Qwen1.5B-L36-MTP4 / trainer_state.json
ZMC2019's picture
Model save
ae0f101 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.994876571960876,
"eval_steps": 500,
"global_step": 6438,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004657661853749418,
"grad_norm": 14.037870104148775,
"learning_rate": 3.8819875776397516e-07,
"loss": 2.6559,
"step": 5
},
{
"epoch": 0.009315323707498836,
"grad_norm": 14.018507695640166,
"learning_rate": 7.763975155279503e-07,
"loss": 2.6439,
"step": 10
},
{
"epoch": 0.013972985561248253,
"grad_norm": 13.74585203517481,
"learning_rate": 1.1645962732919254e-06,
"loss": 2.6109,
"step": 15
},
{
"epoch": 0.018630647414997672,
"grad_norm": 12.26602342821938,
"learning_rate": 1.5527950310559006e-06,
"loss": 2.541,
"step": 20
},
{
"epoch": 0.02328830926874709,
"grad_norm": 8.985034614318803,
"learning_rate": 1.940993788819876e-06,
"loss": 2.365,
"step": 25
},
{
"epoch": 0.027945971122496506,
"grad_norm": 8.518106704276775,
"learning_rate": 2.329192546583851e-06,
"loss": 2.2655,
"step": 30
},
{
"epoch": 0.032603632976245925,
"grad_norm": 7.692662664956911,
"learning_rate": 2.7173913043478263e-06,
"loss": 2.2177,
"step": 35
},
{
"epoch": 0.037261294829995344,
"grad_norm": 2.2695875679058966,
"learning_rate": 3.1055900621118013e-06,
"loss": 2.1003,
"step": 40
},
{
"epoch": 0.04191895668374476,
"grad_norm": 2.7959358163596217,
"learning_rate": 3.4937888198757763e-06,
"loss": 1.9482,
"step": 45
},
{
"epoch": 0.04657661853749418,
"grad_norm": 2.1750434470602715,
"learning_rate": 3.881987577639752e-06,
"loss": 1.9068,
"step": 50
},
{
"epoch": 0.05123428039124359,
"grad_norm": 1.1232980728931912,
"learning_rate": 4.270186335403727e-06,
"loss": 1.8378,
"step": 55
},
{
"epoch": 0.05589194224499301,
"grad_norm": 0.8825405994938238,
"learning_rate": 4.658385093167702e-06,
"loss": 1.7925,
"step": 60
},
{
"epoch": 0.06054960409874243,
"grad_norm": 0.8987047049592403,
"learning_rate": 5.046583850931677e-06,
"loss": 1.7318,
"step": 65
},
{
"epoch": 0.06520726595249185,
"grad_norm": 0.7265996503150173,
"learning_rate": 5.4347826086956525e-06,
"loss": 1.65,
"step": 70
},
{
"epoch": 0.06986492780624126,
"grad_norm": 0.6528101497449033,
"learning_rate": 5.8229813664596275e-06,
"loss": 1.6243,
"step": 75
},
{
"epoch": 0.07452258965999069,
"grad_norm": 0.7224211410313209,
"learning_rate": 6.2111801242236025e-06,
"loss": 1.6201,
"step": 80
},
{
"epoch": 0.0791802515137401,
"grad_norm": 0.5806690237524301,
"learning_rate": 6.5993788819875775e-06,
"loss": 1.5475,
"step": 85
},
{
"epoch": 0.08383791336748952,
"grad_norm": 0.597058825325402,
"learning_rate": 6.9875776397515525e-06,
"loss": 1.5384,
"step": 90
},
{
"epoch": 0.08849557522123894,
"grad_norm": 0.55670574011478,
"learning_rate": 7.375776397515528e-06,
"loss": 1.5362,
"step": 95
},
{
"epoch": 0.09315323707498836,
"grad_norm": 0.5393976696535505,
"learning_rate": 7.763975155279503e-06,
"loss": 1.4973,
"step": 100
},
{
"epoch": 0.09781089892873777,
"grad_norm": 0.5204227077686031,
"learning_rate": 8.15217391304348e-06,
"loss": 1.4845,
"step": 105
},
{
"epoch": 0.10246856078248719,
"grad_norm": 0.5563822350418148,
"learning_rate": 8.540372670807453e-06,
"loss": 1.4439,
"step": 110
},
{
"epoch": 0.10712622263623661,
"grad_norm": 0.5181030795462002,
"learning_rate": 8.92857142857143e-06,
"loss": 1.4305,
"step": 115
},
{
"epoch": 0.11178388448998602,
"grad_norm": 0.5731923823330213,
"learning_rate": 9.316770186335403e-06,
"loss": 1.439,
"step": 120
},
{
"epoch": 0.11644154634373545,
"grad_norm": 0.582681646921711,
"learning_rate": 9.70496894409938e-06,
"loss": 1.3811,
"step": 125
},
{
"epoch": 0.12109920819748486,
"grad_norm": 0.5474901747641734,
"learning_rate": 1.0093167701863353e-05,
"loss": 1.3784,
"step": 130
},
{
"epoch": 0.1257568700512343,
"grad_norm": 0.5932031228323433,
"learning_rate": 1.048136645962733e-05,
"loss": 1.3829,
"step": 135
},
{
"epoch": 0.1304145319049837,
"grad_norm": 0.5513551481294792,
"learning_rate": 1.0869565217391305e-05,
"loss": 1.3528,
"step": 140
},
{
"epoch": 0.1350721937587331,
"grad_norm": 0.5778299522858029,
"learning_rate": 1.1257763975155281e-05,
"loss": 1.3657,
"step": 145
},
{
"epoch": 0.13972985561248252,
"grad_norm": 0.6252635575293012,
"learning_rate": 1.1645962732919255e-05,
"loss": 1.3365,
"step": 150
},
{
"epoch": 0.14438751746623196,
"grad_norm": 0.6131491406232994,
"learning_rate": 1.2034161490683231e-05,
"loss": 1.2873,
"step": 155
},
{
"epoch": 0.14904517931998137,
"grad_norm": 0.6853390178591717,
"learning_rate": 1.2422360248447205e-05,
"loss": 1.2928,
"step": 160
},
{
"epoch": 0.1537028411737308,
"grad_norm": 0.6954345622753815,
"learning_rate": 1.2810559006211181e-05,
"loss": 1.2843,
"step": 165
},
{
"epoch": 0.1583605030274802,
"grad_norm": 0.7535183287534551,
"learning_rate": 1.3198757763975155e-05,
"loss": 1.2709,
"step": 170
},
{
"epoch": 0.1630181648812296,
"grad_norm": 0.6598685647547688,
"learning_rate": 1.3586956521739131e-05,
"loss": 1.2856,
"step": 175
},
{
"epoch": 0.16767582673497905,
"grad_norm": 0.6918935778616571,
"learning_rate": 1.3975155279503105e-05,
"loss": 1.2668,
"step": 180
},
{
"epoch": 0.17233348858872846,
"grad_norm": 0.6301889269605553,
"learning_rate": 1.4363354037267083e-05,
"loss": 1.2513,
"step": 185
},
{
"epoch": 0.17699115044247787,
"grad_norm": 0.8764303684748792,
"learning_rate": 1.4751552795031057e-05,
"loss": 1.2342,
"step": 190
},
{
"epoch": 0.18164881229622729,
"grad_norm": 0.8939970938191784,
"learning_rate": 1.5139751552795031e-05,
"loss": 1.2355,
"step": 195
},
{
"epoch": 0.18630647414997673,
"grad_norm": 0.6866794293645567,
"learning_rate": 1.5527950310559007e-05,
"loss": 1.1672,
"step": 200
},
{
"epoch": 0.19096413600372614,
"grad_norm": 0.649144682338704,
"learning_rate": 1.5916149068322984e-05,
"loss": 1.2104,
"step": 205
},
{
"epoch": 0.19562179785747555,
"grad_norm": 0.8218693043165641,
"learning_rate": 1.630434782608696e-05,
"loss": 1.1818,
"step": 210
},
{
"epoch": 0.20027945971122496,
"grad_norm": 0.7430490952040086,
"learning_rate": 1.6692546583850933e-05,
"loss": 1.1996,
"step": 215
},
{
"epoch": 0.20493712156497437,
"grad_norm": 0.8125678825371954,
"learning_rate": 1.7080745341614907e-05,
"loss": 1.1879,
"step": 220
},
{
"epoch": 0.2095947834187238,
"grad_norm": 0.676480033060189,
"learning_rate": 1.7468944099378884e-05,
"loss": 1.1874,
"step": 225
},
{
"epoch": 0.21425244527247322,
"grad_norm": 0.8515544361485938,
"learning_rate": 1.785714285714286e-05,
"loss": 1.1789,
"step": 230
},
{
"epoch": 0.21891010712622264,
"grad_norm": 0.6704690025578643,
"learning_rate": 1.8245341614906833e-05,
"loss": 1.1653,
"step": 235
},
{
"epoch": 0.22356776897997205,
"grad_norm": 0.8706610468125755,
"learning_rate": 1.8633540372670807e-05,
"loss": 1.1587,
"step": 240
},
{
"epoch": 0.22822543083372146,
"grad_norm": 0.5898112084526719,
"learning_rate": 1.9021739130434784e-05,
"loss": 1.133,
"step": 245
},
{
"epoch": 0.2328830926874709,
"grad_norm": 1.106503362987722,
"learning_rate": 1.940993788819876e-05,
"loss": 1.1191,
"step": 250
},
{
"epoch": 0.2375407545412203,
"grad_norm": 1.0286327080054072,
"learning_rate": 1.9798136645962733e-05,
"loss": 1.1433,
"step": 255
},
{
"epoch": 0.24219841639496972,
"grad_norm": 0.8407919099301676,
"learning_rate": 2.0186335403726707e-05,
"loss": 1.1506,
"step": 260
},
{
"epoch": 0.24685607824871914,
"grad_norm": 0.8654728309548365,
"learning_rate": 2.0574534161490684e-05,
"loss": 1.1308,
"step": 265
},
{
"epoch": 0.2515137401024686,
"grad_norm": 0.7695850121818572,
"learning_rate": 2.096273291925466e-05,
"loss": 1.1041,
"step": 270
},
{
"epoch": 0.25617140195621796,
"grad_norm": 0.7910585274047307,
"learning_rate": 2.1350931677018636e-05,
"loss": 1.1376,
"step": 275
},
{
"epoch": 0.2608290638099674,
"grad_norm": 0.7458734894767046,
"learning_rate": 2.173913043478261e-05,
"loss": 1.1132,
"step": 280
},
{
"epoch": 0.26548672566371684,
"grad_norm": 0.8187339847365924,
"learning_rate": 2.2127329192546584e-05,
"loss": 1.1015,
"step": 285
},
{
"epoch": 0.2701443875174662,
"grad_norm": 0.763447142888627,
"learning_rate": 2.2515527950310562e-05,
"loss": 1.096,
"step": 290
},
{
"epoch": 0.27480204937121566,
"grad_norm": 0.7934188486978877,
"learning_rate": 2.2903726708074536e-05,
"loss": 1.0807,
"step": 295
},
{
"epoch": 0.27945971122496505,
"grad_norm": 0.788453952770843,
"learning_rate": 2.329192546583851e-05,
"loss": 1.1043,
"step": 300
},
{
"epoch": 0.2841173730787145,
"grad_norm": 0.7047893933708144,
"learning_rate": 2.3680124223602484e-05,
"loss": 1.1136,
"step": 305
},
{
"epoch": 0.2887750349324639,
"grad_norm": 0.795257449558012,
"learning_rate": 2.4068322981366462e-05,
"loss": 1.0977,
"step": 310
},
{
"epoch": 0.2934326967862133,
"grad_norm": 0.6915051544535833,
"learning_rate": 2.4456521739130436e-05,
"loss": 1.0722,
"step": 315
},
{
"epoch": 0.29809035863996275,
"grad_norm": 0.7265299997807345,
"learning_rate": 2.484472049689441e-05,
"loss": 1.0911,
"step": 320
},
{
"epoch": 0.30274802049371213,
"grad_norm": 0.7601977613919204,
"learning_rate": 2.5232919254658388e-05,
"loss": 1.0928,
"step": 325
},
{
"epoch": 0.3074056823474616,
"grad_norm": 0.8099690160950118,
"learning_rate": 2.5621118012422362e-05,
"loss": 1.0879,
"step": 330
},
{
"epoch": 0.312063344201211,
"grad_norm": 0.6886893057959771,
"learning_rate": 2.6009316770186336e-05,
"loss": 1.0453,
"step": 335
},
{
"epoch": 0.3167210060549604,
"grad_norm": 0.7367664406007456,
"learning_rate": 2.639751552795031e-05,
"loss": 1.0748,
"step": 340
},
{
"epoch": 0.32137866790870984,
"grad_norm": 0.6783866452713414,
"learning_rate": 2.6785714285714288e-05,
"loss": 1.0507,
"step": 345
},
{
"epoch": 0.3260363297624592,
"grad_norm": 0.9161119691045607,
"learning_rate": 2.7173913043478262e-05,
"loss": 1.054,
"step": 350
},
{
"epoch": 0.33069399161620866,
"grad_norm": 0.7644762711765951,
"learning_rate": 2.7562111801242236e-05,
"loss": 1.0447,
"step": 355
},
{
"epoch": 0.3353516534699581,
"grad_norm": 0.7491094719090783,
"learning_rate": 2.795031055900621e-05,
"loss": 1.0648,
"step": 360
},
{
"epoch": 0.3400093153237075,
"grad_norm": 0.7464676647377744,
"learning_rate": 2.833850931677019e-05,
"loss": 1.0619,
"step": 365
},
{
"epoch": 0.3446669771774569,
"grad_norm": 0.8302793649617403,
"learning_rate": 2.8726708074534165e-05,
"loss": 1.0842,
"step": 370
},
{
"epoch": 0.3493246390312063,
"grad_norm": 0.7553615087579988,
"learning_rate": 2.911490683229814e-05,
"loss": 1.0676,
"step": 375
},
{
"epoch": 0.35398230088495575,
"grad_norm": 0.7122417970022175,
"learning_rate": 2.9503105590062114e-05,
"loss": 1.04,
"step": 380
},
{
"epoch": 0.3586399627387052,
"grad_norm": 0.7963778926214264,
"learning_rate": 2.9891304347826088e-05,
"loss": 1.0258,
"step": 385
},
{
"epoch": 0.36329762459245457,
"grad_norm": 0.8039906902669227,
"learning_rate": 3.0279503105590062e-05,
"loss": 1.0318,
"step": 390
},
{
"epoch": 0.367955286446204,
"grad_norm": 0.828586879177044,
"learning_rate": 3.066770186335404e-05,
"loss": 1.0176,
"step": 395
},
{
"epoch": 0.37261294829995345,
"grad_norm": 0.7741728323429156,
"learning_rate": 3.1055900621118014e-05,
"loss": 1.0147,
"step": 400
},
{
"epoch": 0.37727061015370283,
"grad_norm": 0.7854840295592591,
"learning_rate": 3.1444099378881995e-05,
"loss": 1.0193,
"step": 405
},
{
"epoch": 0.3819282720074523,
"grad_norm": 0.7952699696224765,
"learning_rate": 3.183229813664597e-05,
"loss": 1.0299,
"step": 410
},
{
"epoch": 0.38658593386120166,
"grad_norm": 0.8433374305657526,
"learning_rate": 3.222049689440994e-05,
"loss": 1.0284,
"step": 415
},
{
"epoch": 0.3912435957149511,
"grad_norm": 0.8256286197983435,
"learning_rate": 3.260869565217392e-05,
"loss": 1.0079,
"step": 420
},
{
"epoch": 0.39590125756870054,
"grad_norm": 0.7671147431831079,
"learning_rate": 3.299689440993789e-05,
"loss": 1.037,
"step": 425
},
{
"epoch": 0.4005589194224499,
"grad_norm": 0.6726151565927365,
"learning_rate": 3.3385093167701865e-05,
"loss": 1.0252,
"step": 430
},
{
"epoch": 0.40521658127619936,
"grad_norm": 0.912517511617104,
"learning_rate": 3.377329192546584e-05,
"loss": 1.0278,
"step": 435
},
{
"epoch": 0.40987424312994875,
"grad_norm": 0.8685516813400533,
"learning_rate": 3.4161490683229814e-05,
"loss": 1.0481,
"step": 440
},
{
"epoch": 0.4145319049836982,
"grad_norm": 0.7671909964824053,
"learning_rate": 3.4549689440993795e-05,
"loss": 0.9964,
"step": 445
},
{
"epoch": 0.4191895668374476,
"grad_norm": 1.0135652385996252,
"learning_rate": 3.493788819875777e-05,
"loss": 1.0028,
"step": 450
},
{
"epoch": 0.423847228691197,
"grad_norm": 0.899973975312835,
"learning_rate": 3.532608695652174e-05,
"loss": 1.0448,
"step": 455
},
{
"epoch": 0.42850489054494645,
"grad_norm": 0.9156532713655958,
"learning_rate": 3.571428571428572e-05,
"loss": 1.0024,
"step": 460
},
{
"epoch": 0.43316255239869583,
"grad_norm": 1.1365454340055587,
"learning_rate": 3.610248447204969e-05,
"loss": 1.0213,
"step": 465
},
{
"epoch": 0.43782021425244527,
"grad_norm": 0.9918128476101238,
"learning_rate": 3.6490683229813665e-05,
"loss": 1.002,
"step": 470
},
{
"epoch": 0.4424778761061947,
"grad_norm": 0.9151718575356949,
"learning_rate": 3.687888198757764e-05,
"loss": 1.0194,
"step": 475
},
{
"epoch": 0.4471355379599441,
"grad_norm": 0.742255103116907,
"learning_rate": 3.7267080745341614e-05,
"loss": 0.9891,
"step": 480
},
{
"epoch": 0.45179319981369354,
"grad_norm": 0.8586376934224837,
"learning_rate": 3.765527950310559e-05,
"loss": 0.9955,
"step": 485
},
{
"epoch": 0.4564508616674429,
"grad_norm": 0.9103531228057,
"learning_rate": 3.804347826086957e-05,
"loss": 1.0018,
"step": 490
},
{
"epoch": 0.46110852352119236,
"grad_norm": 0.7448983584328042,
"learning_rate": 3.843167701863354e-05,
"loss": 0.9823,
"step": 495
},
{
"epoch": 0.4657661853749418,
"grad_norm": 0.7349324323567711,
"learning_rate": 3.881987577639752e-05,
"loss": 0.9967,
"step": 500
},
{
"epoch": 0.4704238472286912,
"grad_norm": 0.695545445845715,
"learning_rate": 3.920807453416149e-05,
"loss": 0.983,
"step": 505
},
{
"epoch": 0.4750815090824406,
"grad_norm": 0.9785822896697502,
"learning_rate": 3.9596273291925465e-05,
"loss": 0.9726,
"step": 510
},
{
"epoch": 0.47973917093619,
"grad_norm": 1.0555748680465804,
"learning_rate": 3.998447204968944e-05,
"loss": 0.9996,
"step": 515
},
{
"epoch": 0.48439683278993945,
"grad_norm": 0.966587096450481,
"learning_rate": 4.0372670807453414e-05,
"loss": 0.9898,
"step": 520
},
{
"epoch": 0.4890544946436889,
"grad_norm": 0.910161503630037,
"learning_rate": 4.076086956521739e-05,
"loss": 1.0064,
"step": 525
},
{
"epoch": 0.49371215649743827,
"grad_norm": 0.9152897977721635,
"learning_rate": 4.114906832298137e-05,
"loss": 0.9759,
"step": 530
},
{
"epoch": 0.4983698183511877,
"grad_norm": 0.8668603504226737,
"learning_rate": 4.153726708074534e-05,
"loss": 0.9891,
"step": 535
},
{
"epoch": 0.5030274802049371,
"grad_norm": 0.8468939833128593,
"learning_rate": 4.192546583850932e-05,
"loss": 0.9987,
"step": 540
},
{
"epoch": 0.5076851420586865,
"grad_norm": 0.731218604929643,
"learning_rate": 4.23136645962733e-05,
"loss": 0.9713,
"step": 545
},
{
"epoch": 0.5123428039124359,
"grad_norm": 0.9843894198247464,
"learning_rate": 4.270186335403727e-05,
"loss": 1.0025,
"step": 550
},
{
"epoch": 0.5170004657661854,
"grad_norm": 0.7631176615609254,
"learning_rate": 4.3090062111801246e-05,
"loss": 0.9754,
"step": 555
},
{
"epoch": 0.5216581276199348,
"grad_norm": 0.9494172654241818,
"learning_rate": 4.347826086956522e-05,
"loss": 0.9847,
"step": 560
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.758600818741894,
"learning_rate": 4.3866459627329195e-05,
"loss": 0.981,
"step": 565
},
{
"epoch": 0.5309734513274337,
"grad_norm": 0.6523938368399681,
"learning_rate": 4.425465838509317e-05,
"loss": 0.9766,
"step": 570
},
{
"epoch": 0.5356311131811831,
"grad_norm": 0.8196799702305156,
"learning_rate": 4.464285714285715e-05,
"loss": 0.9726,
"step": 575
},
{
"epoch": 0.5402887750349324,
"grad_norm": 0.7875301746371101,
"learning_rate": 4.5031055900621124e-05,
"loss": 0.9716,
"step": 580
},
{
"epoch": 0.5449464368886818,
"grad_norm": 0.9366890391534561,
"learning_rate": 4.54192546583851e-05,
"loss": 0.9639,
"step": 585
},
{
"epoch": 0.5496040987424313,
"grad_norm": 0.8159290575760685,
"learning_rate": 4.580745341614907e-05,
"loss": 0.9959,
"step": 590
},
{
"epoch": 0.5542617605961807,
"grad_norm": 0.7904038693901743,
"learning_rate": 4.6195652173913046e-05,
"loss": 0.9777,
"step": 595
},
{
"epoch": 0.5589194224499301,
"grad_norm": 0.8820141458513885,
"learning_rate": 4.658385093167702e-05,
"loss": 0.987,
"step": 600
},
{
"epoch": 0.5635770843036796,
"grad_norm": 0.6581899070095203,
"learning_rate": 4.6972049689440995e-05,
"loss": 0.9313,
"step": 605
},
{
"epoch": 0.568234746157429,
"grad_norm": 0.6921366906956813,
"learning_rate": 4.736024844720497e-05,
"loss": 0.9711,
"step": 610
},
{
"epoch": 0.5728924080111784,
"grad_norm": 1.0400083726029987,
"learning_rate": 4.774844720496895e-05,
"loss": 0.9567,
"step": 615
},
{
"epoch": 0.5775500698649279,
"grad_norm": 1.0259646153662887,
"learning_rate": 4.8136645962732924e-05,
"loss": 0.9636,
"step": 620
},
{
"epoch": 0.5822077317186772,
"grad_norm": 0.8898990676451768,
"learning_rate": 4.85248447204969e-05,
"loss": 0.9842,
"step": 625
},
{
"epoch": 0.5868653935724266,
"grad_norm": 0.9603886153911195,
"learning_rate": 4.891304347826087e-05,
"loss": 0.9552,
"step": 630
},
{
"epoch": 0.5915230554261761,
"grad_norm": 0.7643568136993506,
"learning_rate": 4.9301242236024846e-05,
"loss": 0.951,
"step": 635
},
{
"epoch": 0.5961807172799255,
"grad_norm": 0.7613193066594093,
"learning_rate": 4.968944099378882e-05,
"loss": 0.9648,
"step": 640
},
{
"epoch": 0.6008383791336749,
"grad_norm": 0.8024613228030518,
"learning_rate": 4.999137038315499e-05,
"loss": 0.963,
"step": 645
},
{
"epoch": 0.6054960409874243,
"grad_norm": 0.8661787138644403,
"learning_rate": 4.994822229892993e-05,
"loss": 0.9544,
"step": 650
},
{
"epoch": 0.6101537028411738,
"grad_norm": 0.9059824800026692,
"learning_rate": 4.9905074214704875e-05,
"loss": 0.9679,
"step": 655
},
{
"epoch": 0.6148113646949231,
"grad_norm": 0.8439746120342664,
"learning_rate": 4.986192613047981e-05,
"loss": 0.9739,
"step": 660
},
{
"epoch": 0.6194690265486725,
"grad_norm": 1.0683925183560257,
"learning_rate": 4.9818778046254746e-05,
"loss": 0.9585,
"step": 665
},
{
"epoch": 0.624126688402422,
"grad_norm": 0.6613910069447578,
"learning_rate": 4.977562996202969e-05,
"loss": 0.9645,
"step": 670
},
{
"epoch": 0.6287843502561714,
"grad_norm": 0.7575403562844513,
"learning_rate": 4.9732481877804624e-05,
"loss": 0.9441,
"step": 675
},
{
"epoch": 0.6334420121099208,
"grad_norm": 0.7124161020635701,
"learning_rate": 4.968933379357957e-05,
"loss": 0.9522,
"step": 680
},
{
"epoch": 0.6380996739636703,
"grad_norm": 0.7797416220084484,
"learning_rate": 4.964618570935451e-05,
"loss": 0.9767,
"step": 685
},
{
"epoch": 0.6427573358174197,
"grad_norm": 0.7611566850882846,
"learning_rate": 4.9603037625129445e-05,
"loss": 0.9508,
"step": 690
},
{
"epoch": 0.6474149976711691,
"grad_norm": 0.6799999335018179,
"learning_rate": 4.955988954090439e-05,
"loss": 0.9512,
"step": 695
},
{
"epoch": 0.6520726595249184,
"grad_norm": 0.8631080619144671,
"learning_rate": 4.951674145667933e-05,
"loss": 0.9445,
"step": 700
},
{
"epoch": 0.6567303213786679,
"grad_norm": 0.7799699034483604,
"learning_rate": 4.9473593372454266e-05,
"loss": 0.9364,
"step": 705
},
{
"epoch": 0.6613879832324173,
"grad_norm": 0.8550493933043184,
"learning_rate": 4.94304452882292e-05,
"loss": 0.9609,
"step": 710
},
{
"epoch": 0.6660456450861667,
"grad_norm": 0.7299471562792063,
"learning_rate": 4.9387297204004144e-05,
"loss": 0.9489,
"step": 715
},
{
"epoch": 0.6707033069399162,
"grad_norm": 0.8935263071636578,
"learning_rate": 4.934414911977908e-05,
"loss": 0.9283,
"step": 720
},
{
"epoch": 0.6753609687936656,
"grad_norm": 0.6888098427802909,
"learning_rate": 4.930100103555402e-05,
"loss": 0.939,
"step": 725
},
{
"epoch": 0.680018630647415,
"grad_norm": 0.7995262285128696,
"learning_rate": 4.9257852951328965e-05,
"loss": 0.9539,
"step": 730
},
{
"epoch": 0.6846762925011645,
"grad_norm": 0.7947129512554866,
"learning_rate": 4.921470486710391e-05,
"loss": 0.9293,
"step": 735
},
{
"epoch": 0.6893339543549138,
"grad_norm": 0.779585099199785,
"learning_rate": 4.917155678287884e-05,
"loss": 0.9405,
"step": 740
},
{
"epoch": 0.6939916162086632,
"grad_norm": 0.8303446372378965,
"learning_rate": 4.9128408698653785e-05,
"loss": 0.9314,
"step": 745
},
{
"epoch": 0.6986492780624126,
"grad_norm": 1.139024040895227,
"learning_rate": 4.908526061442872e-05,
"loss": 0.9389,
"step": 750
},
{
"epoch": 0.7033069399161621,
"grad_norm": 0.7605700848959173,
"learning_rate": 4.904211253020366e-05,
"loss": 0.9178,
"step": 755
},
{
"epoch": 0.7079646017699115,
"grad_norm": 0.688563607175027,
"learning_rate": 4.89989644459786e-05,
"loss": 0.9215,
"step": 760
},
{
"epoch": 0.7126222636236609,
"grad_norm": 0.6181695501552628,
"learning_rate": 4.895581636175354e-05,
"loss": 0.9419,
"step": 765
},
{
"epoch": 0.7172799254774104,
"grad_norm": 0.6622209131129708,
"learning_rate": 4.891266827752848e-05,
"loss": 0.9408,
"step": 770
},
{
"epoch": 0.7219375873311598,
"grad_norm": 0.6582361407570896,
"learning_rate": 4.886952019330342e-05,
"loss": 0.915,
"step": 775
},
{
"epoch": 0.7265952491849091,
"grad_norm": 0.6803939066291419,
"learning_rate": 4.882637210907836e-05,
"loss": 0.9487,
"step": 780
},
{
"epoch": 0.7312529110386586,
"grad_norm": 0.6720553599662832,
"learning_rate": 4.87832240248533e-05,
"loss": 0.9369,
"step": 785
},
{
"epoch": 0.735910572892408,
"grad_norm": 0.7316521366862686,
"learning_rate": 4.874007594062824e-05,
"loss": 0.9198,
"step": 790
},
{
"epoch": 0.7405682347461574,
"grad_norm": 0.5950973796307311,
"learning_rate": 4.8696927856403176e-05,
"loss": 0.9336,
"step": 795
},
{
"epoch": 0.7452258965999069,
"grad_norm": 0.733870802001813,
"learning_rate": 4.865377977217811e-05,
"loss": 0.9235,
"step": 800
},
{
"epoch": 0.7498835584536563,
"grad_norm": 0.6129379811403421,
"learning_rate": 4.8610631687953055e-05,
"loss": 0.9569,
"step": 805
},
{
"epoch": 0.7545412203074057,
"grad_norm": 0.591475313672276,
"learning_rate": 4.8567483603728e-05,
"loss": 0.9244,
"step": 810
},
{
"epoch": 0.759198882161155,
"grad_norm": 0.6044956006930611,
"learning_rate": 4.852433551950293e-05,
"loss": 0.9314,
"step": 815
},
{
"epoch": 0.7638565440149045,
"grad_norm": 0.7664178521354441,
"learning_rate": 4.8481187435277875e-05,
"loss": 0.9293,
"step": 820
},
{
"epoch": 0.7685142058686539,
"grad_norm": 0.7107416917500197,
"learning_rate": 4.843803935105282e-05,
"loss": 0.9055,
"step": 825
},
{
"epoch": 0.7731718677224033,
"grad_norm": 0.7618528783533417,
"learning_rate": 4.839489126682776e-05,
"loss": 0.9209,
"step": 830
},
{
"epoch": 0.7778295295761528,
"grad_norm": 0.7258800949018822,
"learning_rate": 4.8351743182602696e-05,
"loss": 0.9345,
"step": 835
},
{
"epoch": 0.7824871914299022,
"grad_norm": 0.6783953124506613,
"learning_rate": 4.830859509837763e-05,
"loss": 0.904,
"step": 840
},
{
"epoch": 0.7871448532836516,
"grad_norm": 0.6836122062349896,
"learning_rate": 4.8265447014152574e-05,
"loss": 0.918,
"step": 845
},
{
"epoch": 0.7918025151374011,
"grad_norm": 0.6507985445046874,
"learning_rate": 4.822229892992751e-05,
"loss": 0.9203,
"step": 850
},
{
"epoch": 0.7964601769911505,
"grad_norm": 0.6643585008530555,
"learning_rate": 4.817915084570245e-05,
"loss": 0.9002,
"step": 855
},
{
"epoch": 0.8011178388448998,
"grad_norm": 0.6054726374838411,
"learning_rate": 4.8136002761477395e-05,
"loss": 0.9094,
"step": 860
},
{
"epoch": 0.8057755006986492,
"grad_norm": 0.650341215526472,
"learning_rate": 4.809285467725233e-05,
"loss": 0.9123,
"step": 865
},
{
"epoch": 0.8104331625523987,
"grad_norm": 0.6506499441912863,
"learning_rate": 4.804970659302727e-05,
"loss": 0.8989,
"step": 870
},
{
"epoch": 0.8150908244061481,
"grad_norm": 0.8580048917270087,
"learning_rate": 4.8006558508802216e-05,
"loss": 0.9226,
"step": 875
},
{
"epoch": 0.8197484862598975,
"grad_norm": 0.6055892662420432,
"learning_rate": 4.796341042457715e-05,
"loss": 0.8859,
"step": 880
},
{
"epoch": 0.824406148113647,
"grad_norm": 0.7139422196799956,
"learning_rate": 4.792026234035209e-05,
"loss": 0.9232,
"step": 885
},
{
"epoch": 0.8290638099673964,
"grad_norm": 0.6089661735907969,
"learning_rate": 4.787711425612703e-05,
"loss": 0.919,
"step": 890
},
{
"epoch": 0.8337214718211458,
"grad_norm": 0.580454247375901,
"learning_rate": 4.7833966171901966e-05,
"loss": 0.9171,
"step": 895
},
{
"epoch": 0.8383791336748952,
"grad_norm": 0.5849797012971689,
"learning_rate": 4.779081808767691e-05,
"loss": 0.9055,
"step": 900
},
{
"epoch": 0.8430367955286446,
"grad_norm": 0.6008467270294285,
"learning_rate": 4.774767000345185e-05,
"loss": 0.914,
"step": 905
},
{
"epoch": 0.847694457382394,
"grad_norm": 0.5537296928970568,
"learning_rate": 4.770452191922679e-05,
"loss": 0.9286,
"step": 910
},
{
"epoch": 0.8523521192361434,
"grad_norm": 0.6492423617592114,
"learning_rate": 4.766137383500173e-05,
"loss": 0.9057,
"step": 915
},
{
"epoch": 0.8570097810898929,
"grad_norm": 0.62560716037305,
"learning_rate": 4.761822575077667e-05,
"loss": 0.9245,
"step": 920
},
{
"epoch": 0.8616674429436423,
"grad_norm": 0.7082200597011241,
"learning_rate": 4.757507766655161e-05,
"loss": 0.9103,
"step": 925
},
{
"epoch": 0.8663251047973917,
"grad_norm": 0.6957809388430001,
"learning_rate": 4.753192958232654e-05,
"loss": 0.9158,
"step": 930
},
{
"epoch": 0.8709827666511412,
"grad_norm": 0.5595395190034976,
"learning_rate": 4.7488781498101485e-05,
"loss": 0.8892,
"step": 935
},
{
"epoch": 0.8756404285048905,
"grad_norm": 0.7398962569929148,
"learning_rate": 4.744563341387643e-05,
"loss": 0.8897,
"step": 940
},
{
"epoch": 0.8802980903586399,
"grad_norm": 0.6298362220627544,
"learning_rate": 4.7402485329651363e-05,
"loss": 0.8979,
"step": 945
},
{
"epoch": 0.8849557522123894,
"grad_norm": 0.6392904649774634,
"learning_rate": 4.7359337245426306e-05,
"loss": 0.8749,
"step": 950
},
{
"epoch": 0.8896134140661388,
"grad_norm": 0.5910452010752443,
"learning_rate": 4.731618916120125e-05,
"loss": 0.9037,
"step": 955
},
{
"epoch": 0.8942710759198882,
"grad_norm": 0.5724046276347944,
"learning_rate": 4.7273041076976184e-05,
"loss": 0.9139,
"step": 960
},
{
"epoch": 0.8989287377736377,
"grad_norm": 0.6169462890930449,
"learning_rate": 4.722989299275113e-05,
"loss": 0.903,
"step": 965
},
{
"epoch": 0.9035863996273871,
"grad_norm": 0.5556544386361997,
"learning_rate": 4.718674490852606e-05,
"loss": 0.9093,
"step": 970
},
{
"epoch": 0.9082440614811365,
"grad_norm": 0.6590333956524478,
"learning_rate": 4.7143596824301e-05,
"loss": 0.8983,
"step": 975
},
{
"epoch": 0.9129017233348858,
"grad_norm": 0.748488058996119,
"learning_rate": 4.710044874007594e-05,
"loss": 0.9043,
"step": 980
},
{
"epoch": 0.9175593851886353,
"grad_norm": 0.606400629600308,
"learning_rate": 4.705730065585088e-05,
"loss": 0.9128,
"step": 985
},
{
"epoch": 0.9222170470423847,
"grad_norm": 0.5852937118094048,
"learning_rate": 4.7014152571625826e-05,
"loss": 0.8554,
"step": 990
},
{
"epoch": 0.9268747088961341,
"grad_norm": 0.5476277605999371,
"learning_rate": 4.697100448740076e-05,
"loss": 0.8878,
"step": 995
},
{
"epoch": 0.9315323707498836,
"grad_norm": 0.6330430492588804,
"learning_rate": 4.6927856403175704e-05,
"loss": 0.8924,
"step": 1000
},
{
"epoch": 0.936190032603633,
"grad_norm": 0.7023864398254552,
"learning_rate": 4.6884708318950646e-05,
"loss": 0.9133,
"step": 1005
},
{
"epoch": 0.9408476944573824,
"grad_norm": 0.5921577951396615,
"learning_rate": 4.684156023472558e-05,
"loss": 0.8913,
"step": 1010
},
{
"epoch": 0.9455053563111319,
"grad_norm": 0.6151573442083585,
"learning_rate": 4.679841215050052e-05,
"loss": 0.8973,
"step": 1015
},
{
"epoch": 0.9501630181648812,
"grad_norm": 0.5967475819345346,
"learning_rate": 4.675526406627546e-05,
"loss": 0.9138,
"step": 1020
},
{
"epoch": 0.9548206800186306,
"grad_norm": 0.6013234118124504,
"learning_rate": 4.6712115982050396e-05,
"loss": 0.8966,
"step": 1025
},
{
"epoch": 0.95947834187238,
"grad_norm": 0.6608553067855343,
"learning_rate": 4.666896789782534e-05,
"loss": 0.88,
"step": 1030
},
{
"epoch": 0.9641360037261295,
"grad_norm": 0.6695319084789703,
"learning_rate": 4.662581981360028e-05,
"loss": 0.887,
"step": 1035
},
{
"epoch": 0.9687936655798789,
"grad_norm": 0.5744412033836418,
"learning_rate": 4.658267172937522e-05,
"loss": 0.8807,
"step": 1040
},
{
"epoch": 0.9734513274336283,
"grad_norm": 0.5550351398683305,
"learning_rate": 4.653952364515016e-05,
"loss": 0.8856,
"step": 1045
},
{
"epoch": 0.9781089892873778,
"grad_norm": 0.5022448639615988,
"learning_rate": 4.64963755609251e-05,
"loss": 0.8831,
"step": 1050
},
{
"epoch": 0.9827666511411272,
"grad_norm": 0.5621280358732187,
"learning_rate": 4.645322747670004e-05,
"loss": 0.8882,
"step": 1055
},
{
"epoch": 0.9874243129948765,
"grad_norm": 0.5680029674592303,
"learning_rate": 4.641007939247497e-05,
"loss": 0.8915,
"step": 1060
},
{
"epoch": 0.992081974848626,
"grad_norm": 0.6056783632384961,
"learning_rate": 4.6366931308249916e-05,
"loss": 0.8898,
"step": 1065
},
{
"epoch": 0.9967396367023754,
"grad_norm": 0.5456430939985069,
"learning_rate": 4.632378322402486e-05,
"loss": 0.87,
"step": 1070
},
{
"epoch": 1.00093153237075,
"grad_norm": 0.9221112959963783,
"learning_rate": 4.6280635139799794e-05,
"loss": 0.9038,
"step": 1075
},
{
"epoch": 1.0055891942244992,
"grad_norm": 0.6363659450908413,
"learning_rate": 4.6237487055574736e-05,
"loss": 0.8434,
"step": 1080
},
{
"epoch": 1.0102468560782487,
"grad_norm": 0.6126175571817937,
"learning_rate": 4.619433897134968e-05,
"loss": 0.8165,
"step": 1085
},
{
"epoch": 1.0149045179319982,
"grad_norm": 0.7222816091274098,
"learning_rate": 4.6151190887124615e-05,
"loss": 0.8534,
"step": 1090
},
{
"epoch": 1.0195621797857475,
"grad_norm": 0.5536745759404794,
"learning_rate": 4.610804280289955e-05,
"loss": 0.8433,
"step": 1095
},
{
"epoch": 1.024219841639497,
"grad_norm": 0.6558844604399796,
"learning_rate": 4.606489471867449e-05,
"loss": 0.8482,
"step": 1100
},
{
"epoch": 1.0288775034932465,
"grad_norm": 0.7025611932544967,
"learning_rate": 4.602174663444943e-05,
"loss": 0.8425,
"step": 1105
},
{
"epoch": 1.0335351653469957,
"grad_norm": 0.5360081102807046,
"learning_rate": 4.597859855022437e-05,
"loss": 0.8374,
"step": 1110
},
{
"epoch": 1.0381928272007452,
"grad_norm": 0.5624535132798797,
"learning_rate": 4.5935450465999314e-05,
"loss": 0.8244,
"step": 1115
},
{
"epoch": 1.0428504890544947,
"grad_norm": 0.6292956541367164,
"learning_rate": 4.589230238177425e-05,
"loss": 0.8411,
"step": 1120
},
{
"epoch": 1.047508150908244,
"grad_norm": 0.5677787694167703,
"learning_rate": 4.584915429754919e-05,
"loss": 0.8507,
"step": 1125
},
{
"epoch": 1.0521658127619935,
"grad_norm": 0.5852407402420233,
"learning_rate": 4.5806006213324134e-05,
"loss": 0.8293,
"step": 1130
},
{
"epoch": 1.056823474615743,
"grad_norm": 0.5286627972800246,
"learning_rate": 4.576285812909907e-05,
"loss": 0.8402,
"step": 1135
},
{
"epoch": 1.0614811364694923,
"grad_norm": 0.565583073571462,
"learning_rate": 4.5719710044874006e-05,
"loss": 0.842,
"step": 1140
},
{
"epoch": 1.0661387983232418,
"grad_norm": 0.6235025356308691,
"learning_rate": 4.567656196064895e-05,
"loss": 0.8384,
"step": 1145
},
{
"epoch": 1.0707964601769913,
"grad_norm": 0.5676116085646312,
"learning_rate": 4.563341387642389e-05,
"loss": 0.838,
"step": 1150
},
{
"epoch": 1.0754541220307405,
"grad_norm": 0.7437050533342857,
"learning_rate": 4.5590265792198827e-05,
"loss": 0.8434,
"step": 1155
},
{
"epoch": 1.08011178388449,
"grad_norm": 0.7179223471862046,
"learning_rate": 4.554711770797377e-05,
"loss": 0.8587,
"step": 1160
},
{
"epoch": 1.0847694457382393,
"grad_norm": 0.5402283961364768,
"learning_rate": 4.550396962374871e-05,
"loss": 0.8469,
"step": 1165
},
{
"epoch": 1.0894271075919888,
"grad_norm": 0.602625129953774,
"learning_rate": 4.546082153952365e-05,
"loss": 0.8432,
"step": 1170
},
{
"epoch": 1.0940847694457383,
"grad_norm": 0.5631553474157106,
"learning_rate": 4.541767345529859e-05,
"loss": 0.8477,
"step": 1175
},
{
"epoch": 1.0987424312994876,
"grad_norm": 0.6162837263832036,
"learning_rate": 4.5374525371073526e-05,
"loss": 0.8314,
"step": 1180
},
{
"epoch": 1.103400093153237,
"grad_norm": 0.5696876350925066,
"learning_rate": 4.533137728684846e-05,
"loss": 0.8268,
"step": 1185
},
{
"epoch": 1.1080577550069866,
"grad_norm": 0.5741552121659874,
"learning_rate": 4.5288229202623404e-05,
"loss": 0.797,
"step": 1190
},
{
"epoch": 1.1127154168607358,
"grad_norm": 0.5868300402624936,
"learning_rate": 4.5245081118398346e-05,
"loss": 0.8177,
"step": 1195
},
{
"epoch": 1.1173730787144853,
"grad_norm": 0.575094872638148,
"learning_rate": 4.520193303417328e-05,
"loss": 0.8297,
"step": 1200
},
{
"epoch": 1.1220307405682348,
"grad_norm": 0.6424040008443889,
"learning_rate": 4.5158784949948224e-05,
"loss": 0.8511,
"step": 1205
},
{
"epoch": 1.126688402421984,
"grad_norm": 0.5915975477381874,
"learning_rate": 4.511563686572317e-05,
"loss": 0.8421,
"step": 1210
},
{
"epoch": 1.1313460642757336,
"grad_norm": 0.6551217680153546,
"learning_rate": 4.50724887814981e-05,
"loss": 0.8318,
"step": 1215
},
{
"epoch": 1.136003726129483,
"grad_norm": 0.606747965007546,
"learning_rate": 4.5029340697273045e-05,
"loss": 0.8586,
"step": 1220
},
{
"epoch": 1.1406613879832324,
"grad_norm": 0.5836907149894681,
"learning_rate": 4.498619261304798e-05,
"loss": 0.8292,
"step": 1225
},
{
"epoch": 1.1453190498369819,
"grad_norm": 0.5472296611267455,
"learning_rate": 4.4943044528822923e-05,
"loss": 0.8288,
"step": 1230
},
{
"epoch": 1.1499767116907313,
"grad_norm": 0.6576237995587761,
"learning_rate": 4.489989644459786e-05,
"loss": 0.8345,
"step": 1235
},
{
"epoch": 1.1546343735444806,
"grad_norm": 0.6057108499428023,
"learning_rate": 4.48567483603728e-05,
"loss": 0.8506,
"step": 1240
},
{
"epoch": 1.1592920353982301,
"grad_norm": 0.6564626329492939,
"learning_rate": 4.4813600276147744e-05,
"loss": 0.8415,
"step": 1245
},
{
"epoch": 1.1639496972519794,
"grad_norm": 0.6820004841637497,
"learning_rate": 4.477045219192268e-05,
"loss": 0.8611,
"step": 1250
},
{
"epoch": 1.1686073591057289,
"grad_norm": 0.6171342695838625,
"learning_rate": 4.472730410769762e-05,
"loss": 0.8402,
"step": 1255
},
{
"epoch": 1.1732650209594784,
"grad_norm": 0.5688972110628895,
"learning_rate": 4.4684156023472565e-05,
"loss": 0.8505,
"step": 1260
},
{
"epoch": 1.1779226828132279,
"grad_norm": 0.5891759900853567,
"learning_rate": 4.46410079392475e-05,
"loss": 0.8278,
"step": 1265
},
{
"epoch": 1.1825803446669771,
"grad_norm": 0.6977683673147888,
"learning_rate": 4.4597859855022436e-05,
"loss": 0.8168,
"step": 1270
},
{
"epoch": 1.1872380065207266,
"grad_norm": 0.6546497320070066,
"learning_rate": 4.455471177079738e-05,
"loss": 0.8269,
"step": 1275
},
{
"epoch": 1.191895668374476,
"grad_norm": 0.6048800604194604,
"learning_rate": 4.4511563686572315e-05,
"loss": 0.843,
"step": 1280
},
{
"epoch": 1.1965533302282254,
"grad_norm": 0.6978477920880507,
"learning_rate": 4.446841560234726e-05,
"loss": 0.8457,
"step": 1285
},
{
"epoch": 1.201210992081975,
"grad_norm": 0.815572653341471,
"learning_rate": 4.44252675181222e-05,
"loss": 0.8208,
"step": 1290
},
{
"epoch": 1.2058686539357242,
"grad_norm": 0.6579935666327247,
"learning_rate": 4.4382119433897135e-05,
"loss": 0.8394,
"step": 1295
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.5995063193473495,
"learning_rate": 4.433897134967208e-05,
"loss": 0.8275,
"step": 1300
},
{
"epoch": 1.2151839776432232,
"grad_norm": 0.6609879496434774,
"learning_rate": 4.429582326544702e-05,
"loss": 0.846,
"step": 1305
},
{
"epoch": 1.2198416394969724,
"grad_norm": 0.5789903731200582,
"learning_rate": 4.4252675181221956e-05,
"loss": 0.8503,
"step": 1310
},
{
"epoch": 1.224499301350722,
"grad_norm": 0.5432938262471018,
"learning_rate": 4.420952709699689e-05,
"loss": 0.8335,
"step": 1315
},
{
"epoch": 1.2291569632044714,
"grad_norm": 0.9738307220223945,
"learning_rate": 4.4166379012771834e-05,
"loss": 0.8536,
"step": 1320
},
{
"epoch": 1.2338146250582207,
"grad_norm": 0.6601401078017379,
"learning_rate": 4.412323092854678e-05,
"loss": 0.8287,
"step": 1325
},
{
"epoch": 1.2384722869119702,
"grad_norm": 0.8454838556235085,
"learning_rate": 4.408008284432171e-05,
"loss": 0.847,
"step": 1330
},
{
"epoch": 1.2431299487657197,
"grad_norm": 0.7189640931574706,
"learning_rate": 4.4036934760096655e-05,
"loss": 0.8341,
"step": 1335
},
{
"epoch": 1.247787610619469,
"grad_norm": 0.65499438366459,
"learning_rate": 4.39937866758716e-05,
"loss": 0.8296,
"step": 1340
},
{
"epoch": 1.2524452724732185,
"grad_norm": 0.5293078036494245,
"learning_rate": 4.395063859164653e-05,
"loss": 0.8123,
"step": 1345
},
{
"epoch": 1.257102934326968,
"grad_norm": 0.4988808471368861,
"learning_rate": 4.3907490507421476e-05,
"loss": 0.8299,
"step": 1350
},
{
"epoch": 1.2617605961807172,
"grad_norm": 1.0174215574294028,
"learning_rate": 4.386434242319641e-05,
"loss": 0.8237,
"step": 1355
},
{
"epoch": 1.2664182580344667,
"grad_norm": 0.5529007964457375,
"learning_rate": 4.382119433897135e-05,
"loss": 0.8091,
"step": 1360
},
{
"epoch": 1.271075919888216,
"grad_norm": 0.5429714398441454,
"learning_rate": 4.377804625474629e-05,
"loss": 0.8399,
"step": 1365
},
{
"epoch": 1.2757335817419655,
"grad_norm": 0.5963112183610055,
"learning_rate": 4.373489817052123e-05,
"loss": 0.8297,
"step": 1370
},
{
"epoch": 1.280391243595715,
"grad_norm": 0.6012562114727978,
"learning_rate": 4.369175008629617e-05,
"loss": 0.8292,
"step": 1375
},
{
"epoch": 1.2850489054494645,
"grad_norm": 0.5118548369236212,
"learning_rate": 4.364860200207111e-05,
"loss": 0.8291,
"step": 1380
},
{
"epoch": 1.2897065673032138,
"grad_norm": 0.558834446477445,
"learning_rate": 4.360545391784605e-05,
"loss": 0.829,
"step": 1385
},
{
"epoch": 1.2943642291569633,
"grad_norm": 0.7487565636309964,
"learning_rate": 4.356230583362099e-05,
"loss": 0.8453,
"step": 1390
},
{
"epoch": 1.2990218910107125,
"grad_norm": 0.5866609927345416,
"learning_rate": 4.351915774939593e-05,
"loss": 0.8177,
"step": 1395
},
{
"epoch": 1.303679552864462,
"grad_norm": 0.7832341341720332,
"learning_rate": 4.347600966517087e-05,
"loss": 0.8437,
"step": 1400
},
{
"epoch": 1.3083372147182115,
"grad_norm": 0.5539507604974554,
"learning_rate": 4.343286158094581e-05,
"loss": 0.8251,
"step": 1405
},
{
"epoch": 1.312994876571961,
"grad_norm": 0.48034968028596836,
"learning_rate": 4.3389713496720745e-05,
"loss": 0.8157,
"step": 1410
},
{
"epoch": 1.3176525384257103,
"grad_norm": 0.5146231089706201,
"learning_rate": 4.334656541249569e-05,
"loss": 0.851,
"step": 1415
},
{
"epoch": 1.3223102002794598,
"grad_norm": 0.5596868182553039,
"learning_rate": 4.330341732827063e-05,
"loss": 0.8336,
"step": 1420
},
{
"epoch": 1.326967862133209,
"grad_norm": 0.6983008027531773,
"learning_rate": 4.3260269244045566e-05,
"loss": 0.82,
"step": 1425
},
{
"epoch": 1.3316255239869585,
"grad_norm": 0.744496091683945,
"learning_rate": 4.321712115982051e-05,
"loss": 0.8193,
"step": 1430
},
{
"epoch": 1.336283185840708,
"grad_norm": 0.5150047650943387,
"learning_rate": 4.317397307559545e-05,
"loss": 0.8248,
"step": 1435
},
{
"epoch": 1.3409408476944573,
"grad_norm": 0.5484843193419154,
"learning_rate": 4.3130824991370387e-05,
"loss": 0.8301,
"step": 1440
},
{
"epoch": 1.3455985095482068,
"grad_norm": 0.5344563616800695,
"learning_rate": 4.308767690714532e-05,
"loss": 0.846,
"step": 1445
},
{
"epoch": 1.350256171401956,
"grad_norm": 0.5727837145848953,
"learning_rate": 4.3044528822920265e-05,
"loss": 0.8254,
"step": 1450
},
{
"epoch": 1.3549138332557056,
"grad_norm": 0.636733485984348,
"learning_rate": 4.30013807386952e-05,
"loss": 0.8556,
"step": 1455
},
{
"epoch": 1.359571495109455,
"grad_norm": 1.9749022217888352,
"learning_rate": 4.295823265447014e-05,
"loss": 0.8287,
"step": 1460
},
{
"epoch": 1.3642291569632046,
"grad_norm": 0.6137468932762344,
"learning_rate": 4.2915084570245085e-05,
"loss": 0.8247,
"step": 1465
},
{
"epoch": 1.3688868188169538,
"grad_norm": 0.5057907698773128,
"learning_rate": 4.287193648602002e-05,
"loss": 0.8304,
"step": 1470
},
{
"epoch": 1.3735444806707033,
"grad_norm": 0.7818389700842573,
"learning_rate": 4.2828788401794964e-05,
"loss": 0.8176,
"step": 1475
},
{
"epoch": 1.3782021425244526,
"grad_norm": 0.5323352228644667,
"learning_rate": 4.27856403175699e-05,
"loss": 0.8183,
"step": 1480
},
{
"epoch": 1.382859804378202,
"grad_norm": 0.637670077434287,
"learning_rate": 4.274249223334484e-05,
"loss": 0.8269,
"step": 1485
},
{
"epoch": 1.3875174662319516,
"grad_norm": 0.7043428559259748,
"learning_rate": 4.269934414911978e-05,
"loss": 0.837,
"step": 1490
},
{
"epoch": 1.392175128085701,
"grad_norm": 0.5064589289765138,
"learning_rate": 4.265619606489472e-05,
"loss": 0.8351,
"step": 1495
},
{
"epoch": 1.3968327899394504,
"grad_norm": 0.6623292044979259,
"learning_rate": 4.261304798066966e-05,
"loss": 0.9041,
"step": 1500
},
{
"epoch": 1.4014904517931999,
"grad_norm": 0.5186037454741974,
"learning_rate": 4.25698998964446e-05,
"loss": 0.8478,
"step": 1505
},
{
"epoch": 1.4061481136469491,
"grad_norm": 4.197835799315802,
"learning_rate": 4.252675181221954e-05,
"loss": 0.8281,
"step": 1510
},
{
"epoch": 1.4108057755006986,
"grad_norm": 0.7873855530523827,
"learning_rate": 4.248360372799448e-05,
"loss": 0.8473,
"step": 1515
},
{
"epoch": 1.4154634373544481,
"grad_norm": 0.6809163327185851,
"learning_rate": 4.244045564376942e-05,
"loss": 0.8224,
"step": 1520
},
{
"epoch": 1.4201210992081974,
"grad_norm": 0.6235694245025729,
"learning_rate": 4.2397307559544355e-05,
"loss": 0.8207,
"step": 1525
},
{
"epoch": 1.424778761061947,
"grad_norm": 0.6402875688149763,
"learning_rate": 4.23541594753193e-05,
"loss": 0.8077,
"step": 1530
},
{
"epoch": 1.4294364229156964,
"grad_norm": 0.4740272476367251,
"learning_rate": 4.231101139109423e-05,
"loss": 0.8356,
"step": 1535
},
{
"epoch": 1.4340940847694457,
"grad_norm": 0.6099287192468218,
"learning_rate": 4.2267863306869176e-05,
"loss": 0.8182,
"step": 1540
},
{
"epoch": 1.4387517466231952,
"grad_norm": 0.5545835645045659,
"learning_rate": 4.222471522264412e-05,
"loss": 0.8192,
"step": 1545
},
{
"epoch": 1.4434094084769447,
"grad_norm": 0.5409026812594355,
"learning_rate": 4.2181567138419054e-05,
"loss": 0.8085,
"step": 1550
},
{
"epoch": 1.448067070330694,
"grad_norm": 0.5436612116893749,
"learning_rate": 4.2138419054193996e-05,
"loss": 0.8024,
"step": 1555
},
{
"epoch": 1.4527247321844434,
"grad_norm": 0.5495238333319105,
"learning_rate": 4.209527096996894e-05,
"loss": 0.8157,
"step": 1560
},
{
"epoch": 1.4573823940381927,
"grad_norm": 0.5624828440338228,
"learning_rate": 4.2052122885743875e-05,
"loss": 0.8337,
"step": 1565
},
{
"epoch": 1.4620400558919422,
"grad_norm": 0.5658622352956784,
"learning_rate": 4.200897480151881e-05,
"loss": 0.8604,
"step": 1570
},
{
"epoch": 1.4666977177456917,
"grad_norm": 0.49705242407659034,
"learning_rate": 4.196582671729375e-05,
"loss": 0.8203,
"step": 1575
},
{
"epoch": 1.4713553795994412,
"grad_norm": 0.4961979991263778,
"learning_rate": 4.1922678633068695e-05,
"loss": 0.8002,
"step": 1580
},
{
"epoch": 1.4760130414531905,
"grad_norm": 0.4834846803784741,
"learning_rate": 4.187953054884363e-05,
"loss": 0.8162,
"step": 1585
},
{
"epoch": 1.48067070330694,
"grad_norm": 0.525816662994698,
"learning_rate": 4.1836382464618573e-05,
"loss": 0.8156,
"step": 1590
},
{
"epoch": 1.4853283651606892,
"grad_norm": 0.5214766244167187,
"learning_rate": 4.1793234380393516e-05,
"loss": 0.8091,
"step": 1595
},
{
"epoch": 1.4899860270144387,
"grad_norm": 0.4556881067346665,
"learning_rate": 4.175008629616845e-05,
"loss": 0.8226,
"step": 1600
},
{
"epoch": 1.4946436888681882,
"grad_norm": 0.4621437528857733,
"learning_rate": 4.1706938211943394e-05,
"loss": 0.8068,
"step": 1605
},
{
"epoch": 1.4993013507219377,
"grad_norm": 0.5285179894725354,
"learning_rate": 4.166379012771833e-05,
"loss": 0.8173,
"step": 1610
},
{
"epoch": 1.503959012575687,
"grad_norm": 0.5260089002965593,
"learning_rate": 4.1620642043493266e-05,
"loss": 0.8198,
"step": 1615
},
{
"epoch": 1.5086166744294365,
"grad_norm": 0.5295370263008858,
"learning_rate": 4.157749395926821e-05,
"loss": 0.8154,
"step": 1620
},
{
"epoch": 1.5132743362831858,
"grad_norm": 0.4869530399329456,
"learning_rate": 4.153434587504315e-05,
"loss": 0.8249,
"step": 1625
},
{
"epoch": 1.5179319981369352,
"grad_norm": 0.6067836521311675,
"learning_rate": 4.1491197790818086e-05,
"loss": 0.8197,
"step": 1630
},
{
"epoch": 1.5225896599906847,
"grad_norm": 0.5303238410825005,
"learning_rate": 4.144804970659303e-05,
"loss": 0.8411,
"step": 1635
},
{
"epoch": 1.5272473218444342,
"grad_norm": 0.5262674711845159,
"learning_rate": 4.140490162236797e-05,
"loss": 0.8248,
"step": 1640
},
{
"epoch": 1.5319049836981835,
"grad_norm": 0.4789682783106058,
"learning_rate": 4.1361753538142914e-05,
"loss": 0.8351,
"step": 1645
},
{
"epoch": 1.5365626455519328,
"grad_norm": 0.5368803341047521,
"learning_rate": 4.131860545391785e-05,
"loss": 0.8064,
"step": 1650
},
{
"epoch": 1.5412203074056823,
"grad_norm": 0.5179417938574834,
"learning_rate": 4.1275457369692785e-05,
"loss": 0.802,
"step": 1655
},
{
"epoch": 1.5458779692594318,
"grad_norm": 0.5668282815058873,
"learning_rate": 4.123230928546773e-05,
"loss": 0.7992,
"step": 1660
},
{
"epoch": 1.5505356311131813,
"grad_norm": 0.5678868939249553,
"learning_rate": 4.1189161201242664e-05,
"loss": 0.8255,
"step": 1665
},
{
"epoch": 1.5551932929669308,
"grad_norm": 2.284864352356659,
"learning_rate": 4.1146013117017606e-05,
"loss": 0.8147,
"step": 1670
},
{
"epoch": 1.55985095482068,
"grad_norm": 0.5386569746518188,
"learning_rate": 4.110286503279255e-05,
"loss": 0.8085,
"step": 1675
},
{
"epoch": 1.5645086166744293,
"grad_norm": 0.5364020323012693,
"learning_rate": 4.1059716948567484e-05,
"loss": 0.8253,
"step": 1680
},
{
"epoch": 1.5691662785281788,
"grad_norm": 0.6168960795213244,
"learning_rate": 4.101656886434243e-05,
"loss": 0.8035,
"step": 1685
},
{
"epoch": 1.5738239403819283,
"grad_norm": 0.4846106072381326,
"learning_rate": 4.097342078011737e-05,
"loss": 0.8179,
"step": 1690
},
{
"epoch": 1.5784816022356778,
"grad_norm": 0.5532441305001439,
"learning_rate": 4.0930272695892305e-05,
"loss": 0.8117,
"step": 1695
},
{
"epoch": 1.583139264089427,
"grad_norm": 0.4839401425043646,
"learning_rate": 4.088712461166724e-05,
"loss": 0.8208,
"step": 1700
},
{
"epoch": 1.5877969259431766,
"grad_norm": 0.4659417284258822,
"learning_rate": 4.084397652744218e-05,
"loss": 0.8101,
"step": 1705
},
{
"epoch": 1.5924545877969258,
"grad_norm": 0.5536360542493969,
"learning_rate": 4.080082844321712e-05,
"loss": 0.8183,
"step": 1710
},
{
"epoch": 1.5971122496506753,
"grad_norm": 0.613015733329207,
"learning_rate": 4.075768035899206e-05,
"loss": 0.8345,
"step": 1715
},
{
"epoch": 1.6017699115044248,
"grad_norm": 0.5037047505496149,
"learning_rate": 4.0714532274767004e-05,
"loss": 0.8043,
"step": 1720
},
{
"epoch": 1.6064275733581743,
"grad_norm": 0.5314368843138948,
"learning_rate": 4.0671384190541946e-05,
"loss": 0.8093,
"step": 1725
},
{
"epoch": 1.6110852352119236,
"grad_norm": 0.5472689278490457,
"learning_rate": 4.062823610631688e-05,
"loss": 0.8339,
"step": 1730
},
{
"epoch": 1.6157428970656729,
"grad_norm": 0.6018155674442411,
"learning_rate": 4.0585088022091825e-05,
"loss": 0.8264,
"step": 1735
},
{
"epoch": 1.6204005589194224,
"grad_norm": 0.49294301703177756,
"learning_rate": 4.054193993786676e-05,
"loss": 0.7908,
"step": 1740
},
{
"epoch": 1.6250582207731719,
"grad_norm": 0.4867952107412176,
"learning_rate": 4.0498791853641696e-05,
"loss": 0.8535,
"step": 1745
},
{
"epoch": 1.6297158826269214,
"grad_norm": 0.5458001464255956,
"learning_rate": 4.045564376941664e-05,
"loss": 0.8033,
"step": 1750
},
{
"epoch": 1.6343735444806708,
"grad_norm": 0.4903678189049089,
"learning_rate": 4.041249568519158e-05,
"loss": 0.8336,
"step": 1755
},
{
"epoch": 1.6390312063344201,
"grad_norm": 0.48052156111025546,
"learning_rate": 4.036934760096652e-05,
"loss": 0.7971,
"step": 1760
},
{
"epoch": 1.6436888681881694,
"grad_norm": 0.4775423173877248,
"learning_rate": 4.032619951674146e-05,
"loss": 0.8074,
"step": 1765
},
{
"epoch": 1.648346530041919,
"grad_norm": 0.5512406872892543,
"learning_rate": 4.02830514325164e-05,
"loss": 0.7981,
"step": 1770
},
{
"epoch": 1.6530041918956684,
"grad_norm": 0.47524257875404746,
"learning_rate": 4.023990334829134e-05,
"loss": 0.8297,
"step": 1775
},
{
"epoch": 1.6576618537494179,
"grad_norm": 0.4729453416458736,
"learning_rate": 4.019675526406628e-05,
"loss": 0.7991,
"step": 1780
},
{
"epoch": 1.6623195156031674,
"grad_norm": 0.5069795573310173,
"learning_rate": 4.0153607179841216e-05,
"loss": 0.8202,
"step": 1785
},
{
"epoch": 1.6669771774569166,
"grad_norm": 0.4950494671876146,
"learning_rate": 4.011045909561615e-05,
"loss": 0.8024,
"step": 1790
},
{
"epoch": 1.671634839310666,
"grad_norm": 0.581784405111925,
"learning_rate": 4.0067311011391094e-05,
"loss": 0.7972,
"step": 1795
},
{
"epoch": 1.6762925011644154,
"grad_norm": 0.4427443809477199,
"learning_rate": 4.0024162927166037e-05,
"loss": 0.798,
"step": 1800
},
{
"epoch": 1.680950163018165,
"grad_norm": 0.5630812899708987,
"learning_rate": 3.998101484294098e-05,
"loss": 0.8084,
"step": 1805
},
{
"epoch": 1.6856078248719144,
"grad_norm": 0.47170620633552024,
"learning_rate": 3.9937866758715915e-05,
"loss": 0.8061,
"step": 1810
},
{
"epoch": 1.6902654867256637,
"grad_norm": 0.484745803492533,
"learning_rate": 3.989471867449086e-05,
"loss": 0.8085,
"step": 1815
},
{
"epoch": 1.6949231485794132,
"grad_norm": 0.5256720082001222,
"learning_rate": 3.98515705902658e-05,
"loss": 0.7787,
"step": 1820
},
{
"epoch": 1.6995808104331624,
"grad_norm": 0.5203870132810984,
"learning_rate": 3.9808422506040736e-05,
"loss": 0.7886,
"step": 1825
},
{
"epoch": 1.704238472286912,
"grad_norm": 0.5045427269257229,
"learning_rate": 3.976527442181567e-05,
"loss": 0.7939,
"step": 1830
},
{
"epoch": 1.7088961341406614,
"grad_norm": 0.5169331146428379,
"learning_rate": 3.9722126337590614e-05,
"loss": 0.81,
"step": 1835
},
{
"epoch": 1.713553795994411,
"grad_norm": 0.5431794591639149,
"learning_rate": 3.967897825336555e-05,
"loss": 0.8029,
"step": 1840
},
{
"epoch": 1.7182114578481602,
"grad_norm": 0.5477208936261382,
"learning_rate": 3.963583016914049e-05,
"loss": 0.8069,
"step": 1845
},
{
"epoch": 1.7228691197019095,
"grad_norm": 0.6041641324030878,
"learning_rate": 3.9592682084915434e-05,
"loss": 0.8258,
"step": 1850
},
{
"epoch": 1.727526781555659,
"grad_norm": 0.5090806504384892,
"learning_rate": 3.954953400069037e-05,
"loss": 0.7881,
"step": 1855
},
{
"epoch": 1.7321844434094085,
"grad_norm": 0.49889036648991425,
"learning_rate": 3.950638591646531e-05,
"loss": 0.8397,
"step": 1860
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.4935177448464897,
"learning_rate": 3.9463237832240255e-05,
"loss": 0.7969,
"step": 1865
},
{
"epoch": 1.7414997671169075,
"grad_norm": 0.47140321745349795,
"learning_rate": 3.942008974801519e-05,
"loss": 0.8106,
"step": 1870
},
{
"epoch": 1.7461574289706567,
"grad_norm": 0.469571803119941,
"learning_rate": 3.937694166379013e-05,
"loss": 0.8384,
"step": 1875
},
{
"epoch": 1.750815090824406,
"grad_norm": 0.5050432985953494,
"learning_rate": 3.933379357956507e-05,
"loss": 0.8378,
"step": 1880
},
{
"epoch": 1.7554727526781555,
"grad_norm": 0.563820523969095,
"learning_rate": 3.9290645495340005e-05,
"loss": 0.8234,
"step": 1885
},
{
"epoch": 1.760130414531905,
"grad_norm": 0.5849722918534452,
"learning_rate": 3.924749741111495e-05,
"loss": 0.8135,
"step": 1890
},
{
"epoch": 1.7647880763856545,
"grad_norm": 0.5619437429386521,
"learning_rate": 3.920434932688989e-05,
"loss": 0.8321,
"step": 1895
},
{
"epoch": 1.7694457382394038,
"grad_norm": 0.5541650724131253,
"learning_rate": 3.916120124266483e-05,
"loss": 0.8122,
"step": 1900
},
{
"epoch": 1.7741034000931533,
"grad_norm": 0.4790373788753262,
"learning_rate": 3.911805315843977e-05,
"loss": 0.8107,
"step": 1905
},
{
"epoch": 1.7787610619469025,
"grad_norm": 0.48571446107425204,
"learning_rate": 3.9074905074214704e-05,
"loss": 0.7954,
"step": 1910
},
{
"epoch": 1.783418723800652,
"grad_norm": 0.4554059775758542,
"learning_rate": 3.9031756989989646e-05,
"loss": 0.7812,
"step": 1915
},
{
"epoch": 1.7880763856544015,
"grad_norm": 0.5130980660028385,
"learning_rate": 3.898860890576458e-05,
"loss": 0.7971,
"step": 1920
},
{
"epoch": 1.792734047508151,
"grad_norm": 0.5185882487044102,
"learning_rate": 3.8945460821539525e-05,
"loss": 0.8086,
"step": 1925
},
{
"epoch": 1.7973917093619003,
"grad_norm": 0.5103912287550575,
"learning_rate": 3.890231273731447e-05,
"loss": 0.7979,
"step": 1930
},
{
"epoch": 1.8020493712156498,
"grad_norm": 0.462819612155367,
"learning_rate": 3.88591646530894e-05,
"loss": 0.7848,
"step": 1935
},
{
"epoch": 1.806707033069399,
"grad_norm": 0.4664370966792569,
"learning_rate": 3.8816016568864345e-05,
"loss": 0.7909,
"step": 1940
},
{
"epoch": 1.8113646949231486,
"grad_norm": 0.4543630912158898,
"learning_rate": 3.877286848463929e-05,
"loss": 0.8142,
"step": 1945
},
{
"epoch": 1.816022356776898,
"grad_norm": 0.46997582522664094,
"learning_rate": 3.8729720400414224e-05,
"loss": 0.7964,
"step": 1950
},
{
"epoch": 1.8206800186306475,
"grad_norm": 0.4608432519022556,
"learning_rate": 3.868657231618916e-05,
"loss": 0.7993,
"step": 1955
},
{
"epoch": 1.8253376804843968,
"grad_norm": 0.48943547745302063,
"learning_rate": 3.86434242319641e-05,
"loss": 0.8045,
"step": 1960
},
{
"epoch": 1.829995342338146,
"grad_norm": 0.5034658417649208,
"learning_rate": 3.860027614773904e-05,
"loss": 0.8074,
"step": 1965
},
{
"epoch": 1.8346530041918956,
"grad_norm": 0.5079636831645327,
"learning_rate": 3.855712806351398e-05,
"loss": 0.8014,
"step": 1970
},
{
"epoch": 1.839310666045645,
"grad_norm": 0.5350620373119227,
"learning_rate": 3.851397997928892e-05,
"loss": 0.8108,
"step": 1975
},
{
"epoch": 1.8439683278993946,
"grad_norm": 0.4876171067090009,
"learning_rate": 3.8470831895063865e-05,
"loss": 0.7971,
"step": 1980
},
{
"epoch": 1.848625989753144,
"grad_norm": 0.5792114882681738,
"learning_rate": 3.84276838108388e-05,
"loss": 0.8266,
"step": 1985
},
{
"epoch": 1.8532836516068933,
"grad_norm": 0.5633248691401386,
"learning_rate": 3.838453572661374e-05,
"loss": 0.8094,
"step": 1990
},
{
"epoch": 1.8579413134606426,
"grad_norm": 0.4635065065769153,
"learning_rate": 3.834138764238868e-05,
"loss": 0.7795,
"step": 1995
},
{
"epoch": 1.8625989753143921,
"grad_norm": 0.5211367454249431,
"learning_rate": 3.8298239558163615e-05,
"loss": 0.803,
"step": 2000
},
{
"epoch": 1.8672566371681416,
"grad_norm": 0.5244552884735643,
"learning_rate": 3.825509147393856e-05,
"loss": 0.8218,
"step": 2005
},
{
"epoch": 1.871914299021891,
"grad_norm": 0.4367703254486583,
"learning_rate": 3.82119433897135e-05,
"loss": 0.7895,
"step": 2010
},
{
"epoch": 1.8765719608756404,
"grad_norm": 0.4365293713310002,
"learning_rate": 3.8168795305488435e-05,
"loss": 0.8254,
"step": 2015
},
{
"epoch": 1.8812296227293899,
"grad_norm": 0.4401620684419659,
"learning_rate": 3.812564722126338e-05,
"loss": 0.7918,
"step": 2020
},
{
"epoch": 1.8858872845831391,
"grad_norm": 0.47779920006319,
"learning_rate": 3.808249913703832e-05,
"loss": 0.8029,
"step": 2025
},
{
"epoch": 1.8905449464368886,
"grad_norm": 0.5555355624639677,
"learning_rate": 3.8039351052813256e-05,
"loss": 0.796,
"step": 2030
},
{
"epoch": 1.8952026082906381,
"grad_norm": 0.49458336406404724,
"learning_rate": 3.79962029685882e-05,
"loss": 0.789,
"step": 2035
},
{
"epoch": 1.8998602701443876,
"grad_norm": 0.5146735090214605,
"learning_rate": 3.7953054884363134e-05,
"loss": 0.8121,
"step": 2040
},
{
"epoch": 1.904517931998137,
"grad_norm": 0.4494321925050035,
"learning_rate": 3.790990680013807e-05,
"loss": 0.8214,
"step": 2045
},
{
"epoch": 1.9091755938518864,
"grad_norm": 0.5050066179121853,
"learning_rate": 3.786675871591301e-05,
"loss": 0.7987,
"step": 2050
},
{
"epoch": 1.9138332557056357,
"grad_norm": 0.47989526540598243,
"learning_rate": 3.7823610631687955e-05,
"loss": 0.8003,
"step": 2055
},
{
"epoch": 1.9184909175593852,
"grad_norm": 0.48483535452469545,
"learning_rate": 3.77804625474629e-05,
"loss": 0.7898,
"step": 2060
},
{
"epoch": 1.9231485794131347,
"grad_norm": 0.4687766430461443,
"learning_rate": 3.773731446323783e-05,
"loss": 0.7986,
"step": 2065
},
{
"epoch": 1.9278062412668842,
"grad_norm": 0.4752511382396736,
"learning_rate": 3.7694166379012776e-05,
"loss": 0.8032,
"step": 2070
},
{
"epoch": 1.9324639031206334,
"grad_norm": 0.5009445413481693,
"learning_rate": 3.765101829478772e-05,
"loss": 0.7997,
"step": 2075
},
{
"epoch": 1.9371215649743827,
"grad_norm": 0.4787984334128784,
"learning_rate": 3.7607870210562654e-05,
"loss": 0.7837,
"step": 2080
},
{
"epoch": 1.9417792268281322,
"grad_norm": 0.49842411743772097,
"learning_rate": 3.756472212633759e-05,
"loss": 0.8119,
"step": 2085
},
{
"epoch": 1.9464368886818817,
"grad_norm": 0.5129102375544244,
"learning_rate": 3.752157404211253e-05,
"loss": 0.7717,
"step": 2090
},
{
"epoch": 1.9510945505356312,
"grad_norm": 0.4983269320057428,
"learning_rate": 3.747842595788747e-05,
"loss": 0.7932,
"step": 2095
},
{
"epoch": 1.9557522123893807,
"grad_norm": 0.5694242293113918,
"learning_rate": 3.743527787366241e-05,
"loss": 0.7911,
"step": 2100
},
{
"epoch": 1.96040987424313,
"grad_norm": 0.47125598634785504,
"learning_rate": 3.739212978943735e-05,
"loss": 0.8082,
"step": 2105
},
{
"epoch": 1.9650675360968792,
"grad_norm": 0.4686243867209378,
"learning_rate": 3.734898170521229e-05,
"loss": 0.7781,
"step": 2110
},
{
"epoch": 1.9697251979506287,
"grad_norm": 0.4900196821966246,
"learning_rate": 3.730583362098723e-05,
"loss": 0.7998,
"step": 2115
},
{
"epoch": 1.9743828598043782,
"grad_norm": 0.5097124271224043,
"learning_rate": 3.7262685536762174e-05,
"loss": 0.8051,
"step": 2120
},
{
"epoch": 1.9790405216581277,
"grad_norm": 0.4957311238066013,
"learning_rate": 3.721953745253711e-05,
"loss": 0.7859,
"step": 2125
},
{
"epoch": 1.983698183511877,
"grad_norm": 0.4697294046892151,
"learning_rate": 3.7176389368312045e-05,
"loss": 0.7934,
"step": 2130
},
{
"epoch": 1.9883558453656265,
"grad_norm": 0.49474241334656344,
"learning_rate": 3.713324128408699e-05,
"loss": 0.8048,
"step": 2135
},
{
"epoch": 1.9930135072193758,
"grad_norm": 0.46310710184268955,
"learning_rate": 3.709009319986193e-05,
"loss": 0.8022,
"step": 2140
},
{
"epoch": 1.9976711690731253,
"grad_norm": 0.4756968698386488,
"learning_rate": 3.7046945115636866e-05,
"loss": 0.8023,
"step": 2145
},
{
"epoch": 2.0018630647415,
"grad_norm": 0.527343779742018,
"learning_rate": 3.700379703141181e-05,
"loss": 0.7902,
"step": 2150
},
{
"epoch": 2.0065207265952494,
"grad_norm": 0.5306054657790139,
"learning_rate": 3.696064894718675e-05,
"loss": 0.7417,
"step": 2155
},
{
"epoch": 2.0111783884489984,
"grad_norm": 0.4980162319421436,
"learning_rate": 3.6917500862961687e-05,
"loss": 0.7281,
"step": 2160
},
{
"epoch": 2.015836050302748,
"grad_norm": 0.49663031109205596,
"learning_rate": 3.687435277873663e-05,
"loss": 0.7391,
"step": 2165
},
{
"epoch": 2.0204937121564974,
"grad_norm": 0.5451837034273049,
"learning_rate": 3.6831204694511565e-05,
"loss": 0.7146,
"step": 2170
},
{
"epoch": 2.025151374010247,
"grad_norm": 0.509552114169026,
"learning_rate": 3.67880566102865e-05,
"loss": 0.7549,
"step": 2175
},
{
"epoch": 2.0298090358639964,
"grad_norm": 0.47803133974312734,
"learning_rate": 3.674490852606144e-05,
"loss": 0.7273,
"step": 2180
},
{
"epoch": 2.034466697717746,
"grad_norm": 0.5043879767130786,
"learning_rate": 3.6701760441836386e-05,
"loss": 0.7476,
"step": 2185
},
{
"epoch": 2.039124359571495,
"grad_norm": 0.5383186227544483,
"learning_rate": 3.665861235761132e-05,
"loss": 0.739,
"step": 2190
},
{
"epoch": 2.0437820214252445,
"grad_norm": 0.49461761719203784,
"learning_rate": 3.6615464273386264e-05,
"loss": 0.7376,
"step": 2195
},
{
"epoch": 2.048439683278994,
"grad_norm": 0.5533513553704322,
"learning_rate": 3.6572316189161206e-05,
"loss": 0.7344,
"step": 2200
},
{
"epoch": 2.0530973451327434,
"grad_norm": 0.48201139067056653,
"learning_rate": 3.652916810493614e-05,
"loss": 0.7452,
"step": 2205
},
{
"epoch": 2.057755006986493,
"grad_norm": 0.4956778283399664,
"learning_rate": 3.6486020020711085e-05,
"loss": 0.7213,
"step": 2210
},
{
"epoch": 2.062412668840242,
"grad_norm": 0.5511345017727285,
"learning_rate": 3.644287193648602e-05,
"loss": 0.7346,
"step": 2215
},
{
"epoch": 2.0670703306939915,
"grad_norm": 0.4657556929449439,
"learning_rate": 3.639972385226096e-05,
"loss": 0.7404,
"step": 2220
},
{
"epoch": 2.071727992547741,
"grad_norm": 0.5089880530513546,
"learning_rate": 3.63565757680359e-05,
"loss": 0.7359,
"step": 2225
},
{
"epoch": 2.0763856544014905,
"grad_norm": 0.5058181267132648,
"learning_rate": 3.631342768381084e-05,
"loss": 0.7377,
"step": 2230
},
{
"epoch": 2.08104331625524,
"grad_norm": 0.507478279733912,
"learning_rate": 3.6270279599585783e-05,
"loss": 0.7287,
"step": 2235
},
{
"epoch": 2.0857009781089895,
"grad_norm": 0.49894629085867603,
"learning_rate": 3.622713151536072e-05,
"loss": 0.7436,
"step": 2240
},
{
"epoch": 2.0903586399627385,
"grad_norm": 0.47534602756485905,
"learning_rate": 3.618398343113566e-05,
"loss": 0.7434,
"step": 2245
},
{
"epoch": 2.095016301816488,
"grad_norm": 0.48030281583814866,
"learning_rate": 3.6140835346910604e-05,
"loss": 0.7289,
"step": 2250
},
{
"epoch": 2.0996739636702375,
"grad_norm": 0.5258046969040029,
"learning_rate": 3.609768726268554e-05,
"loss": 0.7533,
"step": 2255
},
{
"epoch": 2.104331625523987,
"grad_norm": 0.5136907462435186,
"learning_rate": 3.6054539178460476e-05,
"loss": 0.7276,
"step": 2260
},
{
"epoch": 2.1089892873777365,
"grad_norm": 0.4810156623610355,
"learning_rate": 3.601139109423542e-05,
"loss": 0.756,
"step": 2265
},
{
"epoch": 2.113646949231486,
"grad_norm": 0.47960283436866896,
"learning_rate": 3.5968243010010354e-05,
"loss": 0.7346,
"step": 2270
},
{
"epoch": 2.118304611085235,
"grad_norm": 0.45763532355865844,
"learning_rate": 3.5925094925785296e-05,
"loss": 0.7377,
"step": 2275
},
{
"epoch": 2.1229622729389845,
"grad_norm": 0.504308557124316,
"learning_rate": 3.588194684156024e-05,
"loss": 0.7372,
"step": 2280
},
{
"epoch": 2.127619934792734,
"grad_norm": 0.5033911076958377,
"learning_rate": 3.5838798757335175e-05,
"loss": 0.7439,
"step": 2285
},
{
"epoch": 2.1322775966464835,
"grad_norm": 0.5592257193363861,
"learning_rate": 3.579565067311012e-05,
"loss": 0.7429,
"step": 2290
},
{
"epoch": 2.136935258500233,
"grad_norm": 0.5060165047760968,
"learning_rate": 3.575250258888505e-05,
"loss": 0.7538,
"step": 2295
},
{
"epoch": 2.1415929203539825,
"grad_norm": 0.4365692283959372,
"learning_rate": 3.5709354504659995e-05,
"loss": 0.7556,
"step": 2300
},
{
"epoch": 2.1462505822077316,
"grad_norm": 0.487649630360977,
"learning_rate": 3.566620642043493e-05,
"loss": 0.7397,
"step": 2305
},
{
"epoch": 2.150908244061481,
"grad_norm": 0.4475405819347315,
"learning_rate": 3.5623058336209874e-05,
"loss": 0.7112,
"step": 2310
},
{
"epoch": 2.1555659059152306,
"grad_norm": 0.4722891130226953,
"learning_rate": 3.5579910251984816e-05,
"loss": 0.7258,
"step": 2315
},
{
"epoch": 2.16022356776898,
"grad_norm": 0.4888175714369976,
"learning_rate": 3.553676216775975e-05,
"loss": 0.7291,
"step": 2320
},
{
"epoch": 2.1648812296227296,
"grad_norm": 0.5149403178372757,
"learning_rate": 3.5493614083534694e-05,
"loss": 0.7481,
"step": 2325
},
{
"epoch": 2.1695388914764786,
"grad_norm": 0.4656812561283848,
"learning_rate": 3.545046599930964e-05,
"loss": 0.7508,
"step": 2330
},
{
"epoch": 2.174196553330228,
"grad_norm": 0.5137790034694885,
"learning_rate": 3.540731791508457e-05,
"loss": 0.7233,
"step": 2335
},
{
"epoch": 2.1788542151839776,
"grad_norm": 0.4553249699224624,
"learning_rate": 3.536416983085951e-05,
"loss": 0.7452,
"step": 2340
},
{
"epoch": 2.183511877037727,
"grad_norm": 0.5039975669698809,
"learning_rate": 3.532102174663445e-05,
"loss": 0.7418,
"step": 2345
},
{
"epoch": 2.1881695388914766,
"grad_norm": 0.5019742558189558,
"learning_rate": 3.5277873662409386e-05,
"loss": 0.7434,
"step": 2350
},
{
"epoch": 2.192827200745226,
"grad_norm": 0.4976202334401095,
"learning_rate": 3.523472557818433e-05,
"loss": 0.7286,
"step": 2355
},
{
"epoch": 2.197484862598975,
"grad_norm": 0.46894087915981364,
"learning_rate": 3.519157749395927e-05,
"loss": 0.7463,
"step": 2360
},
{
"epoch": 2.2021425244527246,
"grad_norm": 0.43387707760639765,
"learning_rate": 3.514842940973421e-05,
"loss": 0.7411,
"step": 2365
},
{
"epoch": 2.206800186306474,
"grad_norm": 0.49150140111118595,
"learning_rate": 3.510528132550915e-05,
"loss": 0.7496,
"step": 2370
},
{
"epoch": 2.2114578481602236,
"grad_norm": 0.5585531969734789,
"learning_rate": 3.506213324128409e-05,
"loss": 0.7495,
"step": 2375
},
{
"epoch": 2.216115510013973,
"grad_norm": 0.5153195977847673,
"learning_rate": 3.501898515705903e-05,
"loss": 0.7333,
"step": 2380
},
{
"epoch": 2.2207731718677226,
"grad_norm": 0.4492901284094745,
"learning_rate": 3.4975837072833964e-05,
"loss": 0.7596,
"step": 2385
},
{
"epoch": 2.2254308337214717,
"grad_norm": 0.44696866001250674,
"learning_rate": 3.4932688988608906e-05,
"loss": 0.74,
"step": 2390
},
{
"epoch": 2.230088495575221,
"grad_norm": 0.5220507714567176,
"learning_rate": 3.488954090438385e-05,
"loss": 0.7391,
"step": 2395
},
{
"epoch": 2.2347461574289706,
"grad_norm": 0.47887225464445543,
"learning_rate": 3.4846392820158784e-05,
"loss": 0.7243,
"step": 2400
},
{
"epoch": 2.23940381928272,
"grad_norm": 0.4811847900886404,
"learning_rate": 3.480324473593373e-05,
"loss": 0.7418,
"step": 2405
},
{
"epoch": 2.2440614811364696,
"grad_norm": 0.5272801320352328,
"learning_rate": 3.476009665170867e-05,
"loss": 0.74,
"step": 2410
},
{
"epoch": 2.248719142990219,
"grad_norm": 0.5307464586993595,
"learning_rate": 3.4716948567483605e-05,
"loss": 0.7324,
"step": 2415
},
{
"epoch": 2.253376804843968,
"grad_norm": 0.49901292595440055,
"learning_rate": 3.467380048325855e-05,
"loss": 0.7113,
"step": 2420
},
{
"epoch": 2.2580344666977177,
"grad_norm": 0.5070314821761425,
"learning_rate": 3.463065239903348e-05,
"loss": 0.7562,
"step": 2425
},
{
"epoch": 2.262692128551467,
"grad_norm": 0.44427974244795587,
"learning_rate": 3.458750431480842e-05,
"loss": 0.7242,
"step": 2430
},
{
"epoch": 2.2673497904052167,
"grad_norm": 0.4771303989768548,
"learning_rate": 3.454435623058336e-05,
"loss": 0.7391,
"step": 2435
},
{
"epoch": 2.272007452258966,
"grad_norm": 0.49142003038405735,
"learning_rate": 3.4501208146358304e-05,
"loss": 0.7359,
"step": 2440
},
{
"epoch": 2.276665114112715,
"grad_norm": 0.5101425248743833,
"learning_rate": 3.445806006213324e-05,
"loss": 0.7337,
"step": 2445
},
{
"epoch": 2.2813227759664647,
"grad_norm": 0.4469242564135119,
"learning_rate": 3.441491197790818e-05,
"loss": 0.7328,
"step": 2450
},
{
"epoch": 2.285980437820214,
"grad_norm": 0.5151260716573384,
"learning_rate": 3.4371763893683125e-05,
"loss": 0.7285,
"step": 2455
},
{
"epoch": 2.2906380996739637,
"grad_norm": 0.6197779956105796,
"learning_rate": 3.432861580945806e-05,
"loss": 0.7195,
"step": 2460
},
{
"epoch": 2.295295761527713,
"grad_norm": 0.5338226806005612,
"learning_rate": 3.4285467725233e-05,
"loss": 0.7581,
"step": 2465
},
{
"epoch": 2.2999534233814627,
"grad_norm": 0.47245608705220976,
"learning_rate": 3.424231964100794e-05,
"loss": 0.7431,
"step": 2470
},
{
"epoch": 2.3046110852352117,
"grad_norm": 0.5201411523481256,
"learning_rate": 3.419917155678288e-05,
"loss": 0.7482,
"step": 2475
},
{
"epoch": 2.3092687470889612,
"grad_norm": 0.4908970760918655,
"learning_rate": 3.415602347255782e-05,
"loss": 0.7355,
"step": 2480
},
{
"epoch": 2.3139264089427107,
"grad_norm": 0.4931573032699934,
"learning_rate": 3.411287538833276e-05,
"loss": 0.7569,
"step": 2485
},
{
"epoch": 2.3185840707964602,
"grad_norm": 0.47470883095660726,
"learning_rate": 3.40697273041077e-05,
"loss": 0.756,
"step": 2490
},
{
"epoch": 2.3232417326502097,
"grad_norm": 0.4878991939271055,
"learning_rate": 3.402657921988264e-05,
"loss": 0.7369,
"step": 2495
},
{
"epoch": 2.3278993945039588,
"grad_norm": 0.4594200319210018,
"learning_rate": 3.398343113565758e-05,
"loss": 0.7142,
"step": 2500
},
{
"epoch": 2.3325570563577083,
"grad_norm": 0.46387141485413524,
"learning_rate": 3.394028305143252e-05,
"loss": 0.7345,
"step": 2505
},
{
"epoch": 2.3372147182114578,
"grad_norm": 0.531841628390487,
"learning_rate": 3.389713496720746e-05,
"loss": 0.7383,
"step": 2510
},
{
"epoch": 2.3418723800652073,
"grad_norm": 0.46190990050373143,
"learning_rate": 3.3853986882982394e-05,
"loss": 0.7381,
"step": 2515
},
{
"epoch": 2.3465300419189568,
"grad_norm": 0.4862616056924088,
"learning_rate": 3.381083879875734e-05,
"loss": 0.7447,
"step": 2520
},
{
"epoch": 2.3511877037727063,
"grad_norm": 0.46877636975068077,
"learning_rate": 3.376769071453227e-05,
"loss": 0.7409,
"step": 2525
},
{
"epoch": 2.3558453656264557,
"grad_norm": 0.47736005453793845,
"learning_rate": 3.3724542630307215e-05,
"loss": 0.74,
"step": 2530
},
{
"epoch": 2.360503027480205,
"grad_norm": 0.45506496994603135,
"learning_rate": 3.368139454608216e-05,
"loss": 0.7189,
"step": 2535
},
{
"epoch": 2.3651606893339543,
"grad_norm": 0.43838587525237144,
"learning_rate": 3.363824646185709e-05,
"loss": 0.7168,
"step": 2540
},
{
"epoch": 2.369818351187704,
"grad_norm": 0.44812953360079566,
"learning_rate": 3.3595098377632036e-05,
"loss": 0.7436,
"step": 2545
},
{
"epoch": 2.3744760130414533,
"grad_norm": 0.45612583541921564,
"learning_rate": 3.355195029340698e-05,
"loss": 0.7422,
"step": 2550
},
{
"epoch": 2.3791336748952028,
"grad_norm": 0.47178312549781826,
"learning_rate": 3.3508802209181914e-05,
"loss": 0.7269,
"step": 2555
},
{
"epoch": 2.383791336748952,
"grad_norm": 0.48551060780875127,
"learning_rate": 3.346565412495685e-05,
"loss": 0.7337,
"step": 2560
},
{
"epoch": 2.3884489986027013,
"grad_norm": 0.44976638729567375,
"learning_rate": 3.342250604073179e-05,
"loss": 0.7372,
"step": 2565
},
{
"epoch": 2.393106660456451,
"grad_norm": 0.43474437953643713,
"learning_rate": 3.3379357956506735e-05,
"loss": 0.7377,
"step": 2570
},
{
"epoch": 2.3977643223102003,
"grad_norm": 0.49975414042600347,
"learning_rate": 3.333620987228167e-05,
"loss": 0.7386,
"step": 2575
},
{
"epoch": 2.40242198416395,
"grad_norm": 0.5001674134152022,
"learning_rate": 3.329306178805661e-05,
"loss": 0.7241,
"step": 2580
},
{
"epoch": 2.4070796460176993,
"grad_norm": 0.5029118589595981,
"learning_rate": 3.3249913703831555e-05,
"loss": 0.7256,
"step": 2585
},
{
"epoch": 2.4117373078714484,
"grad_norm": 0.46960690178551306,
"learning_rate": 3.320676561960649e-05,
"loss": 0.7154,
"step": 2590
},
{
"epoch": 2.416394969725198,
"grad_norm": 0.4531885637486837,
"learning_rate": 3.3163617535381434e-05,
"loss": 0.7499,
"step": 2595
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.47476110228618335,
"learning_rate": 3.312046945115637e-05,
"loss": 0.7428,
"step": 2600
},
{
"epoch": 2.425710293432697,
"grad_norm": 0.4523064808621614,
"learning_rate": 3.3077321366931305e-05,
"loss": 0.7277,
"step": 2605
},
{
"epoch": 2.4303679552864463,
"grad_norm": 0.45178488039892994,
"learning_rate": 3.303417328270625e-05,
"loss": 0.7188,
"step": 2610
},
{
"epoch": 2.4350256171401954,
"grad_norm": 0.49014947823403127,
"learning_rate": 3.299102519848119e-05,
"loss": 0.7252,
"step": 2615
},
{
"epoch": 2.439683278993945,
"grad_norm": 0.4948237045149303,
"learning_rate": 3.2947877114256126e-05,
"loss": 0.7345,
"step": 2620
},
{
"epoch": 2.4443409408476944,
"grad_norm": 0.5370877028222926,
"learning_rate": 3.290472903003107e-05,
"loss": 0.7313,
"step": 2625
},
{
"epoch": 2.448998602701444,
"grad_norm": 0.46235224775820505,
"learning_rate": 3.286158094580601e-05,
"loss": 0.7203,
"step": 2630
},
{
"epoch": 2.4536562645551934,
"grad_norm": 0.44349107580473884,
"learning_rate": 3.281843286158095e-05,
"loss": 0.7427,
"step": 2635
},
{
"epoch": 2.458313926408943,
"grad_norm": 0.5208697307440874,
"learning_rate": 3.277528477735589e-05,
"loss": 0.7405,
"step": 2640
},
{
"epoch": 2.4629715882626924,
"grad_norm": 0.5024334402123977,
"learning_rate": 3.2732136693130825e-05,
"loss": 0.7398,
"step": 2645
},
{
"epoch": 2.4676292501164414,
"grad_norm": 0.4790235899407453,
"learning_rate": 3.268898860890577e-05,
"loss": 0.7254,
"step": 2650
},
{
"epoch": 2.472286911970191,
"grad_norm": 0.5322015962737839,
"learning_rate": 3.26458405246807e-05,
"loss": 0.738,
"step": 2655
},
{
"epoch": 2.4769445738239404,
"grad_norm": 0.4569166889476487,
"learning_rate": 3.2602692440455645e-05,
"loss": 0.7265,
"step": 2660
},
{
"epoch": 2.48160223567769,
"grad_norm": 0.48054196038218905,
"learning_rate": 3.255954435623059e-05,
"loss": 0.7266,
"step": 2665
},
{
"epoch": 2.4862598975314394,
"grad_norm": 0.5641082189572171,
"learning_rate": 3.2516396272005524e-05,
"loss": 0.7388,
"step": 2670
},
{
"epoch": 2.4909175593851884,
"grad_norm": 0.5662978894220593,
"learning_rate": 3.2473248187780466e-05,
"loss": 0.7235,
"step": 2675
},
{
"epoch": 2.495575221238938,
"grad_norm": 0.5391118159361664,
"learning_rate": 3.243010010355541e-05,
"loss": 0.7273,
"step": 2680
},
{
"epoch": 2.5002328830926874,
"grad_norm": 0.5362127590103279,
"learning_rate": 3.2386952019330344e-05,
"loss": 0.7336,
"step": 2685
},
{
"epoch": 2.504890544946437,
"grad_norm": 0.4748435027401365,
"learning_rate": 3.234380393510528e-05,
"loss": 0.7096,
"step": 2690
},
{
"epoch": 2.5095482068001864,
"grad_norm": 0.5033991418235945,
"learning_rate": 3.230065585088022e-05,
"loss": 0.7303,
"step": 2695
},
{
"epoch": 2.514205868653936,
"grad_norm": 0.43440064872955475,
"learning_rate": 3.225750776665516e-05,
"loss": 0.7216,
"step": 2700
},
{
"epoch": 2.5188635305076854,
"grad_norm": 0.5577535793990844,
"learning_rate": 3.22143596824301e-05,
"loss": 0.7426,
"step": 2705
},
{
"epoch": 2.5235211923614345,
"grad_norm": 0.48251183034179124,
"learning_rate": 3.217121159820504e-05,
"loss": 0.7272,
"step": 2710
},
{
"epoch": 2.528178854215184,
"grad_norm": 0.46412438688592356,
"learning_rate": 3.2128063513979986e-05,
"loss": 0.7425,
"step": 2715
},
{
"epoch": 2.5328365160689335,
"grad_norm": 0.4495342474637197,
"learning_rate": 3.208491542975492e-05,
"loss": 0.762,
"step": 2720
},
{
"epoch": 2.537494177922683,
"grad_norm": 0.44955454953760104,
"learning_rate": 3.204176734552986e-05,
"loss": 0.7301,
"step": 2725
},
{
"epoch": 2.542151839776432,
"grad_norm": 0.4726584773035974,
"learning_rate": 3.19986192613048e-05,
"loss": 0.7504,
"step": 2730
},
{
"epoch": 2.5468095016301815,
"grad_norm": 0.45736686524629533,
"learning_rate": 3.1955471177079735e-05,
"loss": 0.7263,
"step": 2735
},
{
"epoch": 2.551467163483931,
"grad_norm": 0.44559915699792074,
"learning_rate": 3.191232309285468e-05,
"loss": 0.7257,
"step": 2740
},
{
"epoch": 2.5561248253376805,
"grad_norm": 0.4487038810087417,
"learning_rate": 3.186917500862962e-05,
"loss": 0.7237,
"step": 2745
},
{
"epoch": 2.56078248719143,
"grad_norm": 0.41389475679174176,
"learning_rate": 3.1826026924404556e-05,
"loss": 0.7238,
"step": 2750
},
{
"epoch": 2.5654401490451795,
"grad_norm": 0.4786476587851291,
"learning_rate": 3.17828788401795e-05,
"loss": 0.7155,
"step": 2755
},
{
"epoch": 2.570097810898929,
"grad_norm": 0.489442857141447,
"learning_rate": 3.173973075595444e-05,
"loss": 0.7372,
"step": 2760
},
{
"epoch": 2.574755472752678,
"grad_norm": 0.5094148417556145,
"learning_rate": 3.169658267172938e-05,
"loss": 0.724,
"step": 2765
},
{
"epoch": 2.5794131346064275,
"grad_norm": 0.5051284189770395,
"learning_rate": 3.165343458750431e-05,
"loss": 0.7293,
"step": 2770
},
{
"epoch": 2.584070796460177,
"grad_norm": 0.5091129444887145,
"learning_rate": 3.1610286503279255e-05,
"loss": 0.7239,
"step": 2775
},
{
"epoch": 2.5887284583139265,
"grad_norm": 0.47972755418083146,
"learning_rate": 3.156713841905419e-05,
"loss": 0.7344,
"step": 2780
},
{
"epoch": 2.5933861201676756,
"grad_norm": 0.4345209631870362,
"learning_rate": 3.152399033482913e-05,
"loss": 0.7455,
"step": 2785
},
{
"epoch": 2.598043782021425,
"grad_norm": 0.4646896409021501,
"learning_rate": 3.1480842250604076e-05,
"loss": 0.7449,
"step": 2790
},
{
"epoch": 2.6027014438751745,
"grad_norm": 0.4536688273152399,
"learning_rate": 3.143769416637902e-05,
"loss": 0.7289,
"step": 2795
},
{
"epoch": 2.607359105728924,
"grad_norm": 0.47474413012505823,
"learning_rate": 3.1394546082153954e-05,
"loss": 0.7296,
"step": 2800
},
{
"epoch": 2.6120167675826735,
"grad_norm": 0.4493379121694987,
"learning_rate": 3.13513979979289e-05,
"loss": 0.7222,
"step": 2805
},
{
"epoch": 2.616674429436423,
"grad_norm": 0.4800703688735423,
"learning_rate": 3.130824991370383e-05,
"loss": 0.736,
"step": 2810
},
{
"epoch": 2.6213320912901725,
"grad_norm": 0.45499876944819995,
"learning_rate": 3.126510182947877e-05,
"loss": 0.724,
"step": 2815
},
{
"epoch": 2.625989753143922,
"grad_norm": 0.47590616696249266,
"learning_rate": 3.122195374525371e-05,
"loss": 0.7423,
"step": 2820
},
{
"epoch": 2.630647414997671,
"grad_norm": 0.4607628535501758,
"learning_rate": 3.117880566102865e-05,
"loss": 0.743,
"step": 2825
},
{
"epoch": 2.6353050768514206,
"grad_norm": 0.6908597333736273,
"learning_rate": 3.113565757680359e-05,
"loss": 0.725,
"step": 2830
},
{
"epoch": 2.63996273870517,
"grad_norm": 0.4523479184212291,
"learning_rate": 3.109250949257853e-05,
"loss": 0.7449,
"step": 2835
},
{
"epoch": 2.6446204005589196,
"grad_norm": 0.46655610331143293,
"learning_rate": 3.1049361408353474e-05,
"loss": 0.7503,
"step": 2840
},
{
"epoch": 2.6492780624126686,
"grad_norm": 0.495804001874504,
"learning_rate": 3.100621332412841e-05,
"loss": 0.7399,
"step": 2845
},
{
"epoch": 2.653935724266418,
"grad_norm": 0.4861510474965587,
"learning_rate": 3.096306523990335e-05,
"loss": 0.7416,
"step": 2850
},
{
"epoch": 2.6585933861201676,
"grad_norm": 0.43741153602282423,
"learning_rate": 3.091991715567829e-05,
"loss": 0.7293,
"step": 2855
},
{
"epoch": 2.663251047973917,
"grad_norm": 0.48413889421676215,
"learning_rate": 3.0876769071453223e-05,
"loss": 0.7422,
"step": 2860
},
{
"epoch": 2.6679087098276666,
"grad_norm": 0.4977917714734477,
"learning_rate": 3.0833620987228166e-05,
"loss": 0.7391,
"step": 2865
},
{
"epoch": 2.672566371681416,
"grad_norm": 0.4931812204297124,
"learning_rate": 3.079047290300311e-05,
"loss": 0.7235,
"step": 2870
},
{
"epoch": 2.6772240335351656,
"grad_norm": 0.42330639560301353,
"learning_rate": 3.074732481877805e-05,
"loss": 0.7322,
"step": 2875
},
{
"epoch": 2.6818816953889146,
"grad_norm": 0.4580606866066788,
"learning_rate": 3.070417673455299e-05,
"loss": 0.7199,
"step": 2880
},
{
"epoch": 2.686539357242664,
"grad_norm": 0.4685523411063327,
"learning_rate": 3.066102865032793e-05,
"loss": 0.7497,
"step": 2885
},
{
"epoch": 2.6911970190964136,
"grad_norm": 0.43758724338416044,
"learning_rate": 3.061788056610287e-05,
"loss": 0.7241,
"step": 2890
},
{
"epoch": 2.695854680950163,
"grad_norm": 0.5139581505046792,
"learning_rate": 3.057473248187781e-05,
"loss": 0.7387,
"step": 2895
},
{
"epoch": 2.700512342803912,
"grad_norm": 0.4868622670145385,
"learning_rate": 3.053158439765274e-05,
"loss": 0.7288,
"step": 2900
},
{
"epoch": 2.7051700046576617,
"grad_norm": 0.472071906864403,
"learning_rate": 3.048843631342769e-05,
"loss": 0.72,
"step": 2905
},
{
"epoch": 2.709827666511411,
"grad_norm": 0.461171899159895,
"learning_rate": 3.0445288229202625e-05,
"loss": 0.7256,
"step": 2910
},
{
"epoch": 2.7144853283651607,
"grad_norm": 0.4381661896304001,
"learning_rate": 3.0402140144977564e-05,
"loss": 0.7412,
"step": 2915
},
{
"epoch": 2.71914299021891,
"grad_norm": 0.4671330057430109,
"learning_rate": 3.0358992060752506e-05,
"loss": 0.7253,
"step": 2920
},
{
"epoch": 2.7238006520726596,
"grad_norm": 0.5346735807031436,
"learning_rate": 3.0315843976527442e-05,
"loss": 0.7486,
"step": 2925
},
{
"epoch": 2.728458313926409,
"grad_norm": 0.4354706109435107,
"learning_rate": 3.027269589230238e-05,
"loss": 0.7472,
"step": 2930
},
{
"epoch": 2.7331159757801586,
"grad_norm": 0.4437288684321577,
"learning_rate": 3.0229547808077324e-05,
"loss": 0.7269,
"step": 2935
},
{
"epoch": 2.7377736376339077,
"grad_norm": 0.5218710146843476,
"learning_rate": 3.018639972385226e-05,
"loss": 0.7287,
"step": 2940
},
{
"epoch": 2.742431299487657,
"grad_norm": 0.4459758133871711,
"learning_rate": 3.0143251639627202e-05,
"loss": 0.7414,
"step": 2945
},
{
"epoch": 2.7470889613414067,
"grad_norm": 0.46373131985846017,
"learning_rate": 3.010010355540214e-05,
"loss": 0.7273,
"step": 2950
},
{
"epoch": 2.751746623195156,
"grad_norm": 0.45258002027378524,
"learning_rate": 3.0056955471177077e-05,
"loss": 0.7207,
"step": 2955
},
{
"epoch": 2.7564042850489052,
"grad_norm": 0.4742833473816359,
"learning_rate": 3.001380738695202e-05,
"loss": 0.7327,
"step": 2960
},
{
"epoch": 2.7610619469026547,
"grad_norm": 0.4403546844429926,
"learning_rate": 2.9970659302726962e-05,
"loss": 0.73,
"step": 2965
},
{
"epoch": 2.765719608756404,
"grad_norm": 0.442379107371057,
"learning_rate": 2.99275112185019e-05,
"loss": 0.7343,
"step": 2970
},
{
"epoch": 2.7703772706101537,
"grad_norm": 0.4789864727290059,
"learning_rate": 2.9884363134276837e-05,
"loss": 0.7193,
"step": 2975
},
{
"epoch": 2.775034932463903,
"grad_norm": 0.488186649733138,
"learning_rate": 2.984121505005178e-05,
"loss": 0.7224,
"step": 2980
},
{
"epoch": 2.7796925943176527,
"grad_norm": 0.518256732525424,
"learning_rate": 2.979806696582672e-05,
"loss": 0.7403,
"step": 2985
},
{
"epoch": 2.784350256171402,
"grad_norm": 0.4382664967456426,
"learning_rate": 2.9754918881601657e-05,
"loss": 0.7272,
"step": 2990
},
{
"epoch": 2.7890079180251512,
"grad_norm": 0.47112665618864485,
"learning_rate": 2.9711770797376596e-05,
"loss": 0.7242,
"step": 2995
},
{
"epoch": 2.7936655798789007,
"grad_norm": 0.4338350589114092,
"learning_rate": 2.966862271315154e-05,
"loss": 0.7427,
"step": 3000
},
{
"epoch": 2.7983232417326502,
"grad_norm": 0.42042818529701553,
"learning_rate": 2.9625474628926475e-05,
"loss": 0.7185,
"step": 3005
},
{
"epoch": 2.8029809035863997,
"grad_norm": 0.4517788331872833,
"learning_rate": 2.9582326544701417e-05,
"loss": 0.7448,
"step": 3010
},
{
"epoch": 2.807638565440149,
"grad_norm": 0.46640580860154485,
"learning_rate": 2.9539178460476356e-05,
"loss": 0.7069,
"step": 3015
},
{
"epoch": 2.8122962272938983,
"grad_norm": 0.4720403640791126,
"learning_rate": 2.9496030376251292e-05,
"loss": 0.7383,
"step": 3020
},
{
"epoch": 2.8169538891476478,
"grad_norm": 0.45321007730262314,
"learning_rate": 2.9452882292026235e-05,
"loss": 0.7308,
"step": 3025
},
{
"epoch": 2.8216115510013973,
"grad_norm": 0.5182778932168417,
"learning_rate": 2.9409734207801177e-05,
"loss": 0.7372,
"step": 3030
},
{
"epoch": 2.8262692128551468,
"grad_norm": 0.4939735503668823,
"learning_rate": 2.9366586123576113e-05,
"loss": 0.7349,
"step": 3035
},
{
"epoch": 2.8309268747088963,
"grad_norm": 0.4955354642242852,
"learning_rate": 2.9323438039351052e-05,
"loss": 0.7522,
"step": 3040
},
{
"epoch": 2.8355845365626458,
"grad_norm": 0.4389893298550623,
"learning_rate": 2.9280289955125994e-05,
"loss": 0.7277,
"step": 3045
},
{
"epoch": 2.840242198416395,
"grad_norm": 0.4555715150024362,
"learning_rate": 2.9237141870900937e-05,
"loss": 0.7364,
"step": 3050
},
{
"epoch": 2.8448998602701443,
"grad_norm": 0.4906466706375227,
"learning_rate": 2.9193993786675873e-05,
"loss": 0.7401,
"step": 3055
},
{
"epoch": 2.849557522123894,
"grad_norm": 0.4743288708851488,
"learning_rate": 2.9150845702450812e-05,
"loss": 0.7424,
"step": 3060
},
{
"epoch": 2.8542151839776433,
"grad_norm": 0.4720591375600582,
"learning_rate": 2.9107697618225754e-05,
"loss": 0.7303,
"step": 3065
},
{
"epoch": 2.858872845831393,
"grad_norm": 0.4538021780231351,
"learning_rate": 2.906454953400069e-05,
"loss": 0.7142,
"step": 3070
},
{
"epoch": 2.863530507685142,
"grad_norm": 0.4659413361395233,
"learning_rate": 2.9021401449775632e-05,
"loss": 0.7336,
"step": 3075
},
{
"epoch": 2.8681881695388913,
"grad_norm": 0.46579867620172544,
"learning_rate": 2.897825336555057e-05,
"loss": 0.7139,
"step": 3080
},
{
"epoch": 2.872845831392641,
"grad_norm": 0.4558032880562939,
"learning_rate": 2.8935105281325507e-05,
"loss": 0.7329,
"step": 3085
},
{
"epoch": 2.8775034932463903,
"grad_norm": 0.43377128416649396,
"learning_rate": 2.889195719710045e-05,
"loss": 0.7429,
"step": 3090
},
{
"epoch": 2.88216115510014,
"grad_norm": 0.5083131527805567,
"learning_rate": 2.8848809112875392e-05,
"loss": 0.7187,
"step": 3095
},
{
"epoch": 2.8868188169538893,
"grad_norm": 0.4500221460901065,
"learning_rate": 2.8805661028650328e-05,
"loss": 0.7313,
"step": 3100
},
{
"epoch": 2.891476478807639,
"grad_norm": 0.4792267346394915,
"learning_rate": 2.8762512944425267e-05,
"loss": 0.7422,
"step": 3105
},
{
"epoch": 2.896134140661388,
"grad_norm": 0.4756004802811119,
"learning_rate": 2.871936486020021e-05,
"loss": 0.7208,
"step": 3110
},
{
"epoch": 2.9007918025151374,
"grad_norm": 0.5367158641267008,
"learning_rate": 2.8676216775975145e-05,
"loss": 0.7139,
"step": 3115
},
{
"epoch": 2.905449464368887,
"grad_norm": 0.48984639642406336,
"learning_rate": 2.8633068691750088e-05,
"loss": 0.7287,
"step": 3120
},
{
"epoch": 2.9101071262226363,
"grad_norm": 0.44180556724784487,
"learning_rate": 2.8589920607525027e-05,
"loss": 0.7391,
"step": 3125
},
{
"epoch": 2.9147647880763854,
"grad_norm": 0.4036112417048904,
"learning_rate": 2.854677252329997e-05,
"loss": 0.708,
"step": 3130
},
{
"epoch": 2.919422449930135,
"grad_norm": 0.4257374710769896,
"learning_rate": 2.8503624439074905e-05,
"loss": 0.718,
"step": 3135
},
{
"epoch": 2.9240801117838844,
"grad_norm": 0.4129638150505555,
"learning_rate": 2.8460476354849848e-05,
"loss": 0.7401,
"step": 3140
},
{
"epoch": 2.928737773637634,
"grad_norm": 0.45338592438503983,
"learning_rate": 2.8417328270624787e-05,
"loss": 0.7442,
"step": 3145
},
{
"epoch": 2.9333954354913834,
"grad_norm": 0.4527231664247476,
"learning_rate": 2.8374180186399723e-05,
"loss": 0.7401,
"step": 3150
},
{
"epoch": 2.938053097345133,
"grad_norm": 0.5703778818207271,
"learning_rate": 2.8331032102174665e-05,
"loss": 0.713,
"step": 3155
},
{
"epoch": 2.9427107591988824,
"grad_norm": 0.46786631290538294,
"learning_rate": 2.8287884017949608e-05,
"loss": 0.7337,
"step": 3160
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.44571913627760557,
"learning_rate": 2.8244735933724543e-05,
"loss": 0.7358,
"step": 3165
},
{
"epoch": 2.952026082906381,
"grad_norm": 0.42610446275618197,
"learning_rate": 2.8201587849499482e-05,
"loss": 0.7281,
"step": 3170
},
{
"epoch": 2.9566837447601304,
"grad_norm": 0.41660557764054273,
"learning_rate": 2.8158439765274425e-05,
"loss": 0.7393,
"step": 3175
},
{
"epoch": 2.96134140661388,
"grad_norm": 0.4224639364977561,
"learning_rate": 2.811529168104936e-05,
"loss": 0.7211,
"step": 3180
},
{
"epoch": 2.9659990684676294,
"grad_norm": 0.46634643211770355,
"learning_rate": 2.8072143596824303e-05,
"loss": 0.7325,
"step": 3185
},
{
"epoch": 2.9706567303213784,
"grad_norm": 0.4519644958202247,
"learning_rate": 2.8028995512599242e-05,
"loss": 0.7122,
"step": 3190
},
{
"epoch": 2.975314392175128,
"grad_norm": 0.4138426548649053,
"learning_rate": 2.7985847428374178e-05,
"loss": 0.7405,
"step": 3195
},
{
"epoch": 2.9799720540288774,
"grad_norm": 0.44704044906136875,
"learning_rate": 2.794269934414912e-05,
"loss": 0.7371,
"step": 3200
},
{
"epoch": 2.984629715882627,
"grad_norm": 0.4577403564712834,
"learning_rate": 2.7899551259924063e-05,
"loss": 0.748,
"step": 3205
},
{
"epoch": 2.9892873777363764,
"grad_norm": 0.44734870917193764,
"learning_rate": 2.7856403175699002e-05,
"loss": 0.7026,
"step": 3210
},
{
"epoch": 2.993945039590126,
"grad_norm": 0.44948584194107366,
"learning_rate": 2.7813255091473938e-05,
"loss": 0.7045,
"step": 3215
},
{
"epoch": 2.9986027014438754,
"grad_norm": 0.44929280385567383,
"learning_rate": 2.777010700724888e-05,
"loss": 0.737,
"step": 3220
},
{
"epoch": 3.0027945971122496,
"grad_norm": 0.5649620824470394,
"learning_rate": 2.7726958923023823e-05,
"loss": 0.6761,
"step": 3225
},
{
"epoch": 3.007452258965999,
"grad_norm": 0.4955694235403784,
"learning_rate": 2.768381083879876e-05,
"loss": 0.655,
"step": 3230
},
{
"epoch": 3.0121099208197486,
"grad_norm": 0.5520902971013796,
"learning_rate": 2.7640662754573698e-05,
"loss": 0.6628,
"step": 3235
},
{
"epoch": 3.016767582673498,
"grad_norm": 0.4274420618363629,
"learning_rate": 2.759751467034864e-05,
"loss": 0.6445,
"step": 3240
},
{
"epoch": 3.021425244527247,
"grad_norm": 0.5411731386852289,
"learning_rate": 2.7554366586123576e-05,
"loss": 0.6606,
"step": 3245
},
{
"epoch": 3.0260829063809966,
"grad_norm": 0.44698573945453757,
"learning_rate": 2.751121850189852e-05,
"loss": 0.6648,
"step": 3250
},
{
"epoch": 3.030740568234746,
"grad_norm": 0.438420990691833,
"learning_rate": 2.7468070417673457e-05,
"loss": 0.6663,
"step": 3255
},
{
"epoch": 3.0353982300884956,
"grad_norm": 0.455229642854329,
"learning_rate": 2.7424922333448393e-05,
"loss": 0.6683,
"step": 3260
},
{
"epoch": 3.040055891942245,
"grad_norm": 0.49190199547309743,
"learning_rate": 2.7381774249223336e-05,
"loss": 0.6731,
"step": 3265
},
{
"epoch": 3.0447135537959946,
"grad_norm": 0.4487013040424457,
"learning_rate": 2.7338626164998278e-05,
"loss": 0.6528,
"step": 3270
},
{
"epoch": 3.0493712156497437,
"grad_norm": 0.43437068343719426,
"learning_rate": 2.7295478080773214e-05,
"loss": 0.6607,
"step": 3275
},
{
"epoch": 3.054028877503493,
"grad_norm": 0.4582623419182831,
"learning_rate": 2.7252329996548153e-05,
"loss": 0.65,
"step": 3280
},
{
"epoch": 3.0586865393572427,
"grad_norm": 0.46784851131740257,
"learning_rate": 2.7209181912323096e-05,
"loss": 0.6817,
"step": 3285
},
{
"epoch": 3.063344201210992,
"grad_norm": 0.4618365438998452,
"learning_rate": 2.7166033828098038e-05,
"loss": 0.659,
"step": 3290
},
{
"epoch": 3.0680018630647417,
"grad_norm": 0.5051196818431409,
"learning_rate": 2.7122885743872974e-05,
"loss": 0.6664,
"step": 3295
},
{
"epoch": 3.0726595249184907,
"grad_norm": 0.5278805351509813,
"learning_rate": 2.7079737659647913e-05,
"loss": 0.655,
"step": 3300
},
{
"epoch": 3.07731718677224,
"grad_norm": 0.5238205105124322,
"learning_rate": 2.7036589575422855e-05,
"loss": 0.6501,
"step": 3305
},
{
"epoch": 3.0819748486259897,
"grad_norm": 0.5238389473428904,
"learning_rate": 2.699344149119779e-05,
"loss": 0.6587,
"step": 3310
},
{
"epoch": 3.086632510479739,
"grad_norm": 0.4956542427920923,
"learning_rate": 2.6950293406972734e-05,
"loss": 0.652,
"step": 3315
},
{
"epoch": 3.0912901723334887,
"grad_norm": 0.4323544530467572,
"learning_rate": 2.6907145322747673e-05,
"loss": 0.6423,
"step": 3320
},
{
"epoch": 3.095947834187238,
"grad_norm": 0.4501450597084916,
"learning_rate": 2.686399723852261e-05,
"loss": 0.6727,
"step": 3325
},
{
"epoch": 3.1006054960409872,
"grad_norm": 0.5342968111952526,
"learning_rate": 2.682084915429755e-05,
"loss": 0.6613,
"step": 3330
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.4952428719045549,
"learning_rate": 2.677770107007249e-05,
"loss": 0.6539,
"step": 3335
},
{
"epoch": 3.109920819748486,
"grad_norm": 0.4464148435601403,
"learning_rate": 2.673455298584743e-05,
"loss": 0.6496,
"step": 3340
},
{
"epoch": 3.1145784816022357,
"grad_norm": 0.42423733730735846,
"learning_rate": 2.6691404901622368e-05,
"loss": 0.662,
"step": 3345
},
{
"epoch": 3.119236143455985,
"grad_norm": 0.4526284930230991,
"learning_rate": 2.664825681739731e-05,
"loss": 0.6862,
"step": 3350
},
{
"epoch": 3.1238938053097347,
"grad_norm": 0.4327570355006767,
"learning_rate": 2.6605108733172247e-05,
"loss": 0.6616,
"step": 3355
},
{
"epoch": 3.1285514671634838,
"grad_norm": 0.4180247578911309,
"learning_rate": 2.6561960648947186e-05,
"loss": 0.6613,
"step": 3360
},
{
"epoch": 3.1332091290172333,
"grad_norm": 0.48989064474286764,
"learning_rate": 2.6518812564722128e-05,
"loss": 0.673,
"step": 3365
},
{
"epoch": 3.1378667908709827,
"grad_norm": 0.45013987334870337,
"learning_rate": 2.647566448049707e-05,
"loss": 0.6647,
"step": 3370
},
{
"epoch": 3.1425244527247322,
"grad_norm": 0.4306897032201952,
"learning_rate": 2.6432516396272006e-05,
"loss": 0.663,
"step": 3375
},
{
"epoch": 3.1471821145784817,
"grad_norm": 0.45218980570674416,
"learning_rate": 2.6389368312046945e-05,
"loss": 0.6626,
"step": 3380
},
{
"epoch": 3.1518397764322312,
"grad_norm": 0.4269558776822659,
"learning_rate": 2.6346220227821888e-05,
"loss": 0.6587,
"step": 3385
},
{
"epoch": 3.1564974382859803,
"grad_norm": 0.4570131615052829,
"learning_rate": 2.6303072143596824e-05,
"loss": 0.6681,
"step": 3390
},
{
"epoch": 3.16115510013973,
"grad_norm": 0.41909205537848343,
"learning_rate": 2.6259924059371766e-05,
"loss": 0.6512,
"step": 3395
},
{
"epoch": 3.1658127619934793,
"grad_norm": 0.45070558039326897,
"learning_rate": 2.6216775975146705e-05,
"loss": 0.6688,
"step": 3400
},
{
"epoch": 3.1704704238472288,
"grad_norm": 0.4418832834353741,
"learning_rate": 2.617362789092164e-05,
"loss": 0.6742,
"step": 3405
},
{
"epoch": 3.1751280857009783,
"grad_norm": 0.43144075729659026,
"learning_rate": 2.6130479806696584e-05,
"loss": 0.6819,
"step": 3410
},
{
"epoch": 3.1797857475547273,
"grad_norm": 0.4461048208336212,
"learning_rate": 2.6087331722471526e-05,
"loss": 0.6578,
"step": 3415
},
{
"epoch": 3.184443409408477,
"grad_norm": 0.495460849711094,
"learning_rate": 2.6044183638246462e-05,
"loss": 0.6808,
"step": 3420
},
{
"epoch": 3.1891010712622263,
"grad_norm": 0.5193560885811147,
"learning_rate": 2.60010355540214e-05,
"loss": 0.6629,
"step": 3425
},
{
"epoch": 3.193758733115976,
"grad_norm": 0.520924441326943,
"learning_rate": 2.5957887469796343e-05,
"loss": 0.6854,
"step": 3430
},
{
"epoch": 3.1984163949697253,
"grad_norm": 0.4557258318037281,
"learning_rate": 2.591473938557128e-05,
"loss": 0.6797,
"step": 3435
},
{
"epoch": 3.203074056823475,
"grad_norm": 0.48141994845392755,
"learning_rate": 2.587159130134622e-05,
"loss": 0.6728,
"step": 3440
},
{
"epoch": 3.207731718677224,
"grad_norm": 0.4813706001238264,
"learning_rate": 2.582844321712116e-05,
"loss": 0.6529,
"step": 3445
},
{
"epoch": 3.2123893805309733,
"grad_norm": 0.4510461852751938,
"learning_rate": 2.5785295132896096e-05,
"loss": 0.6669,
"step": 3450
},
{
"epoch": 3.217047042384723,
"grad_norm": 0.4602528355138011,
"learning_rate": 2.574214704867104e-05,
"loss": 0.6725,
"step": 3455
},
{
"epoch": 3.2217047042384723,
"grad_norm": 0.4633154565224799,
"learning_rate": 2.569899896444598e-05,
"loss": 0.6692,
"step": 3460
},
{
"epoch": 3.226362366092222,
"grad_norm": 0.47600755891130275,
"learning_rate": 2.565585088022092e-05,
"loss": 0.6613,
"step": 3465
},
{
"epoch": 3.2310200279459713,
"grad_norm": 0.4882939190293464,
"learning_rate": 2.5612702795995856e-05,
"loss": 0.674,
"step": 3470
},
{
"epoch": 3.2356776897997204,
"grad_norm": 0.4532071802883846,
"learning_rate": 2.55695547117708e-05,
"loss": 0.6606,
"step": 3475
},
{
"epoch": 3.24033535165347,
"grad_norm": 0.4296635507904013,
"learning_rate": 2.552640662754574e-05,
"loss": 0.67,
"step": 3480
},
{
"epoch": 3.2449930135072194,
"grad_norm": 0.5017255136543564,
"learning_rate": 2.5483258543320677e-05,
"loss": 0.6871,
"step": 3485
},
{
"epoch": 3.249650675360969,
"grad_norm": 0.4948112197113389,
"learning_rate": 2.5440110459095616e-05,
"loss": 0.693,
"step": 3490
},
{
"epoch": 3.2543083372147183,
"grad_norm": 0.49247917585392487,
"learning_rate": 2.539696237487056e-05,
"loss": 0.6876,
"step": 3495
},
{
"epoch": 3.258965999068468,
"grad_norm": 0.5347369741382454,
"learning_rate": 2.5353814290645494e-05,
"loss": 0.6849,
"step": 3500
},
{
"epoch": 3.263623660922217,
"grad_norm": 0.495732345471852,
"learning_rate": 2.5310666206420437e-05,
"loss": 0.6908,
"step": 3505
},
{
"epoch": 3.2682813227759664,
"grad_norm": 0.45392823221334166,
"learning_rate": 2.5267518122195376e-05,
"loss": 0.6901,
"step": 3510
},
{
"epoch": 3.272938984629716,
"grad_norm": 0.44957677688338854,
"learning_rate": 2.5224370037970312e-05,
"loss": 0.6776,
"step": 3515
},
{
"epoch": 3.2775966464834654,
"grad_norm": 0.448342787281423,
"learning_rate": 2.5181221953745254e-05,
"loss": 0.6751,
"step": 3520
},
{
"epoch": 3.282254308337215,
"grad_norm": 0.4753362927286903,
"learning_rate": 2.5138073869520197e-05,
"loss": 0.7008,
"step": 3525
},
{
"epoch": 3.286911970190964,
"grad_norm": 0.4303570706933123,
"learning_rate": 2.5094925785295132e-05,
"loss": 0.6564,
"step": 3530
},
{
"epoch": 3.2915696320447134,
"grad_norm": 0.4566321018828311,
"learning_rate": 2.505177770107007e-05,
"loss": 0.6548,
"step": 3535
},
{
"epoch": 3.296227293898463,
"grad_norm": 0.5013218048356842,
"learning_rate": 2.5008629616845014e-05,
"loss": 0.6602,
"step": 3540
},
{
"epoch": 3.3008849557522124,
"grad_norm": 0.48919607850045277,
"learning_rate": 2.4965481532619953e-05,
"loss": 0.6775,
"step": 3545
},
{
"epoch": 3.305542617605962,
"grad_norm": 0.48623977714687,
"learning_rate": 2.4922333448394892e-05,
"loss": 0.6658,
"step": 3550
},
{
"epoch": 3.3102002794597114,
"grad_norm": 0.49148464778598894,
"learning_rate": 2.487918536416983e-05,
"loss": 0.6521,
"step": 3555
},
{
"epoch": 3.3148579413134605,
"grad_norm": 0.4893622152875232,
"learning_rate": 2.483603727994477e-05,
"loss": 0.6684,
"step": 3560
},
{
"epoch": 3.31951560316721,
"grad_norm": 0.4536686442042023,
"learning_rate": 2.4792889195719713e-05,
"loss": 0.6689,
"step": 3565
},
{
"epoch": 3.3241732650209594,
"grad_norm": 0.4736127984225089,
"learning_rate": 2.4749741111494652e-05,
"loss": 0.6622,
"step": 3570
},
{
"epoch": 3.328830926874709,
"grad_norm": 0.46628026872899403,
"learning_rate": 2.4706593027269588e-05,
"loss": 0.6813,
"step": 3575
},
{
"epoch": 3.3334885887284584,
"grad_norm": 0.4999633681973708,
"learning_rate": 2.466344494304453e-05,
"loss": 0.6731,
"step": 3580
},
{
"epoch": 3.3381462505822075,
"grad_norm": 0.5641043771334887,
"learning_rate": 2.462029685881947e-05,
"loss": 0.6767,
"step": 3585
},
{
"epoch": 3.342803912435957,
"grad_norm": 0.45714731895877286,
"learning_rate": 2.457714877459441e-05,
"loss": 0.6643,
"step": 3590
},
{
"epoch": 3.3474615742897065,
"grad_norm": 0.4334973778339274,
"learning_rate": 2.4534000690369348e-05,
"loss": 0.6614,
"step": 3595
},
{
"epoch": 3.352119236143456,
"grad_norm": 0.4886669385610417,
"learning_rate": 2.4490852606144287e-05,
"loss": 0.664,
"step": 3600
},
{
"epoch": 3.3567768979972055,
"grad_norm": 0.45151461746127364,
"learning_rate": 2.444770452191923e-05,
"loss": 0.6492,
"step": 3605
},
{
"epoch": 3.361434559850955,
"grad_norm": 0.4532080460532112,
"learning_rate": 2.440455643769417e-05,
"loss": 0.6684,
"step": 3610
},
{
"epoch": 3.3660922217047045,
"grad_norm": 0.4607883027240199,
"learning_rate": 2.4361408353469108e-05,
"loss": 0.6627,
"step": 3615
},
{
"epoch": 3.3707498835584535,
"grad_norm": 0.5184420407083233,
"learning_rate": 2.4318260269244047e-05,
"loss": 0.6633,
"step": 3620
},
{
"epoch": 3.375407545412203,
"grad_norm": 0.42981832962189226,
"learning_rate": 2.4275112185018986e-05,
"loss": 0.6794,
"step": 3625
},
{
"epoch": 3.3800652072659525,
"grad_norm": 0.4696743898481451,
"learning_rate": 2.4231964100793925e-05,
"loss": 0.6665,
"step": 3630
},
{
"epoch": 3.384722869119702,
"grad_norm": 0.4471999488266161,
"learning_rate": 2.4188816016568867e-05,
"loss": 0.6842,
"step": 3635
},
{
"epoch": 3.3893805309734515,
"grad_norm": 0.42037921982570414,
"learning_rate": 2.4145667932343803e-05,
"loss": 0.6449,
"step": 3640
},
{
"epoch": 3.3940381928272005,
"grad_norm": 0.452619450283994,
"learning_rate": 2.4102519848118746e-05,
"loss": 0.6728,
"step": 3645
},
{
"epoch": 3.39869585468095,
"grad_norm": 0.42839450229764453,
"learning_rate": 2.4059371763893685e-05,
"loss": 0.6549,
"step": 3650
},
{
"epoch": 3.4033535165346995,
"grad_norm": 0.4358545473409289,
"learning_rate": 2.4016223679668624e-05,
"loss": 0.6831,
"step": 3655
},
{
"epoch": 3.408011178388449,
"grad_norm": 0.44463467097834486,
"learning_rate": 2.3973075595443563e-05,
"loss": 0.6588,
"step": 3660
},
{
"epoch": 3.4126688402421985,
"grad_norm": 0.44918393535421824,
"learning_rate": 2.3929927511218502e-05,
"loss": 0.6592,
"step": 3665
},
{
"epoch": 3.417326502095948,
"grad_norm": 0.45572870889698214,
"learning_rate": 2.388677942699344e-05,
"loss": 0.6847,
"step": 3670
},
{
"epoch": 3.421984163949697,
"grad_norm": 0.5312757889979894,
"learning_rate": 2.3843631342768384e-05,
"loss": 0.6639,
"step": 3675
},
{
"epoch": 3.4266418258034466,
"grad_norm": 0.5266603067472917,
"learning_rate": 2.3800483258543323e-05,
"loss": 0.6482,
"step": 3680
},
{
"epoch": 3.431299487657196,
"grad_norm": 0.4584084291790257,
"learning_rate": 2.3757335174318262e-05,
"loss": 0.6757,
"step": 3685
},
{
"epoch": 3.4359571495109456,
"grad_norm": 0.4304692509027962,
"learning_rate": 2.37141870900932e-05,
"loss": 0.6596,
"step": 3690
},
{
"epoch": 3.440614811364695,
"grad_norm": 0.4655682089191918,
"learning_rate": 2.367103900586814e-05,
"loss": 0.6664,
"step": 3695
},
{
"epoch": 3.445272473218444,
"grad_norm": 0.4473286471345806,
"learning_rate": 2.3627890921643083e-05,
"loss": 0.6814,
"step": 3700
},
{
"epoch": 3.4499301350721936,
"grad_norm": 0.4342461882548441,
"learning_rate": 2.358474283741802e-05,
"loss": 0.6859,
"step": 3705
},
{
"epoch": 3.454587796925943,
"grad_norm": 0.4634466813676635,
"learning_rate": 2.3541594753192957e-05,
"loss": 0.675,
"step": 3710
},
{
"epoch": 3.4592454587796926,
"grad_norm": 0.463462899087277,
"learning_rate": 2.34984466689679e-05,
"loss": 0.6581,
"step": 3715
},
{
"epoch": 3.463903120633442,
"grad_norm": 0.4718492791389515,
"learning_rate": 2.345529858474284e-05,
"loss": 0.6633,
"step": 3720
},
{
"epoch": 3.4685607824871916,
"grad_norm": 0.46408016382949474,
"learning_rate": 2.3412150500517778e-05,
"loss": 0.6803,
"step": 3725
},
{
"epoch": 3.473218444340941,
"grad_norm": 0.4515706039077529,
"learning_rate": 2.3369002416292717e-05,
"loss": 0.672,
"step": 3730
},
{
"epoch": 3.47787610619469,
"grad_norm": 0.4546481906424658,
"learning_rate": 2.3325854332067656e-05,
"loss": 0.6788,
"step": 3735
},
{
"epoch": 3.4825337680484396,
"grad_norm": 0.45260779635281945,
"learning_rate": 2.32827062478426e-05,
"loss": 0.6621,
"step": 3740
},
{
"epoch": 3.487191429902189,
"grad_norm": 0.4465723090136659,
"learning_rate": 2.3239558163617538e-05,
"loss": 0.6711,
"step": 3745
},
{
"epoch": 3.4918490917559386,
"grad_norm": 0.4503456117275582,
"learning_rate": 2.3196410079392474e-05,
"loss": 0.6787,
"step": 3750
},
{
"epoch": 3.496506753609688,
"grad_norm": 0.4488480782408311,
"learning_rate": 2.3153261995167416e-05,
"loss": 0.6715,
"step": 3755
},
{
"epoch": 3.501164415463437,
"grad_norm": 0.4807688539808926,
"learning_rate": 2.3110113910942355e-05,
"loss": 0.6691,
"step": 3760
},
{
"epoch": 3.5058220773171866,
"grad_norm": 0.5167283934755129,
"learning_rate": 2.3066965826717294e-05,
"loss": 0.6556,
"step": 3765
},
{
"epoch": 3.510479739170936,
"grad_norm": 0.5540283763976265,
"learning_rate": 2.3023817742492234e-05,
"loss": 0.6796,
"step": 3770
},
{
"epoch": 3.5151374010246856,
"grad_norm": 0.42946681800570363,
"learning_rate": 2.2980669658267173e-05,
"loss": 0.659,
"step": 3775
},
{
"epoch": 3.519795062878435,
"grad_norm": 0.42051099266667324,
"learning_rate": 2.2937521574042115e-05,
"loss": 0.6409,
"step": 3780
},
{
"epoch": 3.5244527247321846,
"grad_norm": 22.372062876041408,
"learning_rate": 2.2894373489817054e-05,
"loss": 0.8085,
"step": 3785
},
{
"epoch": 3.529110386585934,
"grad_norm": 0.4376085550930596,
"learning_rate": 2.285122540559199e-05,
"loss": 0.6663,
"step": 3790
},
{
"epoch": 3.533768048439683,
"grad_norm": 0.4721290842606934,
"learning_rate": 2.2808077321366933e-05,
"loss": 0.6597,
"step": 3795
},
{
"epoch": 3.5384257102934327,
"grad_norm": 0.4524333273382029,
"learning_rate": 2.276492923714187e-05,
"loss": 0.6813,
"step": 3800
},
{
"epoch": 3.543083372147182,
"grad_norm": 0.4683136917340155,
"learning_rate": 2.2721781152916814e-05,
"loss": 0.6913,
"step": 3805
},
{
"epoch": 3.5477410340009317,
"grad_norm": 0.44767173362613494,
"learning_rate": 2.267863306869175e-05,
"loss": 0.6795,
"step": 3810
},
{
"epoch": 3.5523986958546807,
"grad_norm": 0.42108561845735404,
"learning_rate": 2.263548498446669e-05,
"loss": 0.678,
"step": 3815
},
{
"epoch": 3.55705635770843,
"grad_norm": 0.42298340047101063,
"learning_rate": 2.259233690024163e-05,
"loss": 0.6593,
"step": 3820
},
{
"epoch": 3.5617140195621797,
"grad_norm": 0.4416808046947552,
"learning_rate": 2.254918881601657e-05,
"loss": 0.6868,
"step": 3825
},
{
"epoch": 3.566371681415929,
"grad_norm": 0.4488088247207295,
"learning_rate": 2.250604073179151e-05,
"loss": 0.6741,
"step": 3830
},
{
"epoch": 3.5710293432696787,
"grad_norm": 0.4212039178070685,
"learning_rate": 2.246289264756645e-05,
"loss": 0.6811,
"step": 3835
},
{
"epoch": 3.575687005123428,
"grad_norm": 0.44260327570897895,
"learning_rate": 2.2419744563341388e-05,
"loss": 0.6686,
"step": 3840
},
{
"epoch": 3.5803446669771777,
"grad_norm": 0.4438396142882864,
"learning_rate": 2.237659647911633e-05,
"loss": 0.659,
"step": 3845
},
{
"epoch": 3.5850023288309267,
"grad_norm": 0.44951766788920505,
"learning_rate": 2.233344839489127e-05,
"loss": 0.6702,
"step": 3850
},
{
"epoch": 3.5896599906846762,
"grad_norm": 0.4313373929106834,
"learning_rate": 2.2290300310666205e-05,
"loss": 0.6851,
"step": 3855
},
{
"epoch": 3.5943176525384257,
"grad_norm": 0.4271227216658829,
"learning_rate": 2.2247152226441148e-05,
"loss": 0.6717,
"step": 3860
},
{
"epoch": 3.598975314392175,
"grad_norm": 0.4648664337839032,
"learning_rate": 2.2204004142216087e-05,
"loss": 0.6823,
"step": 3865
},
{
"epoch": 3.6036329762459243,
"grad_norm": 0.4473468080782318,
"learning_rate": 2.2160856057991026e-05,
"loss": 0.6735,
"step": 3870
},
{
"epoch": 3.6082906380996738,
"grad_norm": 0.43996342525349236,
"learning_rate": 2.2117707973765965e-05,
"loss": 0.6734,
"step": 3875
},
{
"epoch": 3.6129482999534233,
"grad_norm": 0.4291942243921852,
"learning_rate": 2.2074559889540904e-05,
"loss": 0.6789,
"step": 3880
},
{
"epoch": 3.6176059618071728,
"grad_norm": 0.4765745278898558,
"learning_rate": 2.2031411805315847e-05,
"loss": 0.668,
"step": 3885
},
{
"epoch": 3.6222636236609222,
"grad_norm": 0.41815398973713275,
"learning_rate": 2.1988263721090786e-05,
"loss": 0.6682,
"step": 3890
},
{
"epoch": 3.6269212855146717,
"grad_norm": 0.43719379830239535,
"learning_rate": 2.1945115636865725e-05,
"loss": 0.6753,
"step": 3895
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.4006880656840204,
"learning_rate": 2.1901967552640664e-05,
"loss": 0.6505,
"step": 3900
},
{
"epoch": 3.6362366092221707,
"grad_norm": 0.4121265038250903,
"learning_rate": 2.1858819468415603e-05,
"loss": 0.6857,
"step": 3905
},
{
"epoch": 3.64089427107592,
"grad_norm": 0.4326970184912615,
"learning_rate": 2.1815671384190542e-05,
"loss": 0.6683,
"step": 3910
},
{
"epoch": 3.6455519329296693,
"grad_norm": 0.4201491695924619,
"learning_rate": 2.1772523299965485e-05,
"loss": 0.6771,
"step": 3915
},
{
"epoch": 3.6502095947834188,
"grad_norm": 0.43258507023687753,
"learning_rate": 2.172937521574042e-05,
"loss": 0.6584,
"step": 3920
},
{
"epoch": 3.6548672566371683,
"grad_norm": 0.4585906560215873,
"learning_rate": 2.1686227131515363e-05,
"loss": 0.6373,
"step": 3925
},
{
"epoch": 3.6595249184909173,
"grad_norm": 0.43713677266751094,
"learning_rate": 2.1643079047290302e-05,
"loss": 0.6701,
"step": 3930
},
{
"epoch": 3.664182580344667,
"grad_norm": 0.452764864256112,
"learning_rate": 2.159993096306524e-05,
"loss": 0.6669,
"step": 3935
},
{
"epoch": 3.6688402421984163,
"grad_norm": 0.41850536874968525,
"learning_rate": 2.155678287884018e-05,
"loss": 0.6703,
"step": 3940
},
{
"epoch": 3.673497904052166,
"grad_norm": 0.4785367330909523,
"learning_rate": 2.151363479461512e-05,
"loss": 0.6653,
"step": 3945
},
{
"epoch": 3.6781555659059153,
"grad_norm": 0.42672653887627465,
"learning_rate": 2.147048671039006e-05,
"loss": 0.6523,
"step": 3950
},
{
"epoch": 3.682813227759665,
"grad_norm": 0.38619011006304327,
"learning_rate": 2.1427338626165e-05,
"loss": 0.6766,
"step": 3955
},
{
"epoch": 3.6874708896134143,
"grad_norm": 0.48329034915930996,
"learning_rate": 2.1384190541939937e-05,
"loss": 0.6794,
"step": 3960
},
{
"epoch": 3.6921285514671633,
"grad_norm": 0.41716760703725536,
"learning_rate": 2.134104245771488e-05,
"loss": 0.6674,
"step": 3965
},
{
"epoch": 3.696786213320913,
"grad_norm": 0.46039645375421206,
"learning_rate": 2.129789437348982e-05,
"loss": 0.6829,
"step": 3970
},
{
"epoch": 3.7014438751746623,
"grad_norm": 0.41626180642112476,
"learning_rate": 2.1254746289264758e-05,
"loss": 0.648,
"step": 3975
},
{
"epoch": 3.706101537028412,
"grad_norm": 0.447684709770321,
"learning_rate": 2.1211598205039697e-05,
"loss": 0.6604,
"step": 3980
},
{
"epoch": 3.710759198882161,
"grad_norm": 0.41527686884250226,
"learning_rate": 2.1168450120814636e-05,
"loss": 0.6944,
"step": 3985
},
{
"epoch": 3.7154168607359104,
"grad_norm": 0.4673707506990223,
"learning_rate": 2.1125302036589575e-05,
"loss": 0.6715,
"step": 3990
},
{
"epoch": 3.72007452258966,
"grad_norm": 0.4259681640946891,
"learning_rate": 2.1082153952364517e-05,
"loss": 0.6609,
"step": 3995
},
{
"epoch": 3.7247321844434094,
"grad_norm": 0.44578301736933307,
"learning_rate": 2.1039005868139457e-05,
"loss": 0.6669,
"step": 4000
},
{
"epoch": 3.729389846297159,
"grad_norm": 0.4201611550934968,
"learning_rate": 2.0995857783914392e-05,
"loss": 0.6676,
"step": 4005
},
{
"epoch": 3.7340475081509084,
"grad_norm": 0.4451294722121708,
"learning_rate": 2.0952709699689335e-05,
"loss": 0.6807,
"step": 4010
},
{
"epoch": 3.738705170004658,
"grad_norm": 0.39293811743802215,
"learning_rate": 2.0909561615464274e-05,
"loss": 0.6537,
"step": 4015
},
{
"epoch": 3.7433628318584073,
"grad_norm": 0.425727834151418,
"learning_rate": 2.0866413531239216e-05,
"loss": 0.6812,
"step": 4020
},
{
"epoch": 3.7480204937121564,
"grad_norm": 0.43571368169120256,
"learning_rate": 2.0823265447014152e-05,
"loss": 0.6749,
"step": 4025
},
{
"epoch": 3.752678155565906,
"grad_norm": 0.40642489831038664,
"learning_rate": 2.078011736278909e-05,
"loss": 0.6701,
"step": 4030
},
{
"epoch": 3.7573358174196554,
"grad_norm": 0.46216496870944207,
"learning_rate": 2.0736969278564034e-05,
"loss": 0.6635,
"step": 4035
},
{
"epoch": 3.761993479273405,
"grad_norm": 0.4365114089567727,
"learning_rate": 2.0693821194338973e-05,
"loss": 0.6608,
"step": 4040
},
{
"epoch": 3.766651141127154,
"grad_norm": 0.43741874072599796,
"learning_rate": 2.0650673110113912e-05,
"loss": 0.6885,
"step": 4045
},
{
"epoch": 3.7713088029809034,
"grad_norm": 0.42096705297972964,
"learning_rate": 2.060752502588885e-05,
"loss": 0.6802,
"step": 4050
},
{
"epoch": 3.775966464834653,
"grad_norm": 0.49406108041678903,
"learning_rate": 2.056437694166379e-05,
"loss": 0.6843,
"step": 4055
},
{
"epoch": 3.7806241266884024,
"grad_norm": 0.42050323757733654,
"learning_rate": 2.0521228857438733e-05,
"loss": 0.6659,
"step": 4060
},
{
"epoch": 3.785281788542152,
"grad_norm": 0.43875536709277346,
"learning_rate": 2.0478080773213672e-05,
"loss": 0.6593,
"step": 4065
},
{
"epoch": 3.7899394503959014,
"grad_norm": 0.4551758764550099,
"learning_rate": 2.0434932688988608e-05,
"loss": 0.6676,
"step": 4070
},
{
"epoch": 3.794597112249651,
"grad_norm": 0.43875906331559933,
"learning_rate": 2.039178460476355e-05,
"loss": 0.6633,
"step": 4075
},
{
"epoch": 3.7992547741034,
"grad_norm": 0.49522270487134434,
"learning_rate": 2.034863652053849e-05,
"loss": 0.6723,
"step": 4080
},
{
"epoch": 3.8039124359571495,
"grad_norm": 0.4255138431861227,
"learning_rate": 2.0305488436313428e-05,
"loss": 0.6707,
"step": 4085
},
{
"epoch": 3.808570097810899,
"grad_norm": 0.4249333456739798,
"learning_rate": 2.0262340352088367e-05,
"loss": 0.6672,
"step": 4090
},
{
"epoch": 3.8132277596646484,
"grad_norm": 0.4222583096651918,
"learning_rate": 2.0219192267863306e-05,
"loss": 0.6578,
"step": 4095
},
{
"epoch": 3.8178854215183975,
"grad_norm": 0.46110284243808286,
"learning_rate": 2.017604418363825e-05,
"loss": 0.665,
"step": 4100
},
{
"epoch": 3.822543083372147,
"grad_norm": 0.439490642784066,
"learning_rate": 2.0132896099413188e-05,
"loss": 0.6601,
"step": 4105
},
{
"epoch": 3.8272007452258965,
"grad_norm": 0.44224347985960144,
"learning_rate": 2.0089748015188127e-05,
"loss": 0.6647,
"step": 4110
},
{
"epoch": 3.831858407079646,
"grad_norm": 0.43734391875827633,
"learning_rate": 2.0046599930963066e-05,
"loss": 0.6789,
"step": 4115
},
{
"epoch": 3.8365160689333955,
"grad_norm": 0.44541776084227985,
"learning_rate": 2.0003451846738005e-05,
"loss": 0.6567,
"step": 4120
},
{
"epoch": 3.841173730787145,
"grad_norm": 0.4351663828609302,
"learning_rate": 1.9960303762512945e-05,
"loss": 0.6774,
"step": 4125
},
{
"epoch": 3.8458313926408945,
"grad_norm": 0.46212604143724856,
"learning_rate": 1.9917155678287887e-05,
"loss": 0.6816,
"step": 4130
},
{
"epoch": 3.850489054494644,
"grad_norm": 0.4827514548611152,
"learning_rate": 1.9874007594062823e-05,
"loss": 0.6771,
"step": 4135
},
{
"epoch": 3.855146716348393,
"grad_norm": 0.466359350965105,
"learning_rate": 1.9830859509837765e-05,
"loss": 0.6727,
"step": 4140
},
{
"epoch": 3.8598043782021425,
"grad_norm": 0.4288376838447238,
"learning_rate": 1.9787711425612704e-05,
"loss": 0.6782,
"step": 4145
},
{
"epoch": 3.864462040055892,
"grad_norm": 0.4221479927446133,
"learning_rate": 1.9744563341387643e-05,
"loss": 0.6582,
"step": 4150
},
{
"epoch": 3.8691197019096415,
"grad_norm": 0.42808299688301105,
"learning_rate": 1.9701415257162583e-05,
"loss": 0.6633,
"step": 4155
},
{
"epoch": 3.8737773637633905,
"grad_norm": 0.4385868446482319,
"learning_rate": 1.9658267172937522e-05,
"loss": 0.6957,
"step": 4160
},
{
"epoch": 3.87843502561714,
"grad_norm": 0.46664060467404894,
"learning_rate": 1.961511908871246e-05,
"loss": 0.6815,
"step": 4165
},
{
"epoch": 3.8830926874708895,
"grad_norm": 0.4415348584856722,
"learning_rate": 1.9571971004487403e-05,
"loss": 0.6647,
"step": 4170
},
{
"epoch": 3.887750349324639,
"grad_norm": 0.449942040491185,
"learning_rate": 1.952882292026234e-05,
"loss": 0.68,
"step": 4175
},
{
"epoch": 3.8924080111783885,
"grad_norm": 0.4414098811625752,
"learning_rate": 1.948567483603728e-05,
"loss": 0.671,
"step": 4180
},
{
"epoch": 3.897065673032138,
"grad_norm": 0.42891467747530376,
"learning_rate": 1.944252675181222e-05,
"loss": 0.6732,
"step": 4185
},
{
"epoch": 3.9017233348858875,
"grad_norm": 0.41584348048954695,
"learning_rate": 1.939937866758716e-05,
"loss": 0.6689,
"step": 4190
},
{
"epoch": 3.9063809967396366,
"grad_norm": 0.43294634671450094,
"learning_rate": 1.93562305833621e-05,
"loss": 0.6562,
"step": 4195
},
{
"epoch": 3.911038658593386,
"grad_norm": 0.43238591716188074,
"learning_rate": 1.9313082499137038e-05,
"loss": 0.6619,
"step": 4200
},
{
"epoch": 3.9156963204471356,
"grad_norm": 0.417950730256557,
"learning_rate": 1.9269934414911977e-05,
"loss": 0.6711,
"step": 4205
},
{
"epoch": 3.920353982300885,
"grad_norm": 0.40509888786084697,
"learning_rate": 1.922678633068692e-05,
"loss": 0.6623,
"step": 4210
},
{
"epoch": 3.925011644154634,
"grad_norm": 0.4243068617748856,
"learning_rate": 1.918363824646186e-05,
"loss": 0.6583,
"step": 4215
},
{
"epoch": 3.9296693060083836,
"grad_norm": 0.41822911394479817,
"learning_rate": 1.9140490162236798e-05,
"loss": 0.6645,
"step": 4220
},
{
"epoch": 3.934326967862133,
"grad_norm": 0.4377285415892218,
"learning_rate": 1.9097342078011737e-05,
"loss": 0.6603,
"step": 4225
},
{
"epoch": 3.9389846297158826,
"grad_norm": 0.4141650537612767,
"learning_rate": 1.9054193993786676e-05,
"loss": 0.6673,
"step": 4230
},
{
"epoch": 3.943642291569632,
"grad_norm": 0.4382605865395,
"learning_rate": 1.901104590956162e-05,
"loss": 0.6643,
"step": 4235
},
{
"epoch": 3.9482999534233816,
"grad_norm": 0.4202680156252619,
"learning_rate": 1.8967897825336554e-05,
"loss": 0.6587,
"step": 4240
},
{
"epoch": 3.952957615277131,
"grad_norm": 0.40387812732388845,
"learning_rate": 1.8924749741111493e-05,
"loss": 0.6601,
"step": 4245
},
{
"epoch": 3.9576152771308806,
"grad_norm": 0.4156926899477138,
"learning_rate": 1.8881601656886436e-05,
"loss": 0.6676,
"step": 4250
},
{
"epoch": 3.9622729389846296,
"grad_norm": 0.42946870450766167,
"learning_rate": 1.8838453572661375e-05,
"loss": 0.6683,
"step": 4255
},
{
"epoch": 3.966930600838379,
"grad_norm": 0.43212939224797026,
"learning_rate": 1.8795305488436314e-05,
"loss": 0.6792,
"step": 4260
},
{
"epoch": 3.9715882626921286,
"grad_norm": 0.43345652576381805,
"learning_rate": 1.8752157404211253e-05,
"loss": 0.678,
"step": 4265
},
{
"epoch": 3.976245924545878,
"grad_norm": 0.40339649642482517,
"learning_rate": 1.8709009319986192e-05,
"loss": 0.6699,
"step": 4270
},
{
"epoch": 3.980903586399627,
"grad_norm": 0.4169258845494642,
"learning_rate": 1.8665861235761135e-05,
"loss": 0.6817,
"step": 4275
},
{
"epoch": 3.9855612482533767,
"grad_norm": 0.42456347826037305,
"learning_rate": 1.8622713151536074e-05,
"loss": 0.6763,
"step": 4280
},
{
"epoch": 3.990218910107126,
"grad_norm": 0.4181412977181726,
"learning_rate": 1.857956506731101e-05,
"loss": 0.6579,
"step": 4285
},
{
"epoch": 3.9948765719608756,
"grad_norm": 0.41303621027879367,
"learning_rate": 1.8536416983085952e-05,
"loss": 0.6697,
"step": 4290
},
{
"epoch": 3.999534233814625,
"grad_norm": 0.4264209763127758,
"learning_rate": 1.849326889886089e-05,
"loss": 0.673,
"step": 4295
},
{
"epoch": 4.003726129483,
"grad_norm": 0.4806819361610478,
"learning_rate": 1.8450120814635834e-05,
"loss": 0.612,
"step": 4300
},
{
"epoch": 4.008383791336749,
"grad_norm": 0.4636853968999321,
"learning_rate": 1.840697273041077e-05,
"loss": 0.5965,
"step": 4305
},
{
"epoch": 4.013041453190499,
"grad_norm": 0.47840553026490606,
"learning_rate": 1.836382464618571e-05,
"loss": 0.605,
"step": 4310
},
{
"epoch": 4.017699115044247,
"grad_norm": 0.4368148682865373,
"learning_rate": 1.832067656196065e-05,
"loss": 0.5943,
"step": 4315
},
{
"epoch": 4.022356776897997,
"grad_norm": 0.4507347734680539,
"learning_rate": 1.827752847773559e-05,
"loss": 0.5999,
"step": 4320
},
{
"epoch": 4.027014438751746,
"grad_norm": 0.4473462336632657,
"learning_rate": 1.823438039351053e-05,
"loss": 0.5837,
"step": 4325
},
{
"epoch": 4.031672100605496,
"grad_norm": 0.45940433656549884,
"learning_rate": 1.819123230928547e-05,
"loss": 0.6025,
"step": 4330
},
{
"epoch": 4.036329762459245,
"grad_norm": 0.4373918323359047,
"learning_rate": 1.8148084225060408e-05,
"loss": 0.6083,
"step": 4335
},
{
"epoch": 4.040987424312995,
"grad_norm": 0.45627386533438385,
"learning_rate": 1.810493614083535e-05,
"loss": 0.6057,
"step": 4340
},
{
"epoch": 4.045645086166744,
"grad_norm": 0.4158899609247656,
"learning_rate": 1.806178805661029e-05,
"loss": 0.5976,
"step": 4345
},
{
"epoch": 4.050302748020494,
"grad_norm": 0.44858142598667394,
"learning_rate": 1.8018639972385225e-05,
"loss": 0.6072,
"step": 4350
},
{
"epoch": 4.054960409874243,
"grad_norm": 0.4319843019181772,
"learning_rate": 1.7975491888160167e-05,
"loss": 0.6168,
"step": 4355
},
{
"epoch": 4.059618071727993,
"grad_norm": 0.45788706950464525,
"learning_rate": 1.7932343803935107e-05,
"loss": 0.6216,
"step": 4360
},
{
"epoch": 4.064275733581742,
"grad_norm": 0.44601975328923293,
"learning_rate": 1.7889195719710046e-05,
"loss": 0.6102,
"step": 4365
},
{
"epoch": 4.068933395435492,
"grad_norm": 0.4264885542380311,
"learning_rate": 1.7846047635484985e-05,
"loss": 0.6101,
"step": 4370
},
{
"epoch": 4.07359105728924,
"grad_norm": 0.484392380924622,
"learning_rate": 1.7802899551259924e-05,
"loss": 0.6034,
"step": 4375
},
{
"epoch": 4.07824871914299,
"grad_norm": 0.4616445369154263,
"learning_rate": 1.7759751467034866e-05,
"loss": 0.601,
"step": 4380
},
{
"epoch": 4.082906380996739,
"grad_norm": 0.44094231922096266,
"learning_rate": 1.7716603382809806e-05,
"loss": 0.5995,
"step": 4385
},
{
"epoch": 4.087564042850489,
"grad_norm": 0.42668516125422923,
"learning_rate": 1.767345529858474e-05,
"loss": 0.6134,
"step": 4390
},
{
"epoch": 4.092221704704238,
"grad_norm": 0.43374717210175096,
"learning_rate": 1.7630307214359684e-05,
"loss": 0.6061,
"step": 4395
},
{
"epoch": 4.096879366557988,
"grad_norm": 0.46240607909102194,
"learning_rate": 1.7587159130134623e-05,
"loss": 0.6074,
"step": 4400
},
{
"epoch": 4.101537028411737,
"grad_norm": 0.4541018188451341,
"learning_rate": 1.7544011045909562e-05,
"loss": 0.6072,
"step": 4405
},
{
"epoch": 4.106194690265487,
"grad_norm": 0.458832120215879,
"learning_rate": 1.75008629616845e-05,
"loss": 0.6118,
"step": 4410
},
{
"epoch": 4.110852352119236,
"grad_norm": 0.4540320628873728,
"learning_rate": 1.745771487745944e-05,
"loss": 0.6022,
"step": 4415
},
{
"epoch": 4.115510013972986,
"grad_norm": 0.4961044448323278,
"learning_rate": 1.7414566793234383e-05,
"loss": 0.6226,
"step": 4420
},
{
"epoch": 4.120167675826735,
"grad_norm": 0.4214996889905062,
"learning_rate": 1.7371418709009322e-05,
"loss": 0.611,
"step": 4425
},
{
"epoch": 4.124825337680484,
"grad_norm": 0.4336548232531617,
"learning_rate": 1.732827062478426e-05,
"loss": 0.6198,
"step": 4430
},
{
"epoch": 4.1294829995342335,
"grad_norm": 0.46168529407914377,
"learning_rate": 1.72851225405592e-05,
"loss": 0.5939,
"step": 4435
},
{
"epoch": 4.134140661387983,
"grad_norm": 0.47290688120512087,
"learning_rate": 1.724197445633414e-05,
"loss": 0.6097,
"step": 4440
},
{
"epoch": 4.1387983232417325,
"grad_norm": 0.4755058837401803,
"learning_rate": 1.7198826372109078e-05,
"loss": 0.6138,
"step": 4445
},
{
"epoch": 4.143455985095482,
"grad_norm": 0.41589445408296166,
"learning_rate": 1.715567828788402e-05,
"loss": 0.6109,
"step": 4450
},
{
"epoch": 4.1481136469492315,
"grad_norm": 0.4498672135606939,
"learning_rate": 1.7112530203658957e-05,
"loss": 0.6083,
"step": 4455
},
{
"epoch": 4.152771308802981,
"grad_norm": 0.4766526068840867,
"learning_rate": 1.70693821194339e-05,
"loss": 0.6085,
"step": 4460
},
{
"epoch": 4.1574289706567304,
"grad_norm": 0.4295528412479077,
"learning_rate": 1.7026234035208838e-05,
"loss": 0.6225,
"step": 4465
},
{
"epoch": 4.16208663251048,
"grad_norm": 0.4298764761727921,
"learning_rate": 1.6983085950983777e-05,
"loss": 0.6221,
"step": 4470
},
{
"epoch": 4.166744294364229,
"grad_norm": 0.4591338614641095,
"learning_rate": 1.6939937866758716e-05,
"loss": 0.6085,
"step": 4475
},
{
"epoch": 4.171401956217979,
"grad_norm": 0.44174712200018607,
"learning_rate": 1.6896789782533655e-05,
"loss": 0.6066,
"step": 4480
},
{
"epoch": 4.1760596180717275,
"grad_norm": 0.41101833546266026,
"learning_rate": 1.6853641698308595e-05,
"loss": 0.5892,
"step": 4485
},
{
"epoch": 4.180717279925477,
"grad_norm": 0.4256851452409376,
"learning_rate": 1.6810493614083537e-05,
"loss": 0.5994,
"step": 4490
},
{
"epoch": 4.1853749417792265,
"grad_norm": 0.42254805258229255,
"learning_rate": 1.6767345529858476e-05,
"loss": 0.6137,
"step": 4495
},
{
"epoch": 4.190032603632976,
"grad_norm": 0.42742224570110604,
"learning_rate": 1.6724197445633415e-05,
"loss": 0.6209,
"step": 4500
},
{
"epoch": 4.1946902654867255,
"grad_norm": 0.45649771343928924,
"learning_rate": 1.6681049361408354e-05,
"loss": 0.5993,
"step": 4505
},
{
"epoch": 4.199347927340475,
"grad_norm": 0.49288033709364676,
"learning_rate": 1.6637901277183294e-05,
"loss": 0.6159,
"step": 4510
},
{
"epoch": 4.2040055891942245,
"grad_norm": 0.4503495359867892,
"learning_rate": 1.6594753192958236e-05,
"loss": 0.6108,
"step": 4515
},
{
"epoch": 4.208663251047974,
"grad_norm": 0.4808981005373163,
"learning_rate": 1.6551605108733172e-05,
"loss": 0.6112,
"step": 4520
},
{
"epoch": 4.2133209129017235,
"grad_norm": 0.42439860467905544,
"learning_rate": 1.650845702450811e-05,
"loss": 0.6039,
"step": 4525
},
{
"epoch": 4.217978574755473,
"grad_norm": 0.4993795950407368,
"learning_rate": 1.6465308940283053e-05,
"loss": 0.6071,
"step": 4530
},
{
"epoch": 4.2226362366092225,
"grad_norm": 0.43029625292079626,
"learning_rate": 1.6422160856057992e-05,
"loss": 0.6083,
"step": 4535
},
{
"epoch": 4.227293898462972,
"grad_norm": 0.4423811846646929,
"learning_rate": 1.637901277183293e-05,
"loss": 0.6143,
"step": 4540
},
{
"epoch": 4.231951560316721,
"grad_norm": 0.49407362582852016,
"learning_rate": 1.633586468760787e-05,
"loss": 0.6145,
"step": 4545
},
{
"epoch": 4.23660922217047,
"grad_norm": 0.45352381389441976,
"learning_rate": 1.629271660338281e-05,
"loss": 0.6092,
"step": 4550
},
{
"epoch": 4.24126688402422,
"grad_norm": 0.4867561002947084,
"learning_rate": 1.6249568519157752e-05,
"loss": 0.6188,
"step": 4555
},
{
"epoch": 4.245924545877969,
"grad_norm": 0.47489599095147444,
"learning_rate": 1.620642043493269e-05,
"loss": 0.6247,
"step": 4560
},
{
"epoch": 4.250582207731719,
"grad_norm": 0.4838388077146676,
"learning_rate": 1.6163272350707627e-05,
"loss": 0.6098,
"step": 4565
},
{
"epoch": 4.255239869585468,
"grad_norm": 0.49923583496165636,
"learning_rate": 1.612012426648257e-05,
"loss": 0.6239,
"step": 4570
},
{
"epoch": 4.259897531439218,
"grad_norm": 0.4203945221613941,
"learning_rate": 1.607697618225751e-05,
"loss": 0.6216,
"step": 4575
},
{
"epoch": 4.264555193292967,
"grad_norm": 0.446133559000005,
"learning_rate": 1.6033828098032448e-05,
"loss": 0.6192,
"step": 4580
},
{
"epoch": 4.269212855146717,
"grad_norm": 0.4180265005539548,
"learning_rate": 1.5990680013807387e-05,
"loss": 0.6143,
"step": 4585
},
{
"epoch": 4.273870517000466,
"grad_norm": 0.4525922240195107,
"learning_rate": 1.5947531929582326e-05,
"loss": 0.6081,
"step": 4590
},
{
"epoch": 4.2785281788542155,
"grad_norm": 0.45234994843255383,
"learning_rate": 1.590438384535727e-05,
"loss": 0.6178,
"step": 4595
},
{
"epoch": 4.283185840707965,
"grad_norm": 0.4494375930375664,
"learning_rate": 1.5861235761132208e-05,
"loss": 0.6198,
"step": 4600
},
{
"epoch": 4.287843502561714,
"grad_norm": 0.4225621793308862,
"learning_rate": 1.5818087676907143e-05,
"loss": 0.6271,
"step": 4605
},
{
"epoch": 4.292501164415463,
"grad_norm": 0.46236415865283903,
"learning_rate": 1.5774939592682086e-05,
"loss": 0.6107,
"step": 4610
},
{
"epoch": 4.297158826269213,
"grad_norm": 0.46052115118679277,
"learning_rate": 1.5731791508457025e-05,
"loss": 0.6102,
"step": 4615
},
{
"epoch": 4.301816488122962,
"grad_norm": 0.46117671507013813,
"learning_rate": 1.5688643424231964e-05,
"loss": 0.6049,
"step": 4620
},
{
"epoch": 4.306474149976712,
"grad_norm": 0.4407496966658314,
"learning_rate": 1.5645495340006903e-05,
"loss": 0.6126,
"step": 4625
},
{
"epoch": 4.311131811830461,
"grad_norm": 0.4457616926070442,
"learning_rate": 1.5602347255781842e-05,
"loss": 0.619,
"step": 4630
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.45136738284407385,
"learning_rate": 1.5559199171556785e-05,
"loss": 0.6102,
"step": 4635
},
{
"epoch": 4.32044713553796,
"grad_norm": 0.4380548820490039,
"learning_rate": 1.5516051087331724e-05,
"loss": 0.6095,
"step": 4640
},
{
"epoch": 4.32510479739171,
"grad_norm": 0.44183287095699714,
"learning_rate": 1.5472903003106663e-05,
"loss": 0.6067,
"step": 4645
},
{
"epoch": 4.329762459245459,
"grad_norm": 0.43123650201876623,
"learning_rate": 1.5429754918881602e-05,
"loss": 0.6329,
"step": 4650
},
{
"epoch": 4.334420121099209,
"grad_norm": 0.4177619149141171,
"learning_rate": 1.538660683465654e-05,
"loss": 0.6097,
"step": 4655
},
{
"epoch": 4.339077782952957,
"grad_norm": 0.4352353479215684,
"learning_rate": 1.534345875043148e-05,
"loss": 0.6161,
"step": 4660
},
{
"epoch": 4.343735444806707,
"grad_norm": 0.4605489056353871,
"learning_rate": 1.5300310666206423e-05,
"loss": 0.6272,
"step": 4665
},
{
"epoch": 4.348393106660456,
"grad_norm": 0.43609413400101865,
"learning_rate": 1.525716258198136e-05,
"loss": 0.6194,
"step": 4670
},
{
"epoch": 4.353050768514206,
"grad_norm": 0.43918268795029974,
"learning_rate": 1.5214014497756301e-05,
"loss": 0.6012,
"step": 4675
},
{
"epoch": 4.357708430367955,
"grad_norm": 0.4554666556635843,
"learning_rate": 1.517086641353124e-05,
"loss": 0.6248,
"step": 4680
},
{
"epoch": 4.362366092221705,
"grad_norm": 0.44745527897814913,
"learning_rate": 1.5127718329306178e-05,
"loss": 0.6052,
"step": 4685
},
{
"epoch": 4.367023754075454,
"grad_norm": 0.4566195590924305,
"learning_rate": 1.508457024508112e-05,
"loss": 0.6147,
"step": 4690
},
{
"epoch": 4.371681415929204,
"grad_norm": 0.4295277655025628,
"learning_rate": 1.5041422160856058e-05,
"loss": 0.6056,
"step": 4695
},
{
"epoch": 4.376339077782953,
"grad_norm": 0.4762038397304181,
"learning_rate": 1.4998274076630997e-05,
"loss": 0.6267,
"step": 4700
},
{
"epoch": 4.380996739636703,
"grad_norm": 0.43161468066172953,
"learning_rate": 1.4955125992405938e-05,
"loss": 0.6093,
"step": 4705
},
{
"epoch": 4.385654401490452,
"grad_norm": 0.4537790846607464,
"learning_rate": 1.4911977908180877e-05,
"loss": 0.627,
"step": 4710
},
{
"epoch": 4.390312063344201,
"grad_norm": 0.47628975099139476,
"learning_rate": 1.4868829823955818e-05,
"loss": 0.599,
"step": 4715
},
{
"epoch": 4.39496972519795,
"grad_norm": 0.4284210913770137,
"learning_rate": 1.4825681739730757e-05,
"loss": 0.6047,
"step": 4720
},
{
"epoch": 4.3996273870517,
"grad_norm": 0.4623777997484651,
"learning_rate": 1.4782533655505696e-05,
"loss": 0.5996,
"step": 4725
},
{
"epoch": 4.404285048905449,
"grad_norm": 0.44782456058740255,
"learning_rate": 1.4739385571280637e-05,
"loss": 0.6242,
"step": 4730
},
{
"epoch": 4.408942710759199,
"grad_norm": 0.42865704978519364,
"learning_rate": 1.4696237487055576e-05,
"loss": 0.6216,
"step": 4735
},
{
"epoch": 4.413600372612948,
"grad_norm": 0.42258284119468553,
"learning_rate": 1.4653089402830513e-05,
"loss": 0.5949,
"step": 4740
},
{
"epoch": 4.418258034466698,
"grad_norm": 0.4404360437172725,
"learning_rate": 1.4609941318605456e-05,
"loss": 0.6142,
"step": 4745
},
{
"epoch": 4.422915696320447,
"grad_norm": 0.4471745958531887,
"learning_rate": 1.4566793234380393e-05,
"loss": 0.6046,
"step": 4750
},
{
"epoch": 4.427573358174197,
"grad_norm": 0.45193754607906805,
"learning_rate": 1.4523645150155336e-05,
"loss": 0.6301,
"step": 4755
},
{
"epoch": 4.432231020027946,
"grad_norm": 0.4378890697332401,
"learning_rate": 1.4480497065930273e-05,
"loss": 0.613,
"step": 4760
},
{
"epoch": 4.436888681881696,
"grad_norm": 0.40172229970174256,
"learning_rate": 1.4437348981705212e-05,
"loss": 0.6126,
"step": 4765
},
{
"epoch": 4.441546343735445,
"grad_norm": 0.44671360931171006,
"learning_rate": 1.4394200897480153e-05,
"loss": 0.5993,
"step": 4770
},
{
"epoch": 4.446204005589194,
"grad_norm": 0.4208530812149307,
"learning_rate": 1.4351052813255092e-05,
"loss": 0.6117,
"step": 4775
},
{
"epoch": 4.450861667442943,
"grad_norm": 0.4566343802538676,
"learning_rate": 1.4307904729030031e-05,
"loss": 0.6165,
"step": 4780
},
{
"epoch": 4.455519329296693,
"grad_norm": 0.43920444604920234,
"learning_rate": 1.4264756644804972e-05,
"loss": 0.6244,
"step": 4785
},
{
"epoch": 4.460176991150442,
"grad_norm": 0.4410914601766885,
"learning_rate": 1.4221608560579911e-05,
"loss": 0.6096,
"step": 4790
},
{
"epoch": 4.464834653004192,
"grad_norm": 0.47815149847379673,
"learning_rate": 1.4178460476354852e-05,
"loss": 0.6144,
"step": 4795
},
{
"epoch": 4.469492314857941,
"grad_norm": 0.4393143547044593,
"learning_rate": 1.4135312392129791e-05,
"loss": 0.6212,
"step": 4800
},
{
"epoch": 4.474149976711691,
"grad_norm": 0.4326176715876059,
"learning_rate": 1.4092164307904728e-05,
"loss": 0.6182,
"step": 4805
},
{
"epoch": 4.47880763856544,
"grad_norm": 0.45343036326683683,
"learning_rate": 1.404901622367967e-05,
"loss": 0.6116,
"step": 4810
},
{
"epoch": 4.48346530041919,
"grad_norm": 0.41682266432299153,
"learning_rate": 1.4005868139454608e-05,
"loss": 0.6026,
"step": 4815
},
{
"epoch": 4.488122962272939,
"grad_norm": 0.4493369452130179,
"learning_rate": 1.3962720055229547e-05,
"loss": 0.6237,
"step": 4820
},
{
"epoch": 4.492780624126689,
"grad_norm": 0.4337623071265652,
"learning_rate": 1.3919571971004488e-05,
"loss": 0.6161,
"step": 4825
},
{
"epoch": 4.497438285980438,
"grad_norm": 0.43658371794237966,
"learning_rate": 1.3876423886779427e-05,
"loss": 0.6259,
"step": 4830
},
{
"epoch": 4.502095947834187,
"grad_norm": 0.43087708114430684,
"learning_rate": 1.3833275802554368e-05,
"loss": 0.615,
"step": 4835
},
{
"epoch": 4.506753609687936,
"grad_norm": 0.4308877230714715,
"learning_rate": 1.3790127718329307e-05,
"loss": 0.6178,
"step": 4840
},
{
"epoch": 4.511411271541686,
"grad_norm": 0.4345512713590378,
"learning_rate": 1.3746979634104246e-05,
"loss": 0.6135,
"step": 4845
},
{
"epoch": 4.516068933395435,
"grad_norm": 0.42499223546672094,
"learning_rate": 1.3703831549879187e-05,
"loss": 0.6011,
"step": 4850
},
{
"epoch": 4.520726595249185,
"grad_norm": 0.4297293994576531,
"learning_rate": 1.3660683465654126e-05,
"loss": 0.6155,
"step": 4855
},
{
"epoch": 4.525384257102934,
"grad_norm": 0.44676211626706,
"learning_rate": 1.3617535381429064e-05,
"loss": 0.6026,
"step": 4860
},
{
"epoch": 4.530041918956684,
"grad_norm": 0.41451280302413623,
"learning_rate": 1.3574387297204006e-05,
"loss": 0.6135,
"step": 4865
},
{
"epoch": 4.534699580810433,
"grad_norm": 0.41024444397996657,
"learning_rate": 1.3531239212978944e-05,
"loss": 0.6172,
"step": 4870
},
{
"epoch": 4.539357242664183,
"grad_norm": 0.4249317938403838,
"learning_rate": 1.3488091128753886e-05,
"loss": 0.6206,
"step": 4875
},
{
"epoch": 4.544014904517932,
"grad_norm": 0.4288584553452124,
"learning_rate": 1.3444943044528824e-05,
"loss": 0.6187,
"step": 4880
},
{
"epoch": 4.548672566371682,
"grad_norm": 0.42815344099306313,
"learning_rate": 1.3401794960303763e-05,
"loss": 0.6162,
"step": 4885
},
{
"epoch": 4.55333022822543,
"grad_norm": 0.4476784897625909,
"learning_rate": 1.3358646876078703e-05,
"loss": 0.612,
"step": 4890
},
{
"epoch": 4.55798789007918,
"grad_norm": 0.4146360781064623,
"learning_rate": 1.3315498791853643e-05,
"loss": 0.608,
"step": 4895
},
{
"epoch": 4.562645551932929,
"grad_norm": 0.42180653622806297,
"learning_rate": 1.327235070762858e-05,
"loss": 0.5984,
"step": 4900
},
{
"epoch": 4.567303213786679,
"grad_norm": 0.42786143566290896,
"learning_rate": 1.3229202623403522e-05,
"loss": 0.616,
"step": 4905
},
{
"epoch": 4.571960875640428,
"grad_norm": 0.4337556272151318,
"learning_rate": 1.318605453917846e-05,
"loss": 0.6104,
"step": 4910
},
{
"epoch": 4.576618537494178,
"grad_norm": 0.43813645125024153,
"learning_rate": 1.3142906454953402e-05,
"loss": 0.6059,
"step": 4915
},
{
"epoch": 4.581276199347927,
"grad_norm": 0.44397103218263223,
"learning_rate": 1.309975837072834e-05,
"loss": 0.6067,
"step": 4920
},
{
"epoch": 4.585933861201677,
"grad_norm": 0.4206021176834172,
"learning_rate": 1.3056610286503279e-05,
"loss": 0.6188,
"step": 4925
},
{
"epoch": 4.590591523055426,
"grad_norm": 0.4155111249198103,
"learning_rate": 1.301346220227822e-05,
"loss": 0.6044,
"step": 4930
},
{
"epoch": 4.595249184909176,
"grad_norm": 0.4077942507751532,
"learning_rate": 1.2970314118053159e-05,
"loss": 0.6195,
"step": 4935
},
{
"epoch": 4.599906846762925,
"grad_norm": 0.40937985817451483,
"learning_rate": 1.2927166033828098e-05,
"loss": 0.6046,
"step": 4940
},
{
"epoch": 4.604564508616674,
"grad_norm": 0.42556174213849096,
"learning_rate": 1.2884017949603039e-05,
"loss": 0.6312,
"step": 4945
},
{
"epoch": 4.6092221704704235,
"grad_norm": 0.42998849734417294,
"learning_rate": 1.2840869865377978e-05,
"loss": 0.6057,
"step": 4950
},
{
"epoch": 4.613879832324173,
"grad_norm": 0.41004193899090896,
"learning_rate": 1.2797721781152919e-05,
"loss": 0.6198,
"step": 4955
},
{
"epoch": 4.6185374941779225,
"grad_norm": 0.42096090582338275,
"learning_rate": 1.2754573696927858e-05,
"loss": 0.6039,
"step": 4960
},
{
"epoch": 4.623195156031672,
"grad_norm": 0.42400995011780857,
"learning_rate": 1.2711425612702795e-05,
"loss": 0.6266,
"step": 4965
},
{
"epoch": 4.6278528178854215,
"grad_norm": 0.42040078255527125,
"learning_rate": 1.2668277528477738e-05,
"loss": 0.6228,
"step": 4970
},
{
"epoch": 4.632510479739171,
"grad_norm": 0.4431900153893484,
"learning_rate": 1.2625129444252675e-05,
"loss": 0.5979,
"step": 4975
},
{
"epoch": 4.6371681415929205,
"grad_norm": 0.4212544728790483,
"learning_rate": 1.2581981360027614e-05,
"loss": 0.603,
"step": 4980
},
{
"epoch": 4.64182580344667,
"grad_norm": 0.437403098140137,
"learning_rate": 1.2538833275802555e-05,
"loss": 0.6282,
"step": 4985
},
{
"epoch": 4.6464834653004194,
"grad_norm": 0.42245841231912035,
"learning_rate": 1.2495685191577494e-05,
"loss": 0.6025,
"step": 4990
},
{
"epoch": 4.651141127154169,
"grad_norm": 0.4195991173380262,
"learning_rate": 1.2452537107352433e-05,
"loss": 0.6155,
"step": 4995
},
{
"epoch": 4.6557987890079175,
"grad_norm": 0.42896342530484133,
"learning_rate": 1.2409389023127374e-05,
"loss": 0.6172,
"step": 5000
},
{
"epoch": 4.660456450861668,
"grad_norm": 0.4519927520866544,
"learning_rate": 1.2366240938902313e-05,
"loss": 0.6151,
"step": 5005
},
{
"epoch": 4.6651141127154165,
"grad_norm": 0.4154772218254218,
"learning_rate": 1.2323092854677252e-05,
"loss": 0.6231,
"step": 5010
},
{
"epoch": 4.669771774569166,
"grad_norm": 0.41189543902219217,
"learning_rate": 1.2279944770452193e-05,
"loss": 0.6129,
"step": 5015
},
{
"epoch": 4.6744294364229155,
"grad_norm": 0.41834462910543274,
"learning_rate": 1.2236796686227132e-05,
"loss": 0.6144,
"step": 5020
},
{
"epoch": 4.679087098276665,
"grad_norm": 0.4276803315518868,
"learning_rate": 1.2193648602002073e-05,
"loss": 0.6193,
"step": 5025
},
{
"epoch": 4.6837447601304145,
"grad_norm": 0.4471592285620113,
"learning_rate": 1.215050051777701e-05,
"loss": 0.6115,
"step": 5030
},
{
"epoch": 4.688402421984164,
"grad_norm": 0.4280179038134522,
"learning_rate": 1.2107352433551951e-05,
"loss": 0.6118,
"step": 5035
},
{
"epoch": 4.6930600838379135,
"grad_norm": 0.4156278571195452,
"learning_rate": 1.206420434932689e-05,
"loss": 0.6085,
"step": 5040
},
{
"epoch": 4.697717745691663,
"grad_norm": 0.4389936054467469,
"learning_rate": 1.202105626510183e-05,
"loss": 0.6016,
"step": 5045
},
{
"epoch": 4.7023754075454125,
"grad_norm": 0.4398668606342934,
"learning_rate": 1.1977908180876769e-05,
"loss": 0.6053,
"step": 5050
},
{
"epoch": 4.707033069399162,
"grad_norm": 0.4152682167746281,
"learning_rate": 1.193476009665171e-05,
"loss": 0.6069,
"step": 5055
},
{
"epoch": 4.7116907312529115,
"grad_norm": 0.42222398978193676,
"learning_rate": 1.1891612012426649e-05,
"loss": 0.6076,
"step": 5060
},
{
"epoch": 4.71634839310666,
"grad_norm": 0.43524199728540414,
"learning_rate": 1.1848463928201588e-05,
"loss": 0.6008,
"step": 5065
},
{
"epoch": 4.72100605496041,
"grad_norm": 0.4110464798198264,
"learning_rate": 1.1805315843976528e-05,
"loss": 0.6073,
"step": 5070
},
{
"epoch": 4.725663716814159,
"grad_norm": 0.43391534282606625,
"learning_rate": 1.1762167759751468e-05,
"loss": 0.6187,
"step": 5075
},
{
"epoch": 4.730321378667909,
"grad_norm": 0.4344268574647356,
"learning_rate": 1.1719019675526408e-05,
"loss": 0.6183,
"step": 5080
},
{
"epoch": 4.734979040521658,
"grad_norm": 0.46079886577181006,
"learning_rate": 1.1675871591301346e-05,
"loss": 0.6032,
"step": 5085
},
{
"epoch": 4.739636702375408,
"grad_norm": 0.4095438317917713,
"learning_rate": 1.1632723507076287e-05,
"loss": 0.6108,
"step": 5090
},
{
"epoch": 4.744294364229157,
"grad_norm": 0.4192945945524116,
"learning_rate": 1.1589575422851226e-05,
"loss": 0.6014,
"step": 5095
},
{
"epoch": 4.748952026082907,
"grad_norm": 0.425187657232121,
"learning_rate": 1.1546427338626167e-05,
"loss": 0.6149,
"step": 5100
},
{
"epoch": 4.753609687936656,
"grad_norm": 0.4314640586451845,
"learning_rate": 1.1503279254401104e-05,
"loss": 0.617,
"step": 5105
},
{
"epoch": 4.7582673497904056,
"grad_norm": 0.4226989429033583,
"learning_rate": 1.1460131170176045e-05,
"loss": 0.6161,
"step": 5110
},
{
"epoch": 4.762925011644155,
"grad_norm": 0.4609333444144553,
"learning_rate": 1.1416983085950984e-05,
"loss": 0.6114,
"step": 5115
},
{
"epoch": 4.767582673497904,
"grad_norm": 0.41768650972615995,
"learning_rate": 1.1373835001725925e-05,
"loss": 0.6002,
"step": 5120
},
{
"epoch": 4.772240335351653,
"grad_norm": 0.43194965243676026,
"learning_rate": 1.1330686917500862e-05,
"loss": 0.6113,
"step": 5125
},
{
"epoch": 4.776897997205403,
"grad_norm": 0.4095581873851778,
"learning_rate": 1.1287538833275803e-05,
"loss": 0.6164,
"step": 5130
},
{
"epoch": 4.781555659059152,
"grad_norm": 0.4313289478900628,
"learning_rate": 1.1244390749050742e-05,
"loss": 0.634,
"step": 5135
},
{
"epoch": 4.786213320912902,
"grad_norm": 0.4127155972701948,
"learning_rate": 1.1201242664825683e-05,
"loss": 0.6017,
"step": 5140
},
{
"epoch": 4.790870982766651,
"grad_norm": 0.39603431972121866,
"learning_rate": 1.1158094580600622e-05,
"loss": 0.6125,
"step": 5145
},
{
"epoch": 4.795528644620401,
"grad_norm": 0.4342284527499743,
"learning_rate": 1.1114946496375561e-05,
"loss": 0.5879,
"step": 5150
},
{
"epoch": 4.80018630647415,
"grad_norm": 0.41941974920430475,
"learning_rate": 1.1071798412150502e-05,
"loss": 0.5973,
"step": 5155
},
{
"epoch": 4.8048439683279,
"grad_norm": 0.47802974385560876,
"learning_rate": 1.1028650327925441e-05,
"loss": 0.5995,
"step": 5160
},
{
"epoch": 4.809501630181649,
"grad_norm": 0.3965798387548902,
"learning_rate": 1.098550224370038e-05,
"loss": 0.6184,
"step": 5165
},
{
"epoch": 4.814159292035399,
"grad_norm": 0.43801160852058696,
"learning_rate": 1.094235415947532e-05,
"loss": 0.6203,
"step": 5170
},
{
"epoch": 4.818816953889147,
"grad_norm": 0.4241260906344107,
"learning_rate": 1.089920607525026e-05,
"loss": 0.6272,
"step": 5175
},
{
"epoch": 4.823474615742897,
"grad_norm": 0.4105124110466843,
"learning_rate": 1.0856057991025199e-05,
"loss": 0.6052,
"step": 5180
},
{
"epoch": 4.828132277596646,
"grad_norm": 0.425415723767141,
"learning_rate": 1.0812909906800138e-05,
"loss": 0.6038,
"step": 5185
},
{
"epoch": 4.832789939450396,
"grad_norm": 0.47310761948520047,
"learning_rate": 1.0769761822575077e-05,
"loss": 0.629,
"step": 5190
},
{
"epoch": 4.837447601304145,
"grad_norm": 0.4083624743588117,
"learning_rate": 1.0726613738350018e-05,
"loss": 0.6096,
"step": 5195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.3993868337304555,
"learning_rate": 1.0683465654124957e-05,
"loss": 0.6014,
"step": 5200
},
{
"epoch": 4.846762925011644,
"grad_norm": 0.4206135006510575,
"learning_rate": 1.0640317569899896e-05,
"loss": 0.6206,
"step": 5205
},
{
"epoch": 4.851420586865394,
"grad_norm": 0.43557861630634326,
"learning_rate": 1.0597169485674835e-05,
"loss": 0.6261,
"step": 5210
},
{
"epoch": 4.856078248719143,
"grad_norm": 0.4317307534235413,
"learning_rate": 1.0554021401449776e-05,
"loss": 0.6129,
"step": 5215
},
{
"epoch": 4.860735910572893,
"grad_norm": 0.41079339176887714,
"learning_rate": 1.0510873317224715e-05,
"loss": 0.6159,
"step": 5220
},
{
"epoch": 4.865393572426642,
"grad_norm": 0.41800514370168124,
"learning_rate": 1.0467725232999655e-05,
"loss": 0.6007,
"step": 5225
},
{
"epoch": 4.870051234280391,
"grad_norm": 0.42937334307178815,
"learning_rate": 1.0424577148774595e-05,
"loss": 0.6134,
"step": 5230
},
{
"epoch": 4.874708896134141,
"grad_norm": 0.45733900734235955,
"learning_rate": 1.0381429064549534e-05,
"loss": 0.6238,
"step": 5235
},
{
"epoch": 4.87936655798789,
"grad_norm": 0.4418166298473757,
"learning_rate": 1.0338280980324475e-05,
"loss": 0.6091,
"step": 5240
},
{
"epoch": 4.884024219841639,
"grad_norm": 0.4176075023708318,
"learning_rate": 1.0295132896099413e-05,
"loss": 0.6327,
"step": 5245
},
{
"epoch": 4.888681881695389,
"grad_norm": 0.45132383481251026,
"learning_rate": 1.0251984811874353e-05,
"loss": 0.6271,
"step": 5250
},
{
"epoch": 4.893339543549138,
"grad_norm": 0.4093678308993761,
"learning_rate": 1.0208836727649293e-05,
"loss": 0.6118,
"step": 5255
},
{
"epoch": 4.897997205402888,
"grad_norm": 0.4358137402583084,
"learning_rate": 1.0165688643424233e-05,
"loss": 0.5982,
"step": 5260
},
{
"epoch": 4.902654867256637,
"grad_norm": 0.40661293305643104,
"learning_rate": 1.012254055919917e-05,
"loss": 0.6342,
"step": 5265
},
{
"epoch": 4.907312529110387,
"grad_norm": 0.42806269456221924,
"learning_rate": 1.0079392474974112e-05,
"loss": 0.6172,
"step": 5270
},
{
"epoch": 4.911970190964136,
"grad_norm": 0.41622182265496577,
"learning_rate": 1.003624439074905e-05,
"loss": 0.6149,
"step": 5275
},
{
"epoch": 4.916627852817886,
"grad_norm": 0.4183378435998186,
"learning_rate": 9.993096306523992e-06,
"loss": 0.6153,
"step": 5280
},
{
"epoch": 4.921285514671635,
"grad_norm": 0.422784188546255,
"learning_rate": 9.94994822229893e-06,
"loss": 0.6084,
"step": 5285
},
{
"epoch": 4.925943176525385,
"grad_norm": 0.428353990517149,
"learning_rate": 9.90680013807387e-06,
"loss": 0.6241,
"step": 5290
},
{
"epoch": 4.930600838379133,
"grad_norm": 0.41604173789317267,
"learning_rate": 9.86365205384881e-06,
"loss": 0.605,
"step": 5295
},
{
"epoch": 4.935258500232883,
"grad_norm": 0.39758808242607707,
"learning_rate": 9.82050396962375e-06,
"loss": 0.6116,
"step": 5300
},
{
"epoch": 4.939916162086632,
"grad_norm": 0.4253286290481327,
"learning_rate": 9.777355885398689e-06,
"loss": 0.6197,
"step": 5305
},
{
"epoch": 4.944573823940382,
"grad_norm": 0.42181425213043905,
"learning_rate": 9.734207801173628e-06,
"loss": 0.6117,
"step": 5310
},
{
"epoch": 4.949231485794131,
"grad_norm": 0.4134798514571126,
"learning_rate": 9.691059716948569e-06,
"loss": 0.5979,
"step": 5315
},
{
"epoch": 4.953889147647881,
"grad_norm": 0.4098416893467789,
"learning_rate": 9.647911632723508e-06,
"loss": 0.6119,
"step": 5320
},
{
"epoch": 4.95854680950163,
"grad_norm": 0.45216033261906713,
"learning_rate": 9.604763548498447e-06,
"loss": 0.6189,
"step": 5325
},
{
"epoch": 4.96320447135538,
"grad_norm": 0.44225564927802763,
"learning_rate": 9.561615464273386e-06,
"loss": 0.6181,
"step": 5330
},
{
"epoch": 4.967862133209129,
"grad_norm": 0.4078319297335572,
"learning_rate": 9.518467380048327e-06,
"loss": 0.5954,
"step": 5335
},
{
"epoch": 4.972519795062879,
"grad_norm": 0.4118776896593466,
"learning_rate": 9.475319295823266e-06,
"loss": 0.6138,
"step": 5340
},
{
"epoch": 4.977177456916628,
"grad_norm": 0.4211790875068119,
"learning_rate": 9.432171211598205e-06,
"loss": 0.6049,
"step": 5345
},
{
"epoch": 4.981835118770377,
"grad_norm": 0.40842310789076874,
"learning_rate": 9.389023127373144e-06,
"loss": 0.5949,
"step": 5350
},
{
"epoch": 4.986492780624126,
"grad_norm": 0.40580311588149054,
"learning_rate": 9.345875043148085e-06,
"loss": 0.6134,
"step": 5355
},
{
"epoch": 4.991150442477876,
"grad_norm": 0.4323753404756848,
"learning_rate": 9.302726958923024e-06,
"loss": 0.5883,
"step": 5360
},
{
"epoch": 4.995808104331625,
"grad_norm": 0.4064246433001648,
"learning_rate": 9.259578874697963e-06,
"loss": 0.6183,
"step": 5365
},
{
"epoch": 5.0,
"grad_norm": 0.4189595295526227,
"learning_rate": 9.216430790472904e-06,
"loss": 0.5964,
"step": 5370
},
{
"epoch": 5.0046576618537495,
"grad_norm": 0.45660893906293887,
"learning_rate": 9.173282706247843e-06,
"loss": 0.5727,
"step": 5375
},
{
"epoch": 5.009315323707499,
"grad_norm": 0.48950393937721143,
"learning_rate": 9.130134622022784e-06,
"loss": 0.5716,
"step": 5380
},
{
"epoch": 5.0139729855612485,
"grad_norm": 0.46305435866792793,
"learning_rate": 9.086986537797721e-06,
"loss": 0.5652,
"step": 5385
},
{
"epoch": 5.018630647414998,
"grad_norm": 0.46351578346695144,
"learning_rate": 9.043838453572662e-06,
"loss": 0.5831,
"step": 5390
},
{
"epoch": 5.0232883092687475,
"grad_norm": 0.43383882806850615,
"learning_rate": 9.000690369347601e-06,
"loss": 0.5556,
"step": 5395
},
{
"epoch": 5.027945971122496,
"grad_norm": 0.4283268226889008,
"learning_rate": 8.957542285122542e-06,
"loss": 0.5523,
"step": 5400
},
{
"epoch": 5.032603632976246,
"grad_norm": 0.429645710696869,
"learning_rate": 8.91439420089748e-06,
"loss": 0.5574,
"step": 5405
},
{
"epoch": 5.037261294829995,
"grad_norm": 0.4094166465952819,
"learning_rate": 8.87124611667242e-06,
"loss": 0.5597,
"step": 5410
},
{
"epoch": 5.041918956683745,
"grad_norm": 0.4712850595020832,
"learning_rate": 8.82809803244736e-06,
"loss": 0.5609,
"step": 5415
},
{
"epoch": 5.046576618537494,
"grad_norm": 0.4246892863915132,
"learning_rate": 8.7849499482223e-06,
"loss": 0.5457,
"step": 5420
},
{
"epoch": 5.051234280391244,
"grad_norm": 0.4187158662633382,
"learning_rate": 8.741801863997238e-06,
"loss": 0.5606,
"step": 5425
},
{
"epoch": 5.055891942244993,
"grad_norm": 0.41213110304851663,
"learning_rate": 8.698653779772179e-06,
"loss": 0.5716,
"step": 5430
},
{
"epoch": 5.0605496040987425,
"grad_norm": 0.4153843626149877,
"learning_rate": 8.655505695547118e-06,
"loss": 0.5389,
"step": 5435
},
{
"epoch": 5.065207265952492,
"grad_norm": 0.40899362039500287,
"learning_rate": 8.612357611322058e-06,
"loss": 0.5449,
"step": 5440
},
{
"epoch": 5.0698649278062415,
"grad_norm": 0.4246192080492623,
"learning_rate": 8.569209527096998e-06,
"loss": 0.5508,
"step": 5445
},
{
"epoch": 5.074522589659991,
"grad_norm": 0.40779228236908316,
"learning_rate": 8.526061442871937e-06,
"loss": 0.5756,
"step": 5450
},
{
"epoch": 5.0791802515137405,
"grad_norm": 0.45516682878989534,
"learning_rate": 8.482913358646877e-06,
"loss": 0.5734,
"step": 5455
},
{
"epoch": 5.083837913367489,
"grad_norm": 0.3998404619097546,
"learning_rate": 8.439765274421817e-06,
"loss": 0.5394,
"step": 5460
},
{
"epoch": 5.088495575221239,
"grad_norm": 0.4340401362574142,
"learning_rate": 8.396617190196756e-06,
"loss": 0.5634,
"step": 5465
},
{
"epoch": 5.093153237074988,
"grad_norm": 0.44887014864072095,
"learning_rate": 8.353469105971695e-06,
"loss": 0.5702,
"step": 5470
},
{
"epoch": 5.097810898928738,
"grad_norm": 0.42103357271725655,
"learning_rate": 8.310321021746636e-06,
"loss": 0.569,
"step": 5475
},
{
"epoch": 5.102468560782487,
"grad_norm": 0.4425626777811027,
"learning_rate": 8.267172937521575e-06,
"loss": 0.5801,
"step": 5480
},
{
"epoch": 5.107126222636237,
"grad_norm": 0.43103556729424025,
"learning_rate": 8.224024853296514e-06,
"loss": 0.562,
"step": 5485
},
{
"epoch": 5.111783884489986,
"grad_norm": 0.4288766363689286,
"learning_rate": 8.180876769071453e-06,
"loss": 0.5549,
"step": 5490
},
{
"epoch": 5.116441546343736,
"grad_norm": 0.4411211103107296,
"learning_rate": 8.137728684846394e-06,
"loss": 0.5782,
"step": 5495
},
{
"epoch": 5.121099208197485,
"grad_norm": 0.44590845105225113,
"learning_rate": 8.094580600621333e-06,
"loss": 0.5677,
"step": 5500
},
{
"epoch": 5.125756870051235,
"grad_norm": 0.42044193491024917,
"learning_rate": 8.051432516396272e-06,
"loss": 0.5647,
"step": 5505
},
{
"epoch": 5.130414531904984,
"grad_norm": 0.4220746510226187,
"learning_rate": 8.008284432171211e-06,
"loss": 0.5794,
"step": 5510
},
{
"epoch": 5.135072193758733,
"grad_norm": 0.41475894763305793,
"learning_rate": 7.965136347946152e-06,
"loss": 0.5517,
"step": 5515
},
{
"epoch": 5.139729855612482,
"grad_norm": 0.42827780055854336,
"learning_rate": 7.921988263721091e-06,
"loss": 0.5737,
"step": 5520
},
{
"epoch": 5.144387517466232,
"grad_norm": 0.41403286591032884,
"learning_rate": 7.87884017949603e-06,
"loss": 0.563,
"step": 5525
},
{
"epoch": 5.149045179319981,
"grad_norm": 0.42984527591895905,
"learning_rate": 7.835692095270971e-06,
"loss": 0.5561,
"step": 5530
},
{
"epoch": 5.153702841173731,
"grad_norm": 0.42945380057609706,
"learning_rate": 7.79254401104591e-06,
"loss": 0.5666,
"step": 5535
},
{
"epoch": 5.15836050302748,
"grad_norm": 0.4432460499019077,
"learning_rate": 7.749395926820851e-06,
"loss": 0.5615,
"step": 5540
},
{
"epoch": 5.16301816488123,
"grad_norm": 0.42159801759717264,
"learning_rate": 7.706247842595788e-06,
"loss": 0.5604,
"step": 5545
},
{
"epoch": 5.167675826734979,
"grad_norm": 0.4375171715557641,
"learning_rate": 7.663099758370729e-06,
"loss": 0.5708,
"step": 5550
},
{
"epoch": 5.172333488588729,
"grad_norm": 0.42874893445419604,
"learning_rate": 7.619951674145669e-06,
"loss": 0.5676,
"step": 5555
},
{
"epoch": 5.176991150442478,
"grad_norm": 0.42541474226000026,
"learning_rate": 7.576803589920608e-06,
"loss": 0.5902,
"step": 5560
},
{
"epoch": 5.181648812296228,
"grad_norm": 0.4302044467345254,
"learning_rate": 7.533655505695547e-06,
"loss": 0.5623,
"step": 5565
},
{
"epoch": 5.186306474149977,
"grad_norm": 0.4124995234442213,
"learning_rate": 7.490507421470487e-06,
"loss": 0.5612,
"step": 5570
},
{
"epoch": 5.190964136003726,
"grad_norm": 0.40565576085511523,
"learning_rate": 7.447359337245427e-06,
"loss": 0.5681,
"step": 5575
},
{
"epoch": 5.195621797857475,
"grad_norm": 0.41713465634842206,
"learning_rate": 7.4042112530203655e-06,
"loss": 0.5571,
"step": 5580
},
{
"epoch": 5.200279459711225,
"grad_norm": 0.41967547106733516,
"learning_rate": 7.361063168795305e-06,
"loss": 0.5639,
"step": 5585
},
{
"epoch": 5.204937121564974,
"grad_norm": 0.4506892995367639,
"learning_rate": 7.317915084570245e-06,
"loss": 0.5677,
"step": 5590
},
{
"epoch": 5.209594783418724,
"grad_norm": 0.4258131849898496,
"learning_rate": 7.274767000345185e-06,
"loss": 0.5683,
"step": 5595
},
{
"epoch": 5.214252445272473,
"grad_norm": 0.4204658172936071,
"learning_rate": 7.231618916120124e-06,
"loss": 0.5745,
"step": 5600
},
{
"epoch": 5.218910107126223,
"grad_norm": 0.4336757745199986,
"learning_rate": 7.1884708318950636e-06,
"loss": 0.5739,
"step": 5605
},
{
"epoch": 5.223567768979972,
"grad_norm": 0.42950896150032264,
"learning_rate": 7.1453227476700035e-06,
"loss": 0.5782,
"step": 5610
},
{
"epoch": 5.228225430833722,
"grad_norm": 0.4374368910234369,
"learning_rate": 7.1021746634449435e-06,
"loss": 0.5584,
"step": 5615
},
{
"epoch": 5.232883092687471,
"grad_norm": 0.42566321904544496,
"learning_rate": 7.059026579219883e-06,
"loss": 0.5612,
"step": 5620
},
{
"epoch": 5.237540754541221,
"grad_norm": 0.42008279382378816,
"learning_rate": 7.0158784949948226e-06,
"loss": 0.5476,
"step": 5625
},
{
"epoch": 5.242198416394969,
"grad_norm": 0.449770356710987,
"learning_rate": 6.9727304107697625e-06,
"loss": 0.5578,
"step": 5630
},
{
"epoch": 5.246856078248719,
"grad_norm": 0.4445174977744607,
"learning_rate": 6.9295823265447025e-06,
"loss": 0.5685,
"step": 5635
},
{
"epoch": 5.251513740102468,
"grad_norm": 0.4117334382190548,
"learning_rate": 6.886434242319641e-06,
"loss": 0.5547,
"step": 5640
},
{
"epoch": 5.256171401956218,
"grad_norm": 0.4016495989823376,
"learning_rate": 6.843286158094581e-06,
"loss": 0.5613,
"step": 5645
},
{
"epoch": 5.260829063809967,
"grad_norm": 0.44309336309542724,
"learning_rate": 6.800138073869521e-06,
"loss": 0.5637,
"step": 5650
},
{
"epoch": 5.265486725663717,
"grad_norm": 0.4278304988266007,
"learning_rate": 6.756989989644461e-06,
"loss": 0.5587,
"step": 5655
},
{
"epoch": 5.270144387517466,
"grad_norm": 0.4277346963877531,
"learning_rate": 6.713841905419399e-06,
"loss": 0.5575,
"step": 5660
},
{
"epoch": 5.274802049371216,
"grad_norm": 0.40625036184582464,
"learning_rate": 6.670693821194339e-06,
"loss": 0.5679,
"step": 5665
},
{
"epoch": 5.279459711224965,
"grad_norm": 0.4213874882571437,
"learning_rate": 6.627545736969279e-06,
"loss": 0.5592,
"step": 5670
},
{
"epoch": 5.284117373078715,
"grad_norm": 0.41154248379143205,
"learning_rate": 6.584397652744219e-06,
"loss": 0.5511,
"step": 5675
},
{
"epoch": 5.288775034932464,
"grad_norm": 0.44036501495565056,
"learning_rate": 6.541249568519157e-06,
"loss": 0.5765,
"step": 5680
},
{
"epoch": 5.293432696786214,
"grad_norm": 0.41669603634270275,
"learning_rate": 6.498101484294097e-06,
"loss": 0.5635,
"step": 5685
},
{
"epoch": 5.298090358639962,
"grad_norm": 0.4035117456098896,
"learning_rate": 6.454953400069037e-06,
"loss": 0.5676,
"step": 5690
},
{
"epoch": 5.302748020493712,
"grad_norm": 0.4105214591482102,
"learning_rate": 6.411805315843977e-06,
"loss": 0.5653,
"step": 5695
},
{
"epoch": 5.307405682347461,
"grad_norm": 0.40962030817642703,
"learning_rate": 6.368657231618916e-06,
"loss": 0.5576,
"step": 5700
},
{
"epoch": 5.312063344201211,
"grad_norm": 0.41377230715840563,
"learning_rate": 6.325509147393856e-06,
"loss": 0.5571,
"step": 5705
},
{
"epoch": 5.31672100605496,
"grad_norm": 0.4333285787168063,
"learning_rate": 6.282361063168796e-06,
"loss": 0.5671,
"step": 5710
},
{
"epoch": 5.32137866790871,
"grad_norm": 0.42261550816877297,
"learning_rate": 6.239212978943735e-06,
"loss": 0.5661,
"step": 5715
},
{
"epoch": 5.326036329762459,
"grad_norm": 0.4288493148116201,
"learning_rate": 6.196064894718675e-06,
"loss": 0.565,
"step": 5720
},
{
"epoch": 5.330693991616209,
"grad_norm": 0.44420344734015876,
"learning_rate": 6.152916810493614e-06,
"loss": 0.5701,
"step": 5725
},
{
"epoch": 5.335351653469958,
"grad_norm": 0.438366969446799,
"learning_rate": 6.109768726268554e-06,
"loss": 0.5734,
"step": 5730
},
{
"epoch": 5.340009315323708,
"grad_norm": 0.41601777699539805,
"learning_rate": 6.066620642043493e-06,
"loss": 0.5668,
"step": 5735
},
{
"epoch": 5.344666977177457,
"grad_norm": 0.42093441496169604,
"learning_rate": 6.023472557818433e-06,
"loss": 0.5707,
"step": 5740
},
{
"epoch": 5.349324639031206,
"grad_norm": 0.44402461374564856,
"learning_rate": 5.980324473593372e-06,
"loss": 0.5611,
"step": 5745
},
{
"epoch": 5.353982300884955,
"grad_norm": 0.41520612597240397,
"learning_rate": 5.937176389368312e-06,
"loss": 0.5575,
"step": 5750
},
{
"epoch": 5.358639962738705,
"grad_norm": 0.429765461012515,
"learning_rate": 5.894028305143251e-06,
"loss": 0.56,
"step": 5755
},
{
"epoch": 5.363297624592454,
"grad_norm": 0.4122117030859129,
"learning_rate": 5.850880220918191e-06,
"loss": 0.5616,
"step": 5760
},
{
"epoch": 5.367955286446204,
"grad_norm": 0.42948434368378274,
"learning_rate": 5.807732136693131e-06,
"loss": 0.5609,
"step": 5765
},
{
"epoch": 5.372612948299953,
"grad_norm": 0.42924647584016984,
"learning_rate": 5.76458405246807e-06,
"loss": 0.5665,
"step": 5770
},
{
"epoch": 5.377270610153703,
"grad_norm": 0.4065232953187284,
"learning_rate": 5.72143596824301e-06,
"loss": 0.5611,
"step": 5775
},
{
"epoch": 5.381928272007452,
"grad_norm": 0.4269979309846978,
"learning_rate": 5.67828788401795e-06,
"loss": 0.5615,
"step": 5780
},
{
"epoch": 5.386585933861202,
"grad_norm": 0.43683618580909445,
"learning_rate": 5.6351397997928894e-06,
"loss": 0.5653,
"step": 5785
},
{
"epoch": 5.391243595714951,
"grad_norm": 0.4138395029946517,
"learning_rate": 5.591991715567829e-06,
"loss": 0.5802,
"step": 5790
},
{
"epoch": 5.395901257568701,
"grad_norm": 0.39958490348240855,
"learning_rate": 5.5488436313427685e-06,
"loss": 0.573,
"step": 5795
},
{
"epoch": 5.4005589194224495,
"grad_norm": 0.4213641587451733,
"learning_rate": 5.5056955471177085e-06,
"loss": 0.561,
"step": 5800
},
{
"epoch": 5.405216581276199,
"grad_norm": 0.4301783875024671,
"learning_rate": 5.462547462892648e-06,
"loss": 0.5911,
"step": 5805
},
{
"epoch": 5.4098742431299485,
"grad_norm": 0.42552211116782834,
"learning_rate": 5.4193993786675876e-06,
"loss": 0.5581,
"step": 5810
},
{
"epoch": 5.414531904983698,
"grad_norm": 0.4178598551406305,
"learning_rate": 5.376251294442527e-06,
"loss": 0.5568,
"step": 5815
},
{
"epoch": 5.4191895668374475,
"grad_norm": 0.39499266804796795,
"learning_rate": 5.333103210217467e-06,
"loss": 0.5573,
"step": 5820
},
{
"epoch": 5.423847228691197,
"grad_norm": 0.4119408264424358,
"learning_rate": 5.289955125992406e-06,
"loss": 0.5637,
"step": 5825
},
{
"epoch": 5.4285048905449464,
"grad_norm": 0.4459547056976113,
"learning_rate": 5.246807041767346e-06,
"loss": 0.5585,
"step": 5830
},
{
"epoch": 5.433162552398696,
"grad_norm": 0.41872986179375277,
"learning_rate": 5.203658957542285e-06,
"loss": 0.5613,
"step": 5835
},
{
"epoch": 5.437820214252445,
"grad_norm": 0.4264565493001382,
"learning_rate": 5.160510873317225e-06,
"loss": 0.5826,
"step": 5840
},
{
"epoch": 5.442477876106195,
"grad_norm": 0.422871863759466,
"learning_rate": 5.117362789092165e-06,
"loss": 0.5572,
"step": 5845
},
{
"epoch": 5.447135537959944,
"grad_norm": 0.42489006212816244,
"learning_rate": 5.074214704867105e-06,
"loss": 0.5717,
"step": 5850
},
{
"epoch": 5.451793199813694,
"grad_norm": 0.42662171969491214,
"learning_rate": 5.031066620642044e-06,
"loss": 0.5692,
"step": 5855
},
{
"epoch": 5.4564508616674425,
"grad_norm": 0.408395244062005,
"learning_rate": 4.987918536416984e-06,
"loss": 0.5663,
"step": 5860
},
{
"epoch": 5.461108523521192,
"grad_norm": 0.44099787038065397,
"learning_rate": 4.944770452191923e-06,
"loss": 0.5857,
"step": 5865
},
{
"epoch": 5.4657661853749415,
"grad_norm": 0.42041261750553266,
"learning_rate": 4.901622367966863e-06,
"loss": 0.5652,
"step": 5870
},
{
"epoch": 5.470423847228691,
"grad_norm": 0.45645506662310636,
"learning_rate": 4.858474283741802e-06,
"loss": 0.5386,
"step": 5875
},
{
"epoch": 5.4750815090824405,
"grad_norm": 0.41592940114800914,
"learning_rate": 4.815326199516742e-06,
"loss": 0.5596,
"step": 5880
},
{
"epoch": 5.47973917093619,
"grad_norm": 0.4294223606896892,
"learning_rate": 4.772178115291681e-06,
"loss": 0.5634,
"step": 5885
},
{
"epoch": 5.4843968327899395,
"grad_norm": 0.40597268155861205,
"learning_rate": 4.729030031066621e-06,
"loss": 0.5719,
"step": 5890
},
{
"epoch": 5.489054494643689,
"grad_norm": 0.4028982988870574,
"learning_rate": 4.68588194684156e-06,
"loss": 0.5736,
"step": 5895
},
{
"epoch": 5.4937121564974385,
"grad_norm": 0.40186298121389796,
"learning_rate": 4.6427338626165e-06,
"loss": 0.5599,
"step": 5900
},
{
"epoch": 5.498369818351188,
"grad_norm": 0.42303591732933404,
"learning_rate": 4.599585778391439e-06,
"loss": 0.5818,
"step": 5905
},
{
"epoch": 5.5030274802049375,
"grad_norm": 0.4234758874100327,
"learning_rate": 4.556437694166379e-06,
"loss": 0.5542,
"step": 5910
},
{
"epoch": 5.507685142058687,
"grad_norm": 0.43047017308758134,
"learning_rate": 4.513289609941319e-06,
"loss": 0.5635,
"step": 5915
},
{
"epoch": 5.512342803912436,
"grad_norm": 0.40987555865691677,
"learning_rate": 4.470141525716258e-06,
"loss": 0.5559,
"step": 5920
},
{
"epoch": 5.517000465766185,
"grad_norm": 0.3863516212833326,
"learning_rate": 4.426993441491198e-06,
"loss": 0.5685,
"step": 5925
},
{
"epoch": 5.521658127619935,
"grad_norm": 0.4105942607404805,
"learning_rate": 4.383845357266138e-06,
"loss": 0.559,
"step": 5930
},
{
"epoch": 5.526315789473684,
"grad_norm": 0.39860381357892277,
"learning_rate": 4.340697273041077e-06,
"loss": 0.5617,
"step": 5935
},
{
"epoch": 5.530973451327434,
"grad_norm": 0.41665052343649345,
"learning_rate": 4.297549188816017e-06,
"loss": 0.5626,
"step": 5940
},
{
"epoch": 5.535631113181183,
"grad_norm": 0.4095867788678644,
"learning_rate": 4.254401104590956e-06,
"loss": 0.5653,
"step": 5945
},
{
"epoch": 5.5402887750349326,
"grad_norm": 0.4195303349878415,
"learning_rate": 4.211253020365896e-06,
"loss": 0.5839,
"step": 5950
},
{
"epoch": 5.544946436888682,
"grad_norm": 0.4120068956230269,
"learning_rate": 4.168104936140835e-06,
"loss": 0.5438,
"step": 5955
},
{
"epoch": 5.5496040987424315,
"grad_norm": 0.4196848830981198,
"learning_rate": 4.124956851915775e-06,
"loss": 0.5646,
"step": 5960
},
{
"epoch": 5.554261760596181,
"grad_norm": 0.40960812480651215,
"learning_rate": 4.0818087676907145e-06,
"loss": 0.5621,
"step": 5965
},
{
"epoch": 5.5589194224499305,
"grad_norm": 0.4173392960919226,
"learning_rate": 4.0386606834656544e-06,
"loss": 0.5564,
"step": 5970
},
{
"epoch": 5.563577084303679,
"grad_norm": 0.4212717258387222,
"learning_rate": 3.9955125992405935e-06,
"loss": 0.5678,
"step": 5975
},
{
"epoch": 5.568234746157429,
"grad_norm": 0.42550419511681836,
"learning_rate": 3.9523645150155335e-06,
"loss": 0.5662,
"step": 5980
},
{
"epoch": 5.572892408011178,
"grad_norm": 0.41623348722575887,
"learning_rate": 3.909216430790473e-06,
"loss": 0.5643,
"step": 5985
},
{
"epoch": 5.577550069864928,
"grad_norm": 0.41015141306127173,
"learning_rate": 3.8660683465654126e-06,
"loss": 0.5576,
"step": 5990
},
{
"epoch": 5.582207731718677,
"grad_norm": 0.42500858113635887,
"learning_rate": 3.8229202623403525e-06,
"loss": 0.5587,
"step": 5995
},
{
"epoch": 5.586865393572427,
"grad_norm": 0.4320183407766433,
"learning_rate": 3.779772178115292e-06,
"loss": 0.5761,
"step": 6000
},
{
"epoch": 5.591523055426176,
"grad_norm": 0.40912841983889053,
"learning_rate": 3.736624093890231e-06,
"loss": 0.5715,
"step": 6005
},
{
"epoch": 5.596180717279926,
"grad_norm": 0.40709797406945314,
"learning_rate": 3.693476009665171e-06,
"loss": 0.5675,
"step": 6010
},
{
"epoch": 5.600838379133675,
"grad_norm": 0.41347940747913237,
"learning_rate": 3.6503279254401107e-06,
"loss": 0.5723,
"step": 6015
},
{
"epoch": 5.605496040987425,
"grad_norm": 0.4169211842077042,
"learning_rate": 3.6071798412150506e-06,
"loss": 0.5585,
"step": 6020
},
{
"epoch": 5.610153702841174,
"grad_norm": 0.42537558386566227,
"learning_rate": 3.5640317569899898e-06,
"loss": 0.5531,
"step": 6025
},
{
"epoch": 5.614811364694923,
"grad_norm": 0.4305352425559143,
"learning_rate": 3.5208836727649297e-06,
"loss": 0.5604,
"step": 6030
},
{
"epoch": 5.619469026548672,
"grad_norm": 0.4036448018189107,
"learning_rate": 3.477735588539869e-06,
"loss": 0.5539,
"step": 6035
},
{
"epoch": 5.624126688402422,
"grad_norm": 0.41741840345951775,
"learning_rate": 3.434587504314809e-06,
"loss": 0.5563,
"step": 6040
},
{
"epoch": 5.628784350256171,
"grad_norm": 0.4033377591192777,
"learning_rate": 3.3914394200897483e-06,
"loss": 0.5645,
"step": 6045
},
{
"epoch": 5.633442012109921,
"grad_norm": 0.4171782665393845,
"learning_rate": 3.348291335864688e-06,
"loss": 0.5797,
"step": 6050
},
{
"epoch": 5.63809967396367,
"grad_norm": 0.39433610011537246,
"learning_rate": 3.3051432516396274e-06,
"loss": 0.5583,
"step": 6055
},
{
"epoch": 5.64275733581742,
"grad_norm": 0.4103448842108257,
"learning_rate": 3.2619951674145674e-06,
"loss": 0.5625,
"step": 6060
},
{
"epoch": 5.647414997671169,
"grad_norm": 0.3889964864956959,
"learning_rate": 3.2188470831895065e-06,
"loss": 0.5574,
"step": 6065
},
{
"epoch": 5.652072659524919,
"grad_norm": 0.39757072994654735,
"learning_rate": 3.1756989989644464e-06,
"loss": 0.5617,
"step": 6070
},
{
"epoch": 5.656730321378668,
"grad_norm": 0.40262036958192626,
"learning_rate": 3.1325509147393856e-06,
"loss": 0.5668,
"step": 6075
},
{
"epoch": 5.661387983232418,
"grad_norm": 0.39979293172348945,
"learning_rate": 3.089402830514325e-06,
"loss": 0.5684,
"step": 6080
},
{
"epoch": 5.666045645086166,
"grad_norm": 0.42639437886086473,
"learning_rate": 3.046254746289265e-06,
"loss": 0.5749,
"step": 6085
},
{
"epoch": 5.670703306939917,
"grad_norm": 0.4038998352505134,
"learning_rate": 3.0031066620642046e-06,
"loss": 0.5505,
"step": 6090
},
{
"epoch": 5.675360968793665,
"grad_norm": 0.42105952397990837,
"learning_rate": 2.959958577839144e-06,
"loss": 0.5506,
"step": 6095
},
{
"epoch": 5.680018630647415,
"grad_norm": 0.3907174988280262,
"learning_rate": 2.9168104936140837e-06,
"loss": 0.5632,
"step": 6100
},
{
"epoch": 5.684676292501164,
"grad_norm": 0.4142483675415474,
"learning_rate": 2.873662409389023e-06,
"loss": 0.5695,
"step": 6105
},
{
"epoch": 5.689333954354914,
"grad_norm": 0.41194771682758324,
"learning_rate": 2.8305143251639627e-06,
"loss": 0.5645,
"step": 6110
},
{
"epoch": 5.693991616208663,
"grad_norm": 0.41453421686895764,
"learning_rate": 2.7873662409389023e-06,
"loss": 0.5572,
"step": 6115
},
{
"epoch": 5.698649278062413,
"grad_norm": 0.4065816283780963,
"learning_rate": 2.7442181567138422e-06,
"loss": 0.5704,
"step": 6120
},
{
"epoch": 5.703306939916162,
"grad_norm": 0.42769337257365975,
"learning_rate": 2.7010700724887818e-06,
"loss": 0.5698,
"step": 6125
},
{
"epoch": 5.707964601769912,
"grad_norm": 0.3743277260802833,
"learning_rate": 2.6579219882637213e-06,
"loss": 0.5562,
"step": 6130
},
{
"epoch": 5.712622263623661,
"grad_norm": 0.4035379866637785,
"learning_rate": 2.614773904038661e-06,
"loss": 0.5655,
"step": 6135
},
{
"epoch": 5.717279925477411,
"grad_norm": 0.41488838089367225,
"learning_rate": 2.5716258198136004e-06,
"loss": 0.568,
"step": 6140
},
{
"epoch": 5.72193758733116,
"grad_norm": 0.41383472645868846,
"learning_rate": 2.52847773558854e-06,
"loss": 0.5565,
"step": 6145
},
{
"epoch": 5.726595249184909,
"grad_norm": 0.4054906747465409,
"learning_rate": 2.4853296513634795e-06,
"loss": 0.5666,
"step": 6150
},
{
"epoch": 5.731252911038658,
"grad_norm": 0.40137154610973613,
"learning_rate": 2.442181567138419e-06,
"loss": 0.5759,
"step": 6155
},
{
"epoch": 5.735910572892408,
"grad_norm": 0.4070737264766523,
"learning_rate": 2.399033482913359e-06,
"loss": 0.5726,
"step": 6160
},
{
"epoch": 5.740568234746157,
"grad_norm": 0.4007959321109693,
"learning_rate": 2.3558853986882985e-06,
"loss": 0.5746,
"step": 6165
},
{
"epoch": 5.745225896599907,
"grad_norm": 0.40888165957386624,
"learning_rate": 2.312737314463238e-06,
"loss": 0.5524,
"step": 6170
},
{
"epoch": 5.749883558453656,
"grad_norm": 0.4035306077364339,
"learning_rate": 2.2695892302381776e-06,
"loss": 0.5673,
"step": 6175
},
{
"epoch": 5.754541220307406,
"grad_norm": 0.39502013880626413,
"learning_rate": 2.226441146013117e-06,
"loss": 0.5655,
"step": 6180
},
{
"epoch": 5.759198882161155,
"grad_norm": 0.43395568549444674,
"learning_rate": 2.1832930617880566e-06,
"loss": 0.5632,
"step": 6185
},
{
"epoch": 5.763856544014905,
"grad_norm": 0.39420045671889875,
"learning_rate": 2.140144977562996e-06,
"loss": 0.5551,
"step": 6190
},
{
"epoch": 5.768514205868654,
"grad_norm": 0.39867593169108606,
"learning_rate": 2.096996893337936e-06,
"loss": 0.5742,
"step": 6195
},
{
"epoch": 5.773171867722404,
"grad_norm": 0.40156836437794047,
"learning_rate": 2.0538488091128757e-06,
"loss": 0.5699,
"step": 6200
},
{
"epoch": 5.777829529576152,
"grad_norm": 0.4048010544032807,
"learning_rate": 2.010700724887815e-06,
"loss": 0.5756,
"step": 6205
},
{
"epoch": 5.782487191429902,
"grad_norm": 0.38549911867624526,
"learning_rate": 1.9675526406627547e-06,
"loss": 0.5579,
"step": 6210
},
{
"epoch": 5.787144853283651,
"grad_norm": 0.39742374128807434,
"learning_rate": 1.9244045564376943e-06,
"loss": 0.557,
"step": 6215
},
{
"epoch": 5.791802515137401,
"grad_norm": 0.40433000457181545,
"learning_rate": 1.8812564722126338e-06,
"loss": 0.5574,
"step": 6220
},
{
"epoch": 5.79646017699115,
"grad_norm": 0.3941441139360801,
"learning_rate": 1.8381083879875736e-06,
"loss": 0.5682,
"step": 6225
},
{
"epoch": 5.8011178388449,
"grad_norm": 0.3992145826472066,
"learning_rate": 1.794960303762513e-06,
"loss": 0.5807,
"step": 6230
},
{
"epoch": 5.805775500698649,
"grad_norm": 0.40959722802929466,
"learning_rate": 1.7518122195374526e-06,
"loss": 0.5724,
"step": 6235
},
{
"epoch": 5.810433162552399,
"grad_norm": 0.40172842565861966,
"learning_rate": 1.7086641353123924e-06,
"loss": 0.5643,
"step": 6240
},
{
"epoch": 5.815090824406148,
"grad_norm": 0.394640153858509,
"learning_rate": 1.665516051087332e-06,
"loss": 0.5621,
"step": 6245
},
{
"epoch": 5.819748486259898,
"grad_norm": 0.4033965017648463,
"learning_rate": 1.6223679668622715e-06,
"loss": 0.5673,
"step": 6250
},
{
"epoch": 5.824406148113647,
"grad_norm": 0.398166753004289,
"learning_rate": 1.579219882637211e-06,
"loss": 0.5815,
"step": 6255
},
{
"epoch": 5.829063809967396,
"grad_norm": 0.40136934344268105,
"learning_rate": 1.5360717984121505e-06,
"loss": 0.5542,
"step": 6260
},
{
"epoch": 5.833721471821145,
"grad_norm": 0.42153262160447147,
"learning_rate": 1.49292371418709e-06,
"loss": 0.5644,
"step": 6265
},
{
"epoch": 5.838379133674895,
"grad_norm": 0.4013430168188573,
"learning_rate": 1.4497756299620296e-06,
"loss": 0.5807,
"step": 6270
},
{
"epoch": 5.843036795528644,
"grad_norm": 0.40144159950620084,
"learning_rate": 1.4066275457369694e-06,
"loss": 0.5672,
"step": 6275
},
{
"epoch": 5.847694457382394,
"grad_norm": 0.3905704278699857,
"learning_rate": 1.363479461511909e-06,
"loss": 0.5643,
"step": 6280
},
{
"epoch": 5.852352119236143,
"grad_norm": 0.38944035369021035,
"learning_rate": 1.3203313772868484e-06,
"loss": 0.5686,
"step": 6285
},
{
"epoch": 5.857009781089893,
"grad_norm": 0.3916599156858741,
"learning_rate": 1.277183293061788e-06,
"loss": 0.5562,
"step": 6290
},
{
"epoch": 5.861667442943642,
"grad_norm": 0.3736920720767908,
"learning_rate": 1.2340352088367277e-06,
"loss": 0.5545,
"step": 6295
},
{
"epoch": 5.866325104797392,
"grad_norm": 0.3921475156350814,
"learning_rate": 1.1908871246116673e-06,
"loss": 0.5569,
"step": 6300
},
{
"epoch": 5.870982766651141,
"grad_norm": 0.39718217346161483,
"learning_rate": 1.1477390403866068e-06,
"loss": 0.5473,
"step": 6305
},
{
"epoch": 5.875640428504891,
"grad_norm": 0.3994935700401086,
"learning_rate": 1.1045909561615463e-06,
"loss": 0.5736,
"step": 6310
},
{
"epoch": 5.8802980903586395,
"grad_norm": 0.40030906526793253,
"learning_rate": 1.061442871936486e-06,
"loss": 0.5441,
"step": 6315
},
{
"epoch": 5.88495575221239,
"grad_norm": 0.38817914033940354,
"learning_rate": 1.0182947877114256e-06,
"loss": 0.5584,
"step": 6320
},
{
"epoch": 5.8896134140661385,
"grad_norm": 0.39461128334067946,
"learning_rate": 9.751467034863652e-07,
"loss": 0.5598,
"step": 6325
},
{
"epoch": 5.894271075919888,
"grad_norm": 0.4000216002515271,
"learning_rate": 9.319986192613048e-07,
"loss": 0.5767,
"step": 6330
},
{
"epoch": 5.8989287377736375,
"grad_norm": 0.405724680285917,
"learning_rate": 8.888505350362444e-07,
"loss": 0.5532,
"step": 6335
},
{
"epoch": 5.903586399627387,
"grad_norm": 0.3966245206781564,
"learning_rate": 8.45702450811184e-07,
"loss": 0.5548,
"step": 6340
},
{
"epoch": 5.9082440614811365,
"grad_norm": 0.3985605889076508,
"learning_rate": 8.025543665861236e-07,
"loss": 0.5435,
"step": 6345
},
{
"epoch": 5.912901723334886,
"grad_norm": 0.3911890676900607,
"learning_rate": 7.594062823610632e-07,
"loss": 0.563,
"step": 6350
},
{
"epoch": 5.9175593851886354,
"grad_norm": 0.401965871420791,
"learning_rate": 7.162581981360028e-07,
"loss": 0.5627,
"step": 6355
},
{
"epoch": 5.922217047042385,
"grad_norm": 0.39202956971696346,
"learning_rate": 6.731101139109423e-07,
"loss": 0.5494,
"step": 6360
},
{
"epoch": 5.926874708896134,
"grad_norm": 0.40217414438035676,
"learning_rate": 6.29962029685882e-07,
"loss": 0.5716,
"step": 6365
},
{
"epoch": 5.931532370749884,
"grad_norm": 0.39499893054659513,
"learning_rate": 5.868139454608215e-07,
"loss": 0.5495,
"step": 6370
},
{
"epoch": 5.936190032603633,
"grad_norm": 0.4016719588732879,
"learning_rate": 5.436658612357612e-07,
"loss": 0.5802,
"step": 6375
},
{
"epoch": 5.940847694457382,
"grad_norm": 0.39553277802822207,
"learning_rate": 5.005177770107007e-07,
"loss": 0.5606,
"step": 6380
},
{
"epoch": 5.9455053563111315,
"grad_norm": 0.39850656808167323,
"learning_rate": 4.5736969278564034e-07,
"loss": 0.5679,
"step": 6385
},
{
"epoch": 5.950163018164881,
"grad_norm": 0.401297459288811,
"learning_rate": 4.142216085605799e-07,
"loss": 0.5821,
"step": 6390
},
{
"epoch": 5.9548206800186305,
"grad_norm": 0.3972992243477489,
"learning_rate": 3.710735243355195e-07,
"loss": 0.5721,
"step": 6395
},
{
"epoch": 5.95947834187238,
"grad_norm": 0.3958166758264783,
"learning_rate": 3.279254401104591e-07,
"loss": 0.5678,
"step": 6400
},
{
"epoch": 5.9641360037261295,
"grad_norm": 0.3834453135346943,
"learning_rate": 2.847773558853987e-07,
"loss": 0.5566,
"step": 6405
},
{
"epoch": 5.968793665579879,
"grad_norm": 0.39085792149862253,
"learning_rate": 2.416292716603383e-07,
"loss": 0.5616,
"step": 6410
},
{
"epoch": 5.9734513274336285,
"grad_norm": 0.3816402087868799,
"learning_rate": 1.984811874352779e-07,
"loss": 0.5426,
"step": 6415
},
{
"epoch": 5.978108989287378,
"grad_norm": 0.3949736427696376,
"learning_rate": 1.5533310321021747e-07,
"loss": 0.549,
"step": 6420
},
{
"epoch": 5.9827666511411275,
"grad_norm": 0.38007476087612035,
"learning_rate": 1.1218501898515707e-07,
"loss": 0.5606,
"step": 6425
},
{
"epoch": 5.987424312994877,
"grad_norm": 0.3925924080479119,
"learning_rate": 6.903693476009665e-08,
"loss": 0.5742,
"step": 6430
},
{
"epoch": 5.992081974848626,
"grad_norm": 0.38285900233997977,
"learning_rate": 2.5888850535036245e-08,
"loss": 0.5667,
"step": 6435
},
{
"epoch": 5.994876571960876,
"step": 6438,
"total_flos": 3121889937063936.0,
"train_loss": 0.7511046439751812,
"train_runtime": 81013.2052,
"train_samples_per_second": 1.272,
"train_steps_per_second": 0.079
}
],
"logging_steps": 5,
"max_steps": 6438,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3121889937063936.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}