9b-94 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
ae4999c 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": 7.694468975067139,
"learning_rate": 1.5789473684210527e-07,
"loss": 1.8082377910614014,
"step": 2
},
{
"epoch": 0.008456659619450317,
"grad_norm": 4.224132537841797,
"learning_rate": 4.736842105263158e-07,
"loss": 2.105644941329956,
"step": 4
},
{
"epoch": 0.012684989429175475,
"grad_norm": 40.48551940917969,
"learning_rate": 7.894736842105263e-07,
"loss": 2.1039841175079346,
"step": 6
},
{
"epoch": 0.016913319238900635,
"grad_norm": 4.266447067260742,
"learning_rate": 1.1052631578947367e-06,
"loss": 1.9520416259765625,
"step": 8
},
{
"epoch": 0.021141649048625793,
"grad_norm": 14.256462097167969,
"learning_rate": 1.4210526315789475e-06,
"loss": 1.9604310989379883,
"step": 10
},
{
"epoch": 0.02536997885835095,
"grad_norm": 9.096770286560059,
"learning_rate": 1.7368421052631579e-06,
"loss": 1.6455304622650146,
"step": 12
},
{
"epoch": 0.02959830866807611,
"grad_norm": 5.048446178436279,
"learning_rate": 2.0526315789473687e-06,
"loss": 1.5505057573318481,
"step": 14
},
{
"epoch": 0.03382663847780127,
"grad_norm": 14.58578872680664,
"learning_rate": 2.368421052631579e-06,
"loss": 1.6621551513671875,
"step": 16
},
{
"epoch": 0.03805496828752643,
"grad_norm": 5.253897666931152,
"learning_rate": 2.6842105263157895e-06,
"loss": 1.6740902662277222,
"step": 18
},
{
"epoch": 0.042283298097251586,
"grad_norm": 23.081262588500977,
"learning_rate": 3e-06,
"loss": 1.3902744054794312,
"step": 20
},
{
"epoch": 0.046511627906976744,
"grad_norm": 9.22622013092041,
"learning_rate": 3.3157894736842107e-06,
"loss": 1.2998855113983154,
"step": 22
},
{
"epoch": 0.0507399577167019,
"grad_norm": 5.467698574066162,
"learning_rate": 3.6315789473684213e-06,
"loss": 1.7007815837860107,
"step": 24
},
{
"epoch": 0.05496828752642706,
"grad_norm": 12.720537185668945,
"learning_rate": 3.9473684210526315e-06,
"loss": 1.308732509613037,
"step": 26
},
{
"epoch": 0.05919661733615222,
"grad_norm": 2.6377975940704346,
"learning_rate": 4.2631578947368425e-06,
"loss": 1.552311897277832,
"step": 28
},
{
"epoch": 0.06342494714587738,
"grad_norm": 5.389072418212891,
"learning_rate": 4.578947368421053e-06,
"loss": 1.2000977993011475,
"step": 30
},
{
"epoch": 0.06765327695560254,
"grad_norm": 12.113367080688477,
"learning_rate": 4.894736842105264e-06,
"loss": 0.925064742565155,
"step": 32
},
{
"epoch": 0.07188160676532769,
"grad_norm": 4.143190860748291,
"learning_rate": 5.210526315789474e-06,
"loss": 1.124603033065796,
"step": 34
},
{
"epoch": 0.07610993657505286,
"grad_norm": 65.45590209960938,
"learning_rate": 5.526315789473684e-06,
"loss": 1.3079326152801514,
"step": 36
},
{
"epoch": 0.080338266384778,
"grad_norm": 6.632326126098633,
"learning_rate": 5.842105263157895e-06,
"loss": 1.213045597076416,
"step": 38
},
{
"epoch": 0.08456659619450317,
"grad_norm": 13.148003578186035,
"learning_rate": 6.157894736842105e-06,
"loss": 1.1457359790802002,
"step": 40
},
{
"epoch": 0.08879492600422834,
"grad_norm": 9.391830444335938,
"learning_rate": 6.473684210526316e-06,
"loss": 1.6578178405761719,
"step": 42
},
{
"epoch": 0.09302325581395349,
"grad_norm": 12.186043739318848,
"learning_rate": 6.7894736842105264e-06,
"loss": 1.530977487564087,
"step": 44
},
{
"epoch": 0.09725158562367865,
"grad_norm": 4.1364030838012695,
"learning_rate": 7.105263157894737e-06,
"loss": 1.208533525466919,
"step": 46
},
{
"epoch": 0.1014799154334038,
"grad_norm": 4.0209150314331055,
"learning_rate": 7.421052631578948e-06,
"loss": 1.4389941692352295,
"step": 48
},
{
"epoch": 0.10570824524312897,
"grad_norm": 5.241251468658447,
"learning_rate": 7.73684210526316e-06,
"loss": 1.4688036441802979,
"step": 50
},
{
"epoch": 0.10993657505285412,
"grad_norm": 15.264612197875977,
"learning_rate": 8.052631578947368e-06,
"loss": 0.6938844323158264,
"step": 52
},
{
"epoch": 0.11416490486257928,
"grad_norm": 2.7024848461151123,
"learning_rate": 8.368421052631578e-06,
"loss": 1.0089298486709595,
"step": 54
},
{
"epoch": 0.11839323467230443,
"grad_norm": 6.740093231201172,
"learning_rate": 8.68421052631579e-06,
"loss": 1.3316614627838135,
"step": 56
},
{
"epoch": 0.1226215644820296,
"grad_norm": 5.489269733428955,
"learning_rate": 9e-06,
"loss": 1.342879295349121,
"step": 58
},
{
"epoch": 0.12684989429175475,
"grad_norm": 7.907677173614502,
"learning_rate": 9.315789473684212e-06,
"loss": 1.3620566129684448,
"step": 60
},
{
"epoch": 0.13107822410147993,
"grad_norm": 8.113728523254395,
"learning_rate": 9.63157894736842e-06,
"loss": 0.7701879143714905,
"step": 62
},
{
"epoch": 0.13530655391120508,
"grad_norm": 4.900993347167969,
"learning_rate": 9.94736842105263e-06,
"loss": 0.9912793636322021,
"step": 64
},
{
"epoch": 0.13953488372093023,
"grad_norm": 3.029409170150757,
"learning_rate": 1.0263157894736843e-05,
"loss": 1.304231882095337,
"step": 66
},
{
"epoch": 0.14376321353065538,
"grad_norm": 3.575608253479004,
"learning_rate": 1.0578947368421053e-05,
"loss": 1.3269106149673462,
"step": 68
},
{
"epoch": 0.14799154334038056,
"grad_norm": 3.806364059448242,
"learning_rate": 1.0894736842105265e-05,
"loss": 1.3366682529449463,
"step": 70
},
{
"epoch": 0.1522198731501057,
"grad_norm": 2.623356342315674,
"learning_rate": 1.1210526315789473e-05,
"loss": 1.3186771869659424,
"step": 72
},
{
"epoch": 0.15644820295983086,
"grad_norm": 3.0762627124786377,
"learning_rate": 1.1526315789473683e-05,
"loss": 0.9637660384178162,
"step": 74
},
{
"epoch": 0.160676532769556,
"grad_norm": 2.766496419906616,
"learning_rate": 1.1842105263157895e-05,
"loss": 0.8372132182121277,
"step": 76
},
{
"epoch": 0.1649048625792812,
"grad_norm": 3.4211647510528564,
"learning_rate": 1.2157894736842105e-05,
"loss": 1.2967195510864258,
"step": 78
},
{
"epoch": 0.16913319238900634,
"grad_norm": 19.049074172973633,
"learning_rate": 1.2473684210526317e-05,
"loss": 1.109818458557129,
"step": 80
},
{
"epoch": 0.1733615221987315,
"grad_norm": 4.257162570953369,
"learning_rate": 1.2789473684210526e-05,
"loss": 1.0717711448669434,
"step": 82
},
{
"epoch": 0.17758985200845667,
"grad_norm": 3.1085259914398193,
"learning_rate": 1.3105263157894738e-05,
"loss": 0.8620362281799316,
"step": 84
},
{
"epoch": 0.18181818181818182,
"grad_norm": 6.120578289031982,
"learning_rate": 1.3421052631578948e-05,
"loss": 1.7329645156860352,
"step": 86
},
{
"epoch": 0.18604651162790697,
"grad_norm": 4.158970355987549,
"learning_rate": 1.3736842105263158e-05,
"loss": 1.2780858278274536,
"step": 88
},
{
"epoch": 0.19027484143763213,
"grad_norm": 10.176080703735352,
"learning_rate": 1.405263157894737e-05,
"loss": 1.0034303665161133,
"step": 90
},
{
"epoch": 0.1945031712473573,
"grad_norm": 6.470720291137695,
"learning_rate": 1.4368421052631578e-05,
"loss": 1.3009271621704102,
"step": 92
},
{
"epoch": 0.19873150105708245,
"grad_norm": 4.8003740310668945,
"learning_rate": 1.468421052631579e-05,
"loss": 0.9801025390625,
"step": 94
},
{
"epoch": 0.2029598308668076,
"grad_norm": 11.697628021240234,
"learning_rate": 1.5e-05,
"loss": 1.2845995426177979,
"step": 96
},
{
"epoch": 0.20718816067653276,
"grad_norm": 2.799666166305542,
"learning_rate": 1.499995873926907e-05,
"loss": 1.361710548400879,
"step": 98
},
{
"epoch": 0.21141649048625794,
"grad_norm": 4.723179817199707,
"learning_rate": 1.49998349575807e-05,
"loss": 1.3086001873016357,
"step": 100
},
{
"epoch": 0.2156448202959831,
"grad_norm": 2.69785213470459,
"learning_rate": 1.499962865644818e-05,
"loss": 1.2330563068389893,
"step": 102
},
{
"epoch": 0.21987315010570824,
"grad_norm": 5.640810966491699,
"learning_rate": 1.4999339838393618e-05,
"loss": 1.3007533550262451,
"step": 104
},
{
"epoch": 0.22410147991543342,
"grad_norm": 3.2364342212677,
"learning_rate": 1.4998968506947934e-05,
"loss": 1.2876789569854736,
"step": 106
},
{
"epoch": 0.22832980972515857,
"grad_norm": 4.499763488769531,
"learning_rate": 1.4998514666650803e-05,
"loss": 1.0546846389770508,
"step": 108
},
{
"epoch": 0.23255813953488372,
"grad_norm": 6.636302471160889,
"learning_rate": 1.4997978323050603e-05,
"loss": 1.274618148803711,
"step": 110
},
{
"epoch": 0.23678646934460887,
"grad_norm": 6.544074535369873,
"learning_rate": 1.4997359482704357e-05,
"loss": 1.3548552989959717,
"step": 112
},
{
"epoch": 0.24101479915433405,
"grad_norm": 3.450063705444336,
"learning_rate": 1.4996658153177625e-05,
"loss": 1.6124544143676758,
"step": 114
},
{
"epoch": 0.2452431289640592,
"grad_norm": 4.983824729919434,
"learning_rate": 1.4995874343044454e-05,
"loss": 1.3115026950836182,
"step": 116
},
{
"epoch": 0.24947145877378435,
"grad_norm": 4.061642169952393,
"learning_rate": 1.4995008061887229e-05,
"loss": 1.284308671951294,
"step": 118
},
{
"epoch": 0.2536997885835095,
"grad_norm": 3.3890793323516846,
"learning_rate": 1.4994059320296592e-05,
"loss": 0.8836103677749634,
"step": 120
},
{
"epoch": 0.25792811839323465,
"grad_norm": 4.161872863769531,
"learning_rate": 1.4993028129871286e-05,
"loss": 1.3136240243911743,
"step": 122
},
{
"epoch": 0.26215644820295986,
"grad_norm": 3.673124074935913,
"learning_rate": 1.499191450321803e-05,
"loss": 1.1257632970809937,
"step": 124
},
{
"epoch": 0.266384778012685,
"grad_norm": 4.10827112197876,
"learning_rate": 1.4990718453951355e-05,
"loss": 1.3352493047714233,
"step": 126
},
{
"epoch": 0.27061310782241016,
"grad_norm": 5.97629451751709,
"learning_rate": 1.4989439996693445e-05,
"loss": 1.0660347938537598,
"step": 128
},
{
"epoch": 0.2748414376321353,
"grad_norm": 3.9054954051971436,
"learning_rate": 1.4988079147073955e-05,
"loss": 0.9676834344863892,
"step": 130
},
{
"epoch": 0.27906976744186046,
"grad_norm": 10.937294006347656,
"learning_rate": 1.4986635921729819e-05,
"loss": 1.2509698867797852,
"step": 132
},
{
"epoch": 0.2832980972515856,
"grad_norm": 16.520305633544922,
"learning_rate": 1.4985110338305041e-05,
"loss": 0.8924391269683838,
"step": 134
},
{
"epoch": 0.28752642706131076,
"grad_norm": 9.822235107421875,
"learning_rate": 1.49835024154505e-05,
"loss": 1.1069753170013428,
"step": 136
},
{
"epoch": 0.2917547568710359,
"grad_norm": 12.208863258361816,
"learning_rate": 1.4981812172823692e-05,
"loss": 1.3648462295532227,
"step": 138
},
{
"epoch": 0.2959830866807611,
"grad_norm": 3.7847044467926025,
"learning_rate": 1.4980039631088512e-05,
"loss": 0.9196479320526123,
"step": 140
},
{
"epoch": 0.30021141649048627,
"grad_norm": 5.275660037994385,
"learning_rate": 1.4978184811914998e-05,
"loss": 1.5483441352844238,
"step": 142
},
{
"epoch": 0.3044397463002114,
"grad_norm": 11.417383193969727,
"learning_rate": 1.497624773797906e-05,
"loss": 1.1038883924484253,
"step": 144
},
{
"epoch": 0.3086680761099366,
"grad_norm": 5.180784225463867,
"learning_rate": 1.4974228432962202e-05,
"loss": 0.6889899373054504,
"step": 146
},
{
"epoch": 0.3128964059196617,
"grad_norm": 4.184180736541748,
"learning_rate": 1.4972126921551245e-05,
"loss": 1.2477037906646729,
"step": 148
},
{
"epoch": 0.3171247357293869,
"grad_norm": 3.821504831314087,
"learning_rate": 1.4969943229438003e-05,
"loss": 1.2588844299316406,
"step": 150
},
{
"epoch": 0.321353065539112,
"grad_norm": 3.584263801574707,
"learning_rate": 1.4967677383318993e-05,
"loss": 1.1803827285766602,
"step": 152
},
{
"epoch": 0.32558139534883723,
"grad_norm": 8.572141647338867,
"learning_rate": 1.4965329410895095e-05,
"loss": 0.7253922820091248,
"step": 154
},
{
"epoch": 0.3298097251585624,
"grad_norm": 3.9199986457824707,
"learning_rate": 1.496289934087121e-05,
"loss": 1.2379381656646729,
"step": 156
},
{
"epoch": 0.33403805496828753,
"grad_norm": 5.235530853271484,
"learning_rate": 1.4960387202955927e-05,
"loss": 1.2606801986694336,
"step": 158
},
{
"epoch": 0.3382663847780127,
"grad_norm": 3.5799858570098877,
"learning_rate": 1.4957793027861142e-05,
"loss": 1.2408926486968994,
"step": 160
},
{
"epoch": 0.34249471458773784,
"grad_norm": 11.666617393493652,
"learning_rate": 1.4955116847301678e-05,
"loss": 1.127700924873352,
"step": 162
},
{
"epoch": 0.346723044397463,
"grad_norm": 4.9434380531311035,
"learning_rate": 1.4952358693994928e-05,
"loss": 1.0459409952163696,
"step": 164
},
{
"epoch": 0.35095137420718814,
"grad_norm": 3.040898323059082,
"learning_rate": 1.494951860166042e-05,
"loss": 0.8981512784957886,
"step": 166
},
{
"epoch": 0.35517970401691334,
"grad_norm": 17.78899383544922,
"learning_rate": 1.4946596605019421e-05,
"loss": 1.183695912361145,
"step": 168
},
{
"epoch": 0.3594080338266385,
"grad_norm": 3.662290573120117,
"learning_rate": 1.494359273979452e-05,
"loss": 1.1915611028671265,
"step": 170
},
{
"epoch": 0.36363636363636365,
"grad_norm": 2.9901981353759766,
"learning_rate": 1.4940507042709171e-05,
"loss": 1.0496366024017334,
"step": 172
},
{
"epoch": 0.3678646934460888,
"grad_norm": 4.738712310791016,
"learning_rate": 1.4937339551487265e-05,
"loss": 1.3456393480300903,
"step": 174
},
{
"epoch": 0.37209302325581395,
"grad_norm": 5.071211338043213,
"learning_rate": 1.493409030485265e-05,
"loss": 1.022275447845459,
"step": 176
},
{
"epoch": 0.3763213530655391,
"grad_norm": 6.975468158721924,
"learning_rate": 1.4930759342528674e-05,
"loss": 0.9057396650314331,
"step": 178
},
{
"epoch": 0.38054968287526425,
"grad_norm": 2.4394216537475586,
"learning_rate": 1.492734670523769e-05,
"loss": 1.2772858142852783,
"step": 180
},
{
"epoch": 0.38477801268498946,
"grad_norm": 3.5154824256896973,
"learning_rate": 1.4923852434700561e-05,
"loss": 1.5138930082321167,
"step": 182
},
{
"epoch": 0.3890063424947146,
"grad_norm": 4.25987434387207,
"learning_rate": 1.4920276573636144e-05,
"loss": 1.319218635559082,
"step": 184
},
{
"epoch": 0.39323467230443976,
"grad_norm": 5.062972545623779,
"learning_rate": 1.4916619165760783e-05,
"loss": 1.0844130516052246,
"step": 186
},
{
"epoch": 0.3974630021141649,
"grad_norm": 3.5910322666168213,
"learning_rate": 1.491288025578775e-05,
"loss": 0.9806352853775024,
"step": 188
},
{
"epoch": 0.40169133192389006,
"grad_norm": 7.523984432220459,
"learning_rate": 1.4909059889426732e-05,
"loss": 1.0841087102890015,
"step": 190
},
{
"epoch": 0.4059196617336152,
"grad_norm": 3.9168834686279297,
"learning_rate": 1.4905158113383237e-05,
"loss": 1.300925374031067,
"step": 192
},
{
"epoch": 0.41014799154334036,
"grad_norm": 1.194130301475525,
"learning_rate": 1.4901174975358043e-05,
"loss": 1.0178430080413818,
"step": 194
},
{
"epoch": 0.4143763213530655,
"grad_norm": 3.8591973781585693,
"learning_rate": 1.4897110524046617e-05,
"loss": 1.2843847274780273,
"step": 196
},
{
"epoch": 0.4186046511627907,
"grad_norm": 13.075984954833984,
"learning_rate": 1.4892964809138505e-05,
"loss": 1.1163668632507324,
"step": 198
},
{
"epoch": 0.42283298097251587,
"grad_norm": 3.885435104370117,
"learning_rate": 1.4888737881316741e-05,
"loss": 1.1895148754119873,
"step": 200
},
{
"epoch": 0.427061310782241,
"grad_norm": 8.137332916259766,
"learning_rate": 1.4884429792257216e-05,
"loss": 0.5709080696105957,
"step": 202
},
{
"epoch": 0.4312896405919662,
"grad_norm": 3.502025842666626,
"learning_rate": 1.4880040594628045e-05,
"loss": 1.2324035167694092,
"step": 204
},
{
"epoch": 0.4355179704016913,
"grad_norm": 9.34576416015625,
"learning_rate": 1.4875570342088935e-05,
"loss": 1.3080203533172607,
"step": 206
},
{
"epoch": 0.4397463002114165,
"grad_norm": 13.370067596435547,
"learning_rate": 1.487101908929052e-05,
"loss": 0.762572169303894,
"step": 208
},
{
"epoch": 0.4439746300211416,
"grad_norm": 4.07540225982666,
"learning_rate": 1.4866386891873693e-05,
"loss": 1.2926549911499023,
"step": 210
},
{
"epoch": 0.44820295983086683,
"grad_norm": 4.890863418579102,
"learning_rate": 1.486167380646893e-05,
"loss": 1.2314975261688232,
"step": 212
},
{
"epoch": 0.452431289640592,
"grad_norm": 8.476245880126953,
"learning_rate": 1.4856879890695592e-05,
"loss": 1.419435977935791,
"step": 214
},
{
"epoch": 0.45665961945031713,
"grad_norm": 4.6727705001831055,
"learning_rate": 1.4852005203161227e-05,
"loss": 0.6406538486480713,
"step": 216
},
{
"epoch": 0.4608879492600423,
"grad_norm": 5.0999650955200195,
"learning_rate": 1.4847049803460847e-05,
"loss": 1.0359560251235962,
"step": 218
},
{
"epoch": 0.46511627906976744,
"grad_norm": 4.5378193855285645,
"learning_rate": 1.4842013752176204e-05,
"loss": 1.2910044193267822,
"step": 220
},
{
"epoch": 0.4693446088794926,
"grad_norm": 2.866401433944702,
"learning_rate": 1.4836897110875048e-05,
"loss": 1.2315150499343872,
"step": 222
},
{
"epoch": 0.47357293868921774,
"grad_norm": 3.354686737060547,
"learning_rate": 1.4831699942110378e-05,
"loss": 1.3348431587219238,
"step": 224
},
{
"epoch": 0.47780126849894294,
"grad_norm": 2.757904529571533,
"learning_rate": 1.4826422309419665e-05,
"loss": 1.2508039474487305,
"step": 226
},
{
"epoch": 0.4820295983086681,
"grad_norm": 3.680971622467041,
"learning_rate": 1.4821064277324092e-05,
"loss": 1.234609603881836,
"step": 228
},
{
"epoch": 0.48625792811839325,
"grad_norm": 9.203660011291504,
"learning_rate": 1.4815625911327753e-05,
"loss": 0.950295090675354,
"step": 230
},
{
"epoch": 0.4904862579281184,
"grad_norm": 3.0646936893463135,
"learning_rate": 1.4810107277916858e-05,
"loss": 1.2634902000427246,
"step": 232
},
{
"epoch": 0.49471458773784355,
"grad_norm": 7.7952141761779785,
"learning_rate": 1.4804508444558917e-05,
"loss": 0.9855551719665527,
"step": 234
},
{
"epoch": 0.4989429175475687,
"grad_norm": 8.720172882080078,
"learning_rate": 1.479882947970192e-05,
"loss": 0.8855298757553101,
"step": 236
},
{
"epoch": 0.5031712473572939,
"grad_norm": 3.9728446006774902,
"learning_rate": 1.4793070452773488e-05,
"loss": 0.8751131296157837,
"step": 238
},
{
"epoch": 0.507399577167019,
"grad_norm": 4.010256767272949,
"learning_rate": 1.4787231434180044e-05,
"loss": 0.9102276563644409,
"step": 240
},
{
"epoch": 0.5116279069767442,
"grad_norm": 7.615983009338379,
"learning_rate": 1.4781312495305934e-05,
"loss": 1.312921166419983,
"step": 242
},
{
"epoch": 0.5158562367864693,
"grad_norm": 7.864718437194824,
"learning_rate": 1.4775313708512563e-05,
"loss": 1.2797123193740845,
"step": 244
},
{
"epoch": 0.5200845665961945,
"grad_norm": 5.20532751083374,
"learning_rate": 1.4769235147137514e-05,
"loss": 1.285362958908081,
"step": 246
},
{
"epoch": 0.5243128964059197,
"grad_norm": 3.5454776287078857,
"learning_rate": 1.4763076885493637e-05,
"loss": 1.2366383075714111,
"step": 248
},
{
"epoch": 0.5285412262156448,
"grad_norm": 8.903877258300781,
"learning_rate": 1.4756838998868158e-05,
"loss": 1.1922203302383423,
"step": 250
},
{
"epoch": 0.53276955602537,
"grad_norm": 24.63131332397461,
"learning_rate": 1.4750521563521741e-05,
"loss": 1.1829485893249512,
"step": 252
},
{
"epoch": 0.5369978858350951,
"grad_norm": 6.238285064697266,
"learning_rate": 1.474412465668758e-05,
"loss": 1.7425312995910645,
"step": 254
},
{
"epoch": 0.5412262156448203,
"grad_norm": 2.712554454803467,
"learning_rate": 1.4737648356570426e-05,
"loss": 1.2178256511688232,
"step": 256
},
{
"epoch": 0.5454545454545454,
"grad_norm": 2.992905855178833,
"learning_rate": 1.4731092742345652e-05,
"loss": 1.3663074970245361,
"step": 258
},
{
"epoch": 0.5496828752642706,
"grad_norm": 3.578605890274048,
"learning_rate": 1.472445789415827e-05,
"loss": 1.017599105834961,
"step": 260
},
{
"epoch": 0.5539112050739958,
"grad_norm": 2.6858808994293213,
"learning_rate": 1.4717743893121974e-05,
"loss": 1.1998910903930664,
"step": 262
},
{
"epoch": 0.5581395348837209,
"grad_norm": 7.4404730796813965,
"learning_rate": 1.4710950821318115e-05,
"loss": 1.060891032218933,
"step": 264
},
{
"epoch": 0.5623678646934461,
"grad_norm": 5.006414413452148,
"learning_rate": 1.4704078761794731e-05,
"loss": 1.5014731884002686,
"step": 266
},
{
"epoch": 0.5665961945031712,
"grad_norm": 22.558250427246094,
"learning_rate": 1.4697127798565508e-05,
"loss": 0.5403841733932495,
"step": 268
},
{
"epoch": 0.5708245243128964,
"grad_norm": 7.220975875854492,
"learning_rate": 1.4690098016608766e-05,
"loss": 1.0778307914733887,
"step": 270
},
{
"epoch": 0.5750528541226215,
"grad_norm": 7.279621601104736,
"learning_rate": 1.4682989501866411e-05,
"loss": 1.1842423677444458,
"step": 272
},
{
"epoch": 0.5792811839323467,
"grad_norm": 2.9293410778045654,
"learning_rate": 1.4675802341242893e-05,
"loss": 0.9627125263214111,
"step": 274
},
{
"epoch": 0.5835095137420718,
"grad_norm": 4.534882068634033,
"learning_rate": 1.466853662260414e-05,
"loss": 0.915324330329895,
"step": 276
},
{
"epoch": 0.587737843551797,
"grad_norm": 1.1912890672683716,
"learning_rate": 1.4661192434776478e-05,
"loss": 0.9031130075454712,
"step": 278
},
{
"epoch": 0.5919661733615222,
"grad_norm": 3.6678357124328613,
"learning_rate": 1.4653769867545556e-05,
"loss": 1.2389652729034424,
"step": 280
},
{
"epoch": 0.5961945031712473,
"grad_norm": 4.153764247894287,
"learning_rate": 1.4646269011655239e-05,
"loss": 1.4591269493103027,
"step": 282
},
{
"epoch": 0.6004228329809725,
"grad_norm": 1.8016067743301392,
"learning_rate": 1.4638689958806506e-05,
"loss": 0.9679784774780273,
"step": 284
},
{
"epoch": 0.6046511627906976,
"grad_norm": 6.767185688018799,
"learning_rate": 1.4631032801656318e-05,
"loss": 1.2702375650405884,
"step": 286
},
{
"epoch": 0.6088794926004228,
"grad_norm": 9.04157543182373,
"learning_rate": 1.4623297633816507e-05,
"loss": 0.4504528045654297,
"step": 288
},
{
"epoch": 0.6131078224101479,
"grad_norm": 2.0485124588012695,
"learning_rate": 1.4615484549852597e-05,
"loss": 1.0178403854370117,
"step": 290
},
{
"epoch": 0.6173361522198731,
"grad_norm": 3.4373555183410645,
"learning_rate": 1.4607593645282686e-05,
"loss": 0.890856146812439,
"step": 292
},
{
"epoch": 0.6215644820295984,
"grad_norm": 3.560213088989258,
"learning_rate": 1.4599625016576245e-05,
"loss": 0.7663946747779846,
"step": 294
},
{
"epoch": 0.6257928118393234,
"grad_norm": 7.024014472961426,
"learning_rate": 1.4591578761152972e-05,
"loss": 0.7603721022605896,
"step": 296
},
{
"epoch": 0.6300211416490487,
"grad_norm": 3.070758819580078,
"learning_rate": 1.4583454977381563e-05,
"loss": 1.2674808502197266,
"step": 298
},
{
"epoch": 0.6342494714587738,
"grad_norm": 4.765546798706055,
"learning_rate": 1.4575253764578541e-05,
"loss": 1.2386133670806885,
"step": 300
},
{
"epoch": 0.638477801268499,
"grad_norm": 3.3891117572784424,
"learning_rate": 1.4566975223007029e-05,
"loss": 0.7391263246536255,
"step": 302
},
{
"epoch": 0.642706131078224,
"grad_norm": 11.069788932800293,
"learning_rate": 1.4558619453875524e-05,
"loss": 0.91716068983078,
"step": 304
},
{
"epoch": 0.6469344608879493,
"grad_norm": 2.78773832321167,
"learning_rate": 1.4550186559336655e-05,
"loss": 0.9525253772735596,
"step": 306
},
{
"epoch": 0.6511627906976745,
"grad_norm": 3.1941332817077637,
"learning_rate": 1.4541676642485954e-05,
"loss": 1.2245872020721436,
"step": 308
},
{
"epoch": 0.6553911205073996,
"grad_norm": 3.0903007984161377,
"learning_rate": 1.4533089807360568e-05,
"loss": 1.193055510520935,
"step": 310
},
{
"epoch": 0.6596194503171248,
"grad_norm": 3.724569320678711,
"learning_rate": 1.452442615893801e-05,
"loss": 1.2031176090240479,
"step": 312
},
{
"epoch": 0.6638477801268499,
"grad_norm": 5.010465145111084,
"learning_rate": 1.4515685803134858e-05,
"loss": 1.2304129600524902,
"step": 314
},
{
"epoch": 0.6680761099365751,
"grad_norm": 1.9797773361206055,
"learning_rate": 1.4506868846805476e-05,
"loss": 1.0927329063415527,
"step": 316
},
{
"epoch": 0.6723044397463002,
"grad_norm": 5.945505142211914,
"learning_rate": 1.4497975397740696e-05,
"loss": 0.7723104357719421,
"step": 318
},
{
"epoch": 0.6765327695560254,
"grad_norm": 3.638169765472412,
"learning_rate": 1.4489005564666508e-05,
"loss": 1.24289870262146,
"step": 320
},
{
"epoch": 0.6807610993657506,
"grad_norm": 5.947479248046875,
"learning_rate": 1.4479959457242723e-05,
"loss": 1.3015351295471191,
"step": 322
},
{
"epoch": 0.6849894291754757,
"grad_norm": 3.7739808559417725,
"learning_rate": 1.4470837186061642e-05,
"loss": 1.1160858869552612,
"step": 324
},
{
"epoch": 0.6892177589852009,
"grad_norm": 2.759521722793579,
"learning_rate": 1.4461638862646691e-05,
"loss": 1.2636997699737549,
"step": 326
},
{
"epoch": 0.693446088794926,
"grad_norm": 5.39089822769165,
"learning_rate": 1.4452364599451069e-05,
"loss": 1.161313772201538,
"step": 328
},
{
"epoch": 0.6976744186046512,
"grad_norm": 6.9409403800964355,
"learning_rate": 1.4443014509856374e-05,
"loss": 0.9155374765396118,
"step": 330
},
{
"epoch": 0.7019027484143763,
"grad_norm": 3.952794313430786,
"learning_rate": 1.4433588708171203e-05,
"loss": 1.2246153354644775,
"step": 332
},
{
"epoch": 0.7061310782241015,
"grad_norm": 14.894582748413086,
"learning_rate": 1.4424087309629773e-05,
"loss": 1.1807254552841187,
"step": 334
},
{
"epoch": 0.7103594080338267,
"grad_norm": 2.6782469749450684,
"learning_rate": 1.4414510430390493e-05,
"loss": 1.2114849090576172,
"step": 336
},
{
"epoch": 0.7145877378435518,
"grad_norm": 5.351640224456787,
"learning_rate": 1.4404858187534565e-05,
"loss": 1.3075833320617676,
"step": 338
},
{
"epoch": 0.718816067653277,
"grad_norm": 42.92411804199219,
"learning_rate": 1.4395130699064532e-05,
"loss": 1.1994694471359253,
"step": 340
},
{
"epoch": 0.7230443974630021,
"grad_norm": 6.069267272949219,
"learning_rate": 1.438532808390285e-05,
"loss": 1.12200927734375,
"step": 342
},
{
"epoch": 0.7272727272727273,
"grad_norm": 21.054357528686523,
"learning_rate": 1.4375450461890427e-05,
"loss": 0.9327563643455505,
"step": 344
},
{
"epoch": 0.7315010570824524,
"grad_norm": 11.144492149353027,
"learning_rate": 1.4365497953785162e-05,
"loss": 0.8434457778930664,
"step": 346
},
{
"epoch": 0.7357293868921776,
"grad_norm": 6.523318290710449,
"learning_rate": 1.4355470681260463e-05,
"loss": 0.9562816619873047,
"step": 348
},
{
"epoch": 0.7399577167019028,
"grad_norm": 2.910780906677246,
"learning_rate": 1.4345368766903767e-05,
"loss": 1.322003960609436,
"step": 350
},
{
"epoch": 0.7441860465116279,
"grad_norm": 7.639605522155762,
"learning_rate": 1.4335192334215031e-05,
"loss": 1.1579456329345703,
"step": 352
},
{
"epoch": 0.7484143763213531,
"grad_norm": 8.27808952331543,
"learning_rate": 1.432494150760524e-05,
"loss": 0.7133760452270508,
"step": 354
},
{
"epoch": 0.7526427061310782,
"grad_norm": 7.430663108825684,
"learning_rate": 1.4314616412394867e-05,
"loss": 1.0546461343765259,
"step": 356
},
{
"epoch": 0.7568710359408034,
"grad_norm": 2.629093885421753,
"learning_rate": 1.4304217174812343e-05,
"loss": 1.2102062702178955,
"step": 358
},
{
"epoch": 0.7610993657505285,
"grad_norm": 8.667981147766113,
"learning_rate": 1.429374392199253e-05,
"loss": 0.7335385680198669,
"step": 360
},
{
"epoch": 0.7653276955602537,
"grad_norm": 8.128499984741211,
"learning_rate": 1.4283196781975148e-05,
"loss": 1.0875585079193115,
"step": 362
},
{
"epoch": 0.7695560253699789,
"grad_norm": 186.99935913085938,
"learning_rate": 1.4272575883703227e-05,
"loss": 1.0409858226776123,
"step": 364
},
{
"epoch": 0.773784355179704,
"grad_norm": 3.1076133251190186,
"learning_rate": 1.4261881357021509e-05,
"loss": 1.2671312093734741,
"step": 366
},
{
"epoch": 0.7780126849894292,
"grad_norm": 4.150426864624023,
"learning_rate": 1.4251113332674884e-05,
"loss": 1.2061045169830322,
"step": 368
},
{
"epoch": 0.7822410147991543,
"grad_norm": 7.632771015167236,
"learning_rate": 1.4240271942306772e-05,
"loss": 0.881278932094574,
"step": 370
},
{
"epoch": 0.7864693446088795,
"grad_norm": 3.4033656120300293,
"learning_rate": 1.4229357318457529e-05,
"loss": 1.1870850324630737,
"step": 372
},
{
"epoch": 0.7906976744186046,
"grad_norm": 5.540162086486816,
"learning_rate": 1.4218369594562817e-05,
"loss": 1.2046070098876953,
"step": 374
},
{
"epoch": 0.7949260042283298,
"grad_norm": 8.602273941040039,
"learning_rate": 1.420730890495197e-05,
"loss": 1.2873551845550537,
"step": 376
},
{
"epoch": 0.7991543340380549,
"grad_norm": 5.199553489685059,
"learning_rate": 1.419617538484637e-05,
"loss": 0.9428425431251526,
"step": 378
},
{
"epoch": 0.8033826638477801,
"grad_norm": 5.882444381713867,
"learning_rate": 1.4184969170357766e-05,
"loss": 1.2508769035339355,
"step": 380
},
{
"epoch": 0.8076109936575053,
"grad_norm": 3.6772570610046387,
"learning_rate": 1.4173690398486635e-05,
"loss": 1.166191816329956,
"step": 382
},
{
"epoch": 0.8118393234672304,
"grad_norm": 3.7890865802764893,
"learning_rate": 1.4162339207120496e-05,
"loss": 0.9958459138870239,
"step": 384
},
{
"epoch": 0.8160676532769556,
"grad_norm": 5.569543361663818,
"learning_rate": 1.4150915735032221e-05,
"loss": 1.018789529800415,
"step": 386
},
{
"epoch": 0.8202959830866807,
"grad_norm": 3.23010516166687,
"learning_rate": 1.4139420121878348e-05,
"loss": 1.4793826341629028,
"step": 388
},
{
"epoch": 0.8245243128964059,
"grad_norm": 8.0390043258667,
"learning_rate": 1.4127852508197365e-05,
"loss": 1.171407699584961,
"step": 390
},
{
"epoch": 0.828752642706131,
"grad_norm": 6.130298614501953,
"learning_rate": 1.4116213035408e-05,
"loss": 0.9279772043228149,
"step": 392
},
{
"epoch": 0.8329809725158562,
"grad_norm": 3.387218713760376,
"learning_rate": 1.410450184580748e-05,
"loss": 1.2349982261657715,
"step": 394
},
{
"epoch": 0.8372093023255814,
"grad_norm": 1.0166298151016235,
"learning_rate": 1.4092719082569803e-05,
"loss": 0.8054320812225342,
"step": 396
},
{
"epoch": 0.8414376321353065,
"grad_norm": 2.6399528980255127,
"learning_rate": 1.4080864889743983e-05,
"loss": 0.7137354016304016,
"step": 398
},
{
"epoch": 0.8456659619450317,
"grad_norm": 5.8082780838012695,
"learning_rate": 1.4068939412252294e-05,
"loss": 1.1254661083221436,
"step": 400
},
{
"epoch": 0.8498942917547568,
"grad_norm": 4.094141006469727,
"learning_rate": 1.4056942795888487e-05,
"loss": 0.9756426811218262,
"step": 402
},
{
"epoch": 0.854122621564482,
"grad_norm": 2.5022029876708984,
"learning_rate": 1.4044875187316015e-05,
"loss": 0.8390086889266968,
"step": 404
},
{
"epoch": 0.8583509513742071,
"grad_norm": 4.811787128448486,
"learning_rate": 1.4032736734066244e-05,
"loss": 0.9835063219070435,
"step": 406
},
{
"epoch": 0.8625792811839323,
"grad_norm": 11.851859092712402,
"learning_rate": 1.402052758453664e-05,
"loss": 0.996360182762146,
"step": 408
},
{
"epoch": 0.8668076109936576,
"grad_norm": 3.006216049194336,
"learning_rate": 1.400824788798896e-05,
"loss": 0.8563735485076904,
"step": 410
},
{
"epoch": 0.8710359408033826,
"grad_norm": 5.611896991729736,
"learning_rate": 1.399589779454743e-05,
"loss": 1.1601784229278564,
"step": 412
},
{
"epoch": 0.8752642706131079,
"grad_norm": 5.7954421043396,
"learning_rate": 1.3983477455196905e-05,
"loss": 1.065341830253601,
"step": 414
},
{
"epoch": 0.879492600422833,
"grad_norm": 5.784130573272705,
"learning_rate": 1.3970987021781022e-05,
"loss": 1.0900375843048096,
"step": 416
},
{
"epoch": 0.8837209302325582,
"grad_norm": 4.542938709259033,
"learning_rate": 1.3958426647000346e-05,
"loss": 1.213735580444336,
"step": 418
},
{
"epoch": 0.8879492600422833,
"grad_norm": 7.152516841888428,
"learning_rate": 1.3945796484410508e-05,
"loss": 1.1278403997421265,
"step": 420
},
{
"epoch": 0.8921775898520085,
"grad_norm": 7.564744472503662,
"learning_rate": 1.3933096688420315e-05,
"loss": 0.9469781517982483,
"step": 422
},
{
"epoch": 0.8964059196617337,
"grad_norm": 3.915074110031128,
"learning_rate": 1.3920327414289881e-05,
"loss": 1.2209410667419434,
"step": 424
},
{
"epoch": 0.9006342494714588,
"grad_norm": 2.8434088230133057,
"learning_rate": 1.3907488818128711e-05,
"loss": 1.1207141876220703,
"step": 426
},
{
"epoch": 0.904862579281184,
"grad_norm": 6.559962749481201,
"learning_rate": 1.3894581056893797e-05,
"loss": 0.9067251086235046,
"step": 428
},
{
"epoch": 0.9090909090909091,
"grad_norm": 2.8538448810577393,
"learning_rate": 1.388160428838771e-05,
"loss": 1.05842924118042,
"step": 430
},
{
"epoch": 0.9133192389006343,
"grad_norm": 4.789990425109863,
"learning_rate": 1.386855867125666e-05,
"loss": 1.2559375762939453,
"step": 432
},
{
"epoch": 0.9175475687103594,
"grad_norm": 9.767813682556152,
"learning_rate": 1.3855444364988552e-05,
"loss": 0.5389173030853271,
"step": 434
},
{
"epoch": 0.9217758985200846,
"grad_norm": 11.294173240661621,
"learning_rate": 1.3842261529911054e-05,
"loss": 0.38379937410354614,
"step": 436
},
{
"epoch": 0.9260042283298098,
"grad_norm": 2.282783031463623,
"learning_rate": 1.382901032718962e-05,
"loss": 0.9439561367034912,
"step": 438
},
{
"epoch": 0.9302325581395349,
"grad_norm": 2.3570144176483154,
"learning_rate": 1.3815690918825522e-05,
"loss": 1.1267650127410889,
"step": 440
},
{
"epoch": 0.9344608879492601,
"grad_norm": 11.564787864685059,
"learning_rate": 1.3802303467653885e-05,
"loss": 0.7750686407089233,
"step": 442
},
{
"epoch": 0.9386892177589852,
"grad_norm": 4.886003017425537,
"learning_rate": 1.3788848137341675e-05,
"loss": 1.245846152305603,
"step": 444
},
{
"epoch": 0.9429175475687104,
"grad_norm": 5.777932167053223,
"learning_rate": 1.3775325092385711e-05,
"loss": 1.1271302700042725,
"step": 446
},
{
"epoch": 0.9471458773784355,
"grad_norm": 32.36698532104492,
"learning_rate": 1.3761734498110652e-05,
"loss": 1.1175425052642822,
"step": 448
},
{
"epoch": 0.9513742071881607,
"grad_norm": 13.74611759185791,
"learning_rate": 1.3748076520666974e-05,
"loss": 0.8335492610931396,
"step": 450
},
{
"epoch": 0.9556025369978859,
"grad_norm": 3.217629909515381,
"learning_rate": 1.3734351327028936e-05,
"loss": 0.8984948992729187,
"step": 452
},
{
"epoch": 0.959830866807611,
"grad_norm": 20.633712768554688,
"learning_rate": 1.3720559084992541e-05,
"loss": 1.3068127632141113,
"step": 454
},
{
"epoch": 0.9640591966173362,
"grad_norm": 9.706938743591309,
"learning_rate": 1.3706699963173498e-05,
"loss": 0.6966771483421326,
"step": 456
},
{
"epoch": 0.9682875264270613,
"grad_norm": 35.36222457885742,
"learning_rate": 1.3692774131005132e-05,
"loss": 0.8925000429153442,
"step": 458
},
{
"epoch": 0.9725158562367865,
"grad_norm": 6.042158603668213,
"learning_rate": 1.3678781758736337e-05,
"loss": 1.2271270751953125,
"step": 460
},
{
"epoch": 0.9767441860465116,
"grad_norm": 3.6695802211761475,
"learning_rate": 1.3664723017429491e-05,
"loss": 1.2088079452514648,
"step": 462
},
{
"epoch": 0.9809725158562368,
"grad_norm": 4.8280158042907715,
"learning_rate": 1.3650598078958347e-05,
"loss": 0.7629251480102539,
"step": 464
},
{
"epoch": 0.985200845665962,
"grad_norm": 2.760213851928711,
"learning_rate": 1.3636407116005958e-05,
"loss": 1.248457431793213,
"step": 466
},
{
"epoch": 0.9894291754756871,
"grad_norm": 7.934692859649658,
"learning_rate": 1.3622150302062543e-05,
"loss": 1.0029637813568115,
"step": 468
},
{
"epoch": 0.9936575052854123,
"grad_norm": 4.111275672912598,
"learning_rate": 1.3607827811423387e-05,
"loss": 1.2393298149108887,
"step": 470
},
{
"epoch": 0.9978858350951374,
"grad_norm": 3.1085457801818848,
"learning_rate": 1.3593439819186684e-05,
"loss": 1.2449924945831299,
"step": 472
},
{
"epoch": 1.0021141649048626,
"grad_norm": 4.671144485473633,
"learning_rate": 1.3578986501251426e-05,
"loss": 0.9800138473510742,
"step": 474
},
{
"epoch": 1.0063424947145878,
"grad_norm": 5.851824760437012,
"learning_rate": 1.3564468034315229e-05,
"loss": 1.0923573970794678,
"step": 476
},
{
"epoch": 1.0105708245243128,
"grad_norm": 15.939520835876465,
"learning_rate": 1.3549884595872185e-05,
"loss": 0.4250287115573883,
"step": 478
},
{
"epoch": 1.014799154334038,
"grad_norm": 4.853209972381592,
"learning_rate": 1.3535236364210687e-05,
"loss": 0.7025294303894043,
"step": 480
},
{
"epoch": 1.0190274841437632,
"grad_norm": 6.012503623962402,
"learning_rate": 1.3520523518411252e-05,
"loss": 0.6440640687942505,
"step": 482
},
{
"epoch": 1.0232558139534884,
"grad_norm": 3.6974568367004395,
"learning_rate": 1.350574623834433e-05,
"loss": 0.9773796200752258,
"step": 484
},
{
"epoch": 1.0274841437632136,
"grad_norm": 6.0878400802612305,
"learning_rate": 1.349090470466811e-05,
"loss": 0.9877686500549316,
"step": 486
},
{
"epoch": 1.0317124735729386,
"grad_norm": 8.939531326293945,
"learning_rate": 1.3475999098826302e-05,
"loss": 0.9277652502059937,
"step": 488
},
{
"epoch": 1.0359408033826638,
"grad_norm": 14.893548011779785,
"learning_rate": 1.3461029603045926e-05,
"loss": 0.813503623008728,
"step": 490
},
{
"epoch": 1.040169133192389,
"grad_norm": 10.009620666503906,
"learning_rate": 1.344599640033508e-05,
"loss": 0.6832774877548218,
"step": 492
},
{
"epoch": 1.0443974630021142,
"grad_norm": 3.354914903640747,
"learning_rate": 1.3430899674480709e-05,
"loss": 1.1384618282318115,
"step": 494
},
{
"epoch": 1.0486257928118394,
"grad_norm": 2.5703911781311035,
"learning_rate": 1.3415739610046352e-05,
"loss": 0.9616219401359558,
"step": 496
},
{
"epoch": 1.0528541226215644,
"grad_norm": 2.797065019607544,
"learning_rate": 1.3400516392369888e-05,
"loss": 0.800805926322937,
"step": 498
},
{
"epoch": 1.0570824524312896,
"grad_norm": 3.4388198852539062,
"learning_rate": 1.3385230207561267e-05,
"loss": 0.8043174147605896,
"step": 500
},
{
"epoch": 1.0613107822410148,
"grad_norm": 2.472867012023926,
"learning_rate": 1.336988124250024e-05,
"loss": 1.0902597904205322,
"step": 502
},
{
"epoch": 1.06553911205074,
"grad_norm": 2.6906862258911133,
"learning_rate": 1.3354469684834074e-05,
"loss": 0.9202855825424194,
"step": 504
},
{
"epoch": 1.069767441860465,
"grad_norm": 11.038729667663574,
"learning_rate": 1.3338995722975248e-05,
"loss": 0.7815649509429932,
"step": 506
},
{
"epoch": 1.0739957716701902,
"grad_norm": 2.476713180541992,
"learning_rate": 1.332345954609917e-05,
"loss": 0.7831959128379822,
"step": 508
},
{
"epoch": 1.0782241014799154,
"grad_norm": 10.318900108337402,
"learning_rate": 1.3307861344141838e-05,
"loss": 0.8936349153518677,
"step": 510
},
{
"epoch": 1.0824524312896406,
"grad_norm": 8.494014739990234,
"learning_rate": 1.3292201307797538e-05,
"loss": 1.0179133415222168,
"step": 512
},
{
"epoch": 1.0866807610993658,
"grad_norm": 5.610970497131348,
"learning_rate": 1.3276479628516509e-05,
"loss": 1.075989842414856,
"step": 514
},
{
"epoch": 1.0909090909090908,
"grad_norm": 3.0615196228027344,
"learning_rate": 1.3260696498502594e-05,
"loss": 0.8677607774734497,
"step": 516
},
{
"epoch": 1.095137420718816,
"grad_norm": 3.442467212677002,
"learning_rate": 1.3244852110710897e-05,
"loss": 0.808364987373352,
"step": 518
},
{
"epoch": 1.0993657505285412,
"grad_norm": 12.80532169342041,
"learning_rate": 1.3228946658845429e-05,
"loss": 0.6996243000030518,
"step": 520
},
{
"epoch": 1.1035940803382664,
"grad_norm": 2.746342182159424,
"learning_rate": 1.3212980337356727e-05,
"loss": 1.1934597492218018,
"step": 522
},
{
"epoch": 1.1078224101479917,
"grad_norm": 20.217182159423828,
"learning_rate": 1.3196953341439487e-05,
"loss": 0.7237528562545776,
"step": 524
},
{
"epoch": 1.1120507399577166,
"grad_norm": 3.731555223464966,
"learning_rate": 1.3180865867030169e-05,
"loss": 1.2389192581176758,
"step": 526
},
{
"epoch": 1.1162790697674418,
"grad_norm": 2.772258758544922,
"learning_rate": 1.3164718110804614e-05,
"loss": 0.7679740190505981,
"step": 528
},
{
"epoch": 1.120507399577167,
"grad_norm": 6.68835973739624,
"learning_rate": 1.314851027017563e-05,
"loss": 0.5018873810768127,
"step": 530
},
{
"epoch": 1.1247357293868923,
"grad_norm": 3.607449531555176,
"learning_rate": 1.3132242543290577e-05,
"loss": 0.9644540548324585,
"step": 532
},
{
"epoch": 1.1289640591966172,
"grad_norm": 2.7888481616973877,
"learning_rate": 1.3115915129028958e-05,
"loss": 0.7334082126617432,
"step": 534
},
{
"epoch": 1.1331923890063424,
"grad_norm": 2.655278444290161,
"learning_rate": 1.3099528226999963e-05,
"loss": 0.8843320608139038,
"step": 536
},
{
"epoch": 1.1374207188160677,
"grad_norm": 25.368104934692383,
"learning_rate": 1.308308203754006e-05,
"loss": 0.6815811991691589,
"step": 538
},
{
"epoch": 1.1416490486257929,
"grad_norm": 6.247807025909424,
"learning_rate": 1.3066576761710523e-05,
"loss": 0.6804174780845642,
"step": 540
},
{
"epoch": 1.145877378435518,
"grad_norm": 2.6505532264709473,
"learning_rate": 1.3050012601294974e-05,
"loss": 1.007495403289795,
"step": 542
},
{
"epoch": 1.150105708245243,
"grad_norm": 10.71812629699707,
"learning_rate": 1.3033389758796933e-05,
"loss": 0.7838773727416992,
"step": 544
},
{
"epoch": 1.1543340380549683,
"grad_norm": 4.875476837158203,
"learning_rate": 1.301670843743733e-05,
"loss": 1.111068606376648,
"step": 546
},
{
"epoch": 1.1585623678646935,
"grad_norm": 4.563485145568848,
"learning_rate": 1.2999968841152018e-05,
"loss": 1.2227811813354492,
"step": 548
},
{
"epoch": 1.1627906976744187,
"grad_norm": 4.7432942390441895,
"learning_rate": 1.2983171174589283e-05,
"loss": 0.9843556880950928,
"step": 550
},
{
"epoch": 1.1670190274841437,
"grad_norm": 3.407167673110962,
"learning_rate": 1.2966315643107354e-05,
"loss": 1.002907395362854,
"step": 552
},
{
"epoch": 1.1712473572938689,
"grad_norm": 4.063988208770752,
"learning_rate": 1.294940245277187e-05,
"loss": 1.0935487747192383,
"step": 554
},
{
"epoch": 1.175475687103594,
"grad_norm": 4.877832889556885,
"learning_rate": 1.2932431810353383e-05,
"loss": 0.9643925428390503,
"step": 556
},
{
"epoch": 1.1797040169133193,
"grad_norm": 2.247758388519287,
"learning_rate": 1.2915403923324808e-05,
"loss": 1.0302636623382568,
"step": 558
},
{
"epoch": 1.1839323467230445,
"grad_norm": 5.0767974853515625,
"learning_rate": 1.289831899985891e-05,
"loss": 0.8964630961418152,
"step": 560
},
{
"epoch": 1.1881606765327695,
"grad_norm": 2.1838598251342773,
"learning_rate": 1.2881177248825742e-05,
"loss": 0.6631878614425659,
"step": 562
},
{
"epoch": 1.1923890063424947,
"grad_norm": 4.372375011444092,
"learning_rate": 1.2863978879790097e-05,
"loss": 0.9782743453979492,
"step": 564
},
{
"epoch": 1.1966173361522199,
"grad_norm": 7.168341636657715,
"learning_rate": 1.2846724103008943e-05,
"loss": 0.821781575679779,
"step": 566
},
{
"epoch": 1.200845665961945,
"grad_norm": 6.865265369415283,
"learning_rate": 1.2829413129428864e-05,
"loss": 1.1108343601226807,
"step": 568
},
{
"epoch": 1.20507399577167,
"grad_norm": 16.99683952331543,
"learning_rate": 1.2812046170683459e-05,
"loss": 1.178511619567871,
"step": 570
},
{
"epoch": 1.2093023255813953,
"grad_norm": 10.085429191589355,
"learning_rate": 1.2794623439090783e-05,
"loss": 0.9746255874633789,
"step": 572
},
{
"epoch": 1.2135306553911205,
"grad_norm": 2.6181468963623047,
"learning_rate": 1.2777145147650728e-05,
"loss": 0.7305482625961304,
"step": 574
},
{
"epoch": 1.2177589852008457,
"grad_norm": 5.700075149536133,
"learning_rate": 1.2759611510042426e-05,
"loss": 0.8339565396308899,
"step": 576
},
{
"epoch": 1.221987315010571,
"grad_norm": 2.406853437423706,
"learning_rate": 1.2742022740621643e-05,
"loss": 0.8000067472457886,
"step": 578
},
{
"epoch": 1.226215644820296,
"grad_norm": 4.314510345458984,
"learning_rate": 1.2724379054418152e-05,
"loss": 0.7380365133285522,
"step": 580
},
{
"epoch": 1.230443974630021,
"grad_norm": 3.731355905532837,
"learning_rate": 1.2706680667133107e-05,
"loss": 1.0167348384857178,
"step": 582
},
{
"epoch": 1.2346723044397463,
"grad_norm": 9.207245826721191,
"learning_rate": 1.26889277951364e-05,
"loss": 0.8219566941261292,
"step": 584
},
{
"epoch": 1.2389006342494715,
"grad_norm": 20.01541519165039,
"learning_rate": 1.2671120655464021e-05,
"loss": 0.8420779705047607,
"step": 586
},
{
"epoch": 1.2431289640591967,
"grad_norm": 3.9660184383392334,
"learning_rate": 1.2653259465815411e-05,
"loss": 0.9036701321601868,
"step": 588
},
{
"epoch": 1.2473572938689217,
"grad_norm": 1.908424973487854,
"learning_rate": 1.263534444455079e-05,
"loss": 0.6175505518913269,
"step": 590
},
{
"epoch": 1.251585623678647,
"grad_norm": 2.081923246383667,
"learning_rate": 1.2617375810688487e-05,
"loss": 0.8378012180328369,
"step": 592
},
{
"epoch": 1.255813953488372,
"grad_norm": 31.758546829223633,
"learning_rate": 1.2599353783902274e-05,
"loss": 1.101624846458435,
"step": 594
},
{
"epoch": 1.2600422832980973,
"grad_norm": 2.6002578735351562,
"learning_rate": 1.2581278584518666e-05,
"loss": 0.7061848640441895,
"step": 596
},
{
"epoch": 1.2642706131078225,
"grad_norm": 4.041270732879639,
"learning_rate": 1.2563150433514236e-05,
"loss": 1.0772545337677002,
"step": 598
},
{
"epoch": 1.2684989429175475,
"grad_norm": 5.1957573890686035,
"learning_rate": 1.2544969552512918e-05,
"loss": 1.2544552087783813,
"step": 600
},
{
"epoch": 1.2727272727272727,
"grad_norm": 2.725302219390869,
"learning_rate": 1.252673616378328e-05,
"loss": 0.8749057054519653,
"step": 602
},
{
"epoch": 1.276955602536998,
"grad_norm": 2.6818063259124756,
"learning_rate": 1.2508450490235831e-05,
"loss": 0.9962588548660278,
"step": 604
},
{
"epoch": 1.2811839323467231,
"grad_norm": 20.071857452392578,
"learning_rate": 1.2490112755420276e-05,
"loss": 0.7794040441513062,
"step": 606
},
{
"epoch": 1.285412262156448,
"grad_norm": 4.112557888031006,
"learning_rate": 1.2471723183522784e-05,
"loss": 1.0595139265060425,
"step": 608
},
{
"epoch": 1.2896405919661733,
"grad_norm": 3.070066213607788,
"learning_rate": 1.2453281999363266e-05,
"loss": 0.7081382274627686,
"step": 610
},
{
"epoch": 1.2938689217758985,
"grad_norm": 6.5331130027771,
"learning_rate": 1.2434789428392605e-05,
"loss": 0.6282811760902405,
"step": 612
},
{
"epoch": 1.2980972515856237,
"grad_norm": 2.448627233505249,
"learning_rate": 1.241624569668991e-05,
"loss": 0.8275872468948364,
"step": 614
},
{
"epoch": 1.302325581395349,
"grad_norm": 6.834640979766846,
"learning_rate": 1.2397651030959753e-05,
"loss": 0.492057204246521,
"step": 616
},
{
"epoch": 1.306553911205074,
"grad_norm": 30.322317123413086,
"learning_rate": 1.2379005658529392e-05,
"loss": 0.9143649339675903,
"step": 618
},
{
"epoch": 1.3107822410147991,
"grad_norm": 6.323922157287598,
"learning_rate": 1.2360309807345995e-05,
"loss": 1.105822205543518,
"step": 620
},
{
"epoch": 1.3150105708245243,
"grad_norm": 2.325181245803833,
"learning_rate": 1.2341563705973856e-05,
"loss": 0.9550759196281433,
"step": 622
},
{
"epoch": 1.3192389006342495,
"grad_norm": 5.629472255706787,
"learning_rate": 1.2322767583591596e-05,
"loss": 0.7337983846664429,
"step": 624
},
{
"epoch": 1.3234672304439745,
"grad_norm": 3.1637933254241943,
"learning_rate": 1.2303921669989356e-05,
"loss": 1.0394093990325928,
"step": 626
},
{
"epoch": 1.3276955602536997,
"grad_norm": 7.200225353240967,
"learning_rate": 1.228502619556601e-05,
"loss": 1.1816655397415161,
"step": 628
},
{
"epoch": 1.331923890063425,
"grad_norm": 5.452358722686768,
"learning_rate": 1.226608139132632e-05,
"loss": 0.9340100288391113,
"step": 630
},
{
"epoch": 1.3361522198731501,
"grad_norm": 3.514946937561035,
"learning_rate": 1.2247087488878131e-05,
"loss": 0.5858926773071289,
"step": 632
},
{
"epoch": 1.3403805496828753,
"grad_norm": 2.9155478477478027,
"learning_rate": 1.2228044720429527e-05,
"loss": 0.7269644737243652,
"step": 634
},
{
"epoch": 1.3446088794926006,
"grad_norm": 8.57668399810791,
"learning_rate": 1.2208953318786009e-05,
"loss": 0.6660975217819214,
"step": 636
},
{
"epoch": 1.3488372093023255,
"grad_norm": 40.965423583984375,
"learning_rate": 1.2189813517347628e-05,
"loss": 0.9851181507110596,
"step": 638
},
{
"epoch": 1.3530655391120507,
"grad_norm": 2.95336651802063,
"learning_rate": 1.2170625550106147e-05,
"loss": 0.8330931067466736,
"step": 640
},
{
"epoch": 1.357293868921776,
"grad_norm": 2.1522130966186523,
"learning_rate": 1.2151389651642174e-05,
"loss": 1.1999034881591797,
"step": 642
},
{
"epoch": 1.361522198731501,
"grad_norm": 3.892958164215088,
"learning_rate": 1.2132106057122298e-05,
"loss": 1.0367488861083984,
"step": 644
},
{
"epoch": 1.3657505285412261,
"grad_norm": 5.883722305297852,
"learning_rate": 1.2112775002296207e-05,
"loss": 0.7198834419250488,
"step": 646
},
{
"epoch": 1.3699788583509513,
"grad_norm": 9.357966423034668,
"learning_rate": 1.2093396723493814e-05,
"loss": 0.7189715504646301,
"step": 648
},
{
"epoch": 1.3742071881606766,
"grad_norm": 11.686338424682617,
"learning_rate": 1.2073971457622361e-05,
"loss": 0.7784378528594971,
"step": 650
},
{
"epoch": 1.3784355179704018,
"grad_norm": 2.431422472000122,
"learning_rate": 1.2054499442163533e-05,
"loss": 0.7387130260467529,
"step": 652
},
{
"epoch": 1.382663847780127,
"grad_norm": 3.942812204360962,
"learning_rate": 1.2034980915170535e-05,
"loss": 1.1472055912017822,
"step": 654
},
{
"epoch": 1.386892177589852,
"grad_norm": 3.257403612136841,
"learning_rate": 1.2015416115265201e-05,
"loss": 0.9937019348144531,
"step": 656
},
{
"epoch": 1.3911205073995772,
"grad_norm": 8.456377983093262,
"learning_rate": 1.1995805281635077e-05,
"loss": 0.8755460381507874,
"step": 658
},
{
"epoch": 1.3953488372093024,
"grad_norm": 2.4463460445404053,
"learning_rate": 1.1976148654030476e-05,
"loss": 0.9591925144195557,
"step": 660
},
{
"epoch": 1.3995771670190273,
"grad_norm": 2.219970941543579,
"learning_rate": 1.1956446472761575e-05,
"loss": 1.027714490890503,
"step": 662
},
{
"epoch": 1.4038054968287526,
"grad_norm": 0.7301704287528992,
"learning_rate": 1.193669897869545e-05,
"loss": 0.7829005122184753,
"step": 664
},
{
"epoch": 1.4080338266384778,
"grad_norm": 5.239259719848633,
"learning_rate": 1.191690641325315e-05,
"loss": 1.2836198806762695,
"step": 666
},
{
"epoch": 1.412262156448203,
"grad_norm": 7.135112762451172,
"learning_rate": 1.1897069018406743e-05,
"loss": 1.0177743434906006,
"step": 668
},
{
"epoch": 1.4164904862579282,
"grad_norm": 1.8965673446655273,
"learning_rate": 1.1877187036676345e-05,
"loss": 0.5985036492347717,
"step": 670
},
{
"epoch": 1.4207188160676534,
"grad_norm": 2.7872605323791504,
"learning_rate": 1.1857260711127173e-05,
"loss": 1.022948980331421,
"step": 672
},
{
"epoch": 1.4249471458773784,
"grad_norm": 3.7553322315216064,
"learning_rate": 1.1837290285366566e-05,
"loss": 0.5952565670013428,
"step": 674
},
{
"epoch": 1.4291754756871036,
"grad_norm": 8.235332489013672,
"learning_rate": 1.1817276003540995e-05,
"loss": 0.6997470855712891,
"step": 676
},
{
"epoch": 1.4334038054968288,
"grad_norm": 6.720636367797852,
"learning_rate": 1.1797218110333103e-05,
"loss": 1.2756273746490479,
"step": 678
},
{
"epoch": 1.437632135306554,
"grad_norm": 3.039116859436035,
"learning_rate": 1.177711685095869e-05,
"loss": 1.114811897277832,
"step": 680
},
{
"epoch": 1.441860465116279,
"grad_norm": 5.3889079093933105,
"learning_rate": 1.1756972471163728e-05,
"loss": 1.1164124011993408,
"step": 682
},
{
"epoch": 1.4460887949260042,
"grad_norm": 3.7277066707611084,
"learning_rate": 1.1736785217221353e-05,
"loss": 1.078765630722046,
"step": 684
},
{
"epoch": 1.4503171247357294,
"grad_norm": 10.488996505737305,
"learning_rate": 1.1716555335928856e-05,
"loss": 1.343940258026123,
"step": 686
},
{
"epoch": 1.4545454545454546,
"grad_norm": 3.4089832305908203,
"learning_rate": 1.1696283074604662e-05,
"loss": 1.007319688796997,
"step": 688
},
{
"epoch": 1.4587737843551798,
"grad_norm": 20.99650764465332,
"learning_rate": 1.1675968681085309e-05,
"loss": 0.5235151052474976,
"step": 690
},
{
"epoch": 1.463002114164905,
"grad_norm": 3.5375492572784424,
"learning_rate": 1.165561240372242e-05,
"loss": 1.1756583452224731,
"step": 692
},
{
"epoch": 1.46723044397463,
"grad_norm": 3.020423412322998,
"learning_rate": 1.1635214491379666e-05,
"loss": 0.5886895060539246,
"step": 694
},
{
"epoch": 1.4714587737843552,
"grad_norm": 3.2611489295959473,
"learning_rate": 1.1614775193429719e-05,
"loss": 0.9970772862434387,
"step": 696
},
{
"epoch": 1.4756871035940804,
"grad_norm": 4.714967727661133,
"learning_rate": 1.159429475975121e-05,
"loss": 1.101649284362793,
"step": 698
},
{
"epoch": 1.4799154334038054,
"grad_norm": 3.6968319416046143,
"learning_rate": 1.1573773440725663e-05,
"loss": 1.002673625946045,
"step": 700
},
{
"epoch": 1.4841437632135306,
"grad_norm": 2.209737777709961,
"learning_rate": 1.1553211487234456e-05,
"loss": 1.0193003416061401,
"step": 702
},
{
"epoch": 1.4883720930232558,
"grad_norm": 3.091419219970703,
"learning_rate": 1.1532609150655723e-05,
"loss": 0.9798566102981567,
"step": 704
},
{
"epoch": 1.492600422832981,
"grad_norm": 2.548701286315918,
"learning_rate": 1.1511966682861316e-05,
"loss": 0.7223621606826782,
"step": 706
},
{
"epoch": 1.4968287526427062,
"grad_norm": 1.5869474411010742,
"learning_rate": 1.1491284336213685e-05,
"loss": 0.25400954484939575,
"step": 708
},
{
"epoch": 1.5010570824524314,
"grad_norm": 4.244919776916504,
"learning_rate": 1.1470562363562834e-05,
"loss": 0.6636357307434082,
"step": 710
},
{
"epoch": 1.5052854122621564,
"grad_norm": 5.597408771514893,
"learning_rate": 1.1449801018243205e-05,
"loss": 0.5520629286766052,
"step": 712
},
{
"epoch": 1.5095137420718816,
"grad_norm": 3.0784873962402344,
"learning_rate": 1.1429000554070582e-05,
"loss": 0.9222792387008667,
"step": 714
},
{
"epoch": 1.5137420718816068,
"grad_norm": 3.9098949432373047,
"learning_rate": 1.1408161225339e-05,
"loss": 1.0374963283538818,
"step": 716
},
{
"epoch": 1.5179704016913318,
"grad_norm": 3.4602458477020264,
"learning_rate": 1.138728328681762e-05,
"loss": 0.9747165441513062,
"step": 718
},
{
"epoch": 1.522198731501057,
"grad_norm": 1.7642043828964233,
"learning_rate": 1.1366366993747633e-05,
"loss": 1.0247418880462646,
"step": 720
},
{
"epoch": 1.5264270613107822,
"grad_norm": 6.275250434875488,
"learning_rate": 1.1345412601839121e-05,
"loss": 0.9292112588882446,
"step": 722
},
{
"epoch": 1.5306553911205074,
"grad_norm": 3.356963872909546,
"learning_rate": 1.1324420367267947e-05,
"loss": 1.0227938890457153,
"step": 724
},
{
"epoch": 1.5348837209302326,
"grad_norm": 3.6875593662261963,
"learning_rate": 1.1303390546672611e-05,
"loss": 0.8873927593231201,
"step": 726
},
{
"epoch": 1.5391120507399578,
"grad_norm": 4.997678279876709,
"learning_rate": 1.128232339715112e-05,
"loss": 0.6472434401512146,
"step": 728
},
{
"epoch": 1.543340380549683,
"grad_norm": 6.311956882476807,
"learning_rate": 1.1261219176257841e-05,
"loss": 0.8464550971984863,
"step": 730
},
{
"epoch": 1.547568710359408,
"grad_norm": 7.914978981018066,
"learning_rate": 1.124007814200035e-05,
"loss": 0.9824374914169312,
"step": 732
},
{
"epoch": 1.5517970401691332,
"grad_norm": 2.399554967880249,
"learning_rate": 1.1218900552836286e-05,
"loss": 0.27934643626213074,
"step": 734
},
{
"epoch": 1.5560253699788582,
"grad_norm": 10.44113826751709,
"learning_rate": 1.1197686667670186e-05,
"loss": 0.9760729670524597,
"step": 736
},
{
"epoch": 1.5602536997885834,
"grad_norm": 1.8871999979019165,
"learning_rate": 1.1176436745850315e-05,
"loss": 0.6704195141792297,
"step": 738
},
{
"epoch": 1.5644820295983086,
"grad_norm": 2.2626752853393555,
"learning_rate": 1.1155151047165512e-05,
"loss": 1.036656379699707,
"step": 740
},
{
"epoch": 1.5687103594080338,
"grad_norm": 7.728819847106934,
"learning_rate": 1.1133829831841988e-05,
"loss": 0.8059333562850952,
"step": 742
},
{
"epoch": 1.572938689217759,
"grad_norm": 2.929497480392456,
"learning_rate": 1.1112473360540172e-05,
"loss": 1.0583416223526,
"step": 744
},
{
"epoch": 1.5771670190274842,
"grad_norm": 5.286141872406006,
"learning_rate": 1.1091081894351504e-05,
"loss": 1.1481748819351196,
"step": 746
},
{
"epoch": 1.5813953488372094,
"grad_norm": 3.026707410812378,
"learning_rate": 1.1069655694795247e-05,
"loss": 0.8132309317588806,
"step": 748
},
{
"epoch": 1.5856236786469344,
"grad_norm": 1.7127267122268677,
"learning_rate": 1.1048195023815304e-05,
"loss": 0.06812727451324463,
"step": 750
},
{
"epoch": 1.5898520084566596,
"grad_norm": 11.426569938659668,
"learning_rate": 1.1026700143776998e-05,
"loss": 0.6742124557495117,
"step": 752
},
{
"epoch": 1.5940803382663846,
"grad_norm": 3.724337577819824,
"learning_rate": 1.1005171317463876e-05,
"loss": 1.1136109828948975,
"step": 754
},
{
"epoch": 1.5983086680761098,
"grad_norm": 14.46644115447998,
"learning_rate": 1.0983608808074485e-05,
"loss": 1.0555235147476196,
"step": 756
},
{
"epoch": 1.602536997885835,
"grad_norm": 2.902543783187866,
"learning_rate": 1.0962012879219166e-05,
"loss": 1.0864543914794922,
"step": 758
},
{
"epoch": 1.6067653276955602,
"grad_norm": 2.2856698036193848,
"learning_rate": 1.094038379491683e-05,
"loss": 1.0354400873184204,
"step": 760
},
{
"epoch": 1.6109936575052854,
"grad_norm": 5.503824710845947,
"learning_rate": 1.0918721819591716e-05,
"loss": 1.0819224119186401,
"step": 762
},
{
"epoch": 1.6152219873150107,
"grad_norm": 4.914931297302246,
"learning_rate": 1.089702721807018e-05,
"loss": 1.0502334833145142,
"step": 764
},
{
"epoch": 1.6194503171247359,
"grad_norm": 53.18983459472656,
"learning_rate": 1.087530025557744e-05,
"loss": 0.8511941432952881,
"step": 766
},
{
"epoch": 1.6236786469344608,
"grad_norm": 4.677562236785889,
"learning_rate": 1.0853541197734342e-05,
"loss": 1.0318713188171387,
"step": 768
},
{
"epoch": 1.627906976744186,
"grad_norm": 8.187718391418457,
"learning_rate": 1.0831750310554107e-05,
"loss": 1.1383025646209717,
"step": 770
},
{
"epoch": 1.6321353065539113,
"grad_norm": 3.5982542037963867,
"learning_rate": 1.0809927860439087e-05,
"loss": 0.9839271903038025,
"step": 772
},
{
"epoch": 1.6363636363636362,
"grad_norm": 2.9337029457092285,
"learning_rate": 1.0788074114177505e-05,
"loss": 0.8683944940567017,
"step": 774
},
{
"epoch": 1.6405919661733614,
"grad_norm": 4.334524631500244,
"learning_rate": 1.0766189338940183e-05,
"loss": 1.081053376197815,
"step": 776
},
{
"epoch": 1.6448202959830867,
"grad_norm": 4.9040913581848145,
"learning_rate": 1.0744273802277298e-05,
"loss": 0.666792631149292,
"step": 778
},
{
"epoch": 1.6490486257928119,
"grad_norm": 3.91910457611084,
"learning_rate": 1.0722327772115085e-05,
"loss": 1.0052895545959473,
"step": 780
},
{
"epoch": 1.653276955602537,
"grad_norm": 4.458645820617676,
"learning_rate": 1.0700351516752581e-05,
"loss": 0.795367956161499,
"step": 782
},
{
"epoch": 1.6575052854122623,
"grad_norm": 3.319808006286621,
"learning_rate": 1.0678345304858339e-05,
"loss": 1.1920814514160156,
"step": 784
},
{
"epoch": 1.6617336152219875,
"grad_norm": 9.8060884475708,
"learning_rate": 1.0656309405467135e-05,
"loss": 0.47730082273483276,
"step": 786
},
{
"epoch": 1.6659619450317125,
"grad_norm": 1.8717176914215088,
"learning_rate": 1.0634244087976699e-05,
"loss": 1.0296953916549683,
"step": 788
},
{
"epoch": 1.6701902748414377,
"grad_norm": 2.6735174655914307,
"learning_rate": 1.0612149622144399e-05,
"loss": 0.7138187289237976,
"step": 790
},
{
"epoch": 1.6744186046511627,
"grad_norm": 2.331336736679077,
"learning_rate": 1.0590026278083955e-05,
"loss": 0.28828227519989014,
"step": 792
},
{
"epoch": 1.6786469344608879,
"grad_norm": 5.77733850479126,
"learning_rate": 1.0567874326262141e-05,
"loss": 0.9935804605484009,
"step": 794
},
{
"epoch": 1.682875264270613,
"grad_norm": 2.2524654865264893,
"learning_rate": 1.054569403749546e-05,
"loss": 0.6224902868270874,
"step": 796
},
{
"epoch": 1.6871035940803383,
"grad_norm": 2.1855461597442627,
"learning_rate": 1.0523485682946862e-05,
"loss": 1.1767523288726807,
"step": 798
},
{
"epoch": 1.6913319238900635,
"grad_norm": 2.8387317657470703,
"learning_rate": 1.05012495341224e-05,
"loss": 1.0280368328094482,
"step": 800
},
{
"epoch": 1.6955602536997887,
"grad_norm": 27.568706512451172,
"learning_rate": 1.0478985862867923e-05,
"loss": 0.5349442362785339,
"step": 802
},
{
"epoch": 1.699788583509514,
"grad_norm": 2.9083518981933594,
"learning_rate": 1.0456694941365764e-05,
"loss": 1.076808214187622,
"step": 804
},
{
"epoch": 1.7040169133192389,
"grad_norm": 5.691309928894043,
"learning_rate": 1.0434377042131388e-05,
"loss": 0.7902665138244629,
"step": 806
},
{
"epoch": 1.708245243128964,
"grad_norm": 4.423396587371826,
"learning_rate": 1.0412032438010084e-05,
"loss": 0.7379248142242432,
"step": 808
},
{
"epoch": 1.712473572938689,
"grad_norm": 2.52016544342041,
"learning_rate": 1.038966140217361e-05,
"loss": 1.1066546440124512,
"step": 810
},
{
"epoch": 1.7167019027484143,
"grad_norm": 12.179728507995605,
"learning_rate": 1.0367264208116872e-05,
"loss": 1.3080463409423828,
"step": 812
},
{
"epoch": 1.7209302325581395,
"grad_norm": 2.4812378883361816,
"learning_rate": 1.0344841129654561e-05,
"loss": 1.0156711339950562,
"step": 814
},
{
"epoch": 1.7251585623678647,
"grad_norm": 4.029164791107178,
"learning_rate": 1.0322392440917821e-05,
"loss": 1.0871458053588867,
"step": 816
},
{
"epoch": 1.72938689217759,
"grad_norm": 3.927384853363037,
"learning_rate": 1.0299918416350892e-05,
"loss": 0.8827550411224365,
"step": 818
},
{
"epoch": 1.733615221987315,
"grad_norm": 5.5373759269714355,
"learning_rate": 1.027741933070775e-05,
"loss": 0.9537783265113831,
"step": 820
},
{
"epoch": 1.7378435517970403,
"grad_norm": 4.271606922149658,
"learning_rate": 1.0254895459048762e-05,
"loss": 0.689419686794281,
"step": 822
},
{
"epoch": 1.7420718816067653,
"grad_norm": 1.294979453086853,
"learning_rate": 1.0232347076737304e-05,
"loss": 0.5793521404266357,
"step": 824
},
{
"epoch": 1.7463002114164905,
"grad_norm": 17.122577667236328,
"learning_rate": 1.0209774459436412e-05,
"loss": 0.9610742926597595,
"step": 826
},
{
"epoch": 1.7505285412262155,
"grad_norm": 3.9229793548583984,
"learning_rate": 1.0187177883105397e-05,
"loss": 0.8773674964904785,
"step": 828
},
{
"epoch": 1.7547568710359407,
"grad_norm": 11.299139976501465,
"learning_rate": 1.0164557623996484e-05,
"loss": 0.6969875693321228,
"step": 830
},
{
"epoch": 1.758985200845666,
"grad_norm": 3.3945024013519287,
"learning_rate": 1.0141913958651433e-05,
"loss": 1.096483826637268,
"step": 832
},
{
"epoch": 1.763213530655391,
"grad_norm": 3.2834410667419434,
"learning_rate": 1.0119247163898143e-05,
"loss": 0.72940993309021,
"step": 834
},
{
"epoch": 1.7674418604651163,
"grad_norm": 3.5844316482543945,
"learning_rate": 1.0096557516847292e-05,
"loss": 0.7658038139343262,
"step": 836
},
{
"epoch": 1.7716701902748415,
"grad_norm": 2.6991755962371826,
"learning_rate": 1.0073845294888928e-05,
"loss": 1.0644285678863525,
"step": 838
},
{
"epoch": 1.7758985200845667,
"grad_norm": 2.1810121536254883,
"learning_rate": 1.005111077568909e-05,
"loss": 0.7696046829223633,
"step": 840
},
{
"epoch": 1.7801268498942917,
"grad_norm": 5.519890785217285,
"learning_rate": 1.0028354237186413e-05,
"loss": 0.3710637092590332,
"step": 842
},
{
"epoch": 1.784355179704017,
"grad_norm": 2.480895757675171,
"learning_rate": 1.0005575957588714e-05,
"loss": 0.6643533706665039,
"step": 844
},
{
"epoch": 1.7885835095137421,
"grad_norm": 12.730748176574707,
"learning_rate": 9.982776215369619e-06,
"loss": 0.9776512980461121,
"step": 846
},
{
"epoch": 1.792811839323467,
"grad_norm": 3.686861515045166,
"learning_rate": 9.959955289265134e-06,
"loss": 0.7493982911109924,
"step": 848
},
{
"epoch": 1.7970401691331923,
"grad_norm": 7.337475299835205,
"learning_rate": 9.937113458270251e-06,
"loss": 0.6792470216751099,
"step": 850
},
{
"epoch": 1.8012684989429175,
"grad_norm": 2.533167839050293,
"learning_rate": 9.914251001635533e-06,
"loss": 0.8989590406417847,
"step": 852
},
{
"epoch": 1.8054968287526427,
"grad_norm": 11.006254196166992,
"learning_rate": 9.89136819886369e-06,
"loss": 0.6507301926612854,
"step": 854
},
{
"epoch": 1.809725158562368,
"grad_norm": 3.517608404159546,
"learning_rate": 9.868465329706187e-06,
"loss": 1.0075135231018066,
"step": 856
},
{
"epoch": 1.8139534883720931,
"grad_norm": 7.212334632873535,
"learning_rate": 9.845542674159795e-06,
"loss": 0.6750871539115906,
"step": 858
},
{
"epoch": 1.8181818181818183,
"grad_norm": 17.240829467773438,
"learning_rate": 9.822600512463191e-06,
"loss": 0.9815979599952698,
"step": 860
},
{
"epoch": 1.8224101479915433,
"grad_norm": 3.5664961338043213,
"learning_rate": 9.799639125093517e-06,
"loss": 0.9762383699417114,
"step": 862
},
{
"epoch": 1.8266384778012685,
"grad_norm": 2.7361466884613037,
"learning_rate": 9.776658792762961e-06,
"loss": 0.5271580815315247,
"step": 864
},
{
"epoch": 1.8308668076109935,
"grad_norm": 2.0726253986358643,
"learning_rate": 9.753659796415316e-06,
"loss": 0.7146393060684204,
"step": 866
},
{
"epoch": 1.8350951374207187,
"grad_norm": 3.1244375705718994,
"learning_rate": 9.730642417222552e-06,
"loss": 1.0803109407424927,
"step": 868
},
{
"epoch": 1.839323467230444,
"grad_norm": 2.640867233276367,
"learning_rate": 9.707606936581379e-06,
"loss": 1.0314481258392334,
"step": 870
},
{
"epoch": 1.8435517970401691,
"grad_norm": 0.9036064147949219,
"learning_rate": 9.684553636109804e-06,
"loss": 0.6167151927947998,
"step": 872
},
{
"epoch": 1.8477801268498943,
"grad_norm": 2.197615385055542,
"learning_rate": 9.66148279764368e-06,
"loss": 1.0580313205718994,
"step": 874
},
{
"epoch": 1.8520084566596196,
"grad_norm": 8.674159049987793,
"learning_rate": 9.638394703233283e-06,
"loss": 0.8300498723983765,
"step": 876
},
{
"epoch": 1.8562367864693448,
"grad_norm": 13.859265327453613,
"learning_rate": 9.61528963513984e-06,
"loss": 0.4127427339553833,
"step": 878
},
{
"epoch": 1.8604651162790697,
"grad_norm": 23.210044860839844,
"learning_rate": 9.592167875832095e-06,
"loss": 0.7470444440841675,
"step": 880
},
{
"epoch": 1.864693446088795,
"grad_norm": 2.505610942840576,
"learning_rate": 9.56902970798284e-06,
"loss": 1.090230941772461,
"step": 882
},
{
"epoch": 1.86892177589852,
"grad_norm": 2.9492130279541016,
"learning_rate": 9.54587541446547e-06,
"loss": 0.7710232734680176,
"step": 884
},
{
"epoch": 1.8731501057082451,
"grad_norm": 3.537020206451416,
"learning_rate": 9.522705278350537e-06,
"loss": 1.0183334350585938,
"step": 886
},
{
"epoch": 1.8773784355179703,
"grad_norm": 2.785041093826294,
"learning_rate": 9.499519582902253e-06,
"loss": 1.1000419855117798,
"step": 888
},
{
"epoch": 1.8816067653276956,
"grad_norm": 8.031927108764648,
"learning_rate": 9.476318611575066e-06,
"loss": 0.6185715794563293,
"step": 890
},
{
"epoch": 1.8858350951374208,
"grad_norm": 13.80721378326416,
"learning_rate": 9.45310264801017e-06,
"loss": 0.7824044227600098,
"step": 892
},
{
"epoch": 1.890063424947146,
"grad_norm": 5.304888725280762,
"learning_rate": 9.429871976032045e-06,
"loss": 1.0168921947479248,
"step": 894
},
{
"epoch": 1.8942917547568712,
"grad_norm": 4.4846906661987305,
"learning_rate": 9.40662687964499e-06,
"loss": 1.3741871118545532,
"step": 896
},
{
"epoch": 1.8985200845665962,
"grad_norm": 3.269422769546509,
"learning_rate": 9.383367643029648e-06,
"loss": 0.9915837049484253,
"step": 898
},
{
"epoch": 1.9027484143763214,
"grad_norm": 7.235335826873779,
"learning_rate": 9.360094550539532e-06,
"loss": 0.49685949087142944,
"step": 900
},
{
"epoch": 1.9069767441860463,
"grad_norm": 3.386599540710449,
"learning_rate": 9.336807886697548e-06,
"loss": 1.0478886365890503,
"step": 902
},
{
"epoch": 1.9112050739957716,
"grad_norm": 6.171862602233887,
"learning_rate": 9.313507936192513e-06,
"loss": 0.4817698001861572,
"step": 904
},
{
"epoch": 1.9154334038054968,
"grad_norm": 2.7828140258789062,
"learning_rate": 9.290194983875687e-06,
"loss": 0.8664864301681519,
"step": 906
},
{
"epoch": 1.919661733615222,
"grad_norm": 5.921950340270996,
"learning_rate": 9.266869314757277e-06,
"loss": 1.2135703563690186,
"step": 908
},
{
"epoch": 1.9238900634249472,
"grad_norm": 2.893843173980713,
"learning_rate": 9.243531214002959e-06,
"loss": 0.7672044634819031,
"step": 910
},
{
"epoch": 1.9281183932346724,
"grad_norm": 2.384639024734497,
"learning_rate": 9.22018096693039e-06,
"loss": 1.076727032661438,
"step": 912
},
{
"epoch": 1.9323467230443976,
"grad_norm": 12.744176864624023,
"learning_rate": 9.196818859005725e-06,
"loss": 0.7033243775367737,
"step": 914
},
{
"epoch": 1.9365750528541226,
"grad_norm": 5.475032806396484,
"learning_rate": 9.17344517584011e-06,
"loss": 1.0084125995635986,
"step": 916
},
{
"epoch": 1.9408033826638478,
"grad_norm": 2.289100170135498,
"learning_rate": 9.150060203186227e-06,
"loss": 0.6431819200515747,
"step": 918
},
{
"epoch": 1.945031712473573,
"grad_norm": 12.430115699768066,
"learning_rate": 9.126664226934752e-06,
"loss": 0.8973474502563477,
"step": 920
},
{
"epoch": 1.949260042283298,
"grad_norm": 3.9509570598602295,
"learning_rate": 9.103257533110894e-06,
"loss": 0.9589901566505432,
"step": 922
},
{
"epoch": 1.9534883720930232,
"grad_norm": 2.3002769947052,
"learning_rate": 9.079840407870894e-06,
"loss": 0.9648246169090271,
"step": 924
},
{
"epoch": 1.9577167019027484,
"grad_norm": 0.7922413349151611,
"learning_rate": 9.05641313749851e-06,
"loss": 0.6890993118286133,
"step": 926
},
{
"epoch": 1.9619450317124736,
"grad_norm": 2.6043052673339844,
"learning_rate": 9.03297600840154e-06,
"loss": 1.1835768222808838,
"step": 928
},
{
"epoch": 1.9661733615221988,
"grad_norm": 4.483892440795898,
"learning_rate": 9.0095293071083e-06,
"loss": 0.7545671463012695,
"step": 930
},
{
"epoch": 1.970401691331924,
"grad_norm": 3.2555809020996094,
"learning_rate": 8.986073320264131e-06,
"loss": 1.2224719524383545,
"step": 932
},
{
"epoch": 1.9746300211416492,
"grad_norm": 23.03080940246582,
"learning_rate": 8.962608334627904e-06,
"loss": 0.7920926213264465,
"step": 934
},
{
"epoch": 1.9788583509513742,
"grad_norm": 4.875718116760254,
"learning_rate": 8.939134637068488e-06,
"loss": 0.6524693965911865,
"step": 936
},
{
"epoch": 1.9830866807610994,
"grad_norm": 4.2743449211120605,
"learning_rate": 8.915652514561267e-06,
"loss": 0.738945484161377,
"step": 938
},
{
"epoch": 1.9873150105708244,
"grad_norm": 3.230198621749878,
"learning_rate": 8.892162254184628e-06,
"loss": 0.9962481260299683,
"step": 940
},
{
"epoch": 1.9915433403805496,
"grad_norm": 3.018540143966675,
"learning_rate": 8.868664143116435e-06,
"loss": 1.0108933448791504,
"step": 942
},
{
"epoch": 1.9957716701902748,
"grad_norm": 4.269857406616211,
"learning_rate": 8.845158468630539e-06,
"loss": 1.061213493347168,
"step": 944
},
{
"epoch": 2.0,
"grad_norm": 4.223613739013672,
"learning_rate": 8.821645518093257e-06,
"loss": 0.46159759163856506,
"step": 946
},
{
"epoch": 2.004228329809725,
"grad_norm": 11.974381446838379,
"learning_rate": 8.798125578959852e-06,
"loss": 0.7713661193847656,
"step": 948
},
{
"epoch": 2.0084566596194504,
"grad_norm": 2.5552642345428467,
"learning_rate": 8.77459893877103e-06,
"loss": 0.7751846313476562,
"step": 950
},
{
"epoch": 2.0126849894291756,
"grad_norm": 4.3100690841674805,
"learning_rate": 8.75106588514942e-06,
"loss": 0.7624510526657104,
"step": 952
},
{
"epoch": 2.016913319238901,
"grad_norm": 3.5014708042144775,
"learning_rate": 8.72752670579606e-06,
"loss": 0.9763152003288269,
"step": 954
},
{
"epoch": 2.0211416490486256,
"grad_norm": 6.179083347320557,
"learning_rate": 8.703981688486867e-06,
"loss": 0.9362106323242188,
"step": 956
},
{
"epoch": 2.025369978858351,
"grad_norm": 4.7563347816467285,
"learning_rate": 8.680431121069144e-06,
"loss": 0.6679912805557251,
"step": 958
},
{
"epoch": 2.029598308668076,
"grad_norm": 3.954463005065918,
"learning_rate": 8.656875291458037e-06,
"loss": 0.9013761281967163,
"step": 960
},
{
"epoch": 2.033826638477801,
"grad_norm": 5.922805309295654,
"learning_rate": 8.633314487633026e-06,
"loss": 0.836639940738678,
"step": 962
},
{
"epoch": 2.0380549682875264,
"grad_norm": 7.306420803070068,
"learning_rate": 8.609748997634405e-06,
"loss": 0.7415498495101929,
"step": 964
},
{
"epoch": 2.0422832980972516,
"grad_norm": 3.4032561779022217,
"learning_rate": 8.586179109559756e-06,
"loss": 0.6187089085578918,
"step": 966
},
{
"epoch": 2.046511627906977,
"grad_norm": 12.172759056091309,
"learning_rate": 8.56260511156043e-06,
"loss": 0.6059896349906921,
"step": 968
},
{
"epoch": 2.050739957716702,
"grad_norm": 3.80167555809021,
"learning_rate": 8.539027291838022e-06,
"loss": 0.9446553587913513,
"step": 970
},
{
"epoch": 2.0549682875264272,
"grad_norm": 8.426054000854492,
"learning_rate": 8.515445938640856e-06,
"loss": 0.4593411684036255,
"step": 972
},
{
"epoch": 2.059196617336152,
"grad_norm": 2.598299264907837,
"learning_rate": 8.49186134026044e-06,
"loss": 0.8528721332550049,
"step": 974
},
{
"epoch": 2.063424947145877,
"grad_norm": 3.0406436920166016,
"learning_rate": 8.468273785027973e-06,
"loss": 0.7405988574028015,
"step": 976
},
{
"epoch": 2.0676532769556024,
"grad_norm": 3.550297737121582,
"learning_rate": 8.444683561310793e-06,
"loss": 0.8227701187133789,
"step": 978
},
{
"epoch": 2.0718816067653276,
"grad_norm": 8.660592079162598,
"learning_rate": 8.421090957508857e-06,
"loss": 0.638812780380249,
"step": 980
},
{
"epoch": 2.076109936575053,
"grad_norm": 30.22115135192871,
"learning_rate": 8.39749626205123e-06,
"loss": 0.6441130638122559,
"step": 982
},
{
"epoch": 2.080338266384778,
"grad_norm": 2.1604866981506348,
"learning_rate": 8.37389976339254e-06,
"loss": 0.6124245524406433,
"step": 984
},
{
"epoch": 2.0845665961945032,
"grad_norm": 3.387357711791992,
"learning_rate": 8.350301750009467e-06,
"loss": 0.7291161417961121,
"step": 986
},
{
"epoch": 2.0887949260042284,
"grad_norm": 4.107940196990967,
"learning_rate": 8.3267025103972e-06,
"loss": 0.868223249912262,
"step": 988
},
{
"epoch": 2.0930232558139537,
"grad_norm": 4.35333776473999,
"learning_rate": 8.30310233306593e-06,
"loss": 0.8408278226852417,
"step": 990
},
{
"epoch": 2.097251585623679,
"grad_norm": 8.392865180969238,
"learning_rate": 8.279501506537304e-06,
"loss": 0.5695615410804749,
"step": 992
},
{
"epoch": 2.1014799154334036,
"grad_norm": 21.974660873413086,
"learning_rate": 8.255900319340908e-06,
"loss": 0.5733252763748169,
"step": 994
},
{
"epoch": 2.105708245243129,
"grad_norm": 3.524529457092285,
"learning_rate": 8.232299060010737e-06,
"loss": 0.5694207549095154,
"step": 996
},
{
"epoch": 2.109936575052854,
"grad_norm": 3.8602561950683594,
"learning_rate": 8.208698017081669e-06,
"loss": 0.6829028129577637,
"step": 998
},
{
"epoch": 2.1141649048625792,
"grad_norm": 3.903876543045044,
"learning_rate": 8.185097479085937e-06,
"loss": 0.7495624423027039,
"step": 1000
},
{
"epoch": 2.1183932346723044,
"grad_norm": 6.39402961730957,
"learning_rate": 8.161497734549597e-06,
"loss": 0.25603872537612915,
"step": 1002
},
{
"epoch": 2.1226215644820297,
"grad_norm": 5.94532585144043,
"learning_rate": 8.137899071989008e-06,
"loss": 0.6226204633712769,
"step": 1004
},
{
"epoch": 2.126849894291755,
"grad_norm": 6.8498125076293945,
"learning_rate": 8.1143017799073e-06,
"loss": 0.7593845129013062,
"step": 1006
},
{
"epoch": 2.13107822410148,
"grad_norm": 5.664738178253174,
"learning_rate": 8.090706146790852e-06,
"loss": 0.8689855337142944,
"step": 1008
},
{
"epoch": 2.1353065539112053,
"grad_norm": 2.65742564201355,
"learning_rate": 8.067112461105758e-06,
"loss": 0.36753150820732117,
"step": 1010
},
{
"epoch": 2.13953488372093,
"grad_norm": 3.0979502201080322,
"learning_rate": 8.043521011294303e-06,
"loss": 0.4595454931259155,
"step": 1012
},
{
"epoch": 2.1437632135306552,
"grad_norm": 3.5513947010040283,
"learning_rate": 8.01993208577144e-06,
"loss": 0.4619039297103882,
"step": 1014
},
{
"epoch": 2.1479915433403804,
"grad_norm": 5.999022483825684,
"learning_rate": 7.996345972921259e-06,
"loss": 0.6711839437484741,
"step": 1016
},
{
"epoch": 2.1522198731501057,
"grad_norm": 3.678095579147339,
"learning_rate": 7.972762961093464e-06,
"loss": 0.7942661643028259,
"step": 1018
},
{
"epoch": 2.156448202959831,
"grad_norm": 4.207014560699463,
"learning_rate": 7.949183338599856e-06,
"loss": 0.6157369613647461,
"step": 1020
},
{
"epoch": 2.160676532769556,
"grad_norm": 5.344851493835449,
"learning_rate": 7.925607393710787e-06,
"loss": 0.5308436751365662,
"step": 1022
},
{
"epoch": 2.1649048625792813,
"grad_norm": 4.248290538787842,
"learning_rate": 7.902035414651658e-06,
"loss": 0.7235709428787231,
"step": 1024
},
{
"epoch": 2.1691331923890065,
"grad_norm": 2.3185105323791504,
"learning_rate": 7.878467689599383e-06,
"loss": 0.8228424787521362,
"step": 1026
},
{
"epoch": 2.1733615221987317,
"grad_norm": 3.386425495147705,
"learning_rate": 7.854904506678867e-06,
"loss": 0.7933652400970459,
"step": 1028
},
{
"epoch": 2.177589852008457,
"grad_norm": 4.166110038757324,
"learning_rate": 7.831346153959488e-06,
"loss": 0.6963284015655518,
"step": 1030
},
{
"epoch": 2.1818181818181817,
"grad_norm": 6.0741987228393555,
"learning_rate": 7.807792919451577e-06,
"loss": 0.9691816568374634,
"step": 1032
},
{
"epoch": 2.186046511627907,
"grad_norm": 4.15472936630249,
"learning_rate": 7.784245091102887e-06,
"loss": 1.0107653141021729,
"step": 1034
},
{
"epoch": 2.190274841437632,
"grad_norm": 2.812757968902588,
"learning_rate": 7.760702956795074e-06,
"loss": 0.5595121383666992,
"step": 1036
},
{
"epoch": 2.1945031712473573,
"grad_norm": 10.084578514099121,
"learning_rate": 7.737166804340196e-06,
"loss": 0.33979323506355286,
"step": 1038
},
{
"epoch": 2.1987315010570825,
"grad_norm": 3.1473519802093506,
"learning_rate": 7.713636921477167e-06,
"loss": 0.8217160701751709,
"step": 1040
},
{
"epoch": 2.2029598308668077,
"grad_norm": 10.61715316772461,
"learning_rate": 7.690113595868264e-06,
"loss": 0.31587648391723633,
"step": 1042
},
{
"epoch": 2.207188160676533,
"grad_norm": 6.740917205810547,
"learning_rate": 7.666597115095591e-06,
"loss": 0.7981530427932739,
"step": 1044
},
{
"epoch": 2.211416490486258,
"grad_norm": 4.512329578399658,
"learning_rate": 7.64308776665757e-06,
"loss": 0.428538054227829,
"step": 1046
},
{
"epoch": 2.2156448202959833,
"grad_norm": 5.668430805206299,
"learning_rate": 7.619585837965435e-06,
"loss": 0.6565294861793518,
"step": 1048
},
{
"epoch": 2.219873150105708,
"grad_norm": 8.00012493133545,
"learning_rate": 7.596091616339704e-06,
"loss": 0.5196812748908997,
"step": 1050
},
{
"epoch": 2.2241014799154333,
"grad_norm": 3.3288419246673584,
"learning_rate": 7.572605389006674e-06,
"loss": 0.6543052792549133,
"step": 1052
},
{
"epoch": 2.2283298097251585,
"grad_norm": 5.0379204750061035,
"learning_rate": 7.549127443094913e-06,
"loss": 0.78211510181427,
"step": 1054
},
{
"epoch": 2.2325581395348837,
"grad_norm": 2.536381483078003,
"learning_rate": 7.525658065631742e-06,
"loss": 0.7623010873794556,
"step": 1056
},
{
"epoch": 2.236786469344609,
"grad_norm": 0.8982001543045044,
"learning_rate": 7.502197543539728e-06,
"loss": 0.5135959386825562,
"step": 1058
},
{
"epoch": 2.241014799154334,
"grad_norm": 2.4977076053619385,
"learning_rate": 7.47874616363318e-06,
"loss": 0.708351731300354,
"step": 1060
},
{
"epoch": 2.2452431289640593,
"grad_norm": 3.0987396240234375,
"learning_rate": 7.455304212614639e-06,
"loss": 0.8210294246673584,
"step": 1062
},
{
"epoch": 2.2494714587737845,
"grad_norm": 5.032050609588623,
"learning_rate": 7.431871977071373e-06,
"loss": 0.6369508504867554,
"step": 1064
},
{
"epoch": 2.2536997885835097,
"grad_norm": 8.411706924438477,
"learning_rate": 7.408449743471876e-06,
"loss": 0.48274511098861694,
"step": 1066
},
{
"epoch": 2.2579281183932345,
"grad_norm": 3.858473539352417,
"learning_rate": 7.385037798162367e-06,
"loss": 0.7990399599075317,
"step": 1068
},
{
"epoch": 2.2621564482029597,
"grad_norm": 17.01264762878418,
"learning_rate": 7.361636427363279e-06,
"loss": 0.531071126461029,
"step": 1070
},
{
"epoch": 2.266384778012685,
"grad_norm": 3.482454776763916,
"learning_rate": 7.338245917165771e-06,
"loss": 0.8320953249931335,
"step": 1072
},
{
"epoch": 2.27061310782241,
"grad_norm": 1.3474791049957275,
"learning_rate": 7.314866553528231e-06,
"loss": 0.5014660358428955,
"step": 1074
},
{
"epoch": 2.2748414376321353,
"grad_norm": 5.595104217529297,
"learning_rate": 7.291498622272774e-06,
"loss": 0.5741405487060547,
"step": 1076
},
{
"epoch": 2.2790697674418605,
"grad_norm": 22.88589096069336,
"learning_rate": 7.268142409081741e-06,
"loss": 0.2660031318664551,
"step": 1078
},
{
"epoch": 2.2832980972515857,
"grad_norm": 1.1417269706726074,
"learning_rate": 7.244798199494229e-06,
"loss": 0.13065798580646515,
"step": 1080
},
{
"epoch": 2.287526427061311,
"grad_norm": 0.5472865104675293,
"learning_rate": 7.221466278902575e-06,
"loss": 0.4948457181453705,
"step": 1082
},
{
"epoch": 2.291754756871036,
"grad_norm": 2.980639934539795,
"learning_rate": 7.198146932548882e-06,
"loss": 0.31336259841918945,
"step": 1084
},
{
"epoch": 2.295983086680761,
"grad_norm": 4.62089729309082,
"learning_rate": 7.174840445521527e-06,
"loss": 0.15869992971420288,
"step": 1086
},
{
"epoch": 2.300211416490486,
"grad_norm": 5.934708595275879,
"learning_rate": 7.151547102751679e-06,
"loss": 0.6501063108444214,
"step": 1088
},
{
"epoch": 2.3044397463002113,
"grad_norm": 2.87433123588562,
"learning_rate": 7.128267189009815e-06,
"loss": 0.7624810934066772,
"step": 1090
},
{
"epoch": 2.3086680761099365,
"grad_norm": 4.089123249053955,
"learning_rate": 7.1050009889022285e-06,
"loss": 0.5524851083755493,
"step": 1092
},
{
"epoch": 2.3128964059196617,
"grad_norm": 4.505253791809082,
"learning_rate": 7.081748786867563e-06,
"loss": 0.7651753425598145,
"step": 1094
},
{
"epoch": 2.317124735729387,
"grad_norm": 3.3210437297821045,
"learning_rate": 7.058510867173335e-06,
"loss": 0.5861210823059082,
"step": 1096
},
{
"epoch": 2.321353065539112,
"grad_norm": 9.441654205322266,
"learning_rate": 7.035287513912446e-06,
"loss": 0.7587156891822815,
"step": 1098
},
{
"epoch": 2.3255813953488373,
"grad_norm": 4.8436150550842285,
"learning_rate": 7.012079010999715e-06,
"loss": 0.5002784132957458,
"step": 1100
},
{
"epoch": 2.3298097251585626,
"grad_norm": 4.722902774810791,
"learning_rate": 6.9888856421684235e-06,
"loss": 0.6835054755210876,
"step": 1102
},
{
"epoch": 2.3340380549682873,
"grad_norm": 6.3986968994140625,
"learning_rate": 6.965707690966817e-06,
"loss": 0.800835132598877,
"step": 1104
},
{
"epoch": 2.3382663847780125,
"grad_norm": 2.611933946609497,
"learning_rate": 6.94254544075466e-06,
"loss": 0.744974672794342,
"step": 1106
},
{
"epoch": 2.3424947145877377,
"grad_norm": 2.5357513427734375,
"learning_rate": 6.919399174699772e-06,
"loss": 0.8622950315475464,
"step": 1108
},
{
"epoch": 2.346723044397463,
"grad_norm": 3.6918015480041504,
"learning_rate": 6.896269175774555e-06,
"loss": 0.9981531500816345,
"step": 1110
},
{
"epoch": 2.350951374207188,
"grad_norm": 1.6883232593536377,
"learning_rate": 6.8731557267525306e-06,
"loss": 0.5184407234191895,
"step": 1112
},
{
"epoch": 2.3551797040169133,
"grad_norm": 7.713250637054443,
"learning_rate": 6.850059110204907e-06,
"loss": 0.6412698030471802,
"step": 1114
},
{
"epoch": 2.3594080338266386,
"grad_norm": 14.632420539855957,
"learning_rate": 6.8269796084971e-06,
"loss": 0.483031690120697,
"step": 1116
},
{
"epoch": 2.3636363636363638,
"grad_norm": 10.526875495910645,
"learning_rate": 6.803917503785286e-06,
"loss": 1.0426650047302246,
"step": 1118
},
{
"epoch": 2.367864693446089,
"grad_norm": 2.9324185848236084,
"learning_rate": 6.780873078012963e-06,
"loss": 0.9922658205032349,
"step": 1120
},
{
"epoch": 2.3720930232558137,
"grad_norm": 3.5904834270477295,
"learning_rate": 6.7578466129074925e-06,
"loss": 0.672528862953186,
"step": 1122
},
{
"epoch": 2.376321353065539,
"grad_norm": 5.510763168334961,
"learning_rate": 6.734838389976669e-06,
"loss": 0.5631533861160278,
"step": 1124
},
{
"epoch": 2.380549682875264,
"grad_norm": 11.611955642700195,
"learning_rate": 6.71184869050525e-06,
"loss": 0.7926510572433472,
"step": 1126
},
{
"epoch": 2.3847780126849893,
"grad_norm": 17.06283187866211,
"learning_rate": 6.688877795551555e-06,
"loss": 0.5594879388809204,
"step": 1128
},
{
"epoch": 2.3890063424947146,
"grad_norm": 3.918444871902466,
"learning_rate": 6.665925985944006e-06,
"loss": 0.6550354957580566,
"step": 1130
},
{
"epoch": 2.3932346723044398,
"grad_norm": 46.8716926574707,
"learning_rate": 6.642993542277688e-06,
"loss": 0.5268495678901672,
"step": 1132
},
{
"epoch": 2.397463002114165,
"grad_norm": 3.0828797817230225,
"learning_rate": 6.620080744910945e-06,
"loss": 0.826082706451416,
"step": 1134
},
{
"epoch": 2.40169133192389,
"grad_norm": 10.348783493041992,
"learning_rate": 6.597187873961926e-06,
"loss": 0.6743037700653076,
"step": 1136
},
{
"epoch": 2.4059196617336154,
"grad_norm": 12.417279243469238,
"learning_rate": 6.5743152093051795e-06,
"loss": 0.6979293823242188,
"step": 1138
},
{
"epoch": 2.41014799154334,
"grad_norm": 2.407043933868408,
"learning_rate": 6.551463030568219e-06,
"loss": 0.8857864141464233,
"step": 1140
},
{
"epoch": 2.4143763213530653,
"grad_norm": 2.7346420288085938,
"learning_rate": 6.528631617128107e-06,
"loss": 0.863304853439331,
"step": 1142
},
{
"epoch": 2.4186046511627906,
"grad_norm": 6.259657382965088,
"learning_rate": 6.505821248108051e-06,
"loss": 0.45249801874160767,
"step": 1144
},
{
"epoch": 2.4228329809725158,
"grad_norm": 15.594244003295898,
"learning_rate": 6.4830322023739745e-06,
"loss": 0.6080983877182007,
"step": 1146
},
{
"epoch": 2.427061310782241,
"grad_norm": 2.611605644226074,
"learning_rate": 6.4602647585311134e-06,
"loss": 0.5861684083938599,
"step": 1148
},
{
"epoch": 2.431289640591966,
"grad_norm": 3.322051763534546,
"learning_rate": 6.437519194920624e-06,
"loss": 0.46751779317855835,
"step": 1150
},
{
"epoch": 2.4355179704016914,
"grad_norm": 2.1621007919311523,
"learning_rate": 6.414795789616152e-06,
"loss": 0.4653051793575287,
"step": 1152
},
{
"epoch": 2.4397463002114166,
"grad_norm": 2.70967698097229,
"learning_rate": 6.3920948204204574e-06,
"loss": 0.47130006551742554,
"step": 1154
},
{
"epoch": 2.443974630021142,
"grad_norm": 2.744441032409668,
"learning_rate": 6.369416564862013e-06,
"loss": 0.7375286221504211,
"step": 1156
},
{
"epoch": 2.448202959830867,
"grad_norm": 3.5042006969451904,
"learning_rate": 6.346761300191603e-06,
"loss": 0.43085339665412903,
"step": 1158
},
{
"epoch": 2.452431289640592,
"grad_norm": 2.566814661026001,
"learning_rate": 6.324129303378936e-06,
"loss": 0.7773959040641785,
"step": 1160
},
{
"epoch": 2.456659619450317,
"grad_norm": 3.937950849533081,
"learning_rate": 6.30152085110927e-06,
"loss": 0.8399026393890381,
"step": 1162
},
{
"epoch": 2.460887949260042,
"grad_norm": 2.5817205905914307,
"learning_rate": 6.278936219780016e-06,
"loss": 0.5889649391174316,
"step": 1164
},
{
"epoch": 2.4651162790697674,
"grad_norm": 5.135617733001709,
"learning_rate": 6.256375685497364e-06,
"loss": 0.8806902170181274,
"step": 1166
},
{
"epoch": 2.4693446088794926,
"grad_norm": 2.385249376296997,
"learning_rate": 6.233839524072911e-06,
"loss": 0.8959009647369385,
"step": 1168
},
{
"epoch": 2.473572938689218,
"grad_norm": 5.280783653259277,
"learning_rate": 6.211328011020282e-06,
"loss": 0.5888158082962036,
"step": 1170
},
{
"epoch": 2.477801268498943,
"grad_norm": 9.050507545471191,
"learning_rate": 6.188841421551775e-06,
"loss": 0.6950839161872864,
"step": 1172
},
{
"epoch": 2.482029598308668,
"grad_norm": 6.435463905334473,
"learning_rate": 6.166380030574974e-06,
"loss": 0.5979903936386108,
"step": 1174
},
{
"epoch": 2.4862579281183934,
"grad_norm": 2.2374320030212402,
"learning_rate": 6.1439441126894066e-06,
"loss": 0.544196605682373,
"step": 1176
},
{
"epoch": 2.4904862579281186,
"grad_norm": 4.943401336669922,
"learning_rate": 6.12153394218319e-06,
"loss": 0.6882659196853638,
"step": 1178
},
{
"epoch": 2.4947145877378434,
"grad_norm": 6.256656169891357,
"learning_rate": 6.099149793029655e-06,
"loss": 0.503983736038208,
"step": 1180
},
{
"epoch": 2.4989429175475686,
"grad_norm": 1.7354086637496948,
"learning_rate": 6.076791938884021e-06,
"loss": 0.37795916199684143,
"step": 1182
},
{
"epoch": 2.503171247357294,
"grad_norm": 1.8148046731948853,
"learning_rate": 6.054460653080039e-06,
"loss": 0.11189557611942291,
"step": 1184
},
{
"epoch": 2.507399577167019,
"grad_norm": 5.922695159912109,
"learning_rate": 6.032156208626644e-06,
"loss": 0.8841594457626343,
"step": 1186
},
{
"epoch": 2.511627906976744,
"grad_norm": 27.15448760986328,
"learning_rate": 6.009878878204636e-06,
"loss": 0.6559537649154663,
"step": 1188
},
{
"epoch": 2.5158562367864694,
"grad_norm": 14.289588928222656,
"learning_rate": 5.987628934163326e-06,
"loss": 0.6222526431083679,
"step": 1190
},
{
"epoch": 2.5200845665961946,
"grad_norm": 4.882129669189453,
"learning_rate": 5.9654066485172255e-06,
"loss": 0.9310466051101685,
"step": 1192
},
{
"epoch": 2.52431289640592,
"grad_norm": 2.318420886993408,
"learning_rate": 5.943212292942697e-06,
"loss": 0.5784898400306702,
"step": 1194
},
{
"epoch": 2.528541226215645,
"grad_norm": 18.058345794677734,
"learning_rate": 5.921046138774655e-06,
"loss": 0.7834794521331787,
"step": 1196
},
{
"epoch": 2.53276955602537,
"grad_norm": 4.75090217590332,
"learning_rate": 5.898908457003243e-06,
"loss": 0.7496220469474792,
"step": 1198
},
{
"epoch": 2.536997885835095,
"grad_norm": 83.07051086425781,
"learning_rate": 5.876799518270515e-06,
"loss": 0.6990331411361694,
"step": 1200
},
{
"epoch": 2.54122621564482,
"grad_norm": 3.8002798557281494,
"learning_rate": 5.854719592867123e-06,
"loss": 0.8437431454658508,
"step": 1202
},
{
"epoch": 2.5454545454545454,
"grad_norm": 6.901557445526123,
"learning_rate": 5.832668950729036e-06,
"loss": 0.27411916851997375,
"step": 1204
},
{
"epoch": 2.5496828752642706,
"grad_norm": 2.845240354537964,
"learning_rate": 5.810647861434207e-06,
"loss": 0.8636689186096191,
"step": 1206
},
{
"epoch": 2.553911205073996,
"grad_norm": 3.0194482803344727,
"learning_rate": 5.788656594199295e-06,
"loss": 0.8075793981552124,
"step": 1208
},
{
"epoch": 2.558139534883721,
"grad_norm": 3.082803249359131,
"learning_rate": 5.766695417876385e-06,
"loss": 0.9541327953338623,
"step": 1210
},
{
"epoch": 2.5623678646934462,
"grad_norm": 25.209640502929688,
"learning_rate": 5.7447646009496814e-06,
"loss": 0.5687824487686157,
"step": 1212
},
{
"epoch": 2.5665961945031714,
"grad_norm": 3.5537517070770264,
"learning_rate": 5.722864411532224e-06,
"loss": 0.731854259967804,
"step": 1214
},
{
"epoch": 2.570824524312896,
"grad_norm": 6.745619297027588,
"learning_rate": 5.700995117362633e-06,
"loss": 0.5729124546051025,
"step": 1216
},
{
"epoch": 2.5750528541226214,
"grad_norm": 4.356179237365723,
"learning_rate": 5.679156985801816e-06,
"loss": 0.5397675633430481,
"step": 1218
},
{
"epoch": 2.5792811839323466,
"grad_norm": 17.691112518310547,
"learning_rate": 5.6573502838297005e-06,
"loss": 0.4180915951728821,
"step": 1220
},
{
"epoch": 2.583509513742072,
"grad_norm": 9.041176795959473,
"learning_rate": 5.635575278041984e-06,
"loss": 0.5155890583992004,
"step": 1222
},
{
"epoch": 2.587737843551797,
"grad_norm": 11.912383079528809,
"learning_rate": 5.613832234646856e-06,
"loss": 0.6245535612106323,
"step": 1224
},
{
"epoch": 2.5919661733615222,
"grad_norm": 3.5519378185272217,
"learning_rate": 5.592121419461765e-06,
"loss": 0.27122265100479126,
"step": 1226
},
{
"epoch": 2.5961945031712474,
"grad_norm": 2.705690860748291,
"learning_rate": 5.570443097910141e-06,
"loss": 0.6669236421585083,
"step": 1228
},
{
"epoch": 2.6004228329809727,
"grad_norm": 4.374434471130371,
"learning_rate": 5.548797535018177e-06,
"loss": 0.7499839067459106,
"step": 1230
},
{
"epoch": 2.604651162790698,
"grad_norm": 7.572882175445557,
"learning_rate": 5.527184995411577e-06,
"loss": 0.5752793550491333,
"step": 1232
},
{
"epoch": 2.6088794926004226,
"grad_norm": 8.03064250946045,
"learning_rate": 5.505605743312321e-06,
"loss": 0.5467522144317627,
"step": 1234
},
{
"epoch": 2.613107822410148,
"grad_norm": 2.6537134647369385,
"learning_rate": 5.484060042535431e-06,
"loss": 0.7529774308204651,
"step": 1236
},
{
"epoch": 2.617336152219873,
"grad_norm": 5.498042106628418,
"learning_rate": 5.46254815648576e-06,
"loss": 0.8222641348838806,
"step": 1238
},
{
"epoch": 2.6215644820295982,
"grad_norm": 5.9159440994262695,
"learning_rate": 5.4410703481547604e-06,
"loss": 0.5579326748847961,
"step": 1240
},
{
"epoch": 2.6257928118393234,
"grad_norm": 4.791767120361328,
"learning_rate": 5.419626880117257e-06,
"loss": 0.7740464806556702,
"step": 1242
},
{
"epoch": 2.6300211416490487,
"grad_norm": 4.887804985046387,
"learning_rate": 5.3982180145282705e-06,
"loss": 0.809455156326294,
"step": 1244
},
{
"epoch": 2.634249471458774,
"grad_norm": 15.179192543029785,
"learning_rate": 5.376844013119783e-06,
"loss": 0.5088845491409302,
"step": 1246
},
{
"epoch": 2.638477801268499,
"grad_norm": 17.6065673828125,
"learning_rate": 5.355505137197547e-06,
"loss": 0.4826335310935974,
"step": 1248
},
{
"epoch": 2.6427061310782243,
"grad_norm": 3.4311397075653076,
"learning_rate": 5.334201647637893e-06,
"loss": 0.8478144407272339,
"step": 1250
},
{
"epoch": 2.646934460887949,
"grad_norm": 18.74444580078125,
"learning_rate": 5.312933804884545e-06,
"loss": 0.3134567141532898,
"step": 1252
},
{
"epoch": 2.6511627906976747,
"grad_norm": 19.681562423706055,
"learning_rate": 5.291701868945415e-06,
"loss": 0.4650720953941345,
"step": 1254
},
{
"epoch": 2.6553911205073994,
"grad_norm": 2.7126781940460205,
"learning_rate": 5.270506099389458e-06,
"loss": 0.7491689920425415,
"step": 1256
},
{
"epoch": 2.6596194503171247,
"grad_norm": 2.6843392848968506,
"learning_rate": 5.249346755343472e-06,
"loss": 0.2267875224351883,
"step": 1258
},
{
"epoch": 2.66384778012685,
"grad_norm": 3.666651725769043,
"learning_rate": 5.228224095488936e-06,
"loss": 0.3900783658027649,
"step": 1260
},
{
"epoch": 2.668076109936575,
"grad_norm": 8.007824897766113,
"learning_rate": 5.207138378058854e-06,
"loss": 0.6020317077636719,
"step": 1262
},
{
"epoch": 2.6723044397463003,
"grad_norm": 16.254152297973633,
"learning_rate": 5.1860898608345945e-06,
"loss": 0.5172767639160156,
"step": 1264
},
{
"epoch": 2.6765327695560255,
"grad_norm": 8.599892616271973,
"learning_rate": 5.165078801142734e-06,
"loss": 0.3476654887199402,
"step": 1266
},
{
"epoch": 2.6807610993657507,
"grad_norm": 2.5561442375183105,
"learning_rate": 5.144105455851921e-06,
"loss": 0.5237537622451782,
"step": 1268
},
{
"epoch": 2.6849894291754755,
"grad_norm": 5.838195323944092,
"learning_rate": 5.123170081369727e-06,
"loss": 0.6949069499969482,
"step": 1270
},
{
"epoch": 2.689217758985201,
"grad_norm": 47.40632629394531,
"learning_rate": 5.102272933639516e-06,
"loss": 0.534892201423645,
"step": 1272
},
{
"epoch": 2.693446088794926,
"grad_norm": 1.3106054067611694,
"learning_rate": 5.081414268137314e-06,
"loss": 0.26238980889320374,
"step": 1274
},
{
"epoch": 2.697674418604651,
"grad_norm": 4.7548298835754395,
"learning_rate": 5.060594339868688e-06,
"loss": 0.3718658685684204,
"step": 1276
},
{
"epoch": 2.7019027484143763,
"grad_norm": 3.828075647354126,
"learning_rate": 5.039813403365626e-06,
"loss": 0.8566993474960327,
"step": 1278
},
{
"epoch": 2.7061310782241015,
"grad_norm": 7.661798477172852,
"learning_rate": 5.019071712683423e-06,
"loss": 0.5238276720046997,
"step": 1280
},
{
"epoch": 2.7103594080338267,
"grad_norm": 4.937265872955322,
"learning_rate": 4.998369521397585e-06,
"loss": 0.7930333614349365,
"step": 1282
},
{
"epoch": 2.714587737843552,
"grad_norm": 0.972127377986908,
"learning_rate": 4.977707082600714e-06,
"loss": 0.48816972970962524,
"step": 1284
},
{
"epoch": 2.718816067653277,
"grad_norm": 0.611610472202301,
"learning_rate": 4.957084648899424e-06,
"loss": 0.7112208604812622,
"step": 1286
},
{
"epoch": 2.723044397463002,
"grad_norm": 5.758712291717529,
"learning_rate": 4.936502472411261e-06,
"loss": 0.4793018400669098,
"step": 1288
},
{
"epoch": 2.7272727272727275,
"grad_norm": 2.5748517513275146,
"learning_rate": 4.9159608047615925e-06,
"loss": 0.770174503326416,
"step": 1290
},
{
"epoch": 2.7315010570824523,
"grad_norm": 2.6540465354919434,
"learning_rate": 4.895459897080558e-06,
"loss": 0.8726343512535095,
"step": 1292
},
{
"epoch": 2.7357293868921775,
"grad_norm": 5.4259724617004395,
"learning_rate": 4.875000000000002e-06,
"loss": 0.7556043267250061,
"step": 1294
},
{
"epoch": 2.7399577167019027,
"grad_norm": 2.2050538063049316,
"learning_rate": 4.854581363650374e-06,
"loss": 0.22899200022220612,
"step": 1296
},
{
"epoch": 2.744186046511628,
"grad_norm": 2.445115089416504,
"learning_rate": 4.834204237657707e-06,
"loss": 0.47289910912513733,
"step": 1298
},
{
"epoch": 2.748414376321353,
"grad_norm": 2.6427910327911377,
"learning_rate": 4.813868871140562e-06,
"loss": 0.6392954587936401,
"step": 1300
},
{
"epoch": 2.7526427061310783,
"grad_norm": 4.095657825469971,
"learning_rate": 4.793575512706958e-06,
"loss": 0.5735055208206177,
"step": 1302
},
{
"epoch": 2.7568710359408035,
"grad_norm": 3.827575206756592,
"learning_rate": 4.7733244104513524e-06,
"loss": 0.5476028919219971,
"step": 1304
},
{
"epoch": 2.7610993657505283,
"grad_norm": 8.736998558044434,
"learning_rate": 4.753115811951612e-06,
"loss": 0.6437253952026367,
"step": 1306
},
{
"epoch": 2.765327695560254,
"grad_norm": 4.04536247253418,
"learning_rate": 4.732949964265975e-06,
"loss": 0.7658560872077942,
"step": 1308
},
{
"epoch": 2.7695560253699787,
"grad_norm": 3.036914587020874,
"learning_rate": 4.71282711393002e-06,
"loss": 0.8213181495666504,
"step": 1310
},
{
"epoch": 2.773784355179704,
"grad_norm": 2.982767343521118,
"learning_rate": 4.692747506953686e-06,
"loss": 0.6892085671424866,
"step": 1312
},
{
"epoch": 2.778012684989429,
"grad_norm": 5.73858118057251,
"learning_rate": 4.672711388818235e-06,
"loss": 0.41903311014175415,
"step": 1314
},
{
"epoch": 2.7822410147991543,
"grad_norm": 3.6708431243896484,
"learning_rate": 4.652719004473257e-06,
"loss": 0.7566148638725281,
"step": 1316
},
{
"epoch": 2.7864693446088795,
"grad_norm": 1.1564128398895264,
"learning_rate": 4.632770598333681e-06,
"loss": 0.4328104257583618,
"step": 1318
},
{
"epoch": 2.7906976744186047,
"grad_norm": 3.038012742996216,
"learning_rate": 4.612866414276793e-06,
"loss": 0.6838144063949585,
"step": 1320
},
{
"epoch": 2.79492600422833,
"grad_norm": 7.835508823394775,
"learning_rate": 4.593006695639241e-06,
"loss": 0.2752630114555359,
"step": 1322
},
{
"epoch": 2.7991543340380547,
"grad_norm": 2.6067874431610107,
"learning_rate": 4.573191685214053e-06,
"loss": 0.8349895477294922,
"step": 1324
},
{
"epoch": 2.8033826638477803,
"grad_norm": 20.813175201416016,
"learning_rate": 4.5534216252477015e-06,
"loss": 0.24655112624168396,
"step": 1326
},
{
"epoch": 2.807610993657505,
"grad_norm": 4.950113296508789,
"learning_rate": 4.533696757437112e-06,
"loss": 0.3304949104785919,
"step": 1328
},
{
"epoch": 2.8118393234672303,
"grad_norm": 2.336275815963745,
"learning_rate": 4.514017322926712e-06,
"loss": 0.5199775695800781,
"step": 1330
},
{
"epoch": 2.8160676532769555,
"grad_norm": 3.1804957389831543,
"learning_rate": 4.494383562305501e-06,
"loss": 0.8000071048736572,
"step": 1332
},
{
"epoch": 2.8202959830866807,
"grad_norm": 59.97634506225586,
"learning_rate": 4.474795715604091e-06,
"loss": 0.6024331450462341,
"step": 1334
},
{
"epoch": 2.824524312896406,
"grad_norm": 3.2196907997131348,
"learning_rate": 4.45525402229177e-06,
"loss": 0.8493744134902954,
"step": 1336
},
{
"epoch": 2.828752642706131,
"grad_norm": 3.6626698970794678,
"learning_rate": 4.435758721273596e-06,
"loss": 0.884376049041748,
"step": 1338
},
{
"epoch": 2.8329809725158563,
"grad_norm": 6.249343395233154,
"learning_rate": 4.416310050887456e-06,
"loss": 0.625389575958252,
"step": 1340
},
{
"epoch": 2.8372093023255816,
"grad_norm": 2.2683088779449463,
"learning_rate": 4.396908248901157e-06,
"loss": 0.8382396101951599,
"step": 1342
},
{
"epoch": 2.8414376321353068,
"grad_norm": 2.673769235610962,
"learning_rate": 4.377553552509526e-06,
"loss": 0.610434889793396,
"step": 1344
},
{
"epoch": 2.8456659619450315,
"grad_norm": 22.46831512451172,
"learning_rate": 4.3582461983315e-06,
"loss": 0.701159656047821,
"step": 1346
},
{
"epoch": 2.8498942917547567,
"grad_norm": 3.0735974311828613,
"learning_rate": 4.33898642240724e-06,
"loss": 0.7810471057891846,
"step": 1348
},
{
"epoch": 2.854122621564482,
"grad_norm": 6.111540794372559,
"learning_rate": 4.319774460195247e-06,
"loss": 0.4749992787837982,
"step": 1350
},
{
"epoch": 2.858350951374207,
"grad_norm": 5.31748104095459,
"learning_rate": 4.300610546569476e-06,
"loss": 0.7766779661178589,
"step": 1352
},
{
"epoch": 2.8625792811839323,
"grad_norm": 1.829461693763733,
"learning_rate": 4.281494915816471e-06,
"loss": 0.695534348487854,
"step": 1354
},
{
"epoch": 2.8668076109936576,
"grad_norm": 6.927337646484375,
"learning_rate": 4.262427801632496e-06,
"loss": 0.3583388328552246,
"step": 1356
},
{
"epoch": 2.8710359408033828,
"grad_norm": 4.506462097167969,
"learning_rate": 4.2434094371206845e-06,
"loss": 0.7241630554199219,
"step": 1358
},
{
"epoch": 2.875264270613108,
"grad_norm": 5.312159538269043,
"learning_rate": 4.2244400547881815e-06,
"loss": 0.418176531791687,
"step": 1360
},
{
"epoch": 2.879492600422833,
"grad_norm": 1.7675882577896118,
"learning_rate": 4.205519886543308e-06,
"loss": 0.5930696725845337,
"step": 1362
},
{
"epoch": 2.883720930232558,
"grad_norm": 95.4727554321289,
"learning_rate": 4.186649163692724e-06,
"loss": 0.9458314180374146,
"step": 1364
},
{
"epoch": 2.887949260042283,
"grad_norm": 6.941816806793213,
"learning_rate": 4.167828116938596e-06,
"loss": 0.7628875970840454,
"step": 1366
},
{
"epoch": 2.8921775898520083,
"grad_norm": 2.6443560123443604,
"learning_rate": 4.149056976375781e-06,
"loss": 0.7894702553749084,
"step": 1368
},
{
"epoch": 2.8964059196617336,
"grad_norm": 2.783626079559326,
"learning_rate": 4.130335971489016e-06,
"loss": 0.8570024967193604,
"step": 1370
},
{
"epoch": 2.9006342494714588,
"grad_norm": 3.836874485015869,
"learning_rate": 4.111665331150106e-06,
"loss": 0.8969783782958984,
"step": 1372
},
{
"epoch": 2.904862579281184,
"grad_norm": 2.837597608566284,
"learning_rate": 4.093045283615127e-06,
"loss": 0.7379070520401001,
"step": 1374
},
{
"epoch": 2.909090909090909,
"grad_norm": 2.364428758621216,
"learning_rate": 4.074476056521649e-06,
"loss": 0.8580070734024048,
"step": 1376
},
{
"epoch": 2.9133192389006344,
"grad_norm": 5.514664649963379,
"learning_rate": 4.055957876885926e-06,
"loss": 0.488551527261734,
"step": 1378
},
{
"epoch": 2.9175475687103596,
"grad_norm": 3.8790414333343506,
"learning_rate": 4.037490971100145e-06,
"loss": 0.5080937743186951,
"step": 1380
},
{
"epoch": 2.9217758985200843,
"grad_norm": 3.09446382522583,
"learning_rate": 4.019075564929654e-06,
"loss": 0.8675774931907654,
"step": 1382
},
{
"epoch": 2.92600422832981,
"grad_norm": 5.416377544403076,
"learning_rate": 4.000711883510186e-06,
"loss": 0.6472148299217224,
"step": 1384
},
{
"epoch": 2.9302325581395348,
"grad_norm": 2.695937395095825,
"learning_rate": 3.982400151345126e-06,
"loss": 0.5023255944252014,
"step": 1386
},
{
"epoch": 2.93446088794926,
"grad_norm": 3.863544225692749,
"learning_rate": 3.964140592302755e-06,
"loss": 0.7238068580627441,
"step": 1388
},
{
"epoch": 2.938689217758985,
"grad_norm": 0.5972089171409607,
"learning_rate": 3.9459334296135236e-06,
"loss": 0.41500455141067505,
"step": 1390
},
{
"epoch": 2.9429175475687104,
"grad_norm": 2.5893378257751465,
"learning_rate": 3.927778885867304e-06,
"loss": 0.8614441752433777,
"step": 1392
},
{
"epoch": 2.9471458773784356,
"grad_norm": 1.21278715133667,
"learning_rate": 3.909677183010689e-06,
"loss": 0.1820823848247528,
"step": 1394
},
{
"epoch": 2.951374207188161,
"grad_norm": 4.897063732147217,
"learning_rate": 3.891628542344273e-06,
"loss": 0.45394155383110046,
"step": 1396
},
{
"epoch": 2.955602536997886,
"grad_norm": 98.68574523925781,
"learning_rate": 3.873633184519934e-06,
"loss": 0.7896195650100708,
"step": 1398
},
{
"epoch": 2.9598308668076108,
"grad_norm": 2.2949140071868896,
"learning_rate": 3.855691329538151e-06,
"loss": 1.0011615753173828,
"step": 1400
},
{
"epoch": 2.9640591966173364,
"grad_norm": 9.6893892288208,
"learning_rate": 3.837803196745316e-06,
"loss": 0.6197819709777832,
"step": 1402
},
{
"epoch": 2.968287526427061,
"grad_norm": 16.134605407714844,
"learning_rate": 3.819969004831031e-06,
"loss": 0.4888783395290375,
"step": 1404
},
{
"epoch": 2.9725158562367864,
"grad_norm": 6.059750556945801,
"learning_rate": 3.8021889718254596e-06,
"loss": 0.7471818923950195,
"step": 1406
},
{
"epoch": 2.9767441860465116,
"grad_norm": 2.5034637451171875,
"learning_rate": 3.7844633150966528e-06,
"loss": 0.688784122467041,
"step": 1408
},
{
"epoch": 2.980972515856237,
"grad_norm": 6.228513717651367,
"learning_rate": 3.766792251347883e-06,
"loss": 0.5776879787445068,
"step": 1410
},
{
"epoch": 2.985200845665962,
"grad_norm": 3.410459280014038,
"learning_rate": 3.749175996615e-06,
"loss": 0.81548011302948,
"step": 1412
},
{
"epoch": 2.989429175475687,
"grad_norm": 2.448632001876831,
"learning_rate": 3.7316147662638003e-06,
"loss": 0.5467603206634521,
"step": 1414
},
{
"epoch": 2.9936575052854124,
"grad_norm": 11.337075233459473,
"learning_rate": 3.7141087749873817e-06,
"loss": 0.6863502264022827,
"step": 1416
},
{
"epoch": 2.997885835095137,
"grad_norm": 2.8512160778045654,
"learning_rate": 3.696658236803512e-06,
"loss": 0.45499980449676514,
"step": 1418
},
{
"epoch": 3.0021141649048624,
"grad_norm": 3.1439597606658936,
"learning_rate": 3.6792633650520413e-06,
"loss": 0.7144482135772705,
"step": 1420
},
{
"epoch": 3.0063424947145876,
"grad_norm": 5.3563690185546875,
"learning_rate": 3.6619243723922596e-06,
"loss": 0.4187362492084503,
"step": 1422
},
{
"epoch": 3.010570824524313,
"grad_norm": 6.3570685386657715,
"learning_rate": 3.644641470800321e-06,
"loss": 0.5987511873245239,
"step": 1424
},
{
"epoch": 3.014799154334038,
"grad_norm": 8.241591453552246,
"learning_rate": 3.6274148715666392e-06,
"loss": 0.2966862618923187,
"step": 1426
},
{
"epoch": 3.019027484143763,
"grad_norm": 3.058715343475342,
"learning_rate": 3.6102447852933117e-06,
"loss": 0.3856967091560364,
"step": 1428
},
{
"epoch": 3.0232558139534884,
"grad_norm": 2.451725482940674,
"learning_rate": 3.593131421891541e-06,
"loss": 0.6055225133895874,
"step": 1430
},
{
"epoch": 3.0274841437632136,
"grad_norm": 3.799666404724121,
"learning_rate": 3.5760749905790703e-06,
"loss": 0.6629390716552734,
"step": 1432
},
{
"epoch": 3.031712473572939,
"grad_norm": 4.413266658782959,
"learning_rate": 3.559075699877624e-06,
"loss": 0.4798010587692261,
"step": 1434
},
{
"epoch": 3.035940803382664,
"grad_norm": 12.168867111206055,
"learning_rate": 3.5421337576103625e-06,
"loss": 0.3695339858531952,
"step": 1436
},
{
"epoch": 3.040169133192389,
"grad_norm": 2.4337873458862305,
"learning_rate": 3.525249370899335e-06,
"loss": 0.12947264313697815,
"step": 1438
},
{
"epoch": 3.044397463002114,
"grad_norm": 3.6875813007354736,
"learning_rate": 3.5084227461629525e-06,
"loss": 0.7691330909729004,
"step": 1440
},
{
"epoch": 3.048625792811839,
"grad_norm": 4.5679426193237305,
"learning_rate": 3.4916540891134636e-06,
"loss": 0.40797170996665955,
"step": 1442
},
{
"epoch": 3.0528541226215644,
"grad_norm": 88.38854217529297,
"learning_rate": 3.4749436047544367e-06,
"loss": 0.14900317788124084,
"step": 1444
},
{
"epoch": 3.0570824524312896,
"grad_norm": 5.826979637145996,
"learning_rate": 3.458291497378257e-06,
"loss": 0.3735783100128174,
"step": 1446
},
{
"epoch": 3.061310782241015,
"grad_norm": 26.2069149017334,
"learning_rate": 3.4416979705636272e-06,
"loss": 0.5161640048027039,
"step": 1448
},
{
"epoch": 3.06553911205074,
"grad_norm": 2.800994396209717,
"learning_rate": 3.4251632271730805e-06,
"loss": 0.30497026443481445,
"step": 1450
},
{
"epoch": 3.0697674418604652,
"grad_norm": 4.793062210083008,
"learning_rate": 3.4086874693504953e-06,
"loss": 0.4791848659515381,
"step": 1452
},
{
"epoch": 3.0739957716701904,
"grad_norm": 2.5238945484161377,
"learning_rate": 3.392270898518632e-06,
"loss": 0.6181557178497314,
"step": 1454
},
{
"epoch": 3.0782241014799157,
"grad_norm": 1.6508667469024658,
"learning_rate": 3.375913715376664e-06,
"loss": 0.12339607626199722,
"step": 1456
},
{
"epoch": 3.0824524312896404,
"grad_norm": 2.436734914779663,
"learning_rate": 3.3596161198977267e-06,
"loss": 0.43841099739074707,
"step": 1458
},
{
"epoch": 3.0866807610993656,
"grad_norm": 3.9307093620300293,
"learning_rate": 3.343378311326473e-06,
"loss": 0.6221660375595093,
"step": 1460
},
{
"epoch": 3.090909090909091,
"grad_norm": 8.085297584533691,
"learning_rate": 3.327200488176636e-06,
"loss": 0.39823031425476074,
"step": 1462
},
{
"epoch": 3.095137420718816,
"grad_norm": 2.621410846710205,
"learning_rate": 3.3110828482286036e-06,
"loss": 0.33392971754074097,
"step": 1464
},
{
"epoch": 3.0993657505285412,
"grad_norm": 2.6769158840179443,
"learning_rate": 3.2950255885269988e-06,
"loss": 0.5379525423049927,
"step": 1466
},
{
"epoch": 3.1035940803382664,
"grad_norm": 2.421057939529419,
"learning_rate": 3.279028905378274e-06,
"loss": 0.5930631160736084,
"step": 1468
},
{
"epoch": 3.1078224101479917,
"grad_norm": 3.5492327213287354,
"learning_rate": 3.2630929943483033e-06,
"loss": 0.4956851005554199,
"step": 1470
},
{
"epoch": 3.112050739957717,
"grad_norm": 6.835333824157715,
"learning_rate": 3.2472180502600095e-06,
"loss": 0.45605579018592834,
"step": 1472
},
{
"epoch": 3.116279069767442,
"grad_norm": 3.2966368198394775,
"learning_rate": 3.2314042671909547e-06,
"loss": 0.5774145126342773,
"step": 1474
},
{
"epoch": 3.120507399577167,
"grad_norm": 5.298715591430664,
"learning_rate": 3.2156518384709914e-06,
"loss": 0.6050201654434204,
"step": 1476
},
{
"epoch": 3.124735729386892,
"grad_norm": 8.961942672729492,
"learning_rate": 3.1999609566798935e-06,
"loss": 0.20493540167808533,
"step": 1478
},
{
"epoch": 3.1289640591966172,
"grad_norm": 5.5406317710876465,
"learning_rate": 3.1843318136449904e-06,
"loss": 0.6007043719291687,
"step": 1480
},
{
"epoch": 3.1331923890063424,
"grad_norm": 3.2629332542419434,
"learning_rate": 3.168764600438834e-06,
"loss": 0.66721111536026,
"step": 1482
},
{
"epoch": 3.1374207188160677,
"grad_norm": 2.272597312927246,
"learning_rate": 3.1532595073768675e-06,
"loss": 0.1713816523551941,
"step": 1484
},
{
"epoch": 3.141649048625793,
"grad_norm": 5.638473987579346,
"learning_rate": 3.1378167240150746e-06,
"loss": 0.5887155532836914,
"step": 1486
},
{
"epoch": 3.145877378435518,
"grad_norm": 2.7961437702178955,
"learning_rate": 3.122436439147688e-06,
"loss": 0.777740478515625,
"step": 1488
},
{
"epoch": 3.1501057082452433,
"grad_norm": 14.388794898986816,
"learning_rate": 3.1071188408048715e-06,
"loss": 0.6065592765808105,
"step": 1490
},
{
"epoch": 3.1543340380549685,
"grad_norm": 8.867056846618652,
"learning_rate": 3.091864116250421e-06,
"loss": 0.42025887966156006,
"step": 1492
},
{
"epoch": 3.1585623678646932,
"grad_norm": 2.7912538051605225,
"learning_rate": 3.076672451979466e-06,
"loss": 0.25820159912109375,
"step": 1494
},
{
"epoch": 3.1627906976744184,
"grad_norm": 0.1891254335641861,
"learning_rate": 3.061544033716211e-06,
"loss": 0.3286079466342926,
"step": 1496
},
{
"epoch": 3.1670190274841437,
"grad_norm": 2.4893414974212646,
"learning_rate": 3.0464790464116474e-06,
"loss": 0.501311719417572,
"step": 1498
},
{
"epoch": 3.171247357293869,
"grad_norm": 17.540361404418945,
"learning_rate": 3.0314776742412915e-06,
"loss": 0.6890122890472412,
"step": 1500
},
{
"epoch": 3.175475687103594,
"grad_norm": 5.458293914794922,
"learning_rate": 3.0165401006029515e-06,
"loss": 0.14551039040088654,
"step": 1502
},
{
"epoch": 3.1797040169133193,
"grad_norm": 2.249338388442993,
"learning_rate": 3.0016665081144643e-06,
"loss": 0.22969891130924225,
"step": 1504
},
{
"epoch": 3.1839323467230445,
"grad_norm": 5.115957736968994,
"learning_rate": 2.9868570786114753e-06,
"loss": 0.5196421146392822,
"step": 1506
},
{
"epoch": 3.1881606765327697,
"grad_norm": 6.169950485229492,
"learning_rate": 2.9721119931452044e-06,
"loss": 0.27555975317955017,
"step": 1508
},
{
"epoch": 3.192389006342495,
"grad_norm": 26.197885513305664,
"learning_rate": 2.9574314319802505e-06,
"loss": 0.14574167132377625,
"step": 1510
},
{
"epoch": 3.1966173361522197,
"grad_norm": 0.9431304335594177,
"learning_rate": 2.942815574592373e-06,
"loss": 0.06069979816675186,
"step": 1512
},
{
"epoch": 3.200845665961945,
"grad_norm": 22.147968292236328,
"learning_rate": 2.928264599666294e-06,
"loss": 0.7088733315467834,
"step": 1514
},
{
"epoch": 3.20507399577167,
"grad_norm": 3.5478057861328125,
"learning_rate": 2.9137786850935336e-06,
"loss": 0.5723929405212402,
"step": 1516
},
{
"epoch": 3.2093023255813953,
"grad_norm": 5.536271095275879,
"learning_rate": 2.8993580079702178e-06,
"loss": 0.34792324900627136,
"step": 1518
},
{
"epoch": 3.2135306553911205,
"grad_norm": 9.223129272460938,
"learning_rate": 2.8850027445949116e-06,
"loss": 0.4138156473636627,
"step": 1520
},
{
"epoch": 3.2177589852008457,
"grad_norm": 3.8463335037231445,
"learning_rate": 2.8707130704664836e-06,
"loss": 0.5544077157974243,
"step": 1522
},
{
"epoch": 3.221987315010571,
"grad_norm": 10.065972328186035,
"learning_rate": 2.856489160281939e-06,
"loss": 0.44332054257392883,
"step": 1524
},
{
"epoch": 3.226215644820296,
"grad_norm": 3.8914177417755127,
"learning_rate": 2.842331187934293e-06,
"loss": 0.5320138931274414,
"step": 1526
},
{
"epoch": 3.2304439746300213,
"grad_norm": 11.092885971069336,
"learning_rate": 2.828239326510448e-06,
"loss": 0.09144806861877441,
"step": 1528
},
{
"epoch": 3.234672304439746,
"grad_norm": 7.451350212097168,
"learning_rate": 2.814213748289069e-06,
"loss": 0.609919548034668,
"step": 1530
},
{
"epoch": 3.2389006342494713,
"grad_norm": 4.692660808563232,
"learning_rate": 2.8002546247384868e-06,
"loss": 0.4236515760421753,
"step": 1532
},
{
"epoch": 3.2431289640591965,
"grad_norm": 3.8960964679718018,
"learning_rate": 2.786362126514594e-06,
"loss": 0.5734156966209412,
"step": 1534
},
{
"epoch": 3.2473572938689217,
"grad_norm": 5.420835494995117,
"learning_rate": 2.7725364234587617e-06,
"loss": 0.347150981426239,
"step": 1536
},
{
"epoch": 3.251585623678647,
"grad_norm": 6.251684188842773,
"learning_rate": 2.7587776845957662e-06,
"loss": 0.4469456076622009,
"step": 1538
},
{
"epoch": 3.255813953488372,
"grad_norm": 21.507543563842773,
"learning_rate": 2.7450860781317173e-06,
"loss": 0.3971479535102844,
"step": 1540
},
{
"epoch": 3.2600422832980973,
"grad_norm": 7.660346984863281,
"learning_rate": 2.7314617714520043e-06,
"loss": 0.8031713962554932,
"step": 1542
},
{
"epoch": 3.2642706131078225,
"grad_norm": 2.2891054153442383,
"learning_rate": 2.7179049311192504e-06,
"loss": 0.48143908381462097,
"step": 1544
},
{
"epoch": 3.2684989429175477,
"grad_norm": 4.596741676330566,
"learning_rate": 2.7044157228712737e-06,
"loss": 0.6226749420166016,
"step": 1546
},
{
"epoch": 3.2727272727272725,
"grad_norm": 1.4002420902252197,
"learning_rate": 2.690994311619067e-06,
"loss": 0.11138340830802917,
"step": 1548
},
{
"epoch": 3.276955602536998,
"grad_norm": 4.153841495513916,
"learning_rate": 2.6776408614447735e-06,
"loss": 0.6336830854415894,
"step": 1550
},
{
"epoch": 3.281183932346723,
"grad_norm": 2.388697624206543,
"learning_rate": 2.664355535599685e-06,
"loss": 0.546576976776123,
"step": 1552
},
{
"epoch": 3.285412262156448,
"grad_norm": 0.29166239500045776,
"learning_rate": 2.651138496502252e-06,
"loss": 0.033531758934259415,
"step": 1554
},
{
"epoch": 3.2896405919661733,
"grad_norm": 4.946993350982666,
"learning_rate": 2.6379899057360827e-06,
"loss": 0.5425522327423096,
"step": 1556
},
{
"epoch": 3.2938689217758985,
"grad_norm": 3.1549251079559326,
"learning_rate": 2.6249099240479836e-06,
"loss": 0.32674214243888855,
"step": 1558
},
{
"epoch": 3.2980972515856237,
"grad_norm": 8.850679397583008,
"learning_rate": 2.611898711345988e-06,
"loss": 0.6610650420188904,
"step": 1560
},
{
"epoch": 3.302325581395349,
"grad_norm": 6.184957027435303,
"learning_rate": 2.598956426697395e-06,
"loss": 0.5154082775115967,
"step": 1562
},
{
"epoch": 3.306553911205074,
"grad_norm": 3.3899574279785156,
"learning_rate": 2.586083228326834e-06,
"loss": 0.2964227795600891,
"step": 1564
},
{
"epoch": 3.3107822410147993,
"grad_norm": 4.757692813873291,
"learning_rate": 2.5732792736143315e-06,
"loss": 0.37677574157714844,
"step": 1566
},
{
"epoch": 3.3150105708245245,
"grad_norm": 0.7952932715415955,
"learning_rate": 2.5605447190933735e-06,
"loss": 0.11109252274036407,
"step": 1568
},
{
"epoch": 3.3192389006342493,
"grad_norm": 4.223363876342773,
"learning_rate": 2.547879720449005e-06,
"loss": 0.42762500047683716,
"step": 1570
},
{
"epoch": 3.3234672304439745,
"grad_norm": 7.4136528968811035,
"learning_rate": 2.5352844325159256e-06,
"loss": 0.367853045463562,
"step": 1572
},
{
"epoch": 3.3276955602536997,
"grad_norm": 7.674802780151367,
"learning_rate": 2.522759009276592e-06,
"loss": 0.0927632600069046,
"step": 1574
},
{
"epoch": 3.331923890063425,
"grad_norm": 10.520106315612793,
"learning_rate": 2.51030360385933e-06,
"loss": 0.21945616602897644,
"step": 1576
},
{
"epoch": 3.33615221987315,
"grad_norm": 4.3439178466796875,
"learning_rate": 2.4979183685364768e-06,
"loss": 0.423862487077713,
"step": 1578
},
{
"epoch": 3.3403805496828753,
"grad_norm": 2.7911365032196045,
"learning_rate": 2.485603454722514e-06,
"loss": 0.22814875841140747,
"step": 1580
},
{
"epoch": 3.3446088794926006,
"grad_norm": 8.745915412902832,
"learning_rate": 2.4733590129722057e-06,
"loss": 0.6174420118331909,
"step": 1582
},
{
"epoch": 3.3488372093023258,
"grad_norm": 18.10921287536621,
"learning_rate": 2.4611851929787727e-06,
"loss": 0.4750524163246155,
"step": 1584
},
{
"epoch": 3.353065539112051,
"grad_norm": 8.06883716583252,
"learning_rate": 2.449082143572057e-06,
"loss": 0.3762156069278717,
"step": 1586
},
{
"epoch": 3.3572938689217757,
"grad_norm": 7.110628128051758,
"learning_rate": 2.4370500127167007e-06,
"loss": 0.5135639309883118,
"step": 1588
},
{
"epoch": 3.361522198731501,
"grad_norm": 3.898881673812866,
"learning_rate": 2.425088947510333e-06,
"loss": 0.15028291940689087,
"step": 1590
},
{
"epoch": 3.365750528541226,
"grad_norm": 3.6705503463745117,
"learning_rate": 2.4131990941817855e-06,
"loss": 0.46352654695510864,
"step": 1592
},
{
"epoch": 3.3699788583509513,
"grad_norm": 4.805996894836426,
"learning_rate": 2.4013805980892935e-06,
"loss": 0.2101811170578003,
"step": 1594
},
{
"epoch": 3.3742071881606766,
"grad_norm": 11.132594108581543,
"learning_rate": 2.3896336037187134e-06,
"loss": 0.4875684976577759,
"step": 1596
},
{
"epoch": 3.3784355179704018,
"grad_norm": 8.26980972290039,
"learning_rate": 2.3779582546817767e-06,
"loss": 0.46042075753211975,
"step": 1598
},
{
"epoch": 3.382663847780127,
"grad_norm": 1.4917199611663818,
"learning_rate": 2.366354693714316e-06,
"loss": 0.1829816997051239,
"step": 1600
},
{
"epoch": 3.386892177589852,
"grad_norm": 7.070324897766113,
"learning_rate": 2.3548230626745206e-06,
"loss": 0.3791201114654541,
"step": 1602
},
{
"epoch": 3.3911205073995774,
"grad_norm": 4.70546293258667,
"learning_rate": 2.3433635025412175e-06,
"loss": 0.6322838664054871,
"step": 1604
},
{
"epoch": 3.395348837209302,
"grad_norm": 3.3470864295959473,
"learning_rate": 2.3319761534121324e-06,
"loss": 0.5865774154663086,
"step": 1606
},
{
"epoch": 3.3995771670190273,
"grad_norm": 11.541276931762695,
"learning_rate": 2.320661154502184e-06,
"loss": 0.266635000705719,
"step": 1608
},
{
"epoch": 3.4038054968287526,
"grad_norm": 5.313563346862793,
"learning_rate": 2.309418644141778e-06,
"loss": 0.6351131796836853,
"step": 1610
},
{
"epoch": 3.4080338266384778,
"grad_norm": 3.4072742462158203,
"learning_rate": 2.298248759775121e-06,
"loss": 0.5530030727386475,
"step": 1612
},
{
"epoch": 3.412262156448203,
"grad_norm": 4.182117938995361,
"learning_rate": 2.287151637958538e-06,
"loss": 0.3355867266654968,
"step": 1614
},
{
"epoch": 3.416490486257928,
"grad_norm": 4.2684221267700195,
"learning_rate": 2.276127414358801e-06,
"loss": 0.5294431447982788,
"step": 1616
},
{
"epoch": 3.4207188160676534,
"grad_norm": 3.2420287132263184,
"learning_rate": 2.265176223751472e-06,
"loss": 0.5649456977844238,
"step": 1618
},
{
"epoch": 3.4249471458773786,
"grad_norm": 3.103886127471924,
"learning_rate": 2.254298200019256e-06,
"loss": 0.5623273849487305,
"step": 1620
},
{
"epoch": 3.429175475687104,
"grad_norm": 43.18498229980469,
"learning_rate": 2.243493476150364e-06,
"loss": 0.3995000123977661,
"step": 1622
},
{
"epoch": 3.4334038054968286,
"grad_norm": 5.148665428161621,
"learning_rate": 2.2327621842368864e-06,
"loss": 0.5599923133850098,
"step": 1624
},
{
"epoch": 3.4376321353065538,
"grad_norm": 8.19351863861084,
"learning_rate": 2.222104455473177e-06,
"loss": 0.3380429148674011,
"step": 1626
},
{
"epoch": 3.441860465116279,
"grad_norm": 2.9326305389404297,
"learning_rate": 2.2115204201542526e-06,
"loss": 0.702970027923584,
"step": 1628
},
{
"epoch": 3.446088794926004,
"grad_norm": 7.2235307693481445,
"learning_rate": 2.2010102076741976e-06,
"loss": 0.2136242389678955,
"step": 1630
},
{
"epoch": 3.4503171247357294,
"grad_norm": 5.567283630371094,
"learning_rate": 2.190573946524581e-06,
"loss": 0.3858000934123993,
"step": 1632
},
{
"epoch": 3.4545454545454546,
"grad_norm": 3.113748550415039,
"learning_rate": 2.1802117642928894e-06,
"loss": 0.34387698769569397,
"step": 1634
},
{
"epoch": 3.45877378435518,
"grad_norm": 3.291321039199829,
"learning_rate": 2.169923787660963e-06,
"loss": 0.5708330869674683,
"step": 1636
},
{
"epoch": 3.463002114164905,
"grad_norm": 0.22893865406513214,
"learning_rate": 2.1597101424034486e-06,
"loss": 0.13088123500347137,
"step": 1638
},
{
"epoch": 3.46723044397463,
"grad_norm": 3.1485464572906494,
"learning_rate": 2.149570953386265e-06,
"loss": 0.28203028440475464,
"step": 1640
},
{
"epoch": 3.471458773784355,
"grad_norm": 1.1462241411209106,
"learning_rate": 2.1395063445650702e-06,
"loss": 0.06495456397533417,
"step": 1642
},
{
"epoch": 3.47568710359408,
"grad_norm": 4.584822654724121,
"learning_rate": 2.1295164389837494e-06,
"loss": 0.4998210370540619,
"step": 1644
},
{
"epoch": 3.4799154334038054,
"grad_norm": 3.43422269821167,
"learning_rate": 2.1196013587729142e-06,
"loss": 0.49499210715293884,
"step": 1646
},
{
"epoch": 3.4841437632135306,
"grad_norm": 0.5868573188781738,
"learning_rate": 2.109761225148402e-06,
"loss": 0.3613981008529663,
"step": 1648
},
{
"epoch": 3.488372093023256,
"grad_norm": 3.685621500015259,
"learning_rate": 2.0999961584098e-06,
"loss": 0.3973362445831299,
"step": 1650
},
{
"epoch": 3.492600422832981,
"grad_norm": 7.215812683105469,
"learning_rate": 2.0903062779389724e-06,
"loss": 0.3681108355522156,
"step": 1652
},
{
"epoch": 3.496828752642706,
"grad_norm": 9.877945899963379,
"learning_rate": 2.0806917021986e-06,
"loss": 0.45029395818710327,
"step": 1654
},
{
"epoch": 3.5010570824524314,
"grad_norm": 3.5222997665405273,
"learning_rate": 2.071152548730737e-06,
"loss": 0.5927355289459229,
"step": 1656
},
{
"epoch": 3.5052854122621566,
"grad_norm": 0.48773089051246643,
"learning_rate": 2.061688934155364e-06,
"loss": 0.2075621485710144,
"step": 1658
},
{
"epoch": 3.5095137420718814,
"grad_norm": 2.8444597721099854,
"learning_rate": 2.052300974168971e-06,
"loss": 0.3432048261165619,
"step": 1660
},
{
"epoch": 3.513742071881607,
"grad_norm": 5.066531181335449,
"learning_rate": 2.042988783543144e-06,
"loss": 0.31335684657096863,
"step": 1662
},
{
"epoch": 3.517970401691332,
"grad_norm": 5.156900405883789,
"learning_rate": 2.0337524761231517e-06,
"loss": 0.4727555513381958,
"step": 1664
},
{
"epoch": 3.522198731501057,
"grad_norm": 6.01795768737793,
"learning_rate": 2.024592164826562e-06,
"loss": 0.7001634836196899,
"step": 1666
},
{
"epoch": 3.526427061310782,
"grad_norm": 2.102504014968872,
"learning_rate": 2.015507961641865e-06,
"loss": 0.13448987901210785,
"step": 1668
},
{
"epoch": 3.5306553911205074,
"grad_norm": 0.33504489064216614,
"learning_rate": 2.0064999776270894e-06,
"loss": 0.3235025405883789,
"step": 1670
},
{
"epoch": 3.5348837209302326,
"grad_norm": 6.815194606781006,
"learning_rate": 1.9975683229084617e-06,
"loss": 0.3820517659187317,
"step": 1672
},
{
"epoch": 3.539112050739958,
"grad_norm": 3.395521402359009,
"learning_rate": 1.9887131066790485e-06,
"loss": 0.5702619552612305,
"step": 1674
},
{
"epoch": 3.543340380549683,
"grad_norm": 3.1102519035339355,
"learning_rate": 1.979934437197428e-06,
"loss": 0.2866312861442566,
"step": 1676
},
{
"epoch": 3.547568710359408,
"grad_norm": 4.02962589263916,
"learning_rate": 1.9712324217863566e-06,
"loss": 0.5964412689208984,
"step": 1678
},
{
"epoch": 3.5517970401691334,
"grad_norm": 1.861243724822998,
"learning_rate": 1.962607166831473e-06,
"loss": 0.31753939390182495,
"step": 1680
},
{
"epoch": 3.556025369978858,
"grad_norm": 3.3818070888519287,
"learning_rate": 1.9540587777799843e-06,
"loss": 0.36724668741226196,
"step": 1682
},
{
"epoch": 3.5602536997885834,
"grad_norm": 5.534794807434082,
"learning_rate": 1.945587359139377e-06,
"loss": 0.4233294129371643,
"step": 1684
},
{
"epoch": 3.5644820295983086,
"grad_norm": 28.959678649902344,
"learning_rate": 1.9371930144761517e-06,
"loss": 0.0679631233215332,
"step": 1686
},
{
"epoch": 3.568710359408034,
"grad_norm": 5.814289569854736,
"learning_rate": 1.9288758464145427e-06,
"loss": 0.4481062591075897,
"step": 1688
},
{
"epoch": 3.572938689217759,
"grad_norm": 0.8135766983032227,
"learning_rate": 1.920635956635271e-06,
"loss": 0.02256624586880207,
"step": 1690
},
{
"epoch": 3.5771670190274842,
"grad_norm": 9.225582122802734,
"learning_rate": 1.9124734458743e-06,
"loss": 0.2678327262401581,
"step": 1692
},
{
"epoch": 3.5813953488372094,
"grad_norm": 7.333009719848633,
"learning_rate": 1.9043884139216031e-06,
"loss": 0.4615964889526367,
"step": 1694
},
{
"epoch": 3.585623678646934,
"grad_norm": 5.408736228942871,
"learning_rate": 1.8963809596199432e-06,
"loss": 0.19673492014408112,
"step": 1696
},
{
"epoch": 3.58985200845666,
"grad_norm": 3.9907760620117188,
"learning_rate": 1.8884511808636671e-06,
"loss": 0.4891815483570099,
"step": 1698
},
{
"epoch": 3.5940803382663846,
"grad_norm": 4.118829727172852,
"learning_rate": 1.8805991745975062e-06,
"loss": 0.8720452785491943,
"step": 1700
},
{
"epoch": 3.59830866807611,
"grad_norm": 12.184479713439941,
"learning_rate": 1.8728250368153906e-06,
"loss": 0.5542669892311096,
"step": 1702
},
{
"epoch": 3.602536997885835,
"grad_norm": 8.63931655883789,
"learning_rate": 1.8651288625592794e-06,
"loss": 0.20688292384147644,
"step": 1704
},
{
"epoch": 3.6067653276955602,
"grad_norm": 2.81652569770813,
"learning_rate": 1.8575107459179937e-06,
"loss": 0.2420281618833542,
"step": 1706
},
{
"epoch": 3.6109936575052854,
"grad_norm": 5.55596399307251,
"learning_rate": 1.8499707800260706e-06,
"loss": 0.6276252269744873,
"step": 1708
},
{
"epoch": 3.6152219873150107,
"grad_norm": 10.919977188110352,
"learning_rate": 1.8425090570626223e-06,
"loss": 0.4683901071548462,
"step": 1710
},
{
"epoch": 3.619450317124736,
"grad_norm": 3.751389980316162,
"learning_rate": 1.8351256682502103e-06,
"loss": 0.24191835522651672,
"step": 1712
},
{
"epoch": 3.6236786469344606,
"grad_norm": 1.3749680519104004,
"learning_rate": 1.827820703853728e-06,
"loss": 0.1549309939146042,
"step": 1714
},
{
"epoch": 3.6279069767441863,
"grad_norm": 7.281813144683838,
"learning_rate": 1.8205942531793012e-06,
"loss": 0.4165268540382385,
"step": 1716
},
{
"epoch": 3.632135306553911,
"grad_norm": 7.609879493713379,
"learning_rate": 1.813446404573192e-06,
"loss": 0.2930583357810974,
"step": 1718
},
{
"epoch": 3.6363636363636362,
"grad_norm": 4.279430866241455,
"learning_rate": 1.8063772454207212e-06,
"loss": 0.15954896807670593,
"step": 1720
},
{
"epoch": 3.6405919661733614,
"grad_norm": 4.053138256072998,
"learning_rate": 1.7993868621451993e-06,
"loss": 0.4065016806125641,
"step": 1722
},
{
"epoch": 3.6448202959830867,
"grad_norm": 3.9610989093780518,
"learning_rate": 1.7924753402068716e-06,
"loss": 0.48019081354141235,
"step": 1724
},
{
"epoch": 3.649048625792812,
"grad_norm": 3.0792617797851562,
"learning_rate": 1.7856427641018695e-06,
"loss": 0.6954116821289062,
"step": 1726
},
{
"epoch": 3.653276955602537,
"grad_norm": 5.730588436126709,
"learning_rate": 1.7788892173611808e-06,
"loss": 0.7138341665267944,
"step": 1728
},
{
"epoch": 3.6575052854122623,
"grad_norm": 9.384326934814453,
"learning_rate": 1.7722147825496282e-06,
"loss": 0.6630492806434631,
"step": 1730
},
{
"epoch": 3.6617336152219875,
"grad_norm": 4.569153308868408,
"learning_rate": 1.7656195412648589e-06,
"loss": 0.41630300879478455,
"step": 1732
},
{
"epoch": 3.6659619450317127,
"grad_norm": 5.186647891998291,
"learning_rate": 1.7591035741363465e-06,
"loss": 0.6005408763885498,
"step": 1734
},
{
"epoch": 3.6701902748414374,
"grad_norm": 4.728799819946289,
"learning_rate": 1.7526669608244076e-06,
"loss": 0.46103861927986145,
"step": 1736
},
{
"epoch": 3.6744186046511627,
"grad_norm": 6.2252984046936035,
"learning_rate": 1.7463097800192268e-06,
"loss": 0.08872466534376144,
"step": 1738
},
{
"epoch": 3.678646934460888,
"grad_norm": 3.8498623371124268,
"learning_rate": 1.7400321094398917e-06,
"loss": 0.2832813858985901,
"step": 1740
},
{
"epoch": 3.682875264270613,
"grad_norm": 5.854602813720703,
"learning_rate": 1.733834025833447e-06,
"loss": 0.26969748735427856,
"step": 1742
},
{
"epoch": 3.6871035940803383,
"grad_norm": 5.287208557128906,
"learning_rate": 1.7277156049739573e-06,
"loss": 0.5257506370544434,
"step": 1744
},
{
"epoch": 3.6913319238900635,
"grad_norm": 0.405897855758667,
"learning_rate": 1.7216769216615738e-06,
"loss": 0.2664951682090759,
"step": 1746
},
{
"epoch": 3.6955602536997887,
"grad_norm": 11.106414794921875,
"learning_rate": 1.7157180497216258e-06,
"loss": 0.2913108468055725,
"step": 1748
},
{
"epoch": 3.699788583509514,
"grad_norm": 4.00687837600708,
"learning_rate": 1.7098390620037193e-06,
"loss": 0.9774203896522522,
"step": 1750
},
{
"epoch": 3.704016913319239,
"grad_norm": 4.831068992614746,
"learning_rate": 1.7040400303808388e-06,
"loss": 0.1801181137561798,
"step": 1752
},
{
"epoch": 3.708245243128964,
"grad_norm": 3.756962299346924,
"learning_rate": 1.6983210257484758e-06,
"loss": 0.3975505828857422,
"step": 1754
},
{
"epoch": 3.712473572938689,
"grad_norm": 5.675543785095215,
"learning_rate": 1.69268211802376e-06,
"loss": 0.4291342496871948,
"step": 1756
},
{
"epoch": 3.7167019027484143,
"grad_norm": 4.510821342468262,
"learning_rate": 1.6871233761446057e-06,
"loss": 0.30917859077453613,
"step": 1758
},
{
"epoch": 3.7209302325581395,
"grad_norm": 1.0369709730148315,
"learning_rate": 1.6816448680688613e-06,
"loss": 0.24939927458763123,
"step": 1760
},
{
"epoch": 3.7251585623678647,
"grad_norm": 4.5799031257629395,
"learning_rate": 1.6762466607734931e-06,
"loss": 0.4708609879016876,
"step": 1762
},
{
"epoch": 3.72938689217759,
"grad_norm": 2.7423105239868164,
"learning_rate": 1.6709288202537525e-06,
"loss": 0.3550443947315216,
"step": 1764
},
{
"epoch": 3.733615221987315,
"grad_norm": 14.882494926452637,
"learning_rate": 1.665691411522376e-06,
"loss": 0.5527515411376953,
"step": 1766
},
{
"epoch": 3.7378435517970403,
"grad_norm": 3.920135498046875,
"learning_rate": 1.6605344986087893e-06,
"loss": 0.42553144693374634,
"step": 1768
},
{
"epoch": 3.7420718816067655,
"grad_norm": 0.629116952419281,
"learning_rate": 1.655458144558326e-06,
"loss": 0.08822059631347656,
"step": 1770
},
{
"epoch": 3.7463002114164903,
"grad_norm": 4.265275478363037,
"learning_rate": 1.6504624114314536e-06,
"loss": 0.3891627788543701,
"step": 1772
},
{
"epoch": 3.7505285412262155,
"grad_norm": 9.811651229858398,
"learning_rate": 1.6455473603030148e-06,
"loss": 0.6639033555984497,
"step": 1774
},
{
"epoch": 3.7547568710359407,
"grad_norm": 5.193588733673096,
"learning_rate": 1.6407130512614863e-06,
"loss": 0.50983726978302,
"step": 1776
},
{
"epoch": 3.758985200845666,
"grad_norm": 4.611268520355225,
"learning_rate": 1.6359595434082377e-06,
"loss": 0.5774052739143372,
"step": 1778
},
{
"epoch": 3.763213530655391,
"grad_norm": 6.4507622718811035,
"learning_rate": 1.6312868948568114e-06,
"loss": 0.6459259986877441,
"step": 1780
},
{
"epoch": 3.7674418604651163,
"grad_norm": 3.9018239974975586,
"learning_rate": 1.6266951627322143e-06,
"loss": 0.6416741609573364,
"step": 1782
},
{
"epoch": 3.7716701902748415,
"grad_norm": 12.602282524108887,
"learning_rate": 1.622184403170217e-06,
"loss": 0.13825801014900208,
"step": 1784
},
{
"epoch": 3.7758985200845667,
"grad_norm": 3.8429384231567383,
"learning_rate": 1.6177546713166654e-06,
"loss": 0.3972879946231842,
"step": 1786
},
{
"epoch": 3.780126849894292,
"grad_norm": 5.463750839233398,
"learning_rate": 1.6134060213268122e-06,
"loss": 0.2276698350906372,
"step": 1788
},
{
"epoch": 3.7843551797040167,
"grad_norm": 4.239009380340576,
"learning_rate": 1.609138506364651e-06,
"loss": 0.06166570633649826,
"step": 1790
},
{
"epoch": 3.7885835095137423,
"grad_norm": 0.6214521527290344,
"learning_rate": 1.6049521786022663e-06,
"loss": 0.3091050386428833,
"step": 1792
},
{
"epoch": 3.792811839323467,
"grad_norm": 3.232224225997925,
"learning_rate": 1.6008470892191971e-06,
"loss": 0.6148475408554077,
"step": 1794
},
{
"epoch": 3.7970401691331923,
"grad_norm": 8.090229034423828,
"learning_rate": 1.5968232884018094e-06,
"loss": 0.2800947427749634,
"step": 1796
},
{
"epoch": 3.8012684989429175,
"grad_norm": 28.455608367919922,
"learning_rate": 1.592880825342686e-06,
"loss": 0.5017866492271423,
"step": 1798
},
{
"epoch": 3.8054968287526427,
"grad_norm": 4.026337623596191,
"learning_rate": 1.5890197482400214e-06,
"loss": 0.6740830540657043,
"step": 1800
},
{
"epoch": 3.809725158562368,
"grad_norm": 3.508122444152832,
"learning_rate": 1.5852401042970336e-06,
"loss": 0.3761969208717346,
"step": 1802
},
{
"epoch": 3.813953488372093,
"grad_norm": 3.400318145751953,
"learning_rate": 1.5815419397213902e-06,
"loss": 0.5787155032157898,
"step": 1804
},
{
"epoch": 3.8181818181818183,
"grad_norm": 4.089182376861572,
"learning_rate": 1.577925299724637e-06,
"loss": 0.5465168356895447,
"step": 1806
},
{
"epoch": 3.822410147991543,
"grad_norm": 41.6027717590332,
"learning_rate": 1.5743902285216514e-06,
"loss": 0.1728125363588333,
"step": 1808
},
{
"epoch": 3.8266384778012688,
"grad_norm": 3.3004062175750732,
"learning_rate": 1.5709367693300995e-06,
"loss": 0.5649695992469788,
"step": 1810
},
{
"epoch": 3.8308668076109935,
"grad_norm": 4.2993669509887695,
"learning_rate": 1.5675649643699071e-06,
"loss": 0.6549540758132935,
"step": 1812
},
{
"epoch": 3.8350951374207187,
"grad_norm": 3.4883182048797607,
"learning_rate": 1.5642748548627444e-06,
"loss": 0.5488921999931335,
"step": 1814
},
{
"epoch": 3.839323467230444,
"grad_norm": 3.6274850368499756,
"learning_rate": 1.5610664810315224e-06,
"loss": 0.3673040270805359,
"step": 1816
},
{
"epoch": 3.843551797040169,
"grad_norm": 3.024437665939331,
"learning_rate": 1.5579398820998995e-06,
"loss": 0.3205223083496094,
"step": 1818
},
{
"epoch": 3.8477801268498943,
"grad_norm": 3.776864528656006,
"learning_rate": 1.5548950962918043e-06,
"loss": 0.29514163732528687,
"step": 1820
},
{
"epoch": 3.8520084566596196,
"grad_norm": 0.9230905771255493,
"learning_rate": 1.5519321608309674e-06,
"loss": 0.3962177038192749,
"step": 1822
},
{
"epoch": 3.8562367864693448,
"grad_norm": 3.7600820064544678,
"learning_rate": 1.549051111940463e-06,
"loss": 0.3158684968948364,
"step": 1824
},
{
"epoch": 3.8604651162790695,
"grad_norm": 8.070271492004395,
"learning_rate": 1.5462519848422735e-06,
"loss": 0.33815300464630127,
"step": 1826
},
{
"epoch": 3.864693446088795,
"grad_norm": 5.189460754394531,
"learning_rate": 1.5435348137568503e-06,
"loss": 0.5764117240905762,
"step": 1828
},
{
"epoch": 3.86892177589852,
"grad_norm": 3.977822780609131,
"learning_rate": 1.540899631902703e-06,
"loss": 0.6190783381462097,
"step": 1830
},
{
"epoch": 3.873150105708245,
"grad_norm": 19.55247688293457,
"learning_rate": 1.5383464714959885e-06,
"loss": 0.3775050640106201,
"step": 1832
},
{
"epoch": 3.8773784355179703,
"grad_norm": 5.020406246185303,
"learning_rate": 1.5358753637501168e-06,
"loss": 0.28668826818466187,
"step": 1834
},
{
"epoch": 3.8816067653276956,
"grad_norm": 0.40400922298431396,
"learning_rate": 1.5334863388753753e-06,
"loss": 0.27497953176498413,
"step": 1836
},
{
"epoch": 3.8858350951374208,
"grad_norm": 2.1234359741210938,
"learning_rate": 1.5311794260785508e-06,
"loss": 0.26325470209121704,
"step": 1838
},
{
"epoch": 3.890063424947146,
"grad_norm": 2.732037305831909,
"learning_rate": 1.5289546535625803e-06,
"loss": 0.3790256381034851,
"step": 1840
},
{
"epoch": 3.894291754756871,
"grad_norm": 7.596894264221191,
"learning_rate": 1.5268120485262017e-06,
"loss": 0.46870318055152893,
"step": 1842
},
{
"epoch": 3.898520084566596,
"grad_norm": 2.9562246799468994,
"learning_rate": 1.5247516371636211e-06,
"loss": 0.32340067625045776,
"step": 1844
},
{
"epoch": 3.9027484143763216,
"grad_norm": 18.76889419555664,
"learning_rate": 1.522773444664197e-06,
"loss": 0.3392444849014282,
"step": 1846
},
{
"epoch": 3.9069767441860463,
"grad_norm": 3.1390750408172607,
"learning_rate": 1.5208774952121262e-06,
"loss": 0.10644078254699707,
"step": 1848
},
{
"epoch": 3.9112050739957716,
"grad_norm": 2.6275129318237305,
"learning_rate": 1.5190638119861523e-06,
"loss": 0.5070509910583496,
"step": 1850
},
{
"epoch": 3.9154334038054968,
"grad_norm": 1.2279460430145264,
"learning_rate": 1.5173324171592825e-06,
"loss": 0.32292094826698303,
"step": 1852
},
{
"epoch": 3.919661733615222,
"grad_norm": 3.8160648345947266,
"learning_rate": 1.5156833318985132e-06,
"loss": 0.568835973739624,
"step": 1854
},
{
"epoch": 3.923890063424947,
"grad_norm": 5.367692470550537,
"learning_rate": 1.514116576364574e-06,
"loss": 0.22992059588432312,
"step": 1856
},
{
"epoch": 3.9281183932346724,
"grad_norm": 5.4259185791015625,
"learning_rate": 1.5126321697116824e-06,
"loss": 0.5054709911346436,
"step": 1858
},
{
"epoch": 3.9323467230443976,
"grad_norm": 8.427145957946777,
"learning_rate": 1.511230130087305e-06,
"loss": 0.21720188856124878,
"step": 1860
},
{
"epoch": 3.9365750528541223,
"grad_norm": 18.63742446899414,
"learning_rate": 1.509910474631939e-06,
"loss": 0.8239214420318604,
"step": 1862
},
{
"epoch": 3.940803382663848,
"grad_norm": 3.083559513092041,
"learning_rate": 1.5086732194789031e-06,
"loss": 0.36028850078582764,
"step": 1864
},
{
"epoch": 3.9450317124735728,
"grad_norm": 5.646785736083984,
"learning_rate": 1.507518379754139e-06,
"loss": 0.5543577671051025,
"step": 1866
},
{
"epoch": 3.949260042283298,
"grad_norm": 0.31760188937187195,
"learning_rate": 1.5064459695760255e-06,
"loss": 0.01628248207271099,
"step": 1868
},
{
"epoch": 3.953488372093023,
"grad_norm": 3.5817031860351562,
"learning_rate": 1.5054560020552078e-06,
"loss": 0.5692353248596191,
"step": 1870
},
{
"epoch": 3.9577167019027484,
"grad_norm": 5.22877311706543,
"learning_rate": 1.504548489294438e-06,
"loss": 0.3443927764892578,
"step": 1872
},
{
"epoch": 3.9619450317124736,
"grad_norm": 5.386042594909668,
"learning_rate": 1.5037234423884224e-06,
"loss": 0.4191185235977173,
"step": 1874
},
{
"epoch": 3.966173361522199,
"grad_norm": 3.5231735706329346,
"learning_rate": 1.5029808714236927e-06,
"loss": 0.5892931222915649,
"step": 1876
},
{
"epoch": 3.970401691331924,
"grad_norm": 3.3802387714385986,
"learning_rate": 1.502320785478475e-06,
"loss": 0.12032198905944824,
"step": 1878
},
{
"epoch": 3.974630021141649,
"grad_norm": 3.0940866470336914,
"learning_rate": 1.5017431926225876e-06,
"loss": 0.2673117518424988,
"step": 1880
},
{
"epoch": 3.9788583509513744,
"grad_norm": 6.358997821807861,
"learning_rate": 1.5012480999173337e-06,
"loss": 0.520773708820343,
"step": 1882
},
{
"epoch": 3.983086680761099,
"grad_norm": 3.1239089965820312,
"learning_rate": 1.5008355134154204e-06,
"loss": 0.7278933525085449,
"step": 1884
},
{
"epoch": 3.9873150105708244,
"grad_norm": 2.787248373031616,
"learning_rate": 1.500505438160885e-06,
"loss": 0.4279649257659912,
"step": 1886
},
{
"epoch": 3.9915433403805496,
"grad_norm": 2.290093421936035,
"learning_rate": 1.500257878189028e-06,
"loss": 0.5085483193397522,
"step": 1888
},
{
"epoch": 3.995771670190275,
"grad_norm": 3.092968702316284,
"learning_rate": 1.5000928365263716e-06,
"loss": 0.5331004858016968,
"step": 1890
},
{
"epoch": 4.0,
"grad_norm": 4.6352009773254395,
"learning_rate": 1.5000103151906145e-06,
"loss": 0.2658572196960449,
"step": 1892
},
{
"epoch": 4.0,
"step": 1892,
"total_flos": 3.6314083289926205e+18,
"train_loss": 0.773373347370399,
"train_runtime": 4468.5675,
"train_samples_per_second": 12.702,
"train_steps_per_second": 0.423
}
],
"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": 3.6314083289926205e+18,
"train_batch_size": 5,
"trial_name": null,
"trial_params": null
}