9b-98 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
628066d 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": 2.775197744369507,
"learning_rate": 1.0526315789473685e-07,
"loss": 1.8237788677215576,
"step": 2
},
{
"epoch": 0.008456659619450317,
"grad_norm": 1.4798002243041992,
"learning_rate": 3.1578947368421055e-07,
"loss": 2.2126944065093994,
"step": 4
},
{
"epoch": 0.012684989429175475,
"grad_norm": 6.348167896270752,
"learning_rate": 5.263157894736843e-07,
"loss": 2.2040843963623047,
"step": 6
},
{
"epoch": 0.016913319238900635,
"grad_norm": 1.6810775995254517,
"learning_rate": 7.368421052631579e-07,
"loss": 1.9967679977416992,
"step": 8
},
{
"epoch": 0.021141649048625793,
"grad_norm": 8.419306755065918,
"learning_rate": 9.473684210526317e-07,
"loss": 2.010589838027954,
"step": 10
},
{
"epoch": 0.02536997885835095,
"grad_norm": 12.916945457458496,
"learning_rate": 1.1578947368421053e-06,
"loss": 1.634893536567688,
"step": 12
},
{
"epoch": 0.02959830866807611,
"grad_norm": 4.631015777587891,
"learning_rate": 1.3684210526315791e-06,
"loss": 1.6433416604995728,
"step": 14
},
{
"epoch": 0.03382663847780127,
"grad_norm": 1.9718029499053955,
"learning_rate": 1.5789473684210526e-06,
"loss": 1.7679979801177979,
"step": 16
},
{
"epoch": 0.03805496828752643,
"grad_norm": 1.3308587074279785,
"learning_rate": 1.7894736842105265e-06,
"loss": 1.7944492101669312,
"step": 18
},
{
"epoch": 0.042283298097251586,
"grad_norm": 3.826141834259033,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.6710608005523682,
"step": 20
},
{
"epoch": 0.046511627906976744,
"grad_norm": 5.539140224456787,
"learning_rate": 2.2105263157894738e-06,
"loss": 1.5058794021606445,
"step": 22
},
{
"epoch": 0.0507399577167019,
"grad_norm": 2.2339963912963867,
"learning_rate": 2.4210526315789477e-06,
"loss": 1.774712324142456,
"step": 24
},
{
"epoch": 0.05496828752642706,
"grad_norm": 2.276745319366455,
"learning_rate": 2.631578947368421e-06,
"loss": 1.3508614301681519,
"step": 26
},
{
"epoch": 0.05919661733615222,
"grad_norm": 0.9598667025566101,
"learning_rate": 2.842105263157895e-06,
"loss": 1.6527801752090454,
"step": 28
},
{
"epoch": 0.06342494714587738,
"grad_norm": 1.8563207387924194,
"learning_rate": 3.052631578947369e-06,
"loss": 1.2124661207199097,
"step": 30
},
{
"epoch": 0.06765327695560254,
"grad_norm": 3.4228968620300293,
"learning_rate": 3.2631578947368423e-06,
"loss": 0.9826474785804749,
"step": 32
},
{
"epoch": 0.07188160676532769,
"grad_norm": 1.2923866510391235,
"learning_rate": 3.473684210526316e-06,
"loss": 1.1629663705825806,
"step": 34
},
{
"epoch": 0.07610993657505286,
"grad_norm": 1.7601325511932373,
"learning_rate": 3.6842105263157896e-06,
"loss": 1.352150797843933,
"step": 36
},
{
"epoch": 0.080338266384778,
"grad_norm": 12.143952369689941,
"learning_rate": 3.894736842105263e-06,
"loss": 1.2857341766357422,
"step": 38
},
{
"epoch": 0.08456659619450317,
"grad_norm": 3.5520753860473633,
"learning_rate": 4.105263157894737e-06,
"loss": 1.208201289176941,
"step": 40
},
{
"epoch": 0.08879492600422834,
"grad_norm": 3.265651226043701,
"learning_rate": 4.315789473684211e-06,
"loss": 1.697877049446106,
"step": 42
},
{
"epoch": 0.09302325581395349,
"grad_norm": 2.618696451187134,
"learning_rate": 4.526315789473685e-06,
"loss": 1.5967975854873657,
"step": 44
},
{
"epoch": 0.09725158562367865,
"grad_norm": 1.083250641822815,
"learning_rate": 4.736842105263158e-06,
"loss": 1.266770601272583,
"step": 46
},
{
"epoch": 0.1014799154334038,
"grad_norm": 0.9137223958969116,
"learning_rate": 4.947368421052632e-06,
"loss": 1.5009700059890747,
"step": 48
},
{
"epoch": 0.10570824524312897,
"grad_norm": 2.8655412197113037,
"learning_rate": 5.157894736842106e-06,
"loss": 1.4657362699508667,
"step": 50
},
{
"epoch": 0.10993657505285412,
"grad_norm": 2.6313316822052,
"learning_rate": 5.36842105263158e-06,
"loss": 0.7068197131156921,
"step": 52
},
{
"epoch": 0.11416490486257928,
"grad_norm": 2.766634702682495,
"learning_rate": 5.578947368421052e-06,
"loss": 1.0510985851287842,
"step": 54
},
{
"epoch": 0.11839323467230443,
"grad_norm": 1.289494276046753,
"learning_rate": 5.789473684210527e-06,
"loss": 1.3679696321487427,
"step": 56
},
{
"epoch": 0.1226215644820296,
"grad_norm": 1.4662611484527588,
"learning_rate": 6e-06,
"loss": 1.387304663658142,
"step": 58
},
{
"epoch": 0.12684989429175475,
"grad_norm": 1.6378839015960693,
"learning_rate": 6.2105263157894745e-06,
"loss": 1.3992042541503906,
"step": 60
},
{
"epoch": 0.13107822410147993,
"grad_norm": 2.2634947299957275,
"learning_rate": 6.421052631578948e-06,
"loss": 0.77618408203125,
"step": 62
},
{
"epoch": 0.13530655391120508,
"grad_norm": 1.5557911396026611,
"learning_rate": 6.631578947368421e-06,
"loss": 1.0436253547668457,
"step": 64
},
{
"epoch": 0.13953488372093023,
"grad_norm": 0.8463269472122192,
"learning_rate": 6.842105263157896e-06,
"loss": 1.3485163450241089,
"step": 66
},
{
"epoch": 0.14376321353065538,
"grad_norm": 0.8468639254570007,
"learning_rate": 7.052631578947369e-06,
"loss": 1.3779118061065674,
"step": 68
},
{
"epoch": 0.14799154334038056,
"grad_norm": 1.1941685676574707,
"learning_rate": 7.263157894736843e-06,
"loss": 1.3528388738632202,
"step": 70
},
{
"epoch": 0.1522198731501057,
"grad_norm": 2.4732611179351807,
"learning_rate": 7.473684210526316e-06,
"loss": 1.3456642627716064,
"step": 72
},
{
"epoch": 0.15644820295983086,
"grad_norm": 0.8812842965126038,
"learning_rate": 7.68421052631579e-06,
"loss": 0.9433136582374573,
"step": 74
},
{
"epoch": 0.160676532769556,
"grad_norm": 1.154260277748108,
"learning_rate": 7.894736842105265e-06,
"loss": 0.8495628237724304,
"step": 76
},
{
"epoch": 0.1649048625792812,
"grad_norm": 2.192192792892456,
"learning_rate": 8.105263157894736e-06,
"loss": 1.2864881753921509,
"step": 78
},
{
"epoch": 0.16913319238900634,
"grad_norm": 2.8072850704193115,
"learning_rate": 8.315789473684212e-06,
"loss": 1.0948576927185059,
"step": 80
},
{
"epoch": 0.1733615221987315,
"grad_norm": 1.465378999710083,
"learning_rate": 8.526315789473685e-06,
"loss": 1.0818125009536743,
"step": 82
},
{
"epoch": 0.17758985200845667,
"grad_norm": 0.7715460062026978,
"learning_rate": 8.736842105263158e-06,
"loss": 0.8977410793304443,
"step": 84
},
{
"epoch": 0.18181818181818182,
"grad_norm": 1.8554987907409668,
"learning_rate": 8.947368421052632e-06,
"loss": 1.7141196727752686,
"step": 86
},
{
"epoch": 0.18604651162790697,
"grad_norm": 0.9680240750312805,
"learning_rate": 9.157894736842105e-06,
"loss": 1.331464171409607,
"step": 88
},
{
"epoch": 0.19027484143763213,
"grad_norm": 2.164674758911133,
"learning_rate": 9.36842105263158e-06,
"loss": 0.9962979555130005,
"step": 90
},
{
"epoch": 0.1945031712473573,
"grad_norm": 0.9289390444755554,
"learning_rate": 9.578947368421054e-06,
"loss": 1.3149358034133911,
"step": 92
},
{
"epoch": 0.19873150105708245,
"grad_norm": 0.8665067553520203,
"learning_rate": 9.789473684210527e-06,
"loss": 0.9918832778930664,
"step": 94
},
{
"epoch": 0.2029598308668076,
"grad_norm": 1.5449823141098022,
"learning_rate": 1e-05,
"loss": 1.2967091798782349,
"step": 96
},
{
"epoch": 0.20718816067653276,
"grad_norm": 1.5481503009796143,
"learning_rate": 9.999972492846047e-06,
"loss": 1.3855273723602295,
"step": 98
},
{
"epoch": 0.21141649048625794,
"grad_norm": 0.8566465377807617,
"learning_rate": 9.999889971720468e-06,
"loss": 1.3273911476135254,
"step": 100
},
{
"epoch": 0.2156448202959831,
"grad_norm": 0.6091927886009216,
"learning_rate": 9.999752437632121e-06,
"loss": 1.2459510564804077,
"step": 102
},
{
"epoch": 0.21987315010570824,
"grad_norm": 1.5742939710617065,
"learning_rate": 9.999559892262412e-06,
"loss": 1.3078598976135254,
"step": 104
},
{
"epoch": 0.22410147991543342,
"grad_norm": 1.2861239910125732,
"learning_rate": 9.999312337965291e-06,
"loss": 1.2986016273498535,
"step": 106
},
{
"epoch": 0.22832980972515857,
"grad_norm": 0.9632868766784668,
"learning_rate": 9.999009777767202e-06,
"loss": 1.038223147392273,
"step": 108
},
{
"epoch": 0.23255813953488372,
"grad_norm": 1.206147313117981,
"learning_rate": 9.99865221536707e-06,
"loss": 1.2848056554794312,
"step": 110
},
{
"epoch": 0.23678646934460887,
"grad_norm": 4.335933208465576,
"learning_rate": 9.998239655136237e-06,
"loss": 1.3580799102783203,
"step": 112
},
{
"epoch": 0.24101479915433405,
"grad_norm": 0.9401216506958008,
"learning_rate": 9.997772102118417e-06,
"loss": 1.5649545192718506,
"step": 114
},
{
"epoch": 0.2452431289640592,
"grad_norm": 1.428720474243164,
"learning_rate": 9.997249562029637e-06,
"loss": 1.3351342678070068,
"step": 116
},
{
"epoch": 0.24947145877378435,
"grad_norm": 1.440656065940857,
"learning_rate": 9.996672041258154e-06,
"loss": 1.341488242149353,
"step": 118
},
{
"epoch": 0.2536997885835095,
"grad_norm": 1.2661004066467285,
"learning_rate": 9.996039546864395e-06,
"loss": 0.843405544757843,
"step": 120
},
{
"epoch": 0.25792811839323465,
"grad_norm": 1.0616695880889893,
"learning_rate": 9.995352086580858e-06,
"loss": 1.3061366081237793,
"step": 122
},
{
"epoch": 0.26215644820295986,
"grad_norm": 0.8607848882675171,
"learning_rate": 9.994609668812022e-06,
"loss": 1.1474379301071167,
"step": 124
},
{
"epoch": 0.266384778012685,
"grad_norm": 3.890989065170288,
"learning_rate": 9.993812302634237e-06,
"loss": 1.3460017442703247,
"step": 126
},
{
"epoch": 0.27061310782241016,
"grad_norm": 0.9614801406860352,
"learning_rate": 9.99295999779563e-06,
"loss": 0.9562265276908875,
"step": 128
},
{
"epoch": 0.2748414376321353,
"grad_norm": 1.4079248905181885,
"learning_rate": 9.99205276471597e-06,
"loss": 0.9662185311317444,
"step": 130
},
{
"epoch": 0.27906976744186046,
"grad_norm": 1.5130228996276855,
"learning_rate": 9.991090614486546e-06,
"loss": 1.2819286584854126,
"step": 132
},
{
"epoch": 0.2832980972515856,
"grad_norm": 12.365251541137695,
"learning_rate": 9.990073558870028e-06,
"loss": 0.8340978622436523,
"step": 134
},
{
"epoch": 0.28752642706131076,
"grad_norm": 4.939059734344482,
"learning_rate": 9.989001610300333e-06,
"loss": 1.1327745914459229,
"step": 136
},
{
"epoch": 0.2917547568710359,
"grad_norm": 2.3762025833129883,
"learning_rate": 9.987874781882461e-06,
"loss": 1.3777695894241333,
"step": 138
},
{
"epoch": 0.2959830866807611,
"grad_norm": 1.5781553983688354,
"learning_rate": 9.986693087392341e-06,
"loss": 0.9029810428619385,
"step": 140
},
{
"epoch": 0.30021141649048627,
"grad_norm": 0.9379784464836121,
"learning_rate": 9.985456541276667e-06,
"loss": 1.5644371509552002,
"step": 142
},
{
"epoch": 0.3044397463002114,
"grad_norm": 3.3895201683044434,
"learning_rate": 9.984165158652708e-06,
"loss": 1.1003453731536865,
"step": 144
},
{
"epoch": 0.3086680761099366,
"grad_norm": 1.8978742361068726,
"learning_rate": 9.982818955308135e-06,
"loss": 0.688273549079895,
"step": 146
},
{
"epoch": 0.3128964059196617,
"grad_norm": 1.6199144124984741,
"learning_rate": 9.98141794770083e-06,
"loss": 1.2452049255371094,
"step": 148
},
{
"epoch": 0.3171247357293869,
"grad_norm": 1.4302568435668945,
"learning_rate": 9.979962152958669e-06,
"loss": 1.2656316757202148,
"step": 150
},
{
"epoch": 0.321353065539112,
"grad_norm": 0.9690020084381104,
"learning_rate": 9.97845158887933e-06,
"loss": 1.1776196956634521,
"step": 152
},
{
"epoch": 0.32558139534883723,
"grad_norm": 3.208920955657959,
"learning_rate": 9.976886273930063e-06,
"loss": 0.7534308433532715,
"step": 154
},
{
"epoch": 0.3298097251585624,
"grad_norm": 1.0861670970916748,
"learning_rate": 9.975266227247474e-06,
"loss": 1.2519267797470093,
"step": 156
},
{
"epoch": 0.33403805496828753,
"grad_norm": 1.1383721828460693,
"learning_rate": 9.973591468637286e-06,
"loss": 1.2700384855270386,
"step": 158
},
{
"epoch": 0.3382663847780127,
"grad_norm": 1.0125421285629272,
"learning_rate": 9.971862018574094e-06,
"loss": 1.2574166059494019,
"step": 160
},
{
"epoch": 0.34249471458773784,
"grad_norm": 3.1870298385620117,
"learning_rate": 9.97007789820112e-06,
"loss": 1.0946229696273804,
"step": 162
},
{
"epoch": 0.346723044397463,
"grad_norm": 1.6475019454956055,
"learning_rate": 9.968239129329953e-06,
"loss": 0.9601567983627319,
"step": 164
},
{
"epoch": 0.35095137420718814,
"grad_norm": 1.3436022996902466,
"learning_rate": 9.96634573444028e-06,
"loss": 0.8783005475997925,
"step": 166
},
{
"epoch": 0.35517970401691334,
"grad_norm": 1.2205567359924316,
"learning_rate": 9.964397736679615e-06,
"loss": 1.2077776193618774,
"step": 168
},
{
"epoch": 0.3594080338266385,
"grad_norm": 0.8040755987167358,
"learning_rate": 9.962395159863014e-06,
"loss": 1.2072582244873047,
"step": 170
},
{
"epoch": 0.36363636363636365,
"grad_norm": 1.1201058626174927,
"learning_rate": 9.960338028472781e-06,
"loss": 1.0414552688598633,
"step": 172
},
{
"epoch": 0.3678646934460888,
"grad_norm": 1.1869196891784668,
"learning_rate": 9.958226367658176e-06,
"loss": 1.3537276983261108,
"step": 174
},
{
"epoch": 0.37209302325581395,
"grad_norm": 1.296112060546875,
"learning_rate": 9.9560602032351e-06,
"loss": 1.044026255607605,
"step": 176
},
{
"epoch": 0.3763213530655391,
"grad_norm": 3.114501714706421,
"learning_rate": 9.953839561685783e-06,
"loss": 0.9390140175819397,
"step": 178
},
{
"epoch": 0.38054968287526425,
"grad_norm": 0.7197378277778625,
"learning_rate": 9.95156447015846e-06,
"loss": 1.2823635339736938,
"step": 180
},
{
"epoch": 0.38477801268498946,
"grad_norm": 0.753393828868866,
"learning_rate": 9.94923495646704e-06,
"loss": 1.531618595123291,
"step": 182
},
{
"epoch": 0.3890063424947146,
"grad_norm": 0.8798313736915588,
"learning_rate": 9.946851049090764e-06,
"loss": 1.3364548683166504,
"step": 184
},
{
"epoch": 0.39323467230443976,
"grad_norm": 1.4332983493804932,
"learning_rate": 9.944412777173854e-06,
"loss": 1.1219594478607178,
"step": 186
},
{
"epoch": 0.3974630021141649,
"grad_norm": 0.7825232148170471,
"learning_rate": 9.941920170525168e-06,
"loss": 0.9862596392631531,
"step": 188
},
{
"epoch": 0.40169133192389006,
"grad_norm": 2.6555864810943604,
"learning_rate": 9.939373259617822e-06,
"loss": 1.0160772800445557,
"step": 190
},
{
"epoch": 0.4059196617336152,
"grad_norm": 1.148439645767212,
"learning_rate": 9.936772075588825e-06,
"loss": 1.2915797233581543,
"step": 192
},
{
"epoch": 0.41014799154334036,
"grad_norm": 0.5568650960922241,
"learning_rate": 9.934116650238697e-06,
"loss": 0.9997367858886719,
"step": 194
},
{
"epoch": 0.4143763213530655,
"grad_norm": 1.3544172048568726,
"learning_rate": 9.931407016031078e-06,
"loss": 1.2790859937667847,
"step": 196
},
{
"epoch": 0.4186046511627907,
"grad_norm": 1.345781922340393,
"learning_rate": 9.928643206092338e-06,
"loss": 1.075522780418396,
"step": 198
},
{
"epoch": 0.42283298097251587,
"grad_norm": 0.9323049783706665,
"learning_rate": 9.925825254211161e-06,
"loss": 1.1856673955917358,
"step": 200
},
{
"epoch": 0.427061310782241,
"grad_norm": 0.6949234008789062,
"learning_rate": 9.922953194838143e-06,
"loss": 0.5851123332977295,
"step": 202
},
{
"epoch": 0.4312896405919662,
"grad_norm": 0.8129012584686279,
"learning_rate": 9.920027063085363e-06,
"loss": 1.2291486263275146,
"step": 204
},
{
"epoch": 0.4355179704016913,
"grad_norm": 2.274306058883667,
"learning_rate": 9.917046894725957e-06,
"loss": 1.312781810760498,
"step": 206
},
{
"epoch": 0.4397463002114165,
"grad_norm": 2.1606812477111816,
"learning_rate": 9.91401272619368e-06,
"loss": 0.7842003703117371,
"step": 208
},
{
"epoch": 0.4439746300211416,
"grad_norm": 1.6500720977783203,
"learning_rate": 9.910924594582463e-06,
"loss": 1.285497784614563,
"step": 210
},
{
"epoch": 0.44820295983086683,
"grad_norm": 1.3784054517745972,
"learning_rate": 9.907782537645955e-06,
"loss": 1.2085553407669067,
"step": 212
},
{
"epoch": 0.452431289640592,
"grad_norm": 1.40472412109375,
"learning_rate": 9.904586593797062e-06,
"loss": 1.4150958061218262,
"step": 214
},
{
"epoch": 0.45665961945031713,
"grad_norm": 1.685820460319519,
"learning_rate": 9.901336802107485e-06,
"loss": 0.6261395215988159,
"step": 216
},
{
"epoch": 0.4608879492600423,
"grad_norm": 1.414623498916626,
"learning_rate": 9.89803320230723e-06,
"loss": 0.9964258074760437,
"step": 218
},
{
"epoch": 0.46511627906976744,
"grad_norm": 1.4940500259399414,
"learning_rate": 9.894675834784136e-06,
"loss": 1.265001893043518,
"step": 220
},
{
"epoch": 0.4693446088794926,
"grad_norm": 1.1460466384887695,
"learning_rate": 9.891264740583367e-06,
"loss": 1.2316094636917114,
"step": 222
},
{
"epoch": 0.47357293868921774,
"grad_norm": 1.180639386177063,
"learning_rate": 9.88779996140692e-06,
"loss": 1.3445286750793457,
"step": 224
},
{
"epoch": 0.47780126849894294,
"grad_norm": 1.6520229578018188,
"learning_rate": 9.884281539613109e-06,
"loss": 1.245647668838501,
"step": 226
},
{
"epoch": 0.4820295983086681,
"grad_norm": 0.9897115230560303,
"learning_rate": 9.880709518216061e-06,
"loss": 1.23969304561615,
"step": 228
},
{
"epoch": 0.48625792811839325,
"grad_norm": 0.8484343886375427,
"learning_rate": 9.87708394088517e-06,
"loss": 0.9211059808731079,
"step": 230
},
{
"epoch": 0.4904862579281184,
"grad_norm": 1.482115387916565,
"learning_rate": 9.873404851944573e-06,
"loss": 1.2463995218276978,
"step": 232
},
{
"epoch": 0.49471458773784355,
"grad_norm": 13.69424057006836,
"learning_rate": 9.869672296372611e-06,
"loss": 0.9416562914848328,
"step": 234
},
{
"epoch": 0.4989429175475687,
"grad_norm": 2.960153818130493,
"learning_rate": 9.86588631980128e-06,
"loss": 0.8898116946220398,
"step": 236
},
{
"epoch": 0.5031712473572939,
"grad_norm": 15.054459571838379,
"learning_rate": 9.86204696851566e-06,
"loss": 0.8407182693481445,
"step": 238
},
{
"epoch": 0.507399577167019,
"grad_norm": 1.030059576034546,
"learning_rate": 9.858154289453363e-06,
"loss": 0.9114965200424194,
"step": 240
},
{
"epoch": 0.5116279069767442,
"grad_norm": 1.0919334888458252,
"learning_rate": 9.854208330203958e-06,
"loss": 1.3099021911621094,
"step": 242
},
{
"epoch": 0.5158562367864693,
"grad_norm": 0.9385789632797241,
"learning_rate": 9.850209139008377e-06,
"loss": 1.3302651643753052,
"step": 244
},
{
"epoch": 0.5200845665961945,
"grad_norm": 21.08572006225586,
"learning_rate": 9.846156764758344e-06,
"loss": 1.2784703969955444,
"step": 246
},
{
"epoch": 0.5243128964059197,
"grad_norm": 0.9877079725265503,
"learning_rate": 9.842051256995759e-06,
"loss": 1.2360607385635376,
"step": 248
},
{
"epoch": 0.5285412262156448,
"grad_norm": 0.7909377217292786,
"learning_rate": 9.837892665912105e-06,
"loss": 1.1820871829986572,
"step": 250
},
{
"epoch": 0.53276955602537,
"grad_norm": 3.7115564346313477,
"learning_rate": 9.833681042347827e-06,
"loss": 1.1672182083129883,
"step": 252
},
{
"epoch": 0.5369978858350951,
"grad_norm": 1.4700535535812378,
"learning_rate": 9.829416437791721e-06,
"loss": 1.736332654953003,
"step": 254
},
{
"epoch": 0.5412262156448203,
"grad_norm": 1.021111011505127,
"learning_rate": 9.825098904380284e-06,
"loss": 1.2121504545211792,
"step": 256
},
{
"epoch": 0.5454545454545454,
"grad_norm": 2.407238006591797,
"learning_rate": 9.820728494897101e-06,
"loss": 1.3981647491455078,
"step": 258
},
{
"epoch": 0.5496828752642706,
"grad_norm": 2.120471477508545,
"learning_rate": 9.816305262772182e-06,
"loss": 0.9960198998451233,
"step": 260
},
{
"epoch": 0.5539112050739958,
"grad_norm": 2.1179914474487305,
"learning_rate": 9.811829262081315e-06,
"loss": 1.1999456882476807,
"step": 262
},
{
"epoch": 0.5581395348837209,
"grad_norm": 4.078908920288086,
"learning_rate": 9.807300547545411e-06,
"loss": 1.051332950592041,
"step": 264
},
{
"epoch": 0.5623678646934461,
"grad_norm": 1.0971733331680298,
"learning_rate": 9.802719174529821e-06,
"loss": 1.4946622848510742,
"step": 266
},
{
"epoch": 0.5665961945031712,
"grad_norm": 2.7961134910583496,
"learning_rate": 9.798085199043672e-06,
"loss": 0.5385439991950989,
"step": 268
},
{
"epoch": 0.5708245243128964,
"grad_norm": 3.0393545627593994,
"learning_rate": 9.793398677739177e-06,
"loss": 0.9961667060852051,
"step": 270
},
{
"epoch": 0.5750528541226215,
"grad_norm": 2.006559371948242,
"learning_rate": 9.788659667910942e-06,
"loss": 1.1786088943481445,
"step": 272
},
{
"epoch": 0.5792811839323467,
"grad_norm": 0.9761899709701538,
"learning_rate": 9.783868227495262e-06,
"loss": 0.9596310257911682,
"step": 274
},
{
"epoch": 0.5835095137420718,
"grad_norm": 1.074391484260559,
"learning_rate": 9.779024415069427e-06,
"loss": 0.9489659070968628,
"step": 276
},
{
"epoch": 0.587737843551797,
"grad_norm": 0.47903573513031006,
"learning_rate": 9.774128289850987e-06,
"loss": 0.8689467310905457,
"step": 278
},
{
"epoch": 0.5919661733615222,
"grad_norm": 0.875600278377533,
"learning_rate": 9.769179911697037e-06,
"loss": 1.2291432619094849,
"step": 280
},
{
"epoch": 0.5961945031712473,
"grad_norm": 1.177699327468872,
"learning_rate": 9.764179341103493e-06,
"loss": 1.4657536745071411,
"step": 282
},
{
"epoch": 0.6004228329809725,
"grad_norm": 0.4103357493877411,
"learning_rate": 9.759126639204337e-06,
"loss": 0.9372889995574951,
"step": 284
},
{
"epoch": 0.6046511627906976,
"grad_norm": 0.9785255789756775,
"learning_rate": 9.75402186777088e-06,
"loss": 1.2957736253738403,
"step": 286
},
{
"epoch": 0.6088794926004228,
"grad_norm": 1.252195119857788,
"learning_rate": 9.748865089211004e-06,
"loss": 0.42847904562950134,
"step": 288
},
{
"epoch": 0.6131078224101479,
"grad_norm": 0.8626077771186829,
"learning_rate": 9.743656366568398e-06,
"loss": 0.9476506114006042,
"step": 290
},
{
"epoch": 0.6173361522198731,
"grad_norm": 0.7865381240844727,
"learning_rate": 9.738395763521792e-06,
"loss": 0.8844329714775085,
"step": 292
},
{
"epoch": 0.6215644820295984,
"grad_norm": 7.1966962814331055,
"learning_rate": 9.733083344384164e-06,
"loss": 0.7337439060211182,
"step": 294
},
{
"epoch": 0.6257928118393234,
"grad_norm": 1.26055109500885,
"learning_rate": 9.727719174101981e-06,
"loss": 0.7471031546592712,
"step": 296
},
{
"epoch": 0.6300211416490487,
"grad_norm": 1.9840526580810547,
"learning_rate": 9.722303318254376e-06,
"loss": 1.2770583629608154,
"step": 298
},
{
"epoch": 0.6342494714587738,
"grad_norm": 0.836051344871521,
"learning_rate": 9.71683584305236e-06,
"loss": 1.2418497800827026,
"step": 300
},
{
"epoch": 0.638477801268499,
"grad_norm": 1.838856816291809,
"learning_rate": 9.71131681533802e-06,
"loss": 0.7396769523620605,
"step": 302
},
{
"epoch": 0.642706131078224,
"grad_norm": 1.614635705947876,
"learning_rate": 9.705746302583682e-06,
"loss": 0.8491560220718384,
"step": 304
},
{
"epoch": 0.6469344608879493,
"grad_norm": 0.7019956111907959,
"learning_rate": 9.700124372891104e-06,
"loss": 0.9127112030982971,
"step": 306
},
{
"epoch": 0.6511627906976745,
"grad_norm": 1.1646689176559448,
"learning_rate": 9.694451094990636e-06,
"loss": 1.23308265209198,
"step": 308
},
{
"epoch": 0.6553911205073996,
"grad_norm": 1.3276031017303467,
"learning_rate": 9.68872653824038e-06,
"loss": 1.204702615737915,
"step": 310
},
{
"epoch": 0.6596194503171248,
"grad_norm": 0.9950634241104126,
"learning_rate": 9.68295077262534e-06,
"loss": 1.206320881843567,
"step": 312
},
{
"epoch": 0.6638477801268499,
"grad_norm": 1.1397180557250977,
"learning_rate": 9.677123868756573e-06,
"loss": 1.227231740951538,
"step": 314
},
{
"epoch": 0.6680761099365751,
"grad_norm": 0.7439650297164917,
"learning_rate": 9.671245897870318e-06,
"loss": 1.0743789672851562,
"step": 316
},
{
"epoch": 0.6723044397463002,
"grad_norm": 2.4653406143188477,
"learning_rate": 9.665316931827131e-06,
"loss": 0.7797743082046509,
"step": 318
},
{
"epoch": 0.6765327695560254,
"grad_norm": 1.2759696245193481,
"learning_rate": 9.659337043111006e-06,
"loss": 1.2426550388336182,
"step": 320
},
{
"epoch": 0.6807610993657506,
"grad_norm": 6.864377021789551,
"learning_rate": 9.653306304828483e-06,
"loss": 1.3053348064422607,
"step": 322
},
{
"epoch": 0.6849894291754757,
"grad_norm": 1.2228835821151733,
"learning_rate": 9.647224790707762e-06,
"loss": 1.0797638893127441,
"step": 324
},
{
"epoch": 0.6892177589852009,
"grad_norm": 1.0346611738204956,
"learning_rate": 9.641092575097795e-06,
"loss": 1.2647968530654907,
"step": 326
},
{
"epoch": 0.693446088794926,
"grad_norm": 4.842899322509766,
"learning_rate": 9.63490973296738e-06,
"loss": 1.1791775226593018,
"step": 328
},
{
"epoch": 0.6976744186046512,
"grad_norm": 3.4970157146453857,
"learning_rate": 9.62867633990425e-06,
"loss": 0.8624451756477356,
"step": 330
},
{
"epoch": 0.7019027484143763,
"grad_norm": 0.8797374963760376,
"learning_rate": 9.622392472114136e-06,
"loss": 1.333149790763855,
"step": 332
},
{
"epoch": 0.7061310782241015,
"grad_norm": 2.034334182739258,
"learning_rate": 9.61605820641985e-06,
"loss": 1.2266896963119507,
"step": 334
},
{
"epoch": 0.7103594080338267,
"grad_norm": 0.934144139289856,
"learning_rate": 9.60967362026033e-06,
"loss": 1.207449197769165,
"step": 336
},
{
"epoch": 0.7145877378435518,
"grad_norm": 1.0850398540496826,
"learning_rate": 9.60323879168971e-06,
"loss": 1.3129963874816895,
"step": 338
},
{
"epoch": 0.718816067653277,
"grad_norm": 2.163994550704956,
"learning_rate": 9.596753799376355e-06,
"loss": 1.2012054920196533,
"step": 340
},
{
"epoch": 0.7230443974630021,
"grad_norm": 1.8449740409851074,
"learning_rate": 9.5902187226019e-06,
"loss": 1.1369110345840454,
"step": 342
},
{
"epoch": 0.7272727272727273,
"grad_norm": 3.044912815093994,
"learning_rate": 9.583633641260285e-06,
"loss": 0.8962236642837524,
"step": 344
},
{
"epoch": 0.7315010570824524,
"grad_norm": 6.536602973937988,
"learning_rate": 9.576998635856774e-06,
"loss": 0.8463191986083984,
"step": 346
},
{
"epoch": 0.7357293868921776,
"grad_norm": 1.3714725971221924,
"learning_rate": 9.570313787506974e-06,
"loss": 0.9794567823410034,
"step": 348
},
{
"epoch": 0.7399577167019028,
"grad_norm": 1.8909012079238892,
"learning_rate": 9.563579177935844e-06,
"loss": 1.310282826423645,
"step": 350
},
{
"epoch": 0.7441860465116279,
"grad_norm": 1.8178564310073853,
"learning_rate": 9.556794889476687e-06,
"loss": 1.1669079065322876,
"step": 352
},
{
"epoch": 0.7484143763213531,
"grad_norm": 4.068032264709473,
"learning_rate": 9.549961005070161e-06,
"loss": 0.711179256439209,
"step": 354
},
{
"epoch": 0.7526427061310782,
"grad_norm": 2.254777193069458,
"learning_rate": 9.543077608263245e-06,
"loss": 1.0197420120239258,
"step": 356
},
{
"epoch": 0.7568710359408034,
"grad_norm": 1.0370128154754639,
"learning_rate": 9.536144783208229e-06,
"loss": 1.2137476205825806,
"step": 358
},
{
"epoch": 0.7610993657505285,
"grad_norm": 1.7453880310058594,
"learning_rate": 9.529162614661687e-06,
"loss": 0.7597782015800476,
"step": 360
},
{
"epoch": 0.7653276955602537,
"grad_norm": 1.1258838176727295,
"learning_rate": 9.522131187983433e-06,
"loss": 1.1039535999298096,
"step": 362
},
{
"epoch": 0.7695560253699789,
"grad_norm": 4.441952705383301,
"learning_rate": 9.515050589135485e-06,
"loss": 1.0632761716842651,
"step": 364
},
{
"epoch": 0.773784355179704,
"grad_norm": 2.047109603881836,
"learning_rate": 9.507920904681006e-06,
"loss": 1.2666383981704712,
"step": 366
},
{
"epoch": 0.7780126849894292,
"grad_norm": 1.1188606023788452,
"learning_rate": 9.500742221783257e-06,
"loss": 1.2113542556762695,
"step": 368
},
{
"epoch": 0.7822410147991543,
"grad_norm": 1.4288474321365356,
"learning_rate": 9.493514628204516e-06,
"loss": 0.8752796649932861,
"step": 370
},
{
"epoch": 0.7864693446088795,
"grad_norm": 1.1664959192276,
"learning_rate": 9.48623821230502e-06,
"loss": 1.1909871101379395,
"step": 372
},
{
"epoch": 0.7906976744186046,
"grad_norm": 1.1087419986724854,
"learning_rate": 9.478913063041878e-06,
"loss": 1.2108728885650635,
"step": 374
},
{
"epoch": 0.7949260042283298,
"grad_norm": 0.8613646030426025,
"learning_rate": 9.471539269967982e-06,
"loss": 1.2900152206420898,
"step": 376
},
{
"epoch": 0.7991543340380549,
"grad_norm": 1.0001124143600464,
"learning_rate": 9.464116923230913e-06,
"loss": 0.9615880250930786,
"step": 378
},
{
"epoch": 0.8033826638477801,
"grad_norm": 1.690933108329773,
"learning_rate": 9.456646113571844e-06,
"loss": 1.306993007659912,
"step": 380
},
{
"epoch": 0.8076109936575053,
"grad_norm": 0.9221349358558655,
"learning_rate": 9.449126932324424e-06,
"loss": 1.1692599058151245,
"step": 382
},
{
"epoch": 0.8118393234672304,
"grad_norm": 0.9589253664016724,
"learning_rate": 9.441559471413665e-06,
"loss": 0.9634546041488647,
"step": 384
},
{
"epoch": 0.8160676532769556,
"grad_norm": 2.1232404708862305,
"learning_rate": 9.433943823354816e-06,
"loss": 0.9619753360748291,
"step": 386
},
{
"epoch": 0.8202959830866807,
"grad_norm": 1.2578006982803345,
"learning_rate": 9.426280081252232e-06,
"loss": 1.4784586429595947,
"step": 388
},
{
"epoch": 0.8245243128964059,
"grad_norm": 2.922351121902466,
"learning_rate": 9.418568338798245e-06,
"loss": 1.1491820812225342,
"step": 390
},
{
"epoch": 0.828752642706131,
"grad_norm": 2.3026421070098877,
"learning_rate": 9.410808690271999e-06,
"loss": 0.9402557611465454,
"step": 392
},
{
"epoch": 0.8329809725158562,
"grad_norm": 1.1002447605133057,
"learning_rate": 9.40300123053832e-06,
"loss": 1.2377575635910034,
"step": 394
},
{
"epoch": 0.8372093023255814,
"grad_norm": 0.28478723764419556,
"learning_rate": 9.395146055046535e-06,
"loss": 0.7619230151176453,
"step": 396
},
{
"epoch": 0.8414376321353065,
"grad_norm": 0.699545681476593,
"learning_rate": 9.387243259829323e-06,
"loss": 0.6848609447479248,
"step": 398
},
{
"epoch": 0.8456659619450317,
"grad_norm": 1.4446637630462646,
"learning_rate": 9.37929294150153e-06,
"loss": 1.128559947013855,
"step": 400
},
{
"epoch": 0.8498942917547568,
"grad_norm": 1.7312356233596802,
"learning_rate": 9.371295197258992e-06,
"loss": 0.9686368703842163,
"step": 402
},
{
"epoch": 0.854122621564482,
"grad_norm": 1.6062495708465576,
"learning_rate": 9.363250124877345e-06,
"loss": 0.8305947184562683,
"step": 404
},
{
"epoch": 0.8583509513742071,
"grad_norm": 3.7103312015533447,
"learning_rate": 9.35515782271083e-06,
"loss": 0.994686484336853,
"step": 406
},
{
"epoch": 0.8625792811839323,
"grad_norm": 0.8811690807342529,
"learning_rate": 9.347018389691093e-06,
"loss": 1.0020716190338135,
"step": 408
},
{
"epoch": 0.8668076109936576,
"grad_norm": 2.8474152088165283,
"learning_rate": 9.338831925325974e-06,
"loss": 0.8986063003540039,
"step": 410
},
{
"epoch": 0.8710359408033826,
"grad_norm": 1.6215959787368774,
"learning_rate": 9.330598529698287e-06,
"loss": 1.1754546165466309,
"step": 412
},
{
"epoch": 0.8752642706131079,
"grad_norm": 2.0055744647979736,
"learning_rate": 9.322318303464603e-06,
"loss": 1.0183253288269043,
"step": 414
},
{
"epoch": 0.879492600422833,
"grad_norm": 4.045909881591797,
"learning_rate": 9.313991347854015e-06,
"loss": 1.1169545650482178,
"step": 416
},
{
"epoch": 0.8837209302325582,
"grad_norm": 1.152048945426941,
"learning_rate": 9.305617764666897e-06,
"loss": 1.2168593406677246,
"step": 418
},
{
"epoch": 0.8879492600422833,
"grad_norm": 1.2169914245605469,
"learning_rate": 9.297197656273672e-06,
"loss": 1.3040281534194946,
"step": 420
},
{
"epoch": 0.8921775898520085,
"grad_norm": 2.7920634746551514,
"learning_rate": 9.288731125613544e-06,
"loss": 0.9540579915046692,
"step": 422
},
{
"epoch": 0.8964059196617337,
"grad_norm": 1.1594703197479248,
"learning_rate": 9.280218276193255e-06,
"loss": 1.2285897731781006,
"step": 424
},
{
"epoch": 0.9006342494714588,
"grad_norm": 0.8686127066612244,
"learning_rate": 9.271659212085807e-06,
"loss": 1.0948278903961182,
"step": 426
},
{
"epoch": 0.904862579281184,
"grad_norm": 2.1172947883605957,
"learning_rate": 9.263054037929199e-06,
"loss": 0.876900851726532,
"step": 428
},
{
"epoch": 0.9090909090909091,
"grad_norm": 1.7088205814361572,
"learning_rate": 9.254402858925142e-06,
"loss": 1.1154334545135498,
"step": 430
},
{
"epoch": 0.9133192389006343,
"grad_norm": 1.0164040327072144,
"learning_rate": 9.245705780837774e-06,
"loss": 1.2577118873596191,
"step": 432
},
{
"epoch": 0.9175475687103594,
"grad_norm": 3.4467175006866455,
"learning_rate": 9.236962909992368e-06,
"loss": 0.5238693356513977,
"step": 434
},
{
"epoch": 0.9217758985200846,
"grad_norm": 4.376553535461426,
"learning_rate": 9.228174353274037e-06,
"loss": 0.3832204043865204,
"step": 436
},
{
"epoch": 0.9260042283298098,
"grad_norm": 0.7531916499137878,
"learning_rate": 9.219340218126413e-06,
"loss": 0.9414502382278442,
"step": 438
},
{
"epoch": 0.9302325581395349,
"grad_norm": 1.0536175966262817,
"learning_rate": 9.210460612550348e-06,
"loss": 1.1374485492706299,
"step": 440
},
{
"epoch": 0.9344608879492601,
"grad_norm": 4.08363676071167,
"learning_rate": 9.20153564510259e-06,
"loss": 0.7624293565750122,
"step": 442
},
{
"epoch": 0.9386892177589852,
"grad_norm": 0.954616129398346,
"learning_rate": 9.19256542489445e-06,
"loss": 1.253175139427185,
"step": 444
},
{
"epoch": 0.9429175475687104,
"grad_norm": 0.7408168911933899,
"learning_rate": 9.183550061590476e-06,
"loss": 1.0929017066955566,
"step": 446
},
{
"epoch": 0.9471458773784355,
"grad_norm": 2.451781749725342,
"learning_rate": 9.174489665407103e-06,
"loss": 1.0880546569824219,
"step": 448
},
{
"epoch": 0.9513742071881607,
"grad_norm": 4.312840461730957,
"learning_rate": 9.165384347111317e-06,
"loss": 0.7974986433982849,
"step": 450
},
{
"epoch": 0.9556025369978859,
"grad_norm": 0.8640457987785339,
"learning_rate": 9.156234218019291e-06,
"loss": 0.9814733862876892,
"step": 452
},
{
"epoch": 0.959830866807611,
"grad_norm": 2.013702869415283,
"learning_rate": 9.147039389995029e-06,
"loss": 1.3165922164916992,
"step": 454
},
{
"epoch": 0.9640591966173362,
"grad_norm": 4.8829216957092285,
"learning_rate": 9.137799975448998e-06,
"loss": 0.6675208806991577,
"step": 456
},
{
"epoch": 0.9682875264270613,
"grad_norm": 5.20572566986084,
"learning_rate": 9.128516087336755e-06,
"loss": 0.8903831839561462,
"step": 458
},
{
"epoch": 0.9725158562367865,
"grad_norm": 1.5217186212539673,
"learning_rate": 9.119187839157559e-06,
"loss": 1.2241164445877075,
"step": 460
},
{
"epoch": 0.9767441860465116,
"grad_norm": 1.8690365552902222,
"learning_rate": 9.109815344952995e-06,
"loss": 1.2088654041290283,
"step": 462
},
{
"epoch": 0.9809725158562368,
"grad_norm": 2.7457756996154785,
"learning_rate": 9.100398719305566e-06,
"loss": 0.7155973315238953,
"step": 464
},
{
"epoch": 0.985200845665962,
"grad_norm": 0.9488766193389893,
"learning_rate": 9.090938077337305e-06,
"loss": 1.1860933303833008,
"step": 466
},
{
"epoch": 0.9894291754756871,
"grad_norm": 2.899150848388672,
"learning_rate": 9.081433534708363e-06,
"loss": 0.9845510125160217,
"step": 468
},
{
"epoch": 0.9936575052854123,
"grad_norm": 1.8763692378997803,
"learning_rate": 9.071885207615591e-06,
"loss": 1.2428233623504639,
"step": 470
},
{
"epoch": 0.9978858350951374,
"grad_norm": 0.9538397789001465,
"learning_rate": 9.062293212791123e-06,
"loss": 1.2468990087509155,
"step": 472
},
{
"epoch": 1.0021141649048626,
"grad_norm": 1.4480160474777222,
"learning_rate": 9.052657667500951e-06,
"loss": 0.9474197030067444,
"step": 474
},
{
"epoch": 1.0063424947145878,
"grad_norm": 0.8437069654464722,
"learning_rate": 9.042978689543487e-06,
"loss": 1.1090214252471924,
"step": 476
},
{
"epoch": 1.0105708245243128,
"grad_norm": 3.502439498901367,
"learning_rate": 9.033256397248124e-06,
"loss": 0.4405559003353119,
"step": 478
},
{
"epoch": 1.014799154334038,
"grad_norm": 3.1259708404541016,
"learning_rate": 9.02349090947379e-06,
"loss": 0.7095645070075989,
"step": 480
},
{
"epoch": 1.0190274841437632,
"grad_norm": 3.914839267730713,
"learning_rate": 9.013682345607502e-06,
"loss": 0.6677669286727905,
"step": 482
},
{
"epoch": 1.0232558139534884,
"grad_norm": 0.8538970947265625,
"learning_rate": 9.003830825562887e-06,
"loss": 0.9967820644378662,
"step": 484
},
{
"epoch": 1.0274841437632136,
"grad_norm": 2.6352531909942627,
"learning_rate": 8.99393646977874e-06,
"loss": 1.0178574323654175,
"step": 486
},
{
"epoch": 1.0317124735729386,
"grad_norm": 1.18100905418396,
"learning_rate": 8.983999399217536e-06,
"loss": 0.8780731558799744,
"step": 488
},
{
"epoch": 1.0359408033826638,
"grad_norm": 2.754648208618164,
"learning_rate": 8.974019735363952e-06,
"loss": 0.8479733467102051,
"step": 490
},
{
"epoch": 1.040169133192389,
"grad_norm": 2.1098480224609375,
"learning_rate": 8.963997600223387e-06,
"loss": 0.7347881197929382,
"step": 492
},
{
"epoch": 1.0443974630021142,
"grad_norm": 1.1570031642913818,
"learning_rate": 8.953933116320474e-06,
"loss": 1.124723196029663,
"step": 494
},
{
"epoch": 1.0486257928118394,
"grad_norm": 1.4699747562408447,
"learning_rate": 8.943826406697569e-06,
"loss": 0.9707335233688354,
"step": 496
},
{
"epoch": 1.0528541226215644,
"grad_norm": 1.2355912923812866,
"learning_rate": 8.93367759491326e-06,
"loss": 0.7460411787033081,
"step": 498
},
{
"epoch": 1.0570824524312896,
"grad_norm": 1.862128496170044,
"learning_rate": 8.923486805040845e-06,
"loss": 0.8815531730651855,
"step": 500
},
{
"epoch": 1.0613107822410148,
"grad_norm": 0.9728148579597473,
"learning_rate": 8.913254161666828e-06,
"loss": 1.1246565580368042,
"step": 502
},
{
"epoch": 1.06553911205074,
"grad_norm": 1.5676933526992798,
"learning_rate": 8.902979789889382e-06,
"loss": 1.0279873609542847,
"step": 504
},
{
"epoch": 1.069767441860465,
"grad_norm": 5.227234840393066,
"learning_rate": 8.892663815316833e-06,
"loss": 0.8011895418167114,
"step": 506
},
{
"epoch": 1.0739957716701902,
"grad_norm": 1.3110841512680054,
"learning_rate": 8.882306364066114e-06,
"loss": 0.7954361438751221,
"step": 508
},
{
"epoch": 1.0782241014799154,
"grad_norm": 4.763819217681885,
"learning_rate": 8.871907562761226e-06,
"loss": 0.9455935955047607,
"step": 510
},
{
"epoch": 1.0824524312896406,
"grad_norm": 1.2440690994262695,
"learning_rate": 8.861467538531692e-06,
"loss": 0.9990676641464233,
"step": 512
},
{
"epoch": 1.0866807610993658,
"grad_norm": 1.9285670518875122,
"learning_rate": 8.850986419011006e-06,
"loss": 1.0558518171310425,
"step": 514
},
{
"epoch": 1.0909090909090908,
"grad_norm": 1.0552536249160767,
"learning_rate": 8.840464332335062e-06,
"loss": 0.8359754085540771,
"step": 516
},
{
"epoch": 1.095137420718816,
"grad_norm": 1.1247268915176392,
"learning_rate": 8.829901407140599e-06,
"loss": 0.8400765657424927,
"step": 518
},
{
"epoch": 1.0993657505285412,
"grad_norm": 4.692409992218018,
"learning_rate": 8.81929777256362e-06,
"loss": 0.6272275447845459,
"step": 520
},
{
"epoch": 1.1035940803382664,
"grad_norm": 1.560003638267517,
"learning_rate": 8.808653558237818e-06,
"loss": 1.3277077674865723,
"step": 522
},
{
"epoch": 1.1078224101479917,
"grad_norm": 37.027408599853516,
"learning_rate": 8.797968894292992e-06,
"loss": 0.7344569563865662,
"step": 524
},
{
"epoch": 1.1120507399577166,
"grad_norm": 1.5640825033187866,
"learning_rate": 8.787243911353447e-06,
"loss": 1.2844988107681274,
"step": 526
},
{
"epoch": 1.1162790697674418,
"grad_norm": 1.1317330598831177,
"learning_rate": 8.776478740536409e-06,
"loss": 0.8073318600654602,
"step": 528
},
{
"epoch": 1.120507399577167,
"grad_norm": 8.045904159545898,
"learning_rate": 8.76567351345042e-06,
"loss": 0.49239689111709595,
"step": 530
},
{
"epoch": 1.1247357293868923,
"grad_norm": 1.1490347385406494,
"learning_rate": 8.754828362193719e-06,
"loss": 1.0262296199798584,
"step": 532
},
{
"epoch": 1.1289640591966172,
"grad_norm": 0.7461928725242615,
"learning_rate": 8.743943419352638e-06,
"loss": 0.787236213684082,
"step": 534
},
{
"epoch": 1.1331923890063424,
"grad_norm": 0.7733618021011353,
"learning_rate": 8.733018817999977e-06,
"loss": 0.935680627822876,
"step": 536
},
{
"epoch": 1.1374207188160677,
"grad_norm": 1.946588158607483,
"learning_rate": 8.722054691693374e-06,
"loss": 0.8061190843582153,
"step": 538
},
{
"epoch": 1.1416490486257929,
"grad_norm": 4.8798747062683105,
"learning_rate": 8.711051174473682e-06,
"loss": 0.7387879490852356,
"step": 540
},
{
"epoch": 1.145877378435518,
"grad_norm": 2.2170217037200928,
"learning_rate": 8.700008400863316e-06,
"loss": 1.0502710342407227,
"step": 542
},
{
"epoch": 1.150105708245243,
"grad_norm": 3.5251572132110596,
"learning_rate": 8.688926505864624e-06,
"loss": 0.7766356468200684,
"step": 544
},
{
"epoch": 1.1543340380549683,
"grad_norm": 0.8779979944229126,
"learning_rate": 8.67780562495822e-06,
"loss": 1.1293692588806152,
"step": 546
},
{
"epoch": 1.1585623678646935,
"grad_norm": 2.156127691268921,
"learning_rate": 8.666645894101346e-06,
"loss": 1.241573691368103,
"step": 548
},
{
"epoch": 1.1627906976744187,
"grad_norm": 2.2766292095184326,
"learning_rate": 8.655447449726189e-06,
"loss": 0.7699335217475891,
"step": 550
},
{
"epoch": 1.1670190274841437,
"grad_norm": 0.8864946365356445,
"learning_rate": 8.644210428738237e-06,
"loss": 1.0497722625732422,
"step": 552
},
{
"epoch": 1.1712473572938689,
"grad_norm": 1.126464605331421,
"learning_rate": 8.632934968514582e-06,
"loss": 1.1691752672195435,
"step": 554
},
{
"epoch": 1.175475687103594,
"grad_norm": 1.1422570943832397,
"learning_rate": 8.621621206902256e-06,
"loss": 1.0135767459869385,
"step": 556
},
{
"epoch": 1.1797040169133193,
"grad_norm": 1.1006174087524414,
"learning_rate": 8.61026928221654e-06,
"loss": 1.017207384109497,
"step": 558
},
{
"epoch": 1.1839323467230445,
"grad_norm": 0.616320013999939,
"learning_rate": 8.598879333239274e-06,
"loss": 0.9789807200431824,
"step": 560
},
{
"epoch": 1.1881606765327695,
"grad_norm": 1.546005368232727,
"learning_rate": 8.587451499217161e-06,
"loss": 0.6851304173469543,
"step": 562
},
{
"epoch": 1.1923890063424947,
"grad_norm": 1.207463026046753,
"learning_rate": 8.575985919860065e-06,
"loss": 1.0593668222427368,
"step": 564
},
{
"epoch": 1.1966173361522199,
"grad_norm": 2.237175226211548,
"learning_rate": 8.564482735339296e-06,
"loss": 0.9086387753486633,
"step": 566
},
{
"epoch": 1.200845665961945,
"grad_norm": 1.6928720474243164,
"learning_rate": 8.552942086285909e-06,
"loss": 1.2082197666168213,
"step": 568
},
{
"epoch": 1.20507399577167,
"grad_norm": 1.5580166578292847,
"learning_rate": 8.541364113788974e-06,
"loss": 1.1556533575057983,
"step": 570
},
{
"epoch": 1.2093023255813953,
"grad_norm": 12.289173126220703,
"learning_rate": 8.529748959393855e-06,
"loss": 1.0296770334243774,
"step": 572
},
{
"epoch": 1.2135306553911205,
"grad_norm": 0.9645960927009583,
"learning_rate": 8.518096765100485e-06,
"loss": 0.7730268836021423,
"step": 574
},
{
"epoch": 1.2177589852008457,
"grad_norm": 2.6180851459503174,
"learning_rate": 8.50640767336162e-06,
"loss": 0.7640129327774048,
"step": 576
},
{
"epoch": 1.221987315010571,
"grad_norm": 2.6575820446014404,
"learning_rate": 8.494681827081096e-06,
"loss": 0.9303947687149048,
"step": 578
},
{
"epoch": 1.226215644820296,
"grad_norm": 1.625160813331604,
"learning_rate": 8.482919369612103e-06,
"loss": 0.7814822196960449,
"step": 580
},
{
"epoch": 1.230443974630021,
"grad_norm": 1.0828288793563843,
"learning_rate": 8.471120444755405e-06,
"loss": 1.1155931949615479,
"step": 582
},
{
"epoch": 1.2346723044397463,
"grad_norm": 2.7364377975463867,
"learning_rate": 8.4592851967576e-06,
"loss": 0.9870080947875977,
"step": 584
},
{
"epoch": 1.2389006342494715,
"grad_norm": 3.5304481983184814,
"learning_rate": 8.447413770309349e-06,
"loss": 0.7886775135993958,
"step": 586
},
{
"epoch": 1.2431289640591967,
"grad_norm": 0.9153185486793518,
"learning_rate": 8.435506310543607e-06,
"loss": 0.8513044714927673,
"step": 588
},
{
"epoch": 1.2473572938689217,
"grad_norm": 0.764589250087738,
"learning_rate": 8.42356296303386e-06,
"loss": 0.6052752733230591,
"step": 590
},
{
"epoch": 1.251585623678647,
"grad_norm": 0.6078240871429443,
"learning_rate": 8.411583873792325e-06,
"loss": 0.9022788405418396,
"step": 592
},
{
"epoch": 1.255813953488372,
"grad_norm": 1.1883888244628906,
"learning_rate": 8.399569189268182e-06,
"loss": 1.0751527547836304,
"step": 594
},
{
"epoch": 1.2600422832980973,
"grad_norm": 1.4856717586517334,
"learning_rate": 8.387519056345778e-06,
"loss": 0.714518129825592,
"step": 596
},
{
"epoch": 1.2642706131078225,
"grad_norm": 0.8114223480224609,
"learning_rate": 8.375433622342824e-06,
"loss": 1.0766490697860718,
"step": 598
},
{
"epoch": 1.2684989429175475,
"grad_norm": 1.6652065515518188,
"learning_rate": 8.363313035008613e-06,
"loss": 1.2978875637054443,
"step": 600
},
{
"epoch": 1.2727272727272727,
"grad_norm": 0.7716587781906128,
"learning_rate": 8.351157442522187e-06,
"loss": 0.9173111915588379,
"step": 602
},
{
"epoch": 1.276955602536998,
"grad_norm": 2.667154550552368,
"learning_rate": 8.338966993490554e-06,
"loss": 1.0912063121795654,
"step": 604
},
{
"epoch": 1.2811839323467231,
"grad_norm": 1.6391500234603882,
"learning_rate": 8.326741836946852e-06,
"loss": 0.7595437169075012,
"step": 606
},
{
"epoch": 1.285412262156448,
"grad_norm": 1.0511261224746704,
"learning_rate": 8.314482122348523e-06,
"loss": 1.0843315124511719,
"step": 608
},
{
"epoch": 1.2896405919661733,
"grad_norm": 0.8262497782707214,
"learning_rate": 8.30218799957551e-06,
"loss": 0.7324892282485962,
"step": 610
},
{
"epoch": 1.2938689217758985,
"grad_norm": 0.9290772080421448,
"learning_rate": 8.289859618928404e-06,
"loss": 0.6603423357009888,
"step": 612
},
{
"epoch": 1.2980972515856237,
"grad_norm": 0.9745301604270935,
"learning_rate": 8.277497131126608e-06,
"loss": 0.8386492133140564,
"step": 614
},
{
"epoch": 1.302325581395349,
"grad_norm": 3.0962748527526855,
"learning_rate": 8.265100687306503e-06,
"loss": 0.5407689213752747,
"step": 616
},
{
"epoch": 1.306553911205074,
"grad_norm": 1.9692397117614746,
"learning_rate": 8.252670439019595e-06,
"loss": 0.9149762988090515,
"step": 618
},
{
"epoch": 1.3107822410147991,
"grad_norm": 1.0410351753234863,
"learning_rate": 8.240206538230663e-06,
"loss": 1.092056155204773,
"step": 620
},
{
"epoch": 1.3150105708245243,
"grad_norm": 0.8065683841705322,
"learning_rate": 8.227709137315905e-06,
"loss": 1.0072970390319824,
"step": 622
},
{
"epoch": 1.3192389006342495,
"grad_norm": 5.005168914794922,
"learning_rate": 8.215178389061064e-06,
"loss": 0.6487065553665161,
"step": 624
},
{
"epoch": 1.3234672304439745,
"grad_norm": 0.8799970149993896,
"learning_rate": 8.202614446659572e-06,
"loss": 1.1100835800170898,
"step": 626
},
{
"epoch": 1.3276955602536997,
"grad_norm": 1.4324008226394653,
"learning_rate": 8.190017463710675e-06,
"loss": 1.2896560430526733,
"step": 628
},
{
"epoch": 1.331923890063425,
"grad_norm": 0.9906747341156006,
"learning_rate": 8.177387594217548e-06,
"loss": 0.9537054300308228,
"step": 630
},
{
"epoch": 1.3361522198731501,
"grad_norm": 0.8879162073135376,
"learning_rate": 8.164724992585421e-06,
"loss": 0.5851172804832458,
"step": 632
},
{
"epoch": 1.3403805496828753,
"grad_norm": 0.9549749493598938,
"learning_rate": 8.152029813619685e-06,
"loss": 0.7516475915908813,
"step": 634
},
{
"epoch": 1.3446088794926006,
"grad_norm": 0.8472812175750732,
"learning_rate": 8.139302212524006e-06,
"loss": 0.7518019676208496,
"step": 636
},
{
"epoch": 1.3488372093023255,
"grad_norm": 6.719564914703369,
"learning_rate": 8.126542344898419e-06,
"loss": 0.9743859767913818,
"step": 638
},
{
"epoch": 1.3530655391120507,
"grad_norm": 0.49654120206832886,
"learning_rate": 8.113750366737431e-06,
"loss": 0.7649727463722229,
"step": 640
},
{
"epoch": 1.357293868921776,
"grad_norm": 0.7154551148414612,
"learning_rate": 8.100926434428116e-06,
"loss": 1.1713658571243286,
"step": 642
},
{
"epoch": 1.361522198731501,
"grad_norm": 1.369627594947815,
"learning_rate": 8.0880707047482e-06,
"loss": 1.0085299015045166,
"step": 644
},
{
"epoch": 1.3657505285412261,
"grad_norm": 4.835188865661621,
"learning_rate": 8.075183334864139e-06,
"loss": 0.7513466477394104,
"step": 646
},
{
"epoch": 1.3699788583509513,
"grad_norm": 4.035396575927734,
"learning_rate": 8.062264482329209e-06,
"loss": 0.7092136740684509,
"step": 648
},
{
"epoch": 1.3742071881606766,
"grad_norm": 4.148628234863281,
"learning_rate": 8.049314305081575e-06,
"loss": 0.7393103241920471,
"step": 650
},
{
"epoch": 1.3784355179704018,
"grad_norm": 5.343515396118164,
"learning_rate": 8.036332961442355e-06,
"loss": 0.7626214027404785,
"step": 652
},
{
"epoch": 1.382663847780127,
"grad_norm": 0.8783438801765442,
"learning_rate": 8.023320610113691e-06,
"loss": 1.1477959156036377,
"step": 654
},
{
"epoch": 1.386892177589852,
"grad_norm": 0.8855472207069397,
"learning_rate": 8.010277410176801e-06,
"loss": 0.9967976212501526,
"step": 656
},
{
"epoch": 1.3911205073995772,
"grad_norm": 2.955355644226074,
"learning_rate": 7.997203521090053e-06,
"loss": 0.8183122873306274,
"step": 658
},
{
"epoch": 1.3953488372093024,
"grad_norm": 1.0300184488296509,
"learning_rate": 7.984099102686986e-06,
"loss": 0.9316401481628418,
"step": 660
},
{
"epoch": 1.3995771670190273,
"grad_norm": 0.7101503610610962,
"learning_rate": 7.970964315174384e-06,
"loss": 1.0914911031723022,
"step": 662
},
{
"epoch": 1.4038054968287526,
"grad_norm": 0.23997701704502106,
"learning_rate": 7.9577993191303e-06,
"loss": 0.9474119544029236,
"step": 664
},
{
"epoch": 1.4080338266384778,
"grad_norm": 1.3534051179885864,
"learning_rate": 7.944604275502102e-06,
"loss": 1.416314959526062,
"step": 666
},
{
"epoch": 1.412262156448203,
"grad_norm": 0.98891681432724,
"learning_rate": 7.931379345604495e-06,
"loss": 1.0960770845413208,
"step": 668
},
{
"epoch": 1.4164904862579282,
"grad_norm": 1.527935266494751,
"learning_rate": 7.918124691117563e-06,
"loss": 0.7882940173149109,
"step": 670
},
{
"epoch": 1.4207188160676534,
"grad_norm": 1.008180022239685,
"learning_rate": 7.904840474084783e-06,
"loss": 1.0377100706100464,
"step": 672
},
{
"epoch": 1.4249471458773784,
"grad_norm": 0.8609725832939148,
"learning_rate": 7.891526856911046e-06,
"loss": 0.5663607120513916,
"step": 674
},
{
"epoch": 1.4291754756871036,
"grad_norm": 1.3604178428649902,
"learning_rate": 7.878184002360664e-06,
"loss": 0.7705097198486328,
"step": 676
},
{
"epoch": 1.4334038054968288,
"grad_norm": 0.8752778768539429,
"learning_rate": 7.864812073555402e-06,
"loss": 1.2220746278762817,
"step": 678
},
{
"epoch": 1.437632135306554,
"grad_norm": 1.1921929121017456,
"learning_rate": 7.851411233972461e-06,
"loss": 1.133432149887085,
"step": 680
},
{
"epoch": 1.441860465116279,
"grad_norm": 1.7379199266433716,
"learning_rate": 7.837981647442486e-06,
"loss": 1.1666901111602783,
"step": 682
},
{
"epoch": 1.4460887949260042,
"grad_norm": 0.9507070779800415,
"learning_rate": 7.82452347814757e-06,
"loss": 1.0301626920700073,
"step": 684
},
{
"epoch": 1.4503171247357294,
"grad_norm": 3.450653314590454,
"learning_rate": 7.811036890619239e-06,
"loss": 1.3803545236587524,
"step": 686
},
{
"epoch": 1.4545454545454546,
"grad_norm": 0.8120247721672058,
"learning_rate": 7.797522049736441e-06,
"loss": 1.085930585861206,
"step": 688
},
{
"epoch": 1.4587737843551798,
"grad_norm": 1.3309519290924072,
"learning_rate": 7.78397912072354e-06,
"loss": 0.4181835353374481,
"step": 690
},
{
"epoch": 1.463002114164905,
"grad_norm": 0.782172441482544,
"learning_rate": 7.770408269148281e-06,
"loss": 1.1031650304794312,
"step": 692
},
{
"epoch": 1.46723044397463,
"grad_norm": 1.943954348564148,
"learning_rate": 7.756809660919778e-06,
"loss": 0.6692284941673279,
"step": 694
},
{
"epoch": 1.4714587737843552,
"grad_norm": 1.5172237157821655,
"learning_rate": 7.74318346228648e-06,
"loss": 1.0448495149612427,
"step": 696
},
{
"epoch": 1.4756871035940804,
"grad_norm": 1.2889461517333984,
"learning_rate": 7.72952983983414e-06,
"loss": 1.0857034921646118,
"step": 698
},
{
"epoch": 1.4799154334038054,
"grad_norm": 4.070534706115723,
"learning_rate": 7.715848960483776e-06,
"loss": 0.9387965202331543,
"step": 700
},
{
"epoch": 1.4841437632135306,
"grad_norm": 0.6505371332168579,
"learning_rate": 7.702140991489637e-06,
"loss": 1.0458344221115112,
"step": 702
},
{
"epoch": 1.4883720930232558,
"grad_norm": 0.7632599472999573,
"learning_rate": 7.688406100437149e-06,
"loss": 0.9895470142364502,
"step": 704
},
{
"epoch": 1.492600422832981,
"grad_norm": 0.6682064533233643,
"learning_rate": 7.674644455240878e-06,
"loss": 0.7049736380577087,
"step": 706
},
{
"epoch": 1.4968287526427062,
"grad_norm": 0.5240515470504761,
"learning_rate": 7.660856224142456e-06,
"loss": 0.2763656973838806,
"step": 708
},
{
"epoch": 1.5010570824524314,
"grad_norm": 8.019431114196777,
"learning_rate": 7.647041575708555e-06,
"loss": 0.6486303210258484,
"step": 710
},
{
"epoch": 1.5052854122621564,
"grad_norm": 1.4818682670593262,
"learning_rate": 7.633200678828804e-06,
"loss": 0.5208026766777039,
"step": 712
},
{
"epoch": 1.5095137420718816,
"grad_norm": 1.9431403875350952,
"learning_rate": 7.6193337027137215e-06,
"loss": 0.8331357836723328,
"step": 714
},
{
"epoch": 1.5137420718816068,
"grad_norm": 1.2872241735458374,
"learning_rate": 7.605440816892667e-06,
"loss": 1.033551812171936,
"step": 716
},
{
"epoch": 1.5179704016913318,
"grad_norm": 1.0277150869369507,
"learning_rate": 7.591522191211746e-06,
"loss": 0.9546146988868713,
"step": 718
},
{
"epoch": 1.522198731501057,
"grad_norm": 0.4126553237438202,
"learning_rate": 7.577577995831755e-06,
"loss": 0.8476111888885498,
"step": 720
},
{
"epoch": 1.5264270613107822,
"grad_norm": 1.9305791854858398,
"learning_rate": 7.563608401226081e-06,
"loss": 0.8556215167045593,
"step": 722
},
{
"epoch": 1.5306553911205074,
"grad_norm": 0.8215469717979431,
"learning_rate": 7.549613578178632e-06,
"loss": 0.9912751913070679,
"step": 724
},
{
"epoch": 1.5348837209302326,
"grad_norm": 1.11053466796875,
"learning_rate": 7.535593697781741e-06,
"loss": 0.8932139873504639,
"step": 726
},
{
"epoch": 1.5391120507399578,
"grad_norm": 2.1221063137054443,
"learning_rate": 7.52154893143408e-06,
"loss": 0.7093181014060974,
"step": 728
},
{
"epoch": 1.543340380549683,
"grad_norm": 1.3867255449295044,
"learning_rate": 7.50747945083856e-06,
"loss": 0.8107166290283203,
"step": 730
},
{
"epoch": 1.547568710359408,
"grad_norm": 0.6168336272239685,
"learning_rate": 7.493385428000234e-06,
"loss": 1.0184130668640137,
"step": 732
},
{
"epoch": 1.5517970401691332,
"grad_norm": 0.5808829069137573,
"learning_rate": 7.479267035224191e-06,
"loss": 0.23352989554405212,
"step": 734
},
{
"epoch": 1.5560253699788582,
"grad_norm": 0.7862116098403931,
"learning_rate": 7.465124445113459e-06,
"loss": 0.9731864929199219,
"step": 736
},
{
"epoch": 1.5602536997885834,
"grad_norm": 0.4767437279224396,
"learning_rate": 7.450957830566878e-06,
"loss": 0.7167576551437378,
"step": 738
},
{
"epoch": 1.5644820295983086,
"grad_norm": 0.957775890827179,
"learning_rate": 7.436767364777009e-06,
"loss": 1.0845588445663452,
"step": 740
},
{
"epoch": 1.5687103594080338,
"grad_norm": 2.063708782196045,
"learning_rate": 7.422553221227993e-06,
"loss": 0.7711994647979736,
"step": 742
},
{
"epoch": 1.572938689217759,
"grad_norm": 0.9038634300231934,
"learning_rate": 7.4083155736934484e-06,
"loss": 1.0538545846939087,
"step": 744
},
{
"epoch": 1.5771670190274842,
"grad_norm": 1.7808231115341187,
"learning_rate": 7.394054596234336e-06,
"loss": 1.179479718208313,
"step": 746
},
{
"epoch": 1.5813953488372094,
"grad_norm": 1.2698373794555664,
"learning_rate": 7.379770463196831e-06,
"loss": 0.9535448551177979,
"step": 748
},
{
"epoch": 1.5856236786469344,
"grad_norm": 0.7018219828605652,
"learning_rate": 7.365463349210203e-06,
"loss": 0.05907959118485451,
"step": 750
},
{
"epoch": 1.5898520084566596,
"grad_norm": 1.2816579341888428,
"learning_rate": 7.351133429184666e-06,
"loss": 0.8288397192955017,
"step": 752
},
{
"epoch": 1.5940803382663846,
"grad_norm": 1.3558180332183838,
"learning_rate": 7.336780878309251e-06,
"loss": 1.0991040468215942,
"step": 754
},
{
"epoch": 1.5983086680761098,
"grad_norm": 0.9416487216949463,
"learning_rate": 7.322405872049657e-06,
"loss": 1.0753154754638672,
"step": 756
},
{
"epoch": 1.602536997885835,
"grad_norm": 0.8170706033706665,
"learning_rate": 7.308008586146111e-06,
"loss": 1.1325535774230957,
"step": 758
},
{
"epoch": 1.6067653276955602,
"grad_norm": 0.7141672968864441,
"learning_rate": 7.293589196611221e-06,
"loss": 1.0982680320739746,
"step": 760
},
{
"epoch": 1.6109936575052854,
"grad_norm": 0.9254981875419617,
"learning_rate": 7.279147879727811e-06,
"loss": 1.0790622234344482,
"step": 762
},
{
"epoch": 1.6152219873150107,
"grad_norm": 1.559340000152588,
"learning_rate": 7.264684812046787e-06,
"loss": 0.9687791466712952,
"step": 764
},
{
"epoch": 1.6194503171247359,
"grad_norm": 7.50747537612915,
"learning_rate": 7.250200170384961e-06,
"loss": 0.7920904159545898,
"step": 766
},
{
"epoch": 1.6236786469344608,
"grad_norm": 1.4139152765274048,
"learning_rate": 7.235694131822894e-06,
"loss": 1.0722813606262207,
"step": 768
},
{
"epoch": 1.627906976744186,
"grad_norm": 1.4232721328735352,
"learning_rate": 7.221166873702738e-06,
"loss": 1.071115255355835,
"step": 770
},
{
"epoch": 1.6321353065539113,
"grad_norm": 0.8064752221107483,
"learning_rate": 7.206618573626058e-06,
"loss": 1.0536168813705444,
"step": 772
},
{
"epoch": 1.6363636363636362,
"grad_norm": 0.8003419041633606,
"learning_rate": 7.19204940945167e-06,
"loss": 0.7976323366165161,
"step": 774
},
{
"epoch": 1.6405919661733614,
"grad_norm": 0.991787850856781,
"learning_rate": 7.177459559293456e-06,
"loss": 1.0970121622085571,
"step": 776
},
{
"epoch": 1.6448202959830867,
"grad_norm": 0.8416776657104492,
"learning_rate": 7.162849201518198e-06,
"loss": 0.6580658555030823,
"step": 778
},
{
"epoch": 1.6490486257928119,
"grad_norm": 0.9339558482170105,
"learning_rate": 7.1482185147433905e-06,
"loss": 1.0272672176361084,
"step": 780
},
{
"epoch": 1.653276955602537,
"grad_norm": 1.2267097234725952,
"learning_rate": 7.133567677835055e-06,
"loss": 0.7762124538421631,
"step": 782
},
{
"epoch": 1.6575052854122623,
"grad_norm": 10.771578788757324,
"learning_rate": 7.11889686990556e-06,
"loss": 1.1939367055892944,
"step": 784
},
{
"epoch": 1.6617336152219875,
"grad_norm": 1.6551384925842285,
"learning_rate": 7.104206270311423e-06,
"loss": 0.5133641958236694,
"step": 786
},
{
"epoch": 1.6659619450317125,
"grad_norm": 1.1092427968978882,
"learning_rate": 7.089496058651133e-06,
"loss": 1.1264673471450806,
"step": 788
},
{
"epoch": 1.6701902748414377,
"grad_norm": 0.737005889415741,
"learning_rate": 7.0747664147629335e-06,
"loss": 0.7446142435073853,
"step": 790
},
{
"epoch": 1.6744186046511627,
"grad_norm": 0.650080144405365,
"learning_rate": 7.060017518722637e-06,
"loss": 0.3989156186580658,
"step": 792
},
{
"epoch": 1.6786469344608879,
"grad_norm": 0.7440600395202637,
"learning_rate": 7.0452495508414276e-06,
"loss": 1.0357478857040405,
"step": 794
},
{
"epoch": 1.682875264270613,
"grad_norm": 0.6777753829956055,
"learning_rate": 7.030462691663641e-06,
"loss": 0.5842429995536804,
"step": 796
},
{
"epoch": 1.6871035940803383,
"grad_norm": 0.8586596846580505,
"learning_rate": 7.015657121964575e-06,
"loss": 1.1753922700881958,
"step": 798
},
{
"epoch": 1.6913319238900635,
"grad_norm": 2.220923662185669,
"learning_rate": 7.000833022748267e-06,
"loss": 1.0605896711349487,
"step": 800
},
{
"epoch": 1.6955602536997887,
"grad_norm": 2.1386070251464844,
"learning_rate": 6.985990575245283e-06,
"loss": 0.5545616745948792,
"step": 802
},
{
"epoch": 1.699788583509514,
"grad_norm": 0.8807734847068787,
"learning_rate": 6.97112996091051e-06,
"loss": 1.037501573562622,
"step": 804
},
{
"epoch": 1.7040169133192389,
"grad_norm": 2.373481512069702,
"learning_rate": 6.956251361420926e-06,
"loss": 0.7858154773712158,
"step": 806
},
{
"epoch": 1.708245243128964,
"grad_norm": 1.4688576459884644,
"learning_rate": 6.94135495867339e-06,
"loss": 0.7295252680778503,
"step": 808
},
{
"epoch": 1.712473572938689,
"grad_norm": 0.8609589338302612,
"learning_rate": 6.926440934782407e-06,
"loss": 1.0780566930770874,
"step": 810
},
{
"epoch": 1.7167019027484143,
"grad_norm": 2.438175916671753,
"learning_rate": 6.9115094720779155e-06,
"loss": 1.2928038835525513,
"step": 812
},
{
"epoch": 1.7209302325581395,
"grad_norm": 1.2204934358596802,
"learning_rate": 6.896560753103041e-06,
"loss": 1.0531013011932373,
"step": 814
},
{
"epoch": 1.7251585623678647,
"grad_norm": 1.0662896633148193,
"learning_rate": 6.881594960611881e-06,
"loss": 1.122856616973877,
"step": 816
},
{
"epoch": 1.72938689217759,
"grad_norm": 1.3619959354400635,
"learning_rate": 6.866612277567261e-06,
"loss": 0.8972553610801697,
"step": 818
},
{
"epoch": 1.733615221987315,
"grad_norm": 6.620126247406006,
"learning_rate": 6.851612887138501e-06,
"loss": 1.0169367790222168,
"step": 820
},
{
"epoch": 1.7378435517970403,
"grad_norm": 1.0335197448730469,
"learning_rate": 6.836596972699175e-06,
"loss": 0.6592488288879395,
"step": 822
},
{
"epoch": 1.7420718816067653,
"grad_norm": 0.4447021484375,
"learning_rate": 6.82156471782487e-06,
"loss": 0.6091707348823547,
"step": 824
},
{
"epoch": 1.7463002114164905,
"grad_norm": 32.608062744140625,
"learning_rate": 6.806516306290942e-06,
"loss": 0.8855757713317871,
"step": 826
},
{
"epoch": 1.7505285412262155,
"grad_norm": 0.9054672718048096,
"learning_rate": 6.791451922070265e-06,
"loss": 0.898021399974823,
"step": 828
},
{
"epoch": 1.7547568710359407,
"grad_norm": 1.9797751903533936,
"learning_rate": 6.77637174933099e-06,
"loss": 0.7365894913673401,
"step": 830
},
{
"epoch": 1.758985200845666,
"grad_norm": 2.273200035095215,
"learning_rate": 6.7612759724342895e-06,
"loss": 1.0903867483139038,
"step": 832
},
{
"epoch": 1.763213530655391,
"grad_norm": 5.753444671630859,
"learning_rate": 6.746164775932096e-06,
"loss": 0.7640259861946106,
"step": 834
},
{
"epoch": 1.7674418604651163,
"grad_norm": 0.8965550661087036,
"learning_rate": 6.731038344564862e-06,
"loss": 0.8017289638519287,
"step": 836
},
{
"epoch": 1.7716701902748415,
"grad_norm": 1.181828498840332,
"learning_rate": 6.715896863259286e-06,
"loss": 1.1074180603027344,
"step": 838
},
{
"epoch": 1.7758985200845667,
"grad_norm": 0.9695769548416138,
"learning_rate": 6.700740517126061e-06,
"loss": 0.832443356513977,
"step": 840
},
{
"epoch": 1.7801268498942917,
"grad_norm": 1.968226671218872,
"learning_rate": 6.685569491457608e-06,
"loss": 0.3589765727519989,
"step": 842
},
{
"epoch": 1.784355179704017,
"grad_norm": 1.0119850635528564,
"learning_rate": 6.670383971725809e-06,
"loss": 0.6536914110183716,
"step": 844
},
{
"epoch": 1.7885835095137421,
"grad_norm": 3.755156993865967,
"learning_rate": 6.655184143579746e-06,
"loss": 0.9794473648071289,
"step": 846
},
{
"epoch": 1.792811839323467,
"grad_norm": 1.3339072465896606,
"learning_rate": 6.639970192843423e-06,
"loss": 0.7290109992027283,
"step": 848
},
{
"epoch": 1.7970401691331923,
"grad_norm": 24.011920928955078,
"learning_rate": 6.624742305513502e-06,
"loss": 0.7500961422920227,
"step": 850
},
{
"epoch": 1.8012684989429175,
"grad_norm": 1.2864532470703125,
"learning_rate": 6.609500667757022e-06,
"loss": 0.9486823081970215,
"step": 852
},
{
"epoch": 1.8054968287526427,
"grad_norm": 0.7961009740829468,
"learning_rate": 6.594245465909127e-06,
"loss": 0.8014146089553833,
"step": 854
},
{
"epoch": 1.809725158562368,
"grad_norm": 1.1502172946929932,
"learning_rate": 6.578976886470791e-06,
"loss": 1.0471374988555908,
"step": 856
},
{
"epoch": 1.8139534883720931,
"grad_norm": 1.030369520187378,
"learning_rate": 6.56369511610653e-06,
"loss": 0.6216817498207092,
"step": 858
},
{
"epoch": 1.8181818181818183,
"grad_norm": 0.8045199513435364,
"learning_rate": 6.548400341642128e-06,
"loss": 0.971676230430603,
"step": 860
},
{
"epoch": 1.8224101479915433,
"grad_norm": 0.9689712524414062,
"learning_rate": 6.533092750062345e-06,
"loss": 1.0254318714141846,
"step": 862
},
{
"epoch": 1.8266384778012685,
"grad_norm": 0.9046330451965332,
"learning_rate": 6.5177725285086416e-06,
"loss": 0.49674558639526367,
"step": 864
},
{
"epoch": 1.8308668076109935,
"grad_norm": 0.7043809294700623,
"learning_rate": 6.5024398642768785e-06,
"loss": 0.7581978440284729,
"step": 866
},
{
"epoch": 1.8350951374207187,
"grad_norm": 0.7090198993682861,
"learning_rate": 6.487094944815036e-06,
"loss": 1.0695147514343262,
"step": 868
},
{
"epoch": 1.839323467230444,
"grad_norm": 0.7204838395118713,
"learning_rate": 6.47173795772092e-06,
"loss": 1.0842490196228027,
"step": 870
},
{
"epoch": 1.8435517970401691,
"grad_norm": 0.5262123346328735,
"learning_rate": 6.456369090739869e-06,
"loss": 0.7069533467292786,
"step": 872
},
{
"epoch": 1.8477801268498943,
"grad_norm": 0.8653817176818848,
"learning_rate": 6.440988531762454e-06,
"loss": 1.0705362558364868,
"step": 874
},
{
"epoch": 1.8520084566596196,
"grad_norm": 3.169382333755493,
"learning_rate": 6.42559646882219e-06,
"loss": 0.9032923579216003,
"step": 876
},
{
"epoch": 1.8562367864693448,
"grad_norm": 1.8638420104980469,
"learning_rate": 6.410193090093227e-06,
"loss": 0.32960423827171326,
"step": 878
},
{
"epoch": 1.8604651162790697,
"grad_norm": 5.069002628326416,
"learning_rate": 6.394778583888064e-06,
"loss": 0.7737627625465393,
"step": 880
},
{
"epoch": 1.864693446088795,
"grad_norm": 1.5018755197525024,
"learning_rate": 6.379353138655227e-06,
"loss": 1.071684718132019,
"step": 882
},
{
"epoch": 1.86892177589852,
"grad_norm": 1.680917501449585,
"learning_rate": 6.3639169429769805e-06,
"loss": 0.7290086150169373,
"step": 884
},
{
"epoch": 1.8731501057082451,
"grad_norm": 1.9813584089279175,
"learning_rate": 6.348470185567025e-06,
"loss": 1.088840126991272,
"step": 886
},
{
"epoch": 1.8773784355179703,
"grad_norm": 0.7224422097206116,
"learning_rate": 6.333013055268169e-06,
"loss": 1.1129541397094727,
"step": 888
},
{
"epoch": 1.8816067653276956,
"grad_norm": 2.9181861877441406,
"learning_rate": 6.317545741050044e-06,
"loss": 0.5802443027496338,
"step": 890
},
{
"epoch": 1.8858350951374208,
"grad_norm": 16.069835662841797,
"learning_rate": 6.3020684320067805e-06,
"loss": 0.7791861295700073,
"step": 892
},
{
"epoch": 1.890063424947146,
"grad_norm": 1.2361421585083008,
"learning_rate": 6.2865813173546976e-06,
"loss": 1.1204155683517456,
"step": 894
},
{
"epoch": 1.8942917547568712,
"grad_norm": 1.4849683046340942,
"learning_rate": 6.271084586429994e-06,
"loss": 1.4051682949066162,
"step": 896
},
{
"epoch": 1.8985200845665962,
"grad_norm": 0.8966804146766663,
"learning_rate": 6.255578428686433e-06,
"loss": 1.0661479234695435,
"step": 898
},
{
"epoch": 1.9027484143763214,
"grad_norm": 4.626205921173096,
"learning_rate": 6.240063033693023e-06,
"loss": 0.49465489387512207,
"step": 900
},
{
"epoch": 1.9069767441860463,
"grad_norm": 1.3005685806274414,
"learning_rate": 6.2245385911316994e-06,
"loss": 1.050133228302002,
"step": 902
},
{
"epoch": 1.9112050739957716,
"grad_norm": 11.314348220825195,
"learning_rate": 6.2090052907950096e-06,
"loss": 0.5165952444076538,
"step": 904
},
{
"epoch": 1.9154334038054968,
"grad_norm": 1.0358664989471436,
"learning_rate": 6.193463322583792e-06,
"loss": 0.8667086362838745,
"step": 906
},
{
"epoch": 1.919661733615222,
"grad_norm": 1.3207107782363892,
"learning_rate": 6.177912876504851e-06,
"loss": 1.175122857093811,
"step": 908
},
{
"epoch": 1.9238900634249472,
"grad_norm": 1.2591102123260498,
"learning_rate": 6.16235414266864e-06,
"loss": 0.7736162543296814,
"step": 910
},
{
"epoch": 1.9281183932346724,
"grad_norm": 0.745734453201294,
"learning_rate": 6.146787311286927e-06,
"loss": 0.9956371188163757,
"step": 912
},
{
"epoch": 1.9323467230443976,
"grad_norm": 7.5462260246276855,
"learning_rate": 6.131212572670483e-06,
"loss": 0.7533048987388611,
"step": 914
},
{
"epoch": 1.9365750528541226,
"grad_norm": 0.9792137145996094,
"learning_rate": 6.115630117226741e-06,
"loss": 1.1065865755081177,
"step": 916
},
{
"epoch": 1.9408033826638478,
"grad_norm": 1.2203872203826904,
"learning_rate": 6.1000401354574855e-06,
"loss": 0.6845130324363708,
"step": 918
},
{
"epoch": 1.945031712473573,
"grad_norm": 3.2082433700561523,
"learning_rate": 6.084442817956502e-06,
"loss": 0.8848755955696106,
"step": 920
},
{
"epoch": 1.949260042283298,
"grad_norm": 1.702277421951294,
"learning_rate": 6.0688383554072625e-06,
"loss": 0.9052953124046326,
"step": 922
},
{
"epoch": 1.9534883720930232,
"grad_norm": 0.7503824830055237,
"learning_rate": 6.053226938580596e-06,
"loss": 0.9683996438980103,
"step": 924
},
{
"epoch": 1.9577167019027484,
"grad_norm": 0.7164794206619263,
"learning_rate": 6.03760875833234e-06,
"loss": 0.8380364179611206,
"step": 926
},
{
"epoch": 1.9619450317124736,
"grad_norm": 0.9604968428611755,
"learning_rate": 6.021984005601027e-06,
"loss": 1.1545363664627075,
"step": 928
},
{
"epoch": 1.9661733615221988,
"grad_norm": 1.1623061895370483,
"learning_rate": 6.006352871405534e-06,
"loss": 0.8529947400093079,
"step": 930
},
{
"epoch": 1.970401691331924,
"grad_norm": 0.8874978423118591,
"learning_rate": 5.990715546842755e-06,
"loss": 1.216421127319336,
"step": 932
},
{
"epoch": 1.9746300211416492,
"grad_norm": 2.1490790843963623,
"learning_rate": 5.9750722230852694e-06,
"loss": 0.8838400840759277,
"step": 934
},
{
"epoch": 1.9788583509513742,
"grad_norm": 0.9013667702674866,
"learning_rate": 5.959423091378992e-06,
"loss": 0.6700854301452637,
"step": 936
},
{
"epoch": 1.9830866807610994,
"grad_norm": 1.8630766868591309,
"learning_rate": 5.943768343040845e-06,
"loss": 0.7751465439796448,
"step": 938
},
{
"epoch": 1.9873150105708244,
"grad_norm": 0.8277192711830139,
"learning_rate": 5.928108169456419e-06,
"loss": 1.0222362279891968,
"step": 940
},
{
"epoch": 1.9915433403805496,
"grad_norm": 1.315276861190796,
"learning_rate": 5.912442762077624e-06,
"loss": 1.053969144821167,
"step": 942
},
{
"epoch": 1.9957716701902748,
"grad_norm": 2.6161696910858154,
"learning_rate": 5.89677231242036e-06,
"loss": 1.0975896120071411,
"step": 944
},
{
"epoch": 2.0,
"grad_norm": 1.3356083631515503,
"learning_rate": 5.8810970120621714e-06,
"loss": 0.4402746260166168,
"step": 946
},
{
"epoch": 2.004228329809725,
"grad_norm": 3.0463569164276123,
"learning_rate": 5.8654170526399014e-06,
"loss": 0.79519122838974,
"step": 948
},
{
"epoch": 2.0084566596194504,
"grad_norm": 0.7577338814735413,
"learning_rate": 5.849732625847354e-06,
"loss": 0.9142102003097534,
"step": 950
},
{
"epoch": 2.0126849894291756,
"grad_norm": 2.497982978820801,
"learning_rate": 5.834043923432947e-06,
"loss": 0.8476440906524658,
"step": 952
},
{
"epoch": 2.016913319238901,
"grad_norm": 2.7633721828460693,
"learning_rate": 5.818351137197373e-06,
"loss": 1.050397515296936,
"step": 954
},
{
"epoch": 2.0211416490486256,
"grad_norm": 2.370466947555542,
"learning_rate": 5.802654458991245e-06,
"loss": 1.0401432514190674,
"step": 956
},
{
"epoch": 2.025369978858351,
"grad_norm": 1.1885958909988403,
"learning_rate": 5.786954080712763e-06,
"loss": 0.6956438422203064,
"step": 958
},
{
"epoch": 2.029598308668076,
"grad_norm": 0.9322492480278015,
"learning_rate": 5.7712501943053576e-06,
"loss": 0.8916679620742798,
"step": 960
},
{
"epoch": 2.033826638477801,
"grad_norm": 1.8816481828689575,
"learning_rate": 5.755542991755352e-06,
"loss": 0.9260307550430298,
"step": 962
},
{
"epoch": 2.0380549682875264,
"grad_norm": 3.0193986892700195,
"learning_rate": 5.739832665089604e-06,
"loss": 0.7821993231773376,
"step": 964
},
{
"epoch": 2.0422832980972516,
"grad_norm": 1.2901606559753418,
"learning_rate": 5.724119406373171e-06,
"loss": 0.63918137550354,
"step": 966
},
{
"epoch": 2.046511627906977,
"grad_norm": 6.436103820800781,
"learning_rate": 5.708403407706954e-06,
"loss": 0.860115647315979,
"step": 968
},
{
"epoch": 2.050739957716702,
"grad_norm": 1.1011855602264404,
"learning_rate": 5.692684861225349e-06,
"loss": 0.939681351184845,
"step": 970
},
{
"epoch": 2.0549682875264272,
"grad_norm": 3.379621982574463,
"learning_rate": 5.676963959093904e-06,
"loss": 0.4519176781177521,
"step": 972
},
{
"epoch": 2.059196617336152,
"grad_norm": 1.3900830745697021,
"learning_rate": 5.66124089350696e-06,
"loss": 0.8939884305000305,
"step": 974
},
{
"epoch": 2.063424947145877,
"grad_norm": 1.4225982427597046,
"learning_rate": 5.645515856685316e-06,
"loss": 0.9207086563110352,
"step": 976
},
{
"epoch": 2.0676532769556024,
"grad_norm": 1.0494661331176758,
"learning_rate": 5.629789040873862e-06,
"loss": 0.930915892124176,
"step": 978
},
{
"epoch": 2.0718816067653276,
"grad_norm": 2.1130642890930176,
"learning_rate": 5.614060638339238e-06,
"loss": 0.7509907484054565,
"step": 980
},
{
"epoch": 2.076109936575053,
"grad_norm": 2.3282723426818848,
"learning_rate": 5.598330841367487e-06,
"loss": 0.7202107310295105,
"step": 982
},
{
"epoch": 2.080338266384778,
"grad_norm": 0.8543728590011597,
"learning_rate": 5.582599842261693e-06,
"loss": 0.6598795652389526,
"step": 984
},
{
"epoch": 2.0845665961945032,
"grad_norm": 1.1249150037765503,
"learning_rate": 5.566867833339644e-06,
"loss": 0.7731543183326721,
"step": 986
},
{
"epoch": 2.0887949260042284,
"grad_norm": 0.9826174378395081,
"learning_rate": 5.551135006931467e-06,
"loss": 0.8880943655967712,
"step": 988
},
{
"epoch": 2.0930232558139537,
"grad_norm": 1.3232882022857666,
"learning_rate": 5.535401555377287e-06,
"loss": 0.8624506592750549,
"step": 990
},
{
"epoch": 2.097251585623679,
"grad_norm": 3.0641980171203613,
"learning_rate": 5.51966767102487e-06,
"loss": 0.5808882713317871,
"step": 992
},
{
"epoch": 2.1014799154334036,
"grad_norm": 5.181814193725586,
"learning_rate": 5.503933546227273e-06,
"loss": 0.6591153740882874,
"step": 994
},
{
"epoch": 2.105708245243129,
"grad_norm": 0.8502072095870972,
"learning_rate": 5.4881993733404916e-06,
"loss": 0.6375114917755127,
"step": 996
},
{
"epoch": 2.109936575052854,
"grad_norm": 0.9985566735267639,
"learning_rate": 5.472465344721113e-06,
"loss": 0.896757185459137,
"step": 998
},
{
"epoch": 2.1141649048625792,
"grad_norm": 1.097866177558899,
"learning_rate": 5.456731652723959e-06,
"loss": 0.8528929352760315,
"step": 1000
},
{
"epoch": 2.1183932346723044,
"grad_norm": 0.9444407820701599,
"learning_rate": 5.440998489699731e-06,
"loss": 0.2696181535720825,
"step": 1002
},
{
"epoch": 2.1226215644820297,
"grad_norm": 3.740973949432373,
"learning_rate": 5.425266047992672e-06,
"loss": 0.6543110609054565,
"step": 1004
},
{
"epoch": 2.126849894291755,
"grad_norm": 0.9048804640769958,
"learning_rate": 5.4095345199382e-06,
"loss": 0.9101112484931946,
"step": 1006
},
{
"epoch": 2.13107822410148,
"grad_norm": 2.3448641300201416,
"learning_rate": 5.393804097860568e-06,
"loss": 0.847807765007019,
"step": 1008
},
{
"epoch": 2.1353065539112053,
"grad_norm": 0.8529036045074463,
"learning_rate": 5.3780749740705065e-06,
"loss": 0.5853134393692017,
"step": 1010
},
{
"epoch": 2.13953488372093,
"grad_norm": 1.4966481924057007,
"learning_rate": 5.362347340862869e-06,
"loss": 0.49149683117866516,
"step": 1012
},
{
"epoch": 2.1437632135306552,
"grad_norm": 0.9276267290115356,
"learning_rate": 5.3466213905142935e-06,
"loss": 0.541230320930481,
"step": 1014
},
{
"epoch": 2.1479915433403804,
"grad_norm": 2.2109150886535645,
"learning_rate": 5.3308973152808395e-06,
"loss": 0.7555081844329834,
"step": 1016
},
{
"epoch": 2.1522198731501057,
"grad_norm": 2.556731700897217,
"learning_rate": 5.315175307395644e-06,
"loss": 0.8441729545593262,
"step": 1018
},
{
"epoch": 2.156448202959831,
"grad_norm": 2.1739888191223145,
"learning_rate": 5.299455559066571e-06,
"loss": 0.6941192150115967,
"step": 1020
},
{
"epoch": 2.160676532769556,
"grad_norm": 1.1940555572509766,
"learning_rate": 5.283738262473859e-06,
"loss": 0.6465023756027222,
"step": 1022
},
{
"epoch": 2.1649048625792813,
"grad_norm": 3.0477652549743652,
"learning_rate": 5.268023609767772e-06,
"loss": 0.8692419528961182,
"step": 1024
},
{
"epoch": 2.1691331923890065,
"grad_norm": 1.2980542182922363,
"learning_rate": 5.2523117930662545e-06,
"loss": 0.9159978032112122,
"step": 1026
},
{
"epoch": 2.1733615221987317,
"grad_norm": 2.0734403133392334,
"learning_rate": 5.2366030044525785e-06,
"loss": 0.9289013743400574,
"step": 1028
},
{
"epoch": 2.177589852008457,
"grad_norm": 1.2459543943405151,
"learning_rate": 5.220897435972992e-06,
"loss": 0.8755097389221191,
"step": 1030
},
{
"epoch": 2.1818181818181817,
"grad_norm": 1.324737310409546,
"learning_rate": 5.205195279634385e-06,
"loss": 0.9681872129440308,
"step": 1032
},
{
"epoch": 2.186046511627907,
"grad_norm": 1.166222333908081,
"learning_rate": 5.1894967274019245e-06,
"loss": 0.9441096186637878,
"step": 1034
},
{
"epoch": 2.190274841437632,
"grad_norm": 1.2177082300186157,
"learning_rate": 5.173801971196716e-06,
"loss": 0.6271236538887024,
"step": 1036
},
{
"epoch": 2.1945031712473573,
"grad_norm": 2.871882438659668,
"learning_rate": 5.158111202893464e-06,
"loss": 0.2950773537158966,
"step": 1038
},
{
"epoch": 2.1987315010570825,
"grad_norm": 1.143522024154663,
"learning_rate": 5.142424614318112e-06,
"loss": 0.938387393951416,
"step": 1040
},
{
"epoch": 2.2029598308668077,
"grad_norm": 48.49791717529297,
"learning_rate": 5.12674239724551e-06,
"loss": 0.34457671642303467,
"step": 1042
},
{
"epoch": 2.207188160676533,
"grad_norm": 0.733287513256073,
"learning_rate": 5.111064743397061e-06,
"loss": 0.9282559156417847,
"step": 1044
},
{
"epoch": 2.211416490486258,
"grad_norm": 1.1249736547470093,
"learning_rate": 5.09539184443838e-06,
"loss": 0.4425432085990906,
"step": 1046
},
{
"epoch": 2.2156448202959833,
"grad_norm": 2.6254959106445312,
"learning_rate": 5.079723891976956e-06,
"loss": 0.7763065099716187,
"step": 1048
},
{
"epoch": 2.219873150105708,
"grad_norm": 25.987655639648438,
"learning_rate": 5.064061077559803e-06,
"loss": 0.6502024531364441,
"step": 1050
},
{
"epoch": 2.2241014799154333,
"grad_norm": 4.657992839813232,
"learning_rate": 5.048403592671116e-06,
"loss": 0.8349530100822449,
"step": 1052
},
{
"epoch": 2.2283298097251585,
"grad_norm": 1.2359461784362793,
"learning_rate": 5.032751628729943e-06,
"loss": 0.9067742228507996,
"step": 1054
},
{
"epoch": 2.2325581395348837,
"grad_norm": 0.6538547277450562,
"learning_rate": 5.017105377087829e-06,
"loss": 0.9318907856941223,
"step": 1056
},
{
"epoch": 2.236786469344609,
"grad_norm": 0.634844183921814,
"learning_rate": 5.0014650290264854e-06,
"loss": 0.5932443141937256,
"step": 1058
},
{
"epoch": 2.241014799154334,
"grad_norm": 1.2073909044265747,
"learning_rate": 4.9858307757554536e-06,
"loss": 0.7449318170547485,
"step": 1060
},
{
"epoch": 2.2452431289640593,
"grad_norm": 0.8729775547981262,
"learning_rate": 4.97020280840976e-06,
"loss": 0.8536829948425293,
"step": 1062
},
{
"epoch": 2.2494714587737845,
"grad_norm": 2.562119245529175,
"learning_rate": 4.954581318047582e-06,
"loss": 0.8429497480392456,
"step": 1064
},
{
"epoch": 2.2536997885835097,
"grad_norm": 1.2921231985092163,
"learning_rate": 4.938966495647918e-06,
"loss": 0.4837799668312073,
"step": 1066
},
{
"epoch": 2.2579281183932345,
"grad_norm": 1.5965522527694702,
"learning_rate": 4.923358532108244e-06,
"loss": 0.7402957081794739,
"step": 1068
},
{
"epoch": 2.2621564482029597,
"grad_norm": 15.080484390258789,
"learning_rate": 4.907757618242186e-06,
"loss": 0.7502317428588867,
"step": 1070
},
{
"epoch": 2.266384778012685,
"grad_norm": 1.1392362117767334,
"learning_rate": 4.892163944777181e-06,
"loss": 0.9198048114776611,
"step": 1072
},
{
"epoch": 2.27061310782241,
"grad_norm": 0.45098453760147095,
"learning_rate": 4.876577702352155e-06,
"loss": 0.6175814270973206,
"step": 1074
},
{
"epoch": 2.2748414376321353,
"grad_norm": 4.194945812225342,
"learning_rate": 4.860999081515183e-06,
"loss": 0.7774248123168945,
"step": 1076
},
{
"epoch": 2.2790697674418605,
"grad_norm": 4.153860569000244,
"learning_rate": 4.845428272721161e-06,
"loss": 0.2229873090982437,
"step": 1078
},
{
"epoch": 2.2832980972515857,
"grad_norm": 0.6728289723396301,
"learning_rate": 4.8298654663294865e-06,
"loss": 0.1489698737859726,
"step": 1080
},
{
"epoch": 2.287526427061311,
"grad_norm": 1.4065691232681274,
"learning_rate": 4.814310852601717e-06,
"loss": 0.48190781474113464,
"step": 1082
},
{
"epoch": 2.291754756871036,
"grad_norm": 0.9727882146835327,
"learning_rate": 4.7987646216992545e-06,
"loss": 0.32167869806289673,
"step": 1084
},
{
"epoch": 2.295983086680761,
"grad_norm": 0.9208230376243591,
"learning_rate": 4.783226963681018e-06,
"loss": 0.14745894074440002,
"step": 1086
},
{
"epoch": 2.300211416490486,
"grad_norm": 5.599506855010986,
"learning_rate": 4.76769806850112e-06,
"loss": 0.8546797633171082,
"step": 1088
},
{
"epoch": 2.3044397463002113,
"grad_norm": 1.1775867938995361,
"learning_rate": 4.752178126006544e-06,
"loss": 0.8449016213417053,
"step": 1090
},
{
"epoch": 2.3086680761099365,
"grad_norm": 4.704752445220947,
"learning_rate": 4.736667325934819e-06,
"loss": 0.5860316753387451,
"step": 1092
},
{
"epoch": 2.3128964059196617,
"grad_norm": 1.0056744813919067,
"learning_rate": 4.721165857911709e-06,
"loss": 0.8732479214668274,
"step": 1094
},
{
"epoch": 2.317124735729387,
"grad_norm": 1.3367846012115479,
"learning_rate": 4.705673911448891e-06,
"loss": 0.8070216178894043,
"step": 1096
},
{
"epoch": 2.321353065539112,
"grad_norm": 0.9523402452468872,
"learning_rate": 4.690191675941631e-06,
"loss": 0.8223982453346252,
"step": 1098
},
{
"epoch": 2.3255813953488373,
"grad_norm": 6.13200569152832,
"learning_rate": 4.674719340666477e-06,
"loss": 0.5391439199447632,
"step": 1100
},
{
"epoch": 2.3298097251585626,
"grad_norm": 1.171050786972046,
"learning_rate": 4.659257094778949e-06,
"loss": 0.6862545013427734,
"step": 1102
},
{
"epoch": 2.3340380549682873,
"grad_norm": 1.9349039793014526,
"learning_rate": 4.643805127311212e-06,
"loss": 0.9349678754806519,
"step": 1104
},
{
"epoch": 2.3382663847780125,
"grad_norm": 1.0332423448562622,
"learning_rate": 4.628363627169774e-06,
"loss": 0.8068013787269592,
"step": 1106
},
{
"epoch": 2.3424947145877377,
"grad_norm": 1.2433507442474365,
"learning_rate": 4.612932783133182e-06,
"loss": 0.8333487510681152,
"step": 1108
},
{
"epoch": 2.346723044397463,
"grad_norm": 1.800581455230713,
"learning_rate": 4.5975127838497036e-06,
"loss": 0.981497585773468,
"step": 1110
},
{
"epoch": 2.350951374207188,
"grad_norm": 0.3760753870010376,
"learning_rate": 4.582103817835021e-06,
"loss": 0.5332903265953064,
"step": 1112
},
{
"epoch": 2.3551797040169133,
"grad_norm": 2.77357816696167,
"learning_rate": 4.566706073469938e-06,
"loss": 0.7114543318748474,
"step": 1114
},
{
"epoch": 2.3594080338266386,
"grad_norm": 1.0361895561218262,
"learning_rate": 4.551319738998067e-06,
"loss": 0.5817387104034424,
"step": 1116
},
{
"epoch": 2.3636363636363638,
"grad_norm": 2.930225133895874,
"learning_rate": 4.535945002523524e-06,
"loss": 1.0342414379119873,
"step": 1118
},
{
"epoch": 2.367864693446089,
"grad_norm": 1.0692269802093506,
"learning_rate": 4.5205820520086415e-06,
"loss": 0.8719377517700195,
"step": 1120
},
{
"epoch": 2.3720930232558137,
"grad_norm": 1.1443272829055786,
"learning_rate": 4.505231075271662e-06,
"loss": 0.8489775061607361,
"step": 1122
},
{
"epoch": 2.376321353065539,
"grad_norm": 1.1029630899429321,
"learning_rate": 4.489892259984446e-06,
"loss": 0.6426911950111389,
"step": 1124
},
{
"epoch": 2.380549682875264,
"grad_norm": 2.715700149536133,
"learning_rate": 4.474565793670167e-06,
"loss": 0.8776054382324219,
"step": 1126
},
{
"epoch": 2.3847780126849893,
"grad_norm": 1.622450351715088,
"learning_rate": 4.459251863701038e-06,
"loss": 0.6117690801620483,
"step": 1128
},
{
"epoch": 2.3890063424947146,
"grad_norm": 1.326261281967163,
"learning_rate": 4.443950657296004e-06,
"loss": 0.6771716475486755,
"step": 1130
},
{
"epoch": 2.3932346723044398,
"grad_norm": 2.561893939971924,
"learning_rate": 4.4286623615184585e-06,
"loss": 0.5820841789245605,
"step": 1132
},
{
"epoch": 2.397463002114165,
"grad_norm": 1.1955907344818115,
"learning_rate": 4.413387163273964e-06,
"loss": 0.9275699853897095,
"step": 1134
},
{
"epoch": 2.40169133192389,
"grad_norm": 3.236022710800171,
"learning_rate": 4.3981252493079506e-06,
"loss": 0.6621242761611938,
"step": 1136
},
{
"epoch": 2.4059196617336154,
"grad_norm": 1.612190842628479,
"learning_rate": 4.382876806203453e-06,
"loss": 0.7824278473854065,
"step": 1138
},
{
"epoch": 2.41014799154334,
"grad_norm": 1.1192189455032349,
"learning_rate": 4.3676420203788125e-06,
"loss": 0.8992093801498413,
"step": 1140
},
{
"epoch": 2.4143763213530653,
"grad_norm": 1.2494477033615112,
"learning_rate": 4.352421078085405e-06,
"loss": 0.9680632948875427,
"step": 1142
},
{
"epoch": 2.4186046511627906,
"grad_norm": 1.7657294273376465,
"learning_rate": 4.337214165405368e-06,
"loss": 0.4477839767932892,
"step": 1144
},
{
"epoch": 2.4228329809725158,
"grad_norm": 3.7850959300994873,
"learning_rate": 4.322021468249316e-06,
"loss": 0.6704675555229187,
"step": 1146
},
{
"epoch": 2.427061310782241,
"grad_norm": 1.015831470489502,
"learning_rate": 4.306843172354076e-06,
"loss": 0.6930139660835266,
"step": 1148
},
{
"epoch": 2.431289640591966,
"grad_norm": 2.0889852046966553,
"learning_rate": 4.291679463280416e-06,
"loss": 0.46937960386276245,
"step": 1150
},
{
"epoch": 2.4355179704016914,
"grad_norm": 0.3903566896915436,
"learning_rate": 4.276530526410768e-06,
"loss": 0.5279465317726135,
"step": 1152
},
{
"epoch": 2.4397463002114166,
"grad_norm": 0.8405472636222839,
"learning_rate": 4.261396546946972e-06,
"loss": 0.6472615599632263,
"step": 1154
},
{
"epoch": 2.443974630021142,
"grad_norm": 2.189234733581543,
"learning_rate": 4.246277709908009e-06,
"loss": 0.8786381483078003,
"step": 1156
},
{
"epoch": 2.448202959830867,
"grad_norm": 1.407827615737915,
"learning_rate": 4.231174200127736e-06,
"loss": 0.4858361780643463,
"step": 1158
},
{
"epoch": 2.452431289640592,
"grad_norm": 0.9727794528007507,
"learning_rate": 4.216086202252624e-06,
"loss": 0.9363426566123962,
"step": 1160
},
{
"epoch": 2.456659619450317,
"grad_norm": 1.078275203704834,
"learning_rate": 4.201013900739513e-06,
"loss": 0.888515830039978,
"step": 1162
},
{
"epoch": 2.460887949260042,
"grad_norm": 1.4499064683914185,
"learning_rate": 4.185957479853344e-06,
"loss": 0.6493883728981018,
"step": 1164
},
{
"epoch": 2.4651162790697674,
"grad_norm": 1.943841576576233,
"learning_rate": 4.1709171236649096e-06,
"loss": 1.0317200422286987,
"step": 1166
},
{
"epoch": 2.4693446088794926,
"grad_norm": 1.0731745958328247,
"learning_rate": 4.155893016048607e-06,
"loss": 1.0403978824615479,
"step": 1168
},
{
"epoch": 2.473572938689218,
"grad_norm": 1.503381609916687,
"learning_rate": 4.140885340680189e-06,
"loss": 0.6530440449714661,
"step": 1170
},
{
"epoch": 2.477801268498943,
"grad_norm": 36.47575378417969,
"learning_rate": 4.125894281034517e-06,
"loss": 0.7441340684890747,
"step": 1172
},
{
"epoch": 2.482029598308668,
"grad_norm": 0.9915839433670044,
"learning_rate": 4.110920020383316e-06,
"loss": 0.662432074546814,
"step": 1174
},
{
"epoch": 2.4862579281183934,
"grad_norm": 1.4273513555526733,
"learning_rate": 4.0959627417929386e-06,
"loss": 0.6897416710853577,
"step": 1176
},
{
"epoch": 2.4904862579281186,
"grad_norm": 1.7252953052520752,
"learning_rate": 4.081022628122127e-06,
"loss": 0.6987327337265015,
"step": 1178
},
{
"epoch": 2.4947145877378434,
"grad_norm": 3.7639214992523193,
"learning_rate": 4.06609986201977e-06,
"loss": 0.5576756000518799,
"step": 1180
},
{
"epoch": 2.4989429175475686,
"grad_norm": 1.0913081169128418,
"learning_rate": 4.051194625922681e-06,
"loss": 0.35905197262763977,
"step": 1182
},
{
"epoch": 2.503171247357294,
"grad_norm": 1.1467711925506592,
"learning_rate": 4.036307102053359e-06,
"loss": 0.11249269545078278,
"step": 1184
},
{
"epoch": 2.507399577167019,
"grad_norm": 1.0933419466018677,
"learning_rate": 4.021437472417763e-06,
"loss": 0.9699214100837708,
"step": 1186
},
{
"epoch": 2.511627906976744,
"grad_norm": 2.807231903076172,
"learning_rate": 4.006585918803091e-06,
"loss": 0.7549931406974792,
"step": 1188
},
{
"epoch": 2.5158562367864694,
"grad_norm": 4.363275051116943,
"learning_rate": 3.991752622775551e-06,
"loss": 0.7365719079971313,
"step": 1190
},
{
"epoch": 2.5200845665961946,
"grad_norm": 1.0322295427322388,
"learning_rate": 3.97693776567815e-06,
"loss": 0.8650130033493042,
"step": 1192
},
{
"epoch": 2.52431289640592,
"grad_norm": 0.6582368612289429,
"learning_rate": 3.962141528628465e-06,
"loss": 0.5743870139122009,
"step": 1194
},
{
"epoch": 2.528541226215645,
"grad_norm": 2.5764482021331787,
"learning_rate": 3.947364092516437e-06,
"loss": 0.8143431544303894,
"step": 1196
},
{
"epoch": 2.53276955602537,
"grad_norm": 1.182889699935913,
"learning_rate": 3.932605638002163e-06,
"loss": 0.8320148587226868,
"step": 1198
},
{
"epoch": 2.536997885835095,
"grad_norm": 2.314754009246826,
"learning_rate": 3.917866345513677e-06,
"loss": 0.6690659523010254,
"step": 1200
},
{
"epoch": 2.54122621564482,
"grad_norm": 1.5246949195861816,
"learning_rate": 3.903146395244749e-06,
"loss": 0.9002770185470581,
"step": 1202
},
{
"epoch": 2.5454545454545454,
"grad_norm": 2.0140490531921387,
"learning_rate": 3.888445967152691e-06,
"loss": 0.3049054741859436,
"step": 1204
},
{
"epoch": 2.5496828752642706,
"grad_norm": 1.0781471729278564,
"learning_rate": 3.873765240956138e-06,
"loss": 0.9467779397964478,
"step": 1206
},
{
"epoch": 2.553911205073996,
"grad_norm": 1.1137040853500366,
"learning_rate": 3.859104396132863e-06,
"loss": 0.9237363934516907,
"step": 1208
},
{
"epoch": 2.558139534883721,
"grad_norm": 0.9697665572166443,
"learning_rate": 3.84446361191759e-06,
"loss": 0.9140211939811707,
"step": 1210
},
{
"epoch": 2.5623678646934462,
"grad_norm": 2.0526888370513916,
"learning_rate": 3.829843067299788e-06,
"loss": 0.6337721347808838,
"step": 1212
},
{
"epoch": 2.5665961945031714,
"grad_norm": 1.277137041091919,
"learning_rate": 3.8152429410214834e-06,
"loss": 0.8417494297027588,
"step": 1214
},
{
"epoch": 2.570824524312896,
"grad_norm": 3.570134162902832,
"learning_rate": 3.8006634115750886e-06,
"loss": 0.5440153479576111,
"step": 1216
},
{
"epoch": 2.5750528541226214,
"grad_norm": 3.0272088050842285,
"learning_rate": 3.7861046572012105e-06,
"loss": 0.5331149697303772,
"step": 1218
},
{
"epoch": 2.5792811839323466,
"grad_norm": 2.8399817943573,
"learning_rate": 3.771566855886467e-06,
"loss": 0.47078245878219604,
"step": 1220
},
{
"epoch": 2.583509513742072,
"grad_norm": 1.5652203559875488,
"learning_rate": 3.7570501853613227e-06,
"loss": 0.49988022446632385,
"step": 1222
},
{
"epoch": 2.587737843551797,
"grad_norm": 2.8541760444641113,
"learning_rate": 3.742554823097905e-06,
"loss": 0.7600050568580627,
"step": 1224
},
{
"epoch": 2.5919661733615222,
"grad_norm": 0.38807129859924316,
"learning_rate": 3.7280809463078436e-06,
"loss": 0.35192567110061646,
"step": 1226
},
{
"epoch": 2.5961945031712474,
"grad_norm": 2.45103120803833,
"learning_rate": 3.7136287319400944e-06,
"loss": 0.6349934935569763,
"step": 1228
},
{
"epoch": 2.6004228329809727,
"grad_norm": 2.1959896087646484,
"learning_rate": 3.6991983566787847e-06,
"loss": 0.7630847692489624,
"step": 1230
},
{
"epoch": 2.604651162790698,
"grad_norm": 4.74827241897583,
"learning_rate": 3.6847899969410515e-06,
"loss": 0.6456127762794495,
"step": 1232
},
{
"epoch": 2.6088794926004226,
"grad_norm": 2.280721426010132,
"learning_rate": 3.6704038288748812e-06,
"loss": 0.6520678400993347,
"step": 1234
},
{
"epoch": 2.613107822410148,
"grad_norm": 1.1855796575546265,
"learning_rate": 3.6560400283569543e-06,
"loss": 0.7459301948547363,
"step": 1236
},
{
"epoch": 2.617336152219873,
"grad_norm": 1.5799105167388916,
"learning_rate": 3.641698770990507e-06,
"loss": 0.9225113391876221,
"step": 1238
},
{
"epoch": 2.6215644820295982,
"grad_norm": 9.281462669372559,
"learning_rate": 3.6273802321031738e-06,
"loss": 0.6870522499084473,
"step": 1240
},
{
"epoch": 2.6257928118393234,
"grad_norm": 1.215850830078125,
"learning_rate": 3.6130845867448384e-06,
"loss": 0.768366277217865,
"step": 1242
},
{
"epoch": 2.6300211416490487,
"grad_norm": 1.9531065225601196,
"learning_rate": 3.598812009685514e-06,
"loss": 0.8747004866600037,
"step": 1244
},
{
"epoch": 2.634249471458774,
"grad_norm": 3.6630496978759766,
"learning_rate": 3.584562675413189e-06,
"loss": 0.5971186757087708,
"step": 1246
},
{
"epoch": 2.638477801268499,
"grad_norm": 1.4509460926055908,
"learning_rate": 3.5703367581316985e-06,
"loss": 0.39166009426116943,
"step": 1248
},
{
"epoch": 2.6427061310782243,
"grad_norm": 1.0395610332489014,
"learning_rate": 3.5561344317585957e-06,
"loss": 0.9035303592681885,
"step": 1250
},
{
"epoch": 2.646934460887949,
"grad_norm": 3.8952317237854004,
"learning_rate": 3.5419558699230305e-06,
"loss": 0.4824777841567993,
"step": 1252
},
{
"epoch": 2.6511627906976747,
"grad_norm": 4.396440029144287,
"learning_rate": 3.52780124596361e-06,
"loss": 0.5573374032974243,
"step": 1254
},
{
"epoch": 2.6553911205073994,
"grad_norm": 0.675754964351654,
"learning_rate": 3.5136707329263055e-06,
"loss": 0.8686530590057373,
"step": 1256
},
{
"epoch": 2.6596194503171247,
"grad_norm": 1.2980667352676392,
"learning_rate": 3.499564503562315e-06,
"loss": 0.27861249446868896,
"step": 1258
},
{
"epoch": 2.66384778012685,
"grad_norm": 0.6087518930435181,
"learning_rate": 3.4854827303259575e-06,
"loss": 0.4511662721633911,
"step": 1260
},
{
"epoch": 2.668076109936575,
"grad_norm": 3.3265137672424316,
"learning_rate": 3.47142558537257e-06,
"loss": 0.6650041341781616,
"step": 1262
},
{
"epoch": 2.6723044397463003,
"grad_norm": 2.1935572624206543,
"learning_rate": 3.4573932405563966e-06,
"loss": 0.5863999128341675,
"step": 1264
},
{
"epoch": 2.6765327695560255,
"grad_norm": 3.012540340423584,
"learning_rate": 3.4433858674284894e-06,
"loss": 0.2797367572784424,
"step": 1266
},
{
"epoch": 2.6807610993657507,
"grad_norm": 0.9022391438484192,
"learning_rate": 3.429403637234614e-06,
"loss": 0.7000196576118469,
"step": 1268
},
{
"epoch": 2.6849894291754755,
"grad_norm": 2.1052815914154053,
"learning_rate": 3.4154467209131515e-06,
"loss": 0.8440768122673035,
"step": 1270
},
{
"epoch": 2.689217758985201,
"grad_norm": 4.980707168579102,
"learning_rate": 3.401515289093011e-06,
"loss": 0.6595039963722229,
"step": 1272
},
{
"epoch": 2.693446088794926,
"grad_norm": 0.3431156277656555,
"learning_rate": 3.3876095120915433e-06,
"loss": 0.40372785925865173,
"step": 1274
},
{
"epoch": 2.697674418604651,
"grad_norm": 3.0530407428741455,
"learning_rate": 3.3737295599124592e-06,
"loss": 0.44936662912368774,
"step": 1276
},
{
"epoch": 2.7019027484143763,
"grad_norm": 1.2963279485702515,
"learning_rate": 3.3598756022437507e-06,
"loss": 0.8797158598899841,
"step": 1278
},
{
"epoch": 2.7061310782241015,
"grad_norm": 2.556929349899292,
"learning_rate": 3.3460478084556157e-06,
"loss": 0.5719730257987976,
"step": 1280
},
{
"epoch": 2.7103594080338267,
"grad_norm": 1.463066816329956,
"learning_rate": 3.33224634759839e-06,
"loss": 0.9499182105064392,
"step": 1282
},
{
"epoch": 2.714587737843552,
"grad_norm": 0.4782409369945526,
"learning_rate": 3.318471388400476e-06,
"loss": 0.6272532939910889,
"step": 1284
},
{
"epoch": 2.718816067653277,
"grad_norm": 0.25632527470588684,
"learning_rate": 3.3047230992662834e-06,
"loss": 0.9857101440429688,
"step": 1286
},
{
"epoch": 2.723044397463002,
"grad_norm": 0.8656527996063232,
"learning_rate": 3.291001648274175e-06,
"loss": 0.44188177585601807,
"step": 1288
},
{
"epoch": 2.7272727272727275,
"grad_norm": 0.9704158306121826,
"learning_rate": 3.277307203174395e-06,
"loss": 0.8773584365844727,
"step": 1290
},
{
"epoch": 2.7315010570824523,
"grad_norm": 1.1535475254058838,
"learning_rate": 3.263639931387039e-06,
"loss": 0.9490036964416504,
"step": 1292
},
{
"epoch": 2.7357293868921775,
"grad_norm": 1.4720072746276855,
"learning_rate": 3.2500000000000015e-06,
"loss": 0.8431513905525208,
"step": 1294
},
{
"epoch": 2.7399577167019027,
"grad_norm": 1.7266931533813477,
"learning_rate": 3.236387575766916e-06,
"loss": 0.2813695967197418,
"step": 1296
},
{
"epoch": 2.744186046511628,
"grad_norm": 1.2458494901657104,
"learning_rate": 3.222802825105138e-06,
"loss": 0.5231180787086487,
"step": 1298
},
{
"epoch": 2.748414376321353,
"grad_norm": 0.9862059354782104,
"learning_rate": 3.209245914093708e-06,
"loss": 0.6971995830535889,
"step": 1300
},
{
"epoch": 2.7526427061310783,
"grad_norm": 1.3058788776397705,
"learning_rate": 3.1957170084713053e-06,
"loss": 0.7326510548591614,
"step": 1302
},
{
"epoch": 2.7568710359408035,
"grad_norm": 0.609324038028717,
"learning_rate": 3.182216273634235e-06,
"loss": 0.5214690566062927,
"step": 1304
},
{
"epoch": 2.7610993657505283,
"grad_norm": 6.1062798500061035,
"learning_rate": 3.1687438746344086e-06,
"loss": 0.667678952217102,
"step": 1306
},
{
"epoch": 2.765327695560254,
"grad_norm": 1.509377121925354,
"learning_rate": 3.1552999761773166e-06,
"loss": 0.7728456854820251,
"step": 1308
},
{
"epoch": 2.7695560253699787,
"grad_norm": 1.2849303483963013,
"learning_rate": 3.141884742620013e-06,
"loss": 0.8508955836296082,
"step": 1310
},
{
"epoch": 2.773784355179704,
"grad_norm": 1.50644052028656,
"learning_rate": 3.1284983379691247e-06,
"loss": 0.8556861281394958,
"step": 1312
},
{
"epoch": 2.778012684989429,
"grad_norm": 3.093153238296509,
"learning_rate": 3.1151409258788235e-06,
"loss": 0.4471794366836548,
"step": 1314
},
{
"epoch": 2.7822410147991543,
"grad_norm": 1.4079885482788086,
"learning_rate": 3.101812669648838e-06,
"loss": 0.8230998516082764,
"step": 1316
},
{
"epoch": 2.7864693446088795,
"grad_norm": 0.3594854176044464,
"learning_rate": 3.0885137322224546e-06,
"loss": 0.5559835433959961,
"step": 1318
},
{
"epoch": 2.7906976744186047,
"grad_norm": 6.585972785949707,
"learning_rate": 3.075244276184529e-06,
"loss": 0.8112860321998596,
"step": 1320
},
{
"epoch": 2.79492600422833,
"grad_norm": 2.1540377140045166,
"learning_rate": 3.0620044637594936e-06,
"loss": 0.3771233856678009,
"step": 1322
},
{
"epoch": 2.7991543340380547,
"grad_norm": 1.568415641784668,
"learning_rate": 3.0487944568093685e-06,
"loss": 1.0688905715942383,
"step": 1324
},
{
"epoch": 2.8033826638477803,
"grad_norm": 3.4723870754241943,
"learning_rate": 3.035614416831801e-06,
"loss": 0.20393051207065582,
"step": 1326
},
{
"epoch": 2.807610993657505,
"grad_norm": 2.7313528060913086,
"learning_rate": 3.0224645049580746e-06,
"loss": 0.3667452037334442,
"step": 1328
},
{
"epoch": 2.8118393234672303,
"grad_norm": 0.9869865775108337,
"learning_rate": 3.0093448819511416e-06,
"loss": 0.5145280361175537,
"step": 1330
},
{
"epoch": 2.8160676532769555,
"grad_norm": 0.985115647315979,
"learning_rate": 2.9962557082036677e-06,
"loss": 0.8460598587989807,
"step": 1332
},
{
"epoch": 2.8202959830866807,
"grad_norm": 0.8554550409317017,
"learning_rate": 2.9831971437360606e-06,
"loss": 0.6342540383338928,
"step": 1334
},
{
"epoch": 2.824524312896406,
"grad_norm": 0.9474649429321289,
"learning_rate": 2.9701693481945132e-06,
"loss": 0.9361258149147034,
"step": 1336
},
{
"epoch": 2.828752642706131,
"grad_norm": 1.0687966346740723,
"learning_rate": 2.957172480849064e-06,
"loss": 1.0203008651733398,
"step": 1338
},
{
"epoch": 2.8329809725158563,
"grad_norm": 0.9894439578056335,
"learning_rate": 2.9442067005916375e-06,
"loss": 0.6205085515975952,
"step": 1340
},
{
"epoch": 2.8372093023255816,
"grad_norm": 1.2212331295013428,
"learning_rate": 2.931272165934105e-06,
"loss": 0.9145471453666687,
"step": 1342
},
{
"epoch": 2.8414376321353068,
"grad_norm": 1.4753201007843018,
"learning_rate": 2.918369035006351e-06,
"loss": 0.6152521371841431,
"step": 1344
},
{
"epoch": 2.8456659619450315,
"grad_norm": 2.1296560764312744,
"learning_rate": 2.9054974655543335e-06,
"loss": 0.5806496143341064,
"step": 1346
},
{
"epoch": 2.8498942917547567,
"grad_norm": 0.9690008163452148,
"learning_rate": 2.89265761493816e-06,
"loss": 0.8748153448104858,
"step": 1348
},
{
"epoch": 2.854122621564482,
"grad_norm": 3.9232616424560547,
"learning_rate": 2.8798496401301644e-06,
"loss": 0.5031945109367371,
"step": 1350
},
{
"epoch": 2.858350951374207,
"grad_norm": 1.3148808479309082,
"learning_rate": 2.8670736977129843e-06,
"loss": 0.7939937114715576,
"step": 1352
},
{
"epoch": 2.8625792811839323,
"grad_norm": 0.8906443119049072,
"learning_rate": 2.854329943877647e-06,
"loss": 0.8754093050956726,
"step": 1354
},
{
"epoch": 2.8668076109936576,
"grad_norm": 10.015877723693848,
"learning_rate": 2.841618534421664e-06,
"loss": 0.4468953609466553,
"step": 1356
},
{
"epoch": 2.8710359408033828,
"grad_norm": 0.9385186433792114,
"learning_rate": 2.8289396247471227e-06,
"loss": 0.8845570683479309,
"step": 1358
},
{
"epoch": 2.875264270613108,
"grad_norm": 1.8789618015289307,
"learning_rate": 2.816293369858788e-06,
"loss": 0.6173380017280579,
"step": 1360
},
{
"epoch": 2.879492600422833,
"grad_norm": 1.8747756481170654,
"learning_rate": 2.803679924362206e-06,
"loss": 0.6125428080558777,
"step": 1362
},
{
"epoch": 2.883720930232558,
"grad_norm": 0.895305871963501,
"learning_rate": 2.7910994424618165e-06,
"loss": 0.8898204565048218,
"step": 1364
},
{
"epoch": 2.887949260042283,
"grad_norm": 3.120408058166504,
"learning_rate": 2.778552077959064e-06,
"loss": 0.6943051218986511,
"step": 1366
},
{
"epoch": 2.8921775898520083,
"grad_norm": 1.0748764276504517,
"learning_rate": 2.766037984250521e-06,
"loss": 0.9036406874656677,
"step": 1368
},
{
"epoch": 2.8964059196617336,
"grad_norm": 1.6608673334121704,
"learning_rate": 2.753557314326011e-06,
"loss": 0.8702334761619568,
"step": 1370
},
{
"epoch": 2.9006342494714588,
"grad_norm": 0.9324423670768738,
"learning_rate": 2.7411102207667373e-06,
"loss": 0.9293439984321594,
"step": 1372
},
{
"epoch": 2.904862579281184,
"grad_norm": 8.136580467224121,
"learning_rate": 2.728696855743418e-06,
"loss": 0.7172396183013916,
"step": 1374
},
{
"epoch": 2.909090909090909,
"grad_norm": 0.8305571675300598,
"learning_rate": 2.7163173710144323e-06,
"loss": 0.8969225883483887,
"step": 1376
},
{
"epoch": 2.9133192389006344,
"grad_norm": 2.3936431407928467,
"learning_rate": 2.703971917923951e-06,
"loss": 0.6219802498817444,
"step": 1378
},
{
"epoch": 2.9175475687103596,
"grad_norm": 1.2728478908538818,
"learning_rate": 2.6916606474000965e-06,
"loss": 0.6392749547958374,
"step": 1380
},
{
"epoch": 2.9217758985200843,
"grad_norm": 1.4852479696273804,
"learning_rate": 2.679383709953103e-06,
"loss": 0.9415355324745178,
"step": 1382
},
{
"epoch": 2.92600422832981,
"grad_norm": 2.27649188041687,
"learning_rate": 2.6671412556734578e-06,
"loss": 0.7569323182106018,
"step": 1384
},
{
"epoch": 2.9302325581395348,
"grad_norm": 1.325685739517212,
"learning_rate": 2.6549334342300844e-06,
"loss": 0.7807685136795044,
"step": 1386
},
{
"epoch": 2.93446088794926,
"grad_norm": 0.9802981019020081,
"learning_rate": 2.642760394868503e-06,
"loss": 0.9013978242874146,
"step": 1388
},
{
"epoch": 2.938689217758985,
"grad_norm": 0.4411501884460449,
"learning_rate": 2.6306222864090157e-06,
"loss": 0.6486958265304565,
"step": 1390
},
{
"epoch": 2.9429175475687104,
"grad_norm": 1.4360910654067993,
"learning_rate": 2.61851925724487e-06,
"loss": 0.8538187146186829,
"step": 1392
},
{
"epoch": 2.9471458773784356,
"grad_norm": 0.8863372206687927,
"learning_rate": 2.606451455340459e-06,
"loss": 0.2920781373977661,
"step": 1394
},
{
"epoch": 2.951374207188161,
"grad_norm": 0.8407305479049683,
"learning_rate": 2.5944190282295155e-06,
"loss": 0.4819478392601013,
"step": 1396
},
{
"epoch": 2.955602536997886,
"grad_norm": 1.3873281478881836,
"learning_rate": 2.5824221230132895e-06,
"loss": 0.8476042747497559,
"step": 1398
},
{
"epoch": 2.9598308668076108,
"grad_norm": 0.8453152775764465,
"learning_rate": 2.5704608863587676e-06,
"loss": 1.1456267833709717,
"step": 1400
},
{
"epoch": 2.9640591966173364,
"grad_norm": 2.8339626789093018,
"learning_rate": 2.558535464496877e-06,
"loss": 0.7181233763694763,
"step": 1402
},
{
"epoch": 2.968287526427061,
"grad_norm": 1.3199498653411865,
"learning_rate": 2.5466460032206872e-06,
"loss": 0.44982028007507324,
"step": 1404
},
{
"epoch": 2.9725158562367864,
"grad_norm": 4.2923808097839355,
"learning_rate": 2.53479264788364e-06,
"loss": 0.8255422711372375,
"step": 1406
},
{
"epoch": 2.9767441860465116,
"grad_norm": 0.9002554416656494,
"learning_rate": 2.522975543397769e-06,
"loss": 0.7036513090133667,
"step": 1408
},
{
"epoch": 2.980972515856237,
"grad_norm": 3.6176161766052246,
"learning_rate": 2.511194834231922e-06,
"loss": 0.6245020031929016,
"step": 1410
},
{
"epoch": 2.985200845665962,
"grad_norm": 2.2563908100128174,
"learning_rate": 2.49945066441e-06,
"loss": 0.8749826550483704,
"step": 1412
},
{
"epoch": 2.989429175475687,
"grad_norm": 0.8490020036697388,
"learning_rate": 2.4877431775092004e-06,
"loss": 0.6765884160995483,
"step": 1414
},
{
"epoch": 2.9936575052854124,
"grad_norm": 4.0843095779418945,
"learning_rate": 2.4760725166582544e-06,
"loss": 0.7786858677864075,
"step": 1416
},
{
"epoch": 2.997885835095137,
"grad_norm": 1.2769992351531982,
"learning_rate": 2.464438824535675e-06,
"loss": 0.46983152627944946,
"step": 1418
},
{
"epoch": 3.0021141649048624,
"grad_norm": 1.3773553371429443,
"learning_rate": 2.452842243368028e-06,
"loss": 0.8471871018409729,
"step": 1420
},
{
"epoch": 3.0063424947145876,
"grad_norm": 1.1417287588119507,
"learning_rate": 2.4412829149281735e-06,
"loss": 0.47468480467796326,
"step": 1422
},
{
"epoch": 3.010570824524313,
"grad_norm": 2.79913067817688,
"learning_rate": 2.4297609805335472e-06,
"loss": 0.8641015887260437,
"step": 1424
},
{
"epoch": 3.014799154334038,
"grad_norm": 3.591440439224243,
"learning_rate": 2.4182765810444263e-06,
"loss": 0.38288751244544983,
"step": 1426
},
{
"epoch": 3.019027484143763,
"grad_norm": 0.9736557602882385,
"learning_rate": 2.4068298568622083e-06,
"loss": 0.4641028344631195,
"step": 1428
},
{
"epoch": 3.0232558139534884,
"grad_norm": 2.386687755584717,
"learning_rate": 2.395420947927694e-06,
"loss": 0.6623141765594482,
"step": 1430
},
{
"epoch": 3.0274841437632136,
"grad_norm": 1.4042739868164062,
"learning_rate": 2.3840499937193804e-06,
"loss": 0.7729113698005676,
"step": 1432
},
{
"epoch": 3.031712473572939,
"grad_norm": 0.9072668552398682,
"learning_rate": 2.3727171332517496e-06,
"loss": 0.6452758312225342,
"step": 1434
},
{
"epoch": 3.035940803382664,
"grad_norm": 3.2964706420898438,
"learning_rate": 2.361422505073575e-06,
"loss": 0.46589648723602295,
"step": 1436
},
{
"epoch": 3.040169133192389,
"grad_norm": 2.0871942043304443,
"learning_rate": 2.3501662472662233e-06,
"loss": 0.10092175006866455,
"step": 1438
},
{
"epoch": 3.044397463002114,
"grad_norm": 1.5289496183395386,
"learning_rate": 2.3389484974419685e-06,
"loss": 0.9250109195709229,
"step": 1440
},
{
"epoch": 3.048625792811839,
"grad_norm": 4.423540115356445,
"learning_rate": 2.327769392742309e-06,
"loss": 0.4723988473415375,
"step": 1442
},
{
"epoch": 3.0528541226215644,
"grad_norm": 5.82054328918457,
"learning_rate": 2.316629069836291e-06,
"loss": 0.1218714565038681,
"step": 1444
},
{
"epoch": 3.0570824524312896,
"grad_norm": 2.010028600692749,
"learning_rate": 2.305527664918838e-06,
"loss": 0.4688637852668762,
"step": 1446
},
{
"epoch": 3.061310782241015,
"grad_norm": 4.10849142074585,
"learning_rate": 2.294465313709085e-06,
"loss": 0.5825417637825012,
"step": 1448
},
{
"epoch": 3.06553911205074,
"grad_norm": 1.0164453983306885,
"learning_rate": 2.2834421514487205e-06,
"loss": 0.5816891193389893,
"step": 1450
},
{
"epoch": 3.0697674418604652,
"grad_norm": 2.739476442337036,
"learning_rate": 2.27245831290033e-06,
"loss": 0.5461285710334778,
"step": 1452
},
{
"epoch": 3.0739957716701904,
"grad_norm": 1.016365885734558,
"learning_rate": 2.261513932345755e-06,
"loss": 0.665695071220398,
"step": 1454
},
{
"epoch": 3.0782241014799157,
"grad_norm": 0.3457100987434387,
"learning_rate": 2.250609143584443e-06,
"loss": 0.17615148425102234,
"step": 1456
},
{
"epoch": 3.0824524312896404,
"grad_norm": 0.718634843826294,
"learning_rate": 2.239744079931818e-06,
"loss": 0.4743122160434723,
"step": 1458
},
{
"epoch": 3.0866807610993656,
"grad_norm": 1.379178762435913,
"learning_rate": 2.228918874217649e-06,
"loss": 0.6308706998825073,
"step": 1460
},
{
"epoch": 3.090909090909091,
"grad_norm": 1.1627405881881714,
"learning_rate": 2.2181336587844244e-06,
"loss": 0.663172721862793,
"step": 1462
},
{
"epoch": 3.095137420718816,
"grad_norm": 1.101618766784668,
"learning_rate": 2.207388565485736e-06,
"loss": 0.4676929712295532,
"step": 1464
},
{
"epoch": 3.0993657505285412,
"grad_norm": 1.8849793672561646,
"learning_rate": 2.196683725684666e-06,
"loss": 0.6676308512687683,
"step": 1466
},
{
"epoch": 3.1035940803382664,
"grad_norm": 1.828122854232788,
"learning_rate": 2.1860192702521826e-06,
"loss": 0.786160409450531,
"step": 1468
},
{
"epoch": 3.1078224101479917,
"grad_norm": 1.23301100730896,
"learning_rate": 2.1753953295655354e-06,
"loss": 0.7322853803634644,
"step": 1470
},
{
"epoch": 3.112050739957717,
"grad_norm": 1.852187156677246,
"learning_rate": 2.164812033506673e-06,
"loss": 0.4200623035430908,
"step": 1472
},
{
"epoch": 3.116279069767442,
"grad_norm": 1.7936407327651978,
"learning_rate": 2.1542695114606367e-06,
"loss": 0.7873924970626831,
"step": 1474
},
{
"epoch": 3.120507399577167,
"grad_norm": 0.889731228351593,
"learning_rate": 2.1437678923139944e-06,
"loss": 0.6614270806312561,
"step": 1476
},
{
"epoch": 3.124735729386892,
"grad_norm": 2.366644859313965,
"learning_rate": 2.1333073044532628e-06,
"loss": 0.21773655712604523,
"step": 1478
},
{
"epoch": 3.1289640591966172,
"grad_norm": 1.470626711845398,
"learning_rate": 2.122887875763327e-06,
"loss": 0.7388585209846497,
"step": 1480
},
{
"epoch": 3.1331923890063424,
"grad_norm": 1.105784296989441,
"learning_rate": 2.1125097336258897e-06,
"loss": 0.6820505261421204,
"step": 1482
},
{
"epoch": 3.1374207188160677,
"grad_norm": 12.667024612426758,
"learning_rate": 2.1021730049179117e-06,
"loss": 0.1290522664785385,
"step": 1484
},
{
"epoch": 3.141649048625793,
"grad_norm": 4.755002498626709,
"learning_rate": 2.09187781601005e-06,
"loss": 0.8045941591262817,
"step": 1486
},
{
"epoch": 3.145877378435518,
"grad_norm": 5.025367259979248,
"learning_rate": 2.0816242927651254e-06,
"loss": 0.8704447150230408,
"step": 1488
},
{
"epoch": 3.1501057082452433,
"grad_norm": 4.53761625289917,
"learning_rate": 2.071412560536581e-06,
"loss": 0.6323794722557068,
"step": 1490
},
{
"epoch": 3.1543340380549685,
"grad_norm": 9.234023094177246,
"learning_rate": 2.0612427441669475e-06,
"loss": 0.5127988457679749,
"step": 1492
},
{
"epoch": 3.1585623678646932,
"grad_norm": 1.335408329963684,
"learning_rate": 2.051114967986311e-06,
"loss": 0.35932689905166626,
"step": 1494
},
{
"epoch": 3.1627906976744184,
"grad_norm": 0.22652487456798553,
"learning_rate": 2.0410293558108074e-06,
"loss": 0.5563127398490906,
"step": 1496
},
{
"epoch": 3.1670190274841437,
"grad_norm": 1.1263883113861084,
"learning_rate": 2.0309860309410986e-06,
"loss": 0.7676534056663513,
"step": 1498
},
{
"epoch": 3.171247357293869,
"grad_norm": 0.9774248003959656,
"learning_rate": 2.0209851161608613e-06,
"loss": 0.8644935488700867,
"step": 1500
},
{
"epoch": 3.175475687103594,
"grad_norm": 9.442337036132812,
"learning_rate": 2.011026733735301e-06,
"loss": 0.2904767096042633,
"step": 1502
},
{
"epoch": 3.1797040169133193,
"grad_norm": 1.4310801029205322,
"learning_rate": 2.001111005409643e-06,
"loss": 0.38165605068206787,
"step": 1504
},
{
"epoch": 3.1839323467230445,
"grad_norm": 3.6480040550231934,
"learning_rate": 1.9912380524076503e-06,
"loss": 0.5224056243896484,
"step": 1506
},
{
"epoch": 3.1881606765327697,
"grad_norm": 1.7104531526565552,
"learning_rate": 1.981407995430136e-06,
"loss": 0.3544110059738159,
"step": 1508
},
{
"epoch": 3.192389006342495,
"grad_norm": 4.724207878112793,
"learning_rate": 1.9716209546535e-06,
"loss": 0.16682866215705872,
"step": 1510
},
{
"epoch": 3.1966173361522197,
"grad_norm": 1.0528193712234497,
"learning_rate": 1.9618770497282487e-06,
"loss": 0.060837581753730774,
"step": 1512
},
{
"epoch": 3.200845665961945,
"grad_norm": 2.50264835357666,
"learning_rate": 1.9521763997775294e-06,
"loss": 0.6747430562973022,
"step": 1514
},
{
"epoch": 3.20507399577167,
"grad_norm": 1.9424082040786743,
"learning_rate": 1.942519123395689e-06,
"loss": 0.7167675495147705,
"step": 1516
},
{
"epoch": 3.2093023255813953,
"grad_norm": 0.9680429697036743,
"learning_rate": 1.932905338646812e-06,
"loss": 0.40749016404151917,
"step": 1518
},
{
"epoch": 3.2135306553911205,
"grad_norm": 0.6546218991279602,
"learning_rate": 1.9233351630632747e-06,
"loss": 0.7252275943756104,
"step": 1520
},
{
"epoch": 3.2177589852008457,
"grad_norm": 1.42070472240448,
"learning_rate": 1.9138087136443227e-06,
"loss": 0.7453226447105408,
"step": 1522
},
{
"epoch": 3.221987315010571,
"grad_norm": 2.68642520904541,
"learning_rate": 1.904326106854626e-06,
"loss": 0.810870885848999,
"step": 1524
},
{
"epoch": 3.226215644820296,
"grad_norm": 1.0590064525604248,
"learning_rate": 1.8948874586228622e-06,
"loss": 0.7173471450805664,
"step": 1526
},
{
"epoch": 3.2304439746300213,
"grad_norm": 7.953322887420654,
"learning_rate": 1.8854928843402987e-06,
"loss": 0.14129628241062164,
"step": 1528
},
{
"epoch": 3.234672304439746,
"grad_norm": 2.043560266494751,
"learning_rate": 1.8761424988593796e-06,
"loss": 0.6986078023910522,
"step": 1530
},
{
"epoch": 3.2389006342494713,
"grad_norm": 1.5021138191223145,
"learning_rate": 1.8668364164923248e-06,
"loss": 0.6416775584220886,
"step": 1532
},
{
"epoch": 3.2431289640591965,
"grad_norm": 1.1615984439849854,
"learning_rate": 1.8575747510097294e-06,
"loss": 0.7424706220626831,
"step": 1534
},
{
"epoch": 3.2473572938689217,
"grad_norm": 1.7301143407821655,
"learning_rate": 1.8483576156391747e-06,
"loss": 0.575734555721283,
"step": 1536
},
{
"epoch": 3.251585623678647,
"grad_norm": 1.3022019863128662,
"learning_rate": 1.8391851230638443e-06,
"loss": 0.6326171159744263,
"step": 1538
},
{
"epoch": 3.255813953488372,
"grad_norm": 1.426544189453125,
"learning_rate": 1.8300573854211452e-06,
"loss": 0.43459293246269226,
"step": 1540
},
{
"epoch": 3.2600422832980973,
"grad_norm": 2.986692428588867,
"learning_rate": 1.8209745143013363e-06,
"loss": 0.9323794841766357,
"step": 1542
},
{
"epoch": 3.2642706131078225,
"grad_norm": 1.1637287139892578,
"learning_rate": 1.811936620746167e-06,
"loss": 0.8775114417076111,
"step": 1544
},
{
"epoch": 3.2684989429175477,
"grad_norm": 2.207827091217041,
"learning_rate": 1.8029438152475159e-06,
"loss": 0.6728221774101257,
"step": 1546
},
{
"epoch": 3.2727272727272725,
"grad_norm": 1.165825605392456,
"learning_rate": 1.7939962077460448e-06,
"loss": 0.19605638086795807,
"step": 1548
},
{
"epoch": 3.276955602536998,
"grad_norm": 2.7155909538269043,
"learning_rate": 1.785093907629849e-06,
"loss": 0.6820244789123535,
"step": 1550
},
{
"epoch": 3.281183932346723,
"grad_norm": 1.0189692974090576,
"learning_rate": 1.7762370237331233e-06,
"loss": 0.7519944310188293,
"step": 1552
},
{
"epoch": 3.285412262156448,
"grad_norm": 0.08077098429203033,
"learning_rate": 1.7674256643348348e-06,
"loss": 0.11091545224189758,
"step": 1554
},
{
"epoch": 3.2896405919661733,
"grad_norm": 1.0770270824432373,
"learning_rate": 1.7586599371573887e-06,
"loss": 0.7418533563613892,
"step": 1556
},
{
"epoch": 3.2938689217758985,
"grad_norm": 2.1607844829559326,
"learning_rate": 1.7499399493653225e-06,
"loss": 0.4018825888633728,
"step": 1558
},
{
"epoch": 3.2980972515856237,
"grad_norm": 8.062464714050293,
"learning_rate": 1.7412658075639921e-06,
"loss": 0.9178929924964905,
"step": 1560
},
{
"epoch": 3.302325581395349,
"grad_norm": 3.0719194412231445,
"learning_rate": 1.7326376177982635e-06,
"loss": 0.6499283313751221,
"step": 1562
},
{
"epoch": 3.306553911205074,
"grad_norm": 2.9934604167938232,
"learning_rate": 1.7240554855512225e-06,
"loss": 0.4016306400299072,
"step": 1564
},
{
"epoch": 3.3107822410147993,
"grad_norm": 1.0597639083862305,
"learning_rate": 1.7155195157428878e-06,
"loss": 0.4314196705818176,
"step": 1566
},
{
"epoch": 3.3150105708245245,
"grad_norm": 0.38444390892982483,
"learning_rate": 1.707029812728916e-06,
"loss": 0.11385161429643631,
"step": 1568
},
{
"epoch": 3.3192389006342493,
"grad_norm": 1.8594019412994385,
"learning_rate": 1.6985864802993366e-06,
"loss": 0.4924876093864441,
"step": 1570
},
{
"epoch": 3.3234672304439745,
"grad_norm": 30.36981964111328,
"learning_rate": 1.690189621677284e-06,
"loss": 0.5334683656692505,
"step": 1572
},
{
"epoch": 3.3276955602536997,
"grad_norm": 1.36484694480896,
"learning_rate": 1.681839339517728e-06,
"loss": 0.119308240711689,
"step": 1574
},
{
"epoch": 3.331923890063425,
"grad_norm": 2.364935874938965,
"learning_rate": 1.67353573590622e-06,
"loss": 0.2795443534851074,
"step": 1576
},
{
"epoch": 3.33615221987315,
"grad_norm": 1.1019829511642456,
"learning_rate": 1.6652789123576513e-06,
"loss": 0.41667982935905457,
"step": 1578
},
{
"epoch": 3.3403805496828753,
"grad_norm": 0.9568066000938416,
"learning_rate": 1.6570689698150094e-06,
"loss": 0.42778220772743225,
"step": 1580
},
{
"epoch": 3.3446088794926006,
"grad_norm": 1.1865266561508179,
"learning_rate": 1.6489060086481373e-06,
"loss": 0.8351420164108276,
"step": 1582
},
{
"epoch": 3.3488372093023258,
"grad_norm": 4.144891738891602,
"learning_rate": 1.6407901286525152e-06,
"loss": 0.5559819936752319,
"step": 1584
},
{
"epoch": 3.353065539112051,
"grad_norm": 1.0440059900283813,
"learning_rate": 1.6327214290480382e-06,
"loss": 0.47431856393814087,
"step": 1586
},
{
"epoch": 3.3572938689217757,
"grad_norm": 6.686224937438965,
"learning_rate": 1.6247000084778006e-06,
"loss": 0.8340229988098145,
"step": 1588
},
{
"epoch": 3.361522198731501,
"grad_norm": 1.5362783670425415,
"learning_rate": 1.6167259650068887e-06,
"loss": 0.23965230584144592,
"step": 1590
},
{
"epoch": 3.365750528541226,
"grad_norm": 1.4101192951202393,
"learning_rate": 1.6087993961211904e-06,
"loss": 0.5966681838035583,
"step": 1592
},
{
"epoch": 3.3699788583509513,
"grad_norm": 5.704249858856201,
"learning_rate": 1.6009203987261957e-06,
"loss": 0.3519611358642578,
"step": 1594
},
{
"epoch": 3.3742071881606766,
"grad_norm": 1.013566493988037,
"learning_rate": 1.593089069145809e-06,
"loss": 0.7356735467910767,
"step": 1596
},
{
"epoch": 3.3784355179704018,
"grad_norm": 2.624998092651367,
"learning_rate": 1.5853055031211844e-06,
"loss": 0.4649814963340759,
"step": 1598
},
{
"epoch": 3.382663847780127,
"grad_norm": 0.9976376295089722,
"learning_rate": 1.577569795809544e-06,
"loss": 0.2158581018447876,
"step": 1600
},
{
"epoch": 3.386892177589852,
"grad_norm": 1.775536298751831,
"learning_rate": 1.5698820417830137e-06,
"loss": 0.5331489443778992,
"step": 1602
},
{
"epoch": 3.3911205073995774,
"grad_norm": 1.8955438137054443,
"learning_rate": 1.5622423350274784e-06,
"loss": 0.7937752604484558,
"step": 1604
},
{
"epoch": 3.395348837209302,
"grad_norm": 0.8109870553016663,
"learning_rate": 1.5546507689414217e-06,
"loss": 0.7626457214355469,
"step": 1606
},
{
"epoch": 3.3995771670190273,
"grad_norm": 3.7128920555114746,
"learning_rate": 1.5471074363347894e-06,
"loss": 0.4144952893257141,
"step": 1608
},
{
"epoch": 3.4038054968287526,
"grad_norm": 1.1224002838134766,
"learning_rate": 1.539612429427852e-06,
"loss": 0.754238486289978,
"step": 1610
},
{
"epoch": 3.4080338266384778,
"grad_norm": 2.560854911804199,
"learning_rate": 1.5321658398500808e-06,
"loss": 0.8786121606826782,
"step": 1612
},
{
"epoch": 3.412262156448203,
"grad_norm": 1.978441834449768,
"learning_rate": 1.5247677586390255e-06,
"loss": 0.5892844796180725,
"step": 1614
},
{
"epoch": 3.416490486257928,
"grad_norm": 1.3254951238632202,
"learning_rate": 1.5174182762392007e-06,
"loss": 0.7660162448883057,
"step": 1616
},
{
"epoch": 3.4207188160676534,
"grad_norm": 1.1824746131896973,
"learning_rate": 1.5101174825009815e-06,
"loss": 0.7800488471984863,
"step": 1618
},
{
"epoch": 3.4249471458773786,
"grad_norm": 1.4118192195892334,
"learning_rate": 1.5028654666795042e-06,
"loss": 0.7303675413131714,
"step": 1620
},
{
"epoch": 3.429175475687104,
"grad_norm": 8.481801986694336,
"learning_rate": 1.495662317433576e-06,
"loss": 0.40604886412620544,
"step": 1622
},
{
"epoch": 3.4334038054968286,
"grad_norm": 1.8304216861724854,
"learning_rate": 1.4885081228245909e-06,
"loss": 0.6993787288665771,
"step": 1624
},
{
"epoch": 3.4376321353065538,
"grad_norm": 2.053462266921997,
"learning_rate": 1.4814029703154515e-06,
"loss": 0.5150508284568787,
"step": 1626
},
{
"epoch": 3.441860465116279,
"grad_norm": 0.7892867922782898,
"learning_rate": 1.4743469467695017e-06,
"loss": 0.7701771855354309,
"step": 1628
},
{
"epoch": 3.446088794926004,
"grad_norm": 2.034884452819824,
"learning_rate": 1.4673401384494651e-06,
"loss": 0.5160723328590393,
"step": 1630
},
{
"epoch": 3.4503171247357294,
"grad_norm": 1.8679616451263428,
"learning_rate": 1.4603826310163874e-06,
"loss": 0.7052832245826721,
"step": 1632
},
{
"epoch": 3.4545454545454546,
"grad_norm": 1.2530237436294556,
"learning_rate": 1.453474509528593e-06,
"loss": 0.6002809405326843,
"step": 1634
},
{
"epoch": 3.45877378435518,
"grad_norm": 1.730436086654663,
"learning_rate": 1.446615858440642e-06,
"loss": 0.67807936668396,
"step": 1636
},
{
"epoch": 3.463002114164905,
"grad_norm": 0.352256715297699,
"learning_rate": 1.4398067616022993e-06,
"loss": 0.36192649602890015,
"step": 1638
},
{
"epoch": 3.46723044397463,
"grad_norm": 0.8249413967132568,
"learning_rate": 1.43304730225751e-06,
"loss": 0.44184619188308716,
"step": 1640
},
{
"epoch": 3.471458773784355,
"grad_norm": 0.8460271954536438,
"learning_rate": 1.4263375630433802e-06,
"loss": 0.13542139530181885,
"step": 1642
},
{
"epoch": 3.47568710359408,
"grad_norm": 1.2860581874847412,
"learning_rate": 1.4196776259891665e-06,
"loss": 0.6818434000015259,
"step": 1644
},
{
"epoch": 3.4799154334038054,
"grad_norm": 5.005012512207031,
"learning_rate": 1.4130675725152762e-06,
"loss": 0.5961543917655945,
"step": 1646
},
{
"epoch": 3.4841437632135306,
"grad_norm": 0.3105161786079407,
"learning_rate": 1.4065074834322677e-06,
"loss": 0.3956964612007141,
"step": 1648
},
{
"epoch": 3.488372093023256,
"grad_norm": 1.3457845449447632,
"learning_rate": 1.3999974389398669e-06,
"loss": 0.39709538221359253,
"step": 1650
},
{
"epoch": 3.492600422832981,
"grad_norm": 5.773613929748535,
"learning_rate": 1.3935375186259817e-06,
"loss": 0.44266894459724426,
"step": 1652
},
{
"epoch": 3.496828752642706,
"grad_norm": 4.148026466369629,
"learning_rate": 1.3871278014657332e-06,
"loss": 0.6778226494789124,
"step": 1654
},
{
"epoch": 3.5010570824524314,
"grad_norm": 1.255881667137146,
"learning_rate": 1.3807683658204915e-06,
"loss": 0.7524371147155762,
"step": 1656
},
{
"epoch": 3.5052854122621566,
"grad_norm": 0.27654775977134705,
"learning_rate": 1.3744592894369093e-06,
"loss": 0.5010100603103638,
"step": 1658
},
{
"epoch": 3.5095137420718814,
"grad_norm": 2.067751407623291,
"learning_rate": 1.3682006494459806e-06,
"loss": 0.6396808624267578,
"step": 1660
},
{
"epoch": 3.513742071881607,
"grad_norm": 2.0158936977386475,
"learning_rate": 1.361992522362096e-06,
"loss": 0.4932281970977783,
"step": 1662
},
{
"epoch": 3.517970401691332,
"grad_norm": 0.9829212427139282,
"learning_rate": 1.355834984082101e-06,
"loss": 0.801659107208252,
"step": 1664
},
{
"epoch": 3.522198731501057,
"grad_norm": 1.7291293144226074,
"learning_rate": 1.3497281098843746e-06,
"loss": 0.8053663372993469,
"step": 1666
},
{
"epoch": 3.526427061310782,
"grad_norm": 0.40896299481391907,
"learning_rate": 1.34367197442791e-06,
"loss": 0.10511410981416702,
"step": 1668
},
{
"epoch": 3.5306553911205074,
"grad_norm": 0.2801949381828308,
"learning_rate": 1.3376666517513931e-06,
"loss": 0.5413160920143127,
"step": 1670
},
{
"epoch": 3.5348837209302326,
"grad_norm": 2.4061994552612305,
"learning_rate": 1.3317122152723078e-06,
"loss": 0.4546211361885071,
"step": 1672
},
{
"epoch": 3.539112050739958,
"grad_norm": 1.1304322481155396,
"learning_rate": 1.3258087377860323e-06,
"loss": 0.695918083190918,
"step": 1674
},
{
"epoch": 3.543340380549683,
"grad_norm": 1.494205355644226,
"learning_rate": 1.3199562914649519e-06,
"loss": 0.4167768955230713,
"step": 1676
},
{
"epoch": 3.547568710359408,
"grad_norm": 1.2814868688583374,
"learning_rate": 1.314154947857571e-06,
"loss": 0.694580614566803,
"step": 1678
},
{
"epoch": 3.5517970401691334,
"grad_norm": 0.939893901348114,
"learning_rate": 1.3084047778876487e-06,
"loss": 0.5080829858779907,
"step": 1680
},
{
"epoch": 3.556025369978858,
"grad_norm": 0.9304549694061279,
"learning_rate": 1.3027058518533228e-06,
"loss": 0.42775458097457886,
"step": 1682
},
{
"epoch": 3.5602536997885834,
"grad_norm": 1.0786170959472656,
"learning_rate": 1.2970582394262516e-06,
"loss": 0.41797205805778503,
"step": 1684
},
{
"epoch": 3.5644820295983086,
"grad_norm": 2.6251063346862793,
"learning_rate": 1.291462009650768e-06,
"loss": 0.267993301153183,
"step": 1686
},
{
"epoch": 3.568710359408034,
"grad_norm": 2.486849069595337,
"learning_rate": 1.2859172309430285e-06,
"loss": 0.5101721286773682,
"step": 1688
},
{
"epoch": 3.572938689217759,
"grad_norm": 0.12106774002313614,
"learning_rate": 1.280423971090181e-06,
"loss": 0.16216865181922913,
"step": 1690
},
{
"epoch": 3.5771670190274842,
"grad_norm": 1.5760449171066284,
"learning_rate": 1.2749822972495333e-06,
"loss": 0.2612428069114685,
"step": 1692
},
{
"epoch": 3.5813953488372094,
"grad_norm": 2.232821464538574,
"learning_rate": 1.2695922759477356e-06,
"loss": 0.45810049772262573,
"step": 1694
},
{
"epoch": 3.585623678646934,
"grad_norm": 2.3002195358276367,
"learning_rate": 1.2642539730799622e-06,
"loss": 0.3826388716697693,
"step": 1696
},
{
"epoch": 3.58985200845666,
"grad_norm": 1.0522129535675049,
"learning_rate": 1.2589674539091115e-06,
"loss": 0.728545069694519,
"step": 1698
},
{
"epoch": 3.5940803382663846,
"grad_norm": 1.242369294166565,
"learning_rate": 1.2537327830650042e-06,
"loss": 0.8620517253875732,
"step": 1700
},
{
"epoch": 3.59830866807611,
"grad_norm": 3.118849754333496,
"learning_rate": 1.248550024543594e-06,
"loss": 0.8622214794158936,
"step": 1702
},
{
"epoch": 3.602536997885835,
"grad_norm": 3.0633206367492676,
"learning_rate": 1.2434192417061862e-06,
"loss": 0.32692381739616394,
"step": 1704
},
{
"epoch": 3.6067653276955602,
"grad_norm": 1.0920906066894531,
"learning_rate": 1.2383404972786625e-06,
"loss": 0.35319578647613525,
"step": 1706
},
{
"epoch": 3.6109936575052854,
"grad_norm": 2.205411434173584,
"learning_rate": 1.2333138533507139e-06,
"loss": 0.6222621202468872,
"step": 1708
},
{
"epoch": 3.6152219873150107,
"grad_norm": 3.191011428833008,
"learning_rate": 1.2283393713750816e-06,
"loss": 0.5288694500923157,
"step": 1710
},
{
"epoch": 3.619450317124736,
"grad_norm": 1.289028525352478,
"learning_rate": 1.2234171121668069e-06,
"loss": 0.4993457496166229,
"step": 1712
},
{
"epoch": 3.6236786469344606,
"grad_norm": 0.47990837693214417,
"learning_rate": 1.2185471359024855e-06,
"loss": 0.23633258044719696,
"step": 1714
},
{
"epoch": 3.6279069767441863,
"grad_norm": 7.445574760437012,
"learning_rate": 1.2137295021195341e-06,
"loss": 0.6363295316696167,
"step": 1716
},
{
"epoch": 3.632135306553911,
"grad_norm": 0.29476073384284973,
"learning_rate": 1.2089642697154613e-06,
"loss": 0.3790207505226135,
"step": 1718
},
{
"epoch": 3.6363636363636362,
"grad_norm": 3.0504887104034424,
"learning_rate": 1.2042514969471475e-06,
"loss": 0.4065771996974945,
"step": 1720
},
{
"epoch": 3.6405919661733614,
"grad_norm": 1.1041029691696167,
"learning_rate": 1.1995912414301329e-06,
"loss": 0.576541006565094,
"step": 1722
},
{
"epoch": 3.6448202959830867,
"grad_norm": 1.2087101936340332,
"learning_rate": 1.1949835601379144e-06,
"loss": 0.49673745036125183,
"step": 1724
},
{
"epoch": 3.649048625792812,
"grad_norm": 1.4842736721038818,
"learning_rate": 1.1904285094012463e-06,
"loss": 0.7778873443603516,
"step": 1726
},
{
"epoch": 3.653276955602537,
"grad_norm": 1.3503435850143433,
"learning_rate": 1.185926144907454e-06,
"loss": 0.8980062007904053,
"step": 1728
},
{
"epoch": 3.6575052854122623,
"grad_norm": 4.251342296600342,
"learning_rate": 1.1814765216997522e-06,
"loss": 0.7834890484809875,
"step": 1730
},
{
"epoch": 3.6617336152219875,
"grad_norm": 2.6598100662231445,
"learning_rate": 1.1770796941765727e-06,
"loss": 0.6494299173355103,
"step": 1732
},
{
"epoch": 3.6659619450317127,
"grad_norm": 1.4917654991149902,
"learning_rate": 1.1727357160908978e-06,
"loss": 0.6872870922088623,
"step": 1734
},
{
"epoch": 3.6701902748414374,
"grad_norm": 2.6213467121124268,
"learning_rate": 1.1684446405496052e-06,
"loss": 0.742832601070404,
"step": 1736
},
{
"epoch": 3.6744186046511627,
"grad_norm": 15.825387954711914,
"learning_rate": 1.1642065200128179e-06,
"loss": 0.16883477568626404,
"step": 1738
},
{
"epoch": 3.678646934460888,
"grad_norm": 1.96060311794281,
"learning_rate": 1.1600214062932613e-06,
"loss": 0.396617591381073,
"step": 1740
},
{
"epoch": 3.682875264270613,
"grad_norm": 1.4047915935516357,
"learning_rate": 1.1558893505556315e-06,
"loss": 0.29157963395118713,
"step": 1742
},
{
"epoch": 3.6871035940803383,
"grad_norm": 0.7522944211959839,
"learning_rate": 1.1518104033159715e-06,
"loss": 0.7408952713012695,
"step": 1744
},
{
"epoch": 3.6913319238900635,
"grad_norm": 0.21545825898647308,
"learning_rate": 1.1477846144410493e-06,
"loss": 0.39630720019340515,
"step": 1746
},
{
"epoch": 3.6955602536997887,
"grad_norm": 5.064393997192383,
"learning_rate": 1.1438120331477506e-06,
"loss": 0.4046843647956848,
"step": 1748
},
{
"epoch": 3.699788583509514,
"grad_norm": 1.2702808380126953,
"learning_rate": 1.1398927080024796e-06,
"loss": 0.8094220161437988,
"step": 1750
},
{
"epoch": 3.704016913319239,
"grad_norm": 1.8352627754211426,
"learning_rate": 1.1360266869205593e-06,
"loss": 0.2824421226978302,
"step": 1752
},
{
"epoch": 3.708245243128964,
"grad_norm": 2.757002115249634,
"learning_rate": 1.1322140171656507e-06,
"loss": 0.48797643184661865,
"step": 1754
},
{
"epoch": 3.712473572938689,
"grad_norm": 2.665019989013672,
"learning_rate": 1.1284547453491734e-06,
"loss": 0.5660792589187622,
"step": 1756
},
{
"epoch": 3.7167019027484143,
"grad_norm": 2.2498669624328613,
"learning_rate": 1.1247489174297372e-06,
"loss": 0.3723340630531311,
"step": 1758
},
{
"epoch": 3.7209302325581395,
"grad_norm": 0.4733290672302246,
"learning_rate": 1.1210965787125743e-06,
"loss": 0.33764639496803284,
"step": 1760
},
{
"epoch": 3.7251585623678647,
"grad_norm": 1.1974353790283203,
"learning_rate": 1.1174977738489955e-06,
"loss": 0.6515315175056458,
"step": 1762
},
{
"epoch": 3.72938689217759,
"grad_norm": 1.2089109420776367,
"learning_rate": 1.113952546835835e-06,
"loss": 0.4179830849170685,
"step": 1764
},
{
"epoch": 3.733615221987315,
"grad_norm": 1.974173903465271,
"learning_rate": 1.1104609410149174e-06,
"loss": 0.7109867930412292,
"step": 1766
},
{
"epoch": 3.7378435517970403,
"grad_norm": 1.2884228229522705,
"learning_rate": 1.1070229990725262e-06,
"loss": 0.6387155652046204,
"step": 1768
},
{
"epoch": 3.7420718816067655,
"grad_norm": 0.3562011420726776,
"learning_rate": 1.1036387630388841e-06,
"loss": 0.13756021857261658,
"step": 1770
},
{
"epoch": 3.7463002114164903,
"grad_norm": 1.5834155082702637,
"learning_rate": 1.100308274287636e-06,
"loss": 0.42179226875305176,
"step": 1772
},
{
"epoch": 3.7505285412262155,
"grad_norm": 2.7772417068481445,
"learning_rate": 1.0970315735353434e-06,
"loss": 0.9205307960510254,
"step": 1774
},
{
"epoch": 3.7547568710359407,
"grad_norm": 2.1913790702819824,
"learning_rate": 1.0938087008409909e-06,
"loss": 0.7159267663955688,
"step": 1776
},
{
"epoch": 3.758985200845666,
"grad_norm": 0.8040708303451538,
"learning_rate": 1.0906396956054919e-06,
"loss": 0.7789302468299866,
"step": 1778
},
{
"epoch": 3.763213530655391,
"grad_norm": 1.675683856010437,
"learning_rate": 1.0875245965712077e-06,
"loss": 0.9039272665977478,
"step": 1780
},
{
"epoch": 3.7674418604651163,
"grad_norm": 1.8900232315063477,
"learning_rate": 1.0844634418214763e-06,
"loss": 0.6308835744857788,
"step": 1782
},
{
"epoch": 3.7716701902748415,
"grad_norm": 2.1087892055511475,
"learning_rate": 1.0814562687801448e-06,
"loss": 0.15215148031711578,
"step": 1784
},
{
"epoch": 3.7758985200845667,
"grad_norm": 1.7626923322677612,
"learning_rate": 1.0785031142111103e-06,
"loss": 0.5002808570861816,
"step": 1786
},
{
"epoch": 3.780126849894292,
"grad_norm": 4.033579349517822,
"learning_rate": 1.075604014217875e-06,
"loss": 0.4014955759048462,
"step": 1788
},
{
"epoch": 3.7843551797040167,
"grad_norm": 2.502012014389038,
"learning_rate": 1.0727590042431009e-06,
"loss": 0.11565636843442917,
"step": 1790
},
{
"epoch": 3.7885835095137423,
"grad_norm": 0.8418030142784119,
"learning_rate": 1.0699681190681777e-06,
"loss": 0.5704522728919983,
"step": 1792
},
{
"epoch": 3.792811839323467,
"grad_norm": 2.600451707839966,
"learning_rate": 1.067231392812798e-06,
"loss": 0.706144392490387,
"step": 1794
},
{
"epoch": 3.7970401691331923,
"grad_norm": 1.5629241466522217,
"learning_rate": 1.0645488589345396e-06,
"loss": 0.37902742624282837,
"step": 1796
},
{
"epoch": 3.8012684989429175,
"grad_norm": 3.743875503540039,
"learning_rate": 1.0619205502284574e-06,
"loss": 0.7031501531600952,
"step": 1798
},
{
"epoch": 3.8054968287526427,
"grad_norm": 1.4502253532409668,
"learning_rate": 1.0593464988266811e-06,
"loss": 0.7414308786392212,
"step": 1800
},
{
"epoch": 3.809725158562368,
"grad_norm": 1.276214838027954,
"learning_rate": 1.0568267361980225e-06,
"loss": 0.4612909257411957,
"step": 1802
},
{
"epoch": 3.813953488372093,
"grad_norm": 1.3008630275726318,
"learning_rate": 1.0543612931475935e-06,
"loss": 0.6613439917564392,
"step": 1804
},
{
"epoch": 3.8181818181818183,
"grad_norm": 1.8217835426330566,
"learning_rate": 1.0519501998164247e-06,
"loss": 0.7989141345024109,
"step": 1806
},
{
"epoch": 3.822410147991543,
"grad_norm": 3.0844297409057617,
"learning_rate": 1.049593485681101e-06,
"loss": 0.12782138586044312,
"step": 1808
},
{
"epoch": 3.8266384778012688,
"grad_norm": 1.7443443536758423,
"learning_rate": 1.0472911795533997e-06,
"loss": 0.8287963271141052,
"step": 1810
},
{
"epoch": 3.8308668076109935,
"grad_norm": 1.4140102863311768,
"learning_rate": 1.045043309579938e-06,
"loss": 0.9187198281288147,
"step": 1812
},
{
"epoch": 3.8350951374207187,
"grad_norm": 1.0015143156051636,
"learning_rate": 1.0428499032418297e-06,
"loss": 0.6553718447685242,
"step": 1814
},
{
"epoch": 3.839323467230444,
"grad_norm": 1.4465429782867432,
"learning_rate": 1.0407109873543484e-06,
"loss": 0.47061508893966675,
"step": 1816
},
{
"epoch": 3.843551797040169,
"grad_norm": 1.9479564428329468,
"learning_rate": 1.0386265880665996e-06,
"loss": 0.4079810380935669,
"step": 1818
},
{
"epoch": 3.8477801268498943,
"grad_norm": 1.1661425828933716,
"learning_rate": 1.036596730861203e-06,
"loss": 0.4138513505458832,
"step": 1820
},
{
"epoch": 3.8520084566596196,
"grad_norm": 0.310879111289978,
"learning_rate": 1.0346214405539783e-06,
"loss": 0.489185094833374,
"step": 1822
},
{
"epoch": 3.8562367864693448,
"grad_norm": 1.064117193222046,
"learning_rate": 1.0327007412936419e-06,
"loss": 0.41911089420318604,
"step": 1824
},
{
"epoch": 3.8604651162790695,
"grad_norm": 21.91782569885254,
"learning_rate": 1.0308346565615158e-06,
"loss": 0.45389291644096375,
"step": 1826
},
{
"epoch": 3.864693446088795,
"grad_norm": 1.1133397817611694,
"learning_rate": 1.0290232091712337e-06,
"loss": 0.7835782766342163,
"step": 1828
},
{
"epoch": 3.86892177589852,
"grad_norm": 1.3075751066207886,
"learning_rate": 1.0272664212684687e-06,
"loss": 0.7626776099205017,
"step": 1830
},
{
"epoch": 3.873150105708245,
"grad_norm": 1.386792778968811,
"learning_rate": 1.025564314330659e-06,
"loss": 0.48643749952316284,
"step": 1832
},
{
"epoch": 3.8773784355179703,
"grad_norm": 0.9850978255271912,
"learning_rate": 1.0239169091667447e-06,
"loss": 0.37607988715171814,
"step": 1834
},
{
"epoch": 3.8816067653276956,
"grad_norm": 0.2896578907966614,
"learning_rate": 1.022324225916917e-06,
"loss": 0.38754960894584656,
"step": 1836
},
{
"epoch": 3.8858350951374208,
"grad_norm": 0.40261310338974,
"learning_rate": 1.0207862840523672e-06,
"loss": 0.5298827886581421,
"step": 1838
},
{
"epoch": 3.890063424947146,
"grad_norm": 1.5322794914245605,
"learning_rate": 1.0193031023750534e-06,
"loss": 0.6698333024978638,
"step": 1840
},
{
"epoch": 3.894291754756871,
"grad_norm": 1.147697925567627,
"learning_rate": 1.017874699017468e-06,
"loss": 0.7189450263977051,
"step": 1842
},
{
"epoch": 3.898520084566596,
"grad_norm": 1.7302483320236206,
"learning_rate": 1.0165010914424142e-06,
"loss": 0.41126322746276855,
"step": 1844
},
{
"epoch": 3.9027484143763216,
"grad_norm": 7.678551197052002,
"learning_rate": 1.015182296442798e-06,
"loss": 0.44942396879196167,
"step": 1846
},
{
"epoch": 3.9069767441860463,
"grad_norm": 1.2715246677398682,
"learning_rate": 1.0139183301414176e-06,
"loss": 0.08894074708223343,
"step": 1848
},
{
"epoch": 3.9112050739957716,
"grad_norm": 1.0519750118255615,
"learning_rate": 1.0127092079907682e-06,
"loss": 0.6557559967041016,
"step": 1850
},
{
"epoch": 3.9154334038054968,
"grad_norm": 0.5320864319801331,
"learning_rate": 1.0115549447728551e-06,
"loss": 0.4044731557369232,
"step": 1852
},
{
"epoch": 3.919661733615222,
"grad_norm": 0.8874520063400269,
"learning_rate": 1.0104555545990088e-06,
"loss": 0.6532144546508789,
"step": 1854
},
{
"epoch": 3.923890063424947,
"grad_norm": 2.0306191444396973,
"learning_rate": 1.009411050909716e-06,
"loss": 0.42025789618492126,
"step": 1856
},
{
"epoch": 3.9281183932346724,
"grad_norm": 1.6064660549163818,
"learning_rate": 1.008421446474455e-06,
"loss": 0.5511633157730103,
"step": 1858
},
{
"epoch": 3.9323467230443976,
"grad_norm": 4.159409523010254,
"learning_rate": 1.0074867533915367e-06,
"loss": 0.30626317858695984,
"step": 1860
},
{
"epoch": 3.9365750528541223,
"grad_norm": 4.066617488861084,
"learning_rate": 1.0066069830879593e-06,
"loss": 0.8771426677703857,
"step": 1862
},
{
"epoch": 3.940803382663848,
"grad_norm": 1.2944949865341187,
"learning_rate": 1.0057821463192687e-06,
"loss": 0.7035272121429443,
"step": 1864
},
{
"epoch": 3.9450317124735728,
"grad_norm": 1.8211544752120972,
"learning_rate": 1.005012253169426e-06,
"loss": 0.6347234845161438,
"step": 1866
},
{
"epoch": 3.949260042283298,
"grad_norm": 0.0695076733827591,
"learning_rate": 1.0042973130506839e-06,
"loss": 0.0017112145433202386,
"step": 1868
},
{
"epoch": 3.953488372093023,
"grad_norm": 2.2684385776519775,
"learning_rate": 1.003637334703472e-06,
"loss": 0.8074597120285034,
"step": 1870
},
{
"epoch": 3.9577167019027484,
"grad_norm": 3.7104640007019043,
"learning_rate": 1.0030323261962922e-06,
"loss": 0.4495319426059723,
"step": 1872
},
{
"epoch": 3.9619450317124736,
"grad_norm": 1.141530156135559,
"learning_rate": 1.002482294925615e-06,
"loss": 0.4483862519264221,
"step": 1874
},
{
"epoch": 3.966173361522199,
"grad_norm": 1.891880750656128,
"learning_rate": 1.0019872476157952e-06,
"loss": 0.7255292534828186,
"step": 1876
},
{
"epoch": 3.970401691331924,
"grad_norm": 6.3983964920043945,
"learning_rate": 1.0015471903189834e-06,
"loss": 0.24238592386245728,
"step": 1878
},
{
"epoch": 3.974630021141649,
"grad_norm": 0.8884820342063904,
"learning_rate": 1.0011621284150583e-06,
"loss": 0.3389042317867279,
"step": 1880
},
{
"epoch": 3.9788583509513744,
"grad_norm": 6.9358954429626465,
"learning_rate": 1.000832066611556e-06,
"loss": 0.5568378567695618,
"step": 1882
},
{
"epoch": 3.983086680761099,
"grad_norm": 1.2731667757034302,
"learning_rate": 1.0005570089436137e-06,
"loss": 0.8160393238067627,
"step": 1884
},
{
"epoch": 3.9873150105708244,
"grad_norm": 0.9162022471427917,
"learning_rate": 1.0003369587739234e-06,
"loss": 0.5509255528450012,
"step": 1886
},
{
"epoch": 3.9915433403805496,
"grad_norm": 4.305130958557129,
"learning_rate": 1.0001719187926854e-06,
"loss": 0.7377826571464539,
"step": 1888
},
{
"epoch": 3.995771670190275,
"grad_norm": 1.0448815822601318,
"learning_rate": 1.0000618910175811e-06,
"loss": 0.6807811856269836,
"step": 1890
},
{
"epoch": 4.0,
"grad_norm": 1.0182654857635498,
"learning_rate": 1.000006876793743e-06,
"loss": 0.39140379428863525,
"step": 1892
},
{
"epoch": 4.0,
"step": 1892,
"total_flos": 3.5845266747435254e+18,
"train_loss": 0.8288866212109377,
"train_runtime": 9151.7368,
"train_samples_per_second": 6.202,
"train_steps_per_second": 0.207
}
],
"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
}