meta_no_pipeline_30k_32B / trainer_state.json
ssu53's picture
Upload model
a5b8013 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": 3.8814706034180015,
"learning_rate": 3.2258064516129035e-07,
"loss": 0.9756,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 4.12180455785843,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0274,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 3.925111993613436,
"learning_rate": 9.67741935483871e-07,
"loss": 0.9901,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 3.837669157236785,
"learning_rate": 1.2903225806451614e-06,
"loss": 0.9835,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 3.498877313756905,
"learning_rate": 1.6129032258064516e-06,
"loss": 0.9639,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 3.6936969551795182,
"learning_rate": 1.935483870967742e-06,
"loss": 0.9499,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 2.9165120723040925,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.9064,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 2.948659961922266,
"learning_rate": 2.580645161290323e-06,
"loss": 0.9261,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 2.188594772539154,
"learning_rate": 2.903225806451613e-06,
"loss": 0.8817,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 1.9505117710666822,
"learning_rate": 3.225806451612903e-06,
"loss": 0.891,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 1.763453845063286,
"learning_rate": 3.548387096774194e-06,
"loss": 0.8536,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 1.4198526555754307,
"learning_rate": 3.870967741935484e-06,
"loss": 0.8251,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 1.3029206191417797,
"learning_rate": 4.193548387096774e-06,
"loss": 0.8179,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 1.2604830600274026,
"learning_rate": 4.516129032258065e-06,
"loss": 0.8124,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 1.2065789803410185,
"learning_rate": 4.838709677419355e-06,
"loss": 0.8413,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 1.1149823273189958,
"learning_rate": 5.161290322580646e-06,
"loss": 0.7683,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 1.2357683220854039,
"learning_rate": 5.483870967741935e-06,
"loss": 0.761,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 1.0805331086023167,
"learning_rate": 5.806451612903226e-06,
"loss": 0.7581,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 0.848284212502348,
"learning_rate": 6.129032258064517e-06,
"loss": 0.7291,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 0.87844887956233,
"learning_rate": 6.451612903225806e-06,
"loss": 0.711,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 0.7797900114402678,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7306,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 0.871258561883303,
"learning_rate": 7.096774193548388e-06,
"loss": 0.7179,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 0.8415008772017533,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.689,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 0.7395352287579315,
"learning_rate": 7.741935483870968e-06,
"loss": 0.671,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 0.7644039766565234,
"learning_rate": 8.064516129032258e-06,
"loss": 0.679,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.6264835822200987,
"learning_rate": 8.387096774193549e-06,
"loss": 0.6281,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 0.6447338698959464,
"learning_rate": 8.70967741935484e-06,
"loss": 0.6863,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 0.6110058604828769,
"learning_rate": 9.03225806451613e-06,
"loss": 0.6749,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 0.5042148617192856,
"learning_rate": 9.35483870967742e-06,
"loss": 0.6524,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 0.5337478280266854,
"learning_rate": 9.67741935483871e-06,
"loss": 0.646,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 0.4852151598738288,
"learning_rate": 1e-05,
"loss": 0.6246,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 0.46707516863027926,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.6218,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 0.528676943739116,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.6306,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 0.5184708117459765,
"learning_rate": 1.096774193548387e-05,
"loss": 0.6315,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 0.4884861076774194,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.6218,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 0.46386750313826103,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6107,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 0.45820765664012425,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.6124,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.4633422013856627,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6238,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.4668079189834338,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.6219,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 0.4377464237213352,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6143,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 0.47255476427258364,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6268,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 0.5134927137181531,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6156,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 0.48587418375096864,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.6048,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 0.4665085047331668,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6009,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 0.532397384810035,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6297,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 0.4319285837092898,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6031,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 0.4366334797957311,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6004,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 0.4591284856668385,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.5899,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 0.5106827671308912,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6249,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 0.4247813497620674,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.5763,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 0.47199985877928824,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6159,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.4980423978752336,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6114,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 0.4152320063999814,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.5798,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 0.44946603226090287,
"learning_rate": 1.741935483870968e-05,
"loss": 0.5922,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 0.4104953143372891,
"learning_rate": 1.774193548387097e-05,
"loss": 0.5464,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 0.48469160711407705,
"learning_rate": 1.806451612903226e-05,
"loss": 0.622,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.4057914124534525,
"learning_rate": 1.838709677419355e-05,
"loss": 0.5875,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 0.5011222497467952,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 0.46492739352660434,
"learning_rate": 1.903225806451613e-05,
"loss": 0.5475,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 0.41846703371710475,
"learning_rate": 1.935483870967742e-05,
"loss": 0.5611,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 0.4224130157133128,
"learning_rate": 1.967741935483871e-05,
"loss": 0.5767,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 0.4759001978065436,
"learning_rate": 2e-05,
"loss": 0.6219,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 0.4606436001797547,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.5851,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 0.5290061343625666,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.5676,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.47995054313291297,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.5826,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 0.5060741296811404,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.5785,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 0.5904516895532496,
"learning_rate": 2.161290322580645e-05,
"loss": 0.5507,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 0.48707911708288054,
"learning_rate": 2.193548387096774e-05,
"loss": 0.5648,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 0.5888296804622808,
"learning_rate": 2.225806451612903e-05,
"loss": 0.5781,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 0.5825082869106455,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.5878,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 0.5407839063711196,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6011,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 0.5995396689101684,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.5779,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 0.6430488131662697,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.5916,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 0.4523813185279133,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.573,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 0.6588093557655494,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.6105,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.48679158537170164,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.5696,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 0.45394283848534545,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.5865,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.49656200045857685,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.581,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 0.4866639182038764,
"learning_rate": 2.548387096774194e-05,
"loss": 0.5777,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 0.5703677535089352,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6066,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 0.4245697069442014,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.5779,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 0.4844836995259489,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.6001,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.535822507716651,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.5679,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 0.5088754857086452,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.5764,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 0.5023168541993437,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.5529,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 0.49498233456481894,
"learning_rate": 2.774193548387097e-05,
"loss": 0.5649,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 0.46726429134060576,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.5905,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 0.44354368067556005,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.5332,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 0.45476561151759554,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.5751,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 0.49340731051673276,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.5597,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.4576246827692569,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.5476,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 0.47751427697663007,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.5293,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 0.4357269151533143,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.5873,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 0.5351175280375254,
"learning_rate": 3.032258064516129e-05,
"loss": 0.577,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 0.4284511330212561,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.5565,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 0.53885828372451,
"learning_rate": 3.096774193548387e-05,
"loss": 0.5898,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.4580583915653527,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.5622,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 0.48691117098529124,
"learning_rate": 3.161290322580645e-05,
"loss": 0.5672,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 0.44704391131124904,
"learning_rate": 3.193548387096774e-05,
"loss": 0.5632,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 0.4233441026169608,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.5743,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 0.4275645670051757,
"learning_rate": 3.258064516129033e-05,
"loss": 0.5389,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 0.40707541777438944,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.5569,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 0.46823925541066874,
"learning_rate": 3.322580645161291e-05,
"loss": 0.5888,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.4347214311197812,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.5505,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 0.4744977209025911,
"learning_rate": 3.387096774193549e-05,
"loss": 0.5724,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 0.45942953258676617,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.5789,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 0.4769876507430913,
"learning_rate": 3.451612903225807e-05,
"loss": 0.5602,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.47237074986692135,
"learning_rate": 3.483870967741936e-05,
"loss": 0.565,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 0.4662982161357429,
"learning_rate": 3.516129032258065e-05,
"loss": 0.5704,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 0.45918399726543613,
"learning_rate": 3.548387096774194e-05,
"loss": 0.5559,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 0.5362403487911285,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.5969,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 0.5210411060449173,
"learning_rate": 3.612903225806452e-05,
"loss": 0.5558,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.5070682530710503,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.5599,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.5467889745074546,
"learning_rate": 3.67741935483871e-05,
"loss": 0.5728,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 0.4466853280106437,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.5433,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.5788006683388244,
"learning_rate": 3.741935483870968e-05,
"loss": 0.5516,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.48372515766964225,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.5831,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.5253098827826163,
"learning_rate": 3.806451612903226e-05,
"loss": 0.5575,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.45501664454423063,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.5478,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.4683444843899017,
"learning_rate": 3.870967741935484e-05,
"loss": 0.5663,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 0.51634413602177,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.5816,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 0.43444042193949195,
"learning_rate": 3.935483870967742e-05,
"loss": 0.5728,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 0.518342982503112,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.5616,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 0.564876180560867,
"learning_rate": 4e-05,
"loss": 0.5955,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 0.4763981793213528,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.5777,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 0.6081843564830384,
"learning_rate": 3.999968016170525e-05,
"loss": 0.5561,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 0.4618967250925891,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.5607,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.5266972433649928,
"learning_rate": 3.999872065705065e-05,
"loss": 0.562,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.4927412280531229,
"learning_rate": 3.999800103862945e-05,
"loss": 0.594,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.4872081920850752,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.5593,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 0.4411992392583207,
"learning_rate": 3.999608209836944e-05,
"loss": 0.5577,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 0.47343516267005176,
"learning_rate": 3.999488279187443e-05,
"loss": 0.5365,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.4257090454806133,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.5424,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 0.49692604529840384,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.5626,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 0.47224940730415305,
"learning_rate": 3.999032564583976e-05,
"loss": 0.5431,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 0.42529518856017906,
"learning_rate": 3.998848689546591e-05,
"loss": 0.5653,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 0.5016133178374419,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.5691,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 0.44317860683451515,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.558,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 0.4317487987886679,
"learning_rate": 3.998201174509306e-05,
"loss": 0.5556,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.5479284140798366,
"learning_rate": 3.997953378607961e-05,
"loss": 0.5613,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.4826854926682329,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.554,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 0.4958279711947429,
"learning_rate": 3.997409862168454e-05,
"loss": 0.5482,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.4573482015449114,
"learning_rate": 3.997114145976237e-05,
"loss": 0.5736,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 0.5006358879890985,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.5605,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.4392091801903901,
"learning_rate": 3.996474809469545e-05,
"loss": 0.5744,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.4627355320183437,
"learning_rate": 3.996131194267189e-05,
"loss": 0.577,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 0.42512252301355513,
"learning_rate": 3.995771618052995e-05,
"loss": 0.5432,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 0.41449211504725686,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.555,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 0.45774949751728017,
"learning_rate": 3.995004594217343e-05,
"loss": 0.5726,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 0.4925693142694618,
"learning_rate": 3.994597152728987e-05,
"loss": 0.5412,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 0.414001533554994,
"learning_rate": 3.994173762494947e-05,
"loss": 0.5467,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.5211470137073257,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.5442,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 0.4508866940514989,
"learning_rate": 3.993279149458988e-05,
"loss": 0.5427,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.5201794509433897,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.5479,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.4275576262656416,
"learning_rate": 3.992320783722616e-05,
"loss": 0.5356,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.47719877290396373,
"learning_rate": 3.991817703090965e-05,
"loss": 0.563,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.4223294307139996,
"learning_rate": 3.991298695938038e-05,
"loss": 0.5728,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.4091169912307533,
"learning_rate": 3.9907637664138e-05,
"loss": 0.5433,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.4244972682717285,
"learning_rate": 3.990212918795534e-05,
"loss": 0.5599,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.40831152155372563,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.5667,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.42573618315654754,
"learning_rate": 3.989063487022416e-05,
"loss": 0.5886,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 0.3975711304194636,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.5724,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 0.44790976661666787,
"learning_rate": 3.987850437381915e-05,
"loss": 0.5617,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 0.4218844963056509,
"learning_rate": 3.987220067906313e-05,
"loss": 0.5672,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 0.44202841263210757,
"learning_rate": 3.986573808672002e-05,
"loss": 0.5581,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.4856911335142941,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.5555,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 0.3838221260751295,
"learning_rate": 3.985233641724152e-05,
"loss": 0.5221,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 0.4530907733171599,
"learning_rate": 3.984539744726554e-05,
"loss": 0.5607,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.5074990009873959,
"learning_rate": 3.983829979402038e-05,
"loss": 0.5597,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 0.45781053698490876,
"learning_rate": 3.983104351425871e-05,
"loss": 0.5581,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.47135851726588246,
"learning_rate": 3.982362866600154e-05,
"loss": 0.5469,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.47640412764708323,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.5657,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.40106872425971785,
"learning_rate": 3.980832350242388e-05,
"loss": 0.5588,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 0.5426375020518981,
"learning_rate": 3.980043330948306e-05,
"loss": 0.5575,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.4477669081927397,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.5696,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.4880821946516913,
"learning_rate": 3.978417801674579e-05,
"loss": 0.5524,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 0.5011496745885547,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.5613,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.4357091483345441,
"learning_rate": 3.976728995023254e-05,
"loss": 0.5259,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.39474781961780614,
"learning_rate": 3.975860879481514e-05,
"loss": 0.5495,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.4314756458266692,
"learning_rate": 3.974976965008833e-05,
"loss": 0.5423,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.416348739321978,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.5796,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.4177535694763515,
"learning_rate": 3.973161767667946e-05,
"loss": 0.5558,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.5101656630719539,
"learning_rate": 3.972230499314009e-05,
"loss": 0.5355,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.4279659687516323,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.5781,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 0.5034245034546916,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.5585,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 0.46419339682016125,
"learning_rate": 3.969342105253098e-05,
"loss": 0.576,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 0.4124795591002647,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.5115,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.4111993157179236,
"learning_rate": 3.967337762346568e-05,
"loss": 0.5471,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 0.5267006552370684,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.5599,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.41361429876147243,
"learning_rate": 3.965270496444528e-05,
"loss": 0.5657,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 0.48356475481861766,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.5558,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 0.4014868363833699,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.5704,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 0.4395769139847585,
"learning_rate": 3.962051762160151e-05,
"loss": 0.5606,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 0.3853928781175799,
"learning_rate": 3.960947462140638e-05,
"loss": 0.5444,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.4624432978668155,
"learning_rate": 3.959827482437472e-05,
"loss": 0.5716,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.35490886398614024,
"learning_rate": 3.958691832005981e-05,
"loss": 0.5448,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 0.4677524786039117,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.5647,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 0.42693683915897057,
"learning_rate": 3.956373555405774e-05,
"loss": 0.5631,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.490259811019483,
"learning_rate": 3.955190947773937e-05,
"loss": 0.532,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 0.4596909112753546,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.5726,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 0.5257652678919216,
"learning_rate": 3.952778841127214e-05,
"loss": 0.5388,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.4516322892291902,
"learning_rate": 3.95154936139947e-05,
"loss": 0.5524,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 0.5327996408337012,
"learning_rate": 3.950304277135034e-05,
"loss": 0.5556,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 0.39962699478635816,
"learning_rate": 3.94904359828957e-05,
"loss": 0.5392,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 0.4405464652925951,
"learning_rate": 3.94776733494343e-05,
"loss": 0.5465,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 0.4085398368644957,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.5324,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 0.4077033850550244,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.5437,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.3848471383356622,
"learning_rate": 3.943845140573213e-05,
"loss": 0.5622,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.43647771145220304,
"learning_rate": 3.942506642518952e-05,
"loss": 0.5671,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.37371978511462395,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.5741,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 0.4732652084503298,
"learning_rate": 3.939783060543167e-05,
"loss": 0.5526,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 0.4560339553765893,
"learning_rate": 3.938397998399333e-05,
"loss": 0.5421,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.37424329390124816,
"learning_rate": 3.936997436876756e-05,
"loss": 0.5562,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 0.5128987818306787,
"learning_rate": 3.935581387174289e-05,
"loss": 0.5539,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 0.37322876494714985,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.5566,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.4060825347561096,
"learning_rate": 3.932702868644224e-05,
"loss": 0.547,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.45453599913678394,
"learning_rate": 3.931240422833181e-05,
"loss": 0.5716,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.42011921084987464,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.543,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.4609758263132003,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.5574,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.4488902290731712,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.5293,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.41773317012349287,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.5447,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 0.4219435680236813,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.5146,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.4635366765148105,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.5371,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 0.4564213341282917,
"learning_rate": 3.92057159243286e-05,
"loss": 0.5454,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 0.43589959797644207,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.5439,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.436918635554428,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.5323,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.4099414390035029,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.5547,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.41154529391654304,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5466,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 0.38826364902789445,
"learning_rate": 3.912490053436079e-05,
"loss": 0.5383,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 0.37658125044221796,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.5413,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.40643785422705536,
"learning_rate": 3.909150300975178e-05,
"loss": 0.5388,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.4204837444255185,
"learning_rate": 3.907457519765642e-05,
"loss": 0.5476,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.39718924610502254,
"learning_rate": 3.905749486576607e-05,
"loss": 0.5186,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.40936183034258306,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5704,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.34169698436093515,
"learning_rate": 3.902287719011435e-05,
"loss": 0.5526,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 0.3966335983914938,
"learning_rate": 3.900534012315499e-05,
"loss": 0.5494,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 0.3831199272198189,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.5063,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 0.4606290641747254,
"learning_rate": 3.896981023209777e-05,
"loss": 0.5509,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 0.4001471561512102,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.5506,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 0.41212551219748617,
"learning_rate": 3.893367361386491e-05,
"loss": 0.5434,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 0.4020919583308468,
"learning_rate": 3.891537814248417e-05,
"loss": 0.4953,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.4110788725427374,
"learning_rate": 3.889693142424384e-05,
"loss": 0.5397,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 0.4220683154051289,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.5587,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 0.4252913983632708,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5635,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.3546547776651679,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.5445,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.481190375423052,
"learning_rate": 3.882163505079171e-05,
"loss": 0.5426,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.4171393318215334,
"learning_rate": 3.880243433489132e-05,
"loss": 0.505,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.5507778760862126,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4063,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.5438962338085048,
"learning_rate": 3.876358202652936e-05,
"loss": 0.4201,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.6270783135644249,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.412,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.5615796761991196,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.3835,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.5273980761840045,
"learning_rate": 3.870417871154801e-05,
"loss": 0.3896,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.5234792532983567,
"learning_rate": 3.868407827802192e-05,
"loss": 0.3839,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.5713073049730966,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.4172,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.5780962574527349,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4004,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.4466283575195174,
"learning_rate": 3.862288124195319e-05,
"loss": 0.3871,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.5202239453853866,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.4225,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 0.5257172397489868,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.3771,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 0.47278414237113736,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.3981,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.5456937374883605,
"learning_rate": 3.853920126717703e-05,
"loss": 0.3893,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 0.5146114478739794,
"learning_rate": 3.851791025492574e-05,
"loss": 0.3964,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.485188116767974,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.3852,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.45476169137925654,
"learning_rate": 3.847488419569866e-05,
"loss": 0.3818,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.4325485610930113,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.3915,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.4406582197186754,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.3969,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.4460944150768255,
"learning_rate": 3.840923760917199e-05,
"loss": 0.3752,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.4383310122137117,
"learning_rate": 3.838706077964473e-05,
"loss": 0.3941,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.46404344991378926,
"learning_rate": 3.836473692766944e-05,
"loss": 0.4005,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.3603868655323874,
"learning_rate": 3.834226623174705e-05,
"loss": 0.3564,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.40091293883804097,
"learning_rate": 3.831964887155264e-05,
"loss": 0.4108,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.4071721768270374,
"learning_rate": 3.829688502793403e-05,
"loss": 0.4011,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.40602630239237814,
"learning_rate": 3.827397488291029e-05,
"loss": 0.3942,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.3822089365092082,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.3786,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.45741454158485867,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4148,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.4185402630359577,
"learning_rate": 3.820436847713776e-05,
"loss": 0.3835,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 0.4501157368224525,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.3923,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.5632066271769794,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.414,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.42505381705996353,
"learning_rate": 3.813345202354145e-05,
"loss": 0.3863,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 0.48871854979463053,
"learning_rate": 3.810952296329563e-05,
"loss": 0.3935,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.49853100208683093,
"learning_rate": 3.808544909978676e-05,
"loss": 0.3879,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.38565147585968634,
"learning_rate": 3.806123062550882e-05,
"loss": 0.4263,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.4611043047708484,
"learning_rate": 3.803686773411208e-05,
"loss": 0.3816,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 0.39785865708338475,
"learning_rate": 3.801236062040158e-05,
"loss": 0.397,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 0.41811526835811935,
"learning_rate": 3.798770948033553e-05,
"loss": 0.394,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.3926683484569302,
"learning_rate": 3.796291451102381e-05,
"loss": 0.3995,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.3901540250534727,
"learning_rate": 3.793797591072631e-05,
"loss": 0.3702,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.3869780613353573,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4054,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.40334663462530995,
"learning_rate": 3.788766861595443e-05,
"loss": 0.4103,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 0.3830366119856383,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.3852,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 0.3981286264595424,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4084,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.4327285885197706,
"learning_rate": 3.781113546385265e-05,
"loss": 0.3781,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 0.4229835645102718,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.3985,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.3525038608941972,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.3941,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.4147982282657617,
"learning_rate": 3.77333205623025e-05,
"loss": 0.3894,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.36788374983962213,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.3827,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.35146821748608853,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.3811,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 0.4018737986177741,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4255,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.3605948636377717,
"learning_rate": 3.762758321970216e-05,
"loss": 0.378,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.4290061042681929,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.4045,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.45263705481736777,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4086,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 0.3544822144490669,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.3794,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.40491725447532695,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.3882,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.36966026379346484,
"learning_rate": 3.749224181796305e-05,
"loss": 0.388,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.39766468254941506,
"learning_rate": 3.746475306375545e-05,
"loss": 0.4239,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 0.3464465705398607,
"learning_rate": 3.743712466184774e-05,
"loss": 0.3791,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 0.3679042973862517,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.3875,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.4034870152255123,
"learning_rate": 3.73814497997107e-05,
"loss": 0.3959,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 0.41962456966416267,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.3889,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.37782354297174886,
"learning_rate": 3.732521901224724e-05,
"loss": 0.4143,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.44565800111602555,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.3994,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.38082909843334306,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.39,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.4130405134275214,
"learning_rate": 3.723983441007882e-05,
"loss": 0.3894,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.37364809769457985,
"learning_rate": 3.721109687296781e-05,
"loss": 0.3915,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.47996440638463883,
"learning_rate": 3.718222171638484e-05,
"loss": 0.3961,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.3326446787174454,
"learning_rate": 3.715320917121488e-05,
"loss": 0.4146,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.42496798743767095,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4085,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.3464000005289976,
"learning_rate": 3.709477284414488e-05,
"loss": 0.373,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 0.39959815195930004,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.3774,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 0.357913520383353,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.417,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.35758914271114367,
"learning_rate": 3.70060937743295e-05,
"loss": 0.4012,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.3327441891648363,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.3798,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 0.405558008408539,
"learning_rate": 3.694629409915555e-05,
"loss": 0.4121,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.3376047391106543,
"learning_rate": 3.691619088858401e-05,
"loss": 0.389,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.3372606371224359,
"learning_rate": 3.68859524166009e-05,
"loss": 0.3937,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.34690275556204875,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.402,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.34954619418235805,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.3923,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 0.3202581862386034,
"learning_rate": 3.67944278554384e-05,
"loss": 0.3948,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 0.36306860752935716,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4054,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.3537472741600283,
"learning_rate": 3.673273963576791e-05,
"loss": 0.376,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 0.38891159260287667,
"learning_rate": 3.670169471054001e-05,
"loss": 0.3926,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.34903651512140865,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4008,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.39099036065301507,
"learning_rate": 3.663920447046888e-05,
"loss": 0.395,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.323792840760939,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.3998,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.3798776699607233,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4096,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.38852431287199557,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4225,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.34168103221672674,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.3892,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.37500235849972763,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.4012,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.44792905028837243,
"learning_rate": 3.644854871713611e-05,
"loss": 0.4133,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.31822374860286085,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.3912,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 0.4049072198658537,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4124,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.3649603436382293,
"learning_rate": 3.635144184933412e-05,
"loss": 0.415,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.38090139078973784,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.3981,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.34710909225800435,
"learning_rate": 3.628604977762752e-05,
"loss": 0.3877,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.40612753194938384,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.3968,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.33299388273847813,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.3955,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.3382314025039824,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4166,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.3332616853131732,
"learning_rate": 3.615370507164654e-05,
"loss": 0.4113,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.37274184679432076,
"learning_rate": 3.61202953196805e-05,
"loss": 0.3798,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 0.38110912551346265,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.3893,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.3604095836272426,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.3902,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.3985316447418196,
"learning_rate": 3.601929375279662e-05,
"loss": 0.4112,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.37053284149817395,
"learning_rate": 3.598537002418012e-05,
"loss": 0.3772,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.44770496209119187,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4159,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 0.3756219600603902,
"learning_rate": 3.591713938344416e-05,
"loss": 0.3909,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.38022713970230076,
"learning_rate": 3.588283301689507e-05,
"loss": 0.3587,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.3687290731724133,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.4045,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.3692425120606892,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.3953,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.4090171417458383,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.413,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.3682611781705827,
"learning_rate": 3.574434032194994e-05,
"loss": 0.421,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.3940171504933531,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4133,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.363253914773402,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.3933,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.3837466994059811,
"learning_rate": 3.563914797637216e-05,
"loss": 0.392,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.46146180099027695,
"learning_rate": 3.560383338592649e-05,
"loss": 0.4211,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.38552038647708403,
"learning_rate": 3.556839402764487e-05,
"loss": 0.3937,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 0.4181176172606353,
"learning_rate": 3.553283018489946e-05,
"loss": 0.3806,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.36629484937122136,
"learning_rate": 3.54971421420578e-05,
"loss": 0.367,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.3495567171080261,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.4027,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 0.3627665747674033,
"learning_rate": 3.54253945985191e-05,
"loss": 0.4235,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.3673869981795747,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.3877,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 0.4444159249366251,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.3933,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.3795532291620352,
"learning_rate": 3.531684894865866e-05,
"loss": 0.401,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.4780759641785068,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4081,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.37800602392152144,
"learning_rate": 3.524387233431892e-05,
"loss": 0.4148,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.47444357411949306,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4269,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.3750916868578045,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.3809,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 0.40639906767064,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4048,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.36756967724669487,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.3889,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.41050925290547435,
"learning_rate": 3.505930287892533e-05,
"loss": 0.4157,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.4098612083915544,
"learning_rate": 3.502202656100099e-05,
"loss": 0.4144,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.3631457467618242,
"learning_rate": 3.498463012735256e-05,
"loss": 0.423,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.4186139936972417,
"learning_rate": 3.494711387700094e-05,
"loss": 0.3935,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.33196359726244246,
"learning_rate": 3.490947810992505e-05,
"loss": 0.3724,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 0.3842052920508146,
"learning_rate": 3.48717231270595e-05,
"loss": 0.3914,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.3567160991034479,
"learning_rate": 3.483384923029211e-05,
"loss": 0.4118,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.40214034055278997,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4066,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 0.33246716544843513,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.3765,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.36931995634794385,
"learning_rate": 3.471951708970527e-05,
"loss": 0.4019,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.3774105944818115,
"learning_rate": 3.468117057518921e-05,
"loss": 0.4011,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 0.3247954618235457,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.3969,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.3810527058557629,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4038,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.3968219913880893,
"learning_rate": 3.456542792131039e-05,
"loss": 0.403,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.3752369717895491,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.4047,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.39353771189652775,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4097,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 0.36424382666482297,
"learning_rate": 3.44486370900573e-05,
"loss": 0.3736,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.35766251291910806,
"learning_rate": 3.440947533516396e-05,
"loss": 0.3764,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.3704058104974219,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.4158,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 0.36456378773051884,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4289,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.39613300265075824,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4055,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.370073527741735,
"learning_rate": 3.425167928295015e-05,
"loss": 0.4023,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.36871056731874874,
"learning_rate": 3.421194458890411e-05,
"loss": 0.3971,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 0.3356306037077937,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.4007,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.36639523984354266,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.393,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.3498697132107335,
"learning_rate": 3.409205995219561e-05,
"loss": 0.4061,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 0.33409683082505465,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.4018,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 0.37830742940795503,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4024,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 0.37100393324594777,
"learning_rate": 3.397116120327025e-05,
"loss": 0.3821,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.37689591289327384,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4085,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 0.36525066954878865,
"learning_rate": 3.389000293509524e-05,
"loss": 0.3856,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 0.36603120428520575,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4108,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.3378565517929529,
"learning_rate": 3.380840041143494e-05,
"loss": 0.3949,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.3856479017248662,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4266,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.32702781546495696,
"learning_rate": 3.372635624225055e-05,
"loss": 0.3844,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.3593381566604198,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.3901,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.3144476487997702,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.405,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.37612179702796916,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.3939,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.3448977630039844,
"learning_rate": 3.356095347769799e-05,
"loss": 0.3871,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.34746035291357236,
"learning_rate": 3.351933087522552e-05,
"loss": 0.3984,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.33291271828956787,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4022,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.3730910629619769,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.4015,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.3504906772513143,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.4051,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 0.3693195287314728,
"learning_rate": 3.335176280432308e-05,
"loss": 0.3895,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.3661034874555027,
"learning_rate": 3.330960304618295e-05,
"loss": 0.4059,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.3645347438109002,
"learning_rate": 3.326733686481153e-05,
"loss": 0.4019,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.3556630646273687,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4004,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.35861634867939296,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4098,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.3444146989357686,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4132,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.35327401172872397,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4238,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.33153841949818846,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4024,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.35055534724381127,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4103,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.35282510207764806,
"learning_rate": 3.296852223985148e-05,
"loss": 0.3952,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 0.3658262278654931,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.3991,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.3778214870407826,
"learning_rate": 3.288220822900796e-05,
"loss": 0.3934,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.3795441072638457,
"learning_rate": 3.283889654214788e-05,
"loss": 0.3954,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.35871025344517954,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4048,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.3710335960271525,
"learning_rate": 3.275196553714379e-05,
"loss": 0.3829,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.34371406919498176,
"learning_rate": 3.270834691409778e-05,
"loss": 0.3848,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.34387109694399287,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4055,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.3760822773376088,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4031,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.3652546943156867,
"learning_rate": 3.257688275049323e-05,
"loss": 0.3978,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 0.38650140541011496,
"learning_rate": 3.253285976578535e-05,
"loss": 0.386,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.39879241300370105,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.3877,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.3284748904644566,
"learning_rate": 3.244451351193935e-05,
"loss": 0.3973,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.38432135615041924,
"learning_rate": 3.240019094921551e-05,
"loss": 0.3958,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.34346324252486604,
"learning_rate": 3.235576923489527e-05,
"loss": 0.387,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.37373418929342134,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.4056,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.33568735760377966,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4042,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 0.3237841857496589,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.4059,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.3852823969619865,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.399,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 0.33233396154974704,
"learning_rate": 3.21321858487569e-05,
"loss": 0.4046,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.33584988218194195,
"learning_rate": 3.208717671149931e-05,
"loss": 0.3881,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.35219523569769917,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4067,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 0.3471104905064542,
"learning_rate": 3.19968688514201e-05,
"loss": 0.3985,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 0.33706209083978306,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4143,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.3498235827205893,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4237,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.33894071235286355,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4103,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.3208369064183869,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4046,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.333598298793165,
"learning_rate": 3.176942683684842e-05,
"loss": 0.3849,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.32174217037911773,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.3986,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.3397854754141119,
"learning_rate": 3.167778872346327e-05,
"loss": 0.4089,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.3224907508821308,
"learning_rate": 3.163182942046815e-05,
"loss": 0.391,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.3154824537668621,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4004,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.3527993459127271,
"learning_rate": 3.153963215931669e-05,
"loss": 0.3763,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.3397130206604165,
"learning_rate": 3.14933949383672e-05,
"loss": 0.4087,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.3241175314039794,
"learning_rate": 3.144706581653805e-05,
"loss": 0.3854,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.3826698806807725,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4119,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.3178873168250489,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4006,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 0.4202845354710102,
"learning_rate": 3.130753075389327e-05,
"loss": 0.3882,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.33144042192840456,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.3983,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.34194608133070425,
"learning_rate": 3.121405468537547e-05,
"loss": 0.3903,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.3461327498347712,
"learning_rate": 3.116718196315488e-05,
"loss": 0.387,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.36827068350018827,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4097,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 0.3339991467361592,
"learning_rate": 3.107316901675252e-05,
"loss": 0.3964,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.3670128516677621,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4011,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.36083649043580607,
"learning_rate": 3.097880190800057e-05,
"loss": 0.4022,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.3475858465161063,
"learning_rate": 3.093148648549788e-05,
"loss": 0.3998,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.33559889878639904,
"learning_rate": 3.088408365512055e-05,
"loss": 0.3804,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.36847503408219856,
"learning_rate": 3.083659379590034e-05,
"loss": 0.4219,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.34879449986381944,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.4087,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.3349034673707377,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4013,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.3679558537118173,
"learning_rate": 3.069360584592015e-05,
"loss": 0.3968,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.36296797919173407,
"learning_rate": 3.064577167551816e-05,
"loss": 0.3759,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 0.3402733659246349,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4008,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.4282263170078463,
"learning_rate": 3.054984834795566e-05,
"loss": 0.3882,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.31100259273696723,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.389,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.3820015496477748,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.3862,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.4140770220291767,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4186,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.32942085373446295,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.3861,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.3541866227777258,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.3834,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.36476970535539244,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4267,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.3594374243245046,
"learning_rate": 3.021147971398632e-05,
"loss": 0.3828,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.34392273826724373,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.425,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.3410889150529019,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.3966,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.31751138324484135,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4026,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.35815318618323494,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4031,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.44980966869806294,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.2753,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.35552278312025376,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.255,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.5134341821547689,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.2471,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.3434428464854084,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.2525,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.3944227083269992,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2412,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.34640775303052723,
"learning_rate": 2.972118904869376e-05,
"loss": 0.2442,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.36128749103190966,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.2217,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.3429726368880411,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.2298,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.3311810380943305,
"learning_rate": 2.95725682050846e-05,
"loss": 0.244,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.33347022951039135,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.2277,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.30064795460047644,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.2555,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.36286742360429153,
"learning_rate": 2.942325848714505e-05,
"loss": 0.2474,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.30942056625993425,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.2765,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.32120338480518557,
"learning_rate": 2.932334127129686e-05,
"loss": 0.241,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.29554776475769756,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.2649,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.3600920956727756,
"learning_rate": 2.922312585929131e-05,
"loss": 0.2622,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.32880990753219685,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.2424,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.2932999342995316,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.2299,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.354118857010248,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.2654,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.31474095977863353,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.2296,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.3060506209123143,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.2635,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 0.32265017354798975,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.2466,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.3177053649678313,
"learning_rate": 2.887006998351372e-05,
"loss": 0.2695,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.2971789104986174,
"learning_rate": 2.881934649605535e-05,
"loss": 0.2441,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.2818324617005397,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.2483,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.28086314050609923,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.2346,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.28809183103305375,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2396,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.2739735888658174,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.2412,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.2539149356461799,
"learning_rate": 2.856467940121025e-05,
"loss": 0.2545,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.2979646859889247,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2427,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.28533022927438484,
"learning_rate": 2.846233032790149e-05,
"loss": 0.2597,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.25764139965383437,
"learning_rate": 2.84110540896642e-05,
"loss": 0.2346,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.31189377827969444,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.2403,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.2635990809046953,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2543,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.280316214352365,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.252,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.27052334236281744,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.2549,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.30712698914712605,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.2194,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.2809564340368507,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.2555,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 0.2865321284143028,
"learning_rate": 2.805026032173773e-05,
"loss": 0.2561,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.2873204928817813,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.2476,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.2672883143497271,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.2442,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.2987249591636575,
"learning_rate": 2.7894660620408e-05,
"loss": 0.2562,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 0.2832435347615956,
"learning_rate": 2.784266724883173e-05,
"loss": 0.2459,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.29302095475996515,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.2441,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.2633114672423965,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.2392,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.2800689566197326,
"learning_rate": 2.768631254100716e-05,
"loss": 0.2529,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.29924100059705167,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.2272,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.26250726515110295,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.246,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.28357995055388313,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.2439,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.27821130308991754,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.2331,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.3423575701957564,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2626,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.2798989473305917,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.2662,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.31724025077448853,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.2228,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.2685977909522443,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.237,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.31524121748382966,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.2411,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.28583371679756764,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.2424,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.30580171911887905,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.232,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.29393174437802155,
"learning_rate": 2.705547548999996e-05,
"loss": 0.2643,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.2898792329850466,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.2689,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.280562617132628,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.2639,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.29019495185194344,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.2752,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.2792975354487715,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.2578,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.29538829773367237,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.2302,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 0.2787681113816963,
"learning_rate": 2.67369650552155e-05,
"loss": 0.2546,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.2797793532851524,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.2634,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.27789851055823744,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.2382,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.29650956451718063,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.2523,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.2673972720769891,
"learning_rate": 2.652354173593614e-05,
"loss": 0.2514,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.26932976730657354,
"learning_rate": 2.647005443331317e-05,
"loss": 0.2415,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.27377892732345693,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.2412,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.27258405213940284,
"learning_rate": 2.636292505301542e-05,
"loss": 0.2457,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.2815566946109616,
"learning_rate": 2.630928383194431e-05,
"loss": 0.2299,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.26249847985907226,
"learning_rate": 2.625559216200781e-05,
"loss": 0.251,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.29960657562093235,
"learning_rate": 2.620185047252308e-05,
"loss": 0.2297,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.27601075170808304,
"learning_rate": 2.614805919320723e-05,
"loss": 0.2514,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.31864246354155357,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.2684,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.33276003474731414,
"learning_rate": 2.60403295859298e-05,
"loss": 0.2476,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.31492607723344856,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2384,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.31673236821860284,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.2507,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.31265067506631333,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.2558,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.3042578226168231,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.2442,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.29655524567234864,
"learning_rate": 2.577016790136453e-05,
"loss": 0.2587,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.2876768830021647,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.2334,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.29700407841913445,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.2497,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.2955999851605047,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.2338,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.29733527811878147,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.2619,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.2851844079505442,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.2619,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.29405218698731306,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.2435,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.2799252777189737,
"learning_rate": 2.539001591449889e-05,
"loss": 0.2362,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 0.28999418912788427,
"learning_rate": 2.53355326151989e-05,
"loss": 0.2398,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.2854634328718973,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2408,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.2972392709463094,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.2737,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.2778695234035645,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2514,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.28449131641669917,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.2403,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.26758629037439324,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2281,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.3046810725843337,
"learning_rate": 2.500775217241923e-05,
"loss": 0.2819,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.29288344309144043,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.2725,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.2793796561156732,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2321,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.28834633240020485,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.2497,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.2999700766032818,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.2511,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.2774450517433758,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.2508,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.2839561025780153,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.249,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 0.30339161319990726,
"learning_rate": 2.46235264422994e-05,
"loss": 0.2552,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.29566139857646445,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.2329,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.30979383603828986,
"learning_rate": 2.451340834511592e-05,
"loss": 0.2591,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.27233407682712063,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.2502,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.30177750105049583,
"learning_rate": 2.440314589184958e-05,
"loss": 0.2577,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.2985198370668689,
"learning_rate": 2.434796163357541e-05,
"loss": 0.2493,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.29175391548661755,
"learning_rate": 2.429274260911588e-05,
"loss": 0.2221,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.2713660727966892,
"learning_rate": 2.423748926000084e-05,
"loss": 0.2343,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.3108677162822014,
"learning_rate": 2.418220202803459e-05,
"loss": 0.2297,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.2779372482394146,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2513,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.28418349103708135,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.2618,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.27992797954385296,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2306,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.2887389478917912,
"learning_rate": 2.396072311715187e-05,
"loss": 0.248,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.29969573543901556,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2484,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.27693397272079956,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.2561,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.2919250172831843,
"learning_rate": 2.379427985200223e-05,
"loss": 0.2644,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.3065645974500416,
"learning_rate": 2.373873749394972e-05,
"loss": 0.2241,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.2901221331885418,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.2218,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.2951080589040094,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.2649,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.31981569405983734,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.2559,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.2881371229466053,
"learning_rate": 2.351627355800233e-05,
"loss": 0.2641,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.2875226386658525,
"learning_rate": 2.346058617160256e-05,
"loss": 0.2361,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.30378384129349995,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2518,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.301165203879785,
"learning_rate": 2.334912883203469e-05,
"loss": 0.2571,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.28888153749090156,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.24,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.32515320037710405,
"learning_rate": 2.323756437450136e-05,
"loss": 0.2382,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.2743221423830085,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.2566,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.31122179335366157,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.2392,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.29542915571723855,
"learning_rate": 2.307002464849052e-05,
"loss": 0.2462,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.26716261088022747,
"learning_rate": 2.30141283818846e-05,
"loss": 0.2539,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.32176344856021444,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.2427,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 0.2896758146399387,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.2357,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.3225326002999325,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.2325,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.30024815626850165,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.2359,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.278764764237527,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2563,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.29648565162524815,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.2486,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.27427445146333723,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2357,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.28935976232723304,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.2506,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.3005367537468842,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.2571,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.30559442347164745,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.2464,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.309682414508718,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.2518,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.26753113303961296,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.2457,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.2845474447109726,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2248,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.29578000072064287,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.2412,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.2766960327139631,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.2509,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.2766367757583449,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.2344,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.29317070505185167,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.259,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.2735630734002488,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.2281,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.2718456496988286,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.2517,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.2666730988669495,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.2268,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.28276419781356044,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.2486,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.26444917595127315,
"learning_rate": 2.17791064266051e-05,
"loss": 0.243,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 0.27749265372686,
"learning_rate": 2.172276926861741e-05,
"loss": 0.2446,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.27562684158254275,
"learning_rate": 2.166641833541255e-05,
"loss": 0.2466,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.2858114629087196,
"learning_rate": 2.161005407757109e-05,
"loss": 0.2686,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.34598641958138177,
"learning_rate": 2.155367694578013e-05,
"loss": 0.2615,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.27521822442400945,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2372,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.2840798068909672,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.2341,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.28322748049304236,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.2635,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.281796245903486,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.2608,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.27425888398827114,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.2358,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.29329265877609195,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.2431,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.29169744937371966,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.2415,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.25936876583086443,
"learning_rate": 2.110225054402733e-05,
"loss": 0.2398,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.30448209550405664,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.2512,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.30598109999863404,
"learning_rate": 2.09892967807724e-05,
"loss": 0.232,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.2805064703069358,
"learning_rate": 2.093280780774835e-05,
"loss": 0.2421,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.30040764672728415,
"learning_rate": 2.087631137601793e-05,
"loss": 0.2574,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.2761762034023045,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.2296,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.30886446685411195,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.2515,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.32290724658145703,
"learning_rate": 2.070678184630458e-05,
"loss": 0.2496,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.27122593048803995,
"learning_rate": 2.06502600977305e-05,
"loss": 0.2432,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.30084019414414376,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.2784,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.2787792084211904,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.2365,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.27336019799471073,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.2329,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.26120146339595623,
"learning_rate": 2.042412562887594e-05,
"loss": 0.2655,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.254534484704883,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.2707,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.26212390705763666,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.2508,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.27886767762243203,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.2472,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.2879920241679287,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.2456,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.2727526213415045,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.2511,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.28481409750211617,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.2431,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.2556470394772378,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.2596,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 0.2692085529889135,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.247,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.2995387471625604,
"learning_rate": 1.99151687695709e-05,
"loss": 0.2352,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.2544893030848492,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.2356,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.26691623669571646,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.2663,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.2576281096146166,
"learning_rate": 1.974551241345447e-05,
"loss": 0.2574,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.27480495072374034,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.252,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.28939933742492674,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2324,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.26414795450866746,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.2604,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.27100984785777077,
"learning_rate": 1.951933453677638e-05,
"loss": 0.2476,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.2677175855437437,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.2529,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.2594926574930046,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.2358,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.268892757098544,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.2655,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.27738876741425067,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.2409,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.2646583219164703,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.2328,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.2710073156889436,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2341,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.2683119838715539,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.25,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.2800806670497316,
"learning_rate": 1.906719219225165e-05,
"loss": 0.2616,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.2591618310225513,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.2355,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.2730418563363313,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.247,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.26805211360135556,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.2544,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.26152503248121867,
"learning_rate": 1.884128556891707e-05,
"loss": 0.248,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.263746387857871,
"learning_rate": 1.878483094691119e-05,
"loss": 0.2507,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.27636173297142863,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.2522,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.393251447018487,
"learning_rate": 1.867195130360947e-05,
"loss": 0.2518,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.2632148099800927,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.2552,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.2808576316722852,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.2277,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.26464077510951806,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.2327,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.28927337959351773,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.2296,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.2634682259999594,
"learning_rate": 1.838994592242891e-05,
"loss": 0.2371,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.28292542198216986,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.2342,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.2646297508618514,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.2443,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.28981908753204044,
"learning_rate": 1.82208935733949e-05,
"loss": 0.2304,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.2808048104764432,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.2437,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.25724171972895155,
"learning_rate": 1.810826238483892e-05,
"loss": 0.2475,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.27921364107905766,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.2427,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.24887486003337259,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.2486,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.2578455983540125,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.2532,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.2612830194494556,
"learning_rate": 1.788318512320838e-05,
"loss": 0.2566,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.264357771035379,
"learning_rate": 1.782695699829051e-05,
"loss": 0.2337,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.28246137146697586,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.2343,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.2658206174464214,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.2371,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.26811433156006276,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.2379,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.2756005934077426,
"learning_rate": 1.76022227484387e-05,
"loss": 0.2676,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.2699458996475736,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.2368,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.2685811736484852,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.2594,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.27329735003629624,
"learning_rate": 1.743387180005809e-05,
"loss": 0.253,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.25390332630077334,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2251,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.2606001329232142,
"learning_rate": 1.732173968007914e-05,
"loss": 0.2409,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.27995247960350017,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.2333,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.2675954254358945,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2441,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.24752026202621394,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2552,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.3052035685928038,
"learning_rate": 1.709773600686015e-05,
"loss": 0.2602,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.27817230493096545,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.2665,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.2843298731366203,
"learning_rate": 1.69858716181154e-05,
"loss": 0.2363,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.27218622750507165,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2599,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.2655344261918971,
"learning_rate": 1.687410363273884e-05,
"loss": 0.2185,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.2564863759919573,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.2418,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.28232964258208637,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.2568,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.263761390803984,
"learning_rate": 1.67066402299235e-05,
"loss": 0.2455,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.25997549366072054,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.2034,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.27750043179062245,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.2539,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.2571175656063153,
"learning_rate": 1.653941382839745e-05,
"loss": 0.2418,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.39527858400132587,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.1714,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.29453427431688584,
"learning_rate": 1.642806717162757e-05,
"loss": 0.1621,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.21139191512125044,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1298,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.33603755098260524,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.156,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.39263546868474225,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.1825,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.2877610192956902,
"learning_rate": 1.620572014799777e-05,
"loss": 0.1435,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.26090449218881867,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.1713,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.26339255810985185,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.1438,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.25345427029795925,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1488,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.2502098266724432,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.1646,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.23589025517386197,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.1669,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.2614852787874558,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.1412,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.2315525507630847,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.1552,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.2469386162448032,
"learning_rate": 1.576251073999917e-05,
"loss": 0.1753,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.22796337966956468,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.1388,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.23045919174096505,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.152,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.21702465953361347,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.1636,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 0.22641491791962814,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.1456,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.22854893132716717,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.1503,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.2013300269482367,
"learning_rate": 1.543151434154755e-05,
"loss": 0.1574,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.2088434623619834,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.1471,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.20660841717597453,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.1391,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.22700849749426857,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.1419,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.19862066102179543,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.1386,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.20023758152622978,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.1596,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.21169712060738674,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.1491,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.2061018357022915,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.1663,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.20661990815211267,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.1612,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.2240710801938828,
"learning_rate": 1.493751509814374e-05,
"loss": 0.1377,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.22753246120217976,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1696,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.20666155939652062,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.1542,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.209731915691284,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1357,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 0.21197030767939776,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.1371,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.2182091121792086,
"learning_rate": 1.466446738480111e-05,
"loss": 0.1574,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.19938989816872338,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1575,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 0.19226321280346487,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.1541,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.21347611683338955,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.1361,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.21715134353184704,
"learning_rate": 1.444679451900437e-05,
"loss": 0.1446,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.20894184444113534,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.1485,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.20904109466338142,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.1569,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.20719019785657397,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.1298,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.19505148380458326,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.1582,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.2269816275243439,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.1333,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.22073190870523304,
"learning_rate": 1.412162598317602e-05,
"loss": 0.1551,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.20612645141713126,
"learning_rate": 1.406759321421907e-05,
"loss": 0.1381,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.20222178879456812,
"learning_rate": 1.401360788062871e-05,
"loss": 0.165,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.20332824327563248,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.1423,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.1883641064694965,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.136,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.19530142737299352,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.1464,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.20636813136747445,
"learning_rate": 1.379814952747693e-05,
"loss": 0.1423,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.1949364760746513,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.1317,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.22216088986718527,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.1409,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.18920265497922,
"learning_rate": 1.363707494698459e-05,
"loss": 0.1271,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.21678661389787166,
"learning_rate": 1.358348460369265e-05,
"loss": 0.1475,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.20923016600371078,
"learning_rate": 1.352994556668684e-05,
"loss": 0.1727,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.19733949597532244,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.1469,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.19653207008597026,
"learning_rate": 1.342302312350678e-05,
"loss": 0.1589,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.20879717832663747,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.1587,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.20624719159259022,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1768,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.203017977120565,
"learning_rate": 1.326303494478451e-05,
"loss": 0.1569,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.2042463454492776,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.1432,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.19540411010163689,
"learning_rate": 1.315664479122898e-05,
"loss": 0.1452,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.1814339550113446,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.1494,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.18265571812019313,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.1387,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.187535433376629,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.1541,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.1972364698227112,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.1333,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.20265912050766446,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.1601,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.1917287467788,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.1292,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.2062189660796884,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.1521,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.19661185943108908,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.1472,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.2023942193159033,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.1502,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.2078152300732994,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.1321,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.2043640917900557,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.1547,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.20775739513980418,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.1658,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.21777659293427232,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.1936,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.20247486065900935,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.14,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.21114203881124044,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.1541,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.19849074600844438,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.1242,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.19106093821575357,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.1203,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.20966264168655804,
"learning_rate": 1.220938883250288e-05,
"loss": 0.1632,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.19135369169318286,
"learning_rate": 1.215733275116827e-05,
"loss": 0.1465,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.20674160160164068,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.1554,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.18890755527266792,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.1459,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.18972990585090302,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.1425,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.2095008820076673,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.145,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.19195677396042396,
"learning_rate": 1.189800129803139e-05,
"loss": 0.1482,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.19875779019758952,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.1463,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.20074471926218884,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.1389,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.19239249405371248,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.1522,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.202856077244993,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.1596,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.20556061432626446,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.1477,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.2031742235592653,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.1615,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.20802573144298925,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.1599,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.20991114031476157,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.1455,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.20137748913709383,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.1508,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.20136056603079264,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.1386,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 0.2042511001548749,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.1509,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.18692386091119,
"learning_rate": 1.128231163049258e-05,
"loss": 0.1367,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.20121677206093194,
"learning_rate": 1.123144751066259e-05,
"loss": 0.1588,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.19713502293574375,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.1591,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.19806208915165907,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.1871,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.19495746571946304,
"learning_rate": 1.107927745387114e-05,
"loss": 0.1587,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.19229378492203564,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.1345,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.1953005963615259,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.1541,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.20459939436491245,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.1583,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.19279052365833588,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.1403,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.18042589719452537,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.1245,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.19765840354751574,
"learning_rate": 1.077687414070869e-05,
"loss": 0.133,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.21013829918349036,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.1434,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.20635877898257265,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.1586,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.1897709907622457,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.1426,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.19512439035232867,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.1521,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.20582512569261452,
"learning_rate": 1.052689572753247e-05,
"loss": 0.1483,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.1971016162050482,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.1479,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.18431866962714116,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.1284,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.2003917670209446,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.134,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.18930781430075894,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.1418,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.1894113499565902,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1742,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.19082207209679858,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.175,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.18579641501110614,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.1366,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.19981241898227461,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.1402,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.1968090728801054,
"learning_rate": 1.008173988718005e-05,
"loss": 0.1292,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.1829792094175797,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.1689,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.1867266330295612,
"learning_rate": 9.983678617766677e-06,
"loss": 0.1355,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.19937595249993748,
"learning_rate": 9.934767922553788e-06,
"loss": 0.163,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.2080879023156462,
"learning_rate": 9.885937708668371e-06,
"loss": 0.1339,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.1880200027744832,
"learning_rate": 9.837188366555524e-06,
"loss": 0.1251,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.19262983858023774,
"learning_rate": 9.788520286013691e-06,
"loss": 0.1431,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.1907342446492516,
"learning_rate": 9.739933856191537e-06,
"loss": 0.1453,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.1952192387010953,
"learning_rate": 9.691429465584866e-06,
"loss": 0.1625,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.1963698283622041,
"learning_rate": 9.643007502033492e-06,
"loss": 0.133,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.19998946690195527,
"learning_rate": 9.594668352718152e-06,
"loss": 0.1534,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.18117289367690384,
"learning_rate": 9.546412404157383e-06,
"loss": 0.1253,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.1795376923304172,
"learning_rate": 9.498240042204471e-06,
"loss": 0.1596,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.19095484995781653,
"learning_rate": 9.450151652044339e-06,
"loss": 0.1281,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.19184054210295018,
"learning_rate": 9.402147618190472e-06,
"loss": 0.1348,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.19280269776721018,
"learning_rate": 9.354228324481841e-06,
"loss": 0.1448,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.18647195974127764,
"learning_rate": 9.306394154079852e-06,
"loss": 0.1369,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.19779475576076957,
"learning_rate": 9.258645489465254e-06,
"loss": 0.1576,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.19202393163598697,
"learning_rate": 9.21098271243509e-06,
"loss": 0.1495,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.19157813576838714,
"learning_rate": 9.163406204099659e-06,
"loss": 0.1415,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.19535872996446468,
"learning_rate": 9.115916344879452e-06,
"loss": 0.1565,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.1964838375627556,
"learning_rate": 9.068513514502126e-06,
"loss": 0.1369,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.18088112463830466,
"learning_rate": 9.021198091999439e-06,
"loss": 0.1413,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.18783360837805965,
"learning_rate": 8.973970455704251e-06,
"loss": 0.152,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.19349320224191197,
"learning_rate": 8.926830983247495e-06,
"loss": 0.1161,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.1890567829867351,
"learning_rate": 8.879780051555116e-06,
"loss": 0.1211,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.1845038256472456,
"learning_rate": 8.83281803684512e-06,
"loss": 0.1508,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.182314924140796,
"learning_rate": 8.785945314624532e-06,
"loss": 0.1454,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.19139917070845291,
"learning_rate": 8.739162259686385e-06,
"loss": 0.1403,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.20423385700400032,
"learning_rate": 8.692469246106734e-06,
"loss": 0.1321,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.18650312890696652,
"learning_rate": 8.645866647241677e-06,
"loss": 0.1251,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.187738670741468,
"learning_rate": 8.599354835724348e-06,
"loss": 0.1525,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.17821726361198564,
"learning_rate": 8.552934183461958e-06,
"loss": 0.1428,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.19159298739033534,
"learning_rate": 8.506605061632805e-06,
"loss": 0.1758,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.18367891189124927,
"learning_rate": 8.460367840683308e-06,
"loss": 0.148,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.17569943662036244,
"learning_rate": 8.414222890325065e-06,
"loss": 0.1443,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.18659047081563881,
"learning_rate": 8.368170579531855e-06,
"loss": 0.1386,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.18432501634264647,
"learning_rate": 8.322211276536733e-06,
"loss": 0.1679,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.18752705419335508,
"learning_rate": 8.27634534882907e-06,
"loss": 0.1296,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.18893807209898647,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1271,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.18590676173149184,
"learning_rate": 8.184895085497466e-06,
"loss": 0.1563,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.1960595219847967,
"learning_rate": 8.139311481107401e-06,
"loss": 0.1549,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.1848693734803453,
"learning_rate": 8.093822714466686e-06,
"loss": 0.1525,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.18712853420823,
"learning_rate": 8.048429149302271e-06,
"loss": 0.1623,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.1858702608513256,
"learning_rate": 8.003131148579902e-06,
"loss": 0.1424,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.18770237048483696,
"learning_rate": 7.957929074501182e-06,
"loss": 0.1449,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.17632964029567297,
"learning_rate": 7.912823288500695e-06,
"loss": 0.1493,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.1862045526842417,
"learning_rate": 7.867814151243102e-06,
"loss": 0.122,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 0.1960696862668804,
"learning_rate": 7.822902022620267e-06,
"loss": 0.1511,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.19705932819444522,
"learning_rate": 7.778087261748377e-06,
"loss": 0.1369,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.18015790579014085,
"learning_rate": 7.733370226965056e-06,
"loss": 0.1496,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.1849038199567583,
"learning_rate": 7.688751275826527e-06,
"loss": 0.139,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.1836872699226761,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1433,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.192877663358015,
"learning_rate": 7.599809050784502e-06,
"loss": 0.145,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.1854847954683257,
"learning_rate": 7.555486488060661e-06,
"loss": 0.148,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.1741021879142585,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1355,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.19421067183397706,
"learning_rate": 7.467140234214656e-06,
"loss": 0.1439,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.19275107851486523,
"learning_rate": 7.423117249506777e-06,
"loss": 0.1568,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.18681544552664822,
"learning_rate": 7.379194829218231e-06,
"loss": 0.1398,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.1839907707364559,
"learning_rate": 7.335373324551518e-06,
"loss": 0.1309,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.1791188020119541,
"learning_rate": 7.291653085902224e-06,
"loss": 0.1272,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.18693860060858408,
"learning_rate": 7.248034462856213e-06,
"loss": 0.124,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.1864734072083325,
"learning_rate": 7.20451780418683e-06,
"loss": 0.1563,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.18115628263550548,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1586,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.18260694328339838,
"learning_rate": 7.117791770992046e-06,
"loss": 0.142,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.19284706561741627,
"learning_rate": 7.074583089925688e-06,
"loss": 0.1442,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.17649985422286618,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.1612,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.1986238172062214,
"learning_rate": 6.988476126329604e-06,
"loss": 0.1402,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.18961241288079314,
"learning_rate": 6.945578532308857e-06,
"loss": 0.1542,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.18547615344955462,
"learning_rate": 6.902785321094301e-06,
"loss": 0.1365,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.18290864578601312,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.1359,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.1841017596146119,
"learning_rate": 6.817513414939892e-06,
"loss": 0.1368,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.1857059756298316,
"learning_rate": 6.775035401831924e-06,
"loss": 0.1385,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.18443252164706592,
"learning_rate": 6.732663135188473e-06,
"loss": 0.1477,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.18580952867718462,
"learning_rate": 6.690396953817055e-06,
"loss": 0.1702,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.1850548201551399,
"learning_rate": 6.648237195676927e-06,
"loss": 0.1326,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.18850484346574006,
"learning_rate": 6.606184197876395e-06,
"loss": 0.1539,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.18192021505743294,
"learning_rate": 6.564238296670109e-06,
"loss": 0.1314,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.1826139591599639,
"learning_rate": 6.52239982745638e-06,
"loss": 0.1499,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.1893607950687341,
"learning_rate": 6.480669124774484e-06,
"loss": 0.1528,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.19410098812787072,
"learning_rate": 6.439046522302013e-06,
"loss": 0.1417,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.17748274292525337,
"learning_rate": 6.397532352852191e-06,
"loss": 0.1612,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.1805174791819435,
"learning_rate": 6.356126948371202e-06,
"loss": 0.1596,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.19062146899419022,
"learning_rate": 6.314830639935559e-06,
"loss": 0.155,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.18331073770038667,
"learning_rate": 6.273643757749448e-06,
"loss": 0.1376,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.18131802670941324,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1212,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.17915415480367808,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1222,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.19753298326338636,
"learning_rate": 6.150742957589784e-06,
"loss": 0.1524,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.18866259027075216,
"learning_rate": 6.109997064904774e-06,
"loss": 0.155,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.1789776913574906,
"learning_rate": 6.069362236313092e-06,
"loss": 0.1226,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.17911661316407348,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.1467,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.1963782830718025,
"learning_rate": 5.988427070179108e-06,
"loss": 0.149,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.1884635870088772,
"learning_rate": 5.948127379792234e-06,
"loss": 0.1409,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.17710961452638718,
"learning_rate": 5.90794004780439e-06,
"loss": 0.132,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.18259058463598085,
"learning_rate": 5.867865395552405e-06,
"loss": 0.1334,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.1783278132153821,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.1521,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.17155894303428573,
"learning_rate": 5.788055411095892e-06,
"loss": 0.1317,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.17905236824059176,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.1669,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.1832387840647379,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1422,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.17130923519055657,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1405,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.18593989301434127,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1528,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.18214360262353402,
"learning_rate": 5.590524664836043e-06,
"loss": 0.1327,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.17650129037368562,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1403,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.18126281744421238,
"learning_rate": 5.51231668596653e-06,
"loss": 0.1444,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.18624378093652794,
"learning_rate": 5.47338630512007e-06,
"loss": 0.1573,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.18870083379118524,
"learning_rate": 5.434572078689617e-06,
"loss": 0.1348,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.19360957105244064,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1433,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 0.1832515929753099,
"learning_rate": 5.357293329575573e-06,
"loss": 0.1325,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.18265694594618903,
"learning_rate": 5.318829424810797e-06,
"loss": 0.157,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.18527488330273006,
"learning_rate": 5.280482910294728e-06,
"loss": 0.1547,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.17571348254025873,
"learning_rate": 5.242254092645078e-06,
"loss": 0.1687,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.18056076697662632,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.1267,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.1845564542123631,
"learning_rate": 5.166150769707892e-06,
"loss": 0.1304,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.17200837502464367,
"learning_rate": 5.128276872940503e-06,
"loss": 0.1612,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.18198012397679023,
"learning_rate": 5.090521890074953e-06,
"loss": 0.1344,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.1769198849580927,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.1629,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.17268135919335711,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1547,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.18255565451164715,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.1386,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.17725029367774892,
"learning_rate": 4.940697121074675e-06,
"loss": 0.1343,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.19049209359941477,
"learning_rate": 4.903541216934864e-06,
"loss": 0.1668,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.18197408085181732,
"learning_rate": 4.8665060236772e-06,
"loss": 0.1403,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.1812776205777416,
"learning_rate": 4.829591837434102e-06,
"loss": 0.1389,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.19187398559716767,
"learning_rate": 4.792798953370424e-06,
"loss": 0.1273,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.1872665336327696,
"learning_rate": 4.756127665681083e-06,
"loss": 0.1369,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.18410846593325061,
"learning_rate": 4.719578267588725e-06,
"loss": 0.1284,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.16318477717224003,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.1366,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.26936549826247275,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1045,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.21452453995002427,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.23057688135849874,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1088,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.1692365104401245,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.0932,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.11869212528344537,
"learning_rate": 4.502857857942204e-06,
"loss": 0.0943,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.17050627181005287,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1103,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.19501812008290192,
"learning_rate": 4.431605972355137e-06,
"loss": 0.092,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.24701526390198672,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1062,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.22523417668412776,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1078,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.19584363147498962,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1061,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.16753037810688723,
"learning_rate": 4.29059827474257e-06,
"loss": 0.0906,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.1499703325870608,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1101,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.1378296419469806,
"learning_rate": 4.22084697268323e-06,
"loss": 0.0887,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.15811535698970972,
"learning_rate": 4.186160437003535e-06,
"loss": 0.1237,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.1616571957800394,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1006,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.15937444566546843,
"learning_rate": 4.11716698310493e-06,
"loss": 0.123,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.14471555548542286,
"learning_rate": 4.082860616555848e-06,
"loss": 0.0768,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.1464411360238387,
"learning_rate": 4.048681523029125e-06,
"loss": 0.0816,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.14759386211614647,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1042,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.13782933588430218,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1021,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.13955675074718693,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1016,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.15711967869741533,
"learning_rate": 3.913243329737353e-06,
"loss": 0.0951,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.1488276778824095,
"learning_rate": 3.879704680319503e-06,
"loss": 0.1041,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.14391817025768244,
"learning_rate": 3.846294928353471e-06,
"loss": 0.0953,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.1576828467563309,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1245,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.13728919046004656,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.0906,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.1394888594554878,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1012,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.1451680073198867,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1146,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.14786460944634652,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1196,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.145257605754172,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1083,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.12612647382822598,
"learning_rate": 3.616058102895075e-06,
"loss": 0.0857,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.13794205124918874,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1054,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.1485935699450192,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.0867,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.14371287832898874,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1085,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.13393082587637986,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.1024,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.13111990054524006,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1035,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.13321315045118096,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.0921,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.14462119893672862,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1034,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.13792721857485926,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.103,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.13126433585474678,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1087,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.13184103822416923,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.093,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.12302021392192727,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.097,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.13693388379624613,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1145,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.12644949151128987,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1219,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.14827262948091877,
"learning_rate": 3.174929343376374e-06,
"loss": 0.0909,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.1473305699317708,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.1214,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.13904097536862348,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.0913,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.14119208099681518,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.1111,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.13691335847051886,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1009,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.1278074863551623,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.0883,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.12957006637982946,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.0965,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.1326584584552232,
"learning_rate": 2.964210239224696e-06,
"loss": 0.0913,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.13647384665161594,
"learning_rate": 2.934650470499869e-06,
"loss": 0.0993,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.1356539417230338,
"learning_rate": 2.905227155855128e-06,
"loss": 0.0978,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.1293400398065256,
"learning_rate": 2.875940530558519e-06,
"loss": 0.084,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.1246154113012909,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.0939,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.11935607275414799,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.0831,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.1499291136696628,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1264,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.14474384807194157,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1089,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.1464409442600202,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.0998,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.13013147753434193,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1031,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.14153961233661663,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.105,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.1300683164001129,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.1023,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.13448349256646586,
"learning_rate": 2.618550200289303e-06,
"loss": 0.0825,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.14082918485111032,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1121,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.1480134732952289,
"learning_rate": 2.562875338152264e-06,
"loss": 0.107,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.13539964913961072,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1289,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.12227614598713879,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.0909,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.1300369848241089,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1191,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.13007481779791222,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1015,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.1275291977047206,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.0898,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.1396218542674767,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1157,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.13209244091919595,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.0923,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.1289459492784817,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.0868,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.13993484156812094,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1209,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.13723310410008122,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.0818,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.13301611707961405,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1158,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.13125986504000495,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1076,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.13027893336686022,
"learning_rate": 2.214660694699091e-06,
"loss": 0.1038,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.13392901245681685,
"learning_rate": 2.188864536147355e-06,
"loss": 0.0979,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.14331666509120947,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1073,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.13858854440862328,
"learning_rate": 2.137699676264171e-06,
"loss": 0.0872,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.1333366902318421,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.0882,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.13014465106796552,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1044,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.13419742216605887,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1308,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.136410904730083,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1035,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.12785351572969592,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1106,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.13670831509831224,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1096,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.12679279848815062,
"learning_rate": 1.963132265887919e-06,
"loss": 0.0977,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.13443964117520077,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.0886,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.1198127224904076,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.0936,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.1312025628643908,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.0965,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.14250674287251833,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1096,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.14184901312252482,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.0934,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.13600383026746715,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.0783,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.13565860794741674,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1074,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.13027962637292334,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.0957,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.13304202182341318,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.0901,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.13702001685085624,
"learning_rate": 1.726025117089718e-06,
"loss": 0.116,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.134117259795698,
"learning_rate": 1.70311497206598e-06,
"loss": 0.0975,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.13203469918221314,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1092,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.12864430032574814,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1069,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.1248200464433875,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1068,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.13894925649148643,
"learning_rate": 1.612939220355272e-06,
"loss": 0.0963,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.13858334956633261,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1108,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.12311575676846576,
"learning_rate": 1.568732761074121e-06,
"loss": 0.0937,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.13172161955339903,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1013,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.12426175652917408,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.0851,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.13085843950351683,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.0938,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.13989974938086713,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1106,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.1346082973196566,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.0907,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.1361488996208896,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.0898,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.13151636436876502,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.097,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.12656917717003566,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.104,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.12756506371419707,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.0889,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.1247046856536066,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.0928,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.13354140098875678,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1107,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.14275927057277227,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.133,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.13081170937593728,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.1042,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.1392355714811466,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1016,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.13321915208532564,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.105,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.13869316707412288,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1006,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.1313992100913352,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.119,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.13190981335888605,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.0864,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.1322676052588041,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.0975,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.1282106612699005,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1016,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.12556637822176928,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1222,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.12805088912577883,
"learning_rate": 1.121666393356604e-06,
"loss": 0.0882,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.1379007390242921,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.1091,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.1346544575054976,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.0899,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.12564262759269473,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1234,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.13342251331119642,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.098,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.1277129525873884,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1145,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.12794489080995766,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1132,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.13123291313736804,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1038,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.12836713695431082,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1005,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.1255667844785128,
"learning_rate": 9.597378493454013e-07,
"loss": 0.0892,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.12746453827380147,
"learning_rate": 9.425051342339353e-07,
"loss": 0.0748,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.1274673460746527,
"learning_rate": 9.254248023435864e-07,
"loss": 0.0934,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.12871987784980626,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1105,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.12897459471862605,
"learning_rate": 8.917218333021993e-07,
"loss": 0.1071,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.14955653936252045,
"learning_rate": 8.750994656392109e-07,
"loss": 0.0993,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.13780239926673613,
"learning_rate": 8.58630020171265e-07,
"loss": 0.104,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.13759052862903418,
"learning_rate": 8.423136285876099e-07,
"loss": 0.105,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.13123759501157634,
"learning_rate": 8.26150421353682e-07,
"loss": 0.107,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.13271238937783847,
"learning_rate": 8.101405277100549e-07,
"loss": 0.105,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.12404511657005206,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1157,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.14315929069628747,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1079,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.1351184053242749,
"learning_rate": 7.630320023322024e-07,
"loss": 0.0979,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.1383933756717437,
"learning_rate": 7.476366309223903e-07,
"loss": 0.0978,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.13633461931422755,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1019,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.1372381389706038,
"learning_rate": 7.173078341262618e-07,
"loss": 0.0906,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.1348383448455062,
"learning_rate": 7.023746512481989e-07,
"loss": 0.0946,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.12950017205620304,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1035,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.12422854839413518,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1091,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.13563270755990434,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1344,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.12513437975448358,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1131,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.13239463834836157,
"learning_rate": 6.300256312324471e-07,
"loss": 0.0999,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.123056343223172,
"learning_rate": 6.160200160066776e-07,
"loss": 0.0824,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.127188188311107,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1085,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.13825073686381087,
"learning_rate": 5.884738776666044e-07,
"loss": 0.0977,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.12804404673634381,
"learning_rate": 5.749335748104878e-07,
"loss": 0.0753,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.13043258919072492,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1064,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.1301262985162384,
"learning_rate": 5.483190430647289e-07,
"loss": 0.0893,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.1404824788699619,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1064,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.12445379609013366,
"learning_rate": 5.22326650565701e-07,
"loss": 0.0959,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.13270009917755843,
"learning_rate": 5.095640171043048e-07,
"loss": 0.0895,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.13289186906131065,
"learning_rate": 4.969572286496594e-07,
"loss": 0.0893,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.13300822175068697,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1176,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.12953213745834957,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.0864,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.13189214189157483,
"learning_rate": 4.600729351261857e-07,
"loss": 0.0837,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.14083680630043297,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1264,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.1306342622353077,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.0913,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.12978668244510944,
"learning_rate": 4.245948007320588e-07,
"loss": 0.092,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.12264551081482267,
"learning_rate": 4.130816799401971e-07,
"loss": 0.0831,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.1240517005407744,
"learning_rate": 4.017251756252827e-07,
"loss": 0.0951,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.1392599854223085,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1156,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.1261632505802739,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.0798,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.12467575734654572,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1216,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.13090914283005367,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1092,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.12790315892989287,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1228,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.13108094728158248,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1116,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.12038298874660257,
"learning_rate": 3.266223765343224e-07,
"loss": 0.0876,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.13840186688177333,
"learning_rate": 3.16521967718757e-07,
"loss": 0.0887,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.12984692052167313,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.0891,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.12639981557632907,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.0849,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.12938395623193868,
"learning_rate": 2.871653894244597e-07,
"loss": 0.0976,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.12819384021468058,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.097,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.13360290653927623,
"learning_rate": 2.683823233205485e-07,
"loss": 0.0985,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.1276029752019974,
"learning_rate": 2.592274132703287e-07,
"loss": 0.0976,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.12620434840135836,
"learning_rate": 2.502303499116754e-07,
"loss": 0.0864,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.12522291450966055,
"learning_rate": 2.413912051848666e-07,
"loss": 0.0982,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.12429522841027384,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.0838,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.12708358979754256,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.101,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.13029140508307738,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1073,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.12697782859801243,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.0969,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.13412088829496416,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.1052,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.12614883829821508,
"learning_rate": 1.916764975761254e-07,
"loss": 0.104,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.1307378813852029,
"learning_rate": 1.839446914621923e-07,
"loss": 0.0998,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.11934529254382126,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1019,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.1326607294534564,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.0896,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.1287844439650788,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1147,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.13243728519667558,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.0876,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.11911718138927924,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.0832,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.13205760799142105,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.1106,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.1376982348606062,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.0928,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.12113276025555723,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1009,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.1250141692584398,
"learning_rate": 1.214956261808542e-07,
"loss": 0.102,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.1296829487931315,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1204,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.13756713327985962,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.0776,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.14492758164269898,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.0978,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.1395454013921504,
"learning_rate": 9.787081204466253e-08,
"loss": 0.11,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.12153890413903348,
"learning_rate": 9.236233586200627e-08,
"loss": 0.0869,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.13267223014821775,
"learning_rate": 8.70130406196279e-08,
"loss": 0.0925,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.13527080503978062,
"learning_rate": 8.182296909035047e-08,
"loss": 0.102,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.12857833616591316,
"learning_rate": 7.679216277384439e-08,
"loss": 0.0939,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.12552327153593512,
"learning_rate": 7.19206618963053e-08,
"loss": 0.0898,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.13816246028270787,
"learning_rate": 6.720850541012347e-08,
"loss": 0.0992,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.12809301843572066,
"learning_rate": 6.265573099357714e-08,
"loss": 0.0905,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.12413060048069317,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1243,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.135849155916507,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1161,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.12502641342309098,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.0851,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.13127818286437162,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.0964,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.13201427815453273,
"learning_rate": 4.228381947004812e-08,
"loss": 0.076,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.13519790883646396,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1002,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.12467560197526961,
"learning_rate": 3.525190530455636e-08,
"loss": 0.0951,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.13723340918154986,
"learning_rate": 3.197539087475887e-08,
"loss": 0.0836,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.13208223439525235,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.104,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.12565085292348271,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.0944,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.12148675003069724,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.0931,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.14273159187417658,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.0978,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.12047557635512587,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1079,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.13416475911881645,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.0988,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.13714396587569072,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.0917,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.13173412458852785,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.0787,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.13783335620957146,
"learning_rate": 9.67435416023843e-09,
"loss": 0.0981,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.13186973175539024,
"learning_rate": 7.995445897566178e-09,
"loss": 0.0946,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.13201992060348475,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1005,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.12542989932955045,
"learning_rate": 5.11720812557348e-09,
"loss": 0.0903,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.1293532233086812,
"learning_rate": 3.917901630563492e-09,
"loss": 0.0944,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.12151398514881334,
"learning_rate": 2.878483275170485e-09,
"loss": 0.0969,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.1356334554223748,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1257,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.12573149478947893,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.0993,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.12688019594069153,
"learning_rate": 7.196337656112384e-10,
"loss": 0.0973,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.1208179277671015,
"learning_rate": 3.198382947511292e-10,
"loss": 0.0921,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.12635154317205535,
"learning_rate": 7.995973352770137e-11,
"loss": 0.0847,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.12149057970301061,
"learning_rate": 0.0,
"loss": 0.0976,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.4793150418439373e+18,
"train_loss": 0.2973083603961265,
"train_runtime": 81591.2685,
"train_samples_per_second": 1.936,
"train_steps_per_second": 0.015
}
],
"logging_steps": 1,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.4793150418439373e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}