025162146,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.3566,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.39143093208682245,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.3445,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.4405885686630548,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.3501,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.38099541512655755,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.3323,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.5028001258584887,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.3413,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.41270435038032127,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.3298,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.5388412996708446,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.344,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.36115242645876583,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.3349,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.4839508495298222,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.3493,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.37601738974258814,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.3397,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.419010822724959,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.3383,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.38588096759621127,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.3426,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.33198861595851875,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.3615,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.3821479777773096,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.326,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.3266493210674952,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.3429,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.34511723978801306,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.3305,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.35692621785691403,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.3364,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.39755853267805846,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.3407,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.3504015337338433,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.365,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.3884857284913918,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.3474,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.3573782556635476,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.3265,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.3545175967172827,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.3566,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3528141202051488,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.3382,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.41640372759157523,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.3262,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.35186866258192506,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.3265,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.37007953612435024,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.3316,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.3527666324794516,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.3448,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.35757947296457215,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.3239,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.3740674045647777,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.3539,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.3372592007254995,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.3378,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.35012794900051875,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.3492,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.32265863444706866,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.3397,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3717425158125704,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.3564,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.3775512298358198,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.3326,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.32552390546736387,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.342,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.4074791774597016,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.324,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3433129894442334,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.3301,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.34281684620817965,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.3115,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.36975533423863377,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.3321,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.3807531590970537,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.3387,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.3137682766393665,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.3366,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.38153769220112893,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.3344,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.35350292390106774,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.3592,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.35638873929687476,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.337,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3566903588023731,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.3341,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.385021052991219,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.3374,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.36866119566108074,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.3307,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.40566829397516546,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.325,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.35618153532024666,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.3637,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.46768432219842887,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.3326,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.34968363537815766,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.3294,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.4731054727572334,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.3416,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.39907494611328714,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.3331,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.4624285150595924,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.3378,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.3526825337738502,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.35,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.4382034479715699,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.3352,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.3504990227582377,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.3281,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.4377725387813618,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.3464,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.4316848521205376,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.3366,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.4877410145944489,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.3602,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.4523190155331149,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.3269,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.3957105356864238,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.3207,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.5249277057724323,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.3438,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.36499041763324735,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.341,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.471623715621487,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.3568,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.36628126695462626,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.3303,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.47327130110615107,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.3444,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.3504884813212652,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.3362,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.46283541642477516,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.3308,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.3114562898987759,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.3237,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.4075082826416107,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.3231,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.33510287316588844,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.3446,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.3476545506419642,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.3363,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.3722194566846158,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.3601,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.32805730456365917,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.3365,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.34779907662216136,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.3521,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.32008693975311386,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.3232,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.34735751007066107,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.3265,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3427819370767772,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.3497,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.312782317827876,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.3532,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.37542839882538376,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.3104,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.3543846707849228,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.3172,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.3908499291186466,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.3498,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.38147865361482214,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.3485,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.35612384217453924,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.3489,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.4190736103102229,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.3499,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.34943440224314276,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.358,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.39241718942659326,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.3564,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.3436843877365785,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.335,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.31908839178047554,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.3291,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.33000888183595367,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.342,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.3135413519911013,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.3416,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.34086101483396075,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.3543,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.3400217386456065,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.3221,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.34618228553856023,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.3586,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.35336339777243314,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.3458,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.3631081423828395,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.3526,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.3196774001667569,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.3454,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3638155005294562,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.3283,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.33202179000830334,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.3479,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.3330856883274656,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.3393,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.31325651082353667,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.3466,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.34544180746682196,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.3461,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.3465455349890137,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.3368,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.36006527762022833,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.3405,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.3361791584310369,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.349,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.4066752186263788,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.3373,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.3296059170632213,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.3139,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.39587765117549656,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.3342,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.30818404253599285,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.344,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.34516947451551966,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.3691,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.29047498740360767,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.3299,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.3254926054328272,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.3444,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.3276881668747399,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.3388,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.2986593816048259,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.3436,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.3202474370600838,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.3469,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.3614897825928247,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.3449,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.38269618565899244,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.3389,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.3322057376198955,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.3455,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.32492161283981913,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.3363,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.3897367587494612,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.3357,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.3129605927599904,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.3456,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.3726179457595831,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.343,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.3313381772170867,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.3438,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.3360270181074832,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.3458,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.30965171607925,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.3422,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.33592428044761913,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.3653,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.3231795592404518,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.334,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.3180983037850161,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.3502,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.3215667203264101,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.3557,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.34128979250970115,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.3458,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.3398492739468077,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.3392,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3637122347820278,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.3291,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.3491883360293926,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.3226,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.3488798145063822,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.3292,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.3181605604147677,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.3546,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.3282864474457569,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.3467,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.33433625170184644,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.3478,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.3148605003450344,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.3573,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.34794383036638404,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.3425,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.3091980840880252,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.3478,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.3818434414002979,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.3393,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.29820776912621966,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.3527,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.3968089193357377,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.3439,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.3102461910463243,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.3439,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.4104268004950747,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.3354,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.3170261154344354,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.3406,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.32041183399337675,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.3168,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3357768601414828,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.3469,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.3021708413175612,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.3367,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.32641697777038564,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.329,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.2782371809230084,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.3191,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.3071323703807642,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.3612,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.30177934808079443,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.3471,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.33379856813779235,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.3483,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.31333406729516816,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.3397,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.31493139307679135,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.3302,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.3279087848808713,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.3273,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.32051014639315173,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.3394,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.3268777785725164,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.3495,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.367082641382404,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.3601,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3101384589839179,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.3329,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.3371414273036539,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.3316,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.35513285905268027,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.3391,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.3199663990075157,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.3507,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.31156575652792934,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.329,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.27919427094828253,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.3383,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.36556342673210074,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.3458,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.3164759445545137,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.3288,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.32847483674540834,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.3446,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.3159751522985701,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.3603,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.34550402423523224,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.3325,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.3347543383252343,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.3284,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.30631541005918617,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.361,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.37300189756816654,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.3248,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.2944084250484221,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.3417,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.35322369821882166,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.3273,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.2900876244088141,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.345,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.34766357330718023,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.3322,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.3410441872905066,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.3335,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.3456964506349768,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.3398,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.36208211168502874,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.3299,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.37476447497645143,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.3357,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.3755506896814418,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.3558,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.35400080461456956,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.3364,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.33891541089246896,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.3397,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.3356269449246293,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.3279,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.40032514723031626,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.3502,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.2959262866701889,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.3538,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3324774403977879,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.3336,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.3633496763993631,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.3422,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.32287226041317435,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.3366,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.3681491563649062,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.3515,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.3034277001125987,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.3402,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.3556749912161671,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.3441,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.3185566192809673,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.332,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.2942909935401499,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.3521,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.3439735510296535,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.3324,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.32087366540411627,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.3645,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.2919526588712567,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.3478,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.31898444031379863,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.3515,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.29246986431643573,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.3447,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.2852081420536923,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.331,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.302993331875993,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.332,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.316126286158395,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.3489,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.30363883287365034,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.3386,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.2909446047600821,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.3355,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.3234775154334607,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.3517,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.29315916197714315,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.3232,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.288255775854415,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.3461,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.3669862513057425,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.3426,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.28946885528387173,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.3331,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.3331070543872592,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.3406,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.3691371333746214,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.3517,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.3201232938138458,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.3269,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.3571203253341927,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.3495,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.34764589498762705,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.3335,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.35330534459343677,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.3408,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.30028614071090304,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.3151,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.3330210846148145,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.3416,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.2874921733655868,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.3379,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.3322079340240357,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.3386,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.31222739847178865,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.3361,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.30678986621066584,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.3348,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.3115526716976879,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.3481,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.336188146079344,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.3458,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.3476858782225495,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.3533,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.31488698944831794,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.331,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.2827570261895736,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.3317,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.3142360114223916,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.3352,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.3096041774814849,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.3451,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.2912501317153361,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.3267,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.30373963168537355,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.3348,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.28814851652487405,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.3414,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.279427504723316,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.3261,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.31926983899462136,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.3404,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.28874794026750367,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.3293,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.30669762805568185,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.3479,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.31943703646238214,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.3402,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.297148539109757,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.3424,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.3098378341206031,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.3454,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.30418238782915197,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.3433,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.28650590339345233,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.3367,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.26687455044689024,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.326,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.32202634911434486,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.3237,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.38517404741774014,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.2872,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.29147771796254707,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.3248,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.40359857138800204,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.2744,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.3591643455389048,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.2674,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.4004934029126143,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.2933,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.36899838399153945,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.2846,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.3288179510442746,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.2982,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.36584643225654345,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.2876,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.33819578829468816,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.2759,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.3014661978818876,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.2792,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.3208301319381475,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.2876,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.3132797051423753,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.2927,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.2998889794932974,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.2944,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.30200662796750244,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.2721,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.33017737520735957,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.2826,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.31004648388250994,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.2933,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.28878037072513263,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.2873,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.2861629463085056,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.2905,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.2616046661927483,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.2963,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.27190487365113747,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.2774,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.28747948508435306,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.2721,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.2911543552618015,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.2852,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.2823211720001752,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.2955,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.282386119717298,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.2763,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.3212524932645384,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.276,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.27566567402492004,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.2938,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.28777404086433395,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.2699,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.2893644863606055,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.263,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.29650781155583644,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.3039,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.27720812047698395,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.2786,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.2701702451185461,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.2892,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.2921233455655939,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.3055,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.27803071036157706,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.287,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.2720810884946512,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.2737,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.26618077374789406,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.3,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.2739959258821922,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.2694,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.2703660703714418,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.2895,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.2872557502483679,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.2919,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.29389752827832305,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.281,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.33653805518689933,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.2871,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.3005503355994217,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.2661,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.34595930589516577,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.2899,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.29969530400187017,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.2765,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.27658737063695626,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.2595,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.31529454584814187,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.2832,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.2756091432997394,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.291,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.2970314596618258,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.2838,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.2933168913086024,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.2884,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.30011873531453964,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.2784,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.2921703195682315,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.2676,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.26217863090727717,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.2822,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.28344396063248734,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.294,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.2645809038025676,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.2754,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.2593208427525029,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.2906,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.25008234772205484,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.2761,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.27232558611731916,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.3017,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.28894214802364326,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.2654,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.26631719895738354,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.2925,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.3080081434362981,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.3097,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.24963235653925944,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.2887,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.25242077939372076,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.3001,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.2806942110948762,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.2587,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.25303024564506454,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.3019,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.26842479596120544,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.3047,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.2998161414512404,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.2638,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.24886811779250237,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.2829,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.29053912516077207,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.2899,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.26682926848632915,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.2962,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.26494907564499626,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.2829,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.25630758499935996,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.287,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.24818697834209616,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.2827,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.25215563327821794,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.2895,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.2693092053905734,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.2805,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.2541886840621306,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.2675,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.2789435005563363,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.2695,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.27154191058851795,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.2919,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.253565826217212,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.2891,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.3076568259436967,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.2874,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.29796163277167836,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.2968,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.27527034683563684,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.3057,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.2640846874115319,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.272,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.26067627908538504,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.2864,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.25563725661788733,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.2798,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.27527015495609836,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.3126,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.2726036088599807,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.3016,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.27029526929639613,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.28,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.2647722505297886,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.295,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.26532026480706306,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.2678,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.2700337345648133,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.2683,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.2718878921961319,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.2861,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.272251550390087,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.278,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.2720644443186734,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.2993,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.2836691318148551,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.2641,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.27850272055119296,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.2974,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.26025086233411104,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.2953,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.29186714759331345,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.2851,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.28002606022642046,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.2842,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.2678047557235608,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.2583,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.255501766863768,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.2916,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.2912894834880497,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.2835,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.26274023639563,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.2973,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.2641176076124428,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.2906,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.29080436054354475,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.3019,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.28224805545261417,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.285,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.2725998117381138,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.2717,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.2720718838590378,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.2868,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.2843361980292534,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.2728,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.27173790218502525,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.2912,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.27517342044883425,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.3093,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.25300662897756876,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.2927,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.2570322342015494,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.2929,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.2611398258843862,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.2777,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.24464102373461719,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.2766,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.27465397312898676,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.2648,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.2836485128273025,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.2791,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.2800606391479556,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.2815,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.29273662811443274,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.2712,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.2545964249803096,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.2815,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.24789417856797713,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.2941,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.2914665164999631,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.2633,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.27057149718973056,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.2751,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.25674374763048463,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.2823,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.25772702723654506,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.2749,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.24489680979224027,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.2994,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.26163358768840334,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.2916,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.2489290270146054,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.2818,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.24961656456209338,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.28,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.2546100035881642,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.2922,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.25948212787907343,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.272,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.2691194770470863,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.2622,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.2533314257691081,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.2824,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.28180867773684193,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.2846,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.25936806240460514,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.2682,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.25761431642647853,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.2853,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2642765983827174,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.2824,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.27208401274911365,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.2925,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.2333095485975656,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.2869,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.25054392179220525,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.3066,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.273072705251065,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.2961,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.2452797129285561,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.2974,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.2533460361347234,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.2878,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.24182984852802714,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.2923,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.2543981428036953,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.283,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.24856504155977457,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.271,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.2527152685325113,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.2743,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.23279239086424372,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.2867,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.2494585846248131,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.2739,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.24628153034800226,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.2614,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.21803883370515265,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.2969,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2459190494567177,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.31,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.2435353366719146,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.2839,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.24386396757438272,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.2854,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.24680301311827152,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.2838,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.24746612650918282,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.2923,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.24069069900818849,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.2699,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.253363110188959,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.2827,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.24687935211368012,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.2771,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.2449319875824903,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.29,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.24534908713503742,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.2713,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.27076300330588116,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.2633,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.24310610625891896,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.2779,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.2611754051615096,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.3031,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.25939032261250994,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.2728,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.22908003655016146,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.2891,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.24738710584118273,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.286,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.2440437385159429,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.288,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.24588942632814767,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.2818,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.26444660061726133,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.2744,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.27506749930560526,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.2706,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.2504550862991011,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.2743,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.27613321611596353,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.2836,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.2340575479854184,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.2836,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.24256802566423696,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.2723,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.2788686025239232,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.2861,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.2342490206426046,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.2822,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.24097015193540455,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.2884,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.24938215571780833,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.2858,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.23704404414006588,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.2812,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.23937824265602411,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.2721,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.25700348076674906,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.2711,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.2528794375884632,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.2626,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.22639664992258293,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.2815,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.24492821663101624,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.2906,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.2614502142446894,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.2713,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2538181165204462,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.2708,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.23999858407533628,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.2592,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.23264860161713952,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.2679,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.23701035426085196,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.2736,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.26514796794508655,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.2906,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.24317923664598798,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.2663,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.2526369644085148,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.2837,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.25985677331284274,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.263,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.231058625186213,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.2816,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.23653336038634548,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.2837,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2411883330950155,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.3001,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.2494619209391677,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.2901,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.24138011154635494,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.2726,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.2511401566857059,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.2715,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.24817797418106516,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.2902,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.23400429768849018,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.2872,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.2475477143258035,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.2836,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2601989034370704,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.2929,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.2485537235618798,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.275,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.22809598497962352,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.2998,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.2606661272790962,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.2889,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.2567331826791781,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.2791,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.25275384802158624,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.2741,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.25078112386014756,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.2845,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.2573129887121542,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.2794,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.2544451707045002,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.2623,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.2590034630818414,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.2558,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.25885171440327365,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.2785,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.24548474851827062,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.2873,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2379898914703989,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.2992,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.23174076273729438,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.2837,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.24876972460000574,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.2984,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.23685125635475687,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.2684,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.2337850943807669,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.276,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.24797692753869988,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.2814,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.22114719013185008,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.3024,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.22273689661075738,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.2772,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.24168216210685675,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.2779,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.22759332414817934,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.2841,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.22270185195208222,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.2744,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.21820299676640342,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.2896,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.2381482450000703,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.2708,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.23481748806336852,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.2732,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.2186474974640617,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.2805,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.24797828324448973,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.2734,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.24115497064563773,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.3061,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.22720677326217484,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.2806,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.2439932474792294,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.272,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.25355916651359905,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.2811,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.2418710307500886,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.2895,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.2227303835798364,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.2683,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.22079762920986268,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.282,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2539999562198711,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.2765,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.22973001882440494,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.2867,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.23581606097877728,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.2584,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.23781632320101623,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.2974,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.22635999345071142,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.283,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.23708098132289954,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.2898,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2556260983991515,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.2908,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.23402156500974747,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.2887,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.2364254325884876,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.2845,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.23555874922336212,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.2914,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.21392038656818707,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.282,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.3667010615110471,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.2377,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.2463194503905259,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.256,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.22120460087514923,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.2498,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.34084289157641995,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.2633,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.3109441326567144,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.2438,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.261512766209975,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.2345,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.2538487844750554,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.2667,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.29328293325111066,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.2691,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.3069662418748037,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.2505,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.26043262431894315,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.24,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.24497917404650227,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.2304,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.2510955135543333,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.2342,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.3197443381353068,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.2554,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.2480389573857945,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.2567,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.23045572820183702,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.2419,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.24108174201356242,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.2264,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.24561325960276828,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.2613,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.2398369492373426,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.254,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.2337456617833273,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.2742,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.24116172650022338,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.2457,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.21959458435286966,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.2653,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.29012879674307673,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.2614,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.21119628326458373,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.2376,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.22766609418788736,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.2668,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.22168706349801773,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.2625,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.20770940573183627,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.259,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.22346079305716698,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.2453,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.2190255068794934,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.2324,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.21994827321588295,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.241,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.21372586443841546,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.2661,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.20906545522287423,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.2501,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.22150378217362457,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.2492,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.22164988834971538,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.2275,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.22437683217359333,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.2442,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.21225672899993928,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.2434,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.20329887472615682,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.2749,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.21288402775247608,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.2349,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.21009756616791928,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.2497,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.22284403646423817,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.2628,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.19372947582838151,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.2597,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.20261883539384104,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.2375,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.2200922414321404,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.2658,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.2001821172649289,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.2361,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.20799085256760808,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.2508,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.20881049261740428,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.2476,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.21466423061912066,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.2466,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.20130424079403836,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.2428,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.20597781127864637,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.2476,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.20394809594607283,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.239,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.20123605346944523,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.2501,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.20360832315132157,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.2397,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.18972387090612372,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.2274,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.19987618518685824,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.237,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.20761626116647594,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.2369,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.21159112583097106,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.2253,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.20636290847556915,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.2372,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.21400963257642938,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.2428,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.2090011907815471,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.249,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.20419376767603362,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.2481,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.209747356276331,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.248,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.20203173092254922,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.2317,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.20716307590815583,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.2433,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.21522775226685542,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.2296,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.2058780015384947,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.2361,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.20662679442700096,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.2231,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.1997598269491436,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.2521,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.19645858266174776,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.24,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.20171610232544143,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.232,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.19179478434722833,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.2589,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.19321050427316921,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.2287,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.20637633343265788,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.2448,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.20633831657631022,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.2465,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.19694967202989297,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.2357,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.2025260020892305,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.2561,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.19768181721847436,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.2379,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.20339847760379975,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.2296,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.20849081497229555,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.2398,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.2027637294772254,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.2358,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.194610346032025,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.2375,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.19272957915603434,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.2253,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.20650629310841714,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.242,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.1922714790427419,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.2634,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.19993871277682423,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.2508,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.19355337030652972,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.2593,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.21248529608533637,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.2285,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.2003289583028999,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.2331,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.20043237193357002,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.2472,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.20173196527273435,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.2486,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.2015564548559874,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.2462,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.20213657901806112,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.2653,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.20049372902526916,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.2148,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.19522811307753543,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.2583,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.19763446051213676,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.2823,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.19306526311094233,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.2564,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.20570548898880361,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.2246,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.20368101974792124,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.2544,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.20602375004737367,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.2406,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.18864994278105465,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.2399,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.20253426242744424,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.2581,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.19598662569747205,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.2542,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.21731229161930735,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.2407,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.1962090845733424,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.2566,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.19581825288160704,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.2467,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.2044064755908074,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.2486,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.19716039238382246,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.2478,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.20724460697580288,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.2229,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.20353901163323107,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.2362,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.20560981998106212,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.2253,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.20775919003927595,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.2385,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.2068233314618667,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.2291,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.19761555389504845,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.2341,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.19679412461029658,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.2525,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.20948683962188222,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.2433,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.19448292032768844,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.2511,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.19487499030408786,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.2452,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.19667072257816218,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.2486,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.2009632628801362,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.2415,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.20190053609527817,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.2439,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.20796603883638176,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.2284,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.2018897801453353,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.2246,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.2008251547289027,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.2505,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.19554016904194726,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.2565,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.19826238743481928,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.2513,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.20175639931725378,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.2664,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.18716468006580853,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.2679,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.19647688539679753,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.2596,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.20004045779209012,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.2427,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.20160572168178983,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.2469,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.19456116742069215,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.2544,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.19953669415305267,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.2695,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.183257060705743,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.2331,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.1899516604536447,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.2566,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.1988029964086575,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.2428,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.19624855020735563,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.2562,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.21761353026469263,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.2355,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.18861382520814907,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.2617,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.21010136902308957,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.2362,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.19275598004337147,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.2193,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.19604396851613587,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.2347,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.20142033307327956,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.249,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.19168534473373794,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.2308,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.2050829522005664,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.2395,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.19389070118239413,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.2569,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.20595171388643366,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.2162,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.19636893924600238,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.256,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.1906929123180251,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.2323,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.19393795873470246,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.2281,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.18801295979802116,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.2405,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.19849256268356227,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.2398,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.19423484971959334,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.2344,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.1973580766861525,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.2499,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.19666430358297135,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.2511,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.18837821527792017,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.2579,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.21503763102968917,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.243,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.20704839264261252,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.2396,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.1970149716988468,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.249,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.19849560163004948,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.2399,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.19908708036527187,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.2462,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.18932526566270785,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.2513,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.19333577745928046,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.2411,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.1905551995060665,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.2548,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.18674731245395143,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.2489,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.1933778098645824,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.2569,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.19537308548305882,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.2416,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.19288403405650642,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.251,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.19391885813679183,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.2423,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.19433165122614945,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.2657,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.18068497054316265,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.2545,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.2026342388614678,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.2372,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.18387446098316798,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.259,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.19052230738372686,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.2345,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.19260858387689783,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.2588,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.2003215931826391,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.2518,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.18468819522688681,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.2612,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.2094663989249005,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.2399,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.1892121068134879,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.2357,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.19588323872204244,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.2572,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.20703260263378648,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.2372,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.19708253012483218,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.217,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.1898713344181109,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.2535,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.19550835159009608,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.2349,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.18890506477851773,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.2569,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.1925189527550609,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.2489,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.193217891536891,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.2638,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.19905626985341301,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.2406,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.19372707111477547,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.2367,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.18812323245315193,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.2568,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.2007369707929803,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.2523,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.2001251598368832,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.2422,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.19782735121414002,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.2549,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.18906442011559516,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.247,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.19691627147790897,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.2642,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.1873514895915077,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.2403,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.19933891341813664,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.2161,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.19525326055597972,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.2491,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.20331722820105771,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.226,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.20186239896155617,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.2323,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.18779794467632802,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.2568,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.18787367168209793,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.251,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.19795329482724058,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.2454,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.18004820988616232,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.2565,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.19452816603959613,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.2439,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.1878265760448555,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.2626,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.19729062993457017,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.2363,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.1892516974937346,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.227,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.19191570625342066,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.2509,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.18579093178841943,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.2739,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.19819600421487527,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.244,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.19299703095027976,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.2398,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.19555609585356287,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.2528,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.2166415994064366,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.2317,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.19117954032046705,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.251,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.188470856278682,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.2566,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.1894159865500743,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.2434,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.19481275701035947,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.2429,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.19361696750014928,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.2431,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.19242558068480725,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.2538,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.1906061906420456,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.2405,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.18868997768354734,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.2406,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.1918252213244825,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.2351,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2037738553061667,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.2228,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.19213379881511286,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.2558,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.1898030645309521,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.2468,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.1840899890856985,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.2687,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.19636082263809468,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.2414,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.18909203693418877,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.2622,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.19505356324004056,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.2336,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.19461861950998344,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.2512,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.20025253723214317,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.249,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.19231978020668772,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.2632,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.18834092689944532,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.2445,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.19268955572280402,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.2728,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.2021853092288204,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.2432,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.21722177948364976,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.2625,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.19993341542122145,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.2417,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.18607785487476255,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.2238,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.19692152904677387,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.2298,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.18897999638283347,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.2227,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.1943827650296836,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.2248,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.192575861851457,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.2472,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.1918307280183174,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.2443,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.19015478706107844,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.2513,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.19764697821943192,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.2368,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.20239885361763105,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.25,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.19651810902509487,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.2385,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.19304501879227157,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.246,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.18893194079060402,
+      "learning_rate": 0.0,
+      "loss": 0.2466,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 3.2637958953212314e+18,
+      "train_loss": 0.3575047285450615,
+      "train_runtime": 27590.131,
+      "train_samples_per_second": 5.727,
+      "train_steps_per_second": 0.045
+    }
+  ],
+  "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": 3.2637958953212314e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}