9b-103 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
073da04 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1892,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004228329809725159,
"grad_norm": 2.896852493286133,
"learning_rate": 1.0526315789473685e-07,
"loss": 1.8244051933288574,
"step": 2
},
{
"epoch": 0.008456659619450317,
"grad_norm": 1.3089975118637085,
"learning_rate": 3.1578947368421055e-07,
"loss": 2.2067313194274902,
"step": 4
},
{
"epoch": 0.012684989429175475,
"grad_norm": 8.145967483520508,
"learning_rate": 5.263157894736843e-07,
"loss": 2.205047130584717,
"step": 6
},
{
"epoch": 0.016913319238900635,
"grad_norm": 1.1983146667480469,
"learning_rate": 7.368421052631579e-07,
"loss": 1.9966031312942505,
"step": 8
},
{
"epoch": 0.021141649048625793,
"grad_norm": 5.088335990905762,
"learning_rate": 9.473684210526317e-07,
"loss": 2.0129404067993164,
"step": 10
},
{
"epoch": 0.02536997885835095,
"grad_norm": 2.381748914718628,
"learning_rate": 1.1578947368421053e-06,
"loss": 1.6334909200668335,
"step": 12
},
{
"epoch": 0.02959830866807611,
"grad_norm": 1.500184416770935,
"learning_rate": 1.3684210526315791e-06,
"loss": 1.6388685703277588,
"step": 14
},
{
"epoch": 0.03382663847780127,
"grad_norm": 2.3994479179382324,
"learning_rate": 1.5789473684210526e-06,
"loss": 1.758866786956787,
"step": 16
},
{
"epoch": 0.03805496828752643,
"grad_norm": 1.419136881828308,
"learning_rate": 1.7894736842105265e-06,
"loss": 1.7872570753097534,
"step": 18
},
{
"epoch": 0.042283298097251586,
"grad_norm": 4.953869342803955,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.6357908248901367,
"step": 20
},
{
"epoch": 0.046511627906976744,
"grad_norm": 5.676716327667236,
"learning_rate": 2.2105263157894738e-06,
"loss": 1.4728902578353882,
"step": 22
},
{
"epoch": 0.0507399577167019,
"grad_norm": 1.9945367574691772,
"learning_rate": 2.4210526315789477e-06,
"loss": 1.7624297142028809,
"step": 24
},
{
"epoch": 0.05496828752642706,
"grad_norm": 2.850555419921875,
"learning_rate": 2.631578947368421e-06,
"loss": 1.3475571870803833,
"step": 26
},
{
"epoch": 0.05919661733615222,
"grad_norm": 0.9735235571861267,
"learning_rate": 2.842105263157895e-06,
"loss": 1.6519263982772827,
"step": 28
},
{
"epoch": 0.06342494714587738,
"grad_norm": 1.3306273221969604,
"learning_rate": 3.052631578947369e-06,
"loss": 1.2123682498931885,
"step": 30
},
{
"epoch": 0.06765327695560254,
"grad_norm": 1.5061205625534058,
"learning_rate": 3.2631578947368423e-06,
"loss": 0.9824841022491455,
"step": 32
},
{
"epoch": 0.07188160676532769,
"grad_norm": 3.700876235961914,
"learning_rate": 3.473684210526316e-06,
"loss": 1.1644264459609985,
"step": 34
},
{
"epoch": 0.07610993657505286,
"grad_norm": 2.1790671348571777,
"learning_rate": 3.6842105263157896e-06,
"loss": 1.3551349639892578,
"step": 36
},
{
"epoch": 0.080338266384778,
"grad_norm": 1.5091564655303955,
"learning_rate": 3.894736842105263e-06,
"loss": 1.2848972082138062,
"step": 38
},
{
"epoch": 0.08456659619450317,
"grad_norm": 2.5904228687286377,
"learning_rate": 4.105263157894737e-06,
"loss": 1.2055286169052124,
"step": 40
},
{
"epoch": 0.08879492600422834,
"grad_norm": 1.6730945110321045,
"learning_rate": 4.315789473684211e-06,
"loss": 1.6836695671081543,
"step": 42
},
{
"epoch": 0.09302325581395349,
"grad_norm": 7.25777530670166,
"learning_rate": 4.526315789473685e-06,
"loss": 1.5885814428329468,
"step": 44
},
{
"epoch": 0.09725158562367865,
"grad_norm": 1.6787711381912231,
"learning_rate": 4.736842105263158e-06,
"loss": 1.2579834461212158,
"step": 46
},
{
"epoch": 0.1014799154334038,
"grad_norm": 0.7427865862846375,
"learning_rate": 4.947368421052632e-06,
"loss": 1.5057624578475952,
"step": 48
},
{
"epoch": 0.10570824524312897,
"grad_norm": 1.0933090448379517,
"learning_rate": 5.157894736842106e-06,
"loss": 1.475600242614746,
"step": 50
},
{
"epoch": 0.10993657505285412,
"grad_norm": 2.284050703048706,
"learning_rate": 5.36842105263158e-06,
"loss": 0.694992184638977,
"step": 52
},
{
"epoch": 0.11416490486257928,
"grad_norm": 1.065284013748169,
"learning_rate": 5.578947368421052e-06,
"loss": 1.0386056900024414,
"step": 54
},
{
"epoch": 0.11839323467230443,
"grad_norm": 1.5792242288589478,
"learning_rate": 5.789473684210527e-06,
"loss": 1.3667396306991577,
"step": 56
},
{
"epoch": 0.1226215644820296,
"grad_norm": 1.5710326433181763,
"learning_rate": 6e-06,
"loss": 1.3940433263778687,
"step": 58
},
{
"epoch": 0.12684989429175475,
"grad_norm": 1.837470531463623,
"learning_rate": 6.2105263157894745e-06,
"loss": 1.4068245887756348,
"step": 60
},
{
"epoch": 0.13107822410147993,
"grad_norm": 2.3961586952209473,
"learning_rate": 6.421052631578948e-06,
"loss": 0.7671315670013428,
"step": 62
},
{
"epoch": 0.13530655391120508,
"grad_norm": 1.1925511360168457,
"learning_rate": 6.631578947368421e-06,
"loss": 1.03555166721344,
"step": 64
},
{
"epoch": 0.13953488372093023,
"grad_norm": 1.078287959098816,
"learning_rate": 6.842105263157896e-06,
"loss": 1.3419888019561768,
"step": 66
},
{
"epoch": 0.14376321353065538,
"grad_norm": 1.193540334701538,
"learning_rate": 7.052631578947369e-06,
"loss": 1.3759318590164185,
"step": 68
},
{
"epoch": 0.14799154334038056,
"grad_norm": 6.178553581237793,
"learning_rate": 7.263157894736843e-06,
"loss": 1.3569273948669434,
"step": 70
},
{
"epoch": 0.1522198731501057,
"grad_norm": 0.9481953978538513,
"learning_rate": 7.473684210526316e-06,
"loss": 1.3459744453430176,
"step": 72
},
{
"epoch": 0.15644820295983086,
"grad_norm": 1.2028135061264038,
"learning_rate": 7.68421052631579e-06,
"loss": 0.9366351366043091,
"step": 74
},
{
"epoch": 0.160676532769556,
"grad_norm": 1.5675595998764038,
"learning_rate": 7.894736842105265e-06,
"loss": 0.8408719301223755,
"step": 76
},
{
"epoch": 0.1649048625792812,
"grad_norm": 1.4534165859222412,
"learning_rate": 8.105263157894736e-06,
"loss": 1.2801278829574585,
"step": 78
},
{
"epoch": 0.16913319238900634,
"grad_norm": 2.50400710105896,
"learning_rate": 8.315789473684212e-06,
"loss": 1.0849196910858154,
"step": 80
},
{
"epoch": 0.1733615221987315,
"grad_norm": 1.2264550924301147,
"learning_rate": 8.526315789473685e-06,
"loss": 1.0859289169311523,
"step": 82
},
{
"epoch": 0.17758985200845667,
"grad_norm": 0.7737696766853333,
"learning_rate": 8.736842105263158e-06,
"loss": 0.8703932762145996,
"step": 84
},
{
"epoch": 0.18181818181818182,
"grad_norm": 2.462402105331421,
"learning_rate": 8.947368421052632e-06,
"loss": 1.7095427513122559,
"step": 86
},
{
"epoch": 0.18604651162790697,
"grad_norm": 1.2406187057495117,
"learning_rate": 9.157894736842105e-06,
"loss": 1.3062974214553833,
"step": 88
},
{
"epoch": 0.19027484143763213,
"grad_norm": 1.6565141677856445,
"learning_rate": 9.36842105263158e-06,
"loss": 0.9879953861236572,
"step": 90
},
{
"epoch": 0.1945031712473573,
"grad_norm": 1.7372864484786987,
"learning_rate": 9.578947368421054e-06,
"loss": 1.3177132606506348,
"step": 92
},
{
"epoch": 0.19873150105708245,
"grad_norm": 0.9786802530288696,
"learning_rate": 9.789473684210527e-06,
"loss": 0.9800747632980347,
"step": 94
},
{
"epoch": 0.2029598308668076,
"grad_norm": 1.5775426626205444,
"learning_rate": 1e-05,
"loss": 1.2979300022125244,
"step": 96
},
{
"epoch": 0.20718816067653276,
"grad_norm": 2.6896512508392334,
"learning_rate": 9.999972492846047e-06,
"loss": 1.384432315826416,
"step": 98
},
{
"epoch": 0.21141649048625794,
"grad_norm": 1.2727605104446411,
"learning_rate": 9.999889971720468e-06,
"loss": 1.3217734098434448,
"step": 100
},
{
"epoch": 0.2156448202959831,
"grad_norm": 0.9020416736602783,
"learning_rate": 9.999752437632121e-06,
"loss": 1.2460750341415405,
"step": 102
},
{
"epoch": 0.21987315010570824,
"grad_norm": 1.2459571361541748,
"learning_rate": 9.999559892262412e-06,
"loss": 1.3091208934783936,
"step": 104
},
{
"epoch": 0.22410147991543342,
"grad_norm": 0.9687027335166931,
"learning_rate": 9.999312337965291e-06,
"loss": 1.297829508781433,
"step": 106
},
{
"epoch": 0.22832980972515857,
"grad_norm": 2.3220086097717285,
"learning_rate": 9.999009777767202e-06,
"loss": 1.0220335721969604,
"step": 108
},
{
"epoch": 0.23255813953488372,
"grad_norm": 1.1271264553070068,
"learning_rate": 9.99865221536707e-06,
"loss": 1.286104440689087,
"step": 110
},
{
"epoch": 0.23678646934460887,
"grad_norm": 1.8850802183151245,
"learning_rate": 9.998239655136237e-06,
"loss": 1.3547574281692505,
"step": 112
},
{
"epoch": 0.24101479915433405,
"grad_norm": 0.994990348815918,
"learning_rate": 9.997772102118417e-06,
"loss": 1.6156290769577026,
"step": 114
},
{
"epoch": 0.2452431289640592,
"grad_norm": 1.4753178358078003,
"learning_rate": 9.997249562029637e-06,
"loss": 1.3242958784103394,
"step": 116
},
{
"epoch": 0.24947145877378435,
"grad_norm": 4.679366588592529,
"learning_rate": 9.996672041258154e-06,
"loss": 1.3481351137161255,
"step": 118
},
{
"epoch": 0.2536997885835095,
"grad_norm": 1.585523247718811,
"learning_rate": 9.996039546864395e-06,
"loss": 0.8401036858558655,
"step": 120
},
{
"epoch": 0.25792811839323465,
"grad_norm": 1.0143969058990479,
"learning_rate": 9.995352086580858e-06,
"loss": 1.2970356941223145,
"step": 122
},
{
"epoch": 0.26215644820295986,
"grad_norm": 0.7017656564712524,
"learning_rate": 9.994609668812022e-06,
"loss": 1.1422041654586792,
"step": 124
},
{
"epoch": 0.266384778012685,
"grad_norm": 1.5835518836975098,
"learning_rate": 9.993812302634237e-06,
"loss": 1.345003604888916,
"step": 126
},
{
"epoch": 0.27061310782241016,
"grad_norm": 1.2006107568740845,
"learning_rate": 9.99295999779563e-06,
"loss": 0.9308596849441528,
"step": 128
},
{
"epoch": 0.2748414376321353,
"grad_norm": 1.097043514251709,
"learning_rate": 9.99205276471597e-06,
"loss": 0.9848383665084839,
"step": 130
},
{
"epoch": 0.27906976744186046,
"grad_norm": 2.325146436691284,
"learning_rate": 9.991090614486546e-06,
"loss": 1.2695246934890747,
"step": 132
},
{
"epoch": 0.2832980972515856,
"grad_norm": 1.3147951364517212,
"learning_rate": 9.990073558870028e-06,
"loss": 0.8363595604896545,
"step": 134
},
{
"epoch": 0.28752642706131076,
"grad_norm": 2.1097497940063477,
"learning_rate": 9.989001610300333e-06,
"loss": 1.1261337995529175,
"step": 136
},
{
"epoch": 0.2917547568710359,
"grad_norm": 2.1868903636932373,
"learning_rate": 9.987874781882461e-06,
"loss": 1.3797463178634644,
"step": 138
},
{
"epoch": 0.2959830866807611,
"grad_norm": 0.9027886390686035,
"learning_rate": 9.986693087392341e-06,
"loss": 0.9262903332710266,
"step": 140
},
{
"epoch": 0.30021141649048627,
"grad_norm": 0.959275484085083,
"learning_rate": 9.985456541276667e-06,
"loss": 1.563652753829956,
"step": 142
},
{
"epoch": 0.3044397463002114,
"grad_norm": 3.225210666656494,
"learning_rate": 9.984165158652708e-06,
"loss": 1.0929958820343018,
"step": 144
},
{
"epoch": 0.3086680761099366,
"grad_norm": 3.16994309425354,
"learning_rate": 9.982818955308135e-06,
"loss": 0.6909764409065247,
"step": 146
},
{
"epoch": 0.3128964059196617,
"grad_norm": 1.1078425645828247,
"learning_rate": 9.98141794770083e-06,
"loss": 1.2502849102020264,
"step": 148
},
{
"epoch": 0.3171247357293869,
"grad_norm": 1.2160269021987915,
"learning_rate": 9.979962152958669e-06,
"loss": 1.269697666168213,
"step": 150
},
{
"epoch": 0.321353065539112,
"grad_norm": 0.7900623083114624,
"learning_rate": 9.97845158887933e-06,
"loss": 1.1772613525390625,
"step": 152
},
{
"epoch": 0.32558139534883723,
"grad_norm": 2.6791653633117676,
"learning_rate": 9.976886273930063e-06,
"loss": 0.7300588488578796,
"step": 154
},
{
"epoch": 0.3298097251585624,
"grad_norm": 1.1209478378295898,
"learning_rate": 9.975266227247474e-06,
"loss": 1.248189091682434,
"step": 156
},
{
"epoch": 0.33403805496828753,
"grad_norm": 3.0206446647644043,
"learning_rate": 9.973591468637286e-06,
"loss": 1.2687840461730957,
"step": 158
},
{
"epoch": 0.3382663847780127,
"grad_norm": 1.1757704019546509,
"learning_rate": 9.971862018574094e-06,
"loss": 1.2619996070861816,
"step": 160
},
{
"epoch": 0.34249471458773784,
"grad_norm": 1.0647850036621094,
"learning_rate": 9.97007789820112e-06,
"loss": 1.0929667949676514,
"step": 162
},
{
"epoch": 0.346723044397463,
"grad_norm": 0.7367534041404724,
"learning_rate": 9.968239129329953e-06,
"loss": 0.9517663717269897,
"step": 164
},
{
"epoch": 0.35095137420718814,
"grad_norm": 2.4899051189422607,
"learning_rate": 9.96634573444028e-06,
"loss": 0.877385675907135,
"step": 166
},
{
"epoch": 0.35517970401691334,
"grad_norm": 0.8361695408821106,
"learning_rate": 9.964397736679615e-06,
"loss": 1.1926270723342896,
"step": 168
},
{
"epoch": 0.3594080338266385,
"grad_norm": 0.8268985748291016,
"learning_rate": 9.962395159863014e-06,
"loss": 1.206201195716858,
"step": 170
},
{
"epoch": 0.36363636363636365,
"grad_norm": 0.860716700553894,
"learning_rate": 9.960338028472781e-06,
"loss": 1.0416215658187866,
"step": 172
},
{
"epoch": 0.3678646934460888,
"grad_norm": 1.717405915260315,
"learning_rate": 9.958226367658176e-06,
"loss": 1.354285478591919,
"step": 174
},
{
"epoch": 0.37209302325581395,
"grad_norm": 1.2230415344238281,
"learning_rate": 9.9560602032351e-06,
"loss": 1.0286847352981567,
"step": 176
},
{
"epoch": 0.3763213530655391,
"grad_norm": 5.4142889976501465,
"learning_rate": 9.953839561685783e-06,
"loss": 0.8806125521659851,
"step": 178
},
{
"epoch": 0.38054968287526425,
"grad_norm": 0.6565871238708496,
"learning_rate": 9.95156447015846e-06,
"loss": 1.2810680866241455,
"step": 180
},
{
"epoch": 0.38477801268498946,
"grad_norm": 0.767181396484375,
"learning_rate": 9.94923495646704e-06,
"loss": 1.5317225456237793,
"step": 182
},
{
"epoch": 0.3890063424947146,
"grad_norm": 1.2074666023254395,
"learning_rate": 9.946851049090764e-06,
"loss": 1.3414469957351685,
"step": 184
},
{
"epoch": 0.39323467230443976,
"grad_norm": 2.182628870010376,
"learning_rate": 9.944412777173854e-06,
"loss": 1.1146631240844727,
"step": 186
},
{
"epoch": 0.3974630021141649,
"grad_norm": 0.8745899200439453,
"learning_rate": 9.941920170525168e-06,
"loss": 0.9901984333992004,
"step": 188
},
{
"epoch": 0.40169133192389006,
"grad_norm": 1.3491839170455933,
"learning_rate": 9.939373259617822e-06,
"loss": 1.0261520147323608,
"step": 190
},
{
"epoch": 0.4059196617336152,
"grad_norm": 0.9883779883384705,
"learning_rate": 9.936772075588825e-06,
"loss": 1.294724464416504,
"step": 192
},
{
"epoch": 0.41014799154334036,
"grad_norm": 3.7000715732574463,
"learning_rate": 9.934116650238697e-06,
"loss": 1.0057058334350586,
"step": 194
},
{
"epoch": 0.4143763213530655,
"grad_norm": 2.107089042663574,
"learning_rate": 9.931407016031078e-06,
"loss": 1.277288794517517,
"step": 196
},
{
"epoch": 0.4186046511627907,
"grad_norm": 1.0502349138259888,
"learning_rate": 9.928643206092338e-06,
"loss": 1.0722697973251343,
"step": 198
},
{
"epoch": 0.42283298097251587,
"grad_norm": 1.2788581848144531,
"learning_rate": 9.925825254211161e-06,
"loss": 1.1851972341537476,
"step": 200
},
{
"epoch": 0.427061310782241,
"grad_norm": 3.926483631134033,
"learning_rate": 9.922953194838143e-06,
"loss": 0.5971790552139282,
"step": 202
},
{
"epoch": 0.4312896405919662,
"grad_norm": 0.9535334706306458,
"learning_rate": 9.920027063085363e-06,
"loss": 1.2318687438964844,
"step": 204
},
{
"epoch": 0.4355179704016913,
"grad_norm": 1.716219425201416,
"learning_rate": 9.917046894725957e-06,
"loss": 1.2947533130645752,
"step": 206
},
{
"epoch": 0.4397463002114165,
"grad_norm": 2.18400502204895,
"learning_rate": 9.91401272619368e-06,
"loss": 0.7953985333442688,
"step": 208
},
{
"epoch": 0.4439746300211416,
"grad_norm": 1.0759165287017822,
"learning_rate": 9.910924594582463e-06,
"loss": 1.2876534461975098,
"step": 210
},
{
"epoch": 0.44820295983086683,
"grad_norm": 1.947562336921692,
"learning_rate": 9.907782537645955e-06,
"loss": 1.2165019512176514,
"step": 212
},
{
"epoch": 0.452431289640592,
"grad_norm": 1.8069974184036255,
"learning_rate": 9.904586593797062e-06,
"loss": 1.4206544160842896,
"step": 214
},
{
"epoch": 0.45665961945031713,
"grad_norm": 19.4478759765625,
"learning_rate": 9.901336802107485e-06,
"loss": 0.6380563378334045,
"step": 216
},
{
"epoch": 0.4608879492600423,
"grad_norm": 3.2247512340545654,
"learning_rate": 9.89803320230723e-06,
"loss": 1.0073024034500122,
"step": 218
},
{
"epoch": 0.46511627906976744,
"grad_norm": 1.3589563369750977,
"learning_rate": 9.894675834784136e-06,
"loss": 1.2687593698501587,
"step": 220
},
{
"epoch": 0.4693446088794926,
"grad_norm": 0.9938497543334961,
"learning_rate": 9.891264740583367e-06,
"loss": 1.2345340251922607,
"step": 222
},
{
"epoch": 0.47357293868921774,
"grad_norm": 1.7045886516571045,
"learning_rate": 9.88779996140692e-06,
"loss": 1.3495219945907593,
"step": 224
},
{
"epoch": 0.47780126849894294,
"grad_norm": 1.0711443424224854,
"learning_rate": 9.884281539613109e-06,
"loss": 1.246127963066101,
"step": 226
},
{
"epoch": 0.4820295983086681,
"grad_norm": 1.1415208578109741,
"learning_rate": 9.880709518216061e-06,
"loss": 1.2400927543640137,
"step": 228
},
{
"epoch": 0.48625792811839325,
"grad_norm": 0.7870588302612305,
"learning_rate": 9.87708394088517e-06,
"loss": 0.8983721137046814,
"step": 230
},
{
"epoch": 0.4904862579281184,
"grad_norm": 0.8683267831802368,
"learning_rate": 9.873404851944573e-06,
"loss": 1.2452239990234375,
"step": 232
},
{
"epoch": 0.49471458773784355,
"grad_norm": 5.117804527282715,
"learning_rate": 9.869672296372611e-06,
"loss": 0.9248649477958679,
"step": 234
},
{
"epoch": 0.4989429175475687,
"grad_norm": 2.0462708473205566,
"learning_rate": 9.86588631980128e-06,
"loss": 0.876616358757019,
"step": 236
},
{
"epoch": 0.5031712473572939,
"grad_norm": 20.712501525878906,
"learning_rate": 9.86204696851566e-06,
"loss": 0.8371405601501465,
"step": 238
},
{
"epoch": 0.507399577167019,
"grad_norm": 1.034693717956543,
"learning_rate": 9.858154289453363e-06,
"loss": 0.9195130467414856,
"step": 240
},
{
"epoch": 0.5116279069767442,
"grad_norm": 2.491039991378784,
"learning_rate": 9.854208330203958e-06,
"loss": 1.3159518241882324,
"step": 242
},
{
"epoch": 0.5158562367864693,
"grad_norm": 2.147132396697998,
"learning_rate": 9.850209139008377e-06,
"loss": 1.3358653783798218,
"step": 244
},
{
"epoch": 0.5200845665961945,
"grad_norm": 3.794759511947632,
"learning_rate": 9.846156764758344e-06,
"loss": 1.2811285257339478,
"step": 246
},
{
"epoch": 0.5243128964059197,
"grad_norm": 1.548609733581543,
"learning_rate": 9.842051256995759e-06,
"loss": 1.2365938425064087,
"step": 248
},
{
"epoch": 0.5285412262156448,
"grad_norm": 2.255784511566162,
"learning_rate": 9.837892665912105e-06,
"loss": 1.1832764148712158,
"step": 250
},
{
"epoch": 0.53276955602537,
"grad_norm": 4.790487766265869,
"learning_rate": 9.833681042347827e-06,
"loss": 1.1316468715667725,
"step": 252
},
{
"epoch": 0.5369978858350951,
"grad_norm": 2.892882823944092,
"learning_rate": 9.829416437791721e-06,
"loss": 1.7385282516479492,
"step": 254
},
{
"epoch": 0.5412262156448203,
"grad_norm": 1.2317280769348145,
"learning_rate": 9.825098904380284e-06,
"loss": 1.2122973203659058,
"step": 256
},
{
"epoch": 0.5454545454545454,
"grad_norm": 1.31428062915802,
"learning_rate": 9.820728494897101e-06,
"loss": 1.3897485733032227,
"step": 258
},
{
"epoch": 0.5496828752642706,
"grad_norm": 1.081001877784729,
"learning_rate": 9.816305262772182e-06,
"loss": 0.9935633540153503,
"step": 260
},
{
"epoch": 0.5539112050739958,
"grad_norm": 1.1620075702667236,
"learning_rate": 9.811829262081315e-06,
"loss": 1.194795846939087,
"step": 262
},
{
"epoch": 0.5581395348837209,
"grad_norm": 2.897712469100952,
"learning_rate": 9.807300547545411e-06,
"loss": 1.0515440702438354,
"step": 264
},
{
"epoch": 0.5623678646934461,
"grad_norm": 1.5808881521224976,
"learning_rate": 9.802719174529821e-06,
"loss": 1.4962360858917236,
"step": 266
},
{
"epoch": 0.5665961945031712,
"grad_norm": 2.5658390522003174,
"learning_rate": 9.798085199043672e-06,
"loss": 0.5049180388450623,
"step": 268
},
{
"epoch": 0.5708245243128964,
"grad_norm": 3.8659236431121826,
"learning_rate": 9.793398677739177e-06,
"loss": 1.0005030632019043,
"step": 270
},
{
"epoch": 0.5750528541226215,
"grad_norm": 2.149238109588623,
"learning_rate": 9.788659667910942e-06,
"loss": 1.1760281324386597,
"step": 272
},
{
"epoch": 0.5792811839323467,
"grad_norm": 0.8895676732063293,
"learning_rate": 9.783868227495262e-06,
"loss": 0.9585500359535217,
"step": 274
},
{
"epoch": 0.5835095137420718,
"grad_norm": 3.1487722396850586,
"learning_rate": 9.779024415069427e-06,
"loss": 0.947961151599884,
"step": 276
},
{
"epoch": 0.587737843551797,
"grad_norm": 0.6882687211036682,
"learning_rate": 9.774128289850987e-06,
"loss": 0.8679822087287903,
"step": 278
},
{
"epoch": 0.5919661733615222,
"grad_norm": 1.4155272245407104,
"learning_rate": 9.769179911697037e-06,
"loss": 1.2337456941604614,
"step": 280
},
{
"epoch": 0.5961945031712473,
"grad_norm": 1.298057198524475,
"learning_rate": 9.764179341103493e-06,
"loss": 1.4632441997528076,
"step": 282
},
{
"epoch": 0.6004228329809725,
"grad_norm": 0.3873730003833771,
"learning_rate": 9.759126639204337e-06,
"loss": 0.9341024160385132,
"step": 284
},
{
"epoch": 0.6046511627906976,
"grad_norm": 0.8940773606300354,
"learning_rate": 9.75402186777088e-06,
"loss": 1.3100539445877075,
"step": 286
},
{
"epoch": 0.6088794926004228,
"grad_norm": 2.8771839141845703,
"learning_rate": 9.748865089211004e-06,
"loss": 0.42972153425216675,
"step": 288
},
{
"epoch": 0.6131078224101479,
"grad_norm": 1.2507033348083496,
"learning_rate": 9.743656366568398e-06,
"loss": 0.9476989507675171,
"step": 290
},
{
"epoch": 0.6173361522198731,
"grad_norm": 1.004989504814148,
"learning_rate": 9.738395763521792e-06,
"loss": 0.8646316528320312,
"step": 292
},
{
"epoch": 0.6215644820295984,
"grad_norm": 6.049814701080322,
"learning_rate": 9.733083344384164e-06,
"loss": 0.7138922810554504,
"step": 294
},
{
"epoch": 0.6257928118393234,
"grad_norm": 2.484060049057007,
"learning_rate": 9.727719174101981e-06,
"loss": 0.7330291867256165,
"step": 296
},
{
"epoch": 0.6300211416490487,
"grad_norm": 1.278454303741455,
"learning_rate": 9.722303318254376e-06,
"loss": 1.276779294013977,
"step": 298
},
{
"epoch": 0.6342494714587738,
"grad_norm": 1.6446763277053833,
"learning_rate": 9.71683584305236e-06,
"loss": 1.240735411643982,
"step": 300
},
{
"epoch": 0.638477801268499,
"grad_norm": 1.8123787641525269,
"learning_rate": 9.71131681533802e-06,
"loss": 0.718811571598053,
"step": 302
},
{
"epoch": 0.642706131078224,
"grad_norm": 11.416818618774414,
"learning_rate": 9.705746302583682e-06,
"loss": 0.8497800827026367,
"step": 304
},
{
"epoch": 0.6469344608879493,
"grad_norm": 1.5144388675689697,
"learning_rate": 9.700124372891104e-06,
"loss": 0.9179715514183044,
"step": 306
},
{
"epoch": 0.6511627906976745,
"grad_norm": 1.039093017578125,
"learning_rate": 9.694451094990636e-06,
"loss": 1.232336163520813,
"step": 308
},
{
"epoch": 0.6553911205073996,
"grad_norm": 2.932518720626831,
"learning_rate": 9.68872653824038e-06,
"loss": 1.2035396099090576,
"step": 310
},
{
"epoch": 0.6596194503171248,
"grad_norm": 1.2854681015014648,
"learning_rate": 9.68295077262534e-06,
"loss": 1.2047369480133057,
"step": 312
},
{
"epoch": 0.6638477801268499,
"grad_norm": 1.1843533515930176,
"learning_rate": 9.677123868756573e-06,
"loss": 1.2297992706298828,
"step": 314
},
{
"epoch": 0.6680761099365751,
"grad_norm": 1.1218198537826538,
"learning_rate": 9.671245897870318e-06,
"loss": 1.068047285079956,
"step": 316
},
{
"epoch": 0.6723044397463002,
"grad_norm": 2.8922905921936035,
"learning_rate": 9.665316931827131e-06,
"loss": 0.7625845670700073,
"step": 318
},
{
"epoch": 0.6765327695560254,
"grad_norm": 2.6902432441711426,
"learning_rate": 9.659337043111006e-06,
"loss": 1.237903118133545,
"step": 320
},
{
"epoch": 0.6807610993657506,
"grad_norm": 1.871984839439392,
"learning_rate": 9.653306304828483e-06,
"loss": 1.3034955263137817,
"step": 322
},
{
"epoch": 0.6849894291754757,
"grad_norm": 1.214774250984192,
"learning_rate": 9.647224790707762e-06,
"loss": 1.0795611143112183,
"step": 324
},
{
"epoch": 0.6892177589852009,
"grad_norm": 1.0731302499771118,
"learning_rate": 9.641092575097795e-06,
"loss": 1.2623587846755981,
"step": 326
},
{
"epoch": 0.693446088794926,
"grad_norm": 4.90795373916626,
"learning_rate": 9.63490973296738e-06,
"loss": 1.1686569452285767,
"step": 328
},
{
"epoch": 0.6976744186046512,
"grad_norm": 2.681135416030884,
"learning_rate": 9.62867633990425e-06,
"loss": 0.8564498424530029,
"step": 330
},
{
"epoch": 0.7019027484143763,
"grad_norm": 1.212094783782959,
"learning_rate": 9.622392472114136e-06,
"loss": 1.25063157081604,
"step": 332
},
{
"epoch": 0.7061310782241015,
"grad_norm": 2.725762367248535,
"learning_rate": 9.61605820641985e-06,
"loss": 1.2101666927337646,
"step": 334
},
{
"epoch": 0.7103594080338267,
"grad_norm": 1.2129253149032593,
"learning_rate": 9.60967362026033e-06,
"loss": 1.2083511352539062,
"step": 336
},
{
"epoch": 0.7145877378435518,
"grad_norm": 1.555783748626709,
"learning_rate": 9.60323879168971e-06,
"loss": 1.3075604438781738,
"step": 338
},
{
"epoch": 0.718816067653277,
"grad_norm": 3.475314140319824,
"learning_rate": 9.596753799376355e-06,
"loss": 1.19049870967865,
"step": 340
},
{
"epoch": 0.7230443974630021,
"grad_norm": 2.0879061222076416,
"learning_rate": 9.5902187226019e-06,
"loss": 1.1154621839523315,
"step": 342
},
{
"epoch": 0.7272727272727273,
"grad_norm": 8.40845012664795,
"learning_rate": 9.583633641260285e-06,
"loss": 0.9010623097419739,
"step": 344
},
{
"epoch": 0.7315010570824524,
"grad_norm": 8.427756309509277,
"learning_rate": 9.576998635856774e-06,
"loss": 0.8225324153900146,
"step": 346
},
{
"epoch": 0.7357293868921776,
"grad_norm": 2.1473522186279297,
"learning_rate": 9.570313787506974e-06,
"loss": 0.9615020751953125,
"step": 348
},
{
"epoch": 0.7399577167019028,
"grad_norm": 6.062769889831543,
"learning_rate": 9.563579177935844e-06,
"loss": 1.3048603534698486,
"step": 350
},
{
"epoch": 0.7441860465116279,
"grad_norm": 2.07364559173584,
"learning_rate": 9.556794889476687e-06,
"loss": 1.159698724746704,
"step": 352
},
{
"epoch": 0.7484143763213531,
"grad_norm": 2.1313230991363525,
"learning_rate": 9.549961005070161e-06,
"loss": 0.7197738885879517,
"step": 354
},
{
"epoch": 0.7526427061310782,
"grad_norm": 2.408674716949463,
"learning_rate": 9.543077608263245e-06,
"loss": 1.0134952068328857,
"step": 356
},
{
"epoch": 0.7568710359408034,
"grad_norm": 1.957841396331787,
"learning_rate": 9.536144783208229e-06,
"loss": 1.2071951627731323,
"step": 358
},
{
"epoch": 0.7610993657505285,
"grad_norm": 8.179128646850586,
"learning_rate": 9.529162614661687e-06,
"loss": 0.7516147494316101,
"step": 360
},
{
"epoch": 0.7653276955602537,
"grad_norm": 2.55159592628479,
"learning_rate": 9.522131187983433e-06,
"loss": 1.0871793031692505,
"step": 362
},
{
"epoch": 0.7695560253699789,
"grad_norm": 2.5193471908569336,
"learning_rate": 9.515050589135485e-06,
"loss": 1.0482189655303955,
"step": 364
},
{
"epoch": 0.773784355179704,
"grad_norm": 1.7531400918960571,
"learning_rate": 9.507920904681006e-06,
"loss": 1.2599376440048218,
"step": 366
},
{
"epoch": 0.7780126849894292,
"grad_norm": 1.85068941116333,
"learning_rate": 9.500742221783257e-06,
"loss": 1.2061271667480469,
"step": 368
},
{
"epoch": 0.7822410147991543,
"grad_norm": 5.245235443115234,
"learning_rate": 9.493514628204516e-06,
"loss": 0.8552356362342834,
"step": 370
},
{
"epoch": 0.7864693446088795,
"grad_norm": 1.3197134733200073,
"learning_rate": 9.48623821230502e-06,
"loss": 1.1898118257522583,
"step": 372
},
{
"epoch": 0.7906976744186046,
"grad_norm": 1.175612211227417,
"learning_rate": 9.478913063041878e-06,
"loss": 1.2073456048965454,
"step": 374
},
{
"epoch": 0.7949260042283298,
"grad_norm": 2.138735294342041,
"learning_rate": 9.471539269967982e-06,
"loss": 1.290218710899353,
"step": 376
},
{
"epoch": 0.7991543340380549,
"grad_norm": 1.6098554134368896,
"learning_rate": 9.464116923230913e-06,
"loss": 0.9563854932785034,
"step": 378
},
{
"epoch": 0.8033826638477801,
"grad_norm": 2.0989625453948975,
"learning_rate": 9.456646113571844e-06,
"loss": 1.2941992282867432,
"step": 380
},
{
"epoch": 0.8076109936575053,
"grad_norm": 2.350449800491333,
"learning_rate": 9.449126932324424e-06,
"loss": 1.168163776397705,
"step": 382
},
{
"epoch": 0.8118393234672304,
"grad_norm": 1.0263327360153198,
"learning_rate": 9.441559471413665e-06,
"loss": 0.9715777635574341,
"step": 384
},
{
"epoch": 0.8160676532769556,
"grad_norm": 2.5212090015411377,
"learning_rate": 9.433943823354816e-06,
"loss": 0.9501182436943054,
"step": 386
},
{
"epoch": 0.8202959830866807,
"grad_norm": 1.4024837017059326,
"learning_rate": 9.426280081252232e-06,
"loss": 1.4798977375030518,
"step": 388
},
{
"epoch": 0.8245243128964059,
"grad_norm": 6.935363292694092,
"learning_rate": 9.418568338798245e-06,
"loss": 1.1398916244506836,
"step": 390
},
{
"epoch": 0.828752642706131,
"grad_norm": 2.5754780769348145,
"learning_rate": 9.410808690271999e-06,
"loss": 0.9194220900535583,
"step": 392
},
{
"epoch": 0.8329809725158562,
"grad_norm": 1.9557315111160278,
"learning_rate": 9.40300123053832e-06,
"loss": 1.2362662553787231,
"step": 394
},
{
"epoch": 0.8372093023255814,
"grad_norm": 0.47936832904815674,
"learning_rate": 9.395146055046535e-06,
"loss": 0.7501541972160339,
"step": 396
},
{
"epoch": 0.8414376321353065,
"grad_norm": 0.6807716488838196,
"learning_rate": 9.387243259829323e-06,
"loss": 0.6976569294929504,
"step": 398
},
{
"epoch": 0.8456659619450317,
"grad_norm": 6.227804660797119,
"learning_rate": 9.37929294150153e-06,
"loss": 1.1244752407073975,
"step": 400
},
{
"epoch": 0.8498942917547568,
"grad_norm": 1.474228024482727,
"learning_rate": 9.371295197258992e-06,
"loss": 0.9696781039237976,
"step": 402
},
{
"epoch": 0.854122621564482,
"grad_norm": 1.3988255262374878,
"learning_rate": 9.363250124877345e-06,
"loss": 0.8240630030632019,
"step": 404
},
{
"epoch": 0.8583509513742071,
"grad_norm": 2.276531457901001,
"learning_rate": 9.35515782271083e-06,
"loss": 0.9939208626747131,
"step": 406
},
{
"epoch": 0.8625792811839323,
"grad_norm": 1.0749969482421875,
"learning_rate": 9.347018389691093e-06,
"loss": 0.9903083443641663,
"step": 408
},
{
"epoch": 0.8668076109936576,
"grad_norm": 3.1645753383636475,
"learning_rate": 9.338831925325974e-06,
"loss": 0.8789092302322388,
"step": 410
},
{
"epoch": 0.8710359408033826,
"grad_norm": 2.273981809616089,
"learning_rate": 9.330598529698287e-06,
"loss": 1.1768401861190796,
"step": 412
},
{
"epoch": 0.8752642706131079,
"grad_norm": 0.9225611686706543,
"learning_rate": 9.322318303464603e-06,
"loss": 1.0258930921554565,
"step": 414
},
{
"epoch": 0.879492600422833,
"grad_norm": 5.045302391052246,
"learning_rate": 9.313991347854015e-06,
"loss": 1.1129915714263916,
"step": 416
},
{
"epoch": 0.8837209302325582,
"grad_norm": 7.1052565574646,
"learning_rate": 9.305617764666897e-06,
"loss": 1.2124370336532593,
"step": 418
},
{
"epoch": 0.8879492600422833,
"grad_norm": 1.4860345125198364,
"learning_rate": 9.297197656273672e-06,
"loss": 1.188872218132019,
"step": 420
},
{
"epoch": 0.8921775898520085,
"grad_norm": 3.857395648956299,
"learning_rate": 9.288731125613544e-06,
"loss": 0.9384881854057312,
"step": 422
},
{
"epoch": 0.8964059196617337,
"grad_norm": 1.1968402862548828,
"learning_rate": 9.280218276193255e-06,
"loss": 1.2261526584625244,
"step": 424
},
{
"epoch": 0.9006342494714588,
"grad_norm": 0.9379101991653442,
"learning_rate": 9.271659212085807e-06,
"loss": 1.0875624418258667,
"step": 426
},
{
"epoch": 0.904862579281184,
"grad_norm": 3.7680888175964355,
"learning_rate": 9.263054037929199e-06,
"loss": 0.8655368685722351,
"step": 428
},
{
"epoch": 0.9090909090909091,
"grad_norm": 4.634127616882324,
"learning_rate": 9.254402858925142e-06,
"loss": 1.1067538261413574,
"step": 430
},
{
"epoch": 0.9133192389006343,
"grad_norm": 1.4680707454681396,
"learning_rate": 9.245705780837774e-06,
"loss": 1.257076382637024,
"step": 432
},
{
"epoch": 0.9175475687103594,
"grad_norm": 1.9205492734909058,
"learning_rate": 9.236962909992368e-06,
"loss": 0.5450637340545654,
"step": 434
},
{
"epoch": 0.9217758985200846,
"grad_norm": 26.722627639770508,
"learning_rate": 9.228174353274037e-06,
"loss": 0.3882911205291748,
"step": 436
},
{
"epoch": 0.9260042283298098,
"grad_norm": 4.1831488609313965,
"learning_rate": 9.219340218126413e-06,
"loss": 0.9474373459815979,
"step": 438
},
{
"epoch": 0.9302325581395349,
"grad_norm": 0.9313554763793945,
"learning_rate": 9.210460612550348e-06,
"loss": 1.1293978691101074,
"step": 440
},
{
"epoch": 0.9344608879492601,
"grad_norm": 4.188915252685547,
"learning_rate": 9.20153564510259e-06,
"loss": 0.745289146900177,
"step": 442
},
{
"epoch": 0.9386892177589852,
"grad_norm": 1.1278609037399292,
"learning_rate": 9.19256542489445e-06,
"loss": 1.246189832687378,
"step": 444
},
{
"epoch": 0.9429175475687104,
"grad_norm": 1.4961402416229248,
"learning_rate": 9.183550061590476e-06,
"loss": 1.094343662261963,
"step": 446
},
{
"epoch": 0.9471458773784355,
"grad_norm": 4.029289722442627,
"learning_rate": 9.174489665407103e-06,
"loss": 1.089815616607666,
"step": 448
},
{
"epoch": 0.9513742071881607,
"grad_norm": 6.033188343048096,
"learning_rate": 9.165384347111317e-06,
"loss": 0.7955329418182373,
"step": 450
},
{
"epoch": 0.9556025369978859,
"grad_norm": 1.9339733123779297,
"learning_rate": 9.156234218019291e-06,
"loss": 0.9353688359260559,
"step": 452
},
{
"epoch": 0.959830866807611,
"grad_norm": 1.51326584815979,
"learning_rate": 9.147039389995029e-06,
"loss": 1.3155266046524048,
"step": 454
},
{
"epoch": 0.9640591966173362,
"grad_norm": 7.166233539581299,
"learning_rate": 9.137799975448998e-06,
"loss": 0.6612483263015747,
"step": 456
},
{
"epoch": 0.9682875264270613,
"grad_norm": 3.3331398963928223,
"learning_rate": 9.128516087336755e-06,
"loss": 0.8956295847892761,
"step": 458
},
{
"epoch": 0.9725158562367865,
"grad_norm": 2.485344171524048,
"learning_rate": 9.119187839157559e-06,
"loss": 1.227403163909912,
"step": 460
},
{
"epoch": 0.9767441860465116,
"grad_norm": 0.8651933670043945,
"learning_rate": 9.109815344952995e-06,
"loss": 1.2107855081558228,
"step": 462
},
{
"epoch": 0.9809725158562368,
"grad_norm": 1.878743290901184,
"learning_rate": 9.100398719305566e-06,
"loss": 0.7366509437561035,
"step": 464
},
{
"epoch": 0.985200845665962,
"grad_norm": 1.2255029678344727,
"learning_rate": 9.090938077337305e-06,
"loss": 1.1837794780731201,
"step": 466
},
{
"epoch": 0.9894291754756871,
"grad_norm": 2.9394116401672363,
"learning_rate": 9.081433534708363e-06,
"loss": 0.9721347689628601,
"step": 468
},
{
"epoch": 0.9936575052854123,
"grad_norm": 0.9846524596214294,
"learning_rate": 9.071885207615591e-06,
"loss": 1.2441462278366089,
"step": 470
},
{
"epoch": 0.9978858350951374,
"grad_norm": 1.5261141061782837,
"learning_rate": 9.062293212791123e-06,
"loss": 1.2467325925827026,
"step": 472
},
{
"epoch": 1.0021141649048626,
"grad_norm": 1.6327552795410156,
"learning_rate": 9.052657667500951e-06,
"loss": 0.9771024584770203,
"step": 474
},
{
"epoch": 1.0063424947145878,
"grad_norm": 0.9313510060310364,
"learning_rate": 9.042978689543487e-06,
"loss": 1.114615797996521,
"step": 476
},
{
"epoch": 1.0105708245243128,
"grad_norm": 3.062166213989258,
"learning_rate": 9.033256397248124e-06,
"loss": 0.4529619514942169,
"step": 478
},
{
"epoch": 1.014799154334038,
"grad_norm": 2.18306565284729,
"learning_rate": 9.02349090947379e-06,
"loss": 0.7353109121322632,
"step": 480
},
{
"epoch": 1.0190274841437632,
"grad_norm": 1.6017274856567383,
"learning_rate": 9.013682345607502e-06,
"loss": 0.6578851938247681,
"step": 482
},
{
"epoch": 1.0232558139534884,
"grad_norm": 0.9708321690559387,
"learning_rate": 9.003830825562887e-06,
"loss": 1.0006972551345825,
"step": 484
},
{
"epoch": 1.0274841437632136,
"grad_norm": 1.420487403869629,
"learning_rate": 8.99393646977874e-06,
"loss": 1.0055036544799805,
"step": 486
},
{
"epoch": 1.0317124735729386,
"grad_norm": 2.9387149810791016,
"learning_rate": 8.983999399217536e-06,
"loss": 0.8852903246879578,
"step": 488
},
{
"epoch": 1.0359408033826638,
"grad_norm": 2.165130376815796,
"learning_rate": 8.974019735363952e-06,
"loss": 0.8060842752456665,
"step": 490
},
{
"epoch": 1.040169133192389,
"grad_norm": 29.002117156982422,
"learning_rate": 8.963997600223387e-06,
"loss": 0.7252429127693176,
"step": 492
},
{
"epoch": 1.0443974630021142,
"grad_norm": 2.428952217102051,
"learning_rate": 8.953933116320474e-06,
"loss": 1.1315640211105347,
"step": 494
},
{
"epoch": 1.0486257928118394,
"grad_norm": 1.2200968265533447,
"learning_rate": 8.943826406697569e-06,
"loss": 1.0207574367523193,
"step": 496
},
{
"epoch": 1.0528541226215644,
"grad_norm": 0.8859376311302185,
"learning_rate": 8.93367759491326e-06,
"loss": 0.7872439026832581,
"step": 498
},
{
"epoch": 1.0570824524312896,
"grad_norm": 1.045777440071106,
"learning_rate": 8.923486805040845e-06,
"loss": 0.882882297039032,
"step": 500
},
{
"epoch": 1.0613107822410148,
"grad_norm": 0.9387926459312439,
"learning_rate": 8.913254161666828e-06,
"loss": 1.1162152290344238,
"step": 502
},
{
"epoch": 1.06553911205074,
"grad_norm": 2.4976844787597656,
"learning_rate": 8.902979789889382e-06,
"loss": 1.0760377645492554,
"step": 504
},
{
"epoch": 1.069767441860465,
"grad_norm": 7.320751190185547,
"learning_rate": 8.892663815316833e-06,
"loss": 0.8308604955673218,
"step": 506
},
{
"epoch": 1.0739957716701902,
"grad_norm": 1.1178933382034302,
"learning_rate": 8.882306364066114e-06,
"loss": 0.8055670857429504,
"step": 508
},
{
"epoch": 1.0782241014799154,
"grad_norm": 3.726597785949707,
"learning_rate": 8.871907562761226e-06,
"loss": 0.9297540783882141,
"step": 510
},
{
"epoch": 1.0824524312896406,
"grad_norm": 0.8374629616737366,
"learning_rate": 8.861467538531692e-06,
"loss": 0.9856585264205933,
"step": 512
},
{
"epoch": 1.0866807610993658,
"grad_norm": 0.7937526106834412,
"learning_rate": 8.850986419011006e-06,
"loss": 1.1037338972091675,
"step": 514
},
{
"epoch": 1.0909090909090908,
"grad_norm": 1.1389909982681274,
"learning_rate": 8.840464332335062e-06,
"loss": 0.8448182940483093,
"step": 516
},
{
"epoch": 1.095137420718816,
"grad_norm": 1.0539064407348633,
"learning_rate": 8.829901407140599e-06,
"loss": 0.8356651067733765,
"step": 518
},
{
"epoch": 1.0993657505285412,
"grad_norm": 1.0243035554885864,
"learning_rate": 8.81929777256362e-06,
"loss": 0.6140103340148926,
"step": 520
},
{
"epoch": 1.1035940803382664,
"grad_norm": 1.2849695682525635,
"learning_rate": 8.808653558237818e-06,
"loss": 1.3192360401153564,
"step": 522
},
{
"epoch": 1.1078224101479917,
"grad_norm": 2.690241813659668,
"learning_rate": 8.797968894292992e-06,
"loss": 0.7449397444725037,
"step": 524
},
{
"epoch": 1.1120507399577166,
"grad_norm": 1.264523983001709,
"learning_rate": 8.787243911353447e-06,
"loss": 1.3042547702789307,
"step": 526
},
{
"epoch": 1.1162790697674418,
"grad_norm": 1.0589176416397095,
"learning_rate": 8.776478740536409e-06,
"loss": 0.8097569346427917,
"step": 528
},
{
"epoch": 1.120507399577167,
"grad_norm": 4.575756072998047,
"learning_rate": 8.76567351345042e-06,
"loss": 0.4913235902786255,
"step": 530
},
{
"epoch": 1.1247357293868923,
"grad_norm": 0.9360055327415466,
"learning_rate": 8.754828362193719e-06,
"loss": 0.9835745692253113,
"step": 532
},
{
"epoch": 1.1289640591966172,
"grad_norm": 3.9901702404022217,
"learning_rate": 8.743943419352638e-06,
"loss": 0.7979632616043091,
"step": 534
},
{
"epoch": 1.1331923890063424,
"grad_norm": 0.905530571937561,
"learning_rate": 8.733018817999977e-06,
"loss": 0.9963710308074951,
"step": 536
},
{
"epoch": 1.1374207188160677,
"grad_norm": 4.259133815765381,
"learning_rate": 8.722054691693374e-06,
"loss": 0.6634504795074463,
"step": 538
},
{
"epoch": 1.1416490486257929,
"grad_norm": 3.2057507038116455,
"learning_rate": 8.711051174473682e-06,
"loss": 0.7573935389518738,
"step": 540
},
{
"epoch": 1.145877378435518,
"grad_norm": 0.8688735961914062,
"learning_rate": 8.700008400863316e-06,
"loss": 1.0432833433151245,
"step": 542
},
{
"epoch": 1.150105708245243,
"grad_norm": 5.89119815826416,
"learning_rate": 8.688926505864624e-06,
"loss": 0.8422532677650452,
"step": 544
},
{
"epoch": 1.1543340380549683,
"grad_norm": 0.7301739454269409,
"learning_rate": 8.67780562495822e-06,
"loss": 1.154378890991211,
"step": 546
},
{
"epoch": 1.1585623678646935,
"grad_norm": 2.2917237281799316,
"learning_rate": 8.666645894101346e-06,
"loss": 1.2317252159118652,
"step": 548
},
{
"epoch": 1.1627906976744187,
"grad_norm": 1.5925415754318237,
"learning_rate": 8.655447449726189e-06,
"loss": 0.8310880661010742,
"step": 550
},
{
"epoch": 1.1670190274841437,
"grad_norm": 1.0509593486785889,
"learning_rate": 8.644210428738237e-06,
"loss": 1.044269323348999,
"step": 552
},
{
"epoch": 1.1712473572938689,
"grad_norm": 1.313195824623108,
"learning_rate": 8.632934968514582e-06,
"loss": 1.142549753189087,
"step": 554
},
{
"epoch": 1.175475687103594,
"grad_norm": 1.7246289253234863,
"learning_rate": 8.621621206902256e-06,
"loss": 0.9859258532524109,
"step": 556
},
{
"epoch": 1.1797040169133193,
"grad_norm": 0.9654635190963745,
"learning_rate": 8.61026928221654e-06,
"loss": 1.0348000526428223,
"step": 558
},
{
"epoch": 1.1839323467230445,
"grad_norm": 1.2986783981323242,
"learning_rate": 8.598879333239274e-06,
"loss": 0.9646149277687073,
"step": 560
},
{
"epoch": 1.1881606765327695,
"grad_norm": 1.510294795036316,
"learning_rate": 8.587451499217161e-06,
"loss": 0.7072579860687256,
"step": 562
},
{
"epoch": 1.1923890063424947,
"grad_norm": 1.839513897895813,
"learning_rate": 8.575985919860065e-06,
"loss": 1.0070905685424805,
"step": 564
},
{
"epoch": 1.1966173361522199,
"grad_norm": 7.103578567504883,
"learning_rate": 8.564482735339296e-06,
"loss": 0.9337130188941956,
"step": 566
},
{
"epoch": 1.200845665961945,
"grad_norm": 1.3289200067520142,
"learning_rate": 8.552942086285909e-06,
"loss": 1.167741298675537,
"step": 568
},
{
"epoch": 1.20507399577167,
"grad_norm": 3.186591148376465,
"learning_rate": 8.541364113788974e-06,
"loss": 1.2224643230438232,
"step": 570
},
{
"epoch": 1.2093023255813953,
"grad_norm": 1.439972996711731,
"learning_rate": 8.529748959393855e-06,
"loss": 1.039496898651123,
"step": 572
},
{
"epoch": 1.2135306553911205,
"grad_norm": 1.3833200931549072,
"learning_rate": 8.518096765100485e-06,
"loss": 0.7393906712532043,
"step": 574
},
{
"epoch": 1.2177589852008457,
"grad_norm": 2.091217279434204,
"learning_rate": 8.50640767336162e-06,
"loss": 0.8442506790161133,
"step": 576
},
{
"epoch": 1.221987315010571,
"grad_norm": 3.9008047580718994,
"learning_rate": 8.494681827081096e-06,
"loss": 0.9975602030754089,
"step": 578
},
{
"epoch": 1.226215644820296,
"grad_norm": 1.7783403396606445,
"learning_rate": 8.482919369612103e-06,
"loss": 0.7915345430374146,
"step": 580
},
{
"epoch": 1.230443974630021,
"grad_norm": 0.7436689138412476,
"learning_rate": 8.471120444755405e-06,
"loss": 1.054902195930481,
"step": 582
},
{
"epoch": 1.2346723044397463,
"grad_norm": 2.4143307209014893,
"learning_rate": 8.4592851967576e-06,
"loss": 0.9139824509620667,
"step": 584
},
{
"epoch": 1.2389006342494715,
"grad_norm": 1.2315016984939575,
"learning_rate": 8.447413770309349e-06,
"loss": 0.7946511507034302,
"step": 586
},
{
"epoch": 1.2431289640591967,
"grad_norm": 1.110283613204956,
"learning_rate": 8.435506310543607e-06,
"loss": 0.9444854855537415,
"step": 588
},
{
"epoch": 1.2473572938689217,
"grad_norm": 1.010946273803711,
"learning_rate": 8.42356296303386e-06,
"loss": 0.6641428470611572,
"step": 590
},
{
"epoch": 1.251585623678647,
"grad_norm": 0.6903517246246338,
"learning_rate": 8.411583873792325e-06,
"loss": 0.803213894367218,
"step": 592
},
{
"epoch": 1.255813953488372,
"grad_norm": 2.2428789138793945,
"learning_rate": 8.399569189268182e-06,
"loss": 1.047710657119751,
"step": 594
},
{
"epoch": 1.2600422832980973,
"grad_norm": 1.3929978609085083,
"learning_rate": 8.387519056345778e-06,
"loss": 0.6649131178855896,
"step": 596
},
{
"epoch": 1.2642706131078225,
"grad_norm": 1.124123454093933,
"learning_rate": 8.375433622342824e-06,
"loss": 1.0770790576934814,
"step": 598
},
{
"epoch": 1.2684989429175475,
"grad_norm": 1.6450690031051636,
"learning_rate": 8.363313035008613e-06,
"loss": 1.2491010427474976,
"step": 600
},
{
"epoch": 1.2727272727272727,
"grad_norm": 0.8932709693908691,
"learning_rate": 8.351157442522187e-06,
"loss": 0.9033283591270447,
"step": 602
},
{
"epoch": 1.276955602536998,
"grad_norm": 0.915249228477478,
"learning_rate": 8.338966993490554e-06,
"loss": 1.1027641296386719,
"step": 604
},
{
"epoch": 1.2811839323467231,
"grad_norm": 2.09724760055542,
"learning_rate": 8.326741836946852e-06,
"loss": 0.7590098977088928,
"step": 606
},
{
"epoch": 1.285412262156448,
"grad_norm": 1.21298348903656,
"learning_rate": 8.314482122348523e-06,
"loss": 1.124398112297058,
"step": 608
},
{
"epoch": 1.2896405919661733,
"grad_norm": 0.7811681032180786,
"learning_rate": 8.30218799957551e-06,
"loss": 0.7333798408508301,
"step": 610
},
{
"epoch": 1.2938689217758985,
"grad_norm": 1.256026029586792,
"learning_rate": 8.289859618928404e-06,
"loss": 0.6493385434150696,
"step": 612
},
{
"epoch": 1.2980972515856237,
"grad_norm": 1.7578778266906738,
"learning_rate": 8.277497131126608e-06,
"loss": 0.8620043396949768,
"step": 614
},
{
"epoch": 1.302325581395349,
"grad_norm": 8.236961364746094,
"learning_rate": 8.265100687306503e-06,
"loss": 0.5509307980537415,
"step": 616
},
{
"epoch": 1.306553911205074,
"grad_norm": 5.198338508605957,
"learning_rate": 8.252670439019595e-06,
"loss": 0.8596081137657166,
"step": 618
},
{
"epoch": 1.3107822410147991,
"grad_norm": 1.3249956369400024,
"learning_rate": 8.240206538230663e-06,
"loss": 1.1268280744552612,
"step": 620
},
{
"epoch": 1.3150105708245243,
"grad_norm": 0.7324321269989014,
"learning_rate": 8.227709137315905e-06,
"loss": 1.0642417669296265,
"step": 622
},
{
"epoch": 1.3192389006342495,
"grad_norm": 2.9953842163085938,
"learning_rate": 8.215178389061064e-06,
"loss": 0.6910127401351929,
"step": 624
},
{
"epoch": 1.3234672304439745,
"grad_norm": 0.9970926642417908,
"learning_rate": 8.202614446659572e-06,
"loss": 1.0903798341751099,
"step": 626
},
{
"epoch": 1.3276955602536997,
"grad_norm": 3.3420803546905518,
"learning_rate": 8.190017463710675e-06,
"loss": 1.2222436666488647,
"step": 628
},
{
"epoch": 1.331923890063425,
"grad_norm": 0.9811909198760986,
"learning_rate": 8.177387594217548e-06,
"loss": 0.9634477496147156,
"step": 630
},
{
"epoch": 1.3361522198731501,
"grad_norm": 1.1461257934570312,
"learning_rate": 8.164724992585421e-06,
"loss": 0.5971619486808777,
"step": 632
},
{
"epoch": 1.3403805496828753,
"grad_norm": 0.7688302397727966,
"learning_rate": 8.152029813619685e-06,
"loss": 0.681722104549408,
"step": 634
},
{
"epoch": 1.3446088794926006,
"grad_norm": 0.5482288599014282,
"learning_rate": 8.139302212524006e-06,
"loss": 0.7553455233573914,
"step": 636
},
{
"epoch": 1.3488372093023255,
"grad_norm": 5.6167378425598145,
"learning_rate": 8.126542344898419e-06,
"loss": 0.9999579787254333,
"step": 638
},
{
"epoch": 1.3530655391120507,
"grad_norm": 1.6354902982711792,
"learning_rate": 8.113750366737431e-06,
"loss": 0.6705687046051025,
"step": 640
},
{
"epoch": 1.357293868921776,
"grad_norm": 0.7095193862915039,
"learning_rate": 8.100926434428116e-06,
"loss": 1.2020138502120972,
"step": 642
},
{
"epoch": 1.361522198731501,
"grad_norm": 1.59982430934906,
"learning_rate": 8.0880707047482e-06,
"loss": 0.9918311834335327,
"step": 644
},
{
"epoch": 1.3657505285412261,
"grad_norm": 4.373168468475342,
"learning_rate": 8.075183334864139e-06,
"loss": 0.736129879951477,
"step": 646
},
{
"epoch": 1.3699788583509513,
"grad_norm": 1.850530982017517,
"learning_rate": 8.062264482329209e-06,
"loss": 0.72136390209198,
"step": 648
},
{
"epoch": 1.3742071881606766,
"grad_norm": 3.145092010498047,
"learning_rate": 8.049314305081575e-06,
"loss": 0.742005467414856,
"step": 650
},
{
"epoch": 1.3784355179704018,
"grad_norm": 1.972105622291565,
"learning_rate": 8.036332961442355e-06,
"loss": 0.9196822047233582,
"step": 652
},
{
"epoch": 1.382663847780127,
"grad_norm": 1.3466448783874512,
"learning_rate": 8.023320610113691e-06,
"loss": 1.1615943908691406,
"step": 654
},
{
"epoch": 1.386892177589852,
"grad_norm": 1.0326884984970093,
"learning_rate": 8.010277410176801e-06,
"loss": 1.059496283531189,
"step": 656
},
{
"epoch": 1.3911205073995772,
"grad_norm": 1.907827615737915,
"learning_rate": 7.997203521090053e-06,
"loss": 0.8208768367767334,
"step": 658
},
{
"epoch": 1.3953488372093024,
"grad_norm": 0.8312616944313049,
"learning_rate": 7.984099102686986e-06,
"loss": 0.9305316209793091,
"step": 660
},
{
"epoch": 1.3995771670190273,
"grad_norm": 0.637813925743103,
"learning_rate": 7.970964315174384e-06,
"loss": 1.0702272653579712,
"step": 662
},
{
"epoch": 1.4038054968287526,
"grad_norm": 0.27447155117988586,
"learning_rate": 7.9577993191303e-06,
"loss": 0.9092645645141602,
"step": 664
},
{
"epoch": 1.4080338266384778,
"grad_norm": 1.9821829795837402,
"learning_rate": 7.944604275502102e-06,
"loss": 1.3273407220840454,
"step": 666
},
{
"epoch": 1.412262156448203,
"grad_norm": 0.7487687468528748,
"learning_rate": 7.931379345604495e-06,
"loss": 1.063726782798767,
"step": 668
},
{
"epoch": 1.4164904862579282,
"grad_norm": 2.504626989364624,
"learning_rate": 7.918124691117563e-06,
"loss": 0.7032148838043213,
"step": 670
},
{
"epoch": 1.4207188160676534,
"grad_norm": 0.645386815071106,
"learning_rate": 7.904840474084783e-06,
"loss": 1.1364572048187256,
"step": 672
},
{
"epoch": 1.4249471458773784,
"grad_norm": 0.8274882435798645,
"learning_rate": 7.891526856911046e-06,
"loss": 0.5882174968719482,
"step": 674
},
{
"epoch": 1.4291754756871036,
"grad_norm": 1.3122429847717285,
"learning_rate": 7.878184002360664e-06,
"loss": 0.7344486117362976,
"step": 676
},
{
"epoch": 1.4334038054968288,
"grad_norm": 0.8297415971755981,
"learning_rate": 7.864812073555402e-06,
"loss": 1.257372498512268,
"step": 678
},
{
"epoch": 1.437632135306554,
"grad_norm": 0.9948843121528625,
"learning_rate": 7.851411233972461e-06,
"loss": 1.1502310037612915,
"step": 680
},
{
"epoch": 1.441860465116279,
"grad_norm": 1.9943336248397827,
"learning_rate": 7.837981647442486e-06,
"loss": 1.1037263870239258,
"step": 682
},
{
"epoch": 1.4460887949260042,
"grad_norm": 1.4379900693893433,
"learning_rate": 7.82452347814757e-06,
"loss": 1.0680630207061768,
"step": 684
},
{
"epoch": 1.4503171247357294,
"grad_norm": 2.189563512802124,
"learning_rate": 7.811036890619239e-06,
"loss": 1.4032574892044067,
"step": 686
},
{
"epoch": 1.4545454545454546,
"grad_norm": 1.00834321975708,
"learning_rate": 7.797522049736441e-06,
"loss": 1.0359902381896973,
"step": 688
},
{
"epoch": 1.4587737843551798,
"grad_norm": 3.6665594577789307,
"learning_rate": 7.78397912072354e-06,
"loss": 0.4269692003726959,
"step": 690
},
{
"epoch": 1.463002114164905,
"grad_norm": 0.7096814513206482,
"learning_rate": 7.770408269148281e-06,
"loss": 1.1489344835281372,
"step": 692
},
{
"epoch": 1.46723044397463,
"grad_norm": 1.303365707397461,
"learning_rate": 7.756809660919778e-06,
"loss": 0.5354632139205933,
"step": 694
},
{
"epoch": 1.4714587737843552,
"grad_norm": 1.7412745952606201,
"learning_rate": 7.74318346228648e-06,
"loss": 1.0408984422683716,
"step": 696
},
{
"epoch": 1.4756871035940804,
"grad_norm": 0.9331637024879456,
"learning_rate": 7.72952983983414e-06,
"loss": 1.1087017059326172,
"step": 698
},
{
"epoch": 1.4799154334038054,
"grad_norm": 1.4755645990371704,
"learning_rate": 7.715848960483776e-06,
"loss": 0.9330304861068726,
"step": 700
},
{
"epoch": 1.4841437632135306,
"grad_norm": 0.7734619379043579,
"learning_rate": 7.702140991489637e-06,
"loss": 1.0741868019104004,
"step": 702
},
{
"epoch": 1.4883720930232558,
"grad_norm": 0.79962158203125,
"learning_rate": 7.688406100437149e-06,
"loss": 1.0055755376815796,
"step": 704
},
{
"epoch": 1.492600422832981,
"grad_norm": 0.6755836606025696,
"learning_rate": 7.674644455240878e-06,
"loss": 0.7014404535293579,
"step": 706
},
{
"epoch": 1.4968287526427062,
"grad_norm": 0.4090423285961151,
"learning_rate": 7.660856224142456e-06,
"loss": 0.25834783911705017,
"step": 708
},
{
"epoch": 1.5010570824524314,
"grad_norm": 1.8259189128875732,
"learning_rate": 7.647041575708555e-06,
"loss": 0.6055929660797119,
"step": 710
},
{
"epoch": 1.5052854122621564,
"grad_norm": 2.0474252700805664,
"learning_rate": 7.633200678828804e-06,
"loss": 0.55499267578125,
"step": 712
},
{
"epoch": 1.5095137420718816,
"grad_norm": 1.0860965251922607,
"learning_rate": 7.6193337027137215e-06,
"loss": 0.814342200756073,
"step": 714
},
{
"epoch": 1.5137420718816068,
"grad_norm": 1.0963035821914673,
"learning_rate": 7.605440816892667e-06,
"loss": 1.0843336582183838,
"step": 716
},
{
"epoch": 1.5179704016913318,
"grad_norm": 1.1276558637619019,
"learning_rate": 7.591522191211746e-06,
"loss": 0.9576719403266907,
"step": 718
},
{
"epoch": 1.522198731501057,
"grad_norm": 0.4725880026817322,
"learning_rate": 7.577577995831755e-06,
"loss": 0.8506250381469727,
"step": 720
},
{
"epoch": 1.5264270613107822,
"grad_norm": 2.1232128143310547,
"learning_rate": 7.563608401226081e-06,
"loss": 0.850522518157959,
"step": 722
},
{
"epoch": 1.5306553911205074,
"grad_norm": 0.8134534955024719,
"learning_rate": 7.549613578178632e-06,
"loss": 0.8823297619819641,
"step": 724
},
{
"epoch": 1.5348837209302326,
"grad_norm": 1.0317206382751465,
"learning_rate": 7.535593697781741e-06,
"loss": 0.9172330498695374,
"step": 726
},
{
"epoch": 1.5391120507399578,
"grad_norm": 1.1433746814727783,
"learning_rate": 7.52154893143408e-06,
"loss": 0.6046597361564636,
"step": 728
},
{
"epoch": 1.543340380549683,
"grad_norm": 5.408229827880859,
"learning_rate": 7.50747945083856e-06,
"loss": 0.8174023628234863,
"step": 730
},
{
"epoch": 1.547568710359408,
"grad_norm": 0.6964724659919739,
"learning_rate": 7.493385428000234e-06,
"loss": 1.0265309810638428,
"step": 732
},
{
"epoch": 1.5517970401691332,
"grad_norm": 1.0122097730636597,
"learning_rate": 7.479267035224191e-06,
"loss": 0.2555813193321228,
"step": 734
},
{
"epoch": 1.5560253699788582,
"grad_norm": 1.2480603456497192,
"learning_rate": 7.465124445113459e-06,
"loss": 0.9673231244087219,
"step": 736
},
{
"epoch": 1.5602536997885834,
"grad_norm": 0.38378793001174927,
"learning_rate": 7.450957830566878e-06,
"loss": 0.7110365629196167,
"step": 738
},
{
"epoch": 1.5644820295983086,
"grad_norm": 0.5784710645675659,
"learning_rate": 7.436767364777009e-06,
"loss": 1.0822460651397705,
"step": 740
},
{
"epoch": 1.5687103594080338,
"grad_norm": 2.4423210620880127,
"learning_rate": 7.422553221227993e-06,
"loss": 0.7957349419593811,
"step": 742
},
{
"epoch": 1.572938689217759,
"grad_norm": 1.2780396938323975,
"learning_rate": 7.4083155736934484e-06,
"loss": 1.1004595756530762,
"step": 744
},
{
"epoch": 1.5771670190274842,
"grad_norm": 1.350319504737854,
"learning_rate": 7.394054596234336e-06,
"loss": 1.187515377998352,
"step": 746
},
{
"epoch": 1.5813953488372094,
"grad_norm": 1.1918803453445435,
"learning_rate": 7.379770463196831e-06,
"loss": 0.8736639022827148,
"step": 748
},
{
"epoch": 1.5856236786469344,
"grad_norm": 1.0435072183609009,
"learning_rate": 7.365463349210203e-06,
"loss": 0.06134287640452385,
"step": 750
},
{
"epoch": 1.5898520084566596,
"grad_norm": 1.177916169166565,
"learning_rate": 7.351133429184666e-06,
"loss": 0.8378365635871887,
"step": 752
},
{
"epoch": 1.5940803382663846,
"grad_norm": 1.514112949371338,
"learning_rate": 7.336780878309251e-06,
"loss": 1.1457360982894897,
"step": 754
},
{
"epoch": 1.5983086680761098,
"grad_norm": 1.1603827476501465,
"learning_rate": 7.322405872049657e-06,
"loss": 1.079711675643921,
"step": 756
},
{
"epoch": 1.602536997885835,
"grad_norm": 0.8602437973022461,
"learning_rate": 7.308008586146111e-06,
"loss": 1.1031216382980347,
"step": 758
},
{
"epoch": 1.6067653276955602,
"grad_norm": 1.977730393409729,
"learning_rate": 7.293589196611221e-06,
"loss": 1.0956496000289917,
"step": 760
},
{
"epoch": 1.6109936575052854,
"grad_norm": 0.8727390766143799,
"learning_rate": 7.279147879727811e-06,
"loss": 1.0630828142166138,
"step": 762
},
{
"epoch": 1.6152219873150107,
"grad_norm": 1.6057307720184326,
"learning_rate": 7.264684812046787e-06,
"loss": 1.01640784740448,
"step": 764
},
{
"epoch": 1.6194503171247359,
"grad_norm": 8.582555770874023,
"learning_rate": 7.250200170384961e-06,
"loss": 0.8622811436653137,
"step": 766
},
{
"epoch": 1.6236786469344608,
"grad_norm": 0.8186838030815125,
"learning_rate": 7.235694131822894e-06,
"loss": 1.0657317638397217,
"step": 768
},
{
"epoch": 1.627906976744186,
"grad_norm": 1.5353238582611084,
"learning_rate": 7.221166873702738e-06,
"loss": 1.078508734703064,
"step": 770
},
{
"epoch": 1.6321353065539113,
"grad_norm": 0.891819179058075,
"learning_rate": 7.206618573626058e-06,
"loss": 1.070008635520935,
"step": 772
},
{
"epoch": 1.6363636363636362,
"grad_norm": 1.741459608078003,
"learning_rate": 7.19204940945167e-06,
"loss": 0.786920428276062,
"step": 774
},
{
"epoch": 1.6405919661733614,
"grad_norm": 0.9303106069564819,
"learning_rate": 7.177459559293456e-06,
"loss": 1.0890278816223145,
"step": 776
},
{
"epoch": 1.6448202959830867,
"grad_norm": 1.183817982673645,
"learning_rate": 7.162849201518198e-06,
"loss": 0.6884796619415283,
"step": 778
},
{
"epoch": 1.6490486257928119,
"grad_norm": 1.7773730754852295,
"learning_rate": 7.1482185147433905e-06,
"loss": 1.0220835208892822,
"step": 780
},
{
"epoch": 1.653276955602537,
"grad_norm": 0.8902779221534729,
"learning_rate": 7.133567677835055e-06,
"loss": 0.7693580985069275,
"step": 782
},
{
"epoch": 1.6575052854122623,
"grad_norm": 1.2130192518234253,
"learning_rate": 7.11889686990556e-06,
"loss": 1.1889230012893677,
"step": 784
},
{
"epoch": 1.6617336152219875,
"grad_norm": 2.664362907409668,
"learning_rate": 7.104206270311423e-06,
"loss": 0.49840009212493896,
"step": 786
},
{
"epoch": 1.6659619450317125,
"grad_norm": 0.8622718453407288,
"learning_rate": 7.089496058651133e-06,
"loss": 1.0763899087905884,
"step": 788
},
{
"epoch": 1.6701902748414377,
"grad_norm": 0.92386394739151,
"learning_rate": 7.0747664147629335e-06,
"loss": 0.7635555863380432,
"step": 790
},
{
"epoch": 1.6744186046511627,
"grad_norm": 1.7060062885284424,
"learning_rate": 7.060017518722637e-06,
"loss": 0.4200564920902252,
"step": 792
},
{
"epoch": 1.6786469344608879,
"grad_norm": 2.1876258850097656,
"learning_rate": 7.0452495508414276e-06,
"loss": 1.0807533264160156,
"step": 794
},
{
"epoch": 1.682875264270613,
"grad_norm": 0.8581140041351318,
"learning_rate": 7.030462691663641e-06,
"loss": 0.6274865865707397,
"step": 796
},
{
"epoch": 1.6871035940803383,
"grad_norm": 0.7943618893623352,
"learning_rate": 7.015657121964575e-06,
"loss": 1.1716108322143555,
"step": 798
},
{
"epoch": 1.6913319238900635,
"grad_norm": 0.9652339816093445,
"learning_rate": 7.000833022748267e-06,
"loss": 1.0458505153656006,
"step": 800
},
{
"epoch": 1.6955602536997887,
"grad_norm": 2.7819712162017822,
"learning_rate": 6.985990575245283e-06,
"loss": 0.5537415742874146,
"step": 802
},
{
"epoch": 1.699788583509514,
"grad_norm": 0.7498971223831177,
"learning_rate": 6.97112996091051e-06,
"loss": 1.04530930519104,
"step": 804
},
{
"epoch": 1.7040169133192389,
"grad_norm": 6.918619632720947,
"learning_rate": 6.956251361420926e-06,
"loss": 0.7852331399917603,
"step": 806
},
{
"epoch": 1.708245243128964,
"grad_norm": 3.3758692741394043,
"learning_rate": 6.94135495867339e-06,
"loss": 0.7689675688743591,
"step": 808
},
{
"epoch": 1.712473572938689,
"grad_norm": 1.0341968536376953,
"learning_rate": 6.926440934782407e-06,
"loss": 1.0918835401535034,
"step": 810
},
{
"epoch": 1.7167019027484143,
"grad_norm": 9.2674560546875,
"learning_rate": 6.9115094720779155e-06,
"loss": 1.3195767402648926,
"step": 812
},
{
"epoch": 1.7209302325581395,
"grad_norm": 0.7892317771911621,
"learning_rate": 6.896560753103041e-06,
"loss": 1.050921082496643,
"step": 814
},
{
"epoch": 1.7251585623678647,
"grad_norm": 1.4375678300857544,
"learning_rate": 6.881594960611881e-06,
"loss": 1.138192057609558,
"step": 816
},
{
"epoch": 1.72938689217759,
"grad_norm": 0.9784802198410034,
"learning_rate": 6.866612277567261e-06,
"loss": 0.9113901257514954,
"step": 818
},
{
"epoch": 1.733615221987315,
"grad_norm": 12.666766166687012,
"learning_rate": 6.851612887138501e-06,
"loss": 0.9583132266998291,
"step": 820
},
{
"epoch": 1.7378435517970403,
"grad_norm": 3.450157403945923,
"learning_rate": 6.836596972699175e-06,
"loss": 0.6922838687896729,
"step": 822
},
{
"epoch": 1.7420718816067653,
"grad_norm": 0.5298084020614624,
"learning_rate": 6.82156471782487e-06,
"loss": 0.6028211116790771,
"step": 824
},
{
"epoch": 1.7463002114164905,
"grad_norm": 3.4586191177368164,
"learning_rate": 6.806516306290942e-06,
"loss": 0.911165177822113,
"step": 826
},
{
"epoch": 1.7505285412262155,
"grad_norm": 1.0856589078903198,
"learning_rate": 6.791451922070265e-06,
"loss": 0.8932915329933167,
"step": 828
},
{
"epoch": 1.7547568710359407,
"grad_norm": 6.422530174255371,
"learning_rate": 6.77637174933099e-06,
"loss": 0.7858769297599792,
"step": 830
},
{
"epoch": 1.758985200845666,
"grad_norm": 1.1341559886932373,
"learning_rate": 6.7612759724342895e-06,
"loss": 1.1220742464065552,
"step": 832
},
{
"epoch": 1.763213530655391,
"grad_norm": 2.2434422969818115,
"learning_rate": 6.746164775932096e-06,
"loss": 0.7907599806785583,
"step": 834
},
{
"epoch": 1.7674418604651163,
"grad_norm": 1.2704675197601318,
"learning_rate": 6.731038344564862e-06,
"loss": 0.7747330665588379,
"step": 836
},
{
"epoch": 1.7716701902748415,
"grad_norm": 5.313762187957764,
"learning_rate": 6.715896863259286e-06,
"loss": 1.1435916423797607,
"step": 838
},
{
"epoch": 1.7758985200845667,
"grad_norm": 0.977674126625061,
"learning_rate": 6.700740517126061e-06,
"loss": 0.8452804088592529,
"step": 840
},
{
"epoch": 1.7801268498942917,
"grad_norm": 1.3944824934005737,
"learning_rate": 6.685569491457608e-06,
"loss": 0.3716931939125061,
"step": 842
},
{
"epoch": 1.784355179704017,
"grad_norm": 0.833726704120636,
"learning_rate": 6.670383971725809e-06,
"loss": 0.64637690782547,
"step": 844
},
{
"epoch": 1.7885835095137421,
"grad_norm": 3.612642288208008,
"learning_rate": 6.655184143579746e-06,
"loss": 0.9325193762779236,
"step": 846
},
{
"epoch": 1.792811839323467,
"grad_norm": 1.4551070928573608,
"learning_rate": 6.639970192843423e-06,
"loss": 0.7159190773963928,
"step": 848
},
{
"epoch": 1.7970401691331923,
"grad_norm": 1.945599913597107,
"learning_rate": 6.624742305513502e-06,
"loss": 0.684307336807251,
"step": 850
},
{
"epoch": 1.8012684989429175,
"grad_norm": 1.1321507692337036,
"learning_rate": 6.609500667757022e-06,
"loss": 0.9418154358863831,
"step": 852
},
{
"epoch": 1.8054968287526427,
"grad_norm": 2.010143518447876,
"learning_rate": 6.594245465909127e-06,
"loss": 0.8230486512184143,
"step": 854
},
{
"epoch": 1.809725158562368,
"grad_norm": 8.034485816955566,
"learning_rate": 6.578976886470791e-06,
"loss": 1.0074877738952637,
"step": 856
},
{
"epoch": 1.8139534883720931,
"grad_norm": 1.209224820137024,
"learning_rate": 6.56369511610653e-06,
"loss": 0.6451162099838257,
"step": 858
},
{
"epoch": 1.8181818181818183,
"grad_norm": 3.1519598960876465,
"learning_rate": 6.548400341642128e-06,
"loss": 1.0073386430740356,
"step": 860
},
{
"epoch": 1.8224101479915433,
"grad_norm": 1.372086763381958,
"learning_rate": 6.533092750062345e-06,
"loss": 1.040485143661499,
"step": 862
},
{
"epoch": 1.8266384778012685,
"grad_norm": 1.778725504875183,
"learning_rate": 6.5177725285086416e-06,
"loss": 0.5030984878540039,
"step": 864
},
{
"epoch": 1.8308668076109935,
"grad_norm": 0.6900845170021057,
"learning_rate": 6.5024398642768785e-06,
"loss": 0.7327762842178345,
"step": 866
},
{
"epoch": 1.8350951374207187,
"grad_norm": 0.6955262422561646,
"learning_rate": 6.487094944815036e-06,
"loss": 1.0692901611328125,
"step": 868
},
{
"epoch": 1.839323467230444,
"grad_norm": 0.739895224571228,
"learning_rate": 6.47173795772092e-06,
"loss": 1.0716890096664429,
"step": 870
},
{
"epoch": 1.8435517970401691,
"grad_norm": 0.5424486994743347,
"learning_rate": 6.456369090739869e-06,
"loss": 0.8757049441337585,
"step": 872
},
{
"epoch": 1.8477801268498943,
"grad_norm": 1.1952996253967285,
"learning_rate": 6.440988531762454e-06,
"loss": 1.076974630355835,
"step": 874
},
{
"epoch": 1.8520084566596196,
"grad_norm": 6.045002460479736,
"learning_rate": 6.42559646882219e-06,
"loss": 0.9145498871803284,
"step": 876
},
{
"epoch": 1.8562367864693448,
"grad_norm": 12.369074821472168,
"learning_rate": 6.410193090093227e-06,
"loss": 0.41486606001853943,
"step": 878
},
{
"epoch": 1.8604651162790697,
"grad_norm": 10.631083488464355,
"learning_rate": 6.394778583888064e-06,
"loss": 0.6638185381889343,
"step": 880
},
{
"epoch": 1.864693446088795,
"grad_norm": 1.150240421295166,
"learning_rate": 6.379353138655227e-06,
"loss": 1.0596197843551636,
"step": 882
},
{
"epoch": 1.86892177589852,
"grad_norm": 0.6067917943000793,
"learning_rate": 6.3639169429769805e-06,
"loss": 0.7086359262466431,
"step": 884
},
{
"epoch": 1.8731501057082451,
"grad_norm": 1.6421762704849243,
"learning_rate": 6.348470185567025e-06,
"loss": 1.125901222229004,
"step": 886
},
{
"epoch": 1.8773784355179703,
"grad_norm": 1.1221742630004883,
"learning_rate": 6.333013055268169e-06,
"loss": 1.1121491193771362,
"step": 888
},
{
"epoch": 1.8816067653276956,
"grad_norm": 5.079599857330322,
"learning_rate": 6.317545741050044e-06,
"loss": 0.6150096654891968,
"step": 890
},
{
"epoch": 1.8858350951374208,
"grad_norm": 3.326103925704956,
"learning_rate": 6.3020684320067805e-06,
"loss": 0.8114713430404663,
"step": 892
},
{
"epoch": 1.890063424947146,
"grad_norm": 1.2685209512710571,
"learning_rate": 6.2865813173546976e-06,
"loss": 1.1205917596817017,
"step": 894
},
{
"epoch": 1.8942917547568712,
"grad_norm": 2.238971710205078,
"learning_rate": 6.271084586429994e-06,
"loss": 1.3695721626281738,
"step": 896
},
{
"epoch": 1.8985200845665962,
"grad_norm": 1.6322273015975952,
"learning_rate": 6.255578428686433e-06,
"loss": 1.066908597946167,
"step": 898
},
{
"epoch": 1.9027484143763214,
"grad_norm": 26.275907516479492,
"learning_rate": 6.240063033693023e-06,
"loss": 0.49670320749282837,
"step": 900
},
{
"epoch": 1.9069767441860463,
"grad_norm": 0.7586657404899597,
"learning_rate": 6.2245385911316994e-06,
"loss": 1.0753251314163208,
"step": 902
},
{
"epoch": 1.9112050739957716,
"grad_norm": 1.5362434387207031,
"learning_rate": 6.2090052907950096e-06,
"loss": 0.49595069885253906,
"step": 904
},
{
"epoch": 1.9154334038054968,
"grad_norm": 0.9343399405479431,
"learning_rate": 6.193463322583792e-06,
"loss": 0.8677418231964111,
"step": 906
},
{
"epoch": 1.919661733615222,
"grad_norm": 1.4838796854019165,
"learning_rate": 6.177912876504851e-06,
"loss": 1.180745244026184,
"step": 908
},
{
"epoch": 1.9238900634249472,
"grad_norm": 0.7102862596511841,
"learning_rate": 6.16235414266864e-06,
"loss": 0.7499901056289673,
"step": 910
},
{
"epoch": 1.9281183932346724,
"grad_norm": 0.8403708338737488,
"learning_rate": 6.146787311286927e-06,
"loss": 1.0172603130340576,
"step": 912
},
{
"epoch": 1.9323467230443976,
"grad_norm": 1.574170470237732,
"learning_rate": 6.131212572670483e-06,
"loss": 0.6987271308898926,
"step": 914
},
{
"epoch": 1.9365750528541226,
"grad_norm": 1.7023367881774902,
"learning_rate": 6.115630117226741e-06,
"loss": 1.072776436805725,
"step": 916
},
{
"epoch": 1.9408033826638478,
"grad_norm": 1.382447361946106,
"learning_rate": 6.1000401354574855e-06,
"loss": 0.6643326282501221,
"step": 918
},
{
"epoch": 1.945031712473573,
"grad_norm": 2.3708441257476807,
"learning_rate": 6.084442817956502e-06,
"loss": 0.8780905604362488,
"step": 920
},
{
"epoch": 1.949260042283298,
"grad_norm": 18.979156494140625,
"learning_rate": 6.0688383554072625e-06,
"loss": 0.9448009729385376,
"step": 922
},
{
"epoch": 1.9534883720930232,
"grad_norm": 0.6649048328399658,
"learning_rate": 6.053226938580596e-06,
"loss": 0.9952071309089661,
"step": 924
},
{
"epoch": 1.9577167019027484,
"grad_norm": 0.47499391436576843,
"learning_rate": 6.03760875833234e-06,
"loss": 0.8334546089172363,
"step": 926
},
{
"epoch": 1.9619450317124736,
"grad_norm": 1.3545820713043213,
"learning_rate": 6.021984005601027e-06,
"loss": 1.1505999565124512,
"step": 928
},
{
"epoch": 1.9661733615221988,
"grad_norm": 0.9247393608093262,
"learning_rate": 6.006352871405534e-06,
"loss": 0.7951589822769165,
"step": 930
},
{
"epoch": 1.970401691331924,
"grad_norm": 1.1587799787521362,
"learning_rate": 5.990715546842755e-06,
"loss": 1.2130314111709595,
"step": 932
},
{
"epoch": 1.9746300211416492,
"grad_norm": 2.277764320373535,
"learning_rate": 5.9750722230852694e-06,
"loss": 0.825849711894989,
"step": 934
},
{
"epoch": 1.9788583509513742,
"grad_norm": 1.0200295448303223,
"learning_rate": 5.959423091378992e-06,
"loss": 0.8113455772399902,
"step": 936
},
{
"epoch": 1.9830866807610994,
"grad_norm": 8.313407897949219,
"learning_rate": 5.943768343040845e-06,
"loss": 0.7946412563323975,
"step": 938
},
{
"epoch": 1.9873150105708244,
"grad_norm": 0.8906210064888,
"learning_rate": 5.928108169456419e-06,
"loss": 1.020690679550171,
"step": 940
},
{
"epoch": 1.9915433403805496,
"grad_norm": 1.765325903892517,
"learning_rate": 5.912442762077624e-06,
"loss": 1.0702815055847168,
"step": 942
},
{
"epoch": 1.9957716701902748,
"grad_norm": 1.177565336227417,
"learning_rate": 5.89677231242036e-06,
"loss": 1.0964148044586182,
"step": 944
},
{
"epoch": 2.0,
"grad_norm": 1.28970468044281,
"learning_rate": 5.8810970120621714e-06,
"loss": 0.3808255195617676,
"step": 946
},
{
"epoch": 2.004228329809725,
"grad_norm": 3.543426752090454,
"learning_rate": 5.8654170526399014e-06,
"loss": 0.7299563884735107,
"step": 948
},
{
"epoch": 2.0084566596194504,
"grad_norm": 0.9845713973045349,
"learning_rate": 5.849732625847354e-06,
"loss": 0.9211766719818115,
"step": 950
},
{
"epoch": 2.0126849894291756,
"grad_norm": 1.2199254035949707,
"learning_rate": 5.834043923432947e-06,
"loss": 0.7713860273361206,
"step": 952
},
{
"epoch": 2.016913319238901,
"grad_norm": 2.5373926162719727,
"learning_rate": 5.818351137197373e-06,
"loss": 1.0540279150009155,
"step": 954
},
{
"epoch": 2.0211416490486256,
"grad_norm": 1.191591501235962,
"learning_rate": 5.802654458991245e-06,
"loss": 1.0390864610671997,
"step": 956
},
{
"epoch": 2.025369978858351,
"grad_norm": 1.0018290281295776,
"learning_rate": 5.786954080712763e-06,
"loss": 0.6533144116401672,
"step": 958
},
{
"epoch": 2.029598308668076,
"grad_norm": 0.8343505859375,
"learning_rate": 5.7712501943053576e-06,
"loss": 0.9062074422836304,
"step": 960
},
{
"epoch": 2.033826638477801,
"grad_norm": 2.798297643661499,
"learning_rate": 5.755542991755352e-06,
"loss": 0.9095263481140137,
"step": 962
},
{
"epoch": 2.0380549682875264,
"grad_norm": 2.7954442501068115,
"learning_rate": 5.739832665089604e-06,
"loss": 0.7307310104370117,
"step": 964
},
{
"epoch": 2.0422832980972516,
"grad_norm": 1.134375810623169,
"learning_rate": 5.724119406373171e-06,
"loss": 0.622187077999115,
"step": 966
},
{
"epoch": 2.046511627906977,
"grad_norm": 3.2358651161193848,
"learning_rate": 5.708403407706954e-06,
"loss": 0.6713780164718628,
"step": 968
},
{
"epoch": 2.050739957716702,
"grad_norm": 0.7578688263893127,
"learning_rate": 5.692684861225349e-06,
"loss": 0.9283575415611267,
"step": 970
},
{
"epoch": 2.0549682875264272,
"grad_norm": 10.6370210647583,
"learning_rate": 5.676963959093904e-06,
"loss": 0.5982099771499634,
"step": 972
},
{
"epoch": 2.059196617336152,
"grad_norm": 1.1164742708206177,
"learning_rate": 5.66124089350696e-06,
"loss": 0.847841739654541,
"step": 974
},
{
"epoch": 2.063424947145877,
"grad_norm": 1.1554869413375854,
"learning_rate": 5.645515856685316e-06,
"loss": 0.8656512498855591,
"step": 976
},
{
"epoch": 2.0676532769556024,
"grad_norm": 1.6066093444824219,
"learning_rate": 5.629789040873862e-06,
"loss": 0.8995035290718079,
"step": 978
},
{
"epoch": 2.0718816067653276,
"grad_norm": 1.6776983737945557,
"learning_rate": 5.614060638339238e-06,
"loss": 0.7070065140724182,
"step": 980
},
{
"epoch": 2.076109936575053,
"grad_norm": 5.533592700958252,
"learning_rate": 5.598330841367487e-06,
"loss": 0.6764956712722778,
"step": 982
},
{
"epoch": 2.080338266384778,
"grad_norm": 0.7864822149276733,
"learning_rate": 5.582599842261693e-06,
"loss": 0.6749998331069946,
"step": 984
},
{
"epoch": 2.0845665961945032,
"grad_norm": 1.1458630561828613,
"learning_rate": 5.566867833339644e-06,
"loss": 0.7698004841804504,
"step": 986
},
{
"epoch": 2.0887949260042284,
"grad_norm": 1.1800228357315063,
"learning_rate": 5.551135006931467e-06,
"loss": 0.9230565428733826,
"step": 988
},
{
"epoch": 2.0930232558139537,
"grad_norm": 1.0281647443771362,
"learning_rate": 5.535401555377287e-06,
"loss": 0.8932913541793823,
"step": 990
},
{
"epoch": 2.097251585623679,
"grad_norm": 3.039360284805298,
"learning_rate": 5.51966767102487e-06,
"loss": 0.5426384806632996,
"step": 992
},
{
"epoch": 2.1014799154334036,
"grad_norm": 6.610354900360107,
"learning_rate": 5.503933546227273e-06,
"loss": 0.6222950220108032,
"step": 994
},
{
"epoch": 2.105708245243129,
"grad_norm": 0.8800019025802612,
"learning_rate": 5.4881993733404916e-06,
"loss": 0.6532884240150452,
"step": 996
},
{
"epoch": 2.109936575052854,
"grad_norm": 0.7089502215385437,
"learning_rate": 5.472465344721113e-06,
"loss": 0.8455125093460083,
"step": 998
},
{
"epoch": 2.1141649048625792,
"grad_norm": 1.3703752756118774,
"learning_rate": 5.456731652723959e-06,
"loss": 0.971792459487915,
"step": 1000
},
{
"epoch": 2.1183932346723044,
"grad_norm": 1.3223247528076172,
"learning_rate": 5.440998489699731e-06,
"loss": 0.24078744649887085,
"step": 1002
},
{
"epoch": 2.1226215644820297,
"grad_norm": 5.56455659866333,
"learning_rate": 5.425266047992672e-06,
"loss": 0.6732622385025024,
"step": 1004
},
{
"epoch": 2.126849894291755,
"grad_norm": 1.0352561473846436,
"learning_rate": 5.4095345199382e-06,
"loss": 0.8939829468727112,
"step": 1006
},
{
"epoch": 2.13107822410148,
"grad_norm": 4.4449639320373535,
"learning_rate": 5.393804097860568e-06,
"loss": 0.8419816493988037,
"step": 1008
},
{
"epoch": 2.1353065539112053,
"grad_norm": 0.8464795351028442,
"learning_rate": 5.3780749740705065e-06,
"loss": 0.6752206087112427,
"step": 1010
},
{
"epoch": 2.13953488372093,
"grad_norm": 1.4907701015472412,
"learning_rate": 5.362347340862869e-06,
"loss": 0.5165109634399414,
"step": 1012
},
{
"epoch": 2.1437632135306552,
"grad_norm": 0.9565569758415222,
"learning_rate": 5.3466213905142935e-06,
"loss": 0.6269171237945557,
"step": 1014
},
{
"epoch": 2.1479915433403804,
"grad_norm": 1.062368392944336,
"learning_rate": 5.3308973152808395e-06,
"loss": 0.76082444190979,
"step": 1016
},
{
"epoch": 2.1522198731501057,
"grad_norm": 0.949715256690979,
"learning_rate": 5.315175307395644e-06,
"loss": 0.9204181432723999,
"step": 1018
},
{
"epoch": 2.156448202959831,
"grad_norm": 2.2162368297576904,
"learning_rate": 5.299455559066571e-06,
"loss": 0.710947573184967,
"step": 1020
},
{
"epoch": 2.160676532769556,
"grad_norm": 1.2074542045593262,
"learning_rate": 5.283738262473859e-06,
"loss": 0.617449164390564,
"step": 1022
},
{
"epoch": 2.1649048625792813,
"grad_norm": 1.0769579410552979,
"learning_rate": 5.268023609767772e-06,
"loss": 0.8831707239151001,
"step": 1024
},
{
"epoch": 2.1691331923890065,
"grad_norm": 1.0203371047973633,
"learning_rate": 5.2523117930662545e-06,
"loss": 0.876832902431488,
"step": 1026
},
{
"epoch": 2.1733615221987317,
"grad_norm": 1.6831961870193481,
"learning_rate": 5.2366030044525785e-06,
"loss": 0.8998016119003296,
"step": 1028
},
{
"epoch": 2.177589852008457,
"grad_norm": 4.03355073928833,
"learning_rate": 5.220897435972992e-06,
"loss": 0.8626359701156616,
"step": 1030
},
{
"epoch": 2.1818181818181817,
"grad_norm": 0.95368891954422,
"learning_rate": 5.205195279634385e-06,
"loss": 0.8777475357055664,
"step": 1032
},
{
"epoch": 2.186046511627907,
"grad_norm": 1.8358793258666992,
"learning_rate": 5.1894967274019245e-06,
"loss": 1.0198477506637573,
"step": 1034
},
{
"epoch": 2.190274841437632,
"grad_norm": 1.433210849761963,
"learning_rate": 5.173801971196716e-06,
"loss": 0.6466475129127502,
"step": 1036
},
{
"epoch": 2.1945031712473573,
"grad_norm": 3.093515157699585,
"learning_rate": 5.158111202893464e-06,
"loss": 0.3333320915699005,
"step": 1038
},
{
"epoch": 2.1987315010570825,
"grad_norm": 1.1949971914291382,
"learning_rate": 5.142424614318112e-06,
"loss": 0.9545713067054749,
"step": 1040
},
{
"epoch": 2.2029598308668077,
"grad_norm": 13.295270919799805,
"learning_rate": 5.12674239724551e-06,
"loss": 0.32372453808784485,
"step": 1042
},
{
"epoch": 2.207188160676533,
"grad_norm": 0.9597116708755493,
"learning_rate": 5.111064743397061e-06,
"loss": 0.9214589595794678,
"step": 1044
},
{
"epoch": 2.211416490486258,
"grad_norm": 2.5795164108276367,
"learning_rate": 5.09539184443838e-06,
"loss": 0.46609729528427124,
"step": 1046
},
{
"epoch": 2.2156448202959833,
"grad_norm": 3.3308563232421875,
"learning_rate": 5.079723891976956e-06,
"loss": 0.7019236087799072,
"step": 1048
},
{
"epoch": 2.219873150105708,
"grad_norm": 2.7738068103790283,
"learning_rate": 5.064061077559803e-06,
"loss": 0.5528712272644043,
"step": 1050
},
{
"epoch": 2.2241014799154333,
"grad_norm": 1.3890830278396606,
"learning_rate": 5.048403592671116e-06,
"loss": 0.9125267863273621,
"step": 1052
},
{
"epoch": 2.2283298097251585,
"grad_norm": 1.012911081314087,
"learning_rate": 5.032751628729943e-06,
"loss": 0.9098288416862488,
"step": 1054
},
{
"epoch": 2.2325581395348837,
"grad_norm": 0.7264014482498169,
"learning_rate": 5.017105377087829e-06,
"loss": 0.9031175374984741,
"step": 1056
},
{
"epoch": 2.236786469344609,
"grad_norm": 0.3495601415634155,
"learning_rate": 5.0014650290264854e-06,
"loss": 0.658706784248352,
"step": 1058
},
{
"epoch": 2.241014799154334,
"grad_norm": 1.0988720655441284,
"learning_rate": 4.9858307757554536e-06,
"loss": 0.8154771327972412,
"step": 1060
},
{
"epoch": 2.2452431289640593,
"grad_norm": 1.5173739194869995,
"learning_rate": 4.97020280840976e-06,
"loss": 0.8678203821182251,
"step": 1062
},
{
"epoch": 2.2494714587737845,
"grad_norm": 1.5388199090957642,
"learning_rate": 4.954581318047582e-06,
"loss": 0.8165386915206909,
"step": 1064
},
{
"epoch": 2.2536997885835097,
"grad_norm": 1.4251195192337036,
"learning_rate": 4.938966495647918e-06,
"loss": 0.4654849171638489,
"step": 1066
},
{
"epoch": 2.2579281183932345,
"grad_norm": 0.9386773705482483,
"learning_rate": 4.923358532108244e-06,
"loss": 0.7494522333145142,
"step": 1068
},
{
"epoch": 2.2621564482029597,
"grad_norm": 2.6979904174804688,
"learning_rate": 4.907757618242186e-06,
"loss": 0.6902294158935547,
"step": 1070
},
{
"epoch": 2.266384778012685,
"grad_norm": 0.9414252638816833,
"learning_rate": 4.892163944777181e-06,
"loss": 0.8990166187286377,
"step": 1072
},
{
"epoch": 2.27061310782241,
"grad_norm": 1.597653865814209,
"learning_rate": 4.876577702352155e-06,
"loss": 0.7023258209228516,
"step": 1074
},
{
"epoch": 2.2748414376321353,
"grad_norm": 1.6104449033737183,
"learning_rate": 4.860999081515183e-06,
"loss": 0.7002600431442261,
"step": 1076
},
{
"epoch": 2.2790697674418605,
"grad_norm": 2.7923903465270996,
"learning_rate": 4.845428272721161e-06,
"loss": 0.24515867233276367,
"step": 1078
},
{
"epoch": 2.2832980972515857,
"grad_norm": 0.591924250125885,
"learning_rate": 4.8298654663294865e-06,
"loss": 0.16100549697875977,
"step": 1080
},
{
"epoch": 2.287526427061311,
"grad_norm": 1.7447562217712402,
"learning_rate": 4.814310852601717e-06,
"loss": 0.5650036931037903,
"step": 1082
},
{
"epoch": 2.291754756871036,
"grad_norm": 1.2828580141067505,
"learning_rate": 4.7987646216992545e-06,
"loss": 0.2993309497833252,
"step": 1084
},
{
"epoch": 2.295983086680761,
"grad_norm": 1.709354043006897,
"learning_rate": 4.783226963681018e-06,
"loss": 0.14297160506248474,
"step": 1086
},
{
"epoch": 2.300211416490486,
"grad_norm": 1.3488517999649048,
"learning_rate": 4.76769806850112e-06,
"loss": 0.5905637145042419,
"step": 1088
},
{
"epoch": 2.3044397463002113,
"grad_norm": 0.9696362018585205,
"learning_rate": 4.752178126006544e-06,
"loss": 0.896177351474762,
"step": 1090
},
{
"epoch": 2.3086680761099365,
"grad_norm": 0.9389773011207581,
"learning_rate": 4.736667325934819e-06,
"loss": 0.6187787652015686,
"step": 1092
},
{
"epoch": 2.3128964059196617,
"grad_norm": 0.8135393261909485,
"learning_rate": 4.721165857911709e-06,
"loss": 0.9242386817932129,
"step": 1094
},
{
"epoch": 2.317124735729387,
"grad_norm": 0.9435974359512329,
"learning_rate": 4.705673911448891e-06,
"loss": 0.8046470880508423,
"step": 1096
},
{
"epoch": 2.321353065539112,
"grad_norm": 1.2371445894241333,
"learning_rate": 4.690191675941631e-06,
"loss": 0.8578451871871948,
"step": 1098
},
{
"epoch": 2.3255813953488373,
"grad_norm": 1.9206478595733643,
"learning_rate": 4.674719340666477e-06,
"loss": 0.5210774540901184,
"step": 1100
},
{
"epoch": 2.3298097251585626,
"grad_norm": 1.8375836610794067,
"learning_rate": 4.659257094778949e-06,
"loss": 0.5852177739143372,
"step": 1102
},
{
"epoch": 2.3340380549682873,
"grad_norm": 1.4442607164382935,
"learning_rate": 4.643805127311212e-06,
"loss": 0.9156836271286011,
"step": 1104
},
{
"epoch": 2.3382663847780125,
"grad_norm": 1.0903940200805664,
"learning_rate": 4.628363627169774e-06,
"loss": 0.8298915028572083,
"step": 1106
},
{
"epoch": 2.3424947145877377,
"grad_norm": 0.8846871852874756,
"learning_rate": 4.612932783133182e-06,
"loss": 0.9094023108482361,
"step": 1108
},
{
"epoch": 2.346723044397463,
"grad_norm": 2.8649587631225586,
"learning_rate": 4.5975127838497036e-06,
"loss": 1.0473685264587402,
"step": 1110
},
{
"epoch": 2.350951374207188,
"grad_norm": 0.5405344367027283,
"learning_rate": 4.582103817835021e-06,
"loss": 0.5521120429039001,
"step": 1112
},
{
"epoch": 2.3551797040169133,
"grad_norm": 2.530195474624634,
"learning_rate": 4.566706073469938e-06,
"loss": 0.7570968866348267,
"step": 1114
},
{
"epoch": 2.3594080338266386,
"grad_norm": 1.1219722032546997,
"learning_rate": 4.551319738998067e-06,
"loss": 0.5681474804878235,
"step": 1116
},
{
"epoch": 2.3636363636363638,
"grad_norm": 2.5799829959869385,
"learning_rate": 4.535945002523524e-06,
"loss": 1.1965405941009521,
"step": 1118
},
{
"epoch": 2.367864693446089,
"grad_norm": 3.6101481914520264,
"learning_rate": 4.5205820520086415e-06,
"loss": 0.870458722114563,
"step": 1120
},
{
"epoch": 2.3720930232558137,
"grad_norm": 2.5263142585754395,
"learning_rate": 4.505231075271662e-06,
"loss": 0.8029850125312805,
"step": 1122
},
{
"epoch": 2.376321353065539,
"grad_norm": 1.3465577363967896,
"learning_rate": 4.489892259984446e-06,
"loss": 0.6342759728431702,
"step": 1124
},
{
"epoch": 2.380549682875264,
"grad_norm": 3.4513235092163086,
"learning_rate": 4.474565793670167e-06,
"loss": 0.928214430809021,
"step": 1126
},
{
"epoch": 2.3847780126849893,
"grad_norm": 2.604207754135132,
"learning_rate": 4.459251863701038e-06,
"loss": 0.601742148399353,
"step": 1128
},
{
"epoch": 2.3890063424947146,
"grad_norm": 1.0077075958251953,
"learning_rate": 4.443950657296004e-06,
"loss": 0.6908794641494751,
"step": 1130
},
{
"epoch": 2.3932346723044398,
"grad_norm": 3.369401693344116,
"learning_rate": 4.4286623615184585e-06,
"loss": 0.5740441679954529,
"step": 1132
},
{
"epoch": 2.397463002114165,
"grad_norm": 0.8096948266029358,
"learning_rate": 4.413387163273964e-06,
"loss": 0.8965785503387451,
"step": 1134
},
{
"epoch": 2.40169133192389,
"grad_norm": 6.409458637237549,
"learning_rate": 4.3981252493079506e-06,
"loss": 0.5919017791748047,
"step": 1136
},
{
"epoch": 2.4059196617336154,
"grad_norm": 1.5841705799102783,
"learning_rate": 4.382876806203453e-06,
"loss": 0.7373220920562744,
"step": 1138
},
{
"epoch": 2.41014799154334,
"grad_norm": 1.1616737842559814,
"learning_rate": 4.3676420203788125e-06,
"loss": 0.8810633420944214,
"step": 1140
},
{
"epoch": 2.4143763213530653,
"grad_norm": 3.2592217922210693,
"learning_rate": 4.352421078085405e-06,
"loss": 1.011488676071167,
"step": 1142
},
{
"epoch": 2.4186046511627906,
"grad_norm": 1.8064744472503662,
"learning_rate": 4.337214165405368e-06,
"loss": 0.4719372093677521,
"step": 1144
},
{
"epoch": 2.4228329809725158,
"grad_norm": 9.876317024230957,
"learning_rate": 4.322021468249316e-06,
"loss": 0.6632931232452393,
"step": 1146
},
{
"epoch": 2.427061310782241,
"grad_norm": 0.9402738213539124,
"learning_rate": 4.306843172354076e-06,
"loss": 0.6946466565132141,
"step": 1148
},
{
"epoch": 2.431289640591966,
"grad_norm": 1.5364139080047607,
"learning_rate": 4.291679463280416e-06,
"loss": 0.5582535862922668,
"step": 1150
},
{
"epoch": 2.4355179704016914,
"grad_norm": 0.6958863139152527,
"learning_rate": 4.276530526410768e-06,
"loss": 0.551296591758728,
"step": 1152
},
{
"epoch": 2.4397463002114166,
"grad_norm": 0.6311137676239014,
"learning_rate": 4.261396546946972e-06,
"loss": 0.4590912163257599,
"step": 1154
},
{
"epoch": 2.443974630021142,
"grad_norm": 0.7703436017036438,
"learning_rate": 4.246277709908009e-06,
"loss": 0.7646368741989136,
"step": 1156
},
{
"epoch": 2.448202959830867,
"grad_norm": 0.9557932019233704,
"learning_rate": 4.231174200127736e-06,
"loss": 0.5027541518211365,
"step": 1158
},
{
"epoch": 2.452431289640592,
"grad_norm": 1.3317209482192993,
"learning_rate": 4.216086202252624e-06,
"loss": 0.8837262988090515,
"step": 1160
},
{
"epoch": 2.456659619450317,
"grad_norm": 0.8966196179389954,
"learning_rate": 4.201013900739513e-06,
"loss": 0.8745890855789185,
"step": 1162
},
{
"epoch": 2.460887949260042,
"grad_norm": 0.9611190557479858,
"learning_rate": 4.185957479853344e-06,
"loss": 0.6712213754653931,
"step": 1164
},
{
"epoch": 2.4651162790697674,
"grad_norm": 1.8119922876358032,
"learning_rate": 4.1709171236649096e-06,
"loss": 1.0225939750671387,
"step": 1166
},
{
"epoch": 2.4693446088794926,
"grad_norm": 1.9253120422363281,
"learning_rate": 4.155893016048607e-06,
"loss": 1.0690964460372925,
"step": 1168
},
{
"epoch": 2.473572938689218,
"grad_norm": 1.2033724784851074,
"learning_rate": 4.140885340680189e-06,
"loss": 0.6253563165664673,
"step": 1170
},
{
"epoch": 2.477801268498943,
"grad_norm": 2.3583903312683105,
"learning_rate": 4.125894281034517e-06,
"loss": 0.7692979574203491,
"step": 1172
},
{
"epoch": 2.482029598308668,
"grad_norm": 1.6322160959243774,
"learning_rate": 4.110920020383316e-06,
"loss": 0.6338727474212646,
"step": 1174
},
{
"epoch": 2.4862579281183934,
"grad_norm": 1.3538206815719604,
"learning_rate": 4.0959627417929386e-06,
"loss": 0.6970939636230469,
"step": 1176
},
{
"epoch": 2.4904862579281186,
"grad_norm": 4.530666828155518,
"learning_rate": 4.081022628122127e-06,
"loss": 0.7442941665649414,
"step": 1178
},
{
"epoch": 2.4947145877378434,
"grad_norm": 1.4479533433914185,
"learning_rate": 4.06609986201977e-06,
"loss": 0.5736589431762695,
"step": 1180
},
{
"epoch": 2.4989429175475686,
"grad_norm": 1.7269279956817627,
"learning_rate": 4.051194625922681e-06,
"loss": 0.4028988480567932,
"step": 1182
},
{
"epoch": 2.503171247357294,
"grad_norm": 1.1051733493804932,
"learning_rate": 4.036307102053359e-06,
"loss": 0.14083461463451385,
"step": 1184
},
{
"epoch": 2.507399577167019,
"grad_norm": 1.5755764245986938,
"learning_rate": 4.021437472417763e-06,
"loss": 0.9020786285400391,
"step": 1186
},
{
"epoch": 2.511627906976744,
"grad_norm": 2.1714582443237305,
"learning_rate": 4.006585918803091e-06,
"loss": 0.6427817940711975,
"step": 1188
},
{
"epoch": 2.5158562367864694,
"grad_norm": 2.8261818885803223,
"learning_rate": 3.991752622775551e-06,
"loss": 0.6141193509101868,
"step": 1190
},
{
"epoch": 2.5200845665961946,
"grad_norm": 1.4536919593811035,
"learning_rate": 3.97693776567815e-06,
"loss": 0.940744161605835,
"step": 1192
},
{
"epoch": 2.52431289640592,
"grad_norm": 0.6860393285751343,
"learning_rate": 3.962141528628465e-06,
"loss": 0.5753416419029236,
"step": 1194
},
{
"epoch": 2.528541226215645,
"grad_norm": 4.277496814727783,
"learning_rate": 3.947364092516437e-06,
"loss": 0.9056035280227661,
"step": 1196
},
{
"epoch": 2.53276955602537,
"grad_norm": 1.433030366897583,
"learning_rate": 3.932605638002163e-06,
"loss": 0.8586236834526062,
"step": 1198
},
{
"epoch": 2.536997885835095,
"grad_norm": 3.587246894836426,
"learning_rate": 3.917866345513677e-06,
"loss": 0.6541805267333984,
"step": 1200
},
{
"epoch": 2.54122621564482,
"grad_norm": 0.9127486348152161,
"learning_rate": 3.903146395244749e-06,
"loss": 0.8223477005958557,
"step": 1202
},
{
"epoch": 2.5454545454545454,
"grad_norm": 2.6418988704681396,
"learning_rate": 3.888445967152691e-06,
"loss": 0.33483150601387024,
"step": 1204
},
{
"epoch": 2.5496828752642706,
"grad_norm": 1.132583737373352,
"learning_rate": 3.873765240956138e-06,
"loss": 0.9123491048812866,
"step": 1206
},
{
"epoch": 2.553911205073996,
"grad_norm": 0.997033417224884,
"learning_rate": 3.859104396132863e-06,
"loss": 0.9119923710823059,
"step": 1208
},
{
"epoch": 2.558139534883721,
"grad_norm": 0.8720810413360596,
"learning_rate": 3.84446361191759e-06,
"loss": 0.916553258895874,
"step": 1210
},
{
"epoch": 2.5623678646934462,
"grad_norm": 2.4558215141296387,
"learning_rate": 3.829843067299788e-06,
"loss": 0.6590366363525391,
"step": 1212
},
{
"epoch": 2.5665961945031714,
"grad_norm": 0.957917332649231,
"learning_rate": 3.8152429410214834e-06,
"loss": 0.8373352289199829,
"step": 1214
},
{
"epoch": 2.570824524312896,
"grad_norm": 3.4902255535125732,
"learning_rate": 3.8006634115750886e-06,
"loss": 0.5446149706840515,
"step": 1216
},
{
"epoch": 2.5750528541226214,
"grad_norm": 2.0583739280700684,
"learning_rate": 3.7861046572012105e-06,
"loss": 0.5926297307014465,
"step": 1218
},
{
"epoch": 2.5792811839323466,
"grad_norm": 3.7689881324768066,
"learning_rate": 3.771566855886467e-06,
"loss": 0.48180636763572693,
"step": 1220
},
{
"epoch": 2.583509513742072,
"grad_norm": 1.6291465759277344,
"learning_rate": 3.7570501853613227e-06,
"loss": 0.5041254758834839,
"step": 1222
},
{
"epoch": 2.587737843551797,
"grad_norm": 5.261500835418701,
"learning_rate": 3.742554823097905e-06,
"loss": 0.687404215335846,
"step": 1224
},
{
"epoch": 2.5919661733615222,
"grad_norm": 0.5467351675033569,
"learning_rate": 3.7280809463078436e-06,
"loss": 0.35284334421157837,
"step": 1226
},
{
"epoch": 2.5961945031712474,
"grad_norm": 1.3150749206542969,
"learning_rate": 3.7136287319400944e-06,
"loss": 0.669701099395752,
"step": 1228
},
{
"epoch": 2.6004228329809727,
"grad_norm": 2.019007682800293,
"learning_rate": 3.6991983566787847e-06,
"loss": 0.7513763308525085,
"step": 1230
},
{
"epoch": 2.604651162790698,
"grad_norm": 4.118278980255127,
"learning_rate": 3.6847899969410515e-06,
"loss": 0.6786062717437744,
"step": 1232
},
{
"epoch": 2.6088794926004226,
"grad_norm": 1.8718292713165283,
"learning_rate": 3.6704038288748812e-06,
"loss": 0.6505091190338135,
"step": 1234
},
{
"epoch": 2.613107822410148,
"grad_norm": 1.462855339050293,
"learning_rate": 3.6560400283569543e-06,
"loss": 0.7407243251800537,
"step": 1236
},
{
"epoch": 2.617336152219873,
"grad_norm": 1.429216980934143,
"learning_rate": 3.641698770990507e-06,
"loss": 0.818171501159668,
"step": 1238
},
{
"epoch": 2.6215644820295982,
"grad_norm": 0.9993104934692383,
"learning_rate": 3.6273802321031738e-06,
"loss": 0.6348567605018616,
"step": 1240
},
{
"epoch": 2.6257928118393234,
"grad_norm": 0.9296184778213501,
"learning_rate": 3.6130845867448384e-06,
"loss": 0.7757455706596375,
"step": 1242
},
{
"epoch": 2.6300211416490487,
"grad_norm": 0.8016238212585449,
"learning_rate": 3.598812009685514e-06,
"loss": 0.8527669906616211,
"step": 1244
},
{
"epoch": 2.634249471458774,
"grad_norm": 2.526475429534912,
"learning_rate": 3.584562675413189e-06,
"loss": 0.6010987758636475,
"step": 1246
},
{
"epoch": 2.638477801268499,
"grad_norm": 4.499967575073242,
"learning_rate": 3.5703367581316985e-06,
"loss": 0.45468515157699585,
"step": 1248
},
{
"epoch": 2.6427061310782243,
"grad_norm": 2.5104167461395264,
"learning_rate": 3.5561344317585957e-06,
"loss": 0.9026956558227539,
"step": 1250
},
{
"epoch": 2.646934460887949,
"grad_norm": 2.173785924911499,
"learning_rate": 3.5419558699230305e-06,
"loss": 0.4719690680503845,
"step": 1252
},
{
"epoch": 2.6511627906976747,
"grad_norm": 7.44932222366333,
"learning_rate": 3.52780124596361e-06,
"loss": 0.5562307238578796,
"step": 1254
},
{
"epoch": 2.6553911205073994,
"grad_norm": 0.7967355847358704,
"learning_rate": 3.5136707329263055e-06,
"loss": 0.8979150652885437,
"step": 1256
},
{
"epoch": 2.6596194503171247,
"grad_norm": 1.8158342838287354,
"learning_rate": 3.499564503562315e-06,
"loss": 0.2914978265762329,
"step": 1258
},
{
"epoch": 2.66384778012685,
"grad_norm": 0.9016306400299072,
"learning_rate": 3.4854827303259575e-06,
"loss": 0.47832489013671875,
"step": 1260
},
{
"epoch": 2.668076109936575,
"grad_norm": 2.051372528076172,
"learning_rate": 3.47142558537257e-06,
"loss": 0.5821577310562134,
"step": 1262
},
{
"epoch": 2.6723044397463003,
"grad_norm": 2.8656489849090576,
"learning_rate": 3.4573932405563966e-06,
"loss": 0.5456935167312622,
"step": 1264
},
{
"epoch": 2.6765327695560255,
"grad_norm": 2.2788443565368652,
"learning_rate": 3.4433858674284894e-06,
"loss": 0.39581629633903503,
"step": 1266
},
{
"epoch": 2.6807610993657507,
"grad_norm": 1.2009937763214111,
"learning_rate": 3.429403637234614e-06,
"loss": 0.5747796297073364,
"step": 1268
},
{
"epoch": 2.6849894291754755,
"grad_norm": 1.1330604553222656,
"learning_rate": 3.4154467209131515e-06,
"loss": 0.8658438920974731,
"step": 1270
},
{
"epoch": 2.689217758985201,
"grad_norm": 4.2896504402160645,
"learning_rate": 3.401515289093011e-06,
"loss": 0.6409957408905029,
"step": 1272
},
{
"epoch": 2.693446088794926,
"grad_norm": 0.3948211073875427,
"learning_rate": 3.3876095120915433e-06,
"loss": 0.3648834526538849,
"step": 1274
},
{
"epoch": 2.697674418604651,
"grad_norm": 2.79600191116333,
"learning_rate": 3.3737295599124592e-06,
"loss": 0.3996630311012268,
"step": 1276
},
{
"epoch": 2.7019027484143763,
"grad_norm": 1.3608835935592651,
"learning_rate": 3.3598756022437507e-06,
"loss": 0.8879305124282837,
"step": 1278
},
{
"epoch": 2.7061310782241015,
"grad_norm": 2.3612143993377686,
"learning_rate": 3.3460478084556157e-06,
"loss": 0.5910604000091553,
"step": 1280
},
{
"epoch": 2.7103594080338267,
"grad_norm": 1.4367172718048096,
"learning_rate": 3.33224634759839e-06,
"loss": 0.9345384836196899,
"step": 1282
},
{
"epoch": 2.714587737843552,
"grad_norm": 1.6179012060165405,
"learning_rate": 3.318471388400476e-06,
"loss": 0.5907537341117859,
"step": 1284
},
{
"epoch": 2.718816067653277,
"grad_norm": 0.27789580821990967,
"learning_rate": 3.3047230992662834e-06,
"loss": 0.9225029349327087,
"step": 1286
},
{
"epoch": 2.723044397463002,
"grad_norm": 0.762175440788269,
"learning_rate": 3.291001648274175e-06,
"loss": 0.45095983147621155,
"step": 1288
},
{
"epoch": 2.7272727272727275,
"grad_norm": 0.9289544820785522,
"learning_rate": 3.277307203174395e-06,
"loss": 0.8066924810409546,
"step": 1290
},
{
"epoch": 2.7315010570824523,
"grad_norm": 0.8659968972206116,
"learning_rate": 3.263639931387039e-06,
"loss": 0.9190667271614075,
"step": 1292
},
{
"epoch": 2.7357293868921775,
"grad_norm": 1.809422492980957,
"learning_rate": 3.2500000000000015e-06,
"loss": 0.9080826044082642,
"step": 1294
},
{
"epoch": 2.7399577167019027,
"grad_norm": 0.6880254149436951,
"learning_rate": 3.236387575766916e-06,
"loss": 0.3050709068775177,
"step": 1296
},
{
"epoch": 2.744186046511628,
"grad_norm": 0.9908064007759094,
"learning_rate": 3.222802825105138e-06,
"loss": 0.5199543833732605,
"step": 1298
},
{
"epoch": 2.748414376321353,
"grad_norm": 1.6468696594238281,
"learning_rate": 3.209245914093708e-06,
"loss": 0.6294687986373901,
"step": 1300
},
{
"epoch": 2.7526427061310783,
"grad_norm": 1.6277974843978882,
"learning_rate": 3.1957170084713053e-06,
"loss": 0.7647433280944824,
"step": 1302
},
{
"epoch": 2.7568710359408035,
"grad_norm": 0.6971714496612549,
"learning_rate": 3.182216273634235e-06,
"loss": 0.5780515074729919,
"step": 1304
},
{
"epoch": 2.7610993657505283,
"grad_norm": 4.069775581359863,
"learning_rate": 3.1687438746344086e-06,
"loss": 0.7356036305427551,
"step": 1306
},
{
"epoch": 2.765327695560254,
"grad_norm": 2.1779944896698,
"learning_rate": 3.1552999761773166e-06,
"loss": 0.8199219703674316,
"step": 1308
},
{
"epoch": 2.7695560253699787,
"grad_norm": 1.0269923210144043,
"learning_rate": 3.141884742620013e-06,
"loss": 0.8712670803070068,
"step": 1310
},
{
"epoch": 2.773784355179704,
"grad_norm": 1.4878244400024414,
"learning_rate": 3.1284983379691247e-06,
"loss": 0.7457907795906067,
"step": 1312
},
{
"epoch": 2.778012684989429,
"grad_norm": 0.7093844413757324,
"learning_rate": 3.1151409258788235e-06,
"loss": 0.45220234990119934,
"step": 1314
},
{
"epoch": 2.7822410147991543,
"grad_norm": 2.556499719619751,
"learning_rate": 3.101812669648838e-06,
"loss": 0.7896378040313721,
"step": 1316
},
{
"epoch": 2.7864693446088795,
"grad_norm": 0.594222366809845,
"learning_rate": 3.0885137322224546e-06,
"loss": 0.5855066776275635,
"step": 1318
},
{
"epoch": 2.7906976744186047,
"grad_norm": 2.469794988632202,
"learning_rate": 3.075244276184529e-06,
"loss": 0.8110044002532959,
"step": 1320
},
{
"epoch": 2.79492600422833,
"grad_norm": 4.330004692077637,
"learning_rate": 3.0620044637594936e-06,
"loss": 0.3200315237045288,
"step": 1322
},
{
"epoch": 2.7991543340380547,
"grad_norm": 0.8099966645240784,
"learning_rate": 3.0487944568093685e-06,
"loss": 1.002435564994812,
"step": 1324
},
{
"epoch": 2.8033826638477803,
"grad_norm": 2.4305269718170166,
"learning_rate": 3.035614416831801e-06,
"loss": 0.25673067569732666,
"step": 1326
},
{
"epoch": 2.807610993657505,
"grad_norm": 2.854151964187622,
"learning_rate": 3.0224645049580746e-06,
"loss": 0.41537269949913025,
"step": 1328
},
{
"epoch": 2.8118393234672303,
"grad_norm": 0.9382374286651611,
"learning_rate": 3.0093448819511416e-06,
"loss": 0.566652238368988,
"step": 1330
},
{
"epoch": 2.8160676532769555,
"grad_norm": 1.395978569984436,
"learning_rate": 2.9962557082036677e-06,
"loss": 0.9220659136772156,
"step": 1332
},
{
"epoch": 2.8202959830866807,
"grad_norm": 1.0719799995422363,
"learning_rate": 2.9831971437360606e-06,
"loss": 0.6312952041625977,
"step": 1334
},
{
"epoch": 2.824524312896406,
"grad_norm": 20.29490089416504,
"learning_rate": 2.9701693481945132e-06,
"loss": 0.9521694779396057,
"step": 1336
},
{
"epoch": 2.828752642706131,
"grad_norm": 0.7605103850364685,
"learning_rate": 2.957172480849064e-06,
"loss": 0.8885098695755005,
"step": 1338
},
{
"epoch": 2.8329809725158563,
"grad_norm": 2.091263771057129,
"learning_rate": 2.9442067005916375e-06,
"loss": 0.6456158757209778,
"step": 1340
},
{
"epoch": 2.8372093023255816,
"grad_norm": 0.8573316931724548,
"learning_rate": 2.931272165934105e-06,
"loss": 0.8637920618057251,
"step": 1342
},
{
"epoch": 2.8414376321353068,
"grad_norm": 1.4005157947540283,
"learning_rate": 2.918369035006351e-06,
"loss": 0.6166179180145264,
"step": 1344
},
{
"epoch": 2.8456659619450315,
"grad_norm": 1.8555761575698853,
"learning_rate": 2.9054974655543335e-06,
"loss": 0.6378682851791382,
"step": 1346
},
{
"epoch": 2.8498942917547567,
"grad_norm": 0.9346190094947815,
"learning_rate": 2.89265761493816e-06,
"loss": 0.791275143623352,
"step": 1348
},
{
"epoch": 2.854122621564482,
"grad_norm": 3.476630687713623,
"learning_rate": 2.8798496401301644e-06,
"loss": 0.47718340158462524,
"step": 1350
},
{
"epoch": 2.858350951374207,
"grad_norm": 11.810805320739746,
"learning_rate": 2.8670736977129843e-06,
"loss": 0.8192956447601318,
"step": 1352
},
{
"epoch": 2.8625792811839323,
"grad_norm": 1.228657841682434,
"learning_rate": 2.854329943877647e-06,
"loss": 0.9457657337188721,
"step": 1354
},
{
"epoch": 2.8668076109936576,
"grad_norm": 4.220162868499756,
"learning_rate": 2.841618534421664e-06,
"loss": 0.396058052778244,
"step": 1356
},
{
"epoch": 2.8710359408033828,
"grad_norm": 30.189821243286133,
"learning_rate": 2.8289396247471227e-06,
"loss": 0.8925053477287292,
"step": 1358
},
{
"epoch": 2.875264270613108,
"grad_norm": 1.379380464553833,
"learning_rate": 2.816293369858788e-06,
"loss": 0.5891386270523071,
"step": 1360
},
{
"epoch": 2.879492600422833,
"grad_norm": 1.7700012922286987,
"learning_rate": 2.803679924362206e-06,
"loss": 0.6942218542098999,
"step": 1362
},
{
"epoch": 2.883720930232558,
"grad_norm": 1.210906982421875,
"learning_rate": 2.7910994424618165e-06,
"loss": 0.9861003160476685,
"step": 1364
},
{
"epoch": 2.887949260042283,
"grad_norm": 4.873892784118652,
"learning_rate": 2.778552077959064e-06,
"loss": 0.707911491394043,
"step": 1366
},
{
"epoch": 2.8921775898520083,
"grad_norm": 0.9191520810127258,
"learning_rate": 2.766037984250521e-06,
"loss": 0.8886877298355103,
"step": 1368
},
{
"epoch": 2.8964059196617336,
"grad_norm": 3.467663288116455,
"learning_rate": 2.753557314326011e-06,
"loss": 0.8607564568519592,
"step": 1370
},
{
"epoch": 2.9006342494714588,
"grad_norm": 0.999760091304779,
"learning_rate": 2.7411102207667373e-06,
"loss": 0.8891087770462036,
"step": 1372
},
{
"epoch": 2.904862579281184,
"grad_norm": 1.2519502639770508,
"learning_rate": 2.728696855743418e-06,
"loss": 0.8682777881622314,
"step": 1374
},
{
"epoch": 2.909090909090909,
"grad_norm": 1.235428810119629,
"learning_rate": 2.7163173710144323e-06,
"loss": 0.8566583395004272,
"step": 1376
},
{
"epoch": 2.9133192389006344,
"grad_norm": 2.1394336223602295,
"learning_rate": 2.703971917923951e-06,
"loss": 0.6189751029014587,
"step": 1378
},
{
"epoch": 2.9175475687103596,
"grad_norm": 1.1033439636230469,
"learning_rate": 2.6916606474000965e-06,
"loss": 0.6509535312652588,
"step": 1380
},
{
"epoch": 2.9217758985200843,
"grad_norm": 3.1471967697143555,
"learning_rate": 2.679383709953103e-06,
"loss": 1.0116616487503052,
"step": 1382
},
{
"epoch": 2.92600422832981,
"grad_norm": 4.125372409820557,
"learning_rate": 2.6671412556734578e-06,
"loss": 0.8084197044372559,
"step": 1384
},
{
"epoch": 2.9302325581395348,
"grad_norm": 1.1657354831695557,
"learning_rate": 2.6549334342300844e-06,
"loss": 0.6304535269737244,
"step": 1386
},
{
"epoch": 2.93446088794926,
"grad_norm": 0.9015957117080688,
"learning_rate": 2.642760394868503e-06,
"loss": 0.8338552713394165,
"step": 1388
},
{
"epoch": 2.938689217758985,
"grad_norm": 0.33686238527297974,
"learning_rate": 2.6306222864090157e-06,
"loss": 0.6320370435714722,
"step": 1390
},
{
"epoch": 2.9429175475687104,
"grad_norm": 1.191013216972351,
"learning_rate": 2.61851925724487e-06,
"loss": 0.8449854254722595,
"step": 1392
},
{
"epoch": 2.9471458773784356,
"grad_norm": 0.37625300884246826,
"learning_rate": 2.606451455340459e-06,
"loss": 0.2572139799594879,
"step": 1394
},
{
"epoch": 2.951374207188161,
"grad_norm": 0.8979637026786804,
"learning_rate": 2.5944190282295155e-06,
"loss": 0.5207499265670776,
"step": 1396
},
{
"epoch": 2.955602536997886,
"grad_norm": 6.482511520385742,
"learning_rate": 2.5824221230132895e-06,
"loss": 0.8485027551651001,
"step": 1398
},
{
"epoch": 2.9598308668076108,
"grad_norm": 0.8348364233970642,
"learning_rate": 2.5704608863587676e-06,
"loss": 1.0486241579055786,
"step": 1400
},
{
"epoch": 2.9640591966173364,
"grad_norm": 2.5745809078216553,
"learning_rate": 2.558535464496877e-06,
"loss": 0.738484263420105,
"step": 1402
},
{
"epoch": 2.968287526427061,
"grad_norm": 3.472229480743408,
"learning_rate": 2.5466460032206872e-06,
"loss": 0.4392167627811432,
"step": 1404
},
{
"epoch": 2.9725158562367864,
"grad_norm": 4.279146194458008,
"learning_rate": 2.53479264788364e-06,
"loss": 0.8173143267631531,
"step": 1406
},
{
"epoch": 2.9767441860465116,
"grad_norm": 0.9447436332702637,
"learning_rate": 2.522975543397769e-06,
"loss": 0.6879301071166992,
"step": 1408
},
{
"epoch": 2.980972515856237,
"grad_norm": 5.459957122802734,
"learning_rate": 2.511194834231922e-06,
"loss": 0.6252565979957581,
"step": 1410
},
{
"epoch": 2.985200845665962,
"grad_norm": 1.1335101127624512,
"learning_rate": 2.49945066441e-06,
"loss": 0.8397121429443359,
"step": 1412
},
{
"epoch": 2.989429175475687,
"grad_norm": 0.7335910797119141,
"learning_rate": 2.4877431775092004e-06,
"loss": 0.6167458891868591,
"step": 1414
},
{
"epoch": 2.9936575052854124,
"grad_norm": 4.312607288360596,
"learning_rate": 2.4760725166582544e-06,
"loss": 0.7645383477210999,
"step": 1416
},
{
"epoch": 2.997885835095137,
"grad_norm": 1.5329777002334595,
"learning_rate": 2.464438824535675e-06,
"loss": 0.4461861848831177,
"step": 1418
},
{
"epoch": 3.0021141649048624,
"grad_norm": 1.1518954038619995,
"learning_rate": 2.452842243368028e-06,
"loss": 0.8723726868629456,
"step": 1420
},
{
"epoch": 3.0063424947145876,
"grad_norm": 1.157313346862793,
"learning_rate": 2.4412829149281735e-06,
"loss": 0.42229244112968445,
"step": 1422
},
{
"epoch": 3.010570824524313,
"grad_norm": 2.366347074508667,
"learning_rate": 2.4297609805335472e-06,
"loss": 0.7791450023651123,
"step": 1424
},
{
"epoch": 3.014799154334038,
"grad_norm": 7.397970676422119,
"learning_rate": 2.4182765810444263e-06,
"loss": 0.34111690521240234,
"step": 1426
},
{
"epoch": 3.019027484143763,
"grad_norm": 0.9462530612945557,
"learning_rate": 2.4068298568622083e-06,
"loss": 0.41975370049476624,
"step": 1428
},
{
"epoch": 3.0232558139534884,
"grad_norm": 1.4463883638381958,
"learning_rate": 2.395420947927694e-06,
"loss": 0.6927764415740967,
"step": 1430
},
{
"epoch": 3.0274841437632136,
"grad_norm": 1.7992171049118042,
"learning_rate": 2.3840499937193804e-06,
"loss": 0.7748026847839355,
"step": 1432
},
{
"epoch": 3.031712473572939,
"grad_norm": 1.013612985610962,
"learning_rate": 2.3727171332517496e-06,
"loss": 0.6729182004928589,
"step": 1434
},
{
"epoch": 3.035940803382664,
"grad_norm": 8.133994102478027,
"learning_rate": 2.361422505073575e-06,
"loss": 0.5017516016960144,
"step": 1436
},
{
"epoch": 3.040169133192389,
"grad_norm": 0.613357424736023,
"learning_rate": 2.3501662472662233e-06,
"loss": 0.1772938072681427,
"step": 1438
},
{
"epoch": 3.044397463002114,
"grad_norm": 1.2707237005233765,
"learning_rate": 2.3389484974419685e-06,
"loss": 0.7271437048912048,
"step": 1440
},
{
"epoch": 3.048625792811839,
"grad_norm": 1.034716248512268,
"learning_rate": 2.327769392742309e-06,
"loss": 0.4943666458129883,
"step": 1442
},
{
"epoch": 3.0528541226215644,
"grad_norm": 2.7102303504943848,
"learning_rate": 2.316629069836291e-06,
"loss": 0.14087991416454315,
"step": 1444
},
{
"epoch": 3.0570824524312896,
"grad_norm": 2.461373805999756,
"learning_rate": 2.305527664918838e-06,
"loss": 0.4643915295600891,
"step": 1446
},
{
"epoch": 3.061310782241015,
"grad_norm": 5.142198085784912,
"learning_rate": 2.294465313709085e-06,
"loss": 0.6128051280975342,
"step": 1448
},
{
"epoch": 3.06553911205074,
"grad_norm": 1.7347453832626343,
"learning_rate": 2.2834421514487205e-06,
"loss": 0.5520272254943848,
"step": 1450
},
{
"epoch": 3.0697674418604652,
"grad_norm": 19.25787353515625,
"learning_rate": 2.27245831290033e-06,
"loss": 0.5760930776596069,
"step": 1452
},
{
"epoch": 3.0739957716701904,
"grad_norm": 1.2460963726043701,
"learning_rate": 2.261513932345755e-06,
"loss": 0.7124364972114563,
"step": 1454
},
{
"epoch": 3.0782241014799157,
"grad_norm": 0.6032560467720032,
"learning_rate": 2.250609143584443e-06,
"loss": 0.13673415780067444,
"step": 1456
},
{
"epoch": 3.0824524312896404,
"grad_norm": 0.8289851546287537,
"learning_rate": 2.239744079931818e-06,
"loss": 0.45457348227500916,
"step": 1458
},
{
"epoch": 3.0866807610993656,
"grad_norm": 1.3416787385940552,
"learning_rate": 2.228918874217649e-06,
"loss": 0.6314166188240051,
"step": 1460
},
{
"epoch": 3.090909090909091,
"grad_norm": 2.7620086669921875,
"learning_rate": 2.2181336587844244e-06,
"loss": 0.6687834858894348,
"step": 1462
},
{
"epoch": 3.095137420718816,
"grad_norm": 1.2758150100708008,
"learning_rate": 2.207388565485736e-06,
"loss": 0.5059049725532532,
"step": 1464
},
{
"epoch": 3.0993657505285412,
"grad_norm": 2.085491895675659,
"learning_rate": 2.196683725684666e-06,
"loss": 0.5610879063606262,
"step": 1466
},
{
"epoch": 3.1035940803382664,
"grad_norm": 0.99587482213974,
"learning_rate": 2.1860192702521826e-06,
"loss": 0.6807426810264587,
"step": 1468
},
{
"epoch": 3.1078224101479917,
"grad_norm": 1.2479846477508545,
"learning_rate": 2.1753953295655354e-06,
"loss": 0.7462598085403442,
"step": 1470
},
{
"epoch": 3.112050739957717,
"grad_norm": 4.688530445098877,
"learning_rate": 2.164812033506673e-06,
"loss": 0.4493924081325531,
"step": 1472
},
{
"epoch": 3.116279069767442,
"grad_norm": 1.14339280128479,
"learning_rate": 2.1542695114606367e-06,
"loss": 0.7715897560119629,
"step": 1474
},
{
"epoch": 3.120507399577167,
"grad_norm": 0.9763163924217224,
"learning_rate": 2.1437678923139944e-06,
"loss": 0.6441456079483032,
"step": 1476
},
{
"epoch": 3.124735729386892,
"grad_norm": 4.06795597076416,
"learning_rate": 2.1333073044532628e-06,
"loss": 0.27343037724494934,
"step": 1478
},
{
"epoch": 3.1289640591966172,
"grad_norm": 1.8711929321289062,
"learning_rate": 2.122887875763327e-06,
"loss": 0.6824556589126587,
"step": 1480
},
{
"epoch": 3.1331923890063424,
"grad_norm": 0.8963372707366943,
"learning_rate": 2.1125097336258897e-06,
"loss": 0.7033651471138,
"step": 1482
},
{
"epoch": 3.1374207188160677,
"grad_norm": 1.0724904537200928,
"learning_rate": 2.1021730049179117e-06,
"loss": 0.22907228767871857,
"step": 1484
},
{
"epoch": 3.141649048625793,
"grad_norm": 4.497315883636475,
"learning_rate": 2.09187781601005e-06,
"loss": 0.8114728927612305,
"step": 1486
},
{
"epoch": 3.145877378435518,
"grad_norm": 1.3675453662872314,
"learning_rate": 2.0816242927651254e-06,
"loss": 0.8725941777229309,
"step": 1488
},
{
"epoch": 3.1501057082452433,
"grad_norm": 2.9197142124176025,
"learning_rate": 2.071412560536581e-06,
"loss": 0.5963315963745117,
"step": 1490
},
{
"epoch": 3.1543340380549685,
"grad_norm": 2.065478563308716,
"learning_rate": 2.0612427441669475e-06,
"loss": 0.5692322850227356,
"step": 1492
},
{
"epoch": 3.1585623678646932,
"grad_norm": 0.8671579957008362,
"learning_rate": 2.051114967986311e-06,
"loss": 0.33531826734542847,
"step": 1494
},
{
"epoch": 3.1627906976744184,
"grad_norm": 0.18376043438911438,
"learning_rate": 2.0410293558108074e-06,
"loss": 0.5413160920143127,
"step": 1496
},
{
"epoch": 3.1670190274841437,
"grad_norm": 1.024639368057251,
"learning_rate": 2.0309860309410986e-06,
"loss": 0.7540732622146606,
"step": 1498
},
{
"epoch": 3.171247357293869,
"grad_norm": 1.1238852739334106,
"learning_rate": 2.0209851161608613e-06,
"loss": 0.7976906299591064,
"step": 1500
},
{
"epoch": 3.175475687103594,
"grad_norm": 4.7006635665893555,
"learning_rate": 2.011026733735301e-06,
"loss": 0.26561132073402405,
"step": 1502
},
{
"epoch": 3.1797040169133193,
"grad_norm": 0.7230396270751953,
"learning_rate": 2.001111005409643e-06,
"loss": 0.42223721742630005,
"step": 1504
},
{
"epoch": 3.1839323467230445,
"grad_norm": 1.1168904304504395,
"learning_rate": 1.9912380524076503e-06,
"loss": 0.5349246263504028,
"step": 1506
},
{
"epoch": 3.1881606765327697,
"grad_norm": 1.9625024795532227,
"learning_rate": 1.981407995430136e-06,
"loss": 0.27550962567329407,
"step": 1508
},
{
"epoch": 3.192389006342495,
"grad_norm": 8.77584171295166,
"learning_rate": 1.9716209546535e-06,
"loss": 0.25047436356544495,
"step": 1510
},
{
"epoch": 3.1966173361522197,
"grad_norm": 0.20387931168079376,
"learning_rate": 1.9618770497282487e-06,
"loss": 0.04526889696717262,
"step": 1512
},
{
"epoch": 3.200845665961945,
"grad_norm": 5.831273078918457,
"learning_rate": 1.9521763997775294e-06,
"loss": 0.6410486102104187,
"step": 1514
},
{
"epoch": 3.20507399577167,
"grad_norm": 1.461323857307434,
"learning_rate": 1.942519123395689e-06,
"loss": 0.7742519974708557,
"step": 1516
},
{
"epoch": 3.2093023255813953,
"grad_norm": 1.1463195085525513,
"learning_rate": 1.932905338646812e-06,
"loss": 0.4687974452972412,
"step": 1518
},
{
"epoch": 3.2135306553911205,
"grad_norm": 0.8970885872840881,
"learning_rate": 1.9233351630632747e-06,
"loss": 0.7083006501197815,
"step": 1520
},
{
"epoch": 3.2177589852008457,
"grad_norm": 0.9952190518379211,
"learning_rate": 1.9138087136443227e-06,
"loss": 0.7196356654167175,
"step": 1522
},
{
"epoch": 3.221987315010571,
"grad_norm": 2.105109453201294,
"learning_rate": 1.904326106854626e-06,
"loss": 0.725276529788971,
"step": 1524
},
{
"epoch": 3.226215644820296,
"grad_norm": 1.6559922695159912,
"learning_rate": 1.8948874586228622e-06,
"loss": 0.6343616247177124,
"step": 1526
},
{
"epoch": 3.2304439746300213,
"grad_norm": 1.6503781080245972,
"learning_rate": 1.8854928843402987e-06,
"loss": 0.09520567953586578,
"step": 1528
},
{
"epoch": 3.234672304439746,
"grad_norm": 3.222755193710327,
"learning_rate": 1.8761424988593796e-06,
"loss": 0.723050594329834,
"step": 1530
},
{
"epoch": 3.2389006342494713,
"grad_norm": 3.8416454792022705,
"learning_rate": 1.8668364164923248e-06,
"loss": 0.6272533535957336,
"step": 1532
},
{
"epoch": 3.2431289640591965,
"grad_norm": 1.00477135181427,
"learning_rate": 1.8575747510097294e-06,
"loss": 0.6595237255096436,
"step": 1534
},
{
"epoch": 3.2473572938689217,
"grad_norm": 1.9450788497924805,
"learning_rate": 1.8483576156391747e-06,
"loss": 0.47564584016799927,
"step": 1536
},
{
"epoch": 3.251585623678647,
"grad_norm": 1.9304862022399902,
"learning_rate": 1.8391851230638443e-06,
"loss": 0.4504464566707611,
"step": 1538
},
{
"epoch": 3.255813953488372,
"grad_norm": 1.3938456773757935,
"learning_rate": 1.8300573854211452e-06,
"loss": 0.4539117217063904,
"step": 1540
},
{
"epoch": 3.2600422832980973,
"grad_norm": 5.073108196258545,
"learning_rate": 1.8209745143013363e-06,
"loss": 0.9542956948280334,
"step": 1542
},
{
"epoch": 3.2642706131078225,
"grad_norm": 0.8300318717956543,
"learning_rate": 1.811936620746167e-06,
"loss": 0.7863713502883911,
"step": 1544
},
{
"epoch": 3.2684989429175477,
"grad_norm": 12.513225555419922,
"learning_rate": 1.8029438152475159e-06,
"loss": 0.7303227186203003,
"step": 1546
},
{
"epoch": 3.2727272727272725,
"grad_norm": 3.76115345954895,
"learning_rate": 1.7939962077460448e-06,
"loss": 0.1273539662361145,
"step": 1548
},
{
"epoch": 3.276955602536998,
"grad_norm": 7.385823726654053,
"learning_rate": 1.785093907629849e-06,
"loss": 0.6256406307220459,
"step": 1550
},
{
"epoch": 3.281183932346723,
"grad_norm": 1.8356283903121948,
"learning_rate": 1.7762370237331233e-06,
"loss": 0.7494076490402222,
"step": 1552
},
{
"epoch": 3.285412262156448,
"grad_norm": 0.15552738308906555,
"learning_rate": 1.7674256643348348e-06,
"loss": 0.0979471504688263,
"step": 1554
},
{
"epoch": 3.2896405919661733,
"grad_norm": 1.1801795959472656,
"learning_rate": 1.7586599371573887e-06,
"loss": 0.7723276615142822,
"step": 1556
},
{
"epoch": 3.2938689217758985,
"grad_norm": 1.9973111152648926,
"learning_rate": 1.7499399493653225e-06,
"loss": 0.46918126940727234,
"step": 1558
},
{
"epoch": 3.2980972515856237,
"grad_norm": 2.415146827697754,
"learning_rate": 1.7412658075639921e-06,
"loss": 0.9041464328765869,
"step": 1560
},
{
"epoch": 3.302325581395349,
"grad_norm": 5.58949613571167,
"learning_rate": 1.7326376177982635e-06,
"loss": 0.7202844619750977,
"step": 1562
},
{
"epoch": 3.306553911205074,
"grad_norm": 1.0739604234695435,
"learning_rate": 1.7240554855512225e-06,
"loss": 0.4417579770088196,
"step": 1564
},
{
"epoch": 3.3107822410147993,
"grad_norm": 1.1985353231430054,
"learning_rate": 1.7155195157428878e-06,
"loss": 0.6076565384864807,
"step": 1566
},
{
"epoch": 3.3150105708245245,
"grad_norm": 0.2641305923461914,
"learning_rate": 1.707029812728916e-06,
"loss": 0.09184014797210693,
"step": 1568
},
{
"epoch": 3.3192389006342493,
"grad_norm": 1.5718731880187988,
"learning_rate": 1.6985864802993366e-06,
"loss": 0.4400280714035034,
"step": 1570
},
{
"epoch": 3.3234672304439745,
"grad_norm": 4.702225208282471,
"learning_rate": 1.690189621677284e-06,
"loss": 0.5205068588256836,
"step": 1572
},
{
"epoch": 3.3276955602536997,
"grad_norm": 3.0285556316375732,
"learning_rate": 1.681839339517728e-06,
"loss": 0.18540425598621368,
"step": 1574
},
{
"epoch": 3.331923890063425,
"grad_norm": 3.3052446842193604,
"learning_rate": 1.67353573590622e-06,
"loss": 0.3356280028820038,
"step": 1576
},
{
"epoch": 3.33615221987315,
"grad_norm": 1.888217568397522,
"learning_rate": 1.6652789123576513e-06,
"loss": 0.4342425763607025,
"step": 1578
},
{
"epoch": 3.3403805496828753,
"grad_norm": 0.8568776845932007,
"learning_rate": 1.6570689698150094e-06,
"loss": 0.4174615144729614,
"step": 1580
},
{
"epoch": 3.3446088794926006,
"grad_norm": 1.127447485923767,
"learning_rate": 1.6489060086481373e-06,
"loss": 0.8088030815124512,
"step": 1582
},
{
"epoch": 3.3488372093023258,
"grad_norm": 4.583712100982666,
"learning_rate": 1.6407901286525152e-06,
"loss": 0.5413541793823242,
"step": 1584
},
{
"epoch": 3.353065539112051,
"grad_norm": 2.868380308151245,
"learning_rate": 1.6327214290480382e-06,
"loss": 0.538662850856781,
"step": 1586
},
{
"epoch": 3.3572938689217757,
"grad_norm": 3.1145548820495605,
"learning_rate": 1.6247000084778006e-06,
"loss": 0.8664639592170715,
"step": 1588
},
{
"epoch": 3.361522198731501,
"grad_norm": 1.4486368894577026,
"learning_rate": 1.6167259650068887e-06,
"loss": 0.267899751663208,
"step": 1590
},
{
"epoch": 3.365750528541226,
"grad_norm": 0.9725316762924194,
"learning_rate": 1.6087993961211904e-06,
"loss": 0.7432690858840942,
"step": 1592
},
{
"epoch": 3.3699788583509513,
"grad_norm": 0.9898474216461182,
"learning_rate": 1.6009203987261957e-06,
"loss": 0.20856915414333344,
"step": 1594
},
{
"epoch": 3.3742071881606766,
"grad_norm": 0.9847052693367004,
"learning_rate": 1.593089069145809e-06,
"loss": 0.7779243588447571,
"step": 1596
},
{
"epoch": 3.3784355179704018,
"grad_norm": 6.319812774658203,
"learning_rate": 1.5853055031211844e-06,
"loss": 0.548964262008667,
"step": 1598
},
{
"epoch": 3.382663847780127,
"grad_norm": 0.6208621859550476,
"learning_rate": 1.577569795809544e-06,
"loss": 0.2240552306175232,
"step": 1600
},
{
"epoch": 3.386892177589852,
"grad_norm": 3.8312599658966064,
"learning_rate": 1.5698820417830137e-06,
"loss": 0.4887566864490509,
"step": 1602
},
{
"epoch": 3.3911205073995774,
"grad_norm": 1.8255524635314941,
"learning_rate": 1.5622423350274784e-06,
"loss": 0.7597917318344116,
"step": 1604
},
{
"epoch": 3.395348837209302,
"grad_norm": 1.0387070178985596,
"learning_rate": 1.5546507689414217e-06,
"loss": 0.7400175929069519,
"step": 1606
},
{
"epoch": 3.3995771670190273,
"grad_norm": 2.640453338623047,
"learning_rate": 1.5471074363347894e-06,
"loss": 0.5016782879829407,
"step": 1608
},
{
"epoch": 3.4038054968287526,
"grad_norm": 1.0323295593261719,
"learning_rate": 1.539612429427852e-06,
"loss": 0.7418125867843628,
"step": 1610
},
{
"epoch": 3.4080338266384778,
"grad_norm": 3.3146867752075195,
"learning_rate": 1.5321658398500808e-06,
"loss": 0.763083815574646,
"step": 1612
},
{
"epoch": 3.412262156448203,
"grad_norm": 1.306364893913269,
"learning_rate": 1.5247677586390255e-06,
"loss": 0.6090471744537354,
"step": 1614
},
{
"epoch": 3.416490486257928,
"grad_norm": 1.330155372619629,
"learning_rate": 1.5174182762392007e-06,
"loss": 0.6685143113136292,
"step": 1616
},
{
"epoch": 3.4207188160676534,
"grad_norm": 1.0338711738586426,
"learning_rate": 1.5101174825009815e-06,
"loss": 0.9084112048149109,
"step": 1618
},
{
"epoch": 3.4249471458773786,
"grad_norm": 1.2957884073257446,
"learning_rate": 1.5028654666795042e-06,
"loss": 0.6523966193199158,
"step": 1620
},
{
"epoch": 3.429175475687104,
"grad_norm": 56.50241470336914,
"learning_rate": 1.495662317433576e-06,
"loss": 0.41208595037460327,
"step": 1622
},
{
"epoch": 3.4334038054968286,
"grad_norm": 0.9969905614852905,
"learning_rate": 1.4885081228245909e-06,
"loss": 0.7010380029678345,
"step": 1624
},
{
"epoch": 3.4376321353065538,
"grad_norm": 1.4458160400390625,
"learning_rate": 1.4814029703154515e-06,
"loss": 0.4886406362056732,
"step": 1626
},
{
"epoch": 3.441860465116279,
"grad_norm": 0.9273692965507507,
"learning_rate": 1.4743469467695017e-06,
"loss": 0.7973940372467041,
"step": 1628
},
{
"epoch": 3.446088794926004,
"grad_norm": 3.229093551635742,
"learning_rate": 1.4673401384494651e-06,
"loss": 0.46158260107040405,
"step": 1630
},
{
"epoch": 3.4503171247357294,
"grad_norm": 1.0890427827835083,
"learning_rate": 1.4603826310163874e-06,
"loss": 0.6124342679977417,
"step": 1632
},
{
"epoch": 3.4545454545454546,
"grad_norm": 1.319909930229187,
"learning_rate": 1.453474509528593e-06,
"loss": 0.4707542061805725,
"step": 1634
},
{
"epoch": 3.45877378435518,
"grad_norm": 0.9138315320014954,
"learning_rate": 1.446615858440642e-06,
"loss": 0.7025073766708374,
"step": 1636
},
{
"epoch": 3.463002114164905,
"grad_norm": 0.26452329754829407,
"learning_rate": 1.4398067616022993e-06,
"loss": 0.31818482279777527,
"step": 1638
},
{
"epoch": 3.46723044397463,
"grad_norm": 1.0385550260543823,
"learning_rate": 1.43304730225751e-06,
"loss": 0.43809548020362854,
"step": 1640
},
{
"epoch": 3.471458773784355,
"grad_norm": 0.5054123997688293,
"learning_rate": 1.4263375630433802e-06,
"loss": 0.12670187652111053,
"step": 1642
},
{
"epoch": 3.47568710359408,
"grad_norm": 1.5091001987457275,
"learning_rate": 1.4196776259891665e-06,
"loss": 0.792249858379364,
"step": 1644
},
{
"epoch": 3.4799154334038054,
"grad_norm": 1.8900986909866333,
"learning_rate": 1.4130675725152762e-06,
"loss": 0.633941113948822,
"step": 1646
},
{
"epoch": 3.4841437632135306,
"grad_norm": 0.13472701609134674,
"learning_rate": 1.4065074834322677e-06,
"loss": 0.4024646282196045,
"step": 1648
},
{
"epoch": 3.488372093023256,
"grad_norm": 1.4286351203918457,
"learning_rate": 1.3999974389398669e-06,
"loss": 0.38571226596832275,
"step": 1650
},
{
"epoch": 3.492600422832981,
"grad_norm": 3.8543434143066406,
"learning_rate": 1.3935375186259817e-06,
"loss": 0.6013924479484558,
"step": 1652
},
{
"epoch": 3.496828752642706,
"grad_norm": 2.653520345687866,
"learning_rate": 1.3871278014657332e-06,
"loss": 0.532644510269165,
"step": 1654
},
{
"epoch": 3.5010570824524314,
"grad_norm": 1.136316180229187,
"learning_rate": 1.3807683658204915e-06,
"loss": 0.6648343801498413,
"step": 1656
},
{
"epoch": 3.5052854122621566,
"grad_norm": 0.3273751437664032,
"learning_rate": 1.3744592894369093e-06,
"loss": 0.3973644971847534,
"step": 1658
},
{
"epoch": 3.5095137420718814,
"grad_norm": 0.8818485140800476,
"learning_rate": 1.3682006494459806e-06,
"loss": 0.5423086285591125,
"step": 1660
},
{
"epoch": 3.513742071881607,
"grad_norm": 2.183785915374756,
"learning_rate": 1.361992522362096e-06,
"loss": 0.5176860690116882,
"step": 1662
},
{
"epoch": 3.517970401691332,
"grad_norm": 1.6725293397903442,
"learning_rate": 1.355834984082101e-06,
"loss": 0.7665601968765259,
"step": 1664
},
{
"epoch": 3.522198731501057,
"grad_norm": 13.793525695800781,
"learning_rate": 1.3497281098843746e-06,
"loss": 0.7127768397331238,
"step": 1666
},
{
"epoch": 3.526427061310782,
"grad_norm": 0.5776848196983337,
"learning_rate": 1.34367197442791e-06,
"loss": 0.08478391170501709,
"step": 1668
},
{
"epoch": 3.5306553911205074,
"grad_norm": 0.2763535678386688,
"learning_rate": 1.3376666517513931e-06,
"loss": 0.6243268251419067,
"step": 1670
},
{
"epoch": 3.5348837209302326,
"grad_norm": 2.9814653396606445,
"learning_rate": 1.3317122152723078e-06,
"loss": 0.39350929856300354,
"step": 1672
},
{
"epoch": 3.539112050739958,
"grad_norm": 1.023309350013733,
"learning_rate": 1.3258087377860323e-06,
"loss": 0.7441390752792358,
"step": 1674
},
{
"epoch": 3.543340380549683,
"grad_norm": 3.079191207885742,
"learning_rate": 1.3199562914649519e-06,
"loss": 0.4954609274864197,
"step": 1676
},
{
"epoch": 3.547568710359408,
"grad_norm": 0.9636067152023315,
"learning_rate": 1.314154947857571e-06,
"loss": 0.7136701345443726,
"step": 1678
},
{
"epoch": 3.5517970401691334,
"grad_norm": 0.8986470103263855,
"learning_rate": 1.3084047778876487e-06,
"loss": 0.45399096608161926,
"step": 1680
},
{
"epoch": 3.556025369978858,
"grad_norm": 4.64102840423584,
"learning_rate": 1.3027058518533228e-06,
"loss": 0.46073251962661743,
"step": 1682
},
{
"epoch": 3.5602536997885834,
"grad_norm": 1.3546741008758545,
"learning_rate": 1.2970582394262516e-06,
"loss": 0.3964889943599701,
"step": 1684
},
{
"epoch": 3.5644820295983086,
"grad_norm": 4.363003253936768,
"learning_rate": 1.291462009650768e-06,
"loss": 0.2546060085296631,
"step": 1686
},
{
"epoch": 3.568710359408034,
"grad_norm": 1.1613324880599976,
"learning_rate": 1.2859172309430285e-06,
"loss": 0.4963217079639435,
"step": 1688
},
{
"epoch": 3.572938689217759,
"grad_norm": 0.07351069152355194,
"learning_rate": 1.280423971090181e-06,
"loss": 0.15672074258327484,
"step": 1690
},
{
"epoch": 3.5771670190274842,
"grad_norm": 1.7787182331085205,
"learning_rate": 1.2749822972495333e-06,
"loss": 0.3415200412273407,
"step": 1692
},
{
"epoch": 3.5813953488372094,
"grad_norm": 4.3776679039001465,
"learning_rate": 1.2695922759477356e-06,
"loss": 0.5011807084083557,
"step": 1694
},
{
"epoch": 3.585623678646934,
"grad_norm": 3.016995668411255,
"learning_rate": 1.2642539730799622e-06,
"loss": 0.25671106576919556,
"step": 1696
},
{
"epoch": 3.58985200845666,
"grad_norm": 1.1093790531158447,
"learning_rate": 1.2589674539091115e-06,
"loss": 0.7131949067115784,
"step": 1698
},
{
"epoch": 3.5940803382663846,
"grad_norm": 1.0579215288162231,
"learning_rate": 1.2537327830650042e-06,
"loss": 0.8628058433532715,
"step": 1700
},
{
"epoch": 3.59830866807611,
"grad_norm": 3.7192111015319824,
"learning_rate": 1.248550024543594e-06,
"loss": 0.7582141757011414,
"step": 1702
},
{
"epoch": 3.602536997885835,
"grad_norm": 2.5885941982269287,
"learning_rate": 1.2434192417061862e-06,
"loss": 0.3018801808357239,
"step": 1704
},
{
"epoch": 3.6067653276955602,
"grad_norm": 0.898874044418335,
"learning_rate": 1.2383404972786625e-06,
"loss": 0.3458724617958069,
"step": 1706
},
{
"epoch": 3.6109936575052854,
"grad_norm": 2.127307415008545,
"learning_rate": 1.2333138533507139e-06,
"loss": 0.6275134086608887,
"step": 1708
},
{
"epoch": 3.6152219873150107,
"grad_norm": 2.348921060562134,
"learning_rate": 1.2283393713750816e-06,
"loss": 0.47557583451271057,
"step": 1710
},
{
"epoch": 3.619450317124736,
"grad_norm": 1.158357858657837,
"learning_rate": 1.2234171121668069e-06,
"loss": 0.3919115662574768,
"step": 1712
},
{
"epoch": 3.6236786469344606,
"grad_norm": 0.2824644446372986,
"learning_rate": 1.2185471359024855e-06,
"loss": 0.22734051942825317,
"step": 1714
},
{
"epoch": 3.6279069767441863,
"grad_norm": 7.804412841796875,
"learning_rate": 1.2137295021195341e-06,
"loss": 0.6430119276046753,
"step": 1716
},
{
"epoch": 3.632135306553911,
"grad_norm": 0.8730980157852173,
"learning_rate": 1.2089642697154613e-06,
"loss": 0.4239366352558136,
"step": 1718
},
{
"epoch": 3.6363636363636362,
"grad_norm": 4.0313520431518555,
"learning_rate": 1.2042514969471475e-06,
"loss": 0.38808903098106384,
"step": 1720
},
{
"epoch": 3.6405919661733614,
"grad_norm": 1.1828868389129639,
"learning_rate": 1.1995912414301329e-06,
"loss": 0.5600469708442688,
"step": 1722
},
{
"epoch": 3.6448202959830867,
"grad_norm": 2.407282590866089,
"learning_rate": 1.1949835601379144e-06,
"loss": 0.5465320348739624,
"step": 1724
},
{
"epoch": 3.649048625792812,
"grad_norm": 1.004649043083191,
"learning_rate": 1.1904285094012463e-06,
"loss": 0.7993420958518982,
"step": 1726
},
{
"epoch": 3.653276955602537,
"grad_norm": 2.1744253635406494,
"learning_rate": 1.185926144907454e-06,
"loss": 0.7719228863716125,
"step": 1728
},
{
"epoch": 3.6575052854122623,
"grad_norm": 4.4379353523254395,
"learning_rate": 1.1814765216997522e-06,
"loss": 0.6549544334411621,
"step": 1730
},
{
"epoch": 3.6617336152219875,
"grad_norm": 1.209997296333313,
"learning_rate": 1.1770796941765727e-06,
"loss": 0.6419261693954468,
"step": 1732
},
{
"epoch": 3.6659619450317127,
"grad_norm": 1.4616214036941528,
"learning_rate": 1.1727357160908978e-06,
"loss": 0.7665488719940186,
"step": 1734
},
{
"epoch": 3.6701902748414374,
"grad_norm": 0.9722137451171875,
"learning_rate": 1.1684446405496052e-06,
"loss": 0.6962933540344238,
"step": 1736
},
{
"epoch": 3.6744186046511627,
"grad_norm": 4.480654239654541,
"learning_rate": 1.1642065200128179e-06,
"loss": 0.13903529942035675,
"step": 1738
},
{
"epoch": 3.678646934460888,
"grad_norm": 1.3094059228897095,
"learning_rate": 1.1600214062932613e-06,
"loss": 0.41632524132728577,
"step": 1740
},
{
"epoch": 3.682875264270613,
"grad_norm": 7.562599182128906,
"learning_rate": 1.1558893505556315e-06,
"loss": 0.3600577116012573,
"step": 1742
},
{
"epoch": 3.6871035940803383,
"grad_norm": 1.2115343809127808,
"learning_rate": 1.1518104033159715e-06,
"loss": 0.7974560856819153,
"step": 1744
},
{
"epoch": 3.6913319238900635,
"grad_norm": 0.22515630722045898,
"learning_rate": 1.1477846144410493e-06,
"loss": 0.318876713514328,
"step": 1746
},
{
"epoch": 3.6955602536997887,
"grad_norm": 3.837008237838745,
"learning_rate": 1.1438120331477506e-06,
"loss": 0.3402837812900543,
"step": 1748
},
{
"epoch": 3.699788583509514,
"grad_norm": 1.5260231494903564,
"learning_rate": 1.1398927080024796e-06,
"loss": 0.8928380608558655,
"step": 1750
},
{
"epoch": 3.704016913319239,
"grad_norm": 4.764921188354492,
"learning_rate": 1.1360266869205593e-06,
"loss": 0.2772078812122345,
"step": 1752
},
{
"epoch": 3.708245243128964,
"grad_norm": 7.186609268188477,
"learning_rate": 1.1322140171656507e-06,
"loss": 0.5219692587852478,
"step": 1754
},
{
"epoch": 3.712473572938689,
"grad_norm": 0.9503617882728577,
"learning_rate": 1.1284547453491734e-06,
"loss": 0.6985856294631958,
"step": 1756
},
{
"epoch": 3.7167019027484143,
"grad_norm": 4.593424320220947,
"learning_rate": 1.1247489174297372e-06,
"loss": 0.4472501277923584,
"step": 1758
},
{
"epoch": 3.7209302325581395,
"grad_norm": 0.19857680797576904,
"learning_rate": 1.1210965787125743e-06,
"loss": 0.3343963623046875,
"step": 1760
},
{
"epoch": 3.7251585623678647,
"grad_norm": 2.3729851245880127,
"learning_rate": 1.1174977738489955e-06,
"loss": 0.6045705080032349,
"step": 1762
},
{
"epoch": 3.72938689217759,
"grad_norm": 0.7916973829269409,
"learning_rate": 1.113952546835835e-06,
"loss": 0.3852551281452179,
"step": 1764
},
{
"epoch": 3.733615221987315,
"grad_norm": 1.8627989292144775,
"learning_rate": 1.1104609410149174e-06,
"loss": 0.7053631544113159,
"step": 1766
},
{
"epoch": 3.7378435517970403,
"grad_norm": 6.04041862487793,
"learning_rate": 1.1070229990725262e-06,
"loss": 0.5868628621101379,
"step": 1768
},
{
"epoch": 3.7420718816067655,
"grad_norm": 0.16452986001968384,
"learning_rate": 1.1036387630388841e-06,
"loss": 0.1771097481250763,
"step": 1770
},
{
"epoch": 3.7463002114164903,
"grad_norm": 1.3313934803009033,
"learning_rate": 1.100308274287636e-06,
"loss": 0.4322102963924408,
"step": 1772
},
{
"epoch": 3.7505285412262155,
"grad_norm": 2.334596633911133,
"learning_rate": 1.0970315735353434e-06,
"loss": 0.8584076762199402,
"step": 1774
},
{
"epoch": 3.7547568710359407,
"grad_norm": 1.4565469026565552,
"learning_rate": 1.0938087008409909e-06,
"loss": 0.7099536657333374,
"step": 1776
},
{
"epoch": 3.758985200845666,
"grad_norm": 1.444453477859497,
"learning_rate": 1.0906396956054919e-06,
"loss": 0.7511817812919617,
"step": 1778
},
{
"epoch": 3.763213530655391,
"grad_norm": 1.4505425691604614,
"learning_rate": 1.0875245965712077e-06,
"loss": 0.9134868383407593,
"step": 1780
},
{
"epoch": 3.7674418604651163,
"grad_norm": 1.4371905326843262,
"learning_rate": 1.0844634418214763e-06,
"loss": 0.705837607383728,
"step": 1782
},
{
"epoch": 3.7716701902748415,
"grad_norm": 1.7402514219284058,
"learning_rate": 1.0814562687801448e-06,
"loss": 0.1345704197883606,
"step": 1784
},
{
"epoch": 3.7758985200845667,
"grad_norm": 1.482924222946167,
"learning_rate": 1.0785031142111103e-06,
"loss": 0.5008679032325745,
"step": 1786
},
{
"epoch": 3.780126849894292,
"grad_norm": 2.2061705589294434,
"learning_rate": 1.075604014217875e-06,
"loss": 0.4554566740989685,
"step": 1788
},
{
"epoch": 3.7843551797040167,
"grad_norm": 7.964588165283203,
"learning_rate": 1.0727590042431009e-06,
"loss": 0.12333111464977264,
"step": 1790
},
{
"epoch": 3.7885835095137423,
"grad_norm": 0.9754509329795837,
"learning_rate": 1.0699681190681777e-06,
"loss": 0.5811925530433655,
"step": 1792
},
{
"epoch": 3.792811839323467,
"grad_norm": 1.7581695318222046,
"learning_rate": 1.067231392812798e-06,
"loss": 0.71378093957901,
"step": 1794
},
{
"epoch": 3.7970401691331923,
"grad_norm": 1.8098187446594238,
"learning_rate": 1.0645488589345396e-06,
"loss": 0.3802398145198822,
"step": 1796
},
{
"epoch": 3.8012684989429175,
"grad_norm": 3.417020320892334,
"learning_rate": 1.0619205502284574e-06,
"loss": 0.6794342398643494,
"step": 1798
},
{
"epoch": 3.8054968287526427,
"grad_norm": 1.5233476161956787,
"learning_rate": 1.0593464988266811e-06,
"loss": 0.7984983325004578,
"step": 1800
},
{
"epoch": 3.809725158562368,
"grad_norm": 1.29951810836792,
"learning_rate": 1.0568267361980225e-06,
"loss": 0.5276030898094177,
"step": 1802
},
{
"epoch": 3.813953488372093,
"grad_norm": 1.5124551057815552,
"learning_rate": 1.0543612931475935e-06,
"loss": 0.6599000096321106,
"step": 1804
},
{
"epoch": 3.8181818181818183,
"grad_norm": 1.0791144371032715,
"learning_rate": 1.0519501998164247e-06,
"loss": 0.7276206016540527,
"step": 1806
},
{
"epoch": 3.822410147991543,
"grad_norm": 4.975444316864014,
"learning_rate": 1.049593485681101e-06,
"loss": 0.20384155213832855,
"step": 1808
},
{
"epoch": 3.8266384778012688,
"grad_norm": 2.325077533721924,
"learning_rate": 1.0472911795533997e-06,
"loss": 0.8015995025634766,
"step": 1810
},
{
"epoch": 3.8308668076109935,
"grad_norm": 1.7632167339324951,
"learning_rate": 1.045043309579938e-06,
"loss": 0.8200293779373169,
"step": 1812
},
{
"epoch": 3.8350951374207187,
"grad_norm": 0.8766272068023682,
"learning_rate": 1.0428499032418297e-06,
"loss": 0.6026349663734436,
"step": 1814
},
{
"epoch": 3.839323467230444,
"grad_norm": 0.9692689776420593,
"learning_rate": 1.0407109873543484e-06,
"loss": 0.4662686586380005,
"step": 1816
},
{
"epoch": 3.843551797040169,
"grad_norm": 2.7962687015533447,
"learning_rate": 1.0386265880665996e-06,
"loss": 0.4446929097175598,
"step": 1818
},
{
"epoch": 3.8477801268498943,
"grad_norm": 1.7485017776489258,
"learning_rate": 1.036596730861203e-06,
"loss": 0.41421645879745483,
"step": 1820
},
{
"epoch": 3.8520084566596196,
"grad_norm": 2.2581613063812256,
"learning_rate": 1.0346214405539783e-06,
"loss": 0.5091472268104553,
"step": 1822
},
{
"epoch": 3.8562367864693448,
"grad_norm": 0.8584007024765015,
"learning_rate": 1.0327007412936419e-06,
"loss": 0.4677910804748535,
"step": 1824
},
{
"epoch": 3.8604651162790695,
"grad_norm": 11.01543140411377,
"learning_rate": 1.0308346565615158e-06,
"loss": 0.40488898754119873,
"step": 1826
},
{
"epoch": 3.864693446088795,
"grad_norm": 0.91716068983078,
"learning_rate": 1.0290232091712337e-06,
"loss": 0.7792301177978516,
"step": 1828
},
{
"epoch": 3.86892177589852,
"grad_norm": 1.8299479484558105,
"learning_rate": 1.0272664212684687e-06,
"loss": 0.703758955001831,
"step": 1830
},
{
"epoch": 3.873150105708245,
"grad_norm": 1.7577440738677979,
"learning_rate": 1.025564314330659e-06,
"loss": 0.44360318779945374,
"step": 1832
},
{
"epoch": 3.8773784355179703,
"grad_norm": 1.157348871231079,
"learning_rate": 1.0239169091667447e-06,
"loss": 0.3563556373119354,
"step": 1834
},
{
"epoch": 3.8816067653276956,
"grad_norm": 0.5805785655975342,
"learning_rate": 1.022324225916917e-06,
"loss": 0.46799778938293457,
"step": 1836
},
{
"epoch": 3.8858350951374208,
"grad_norm": 0.20997752249240875,
"learning_rate": 1.0207862840523672e-06,
"loss": 0.3553447425365448,
"step": 1838
},
{
"epoch": 3.890063424947146,
"grad_norm": 0.97221440076828,
"learning_rate": 1.0193031023750534e-06,
"loss": 0.5947901606559753,
"step": 1840
},
{
"epoch": 3.894291754756871,
"grad_norm": 1.0569690465927124,
"learning_rate": 1.017874699017468e-06,
"loss": 0.6770817041397095,
"step": 1842
},
{
"epoch": 3.898520084566596,
"grad_norm": 3.0368893146514893,
"learning_rate": 1.0165010914424142e-06,
"loss": 0.44024741649627686,
"step": 1844
},
{
"epoch": 3.9027484143763216,
"grad_norm": 107.50782012939453,
"learning_rate": 1.015182296442798e-06,
"loss": 0.43350571393966675,
"step": 1846
},
{
"epoch": 3.9069767441860463,
"grad_norm": 1.9543638229370117,
"learning_rate": 1.0139183301414176e-06,
"loss": 0.15080475807189941,
"step": 1848
},
{
"epoch": 3.9112050739957716,
"grad_norm": 5.518731117248535,
"learning_rate": 1.0127092079907682e-06,
"loss": 0.6492822170257568,
"step": 1850
},
{
"epoch": 3.9154334038054968,
"grad_norm": 0.7199242115020752,
"learning_rate": 1.0115549447728551e-06,
"loss": 0.5045459866523743,
"step": 1852
},
{
"epoch": 3.919661733615222,
"grad_norm": 0.8023747205734253,
"learning_rate": 1.0104555545990088e-06,
"loss": 0.6962122321128845,
"step": 1854
},
{
"epoch": 3.923890063424947,
"grad_norm": 1.550148367881775,
"learning_rate": 1.009411050909716e-06,
"loss": 0.3795410394668579,
"step": 1856
},
{
"epoch": 3.9281183932346724,
"grad_norm": 1.9650306701660156,
"learning_rate": 1.008421446474455e-06,
"loss": 0.5384699106216431,
"step": 1858
},
{
"epoch": 3.9323467230443976,
"grad_norm": 2.030888795852661,
"learning_rate": 1.0074867533915367e-06,
"loss": 0.21794721484184265,
"step": 1860
},
{
"epoch": 3.9365750528541223,
"grad_norm": 3.3347012996673584,
"learning_rate": 1.0066069830879593e-06,
"loss": 0.814431369304657,
"step": 1862
},
{
"epoch": 3.940803382663848,
"grad_norm": 1.170685887336731,
"learning_rate": 1.0057821463192687e-06,
"loss": 0.6267325282096863,
"step": 1864
},
{
"epoch": 3.9450317124735728,
"grad_norm": 1.4499508142471313,
"learning_rate": 1.005012253169426e-06,
"loss": 0.6560044884681702,
"step": 1866
},
{
"epoch": 3.949260042283298,
"grad_norm": 1.9098694324493408,
"learning_rate": 1.0042973130506839e-06,
"loss": 0.005977078340947628,
"step": 1868
},
{
"epoch": 3.953488372093023,
"grad_norm": 1.770297646522522,
"learning_rate": 1.003637334703472e-06,
"loss": 0.7988382577896118,
"step": 1870
},
{
"epoch": 3.9577167019027484,
"grad_norm": 1.5987740755081177,
"learning_rate": 1.0030323261962922e-06,
"loss": 0.5535221695899963,
"step": 1872
},
{
"epoch": 3.9619450317124736,
"grad_norm": 1.6556613445281982,
"learning_rate": 1.002482294925615e-06,
"loss": 0.4262787103652954,
"step": 1874
},
{
"epoch": 3.966173361522199,
"grad_norm": 0.8752848505973816,
"learning_rate": 1.0019872476157952e-06,
"loss": 0.6988322138786316,
"step": 1876
},
{
"epoch": 3.970401691331924,
"grad_norm": 5.043781757354736,
"learning_rate": 1.0015471903189834e-06,
"loss": 0.2591104805469513,
"step": 1878
},
{
"epoch": 3.974630021141649,
"grad_norm": 1.1084809303283691,
"learning_rate": 1.0011621284150583e-06,
"loss": 0.3260257840156555,
"step": 1880
},
{
"epoch": 3.9788583509513744,
"grad_norm": 1.7564277648925781,
"learning_rate": 1.000832066611556e-06,
"loss": 0.7079311013221741,
"step": 1882
},
{
"epoch": 3.983086680761099,
"grad_norm": 1.4538663625717163,
"learning_rate": 1.0005570089436137e-06,
"loss": 0.8567215800285339,
"step": 1884
},
{
"epoch": 3.9873150105708244,
"grad_norm": 1.0733779668807983,
"learning_rate": 1.0003369587739234e-06,
"loss": 0.5167215466499329,
"step": 1886
},
{
"epoch": 3.9915433403805496,
"grad_norm": 1.339658498764038,
"learning_rate": 1.0001719187926854e-06,
"loss": 0.7091305255889893,
"step": 1888
},
{
"epoch": 3.995771670190275,
"grad_norm": 0.9317363500595093,
"learning_rate": 1.0000618910175811e-06,
"loss": 0.6818731427192688,
"step": 1890
},
{
"epoch": 4.0,
"grad_norm": 1.0739333629608154,
"learning_rate": 1.000006876793743e-06,
"loss": 0.39930370450019836,
"step": 1892
},
{
"epoch": 4.0,
"step": 1892,
"total_flos": 3.5845266747435254e+18,
"train_loss": 0.8258528457846233,
"train_runtime": 9426.6588,
"train_samples_per_second": 6.021,
"train_steps_per_second": 0.201
}
],
"logging_steps": 2,
"max_steps": 1892,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.5845266747435254e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}