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": 9.030930206387373,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.7882,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 8.938580193479318,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.8177,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 8.966091227784,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.7453,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 8.857564467197985,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.7712,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 8.217370982241206,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.7519,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 8.053704016053429,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.7574,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 6.4462369213292545,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.7601,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.862686447934622,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.7091,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.4157908045718983,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.6825,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.9220966381805593,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.6761,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.7221537105665865,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 1.6437,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 6.610803241683604,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 1.6821,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 7.043826419983568,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 1.6667,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 6.4739882857001305,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 1.6958,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 6.305653762045277,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.7011,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 4.579518104068715,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 1.621,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 4.039332537729107,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 1.5583,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 3.200443226148584,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 1.6141,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 2.307042963934966,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 1.5316,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.7394582387145245,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.5469,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.6157815657978978,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 1.497,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.8775635453146464,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 1.5272,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.8800676142409525,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 1.5181,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.7438307922047478,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 1.4663,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.3576893689429792,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.4558,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.205595936001091,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 1.4793,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.3041861637921048,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 1.4837,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.0603950957989177,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 1.5034,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.0187792267543652,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 1.4512,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.9419288343280585,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.459,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.8721184594337709,
+      "learning_rate": 1e-05,
+      "loss": 1.4308,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.8002609998020533,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 1.433,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.6956105050177884,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 1.4153,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.7974796192319732,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 1.3926,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.7856946424570612,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 1.4124,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.6904980859840827,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 1.4153,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6982912342719605,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 1.3835,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.7667290759765923,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 1.4063,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.6403974807538365,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 1.411,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.7668622641953771,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 1.3985,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.6496889197873907,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 1.4099,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.727852669292149,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 1.4057,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.6818522463086555,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 1.3851,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.5907700171274515,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 1.3903,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.7444540746051768,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 1.3557,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.6795999938163146,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 1.397,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.5886533520107822,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 1.3559,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.4855425436711486,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 1.3777,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.6451757428817858,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 1.3664,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.7231128362727212,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 1.3516,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.6855630603278045,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 1.3488,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.7004243234114527,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 1.3499,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.7101720620789739,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 1.3626,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.7334451259024422,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 1.3502,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.6500043239423613,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 1.3665,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.7680857411894683,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 1.3316,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.7139016453766541,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 1.3477,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.831054590899145,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 1.3493,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.8517451648978318,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 1.3261,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.9640839733947868,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 1.2952,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 1.0265071292485406,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 1.3163,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 1.2500200832439137,
+      "learning_rate": 2e-05,
+      "loss": 1.3568,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.8852877328921916,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 1.3416,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 1.2076243683252526,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 1.3329,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 1.2248590493557672,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 1.3284,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.8249246353508212,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 1.342,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.9934342521315579,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 1.2941,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 1.1276110893147107,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 1.3721,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.1733459420807038,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 1.3274,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 1.0622736011481624,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 1.3682,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 1.0545693696569536,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 1.346,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.9410736263506538,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 1.3425,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.1323301498099054,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 1.2938,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 1.1068741139308258,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 1.3278,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.2637319313427957,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 1.3094,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.9658097446160269,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 1.3436,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.329612497589456,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 1.3032,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.0487370877650488,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 1.2915,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.326750257043154,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 1.3253,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.1675734611731108,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 1.3246,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.2972652118813492,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 1.3558,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.2107517375816503,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 1.3251,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.4652276854465311,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 1.3233,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 1.0800072576945179,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 1.3157,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.9602128066801362,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 1.3094,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.3726219210909525,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 1.3321,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 1.26143437863382,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 1.3484,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.526592064602013,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 1.3237,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.3877280033177113,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 1.3084,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.2238167754658587,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 1.3132,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.8312786381745442,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 1.3302,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.1656176779895142,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 1.3484,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.4556249104346335,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.333,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.5460977727291745,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 1.3032,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.2141024212448126,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 1.3104,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.7016057400422202,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 1.3192,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.0046210317770883,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 1.3111,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.635928539163316,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 1.3165,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.1593211758434971,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 1.327,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 2.273669203090114,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 1.2922,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.4992825118806434,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 1.329,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 2.0692033404750094,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 1.3003,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.6068916513831748,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 1.347,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.9296565743758438,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 1.3434,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.581866823019858,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 1.3136,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.9022034403831796,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 1.318,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.7273955045758804,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 1.2895,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.7132206882820382,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 1.3101,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.399543246667716,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 1.3045,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.6591352451637373,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 1.3036,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.5244891042655777,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 1.2881,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.506595901967815,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 1.3207,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.3291219098590488,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 1.3256,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.6152293324173814,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 1.3055,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.118844593635542,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 1.3082,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.8007786467267684,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 1.2835,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.4292087416950727,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 1.2937,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.610913006544316,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 1.3128,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.5064794327223126,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 1.2677,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.4510369530805316,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 1.3323,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.2705393432777854,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 1.2957,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.8961262078047927,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 1.2902,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.1288649875012473,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 1.2708,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 2.119913551367303,
+      "learning_rate": 4e-05,
+      "loss": 1.3029,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.6863473660901738,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 1.2965,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.3893619316463628,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 1.3325,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 1.4686530718377315,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 1.3301,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.1874236590731746,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 1.2978,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 1.5167765361396797,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 1.3176,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 2.329619876689289,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 1.318,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.2945482808720368,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 1.3373,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.978241411105211,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 1.3526,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.9359433441221112,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 1.3104,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 1.4986717321307599,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 1.2817,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.5061690256341853,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 1.3233,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 1.8499918106494107,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 1.3289,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.2192095581270204,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 1.3197,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 1.7221495364300092,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 1.2849,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.0189822878628094,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 1.3246,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 1.380092946379827,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 1.32,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 1.5727418315176025,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 1.3303,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.231881826525147,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 1.2894,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 1.5556302909105124,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 1.2833,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 1.4868580658652515,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 1.2841,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 1.771839251036206,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 1.3468,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.3292027258007582,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 1.2802,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 1.8413062285219879,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 1.3315,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.5966187768991953,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 1.2693,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.8730229789692415,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 1.3269,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 1.4033070348919954,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 1.3324,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.6770152611076137,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 1.3009,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.8573439919171723,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 1.2888,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.6893876459654837,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 1.3117,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 1.1852167581887654,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 1.2915,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 1.7298249019984602,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 1.3225,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.0876231139182695,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 1.3126,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 1.9957548474567202,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 1.2878,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 1.475387872768275,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 1.2944,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 1.7670901903180614,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 1.2937,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 1.6812230934349437,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 1.3126,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 1.5020934935200967,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 1.2844,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 1.6011218177046427,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 1.2876,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 1.3447276228739748,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 1.2645,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 1.2933826880856527,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 1.31,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 1.621402303674452,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 1.2918,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 1.384939681882679,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 1.2931,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.7695428303152076,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 1.2816,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 1.3835649661780987,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 1.3034,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 1.262897077315001,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 1.2877,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 1.232158682005764,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 1.3056,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 1.3572291350203043,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 1.2709,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 1.4023326529822666,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 1.3004,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 1.0979879859699835,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 1.3146,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 1.4571974229288638,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 1.2946,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 1.216792586219381,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 1.2999,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 1.2294636010409346,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 1.3034,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 1.0998599936346862,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 1.3045,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 1.4393671708472808,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 1.2639,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 1.02755161873019,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 1.2624,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 1.3288454108521055,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 1.2912,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 1.2772520943193542,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 1.2727,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.9557031218294096,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 1.2714,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 1.135985004872633,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 1.3209,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 1.3413704344127502,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 1.2924,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 1.4895566184509081,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 1.2927,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.9020493884997116,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 1.2968,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 1.2076603305529012,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 1.3105,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 1.1411947415084296,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 1.287,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 1.205815768416514,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 1.2642,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.0662976553869938,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 1.2646,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.8929738409331912,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 1.3214,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 1.176843131461752,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 1.2986,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 1.650520888355397,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 1.2243,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 1.2343944205572708,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 1.302,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.566659487520427,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 1.2987,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 1.1154181085819548,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 1.2747,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 1.961219075650415,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 1.2634,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.9070478879081352,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 1.2675,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 1.747860387097368,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 1.2968,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 1.482246708631331,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 1.313,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 1.1803553471574262,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 1.2787,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 2.193210345212553,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 1.2587,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 1.2366266849659133,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 1.2901,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 2.4961419381156813,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 1.309,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 2.260412591898833,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 1.2945,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 1.5582840974909866,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 1.2941,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 1.4870512226852668,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 1.3564,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 1.1162112185065507,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 1.3068,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 1.7264388039058787,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 1.2876,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 1.0913954212862211,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 1.3168,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 1.5943590220293145,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 1.3091,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 1.3881195008822214,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 1.3325,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 1.1297617713063615,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 1.3055,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 1.573339225039335,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 1.315,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 1.1871050481555605,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 1.2686,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 1.3198542419829409,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 1.3158,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 1.434821048771116,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 1.3057,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.8720284815308159,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 1.2635,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 1.359540697993979,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 1.2622,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.9653381224839483,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 1.2728,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 1.2434684590104763,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 1.3092,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 1.0228083437803543,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 1.2754,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 1.0938862663293787,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 1.2742,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 1.088638537729834,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 1.2352,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.9095972419821755,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 1.2562,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 1.1958573552177587,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 1.2836,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.7227185723755084,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 1.2922,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.98004472521616,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 1.2974,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 1.4247488373695794,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 1.2969,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.7961993755963954,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 1.2771,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 1.3741962712889642,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 1.2722,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.8886186014211453,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 1.2968,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 1.2575272100330852,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 1.2763,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.7528541949976959,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 1.2753,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 1.1136546868825072,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 1.3009,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.8492791391141405,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 1.2998,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 1.1715834611565379,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 1.259,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.7779883224254815,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 1.2872,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.9285507092934518,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 1.2798,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.7859214143586222,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 1.3154,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.8882728248244741,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 1.2381,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 1.1873192746409427,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 1.3087,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.7726252997352165,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 1.2643,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 1.1497041971385586,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 1.2999,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.853849477160058,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 1.2949,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.7699588073194379,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 1.2796,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6521298519977899,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 1.2937,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.8511623203184888,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 1.2308,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.7309645909887456,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 1.2506,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 1.0434622833741378,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 1.2287,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.9315006801358718,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 1.2684,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 1.1147688928105384,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 1.2446,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 1.1671530502814962,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 1.2611,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.9195327979597557,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 1.206,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.7282114351168681,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 1.2343,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.7028312438607058,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 1.2358,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.5918226270346705,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 1.2472,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.49888682868730544,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 1.2185,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.5624264017516628,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 1.2234,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.5526211169832169,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 1.2392,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.8493079702365706,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 1.2449,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 1.2893287842234582,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 1.2662,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.7302765990444382,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 1.2423,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.486826505234556,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 1.244,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.6000733323030778,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 1.2554,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.64942587911501,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 1.2477,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.7258810780685105,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 1.1918,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.735683216002629,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 1.2569,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.718139476085674,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 1.2245,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.8415696901222136,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 1.2219,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.7895134971209307,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 1.252,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.7959553481167386,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 1.2466,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.8704214721093472,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 1.2472,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 1.2934531600397385,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 1.2523,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.6457442952217125,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 1.2589,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.5839763832946,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 1.2384,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.6292449800495077,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 1.236,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.7647194372308378,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 1.2431,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.8770108501820101,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 1.2428,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.8087693820086388,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 1.2337,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.9560094486031722,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 1.2646,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 1.3468805599069409,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 1.2099,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.6138015446307277,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 1.2911,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.8410397278976852,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 1.2659,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 1.3284665909370375,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 1.1888,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.6705086215979988,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 1.2527,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.8494410303973814,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 1.2142,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 1.170061922535945,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 1.265,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.8150777558064993,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 1.2016,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 1.2646795787211895,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 1.2409,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.8271611804655289,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 1.2322,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.46376088629961054,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 1.2202,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.8827279879118356,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 1.2607,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 1.4330958940407326,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 1.2449,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.7332280149797906,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 1.2319,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 1.141992205876103,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 1.2828,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.9754376140040567,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 1.2366,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 1.3240421880527125,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 1.206,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.6348348491534945,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 1.2533,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 1.142145761851467,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 1.2609,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 1.2675500010611944,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 1.2207,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.7434905757367913,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 1.2132,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.8066397342767041,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 1.2268,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 1.050411671287244,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 1.2701,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.9017137696998904,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 1.213,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.9893518236610617,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 1.2372,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.7648362943834413,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 1.2179,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.6524005907762497,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 1.2324,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.7625048284782779,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 1.2431,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 1.1456547186807795,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 1.2163,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.975260256661977,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 1.2268,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.7456986019659766,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 1.2502,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.8016555340956522,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 1.2406,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.8739784025241338,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 1.2447,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.6264319672225697,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 1.2724,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.533394251167908,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 1.2498,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.6039300304020853,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 1.273,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.5066854796062049,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 1.2007,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.5867863021273344,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 1.2422,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5974441674371264,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 1.2132,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.5670918858384714,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 1.2365,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.7166438364569826,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 1.2098,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.6036026781210585,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 1.245,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.5013015522641503,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 1.2595,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.5878062523271671,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 1.2097,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.4761664474776382,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 1.25,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.677186506528252,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 1.2251,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.6563382955085885,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 1.2077,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.6250762700511,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 1.2491,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.7869544529218331,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 1.2364,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.7211626697098928,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 1.2358,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.5728462162102255,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 1.2587,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.6100709401513006,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 1.2375,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.7774902581623357,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 1.1922,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.6987098374489625,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 1.2548,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.4833335813001157,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 1.2332,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.6825163129978069,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 1.226,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.6551509981784848,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 1.2573,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.7056060906380404,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 1.193,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.5525856479040531,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 1.227,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.5186936343628614,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 1.2301,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.6134649276625018,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 1.2553,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.5767475096967078,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 1.2535,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.6172064281775387,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 1.2,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.6120030408306654,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 1.2389,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.5449969973677251,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 1.214,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.5143422050337456,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 1.2454,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.5137619583143451,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 1.2308,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.783511043671025,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 1.227,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.7422527193088895,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 1.2222,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.6868989576622792,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 1.2503,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.8034417486084775,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 1.226,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.7912800390909331,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 1.2195,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.6306257150235524,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 1.2374,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.5796408374757922,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 1.2338,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.5803222162099835,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 1.2475,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.577384863160168,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 1.2239,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.7781047526954532,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 1.2753,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.772307045514868,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 1.2593,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.5713724829584839,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 1.2372,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.5399911029376778,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 1.218,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.6751863020551994,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 1.2291,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.6737798880189588,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 1.2238,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.5482620481472131,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 1.2633,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.7120743725947122,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 1.239,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.7218534087912991,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 1.1888,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.7408217276138015,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 1.2809,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.8385044482156256,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 1.2406,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.9083294674376017,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 1.2493,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.8778433609657174,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 1.2451,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.6942864567797696,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 1.2637,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.857023028764678,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 1.2267,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.7226076590351608,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 1.2234,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.747801308448537,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 1.2427,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.9530943891838936,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 1.2299,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.837007356876293,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 1.2614,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.6007069457976612,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 1.2724,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.7175865432354925,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 1.2304,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.8388456842592316,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 1.2359,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.5813311273891598,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 1.2365,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.6189836125587014,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 1.2098,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.4542606961682159,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 1.2448,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.6232311818529609,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 1.2121,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.6469137979026196,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 1.2433,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.4984409528302659,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 1.2752,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.610726188683452,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 1.2418,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.4481074588095154,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 1.224,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.521969749018527,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 1.2526,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.5645869445416465,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 1.233,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.5205576619461711,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 1.1867,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.5093767070981804,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 1.244,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.6621347627774146,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 1.2352,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.5268459158592109,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 1.2564,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.6536851953758569,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 1.2366,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.4865031006967166,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 1.2027,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.48774322900627454,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 1.2737,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.6059204857384378,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 1.2371,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.6031196116835728,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 1.2179,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.7051026350877742,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 1.2181,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.46223225819354996,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 1.2452,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.550055531332671,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 1.2264,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.7482351101209902,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 1.2319,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 1.0746186613408275,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 1.1885,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.8820581226243158,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 1.2544,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.547962163518805,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 1.241,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.5678977401674891,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 1.2498,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.8322285695929942,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 1.2451,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.7990492035073242,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 1.2459,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.6941612484044329,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 1.2284,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.5891926113670936,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 1.2705,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.5288462886191063,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 1.2209,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.5547117757822198,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 1.2284,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.4292735724147549,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 1.2428,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.5178329418177852,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 1.2268,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.5295633767911945,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 1.2452,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.632725768643043,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 1.2103,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.48004608780050195,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 1.2014,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.4897757143106434,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 1.2576,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.5840657965020724,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 1.2393,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.6319868856800691,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 1.2214,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.6365953647257562,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 1.2105,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.5050439534568512,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 1.1941,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.48229980158174673,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 1.2483,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.47412382167006906,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 1.2421,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.5422987334634256,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 1.2475,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.5921101643826208,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 1.2201,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.42603442626330085,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 1.2326,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.697812825132924,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 1.2399,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.6587318331350369,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 1.2621,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.6146148365172779,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 1.2097,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.5533548830914354,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 1.2351,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.4796704332124406,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 1.237,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.4556400910910856,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 1.2258,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.5367185177293067,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 1.2176,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.6240333345429688,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 1.2359,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.6518274752771063,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 1.2489,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.5469794856864028,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 1.2403,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.43050546374905047,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 1.2383,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.4929235210410244,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 1.2324,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.6157663406349849,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 1.2344,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.5319023531620397,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 1.2283,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.46678362951351077,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 1.2382,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.4587735345057269,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 1.2495,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.5398172407225594,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 1.1999,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.7652707084450785,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 1.2464,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.7477825496851623,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 1.2255,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.4784178301912035,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 1.224,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.5040251705664606,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 1.2121,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.4546806012806844,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 1.2581,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.6122361689585472,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 1.222,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.5359528566442109,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 1.2256,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.5633419017468977,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 1.2769,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.5375408270243126,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 1.2202,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.4670155669951938,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 1.2428,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.6744924829179295,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 1.2558,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.5711256001324754,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 1.2362,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.4937144003813173,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 1.251,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.5233545584654926,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 1.2367,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.9132204572388604,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 1.2173,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.8936534725513157,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 1.225,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.6192562238099169,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 1.2254,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.5800514634532443,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 1.2143,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.7247434287912944,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 1.2141,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.6100503162729299,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 1.2124,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.5545360242683155,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 1.2256,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.48431804696059333,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 1.2262,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.6466832143322753,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 1.2175,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.5987629209267068,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 1.2351,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.44687807685712777,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 1.225,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.6205259700291076,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 1.218,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.6733145046205903,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 1.2554,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.6037260519937844,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 1.2032,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.458365467075177,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 1.2186,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.4732666297780961,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 1.236,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.4508717333685152,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 1.2311,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.48984577905238713,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 1.2228,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.42707311944654464,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 1.2455,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.40601054959623684,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 1.236,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.42992193864929124,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 1.2461,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.42721757981095926,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 1.2431,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.5717649270776275,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 1.2324,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.46869710500865086,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 1.2013,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.47980895824626085,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 1.23,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.40511390716061835,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 1.2263,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.4688380109988703,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 1.2595,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.4087995215904389,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 1.2252,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.5198844622260129,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 1.1976,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.5820022862722293,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 1.2655,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.4021530991428825,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 1.2128,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.5878669549909395,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 1.2296,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.4518748857170137,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 1.2092,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.38118203882666984,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 1.235,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.5932416465337842,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 1.2229,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.40553891582558527,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 1.2185,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.6431583382106133,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 1.1711,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.5219877356582673,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 1.173,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.7769242281216409,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 1.1717,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.8056953961679806,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 1.1808,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.7149127616736755,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 1.1707,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.4066808139556304,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 1.1849,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.6575624533627618,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 1.1621,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.611037971116873,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 1.1652,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.5292024690960478,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 1.1576,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.5360280915056044,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 1.1694,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.5021495566791789,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 1.1868,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.5660298592833004,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 1.1709,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.4724565022200517,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 1.1353,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.5560385144034908,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 1.1971,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.43552488378439874,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 1.1741,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.5514833311885787,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 1.1519,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.4969572767182537,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 1.1594,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.5180670701773942,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 1.1926,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.569879692465984,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 1.1745,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.5676695508203213,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 1.1632,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.6132363968582042,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 1.1551,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.4782877026769802,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 1.172,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.5246983875751435,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 1.1503,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.41003452481113384,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 1.1887,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.47203820734054847,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 1.1573,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.422780407404314,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 1.1606,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.44105860356932475,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 1.1761,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.5715200624871289,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 1.1626,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.46802670634026283,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 1.1553,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.5155040660981186,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 1.1764,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.4359947276403873,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 1.1606,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.44486990236264506,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 1.1874,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.438565949158543,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 1.1655,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.5023019411223663,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 1.1741,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3580911648605477,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 1.1612,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.522904386967037,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 1.1668,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.5042086046672205,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 1.1821,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.42628462703059655,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 1.1796,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.5141107992736239,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 1.1515,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.39403942180673934,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 1.1586,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.48320586101934654,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 1.1574,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.4045844670716165,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 1.1899,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.42898921488656294,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 1.155,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.3812586375539997,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 1.1681,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.40927593563008713,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 1.1502,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.3606763607240093,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 1.1396,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3904650010534445,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 1.1802,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.37334610603293994,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 1.1659,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3736522112208408,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 1.1811,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.37012530912482516,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 1.1713,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.37789210800297623,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 1.181,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.3573020264603067,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 1.1861,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.37469534970998386,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 1.2065,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.37962548430714144,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 1.1916,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.3185605939001945,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 1.1603,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3961057121721894,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 1.1817,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.46672373587928484,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 1.1748,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.3727388122155398,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 1.1909,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.4459460163390646,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 1.1744,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.37091099326907956,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 1.1591,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.44199303730605327,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 1.1568,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.41407773846220414,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 1.1461,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.32929030790678326,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 1.1628,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.43804643533986387,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 1.1744,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.44376917938095406,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 1.1505,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.34865127116009104,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 1.1808,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.6618146527273452,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 1.1818,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.42807616831587075,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 1.1788,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.4772612886390102,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 1.1501,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.393940606295849,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 1.1722,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.5089583278423777,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 1.1578,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.5234297949095937,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 1.1711,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.4385404208028833,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 1.158,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.4456919295321012,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 1.2,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.47445692229351716,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 1.1717,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.5918852539830328,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 1.2225,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.492293000390233,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 1.1828,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.6798885801710798,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 1.1745,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.46257368896351386,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 1.1207,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.5611228774215784,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 1.169,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.5158849026630361,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 1.1869,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.5905934587792085,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 1.1694,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.43903955788223414,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 1.1825,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.3969672359009106,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 1.1505,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.4854223911432243,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 1.1856,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.484425705725167,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 1.1704,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.5011454867161532,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 1.1845,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.4435965164428659,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 1.1474,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.5163195594663751,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 1.1943,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.3666065734337421,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 1.1402,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.5062582054554314,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 1.2066,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 1.8211160202700492,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 1.1952,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.4950500812439294,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 1.1453,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.48018626471931425,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 1.1489,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.5127925391022496,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 1.1823,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.456675863952111,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 1.1801,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.4411963467102283,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 1.1628,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.4028997796829533,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 1.1687,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.45231484339020245,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 1.1615,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.47737101000004356,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 1.1814,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.35823609318931293,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 1.1673,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.4596458715858679,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 1.1774,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.37331718313435197,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 1.1588,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.5086493602669719,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 1.1594,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.39436487025222017,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 1.1733,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.4827357699856343,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 1.1723,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.31954397540796253,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 1.164,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.461675152042425,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 1.1669,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.39704127935903105,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 1.1752,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.4236393819586705,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 1.206,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.36118861111529793,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 1.1697,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.4242667383001979,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 1.1554,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.4400423383178684,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 1.2057,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.4617821152963008,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 1.1819,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.399042010169328,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 1.1856,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.5042982214740817,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 1.1714,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.36692482723994363,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 1.1693,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.4897259101691858,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 1.1672,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.35755522422861796,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 1.1629,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.40163878282725496,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 1.1805,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.4808311820254564,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 1.1759,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.3561127704072733,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 1.1629,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.4631217402399673,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 1.1603,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.4322374444527544,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 1.1751,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.34462723445532967,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 1.146,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.515295143953138,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 1.1566,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.4918863421310586,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 1.1674,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.3414758044026577,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 1.1674,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.3770280794057243,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 1.1706,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.39174250070012356,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 1.1981,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.3562131599197903,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 1.1686,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.38521794700600615,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 1.1776,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.3498545405355031,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 1.1982,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.45394295813569036,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 1.1667,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.38814656646784573,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 1.2101,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.4455401525993896,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 1.1509,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.4438793240413957,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 1.197,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.41584987308395616,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 1.1994,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.48861426393344465,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 1.1409,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.4344298106995791,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 1.1751,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.45357070847961567,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 1.153,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.41161837411256896,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 1.1495,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.48010024725365447,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 1.1738,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.38869060535626715,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 1.1666,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.5090287467608208,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 1.1452,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.31444031043254267,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 1.2115,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.4512751903638904,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 1.1667,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.39645889769014636,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 1.162,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.39882895613625974,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 1.1754,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.4130736972690018,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 1.1872,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.39152713554415725,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 1.1975,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.40788140204616913,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 1.1346,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.36415141201185935,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 1.1659,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.41890686520066306,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 1.184,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.5740834899038041,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 1.1961,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.4118600477983748,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 1.2071,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.3544012062646382,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 1.1264,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.3949824814878141,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 1.1849,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.42850261988420135,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 1.1666,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.3486097549385856,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 1.1774,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.607670537747197,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 1.1921,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.39931001523359166,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 1.1965,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.32797441782515996,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 1.1749,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3721719752094917,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 1.1704,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.39661390990087986,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 1.1697,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.3447847066031941,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 1.1534,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.4031375548287242,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 1.1988,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.4368420399781265,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 1.1505,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.3283400048076068,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 1.2002,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.4798134277250813,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 1.174,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.34210084472454555,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 1.1429,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.4112974665473675,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 1.1496,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.4075984221575576,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 1.1477,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.36592699844318116,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 1.166,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.437523574906272,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 1.1859,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.324753191346271,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 1.1876,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.4210232082306323,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 1.1762,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.3635511172182,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 1.1705,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.3499081284546352,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 1.1622,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.3840767647715127,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 1.1928,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.33968852229194485,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 1.1548,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.44020950678749904,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 1.1848,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.3518840704097709,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 1.1731,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.38571969155187036,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 1.1568,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.3282612406953876,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 1.1663,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.416665953852374,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 1.1863,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.34028855499734173,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 1.1652,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.45217762492612984,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 1.1361,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.3379241980070863,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 1.1861,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.423726638549872,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 1.1697,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.33544182277271056,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 1.1702,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3268897107577056,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 1.1634,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.3880835238489345,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 1.1814,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.33813763883191933,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 1.1516,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.33285920314526013,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 1.1605,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.4465983462175343,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 1.1433,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.3323353669263386,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 1.1491,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.47963472259352924,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 1.1705,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.34809799864079066,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 1.1368,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.42517124513509086,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 1.1979,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.4486733187006605,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 1.1723,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.3459337276247137,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 1.1776,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.40289942638367904,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 1.1785,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3748277358629111,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 1.1566,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.32263816956940866,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 1.1571,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.4790635790290561,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 1.1672,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.3181117878609447,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 1.1447,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3926121842068166,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 1.1878,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.38818434645876826,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 1.1838,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.34985910937174536,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 1.1699,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.35705586441057124,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 1.1936,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.38196735561741874,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 1.1776,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.3590140634655532,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 1.1619,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.3951693374696341,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 1.171,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.34614153156705807,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 1.1643,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.32568452610875137,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 1.1977,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.388478994222827,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 1.1482,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.3591209184395121,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 1.1912,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.3517943313013251,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 1.1497,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.38678546656440477,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 1.145,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.3238350942610266,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 1.176,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.3924203421219756,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 1.1662,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.3398822995055771,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 1.1645,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.36175825327666244,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 1.1746,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.31235754082876865,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 1.1542,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.3260099993265514,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 1.1749,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.3538539424286909,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 1.1543,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.29517438178107713,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 1.1627,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.43580526136249487,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 1.1528,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.3572076277166705,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 1.1698,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.405489421394285,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 1.1562,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.4610537182384359,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 1.1526,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.2859464810953611,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 1.1783,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.3887314259379158,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 1.1481,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.33131829205607166,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 1.1225,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.30537789386051334,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 1.1642,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.31770202601127406,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 1.157,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.33857797160745196,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 1.1965,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.3327620341751812,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 1.1699,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.2692008694398144,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 1.1672,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.3483492412835819,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 1.1685,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.3473811667717475,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 1.1636,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.3604060495743634,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 1.2006,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.3395783090737618,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 1.1664,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.361072783613949,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 1.1521,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.25867594528570015,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 1.1992,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.32410985279046395,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 1.1786,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.413469911239968,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 1.1131,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.37615437256167744,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 1.0943,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.5289649592833459,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 1.0952,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.5354754755934071,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 1.1293,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.4006161578140124,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 1.103,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.407851541548663,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 1.1192,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.505380269291913,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 1.1381,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.4433515090294291,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 1.1044,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.46040988263248195,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 1.1108,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.4157116327599922,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 1.1079,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.5002309412086695,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 1.1025,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.3704749299710546,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 1.1113,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.4750942187840581,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 1.0852,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.548340110448322,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 1.1179,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.41984592085165146,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 1.0642,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.5333687945398132,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 1.0734,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.5905772521581814,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 1.1396,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.44793645497081175,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 1.1465,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.4794396840083238,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 1.1091,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.45278470315511676,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 1.1186,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.47081540925869597,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 1.1448,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.3847193744269383,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 1.0942,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.5645006183302868,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 1.1019,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.3669104399688113,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 1.1022,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.5796454077135389,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 1.1209,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.3827464665019322,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 1.0958,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.437057062024851,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 1.1087,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.4301891586042612,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 1.1321,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.3326517351727638,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 1.1229,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.4210286070012397,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 1.1109,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.31190185484881566,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 1.0954,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.3617791348027955,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 1.1132,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.356890754787309,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 1.1034,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.30696518925773425,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 1.0821,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.40667748229713285,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 1.1268,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.29019987070890657,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 1.1016,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.3459236591482758,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 1.0667,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.2863085274440821,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 1.1102,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.30322459426943504,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 1.1133,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.35428995399409746,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 1.1245,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.29781241018708027,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 1.1292,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.3697495846922897,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 1.1055,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.3436374154398403,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 1.1345,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.3233537319361711,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 1.1494,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.3106287286803188,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 1.0946,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.3483334627782534,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 1.1221,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.34603532255607217,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 1.1237,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.3214560521358444,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 1.1012,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.37518719778737725,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 1.1276,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.3088160397812162,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 1.1183,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.32223402593070616,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 1.0916,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.35196923972357175,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 1.1239,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.3197312405582444,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 1.1022,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.3419426457474221,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 1.0982,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.31306489768210455,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 1.0908,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.32133753918588653,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 1.0853,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.3508601547260109,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 1.0927,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.3372230021300127,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 1.1003,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.2994801876309547,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 1.0795,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.2881561142639098,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 1.1044,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.33667886247031853,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 1.082,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.32056968724214857,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 1.0998,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 1.6035793877594893,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 1.1433,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.3286719596070736,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 1.1079,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.3053011258001152,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 1.1062,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2850291173157163,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 1.1199,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.28985076973058793,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 1.0988,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.2947089162513779,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 1.1587,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.3104370120426513,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 1.1165,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.28410173531084904,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 1.0562,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.3255373523297444,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 1.1075,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.32681460283257485,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 1.1576,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.33053390241857905,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 1.1217,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.35729901466085734,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 1.1098,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.337103036511329,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 1.0949,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.33380711748036773,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 1.0847,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.3419358990647639,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 1.1322,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.28721967723731323,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 1.1004,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.40080608145079877,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 1.0753,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.3421286726865853,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 1.0809,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.3614766666307591,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 1.1006,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.3777921265108913,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 1.1179,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.338247784608439,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 1.102,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.4005797192787366,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 1.1388,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.40281913386673235,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 1.1141,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.33324026764028775,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 1.1038,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.3402115088515624,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 1.1037,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.29904841615685496,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 1.1161,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.3321462896804648,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 1.1101,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.26604832923015165,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 1.1024,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.2744667176167818,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 1.0674,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.29755356402390837,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 1.1538,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.26064728135150006,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 1.1111,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.3351144750398668,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 1.109,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.34797069478399445,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 1.1336,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.28113319650055363,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 1.0932,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.28716262480180693,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 1.114,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.31994918277426504,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 1.0952,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.3036641398553372,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 1.0954,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.2837518896471969,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 1.0809,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.3144996226649063,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 1.1345,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.29647536639742317,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 1.1008,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.2978803461228483,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 1.0901,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.30489234453233366,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 1.128,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.3134225162044821,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 1.1018,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.25878053637379006,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 1.109,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.30331442919695145,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 1.1425,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.3064198836659915,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 1.1184,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.27123183033490983,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 1.0832,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.27956943260585226,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 1.1149,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.2892284640867995,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 1.1158,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.32736753796106666,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 1.1154,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.27116822415666547,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 1.093,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.3680843336192271,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 1.1144,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.29014820787890466,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 1.1277,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.2692122414335499,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 1.0804,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.32622664897815534,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 1.11,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.31373125491733994,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 1.137,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.272443017922467,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 1.1462,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.33369037755013237,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 1.1326,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.295061925863504,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 1.0928,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2573352145145659,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 1.1078,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.2901849983989046,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 1.1502,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.29731123569589907,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 1.0912,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.2550509302519383,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 1.1293,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.33428822163791083,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 1.0946,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.29163823133066524,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 1.0912,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.28335434188917596,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 1.1135,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.3260754852579643,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 1.095,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.2984641417182733,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 1.1389,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.28469352632154293,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 1.1176,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.2713417851869831,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 1.0994,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.2779101848801293,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 1.1419,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.2736883750297168,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 1.1118,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2477874897381302,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 1.153,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.2841954600586543,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 1.1144,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.27279615637660604,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 1.1254,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.2659323543385336,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 1.1065,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.2866556661590642,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 1.0918,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.2859955779098265,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 1.1126,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.2923871708882322,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 1.134,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.2692538249930711,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 1.1563,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.5046995347453839,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 1.1022,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.24316476315893795,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 1.1309,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.312126477706522,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 1.0738,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.2774766137873788,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 1.1237,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.34248239087953697,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 1.1202,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.26537883294236136,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 1.1151,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.31198361519587187,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 1.1261,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2923205262599721,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 1.1185,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.2607970490427802,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 1.1219,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.2888680849007242,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 1.0853,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.2790107232384785,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 1.1089,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.26562205832168073,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 1.1058,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.3232261702535595,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 1.1111,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.24714741778738433,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 1.1199,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.2975303578689533,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 1.0956,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.29963435429836255,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 1.1333,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.29327262361114337,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 1.1162,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.2820095007048355,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 1.1117,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2869413338354566,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 1.1207,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.3084675715673602,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 1.1284,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.25516393147418553,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 1.1014,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.32255392654840886,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 1.1225,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.294000151792705,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 1.1374,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.33123210145421655,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 1.1454,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.26918318409768194,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 1.1136,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.2551938640318603,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 1.1287,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.25606767815995946,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 1.0841,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.2855900768161887,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 1.1044,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.2763464726539366,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 1.0978,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.2657103337952839,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 1.1225,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.2785154114764045,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 1.0778,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.2610382369794581,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 1.1134,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.26469222779138196,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 1.1192,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.25132244399174275,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 1.1261,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.25808808727554067,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 1.1205,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.2505956352748053,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 1.0972,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.26114017423876834,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 1.1019,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.27326293593459333,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 1.08,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.30471805339529595,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 1.1469,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.2721101074216335,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 1.1175,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.2530590407925833,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 1.1124,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.24313282138594217,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 1.1013,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2771542972431273,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 1.0678,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.2824922433946152,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 1.1316,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.2837513076322965,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 1.0927,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.27487409448262745,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 1.0973,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 3.8089788566742424,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 1.0815,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.3368805190595793,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 1.116,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.2981925051215732,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 1.1525,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.3086808141459779,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 1.1135,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.27879318781134815,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 1.1018,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.26759194998142916,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 1.0851,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2856240031192073,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 1.1078,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.3001666751773145,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 1.1173,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.27326459885038024,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 1.0943,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.2650323216609499,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 1.1059,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.31359689366238636,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 1.1226,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.2980749399840304,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 1.0951,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.25085581222617714,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 1.1088,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.28000962165918425,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 1.1157,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.29901550987539266,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 1.1134,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.2685202790446304,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 1.1152,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.2586325904805729,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 1.0902,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.3017312901189619,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 1.0944,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.23047471743518,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 1.0883,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.24619150214661523,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 1.0964,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 3.1338920670412604,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 1.2441,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.26971421798326584,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 1.1034,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.26544975120050945,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 1.1259,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.25346419201443593,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 1.0887,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.2897938499901517,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 1.1019,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.2829504585050015,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 1.1102,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.27521280233467793,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 1.1236,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.2620365862441391,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 1.1347,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.25240030761711957,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 1.0902,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.2600825749326532,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 1.131,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.23791796474204538,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 1.0818,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.24028486863455678,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 1.1043,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.24308077568035472,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 1.1438,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.2308878691256519,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 1.1142,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.37986614802830904,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 1.1228,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.2483063303022872,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 1.105,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.2278426558822412,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 1.0703,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.24752322913718602,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 1.1095,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.2572802730634506,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 1.135,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.24952668040409945,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 1.0776,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.301180460949787,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 1.0935,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.2586292770456293,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 1.1611,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.27196401190887437,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 1.1204,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.2702496531373556,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 1.12,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.25987026918022227,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 1.1227,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.25354888867477876,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 1.0956,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.2504097028295482,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 1.0752,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.25787270986092903,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 1.1158,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2484973185238423,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 1.1147,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.23607095664667202,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 1.1149,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.26032096341910005,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 1.1212,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.25985748864285735,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 1.1124,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.2622426172412306,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 1.0951,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.23514723936512288,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 1.1299,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.23599190091543817,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 1.1023,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.24730715140732384,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 1.1136,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.26939344081840605,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 1.0932,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.2415334886036034,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 1.1334,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.24658210489419374,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 1.0853,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.38638275094891106,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 1.0853,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.2855773548159178,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 1.0791,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.2878973832638627,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 1.0629,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.3898771756860018,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 1.0517,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.3890873629362933,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 1.0468,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.322688163842315,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 1.0832,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.3504514116669564,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 1.065,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.32892497742266485,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 1.1017,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.2983692018106379,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 1.0401,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.30216133329429157,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 1.0654,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.33132682353705706,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 1.0761,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.37959777702647035,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 1.0383,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.31650828862633207,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 1.0756,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.34260688844044884,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 1.0532,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.31470129346282005,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 1.0586,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.309803207955131,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 1.0737,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.289440009465605,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 1.0578,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.285923509763534,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 1.0858,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.30420686450121526,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 1.0846,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.2853552403999562,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 1.0736,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.2803114234712569,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 1.1147,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.27792452946249957,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 1.0882,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.2834149612575316,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 1.0643,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.2986727812402206,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 1.0672,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.25612924638899437,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 1.0709,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.29284237178497485,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 1.0609,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.2779936422265184,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 1.0729,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.7668815705645337,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 1.0709,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.30591921429037666,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 1.044,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.26052993254626694,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 1.0833,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.2790045822539721,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 1.0402,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.3015611341075878,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 1.092,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.29591844776579085,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 1.0111,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.26148964386428136,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 1.0683,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.24829610097776472,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 1.0828,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.272091068831927,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 1.0767,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.27226466766539187,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 1.0599,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.2469826117564637,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 1.0604,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.26469371443653456,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 1.0653,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.2587647612214328,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 1.042,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.23925393089076064,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 1.0465,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.23393462316332436,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 1.056,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.2564696305379957,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 1.0662,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.2833192135078331,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 1.0614,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.2749220340290399,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 1.0791,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.24450409213821717,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 1.0559,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.24771143692188444,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 1.096,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.2617778835209913,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 1.0935,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.24617206665471938,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 1.1275,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.25321258681333175,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 1.0877,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.24510694007927514,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 1.0392,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.2331982733594197,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 1.0956,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.24338299576562247,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 1.0745,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.2592374526657989,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 1.0518,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.24341580963642906,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 1.0495,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.24571520756272228,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 1.0612,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.24621609784320964,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 1.0834,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.28052849436652094,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 1.1144,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.2811111776302061,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 1.0328,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.23428186459053238,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 1.0461,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.24462698347472184,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 1.1123,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.24297198289164113,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 1.0376,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.26109541749660076,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 1.084,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.2774542962197102,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 1.0423,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.2387796064154447,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 1.0445,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.23604760430438806,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 1.0677,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.2485243492791167,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 1.061,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.24932260575928517,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 1.0621,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.249500345318112,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 1.0722,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2438461991132367,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 1.0393,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.23962957010394967,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 1.0719,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.2536049620320814,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 1.0573,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.21966129846693402,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 1.0639,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.2224357376214079,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 1.0965,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.24471306952933922,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 1.0185,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.23736332669805865,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 1.0763,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.22553987826302482,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 1.0684,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.23191113786807818,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 1.0852,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.22764535605324102,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 1.0628,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.24627976608796948,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 1.0622,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.2268394379675838,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 1.0547,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.2393226703112864,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 1.0665,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.22831294409774058,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 1.0784,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.24091048844028476,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 1.03,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.25672223860583543,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 1.0983,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.26318870954725093,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 1.034,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.2330491415262609,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 1.0752,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.23502097310052666,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 1.0375,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.2466564712160434,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 1.0209,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.25322110860228875,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 1.0483,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.23632170901953647,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 1.0656,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.24177163423907286,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 1.0634,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.2479573937490817,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 1.0261,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.2403618442268095,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 1.0384,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.2358433936328453,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 1.1042,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.250525187820149,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 1.0729,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.23266363146077715,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 1.0534,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.2342554901752827,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 1.0797,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.5371886581337487,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 1.1017,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.23297990774588018,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 1.111,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.23396492589607942,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 1.0779,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 1.3654261527846838,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 1.0842,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.26743556759579407,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 1.0604,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.24566384475034003,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 1.0655,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.2345679806314927,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 1.0641,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.22735586332118546,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 1.0519,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.2371519512487248,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 1.0588,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.2638404312438537,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 1.0768,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.24567098684104813,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 1.0709,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.2361607680143941,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 1.0405,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.2438686506253647,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 1.0408,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.23292040829795177,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 1.0904,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.24165430516748085,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 1.1006,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.21975270162631974,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 1.072,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.23612638005375672,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 1.0834,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.25488099967384087,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 1.0779,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.23952120732475854,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 1.0631,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.2289093992894601,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 1.0618,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.2281571942845917,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 1.0727,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.2654397359274916,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.9989,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.24640801094158923,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 1.0993,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.23691905927567203,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 1.0902,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.23877058730140127,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 1.0249,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.2424895377414378,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 1.0799,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.23962058948605328,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 1.0296,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.22790327409305916,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 1.0479,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.23152110379294139,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 1.094,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.23894978678803624,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 1.0452,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.22637547242391637,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 1.0731,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.2273069449316352,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 1.0876,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.2256277000490742,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 1.0334,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.23811038288974354,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 1.0957,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.22678335907850913,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 1.0653,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2957055895063758,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 1.0634,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.22893979645626386,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 1.0428,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.22456278283554587,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 1.0622,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.2226686789253215,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 1.0596,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.23325927752929043,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 1.0685,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.23488542764460837,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 1.0856,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.22085534016389408,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 1.0427,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.22968125341823944,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 1.057,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.2289495217955637,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 1.0488,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.22643350858370379,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 1.11,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.2339809214874475,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 1.0264,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.23223392542359092,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 1.0928,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.233463674645543,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 1.0804,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.2231178994568372,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 1.0879,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.22708201204249445,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 1.0868,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.21828273464077821,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 1.0619,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.21867416170039666,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 1.0441,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.2241731723532887,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 1.0485,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.22711979422842538,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 1.0648,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.24008835101024548,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 1.0833,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.243287295879531,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 1.0374,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.25042184072783313,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 1.0836,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.23212439232480253,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 1.076,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.2276811569633835,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 1.0859,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.2302882323934781,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 1.0673,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.23852952658436374,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 1.0884,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.22473279420298234,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 1.0303,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.22260493032654408,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 1.0709,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.22331600065675405,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 1.0864,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.22553517718585853,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 1.0527,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.22147973947597338,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 1.0427,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.2168983023421726,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 1.0581,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.2274743812927327,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 1.0839,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.2323282710700288,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 1.0892,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.23137105841732603,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 1.0588,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.22764937345430508,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 1.1036,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.2269561897596253,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 1.0187,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.22010071356148866,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 1.0704,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.21852905245574397,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 1.0629,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.22447166452227144,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 1.0584,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.23155204178660654,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 1.0742,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.229763876633041,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 1.0631,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.22756985940074803,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 1.049,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.22593977626557693,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 1.0953,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.23344135214137182,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 1.0525,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.21347466760854888,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 1.0804,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.22683297566380634,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 1.0393,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.23416777287860954,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 1.0658,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.21743700443473934,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 1.0674,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.222494348688168,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 1.06,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.2194018302181184,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 1.0869,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.22464120864591836,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 1.0448,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.22226677332311792,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 1.0716,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.21399155072311912,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 1.0887,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.21653451226060505,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 1.059,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.21743244509726872,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 1.0722,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.21120063622169463,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 1.0827,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.22150905465772708,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 1.0656,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.21996034929758296,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 1.0942,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.2323612087944544,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 1.0746,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.21006928285151738,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 1.0534,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.22865402024202522,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 1.035,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.2395478484957313,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 1.0363,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.2248586228325117,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 1.0851,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.23828118514002242,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 1.0726,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.24047938947077205,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 1.0812,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.2148703342434829,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 1.1076,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.22279988336747156,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 1.0602,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.2339272606383704,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 1.0602,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.20963656926506266,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 1.0995,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.21559032090215513,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 1.0812,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.22101907219569578,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 1.083,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.21296447881852412,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 1.0793,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.22678112913712628,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 1.0389,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.2162560967095176,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 1.0328,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.2558825560553354,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 1.0734,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.22360054896913992,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 1.0456,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.2744829156834489,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 1.0939,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.22694250963395618,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 1.0806,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.21590758853443007,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 1.052,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.23832799403597107,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 1.0469,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.20873028591223783,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 1.0879,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.2217860126140724,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 1.0359,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.3341860962758871,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 1.0937,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.2199223964109025,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 1.1221,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.23770827401381636,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 1.0228,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.20750622048702372,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 1.08,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2161236020124174,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 1.0608,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.20975828900618673,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 1.0455,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.22025260382658302,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 1.0993,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.21038063146973293,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 1.0761,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.22698714451043214,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 1.0622,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.21392854350108623,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 1.1011,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.21314650598518375,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 1.0619,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.23880623041750362,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 1.071,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.21438274451787806,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 1.0673,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.221471570549488,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 1.0678,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.2293937047330345,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 1.0803,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.23213559238832882,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 1.0725,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.23221581912221503,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 1.0659,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.21655188659813496,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 1.0595,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.21622573705045814,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 1.0934,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.32481262128228705,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 1.088,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.22115149390686456,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 1.0461,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.4447154015508061,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 1.0626,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.20929693890217818,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 1.0694,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.2175030395445064,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 1.0622,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.21128286685530523,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 1.0822,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.21613687572134987,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 1.0663,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.2087543814275933,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 1.0777,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.21756821657964684,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 1.0664,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.23138216026910008,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 1.0428,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.21231430932211712,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 1.0861,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.21966557760630484,
+      "learning_rate": 0.0,
+      "loss": 1.053,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 5.721330610341937e+18,
+      "train_loss": 1.1867979310784744,
+      "train_runtime": 29475.7479,
+      "train_samples_per_second": 5.36,
+      "train_steps_per_second": 0.042
+    }
+  ],
+  "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": 5.721330610341937e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}