diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 8.317266325237346,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.3312,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 8.503685183685702,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.3398,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 8.267228258185929,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.3229,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 8.304705759734484,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.3203,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 7.959174751274234,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.3233,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 7.344147802801531,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.2793,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 6.137223470967539,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.2326,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.723511522826667,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.2356,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.2519278922797215,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.1709,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.8136313286114527,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.135,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.8938369393138936,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 1.1727,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 4.442932798113606,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 1.1288,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.547686875816013,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 1.1135,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 4.592980537941857,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 1.128,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 4.160750840285266,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.0849,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 3.0186748788492848,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 1.0449,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.8010261499950437,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 1.049,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.4833267159358345,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 1.0152,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 2.030325121623024,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 1.0117,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.9221837800825423,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.9912,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.7175617276613944,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.9694,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.4713957095515322,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.97,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.3751538252858935,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.9081,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.2969975598257286,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.9503,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.413663138206846,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.9628,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.3840273409783774,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.9207,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.9321040946901901,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.9236,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.1630461865403963,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.9572,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.03852300382373,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.8911,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.9607527158672855,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.9259,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 1.0046864722277473,
+      "learning_rate": 1e-05,
+      "loss": 0.9151,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 1.005234026642856,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.9161,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.9354619510357154,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.916,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.9965003709993215,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.9261,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.9154261107569445,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.8895,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.9176036441871902,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.8875,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.9392125268383722,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.8984,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 1.1144279030261954,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.8938,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.9762114234981011,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.904,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.920809379086763,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.8754,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 1.0774837379656832,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.8517,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.8884900186442328,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.8602,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.8035404872305616,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.872,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.9473311762208649,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.8889,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.8978861694876803,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.8704,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.7975801931910698,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.8705,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.8665817573223,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.8798,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.8548778719826162,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.8859,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.7705911694456804,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.8878,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.7916918801397863,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.8705,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.9219525307998837,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.8729,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.7900076410625192,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.8553,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.8671687167118515,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.8806,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.7636959720748103,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.8607,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.8287856630839497,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.8664,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.8365610252126814,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.8616,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.7816049723265351,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.8754,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.824040926446627,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.8681,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.8037761163861038,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.8578,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.808857532134553,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.8282,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.8191653600133488,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.8481,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.8438855196788108,
+      "learning_rate": 2e-05,
+      "loss": 0.8849,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.9323339249061516,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.8245,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.8557662007128882,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.8461,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.8894816270205561,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.8552,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.9693039722317263,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.8581,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.9555699507625491,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.8338,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.9263584854405474,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.8527,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.8700246490205676,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.8431,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.924749258738039,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.842,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.7662503231571155,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.8347,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.8736041213613165,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.8508,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.9136475182554112,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.868,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7529935436907144,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.8496,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.9284679848364342,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.835,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.8535485123718598,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.8568,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.7792663800107292,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.8489,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.8188948378576872,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.8222,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.8612847468247322,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.8546,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.8603400359989549,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.8248,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.8765846229945935,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.8932,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.8128777536331232,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.8538,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.7600007122188788,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.8512,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.7818527153637945,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.8571,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.8394682414109575,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.8269,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.9581322569790248,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.8661,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.9622668319926433,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.8467,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.9837939485534648,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.8467,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.8661053751410883,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.8523,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.8711896080246645,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.8567,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.0169299769319962,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.8516,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.869821015573832,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.8358,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.149190736652498,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.8261,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.073267556788138,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.8537,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.9247411428927843,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.8583,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 0.8333848684794566,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.8485,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.0973404334725914,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.8791,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.8775089888303794,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.8385,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.1130667056915884,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.862,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.9409731043262902,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.8646,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.8559085460220172,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.8444,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.9243552729525513,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.8298,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.7560932267368317,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.836,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.8204773119548138,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.8494,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.8199751740435232,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.8405,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.7484441137218122,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.8459,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 0.8270815508662256,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.8363,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.8680807765149323,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.8356,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.839899276352793,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.8415,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.9104482104893366,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.8567,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.9081133318608623,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.8454,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.9254821539076908,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.8542,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 0.8901014042045796,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.8524,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.0237646619459515,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.8601,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.9084932319908384,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.8713,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.8431620906944463,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.8441,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.921488058238823,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.8425,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.8849836954042669,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.8407,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.8526492261676782,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.8362,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.0008857468356644,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.8718,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.8945172778780507,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.8405,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.0468191977827124,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.8608,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.8144539364691342,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.8493,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.8053859381337193,
+      "learning_rate": 4e-05,
+      "loss": 0.8376,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.9447401962755656,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.8587,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.8405969509764863,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.8508,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.9785827310911439,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.8493,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.0738170954616695,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.8555,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.7700758974899988,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.8328,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.045939311194805,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.8419,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.8541395051301053,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.8363,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.7941905877533638,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.8281,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.9033789733479913,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.8665,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.9627681480258935,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.864,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.9647705275853717,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.8529,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.8687377144103025,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.8444,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.0146946661527843,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.8524,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 1.0364123027279895,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.8436,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.8059383793089537,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.8473,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.9864889996293847,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.8351,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.7513213522038149,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.8343,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.8309633185737607,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.8435,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.7646923481273888,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.8338,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.9013950379981674,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.8366,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.7920416421361116,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.8492,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.7909636815534693,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.8287,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.8975961953793797,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.891,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.8364534391199038,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.8717,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.9618703728728887,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.852,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.764790848234075,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.8434,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.8100790147626982,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.8464,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.9224635948929525,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.8525,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.7414175713272108,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.8505,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.8144109663000135,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.8521,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.7784857953738806,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.8212,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.7881065579173319,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.8504,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.8612490946146781,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.8583,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.7905955386799568,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.8126,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.7921271026553273,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.8649,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.8404680091438314,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.8401,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.720216159156432,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.8306,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.9180310062537091,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.8382,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.6966586940306317,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.8531,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.7093789623781596,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.8146,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.7506508779886498,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.8461,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.6974955486368635,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.8559,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.7780276972225949,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.8532,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.7530333346994782,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.8718,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.765682608313324,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.8474,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.8042342725204465,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.8445,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.7911634545636456,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.828,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.8024653596283908,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.8539,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.8502779208425667,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.8376,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.7645599893324334,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.8071,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.7474978572892556,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.8319,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.7569592306941739,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.8447,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.6803070706395027,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.8505,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.7459464311955176,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.8365,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.6975269058777966,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.8278,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.7061312263744265,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.8354,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.7631710968561685,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.8594,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.7254009247885064,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.851,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.6783092328351052,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.8521,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.8153592080455676,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.8577,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.7678080883143795,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.8287,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.7166332600682185,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.8359,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.9559287726964756,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.8528,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.719067293755104,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.8402,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.8328719859023485,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.8514,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.8464170897665483,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.8482,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.7261586746362848,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.8157,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.8535749707452516,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.8459,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.8325754978237382,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.8312,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.8441927764521291,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.8591,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.8631361468267926,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.8293,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.8331111346699132,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.8188,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.8767768479143294,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.8336,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.8534244764779689,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.8428,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.8535117794661334,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.8594,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.7974539270688034,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.8143,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.8079361993191888,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.8388,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.6764011762888001,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.8077,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.8468007732993501,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.8441,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.7028572801724919,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.8332,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.776609217856651,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.8491,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.7077330094791761,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.8474,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.7603980254267301,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.8119,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.7536819353720889,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.8519,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.8066209145507391,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.8496,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.7326386031093653,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.8338,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.7456335390764363,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.8433,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.7606973125699132,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.8315,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.8376515785446722,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.8239,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.7752962319475218,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.8124,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.9632687908148679,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.853,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.8270654985846726,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.8457,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 1.0084454290674203,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.8345,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.752862501928751,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.8386,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 1.0064347752268803,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.8382,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.9207296515363046,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.8182,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.7200109216173287,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.8254,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 1.0071368232798814,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.8396,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.7675583336613683,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.8328,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.7585575477094199,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.8088,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.8759181866268663,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.8526,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.6662856752994805,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.8227,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.8436789942626702,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.8328,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.7148833153180533,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.8277,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.7326366128744768,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.8246,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.7620035044355852,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.8372,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.6829046445787172,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.8126,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.7337835458068518,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.8432,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.7394976194177929,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.81,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.7537851577415293,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.8191,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.7042334116134896,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.8428,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.7135620682147359,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.865,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.7289712784793861,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.8537,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.7185135585636402,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.8245,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.6738658794232055,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.8276,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.6558083506540897,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.814,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.6869719107262157,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.8262,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.6162628268246355,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.8437,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.6493917121775226,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.8279,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.6477211924923749,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.8155,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.6600192660069351,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.8473,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.7262579257120874,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.8043,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6813379102362634,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.8222,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.9769585566386558,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.643,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.7876307641330222,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.6117,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.9590650594301786,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.6195,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 1.0878027426569368,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.6107,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.7883494739761921,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.6316,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.939796577126792,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.5928,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.8298081356818249,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.5989,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.7189347647378177,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.626,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.797055082467583,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.6275,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.7735805752723832,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.6279,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.7844361106410405,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.6002,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.7260495203659698,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.6232,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.7210241237904027,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5839,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.6670723740397873,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.6061,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.732568942037398,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.6187,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.6878284493250366,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.569,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.6855129739900951,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5948,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.6993850540615021,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.6025,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.7468307990970481,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.5927,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.7246633279176229,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.6006,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.7223240880258415,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5952,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.7988598286826962,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.6188,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.7078697420427954,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5684,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.7563428757616993,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.5656,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.7155367918676898,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.6047,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.7752692436276457,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.5831,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.7698331149354595,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5914,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.7763918030745384,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5789,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.7244727567942713,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.591,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.708672478052607,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.5941,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.8505845807819812,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5751,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.7727163700437496,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.5964,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.692473285717392,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5854,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.8468748003458462,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.5754,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.6694605323787937,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.6158,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.7598683763671535,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.6088,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.7115914608913806,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.5839,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.6707544065178296,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5841,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.6734785229473688,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.5898,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.7009811332803,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.6048,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.6788003039053281,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5831,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.7997204910268249,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5768,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.7175332147852699,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5841,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.7709348785956001,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.6194,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.7949252655949829,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.6015,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.8181215740009784,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5982,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.646786032784436,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.6064,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.778163868046242,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5969,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.6658196904272867,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.591,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.718846739337091,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.6038,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.7117283717501496,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.6092,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.6392777032722343,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.598,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.7237257699080947,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5978,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.7008244441170146,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5959,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.6838965380278914,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.6061,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.7149346165175817,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.5913,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.6732943985835237,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.6036,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.6727587620895212,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.5916,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.6768332140369946,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.6013,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.705763882608435,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.6047,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.651148822763441,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.6162,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.7056241843268672,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.6069,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.7643589830226055,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.6214,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.6902041763741645,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5973,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.7439176498152483,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5978,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.6945143748785414,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.6006,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.6528923144887129,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5568,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.7877131187701758,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.5952,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.703836491533215,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.6168,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.7636411541359437,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.6063,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.6950284634633093,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.5995,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.7281763623525239,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5772,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.7659138533281721,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.6012,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.6028291546643537,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5701,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.6457049608556694,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.6276,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.7000823106603438,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.6053,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.7109846699418881,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.6084,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.6816505289612671,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.6236,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.7081748596100046,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.6173,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.6523930928428219,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5779,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.6730070191855079,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.5793,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.635504156591923,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.6039,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.6903158164614315,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.6003,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.6466296952690422,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5653,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.6588806604598376,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.6245,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.7358355768031299,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.625,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.6266895257784146,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.6018,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.7284183004707573,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.6202,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.7347196806045739,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.5877,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.7499609596514704,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.5979,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.7602005676245295,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5833,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.6654247622593918,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.618,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.7403056371374448,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.6156,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.6804542475048265,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.595,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.7499654115647842,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.6388,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.7024400069744512,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.5805,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.6383565241332759,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.6045,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.6415849478651233,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.5872,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.6976201396089605,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5872,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.6280454229362147,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.6027,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.7217684958897189,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.6006,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.7924224743330324,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.6213,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.692898210521517,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.587,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.7956314670087163,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.622,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.6353343639451924,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5874,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.7325652181998005,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5994,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.6481397916621304,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.6144,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.6452482316343778,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.6059,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.7270025208732472,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.6129,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.6622190431410326,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5959,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.757481789071585,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.6164,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.690587101375279,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5951,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.6934182413900896,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.5991,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.6537523563528974,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.589,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.6048472155211783,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5781,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.6429783290833531,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5945,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.6606118144085685,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.6036,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.621668324064701,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.6068,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.6227929851152384,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.6111,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.6581928326733394,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5925,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.6179070168062875,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.6054,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.658539070077783,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5742,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.6653207600445685,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5883,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.7247502065795902,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.6194,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.656532072293359,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.605,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.7405335339182697,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.6116,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.7068294260564405,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.6185,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.6852780728506497,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.5968,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.7284611777953899,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.5985,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.620274786446234,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5822,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.6695605598182882,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.6242,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.6942281660026427,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.6126,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.6550117539335004,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.6075,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.7311972322573091,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.6063,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.661139846234288,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5785,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.7128261099694109,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.622,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.6793087147505383,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5974,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.6563054228365046,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.6106,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.6469796801306348,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5995,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.6936722304772247,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5967,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.699231463938244,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.6297,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.7216515198446283,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.6219,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.6902025979168791,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.6239,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.7067418234420136,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.6306,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.6526611098932509,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5976,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.675947256646748,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.6154,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.6815138183967646,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.6085,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.6000441176773688,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.597,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.6953120199648481,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.6027,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.6801759353282475,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.59,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.6561858315656329,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.5802,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.7178797702838535,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.5979,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.6792200444506362,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.608,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.69190926421629,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.592,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.6669591631104451,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.6008,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.6762236348656292,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.6268,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.6639600100532357,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.6106,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.6300790883884702,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.5954,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.6520947800071641,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5864,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.626337807975485,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.6142,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.6306548650207181,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.6108,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5913821693111564,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.5899,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.6828314597723257,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.6122,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.6232546908890789,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.6015,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.6604556147377137,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.5973,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.6410637410327849,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.6249,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.6437029160699036,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5861,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.722449619552065,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.6204,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.6696065194482955,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.6007,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.5908269534298062,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.6116,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.6911292952610563,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5958,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.6154460374795604,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.5817,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.6290620637808926,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.6099,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.6187245737758569,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.6076,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.6181087108933123,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.5921,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.5884776334808645,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.6093,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.6132351965616768,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.5974,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.6363241466972342,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.6161,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.6165710064440201,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.61,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.6310620757006312,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.6139,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.6036874245977263,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.5871,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.5892180429229356,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.6101,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.615082283862219,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.6003,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.6343372267407972,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.6062,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.6426940155497148,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.5927,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.620706297462618,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.6092,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.6190165056213492,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.6199,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.6100211649082288,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.5981,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.6367674512252818,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.6048,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.6000896916042657,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.6122,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.6664917754242302,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5908,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.5907604460004822,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.6052,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.7215619308385518,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.6261,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.6179238051381747,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.5827,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.7143793274891916,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.6013,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.5934555947223571,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.6094,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.7220618414922209,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.6134,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.6188206532738442,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.5873,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.6812394531779388,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.5925,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.6172939040095606,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.5966,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.6029973075970619,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.594,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.610454807182669,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5958,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.6787737388123438,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.6204,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.6319044332622754,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.6092,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.6904054990980798,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.618,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.6763139391622814,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.6022,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.6285751335191883,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.6093,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.6574990888593066,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.6363,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.614628828505626,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.616,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.6444103665765684,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.6105,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.6466483242855328,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.6191,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.6555949019797458,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.5897,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.6763234924442286,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.5894,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.6910674660927704,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.6198,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.6240070299549759,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.6116,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.6026756327881783,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5987,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.6417997228724908,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.588,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.6255946601736184,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.6233,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.6564301025042713,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.6038,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.6368920639168719,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.6122,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.6371955999306548,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.6164,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.6688754447881288,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.5907,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.6019101009817878,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.598,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.6279213407150995,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5839,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.6177318494366222,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5967,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.7025015017731956,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5973,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.613773420872587,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.6135,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.6989687176942229,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.6305,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.6828031025052463,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.6003,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.6609208753811111,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.6124,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.6372518272212568,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.59,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.6522575385409364,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5982,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.6337228456809596,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5984,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.6101016318446886,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5897,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.6440673720837219,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.6102,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.6096396908654501,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.5903,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.6373841912191792,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.6143,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.6397558588536717,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.6159,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.6302276094306755,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5936,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.6280351434553685,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.601,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.6774767790273825,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.5877,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.6297506342326624,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.6119,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.6349836534032675,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.5975,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.6451142484064966,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.6308,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.6609804805002291,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.6011,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.636646881405976,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.604,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6455749153842469,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5882,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 1.0100569332132192,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.3622,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.6049139164472497,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.3501,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 1.4703922198462882,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.3425,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 1.0526685521160446,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.3521,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.7056449899584533,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.375,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.7333082045805758,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.3459,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.6999359764235211,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.3636,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.6475822646401135,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.3389,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.7223237997665618,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.3548,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.6851155250373244,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.357,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.7069252239012807,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.3453,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.6105704334343713,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.3191,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.6136133450796744,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.3456,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.6187279109217614,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.3179,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.5909476887649604,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.3432,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.5893444610196149,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.3576,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.6527622699156342,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.3529,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.5438539246227063,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.3288,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.6284153246664305,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.3408,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.5777864401365017,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.3236,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.5864554851156442,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.3214,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.5773633769439861,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.3291,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.5787239362968497,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.3153,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.5791256179629743,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.3299,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.5989501005189612,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.3458,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.5894615627972808,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.3375,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.5747610957543748,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.3583,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.5956920064823761,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.3404,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.5664343382224644,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.3237,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.5641512847622789,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.337,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.5961143491011881,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.3432,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.5764810080104836,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.3218,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.5830126842863328,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.341,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.5903770672179628,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.3509,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.6021100922421023,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.3544,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.611836977063549,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.3307,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.5822001067105352,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.3458,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.5615386255618798,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.3558,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.5652170410507947,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.3295,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.5561285055422648,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.3331,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.5586885498870963,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.3679,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.5828944967367151,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.3227,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.5562288700272994,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.3341,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.5364284596401395,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.3631,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.5698699819529844,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.3542,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.5534449904343098,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.334,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.5694066073750684,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.3396,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.6266017505306728,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.331,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.5282620174555605,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.3707,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.585628024044084,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.3524,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.5725176953886092,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.3291,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.5691713838437745,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.3342,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.5756021877464468,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.3569,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.6043457868166304,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.3297,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.5388969975586047,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.3551,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.6180873732915952,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.3459,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.574992441086384,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.3544,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.6026985053955528,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.3389,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.5669379717734958,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.3263,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.5522099580026995,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.3303,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.6041511239672157,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.342,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.5713914293123133,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.3501,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.5885823201256785,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.3537,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.5787643855587323,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.3552,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.5652712637528189,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.3382,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.5439653846932163,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.3376,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.574411016588304,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.3582,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.5336916711135775,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.351,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.5430638022657509,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.3463,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.566166892144922,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.3281,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.5409360501969959,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.332,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.5493221929586596,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.3365,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.5496767307444222,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.343,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.550230906802231,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.3371,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.5776871456285952,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.3346,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.5376990003416026,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.332,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.592300737195329,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.3391,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.5532145030361577,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.3406,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.5797491313963051,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.3597,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.5420978526764325,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.3482,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.571006542840677,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.3379,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.5545906257240174,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.3581,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.5457992215975833,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.3446,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.5420134331495257,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.3483,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.5695352920049744,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.3635,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.5631950106605483,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.3429,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.529901852583191,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.3626,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.5661645264658371,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.3583,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.559063755967253,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.3359,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.5334331733540602,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.3244,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.5468567916868022,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.3386,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.5844534171148766,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.3359,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.529900537257437,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.3478,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.5597769937870536,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.356,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.5432835573563028,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.3509,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.5371736253450327,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.3318,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.5311354850170172,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.3227,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.5523125027678303,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.3656,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.5727483910305652,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.3443,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.5427905796259423,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.3409,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.5467501352735339,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.324,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.5592885290078429,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.3397,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.5563421681592203,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.336,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.5808825911621796,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.3325,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.5622514805451074,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.3584,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.5761251328222391,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.3509,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.5469059886743848,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.3478,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.5565765639824831,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.3505,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.5705670885102199,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.3423,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.5396123811664599,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.3364,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.5271262079887092,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.3207,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.574932412352809,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.3525,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.5584164896132349,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.3542,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.5595697467689577,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.3326,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.5460059642836204,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.3494,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.556350174205973,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.3397,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.558580811585167,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.3382,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.5509196707844438,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.3478,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.5864006420831034,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.3194,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.601053756443427,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.3355,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.5984202967003752,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.3292,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.5925048669101722,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.3509,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.5381409050635372,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.342,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.5912530954288965,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.3379,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.575015732765815,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.3414,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.5593504753031716,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.3486,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.5800758139990531,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.3409,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.5724619847857167,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.3358,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.525760719759908,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.348,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.5969296070533623,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.3651,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.5471372884983361,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.3446,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.5423668294030063,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.3725,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.6064754596270374,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.3466,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.5722127367561147,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.337,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.5720081467474303,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.3504,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.5456755807481088,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.3385,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.5707780255931457,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.3471,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.544127730174404,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.3349,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.5502701303584477,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.3319,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.5475614030716827,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.3485,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.5327790223238197,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.3647,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.5483261235697647,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.3364,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.5754584476218375,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.3562,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.5366254477169853,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.3541,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.5318843532410411,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.3469,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.5562579310816708,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.34,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.5271525074231458,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.3344,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.5324804829703578,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.3392,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.5443601706958678,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.3408,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.555002585713861,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.3434,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.5441698785926171,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.3413,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.525555894184216,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.3403,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.5500724707450356,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.3529,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.5354543620152118,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.3358,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.5531630937196578,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.3247,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.5541596885808804,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.3309,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.5540150009354186,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.3405,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.5534756800015782,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.3489,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.5630360216485198,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.3392,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.5498389781949705,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.3348,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.5576823864202792,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.3553,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.5331597855875816,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.347,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.5454962238537152,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.3599,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.5617072704693394,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.3419,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.5611159579444602,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.3635,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.5382989121420156,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.3473,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.5665623535595592,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.3361,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.5324169265333576,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.3444,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.5323759958142846,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.3443,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.5466375126566442,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.3629,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.5463186605659818,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.3507,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.548354932926347,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.327,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.5229051359706978,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.3391,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.5553185433419713,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.3378,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.5303477968512158,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.3362,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.5465172012954476,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.3499,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.5712631967788888,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.3501,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.5238450338910507,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.3288,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.5374393954315376,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.363,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.5801300834912518,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.3521,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.5800105905045798,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.3279,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.5272422213108993,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.3332,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.5398871591523977,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.3604,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.5462834445754764,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.3258,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.5353077515826153,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.3563,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.5477571306713852,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.3562,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.551081847555596,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.3406,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.562159014543688,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.3443,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.5156048610801632,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.3352,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.5532043521331,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.337,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.5411076195813325,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.3572,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.5644656022224783,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.3478,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.5266803822053154,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.3504,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.5233251068366697,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.3514,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.5318591445000947,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.3575,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.5138788128345183,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.3567,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.5425541533859967,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.3359,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.5247845897514448,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.3392,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.5062829376176121,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.3514,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.5436002916098391,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.3388,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.539704638019593,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.3395,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.5418858866692667,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.3248,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.5405298425395892,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.345,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.5330624822557066,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.335,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.5379434556994598,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.3541,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.5479863841130828,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.3392,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.5473559272915602,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.3691,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.5597980193578911,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.3553,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.5286755272128693,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.3554,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.5228802265301856,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.3461,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.5260846315643228,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.3393,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.5539279333963202,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.333,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.5511018252493167,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.3325,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.5284796307166738,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.3584,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.5186899177973752,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.3392,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.5426654935917469,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.3449,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.5486168853938066,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.3327,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.5376499023802759,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.3456,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.549743023423742,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.3391,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.5459523852588982,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.3269,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.56854064241166,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.3228,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.5615130665998501,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.3425,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.5548867267608762,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.3268,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.5282853798544465,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.322,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.5719761586611201,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.3423,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.5348049522119066,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.3353,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.5723056967736931,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.3403,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.5516944309494233,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.3463,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.5288410692947176,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.3412,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.51956962181056,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.3382,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.5617151934649002,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.3339,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.5160561212870879,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.3434,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.5316236171563291,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.3545,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.5582613955361153,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.3325,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.5152990783800964,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.3611,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.5545781993512116,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.3652,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.5591563697235213,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.3417,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.5457842306967071,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.3153,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.5669790275117264,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.3399,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.5358002621494102,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.3527,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.5505015266766515,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.3509,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.5242607899932558,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.344,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.5054058128238528,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.3389,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.5676686843979326,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.3256,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.5371066883801024,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.3487,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.5521528397204916,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.351,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5513555797974388,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.3244,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.7315355881205179,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.1831,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.5776526719891492,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.1802,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.45520943492784705,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.1621,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.5217337484510097,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.1702,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.7503350732230867,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.1862,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.6489469299171278,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.1656,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.5566878351140722,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.1863,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.4684079901288976,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.179,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.5002123771403766,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.1782,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.49034329805185334,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.1796,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.5186452189845202,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.1846,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.45250978059298436,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.1868,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.4330026213690625,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.176,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.46063632732094434,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.1554,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.4638353366622346,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.1705,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.5283205535955559,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.1735,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.49754071692687596,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.158,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.4777174667423362,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.1663,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.4448047589213148,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.1565,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.45783728546700647,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.1834,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.44656755109416096,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.1999,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.44936772097069516,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.1685,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.4371128686508068,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.1739,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.45930835813841875,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.1664,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.43034587549867104,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.1784,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.4415744532331712,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.1667,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.4311442820068156,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.176,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.4537799145781204,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.1689,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.4318588805880447,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.1502,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.42821032128991493,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.1792,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.43647887564102644,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.1666,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.43245549497853064,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.1796,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.4232037501066364,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.1867,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.4168137263199884,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.1612,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.4298089176089513,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.1647,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.42645486122042237,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.1656,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.41820149310043736,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.1606,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.43828465564400637,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.177,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.4087901274434656,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.1677,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.43903603712779243,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.1965,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.428978999546326,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.1776,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.40582731176022413,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.1577,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.42836770946057534,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.1613,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.42554725707722635,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.168,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.41774916627452235,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.1799,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.4173160736625686,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.1698,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.43103675230185856,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.1778,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.42115053302647837,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.2049,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.4174984166812597,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.1633,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.44456496795155936,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.1865,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.4200239134293697,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.1559,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.4255818882045699,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.1822,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.43245035221051764,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.1872,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.4153599798005993,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.1668,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.42775196132306514,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.1849,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.42131793451701055,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.1711,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.4341820017360081,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.1794,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.42291790629204307,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.1541,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.42466703646681364,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.1632,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.40631018096780264,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.173,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.4273388479247412,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.1793,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.4175227983190945,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.1861,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.4157686585307497,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.1587,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.424043833955625,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.1665,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.42009330510167636,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.164,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.4261301000388626,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.1844,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.4130538650021733,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.1555,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.40814492821139825,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.1676,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.41744128169372585,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.1931,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.45028977758441563,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.1658,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.4110392091310635,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.1818,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.4228774811987171,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.1758,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.3997137937583266,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.1672,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.41383427275789475,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.1745,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.42093620574631707,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.1858,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.42126965530897853,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.1884,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.42685571203858136,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.1497,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.4205708940666879,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.1615,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.42276141710026427,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.1912,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.4187636954827499,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.1607,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.4389961757674339,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.1922,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.423071049582904,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.1674,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.42480371686839286,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.1653,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.422383177721803,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.1745,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.4380843247580319,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.1625,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.43838751322325237,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.1516,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.41842295172073174,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.1761,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.42162416144792464,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.1616,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.4130391917781226,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.1757,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.4076868382259196,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.1746,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.41129600770987673,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.1901,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.4147621705042196,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.1894,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.42425303589250496,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.1638,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.4279275913858474,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.1749,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.4226401456127456,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.1572,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.43902784338508133,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.1724,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.4109325578408664,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.1569,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.41546917406671163,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.1884,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.41190605905240496,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.1705,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.40905166794168324,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.1641,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.39820246933866116,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.1743,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.4101221785189451,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.1493,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.4186605543581948,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.1703,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.4010154864423942,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.1818,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.41095789334033367,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.1749,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.41610780645989187,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.1558,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.43187284524242653,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.1536,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.41094594450282157,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.1663,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.4191673079519829,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.1565,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.42490907341391354,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.155,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.3986721027130004,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.1854,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.41939737851586245,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.1571,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.41641397533597146,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.1791,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.42874039298615557,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.1493,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.41014356575471783,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.1619,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.3987132568895464,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.1546,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.4230356143195093,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.1818,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.41427412234889505,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.1619,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.40180497955651545,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.1775,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.4276180012046157,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.1573,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.41377432432981526,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.1282,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.4196547642919124,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.188,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.4263953791788068,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.1695,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.4202603372727617,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.1671,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.40982793115210686,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.1635,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.41227466268870583,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.1457,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.4166002944819214,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.1949,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.4251824992904411,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.1843,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.4032057143502511,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.1736,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.3990805291922586,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.1626,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.4031948391791632,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.1671,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.42335904821333226,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.1623,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.4067949670063935,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.1883,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.4091995819602747,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.1656,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.3990581233693573,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.1754,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.43032679532791107,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.1624,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.4015506875180742,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.1777,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.39927479244117137,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.1746,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.4131924531066878,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.1806,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.4028249159280207,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.1724,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.40841334813229446,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.1665,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.39816224191923183,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.1704,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.4285672933158267,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.1902,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.4231733284078953,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.1814,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.41522986380896715,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.1597,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.41328341892943277,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.169,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.40909031012493685,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.1613,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.4192183795247589,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.1488,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.4192838110868659,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.157,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.40104852899273485,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.1679,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.42998890663162914,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.1549,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.42443039468231464,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.1695,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.407080502205124,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.1415,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.4035410649391483,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.1455,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.4097362921098931,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.1479,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.4101696301977876,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.1779,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.4156020232011083,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.162,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.42751696720367854,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.164,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.4138188966063661,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.153,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.39920899316567926,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.1755,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.4107034049436486,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.1698,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.42945967726780165,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.1939,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.40563428217976943,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.1775,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.40657954311928446,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.1852,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.40475538485396334,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.1556,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.4146175318842354,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.1833,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.39759667509410335,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.1686,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.4222433477392952,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.1675,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.4032326993712908,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.139,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.4144013028489893,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.158,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.4297737611460044,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.1812,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.41461909360839083,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.1707,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.4158775211383008,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.1661,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.38729136497151373,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.1992,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.41272176142074335,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.1774,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.43274311882867805,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.1786,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.41518836002075044,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.1542,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.40287592987426163,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.162,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.3976471303013802,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.1523,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.40005470327618226,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.1476,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.4082041872144549,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.196,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.4203765927911999,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.1915,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.404727017459677,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.1635,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.39933280806193994,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.1771,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.3966694528633899,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.1823,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.41064772188581705,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.1586,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.4035995165275676,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.1933,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.41233659645496273,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.1501,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.40237184053159886,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.1586,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.4058379938416915,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.1665,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.40481011638482994,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.1701,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.4165397944180761,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.1737,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.41620090195446025,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.2026,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.4011343588172182,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.172,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.40583867399647205,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.1767,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.41037742571791497,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.1708,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.40122162731171807,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.1615,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.4095011488769105,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.1661,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.41492733510435503,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.1542,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.4092766025759163,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.1911,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.40790689468999514,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.1636,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.41234783545210996,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.1518,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.414483835774652,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.148,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.41976596228476104,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.1797,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.4047341460136128,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.1725,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.3944812761939313,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.1424,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.3920060867198696,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.1715,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.4103735641623261,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.1519,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.4041651793386375,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.1608,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.4026965130360659,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.1828,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.399697646363791,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.179,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.39791638757403264,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.1697,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.4080756628621603,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.1664,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.4044744600369016,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.1928,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.39004712643358347,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.1532,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.3956671695268231,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.1511,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.4092641422908669,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.1725,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.403711478906883,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.1588,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.39191796405425644,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.1462,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.40606656641812794,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.1703,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.40410525208239734,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.1471,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.40668043415111027,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.1562,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.4083910631028534,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.1657,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.4038481829037809,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.1618,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.40191091734991635,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.1649,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.39567888790486166,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.1482,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.408565309551293,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.1569,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.4154901198757358,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.1612,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.4086591310872957,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.1519,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.41361402589042284,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.1707,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.39412816610945145,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.1794,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.4095982497696529,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.1559,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.3920419603189226,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.1634,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.40303633214971957,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.1726,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.3948558298664014,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.1699,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.3731209262205102,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.1594,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.40048926244894606,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.1819,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.4012833330299973,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.168,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.39979141808864305,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.164,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.4059710925184172,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.1601,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.41931999931488634,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.1896,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.4006488300976749,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.1607,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.39647180048907577,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.1575,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.41035859837404626,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.1642,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.4040268263068401,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.1594,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.3898689696523461,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.1782,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.381311670832736,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.1599,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.44458439570460806,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.1133,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.40110603312876886,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.107,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.38025673791833176,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.1136,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.314141126037602,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.0967,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.2829448455474104,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.1154,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.2809005890474212,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.0832,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.286874442976991,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.0893,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.28890332875005337,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.1049,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.37543143336349877,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.1004,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.4024801026598761,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.115,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.39674240136146044,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.1254,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.36202815465796556,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.1092,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.3552394367687808,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.0994,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.31174279423636786,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.1099,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.26759783768091605,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.0916,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.27928180294134697,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.0846,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.28918015039537326,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.1135,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.3122015219414348,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.1095,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.289139849549047,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.0887,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.2910214567977694,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.0916,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.3155236044033572,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.117,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.28974762102604773,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.1108,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.2557168340919924,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.0765,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.28912318967379563,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.098,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.2716555463764011,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.1076,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.2977532342609161,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.1117,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.2971917154538049,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.1192,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.29901094241905585,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.1009,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.29152926116367023,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.0975,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.2824454719038769,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.1245,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.27391627991367506,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.0921,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.2666679227517246,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.0863,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.297715395769984,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.1086,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.28361066855232553,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.116,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.28255728995741497,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.1076,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.26870900894034033,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.0942,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.2707793522942157,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.0997,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.271999050015041,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.1155,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.2711673034310951,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.0956,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.26872146639645633,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.0912,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.2682902555607374,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.0905,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.27793071973674166,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.0954,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.25980018522601855,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.073,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.2857102452380538,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.1101,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.27578592223814596,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.0868,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.28041421000773503,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.108,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.28294950875477093,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.1188,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.2683811091831457,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.0849,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.2685545592329291,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.1133,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.262319439886969,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.0965,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.26952395618797004,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.0887,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.281717209805176,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.1027,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.2682940674837509,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.0929,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.2782597790287045,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.1015,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.2745731810252204,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.0979,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.27774303382765764,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.1028,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.2762709203286096,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.0998,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.28081089036351836,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.0942,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.27572426494480484,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.1192,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.268903514892095,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.0914,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.2682814073060583,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.0829,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.27562038602552574,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.0888,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.26236252523332804,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.0873,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.28681702491299227,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.1046,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.27651217839644254,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.1117,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.2748406654643484,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.1103,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.2669477220942348,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.0925,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.2687814231345847,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.0823,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.26647924028108166,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.085,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2627816503481526,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.0825,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.2734932465009017,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.0983,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.28577688476445684,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.093,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.2518332286084469,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.0876,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.2749014938712024,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.1048,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.29250305130351556,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.103,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.2748727995794473,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.0939,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.26717058626790313,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.0814,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.28316095253513607,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.0874,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.2795979693479132,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.1004,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.2657350928821658,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.0859,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.27105233418454944,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.1077,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.28080614939176174,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.0887,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.27876165838596295,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.0982,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.28088976533176685,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.0928,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.2803120487999832,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.0995,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.27486824259703496,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.0953,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.2745373276879817,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.0932,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.2697103579898317,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.0904,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.267004187216695,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.0822,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.2879405194242265,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.109,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.28132423158803965,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.1093,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.27800301179339704,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.0971,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.26689805433362807,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.085,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.27796633626861295,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.0909,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.28096751697148814,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.0932,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.27321073328970263,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.0793,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.26915391123807425,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.1168,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.26342348388623704,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.1014,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.26760029217036474,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.0991,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.27922767525120074,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.0946,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.3018387247001675,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.1107,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.2849802979103152,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.0968,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.2763209298628883,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.1154,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.26981599008886537,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.1106,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.26750102635972584,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.0896,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.27890881259936384,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.085,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.2700856228981793,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.0848,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.25482490428927274,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.0775,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.2756158500212126,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.1016,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.27032710176295793,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.0851,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.2717195094040777,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.1044,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.2728975650891002,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.0914,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.26658543254190764,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.0973,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.26446194637604054,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.0974,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.286128074164045,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.1024,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.27757241751724304,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.1048,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.2802871441595208,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.086,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.2669852234541463,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.0889,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.2725804900543633,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.1018,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.2624121875336081,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.095,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.2777983834180457,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.0941,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.27525851181937183,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.0865,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.26986688369132583,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.0895,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.28503850727165414,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.1063,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.2719819191609234,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.0889,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.27710060051276353,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.1077,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.2781474879742165,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.1167,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.28679254248816977,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.1103,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.2665477318800215,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.0912,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.2784994930413211,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.0984,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.27357106075127674,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.1009,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.26106936686134957,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.0955,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.2817102491890744,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.0984,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2585114571469364,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.0879,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.2614409528640467,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.0919,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.2793068306831759,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.0922,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.2688380059519528,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.0971,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.27125353643504174,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.0958,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.2740306982281169,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.0816,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.2832011422661239,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.0974,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.2680777650092846,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.1043,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.2771020888674819,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.0897,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.28763205271605446,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.1036,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.2943802037675627,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.1117,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.27891708786673164,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.115,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.28204532150144473,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.095,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.25808397234612784,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.0828,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.28149119769564246,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.1111,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.2741298609649574,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.1043,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.2744044358282597,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.0996,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.2740919648014183,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.095,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.2700509189031899,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.0896,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.27018287001601315,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.0936,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.2605665977330654,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.0961,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.2624820852548769,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.1034,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.2660273541433513,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.0932,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.2633171038718111,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.0929,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.261816680976403,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.0981,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.2701430075944942,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.0825,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.2622158289075276,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.0864,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.2839803535127592,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.098,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.2720488831226205,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.1215,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.25198561716989143,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.0746,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.27138258051495856,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.0898,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.2618918180198606,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.087,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.27700594118159444,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.0855,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.2732912263603751,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.1168,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.2984913292788446,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.0915,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.26395580624558834,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.0867,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.2626003368107808,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.0901,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.2594861186437239,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.0726,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.26849213814069,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.1049,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.2747923161134012,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.1263,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.2599847719076777,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.0744,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.26984606488436047,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.0851,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.27679051861585974,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.1139,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.26574143201577805,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.1204,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.26512472337861925,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.1136,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.2711920449621344,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.0933,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.2665604895798042,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.1109,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.267913199583169,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.1098,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.2699424156139772,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.0949,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.278708139027826,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.098,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.26425018689037,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.095,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.2776231970250769,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.1018,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.27202983799866043,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.0919,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.26594845264286493,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.0866,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.2881121374087442,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.109,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.27057089942180584,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.1059,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.26675881345399577,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.0821,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.27018285395072267,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.0878,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.27093368906632614,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.0863,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.27536911567180367,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.1061,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.2582383243027332,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.0739,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.26166693819645664,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.0902,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.2793262665875749,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.1153,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.27285165591685256,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.084,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.26940753436273657,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.0858,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.27236258275845326,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.0917,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.27894307291545906,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.1273,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.2683872405036478,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.0809,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.26739172411016043,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.102,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.25992288709097694,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.1063,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.27707411746592137,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.1129,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.2724245812762965,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.1087,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.27753157241557364,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.0862,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.26878631109578904,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.0966,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.26362496298418586,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.0892,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.28011148978732003,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.0982,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.28142879560947015,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.0978,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.26772502806156095,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.1002,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.2777506914957635,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.1112,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.2634541126307137,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.0823,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.26567395039442654,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.0937,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.2737530867894637,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.0984,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.2789260413295601,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.0871,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.27791670664485957,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.1129,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.26901144199358995,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.0961,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.27686288107152696,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.1172,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.277860273701867,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.104,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.27303344436748406,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.0977,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.25858607296569774,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.0799,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.2733588927800666,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.1212,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.276662283580977,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.0976,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.2784447883190996,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.1062,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.2768088589794918,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.097,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.2760549424429185,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.1066,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.26493881320953444,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.0884,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.2647833090388456,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.0989,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.26214642152725215,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.0837,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.26074407333100785,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.0845,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.2749972320985885,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.085,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.2674883511432337,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.0983,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.26639449450431607,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.0896,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.2739249898883568,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.0955,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.26289761290446073,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.0949,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.27596322540543833,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.1186,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.2803044432141957,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.1107,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.2727614854056601,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.0983,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.2755819038718652,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.1044,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.2738776368323585,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.0918,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.2805220387252598,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.0816,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.26238854627166086,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.0945,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.2556184840625959,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.0772,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.27108769870714927,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.0795,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.26633480804180304,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.0918,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.2681529283123277,
+      "learning_rate": 0.0,
+      "loss": 0.099,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 9.420188652677366e+17,
+      "train_loss": 0.41800092731410193,
+      "train_runtime": 13628.8158,
+      "train_samples_per_second": 11.592,
+      "train_steps_per_second": 0.091
+    }
+  ],
+  "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": 9.420188652677366e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}