9b-96 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
f4bf28a verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1892,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004228329809725159,
"grad_norm": 4.105223178863525,
"learning_rate": 1.0526315789473685e-07,
"loss": 1.8271780014038086,
"step": 2
},
{
"epoch": 0.008456659619450317,
"grad_norm": 1.2513997554779053,
"learning_rate": 3.1578947368421055e-07,
"loss": 2.2135629653930664,
"step": 4
},
{
"epoch": 0.012684989429175475,
"grad_norm": 11.961859703063965,
"learning_rate": 5.263157894736843e-07,
"loss": 2.206533193588257,
"step": 6
},
{
"epoch": 0.016913319238900635,
"grad_norm": 1.6793477535247803,
"learning_rate": 7.368421052631579e-07,
"loss": 1.9962220191955566,
"step": 8
},
{
"epoch": 0.021141649048625793,
"grad_norm": 8.071287155151367,
"learning_rate": 9.473684210526317e-07,
"loss": 2.0124125480651855,
"step": 10
},
{
"epoch": 0.02536997885835095,
"grad_norm": 2.324535608291626,
"learning_rate": 1.1578947368421053e-06,
"loss": 1.6356844902038574,
"step": 12
},
{
"epoch": 0.02959830866807611,
"grad_norm": 1.4699110984802246,
"learning_rate": 1.3684210526315791e-06,
"loss": 1.6418602466583252,
"step": 14
},
{
"epoch": 0.03382663847780127,
"grad_norm": 1.9240294694900513,
"learning_rate": 1.5789473684210526e-06,
"loss": 1.7639219760894775,
"step": 16
},
{
"epoch": 0.03805496828752643,
"grad_norm": 1.6100047826766968,
"learning_rate": 1.7894736842105265e-06,
"loss": 1.7936522960662842,
"step": 18
},
{
"epoch": 0.042283298097251586,
"grad_norm": 5.262239933013916,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.654106855392456,
"step": 20
},
{
"epoch": 0.046511627906976744,
"grad_norm": 4.83996057510376,
"learning_rate": 2.2105263157894738e-06,
"loss": 1.475663423538208,
"step": 22
},
{
"epoch": 0.0507399577167019,
"grad_norm": 2.40704607963562,
"learning_rate": 2.4210526315789477e-06,
"loss": 1.758310317993164,
"step": 24
},
{
"epoch": 0.05496828752642706,
"grad_norm": 3.138162612915039,
"learning_rate": 2.631578947368421e-06,
"loss": 1.3428184986114502,
"step": 26
},
{
"epoch": 0.05919661733615222,
"grad_norm": 0.876343309879303,
"learning_rate": 2.842105263157895e-06,
"loss": 1.6434693336486816,
"step": 28
},
{
"epoch": 0.06342494714587738,
"grad_norm": 1.211041808128357,
"learning_rate": 3.052631578947369e-06,
"loss": 1.211287021636963,
"step": 30
},
{
"epoch": 0.06765327695560254,
"grad_norm": 5.724530220031738,
"learning_rate": 3.2631578947368423e-06,
"loss": 0.9725242257118225,
"step": 32
},
{
"epoch": 0.07188160676532769,
"grad_norm": 1.2683931589126587,
"learning_rate": 3.473684210526316e-06,
"loss": 1.1639443635940552,
"step": 34
},
{
"epoch": 0.07610993657505286,
"grad_norm": 1.168879747390747,
"learning_rate": 3.6842105263157896e-06,
"loss": 1.3444358110427856,
"step": 36
},
{
"epoch": 0.080338266384778,
"grad_norm": 5.301766872406006,
"learning_rate": 3.894736842105263e-06,
"loss": 1.2748759984970093,
"step": 38
},
{
"epoch": 0.08456659619450317,
"grad_norm": 3.4816555976867676,
"learning_rate": 4.105263157894737e-06,
"loss": 1.2032517194747925,
"step": 40
},
{
"epoch": 0.08879492600422834,
"grad_norm": 2.4880456924438477,
"learning_rate": 4.315789473684211e-06,
"loss": 1.6828070878982544,
"step": 42
},
{
"epoch": 0.09302325581395349,
"grad_norm": 1.8762389421463013,
"learning_rate": 4.526315789473685e-06,
"loss": 1.5913126468658447,
"step": 44
},
{
"epoch": 0.09725158562367865,
"grad_norm": 3.2679994106292725,
"learning_rate": 4.736842105263158e-06,
"loss": 1.2594860792160034,
"step": 46
},
{
"epoch": 0.1014799154334038,
"grad_norm": 0.7650827765464783,
"learning_rate": 4.947368421052632e-06,
"loss": 1.496073603630066,
"step": 48
},
{
"epoch": 0.10570824524312897,
"grad_norm": 1.628807544708252,
"learning_rate": 5.157894736842106e-06,
"loss": 1.4602724313735962,
"step": 50
},
{
"epoch": 0.10993657505285412,
"grad_norm": 2.712521553039551,
"learning_rate": 5.36842105263158e-06,
"loss": 0.6989067792892456,
"step": 52
},
{
"epoch": 0.11416490486257928,
"grad_norm": 0.7859939336776733,
"learning_rate": 5.578947368421052e-06,
"loss": 1.037105917930603,
"step": 54
},
{
"epoch": 0.11839323467230443,
"grad_norm": 1.3851209878921509,
"learning_rate": 5.789473684210527e-06,
"loss": 1.3567808866500854,
"step": 56
},
{
"epoch": 0.1226215644820296,
"grad_norm": 1.1566178798675537,
"learning_rate": 6e-06,
"loss": 1.3738900423049927,
"step": 58
},
{
"epoch": 0.12684989429175475,
"grad_norm": 0.9807841777801514,
"learning_rate": 6.2105263157894745e-06,
"loss": 1.3903249502182007,
"step": 60
},
{
"epoch": 0.13107822410147993,
"grad_norm": 2.721123218536377,
"learning_rate": 6.421052631578948e-06,
"loss": 0.769786536693573,
"step": 62
},
{
"epoch": 0.13530655391120508,
"grad_norm": 1.080932855606079,
"learning_rate": 6.631578947368421e-06,
"loss": 1.0346770286560059,
"step": 64
},
{
"epoch": 0.13953488372093023,
"grad_norm": 0.7417203783988953,
"learning_rate": 6.842105263157896e-06,
"loss": 1.333381175994873,
"step": 66
},
{
"epoch": 0.14376321353065538,
"grad_norm": 1.1657756567001343,
"learning_rate": 7.052631578947369e-06,
"loss": 1.364691972732544,
"step": 68
},
{
"epoch": 0.14799154334038056,
"grad_norm": 1.7600194215774536,
"learning_rate": 7.263157894736843e-06,
"loss": 1.3450676202774048,
"step": 70
},
{
"epoch": 0.1522198731501057,
"grad_norm": 1.1755248308181763,
"learning_rate": 7.473684210526316e-06,
"loss": 1.3401083946228027,
"step": 72
},
{
"epoch": 0.15644820295983086,
"grad_norm": 0.9038915038108826,
"learning_rate": 7.68421052631579e-06,
"loss": 0.9320542216300964,
"step": 74
},
{
"epoch": 0.160676532769556,
"grad_norm": 1.20891273021698,
"learning_rate": 7.894736842105265e-06,
"loss": 0.8452156782150269,
"step": 76
},
{
"epoch": 0.1649048625792812,
"grad_norm": 1.9456303119659424,
"learning_rate": 8.105263157894736e-06,
"loss": 1.2799372673034668,
"step": 78
},
{
"epoch": 0.16913319238900634,
"grad_norm": 3.8374783992767334,
"learning_rate": 8.315789473684212e-06,
"loss": 1.0811553001403809,
"step": 80
},
{
"epoch": 0.1733615221987315,
"grad_norm": 1.1618369817733765,
"learning_rate": 8.526315789473685e-06,
"loss": 1.0897631645202637,
"step": 82
},
{
"epoch": 0.17758985200845667,
"grad_norm": 0.6783086657524109,
"learning_rate": 8.736842105263158e-06,
"loss": 0.8950634598731995,
"step": 84
},
{
"epoch": 0.18181818181818182,
"grad_norm": 1.9268966913223267,
"learning_rate": 8.947368421052632e-06,
"loss": 1.711360216140747,
"step": 86
},
{
"epoch": 0.18604651162790697,
"grad_norm": 1.1067707538604736,
"learning_rate": 9.157894736842105e-06,
"loss": 1.302412509918213,
"step": 88
},
{
"epoch": 0.19027484143763213,
"grad_norm": 2.0564188957214355,
"learning_rate": 9.36842105263158e-06,
"loss": 0.9922522306442261,
"step": 90
},
{
"epoch": 0.1945031712473573,
"grad_norm": 0.915518581867218,
"learning_rate": 9.578947368421054e-06,
"loss": 1.3106107711791992,
"step": 92
},
{
"epoch": 0.19873150105708245,
"grad_norm": 0.9490787386894226,
"learning_rate": 9.789473684210527e-06,
"loss": 0.9769561886787415,
"step": 94
},
{
"epoch": 0.2029598308668076,
"grad_norm": 1.47053062915802,
"learning_rate": 1e-05,
"loss": 1.2871181964874268,
"step": 96
},
{
"epoch": 0.20718816067653276,
"grad_norm": 0.8488267064094543,
"learning_rate": 9.999972492846047e-06,
"loss": 1.3827238082885742,
"step": 98
},
{
"epoch": 0.21141649048625794,
"grad_norm": 1.216845154762268,
"learning_rate": 9.999889971720468e-06,
"loss": 1.3255081176757812,
"step": 100
},
{
"epoch": 0.2156448202959831,
"grad_norm": 0.5432454943656921,
"learning_rate": 9.999752437632121e-06,
"loss": 1.242600440979004,
"step": 102
},
{
"epoch": 0.21987315010570824,
"grad_norm": 1.1232975721359253,
"learning_rate": 9.999559892262412e-06,
"loss": 1.3049225807189941,
"step": 104
},
{
"epoch": 0.22410147991543342,
"grad_norm": 1.0933849811553955,
"learning_rate": 9.999312337965291e-06,
"loss": 1.2947238683700562,
"step": 106
},
{
"epoch": 0.22832980972515857,
"grad_norm": 1.5993964672088623,
"learning_rate": 9.999009777767202e-06,
"loss": 1.0347710847854614,
"step": 108
},
{
"epoch": 0.23255813953488372,
"grad_norm": 1.020851731300354,
"learning_rate": 9.99865221536707e-06,
"loss": 1.280065894126892,
"step": 110
},
{
"epoch": 0.23678646934460887,
"grad_norm": 2.3744354248046875,
"learning_rate": 9.998239655136237e-06,
"loss": 1.3588924407958984,
"step": 112
},
{
"epoch": 0.24101479915433405,
"grad_norm": 1.7851115465164185,
"learning_rate": 9.997772102118417e-06,
"loss": 1.4594029188156128,
"step": 114
},
{
"epoch": 0.2452431289640592,
"grad_norm": 1.4111109972000122,
"learning_rate": 9.997249562029637e-06,
"loss": 1.3292509317398071,
"step": 116
},
{
"epoch": 0.24947145877378435,
"grad_norm": 1.4081361293792725,
"learning_rate": 9.996672041258154e-06,
"loss": 1.342902660369873,
"step": 118
},
{
"epoch": 0.2536997885835095,
"grad_norm": 1.5050036907196045,
"learning_rate": 9.996039546864395e-06,
"loss": 0.846265435218811,
"step": 120
},
{
"epoch": 0.25792811839323465,
"grad_norm": 1.0461825132369995,
"learning_rate": 9.995352086580858e-06,
"loss": 1.2984052896499634,
"step": 122
},
{
"epoch": 0.26215644820295986,
"grad_norm": 1.1487557888031006,
"learning_rate": 9.994609668812022e-06,
"loss": 1.1421293020248413,
"step": 124
},
{
"epoch": 0.266384778012685,
"grad_norm": 1.1190460920333862,
"learning_rate": 9.993812302634237e-06,
"loss": 1.3438360691070557,
"step": 126
},
{
"epoch": 0.27061310782241016,
"grad_norm": 1.2282752990722656,
"learning_rate": 9.99295999779563e-06,
"loss": 0.9613946080207825,
"step": 128
},
{
"epoch": 0.2748414376321353,
"grad_norm": 1.2685223817825317,
"learning_rate": 9.99205276471597e-06,
"loss": 0.978096604347229,
"step": 130
},
{
"epoch": 0.27906976744186046,
"grad_norm": 1.4892326593399048,
"learning_rate": 9.991090614486546e-06,
"loss": 1.2701090574264526,
"step": 132
},
{
"epoch": 0.2832980972515856,
"grad_norm": 2.3711142539978027,
"learning_rate": 9.990073558870028e-06,
"loss": 0.8357971906661987,
"step": 134
},
{
"epoch": 0.28752642706131076,
"grad_norm": 10.921707153320312,
"learning_rate": 9.989001610300333e-06,
"loss": 1.111595630645752,
"step": 136
},
{
"epoch": 0.2917547568710359,
"grad_norm": 1.9601994752883911,
"learning_rate": 9.987874781882461e-06,
"loss": 1.3744772672653198,
"step": 138
},
{
"epoch": 0.2959830866807611,
"grad_norm": 1.1790173053741455,
"learning_rate": 9.986693087392341e-06,
"loss": 0.9393970966339111,
"step": 140
},
{
"epoch": 0.30021141649048627,
"grad_norm": 1.0423545837402344,
"learning_rate": 9.985456541276667e-06,
"loss": 1.5640604496002197,
"step": 142
},
{
"epoch": 0.3044397463002114,
"grad_norm": 6.126891613006592,
"learning_rate": 9.984165158652708e-06,
"loss": 1.0898505449295044,
"step": 144
},
{
"epoch": 0.3086680761099366,
"grad_norm": 6.716033935546875,
"learning_rate": 9.982818955308135e-06,
"loss": 0.689599871635437,
"step": 146
},
{
"epoch": 0.3128964059196617,
"grad_norm": 1.7806477546691895,
"learning_rate": 9.98141794770083e-06,
"loss": 1.2405976057052612,
"step": 148
},
{
"epoch": 0.3171247357293869,
"grad_norm": 0.7157177329063416,
"learning_rate": 9.979962152958669e-06,
"loss": 1.2606029510498047,
"step": 150
},
{
"epoch": 0.321353065539112,
"grad_norm": 1.4935792684555054,
"learning_rate": 9.97845158887933e-06,
"loss": 1.175560712814331,
"step": 152
},
{
"epoch": 0.32558139534883723,
"grad_norm": 3.346512794494629,
"learning_rate": 9.976886273930063e-06,
"loss": 0.7485346794128418,
"step": 154
},
{
"epoch": 0.3298097251585624,
"grad_norm": 1.140684723854065,
"learning_rate": 9.975266227247474e-06,
"loss": 1.2501227855682373,
"step": 156
},
{
"epoch": 0.33403805496828753,
"grad_norm": 1.6993314027786255,
"learning_rate": 9.973591468637286e-06,
"loss": 1.2625560760498047,
"step": 158
},
{
"epoch": 0.3382663847780127,
"grad_norm": 1.1285232305526733,
"learning_rate": 9.971862018574094e-06,
"loss": 1.2512822151184082,
"step": 160
},
{
"epoch": 0.34249471458773784,
"grad_norm": 1.880086898803711,
"learning_rate": 9.97007789820112e-06,
"loss": 1.0875740051269531,
"step": 162
},
{
"epoch": 0.346723044397463,
"grad_norm": 3.3863887786865234,
"learning_rate": 9.968239129329953e-06,
"loss": 0.9627373218536377,
"step": 164
},
{
"epoch": 0.35095137420718814,
"grad_norm": 1.1893935203552246,
"learning_rate": 9.96634573444028e-06,
"loss": 0.8808426856994629,
"step": 166
},
{
"epoch": 0.35517970401691334,
"grad_norm": 1.085954189300537,
"learning_rate": 9.964397736679615e-06,
"loss": 1.1921576261520386,
"step": 168
},
{
"epoch": 0.3594080338266385,
"grad_norm": 1.157077431678772,
"learning_rate": 9.962395159863014e-06,
"loss": 1.2048157453536987,
"step": 170
},
{
"epoch": 0.36363636363636365,
"grad_norm": 1.1349718570709229,
"learning_rate": 9.960338028472781e-06,
"loss": 1.0401729345321655,
"step": 172
},
{
"epoch": 0.3678646934460888,
"grad_norm": 1.9062402248382568,
"learning_rate": 9.958226367658176e-06,
"loss": 1.348514437675476,
"step": 174
},
{
"epoch": 0.37209302325581395,
"grad_norm": 2.720722198486328,
"learning_rate": 9.9560602032351e-06,
"loss": 1.0518444776535034,
"step": 176
},
{
"epoch": 0.3763213530655391,
"grad_norm": 1.4164613485336304,
"learning_rate": 9.953839561685783e-06,
"loss": 0.8652096390724182,
"step": 178
},
{
"epoch": 0.38054968287526425,
"grad_norm": 0.7160219550132751,
"learning_rate": 9.95156447015846e-06,
"loss": 1.2788941860198975,
"step": 180
},
{
"epoch": 0.38477801268498946,
"grad_norm": 1.2482486963272095,
"learning_rate": 9.94923495646704e-06,
"loss": 1.517291784286499,
"step": 182
},
{
"epoch": 0.3890063424947146,
"grad_norm": 0.986268162727356,
"learning_rate": 9.946851049090764e-06,
"loss": 1.3300813436508179,
"step": 184
},
{
"epoch": 0.39323467230443976,
"grad_norm": 1.6990145444869995,
"learning_rate": 9.944412777173854e-06,
"loss": 1.111851453781128,
"step": 186
},
{
"epoch": 0.3974630021141649,
"grad_norm": 1.0885865688323975,
"learning_rate": 9.941920170525168e-06,
"loss": 0.9844686985015869,
"step": 188
},
{
"epoch": 0.40169133192389006,
"grad_norm": 2.5655248165130615,
"learning_rate": 9.939373259617822e-06,
"loss": 1.0147863626480103,
"step": 190
},
{
"epoch": 0.4059196617336152,
"grad_norm": 1.623833179473877,
"learning_rate": 9.936772075588825e-06,
"loss": 1.2935187816619873,
"step": 192
},
{
"epoch": 0.41014799154334036,
"grad_norm": 0.5006506443023682,
"learning_rate": 9.934116650238697e-06,
"loss": 0.998938798904419,
"step": 194
},
{
"epoch": 0.4143763213530655,
"grad_norm": 3.403019905090332,
"learning_rate": 9.931407016031078e-06,
"loss": 1.277458667755127,
"step": 196
},
{
"epoch": 0.4186046511627907,
"grad_norm": 1.192270040512085,
"learning_rate": 9.928643206092338e-06,
"loss": 1.0763307809829712,
"step": 198
},
{
"epoch": 0.42283298097251587,
"grad_norm": 0.7673647403717041,
"learning_rate": 9.925825254211161e-06,
"loss": 1.1796720027923584,
"step": 200
},
{
"epoch": 0.427061310782241,
"grad_norm": 0.8333643078804016,
"learning_rate": 9.922953194838143e-06,
"loss": 0.595544159412384,
"step": 202
},
{
"epoch": 0.4312896405919662,
"grad_norm": 0.8162744045257568,
"learning_rate": 9.920027063085363e-06,
"loss": 1.2246547937393188,
"step": 204
},
{
"epoch": 0.4355179704016913,
"grad_norm": 4.353880405426025,
"learning_rate": 9.917046894725957e-06,
"loss": 1.28725266456604,
"step": 206
},
{
"epoch": 0.4397463002114165,
"grad_norm": 4.166569709777832,
"learning_rate": 9.91401272619368e-06,
"loss": 0.7904917001724243,
"step": 208
},
{
"epoch": 0.4439746300211416,
"grad_norm": 1.12864351272583,
"learning_rate": 9.910924594582463e-06,
"loss": 1.2832318544387817,
"step": 210
},
{
"epoch": 0.44820295983086683,
"grad_norm": 1.1492962837219238,
"learning_rate": 9.907782537645955e-06,
"loss": 1.2079336643218994,
"step": 212
},
{
"epoch": 0.452431289640592,
"grad_norm": 1.7490490674972534,
"learning_rate": 9.904586593797062e-06,
"loss": 1.416234016418457,
"step": 214
},
{
"epoch": 0.45665961945031713,
"grad_norm": 0.8846709728240967,
"learning_rate": 9.901336802107485e-06,
"loss": 0.6584013104438782,
"step": 216
},
{
"epoch": 0.4608879492600423,
"grad_norm": 1.1017366647720337,
"learning_rate": 9.89803320230723e-06,
"loss": 1.0077871084213257,
"step": 218
},
{
"epoch": 0.46511627906976744,
"grad_norm": 0.9772974848747253,
"learning_rate": 9.894675834784136e-06,
"loss": 1.2816599607467651,
"step": 220
},
{
"epoch": 0.4693446088794926,
"grad_norm": 1.021376609802246,
"learning_rate": 9.891264740583367e-06,
"loss": 1.2406489849090576,
"step": 222
},
{
"epoch": 0.47357293868921774,
"grad_norm": 1.221967101097107,
"learning_rate": 9.88779996140692e-06,
"loss": 1.3515628576278687,
"step": 224
},
{
"epoch": 0.47780126849894294,
"grad_norm": 1.059727430343628,
"learning_rate": 9.884281539613109e-06,
"loss": 1.2479000091552734,
"step": 226
},
{
"epoch": 0.4820295983086681,
"grad_norm": 0.924156665802002,
"learning_rate": 9.880709518216061e-06,
"loss": 1.244722604751587,
"step": 228
},
{
"epoch": 0.48625792811839325,
"grad_norm": 1.0760767459869385,
"learning_rate": 9.87708394088517e-06,
"loss": 0.9330758452415466,
"step": 230
},
{
"epoch": 0.4904862579281184,
"grad_norm": 0.985787034034729,
"learning_rate": 9.873404851944573e-06,
"loss": 1.2493349313735962,
"step": 232
},
{
"epoch": 0.49471458773784355,
"grad_norm": 2.560053825378418,
"learning_rate": 9.869672296372611e-06,
"loss": 0.960759162902832,
"step": 234
},
{
"epoch": 0.4989429175475687,
"grad_norm": 2.581348419189453,
"learning_rate": 9.86588631980128e-06,
"loss": 0.9124109745025635,
"step": 236
},
{
"epoch": 0.5031712473572939,
"grad_norm": 7.913112163543701,
"learning_rate": 9.86204696851566e-06,
"loss": 0.8496532440185547,
"step": 238
},
{
"epoch": 0.507399577167019,
"grad_norm": 1.270423412322998,
"learning_rate": 9.858154289453363e-06,
"loss": 0.9334291815757751,
"step": 240
},
{
"epoch": 0.5116279069767442,
"grad_norm": 1.3697165250778198,
"learning_rate": 9.854208330203958e-06,
"loss": 1.3218450546264648,
"step": 242
},
{
"epoch": 0.5158562367864693,
"grad_norm": 4.063036918640137,
"learning_rate": 9.850209139008377e-06,
"loss": 1.341506004333496,
"step": 244
},
{
"epoch": 0.5200845665961945,
"grad_norm": 1.9559390544891357,
"learning_rate": 9.846156764758344e-06,
"loss": 1.283069372177124,
"step": 246
},
{
"epoch": 0.5243128964059197,
"grad_norm": 2.6484060287475586,
"learning_rate": 9.842051256995759e-06,
"loss": 1.239783763885498,
"step": 248
},
{
"epoch": 0.5285412262156448,
"grad_norm": 1.0665085315704346,
"learning_rate": 9.837892665912105e-06,
"loss": 1.187054991722107,
"step": 250
},
{
"epoch": 0.53276955602537,
"grad_norm": 6.941689968109131,
"learning_rate": 9.833681042347827e-06,
"loss": 1.162398099899292,
"step": 252
},
{
"epoch": 0.5369978858350951,
"grad_norm": 1.6367284059524536,
"learning_rate": 9.829416437791721e-06,
"loss": 1.741904854774475,
"step": 254
},
{
"epoch": 0.5412262156448203,
"grad_norm": 0.8215171098709106,
"learning_rate": 9.825098904380284e-06,
"loss": 1.2098015546798706,
"step": 256
},
{
"epoch": 0.5454545454545454,
"grad_norm": 2.12286639213562,
"learning_rate": 9.820728494897101e-06,
"loss": 1.3860020637512207,
"step": 258
},
{
"epoch": 0.5496828752642706,
"grad_norm": 1.9879218339920044,
"learning_rate": 9.816305262772182e-06,
"loss": 0.9947566986083984,
"step": 260
},
{
"epoch": 0.5539112050739958,
"grad_norm": 1.455824613571167,
"learning_rate": 9.811829262081315e-06,
"loss": 1.1925487518310547,
"step": 262
},
{
"epoch": 0.5581395348837209,
"grad_norm": 1.1726953983306885,
"learning_rate": 9.807300547545411e-06,
"loss": 1.0478217601776123,
"step": 264
},
{
"epoch": 0.5623678646934461,
"grad_norm": 1.2227460145950317,
"learning_rate": 9.802719174529821e-06,
"loss": 1.4977850914001465,
"step": 266
},
{
"epoch": 0.5665961945031712,
"grad_norm": 2.1803297996520996,
"learning_rate": 9.798085199043672e-06,
"loss": 0.5466461777687073,
"step": 268
},
{
"epoch": 0.5708245243128964,
"grad_norm": 1.7832331657409668,
"learning_rate": 9.793398677739177e-06,
"loss": 1.0137481689453125,
"step": 270
},
{
"epoch": 0.5750528541226215,
"grad_norm": 1.6292634010314941,
"learning_rate": 9.788659667910942e-06,
"loss": 1.1755725145339966,
"step": 272
},
{
"epoch": 0.5792811839323467,
"grad_norm": 1.0001522302627563,
"learning_rate": 9.783868227495262e-06,
"loss": 0.955671489238739,
"step": 274
},
{
"epoch": 0.5835095137420718,
"grad_norm": 2.0499613285064697,
"learning_rate": 9.779024415069427e-06,
"loss": 0.9340962171554565,
"step": 276
},
{
"epoch": 0.587737843551797,
"grad_norm": 0.47681549191474915,
"learning_rate": 9.774128289850987e-06,
"loss": 0.8637338876724243,
"step": 278
},
{
"epoch": 0.5919661733615222,
"grad_norm": 1.2753183841705322,
"learning_rate": 9.769179911697037e-06,
"loss": 1.2291476726531982,
"step": 280
},
{
"epoch": 0.5961945031712473,
"grad_norm": 2.742964506149292,
"learning_rate": 9.764179341103493e-06,
"loss": 1.4667253494262695,
"step": 282
},
{
"epoch": 0.6004228329809725,
"grad_norm": 0.35357359051704407,
"learning_rate": 9.759126639204337e-06,
"loss": 0.9396631717681885,
"step": 284
},
{
"epoch": 0.6046511627906976,
"grad_norm": 1.436333417892456,
"learning_rate": 9.75402186777088e-06,
"loss": 1.2948578596115112,
"step": 286
},
{
"epoch": 0.6088794926004228,
"grad_norm": 0.9579204320907593,
"learning_rate": 9.748865089211004e-06,
"loss": 0.4360913038253784,
"step": 288
},
{
"epoch": 0.6131078224101479,
"grad_norm": 4.398628234863281,
"learning_rate": 9.743656366568398e-06,
"loss": 0.9474999904632568,
"step": 290
},
{
"epoch": 0.6173361522198731,
"grad_norm": 1.3529343605041504,
"learning_rate": 9.738395763521792e-06,
"loss": 0.8793157339096069,
"step": 292
},
{
"epoch": 0.6215644820295984,
"grad_norm": 1.2661802768707275,
"learning_rate": 9.733083344384164e-06,
"loss": 0.7495632171630859,
"step": 294
},
{
"epoch": 0.6257928118393234,
"grad_norm": 1.323224663734436,
"learning_rate": 9.727719174101981e-06,
"loss": 0.7435659170150757,
"step": 296
},
{
"epoch": 0.6300211416490487,
"grad_norm": 1.1257333755493164,
"learning_rate": 9.722303318254376e-06,
"loss": 1.2757129669189453,
"step": 298
},
{
"epoch": 0.6342494714587738,
"grad_norm": 2.0354418754577637,
"learning_rate": 9.71683584305236e-06,
"loss": 1.2378349304199219,
"step": 300
},
{
"epoch": 0.638477801268499,
"grad_norm": 1.3787568807601929,
"learning_rate": 9.71131681533802e-06,
"loss": 0.746528685092926,
"step": 302
},
{
"epoch": 0.642706131078224,
"grad_norm": 4.07403564453125,
"learning_rate": 9.705746302583682e-06,
"loss": 0.8612475395202637,
"step": 304
},
{
"epoch": 0.6469344608879493,
"grad_norm": 1.457455039024353,
"learning_rate": 9.700124372891104e-06,
"loss": 0.9252447485923767,
"step": 306
},
{
"epoch": 0.6511627906976745,
"grad_norm": 1.0487290620803833,
"learning_rate": 9.694451094990636e-06,
"loss": 1.2314348220825195,
"step": 308
},
{
"epoch": 0.6553911205073996,
"grad_norm": 2.5076074600219727,
"learning_rate": 9.68872653824038e-06,
"loss": 1.2077559232711792,
"step": 310
},
{
"epoch": 0.6596194503171248,
"grad_norm": 1.4133344888687134,
"learning_rate": 9.68295077262534e-06,
"loss": 1.2097148895263672,
"step": 312
},
{
"epoch": 0.6638477801268499,
"grad_norm": 1.1078712940216064,
"learning_rate": 9.677123868756573e-06,
"loss": 1.2286163568496704,
"step": 314
},
{
"epoch": 0.6680761099365751,
"grad_norm": 1.0299714803695679,
"learning_rate": 9.671245897870318e-06,
"loss": 1.0676769018173218,
"step": 316
},
{
"epoch": 0.6723044397463002,
"grad_norm": 1.8400555849075317,
"learning_rate": 9.665316931827131e-06,
"loss": 0.7731718420982361,
"step": 318
},
{
"epoch": 0.6765327695560254,
"grad_norm": 1.8411188125610352,
"learning_rate": 9.659337043111006e-06,
"loss": 1.2415677309036255,
"step": 320
},
{
"epoch": 0.6807610993657506,
"grad_norm": 1.0365511178970337,
"learning_rate": 9.653306304828483e-06,
"loss": 1.3031237125396729,
"step": 322
},
{
"epoch": 0.6849894291754757,
"grad_norm": 2.106027364730835,
"learning_rate": 9.647224790707762e-06,
"loss": 1.0795551538467407,
"step": 324
},
{
"epoch": 0.6892177589852009,
"grad_norm": 3.0075478553771973,
"learning_rate": 9.641092575097795e-06,
"loss": 1.2615439891815186,
"step": 326
},
{
"epoch": 0.693446088794926,
"grad_norm": 7.821414470672607,
"learning_rate": 9.63490973296738e-06,
"loss": 1.1690138578414917,
"step": 328
},
{
"epoch": 0.6976744186046512,
"grad_norm": 2.229149103164673,
"learning_rate": 9.62867633990425e-06,
"loss": 0.8666979670524597,
"step": 330
},
{
"epoch": 0.7019027484143763,
"grad_norm": 1.2929322719573975,
"learning_rate": 9.622392472114136e-06,
"loss": 1.2522735595703125,
"step": 332
},
{
"epoch": 0.7061310782241015,
"grad_norm": 2.3682351112365723,
"learning_rate": 9.61605820641985e-06,
"loss": 1.2169489860534668,
"step": 334
},
{
"epoch": 0.7103594080338267,
"grad_norm": 1.423481822013855,
"learning_rate": 9.60967362026033e-06,
"loss": 1.204700231552124,
"step": 336
},
{
"epoch": 0.7145877378435518,
"grad_norm": 1.141951322555542,
"learning_rate": 9.60323879168971e-06,
"loss": 1.3039082288742065,
"step": 338
},
{
"epoch": 0.718816067653277,
"grad_norm": 2.867393970489502,
"learning_rate": 9.596753799376355e-06,
"loss": 1.20618736743927,
"step": 340
},
{
"epoch": 0.7230443974630021,
"grad_norm": 1.9708638191223145,
"learning_rate": 9.5902187226019e-06,
"loss": 1.1258702278137207,
"step": 342
},
{
"epoch": 0.7272727272727273,
"grad_norm": 1.4016013145446777,
"learning_rate": 9.583633641260285e-06,
"loss": 0.909141480922699,
"step": 344
},
{
"epoch": 0.7315010570824524,
"grad_norm": 3.1992838382720947,
"learning_rate": 9.576998635856774e-06,
"loss": 0.8569663166999817,
"step": 346
},
{
"epoch": 0.7357293868921776,
"grad_norm": 1.7542674541473389,
"learning_rate": 9.570313787506974e-06,
"loss": 0.9917646050453186,
"step": 348
},
{
"epoch": 0.7399577167019028,
"grad_norm": 1.1175165176391602,
"learning_rate": 9.563579177935844e-06,
"loss": 1.2983906269073486,
"step": 350
},
{
"epoch": 0.7441860465116279,
"grad_norm": 0.9748227596282959,
"learning_rate": 9.556794889476687e-06,
"loss": 1.1592947244644165,
"step": 352
},
{
"epoch": 0.7484143763213531,
"grad_norm": 2.805424451828003,
"learning_rate": 9.549961005070161e-06,
"loss": 0.7118173837661743,
"step": 354
},
{
"epoch": 0.7526427061310782,
"grad_norm": 1.4081275463104248,
"learning_rate": 9.543077608263245e-06,
"loss": 1.0245922803878784,
"step": 356
},
{
"epoch": 0.7568710359408034,
"grad_norm": 0.8380882143974304,
"learning_rate": 9.536144783208229e-06,
"loss": 1.206932544708252,
"step": 358
},
{
"epoch": 0.7610993657505285,
"grad_norm": 7.674241542816162,
"learning_rate": 9.529162614661687e-06,
"loss": 0.7563204169273376,
"step": 360
},
{
"epoch": 0.7653276955602537,
"grad_norm": 1.5167875289916992,
"learning_rate": 9.522131187983433e-06,
"loss": 1.094705581665039,
"step": 362
},
{
"epoch": 0.7695560253699789,
"grad_norm": 2.942284345626831,
"learning_rate": 9.515050589135485e-06,
"loss": 1.04949951171875,
"step": 364
},
{
"epoch": 0.773784355179704,
"grad_norm": 1.4387823343276978,
"learning_rate": 9.507920904681006e-06,
"loss": 1.2584033012390137,
"step": 366
},
{
"epoch": 0.7780126849894292,
"grad_norm": 1.334485411643982,
"learning_rate": 9.500742221783257e-06,
"loss": 1.2004865407943726,
"step": 368
},
{
"epoch": 0.7822410147991543,
"grad_norm": 2.1047496795654297,
"learning_rate": 9.493514628204516e-06,
"loss": 0.87196284532547,
"step": 370
},
{
"epoch": 0.7864693446088795,
"grad_norm": 3.5297274589538574,
"learning_rate": 9.48623821230502e-06,
"loss": 1.1859652996063232,
"step": 372
},
{
"epoch": 0.7906976744186046,
"grad_norm": 1.4089442491531372,
"learning_rate": 9.478913063041878e-06,
"loss": 1.2051024436950684,
"step": 374
},
{
"epoch": 0.7949260042283298,
"grad_norm": 4.6978302001953125,
"learning_rate": 9.471539269967982e-06,
"loss": 1.2903268337249756,
"step": 376
},
{
"epoch": 0.7991543340380549,
"grad_norm": 2.450777053833008,
"learning_rate": 9.464116923230913e-06,
"loss": 0.9673522710800171,
"step": 378
},
{
"epoch": 0.8033826638477801,
"grad_norm": 1.6850031614303589,
"learning_rate": 9.456646113571844e-06,
"loss": 1.3038182258605957,
"step": 380
},
{
"epoch": 0.8076109936575053,
"grad_norm": 1.281440258026123,
"learning_rate": 9.449126932324424e-06,
"loss": 1.1666007041931152,
"step": 382
},
{
"epoch": 0.8118393234672304,
"grad_norm": 1.2754701375961304,
"learning_rate": 9.441559471413665e-06,
"loss": 0.9718303084373474,
"step": 384
},
{
"epoch": 0.8160676532769556,
"grad_norm": 1.8220703601837158,
"learning_rate": 9.433943823354816e-06,
"loss": 0.9527886509895325,
"step": 386
},
{
"epoch": 0.8202959830866807,
"grad_norm": 0.90151047706604,
"learning_rate": 9.426280081252232e-06,
"loss": 1.4771757125854492,
"step": 388
},
{
"epoch": 0.8245243128964059,
"grad_norm": 6.917442798614502,
"learning_rate": 9.418568338798245e-06,
"loss": 1.149660587310791,
"step": 390
},
{
"epoch": 0.828752642706131,
"grad_norm": 1.9025379419326782,
"learning_rate": 9.410808690271999e-06,
"loss": 0.9310383796691895,
"step": 392
},
{
"epoch": 0.8329809725158562,
"grad_norm": 1.3193538188934326,
"learning_rate": 9.40300123053832e-06,
"loss": 1.2344201803207397,
"step": 394
},
{
"epoch": 0.8372093023255814,
"grad_norm": 0.23406380414962769,
"learning_rate": 9.395146055046535e-06,
"loss": 0.7611551880836487,
"step": 396
},
{
"epoch": 0.8414376321353065,
"grad_norm": 2.347813129425049,
"learning_rate": 9.387243259829323e-06,
"loss": 0.6888359785079956,
"step": 398
},
{
"epoch": 0.8456659619450317,
"grad_norm": 1.8717073202133179,
"learning_rate": 9.37929294150153e-06,
"loss": 1.1256269216537476,
"step": 400
},
{
"epoch": 0.8498942917547568,
"grad_norm": 1.3918894529342651,
"learning_rate": 9.371295197258992e-06,
"loss": 0.9689964652061462,
"step": 402
},
{
"epoch": 0.854122621564482,
"grad_norm": 1.2325738668441772,
"learning_rate": 9.363250124877345e-06,
"loss": 0.8262078762054443,
"step": 404
},
{
"epoch": 0.8583509513742071,
"grad_norm": 1.6549572944641113,
"learning_rate": 9.35515782271083e-06,
"loss": 0.9803326725959778,
"step": 406
},
{
"epoch": 0.8625792811839323,
"grad_norm": 0.7307443618774414,
"learning_rate": 9.347018389691093e-06,
"loss": 0.9950054883956909,
"step": 408
},
{
"epoch": 0.8668076109936576,
"grad_norm": 1.3734959363937378,
"learning_rate": 9.338831925325974e-06,
"loss": 0.8809871077537537,
"step": 410
},
{
"epoch": 0.8710359408033826,
"grad_norm": 12.435693740844727,
"learning_rate": 9.330598529698287e-06,
"loss": 1.1730930805206299,
"step": 412
},
{
"epoch": 0.8752642706131079,
"grad_norm": 1.826246738433838,
"learning_rate": 9.322318303464603e-06,
"loss": 1.020997166633606,
"step": 414
},
{
"epoch": 0.879492600422833,
"grad_norm": 1.7890242338180542,
"learning_rate": 9.313991347854015e-06,
"loss": 1.1168156862258911,
"step": 416
},
{
"epoch": 0.8837209302325582,
"grad_norm": 1.1680430173873901,
"learning_rate": 9.305617764666897e-06,
"loss": 1.2124841213226318,
"step": 418
},
{
"epoch": 0.8879492600422833,
"grad_norm": 1.165459156036377,
"learning_rate": 9.297197656273672e-06,
"loss": 1.2253860235214233,
"step": 420
},
{
"epoch": 0.8921775898520085,
"grad_norm": 35.40209197998047,
"learning_rate": 9.288731125613544e-06,
"loss": 0.9485971927642822,
"step": 422
},
{
"epoch": 0.8964059196617337,
"grad_norm": 1.066243290901184,
"learning_rate": 9.280218276193255e-06,
"loss": 1.2243362665176392,
"step": 424
},
{
"epoch": 0.9006342494714588,
"grad_norm": 0.9742780923843384,
"learning_rate": 9.271659212085807e-06,
"loss": 1.0899114608764648,
"step": 426
},
{
"epoch": 0.904862579281184,
"grad_norm": 7.175337314605713,
"learning_rate": 9.263054037929199e-06,
"loss": 0.8689170479774475,
"step": 428
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.9477011561393738,
"learning_rate": 9.254402858925142e-06,
"loss": 1.1052967309951782,
"step": 430
},
{
"epoch": 0.9133192389006343,
"grad_norm": 2.36238431930542,
"learning_rate": 9.245705780837774e-06,
"loss": 1.2545251846313477,
"step": 432
},
{
"epoch": 0.9175475687103594,
"grad_norm": 2.5181665420532227,
"learning_rate": 9.236962909992368e-06,
"loss": 0.5464211702346802,
"step": 434
},
{
"epoch": 0.9217758985200846,
"grad_norm": 7.958632946014404,
"learning_rate": 9.228174353274037e-06,
"loss": 0.37406569719314575,
"step": 436
},
{
"epoch": 0.9260042283298098,
"grad_norm": 0.7444133758544922,
"learning_rate": 9.219340218126413e-06,
"loss": 0.9482316374778748,
"step": 438
},
{
"epoch": 0.9302325581395349,
"grad_norm": 1.134406566619873,
"learning_rate": 9.210460612550348e-06,
"loss": 1.1393942832946777,
"step": 440
},
{
"epoch": 0.9344608879492601,
"grad_norm": 8.245950698852539,
"learning_rate": 9.20153564510259e-06,
"loss": 0.7637899518013,
"step": 442
},
{
"epoch": 0.9386892177589852,
"grad_norm": 1.4876855611801147,
"learning_rate": 9.19256542489445e-06,
"loss": 1.2526780366897583,
"step": 444
},
{
"epoch": 0.9429175475687104,
"grad_norm": 0.977704644203186,
"learning_rate": 9.183550061590476e-06,
"loss": 1.0953787565231323,
"step": 446
},
{
"epoch": 0.9471458773784355,
"grad_norm": 3.1618423461914062,
"learning_rate": 9.174489665407103e-06,
"loss": 1.115494966506958,
"step": 448
},
{
"epoch": 0.9513742071881607,
"grad_norm": 3.4997429847717285,
"learning_rate": 9.165384347111317e-06,
"loss": 0.8016799092292786,
"step": 450
},
{
"epoch": 0.9556025369978859,
"grad_norm": 2.415138006210327,
"learning_rate": 9.156234218019291e-06,
"loss": 0.923848032951355,
"step": 452
},
{
"epoch": 0.959830866807611,
"grad_norm": 1.7008107900619507,
"learning_rate": 9.147039389995029e-06,
"loss": 1.3221611976623535,
"step": 454
},
{
"epoch": 0.9640591966173362,
"grad_norm": 1.9487804174423218,
"learning_rate": 9.137799975448998e-06,
"loss": 0.699728786945343,
"step": 456
},
{
"epoch": 0.9682875264270613,
"grad_norm": 1.6703341007232666,
"learning_rate": 9.128516087336755e-06,
"loss": 0.8970375061035156,
"step": 458
},
{
"epoch": 0.9725158562367865,
"grad_norm": 1.3661073446273804,
"learning_rate": 9.119187839157559e-06,
"loss": 1.2227510213851929,
"step": 460
},
{
"epoch": 0.9767441860465116,
"grad_norm": 1.6562296152114868,
"learning_rate": 9.109815344952995e-06,
"loss": 1.2037882804870605,
"step": 462
},
{
"epoch": 0.9809725158562368,
"grad_norm": 4.352667808532715,
"learning_rate": 9.100398719305566e-06,
"loss": 0.7387683391571045,
"step": 464
},
{
"epoch": 0.985200845665962,
"grad_norm": 1.662271499633789,
"learning_rate": 9.090938077337305e-06,
"loss": 1.1843740940093994,
"step": 466
},
{
"epoch": 0.9894291754756871,
"grad_norm": 7.772258281707764,
"learning_rate": 9.081433534708363e-06,
"loss": 0.992527425289154,
"step": 468
},
{
"epoch": 0.9936575052854123,
"grad_norm": 1.0815784931182861,
"learning_rate": 9.071885207615591e-06,
"loss": 1.2383248805999756,
"step": 470
},
{
"epoch": 0.9978858350951374,
"grad_norm": 1.6808488368988037,
"learning_rate": 9.062293212791123e-06,
"loss": 1.24074125289917,
"step": 472
},
{
"epoch": 1.0021141649048626,
"grad_norm": 2.4606966972351074,
"learning_rate": 9.052657667500951e-06,
"loss": 0.990084707736969,
"step": 474
},
{
"epoch": 1.0063424947145878,
"grad_norm": 1.1899473667144775,
"learning_rate": 9.042978689543487e-06,
"loss": 1.1488051414489746,
"step": 476
},
{
"epoch": 1.0105708245243128,
"grad_norm": 1.8924003839492798,
"learning_rate": 9.033256397248124e-06,
"loss": 0.46862325072288513,
"step": 478
},
{
"epoch": 1.014799154334038,
"grad_norm": 1.4806647300720215,
"learning_rate": 9.02349090947379e-06,
"loss": 0.7944547533988953,
"step": 480
},
{
"epoch": 1.0190274841437632,
"grad_norm": 1.3868756294250488,
"learning_rate": 9.013682345607502e-06,
"loss": 0.6805719137191772,
"step": 482
},
{
"epoch": 1.0232558139534884,
"grad_norm": 1.2194361686706543,
"learning_rate": 9.003830825562887e-06,
"loss": 0.9927952885627747,
"step": 484
},
{
"epoch": 1.0274841437632136,
"grad_norm": 2.6877315044403076,
"learning_rate": 8.99393646977874e-06,
"loss": 1.0203206539154053,
"step": 486
},
{
"epoch": 1.0317124735729386,
"grad_norm": 2.4886112213134766,
"learning_rate": 8.983999399217536e-06,
"loss": 0.8519719839096069,
"step": 488
},
{
"epoch": 1.0359408033826638,
"grad_norm": 2.133185863494873,
"learning_rate": 8.974019735363952e-06,
"loss": 0.8012784719467163,
"step": 490
},
{
"epoch": 1.040169133192389,
"grad_norm": 1.9001578092575073,
"learning_rate": 8.963997600223387e-06,
"loss": 0.7149202227592468,
"step": 492
},
{
"epoch": 1.0443974630021142,
"grad_norm": 1.2588262557983398,
"learning_rate": 8.953933116320474e-06,
"loss": 1.1228018999099731,
"step": 494
},
{
"epoch": 1.0486257928118394,
"grad_norm": 1.518155813217163,
"learning_rate": 8.943826406697569e-06,
"loss": 0.990390956401825,
"step": 496
},
{
"epoch": 1.0528541226215644,
"grad_norm": 1.0298517942428589,
"learning_rate": 8.93367759491326e-06,
"loss": 0.8367645144462585,
"step": 498
},
{
"epoch": 1.0570824524312896,
"grad_norm": 1.7629401683807373,
"learning_rate": 8.923486805040845e-06,
"loss": 0.8427010178565979,
"step": 500
},
{
"epoch": 1.0613107822410148,
"grad_norm": 1.0507843494415283,
"learning_rate": 8.913254161666828e-06,
"loss": 1.092274785041809,
"step": 502
},
{
"epoch": 1.06553911205074,
"grad_norm": 1.6207677125930786,
"learning_rate": 8.902979789889382e-06,
"loss": 1.0063058137893677,
"step": 504
},
{
"epoch": 1.069767441860465,
"grad_norm": 5.7719197273254395,
"learning_rate": 8.892663815316833e-06,
"loss": 0.8222056031227112,
"step": 506
},
{
"epoch": 1.0739957716701902,
"grad_norm": 0.9822844862937927,
"learning_rate": 8.882306364066114e-06,
"loss": 0.826958417892456,
"step": 508
},
{
"epoch": 1.0782241014799154,
"grad_norm": 5.889886856079102,
"learning_rate": 8.871907562761226e-06,
"loss": 0.9992004632949829,
"step": 510
},
{
"epoch": 1.0824524312896406,
"grad_norm": 1.3127708435058594,
"learning_rate": 8.861467538531692e-06,
"loss": 1.0234159231185913,
"step": 512
},
{
"epoch": 1.0866807610993658,
"grad_norm": 0.7838226556777954,
"learning_rate": 8.850986419011006e-06,
"loss": 1.0984134674072266,
"step": 514
},
{
"epoch": 1.0909090909090908,
"grad_norm": 2.3629720211029053,
"learning_rate": 8.840464332335062e-06,
"loss": 0.8936536312103271,
"step": 516
},
{
"epoch": 1.095137420718816,
"grad_norm": 1.2885105609893799,
"learning_rate": 8.829901407140599e-06,
"loss": 0.8397279977798462,
"step": 518
},
{
"epoch": 1.0993657505285412,
"grad_norm": 1.2271095514297485,
"learning_rate": 8.81929777256362e-06,
"loss": 0.6127222776412964,
"step": 520
},
{
"epoch": 1.1035940803382664,
"grad_norm": 1.2394264936447144,
"learning_rate": 8.808653558237818e-06,
"loss": 1.3062413930892944,
"step": 522
},
{
"epoch": 1.1078224101479917,
"grad_norm": 3.0176913738250732,
"learning_rate": 8.797968894292992e-06,
"loss": 0.7522382140159607,
"step": 524
},
{
"epoch": 1.1120507399577166,
"grad_norm": 1.3185428380966187,
"learning_rate": 8.787243911353447e-06,
"loss": 1.3689192533493042,
"step": 526
},
{
"epoch": 1.1162790697674418,
"grad_norm": 0.8265845775604248,
"learning_rate": 8.776478740536409e-06,
"loss": 0.8091297149658203,
"step": 528
},
{
"epoch": 1.120507399577167,
"grad_norm": 3.068026065826416,
"learning_rate": 8.76567351345042e-06,
"loss": 0.48788371682167053,
"step": 530
},
{
"epoch": 1.1247357293868923,
"grad_norm": 1.3478929996490479,
"learning_rate": 8.754828362193719e-06,
"loss": 0.9735050201416016,
"step": 532
},
{
"epoch": 1.1289640591966172,
"grad_norm": 1.5769144296646118,
"learning_rate": 8.743943419352638e-06,
"loss": 0.7880621552467346,
"step": 534
},
{
"epoch": 1.1331923890063424,
"grad_norm": 0.6606605052947998,
"learning_rate": 8.733018817999977e-06,
"loss": 0.9622098803520203,
"step": 536
},
{
"epoch": 1.1374207188160677,
"grad_norm": 9.970499038696289,
"learning_rate": 8.722054691693374e-06,
"loss": 0.7383369207382202,
"step": 538
},
{
"epoch": 1.1416490486257929,
"grad_norm": 1.727600336074829,
"learning_rate": 8.711051174473682e-06,
"loss": 0.7146698236465454,
"step": 540
},
{
"epoch": 1.145877378435518,
"grad_norm": 0.7654526233673096,
"learning_rate": 8.700008400863316e-06,
"loss": 1.0926287174224854,
"step": 542
},
{
"epoch": 1.150105708245243,
"grad_norm": 24.464750289916992,
"learning_rate": 8.688926505864624e-06,
"loss": 0.7354991436004639,
"step": 544
},
{
"epoch": 1.1543340380549683,
"grad_norm": 0.7699249386787415,
"learning_rate": 8.67780562495822e-06,
"loss": 1.1801013946533203,
"step": 546
},
{
"epoch": 1.1585623678646935,
"grad_norm": 3.897416114807129,
"learning_rate": 8.666645894101346e-06,
"loss": 1.2188596725463867,
"step": 548
},
{
"epoch": 1.1627906976744187,
"grad_norm": 1.4776039123535156,
"learning_rate": 8.655447449726189e-06,
"loss": 0.7939070463180542,
"step": 550
},
{
"epoch": 1.1670190274841437,
"grad_norm": 0.8722915053367615,
"learning_rate": 8.644210428738237e-06,
"loss": 0.9924266934394836,
"step": 552
},
{
"epoch": 1.1712473572938689,
"grad_norm": 0.9086503982543945,
"learning_rate": 8.632934968514582e-06,
"loss": 1.1229876279830933,
"step": 554
},
{
"epoch": 1.175475687103594,
"grad_norm": 2.393411874771118,
"learning_rate": 8.621621206902256e-06,
"loss": 1.1451060771942139,
"step": 556
},
{
"epoch": 1.1797040169133193,
"grad_norm": 1.1969093084335327,
"learning_rate": 8.61026928221654e-06,
"loss": 1.0445747375488281,
"step": 558
},
{
"epoch": 1.1839323467230445,
"grad_norm": 0.7867012023925781,
"learning_rate": 8.598879333239274e-06,
"loss": 0.9547511339187622,
"step": 560
},
{
"epoch": 1.1881606765327695,
"grad_norm": 0.9761852025985718,
"learning_rate": 8.587451499217161e-06,
"loss": 0.654085099697113,
"step": 562
},
{
"epoch": 1.1923890063424947,
"grad_norm": 1.1462132930755615,
"learning_rate": 8.575985919860065e-06,
"loss": 1.0323113203048706,
"step": 564
},
{
"epoch": 1.1966173361522199,
"grad_norm": 6.983845233917236,
"learning_rate": 8.564482735339296e-06,
"loss": 0.9465799331665039,
"step": 566
},
{
"epoch": 1.200845665961945,
"grad_norm": 2.1542341709136963,
"learning_rate": 8.552942086285909e-06,
"loss": 1.1723790168762207,
"step": 568
},
{
"epoch": 1.20507399577167,
"grad_norm": 4.250433444976807,
"learning_rate": 8.541364113788974e-06,
"loss": 1.1649366617202759,
"step": 570
},
{
"epoch": 1.2093023255813953,
"grad_norm": 1.6407173871994019,
"learning_rate": 8.529748959393855e-06,
"loss": 1.0406873226165771,
"step": 572
},
{
"epoch": 1.2135306553911205,
"grad_norm": 1.1917067766189575,
"learning_rate": 8.518096765100485e-06,
"loss": 0.7828882336616516,
"step": 574
},
{
"epoch": 1.2177589852008457,
"grad_norm": 2.0870354175567627,
"learning_rate": 8.50640767336162e-06,
"loss": 0.7890576124191284,
"step": 576
},
{
"epoch": 1.221987315010571,
"grad_norm": 1.4371873140335083,
"learning_rate": 8.494681827081096e-06,
"loss": 0.8771016597747803,
"step": 578
},
{
"epoch": 1.226215644820296,
"grad_norm": 1.9948108196258545,
"learning_rate": 8.482919369612103e-06,
"loss": 0.7651607990264893,
"step": 580
},
{
"epoch": 1.230443974630021,
"grad_norm": 2.2344977855682373,
"learning_rate": 8.471120444755405e-06,
"loss": 1.0366690158843994,
"step": 582
},
{
"epoch": 1.2346723044397463,
"grad_norm": 5.77453088760376,
"learning_rate": 8.4592851967576e-06,
"loss": 1.0104398727416992,
"step": 584
},
{
"epoch": 1.2389006342494715,
"grad_norm": 11.20981502532959,
"learning_rate": 8.447413770309349e-06,
"loss": 0.8929464817047119,
"step": 586
},
{
"epoch": 1.2431289640591967,
"grad_norm": 1.3959273099899292,
"learning_rate": 8.435506310543607e-06,
"loss": 0.9282096028327942,
"step": 588
},
{
"epoch": 1.2473572938689217,
"grad_norm": 1.612087607383728,
"learning_rate": 8.42356296303386e-06,
"loss": 0.6361445188522339,
"step": 590
},
{
"epoch": 1.251585623678647,
"grad_norm": 0.5314251184463501,
"learning_rate": 8.411583873792325e-06,
"loss": 0.8630724549293518,
"step": 592
},
{
"epoch": 1.255813953488372,
"grad_norm": 18.40007209777832,
"learning_rate": 8.399569189268182e-06,
"loss": 1.0672541856765747,
"step": 594
},
{
"epoch": 1.2600422832980973,
"grad_norm": 1.36466646194458,
"learning_rate": 8.387519056345778e-06,
"loss": 0.67390376329422,
"step": 596
},
{
"epoch": 1.2642706131078225,
"grad_norm": 1.1603999137878418,
"learning_rate": 8.375433622342824e-06,
"loss": 1.085272192955017,
"step": 598
},
{
"epoch": 1.2684989429175475,
"grad_norm": 2.584357738494873,
"learning_rate": 8.363313035008613e-06,
"loss": 1.2544325590133667,
"step": 600
},
{
"epoch": 1.2727272727272727,
"grad_norm": 1.2049429416656494,
"learning_rate": 8.351157442522187e-06,
"loss": 0.9232980608940125,
"step": 602
},
{
"epoch": 1.276955602536998,
"grad_norm": 1.2597911357879639,
"learning_rate": 8.338966993490554e-06,
"loss": 1.1165087223052979,
"step": 604
},
{
"epoch": 1.2811839323467231,
"grad_norm": 2.2192118167877197,
"learning_rate": 8.326741836946852e-06,
"loss": 0.721271276473999,
"step": 606
},
{
"epoch": 1.285412262156448,
"grad_norm": 0.9270818829536438,
"learning_rate": 8.314482122348523e-06,
"loss": 1.056286334991455,
"step": 608
},
{
"epoch": 1.2896405919661733,
"grad_norm": 0.9210339188575745,
"learning_rate": 8.30218799957551e-06,
"loss": 0.7497894167900085,
"step": 610
},
{
"epoch": 1.2938689217758985,
"grad_norm": 2.7070858478546143,
"learning_rate": 8.289859618928404e-06,
"loss": 0.6183304786682129,
"step": 612
},
{
"epoch": 1.2980972515856237,
"grad_norm": 1.041532039642334,
"learning_rate": 8.277497131126608e-06,
"loss": 0.895203173160553,
"step": 614
},
{
"epoch": 1.302325581395349,
"grad_norm": 3.0417888164520264,
"learning_rate": 8.265100687306503e-06,
"loss": 0.4850989282131195,
"step": 616
},
{
"epoch": 1.306553911205074,
"grad_norm": 5.1612677574157715,
"learning_rate": 8.252670439019595e-06,
"loss": 0.9282147884368896,
"step": 618
},
{
"epoch": 1.3107822410147991,
"grad_norm": 1.1880724430084229,
"learning_rate": 8.240206538230663e-06,
"loss": 1.089096188545227,
"step": 620
},
{
"epoch": 1.3150105708245243,
"grad_norm": 1.9302226305007935,
"learning_rate": 8.227709137315905e-06,
"loss": 1.0690324306488037,
"step": 622
},
{
"epoch": 1.3192389006342495,
"grad_norm": 8.238303184509277,
"learning_rate": 8.215178389061064e-06,
"loss": 0.6763749718666077,
"step": 624
},
{
"epoch": 1.3234672304439745,
"grad_norm": 1.1396892070770264,
"learning_rate": 8.202614446659572e-06,
"loss": 1.0496219396591187,
"step": 626
},
{
"epoch": 1.3276955602536997,
"grad_norm": 1.584627389907837,
"learning_rate": 8.190017463710675e-06,
"loss": 1.1644724607467651,
"step": 628
},
{
"epoch": 1.331923890063425,
"grad_norm": 1.3790614604949951,
"learning_rate": 8.177387594217548e-06,
"loss": 0.9552294611930847,
"step": 630
},
{
"epoch": 1.3361522198731501,
"grad_norm": 2.27864146232605,
"learning_rate": 8.164724992585421e-06,
"loss": 0.5984556674957275,
"step": 632
},
{
"epoch": 1.3403805496828753,
"grad_norm": 1.3319947719573975,
"learning_rate": 8.152029813619685e-06,
"loss": 0.6957176923751831,
"step": 634
},
{
"epoch": 1.3446088794926006,
"grad_norm": 1.01459538936615,
"learning_rate": 8.139302212524006e-06,
"loss": 0.8053246140480042,
"step": 636
},
{
"epoch": 1.3488372093023255,
"grad_norm": 39.183414459228516,
"learning_rate": 8.126542344898419e-06,
"loss": 0.9739262461662292,
"step": 638
},
{
"epoch": 1.3530655391120507,
"grad_norm": 0.4954521358013153,
"learning_rate": 8.113750366737431e-06,
"loss": 0.6748231053352356,
"step": 640
},
{
"epoch": 1.357293868921776,
"grad_norm": 0.7353281378746033,
"learning_rate": 8.100926434428116e-06,
"loss": 1.1696051359176636,
"step": 642
},
{
"epoch": 1.361522198731501,
"grad_norm": 1.1235963106155396,
"learning_rate": 8.0880707047482e-06,
"loss": 0.9992832541465759,
"step": 644
},
{
"epoch": 1.3657505285412261,
"grad_norm": 10.986451148986816,
"learning_rate": 8.075183334864139e-06,
"loss": 0.7327241897583008,
"step": 646
},
{
"epoch": 1.3699788583509513,
"grad_norm": 11.65727710723877,
"learning_rate": 8.062264482329209e-06,
"loss": 0.6986007690429688,
"step": 648
},
{
"epoch": 1.3742071881606766,
"grad_norm": 6.460354328155518,
"learning_rate": 8.049314305081575e-06,
"loss": 0.7701452374458313,
"step": 650
},
{
"epoch": 1.3784355179704018,
"grad_norm": 1.1553010940551758,
"learning_rate": 8.036332961442355e-06,
"loss": 0.844822347164154,
"step": 652
},
{
"epoch": 1.382663847780127,
"grad_norm": 1.0680776834487915,
"learning_rate": 8.023320610113691e-06,
"loss": 1.1353340148925781,
"step": 654
},
{
"epoch": 1.386892177589852,
"grad_norm": 0.8363878130912781,
"learning_rate": 8.010277410176801e-06,
"loss": 0.9862926602363586,
"step": 656
},
{
"epoch": 1.3911205073995772,
"grad_norm": 3.794541120529175,
"learning_rate": 7.997203521090053e-06,
"loss": 0.8430204391479492,
"step": 658
},
{
"epoch": 1.3953488372093024,
"grad_norm": 1.0263630151748657,
"learning_rate": 7.984099102686986e-06,
"loss": 0.9049514532089233,
"step": 660
},
{
"epoch": 1.3995771670190273,
"grad_norm": 1.4661459922790527,
"learning_rate": 7.970964315174384e-06,
"loss": 1.117480754852295,
"step": 662
},
{
"epoch": 1.4038054968287526,
"grad_norm": 0.2747534513473511,
"learning_rate": 7.9577993191303e-06,
"loss": 0.8609821796417236,
"step": 664
},
{
"epoch": 1.4080338266384778,
"grad_norm": 1.7366811037063599,
"learning_rate": 7.944604275502102e-06,
"loss": 1.3553338050842285,
"step": 666
},
{
"epoch": 1.412262156448203,
"grad_norm": 0.8697457313537598,
"learning_rate": 7.931379345604495e-06,
"loss": 1.081929326057434,
"step": 668
},
{
"epoch": 1.4164904862579282,
"grad_norm": 3.104210138320923,
"learning_rate": 7.918124691117563e-06,
"loss": 0.7276521325111389,
"step": 670
},
{
"epoch": 1.4207188160676534,
"grad_norm": 0.9394916296005249,
"learning_rate": 7.904840474084783e-06,
"loss": 1.0517014265060425,
"step": 672
},
{
"epoch": 1.4249471458773784,
"grad_norm": 0.6771994829177856,
"learning_rate": 7.891526856911046e-06,
"loss": 0.5833929181098938,
"step": 674
},
{
"epoch": 1.4291754756871036,
"grad_norm": 2.0351264476776123,
"learning_rate": 7.878184002360664e-06,
"loss": 0.7924460172653198,
"step": 676
},
{
"epoch": 1.4334038054968288,
"grad_norm": 0.8329294323921204,
"learning_rate": 7.864812073555402e-06,
"loss": 1.2809078693389893,
"step": 678
},
{
"epoch": 1.437632135306554,
"grad_norm": 1.223764181137085,
"learning_rate": 7.851411233972461e-06,
"loss": 1.1325117349624634,
"step": 680
},
{
"epoch": 1.441860465116279,
"grad_norm": 3.7025115489959717,
"learning_rate": 7.837981647442486e-06,
"loss": 1.109671950340271,
"step": 682
},
{
"epoch": 1.4460887949260042,
"grad_norm": 0.7311835885047913,
"learning_rate": 7.82452347814757e-06,
"loss": 1.0558146238327026,
"step": 684
},
{
"epoch": 1.4503171247357294,
"grad_norm": 1.8109835386276245,
"learning_rate": 7.811036890619239e-06,
"loss": 1.3652693033218384,
"step": 686
},
{
"epoch": 1.4545454545454546,
"grad_norm": 1.0350303649902344,
"learning_rate": 7.797522049736441e-06,
"loss": 1.083454966545105,
"step": 688
},
{
"epoch": 1.4587737843551798,
"grad_norm": 2.478970527648926,
"learning_rate": 7.78397912072354e-06,
"loss": 0.43894389271736145,
"step": 690
},
{
"epoch": 1.463002114164905,
"grad_norm": 0.7424451112747192,
"learning_rate": 7.770408269148281e-06,
"loss": 1.0906665325164795,
"step": 692
},
{
"epoch": 1.46723044397463,
"grad_norm": 2.7369964122772217,
"learning_rate": 7.756809660919778e-06,
"loss": 0.5293806791305542,
"step": 694
},
{
"epoch": 1.4714587737843552,
"grad_norm": 1.301570177078247,
"learning_rate": 7.74318346228648e-06,
"loss": 1.0440341234207153,
"step": 696
},
{
"epoch": 1.4756871035940804,
"grad_norm": 1.2320626974105835,
"learning_rate": 7.72952983983414e-06,
"loss": 1.0965993404388428,
"step": 698
},
{
"epoch": 1.4799154334038054,
"grad_norm": 3.4229559898376465,
"learning_rate": 7.715848960483776e-06,
"loss": 0.9500837326049805,
"step": 700
},
{
"epoch": 1.4841437632135306,
"grad_norm": 1.0504419803619385,
"learning_rate": 7.702140991489637e-06,
"loss": 1.076796054840088,
"step": 702
},
{
"epoch": 1.4883720930232558,
"grad_norm": 0.8435136079788208,
"learning_rate": 7.688406100437149e-06,
"loss": 0.9962727427482605,
"step": 704
},
{
"epoch": 1.492600422832981,
"grad_norm": 0.9324005842208862,
"learning_rate": 7.674644455240878e-06,
"loss": 0.7798671126365662,
"step": 706
},
{
"epoch": 1.4968287526427062,
"grad_norm": 1.25261652469635,
"learning_rate": 7.660856224142456e-06,
"loss": 0.2566996216773987,
"step": 708
},
{
"epoch": 1.5010570824524314,
"grad_norm": 4.27152156829834,
"learning_rate": 7.647041575708555e-06,
"loss": 0.6202626824378967,
"step": 710
},
{
"epoch": 1.5052854122621564,
"grad_norm": 1.4225716590881348,
"learning_rate": 7.633200678828804e-06,
"loss": 0.5293870568275452,
"step": 712
},
{
"epoch": 1.5095137420718816,
"grad_norm": 2.0559539794921875,
"learning_rate": 7.6193337027137215e-06,
"loss": 0.9094487428665161,
"step": 714
},
{
"epoch": 1.5137420718816068,
"grad_norm": 0.8616843819618225,
"learning_rate": 7.605440816892667e-06,
"loss": 1.095259666442871,
"step": 716
},
{
"epoch": 1.5179704016913318,
"grad_norm": 2.8545007705688477,
"learning_rate": 7.591522191211746e-06,
"loss": 0.9499167203903198,
"step": 718
},
{
"epoch": 1.522198731501057,
"grad_norm": 0.3324730694293976,
"learning_rate": 7.577577995831755e-06,
"loss": 0.8667873740196228,
"step": 720
},
{
"epoch": 1.5264270613107822,
"grad_norm": 3.2699382305145264,
"learning_rate": 7.563608401226081e-06,
"loss": 0.8496268391609192,
"step": 722
},
{
"epoch": 1.5306553911205074,
"grad_norm": 2.7426388263702393,
"learning_rate": 7.549613578178632e-06,
"loss": 0.8953691720962524,
"step": 724
},
{
"epoch": 1.5348837209302326,
"grad_norm": 1.2888638973236084,
"learning_rate": 7.535593697781741e-06,
"loss": 0.9176872968673706,
"step": 726
},
{
"epoch": 1.5391120507399578,
"grad_norm": 1.22314453125,
"learning_rate": 7.52154893143408e-06,
"loss": 0.6477530002593994,
"step": 728
},
{
"epoch": 1.543340380549683,
"grad_norm": 2.0176548957824707,
"learning_rate": 7.50747945083856e-06,
"loss": 0.810649573802948,
"step": 730
},
{
"epoch": 1.547568710359408,
"grad_norm": 1.8671491146087646,
"learning_rate": 7.493385428000234e-06,
"loss": 1.0150587558746338,
"step": 732
},
{
"epoch": 1.5517970401691332,
"grad_norm": 0.4674511253833771,
"learning_rate": 7.479267035224191e-06,
"loss": 0.2408760040998459,
"step": 734
},
{
"epoch": 1.5560253699788582,
"grad_norm": 1.1424227952957153,
"learning_rate": 7.465124445113459e-06,
"loss": 0.9625402092933655,
"step": 736
},
{
"epoch": 1.5602536997885834,
"grad_norm": 0.3999347984790802,
"learning_rate": 7.450957830566878e-06,
"loss": 0.7018699645996094,
"step": 738
},
{
"epoch": 1.5644820295983086,
"grad_norm": 0.6436890363693237,
"learning_rate": 7.436767364777009e-06,
"loss": 1.0802757740020752,
"step": 740
},
{
"epoch": 1.5687103594080338,
"grad_norm": 1.5571529865264893,
"learning_rate": 7.422553221227993e-06,
"loss": 0.709961473941803,
"step": 742
},
{
"epoch": 1.572938689217759,
"grad_norm": 2.840071201324463,
"learning_rate": 7.4083155736934484e-06,
"loss": 1.1563467979431152,
"step": 744
},
{
"epoch": 1.5771670190274842,
"grad_norm": 2.561483144760132,
"learning_rate": 7.394054596234336e-06,
"loss": 1.1827778816223145,
"step": 746
},
{
"epoch": 1.5813953488372094,
"grad_norm": 5.667703151702881,
"learning_rate": 7.379770463196831e-06,
"loss": 0.9164658188819885,
"step": 748
},
{
"epoch": 1.5856236786469344,
"grad_norm": 0.1942562460899353,
"learning_rate": 7.365463349210203e-06,
"loss": 0.0511607900261879,
"step": 750
},
{
"epoch": 1.5898520084566596,
"grad_norm": 2.0378458499908447,
"learning_rate": 7.351133429184666e-06,
"loss": 0.8130763173103333,
"step": 752
},
{
"epoch": 1.5940803382663846,
"grad_norm": 1.423600673675537,
"learning_rate": 7.336780878309251e-06,
"loss": 1.1934304237365723,
"step": 754
},
{
"epoch": 1.5983086680761098,
"grad_norm": 0.8461257815361023,
"learning_rate": 7.322405872049657e-06,
"loss": 1.077623963356018,
"step": 756
},
{
"epoch": 1.602536997885835,
"grad_norm": 0.8712191581726074,
"learning_rate": 7.308008586146111e-06,
"loss": 1.112871766090393,
"step": 758
},
{
"epoch": 1.6067653276955602,
"grad_norm": 1.5069035291671753,
"learning_rate": 7.293589196611221e-06,
"loss": 1.0710359811782837,
"step": 760
},
{
"epoch": 1.6109936575052854,
"grad_norm": 1.838446021080017,
"learning_rate": 7.279147879727811e-06,
"loss": 1.093487024307251,
"step": 762
},
{
"epoch": 1.6152219873150107,
"grad_norm": 1.709586501121521,
"learning_rate": 7.264684812046787e-06,
"loss": 1.0070074796676636,
"step": 764
},
{
"epoch": 1.6194503171247359,
"grad_norm": 2.432494640350342,
"learning_rate": 7.250200170384961e-06,
"loss": 0.8508090376853943,
"step": 766
},
{
"epoch": 1.6236786469344608,
"grad_norm": 1.0690550804138184,
"learning_rate": 7.235694131822894e-06,
"loss": 1.067232370376587,
"step": 768
},
{
"epoch": 1.627906976744186,
"grad_norm": 1.4697656631469727,
"learning_rate": 7.221166873702738e-06,
"loss": 1.0529597997665405,
"step": 770
},
{
"epoch": 1.6321353065539113,
"grad_norm": 1.0028948783874512,
"learning_rate": 7.206618573626058e-06,
"loss": 1.0673527717590332,
"step": 772
},
{
"epoch": 1.6363636363636362,
"grad_norm": 0.8941748738288879,
"learning_rate": 7.19204940945167e-06,
"loss": 0.7840203046798706,
"step": 774
},
{
"epoch": 1.6405919661733614,
"grad_norm": 1.9076416492462158,
"learning_rate": 7.177459559293456e-06,
"loss": 1.0982592105865479,
"step": 776
},
{
"epoch": 1.6448202959830867,
"grad_norm": 1.9868332147598267,
"learning_rate": 7.162849201518198e-06,
"loss": 0.659265398979187,
"step": 778
},
{
"epoch": 1.6490486257928119,
"grad_norm": 1.0087363719940186,
"learning_rate": 7.1482185147433905e-06,
"loss": 1.0045416355133057,
"step": 780
},
{
"epoch": 1.653276955602537,
"grad_norm": 1.943669080734253,
"learning_rate": 7.133567677835055e-06,
"loss": 0.7921465039253235,
"step": 782
},
{
"epoch": 1.6575052854122623,
"grad_norm": 1.210842490196228,
"learning_rate": 7.11889686990556e-06,
"loss": 1.1942638158798218,
"step": 784
},
{
"epoch": 1.6617336152219875,
"grad_norm": 1.6035552024841309,
"learning_rate": 7.104206270311423e-06,
"loss": 0.4886111915111542,
"step": 786
},
{
"epoch": 1.6659619450317125,
"grad_norm": 0.9401936531066895,
"learning_rate": 7.089496058651133e-06,
"loss": 1.078162670135498,
"step": 788
},
{
"epoch": 1.6701902748414377,
"grad_norm": 0.9160062670707703,
"learning_rate": 7.0747664147629335e-06,
"loss": 0.7416279315948486,
"step": 790
},
{
"epoch": 1.6744186046511627,
"grad_norm": 0.7420851588249207,
"learning_rate": 7.060017518722637e-06,
"loss": 0.4208860993385315,
"step": 792
},
{
"epoch": 1.6786469344608879,
"grad_norm": 1.0709822177886963,
"learning_rate": 7.0452495508414276e-06,
"loss": 1.0613205432891846,
"step": 794
},
{
"epoch": 1.682875264270613,
"grad_norm": 0.4770656228065491,
"learning_rate": 7.030462691663641e-06,
"loss": 0.594983696937561,
"step": 796
},
{
"epoch": 1.6871035940803383,
"grad_norm": 1.4194773435592651,
"learning_rate": 7.015657121964575e-06,
"loss": 1.1746389865875244,
"step": 798
},
{
"epoch": 1.6913319238900635,
"grad_norm": 1.1806695461273193,
"learning_rate": 7.000833022748267e-06,
"loss": 1.056938648223877,
"step": 800
},
{
"epoch": 1.6955602536997887,
"grad_norm": 7.539623737335205,
"learning_rate": 6.985990575245283e-06,
"loss": 0.5604655146598816,
"step": 802
},
{
"epoch": 1.699788583509514,
"grad_norm": 0.9911687970161438,
"learning_rate": 6.97112996091051e-06,
"loss": 1.041745901107788,
"step": 804
},
{
"epoch": 1.7040169133192389,
"grad_norm": 2.0413644313812256,
"learning_rate": 6.956251361420926e-06,
"loss": 0.7910910248756409,
"step": 806
},
{
"epoch": 1.708245243128964,
"grad_norm": 26.18593978881836,
"learning_rate": 6.94135495867339e-06,
"loss": 0.748640775680542,
"step": 808
},
{
"epoch": 1.712473572938689,
"grad_norm": 0.9232349991798401,
"learning_rate": 6.926440934782407e-06,
"loss": 1.0842905044555664,
"step": 810
},
{
"epoch": 1.7167019027484143,
"grad_norm": 2.251222610473633,
"learning_rate": 6.9115094720779155e-06,
"loss": 1.3004353046417236,
"step": 812
},
{
"epoch": 1.7209302325581395,
"grad_norm": 1.0462383031845093,
"learning_rate": 6.896560753103041e-06,
"loss": 1.0440006256103516,
"step": 814
},
{
"epoch": 1.7251585623678647,
"grad_norm": 0.8429021239280701,
"learning_rate": 6.881594960611881e-06,
"loss": 1.1201937198638916,
"step": 816
},
{
"epoch": 1.72938689217759,
"grad_norm": 0.9414411187171936,
"learning_rate": 6.866612277567261e-06,
"loss": 0.8856564164161682,
"step": 818
},
{
"epoch": 1.733615221987315,
"grad_norm": 2.5502281188964844,
"learning_rate": 6.851612887138501e-06,
"loss": 0.9899932742118835,
"step": 820
},
{
"epoch": 1.7378435517970403,
"grad_norm": 0.7951850295066833,
"learning_rate": 6.836596972699175e-06,
"loss": 0.685820996761322,
"step": 822
},
{
"epoch": 1.7420718816067653,
"grad_norm": 2.9686474800109863,
"learning_rate": 6.82156471782487e-06,
"loss": 0.792769730091095,
"step": 824
},
{
"epoch": 1.7463002114164905,
"grad_norm": 3.299290657043457,
"learning_rate": 6.806516306290942e-06,
"loss": 0.9077906012535095,
"step": 826
},
{
"epoch": 1.7505285412262155,
"grad_norm": 0.7948803305625916,
"learning_rate": 6.791451922070265e-06,
"loss": 0.9383656978607178,
"step": 828
},
{
"epoch": 1.7547568710359407,
"grad_norm": 1.8922616243362427,
"learning_rate": 6.77637174933099e-06,
"loss": 0.7109253406524658,
"step": 830
},
{
"epoch": 1.758985200845666,
"grad_norm": 1.5757509469985962,
"learning_rate": 6.7612759724342895e-06,
"loss": 1.121849536895752,
"step": 832
},
{
"epoch": 1.763213530655391,
"grad_norm": 3.7704286575317383,
"learning_rate": 6.746164775932096e-06,
"loss": 0.7459769248962402,
"step": 834
},
{
"epoch": 1.7674418604651163,
"grad_norm": 4.336609363555908,
"learning_rate": 6.731038344564862e-06,
"loss": 0.7917211651802063,
"step": 836
},
{
"epoch": 1.7716701902748415,
"grad_norm": 1.2018744945526123,
"learning_rate": 6.715896863259286e-06,
"loss": 1.0897058248519897,
"step": 838
},
{
"epoch": 1.7758985200845667,
"grad_norm": 1.6032429933547974,
"learning_rate": 6.700740517126061e-06,
"loss": 0.7954162955284119,
"step": 840
},
{
"epoch": 1.7801268498942917,
"grad_norm": 3.5267279148101807,
"learning_rate": 6.685569491457608e-06,
"loss": 0.4790779650211334,
"step": 842
},
{
"epoch": 1.784355179704017,
"grad_norm": 2.2513656616210938,
"learning_rate": 6.670383971725809e-06,
"loss": 0.7167988419532776,
"step": 844
},
{
"epoch": 1.7885835095137421,
"grad_norm": 2.1263468265533447,
"learning_rate": 6.655184143579746e-06,
"loss": 0.9691937565803528,
"step": 846
},
{
"epoch": 1.792811839323467,
"grad_norm": 1.0709644556045532,
"learning_rate": 6.639970192843423e-06,
"loss": 0.7088806629180908,
"step": 848
},
{
"epoch": 1.7970401691331923,
"grad_norm": 1.2077969312667847,
"learning_rate": 6.624742305513502e-06,
"loss": 0.7262135148048401,
"step": 850
},
{
"epoch": 1.8012684989429175,
"grad_norm": 1.6867655515670776,
"learning_rate": 6.609500667757022e-06,
"loss": 0.9637296795845032,
"step": 852
},
{
"epoch": 1.8054968287526427,
"grad_norm": 0.7157866358757019,
"learning_rate": 6.594245465909127e-06,
"loss": 0.7437013387680054,
"step": 854
},
{
"epoch": 1.809725158562368,
"grad_norm": 0.7874408960342407,
"learning_rate": 6.578976886470791e-06,
"loss": 1.0531718730926514,
"step": 856
},
{
"epoch": 1.8139534883720931,
"grad_norm": 6.787167549133301,
"learning_rate": 6.56369511610653e-06,
"loss": 0.6696254014968872,
"step": 858
},
{
"epoch": 1.8181818181818183,
"grad_norm": 0.8861216902732849,
"learning_rate": 6.548400341642128e-06,
"loss": 0.9765800833702087,
"step": 860
},
{
"epoch": 1.8224101479915433,
"grad_norm": 0.6991898417472839,
"learning_rate": 6.533092750062345e-06,
"loss": 1.0134353637695312,
"step": 862
},
{
"epoch": 1.8266384778012685,
"grad_norm": 3.4546828269958496,
"learning_rate": 6.5177725285086416e-06,
"loss": 0.5766164660453796,
"step": 864
},
{
"epoch": 1.8308668076109935,
"grad_norm": 0.8045233488082886,
"learning_rate": 6.5024398642768785e-06,
"loss": 0.7740400433540344,
"step": 866
},
{
"epoch": 1.8350951374207187,
"grad_norm": 0.8292525410652161,
"learning_rate": 6.487094944815036e-06,
"loss": 1.0496159791946411,
"step": 868
},
{
"epoch": 1.839323467230444,
"grad_norm": 1.30192232131958,
"learning_rate": 6.47173795772092e-06,
"loss": 1.0689643621444702,
"step": 870
},
{
"epoch": 1.8435517970401691,
"grad_norm": 0.4630022346973419,
"learning_rate": 6.456369090739869e-06,
"loss": 0.7175701856613159,
"step": 872
},
{
"epoch": 1.8477801268498943,
"grad_norm": 1.4346680641174316,
"learning_rate": 6.440988531762454e-06,
"loss": 1.0404019355773926,
"step": 874
},
{
"epoch": 1.8520084566596196,
"grad_norm": 2.4394078254699707,
"learning_rate": 6.42559646882219e-06,
"loss": 0.9022433161735535,
"step": 876
},
{
"epoch": 1.8562367864693448,
"grad_norm": 3.233065128326416,
"learning_rate": 6.410193090093227e-06,
"loss": 0.34865736961364746,
"step": 878
},
{
"epoch": 1.8604651162790697,
"grad_norm": 6.990861415863037,
"learning_rate": 6.394778583888064e-06,
"loss": 0.7052128911018372,
"step": 880
},
{
"epoch": 1.864693446088795,
"grad_norm": 1.999719262123108,
"learning_rate": 6.379353138655227e-06,
"loss": 1.0530213117599487,
"step": 882
},
{
"epoch": 1.86892177589852,
"grad_norm": 1.1507999897003174,
"learning_rate": 6.3639169429769805e-06,
"loss": 0.7579324245452881,
"step": 884
},
{
"epoch": 1.8731501057082451,
"grad_norm": 1.1435511112213135,
"learning_rate": 6.348470185567025e-06,
"loss": 1.048447608947754,
"step": 886
},
{
"epoch": 1.8773784355179703,
"grad_norm": 0.7901480197906494,
"learning_rate": 6.333013055268169e-06,
"loss": 1.1018636226654053,
"step": 888
},
{
"epoch": 1.8816067653276956,
"grad_norm": 1.6382914781570435,
"learning_rate": 6.317545741050044e-06,
"loss": 0.5727537870407104,
"step": 890
},
{
"epoch": 1.8858350951374208,
"grad_norm": 2.474337577819824,
"learning_rate": 6.3020684320067805e-06,
"loss": 0.7920655012130737,
"step": 892
},
{
"epoch": 1.890063424947146,
"grad_norm": 6.634578704833984,
"learning_rate": 6.2865813173546976e-06,
"loss": 1.0641400814056396,
"step": 894
},
{
"epoch": 1.8942917547568712,
"grad_norm": 3.7671637535095215,
"learning_rate": 6.271084586429994e-06,
"loss": 1.4103152751922607,
"step": 896
},
{
"epoch": 1.8985200845665962,
"grad_norm": 1.6423215866088867,
"learning_rate": 6.255578428686433e-06,
"loss": 1.0158220529556274,
"step": 898
},
{
"epoch": 1.9027484143763214,
"grad_norm": 1.880095362663269,
"learning_rate": 6.240063033693023e-06,
"loss": 0.4871620833873749,
"step": 900
},
{
"epoch": 1.9069767441860463,
"grad_norm": 1.7650635242462158,
"learning_rate": 6.2245385911316994e-06,
"loss": 1.0664781332015991,
"step": 902
},
{
"epoch": 1.9112050739957716,
"grad_norm": 7.110355377197266,
"learning_rate": 6.2090052907950096e-06,
"loss": 0.5222395658493042,
"step": 904
},
{
"epoch": 1.9154334038054968,
"grad_norm": 1.1369906663894653,
"learning_rate": 6.193463322583792e-06,
"loss": 0.8623431921005249,
"step": 906
},
{
"epoch": 1.919661733615222,
"grad_norm": 1.1148680448532104,
"learning_rate": 6.177912876504851e-06,
"loss": 1.2079038619995117,
"step": 908
},
{
"epoch": 1.9238900634249472,
"grad_norm": 1.4633415937423706,
"learning_rate": 6.16235414266864e-06,
"loss": 0.7764965295791626,
"step": 910
},
{
"epoch": 1.9281183932346724,
"grad_norm": 1.5081416368484497,
"learning_rate": 6.146787311286927e-06,
"loss": 1.02378511428833,
"step": 912
},
{
"epoch": 1.9323467230443976,
"grad_norm": 4.188223838806152,
"learning_rate": 6.131212572670483e-06,
"loss": 0.6803139448165894,
"step": 914
},
{
"epoch": 1.9365750528541226,
"grad_norm": 0.9626418948173523,
"learning_rate": 6.115630117226741e-06,
"loss": 1.0788558721542358,
"step": 916
},
{
"epoch": 1.9408033826638478,
"grad_norm": 0.7504255771636963,
"learning_rate": 6.1000401354574855e-06,
"loss": 0.6543954014778137,
"step": 918
},
{
"epoch": 1.945031712473573,
"grad_norm": 1.2343472242355347,
"learning_rate": 6.084442817956502e-06,
"loss": 0.9022839665412903,
"step": 920
},
{
"epoch": 1.949260042283298,
"grad_norm": 2.5817534923553467,
"learning_rate": 6.0688383554072625e-06,
"loss": 0.9855140447616577,
"step": 922
},
{
"epoch": 1.9534883720930232,
"grad_norm": 0.7275571823120117,
"learning_rate": 6.053226938580596e-06,
"loss": 0.977034330368042,
"step": 924
},
{
"epoch": 1.9577167019027484,
"grad_norm": 0.38651394844055176,
"learning_rate": 6.03760875833234e-06,
"loss": 0.8864318132400513,
"step": 926
},
{
"epoch": 1.9619450317124736,
"grad_norm": 1.135390043258667,
"learning_rate": 6.021984005601027e-06,
"loss": 1.147244930267334,
"step": 928
},
{
"epoch": 1.9661733615221988,
"grad_norm": 1.0886331796646118,
"learning_rate": 6.006352871405534e-06,
"loss": 0.8382534384727478,
"step": 930
},
{
"epoch": 1.970401691331924,
"grad_norm": 1.2228710651397705,
"learning_rate": 5.990715546842755e-06,
"loss": 1.211573839187622,
"step": 932
},
{
"epoch": 1.9746300211416492,
"grad_norm": 1.6050304174423218,
"learning_rate": 5.9750722230852694e-06,
"loss": 0.8218317627906799,
"step": 934
},
{
"epoch": 1.9788583509513742,
"grad_norm": 0.9150649905204773,
"learning_rate": 5.959423091378992e-06,
"loss": 0.6709915399551392,
"step": 936
},
{
"epoch": 1.9830866807610994,
"grad_norm": 3.3599793910980225,
"learning_rate": 5.943768343040845e-06,
"loss": 0.8266628980636597,
"step": 938
},
{
"epoch": 1.9873150105708244,
"grad_norm": 2.21895432472229,
"learning_rate": 5.928108169456419e-06,
"loss": 1.063590407371521,
"step": 940
},
{
"epoch": 1.9915433403805496,
"grad_norm": 0.7361075282096863,
"learning_rate": 5.912442762077624e-06,
"loss": 1.058224081993103,
"step": 942
},
{
"epoch": 1.9957716701902748,
"grad_norm": 0.9331009387969971,
"learning_rate": 5.89677231242036e-06,
"loss": 1.0889805555343628,
"step": 944
},
{
"epoch": 2.0,
"grad_norm": 1.483725666999817,
"learning_rate": 5.8810970120621714e-06,
"loss": 0.46498048305511475,
"step": 946
},
{
"epoch": 2.004228329809725,
"grad_norm": 19.712791442871094,
"learning_rate": 5.8654170526399014e-06,
"loss": 0.7300288677215576,
"step": 948
},
{
"epoch": 2.0084566596194504,
"grad_norm": 0.7668775916099548,
"learning_rate": 5.849732625847354e-06,
"loss": 0.9317559003829956,
"step": 950
},
{
"epoch": 2.0126849894291756,
"grad_norm": 1.4207711219787598,
"learning_rate": 5.834043923432947e-06,
"loss": 0.874889612197876,
"step": 952
},
{
"epoch": 2.016913319238901,
"grad_norm": 1.6520004272460938,
"learning_rate": 5.818351137197373e-06,
"loss": 1.0628215074539185,
"step": 954
},
{
"epoch": 2.0211416490486256,
"grad_norm": 3.5480082035064697,
"learning_rate": 5.802654458991245e-06,
"loss": 1.189589500427246,
"step": 956
},
{
"epoch": 2.025369978858351,
"grad_norm": 0.9280943870544434,
"learning_rate": 5.786954080712763e-06,
"loss": 0.6518001556396484,
"step": 958
},
{
"epoch": 2.029598308668076,
"grad_norm": 1.1712863445281982,
"learning_rate": 5.7712501943053576e-06,
"loss": 0.8779881000518799,
"step": 960
},
{
"epoch": 2.033826638477801,
"grad_norm": 1.3695743083953857,
"learning_rate": 5.755542991755352e-06,
"loss": 0.8559538722038269,
"step": 962
},
{
"epoch": 2.0380549682875264,
"grad_norm": 6.934473514556885,
"learning_rate": 5.739832665089604e-06,
"loss": 0.8010875582695007,
"step": 964
},
{
"epoch": 2.0422832980972516,
"grad_norm": 1.1178871393203735,
"learning_rate": 5.724119406373171e-06,
"loss": 0.5907366275787354,
"step": 966
},
{
"epoch": 2.046511627906977,
"grad_norm": 3.8515806198120117,
"learning_rate": 5.708403407706954e-06,
"loss": 0.7840123772621155,
"step": 968
},
{
"epoch": 2.050739957716702,
"grad_norm": 1.211416244506836,
"learning_rate": 5.692684861225349e-06,
"loss": 0.9240961670875549,
"step": 970
},
{
"epoch": 2.0549682875264272,
"grad_norm": 2.7669265270233154,
"learning_rate": 5.676963959093904e-06,
"loss": 0.6099101305007935,
"step": 972
},
{
"epoch": 2.059196617336152,
"grad_norm": 1.5010597705841064,
"learning_rate": 5.66124089350696e-06,
"loss": 0.8929619193077087,
"step": 974
},
{
"epoch": 2.063424947145877,
"grad_norm": 1.254497766494751,
"learning_rate": 5.645515856685316e-06,
"loss": 0.9122307300567627,
"step": 976
},
{
"epoch": 2.0676532769556024,
"grad_norm": 0.8442611694335938,
"learning_rate": 5.629789040873862e-06,
"loss": 0.9118245244026184,
"step": 978
},
{
"epoch": 2.0718816067653276,
"grad_norm": 2.1886403560638428,
"learning_rate": 5.614060638339238e-06,
"loss": 0.6734197735786438,
"step": 980
},
{
"epoch": 2.076109936575053,
"grad_norm": 4.2747578620910645,
"learning_rate": 5.598330841367487e-06,
"loss": 0.6575128436088562,
"step": 982
},
{
"epoch": 2.080338266384778,
"grad_norm": 1.1895456314086914,
"learning_rate": 5.582599842261693e-06,
"loss": 0.6742628216743469,
"step": 984
},
{
"epoch": 2.0845665961945032,
"grad_norm": 1.383929967880249,
"learning_rate": 5.566867833339644e-06,
"loss": 0.8560068607330322,
"step": 986
},
{
"epoch": 2.0887949260042284,
"grad_norm": 1.5862306356430054,
"learning_rate": 5.551135006931467e-06,
"loss": 0.9456977248191833,
"step": 988
},
{
"epoch": 2.0930232558139537,
"grad_norm": 1.399377703666687,
"learning_rate": 5.535401555377287e-06,
"loss": 0.8707787990570068,
"step": 990
},
{
"epoch": 2.097251585623679,
"grad_norm": 3.597984790802002,
"learning_rate": 5.51966767102487e-06,
"loss": 0.5525308847427368,
"step": 992
},
{
"epoch": 2.1014799154334036,
"grad_norm": 2.3996775150299072,
"learning_rate": 5.503933546227273e-06,
"loss": 0.7161447405815125,
"step": 994
},
{
"epoch": 2.105708245243129,
"grad_norm": 1.079032301902771,
"learning_rate": 5.4881993733404916e-06,
"loss": 0.7069299221038818,
"step": 996
},
{
"epoch": 2.109936575052854,
"grad_norm": 0.9514685273170471,
"learning_rate": 5.472465344721113e-06,
"loss": 0.8299159407615662,
"step": 998
},
{
"epoch": 2.1141649048625792,
"grad_norm": 1.0228517055511475,
"learning_rate": 5.456731652723959e-06,
"loss": 0.9103403091430664,
"step": 1000
},
{
"epoch": 2.1183932346723044,
"grad_norm": 0.5176165699958801,
"learning_rate": 5.440998489699731e-06,
"loss": 0.29888591170310974,
"step": 1002
},
{
"epoch": 2.1226215644820297,
"grad_norm": 13.821861267089844,
"learning_rate": 5.425266047992672e-06,
"loss": 0.703010618686676,
"step": 1004
},
{
"epoch": 2.126849894291755,
"grad_norm": 1.331196665763855,
"learning_rate": 5.4095345199382e-06,
"loss": 0.966987133026123,
"step": 1006
},
{
"epoch": 2.13107822410148,
"grad_norm": 1.162968635559082,
"learning_rate": 5.393804097860568e-06,
"loss": 0.8339225053787231,
"step": 1008
},
{
"epoch": 2.1353065539112053,
"grad_norm": 0.7311312556266785,
"learning_rate": 5.3780749740705065e-06,
"loss": 0.5464614033699036,
"step": 1010
},
{
"epoch": 2.13953488372093,
"grad_norm": 2.364105701446533,
"learning_rate": 5.362347340862869e-06,
"loss": 0.5313394069671631,
"step": 1012
},
{
"epoch": 2.1437632135306552,
"grad_norm": 0.7674338817596436,
"learning_rate": 5.3466213905142935e-06,
"loss": 0.5691075325012207,
"step": 1014
},
{
"epoch": 2.1479915433403804,
"grad_norm": 1.1573779582977295,
"learning_rate": 5.3308973152808395e-06,
"loss": 0.788046658039093,
"step": 1016
},
{
"epoch": 2.1522198731501057,
"grad_norm": 26.79161262512207,
"learning_rate": 5.315175307395644e-06,
"loss": 0.8760414719581604,
"step": 1018
},
{
"epoch": 2.156448202959831,
"grad_norm": 2.838092803955078,
"learning_rate": 5.299455559066571e-06,
"loss": 0.753994345664978,
"step": 1020
},
{
"epoch": 2.160676532769556,
"grad_norm": 1.247550368309021,
"learning_rate": 5.283738262473859e-06,
"loss": 0.6406781077384949,
"step": 1022
},
{
"epoch": 2.1649048625792813,
"grad_norm": 1.8801127672195435,
"learning_rate": 5.268023609767772e-06,
"loss": 0.9306709170341492,
"step": 1024
},
{
"epoch": 2.1691331923890065,
"grad_norm": 1.0044602155685425,
"learning_rate": 5.2523117930662545e-06,
"loss": 0.8799101114273071,
"step": 1026
},
{
"epoch": 2.1733615221987317,
"grad_norm": 2.124178171157837,
"learning_rate": 5.2366030044525785e-06,
"loss": 0.8743594288825989,
"step": 1028
},
{
"epoch": 2.177589852008457,
"grad_norm": 0.8864532113075256,
"learning_rate": 5.220897435972992e-06,
"loss": 0.8489775657653809,
"step": 1030
},
{
"epoch": 2.1818181818181817,
"grad_norm": 1.3388663530349731,
"learning_rate": 5.205195279634385e-06,
"loss": 0.9601937532424927,
"step": 1032
},
{
"epoch": 2.186046511627907,
"grad_norm": 1.6758944988250732,
"learning_rate": 5.1894967274019245e-06,
"loss": 0.9500871896743774,
"step": 1034
},
{
"epoch": 2.190274841437632,
"grad_norm": 1.0116097927093506,
"learning_rate": 5.173801971196716e-06,
"loss": 0.6621299386024475,
"step": 1036
},
{
"epoch": 2.1945031712473573,
"grad_norm": 3.2613332271575928,
"learning_rate": 5.158111202893464e-06,
"loss": 0.4370371401309967,
"step": 1038
},
{
"epoch": 2.1987315010570825,
"grad_norm": 1.4374898672103882,
"learning_rate": 5.142424614318112e-06,
"loss": 0.9756577014923096,
"step": 1040
},
{
"epoch": 2.2029598308668077,
"grad_norm": 4.267157077789307,
"learning_rate": 5.12674239724551e-06,
"loss": 0.31933093070983887,
"step": 1042
},
{
"epoch": 2.207188160676533,
"grad_norm": 0.9203890562057495,
"learning_rate": 5.111064743397061e-06,
"loss": 0.90626060962677,
"step": 1044
},
{
"epoch": 2.211416490486258,
"grad_norm": 0.7490631341934204,
"learning_rate": 5.09539184443838e-06,
"loss": 0.5912353992462158,
"step": 1046
},
{
"epoch": 2.2156448202959833,
"grad_norm": 1.9676830768585205,
"learning_rate": 5.079723891976956e-06,
"loss": 0.6902493834495544,
"step": 1048
},
{
"epoch": 2.219873150105708,
"grad_norm": 2.5628654956817627,
"learning_rate": 5.064061077559803e-06,
"loss": 0.5639207363128662,
"step": 1050
},
{
"epoch": 2.2241014799154333,
"grad_norm": 1.2823482751846313,
"learning_rate": 5.048403592671116e-06,
"loss": 0.7992308735847473,
"step": 1052
},
{
"epoch": 2.2283298097251585,
"grad_norm": 1.024621844291687,
"learning_rate": 5.032751628729943e-06,
"loss": 0.8880590200424194,
"step": 1054
},
{
"epoch": 2.2325581395348837,
"grad_norm": 1.0416924953460693,
"learning_rate": 5.017105377087829e-06,
"loss": 0.9065860509872437,
"step": 1056
},
{
"epoch": 2.236786469344609,
"grad_norm": 0.31122004985809326,
"learning_rate": 5.0014650290264854e-06,
"loss": 0.6004120111465454,
"step": 1058
},
{
"epoch": 2.241014799154334,
"grad_norm": 0.9880802035331726,
"learning_rate": 4.9858307757554536e-06,
"loss": 0.694818913936615,
"step": 1060
},
{
"epoch": 2.2452431289640593,
"grad_norm": 3.2096593379974365,
"learning_rate": 4.97020280840976e-06,
"loss": 1.0144774913787842,
"step": 1062
},
{
"epoch": 2.2494714587737845,
"grad_norm": 1.538493037223816,
"learning_rate": 4.954581318047582e-06,
"loss": 0.8697788119316101,
"step": 1064
},
{
"epoch": 2.2536997885835097,
"grad_norm": 2.093294382095337,
"learning_rate": 4.938966495647918e-06,
"loss": 0.5204666256904602,
"step": 1066
},
{
"epoch": 2.2579281183932345,
"grad_norm": 1.0571445226669312,
"learning_rate": 4.923358532108244e-06,
"loss": 0.796515703201294,
"step": 1068
},
{
"epoch": 2.2621564482029597,
"grad_norm": 2.4674012660980225,
"learning_rate": 4.907757618242186e-06,
"loss": 0.6934959292411804,
"step": 1070
},
{
"epoch": 2.266384778012685,
"grad_norm": 0.975052535533905,
"learning_rate": 4.892163944777181e-06,
"loss": 0.9722970724105835,
"step": 1072
},
{
"epoch": 2.27061310782241,
"grad_norm": 0.8109495043754578,
"learning_rate": 4.876577702352155e-06,
"loss": 0.6717290878295898,
"step": 1074
},
{
"epoch": 2.2748414376321353,
"grad_norm": 4.0518670082092285,
"learning_rate": 4.860999081515183e-06,
"loss": 0.7028070688247681,
"step": 1076
},
{
"epoch": 2.2790697674418605,
"grad_norm": 2.1142730712890625,
"learning_rate": 4.845428272721161e-06,
"loss": 0.2363855242729187,
"step": 1078
},
{
"epoch": 2.2832980972515857,
"grad_norm": 0.5239435434341431,
"learning_rate": 4.8298654663294865e-06,
"loss": 0.11930263787508011,
"step": 1080
},
{
"epoch": 2.287526427061311,
"grad_norm": 0.1596202552318573,
"learning_rate": 4.814310852601717e-06,
"loss": 0.4791639745235443,
"step": 1082
},
{
"epoch": 2.291754756871036,
"grad_norm": 2.122868776321411,
"learning_rate": 4.7987646216992545e-06,
"loss": 0.31743818521499634,
"step": 1084
},
{
"epoch": 2.295983086680761,
"grad_norm": 0.71204674243927,
"learning_rate": 4.783226963681018e-06,
"loss": 0.15153108537197113,
"step": 1086
},
{
"epoch": 2.300211416490486,
"grad_norm": 4.329046249389648,
"learning_rate": 4.76769806850112e-06,
"loss": 0.6926787495613098,
"step": 1088
},
{
"epoch": 2.3044397463002113,
"grad_norm": 1.0494717359542847,
"learning_rate": 4.752178126006544e-06,
"loss": 0.9127827882766724,
"step": 1090
},
{
"epoch": 2.3086680761099365,
"grad_norm": 3.4832351207733154,
"learning_rate": 4.736667325934819e-06,
"loss": 0.5838255882263184,
"step": 1092
},
{
"epoch": 2.3128964059196617,
"grad_norm": 0.9454461336135864,
"learning_rate": 4.721165857911709e-06,
"loss": 0.8850477933883667,
"step": 1094
},
{
"epoch": 2.317124735729387,
"grad_norm": 0.823363184928894,
"learning_rate": 4.705673911448891e-06,
"loss": 0.7873114943504333,
"step": 1096
},
{
"epoch": 2.321353065539112,
"grad_norm": 1.3021113872528076,
"learning_rate": 4.690191675941631e-06,
"loss": 0.865074634552002,
"step": 1098
},
{
"epoch": 2.3255813953488373,
"grad_norm": 8.062743186950684,
"learning_rate": 4.674719340666477e-06,
"loss": 0.5237413048744202,
"step": 1100
},
{
"epoch": 2.3298097251585626,
"grad_norm": 0.9624851942062378,
"learning_rate": 4.659257094778949e-06,
"loss": 0.6671454906463623,
"step": 1102
},
{
"epoch": 2.3340380549682873,
"grad_norm": 1.3742395639419556,
"learning_rate": 4.643805127311212e-06,
"loss": 0.9226606488227844,
"step": 1104
},
{
"epoch": 2.3382663847780125,
"grad_norm": 0.9934477210044861,
"learning_rate": 4.628363627169774e-06,
"loss": 0.8610000014305115,
"step": 1106
},
{
"epoch": 2.3424947145877377,
"grad_norm": 0.9795897603034973,
"learning_rate": 4.612932783133182e-06,
"loss": 0.859033465385437,
"step": 1108
},
{
"epoch": 2.346723044397463,
"grad_norm": 1.6531788110733032,
"learning_rate": 4.5975127838497036e-06,
"loss": 0.9669740796089172,
"step": 1110
},
{
"epoch": 2.350951374207188,
"grad_norm": 1.266805648803711,
"learning_rate": 4.582103817835021e-06,
"loss": 0.5835067629814148,
"step": 1112
},
{
"epoch": 2.3551797040169133,
"grad_norm": 2.969149351119995,
"learning_rate": 4.566706073469938e-06,
"loss": 0.8170666098594666,
"step": 1114
},
{
"epoch": 2.3594080338266386,
"grad_norm": 1.5357444286346436,
"learning_rate": 4.551319738998067e-06,
"loss": 0.6888680458068848,
"step": 1116
},
{
"epoch": 2.3636363636363638,
"grad_norm": 2.480497121810913,
"learning_rate": 4.535945002523524e-06,
"loss": 1.0695511102676392,
"step": 1118
},
{
"epoch": 2.367864693446089,
"grad_norm": 0.9709950089454651,
"learning_rate": 4.5205820520086415e-06,
"loss": 0.9124900102615356,
"step": 1120
},
{
"epoch": 2.3720930232558137,
"grad_norm": 6.447284698486328,
"learning_rate": 4.505231075271662e-06,
"loss": 0.7588623762130737,
"step": 1122
},
{
"epoch": 2.376321353065539,
"grad_norm": 1.0373855829238892,
"learning_rate": 4.489892259984446e-06,
"loss": 0.6135239005088806,
"step": 1124
},
{
"epoch": 2.380549682875264,
"grad_norm": 3.688616991043091,
"learning_rate": 4.474565793670167e-06,
"loss": 0.9221339225769043,
"step": 1126
},
{
"epoch": 2.3847780126849893,
"grad_norm": 2.0843348503112793,
"learning_rate": 4.459251863701038e-06,
"loss": 0.6196430921554565,
"step": 1128
},
{
"epoch": 2.3890063424947146,
"grad_norm": 2.2064480781555176,
"learning_rate": 4.443950657296004e-06,
"loss": 0.7247291207313538,
"step": 1130
},
{
"epoch": 2.3932346723044398,
"grad_norm": 2.03659987449646,
"learning_rate": 4.4286623615184585e-06,
"loss": 0.5389364361763,
"step": 1132
},
{
"epoch": 2.397463002114165,
"grad_norm": 0.9074171781539917,
"learning_rate": 4.413387163273964e-06,
"loss": 0.9406471252441406,
"step": 1134
},
{
"epoch": 2.40169133192389,
"grad_norm": 5.236852645874023,
"learning_rate": 4.3981252493079506e-06,
"loss": 0.6433354616165161,
"step": 1136
},
{
"epoch": 2.4059196617336154,
"grad_norm": 10.138160705566406,
"learning_rate": 4.382876806203453e-06,
"loss": 0.7463909983634949,
"step": 1138
},
{
"epoch": 2.41014799154334,
"grad_norm": 1.043208360671997,
"learning_rate": 4.3676420203788125e-06,
"loss": 0.88657546043396,
"step": 1140
},
{
"epoch": 2.4143763213530653,
"grad_norm": 0.8148168325424194,
"learning_rate": 4.352421078085405e-06,
"loss": 0.8769685626029968,
"step": 1142
},
{
"epoch": 2.4186046511627906,
"grad_norm": 1.4624063968658447,
"learning_rate": 4.337214165405368e-06,
"loss": 0.5223655700683594,
"step": 1144
},
{
"epoch": 2.4228329809725158,
"grad_norm": 3.0132341384887695,
"learning_rate": 4.322021468249316e-06,
"loss": 0.7219529747962952,
"step": 1146
},
{
"epoch": 2.427061310782241,
"grad_norm": 1.8713390827178955,
"learning_rate": 4.306843172354076e-06,
"loss": 0.7166211605072021,
"step": 1148
},
{
"epoch": 2.431289640591966,
"grad_norm": 1.1524786949157715,
"learning_rate": 4.291679463280416e-06,
"loss": 0.466641366481781,
"step": 1150
},
{
"epoch": 2.4355179704016914,
"grad_norm": 0.8421763181686401,
"learning_rate": 4.276530526410768e-06,
"loss": 0.5435295104980469,
"step": 1152
},
{
"epoch": 2.4397463002114166,
"grad_norm": 1.090937852859497,
"learning_rate": 4.261396546946972e-06,
"loss": 0.5777373313903809,
"step": 1154
},
{
"epoch": 2.443974630021142,
"grad_norm": 1.1153584718704224,
"learning_rate": 4.246277709908009e-06,
"loss": 0.8293700218200684,
"step": 1156
},
{
"epoch": 2.448202959830867,
"grad_norm": 1.2401615381240845,
"learning_rate": 4.231174200127736e-06,
"loss": 0.5208355188369751,
"step": 1158
},
{
"epoch": 2.452431289640592,
"grad_norm": 0.6799382567405701,
"learning_rate": 4.216086202252624e-06,
"loss": 0.8532727956771851,
"step": 1160
},
{
"epoch": 2.456659619450317,
"grad_norm": 1.3806763887405396,
"learning_rate": 4.201013900739513e-06,
"loss": 0.886535108089447,
"step": 1162
},
{
"epoch": 2.460887949260042,
"grad_norm": 1.3361854553222656,
"learning_rate": 4.185957479853344e-06,
"loss": 0.695877194404602,
"step": 1164
},
{
"epoch": 2.4651162790697674,
"grad_norm": 1.512601375579834,
"learning_rate": 4.1709171236649096e-06,
"loss": 1.0502127408981323,
"step": 1166
},
{
"epoch": 2.4693446088794926,
"grad_norm": 2.1257808208465576,
"learning_rate": 4.155893016048607e-06,
"loss": 1.0154898166656494,
"step": 1168
},
{
"epoch": 2.473572938689218,
"grad_norm": 0.8388662934303284,
"learning_rate": 4.140885340680189e-06,
"loss": 0.612666130065918,
"step": 1170
},
{
"epoch": 2.477801268498943,
"grad_norm": 2.466283082962036,
"learning_rate": 4.125894281034517e-06,
"loss": 0.7260602116584778,
"step": 1172
},
{
"epoch": 2.482029598308668,
"grad_norm": 0.88197922706604,
"learning_rate": 4.110920020383316e-06,
"loss": 0.6661515235900879,
"step": 1174
},
{
"epoch": 2.4862579281183934,
"grad_norm": 1.7739195823669434,
"learning_rate": 4.0959627417929386e-06,
"loss": 0.7041642665863037,
"step": 1176
},
{
"epoch": 2.4904862579281186,
"grad_norm": 2.1739261150360107,
"learning_rate": 4.081022628122127e-06,
"loss": 0.8279573321342468,
"step": 1178
},
{
"epoch": 2.4947145877378434,
"grad_norm": 7.311237335205078,
"learning_rate": 4.06609986201977e-06,
"loss": 0.579248309135437,
"step": 1180
},
{
"epoch": 2.4989429175475686,
"grad_norm": 0.5175657272338867,
"learning_rate": 4.051194625922681e-06,
"loss": 0.355604887008667,
"step": 1182
},
{
"epoch": 2.503171247357294,
"grad_norm": 0.7100090384483337,
"learning_rate": 4.036307102053359e-06,
"loss": 0.11728010326623917,
"step": 1184
},
{
"epoch": 2.507399577167019,
"grad_norm": 1.0904265642166138,
"learning_rate": 4.021437472417763e-06,
"loss": 0.906569242477417,
"step": 1186
},
{
"epoch": 2.511627906976744,
"grad_norm": 34.20032501220703,
"learning_rate": 4.006585918803091e-06,
"loss": 0.657234787940979,
"step": 1188
},
{
"epoch": 2.5158562367864694,
"grad_norm": 2.3317790031433105,
"learning_rate": 3.991752622775551e-06,
"loss": 0.6361251473426819,
"step": 1190
},
{
"epoch": 2.5200845665961946,
"grad_norm": 0.9664400815963745,
"learning_rate": 3.97693776567815e-06,
"loss": 0.8825923800468445,
"step": 1192
},
{
"epoch": 2.52431289640592,
"grad_norm": 0.8697760701179504,
"learning_rate": 3.962141528628465e-06,
"loss": 0.5446211695671082,
"step": 1194
},
{
"epoch": 2.528541226215645,
"grad_norm": 2.98101544380188,
"learning_rate": 3.947364092516437e-06,
"loss": 0.9272128343582153,
"step": 1196
},
{
"epoch": 2.53276955602537,
"grad_norm": 0.8588328957557678,
"learning_rate": 3.932605638002163e-06,
"loss": 0.84901362657547,
"step": 1198
},
{
"epoch": 2.536997885835095,
"grad_norm": 2.9067835807800293,
"learning_rate": 3.917866345513677e-06,
"loss": 0.7447222471237183,
"step": 1200
},
{
"epoch": 2.54122621564482,
"grad_norm": 1.581717610359192,
"learning_rate": 3.903146395244749e-06,
"loss": 0.8643378019332886,
"step": 1202
},
{
"epoch": 2.5454545454545454,
"grad_norm": 3.1515986919403076,
"learning_rate": 3.888445967152691e-06,
"loss": 0.29685986042022705,
"step": 1204
},
{
"epoch": 2.5496828752642706,
"grad_norm": 1.3771530389785767,
"learning_rate": 3.873765240956138e-06,
"loss": 0.9426017999649048,
"step": 1206
},
{
"epoch": 2.553911205073996,
"grad_norm": 1.1528037786483765,
"learning_rate": 3.859104396132863e-06,
"loss": 0.9731027483940125,
"step": 1208
},
{
"epoch": 2.558139534883721,
"grad_norm": 1.3156410455703735,
"learning_rate": 3.84446361191759e-06,
"loss": 0.8804532885551453,
"step": 1210
},
{
"epoch": 2.5623678646934462,
"grad_norm": 1.5303044319152832,
"learning_rate": 3.829843067299788e-06,
"loss": 0.6254847645759583,
"step": 1212
},
{
"epoch": 2.5665961945031714,
"grad_norm": 1.4074640274047852,
"learning_rate": 3.8152429410214834e-06,
"loss": 0.9209861755371094,
"step": 1214
},
{
"epoch": 2.570824524312896,
"grad_norm": 1.7678253650665283,
"learning_rate": 3.8006634115750886e-06,
"loss": 0.5832459926605225,
"step": 1216
},
{
"epoch": 2.5750528541226214,
"grad_norm": 2.1125075817108154,
"learning_rate": 3.7861046572012105e-06,
"loss": 0.4877403974533081,
"step": 1218
},
{
"epoch": 2.5792811839323466,
"grad_norm": 3.4932961463928223,
"learning_rate": 3.771566855886467e-06,
"loss": 0.5127173066139221,
"step": 1220
},
{
"epoch": 2.583509513742072,
"grad_norm": 1.784134864807129,
"learning_rate": 3.7570501853613227e-06,
"loss": 0.49880778789520264,
"step": 1222
},
{
"epoch": 2.587737843551797,
"grad_norm": 6.833421230316162,
"learning_rate": 3.742554823097905e-06,
"loss": 0.7275419235229492,
"step": 1224
},
{
"epoch": 2.5919661733615222,
"grad_norm": 0.4500092566013336,
"learning_rate": 3.7280809463078436e-06,
"loss": 0.33249756693840027,
"step": 1226
},
{
"epoch": 2.5961945031712474,
"grad_norm": 1.6668881177902222,
"learning_rate": 3.7136287319400944e-06,
"loss": 0.7344802021980286,
"step": 1228
},
{
"epoch": 2.6004228329809727,
"grad_norm": 1.5321757793426514,
"learning_rate": 3.6991983566787847e-06,
"loss": 0.7713485956192017,
"step": 1230
},
{
"epoch": 2.604651162790698,
"grad_norm": 1.5511820316314697,
"learning_rate": 3.6847899969410515e-06,
"loss": 0.5699136853218079,
"step": 1232
},
{
"epoch": 2.6088794926004226,
"grad_norm": 2.480119466781616,
"learning_rate": 3.6704038288748812e-06,
"loss": 0.664405345916748,
"step": 1234
},
{
"epoch": 2.613107822410148,
"grad_norm": 1.0433802604675293,
"learning_rate": 3.6560400283569543e-06,
"loss": 0.8138096928596497,
"step": 1236
},
{
"epoch": 2.617336152219873,
"grad_norm": 1.3131661415100098,
"learning_rate": 3.641698770990507e-06,
"loss": 0.8590487241744995,
"step": 1238
},
{
"epoch": 2.6215644820295982,
"grad_norm": 1.2021374702453613,
"learning_rate": 3.6273802321031738e-06,
"loss": 0.6605381369590759,
"step": 1240
},
{
"epoch": 2.6257928118393234,
"grad_norm": 1.2149405479431152,
"learning_rate": 3.6130845867448384e-06,
"loss": 0.7428333163261414,
"step": 1242
},
{
"epoch": 2.6300211416490487,
"grad_norm": 3.560697078704834,
"learning_rate": 3.598812009685514e-06,
"loss": 0.8363604545593262,
"step": 1244
},
{
"epoch": 2.634249471458774,
"grad_norm": 5.153426647186279,
"learning_rate": 3.584562675413189e-06,
"loss": 0.5476874113082886,
"step": 1246
},
{
"epoch": 2.638477801268499,
"grad_norm": 18.06144905090332,
"learning_rate": 3.5703367581316985e-06,
"loss": 0.445987343788147,
"step": 1248
},
{
"epoch": 2.6427061310782243,
"grad_norm": 1.899404764175415,
"learning_rate": 3.5561344317585957e-06,
"loss": 0.9049619436264038,
"step": 1250
},
{
"epoch": 2.646934460887949,
"grad_norm": 7.480179309844971,
"learning_rate": 3.5419558699230305e-06,
"loss": 0.3996701240539551,
"step": 1252
},
{
"epoch": 2.6511627906976747,
"grad_norm": 4.1430439949035645,
"learning_rate": 3.52780124596361e-06,
"loss": 0.5693390369415283,
"step": 1254
},
{
"epoch": 2.6553911205073994,
"grad_norm": 0.9929502606391907,
"learning_rate": 3.5136707329263055e-06,
"loss": 0.9592902064323425,
"step": 1256
},
{
"epoch": 2.6596194503171247,
"grad_norm": 0.9065409302711487,
"learning_rate": 3.499564503562315e-06,
"loss": 0.19373737275600433,
"step": 1258
},
{
"epoch": 2.66384778012685,
"grad_norm": 1.1277961730957031,
"learning_rate": 3.4854827303259575e-06,
"loss": 0.4543590247631073,
"step": 1260
},
{
"epoch": 2.668076109936575,
"grad_norm": 2.8574512004852295,
"learning_rate": 3.47142558537257e-06,
"loss": 0.6632375717163086,
"step": 1262
},
{
"epoch": 2.6723044397463003,
"grad_norm": 3.951639175415039,
"learning_rate": 3.4573932405563966e-06,
"loss": 0.7306318283081055,
"step": 1264
},
{
"epoch": 2.6765327695560255,
"grad_norm": 7.889729976654053,
"learning_rate": 3.4433858674284894e-06,
"loss": 0.3242054283618927,
"step": 1266
},
{
"epoch": 2.6807610993657507,
"grad_norm": 1.3038914203643799,
"learning_rate": 3.429403637234614e-06,
"loss": 0.6848164796829224,
"step": 1268
},
{
"epoch": 2.6849894291754755,
"grad_norm": 1.040703535079956,
"learning_rate": 3.4154467209131515e-06,
"loss": 0.9264628887176514,
"step": 1270
},
{
"epoch": 2.689217758985201,
"grad_norm": 10.250706672668457,
"learning_rate": 3.401515289093011e-06,
"loss": 0.6427375674247742,
"step": 1272
},
{
"epoch": 2.693446088794926,
"grad_norm": 0.6268860101699829,
"learning_rate": 3.3876095120915433e-06,
"loss": 0.467488169670105,
"step": 1274
},
{
"epoch": 2.697674418604651,
"grad_norm": 1.3762441873550415,
"learning_rate": 3.3737295599124592e-06,
"loss": 0.498369038105011,
"step": 1276
},
{
"epoch": 2.7019027484143763,
"grad_norm": 1.6863068342208862,
"learning_rate": 3.3598756022437507e-06,
"loss": 0.8639737367630005,
"step": 1278
},
{
"epoch": 2.7061310782241015,
"grad_norm": 1.236923098564148,
"learning_rate": 3.3460478084556157e-06,
"loss": 0.6078754663467407,
"step": 1280
},
{
"epoch": 2.7103594080338267,
"grad_norm": 0.8999261856079102,
"learning_rate": 3.33224634759839e-06,
"loss": 0.947062611579895,
"step": 1282
},
{
"epoch": 2.714587737843552,
"grad_norm": 0.36551353335380554,
"learning_rate": 3.318471388400476e-06,
"loss": 0.6062276363372803,
"step": 1284
},
{
"epoch": 2.718816067653277,
"grad_norm": 0.2991828918457031,
"learning_rate": 3.3047230992662834e-06,
"loss": 1.009516954421997,
"step": 1286
},
{
"epoch": 2.723044397463002,
"grad_norm": 0.7167829871177673,
"learning_rate": 3.291001648274175e-06,
"loss": 0.44804781675338745,
"step": 1288
},
{
"epoch": 2.7272727272727275,
"grad_norm": 1.2276729345321655,
"learning_rate": 3.277307203174395e-06,
"loss": 0.7913851141929626,
"step": 1290
},
{
"epoch": 2.7315010570824523,
"grad_norm": 0.8791899085044861,
"learning_rate": 3.263639931387039e-06,
"loss": 0.8930345773696899,
"step": 1292
},
{
"epoch": 2.7357293868921775,
"grad_norm": 1.1345866918563843,
"learning_rate": 3.2500000000000015e-06,
"loss": 0.89072585105896,
"step": 1294
},
{
"epoch": 2.7399577167019027,
"grad_norm": 0.5883175730705261,
"learning_rate": 3.236387575766916e-06,
"loss": 0.3013880252838135,
"step": 1296
},
{
"epoch": 2.744186046511628,
"grad_norm": 1.4821778535842896,
"learning_rate": 3.222802825105138e-06,
"loss": 0.4950874149799347,
"step": 1298
},
{
"epoch": 2.748414376321353,
"grad_norm": 1.7014998197555542,
"learning_rate": 3.209245914093708e-06,
"loss": 0.5958110094070435,
"step": 1300
},
{
"epoch": 2.7526427061310783,
"grad_norm": 1.4416790008544922,
"learning_rate": 3.1957170084713053e-06,
"loss": 0.7179307341575623,
"step": 1302
},
{
"epoch": 2.7568710359408035,
"grad_norm": 0.8473976254463196,
"learning_rate": 3.182216273634235e-06,
"loss": 0.5455998182296753,
"step": 1304
},
{
"epoch": 2.7610993657505283,
"grad_norm": 3.5670745372772217,
"learning_rate": 3.1687438746344086e-06,
"loss": 0.7428355813026428,
"step": 1306
},
{
"epoch": 2.765327695560254,
"grad_norm": 24.615921020507812,
"learning_rate": 3.1552999761773166e-06,
"loss": 0.7578653693199158,
"step": 1308
},
{
"epoch": 2.7695560253699787,
"grad_norm": 0.8577094674110413,
"learning_rate": 3.141884742620013e-06,
"loss": 0.8551408052444458,
"step": 1310
},
{
"epoch": 2.773784355179704,
"grad_norm": 1.1516927480697632,
"learning_rate": 3.1284983379691247e-06,
"loss": 0.8440461754798889,
"step": 1312
},
{
"epoch": 2.778012684989429,
"grad_norm": 0.4825911223888397,
"learning_rate": 3.1151409258788235e-06,
"loss": 0.48459291458129883,
"step": 1314
},
{
"epoch": 2.7822410147991543,
"grad_norm": 1.0467877388000488,
"learning_rate": 3.101812669648838e-06,
"loss": 0.8143327832221985,
"step": 1316
},
{
"epoch": 2.7864693446088795,
"grad_norm": 0.39678460359573364,
"learning_rate": 3.0885137322224546e-06,
"loss": 0.5902388095855713,
"step": 1318
},
{
"epoch": 2.7906976744186047,
"grad_norm": 1.3878825902938843,
"learning_rate": 3.075244276184529e-06,
"loss": 0.8759530782699585,
"step": 1320
},
{
"epoch": 2.79492600422833,
"grad_norm": 3.2891933917999268,
"learning_rate": 3.0620044637594936e-06,
"loss": 0.4169163107872009,
"step": 1322
},
{
"epoch": 2.7991543340380547,
"grad_norm": 1.8422366380691528,
"learning_rate": 3.0487944568093685e-06,
"loss": 0.9175578951835632,
"step": 1324
},
{
"epoch": 2.8033826638477803,
"grad_norm": 2.5718417167663574,
"learning_rate": 3.035614416831801e-06,
"loss": 0.20231308043003082,
"step": 1326
},
{
"epoch": 2.807610993657505,
"grad_norm": 1.7762324810028076,
"learning_rate": 3.0224645049580746e-06,
"loss": 0.33037495613098145,
"step": 1328
},
{
"epoch": 2.8118393234672303,
"grad_norm": 0.8373496532440186,
"learning_rate": 3.0093448819511416e-06,
"loss": 0.5284480452537537,
"step": 1330
},
{
"epoch": 2.8160676532769555,
"grad_norm": 1.5008878707885742,
"learning_rate": 2.9962557082036677e-06,
"loss": 0.916854977607727,
"step": 1332
},
{
"epoch": 2.8202959830866807,
"grad_norm": 1.02738356590271,
"learning_rate": 2.9831971437360606e-06,
"loss": 0.5735000967979431,
"step": 1334
},
{
"epoch": 2.824524312896406,
"grad_norm": 1.2757019996643066,
"learning_rate": 2.9701693481945132e-06,
"loss": 0.8907558917999268,
"step": 1336
},
{
"epoch": 2.828752642706131,
"grad_norm": 1.360640525817871,
"learning_rate": 2.957172480849064e-06,
"loss": 0.9124787449836731,
"step": 1338
},
{
"epoch": 2.8329809725158563,
"grad_norm": 1.057644009590149,
"learning_rate": 2.9442067005916375e-06,
"loss": 0.7098805904388428,
"step": 1340
},
{
"epoch": 2.8372093023255816,
"grad_norm": 1.3314310312271118,
"learning_rate": 2.931272165934105e-06,
"loss": 0.9696467518806458,
"step": 1342
},
{
"epoch": 2.8414376321353068,
"grad_norm": 0.9993709921836853,
"learning_rate": 2.918369035006351e-06,
"loss": 0.5990903973579407,
"step": 1344
},
{
"epoch": 2.8456659619450315,
"grad_norm": 2.0502212047576904,
"learning_rate": 2.9054974655543335e-06,
"loss": 0.6769232153892517,
"step": 1346
},
{
"epoch": 2.8498942917547567,
"grad_norm": 0.7910188436508179,
"learning_rate": 2.89265761493816e-06,
"loss": 0.8802707195281982,
"step": 1348
},
{
"epoch": 2.854122621564482,
"grad_norm": 5.838414669036865,
"learning_rate": 2.8798496401301644e-06,
"loss": 0.5067776441574097,
"step": 1350
},
{
"epoch": 2.858350951374207,
"grad_norm": 1.9082633256912231,
"learning_rate": 2.8670736977129843e-06,
"loss": 0.8332048654556274,
"step": 1352
},
{
"epoch": 2.8625792811839323,
"grad_norm": 1.0120817422866821,
"learning_rate": 2.854329943877647e-06,
"loss": 0.8753684163093567,
"step": 1354
},
{
"epoch": 2.8668076109936576,
"grad_norm": 2.5912153720855713,
"learning_rate": 2.841618534421664e-06,
"loss": 0.2954685091972351,
"step": 1356
},
{
"epoch": 2.8710359408033828,
"grad_norm": 0.930406391620636,
"learning_rate": 2.8289396247471227e-06,
"loss": 0.9577110409736633,
"step": 1358
},
{
"epoch": 2.875264270613108,
"grad_norm": 10.650595664978027,
"learning_rate": 2.816293369858788e-06,
"loss": 0.5425952076911926,
"step": 1360
},
{
"epoch": 2.879492600422833,
"grad_norm": 0.9867213368415833,
"learning_rate": 2.803679924362206e-06,
"loss": 0.6205761432647705,
"step": 1362
},
{
"epoch": 2.883720930232558,
"grad_norm": 1.0803779363632202,
"learning_rate": 2.7910994424618165e-06,
"loss": 0.9687942266464233,
"step": 1364
},
{
"epoch": 2.887949260042283,
"grad_norm": 1.8729515075683594,
"learning_rate": 2.778552077959064e-06,
"loss": 0.8614194989204407,
"step": 1366
},
{
"epoch": 2.8921775898520083,
"grad_norm": 1.3931171894073486,
"learning_rate": 2.766037984250521e-06,
"loss": 0.886341392993927,
"step": 1368
},
{
"epoch": 2.8964059196617336,
"grad_norm": 1.0148799419403076,
"learning_rate": 2.753557314326011e-06,
"loss": 0.908317506313324,
"step": 1370
},
{
"epoch": 2.9006342494714588,
"grad_norm": 4.6803765296936035,
"learning_rate": 2.7411102207667373e-06,
"loss": 0.9467305541038513,
"step": 1372
},
{
"epoch": 2.904862579281184,
"grad_norm": 3.198029041290283,
"learning_rate": 2.728696855743418e-06,
"loss": 0.8886833786964417,
"step": 1374
},
{
"epoch": 2.909090909090909,
"grad_norm": 1.0209606885910034,
"learning_rate": 2.7163173710144323e-06,
"loss": 0.8771613240242004,
"step": 1376
},
{
"epoch": 2.9133192389006344,
"grad_norm": 2.2020952701568604,
"learning_rate": 2.703971917923951e-06,
"loss": 0.5990895628929138,
"step": 1378
},
{
"epoch": 2.9175475687103596,
"grad_norm": 0.9331658482551575,
"learning_rate": 2.6916606474000965e-06,
"loss": 0.6395297646522522,
"step": 1380
},
{
"epoch": 2.9217758985200843,
"grad_norm": 1.51901113986969,
"learning_rate": 2.679383709953103e-06,
"loss": 0.8922448754310608,
"step": 1382
},
{
"epoch": 2.92600422832981,
"grad_norm": 2.52286434173584,
"learning_rate": 2.6671412556734578e-06,
"loss": 0.7213451862335205,
"step": 1384
},
{
"epoch": 2.9302325581395348,
"grad_norm": 0.9286597967147827,
"learning_rate": 2.6549334342300844e-06,
"loss": 0.5824087262153625,
"step": 1386
},
{
"epoch": 2.93446088794926,
"grad_norm": 1.4062398672103882,
"learning_rate": 2.642760394868503e-06,
"loss": 0.852761447429657,
"step": 1388
},
{
"epoch": 2.938689217758985,
"grad_norm": 0.27306583523750305,
"learning_rate": 2.6306222864090157e-06,
"loss": 0.6247194409370422,
"step": 1390
},
{
"epoch": 2.9429175475687104,
"grad_norm": 1.4418777227401733,
"learning_rate": 2.61851925724487e-06,
"loss": 0.8824759125709534,
"step": 1392
},
{
"epoch": 2.9471458773784356,
"grad_norm": 0.35564297437667847,
"learning_rate": 2.606451455340459e-06,
"loss": 0.23137271404266357,
"step": 1394
},
{
"epoch": 2.951374207188161,
"grad_norm": 0.9093311429023743,
"learning_rate": 2.5944190282295155e-06,
"loss": 0.47480979561805725,
"step": 1396
},
{
"epoch": 2.955602536997886,
"grad_norm": 1.6398972272872925,
"learning_rate": 2.5824221230132895e-06,
"loss": 0.9465433359146118,
"step": 1398
},
{
"epoch": 2.9598308668076108,
"grad_norm": 1.07735276222229,
"learning_rate": 2.5704608863587676e-06,
"loss": 1.004073977470398,
"step": 1400
},
{
"epoch": 2.9640591966173364,
"grad_norm": 2.2225751876831055,
"learning_rate": 2.558535464496877e-06,
"loss": 0.7205920219421387,
"step": 1402
},
{
"epoch": 2.968287526427061,
"grad_norm": 0.7435517311096191,
"learning_rate": 2.5466460032206872e-06,
"loss": 0.47887614369392395,
"step": 1404
},
{
"epoch": 2.9725158562367864,
"grad_norm": 2.481041431427002,
"learning_rate": 2.53479264788364e-06,
"loss": 0.8457914590835571,
"step": 1406
},
{
"epoch": 2.9767441860465116,
"grad_norm": 1.012760877609253,
"learning_rate": 2.522975543397769e-06,
"loss": 0.6757799386978149,
"step": 1408
},
{
"epoch": 2.980972515856237,
"grad_norm": 3.697950839996338,
"learning_rate": 2.511194834231922e-06,
"loss": 0.715408444404602,
"step": 1410
},
{
"epoch": 2.985200845665962,
"grad_norm": 1.019342303276062,
"learning_rate": 2.49945066441e-06,
"loss": 0.9022645950317383,
"step": 1412
},
{
"epoch": 2.989429175475687,
"grad_norm": 5.096213340759277,
"learning_rate": 2.4877431775092004e-06,
"loss": 0.6168473958969116,
"step": 1414
},
{
"epoch": 2.9936575052854124,
"grad_norm": 2.5709292888641357,
"learning_rate": 2.4760725166582544e-06,
"loss": 0.7178599834442139,
"step": 1416
},
{
"epoch": 2.997885835095137,
"grad_norm": 2.9766063690185547,
"learning_rate": 2.464438824535675e-06,
"loss": 0.46865221858024597,
"step": 1418
},
{
"epoch": 3.0021141649048624,
"grad_norm": 0.9445246458053589,
"learning_rate": 2.452842243368028e-06,
"loss": 0.8217405676841736,
"step": 1420
},
{
"epoch": 3.0063424947145876,
"grad_norm": 2.723698377609253,
"learning_rate": 2.4412829149281735e-06,
"loss": 0.5070158839225769,
"step": 1422
},
{
"epoch": 3.010570824524313,
"grad_norm": 2.3810276985168457,
"learning_rate": 2.4297609805335472e-06,
"loss": 0.9459589123725891,
"step": 1424
},
{
"epoch": 3.014799154334038,
"grad_norm": 3.9616851806640625,
"learning_rate": 2.4182765810444263e-06,
"loss": 0.4811483323574066,
"step": 1426
},
{
"epoch": 3.019027484143763,
"grad_norm": 0.9534226655960083,
"learning_rate": 2.4068298568622083e-06,
"loss": 0.4378834664821625,
"step": 1428
},
{
"epoch": 3.0232558139534884,
"grad_norm": 1.0616681575775146,
"learning_rate": 2.395420947927694e-06,
"loss": 0.6399895548820496,
"step": 1430
},
{
"epoch": 3.0274841437632136,
"grad_norm": 1.937260627746582,
"learning_rate": 2.3840499937193804e-06,
"loss": 0.7214668989181519,
"step": 1432
},
{
"epoch": 3.031712473572939,
"grad_norm": 0.9942629337310791,
"learning_rate": 2.3727171332517496e-06,
"loss": 0.7045053839683533,
"step": 1434
},
{
"epoch": 3.035940803382664,
"grad_norm": 4.3150434494018555,
"learning_rate": 2.361422505073575e-06,
"loss": 0.48319727182388306,
"step": 1436
},
{
"epoch": 3.040169133192389,
"grad_norm": 0.6073245406150818,
"learning_rate": 2.3501662472662233e-06,
"loss": 0.14261965453624725,
"step": 1438
},
{
"epoch": 3.044397463002114,
"grad_norm": 1.5740426778793335,
"learning_rate": 2.3389484974419685e-06,
"loss": 0.824199378490448,
"step": 1440
},
{
"epoch": 3.048625792811839,
"grad_norm": 1.0348553657531738,
"learning_rate": 2.327769392742309e-06,
"loss": 0.5016360282897949,
"step": 1442
},
{
"epoch": 3.0528541226215644,
"grad_norm": 3.911590576171875,
"learning_rate": 2.316629069836291e-06,
"loss": 0.1290048062801361,
"step": 1444
},
{
"epoch": 3.0570824524312896,
"grad_norm": 1.6653960943222046,
"learning_rate": 2.305527664918838e-06,
"loss": 0.4502914547920227,
"step": 1446
},
{
"epoch": 3.061310782241015,
"grad_norm": 2.256114959716797,
"learning_rate": 2.294465313709085e-06,
"loss": 0.5582813024520874,
"step": 1448
},
{
"epoch": 3.06553911205074,
"grad_norm": 1.387699842453003,
"learning_rate": 2.2834421514487205e-06,
"loss": 0.4848635792732239,
"step": 1450
},
{
"epoch": 3.0697674418604652,
"grad_norm": 2.634533166885376,
"learning_rate": 2.27245831290033e-06,
"loss": 0.5259836912155151,
"step": 1452
},
{
"epoch": 3.0739957716701904,
"grad_norm": 1.2000439167022705,
"learning_rate": 2.261513932345755e-06,
"loss": 0.6875229477882385,
"step": 1454
},
{
"epoch": 3.0782241014799157,
"grad_norm": 0.343357652425766,
"learning_rate": 2.250609143584443e-06,
"loss": 0.17928436398506165,
"step": 1456
},
{
"epoch": 3.0824524312896404,
"grad_norm": 1.1093839406967163,
"learning_rate": 2.239744079931818e-06,
"loss": 0.5352936387062073,
"step": 1458
},
{
"epoch": 3.0866807610993656,
"grad_norm": 1.237159013748169,
"learning_rate": 2.228918874217649e-06,
"loss": 0.6347687244415283,
"step": 1460
},
{
"epoch": 3.090909090909091,
"grad_norm": 1.1333320140838623,
"learning_rate": 2.2181336587844244e-06,
"loss": 0.8107811212539673,
"step": 1462
},
{
"epoch": 3.095137420718816,
"grad_norm": 0.8327125310897827,
"learning_rate": 2.207388565485736e-06,
"loss": 0.47892993688583374,
"step": 1464
},
{
"epoch": 3.0993657505285412,
"grad_norm": 6.478003978729248,
"learning_rate": 2.196683725684666e-06,
"loss": 0.5957081913948059,
"step": 1466
},
{
"epoch": 3.1035940803382664,
"grad_norm": 1.1537619829177856,
"learning_rate": 2.1860192702521826e-06,
"loss": 0.7536590099334717,
"step": 1468
},
{
"epoch": 3.1078224101479917,
"grad_norm": 1.094152808189392,
"learning_rate": 2.1753953295655354e-06,
"loss": 0.8047918677330017,
"step": 1470
},
{
"epoch": 3.112050739957717,
"grad_norm": 1.9959810972213745,
"learning_rate": 2.164812033506673e-06,
"loss": 0.49085742235183716,
"step": 1472
},
{
"epoch": 3.116279069767442,
"grad_norm": 0.9419063925743103,
"learning_rate": 2.1542695114606367e-06,
"loss": 0.8454470634460449,
"step": 1474
},
{
"epoch": 3.120507399577167,
"grad_norm": 0.8833346366882324,
"learning_rate": 2.1437678923139944e-06,
"loss": 0.6828165054321289,
"step": 1476
},
{
"epoch": 3.124735729386892,
"grad_norm": 3.811008930206299,
"learning_rate": 2.1333073044532628e-06,
"loss": 0.23634839057922363,
"step": 1478
},
{
"epoch": 3.1289640591966172,
"grad_norm": 1.1801276206970215,
"learning_rate": 2.122887875763327e-06,
"loss": 0.74394291639328,
"step": 1480
},
{
"epoch": 3.1331923890063424,
"grad_norm": 0.9390035271644592,
"learning_rate": 2.1125097336258897e-06,
"loss": 0.7065632939338684,
"step": 1482
},
{
"epoch": 3.1374207188160677,
"grad_norm": 1.122446894645691,
"learning_rate": 2.1021730049179117e-06,
"loss": 0.2781147360801697,
"step": 1484
},
{
"epoch": 3.141649048625793,
"grad_norm": 2.646714687347412,
"learning_rate": 2.09187781601005e-06,
"loss": 0.6746383905410767,
"step": 1486
},
{
"epoch": 3.145877378435518,
"grad_norm": 0.6934086680412292,
"learning_rate": 2.0816242927651254e-06,
"loss": 0.8090722560882568,
"step": 1488
},
{
"epoch": 3.1501057082452433,
"grad_norm": 9.405001640319824,
"learning_rate": 2.071412560536581e-06,
"loss": 0.6543447971343994,
"step": 1490
},
{
"epoch": 3.1543340380549685,
"grad_norm": 3.9425649642944336,
"learning_rate": 2.0612427441669475e-06,
"loss": 0.5662192106246948,
"step": 1492
},
{
"epoch": 3.1585623678646932,
"grad_norm": 1.1375287771224976,
"learning_rate": 2.051114967986311e-06,
"loss": 0.38726744055747986,
"step": 1494
},
{
"epoch": 3.1627906976744184,
"grad_norm": 0.19144238531589508,
"learning_rate": 2.0410293558108074e-06,
"loss": 0.5012381076812744,
"step": 1496
},
{
"epoch": 3.1670190274841437,
"grad_norm": 0.841063916683197,
"learning_rate": 2.0309860309410986e-06,
"loss": 0.751567006111145,
"step": 1498
},
{
"epoch": 3.171247357293869,
"grad_norm": 0.9299562573432922,
"learning_rate": 2.0209851161608613e-06,
"loss": 0.8334060311317444,
"step": 1500
},
{
"epoch": 3.175475687103594,
"grad_norm": 3.2050392627716064,
"learning_rate": 2.011026733735301e-06,
"loss": 0.20530392229557037,
"step": 1502
},
{
"epoch": 3.1797040169133193,
"grad_norm": 3.570608615875244,
"learning_rate": 2.001111005409643e-06,
"loss": 0.4520590007305145,
"step": 1504
},
{
"epoch": 3.1839323467230445,
"grad_norm": 1.0268369913101196,
"learning_rate": 1.9912380524076503e-06,
"loss": 0.6378798484802246,
"step": 1506
},
{
"epoch": 3.1881606765327697,
"grad_norm": 2.796509027481079,
"learning_rate": 1.981407995430136e-06,
"loss": 0.3881588578224182,
"step": 1508
},
{
"epoch": 3.192389006342495,
"grad_norm": 12.317697525024414,
"learning_rate": 1.9716209546535e-06,
"loss": 0.2764635980129242,
"step": 1510
},
{
"epoch": 3.1966173361522197,
"grad_norm": 0.27015185356140137,
"learning_rate": 1.9618770497282487e-06,
"loss": 0.04972046613693237,
"step": 1512
},
{
"epoch": 3.200845665961945,
"grad_norm": 2.4296958446502686,
"learning_rate": 1.9521763997775294e-06,
"loss": 0.6549552083015442,
"step": 1514
},
{
"epoch": 3.20507399577167,
"grad_norm": 2.094703435897827,
"learning_rate": 1.942519123395689e-06,
"loss": 0.8088480234146118,
"step": 1516
},
{
"epoch": 3.2093023255813953,
"grad_norm": 1.2841097116470337,
"learning_rate": 1.932905338646812e-06,
"loss": 0.47878897190093994,
"step": 1518
},
{
"epoch": 3.2135306553911205,
"grad_norm": 0.7781525254249573,
"learning_rate": 1.9233351630632747e-06,
"loss": 0.5921374559402466,
"step": 1520
},
{
"epoch": 3.2177589852008457,
"grad_norm": 2.89263916015625,
"learning_rate": 1.9138087136443227e-06,
"loss": 0.6780604124069214,
"step": 1522
},
{
"epoch": 3.221987315010571,
"grad_norm": 3.3082852363586426,
"learning_rate": 1.904326106854626e-06,
"loss": 0.8326728940010071,
"step": 1524
},
{
"epoch": 3.226215644820296,
"grad_norm": 1.7145084142684937,
"learning_rate": 1.8948874586228622e-06,
"loss": 0.833214521408081,
"step": 1526
},
{
"epoch": 3.2304439746300213,
"grad_norm": 5.512892246246338,
"learning_rate": 1.8854928843402987e-06,
"loss": 0.11821231991052628,
"step": 1528
},
{
"epoch": 3.234672304439746,
"grad_norm": 1.682637095451355,
"learning_rate": 1.8761424988593796e-06,
"loss": 0.6847789287567139,
"step": 1530
},
{
"epoch": 3.2389006342494713,
"grad_norm": 2.431734561920166,
"learning_rate": 1.8668364164923248e-06,
"loss": 0.6516000032424927,
"step": 1532
},
{
"epoch": 3.2431289640591965,
"grad_norm": 1.9163728952407837,
"learning_rate": 1.8575747510097294e-06,
"loss": 0.7170204520225525,
"step": 1534
},
{
"epoch": 3.2473572938689217,
"grad_norm": 1.9188907146453857,
"learning_rate": 1.8483576156391747e-06,
"loss": 0.5550706386566162,
"step": 1536
},
{
"epoch": 3.251585623678647,
"grad_norm": 0.919139564037323,
"learning_rate": 1.8391851230638443e-06,
"loss": 0.5619450211524963,
"step": 1538
},
{
"epoch": 3.255813953488372,
"grad_norm": 1.158917784690857,
"learning_rate": 1.8300573854211452e-06,
"loss": 0.40066978335380554,
"step": 1540
},
{
"epoch": 3.2600422832980973,
"grad_norm": 3.2770321369171143,
"learning_rate": 1.8209745143013363e-06,
"loss": 0.9334087371826172,
"step": 1542
},
{
"epoch": 3.2642706131078225,
"grad_norm": 0.972515344619751,
"learning_rate": 1.811936620746167e-06,
"loss": 0.7332847118377686,
"step": 1544
},
{
"epoch": 3.2684989429175477,
"grad_norm": 1.4344654083251953,
"learning_rate": 1.8029438152475159e-06,
"loss": 0.658280611038208,
"step": 1546
},
{
"epoch": 3.2727272727272725,
"grad_norm": 1.1592069864273071,
"learning_rate": 1.7939962077460448e-06,
"loss": 0.1725965291261673,
"step": 1548
},
{
"epoch": 3.276955602536998,
"grad_norm": 2.889612913131714,
"learning_rate": 1.785093907629849e-06,
"loss": 0.7007735967636108,
"step": 1550
},
{
"epoch": 3.281183932346723,
"grad_norm": 1.1329373121261597,
"learning_rate": 1.7762370237331233e-06,
"loss": 0.7469728589057922,
"step": 1552
},
{
"epoch": 3.285412262156448,
"grad_norm": 0.0801200270652771,
"learning_rate": 1.7674256643348348e-06,
"loss": 0.07060103863477707,
"step": 1554
},
{
"epoch": 3.2896405919661733,
"grad_norm": 1.400202751159668,
"learning_rate": 1.7586599371573887e-06,
"loss": 0.7028331756591797,
"step": 1556
},
{
"epoch": 3.2938689217758985,
"grad_norm": 1.2305015325546265,
"learning_rate": 1.7499399493653225e-06,
"loss": 0.4038763642311096,
"step": 1558
},
{
"epoch": 3.2980972515856237,
"grad_norm": 3.4097230434417725,
"learning_rate": 1.7412658075639921e-06,
"loss": 0.8855617642402649,
"step": 1560
},
{
"epoch": 3.302325581395349,
"grad_norm": 13.10532283782959,
"learning_rate": 1.7326376177982635e-06,
"loss": 0.6156935691833496,
"step": 1562
},
{
"epoch": 3.306553911205074,
"grad_norm": 0.9922959208488464,
"learning_rate": 1.7240554855512225e-06,
"loss": 0.3985689878463745,
"step": 1564
},
{
"epoch": 3.3107822410147993,
"grad_norm": 1.160326600074768,
"learning_rate": 1.7155195157428878e-06,
"loss": 0.540247917175293,
"step": 1566
},
{
"epoch": 3.3150105708245245,
"grad_norm": 0.7250515818595886,
"learning_rate": 1.707029812728916e-06,
"loss": 0.10611023753881454,
"step": 1568
},
{
"epoch": 3.3192389006342493,
"grad_norm": 2.959808111190796,
"learning_rate": 1.6985864802993366e-06,
"loss": 0.43417036533355713,
"step": 1570
},
{
"epoch": 3.3234672304439745,
"grad_norm": 37.109107971191406,
"learning_rate": 1.690189621677284e-06,
"loss": 0.6207136511802673,
"step": 1572
},
{
"epoch": 3.3276955602536997,
"grad_norm": 1.7058624029159546,
"learning_rate": 1.681839339517728e-06,
"loss": 0.09287132322788239,
"step": 1574
},
{
"epoch": 3.331923890063425,
"grad_norm": 3.4163124561309814,
"learning_rate": 1.67353573590622e-06,
"loss": 0.3172765076160431,
"step": 1576
},
{
"epoch": 3.33615221987315,
"grad_norm": 1.0825945138931274,
"learning_rate": 1.6652789123576513e-06,
"loss": 0.4265225827693939,
"step": 1578
},
{
"epoch": 3.3403805496828753,
"grad_norm": 1.9592632055282593,
"learning_rate": 1.6570689698150094e-06,
"loss": 0.4097139537334442,
"step": 1580
},
{
"epoch": 3.3446088794926006,
"grad_norm": 0.8935739994049072,
"learning_rate": 1.6489060086481373e-06,
"loss": 0.9140397310256958,
"step": 1582
},
{
"epoch": 3.3488372093023258,
"grad_norm": 5.108261585235596,
"learning_rate": 1.6407901286525152e-06,
"loss": 0.6245816946029663,
"step": 1584
},
{
"epoch": 3.353065539112051,
"grad_norm": 2.5040066242218018,
"learning_rate": 1.6327214290480382e-06,
"loss": 0.5924360156059265,
"step": 1586
},
{
"epoch": 3.3572938689217757,
"grad_norm": 2.757362127304077,
"learning_rate": 1.6247000084778006e-06,
"loss": 0.7638341188430786,
"step": 1588
},
{
"epoch": 3.361522198731501,
"grad_norm": 1.3833210468292236,
"learning_rate": 1.6167259650068887e-06,
"loss": 0.21210473775863647,
"step": 1590
},
{
"epoch": 3.365750528541226,
"grad_norm": 1.145029067993164,
"learning_rate": 1.6087993961211904e-06,
"loss": 0.6189537644386292,
"step": 1592
},
{
"epoch": 3.3699788583509513,
"grad_norm": 1.1362985372543335,
"learning_rate": 1.6009203987261957e-06,
"loss": 0.33132678270339966,
"step": 1594
},
{
"epoch": 3.3742071881606766,
"grad_norm": 1.0586540699005127,
"learning_rate": 1.593089069145809e-06,
"loss": 0.7689797282218933,
"step": 1596
},
{
"epoch": 3.3784355179704018,
"grad_norm": 3.164792537689209,
"learning_rate": 1.5853055031211844e-06,
"loss": 0.4815041124820709,
"step": 1598
},
{
"epoch": 3.382663847780127,
"grad_norm": 1.0360368490219116,
"learning_rate": 1.577569795809544e-06,
"loss": 0.21347355842590332,
"step": 1600
},
{
"epoch": 3.386892177589852,
"grad_norm": 3.4981889724731445,
"learning_rate": 1.5698820417830137e-06,
"loss": 0.5095347166061401,
"step": 1602
},
{
"epoch": 3.3911205073995774,
"grad_norm": 1.2895543575286865,
"learning_rate": 1.5622423350274784e-06,
"loss": 0.7505019903182983,
"step": 1604
},
{
"epoch": 3.395348837209302,
"grad_norm": 1.477231502532959,
"learning_rate": 1.5546507689414217e-06,
"loss": 0.7805213332176208,
"step": 1606
},
{
"epoch": 3.3995771670190273,
"grad_norm": 2.5902810096740723,
"learning_rate": 1.5471074363347894e-06,
"loss": 0.4095946252346039,
"step": 1608
},
{
"epoch": 3.4038054968287526,
"grad_norm": 1.3147857189178467,
"learning_rate": 1.539612429427852e-06,
"loss": 0.7399739027023315,
"step": 1610
},
{
"epoch": 3.4080338266384778,
"grad_norm": 1.1041135787963867,
"learning_rate": 1.5321658398500808e-06,
"loss": 0.6881253719329834,
"step": 1612
},
{
"epoch": 3.412262156448203,
"grad_norm": 1.176977515220642,
"learning_rate": 1.5247677586390255e-06,
"loss": 0.5819146037101746,
"step": 1614
},
{
"epoch": 3.416490486257928,
"grad_norm": 5.668468475341797,
"learning_rate": 1.5174182762392007e-06,
"loss": 0.8557113409042358,
"step": 1616
},
{
"epoch": 3.4207188160676534,
"grad_norm": 1.155307650566101,
"learning_rate": 1.5101174825009815e-06,
"loss": 0.7553771138191223,
"step": 1618
},
{
"epoch": 3.4249471458773786,
"grad_norm": 1.0186285972595215,
"learning_rate": 1.5028654666795042e-06,
"loss": 0.7405768036842346,
"step": 1620
},
{
"epoch": 3.429175475687104,
"grad_norm": 1.8877136707305908,
"learning_rate": 1.495662317433576e-06,
"loss": 0.382840633392334,
"step": 1622
},
{
"epoch": 3.4334038054968286,
"grad_norm": 0.8843347430229187,
"learning_rate": 1.4885081228245909e-06,
"loss": 0.6595534086227417,
"step": 1624
},
{
"epoch": 3.4376321353065538,
"grad_norm": 1.4666378498077393,
"learning_rate": 1.4814029703154515e-06,
"loss": 0.4639345407485962,
"step": 1626
},
{
"epoch": 3.441860465116279,
"grad_norm": 1.579208254814148,
"learning_rate": 1.4743469467695017e-06,
"loss": 0.7690479755401611,
"step": 1628
},
{
"epoch": 3.446088794926004,
"grad_norm": 3.049574136734009,
"learning_rate": 1.4673401384494651e-06,
"loss": 0.44521260261535645,
"step": 1630
},
{
"epoch": 3.4503171247357294,
"grad_norm": 1.2153685092926025,
"learning_rate": 1.4603826310163874e-06,
"loss": 0.6739133596420288,
"step": 1632
},
{
"epoch": 3.4545454545454546,
"grad_norm": 1.456907033920288,
"learning_rate": 1.453474509528593e-06,
"loss": 0.4843824505805969,
"step": 1634
},
{
"epoch": 3.45877378435518,
"grad_norm": 1.242173671722412,
"learning_rate": 1.446615858440642e-06,
"loss": 0.7685821652412415,
"step": 1636
},
{
"epoch": 3.463002114164905,
"grad_norm": 0.45140162110328674,
"learning_rate": 1.4398067616022993e-06,
"loss": 0.32386642694473267,
"step": 1638
},
{
"epoch": 3.46723044397463,
"grad_norm": 1.1784615516662598,
"learning_rate": 1.43304730225751e-06,
"loss": 0.4159007966518402,
"step": 1640
},
{
"epoch": 3.471458773784355,
"grad_norm": 0.42984771728515625,
"learning_rate": 1.4263375630433802e-06,
"loss": 0.10018344223499298,
"step": 1642
},
{
"epoch": 3.47568710359408,
"grad_norm": 1.5950305461883545,
"learning_rate": 1.4196776259891665e-06,
"loss": 0.6593277454376221,
"step": 1644
},
{
"epoch": 3.4799154334038054,
"grad_norm": 1.384057879447937,
"learning_rate": 1.4130675725152762e-06,
"loss": 0.6950880289077759,
"step": 1646
},
{
"epoch": 3.4841437632135306,
"grad_norm": 0.5730490684509277,
"learning_rate": 1.4065074834322677e-06,
"loss": 0.4282079339027405,
"step": 1648
},
{
"epoch": 3.488372093023256,
"grad_norm": 1.2524633407592773,
"learning_rate": 1.3999974389398669e-06,
"loss": 0.4726506471633911,
"step": 1650
},
{
"epoch": 3.492600422832981,
"grad_norm": 2.6266963481903076,
"learning_rate": 1.3935375186259817e-06,
"loss": 0.46695253252983093,
"step": 1652
},
{
"epoch": 3.496828752642706,
"grad_norm": 3.294515609741211,
"learning_rate": 1.3871278014657332e-06,
"loss": 0.6511263847351074,
"step": 1654
},
{
"epoch": 3.5010570824524314,
"grad_norm": 2.2993924617767334,
"learning_rate": 1.3807683658204915e-06,
"loss": 0.7142579555511475,
"step": 1656
},
{
"epoch": 3.5052854122621566,
"grad_norm": 0.33005812764167786,
"learning_rate": 1.3744592894369093e-06,
"loss": 0.4682614803314209,
"step": 1658
},
{
"epoch": 3.5095137420718814,
"grad_norm": 1.0756093263626099,
"learning_rate": 1.3682006494459806e-06,
"loss": 0.46230700612068176,
"step": 1660
},
{
"epoch": 3.513742071881607,
"grad_norm": 2.6771621704101562,
"learning_rate": 1.361992522362096e-06,
"loss": 0.4901902675628662,
"step": 1662
},
{
"epoch": 3.517970401691332,
"grad_norm": 1.3595420122146606,
"learning_rate": 1.355834984082101e-06,
"loss": 0.7445568442344666,
"step": 1664
},
{
"epoch": 3.522198731501057,
"grad_norm": 1.7863283157348633,
"learning_rate": 1.3497281098843746e-06,
"loss": 0.7124844789505005,
"step": 1666
},
{
"epoch": 3.526427061310782,
"grad_norm": 1.0789310932159424,
"learning_rate": 1.34367197442791e-06,
"loss": 0.11922147870063782,
"step": 1668
},
{
"epoch": 3.5306553911205074,
"grad_norm": 0.4547461271286011,
"learning_rate": 1.3376666517513931e-06,
"loss": 0.5818626880645752,
"step": 1670
},
{
"epoch": 3.5348837209302326,
"grad_norm": 1.8149471282958984,
"learning_rate": 1.3317122152723078e-06,
"loss": 0.40404754877090454,
"step": 1672
},
{
"epoch": 3.539112050739958,
"grad_norm": 1.0777785778045654,
"learning_rate": 1.3258087377860323e-06,
"loss": 0.6738236546516418,
"step": 1674
},
{
"epoch": 3.543340380549683,
"grad_norm": 4.997625350952148,
"learning_rate": 1.3199562914649519e-06,
"loss": 0.40284815430641174,
"step": 1676
},
{
"epoch": 3.547568710359408,
"grad_norm": 1.375443696975708,
"learning_rate": 1.314154947857571e-06,
"loss": 0.6956469416618347,
"step": 1678
},
{
"epoch": 3.5517970401691334,
"grad_norm": 1.210405707359314,
"learning_rate": 1.3084047778876487e-06,
"loss": 0.4498561918735504,
"step": 1680
},
{
"epoch": 3.556025369978858,
"grad_norm": 1.4606341123580933,
"learning_rate": 1.3027058518533228e-06,
"loss": 0.5174427628517151,
"step": 1682
},
{
"epoch": 3.5602536997885834,
"grad_norm": 1.4805978536605835,
"learning_rate": 1.2970582394262516e-06,
"loss": 0.48671719431877136,
"step": 1684
},
{
"epoch": 3.5644820295983086,
"grad_norm": 4.511134147644043,
"learning_rate": 1.291462009650768e-06,
"loss": 0.11353273689746857,
"step": 1686
},
{
"epoch": 3.568710359408034,
"grad_norm": 1.7886998653411865,
"learning_rate": 1.2859172309430285e-06,
"loss": 0.5367018580436707,
"step": 1688
},
{
"epoch": 3.572938689217759,
"grad_norm": 0.24675719439983368,
"learning_rate": 1.280423971090181e-06,
"loss": 0.1590273529291153,
"step": 1690
},
{
"epoch": 3.5771670190274842,
"grad_norm": 2.029365301132202,
"learning_rate": 1.2749822972495333e-06,
"loss": 0.2917565405368805,
"step": 1692
},
{
"epoch": 3.5813953488372094,
"grad_norm": 2.3119266033172607,
"learning_rate": 1.2695922759477356e-06,
"loss": 0.473431795835495,
"step": 1694
},
{
"epoch": 3.585623678646934,
"grad_norm": 1.175053596496582,
"learning_rate": 1.2642539730799622e-06,
"loss": 0.2605957090854645,
"step": 1696
},
{
"epoch": 3.58985200845666,
"grad_norm": 1.1448858976364136,
"learning_rate": 1.2589674539091115e-06,
"loss": 0.7016085386276245,
"step": 1698
},
{
"epoch": 3.5940803382663846,
"grad_norm": 1.5570433139801025,
"learning_rate": 1.2537327830650042e-06,
"loss": 0.8868935704231262,
"step": 1700
},
{
"epoch": 3.59830866807611,
"grad_norm": 5.23875093460083,
"learning_rate": 1.248550024543594e-06,
"loss": 0.6247466206550598,
"step": 1702
},
{
"epoch": 3.602536997885835,
"grad_norm": 2.257566452026367,
"learning_rate": 1.2434192417061862e-06,
"loss": 0.31777775287628174,
"step": 1704
},
{
"epoch": 3.6067653276955602,
"grad_norm": 1.2846757173538208,
"learning_rate": 1.2383404972786625e-06,
"loss": 0.4019903540611267,
"step": 1706
},
{
"epoch": 3.6109936575052854,
"grad_norm": 2.253659963607788,
"learning_rate": 1.2333138533507139e-06,
"loss": 0.5987856984138489,
"step": 1708
},
{
"epoch": 3.6152219873150107,
"grad_norm": 4.348722457885742,
"learning_rate": 1.2283393713750816e-06,
"loss": 0.5229067802429199,
"step": 1710
},
{
"epoch": 3.619450317124736,
"grad_norm": 2.032782793045044,
"learning_rate": 1.2234171121668069e-06,
"loss": 0.46593305468559265,
"step": 1712
},
{
"epoch": 3.6236786469344606,
"grad_norm": 0.3673475682735443,
"learning_rate": 1.2185471359024855e-06,
"loss": 0.28558778762817383,
"step": 1714
},
{
"epoch": 3.6279069767441863,
"grad_norm": 4.834332466125488,
"learning_rate": 1.2137295021195341e-06,
"loss": 0.6039106845855713,
"step": 1716
},
{
"epoch": 3.632135306553911,
"grad_norm": 0.4229423403739929,
"learning_rate": 1.2089642697154613e-06,
"loss": 0.3675139844417572,
"step": 1718
},
{
"epoch": 3.6363636363636362,
"grad_norm": 2.2572038173675537,
"learning_rate": 1.2042514969471475e-06,
"loss": 0.3869643807411194,
"step": 1720
},
{
"epoch": 3.6405919661733614,
"grad_norm": 1.2497291564941406,
"learning_rate": 1.1995912414301329e-06,
"loss": 0.5978645086288452,
"step": 1722
},
{
"epoch": 3.6448202959830867,
"grad_norm": 1.2579014301300049,
"learning_rate": 1.1949835601379144e-06,
"loss": 0.6022518873214722,
"step": 1724
},
{
"epoch": 3.649048625792812,
"grad_norm": 1.0214322805404663,
"learning_rate": 1.1904285094012463e-06,
"loss": 0.7669554352760315,
"step": 1726
},
{
"epoch": 3.653276955602537,
"grad_norm": 3.1058895587921143,
"learning_rate": 1.185926144907454e-06,
"loss": 0.7332088351249695,
"step": 1728
},
{
"epoch": 3.6575052854122623,
"grad_norm": 12.841395378112793,
"learning_rate": 1.1814765216997522e-06,
"loss": 0.680499792098999,
"step": 1730
},
{
"epoch": 3.6617336152219875,
"grad_norm": 1.4355413913726807,
"learning_rate": 1.1770796941765727e-06,
"loss": 0.6887603998184204,
"step": 1732
},
{
"epoch": 3.6659619450317127,
"grad_norm": 1.9887467622756958,
"learning_rate": 1.1727357160908978e-06,
"loss": 0.6664218902587891,
"step": 1734
},
{
"epoch": 3.6701902748414374,
"grad_norm": 1.1173006296157837,
"learning_rate": 1.1684446405496052e-06,
"loss": 0.7440875172615051,
"step": 1736
},
{
"epoch": 3.6744186046511627,
"grad_norm": 3.4305248260498047,
"learning_rate": 1.1642065200128179e-06,
"loss": 0.19567635655403137,
"step": 1738
},
{
"epoch": 3.678646934460888,
"grad_norm": 1.9036072492599487,
"learning_rate": 1.1600214062932613e-06,
"loss": 0.41257956624031067,
"step": 1740
},
{
"epoch": 3.682875264270613,
"grad_norm": 1.285462737083435,
"learning_rate": 1.1558893505556315e-06,
"loss": 0.24330833554267883,
"step": 1742
},
{
"epoch": 3.6871035940803383,
"grad_norm": 1.1461507081985474,
"learning_rate": 1.1518104033159715e-06,
"loss": 0.6836899518966675,
"step": 1744
},
{
"epoch": 3.6913319238900635,
"grad_norm": 0.5551213622093201,
"learning_rate": 1.1477846144410493e-06,
"loss": 0.3795410990715027,
"step": 1746
},
{
"epoch": 3.6955602536997887,
"grad_norm": 4.269996643066406,
"learning_rate": 1.1438120331477506e-06,
"loss": 0.2996211051940918,
"step": 1748
},
{
"epoch": 3.699788583509514,
"grad_norm": 1.9732766151428223,
"learning_rate": 1.1398927080024796e-06,
"loss": 0.7808722257614136,
"step": 1750
},
{
"epoch": 3.704016913319239,
"grad_norm": 1.7274233102798462,
"learning_rate": 1.1360266869205593e-06,
"loss": 0.28832098841667175,
"step": 1752
},
{
"epoch": 3.708245243128964,
"grad_norm": 4.9157395362854,
"learning_rate": 1.1322140171656507e-06,
"loss": 0.5019751191139221,
"step": 1754
},
{
"epoch": 3.712473572938689,
"grad_norm": 1.1272751092910767,
"learning_rate": 1.1284547453491734e-06,
"loss": 0.6341027021408081,
"step": 1756
},
{
"epoch": 3.7167019027484143,
"grad_norm": 2.441051483154297,
"learning_rate": 1.1247489174297372e-06,
"loss": 0.39409759640693665,
"step": 1758
},
{
"epoch": 3.7209302325581395,
"grad_norm": 0.27103060483932495,
"learning_rate": 1.1210965787125743e-06,
"loss": 0.36354658007621765,
"step": 1760
},
{
"epoch": 3.7251585623678647,
"grad_norm": 2.6101622581481934,
"learning_rate": 1.1174977738489955e-06,
"loss": 0.6699774861335754,
"step": 1762
},
{
"epoch": 3.72938689217759,
"grad_norm": 1.1393935680389404,
"learning_rate": 1.113952546835835e-06,
"loss": 0.44287145137786865,
"step": 1764
},
{
"epoch": 3.733615221987315,
"grad_norm": 1.5996503829956055,
"learning_rate": 1.1104609410149174e-06,
"loss": 0.6402109265327454,
"step": 1766
},
{
"epoch": 3.7378435517970403,
"grad_norm": 1.5576012134552002,
"learning_rate": 1.1070229990725262e-06,
"loss": 0.6655474305152893,
"step": 1768
},
{
"epoch": 3.7420718816067655,
"grad_norm": 0.2630586326122284,
"learning_rate": 1.1036387630388841e-06,
"loss": 0.13032640516757965,
"step": 1770
},
{
"epoch": 3.7463002114164903,
"grad_norm": 1.4451303482055664,
"learning_rate": 1.100308274287636e-06,
"loss": 0.4177436828613281,
"step": 1772
},
{
"epoch": 3.7505285412262155,
"grad_norm": 3.356581926345825,
"learning_rate": 1.0970315735353434e-06,
"loss": 0.9017882347106934,
"step": 1774
},
{
"epoch": 3.7547568710359407,
"grad_norm": 1.294739007949829,
"learning_rate": 1.0938087008409909e-06,
"loss": 0.6469268202781677,
"step": 1776
},
{
"epoch": 3.758985200845666,
"grad_norm": 0.8014981150627136,
"learning_rate": 1.0906396956054919e-06,
"loss": 0.7925536036491394,
"step": 1778
},
{
"epoch": 3.763213530655391,
"grad_norm": 1.169784665107727,
"learning_rate": 1.0875245965712077e-06,
"loss": 0.7714580297470093,
"step": 1780
},
{
"epoch": 3.7674418604651163,
"grad_norm": 1.15401029586792,
"learning_rate": 1.0844634418214763e-06,
"loss": 0.7637167572975159,
"step": 1782
},
{
"epoch": 3.7716701902748415,
"grad_norm": 1.546501874923706,
"learning_rate": 1.0814562687801448e-06,
"loss": 0.1087595522403717,
"step": 1784
},
{
"epoch": 3.7758985200845667,
"grad_norm": 1.8229179382324219,
"learning_rate": 1.0785031142111103e-06,
"loss": 0.5062063932418823,
"step": 1786
},
{
"epoch": 3.780126849894292,
"grad_norm": 3.249974489212036,
"learning_rate": 1.075604014217875e-06,
"loss": 0.42154696583747864,
"step": 1788
},
{
"epoch": 3.7843551797040167,
"grad_norm": 3.055185556411743,
"learning_rate": 1.0727590042431009e-06,
"loss": 0.15182873606681824,
"step": 1790
},
{
"epoch": 3.7885835095137423,
"grad_norm": 0.2994537055492401,
"learning_rate": 1.0699681190681777e-06,
"loss": 0.572277307510376,
"step": 1792
},
{
"epoch": 3.792811839323467,
"grad_norm": 1.006197452545166,
"learning_rate": 1.067231392812798e-06,
"loss": 0.7577234506607056,
"step": 1794
},
{
"epoch": 3.7970401691331923,
"grad_norm": 21.956138610839844,
"learning_rate": 1.0645488589345396e-06,
"loss": 0.4559324383735657,
"step": 1796
},
{
"epoch": 3.8012684989429175,
"grad_norm": 3.4596710205078125,
"learning_rate": 1.0619205502284574e-06,
"loss": 0.7389811873435974,
"step": 1798
},
{
"epoch": 3.8054968287526427,
"grad_norm": 1.5893816947937012,
"learning_rate": 1.0593464988266811e-06,
"loss": 0.7755043506622314,
"step": 1800
},
{
"epoch": 3.809725158562368,
"grad_norm": 1.5835062265396118,
"learning_rate": 1.0568267361980225e-06,
"loss": 0.45009055733680725,
"step": 1802
},
{
"epoch": 3.813953488372093,
"grad_norm": 2.99670672416687,
"learning_rate": 1.0543612931475935e-06,
"loss": 0.649986743927002,
"step": 1804
},
{
"epoch": 3.8181818181818183,
"grad_norm": 1.015834093093872,
"learning_rate": 1.0519501998164247e-06,
"loss": 0.7480624914169312,
"step": 1806
},
{
"epoch": 3.822410147991543,
"grad_norm": 3.341177225112915,
"learning_rate": 1.049593485681101e-06,
"loss": 0.1723867803812027,
"step": 1808
},
{
"epoch": 3.8266384778012688,
"grad_norm": 1.422512412071228,
"learning_rate": 1.0472911795533997e-06,
"loss": 0.7097164988517761,
"step": 1810
},
{
"epoch": 3.8308668076109935,
"grad_norm": 2.261986255645752,
"learning_rate": 1.045043309579938e-06,
"loss": 0.8681549429893494,
"step": 1812
},
{
"epoch": 3.8350951374207187,
"grad_norm": 1.0599764585494995,
"learning_rate": 1.0428499032418297e-06,
"loss": 0.6682579517364502,
"step": 1814
},
{
"epoch": 3.839323467230444,
"grad_norm": 1.041394829750061,
"learning_rate": 1.0407109873543484e-06,
"loss": 0.5069891810417175,
"step": 1816
},
{
"epoch": 3.843551797040169,
"grad_norm": 10.299745559692383,
"learning_rate": 1.0386265880665996e-06,
"loss": 0.4282352328300476,
"step": 1818
},
{
"epoch": 3.8477801268498943,
"grad_norm": 1.0352288484573364,
"learning_rate": 1.036596730861203e-06,
"loss": 0.48183050751686096,
"step": 1820
},
{
"epoch": 3.8520084566596196,
"grad_norm": 0.59514319896698,
"learning_rate": 1.0346214405539783e-06,
"loss": 0.4891260862350464,
"step": 1822
},
{
"epoch": 3.8562367864693448,
"grad_norm": 0.9341387152671814,
"learning_rate": 1.0327007412936419e-06,
"loss": 0.4701923429965973,
"step": 1824
},
{
"epoch": 3.8604651162790695,
"grad_norm": 2.6931662559509277,
"learning_rate": 1.0308346565615158e-06,
"loss": 0.43026942014694214,
"step": 1826
},
{
"epoch": 3.864693446088795,
"grad_norm": 1.1196422576904297,
"learning_rate": 1.0290232091712337e-06,
"loss": 0.84291672706604,
"step": 1828
},
{
"epoch": 3.86892177589852,
"grad_norm": 1.5603212118148804,
"learning_rate": 1.0272664212684687e-06,
"loss": 0.8373754620552063,
"step": 1830
},
{
"epoch": 3.873150105708245,
"grad_norm": 3.0397791862487793,
"learning_rate": 1.025564314330659e-06,
"loss": 0.5141012668609619,
"step": 1832
},
{
"epoch": 3.8773784355179703,
"grad_norm": 1.2285468578338623,
"learning_rate": 1.0239169091667447e-06,
"loss": 0.41688698530197144,
"step": 1834
},
{
"epoch": 3.8816067653276956,
"grad_norm": 0.2671162784099579,
"learning_rate": 1.022324225916917e-06,
"loss": 0.32562455534935,
"step": 1836
},
{
"epoch": 3.8858350951374208,
"grad_norm": 0.14366553723812103,
"learning_rate": 1.0207862840523672e-06,
"loss": 0.4081572890281677,
"step": 1838
},
{
"epoch": 3.890063424947146,
"grad_norm": 1.101186752319336,
"learning_rate": 1.0193031023750534e-06,
"loss": 0.6767001748085022,
"step": 1840
},
{
"epoch": 3.894291754756871,
"grad_norm": 1.303435206413269,
"learning_rate": 1.017874699017468e-06,
"loss": 0.6841424107551575,
"step": 1842
},
{
"epoch": 3.898520084566596,
"grad_norm": 1.075760006904602,
"learning_rate": 1.0165010914424142e-06,
"loss": 0.46147653460502625,
"step": 1844
},
{
"epoch": 3.9027484143763216,
"grad_norm": 3.147409677505493,
"learning_rate": 1.015182296442798e-06,
"loss": 0.43763986229896545,
"step": 1846
},
{
"epoch": 3.9069767441860463,
"grad_norm": 3.417548418045044,
"learning_rate": 1.0139183301414176e-06,
"loss": 0.08393225073814392,
"step": 1848
},
{
"epoch": 3.9112050739957716,
"grad_norm": 1.169316291809082,
"learning_rate": 1.0127092079907682e-06,
"loss": 0.7433806657791138,
"step": 1850
},
{
"epoch": 3.9154334038054968,
"grad_norm": 0.559994637966156,
"learning_rate": 1.0115549447728551e-06,
"loss": 0.45681852102279663,
"step": 1852
},
{
"epoch": 3.919661733615222,
"grad_norm": 2.435063123703003,
"learning_rate": 1.0104555545990088e-06,
"loss": 0.6039175987243652,
"step": 1854
},
{
"epoch": 3.923890063424947,
"grad_norm": 1.3974052667617798,
"learning_rate": 1.009411050909716e-06,
"loss": 0.5208229422569275,
"step": 1856
},
{
"epoch": 3.9281183932346724,
"grad_norm": 4.605043888092041,
"learning_rate": 1.008421446474455e-06,
"loss": 0.5826939940452576,
"step": 1858
},
{
"epoch": 3.9323467230443976,
"grad_norm": 4.9776692390441895,
"learning_rate": 1.0074867533915367e-06,
"loss": 0.3075219988822937,
"step": 1860
},
{
"epoch": 3.9365750528541223,
"grad_norm": 3.256028413772583,
"learning_rate": 1.0066069830879593e-06,
"loss": 0.7073742151260376,
"step": 1862
},
{
"epoch": 3.940803382663848,
"grad_norm": 2.045060157775879,
"learning_rate": 1.0057821463192687e-06,
"loss": 0.6516786813735962,
"step": 1864
},
{
"epoch": 3.9450317124735728,
"grad_norm": 1.2108416557312012,
"learning_rate": 1.005012253169426e-06,
"loss": 0.7207849621772766,
"step": 1866
},
{
"epoch": 3.949260042283298,
"grad_norm": 0.2408219277858734,
"learning_rate": 1.0042973130506839e-06,
"loss": 0.01326755527406931,
"step": 1868
},
{
"epoch": 3.953488372093023,
"grad_norm": 1.5168226957321167,
"learning_rate": 1.003637334703472e-06,
"loss": 0.7974779009819031,
"step": 1870
},
{
"epoch": 3.9577167019027484,
"grad_norm": 1.1934226751327515,
"learning_rate": 1.0030323261962922e-06,
"loss": 0.5943294167518616,
"step": 1872
},
{
"epoch": 3.9619450317124736,
"grad_norm": 2.1566379070281982,
"learning_rate": 1.002482294925615e-06,
"loss": 0.5431046485900879,
"step": 1874
},
{
"epoch": 3.966173361522199,
"grad_norm": 1.2434756755828857,
"learning_rate": 1.0019872476157952e-06,
"loss": 0.7903445959091187,
"step": 1876
},
{
"epoch": 3.970401691331924,
"grad_norm": 7.063718318939209,
"learning_rate": 1.0015471903189834e-06,
"loss": 0.28139787912368774,
"step": 1878
},
{
"epoch": 3.974630021141649,
"grad_norm": 1.1494590044021606,
"learning_rate": 1.0011621284150583e-06,
"loss": 0.4004431962966919,
"step": 1880
},
{
"epoch": 3.9788583509513744,
"grad_norm": 2.593742609024048,
"learning_rate": 1.000832066611556e-06,
"loss": 0.5509825944900513,
"step": 1882
},
{
"epoch": 3.983086680761099,
"grad_norm": 3.3344368934631348,
"learning_rate": 1.0005570089436137e-06,
"loss": 0.8007972836494446,
"step": 1884
},
{
"epoch": 3.9873150105708244,
"grad_norm": 3.233119487762451,
"learning_rate": 1.0003369587739234e-06,
"loss": 0.5308348536491394,
"step": 1886
},
{
"epoch": 3.9915433403805496,
"grad_norm": 1.544953465461731,
"learning_rate": 1.0001719187926854e-06,
"loss": 0.7361894249916077,
"step": 1888
},
{
"epoch": 3.995771670190275,
"grad_norm": 0.894606351852417,
"learning_rate": 1.0000618910175811e-06,
"loss": 0.7119108438491821,
"step": 1890
},
{
"epoch": 4.0,
"grad_norm": 1.9086326360702515,
"learning_rate": 1.000006876793743e-06,
"loss": 0.3711016774177551,
"step": 1892
},
{
"epoch": 4.0,
"step": 1892,
"total_flos": 3.5845266747435254e+18,
"train_loss": 0.8291907650812373,
"train_runtime": 8889.4262,
"train_samples_per_second": 6.385,
"train_steps_per_second": 0.213
}
],
"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": 3.5845266747435254e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}