9b-95 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
a6ecfcf 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": 10.896536827087402,
"learning_rate": 1.0526315789473685e-07,
"loss": 1.8247134685516357,
"step": 2
},
{
"epoch": 0.008456659619450317,
"grad_norm": 3.8788363933563232,
"learning_rate": 3.1578947368421055e-07,
"loss": 2.20851993560791,
"step": 4
},
{
"epoch": 0.012684989429175475,
"grad_norm": 11.286669731140137,
"learning_rate": 5.263157894736843e-07,
"loss": 2.2114529609680176,
"step": 6
},
{
"epoch": 0.016913319238900635,
"grad_norm": 3.0939419269561768,
"learning_rate": 7.368421052631579e-07,
"loss": 1.996035099029541,
"step": 8
},
{
"epoch": 0.021141649048625793,
"grad_norm": 10.892350196838379,
"learning_rate": 9.473684210526317e-07,
"loss": 2.012850284576416,
"step": 10
},
{
"epoch": 0.02536997885835095,
"grad_norm": 5.149282932281494,
"learning_rate": 1.1578947368421053e-06,
"loss": 1.633617639541626,
"step": 12
},
{
"epoch": 0.02959830866807611,
"grad_norm": 2.833566665649414,
"learning_rate": 1.3684210526315791e-06,
"loss": 1.6395502090454102,
"step": 14
},
{
"epoch": 0.03382663847780127,
"grad_norm": 4.205920219421387,
"learning_rate": 1.5789473684210526e-06,
"loss": 1.7606263160705566,
"step": 16
},
{
"epoch": 0.03805496828752643,
"grad_norm": 2.306366443634033,
"learning_rate": 1.7894736842105265e-06,
"loss": 1.7853772640228271,
"step": 18
},
{
"epoch": 0.042283298097251586,
"grad_norm": 92.56199645996094,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.6367794275283813,
"step": 20
},
{
"epoch": 0.046511627906976744,
"grad_norm": 9.580833435058594,
"learning_rate": 2.2105263157894738e-06,
"loss": 1.5111497640609741,
"step": 22
},
{
"epoch": 0.0507399577167019,
"grad_norm": 3.3491334915161133,
"learning_rate": 2.4210526315789477e-06,
"loss": 1.7626335620880127,
"step": 24
},
{
"epoch": 0.05496828752642706,
"grad_norm": 16.203901290893555,
"learning_rate": 2.631578947368421e-06,
"loss": 1.3580601215362549,
"step": 26
},
{
"epoch": 0.05919661733615222,
"grad_norm": 2.9515111446380615,
"learning_rate": 2.842105263157895e-06,
"loss": 1.6547349691390991,
"step": 28
},
{
"epoch": 0.06342494714587738,
"grad_norm": 2.1646769046783447,
"learning_rate": 3.052631578947369e-06,
"loss": 1.2189676761627197,
"step": 30
},
{
"epoch": 0.06765327695560254,
"grad_norm": 4.090942859649658,
"learning_rate": 3.2631578947368423e-06,
"loss": 1.0017390251159668,
"step": 32
},
{
"epoch": 0.07188160676532769,
"grad_norm": 2.5726029872894287,
"learning_rate": 3.473684210526316e-06,
"loss": 1.1753416061401367,
"step": 34
},
{
"epoch": 0.07610993657505286,
"grad_norm": 3.7985353469848633,
"learning_rate": 3.6842105263157896e-06,
"loss": 1.3560242652893066,
"step": 36
},
{
"epoch": 0.080338266384778,
"grad_norm": 4.084521770477295,
"learning_rate": 3.894736842105263e-06,
"loss": 1.2818329334259033,
"step": 38
},
{
"epoch": 0.08456659619450317,
"grad_norm": 3.3570821285247803,
"learning_rate": 4.105263157894737e-06,
"loss": 1.2094354629516602,
"step": 40
},
{
"epoch": 0.08879492600422834,
"grad_norm": 8.079401016235352,
"learning_rate": 4.315789473684211e-06,
"loss": 1.6799595355987549,
"step": 42
},
{
"epoch": 0.09302325581395349,
"grad_norm": 5.791478157043457,
"learning_rate": 4.526315789473685e-06,
"loss": 1.5857820510864258,
"step": 44
},
{
"epoch": 0.09725158562367865,
"grad_norm": 2.5073132514953613,
"learning_rate": 4.736842105263158e-06,
"loss": 1.2571828365325928,
"step": 46
},
{
"epoch": 0.1014799154334038,
"grad_norm": 1.780792236328125,
"learning_rate": 4.947368421052632e-06,
"loss": 1.496708631515503,
"step": 48
},
{
"epoch": 0.10570824524312897,
"grad_norm": 3.988499879837036,
"learning_rate": 5.157894736842106e-06,
"loss": 1.4643332958221436,
"step": 50
},
{
"epoch": 0.10993657505285412,
"grad_norm": 5.822750568389893,
"learning_rate": 5.36842105263158e-06,
"loss": 0.7182466983795166,
"step": 52
},
{
"epoch": 0.11416490486257928,
"grad_norm": 1.6763859987258911,
"learning_rate": 5.578947368421052e-06,
"loss": 1.0386135578155518,
"step": 54
},
{
"epoch": 0.11839323467230443,
"grad_norm": 2.901183605194092,
"learning_rate": 5.789473684210527e-06,
"loss": 1.3592360019683838,
"step": 56
},
{
"epoch": 0.1226215644820296,
"grad_norm": 2.61159348487854,
"learning_rate": 6e-06,
"loss": 1.3809711933135986,
"step": 58
},
{
"epoch": 0.12684989429175475,
"grad_norm": 2.7325196266174316,
"learning_rate": 6.2105263157894745e-06,
"loss": 1.3987815380096436,
"step": 60
},
{
"epoch": 0.13107822410147993,
"grad_norm": 12.265192985534668,
"learning_rate": 6.421052631578948e-06,
"loss": 0.767935037612915,
"step": 62
},
{
"epoch": 0.13530655391120508,
"grad_norm": 2.258152484893799,
"learning_rate": 6.631578947368421e-06,
"loss": 1.0339128971099854,
"step": 64
},
{
"epoch": 0.13953488372093023,
"grad_norm": 1.4521725177764893,
"learning_rate": 6.842105263157896e-06,
"loss": 1.3401219844818115,
"step": 66
},
{
"epoch": 0.14376321353065538,
"grad_norm": 3.8978211879730225,
"learning_rate": 7.052631578947369e-06,
"loss": 1.372114896774292,
"step": 68
},
{
"epoch": 0.14799154334038056,
"grad_norm": 2.7399699687957764,
"learning_rate": 7.263157894736843e-06,
"loss": 1.3536897897720337,
"step": 70
},
{
"epoch": 0.1522198731501057,
"grad_norm": 4.342154502868652,
"learning_rate": 7.473684210526316e-06,
"loss": 1.3445250988006592,
"step": 72
},
{
"epoch": 0.15644820295983086,
"grad_norm": 2.0914132595062256,
"learning_rate": 7.68421052631579e-06,
"loss": 0.9299964904785156,
"step": 74
},
{
"epoch": 0.160676532769556,
"grad_norm": 2.9866585731506348,
"learning_rate": 7.894736842105265e-06,
"loss": 0.8379074335098267,
"step": 76
},
{
"epoch": 0.1649048625792812,
"grad_norm": 6.620597839355469,
"learning_rate": 8.105263157894736e-06,
"loss": 1.283897876739502,
"step": 78
},
{
"epoch": 0.16913319238900634,
"grad_norm": 4.500240802764893,
"learning_rate": 8.315789473684212e-06,
"loss": 1.0959813594818115,
"step": 80
},
{
"epoch": 0.1733615221987315,
"grad_norm": 3.408900737762451,
"learning_rate": 8.526315789473685e-06,
"loss": 1.0993692874908447,
"step": 82
},
{
"epoch": 0.17758985200845667,
"grad_norm": 1.4808907508850098,
"learning_rate": 8.736842105263158e-06,
"loss": 0.8934997320175171,
"step": 84
},
{
"epoch": 0.18181818181818182,
"grad_norm": 8.829757690429688,
"learning_rate": 8.947368421052632e-06,
"loss": 1.7145129442214966,
"step": 86
},
{
"epoch": 0.18604651162790697,
"grad_norm": 2.505706548690796,
"learning_rate": 9.157894736842105e-06,
"loss": 1.309715986251831,
"step": 88
},
{
"epoch": 0.19027484143763213,
"grad_norm": 3.6283257007598877,
"learning_rate": 9.36842105263158e-06,
"loss": 0.9996681213378906,
"step": 90
},
{
"epoch": 0.1945031712473573,
"grad_norm": 2.294433355331421,
"learning_rate": 9.578947368421054e-06,
"loss": 1.3151381015777588,
"step": 92
},
{
"epoch": 0.19873150105708245,
"grad_norm": 1.9376956224441528,
"learning_rate": 9.789473684210527e-06,
"loss": 0.9763509035110474,
"step": 94
},
{
"epoch": 0.2029598308668076,
"grad_norm": 1.9883124828338623,
"learning_rate": 1e-05,
"loss": 1.2948415279388428,
"step": 96
},
{
"epoch": 0.20718816067653276,
"grad_norm": 2.774390697479248,
"learning_rate": 9.999972492846047e-06,
"loss": 1.3881323337554932,
"step": 98
},
{
"epoch": 0.21141649048625794,
"grad_norm": 1.9526312351226807,
"learning_rate": 9.999889971720468e-06,
"loss": 1.3291699886322021,
"step": 100
},
{
"epoch": 0.2156448202959831,
"grad_norm": 2.220778226852417,
"learning_rate": 9.999752437632121e-06,
"loss": 1.2488865852355957,
"step": 102
},
{
"epoch": 0.21987315010570824,
"grad_norm": 2.3767788410186768,
"learning_rate": 9.999559892262412e-06,
"loss": 1.3084771633148193,
"step": 104
},
{
"epoch": 0.22410147991543342,
"grad_norm": 2.3884994983673096,
"learning_rate": 9.999312337965291e-06,
"loss": 1.2973334789276123,
"step": 106
},
{
"epoch": 0.22832980972515857,
"grad_norm": 2.914511203765869,
"learning_rate": 9.999009777767202e-06,
"loss": 1.0216310024261475,
"step": 108
},
{
"epoch": 0.23255813953488372,
"grad_norm": 4.819868564605713,
"learning_rate": 9.99865221536707e-06,
"loss": 1.2858015298843384,
"step": 110
},
{
"epoch": 0.23678646934460887,
"grad_norm": 4.666462421417236,
"learning_rate": 9.998239655136237e-06,
"loss": 1.3633902072906494,
"step": 112
},
{
"epoch": 0.24101479915433405,
"grad_norm": 2.384765386581421,
"learning_rate": 9.997772102118417e-06,
"loss": 1.5496187210083008,
"step": 114
},
{
"epoch": 0.2452431289640592,
"grad_norm": 2.4982454776763916,
"learning_rate": 9.997249562029637e-06,
"loss": 1.331103801727295,
"step": 116
},
{
"epoch": 0.24947145877378435,
"grad_norm": 5.8745436668396,
"learning_rate": 9.996672041258154e-06,
"loss": 1.348480224609375,
"step": 118
},
{
"epoch": 0.2536997885835095,
"grad_norm": 6.671241283416748,
"learning_rate": 9.996039546864395e-06,
"loss": 0.8419983386993408,
"step": 120
},
{
"epoch": 0.25792811839323465,
"grad_norm": 3.24928879737854,
"learning_rate": 9.995352086580858e-06,
"loss": 1.3001736402511597,
"step": 122
},
{
"epoch": 0.26215644820295986,
"grad_norm": 3.22674822807312,
"learning_rate": 9.994609668812022e-06,
"loss": 1.1446820497512817,
"step": 124
},
{
"epoch": 0.266384778012685,
"grad_norm": 2.3322103023529053,
"learning_rate": 9.993812302634237e-06,
"loss": 1.347560167312622,
"step": 126
},
{
"epoch": 0.27061310782241016,
"grad_norm": 3.9049975872039795,
"learning_rate": 9.99295999779563e-06,
"loss": 0.9670805931091309,
"step": 128
},
{
"epoch": 0.2748414376321353,
"grad_norm": 5.073225021362305,
"learning_rate": 9.99205276471597e-06,
"loss": 0.9753428101539612,
"step": 130
},
{
"epoch": 0.27906976744186046,
"grad_norm": 4.075209617614746,
"learning_rate": 9.991090614486546e-06,
"loss": 1.2781922817230225,
"step": 132
},
{
"epoch": 0.2832980972515856,
"grad_norm": 1.9618695974349976,
"learning_rate": 9.990073558870028e-06,
"loss": 0.8443728685379028,
"step": 134
},
{
"epoch": 0.28752642706131076,
"grad_norm": 22.899608612060547,
"learning_rate": 9.989001610300333e-06,
"loss": 1.1221332550048828,
"step": 136
},
{
"epoch": 0.2917547568710359,
"grad_norm": 3.2605161666870117,
"learning_rate": 9.987874781882461e-06,
"loss": 1.3802692890167236,
"step": 138
},
{
"epoch": 0.2959830866807611,
"grad_norm": 2.035184860229492,
"learning_rate": 9.986693087392341e-06,
"loss": 0.9475889801979065,
"step": 140
},
{
"epoch": 0.30021141649048627,
"grad_norm": 2.356562852859497,
"learning_rate": 9.985456541276667e-06,
"loss": 1.56862473487854,
"step": 142
},
{
"epoch": 0.3044397463002114,
"grad_norm": 3.441044569015503,
"learning_rate": 9.984165158652708e-06,
"loss": 1.1040163040161133,
"step": 144
},
{
"epoch": 0.3086680761099366,
"grad_norm": 3.976034641265869,
"learning_rate": 9.982818955308135e-06,
"loss": 0.7072302103042603,
"step": 146
},
{
"epoch": 0.3128964059196617,
"grad_norm": 2.6471433639526367,
"learning_rate": 9.98141794770083e-06,
"loss": 1.2470674514770508,
"step": 148
},
{
"epoch": 0.3171247357293869,
"grad_norm": 1.4514168500900269,
"learning_rate": 9.979962152958669e-06,
"loss": 1.264695644378662,
"step": 150
},
{
"epoch": 0.321353065539112,
"grad_norm": 2.264636278152466,
"learning_rate": 9.97845158887933e-06,
"loss": 1.1848430633544922,
"step": 152
},
{
"epoch": 0.32558139534883723,
"grad_norm": 3.5231881141662598,
"learning_rate": 9.976886273930063e-06,
"loss": 0.7380882501602173,
"step": 154
},
{
"epoch": 0.3298097251585624,
"grad_norm": 1.9884917736053467,
"learning_rate": 9.975266227247474e-06,
"loss": 1.2512907981872559,
"step": 156
},
{
"epoch": 0.33403805496828753,
"grad_norm": 1.8444377183914185,
"learning_rate": 9.973591468637286e-06,
"loss": 1.269956350326538,
"step": 158
},
{
"epoch": 0.3382663847780127,
"grad_norm": 2.474635601043701,
"learning_rate": 9.971862018574094e-06,
"loss": 1.2545301914215088,
"step": 160
},
{
"epoch": 0.34249471458773784,
"grad_norm": 6.804418087005615,
"learning_rate": 9.97007789820112e-06,
"loss": 1.0933313369750977,
"step": 162
},
{
"epoch": 0.346723044397463,
"grad_norm": 1.8275090456008911,
"learning_rate": 9.968239129329953e-06,
"loss": 0.9640693664550781,
"step": 164
},
{
"epoch": 0.35095137420718814,
"grad_norm": 3.3783135414123535,
"learning_rate": 9.96634573444028e-06,
"loss": 0.8785189390182495,
"step": 166
},
{
"epoch": 0.35517970401691334,
"grad_norm": 2.1326019763946533,
"learning_rate": 9.964397736679615e-06,
"loss": 1.1976444721221924,
"step": 168
},
{
"epoch": 0.3594080338266385,
"grad_norm": 2.054081916809082,
"learning_rate": 9.962395159863014e-06,
"loss": 1.2078447341918945,
"step": 170
},
{
"epoch": 0.36363636363636365,
"grad_norm": 2.318174123764038,
"learning_rate": 9.960338028472781e-06,
"loss": 1.044271469116211,
"step": 172
},
{
"epoch": 0.3678646934460888,
"grad_norm": 2.973925828933716,
"learning_rate": 9.958226367658176e-06,
"loss": 1.3558095693588257,
"step": 174
},
{
"epoch": 0.37209302325581395,
"grad_norm": 3.0453221797943115,
"learning_rate": 9.9560602032351e-06,
"loss": 1.0375547409057617,
"step": 176
},
{
"epoch": 0.3763213530655391,
"grad_norm": 3.7671561241149902,
"learning_rate": 9.953839561685783e-06,
"loss": 0.8698004484176636,
"step": 178
},
{
"epoch": 0.38054968287526425,
"grad_norm": 1.4240305423736572,
"learning_rate": 9.95156447015846e-06,
"loss": 1.2858734130859375,
"step": 180
},
{
"epoch": 0.38477801268498946,
"grad_norm": 2.968477964401245,
"learning_rate": 9.94923495646704e-06,
"loss": 1.529628038406372,
"step": 182
},
{
"epoch": 0.3890063424947146,
"grad_norm": 4.791656970977783,
"learning_rate": 9.946851049090764e-06,
"loss": 1.3397505283355713,
"step": 184
},
{
"epoch": 0.39323467230443976,
"grad_norm": 2.928767442703247,
"learning_rate": 9.944412777173854e-06,
"loss": 1.1124895811080933,
"step": 186
},
{
"epoch": 0.3974630021141649,
"grad_norm": 2.318756103515625,
"learning_rate": 9.941920170525168e-06,
"loss": 0.987852931022644,
"step": 188
},
{
"epoch": 0.40169133192389006,
"grad_norm": 2.240910768508911,
"learning_rate": 9.939373259617822e-06,
"loss": 1.0128720998764038,
"step": 190
},
{
"epoch": 0.4059196617336152,
"grad_norm": 11.256593704223633,
"learning_rate": 9.936772075588825e-06,
"loss": 1.2900577783584595,
"step": 192
},
{
"epoch": 0.41014799154334036,
"grad_norm": 1.2144771814346313,
"learning_rate": 9.934116650238697e-06,
"loss": 1.0027204751968384,
"step": 194
},
{
"epoch": 0.4143763213530655,
"grad_norm": 2.8316121101379395,
"learning_rate": 9.931407016031078e-06,
"loss": 1.2830491065979004,
"step": 196
},
{
"epoch": 0.4186046511627907,
"grad_norm": 2.720400333404541,
"learning_rate": 9.928643206092338e-06,
"loss": 1.0755822658538818,
"step": 198
},
{
"epoch": 0.42283298097251587,
"grad_norm": 2.378450870513916,
"learning_rate": 9.925825254211161e-06,
"loss": 1.1814831495285034,
"step": 200
},
{
"epoch": 0.427061310782241,
"grad_norm": 4.712563991546631,
"learning_rate": 9.922953194838143e-06,
"loss": 0.591817319393158,
"step": 202
},
{
"epoch": 0.4312896405919662,
"grad_norm": 1.861677646636963,
"learning_rate": 9.920027063085363e-06,
"loss": 1.227463960647583,
"step": 204
},
{
"epoch": 0.4355179704016913,
"grad_norm": 9.449495315551758,
"learning_rate": 9.917046894725957e-06,
"loss": 1.3465194702148438,
"step": 206
},
{
"epoch": 0.4397463002114165,
"grad_norm": 3.934396266937256,
"learning_rate": 9.91401272619368e-06,
"loss": 0.7927879691123962,
"step": 208
},
{
"epoch": 0.4439746300211416,
"grad_norm": 1.9336028099060059,
"learning_rate": 9.910924594582463e-06,
"loss": 1.2989598512649536,
"step": 210
},
{
"epoch": 0.44820295983086683,
"grad_norm": 2.2313694953918457,
"learning_rate": 9.907782537645955e-06,
"loss": 1.2172267436981201,
"step": 212
},
{
"epoch": 0.452431289640592,
"grad_norm": 3.672051191329956,
"learning_rate": 9.904586593797062e-06,
"loss": 1.4167053699493408,
"step": 214
},
{
"epoch": 0.45665961945031713,
"grad_norm": 5.1475443840026855,
"learning_rate": 9.901336802107485e-06,
"loss": 0.6625733375549316,
"step": 216
},
{
"epoch": 0.4608879492600423,
"grad_norm": 3.3248605728149414,
"learning_rate": 9.89803320230723e-06,
"loss": 1.0032970905303955,
"step": 218
},
{
"epoch": 0.46511627906976744,
"grad_norm": 1.614444613456726,
"learning_rate": 9.894675834784136e-06,
"loss": 1.270370364189148,
"step": 220
},
{
"epoch": 0.4693446088794926,
"grad_norm": 2.2117111682891846,
"learning_rate": 9.891264740583367e-06,
"loss": 1.2303348779678345,
"step": 222
},
{
"epoch": 0.47357293868921774,
"grad_norm": 1.7861377000808716,
"learning_rate": 9.88779996140692e-06,
"loss": 1.3521385192871094,
"step": 224
},
{
"epoch": 0.47780126849894294,
"grad_norm": 1.8385097980499268,
"learning_rate": 9.884281539613109e-06,
"loss": 1.2516591548919678,
"step": 226
},
{
"epoch": 0.4820295983086681,
"grad_norm": 2.976051092147827,
"learning_rate": 9.880709518216061e-06,
"loss": 1.2431071996688843,
"step": 228
},
{
"epoch": 0.48625792811839325,
"grad_norm": 2.1636009216308594,
"learning_rate": 9.87708394088517e-06,
"loss": 0.9156869649887085,
"step": 230
},
{
"epoch": 0.4904862579281184,
"grad_norm": 3.7103846073150635,
"learning_rate": 9.873404851944573e-06,
"loss": 1.2507215738296509,
"step": 232
},
{
"epoch": 0.49471458773784355,
"grad_norm": 7.2142181396484375,
"learning_rate": 9.869672296372611e-06,
"loss": 0.9367599487304688,
"step": 234
},
{
"epoch": 0.4989429175475687,
"grad_norm": 5.977535724639893,
"learning_rate": 9.86588631980128e-06,
"loss": 0.8768765926361084,
"step": 236
},
{
"epoch": 0.5031712473572939,
"grad_norm": 3.2464869022369385,
"learning_rate": 9.86204696851566e-06,
"loss": 0.8368737697601318,
"step": 238
},
{
"epoch": 0.507399577167019,
"grad_norm": 2.2248594760894775,
"learning_rate": 9.858154289453363e-06,
"loss": 0.9142512679100037,
"step": 240
},
{
"epoch": 0.5116279069767442,
"grad_norm": 1.6086980104446411,
"learning_rate": 9.854208330203958e-06,
"loss": 1.3154537677764893,
"step": 242
},
{
"epoch": 0.5158562367864693,
"grad_norm": 32.39604187011719,
"learning_rate": 9.850209139008377e-06,
"loss": 1.329585313796997,
"step": 244
},
{
"epoch": 0.5200845665961945,
"grad_norm": 2.7816455364227295,
"learning_rate": 9.846156764758344e-06,
"loss": 1.2880175113677979,
"step": 246
},
{
"epoch": 0.5243128964059197,
"grad_norm": 2.4788341522216797,
"learning_rate": 9.842051256995759e-06,
"loss": 1.2413147687911987,
"step": 248
},
{
"epoch": 0.5285412262156448,
"grad_norm": 2.6546924114227295,
"learning_rate": 9.837892665912105e-06,
"loss": 1.1887092590332031,
"step": 250
},
{
"epoch": 0.53276955602537,
"grad_norm": 6.893032073974609,
"learning_rate": 9.833681042347827e-06,
"loss": 1.1574716567993164,
"step": 252
},
{
"epoch": 0.5369978858350951,
"grad_norm": 4.552901744842529,
"learning_rate": 9.829416437791721e-06,
"loss": 1.7418889999389648,
"step": 254
},
{
"epoch": 0.5412262156448203,
"grad_norm": 8.292205810546875,
"learning_rate": 9.825098904380284e-06,
"loss": 1.2168399095535278,
"step": 256
},
{
"epoch": 0.5454545454545454,
"grad_norm": 2.8239099979400635,
"learning_rate": 9.820728494897101e-06,
"loss": 1.395504355430603,
"step": 258
},
{
"epoch": 0.5496828752642706,
"grad_norm": 3.6764020919799805,
"learning_rate": 9.816305262772182e-06,
"loss": 1.0056312084197998,
"step": 260
},
{
"epoch": 0.5539112050739958,
"grad_norm": 1.5623589754104614,
"learning_rate": 9.811829262081315e-06,
"loss": 1.2014060020446777,
"step": 262
},
{
"epoch": 0.5581395348837209,
"grad_norm": 1.8238946199417114,
"learning_rate": 9.807300547545411e-06,
"loss": 1.0612789392471313,
"step": 264
},
{
"epoch": 0.5623678646934461,
"grad_norm": 2.094256639480591,
"learning_rate": 9.802719174529821e-06,
"loss": 1.498347520828247,
"step": 266
},
{
"epoch": 0.5665961945031712,
"grad_norm": 5.895513534545898,
"learning_rate": 9.798085199043672e-06,
"loss": 0.5114981532096863,
"step": 268
},
{
"epoch": 0.5708245243128964,
"grad_norm": 3.2913691997528076,
"learning_rate": 9.793398677739177e-06,
"loss": 1.004267930984497,
"step": 270
},
{
"epoch": 0.5750528541226215,
"grad_norm": 5.495273590087891,
"learning_rate": 9.788659667910942e-06,
"loss": 1.1760196685791016,
"step": 272
},
{
"epoch": 0.5792811839323467,
"grad_norm": 1.9791826009750366,
"learning_rate": 9.783868227495262e-06,
"loss": 0.9563435316085815,
"step": 274
},
{
"epoch": 0.5835095137420718,
"grad_norm": 1.8265832662582397,
"learning_rate": 9.779024415069427e-06,
"loss": 0.9386124610900879,
"step": 276
},
{
"epoch": 0.587737843551797,
"grad_norm": 1.5713542699813843,
"learning_rate": 9.774128289850987e-06,
"loss": 0.8636568188667297,
"step": 278
},
{
"epoch": 0.5919661733615222,
"grad_norm": 2.50057053565979,
"learning_rate": 9.769179911697037e-06,
"loss": 1.2353553771972656,
"step": 280
},
{
"epoch": 0.5961945031712473,
"grad_norm": 1.8529777526855469,
"learning_rate": 9.764179341103493e-06,
"loss": 1.4679677486419678,
"step": 282
},
{
"epoch": 0.6004228329809725,
"grad_norm": 0.5896320939064026,
"learning_rate": 9.759126639204337e-06,
"loss": 0.9368923902511597,
"step": 284
},
{
"epoch": 0.6046511627906976,
"grad_norm": 2.015765428543091,
"learning_rate": 9.75402186777088e-06,
"loss": 1.3021734952926636,
"step": 286
},
{
"epoch": 0.6088794926004228,
"grad_norm": 2.4573307037353516,
"learning_rate": 9.748865089211004e-06,
"loss": 0.4425542950630188,
"step": 288
},
{
"epoch": 0.6131078224101479,
"grad_norm": 1.880562663078308,
"learning_rate": 9.743656366568398e-06,
"loss": 0.948570728302002,
"step": 290
},
{
"epoch": 0.6173361522198731,
"grad_norm": 2.8457798957824707,
"learning_rate": 9.738395763521792e-06,
"loss": 0.885778546333313,
"step": 292
},
{
"epoch": 0.6215644820295984,
"grad_norm": 3.6312835216522217,
"learning_rate": 9.733083344384164e-06,
"loss": 0.7335525155067444,
"step": 294
},
{
"epoch": 0.6257928118393234,
"grad_norm": 1.7884278297424316,
"learning_rate": 9.727719174101981e-06,
"loss": 0.7348524332046509,
"step": 296
},
{
"epoch": 0.6300211416490487,
"grad_norm": 3.6789941787719727,
"learning_rate": 9.722303318254376e-06,
"loss": 1.279665231704712,
"step": 298
},
{
"epoch": 0.6342494714587738,
"grad_norm": 1.615843415260315,
"learning_rate": 9.71683584305236e-06,
"loss": 1.2439656257629395,
"step": 300
},
{
"epoch": 0.638477801268499,
"grad_norm": 3.3052875995635986,
"learning_rate": 9.71131681533802e-06,
"loss": 0.7251933813095093,
"step": 302
},
{
"epoch": 0.642706131078224,
"grad_norm": 4.2148613929748535,
"learning_rate": 9.705746302583682e-06,
"loss": 0.8468716144561768,
"step": 304
},
{
"epoch": 0.6469344608879493,
"grad_norm": 1.321298599243164,
"learning_rate": 9.700124372891104e-06,
"loss": 0.9158776998519897,
"step": 306
},
{
"epoch": 0.6511627906976745,
"grad_norm": 1.5596500635147095,
"learning_rate": 9.694451094990636e-06,
"loss": 1.2325252294540405,
"step": 308
},
{
"epoch": 0.6553911205073996,
"grad_norm": 2.505157947540283,
"learning_rate": 9.68872653824038e-06,
"loss": 1.205551028251648,
"step": 310
},
{
"epoch": 0.6596194503171248,
"grad_norm": 1.986786127090454,
"learning_rate": 9.68295077262534e-06,
"loss": 1.208229899406433,
"step": 312
},
{
"epoch": 0.6638477801268499,
"grad_norm": 2.7583141326904297,
"learning_rate": 9.677123868756573e-06,
"loss": 1.2330412864685059,
"step": 314
},
{
"epoch": 0.6680761099365751,
"grad_norm": 1.749959111213684,
"learning_rate": 9.671245897870318e-06,
"loss": 1.0731127262115479,
"step": 316
},
{
"epoch": 0.6723044397463002,
"grad_norm": 8.653021812438965,
"learning_rate": 9.665316931827131e-06,
"loss": 0.7769542932510376,
"step": 318
},
{
"epoch": 0.6765327695560254,
"grad_norm": 3.7128660678863525,
"learning_rate": 9.659337043111006e-06,
"loss": 1.2503316402435303,
"step": 320
},
{
"epoch": 0.6807610993657506,
"grad_norm": 2.3261120319366455,
"learning_rate": 9.653306304828483e-06,
"loss": 1.3103998899459839,
"step": 322
},
{
"epoch": 0.6849894291754757,
"grad_norm": 1.6073334217071533,
"learning_rate": 9.647224790707762e-06,
"loss": 1.0835018157958984,
"step": 324
},
{
"epoch": 0.6892177589852009,
"grad_norm": 1.6699814796447754,
"learning_rate": 9.641092575097795e-06,
"loss": 1.2663311958312988,
"step": 326
},
{
"epoch": 0.693446088794926,
"grad_norm": 5.467227458953857,
"learning_rate": 9.63490973296738e-06,
"loss": 1.1797797679901123,
"step": 328
},
{
"epoch": 0.6976744186046512,
"grad_norm": 8.325352668762207,
"learning_rate": 9.62867633990425e-06,
"loss": 0.8772717118263245,
"step": 330
},
{
"epoch": 0.7019027484143763,
"grad_norm": 3.8954403400421143,
"learning_rate": 9.622392472114136e-06,
"loss": 1.3020591735839844,
"step": 332
},
{
"epoch": 0.7061310782241015,
"grad_norm": 3.421680212020874,
"learning_rate": 9.61605820641985e-06,
"loss": 1.2331137657165527,
"step": 334
},
{
"epoch": 0.7103594080338267,
"grad_norm": 2.9216105937957764,
"learning_rate": 9.60967362026033e-06,
"loss": 1.216722011566162,
"step": 336
},
{
"epoch": 0.7145877378435518,
"grad_norm": 3.079737424850464,
"learning_rate": 9.60323879168971e-06,
"loss": 1.320237159729004,
"step": 338
},
{
"epoch": 0.718816067653277,
"grad_norm": 21.207590103149414,
"learning_rate": 9.596753799376355e-06,
"loss": 1.2155309915542603,
"step": 340
},
{
"epoch": 0.7230443974630021,
"grad_norm": 7.110653400421143,
"learning_rate": 9.5902187226019e-06,
"loss": 1.1356801986694336,
"step": 342
},
{
"epoch": 0.7272727272727273,
"grad_norm": 4.04885721206665,
"learning_rate": 9.583633641260285e-06,
"loss": 0.919850766658783,
"step": 344
},
{
"epoch": 0.7315010570824524,
"grad_norm": 21.629915237426758,
"learning_rate": 9.576998635856774e-06,
"loss": 0.8443883657455444,
"step": 346
},
{
"epoch": 0.7357293868921776,
"grad_norm": 4.032278537750244,
"learning_rate": 9.570313787506974e-06,
"loss": 0.9793745279312134,
"step": 348
},
{
"epoch": 0.7399577167019028,
"grad_norm": 2.9799811840057373,
"learning_rate": 9.563579177935844e-06,
"loss": 1.31476628780365,
"step": 350
},
{
"epoch": 0.7441860465116279,
"grad_norm": 4.183584690093994,
"learning_rate": 9.556794889476687e-06,
"loss": 1.1729021072387695,
"step": 352
},
{
"epoch": 0.7484143763213531,
"grad_norm": 8.587214469909668,
"learning_rate": 9.549961005070161e-06,
"loss": 0.711304247379303,
"step": 354
},
{
"epoch": 0.7526427061310782,
"grad_norm": 3.3955609798431396,
"learning_rate": 9.543077608263245e-06,
"loss": 1.023630142211914,
"step": 356
},
{
"epoch": 0.7568710359408034,
"grad_norm": 9.17407512664795,
"learning_rate": 9.536144783208229e-06,
"loss": 1.2122881412506104,
"step": 358
},
{
"epoch": 0.7610993657505285,
"grad_norm": 9.190521240234375,
"learning_rate": 9.529162614661687e-06,
"loss": 0.7622256278991699,
"step": 360
},
{
"epoch": 0.7653276955602537,
"grad_norm": 2.661991596221924,
"learning_rate": 9.522131187983433e-06,
"loss": 1.0938384532928467,
"step": 362
},
{
"epoch": 0.7695560253699789,
"grad_norm": 6.607813358306885,
"learning_rate": 9.515050589135485e-06,
"loss": 1.0740866661071777,
"step": 364
},
{
"epoch": 0.773784355179704,
"grad_norm": 1.7542043924331665,
"learning_rate": 9.507920904681006e-06,
"loss": 1.2673420906066895,
"step": 366
},
{
"epoch": 0.7780126849894292,
"grad_norm": 4.646839141845703,
"learning_rate": 9.500742221783257e-06,
"loss": 1.2059378623962402,
"step": 368
},
{
"epoch": 0.7822410147991543,
"grad_norm": 7.106878757476807,
"learning_rate": 9.493514628204516e-06,
"loss": 0.8690844774246216,
"step": 370
},
{
"epoch": 0.7864693446088795,
"grad_norm": 2.100872039794922,
"learning_rate": 9.48623821230502e-06,
"loss": 1.1936146020889282,
"step": 372
},
{
"epoch": 0.7906976744186046,
"grad_norm": 4.069069862365723,
"learning_rate": 9.478913063041878e-06,
"loss": 1.213336706161499,
"step": 374
},
{
"epoch": 0.7949260042283298,
"grad_norm": 2.4896552562713623,
"learning_rate": 9.471539269967982e-06,
"loss": 1.2931339740753174,
"step": 376
},
{
"epoch": 0.7991543340380549,
"grad_norm": 2.4918503761291504,
"learning_rate": 9.464116923230913e-06,
"loss": 0.9708983898162842,
"step": 378
},
{
"epoch": 0.8033826638477801,
"grad_norm": 7.5798773765563965,
"learning_rate": 9.456646113571844e-06,
"loss": 1.3075306415557861,
"step": 380
},
{
"epoch": 0.8076109936575053,
"grad_norm": 3.540278434753418,
"learning_rate": 9.449126932324424e-06,
"loss": 1.1696829795837402,
"step": 382
},
{
"epoch": 0.8118393234672304,
"grad_norm": 2.064589738845825,
"learning_rate": 9.441559471413665e-06,
"loss": 0.9704643487930298,
"step": 384
},
{
"epoch": 0.8160676532769556,
"grad_norm": 5.516364097595215,
"learning_rate": 9.433943823354816e-06,
"loss": 0.9623307585716248,
"step": 386
},
{
"epoch": 0.8202959830866807,
"grad_norm": 3.5701675415039062,
"learning_rate": 9.426280081252232e-06,
"loss": 1.4792554378509521,
"step": 388
},
{
"epoch": 0.8245243128964059,
"grad_norm": 3.647026777267456,
"learning_rate": 9.418568338798245e-06,
"loss": 1.1559076309204102,
"step": 390
},
{
"epoch": 0.828752642706131,
"grad_norm": 7.64189338684082,
"learning_rate": 9.410808690271999e-06,
"loss": 0.9373843669891357,
"step": 392
},
{
"epoch": 0.8329809725158562,
"grad_norm": 1.6585752964019775,
"learning_rate": 9.40300123053832e-06,
"loss": 1.23500394821167,
"step": 394
},
{
"epoch": 0.8372093023255814,
"grad_norm": 5.4211602210998535,
"learning_rate": 9.395146055046535e-06,
"loss": 0.767842710018158,
"step": 396
},
{
"epoch": 0.8414376321353065,
"grad_norm": 4.376067161560059,
"learning_rate": 9.387243259829323e-06,
"loss": 0.6949623823165894,
"step": 398
},
{
"epoch": 0.8456659619450317,
"grad_norm": 4.3541998863220215,
"learning_rate": 9.37929294150153e-06,
"loss": 1.129615306854248,
"step": 400
},
{
"epoch": 0.8498942917547568,
"grad_norm": 2.610339403152466,
"learning_rate": 9.371295197258992e-06,
"loss": 0.9685183763504028,
"step": 402
},
{
"epoch": 0.854122621564482,
"grad_norm": 1.6165143251419067,
"learning_rate": 9.363250124877345e-06,
"loss": 0.8246805667877197,
"step": 404
},
{
"epoch": 0.8583509513742071,
"grad_norm": 94.51852416992188,
"learning_rate": 9.35515782271083e-06,
"loss": 1.0025086402893066,
"step": 406
},
{
"epoch": 0.8625792811839323,
"grad_norm": 1.6516491174697876,
"learning_rate": 9.347018389691093e-06,
"loss": 0.9830936789512634,
"step": 408
},
{
"epoch": 0.8668076109936576,
"grad_norm": 3.0675344467163086,
"learning_rate": 9.338831925325974e-06,
"loss": 0.8947502374649048,
"step": 410
},
{
"epoch": 0.8710359408033826,
"grad_norm": 5.725985527038574,
"learning_rate": 9.330598529698287e-06,
"loss": 1.1744086742401123,
"step": 412
},
{
"epoch": 0.8752642706131079,
"grad_norm": 1.9330586194992065,
"learning_rate": 9.322318303464603e-06,
"loss": 1.0238213539123535,
"step": 414
},
{
"epoch": 0.879492600422833,
"grad_norm": 5.194445610046387,
"learning_rate": 9.313991347854015e-06,
"loss": 1.1272504329681396,
"step": 416
},
{
"epoch": 0.8837209302325582,
"grad_norm": 2.548227071762085,
"learning_rate": 9.305617764666897e-06,
"loss": 1.214980125427246,
"step": 418
},
{
"epoch": 0.8879492600422833,
"grad_norm": 2.423969268798828,
"learning_rate": 9.297197656273672e-06,
"loss": 1.2046895027160645,
"step": 420
},
{
"epoch": 0.8921775898520085,
"grad_norm": 3.2897918224334717,
"learning_rate": 9.288731125613544e-06,
"loss": 0.9565107822418213,
"step": 422
},
{
"epoch": 0.8964059196617337,
"grad_norm": 3.1844727993011475,
"learning_rate": 9.280218276193255e-06,
"loss": 1.2297732830047607,
"step": 424
},
{
"epoch": 0.9006342494714588,
"grad_norm": 1.984645128250122,
"learning_rate": 9.271659212085807e-06,
"loss": 1.1000995635986328,
"step": 426
},
{
"epoch": 0.904862579281184,
"grad_norm": 4.150125503540039,
"learning_rate": 9.263054037929199e-06,
"loss": 0.8711916208267212,
"step": 428
},
{
"epoch": 0.9090909090909091,
"grad_norm": 2.3658905029296875,
"learning_rate": 9.254402858925142e-06,
"loss": 1.1160526275634766,
"step": 430
},
{
"epoch": 0.9133192389006343,
"grad_norm": 9.74537467956543,
"learning_rate": 9.245705780837774e-06,
"loss": 1.2579355239868164,
"step": 432
},
{
"epoch": 0.9175475687103594,
"grad_norm": 6.627381324768066,
"learning_rate": 9.236962909992368e-06,
"loss": 0.5366116166114807,
"step": 434
},
{
"epoch": 0.9217758985200846,
"grad_norm": 10.433563232421875,
"learning_rate": 9.228174353274037e-06,
"loss": 0.37907874584198,
"step": 436
},
{
"epoch": 0.9260042283298098,
"grad_norm": 1.1070983409881592,
"learning_rate": 9.219340218126413e-06,
"loss": 0.9512419104576111,
"step": 438
},
{
"epoch": 0.9302325581395349,
"grad_norm": 2.009080648422241,
"learning_rate": 9.210460612550348e-06,
"loss": 1.140458345413208,
"step": 440
},
{
"epoch": 0.9344608879492601,
"grad_norm": 6.6255364418029785,
"learning_rate": 9.20153564510259e-06,
"loss": 0.7578648328781128,
"step": 442
},
{
"epoch": 0.9386892177589852,
"grad_norm": 2.0459015369415283,
"learning_rate": 9.19256542489445e-06,
"loss": 1.2609732151031494,
"step": 444
},
{
"epoch": 0.9429175475687104,
"grad_norm": 2.3392221927642822,
"learning_rate": 9.183550061590476e-06,
"loss": 1.0997345447540283,
"step": 446
},
{
"epoch": 0.9471458773784355,
"grad_norm": 8.233269691467285,
"learning_rate": 9.174489665407103e-06,
"loss": 1.0911047458648682,
"step": 448
},
{
"epoch": 0.9513742071881607,
"grad_norm": 4.585381507873535,
"learning_rate": 9.165384347111317e-06,
"loss": 0.7960319519042969,
"step": 450
},
{
"epoch": 0.9556025369978859,
"grad_norm": 2.489595890045166,
"learning_rate": 9.156234218019291e-06,
"loss": 0.9444265961647034,
"step": 452
},
{
"epoch": 0.959830866807611,
"grad_norm": 2.897430896759033,
"learning_rate": 9.147039389995029e-06,
"loss": 1.32359778881073,
"step": 454
},
{
"epoch": 0.9640591966173362,
"grad_norm": 4.780783653259277,
"learning_rate": 9.137799975448998e-06,
"loss": 0.6721425652503967,
"step": 456
},
{
"epoch": 0.9682875264270613,
"grad_norm": 4.538032531738281,
"learning_rate": 9.128516087336755e-06,
"loss": 0.8903247117996216,
"step": 458
},
{
"epoch": 0.9725158562367865,
"grad_norm": 1.971453070640564,
"learning_rate": 9.119187839157559e-06,
"loss": 1.2259626388549805,
"step": 460
},
{
"epoch": 0.9767441860465116,
"grad_norm": 2.368119955062866,
"learning_rate": 9.109815344952995e-06,
"loss": 1.2094054222106934,
"step": 462
},
{
"epoch": 0.9809725158562368,
"grad_norm": 3.0687880516052246,
"learning_rate": 9.100398719305566e-06,
"loss": 0.7205953001976013,
"step": 464
},
{
"epoch": 0.985200845665962,
"grad_norm": 2.1546010971069336,
"learning_rate": 9.090938077337305e-06,
"loss": 1.1862034797668457,
"step": 466
},
{
"epoch": 0.9894291754756871,
"grad_norm": 6.3788042068481445,
"learning_rate": 9.081433534708363e-06,
"loss": 0.9834831953048706,
"step": 468
},
{
"epoch": 0.9936575052854123,
"grad_norm": 3.063295841217041,
"learning_rate": 9.071885207615591e-06,
"loss": 1.2414488792419434,
"step": 470
},
{
"epoch": 0.9978858350951374,
"grad_norm": 1.5341005325317383,
"learning_rate": 9.062293212791123e-06,
"loss": 1.248815894126892,
"step": 472
},
{
"epoch": 1.0021141649048626,
"grad_norm": 3.8180720806121826,
"learning_rate": 9.052657667500951e-06,
"loss": 0.9817138910293579,
"step": 474
},
{
"epoch": 1.0063424947145878,
"grad_norm": 1.397757649421692,
"learning_rate": 9.042978689543487e-06,
"loss": 1.1698315143585205,
"step": 476
},
{
"epoch": 1.0105708245243128,
"grad_norm": 18.28521728515625,
"learning_rate": 9.033256397248124e-06,
"loss": 0.4236201047897339,
"step": 478
},
{
"epoch": 1.014799154334038,
"grad_norm": 8.285704612731934,
"learning_rate": 9.02349090947379e-06,
"loss": 0.7518464922904968,
"step": 480
},
{
"epoch": 1.0190274841437632,
"grad_norm": 17.97257423400879,
"learning_rate": 9.013682345607502e-06,
"loss": 0.7273259162902832,
"step": 482
},
{
"epoch": 1.0232558139534884,
"grad_norm": 1.4044394493103027,
"learning_rate": 9.003830825562887e-06,
"loss": 0.971366286277771,
"step": 484
},
{
"epoch": 1.0274841437632136,
"grad_norm": 2.838318347930908,
"learning_rate": 8.99393646977874e-06,
"loss": 1.079406499862671,
"step": 486
},
{
"epoch": 1.0317124735729386,
"grad_norm": 3.7370786666870117,
"learning_rate": 8.983999399217536e-06,
"loss": 0.903756856918335,
"step": 488
},
{
"epoch": 1.0359408033826638,
"grad_norm": 4.059202671051025,
"learning_rate": 8.974019735363952e-06,
"loss": 0.8265916705131531,
"step": 490
},
{
"epoch": 1.040169133192389,
"grad_norm": 8.024304389953613,
"learning_rate": 8.963997600223387e-06,
"loss": 0.663327157497406,
"step": 492
},
{
"epoch": 1.0443974630021142,
"grad_norm": 2.1461918354034424,
"learning_rate": 8.953933116320474e-06,
"loss": 1.1289945840835571,
"step": 494
},
{
"epoch": 1.0486257928118394,
"grad_norm": 1.5507380962371826,
"learning_rate": 8.943826406697569e-06,
"loss": 0.961588442325592,
"step": 496
},
{
"epoch": 1.0528541226215644,
"grad_norm": 1.8719419240951538,
"learning_rate": 8.93367759491326e-06,
"loss": 0.7690442800521851,
"step": 498
},
{
"epoch": 1.0570824524312896,
"grad_norm": 1.5177143812179565,
"learning_rate": 8.923486805040845e-06,
"loss": 0.8456506133079529,
"step": 500
},
{
"epoch": 1.0613107822410148,
"grad_norm": 1.8179266452789307,
"learning_rate": 8.913254161666828e-06,
"loss": 1.089775800704956,
"step": 502
},
{
"epoch": 1.06553911205074,
"grad_norm": 1.7904376983642578,
"learning_rate": 8.902979789889382e-06,
"loss": 0.994270920753479,
"step": 504
},
{
"epoch": 1.069767441860465,
"grad_norm": 8.080439567565918,
"learning_rate": 8.892663815316833e-06,
"loss": 0.8038026690483093,
"step": 506
},
{
"epoch": 1.0739957716701902,
"grad_norm": 1.8180094957351685,
"learning_rate": 8.882306364066114e-06,
"loss": 0.7990589141845703,
"step": 508
},
{
"epoch": 1.0782241014799154,
"grad_norm": 3.771695613861084,
"learning_rate": 8.871907562761226e-06,
"loss": 0.9725049734115601,
"step": 510
},
{
"epoch": 1.0824524312896406,
"grad_norm": 2.823824882507324,
"learning_rate": 8.861467538531692e-06,
"loss": 1.0080442428588867,
"step": 512
},
{
"epoch": 1.0866807610993658,
"grad_norm": 2.8108537197113037,
"learning_rate": 8.850986419011006e-06,
"loss": 1.103949785232544,
"step": 514
},
{
"epoch": 1.0909090909090908,
"grad_norm": 2.1794841289520264,
"learning_rate": 8.840464332335062e-06,
"loss": 0.841881513595581,
"step": 516
},
{
"epoch": 1.095137420718816,
"grad_norm": 1.8271797895431519,
"learning_rate": 8.829901407140599e-06,
"loss": 0.7940613031387329,
"step": 518
},
{
"epoch": 1.0993657505285412,
"grad_norm": 3.0446231365203857,
"learning_rate": 8.81929777256362e-06,
"loss": 0.6438655257225037,
"step": 520
},
{
"epoch": 1.1035940803382664,
"grad_norm": 2.167773485183716,
"learning_rate": 8.808653558237818e-06,
"loss": 1.3017442226409912,
"step": 522
},
{
"epoch": 1.1078224101479917,
"grad_norm": 8.364603042602539,
"learning_rate": 8.797968894292992e-06,
"loss": 0.8034107089042664,
"step": 524
},
{
"epoch": 1.1120507399577166,
"grad_norm": 2.233309268951416,
"learning_rate": 8.787243911353447e-06,
"loss": 1.2599952220916748,
"step": 526
},
{
"epoch": 1.1162790697674418,
"grad_norm": 1.5485566854476929,
"learning_rate": 8.776478740536409e-06,
"loss": 0.8181761503219604,
"step": 528
},
{
"epoch": 1.120507399577167,
"grad_norm": 7.669973373413086,
"learning_rate": 8.76567351345042e-06,
"loss": 0.5487937927246094,
"step": 530
},
{
"epoch": 1.1247357293868923,
"grad_norm": 1.840721607208252,
"learning_rate": 8.754828362193719e-06,
"loss": 1.0042545795440674,
"step": 532
},
{
"epoch": 1.1289640591966172,
"grad_norm": 1.6549917459487915,
"learning_rate": 8.743943419352638e-06,
"loss": 0.8112912178039551,
"step": 534
},
{
"epoch": 1.1331923890063424,
"grad_norm": 1.3300645351409912,
"learning_rate": 8.733018817999977e-06,
"loss": 0.9182192087173462,
"step": 536
},
{
"epoch": 1.1374207188160677,
"grad_norm": 9.7007474899292,
"learning_rate": 8.722054691693374e-06,
"loss": 0.7612076997756958,
"step": 538
},
{
"epoch": 1.1416490486257929,
"grad_norm": 3.460620164871216,
"learning_rate": 8.711051174473682e-06,
"loss": 0.7255265712738037,
"step": 540
},
{
"epoch": 1.145877378435518,
"grad_norm": 2.5219502449035645,
"learning_rate": 8.700008400863316e-06,
"loss": 1.0871713161468506,
"step": 542
},
{
"epoch": 1.150105708245243,
"grad_norm": 2.8037145137786865,
"learning_rate": 8.688926505864624e-06,
"loss": 0.7969340085983276,
"step": 544
},
{
"epoch": 1.1543340380549683,
"grad_norm": 1.5108563899993896,
"learning_rate": 8.67780562495822e-06,
"loss": 1.14487886428833,
"step": 546
},
{
"epoch": 1.1585623678646935,
"grad_norm": 32.67329025268555,
"learning_rate": 8.666645894101346e-06,
"loss": 1.2605738639831543,
"step": 548
},
{
"epoch": 1.1627906976744187,
"grad_norm": 2.3462131023406982,
"learning_rate": 8.655447449726189e-06,
"loss": 0.8077313899993896,
"step": 550
},
{
"epoch": 1.1670190274841437,
"grad_norm": 2.3685801029205322,
"learning_rate": 8.644210428738237e-06,
"loss": 1.0549979209899902,
"step": 552
},
{
"epoch": 1.1712473572938689,
"grad_norm": 3.0508532524108887,
"learning_rate": 8.632934968514582e-06,
"loss": 1.1402173042297363,
"step": 554
},
{
"epoch": 1.175475687103594,
"grad_norm": 4.525078296661377,
"learning_rate": 8.621621206902256e-06,
"loss": 1.074657678604126,
"step": 556
},
{
"epoch": 1.1797040169133193,
"grad_norm": 2.239074945449829,
"learning_rate": 8.61026928221654e-06,
"loss": 1.0447940826416016,
"step": 558
},
{
"epoch": 1.1839323467230445,
"grad_norm": 1.6043452024459839,
"learning_rate": 8.598879333239274e-06,
"loss": 0.9755728244781494,
"step": 560
},
{
"epoch": 1.1881606765327695,
"grad_norm": 1.657724142074585,
"learning_rate": 8.587451499217161e-06,
"loss": 0.720063328742981,
"step": 562
},
{
"epoch": 1.1923890063424947,
"grad_norm": 5.7902512550354,
"learning_rate": 8.575985919860065e-06,
"loss": 1.048901081085205,
"step": 564
},
{
"epoch": 1.1966173361522199,
"grad_norm": 5.778290748596191,
"learning_rate": 8.564482735339296e-06,
"loss": 0.7837084531784058,
"step": 566
},
{
"epoch": 1.200845665961945,
"grad_norm": 2.2866885662078857,
"learning_rate": 8.552942086285909e-06,
"loss": 1.2207040786743164,
"step": 568
},
{
"epoch": 1.20507399577167,
"grad_norm": 7.162411212921143,
"learning_rate": 8.541364113788974e-06,
"loss": 1.1718227863311768,
"step": 570
},
{
"epoch": 1.2093023255813953,
"grad_norm": 3.0441741943359375,
"learning_rate": 8.529748959393855e-06,
"loss": 1.0734467506408691,
"step": 572
},
{
"epoch": 1.2135306553911205,
"grad_norm": 5.833317279815674,
"learning_rate": 8.518096765100485e-06,
"loss": 0.7699643969535828,
"step": 574
},
{
"epoch": 1.2177589852008457,
"grad_norm": 5.812613010406494,
"learning_rate": 8.50640767336162e-06,
"loss": 0.816740095615387,
"step": 576
},
{
"epoch": 1.221987315010571,
"grad_norm": 2.8499767780303955,
"learning_rate": 8.494681827081096e-06,
"loss": 0.9392343759536743,
"step": 578
},
{
"epoch": 1.226215644820296,
"grad_norm": 6.1556501388549805,
"learning_rate": 8.482919369612103e-06,
"loss": 0.7733572125434875,
"step": 580
},
{
"epoch": 1.230443974630021,
"grad_norm": 1.858975887298584,
"learning_rate": 8.471120444755405e-06,
"loss": 1.0550358295440674,
"step": 582
},
{
"epoch": 1.2346723044397463,
"grad_norm": 12.825976371765137,
"learning_rate": 8.4592851967576e-06,
"loss": 0.9179757833480835,
"step": 584
},
{
"epoch": 1.2389006342494715,
"grad_norm": 5.641286849975586,
"learning_rate": 8.447413770309349e-06,
"loss": 0.8434258699417114,
"step": 586
},
{
"epoch": 1.2431289640591967,
"grad_norm": 2.1304309368133545,
"learning_rate": 8.435506310543607e-06,
"loss": 0.9091084599494934,
"step": 588
},
{
"epoch": 1.2473572938689217,
"grad_norm": 1.647655963897705,
"learning_rate": 8.42356296303386e-06,
"loss": 0.5901362299919128,
"step": 590
},
{
"epoch": 1.251585623678647,
"grad_norm": 1.907537579536438,
"learning_rate": 8.411583873792325e-06,
"loss": 0.8266245126724243,
"step": 592
},
{
"epoch": 1.255813953488372,
"grad_norm": 4.309906005859375,
"learning_rate": 8.399569189268182e-06,
"loss": 1.0732121467590332,
"step": 594
},
{
"epoch": 1.2600422832980973,
"grad_norm": 2.548267126083374,
"learning_rate": 8.387519056345778e-06,
"loss": 0.7040244340896606,
"step": 596
},
{
"epoch": 1.2642706131078225,
"grad_norm": 2.0476815700531006,
"learning_rate": 8.375433622342824e-06,
"loss": 1.1000163555145264,
"step": 598
},
{
"epoch": 1.2684989429175475,
"grad_norm": 7.119895935058594,
"learning_rate": 8.363313035008613e-06,
"loss": 1.2506763935089111,
"step": 600
},
{
"epoch": 1.2727272727272727,
"grad_norm": 2.059234142303467,
"learning_rate": 8.351157442522187e-06,
"loss": 0.8936987519264221,
"step": 602
},
{
"epoch": 1.276955602536998,
"grad_norm": 2.784914255142212,
"learning_rate": 8.338966993490554e-06,
"loss": 1.044416904449463,
"step": 604
},
{
"epoch": 1.2811839323467231,
"grad_norm": 5.724180698394775,
"learning_rate": 8.326741836946852e-06,
"loss": 0.7338106036186218,
"step": 606
},
{
"epoch": 1.285412262156448,
"grad_norm": 2.882368326187134,
"learning_rate": 8.314482122348523e-06,
"loss": 1.0866812467575073,
"step": 608
},
{
"epoch": 1.2896405919661733,
"grad_norm": 4.481686115264893,
"learning_rate": 8.30218799957551e-06,
"loss": 0.7193027138710022,
"step": 610
},
{
"epoch": 1.2938689217758985,
"grad_norm": 2.474071741104126,
"learning_rate": 8.289859618928404e-06,
"loss": 0.6360830068588257,
"step": 612
},
{
"epoch": 1.2980972515856237,
"grad_norm": 1.6604639291763306,
"learning_rate": 8.277497131126608e-06,
"loss": 0.9158079028129578,
"step": 614
},
{
"epoch": 1.302325581395349,
"grad_norm": 5.723630428314209,
"learning_rate": 8.265100687306503e-06,
"loss": 0.5168040990829468,
"step": 616
},
{
"epoch": 1.306553911205074,
"grad_norm": 5.876730918884277,
"learning_rate": 8.252670439019595e-06,
"loss": 0.9449325799942017,
"step": 618
},
{
"epoch": 1.3107822410147991,
"grad_norm": 2.145705223083496,
"learning_rate": 8.240206538230663e-06,
"loss": 1.0762653350830078,
"step": 620
},
{
"epoch": 1.3150105708245243,
"grad_norm": 1.9223785400390625,
"learning_rate": 8.227709137315905e-06,
"loss": 0.9964494109153748,
"step": 622
},
{
"epoch": 1.3192389006342495,
"grad_norm": 4.805337905883789,
"learning_rate": 8.215178389061064e-06,
"loss": 0.7304139137268066,
"step": 624
},
{
"epoch": 1.3234672304439745,
"grad_norm": 1.7315987348556519,
"learning_rate": 8.202614446659572e-06,
"loss": 1.1091506481170654,
"step": 626
},
{
"epoch": 1.3276955602536997,
"grad_norm": 3.6356797218322754,
"learning_rate": 8.190017463710675e-06,
"loss": 1.2163007259368896,
"step": 628
},
{
"epoch": 1.331923890063425,
"grad_norm": 3.0900719165802,
"learning_rate": 8.177387594217548e-06,
"loss": 1.033562421798706,
"step": 630
},
{
"epoch": 1.3361522198731501,
"grad_norm": 12.784806251525879,
"learning_rate": 8.164724992585421e-06,
"loss": 0.6766372323036194,
"step": 632
},
{
"epoch": 1.3403805496828753,
"grad_norm": 1.8737220764160156,
"learning_rate": 8.152029813619685e-06,
"loss": 0.7295541763305664,
"step": 634
},
{
"epoch": 1.3446088794926006,
"grad_norm": 1.5760674476623535,
"learning_rate": 8.139302212524006e-06,
"loss": 0.7528289556503296,
"step": 636
},
{
"epoch": 1.3488372093023255,
"grad_norm": 29.73055648803711,
"learning_rate": 8.126542344898419e-06,
"loss": 0.9538460969924927,
"step": 638
},
{
"epoch": 1.3530655391120507,
"grad_norm": 1.477025032043457,
"learning_rate": 8.113750366737431e-06,
"loss": 0.7379294037818909,
"step": 640
},
{
"epoch": 1.357293868921776,
"grad_norm": 2.8461596965789795,
"learning_rate": 8.100926434428116e-06,
"loss": 1.1783721446990967,
"step": 642
},
{
"epoch": 1.361522198731501,
"grad_norm": 3.098938465118408,
"learning_rate": 8.0880707047482e-06,
"loss": 1.0101354122161865,
"step": 644
},
{
"epoch": 1.3657505285412261,
"grad_norm": 10.078448295593262,
"learning_rate": 8.075183334864139e-06,
"loss": 0.7559187412261963,
"step": 646
},
{
"epoch": 1.3699788583509513,
"grad_norm": 5.03165864944458,
"learning_rate": 8.062264482329209e-06,
"loss": 0.7427597641944885,
"step": 648
},
{
"epoch": 1.3742071881606766,
"grad_norm": 7.7786054611206055,
"learning_rate": 8.049314305081575e-06,
"loss": 0.7390559911727905,
"step": 650
},
{
"epoch": 1.3784355179704018,
"grad_norm": 2.802616834640503,
"learning_rate": 8.036332961442355e-06,
"loss": 0.7865035533905029,
"step": 652
},
{
"epoch": 1.382663847780127,
"grad_norm": 4.577014923095703,
"learning_rate": 8.023320610113691e-06,
"loss": 1.154538869857788,
"step": 654
},
{
"epoch": 1.386892177589852,
"grad_norm": 1.7306524515151978,
"learning_rate": 8.010277410176801e-06,
"loss": 1.0215108394622803,
"step": 656
},
{
"epoch": 1.3911205073995772,
"grad_norm": 5.207336902618408,
"learning_rate": 7.997203521090053e-06,
"loss": 0.8093292713165283,
"step": 658
},
{
"epoch": 1.3953488372093024,
"grad_norm": 2.564222574234009,
"learning_rate": 7.984099102686986e-06,
"loss": 0.9318503141403198,
"step": 660
},
{
"epoch": 1.3995771670190273,
"grad_norm": 1.5019357204437256,
"learning_rate": 7.970964315174384e-06,
"loss": 1.101630449295044,
"step": 662
},
{
"epoch": 1.4038054968287526,
"grad_norm": 0.5547475814819336,
"learning_rate": 7.9577993191303e-06,
"loss": 0.8165593147277832,
"step": 664
},
{
"epoch": 1.4080338266384778,
"grad_norm": 4.846535682678223,
"learning_rate": 7.944604275502102e-06,
"loss": 1.2657307386398315,
"step": 666
},
{
"epoch": 1.412262156448203,
"grad_norm": 1.7670280933380127,
"learning_rate": 7.931379345604495e-06,
"loss": 1.069098949432373,
"step": 668
},
{
"epoch": 1.4164904862579282,
"grad_norm": 2.6535050868988037,
"learning_rate": 7.918124691117563e-06,
"loss": 0.7185813188552856,
"step": 670
},
{
"epoch": 1.4207188160676534,
"grad_norm": 4.082168102264404,
"learning_rate": 7.904840474084783e-06,
"loss": 1.0205265283584595,
"step": 672
},
{
"epoch": 1.4249471458773784,
"grad_norm": 1.9456113576889038,
"learning_rate": 7.891526856911046e-06,
"loss": 0.5637375116348267,
"step": 674
},
{
"epoch": 1.4291754756871036,
"grad_norm": 5.078863143920898,
"learning_rate": 7.878184002360664e-06,
"loss": 0.8056528568267822,
"step": 676
},
{
"epoch": 1.4334038054968288,
"grad_norm": 2.015641689300537,
"learning_rate": 7.864812073555402e-06,
"loss": 1.2710548639297485,
"step": 678
},
{
"epoch": 1.437632135306554,
"grad_norm": 1.927955150604248,
"learning_rate": 7.851411233972461e-06,
"loss": 1.1538954973220825,
"step": 680
},
{
"epoch": 1.441860465116279,
"grad_norm": 12.046246528625488,
"learning_rate": 7.837981647442486e-06,
"loss": 1.1251258850097656,
"step": 682
},
{
"epoch": 1.4460887949260042,
"grad_norm": 4.336371898651123,
"learning_rate": 7.82452347814757e-06,
"loss": 1.0136600732803345,
"step": 684
},
{
"epoch": 1.4503171247357294,
"grad_norm": 4.732032775878906,
"learning_rate": 7.811036890619239e-06,
"loss": 1.4260122776031494,
"step": 686
},
{
"epoch": 1.4545454545454546,
"grad_norm": 2.917357921600342,
"learning_rate": 7.797522049736441e-06,
"loss": 1.0499993562698364,
"step": 688
},
{
"epoch": 1.4587737843551798,
"grad_norm": 4.521021366119385,
"learning_rate": 7.78397912072354e-06,
"loss": 0.41519081592559814,
"step": 690
},
{
"epoch": 1.463002114164905,
"grad_norm": 2.002058982849121,
"learning_rate": 7.770408269148281e-06,
"loss": 1.1246000528335571,
"step": 692
},
{
"epoch": 1.46723044397463,
"grad_norm": 4.08298921585083,
"learning_rate": 7.756809660919778e-06,
"loss": 0.6873543858528137,
"step": 694
},
{
"epoch": 1.4714587737843552,
"grad_norm": 2.4843642711639404,
"learning_rate": 7.74318346228648e-06,
"loss": 1.0610965490341187,
"step": 696
},
{
"epoch": 1.4756871035940804,
"grad_norm": 2.4317469596862793,
"learning_rate": 7.72952983983414e-06,
"loss": 1.0578041076660156,
"step": 698
},
{
"epoch": 1.4799154334038054,
"grad_norm": 4.874222755432129,
"learning_rate": 7.715848960483776e-06,
"loss": 0.9403927326202393,
"step": 700
},
{
"epoch": 1.4841437632135306,
"grad_norm": 1.7970054149627686,
"learning_rate": 7.702140991489637e-06,
"loss": 0.9910249710083008,
"step": 702
},
{
"epoch": 1.4883720930232558,
"grad_norm": 2.068347454071045,
"learning_rate": 7.688406100437149e-06,
"loss": 0.9890831708908081,
"step": 704
},
{
"epoch": 1.492600422832981,
"grad_norm": 1.6065139770507812,
"learning_rate": 7.674644455240878e-06,
"loss": 0.7872288227081299,
"step": 706
},
{
"epoch": 1.4968287526427062,
"grad_norm": 1.7556036710739136,
"learning_rate": 7.660856224142456e-06,
"loss": 0.2610781192779541,
"step": 708
},
{
"epoch": 1.5010570824524314,
"grad_norm": 3.3843164443969727,
"learning_rate": 7.647041575708555e-06,
"loss": 0.6090059280395508,
"step": 710
},
{
"epoch": 1.5052854122621564,
"grad_norm": 15.1498441696167,
"learning_rate": 7.633200678828804e-06,
"loss": 0.5418379306793213,
"step": 712
},
{
"epoch": 1.5095137420718816,
"grad_norm": 2.641587734222412,
"learning_rate": 7.6193337027137215e-06,
"loss": 0.9955928325653076,
"step": 714
},
{
"epoch": 1.5137420718816068,
"grad_norm": 2.793987989425659,
"learning_rate": 7.605440816892667e-06,
"loss": 1.046097755432129,
"step": 716
},
{
"epoch": 1.5179704016913318,
"grad_norm": 2.7375261783599854,
"learning_rate": 7.591522191211746e-06,
"loss": 0.9742239713668823,
"step": 718
},
{
"epoch": 1.522198731501057,
"grad_norm": 0.5911352634429932,
"learning_rate": 7.577577995831755e-06,
"loss": 0.7911907434463501,
"step": 720
},
{
"epoch": 1.5264270613107822,
"grad_norm": 3.520672082901001,
"learning_rate": 7.563608401226081e-06,
"loss": 0.8702362775802612,
"step": 722
},
{
"epoch": 1.5306553911205074,
"grad_norm": 3.057141065597534,
"learning_rate": 7.549613578178632e-06,
"loss": 0.930722713470459,
"step": 724
},
{
"epoch": 1.5348837209302326,
"grad_norm": 3.2798831462860107,
"learning_rate": 7.535593697781741e-06,
"loss": 0.868494987487793,
"step": 726
},
{
"epoch": 1.5391120507399578,
"grad_norm": 7.597347259521484,
"learning_rate": 7.52154893143408e-06,
"loss": 0.6464821100234985,
"step": 728
},
{
"epoch": 1.543340380549683,
"grad_norm": 5.931199550628662,
"learning_rate": 7.50747945083856e-06,
"loss": 0.8129299283027649,
"step": 730
},
{
"epoch": 1.547568710359408,
"grad_norm": 2.386497974395752,
"learning_rate": 7.493385428000234e-06,
"loss": 1.0200014114379883,
"step": 732
},
{
"epoch": 1.5517970401691332,
"grad_norm": 1.0424129962921143,
"learning_rate": 7.479267035224191e-06,
"loss": 0.25139543414115906,
"step": 734
},
{
"epoch": 1.5560253699788582,
"grad_norm": 2.1804757118225098,
"learning_rate": 7.465124445113459e-06,
"loss": 0.9722799062728882,
"step": 736
},
{
"epoch": 1.5602536997885834,
"grad_norm": 0.7487244606018066,
"learning_rate": 7.450957830566878e-06,
"loss": 0.6949633359909058,
"step": 738
},
{
"epoch": 1.5644820295983086,
"grad_norm": 2.06032657623291,
"learning_rate": 7.436767364777009e-06,
"loss": 1.0798356533050537,
"step": 740
},
{
"epoch": 1.5687103594080338,
"grad_norm": 3.9733195304870605,
"learning_rate": 7.422553221227993e-06,
"loss": 0.7741891145706177,
"step": 742
},
{
"epoch": 1.572938689217759,
"grad_norm": 17.24365997314453,
"learning_rate": 7.4083155736934484e-06,
"loss": 1.1810104846954346,
"step": 744
},
{
"epoch": 1.5771670190274842,
"grad_norm": 3.310624837875366,
"learning_rate": 7.394054596234336e-06,
"loss": 1.1740543842315674,
"step": 746
},
{
"epoch": 1.5813953488372094,
"grad_norm": 2.394580841064453,
"learning_rate": 7.379770463196831e-06,
"loss": 0.9907056093215942,
"step": 748
},
{
"epoch": 1.5856236786469344,
"grad_norm": 1.1957483291625977,
"learning_rate": 7.365463349210203e-06,
"loss": 0.06049314886331558,
"step": 750
},
{
"epoch": 1.5898520084566596,
"grad_norm": 10.840557098388672,
"learning_rate": 7.351133429184666e-06,
"loss": 0.7954455614089966,
"step": 752
},
{
"epoch": 1.5940803382663846,
"grad_norm": 3.0723681449890137,
"learning_rate": 7.336780878309251e-06,
"loss": 1.1361210346221924,
"step": 754
},
{
"epoch": 1.5983086680761098,
"grad_norm": 3.554576873779297,
"learning_rate": 7.322405872049657e-06,
"loss": 1.0744367837905884,
"step": 756
},
{
"epoch": 1.602536997885835,
"grad_norm": 2.628537893295288,
"learning_rate": 7.308008586146111e-06,
"loss": 1.1632356643676758,
"step": 758
},
{
"epoch": 1.6067653276955602,
"grad_norm": 1.3682700395584106,
"learning_rate": 7.293589196611221e-06,
"loss": 1.022247076034546,
"step": 760
},
{
"epoch": 1.6109936575052854,
"grad_norm": 2.051743745803833,
"learning_rate": 7.279147879727811e-06,
"loss": 1.1260600090026855,
"step": 762
},
{
"epoch": 1.6152219873150107,
"grad_norm": 4.048652648925781,
"learning_rate": 7.264684812046787e-06,
"loss": 1.0169697999954224,
"step": 764
},
{
"epoch": 1.6194503171247359,
"grad_norm": 33.26997756958008,
"learning_rate": 7.250200170384961e-06,
"loss": 0.8381563425064087,
"step": 766
},
{
"epoch": 1.6236786469344608,
"grad_norm": 2.187502145767212,
"learning_rate": 7.235694131822894e-06,
"loss": 1.0925520658493042,
"step": 768
},
{
"epoch": 1.627906976744186,
"grad_norm": 4.692344665527344,
"learning_rate": 7.221166873702738e-06,
"loss": 1.0759539604187012,
"step": 770
},
{
"epoch": 1.6321353065539113,
"grad_norm": 1.536486268043518,
"learning_rate": 7.206618573626058e-06,
"loss": 1.0734047889709473,
"step": 772
},
{
"epoch": 1.6363636363636362,
"grad_norm": 1.413970947265625,
"learning_rate": 7.19204940945167e-06,
"loss": 0.7571790218353271,
"step": 774
},
{
"epoch": 1.6405919661733614,
"grad_norm": 3.0673563480377197,
"learning_rate": 7.177459559293456e-06,
"loss": 1.0992330312728882,
"step": 776
},
{
"epoch": 1.6448202959830867,
"grad_norm": 5.289725303649902,
"learning_rate": 7.162849201518198e-06,
"loss": 0.6651745438575745,
"step": 778
},
{
"epoch": 1.6490486257928119,
"grad_norm": 1.7714585065841675,
"learning_rate": 7.1482185147433905e-06,
"loss": 1.0166174173355103,
"step": 780
},
{
"epoch": 1.653276955602537,
"grad_norm": 12.672255516052246,
"learning_rate": 7.133567677835055e-06,
"loss": 0.7810131311416626,
"step": 782
},
{
"epoch": 1.6575052854122623,
"grad_norm": 1.8811969757080078,
"learning_rate": 7.11889686990556e-06,
"loss": 1.1926665306091309,
"step": 784
},
{
"epoch": 1.6617336152219875,
"grad_norm": 6.557992458343506,
"learning_rate": 7.104206270311423e-06,
"loss": 0.4603302776813507,
"step": 786
},
{
"epoch": 1.6659619450317125,
"grad_norm": 2.2292654514312744,
"learning_rate": 7.089496058651133e-06,
"loss": 1.1104487180709839,
"step": 788
},
{
"epoch": 1.6701902748414377,
"grad_norm": 2.4697046279907227,
"learning_rate": 7.0747664147629335e-06,
"loss": 0.7627810835838318,
"step": 790
},
{
"epoch": 1.6744186046511627,
"grad_norm": 1.0306406021118164,
"learning_rate": 7.060017518722637e-06,
"loss": 0.5201520323753357,
"step": 792
},
{
"epoch": 1.6786469344608879,
"grad_norm": 1.7865716218948364,
"learning_rate": 7.0452495508414276e-06,
"loss": 1.005384922027588,
"step": 794
},
{
"epoch": 1.682875264270613,
"grad_norm": 1.0506714582443237,
"learning_rate": 7.030462691663641e-06,
"loss": 0.6256891489028931,
"step": 796
},
{
"epoch": 1.6871035940803383,
"grad_norm": 1.5216120481491089,
"learning_rate": 7.015657121964575e-06,
"loss": 1.1837128400802612,
"step": 798
},
{
"epoch": 1.6913319238900635,
"grad_norm": 4.1076979637146,
"learning_rate": 7.000833022748267e-06,
"loss": 1.0919930934906006,
"step": 800
},
{
"epoch": 1.6955602536997887,
"grad_norm": 12.374058723449707,
"learning_rate": 6.985990575245283e-06,
"loss": 0.5114182233810425,
"step": 802
},
{
"epoch": 1.699788583509514,
"grad_norm": 1.2405080795288086,
"learning_rate": 6.97112996091051e-06,
"loss": 1.0521721839904785,
"step": 804
},
{
"epoch": 1.7040169133192389,
"grad_norm": 13.187895774841309,
"learning_rate": 6.956251361420926e-06,
"loss": 0.8074842691421509,
"step": 806
},
{
"epoch": 1.708245243128964,
"grad_norm": 7.946383953094482,
"learning_rate": 6.94135495867339e-06,
"loss": 0.6991731524467468,
"step": 808
},
{
"epoch": 1.712473572938689,
"grad_norm": 3.7994225025177,
"learning_rate": 6.926440934782407e-06,
"loss": 1.0886170864105225,
"step": 810
},
{
"epoch": 1.7167019027484143,
"grad_norm": 3.3649086952209473,
"learning_rate": 6.9115094720779155e-06,
"loss": 1.3119444847106934,
"step": 812
},
{
"epoch": 1.7209302325581395,
"grad_norm": 2.5864572525024414,
"learning_rate": 6.896560753103041e-06,
"loss": 1.0521005392074585,
"step": 814
},
{
"epoch": 1.7251585623678647,
"grad_norm": 2.432727098464966,
"learning_rate": 6.881594960611881e-06,
"loss": 1.120629072189331,
"step": 816
},
{
"epoch": 1.72938689217759,
"grad_norm": 2.5775856971740723,
"learning_rate": 6.866612277567261e-06,
"loss": 0.89154052734375,
"step": 818
},
{
"epoch": 1.733615221987315,
"grad_norm": 6.314211845397949,
"learning_rate": 6.851612887138501e-06,
"loss": 0.9600451588630676,
"step": 820
},
{
"epoch": 1.7378435517970403,
"grad_norm": 2.48111629486084,
"learning_rate": 6.836596972699175e-06,
"loss": 0.7135413885116577,
"step": 822
},
{
"epoch": 1.7420718816067653,
"grad_norm": 1.5427579879760742,
"learning_rate": 6.82156471782487e-06,
"loss": 0.5749027132987976,
"step": 824
},
{
"epoch": 1.7463002114164905,
"grad_norm": 50.70668029785156,
"learning_rate": 6.806516306290942e-06,
"loss": 0.9154766798019409,
"step": 826
},
{
"epoch": 1.7505285412262155,
"grad_norm": 1.7229548692703247,
"learning_rate": 6.791451922070265e-06,
"loss": 0.9088847637176514,
"step": 828
},
{
"epoch": 1.7547568710359407,
"grad_norm": 5.488432884216309,
"learning_rate": 6.77637174933099e-06,
"loss": 0.7854145765304565,
"step": 830
},
{
"epoch": 1.758985200845666,
"grad_norm": 3.098795175552368,
"learning_rate": 6.7612759724342895e-06,
"loss": 1.1325592994689941,
"step": 832
},
{
"epoch": 1.763213530655391,
"grad_norm": 8.12088394165039,
"learning_rate": 6.746164775932096e-06,
"loss": 0.7300012111663818,
"step": 834
},
{
"epoch": 1.7674418604651163,
"grad_norm": 1.5795220136642456,
"learning_rate": 6.731038344564862e-06,
"loss": 0.8145651817321777,
"step": 836
},
{
"epoch": 1.7716701902748415,
"grad_norm": 3.577568292617798,
"learning_rate": 6.715896863259286e-06,
"loss": 1.0880582332611084,
"step": 838
},
{
"epoch": 1.7758985200845667,
"grad_norm": 1.6903910636901855,
"learning_rate": 6.700740517126061e-06,
"loss": 0.8389778137207031,
"step": 840
},
{
"epoch": 1.7801268498942917,
"grad_norm": 17.495668411254883,
"learning_rate": 6.685569491457608e-06,
"loss": 0.3770165741443634,
"step": 842
},
{
"epoch": 1.784355179704017,
"grad_norm": 1.8477803468704224,
"learning_rate": 6.670383971725809e-06,
"loss": 0.7221954464912415,
"step": 844
},
{
"epoch": 1.7885835095137421,
"grad_norm": 15.788907051086426,
"learning_rate": 6.655184143579746e-06,
"loss": 0.9943246841430664,
"step": 846
},
{
"epoch": 1.792811839323467,
"grad_norm": 2.3214402198791504,
"learning_rate": 6.639970192843423e-06,
"loss": 0.7315882444381714,
"step": 848
},
{
"epoch": 1.7970401691331923,
"grad_norm": 5.237164497375488,
"learning_rate": 6.624742305513502e-06,
"loss": 0.6889032125473022,
"step": 850
},
{
"epoch": 1.8012684989429175,
"grad_norm": 2.5360910892486572,
"learning_rate": 6.609500667757022e-06,
"loss": 0.928741991519928,
"step": 852
},
{
"epoch": 1.8054968287526427,
"grad_norm": 4.8166937828063965,
"learning_rate": 6.594245465909127e-06,
"loss": 0.8700690269470215,
"step": 854
},
{
"epoch": 1.809725158562368,
"grad_norm": 1.7509115934371948,
"learning_rate": 6.578976886470791e-06,
"loss": 1.0757640600204468,
"step": 856
},
{
"epoch": 1.8139534883720931,
"grad_norm": 1.636606216430664,
"learning_rate": 6.56369511610653e-06,
"loss": 0.6516000628471375,
"step": 858
},
{
"epoch": 1.8181818181818183,
"grad_norm": 36.33342361450195,
"learning_rate": 6.548400341642128e-06,
"loss": 1.000751256942749,
"step": 860
},
{
"epoch": 1.8224101479915433,
"grad_norm": 1.3217778205871582,
"learning_rate": 6.533092750062345e-06,
"loss": 0.9945152997970581,
"step": 862
},
{
"epoch": 1.8266384778012685,
"grad_norm": 2.364422559738159,
"learning_rate": 6.5177725285086416e-06,
"loss": 0.5224348902702332,
"step": 864
},
{
"epoch": 1.8308668076109935,
"grad_norm": 1.4368764162063599,
"learning_rate": 6.5024398642768785e-06,
"loss": 0.7363207936286926,
"step": 866
},
{
"epoch": 1.8350951374207187,
"grad_norm": 1.170482873916626,
"learning_rate": 6.487094944815036e-06,
"loss": 1.0560847520828247,
"step": 868
},
{
"epoch": 1.839323467230444,
"grad_norm": 2.150615692138672,
"learning_rate": 6.47173795772092e-06,
"loss": 1.065834879875183,
"step": 870
},
{
"epoch": 1.8435517970401691,
"grad_norm": 2.517181396484375,
"learning_rate": 6.456369090739869e-06,
"loss": 0.6953645944595337,
"step": 872
},
{
"epoch": 1.8477801268498943,
"grad_norm": 1.9154415130615234,
"learning_rate": 6.440988531762454e-06,
"loss": 1.0712156295776367,
"step": 874
},
{
"epoch": 1.8520084566596196,
"grad_norm": 10.783659934997559,
"learning_rate": 6.42559646882219e-06,
"loss": 0.9156849384307861,
"step": 876
},
{
"epoch": 1.8562367864693448,
"grad_norm": 1.9548885822296143,
"learning_rate": 6.410193090093227e-06,
"loss": 0.2907255291938782,
"step": 878
},
{
"epoch": 1.8604651162790697,
"grad_norm": 4.9688191413879395,
"learning_rate": 6.394778583888064e-06,
"loss": 0.727829098701477,
"step": 880
},
{
"epoch": 1.864693446088795,
"grad_norm": 1.9172627925872803,
"learning_rate": 6.379353138655227e-06,
"loss": 1.0950987339019775,
"step": 882
},
{
"epoch": 1.86892177589852,
"grad_norm": 1.9341951608657837,
"learning_rate": 6.3639169429769805e-06,
"loss": 0.7468138337135315,
"step": 884
},
{
"epoch": 1.8731501057082451,
"grad_norm": 3.5540969371795654,
"learning_rate": 6.348470185567025e-06,
"loss": 1.1177680492401123,
"step": 886
},
{
"epoch": 1.8773784355179703,
"grad_norm": 2.0135140419006348,
"learning_rate": 6.333013055268169e-06,
"loss": 1.1278119087219238,
"step": 888
},
{
"epoch": 1.8816067653276956,
"grad_norm": 7.849596977233887,
"learning_rate": 6.317545741050044e-06,
"loss": 0.5701417326927185,
"step": 890
},
{
"epoch": 1.8858350951374208,
"grad_norm": 111.09715270996094,
"learning_rate": 6.3020684320067805e-06,
"loss": 0.7751279473304749,
"step": 892
},
{
"epoch": 1.890063424947146,
"grad_norm": 7.196433067321777,
"learning_rate": 6.2865813173546976e-06,
"loss": 1.029491662979126,
"step": 894
},
{
"epoch": 1.8942917547568712,
"grad_norm": 9.199488639831543,
"learning_rate": 6.271084586429994e-06,
"loss": 1.438428282737732,
"step": 896
},
{
"epoch": 1.8985200845665962,
"grad_norm": 1.8875133991241455,
"learning_rate": 6.255578428686433e-06,
"loss": 1.0397366285324097,
"step": 898
},
{
"epoch": 1.9027484143763214,
"grad_norm": 8.345558166503906,
"learning_rate": 6.240063033693023e-06,
"loss": 0.49346959590911865,
"step": 900
},
{
"epoch": 1.9069767441860463,
"grad_norm": 6.782925605773926,
"learning_rate": 6.2245385911316994e-06,
"loss": 1.0342949628829956,
"step": 902
},
{
"epoch": 1.9112050739957716,
"grad_norm": 49.524024963378906,
"learning_rate": 6.2090052907950096e-06,
"loss": 0.49469444155693054,
"step": 904
},
{
"epoch": 1.9154334038054968,
"grad_norm": 2.50970196723938,
"learning_rate": 6.193463322583792e-06,
"loss": 0.8747320771217346,
"step": 906
},
{
"epoch": 1.919661733615222,
"grad_norm": 2.5345299243927,
"learning_rate": 6.177912876504851e-06,
"loss": 1.2518303394317627,
"step": 908
},
{
"epoch": 1.9238900634249472,
"grad_norm": 1.52976655960083,
"learning_rate": 6.16235414266864e-06,
"loss": 0.79448401927948,
"step": 910
},
{
"epoch": 1.9281183932346724,
"grad_norm": 1.8579740524291992,
"learning_rate": 6.146787311286927e-06,
"loss": 1.0500165224075317,
"step": 912
},
{
"epoch": 1.9323467230443976,
"grad_norm": 3.2406206130981445,
"learning_rate": 6.131212572670483e-06,
"loss": 0.7119067311286926,
"step": 914
},
{
"epoch": 1.9365750528541226,
"grad_norm": 1.7776092290878296,
"learning_rate": 6.115630117226741e-06,
"loss": 1.1076221466064453,
"step": 916
},
{
"epoch": 1.9408033826638478,
"grad_norm": 2.010603189468384,
"learning_rate": 6.1000401354574855e-06,
"loss": 0.6314924955368042,
"step": 918
},
{
"epoch": 1.945031712473573,
"grad_norm": 4.036433696746826,
"learning_rate": 6.084442817956502e-06,
"loss": 0.8945176601409912,
"step": 920
},
{
"epoch": 1.949260042283298,
"grad_norm": 2.4197325706481934,
"learning_rate": 6.0688383554072625e-06,
"loss": 0.999203622341156,
"step": 922
},
{
"epoch": 1.9534883720930232,
"grad_norm": 2.1143195629119873,
"learning_rate": 6.053226938580596e-06,
"loss": 1.0023666620254517,
"step": 924
},
{
"epoch": 1.9577167019027484,
"grad_norm": 1.326448917388916,
"learning_rate": 6.03760875833234e-06,
"loss": 0.8744796514511108,
"step": 926
},
{
"epoch": 1.9619450317124736,
"grad_norm": 1.7432910203933716,
"learning_rate": 6.021984005601027e-06,
"loss": 1.1491549015045166,
"step": 928
},
{
"epoch": 1.9661733615221988,
"grad_norm": 2.293018341064453,
"learning_rate": 6.006352871405534e-06,
"loss": 0.7791812419891357,
"step": 930
},
{
"epoch": 1.970401691331924,
"grad_norm": 3.8080272674560547,
"learning_rate": 5.990715546842755e-06,
"loss": 1.2125418186187744,
"step": 932
},
{
"epoch": 1.9746300211416492,
"grad_norm": 5.57804012298584,
"learning_rate": 5.9750722230852694e-06,
"loss": 0.8035264015197754,
"step": 934
},
{
"epoch": 1.9788583509513742,
"grad_norm": 2.6657679080963135,
"learning_rate": 5.959423091378992e-06,
"loss": 0.6753436326980591,
"step": 936
},
{
"epoch": 1.9830866807610994,
"grad_norm": 4.782979488372803,
"learning_rate": 5.943768343040845e-06,
"loss": 0.8794345855712891,
"step": 938
},
{
"epoch": 1.9873150105708244,
"grad_norm": 1.6203317642211914,
"learning_rate": 5.928108169456419e-06,
"loss": 1.0244265794754028,
"step": 940
},
{
"epoch": 1.9915433403805496,
"grad_norm": 1.8124064207077026,
"learning_rate": 5.912442762077624e-06,
"loss": 1.0524667501449585,
"step": 942
},
{
"epoch": 1.9957716701902748,
"grad_norm": 2.375265121459961,
"learning_rate": 5.89677231242036e-06,
"loss": 1.0999082326889038,
"step": 944
},
{
"epoch": 2.0,
"grad_norm": 1.9802249670028687,
"learning_rate": 5.8810970120621714e-06,
"loss": 0.42669975757598877,
"step": 946
},
{
"epoch": 2.004228329809725,
"grad_norm": 10.09327507019043,
"learning_rate": 5.8654170526399014e-06,
"loss": 0.7893583178520203,
"step": 948
},
{
"epoch": 2.0084566596194504,
"grad_norm": 1.1274702548980713,
"learning_rate": 5.849732625847354e-06,
"loss": 0.903694748878479,
"step": 950
},
{
"epoch": 2.0126849894291756,
"grad_norm": 3.7474493980407715,
"learning_rate": 5.834043923432947e-06,
"loss": 1.0042245388031006,
"step": 952
},
{
"epoch": 2.016913319238901,
"grad_norm": 2.3496716022491455,
"learning_rate": 5.818351137197373e-06,
"loss": 1.0462510585784912,
"step": 954
},
{
"epoch": 2.0211416490486256,
"grad_norm": 3.5694615840911865,
"learning_rate": 5.802654458991245e-06,
"loss": 1.156407117843628,
"step": 956
},
{
"epoch": 2.025369978858351,
"grad_norm": 2.089592695236206,
"learning_rate": 5.786954080712763e-06,
"loss": 0.6736660003662109,
"step": 958
},
{
"epoch": 2.029598308668076,
"grad_norm": 2.3547141551971436,
"learning_rate": 5.7712501943053576e-06,
"loss": 0.9282336831092834,
"step": 960
},
{
"epoch": 2.033826638477801,
"grad_norm": 3.8958959579467773,
"learning_rate": 5.755542991755352e-06,
"loss": 0.8411227464675903,
"step": 962
},
{
"epoch": 2.0380549682875264,
"grad_norm": 7.480360984802246,
"learning_rate": 5.739832665089604e-06,
"loss": 0.7590432167053223,
"step": 964
},
{
"epoch": 2.0422832980972516,
"grad_norm": 4.321681022644043,
"learning_rate": 5.724119406373171e-06,
"loss": 0.6027129888534546,
"step": 966
},
{
"epoch": 2.046511627906977,
"grad_norm": 9.939786911010742,
"learning_rate": 5.708403407706954e-06,
"loss": 0.692194938659668,
"step": 968
},
{
"epoch": 2.050739957716702,
"grad_norm": 1.7376298904418945,
"learning_rate": 5.692684861225349e-06,
"loss": 0.9862563610076904,
"step": 970
},
{
"epoch": 2.0549682875264272,
"grad_norm": 9.941186904907227,
"learning_rate": 5.676963959093904e-06,
"loss": 0.46941280364990234,
"step": 972
},
{
"epoch": 2.059196617336152,
"grad_norm": 2.4498038291931152,
"learning_rate": 5.66124089350696e-06,
"loss": 0.8682308197021484,
"step": 974
},
{
"epoch": 2.063424947145877,
"grad_norm": 1.8821032047271729,
"learning_rate": 5.645515856685316e-06,
"loss": 0.8273845911026001,
"step": 976
},
{
"epoch": 2.0676532769556024,
"grad_norm": 2.1382665634155273,
"learning_rate": 5.629789040873862e-06,
"loss": 0.904760479927063,
"step": 978
},
{
"epoch": 2.0718816067653276,
"grad_norm": 2.633039712905884,
"learning_rate": 5.614060638339238e-06,
"loss": 0.6572993993759155,
"step": 980
},
{
"epoch": 2.076109936575053,
"grad_norm": 14.35608196258545,
"learning_rate": 5.598330841367487e-06,
"loss": 0.6795235872268677,
"step": 982
},
{
"epoch": 2.080338266384778,
"grad_norm": 1.8717094659805298,
"learning_rate": 5.582599842261693e-06,
"loss": 0.6491124629974365,
"step": 984
},
{
"epoch": 2.0845665961945032,
"grad_norm": 9.967153549194336,
"learning_rate": 5.566867833339644e-06,
"loss": 0.8446928858757019,
"step": 986
},
{
"epoch": 2.0887949260042284,
"grad_norm": 3.6088359355926514,
"learning_rate": 5.551135006931467e-06,
"loss": 0.9382858276367188,
"step": 988
},
{
"epoch": 2.0930232558139537,
"grad_norm": 1.9751594066619873,
"learning_rate": 5.535401555377287e-06,
"loss": 0.8613487482070923,
"step": 990
},
{
"epoch": 2.097251585623679,
"grad_norm": 4.570734977722168,
"learning_rate": 5.51966767102487e-06,
"loss": 0.5753958225250244,
"step": 992
},
{
"epoch": 2.1014799154334036,
"grad_norm": 7.644152641296387,
"learning_rate": 5.503933546227273e-06,
"loss": 0.6314504146575928,
"step": 994
},
{
"epoch": 2.105708245243129,
"grad_norm": 1.3914270401000977,
"learning_rate": 5.4881993733404916e-06,
"loss": 0.8304003477096558,
"step": 996
},
{
"epoch": 2.109936575052854,
"grad_norm": 3.2083687782287598,
"learning_rate": 5.472465344721113e-06,
"loss": 0.9095271825790405,
"step": 998
},
{
"epoch": 2.1141649048625792,
"grad_norm": 3.6328563690185547,
"learning_rate": 5.456731652723959e-06,
"loss": 0.8138751983642578,
"step": 1000
},
{
"epoch": 2.1183932346723044,
"grad_norm": 1.3322921991348267,
"learning_rate": 5.440998489699731e-06,
"loss": 0.23458953201770782,
"step": 1002
},
{
"epoch": 2.1226215644820297,
"grad_norm": 5.287220478057861,
"learning_rate": 5.425266047992672e-06,
"loss": 0.6508815288543701,
"step": 1004
},
{
"epoch": 2.126849894291755,
"grad_norm": 1.9631524085998535,
"learning_rate": 5.4095345199382e-06,
"loss": 0.808562159538269,
"step": 1006
},
{
"epoch": 2.13107822410148,
"grad_norm": 10.157617568969727,
"learning_rate": 5.393804097860568e-06,
"loss": 0.8550094366073608,
"step": 1008
},
{
"epoch": 2.1353065539112053,
"grad_norm": 1.7904855012893677,
"learning_rate": 5.3780749740705065e-06,
"loss": 0.5485635995864868,
"step": 1010
},
{
"epoch": 2.13953488372093,
"grad_norm": 7.756897449493408,
"learning_rate": 5.362347340862869e-06,
"loss": 0.565731406211853,
"step": 1012
},
{
"epoch": 2.1437632135306552,
"grad_norm": 2.28004789352417,
"learning_rate": 5.3466213905142935e-06,
"loss": 0.6423985958099365,
"step": 1014
},
{
"epoch": 2.1479915433403804,
"grad_norm": 2.1014184951782227,
"learning_rate": 5.3308973152808395e-06,
"loss": 0.7800198793411255,
"step": 1016
},
{
"epoch": 2.1522198731501057,
"grad_norm": 2.7697136402130127,
"learning_rate": 5.315175307395644e-06,
"loss": 0.8784196972846985,
"step": 1018
},
{
"epoch": 2.156448202959831,
"grad_norm": 4.86939811706543,
"learning_rate": 5.299455559066571e-06,
"loss": 0.75307297706604,
"step": 1020
},
{
"epoch": 2.160676532769556,
"grad_norm": 2.4742865562438965,
"learning_rate": 5.283738262473859e-06,
"loss": 0.6117876768112183,
"step": 1022
},
{
"epoch": 2.1649048625792813,
"grad_norm": 1.715795874595642,
"learning_rate": 5.268023609767772e-06,
"loss": 0.8438136577606201,
"step": 1024
},
{
"epoch": 2.1691331923890065,
"grad_norm": 8.555145263671875,
"learning_rate": 5.2523117930662545e-06,
"loss": 0.8897883892059326,
"step": 1026
},
{
"epoch": 2.1733615221987317,
"grad_norm": 3.315589189529419,
"learning_rate": 5.2366030044525785e-06,
"loss": 0.8744567036628723,
"step": 1028
},
{
"epoch": 2.177589852008457,
"grad_norm": 1.9522734880447388,
"learning_rate": 5.220897435972992e-06,
"loss": 0.9510366916656494,
"step": 1030
},
{
"epoch": 2.1818181818181817,
"grad_norm": 2.269469738006592,
"learning_rate": 5.205195279634385e-06,
"loss": 0.9322319030761719,
"step": 1032
},
{
"epoch": 2.186046511627907,
"grad_norm": 2.910616636276245,
"learning_rate": 5.1894967274019245e-06,
"loss": 1.00469970703125,
"step": 1034
},
{
"epoch": 2.190274841437632,
"grad_norm": 1.7277647256851196,
"learning_rate": 5.173801971196716e-06,
"loss": 0.5041213035583496,
"step": 1036
},
{
"epoch": 2.1945031712473573,
"grad_norm": 9.523260116577148,
"learning_rate": 5.158111202893464e-06,
"loss": 0.41243860125541687,
"step": 1038
},
{
"epoch": 2.1987315010570825,
"grad_norm": 5.0023908615112305,
"learning_rate": 5.142424614318112e-06,
"loss": 1.0433340072631836,
"step": 1040
},
{
"epoch": 2.2029598308668077,
"grad_norm": 34.82558822631836,
"learning_rate": 5.12674239724551e-06,
"loss": 0.3530353307723999,
"step": 1042
},
{
"epoch": 2.207188160676533,
"grad_norm": 1.6512815952301025,
"learning_rate": 5.111064743397061e-06,
"loss": 0.8830319046974182,
"step": 1044
},
{
"epoch": 2.211416490486258,
"grad_norm": 2.0978894233703613,
"learning_rate": 5.09539184443838e-06,
"loss": 0.41917169094085693,
"step": 1046
},
{
"epoch": 2.2156448202959833,
"grad_norm": 6.901039123535156,
"learning_rate": 5.079723891976956e-06,
"loss": 0.7787059545516968,
"step": 1048
},
{
"epoch": 2.219873150105708,
"grad_norm": 4.032864093780518,
"learning_rate": 5.064061077559803e-06,
"loss": 0.5900967121124268,
"step": 1050
},
{
"epoch": 2.2241014799154333,
"grad_norm": 3.0690975189208984,
"learning_rate": 5.048403592671116e-06,
"loss": 0.7773241996765137,
"step": 1052
},
{
"epoch": 2.2283298097251585,
"grad_norm": 3.30556321144104,
"learning_rate": 5.032751628729943e-06,
"loss": 0.9232714772224426,
"step": 1054
},
{
"epoch": 2.2325581395348837,
"grad_norm": 2.021285057067871,
"learning_rate": 5.017105377087829e-06,
"loss": 0.9250907897949219,
"step": 1056
},
{
"epoch": 2.236786469344609,
"grad_norm": 0.4446614980697632,
"learning_rate": 5.0014650290264854e-06,
"loss": 0.5066299438476562,
"step": 1058
},
{
"epoch": 2.241014799154334,
"grad_norm": 2.0930335521698,
"learning_rate": 4.9858307757554536e-06,
"loss": 0.8940294981002808,
"step": 1060
},
{
"epoch": 2.2452431289640593,
"grad_norm": 3.3482418060302734,
"learning_rate": 4.97020280840976e-06,
"loss": 1.0772144794464111,
"step": 1062
},
{
"epoch": 2.2494714587737845,
"grad_norm": 2.231581211090088,
"learning_rate": 4.954581318047582e-06,
"loss": 0.8229595422744751,
"step": 1064
},
{
"epoch": 2.2536997885835097,
"grad_norm": 3.4601523876190186,
"learning_rate": 4.938966495647918e-06,
"loss": 0.4687274098396301,
"step": 1066
},
{
"epoch": 2.2579281183932345,
"grad_norm": 2.9063169956207275,
"learning_rate": 4.923358532108244e-06,
"loss": 0.7931013107299805,
"step": 1068
},
{
"epoch": 2.2621564482029597,
"grad_norm": 9.824275016784668,
"learning_rate": 4.907757618242186e-06,
"loss": 0.7804174423217773,
"step": 1070
},
{
"epoch": 2.266384778012685,
"grad_norm": 2.151869297027588,
"learning_rate": 4.892163944777181e-06,
"loss": 0.931074857711792,
"step": 1072
},
{
"epoch": 2.27061310782241,
"grad_norm": 1.5082589387893677,
"learning_rate": 4.876577702352155e-06,
"loss": 0.6072179079055786,
"step": 1074
},
{
"epoch": 2.2748414376321353,
"grad_norm": 6.300508975982666,
"learning_rate": 4.860999081515183e-06,
"loss": 0.792299211025238,
"step": 1076
},
{
"epoch": 2.2790697674418605,
"grad_norm": 12.900514602661133,
"learning_rate": 4.845428272721161e-06,
"loss": 0.24233265221118927,
"step": 1078
},
{
"epoch": 2.2832980972515857,
"grad_norm": 1.6175845861434937,
"learning_rate": 4.8298654663294865e-06,
"loss": 0.13425399363040924,
"step": 1080
},
{
"epoch": 2.287526427061311,
"grad_norm": 0.911220133304596,
"learning_rate": 4.814310852601717e-06,
"loss": 0.5192358493804932,
"step": 1082
},
{
"epoch": 2.291754756871036,
"grad_norm": 2.0668821334838867,
"learning_rate": 4.7987646216992545e-06,
"loss": 0.32165616750717163,
"step": 1084
},
{
"epoch": 2.295983086680761,
"grad_norm": 3.3673295974731445,
"learning_rate": 4.783226963681018e-06,
"loss": 0.1494128406047821,
"step": 1086
},
{
"epoch": 2.300211416490486,
"grad_norm": 12.6102876663208,
"learning_rate": 4.76769806850112e-06,
"loss": 0.5317608118057251,
"step": 1088
},
{
"epoch": 2.3044397463002113,
"grad_norm": 2.655872106552124,
"learning_rate": 4.752178126006544e-06,
"loss": 0.8661559224128723,
"step": 1090
},
{
"epoch": 2.3086680761099365,
"grad_norm": 6.036033630371094,
"learning_rate": 4.736667325934819e-06,
"loss": 0.5884255170822144,
"step": 1092
},
{
"epoch": 2.3128964059196617,
"grad_norm": 1.9851877689361572,
"learning_rate": 4.721165857911709e-06,
"loss": 0.9707424640655518,
"step": 1094
},
{
"epoch": 2.317124735729387,
"grad_norm": 1.6478338241577148,
"learning_rate": 4.705673911448891e-06,
"loss": 0.6345282793045044,
"step": 1096
},
{
"epoch": 2.321353065539112,
"grad_norm": 2.4139904975891113,
"learning_rate": 4.690191675941631e-06,
"loss": 0.8481141328811646,
"step": 1098
},
{
"epoch": 2.3255813953488373,
"grad_norm": 5.668582916259766,
"learning_rate": 4.674719340666477e-06,
"loss": 0.4993610978126526,
"step": 1100
},
{
"epoch": 2.3298097251585626,
"grad_norm": 2.046574354171753,
"learning_rate": 4.659257094778949e-06,
"loss": 0.6641027331352234,
"step": 1102
},
{
"epoch": 2.3340380549682873,
"grad_norm": 3.0015952587127686,
"learning_rate": 4.643805127311212e-06,
"loss": 0.9833755493164062,
"step": 1104
},
{
"epoch": 2.3382663847780125,
"grad_norm": 2.820047616958618,
"learning_rate": 4.628363627169774e-06,
"loss": 0.8555904030799866,
"step": 1106
},
{
"epoch": 2.3424947145877377,
"grad_norm": 1.935187578201294,
"learning_rate": 4.612932783133182e-06,
"loss": 0.8578470945358276,
"step": 1108
},
{
"epoch": 2.346723044397463,
"grad_norm": 3.994123697280884,
"learning_rate": 4.5975127838497036e-06,
"loss": 0.9445978999137878,
"step": 1110
},
{
"epoch": 2.350951374207188,
"grad_norm": 1.2030290365219116,
"learning_rate": 4.582103817835021e-06,
"loss": 0.6240684986114502,
"step": 1112
},
{
"epoch": 2.3551797040169133,
"grad_norm": 5.201540946960449,
"learning_rate": 4.566706073469938e-06,
"loss": 0.7566081285476685,
"step": 1114
},
{
"epoch": 2.3594080338266386,
"grad_norm": 2.5575149059295654,
"learning_rate": 4.551319738998067e-06,
"loss": 0.614561140537262,
"step": 1116
},
{
"epoch": 2.3636363636363638,
"grad_norm": 6.1345086097717285,
"learning_rate": 4.535945002523524e-06,
"loss": 1.0704319477081299,
"step": 1118
},
{
"epoch": 2.367864693446089,
"grad_norm": 2.4230992794036865,
"learning_rate": 4.5205820520086415e-06,
"loss": 0.8580273389816284,
"step": 1120
},
{
"epoch": 2.3720930232558137,
"grad_norm": 2.386030673980713,
"learning_rate": 4.505231075271662e-06,
"loss": 0.7698273658752441,
"step": 1122
},
{
"epoch": 2.376321353065539,
"grad_norm": 2.348827362060547,
"learning_rate": 4.489892259984446e-06,
"loss": 0.5700345635414124,
"step": 1124
},
{
"epoch": 2.380549682875264,
"grad_norm": 10.323647499084473,
"learning_rate": 4.474565793670167e-06,
"loss": 0.9251955151557922,
"step": 1126
},
{
"epoch": 2.3847780126849893,
"grad_norm": 7.83730411529541,
"learning_rate": 4.459251863701038e-06,
"loss": 0.6038002371788025,
"step": 1128
},
{
"epoch": 2.3890063424947146,
"grad_norm": 2.6007416248321533,
"learning_rate": 4.443950657296004e-06,
"loss": 0.7854123115539551,
"step": 1130
},
{
"epoch": 2.3932346723044398,
"grad_norm": 7.870021343231201,
"learning_rate": 4.4286623615184585e-06,
"loss": 0.6616194248199463,
"step": 1132
},
{
"epoch": 2.397463002114165,
"grad_norm": 2.1885952949523926,
"learning_rate": 4.413387163273964e-06,
"loss": 0.9212248921394348,
"step": 1134
},
{
"epoch": 2.40169133192389,
"grad_norm": 4.473540306091309,
"learning_rate": 4.3981252493079506e-06,
"loss": 0.6231296062469482,
"step": 1136
},
{
"epoch": 2.4059196617336154,
"grad_norm": 4.36040735244751,
"learning_rate": 4.382876806203453e-06,
"loss": 0.7281938791275024,
"step": 1138
},
{
"epoch": 2.41014799154334,
"grad_norm": 2.011646032333374,
"learning_rate": 4.3676420203788125e-06,
"loss": 0.878029465675354,
"step": 1140
},
{
"epoch": 2.4143763213530653,
"grad_norm": 1.9183878898620605,
"learning_rate": 4.352421078085405e-06,
"loss": 1.0274100303649902,
"step": 1142
},
{
"epoch": 2.4186046511627906,
"grad_norm": 11.932039260864258,
"learning_rate": 4.337214165405368e-06,
"loss": 0.4374099373817444,
"step": 1144
},
{
"epoch": 2.4228329809725158,
"grad_norm": 12.280984878540039,
"learning_rate": 4.322021468249316e-06,
"loss": 0.6840062141418457,
"step": 1146
},
{
"epoch": 2.427061310782241,
"grad_norm": 2.0083425045013428,
"learning_rate": 4.306843172354076e-06,
"loss": 0.6773014068603516,
"step": 1148
},
{
"epoch": 2.431289640591966,
"grad_norm": 3.0725865364074707,
"learning_rate": 4.291679463280416e-06,
"loss": 0.5143462419509888,
"step": 1150
},
{
"epoch": 2.4355179704016914,
"grad_norm": 9.359436988830566,
"learning_rate": 4.276530526410768e-06,
"loss": 0.5853961706161499,
"step": 1152
},
{
"epoch": 2.4397463002114166,
"grad_norm": 2.0795247554779053,
"learning_rate": 4.261396546946972e-06,
"loss": 0.4346202611923218,
"step": 1154
},
{
"epoch": 2.443974630021142,
"grad_norm": 2.1808338165283203,
"learning_rate": 4.246277709908009e-06,
"loss": 0.9142916202545166,
"step": 1156
},
{
"epoch": 2.448202959830867,
"grad_norm": 2.646580457687378,
"learning_rate": 4.231174200127736e-06,
"loss": 0.48341384530067444,
"step": 1158
},
{
"epoch": 2.452431289640592,
"grad_norm": 1.7327146530151367,
"learning_rate": 4.216086202252624e-06,
"loss": 0.8653247356414795,
"step": 1160
},
{
"epoch": 2.456659619450317,
"grad_norm": 2.3021678924560547,
"learning_rate": 4.201013900739513e-06,
"loss": 0.8690458536148071,
"step": 1162
},
{
"epoch": 2.460887949260042,
"grad_norm": 1.7856725454330444,
"learning_rate": 4.185957479853344e-06,
"loss": 0.6429218053817749,
"step": 1164
},
{
"epoch": 2.4651162790697674,
"grad_norm": 2.6286377906799316,
"learning_rate": 4.1709171236649096e-06,
"loss": 0.9498593807220459,
"step": 1166
},
{
"epoch": 2.4693446088794926,
"grad_norm": 2.5875446796417236,
"learning_rate": 4.155893016048607e-06,
"loss": 1.0248832702636719,
"step": 1168
},
{
"epoch": 2.473572938689218,
"grad_norm": 7.152924060821533,
"learning_rate": 4.140885340680189e-06,
"loss": 0.7057082653045654,
"step": 1170
},
{
"epoch": 2.477801268498943,
"grad_norm": 14.405369758605957,
"learning_rate": 4.125894281034517e-06,
"loss": 0.843054473400116,
"step": 1172
},
{
"epoch": 2.482029598308668,
"grad_norm": 1.7444531917572021,
"learning_rate": 4.110920020383316e-06,
"loss": 0.6087920665740967,
"step": 1174
},
{
"epoch": 2.4862579281183934,
"grad_norm": 2.6002697944641113,
"learning_rate": 4.0959627417929386e-06,
"loss": 0.6467411518096924,
"step": 1176
},
{
"epoch": 2.4904862579281186,
"grad_norm": 2.935264825820923,
"learning_rate": 4.081022628122127e-06,
"loss": 0.8234118223190308,
"step": 1178
},
{
"epoch": 2.4947145877378434,
"grad_norm": 16.244186401367188,
"learning_rate": 4.06609986201977e-06,
"loss": 0.562981128692627,
"step": 1180
},
{
"epoch": 2.4989429175475686,
"grad_norm": 1.124654769897461,
"learning_rate": 4.051194625922681e-06,
"loss": 0.36032021045684814,
"step": 1182
},
{
"epoch": 2.503171247357294,
"grad_norm": 1.4934176206588745,
"learning_rate": 4.036307102053359e-06,
"loss": 0.1417439877986908,
"step": 1184
},
{
"epoch": 2.507399577167019,
"grad_norm": 7.969832420349121,
"learning_rate": 4.021437472417763e-06,
"loss": 0.9294497966766357,
"step": 1186
},
{
"epoch": 2.511627906976744,
"grad_norm": 8.113000869750977,
"learning_rate": 4.006585918803091e-06,
"loss": 0.7197484970092773,
"step": 1188
},
{
"epoch": 2.5158562367864694,
"grad_norm": 7.155160427093506,
"learning_rate": 3.991752622775551e-06,
"loss": 0.6611873507499695,
"step": 1190
},
{
"epoch": 2.5200845665961946,
"grad_norm": 1.6746045351028442,
"learning_rate": 3.97693776567815e-06,
"loss": 0.8050599098205566,
"step": 1192
},
{
"epoch": 2.52431289640592,
"grad_norm": 5.096535682678223,
"learning_rate": 3.962141528628465e-06,
"loss": 0.5711352825164795,
"step": 1194
},
{
"epoch": 2.528541226215645,
"grad_norm": 10.482152938842773,
"learning_rate": 3.947364092516437e-06,
"loss": 0.8540534973144531,
"step": 1196
},
{
"epoch": 2.53276955602537,
"grad_norm": 1.8273385763168335,
"learning_rate": 3.932605638002163e-06,
"loss": 0.8409388661384583,
"step": 1198
},
{
"epoch": 2.536997885835095,
"grad_norm": 9.217585563659668,
"learning_rate": 3.917866345513677e-06,
"loss": 0.6928367614746094,
"step": 1200
},
{
"epoch": 2.54122621564482,
"grad_norm": 2.1136229038238525,
"learning_rate": 3.903146395244749e-06,
"loss": 0.9005012512207031,
"step": 1202
},
{
"epoch": 2.5454545454545454,
"grad_norm": 5.702454090118408,
"learning_rate": 3.888445967152691e-06,
"loss": 0.28177410364151,
"step": 1204
},
{
"epoch": 2.5496828752642706,
"grad_norm": 2.86281418800354,
"learning_rate": 3.873765240956138e-06,
"loss": 0.92289137840271,
"step": 1206
},
{
"epoch": 2.553911205073996,
"grad_norm": 6.114447593688965,
"learning_rate": 3.859104396132863e-06,
"loss": 0.956689715385437,
"step": 1208
},
{
"epoch": 2.558139534883721,
"grad_norm": 2.0390806198120117,
"learning_rate": 3.84446361191759e-06,
"loss": 0.8528469204902649,
"step": 1210
},
{
"epoch": 2.5623678646934462,
"grad_norm": 5.200840473175049,
"learning_rate": 3.829843067299788e-06,
"loss": 0.6572003364562988,
"step": 1212
},
{
"epoch": 2.5665961945031714,
"grad_norm": 2.567246437072754,
"learning_rate": 3.8152429410214834e-06,
"loss": 0.8881071209907532,
"step": 1214
},
{
"epoch": 2.570824524312896,
"grad_norm": 6.34989070892334,
"learning_rate": 3.8006634115750886e-06,
"loss": 0.5154778361320496,
"step": 1216
},
{
"epoch": 2.5750528541226214,
"grad_norm": 3.881256103515625,
"learning_rate": 3.7861046572012105e-06,
"loss": 0.5178666114807129,
"step": 1218
},
{
"epoch": 2.5792811839323466,
"grad_norm": 6.826540946960449,
"learning_rate": 3.771566855886467e-06,
"loss": 0.48407042026519775,
"step": 1220
},
{
"epoch": 2.583509513742072,
"grad_norm": 9.729143142700195,
"learning_rate": 3.7570501853613227e-06,
"loss": 0.5292811393737793,
"step": 1222
},
{
"epoch": 2.587737843551797,
"grad_norm": 7.183631896972656,
"learning_rate": 3.742554823097905e-06,
"loss": 0.7336196899414062,
"step": 1224
},
{
"epoch": 2.5919661733615222,
"grad_norm": 0.5376298427581787,
"learning_rate": 3.7280809463078436e-06,
"loss": 0.27503812313079834,
"step": 1226
},
{
"epoch": 2.5961945031712474,
"grad_norm": 1.9828342199325562,
"learning_rate": 3.7136287319400944e-06,
"loss": 0.7055711150169373,
"step": 1228
},
{
"epoch": 2.6004228329809727,
"grad_norm": 13.654889106750488,
"learning_rate": 3.6991983566787847e-06,
"loss": 0.6512123942375183,
"step": 1230
},
{
"epoch": 2.604651162790698,
"grad_norm": 12.71254825592041,
"learning_rate": 3.6847899969410515e-06,
"loss": 0.649929940700531,
"step": 1232
},
{
"epoch": 2.6088794926004226,
"grad_norm": 4.917657375335693,
"learning_rate": 3.6704038288748812e-06,
"loss": 0.7236469984054565,
"step": 1234
},
{
"epoch": 2.613107822410148,
"grad_norm": 2.205489158630371,
"learning_rate": 3.6560400283569543e-06,
"loss": 0.7024824619293213,
"step": 1236
},
{
"epoch": 2.617336152219873,
"grad_norm": 18.51865005493164,
"learning_rate": 3.641698770990507e-06,
"loss": 0.8929089307785034,
"step": 1238
},
{
"epoch": 2.6215644820295982,
"grad_norm": 2.753260612487793,
"learning_rate": 3.6273802321031738e-06,
"loss": 0.7072829008102417,
"step": 1240
},
{
"epoch": 2.6257928118393234,
"grad_norm": 1.606559157371521,
"learning_rate": 3.6130845867448384e-06,
"loss": 0.8024892807006836,
"step": 1242
},
{
"epoch": 2.6300211416490487,
"grad_norm": 2.0973849296569824,
"learning_rate": 3.598812009685514e-06,
"loss": 0.9532315731048584,
"step": 1244
},
{
"epoch": 2.634249471458774,
"grad_norm": 16.66941261291504,
"learning_rate": 3.584562675413189e-06,
"loss": 0.5999901294708252,
"step": 1246
},
{
"epoch": 2.638477801268499,
"grad_norm": 7.971654891967773,
"learning_rate": 3.5703367581316985e-06,
"loss": 0.491814523935318,
"step": 1248
},
{
"epoch": 2.6427061310782243,
"grad_norm": 1.9864351749420166,
"learning_rate": 3.5561344317585957e-06,
"loss": 0.8885899782180786,
"step": 1250
},
{
"epoch": 2.646934460887949,
"grad_norm": 10.337087631225586,
"learning_rate": 3.5419558699230305e-06,
"loss": 0.41249626874923706,
"step": 1252
},
{
"epoch": 2.6511627906976747,
"grad_norm": 10.842833518981934,
"learning_rate": 3.52780124596361e-06,
"loss": 0.637297511100769,
"step": 1254
},
{
"epoch": 2.6553911205073994,
"grad_norm": 2.5052878856658936,
"learning_rate": 3.5136707329263055e-06,
"loss": 0.9081771373748779,
"step": 1256
},
{
"epoch": 2.6596194503171247,
"grad_norm": 2.8900225162506104,
"learning_rate": 3.499564503562315e-06,
"loss": 0.2507639527320862,
"step": 1258
},
{
"epoch": 2.66384778012685,
"grad_norm": 3.3573076725006104,
"learning_rate": 3.4854827303259575e-06,
"loss": 0.41150540113449097,
"step": 1260
},
{
"epoch": 2.668076109936575,
"grad_norm": 11.89966106414795,
"learning_rate": 3.47142558537257e-06,
"loss": 0.6305783987045288,
"step": 1262
},
{
"epoch": 2.6723044397463003,
"grad_norm": 14.18155574798584,
"learning_rate": 3.4573932405563966e-06,
"loss": 0.5670468211174011,
"step": 1264
},
{
"epoch": 2.6765327695560255,
"grad_norm": 12.071949005126953,
"learning_rate": 3.4433858674284894e-06,
"loss": 0.3618084788322449,
"step": 1266
},
{
"epoch": 2.6807610993657507,
"grad_norm": 2.5437850952148438,
"learning_rate": 3.429403637234614e-06,
"loss": 0.7219557762145996,
"step": 1268
},
{
"epoch": 2.6849894291754755,
"grad_norm": 2.411898136138916,
"learning_rate": 3.4154467209131515e-06,
"loss": 0.869693398475647,
"step": 1270
},
{
"epoch": 2.689217758985201,
"grad_norm": 22.246299743652344,
"learning_rate": 3.401515289093011e-06,
"loss": 0.6177300810813904,
"step": 1272
},
{
"epoch": 2.693446088794926,
"grad_norm": 5.75520133972168,
"learning_rate": 3.3876095120915433e-06,
"loss": 0.4074831008911133,
"step": 1274
},
{
"epoch": 2.697674418604651,
"grad_norm": 4.738722324371338,
"learning_rate": 3.3737295599124592e-06,
"loss": 0.45129725337028503,
"step": 1276
},
{
"epoch": 2.7019027484143763,
"grad_norm": 3.205409049987793,
"learning_rate": 3.3598756022437507e-06,
"loss": 0.8543651103973389,
"step": 1278
},
{
"epoch": 2.7061310782241015,
"grad_norm": 13.158727645874023,
"learning_rate": 3.3460478084556157e-06,
"loss": 0.671482503414154,
"step": 1280
},
{
"epoch": 2.7103594080338267,
"grad_norm": 1.7831779718399048,
"learning_rate": 3.33224634759839e-06,
"loss": 1.0254371166229248,
"step": 1282
},
{
"epoch": 2.714587737843552,
"grad_norm": 1.0492534637451172,
"learning_rate": 3.318471388400476e-06,
"loss": 0.6375106573104858,
"step": 1284
},
{
"epoch": 2.718816067653277,
"grad_norm": 1.1835532188415527,
"learning_rate": 3.3047230992662834e-06,
"loss": 0.959743857383728,
"step": 1286
},
{
"epoch": 2.723044397463002,
"grad_norm": 1.9402508735656738,
"learning_rate": 3.291001648274175e-06,
"loss": 0.43162545561790466,
"step": 1288
},
{
"epoch": 2.7272727272727275,
"grad_norm": 7.228184700012207,
"learning_rate": 3.277307203174395e-06,
"loss": 0.837255597114563,
"step": 1290
},
{
"epoch": 2.7315010570824523,
"grad_norm": 2.536939859390259,
"learning_rate": 3.263639931387039e-06,
"loss": 0.9474502801895142,
"step": 1292
},
{
"epoch": 2.7357293868921775,
"grad_norm": 4.318763732910156,
"learning_rate": 3.2500000000000015e-06,
"loss": 1.027559518814087,
"step": 1294
},
{
"epoch": 2.7399577167019027,
"grad_norm": 2.448118209838867,
"learning_rate": 3.236387575766916e-06,
"loss": 0.3040839433670044,
"step": 1296
},
{
"epoch": 2.744186046511628,
"grad_norm": 2.2139651775360107,
"learning_rate": 3.222802825105138e-06,
"loss": 0.5131670236587524,
"step": 1298
},
{
"epoch": 2.748414376321353,
"grad_norm": 2.004253387451172,
"learning_rate": 3.209245914093708e-06,
"loss": 0.6601057648658752,
"step": 1300
},
{
"epoch": 2.7526427061310783,
"grad_norm": 4.516692638397217,
"learning_rate": 3.1957170084713053e-06,
"loss": 0.6221194267272949,
"step": 1302
},
{
"epoch": 2.7568710359408035,
"grad_norm": 1.7300586700439453,
"learning_rate": 3.182216273634235e-06,
"loss": 0.6846257448196411,
"step": 1304
},
{
"epoch": 2.7610993657505283,
"grad_norm": 17.494178771972656,
"learning_rate": 3.1687438746344086e-06,
"loss": 0.7623063325881958,
"step": 1306
},
{
"epoch": 2.765327695560254,
"grad_norm": 5.773592948913574,
"learning_rate": 3.1552999761773166e-06,
"loss": 0.8817945718765259,
"step": 1308
},
{
"epoch": 2.7695560253699787,
"grad_norm": 3.4069459438323975,
"learning_rate": 3.141884742620013e-06,
"loss": 0.891013503074646,
"step": 1310
},
{
"epoch": 2.773784355179704,
"grad_norm": 2.4098305702209473,
"learning_rate": 3.1284983379691247e-06,
"loss": 0.7703976631164551,
"step": 1312
},
{
"epoch": 2.778012684989429,
"grad_norm": 9.07870864868164,
"learning_rate": 3.1151409258788235e-06,
"loss": 0.4413790702819824,
"step": 1314
},
{
"epoch": 2.7822410147991543,
"grad_norm": 2.415982961654663,
"learning_rate": 3.101812669648838e-06,
"loss": 0.8148735165596008,
"step": 1316
},
{
"epoch": 2.7864693446088795,
"grad_norm": 0.6904446482658386,
"learning_rate": 3.0885137322224546e-06,
"loss": 0.6483764052391052,
"step": 1318
},
{
"epoch": 2.7906976744186047,
"grad_norm": 3.2453792095184326,
"learning_rate": 3.075244276184529e-06,
"loss": 0.8359994292259216,
"step": 1320
},
{
"epoch": 2.79492600422833,
"grad_norm": 8.542537689208984,
"learning_rate": 3.0620044637594936e-06,
"loss": 0.4271160662174225,
"step": 1322
},
{
"epoch": 2.7991543340380547,
"grad_norm": 2.9144785404205322,
"learning_rate": 3.0487944568093685e-06,
"loss": 0.9177171587944031,
"step": 1324
},
{
"epoch": 2.8033826638477803,
"grad_norm": 10.486828804016113,
"learning_rate": 3.035614416831801e-06,
"loss": 0.25211918354034424,
"step": 1326
},
{
"epoch": 2.807610993657505,
"grad_norm": 11.471001625061035,
"learning_rate": 3.0224645049580746e-06,
"loss": 0.37875235080718994,
"step": 1328
},
{
"epoch": 2.8118393234672303,
"grad_norm": 2.8966259956359863,
"learning_rate": 3.0093448819511416e-06,
"loss": 0.5185975432395935,
"step": 1330
},
{
"epoch": 2.8160676532769555,
"grad_norm": 2.494509696960449,
"learning_rate": 2.9962557082036677e-06,
"loss": 0.8701005578041077,
"step": 1332
},
{
"epoch": 2.8202959830866807,
"grad_norm": 2.3917300701141357,
"learning_rate": 2.9831971437360606e-06,
"loss": 0.6798748970031738,
"step": 1334
},
{
"epoch": 2.824524312896406,
"grad_norm": 2.1049065589904785,
"learning_rate": 2.9701693481945132e-06,
"loss": 0.9501597881317139,
"step": 1336
},
{
"epoch": 2.828752642706131,
"grad_norm": 2.220501661300659,
"learning_rate": 2.957172480849064e-06,
"loss": 0.8798794746398926,
"step": 1338
},
{
"epoch": 2.8329809725158563,
"grad_norm": 9.722474098205566,
"learning_rate": 2.9442067005916375e-06,
"loss": 0.6656410694122314,
"step": 1340
},
{
"epoch": 2.8372093023255816,
"grad_norm": 1.7491201162338257,
"learning_rate": 2.931272165934105e-06,
"loss": 0.9249944686889648,
"step": 1342
},
{
"epoch": 2.8414376321353068,
"grad_norm": 1.7615270614624023,
"learning_rate": 2.918369035006351e-06,
"loss": 0.664387583732605,
"step": 1344
},
{
"epoch": 2.8456659619450315,
"grad_norm": 3.221372604370117,
"learning_rate": 2.9054974655543335e-06,
"loss": 0.6025054454803467,
"step": 1346
},
{
"epoch": 2.8498942917547567,
"grad_norm": 1.7131472826004028,
"learning_rate": 2.89265761493816e-06,
"loss": 0.7805812358856201,
"step": 1348
},
{
"epoch": 2.854122621564482,
"grad_norm": 34.04362869262695,
"learning_rate": 2.8798496401301644e-06,
"loss": 0.5436214208602905,
"step": 1350
},
{
"epoch": 2.858350951374207,
"grad_norm": 3.345407009124756,
"learning_rate": 2.8670736977129843e-06,
"loss": 0.9029253721237183,
"step": 1352
},
{
"epoch": 2.8625792811839323,
"grad_norm": 1.611226201057434,
"learning_rate": 2.854329943877647e-06,
"loss": 0.8401460647583008,
"step": 1354
},
{
"epoch": 2.8668076109936576,
"grad_norm": 11.21456241607666,
"learning_rate": 2.841618534421664e-06,
"loss": 0.43195241689682007,
"step": 1356
},
{
"epoch": 2.8710359408033828,
"grad_norm": 2.103454113006592,
"learning_rate": 2.8289396247471227e-06,
"loss": 0.9341262578964233,
"step": 1358
},
{
"epoch": 2.875264270613108,
"grad_norm": 39.83855438232422,
"learning_rate": 2.816293369858788e-06,
"loss": 0.6406150460243225,
"step": 1360
},
{
"epoch": 2.879492600422833,
"grad_norm": 3.1482346057891846,
"learning_rate": 2.803679924362206e-06,
"loss": 0.5738825798034668,
"step": 1362
},
{
"epoch": 2.883720930232558,
"grad_norm": 2.0426697731018066,
"learning_rate": 2.7910994424618165e-06,
"loss": 0.952833890914917,
"step": 1364
},
{
"epoch": 2.887949260042283,
"grad_norm": 4.339751720428467,
"learning_rate": 2.778552077959064e-06,
"loss": 0.7402210235595703,
"step": 1366
},
{
"epoch": 2.8921775898520083,
"grad_norm": 2.1708905696868896,
"learning_rate": 2.766037984250521e-06,
"loss": 0.911862313747406,
"step": 1368
},
{
"epoch": 2.8964059196617336,
"grad_norm": 2.2187283039093018,
"learning_rate": 2.753557314326011e-06,
"loss": 0.8866889476776123,
"step": 1370
},
{
"epoch": 2.9006342494714588,
"grad_norm": 1.8008352518081665,
"learning_rate": 2.7411102207667373e-06,
"loss": 0.9124747514724731,
"step": 1372
},
{
"epoch": 2.904862579281184,
"grad_norm": 1.973971962928772,
"learning_rate": 2.728696855743418e-06,
"loss": 0.7285431623458862,
"step": 1374
},
{
"epoch": 2.909090909090909,
"grad_norm": 2.3564419746398926,
"learning_rate": 2.7163173710144323e-06,
"loss": 0.8938196897506714,
"step": 1376
},
{
"epoch": 2.9133192389006344,
"grad_norm": 10.1366605758667,
"learning_rate": 2.703971917923951e-06,
"loss": 0.6399440765380859,
"step": 1378
},
{
"epoch": 2.9175475687103596,
"grad_norm": 2.226811170578003,
"learning_rate": 2.6916606474000965e-06,
"loss": 0.6401858329772949,
"step": 1380
},
{
"epoch": 2.9217758985200843,
"grad_norm": 8.00619888305664,
"learning_rate": 2.679383709953103e-06,
"loss": 0.9682914018630981,
"step": 1382
},
{
"epoch": 2.92600422832981,
"grad_norm": 5.602720260620117,
"learning_rate": 2.6671412556734578e-06,
"loss": 0.7572795748710632,
"step": 1384
},
{
"epoch": 2.9302325581395348,
"grad_norm": 1.859739065170288,
"learning_rate": 2.6549334342300844e-06,
"loss": 0.5803749561309814,
"step": 1386
},
{
"epoch": 2.93446088794926,
"grad_norm": 4.891067981719971,
"learning_rate": 2.642760394868503e-06,
"loss": 0.8840711712837219,
"step": 1388
},
{
"epoch": 2.938689217758985,
"grad_norm": 1.0822590589523315,
"learning_rate": 2.6306222864090157e-06,
"loss": 0.5688039064407349,
"step": 1390
},
{
"epoch": 2.9429175475687104,
"grad_norm": 3.1064813137054443,
"learning_rate": 2.61851925724487e-06,
"loss": 0.9503481984138489,
"step": 1392
},
{
"epoch": 2.9471458773784356,
"grad_norm": 0.7669625282287598,
"learning_rate": 2.606451455340459e-06,
"loss": 0.22224077582359314,
"step": 1394
},
{
"epoch": 2.951374207188161,
"grad_norm": 1.769733190536499,
"learning_rate": 2.5944190282295155e-06,
"loss": 0.46695834398269653,
"step": 1396
},
{
"epoch": 2.955602536997886,
"grad_norm": 4.761196136474609,
"learning_rate": 2.5824221230132895e-06,
"loss": 0.8505163192749023,
"step": 1398
},
{
"epoch": 2.9598308668076108,
"grad_norm": 1.788468837738037,
"learning_rate": 2.5704608863587676e-06,
"loss": 1.0935367345809937,
"step": 1400
},
{
"epoch": 2.9640591966173364,
"grad_norm": 8.347506523132324,
"learning_rate": 2.558535464496877e-06,
"loss": 0.7321504354476929,
"step": 1402
},
{
"epoch": 2.968287526427061,
"grad_norm": 1.75548255443573,
"learning_rate": 2.5466460032206872e-06,
"loss": 0.45597395300865173,
"step": 1404
},
{
"epoch": 2.9725158562367864,
"grad_norm": 4.281891345977783,
"learning_rate": 2.53479264788364e-06,
"loss": 0.7983198761940002,
"step": 1406
},
{
"epoch": 2.9767441860465116,
"grad_norm": 2.011554479598999,
"learning_rate": 2.522975543397769e-06,
"loss": 0.6962990164756775,
"step": 1408
},
{
"epoch": 2.980972515856237,
"grad_norm": 6.949583053588867,
"learning_rate": 2.511194834231922e-06,
"loss": 0.7349997162818909,
"step": 1410
},
{
"epoch": 2.985200845665962,
"grad_norm": 2.0354180335998535,
"learning_rate": 2.49945066441e-06,
"loss": 0.8892476558685303,
"step": 1412
},
{
"epoch": 2.989429175475687,
"grad_norm": 3.0393824577331543,
"learning_rate": 2.4877431775092004e-06,
"loss": 0.5271695852279663,
"step": 1414
},
{
"epoch": 2.9936575052854124,
"grad_norm": 9.330391883850098,
"learning_rate": 2.4760725166582544e-06,
"loss": 0.8671320676803589,
"step": 1416
},
{
"epoch": 2.997885835095137,
"grad_norm": 4.359357833862305,
"learning_rate": 2.464438824535675e-06,
"loss": 0.4633011221885681,
"step": 1418
},
{
"epoch": 3.0021141649048624,
"grad_norm": 2.4763078689575195,
"learning_rate": 2.452842243368028e-06,
"loss": 0.8404690027236938,
"step": 1420
},
{
"epoch": 3.0063424947145876,
"grad_norm": 4.898531913757324,
"learning_rate": 2.4412829149281735e-06,
"loss": 0.471511572599411,
"step": 1422
},
{
"epoch": 3.010570824524313,
"grad_norm": 3.776346206665039,
"learning_rate": 2.4297609805335472e-06,
"loss": 0.8590734601020813,
"step": 1424
},
{
"epoch": 3.014799154334038,
"grad_norm": 11.123237609863281,
"learning_rate": 2.4182765810444263e-06,
"loss": 0.4775305688381195,
"step": 1426
},
{
"epoch": 3.019027484143763,
"grad_norm": 2.5934579372406006,
"learning_rate": 2.4068298568622083e-06,
"loss": 0.4365171194076538,
"step": 1428
},
{
"epoch": 3.0232558139534884,
"grad_norm": 1.5605335235595703,
"learning_rate": 2.395420947927694e-06,
"loss": 0.64106285572052,
"step": 1430
},
{
"epoch": 3.0274841437632136,
"grad_norm": 2.6134045124053955,
"learning_rate": 2.3840499937193804e-06,
"loss": 0.7672260403633118,
"step": 1432
},
{
"epoch": 3.031712473572939,
"grad_norm": 1.9986504316329956,
"learning_rate": 2.3727171332517496e-06,
"loss": 0.719465970993042,
"step": 1434
},
{
"epoch": 3.035940803382664,
"grad_norm": 10.15622329711914,
"learning_rate": 2.361422505073575e-06,
"loss": 0.47842538356781006,
"step": 1436
},
{
"epoch": 3.040169133192389,
"grad_norm": 0.6120308637619019,
"learning_rate": 2.3501662472662233e-06,
"loss": 0.1397717446088791,
"step": 1438
},
{
"epoch": 3.044397463002114,
"grad_norm": 4.52173376083374,
"learning_rate": 2.3389484974419685e-06,
"loss": 0.7664233446121216,
"step": 1440
},
{
"epoch": 3.048625792811839,
"grad_norm": 2.3592329025268555,
"learning_rate": 2.327769392742309e-06,
"loss": 0.36135226488113403,
"step": 1442
},
{
"epoch": 3.0528541226215644,
"grad_norm": 6.51743745803833,
"learning_rate": 2.316629069836291e-06,
"loss": 0.18552279472351074,
"step": 1444
},
{
"epoch": 3.0570824524312896,
"grad_norm": 6.062836647033691,
"learning_rate": 2.305527664918838e-06,
"loss": 0.5992299318313599,
"step": 1446
},
{
"epoch": 3.061310782241015,
"grad_norm": 19.794702529907227,
"learning_rate": 2.294465313709085e-06,
"loss": 0.6679155826568604,
"step": 1448
},
{
"epoch": 3.06553911205074,
"grad_norm": 2.451261281967163,
"learning_rate": 2.2834421514487205e-06,
"loss": 0.49749624729156494,
"step": 1450
},
{
"epoch": 3.0697674418604652,
"grad_norm": 8.197314262390137,
"learning_rate": 2.27245831290033e-06,
"loss": 0.6167775988578796,
"step": 1452
},
{
"epoch": 3.0739957716701904,
"grad_norm": 2.205322027206421,
"learning_rate": 2.261513932345755e-06,
"loss": 0.7145153284072876,
"step": 1454
},
{
"epoch": 3.0782241014799157,
"grad_norm": 3.412233829498291,
"learning_rate": 2.250609143584443e-06,
"loss": 0.28127458691596985,
"step": 1456
},
{
"epoch": 3.0824524312896404,
"grad_norm": 2.0477826595306396,
"learning_rate": 2.239744079931818e-06,
"loss": 0.5249584913253784,
"step": 1458
},
{
"epoch": 3.0866807610993656,
"grad_norm": 3.4058339595794678,
"learning_rate": 2.228918874217649e-06,
"loss": 0.6693753600120544,
"step": 1460
},
{
"epoch": 3.090909090909091,
"grad_norm": 3.464690923690796,
"learning_rate": 2.2181336587844244e-06,
"loss": 0.7037372589111328,
"step": 1462
},
{
"epoch": 3.095137420718816,
"grad_norm": 2.2315657138824463,
"learning_rate": 2.207388565485736e-06,
"loss": 0.4602961540222168,
"step": 1464
},
{
"epoch": 3.0993657505285412,
"grad_norm": 2.447293758392334,
"learning_rate": 2.196683725684666e-06,
"loss": 0.6293976902961731,
"step": 1466
},
{
"epoch": 3.1035940803382664,
"grad_norm": 2.123246431350708,
"learning_rate": 2.1860192702521826e-06,
"loss": 0.7898637056350708,
"step": 1468
},
{
"epoch": 3.1078224101479917,
"grad_norm": 35.95222854614258,
"learning_rate": 2.1753953295655354e-06,
"loss": 0.5975048542022705,
"step": 1470
},
{
"epoch": 3.112050739957717,
"grad_norm": 5.03873872756958,
"learning_rate": 2.164812033506673e-06,
"loss": 0.5046147108078003,
"step": 1472
},
{
"epoch": 3.116279069767442,
"grad_norm": 8.631458282470703,
"learning_rate": 2.1542695114606367e-06,
"loss": 0.787243127822876,
"step": 1474
},
{
"epoch": 3.120507399577167,
"grad_norm": 1.7753877639770508,
"learning_rate": 2.1437678923139944e-06,
"loss": 0.6715502738952637,
"step": 1476
},
{
"epoch": 3.124735729386892,
"grad_norm": 18.50299644470215,
"learning_rate": 2.1333073044532628e-06,
"loss": 0.2992890775203705,
"step": 1478
},
{
"epoch": 3.1289640591966172,
"grad_norm": 3.513249158859253,
"learning_rate": 2.122887875763327e-06,
"loss": 0.7942399382591248,
"step": 1480
},
{
"epoch": 3.1331923890063424,
"grad_norm": 4.814080238342285,
"learning_rate": 2.1125097336258897e-06,
"loss": 0.680347740650177,
"step": 1482
},
{
"epoch": 3.1374207188160677,
"grad_norm": 4.051783084869385,
"learning_rate": 2.1021730049179117e-06,
"loss": 0.24208950996398926,
"step": 1484
},
{
"epoch": 3.141649048625793,
"grad_norm": 11.636868476867676,
"learning_rate": 2.09187781601005e-06,
"loss": 0.8715755939483643,
"step": 1486
},
{
"epoch": 3.145877378435518,
"grad_norm": 7.1174492835998535,
"learning_rate": 2.0816242927651254e-06,
"loss": 0.8316367268562317,
"step": 1488
},
{
"epoch": 3.1501057082452433,
"grad_norm": 6.379950523376465,
"learning_rate": 2.071412560536581e-06,
"loss": 0.6412270069122314,
"step": 1490
},
{
"epoch": 3.1543340380549685,
"grad_norm": 11.442061424255371,
"learning_rate": 2.0612427441669475e-06,
"loss": 0.61041259765625,
"step": 1492
},
{
"epoch": 3.1585623678646932,
"grad_norm": 3.2749340534210205,
"learning_rate": 2.051114967986311e-06,
"loss": 0.3159480094909668,
"step": 1494
},
{
"epoch": 3.1627906976744184,
"grad_norm": 0.2995913326740265,
"learning_rate": 2.0410293558108074e-06,
"loss": 0.463113397359848,
"step": 1496
},
{
"epoch": 3.1670190274841437,
"grad_norm": 1.7154912948608398,
"learning_rate": 2.0309860309410986e-06,
"loss": 0.5846034288406372,
"step": 1498
},
{
"epoch": 3.171247357293869,
"grad_norm": 2.439157724380493,
"learning_rate": 2.0209851161608613e-06,
"loss": 0.7909647226333618,
"step": 1500
},
{
"epoch": 3.175475687103594,
"grad_norm": 9.485970497131348,
"learning_rate": 2.011026733735301e-06,
"loss": 0.2775943875312805,
"step": 1502
},
{
"epoch": 3.1797040169133193,
"grad_norm": 1.9676028490066528,
"learning_rate": 2.001111005409643e-06,
"loss": 0.311935693025589,
"step": 1504
},
{
"epoch": 3.1839323467230445,
"grad_norm": 2.517533779144287,
"learning_rate": 1.9912380524076503e-06,
"loss": 0.5752232670783997,
"step": 1506
},
{
"epoch": 3.1881606765327697,
"grad_norm": 93.28540802001953,
"learning_rate": 1.981407995430136e-06,
"loss": 0.3040233254432678,
"step": 1508
},
{
"epoch": 3.192389006342495,
"grad_norm": 16.651472091674805,
"learning_rate": 1.9716209546535e-06,
"loss": 0.18534038960933685,
"step": 1510
},
{
"epoch": 3.1966173361522197,
"grad_norm": 0.5653613209724426,
"learning_rate": 1.9618770497282487e-06,
"loss": 0.061203911900520325,
"step": 1512
},
{
"epoch": 3.200845665961945,
"grad_norm": 6.109493255615234,
"learning_rate": 1.9521763997775294e-06,
"loss": 0.5888121128082275,
"step": 1514
},
{
"epoch": 3.20507399577167,
"grad_norm": 2.6389000415802,
"learning_rate": 1.942519123395689e-06,
"loss": 0.7307941913604736,
"step": 1516
},
{
"epoch": 3.2093023255813953,
"grad_norm": 3.343299627304077,
"learning_rate": 1.932905338646812e-06,
"loss": 0.48864707350730896,
"step": 1518
},
{
"epoch": 3.2135306553911205,
"grad_norm": 1.9590729475021362,
"learning_rate": 1.9233351630632747e-06,
"loss": 0.6144390106201172,
"step": 1520
},
{
"epoch": 3.2177589852008457,
"grad_norm": 2.438772439956665,
"learning_rate": 1.9138087136443227e-06,
"loss": 0.8172694444656372,
"step": 1522
},
{
"epoch": 3.221987315010571,
"grad_norm": 5.089728832244873,
"learning_rate": 1.904326106854626e-06,
"loss": 0.6827787160873413,
"step": 1524
},
{
"epoch": 3.226215644820296,
"grad_norm": 2.1171646118164062,
"learning_rate": 1.8948874586228622e-06,
"loss": 0.6316916942596436,
"step": 1526
},
{
"epoch": 3.2304439746300213,
"grad_norm": 8.41878604888916,
"learning_rate": 1.8854928843402987e-06,
"loss": 0.17520464956760406,
"step": 1528
},
{
"epoch": 3.234672304439746,
"grad_norm": 5.3887553215026855,
"learning_rate": 1.8761424988593796e-06,
"loss": 0.7151615619659424,
"step": 1530
},
{
"epoch": 3.2389006342494713,
"grad_norm": 4.3347249031066895,
"learning_rate": 1.8668364164923248e-06,
"loss": 0.7036644220352173,
"step": 1532
},
{
"epoch": 3.2431289640591965,
"grad_norm": 2.4792866706848145,
"learning_rate": 1.8575747510097294e-06,
"loss": 0.7002212405204773,
"step": 1534
},
{
"epoch": 3.2473572938689217,
"grad_norm": 3.4107108116149902,
"learning_rate": 1.8483576156391747e-06,
"loss": 0.5895048975944519,
"step": 1536
},
{
"epoch": 3.251585623678647,
"grad_norm": 2.342998743057251,
"learning_rate": 1.8391851230638443e-06,
"loss": 0.584649920463562,
"step": 1538
},
{
"epoch": 3.255813953488372,
"grad_norm": 2.1657042503356934,
"learning_rate": 1.8300573854211452e-06,
"loss": 0.36975300312042236,
"step": 1540
},
{
"epoch": 3.2600422832980973,
"grad_norm": 10.384785652160645,
"learning_rate": 1.8209745143013363e-06,
"loss": 0.9986827373504639,
"step": 1542
},
{
"epoch": 3.2642706131078225,
"grad_norm": 1.9336402416229248,
"learning_rate": 1.811936620746167e-06,
"loss": 0.7489902973175049,
"step": 1544
},
{
"epoch": 3.2684989429175477,
"grad_norm": 3.454608201980591,
"learning_rate": 1.8029438152475159e-06,
"loss": 0.7176718711853027,
"step": 1546
},
{
"epoch": 3.2727272727272725,
"grad_norm": 2.3592514991760254,
"learning_rate": 1.7939962077460448e-06,
"loss": 0.22781792283058167,
"step": 1548
},
{
"epoch": 3.276955602536998,
"grad_norm": 4.563205242156982,
"learning_rate": 1.785093907629849e-06,
"loss": 0.7862043380737305,
"step": 1550
},
{
"epoch": 3.281183932346723,
"grad_norm": 2.5415520668029785,
"learning_rate": 1.7762370237331233e-06,
"loss": 0.7220112085342407,
"step": 1552
},
{
"epoch": 3.285412262156448,
"grad_norm": 0.19669117033481598,
"learning_rate": 1.7674256643348348e-06,
"loss": 0.0445966012775898,
"step": 1554
},
{
"epoch": 3.2896405919661733,
"grad_norm": 1.9330148696899414,
"learning_rate": 1.7586599371573887e-06,
"loss": 0.7044053077697754,
"step": 1556
},
{
"epoch": 3.2938689217758985,
"grad_norm": 3.323208808898926,
"learning_rate": 1.7499399493653225e-06,
"loss": 0.4170236587524414,
"step": 1558
},
{
"epoch": 3.2980972515856237,
"grad_norm": 8.253467559814453,
"learning_rate": 1.7412658075639921e-06,
"loss": 0.8996248245239258,
"step": 1560
},
{
"epoch": 3.302325581395349,
"grad_norm": 6.560991287231445,
"learning_rate": 1.7326376177982635e-06,
"loss": 0.6192322373390198,
"step": 1562
},
{
"epoch": 3.306553911205074,
"grad_norm": 2.806262254714966,
"learning_rate": 1.7240554855512225e-06,
"loss": 0.37417668104171753,
"step": 1564
},
{
"epoch": 3.3107822410147993,
"grad_norm": 2.6199615001678467,
"learning_rate": 1.7155195157428878e-06,
"loss": 0.48499739170074463,
"step": 1566
},
{
"epoch": 3.3150105708245245,
"grad_norm": 1.073570966720581,
"learning_rate": 1.707029812728916e-06,
"loss": 0.09887000173330307,
"step": 1568
},
{
"epoch": 3.3192389006342493,
"grad_norm": 6.881595134735107,
"learning_rate": 1.6985864802993366e-06,
"loss": 0.44235411286354065,
"step": 1570
},
{
"epoch": 3.3234672304439745,
"grad_norm": 29.780031204223633,
"learning_rate": 1.690189621677284e-06,
"loss": 0.5956472754478455,
"step": 1572
},
{
"epoch": 3.3276955602536997,
"grad_norm": 9.961191177368164,
"learning_rate": 1.681839339517728e-06,
"loss": 0.24543781578540802,
"step": 1574
},
{
"epoch": 3.331923890063425,
"grad_norm": 7.7419867515563965,
"learning_rate": 1.67353573590622e-06,
"loss": 0.2815524935722351,
"step": 1576
},
{
"epoch": 3.33615221987315,
"grad_norm": 1.9094469547271729,
"learning_rate": 1.6652789123576513e-06,
"loss": 0.41532450914382935,
"step": 1578
},
{
"epoch": 3.3403805496828753,
"grad_norm": 2.138737916946411,
"learning_rate": 1.6570689698150094e-06,
"loss": 0.40520399808883667,
"step": 1580
},
{
"epoch": 3.3446088794926006,
"grad_norm": 3.7974910736083984,
"learning_rate": 1.6489060086481373e-06,
"loss": 0.7991803884506226,
"step": 1582
},
{
"epoch": 3.3488372093023258,
"grad_norm": 12.797378540039062,
"learning_rate": 1.6407901286525152e-06,
"loss": 0.5573571920394897,
"step": 1584
},
{
"epoch": 3.353065539112051,
"grad_norm": 5.1685709953308105,
"learning_rate": 1.6327214290480382e-06,
"loss": 0.4989955425262451,
"step": 1586
},
{
"epoch": 3.3572938689217757,
"grad_norm": 8.31866455078125,
"learning_rate": 1.6247000084778006e-06,
"loss": 0.7689661979675293,
"step": 1588
},
{
"epoch": 3.361522198731501,
"grad_norm": 2.3627467155456543,
"learning_rate": 1.6167259650068887e-06,
"loss": 0.28058159351348877,
"step": 1590
},
{
"epoch": 3.365750528541226,
"grad_norm": 2.8095290660858154,
"learning_rate": 1.6087993961211904e-06,
"loss": 0.6468945741653442,
"step": 1592
},
{
"epoch": 3.3699788583509513,
"grad_norm": 5.985776901245117,
"learning_rate": 1.6009203987261957e-06,
"loss": 0.19938525557518005,
"step": 1594
},
{
"epoch": 3.3742071881606766,
"grad_norm": 2.029069185256958,
"learning_rate": 1.593089069145809e-06,
"loss": 0.7038938999176025,
"step": 1596
},
{
"epoch": 3.3784355179704018,
"grad_norm": 13.093486785888672,
"learning_rate": 1.5853055031211844e-06,
"loss": 0.536240816116333,
"step": 1598
},
{
"epoch": 3.382663847780127,
"grad_norm": 1.8641756772994995,
"learning_rate": 1.577569795809544e-06,
"loss": 0.2785604000091553,
"step": 1600
},
{
"epoch": 3.386892177589852,
"grad_norm": 3.5723180770874023,
"learning_rate": 1.5698820417830137e-06,
"loss": 0.49493706226348877,
"step": 1602
},
{
"epoch": 3.3911205073995774,
"grad_norm": 2.9709901809692383,
"learning_rate": 1.5622423350274784e-06,
"loss": 0.7168706059455872,
"step": 1604
},
{
"epoch": 3.395348837209302,
"grad_norm": 2.5602550506591797,
"learning_rate": 1.5546507689414217e-06,
"loss": 0.7720810174942017,
"step": 1606
},
{
"epoch": 3.3995771670190273,
"grad_norm": 5.38353157043457,
"learning_rate": 1.5471074363347894e-06,
"loss": 0.5143139958381653,
"step": 1608
},
{
"epoch": 3.4038054968287526,
"grad_norm": 2.45029878616333,
"learning_rate": 1.539612429427852e-06,
"loss": 0.7599624991416931,
"step": 1610
},
{
"epoch": 3.4080338266384778,
"grad_norm": 3.154144525527954,
"learning_rate": 1.5321658398500808e-06,
"loss": 0.7236141562461853,
"step": 1612
},
{
"epoch": 3.412262156448203,
"grad_norm": 4.617706298828125,
"learning_rate": 1.5247677586390255e-06,
"loss": 0.5125030279159546,
"step": 1614
},
{
"epoch": 3.416490486257928,
"grad_norm": 2.2036054134368896,
"learning_rate": 1.5174182762392007e-06,
"loss": 0.7255399227142334,
"step": 1616
},
{
"epoch": 3.4207188160676534,
"grad_norm": 1.9959144592285156,
"learning_rate": 1.5101174825009815e-06,
"loss": 0.8192509412765503,
"step": 1618
},
{
"epoch": 3.4249471458773786,
"grad_norm": 9.368083000183105,
"learning_rate": 1.5028654666795042e-06,
"loss": 0.7886937260627747,
"step": 1620
},
{
"epoch": 3.429175475687104,
"grad_norm": 5.33017635345459,
"learning_rate": 1.495662317433576e-06,
"loss": 0.4347168803215027,
"step": 1622
},
{
"epoch": 3.4334038054968286,
"grad_norm": 2.857900381088257,
"learning_rate": 1.4885081228245909e-06,
"loss": 0.8163563013076782,
"step": 1624
},
{
"epoch": 3.4376321353065538,
"grad_norm": 3.188650369644165,
"learning_rate": 1.4814029703154515e-06,
"loss": 0.6366214752197266,
"step": 1626
},
{
"epoch": 3.441860465116279,
"grad_norm": 2.068962812423706,
"learning_rate": 1.4743469467695017e-06,
"loss": 0.84031742811203,
"step": 1628
},
{
"epoch": 3.446088794926004,
"grad_norm": 4.269415378570557,
"learning_rate": 1.4673401384494651e-06,
"loss": 0.4532480239868164,
"step": 1630
},
{
"epoch": 3.4503171247357294,
"grad_norm": 3.9718689918518066,
"learning_rate": 1.4603826310163874e-06,
"loss": 0.6922460794448853,
"step": 1632
},
{
"epoch": 3.4545454545454546,
"grad_norm": 2.5951242446899414,
"learning_rate": 1.453474509528593e-06,
"loss": 0.5470925569534302,
"step": 1634
},
{
"epoch": 3.45877378435518,
"grad_norm": 2.000901222229004,
"learning_rate": 1.446615858440642e-06,
"loss": 0.7648042440414429,
"step": 1636
},
{
"epoch": 3.463002114164905,
"grad_norm": 0.5785875916481018,
"learning_rate": 1.4398067616022993e-06,
"loss": 0.2775920033454895,
"step": 1638
},
{
"epoch": 3.46723044397463,
"grad_norm": 5.065203666687012,
"learning_rate": 1.43304730225751e-06,
"loss": 0.4461028575897217,
"step": 1640
},
{
"epoch": 3.471458773784355,
"grad_norm": 1.031956672668457,
"learning_rate": 1.4263375630433802e-06,
"loss": 0.16223891079425812,
"step": 1642
},
{
"epoch": 3.47568710359408,
"grad_norm": 2.613997220993042,
"learning_rate": 1.4196776259891665e-06,
"loss": 0.6519019603729248,
"step": 1644
},
{
"epoch": 3.4799154334038054,
"grad_norm": 3.478142023086548,
"learning_rate": 1.4130675725152762e-06,
"loss": 0.7469334602355957,
"step": 1646
},
{
"epoch": 3.4841437632135306,
"grad_norm": 0.19091814756393433,
"learning_rate": 1.4065074834322677e-06,
"loss": 0.3862321376800537,
"step": 1648
},
{
"epoch": 3.488372093023256,
"grad_norm": 2.608109474182129,
"learning_rate": 1.3999974389398669e-06,
"loss": 0.3519095480442047,
"step": 1650
},
{
"epoch": 3.492600422832981,
"grad_norm": 31.108787536621094,
"learning_rate": 1.3935375186259817e-06,
"loss": 0.5058547854423523,
"step": 1652
},
{
"epoch": 3.496828752642706,
"grad_norm": 5.511758327484131,
"learning_rate": 1.3871278014657332e-06,
"loss": 0.8334790468215942,
"step": 1654
},
{
"epoch": 3.5010570824524314,
"grad_norm": 4.244947910308838,
"learning_rate": 1.3807683658204915e-06,
"loss": 0.6136918663978577,
"step": 1656
},
{
"epoch": 3.5052854122621566,
"grad_norm": 0.6166185140609741,
"learning_rate": 1.3744592894369093e-06,
"loss": 0.37619149684906006,
"step": 1658
},
{
"epoch": 3.5095137420718814,
"grad_norm": 1.480721354484558,
"learning_rate": 1.3682006494459806e-06,
"loss": 0.5119225382804871,
"step": 1660
},
{
"epoch": 3.513742071881607,
"grad_norm": 4.339591026306152,
"learning_rate": 1.361992522362096e-06,
"loss": 0.4678526222705841,
"step": 1662
},
{
"epoch": 3.517970401691332,
"grad_norm": 2.2856690883636475,
"learning_rate": 1.355834984082101e-06,
"loss": 0.6813148856163025,
"step": 1664
},
{
"epoch": 3.522198731501057,
"grad_norm": 4.048765182495117,
"learning_rate": 1.3497281098843746e-06,
"loss": 0.8450045585632324,
"step": 1666
},
{
"epoch": 3.526427061310782,
"grad_norm": 0.35447490215301514,
"learning_rate": 1.34367197442791e-06,
"loss": 0.11628832668066025,
"step": 1668
},
{
"epoch": 3.5306553911205074,
"grad_norm": 0.5413293242454529,
"learning_rate": 1.3376666517513931e-06,
"loss": 0.5976459980010986,
"step": 1670
},
{
"epoch": 3.5348837209302326,
"grad_norm": 2.9492340087890625,
"learning_rate": 1.3317122152723078e-06,
"loss": 0.45478200912475586,
"step": 1672
},
{
"epoch": 3.539112050739958,
"grad_norm": 7.3628644943237305,
"learning_rate": 1.3258087377860323e-06,
"loss": 0.7129542231559753,
"step": 1674
},
{
"epoch": 3.543340380549683,
"grad_norm": 5.348959922790527,
"learning_rate": 1.3199562914649519e-06,
"loss": 0.5657470226287842,
"step": 1676
},
{
"epoch": 3.547568710359408,
"grad_norm": 2.3247578144073486,
"learning_rate": 1.314154947857571e-06,
"loss": 0.5544891357421875,
"step": 1678
},
{
"epoch": 3.5517970401691334,
"grad_norm": 1.6387786865234375,
"learning_rate": 1.3084047778876487e-06,
"loss": 0.3952558636665344,
"step": 1680
},
{
"epoch": 3.556025369978858,
"grad_norm": 2.0331220626831055,
"learning_rate": 1.3027058518533228e-06,
"loss": 0.5185011625289917,
"step": 1682
},
{
"epoch": 3.5602536997885834,
"grad_norm": 3.5141680240631104,
"learning_rate": 1.2970582394262516e-06,
"loss": 0.42228052020072937,
"step": 1684
},
{
"epoch": 3.5644820295983086,
"grad_norm": 28.404787063598633,
"learning_rate": 1.291462009650768e-06,
"loss": 0.21207627654075623,
"step": 1686
},
{
"epoch": 3.568710359408034,
"grad_norm": 3.038224458694458,
"learning_rate": 1.2859172309430285e-06,
"loss": 0.564752459526062,
"step": 1688
},
{
"epoch": 3.572938689217759,
"grad_norm": 0.07917769253253937,
"learning_rate": 1.280423971090181e-06,
"loss": 0.15166008472442627,
"step": 1690
},
{
"epoch": 3.5771670190274842,
"grad_norm": 3.0425899028778076,
"learning_rate": 1.2749822972495333e-06,
"loss": 0.3606494069099426,
"step": 1692
},
{
"epoch": 3.5813953488372094,
"grad_norm": 22.190839767456055,
"learning_rate": 1.2695922759477356e-06,
"loss": 0.43272876739501953,
"step": 1694
},
{
"epoch": 3.585623678646934,
"grad_norm": 4.985686302185059,
"learning_rate": 1.2642539730799622e-06,
"loss": 0.3359771966934204,
"step": 1696
},
{
"epoch": 3.58985200845666,
"grad_norm": 2.5660486221313477,
"learning_rate": 1.2589674539091115e-06,
"loss": 0.7319288849830627,
"step": 1698
},
{
"epoch": 3.5940803382663846,
"grad_norm": 2.467815637588501,
"learning_rate": 1.2537327830650042e-06,
"loss": 1.010713815689087,
"step": 1700
},
{
"epoch": 3.59830866807611,
"grad_norm": 10.793841361999512,
"learning_rate": 1.248550024543594e-06,
"loss": 0.7237288951873779,
"step": 1702
},
{
"epoch": 3.602536997885835,
"grad_norm": 31.189842224121094,
"learning_rate": 1.2434192417061862e-06,
"loss": 0.40026915073394775,
"step": 1704
},
{
"epoch": 3.6067653276955602,
"grad_norm": 2.291203737258911,
"learning_rate": 1.2383404972786625e-06,
"loss": 0.3644050657749176,
"step": 1706
},
{
"epoch": 3.6109936575052854,
"grad_norm": 13.508527755737305,
"learning_rate": 1.2333138533507139e-06,
"loss": 0.6461091637611389,
"step": 1708
},
{
"epoch": 3.6152219873150107,
"grad_norm": 119.67933654785156,
"learning_rate": 1.2283393713750816e-06,
"loss": 0.5887095928192139,
"step": 1710
},
{
"epoch": 3.619450317124736,
"grad_norm": 2.7162024974823,
"learning_rate": 1.2234171121668069e-06,
"loss": 0.37304478883743286,
"step": 1712
},
{
"epoch": 3.6236786469344606,
"grad_norm": 0.8481037616729736,
"learning_rate": 1.2185471359024855e-06,
"loss": 0.23839226365089417,
"step": 1714
},
{
"epoch": 3.6279069767441863,
"grad_norm": 4.298794269561768,
"learning_rate": 1.2137295021195341e-06,
"loss": 0.5106943845748901,
"step": 1716
},
{
"epoch": 3.632135306553911,
"grad_norm": 0.8728058934211731,
"learning_rate": 1.2089642697154613e-06,
"loss": 0.4261704087257385,
"step": 1718
},
{
"epoch": 3.6363636363636362,
"grad_norm": 42.821006774902344,
"learning_rate": 1.2042514969471475e-06,
"loss": 0.5185738205909729,
"step": 1720
},
{
"epoch": 3.6405919661733614,
"grad_norm": 2.976154088973999,
"learning_rate": 1.1995912414301329e-06,
"loss": 0.667026162147522,
"step": 1722
},
{
"epoch": 3.6448202959830867,
"grad_norm": 4.096918106079102,
"learning_rate": 1.1949835601379144e-06,
"loss": 0.5285245180130005,
"step": 1724
},
{
"epoch": 3.649048625792812,
"grad_norm": 8.217414855957031,
"learning_rate": 1.1904285094012463e-06,
"loss": 0.7697716951370239,
"step": 1726
},
{
"epoch": 3.653276955602537,
"grad_norm": 2.2785675525665283,
"learning_rate": 1.185926144907454e-06,
"loss": 0.7511149644851685,
"step": 1728
},
{
"epoch": 3.6575052854122623,
"grad_norm": 20.448144912719727,
"learning_rate": 1.1814765216997522e-06,
"loss": 0.8001516461372375,
"step": 1730
},
{
"epoch": 3.6617336152219875,
"grad_norm": 4.07733154296875,
"learning_rate": 1.1770796941765727e-06,
"loss": 0.7745957374572754,
"step": 1732
},
{
"epoch": 3.6659619450317127,
"grad_norm": 3.162168502807617,
"learning_rate": 1.1727357160908978e-06,
"loss": 0.6833571195602417,
"step": 1734
},
{
"epoch": 3.6701902748414374,
"grad_norm": 2.275343418121338,
"learning_rate": 1.1684446405496052e-06,
"loss": 0.7999974489212036,
"step": 1736
},
{
"epoch": 3.6744186046511627,
"grad_norm": 9.438085556030273,
"learning_rate": 1.1642065200128179e-06,
"loss": 0.15023943781852722,
"step": 1738
},
{
"epoch": 3.678646934460888,
"grad_norm": 4.421003341674805,
"learning_rate": 1.1600214062932613e-06,
"loss": 0.39063918590545654,
"step": 1740
},
{
"epoch": 3.682875264270613,
"grad_norm": 5.524152755737305,
"learning_rate": 1.1558893505556315e-06,
"loss": 0.36235326528549194,
"step": 1742
},
{
"epoch": 3.6871035940803383,
"grad_norm": 2.831690549850464,
"learning_rate": 1.1518104033159715e-06,
"loss": 0.8054654598236084,
"step": 1744
},
{
"epoch": 3.6913319238900635,
"grad_norm": 0.30287468433380127,
"learning_rate": 1.1477846144410493e-06,
"loss": 0.34930431842803955,
"step": 1746
},
{
"epoch": 3.6955602536997887,
"grad_norm": 11.15501880645752,
"learning_rate": 1.1438120331477506e-06,
"loss": 0.48356789350509644,
"step": 1748
},
{
"epoch": 3.699788583509514,
"grad_norm": 5.449039459228516,
"learning_rate": 1.1398927080024796e-06,
"loss": 0.9134684801101685,
"step": 1750
},
{
"epoch": 3.704016913319239,
"grad_norm": 6.509265899658203,
"learning_rate": 1.1360266869205593e-06,
"loss": 0.299058198928833,
"step": 1752
},
{
"epoch": 3.708245243128964,
"grad_norm": 5.66325044631958,
"learning_rate": 1.1322140171656507e-06,
"loss": 0.4870682954788208,
"step": 1754
},
{
"epoch": 3.712473572938689,
"grad_norm": 1.768808364868164,
"learning_rate": 1.1284547453491734e-06,
"loss": 0.6303389072418213,
"step": 1756
},
{
"epoch": 3.7167019027484143,
"grad_norm": 5.996204853057861,
"learning_rate": 1.1247489174297372e-06,
"loss": 0.4302939474582672,
"step": 1758
},
{
"epoch": 3.7209302325581395,
"grad_norm": 0.5048099160194397,
"learning_rate": 1.1210965787125743e-06,
"loss": 0.33148616552352905,
"step": 1760
},
{
"epoch": 3.7251585623678647,
"grad_norm": 2.1203019618988037,
"learning_rate": 1.1174977738489955e-06,
"loss": 0.6639598608016968,
"step": 1762
},
{
"epoch": 3.72938689217759,
"grad_norm": 2.360515832901001,
"learning_rate": 1.113952546835835e-06,
"loss": 0.4429587721824646,
"step": 1764
},
{
"epoch": 3.733615221987315,
"grad_norm": 5.73626708984375,
"learning_rate": 1.1104609410149174e-06,
"loss": 0.6377854347229004,
"step": 1766
},
{
"epoch": 3.7378435517970403,
"grad_norm": 3.096515417098999,
"learning_rate": 1.1070229990725262e-06,
"loss": 0.6642612218856812,
"step": 1768
},
{
"epoch": 3.7420718816067655,
"grad_norm": 1.1408841609954834,
"learning_rate": 1.1036387630388841e-06,
"loss": 0.09178274869918823,
"step": 1770
},
{
"epoch": 3.7463002114164903,
"grad_norm": 3.048360824584961,
"learning_rate": 1.100308274287636e-06,
"loss": 0.4178808331489563,
"step": 1772
},
{
"epoch": 3.7505285412262155,
"grad_norm": 4.978333473205566,
"learning_rate": 1.0970315735353434e-06,
"loss": 0.878942608833313,
"step": 1774
},
{
"epoch": 3.7547568710359407,
"grad_norm": 3.355879068374634,
"learning_rate": 1.0938087008409909e-06,
"loss": 0.8126663565635681,
"step": 1776
},
{
"epoch": 3.758985200845666,
"grad_norm": 1.848779559135437,
"learning_rate": 1.0906396956054919e-06,
"loss": 0.6619945764541626,
"step": 1778
},
{
"epoch": 3.763213530655391,
"grad_norm": 2.604274272918701,
"learning_rate": 1.0875245965712077e-06,
"loss": 0.809323251247406,
"step": 1780
},
{
"epoch": 3.7674418604651163,
"grad_norm": 2.428056001663208,
"learning_rate": 1.0844634418214763e-06,
"loss": 0.6543502807617188,
"step": 1782
},
{
"epoch": 3.7716701902748415,
"grad_norm": 7.606510639190674,
"learning_rate": 1.0814562687801448e-06,
"loss": 0.15582875907421112,
"step": 1784
},
{
"epoch": 3.7758985200845667,
"grad_norm": 5.435375213623047,
"learning_rate": 1.0785031142111103e-06,
"loss": 0.4742170572280884,
"step": 1786
},
{
"epoch": 3.780126849894292,
"grad_norm": 26.589111328125,
"learning_rate": 1.075604014217875e-06,
"loss": 0.413571298122406,
"step": 1788
},
{
"epoch": 3.7843551797040167,
"grad_norm": 9.778313636779785,
"learning_rate": 1.0727590042431009e-06,
"loss": 0.2072286307811737,
"step": 1790
},
{
"epoch": 3.7885835095137423,
"grad_norm": 0.5479719042778015,
"learning_rate": 1.0699681190681777e-06,
"loss": 0.4818591773509979,
"step": 1792
},
{
"epoch": 3.792811839323467,
"grad_norm": 2.1295125484466553,
"learning_rate": 1.067231392812798e-06,
"loss": 0.732568621635437,
"step": 1794
},
{
"epoch": 3.7970401691331923,
"grad_norm": 3.629566192626953,
"learning_rate": 1.0645488589345396e-06,
"loss": 0.5533275604248047,
"step": 1796
},
{
"epoch": 3.8012684989429175,
"grad_norm": 6.110041618347168,
"learning_rate": 1.0619205502284574e-06,
"loss": 0.8026322722434998,
"step": 1798
},
{
"epoch": 3.8054968287526427,
"grad_norm": 2.935824394226074,
"learning_rate": 1.0593464988266811e-06,
"loss": 0.7134299278259277,
"step": 1800
},
{
"epoch": 3.809725158562368,
"grad_norm": 2.8795206546783447,
"learning_rate": 1.0568267361980225e-06,
"loss": 0.5710768699645996,
"step": 1802
},
{
"epoch": 3.813953488372093,
"grad_norm": 2.6274962425231934,
"learning_rate": 1.0543612931475935e-06,
"loss": 0.7541733384132385,
"step": 1804
},
{
"epoch": 3.8181818181818183,
"grad_norm": 2.3090546131134033,
"learning_rate": 1.0519501998164247e-06,
"loss": 0.873486340045929,
"step": 1806
},
{
"epoch": 3.822410147991543,
"grad_norm": 275.98785400390625,
"learning_rate": 1.049593485681101e-06,
"loss": 0.170283704996109,
"step": 1808
},
{
"epoch": 3.8266384778012688,
"grad_norm": 3.3445799350738525,
"learning_rate": 1.0472911795533997e-06,
"loss": 0.7886514067649841,
"step": 1810
},
{
"epoch": 3.8308668076109935,
"grad_norm": 3.1293160915374756,
"learning_rate": 1.045043309579938e-06,
"loss": 0.8562047481536865,
"step": 1812
},
{
"epoch": 3.8350951374207187,
"grad_norm": 6.491039752960205,
"learning_rate": 1.0428499032418297e-06,
"loss": 0.6334226131439209,
"step": 1814
},
{
"epoch": 3.839323467230444,
"grad_norm": 2.520296573638916,
"learning_rate": 1.0407109873543484e-06,
"loss": 0.569333553314209,
"step": 1816
},
{
"epoch": 3.843551797040169,
"grad_norm": 12.923444747924805,
"learning_rate": 1.0386265880665996e-06,
"loss": 0.42684608697891235,
"step": 1818
},
{
"epoch": 3.8477801268498943,
"grad_norm": 2.3961009979248047,
"learning_rate": 1.036596730861203e-06,
"loss": 0.44907671213150024,
"step": 1820
},
{
"epoch": 3.8520084566596196,
"grad_norm": 0.7877438068389893,
"learning_rate": 1.0346214405539783e-06,
"loss": 0.4558654725551605,
"step": 1822
},
{
"epoch": 3.8562367864693448,
"grad_norm": 9.517936706542969,
"learning_rate": 1.0327007412936419e-06,
"loss": 0.4603421688079834,
"step": 1824
},
{
"epoch": 3.8604651162790695,
"grad_norm": 14.035871505737305,
"learning_rate": 1.0308346565615158e-06,
"loss": 0.4513104259967804,
"step": 1826
},
{
"epoch": 3.864693446088795,
"grad_norm": 1.6873284578323364,
"learning_rate": 1.0290232091712337e-06,
"loss": 0.7685039639472961,
"step": 1828
},
{
"epoch": 3.86892177589852,
"grad_norm": 3.751678943634033,
"learning_rate": 1.0272664212684687e-06,
"loss": 0.7587875723838806,
"step": 1830
},
{
"epoch": 3.873150105708245,
"grad_norm": 3.94510817527771,
"learning_rate": 1.025564314330659e-06,
"loss": 0.48427051305770874,
"step": 1832
},
{
"epoch": 3.8773784355179703,
"grad_norm": 1.725934386253357,
"learning_rate": 1.0239169091667447e-06,
"loss": 0.38323408365249634,
"step": 1834
},
{
"epoch": 3.8816067653276956,
"grad_norm": 0.9607004523277283,
"learning_rate": 1.022324225916917e-06,
"loss": 0.44224637746810913,
"step": 1836
},
{
"epoch": 3.8858350951374208,
"grad_norm": 0.3913181722164154,
"learning_rate": 1.0207862840523672e-06,
"loss": 0.4007515609264374,
"step": 1838
},
{
"epoch": 3.890063424947146,
"grad_norm": 2.630666494369507,
"learning_rate": 1.0193031023750534e-06,
"loss": 0.648723840713501,
"step": 1840
},
{
"epoch": 3.894291754756871,
"grad_norm": 2.862149953842163,
"learning_rate": 1.017874699017468e-06,
"loss": 0.6909283995628357,
"step": 1842
},
{
"epoch": 3.898520084566596,
"grad_norm": 2.604625940322876,
"learning_rate": 1.0165010914424142e-06,
"loss": 0.45135021209716797,
"step": 1844
},
{
"epoch": 3.9027484143763216,
"grad_norm": 30.206815719604492,
"learning_rate": 1.015182296442798e-06,
"loss": 0.5103435516357422,
"step": 1846
},
{
"epoch": 3.9069767441860463,
"grad_norm": 11.992305755615234,
"learning_rate": 1.0139183301414176e-06,
"loss": 0.1478520631790161,
"step": 1848
},
{
"epoch": 3.9112050739957716,
"grad_norm": 1.9497147798538208,
"learning_rate": 1.0127092079907682e-06,
"loss": 0.6171524524688721,
"step": 1850
},
{
"epoch": 3.9154334038054968,
"grad_norm": 1.3044767379760742,
"learning_rate": 1.0115549447728551e-06,
"loss": 0.41875213384628296,
"step": 1852
},
{
"epoch": 3.919661733615222,
"grad_norm": 12.343912124633789,
"learning_rate": 1.0104555545990088e-06,
"loss": 0.615163266658783,
"step": 1854
},
{
"epoch": 3.923890063424947,
"grad_norm": 2.4850447177886963,
"learning_rate": 1.009411050909716e-06,
"loss": 0.38548392057418823,
"step": 1856
},
{
"epoch": 3.9281183932346724,
"grad_norm": 8.50257396697998,
"learning_rate": 1.008421446474455e-06,
"loss": 0.6773999929428101,
"step": 1858
},
{
"epoch": 3.9323467230443976,
"grad_norm": 15.77463150024414,
"learning_rate": 1.0074867533915367e-06,
"loss": 0.39413416385650635,
"step": 1860
},
{
"epoch": 3.9365750528541223,
"grad_norm": 12.434320449829102,
"learning_rate": 1.0066069830879593e-06,
"loss": 0.8713715076446533,
"step": 1862
},
{
"epoch": 3.940803382663848,
"grad_norm": 2.4119017124176025,
"learning_rate": 1.0057821463192687e-06,
"loss": 0.7903198003768921,
"step": 1864
},
{
"epoch": 3.9450317124735728,
"grad_norm": 2.082385301589966,
"learning_rate": 1.005012253169426e-06,
"loss": 0.6907875537872314,
"step": 1866
},
{
"epoch": 3.949260042283298,
"grad_norm": 0.17357464134693146,
"learning_rate": 1.0042973130506839e-06,
"loss": 0.007752852980047464,
"step": 1868
},
{
"epoch": 3.953488372093023,
"grad_norm": 2.7287681102752686,
"learning_rate": 1.003637334703472e-06,
"loss": 0.6832239627838135,
"step": 1870
},
{
"epoch": 3.9577167019027484,
"grad_norm": 2.661362886428833,
"learning_rate": 1.0030323261962922e-06,
"loss": 0.4995976686477661,
"step": 1872
},
{
"epoch": 3.9619450317124736,
"grad_norm": 4.728374481201172,
"learning_rate": 1.002482294925615e-06,
"loss": 0.45964568853378296,
"step": 1874
},
{
"epoch": 3.966173361522199,
"grad_norm": 2.4203152656555176,
"learning_rate": 1.0019872476157952e-06,
"loss": 0.5990220904350281,
"step": 1876
},
{
"epoch": 3.970401691331924,
"grad_norm": 11.37574291229248,
"learning_rate": 1.0015471903189834e-06,
"loss": 0.2836225926876068,
"step": 1878
},
{
"epoch": 3.974630021141649,
"grad_norm": 2.8287785053253174,
"learning_rate": 1.0011621284150583e-06,
"loss": 0.4260450005531311,
"step": 1880
},
{
"epoch": 3.9788583509513744,
"grad_norm": 3.307931661605835,
"learning_rate": 1.000832066611556e-06,
"loss": 0.7673491835594177,
"step": 1882
},
{
"epoch": 3.983086680761099,
"grad_norm": 2.6805083751678467,
"learning_rate": 1.0005570089436137e-06,
"loss": 0.8376491665840149,
"step": 1884
},
{
"epoch": 3.9873150105708244,
"grad_norm": 2.169524669647217,
"learning_rate": 1.0003369587739234e-06,
"loss": 0.6630700826644897,
"step": 1886
},
{
"epoch": 3.9915433403805496,
"grad_norm": 1.7867133617401123,
"learning_rate": 1.0001719187926854e-06,
"loss": 0.6671386957168579,
"step": 1888
},
{
"epoch": 3.995771670190275,
"grad_norm": 2.7854933738708496,
"learning_rate": 1.0000618910175811e-06,
"loss": 0.6984207630157471,
"step": 1890
},
{
"epoch": 4.0,
"grad_norm": 3.298140287399292,
"learning_rate": 1.000006876793743e-06,
"loss": 0.31056082248687744,
"step": 1892
},
{
"epoch": 4.0,
"step": 1892,
"total_flos": 3.5845266746764165e+18,
"train_loss": 0.8317920884405938,
"train_runtime": 4572.6706,
"train_samples_per_second": 12.413,
"train_steps_per_second": 0.414
}
],
"logging_steps": 2,
"max_steps": 1892,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.5845266746764165e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}