atsuki-yamaguchi's picture
Upload folder using huggingface_hub
e330393 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.15573493913359462,
"eval_steps": 500,
"global_step": 6000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 55.36319351196289,
"learning_rate": 5e-06,
"loss": 9.4221,
"step": 5
},
{
"epoch": 0.0,
"grad_norm": 27.91165542602539,
"learning_rate": 1e-05,
"loss": 8.9057,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 13.735514640808105,
"learning_rate": 1.5e-05,
"loss": 7.7782,
"step": 15
},
{
"epoch": 0.0,
"grad_norm": 8.841108322143555,
"learning_rate": 2e-05,
"loss": 7.1321,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 6.489623546600342,
"learning_rate": 2.5e-05,
"loss": 6.7408,
"step": 25
},
{
"epoch": 0.0,
"grad_norm": 6.439656734466553,
"learning_rate": 3e-05,
"loss": 6.3743,
"step": 30
},
{
"epoch": 0.0,
"grad_norm": 5.118189334869385,
"learning_rate": 3.5e-05,
"loss": 6.2465,
"step": 35
},
{
"epoch": 0.0,
"grad_norm": 3.616164207458496,
"learning_rate": 4e-05,
"loss": 6.0074,
"step": 40
},
{
"epoch": 0.0,
"grad_norm": 4.094961643218994,
"learning_rate": 4.5e-05,
"loss": 5.7808,
"step": 45
},
{
"epoch": 0.0,
"grad_norm": 3.927417039871216,
"learning_rate": 5e-05,
"loss": 5.579,
"step": 50
},
{
"epoch": 0.0,
"grad_norm": 3.7323696613311768,
"learning_rate": 5.500000000000001e-05,
"loss": 5.3895,
"step": 55
},
{
"epoch": 0.0,
"grad_norm": 3.6021604537963867,
"learning_rate": 6e-05,
"loss": 5.1477,
"step": 60
},
{
"epoch": 0.0,
"grad_norm": 3.5518462657928467,
"learning_rate": 6.500000000000001e-05,
"loss": 4.9691,
"step": 65
},
{
"epoch": 0.0,
"grad_norm": 2.8650312423706055,
"learning_rate": 7e-05,
"loss": 4.809,
"step": 70
},
{
"epoch": 0.0,
"grad_norm": 3.923182725906372,
"learning_rate": 7.500000000000001e-05,
"loss": 4.6982,
"step": 75
},
{
"epoch": 0.0,
"grad_norm": 2.9143548011779785,
"learning_rate": 8e-05,
"loss": 4.4536,
"step": 80
},
{
"epoch": 0.0,
"grad_norm": 3.8249118328094482,
"learning_rate": 8.5e-05,
"loss": 4.4123,
"step": 85
},
{
"epoch": 0.0,
"grad_norm": 2.8610007762908936,
"learning_rate": 9e-05,
"loss": 4.1761,
"step": 90
},
{
"epoch": 0.0,
"grad_norm": 2.5008883476257324,
"learning_rate": 9.5e-05,
"loss": 4.2017,
"step": 95
},
{
"epoch": 0.0,
"grad_norm": 2.3759865760803223,
"learning_rate": 0.0001,
"loss": 4.0345,
"step": 100
},
{
"epoch": 0.0,
"grad_norm": 2.382518768310547,
"learning_rate": 9.999999582259295e-05,
"loss": 4.0203,
"step": 105
},
{
"epoch": 0.0,
"grad_norm": 2.3481087684631348,
"learning_rate": 9.999998329037248e-05,
"loss": 3.9503,
"step": 110
},
{
"epoch": 0.0,
"grad_norm": 2.6210474967956543,
"learning_rate": 9.999996240334068e-05,
"loss": 3.8307,
"step": 115
},
{
"epoch": 0.0,
"grad_norm": 2.382927656173706,
"learning_rate": 9.999993316150106e-05,
"loss": 3.9693,
"step": 120
},
{
"epoch": 0.0,
"grad_norm": 2.484893798828125,
"learning_rate": 9.999989556485852e-05,
"loss": 3.8611,
"step": 125
},
{
"epoch": 0.0,
"grad_norm": 2.187300443649292,
"learning_rate": 9.999984961341928e-05,
"loss": 3.8539,
"step": 130
},
{
"epoch": 0.0,
"grad_norm": 2.034395933151245,
"learning_rate": 9.999979530719108e-05,
"loss": 3.7403,
"step": 135
},
{
"epoch": 0.0,
"grad_norm": 2.1583139896392822,
"learning_rate": 9.999973264618297e-05,
"loss": 3.7363,
"step": 140
},
{
"epoch": 0.0,
"grad_norm": 2.08280348777771,
"learning_rate": 9.999966163040541e-05,
"loss": 3.6022,
"step": 145
},
{
"epoch": 0.0,
"grad_norm": 1.8879830837249756,
"learning_rate": 9.99995822598703e-05,
"loss": 3.6186,
"step": 150
},
{
"epoch": 0.0,
"grad_norm": 2.079864740371704,
"learning_rate": 9.999949453459088e-05,
"loss": 3.612,
"step": 155
},
{
"epoch": 0.0,
"grad_norm": 1.8108304738998413,
"learning_rate": 9.999939845458179e-05,
"loss": 3.6519,
"step": 160
},
{
"epoch": 0.0,
"grad_norm": 1.9264494180679321,
"learning_rate": 9.999929401985911e-05,
"loss": 3.5866,
"step": 165
},
{
"epoch": 0.0,
"grad_norm": 1.9425358772277832,
"learning_rate": 9.99991812304403e-05,
"loss": 3.5623,
"step": 170
},
{
"epoch": 0.0,
"grad_norm": 1.8959733247756958,
"learning_rate": 9.999906008634416e-05,
"loss": 3.4725,
"step": 175
},
{
"epoch": 0.0,
"grad_norm": 7.430732250213623,
"learning_rate": 9.9998930587591e-05,
"loss": 3.3967,
"step": 180
},
{
"epoch": 0.0,
"grad_norm": 1.9701666831970215,
"learning_rate": 9.99987927342024e-05,
"loss": 3.4269,
"step": 185
},
{
"epoch": 0.0,
"grad_norm": 1.8052351474761963,
"learning_rate": 9.999864652620143e-05,
"loss": 3.5281,
"step": 190
},
{
"epoch": 0.01,
"grad_norm": 1.6505353450775146,
"learning_rate": 9.999849196361251e-05,
"loss": 3.5391,
"step": 195
},
{
"epoch": 0.01,
"grad_norm": 2.7946834564208984,
"learning_rate": 9.999832904646147e-05,
"loss": 3.4091,
"step": 200
},
{
"epoch": 0.01,
"grad_norm": 1.8914915323257446,
"learning_rate": 9.99981577747755e-05,
"loss": 3.3459,
"step": 205
},
{
"epoch": 0.01,
"grad_norm": 1.6807687282562256,
"learning_rate": 9.999797814858327e-05,
"loss": 3.4312,
"step": 210
},
{
"epoch": 0.01,
"grad_norm": 1.8404138088226318,
"learning_rate": 9.999779016791477e-05,
"loss": 3.3887,
"step": 215
},
{
"epoch": 0.01,
"grad_norm": 1.6583664417266846,
"learning_rate": 9.999759383280141e-05,
"loss": 3.4189,
"step": 220
},
{
"epoch": 0.01,
"grad_norm": 1.7972958087921143,
"learning_rate": 9.9997389143276e-05,
"loss": 3.2839,
"step": 225
},
{
"epoch": 0.01,
"grad_norm": 1.6305922269821167,
"learning_rate": 9.999717609937275e-05,
"loss": 3.3237,
"step": 230
},
{
"epoch": 0.01,
"grad_norm": 1.7618087530136108,
"learning_rate": 9.999695470112725e-05,
"loss": 3.2777,
"step": 235
},
{
"epoch": 0.01,
"grad_norm": 1.7651971578598022,
"learning_rate": 9.99967249485765e-05,
"loss": 3.2781,
"step": 240
},
{
"epoch": 0.01,
"grad_norm": 1.6341221332550049,
"learning_rate": 9.999648684175888e-05,
"loss": 3.3033,
"step": 245
},
{
"epoch": 0.01,
"grad_norm": 1.7574623823165894,
"learning_rate": 9.999624038071418e-05,
"loss": 3.2653,
"step": 250
},
{
"epoch": 0.01,
"grad_norm": 1.578574299812317,
"learning_rate": 9.99959855654836e-05,
"loss": 3.3131,
"step": 255
},
{
"epoch": 0.01,
"grad_norm": 1.6800299882888794,
"learning_rate": 9.999572239610968e-05,
"loss": 3.2099,
"step": 260
},
{
"epoch": 0.01,
"grad_norm": 1.4835729598999023,
"learning_rate": 9.999545087263645e-05,
"loss": 3.1283,
"step": 265
},
{
"epoch": 0.01,
"grad_norm": 1.5694025754928589,
"learning_rate": 9.999517099510923e-05,
"loss": 3.1847,
"step": 270
},
{
"epoch": 0.01,
"grad_norm": 1.4740264415740967,
"learning_rate": 9.999488276357481e-05,
"loss": 3.1748,
"step": 275
},
{
"epoch": 0.01,
"grad_norm": 1.642398715019226,
"learning_rate": 9.999458617808137e-05,
"loss": 3.2423,
"step": 280
},
{
"epoch": 0.01,
"grad_norm": 1.9440889358520508,
"learning_rate": 9.999428123867843e-05,
"loss": 3.2175,
"step": 285
},
{
"epoch": 0.01,
"grad_norm": 1.5306422710418701,
"learning_rate": 9.999396794541695e-05,
"loss": 3.1484,
"step": 290
},
{
"epoch": 0.01,
"grad_norm": 1.5568021535873413,
"learning_rate": 9.999364629834931e-05,
"loss": 3.2437,
"step": 295
},
{
"epoch": 0.01,
"grad_norm": 1.5204740762710571,
"learning_rate": 9.999331629752924e-05,
"loss": 3.2234,
"step": 300
},
{
"epoch": 0.01,
"grad_norm": 1.4896517992019653,
"learning_rate": 9.999297794301188e-05,
"loss": 3.1452,
"step": 305
},
{
"epoch": 0.01,
"grad_norm": 1.6853694915771484,
"learning_rate": 9.999263123485376e-05,
"loss": 3.2063,
"step": 310
},
{
"epoch": 0.01,
"grad_norm": 1.6633198261260986,
"learning_rate": 9.999227617311282e-05,
"loss": 3.1835,
"step": 315
},
{
"epoch": 0.01,
"grad_norm": 1.6276299953460693,
"learning_rate": 9.99919127578484e-05,
"loss": 3.163,
"step": 320
},
{
"epoch": 0.01,
"grad_norm": 1.4966531991958618,
"learning_rate": 9.999154098912122e-05,
"loss": 3.1888,
"step": 325
},
{
"epoch": 0.01,
"grad_norm": 1.4910067319869995,
"learning_rate": 9.999116086699338e-05,
"loss": 3.1074,
"step": 330
},
{
"epoch": 0.01,
"grad_norm": 1.4052242040634155,
"learning_rate": 9.999077239152843e-05,
"loss": 3.1198,
"step": 335
},
{
"epoch": 0.01,
"grad_norm": 1.3959842920303345,
"learning_rate": 9.999037556279127e-05,
"loss": 3.1288,
"step": 340
},
{
"epoch": 0.01,
"grad_norm": 1.496795892715454,
"learning_rate": 9.998997038084818e-05,
"loss": 3.1105,
"step": 345
},
{
"epoch": 0.01,
"grad_norm": 1.375389575958252,
"learning_rate": 9.998955684576692e-05,
"loss": 3.1137,
"step": 350
},
{
"epoch": 0.01,
"grad_norm": 1.671631932258606,
"learning_rate": 9.998913495761654e-05,
"loss": 3.145,
"step": 355
},
{
"epoch": 0.01,
"grad_norm": 1.4133590459823608,
"learning_rate": 9.998870471646757e-05,
"loss": 3.0291,
"step": 360
},
{
"epoch": 0.01,
"grad_norm": 1.4495996236801147,
"learning_rate": 9.99882661223919e-05,
"loss": 3.0753,
"step": 365
},
{
"epoch": 0.01,
"grad_norm": 1.43621027469635,
"learning_rate": 9.998781917546278e-05,
"loss": 3.0964,
"step": 370
},
{
"epoch": 0.01,
"grad_norm": 1.5507177114486694,
"learning_rate": 9.998736387575493e-05,
"loss": 3.0962,
"step": 375
},
{
"epoch": 0.01,
"grad_norm": 1.3702166080474854,
"learning_rate": 9.998690022334442e-05,
"loss": 3.0427,
"step": 380
},
{
"epoch": 0.01,
"grad_norm": 1.4262795448303223,
"learning_rate": 9.998642821830873e-05,
"loss": 3.0842,
"step": 385
},
{
"epoch": 0.01,
"grad_norm": 1.4134531021118164,
"learning_rate": 9.99859478607267e-05,
"loss": 3.0912,
"step": 390
},
{
"epoch": 0.01,
"grad_norm": 1.4482907056808472,
"learning_rate": 9.998545915067864e-05,
"loss": 3.1275,
"step": 395
},
{
"epoch": 0.01,
"grad_norm": 1.5179977416992188,
"learning_rate": 9.998496208824618e-05,
"loss": 3.1181,
"step": 400
},
{
"epoch": 0.01,
"grad_norm": 1.4615247249603271,
"learning_rate": 9.99844566735124e-05,
"loss": 3.068,
"step": 405
},
{
"epoch": 0.01,
"grad_norm": 1.496274709701538,
"learning_rate": 9.998394290656172e-05,
"loss": 3.0691,
"step": 410
},
{
"epoch": 0.01,
"grad_norm": 1.3730614185333252,
"learning_rate": 9.998342078748002e-05,
"loss": 3.1004,
"step": 415
},
{
"epoch": 0.01,
"grad_norm": 1.4433605670928955,
"learning_rate": 9.998289031635453e-05,
"loss": 3.0329,
"step": 420
},
{
"epoch": 0.01,
"grad_norm": 1.3417984247207642,
"learning_rate": 9.998235149327389e-05,
"loss": 3.0998,
"step": 425
},
{
"epoch": 0.01,
"grad_norm": 1.3686726093292236,
"learning_rate": 9.998180431832816e-05,
"loss": 3.0269,
"step": 430
},
{
"epoch": 0.01,
"grad_norm": 1.3730640411376953,
"learning_rate": 9.998124879160872e-05,
"loss": 3.0793,
"step": 435
},
{
"epoch": 0.01,
"grad_norm": 1.4302995204925537,
"learning_rate": 9.998068491320844e-05,
"loss": 2.9739,
"step": 440
},
{
"epoch": 0.01,
"grad_norm": 1.3892772197723389,
"learning_rate": 9.998011268322152e-05,
"loss": 3.0802,
"step": 445
},
{
"epoch": 0.01,
"grad_norm": 1.3409262895584106,
"learning_rate": 9.997953210174361e-05,
"loss": 2.9613,
"step": 450
},
{
"epoch": 0.01,
"grad_norm": 1.3266420364379883,
"learning_rate": 9.997894316887167e-05,
"loss": 2.9777,
"step": 455
},
{
"epoch": 0.01,
"grad_norm": 1.3892908096313477,
"learning_rate": 9.997834588470414e-05,
"loss": 3.0595,
"step": 460
},
{
"epoch": 0.01,
"grad_norm": 1.478256344795227,
"learning_rate": 9.997774024934082e-05,
"loss": 3.0467,
"step": 465
},
{
"epoch": 0.01,
"grad_norm": 1.3662242889404297,
"learning_rate": 9.997712626288292e-05,
"loss": 3.0275,
"step": 470
},
{
"epoch": 0.01,
"grad_norm": 1.574994444847107,
"learning_rate": 9.997650392543303e-05,
"loss": 3.0277,
"step": 475
},
{
"epoch": 0.01,
"grad_norm": 1.3383744955062866,
"learning_rate": 9.997587323709512e-05,
"loss": 2.9781,
"step": 480
},
{
"epoch": 0.01,
"grad_norm": 1.364506483078003,
"learning_rate": 9.997523419797459e-05,
"loss": 2.9935,
"step": 485
},
{
"epoch": 0.01,
"grad_norm": 1.289879560470581,
"learning_rate": 9.997458680817824e-05,
"loss": 2.9654,
"step": 490
},
{
"epoch": 0.01,
"grad_norm": 1.4652094841003418,
"learning_rate": 9.997393106781421e-05,
"loss": 2.9785,
"step": 495
},
{
"epoch": 0.01,
"grad_norm": 1.261389970779419,
"learning_rate": 9.99732669769921e-05,
"loss": 2.947,
"step": 500
},
{
"epoch": 0.01,
"grad_norm": 1.3607465028762817,
"learning_rate": 9.997259453582286e-05,
"loss": 2.8793,
"step": 505
},
{
"epoch": 0.01,
"grad_norm": 1.522127389907837,
"learning_rate": 9.997191374441887e-05,
"loss": 3.0362,
"step": 510
},
{
"epoch": 0.01,
"grad_norm": 1.3926830291748047,
"learning_rate": 9.997122460289387e-05,
"loss": 2.8733,
"step": 515
},
{
"epoch": 0.01,
"grad_norm": 1.3031957149505615,
"learning_rate": 9.997052711136301e-05,
"loss": 2.9601,
"step": 520
},
{
"epoch": 0.01,
"grad_norm": 1.4590163230895996,
"learning_rate": 9.996982126994287e-05,
"loss": 2.9509,
"step": 525
},
{
"epoch": 0.01,
"grad_norm": 1.5168179273605347,
"learning_rate": 9.996910707875136e-05,
"loss": 2.9095,
"step": 530
},
{
"epoch": 0.01,
"grad_norm": 1.2323592901229858,
"learning_rate": 9.996838453790784e-05,
"loss": 2.908,
"step": 535
},
{
"epoch": 0.01,
"grad_norm": 1.3900178670883179,
"learning_rate": 9.996765364753302e-05,
"loss": 2.9906,
"step": 540
},
{
"epoch": 0.01,
"grad_norm": 1.2697597742080688,
"learning_rate": 9.996691440774906e-05,
"loss": 2.9726,
"step": 545
},
{
"epoch": 0.01,
"grad_norm": 1.8349788188934326,
"learning_rate": 9.996616681867945e-05,
"loss": 2.9081,
"step": 550
},
{
"epoch": 0.01,
"grad_norm": 1.2968906164169312,
"learning_rate": 9.996541088044913e-05,
"loss": 2.9679,
"step": 555
},
{
"epoch": 0.01,
"grad_norm": 1.2811602354049683,
"learning_rate": 9.996464659318442e-05,
"loss": 2.9334,
"step": 560
},
{
"epoch": 0.01,
"grad_norm": 1.3253918886184692,
"learning_rate": 9.996387395701302e-05,
"loss": 2.9377,
"step": 565
},
{
"epoch": 0.01,
"grad_norm": 1.3311257362365723,
"learning_rate": 9.996309297206403e-05,
"loss": 2.8969,
"step": 570
},
{
"epoch": 0.01,
"grad_norm": 2.182140350341797,
"learning_rate": 9.996230363846797e-05,
"loss": 2.9682,
"step": 575
},
{
"epoch": 0.02,
"grad_norm": 1.3210792541503906,
"learning_rate": 9.996150595635671e-05,
"loss": 2.8781,
"step": 580
},
{
"epoch": 0.02,
"grad_norm": 1.3530970811843872,
"learning_rate": 9.996069992586355e-05,
"loss": 2.9088,
"step": 585
},
{
"epoch": 0.02,
"grad_norm": 1.6767858266830444,
"learning_rate": 9.995988554712318e-05,
"loss": 2.9538,
"step": 590
},
{
"epoch": 0.02,
"grad_norm": 1.2738933563232422,
"learning_rate": 9.995906282027166e-05,
"loss": 2.9646,
"step": 595
},
{
"epoch": 0.02,
"grad_norm": 1.364833950996399,
"learning_rate": 9.995823174544649e-05,
"loss": 2.9113,
"step": 600
},
{
"epoch": 0.02,
"grad_norm": 1.3185209035873413,
"learning_rate": 9.995739232278654e-05,
"loss": 2.9173,
"step": 605
},
{
"epoch": 0.02,
"grad_norm": 1.2443362474441528,
"learning_rate": 9.995654455243204e-05,
"loss": 2.9068,
"step": 610
},
{
"epoch": 0.02,
"grad_norm": 1.3306407928466797,
"learning_rate": 9.995568843452468e-05,
"loss": 2.9178,
"step": 615
},
{
"epoch": 0.02,
"grad_norm": 1.3771353960037231,
"learning_rate": 9.99548239692075e-05,
"loss": 2.8165,
"step": 620
},
{
"epoch": 0.02,
"grad_norm": 1.243173599243164,
"learning_rate": 9.995395115662498e-05,
"loss": 2.9069,
"step": 625
},
{
"epoch": 0.02,
"grad_norm": 1.3304129838943481,
"learning_rate": 9.995306999692293e-05,
"loss": 2.9039,
"step": 630
},
{
"epoch": 0.02,
"grad_norm": 1.2537018060684204,
"learning_rate": 9.995218049024858e-05,
"loss": 2.814,
"step": 635
},
{
"epoch": 0.02,
"grad_norm": 1.2215917110443115,
"learning_rate": 9.995128263675058e-05,
"loss": 2.8736,
"step": 640
},
{
"epoch": 0.02,
"grad_norm": 1.3328001499176025,
"learning_rate": 9.995037643657899e-05,
"loss": 2.8717,
"step": 645
},
{
"epoch": 0.02,
"grad_norm": 1.3177193403244019,
"learning_rate": 9.994946188988516e-05,
"loss": 2.9266,
"step": 650
},
{
"epoch": 0.02,
"grad_norm": 1.613402247428894,
"learning_rate": 9.994853899682197e-05,
"loss": 2.8758,
"step": 655
},
{
"epoch": 0.02,
"grad_norm": 1.456215739250183,
"learning_rate": 9.994760775754362e-05,
"loss": 2.9231,
"step": 660
},
{
"epoch": 0.02,
"grad_norm": 1.2243021726608276,
"learning_rate": 9.994666817220571e-05,
"loss": 2.9141,
"step": 665
},
{
"epoch": 0.02,
"grad_norm": 1.4117437601089478,
"learning_rate": 9.994572024096521e-05,
"loss": 2.9128,
"step": 670
},
{
"epoch": 0.02,
"grad_norm": 1.2661588191986084,
"learning_rate": 9.994476396398055e-05,
"loss": 2.9202,
"step": 675
},
{
"epoch": 0.02,
"grad_norm": 1.3185731172561646,
"learning_rate": 9.994379934141153e-05,
"loss": 2.9442,
"step": 680
},
{
"epoch": 0.02,
"grad_norm": 1.2695297002792358,
"learning_rate": 9.994282637341932e-05,
"loss": 2.8368,
"step": 685
},
{
"epoch": 0.02,
"grad_norm": 1.2079784870147705,
"learning_rate": 9.994184506016651e-05,
"loss": 2.938,
"step": 690
},
{
"epoch": 0.02,
"grad_norm": 1.2556864023208618,
"learning_rate": 9.994085540181703e-05,
"loss": 2.8475,
"step": 695
},
{
"epoch": 0.02,
"grad_norm": 1.3012479543685913,
"learning_rate": 9.99398573985363e-05,
"loss": 2.9277,
"step": 700
},
{
"epoch": 0.02,
"grad_norm": 1.1829986572265625,
"learning_rate": 9.993885105049107e-05,
"loss": 2.8959,
"step": 705
},
{
"epoch": 0.02,
"grad_norm": 1.2905986309051514,
"learning_rate": 9.993783635784947e-05,
"loss": 2.8247,
"step": 710
},
{
"epoch": 0.02,
"grad_norm": 1.2108914852142334,
"learning_rate": 9.993681332078108e-05,
"loss": 2.8976,
"step": 715
},
{
"epoch": 0.02,
"grad_norm": 1.1731071472167969,
"learning_rate": 9.993578193945685e-05,
"loss": 2.8187,
"step": 720
},
{
"epoch": 0.02,
"grad_norm": 1.2230644226074219,
"learning_rate": 9.99347422140491e-05,
"loss": 2.8944,
"step": 725
},
{
"epoch": 0.02,
"grad_norm": 1.295494794845581,
"learning_rate": 9.993369414473157e-05,
"loss": 2.8973,
"step": 730
},
{
"epoch": 0.02,
"grad_norm": 1.233834981918335,
"learning_rate": 9.993263773167939e-05,
"loss": 2.8503,
"step": 735
},
{
"epoch": 0.02,
"grad_norm": 1.2504172325134277,
"learning_rate": 9.993157297506907e-05,
"loss": 2.831,
"step": 740
},
{
"epoch": 0.02,
"grad_norm": 1.1807881593704224,
"learning_rate": 9.993049987507856e-05,
"loss": 2.8343,
"step": 745
},
{
"epoch": 0.02,
"grad_norm": 1.1777859926223755,
"learning_rate": 9.992941843188715e-05,
"loss": 2.8423,
"step": 750
},
{
"epoch": 0.02,
"grad_norm": 1.2484335899353027,
"learning_rate": 9.992832864567554e-05,
"loss": 2.8502,
"step": 755
},
{
"epoch": 0.02,
"grad_norm": 1.202521562576294,
"learning_rate": 9.992723051662584e-05,
"loss": 2.8668,
"step": 760
},
{
"epoch": 0.02,
"grad_norm": 1.27983820438385,
"learning_rate": 9.992612404492153e-05,
"loss": 2.7498,
"step": 765
},
{
"epoch": 0.02,
"grad_norm": 1.2180736064910889,
"learning_rate": 9.992500923074752e-05,
"loss": 2.858,
"step": 770
},
{
"epoch": 0.02,
"grad_norm": 1.239127516746521,
"learning_rate": 9.992388607429006e-05,
"loss": 2.8574,
"step": 775
},
{
"epoch": 0.02,
"grad_norm": 1.2120908498764038,
"learning_rate": 9.992275457573685e-05,
"loss": 2.8604,
"step": 780
},
{
"epoch": 0.02,
"grad_norm": 1.390487551689148,
"learning_rate": 9.992161473527695e-05,
"loss": 2.832,
"step": 785
},
{
"epoch": 0.02,
"grad_norm": 1.1785805225372314,
"learning_rate": 9.992046655310083e-05,
"loss": 2.8745,
"step": 790
},
{
"epoch": 0.02,
"grad_norm": 1.1167683601379395,
"learning_rate": 9.991931002940034e-05,
"loss": 2.8348,
"step": 795
},
{
"epoch": 0.02,
"grad_norm": 1.2723276615142822,
"learning_rate": 9.991814516436873e-05,
"loss": 2.7676,
"step": 800
},
{
"epoch": 0.02,
"grad_norm": 1.2031564712524414,
"learning_rate": 9.991697195820065e-05,
"loss": 2.8553,
"step": 805
},
{
"epoch": 0.02,
"grad_norm": 1.2578195333480835,
"learning_rate": 9.991579041109212e-05,
"loss": 2.7428,
"step": 810
},
{
"epoch": 0.02,
"grad_norm": 1.2190525531768799,
"learning_rate": 9.991460052324061e-05,
"loss": 2.8459,
"step": 815
},
{
"epoch": 0.02,
"grad_norm": 1.3070708513259888,
"learning_rate": 9.991340229484491e-05,
"loss": 2.8011,
"step": 820
},
{
"epoch": 0.02,
"grad_norm": 1.2067310810089111,
"learning_rate": 9.991219572610526e-05,
"loss": 2.8396,
"step": 825
},
{
"epoch": 0.02,
"grad_norm": 1.126727819442749,
"learning_rate": 9.991098081722325e-05,
"loss": 2.7219,
"step": 830
},
{
"epoch": 0.02,
"grad_norm": 1.1925138235092163,
"learning_rate": 9.990975756840193e-05,
"loss": 2.8581,
"step": 835
},
{
"epoch": 0.02,
"grad_norm": 1.2076767683029175,
"learning_rate": 9.990852597984566e-05,
"loss": 2.814,
"step": 840
},
{
"epoch": 0.02,
"grad_norm": 1.14447820186615,
"learning_rate": 9.990728605176025e-05,
"loss": 2.8177,
"step": 845
},
{
"epoch": 0.02,
"grad_norm": 1.2271071672439575,
"learning_rate": 9.990603778435288e-05,
"loss": 2.78,
"step": 850
},
{
"epoch": 0.02,
"grad_norm": 1.1659224033355713,
"learning_rate": 9.990478117783214e-05,
"loss": 2.7965,
"step": 855
},
{
"epoch": 0.02,
"grad_norm": 1.1322077512741089,
"learning_rate": 9.990351623240799e-05,
"loss": 2.8132,
"step": 860
},
{
"epoch": 0.02,
"grad_norm": 1.1617186069488525,
"learning_rate": 9.990224294829181e-05,
"loss": 2.8062,
"step": 865
},
{
"epoch": 0.02,
"grad_norm": 1.157658576965332,
"learning_rate": 9.990096132569636e-05,
"loss": 2.825,
"step": 870
},
{
"epoch": 0.02,
"grad_norm": 1.1994798183441162,
"learning_rate": 9.98996713648358e-05,
"loss": 2.7718,
"step": 875
},
{
"epoch": 0.02,
"grad_norm": 1.1695261001586914,
"learning_rate": 9.989837306592567e-05,
"loss": 2.7275,
"step": 880
},
{
"epoch": 0.02,
"grad_norm": 1.1915322542190552,
"learning_rate": 9.989706642918291e-05,
"loss": 2.8258,
"step": 885
},
{
"epoch": 0.02,
"grad_norm": 1.2338296175003052,
"learning_rate": 9.989575145482583e-05,
"loss": 2.8189,
"step": 890
},
{
"epoch": 0.02,
"grad_norm": 1.2801344394683838,
"learning_rate": 9.989442814307422e-05,
"loss": 2.8105,
"step": 895
},
{
"epoch": 0.02,
"grad_norm": 1.1882652044296265,
"learning_rate": 9.989309649414914e-05,
"loss": 2.7951,
"step": 900
},
{
"epoch": 0.02,
"grad_norm": 1.1454718112945557,
"learning_rate": 9.989175650827314e-05,
"loss": 2.7869,
"step": 905
},
{
"epoch": 0.02,
"grad_norm": 1.2468279600143433,
"learning_rate": 9.98904081856701e-05,
"loss": 2.9166,
"step": 910
},
{
"epoch": 0.02,
"grad_norm": 1.1556510925292969,
"learning_rate": 9.988905152656534e-05,
"loss": 2.8006,
"step": 915
},
{
"epoch": 0.02,
"grad_norm": 1.1373000144958496,
"learning_rate": 9.988768653118554e-05,
"loss": 2.7448,
"step": 920
},
{
"epoch": 0.02,
"grad_norm": 1.1404756307601929,
"learning_rate": 9.988631319975881e-05,
"loss": 2.8032,
"step": 925
},
{
"epoch": 0.02,
"grad_norm": 1.2063539028167725,
"learning_rate": 9.988493153251459e-05,
"loss": 2.7408,
"step": 930
},
{
"epoch": 0.02,
"grad_norm": 1.164697289466858,
"learning_rate": 9.988354152968379e-05,
"loss": 2.7275,
"step": 935
},
{
"epoch": 0.02,
"grad_norm": 1.2604222297668457,
"learning_rate": 9.988214319149865e-05,
"loss": 2.771,
"step": 940
},
{
"epoch": 0.02,
"grad_norm": 1.2175700664520264,
"learning_rate": 9.988073651819281e-05,
"loss": 2.7873,
"step": 945
},
{
"epoch": 0.02,
"grad_norm": 1.3100277185440063,
"learning_rate": 9.987932151000137e-05,
"loss": 2.7259,
"step": 950
},
{
"epoch": 0.02,
"grad_norm": 1.1838854551315308,
"learning_rate": 9.987789816716073e-05,
"loss": 2.7419,
"step": 955
},
{
"epoch": 0.02,
"grad_norm": 1.1172375679016113,
"learning_rate": 9.987646648990876e-05,
"loss": 2.7742,
"step": 960
},
{
"epoch": 0.03,
"grad_norm": 1.2150076627731323,
"learning_rate": 9.987502647848466e-05,
"loss": 2.8085,
"step": 965
},
{
"epoch": 0.03,
"grad_norm": 1.1260734796524048,
"learning_rate": 9.987357813312905e-05,
"loss": 2.8057,
"step": 970
},
{
"epoch": 0.03,
"grad_norm": 1.1067372560501099,
"learning_rate": 9.987212145408395e-05,
"loss": 2.6707,
"step": 975
},
{
"epoch": 0.03,
"grad_norm": 1.1513421535491943,
"learning_rate": 9.987065644159278e-05,
"loss": 2.8202,
"step": 980
},
{
"epoch": 0.03,
"grad_norm": 1.1357698440551758,
"learning_rate": 9.986918309590031e-05,
"loss": 2.7249,
"step": 985
},
{
"epoch": 0.03,
"grad_norm": 1.2032572031021118,
"learning_rate": 9.986770141725277e-05,
"loss": 2.7824,
"step": 990
},
{
"epoch": 0.03,
"grad_norm": 1.1099467277526855,
"learning_rate": 9.986621140589772e-05,
"loss": 2.7022,
"step": 995
},
{
"epoch": 0.03,
"grad_norm": 1.1485451459884644,
"learning_rate": 9.986471306208412e-05,
"loss": 2.7115,
"step": 1000
},
{
"epoch": 0.03,
"grad_norm": 1.0639472007751465,
"learning_rate": 9.986320638606235e-05,
"loss": 2.8501,
"step": 1005
},
{
"epoch": 0.03,
"grad_norm": 1.1850502490997314,
"learning_rate": 9.986169137808419e-05,
"loss": 2.7673,
"step": 1010
},
{
"epoch": 0.03,
"grad_norm": 1.1512163877487183,
"learning_rate": 9.986016803840275e-05,
"loss": 2.7822,
"step": 1015
},
{
"epoch": 0.03,
"grad_norm": 1.2584794759750366,
"learning_rate": 9.985863636727262e-05,
"loss": 2.8277,
"step": 1020
},
{
"epoch": 0.03,
"grad_norm": 1.2487725019454956,
"learning_rate": 9.985709636494971e-05,
"loss": 2.7301,
"step": 1025
},
{
"epoch": 0.03,
"grad_norm": 1.225159764289856,
"learning_rate": 9.985554803169134e-05,
"loss": 2.7067,
"step": 1030
},
{
"epoch": 0.03,
"grad_norm": 1.0685861110687256,
"learning_rate": 9.985399136775627e-05,
"loss": 2.7843,
"step": 1035
},
{
"epoch": 0.03,
"grad_norm": 2.2430877685546875,
"learning_rate": 9.985242637340458e-05,
"loss": 2.7641,
"step": 1040
},
{
"epoch": 0.03,
"grad_norm": 1.149694800376892,
"learning_rate": 9.985085304889778e-05,
"loss": 2.77,
"step": 1045
},
{
"epoch": 0.03,
"grad_norm": 1.190642237663269,
"learning_rate": 9.984927139449877e-05,
"loss": 2.7659,
"step": 1050
},
{
"epoch": 0.03,
"grad_norm": 1.118334174156189,
"learning_rate": 9.984768141047184e-05,
"loss": 2.7879,
"step": 1055
},
{
"epoch": 0.03,
"grad_norm": 1.1132348775863647,
"learning_rate": 9.984608309708266e-05,
"loss": 2.6713,
"step": 1060
},
{
"epoch": 0.03,
"grad_norm": 1.097541093826294,
"learning_rate": 9.984447645459831e-05,
"loss": 2.7828,
"step": 1065
},
{
"epoch": 0.03,
"grad_norm": 1.1258339881896973,
"learning_rate": 9.984286148328728e-05,
"loss": 2.6871,
"step": 1070
},
{
"epoch": 0.03,
"grad_norm": 1.249412178993225,
"learning_rate": 9.984123818341937e-05,
"loss": 2.7012,
"step": 1075
},
{
"epoch": 0.03,
"grad_norm": 1.095870018005371,
"learning_rate": 9.983960655526587e-05,
"loss": 2.7095,
"step": 1080
},
{
"epoch": 0.03,
"grad_norm": 1.1016465425491333,
"learning_rate": 9.983796659909941e-05,
"loss": 2.6782,
"step": 1085
},
{
"epoch": 0.03,
"grad_norm": 1.5790003538131714,
"learning_rate": 9.983631831519401e-05,
"loss": 2.7387,
"step": 1090
},
{
"epoch": 0.03,
"grad_norm": 1.1821959018707275,
"learning_rate": 9.983466170382511e-05,
"loss": 2.7154,
"step": 1095
},
{
"epoch": 0.03,
"grad_norm": 1.2674894332885742,
"learning_rate": 9.983299676526949e-05,
"loss": 2.8079,
"step": 1100
},
{
"epoch": 0.03,
"grad_norm": 1.1285605430603027,
"learning_rate": 9.98313234998054e-05,
"loss": 2.7036,
"step": 1105
},
{
"epoch": 0.03,
"grad_norm": 1.2384120225906372,
"learning_rate": 9.982964190771242e-05,
"loss": 2.7555,
"step": 1110
},
{
"epoch": 0.03,
"grad_norm": 1.0747594833374023,
"learning_rate": 9.98279519892715e-05,
"loss": 2.7543,
"step": 1115
},
{
"epoch": 0.03,
"grad_norm": 1.082236647605896,
"learning_rate": 9.982625374476508e-05,
"loss": 2.6533,
"step": 1120
},
{
"epoch": 0.03,
"grad_norm": 1.2081153392791748,
"learning_rate": 9.98245471744769e-05,
"loss": 2.7916,
"step": 1125
},
{
"epoch": 0.03,
"grad_norm": 1.1127433776855469,
"learning_rate": 9.982283227869211e-05,
"loss": 2.7471,
"step": 1130
},
{
"epoch": 0.03,
"grad_norm": 1.0566129684448242,
"learning_rate": 9.982110905769729e-05,
"loss": 2.7246,
"step": 1135
},
{
"epoch": 0.03,
"grad_norm": 1.086361050605774,
"learning_rate": 9.981937751178037e-05,
"loss": 2.7524,
"step": 1140
},
{
"epoch": 0.03,
"grad_norm": 1.1175718307495117,
"learning_rate": 9.981763764123067e-05,
"loss": 2.7541,
"step": 1145
},
{
"epoch": 0.03,
"grad_norm": 1.0751663446426392,
"learning_rate": 9.981588944633893e-05,
"loss": 2.7445,
"step": 1150
},
{
"epoch": 0.03,
"grad_norm": 1.120162844657898,
"learning_rate": 9.981413292739727e-05,
"loss": 2.6706,
"step": 1155
},
{
"epoch": 0.03,
"grad_norm": 1.352687120437622,
"learning_rate": 9.981236808469922e-05,
"loss": 2.7248,
"step": 1160
},
{
"epoch": 0.03,
"grad_norm": 1.0683619976043701,
"learning_rate": 9.981059491853964e-05,
"loss": 2.6547,
"step": 1165
},
{
"epoch": 0.03,
"grad_norm": 1.2855300903320312,
"learning_rate": 9.980881342921482e-05,
"loss": 2.7148,
"step": 1170
},
{
"epoch": 0.03,
"grad_norm": 1.0404140949249268,
"learning_rate": 9.980702361702246e-05,
"loss": 2.7152,
"step": 1175
},
{
"epoch": 0.03,
"grad_norm": 1.2300519943237305,
"learning_rate": 9.980522548226162e-05,
"loss": 2.7274,
"step": 1180
},
{
"epoch": 0.03,
"grad_norm": 1.1546586751937866,
"learning_rate": 9.980341902523279e-05,
"loss": 2.731,
"step": 1185
},
{
"epoch": 0.03,
"grad_norm": 1.0849721431732178,
"learning_rate": 9.980160424623776e-05,
"loss": 2.7581,
"step": 1190
},
{
"epoch": 0.03,
"grad_norm": 1.2501624822616577,
"learning_rate": 9.979978114557984e-05,
"loss": 2.7452,
"step": 1195
},
{
"epoch": 0.03,
"grad_norm": 1.185105800628662,
"learning_rate": 9.979794972356362e-05,
"loss": 2.7429,
"step": 1200
},
{
"epoch": 0.03,
"grad_norm": 1.1640076637268066,
"learning_rate": 9.979610998049516e-05,
"loss": 2.6895,
"step": 1205
},
{
"epoch": 0.03,
"grad_norm": 1.132845163345337,
"learning_rate": 9.979426191668182e-05,
"loss": 2.6959,
"step": 1210
},
{
"epoch": 0.03,
"grad_norm": 1.0936648845672607,
"learning_rate": 9.979240553243246e-05,
"loss": 2.7422,
"step": 1215
},
{
"epoch": 0.03,
"grad_norm": 1.073645830154419,
"learning_rate": 9.979054082805726e-05,
"loss": 2.6922,
"step": 1220
},
{
"epoch": 0.03,
"grad_norm": 1.0646535158157349,
"learning_rate": 9.978866780386777e-05,
"loss": 2.7493,
"step": 1225
},
{
"epoch": 0.03,
"grad_norm": 1.0982900857925415,
"learning_rate": 9.978678646017701e-05,
"loss": 2.7078,
"step": 1230
},
{
"epoch": 0.03,
"grad_norm": 1.082937240600586,
"learning_rate": 9.978489679729934e-05,
"loss": 2.7498,
"step": 1235
},
{
"epoch": 0.03,
"grad_norm": 1.1154654026031494,
"learning_rate": 9.978299881555048e-05,
"loss": 2.7255,
"step": 1240
},
{
"epoch": 0.03,
"grad_norm": 1.087805151939392,
"learning_rate": 9.97810925152476e-05,
"loss": 2.712,
"step": 1245
},
{
"epoch": 0.03,
"grad_norm": 1.0542070865631104,
"learning_rate": 9.977917789670925e-05,
"loss": 2.7996,
"step": 1250
},
{
"epoch": 0.03,
"grad_norm": 1.0711954832077026,
"learning_rate": 9.977725496025535e-05,
"loss": 2.6916,
"step": 1255
},
{
"epoch": 0.03,
"grad_norm": 1.0700037479400635,
"learning_rate": 9.977532370620718e-05,
"loss": 2.7829,
"step": 1260
},
{
"epoch": 0.03,
"grad_norm": 1.1441583633422852,
"learning_rate": 9.977338413488748e-05,
"loss": 2.7586,
"step": 1265
},
{
"epoch": 0.03,
"grad_norm": 1.120880365371704,
"learning_rate": 9.977143624662035e-05,
"loss": 2.7901,
"step": 1270
},
{
"epoch": 0.03,
"grad_norm": 1.024007797241211,
"learning_rate": 9.976948004173125e-05,
"loss": 2.691,
"step": 1275
},
{
"epoch": 0.03,
"grad_norm": 1.030281662940979,
"learning_rate": 9.976751552054706e-05,
"loss": 2.6439,
"step": 1280
},
{
"epoch": 0.03,
"grad_norm": 1.0572916269302368,
"learning_rate": 9.976554268339607e-05,
"loss": 2.6485,
"step": 1285
},
{
"epoch": 0.03,
"grad_norm": 1.0543501377105713,
"learning_rate": 9.97635615306079e-05,
"loss": 2.7154,
"step": 1290
},
{
"epoch": 0.03,
"grad_norm": 1.255071997642517,
"learning_rate": 9.976157206251361e-05,
"loss": 2.6864,
"step": 1295
},
{
"epoch": 0.03,
"grad_norm": 1.1395295858383179,
"learning_rate": 9.975957427944563e-05,
"loss": 2.7154,
"step": 1300
},
{
"epoch": 0.03,
"grad_norm": 1.0557039976119995,
"learning_rate": 9.975756818173778e-05,
"loss": 2.6751,
"step": 1305
},
{
"epoch": 0.03,
"grad_norm": 1.1041921377182007,
"learning_rate": 9.975555376972529e-05,
"loss": 2.7196,
"step": 1310
},
{
"epoch": 0.03,
"grad_norm": 1.1137011051177979,
"learning_rate": 9.975353104374474e-05,
"loss": 2.7197,
"step": 1315
},
{
"epoch": 0.03,
"grad_norm": 1.1169722080230713,
"learning_rate": 9.975150000413411e-05,
"loss": 2.6881,
"step": 1320
},
{
"epoch": 0.03,
"grad_norm": 1.0851343870162964,
"learning_rate": 9.97494606512328e-05,
"loss": 2.7206,
"step": 1325
},
{
"epoch": 0.03,
"grad_norm": 1.1564364433288574,
"learning_rate": 9.974741298538158e-05,
"loss": 2.7584,
"step": 1330
},
{
"epoch": 0.03,
"grad_norm": 1.161625623703003,
"learning_rate": 9.974535700692257e-05,
"loss": 2.708,
"step": 1335
},
{
"epoch": 0.03,
"grad_norm": 1.105658769607544,
"learning_rate": 9.974329271619938e-05,
"loss": 2.6944,
"step": 1340
},
{
"epoch": 0.03,
"grad_norm": 1.138667106628418,
"learning_rate": 9.97412201135569e-05,
"loss": 2.7851,
"step": 1345
},
{
"epoch": 0.04,
"grad_norm": 1.0350295305252075,
"learning_rate": 9.973913919934147e-05,
"loss": 2.6217,
"step": 1350
},
{
"epoch": 0.04,
"grad_norm": 1.0782681703567505,
"learning_rate": 9.973704997390079e-05,
"loss": 2.694,
"step": 1355
},
{
"epoch": 0.04,
"grad_norm": 1.0481685400009155,
"learning_rate": 9.973495243758398e-05,
"loss": 2.7256,
"step": 1360
},
{
"epoch": 0.04,
"grad_norm": 1.1586180925369263,
"learning_rate": 9.973284659074151e-05,
"loss": 2.6731,
"step": 1365
},
{
"epoch": 0.04,
"grad_norm": 1.0626786947250366,
"learning_rate": 9.973073243372528e-05,
"loss": 2.5991,
"step": 1370
},
{
"epoch": 0.04,
"grad_norm": 1.0215508937835693,
"learning_rate": 9.972860996688853e-05,
"loss": 2.7054,
"step": 1375
},
{
"epoch": 0.04,
"grad_norm": 1.1395924091339111,
"learning_rate": 9.972647919058595e-05,
"loss": 2.7566,
"step": 1380
},
{
"epoch": 0.04,
"grad_norm": 1.0663546323776245,
"learning_rate": 9.972434010517358e-05,
"loss": 2.718,
"step": 1385
},
{
"epoch": 0.04,
"grad_norm": 1.0758408308029175,
"learning_rate": 9.972219271100882e-05,
"loss": 2.7177,
"step": 1390
},
{
"epoch": 0.04,
"grad_norm": 1.0529687404632568,
"learning_rate": 9.972003700845052e-05,
"loss": 2.7184,
"step": 1395
},
{
"epoch": 0.04,
"grad_norm": 1.0882738828659058,
"learning_rate": 9.971787299785888e-05,
"loss": 2.7458,
"step": 1400
},
{
"epoch": 0.04,
"grad_norm": 1.0540826320648193,
"learning_rate": 9.97157006795955e-05,
"loss": 2.7252,
"step": 1405
},
{
"epoch": 0.04,
"grad_norm": 1.0113714933395386,
"learning_rate": 9.971352005402338e-05,
"loss": 2.6128,
"step": 1410
},
{
"epoch": 0.04,
"grad_norm": 1.0579109191894531,
"learning_rate": 9.971133112150688e-05,
"loss": 2.6425,
"step": 1415
},
{
"epoch": 0.04,
"grad_norm": 1.0699785947799683,
"learning_rate": 9.970913388241177e-05,
"loss": 2.7385,
"step": 1420
},
{
"epoch": 0.04,
"grad_norm": 1.0873126983642578,
"learning_rate": 9.970692833710517e-05,
"loss": 2.6586,
"step": 1425
},
{
"epoch": 0.04,
"grad_norm": 1.055281162261963,
"learning_rate": 9.970471448595566e-05,
"loss": 2.6867,
"step": 1430
},
{
"epoch": 0.04,
"grad_norm": 1.0569781064987183,
"learning_rate": 9.970249232933315e-05,
"loss": 2.6978,
"step": 1435
},
{
"epoch": 0.04,
"grad_norm": 1.0210258960723877,
"learning_rate": 9.970026186760896e-05,
"loss": 2.6805,
"step": 1440
},
{
"epoch": 0.04,
"grad_norm": 1.04639732837677,
"learning_rate": 9.969802310115577e-05,
"loss": 2.6975,
"step": 1445
},
{
"epoch": 0.04,
"grad_norm": 1.0491751432418823,
"learning_rate": 9.96957760303477e-05,
"loss": 2.6556,
"step": 1450
},
{
"epoch": 0.04,
"grad_norm": 1.0513205528259277,
"learning_rate": 9.969352065556022e-05,
"loss": 2.6533,
"step": 1455
},
{
"epoch": 0.04,
"grad_norm": 1.0667250156402588,
"learning_rate": 9.969125697717017e-05,
"loss": 2.649,
"step": 1460
},
{
"epoch": 0.04,
"grad_norm": 0.9689574837684631,
"learning_rate": 9.968898499555582e-05,
"loss": 2.7475,
"step": 1465
},
{
"epoch": 0.04,
"grad_norm": 1.0673844814300537,
"learning_rate": 9.968670471109682e-05,
"loss": 2.7462,
"step": 1470
},
{
"epoch": 0.04,
"grad_norm": 1.357845664024353,
"learning_rate": 9.968441612417418e-05,
"loss": 2.6824,
"step": 1475
},
{
"epoch": 0.04,
"grad_norm": 0.9657119512557983,
"learning_rate": 9.968211923517032e-05,
"loss": 2.5873,
"step": 1480
},
{
"epoch": 0.04,
"grad_norm": 1.0901970863342285,
"learning_rate": 9.967981404446905e-05,
"loss": 2.668,
"step": 1485
},
{
"epoch": 0.04,
"grad_norm": 1.0867797136306763,
"learning_rate": 9.967750055245555e-05,
"loss": 2.7181,
"step": 1490
},
{
"epoch": 0.04,
"grad_norm": 1.000166416168213,
"learning_rate": 9.967517875951638e-05,
"loss": 2.6284,
"step": 1495
},
{
"epoch": 0.04,
"grad_norm": 1.076178789138794,
"learning_rate": 9.967284866603955e-05,
"loss": 2.5699,
"step": 1500
},
{
"epoch": 0.04,
"grad_norm": 1.0509229898452759,
"learning_rate": 9.967051027241436e-05,
"loss": 2.6827,
"step": 1505
},
{
"epoch": 0.04,
"grad_norm": 1.1289868354797363,
"learning_rate": 9.966816357903155e-05,
"loss": 2.6477,
"step": 1510
},
{
"epoch": 0.04,
"grad_norm": 1.0390571355819702,
"learning_rate": 9.96658085862833e-05,
"loss": 2.6951,
"step": 1515
},
{
"epoch": 0.04,
"grad_norm": 1.0986109972000122,
"learning_rate": 9.966344529456305e-05,
"loss": 2.6713,
"step": 1520
},
{
"epoch": 0.04,
"grad_norm": 1.070153832435608,
"learning_rate": 9.966107370426573e-05,
"loss": 2.6059,
"step": 1525
},
{
"epoch": 0.04,
"grad_norm": 1.0367134809494019,
"learning_rate": 9.965869381578765e-05,
"loss": 2.6963,
"step": 1530
},
{
"epoch": 0.04,
"grad_norm": 1.0132942199707031,
"learning_rate": 9.965630562952643e-05,
"loss": 2.626,
"step": 1535
},
{
"epoch": 0.04,
"grad_norm": 1.0611900091171265,
"learning_rate": 9.965390914588115e-05,
"loss": 2.7066,
"step": 1540
},
{
"epoch": 0.04,
"grad_norm": 1.0624221563339233,
"learning_rate": 9.965150436525225e-05,
"loss": 2.6816,
"step": 1545
},
{
"epoch": 0.04,
"grad_norm": 0.9357782602310181,
"learning_rate": 9.964909128804159e-05,
"loss": 2.6227,
"step": 1550
},
{
"epoch": 0.04,
"grad_norm": 0.9894802570343018,
"learning_rate": 9.964666991465234e-05,
"loss": 2.6206,
"step": 1555
},
{
"epoch": 0.04,
"grad_norm": 1.0641796588897705,
"learning_rate": 9.964424024548914e-05,
"loss": 2.7017,
"step": 1560
},
{
"epoch": 0.04,
"grad_norm": 1.05448317527771,
"learning_rate": 9.964180228095793e-05,
"loss": 2.6346,
"step": 1565
},
{
"epoch": 0.04,
"grad_norm": 1.0907766819000244,
"learning_rate": 9.963935602146612e-05,
"loss": 2.7377,
"step": 1570
},
{
"epoch": 0.04,
"grad_norm": 1.0578776597976685,
"learning_rate": 9.963690146742248e-05,
"loss": 2.692,
"step": 1575
},
{
"epoch": 0.04,
"grad_norm": 1.0521148443222046,
"learning_rate": 9.963443861923715e-05,
"loss": 2.6683,
"step": 1580
},
{
"epoch": 0.04,
"grad_norm": 0.9773673415184021,
"learning_rate": 9.963196747732164e-05,
"loss": 2.6175,
"step": 1585
},
{
"epoch": 0.04,
"grad_norm": 1.0065064430236816,
"learning_rate": 9.96294880420889e-05,
"loss": 2.7257,
"step": 1590
},
{
"epoch": 0.04,
"grad_norm": 1.0831990242004395,
"learning_rate": 9.96270003139532e-05,
"loss": 2.7237,
"step": 1595
},
{
"epoch": 0.04,
"grad_norm": 0.9820133447647095,
"learning_rate": 9.962450429333026e-05,
"loss": 2.6531,
"step": 1600
},
{
"epoch": 0.04,
"grad_norm": 1.020117163658142,
"learning_rate": 9.962199998063715e-05,
"loss": 2.6913,
"step": 1605
},
{
"epoch": 0.04,
"grad_norm": 1.0745240449905396,
"learning_rate": 9.961948737629231e-05,
"loss": 2.7018,
"step": 1610
},
{
"epoch": 0.04,
"grad_norm": 1.0728957653045654,
"learning_rate": 9.961696648071561e-05,
"loss": 2.6371,
"step": 1615
},
{
"epoch": 0.04,
"grad_norm": 1.0365451574325562,
"learning_rate": 9.961443729432828e-05,
"loss": 2.5295,
"step": 1620
},
{
"epoch": 0.04,
"grad_norm": 1.0331250429153442,
"learning_rate": 9.961189981755294e-05,
"loss": 2.6668,
"step": 1625
},
{
"epoch": 0.04,
"grad_norm": 1.0101702213287354,
"learning_rate": 9.960935405081357e-05,
"loss": 2.6852,
"step": 1630
},
{
"epoch": 0.04,
"grad_norm": 0.9997202157974243,
"learning_rate": 9.960679999453559e-05,
"loss": 2.6531,
"step": 1635
},
{
"epoch": 0.04,
"grad_norm": 0.998546838760376,
"learning_rate": 9.960423764914575e-05,
"loss": 2.6448,
"step": 1640
},
{
"epoch": 0.04,
"grad_norm": 1.0364885330200195,
"learning_rate": 9.960166701507221e-05,
"loss": 2.613,
"step": 1645
},
{
"epoch": 0.04,
"grad_norm": 1.0382848978042603,
"learning_rate": 9.959908809274453e-05,
"loss": 2.6476,
"step": 1650
},
{
"epoch": 0.04,
"grad_norm": 0.9934050440788269,
"learning_rate": 9.959650088259362e-05,
"loss": 2.6109,
"step": 1655
},
{
"epoch": 0.04,
"grad_norm": 1.0149688720703125,
"learning_rate": 9.959390538505181e-05,
"loss": 2.7287,
"step": 1660
},
{
"epoch": 0.04,
"grad_norm": 1.0174179077148438,
"learning_rate": 9.959130160055279e-05,
"loss": 2.6373,
"step": 1665
},
{
"epoch": 0.04,
"grad_norm": 1.1122881174087524,
"learning_rate": 9.958868952953163e-05,
"loss": 2.6561,
"step": 1670
},
{
"epoch": 0.04,
"grad_norm": 1.066658854484558,
"learning_rate": 9.95860691724248e-05,
"loss": 2.6799,
"step": 1675
},
{
"epoch": 0.04,
"grad_norm": 1.0403757095336914,
"learning_rate": 9.958344052967018e-05,
"loss": 2.7096,
"step": 1680
},
{
"epoch": 0.04,
"grad_norm": 1.0431218147277832,
"learning_rate": 9.958080360170698e-05,
"loss": 2.6243,
"step": 1685
},
{
"epoch": 0.04,
"grad_norm": 1.0609461069107056,
"learning_rate": 9.957815838897585e-05,
"loss": 2.6333,
"step": 1690
},
{
"epoch": 0.04,
"grad_norm": 1.0334645509719849,
"learning_rate": 9.957550489191874e-05,
"loss": 2.6283,
"step": 1695
},
{
"epoch": 0.04,
"grad_norm": 1.0082629919052124,
"learning_rate": 9.957284311097908e-05,
"loss": 2.6544,
"step": 1700
},
{
"epoch": 0.04,
"grad_norm": 1.0467709302902222,
"learning_rate": 9.957017304660164e-05,
"loss": 2.641,
"step": 1705
},
{
"epoch": 0.04,
"grad_norm": 1.0427700281143188,
"learning_rate": 9.956749469923258e-05,
"loss": 2.6282,
"step": 1710
},
{
"epoch": 0.04,
"grad_norm": 0.9910598397254944,
"learning_rate": 9.956480806931942e-05,
"loss": 2.6226,
"step": 1715
},
{
"epoch": 0.04,
"grad_norm": 1.0921964645385742,
"learning_rate": 9.956211315731112e-05,
"loss": 2.5359,
"step": 1720
},
{
"epoch": 0.04,
"grad_norm": 1.0623970031738281,
"learning_rate": 9.955940996365796e-05,
"loss": 2.6464,
"step": 1725
},
{
"epoch": 0.04,
"grad_norm": 1.0595535039901733,
"learning_rate": 9.955669848881166e-05,
"loss": 2.7279,
"step": 1730
},
{
"epoch": 0.05,
"grad_norm": 1.102903962135315,
"learning_rate": 9.955397873322527e-05,
"loss": 2.6894,
"step": 1735
},
{
"epoch": 0.05,
"grad_norm": 1.0501497983932495,
"learning_rate": 9.955125069735327e-05,
"loss": 2.5303,
"step": 1740
},
{
"epoch": 0.05,
"grad_norm": 1.0027271509170532,
"learning_rate": 9.95485143816515e-05,
"loss": 2.6508,
"step": 1745
},
{
"epoch": 0.05,
"grad_norm": 1.0246371030807495,
"learning_rate": 9.95457697865772e-05,
"loss": 2.6085,
"step": 1750
},
{
"epoch": 0.05,
"grad_norm": 1.0341800451278687,
"learning_rate": 9.954301691258894e-05,
"loss": 2.7007,
"step": 1755
},
{
"epoch": 0.05,
"grad_norm": 0.9934285879135132,
"learning_rate": 9.954025576014674e-05,
"loss": 2.5996,
"step": 1760
},
{
"epoch": 0.05,
"grad_norm": 1.0397307872772217,
"learning_rate": 9.953748632971201e-05,
"loss": 2.6274,
"step": 1765
},
{
"epoch": 0.05,
"grad_norm": 0.9581106305122375,
"learning_rate": 9.953470862174748e-05,
"loss": 2.6644,
"step": 1770
},
{
"epoch": 0.05,
"grad_norm": 1.018480658531189,
"learning_rate": 9.953192263671728e-05,
"loss": 2.599,
"step": 1775
},
{
"epoch": 0.05,
"grad_norm": 1.014201283454895,
"learning_rate": 9.952912837508697e-05,
"loss": 2.6223,
"step": 1780
},
{
"epoch": 0.05,
"grad_norm": 1.0199346542358398,
"learning_rate": 9.952632583732343e-05,
"loss": 2.6546,
"step": 1785
},
{
"epoch": 0.05,
"grad_norm": 1.0939984321594238,
"learning_rate": 9.952351502389498e-05,
"loss": 2.6142,
"step": 1790
},
{
"epoch": 0.05,
"grad_norm": 1.0241179466247559,
"learning_rate": 9.95206959352713e-05,
"loss": 2.6894,
"step": 1795
},
{
"epoch": 0.05,
"grad_norm": 0.976097822189331,
"learning_rate": 9.951786857192341e-05,
"loss": 2.6713,
"step": 1800
},
{
"epoch": 0.05,
"grad_norm": 0.9381155371665955,
"learning_rate": 9.951503293432378e-05,
"loss": 2.5481,
"step": 1805
},
{
"epoch": 0.05,
"grad_norm": 0.9621322751045227,
"learning_rate": 9.951218902294625e-05,
"loss": 2.5711,
"step": 1810
},
{
"epoch": 0.05,
"grad_norm": 0.9531965255737305,
"learning_rate": 9.9509336838266e-05,
"loss": 2.6877,
"step": 1815
},
{
"epoch": 0.05,
"grad_norm": 1.0202903747558594,
"learning_rate": 9.950647638075963e-05,
"loss": 2.6071,
"step": 1820
},
{
"epoch": 0.05,
"grad_norm": 1.4851258993148804,
"learning_rate": 9.950360765090511e-05,
"loss": 2.5837,
"step": 1825
},
{
"epoch": 0.05,
"grad_norm": 1.0187565088272095,
"learning_rate": 9.950073064918179e-05,
"loss": 2.5144,
"step": 1830
},
{
"epoch": 0.05,
"grad_norm": 0.9899535179138184,
"learning_rate": 9.949784537607042e-05,
"loss": 2.5933,
"step": 1835
},
{
"epoch": 0.05,
"grad_norm": 1.031090259552002,
"learning_rate": 9.949495183205311e-05,
"loss": 2.6084,
"step": 1840
},
{
"epoch": 0.05,
"grad_norm": 0.9576619267463684,
"learning_rate": 9.949205001761334e-05,
"loss": 2.6241,
"step": 1845
},
{
"epoch": 0.05,
"grad_norm": 1.0009676218032837,
"learning_rate": 9.948913993323603e-05,
"loss": 2.6472,
"step": 1850
},
{
"epoch": 0.05,
"grad_norm": 1.0144001245498657,
"learning_rate": 9.948622157940744e-05,
"loss": 2.5833,
"step": 1855
},
{
"epoch": 0.05,
"grad_norm": 0.9944317936897278,
"learning_rate": 9.948329495661517e-05,
"loss": 2.6445,
"step": 1860
},
{
"epoch": 0.05,
"grad_norm": 0.9442741274833679,
"learning_rate": 9.94803600653483e-05,
"loss": 2.6692,
"step": 1865
},
{
"epoch": 0.05,
"grad_norm": 0.9770987629890442,
"learning_rate": 9.947741690609722e-05,
"loss": 2.64,
"step": 1870
},
{
"epoch": 0.05,
"grad_norm": 0.9978055357933044,
"learning_rate": 9.94744654793537e-05,
"loss": 2.6514,
"step": 1875
},
{
"epoch": 0.05,
"grad_norm": 0.9885908365249634,
"learning_rate": 9.947150578561095e-05,
"loss": 2.614,
"step": 1880
},
{
"epoch": 0.05,
"grad_norm": 1.0056266784667969,
"learning_rate": 9.946853782536351e-05,
"loss": 2.6684,
"step": 1885
},
{
"epoch": 0.05,
"grad_norm": 0.9736636281013489,
"learning_rate": 9.946556159910732e-05,
"loss": 2.6202,
"step": 1890
},
{
"epoch": 0.05,
"grad_norm": 0.9233741164207458,
"learning_rate": 9.946257710733966e-05,
"loss": 2.5777,
"step": 1895
},
{
"epoch": 0.05,
"grad_norm": 0.9845104813575745,
"learning_rate": 9.945958435055927e-05,
"loss": 2.6377,
"step": 1900
},
{
"epoch": 0.05,
"grad_norm": 0.9074547290802002,
"learning_rate": 9.945658332926622e-05,
"loss": 2.4892,
"step": 1905
},
{
"epoch": 0.05,
"grad_norm": 1.0190004110336304,
"learning_rate": 9.945357404396197e-05,
"loss": 2.6428,
"step": 1910
},
{
"epoch": 0.05,
"grad_norm": 0.90732342004776,
"learning_rate": 9.945055649514934e-05,
"loss": 2.5557,
"step": 1915
},
{
"epoch": 0.05,
"grad_norm": 0.9683628082275391,
"learning_rate": 9.944753068333256e-05,
"loss": 2.5517,
"step": 1920
},
{
"epoch": 0.05,
"grad_norm": 1.0072124004364014,
"learning_rate": 9.944449660901725e-05,
"loss": 2.6563,
"step": 1925
},
{
"epoch": 0.05,
"grad_norm": 1.024245023727417,
"learning_rate": 9.944145427271036e-05,
"loss": 2.6648,
"step": 1930
},
{
"epoch": 0.05,
"grad_norm": 0.953263521194458,
"learning_rate": 9.94384036749203e-05,
"loss": 2.5581,
"step": 1935
},
{
"epoch": 0.05,
"grad_norm": 0.9210425019264221,
"learning_rate": 9.943534481615677e-05,
"loss": 2.5944,
"step": 1940
},
{
"epoch": 0.05,
"grad_norm": 1.1062475442886353,
"learning_rate": 9.943227769693091e-05,
"loss": 2.6189,
"step": 1945
},
{
"epoch": 0.05,
"grad_norm": 0.9907572865486145,
"learning_rate": 9.942920231775524e-05,
"loss": 2.5816,
"step": 1950
},
{
"epoch": 0.05,
"grad_norm": 1.0144506692886353,
"learning_rate": 9.942611867914363e-05,
"loss": 2.5439,
"step": 1955
},
{
"epoch": 0.05,
"grad_norm": 0.982017993927002,
"learning_rate": 9.942302678161133e-05,
"loss": 2.6889,
"step": 1960
},
{
"epoch": 0.05,
"grad_norm": 0.9979785680770874,
"learning_rate": 9.9419926625675e-05,
"loss": 2.6586,
"step": 1965
},
{
"epoch": 0.05,
"grad_norm": 0.9756157994270325,
"learning_rate": 9.941681821185265e-05,
"loss": 2.5679,
"step": 1970
},
{
"epoch": 0.05,
"grad_norm": 0.9914107918739319,
"learning_rate": 9.941370154066371e-05,
"loss": 2.6007,
"step": 1975
},
{
"epoch": 0.05,
"grad_norm": 1.0042197704315186,
"learning_rate": 9.941057661262895e-05,
"loss": 2.5106,
"step": 1980
},
{
"epoch": 0.05,
"grad_norm": 0.9883943796157837,
"learning_rate": 9.940744342827055e-05,
"loss": 2.6583,
"step": 1985
},
{
"epoch": 0.05,
"grad_norm": 0.9845190048217773,
"learning_rate": 9.940430198811202e-05,
"loss": 2.5951,
"step": 1990
},
{
"epoch": 0.05,
"grad_norm": 0.9730456471443176,
"learning_rate": 9.940115229267831e-05,
"loss": 2.5845,
"step": 1995
},
{
"epoch": 0.05,
"grad_norm": 0.9681295156478882,
"learning_rate": 9.939799434249571e-05,
"loss": 2.6608,
"step": 2000
},
{
"epoch": 0.05,
"grad_norm": 0.9918345212936401,
"learning_rate": 9.93948281380919e-05,
"loss": 2.6166,
"step": 2005
},
{
"epoch": 0.05,
"grad_norm": 0.9488073587417603,
"learning_rate": 9.939165367999597e-05,
"loss": 2.5195,
"step": 2010
},
{
"epoch": 0.05,
"grad_norm": 0.9921746253967285,
"learning_rate": 9.938847096873831e-05,
"loss": 2.63,
"step": 2015
},
{
"epoch": 0.05,
"grad_norm": 0.9088597297668457,
"learning_rate": 9.938528000485078e-05,
"loss": 2.5493,
"step": 2020
},
{
"epoch": 0.05,
"grad_norm": 0.9646790027618408,
"learning_rate": 9.938208078886655e-05,
"loss": 2.5548,
"step": 2025
},
{
"epoch": 0.05,
"grad_norm": 1.0113939046859741,
"learning_rate": 9.937887332132023e-05,
"loss": 2.6616,
"step": 2030
},
{
"epoch": 0.05,
"grad_norm": 0.9548825025558472,
"learning_rate": 9.937565760274776e-05,
"loss": 2.58,
"step": 2035
},
{
"epoch": 0.05,
"grad_norm": 1.022743821144104,
"learning_rate": 9.937243363368645e-05,
"loss": 2.6351,
"step": 2040
},
{
"epoch": 0.05,
"grad_norm": 0.9962561130523682,
"learning_rate": 9.936920141467505e-05,
"loss": 2.6268,
"step": 2045
},
{
"epoch": 0.05,
"grad_norm": 0.9401412606239319,
"learning_rate": 9.936596094625364e-05,
"loss": 2.5574,
"step": 2050
},
{
"epoch": 0.05,
"grad_norm": 1.0102975368499756,
"learning_rate": 9.936271222896367e-05,
"loss": 2.6472,
"step": 2055
},
{
"epoch": 0.05,
"grad_norm": 1.051788330078125,
"learning_rate": 9.935945526334803e-05,
"loss": 2.5647,
"step": 2060
},
{
"epoch": 0.05,
"grad_norm": 0.9676756262779236,
"learning_rate": 9.935619004995089e-05,
"loss": 2.5533,
"step": 2065
},
{
"epoch": 0.05,
"grad_norm": 0.9770815968513489,
"learning_rate": 9.935291658931791e-05,
"loss": 2.573,
"step": 2070
},
{
"epoch": 0.05,
"grad_norm": 0.9599032402038574,
"learning_rate": 9.934963488199604e-05,
"loss": 2.5187,
"step": 2075
},
{
"epoch": 0.05,
"grad_norm": 1.0534309148788452,
"learning_rate": 9.934634492853366e-05,
"loss": 2.679,
"step": 2080
},
{
"epoch": 0.05,
"grad_norm": 0.9907177090644836,
"learning_rate": 9.93430467294805e-05,
"loss": 2.6515,
"step": 2085
},
{
"epoch": 0.05,
"grad_norm": 0.9733906388282776,
"learning_rate": 9.933974028538768e-05,
"loss": 2.5326,
"step": 2090
},
{
"epoch": 0.05,
"grad_norm": 0.986055850982666,
"learning_rate": 9.933642559680769e-05,
"loss": 2.5514,
"step": 2095
},
{
"epoch": 0.05,
"grad_norm": 0.9827683568000793,
"learning_rate": 9.93331026642944e-05,
"loss": 2.5879,
"step": 2100
},
{
"epoch": 0.05,
"grad_norm": 0.9323362708091736,
"learning_rate": 9.932977148840307e-05,
"loss": 2.6258,
"step": 2105
},
{
"epoch": 0.05,
"grad_norm": 0.9352030158042908,
"learning_rate": 9.93264320696903e-05,
"loss": 2.5316,
"step": 2110
},
{
"epoch": 0.05,
"grad_norm": 0.9449790120124817,
"learning_rate": 9.932308440871413e-05,
"loss": 2.6239,
"step": 2115
},
{
"epoch": 0.06,
"grad_norm": 0.9402400255203247,
"learning_rate": 9.931972850603394e-05,
"loss": 2.5916,
"step": 2120
},
{
"epoch": 0.06,
"grad_norm": 0.9750306606292725,
"learning_rate": 9.931636436221048e-05,
"loss": 2.5318,
"step": 2125
},
{
"epoch": 0.06,
"grad_norm": 0.9988149404525757,
"learning_rate": 9.931299197780586e-05,
"loss": 2.4978,
"step": 2130
},
{
"epoch": 0.06,
"grad_norm": 0.9912875890731812,
"learning_rate": 9.930961135338361e-05,
"loss": 2.5713,
"step": 2135
},
{
"epoch": 0.06,
"grad_norm": 0.9559140801429749,
"learning_rate": 9.930622248950864e-05,
"loss": 2.5411,
"step": 2140
},
{
"epoch": 0.06,
"grad_norm": 0.9317615032196045,
"learning_rate": 9.93028253867472e-05,
"loss": 2.4905,
"step": 2145
},
{
"epoch": 0.06,
"grad_norm": 0.9385008811950684,
"learning_rate": 9.929942004566695e-05,
"loss": 2.4653,
"step": 2150
},
{
"epoch": 0.06,
"grad_norm": 0.9230170249938965,
"learning_rate": 9.929600646683686e-05,
"loss": 2.5299,
"step": 2155
},
{
"epoch": 0.06,
"grad_norm": 0.9593597650527954,
"learning_rate": 9.929258465082738e-05,
"loss": 2.6441,
"step": 2160
},
{
"epoch": 0.06,
"grad_norm": 1.0900110006332397,
"learning_rate": 9.928915459821027e-05,
"loss": 2.5509,
"step": 2165
},
{
"epoch": 0.06,
"grad_norm": 0.9563053846359253,
"learning_rate": 9.928571630955865e-05,
"loss": 2.519,
"step": 2170
},
{
"epoch": 0.06,
"grad_norm": 1.0250791311264038,
"learning_rate": 9.928226978544708e-05,
"loss": 2.5907,
"step": 2175
},
{
"epoch": 0.06,
"grad_norm": 0.9980397820472717,
"learning_rate": 9.927881502645146e-05,
"loss": 2.5172,
"step": 2180
},
{
"epoch": 0.06,
"grad_norm": 0.9734475016593933,
"learning_rate": 9.927535203314904e-05,
"loss": 2.6015,
"step": 2185
},
{
"epoch": 0.06,
"grad_norm": 0.9620555639266968,
"learning_rate": 9.927188080611849e-05,
"loss": 2.6088,
"step": 2190
},
{
"epoch": 0.06,
"grad_norm": 0.9745156168937683,
"learning_rate": 9.926840134593984e-05,
"loss": 2.5507,
"step": 2195
},
{
"epoch": 0.06,
"grad_norm": 1.027012586593628,
"learning_rate": 9.92649136531945e-05,
"loss": 2.6308,
"step": 2200
},
{
"epoch": 0.06,
"grad_norm": 1.0029487609863281,
"learning_rate": 9.926141772846525e-05,
"loss": 2.587,
"step": 2205
},
{
"epoch": 0.06,
"grad_norm": 1.0114673376083374,
"learning_rate": 9.925791357233623e-05,
"loss": 2.5719,
"step": 2210
},
{
"epoch": 0.06,
"grad_norm": 0.9250192642211914,
"learning_rate": 9.925440118539298e-05,
"loss": 2.5899,
"step": 2215
},
{
"epoch": 0.06,
"grad_norm": 0.9559205770492554,
"learning_rate": 9.925088056822241e-05,
"loss": 2.5606,
"step": 2220
},
{
"epoch": 0.06,
"grad_norm": 0.9945977330207825,
"learning_rate": 9.924735172141281e-05,
"loss": 2.6745,
"step": 2225
},
{
"epoch": 0.06,
"grad_norm": 0.9587653279304504,
"learning_rate": 9.924381464555381e-05,
"loss": 2.6787,
"step": 2230
},
{
"epoch": 0.06,
"grad_norm": 0.9735569357872009,
"learning_rate": 9.924026934123647e-05,
"loss": 2.5446,
"step": 2235
},
{
"epoch": 0.06,
"grad_norm": 0.9250222444534302,
"learning_rate": 9.92367158090532e-05,
"loss": 2.5926,
"step": 2240
},
{
"epoch": 0.06,
"grad_norm": 1.0071675777435303,
"learning_rate": 9.923315404959775e-05,
"loss": 2.5449,
"step": 2245
},
{
"epoch": 0.06,
"grad_norm": 1.00709068775177,
"learning_rate": 9.92295840634653e-05,
"loss": 2.6383,
"step": 2250
},
{
"epoch": 0.06,
"grad_norm": 1.0009005069732666,
"learning_rate": 9.922600585125237e-05,
"loss": 2.6007,
"step": 2255
},
{
"epoch": 0.06,
"grad_norm": 0.9267144203186035,
"learning_rate": 9.922241941355688e-05,
"loss": 2.5785,
"step": 2260
},
{
"epoch": 0.06,
"grad_norm": 0.9651312828063965,
"learning_rate": 9.921882475097811e-05,
"loss": 2.6067,
"step": 2265
},
{
"epoch": 0.06,
"grad_norm": 0.9292260408401489,
"learning_rate": 9.921522186411671e-05,
"loss": 2.6059,
"step": 2270
},
{
"epoch": 0.06,
"grad_norm": 0.9775951504707336,
"learning_rate": 9.92116107535747e-05,
"loss": 2.5775,
"step": 2275
},
{
"epoch": 0.06,
"grad_norm": 0.9640180468559265,
"learning_rate": 9.920799141995549e-05,
"loss": 2.5632,
"step": 2280
},
{
"epoch": 0.06,
"grad_norm": 1.0346662998199463,
"learning_rate": 9.920436386386387e-05,
"loss": 2.564,
"step": 2285
},
{
"epoch": 0.06,
"grad_norm": 0.9629662036895752,
"learning_rate": 9.920072808590596e-05,
"loss": 2.5818,
"step": 2290
},
{
"epoch": 0.06,
"grad_norm": 0.9469580054283142,
"learning_rate": 9.91970840866893e-05,
"loss": 2.5799,
"step": 2295
},
{
"epoch": 0.06,
"grad_norm": 0.9726023077964783,
"learning_rate": 9.919343186682282e-05,
"loss": 2.5519,
"step": 2300
},
{
"epoch": 0.06,
"grad_norm": 0.9784693121910095,
"learning_rate": 9.918977142691675e-05,
"loss": 2.6077,
"step": 2305
},
{
"epoch": 0.06,
"grad_norm": 0.9587064981460571,
"learning_rate": 9.918610276758273e-05,
"loss": 2.5552,
"step": 2310
},
{
"epoch": 0.06,
"grad_norm": 0.9538026452064514,
"learning_rate": 9.918242588943383e-05,
"loss": 2.5475,
"step": 2315
},
{
"epoch": 0.06,
"grad_norm": 0.9905738830566406,
"learning_rate": 9.91787407930844e-05,
"loss": 2.5254,
"step": 2320
},
{
"epoch": 0.06,
"grad_norm": 0.9520058631896973,
"learning_rate": 9.917504747915023e-05,
"loss": 2.5783,
"step": 2325
},
{
"epoch": 0.06,
"grad_norm": 0.9458277821540833,
"learning_rate": 9.917134594824843e-05,
"loss": 2.6254,
"step": 2330
},
{
"epoch": 0.06,
"grad_norm": 0.9688752889633179,
"learning_rate": 9.916763620099754e-05,
"loss": 2.583,
"step": 2335
},
{
"epoch": 0.06,
"grad_norm": 1.0020638704299927,
"learning_rate": 9.916391823801743e-05,
"loss": 2.5378,
"step": 2340
},
{
"epoch": 0.06,
"grad_norm": 0.9028971195220947,
"learning_rate": 9.916019205992937e-05,
"loss": 2.5889,
"step": 2345
},
{
"epoch": 0.06,
"grad_norm": 0.9501419067382812,
"learning_rate": 9.915645766735597e-05,
"loss": 2.5432,
"step": 2350
},
{
"epoch": 0.06,
"grad_norm": 0.9561126828193665,
"learning_rate": 9.915271506092126e-05,
"loss": 2.5796,
"step": 2355
},
{
"epoch": 0.06,
"grad_norm": 0.9839569926261902,
"learning_rate": 9.91489642412506e-05,
"loss": 2.5911,
"step": 2360
},
{
"epoch": 0.06,
"grad_norm": 0.9461435675621033,
"learning_rate": 9.914520520897072e-05,
"loss": 2.5608,
"step": 2365
},
{
"epoch": 0.06,
"grad_norm": 0.9630075097084045,
"learning_rate": 9.914143796470978e-05,
"loss": 2.6011,
"step": 2370
},
{
"epoch": 0.06,
"grad_norm": 0.9186975955963135,
"learning_rate": 9.913766250909726e-05,
"loss": 2.4484,
"step": 2375
},
{
"epoch": 0.06,
"grad_norm": 0.9484267234802246,
"learning_rate": 9.9133878842764e-05,
"loss": 2.6058,
"step": 2380
},
{
"epoch": 0.06,
"grad_norm": 0.9428638219833374,
"learning_rate": 9.913008696634226e-05,
"loss": 2.5728,
"step": 2385
},
{
"epoch": 0.06,
"grad_norm": 0.9141082763671875,
"learning_rate": 9.912628688046564e-05,
"loss": 2.5281,
"step": 2390
},
{
"epoch": 0.06,
"grad_norm": 0.9666171669960022,
"learning_rate": 9.912247858576914e-05,
"loss": 2.5161,
"step": 2395
},
{
"epoch": 0.06,
"grad_norm": 0.9425268173217773,
"learning_rate": 9.911866208288906e-05,
"loss": 2.5178,
"step": 2400
},
{
"epoch": 0.06,
"grad_norm": 0.9157416820526123,
"learning_rate": 9.911483737246319e-05,
"loss": 2.569,
"step": 2405
},
{
"epoch": 0.06,
"grad_norm": 0.9921411871910095,
"learning_rate": 9.911100445513058e-05,
"loss": 2.5222,
"step": 2410
},
{
"epoch": 0.06,
"grad_norm": 0.9465234875679016,
"learning_rate": 9.910716333153169e-05,
"loss": 2.5532,
"step": 2415
},
{
"epoch": 0.06,
"grad_norm": 0.9045655131340027,
"learning_rate": 9.910331400230841e-05,
"loss": 2.5975,
"step": 2420
},
{
"epoch": 0.06,
"grad_norm": 0.9025008082389832,
"learning_rate": 9.90994564681039e-05,
"loss": 2.521,
"step": 2425
},
{
"epoch": 0.06,
"grad_norm": 0.9299361705780029,
"learning_rate": 9.909559072956275e-05,
"loss": 2.5503,
"step": 2430
},
{
"epoch": 0.06,
"grad_norm": 0.978778600692749,
"learning_rate": 9.909171678733091e-05,
"loss": 2.5419,
"step": 2435
},
{
"epoch": 0.06,
"grad_norm": 0.986125648021698,
"learning_rate": 9.908783464205573e-05,
"loss": 2.632,
"step": 2440
},
{
"epoch": 0.06,
"grad_norm": 0.9413051009178162,
"learning_rate": 9.908394429438585e-05,
"loss": 2.5642,
"step": 2445
},
{
"epoch": 0.06,
"grad_norm": 0.9189814925193787,
"learning_rate": 9.908004574497139e-05,
"loss": 2.5817,
"step": 2450
},
{
"epoch": 0.06,
"grad_norm": 0.9355887174606323,
"learning_rate": 9.907613899446372e-05,
"loss": 2.5239,
"step": 2455
},
{
"epoch": 0.06,
"grad_norm": 0.908140242099762,
"learning_rate": 9.90722240435157e-05,
"loss": 2.5891,
"step": 2460
},
{
"epoch": 0.06,
"grad_norm": 0.9359908103942871,
"learning_rate": 9.906830089278148e-05,
"loss": 2.562,
"step": 2465
},
{
"epoch": 0.06,
"grad_norm": 0.9403198957443237,
"learning_rate": 9.906436954291661e-05,
"loss": 2.6108,
"step": 2470
},
{
"epoch": 0.06,
"grad_norm": 0.9441424012184143,
"learning_rate": 9.906042999457798e-05,
"loss": 2.6204,
"step": 2475
},
{
"epoch": 0.06,
"grad_norm": 0.9558393359184265,
"learning_rate": 9.905648224842391e-05,
"loss": 2.5216,
"step": 2480
},
{
"epoch": 0.06,
"grad_norm": 0.9871098399162292,
"learning_rate": 9.905252630511405e-05,
"loss": 2.5241,
"step": 2485
},
{
"epoch": 0.06,
"grad_norm": 0.9927551746368408,
"learning_rate": 9.904856216530938e-05,
"loss": 2.5355,
"step": 2490
},
{
"epoch": 0.06,
"grad_norm": 0.9151021242141724,
"learning_rate": 9.904458982967235e-05,
"loss": 2.564,
"step": 2495
},
{
"epoch": 0.06,
"grad_norm": 0.9128942489624023,
"learning_rate": 9.904060929886668e-05,
"loss": 2.6145,
"step": 2500
},
{
"epoch": 0.07,
"grad_norm": 0.895881712436676,
"learning_rate": 9.903662057355751e-05,
"loss": 2.5112,
"step": 2505
},
{
"epoch": 0.07,
"grad_norm": 0.9313883185386658,
"learning_rate": 9.903262365441137e-05,
"loss": 2.4995,
"step": 2510
},
{
"epoch": 0.07,
"grad_norm": 0.9815142154693604,
"learning_rate": 9.90286185420961e-05,
"loss": 2.5393,
"step": 2515
},
{
"epoch": 0.07,
"grad_norm": 0.8782598376274109,
"learning_rate": 9.902460523728094e-05,
"loss": 2.5062,
"step": 2520
},
{
"epoch": 0.07,
"grad_norm": 0.9691426157951355,
"learning_rate": 9.902058374063651e-05,
"loss": 2.5584,
"step": 2525
},
{
"epoch": 0.07,
"grad_norm": 0.9340393543243408,
"learning_rate": 9.90165540528348e-05,
"loss": 2.517,
"step": 2530
},
{
"epoch": 0.07,
"grad_norm": 0.914356529712677,
"learning_rate": 9.901251617454913e-05,
"loss": 2.5457,
"step": 2535
},
{
"epoch": 0.07,
"grad_norm": 0.9028905630111694,
"learning_rate": 9.900847010645422e-05,
"loss": 2.4888,
"step": 2540
},
{
"epoch": 0.07,
"grad_norm": 0.9713951349258423,
"learning_rate": 9.900441584922616e-05,
"loss": 2.564,
"step": 2545
},
{
"epoch": 0.07,
"grad_norm": 0.9613834023475647,
"learning_rate": 9.90003534035424e-05,
"loss": 2.5052,
"step": 2550
},
{
"epoch": 0.07,
"grad_norm": 0.9060428738594055,
"learning_rate": 9.899628277008175e-05,
"loss": 2.5113,
"step": 2555
},
{
"epoch": 0.07,
"grad_norm": 0.9104215502738953,
"learning_rate": 9.899220394952441e-05,
"loss": 2.5277,
"step": 2560
},
{
"epoch": 0.07,
"grad_norm": 1.1357563734054565,
"learning_rate": 9.898811694255195e-05,
"loss": 2.554,
"step": 2565
},
{
"epoch": 0.07,
"grad_norm": 0.9427745342254639,
"learning_rate": 9.898402174984726e-05,
"loss": 2.51,
"step": 2570
},
{
"epoch": 0.07,
"grad_norm": 0.9076730012893677,
"learning_rate": 9.897991837209464e-05,
"loss": 2.5918,
"step": 2575
},
{
"epoch": 0.07,
"grad_norm": 0.8863185048103333,
"learning_rate": 9.897580680997977e-05,
"loss": 2.5269,
"step": 2580
},
{
"epoch": 0.07,
"grad_norm": 1.6225123405456543,
"learning_rate": 9.897168706418965e-05,
"loss": 2.5386,
"step": 2585
},
{
"epoch": 0.07,
"grad_norm": 0.9163520932197571,
"learning_rate": 9.89675591354127e-05,
"loss": 2.5432,
"step": 2590
},
{
"epoch": 0.07,
"grad_norm": 0.9534482359886169,
"learning_rate": 9.896342302433866e-05,
"loss": 2.5135,
"step": 2595
},
{
"epoch": 0.07,
"grad_norm": 0.9574959874153137,
"learning_rate": 9.895927873165867e-05,
"loss": 2.5306,
"step": 2600
},
{
"epoch": 0.07,
"grad_norm": 0.9217464327812195,
"learning_rate": 9.895512625806522e-05,
"loss": 2.5216,
"step": 2605
},
{
"epoch": 0.07,
"grad_norm": 0.8753473162651062,
"learning_rate": 9.895096560425218e-05,
"loss": 2.6003,
"step": 2610
},
{
"epoch": 0.07,
"grad_norm": 0.9147908091545105,
"learning_rate": 9.894679677091477e-05,
"loss": 2.552,
"step": 2615
},
{
"epoch": 0.07,
"grad_norm": 0.9207434058189392,
"learning_rate": 9.89426197587496e-05,
"loss": 2.6417,
"step": 2620
},
{
"epoch": 0.07,
"grad_norm": 0.9126414060592651,
"learning_rate": 9.893843456845462e-05,
"loss": 2.6052,
"step": 2625
},
{
"epoch": 0.07,
"grad_norm": 0.9393258094787598,
"learning_rate": 9.893424120072916e-05,
"loss": 2.6097,
"step": 2630
},
{
"epoch": 0.07,
"grad_norm": 0.883754551410675,
"learning_rate": 9.893003965627394e-05,
"loss": 2.4747,
"step": 2635
},
{
"epoch": 0.07,
"grad_norm": 0.9356350302696228,
"learning_rate": 9.8925829935791e-05,
"loss": 2.4413,
"step": 2640
},
{
"epoch": 0.07,
"grad_norm": 0.9513151049613953,
"learning_rate": 9.892161203998375e-05,
"loss": 2.6106,
"step": 2645
},
{
"epoch": 0.07,
"grad_norm": 0.9299737811088562,
"learning_rate": 9.891738596955703e-05,
"loss": 2.4982,
"step": 2650
},
{
"epoch": 0.07,
"grad_norm": 0.8935500979423523,
"learning_rate": 9.891315172521696e-05,
"loss": 2.5015,
"step": 2655
},
{
"epoch": 0.07,
"grad_norm": 0.9148421883583069,
"learning_rate": 9.890890930767111e-05,
"loss": 2.5257,
"step": 2660
},
{
"epoch": 0.07,
"grad_norm": 0.8721731901168823,
"learning_rate": 9.890465871762832e-05,
"loss": 2.5325,
"step": 2665
},
{
"epoch": 0.07,
"grad_norm": 0.9183290004730225,
"learning_rate": 9.89003999557989e-05,
"loss": 2.5504,
"step": 2670
},
{
"epoch": 0.07,
"grad_norm": 0.9214638471603394,
"learning_rate": 9.889613302289443e-05,
"loss": 2.5012,
"step": 2675
},
{
"epoch": 0.07,
"grad_norm": 0.8915846943855286,
"learning_rate": 9.889185791962792e-05,
"loss": 2.5824,
"step": 2680
},
{
"epoch": 0.07,
"grad_norm": 0.999625027179718,
"learning_rate": 9.888757464671372e-05,
"loss": 2.6263,
"step": 2685
},
{
"epoch": 0.07,
"grad_norm": 1.0056579113006592,
"learning_rate": 9.888328320486755e-05,
"loss": 2.4858,
"step": 2690
},
{
"epoch": 0.07,
"grad_norm": 0.9436735510826111,
"learning_rate": 9.88789835948065e-05,
"loss": 2.5627,
"step": 2695
},
{
"epoch": 0.07,
"grad_norm": 0.940756618976593,
"learning_rate": 9.887467581724901e-05,
"loss": 2.5039,
"step": 2700
},
{
"epoch": 0.07,
"grad_norm": 0.9503828883171082,
"learning_rate": 9.887035987291488e-05,
"loss": 2.5628,
"step": 2705
},
{
"epoch": 0.07,
"grad_norm": 0.9292907118797302,
"learning_rate": 9.88660357625253e-05,
"loss": 2.5443,
"step": 2710
},
{
"epoch": 0.07,
"grad_norm": 0.9516581296920776,
"learning_rate": 9.886170348680284e-05,
"loss": 2.5781,
"step": 2715
},
{
"epoch": 0.07,
"grad_norm": 0.9502872824668884,
"learning_rate": 9.885736304647139e-05,
"loss": 2.5028,
"step": 2720
},
{
"epoch": 0.07,
"grad_norm": 0.9478536248207092,
"learning_rate": 9.885301444225619e-05,
"loss": 2.5203,
"step": 2725
},
{
"epoch": 0.07,
"grad_norm": 0.8880490064620972,
"learning_rate": 9.884865767488391e-05,
"loss": 2.6008,
"step": 2730
},
{
"epoch": 0.07,
"grad_norm": 0.9714463353157043,
"learning_rate": 9.884429274508256e-05,
"loss": 2.5529,
"step": 2735
},
{
"epoch": 0.07,
"grad_norm": 0.870680034160614,
"learning_rate": 9.883991965358146e-05,
"loss": 2.4635,
"step": 2740
},
{
"epoch": 0.07,
"grad_norm": 0.9662335515022278,
"learning_rate": 9.883553840111136e-05,
"loss": 2.4973,
"step": 2745
},
{
"epoch": 0.07,
"grad_norm": 0.931488573551178,
"learning_rate": 9.883114898840437e-05,
"loss": 2.5733,
"step": 2750
},
{
"epoch": 0.07,
"grad_norm": 0.9375545978546143,
"learning_rate": 9.882675141619391e-05,
"loss": 2.5557,
"step": 2755
},
{
"epoch": 0.07,
"grad_norm": 0.9129160046577454,
"learning_rate": 9.882234568521482e-05,
"loss": 2.5252,
"step": 2760
},
{
"epoch": 0.07,
"grad_norm": 0.9213568568229675,
"learning_rate": 9.881793179620328e-05,
"loss": 2.4686,
"step": 2765
},
{
"epoch": 0.07,
"grad_norm": 0.9021924734115601,
"learning_rate": 9.881350974989682e-05,
"loss": 2.5277,
"step": 2770
},
{
"epoch": 0.07,
"grad_norm": 0.9493595361709595,
"learning_rate": 9.880907954703437e-05,
"loss": 2.44,
"step": 2775
},
{
"epoch": 0.07,
"grad_norm": 0.9300187826156616,
"learning_rate": 9.880464118835619e-05,
"loss": 2.5462,
"step": 2780
},
{
"epoch": 0.07,
"grad_norm": 0.8968974351882935,
"learning_rate": 9.880019467460391e-05,
"loss": 2.5446,
"step": 2785
},
{
"epoch": 0.07,
"grad_norm": 0.9221452474594116,
"learning_rate": 9.879574000652052e-05,
"loss": 2.5533,
"step": 2790
},
{
"epoch": 0.07,
"grad_norm": 0.963124692440033,
"learning_rate": 9.87912771848504e-05,
"loss": 2.5321,
"step": 2795
},
{
"epoch": 0.07,
"grad_norm": 0.9020286202430725,
"learning_rate": 9.878680621033925e-05,
"loss": 2.5018,
"step": 2800
},
{
"epoch": 0.07,
"grad_norm": 0.9417125582695007,
"learning_rate": 9.878232708373415e-05,
"loss": 2.5537,
"step": 2805
},
{
"epoch": 0.07,
"grad_norm": 0.9195294976234436,
"learning_rate": 9.877783980578357e-05,
"loss": 2.5884,
"step": 2810
},
{
"epoch": 0.07,
"grad_norm": 0.8789960145950317,
"learning_rate": 9.877334437723728e-05,
"loss": 2.4671,
"step": 2815
},
{
"epoch": 0.07,
"grad_norm": 0.8929395079612732,
"learning_rate": 9.87688407988465e-05,
"loss": 2.5051,
"step": 2820
},
{
"epoch": 0.07,
"grad_norm": 0.923984944820404,
"learning_rate": 9.876432907136372e-05,
"loss": 2.493,
"step": 2825
},
{
"epoch": 0.07,
"grad_norm": 0.9125444293022156,
"learning_rate": 9.875980919554287e-05,
"loss": 2.5633,
"step": 2830
},
{
"epoch": 0.07,
"grad_norm": 0.9064885973930359,
"learning_rate": 9.875528117213915e-05,
"loss": 2.4947,
"step": 2835
},
{
"epoch": 0.07,
"grad_norm": 0.8959868550300598,
"learning_rate": 9.875074500190922e-05,
"loss": 2.5006,
"step": 2840
},
{
"epoch": 0.07,
"grad_norm": 0.9482019543647766,
"learning_rate": 9.874620068561104e-05,
"loss": 2.5038,
"step": 2845
},
{
"epoch": 0.07,
"grad_norm": 0.9454628825187683,
"learning_rate": 9.874164822400396e-05,
"loss": 2.5277,
"step": 2850
},
{
"epoch": 0.07,
"grad_norm": 0.9381676912307739,
"learning_rate": 9.873708761784866e-05,
"loss": 2.5807,
"step": 2855
},
{
"epoch": 0.07,
"grad_norm": 0.958092212677002,
"learning_rate": 9.873251886790722e-05,
"loss": 2.6152,
"step": 2860
},
{
"epoch": 0.07,
"grad_norm": 0.8879554867744446,
"learning_rate": 9.872794197494307e-05,
"loss": 2.5088,
"step": 2865
},
{
"epoch": 0.07,
"grad_norm": 0.9326107501983643,
"learning_rate": 9.872335693972094e-05,
"loss": 2.5495,
"step": 2870
},
{
"epoch": 0.07,
"grad_norm": 0.8932898044586182,
"learning_rate": 9.871876376300702e-05,
"loss": 2.4966,
"step": 2875
},
{
"epoch": 0.07,
"grad_norm": 0.8879512548446655,
"learning_rate": 9.871416244556881e-05,
"loss": 2.5243,
"step": 2880
},
{
"epoch": 0.07,
"grad_norm": 0.906761646270752,
"learning_rate": 9.870955298817517e-05,
"loss": 2.4751,
"step": 2885
},
{
"epoch": 0.08,
"grad_norm": 0.938980758190155,
"learning_rate": 9.87049353915963e-05,
"loss": 2.4671,
"step": 2890
},
{
"epoch": 0.08,
"grad_norm": 0.835111677646637,
"learning_rate": 9.87003096566038e-05,
"loss": 2.4957,
"step": 2895
},
{
"epoch": 0.08,
"grad_norm": 0.9268792867660522,
"learning_rate": 9.869567578397062e-05,
"loss": 2.5514,
"step": 2900
},
{
"epoch": 0.08,
"grad_norm": 0.9292694926261902,
"learning_rate": 9.869103377447105e-05,
"loss": 2.5926,
"step": 2905
},
{
"epoch": 0.08,
"grad_norm": 0.916403591632843,
"learning_rate": 9.868638362888076e-05,
"loss": 2.4919,
"step": 2910
},
{
"epoch": 0.08,
"grad_norm": 0.9362393617630005,
"learning_rate": 9.868172534797678e-05,
"loss": 2.6159,
"step": 2915
},
{
"epoch": 0.08,
"grad_norm": 0.936874508857727,
"learning_rate": 9.867705893253747e-05,
"loss": 2.4138,
"step": 2920
},
{
"epoch": 0.08,
"grad_norm": 0.915784478187561,
"learning_rate": 9.867238438334259e-05,
"loss": 2.5141,
"step": 2925
},
{
"epoch": 0.08,
"grad_norm": 0.9194604754447937,
"learning_rate": 9.866770170117324e-05,
"loss": 2.5244,
"step": 2930
},
{
"epoch": 0.08,
"grad_norm": 1.0297510623931885,
"learning_rate": 9.866301088681185e-05,
"loss": 2.4862,
"step": 2935
},
{
"epoch": 0.08,
"grad_norm": 0.9317732453346252,
"learning_rate": 9.865831194104227e-05,
"loss": 2.4881,
"step": 2940
},
{
"epoch": 0.08,
"grad_norm": 0.8655943274497986,
"learning_rate": 9.865360486464967e-05,
"loss": 2.4824,
"step": 2945
},
{
"epoch": 0.08,
"grad_norm": 1.000765323638916,
"learning_rate": 9.864888965842056e-05,
"loss": 2.5053,
"step": 2950
},
{
"epoch": 0.08,
"grad_norm": 0.980956494808197,
"learning_rate": 9.864416632314287e-05,
"loss": 2.558,
"step": 2955
},
{
"epoch": 0.08,
"grad_norm": 0.9109508991241455,
"learning_rate": 9.863943485960583e-05,
"loss": 2.4641,
"step": 2960
},
{
"epoch": 0.08,
"grad_norm": 0.9524402618408203,
"learning_rate": 9.863469526860005e-05,
"loss": 2.5327,
"step": 2965
},
{
"epoch": 0.08,
"grad_norm": 0.9533178210258484,
"learning_rate": 9.862994755091752e-05,
"loss": 2.5039,
"step": 2970
},
{
"epoch": 0.08,
"grad_norm": 0.9892792105674744,
"learning_rate": 9.862519170735154e-05,
"loss": 2.4604,
"step": 2975
},
{
"epoch": 0.08,
"grad_norm": 1.1549103260040283,
"learning_rate": 9.86204277386968e-05,
"loss": 2.5456,
"step": 2980
},
{
"epoch": 0.08,
"grad_norm": 0.885002851486206,
"learning_rate": 9.861565564574935e-05,
"loss": 2.437,
"step": 2985
},
{
"epoch": 0.08,
"grad_norm": 0.938525915145874,
"learning_rate": 9.861087542930658e-05,
"loss": 2.4736,
"step": 2990
},
{
"epoch": 0.08,
"grad_norm": 1.1180083751678467,
"learning_rate": 9.860608709016726e-05,
"loss": 2.493,
"step": 2995
},
{
"epoch": 0.08,
"grad_norm": 0.8916651606559753,
"learning_rate": 9.860129062913147e-05,
"loss": 2.5107,
"step": 3000
},
{
"epoch": 0.08,
"grad_norm": 0.8922061324119568,
"learning_rate": 9.859648604700072e-05,
"loss": 2.5162,
"step": 3005
},
{
"epoch": 0.08,
"grad_norm": 0.9099216461181641,
"learning_rate": 9.859167334457781e-05,
"loss": 2.5103,
"step": 3010
},
{
"epoch": 0.08,
"grad_norm": 0.9527703523635864,
"learning_rate": 9.858685252266696e-05,
"loss": 2.5104,
"step": 3015
},
{
"epoch": 0.08,
"grad_norm": 3.305190086364746,
"learning_rate": 9.858202358207367e-05,
"loss": 2.5366,
"step": 3020
},
{
"epoch": 0.08,
"grad_norm": 0.8974258303642273,
"learning_rate": 9.857718652360486e-05,
"loss": 2.4352,
"step": 3025
},
{
"epoch": 0.08,
"grad_norm": 0.9247989058494568,
"learning_rate": 9.857234134806878e-05,
"loss": 2.5042,
"step": 3030
},
{
"epoch": 0.08,
"grad_norm": 0.9463310837745667,
"learning_rate": 9.856748805627503e-05,
"loss": 2.4991,
"step": 3035
},
{
"epoch": 0.08,
"grad_norm": 0.9005513191223145,
"learning_rate": 9.856262664903462e-05,
"loss": 2.5153,
"step": 3040
},
{
"epoch": 0.08,
"grad_norm": 0.8846108913421631,
"learning_rate": 9.855775712715981e-05,
"loss": 2.4064,
"step": 3045
},
{
"epoch": 0.08,
"grad_norm": 0.9176281094551086,
"learning_rate": 9.855287949146432e-05,
"loss": 2.5062,
"step": 3050
},
{
"epoch": 0.08,
"grad_norm": 0.8930521011352539,
"learning_rate": 9.854799374276318e-05,
"loss": 2.5128,
"step": 3055
},
{
"epoch": 0.08,
"grad_norm": 0.9506736993789673,
"learning_rate": 9.854309988187277e-05,
"loss": 2.5052,
"step": 3060
},
{
"epoch": 0.08,
"grad_norm": 0.9455916285514832,
"learning_rate": 9.853819790961086e-05,
"loss": 2.4169,
"step": 3065
},
{
"epoch": 0.08,
"grad_norm": 0.9309181571006775,
"learning_rate": 9.853328782679652e-05,
"loss": 2.5536,
"step": 3070
},
{
"epoch": 0.08,
"grad_norm": 0.9409065246582031,
"learning_rate": 9.852836963425021e-05,
"loss": 2.4852,
"step": 3075
},
{
"epoch": 0.08,
"grad_norm": 0.9356309771537781,
"learning_rate": 9.852344333279375e-05,
"loss": 2.4955,
"step": 3080
},
{
"epoch": 0.08,
"grad_norm": 0.981947660446167,
"learning_rate": 9.851850892325032e-05,
"loss": 2.4619,
"step": 3085
},
{
"epoch": 0.08,
"grad_norm": 0.8765103220939636,
"learning_rate": 9.851356640644443e-05,
"loss": 2.5426,
"step": 3090
},
{
"epoch": 0.08,
"grad_norm": 0.92372727394104,
"learning_rate": 9.850861578320196e-05,
"loss": 2.482,
"step": 3095
},
{
"epoch": 0.08,
"grad_norm": 0.8679202198982239,
"learning_rate": 9.850365705435013e-05,
"loss": 2.4641,
"step": 3100
},
{
"epoch": 0.08,
"grad_norm": 0.8970381617546082,
"learning_rate": 9.849869022071753e-05,
"loss": 2.4028,
"step": 3105
},
{
"epoch": 0.08,
"grad_norm": 0.9081596732139587,
"learning_rate": 9.849371528313411e-05,
"loss": 2.5671,
"step": 3110
},
{
"epoch": 0.08,
"grad_norm": 0.8779835104942322,
"learning_rate": 9.848873224243114e-05,
"loss": 2.4697,
"step": 3115
},
{
"epoch": 0.08,
"grad_norm": 0.8972172141075134,
"learning_rate": 9.848374109944128e-05,
"loss": 2.4688,
"step": 3120
},
{
"epoch": 0.08,
"grad_norm": 0.8907674551010132,
"learning_rate": 9.847874185499855e-05,
"loss": 2.5049,
"step": 3125
},
{
"epoch": 0.08,
"grad_norm": 0.8934869766235352,
"learning_rate": 9.847373450993828e-05,
"loss": 2.4828,
"step": 3130
},
{
"epoch": 0.08,
"grad_norm": 0.9090421199798584,
"learning_rate": 9.846871906509719e-05,
"loss": 2.4807,
"step": 3135
},
{
"epoch": 0.08,
"grad_norm": 0.9558232426643372,
"learning_rate": 9.846369552131334e-05,
"loss": 2.4555,
"step": 3140
},
{
"epoch": 0.08,
"grad_norm": 0.8783603310585022,
"learning_rate": 9.845866387942614e-05,
"loss": 2.5484,
"step": 3145
},
{
"epoch": 0.08,
"grad_norm": 0.8915989398956299,
"learning_rate": 9.845362414027637e-05,
"loss": 2.5269,
"step": 3150
},
{
"epoch": 0.08,
"grad_norm": 0.9381890892982483,
"learning_rate": 9.844857630470614e-05,
"loss": 2.4931,
"step": 3155
},
{
"epoch": 0.08,
"grad_norm": 0.9149619936943054,
"learning_rate": 9.844352037355893e-05,
"loss": 2.4937,
"step": 3160
},
{
"epoch": 0.08,
"grad_norm": 0.9246009588241577,
"learning_rate": 9.843845634767957e-05,
"loss": 2.5451,
"step": 3165
},
{
"epoch": 0.08,
"grad_norm": 0.9666916131973267,
"learning_rate": 9.843338422791424e-05,
"loss": 2.5083,
"step": 3170
},
{
"epoch": 0.08,
"grad_norm": 0.9277740120887756,
"learning_rate": 9.842830401511048e-05,
"loss": 2.5094,
"step": 3175
},
{
"epoch": 0.08,
"grad_norm": 0.9361172318458557,
"learning_rate": 9.842321571011714e-05,
"loss": 2.5089,
"step": 3180
},
{
"epoch": 0.08,
"grad_norm": 0.8992505669593811,
"learning_rate": 9.841811931378448e-05,
"loss": 2.5578,
"step": 3185
},
{
"epoch": 0.08,
"grad_norm": 0.8703805208206177,
"learning_rate": 9.841301482696413e-05,
"loss": 2.4495,
"step": 3190
},
{
"epoch": 0.08,
"grad_norm": 0.8690852522850037,
"learning_rate": 9.840790225050896e-05,
"loss": 2.4756,
"step": 3195
},
{
"epoch": 0.08,
"grad_norm": 0.8974294662475586,
"learning_rate": 9.840278158527328e-05,
"loss": 2.51,
"step": 3200
},
{
"epoch": 0.08,
"grad_norm": 0.8216846585273743,
"learning_rate": 9.839765283211278e-05,
"loss": 2.4813,
"step": 3205
},
{
"epoch": 0.08,
"grad_norm": 0.8770012855529785,
"learning_rate": 9.83925159918844e-05,
"loss": 2.4648,
"step": 3210
},
{
"epoch": 0.08,
"grad_norm": 0.9186285138130188,
"learning_rate": 9.838737106544651e-05,
"loss": 2.5983,
"step": 3215
},
{
"epoch": 0.08,
"grad_norm": 0.890736997127533,
"learning_rate": 9.838221805365881e-05,
"loss": 2.4124,
"step": 3220
},
{
"epoch": 0.08,
"grad_norm": 1.006027340888977,
"learning_rate": 9.837705695738236e-05,
"loss": 2.463,
"step": 3225
},
{
"epoch": 0.08,
"grad_norm": 0.8998842835426331,
"learning_rate": 9.837188777747954e-05,
"loss": 2.5058,
"step": 3230
},
{
"epoch": 0.08,
"grad_norm": 1.0391632318496704,
"learning_rate": 9.836671051481411e-05,
"loss": 2.473,
"step": 3235
},
{
"epoch": 0.08,
"grad_norm": 0.9052726030349731,
"learning_rate": 9.836152517025116e-05,
"loss": 2.5424,
"step": 3240
},
{
"epoch": 0.08,
"grad_norm": 0.9223129153251648,
"learning_rate": 9.835633174465715e-05,
"loss": 2.5392,
"step": 3245
},
{
"epoch": 0.08,
"grad_norm": 0.8875700831413269,
"learning_rate": 9.83511302388999e-05,
"loss": 2.4697,
"step": 3250
},
{
"epoch": 0.08,
"grad_norm": 0.981818675994873,
"learning_rate": 9.834592065384853e-05,
"loss": 2.5728,
"step": 3255
},
{
"epoch": 0.08,
"grad_norm": 0.9348360300064087,
"learning_rate": 9.834070299037357e-05,
"loss": 2.416,
"step": 3260
},
{
"epoch": 0.08,
"grad_norm": 0.9254019856452942,
"learning_rate": 9.833547724934687e-05,
"loss": 2.3897,
"step": 3265
},
{
"epoch": 0.08,
"grad_norm": 0.8724560141563416,
"learning_rate": 9.833024343164159e-05,
"loss": 2.3898,
"step": 3270
},
{
"epoch": 0.09,
"grad_norm": 0.8944783806800842,
"learning_rate": 9.832500153813234e-05,
"loss": 2.439,
"step": 3275
},
{
"epoch": 0.09,
"grad_norm": 0.8892961740493774,
"learning_rate": 9.831975156969498e-05,
"loss": 2.4822,
"step": 3280
},
{
"epoch": 0.09,
"grad_norm": 0.9107081294059753,
"learning_rate": 9.831449352720677e-05,
"loss": 2.4631,
"step": 3285
},
{
"epoch": 0.09,
"grad_norm": 0.8823145031929016,
"learning_rate": 9.830922741154631e-05,
"loss": 2.4825,
"step": 3290
},
{
"epoch": 0.09,
"grad_norm": 0.9482172131538391,
"learning_rate": 9.830395322359357e-05,
"loss": 2.5407,
"step": 3295
},
{
"epoch": 0.09,
"grad_norm": 0.9230644702911377,
"learning_rate": 9.829867096422983e-05,
"loss": 2.3988,
"step": 3300
},
{
"epoch": 0.09,
"grad_norm": 0.8623743653297424,
"learning_rate": 9.829338063433773e-05,
"loss": 2.4566,
"step": 3305
},
{
"epoch": 0.09,
"grad_norm": 0.9412261843681335,
"learning_rate": 9.828808223480126e-05,
"loss": 2.4991,
"step": 3310
},
{
"epoch": 0.09,
"grad_norm": 0.9355928897857666,
"learning_rate": 9.828277576650578e-05,
"loss": 2.5883,
"step": 3315
},
{
"epoch": 0.09,
"grad_norm": 0.950316846370697,
"learning_rate": 9.827746123033796e-05,
"loss": 2.4964,
"step": 3320
},
{
"epoch": 0.09,
"grad_norm": 0.8686743974685669,
"learning_rate": 9.827213862718587e-05,
"loss": 2.4811,
"step": 3325
},
{
"epoch": 0.09,
"grad_norm": 0.9092966914176941,
"learning_rate": 9.826680795793887e-05,
"loss": 2.4558,
"step": 3330
},
{
"epoch": 0.09,
"grad_norm": 0.8906340599060059,
"learning_rate": 9.826146922348773e-05,
"loss": 2.4707,
"step": 3335
},
{
"epoch": 0.09,
"grad_norm": 0.9468456506729126,
"learning_rate": 9.825612242472448e-05,
"loss": 2.407,
"step": 3340
},
{
"epoch": 0.09,
"grad_norm": 0.9757272601127625,
"learning_rate": 9.82507675625426e-05,
"loss": 2.4777,
"step": 3345
},
{
"epoch": 0.09,
"grad_norm": 1.0306956768035889,
"learning_rate": 9.824540463783682e-05,
"loss": 2.4418,
"step": 3350
},
{
"epoch": 0.09,
"grad_norm": 0.9280858039855957,
"learning_rate": 9.824003365150329e-05,
"loss": 2.51,
"step": 3355
},
{
"epoch": 0.09,
"grad_norm": 0.9034349918365479,
"learning_rate": 9.82346546044395e-05,
"loss": 2.47,
"step": 3360
},
{
"epoch": 0.09,
"grad_norm": 0.9494485855102539,
"learning_rate": 9.822926749754424e-05,
"loss": 2.4501,
"step": 3365
},
{
"epoch": 0.09,
"grad_norm": 0.9161350727081299,
"learning_rate": 9.822387233171771e-05,
"loss": 2.5873,
"step": 3370
},
{
"epoch": 0.09,
"grad_norm": 0.928499698638916,
"learning_rate": 9.821846910786136e-05,
"loss": 2.4474,
"step": 3375
},
{
"epoch": 0.09,
"grad_norm": 0.8854020833969116,
"learning_rate": 9.821305782687812e-05,
"loss": 2.4803,
"step": 3380
},
{
"epoch": 0.09,
"grad_norm": 0.9661669731140137,
"learning_rate": 9.820763848967215e-05,
"loss": 2.4867,
"step": 3385
},
{
"epoch": 0.09,
"grad_norm": 0.9096003770828247,
"learning_rate": 9.820221109714902e-05,
"loss": 2.4514,
"step": 3390
},
{
"epoch": 0.09,
"grad_norm": 0.9703332185745239,
"learning_rate": 9.81967756502156e-05,
"loss": 2.4651,
"step": 3395
},
{
"epoch": 0.09,
"grad_norm": 0.8770759105682373,
"learning_rate": 9.819133214978017e-05,
"loss": 2.4688,
"step": 3400
},
{
"epoch": 0.09,
"grad_norm": 0.8933206796646118,
"learning_rate": 9.818588059675229e-05,
"loss": 2.4277,
"step": 3405
},
{
"epoch": 0.09,
"grad_norm": 0.9417106509208679,
"learning_rate": 9.818042099204292e-05,
"loss": 2.4943,
"step": 3410
},
{
"epoch": 0.09,
"grad_norm": 0.9198511838912964,
"learning_rate": 9.817495333656433e-05,
"loss": 2.3686,
"step": 3415
},
{
"epoch": 0.09,
"grad_norm": 0.9361252784729004,
"learning_rate": 9.816947763123013e-05,
"loss": 2.4634,
"step": 3420
},
{
"epoch": 0.09,
"grad_norm": 0.9445791840553284,
"learning_rate": 9.816399387695529e-05,
"loss": 2.551,
"step": 3425
},
{
"epoch": 0.09,
"grad_norm": 0.8754252791404724,
"learning_rate": 9.815850207465616e-05,
"loss": 2.5438,
"step": 3430
},
{
"epoch": 0.09,
"grad_norm": 0.9107419848442078,
"learning_rate": 9.815300222525035e-05,
"loss": 2.4719,
"step": 3435
},
{
"epoch": 0.09,
"grad_norm": 0.8872767686843872,
"learning_rate": 9.814749432965691e-05,
"loss": 2.4689,
"step": 3440
},
{
"epoch": 0.09,
"grad_norm": 0.8860660195350647,
"learning_rate": 9.814197838879615e-05,
"loss": 2.4725,
"step": 3445
},
{
"epoch": 0.09,
"grad_norm": 0.8937519192695618,
"learning_rate": 9.813645440358981e-05,
"loss": 2.5964,
"step": 3450
},
{
"epoch": 0.09,
"grad_norm": 0.9341673851013184,
"learning_rate": 9.813092237496087e-05,
"loss": 2.5596,
"step": 3455
},
{
"epoch": 0.09,
"grad_norm": 0.9140751361846924,
"learning_rate": 9.812538230383376e-05,
"loss": 2.5696,
"step": 3460
},
{
"epoch": 0.09,
"grad_norm": 0.9442065358161926,
"learning_rate": 9.811983419113418e-05,
"loss": 2.5314,
"step": 3465
},
{
"epoch": 0.09,
"grad_norm": 0.8336922526359558,
"learning_rate": 9.81142780377892e-05,
"loss": 2.4158,
"step": 3470
},
{
"epoch": 0.09,
"grad_norm": 0.9034547805786133,
"learning_rate": 9.810871384472724e-05,
"loss": 2.4766,
"step": 3475
},
{
"epoch": 0.09,
"grad_norm": 0.8900104761123657,
"learning_rate": 9.810314161287806e-05,
"loss": 2.5828,
"step": 3480
},
{
"epoch": 0.09,
"grad_norm": 0.907195508480072,
"learning_rate": 9.809756134317276e-05,
"loss": 2.4788,
"step": 3485
},
{
"epoch": 0.09,
"grad_norm": 0.9055370688438416,
"learning_rate": 9.809197303654376e-05,
"loss": 2.5649,
"step": 3490
},
{
"epoch": 0.09,
"grad_norm": 0.9177443981170654,
"learning_rate": 9.808637669392487e-05,
"loss": 2.4982,
"step": 3495
},
{
"epoch": 0.09,
"grad_norm": 0.8704044818878174,
"learning_rate": 9.808077231625119e-05,
"loss": 2.4168,
"step": 3500
},
{
"epoch": 0.09,
"grad_norm": 0.9013683795928955,
"learning_rate": 9.807515990445922e-05,
"loss": 2.4769,
"step": 3505
},
{
"epoch": 0.09,
"grad_norm": 0.8629633188247681,
"learning_rate": 9.806953945948676e-05,
"loss": 2.4896,
"step": 3510
},
{
"epoch": 0.09,
"grad_norm": 0.9063811898231506,
"learning_rate": 9.806391098227297e-05,
"loss": 2.4587,
"step": 3515
},
{
"epoch": 0.09,
"grad_norm": 0.8828935623168945,
"learning_rate": 9.805827447375835e-05,
"loss": 2.4506,
"step": 3520
},
{
"epoch": 0.09,
"grad_norm": 0.9073445796966553,
"learning_rate": 9.805262993488473e-05,
"loss": 2.5141,
"step": 3525
},
{
"epoch": 0.09,
"grad_norm": 0.8797391057014465,
"learning_rate": 9.804697736659532e-05,
"loss": 2.4692,
"step": 3530
},
{
"epoch": 0.09,
"grad_norm": 0.935692310333252,
"learning_rate": 9.804131676983458e-05,
"loss": 2.4164,
"step": 3535
},
{
"epoch": 0.09,
"grad_norm": 0.9069603085517883,
"learning_rate": 9.803564814554844e-05,
"loss": 2.4417,
"step": 3540
},
{
"epoch": 0.09,
"grad_norm": 0.9014777541160583,
"learning_rate": 9.802997149468408e-05,
"loss": 2.4634,
"step": 3545
},
{
"epoch": 0.09,
"grad_norm": 0.8588182926177979,
"learning_rate": 9.802428681819003e-05,
"loss": 2.4765,
"step": 3550
},
{
"epoch": 0.09,
"grad_norm": 0.8845542669296265,
"learning_rate": 9.801859411701621e-05,
"loss": 2.4248,
"step": 3555
},
{
"epoch": 0.09,
"grad_norm": 0.9298128485679626,
"learning_rate": 9.801289339211384e-05,
"loss": 2.4301,
"step": 3560
},
{
"epoch": 0.09,
"grad_norm": 0.9109869599342346,
"learning_rate": 9.800718464443547e-05,
"loss": 2.51,
"step": 3565
},
{
"epoch": 0.09,
"grad_norm": 0.9425008893013,
"learning_rate": 9.800146787493503e-05,
"loss": 2.4605,
"step": 3570
},
{
"epoch": 0.09,
"grad_norm": 0.9391404390335083,
"learning_rate": 9.799574308456775e-05,
"loss": 2.4651,
"step": 3575
},
{
"epoch": 0.09,
"grad_norm": 0.8470202684402466,
"learning_rate": 9.799001027429026e-05,
"loss": 2.4931,
"step": 3580
},
{
"epoch": 0.09,
"grad_norm": 0.8704414367675781,
"learning_rate": 9.798426944506048e-05,
"loss": 2.4883,
"step": 3585
},
{
"epoch": 0.09,
"grad_norm": 0.9353266358375549,
"learning_rate": 9.797852059783766e-05,
"loss": 2.359,
"step": 3590
},
{
"epoch": 0.09,
"grad_norm": 0.8455348610877991,
"learning_rate": 9.79727637335824e-05,
"loss": 2.4422,
"step": 3595
},
{
"epoch": 0.09,
"grad_norm": 0.9137890934944153,
"learning_rate": 9.79669988532567e-05,
"loss": 2.4308,
"step": 3600
},
{
"epoch": 0.09,
"grad_norm": 0.8620888590812683,
"learning_rate": 9.79612259578238e-05,
"loss": 2.4131,
"step": 3605
},
{
"epoch": 0.09,
"grad_norm": 0.9356308579444885,
"learning_rate": 9.795544504824836e-05,
"loss": 2.4935,
"step": 3610
},
{
"epoch": 0.09,
"grad_norm": 0.8894330263137817,
"learning_rate": 9.794965612549635e-05,
"loss": 2.5373,
"step": 3615
},
{
"epoch": 0.09,
"grad_norm": 0.8981727361679077,
"learning_rate": 9.794385919053506e-05,
"loss": 2.4903,
"step": 3620
},
{
"epoch": 0.09,
"grad_norm": 0.8943347334861755,
"learning_rate": 9.793805424433313e-05,
"loss": 2.518,
"step": 3625
},
{
"epoch": 0.09,
"grad_norm": 0.9402914643287659,
"learning_rate": 9.793224128786058e-05,
"loss": 2.4119,
"step": 3630
},
{
"epoch": 0.09,
"grad_norm": 0.9275897145271301,
"learning_rate": 9.79264203220887e-05,
"loss": 2.5328,
"step": 3635
},
{
"epoch": 0.09,
"grad_norm": 0.8831329345703125,
"learning_rate": 9.792059134799016e-05,
"loss": 2.3997,
"step": 3640
},
{
"epoch": 0.09,
"grad_norm": 0.9803993105888367,
"learning_rate": 9.791475436653896e-05,
"loss": 2.4541,
"step": 3645
},
{
"epoch": 0.09,
"grad_norm": 0.9330896735191345,
"learning_rate": 9.790890937871046e-05,
"loss": 2.4193,
"step": 3650
},
{
"epoch": 0.09,
"grad_norm": 0.9279419183731079,
"learning_rate": 9.79030563854813e-05,
"loss": 2.4356,
"step": 3655
},
{
"epoch": 0.09,
"grad_norm": 0.9156128764152527,
"learning_rate": 9.789719538782952e-05,
"loss": 2.4946,
"step": 3660
},
{
"epoch": 0.1,
"grad_norm": 1.1419275999069214,
"learning_rate": 9.789132638673445e-05,
"loss": 2.4383,
"step": 3665
},
{
"epoch": 0.1,
"grad_norm": 0.8956980109214783,
"learning_rate": 9.78854493831768e-05,
"loss": 2.4726,
"step": 3670
},
{
"epoch": 0.1,
"grad_norm": 0.8836015462875366,
"learning_rate": 9.787956437813859e-05,
"loss": 2.46,
"step": 3675
},
{
"epoch": 0.1,
"grad_norm": 0.9386027455329895,
"learning_rate": 9.787367137260317e-05,
"loss": 2.448,
"step": 3680
},
{
"epoch": 0.1,
"grad_norm": 0.9100561141967773,
"learning_rate": 9.786777036755525e-05,
"loss": 2.4759,
"step": 3685
},
{
"epoch": 0.1,
"grad_norm": 0.9176651239395142,
"learning_rate": 9.786186136398085e-05,
"loss": 2.501,
"step": 3690
},
{
"epoch": 0.1,
"grad_norm": 0.8947619199752808,
"learning_rate": 9.785594436286738e-05,
"loss": 2.4448,
"step": 3695
},
{
"epoch": 0.1,
"grad_norm": 0.8647358417510986,
"learning_rate": 9.78500193652035e-05,
"loss": 2.5247,
"step": 3700
},
{
"epoch": 0.1,
"grad_norm": 0.8395596146583557,
"learning_rate": 9.784408637197929e-05,
"loss": 2.4679,
"step": 3705
},
{
"epoch": 0.1,
"grad_norm": 0.8445679545402527,
"learning_rate": 9.783814538418612e-05,
"loss": 2.4673,
"step": 3710
},
{
"epoch": 0.1,
"grad_norm": 0.8595328330993652,
"learning_rate": 9.783219640281671e-05,
"loss": 2.3584,
"step": 3715
},
{
"epoch": 0.1,
"grad_norm": 0.8912994861602783,
"learning_rate": 9.78262394288651e-05,
"loss": 2.5537,
"step": 3720
},
{
"epoch": 0.1,
"grad_norm": 0.8637709021568298,
"learning_rate": 9.782027446332671e-05,
"loss": 2.5305,
"step": 3725
},
{
"epoch": 0.1,
"grad_norm": 0.9240624904632568,
"learning_rate": 9.781430150719822e-05,
"loss": 2.5246,
"step": 3730
},
{
"epoch": 0.1,
"grad_norm": 0.8442887663841248,
"learning_rate": 9.780832056147772e-05,
"loss": 2.4098,
"step": 3735
},
{
"epoch": 0.1,
"grad_norm": 0.9382492303848267,
"learning_rate": 9.78023316271646e-05,
"loss": 2.3824,
"step": 3740
},
{
"epoch": 0.1,
"grad_norm": 0.8716952800750732,
"learning_rate": 9.779633470525957e-05,
"loss": 2.4818,
"step": 3745
},
{
"epoch": 0.1,
"grad_norm": 0.9429145455360413,
"learning_rate": 9.77903297967647e-05,
"loss": 2.5059,
"step": 3750
},
{
"epoch": 0.1,
"grad_norm": 0.9276791214942932,
"learning_rate": 9.778431690268341e-05,
"loss": 2.5209,
"step": 3755
},
{
"epoch": 0.1,
"grad_norm": 0.9431637525558472,
"learning_rate": 9.777829602402043e-05,
"loss": 2.3736,
"step": 3760
},
{
"epoch": 0.1,
"grad_norm": 0.8518359065055847,
"learning_rate": 9.777226716178178e-05,
"loss": 2.5441,
"step": 3765
},
{
"epoch": 0.1,
"grad_norm": 0.8934125900268555,
"learning_rate": 9.776623031697492e-05,
"loss": 2.4303,
"step": 3770
},
{
"epoch": 0.1,
"grad_norm": 0.938008725643158,
"learning_rate": 9.776018549060855e-05,
"loss": 2.475,
"step": 3775
},
{
"epoch": 0.1,
"grad_norm": 0.886810839176178,
"learning_rate": 9.775413268369274e-05,
"loss": 2.4628,
"step": 3780
},
{
"epoch": 0.1,
"grad_norm": 0.9275068044662476,
"learning_rate": 9.77480718972389e-05,
"loss": 2.4518,
"step": 3785
},
{
"epoch": 0.1,
"grad_norm": 0.8837240934371948,
"learning_rate": 9.774200313225976e-05,
"loss": 2.4289,
"step": 3790
},
{
"epoch": 0.1,
"grad_norm": 0.9412903189659119,
"learning_rate": 9.773592638976939e-05,
"loss": 2.476,
"step": 3795
},
{
"epoch": 0.1,
"grad_norm": 0.9908060431480408,
"learning_rate": 9.77298416707832e-05,
"loss": 2.5368,
"step": 3800
},
{
"epoch": 0.1,
"grad_norm": 0.9376251697540283,
"learning_rate": 9.772374897631791e-05,
"loss": 2.5137,
"step": 3805
},
{
"epoch": 0.1,
"grad_norm": 0.8808382749557495,
"learning_rate": 9.771764830739159e-05,
"loss": 2.4702,
"step": 3810
},
{
"epoch": 0.1,
"grad_norm": 0.940961480140686,
"learning_rate": 9.771153966502365e-05,
"loss": 2.4038,
"step": 3815
},
{
"epoch": 0.1,
"grad_norm": 0.8861954212188721,
"learning_rate": 9.77054230502348e-05,
"loss": 2.4059,
"step": 3820
},
{
"epoch": 0.1,
"grad_norm": 0.8724052906036377,
"learning_rate": 9.769929846404711e-05,
"loss": 2.4054,
"step": 3825
},
{
"epoch": 0.1,
"grad_norm": 0.9138715863227844,
"learning_rate": 9.769316590748401e-05,
"loss": 2.4758,
"step": 3830
},
{
"epoch": 0.1,
"grad_norm": 0.8826479911804199,
"learning_rate": 9.768702538157018e-05,
"loss": 2.4798,
"step": 3835
},
{
"epoch": 0.1,
"grad_norm": 0.864983081817627,
"learning_rate": 9.76808768873317e-05,
"loss": 2.4289,
"step": 3840
},
{
"epoch": 0.1,
"grad_norm": 0.8833311200141907,
"learning_rate": 9.767472042579595e-05,
"loss": 2.5038,
"step": 3845
},
{
"epoch": 0.1,
"grad_norm": 0.8873911499977112,
"learning_rate": 9.766855599799168e-05,
"loss": 2.5109,
"step": 3850
},
{
"epoch": 0.1,
"grad_norm": 0.9640964865684509,
"learning_rate": 9.76623836049489e-05,
"loss": 2.5075,
"step": 3855
},
{
"epoch": 0.1,
"grad_norm": 0.9603625535964966,
"learning_rate": 9.765620324769903e-05,
"loss": 2.4857,
"step": 3860
},
{
"epoch": 0.1,
"grad_norm": 0.8782946467399597,
"learning_rate": 9.765001492727476e-05,
"loss": 2.4161,
"step": 3865
},
{
"epoch": 0.1,
"grad_norm": 0.9399060010910034,
"learning_rate": 9.764381864471017e-05,
"loss": 2.4789,
"step": 3870
},
{
"epoch": 0.1,
"grad_norm": 0.8807359933853149,
"learning_rate": 9.76376144010406e-05,
"loss": 2.4464,
"step": 3875
},
{
"epoch": 0.1,
"grad_norm": 0.9333061575889587,
"learning_rate": 9.763140219730278e-05,
"loss": 2.5241,
"step": 3880
},
{
"epoch": 0.1,
"grad_norm": 0.8983426690101624,
"learning_rate": 9.762518203453472e-05,
"loss": 2.3449,
"step": 3885
},
{
"epoch": 0.1,
"grad_norm": 0.8563090562820435,
"learning_rate": 9.761895391377581e-05,
"loss": 2.4672,
"step": 3890
},
{
"epoch": 0.1,
"grad_norm": 0.8848186135292053,
"learning_rate": 9.761271783606674e-05,
"loss": 2.3881,
"step": 3895
},
{
"epoch": 0.1,
"grad_norm": 0.8474920988082886,
"learning_rate": 9.760647380244952e-05,
"loss": 2.4399,
"step": 3900
},
{
"epoch": 0.1,
"grad_norm": 0.8754302859306335,
"learning_rate": 9.760022181396753e-05,
"loss": 2.3654,
"step": 3905
},
{
"epoch": 0.1,
"grad_norm": 0.903482973575592,
"learning_rate": 9.759396187166543e-05,
"loss": 2.4485,
"step": 3910
},
{
"epoch": 0.1,
"grad_norm": 0.883611261844635,
"learning_rate": 9.758769397658926e-05,
"loss": 2.488,
"step": 3915
},
{
"epoch": 0.1,
"grad_norm": 0.9298650622367859,
"learning_rate": 9.758141812978634e-05,
"loss": 2.4176,
"step": 3920
},
{
"epoch": 0.1,
"grad_norm": 0.9373406171798706,
"learning_rate": 9.757513433230534e-05,
"loss": 2.4534,
"step": 3925
},
{
"epoch": 0.1,
"grad_norm": 0.8700555562973022,
"learning_rate": 9.756884258519627e-05,
"loss": 2.4708,
"step": 3930
},
{
"epoch": 0.1,
"grad_norm": 0.916092038154602,
"learning_rate": 9.756254288951046e-05,
"loss": 2.5497,
"step": 3935
},
{
"epoch": 0.1,
"grad_norm": 0.8875973224639893,
"learning_rate": 9.755623524630055e-05,
"loss": 2.3652,
"step": 3940
},
{
"epoch": 0.1,
"grad_norm": 0.84308922290802,
"learning_rate": 9.754991965662055e-05,
"loss": 2.5332,
"step": 3945
},
{
"epoch": 0.1,
"grad_norm": 0.9055904150009155,
"learning_rate": 9.754359612152574e-05,
"loss": 2.4008,
"step": 3950
},
{
"epoch": 0.1,
"grad_norm": 0.8861605525016785,
"learning_rate": 9.753726464207278e-05,
"loss": 2.5826,
"step": 3955
},
{
"epoch": 0.1,
"grad_norm": 0.8517809510231018,
"learning_rate": 9.753092521931963e-05,
"loss": 2.4587,
"step": 3960
},
{
"epoch": 0.1,
"grad_norm": 0.8306174874305725,
"learning_rate": 9.752457785432558e-05,
"loss": 2.417,
"step": 3965
},
{
"epoch": 0.1,
"grad_norm": 0.882463812828064,
"learning_rate": 9.751822254815126e-05,
"loss": 2.3809,
"step": 3970
},
{
"epoch": 0.1,
"grad_norm": 0.8744756579399109,
"learning_rate": 9.751185930185861e-05,
"loss": 2.4584,
"step": 3975
},
{
"epoch": 0.1,
"grad_norm": 0.8309944868087769,
"learning_rate": 9.750548811651091e-05,
"loss": 2.3931,
"step": 3980
},
{
"epoch": 0.1,
"grad_norm": 0.8728353381156921,
"learning_rate": 9.749910899317276e-05,
"loss": 2.5379,
"step": 3985
},
{
"epoch": 0.1,
"grad_norm": 0.8816731572151184,
"learning_rate": 9.74927219329101e-05,
"loss": 2.4856,
"step": 3990
},
{
"epoch": 0.1,
"grad_norm": 0.9452540278434753,
"learning_rate": 9.748632693679016e-05,
"loss": 2.5001,
"step": 3995
},
{
"epoch": 0.1,
"grad_norm": 0.8715027570724487,
"learning_rate": 9.747992400588153e-05,
"loss": 2.4653,
"step": 4000
},
{
"epoch": 0.1,
"grad_norm": 0.9610370397567749,
"learning_rate": 9.747351314125413e-05,
"loss": 2.3837,
"step": 4005
},
{
"epoch": 0.1,
"grad_norm": 0.8891305327415466,
"learning_rate": 9.746709434397916e-05,
"loss": 2.5148,
"step": 4010
},
{
"epoch": 0.1,
"grad_norm": 0.9000439047813416,
"learning_rate": 9.746066761512921e-05,
"loss": 2.4934,
"step": 4015
},
{
"epoch": 0.1,
"grad_norm": 0.8105304837226868,
"learning_rate": 9.745423295577815e-05,
"loss": 2.4136,
"step": 4020
},
{
"epoch": 0.1,
"grad_norm": 0.9107775688171387,
"learning_rate": 9.744779036700118e-05,
"loss": 2.3474,
"step": 4025
},
{
"epoch": 0.1,
"grad_norm": 0.8694551587104797,
"learning_rate": 9.744133984987485e-05,
"loss": 2.4947,
"step": 4030
},
{
"epoch": 0.1,
"grad_norm": 0.8516966700553894,
"learning_rate": 9.743488140547701e-05,
"loss": 2.423,
"step": 4035
},
{
"epoch": 0.1,
"grad_norm": 0.8018931150436401,
"learning_rate": 9.742841503488682e-05,
"loss": 2.4356,
"step": 4040
},
{
"epoch": 0.1,
"grad_norm": 0.842832624912262,
"learning_rate": 9.742194073918482e-05,
"loss": 2.4577,
"step": 4045
},
{
"epoch": 0.11,
"grad_norm": 0.8792210221290588,
"learning_rate": 9.741545851945282e-05,
"loss": 2.4573,
"step": 4050
},
{
"epoch": 0.11,
"grad_norm": 0.8316389322280884,
"learning_rate": 9.740896837677398e-05,
"loss": 2.5143,
"step": 4055
},
{
"epoch": 0.11,
"grad_norm": 0.896112322807312,
"learning_rate": 9.740247031223279e-05,
"loss": 2.4563,
"step": 4060
},
{
"epoch": 0.11,
"grad_norm": 0.8676568865776062,
"learning_rate": 9.739596432691502e-05,
"loss": 2.4799,
"step": 4065
},
{
"epoch": 0.11,
"grad_norm": 0.942961573600769,
"learning_rate": 9.738945042190785e-05,
"loss": 2.3602,
"step": 4070
},
{
"epoch": 0.11,
"grad_norm": 0.8567558526992798,
"learning_rate": 9.738292859829968e-05,
"loss": 2.4668,
"step": 4075
},
{
"epoch": 0.11,
"grad_norm": 0.8903143405914307,
"learning_rate": 9.73763988571803e-05,
"loss": 2.4341,
"step": 4080
},
{
"epoch": 0.11,
"grad_norm": 0.8469401001930237,
"learning_rate": 9.73698611996408e-05,
"loss": 2.4109,
"step": 4085
},
{
"epoch": 0.11,
"grad_norm": 0.8484046459197998,
"learning_rate": 9.736331562677361e-05,
"loss": 2.4684,
"step": 4090
},
{
"epoch": 0.11,
"grad_norm": 0.9134167432785034,
"learning_rate": 9.735676213967247e-05,
"loss": 2.4319,
"step": 4095
},
{
"epoch": 0.11,
"grad_norm": 0.8421111106872559,
"learning_rate": 9.735020073943244e-05,
"loss": 2.4193,
"step": 4100
},
{
"epoch": 0.11,
"grad_norm": 0.8343115448951721,
"learning_rate": 9.734363142714988e-05,
"loss": 2.4259,
"step": 4105
},
{
"epoch": 0.11,
"grad_norm": 0.878003716468811,
"learning_rate": 9.733705420392254e-05,
"loss": 2.4307,
"step": 4110
},
{
"epoch": 0.11,
"grad_norm": 0.8623335957527161,
"learning_rate": 9.733046907084944e-05,
"loss": 2.3984,
"step": 4115
},
{
"epoch": 0.11,
"grad_norm": 0.8645914196968079,
"learning_rate": 9.732387602903091e-05,
"loss": 2.5035,
"step": 4120
},
{
"epoch": 0.11,
"grad_norm": 0.8988960981369019,
"learning_rate": 9.731727507956862e-05,
"loss": 2.4995,
"step": 4125
},
{
"epoch": 0.11,
"grad_norm": 0.8828436732292175,
"learning_rate": 9.73106662235656e-05,
"loss": 2.4481,
"step": 4130
},
{
"epoch": 0.11,
"grad_norm": 0.848412811756134,
"learning_rate": 9.730404946212614e-05,
"loss": 2.383,
"step": 4135
},
{
"epoch": 0.11,
"grad_norm": 0.8317053318023682,
"learning_rate": 9.729742479635589e-05,
"loss": 2.3949,
"step": 4140
},
{
"epoch": 0.11,
"grad_norm": 0.894101619720459,
"learning_rate": 9.729079222736177e-05,
"loss": 2.4875,
"step": 4145
},
{
"epoch": 0.11,
"grad_norm": 0.8426854610443115,
"learning_rate": 9.728415175625211e-05,
"loss": 2.504,
"step": 4150
},
{
"epoch": 0.11,
"grad_norm": 0.8569448590278625,
"learning_rate": 9.727750338413647e-05,
"loss": 2.4974,
"step": 4155
},
{
"epoch": 0.11,
"grad_norm": 0.8657938838005066,
"learning_rate": 9.727084711212579e-05,
"loss": 2.4431,
"step": 4160
},
{
"epoch": 0.11,
"grad_norm": 0.8954491019248962,
"learning_rate": 9.726418294133229e-05,
"loss": 2.4476,
"step": 4165
},
{
"epoch": 0.11,
"grad_norm": 1.3556360006332397,
"learning_rate": 9.725751087286954e-05,
"loss": 2.3948,
"step": 4170
},
{
"epoch": 0.11,
"grad_norm": 0.8271290063858032,
"learning_rate": 9.725083090785242e-05,
"loss": 2.3922,
"step": 4175
},
{
"epoch": 0.11,
"grad_norm": 0.8866262435913086,
"learning_rate": 9.724414304739712e-05,
"loss": 2.4544,
"step": 4180
},
{
"epoch": 0.11,
"grad_norm": 0.8544206619262695,
"learning_rate": 9.723744729262116e-05,
"loss": 2.4047,
"step": 4185
},
{
"epoch": 0.11,
"grad_norm": 0.8622026443481445,
"learning_rate": 9.723074364464337e-05,
"loss": 2.5111,
"step": 4190
},
{
"epoch": 0.11,
"grad_norm": 0.911665141582489,
"learning_rate": 9.722403210458393e-05,
"loss": 2.4327,
"step": 4195
},
{
"epoch": 0.11,
"grad_norm": 0.9319663643836975,
"learning_rate": 9.721731267356426e-05,
"loss": 2.5004,
"step": 4200
},
{
"epoch": 0.11,
"grad_norm": 0.9587587118148804,
"learning_rate": 9.721058535270722e-05,
"loss": 2.4774,
"step": 4205
},
{
"epoch": 0.11,
"grad_norm": 0.9479362964630127,
"learning_rate": 9.720385014313686e-05,
"loss": 2.5569,
"step": 4210
},
{
"epoch": 0.11,
"grad_norm": 0.8598310351371765,
"learning_rate": 9.719710704597865e-05,
"loss": 2.5221,
"step": 4215
},
{
"epoch": 0.11,
"grad_norm": 0.8738716244697571,
"learning_rate": 9.719035606235931e-05,
"loss": 2.3953,
"step": 4220
},
{
"epoch": 0.11,
"grad_norm": 0.9036585688591003,
"learning_rate": 9.718359719340693e-05,
"loss": 2.4099,
"step": 4225
},
{
"epoch": 0.11,
"grad_norm": 0.9076921343803406,
"learning_rate": 9.717683044025086e-05,
"loss": 2.3934,
"step": 4230
},
{
"epoch": 0.11,
"grad_norm": 0.8633479475975037,
"learning_rate": 9.717005580402184e-05,
"loss": 2.4053,
"step": 4235
},
{
"epoch": 0.11,
"grad_norm": 0.9014074206352234,
"learning_rate": 9.716327328585183e-05,
"loss": 2.4764,
"step": 4240
},
{
"epoch": 0.11,
"grad_norm": 0.8797693252563477,
"learning_rate": 9.715648288687421e-05,
"loss": 2.4457,
"step": 4245
},
{
"epoch": 0.11,
"grad_norm": 0.8739760518074036,
"learning_rate": 9.714968460822363e-05,
"loss": 2.3853,
"step": 4250
},
{
"epoch": 0.11,
"grad_norm": 0.8975856304168701,
"learning_rate": 9.714287845103602e-05,
"loss": 2.499,
"step": 4255
},
{
"epoch": 0.11,
"grad_norm": 0.8922436833381653,
"learning_rate": 9.71360644164487e-05,
"loss": 2.5114,
"step": 4260
},
{
"epoch": 0.11,
"grad_norm": 0.8157884478569031,
"learning_rate": 9.712924250560025e-05,
"loss": 2.4149,
"step": 4265
},
{
"epoch": 0.11,
"grad_norm": 0.8854585289955139,
"learning_rate": 9.712241271963059e-05,
"loss": 2.4398,
"step": 4270
},
{
"epoch": 0.11,
"grad_norm": 0.952923059463501,
"learning_rate": 9.711557505968097e-05,
"loss": 2.4231,
"step": 4275
},
{
"epoch": 0.11,
"grad_norm": 0.8512819409370422,
"learning_rate": 9.710872952689391e-05,
"loss": 2.4081,
"step": 4280
},
{
"epoch": 0.11,
"grad_norm": 0.8664557933807373,
"learning_rate": 9.71018761224133e-05,
"loss": 2.4526,
"step": 4285
},
{
"epoch": 0.11,
"grad_norm": 0.8743461966514587,
"learning_rate": 9.709501484738429e-05,
"loss": 2.4356,
"step": 4290
},
{
"epoch": 0.11,
"grad_norm": 0.894364595413208,
"learning_rate": 9.70881457029534e-05,
"loss": 2.4801,
"step": 4295
},
{
"epoch": 0.11,
"grad_norm": 0.8613756895065308,
"learning_rate": 9.708126869026841e-05,
"loss": 2.4472,
"step": 4300
},
{
"epoch": 0.11,
"grad_norm": 0.8578716516494751,
"learning_rate": 9.707438381047846e-05,
"loss": 2.4839,
"step": 4305
},
{
"epoch": 0.11,
"grad_norm": 0.8649423122406006,
"learning_rate": 9.7067491064734e-05,
"loss": 2.4176,
"step": 4310
},
{
"epoch": 0.11,
"grad_norm": 0.8754303455352783,
"learning_rate": 9.706059045418676e-05,
"loss": 2.4705,
"step": 4315
},
{
"epoch": 0.11,
"grad_norm": 0.856622576713562,
"learning_rate": 9.705368197998982e-05,
"loss": 2.3888,
"step": 4320
},
{
"epoch": 0.11,
"grad_norm": 0.9290913343429565,
"learning_rate": 9.704676564329756e-05,
"loss": 2.5074,
"step": 4325
},
{
"epoch": 0.11,
"grad_norm": 0.8195645213127136,
"learning_rate": 9.703984144526565e-05,
"loss": 2.4148,
"step": 4330
},
{
"epoch": 0.11,
"grad_norm": 0.8634433746337891,
"learning_rate": 9.703290938705114e-05,
"loss": 2.441,
"step": 4335
},
{
"epoch": 0.11,
"grad_norm": 0.900746762752533,
"learning_rate": 9.702596946981232e-05,
"loss": 2.3907,
"step": 4340
},
{
"epoch": 0.11,
"grad_norm": 0.8529593348503113,
"learning_rate": 9.701902169470883e-05,
"loss": 2.5076,
"step": 4345
},
{
"epoch": 0.11,
"grad_norm": 0.9410119652748108,
"learning_rate": 9.701206606290164e-05,
"loss": 2.4713,
"step": 4350
},
{
"epoch": 0.11,
"grad_norm": 0.896649956703186,
"learning_rate": 9.700510257555295e-05,
"loss": 2.4803,
"step": 4355
},
{
"epoch": 0.11,
"grad_norm": 0.8850135207176208,
"learning_rate": 9.69981312338264e-05,
"loss": 2.4499,
"step": 4360
},
{
"epoch": 0.11,
"grad_norm": 0.9270910620689392,
"learning_rate": 9.699115203888686e-05,
"loss": 2.413,
"step": 4365
},
{
"epoch": 0.11,
"grad_norm": 0.9040417075157166,
"learning_rate": 9.69841649919005e-05,
"loss": 2.366,
"step": 4370
},
{
"epoch": 0.11,
"grad_norm": 0.884899377822876,
"learning_rate": 9.697717009403485e-05,
"loss": 2.509,
"step": 4375
},
{
"epoch": 0.11,
"grad_norm": 0.8952233791351318,
"learning_rate": 9.697016734645873e-05,
"loss": 2.4336,
"step": 4380
},
{
"epoch": 0.11,
"grad_norm": 0.8898236155509949,
"learning_rate": 9.696315675034227e-05,
"loss": 2.4442,
"step": 4385
},
{
"epoch": 0.11,
"grad_norm": 0.8343461751937866,
"learning_rate": 9.69561383068569e-05,
"loss": 2.3207,
"step": 4390
},
{
"epoch": 0.11,
"grad_norm": 0.8294013142585754,
"learning_rate": 9.694911201717541e-05,
"loss": 2.4039,
"step": 4395
},
{
"epoch": 0.11,
"grad_norm": 0.8741182088851929,
"learning_rate": 9.694207788247186e-05,
"loss": 2.5483,
"step": 4400
},
{
"epoch": 0.11,
"grad_norm": 0.9135164022445679,
"learning_rate": 9.693503590392159e-05,
"loss": 2.4893,
"step": 4405
},
{
"epoch": 0.11,
"grad_norm": 0.8335906267166138,
"learning_rate": 9.692798608270132e-05,
"loss": 2.3861,
"step": 4410
},
{
"epoch": 0.11,
"grad_norm": 0.845115602016449,
"learning_rate": 9.692092841998904e-05,
"loss": 2.4267,
"step": 4415
},
{
"epoch": 0.11,
"grad_norm": 0.8781867623329163,
"learning_rate": 9.691386291696407e-05,
"loss": 2.4369,
"step": 4420
},
{
"epoch": 0.11,
"grad_norm": 0.8892547488212585,
"learning_rate": 9.690678957480702e-05,
"loss": 2.3587,
"step": 4425
},
{
"epoch": 0.11,
"grad_norm": 0.8194572329521179,
"learning_rate": 9.689970839469981e-05,
"loss": 2.4063,
"step": 4430
},
{
"epoch": 0.12,
"grad_norm": 0.8236936330795288,
"learning_rate": 9.689261937782571e-05,
"loss": 2.4227,
"step": 4435
},
{
"epoch": 0.12,
"grad_norm": 0.8684244751930237,
"learning_rate": 9.688552252536924e-05,
"loss": 2.4057,
"step": 4440
},
{
"epoch": 0.12,
"grad_norm": 0.8150362372398376,
"learning_rate": 9.687841783851626e-05,
"loss": 2.3773,
"step": 4445
},
{
"epoch": 0.12,
"grad_norm": 0.8659597635269165,
"learning_rate": 9.687130531845394e-05,
"loss": 2.4704,
"step": 4450
},
{
"epoch": 0.12,
"grad_norm": 0.9055734872817993,
"learning_rate": 9.686418496637076e-05,
"loss": 2.4484,
"step": 4455
},
{
"epoch": 0.12,
"grad_norm": 0.8882852792739868,
"learning_rate": 9.685705678345652e-05,
"loss": 2.4374,
"step": 4460
},
{
"epoch": 0.12,
"grad_norm": 0.9285370707511902,
"learning_rate": 9.684992077090227e-05,
"loss": 2.4187,
"step": 4465
},
{
"epoch": 0.12,
"grad_norm": 0.9069604277610779,
"learning_rate": 9.684277692990045e-05,
"loss": 2.3922,
"step": 4470
},
{
"epoch": 0.12,
"grad_norm": 0.8838167190551758,
"learning_rate": 9.683562526164475e-05,
"loss": 2.4923,
"step": 4475
},
{
"epoch": 0.12,
"grad_norm": 0.8549697995185852,
"learning_rate": 9.68284657673302e-05,
"loss": 2.4606,
"step": 4480
},
{
"epoch": 0.12,
"grad_norm": 0.9082733988761902,
"learning_rate": 9.682129844815311e-05,
"loss": 2.4449,
"step": 4485
},
{
"epoch": 0.12,
"grad_norm": 0.8203231692314148,
"learning_rate": 9.681412330531112e-05,
"loss": 2.3399,
"step": 4490
},
{
"epoch": 0.12,
"grad_norm": 0.914690375328064,
"learning_rate": 9.680694034000318e-05,
"loss": 2.4168,
"step": 4495
},
{
"epoch": 0.12,
"grad_norm": 0.8260876536369324,
"learning_rate": 9.67997495534295e-05,
"loss": 2.4349,
"step": 4500
},
{
"epoch": 0.12,
"grad_norm": 0.9162200093269348,
"learning_rate": 9.679255094679169e-05,
"loss": 2.3601,
"step": 4505
},
{
"epoch": 0.12,
"grad_norm": 0.8778248429298401,
"learning_rate": 9.678534452129256e-05,
"loss": 2.3302,
"step": 4510
},
{
"epoch": 0.12,
"grad_norm": 0.8993799090385437,
"learning_rate": 9.677813027813632e-05,
"loss": 2.4566,
"step": 4515
},
{
"epoch": 0.12,
"grad_norm": 1.0065888166427612,
"learning_rate": 9.677090821852842e-05,
"loss": 2.4626,
"step": 4520
},
{
"epoch": 0.12,
"grad_norm": 0.8669677376747131,
"learning_rate": 9.676367834367562e-05,
"loss": 2.3765,
"step": 4525
},
{
"epoch": 0.12,
"grad_norm": 0.895280659198761,
"learning_rate": 9.675644065478602e-05,
"loss": 2.4088,
"step": 4530
},
{
"epoch": 0.12,
"grad_norm": 0.8988037109375,
"learning_rate": 9.674919515306904e-05,
"loss": 2.3929,
"step": 4535
},
{
"epoch": 0.12,
"grad_norm": 0.9481714367866516,
"learning_rate": 9.674194183973533e-05,
"loss": 2.4354,
"step": 4540
},
{
"epoch": 0.12,
"grad_norm": 0.9057742357254028,
"learning_rate": 9.673468071599694e-05,
"loss": 2.4485,
"step": 4545
},
{
"epoch": 0.12,
"grad_norm": 0.8659425973892212,
"learning_rate": 9.672741178306713e-05,
"loss": 2.3546,
"step": 4550
},
{
"epoch": 0.12,
"grad_norm": 0.8537487983703613,
"learning_rate": 9.672013504216054e-05,
"loss": 2.4243,
"step": 4555
},
{
"epoch": 0.12,
"grad_norm": 0.8243193626403809,
"learning_rate": 9.671285049449307e-05,
"loss": 2.4512,
"step": 4560
},
{
"epoch": 0.12,
"grad_norm": 0.9560414552688599,
"learning_rate": 9.670555814128197e-05,
"loss": 2.5011,
"step": 4565
},
{
"epoch": 0.12,
"grad_norm": 0.8287144303321838,
"learning_rate": 9.669825798374574e-05,
"loss": 2.4163,
"step": 4570
},
{
"epoch": 0.12,
"grad_norm": 0.9361438751220703,
"learning_rate": 9.66909500231042e-05,
"loss": 2.4068,
"step": 4575
},
{
"epoch": 0.12,
"grad_norm": 0.8917521834373474,
"learning_rate": 9.668363426057852e-05,
"loss": 2.4603,
"step": 4580
},
{
"epoch": 0.12,
"grad_norm": 0.8446766138076782,
"learning_rate": 9.66763106973911e-05,
"loss": 2.4542,
"step": 4585
},
{
"epoch": 0.12,
"grad_norm": 0.9952759146690369,
"learning_rate": 9.666897933476568e-05,
"loss": 2.3388,
"step": 4590
},
{
"epoch": 0.12,
"grad_norm": 0.8720665574073792,
"learning_rate": 9.666164017392734e-05,
"loss": 2.3601,
"step": 4595
},
{
"epoch": 0.12,
"grad_norm": 0.8366141319274902,
"learning_rate": 9.665429321610239e-05,
"loss": 2.4541,
"step": 4600
},
{
"epoch": 0.12,
"grad_norm": 0.882962703704834,
"learning_rate": 9.66469384625185e-05,
"loss": 2.4557,
"step": 4605
},
{
"epoch": 0.12,
"grad_norm": 0.9124892354011536,
"learning_rate": 9.66395759144046e-05,
"loss": 2.5347,
"step": 4610
},
{
"epoch": 0.12,
"grad_norm": 0.8681383728981018,
"learning_rate": 9.6632205572991e-05,
"loss": 2.4696,
"step": 4615
},
{
"epoch": 0.12,
"grad_norm": 0.8890042304992676,
"learning_rate": 9.662482743950918e-05,
"loss": 2.4661,
"step": 4620
},
{
"epoch": 0.12,
"grad_norm": 0.9456043243408203,
"learning_rate": 9.661744151519206e-05,
"loss": 2.3454,
"step": 4625
},
{
"epoch": 0.12,
"grad_norm": 0.8631029725074768,
"learning_rate": 9.661004780127373e-05,
"loss": 2.3083,
"step": 4630
},
{
"epoch": 0.12,
"grad_norm": 0.8786954879760742,
"learning_rate": 9.660264629898974e-05,
"loss": 2.394,
"step": 4635
},
{
"epoch": 0.12,
"grad_norm": 0.8512616753578186,
"learning_rate": 9.65952370095768e-05,
"loss": 2.3873,
"step": 4640
},
{
"epoch": 0.12,
"grad_norm": 0.8391211628913879,
"learning_rate": 9.6587819934273e-05,
"loss": 2.3929,
"step": 4645
},
{
"epoch": 0.12,
"grad_norm": 0.8756142258644104,
"learning_rate": 9.658039507431768e-05,
"loss": 2.5008,
"step": 4650
},
{
"epoch": 0.12,
"grad_norm": 0.8577821850776672,
"learning_rate": 9.657296243095151e-05,
"loss": 2.4363,
"step": 4655
},
{
"epoch": 0.12,
"grad_norm": 0.8333130478858948,
"learning_rate": 9.656552200541648e-05,
"loss": 2.4222,
"step": 4660
},
{
"epoch": 0.12,
"grad_norm": 0.8738192915916443,
"learning_rate": 9.655807379895583e-05,
"loss": 2.519,
"step": 4665
},
{
"epoch": 0.12,
"grad_norm": 0.885319173336029,
"learning_rate": 9.655061781281413e-05,
"loss": 2.3858,
"step": 4670
},
{
"epoch": 0.12,
"grad_norm": 0.8956149220466614,
"learning_rate": 9.654315404823729e-05,
"loss": 2.4966,
"step": 4675
},
{
"epoch": 0.12,
"grad_norm": 0.8953343629837036,
"learning_rate": 9.653568250647242e-05,
"loss": 2.4133,
"step": 4680
},
{
"epoch": 0.12,
"grad_norm": 0.8710033297538757,
"learning_rate": 9.652820318876802e-05,
"loss": 2.4624,
"step": 4685
},
{
"epoch": 0.12,
"grad_norm": 0.7936209440231323,
"learning_rate": 9.652071609637385e-05,
"loss": 2.4319,
"step": 4690
},
{
"epoch": 0.12,
"grad_norm": 0.8369600772857666,
"learning_rate": 9.651322123054097e-05,
"loss": 2.3919,
"step": 4695
},
{
"epoch": 0.12,
"grad_norm": 0.8471620082855225,
"learning_rate": 9.650571859252175e-05,
"loss": 2.332,
"step": 4700
},
{
"epoch": 0.12,
"grad_norm": 0.8544102907180786,
"learning_rate": 9.649820818356985e-05,
"loss": 2.4242,
"step": 4705
},
{
"epoch": 0.12,
"grad_norm": 0.8161907196044922,
"learning_rate": 9.649069000494025e-05,
"loss": 2.5108,
"step": 4710
},
{
"epoch": 0.12,
"grad_norm": 0.8195204138755798,
"learning_rate": 9.648316405788916e-05,
"loss": 2.4493,
"step": 4715
},
{
"epoch": 0.12,
"grad_norm": 0.8137380480766296,
"learning_rate": 9.647563034367419e-05,
"loss": 2.3958,
"step": 4720
},
{
"epoch": 0.12,
"grad_norm": 0.8686335682868958,
"learning_rate": 9.646808886355417e-05,
"loss": 2.3909,
"step": 4725
},
{
"epoch": 0.12,
"grad_norm": 0.880024790763855,
"learning_rate": 9.646053961878925e-05,
"loss": 2.3578,
"step": 4730
},
{
"epoch": 0.12,
"grad_norm": 0.8249338865280151,
"learning_rate": 9.64529826106409e-05,
"loss": 2.3982,
"step": 4735
},
{
"epoch": 0.12,
"grad_norm": 0.893627941608429,
"learning_rate": 9.644541784037186e-05,
"loss": 2.4148,
"step": 4740
},
{
"epoch": 0.12,
"grad_norm": 0.8667566180229187,
"learning_rate": 9.643784530924616e-05,
"loss": 2.4694,
"step": 4745
},
{
"epoch": 0.12,
"grad_norm": 0.8830007314682007,
"learning_rate": 9.643026501852916e-05,
"loss": 2.4399,
"step": 4750
},
{
"epoch": 0.12,
"grad_norm": 0.8822873830795288,
"learning_rate": 9.64226769694875e-05,
"loss": 2.3753,
"step": 4755
},
{
"epoch": 0.12,
"grad_norm": 0.8705997467041016,
"learning_rate": 9.641508116338908e-05,
"loss": 2.4576,
"step": 4760
},
{
"epoch": 0.12,
"grad_norm": 0.7943586707115173,
"learning_rate": 9.640747760150319e-05,
"loss": 2.5037,
"step": 4765
},
{
"epoch": 0.12,
"grad_norm": 0.8014355301856995,
"learning_rate": 9.639986628510032e-05,
"loss": 2.4721,
"step": 4770
},
{
"epoch": 0.12,
"grad_norm": 1.5762988328933716,
"learning_rate": 9.639224721545229e-05,
"loss": 2.4097,
"step": 4775
},
{
"epoch": 0.12,
"grad_norm": 0.8363265991210938,
"learning_rate": 9.638462039383223e-05,
"loss": 2.3963,
"step": 4780
},
{
"epoch": 0.12,
"grad_norm": 0.8247116804122925,
"learning_rate": 9.637698582151454e-05,
"loss": 2.3917,
"step": 4785
},
{
"epoch": 0.12,
"grad_norm": 0.8845296502113342,
"learning_rate": 9.636934349977493e-05,
"loss": 2.4309,
"step": 4790
},
{
"epoch": 0.12,
"grad_norm": 0.8529078960418701,
"learning_rate": 9.636169342989043e-05,
"loss": 2.4233,
"step": 4795
},
{
"epoch": 0.12,
"grad_norm": 0.8616467118263245,
"learning_rate": 9.635403561313932e-05,
"loss": 2.4233,
"step": 4800
},
{
"epoch": 0.12,
"grad_norm": 0.7895869612693787,
"learning_rate": 9.634637005080119e-05,
"loss": 2.4025,
"step": 4805
},
{
"epoch": 0.12,
"grad_norm": 0.8492492437362671,
"learning_rate": 9.633869674415693e-05,
"loss": 2.4591,
"step": 4810
},
{
"epoch": 0.12,
"grad_norm": 0.876369059085846,
"learning_rate": 9.633101569448873e-05,
"loss": 2.4433,
"step": 4815
},
{
"epoch": 0.13,
"grad_norm": 0.8298085331916809,
"learning_rate": 9.632332690308004e-05,
"loss": 2.4239,
"step": 4820
},
{
"epoch": 0.13,
"grad_norm": 0.8470132350921631,
"learning_rate": 9.631563037121565e-05,
"loss": 2.4,
"step": 4825
},
{
"epoch": 0.13,
"grad_norm": 0.8233386278152466,
"learning_rate": 9.63079261001816e-05,
"loss": 2.3847,
"step": 4830
},
{
"epoch": 0.13,
"grad_norm": 0.8454317450523376,
"learning_rate": 9.630021409126529e-05,
"loss": 2.4764,
"step": 4835
},
{
"epoch": 0.13,
"grad_norm": 0.855034589767456,
"learning_rate": 9.629249434575531e-05,
"loss": 2.4442,
"step": 4840
},
{
"epoch": 0.13,
"grad_norm": 0.8333593606948853,
"learning_rate": 9.628476686494164e-05,
"loss": 2.5264,
"step": 4845
},
{
"epoch": 0.13,
"grad_norm": 0.9216281771659851,
"learning_rate": 9.627703165011551e-05,
"loss": 2.325,
"step": 4850
},
{
"epoch": 0.13,
"grad_norm": 0.859026312828064,
"learning_rate": 9.626928870256944e-05,
"loss": 2.3986,
"step": 4855
},
{
"epoch": 0.13,
"grad_norm": 0.8586352467536926,
"learning_rate": 9.626153802359724e-05,
"loss": 2.4059,
"step": 4860
},
{
"epoch": 0.13,
"grad_norm": 0.8783639073371887,
"learning_rate": 9.625377961449402e-05,
"loss": 2.3558,
"step": 4865
},
{
"epoch": 0.13,
"grad_norm": 0.8293476104736328,
"learning_rate": 9.62460134765562e-05,
"loss": 2.3513,
"step": 4870
},
{
"epoch": 0.13,
"grad_norm": 0.8316926956176758,
"learning_rate": 9.623823961108145e-05,
"loss": 2.3654,
"step": 4875
},
{
"epoch": 0.13,
"grad_norm": 0.8805388808250427,
"learning_rate": 9.623045801936876e-05,
"loss": 2.3202,
"step": 4880
},
{
"epoch": 0.13,
"grad_norm": 0.8457925319671631,
"learning_rate": 9.622266870271842e-05,
"loss": 2.3897,
"step": 4885
},
{
"epoch": 0.13,
"grad_norm": 0.8593377470970154,
"learning_rate": 9.6214871662432e-05,
"loss": 2.4816,
"step": 4890
},
{
"epoch": 0.13,
"grad_norm": 0.8373056650161743,
"learning_rate": 9.620706689981232e-05,
"loss": 2.5092,
"step": 4895
},
{
"epoch": 0.13,
"grad_norm": 0.8640280365943909,
"learning_rate": 9.619925441616354e-05,
"loss": 2.4978,
"step": 4900
},
{
"epoch": 0.13,
"grad_norm": 0.9332360029220581,
"learning_rate": 9.619143421279112e-05,
"loss": 2.4424,
"step": 4905
},
{
"epoch": 0.13,
"grad_norm": 0.8037847876548767,
"learning_rate": 9.618360629100176e-05,
"loss": 2.3765,
"step": 4910
},
{
"epoch": 0.13,
"grad_norm": 0.8432718515396118,
"learning_rate": 9.61757706521035e-05,
"loss": 2.4523,
"step": 4915
},
{
"epoch": 0.13,
"grad_norm": 0.918671727180481,
"learning_rate": 9.616792729740564e-05,
"loss": 2.4179,
"step": 4920
},
{
"epoch": 0.13,
"grad_norm": 0.8677783608436584,
"learning_rate": 9.616007622821877e-05,
"loss": 2.4092,
"step": 4925
},
{
"epoch": 0.13,
"grad_norm": 0.920523464679718,
"learning_rate": 9.615221744585476e-05,
"loss": 2.2914,
"step": 4930
},
{
"epoch": 0.13,
"grad_norm": 0.8196969032287598,
"learning_rate": 9.614435095162679e-05,
"loss": 2.3849,
"step": 4935
},
{
"epoch": 0.13,
"grad_norm": 0.8105288743972778,
"learning_rate": 9.613647674684936e-05,
"loss": 2.4038,
"step": 4940
},
{
"epoch": 0.13,
"grad_norm": 0.8498440980911255,
"learning_rate": 9.612859483283818e-05,
"loss": 2.4121,
"step": 4945
},
{
"epoch": 0.13,
"grad_norm": 0.8169273734092712,
"learning_rate": 9.612070521091029e-05,
"loss": 2.4326,
"step": 4950
},
{
"epoch": 0.13,
"grad_norm": 0.8756953477859497,
"learning_rate": 9.611280788238403e-05,
"loss": 2.4162,
"step": 4955
},
{
"epoch": 0.13,
"grad_norm": 1.4346004724502563,
"learning_rate": 9.610490284857901e-05,
"loss": 2.4393,
"step": 4960
},
{
"epoch": 0.13,
"grad_norm": 0.8208816647529602,
"learning_rate": 9.609699011081613e-05,
"loss": 2.4174,
"step": 4965
},
{
"epoch": 0.13,
"grad_norm": 0.863175094127655,
"learning_rate": 9.608906967041757e-05,
"loss": 2.3892,
"step": 4970
},
{
"epoch": 0.13,
"grad_norm": 0.8554673790931702,
"learning_rate": 9.608114152870684e-05,
"loss": 2.5032,
"step": 4975
},
{
"epoch": 0.13,
"grad_norm": 0.9026063084602356,
"learning_rate": 9.607320568700866e-05,
"loss": 2.4662,
"step": 4980
},
{
"epoch": 0.13,
"grad_norm": 0.8296844363212585,
"learning_rate": 9.606526214664911e-05,
"loss": 2.5421,
"step": 4985
},
{
"epoch": 0.13,
"grad_norm": 0.8953242301940918,
"learning_rate": 9.605731090895551e-05,
"loss": 2.3417,
"step": 4990
},
{
"epoch": 0.13,
"grad_norm": 0.820361316204071,
"learning_rate": 9.604935197525649e-05,
"loss": 2.3605,
"step": 4995
},
{
"epoch": 0.13,
"grad_norm": 0.7860215306282043,
"learning_rate": 9.604138534688196e-05,
"loss": 2.4265,
"step": 5000
},
{
"epoch": 0.13,
"grad_norm": 0.9001255035400391,
"learning_rate": 9.60334110251631e-05,
"loss": 2.4446,
"step": 5005
},
{
"epoch": 0.13,
"grad_norm": 0.8390194177627563,
"learning_rate": 9.602542901143241e-05,
"loss": 2.3931,
"step": 5010
},
{
"epoch": 0.13,
"grad_norm": 0.9073073863983154,
"learning_rate": 9.601743930702363e-05,
"loss": 2.4316,
"step": 5015
},
{
"epoch": 0.13,
"grad_norm": 0.9029976725578308,
"learning_rate": 9.600944191327184e-05,
"loss": 2.3532,
"step": 5020
},
{
"epoch": 0.13,
"grad_norm": 0.8546518683433533,
"learning_rate": 9.600143683151335e-05,
"loss": 2.4221,
"step": 5025
},
{
"epoch": 0.13,
"grad_norm": 0.992438554763794,
"learning_rate": 9.59934240630858e-05,
"loss": 2.375,
"step": 5030
},
{
"epoch": 0.13,
"grad_norm": 0.7874085903167725,
"learning_rate": 9.598540360932806e-05,
"loss": 2.4026,
"step": 5035
},
{
"epoch": 0.13,
"grad_norm": 0.9013950228691101,
"learning_rate": 9.597737547158035e-05,
"loss": 2.3942,
"step": 5040
},
{
"epoch": 0.13,
"grad_norm": 0.8906729817390442,
"learning_rate": 9.596933965118413e-05,
"loss": 2.4734,
"step": 5045
},
{
"epoch": 0.13,
"grad_norm": 0.8925601243972778,
"learning_rate": 9.596129614948217e-05,
"loss": 2.3457,
"step": 5050
},
{
"epoch": 0.13,
"grad_norm": 0.8798192143440247,
"learning_rate": 9.595324496781849e-05,
"loss": 2.4337,
"step": 5055
},
{
"epoch": 0.13,
"grad_norm": 0.8773041367530823,
"learning_rate": 9.594518610753841e-05,
"loss": 2.4753,
"step": 5060
},
{
"epoch": 0.13,
"grad_norm": 0.9097132086753845,
"learning_rate": 9.593711956998855e-05,
"loss": 2.3639,
"step": 5065
},
{
"epoch": 0.13,
"grad_norm": 0.8333487510681152,
"learning_rate": 9.59290453565168e-05,
"loss": 2.4109,
"step": 5070
},
{
"epoch": 0.13,
"grad_norm": 0.8247542977333069,
"learning_rate": 9.592096346847232e-05,
"loss": 2.3068,
"step": 5075
},
{
"epoch": 0.13,
"grad_norm": 0.8491074442863464,
"learning_rate": 9.591287390720555e-05,
"loss": 2.4586,
"step": 5080
},
{
"epoch": 0.13,
"grad_norm": 0.8207966089248657,
"learning_rate": 9.590477667406828e-05,
"loss": 2.2868,
"step": 5085
},
{
"epoch": 0.13,
"grad_norm": 0.8393003344535828,
"learning_rate": 9.589667177041347e-05,
"loss": 2.3409,
"step": 5090
},
{
"epoch": 0.13,
"grad_norm": 0.9081759452819824,
"learning_rate": 9.588855919759544e-05,
"loss": 2.4144,
"step": 5095
},
{
"epoch": 0.13,
"grad_norm": 0.831897497177124,
"learning_rate": 9.588043895696978e-05,
"loss": 2.4398,
"step": 5100
},
{
"epoch": 0.13,
"grad_norm": 0.8371878862380981,
"learning_rate": 9.587231104989334e-05,
"loss": 2.3835,
"step": 5105
},
{
"epoch": 0.13,
"grad_norm": 0.8361806273460388,
"learning_rate": 9.586417547772426e-05,
"loss": 2.4154,
"step": 5110
},
{
"epoch": 0.13,
"grad_norm": 0.8459285497665405,
"learning_rate": 9.585603224182196e-05,
"loss": 2.4353,
"step": 5115
},
{
"epoch": 0.13,
"grad_norm": 0.8751489520072937,
"learning_rate": 9.584788134354718e-05,
"loss": 2.3827,
"step": 5120
},
{
"epoch": 0.13,
"grad_norm": 0.8890305757522583,
"learning_rate": 9.583972278426186e-05,
"loss": 2.3462,
"step": 5125
},
{
"epoch": 0.13,
"grad_norm": 0.8343215584754944,
"learning_rate": 9.583155656532929e-05,
"loss": 2.4225,
"step": 5130
},
{
"epoch": 0.13,
"grad_norm": 0.8756371140480042,
"learning_rate": 9.582338268811402e-05,
"loss": 2.4842,
"step": 5135
},
{
"epoch": 0.13,
"grad_norm": 0.8187376260757446,
"learning_rate": 9.581520115398185e-05,
"loss": 2.4373,
"step": 5140
},
{
"epoch": 0.13,
"grad_norm": 0.8343190550804138,
"learning_rate": 9.580701196429991e-05,
"loss": 2.3284,
"step": 5145
},
{
"epoch": 0.13,
"grad_norm": 0.921565592288971,
"learning_rate": 9.579881512043655e-05,
"loss": 2.4031,
"step": 5150
},
{
"epoch": 0.13,
"grad_norm": 0.8414582014083862,
"learning_rate": 9.579061062376146e-05,
"loss": 2.4685,
"step": 5155
},
{
"epoch": 0.13,
"grad_norm": 0.8647580146789551,
"learning_rate": 9.578239847564559e-05,
"loss": 2.4379,
"step": 5160
},
{
"epoch": 0.13,
"grad_norm": 0.8450123071670532,
"learning_rate": 9.577417867746113e-05,
"loss": 2.4694,
"step": 5165
},
{
"epoch": 0.13,
"grad_norm": 1.0390018224716187,
"learning_rate": 9.576595123058158e-05,
"loss": 2.4803,
"step": 5170
},
{
"epoch": 0.13,
"grad_norm": 0.7833945751190186,
"learning_rate": 9.575771613638175e-05,
"loss": 2.4428,
"step": 5175
},
{
"epoch": 0.13,
"grad_norm": 0.8745269179344177,
"learning_rate": 9.574947339623765e-05,
"loss": 2.3924,
"step": 5180
},
{
"epoch": 0.13,
"grad_norm": 0.8687005043029785,
"learning_rate": 9.574122301152665e-05,
"loss": 2.3626,
"step": 5185
},
{
"epoch": 0.13,
"grad_norm": 0.8375335335731506,
"learning_rate": 9.573296498362733e-05,
"loss": 2.4774,
"step": 5190
},
{
"epoch": 0.13,
"grad_norm": 0.8410646915435791,
"learning_rate": 9.572469931391959e-05,
"loss": 2.2626,
"step": 5195
},
{
"epoch": 0.13,
"grad_norm": 0.8926162123680115,
"learning_rate": 9.571642600378459e-05,
"loss": 2.409,
"step": 5200
},
{
"epoch": 0.14,
"grad_norm": 0.8314776420593262,
"learning_rate": 9.570814505460475e-05,
"loss": 2.4115,
"step": 5205
},
{
"epoch": 0.14,
"grad_norm": 0.9310240149497986,
"learning_rate": 9.569985646776382e-05,
"loss": 2.4583,
"step": 5210
},
{
"epoch": 0.14,
"grad_norm": 0.8787447810173035,
"learning_rate": 9.569156024464677e-05,
"loss": 2.3867,
"step": 5215
},
{
"epoch": 0.14,
"grad_norm": 0.8731285929679871,
"learning_rate": 9.568325638663989e-05,
"loss": 2.4226,
"step": 5220
},
{
"epoch": 0.14,
"grad_norm": 0.8139272928237915,
"learning_rate": 9.567494489513068e-05,
"loss": 2.4497,
"step": 5225
},
{
"epoch": 0.14,
"grad_norm": 0.8846359252929688,
"learning_rate": 9.5666625771508e-05,
"loss": 2.4603,
"step": 5230
},
{
"epoch": 0.14,
"grad_norm": 0.8469163775444031,
"learning_rate": 9.565829901716194e-05,
"loss": 2.4448,
"step": 5235
},
{
"epoch": 0.14,
"grad_norm": 0.8273356556892395,
"learning_rate": 9.564996463348384e-05,
"loss": 2.3304,
"step": 5240
},
{
"epoch": 0.14,
"grad_norm": 0.9029586315155029,
"learning_rate": 9.564162262186638e-05,
"loss": 2.4431,
"step": 5245
},
{
"epoch": 0.14,
"grad_norm": 0.8696768283843994,
"learning_rate": 9.563327298370347e-05,
"loss": 2.4398,
"step": 5250
},
{
"epoch": 0.14,
"grad_norm": 0.9131995439529419,
"learning_rate": 9.56249157203903e-05,
"loss": 2.4918,
"step": 5255
},
{
"epoch": 0.14,
"grad_norm": 1.076788306236267,
"learning_rate": 9.561655083332333e-05,
"loss": 2.4044,
"step": 5260
},
{
"epoch": 0.14,
"grad_norm": 0.8568449020385742,
"learning_rate": 9.56081783239003e-05,
"loss": 2.4635,
"step": 5265
},
{
"epoch": 0.14,
"grad_norm": 0.9562225937843323,
"learning_rate": 9.559979819352024e-05,
"loss": 2.4157,
"step": 5270
},
{
"epoch": 0.14,
"grad_norm": 0.8881519436836243,
"learning_rate": 9.559141044358342e-05,
"loss": 2.3643,
"step": 5275
},
{
"epoch": 0.14,
"grad_norm": 0.8735615611076355,
"learning_rate": 9.558301507549143e-05,
"loss": 2.4874,
"step": 5280
},
{
"epoch": 0.14,
"grad_norm": 0.8221045136451721,
"learning_rate": 9.557461209064708e-05,
"loss": 2.4089,
"step": 5285
},
{
"epoch": 0.14,
"grad_norm": 0.9125104546546936,
"learning_rate": 9.556620149045447e-05,
"loss": 2.4367,
"step": 5290
},
{
"epoch": 0.14,
"grad_norm": 0.8167206048965454,
"learning_rate": 9.555778327631903e-05,
"loss": 2.4211,
"step": 5295
},
{
"epoch": 0.14,
"grad_norm": 0.9294907450675964,
"learning_rate": 9.554935744964734e-05,
"loss": 2.4991,
"step": 5300
},
{
"epoch": 0.14,
"grad_norm": 0.7857155203819275,
"learning_rate": 9.55409240118474e-05,
"loss": 2.3603,
"step": 5305
},
{
"epoch": 0.14,
"grad_norm": 0.8441645503044128,
"learning_rate": 9.553248296432835e-05,
"loss": 2.375,
"step": 5310
},
{
"epoch": 0.14,
"grad_norm": 0.939723014831543,
"learning_rate": 9.552403430850068e-05,
"loss": 2.4186,
"step": 5315
},
{
"epoch": 0.14,
"grad_norm": 0.8851372599601746,
"learning_rate": 9.551557804577612e-05,
"loss": 2.448,
"step": 5320
},
{
"epoch": 0.14,
"grad_norm": 0.8540197014808655,
"learning_rate": 9.550711417756767e-05,
"loss": 2.3335,
"step": 5325
},
{
"epoch": 0.14,
"grad_norm": 0.868716835975647,
"learning_rate": 9.549864270528965e-05,
"loss": 2.4521,
"step": 5330
},
{
"epoch": 0.14,
"grad_norm": 0.8275147080421448,
"learning_rate": 9.549016363035758e-05,
"loss": 2.4463,
"step": 5335
},
{
"epoch": 0.14,
"grad_norm": 0.8285613059997559,
"learning_rate": 9.548167695418828e-05,
"loss": 2.3492,
"step": 5340
},
{
"epoch": 0.14,
"grad_norm": 0.8636640906333923,
"learning_rate": 9.547318267819986e-05,
"loss": 2.3231,
"step": 5345
},
{
"epoch": 0.14,
"grad_norm": 0.8481895923614502,
"learning_rate": 9.546468080381167e-05,
"loss": 2.3599,
"step": 5350
},
{
"epoch": 0.14,
"grad_norm": 0.8118917942047119,
"learning_rate": 9.545617133244436e-05,
"loss": 2.3275,
"step": 5355
},
{
"epoch": 0.14,
"grad_norm": 0.8362302780151367,
"learning_rate": 9.544765426551981e-05,
"loss": 2.4139,
"step": 5360
},
{
"epoch": 0.14,
"grad_norm": 0.8462435007095337,
"learning_rate": 9.543912960446117e-05,
"loss": 2.3218,
"step": 5365
},
{
"epoch": 0.14,
"grad_norm": 0.84576416015625,
"learning_rate": 9.543059735069293e-05,
"loss": 2.3285,
"step": 5370
},
{
"epoch": 0.14,
"grad_norm": 0.8824775815010071,
"learning_rate": 9.542205750564077e-05,
"loss": 2.3518,
"step": 5375
},
{
"epoch": 0.14,
"grad_norm": 0.8731521964073181,
"learning_rate": 9.541351007073167e-05,
"loss": 2.4394,
"step": 5380
},
{
"epoch": 0.14,
"grad_norm": 0.891345739364624,
"learning_rate": 9.540495504739387e-05,
"loss": 2.3309,
"step": 5385
},
{
"epoch": 0.14,
"grad_norm": 0.9005354046821594,
"learning_rate": 9.53963924370569e-05,
"loss": 2.3669,
"step": 5390
},
{
"epoch": 0.14,
"grad_norm": 0.8092679381370544,
"learning_rate": 9.538782224115151e-05,
"loss": 2.508,
"step": 5395
},
{
"epoch": 0.14,
"grad_norm": 0.9163539409637451,
"learning_rate": 9.537924446110976e-05,
"loss": 2.4135,
"step": 5400
},
{
"epoch": 0.14,
"grad_norm": 0.9057278633117676,
"learning_rate": 9.537065909836497e-05,
"loss": 2.3537,
"step": 5405
},
{
"epoch": 0.14,
"grad_norm": 0.8037895560264587,
"learning_rate": 9.536206615435174e-05,
"loss": 2.3672,
"step": 5410
},
{
"epoch": 0.14,
"grad_norm": 0.8482769727706909,
"learning_rate": 9.535346563050591e-05,
"loss": 2.3835,
"step": 5415
},
{
"epoch": 0.14,
"grad_norm": 0.881119966506958,
"learning_rate": 9.534485752826456e-05,
"loss": 2.4201,
"step": 5420
},
{
"epoch": 0.14,
"grad_norm": 0.8554364442825317,
"learning_rate": 9.53362418490661e-05,
"loss": 2.3647,
"step": 5425
},
{
"epoch": 0.14,
"grad_norm": 0.8279440999031067,
"learning_rate": 9.53276185943502e-05,
"loss": 2.4034,
"step": 5430
},
{
"epoch": 0.14,
"grad_norm": 0.8798325061798096,
"learning_rate": 9.531898776555773e-05,
"loss": 2.4637,
"step": 5435
},
{
"epoch": 0.14,
"grad_norm": 0.9144737720489502,
"learning_rate": 9.531034936413091e-05,
"loss": 2.3528,
"step": 5440
},
{
"epoch": 0.14,
"grad_norm": 0.9441506266593933,
"learning_rate": 9.530170339151314e-05,
"loss": 2.3386,
"step": 5445
},
{
"epoch": 0.14,
"grad_norm": 0.8675415515899658,
"learning_rate": 9.529304984914917e-05,
"loss": 2.3789,
"step": 5450
},
{
"epoch": 0.14,
"grad_norm": 0.8356438279151917,
"learning_rate": 9.528438873848496e-05,
"loss": 2.3441,
"step": 5455
},
{
"epoch": 0.14,
"grad_norm": 0.8329806327819824,
"learning_rate": 9.527572006096773e-05,
"loss": 2.4517,
"step": 5460
},
{
"epoch": 0.14,
"grad_norm": 0.8664317727088928,
"learning_rate": 9.526704381804601e-05,
"loss": 2.451,
"step": 5465
},
{
"epoch": 0.14,
"grad_norm": 0.8535704612731934,
"learning_rate": 9.525836001116957e-05,
"loss": 2.4739,
"step": 5470
},
{
"epoch": 0.14,
"grad_norm": 0.8236183524131775,
"learning_rate": 9.524966864178943e-05,
"loss": 2.3595,
"step": 5475
},
{
"epoch": 0.14,
"grad_norm": 0.8680240511894226,
"learning_rate": 9.524096971135789e-05,
"loss": 2.4479,
"step": 5480
},
{
"epoch": 0.14,
"grad_norm": 0.8324035406112671,
"learning_rate": 9.52322632213285e-05,
"loss": 2.4416,
"step": 5485
},
{
"epoch": 0.14,
"grad_norm": 0.8540217280387878,
"learning_rate": 9.52235491731561e-05,
"loss": 2.4568,
"step": 5490
},
{
"epoch": 0.14,
"grad_norm": 0.860846221446991,
"learning_rate": 9.521482756829675e-05,
"loss": 2.3877,
"step": 5495
},
{
"epoch": 0.14,
"grad_norm": 0.885197103023529,
"learning_rate": 9.520609840820782e-05,
"loss": 2.3848,
"step": 5500
},
{
"epoch": 0.14,
"grad_norm": 0.9146736264228821,
"learning_rate": 9.51973616943479e-05,
"loss": 2.369,
"step": 5505
},
{
"epoch": 0.14,
"grad_norm": 0.8666872382164001,
"learning_rate": 9.518861742817689e-05,
"loss": 2.4614,
"step": 5510
},
{
"epoch": 0.14,
"grad_norm": 0.813156247138977,
"learning_rate": 9.51798656111559e-05,
"loss": 2.2819,
"step": 5515
},
{
"epoch": 0.14,
"grad_norm": 0.8812664151191711,
"learning_rate": 9.517110624474733e-05,
"loss": 2.3471,
"step": 5520
},
{
"epoch": 0.14,
"grad_norm": 0.8435032963752747,
"learning_rate": 9.516233933041485e-05,
"loss": 2.434,
"step": 5525
},
{
"epoch": 0.14,
"grad_norm": 0.810698390007019,
"learning_rate": 9.515356486962337e-05,
"loss": 2.401,
"step": 5530
},
{
"epoch": 0.14,
"grad_norm": 0.798914909362793,
"learning_rate": 9.514478286383907e-05,
"loss": 2.4458,
"step": 5535
},
{
"epoch": 0.14,
"grad_norm": 0.8657212257385254,
"learning_rate": 9.513599331452939e-05,
"loss": 2.3425,
"step": 5540
},
{
"epoch": 0.14,
"grad_norm": 0.8055309653282166,
"learning_rate": 9.512719622316304e-05,
"loss": 2.3766,
"step": 5545
},
{
"epoch": 0.14,
"grad_norm": 0.8028724193572998,
"learning_rate": 9.511839159120997e-05,
"loss": 2.4554,
"step": 5550
},
{
"epoch": 0.14,
"grad_norm": 0.8116658926010132,
"learning_rate": 9.51095794201414e-05,
"loss": 2.344,
"step": 5555
},
{
"epoch": 0.14,
"grad_norm": 0.8904930949211121,
"learning_rate": 9.510075971142983e-05,
"loss": 2.4775,
"step": 5560
},
{
"epoch": 0.14,
"grad_norm": 0.8095250725746155,
"learning_rate": 9.509193246654896e-05,
"loss": 2.2933,
"step": 5565
},
{
"epoch": 0.14,
"grad_norm": 0.8514060974121094,
"learning_rate": 9.508309768697384e-05,
"loss": 2.3374,
"step": 5570
},
{
"epoch": 0.14,
"grad_norm": 0.8403069376945496,
"learning_rate": 9.507425537418068e-05,
"loss": 2.4266,
"step": 5575
},
{
"epoch": 0.14,
"grad_norm": 0.8308573365211487,
"learning_rate": 9.506540552964704e-05,
"loss": 2.3418,
"step": 5580
},
{
"epoch": 0.14,
"grad_norm": 0.8367452621459961,
"learning_rate": 9.505654815485168e-05,
"loss": 2.3762,
"step": 5585
},
{
"epoch": 0.15,
"grad_norm": 0.8345729112625122,
"learning_rate": 9.504768325127464e-05,
"loss": 2.3245,
"step": 5590
},
{
"epoch": 0.15,
"grad_norm": 0.8398026823997498,
"learning_rate": 9.503881082039719e-05,
"loss": 2.3968,
"step": 5595
},
{
"epoch": 0.15,
"grad_norm": 0.8237159252166748,
"learning_rate": 9.50299308637019e-05,
"loss": 2.4709,
"step": 5600
},
{
"epoch": 0.15,
"grad_norm": 0.9180028438568115,
"learning_rate": 9.502104338267257e-05,
"loss": 2.4171,
"step": 5605
},
{
"epoch": 0.15,
"grad_norm": 0.9185792207717896,
"learning_rate": 9.501214837879428e-05,
"loss": 2.3786,
"step": 5610
},
{
"epoch": 0.15,
"grad_norm": 0.8425090909004211,
"learning_rate": 9.500324585355333e-05,
"loss": 2.4082,
"step": 5615
},
{
"epoch": 0.15,
"grad_norm": 0.833187460899353,
"learning_rate": 9.499433580843732e-05,
"loss": 2.3611,
"step": 5620
},
{
"epoch": 0.15,
"grad_norm": 0.8175890445709229,
"learning_rate": 9.498541824493506e-05,
"loss": 2.3007,
"step": 5625
},
{
"epoch": 0.15,
"grad_norm": 0.8270918130874634,
"learning_rate": 9.497649316453667e-05,
"loss": 2.3619,
"step": 5630
},
{
"epoch": 0.15,
"grad_norm": 0.8552674651145935,
"learning_rate": 9.496756056873348e-05,
"loss": 2.3625,
"step": 5635
},
{
"epoch": 0.15,
"grad_norm": 0.8825513124465942,
"learning_rate": 9.49586204590181e-05,
"loss": 2.4104,
"step": 5640
},
{
"epoch": 0.15,
"grad_norm": 0.8108384013175964,
"learning_rate": 9.494967283688438e-05,
"loss": 2.3842,
"step": 5645
},
{
"epoch": 0.15,
"grad_norm": 0.8291341066360474,
"learning_rate": 9.494071770382745e-05,
"loss": 2.437,
"step": 5650
},
{
"epoch": 0.15,
"grad_norm": 0.8763828873634338,
"learning_rate": 9.493175506134367e-05,
"loss": 2.3643,
"step": 5655
},
{
"epoch": 0.15,
"grad_norm": 0.8425805568695068,
"learning_rate": 9.492278491093067e-05,
"loss": 2.4306,
"step": 5660
},
{
"epoch": 0.15,
"grad_norm": 0.8033353090286255,
"learning_rate": 9.491380725408731e-05,
"loss": 2.383,
"step": 5665
},
{
"epoch": 0.15,
"grad_norm": 0.8733636140823364,
"learning_rate": 9.490482209231374e-05,
"loss": 2.3754,
"step": 5670
},
{
"epoch": 0.15,
"grad_norm": 0.8381609320640564,
"learning_rate": 9.489582942711135e-05,
"loss": 2.39,
"step": 5675
},
{
"epoch": 0.15,
"grad_norm": 0.8547597527503967,
"learning_rate": 9.488682925998277e-05,
"loss": 2.3022,
"step": 5680
},
{
"epoch": 0.15,
"grad_norm": 0.9090614914894104,
"learning_rate": 9.48778215924319e-05,
"loss": 2.3305,
"step": 5685
},
{
"epoch": 0.15,
"grad_norm": 0.8316418528556824,
"learning_rate": 9.48688064259639e-05,
"loss": 2.495,
"step": 5690
},
{
"epoch": 0.15,
"grad_norm": 0.8660643696784973,
"learning_rate": 9.485978376208513e-05,
"loss": 2.4094,
"step": 5695
},
{
"epoch": 0.15,
"grad_norm": 0.8256070613861084,
"learning_rate": 9.485075360230328e-05,
"loss": 2.4195,
"step": 5700
},
{
"epoch": 0.15,
"grad_norm": 0.8589950203895569,
"learning_rate": 9.484171594812728e-05,
"loss": 2.4616,
"step": 5705
},
{
"epoch": 0.15,
"grad_norm": 0.8481250405311584,
"learning_rate": 9.483267080106723e-05,
"loss": 2.4128,
"step": 5710
},
{
"epoch": 0.15,
"grad_norm": 0.8542687892913818,
"learning_rate": 9.482361816263456e-05,
"loss": 2.4031,
"step": 5715
},
{
"epoch": 0.15,
"grad_norm": 0.8733797669410706,
"learning_rate": 9.481455803434195e-05,
"loss": 2.3558,
"step": 5720
},
{
"epoch": 0.15,
"grad_norm": 0.8200867176055908,
"learning_rate": 9.48054904177033e-05,
"loss": 2.291,
"step": 5725
},
{
"epoch": 0.15,
"grad_norm": 0.7754344940185547,
"learning_rate": 9.479641531423377e-05,
"loss": 2.395,
"step": 5730
},
{
"epoch": 0.15,
"grad_norm": 0.8544303774833679,
"learning_rate": 9.47873327254498e-05,
"loss": 2.3965,
"step": 5735
},
{
"epoch": 0.15,
"grad_norm": 0.8686069846153259,
"learning_rate": 9.477824265286904e-05,
"loss": 2.3689,
"step": 5740
},
{
"epoch": 0.15,
"grad_norm": 0.8045094013214111,
"learning_rate": 9.476914509801039e-05,
"loss": 2.2953,
"step": 5745
},
{
"epoch": 0.15,
"grad_norm": 0.856065034866333,
"learning_rate": 9.476004006239406e-05,
"loss": 2.3688,
"step": 5750
},
{
"epoch": 0.15,
"grad_norm": 0.8922398090362549,
"learning_rate": 9.475092754754142e-05,
"loss": 2.3764,
"step": 5755
},
{
"epoch": 0.15,
"grad_norm": 0.8231756687164307,
"learning_rate": 9.474180755497517e-05,
"loss": 2.3771,
"step": 5760
},
{
"epoch": 0.15,
"grad_norm": 0.8228115439414978,
"learning_rate": 9.473268008621922e-05,
"loss": 2.4069,
"step": 5765
},
{
"epoch": 0.15,
"grad_norm": 0.8378354907035828,
"learning_rate": 9.472354514279874e-05,
"loss": 2.4223,
"step": 5770
},
{
"epoch": 0.15,
"grad_norm": 0.8504692912101746,
"learning_rate": 9.471440272624012e-05,
"loss": 2.3624,
"step": 5775
},
{
"epoch": 0.15,
"grad_norm": 0.854158341884613,
"learning_rate": 9.470525283807105e-05,
"loss": 2.4345,
"step": 5780
},
{
"epoch": 0.15,
"grad_norm": 0.7879740595817566,
"learning_rate": 9.469609547982044e-05,
"loss": 2.355,
"step": 5785
},
{
"epoch": 0.15,
"grad_norm": 0.8204787969589233,
"learning_rate": 9.468693065301843e-05,
"loss": 2.3315,
"step": 5790
},
{
"epoch": 0.15,
"grad_norm": 0.8343956470489502,
"learning_rate": 9.467775835919646e-05,
"loss": 2.4381,
"step": 5795
},
{
"epoch": 0.15,
"grad_norm": 0.7959950566291809,
"learning_rate": 9.466857859988715e-05,
"loss": 2.3968,
"step": 5800
},
{
"epoch": 0.15,
"grad_norm": 0.8729004859924316,
"learning_rate": 9.465939137662444e-05,
"loss": 2.4851,
"step": 5805
},
{
"epoch": 0.15,
"grad_norm": 0.8949657678604126,
"learning_rate": 9.465019669094346e-05,
"loss": 2.3684,
"step": 5810
},
{
"epoch": 0.15,
"grad_norm": 0.8578696250915527,
"learning_rate": 9.464099454438062e-05,
"loss": 2.2861,
"step": 5815
},
{
"epoch": 0.15,
"grad_norm": 0.8324920535087585,
"learning_rate": 9.463178493847355e-05,
"loss": 2.3758,
"step": 5820
},
{
"epoch": 0.15,
"grad_norm": 0.8466063141822815,
"learning_rate": 9.462256787476113e-05,
"loss": 2.3744,
"step": 5825
},
{
"epoch": 0.15,
"grad_norm": 0.8376399874687195,
"learning_rate": 9.461334335478354e-05,
"loss": 2.3634,
"step": 5830
},
{
"epoch": 0.15,
"grad_norm": 1.0192224979400635,
"learning_rate": 9.460411138008214e-05,
"loss": 2.3808,
"step": 5835
},
{
"epoch": 0.15,
"grad_norm": 0.8341616988182068,
"learning_rate": 9.459487195219954e-05,
"loss": 2.4642,
"step": 5840
},
{
"epoch": 0.15,
"grad_norm": 0.8986044526100159,
"learning_rate": 9.458562507267963e-05,
"loss": 2.4415,
"step": 5845
},
{
"epoch": 0.15,
"grad_norm": 0.8127945065498352,
"learning_rate": 9.457637074306754e-05,
"loss": 2.3559,
"step": 5850
},
{
"epoch": 0.15,
"grad_norm": 0.8740310072898865,
"learning_rate": 9.456710896490961e-05,
"loss": 2.3456,
"step": 5855
},
{
"epoch": 0.15,
"grad_norm": 0.8706247210502625,
"learning_rate": 9.455783973975348e-05,
"loss": 2.3633,
"step": 5860
},
{
"epoch": 0.15,
"grad_norm": 0.8536202907562256,
"learning_rate": 9.454856306914797e-05,
"loss": 2.3113,
"step": 5865
},
{
"epoch": 0.15,
"grad_norm": 0.758399486541748,
"learning_rate": 9.453927895464321e-05,
"loss": 2.3561,
"step": 5870
},
{
"epoch": 0.15,
"grad_norm": 0.871759295463562,
"learning_rate": 9.452998739779051e-05,
"loss": 2.3648,
"step": 5875
},
{
"epoch": 0.15,
"grad_norm": 0.8724145293235779,
"learning_rate": 9.452068840014246e-05,
"loss": 2.4498,
"step": 5880
},
{
"epoch": 0.15,
"grad_norm": 0.8810963034629822,
"learning_rate": 9.451138196325291e-05,
"loss": 2.185,
"step": 5885
},
{
"epoch": 0.15,
"grad_norm": 0.8311077952384949,
"learning_rate": 9.45020680886769e-05,
"loss": 2.4236,
"step": 5890
},
{
"epoch": 0.15,
"grad_norm": 0.8102297782897949,
"learning_rate": 9.449274677797078e-05,
"loss": 2.3686,
"step": 5895
},
{
"epoch": 0.15,
"grad_norm": 0.857153594493866,
"learning_rate": 9.448341803269208e-05,
"loss": 2.4149,
"step": 5900
},
{
"epoch": 0.15,
"grad_norm": 0.8286229968070984,
"learning_rate": 9.44740818543996e-05,
"loss": 2.3668,
"step": 5905
},
{
"epoch": 0.15,
"grad_norm": 0.9215847253799438,
"learning_rate": 9.44647382446534e-05,
"loss": 2.4859,
"step": 5910
},
{
"epoch": 0.15,
"grad_norm": 0.8779950141906738,
"learning_rate": 9.445538720501474e-05,
"loss": 2.3774,
"step": 5915
},
{
"epoch": 0.15,
"grad_norm": 0.8534978032112122,
"learning_rate": 9.444602873704616e-05,
"loss": 2.4151,
"step": 5920
},
{
"epoch": 0.15,
"grad_norm": 0.83034348487854,
"learning_rate": 9.443666284231141e-05,
"loss": 2.359,
"step": 5925
},
{
"epoch": 0.15,
"grad_norm": 0.8255096077919006,
"learning_rate": 9.44272895223755e-05,
"loss": 2.3566,
"step": 5930
},
{
"epoch": 0.15,
"grad_norm": 0.8525983691215515,
"learning_rate": 9.441790877880469e-05,
"loss": 2.3946,
"step": 5935
},
{
"epoch": 0.15,
"grad_norm": 0.8380722403526306,
"learning_rate": 9.440852061316645e-05,
"loss": 2.3668,
"step": 5940
},
{
"epoch": 0.15,
"grad_norm": 0.8316784501075745,
"learning_rate": 9.439912502702952e-05,
"loss": 2.4024,
"step": 5945
},
{
"epoch": 0.15,
"grad_norm": 0.8253279328346252,
"learning_rate": 9.438972202196387e-05,
"loss": 2.4858,
"step": 5950
},
{
"epoch": 0.15,
"grad_norm": 0.8625275492668152,
"learning_rate": 9.43803115995407e-05,
"loss": 2.3946,
"step": 5955
},
{
"epoch": 0.15,
"grad_norm": 0.8443737030029297,
"learning_rate": 9.437089376133246e-05,
"loss": 2.3721,
"step": 5960
},
{
"epoch": 0.15,
"grad_norm": 0.8400998711585999,
"learning_rate": 9.436146850891283e-05,
"loss": 2.386,
"step": 5965
},
{
"epoch": 0.15,
"grad_norm": 0.8208600282669067,
"learning_rate": 9.435203584385673e-05,
"loss": 2.3492,
"step": 5970
},
{
"epoch": 0.16,
"grad_norm": 0.8670238852500916,
"learning_rate": 9.434259576774034e-05,
"loss": 2.4416,
"step": 5975
},
{
"epoch": 0.16,
"grad_norm": 0.8226633667945862,
"learning_rate": 9.433314828214103e-05,
"loss": 2.3806,
"step": 5980
},
{
"epoch": 0.16,
"grad_norm": 0.8466386198997498,
"learning_rate": 9.43236933886375e-05,
"loss": 2.3521,
"step": 5985
},
{
"epoch": 0.16,
"grad_norm": 0.9443104863166809,
"learning_rate": 9.431423108880957e-05,
"loss": 2.422,
"step": 5990
},
{
"epoch": 0.16,
"grad_norm": 0.886673092842102,
"learning_rate": 9.430476138423838e-05,
"loss": 2.3856,
"step": 5995
},
{
"epoch": 0.16,
"grad_norm": 0.8009757399559021,
"learning_rate": 9.429528427650627e-05,
"loss": 2.4325,
"step": 6000
}
],
"logging_steps": 5,
"max_steps": 38527,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"total_flos": 1.7134867590414336e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}