9b-102 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
6961c88 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1892,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004228329809725159,
"grad_norm": 4.704381465911865,
"learning_rate": 1.5789473684210527e-07,
"loss": 1.824363112449646,
"step": 2
},
{
"epoch": 0.008456659619450317,
"grad_norm": 3.311551809310913,
"learning_rate": 4.736842105263158e-07,
"loss": 2.2129101753234863,
"step": 4
},
{
"epoch": 0.012684989429175475,
"grad_norm": 14.908106803894043,
"learning_rate": 7.894736842105263e-07,
"loss": 2.206524133682251,
"step": 6
},
{
"epoch": 0.016913319238900635,
"grad_norm": 3.046475648880005,
"learning_rate": 1.1052631578947367e-06,
"loss": 1.9903408288955688,
"step": 8
},
{
"epoch": 0.021141649048625793,
"grad_norm": 28.930889129638672,
"learning_rate": 1.4210526315789475e-06,
"loss": 1.9737353324890137,
"step": 10
},
{
"epoch": 0.02536997885835095,
"grad_norm": 3.491518497467041,
"learning_rate": 1.7368421052631579e-06,
"loss": 1.6203196048736572,
"step": 12
},
{
"epoch": 0.02959830866807611,
"grad_norm": 4.089481830596924,
"learning_rate": 2.0526315789473687e-06,
"loss": 1.5767359733581543,
"step": 14
},
{
"epoch": 0.03382663847780127,
"grad_norm": 3.1909189224243164,
"learning_rate": 2.368421052631579e-06,
"loss": 1.6946314573287964,
"step": 16
},
{
"epoch": 0.03805496828752643,
"grad_norm": 3.0035154819488525,
"learning_rate": 2.6842105263157895e-06,
"loss": 1.7196022272109985,
"step": 18
},
{
"epoch": 0.042283298097251586,
"grad_norm": 21.877323150634766,
"learning_rate": 3e-06,
"loss": 1.3684844970703125,
"step": 20
},
{
"epoch": 0.046511627906976744,
"grad_norm": 9.42421817779541,
"learning_rate": 3.3157894736842107e-06,
"loss": 1.3061422109603882,
"step": 22
},
{
"epoch": 0.0507399577167019,
"grad_norm": 6.530590534210205,
"learning_rate": 3.6315789473684213e-06,
"loss": 1.71840500831604,
"step": 24
},
{
"epoch": 0.05496828752642706,
"grad_norm": 4.663469314575195,
"learning_rate": 3.9473684210526315e-06,
"loss": 1.2755658626556396,
"step": 26
},
{
"epoch": 0.05919661733615222,
"grad_norm": 1.7210355997085571,
"learning_rate": 4.2631578947368425e-06,
"loss": 1.5721380710601807,
"step": 28
},
{
"epoch": 0.06342494714587738,
"grad_norm": 2.700887441635132,
"learning_rate": 4.578947368421053e-06,
"loss": 1.1442614793777466,
"step": 30
},
{
"epoch": 0.06765327695560254,
"grad_norm": 3.250575304031372,
"learning_rate": 4.894736842105264e-06,
"loss": 0.9219419956207275,
"step": 32
},
{
"epoch": 0.07188160676532769,
"grad_norm": 2.205869674682617,
"learning_rate": 5.210526315789474e-06,
"loss": 1.1184953451156616,
"step": 34
},
{
"epoch": 0.07610993657505286,
"grad_norm": 20.262012481689453,
"learning_rate": 5.526315789473684e-06,
"loss": 1.3092234134674072,
"step": 36
},
{
"epoch": 0.080338266384778,
"grad_norm": 10.212276458740234,
"learning_rate": 5.842105263157895e-06,
"loss": 1.2601300477981567,
"step": 38
},
{
"epoch": 0.08456659619450317,
"grad_norm": 17.500431060791016,
"learning_rate": 6.157894736842105e-06,
"loss": 1.170905590057373,
"step": 40
},
{
"epoch": 0.08879492600422834,
"grad_norm": 6.665106773376465,
"learning_rate": 6.473684210526316e-06,
"loss": 1.6788873672485352,
"step": 42
},
{
"epoch": 0.09302325581395349,
"grad_norm": 10.255254745483398,
"learning_rate": 6.7894736842105264e-06,
"loss": 1.5785198211669922,
"step": 44
},
{
"epoch": 0.09725158562367865,
"grad_norm": 2.286841869354248,
"learning_rate": 7.105263157894737e-06,
"loss": 1.2138264179229736,
"step": 46
},
{
"epoch": 0.1014799154334038,
"grad_norm": 2.4699878692626953,
"learning_rate": 7.421052631578948e-06,
"loss": 1.4601056575775146,
"step": 48
},
{
"epoch": 0.10570824524312897,
"grad_norm": 3.7046945095062256,
"learning_rate": 7.73684210526316e-06,
"loss": 1.4360618591308594,
"step": 50
},
{
"epoch": 0.10993657505285412,
"grad_norm": 13.804616928100586,
"learning_rate": 8.052631578947368e-06,
"loss": 0.6696046590805054,
"step": 52
},
{
"epoch": 0.11416490486257928,
"grad_norm": 2.6909313201904297,
"learning_rate": 8.368421052631578e-06,
"loss": 1.0173393487930298,
"step": 54
},
{
"epoch": 0.11839323467230443,
"grad_norm": 1.630982518196106,
"learning_rate": 8.68421052631579e-06,
"loss": 1.3301122188568115,
"step": 56
},
{
"epoch": 0.1226215644820296,
"grad_norm": 2.1182713508605957,
"learning_rate": 9e-06,
"loss": 1.3487412929534912,
"step": 58
},
{
"epoch": 0.12684989429175475,
"grad_norm": 2.5053207874298096,
"learning_rate": 9.315789473684212e-06,
"loss": 1.3728868961334229,
"step": 60
},
{
"epoch": 0.13107822410147993,
"grad_norm": 9.146308898925781,
"learning_rate": 9.63157894736842e-06,
"loss": 0.7595624923706055,
"step": 62
},
{
"epoch": 0.13530655391120508,
"grad_norm": 2.828664541244507,
"learning_rate": 9.94736842105263e-06,
"loss": 1.0184533596038818,
"step": 64
},
{
"epoch": 0.13953488372093023,
"grad_norm": 4.367115497589111,
"learning_rate": 1.0263157894736843e-05,
"loss": 1.3130357265472412,
"step": 66
},
{
"epoch": 0.14376321353065538,
"grad_norm": 3.0600650310516357,
"learning_rate": 1.0578947368421053e-05,
"loss": 1.34865140914917,
"step": 68
},
{
"epoch": 0.14799154334038056,
"grad_norm": 2.592481851577759,
"learning_rate": 1.0894736842105265e-05,
"loss": 1.338028907775879,
"step": 70
},
{
"epoch": 0.1522198731501057,
"grad_norm": 4.736515998840332,
"learning_rate": 1.1210526315789473e-05,
"loss": 1.3265666961669922,
"step": 72
},
{
"epoch": 0.15644820295983086,
"grad_norm": 2.3097026348114014,
"learning_rate": 1.1526315789473683e-05,
"loss": 0.9411308765411377,
"step": 74
},
{
"epoch": 0.160676532769556,
"grad_norm": 4.697860240936279,
"learning_rate": 1.1842105263157895e-05,
"loss": 0.8470487594604492,
"step": 76
},
{
"epoch": 0.1649048625792812,
"grad_norm": 4.811895370483398,
"learning_rate": 1.2157894736842105e-05,
"loss": 1.2756433486938477,
"step": 78
},
{
"epoch": 0.16913319238900634,
"grad_norm": 7.41347599029541,
"learning_rate": 1.2473684210526317e-05,
"loss": 1.0685019493103027,
"step": 80
},
{
"epoch": 0.1733615221987315,
"grad_norm": 1.8728761672973633,
"learning_rate": 1.2789473684210526e-05,
"loss": 1.08561372756958,
"step": 82
},
{
"epoch": 0.17758985200845667,
"grad_norm": 1.7113884687423706,
"learning_rate": 1.3105263157894738e-05,
"loss": 0.847097635269165,
"step": 84
},
{
"epoch": 0.18181818181818182,
"grad_norm": 3.67093563079834,
"learning_rate": 1.3421052631578948e-05,
"loss": 1.6951191425323486,
"step": 86
},
{
"epoch": 0.18604651162790697,
"grad_norm": 3.370368003845215,
"learning_rate": 1.3736842105263158e-05,
"loss": 1.3011367321014404,
"step": 88
},
{
"epoch": 0.19027484143763213,
"grad_norm": 5.550237655639648,
"learning_rate": 1.405263157894737e-05,
"loss": 0.978367805480957,
"step": 90
},
{
"epoch": 0.1945031712473573,
"grad_norm": 1.7489956617355347,
"learning_rate": 1.4368421052631578e-05,
"loss": 1.3048770427703857,
"step": 92
},
{
"epoch": 0.19873150105708245,
"grad_norm": 2.4277842044830322,
"learning_rate": 1.468421052631579e-05,
"loss": 0.9812408685684204,
"step": 94
},
{
"epoch": 0.2029598308668076,
"grad_norm": 2.454376697540283,
"learning_rate": 1.5e-05,
"loss": 1.2909517288208008,
"step": 96
},
{
"epoch": 0.20718816067653276,
"grad_norm": 2.605318307876587,
"learning_rate": 1.499995873926907e-05,
"loss": 1.3840426206588745,
"step": 98
},
{
"epoch": 0.21141649048625794,
"grad_norm": 1.504913091659546,
"learning_rate": 1.49998349575807e-05,
"loss": 1.319006323814392,
"step": 100
},
{
"epoch": 0.2156448202959831,
"grad_norm": 5.363708019256592,
"learning_rate": 1.499962865644818e-05,
"loss": 1.2458685636520386,
"step": 102
},
{
"epoch": 0.21987315010570824,
"grad_norm": 3.041691780090332,
"learning_rate": 1.4999339838393618e-05,
"loss": 1.3088092803955078,
"step": 104
},
{
"epoch": 0.22410147991543342,
"grad_norm": 2.801055669784546,
"learning_rate": 1.4998968506947934e-05,
"loss": 1.2939674854278564,
"step": 106
},
{
"epoch": 0.22832980972515857,
"grad_norm": 1.6846338510513306,
"learning_rate": 1.4998514666650803e-05,
"loss": 1.0377792119979858,
"step": 108
},
{
"epoch": 0.23255813953488372,
"grad_norm": 1.949150800704956,
"learning_rate": 1.4997978323050603e-05,
"loss": 1.2812877893447876,
"step": 110
},
{
"epoch": 0.23678646934460887,
"grad_norm": 5.502281188964844,
"learning_rate": 1.4997359482704357e-05,
"loss": 1.3570351600646973,
"step": 112
},
{
"epoch": 0.24101479915433405,
"grad_norm": 2.349268913269043,
"learning_rate": 1.4996658153177625e-05,
"loss": 1.6063910722732544,
"step": 114
},
{
"epoch": 0.2452431289640592,
"grad_norm": 3.2021586894989014,
"learning_rate": 1.4995874343044454e-05,
"loss": 1.3220144510269165,
"step": 116
},
{
"epoch": 0.24947145877378435,
"grad_norm": 2.5014760494232178,
"learning_rate": 1.4995008061887229e-05,
"loss": 1.3478097915649414,
"step": 118
},
{
"epoch": 0.2536997885835095,
"grad_norm": 6.440486907958984,
"learning_rate": 1.4994059320296592e-05,
"loss": 0.8495758771896362,
"step": 120
},
{
"epoch": 0.25792811839323465,
"grad_norm": 1.8407394886016846,
"learning_rate": 1.4993028129871286e-05,
"loss": 1.3098301887512207,
"step": 122
},
{
"epoch": 0.26215644820295986,
"grad_norm": 1.789825201034546,
"learning_rate": 1.499191450321803e-05,
"loss": 1.1505930423736572,
"step": 124
},
{
"epoch": 0.266384778012685,
"grad_norm": 2.5154478549957275,
"learning_rate": 1.4990718453951355e-05,
"loss": 1.3440117835998535,
"step": 126
},
{
"epoch": 0.27061310782241016,
"grad_norm": 5.759194850921631,
"learning_rate": 1.4989439996693445e-05,
"loss": 1.135366439819336,
"step": 128
},
{
"epoch": 0.2748414376321353,
"grad_norm": 1.8347864151000977,
"learning_rate": 1.4988079147073955e-05,
"loss": 0.9918815493583679,
"step": 130
},
{
"epoch": 0.27906976744186046,
"grad_norm": 2.3062212467193604,
"learning_rate": 1.4986635921729819e-05,
"loss": 1.2845799922943115,
"step": 132
},
{
"epoch": 0.2832980972515856,
"grad_norm": 2.6987996101379395,
"learning_rate": 1.4985110338305041e-05,
"loss": 0.9715274572372437,
"step": 134
},
{
"epoch": 0.28752642706131076,
"grad_norm": 5.780081272125244,
"learning_rate": 1.49835024154505e-05,
"loss": 1.1196620464324951,
"step": 136
},
{
"epoch": 0.2917547568710359,
"grad_norm": 2.331085681915283,
"learning_rate": 1.4981812172823692e-05,
"loss": 1.3799099922180176,
"step": 138
},
{
"epoch": 0.2959830866807611,
"grad_norm": 2.2767205238342285,
"learning_rate": 1.4980039631088512e-05,
"loss": 0.9513759613037109,
"step": 140
},
{
"epoch": 0.30021141649048627,
"grad_norm": 1.6002323627471924,
"learning_rate": 1.4978184811914998e-05,
"loss": 1.5639171600341797,
"step": 142
},
{
"epoch": 0.3044397463002114,
"grad_norm": 4.709345817565918,
"learning_rate": 1.497624773797906e-05,
"loss": 1.0970286130905151,
"step": 144
},
{
"epoch": 0.3086680761099366,
"grad_norm": 2.453108310699463,
"learning_rate": 1.4974228432962202e-05,
"loss": 0.6733556985855103,
"step": 146
},
{
"epoch": 0.3128964059196617,
"grad_norm": 3.1568665504455566,
"learning_rate": 1.4972126921551245e-05,
"loss": 1.2531988620758057,
"step": 148
},
{
"epoch": 0.3171247357293869,
"grad_norm": 1.1754913330078125,
"learning_rate": 1.4969943229438003e-05,
"loss": 1.2762296199798584,
"step": 150
},
{
"epoch": 0.321353065539112,
"grad_norm": 1.3160622119903564,
"learning_rate": 1.4967677383318993e-05,
"loss": 1.192008376121521,
"step": 152
},
{
"epoch": 0.32558139534883723,
"grad_norm": 5.29488468170166,
"learning_rate": 1.4965329410895095e-05,
"loss": 0.7570050954818726,
"step": 154
},
{
"epoch": 0.3298097251585624,
"grad_norm": 2.46903133392334,
"learning_rate": 1.496289934087121e-05,
"loss": 1.2650794982910156,
"step": 156
},
{
"epoch": 0.33403805496828753,
"grad_norm": 2.032607078552246,
"learning_rate": 1.4960387202955927e-05,
"loss": 1.2755355834960938,
"step": 158
},
{
"epoch": 0.3382663847780127,
"grad_norm": 1.7639720439910889,
"learning_rate": 1.4957793027861142e-05,
"loss": 1.2587889432907104,
"step": 160
},
{
"epoch": 0.34249471458773784,
"grad_norm": 2.4041712284088135,
"learning_rate": 1.4955116847301678e-05,
"loss": 1.0974364280700684,
"step": 162
},
{
"epoch": 0.346723044397463,
"grad_norm": 1.8345462083816528,
"learning_rate": 1.4952358693994928e-05,
"loss": 0.979657769203186,
"step": 164
},
{
"epoch": 0.35095137420718814,
"grad_norm": 1.8596234321594238,
"learning_rate": 1.494951860166042e-05,
"loss": 0.8932428359985352,
"step": 166
},
{
"epoch": 0.35517970401691334,
"grad_norm": 1.578384280204773,
"learning_rate": 1.4946596605019421e-05,
"loss": 1.2057310342788696,
"step": 168
},
{
"epoch": 0.3594080338266385,
"grad_norm": 1.5913245677947998,
"learning_rate": 1.494359273979452e-05,
"loss": 1.2144837379455566,
"step": 170
},
{
"epoch": 0.36363636363636365,
"grad_norm": 1.99128258228302,
"learning_rate": 1.4940507042709171e-05,
"loss": 1.056563377380371,
"step": 172
},
{
"epoch": 0.3678646934460888,
"grad_norm": 2.5846242904663086,
"learning_rate": 1.4937339551487265e-05,
"loss": 1.352506399154663,
"step": 174
},
{
"epoch": 0.37209302325581395,
"grad_norm": 1.8220958709716797,
"learning_rate": 1.493409030485265e-05,
"loss": 1.0490646362304688,
"step": 176
},
{
"epoch": 0.3763213530655391,
"grad_norm": 11.952075958251953,
"learning_rate": 1.4930759342528674e-05,
"loss": 0.9132463932037354,
"step": 178
},
{
"epoch": 0.38054968287526425,
"grad_norm": 1.2839536666870117,
"learning_rate": 1.492734670523769e-05,
"loss": 1.287312388420105,
"step": 180
},
{
"epoch": 0.38477801268498946,
"grad_norm": 1.7404998540878296,
"learning_rate": 1.4923852434700561e-05,
"loss": 1.520090103149414,
"step": 182
},
{
"epoch": 0.3890063424947146,
"grad_norm": 1.5424197912216187,
"learning_rate": 1.4920276573636144e-05,
"loss": 1.3454771041870117,
"step": 184
},
{
"epoch": 0.39323467230443976,
"grad_norm": 2.3417012691497803,
"learning_rate": 1.4916619165760783e-05,
"loss": 1.1211519241333008,
"step": 186
},
{
"epoch": 0.3974630021141649,
"grad_norm": 2.4767096042633057,
"learning_rate": 1.491288025578775e-05,
"loss": 0.9932225942611694,
"step": 188
},
{
"epoch": 0.40169133192389006,
"grad_norm": 1.8516297340393066,
"learning_rate": 1.4909059889426732e-05,
"loss": 1.0381616353988647,
"step": 190
},
{
"epoch": 0.4059196617336152,
"grad_norm": 2.8047099113464355,
"learning_rate": 1.4905158113383237e-05,
"loss": 1.3036831617355347,
"step": 192
},
{
"epoch": 0.41014799154334036,
"grad_norm": 1.1113916635513306,
"learning_rate": 1.4901174975358043e-05,
"loss": 1.022596836090088,
"step": 194
},
{
"epoch": 0.4143763213530655,
"grad_norm": 1.854797601699829,
"learning_rate": 1.4897110524046617e-05,
"loss": 1.2877132892608643,
"step": 196
},
{
"epoch": 0.4186046511627907,
"grad_norm": 2.30233097076416,
"learning_rate": 1.4892964809138505e-05,
"loss": 1.0897512435913086,
"step": 198
},
{
"epoch": 0.42283298097251587,
"grad_norm": 2.2684969902038574,
"learning_rate": 1.4888737881316741e-05,
"loss": 1.1889673471450806,
"step": 200
},
{
"epoch": 0.427061310782241,
"grad_norm": 2.5113182067871094,
"learning_rate": 1.4884429792257216e-05,
"loss": 0.604529857635498,
"step": 202
},
{
"epoch": 0.4312896405919662,
"grad_norm": 1.313503623008728,
"learning_rate": 1.4880040594628045e-05,
"loss": 1.231643557548523,
"step": 204
},
{
"epoch": 0.4355179704016913,
"grad_norm": 3.4917216300964355,
"learning_rate": 1.4875570342088935e-05,
"loss": 1.3376781940460205,
"step": 206
},
{
"epoch": 0.4397463002114165,
"grad_norm": 3.937040090560913,
"learning_rate": 1.487101908929052e-05,
"loss": 0.8179441690444946,
"step": 208
},
{
"epoch": 0.4439746300211416,
"grad_norm": 1.756445050239563,
"learning_rate": 1.4866386891873693e-05,
"loss": 1.3150757551193237,
"step": 210
},
{
"epoch": 0.44820295983086683,
"grad_norm": 1.9177740812301636,
"learning_rate": 1.486167380646893e-05,
"loss": 1.2286052703857422,
"step": 212
},
{
"epoch": 0.452431289640592,
"grad_norm": 2.874783992767334,
"learning_rate": 1.4856879890695592e-05,
"loss": 1.4362702369689941,
"step": 214
},
{
"epoch": 0.45665961945031713,
"grad_norm": 2.0687010288238525,
"learning_rate": 1.4852005203161227e-05,
"loss": 0.6649367809295654,
"step": 216
},
{
"epoch": 0.4608879492600423,
"grad_norm": 2.8608791828155518,
"learning_rate": 1.4847049803460847e-05,
"loss": 1.0288950204849243,
"step": 218
},
{
"epoch": 0.46511627906976744,
"grad_norm": 1.2039422988891602,
"learning_rate": 1.4842013752176204e-05,
"loss": 1.289539098739624,
"step": 220
},
{
"epoch": 0.4693446088794926,
"grad_norm": 3.1522738933563232,
"learning_rate": 1.4836897110875048e-05,
"loss": 1.242141842842102,
"step": 222
},
{
"epoch": 0.47357293868921774,
"grad_norm": 1.4230495691299438,
"learning_rate": 1.4831699942110378e-05,
"loss": 1.367640733718872,
"step": 224
},
{
"epoch": 0.47780126849894294,
"grad_norm": 1.2761074304580688,
"learning_rate": 1.4826422309419665e-05,
"loss": 1.2589974403381348,
"step": 226
},
{
"epoch": 0.4820295983086681,
"grad_norm": 3.499668836593628,
"learning_rate": 1.4821064277324092e-05,
"loss": 1.2439829111099243,
"step": 228
},
{
"epoch": 0.48625792811839325,
"grad_norm": 1.3432852029800415,
"learning_rate": 1.4815625911327753e-05,
"loss": 0.9437692165374756,
"step": 230
},
{
"epoch": 0.4904862579281184,
"grad_norm": 2.4091720581054688,
"learning_rate": 1.4810107277916858e-05,
"loss": 1.2602202892303467,
"step": 232
},
{
"epoch": 0.49471458773784355,
"grad_norm": 3.980276584625244,
"learning_rate": 1.4804508444558917e-05,
"loss": 0.943488597869873,
"step": 234
},
{
"epoch": 0.4989429175475687,
"grad_norm": 3.813063621520996,
"learning_rate": 1.479882947970192e-05,
"loss": 0.8944742679595947,
"step": 236
},
{
"epoch": 0.5031712473572939,
"grad_norm": 2.801894187927246,
"learning_rate": 1.4793070452773488e-05,
"loss": 0.8451627492904663,
"step": 238
},
{
"epoch": 0.507399577167019,
"grad_norm": 2.424893856048584,
"learning_rate": 1.4787231434180044e-05,
"loss": 0.9260126948356628,
"step": 240
},
{
"epoch": 0.5116279069767442,
"grad_norm": 2.86238956451416,
"learning_rate": 1.4781312495305934e-05,
"loss": 1.32850980758667,
"step": 242
},
{
"epoch": 0.5158562367864693,
"grad_norm": 5.032546043395996,
"learning_rate": 1.4775313708512563e-05,
"loss": 1.3423174619674683,
"step": 244
},
{
"epoch": 0.5200845665961945,
"grad_norm": 4.0864152908325195,
"learning_rate": 1.4769235147137514e-05,
"loss": 1.2995855808258057,
"step": 246
},
{
"epoch": 0.5243128964059197,
"grad_norm": 2.7449331283569336,
"learning_rate": 1.4763076885493637e-05,
"loss": 1.2420015335083008,
"step": 248
},
{
"epoch": 0.5285412262156448,
"grad_norm": 4.043562412261963,
"learning_rate": 1.4756838998868158e-05,
"loss": 1.188884973526001,
"step": 250
},
{
"epoch": 0.53276955602537,
"grad_norm": 4.929879188537598,
"learning_rate": 1.4750521563521741e-05,
"loss": 1.1663317680358887,
"step": 252
},
{
"epoch": 0.5369978858350951,
"grad_norm": 4.050250053405762,
"learning_rate": 1.474412465668758e-05,
"loss": 1.7435861825942993,
"step": 254
},
{
"epoch": 0.5412262156448203,
"grad_norm": 1.9945145845413208,
"learning_rate": 1.4737648356570426e-05,
"loss": 1.2141636610031128,
"step": 256
},
{
"epoch": 0.5454545454545454,
"grad_norm": 2.514636278152466,
"learning_rate": 1.4731092742345652e-05,
"loss": 1.3950912952423096,
"step": 258
},
{
"epoch": 0.5496828752642706,
"grad_norm": 2.134345293045044,
"learning_rate": 1.472445789415827e-05,
"loss": 0.9992928504943848,
"step": 260
},
{
"epoch": 0.5539112050739958,
"grad_norm": 1.3497594594955444,
"learning_rate": 1.4717743893121974e-05,
"loss": 1.1979846954345703,
"step": 262
},
{
"epoch": 0.5581395348837209,
"grad_norm": 3.4390580654144287,
"learning_rate": 1.4710950821318115e-05,
"loss": 1.0499022006988525,
"step": 264
},
{
"epoch": 0.5623678646934461,
"grad_norm": 4.900990962982178,
"learning_rate": 1.4704078761794731e-05,
"loss": 1.496317744255066,
"step": 266
},
{
"epoch": 0.5665961945031712,
"grad_norm": 5.405950546264648,
"learning_rate": 1.4697127798565508e-05,
"loss": 0.5493452548980713,
"step": 268
},
{
"epoch": 0.5708245243128964,
"grad_norm": 3.4618313312530518,
"learning_rate": 1.4690098016608766e-05,
"loss": 1.029736876487732,
"step": 270
},
{
"epoch": 0.5750528541226215,
"grad_norm": 3.426192283630371,
"learning_rate": 1.4682989501866411e-05,
"loss": 1.1797683238983154,
"step": 272
},
{
"epoch": 0.5792811839323467,
"grad_norm": 2.367126226425171,
"learning_rate": 1.4675802341242893e-05,
"loss": 0.9646555185317993,
"step": 274
},
{
"epoch": 0.5835095137420718,
"grad_norm": 15.31234073638916,
"learning_rate": 1.466853662260414e-05,
"loss": 0.9308096766471863,
"step": 276
},
{
"epoch": 0.587737843551797,
"grad_norm": 0.6522766947746277,
"learning_rate": 1.4661192434776478e-05,
"loss": 0.9009642004966736,
"step": 278
},
{
"epoch": 0.5919661733615222,
"grad_norm": 1.5370508432388306,
"learning_rate": 1.4653769867545556e-05,
"loss": 1.2289916276931763,
"step": 280
},
{
"epoch": 0.5961945031712473,
"grad_norm": 2.7394051551818848,
"learning_rate": 1.4646269011655239e-05,
"loss": 1.4691234827041626,
"step": 282
},
{
"epoch": 0.6004228329809725,
"grad_norm": 0.8963581323623657,
"learning_rate": 1.4638689958806506e-05,
"loss": 0.9581058025360107,
"step": 284
},
{
"epoch": 0.6046511627906976,
"grad_norm": 8.027399063110352,
"learning_rate": 1.4631032801656318e-05,
"loss": 1.311490774154663,
"step": 286
},
{
"epoch": 0.6088794926004228,
"grad_norm": 2.2466981410980225,
"learning_rate": 1.4623297633816507e-05,
"loss": 0.4543798267841339,
"step": 288
},
{
"epoch": 0.6131078224101479,
"grad_norm": 1.17365562915802,
"learning_rate": 1.4615484549852597e-05,
"loss": 0.9570913314819336,
"step": 290
},
{
"epoch": 0.6173361522198731,
"grad_norm": 1.8631973266601562,
"learning_rate": 1.4607593645282686e-05,
"loss": 0.8872200846672058,
"step": 292
},
{
"epoch": 0.6215644820295984,
"grad_norm": 3.455590009689331,
"learning_rate": 1.4599625016576245e-05,
"loss": 0.7154805660247803,
"step": 294
},
{
"epoch": 0.6257928118393234,
"grad_norm": 3.2526419162750244,
"learning_rate": 1.4591578761152972e-05,
"loss": 0.74409019947052,
"step": 296
},
{
"epoch": 0.6300211416490487,
"grad_norm": 1.9088728427886963,
"learning_rate": 1.4583454977381563e-05,
"loss": 1.2749909162521362,
"step": 298
},
{
"epoch": 0.6342494714587738,
"grad_norm": 1.7068229913711548,
"learning_rate": 1.4575253764578541e-05,
"loss": 1.2347133159637451,
"step": 300
},
{
"epoch": 0.638477801268499,
"grad_norm": 3.112821340560913,
"learning_rate": 1.4566975223007029e-05,
"loss": 0.7435523271560669,
"step": 302
},
{
"epoch": 0.642706131078224,
"grad_norm": 5.360762596130371,
"learning_rate": 1.4558619453875524e-05,
"loss": 0.8920886516571045,
"step": 304
},
{
"epoch": 0.6469344608879493,
"grad_norm": 1.2986818552017212,
"learning_rate": 1.4550186559336655e-05,
"loss": 0.9719116687774658,
"step": 306
},
{
"epoch": 0.6511627906976745,
"grad_norm": 1.9221904277801514,
"learning_rate": 1.4541676642485954e-05,
"loss": 1.2280982732772827,
"step": 308
},
{
"epoch": 0.6553911205073996,
"grad_norm": 3.9950499534606934,
"learning_rate": 1.4533089807360568e-05,
"loss": 1.2017021179199219,
"step": 310
},
{
"epoch": 0.6596194503171248,
"grad_norm": 2.393136501312256,
"learning_rate": 1.452442615893801e-05,
"loss": 1.2053108215332031,
"step": 312
},
{
"epoch": 0.6638477801268499,
"grad_norm": 2.202565908432007,
"learning_rate": 1.4515685803134858e-05,
"loss": 1.2293308973312378,
"step": 314
},
{
"epoch": 0.6680761099365751,
"grad_norm": 1.489235520362854,
"learning_rate": 1.4506868846805476e-05,
"loss": 1.0899698734283447,
"step": 316
},
{
"epoch": 0.6723044397463002,
"grad_norm": 3.6901493072509766,
"learning_rate": 1.4497975397740696e-05,
"loss": 0.7973343133926392,
"step": 318
},
{
"epoch": 0.6765327695560254,
"grad_norm": 1.9012552499771118,
"learning_rate": 1.4489005564666508e-05,
"loss": 1.2414870262145996,
"step": 320
},
{
"epoch": 0.6807610993657506,
"grad_norm": 1.2476270198822021,
"learning_rate": 1.4479959457242723e-05,
"loss": 1.3038647174835205,
"step": 322
},
{
"epoch": 0.6849894291754757,
"grad_norm": 3.06807279586792,
"learning_rate": 1.4470837186061642e-05,
"loss": 1.0853257179260254,
"step": 324
},
{
"epoch": 0.6892177589852009,
"grad_norm": 2.1801917552948,
"learning_rate": 1.4461638862646691e-05,
"loss": 1.26482355594635,
"step": 326
},
{
"epoch": 0.693446088794926,
"grad_norm": 3.968803882598877,
"learning_rate": 1.4452364599451069e-05,
"loss": 1.18692946434021,
"step": 328
},
{
"epoch": 0.6976744186046512,
"grad_norm": 4.3209710121154785,
"learning_rate": 1.4443014509856374e-05,
"loss": 0.8941938877105713,
"step": 330
},
{
"epoch": 0.7019027484143763,
"grad_norm": 1.704699158668518,
"learning_rate": 1.4433588708171203e-05,
"loss": 1.2291384935379028,
"step": 332
},
{
"epoch": 0.7061310782241015,
"grad_norm": 4.724952697753906,
"learning_rate": 1.4424087309629773e-05,
"loss": 1.2391948699951172,
"step": 334
},
{
"epoch": 0.7103594080338267,
"grad_norm": 1.2283551692962646,
"learning_rate": 1.4414510430390493e-05,
"loss": 1.223761796951294,
"step": 336
},
{
"epoch": 0.7145877378435518,
"grad_norm": 1.6698932647705078,
"learning_rate": 1.4404858187534565e-05,
"loss": 1.3163425922393799,
"step": 338
},
{
"epoch": 0.718816067653277,
"grad_norm": 4.7933783531188965,
"learning_rate": 1.4395130699064532e-05,
"loss": 1.205067753791809,
"step": 340
},
{
"epoch": 0.7230443974630021,
"grad_norm": 4.818348407745361,
"learning_rate": 1.438532808390285e-05,
"loss": 1.150251865386963,
"step": 342
},
{
"epoch": 0.7272727272727273,
"grad_norm": 3.2410824298858643,
"learning_rate": 1.4375450461890427e-05,
"loss": 0.9574451446533203,
"step": 344
},
{
"epoch": 0.7315010570824524,
"grad_norm": 49.398216247558594,
"learning_rate": 1.4365497953785162e-05,
"loss": 0.847400426864624,
"step": 346
},
{
"epoch": 0.7357293868921776,
"grad_norm": 2.7538836002349854,
"learning_rate": 1.4355470681260463e-05,
"loss": 0.9942716360092163,
"step": 348
},
{
"epoch": 0.7399577167019028,
"grad_norm": 1.8401576280593872,
"learning_rate": 1.4345368766903767e-05,
"loss": 1.3236424922943115,
"step": 350
},
{
"epoch": 0.7441860465116279,
"grad_norm": 2.550213098526001,
"learning_rate": 1.4335192334215031e-05,
"loss": 1.1727192401885986,
"step": 352
},
{
"epoch": 0.7484143763213531,
"grad_norm": 8.171126365661621,
"learning_rate": 1.432494150760524e-05,
"loss": 0.7158507704734802,
"step": 354
},
{
"epoch": 0.7526427061310782,
"grad_norm": 3.501011848449707,
"learning_rate": 1.4314616412394867e-05,
"loss": 1.0459976196289062,
"step": 356
},
{
"epoch": 0.7568710359408034,
"grad_norm": 1.9981505870819092,
"learning_rate": 1.4304217174812343e-05,
"loss": 1.2196413278579712,
"step": 358
},
{
"epoch": 0.7610993657505285,
"grad_norm": 2.736032485961914,
"learning_rate": 1.429374392199253e-05,
"loss": 0.7594531774520874,
"step": 360
},
{
"epoch": 0.7653276955602537,
"grad_norm": 2.3446195125579834,
"learning_rate": 1.4283196781975148e-05,
"loss": 1.1010284423828125,
"step": 362
},
{
"epoch": 0.7695560253699789,
"grad_norm": 3.869166374206543,
"learning_rate": 1.4272575883703227e-05,
"loss": 1.046842336654663,
"step": 364
},
{
"epoch": 0.773784355179704,
"grad_norm": 1.9482706785202026,
"learning_rate": 1.4261881357021509e-05,
"loss": 1.2721140384674072,
"step": 366
},
{
"epoch": 0.7780126849894292,
"grad_norm": 1.73712158203125,
"learning_rate": 1.4251113332674884e-05,
"loss": 1.2102105617523193,
"step": 368
},
{
"epoch": 0.7822410147991543,
"grad_norm": 4.2684326171875,
"learning_rate": 1.4240271942306772e-05,
"loss": 0.8783853054046631,
"step": 370
},
{
"epoch": 0.7864693446088795,
"grad_norm": 2.6612749099731445,
"learning_rate": 1.4229357318457529e-05,
"loss": 1.1931439638137817,
"step": 372
},
{
"epoch": 0.7906976744186046,
"grad_norm": 2.393467903137207,
"learning_rate": 1.4218369594562817e-05,
"loss": 1.2113313674926758,
"step": 374
},
{
"epoch": 0.7949260042283298,
"grad_norm": 2.42254638671875,
"learning_rate": 1.420730890495197e-05,
"loss": 1.297842025756836,
"step": 376
},
{
"epoch": 0.7991543340380549,
"grad_norm": 2.7572309970855713,
"learning_rate": 1.419617538484637e-05,
"loss": 0.9628263711929321,
"step": 378
},
{
"epoch": 0.8033826638477801,
"grad_norm": 5.698837757110596,
"learning_rate": 1.4184969170357766e-05,
"loss": 1.3083126544952393,
"step": 380
},
{
"epoch": 0.8076109936575053,
"grad_norm": 2.313297748565674,
"learning_rate": 1.4173690398486635e-05,
"loss": 1.1714601516723633,
"step": 382
},
{
"epoch": 0.8118393234672304,
"grad_norm": 2.152954339981079,
"learning_rate": 1.4162339207120496e-05,
"loss": 0.9850152730941772,
"step": 384
},
{
"epoch": 0.8160676532769556,
"grad_norm": 7.026889801025391,
"learning_rate": 1.4150915735032221e-05,
"loss": 0.9662938117980957,
"step": 386
},
{
"epoch": 0.8202959830866807,
"grad_norm": 3.9815030097961426,
"learning_rate": 1.4139420121878348e-05,
"loss": 1.482654333114624,
"step": 388
},
{
"epoch": 0.8245243128964059,
"grad_norm": 4.242010116577148,
"learning_rate": 1.4127852508197365e-05,
"loss": 1.1514601707458496,
"step": 390
},
{
"epoch": 0.828752642706131,
"grad_norm": 6.246471405029297,
"learning_rate": 1.4116213035408e-05,
"loss": 0.9328240156173706,
"step": 392
},
{
"epoch": 0.8329809725158562,
"grad_norm": 2.0024123191833496,
"learning_rate": 1.410450184580748e-05,
"loss": 1.2407183647155762,
"step": 394
},
{
"epoch": 0.8372093023255814,
"grad_norm": 0.5145191550254822,
"learning_rate": 1.4092719082569803e-05,
"loss": 0.7694379091262817,
"step": 396
},
{
"epoch": 0.8414376321353065,
"grad_norm": 1.8720133304595947,
"learning_rate": 1.4080864889743983e-05,
"loss": 0.7048227787017822,
"step": 398
},
{
"epoch": 0.8456659619450317,
"grad_norm": 9.628911972045898,
"learning_rate": 1.4068939412252294e-05,
"loss": 1.1295552253723145,
"step": 400
},
{
"epoch": 0.8498942917547568,
"grad_norm": 3.1187636852264404,
"learning_rate": 1.4056942795888487e-05,
"loss": 0.9754788875579834,
"step": 402
},
{
"epoch": 0.854122621564482,
"grad_norm": 1.3408007621765137,
"learning_rate": 1.4044875187316015e-05,
"loss": 0.8358703851699829,
"step": 404
},
{
"epoch": 0.8583509513742071,
"grad_norm": 3.0311524868011475,
"learning_rate": 1.4032736734066244e-05,
"loss": 0.9890801906585693,
"step": 406
},
{
"epoch": 0.8625792811839323,
"grad_norm": 1.2130165100097656,
"learning_rate": 1.402052758453664e-05,
"loss": 0.9809809923171997,
"step": 408
},
{
"epoch": 0.8668076109936576,
"grad_norm": 3.2059152126312256,
"learning_rate": 1.400824788798896e-05,
"loss": 0.8744783401489258,
"step": 410
},
{
"epoch": 0.8710359408033826,
"grad_norm": 2.250013589859009,
"learning_rate": 1.399589779454743e-05,
"loss": 1.1907601356506348,
"step": 412
},
{
"epoch": 0.8752642706131079,
"grad_norm": 3.5985400676727295,
"learning_rate": 1.3983477455196905e-05,
"loss": 1.0356981754302979,
"step": 414
},
{
"epoch": 0.879492600422833,
"grad_norm": 4.509793758392334,
"learning_rate": 1.3970987021781022e-05,
"loss": 1.129422664642334,
"step": 416
},
{
"epoch": 0.8837209302325582,
"grad_norm": 1.585727334022522,
"learning_rate": 1.3958426647000346e-05,
"loss": 1.213663101196289,
"step": 418
},
{
"epoch": 0.8879492600422833,
"grad_norm": 1.8782931566238403,
"learning_rate": 1.3945796484410508e-05,
"loss": 1.1034950017929077,
"step": 420
},
{
"epoch": 0.8921775898520085,
"grad_norm": 6.022802829742432,
"learning_rate": 1.3933096688420315e-05,
"loss": 0.9355335235595703,
"step": 422
},
{
"epoch": 0.8964059196617337,
"grad_norm": 2.1923162937164307,
"learning_rate": 1.3920327414289881e-05,
"loss": 1.2361102104187012,
"step": 424
},
{
"epoch": 0.9006342494714588,
"grad_norm": 1.5879871845245361,
"learning_rate": 1.3907488818128711e-05,
"loss": 1.1032028198242188,
"step": 426
},
{
"epoch": 0.904862579281184,
"grad_norm": 3.2801454067230225,
"learning_rate": 1.3894581056893797e-05,
"loss": 0.8857176303863525,
"step": 428
},
{
"epoch": 0.9090909090909091,
"grad_norm": 4.478187084197998,
"learning_rate": 1.388160428838771e-05,
"loss": 1.114775538444519,
"step": 430
},
{
"epoch": 0.9133192389006343,
"grad_norm": 3.5156190395355225,
"learning_rate": 1.386855867125666e-05,
"loss": 1.2656114101409912,
"step": 432
},
{
"epoch": 0.9175475687103594,
"grad_norm": 5.811722755432129,
"learning_rate": 1.3855444364988552e-05,
"loss": 0.5445010662078857,
"step": 434
},
{
"epoch": 0.9217758985200846,
"grad_norm": 2.738983392715454,
"learning_rate": 1.3842261529911054e-05,
"loss": 0.36763542890548706,
"step": 436
},
{
"epoch": 0.9260042283298098,
"grad_norm": 0.915411651134491,
"learning_rate": 1.382901032718962e-05,
"loss": 0.947887659072876,
"step": 438
},
{
"epoch": 0.9302325581395349,
"grad_norm": 1.0174638032913208,
"learning_rate": 1.3815690918825522e-05,
"loss": 1.1414650678634644,
"step": 440
},
{
"epoch": 0.9344608879492601,
"grad_norm": 3.6919026374816895,
"learning_rate": 1.3802303467653885e-05,
"loss": 0.7328312397003174,
"step": 442
},
{
"epoch": 0.9386892177589852,
"grad_norm": 1.6336296796798706,
"learning_rate": 1.3788848137341675e-05,
"loss": 1.2639272212982178,
"step": 444
},
{
"epoch": 0.9429175475687104,
"grad_norm": 2.6586177349090576,
"learning_rate": 1.3775325092385711e-05,
"loss": 1.1174240112304688,
"step": 446
},
{
"epoch": 0.9471458773784355,
"grad_norm": 6.325781345367432,
"learning_rate": 1.3761734498110652e-05,
"loss": 1.109105110168457,
"step": 448
},
{
"epoch": 0.9513742071881607,
"grad_norm": 12.293950080871582,
"learning_rate": 1.3748076520666974e-05,
"loss": 0.829167902469635,
"step": 450
},
{
"epoch": 0.9556025369978859,
"grad_norm": 2.417140483856201,
"learning_rate": 1.3734351327028936e-05,
"loss": 0.9364444017410278,
"step": 452
},
{
"epoch": 0.959830866807611,
"grad_norm": 3.063101053237915,
"learning_rate": 1.3720559084992541e-05,
"loss": 1.323636531829834,
"step": 454
},
{
"epoch": 0.9640591966173362,
"grad_norm": 4.471503257751465,
"learning_rate": 1.3706699963173498e-05,
"loss": 0.6806684732437134,
"step": 456
},
{
"epoch": 0.9682875264270613,
"grad_norm": 3.676957607269287,
"learning_rate": 1.3692774131005132e-05,
"loss": 0.8966299295425415,
"step": 458
},
{
"epoch": 0.9725158562367865,
"grad_norm": 1.8058805465698242,
"learning_rate": 1.3678781758736337e-05,
"loss": 1.237372875213623,
"step": 460
},
{
"epoch": 0.9767441860465116,
"grad_norm": 1.6501954793930054,
"learning_rate": 1.3664723017429491e-05,
"loss": 1.214278221130371,
"step": 462
},
{
"epoch": 0.9809725158562368,
"grad_norm": 15.648929595947266,
"learning_rate": 1.3650598078958347e-05,
"loss": 0.7633882761001587,
"step": 464
},
{
"epoch": 0.985200845665962,
"grad_norm": 1.6495610475540161,
"learning_rate": 1.3636407116005958e-05,
"loss": 1.1871402263641357,
"step": 466
},
{
"epoch": 0.9894291754756871,
"grad_norm": 7.437229633331299,
"learning_rate": 1.3622150302062543e-05,
"loss": 1.0027024745941162,
"step": 468
},
{
"epoch": 0.9936575052854123,
"grad_norm": 4.10299015045166,
"learning_rate": 1.3607827811423387e-05,
"loss": 1.2489466667175293,
"step": 470
},
{
"epoch": 0.9978858350951374,
"grad_norm": 2.741398572921753,
"learning_rate": 1.3593439819186684e-05,
"loss": 1.2514197826385498,
"step": 472
},
{
"epoch": 1.0021141649048626,
"grad_norm": 3.852450132369995,
"learning_rate": 1.3578986501251426e-05,
"loss": 0.9532628059387207,
"step": 474
},
{
"epoch": 1.0063424947145878,
"grad_norm": 1.6289736032485962,
"learning_rate": 1.3564468034315229e-05,
"loss": 1.0436811447143555,
"step": 476
},
{
"epoch": 1.0105708245243128,
"grad_norm": 4.520769119262695,
"learning_rate": 1.3549884595872185e-05,
"loss": 0.47843262553215027,
"step": 478
},
{
"epoch": 1.014799154334038,
"grad_norm": 3.184231996536255,
"learning_rate": 1.3535236364210687e-05,
"loss": 0.6820696592330933,
"step": 480
},
{
"epoch": 1.0190274841437632,
"grad_norm": 2.425349712371826,
"learning_rate": 1.3520523518411252e-05,
"loss": 0.6973198652267456,
"step": 482
},
{
"epoch": 1.0232558139534884,
"grad_norm": 1.9616705179214478,
"learning_rate": 1.350574623834433e-05,
"loss": 0.8969473838806152,
"step": 484
},
{
"epoch": 1.0274841437632136,
"grad_norm": 4.929019451141357,
"learning_rate": 1.349090470466811e-05,
"loss": 1.0332489013671875,
"step": 486
},
{
"epoch": 1.0317124735729386,
"grad_norm": 1.854992151260376,
"learning_rate": 1.3475999098826302e-05,
"loss": 0.8985217809677124,
"step": 488
},
{
"epoch": 1.0359408033826638,
"grad_norm": 7.443341255187988,
"learning_rate": 1.3461029603045926e-05,
"loss": 0.8490782976150513,
"step": 490
},
{
"epoch": 1.040169133192389,
"grad_norm": 20.886302947998047,
"learning_rate": 1.344599640033508e-05,
"loss": 0.7396003007888794,
"step": 492
},
{
"epoch": 1.0443974630021142,
"grad_norm": 3.4426891803741455,
"learning_rate": 1.3430899674480709e-05,
"loss": 1.1333117485046387,
"step": 494
},
{
"epoch": 1.0486257928118394,
"grad_norm": 1.7844817638397217,
"learning_rate": 1.3415739610046352e-05,
"loss": 1.0422849655151367,
"step": 496
},
{
"epoch": 1.0528541226215644,
"grad_norm": 22.077728271484375,
"learning_rate": 1.3400516392369888e-05,
"loss": 0.7254852652549744,
"step": 498
},
{
"epoch": 1.0570824524312896,
"grad_norm": 2.2929744720458984,
"learning_rate": 1.3385230207561267e-05,
"loss": 0.8586277961730957,
"step": 500
},
{
"epoch": 1.0613107822410148,
"grad_norm": 1.4192532300949097,
"learning_rate": 1.336988124250024e-05,
"loss": 1.0733442306518555,
"step": 502
},
{
"epoch": 1.06553911205074,
"grad_norm": 2.207188606262207,
"learning_rate": 1.3354469684834074e-05,
"loss": 0.9969977140426636,
"step": 504
},
{
"epoch": 1.069767441860465,
"grad_norm": 1.835033655166626,
"learning_rate": 1.3338995722975248e-05,
"loss": 0.7394174337387085,
"step": 506
},
{
"epoch": 1.0739957716701902,
"grad_norm": 1.427870512008667,
"learning_rate": 1.332345954609917e-05,
"loss": 0.7539105415344238,
"step": 508
},
{
"epoch": 1.0782241014799154,
"grad_norm": 35.589717864990234,
"learning_rate": 1.3307861344141838e-05,
"loss": 0.8321530818939209,
"step": 510
},
{
"epoch": 1.0824524312896406,
"grad_norm": 2.133634328842163,
"learning_rate": 1.3292201307797538e-05,
"loss": 0.9340323209762573,
"step": 512
},
{
"epoch": 1.0866807610993658,
"grad_norm": 1.4967067241668701,
"learning_rate": 1.3276479628516509e-05,
"loss": 0.9899167418479919,
"step": 514
},
{
"epoch": 1.0909090909090908,
"grad_norm": 1.8359649181365967,
"learning_rate": 1.3260696498502594e-05,
"loss": 0.8009569644927979,
"step": 516
},
{
"epoch": 1.095137420718816,
"grad_norm": 1.3696305751800537,
"learning_rate": 1.3244852110710897e-05,
"loss": 0.7826880216598511,
"step": 518
},
{
"epoch": 1.0993657505285412,
"grad_norm": 2.6647791862487793,
"learning_rate": 1.3228946658845429e-05,
"loss": 0.6374254822731018,
"step": 520
},
{
"epoch": 1.1035940803382664,
"grad_norm": 2.1591572761535645,
"learning_rate": 1.3212980337356727e-05,
"loss": 1.2522598505020142,
"step": 522
},
{
"epoch": 1.1078224101479917,
"grad_norm": 2.5482373237609863,
"learning_rate": 1.3196953341439487e-05,
"loss": 0.7302322387695312,
"step": 524
},
{
"epoch": 1.1120507399577166,
"grad_norm": 2.32763671875,
"learning_rate": 1.3180865867030169e-05,
"loss": 1.2803655862808228,
"step": 526
},
{
"epoch": 1.1162790697674418,
"grad_norm": 1.3799355030059814,
"learning_rate": 1.3164718110804614e-05,
"loss": 0.8244212865829468,
"step": 528
},
{
"epoch": 1.120507399577167,
"grad_norm": 4.998963356018066,
"learning_rate": 1.314851027017563e-05,
"loss": 0.5071104168891907,
"step": 530
},
{
"epoch": 1.1247357293868923,
"grad_norm": 2.4235212802886963,
"learning_rate": 1.3132242543290577e-05,
"loss": 0.9878979921340942,
"step": 532
},
{
"epoch": 1.1289640591966172,
"grad_norm": 2.0776116847991943,
"learning_rate": 1.3115915129028958e-05,
"loss": 0.735676646232605,
"step": 534
},
{
"epoch": 1.1331923890063424,
"grad_norm": 1.01461923122406,
"learning_rate": 1.3099528226999963e-05,
"loss": 0.9803445339202881,
"step": 536
},
{
"epoch": 1.1374207188160677,
"grad_norm": 15.57121467590332,
"learning_rate": 1.308308203754006e-05,
"loss": 0.7138959169387817,
"step": 538
},
{
"epoch": 1.1416490486257929,
"grad_norm": 3.679962635040283,
"learning_rate": 1.3066576761710523e-05,
"loss": 0.7000383138656616,
"step": 540
},
{
"epoch": 1.145877378435518,
"grad_norm": 2.4361116886138916,
"learning_rate": 1.3050012601294974e-05,
"loss": 1.0052709579467773,
"step": 542
},
{
"epoch": 1.150105708245243,
"grad_norm": 3.9788804054260254,
"learning_rate": 1.3033389758796933e-05,
"loss": 0.7173207402229309,
"step": 544
},
{
"epoch": 1.1543340380549683,
"grad_norm": 10.04750919342041,
"learning_rate": 1.301670843743733e-05,
"loss": 1.1154426336288452,
"step": 546
},
{
"epoch": 1.1585623678646935,
"grad_norm": 3.304776668548584,
"learning_rate": 1.2999968841152018e-05,
"loss": 1.2687135934829712,
"step": 548
},
{
"epoch": 1.1627906976744187,
"grad_norm": 3.297494649887085,
"learning_rate": 1.2983171174589283e-05,
"loss": 0.9368158578872681,
"step": 550
},
{
"epoch": 1.1670190274841437,
"grad_norm": 2.040799856185913,
"learning_rate": 1.2966315643107354e-05,
"loss": 1.014180302619934,
"step": 552
},
{
"epoch": 1.1712473572938689,
"grad_norm": 2.4606637954711914,
"learning_rate": 1.294940245277187e-05,
"loss": 1.143074870109558,
"step": 554
},
{
"epoch": 1.175475687103594,
"grad_norm": 2.0511674880981445,
"learning_rate": 1.2932431810353383e-05,
"loss": 0.9233866930007935,
"step": 556
},
{
"epoch": 1.1797040169133193,
"grad_norm": 2.6982901096343994,
"learning_rate": 1.2915403923324808e-05,
"loss": 0.9669268727302551,
"step": 558
},
{
"epoch": 1.1839323467230445,
"grad_norm": 1.5301854610443115,
"learning_rate": 1.289831899985891e-05,
"loss": 0.9145544171333313,
"step": 560
},
{
"epoch": 1.1881606765327695,
"grad_norm": 1.4341201782226562,
"learning_rate": 1.2881177248825742e-05,
"loss": 0.629718542098999,
"step": 562
},
{
"epoch": 1.1923890063424947,
"grad_norm": 1.6134895086288452,
"learning_rate": 1.2863978879790097e-05,
"loss": 0.9676129221916199,
"step": 564
},
{
"epoch": 1.1966173361522199,
"grad_norm": 4.984455108642578,
"learning_rate": 1.2846724103008943e-05,
"loss": 0.7948588132858276,
"step": 566
},
{
"epoch": 1.200845665961945,
"grad_norm": 3.2547831535339355,
"learning_rate": 1.2829413129428864e-05,
"loss": 1.1670663356781006,
"step": 568
},
{
"epoch": 1.20507399577167,
"grad_norm": 4.860750198364258,
"learning_rate": 1.2812046170683459e-05,
"loss": 1.171966791152954,
"step": 570
},
{
"epoch": 1.2093023255813953,
"grad_norm": 2.6650962829589844,
"learning_rate": 1.2794623439090783e-05,
"loss": 0.9701319932937622,
"step": 572
},
{
"epoch": 1.2135306553911205,
"grad_norm": 1.811000108718872,
"learning_rate": 1.2777145147650728e-05,
"loss": 0.7256935238838196,
"step": 574
},
{
"epoch": 1.2177589852008457,
"grad_norm": 3.8662736415863037,
"learning_rate": 1.2759611510042426e-05,
"loss": 0.8213169574737549,
"step": 576
},
{
"epoch": 1.221987315010571,
"grad_norm": 6.631006240844727,
"learning_rate": 1.2742022740621643e-05,
"loss": 0.9658501744270325,
"step": 578
},
{
"epoch": 1.226215644820296,
"grad_norm": 2.3924484252929688,
"learning_rate": 1.2724379054418152e-05,
"loss": 0.7121154069900513,
"step": 580
},
{
"epoch": 1.230443974630021,
"grad_norm": 2.506420850753784,
"learning_rate": 1.2706680667133107e-05,
"loss": 0.9701933860778809,
"step": 582
},
{
"epoch": 1.2346723044397463,
"grad_norm": 3.6654019355773926,
"learning_rate": 1.26889277951364e-05,
"loss": 0.8110610246658325,
"step": 584
},
{
"epoch": 1.2389006342494715,
"grad_norm": 6.644792556762695,
"learning_rate": 1.2671120655464021e-05,
"loss": 0.7942482829093933,
"step": 586
},
{
"epoch": 1.2431289640591967,
"grad_norm": 1.8644226789474487,
"learning_rate": 1.2653259465815411e-05,
"loss": 0.7716197967529297,
"step": 588
},
{
"epoch": 1.2473572938689217,
"grad_norm": 5.121194839477539,
"learning_rate": 1.263534444455079e-05,
"loss": 0.6083531975746155,
"step": 590
},
{
"epoch": 1.251585623678647,
"grad_norm": 2.7642040252685547,
"learning_rate": 1.2617375810688487e-05,
"loss": 0.7866605520248413,
"step": 592
},
{
"epoch": 1.255813953488372,
"grad_norm": 3.0334649085998535,
"learning_rate": 1.2599353783902274e-05,
"loss": 1.0548347234725952,
"step": 594
},
{
"epoch": 1.2600422832980973,
"grad_norm": 1.5484217405319214,
"learning_rate": 1.2581278584518666e-05,
"loss": 0.6410794258117676,
"step": 596
},
{
"epoch": 1.2642706131078225,
"grad_norm": 1.9719972610473633,
"learning_rate": 1.2563150433514236e-05,
"loss": 1.0915131568908691,
"step": 598
},
{
"epoch": 1.2684989429175475,
"grad_norm": 3.951808214187622,
"learning_rate": 1.2544969552512918e-05,
"loss": 1.2211267948150635,
"step": 600
},
{
"epoch": 1.2727272727272727,
"grad_norm": 1.1500049829483032,
"learning_rate": 1.252673616378328e-05,
"loss": 0.8765637874603271,
"step": 602
},
{
"epoch": 1.276955602536998,
"grad_norm": 1.5278034210205078,
"learning_rate": 1.2508450490235831e-05,
"loss": 1.0353968143463135,
"step": 604
},
{
"epoch": 1.2811839323467231,
"grad_norm": 4.614338397979736,
"learning_rate": 1.2490112755420276e-05,
"loss": 0.702555775642395,
"step": 606
},
{
"epoch": 1.285412262156448,
"grad_norm": 1.860958456993103,
"learning_rate": 1.2471723183522784e-05,
"loss": 1.0188829898834229,
"step": 608
},
{
"epoch": 1.2896405919661733,
"grad_norm": 2.774688243865967,
"learning_rate": 1.2453281999363266e-05,
"loss": 0.7039048671722412,
"step": 610
},
{
"epoch": 1.2938689217758985,
"grad_norm": 2.1111514568328857,
"learning_rate": 1.2434789428392605e-05,
"loss": 0.6525111198425293,
"step": 612
},
{
"epoch": 1.2980972515856237,
"grad_norm": 2.6182940006256104,
"learning_rate": 1.241624569668991e-05,
"loss": 0.8683521747589111,
"step": 614
},
{
"epoch": 1.302325581395349,
"grad_norm": 2.9026830196380615,
"learning_rate": 1.2397651030959753e-05,
"loss": 0.4969242215156555,
"step": 616
},
{
"epoch": 1.306553911205074,
"grad_norm": 4.860730171203613,
"learning_rate": 1.2379005658529392e-05,
"loss": 0.8436973094940186,
"step": 618
},
{
"epoch": 1.3107822410147991,
"grad_norm": 1.8341456651687622,
"learning_rate": 1.2360309807345995e-05,
"loss": 1.0782966613769531,
"step": 620
},
{
"epoch": 1.3150105708245243,
"grad_norm": 3.7281575202941895,
"learning_rate": 1.2341563705973856e-05,
"loss": 0.9950319528579712,
"step": 622
},
{
"epoch": 1.3192389006342495,
"grad_norm": 6.979416370391846,
"learning_rate": 1.2322767583591596e-05,
"loss": 0.6381583213806152,
"step": 624
},
{
"epoch": 1.3234672304439745,
"grad_norm": 1.2973817586898804,
"learning_rate": 1.2303921669989356e-05,
"loss": 1.020182490348816,
"step": 626
},
{
"epoch": 1.3276955602536997,
"grad_norm": 7.5433831214904785,
"learning_rate": 1.228502619556601e-05,
"loss": 1.1785922050476074,
"step": 628
},
{
"epoch": 1.331923890063425,
"grad_norm": 1.7187669277191162,
"learning_rate": 1.226608139132632e-05,
"loss": 0.8999447226524353,
"step": 630
},
{
"epoch": 1.3361522198731501,
"grad_norm": 2.4582080841064453,
"learning_rate": 1.2247087488878131e-05,
"loss": 0.6647698283195496,
"step": 632
},
{
"epoch": 1.3403805496828753,
"grad_norm": 1.678996205329895,
"learning_rate": 1.2228044720429527e-05,
"loss": 0.7422084808349609,
"step": 634
},
{
"epoch": 1.3446088794926006,
"grad_norm": 1.84951913356781,
"learning_rate": 1.2208953318786009e-05,
"loss": 0.7743874788284302,
"step": 636
},
{
"epoch": 1.3488372093023255,
"grad_norm": 6.82949161529541,
"learning_rate": 1.2189813517347628e-05,
"loss": 0.981259822845459,
"step": 638
},
{
"epoch": 1.3530655391120507,
"grad_norm": 1.8509401082992554,
"learning_rate": 1.2170625550106147e-05,
"loss": 0.7372652292251587,
"step": 640
},
{
"epoch": 1.357293868921776,
"grad_norm": 1.097352385520935,
"learning_rate": 1.2151389651642174e-05,
"loss": 1.1971924304962158,
"step": 642
},
{
"epoch": 1.361522198731501,
"grad_norm": 1.1925487518310547,
"learning_rate": 1.2132106057122298e-05,
"loss": 0.9470350742340088,
"step": 644
},
{
"epoch": 1.3657505285412261,
"grad_norm": 4.496725559234619,
"learning_rate": 1.2112775002296207e-05,
"loss": 0.7435013055801392,
"step": 646
},
{
"epoch": 1.3699788583509513,
"grad_norm": 2.971543788909912,
"learning_rate": 1.2093396723493814e-05,
"loss": 0.7268692255020142,
"step": 648
},
{
"epoch": 1.3742071881606766,
"grad_norm": 4.587360858917236,
"learning_rate": 1.2073971457622361e-05,
"loss": 0.8199620246887207,
"step": 650
},
{
"epoch": 1.3784355179704018,
"grad_norm": 1.7165014743804932,
"learning_rate": 1.2054499442163533e-05,
"loss": 0.6943320035934448,
"step": 652
},
{
"epoch": 1.382663847780127,
"grad_norm": 1.7177228927612305,
"learning_rate": 1.2034980915170535e-05,
"loss": 1.1553151607513428,
"step": 654
},
{
"epoch": 1.386892177589852,
"grad_norm": 1.452949047088623,
"learning_rate": 1.2015416115265201e-05,
"loss": 0.9875773787498474,
"step": 656
},
{
"epoch": 1.3911205073995772,
"grad_norm": 3.4342682361602783,
"learning_rate": 1.1995805281635077e-05,
"loss": 0.8191255331039429,
"step": 658
},
{
"epoch": 1.3953488372093024,
"grad_norm": 1.2341606616973877,
"learning_rate": 1.1976148654030476e-05,
"loss": 0.9288697242736816,
"step": 660
},
{
"epoch": 1.3995771670190273,
"grad_norm": 1.4650920629501343,
"learning_rate": 1.1956446472761575e-05,
"loss": 1.0698814392089844,
"step": 662
},
{
"epoch": 1.4038054968287526,
"grad_norm": 0.4060530960559845,
"learning_rate": 1.193669897869545e-05,
"loss": 0.7074742913246155,
"step": 664
},
{
"epoch": 1.4080338266384778,
"grad_norm": 2.6720030307769775,
"learning_rate": 1.191690641325315e-05,
"loss": 1.281816005706787,
"step": 666
},
{
"epoch": 1.412262156448203,
"grad_norm": 1.1645246744155884,
"learning_rate": 1.1897069018406743e-05,
"loss": 1.024438738822937,
"step": 668
},
{
"epoch": 1.4164904862579282,
"grad_norm": 1.2744784355163574,
"learning_rate": 1.1877187036676345e-05,
"loss": 0.6395283937454224,
"step": 670
},
{
"epoch": 1.4207188160676534,
"grad_norm": 1.13683021068573,
"learning_rate": 1.1857260711127173e-05,
"loss": 0.9569779634475708,
"step": 672
},
{
"epoch": 1.4249471458773784,
"grad_norm": 1.6644426584243774,
"learning_rate": 1.1837290285366566e-05,
"loss": 0.5713965892791748,
"step": 674
},
{
"epoch": 1.4291754756871036,
"grad_norm": 3.0694689750671387,
"learning_rate": 1.1817276003540995e-05,
"loss": 0.7255022525787354,
"step": 676
},
{
"epoch": 1.4334038054968288,
"grad_norm": 1.5061650276184082,
"learning_rate": 1.1797218110333103e-05,
"loss": 1.177720308303833,
"step": 678
},
{
"epoch": 1.437632135306554,
"grad_norm": 1.372604489326477,
"learning_rate": 1.177711685095869e-05,
"loss": 1.1841721534729004,
"step": 680
},
{
"epoch": 1.441860465116279,
"grad_norm": 6.546939849853516,
"learning_rate": 1.1756972471163728e-05,
"loss": 1.1385557651519775,
"step": 682
},
{
"epoch": 1.4460887949260042,
"grad_norm": 1.3319751024246216,
"learning_rate": 1.1736785217221353e-05,
"loss": 0.9709225296974182,
"step": 684
},
{
"epoch": 1.4503171247357294,
"grad_norm": 3.989851474761963,
"learning_rate": 1.1716555335928856e-05,
"loss": 1.3431732654571533,
"step": 686
},
{
"epoch": 1.4545454545454546,
"grad_norm": 1.1075016260147095,
"learning_rate": 1.1696283074604662e-05,
"loss": 1.0651729106903076,
"step": 688
},
{
"epoch": 1.4587737843551798,
"grad_norm": 4.182896137237549,
"learning_rate": 1.1675968681085309e-05,
"loss": 0.45350563526153564,
"step": 690
},
{
"epoch": 1.463002114164905,
"grad_norm": 0.982246994972229,
"learning_rate": 1.165561240372242e-05,
"loss": 1.1992805004119873,
"step": 692
},
{
"epoch": 1.46723044397463,
"grad_norm": 1.5104455947875977,
"learning_rate": 1.1635214491379666e-05,
"loss": 0.5286851525306702,
"step": 694
},
{
"epoch": 1.4714587737843552,
"grad_norm": 3.885072946548462,
"learning_rate": 1.1614775193429719e-05,
"loss": 1.0592529773712158,
"step": 696
},
{
"epoch": 1.4756871035940804,
"grad_norm": 1.5037254095077515,
"learning_rate": 1.159429475975121e-05,
"loss": 1.0921900272369385,
"step": 698
},
{
"epoch": 1.4799154334038054,
"grad_norm": 2.0736563205718994,
"learning_rate": 1.1573773440725663e-05,
"loss": 0.916094183921814,
"step": 700
},
{
"epoch": 1.4841437632135306,
"grad_norm": 1.5245025157928467,
"learning_rate": 1.1553211487234456e-05,
"loss": 0.9614948630332947,
"step": 702
},
{
"epoch": 1.4883720930232558,
"grad_norm": 1.0767780542373657,
"learning_rate": 1.1532609150655723e-05,
"loss": 0.9404919147491455,
"step": 704
},
{
"epoch": 1.492600422832981,
"grad_norm": 1.4485445022583008,
"learning_rate": 1.1511966682861316e-05,
"loss": 0.6694232225418091,
"step": 706
},
{
"epoch": 1.4968287526427062,
"grad_norm": 0.820898175239563,
"learning_rate": 1.1491284336213685e-05,
"loss": 0.2923370599746704,
"step": 708
},
{
"epoch": 1.5010570824524314,
"grad_norm": 1.8570421934127808,
"learning_rate": 1.1470562363562834e-05,
"loss": 0.6382849216461182,
"step": 710
},
{
"epoch": 1.5052854122621564,
"grad_norm": 15.203941345214844,
"learning_rate": 1.1449801018243205e-05,
"loss": 0.5295338034629822,
"step": 712
},
{
"epoch": 1.5095137420718816,
"grad_norm": 2.6793768405914307,
"learning_rate": 1.1429000554070582e-05,
"loss": 0.9053052067756653,
"step": 714
},
{
"epoch": 1.5137420718816068,
"grad_norm": 1.2564373016357422,
"learning_rate": 1.1408161225339e-05,
"loss": 1.0036495923995972,
"step": 716
},
{
"epoch": 1.5179704016913318,
"grad_norm": 1.1565278768539429,
"learning_rate": 1.138728328681762e-05,
"loss": 0.8999833464622498,
"step": 718
},
{
"epoch": 1.522198731501057,
"grad_norm": 0.9645926356315613,
"learning_rate": 1.1366366993747633e-05,
"loss": 0.7568279504776001,
"step": 720
},
{
"epoch": 1.5264270613107822,
"grad_norm": 3.0898590087890625,
"learning_rate": 1.1345412601839121e-05,
"loss": 0.8676016330718994,
"step": 722
},
{
"epoch": 1.5306553911205074,
"grad_norm": 1.0732699632644653,
"learning_rate": 1.1324420367267947e-05,
"loss": 0.8509583473205566,
"step": 724
},
{
"epoch": 1.5348837209302326,
"grad_norm": 1.6352031230926514,
"learning_rate": 1.1303390546672611e-05,
"loss": 0.8795561790466309,
"step": 726
},
{
"epoch": 1.5391120507399578,
"grad_norm": 1.9884954690933228,
"learning_rate": 1.128232339715112e-05,
"loss": 0.6407184600830078,
"step": 728
},
{
"epoch": 1.543340380549683,
"grad_norm": 10.214325904846191,
"learning_rate": 1.1261219176257841e-05,
"loss": 0.8139981031417847,
"step": 730
},
{
"epoch": 1.547568710359408,
"grad_norm": 1.1405316591262817,
"learning_rate": 1.124007814200035e-05,
"loss": 0.9641261100769043,
"step": 732
},
{
"epoch": 1.5517970401691332,
"grad_norm": 1.1297786235809326,
"learning_rate": 1.1218900552836286e-05,
"loss": 0.23017296195030212,
"step": 734
},
{
"epoch": 1.5560253699788582,
"grad_norm": 3.067251682281494,
"learning_rate": 1.1197686667670186e-05,
"loss": 0.9786266088485718,
"step": 736
},
{
"epoch": 1.5602536997885834,
"grad_norm": 0.759483277797699,
"learning_rate": 1.1176436745850315e-05,
"loss": 0.6439876556396484,
"step": 738
},
{
"epoch": 1.5644820295983086,
"grad_norm": 1.4957109689712524,
"learning_rate": 1.1155151047165512e-05,
"loss": 1.026474118232727,
"step": 740
},
{
"epoch": 1.5687103594080338,
"grad_norm": 3.6359450817108154,
"learning_rate": 1.1133829831841988e-05,
"loss": 0.7543759942054749,
"step": 742
},
{
"epoch": 1.572938689217759,
"grad_norm": 2.107734441757202,
"learning_rate": 1.1112473360540172e-05,
"loss": 1.04964280128479,
"step": 744
},
{
"epoch": 1.5771670190274842,
"grad_norm": 3.1930649280548096,
"learning_rate": 1.1091081894351504e-05,
"loss": 1.1645326614379883,
"step": 746
},
{
"epoch": 1.5813953488372094,
"grad_norm": 1.797052025794983,
"learning_rate": 1.1069655694795247e-05,
"loss": 0.8213366270065308,
"step": 748
},
{
"epoch": 1.5856236786469344,
"grad_norm": 1.1854742765426636,
"learning_rate": 1.1048195023815304e-05,
"loss": 0.10135793685913086,
"step": 750
},
{
"epoch": 1.5898520084566596,
"grad_norm": 3.660308599472046,
"learning_rate": 1.1026700143776998e-05,
"loss": 0.7644736170768738,
"step": 752
},
{
"epoch": 1.5940803382663846,
"grad_norm": 3.2544729709625244,
"learning_rate": 1.1005171317463876e-05,
"loss": 1.1141167879104614,
"step": 754
},
{
"epoch": 1.5983086680761098,
"grad_norm": 6.199263572692871,
"learning_rate": 1.0983608808074485e-05,
"loss": 1.078895092010498,
"step": 756
},
{
"epoch": 1.602536997885835,
"grad_norm": 3.1188950538635254,
"learning_rate": 1.0962012879219166e-05,
"loss": 1.098402976989746,
"step": 758
},
{
"epoch": 1.6067653276955602,
"grad_norm": 1.6134928464889526,
"learning_rate": 1.094038379491683e-05,
"loss": 0.9861317873001099,
"step": 760
},
{
"epoch": 1.6109936575052854,
"grad_norm": 2.958409547805786,
"learning_rate": 1.0918721819591716e-05,
"loss": 1.022116780281067,
"step": 762
},
{
"epoch": 1.6152219873150107,
"grad_norm": 2.212822437286377,
"learning_rate": 1.089702721807018e-05,
"loss": 0.9593652486801147,
"step": 764
},
{
"epoch": 1.6194503171247359,
"grad_norm": 2.670440435409546,
"learning_rate": 1.087530025557744e-05,
"loss": 0.7969526052474976,
"step": 766
},
{
"epoch": 1.6236786469344608,
"grad_norm": 1.4264942407608032,
"learning_rate": 1.0853541197734342e-05,
"loss": 1.06660795211792,
"step": 768
},
{
"epoch": 1.627906976744186,
"grad_norm": 5.848434925079346,
"learning_rate": 1.0831750310554107e-05,
"loss": 1.1033109426498413,
"step": 770
},
{
"epoch": 1.6321353065539113,
"grad_norm": 1.063353180885315,
"learning_rate": 1.0809927860439087e-05,
"loss": 0.963697075843811,
"step": 772
},
{
"epoch": 1.6363636363636362,
"grad_norm": 1.2380177974700928,
"learning_rate": 1.0788074114177505e-05,
"loss": 0.7598190307617188,
"step": 774
},
{
"epoch": 1.6405919661733614,
"grad_norm": 1.744482398033142,
"learning_rate": 1.0766189338940183e-05,
"loss": 1.0616977214813232,
"step": 776
},
{
"epoch": 1.6448202959830867,
"grad_norm": 1.7904915809631348,
"learning_rate": 1.0744273802277298e-05,
"loss": 0.6911699175834656,
"step": 778
},
{
"epoch": 1.6490486257928119,
"grad_norm": 1.733080506324768,
"learning_rate": 1.0722327772115085e-05,
"loss": 1.017983317375183,
"step": 780
},
{
"epoch": 1.653276955602537,
"grad_norm": 1.3515630960464478,
"learning_rate": 1.0700351516752581e-05,
"loss": 0.7734979391098022,
"step": 782
},
{
"epoch": 1.6575052854122623,
"grad_norm": 1.638495683670044,
"learning_rate": 1.0678345304858339e-05,
"loss": 1.1936254501342773,
"step": 784
},
{
"epoch": 1.6617336152219875,
"grad_norm": 4.533325672149658,
"learning_rate": 1.0656309405467135e-05,
"loss": 0.4745456874370575,
"step": 786
},
{
"epoch": 1.6659619450317125,
"grad_norm": 1.8574708700180054,
"learning_rate": 1.0634244087976699e-05,
"loss": 1.0188617706298828,
"step": 788
},
{
"epoch": 1.6701902748414377,
"grad_norm": 1.523213505744934,
"learning_rate": 1.0612149622144399e-05,
"loss": 0.7065435647964478,
"step": 790
},
{
"epoch": 1.6744186046511627,
"grad_norm": 4.230642795562744,
"learning_rate": 1.0590026278083955e-05,
"loss": 0.3407204747200012,
"step": 792
},
{
"epoch": 1.6786469344608879,
"grad_norm": 1.2451612949371338,
"learning_rate": 1.0567874326262141e-05,
"loss": 0.987960696220398,
"step": 794
},
{
"epoch": 1.682875264270613,
"grad_norm": 3.043879985809326,
"learning_rate": 1.054569403749546e-05,
"loss": 0.6038637161254883,
"step": 796
},
{
"epoch": 1.6871035940803383,
"grad_norm": 1.131476879119873,
"learning_rate": 1.0523485682946862e-05,
"loss": 1.163629174232483,
"step": 798
},
{
"epoch": 1.6913319238900635,
"grad_norm": 1.2595586776733398,
"learning_rate": 1.05012495341224e-05,
"loss": 1.1228001117706299,
"step": 800
},
{
"epoch": 1.6955602536997887,
"grad_norm": 3.9269957542419434,
"learning_rate": 1.0478985862867923e-05,
"loss": 0.5542206168174744,
"step": 802
},
{
"epoch": 1.699788583509514,
"grad_norm": 1.3841426372528076,
"learning_rate": 1.0456694941365764e-05,
"loss": 1.0466166734695435,
"step": 804
},
{
"epoch": 1.7040169133192389,
"grad_norm": 5.384525299072266,
"learning_rate": 1.0434377042131388e-05,
"loss": 0.7406128644943237,
"step": 806
},
{
"epoch": 1.708245243128964,
"grad_norm": 5.958383560180664,
"learning_rate": 1.0412032438010084e-05,
"loss": 0.7035034894943237,
"step": 808
},
{
"epoch": 1.712473572938689,
"grad_norm": 2.011209487915039,
"learning_rate": 1.038966140217361e-05,
"loss": 1.0565907955169678,
"step": 810
},
{
"epoch": 1.7167019027484143,
"grad_norm": 6.048700332641602,
"learning_rate": 1.0367264208116872e-05,
"loss": 1.3097240924835205,
"step": 812
},
{
"epoch": 1.7209302325581395,
"grad_norm": 1.3847702741622925,
"learning_rate": 1.0344841129654561e-05,
"loss": 1.03816556930542,
"step": 814
},
{
"epoch": 1.7251585623678647,
"grad_norm": 2.6086628437042236,
"learning_rate": 1.0322392440917821e-05,
"loss": 1.1049023866653442,
"step": 816
},
{
"epoch": 1.72938689217759,
"grad_norm": 1.1270302534103394,
"learning_rate": 1.0299918416350892e-05,
"loss": 0.8426474332809448,
"step": 818
},
{
"epoch": 1.733615221987315,
"grad_norm": 5.584467887878418,
"learning_rate": 1.027741933070775e-05,
"loss": 1.0087465047836304,
"step": 820
},
{
"epoch": 1.7378435517970403,
"grad_norm": 1.1592305898666382,
"learning_rate": 1.0254895459048762e-05,
"loss": 0.6469322443008423,
"step": 822
},
{
"epoch": 1.7420718816067653,
"grad_norm": 1.016264796257019,
"learning_rate": 1.0232347076737304e-05,
"loss": 0.5781674385070801,
"step": 824
},
{
"epoch": 1.7463002114164905,
"grad_norm": 6.360507488250732,
"learning_rate": 1.0209774459436412e-05,
"loss": 0.8383398652076721,
"step": 826
},
{
"epoch": 1.7505285412262155,
"grad_norm": 1.2847486734390259,
"learning_rate": 1.0187177883105397e-05,
"loss": 0.8656316995620728,
"step": 828
},
{
"epoch": 1.7547568710359407,
"grad_norm": 7.998954772949219,
"learning_rate": 1.0164557623996484e-05,
"loss": 0.7400867938995361,
"step": 830
},
{
"epoch": 1.758985200845666,
"grad_norm": 2.154827117919922,
"learning_rate": 1.0141913958651433e-05,
"loss": 1.0802297592163086,
"step": 832
},
{
"epoch": 1.763213530655391,
"grad_norm": 18.786178588867188,
"learning_rate": 1.0119247163898143e-05,
"loss": 0.7658360600471497,
"step": 834
},
{
"epoch": 1.7674418604651163,
"grad_norm": 1.3450095653533936,
"learning_rate": 1.0096557516847292e-05,
"loss": 0.7721861600875854,
"step": 836
},
{
"epoch": 1.7716701902748415,
"grad_norm": 1.5817804336547852,
"learning_rate": 1.0073845294888928e-05,
"loss": 1.0332725048065186,
"step": 838
},
{
"epoch": 1.7758985200845667,
"grad_norm": 1.091153860092163,
"learning_rate": 1.005111077568909e-05,
"loss": 0.7373213171958923,
"step": 840
},
{
"epoch": 1.7801268498942917,
"grad_norm": 8.042143821716309,
"learning_rate": 1.0028354237186413e-05,
"loss": 0.3466538190841675,
"step": 842
},
{
"epoch": 1.784355179704017,
"grad_norm": 2.642815113067627,
"learning_rate": 1.0005575957588714e-05,
"loss": 0.653435468673706,
"step": 844
},
{
"epoch": 1.7885835095137421,
"grad_norm": 11.255086898803711,
"learning_rate": 9.982776215369619e-06,
"loss": 0.901837944984436,
"step": 846
},
{
"epoch": 1.792811839323467,
"grad_norm": 1.779013991355896,
"learning_rate": 9.959955289265134e-06,
"loss": 0.7240399122238159,
"step": 848
},
{
"epoch": 1.7970401691331923,
"grad_norm": 5.379182815551758,
"learning_rate": 9.937113458270251e-06,
"loss": 0.6619425415992737,
"step": 850
},
{
"epoch": 1.8012684989429175,
"grad_norm": 3.0560693740844727,
"learning_rate": 9.914251001635533e-06,
"loss": 0.9245142936706543,
"step": 852
},
{
"epoch": 1.8054968287526427,
"grad_norm": 1.3373061418533325,
"learning_rate": 9.89136819886369e-06,
"loss": 0.7170429229736328,
"step": 854
},
{
"epoch": 1.809725158562368,
"grad_norm": 1.0971553325653076,
"learning_rate": 9.868465329706187e-06,
"loss": 0.9948302507400513,
"step": 856
},
{
"epoch": 1.8139534883720931,
"grad_norm": 1.444568395614624,
"learning_rate": 9.845542674159795e-06,
"loss": 0.6135228872299194,
"step": 858
},
{
"epoch": 1.8181818181818183,
"grad_norm": 3.6662490367889404,
"learning_rate": 9.822600512463191e-06,
"loss": 0.9966019988059998,
"step": 860
},
{
"epoch": 1.8224101479915433,
"grad_norm": 1.6096436977386475,
"learning_rate": 9.799639125093517e-06,
"loss": 0.980803370475769,
"step": 862
},
{
"epoch": 1.8266384778012685,
"grad_norm": 0.871141254901886,
"learning_rate": 9.776658792762961e-06,
"loss": 0.47347694635391235,
"step": 864
},
{
"epoch": 1.8308668076109935,
"grad_norm": 1.2584807872772217,
"learning_rate": 9.753659796415316e-06,
"loss": 0.6913288831710815,
"step": 866
},
{
"epoch": 1.8350951374207187,
"grad_norm": 1.0984634160995483,
"learning_rate": 9.730642417222552e-06,
"loss": 1.0209715366363525,
"step": 868
},
{
"epoch": 1.839323467230444,
"grad_norm": 1.4410583972930908,
"learning_rate": 9.707606936581379e-06,
"loss": 1.0635502338409424,
"step": 870
},
{
"epoch": 1.8435517970401691,
"grad_norm": 1.0276752710342407,
"learning_rate": 9.684553636109804e-06,
"loss": 0.6426548957824707,
"step": 872
},
{
"epoch": 1.8477801268498943,
"grad_norm": 1.1092467308044434,
"learning_rate": 9.66148279764368e-06,
"loss": 0.9879615902900696,
"step": 874
},
{
"epoch": 1.8520084566596196,
"grad_norm": 11.211944580078125,
"learning_rate": 9.638394703233283e-06,
"loss": 0.8808318376541138,
"step": 876
},
{
"epoch": 1.8562367864693448,
"grad_norm": 6.206437110900879,
"learning_rate": 9.61528963513984e-06,
"loss": 0.3255208432674408,
"step": 878
},
{
"epoch": 1.8604651162790697,
"grad_norm": 3.4275875091552734,
"learning_rate": 9.592167875832095e-06,
"loss": 0.655363917350769,
"step": 880
},
{
"epoch": 1.864693446088795,
"grad_norm": 3.6875438690185547,
"learning_rate": 9.56902970798284e-06,
"loss": 1.0571465492248535,
"step": 882
},
{
"epoch": 1.86892177589852,
"grad_norm": 1.2043085098266602,
"learning_rate": 9.54587541446547e-06,
"loss": 0.6911188364028931,
"step": 884
},
{
"epoch": 1.8731501057082451,
"grad_norm": 2.393512725830078,
"learning_rate": 9.522705278350537e-06,
"loss": 1.0056746006011963,
"step": 886
},
{
"epoch": 1.8773784355179703,
"grad_norm": 1.8014053106307983,
"learning_rate": 9.499519582902253e-06,
"loss": 1.1018199920654297,
"step": 888
},
{
"epoch": 1.8816067653276956,
"grad_norm": 1.8942985534667969,
"learning_rate": 9.476318611575066e-06,
"loss": 0.5265207290649414,
"step": 890
},
{
"epoch": 1.8858350951374208,
"grad_norm": 4.282611846923828,
"learning_rate": 9.45310264801017e-06,
"loss": 0.7664867639541626,
"step": 892
},
{
"epoch": 1.890063424947146,
"grad_norm": 1.7346903085708618,
"learning_rate": 9.429871976032045e-06,
"loss": 1.038188099861145,
"step": 894
},
{
"epoch": 1.8942917547568712,
"grad_norm": 3.938504457473755,
"learning_rate": 9.40662687964499e-06,
"loss": 1.411336064338684,
"step": 896
},
{
"epoch": 1.8985200845665962,
"grad_norm": 1.5503828525543213,
"learning_rate": 9.383367643029648e-06,
"loss": 1.009209156036377,
"step": 898
},
{
"epoch": 1.9027484143763214,
"grad_norm": 2.3528826236724854,
"learning_rate": 9.360094550539532e-06,
"loss": 0.4925802946090698,
"step": 900
},
{
"epoch": 1.9069767441860463,
"grad_norm": 1.2036664485931396,
"learning_rate": 9.336807886697548e-06,
"loss": 1.0337181091308594,
"step": 902
},
{
"epoch": 1.9112050739957716,
"grad_norm": 4.639069080352783,
"learning_rate": 9.313507936192513e-06,
"loss": 0.5048983693122864,
"step": 904
},
{
"epoch": 1.9154334038054968,
"grad_norm": 4.674471855163574,
"learning_rate": 9.290194983875687e-06,
"loss": 0.8788387775421143,
"step": 906
},
{
"epoch": 1.919661733615222,
"grad_norm": 1.6288189888000488,
"learning_rate": 9.266869314757277e-06,
"loss": 1.1300139427185059,
"step": 908
},
{
"epoch": 1.9238900634249472,
"grad_norm": 1.4102869033813477,
"learning_rate": 9.243531214002959e-06,
"loss": 0.8134955763816833,
"step": 910
},
{
"epoch": 1.9281183932346724,
"grad_norm": 1.6341298818588257,
"learning_rate": 9.22018096693039e-06,
"loss": 0.9380933046340942,
"step": 912
},
{
"epoch": 1.9323467230443976,
"grad_norm": 1.9348236322402954,
"learning_rate": 9.196818859005725e-06,
"loss": 0.6617736220359802,
"step": 914
},
{
"epoch": 1.9365750528541226,
"grad_norm": 1.2717820405960083,
"learning_rate": 9.17344517584011e-06,
"loss": 1.0794345140457153,
"step": 916
},
{
"epoch": 1.9408033826638478,
"grad_norm": 1.6110085248947144,
"learning_rate": 9.150060203186227e-06,
"loss": 0.6167547702789307,
"step": 918
},
{
"epoch": 1.945031712473573,
"grad_norm": 2.5592806339263916,
"learning_rate": 9.126664226934752e-06,
"loss": 0.8708864450454712,
"step": 920
},
{
"epoch": 1.949260042283298,
"grad_norm": 3.9090213775634766,
"learning_rate": 9.103257533110894e-06,
"loss": 0.9234741926193237,
"step": 922
},
{
"epoch": 1.9534883720930232,
"grad_norm": 1.3239997625350952,
"learning_rate": 9.079840407870894e-06,
"loss": 0.9471195936203003,
"step": 924
},
{
"epoch": 1.9577167019027484,
"grad_norm": 0.8116739988327026,
"learning_rate": 9.05641313749851e-06,
"loss": 0.7983577251434326,
"step": 926
},
{
"epoch": 1.9619450317124736,
"grad_norm": 3.1777195930480957,
"learning_rate": 9.03297600840154e-06,
"loss": 1.085768461227417,
"step": 928
},
{
"epoch": 1.9661733615221988,
"grad_norm": 1.4461822509765625,
"learning_rate": 9.0095293071083e-06,
"loss": 0.7136105298995972,
"step": 930
},
{
"epoch": 1.970401691331924,
"grad_norm": 1.2198141813278198,
"learning_rate": 8.986073320264131e-06,
"loss": 1.2170066833496094,
"step": 932
},
{
"epoch": 1.9746300211416492,
"grad_norm": 3.118879556655884,
"learning_rate": 8.962608334627904e-06,
"loss": 0.7703591585159302,
"step": 934
},
{
"epoch": 1.9788583509513742,
"grad_norm": 2.456840753555298,
"learning_rate": 8.939134637068488e-06,
"loss": 0.6549793481826782,
"step": 936
},
{
"epoch": 1.9830866807610994,
"grad_norm": 2.0218310356140137,
"learning_rate": 8.915652514561267e-06,
"loss": 0.779523491859436,
"step": 938
},
{
"epoch": 1.9873150105708244,
"grad_norm": 1.5096863508224487,
"learning_rate": 8.892162254184628e-06,
"loss": 0.9950902462005615,
"step": 940
},
{
"epoch": 1.9915433403805496,
"grad_norm": 1.363619327545166,
"learning_rate": 8.868664143116435e-06,
"loss": 1.0220410823822021,
"step": 942
},
{
"epoch": 1.9957716701902748,
"grad_norm": 1.8180921077728271,
"learning_rate": 8.845158468630539e-06,
"loss": 1.0822381973266602,
"step": 944
},
{
"epoch": 2.0,
"grad_norm": 2.787820816040039,
"learning_rate": 8.821645518093257e-06,
"loss": 0.34972119331359863,
"step": 946
},
{
"epoch": 2.004228329809725,
"grad_norm": 24.766681671142578,
"learning_rate": 8.798125578959852e-06,
"loss": 0.6114112138748169,
"step": 948
},
{
"epoch": 2.0084566596194504,
"grad_norm": 1.491407036781311,
"learning_rate": 8.77459893877103e-06,
"loss": 0.8133289813995361,
"step": 950
},
{
"epoch": 2.0126849894291756,
"grad_norm": 3.2587499618530273,
"learning_rate": 8.75106588514942e-06,
"loss": 0.7695038914680481,
"step": 952
},
{
"epoch": 2.016913319238901,
"grad_norm": 3.2523186206817627,
"learning_rate": 8.72752670579606e-06,
"loss": 1.0784715414047241,
"step": 954
},
{
"epoch": 2.0211416490486256,
"grad_norm": 3.879960060119629,
"learning_rate": 8.703981688486867e-06,
"loss": 0.9948104619979858,
"step": 956
},
{
"epoch": 2.025369978858351,
"grad_norm": 2.143833637237549,
"learning_rate": 8.680431121069144e-06,
"loss": 0.6658627986907959,
"step": 958
},
{
"epoch": 2.029598308668076,
"grad_norm": 7.66282844543457,
"learning_rate": 8.656875291458037e-06,
"loss": 0.9412879943847656,
"step": 960
},
{
"epoch": 2.033826638477801,
"grad_norm": 28.617271423339844,
"learning_rate": 8.633314487633026e-06,
"loss": 0.7398244142532349,
"step": 962
},
{
"epoch": 2.0380549682875264,
"grad_norm": 12.662749290466309,
"learning_rate": 8.609748997634405e-06,
"loss": 0.6164466142654419,
"step": 964
},
{
"epoch": 2.0422832980972516,
"grad_norm": 1.8123739957809448,
"learning_rate": 8.586179109559756e-06,
"loss": 0.5718740224838257,
"step": 966
},
{
"epoch": 2.046511627906977,
"grad_norm": 6.212924480438232,
"learning_rate": 8.56260511156043e-06,
"loss": 0.6756606698036194,
"step": 968
},
{
"epoch": 2.050739957716702,
"grad_norm": 3.720684051513672,
"learning_rate": 8.539027291838022e-06,
"loss": 0.7929984927177429,
"step": 970
},
{
"epoch": 2.0549682875264272,
"grad_norm": 8.034695625305176,
"learning_rate": 8.515445938640856e-06,
"loss": 0.4595068097114563,
"step": 972
},
{
"epoch": 2.059196617336152,
"grad_norm": 2.0762386322021484,
"learning_rate": 8.49186134026044e-06,
"loss": 0.8429713249206543,
"step": 974
},
{
"epoch": 2.063424947145877,
"grad_norm": 2.373539686203003,
"learning_rate": 8.468273785027973e-06,
"loss": 0.9035528898239136,
"step": 976
},
{
"epoch": 2.0676532769556024,
"grad_norm": 1.8188412189483643,
"learning_rate": 8.444683561310793e-06,
"loss": 0.7995322346687317,
"step": 978
},
{
"epoch": 2.0718816067653276,
"grad_norm": 2.725459098815918,
"learning_rate": 8.421090957508857e-06,
"loss": 0.694847583770752,
"step": 980
},
{
"epoch": 2.076109936575053,
"grad_norm": 4.285188674926758,
"learning_rate": 8.39749626205123e-06,
"loss": 0.560046911239624,
"step": 982
},
{
"epoch": 2.080338266384778,
"grad_norm": 1.4149065017700195,
"learning_rate": 8.37389976339254e-06,
"loss": 0.534704327583313,
"step": 984
},
{
"epoch": 2.0845665961945032,
"grad_norm": 1.9620476961135864,
"learning_rate": 8.350301750009467e-06,
"loss": 0.8313402533531189,
"step": 986
},
{
"epoch": 2.0887949260042284,
"grad_norm": 2.5083870887756348,
"learning_rate": 8.3267025103972e-06,
"loss": 0.8133327960968018,
"step": 988
},
{
"epoch": 2.0930232558139537,
"grad_norm": 3.950967788696289,
"learning_rate": 8.30310233306593e-06,
"loss": 0.8477989435195923,
"step": 990
},
{
"epoch": 2.097251585623679,
"grad_norm": 2.234276294708252,
"learning_rate": 8.279501506537304e-06,
"loss": 0.5907564759254456,
"step": 992
},
{
"epoch": 2.1014799154334036,
"grad_norm": 6.177698135375977,
"learning_rate": 8.255900319340908e-06,
"loss": 0.545740008354187,
"step": 994
},
{
"epoch": 2.105708245243129,
"grad_norm": 1.6201454401016235,
"learning_rate": 8.232299060010737e-06,
"loss": 0.664707601070404,
"step": 996
},
{
"epoch": 2.109936575052854,
"grad_norm": 1.3820905685424805,
"learning_rate": 8.208698017081669e-06,
"loss": 0.7701420783996582,
"step": 998
},
{
"epoch": 2.1141649048625792,
"grad_norm": 1.5687397718429565,
"learning_rate": 8.185097479085937e-06,
"loss": 0.8425959348678589,
"step": 1000
},
{
"epoch": 2.1183932346723044,
"grad_norm": 1.980387568473816,
"learning_rate": 8.161497734549597e-06,
"loss": 0.3263434171676636,
"step": 1002
},
{
"epoch": 2.1226215644820297,
"grad_norm": 5.675595283508301,
"learning_rate": 8.137899071989008e-06,
"loss": 0.6078680753707886,
"step": 1004
},
{
"epoch": 2.126849894291755,
"grad_norm": 1.2487083673477173,
"learning_rate": 8.1143017799073e-06,
"loss": 0.7753713130950928,
"step": 1006
},
{
"epoch": 2.13107822410148,
"grad_norm": 3.2538247108459473,
"learning_rate": 8.090706146790852e-06,
"loss": 0.7754625678062439,
"step": 1008
},
{
"epoch": 2.1353065539112053,
"grad_norm": 1.3604462146759033,
"learning_rate": 8.067112461105758e-06,
"loss": 0.39474815130233765,
"step": 1010
},
{
"epoch": 2.13953488372093,
"grad_norm": 1.7930740118026733,
"learning_rate": 8.043521011294303e-06,
"loss": 0.42582058906555176,
"step": 1012
},
{
"epoch": 2.1437632135306552,
"grad_norm": 1.4133408069610596,
"learning_rate": 8.01993208577144e-06,
"loss": 0.47508808970451355,
"step": 1014
},
{
"epoch": 2.1479915433403804,
"grad_norm": 3.4921300411224365,
"learning_rate": 7.996345972921259e-06,
"loss": 0.7460800409317017,
"step": 1016
},
{
"epoch": 2.1522198731501057,
"grad_norm": 1.8524998426437378,
"learning_rate": 7.972762961093464e-06,
"loss": 0.7890527248382568,
"step": 1018
},
{
"epoch": 2.156448202959831,
"grad_norm": 2.931391477584839,
"learning_rate": 7.949183338599856e-06,
"loss": 0.5848892331123352,
"step": 1020
},
{
"epoch": 2.160676532769556,
"grad_norm": 1.7194267511367798,
"learning_rate": 7.925607393710787e-06,
"loss": 0.5405454635620117,
"step": 1022
},
{
"epoch": 2.1649048625792813,
"grad_norm": 1.7146341800689697,
"learning_rate": 7.902035414651658e-06,
"loss": 0.7282487750053406,
"step": 1024
},
{
"epoch": 2.1691331923890065,
"grad_norm": 1.6258643865585327,
"learning_rate": 7.878467689599383e-06,
"loss": 0.729529857635498,
"step": 1026
},
{
"epoch": 2.1733615221987317,
"grad_norm": 2.8171350955963135,
"learning_rate": 7.854904506678867e-06,
"loss": 0.726538896560669,
"step": 1028
},
{
"epoch": 2.177589852008457,
"grad_norm": 2.65883469581604,
"learning_rate": 7.831346153959488e-06,
"loss": 0.7592504024505615,
"step": 1030
},
{
"epoch": 2.1818181818181817,
"grad_norm": 1.6712520122528076,
"learning_rate": 7.807792919451577e-06,
"loss": 0.7350326180458069,
"step": 1032
},
{
"epoch": 2.186046511627907,
"grad_norm": 1.674984097480774,
"learning_rate": 7.784245091102887e-06,
"loss": 0.7854993343353271,
"step": 1034
},
{
"epoch": 2.190274841437632,
"grad_norm": 4.203649997711182,
"learning_rate": 7.760702956795074e-06,
"loss": 0.531688928604126,
"step": 1036
},
{
"epoch": 2.1945031712473573,
"grad_norm": 4.289502143859863,
"learning_rate": 7.737166804340196e-06,
"loss": 0.3217397928237915,
"step": 1038
},
{
"epoch": 2.1987315010570825,
"grad_norm": 1.8345874547958374,
"learning_rate": 7.713636921477167e-06,
"loss": 1.013067364692688,
"step": 1040
},
{
"epoch": 2.2029598308668077,
"grad_norm": 7.410592079162598,
"learning_rate": 7.690113595868264e-06,
"loss": 0.2811543941497803,
"step": 1042
},
{
"epoch": 2.207188160676533,
"grad_norm": 2.9365649223327637,
"learning_rate": 7.666597115095591e-06,
"loss": 0.886922299861908,
"step": 1044
},
{
"epoch": 2.211416490486258,
"grad_norm": 1.749424695968628,
"learning_rate": 7.64308776665757e-06,
"loss": 0.37419652938842773,
"step": 1046
},
{
"epoch": 2.2156448202959833,
"grad_norm": 4.167431831359863,
"learning_rate": 7.619585837965435e-06,
"loss": 0.6467190384864807,
"step": 1048
},
{
"epoch": 2.219873150105708,
"grad_norm": 2.9107508659362793,
"learning_rate": 7.596091616339704e-06,
"loss": 0.5514612197875977,
"step": 1050
},
{
"epoch": 2.2241014799154333,
"grad_norm": 5.101666450500488,
"learning_rate": 7.572605389006674e-06,
"loss": 0.7982590198516846,
"step": 1052
},
{
"epoch": 2.2283298097251585,
"grad_norm": 1.5105267763137817,
"learning_rate": 7.549127443094913e-06,
"loss": 0.7884393930435181,
"step": 1054
},
{
"epoch": 2.2325581395348837,
"grad_norm": 1.2124134302139282,
"learning_rate": 7.525658065631742e-06,
"loss": 0.7675731778144836,
"step": 1056
},
{
"epoch": 2.236786469344609,
"grad_norm": 0.5240764617919922,
"learning_rate": 7.502197543539728e-06,
"loss": 0.41273289918899536,
"step": 1058
},
{
"epoch": 2.241014799154334,
"grad_norm": 1.3946956396102905,
"learning_rate": 7.47874616363318e-06,
"loss": 0.5390036106109619,
"step": 1060
},
{
"epoch": 2.2452431289640593,
"grad_norm": 1.6719964742660522,
"learning_rate": 7.455304212614639e-06,
"loss": 0.9307736158370972,
"step": 1062
},
{
"epoch": 2.2494714587737845,
"grad_norm": 3.6784653663635254,
"learning_rate": 7.431871977071373e-06,
"loss": 0.7393144369125366,
"step": 1064
},
{
"epoch": 2.2536997885835097,
"grad_norm": 16.029266357421875,
"learning_rate": 7.408449743471876e-06,
"loss": 0.48589229583740234,
"step": 1066
},
{
"epoch": 2.2579281183932345,
"grad_norm": 3.1889865398406982,
"learning_rate": 7.385037798162367e-06,
"loss": 0.7017960548400879,
"step": 1068
},
{
"epoch": 2.2621564482029597,
"grad_norm": 226.25563049316406,
"learning_rate": 7.361636427363279e-06,
"loss": 0.6053088307380676,
"step": 1070
},
{
"epoch": 2.266384778012685,
"grad_norm": 1.3916795253753662,
"learning_rate": 7.338245917165771e-06,
"loss": 0.8384519815444946,
"step": 1072
},
{
"epoch": 2.27061310782241,
"grad_norm": 1.1973955631256104,
"learning_rate": 7.314866553528231e-06,
"loss": 0.46596986055374146,
"step": 1074
},
{
"epoch": 2.2748414376321353,
"grad_norm": 4.90501594543457,
"learning_rate": 7.291498622272774e-06,
"loss": 0.5769233703613281,
"step": 1076
},
{
"epoch": 2.2790697674418605,
"grad_norm": 3.487074851989746,
"learning_rate": 7.268142409081741e-06,
"loss": 0.27050769329071045,
"step": 1078
},
{
"epoch": 2.2832980972515857,
"grad_norm": 0.8944596648216248,
"learning_rate": 7.244798199494229e-06,
"loss": 0.13462218642234802,
"step": 1080
},
{
"epoch": 2.287526427061311,
"grad_norm": 0.9704837203025818,
"learning_rate": 7.221466278902575e-06,
"loss": 0.3835107088088989,
"step": 1082
},
{
"epoch": 2.291754756871036,
"grad_norm": 1.700408697128296,
"learning_rate": 7.198146932548882e-06,
"loss": 0.31547147035598755,
"step": 1084
},
{
"epoch": 2.295983086680761,
"grad_norm": 2.8504245281219482,
"learning_rate": 7.174840445521527e-06,
"loss": 0.1282675862312317,
"step": 1086
},
{
"epoch": 2.300211416490486,
"grad_norm": 2.286254405975342,
"learning_rate": 7.151547102751679e-06,
"loss": 0.4673500061035156,
"step": 1088
},
{
"epoch": 2.3044397463002113,
"grad_norm": 2.2698566913604736,
"learning_rate": 7.128267189009815e-06,
"loss": 0.7307963371276855,
"step": 1090
},
{
"epoch": 2.3086680761099365,
"grad_norm": 1.4226760864257812,
"learning_rate": 7.1050009889022285e-06,
"loss": 0.5173399448394775,
"step": 1092
},
{
"epoch": 2.3128964059196617,
"grad_norm": 1.411266803741455,
"learning_rate": 7.081748786867563e-06,
"loss": 0.7340717315673828,
"step": 1094
},
{
"epoch": 2.317124735729387,
"grad_norm": 1.8416746854782104,
"learning_rate": 7.058510867173335e-06,
"loss": 0.5856821537017822,
"step": 1096
},
{
"epoch": 2.321353065539112,
"grad_norm": 2.0288174152374268,
"learning_rate": 7.035287513912446e-06,
"loss": 0.7254132032394409,
"step": 1098
},
{
"epoch": 2.3255813953488373,
"grad_norm": 3.963407039642334,
"learning_rate": 7.012079010999715e-06,
"loss": 0.44933539628982544,
"step": 1100
},
{
"epoch": 2.3298097251585626,
"grad_norm": 1.9250879287719727,
"learning_rate": 6.9888856421684235e-06,
"loss": 0.5230907797813416,
"step": 1102
},
{
"epoch": 2.3340380549682873,
"grad_norm": 1.9608304500579834,
"learning_rate": 6.965707690966817e-06,
"loss": 0.8589090704917908,
"step": 1104
},
{
"epoch": 2.3382663847780125,
"grad_norm": 1.8707932233810425,
"learning_rate": 6.94254544075466e-06,
"loss": 0.7368871569633484,
"step": 1106
},
{
"epoch": 2.3424947145877377,
"grad_norm": 3.9660754203796387,
"learning_rate": 6.919399174699772e-06,
"loss": 0.851331353187561,
"step": 1108
},
{
"epoch": 2.346723044397463,
"grad_norm": 3.2200233936309814,
"learning_rate": 6.896269175774555e-06,
"loss": 0.8817049264907837,
"step": 1110
},
{
"epoch": 2.350951374207188,
"grad_norm": 1.1279667615890503,
"learning_rate": 6.8731557267525306e-06,
"loss": 0.4473658800125122,
"step": 1112
},
{
"epoch": 2.3551797040169133,
"grad_norm": 4.966526508331299,
"learning_rate": 6.850059110204907e-06,
"loss": 0.7412988543510437,
"step": 1114
},
{
"epoch": 2.3594080338266386,
"grad_norm": 2.3118531703948975,
"learning_rate": 6.8269796084971e-06,
"loss": 0.4675288796424866,
"step": 1116
},
{
"epoch": 2.3636363636363638,
"grad_norm": 4.307426452636719,
"learning_rate": 6.803917503785286e-06,
"loss": 0.9795839190483093,
"step": 1118
},
{
"epoch": 2.367864693446089,
"grad_norm": 3.040167808532715,
"learning_rate": 6.780873078012963e-06,
"loss": 0.7678594589233398,
"step": 1120
},
{
"epoch": 2.3720930232558137,
"grad_norm": 1.5041873455047607,
"learning_rate": 6.7578466129074925e-06,
"loss": 0.7190483808517456,
"step": 1122
},
{
"epoch": 2.376321353065539,
"grad_norm": 2.039888620376587,
"learning_rate": 6.734838389976669e-06,
"loss": 0.5487040877342224,
"step": 1124
},
{
"epoch": 2.380549682875264,
"grad_norm": 4.9332990646362305,
"learning_rate": 6.71184869050525e-06,
"loss": 0.722797155380249,
"step": 1126
},
{
"epoch": 2.3847780126849893,
"grad_norm": 5.00702428817749,
"learning_rate": 6.688877795551555e-06,
"loss": 0.49754154682159424,
"step": 1128
},
{
"epoch": 2.3890063424947146,
"grad_norm": 2.7888295650482178,
"learning_rate": 6.665925985944006e-06,
"loss": 0.6010280847549438,
"step": 1130
},
{
"epoch": 2.3932346723044398,
"grad_norm": 3.433521032333374,
"learning_rate": 6.642993542277688e-06,
"loss": 0.6383258104324341,
"step": 1132
},
{
"epoch": 2.397463002114165,
"grad_norm": 7.1966400146484375,
"learning_rate": 6.620080744910945e-06,
"loss": 0.7949010133743286,
"step": 1134
},
{
"epoch": 2.40169133192389,
"grad_norm": 8.449408531188965,
"learning_rate": 6.597187873961926e-06,
"loss": 0.50352942943573,
"step": 1136
},
{
"epoch": 2.4059196617336154,
"grad_norm": 2.717254877090454,
"learning_rate": 6.5743152093051795e-06,
"loss": 0.7247951030731201,
"step": 1138
},
{
"epoch": 2.41014799154334,
"grad_norm": 1.7346482276916504,
"learning_rate": 6.551463030568219e-06,
"loss": 0.7071352005004883,
"step": 1140
},
{
"epoch": 2.4143763213530653,
"grad_norm": 4.9090728759765625,
"learning_rate": 6.528631617128107e-06,
"loss": 0.9658875465393066,
"step": 1142
},
{
"epoch": 2.4186046511627906,
"grad_norm": 6.399837493896484,
"learning_rate": 6.505821248108051e-06,
"loss": 0.4537696838378906,
"step": 1144
},
{
"epoch": 2.4228329809725158,
"grad_norm": 5.828219413757324,
"learning_rate": 6.4830322023739745e-06,
"loss": 0.6452757120132446,
"step": 1146
},
{
"epoch": 2.427061310782241,
"grad_norm": 1.6015979051589966,
"learning_rate": 6.4602647585311134e-06,
"loss": 0.5827908515930176,
"step": 1148
},
{
"epoch": 2.431289640591966,
"grad_norm": 2.543041467666626,
"learning_rate": 6.437519194920624e-06,
"loss": 0.46544528007507324,
"step": 1150
},
{
"epoch": 2.4355179704016914,
"grad_norm": 2.145216464996338,
"learning_rate": 6.414795789616152e-06,
"loss": 0.4755411148071289,
"step": 1152
},
{
"epoch": 2.4397463002114166,
"grad_norm": 1.2720885276794434,
"learning_rate": 6.3920948204204574e-06,
"loss": 0.41912633180618286,
"step": 1154
},
{
"epoch": 2.443974630021142,
"grad_norm": 12.633498191833496,
"learning_rate": 6.369416564862013e-06,
"loss": 0.6496953964233398,
"step": 1156
},
{
"epoch": 2.448202959830867,
"grad_norm": 2.8517889976501465,
"learning_rate": 6.346761300191603e-06,
"loss": 0.46596282720565796,
"step": 1158
},
{
"epoch": 2.452431289640592,
"grad_norm": 9.365266799926758,
"learning_rate": 6.324129303378936e-06,
"loss": 0.8000454306602478,
"step": 1160
},
{
"epoch": 2.456659619450317,
"grad_norm": 1.857039213180542,
"learning_rate": 6.30152085110927e-06,
"loss": 0.6486039161682129,
"step": 1162
},
{
"epoch": 2.460887949260042,
"grad_norm": 1.7713621854782104,
"learning_rate": 6.278936219780016e-06,
"loss": 0.5446457862854004,
"step": 1164
},
{
"epoch": 2.4651162790697674,
"grad_norm": 2.1408638954162598,
"learning_rate": 6.256375685497364e-06,
"loss": 0.9129477143287659,
"step": 1166
},
{
"epoch": 2.4693446088794926,
"grad_norm": 1.4627786874771118,
"learning_rate": 6.233839524072911e-06,
"loss": 0.9019725322723389,
"step": 1168
},
{
"epoch": 2.473572938689218,
"grad_norm": 3.729942798614502,
"learning_rate": 6.211328011020282e-06,
"loss": 0.5925071239471436,
"step": 1170
},
{
"epoch": 2.477801268498943,
"grad_norm": 13.676080703735352,
"learning_rate": 6.188841421551775e-06,
"loss": 0.6982713341712952,
"step": 1172
},
{
"epoch": 2.482029598308668,
"grad_norm": 1.7997795343399048,
"learning_rate": 6.166380030574974e-06,
"loss": 0.6115400791168213,
"step": 1174
},
{
"epoch": 2.4862579281183934,
"grad_norm": 1.1815468072891235,
"learning_rate": 6.1439441126894066e-06,
"loss": 0.5125085115432739,
"step": 1176
},
{
"epoch": 2.4904862579281186,
"grad_norm": 3.142874002456665,
"learning_rate": 6.12153394218319e-06,
"loss": 0.595373809337616,
"step": 1178
},
{
"epoch": 2.4947145877378434,
"grad_norm": 3.1829237937927246,
"learning_rate": 6.099149793029655e-06,
"loss": 0.4783131182193756,
"step": 1180
},
{
"epoch": 2.4989429175475686,
"grad_norm": 1.0863850116729736,
"learning_rate": 6.076791938884021e-06,
"loss": 0.28827065229415894,
"step": 1182
},
{
"epoch": 2.503171247357294,
"grad_norm": 3.749490261077881,
"learning_rate": 6.054460653080039e-06,
"loss": 0.08858740329742432,
"step": 1184
},
{
"epoch": 2.507399577167019,
"grad_norm": 2.9994473457336426,
"learning_rate": 6.032156208626644e-06,
"loss": 0.7518813610076904,
"step": 1186
},
{
"epoch": 2.511627906976744,
"grad_norm": 3.145364761352539,
"learning_rate": 6.009878878204636e-06,
"loss": 0.5571533441543579,
"step": 1188
},
{
"epoch": 2.5158562367864694,
"grad_norm": 6.573192119598389,
"learning_rate": 5.987628934163326e-06,
"loss": 0.6388422250747681,
"step": 1190
},
{
"epoch": 2.5200845665961946,
"grad_norm": 2.2647907733917236,
"learning_rate": 5.9654066485172255e-06,
"loss": 0.8509605526924133,
"step": 1192
},
{
"epoch": 2.52431289640592,
"grad_norm": 1.3656373023986816,
"learning_rate": 5.943212292942697e-06,
"loss": 0.6574472188949585,
"step": 1194
},
{
"epoch": 2.528541226215645,
"grad_norm": 5.842274188995361,
"learning_rate": 5.921046138774655e-06,
"loss": 0.8020691871643066,
"step": 1196
},
{
"epoch": 2.53276955602537,
"grad_norm": 1.8467926979064941,
"learning_rate": 5.898908457003243e-06,
"loss": 0.7512873411178589,
"step": 1198
},
{
"epoch": 2.536997885835095,
"grad_norm": 9.482233047485352,
"learning_rate": 5.876799518270515e-06,
"loss": 0.7120176553726196,
"step": 1200
},
{
"epoch": 2.54122621564482,
"grad_norm": 2.270634412765503,
"learning_rate": 5.854719592867123e-06,
"loss": 0.6714361906051636,
"step": 1202
},
{
"epoch": 2.5454545454545454,
"grad_norm": 8.194986343383789,
"learning_rate": 5.832668950729036e-06,
"loss": 0.2864822745323181,
"step": 1204
},
{
"epoch": 2.5496828752642706,
"grad_norm": 1.8582206964492798,
"learning_rate": 5.810647861434207e-06,
"loss": 0.8347625732421875,
"step": 1206
},
{
"epoch": 2.553911205073996,
"grad_norm": 1.62802255153656,
"learning_rate": 5.788656594199295e-06,
"loss": 0.7626522779464722,
"step": 1208
},
{
"epoch": 2.558139534883721,
"grad_norm": 1.6369847059249878,
"learning_rate": 5.766695417876385e-06,
"loss": 0.7506204843521118,
"step": 1210
},
{
"epoch": 2.5623678646934462,
"grad_norm": 2.289973258972168,
"learning_rate": 5.7447646009496814e-06,
"loss": 0.5922738313674927,
"step": 1212
},
{
"epoch": 2.5665961945031714,
"grad_norm": 1.5666455030441284,
"learning_rate": 5.722864411532224e-06,
"loss": 0.5874464511871338,
"step": 1214
},
{
"epoch": 2.570824524312896,
"grad_norm": 4.369112968444824,
"learning_rate": 5.700995117362633e-06,
"loss": 0.43863433599472046,
"step": 1216
},
{
"epoch": 2.5750528541226214,
"grad_norm": 5.651104927062988,
"learning_rate": 5.679156985801816e-06,
"loss": 0.505707859992981,
"step": 1218
},
{
"epoch": 2.5792811839323466,
"grad_norm": 6.160796642303467,
"learning_rate": 5.6573502838297005e-06,
"loss": 0.4164288640022278,
"step": 1220
},
{
"epoch": 2.583509513742072,
"grad_norm": 2.2909469604492188,
"learning_rate": 5.635575278041984e-06,
"loss": 0.5167214870452881,
"step": 1222
},
{
"epoch": 2.587737843551797,
"grad_norm": 4.068335056304932,
"learning_rate": 5.613832234646856e-06,
"loss": 0.5009158253669739,
"step": 1224
},
{
"epoch": 2.5919661733615222,
"grad_norm": 0.6781812906265259,
"learning_rate": 5.592121419461765e-06,
"loss": 0.29349908232688904,
"step": 1226
},
{
"epoch": 2.5961945031712474,
"grad_norm": 3.461779832839966,
"learning_rate": 5.570443097910141e-06,
"loss": 0.6309870481491089,
"step": 1228
},
{
"epoch": 2.6004228329809727,
"grad_norm": 3.883345603942871,
"learning_rate": 5.548797535018177e-06,
"loss": 0.6579529047012329,
"step": 1230
},
{
"epoch": 2.604651162790698,
"grad_norm": 9.5661039352417,
"learning_rate": 5.527184995411577e-06,
"loss": 0.5248879194259644,
"step": 1232
},
{
"epoch": 2.6088794926004226,
"grad_norm": 5.865969657897949,
"learning_rate": 5.505605743312321e-06,
"loss": 0.5632526278495789,
"step": 1234
},
{
"epoch": 2.613107822410148,
"grad_norm": 1.7783787250518799,
"learning_rate": 5.484060042535431e-06,
"loss": 0.608475923538208,
"step": 1236
},
{
"epoch": 2.617336152219873,
"grad_norm": 2.1970081329345703,
"learning_rate": 5.46254815648576e-06,
"loss": 0.7042860984802246,
"step": 1238
},
{
"epoch": 2.6215644820295982,
"grad_norm": 1.8460139036178589,
"learning_rate": 5.4410703481547604e-06,
"loss": 0.6008255481719971,
"step": 1240
},
{
"epoch": 2.6257928118393234,
"grad_norm": 1.5716763734817505,
"learning_rate": 5.419626880117257e-06,
"loss": 0.6618949174880981,
"step": 1242
},
{
"epoch": 2.6300211416490487,
"grad_norm": 3.4065911769866943,
"learning_rate": 5.3982180145282705e-06,
"loss": 0.6927484273910522,
"step": 1244
},
{
"epoch": 2.634249471458774,
"grad_norm": 14.799368858337402,
"learning_rate": 5.376844013119783e-06,
"loss": 0.4522789716720581,
"step": 1246
},
{
"epoch": 2.638477801268499,
"grad_norm": 3.741028070449829,
"learning_rate": 5.355505137197547e-06,
"loss": 0.45216238498687744,
"step": 1248
},
{
"epoch": 2.6427061310782243,
"grad_norm": 1.875082015991211,
"learning_rate": 5.334201647637893e-06,
"loss": 0.8255988359451294,
"step": 1250
},
{
"epoch": 2.646934460887949,
"grad_norm": 6.704344272613525,
"learning_rate": 5.312933804884545e-06,
"loss": 0.3596426844596863,
"step": 1252
},
{
"epoch": 2.6511627906976747,
"grad_norm": 5.3112874031066895,
"learning_rate": 5.291701868945415e-06,
"loss": 0.5540622472763062,
"step": 1254
},
{
"epoch": 2.6553911205073994,
"grad_norm": 1.176771640777588,
"learning_rate": 5.270506099389458e-06,
"loss": 0.699912428855896,
"step": 1256
},
{
"epoch": 2.6596194503171247,
"grad_norm": 1.2517683506011963,
"learning_rate": 5.249346755343472e-06,
"loss": 0.20151984691619873,
"step": 1258
},
{
"epoch": 2.66384778012685,
"grad_norm": 1.0356135368347168,
"learning_rate": 5.228224095488936e-06,
"loss": 0.3519669771194458,
"step": 1260
},
{
"epoch": 2.668076109936575,
"grad_norm": 4.993144512176514,
"learning_rate": 5.207138378058854e-06,
"loss": 0.5228662490844727,
"step": 1262
},
{
"epoch": 2.6723044397463003,
"grad_norm": 5.321039199829102,
"learning_rate": 5.1860898608345945e-06,
"loss": 0.467984676361084,
"step": 1264
},
{
"epoch": 2.6765327695560255,
"grad_norm": 4.311862468719482,
"learning_rate": 5.165078801142734e-06,
"loss": 0.23078812658786774,
"step": 1266
},
{
"epoch": 2.6807610993657507,
"grad_norm": 2.6508641242980957,
"learning_rate": 5.144105455851921e-06,
"loss": 0.44908571243286133,
"step": 1268
},
{
"epoch": 2.6849894291754755,
"grad_norm": 1.5310877561569214,
"learning_rate": 5.123170081369727e-06,
"loss": 0.7106070518493652,
"step": 1270
},
{
"epoch": 2.689217758985201,
"grad_norm": 6.155274391174316,
"learning_rate": 5.102272933639516e-06,
"loss": 0.5381072163581848,
"step": 1272
},
{
"epoch": 2.693446088794926,
"grad_norm": 0.5662539601325989,
"learning_rate": 5.081414268137314e-06,
"loss": 0.2816016972064972,
"step": 1274
},
{
"epoch": 2.697674418604651,
"grad_norm": 4.242975234985352,
"learning_rate": 5.060594339868688e-06,
"loss": 0.4097983241081238,
"step": 1276
},
{
"epoch": 2.7019027484143763,
"grad_norm": 5.412604808807373,
"learning_rate": 5.039813403365626e-06,
"loss": 0.6172899007797241,
"step": 1278
},
{
"epoch": 2.7061310782241015,
"grad_norm": 4.645112037658691,
"learning_rate": 5.019071712683423e-06,
"loss": 0.5603295564651489,
"step": 1280
},
{
"epoch": 2.7103594080338267,
"grad_norm": 1.5817734003067017,
"learning_rate": 4.998369521397585e-06,
"loss": 0.8464901447296143,
"step": 1282
},
{
"epoch": 2.714587737843552,
"grad_norm": 0.9378266334533691,
"learning_rate": 4.977707082600714e-06,
"loss": 0.602013885974884,
"step": 1284
},
{
"epoch": 2.718816067653277,
"grad_norm": 0.3795956075191498,
"learning_rate": 4.957084648899424e-06,
"loss": 0.744926929473877,
"step": 1286
},
{
"epoch": 2.723044397463002,
"grad_norm": 1.9880757331848145,
"learning_rate": 4.936502472411261e-06,
"loss": 0.3527957797050476,
"step": 1288
},
{
"epoch": 2.7272727272727275,
"grad_norm": 1.7355777025222778,
"learning_rate": 4.9159608047615925e-06,
"loss": 0.7091801166534424,
"step": 1290
},
{
"epoch": 2.7315010570824523,
"grad_norm": 1.4596855640411377,
"learning_rate": 4.895459897080558e-06,
"loss": 0.807796835899353,
"step": 1292
},
{
"epoch": 2.7357293868921775,
"grad_norm": 2.226590633392334,
"learning_rate": 4.875000000000002e-06,
"loss": 0.8595592975616455,
"step": 1294
},
{
"epoch": 2.7399577167019027,
"grad_norm": 4.763701915740967,
"learning_rate": 4.854581363650374e-06,
"loss": 0.2207929491996765,
"step": 1296
},
{
"epoch": 2.744186046511628,
"grad_norm": 1.5515803098678589,
"learning_rate": 4.834204237657707e-06,
"loss": 0.45758575201034546,
"step": 1298
},
{
"epoch": 2.748414376321353,
"grad_norm": 1.8392975330352783,
"learning_rate": 4.813868871140562e-06,
"loss": 0.5696815252304077,
"step": 1300
},
{
"epoch": 2.7526427061310783,
"grad_norm": 2.7251267433166504,
"learning_rate": 4.793575512706958e-06,
"loss": 0.548314094543457,
"step": 1302
},
{
"epoch": 2.7568710359408035,
"grad_norm": 0.9378683567047119,
"learning_rate": 4.7733244104513524e-06,
"loss": 0.4568610191345215,
"step": 1304
},
{
"epoch": 2.7610993657505283,
"grad_norm": 7.682252407073975,
"learning_rate": 4.753115811951612e-06,
"loss": 0.6801155805587769,
"step": 1306
},
{
"epoch": 2.765327695560254,
"grad_norm": 2.429574489593506,
"learning_rate": 4.732949964265975e-06,
"loss": 0.7148873805999756,
"step": 1308
},
{
"epoch": 2.7695560253699787,
"grad_norm": 2.616790533065796,
"learning_rate": 4.71282711393002e-06,
"loss": 0.7284020781517029,
"step": 1310
},
{
"epoch": 2.773784355179704,
"grad_norm": 1.7224617004394531,
"learning_rate": 4.692747506953686e-06,
"loss": 0.6191799640655518,
"step": 1312
},
{
"epoch": 2.778012684989429,
"grad_norm": 1.9497318267822266,
"learning_rate": 4.672711388818235e-06,
"loss": 0.34431034326553345,
"step": 1314
},
{
"epoch": 2.7822410147991543,
"grad_norm": 2.1734511852264404,
"learning_rate": 4.652719004473257e-06,
"loss": 0.7559425830841064,
"step": 1316
},
{
"epoch": 2.7864693446088795,
"grad_norm": 0.6739775538444519,
"learning_rate": 4.632770598333681e-06,
"loss": 0.5308979749679565,
"step": 1318
},
{
"epoch": 2.7906976744186047,
"grad_norm": 3.110369920730591,
"learning_rate": 4.612866414276793e-06,
"loss": 0.7475970387458801,
"step": 1320
},
{
"epoch": 2.79492600422833,
"grad_norm": 3.6263904571533203,
"learning_rate": 4.593006695639241e-06,
"loss": 0.3352441191673279,
"step": 1322
},
{
"epoch": 2.7991543340380547,
"grad_norm": 2.3171989917755127,
"learning_rate": 4.573191685214053e-06,
"loss": 0.7618209719657898,
"step": 1324
},
{
"epoch": 2.8033826638477803,
"grad_norm": 4.778399467468262,
"learning_rate": 4.5534216252477015e-06,
"loss": 0.19848856329917908,
"step": 1326
},
{
"epoch": 2.807610993657505,
"grad_norm": 9.885106086730957,
"learning_rate": 4.533696757437112e-06,
"loss": 0.31195956468582153,
"step": 1328
},
{
"epoch": 2.8118393234672303,
"grad_norm": 4.158289909362793,
"learning_rate": 4.514017322926712e-06,
"loss": 0.4557163119316101,
"step": 1330
},
{
"epoch": 2.8160676532769555,
"grad_norm": 1.9819703102111816,
"learning_rate": 4.494383562305501e-06,
"loss": 0.6762011051177979,
"step": 1332
},
{
"epoch": 2.8202959830866807,
"grad_norm": 1.948534369468689,
"learning_rate": 4.474795715604091e-06,
"loss": 0.5135953426361084,
"step": 1334
},
{
"epoch": 2.824524312896406,
"grad_norm": 1.2752959728240967,
"learning_rate": 4.45525402229177e-06,
"loss": 0.8080368041992188,
"step": 1336
},
{
"epoch": 2.828752642706131,
"grad_norm": 1.7740074396133423,
"learning_rate": 4.435758721273596e-06,
"loss": 0.805374264717102,
"step": 1338
},
{
"epoch": 2.8329809725158563,
"grad_norm": 1.5627951622009277,
"learning_rate": 4.416310050887456e-06,
"loss": 0.6002947092056274,
"step": 1340
},
{
"epoch": 2.8372093023255816,
"grad_norm": 2.2222254276275635,
"learning_rate": 4.396908248901157e-06,
"loss": 0.7086814045906067,
"step": 1342
},
{
"epoch": 2.8414376321353068,
"grad_norm": 1.4105923175811768,
"learning_rate": 4.377553552509526e-06,
"loss": 0.4943678379058838,
"step": 1344
},
{
"epoch": 2.8456659619450315,
"grad_norm": 3.5456433296203613,
"learning_rate": 4.3582461983315e-06,
"loss": 0.5853161811828613,
"step": 1346
},
{
"epoch": 2.8498942917547567,
"grad_norm": 1.4400972127914429,
"learning_rate": 4.33898642240724e-06,
"loss": 0.7183468341827393,
"step": 1348
},
{
"epoch": 2.854122621564482,
"grad_norm": 13.601333618164062,
"learning_rate": 4.319774460195247e-06,
"loss": 0.43445509672164917,
"step": 1350
},
{
"epoch": 2.858350951374207,
"grad_norm": 2.2455263137817383,
"learning_rate": 4.300610546569476e-06,
"loss": 0.720789909362793,
"step": 1352
},
{
"epoch": 2.8625792811839323,
"grad_norm": 1.1618309020996094,
"learning_rate": 4.281494915816471e-06,
"loss": 0.7108666896820068,
"step": 1354
},
{
"epoch": 2.8668076109936576,
"grad_norm": 7.198761463165283,
"learning_rate": 4.262427801632496e-06,
"loss": 0.306668758392334,
"step": 1356
},
{
"epoch": 2.8710359408033828,
"grad_norm": 1.7793060541152954,
"learning_rate": 4.2434094371206845e-06,
"loss": 0.7454710006713867,
"step": 1358
},
{
"epoch": 2.875264270613108,
"grad_norm": 2.858522653579712,
"learning_rate": 4.2244400547881815e-06,
"loss": 0.46587949991226196,
"step": 1360
},
{
"epoch": 2.879492600422833,
"grad_norm": 5.804128646850586,
"learning_rate": 4.205519886543308e-06,
"loss": 0.5216749310493469,
"step": 1362
},
{
"epoch": 2.883720930232558,
"grad_norm": 1.7028459310531616,
"learning_rate": 4.186649163692724e-06,
"loss": 0.8213293552398682,
"step": 1364
},
{
"epoch": 2.887949260042283,
"grad_norm": 2.4620273113250732,
"learning_rate": 4.167828116938596e-06,
"loss": 0.6673083305358887,
"step": 1366
},
{
"epoch": 2.8921775898520083,
"grad_norm": 1.4428051710128784,
"learning_rate": 4.149056976375781e-06,
"loss": 0.6986818313598633,
"step": 1368
},
{
"epoch": 2.8964059196617336,
"grad_norm": 1.893155813217163,
"learning_rate": 4.130335971489016e-06,
"loss": 0.7646907567977905,
"step": 1370
},
{
"epoch": 2.9006342494714588,
"grad_norm": 1.2903612852096558,
"learning_rate": 4.111665331150106e-06,
"loss": 0.7394033670425415,
"step": 1372
},
{
"epoch": 2.904862579281184,
"grad_norm": 1.2461528778076172,
"learning_rate": 4.093045283615127e-06,
"loss": 0.6628772020339966,
"step": 1374
},
{
"epoch": 2.909090909090909,
"grad_norm": 1.64785635471344,
"learning_rate": 4.074476056521649e-06,
"loss": 0.7799428701400757,
"step": 1376
},
{
"epoch": 2.9133192389006344,
"grad_norm": 3.3484251499176025,
"learning_rate": 4.055957876885926e-06,
"loss": 0.4393528401851654,
"step": 1378
},
{
"epoch": 2.9175475687103596,
"grad_norm": 2.3519139289855957,
"learning_rate": 4.037490971100145e-06,
"loss": 0.5560283660888672,
"step": 1380
},
{
"epoch": 2.9217758985200843,
"grad_norm": 3.2409229278564453,
"learning_rate": 4.019075564929654e-06,
"loss": 0.9184545278549194,
"step": 1382
},
{
"epoch": 2.92600422832981,
"grad_norm": 4.495751857757568,
"learning_rate": 4.000711883510186e-06,
"loss": 0.5940873026847839,
"step": 1384
},
{
"epoch": 2.9302325581395348,
"grad_norm": 1.513631820678711,
"learning_rate": 3.982400151345126e-06,
"loss": 0.5513443946838379,
"step": 1386
},
{
"epoch": 2.93446088794926,
"grad_norm": 1.6095917224884033,
"learning_rate": 3.964140592302755e-06,
"loss": 0.6538087129592896,
"step": 1388
},
{
"epoch": 2.938689217758985,
"grad_norm": 0.6691470742225647,
"learning_rate": 3.9459334296135236e-06,
"loss": 0.46507909893989563,
"step": 1390
},
{
"epoch": 2.9429175475687104,
"grad_norm": 2.0764143466949463,
"learning_rate": 3.927778885867304e-06,
"loss": 0.9463499784469604,
"step": 1392
},
{
"epoch": 2.9471458773784356,
"grad_norm": 0.7023881673812866,
"learning_rate": 3.909677183010689e-06,
"loss": 0.172732412815094,
"step": 1394
},
{
"epoch": 2.951374207188161,
"grad_norm": 1.7051056623458862,
"learning_rate": 3.891628542344273e-06,
"loss": 0.5550211071968079,
"step": 1396
},
{
"epoch": 2.955602536997886,
"grad_norm": 4.343619346618652,
"learning_rate": 3.873633184519934e-06,
"loss": 0.8420608043670654,
"step": 1398
},
{
"epoch": 2.9598308668076108,
"grad_norm": 4.121436595916748,
"learning_rate": 3.855691329538151e-06,
"loss": 0.8600051403045654,
"step": 1400
},
{
"epoch": 2.9640591966173364,
"grad_norm": 4.33802604675293,
"learning_rate": 3.837803196745316e-06,
"loss": 0.5907847881317139,
"step": 1402
},
{
"epoch": 2.968287526427061,
"grad_norm": 2.516199827194214,
"learning_rate": 3.819969004831031e-06,
"loss": 0.38888847827911377,
"step": 1404
},
{
"epoch": 2.9725158562367864,
"grad_norm": 5.375875949859619,
"learning_rate": 3.8021889718254596e-06,
"loss": 0.7636460065841675,
"step": 1406
},
{
"epoch": 2.9767441860465116,
"grad_norm": 1.3979852199554443,
"learning_rate": 3.7844633150966528e-06,
"loss": 0.5576034784317017,
"step": 1408
},
{
"epoch": 2.980972515856237,
"grad_norm": 4.267658233642578,
"learning_rate": 3.766792251347883e-06,
"loss": 0.537714421749115,
"step": 1410
},
{
"epoch": 2.985200845665962,
"grad_norm": 1.726967215538025,
"learning_rate": 3.749175996615e-06,
"loss": 0.722823441028595,
"step": 1412
},
{
"epoch": 2.989429175475687,
"grad_norm": 1.251090407371521,
"learning_rate": 3.7316147662638003e-06,
"loss": 0.43247348070144653,
"step": 1414
},
{
"epoch": 2.9936575052854124,
"grad_norm": 3.434035301208496,
"learning_rate": 3.7141087749873817e-06,
"loss": 0.6234498023986816,
"step": 1416
},
{
"epoch": 2.997885835095137,
"grad_norm": 1.6687880754470825,
"learning_rate": 3.696658236803512e-06,
"loss": 0.3949212431907654,
"step": 1418
},
{
"epoch": 3.0021141649048624,
"grad_norm": 1.6642099618911743,
"learning_rate": 3.6792633650520413e-06,
"loss": 0.6867222189903259,
"step": 1420
},
{
"epoch": 3.0063424947145876,
"grad_norm": 2.6655778884887695,
"learning_rate": 3.6619243723922596e-06,
"loss": 0.29861384630203247,
"step": 1422
},
{
"epoch": 3.010570824524313,
"grad_norm": 13.024568557739258,
"learning_rate": 3.644641470800321e-06,
"loss": 0.7144566774368286,
"step": 1424
},
{
"epoch": 3.014799154334038,
"grad_norm": 4.102027416229248,
"learning_rate": 3.6274148715666392e-06,
"loss": 0.3139229416847229,
"step": 1426
},
{
"epoch": 3.019027484143763,
"grad_norm": 3.01511287689209,
"learning_rate": 3.6102447852933117e-06,
"loss": 0.4534086287021637,
"step": 1428
},
{
"epoch": 3.0232558139534884,
"grad_norm": 2.0610299110412598,
"learning_rate": 3.593131421891541e-06,
"loss": 0.5477634072303772,
"step": 1430
},
{
"epoch": 3.0274841437632136,
"grad_norm": 1.6937140226364136,
"learning_rate": 3.5760749905790703e-06,
"loss": 0.49359893798828125,
"step": 1432
},
{
"epoch": 3.031712473572939,
"grad_norm": 2.3263511657714844,
"learning_rate": 3.559075699877624e-06,
"loss": 0.5352256298065186,
"step": 1434
},
{
"epoch": 3.035940803382664,
"grad_norm": 6.835305690765381,
"learning_rate": 3.5421337576103625e-06,
"loss": 0.2217904031276703,
"step": 1436
},
{
"epoch": 3.040169133192389,
"grad_norm": 0.5178420543670654,
"learning_rate": 3.525249370899335e-06,
"loss": 0.0617043673992157,
"step": 1438
},
{
"epoch": 3.044397463002114,
"grad_norm": 2.201230525970459,
"learning_rate": 3.5084227461629525e-06,
"loss": 0.5838777422904968,
"step": 1440
},
{
"epoch": 3.048625792811839,
"grad_norm": 1.9562621116638184,
"learning_rate": 3.4916540891134636e-06,
"loss": 0.31589028239250183,
"step": 1442
},
{
"epoch": 3.0528541226215644,
"grad_norm": 7.107705116271973,
"learning_rate": 3.4749436047544367e-06,
"loss": 0.132436603307724,
"step": 1444
},
{
"epoch": 3.0570824524312896,
"grad_norm": 2.205892562866211,
"learning_rate": 3.458291497378257e-06,
"loss": 0.30102184414863586,
"step": 1446
},
{
"epoch": 3.061310782241015,
"grad_norm": 66.15872955322266,
"learning_rate": 3.4416979705636272e-06,
"loss": 0.4792563319206238,
"step": 1448
},
{
"epoch": 3.06553911205074,
"grad_norm": 1.6659890413284302,
"learning_rate": 3.4251632271730805e-06,
"loss": 0.250087171792984,
"step": 1450
},
{
"epoch": 3.0697674418604652,
"grad_norm": 4.044705390930176,
"learning_rate": 3.4086874693504953e-06,
"loss": 0.42401123046875,
"step": 1452
},
{
"epoch": 3.0739957716701904,
"grad_norm": 1.7333321571350098,
"learning_rate": 3.392270898518632e-06,
"loss": 0.4892985224723816,
"step": 1454
},
{
"epoch": 3.0782241014799157,
"grad_norm": 1.4521524906158447,
"learning_rate": 3.375913715376664e-06,
"loss": 0.13367274403572083,
"step": 1456
},
{
"epoch": 3.0824524312896404,
"grad_norm": 1.5507392883300781,
"learning_rate": 3.3596161198977267e-06,
"loss": 0.34912851452827454,
"step": 1458
},
{
"epoch": 3.0866807610993656,
"grad_norm": 1.9308871030807495,
"learning_rate": 3.343378311326473e-06,
"loss": 0.38985931873321533,
"step": 1460
},
{
"epoch": 3.090909090909091,
"grad_norm": 2.1055166721343994,
"learning_rate": 3.327200488176636e-06,
"loss": 0.4916480481624603,
"step": 1462
},
{
"epoch": 3.095137420718816,
"grad_norm": 1.6326617002487183,
"learning_rate": 3.3110828482286036e-06,
"loss": 0.4394315779209137,
"step": 1464
},
{
"epoch": 3.0993657505285412,
"grad_norm": 1.7664871215820312,
"learning_rate": 3.2950255885269988e-06,
"loss": 0.3395063877105713,
"step": 1466
},
{
"epoch": 3.1035940803382664,
"grad_norm": 2.11129093170166,
"learning_rate": 3.279028905378274e-06,
"loss": 0.5383762121200562,
"step": 1468
},
{
"epoch": 3.1078224101479917,
"grad_norm": 1.5842963457107544,
"learning_rate": 3.2630929943483033e-06,
"loss": 0.480863094329834,
"step": 1470
},
{
"epoch": 3.112050739957717,
"grad_norm": 3.5745162963867188,
"learning_rate": 3.2472180502600095e-06,
"loss": 0.4862865209579468,
"step": 1472
},
{
"epoch": 3.116279069767442,
"grad_norm": 1.539161205291748,
"learning_rate": 3.2314042671909547e-06,
"loss": 0.507552981376648,
"step": 1474
},
{
"epoch": 3.120507399577167,
"grad_norm": 1.3328593969345093,
"learning_rate": 3.2156518384709914e-06,
"loss": 0.46964654326438904,
"step": 1476
},
{
"epoch": 3.124735729386892,
"grad_norm": 5.218214511871338,
"learning_rate": 3.1999609566798935e-06,
"loss": 0.12817713618278503,
"step": 1478
},
{
"epoch": 3.1289640591966172,
"grad_norm": 1.8787182569503784,
"learning_rate": 3.1843318136449904e-06,
"loss": 0.42655909061431885,
"step": 1480
},
{
"epoch": 3.1331923890063424,
"grad_norm": 5.574677467346191,
"learning_rate": 3.168764600438834e-06,
"loss": 0.3973861336708069,
"step": 1482
},
{
"epoch": 3.1374207188160677,
"grad_norm": 0.5539373159408569,
"learning_rate": 3.1532595073768675e-06,
"loss": 0.0943363606929779,
"step": 1484
},
{
"epoch": 3.141649048625793,
"grad_norm": 9.655427932739258,
"learning_rate": 3.1378167240150746e-06,
"loss": 0.5281685590744019,
"step": 1486
},
{
"epoch": 3.145877378435518,
"grad_norm": 2.5484557151794434,
"learning_rate": 3.122436439147688e-06,
"loss": 0.7694233655929565,
"step": 1488
},
{
"epoch": 3.1501057082452433,
"grad_norm": 4.505038738250732,
"learning_rate": 3.1071188408048715e-06,
"loss": 0.4504562020301819,
"step": 1490
},
{
"epoch": 3.1543340380549685,
"grad_norm": 3.3813390731811523,
"learning_rate": 3.091864116250421e-06,
"loss": 0.38999781012535095,
"step": 1492
},
{
"epoch": 3.1585623678646932,
"grad_norm": 1.7855579853057861,
"learning_rate": 3.076672451979466e-06,
"loss": 0.2507738471031189,
"step": 1494
},
{
"epoch": 3.1627906976744184,
"grad_norm": 0.26156097650527954,
"learning_rate": 3.061544033716211e-06,
"loss": 0.3308458924293518,
"step": 1496
},
{
"epoch": 3.1670190274841437,
"grad_norm": 1.5556107759475708,
"learning_rate": 3.0464790464116474e-06,
"loss": 0.5818713903427124,
"step": 1498
},
{
"epoch": 3.171247357293869,
"grad_norm": 2.0821709632873535,
"learning_rate": 3.0314776742412915e-06,
"loss": 0.5305293202400208,
"step": 1500
},
{
"epoch": 3.175475687103594,
"grad_norm": 4.033331871032715,
"learning_rate": 3.0165401006029515e-06,
"loss": 0.14770884811878204,
"step": 1502
},
{
"epoch": 3.1797040169133193,
"grad_norm": 1.9315921068191528,
"learning_rate": 3.0016665081144643e-06,
"loss": 0.37355494499206543,
"step": 1504
},
{
"epoch": 3.1839323467230445,
"grad_norm": 1.3936494588851929,
"learning_rate": 2.9868570786114753e-06,
"loss": 0.36658424139022827,
"step": 1506
},
{
"epoch": 3.1881606765327697,
"grad_norm": 3.391876459121704,
"learning_rate": 2.9721119931452044e-06,
"loss": 0.222508043050766,
"step": 1508
},
{
"epoch": 3.192389006342495,
"grad_norm": 4.5538225173950195,
"learning_rate": 2.9574314319802505e-06,
"loss": 0.08852885663509369,
"step": 1510
},
{
"epoch": 3.1966173361522197,
"grad_norm": 1.435149073600769,
"learning_rate": 2.942815574592373e-06,
"loss": 0.04714357852935791,
"step": 1512
},
{
"epoch": 3.200845665961945,
"grad_norm": 3.4955222606658936,
"learning_rate": 2.928264599666294e-06,
"loss": 0.5409808158874512,
"step": 1514
},
{
"epoch": 3.20507399577167,
"grad_norm": 2.0416336059570312,
"learning_rate": 2.9137786850935336e-06,
"loss": 0.49242472648620605,
"step": 1516
},
{
"epoch": 3.2093023255813953,
"grad_norm": 1.5005817413330078,
"learning_rate": 2.8993580079702178e-06,
"loss": 0.3337843418121338,
"step": 1518
},
{
"epoch": 3.2135306553911205,
"grad_norm": 1.5125510692596436,
"learning_rate": 2.8850027445949116e-06,
"loss": 0.557496190071106,
"step": 1520
},
{
"epoch": 3.2177589852008457,
"grad_norm": 3.043395757675171,
"learning_rate": 2.8707130704664836e-06,
"loss": 0.5054546594619751,
"step": 1522
},
{
"epoch": 3.221987315010571,
"grad_norm": 5.112667083740234,
"learning_rate": 2.856489160281939e-06,
"loss": 0.5476666688919067,
"step": 1524
},
{
"epoch": 3.226215644820296,
"grad_norm": 1.8973793983459473,
"learning_rate": 2.842331187934293e-06,
"loss": 0.4390944838523865,
"step": 1526
},
{
"epoch": 3.2304439746300213,
"grad_norm": 6.383437156677246,
"learning_rate": 2.828239326510448e-06,
"loss": 0.06960128247737885,
"step": 1528
},
{
"epoch": 3.234672304439746,
"grad_norm": 3.420788526535034,
"learning_rate": 2.814213748289069e-06,
"loss": 0.6601115465164185,
"step": 1530
},
{
"epoch": 3.2389006342494713,
"grad_norm": 2.9068996906280518,
"learning_rate": 2.8002546247384868e-06,
"loss": 0.4413766860961914,
"step": 1532
},
{
"epoch": 3.2431289640591965,
"grad_norm": 5.704627513885498,
"learning_rate": 2.786362126514594e-06,
"loss": 0.4749358892440796,
"step": 1534
},
{
"epoch": 3.2473572938689217,
"grad_norm": 2.5792603492736816,
"learning_rate": 2.7725364234587617e-06,
"loss": 0.33815979957580566,
"step": 1536
},
{
"epoch": 3.251585623678647,
"grad_norm": 2.1992955207824707,
"learning_rate": 2.7587776845957662e-06,
"loss": 0.30135706067085266,
"step": 1538
},
{
"epoch": 3.255813953488372,
"grad_norm": 2.5813398361206055,
"learning_rate": 2.7450860781317173e-06,
"loss": 0.29713791608810425,
"step": 1540
},
{
"epoch": 3.2600422832980973,
"grad_norm": 4.867682933807373,
"learning_rate": 2.7314617714520043e-06,
"loss": 0.7522240877151489,
"step": 1542
},
{
"epoch": 3.2642706131078225,
"grad_norm": 3.4543728828430176,
"learning_rate": 2.7179049311192504e-06,
"loss": 0.5658772587776184,
"step": 1544
},
{
"epoch": 3.2684989429175477,
"grad_norm": 5.230175018310547,
"learning_rate": 2.7044157228712737e-06,
"loss": 0.5096595287322998,
"step": 1546
},
{
"epoch": 3.2727272727272725,
"grad_norm": 0.8176948428153992,
"learning_rate": 2.690994311619067e-06,
"loss": 0.06993258744478226,
"step": 1548
},
{
"epoch": 3.276955602536998,
"grad_norm": 3.6103427410125732,
"learning_rate": 2.6776408614447735e-06,
"loss": 0.605951189994812,
"step": 1550
},
{
"epoch": 3.281183932346723,
"grad_norm": 2.1291911602020264,
"learning_rate": 2.664355535599685e-06,
"loss": 0.4419100880622864,
"step": 1552
},
{
"epoch": 3.285412262156448,
"grad_norm": 0.44995903968811035,
"learning_rate": 2.651138496502252e-06,
"loss": 0.05339962989091873,
"step": 1554
},
{
"epoch": 3.2896405919661733,
"grad_norm": 2.977468729019165,
"learning_rate": 2.6379899057360827e-06,
"loss": 0.643664538860321,
"step": 1556
},
{
"epoch": 3.2938689217758985,
"grad_norm": 2.5796053409576416,
"learning_rate": 2.6249099240479836e-06,
"loss": 0.2783702611923218,
"step": 1558
},
{
"epoch": 3.2980972515856237,
"grad_norm": 4.654566287994385,
"learning_rate": 2.611898711345988e-06,
"loss": 0.7086867690086365,
"step": 1560
},
{
"epoch": 3.302325581395349,
"grad_norm": 4.6553544998168945,
"learning_rate": 2.598956426697395e-06,
"loss": 0.48968005180358887,
"step": 1562
},
{
"epoch": 3.306553911205074,
"grad_norm": 1.7385183572769165,
"learning_rate": 2.586083228326834e-06,
"loss": 0.28635382652282715,
"step": 1564
},
{
"epoch": 3.3107822410147993,
"grad_norm": 1.9122589826583862,
"learning_rate": 2.5732792736143315e-06,
"loss": 0.38382118940353394,
"step": 1566
},
{
"epoch": 3.3150105708245245,
"grad_norm": 0.32408830523490906,
"learning_rate": 2.5605447190933735e-06,
"loss": 0.0730479285120964,
"step": 1568
},
{
"epoch": 3.3192389006342493,
"grad_norm": 4.453859806060791,
"learning_rate": 2.547879720449005e-06,
"loss": 0.3159065842628479,
"step": 1570
},
{
"epoch": 3.3234672304439745,
"grad_norm": 8.307226181030273,
"learning_rate": 2.5352844325159256e-06,
"loss": 0.418887734413147,
"step": 1572
},
{
"epoch": 3.3276955602536997,
"grad_norm": 5.4259114265441895,
"learning_rate": 2.522759009276592e-06,
"loss": 0.11129888892173767,
"step": 1574
},
{
"epoch": 3.331923890063425,
"grad_norm": 8.971489906311035,
"learning_rate": 2.51030360385933e-06,
"loss": 0.21846887469291687,
"step": 1576
},
{
"epoch": 3.33615221987315,
"grad_norm": 3.2399227619171143,
"learning_rate": 2.4979183685364768e-06,
"loss": 0.3414691686630249,
"step": 1578
},
{
"epoch": 3.3403805496828753,
"grad_norm": 1.4801366329193115,
"learning_rate": 2.485603454722514e-06,
"loss": 0.23538810014724731,
"step": 1580
},
{
"epoch": 3.3446088794926006,
"grad_norm": 2.040708065032959,
"learning_rate": 2.4733590129722057e-06,
"loss": 0.49476367235183716,
"step": 1582
},
{
"epoch": 3.3488372093023258,
"grad_norm": 2.6494998931884766,
"learning_rate": 2.4611851929787727e-06,
"loss": 0.45221155881881714,
"step": 1584
},
{
"epoch": 3.353065539112051,
"grad_norm": 2.2645339965820312,
"learning_rate": 2.449082143572057e-06,
"loss": 0.38120681047439575,
"step": 1586
},
{
"epoch": 3.3572938689217757,
"grad_norm": 5.372982025146484,
"learning_rate": 2.4370500127167007e-06,
"loss": 0.5445793867111206,
"step": 1588
},
{
"epoch": 3.361522198731501,
"grad_norm": 1.7980395555496216,
"learning_rate": 2.425088947510333e-06,
"loss": 0.08176328986883163,
"step": 1590
},
{
"epoch": 3.365750528541226,
"grad_norm": 1.7179235219955444,
"learning_rate": 2.4131990941817855e-06,
"loss": 0.522789716720581,
"step": 1592
},
{
"epoch": 3.3699788583509513,
"grad_norm": 1.3470094203948975,
"learning_rate": 2.4013805980892935e-06,
"loss": 0.12738136947155,
"step": 1594
},
{
"epoch": 3.3742071881606766,
"grad_norm": 1.9271633625030518,
"learning_rate": 2.3896336037187134e-06,
"loss": 0.37803083658218384,
"step": 1596
},
{
"epoch": 3.3784355179704018,
"grad_norm": 12.261544227600098,
"learning_rate": 2.3779582546817767e-06,
"loss": 0.27141934633255005,
"step": 1598
},
{
"epoch": 3.382663847780127,
"grad_norm": 2.1056787967681885,
"learning_rate": 2.366354693714316e-06,
"loss": 0.234742671251297,
"step": 1600
},
{
"epoch": 3.386892177589852,
"grad_norm": 12.533198356628418,
"learning_rate": 2.3548230626745206e-06,
"loss": 0.34299901127815247,
"step": 1602
},
{
"epoch": 3.3911205073995774,
"grad_norm": 2.656111478805542,
"learning_rate": 2.3433635025412175e-06,
"loss": 0.5765751004219055,
"step": 1604
},
{
"epoch": 3.395348837209302,
"grad_norm": 1.5774534940719604,
"learning_rate": 2.3319761534121324e-06,
"loss": 0.45786333084106445,
"step": 1606
},
{
"epoch": 3.3995771670190273,
"grad_norm": 3.5536763668060303,
"learning_rate": 2.320661154502184e-06,
"loss": 0.23404037952423096,
"step": 1608
},
{
"epoch": 3.4038054968287526,
"grad_norm": 1.8982465267181396,
"learning_rate": 2.309418644141778e-06,
"loss": 0.5482504963874817,
"step": 1610
},
{
"epoch": 3.4080338266384778,
"grad_norm": 4.409867763519287,
"learning_rate": 2.298248759775121e-06,
"loss": 0.5976965427398682,
"step": 1612
},
{
"epoch": 3.412262156448203,
"grad_norm": 4.288993835449219,
"learning_rate": 2.287151637958538e-06,
"loss": 0.41971904039382935,
"step": 1614
},
{
"epoch": 3.416490486257928,
"grad_norm": 3.2369632720947266,
"learning_rate": 2.276127414358801e-06,
"loss": 0.5961102247238159,
"step": 1616
},
{
"epoch": 3.4207188160676534,
"grad_norm": 2.226705312728882,
"learning_rate": 2.265176223751472e-06,
"loss": 0.5602580308914185,
"step": 1618
},
{
"epoch": 3.4249471458773786,
"grad_norm": 1.930182695388794,
"learning_rate": 2.254298200019256e-06,
"loss": 0.4523662328720093,
"step": 1620
},
{
"epoch": 3.429175475687104,
"grad_norm": 7.945974826812744,
"learning_rate": 2.243493476150364e-06,
"loss": 0.2813890278339386,
"step": 1622
},
{
"epoch": 3.4334038054968286,
"grad_norm": 2.2771987915039062,
"learning_rate": 2.2327621842368864e-06,
"loss": 0.5364484190940857,
"step": 1624
},
{
"epoch": 3.4376321353065538,
"grad_norm": 2.1954996585845947,
"learning_rate": 2.222104455473177e-06,
"loss": 0.4502740502357483,
"step": 1626
},
{
"epoch": 3.441860465116279,
"grad_norm": 1.8870927095413208,
"learning_rate": 2.2115204201542526e-06,
"loss": 0.6255757808685303,
"step": 1628
},
{
"epoch": 3.446088794926004,
"grad_norm": 10.188657760620117,
"learning_rate": 2.2010102076741976e-06,
"loss": 0.25465184450149536,
"step": 1630
},
{
"epoch": 3.4503171247357294,
"grad_norm": 2.0800275802612305,
"learning_rate": 2.190573946524581e-06,
"loss": 0.3574382960796356,
"step": 1632
},
{
"epoch": 3.4545454545454546,
"grad_norm": 2.158536434173584,
"learning_rate": 2.1802117642928894e-06,
"loss": 0.42494529485702515,
"step": 1634
},
{
"epoch": 3.45877378435518,
"grad_norm": 2.0724036693573,
"learning_rate": 2.169923787660963e-06,
"loss": 0.5070405006408691,
"step": 1636
},
{
"epoch": 3.463002114164905,
"grad_norm": 0.3990979492664337,
"learning_rate": 2.1597101424034486e-06,
"loss": 0.21116462349891663,
"step": 1638
},
{
"epoch": 3.46723044397463,
"grad_norm": 2.11328125,
"learning_rate": 2.149570953386265e-06,
"loss": 0.35788196325302124,
"step": 1640
},
{
"epoch": 3.471458773784355,
"grad_norm": 0.6507092118263245,
"learning_rate": 2.1395063445650702e-06,
"loss": 0.094949871301651,
"step": 1642
},
{
"epoch": 3.47568710359408,
"grad_norm": 1.9196699857711792,
"learning_rate": 2.1295164389837494e-06,
"loss": 0.4157150983810425,
"step": 1644
},
{
"epoch": 3.4799154334038054,
"grad_norm": 2.5179502964019775,
"learning_rate": 2.1196013587729142e-06,
"loss": 0.5065787434577942,
"step": 1646
},
{
"epoch": 3.4841437632135306,
"grad_norm": 0.19388727843761444,
"learning_rate": 2.109761225148402e-06,
"loss": 0.3387356400489807,
"step": 1648
},
{
"epoch": 3.488372093023256,
"grad_norm": 1.8885760307312012,
"learning_rate": 2.0999961584098e-06,
"loss": 0.2609141767024994,
"step": 1650
},
{
"epoch": 3.492600422832981,
"grad_norm": 10.80978012084961,
"learning_rate": 2.0903062779389724e-06,
"loss": 0.311320424079895,
"step": 1652
},
{
"epoch": 3.496828752642706,
"grad_norm": 5.009494304656982,
"learning_rate": 2.0806917021986e-06,
"loss": 0.504407525062561,
"step": 1654
},
{
"epoch": 3.5010570824524314,
"grad_norm": 2.3830902576446533,
"learning_rate": 2.071152548730737e-06,
"loss": 0.4099867045879364,
"step": 1656
},
{
"epoch": 3.5052854122621566,
"grad_norm": 0.48838508129119873,
"learning_rate": 2.061688934155364e-06,
"loss": 0.290943443775177,
"step": 1658
},
{
"epoch": 3.5095137420718814,
"grad_norm": 4.095788478851318,
"learning_rate": 2.052300974168971e-06,
"loss": 0.4981149733066559,
"step": 1660
},
{
"epoch": 3.513742071881607,
"grad_norm": 8.41342544555664,
"learning_rate": 2.042988783543144e-06,
"loss": 0.2673397660255432,
"step": 1662
},
{
"epoch": 3.517970401691332,
"grad_norm": 1.7386679649353027,
"learning_rate": 2.0337524761231517e-06,
"loss": 0.3966340720653534,
"step": 1664
},
{
"epoch": 3.522198731501057,
"grad_norm": 6.722071647644043,
"learning_rate": 2.024592164826562e-06,
"loss": 0.5063920617103577,
"step": 1666
},
{
"epoch": 3.526427061310782,
"grad_norm": 0.6869686841964722,
"learning_rate": 2.015507961641865e-06,
"loss": 0.03437905013561249,
"step": 1668
},
{
"epoch": 3.5306553911205074,
"grad_norm": 0.42433977127075195,
"learning_rate": 2.0064999776270894e-06,
"loss": 0.37125420570373535,
"step": 1670
},
{
"epoch": 3.5348837209302326,
"grad_norm": 2.42488431930542,
"learning_rate": 1.9975683229084617e-06,
"loss": 0.3777064383029938,
"step": 1672
},
{
"epoch": 3.539112050739958,
"grad_norm": 2.969446897506714,
"learning_rate": 1.9887131066790485e-06,
"loss": 0.4842509627342224,
"step": 1674
},
{
"epoch": 3.543340380549683,
"grad_norm": 3.7433881759643555,
"learning_rate": 1.979934437197428e-06,
"loss": 0.3112730383872986,
"step": 1676
},
{
"epoch": 3.547568710359408,
"grad_norm": 1.7907838821411133,
"learning_rate": 1.9712324217863566e-06,
"loss": 0.4308716952800751,
"step": 1678
},
{
"epoch": 3.5517970401691334,
"grad_norm": 1.4491939544677734,
"learning_rate": 1.962607166831473e-06,
"loss": 0.32640641927719116,
"step": 1680
},
{
"epoch": 3.556025369978858,
"grad_norm": 1.7236990928649902,
"learning_rate": 1.9540587777799843e-06,
"loss": 0.23240575194358826,
"step": 1682
},
{
"epoch": 3.5602536997885834,
"grad_norm": 2.8367772102355957,
"learning_rate": 1.945587359139377e-06,
"loss": 0.38225769996643066,
"step": 1684
},
{
"epoch": 3.5644820295983086,
"grad_norm": 7.260848045349121,
"learning_rate": 1.9371930144761517e-06,
"loss": 0.10525190830230713,
"step": 1686
},
{
"epoch": 3.568710359408034,
"grad_norm": 1.7535074949264526,
"learning_rate": 1.9288758464145427e-06,
"loss": 0.43761515617370605,
"step": 1688
},
{
"epoch": 3.572938689217759,
"grad_norm": 0.31258919835090637,
"learning_rate": 1.920635956635271e-06,
"loss": 0.047197841107845306,
"step": 1690
},
{
"epoch": 3.5771670190274842,
"grad_norm": 1.559328556060791,
"learning_rate": 1.9124734458743e-06,
"loss": 0.24445591866970062,
"step": 1692
},
{
"epoch": 3.5813953488372094,
"grad_norm": 7.896991729736328,
"learning_rate": 1.9043884139216031e-06,
"loss": 0.46396341919898987,
"step": 1694
},
{
"epoch": 3.585623678646934,
"grad_norm": 2.968611240386963,
"learning_rate": 1.8963809596199432e-06,
"loss": 0.16597150266170502,
"step": 1696
},
{
"epoch": 3.58985200845666,
"grad_norm": 2.186088800430298,
"learning_rate": 1.8884511808636671e-06,
"loss": 0.43159210681915283,
"step": 1698
},
{
"epoch": 3.5940803382663846,
"grad_norm": 2.2071585655212402,
"learning_rate": 1.8805991745975062e-06,
"loss": 0.6921452283859253,
"step": 1700
},
{
"epoch": 3.59830866807611,
"grad_norm": 5.664018154144287,
"learning_rate": 1.8728250368153906e-06,
"loss": 0.4875612258911133,
"step": 1702
},
{
"epoch": 3.602536997885835,
"grad_norm": 7.896413803100586,
"learning_rate": 1.8651288625592794e-06,
"loss": 0.18402118980884552,
"step": 1704
},
{
"epoch": 3.6067653276955602,
"grad_norm": 1.5156193971633911,
"learning_rate": 1.8575107459179937e-06,
"loss": 0.2941642105579376,
"step": 1706
},
{
"epoch": 3.6109936575052854,
"grad_norm": 5.081479072570801,
"learning_rate": 1.8499707800260706e-06,
"loss": 0.44564810395240784,
"step": 1708
},
{
"epoch": 3.6152219873150107,
"grad_norm": 2.090866804122925,
"learning_rate": 1.8425090570626223e-06,
"loss": 0.3646758794784546,
"step": 1710
},
{
"epoch": 3.619450317124736,
"grad_norm": 2.4264652729034424,
"learning_rate": 1.8351256682502103e-06,
"loss": 0.23779350519180298,
"step": 1712
},
{
"epoch": 3.6236786469344606,
"grad_norm": 0.5099054574966431,
"learning_rate": 1.827820703853728e-06,
"loss": 0.15786030888557434,
"step": 1714
},
{
"epoch": 3.6279069767441863,
"grad_norm": 6.947947025299072,
"learning_rate": 1.8205942531793012e-06,
"loss": 0.3891189694404602,
"step": 1716
},
{
"epoch": 3.632135306553911,
"grad_norm": 0.5348920226097107,
"learning_rate": 1.813446404573192e-06,
"loss": 0.37808191776275635,
"step": 1718
},
{
"epoch": 3.6363636363636362,
"grad_norm": 20.69429588317871,
"learning_rate": 1.8063772454207212e-06,
"loss": 0.16323083639144897,
"step": 1720
},
{
"epoch": 3.6405919661733614,
"grad_norm": 2.15889835357666,
"learning_rate": 1.7993868621451993e-06,
"loss": 0.38081133365631104,
"step": 1722
},
{
"epoch": 3.6448202959830867,
"grad_norm": 2.0530717372894287,
"learning_rate": 1.7924753402068716e-06,
"loss": 0.4285137951374054,
"step": 1724
},
{
"epoch": 3.649048625792812,
"grad_norm": 2.644247055053711,
"learning_rate": 1.7856427641018695e-06,
"loss": 0.5559520721435547,
"step": 1726
},
{
"epoch": 3.653276955602537,
"grad_norm": 3.4544215202331543,
"learning_rate": 1.7788892173611808e-06,
"loss": 0.6821197271347046,
"step": 1728
},
{
"epoch": 3.6575052854122623,
"grad_norm": 41.25182342529297,
"learning_rate": 1.7722147825496282e-06,
"loss": 0.5455248355865479,
"step": 1730
},
{
"epoch": 3.6617336152219875,
"grad_norm": 3.144439220428467,
"learning_rate": 1.7656195412648589e-06,
"loss": 0.43547379970550537,
"step": 1732
},
{
"epoch": 3.6659619450317127,
"grad_norm": 2.2905819416046143,
"learning_rate": 1.7591035741363465e-06,
"loss": 0.41461676359176636,
"step": 1734
},
{
"epoch": 3.6701902748414374,
"grad_norm": 1.994506597518921,
"learning_rate": 1.7526669608244076e-06,
"loss": 0.5644330978393555,
"step": 1736
},
{
"epoch": 3.6744186046511627,
"grad_norm": 3.258462905883789,
"learning_rate": 1.7463097800192268e-06,
"loss": 0.07579860091209412,
"step": 1738
},
{
"epoch": 3.678646934460888,
"grad_norm": 2.389911413192749,
"learning_rate": 1.7400321094398917e-06,
"loss": 0.2575529217720032,
"step": 1740
},
{
"epoch": 3.682875264270613,
"grad_norm": 6.482794761657715,
"learning_rate": 1.733834025833447e-06,
"loss": 0.13686810433864594,
"step": 1742
},
{
"epoch": 3.6871035940803383,
"grad_norm": 2.8362269401550293,
"learning_rate": 1.7277156049739573e-06,
"loss": 0.5093021392822266,
"step": 1744
},
{
"epoch": 3.6913319238900635,
"grad_norm": 0.2309325635433197,
"learning_rate": 1.7216769216615738e-06,
"loss": 0.21552294492721558,
"step": 1746
},
{
"epoch": 3.6955602536997887,
"grad_norm": 3.054949998855591,
"learning_rate": 1.7157180497216258e-06,
"loss": 0.1866249442100525,
"step": 1748
},
{
"epoch": 3.699788583509514,
"grad_norm": 2.1774001121520996,
"learning_rate": 1.7098390620037193e-06,
"loss": 0.6089503169059753,
"step": 1750
},
{
"epoch": 3.704016913319239,
"grad_norm": 3.3051347732543945,
"learning_rate": 1.7040400303808388e-06,
"loss": 0.18538102507591248,
"step": 1752
},
{
"epoch": 3.708245243128964,
"grad_norm": 4.658204078674316,
"learning_rate": 1.6983210257484758e-06,
"loss": 0.3961324989795685,
"step": 1754
},
{
"epoch": 3.712473572938689,
"grad_norm": 1.3823856115341187,
"learning_rate": 1.69268211802376e-06,
"loss": 0.30216479301452637,
"step": 1756
},
{
"epoch": 3.7167019027484143,
"grad_norm": 3.6195321083068848,
"learning_rate": 1.6871233761446057e-06,
"loss": 0.3396352529525757,
"step": 1758
},
{
"epoch": 3.7209302325581395,
"grad_norm": 0.8568614721298218,
"learning_rate": 1.6816448680688613e-06,
"loss": 0.20964425802230835,
"step": 1760
},
{
"epoch": 3.7251585623678647,
"grad_norm": 2.193138599395752,
"learning_rate": 1.6762466607734931e-06,
"loss": 0.50879967212677,
"step": 1762
},
{
"epoch": 3.72938689217759,
"grad_norm": 2.7081847190856934,
"learning_rate": 1.6709288202537525e-06,
"loss": 0.240130215883255,
"step": 1764
},
{
"epoch": 3.733615221987315,
"grad_norm": 4.313446998596191,
"learning_rate": 1.665691411522376e-06,
"loss": 0.5874007940292358,
"step": 1766
},
{
"epoch": 3.7378435517970403,
"grad_norm": 3.1800005435943604,
"learning_rate": 1.6605344986087893e-06,
"loss": 0.3903779089450836,
"step": 1768
},
{
"epoch": 3.7420718816067655,
"grad_norm": 0.5530804395675659,
"learning_rate": 1.655458144558326e-06,
"loss": 0.05097193270921707,
"step": 1770
},
{
"epoch": 3.7463002114164903,
"grad_norm": 3.3892831802368164,
"learning_rate": 1.6504624114314536e-06,
"loss": 0.3092588484287262,
"step": 1772
},
{
"epoch": 3.7505285412262155,
"grad_norm": 6.151468276977539,
"learning_rate": 1.6455473603030148e-06,
"loss": 0.5655317306518555,
"step": 1774
},
{
"epoch": 3.7547568710359407,
"grad_norm": 2.7377429008483887,
"learning_rate": 1.6407130512614863e-06,
"loss": 0.6494180560112,
"step": 1776
},
{
"epoch": 3.758985200845666,
"grad_norm": 1.744395136833191,
"learning_rate": 1.6359595434082377e-06,
"loss": 0.5580757260322571,
"step": 1778
},
{
"epoch": 3.763213530655391,
"grad_norm": 1.9946132898330688,
"learning_rate": 1.6312868948568114e-06,
"loss": 0.5348998308181763,
"step": 1780
},
{
"epoch": 3.7674418604651163,
"grad_norm": 2.5989975929260254,
"learning_rate": 1.6266951627322143e-06,
"loss": 0.4900321960449219,
"step": 1782
},
{
"epoch": 3.7716701902748415,
"grad_norm": 4.668895244598389,
"learning_rate": 1.622184403170217e-06,
"loss": 0.08716302365064621,
"step": 1784
},
{
"epoch": 3.7758985200845667,
"grad_norm": 7.039021968841553,
"learning_rate": 1.6177546713166654e-06,
"loss": 0.4683957099914551,
"step": 1786
},
{
"epoch": 3.780126849894292,
"grad_norm": 3.350433826446533,
"learning_rate": 1.6134060213268122e-06,
"loss": 0.2138897180557251,
"step": 1788
},
{
"epoch": 3.7843551797040167,
"grad_norm": 4.077768325805664,
"learning_rate": 1.609138506364651e-06,
"loss": 0.09039327502250671,
"step": 1790
},
{
"epoch": 3.7885835095137423,
"grad_norm": 1.1139330863952637,
"learning_rate": 1.6049521786022663e-06,
"loss": 0.24710708856582642,
"step": 1792
},
{
"epoch": 3.792811839323467,
"grad_norm": 3.8078513145446777,
"learning_rate": 1.6008470892191971e-06,
"loss": 0.7512996196746826,
"step": 1794
},
{
"epoch": 3.7970401691331923,
"grad_norm": 2.2760603427886963,
"learning_rate": 1.5968232884018094e-06,
"loss": 0.2964702844619751,
"step": 1796
},
{
"epoch": 3.8012684989429175,
"grad_norm": 9.848620414733887,
"learning_rate": 1.592880825342686e-06,
"loss": 0.5475103855133057,
"step": 1798
},
{
"epoch": 3.8054968287526427,
"grad_norm": 2.6793553829193115,
"learning_rate": 1.5890197482400214e-06,
"loss": 0.6825364828109741,
"step": 1800
},
{
"epoch": 3.809725158562368,
"grad_norm": 2.620032787322998,
"learning_rate": 1.5852401042970336e-06,
"loss": 0.3937552571296692,
"step": 1802
},
{
"epoch": 3.813953488372093,
"grad_norm": 1.516156792640686,
"learning_rate": 1.5815419397213902e-06,
"loss": 0.4678071141242981,
"step": 1804
},
{
"epoch": 3.8181818181818183,
"grad_norm": 2.408399820327759,
"learning_rate": 1.577925299724637e-06,
"loss": 0.6306326389312744,
"step": 1806
},
{
"epoch": 3.822410147991543,
"grad_norm": 17.921382904052734,
"learning_rate": 1.5743902285216514e-06,
"loss": 0.11846039444208145,
"step": 1808
},
{
"epoch": 3.8266384778012688,
"grad_norm": 2.338386297225952,
"learning_rate": 1.5709367693300995e-06,
"loss": 0.6535186171531677,
"step": 1810
},
{
"epoch": 3.8308668076109935,
"grad_norm": 4.544635772705078,
"learning_rate": 1.5675649643699071e-06,
"loss": 0.8023751974105835,
"step": 1812
},
{
"epoch": 3.8350951374207187,
"grad_norm": 1.4747891426086426,
"learning_rate": 1.5642748548627444e-06,
"loss": 0.4715295732021332,
"step": 1814
},
{
"epoch": 3.839323467230444,
"grad_norm": 2.4592528343200684,
"learning_rate": 1.5610664810315224e-06,
"loss": 0.31899112462997437,
"step": 1816
},
{
"epoch": 3.843551797040169,
"grad_norm": 3.573084592819214,
"learning_rate": 1.5579398820998995e-06,
"loss": 0.3200571537017822,
"step": 1818
},
{
"epoch": 3.8477801268498943,
"grad_norm": 1.596789836883545,
"learning_rate": 1.5548950962918043e-06,
"loss": 0.21580329537391663,
"step": 1820
},
{
"epoch": 3.8520084566596196,
"grad_norm": 1.002414584159851,
"learning_rate": 1.5519321608309674e-06,
"loss": 0.2545108497142792,
"step": 1822
},
{
"epoch": 3.8562367864693448,
"grad_norm": 1.9474917650222778,
"learning_rate": 1.549051111940463e-06,
"loss": 0.33239108324050903,
"step": 1824
},
{
"epoch": 3.8604651162790695,
"grad_norm": 4.065573692321777,
"learning_rate": 1.5462519848422735e-06,
"loss": 0.25345712900161743,
"step": 1826
},
{
"epoch": 3.864693446088795,
"grad_norm": 5.647709846496582,
"learning_rate": 1.5435348137568503e-06,
"loss": 0.5618326663970947,
"step": 1828
},
{
"epoch": 3.86892177589852,
"grad_norm": 2.1120333671569824,
"learning_rate": 1.540899631902703e-06,
"loss": 0.4603463411331177,
"step": 1830
},
{
"epoch": 3.873150105708245,
"grad_norm": 1.9890505075454712,
"learning_rate": 1.5383464714959885e-06,
"loss": 0.25480324029922485,
"step": 1832
},
{
"epoch": 3.8773784355179703,
"grad_norm": 1.647275447845459,
"learning_rate": 1.5358753637501168e-06,
"loss": 0.16612286865711212,
"step": 1834
},
{
"epoch": 3.8816067653276956,
"grad_norm": 0.4875168800354004,
"learning_rate": 1.5334863388753753e-06,
"loss": 0.2444835752248764,
"step": 1836
},
{
"epoch": 3.8858350951374208,
"grad_norm": 1.1980202198028564,
"learning_rate": 1.5311794260785508e-06,
"loss": 0.300431489944458,
"step": 1838
},
{
"epoch": 3.890063424947146,
"grad_norm": 2.1776413917541504,
"learning_rate": 1.5289546535625803e-06,
"loss": 0.5288171172142029,
"step": 1840
},
{
"epoch": 3.894291754756871,
"grad_norm": 3.7906546592712402,
"learning_rate": 1.5268120485262017e-06,
"loss": 0.5547816753387451,
"step": 1842
},
{
"epoch": 3.898520084566596,
"grad_norm": 2.167079448699951,
"learning_rate": 1.5247516371636211e-06,
"loss": 0.3471487760543823,
"step": 1844
},
{
"epoch": 3.9027484143763216,
"grad_norm": 2.6233794689178467,
"learning_rate": 1.522773444664197e-06,
"loss": 0.4128296971321106,
"step": 1846
},
{
"epoch": 3.9069767441860463,
"grad_norm": 4.710732460021973,
"learning_rate": 1.5208774952121262e-06,
"loss": 0.06081973761320114,
"step": 1848
},
{
"epoch": 3.9112050739957716,
"grad_norm": 1.8623006343841553,
"learning_rate": 1.5190638119861523e-06,
"loss": 0.45480477809906006,
"step": 1850
},
{
"epoch": 3.9154334038054968,
"grad_norm": 1.5831469297409058,
"learning_rate": 1.5173324171592825e-06,
"loss": 0.3575555980205536,
"step": 1852
},
{
"epoch": 3.919661733615222,
"grad_norm": 1.443250060081482,
"learning_rate": 1.5156833318985132e-06,
"loss": 0.480334997177124,
"step": 1854
},
{
"epoch": 3.923890063424947,
"grad_norm": 2.7109949588775635,
"learning_rate": 1.514116576364574e-06,
"loss": 0.228928342461586,
"step": 1856
},
{
"epoch": 3.9281183932346724,
"grad_norm": 3.2303571701049805,
"learning_rate": 1.5126321697116824e-06,
"loss": 0.39826691150665283,
"step": 1858
},
{
"epoch": 3.9323467230443976,
"grad_norm": 3.9181406497955322,
"learning_rate": 1.511230130087305e-06,
"loss": 0.2125285416841507,
"step": 1860
},
{
"epoch": 3.9365750528541223,
"grad_norm": 7.024599552154541,
"learning_rate": 1.509910474631939e-06,
"loss": 0.7361782789230347,
"step": 1862
},
{
"epoch": 3.940803382663848,
"grad_norm": 1.5889050960540771,
"learning_rate": 1.5086732194789031e-06,
"loss": 0.39678746461868286,
"step": 1864
},
{
"epoch": 3.9450317124735728,
"grad_norm": 1.95686936378479,
"learning_rate": 1.507518379754139e-06,
"loss": 0.6197966933250427,
"step": 1866
},
{
"epoch": 3.949260042283298,
"grad_norm": 0.3734652101993561,
"learning_rate": 1.5064459695760255e-06,
"loss": 0.005497870035469532,
"step": 1868
},
{
"epoch": 3.953488372093023,
"grad_norm": 1.909144639968872,
"learning_rate": 1.5054560020552078e-06,
"loss": 0.558434009552002,
"step": 1870
},
{
"epoch": 3.9577167019027484,
"grad_norm": 1.6199933290481567,
"learning_rate": 1.504548489294438e-06,
"loss": 0.27792245149612427,
"step": 1872
},
{
"epoch": 3.9619450317124736,
"grad_norm": 2.369783401489258,
"learning_rate": 1.5037234423884224e-06,
"loss": 0.30478471517562866,
"step": 1874
},
{
"epoch": 3.966173361522199,
"grad_norm": 1.9914082288742065,
"learning_rate": 1.5029808714236927e-06,
"loss": 0.511407732963562,
"step": 1876
},
{
"epoch": 3.970401691331924,
"grad_norm": 4.849050045013428,
"learning_rate": 1.502320785478475e-06,
"loss": 0.09230373799800873,
"step": 1878
},
{
"epoch": 3.974630021141649,
"grad_norm": 1.9327763319015503,
"learning_rate": 1.5017431926225876e-06,
"loss": 0.25974082946777344,
"step": 1880
},
{
"epoch": 3.9788583509513744,
"grad_norm": 4.299111843109131,
"learning_rate": 1.5012480999173337e-06,
"loss": 0.5388734340667725,
"step": 1882
},
{
"epoch": 3.983086680761099,
"grad_norm": 3.1257824897766113,
"learning_rate": 1.5008355134154204e-06,
"loss": 0.7664746046066284,
"step": 1884
},
{
"epoch": 3.9873150105708244,
"grad_norm": 2.4429774284362793,
"learning_rate": 1.500505438160885e-06,
"loss": 0.3411826491355896,
"step": 1886
},
{
"epoch": 3.9915433403805496,
"grad_norm": 11.56125545501709,
"learning_rate": 1.500257878189028e-06,
"loss": 0.4346163868904114,
"step": 1888
},
{
"epoch": 3.995771670190275,
"grad_norm": 1.4551942348480225,
"learning_rate": 1.5000928365263716e-06,
"loss": 0.34119731187820435,
"step": 1890
},
{
"epoch": 4.0,
"grad_norm": 1.9164273738861084,
"learning_rate": 1.5000103151906145e-06,
"loss": 0.20500323176383972,
"step": 1892
},
{
"epoch": 4.0,
"step": 1892,
"total_flos": 3.5845266746764165e+18,
"train_loss": 0.7556913677518341,
"train_runtime": 4810.6403,
"train_samples_per_second": 11.799,
"train_steps_per_second": 0.393
}
],
"logging_steps": 2,
"max_steps": 1892,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.5845266746764165e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}