teacher_code_qwq / trainer_state.json
sedrickkeh's picture
End of training
82b17f9 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 8.603085403261682,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.78,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 8.741264951277044,
"learning_rate": 6.451612903225807e-07,
"loss": 1.7787,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 8.781751429661314,
"learning_rate": 9.67741935483871e-07,
"loss": 1.8285,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 8.718775068175342,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.7792,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 8.07117042364403,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.7379,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 7.858571877580945,
"learning_rate": 1.935483870967742e-06,
"loss": 1.744,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 6.292369419481074,
"learning_rate": 2.2580645161290324e-06,
"loss": 1.7198,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 5.799951876660287,
"learning_rate": 2.580645161290323e-06,
"loss": 1.7127,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 3.3788188758901954,
"learning_rate": 2.903225806451613e-06,
"loss": 1.7042,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 2.903549185727669,
"learning_rate": 3.225806451612903e-06,
"loss": 1.7178,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 2.645342433117499,
"learning_rate": 3.548387096774194e-06,
"loss": 1.6533,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 6.550262701795908,
"learning_rate": 3.870967741935484e-06,
"loss": 1.6468,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 7.106562408469112,
"learning_rate": 4.193548387096774e-06,
"loss": 1.6515,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 6.800852808846026,
"learning_rate": 4.516129032258065e-06,
"loss": 1.7009,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 6.199637208388759,
"learning_rate": 4.838709677419355e-06,
"loss": 1.6333,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 4.65784951976317,
"learning_rate": 5.161290322580646e-06,
"loss": 1.6095,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 4.134225581404042,
"learning_rate": 5.483870967741935e-06,
"loss": 1.5911,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 3.3817225897284042,
"learning_rate": 5.806451612903226e-06,
"loss": 1.5768,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.4535110026413984,
"learning_rate": 6.129032258064517e-06,
"loss": 1.5508,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 1.8358958272276849,
"learning_rate": 6.451612903225806e-06,
"loss": 1.5059,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 1.6445971293643553,
"learning_rate": 6.774193548387097e-06,
"loss": 1.5037,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 2.0013050281602647,
"learning_rate": 7.096774193548388e-06,
"loss": 1.4759,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.978133917719037,
"learning_rate": 7.4193548387096784e-06,
"loss": 1.4558,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.7460768995186635,
"learning_rate": 7.741935483870968e-06,
"loss": 1.5006,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.3914903139314165,
"learning_rate": 8.064516129032258e-06,
"loss": 1.4137,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.2379097454260668,
"learning_rate": 8.387096774193549e-06,
"loss": 1.4508,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.3463714543736562,
"learning_rate": 8.70967741935484e-06,
"loss": 1.4689,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.3454878237869015,
"learning_rate": 9.03225806451613e-06,
"loss": 1.4182,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.0609849042571378,
"learning_rate": 9.35483870967742e-06,
"loss": 1.4478,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.0462137661188797,
"learning_rate": 9.67741935483871e-06,
"loss": 1.428,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.8859063452686565,
"learning_rate": 1e-05,
"loss": 1.4371,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.9168580361279244,
"learning_rate": 1.0322580645161291e-05,
"loss": 1.4507,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.9076521223952301,
"learning_rate": 1.0645161290322582e-05,
"loss": 1.438,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.8877482577454805,
"learning_rate": 1.096774193548387e-05,
"loss": 1.4006,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.9310738862122158,
"learning_rate": 1.1290322580645164e-05,
"loss": 1.4156,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.8600762768651261,
"learning_rate": 1.1612903225806453e-05,
"loss": 1.4142,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.8710309941588975,
"learning_rate": 1.1935483870967743e-05,
"loss": 1.4078,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.7811357343360708,
"learning_rate": 1.2258064516129034e-05,
"loss": 1.3812,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.7824278206897729,
"learning_rate": 1.2580645161290324e-05,
"loss": 1.3998,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.6411875478110354,
"learning_rate": 1.2903225806451613e-05,
"loss": 1.4107,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.8098749695139988,
"learning_rate": 1.3225806451612903e-05,
"loss": 1.3585,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.7015919969126033,
"learning_rate": 1.3548387096774194e-05,
"loss": 1.401,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.9447866767811938,
"learning_rate": 1.3870967741935486e-05,
"loss": 1.3752,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.8231059355271249,
"learning_rate": 1.4193548387096776e-05,
"loss": 1.3829,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.7836455228767852,
"learning_rate": 1.4516129032258066e-05,
"loss": 1.3827,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.8823536858096905,
"learning_rate": 1.4838709677419357e-05,
"loss": 1.3622,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.7675880901087451,
"learning_rate": 1.5161290322580646e-05,
"loss": 1.3845,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.8277596355555185,
"learning_rate": 1.5483870967741936e-05,
"loss": 1.3506,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.9128006618292558,
"learning_rate": 1.5806451612903226e-05,
"loss": 1.3964,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.9494219158760917,
"learning_rate": 1.6129032258064517e-05,
"loss": 1.3571,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.6749618672924329,
"learning_rate": 1.6451612903225807e-05,
"loss": 1.3728,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.9102808200958468,
"learning_rate": 1.6774193548387098e-05,
"loss": 1.3638,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.8951669810615575,
"learning_rate": 1.7096774193548388e-05,
"loss": 1.3783,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 1.1540495114646225,
"learning_rate": 1.741935483870968e-05,
"loss": 1.3347,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.9694900264608782,
"learning_rate": 1.774193548387097e-05,
"loss": 1.3708,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.9055501003363214,
"learning_rate": 1.806451612903226e-05,
"loss": 1.3532,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 1.2521496532904914,
"learning_rate": 1.838709677419355e-05,
"loss": 1.344,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 1.0651818559148714,
"learning_rate": 1.870967741935484e-05,
"loss": 1.3763,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.7965194482398945,
"learning_rate": 1.903225806451613e-05,
"loss": 1.3688,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.0229915037333424,
"learning_rate": 1.935483870967742e-05,
"loss": 1.3557,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 1.6811485648979807,
"learning_rate": 1.967741935483871e-05,
"loss": 1.3503,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.9744311064773146,
"learning_rate": 2e-05,
"loss": 1.3367,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 2.6658126358492438,
"learning_rate": 2.0322580645161292e-05,
"loss": 1.3284,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 2.26499268383885,
"learning_rate": 2.0645161290322582e-05,
"loss": 1.3773,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.4271672234325339,
"learning_rate": 2.0967741935483873e-05,
"loss": 1.3571,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.7922410192687563,
"learning_rate": 2.1290322580645163e-05,
"loss": 1.3539,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.3275642599301256,
"learning_rate": 2.161290322580645e-05,
"loss": 1.3587,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 1.825784963343716,
"learning_rate": 2.193548387096774e-05,
"loss": 1.3238,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.2509833582466303,
"learning_rate": 2.225806451612903e-05,
"loss": 1.3681,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.863266029031137,
"learning_rate": 2.2580645161290328e-05,
"loss": 1.3223,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.4240423940041465,
"learning_rate": 2.2903225806451615e-05,
"loss": 1.3122,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.6612658232231436,
"learning_rate": 2.3225806451612906e-05,
"loss": 1.3202,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.754604475008043,
"learning_rate": 2.3548387096774196e-05,
"loss": 1.3395,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.5112920047528582,
"learning_rate": 2.3870967741935486e-05,
"loss": 1.2936,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.3437523980056878,
"learning_rate": 2.4193548387096777e-05,
"loss": 1.3664,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.329816968979455,
"learning_rate": 2.4516129032258067e-05,
"loss": 1.3301,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.708719073348853,
"learning_rate": 2.4838709677419358e-05,
"loss": 1.3471,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.419945526046128,
"learning_rate": 2.5161290322580648e-05,
"loss": 1.3206,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.6776880933408487,
"learning_rate": 2.548387096774194e-05,
"loss": 1.3152,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.1368850834928528,
"learning_rate": 2.5806451612903226e-05,
"loss": 1.3313,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 1.6853066266562589,
"learning_rate": 2.6129032258064516e-05,
"loss": 1.3467,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.584151507974858,
"learning_rate": 2.6451612903225806e-05,
"loss": 1.341,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 1.3954785513352805,
"learning_rate": 2.6774193548387097e-05,
"loss": 1.3396,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.8316919024811487,
"learning_rate": 2.7096774193548387e-05,
"loss": 1.3108,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.5514912653328052,
"learning_rate": 2.7419354838709678e-05,
"loss": 1.341,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.865819832555678,
"learning_rate": 2.774193548387097e-05,
"loss": 1.3687,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.67774924070574,
"learning_rate": 2.8064516129032262e-05,
"loss": 1.3104,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 1.7212112990486204,
"learning_rate": 2.8387096774193552e-05,
"loss": 1.3287,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.4842165814679233,
"learning_rate": 2.8709677419354843e-05,
"loss": 1.3131,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.6875860596618901,
"learning_rate": 2.9032258064516133e-05,
"loss": 1.3577,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.519692229781204,
"learning_rate": 2.9354838709677423e-05,
"loss": 1.2938,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.9814785875691234,
"learning_rate": 2.9677419354838714e-05,
"loss": 1.2965,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.3791325567331234,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.3229,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.921818277264272,
"learning_rate": 3.032258064516129e-05,
"loss": 1.3225,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.7318842382468118,
"learning_rate": 3.0645161290322585e-05,
"loss": 1.3143,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.7671275006050755,
"learning_rate": 3.096774193548387e-05,
"loss": 1.3175,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 1.3201734883267726,
"learning_rate": 3.1290322580645166e-05,
"loss": 1.3554,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 2.165715595373869,
"learning_rate": 3.161290322580645e-05,
"loss": 1.3445,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 2.0803350735540183,
"learning_rate": 3.193548387096774e-05,
"loss": 1.3115,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.3004016627801487,
"learning_rate": 3.2258064516129034e-05,
"loss": 1.3404,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.910683770023259,
"learning_rate": 3.258064516129033e-05,
"loss": 1.279,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.4908625771636739,
"learning_rate": 3.2903225806451614e-05,
"loss": 1.3215,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.916796897075714,
"learning_rate": 3.322580645161291e-05,
"loss": 1.2919,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 2.547523311715633,
"learning_rate": 3.3548387096774195e-05,
"loss": 1.3331,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.2090192753988134,
"learning_rate": 3.387096774193549e-05,
"loss": 1.3232,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 2.5330344390068493,
"learning_rate": 3.4193548387096776e-05,
"loss": 1.3275,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.8886418085138748,
"learning_rate": 3.451612903225807e-05,
"loss": 1.33,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 2.1999272657495252,
"learning_rate": 3.483870967741936e-05,
"loss": 1.3452,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 2.5386473067181106,
"learning_rate": 3.516129032258065e-05,
"loss": 1.3191,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.8142539550006433,
"learning_rate": 3.548387096774194e-05,
"loss": 1.3321,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 2.326548998777559,
"learning_rate": 3.5806451612903225e-05,
"loss": 1.299,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.8657778949986303,
"learning_rate": 3.612903225806452e-05,
"loss": 1.3284,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 2.007118241399083,
"learning_rate": 3.6451612903225805e-05,
"loss": 1.3496,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.9937549824745804,
"learning_rate": 3.67741935483871e-05,
"loss": 1.324,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.5136722518947423,
"learning_rate": 3.7096774193548386e-05,
"loss": 1.3121,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 2.2240400050488383,
"learning_rate": 3.741935483870968e-05,
"loss": 1.3129,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 2.358333297228358,
"learning_rate": 3.7741935483870974e-05,
"loss": 1.3583,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 1.6347830444851725,
"learning_rate": 3.806451612903226e-05,
"loss": 1.3347,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 2.7730997483545026,
"learning_rate": 3.8387096774193555e-05,
"loss": 1.3263,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 2.2291402304881824,
"learning_rate": 3.870967741935484e-05,
"loss": 1.3213,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 3.083340292392367,
"learning_rate": 3.9032258064516135e-05,
"loss": 1.3529,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.6625528126799893,
"learning_rate": 3.935483870967742e-05,
"loss": 1.3224,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 3.9779573253978495,
"learning_rate": 3.9677419354838716e-05,
"loss": 1.3391,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 3.2406956600822947,
"learning_rate": 4e-05,
"loss": 1.3236,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 2.5908228423980275,
"learning_rate": 3.9999920040266474e-05,
"loss": 1.3311,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 2.7007189088343426,
"learning_rate": 3.999968016170525e-05,
"loss": 1.2878,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 2.1813524506017785,
"learning_rate": 3.9999280366234394e-05,
"loss": 1.3108,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 2.8782283190911664,
"learning_rate": 3.999872065705065e-05,
"loss": 1.3046,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 2.457764189669879,
"learning_rate": 3.999800103862945e-05,
"loss": 1.3077,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.974207258462718,
"learning_rate": 3.9997121516724834e-05,
"loss": 1.3053,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 2.803002609166132,
"learning_rate": 3.999608209836944e-05,
"loss": 1.3178,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 2.350660373823363,
"learning_rate": 3.999488279187443e-05,
"loss": 1.321,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 2.1346733783875793,
"learning_rate": 3.9993523606829425e-05,
"loss": 1.3112,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 2.7334025442804095,
"learning_rate": 3.9992004554102435e-05,
"loss": 1.3129,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.85980618672797,
"learning_rate": 3.999032564583976e-05,
"loss": 1.3137,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.785403701094788,
"learning_rate": 3.998848689546591e-05,
"loss": 1.2873,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.9555465321470193,
"learning_rate": 3.9986488317683496e-05,
"loss": 1.3569,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.5328476357283825,
"learning_rate": 3.9984329928473067e-05,
"loss": 1.3015,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 2.2336667600487647,
"learning_rate": 3.998201174509306e-05,
"loss": 1.3228,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 1.3966724064773388,
"learning_rate": 3.997953378607961e-05,
"loss": 1.3117,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 2.0747492153994833,
"learning_rate": 3.9976896071246403e-05,
"loss": 1.2909,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.4395917324813792,
"learning_rate": 3.997409862168454e-05,
"loss": 1.3087,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 2.5438400069516356,
"learning_rate": 3.997114145976237e-05,
"loss": 1.2764,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 2.2011434332506994,
"learning_rate": 3.9968024609125245e-05,
"loss": 1.2766,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 1.7381753119733614,
"learning_rate": 3.996474809469545e-05,
"loss": 1.3073,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 2.1984470454037353,
"learning_rate": 3.996131194267189e-05,
"loss": 1.3023,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.5879759807429674,
"learning_rate": 3.995771618052995e-05,
"loss": 1.3251,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 2.355329698298091,
"learning_rate": 3.9953960837021256e-05,
"loss": 1.3243,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 1.7219730246934621,
"learning_rate": 3.995004594217343e-05,
"loss": 1.3058,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 2.2026871626164435,
"learning_rate": 3.994597152728987e-05,
"loss": 1.2593,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.8885338615425793,
"learning_rate": 3.994173762494947e-05,
"loss": 1.2879,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 2.1721179523376786,
"learning_rate": 3.9937344269006424e-05,
"loss": 1.3107,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 1.1946064296436745,
"learning_rate": 3.993279149458988e-05,
"loss": 1.3329,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 2.2756433691937894,
"learning_rate": 3.9928079338103695e-05,
"loss": 1.3299,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 1.7755892396992459,
"learning_rate": 3.992320783722616e-05,
"loss": 1.3339,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 2.005127302583674,
"learning_rate": 3.991817703090965e-05,
"loss": 1.289,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 2.3763339637227654,
"learning_rate": 3.991298695938038e-05,
"loss": 1.3685,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 1.8581408210146746,
"learning_rate": 3.9907637664138e-05,
"loss": 1.2886,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 2.522882505074461,
"learning_rate": 3.990212918795534e-05,
"loss": 1.3039,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 2.177570863438563,
"learning_rate": 3.9896461574878036e-05,
"loss": 1.3111,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 2.0479669792987467,
"learning_rate": 3.989063487022416e-05,
"loss": 1.3245,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.8162618746427344,
"learning_rate": 3.9884649120583903e-05,
"loss": 1.3186,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 1.9582249747806268,
"learning_rate": 3.987850437381915e-05,
"loss": 1.2808,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 1.6034401771438846,
"learning_rate": 3.987220067906313e-05,
"loss": 1.2927,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 1.5385773707547152,
"learning_rate": 3.986573808672002e-05,
"loss": 1.3194,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 2.235016036817719,
"learning_rate": 3.9859116648464534e-05,
"loss": 1.3017,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 1.7753544140185231,
"learning_rate": 3.985233641724152e-05,
"loss": 1.3125,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 1.8944190356646935,
"learning_rate": 3.984539744726554e-05,
"loss": 1.2846,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 1.5008812234019508,
"learning_rate": 3.983829979402038e-05,
"loss": 1.3175,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 2.074515117276577,
"learning_rate": 3.983104351425871e-05,
"loss": 1.2939,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 2.2742170924283496,
"learning_rate": 3.982362866600154e-05,
"loss": 1.3201,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 1.2257355032197317,
"learning_rate": 3.9816055308537806e-05,
"loss": 1.3129,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 2.2970593038962615,
"learning_rate": 3.980832350242388e-05,
"loss": 1.331,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.6804374745923547,
"learning_rate": 3.980043330948306e-05,
"loss": 1.3127,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 2.1957831548014553,
"learning_rate": 3.9792384792805124e-05,
"loss": 1.2593,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 1.6733296211240598,
"learning_rate": 3.978417801674579e-05,
"loss": 1.3299,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 2.2701838461326056,
"learning_rate": 3.9775813046926244e-05,
"loss": 1.3394,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 1.4583777918881407,
"learning_rate": 3.976728995023254e-05,
"loss": 1.323,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 1.9957054069839526,
"learning_rate": 3.975860879481514e-05,
"loss": 1.3072,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 1.6804956204676542,
"learning_rate": 3.974976965008833e-05,
"loss": 1.3187,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 2.0601014930711172,
"learning_rate": 3.9740772586729674e-05,
"loss": 1.3213,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 1.464638864968656,
"learning_rate": 3.973161767667946e-05,
"loss": 1.2888,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 2.5248512059052985,
"learning_rate": 3.972230499314009e-05,
"loss": 1.2853,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 1.7767092313860169,
"learning_rate": 3.9712834610575545e-05,
"loss": 1.329,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 1.906408457538698,
"learning_rate": 3.9703206604710746e-05,
"loss": 1.3143,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 1.4879990837647818,
"learning_rate": 3.969342105253098e-05,
"loss": 1.2763,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 1.728504409663229,
"learning_rate": 3.9683478032281246e-05,
"loss": 1.2932,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 1.6295722420023728,
"learning_rate": 3.967337762346568e-05,
"loss": 1.2971,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 1.8982684561373646,
"learning_rate": 3.9663119906846885e-05,
"loss": 1.3147,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.923328209350778,
"learning_rate": 3.965270496444528e-05,
"loss": 1.3072,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 1.8796329345491383,
"learning_rate": 3.9642132879538465e-05,
"loss": 1.3104,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 2.021082350276648,
"learning_rate": 3.9631403736660566e-05,
"loss": 1.2804,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 1.5871071608782763,
"learning_rate": 3.962051762160151e-05,
"loss": 1.2738,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 1.920095131411854,
"learning_rate": 3.960947462140638e-05,
"loss": 1.2736,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.4257320448625483,
"learning_rate": 3.959827482437472e-05,
"loss": 1.2571,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 2.586079188504378,
"learning_rate": 3.958691832005981e-05,
"loss": 1.3034,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 1.1596616980527747,
"learning_rate": 3.9575405199267944e-05,
"loss": 1.2855,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 2.1359601491104447,
"learning_rate": 3.956373555405774e-05,
"loss": 1.3008,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 2.0245173380201518,
"learning_rate": 3.955190947773937e-05,
"loss": 1.3216,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 1.9021165684706876,
"learning_rate": 3.9539927064873824e-05,
"loss": 1.283,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 2.0345923419998924,
"learning_rate": 3.952778841127214e-05,
"loss": 1.2962,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 1.3151789794823892,
"learning_rate": 3.95154936139947e-05,
"loss": 1.2978,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 1.8683600996497662,
"learning_rate": 3.950304277135034e-05,
"loss": 1.313,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.635924256104552,
"learning_rate": 3.94904359828957e-05,
"loss": 1.3019,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 1.3955950665691177,
"learning_rate": 3.94776733494343e-05,
"loss": 1.2965,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 2.17084775540723,
"learning_rate": 3.9464754973015826e-05,
"loss": 1.3108,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 1.4060150769936628,
"learning_rate": 3.9451680956935275e-05,
"loss": 1.3139,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 2.953028327198921,
"learning_rate": 3.943845140573213e-05,
"loss": 1.3061,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 2.2169608203847724,
"learning_rate": 3.942506642518952e-05,
"loss": 1.2809,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 2.950606559064427,
"learning_rate": 3.9411526122333396e-05,
"loss": 1.2832,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 2.4523063283682847,
"learning_rate": 3.939783060543167e-05,
"loss": 1.2676,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 2.0207191001061293,
"learning_rate": 3.938397998399333e-05,
"loss": 1.3004,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 2.1637235406740123,
"learning_rate": 3.936997436876756e-05,
"loss": 1.2873,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 1.7963666947883883,
"learning_rate": 3.935581387174289e-05,
"loss": 1.3213,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 1.714410558246024,
"learning_rate": 3.9341498606146295e-05,
"loss": 1.2941,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 1.8724071071869126,
"learning_rate": 3.932702868644224e-05,
"loss": 1.3215,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 1.7326834983735409,
"learning_rate": 3.931240422833181e-05,
"loss": 1.3187,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 2.0722541112057393,
"learning_rate": 3.9297625348751806e-05,
"loss": 1.2848,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 1.5595326039038413,
"learning_rate": 3.9282692165873744e-05,
"loss": 1.304,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 1.7156043560919596,
"learning_rate": 3.9267604799102945e-05,
"loss": 1.2934,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 1.4795251639258051,
"learning_rate": 3.9252363369077615e-05,
"loss": 1.2875,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 1.7818313298514674,
"learning_rate": 3.9236967997667796e-05,
"loss": 1.2956,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 2.4726420712393464,
"learning_rate": 3.9221418807974496e-05,
"loss": 1.3021,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 1.2841269138757097,
"learning_rate": 3.92057159243286e-05,
"loss": 1.2658,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 1.3440979458374847,
"learning_rate": 3.9189859472289956e-05,
"loss": 1.2662,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 1.6490476353090906,
"learning_rate": 3.9173849578646324e-05,
"loss": 1.2823,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 1.7370790536766458,
"learning_rate": 3.9157686371412395e-05,
"loss": 1.3337,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 1.907582351208461,
"learning_rate": 3.914136997982874e-05,
"loss": 1.2964,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 2.4599566049616333,
"learning_rate": 3.912490053436079e-05,
"loss": 1.2972,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 1.1831650375858422,
"learning_rate": 3.9108278166697805e-05,
"loss": 1.282,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 3.9094144405315565,
"learning_rate": 3.909150300975178e-05,
"loss": 1.3185,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 2.8061006838091385,
"learning_rate": 3.907457519765642e-05,
"loss": 1.308,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 3.165742701254044,
"learning_rate": 3.905749486576607e-05,
"loss": 1.3349,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 2.6919046047628474,
"learning_rate": 3.90402621506546e-05,
"loss": 1.2818,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 2.715371450874169,
"learning_rate": 3.902287719011435e-05,
"loss": 1.3307,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 1.9622904930085325,
"learning_rate": 3.900534012315499e-05,
"loss": 1.3244,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 1.5058955840222048,
"learning_rate": 3.8987651090002454e-05,
"loss": 1.2734,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 3.5206835175770075,
"learning_rate": 3.896981023209777e-05,
"loss": 1.2669,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 2.075270166792292,
"learning_rate": 3.8951817692095966e-05,
"loss": 1.307,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 4.280564395871635,
"learning_rate": 3.893367361386491e-05,
"loss": 1.3375,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 3.4081687881025617,
"learning_rate": 3.891537814248417e-05,
"loss": 1.2962,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 3.343670747467635,
"learning_rate": 3.889693142424384e-05,
"loss": 1.3283,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 3.407109595603751,
"learning_rate": 3.8878333606643405e-05,
"loss": 1.3177,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 2.7944710467562976,
"learning_rate": 3.885958483839049e-05,
"loss": 1.282,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 2.313760865790683,
"learning_rate": 3.8840685269399786e-05,
"loss": 1.2714,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 3.2253815302917213,
"learning_rate": 3.882163505079171e-05,
"loss": 1.2757,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 2.1832171595546215,
"learning_rate": 3.880243433489132e-05,
"loss": 1.2657,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 4.196681696145751,
"learning_rate": 3.878308327522702e-05,
"loss": 1.2465,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 3.9813749614354985,
"learning_rate": 3.876358202652936e-05,
"loss": 1.258,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 2.321073725744671,
"learning_rate": 3.8743930744729835e-05,
"loss": 1.2883,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 2.2547853257675525,
"learning_rate": 3.8724129586959535e-05,
"loss": 1.2557,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 2.77496493243085,
"learning_rate": 3.870417871154801e-05,
"loss": 1.2527,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 1.9555780046651263,
"learning_rate": 3.868407827802192e-05,
"loss": 1.2882,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 4.0247135977136415,
"learning_rate": 3.8663828447103804e-05,
"loss": 1.2564,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 3.662995434018082,
"learning_rate": 3.8643429380710765e-05,
"loss": 1.2755,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 2.3104206301063934,
"learning_rate": 3.862288124195319e-05,
"loss": 1.2557,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 2.1407475448530184,
"learning_rate": 3.8602184195133464e-05,
"loss": 1.2401,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 2.9709238143775294,
"learning_rate": 3.8581338405744604e-05,
"loss": 1.2598,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 2.249566095012782,
"learning_rate": 3.8560344040468996e-05,
"loss": 1.2656,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 3.664197128418057,
"learning_rate": 3.853920126717703e-05,
"loss": 1.2337,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 3.381673754216082,
"learning_rate": 3.851791025492574e-05,
"loss": 1.2375,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 2.392989013275902,
"learning_rate": 3.8496471173957515e-05,
"loss": 1.2492,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 2.2236657499211083,
"learning_rate": 3.847488419569866e-05,
"loss": 1.2528,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 2.9583270099958137,
"learning_rate": 3.8453149492758077e-05,
"loss": 1.2628,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 2.481329049382045,
"learning_rate": 3.8431267238925886e-05,
"loss": 1.2669,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 3.24900422675548,
"learning_rate": 3.840923760917199e-05,
"loss": 1.2419,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 3.004329346682039,
"learning_rate": 3.838706077964473e-05,
"loss": 1.2916,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 2.2631048210910096,
"learning_rate": 3.836473692766944e-05,
"loss": 1.2472,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 2.0034692372053753,
"learning_rate": 3.834226623174705e-05,
"loss": 1.2613,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 3.2729066072780566,
"learning_rate": 3.831964887155264e-05,
"loss": 1.2666,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 2.7868971639117266,
"learning_rate": 3.829688502793403e-05,
"loss": 1.2584,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 2.760441659061895,
"learning_rate": 3.827397488291029e-05,
"loss": 1.2477,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 2.5138362925620688,
"learning_rate": 3.8250918619670343e-05,
"loss": 1.2575,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 2.7200765538253013,
"learning_rate": 3.822771642257145e-05,
"loss": 1.2326,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 2.4642720550062065,
"learning_rate": 3.820436847713776e-05,
"loss": 1.2631,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 2.844754129546703,
"learning_rate": 3.8180874970058826e-05,
"loss": 1.2602,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 2.6029789968473365,
"learning_rate": 3.8157236089188103e-05,
"loss": 1.2347,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 2.593458717485962,
"learning_rate": 3.813345202354145e-05,
"loss": 1.2369,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 2.3744768510970977,
"learning_rate": 3.810952296329563e-05,
"loss": 1.2478,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 2.750073686346052,
"learning_rate": 3.808544909978676e-05,
"loss": 1.2545,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 2.484348586580054,
"learning_rate": 3.806123062550882e-05,
"loss": 1.2426,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 2.7371478514891665,
"learning_rate": 3.803686773411208e-05,
"loss": 1.2405,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 2.6438029027796994,
"learning_rate": 3.801236062040158e-05,
"loss": 1.2706,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 2.5478108260655468,
"learning_rate": 3.798770948033553e-05,
"loss": 1.2679,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 2.275054432961149,
"learning_rate": 3.796291451102381e-05,
"loss": 1.2671,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 2.972425876836733,
"learning_rate": 3.793797591072631e-05,
"loss": 1.2465,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 2.5232219174639723,
"learning_rate": 3.791289387885144e-05,
"loss": 1.2611,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 2.564513341459357,
"learning_rate": 3.788766861595443e-05,
"loss": 1.2376,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 2.261198557343063,
"learning_rate": 3.7862300323735835e-05,
"loss": 1.2782,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 2.9203689477313937,
"learning_rate": 3.783678920503983e-05,
"loss": 1.2447,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 2.5376237118102463,
"learning_rate": 3.781113546385265e-05,
"loss": 1.2596,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 2.581681218688675,
"learning_rate": 3.7785339305300916e-05,
"loss": 1.2426,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 2.3349686169991575,
"learning_rate": 3.7759400935650025e-05,
"loss": 1.3044,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 2.828690864437062,
"learning_rate": 3.77333205623025e-05,
"loss": 1.2301,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 2.5469583848221595,
"learning_rate": 3.7707098393796295e-05,
"loss": 1.2609,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 2.48752557302592,
"learning_rate": 3.7680734639803185e-05,
"loss": 1.2388,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 2.2029178173036112,
"learning_rate": 3.765422951112704e-05,
"loss": 1.2575,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 2.769986181671539,
"learning_rate": 3.762758321970216e-05,
"loss": 1.2721,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 2.397927727115516,
"learning_rate": 3.7600795978591584e-05,
"loss": 1.2508,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 2.522470967275978,
"learning_rate": 3.757386800198538e-05,
"loss": 1.2298,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 2.413963698539984,
"learning_rate": 3.7546799505198925e-05,
"loss": 1.2308,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 2.626054397979285,
"learning_rate": 3.7519590704671197e-05,
"loss": 1.2917,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 2.5365296649985516,
"learning_rate": 3.749224181796305e-05,
"loss": 1.2383,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 2.350253632110115,
"learning_rate": 3.746475306375545e-05,
"loss": 1.2639,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 1.993741919297312,
"learning_rate": 3.743712466184774e-05,
"loss": 1.2487,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 2.848026984933644,
"learning_rate": 3.7409356833155885e-05,
"loss": 1.2377,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 2.4786635567380375,
"learning_rate": 3.73814497997107e-05,
"loss": 1.2512,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 2.5057955518835113,
"learning_rate": 3.7353403784656095e-05,
"loss": 1.2358,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 2.5041386412921156,
"learning_rate": 3.732521901224724e-05,
"loss": 1.2774,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 2.31802693409623,
"learning_rate": 3.7296895707848844e-05,
"loss": 1.2626,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 1.9696326823165398,
"learning_rate": 3.7268434097933275e-05,
"loss": 1.2524,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 2.7163907320571443,
"learning_rate": 3.723983441007882e-05,
"loss": 1.2303,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 2.4454952623077393,
"learning_rate": 3.721109687296781e-05,
"loss": 1.2335,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 2.476968260413085,
"learning_rate": 3.718222171638484e-05,
"loss": 1.2713,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 2.1350098620740523,
"learning_rate": 3.715320917121488e-05,
"loss": 1.2501,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 2.3768531416877643,
"learning_rate": 3.712405946944149e-05,
"loss": 1.2591,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 2.0658265317409086,
"learning_rate": 3.709477284414488e-05,
"loss": 1.2594,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 2.5826977681027796,
"learning_rate": 3.7065349529500136e-05,
"loss": 1.277,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 2.2785792639760007,
"learning_rate": 3.7035789760775305e-05,
"loss": 1.2558,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 2.3178412360844582,
"learning_rate": 3.70060937743295e-05,
"loss": 1.2517,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 1.8204351153104847,
"learning_rate": 3.6976261807611045e-05,
"loss": 1.2318,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 2.8843430191601205,
"learning_rate": 3.694629409915555e-05,
"loss": 1.2618,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 2.6454115487077905,
"learning_rate": 3.691619088858401e-05,
"loss": 1.2914,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 2.03160573662652,
"learning_rate": 3.68859524166009e-05,
"loss": 1.2495,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 2.0026562695526473,
"learning_rate": 3.6855578924992226e-05,
"loss": 1.2408,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 2.5581071573224303,
"learning_rate": 3.6825070656623626e-05,
"loss": 1.2324,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 2.081204598577246,
"learning_rate": 3.67944278554384e-05,
"loss": 1.2415,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 2.45513932925379,
"learning_rate": 3.676365076645557e-05,
"loss": 1.2423,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 2.1748855139028227,
"learning_rate": 3.673273963576791e-05,
"loss": 1.2653,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 2.318416530051789,
"learning_rate": 3.670169471054001e-05,
"loss": 1.2316,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 2.046831563698622,
"learning_rate": 3.6670516239006254e-05,
"loss": 1.2286,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 2.534801271063973,
"learning_rate": 3.663920447046888e-05,
"loss": 1.2446,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 2.2519734900039317,
"learning_rate": 3.6607759655295954e-05,
"loss": 1.2385,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 2.2936246686574524,
"learning_rate": 3.657618204491937e-05,
"loss": 1.2637,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 1.98987872105123,
"learning_rate": 3.6544471891832875e-05,
"loss": 1.2367,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 2.3773700299319525,
"learning_rate": 3.6512629449589983e-05,
"loss": 1.2182,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 2.030036803764162,
"learning_rate": 3.6480654972802045e-05,
"loss": 1.2118,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 2.5178810360006687,
"learning_rate": 3.644854871713611e-05,
"loss": 1.2562,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 2.2706046833394393,
"learning_rate": 3.6416310939312924e-05,
"loss": 1.2829,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 1.9190388323513612,
"learning_rate": 3.638394189710493e-05,
"loss": 1.2485,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 1.5995377500720338,
"learning_rate": 3.635144184933412e-05,
"loss": 1.2425,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 2.621926890063351,
"learning_rate": 3.6318811055869994e-05,
"loss": 1.2519,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 2.2551835745209874,
"learning_rate": 3.628604977762752e-05,
"loss": 1.2374,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 2.1325877350922258,
"learning_rate": 3.6253158276565004e-05,
"loss": 1.2834,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 2.0313552984750634,
"learning_rate": 3.6220136815682016e-05,
"loss": 1.293,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 2.105488732889953,
"learning_rate": 3.618698565901727e-05,
"loss": 1.2113,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 1.740442847754041,
"learning_rate": 3.615370507164654e-05,
"loss": 1.2416,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 2.5195741749981986,
"learning_rate": 3.61202953196805e-05,
"loss": 1.2463,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 2.359772032824461,
"learning_rate": 3.6086756670262654e-05,
"loss": 1.2452,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 1.9184988085187173,
"learning_rate": 3.6053089391567135e-05,
"loss": 1.2349,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 1.7698984012404093,
"learning_rate": 3.601929375279662e-05,
"loss": 1.2244,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 2.2906177644301344,
"learning_rate": 3.598537002418012e-05,
"loss": 1.2365,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 1.8143108466738642,
"learning_rate": 3.5951318476970885e-05,
"loss": 1.2506,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 2.506432041416203,
"learning_rate": 3.591713938344416e-05,
"loss": 1.274,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 2.3542728884640294,
"learning_rate": 3.588283301689507e-05,
"loss": 1.2775,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 2.0240422452341607,
"learning_rate": 3.5848399651636424e-05,
"loss": 1.1925,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 1.8804430230387061,
"learning_rate": 3.5813839562996473e-05,
"loss": 1.262,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 1.8466218169933781,
"learning_rate": 3.5779153027316775e-05,
"loss": 1.2383,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 1.3774539485174093,
"learning_rate": 3.574434032194994e-05,
"loss": 1.2668,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 2.3070414983967833,
"learning_rate": 3.5709401725257436e-05,
"loss": 1.2255,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 1.6300413308908703,
"learning_rate": 3.5674337516607346e-05,
"loss": 1.2163,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 2.474013708915133,
"learning_rate": 3.563914797637216e-05,
"loss": 1.2364,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 2.3846372997297727,
"learning_rate": 3.560383338592649e-05,
"loss": 1.2512,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 1.5921316795638425,
"learning_rate": 3.556839402764487e-05,
"loss": 1.2613,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 1.6968308777592256,
"learning_rate": 3.553283018489946e-05,
"loss": 1.2501,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 1.6981140126349181,
"learning_rate": 3.54971421420578e-05,
"loss": 1.2557,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 1.4431617353611101,
"learning_rate": 3.5461330184480524e-05,
"loss": 1.2505,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 1.8588267142960337,
"learning_rate": 3.54253945985191e-05,
"loss": 1.2549,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 1.4972365506939358,
"learning_rate": 3.5389335671513524e-05,
"loss": 1.2223,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 1.9392262676046843,
"learning_rate": 3.5353153691789986e-05,
"loss": 1.2503,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 1.326102972307925,
"learning_rate": 3.531684894865866e-05,
"loss": 1.2525,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 2.0520294714875207,
"learning_rate": 3.528042173241128e-05,
"loss": 1.2362,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 1.8177060122969717,
"learning_rate": 3.524387233431892e-05,
"loss": 1.2373,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 1.6118933648553968,
"learning_rate": 3.520720104662958e-05,
"loss": 1.2392,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 1.4891486584569904,
"learning_rate": 3.5170408162565904e-05,
"loss": 1.2662,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 1.5079246591287336,
"learning_rate": 3.513349397632281e-05,
"loss": 1.2551,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 1.760346880445196,
"learning_rate": 3.5096458783065145e-05,
"loss": 1.244,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 1.3482016613530872,
"learning_rate": 3.505930287892533e-05,
"loss": 1.2289,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 1.716978141554858,
"learning_rate": 3.502202656100099e-05,
"loss": 1.2433,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 1.4084528447686886,
"learning_rate": 3.498463012735256e-05,
"loss": 1.2515,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 1.282414383035051,
"learning_rate": 3.494711387700094e-05,
"loss": 1.2576,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 1.3674000955899142,
"learning_rate": 3.490947810992505e-05,
"loss": 1.2289,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 1.7480638572697764,
"learning_rate": 3.48717231270595e-05,
"loss": 1.2632,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 1.1852514563364522,
"learning_rate": 3.483384923029211e-05,
"loss": 1.2397,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 1.6404382980424357,
"learning_rate": 3.479585672246156e-05,
"loss": 1.2634,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 1.4099929951151569,
"learning_rate": 3.4757745907354924e-05,
"loss": 1.2566,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 1.6914939621933973,
"learning_rate": 3.471951708970527e-05,
"loss": 1.2491,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 1.5643317087514546,
"learning_rate": 3.468117057518921e-05,
"loss": 1.2649,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 1.6309810248386325,
"learning_rate": 3.4642706670424426e-05,
"loss": 1.2779,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 1.4456115430438028,
"learning_rate": 3.460412568296731e-05,
"loss": 1.2341,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 1.460409914755936,
"learning_rate": 3.456542792131039e-05,
"loss": 1.2232,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 1.6449047952189118,
"learning_rate": 3.4526613694879936e-05,
"loss": 1.2742,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 1.0356076884273857,
"learning_rate": 3.448768331403348e-05,
"loss": 1.2109,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 1.161687474406656,
"learning_rate": 3.44486370900573e-05,
"loss": 1.27,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 1.2120872132152882,
"learning_rate": 3.440947533516396e-05,
"loss": 1.2585,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 1.639024339923923,
"learning_rate": 3.4370198362489816e-05,
"loss": 1.2416,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 1.2188840974169026,
"learning_rate": 3.43308064860925e-05,
"loss": 1.256,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 1.5153934384167227,
"learning_rate": 3.4291300020948393e-05,
"loss": 1.244,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 1.2490025171481416,
"learning_rate": 3.425167928295015e-05,
"loss": 1.2116,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 1.776371490033861,
"learning_rate": 3.421194458890411e-05,
"loss": 1.2402,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 1.6270521010289822,
"learning_rate": 3.4172096256527865e-05,
"loss": 1.2683,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 1.4666858452070128,
"learning_rate": 3.4132134604447595e-05,
"loss": 1.2511,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 1.8243212118146532,
"learning_rate": 3.409205995219561e-05,
"loss": 1.2682,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 1.280337097911277,
"learning_rate": 3.4051872620207765e-05,
"loss": 1.2284,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 2.074306111700939,
"learning_rate": 3.40115729298209e-05,
"loss": 1.2727,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 1.5781228176464006,
"learning_rate": 3.397116120327025e-05,
"loss": 1.2809,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 1.5364966421324449,
"learning_rate": 3.3930637763686916e-05,
"loss": 1.263,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 1.6988504475746415,
"learning_rate": 3.389000293509524e-05,
"loss": 1.2812,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 1.1491599763262346,
"learning_rate": 3.3849257042410225e-05,
"loss": 1.2201,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 1.6214204374353773,
"learning_rate": 3.380840041143494e-05,
"loss": 1.2483,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 1.6298529847393424,
"learning_rate": 3.376743336885792e-05,
"loss": 1.2366,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 1.3025830818629034,
"learning_rate": 3.372635624225055e-05,
"loss": 1.2479,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 1.9713505101387372,
"learning_rate": 3.3685169360064445e-05,
"loss": 1.2422,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 1.6331065148801396,
"learning_rate": 3.3643873051628804e-05,
"loss": 1.2806,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 1.9862846662022038,
"learning_rate": 3.3602467647147814e-05,
"loss": 1.266,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 1.7503070464401271,
"learning_rate": 3.356095347769799e-05,
"loss": 1.2576,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 1.6361331742369134,
"learning_rate": 3.351933087522552e-05,
"loss": 1.2534,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 1.341884955771158,
"learning_rate": 3.3477600172543626e-05,
"loss": 1.2614,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 1.6259764524935392,
"learning_rate": 3.3435761703329894e-05,
"loss": 1.2454,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 1.3375003633926594,
"learning_rate": 3.3393815802123606e-05,
"loss": 1.2031,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 1.3678613646284903,
"learning_rate": 3.335176280432308e-05,
"loss": 1.2488,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 1.4820800809344366,
"learning_rate": 3.330960304618295e-05,
"loss": 1.2397,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 1.032409844809438,
"learning_rate": 3.326733686481153e-05,
"loss": 1.2325,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 1.5247569367793423,
"learning_rate": 3.3224964598168084e-05,
"loss": 1.2529,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 1.1697306709858213,
"learning_rate": 3.318248658506012e-05,
"loss": 1.2267,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 2.0745178978369054,
"learning_rate": 3.313990316514069e-05,
"loss": 1.2821,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 1.8381186733780672,
"learning_rate": 3.309721467890571e-05,
"loss": 1.2581,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 1.6298096428938316,
"learning_rate": 3.305442146769115e-05,
"loss": 1.2458,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 1.7541977268943474,
"learning_rate": 3.30115238736704e-05,
"loss": 1.2558,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 1.4526274906121908,
"learning_rate": 3.296852223985148e-05,
"loss": 1.2472,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 1.5630122353741172,
"learning_rate": 3.2925416910074315e-05,
"loss": 1.2834,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 1.3393391871243556,
"learning_rate": 3.288220822900796e-05,
"loss": 1.2884,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 1.3393440843600255,
"learning_rate": 3.283889654214788e-05,
"loss": 1.2768,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 1.1484553579050427,
"learning_rate": 3.279548219581317e-05,
"loss": 1.2231,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 1.299195850489073,
"learning_rate": 3.275196553714379e-05,
"loss": 1.2533,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 1.150670046552566,
"learning_rate": 3.270834691409778e-05,
"loss": 1.2411,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 1.1559252027291038,
"learning_rate": 3.266462667544849e-05,
"loss": 1.2692,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.7987342194876248,
"learning_rate": 3.262080517078178e-05,
"loss": 1.2508,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 1.158795464312777,
"learning_rate": 3.257688275049323e-05,
"loss": 1.2363,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 1.1390246313786598,
"learning_rate": 3.253285976578535e-05,
"loss": 1.2182,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 1.4017363185214353,
"learning_rate": 3.2488736568664756e-05,
"loss": 1.2453,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 1.0621082894551026,
"learning_rate": 3.244451351193935e-05,
"loss": 1.2791,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 1.0736659992791309,
"learning_rate": 3.240019094921551e-05,
"loss": 1.2334,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.9828255050323857,
"learning_rate": 3.235576923489527e-05,
"loss": 1.2427,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 1.1111699815063434,
"learning_rate": 3.2311248724173476e-05,
"loss": 1.2644,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 1.4460894188835012,
"learning_rate": 3.226662977303494e-05,
"loss": 1.232,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.8794226177292409,
"learning_rate": 3.2221912738251626e-05,
"loss": 1.2727,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 1.325861589193597,
"learning_rate": 3.2177097977379736e-05,
"loss": 1.2552,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 1.5962318218194071,
"learning_rate": 3.21321858487569e-05,
"loss": 1.2707,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.9216987264322036,
"learning_rate": 3.208717671149931e-05,
"loss": 1.2735,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 2.0712238648074703,
"learning_rate": 3.204207092549882e-05,
"loss": 1.286,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 1.248079784404818,
"learning_rate": 3.19968688514201e-05,
"loss": 1.2797,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 2.244734005225525,
"learning_rate": 3.195157085069774e-05,
"loss": 1.2866,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 1.758000929845638,
"learning_rate": 3.190617728553332e-05,
"loss": 1.3084,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 2.1554519020296192,
"learning_rate": 3.1860688518892606e-05,
"loss": 1.2628,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 2.0016637728284983,
"learning_rate": 3.181510491450254e-05,
"loss": 1.2589,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 1.6307879321623384,
"learning_rate": 3.176942683684842e-05,
"loss": 1.271,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 1.7367745289462462,
"learning_rate": 3.1723654651170934e-05,
"loss": 1.2463,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 1.537762324721678,
"learning_rate": 3.167778872346327e-05,
"loss": 1.2033,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 1.5917644969140874,
"learning_rate": 3.163182942046815e-05,
"loss": 1.2266,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 1.4820825761265672,
"learning_rate": 3.158577710967494e-05,
"loss": 1.2857,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 1.739445311630486,
"learning_rate": 3.153963215931669e-05,
"loss": 1.2131,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 1.4832621234178407,
"learning_rate": 3.14933949383672e-05,
"loss": 1.2347,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 1.503794624792782,
"learning_rate": 3.144706581653805e-05,
"loss": 1.2875,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 1.3221458857609447,
"learning_rate": 3.140064516427566e-05,
"loss": 1.2253,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 1.4987318567770558,
"learning_rate": 3.135413335275833e-05,
"loss": 1.2605,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 1.353895925737337,
"learning_rate": 3.130753075389327e-05,
"loss": 1.2328,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 1.0318808233739456,
"learning_rate": 3.1260837740313625e-05,
"loss": 1.2193,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 1.5067960881005966,
"learning_rate": 3.121405468537547e-05,
"loss": 1.2199,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 1.0263825699473974,
"learning_rate": 3.116718196315488e-05,
"loss": 1.2228,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 1.517214835326192,
"learning_rate": 3.1120219948444894e-05,
"loss": 1.2399,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.984334242788725,
"learning_rate": 3.107316901675252e-05,
"loss": 1.2451,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 1.4366100725303248,
"learning_rate": 3.1026029544295745e-05,
"loss": 1.2871,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 1.3405223820196488,
"learning_rate": 3.097880190800057e-05,
"loss": 1.2701,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 1.3167290592404377,
"learning_rate": 3.093148648549788e-05,
"loss": 1.2676,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 1.5703069551976294,
"learning_rate": 3.088408365512055e-05,
"loss": 1.2814,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 1.3669562519615845,
"learning_rate": 3.083659379590034e-05,
"loss": 1.24,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 1.2804319366963024,
"learning_rate": 3.0789017287564913e-05,
"loss": 1.2209,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 1.3808341700497553,
"learning_rate": 3.074135451053475e-05,
"loss": 1.2329,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 1.2689725805784557,
"learning_rate": 3.069360584592015e-05,
"loss": 1.2614,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 1.1851069097870748,
"learning_rate": 3.064577167551816e-05,
"loss": 1.2427,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 1.2354826661349685,
"learning_rate": 3.059785238180953e-05,
"loss": 1.2538,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.9007489605518624,
"learning_rate": 3.054984834795566e-05,
"loss": 1.2705,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 1.4485783284926745,
"learning_rate": 3.0501759957795537e-05,
"loss": 1.2629,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 1.0663374574816413,
"learning_rate": 3.0453587595842627e-05,
"loss": 1.2359,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 1.2825248060964702,
"learning_rate": 3.0405331647281856e-05,
"loss": 1.2076,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 1.2245748449256324,
"learning_rate": 3.0356992497966508e-05,
"loss": 1.2411,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 1.2990177533330995,
"learning_rate": 3.0308570534415137e-05,
"loss": 1.2309,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 1.1826645968554,
"learning_rate": 3.0260066143808467e-05,
"loss": 1.2385,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 1.1367822804166667,
"learning_rate": 3.021147971398632e-05,
"loss": 1.2471,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.9937622752089761,
"learning_rate": 3.0162811633444478e-05,
"loss": 1.2466,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.8308382639936551,
"learning_rate": 3.0114062291331635e-05,
"loss": 1.251,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 1.2298726573294565,
"learning_rate": 3.0065232077446217e-05,
"loss": 1.2553,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 1.055987376515106,
"learning_rate": 3.001632138223332e-05,
"loss": 1.2394,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 1.0901592467675243,
"learning_rate": 2.9967330596781565e-05,
"loss": 1.2057,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.9394488078021765,
"learning_rate": 2.9918260112819952e-05,
"loss": 1.2151,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 1.086375831822422,
"learning_rate": 2.9869110322714778e-05,
"loss": 1.2072,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 1.037692222167649,
"learning_rate": 2.9819881619466447e-05,
"loss": 1.24,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 1.1327648136192157,
"learning_rate": 2.9770574396706364e-05,
"loss": 1.1977,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 1.1790942530477546,
"learning_rate": 2.972118904869376e-05,
"loss": 1.1848,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 1.0326355326035122,
"learning_rate": 2.9671725970312563e-05,
"loss": 1.2242,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 1.1611240898204338,
"learning_rate": 2.9622185557068238e-05,
"loss": 1.2218,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 1.1212072101297663,
"learning_rate": 2.95725682050846e-05,
"loss": 1.1864,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.8989190918059027,
"learning_rate": 2.9522874311100684e-05,
"loss": 1.2047,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 1.1108028860779595,
"learning_rate": 2.9473104272467526e-05,
"loss": 1.1902,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.8033727653204633,
"learning_rate": 2.942325848714505e-05,
"loss": 1.1899,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.8565848480364648,
"learning_rate": 2.9373337353698804e-05,
"loss": 1.1751,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.9303964512744914,
"learning_rate": 2.932334127129686e-05,
"loss": 1.2084,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.889058654934133,
"learning_rate": 2.9273270639706547e-05,
"loss": 1.2216,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.8955644192407646,
"learning_rate": 2.922312585929131e-05,
"loss": 1.1568,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.9956222601790276,
"learning_rate": 2.9172907331007478e-05,
"loss": 1.2202,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 1.6929572015674823,
"learning_rate": 2.9122615456401053e-05,
"loss": 1.19,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.7545329717638711,
"learning_rate": 2.9072250637604535e-05,
"loss": 1.2036,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 2.034293523951985,
"learning_rate": 2.9021813277333667e-05,
"loss": 1.1945,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 1.137506758655233,
"learning_rate": 2.8971303778884244e-05,
"loss": 1.2424,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 2.3319241142946803,
"learning_rate": 2.8920722546128863e-05,
"loss": 1.2112,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 1.9932771768413697,
"learning_rate": 2.887006998351372e-05,
"loss": 1.1897,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 2.02875413867428,
"learning_rate": 2.881934649605535e-05,
"loss": 1.2097,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 2.1914004137269623,
"learning_rate": 2.8768552489337408e-05,
"loss": 1.2278,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 1.5333013027666886,
"learning_rate": 2.8717688369507425e-05,
"loss": 1.2014,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 1.7352974385666964,
"learning_rate": 2.8666754543273535e-05,
"loss": 1.1814,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 1.3333934511685073,
"learning_rate": 2.8615751417901273e-05,
"loss": 1.2018,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 1.7012343922165627,
"learning_rate": 2.856467940121025e-05,
"loss": 1.1835,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 1.0613756520291429,
"learning_rate": 2.8513538901570964e-05,
"loss": 1.1841,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 1.5708995143014164,
"learning_rate": 2.846233032790149e-05,
"loss": 1.2024,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 1.1688513591228433,
"learning_rate": 2.84110540896642e-05,
"loss": 1.2087,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 1.27793513547683,
"learning_rate": 2.8359710596862555e-05,
"loss": 1.2031,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 1.1725303756035073,
"learning_rate": 2.8308300260037734e-05,
"loss": 1.1438,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 1.2383153211067572,
"learning_rate": 2.8256823490265424e-05,
"loss": 1.2114,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 1.1023021462273042,
"learning_rate": 2.8205280699152515e-05,
"loss": 1.2125,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 1.5901982656566034,
"learning_rate": 2.8153672298833777e-05,
"loss": 1.1806,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 1.133898835373747,
"learning_rate": 2.8101998701968615e-05,
"loss": 1.1996,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 1.6221125618700756,
"learning_rate": 2.805026032173773e-05,
"loss": 1.1985,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 1.4749530139180094,
"learning_rate": 2.7998457571839832e-05,
"loss": 1.2186,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 1.0940244982754717,
"learning_rate": 2.7946590866488322e-05,
"loss": 1.2088,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 1.4052139615873853,
"learning_rate": 2.7894660620408e-05,
"loss": 1.1585,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 1.111726335131992,
"learning_rate": 2.784266724883173e-05,
"loss": 1.2116,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.7947471268986862,
"learning_rate": 2.7790611167497128e-05,
"loss": 1.2003,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.8494862963445763,
"learning_rate": 2.7738492792643224e-05,
"loss": 1.1814,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.8868932210449914,
"learning_rate": 2.768631254100716e-05,
"loss": 1.2121,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.9777272756122515,
"learning_rate": 2.7634070829820836e-05,
"loss": 1.1961,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 1.0353930949209826,
"learning_rate": 2.7581768076807592e-05,
"loss": 1.1796,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 1.1794744375305817,
"learning_rate": 2.7529404700178832e-05,
"loss": 1.2098,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.9564504833383075,
"learning_rate": 2.7476981118630735e-05,
"loss": 1.1833,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 1.0670019572026488,
"learning_rate": 2.742449775134085e-05,
"loss": 1.1991,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 1.0931499316588882,
"learning_rate": 2.7371955017964788e-05,
"loss": 1.2096,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.8941695068654483,
"learning_rate": 2.7319353338632845e-05,
"loss": 1.1923,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.6902904624698248,
"learning_rate": 2.7266693133946645e-05,
"loss": 1.2033,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 1.0742241841696913,
"learning_rate": 2.7213974824975785e-05,
"loss": 1.1767,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.7244543183018752,
"learning_rate": 2.7161198833254464e-05,
"loss": 1.1851,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.8526321610003282,
"learning_rate": 2.7108365580778095e-05,
"loss": 1.1904,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.5297015243767798,
"learning_rate": 2.705547548999996e-05,
"loss": 1.1985,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.728419155486756,
"learning_rate": 2.7002528983827817e-05,
"loss": 1.1974,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.8757971946822691,
"learning_rate": 2.6949526485620517e-05,
"loss": 1.1629,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 1.0471753281021923,
"learning_rate": 2.6896468419184623e-05,
"loss": 1.177,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 1.137219475631692,
"learning_rate": 2.6843355208771024e-05,
"loss": 1.2323,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.7226408024498755,
"learning_rate": 2.6790187279071536e-05,
"loss": 1.1965,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.8188695051977395,
"learning_rate": 2.67369650552155e-05,
"loss": 1.1869,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 1.3621691448942137,
"learning_rate": 2.6683688962766407e-05,
"loss": 1.1422,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.711224391862871,
"learning_rate": 2.6630359427718468e-05,
"loss": 1.2051,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.8436046848617059,
"learning_rate": 2.6576976876493225e-05,
"loss": 1.1937,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 1.070642339912311,
"learning_rate": 2.652354173593614e-05,
"loss": 1.1749,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.7553034489789205,
"learning_rate": 2.647005443331317e-05,
"loss": 1.2106,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 1.0488732321866254,
"learning_rate": 2.6416515396307354e-05,
"loss": 1.2393,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.9256185320084538,
"learning_rate": 2.636292505301542e-05,
"loss": 1.2116,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.6899690243739529,
"learning_rate": 2.630928383194431e-05,
"loss": 1.1952,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.9475786520111134,
"learning_rate": 2.625559216200781e-05,
"loss": 1.2322,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.8593426499964417,
"learning_rate": 2.620185047252308e-05,
"loss": 1.1836,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.6389258255149758,
"learning_rate": 2.614805919320723e-05,
"loss": 1.1894,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.6199821757632893,
"learning_rate": 2.6094218754173895e-05,
"loss": 1.2201,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.727859826298201,
"learning_rate": 2.60403295859298e-05,
"loss": 1.2102,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.8930601130544346,
"learning_rate": 2.5986392119371297e-05,
"loss": 1.2437,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 1.0121102244065783,
"learning_rate": 2.5932406785780928e-05,
"loss": 1.1996,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.990686561693887,
"learning_rate": 2.5878374016823987e-05,
"loss": 1.2106,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.8304480694717532,
"learning_rate": 2.5824294244545045e-05,
"loss": 1.1816,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.8605255139007713,
"learning_rate": 2.577016790136453e-05,
"loss": 1.2195,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.5744146387908088,
"learning_rate": 2.5715995420075225e-05,
"loss": 1.2042,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.8328201839717123,
"learning_rate": 2.5661777233838864e-05,
"loss": 1.2058,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 1.1450266035563619,
"learning_rate": 2.5607513776182616e-05,
"loss": 1.2076,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.6324502935566931,
"learning_rate": 2.5553205480995633e-05,
"loss": 1.1929,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.698029978483187,
"learning_rate": 2.5498852782525604e-05,
"loss": 1.2118,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.6038621881644255,
"learning_rate": 2.5444456115375252e-05,
"loss": 1.2132,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.6509383029180433,
"learning_rate": 2.539001591449889e-05,
"loss": 1.256,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.7543214488892146,
"learning_rate": 2.53355326151989e-05,
"loss": 1.1915,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.6686725764339347,
"learning_rate": 2.5281006653122288e-05,
"loss": 1.1829,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.7198420037543756,
"learning_rate": 2.5226438464257217e-05,
"loss": 1.2007,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.9756831391965428,
"learning_rate": 2.5171828484929447e-05,
"loss": 1.2073,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 1.0843629904936434,
"learning_rate": 2.5117177151798934e-05,
"loss": 1.1704,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.7416808663990008,
"learning_rate": 2.506248490185627e-05,
"loss": 1.1854,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.7229683643959104,
"learning_rate": 2.500775217241923e-05,
"loss": 1.1901,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 1.0913709072572606,
"learning_rate": 2.4952979401129266e-05,
"loss": 1.2156,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 1.2162122421271846,
"learning_rate": 2.4898167025948e-05,
"loss": 1.2183,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.7665650927346409,
"learning_rate": 2.4843315485153703e-05,
"loss": 1.1923,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.6356326863147574,
"learning_rate": 2.4788425217337856e-05,
"loss": 1.1906,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.9562391138937477,
"learning_rate": 2.4733496661401572e-05,
"loss": 1.1901,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 1.1283988550051705,
"learning_rate": 2.4678530256552114e-05,
"loss": 1.1831,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.8555470670639433,
"learning_rate": 2.46235264422994e-05,
"loss": 1.2163,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.677378106464659,
"learning_rate": 2.4568485658452452e-05,
"loss": 1.1809,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.7946985594860313,
"learning_rate": 2.451340834511592e-05,
"loss": 1.2175,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.9027232380473076,
"learning_rate": 2.4458294942686528e-05,
"loss": 1.1852,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.6654569466788964,
"learning_rate": 2.440314589184958e-05,
"loss": 1.2007,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.5581708703268474,
"learning_rate": 2.434796163357541e-05,
"loss": 1.1928,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.5831693796136088,
"learning_rate": 2.429274260911588e-05,
"loss": 1.1942,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5711615616140153,
"learning_rate": 2.423748926000084e-05,
"loss": 1.2199,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.6497820470721396,
"learning_rate": 2.418220202803459e-05,
"loss": 1.211,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.7153497139268538,
"learning_rate": 2.412688135529237e-05,
"loss": 1.1779,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.8434980539434074,
"learning_rate": 2.4071527684116807e-05,
"loss": 1.155,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.8201260102124509,
"learning_rate": 2.4016141457114372e-05,
"loss": 1.2475,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.6197916284894635,
"learning_rate": 2.396072311715187e-05,
"loss": 1.2221,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.8852844675640947,
"learning_rate": 2.3905273107352863e-05,
"loss": 1.1786,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.7357620579118229,
"learning_rate": 2.3849791871094157e-05,
"loss": 1.2182,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.8531580429201988,
"learning_rate": 2.379427985200223e-05,
"loss": 1.2053,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.6423270818187268,
"learning_rate": 2.373873749394972e-05,
"loss": 1.2203,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.8178657160000015,
"learning_rate": 2.3683165241051836e-05,
"loss": 1.153,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.9284626056741804,
"learning_rate": 2.3627563537662823e-05,
"loss": 1.2237,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.6273775375337646,
"learning_rate": 2.3571932828372433e-05,
"loss": 1.2089,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.8303629363725178,
"learning_rate": 2.351627355800233e-05,
"loss": 1.2197,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.7005300013860735,
"learning_rate": 2.346058617160256e-05,
"loss": 1.1577,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.7123770735468917,
"learning_rate": 2.3404871114447976e-05,
"loss": 1.2007,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.8062572114700841,
"learning_rate": 2.334912883203469e-05,
"loss": 1.2186,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.8157856719698788,
"learning_rate": 2.3293359770076505e-05,
"loss": 1.1844,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.766435175745772,
"learning_rate": 2.323756437450136e-05,
"loss": 1.175,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.9062264346574604,
"learning_rate": 2.3181743091447754e-05,
"loss": 1.2223,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.6432970853468856,
"learning_rate": 2.3125896367261164e-05,
"loss": 1.2269,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.9641808763943207,
"learning_rate": 2.307002464849052e-05,
"loss": 1.1995,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.7582825097947017,
"learning_rate": 2.30141283818846e-05,
"loss": 1.2286,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.6247443462995386,
"learning_rate": 2.2958208014388462e-05,
"loss": 1.2325,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.7195063473779657,
"learning_rate": 2.2902263993139857e-05,
"loss": 1.1882,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.539570058690455,
"learning_rate": 2.2846296765465708e-05,
"loss": 1.215,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.7713397298274007,
"learning_rate": 2.2790306778878465e-05,
"loss": 1.2368,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.593206514968184,
"learning_rate": 2.273429448107257e-05,
"loss": 1.217,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.6746092071412564,
"learning_rate": 2.2678260319920868e-05,
"loss": 1.1949,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.624828065292495,
"learning_rate": 2.262220474347102e-05,
"loss": 1.2041,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.5282739528071027,
"learning_rate": 2.2566128199941918e-05,
"loss": 1.1852,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.5509884853271455,
"learning_rate": 2.2510031137720104e-05,
"loss": 1.2201,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 2.4086519346691753,
"learning_rate": 2.2453914005356206e-05,
"loss": 1.3057,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.622408242615193,
"learning_rate": 2.2397777251561302e-05,
"loss": 1.1933,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.6641502919308558,
"learning_rate": 2.2341621325203393e-05,
"loss": 1.1782,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.6014499085837061,
"learning_rate": 2.2285446675303764e-05,
"loss": 1.2239,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.6761047654062722,
"learning_rate": 2.2229253751033423e-05,
"loss": 1.1981,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.531065151262307,
"learning_rate": 2.2173043001709492e-05,
"loss": 1.1702,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.5261122951335152,
"learning_rate": 2.2116814876791627e-05,
"loss": 1.1774,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.5148238731245197,
"learning_rate": 2.2060569825878413e-05,
"loss": 1.1789,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.49668666675142265,
"learning_rate": 2.2004308298703777e-05,
"loss": 1.2003,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.5043627193604788,
"learning_rate": 2.1948030745133396e-05,
"loss": 1.1948,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.47908784361824874,
"learning_rate": 2.1891737615161083e-05,
"loss": 1.2002,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.7268259487403044,
"learning_rate": 2.1835429358905218e-05,
"loss": 1.206,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.7070628480539952,
"learning_rate": 2.17791064266051e-05,
"loss": 1.1662,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.8590903890360289,
"learning_rate": 2.172276926861741e-05,
"loss": 1.2274,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.6901377861344234,
"learning_rate": 2.166641833541255e-05,
"loss": 1.1856,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.8450858373288503,
"learning_rate": 2.161005407757109e-05,
"loss": 1.2104,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.7620558694567359,
"learning_rate": 2.155367694578013e-05,
"loss": 1.1762,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.525247295092508,
"learning_rate": 2.1497287390829712e-05,
"loss": 1.2163,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.7274600707812038,
"learning_rate": 2.1440885863609208e-05,
"loss": 1.2261,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.5556955416370187,
"learning_rate": 2.1384472815103737e-05,
"loss": 1.2105,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.683234756094595,
"learning_rate": 2.1328048696390534e-05,
"loss": 1.2154,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.55005350840378,
"learning_rate": 2.1271613958635342e-05,
"loss": 1.2056,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.6226597200000944,
"learning_rate": 2.1215169053088825e-05,
"loss": 1.2052,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.5418338301402982,
"learning_rate": 2.1158714431082937e-05,
"loss": 1.1821,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.6348947707024848,
"learning_rate": 2.110225054402733e-05,
"loss": 1.1765,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.5975126636780733,
"learning_rate": 2.1045777843405747e-05,
"loss": 1.2081,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.5215801897157224,
"learning_rate": 2.09892967807724e-05,
"loss": 1.2056,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.7378195222108167,
"learning_rate": 2.093280780774835e-05,
"loss": 1.1897,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.7105340229006941,
"learning_rate": 2.087631137601793e-05,
"loss": 1.1616,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.5539875711362283,
"learning_rate": 2.0819807937325104e-05,
"loss": 1.2089,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.5798057665895844,
"learning_rate": 2.0763297943469847e-05,
"loss": 1.2001,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.7159010314073236,
"learning_rate": 2.070678184630458e-05,
"loss": 1.1803,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.7243357244850688,
"learning_rate": 2.06502600977305e-05,
"loss": 1.2132,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.5937879520004954,
"learning_rate": 2.0593733149694008e-05,
"loss": 1.2161,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.5897685902854041,
"learning_rate": 2.0537201454183074e-05,
"loss": 1.1821,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.5611735781948396,
"learning_rate": 2.0480665463223626e-05,
"loss": 1.19,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.6174202122993415,
"learning_rate": 2.042412562887594e-05,
"loss": 1.1953,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.5661091260362169,
"learning_rate": 2.0367582403231033e-05,
"loss": 1.2211,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.5970333615142671,
"learning_rate": 2.0311036238407023e-05,
"loss": 1.2012,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 1.0322119557155531,
"learning_rate": 2.0254487586545534e-05,
"loss": 1.1939,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.6528426378147218,
"learning_rate": 2.0197936899808088e-05,
"loss": 1.196,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.9266775858300897,
"learning_rate": 2.0141384630372466e-05,
"loss": 1.2092,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.9463777664009674,
"learning_rate": 2.0084831230429106e-05,
"loss": 1.1779,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.6376020536124962,
"learning_rate": 2.0028277152177492e-05,
"loss": 1.2063,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 1.0077099786105035,
"learning_rate": 1.9971722847822518e-05,
"loss": 1.1765,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.7902236185457001,
"learning_rate": 1.99151687695709e-05,
"loss": 1.2145,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.895558029914492,
"learning_rate": 1.9858615369627537e-05,
"loss": 1.2151,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 1.0582475493388663,
"learning_rate": 1.9802063100191916e-05,
"loss": 1.226,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.999406554942165,
"learning_rate": 1.974551241345447e-05,
"loss": 1.1884,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 1.2532371501605712,
"learning_rate": 1.9688963761592984e-05,
"loss": 1.2028,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 1.1044366164250865,
"learning_rate": 1.963241759676897e-05,
"loss": 1.2392,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 1.3169369348652962,
"learning_rate": 1.9575874371124066e-05,
"loss": 1.164,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 1.2076453699129064,
"learning_rate": 1.951933453677638e-05,
"loss": 1.2437,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 1.3554763084400019,
"learning_rate": 1.9462798545816932e-05,
"loss": 1.1933,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 1.0651032721907039,
"learning_rate": 1.9406266850305996e-05,
"loss": 1.1773,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 1.2075607339892538,
"learning_rate": 1.9349739902269503e-05,
"loss": 1.1816,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 1.039440510587944,
"learning_rate": 1.9293218153695425e-05,
"loss": 1.1803,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 1.1476935096664538,
"learning_rate": 1.9236702056530153e-05,
"loss": 1.172,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.9131712210162294,
"learning_rate": 1.9180192062674907e-05,
"loss": 1.1827,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.9897752922407875,
"learning_rate": 1.9123688623982076e-05,
"loss": 1.1764,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.821372357847034,
"learning_rate": 1.906719219225165e-05,
"loss": 1.2154,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 1.0463797637474601,
"learning_rate": 1.9010703219227603e-05,
"loss": 1.215,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.6416566603650623,
"learning_rate": 1.8954222156594256e-05,
"loss": 1.2143,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.7495139874512576,
"learning_rate": 1.8897749455972673e-05,
"loss": 1.1824,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.7739636128889718,
"learning_rate": 1.884128556891707e-05,
"loss": 1.2007,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.7175126330289512,
"learning_rate": 1.878483094691119e-05,
"loss": 1.2065,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.7228716287157573,
"learning_rate": 1.8728386041364664e-05,
"loss": 1.2069,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.6881205476102028,
"learning_rate": 1.867195130360947e-05,
"loss": 1.1799,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.5656871441735614,
"learning_rate": 1.8615527184896263e-05,
"loss": 1.1977,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.7308792917577789,
"learning_rate": 1.8559114136390795e-05,
"loss": 1.2135,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.5572116714715041,
"learning_rate": 1.8502712609170298e-05,
"loss": 1.2006,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.5177716732588655,
"learning_rate": 1.8446323054219876e-05,
"loss": 1.2042,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.6026318039057721,
"learning_rate": 1.838994592242891e-05,
"loss": 1.1994,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.5671881792168806,
"learning_rate": 1.8333581664587453e-05,
"loss": 1.2322,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.7340920217307121,
"learning_rate": 1.8277230731382593e-05,
"loss": 1.1728,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.6340649490185639,
"learning_rate": 1.82208935733949e-05,
"loss": 1.162,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.6371933269675838,
"learning_rate": 1.8164570641094793e-05,
"loss": 1.1806,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.587557369152816,
"learning_rate": 1.810826238483892e-05,
"loss": 1.1786,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.6415152215536124,
"learning_rate": 1.8051969254866608e-05,
"loss": 1.2045,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.4842004436730326,
"learning_rate": 1.7995691701296226e-05,
"loss": 1.2004,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.6088180239719989,
"learning_rate": 1.7939430174121593e-05,
"loss": 1.1789,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.5298692333628027,
"learning_rate": 1.788318512320838e-05,
"loss": 1.1905,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.6301358310669689,
"learning_rate": 1.782695699829051e-05,
"loss": 1.2,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.549415520075841,
"learning_rate": 1.7770746248966587e-05,
"loss": 1.2132,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.5634815212197226,
"learning_rate": 1.7714553324696243e-05,
"loss": 1.1776,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.766539133400323,
"learning_rate": 1.7658378674796614e-05,
"loss": 1.1982,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.459123683854083,
"learning_rate": 1.76022227484387e-05,
"loss": 1.2056,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.6588283016969438,
"learning_rate": 1.7546085994643807e-05,
"loss": 1.1717,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.501965988998357,
"learning_rate": 1.7489968862279902e-05,
"loss": 1.201,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 2.2584127585048015,
"learning_rate": 1.743387180005809e-05,
"loss": 1.2527,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.6307226536062482,
"learning_rate": 1.737779525652899e-05,
"loss": 1.1889,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.7083447596451333,
"learning_rate": 1.732173968007914e-05,
"loss": 1.2034,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.48967383195039715,
"learning_rate": 1.7265705518927437e-05,
"loss": 1.2126,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.45944795777961045,
"learning_rate": 1.7209693221121542e-05,
"loss": 1.215,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.585942632033885,
"learning_rate": 1.7153703234534302e-05,
"loss": 1.1868,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.49675925621210143,
"learning_rate": 1.709773600686015e-05,
"loss": 1.2041,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.6807334967871748,
"learning_rate": 1.7041791985611548e-05,
"loss": 1.1841,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.5329096909111797,
"learning_rate": 1.69858716181154e-05,
"loss": 1.2109,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.5843676468996211,
"learning_rate": 1.692997535150948e-05,
"loss": 1.2054,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.5787305651355045,
"learning_rate": 1.687410363273884e-05,
"loss": 1.2007,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.4435037571089406,
"learning_rate": 1.6818256908552257e-05,
"loss": 1.1931,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.5684074233517431,
"learning_rate": 1.6762435625498646e-05,
"loss": 1.2099,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.47533378555789,
"learning_rate": 1.67066402299235e-05,
"loss": 1.1769,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.511904818247405,
"learning_rate": 1.6650871167965313e-05,
"loss": 1.1847,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.5290521029677894,
"learning_rate": 1.6595128885552028e-05,
"loss": 1.1532,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.47708120039189045,
"learning_rate": 1.653941382839745e-05,
"loss": 1.201,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.528925931985692,
"learning_rate": 1.6483726441997673e-05,
"loss": 1.1639,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.5945670352737704,
"learning_rate": 1.642806717162757e-05,
"loss": 1.1462,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.5037696365482758,
"learning_rate": 1.637243646233718e-05,
"loss": 1.1068,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.5792936056973678,
"learning_rate": 1.6316834758948174e-05,
"loss": 1.1233,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.5739407854902231,
"learning_rate": 1.6261262506050282e-05,
"loss": 1.1511,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.6042258923340935,
"learning_rate": 1.620572014799777e-05,
"loss": 1.1255,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.6014679630187104,
"learning_rate": 1.6150208128905857e-05,
"loss": 1.122,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.6582239706377584,
"learning_rate": 1.6094726892647147e-05,
"loss": 1.1249,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.7165518039492543,
"learning_rate": 1.6039276882848135e-05,
"loss": 1.164,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.61188849852577,
"learning_rate": 1.5983858542885635e-05,
"loss": 1.1373,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.5287352713547551,
"learning_rate": 1.5928472315883203e-05,
"loss": 1.1145,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.4885272739564263,
"learning_rate": 1.5873118644707633e-05,
"loss": 1.1166,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.47880721814062055,
"learning_rate": 1.5817797971965413e-05,
"loss": 1.12,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.4947658907098916,
"learning_rate": 1.576251073999917e-05,
"loss": 1.153,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.4563887192130925,
"learning_rate": 1.5707257390884126e-05,
"loss": 1.153,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.5961251594056004,
"learning_rate": 1.5652038366424595e-05,
"loss": 1.1334,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.47865692918038283,
"learning_rate": 1.5596854108150424e-05,
"loss": 1.1421,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.48079762921721564,
"learning_rate": 1.5541705057313476e-05,
"loss": 1.1025,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.5369193043571898,
"learning_rate": 1.5486591654884086e-05,
"loss": 1.0974,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.5090192659517753,
"learning_rate": 1.543151434154755e-05,
"loss": 1.1089,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.43103742072211554,
"learning_rate": 1.5376473557700615e-05,
"loss": 1.1309,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.46356089837250564,
"learning_rate": 1.5321469743447893e-05,
"loss": 1.1127,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.42441927487456327,
"learning_rate": 1.5266503338598434e-05,
"loss": 1.1545,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.5272277773152224,
"learning_rate": 1.5211574782662149e-05,
"loss": 1.1303,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.4118279896032549,
"learning_rate": 1.5156684514846307e-05,
"loss": 1.1003,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.47586945862080127,
"learning_rate": 1.5101832974052012e-05,
"loss": 1.1717,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.4824310863607328,
"learning_rate": 1.5047020598870737e-05,
"loss": 1.1536,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.5187236978311888,
"learning_rate": 1.4992247827580778e-05,
"loss": 1.1238,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.47337037241542157,
"learning_rate": 1.493751509814374e-05,
"loss": 1.1052,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.5917595725753492,
"learning_rate": 1.4882822848201073e-05,
"loss": 1.1093,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.49580012214953845,
"learning_rate": 1.4828171515070553e-05,
"loss": 1.147,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.4875786884341751,
"learning_rate": 1.4773561535742793e-05,
"loss": 1.0984,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.49241106022007547,
"learning_rate": 1.4718993346877715e-05,
"loss": 1.1,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.4388538601501506,
"learning_rate": 1.466446738480111e-05,
"loss": 1.1444,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.5021984672759217,
"learning_rate": 1.460998408550112e-05,
"loss": 1.1606,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.42040370158459855,
"learning_rate": 1.4555543884624751e-05,
"loss": 1.1253,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.5322498406092719,
"learning_rate": 1.4501147217474402e-05,
"loss": 1.1212,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.43511811127078853,
"learning_rate": 1.444679451900437e-05,
"loss": 1.1571,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.5353231719544004,
"learning_rate": 1.4392486223817397e-05,
"loss": 1.1119,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.5991353202333854,
"learning_rate": 1.4338222766161143e-05,
"loss": 1.1272,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.4696052715639886,
"learning_rate": 1.4284004579924777e-05,
"loss": 1.1333,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.6288444054627907,
"learning_rate": 1.4229832098635479e-05,
"loss": 1.1569,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.39143015291753863,
"learning_rate": 1.4175705755454963e-05,
"loss": 1.1584,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.4504715518518521,
"learning_rate": 1.412162598317602e-05,
"loss": 1.1192,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.4561547289479086,
"learning_rate": 1.406759321421907e-05,
"loss": 1.146,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.43533326308351045,
"learning_rate": 1.401360788062871e-05,
"loss": 1.1677,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.4908602127349103,
"learning_rate": 1.3959670414070204e-05,
"loss": 1.1318,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.40639556225922663,
"learning_rate": 1.3905781245826108e-05,
"loss": 1.1474,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.41500060837316893,
"learning_rate": 1.3851940806792778e-05,
"loss": 1.1432,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.5274946256649654,
"learning_rate": 1.379814952747693e-05,
"loss": 1.1443,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.5022159298171385,
"learning_rate": 1.3744407837992193e-05,
"loss": 1.1256,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.47095674348543093,
"learning_rate": 1.3690716168055692e-05,
"loss": 1.1293,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.4810420030910108,
"learning_rate": 1.363707494698459e-05,
"loss": 1.1571,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.41766396364015174,
"learning_rate": 1.358348460369265e-05,
"loss": 1.099,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.48142846929604194,
"learning_rate": 1.352994556668684e-05,
"loss": 1.1295,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.4727154845114651,
"learning_rate": 1.3476458264063865e-05,
"loss": 1.1552,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.47950228997876915,
"learning_rate": 1.342302312350678e-05,
"loss": 1.1085,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.4188324871754431,
"learning_rate": 1.3369640572281537e-05,
"loss": 1.095,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.5292165599480817,
"learning_rate": 1.3316311037233596e-05,
"loss": 1.1612,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.4619571982781207,
"learning_rate": 1.326303494478451e-05,
"loss": 1.1321,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.5400180370962684,
"learning_rate": 1.3209812720928472e-05,
"loss": 1.1352,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.4584441106875985,
"learning_rate": 1.315664479122898e-05,
"loss": 1.1171,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.42512835855203085,
"learning_rate": 1.3103531580815378e-05,
"loss": 1.1376,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.4213668247790607,
"learning_rate": 1.3050473514379488e-05,
"loss": 1.1208,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.45857511092901704,
"learning_rate": 1.2997471016172188e-05,
"loss": 1.1598,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.4044971538015996,
"learning_rate": 1.2944524510000042e-05,
"loss": 1.1243,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.4961659221983083,
"learning_rate": 1.2891634419221908e-05,
"loss": 1.1295,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.44474002364803905,
"learning_rate": 1.2838801166745545e-05,
"loss": 1.0971,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.4232879449847028,
"learning_rate": 1.2786025175024216e-05,
"loss": 1.144,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.43712123306912304,
"learning_rate": 1.2733306866053357e-05,
"loss": 1.1451,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.40180488939661685,
"learning_rate": 1.2680646661367163e-05,
"loss": 1.0894,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.45233805007640054,
"learning_rate": 1.2628044982035219e-05,
"loss": 1.1339,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.40818809352254937,
"learning_rate": 1.2575502248659155e-05,
"loss": 1.1143,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.4645391529208088,
"learning_rate": 1.2523018881369269e-05,
"loss": 1.1232,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.4214856935779652,
"learning_rate": 1.2470595299821176e-05,
"loss": 1.1368,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.4205581634407771,
"learning_rate": 1.2418231923192416e-05,
"loss": 1.1377,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.38200446309696945,
"learning_rate": 1.2365929170179165e-05,
"loss": 1.1239,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.4679977963050786,
"learning_rate": 1.2313687458992848e-05,
"loss": 1.1553,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.46239908615956354,
"learning_rate": 1.2261507207356785e-05,
"loss": 1.1238,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.3977963173826309,
"learning_rate": 1.220938883250288e-05,
"loss": 1.1303,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.46588687692362296,
"learning_rate": 1.215733275116827e-05,
"loss": 1.1285,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.4079344937416464,
"learning_rate": 1.2105339379592005e-05,
"loss": 1.1321,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.37572796780492185,
"learning_rate": 1.2053409133511681e-05,
"loss": 1.1058,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.38974028906576647,
"learning_rate": 1.2001542428160173e-05,
"loss": 1.1529,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.37306228293208865,
"learning_rate": 1.1949739678262278e-05,
"loss": 1.1349,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.46676477244134756,
"learning_rate": 1.189800129803139e-05,
"loss": 1.11,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.44227824622131834,
"learning_rate": 1.1846327701166228e-05,
"loss": 1.1108,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.43755434003695454,
"learning_rate": 1.1794719300847493e-05,
"loss": 1.1271,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.43895730717836173,
"learning_rate": 1.1743176509734582e-05,
"loss": 1.1729,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.41106197609452877,
"learning_rate": 1.1691699739962275e-05,
"loss": 1.1288,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.4049760868587107,
"learning_rate": 1.1640289403137456e-05,
"loss": 1.1496,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.3537380253715832,
"learning_rate": 1.1588945910335802e-05,
"loss": 1.1264,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 1.7586630454623104,
"learning_rate": 1.1537669672098517e-05,
"loss": 1.1845,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.5094648100351359,
"learning_rate": 1.1486461098429036e-05,
"loss": 1.1647,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.3898846572838452,
"learning_rate": 1.1435320598789747e-05,
"loss": 1.1341,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.5097898835666126,
"learning_rate": 1.1384248582098736e-05,
"loss": 1.1341,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.47717304648250763,
"learning_rate": 1.1333245456726467e-05,
"loss": 1.1258,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.49158093683131365,
"learning_rate": 1.128231163049258e-05,
"loss": 1.113,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.6173311965263752,
"learning_rate": 1.123144751066259e-05,
"loss": 1.134,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.3983763639656086,
"learning_rate": 1.1180653503944657e-05,
"loss": 1.1054,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.597653494737724,
"learning_rate": 1.1129930016486287e-05,
"loss": 1.1427,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.4542837876201974,
"learning_rate": 1.107927745387114e-05,
"loss": 1.1152,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.558383506075408,
"learning_rate": 1.1028696221115766e-05,
"loss": 1.1013,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.4972947741590551,
"learning_rate": 1.0978186722666341e-05,
"loss": 1.1163,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.4886071923433344,
"learning_rate": 1.0927749362395475e-05,
"loss": 1.1491,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.5354496474584195,
"learning_rate": 1.0877384543598954e-05,
"loss": 1.1159,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.48176829706933133,
"learning_rate": 1.0827092668992529e-05,
"loss": 1.1366,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.5654800663973405,
"learning_rate": 1.077687414070869e-05,
"loss": 1.139,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.49422738606545336,
"learning_rate": 1.0726729360293451e-05,
"loss": 1.1474,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.4962938974768986,
"learning_rate": 1.0676658728703147e-05,
"loss": 1.1111,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.4635283285638908,
"learning_rate": 1.0626662646301198e-05,
"loss": 1.1233,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.5043480629388909,
"learning_rate": 1.0576741512854958e-05,
"loss": 1.1336,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.41347363717152413,
"learning_rate": 1.052689572753247e-05,
"loss": 1.1579,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.505249798433399,
"learning_rate": 1.0477125688899328e-05,
"loss": 1.1137,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.4376680566066679,
"learning_rate": 1.0427431794915406e-05,
"loss": 1.1342,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.4229599391528869,
"learning_rate": 1.0377814442931769e-05,
"loss": 1.1262,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.44071898085755723,
"learning_rate": 1.0328274029687439e-05,
"loss": 1.1346,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.42248467275127294,
"learning_rate": 1.0278810951306253e-05,
"loss": 1.1534,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.43255828932736956,
"learning_rate": 1.0229425603293647e-05,
"loss": 1.1312,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.46007981510987794,
"learning_rate": 1.0180118380533558e-05,
"loss": 1.1764,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.45551693297370094,
"learning_rate": 1.0130889677285225e-05,
"loss": 1.1267,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.37574096648281863,
"learning_rate": 1.008173988718005e-05,
"loss": 1.1635,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.4156648717702261,
"learning_rate": 1.0032669403218439e-05,
"loss": 1.1479,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.38470017636444737,
"learning_rate": 9.983678617766677e-06,
"loss": 1.1341,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.3800441817508743,
"learning_rate": 9.934767922553788e-06,
"loss": 1.1563,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.42801581111525655,
"learning_rate": 9.885937708668371e-06,
"loss": 1.1548,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.44979630272666193,
"learning_rate": 9.837188366555524e-06,
"loss": 1.1478,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.40209170649028697,
"learning_rate": 9.788520286013691e-06,
"loss": 1.1231,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.34100388500202833,
"learning_rate": 9.739933856191537e-06,
"loss": 1.1279,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.5021081114837468,
"learning_rate": 9.691429465584866e-06,
"loss": 1.1601,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.3785097589808829,
"learning_rate": 9.643007502033492e-06,
"loss": 1.1332,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.4127882197570853,
"learning_rate": 9.594668352718152e-06,
"loss": 1.1316,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.37969450007649913,
"learning_rate": 9.546412404157383e-06,
"loss": 1.132,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.47745351116222123,
"learning_rate": 9.498240042204471e-06,
"loss": 1.1291,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.39265342175379697,
"learning_rate": 9.450151652044339e-06,
"loss": 1.1516,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.450631562090042,
"learning_rate": 9.402147618190472e-06,
"loss": 1.107,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.3383324438313689,
"learning_rate": 9.354228324481841e-06,
"loss": 1.1092,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.47588495699910666,
"learning_rate": 9.306394154079852e-06,
"loss": 1.1443,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.3468624109745931,
"learning_rate": 9.258645489465254e-06,
"loss": 1.1415,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.3908261875360874,
"learning_rate": 9.21098271243509e-06,
"loss": 1.1251,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.4131640378514314,
"learning_rate": 9.163406204099659e-06,
"loss": 1.1265,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.360498313607415,
"learning_rate": 9.115916344879452e-06,
"loss": 1.1246,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 1.060898913833296,
"learning_rate": 9.068513514502126e-06,
"loss": 1.126,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.43780027598773785,
"learning_rate": 9.021198091999439e-06,
"loss": 1.1551,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3858889991379542,
"learning_rate": 8.973970455704251e-06,
"loss": 1.1658,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.4451604738412859,
"learning_rate": 8.926830983247495e-06,
"loss": 1.1601,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.41316048134234395,
"learning_rate": 8.879780051555116e-06,
"loss": 1.1607,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.46316091513986407,
"learning_rate": 8.83281803684512e-06,
"loss": 1.1264,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.3818283388208469,
"learning_rate": 8.785945314624532e-06,
"loss": 1.0941,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.43676223852562246,
"learning_rate": 8.739162259686385e-06,
"loss": 1.1503,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.4776160999454151,
"learning_rate": 8.692469246106734e-06,
"loss": 1.1031,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.4215675688509712,
"learning_rate": 8.645866647241677e-06,
"loss": 1.1328,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.41023998438762965,
"learning_rate": 8.599354835724348e-06,
"loss": 1.173,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.3737604988983061,
"learning_rate": 8.552934183461958e-06,
"loss": 1.1275,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.3335079148164336,
"learning_rate": 8.506605061632805e-06,
"loss": 1.1385,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.3774607215667267,
"learning_rate": 8.460367840683308e-06,
"loss": 1.0874,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.3906431518887843,
"learning_rate": 8.414222890325065e-06,
"loss": 1.1082,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.3643537587386391,
"learning_rate": 8.368170579531855e-06,
"loss": 1.1361,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3621022532266236,
"learning_rate": 8.322211276536733e-06,
"loss": 1.1412,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.414024943976658,
"learning_rate": 8.27634534882907e-06,
"loss": 1.1389,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.39596707907910517,
"learning_rate": 8.230573163151585e-06,
"loss": 1.1341,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.3747556597484249,
"learning_rate": 8.184895085497466e-06,
"loss": 1.175,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.41528921961556636,
"learning_rate": 8.139311481107401e-06,
"loss": 1.156,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.41236153988163143,
"learning_rate": 8.093822714466686e-06,
"loss": 1.1662,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.41413581465637206,
"learning_rate": 8.048429149302271e-06,
"loss": 1.136,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.4669644014237305,
"learning_rate": 8.003131148579902e-06,
"loss": 1.1238,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.3631870728321623,
"learning_rate": 7.957929074501182e-06,
"loss": 1.1342,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.43825209616054733,
"learning_rate": 7.912823288500695e-06,
"loss": 1.1049,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.49812195533021525,
"learning_rate": 7.867814151243102e-06,
"loss": 1.1047,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.4036596430268589,
"learning_rate": 7.822902022620267e-06,
"loss": 1.1447,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.48210437062594597,
"learning_rate": 7.778087261748377e-06,
"loss": 1.1173,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.46002050264704447,
"learning_rate": 7.733370226965056e-06,
"loss": 1.1269,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.38367590914498734,
"learning_rate": 7.688751275826527e-06,
"loss": 1.1417,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.44177990727102046,
"learning_rate": 7.644230765104735e-06,
"loss": 1.1346,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.4270835434638795,
"learning_rate": 7.599809050784502e-06,
"loss": 1.1239,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.4132155190629358,
"learning_rate": 7.555486488060661e-06,
"loss": 1.1549,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.4039889836241822,
"learning_rate": 7.511263431335249e-06,
"loss": 1.1423,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.35140237507177285,
"learning_rate": 7.467140234214656e-06,
"loss": 1.1155,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.387594823062278,
"learning_rate": 7.423117249506777e-06,
"loss": 1.1303,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.35566250444862907,
"learning_rate": 7.379194829218231e-06,
"loss": 1.1033,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.40361031451649837,
"learning_rate": 7.335373324551518e-06,
"loss": 1.163,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.3938637771535003,
"learning_rate": 7.291653085902224e-06,
"loss": 1.1764,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.42625117851798605,
"learning_rate": 7.248034462856213e-06,
"loss": 1.1463,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.354972125473517,
"learning_rate": 7.20451780418683e-06,
"loss": 1.1128,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.3859514818558594,
"learning_rate": 7.161103457852128e-06,
"loss": 1.1303,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.38212015762716167,
"learning_rate": 7.117791770992046e-06,
"loss": 1.1141,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.36771953743690505,
"learning_rate": 7.074583089925688e-06,
"loss": 1.1366,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.3623826952872305,
"learning_rate": 7.0314777601485155e-06,
"loss": 1.1228,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.32632027676041925,
"learning_rate": 6.988476126329604e-06,
"loss": 1.0813,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3368575922482779,
"learning_rate": 6.945578532308857e-06,
"loss": 1.1301,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.4021967356525026,
"learning_rate": 6.902785321094301e-06,
"loss": 1.0945,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.3446591080024311,
"learning_rate": 6.8600968348593115e-06,
"loss": 1.175,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.3977833616226488,
"learning_rate": 6.817513414939892e-06,
"loss": 1.1228,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.41308185418475,
"learning_rate": 6.775035401831924e-06,
"loss": 1.112,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.3484454400181044,
"learning_rate": 6.732663135188473e-06,
"loss": 1.15,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.539086589872215,
"learning_rate": 6.690396953817055e-06,
"loss": 1.1184,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.4285406905621482,
"learning_rate": 6.648237195676927e-06,
"loss": 1.1246,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.3695600566670292,
"learning_rate": 6.606184197876395e-06,
"loss": 1.131,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.3590614001171659,
"learning_rate": 6.564238296670109e-06,
"loss": 1.1279,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.33572214926953525,
"learning_rate": 6.52239982745638e-06,
"loss": 1.1349,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.5314648559912775,
"learning_rate": 6.480669124774484e-06,
"loss": 1.113,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.3909757379935955,
"learning_rate": 6.439046522302013e-06,
"loss": 1.1437,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.3808759104835845,
"learning_rate": 6.397532352852191e-06,
"loss": 1.1042,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.4247598983732402,
"learning_rate": 6.356126948371202e-06,
"loss": 1.1354,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.4043133112754407,
"learning_rate": 6.314830639935559e-06,
"loss": 1.1304,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.3292134200478527,
"learning_rate": 6.273643757749448e-06,
"loss": 1.1577,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.440669860007943,
"learning_rate": 6.2325666311420856e-06,
"loss": 1.1637,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.40943717401687024,
"learning_rate": 6.1915995885650675e-06,
"loss": 1.1418,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.34762195003379087,
"learning_rate": 6.150742957589784e-06,
"loss": 1.1216,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.42913451053752155,
"learning_rate": 6.109997064904774e-06,
"loss": 1.1354,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.37230710045157994,
"learning_rate": 6.069362236313092e-06,
"loss": 1.1474,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.3758001362822172,
"learning_rate": 6.0288387967297565e-06,
"loss": 1.1632,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.3556590398546523,
"learning_rate": 5.988427070179108e-06,
"loss": 1.1268,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.43646047062373616,
"learning_rate": 5.948127379792234e-06,
"loss": 1.1578,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.37230844187779877,
"learning_rate": 5.90794004780439e-06,
"loss": 1.1625,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.335417956474066,
"learning_rate": 5.867865395552405e-06,
"loss": 1.1374,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.4158144314504683,
"learning_rate": 5.8279037434721364e-06,
"loss": 1.1481,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.3511649604328322,
"learning_rate": 5.788055411095892e-06,
"loss": 1.1462,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.4258871741798407,
"learning_rate": 5.7483207170498625e-06,
"loss": 1.1437,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.450486769779811,
"learning_rate": 5.708699979051613e-06,
"loss": 1.1455,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.3937522233416574,
"learning_rate": 5.66919351390751e-06,
"loss": 1.1223,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.3894498770511455,
"learning_rate": 5.629801637510188e-06,
"loss": 1.184,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.416058462952189,
"learning_rate": 5.590524664836043e-06,
"loss": 1.1416,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.4090490820017678,
"learning_rate": 5.551362909942708e-06,
"loss": 1.1068,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.4271684163771362,
"learning_rate": 5.51231668596653e-06,
"loss": 1.1137,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.43949172508654744,
"learning_rate": 5.47338630512007e-06,
"loss": 1.1374,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.3799047612464521,
"learning_rate": 5.434572078689617e-06,
"loss": 1.1061,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.41066964176100496,
"learning_rate": 5.395874317032694e-06,
"loss": 1.1281,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.3949704539901898,
"learning_rate": 5.357293329575573e-06,
"loss": 1.175,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.3580607509775383,
"learning_rate": 5.318829424810797e-06,
"loss": 1.1219,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.38234508809337464,
"learning_rate": 5.280482910294728e-06,
"loss": 1.1216,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.35406011311774904,
"learning_rate": 5.242254092645078e-06,
"loss": 1.138,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.45606754032725405,
"learning_rate": 5.2041432775384445e-06,
"loss": 1.1161,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.35914857398759203,
"learning_rate": 5.166150769707892e-06,
"loss": 1.1414,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.3281049558133169,
"learning_rate": 5.128276872940503e-06,
"loss": 1.1094,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.3946567466245035,
"learning_rate": 5.090521890074953e-06,
"loss": 1.1537,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.34343443554432856,
"learning_rate": 5.0528861229990675e-06,
"loss": 1.1596,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.40241539783380104,
"learning_rate": 5.015369872647442e-06,
"loss": 1.1442,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.3495269608296758,
"learning_rate": 4.9779734389990195e-06,
"loss": 1.1514,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.3810131654854716,
"learning_rate": 4.940697121074675e-06,
"loss": 1.1041,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.4036471153900224,
"learning_rate": 4.903541216934864e-06,
"loss": 1.1112,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.3623023974124653,
"learning_rate": 4.8665060236772e-06,
"loss": 1.1336,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.3473898046952608,
"learning_rate": 4.829591837434102e-06,
"loss": 1.12,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.3563222103888518,
"learning_rate": 4.792798953370424e-06,
"loss": 1.1135,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.37840134664340563,
"learning_rate": 4.756127665681083e-06,
"loss": 1.1624,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.40235755435764214,
"learning_rate": 4.719578267588725e-06,
"loss": 1.1383,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.43867323908210093,
"learning_rate": 4.6831510513413525e-06,
"loss": 1.1316,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.41221345670310156,
"learning_rate": 4.6468463082100176e-06,
"loss": 1.1013,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.41135692556074244,
"learning_rate": 4.610664328486485e-06,
"loss": 1.0654,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.4177763398734186,
"learning_rate": 4.5746054014809e-06,
"loss": 1.0741,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.42818607503616535,
"learning_rate": 4.5386698155194766e-06,
"loss": 1.0889,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.4445944164400039,
"learning_rate": 4.502857857942204e-06,
"loss": 1.1031,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.393278385595518,
"learning_rate": 4.467169815100549e-06,
"loss": 1.0843,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.43139577209748414,
"learning_rate": 4.431605972355137e-06,
"loss": 1.119,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.4938775822568143,
"learning_rate": 4.396166614073516e-06,
"loss": 1.1045,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.46083619181442753,
"learning_rate": 4.360852023627846e-06,
"loss": 1.0994,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.4286746114477124,
"learning_rate": 4.325662483392656e-06,
"loss": 1.0889,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.41009149518177995,
"learning_rate": 4.29059827474257e-06,
"loss": 1.0762,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.4226801377900272,
"learning_rate": 4.255659678050063e-06,
"loss": 1.099,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.4340266507703813,
"learning_rate": 4.22084697268323e-06,
"loss": 1.0962,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.36595907665192523,
"learning_rate": 4.186160437003535e-06,
"loss": 1.071,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.5109564568484206,
"learning_rate": 4.151600348363582e-06,
"loss": 1.0679,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.40485447937518654,
"learning_rate": 4.11716698310493e-06,
"loss": 1.1203,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.35905830592326227,
"learning_rate": 4.082860616555848e-06,
"loss": 1.0888,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.38499272923418265,
"learning_rate": 4.048681523029125e-06,
"loss": 1.0713,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.3634609476993962,
"learning_rate": 4.014629975819882e-06,
"loss": 1.0831,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.33818309016810527,
"learning_rate": 3.9807062472033855e-06,
"loss": 1.0739,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.3326646392205685,
"learning_rate": 3.94691060843287e-06,
"loss": 1.0565,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.35197128745538403,
"learning_rate": 3.913243329737353e-06,
"loss": 1.0998,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.31985429041703967,
"learning_rate": 3.879704680319503e-06,
"loss": 1.0764,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.32386941044315054,
"learning_rate": 3.846294928353471e-06,
"loss": 1.088,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.3160917650950608,
"learning_rate": 3.8130143409827345e-06,
"loss": 1.0949,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.3132032512342567,
"learning_rate": 3.7798631843179887e-06,
"loss": 1.0956,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.310211840212496,
"learning_rate": 3.7468417234349973e-06,
"loss": 1.0948,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.34307968359754526,
"learning_rate": 3.7139502223724823e-06,
"loss": 1.1158,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.3314353765015988,
"learning_rate": 3.6811889441300095e-06,
"loss": 1.1183,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.32576005046149137,
"learning_rate": 3.648558150665886e-06,
"loss": 1.0984,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.30693049985726834,
"learning_rate": 3.616058102895075e-06,
"loss": 1.0811,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.2921124796550647,
"learning_rate": 3.583689060687079e-06,
"loss": 1.0893,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.31073784891520123,
"learning_rate": 3.5514512828639023e-06,
"loss": 1.1002,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.33827993389256517,
"learning_rate": 3.51934502719796e-06,
"loss": 1.0576,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.3048030509387693,
"learning_rate": 3.4873705504100185e-06,
"loss": 1.0885,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.3037496315703854,
"learning_rate": 3.4555281081671345e-06,
"loss": 1.0821,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.3144079347957466,
"learning_rate": 3.4238179550806326e-06,
"loss": 1.0973,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.29536690962419965,
"learning_rate": 3.3922403447040497e-06,
"loss": 1.0742,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.3252116904579102,
"learning_rate": 3.3607955295311247e-06,
"loss": 1.0567,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.3095876744556457,
"learning_rate": 3.32948376099375e-06,
"loss": 1.0875,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.32441089529698003,
"learning_rate": 3.2983052894599976e-06,
"loss": 1.1093,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.2809625532066985,
"learning_rate": 3.2672603642320945e-06,
"loss": 1.0593,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.2904284249630206,
"learning_rate": 3.236349233544438e-06,
"loss": 1.0879,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.27919478311157875,
"learning_rate": 3.205572144561604e-06,
"loss": 1.0802,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.2787729742288885,
"learning_rate": 3.174929343376374e-06,
"loss": 1.133,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.30605357448375703,
"learning_rate": 3.1444210750077777e-06,
"loss": 1.0754,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.29176003116204585,
"learning_rate": 3.1140475833991046e-06,
"loss": 1.0593,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.2782083698046387,
"learning_rate": 3.0838091114159895e-06,
"loss": 1.0658,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.3066289023656689,
"learning_rate": 3.0537059008444524e-06,
"loss": 1.1096,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.3327573049944418,
"learning_rate": 3.0237381923889563e-06,
"loss": 1.0828,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2910832478398128,
"learning_rate": 2.9939062256704997e-06,
"loss": 1.0897,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.3149103694003108,
"learning_rate": 2.964210239224696e-06,
"loss": 1.0957,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.28758062785363814,
"learning_rate": 2.934650470499869e-06,
"loss": 1.1005,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.28936782388380544,
"learning_rate": 2.905227155855128e-06,
"loss": 1.0938,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.3288800888607633,
"learning_rate": 2.875940530558519e-06,
"loss": 1.0997,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.27820742810063026,
"learning_rate": 2.8467908287851154e-06,
"loss": 1.1131,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.2909602198534316,
"learning_rate": 2.8177782836151603e-06,
"loss": 1.0962,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.2932084290065858,
"learning_rate": 2.7889031270321874e-06,
"loss": 1.1036,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.314139553543045,
"learning_rate": 2.760165589921184e-06,
"loss": 1.0936,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.28050460147094197,
"learning_rate": 2.7315659020667305e-06,
"loss": 1.1011,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.2778883290392359,
"learning_rate": 2.703104292151164e-06,
"loss": 1.0631,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.2803666317850409,
"learning_rate": 2.6747809877527607e-06,
"loss": 1.1153,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.6045072142267184,
"learning_rate": 2.6465962153439082e-06,
"loss": 1.133,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.29029284094008767,
"learning_rate": 2.618550200289303e-06,
"loss": 1.0974,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.27955127114418393,
"learning_rate": 2.59064316684412e-06,
"loss": 1.102,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.2749826984166119,
"learning_rate": 2.562875338152264e-06,
"loss": 1.1402,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.2764728836882542,
"learning_rate": 2.5352469362445552e-06,
"loss": 1.0788,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.29987066332533785,
"learning_rate": 2.5077581820369523e-06,
"loss": 1.1005,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.28687503267761305,
"learning_rate": 2.480409295328803e-06,
"loss": 1.0979,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.33920852074858104,
"learning_rate": 2.45320049480108e-06,
"loss": 1.0926,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.29308394495841855,
"learning_rate": 2.4261319980146293e-06,
"loss": 1.0687,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.3095047801453578,
"learning_rate": 2.399204021408421e-06,
"loss": 1.1208,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.2954265913940137,
"learning_rate": 2.3724167802978395e-06,
"loss": 1.0787,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.3122300784266337,
"learning_rate": 2.3457704888729625e-06,
"loss": 1.0683,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.31393501456587775,
"learning_rate": 2.3192653601968162e-06,
"loss": 1.0909,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.30959729623239046,
"learning_rate": 2.2929016062037057e-06,
"loss": 1.1128,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.282171438701068,
"learning_rate": 2.2666794376975055e-06,
"loss": 1.1238,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.278072766818358,
"learning_rate": 2.24059906434998e-06,
"loss": 1.1002,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.3194726203005625,
"learning_rate": 2.214660694699091e-06,
"loss": 1.0957,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.27974086001281595,
"learning_rate": 2.188864536147355e-06,
"loss": 1.0855,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.6715783123764163,
"learning_rate": 2.1632107949601757e-06,
"loss": 1.1155,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.28813260915407823,
"learning_rate": 2.137699676264171e-06,
"loss": 1.1057,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.2999412725972485,
"learning_rate": 2.1123313840455738e-06,
"loss": 1.0772,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.2649453403252441,
"learning_rate": 2.08710612114857e-06,
"loss": 1.0883,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.29393279559947144,
"learning_rate": 2.0620240892736953e-06,
"loss": 1.0666,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.31244115770852293,
"learning_rate": 2.0370854889762e-06,
"loss": 1.0792,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.2980230877095427,
"learning_rate": 2.0122905196644726e-06,
"loss": 1.0615,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2871147116578234,
"learning_rate": 1.987639379598427e-06,
"loss": 1.1195,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.2678315470865863,
"learning_rate": 1.963132265887919e-06,
"loss": 1.0732,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.271109581598972,
"learning_rate": 1.9387693744911784e-06,
"loss": 1.1065,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.2895946756224003,
"learning_rate": 1.9145509002132368e-06,
"loss": 1.0464,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.26602041758262057,
"learning_rate": 1.8904770367043723e-06,
"loss": 1.0636,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.28638747590642505,
"learning_rate": 1.8665479764585504e-06,
"loss": 1.0739,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.285188413089278,
"learning_rate": 1.8427639108118977e-06,
"loss": 1.0912,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.26935724284431184,
"learning_rate": 1.8191250299411755e-06,
"loss": 1.1238,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.26087015521191953,
"learning_rate": 1.7956315228622446e-06,
"loss": 1.0838,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.2831475485308561,
"learning_rate": 1.7722835774285553e-06,
"loss": 1.0865,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.28643348454829604,
"learning_rate": 1.7490813803296624e-06,
"loss": 1.0997,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.26466131388186337,
"learning_rate": 1.726025117089718e-06,
"loss": 1.0956,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.27820695341372054,
"learning_rate": 1.70311497206598e-06,
"loss": 1.0901,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.28287079094892054,
"learning_rate": 1.6803511284473638e-06,
"loss": 1.0822,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.2609457355927526,
"learning_rate": 1.6577337682529538e-06,
"loss": 1.0829,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.2985362206216598,
"learning_rate": 1.6352630723305617e-06,
"loss": 1.0828,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.293704750196247,
"learning_rate": 1.612939220355272e-06,
"loss": 1.107,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.2652651304456669,
"learning_rate": 1.5907623908280112e-06,
"loss": 1.0847,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.2689014212068562,
"learning_rate": 1.568732761074121e-06,
"loss": 1.085,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.26332043482444273,
"learning_rate": 1.5468505072419237e-06,
"loss": 1.1108,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.27111317160239984,
"learning_rate": 1.5251158043013448e-06,
"loss": 1.101,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.26104503708871735,
"learning_rate": 1.5035288260424885e-06,
"loss": 1.1153,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 1.970170045459338,
"learning_rate": 1.4820897450742623e-06,
"loss": 1.1633,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.26482429673488317,
"learning_rate": 1.4607987328229767e-06,
"loss": 1.0872,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.31589854150500685,
"learning_rate": 1.4396559595310056e-06,
"loss": 1.0667,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.30372110991674417,
"learning_rate": 1.4186615942553973e-06,
"loss": 1.0682,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.27805189203268643,
"learning_rate": 1.3978158048665424e-06,
"loss": 1.0694,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.2743992741640192,
"learning_rate": 1.3771187580468115e-06,
"loss": 1.1168,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.25734466161179703,
"learning_rate": 1.3565706192892392e-06,
"loss": 1.0664,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.25542331431687965,
"learning_rate": 1.3361715528961993e-06,
"loss": 1.0882,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.2912417830383445,
"learning_rate": 1.3159217219780817e-06,
"loss": 1.1102,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.2715501347553617,
"learning_rate": 1.2958212884519949e-06,
"loss": 1.068,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.28548013911902126,
"learning_rate": 1.2758704130404675e-06,
"loss": 1.0829,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.274400982446564,
"learning_rate": 1.2560692552701738e-06,
"loss": 1.1083,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.26932565665725294,
"learning_rate": 1.2364179734706382e-06,
"loss": 1.0846,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.26445888809274753,
"learning_rate": 1.2169167247729852e-06,
"loss": 1.1074,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.26835418155873264,
"learning_rate": 1.1975656651086864e-06,
"loss": 1.0662,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.2521687724041235,
"learning_rate": 1.1783649492082926e-06,
"loss": 1.0885,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.28409420994598633,
"learning_rate": 1.1593147306002183e-06,
"loss": 1.0577,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.26554970994881927,
"learning_rate": 1.140415161609505e-06,
"loss": 1.0984,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.2725099474248271,
"learning_rate": 1.121666393356604e-06,
"loss": 1.0555,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.276129127383448,
"learning_rate": 1.1030685757561632e-06,
"loss": 1.0807,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.2650653408429626,
"learning_rate": 1.0846218575158373e-06,
"loss": 1.0844,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2691639832358718,
"learning_rate": 1.066326386135097e-06,
"loss": 1.0636,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.29181377240784645,
"learning_rate": 1.0481823079040398e-06,
"loss": 1.078,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.2686012972826775,
"learning_rate": 1.0301897679022344e-06,
"loss": 1.0751,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.2545722796541032,
"learning_rate": 1.0123489099975491e-06,
"loss": 1.0716,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.2739502354172817,
"learning_rate": 9.9465987684501e-07,
"loss": 1.0924,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.24278010922267443,
"learning_rate": 9.771228098856534e-07,
"loss": 1.1212,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.29305257375005833,
"learning_rate": 9.597378493454013e-07,
"loss": 1.0739,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.25062090831303024,
"learning_rate": 9.425051342339353e-07,
"loss": 1.0946,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.26099436064153536,
"learning_rate": 9.254248023435864e-07,
"loss": 1.0833,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.2726994879319825,
"learning_rate": 9.084969902482288e-07,
"loss": 1.0582,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.2880092268886225,
"learning_rate": 8.917218333021993e-07,
"loss": 1.0811,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.2436356457150742,
"learning_rate": 8.750994656392109e-07,
"loss": 1.0719,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.2682990052273749,
"learning_rate": 8.58630020171265e-07,
"loss": 1.0713,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.28641428636619387,
"learning_rate": 8.423136285876099e-07,
"loss": 1.0757,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.2654477742337598,
"learning_rate": 8.26150421353682e-07,
"loss": 1.0672,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.2760190640809688,
"learning_rate": 8.101405277100549e-07,
"loss": 1.0824,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.2847509555727561,
"learning_rate": 7.942840756714077e-07,
"loss": 1.1132,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.2783488166268553,
"learning_rate": 7.785811920255093e-07,
"loss": 1.1004,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.2634648116394111,
"learning_rate": 7.630320023322024e-07,
"loss": 1.0795,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.25996189167566663,
"learning_rate": 7.476366309223903e-07,
"loss": 1.0868,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.2659552219322079,
"learning_rate": 7.323952008970537e-07,
"loss": 1.1103,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.2686611165864897,
"learning_rate": 7.173078341262618e-07,
"loss": 1.0345,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.2519845286528926,
"learning_rate": 7.023746512481989e-07,
"loss": 1.1514,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.25200318293591245,
"learning_rate": 6.875957716681902e-07,
"loss": 1.1068,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.25648179553409883,
"learning_rate": 6.729713135577687e-07,
"loss": 1.0883,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.24396679659504275,
"learning_rate": 6.585013938537143e-07,
"loss": 1.1037,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.2875186312302226,
"learning_rate": 6.441861282571115e-07,
"loss": 1.0938,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.27144003379857595,
"learning_rate": 6.300256312324471e-07,
"loss": 1.1057,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.2925938342533537,
"learning_rate": 6.160200160066776e-07,
"loss": 1.0779,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2716640436629508,
"learning_rate": 6.021693945683327e-07,
"loss": 1.0881,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.2527152135304773,
"learning_rate": 5.884738776666044e-07,
"loss": 1.0777,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.2501286166136105,
"learning_rate": 5.749335748104878e-07,
"loss": 1.0862,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.2526994094576257,
"learning_rate": 5.615485942678778e-07,
"loss": 1.0679,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.2506423609439915,
"learning_rate": 5.483190430647289e-07,
"loss": 1.0665,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.25583512534636643,
"learning_rate": 5.352450269841747e-07,
"loss": 1.0834,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.25244308814978395,
"learning_rate": 5.22326650565701e-07,
"loss": 1.0992,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.2575431041069518,
"learning_rate": 5.095640171043048e-07,
"loss": 1.0855,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.2842418855766519,
"learning_rate": 4.969572286496594e-07,
"loss": 1.082,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.2561439352284069,
"learning_rate": 4.845063860053079e-07,
"loss": 1.0813,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.2713290984804756,
"learning_rate": 4.7221158872785997e-07,
"loss": 1.0889,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.2551197645684873,
"learning_rate": 4.600729351261857e-07,
"loss": 1.0972,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.25358193207951363,
"learning_rate": 4.4809052226063355e-07,
"loss": 1.0971,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.24640934532271624,
"learning_rate": 4.3626444594226046e-07,
"loss": 1.0775,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.2792511154661679,
"learning_rate": 4.245948007320588e-07,
"loss": 1.1196,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.27317291398701143,
"learning_rate": 4.130816799401971e-07,
"loss": 1.0689,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.2411271100622031,
"learning_rate": 4.017251756252827e-07,
"loss": 1.0873,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.26471928120206323,
"learning_rate": 3.9052537859362473e-07,
"loss": 1.0781,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.24281298271253426,
"learning_rate": 3.7948237839849687e-07,
"loss": 1.0997,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.2490547579014643,
"learning_rate": 3.6859626333943797e-07,
"loss": 1.1004,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.2609695293393079,
"learning_rate": 3.578671204615347e-07,
"loss": 1.1094,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.24921739988752298,
"learning_rate": 3.472950355547267e-07,
"loss": 1.1178,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.26008630092368906,
"learning_rate": 3.3688009315312244e-07,
"loss": 1.0843,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.26033849857611513,
"learning_rate": 3.266223765343224e-07,
"loss": 1.1033,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.25760438546595604,
"learning_rate": 3.16521967718757e-07,
"loss": 1.0818,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.2847684879281591,
"learning_rate": 3.0657894746902726e-07,
"loss": 1.0673,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.2446431305553472,
"learning_rate": 2.9679339528925654e-07,
"loss": 1.1009,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.26529285734602254,
"learning_rate": 2.871653894244597e-07,
"loss": 1.0991,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.25973710997709576,
"learning_rate": 2.7769500685991266e-07,
"loss": 1.0987,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.2672422744864518,
"learning_rate": 2.683823233205485e-07,
"loss": 1.0996,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.25772387201817637,
"learning_rate": 2.592274132703287e-07,
"loss": 1.0716,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.2595390932174708,
"learning_rate": 2.502303499116754e-07,
"loss": 1.0921,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.24648069220514837,
"learning_rate": 2.413912051848666e-07,
"loss": 1.1069,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.2454931170863349,
"learning_rate": 2.3271004976746613e-07,
"loss": 1.1014,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.2537261488889785,
"learning_rate": 2.2418695307376393e-07,
"loss": 1.0784,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.24819864951486156,
"learning_rate": 2.1582198325420968e-07,
"loss": 1.0818,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2508349084506116,
"learning_rate": 2.0761520719488447e-07,
"loss": 1.0871,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.2567023383922933,
"learning_rate": 1.9956669051694798e-07,
"loss": 1.1265,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.2373768613205901,
"learning_rate": 1.916764975761254e-07,
"loss": 1.0946,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.25176946594443333,
"learning_rate": 1.839446914621923e-07,
"loss": 1.1137,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.2502469959691376,
"learning_rate": 1.7637133399845962e-07,
"loss": 1.0506,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.24652130342727785,
"learning_rate": 1.6895648574129397e-07,
"loss": 1.1066,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2754573533455405,
"learning_rate": 1.6170020597962465e-07,
"loss": 1.0635,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.24943116939225307,
"learning_rate": 1.5460255273447078e-07,
"loss": 1.0993,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.2539663490359687,
"learning_rate": 1.4766358275847715e-07,
"loss": 1.0845,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.24102891117259034,
"learning_rate": 1.4088335153546573e-07,
"loss": 1.0981,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.2528840021757461,
"learning_rate": 1.3426191327998496e-07,
"loss": 1.093,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.23520505859776003,
"learning_rate": 1.277993209368744e-07,
"loss": 1.1007,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.23618233641263664,
"learning_rate": 1.214956261808542e-07,
"loss": 1.0857,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.2520608844173438,
"learning_rate": 1.1535087941609846e-07,
"loss": 1.0969,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.27537345496888754,
"learning_rate": 1.0936512977583802e-07,
"loss": 1.0793,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.2405088575556249,
"learning_rate": 1.0353842512196732e-07,
"loss": 1.0786,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.23904630201291482,
"learning_rate": 9.787081204466253e-08,
"loss": 1.0911,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.24973924324401148,
"learning_rate": 9.236233586200627e-08,
"loss": 1.1119,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.2487105791971777,
"learning_rate": 8.70130406196279e-08,
"loss": 1.1058,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.24231774379409152,
"learning_rate": 8.182296909035047e-08,
"loss": 1.0867,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.25524092804310977,
"learning_rate": 7.679216277384439e-08,
"loss": 1.0795,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.24874000131417404,
"learning_rate": 7.19206618963053e-08,
"loss": 1.0813,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.284204796776814,
"learning_rate": 6.720850541012347e-08,
"loss": 1.0994,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.24228370186596546,
"learning_rate": 6.265573099357714e-08,
"loss": 1.1006,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.24958104797583483,
"learning_rate": 5.826237505053067e-08,
"loss": 1.1007,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.2438390629578957,
"learning_rate": 5.4028472710137e-08,
"loss": 1.0815,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.2615433907393708,
"learning_rate": 4.9954057826571146e-08,
"loss": 1.0783,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.2395241133205361,
"learning_rate": 4.6039162978743776e-08,
"loss": 1.0655,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.2430944403718584,
"learning_rate": 4.228381947004812e-08,
"loss": 1.111,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.2600350332121252,
"learning_rate": 3.868805732811343e-08,
"loss": 1.1088,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.24472640301938442,
"learning_rate": 3.525190530455636e-08,
"loss": 1.0885,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.2550551535812951,
"learning_rate": 3.197539087475887e-08,
"loss": 1.1179,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.24753718267238733,
"learning_rate": 2.8858540237641786e-08,
"loss": 1.0887,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.2595375196272305,
"learning_rate": 2.5901378315458247e-08,
"loss": 1.0937,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.24695128066474364,
"learning_rate": 2.3103928753598347e-08,
"loss": 1.0566,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.2627019247426801,
"learning_rate": 2.0466213920393718e-08,
"loss": 1.0812,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.2774714690236478,
"learning_rate": 1.7988254906942115e-08,
"loss": 1.1064,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.23240577506332505,
"learning_rate": 1.5670071526936447e-08,
"loss": 1.1228,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.24398280119291169,
"learning_rate": 1.3511682316509344e-08,
"loss": 1.1077,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.22638229721933628,
"learning_rate": 1.1513104534086605e-08,
"loss": 1.0997,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.2513948861501898,
"learning_rate": 9.67435416023843e-09,
"loss": 1.09,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.24155478010069545,
"learning_rate": 7.995445897566178e-09,
"loss": 1.1102,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2610336308831626,
"learning_rate": 6.476393170578021e-09,
"loss": 1.0412,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.25136312336774186,
"learning_rate": 5.11720812557348e-09,
"loss": 1.0869,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2491853125147826,
"learning_rate": 3.917901630563492e-09,
"loss": 1.0901,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.24599031319201908,
"learning_rate": 2.878483275170485e-09,
"loss": 1.1192,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.24092850334891475,
"learning_rate": 1.998961370555108e-09,
"loss": 1.1256,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.25538150351818845,
"learning_rate": 1.2793429493518361e-09,
"loss": 1.0874,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.23206709995771904,
"learning_rate": 7.196337656112384e-10,
"loss": 1.0562,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.25659377367530206,
"learning_rate": 3.198382947511292e-10,
"loss": 1.102,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.25805521764288997,
"learning_rate": 7.995973352770137e-11,
"loss": 1.0834,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.24540319218156467,
"learning_rate": 0.0,
"loss": 1.0997,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 3.6878247667496387e+18,
"train_loss": 1.206290436756273,
"train_runtime": 30610.2836,
"train_samples_per_second": 5.159,
"train_steps_per_second": 0.04
}
],
"logging_steps": 1.0,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.6878247667496387e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}