Files
armv8mac_to_x86_qwen25coder…/trainer_state.json
ModelHub XC f924443151 初始化项目,由ModelHub XC社区提供模型
Model: adpretko/armv8mac_to_x86_qwen25coder_0p5b_full
Source: Original Platform
2026-05-27 19:20:18 +08:00

7477 lines
183 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5000058835296471,
"eval_steps": 500,
"global_step": 10623,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00047068237176847136,
"grad_norm": 5.434264659881592,
"learning_rate": 6.269592476489028e-07,
"loss": 0.5668,
"step": 10
},
{
"epoch": 0.0009413647435369427,
"grad_norm": 4.5453104972839355,
"learning_rate": 1.2539184952978056e-06,
"loss": 0.5687,
"step": 20
},
{
"epoch": 0.0014120471153054141,
"grad_norm": 2.296740770339966,
"learning_rate": 1.8808777429467086e-06,
"loss": 0.4541,
"step": 30
},
{
"epoch": 0.0018827294870738854,
"grad_norm": 1.762229561805725,
"learning_rate": 2.507836990595611e-06,
"loss": 0.3285,
"step": 40
},
{
"epoch": 0.0023534118588423567,
"grad_norm": 1.505702257156372,
"learning_rate": 3.1347962382445144e-06,
"loss": 0.2412,
"step": 50
},
{
"epoch": 0.0028240942306108283,
"grad_norm": 1.377434253692627,
"learning_rate": 3.7617554858934172e-06,
"loss": 0.1832,
"step": 60
},
{
"epoch": 0.0032947766023792993,
"grad_norm": 1.0926605463027954,
"learning_rate": 4.3887147335423205e-06,
"loss": 0.1384,
"step": 70
},
{
"epoch": 0.003765458974147771,
"grad_norm": 1.1871973276138306,
"learning_rate": 5.015673981191222e-06,
"loss": 0.1357,
"step": 80
},
{
"epoch": 0.004236141345916242,
"grad_norm": 1.9962201118469238,
"learning_rate": 5.642633228840125e-06,
"loss": 0.1131,
"step": 90
},
{
"epoch": 0.0047068237176847135,
"grad_norm": 1.1790374517440796,
"learning_rate": 6.269592476489029e-06,
"loss": 0.113,
"step": 100
},
{
"epoch": 0.0051775060894531846,
"grad_norm": 1.3358230590820312,
"learning_rate": 6.896551724137932e-06,
"loss": 0.107,
"step": 110
},
{
"epoch": 0.0056481884612216565,
"grad_norm": 1.050266146659851,
"learning_rate": 7.5235109717868345e-06,
"loss": 0.1112,
"step": 120
},
{
"epoch": 0.006118870832990128,
"grad_norm": 1.7772488594055176,
"learning_rate": 8.150470219435737e-06,
"loss": 0.1031,
"step": 130
},
{
"epoch": 0.006589553204758599,
"grad_norm": 0.9741373062133789,
"learning_rate": 8.777429467084641e-06,
"loss": 0.0991,
"step": 140
},
{
"epoch": 0.00706023557652707,
"grad_norm": 1.69870126247406,
"learning_rate": 9.404388714733543e-06,
"loss": 0.1093,
"step": 150
},
{
"epoch": 0.007530917948295542,
"grad_norm": 1.0989326238632202,
"learning_rate": 1.0031347962382445e-05,
"loss": 0.0877,
"step": 160
},
{
"epoch": 0.008001600320064013,
"grad_norm": 0.9203445315361023,
"learning_rate": 1.0658307210031348e-05,
"loss": 0.0897,
"step": 170
},
{
"epoch": 0.008472282691832485,
"grad_norm": 0.9417378902435303,
"learning_rate": 1.128526645768025e-05,
"loss": 0.0839,
"step": 180
},
{
"epoch": 0.008942965063600955,
"grad_norm": 0.9646998047828674,
"learning_rate": 1.1912225705329154e-05,
"loss": 0.0866,
"step": 190
},
{
"epoch": 0.009413647435369427,
"grad_norm": 0.899651825428009,
"learning_rate": 1.2539184952978058e-05,
"loss": 0.0794,
"step": 200
},
{
"epoch": 0.009884329807137899,
"grad_norm": 0.985142171382904,
"learning_rate": 1.316614420062696e-05,
"loss": 0.0868,
"step": 210
},
{
"epoch": 0.010355012178906369,
"grad_norm": 0.7299129962921143,
"learning_rate": 1.3793103448275863e-05,
"loss": 0.0914,
"step": 220
},
{
"epoch": 0.010825694550674841,
"grad_norm": 0.8193659782409668,
"learning_rate": 1.4420062695924765e-05,
"loss": 0.0787,
"step": 230
},
{
"epoch": 0.011296376922443313,
"grad_norm": 0.8029590845108032,
"learning_rate": 1.5047021943573669e-05,
"loss": 0.0744,
"step": 240
},
{
"epoch": 0.011767059294211783,
"grad_norm": 0.866852343082428,
"learning_rate": 1.567398119122257e-05,
"loss": 0.0724,
"step": 250
},
{
"epoch": 0.012237741665980255,
"grad_norm": 0.6705352663993835,
"learning_rate": 1.6300940438871475e-05,
"loss": 0.077,
"step": 260
},
{
"epoch": 0.012708424037748725,
"grad_norm": 0.908003568649292,
"learning_rate": 1.6927899686520378e-05,
"loss": 0.0688,
"step": 270
},
{
"epoch": 0.013179106409517197,
"grad_norm": 0.9268977642059326,
"learning_rate": 1.7554858934169282e-05,
"loss": 0.0827,
"step": 280
},
{
"epoch": 0.01364978878128567,
"grad_norm": 0.7269849181175232,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.077,
"step": 290
},
{
"epoch": 0.01412047115305414,
"grad_norm": 1.1154515743255615,
"learning_rate": 1.8808777429467086e-05,
"loss": 0.0725,
"step": 300
},
{
"epoch": 0.014591153524822612,
"grad_norm": 0.8751449584960938,
"learning_rate": 1.943573667711599e-05,
"loss": 0.0724,
"step": 310
},
{
"epoch": 0.015061835896591083,
"grad_norm": 1.0344765186309814,
"learning_rate": 1.999999953520864e-05,
"loss": 0.0694,
"step": 320
},
{
"epoch": 0.015532518268359554,
"grad_norm": 0.7169214487075806,
"learning_rate": 1.9999943760297725e-05,
"loss": 0.0643,
"step": 330
},
{
"epoch": 0.016003200640128026,
"grad_norm": 0.8278538584709167,
"learning_rate": 1.99997950277089e-05,
"loss": 0.0694,
"step": 340
},
{
"epoch": 0.016473883011896496,
"grad_norm": 0.6600526571273804,
"learning_rate": 1.999955333882476e-05,
"loss": 0.0746,
"step": 350
},
{
"epoch": 0.01694456538366497,
"grad_norm": 0.7181190252304077,
"learning_rate": 1.9999218695892e-05,
"loss": 0.0645,
"step": 360
},
{
"epoch": 0.01741524775543344,
"grad_norm": 0.555476725101471,
"learning_rate": 1.9998791102021396e-05,
"loss": 0.0663,
"step": 370
},
{
"epoch": 0.01788593012720191,
"grad_norm": 0.5973796248435974,
"learning_rate": 1.999827056118779e-05,
"loss": 0.0601,
"step": 380
},
{
"epoch": 0.018356612498970384,
"grad_norm": 0.5690473318099976,
"learning_rate": 1.9997657078230034e-05,
"loss": 0.0589,
"step": 390
},
{
"epoch": 0.018827294870738854,
"grad_norm": 0.5986892580986023,
"learning_rate": 1.9996950658850956e-05,
"loss": 0.0658,
"step": 400
},
{
"epoch": 0.019297977242507324,
"grad_norm": 0.7275658249855042,
"learning_rate": 1.99961513096173e-05,
"loss": 0.0599,
"step": 410
},
{
"epoch": 0.019768659614275798,
"grad_norm": 0.5199920535087585,
"learning_rate": 1.999525903795968e-05,
"loss": 0.0595,
"step": 420
},
{
"epoch": 0.020239341986044268,
"grad_norm": 0.5136874318122864,
"learning_rate": 1.9994273852172484e-05,
"loss": 0.0577,
"step": 430
},
{
"epoch": 0.020710024357812738,
"grad_norm": 0.6104478240013123,
"learning_rate": 1.9993195761413823e-05,
"loss": 0.0615,
"step": 440
},
{
"epoch": 0.021180706729581212,
"grad_norm": 0.551996648311615,
"learning_rate": 1.999202477570544e-05,
"loss": 0.0567,
"step": 450
},
{
"epoch": 0.021651389101349682,
"grad_norm": 0.6211404800415039,
"learning_rate": 1.9990760905932605e-05,
"loss": 0.0556,
"step": 460
},
{
"epoch": 0.022122071473118152,
"grad_norm": 0.4809170365333557,
"learning_rate": 1.9989404163844026e-05,
"loss": 0.0649,
"step": 470
},
{
"epoch": 0.022592753844886626,
"grad_norm": 0.5183375477790833,
"learning_rate": 1.9987954562051724e-05,
"loss": 0.0498,
"step": 480
},
{
"epoch": 0.023063436216655096,
"grad_norm": 0.5590595006942749,
"learning_rate": 1.998641211403095e-05,
"loss": 0.0579,
"step": 490
},
{
"epoch": 0.023534118588423567,
"grad_norm": 0.5478200912475586,
"learning_rate": 1.9984776834120015e-05,
"loss": 0.0672,
"step": 500
},
{
"epoch": 0.024004800960192037,
"grad_norm": 0.6530250906944275,
"learning_rate": 1.9983048737520186e-05,
"loss": 0.0571,
"step": 510
},
{
"epoch": 0.02447548333196051,
"grad_norm": 0.5334309935569763,
"learning_rate": 1.9981227840295544e-05,
"loss": 0.0547,
"step": 520
},
{
"epoch": 0.02494616570372898,
"grad_norm": 0.6887657046318054,
"learning_rate": 1.9979314159372815e-05,
"loss": 0.0608,
"step": 530
},
{
"epoch": 0.02541684807549745,
"grad_norm": 0.8331181406974792,
"learning_rate": 1.997730771254124e-05,
"loss": 0.0598,
"step": 540
},
{
"epoch": 0.025887530447265925,
"grad_norm": 0.6619186401367188,
"learning_rate": 1.9975208518452384e-05,
"loss": 0.0599,
"step": 550
},
{
"epoch": 0.026358212819034395,
"grad_norm": 0.5086070895195007,
"learning_rate": 1.9973016596619973e-05,
"loss": 0.0586,
"step": 560
},
{
"epoch": 0.026828895190802865,
"grad_norm": 0.4673551023006439,
"learning_rate": 1.9970731967419724e-05,
"loss": 0.0538,
"step": 570
},
{
"epoch": 0.02729957756257134,
"grad_norm": 0.43053972721099854,
"learning_rate": 1.9968354652089142e-05,
"loss": 0.0536,
"step": 580
},
{
"epoch": 0.02777025993433981,
"grad_norm": 0.4711664915084839,
"learning_rate": 1.9965884672727313e-05,
"loss": 0.0552,
"step": 590
},
{
"epoch": 0.02824094230610828,
"grad_norm": 0.6097978353500366,
"learning_rate": 1.996332205229473e-05,
"loss": 0.0552,
"step": 600
},
{
"epoch": 0.028711624677876753,
"grad_norm": 0.5124568939208984,
"learning_rate": 1.9960666814613043e-05,
"loss": 0.0438,
"step": 610
},
{
"epoch": 0.029182307049645223,
"grad_norm": 0.6268178224563599,
"learning_rate": 1.995791898436487e-05,
"loss": 0.0552,
"step": 620
},
{
"epoch": 0.029652989421413693,
"grad_norm": 0.6327022314071655,
"learning_rate": 1.995507858709354e-05,
"loss": 0.0592,
"step": 630
},
{
"epoch": 0.030123671793182167,
"grad_norm": 0.47007712721824646,
"learning_rate": 1.9952145649202885e-05,
"loss": 0.063,
"step": 640
},
{
"epoch": 0.030594354164950637,
"grad_norm": 0.5984840989112854,
"learning_rate": 1.9949120197956956e-05,
"loss": 0.0503,
"step": 650
},
{
"epoch": 0.031065036536719107,
"grad_norm": 0.5706822872161865,
"learning_rate": 1.9946002261479817e-05,
"loss": 0.0552,
"step": 660
},
{
"epoch": 0.03153571890848758,
"grad_norm": 0.5763556361198425,
"learning_rate": 1.9942791868755233e-05,
"loss": 0.0603,
"step": 670
},
{
"epoch": 0.03200640128025605,
"grad_norm": 0.6085019111633301,
"learning_rate": 1.993948904962644e-05,
"loss": 0.0503,
"step": 680
},
{
"epoch": 0.03247708365202452,
"grad_norm": 0.5140377879142761,
"learning_rate": 1.9936093834795853e-05,
"loss": 0.0518,
"step": 690
},
{
"epoch": 0.03294776602379299,
"grad_norm": 0.4948691725730896,
"learning_rate": 1.993260625582478e-05,
"loss": 0.0562,
"step": 700
},
{
"epoch": 0.03341844839556146,
"grad_norm": 0.6141917109489441,
"learning_rate": 1.992902634513312e-05,
"loss": 0.0541,
"step": 710
},
{
"epoch": 0.03388913076732994,
"grad_norm": 0.43543606996536255,
"learning_rate": 1.9925354135999083e-05,
"loss": 0.0446,
"step": 720
},
{
"epoch": 0.03435981313909841,
"grad_norm": 0.5050569176673889,
"learning_rate": 1.992158966255887e-05,
"loss": 0.0523,
"step": 730
},
{
"epoch": 0.03483049551086688,
"grad_norm": 0.44272053241729736,
"learning_rate": 1.9917732959806336e-05,
"loss": 0.0552,
"step": 740
},
{
"epoch": 0.03530117788263535,
"grad_norm": 0.39607861638069153,
"learning_rate": 1.9913784063592708e-05,
"loss": 0.0432,
"step": 750
},
{
"epoch": 0.03577186025440382,
"grad_norm": 0.5186623930931091,
"learning_rate": 1.990974301062621e-05,
"loss": 0.052,
"step": 760
},
{
"epoch": 0.03624254262617229,
"grad_norm": 0.545556902885437,
"learning_rate": 1.990560983847174e-05,
"loss": 0.0468,
"step": 770
},
{
"epoch": 0.03671322499794077,
"grad_norm": 0.5561232566833496,
"learning_rate": 1.9901384585550527e-05,
"loss": 0.0487,
"step": 780
},
{
"epoch": 0.03718390736970924,
"grad_norm": 0.43631163239479065,
"learning_rate": 1.989706729113976e-05,
"loss": 0.0516,
"step": 790
},
{
"epoch": 0.03765458974147771,
"grad_norm": 0.5874558687210083,
"learning_rate": 1.9892657995372227e-05,
"loss": 0.0497,
"step": 800
},
{
"epoch": 0.03812527211324618,
"grad_norm": 0.45937541127204895,
"learning_rate": 1.988815673923596e-05,
"loss": 0.0499,
"step": 810
},
{
"epoch": 0.03859595448501465,
"grad_norm": 0.627604067325592,
"learning_rate": 1.9883563564573815e-05,
"loss": 0.0549,
"step": 820
},
{
"epoch": 0.03906663685678312,
"grad_norm": 0.48297274112701416,
"learning_rate": 1.9878878514083124e-05,
"loss": 0.0453,
"step": 830
},
{
"epoch": 0.039537319228551596,
"grad_norm": 0.5967549085617065,
"learning_rate": 1.9874101631315268e-05,
"loss": 0.0479,
"step": 840
},
{
"epoch": 0.040008001600320066,
"grad_norm": 0.4821009933948517,
"learning_rate": 1.9869232960675292e-05,
"loss": 0.0562,
"step": 850
},
{
"epoch": 0.040478683972088536,
"grad_norm": 0.41397395730018616,
"learning_rate": 1.9864272547421482e-05,
"loss": 0.0476,
"step": 860
},
{
"epoch": 0.040949366343857006,
"grad_norm": 0.4486595392227173,
"learning_rate": 1.985922043766495e-05,
"loss": 0.0528,
"step": 870
},
{
"epoch": 0.041420048715625477,
"grad_norm": 0.5143370032310486,
"learning_rate": 1.9854076678369197e-05,
"loss": 0.0511,
"step": 880
},
{
"epoch": 0.04189073108739395,
"grad_norm": 0.5163377523422241,
"learning_rate": 1.984884131734968e-05,
"loss": 0.0447,
"step": 890
},
{
"epoch": 0.042361413459162424,
"grad_norm": 0.43937593698501587,
"learning_rate": 1.9843514403273378e-05,
"loss": 0.0484,
"step": 900
},
{
"epoch": 0.042832095830930894,
"grad_norm": 0.38790178298950195,
"learning_rate": 1.9838095985658324e-05,
"loss": 0.0487,
"step": 910
},
{
"epoch": 0.043302778202699364,
"grad_norm": 0.6651208400726318,
"learning_rate": 1.9832586114873154e-05,
"loss": 0.06,
"step": 920
},
{
"epoch": 0.043773460574467835,
"grad_norm": 0.35643741488456726,
"learning_rate": 1.9826984842136637e-05,
"loss": 0.0453,
"step": 930
},
{
"epoch": 0.044244142946236305,
"grad_norm": 0.5988783836364746,
"learning_rate": 1.982129221951719e-05,
"loss": 0.0471,
"step": 940
},
{
"epoch": 0.044714825318004775,
"grad_norm": 0.3942262828350067,
"learning_rate": 1.9815508299932417e-05,
"loss": 0.043,
"step": 950
},
{
"epoch": 0.04518550768977325,
"grad_norm": 0.47634977102279663,
"learning_rate": 1.980963313714859e-05,
"loss": 0.0549,
"step": 960
},
{
"epoch": 0.04565619006154172,
"grad_norm": 0.4434799253940582,
"learning_rate": 1.9803666785780165e-05,
"loss": 0.05,
"step": 970
},
{
"epoch": 0.04612687243331019,
"grad_norm": 0.5270854234695435,
"learning_rate": 1.979760930128927e-05,
"loss": 0.0495,
"step": 980
},
{
"epoch": 0.04659755480507866,
"grad_norm": 0.576178252696991,
"learning_rate": 1.979146073998519e-05,
"loss": 0.0468,
"step": 990
},
{
"epoch": 0.04706823717684713,
"grad_norm": 0.4326517581939697,
"learning_rate": 1.9785221159023852e-05,
"loss": 0.0515,
"step": 1000
},
{
"epoch": 0.0475389195486156,
"grad_norm": 0.5435857176780701,
"learning_rate": 1.9778890616407266e-05,
"loss": 0.0511,
"step": 1010
},
{
"epoch": 0.048009601920384073,
"grad_norm": 0.5265092849731445,
"learning_rate": 1.9772469170983026e-05,
"loss": 0.0505,
"step": 1020
},
{
"epoch": 0.04848028429215255,
"grad_norm": 0.4413529634475708,
"learning_rate": 1.976595688244373e-05,
"loss": 0.0446,
"step": 1030
},
{
"epoch": 0.04895096666392102,
"grad_norm": 0.4792312979698181,
"learning_rate": 1.975935381132644e-05,
"loss": 0.0427,
"step": 1040
},
{
"epoch": 0.04942164903568949,
"grad_norm": 0.44915708899497986,
"learning_rate": 1.975266001901212e-05,
"loss": 0.041,
"step": 1050
},
{
"epoch": 0.04989233140745796,
"grad_norm": 0.47843867540359497,
"learning_rate": 1.9745875567725058e-05,
"loss": 0.0487,
"step": 1060
},
{
"epoch": 0.05036301377922643,
"grad_norm": 0.4393898546695709,
"learning_rate": 1.973900052053229e-05,
"loss": 0.0474,
"step": 1070
},
{
"epoch": 0.0508336961509949,
"grad_norm": 0.36783841252326965,
"learning_rate": 1.973203494134302e-05,
"loss": 0.0415,
"step": 1080
},
{
"epoch": 0.05130437852276338,
"grad_norm": 0.49190235137939453,
"learning_rate": 1.9724978894908017e-05,
"loss": 0.0492,
"step": 1090
},
{
"epoch": 0.05177506089453185,
"grad_norm": 0.4830932319164276,
"learning_rate": 1.971783244681902e-05,
"loss": 0.0474,
"step": 1100
},
{
"epoch": 0.05224574326630032,
"grad_norm": 0.43566033244132996,
"learning_rate": 1.9710595663508125e-05,
"loss": 0.0453,
"step": 1110
},
{
"epoch": 0.05271642563806879,
"grad_norm": 0.5832585692405701,
"learning_rate": 1.9703268612247172e-05,
"loss": 0.0528,
"step": 1120
},
{
"epoch": 0.05318710800983726,
"grad_norm": 0.459662526845932,
"learning_rate": 1.9695851361147098e-05,
"loss": 0.046,
"step": 1130
},
{
"epoch": 0.05365779038160573,
"grad_norm": 0.44846659898757935,
"learning_rate": 1.9688343979157348e-05,
"loss": 0.0515,
"step": 1140
},
{
"epoch": 0.05412847275337421,
"grad_norm": 0.4661378264427185,
"learning_rate": 1.968074653606519e-05,
"loss": 0.0451,
"step": 1150
},
{
"epoch": 0.05459915512514268,
"grad_norm": 0.3667706549167633,
"learning_rate": 1.9673059102495084e-05,
"loss": 0.0462,
"step": 1160
},
{
"epoch": 0.05506983749691115,
"grad_norm": 0.5238426327705383,
"learning_rate": 1.9665281749908034e-05,
"loss": 0.0466,
"step": 1170
},
{
"epoch": 0.05554051986867962,
"grad_norm": 0.6112167239189148,
"learning_rate": 1.9657414550600907e-05,
"loss": 0.038,
"step": 1180
},
{
"epoch": 0.05601120224044809,
"grad_norm": 0.5197986364364624,
"learning_rate": 1.964945757770578e-05,
"loss": 0.0449,
"step": 1190
},
{
"epoch": 0.05648188461221656,
"grad_norm": 0.4000227451324463,
"learning_rate": 1.964141090518923e-05,
"loss": 0.042,
"step": 1200
},
{
"epoch": 0.056952566983985035,
"grad_norm": 0.4052744209766388,
"learning_rate": 1.963327460785168e-05,
"loss": 0.0466,
"step": 1210
},
{
"epoch": 0.057423249355753506,
"grad_norm": 0.482947438955307,
"learning_rate": 1.962504876132669e-05,
"loss": 0.049,
"step": 1220
},
{
"epoch": 0.057893931727521976,
"grad_norm": 0.5074743628501892,
"learning_rate": 1.9616733442080253e-05,
"loss": 0.0418,
"step": 1230
},
{
"epoch": 0.058364614099290446,
"grad_norm": 0.3047001361846924,
"learning_rate": 1.960832872741007e-05,
"loss": 0.0391,
"step": 1240
},
{
"epoch": 0.058835296471058916,
"grad_norm": 0.4029083251953125,
"learning_rate": 1.9599834695444863e-05,
"loss": 0.0457,
"step": 1250
},
{
"epoch": 0.059305978842827387,
"grad_norm": 0.4102582335472107,
"learning_rate": 1.959125142514362e-05,
"loss": 0.0414,
"step": 1260
},
{
"epoch": 0.059776661214595864,
"grad_norm": 0.40880703926086426,
"learning_rate": 1.9582578996294882e-05,
"loss": 0.0446,
"step": 1270
},
{
"epoch": 0.060247343586364334,
"grad_norm": 0.3396308422088623,
"learning_rate": 1.957381748951599e-05,
"loss": 0.0429,
"step": 1280
},
{
"epoch": 0.060718025958132804,
"grad_norm": 0.5156334638595581,
"learning_rate": 1.9564966986252326e-05,
"loss": 0.0462,
"step": 1290
},
{
"epoch": 0.061188708329901274,
"grad_norm": 0.3503815829753876,
"learning_rate": 1.9556027568776577e-05,
"loss": 0.0438,
"step": 1300
},
{
"epoch": 0.061659390701669745,
"grad_norm": 0.4745870530605316,
"learning_rate": 1.9546999320187966e-05,
"loss": 0.0414,
"step": 1310
},
{
"epoch": 0.062130073073438215,
"grad_norm": 0.39651715755462646,
"learning_rate": 1.953788232441147e-05,
"loss": 0.0411,
"step": 1320
},
{
"epoch": 0.06260075544520669,
"grad_norm": 0.5001482963562012,
"learning_rate": 1.9528676666197026e-05,
"loss": 0.0449,
"step": 1330
},
{
"epoch": 0.06307143781697516,
"grad_norm": 0.39175641536712646,
"learning_rate": 1.951938243111879e-05,
"loss": 0.0352,
"step": 1340
},
{
"epoch": 0.06354212018874363,
"grad_norm": 0.4424596130847931,
"learning_rate": 1.9509999705574293e-05,
"loss": 0.0391,
"step": 1350
},
{
"epoch": 0.0640128025605121,
"grad_norm": 0.34089693427085876,
"learning_rate": 1.9500528576783667e-05,
"loss": 0.0454,
"step": 1360
},
{
"epoch": 0.06448348493228058,
"grad_norm": 0.5439393520355225,
"learning_rate": 1.9490969132788815e-05,
"loss": 0.0483,
"step": 1370
},
{
"epoch": 0.06495416730404904,
"grad_norm": 0.570132315158844,
"learning_rate": 1.9481321462452617e-05,
"loss": 0.0395,
"step": 1380
},
{
"epoch": 0.06542484967581752,
"grad_norm": 0.4053954780101776,
"learning_rate": 1.9471585655458073e-05,
"loss": 0.0385,
"step": 1390
},
{
"epoch": 0.06589553204758598,
"grad_norm": 0.3130151331424713,
"learning_rate": 1.9461761802307494e-05,
"loss": 0.046,
"step": 1400
},
{
"epoch": 0.06636621441935446,
"grad_norm": 0.4699443578720093,
"learning_rate": 1.945184999432166e-05,
"loss": 0.0401,
"step": 1410
},
{
"epoch": 0.06683689679112292,
"grad_norm": 0.320291668176651,
"learning_rate": 1.9441850323638944e-05,
"loss": 0.043,
"step": 1420
},
{
"epoch": 0.0673075791628914,
"grad_norm": 0.34498924016952515,
"learning_rate": 1.943176288321449e-05,
"loss": 0.0414,
"step": 1430
},
{
"epoch": 0.06777826153465988,
"grad_norm": 0.47924962639808655,
"learning_rate": 1.942158776681933e-05,
"loss": 0.0435,
"step": 1440
},
{
"epoch": 0.06824894390642834,
"grad_norm": 0.3699187636375427,
"learning_rate": 1.941132506903951e-05,
"loss": 0.0422,
"step": 1450
},
{
"epoch": 0.06871962627819682,
"grad_norm": 0.4283713102340698,
"learning_rate": 1.9400974885275226e-05,
"loss": 0.0384,
"step": 1460
},
{
"epoch": 0.06919030864996528,
"grad_norm": 0.3198941648006439,
"learning_rate": 1.9390537311739927e-05,
"loss": 0.0455,
"step": 1470
},
{
"epoch": 0.06966099102173376,
"grad_norm": 0.33609142899513245,
"learning_rate": 1.938001244545941e-05,
"loss": 0.0388,
"step": 1480
},
{
"epoch": 0.07013167339350224,
"grad_norm": 0.4284432828426361,
"learning_rate": 1.9369400384270948e-05,
"loss": 0.0402,
"step": 1490
},
{
"epoch": 0.0706023557652707,
"grad_norm": 0.3967095613479614,
"learning_rate": 1.935870122682235e-05,
"loss": 0.0447,
"step": 1500
},
{
"epoch": 0.07107303813703918,
"grad_norm": 0.4282098114490509,
"learning_rate": 1.934791507257105e-05,
"loss": 0.0376,
"step": 1510
},
{
"epoch": 0.07154372050880764,
"grad_norm": 0.42285192012786865,
"learning_rate": 1.933704202178321e-05,
"loss": 0.0437,
"step": 1520
},
{
"epoch": 0.07201440288057612,
"grad_norm": 0.3372488021850586,
"learning_rate": 1.9326082175532744e-05,
"loss": 0.0391,
"step": 1530
},
{
"epoch": 0.07248508525234458,
"grad_norm": 0.4501575529575348,
"learning_rate": 1.9315035635700412e-05,
"loss": 0.0405,
"step": 1540
},
{
"epoch": 0.07295576762411306,
"grad_norm": 0.37335875630378723,
"learning_rate": 1.9303902504972866e-05,
"loss": 0.0458,
"step": 1550
},
{
"epoch": 0.07342644999588153,
"grad_norm": 0.37942060828208923,
"learning_rate": 1.9292682886841683e-05,
"loss": 0.0424,
"step": 1560
},
{
"epoch": 0.07389713236765,
"grad_norm": 0.3235180675983429,
"learning_rate": 1.9281376885602412e-05,
"loss": 0.0429,
"step": 1570
},
{
"epoch": 0.07436781473941848,
"grad_norm": 0.5821486711502075,
"learning_rate": 1.926998460635361e-05,
"loss": 0.0415,
"step": 1580
},
{
"epoch": 0.07483849711118694,
"grad_norm": 0.4379030466079712,
"learning_rate": 1.9258506154995854e-05,
"loss": 0.0373,
"step": 1590
},
{
"epoch": 0.07530917948295542,
"grad_norm": 0.5411556363105774,
"learning_rate": 1.924694163823076e-05,
"loss": 0.0493,
"step": 1600
},
{
"epoch": 0.07577986185472389,
"grad_norm": 0.4541884660720825,
"learning_rate": 1.9235291163559996e-05,
"loss": 0.0464,
"step": 1610
},
{
"epoch": 0.07625054422649236,
"grad_norm": 0.29541483521461487,
"learning_rate": 1.922355483928428e-05,
"loss": 0.0391,
"step": 1620
},
{
"epoch": 0.07672122659826083,
"grad_norm": 0.4667107164859772,
"learning_rate": 1.9211732774502372e-05,
"loss": 0.0386,
"step": 1630
},
{
"epoch": 0.0771919089700293,
"grad_norm": 0.6333792805671692,
"learning_rate": 1.919982507911006e-05,
"loss": 0.0379,
"step": 1640
},
{
"epoch": 0.07766259134179777,
"grad_norm": 0.41373908519744873,
"learning_rate": 1.918783186379913e-05,
"loss": 0.039,
"step": 1650
},
{
"epoch": 0.07813327371356624,
"grad_norm": 0.42216387391090393,
"learning_rate": 1.917575324005636e-05,
"loss": 0.0398,
"step": 1660
},
{
"epoch": 0.07860395608533471,
"grad_norm": 0.3893977403640747,
"learning_rate": 1.916358932016246e-05,
"loss": 0.0437,
"step": 1670
},
{
"epoch": 0.07907463845710319,
"grad_norm": 0.40155720710754395,
"learning_rate": 1.9151340217191042e-05,
"loss": 0.0408,
"step": 1680
},
{
"epoch": 0.07954532082887165,
"grad_norm": 0.3144453763961792,
"learning_rate": 1.9139006045007567e-05,
"loss": 0.0382,
"step": 1690
},
{
"epoch": 0.08001600320064013,
"grad_norm": 0.31395450234413147,
"learning_rate": 1.9126586918268275e-05,
"loss": 0.0393,
"step": 1700
},
{
"epoch": 0.0804866855724086,
"grad_norm": 0.3450036644935608,
"learning_rate": 1.9114082952419134e-05,
"loss": 0.0397,
"step": 1710
},
{
"epoch": 0.08095736794417707,
"grad_norm": 0.3674869239330292,
"learning_rate": 1.9101494263694764e-05,
"loss": 0.0376,
"step": 1720
},
{
"epoch": 0.08142805031594554,
"grad_norm": 0.42553818225860596,
"learning_rate": 1.9088820969117348e-05,
"loss": 0.038,
"step": 1730
},
{
"epoch": 0.08189873268771401,
"grad_norm": 0.5304335355758667,
"learning_rate": 1.907606318649555e-05,
"loss": 0.0362,
"step": 1740
},
{
"epoch": 0.08236941505948249,
"grad_norm": 0.389398992061615,
"learning_rate": 1.906322103442343e-05,
"loss": 0.0463,
"step": 1750
},
{
"epoch": 0.08284009743125095,
"grad_norm": 0.47735217213630676,
"learning_rate": 1.9050294632279317e-05,
"loss": 0.0422,
"step": 1760
},
{
"epoch": 0.08331077980301943,
"grad_norm": 0.3801446855068207,
"learning_rate": 1.9037284100224714e-05,
"loss": 0.0473,
"step": 1770
},
{
"epoch": 0.0837814621747879,
"grad_norm": 0.4198555052280426,
"learning_rate": 1.9024189559203193e-05,
"loss": 0.0415,
"step": 1780
},
{
"epoch": 0.08425214454655637,
"grad_norm": 0.4885464310646057,
"learning_rate": 1.9011011130939254e-05,
"loss": 0.0356,
"step": 1790
},
{
"epoch": 0.08472282691832485,
"grad_norm": 0.3393060266971588,
"learning_rate": 1.8997748937937188e-05,
"loss": 0.0383,
"step": 1800
},
{
"epoch": 0.08519350929009331,
"grad_norm": 0.46003198623657227,
"learning_rate": 1.8984403103479957e-05,
"loss": 0.0425,
"step": 1810
},
{
"epoch": 0.08566419166186179,
"grad_norm": 0.6728103756904602,
"learning_rate": 1.897097375162804e-05,
"loss": 0.0378,
"step": 1820
},
{
"epoch": 0.08613487403363025,
"grad_norm": 0.397967666387558,
"learning_rate": 1.8957461007218272e-05,
"loss": 0.0378,
"step": 1830
},
{
"epoch": 0.08660555640539873,
"grad_norm": 0.31322258710861206,
"learning_rate": 1.8943864995862692e-05,
"loss": 0.0417,
"step": 1840
},
{
"epoch": 0.08707623877716719,
"grad_norm": 0.3281812071800232,
"learning_rate": 1.8930185843947382e-05,
"loss": 0.0383,
"step": 1850
},
{
"epoch": 0.08754692114893567,
"grad_norm": 0.38150227069854736,
"learning_rate": 1.891642367863127e-05,
"loss": 0.0395,
"step": 1860
},
{
"epoch": 0.08801760352070415,
"grad_norm": 0.34693992137908936,
"learning_rate": 1.8902578627844975e-05,
"loss": 0.0424,
"step": 1870
},
{
"epoch": 0.08848828589247261,
"grad_norm": 0.3454534113407135,
"learning_rate": 1.8888650820289594e-05,
"loss": 0.0425,
"step": 1880
},
{
"epoch": 0.08895896826424109,
"grad_norm": 0.3398469388484955,
"learning_rate": 1.8874640385435515e-05,
"loss": 0.039,
"step": 1890
},
{
"epoch": 0.08942965063600955,
"grad_norm": 0.3686939477920532,
"learning_rate": 1.8860547453521232e-05,
"loss": 0.0376,
"step": 1900
},
{
"epoch": 0.08990033300777803,
"grad_norm": 0.6225370764732361,
"learning_rate": 1.8846372155552095e-05,
"loss": 0.0451,
"step": 1910
},
{
"epoch": 0.0903710153795465,
"grad_norm": 0.47314631938934326,
"learning_rate": 1.8832114623299125e-05,
"loss": 0.0411,
"step": 1920
},
{
"epoch": 0.09084169775131497,
"grad_norm": 0.38306376338005066,
"learning_rate": 1.8817774989297776e-05,
"loss": 0.0426,
"step": 1930
},
{
"epoch": 0.09131238012308344,
"grad_norm": 0.3372356593608856,
"learning_rate": 1.8803353386846708e-05,
"loss": 0.0367,
"step": 1940
},
{
"epoch": 0.09178306249485191,
"grad_norm": 0.2947300672531128,
"learning_rate": 1.878884995000654e-05,
"loss": 0.0414,
"step": 1950
},
{
"epoch": 0.09225374486662039,
"grad_norm": 0.36567285656929016,
"learning_rate": 1.8774264813598614e-05,
"loss": 0.0383,
"step": 1960
},
{
"epoch": 0.09272442723838885,
"grad_norm": 0.3237560987472534,
"learning_rate": 1.875959811320373e-05,
"loss": 0.0363,
"step": 1970
},
{
"epoch": 0.09319510961015733,
"grad_norm": 0.5219818949699402,
"learning_rate": 1.87448499851609e-05,
"loss": 0.0418,
"step": 1980
},
{
"epoch": 0.0936657919819258,
"grad_norm": 0.41103696823120117,
"learning_rate": 1.8730020566566068e-05,
"loss": 0.0412,
"step": 1990
},
{
"epoch": 0.09413647435369427,
"grad_norm": 0.43966978788375854,
"learning_rate": 1.8715109995270836e-05,
"loss": 0.0394,
"step": 2000
},
{
"epoch": 0.09460715672546274,
"grad_norm": 0.3581828474998474,
"learning_rate": 1.8700118409881198e-05,
"loss": 0.0401,
"step": 2010
},
{
"epoch": 0.0950778390972312,
"grad_norm": 0.41103577613830566,
"learning_rate": 1.8685045949756232e-05,
"loss": 0.0412,
"step": 2020
},
{
"epoch": 0.09554852146899968,
"grad_norm": 0.40771934390068054,
"learning_rate": 1.8669892755006816e-05,
"loss": 0.0349,
"step": 2030
},
{
"epoch": 0.09601920384076815,
"grad_norm": 0.34854915738105774,
"learning_rate": 1.865465896649432e-05,
"loss": 0.0336,
"step": 2040
},
{
"epoch": 0.09648988621253662,
"grad_norm": 0.4420962631702423,
"learning_rate": 1.8639344725829302e-05,
"loss": 0.0389,
"step": 2050
},
{
"epoch": 0.0969605685843051,
"grad_norm": 0.679857611656189,
"learning_rate": 1.862395017537019e-05,
"loss": 0.0383,
"step": 2060
},
{
"epoch": 0.09743125095607356,
"grad_norm": 0.3952622413635254,
"learning_rate": 1.860847545822195e-05,
"loss": 0.0343,
"step": 2070
},
{
"epoch": 0.09790193332784204,
"grad_norm": 0.3888999819755554,
"learning_rate": 1.8592920718234775e-05,
"loss": 0.0411,
"step": 2080
},
{
"epoch": 0.0983726156996105,
"grad_norm": 0.3833196461200714,
"learning_rate": 1.8577286100002723e-05,
"loss": 0.0357,
"step": 2090
},
{
"epoch": 0.09884329807137898,
"grad_norm": 0.5164761543273926,
"learning_rate": 1.8561571748862394e-05,
"loss": 0.042,
"step": 2100
},
{
"epoch": 0.09931398044314746,
"grad_norm": 0.406002402305603,
"learning_rate": 1.854577781089157e-05,
"loss": 0.0378,
"step": 2110
},
{
"epoch": 0.09978466281491592,
"grad_norm": 0.3831774592399597,
"learning_rate": 1.8529904432907858e-05,
"loss": 0.0372,
"step": 2120
},
{
"epoch": 0.1002553451866844,
"grad_norm": 0.36822307109832764,
"learning_rate": 1.8513951762467318e-05,
"loss": 0.0426,
"step": 2130
},
{
"epoch": 0.10072602755845286,
"grad_norm": 0.33442118763923645,
"learning_rate": 1.849791994786311e-05,
"loss": 0.0415,
"step": 2140
},
{
"epoch": 0.10119670993022134,
"grad_norm": 0.4427483379840851,
"learning_rate": 1.8481809138124092e-05,
"loss": 0.0354,
"step": 2150
},
{
"epoch": 0.1016673923019898,
"grad_norm": 0.3733539879322052,
"learning_rate": 1.846561948301346e-05,
"loss": 0.0402,
"step": 2160
},
{
"epoch": 0.10213807467375828,
"grad_norm": 0.35552889108657837,
"learning_rate": 1.8449351133027327e-05,
"loss": 0.0329,
"step": 2170
},
{
"epoch": 0.10260875704552676,
"grad_norm": 0.5278207659721375,
"learning_rate": 1.8433004239393353e-05,
"loss": 0.0415,
"step": 2180
},
{
"epoch": 0.10307943941729522,
"grad_norm": 0.2857244312763214,
"learning_rate": 1.8416578954069318e-05,
"loss": 0.0426,
"step": 2190
},
{
"epoch": 0.1035501217890637,
"grad_norm": 0.40666356682777405,
"learning_rate": 1.840007542974172e-05,
"loss": 0.0364,
"step": 2200
},
{
"epoch": 0.10402080416083216,
"grad_norm": 0.32413166761398315,
"learning_rate": 1.838349381982435e-05,
"loss": 0.0369,
"step": 2210
},
{
"epoch": 0.10449148653260064,
"grad_norm": 0.5573388338088989,
"learning_rate": 1.8366834278456872e-05,
"loss": 0.0401,
"step": 2220
},
{
"epoch": 0.10496216890436912,
"grad_norm": 0.396282434463501,
"learning_rate": 1.8350096960503383e-05,
"loss": 0.0337,
"step": 2230
},
{
"epoch": 0.10543285127613758,
"grad_norm": 0.32733696699142456,
"learning_rate": 1.833328202155098e-05,
"loss": 0.0355,
"step": 2240
},
{
"epoch": 0.10590353364790606,
"grad_norm": 0.3989327549934387,
"learning_rate": 1.831638961790831e-05,
"loss": 0.0364,
"step": 2250
},
{
"epoch": 0.10637421601967452,
"grad_norm": 0.34185853600502014,
"learning_rate": 1.8299419906604115e-05,
"loss": 0.0367,
"step": 2260
},
{
"epoch": 0.106844898391443,
"grad_norm": 0.3936305344104767,
"learning_rate": 1.8282373045385786e-05,
"loss": 0.0396,
"step": 2270
},
{
"epoch": 0.10731558076321146,
"grad_norm": 0.46239763498306274,
"learning_rate": 1.8265249192717868e-05,
"loss": 0.0363,
"step": 2280
},
{
"epoch": 0.10778626313497994,
"grad_norm": 0.3478502929210663,
"learning_rate": 1.8248048507780626e-05,
"loss": 0.0394,
"step": 2290
},
{
"epoch": 0.10825694550674841,
"grad_norm": 0.47366324067115784,
"learning_rate": 1.8230771150468517e-05,
"loss": 0.0405,
"step": 2300
},
{
"epoch": 0.10872762787851688,
"grad_norm": 0.38909581303596497,
"learning_rate": 1.821341728138876e-05,
"loss": 0.0334,
"step": 2310
},
{
"epoch": 0.10919831025028535,
"grad_norm": 0.363399863243103,
"learning_rate": 1.819598706185979e-05,
"loss": 0.0333,
"step": 2320
},
{
"epoch": 0.10966899262205382,
"grad_norm": 0.46871137619018555,
"learning_rate": 1.8178480653909795e-05,
"loss": 0.0378,
"step": 2330
},
{
"epoch": 0.1101396749938223,
"grad_norm": 0.29376283288002014,
"learning_rate": 1.8160898220275196e-05,
"loss": 0.0338,
"step": 2340
},
{
"epoch": 0.11061035736559076,
"grad_norm": 0.4058816432952881,
"learning_rate": 1.8143239924399127e-05,
"loss": 0.0371,
"step": 2350
},
{
"epoch": 0.11108103973735924,
"grad_norm": 0.32977092266082764,
"learning_rate": 1.8125505930429936e-05,
"loss": 0.0372,
"step": 2360
},
{
"epoch": 0.11155172210912771,
"grad_norm": 0.37261709570884705,
"learning_rate": 1.810769640321963e-05,
"loss": 0.0337,
"step": 2370
},
{
"epoch": 0.11202240448089618,
"grad_norm": 0.32283252477645874,
"learning_rate": 1.8089811508322382e-05,
"loss": 0.0367,
"step": 2380
},
{
"epoch": 0.11249308685266465,
"grad_norm": 0.31642740964889526,
"learning_rate": 1.8071851411992948e-05,
"loss": 0.0321,
"step": 2390
},
{
"epoch": 0.11296376922443312,
"grad_norm": 0.3958761990070343,
"learning_rate": 1.8053816281185154e-05,
"loss": 0.0342,
"step": 2400
},
{
"epoch": 0.1134344515962016,
"grad_norm": 0.4236665964126587,
"learning_rate": 1.803570628355033e-05,
"loss": 0.0351,
"step": 2410
},
{
"epoch": 0.11390513396797007,
"grad_norm": 0.3757830560207367,
"learning_rate": 1.801752158743576e-05,
"loss": 0.0317,
"step": 2420
},
{
"epoch": 0.11437581633973853,
"grad_norm": 0.3777036666870117,
"learning_rate": 1.7999262361883102e-05,
"loss": 0.038,
"step": 2430
},
{
"epoch": 0.11484649871150701,
"grad_norm": 0.31535372138023376,
"learning_rate": 1.7980928776626833e-05,
"loss": 0.0476,
"step": 2440
},
{
"epoch": 0.11531718108327547,
"grad_norm": 0.38056647777557373,
"learning_rate": 1.796252100209266e-05,
"loss": 0.0347,
"step": 2450
},
{
"epoch": 0.11578786345504395,
"grad_norm": 0.34918466210365295,
"learning_rate": 1.794403920939595e-05,
"loss": 0.0453,
"step": 2460
},
{
"epoch": 0.11625854582681242,
"grad_norm": 0.32286667823791504,
"learning_rate": 1.7925483570340118e-05,
"loss": 0.0356,
"step": 2470
},
{
"epoch": 0.11672922819858089,
"grad_norm": 0.3690074384212494,
"learning_rate": 1.7906854257415048e-05,
"loss": 0.0401,
"step": 2480
},
{
"epoch": 0.11719991057034937,
"grad_norm": 0.533723771572113,
"learning_rate": 1.7888151443795478e-05,
"loss": 0.0347,
"step": 2490
},
{
"epoch": 0.11767059294211783,
"grad_norm": 0.3036756217479706,
"learning_rate": 1.78693753033394e-05,
"loss": 0.0331,
"step": 2500
},
{
"epoch": 0.11814127531388631,
"grad_norm": 0.4041900336742401,
"learning_rate": 1.7850526010586437e-05,
"loss": 0.0357,
"step": 2510
},
{
"epoch": 0.11861195768565477,
"grad_norm": 0.3549298942089081,
"learning_rate": 1.7831603740756223e-05,
"loss": 0.0389,
"step": 2520
},
{
"epoch": 0.11908264005742325,
"grad_norm": 0.5169390439987183,
"learning_rate": 1.7812608669746774e-05,
"loss": 0.0366,
"step": 2530
},
{
"epoch": 0.11955332242919173,
"grad_norm": 0.30028674006462097,
"learning_rate": 1.779354097413285e-05,
"loss": 0.0361,
"step": 2540
},
{
"epoch": 0.12002400480096019,
"grad_norm": 0.3425569534301758,
"learning_rate": 1.777440083116432e-05,
"loss": 0.0328,
"step": 2550
},
{
"epoch": 0.12049468717272867,
"grad_norm": 0.44522297382354736,
"learning_rate": 1.7755188418764517e-05,
"loss": 0.0383,
"step": 2560
},
{
"epoch": 0.12096536954449713,
"grad_norm": 0.34117597341537476,
"learning_rate": 1.7735903915528553e-05,
"loss": 0.0407,
"step": 2570
},
{
"epoch": 0.12143605191626561,
"grad_norm": 0.29245781898498535,
"learning_rate": 1.7716547500721715e-05,
"loss": 0.0386,
"step": 2580
},
{
"epoch": 0.12190673428803407,
"grad_norm": 0.3196702003479004,
"learning_rate": 1.7697119354277746e-05,
"loss": 0.037,
"step": 2590
},
{
"epoch": 0.12237741665980255,
"grad_norm": 0.41022437810897827,
"learning_rate": 1.76776196567972e-05,
"loss": 0.036,
"step": 2600
},
{
"epoch": 0.12284809903157103,
"grad_norm": 0.3285203278064728,
"learning_rate": 1.7658048589545757e-05,
"loss": 0.0376,
"step": 2610
},
{
"epoch": 0.12331878140333949,
"grad_norm": 0.35540851950645447,
"learning_rate": 1.7638406334452535e-05,
"loss": 0.0324,
"step": 2620
},
{
"epoch": 0.12378946377510797,
"grad_norm": 0.3838101029396057,
"learning_rate": 1.7618693074108405e-05,
"loss": 0.0352,
"step": 2630
},
{
"epoch": 0.12426014614687643,
"grad_norm": 0.3557164669036865,
"learning_rate": 1.7598908991764288e-05,
"loss": 0.038,
"step": 2640
},
{
"epoch": 0.1247308285186449,
"grad_norm": 0.40848851203918457,
"learning_rate": 1.7579054271329457e-05,
"loss": 0.035,
"step": 2650
},
{
"epoch": 0.12520151089041337,
"grad_norm": 0.31293219327926636,
"learning_rate": 1.755912909736981e-05,
"loss": 0.0376,
"step": 2660
},
{
"epoch": 0.12567219326218185,
"grad_norm": 0.4112180471420288,
"learning_rate": 1.753913365510619e-05,
"loss": 0.0348,
"step": 2670
},
{
"epoch": 0.12614287563395032,
"grad_norm": 0.41358277201652527,
"learning_rate": 1.751906813041263e-05,
"loss": 0.0364,
"step": 2680
},
{
"epoch": 0.1266135580057188,
"grad_norm": 0.41863253712654114,
"learning_rate": 1.749893270981463e-05,
"loss": 0.0337,
"step": 2690
},
{
"epoch": 0.12708424037748725,
"grad_norm": 0.33883562684059143,
"learning_rate": 1.747872758048744e-05,
"loss": 0.0327,
"step": 2700
},
{
"epoch": 0.12755492274925573,
"grad_norm": 0.4214261472225189,
"learning_rate": 1.745845293025431e-05,
"loss": 0.0348,
"step": 2710
},
{
"epoch": 0.1280256051210242,
"grad_norm": 0.2465086430311203,
"learning_rate": 1.7438108947584737e-05,
"loss": 0.0365,
"step": 2720
},
{
"epoch": 0.12849628749279268,
"grad_norm": 0.4348239004611969,
"learning_rate": 1.7417695821592727e-05,
"loss": 0.0378,
"step": 2730
},
{
"epoch": 0.12896696986456116,
"grad_norm": 0.3477732241153717,
"learning_rate": 1.739721374203502e-05,
"loss": 0.0355,
"step": 2740
},
{
"epoch": 0.1294376522363296,
"grad_norm": 0.3776122033596039,
"learning_rate": 1.7376662899309346e-05,
"loss": 0.0353,
"step": 2750
},
{
"epoch": 0.12990833460809809,
"grad_norm": 0.30901825428009033,
"learning_rate": 1.7356043484452643e-05,
"loss": 0.035,
"step": 2760
},
{
"epoch": 0.13037901697986656,
"grad_norm": 0.3747608959674835,
"learning_rate": 1.733535568913928e-05,
"loss": 0.0338,
"step": 2770
},
{
"epoch": 0.13084969935163504,
"grad_norm": 0.30009716749191284,
"learning_rate": 1.731459970567928e-05,
"loss": 0.0316,
"step": 2780
},
{
"epoch": 0.13132038172340352,
"grad_norm": 0.38694047927856445,
"learning_rate": 1.729377572701653e-05,
"loss": 0.0343,
"step": 2790
},
{
"epoch": 0.13179106409517197,
"grad_norm": 0.40451979637145996,
"learning_rate": 1.7272883946726986e-05,
"loss": 0.0398,
"step": 2800
},
{
"epoch": 0.13226174646694044,
"grad_norm": 0.295534610748291,
"learning_rate": 1.7251924559016885e-05,
"loss": 0.0368,
"step": 2810
},
{
"epoch": 0.13273242883870892,
"grad_norm": 0.2992745637893677,
"learning_rate": 1.7230897758720916e-05,
"loss": 0.0363,
"step": 2820
},
{
"epoch": 0.1332031112104774,
"grad_norm": 0.4039139449596405,
"learning_rate": 1.720980374130044e-05,
"loss": 0.0411,
"step": 2830
},
{
"epoch": 0.13367379358224585,
"grad_norm": 0.34203460812568665,
"learning_rate": 1.7188642702841643e-05,
"loss": 0.0324,
"step": 2840
},
{
"epoch": 0.13414447595401433,
"grad_norm": 0.3400214910507202,
"learning_rate": 1.716741484005373e-05,
"loss": 0.0333,
"step": 2850
},
{
"epoch": 0.1346151583257828,
"grad_norm": 0.3761097192764282,
"learning_rate": 1.7146120350267094e-05,
"loss": 0.0363,
"step": 2860
},
{
"epoch": 0.13508584069755128,
"grad_norm": 0.3634023368358612,
"learning_rate": 1.7124759431431485e-05,
"loss": 0.0338,
"step": 2870
},
{
"epoch": 0.13555652306931976,
"grad_norm": 0.3150032162666321,
"learning_rate": 1.7103332282114156e-05,
"loss": 0.0391,
"step": 2880
},
{
"epoch": 0.1360272054410882,
"grad_norm": 0.2863038182258606,
"learning_rate": 1.7081839101498033e-05,
"loss": 0.0319,
"step": 2890
},
{
"epoch": 0.13649788781285668,
"grad_norm": 0.3113557696342468,
"learning_rate": 1.7060280089379854e-05,
"loss": 0.0326,
"step": 2900
},
{
"epoch": 0.13696857018462516,
"grad_norm": 0.38720476627349854,
"learning_rate": 1.703865544616832e-05,
"loss": 0.0357,
"step": 2910
},
{
"epoch": 0.13743925255639364,
"grad_norm": 0.4730969965457916,
"learning_rate": 1.7016965372882227e-05,
"loss": 0.0304,
"step": 2920
},
{
"epoch": 0.13790993492816211,
"grad_norm": 0.407235711812973,
"learning_rate": 1.6995210071148582e-05,
"loss": 0.0313,
"step": 2930
},
{
"epoch": 0.13838061729993056,
"grad_norm": 1.340832233428955,
"learning_rate": 1.6973389743200764e-05,
"loss": 0.0344,
"step": 2940
},
{
"epoch": 0.13885129967169904,
"grad_norm": 0.33377712965011597,
"learning_rate": 1.6951504591876614e-05,
"loss": 0.0396,
"step": 2950
},
{
"epoch": 0.13932198204346752,
"grad_norm": 0.35978761315345764,
"learning_rate": 1.692955482061656e-05,
"loss": 0.0318,
"step": 2960
},
{
"epoch": 0.139792664415236,
"grad_norm": 0.5117897987365723,
"learning_rate": 1.6907540633461728e-05,
"loss": 0.0353,
"step": 2970
},
{
"epoch": 0.14026334678700447,
"grad_norm": 0.4323929250240326,
"learning_rate": 1.6885462235052038e-05,
"loss": 0.0363,
"step": 2980
},
{
"epoch": 0.14073402915877292,
"grad_norm": 0.32677119970321655,
"learning_rate": 1.6863319830624313e-05,
"loss": 0.0402,
"step": 2990
},
{
"epoch": 0.1412047115305414,
"grad_norm": 0.3986310362815857,
"learning_rate": 1.6841113626010358e-05,
"loss": 0.0385,
"step": 3000
},
{
"epoch": 0.14167539390230988,
"grad_norm": 1.4817543029785156,
"learning_rate": 1.6818843827635052e-05,
"loss": 0.0335,
"step": 3010
},
{
"epoch": 0.14214607627407835,
"grad_norm": 0.3320419490337372,
"learning_rate": 1.679651064251444e-05,
"loss": 0.0333,
"step": 3020
},
{
"epoch": 0.1426167586458468,
"grad_norm": 0.3239348232746124,
"learning_rate": 1.677411427825379e-05,
"loss": 0.0324,
"step": 3030
},
{
"epoch": 0.14308744101761528,
"grad_norm": 0.279315322637558,
"learning_rate": 1.6751654943045672e-05,
"loss": 0.0326,
"step": 3040
},
{
"epoch": 0.14355812338938376,
"grad_norm": 0.40376725792884827,
"learning_rate": 1.672913284566803e-05,
"loss": 0.0324,
"step": 3050
},
{
"epoch": 0.14402880576115223,
"grad_norm": 0.30471786856651306,
"learning_rate": 1.6706548195482222e-05,
"loss": 0.0301,
"step": 3060
},
{
"epoch": 0.1444994881329207,
"grad_norm": 0.2932983636856079,
"learning_rate": 1.66839012024311e-05,
"loss": 0.0328,
"step": 3070
},
{
"epoch": 0.14497017050468916,
"grad_norm": 0.31888076663017273,
"learning_rate": 1.666119207703703e-05,
"loss": 0.0327,
"step": 3080
},
{
"epoch": 0.14544085287645764,
"grad_norm": 0.33213096857070923,
"learning_rate": 1.6638421030399962e-05,
"loss": 0.0347,
"step": 3090
},
{
"epoch": 0.14591153524822612,
"grad_norm": 0.4190642833709717,
"learning_rate": 1.6615588274195445e-05,
"loss": 0.0351,
"step": 3100
},
{
"epoch": 0.1463822176199946,
"grad_norm": 0.3590419590473175,
"learning_rate": 1.6592694020672667e-05,
"loss": 0.0356,
"step": 3110
},
{
"epoch": 0.14685289999176307,
"grad_norm": 0.3238411545753479,
"learning_rate": 1.65697384826525e-05,
"loss": 0.0308,
"step": 3120
},
{
"epoch": 0.14732358236353152,
"grad_norm": 0.3472120463848114,
"learning_rate": 1.6546721873525488e-05,
"loss": 0.0336,
"step": 3130
},
{
"epoch": 0.1477942647353,
"grad_norm": 0.2819247841835022,
"learning_rate": 1.6523644407249893e-05,
"loss": 0.0304,
"step": 3140
},
{
"epoch": 0.14826494710706847,
"grad_norm": 0.3401038944721222,
"learning_rate": 1.6500506298349682e-05,
"loss": 0.0368,
"step": 3150
},
{
"epoch": 0.14873562947883695,
"grad_norm": 0.35624802112579346,
"learning_rate": 1.6477307761912555e-05,
"loss": 0.031,
"step": 3160
},
{
"epoch": 0.14920631185060543,
"grad_norm": 0.3220632076263428,
"learning_rate": 1.645404901358794e-05,
"loss": 0.0273,
"step": 3170
},
{
"epoch": 0.14967699422237388,
"grad_norm": 0.2849304676055908,
"learning_rate": 1.6430730269584963e-05,
"loss": 0.0289,
"step": 3180
},
{
"epoch": 0.15014767659414235,
"grad_norm": 0.3250783681869507,
"learning_rate": 1.6407351746670484e-05,
"loss": 0.0292,
"step": 3190
},
{
"epoch": 0.15061835896591083,
"grad_norm": 0.430021733045578,
"learning_rate": 1.638391366216704e-05,
"loss": 0.0288,
"step": 3200
},
{
"epoch": 0.1510890413376793,
"grad_norm": 0.25285810232162476,
"learning_rate": 1.636041623395085e-05,
"loss": 0.0314,
"step": 3210
},
{
"epoch": 0.15155972370944779,
"grad_norm": 0.3291570544242859,
"learning_rate": 1.6336859680449773e-05,
"loss": 0.0324,
"step": 3220
},
{
"epoch": 0.15203040608121624,
"grad_norm": 0.3136582374572754,
"learning_rate": 1.6313244220641304e-05,
"loss": 0.0325,
"step": 3230
},
{
"epoch": 0.1525010884529847,
"grad_norm": 0.3587929606437683,
"learning_rate": 1.6289570074050492e-05,
"loss": 0.0339,
"step": 3240
},
{
"epoch": 0.1529717708247532,
"grad_norm": 0.31578192114830017,
"learning_rate": 1.626583746074796e-05,
"loss": 0.0334,
"step": 3250
},
{
"epoch": 0.15344245319652167,
"grad_norm": 0.28628358244895935,
"learning_rate": 1.6242046601347796e-05,
"loss": 0.0299,
"step": 3260
},
{
"epoch": 0.15391313556829012,
"grad_norm": 0.362964928150177,
"learning_rate": 1.6218197717005562e-05,
"loss": 0.0323,
"step": 3270
},
{
"epoch": 0.1543838179400586,
"grad_norm": 0.29254281520843506,
"learning_rate": 1.6194291029416188e-05,
"loss": 0.0345,
"step": 3280
},
{
"epoch": 0.15485450031182707,
"grad_norm": 0.32078689336776733,
"learning_rate": 1.617032676081194e-05,
"loss": 0.0361,
"step": 3290
},
{
"epoch": 0.15532518268359555,
"grad_norm": 0.3560759127140045,
"learning_rate": 1.614630513396035e-05,
"loss": 0.0355,
"step": 3300
},
{
"epoch": 0.15579586505536402,
"grad_norm": 0.3040817081928253,
"learning_rate": 1.6122226372162137e-05,
"loss": 0.0318,
"step": 3310
},
{
"epoch": 0.15626654742713247,
"grad_norm": 0.2728009521961212,
"learning_rate": 1.6098090699249144e-05,
"loss": 0.0365,
"step": 3320
},
{
"epoch": 0.15673722979890095,
"grad_norm": 0.43440693616867065,
"learning_rate": 1.607389833958223e-05,
"loss": 0.0333,
"step": 3330
},
{
"epoch": 0.15720791217066943,
"grad_norm": 0.27971121668815613,
"learning_rate": 1.6049649518049234e-05,
"loss": 0.0339,
"step": 3340
},
{
"epoch": 0.1576785945424379,
"grad_norm": 0.28107643127441406,
"learning_rate": 1.6025344460062826e-05,
"loss": 0.0319,
"step": 3350
},
{
"epoch": 0.15814927691420638,
"grad_norm": 0.30143097043037415,
"learning_rate": 1.6000983391558457e-05,
"loss": 0.0349,
"step": 3360
},
{
"epoch": 0.15861995928597483,
"grad_norm": 0.3217163383960724,
"learning_rate": 1.5976566538992237e-05,
"loss": 0.0325,
"step": 3370
},
{
"epoch": 0.1590906416577433,
"grad_norm": 0.2864225506782532,
"learning_rate": 1.5952094129338834e-05,
"loss": 0.031,
"step": 3380
},
{
"epoch": 0.1595613240295118,
"grad_norm": 0.4295887351036072,
"learning_rate": 1.5927566390089362e-05,
"loss": 0.03,
"step": 3390
},
{
"epoch": 0.16003200640128026,
"grad_norm": 0.33816084265708923,
"learning_rate": 1.5902983549249272e-05,
"loss": 0.0412,
"step": 3400
},
{
"epoch": 0.16050268877304874,
"grad_norm": 0.3602800965309143,
"learning_rate": 1.5878345835336232e-05,
"loss": 0.0277,
"step": 3410
},
{
"epoch": 0.1609733711448172,
"grad_norm": 0.3140107989311218,
"learning_rate": 1.5853653477377996e-05,
"loss": 0.0312,
"step": 3420
},
{
"epoch": 0.16144405351658567,
"grad_norm": 0.2669489085674286,
"learning_rate": 1.582890670491028e-05,
"loss": 0.0309,
"step": 3430
},
{
"epoch": 0.16191473588835414,
"grad_norm": 0.3519858121871948,
"learning_rate": 1.5804105747974626e-05,
"loss": 0.0339,
"step": 3440
},
{
"epoch": 0.16238541826012262,
"grad_norm": 0.2792114019393921,
"learning_rate": 1.5779250837116275e-05,
"loss": 0.0294,
"step": 3450
},
{
"epoch": 0.16285610063189107,
"grad_norm": 0.2898792624473572,
"learning_rate": 1.5754342203382003e-05,
"loss": 0.0334,
"step": 3460
},
{
"epoch": 0.16332678300365955,
"grad_norm": 0.24655383825302124,
"learning_rate": 1.5729380078317982e-05,
"loss": 0.0314,
"step": 3470
},
{
"epoch": 0.16379746537542803,
"grad_norm": 0.3090629279613495,
"learning_rate": 1.570436469396764e-05,
"loss": 0.0316,
"step": 3480
},
{
"epoch": 0.1642681477471965,
"grad_norm": 0.42195451259613037,
"learning_rate": 1.567929628286949e-05,
"loss": 0.0345,
"step": 3490
},
{
"epoch": 0.16473883011896498,
"grad_norm": 0.31561776995658875,
"learning_rate": 1.5654175078054965e-05,
"loss": 0.0336,
"step": 3500
},
{
"epoch": 0.16520951249073343,
"grad_norm": 0.3707563281059265,
"learning_rate": 1.562900131304627e-05,
"loss": 0.0337,
"step": 3510
},
{
"epoch": 0.1656801948625019,
"grad_norm": 0.27753663063049316,
"learning_rate": 1.5603775221854195e-05,
"loss": 0.034,
"step": 3520
},
{
"epoch": 0.16615087723427038,
"grad_norm": 0.39314696192741394,
"learning_rate": 1.557849703897594e-05,
"loss": 0.0328,
"step": 3530
},
{
"epoch": 0.16662155960603886,
"grad_norm": 0.2622597813606262,
"learning_rate": 1.5553166999392954e-05,
"loss": 0.0332,
"step": 3540
},
{
"epoch": 0.16709224197780734,
"grad_norm": 0.29595157504081726,
"learning_rate": 1.5527785338568718e-05,
"loss": 0.0318,
"step": 3550
},
{
"epoch": 0.1675629243495758,
"grad_norm": 0.26845335960388184,
"learning_rate": 1.550235229244659e-05,
"loss": 0.0308,
"step": 3560
},
{
"epoch": 0.16803360672134426,
"grad_norm": 0.30363011360168457,
"learning_rate": 1.5476868097447586e-05,
"loss": 0.0336,
"step": 3570
},
{
"epoch": 0.16850428909311274,
"grad_norm": 0.3513653576374054,
"learning_rate": 1.5451332990468202e-05,
"loss": 0.0322,
"step": 3580
},
{
"epoch": 0.16897497146488122,
"grad_norm": 0.4145563840866089,
"learning_rate": 1.5425747208878195e-05,
"loss": 0.0334,
"step": 3590
},
{
"epoch": 0.1694456538366497,
"grad_norm": 0.29433363676071167,
"learning_rate": 1.5400110990518386e-05,
"loss": 0.0331,
"step": 3600
},
{
"epoch": 0.16991633620841815,
"grad_norm": 0.32831501960754395,
"learning_rate": 1.5374424573698453e-05,
"loss": 0.0312,
"step": 3610
},
{
"epoch": 0.17038701858018662,
"grad_norm": 0.3588094711303711,
"learning_rate": 1.5348688197194696e-05,
"loss": 0.0306,
"step": 3620
},
{
"epoch": 0.1708577009519551,
"grad_norm": 0.39610594511032104,
"learning_rate": 1.532290210024785e-05,
"loss": 0.0306,
"step": 3630
},
{
"epoch": 0.17132838332372358,
"grad_norm": 0.3050280213356018,
"learning_rate": 1.529706652256083e-05,
"loss": 0.0335,
"step": 3640
},
{
"epoch": 0.17179906569549203,
"grad_norm": 0.4129210114479065,
"learning_rate": 1.5271181704296513e-05,
"loss": 0.0299,
"step": 3650
},
{
"epoch": 0.1722697480672605,
"grad_norm": 0.2414814978837967,
"learning_rate": 1.5245247886075518e-05,
"loss": 0.0308,
"step": 3660
},
{
"epoch": 0.17274043043902898,
"grad_norm": 0.30157119035720825,
"learning_rate": 1.5219265308973952e-05,
"loss": 0.0254,
"step": 3670
},
{
"epoch": 0.17321111281079746,
"grad_norm": 0.47096705436706543,
"learning_rate": 1.519323421452117e-05,
"loss": 0.0382,
"step": 3680
},
{
"epoch": 0.17368179518256593,
"grad_norm": 0.3166336715221405,
"learning_rate": 1.5167154844697549e-05,
"loss": 0.0285,
"step": 3690
},
{
"epoch": 0.17415247755433438,
"grad_norm": 0.3827841281890869,
"learning_rate": 1.5141027441932217e-05,
"loss": 0.0315,
"step": 3700
},
{
"epoch": 0.17462315992610286,
"grad_norm": 0.3364233672618866,
"learning_rate": 1.5114852249100811e-05,
"loss": 0.033,
"step": 3710
},
{
"epoch": 0.17509384229787134,
"grad_norm": 0.30109116435050964,
"learning_rate": 1.5088629509523207e-05,
"loss": 0.0343,
"step": 3720
},
{
"epoch": 0.17556452466963982,
"grad_norm": 0.4630192816257477,
"learning_rate": 1.5062359466961283e-05,
"loss": 0.0303,
"step": 3730
},
{
"epoch": 0.1760352070414083,
"grad_norm": 0.2511216402053833,
"learning_rate": 1.5036042365616621e-05,
"loss": 0.0306,
"step": 3740
},
{
"epoch": 0.17650588941317674,
"grad_norm": 0.41122281551361084,
"learning_rate": 1.5009678450128263e-05,
"loss": 0.0325,
"step": 3750
},
{
"epoch": 0.17697657178494522,
"grad_norm": 0.40054771304130554,
"learning_rate": 1.498326796557042e-05,
"loss": 0.0312,
"step": 3760
},
{
"epoch": 0.1774472541567137,
"grad_norm": 0.38717585802078247,
"learning_rate": 1.495681115745021e-05,
"loss": 0.0332,
"step": 3770
},
{
"epoch": 0.17791793652848217,
"grad_norm": 0.24846479296684265,
"learning_rate": 1.4930308271705357e-05,
"loss": 0.032,
"step": 3780
},
{
"epoch": 0.17838861890025065,
"grad_norm": 0.3396732211112976,
"learning_rate": 1.4903759554701922e-05,
"loss": 0.0292,
"step": 3790
},
{
"epoch": 0.1788593012720191,
"grad_norm": 0.32007864117622375,
"learning_rate": 1.4877165253231995e-05,
"loss": 0.0292,
"step": 3800
},
{
"epoch": 0.17932998364378758,
"grad_norm": 0.3790198862552643,
"learning_rate": 1.4850525614511427e-05,
"loss": 0.0345,
"step": 3810
},
{
"epoch": 0.17980066601555605,
"grad_norm": 0.2988254725933075,
"learning_rate": 1.4823840886177494e-05,
"loss": 0.0326,
"step": 3820
},
{
"epoch": 0.18027134838732453,
"grad_norm": 0.3227101266384125,
"learning_rate": 1.4797111316286639e-05,
"loss": 0.036,
"step": 3830
},
{
"epoch": 0.180742030759093,
"grad_norm": 0.3604857325553894,
"learning_rate": 1.4770337153312131e-05,
"loss": 0.0281,
"step": 3840
},
{
"epoch": 0.18121271313086146,
"grad_norm": 0.340345174074173,
"learning_rate": 1.474351864614177e-05,
"loss": 0.0295,
"step": 3850
},
{
"epoch": 0.18168339550262994,
"grad_norm": 0.3362833261489868,
"learning_rate": 1.4716656044075577e-05,
"loss": 0.0324,
"step": 3860
},
{
"epoch": 0.1821540778743984,
"grad_norm": 0.3097628653049469,
"learning_rate": 1.468974959682346e-05,
"loss": 0.0347,
"step": 3870
},
{
"epoch": 0.1826247602461669,
"grad_norm": 0.3956356942653656,
"learning_rate": 1.466279955450292e-05,
"loss": 0.0333,
"step": 3880
},
{
"epoch": 0.18309544261793534,
"grad_norm": 0.3538338840007782,
"learning_rate": 1.4635806167636698e-05,
"loss": 0.0308,
"step": 3890
},
{
"epoch": 0.18356612498970382,
"grad_norm": 0.33104780316352844,
"learning_rate": 1.4608769687150459e-05,
"loss": 0.0349,
"step": 3900
},
{
"epoch": 0.1840368073614723,
"grad_norm": 0.28109461069107056,
"learning_rate": 1.4581690364370466e-05,
"loss": 0.0334,
"step": 3910
},
{
"epoch": 0.18450748973324077,
"grad_norm": 0.27506646513938904,
"learning_rate": 1.455456845102123e-05,
"loss": 0.0318,
"step": 3920
},
{
"epoch": 0.18497817210500925,
"grad_norm": 0.38008806109428406,
"learning_rate": 1.4527404199223173e-05,
"loss": 0.0343,
"step": 3930
},
{
"epoch": 0.1854488544767777,
"grad_norm": 0.3823818266391754,
"learning_rate": 1.4500197861490293e-05,
"loss": 0.0345,
"step": 3940
},
{
"epoch": 0.18591953684854617,
"grad_norm": 0.2571784257888794,
"learning_rate": 1.4472949690727813e-05,
"loss": 0.0288,
"step": 3950
},
{
"epoch": 0.18639021922031465,
"grad_norm": 0.2995915710926056,
"learning_rate": 1.4445659940229827e-05,
"loss": 0.0323,
"step": 3960
},
{
"epoch": 0.18686090159208313,
"grad_norm": 0.3813773989677429,
"learning_rate": 1.441832886367694e-05,
"loss": 0.0358,
"step": 3970
},
{
"epoch": 0.1873315839638516,
"grad_norm": 0.3301527202129364,
"learning_rate": 1.4390956715133928e-05,
"loss": 0.031,
"step": 3980
},
{
"epoch": 0.18780226633562006,
"grad_norm": 0.36864516139030457,
"learning_rate": 1.4363543749047354e-05,
"loss": 0.032,
"step": 3990
},
{
"epoch": 0.18827294870738853,
"grad_norm": 0.3669602870941162,
"learning_rate": 1.4336090220243222e-05,
"loss": 0.0257,
"step": 4000
},
{
"epoch": 0.188743631079157,
"grad_norm": 0.27271798253059387,
"learning_rate": 1.4308596383924593e-05,
"loss": 0.0323,
"step": 4010
},
{
"epoch": 0.1892143134509255,
"grad_norm": 0.42792803049087524,
"learning_rate": 1.4281062495669224e-05,
"loss": 0.0348,
"step": 4020
},
{
"epoch": 0.18968499582269396,
"grad_norm": 0.2788338363170624,
"learning_rate": 1.4253488811427188e-05,
"loss": 0.0304,
"step": 4030
},
{
"epoch": 0.1901556781944624,
"grad_norm": 0.24826385080814362,
"learning_rate": 1.4225875587518485e-05,
"loss": 0.0341,
"step": 4040
},
{
"epoch": 0.1906263605662309,
"grad_norm": 0.3821421265602112,
"learning_rate": 1.4198223080630686e-05,
"loss": 0.0335,
"step": 4050
},
{
"epoch": 0.19109704293799937,
"grad_norm": 0.44902199506759644,
"learning_rate": 1.4170531547816513e-05,
"loss": 0.0285,
"step": 4060
},
{
"epoch": 0.19156772530976784,
"grad_norm": 0.32608562707901,
"learning_rate": 1.4142801246491476e-05,
"loss": 0.0306,
"step": 4070
},
{
"epoch": 0.1920384076815363,
"grad_norm": 0.3407364785671234,
"learning_rate": 1.4115032434431461e-05,
"loss": 0.0296,
"step": 4080
},
{
"epoch": 0.19250909005330477,
"grad_norm": 0.26627030968666077,
"learning_rate": 1.4087225369770356e-05,
"loss": 0.0267,
"step": 4090
},
{
"epoch": 0.19297977242507325,
"grad_norm": 0.47843673825263977,
"learning_rate": 1.4059380310997626e-05,
"loss": 0.0305,
"step": 4100
},
{
"epoch": 0.19345045479684173,
"grad_norm": 0.3628087043762207,
"learning_rate": 1.403149751695593e-05,
"loss": 0.0316,
"step": 4110
},
{
"epoch": 0.1939211371686102,
"grad_norm": 0.320494145154953,
"learning_rate": 1.40035772468387e-05,
"loss": 0.0329,
"step": 4120
},
{
"epoch": 0.19439181954037865,
"grad_norm": 0.3068598806858063,
"learning_rate": 1.3975619760187746e-05,
"loss": 0.0307,
"step": 4130
},
{
"epoch": 0.19486250191214713,
"grad_norm": 0.2807197570800781,
"learning_rate": 1.3947625316890836e-05,
"loss": 0.0329,
"step": 4140
},
{
"epoch": 0.1953331842839156,
"grad_norm": 0.30862492322921753,
"learning_rate": 1.3919594177179272e-05,
"loss": 0.0329,
"step": 4150
},
{
"epoch": 0.19580386665568408,
"grad_norm": 0.40224847197532654,
"learning_rate": 1.3891526601625492e-05,
"loss": 0.0284,
"step": 4160
},
{
"epoch": 0.19627454902745256,
"grad_norm": 0.2759397327899933,
"learning_rate": 1.3863422851140624e-05,
"loss": 0.0315,
"step": 4170
},
{
"epoch": 0.196745231399221,
"grad_norm": 0.3810548782348633,
"learning_rate": 1.3835283186972077e-05,
"loss": 0.0326,
"step": 4180
},
{
"epoch": 0.1972159137709895,
"grad_norm": 0.270430326461792,
"learning_rate": 1.3807107870701102e-05,
"loss": 0.0296,
"step": 4190
},
{
"epoch": 0.19768659614275796,
"grad_norm": 0.28438347578048706,
"learning_rate": 1.3778897164240378e-05,
"loss": 0.029,
"step": 4200
},
{
"epoch": 0.19815727851452644,
"grad_norm": 0.38514113426208496,
"learning_rate": 1.3750651329831548e-05,
"loss": 0.0246,
"step": 4210
},
{
"epoch": 0.19862796088629492,
"grad_norm": 0.3501206040382385,
"learning_rate": 1.3722370630042809e-05,
"loss": 0.0276,
"step": 4220
},
{
"epoch": 0.19909864325806337,
"grad_norm": 0.247554212808609,
"learning_rate": 1.369405532776646e-05,
"loss": 0.0265,
"step": 4230
},
{
"epoch": 0.19956932562983185,
"grad_norm": 0.27596163749694824,
"learning_rate": 1.3665705686216457e-05,
"loss": 0.0296,
"step": 4240
},
{
"epoch": 0.20004000800160032,
"grad_norm": 0.3220009505748749,
"learning_rate": 1.3637321968925964e-05,
"loss": 0.0306,
"step": 4250
},
{
"epoch": 0.2005106903733688,
"grad_norm": 0.36187198758125305,
"learning_rate": 1.3608904439744905e-05,
"loss": 0.0361,
"step": 4260
},
{
"epoch": 0.20098137274513725,
"grad_norm": 0.386277973651886,
"learning_rate": 1.3580453362837527e-05,
"loss": 0.03,
"step": 4270
},
{
"epoch": 0.20145205511690573,
"grad_norm": 0.2790426015853882,
"learning_rate": 1.355196900267992e-05,
"loss": 0.0289,
"step": 4280
},
{
"epoch": 0.2019227374886742,
"grad_norm": 0.2859556972980499,
"learning_rate": 1.3523451624057566e-05,
"loss": 0.0276,
"step": 4290
},
{
"epoch": 0.20239341986044268,
"grad_norm": 0.3090481460094452,
"learning_rate": 1.3494901492062889e-05,
"loss": 0.0346,
"step": 4300
},
{
"epoch": 0.20286410223221116,
"grad_norm": 0.3289349675178528,
"learning_rate": 1.346631887209278e-05,
"loss": 0.0279,
"step": 4310
},
{
"epoch": 0.2033347846039796,
"grad_norm": 0.3103124797344208,
"learning_rate": 1.343770402984613e-05,
"loss": 0.0316,
"step": 4320
},
{
"epoch": 0.20380546697574808,
"grad_norm": 0.3548365533351898,
"learning_rate": 1.3409057231321363e-05,
"loss": 0.0318,
"step": 4330
},
{
"epoch": 0.20427614934751656,
"grad_norm": 0.2642829418182373,
"learning_rate": 1.3380378742813964e-05,
"loss": 0.0325,
"step": 4340
},
{
"epoch": 0.20474683171928504,
"grad_norm": 0.2960059642791748,
"learning_rate": 1.3351668830914004e-05,
"loss": 0.0251,
"step": 4350
},
{
"epoch": 0.20521751409105352,
"grad_norm": 0.32227885723114014,
"learning_rate": 1.3322927762503656e-05,
"loss": 0.0266,
"step": 4360
},
{
"epoch": 0.20568819646282197,
"grad_norm": 0.34864169359207153,
"learning_rate": 1.329415580475472e-05,
"loss": 0.0258,
"step": 4370
},
{
"epoch": 0.20615887883459044,
"grad_norm": 0.2689371407032013,
"learning_rate": 1.3265353225126143e-05,
"loss": 0.0309,
"step": 4380
},
{
"epoch": 0.20662956120635892,
"grad_norm": 0.2662619948387146,
"learning_rate": 1.3236520291361516e-05,
"loss": 0.0294,
"step": 4390
},
{
"epoch": 0.2071002435781274,
"grad_norm": 0.29656073451042175,
"learning_rate": 1.3207657271486607e-05,
"loss": 0.0299,
"step": 4400
},
{
"epoch": 0.20757092594989587,
"grad_norm": 0.28156155347824097,
"learning_rate": 1.3178764433806858e-05,
"loss": 0.0297,
"step": 4410
},
{
"epoch": 0.20804160832166432,
"grad_norm": 0.24518205225467682,
"learning_rate": 1.3149842046904885e-05,
"loss": 0.032,
"step": 4420
},
{
"epoch": 0.2085122906934328,
"grad_norm": 0.30979758501052856,
"learning_rate": 1.3120890379637996e-05,
"loss": 0.0283,
"step": 4430
},
{
"epoch": 0.20898297306520128,
"grad_norm": 0.298027366399765,
"learning_rate": 1.3091909701135676e-05,
"loss": 0.0342,
"step": 4440
},
{
"epoch": 0.20945365543696975,
"grad_norm": 0.4737192392349243,
"learning_rate": 1.3062900280797104e-05,
"loss": 0.0325,
"step": 4450
},
{
"epoch": 0.20992433780873823,
"grad_norm": 0.39332106709480286,
"learning_rate": 1.3033862388288628e-05,
"loss": 0.0295,
"step": 4460
},
{
"epoch": 0.21039502018050668,
"grad_norm": 0.35261037945747375,
"learning_rate": 1.3004796293541269e-05,
"loss": 0.0306,
"step": 4470
},
{
"epoch": 0.21086570255227516,
"grad_norm": 0.3438118100166321,
"learning_rate": 1.297570226674822e-05,
"loss": 0.0265,
"step": 4480
},
{
"epoch": 0.21133638492404364,
"grad_norm": 0.3694635033607483,
"learning_rate": 1.294658057836232e-05,
"loss": 0.0323,
"step": 4490
},
{
"epoch": 0.2118070672958121,
"grad_norm": 0.29527580738067627,
"learning_rate": 1.2917431499093538e-05,
"loss": 0.0303,
"step": 4500
},
{
"epoch": 0.21227774966758056,
"grad_norm": 0.31543228030204773,
"learning_rate": 1.288825529990647e-05,
"loss": 0.0326,
"step": 4510
},
{
"epoch": 0.21274843203934904,
"grad_norm": 0.30861562490463257,
"learning_rate": 1.2859052252017824e-05,
"loss": 0.0277,
"step": 4520
},
{
"epoch": 0.21321911441111752,
"grad_norm": 0.3934938609600067,
"learning_rate": 1.2829822626893867e-05,
"loss": 0.0307,
"step": 4530
},
{
"epoch": 0.213689796782886,
"grad_norm": 0.36799925565719604,
"learning_rate": 1.2800566696247943e-05,
"loss": 0.0267,
"step": 4540
},
{
"epoch": 0.21416047915465447,
"grad_norm": 0.3517569899559021,
"learning_rate": 1.2771284732037912e-05,
"loss": 0.0313,
"step": 4550
},
{
"epoch": 0.21463116152642292,
"grad_norm": 0.3911064863204956,
"learning_rate": 1.274197700646365e-05,
"loss": 0.0302,
"step": 4560
},
{
"epoch": 0.2151018438981914,
"grad_norm": 0.3876083493232727,
"learning_rate": 1.2712643791964501e-05,
"loss": 0.0291,
"step": 4570
},
{
"epoch": 0.21557252626995987,
"grad_norm": 0.2826777398586273,
"learning_rate": 1.2683285361216745e-05,
"loss": 0.0248,
"step": 4580
},
{
"epoch": 0.21604320864172835,
"grad_norm": 0.2614319622516632,
"learning_rate": 1.2653901987131074e-05,
"loss": 0.0308,
"step": 4590
},
{
"epoch": 0.21651389101349683,
"grad_norm": 0.3779659867286682,
"learning_rate": 1.262449394285005e-05,
"loss": 0.029,
"step": 4600
},
{
"epoch": 0.21698457338526528,
"grad_norm": 0.2700081169605255,
"learning_rate": 1.2595061501745556e-05,
"loss": 0.0352,
"step": 4610
},
{
"epoch": 0.21745525575703376,
"grad_norm": 0.23789380490779877,
"learning_rate": 1.2565604937416267e-05,
"loss": 0.0231,
"step": 4620
},
{
"epoch": 0.21792593812880223,
"grad_norm": 0.2613758146762848,
"learning_rate": 1.2536124523685114e-05,
"loss": 0.0276,
"step": 4630
},
{
"epoch": 0.2183966205005707,
"grad_norm": 0.3436334729194641,
"learning_rate": 1.2506620534596711e-05,
"loss": 0.0311,
"step": 4640
},
{
"epoch": 0.2188673028723392,
"grad_norm": 0.30586010217666626,
"learning_rate": 1.247709324441483e-05,
"loss": 0.0256,
"step": 4650
},
{
"epoch": 0.21933798524410764,
"grad_norm": 0.33210545778274536,
"learning_rate": 1.2447542927619857e-05,
"loss": 0.028,
"step": 4660
},
{
"epoch": 0.2198086676158761,
"grad_norm": 0.3382388651371002,
"learning_rate": 1.2417969858906214e-05,
"loss": 0.0281,
"step": 4670
},
{
"epoch": 0.2202793499876446,
"grad_norm": 0.25897398591041565,
"learning_rate": 1.2388374313179828e-05,
"loss": 0.0304,
"step": 4680
},
{
"epoch": 0.22075003235941307,
"grad_norm": 0.2604251205921173,
"learning_rate": 1.2358756565555563e-05,
"loss": 0.0274,
"step": 4690
},
{
"epoch": 0.22122071473118152,
"grad_norm": 0.34708860516548157,
"learning_rate": 1.2329116891354677e-05,
"loss": 0.0293,
"step": 4700
},
{
"epoch": 0.22169139710295,
"grad_norm": 0.3701171278953552,
"learning_rate": 1.2299455566102248e-05,
"loss": 0.0298,
"step": 4710
},
{
"epoch": 0.22216207947471847,
"grad_norm": 0.3084346652030945,
"learning_rate": 1.2269772865524612e-05,
"loss": 0.0273,
"step": 4720
},
{
"epoch": 0.22263276184648695,
"grad_norm": 0.3036332130432129,
"learning_rate": 1.2240069065546823e-05,
"loss": 0.0333,
"step": 4730
},
{
"epoch": 0.22310344421825543,
"grad_norm": 0.26620161533355713,
"learning_rate": 1.2210344442290054e-05,
"loss": 0.0257,
"step": 4740
},
{
"epoch": 0.22357412659002388,
"grad_norm": 0.4279276132583618,
"learning_rate": 1.2180599272069058e-05,
"loss": 0.0301,
"step": 4750
},
{
"epoch": 0.22404480896179235,
"grad_norm": 0.38443100452423096,
"learning_rate": 1.215083383138958e-05,
"loss": 0.0255,
"step": 4760
},
{
"epoch": 0.22451549133356083,
"grad_norm": 0.3349287509918213,
"learning_rate": 1.2121048396945807e-05,
"loss": 0.0316,
"step": 4770
},
{
"epoch": 0.2249861737053293,
"grad_norm": 0.27736541628837585,
"learning_rate": 1.2091243245617774e-05,
"loss": 0.0326,
"step": 4780
},
{
"epoch": 0.22545685607709778,
"grad_norm": 0.3288596272468567,
"learning_rate": 1.2061418654468808e-05,
"loss": 0.0274,
"step": 4790
},
{
"epoch": 0.22592753844886623,
"grad_norm": 0.319358766078949,
"learning_rate": 1.203157490074294e-05,
"loss": 0.0315,
"step": 4800
},
{
"epoch": 0.2263982208206347,
"grad_norm": 0.28299087285995483,
"learning_rate": 1.2001712261862335e-05,
"loss": 0.0328,
"step": 4810
},
{
"epoch": 0.2268689031924032,
"grad_norm": 0.36975643038749695,
"learning_rate": 1.1971831015424713e-05,
"loss": 0.0332,
"step": 4820
},
{
"epoch": 0.22733958556417166,
"grad_norm": 0.2754369080066681,
"learning_rate": 1.194193143920076e-05,
"loss": 0.0274,
"step": 4830
},
{
"epoch": 0.22781026793594014,
"grad_norm": 0.29628682136535645,
"learning_rate": 1.1912013811131562e-05,
"loss": 0.0263,
"step": 4840
},
{
"epoch": 0.2282809503077086,
"grad_norm": 0.2725105285644531,
"learning_rate": 1.1882078409326003e-05,
"loss": 0.025,
"step": 4850
},
{
"epoch": 0.22875163267947707,
"grad_norm": 0.34257981181144714,
"learning_rate": 1.1852125512058194e-05,
"loss": 0.0363,
"step": 4860
},
{
"epoch": 0.22922231505124555,
"grad_norm": 0.2495962530374527,
"learning_rate": 1.1822155397764873e-05,
"loss": 0.0294,
"step": 4870
},
{
"epoch": 0.22969299742301402,
"grad_norm": 0.4353019893169403,
"learning_rate": 1.179216834504284e-05,
"loss": 0.0282,
"step": 4880
},
{
"epoch": 0.2301636797947825,
"grad_norm": 0.25559765100479126,
"learning_rate": 1.1762164632646334e-05,
"loss": 0.0286,
"step": 4890
},
{
"epoch": 0.23063436216655095,
"grad_norm": 0.20971710979938507,
"learning_rate": 1.1732144539484467e-05,
"loss": 0.0264,
"step": 4900
},
{
"epoch": 0.23110504453831943,
"grad_norm": 0.31122690439224243,
"learning_rate": 1.1702108344618627e-05,
"loss": 0.027,
"step": 4910
},
{
"epoch": 0.2315757269100879,
"grad_norm": 0.3844875693321228,
"learning_rate": 1.1672056327259876e-05,
"loss": 0.0305,
"step": 4920
},
{
"epoch": 0.23204640928185638,
"grad_norm": 0.26517409086227417,
"learning_rate": 1.1641988766766359e-05,
"loss": 0.0231,
"step": 4930
},
{
"epoch": 0.23251709165362483,
"grad_norm": 0.35916706919670105,
"learning_rate": 1.1611905942640707e-05,
"loss": 0.0259,
"step": 4940
},
{
"epoch": 0.2329877740253933,
"grad_norm": 0.20803475379943848,
"learning_rate": 1.1581808134527443e-05,
"loss": 0.0295,
"step": 4950
},
{
"epoch": 0.23345845639716178,
"grad_norm": 0.24563108384609222,
"learning_rate": 1.1551695622210377e-05,
"loss": 0.0323,
"step": 4960
},
{
"epoch": 0.23392913876893026,
"grad_norm": 0.2711374759674072,
"learning_rate": 1.1521568685610003e-05,
"loss": 0.0267,
"step": 4970
},
{
"epoch": 0.23439982114069874,
"grad_norm": 0.37271764874458313,
"learning_rate": 1.1491427604780898e-05,
"loss": 0.027,
"step": 4980
},
{
"epoch": 0.2348705035124672,
"grad_norm": 0.35648247599601746,
"learning_rate": 1.1461272659909137e-05,
"loss": 0.029,
"step": 4990
},
{
"epoch": 0.23534118588423567,
"grad_norm": 0.4681665897369385,
"learning_rate": 1.1431104131309654e-05,
"loss": 0.033,
"step": 5000
},
{
"epoch": 0.23581186825600414,
"grad_norm": 0.38436639308929443,
"learning_rate": 1.1400922299423663e-05,
"loss": 0.0298,
"step": 5010
},
{
"epoch": 0.23628255062777262,
"grad_norm": 0.31347593665122986,
"learning_rate": 1.1370727444816045e-05,
"loss": 0.0298,
"step": 5020
},
{
"epoch": 0.2367532329995411,
"grad_norm": 0.32839059829711914,
"learning_rate": 1.1340519848172735e-05,
"loss": 0.032,
"step": 5030
},
{
"epoch": 0.23722391537130955,
"grad_norm": 0.28293296694755554,
"learning_rate": 1.1310299790298118e-05,
"loss": 0.0287,
"step": 5040
},
{
"epoch": 0.23769459774307802,
"grad_norm": 0.2852814495563507,
"learning_rate": 1.1280067552112408e-05,
"loss": 0.0253,
"step": 5050
},
{
"epoch": 0.2381652801148465,
"grad_norm": 0.24841518700122833,
"learning_rate": 1.124982341464906e-05,
"loss": 0.0337,
"step": 5060
},
{
"epoch": 0.23863596248661498,
"grad_norm": 0.5379837155342102,
"learning_rate": 1.1219567659052126e-05,
"loss": 0.0262,
"step": 5070
},
{
"epoch": 0.23910664485838345,
"grad_norm": 0.3396419286727905,
"learning_rate": 1.118930056657367e-05,
"loss": 0.022,
"step": 5080
},
{
"epoch": 0.2395773272301519,
"grad_norm": 0.3351978659629822,
"learning_rate": 1.115902241857114e-05,
"loss": 0.0266,
"step": 5090
},
{
"epoch": 0.24004800960192038,
"grad_norm": 0.24719175696372986,
"learning_rate": 1.1128733496504751e-05,
"loss": 0.0295,
"step": 5100
},
{
"epoch": 0.24051869197368886,
"grad_norm": 0.3416783809661865,
"learning_rate": 1.1098434081934871e-05,
"loss": 0.0281,
"step": 5110
},
{
"epoch": 0.24098937434545734,
"grad_norm": 0.360234797000885,
"learning_rate": 1.1068124456519402e-05,
"loss": 0.0277,
"step": 5120
},
{
"epoch": 0.24146005671722579,
"grad_norm": 0.2579905688762665,
"learning_rate": 1.1037804902011175e-05,
"loss": 0.0239,
"step": 5130
},
{
"epoch": 0.24193073908899426,
"grad_norm": 0.3462030291557312,
"learning_rate": 1.1007475700255313e-05,
"loss": 0.032,
"step": 5140
},
{
"epoch": 0.24240142146076274,
"grad_norm": 0.31335413455963135,
"learning_rate": 1.0977137133186613e-05,
"loss": 0.0259,
"step": 5150
},
{
"epoch": 0.24287210383253122,
"grad_norm": 0.3449477553367615,
"learning_rate": 1.094678948282694e-05,
"loss": 0.0294,
"step": 5160
},
{
"epoch": 0.2433427862042997,
"grad_norm": 0.315859854221344,
"learning_rate": 1.0916433031282592e-05,
"loss": 0.0244,
"step": 5170
},
{
"epoch": 0.24381346857606814,
"grad_norm": 0.2666293978691101,
"learning_rate": 1.0886068060741676e-05,
"loss": 0.0246,
"step": 5180
},
{
"epoch": 0.24428415094783662,
"grad_norm": 0.26488998532295227,
"learning_rate": 1.0855694853471499e-05,
"loss": 0.0284,
"step": 5190
},
{
"epoch": 0.2447548333196051,
"grad_norm": 0.28882020711898804,
"learning_rate": 1.0825313691815928e-05,
"loss": 0.0265,
"step": 5200
},
{
"epoch": 0.24522551569137357,
"grad_norm": 0.3325543999671936,
"learning_rate": 1.0794924858192779e-05,
"loss": 0.0266,
"step": 5210
},
{
"epoch": 0.24569619806314205,
"grad_norm": 0.3000918924808502,
"learning_rate": 1.0764528635091179e-05,
"loss": 0.0309,
"step": 5220
},
{
"epoch": 0.2461668804349105,
"grad_norm": 0.35208022594451904,
"learning_rate": 1.0734125305068943e-05,
"loss": 0.0256,
"step": 5230
},
{
"epoch": 0.24663756280667898,
"grad_norm": 0.33380022644996643,
"learning_rate": 1.0703715150749967e-05,
"loss": 0.0262,
"step": 5240
},
{
"epoch": 0.24710824517844746,
"grad_norm": 0.37662094831466675,
"learning_rate": 1.0673298454821567e-05,
"loss": 0.0256,
"step": 5250
},
{
"epoch": 0.24757892755021593,
"grad_norm": 0.365785151720047,
"learning_rate": 1.0642875500031878e-05,
"loss": 0.0349,
"step": 5260
},
{
"epoch": 0.2480496099219844,
"grad_norm": 0.24044851958751678,
"learning_rate": 1.0612446569187214e-05,
"loss": 0.0266,
"step": 5270
},
{
"epoch": 0.24852029229375286,
"grad_norm": 0.29368653893470764,
"learning_rate": 1.058201194514944e-05,
"loss": 0.0297,
"step": 5280
},
{
"epoch": 0.24899097466552134,
"grad_norm": 0.23013925552368164,
"learning_rate": 1.0551571910833344e-05,
"loss": 0.0296,
"step": 5290
},
{
"epoch": 0.2494616570372898,
"grad_norm": 0.28861603140830994,
"learning_rate": 1.0521126749204009e-05,
"loss": 0.0267,
"step": 5300
},
{
"epoch": 0.2499323394090583,
"grad_norm": 0.30180099606513977,
"learning_rate": 1.0490676743274181e-05,
"loss": 0.0277,
"step": 5310
},
{
"epoch": 0.25040302178082674,
"grad_norm": 0.41777345538139343,
"learning_rate": 1.0460222176101635e-05,
"loss": 0.0282,
"step": 5320
},
{
"epoch": 0.25087370415259524,
"grad_norm": 0.2709956765174866,
"learning_rate": 1.0429763330786546e-05,
"loss": 0.0223,
"step": 5330
},
{
"epoch": 0.2513443865243637,
"grad_norm": 0.26691508293151855,
"learning_rate": 1.0399300490468862e-05,
"loss": 0.0283,
"step": 5340
},
{
"epoch": 0.25181506889613214,
"grad_norm": 0.2970271110534668,
"learning_rate": 1.0368833938325667e-05,
"loss": 0.0264,
"step": 5350
},
{
"epoch": 0.25228575126790065,
"grad_norm": 0.33287113904953003,
"learning_rate": 1.0338363957568544e-05,
"loss": 0.029,
"step": 5360
},
{
"epoch": 0.2527564336396691,
"grad_norm": 0.32789820432662964,
"learning_rate": 1.030789083144095e-05,
"loss": 0.0278,
"step": 5370
},
{
"epoch": 0.2532271160114376,
"grad_norm": 0.3171306848526001,
"learning_rate": 1.027741484321559e-05,
"loss": 0.0294,
"step": 5380
},
{
"epoch": 0.25369779838320605,
"grad_norm": 0.24666345119476318,
"learning_rate": 1.024693627619176e-05,
"loss": 0.0298,
"step": 5390
},
{
"epoch": 0.2541684807549745,
"grad_norm": 0.3291226029396057,
"learning_rate": 1.0216455413692738e-05,
"loss": 0.0276,
"step": 5400
},
{
"epoch": 0.254639163126743,
"grad_norm": 0.301148384809494,
"learning_rate": 1.0185972539063139e-05,
"loss": 0.0276,
"step": 5410
},
{
"epoch": 0.25510984549851146,
"grad_norm": 0.2809499502182007,
"learning_rate": 1.0155487935666277e-05,
"loss": 0.0278,
"step": 5420
},
{
"epoch": 0.25558052787027996,
"grad_norm": 0.265817254781723,
"learning_rate": 1.0125001886881543e-05,
"loss": 0.0275,
"step": 5430
},
{
"epoch": 0.2560512102420484,
"grad_norm": 0.36675095558166504,
"learning_rate": 1.0094514676101759e-05,
"loss": 0.0298,
"step": 5440
},
{
"epoch": 0.25652189261381686,
"grad_norm": 0.2585409879684448,
"learning_rate": 1.0064026586730553e-05,
"loss": 0.0262,
"step": 5450
},
{
"epoch": 0.25699257498558536,
"grad_norm": 0.2913855314254761,
"learning_rate": 1.0033537902179716e-05,
"loss": 0.0276,
"step": 5460
},
{
"epoch": 0.2574632573573538,
"grad_norm": 0.2676973044872284,
"learning_rate": 1.0003048905866577e-05,
"loss": 0.0275,
"step": 5470
},
{
"epoch": 0.2579339397291223,
"grad_norm": 0.26815009117126465,
"learning_rate": 9.972559881211353e-06,
"loss": 0.0263,
"step": 5480
},
{
"epoch": 0.25840462210089077,
"grad_norm": 0.32846561074256897,
"learning_rate": 9.942071111634538e-06,
"loss": 0.0321,
"step": 5490
},
{
"epoch": 0.2588753044726592,
"grad_norm": 0.372041791677475,
"learning_rate": 9.91158288055425e-06,
"loss": 0.025,
"step": 5500
},
{
"epoch": 0.2593459868444277,
"grad_norm": 0.3609251081943512,
"learning_rate": 9.88109547138359e-06,
"loss": 0.0263,
"step": 5510
},
{
"epoch": 0.25981666921619617,
"grad_norm": 0.3332839906215668,
"learning_rate": 9.850609167528038e-06,
"loss": 0.0264,
"step": 5520
},
{
"epoch": 0.2602873515879647,
"grad_norm": 0.27293452620506287,
"learning_rate": 9.820124252382784e-06,
"loss": 0.0251,
"step": 5530
},
{
"epoch": 0.2607580339597331,
"grad_norm": 0.38913094997406006,
"learning_rate": 9.789641009330113e-06,
"loss": 0.0274,
"step": 5540
},
{
"epoch": 0.2612287163315016,
"grad_norm": 0.3298009932041168,
"learning_rate": 9.759159721736772e-06,
"loss": 0.0253,
"step": 5550
},
{
"epoch": 0.2616993987032701,
"grad_norm": 0.2737114131450653,
"learning_rate": 9.72868067295132e-06,
"loss": 0.0278,
"step": 5560
},
{
"epoch": 0.26217008107503853,
"grad_norm": 0.25773265957832336,
"learning_rate": 9.698204146301513e-06,
"loss": 0.0272,
"step": 5570
},
{
"epoch": 0.26264076344680704,
"grad_norm": 0.27363988757133484,
"learning_rate": 9.667730425091666e-06,
"loss": 0.025,
"step": 5580
},
{
"epoch": 0.2631114458185755,
"grad_norm": 0.31952935457229614,
"learning_rate": 9.637259792599997e-06,
"loss": 0.0295,
"step": 5590
},
{
"epoch": 0.26358212819034393,
"grad_norm": 0.31474849581718445,
"learning_rate": 9.606792532076028e-06,
"loss": 0.0238,
"step": 5600
},
{
"epoch": 0.26405281056211244,
"grad_norm": 0.2836529016494751,
"learning_rate": 9.576328926737936e-06,
"loss": 0.0306,
"step": 5610
},
{
"epoch": 0.2645234929338809,
"grad_norm": 0.3206133544445038,
"learning_rate": 9.545869259769904e-06,
"loss": 0.0264,
"step": 5620
},
{
"epoch": 0.26499417530564934,
"grad_norm": 0.3353675901889801,
"learning_rate": 9.515413814319524e-06,
"loss": 0.0292,
"step": 5630
},
{
"epoch": 0.26546485767741784,
"grad_norm": 0.2633313238620758,
"learning_rate": 9.484962873495137e-06,
"loss": 0.0274,
"step": 5640
},
{
"epoch": 0.2659355400491863,
"grad_norm": 0.3060275912284851,
"learning_rate": 9.454516720363203e-06,
"loss": 0.0272,
"step": 5650
},
{
"epoch": 0.2664062224209548,
"grad_norm": 0.30678167939186096,
"learning_rate": 9.424075637945692e-06,
"loss": 0.0265,
"step": 5660
},
{
"epoch": 0.26687690479272325,
"grad_norm": 0.3052326738834381,
"learning_rate": 9.393639909217423e-06,
"loss": 0.0274,
"step": 5670
},
{
"epoch": 0.2673475871644917,
"grad_norm": 0.20375502109527588,
"learning_rate": 9.363209817103455e-06,
"loss": 0.0272,
"step": 5680
},
{
"epoch": 0.2678182695362602,
"grad_norm": 0.23818859457969666,
"learning_rate": 9.332785644476452e-06,
"loss": 0.0235,
"step": 5690
},
{
"epoch": 0.26828895190802865,
"grad_norm": 0.3721213638782501,
"learning_rate": 9.302367674154043e-06,
"loss": 0.0239,
"step": 5700
},
{
"epoch": 0.26875963427979715,
"grad_norm": 0.29016318917274475,
"learning_rate": 9.271956188896211e-06,
"loss": 0.0284,
"step": 5710
},
{
"epoch": 0.2692303166515656,
"grad_norm": 0.24450695514678955,
"learning_rate": 9.241551471402654e-06,
"loss": 0.023,
"step": 5720
},
{
"epoch": 0.26970099902333405,
"grad_norm": 0.2905883491039276,
"learning_rate": 9.211153804310146e-06,
"loss": 0.0254,
"step": 5730
},
{
"epoch": 0.27017168139510256,
"grad_norm": 0.2244880646467209,
"learning_rate": 9.180763470189938e-06,
"loss": 0.0286,
"step": 5740
},
{
"epoch": 0.270642363766871,
"grad_norm": 0.3987225890159607,
"learning_rate": 9.15038075154511e-06,
"loss": 0.026,
"step": 5750
},
{
"epoch": 0.2711130461386395,
"grad_norm": 0.3773719072341919,
"learning_rate": 9.120005930807939e-06,
"loss": 0.0288,
"step": 5760
},
{
"epoch": 0.27158372851040796,
"grad_norm": 0.32836681604385376,
"learning_rate": 9.0896392903373e-06,
"loss": 0.0228,
"step": 5770
},
{
"epoch": 0.2720544108821764,
"grad_norm": 0.3625943064689636,
"learning_rate": 9.059281112416017e-06,
"loss": 0.0248,
"step": 5780
},
{
"epoch": 0.2725250932539449,
"grad_norm": 0.38033244013786316,
"learning_rate": 9.028931679248249e-06,
"loss": 0.0274,
"step": 5790
},
{
"epoch": 0.27299577562571337,
"grad_norm": 0.5404322743415833,
"learning_rate": 8.998591272956866e-06,
"loss": 0.0263,
"step": 5800
},
{
"epoch": 0.27346645799748187,
"grad_norm": 0.3065263330936432,
"learning_rate": 8.96826017558083e-06,
"loss": 0.0264,
"step": 5810
},
{
"epoch": 0.2739371403692503,
"grad_norm": 0.2693747878074646,
"learning_rate": 8.937938669072557e-06,
"loss": 0.0233,
"step": 5820
},
{
"epoch": 0.27440782274101877,
"grad_norm": 0.2967570126056671,
"learning_rate": 8.90762703529532e-06,
"loss": 0.029,
"step": 5830
},
{
"epoch": 0.2748785051127873,
"grad_norm": 0.19372406601905823,
"learning_rate": 8.877325556020615e-06,
"loss": 0.025,
"step": 5840
},
{
"epoch": 0.2753491874845557,
"grad_norm": 0.33527156710624695,
"learning_rate": 8.847034512925536e-06,
"loss": 0.0307,
"step": 5850
},
{
"epoch": 0.27581986985632423,
"grad_norm": 0.25079697370529175,
"learning_rate": 8.816754187590175e-06,
"loss": 0.0267,
"step": 5860
},
{
"epoch": 0.2762905522280927,
"grad_norm": 0.28940537571907043,
"learning_rate": 8.786484861494984e-06,
"loss": 0.0268,
"step": 5870
},
{
"epoch": 0.27676123459986113,
"grad_norm": 0.21520528197288513,
"learning_rate": 8.756226816018172e-06,
"loss": 0.024,
"step": 5880
},
{
"epoch": 0.27723191697162963,
"grad_norm": 0.23711997270584106,
"learning_rate": 8.725980332433089e-06,
"loss": 0.0293,
"step": 5890
},
{
"epoch": 0.2777025993433981,
"grad_norm": 0.277659147977829,
"learning_rate": 8.695745691905599e-06,
"loss": 0.0302,
"step": 5900
},
{
"epoch": 0.2781732817151666,
"grad_norm": 0.2978789210319519,
"learning_rate": 8.665523175491484e-06,
"loss": 0.0265,
"step": 5910
},
{
"epoch": 0.27864396408693504,
"grad_norm": 0.36361464858055115,
"learning_rate": 8.635313064133817e-06,
"loss": 0.0288,
"step": 5920
},
{
"epoch": 0.2791146464587035,
"grad_norm": 0.2255810797214508,
"learning_rate": 8.605115638660356e-06,
"loss": 0.026,
"step": 5930
},
{
"epoch": 0.279585328830472,
"grad_norm": 0.2857274115085602,
"learning_rate": 8.57493117978094e-06,
"loss": 0.0256,
"step": 5940
},
{
"epoch": 0.28005601120224044,
"grad_norm": 0.30692043900489807,
"learning_rate": 8.544759968084863e-06,
"loss": 0.0246,
"step": 5950
},
{
"epoch": 0.28052669357400895,
"grad_norm": 0.3743002414703369,
"learning_rate": 8.51460228403828e-06,
"loss": 0.026,
"step": 5960
},
{
"epoch": 0.2809973759457774,
"grad_norm": 0.3723158836364746,
"learning_rate": 8.484458407981601e-06,
"loss": 0.0296,
"step": 5970
},
{
"epoch": 0.28146805831754584,
"grad_norm": 0.34177300333976746,
"learning_rate": 8.454328620126871e-06,
"loss": 0.0255,
"step": 5980
},
{
"epoch": 0.28193874068931435,
"grad_norm": 0.3937112092971802,
"learning_rate": 8.424213200555171e-06,
"loss": 0.0296,
"step": 5990
},
{
"epoch": 0.2824094230610828,
"grad_norm": 0.22977639734745026,
"learning_rate": 8.394112429214032e-06,
"loss": 0.0223,
"step": 6000
},
{
"epoch": 0.2828801054328513,
"grad_norm": 0.3045552968978882,
"learning_rate": 8.364026585914802e-06,
"loss": 0.0258,
"step": 6010
},
{
"epoch": 0.28335078780461975,
"grad_norm": 0.26186221837997437,
"learning_rate": 8.33395595033007e-06,
"loss": 0.0243,
"step": 6020
},
{
"epoch": 0.2838214701763882,
"grad_norm": 0.2754952311515808,
"learning_rate": 8.303900801991052e-06,
"loss": 0.0298,
"step": 6030
},
{
"epoch": 0.2842921525481567,
"grad_norm": 0.2855309844017029,
"learning_rate": 8.273861420285e-06,
"loss": 0.0258,
"step": 6040
},
{
"epoch": 0.28476283491992516,
"grad_norm": 0.2871876358985901,
"learning_rate": 8.243838084452603e-06,
"loss": 0.0267,
"step": 6050
},
{
"epoch": 0.2852335172916936,
"grad_norm": 0.27899205684661865,
"learning_rate": 8.213831073585385e-06,
"loss": 0.0283,
"step": 6060
},
{
"epoch": 0.2857041996634621,
"grad_norm": 0.26127520203590393,
"learning_rate": 8.183840666623123e-06,
"loss": 0.0279,
"step": 6070
},
{
"epoch": 0.28617488203523056,
"grad_norm": 0.2863559424877167,
"learning_rate": 8.153867142351242e-06,
"loss": 0.0255,
"step": 6080
},
{
"epoch": 0.28664556440699906,
"grad_norm": 0.4191440939903259,
"learning_rate": 8.123910779398233e-06,
"loss": 0.028,
"step": 6090
},
{
"epoch": 0.2871162467787675,
"grad_norm": 0.3261810541152954,
"learning_rate": 8.093971856233051e-06,
"loss": 0.0247,
"step": 6100
},
{
"epoch": 0.28758692915053596,
"grad_norm": 0.304670125246048,
"learning_rate": 8.064050651162546e-06,
"loss": 0.0284,
"step": 6110
},
{
"epoch": 0.28805761152230447,
"grad_norm": 0.31193462014198303,
"learning_rate": 8.034147442328852e-06,
"loss": 0.0218,
"step": 6120
},
{
"epoch": 0.2885282938940729,
"grad_norm": 0.2479848712682724,
"learning_rate": 8.004262507706819e-06,
"loss": 0.0264,
"step": 6130
},
{
"epoch": 0.2889989762658414,
"grad_norm": 0.2675256133079529,
"learning_rate": 7.97439612510142e-06,
"loss": 0.0286,
"step": 6140
},
{
"epoch": 0.2894696586376099,
"grad_norm": 0.2461802214384079,
"learning_rate": 7.944548572145178e-06,
"loss": 0.0283,
"step": 6150
},
{
"epoch": 0.2899403410093783,
"grad_norm": 0.30420202016830444,
"learning_rate": 7.914720126295572e-06,
"loss": 0.0261,
"step": 6160
},
{
"epoch": 0.2904110233811468,
"grad_norm": 0.3156309723854065,
"learning_rate": 7.884911064832466e-06,
"loss": 0.0249,
"step": 6170
},
{
"epoch": 0.2908817057529153,
"grad_norm": 0.29043760895729065,
"learning_rate": 7.855121664855535e-06,
"loss": 0.0258,
"step": 6180
},
{
"epoch": 0.2913523881246838,
"grad_norm": 0.21979326009750366,
"learning_rate": 7.825352203281682e-06,
"loss": 0.0238,
"step": 6190
},
{
"epoch": 0.29182307049645223,
"grad_norm": 0.3272196352481842,
"learning_rate": 7.79560295684246e-06,
"loss": 0.0287,
"step": 6200
},
{
"epoch": 0.2922937528682207,
"grad_norm": 0.23340076208114624,
"learning_rate": 7.765874202081516e-06,
"loss": 0.0245,
"step": 6210
},
{
"epoch": 0.2927644352399892,
"grad_norm": 0.34092631936073303,
"learning_rate": 7.736166215352004e-06,
"loss": 0.0247,
"step": 6220
},
{
"epoch": 0.29323511761175763,
"grad_norm": 0.3207491338253021,
"learning_rate": 7.706479272814024e-06,
"loss": 0.0257,
"step": 6230
},
{
"epoch": 0.29370579998352614,
"grad_norm": 0.24749363958835602,
"learning_rate": 7.67681365043205e-06,
"loss": 0.0226,
"step": 6240
},
{
"epoch": 0.2941764823552946,
"grad_norm": 0.1940588653087616,
"learning_rate": 7.64716962397237e-06,
"loss": 0.0211,
"step": 6250
},
{
"epoch": 0.29464716472706304,
"grad_norm": 0.28771379590034485,
"learning_rate": 7.617547469000524e-06,
"loss": 0.0243,
"step": 6260
},
{
"epoch": 0.29511784709883154,
"grad_norm": 0.21900710463523865,
"learning_rate": 7.587947460878731e-06,
"loss": 0.0248,
"step": 6270
},
{
"epoch": 0.2955885294706,
"grad_norm": 0.30860117077827454,
"learning_rate": 7.5583698747633394e-06,
"loss": 0.0239,
"step": 6280
},
{
"epoch": 0.2960592118423685,
"grad_norm": 0.42496931552886963,
"learning_rate": 7.528814985602273e-06,
"loss": 0.0228,
"step": 6290
},
{
"epoch": 0.29652989421413695,
"grad_norm": 0.31019580364227295,
"learning_rate": 7.49928306813246e-06,
"loss": 0.0298,
"step": 6300
},
{
"epoch": 0.2970005765859054,
"grad_norm": 0.30217042565345764,
"learning_rate": 7.4697743968772906e-06,
"loss": 0.0289,
"step": 6310
},
{
"epoch": 0.2974712589576739,
"grad_norm": 0.39460352063179016,
"learning_rate": 7.440289246144067e-06,
"loss": 0.0272,
"step": 6320
},
{
"epoch": 0.29794194132944235,
"grad_norm": 0.25050991773605347,
"learning_rate": 7.410827890021444e-06,
"loss": 0.0226,
"step": 6330
},
{
"epoch": 0.29841262370121086,
"grad_norm": 0.3621877431869507,
"learning_rate": 7.381390602376882e-06,
"loss": 0.0314,
"step": 6340
},
{
"epoch": 0.2988833060729793,
"grad_norm": 0.32842952013015747,
"learning_rate": 7.351977656854118e-06,
"loss": 0.0255,
"step": 6350
},
{
"epoch": 0.29935398844474775,
"grad_norm": 0.3007189631462097,
"learning_rate": 7.322589326870597e-06,
"loss": 0.023,
"step": 6360
},
{
"epoch": 0.29982467081651626,
"grad_norm": 0.26633837819099426,
"learning_rate": 7.293225885614948e-06,
"loss": 0.0274,
"step": 6370
},
{
"epoch": 0.3002953531882847,
"grad_norm": 0.2466638833284378,
"learning_rate": 7.263887606044437e-06,
"loss": 0.0227,
"step": 6380
},
{
"epoch": 0.3007660355600532,
"grad_norm": 0.2613636255264282,
"learning_rate": 7.234574760882431e-06,
"loss": 0.0282,
"step": 6390
},
{
"epoch": 0.30123671793182166,
"grad_norm": 0.26148852705955505,
"learning_rate": 7.205287622615866e-06,
"loss": 0.0214,
"step": 6400
},
{
"epoch": 0.3017074003035901,
"grad_norm": 0.3387831747531891,
"learning_rate": 7.176026463492711e-06,
"loss": 0.0216,
"step": 6410
},
{
"epoch": 0.3021780826753586,
"grad_norm": 0.2681928873062134,
"learning_rate": 7.146791555519431e-06,
"loss": 0.0246,
"step": 6420
},
{
"epoch": 0.30264876504712707,
"grad_norm": 0.2519020140171051,
"learning_rate": 7.117583170458478e-06,
"loss": 0.019,
"step": 6430
},
{
"epoch": 0.30311944741889557,
"grad_norm": 0.3675183057785034,
"learning_rate": 7.0884015798257365e-06,
"loss": 0.029,
"step": 6440
},
{
"epoch": 0.303590129790664,
"grad_norm": 0.33489444851875305,
"learning_rate": 7.059247054888025e-06,
"loss": 0.0257,
"step": 6450
},
{
"epoch": 0.30406081216243247,
"grad_norm": 0.3435460329055786,
"learning_rate": 7.030119866660565e-06,
"loss": 0.0254,
"step": 6460
},
{
"epoch": 0.304531494534201,
"grad_norm": 0.2610439360141754,
"learning_rate": 7.001020285904454e-06,
"loss": 0.0235,
"step": 6470
},
{
"epoch": 0.3050021769059694,
"grad_norm": 0.28317561745643616,
"learning_rate": 6.971948583124159e-06,
"loss": 0.0258,
"step": 6480
},
{
"epoch": 0.3054728592777379,
"grad_norm": 0.2600167989730835,
"learning_rate": 6.9429050285650015e-06,
"loss": 0.0259,
"step": 6490
},
{
"epoch": 0.3059435416495064,
"grad_norm": 0.2560778856277466,
"learning_rate": 6.913889892210631e-06,
"loss": 0.026,
"step": 6500
},
{
"epoch": 0.30641422402127483,
"grad_norm": 0.28762975335121155,
"learning_rate": 6.884903443780541e-06,
"loss": 0.0218,
"step": 6510
},
{
"epoch": 0.30688490639304333,
"grad_norm": 0.32111966609954834,
"learning_rate": 6.8559459527275426e-06,
"loss": 0.0252,
"step": 6520
},
{
"epoch": 0.3073555887648118,
"grad_norm": 0.3172910213470459,
"learning_rate": 6.827017688235255e-06,
"loss": 0.0241,
"step": 6530
},
{
"epoch": 0.30782627113658023,
"grad_norm": 0.29458677768707275,
"learning_rate": 6.798118919215625e-06,
"loss": 0.0258,
"step": 6540
},
{
"epoch": 0.30829695350834874,
"grad_norm": 0.31301990151405334,
"learning_rate": 6.769249914306408e-06,
"loss": 0.0282,
"step": 6550
},
{
"epoch": 0.3087676358801172,
"grad_norm": 0.2761021852493286,
"learning_rate": 6.740410941868678e-06,
"loss": 0.0248,
"step": 6560
},
{
"epoch": 0.3092383182518857,
"grad_norm": 0.28405529260635376,
"learning_rate": 6.711602269984339e-06,
"loss": 0.0259,
"step": 6570
},
{
"epoch": 0.30970900062365414,
"grad_norm": 0.32768476009368896,
"learning_rate": 6.6828241664536145e-06,
"loss": 0.0255,
"step": 6580
},
{
"epoch": 0.3101796829954226,
"grad_norm": 0.25966060161590576,
"learning_rate": 6.65407689879258e-06,
"loss": 0.0229,
"step": 6590
},
{
"epoch": 0.3106503653671911,
"grad_norm": 0.3064074218273163,
"learning_rate": 6.625360734230663e-06,
"loss": 0.0226,
"step": 6600
},
{
"epoch": 0.31112104773895954,
"grad_norm": 0.2384389191865921,
"learning_rate": 6.596675939708166e-06,
"loss": 0.0273,
"step": 6610
},
{
"epoch": 0.31159173011072805,
"grad_norm": 0.27735522389411926,
"learning_rate": 6.5680227818737695e-06,
"loss": 0.022,
"step": 6620
},
{
"epoch": 0.3120624124824965,
"grad_norm": 0.30861416459083557,
"learning_rate": 6.539401527082083e-06,
"loss": 0.026,
"step": 6630
},
{
"epoch": 0.31253309485426495,
"grad_norm": 0.34236130118370056,
"learning_rate": 6.510812441391131e-06,
"loss": 0.0248,
"step": 6640
},
{
"epoch": 0.31300377722603345,
"grad_norm": 0.2688142657279968,
"learning_rate": 6.4822557905599156e-06,
"loss": 0.0226,
"step": 6650
},
{
"epoch": 0.3134744595978019,
"grad_norm": 0.2897597551345825,
"learning_rate": 6.4537318400459295e-06,
"loss": 0.0222,
"step": 6660
},
{
"epoch": 0.3139451419695704,
"grad_norm": 0.22985027730464935,
"learning_rate": 6.425240855002674e-06,
"loss": 0.0231,
"step": 6670
},
{
"epoch": 0.31441582434133886,
"grad_norm": 0.2983686625957489,
"learning_rate": 6.396783100277224e-06,
"loss": 0.0246,
"step": 6680
},
{
"epoch": 0.3148865067131073,
"grad_norm": 0.2697815001010895,
"learning_rate": 6.368358840407754e-06,
"loss": 0.0209,
"step": 6690
},
{
"epoch": 0.3153571890848758,
"grad_norm": 0.2615862488746643,
"learning_rate": 6.339968339621056e-06,
"loss": 0.0229,
"step": 6700
},
{
"epoch": 0.31582787145664426,
"grad_norm": 0.38299912214279175,
"learning_rate": 6.311611861830129e-06,
"loss": 0.0304,
"step": 6710
},
{
"epoch": 0.31629855382841277,
"grad_norm": 0.25474509596824646,
"learning_rate": 6.283289670631684e-06,
"loss": 0.0221,
"step": 6720
},
{
"epoch": 0.3167692362001812,
"grad_norm": 0.2385791540145874,
"learning_rate": 6.2550020293037095e-06,
"loss": 0.023,
"step": 6730
},
{
"epoch": 0.31723991857194966,
"grad_norm": 0.2175077646970749,
"learning_rate": 6.2267492008030395e-06,
"loss": 0.0259,
"step": 6740
},
{
"epoch": 0.31771060094371817,
"grad_norm": 0.28159549832344055,
"learning_rate": 6.198531447762875e-06,
"loss": 0.0217,
"step": 6750
},
{
"epoch": 0.3181812833154866,
"grad_norm": 0.40623795986175537,
"learning_rate": 6.1703490324903745e-06,
"loss": 0.023,
"step": 6760
},
{
"epoch": 0.3186519656872551,
"grad_norm": 0.2421952337026596,
"learning_rate": 6.142202216964204e-06,
"loss": 0.0205,
"step": 6770
},
{
"epoch": 0.3191226480590236,
"grad_norm": 0.3194625675678253,
"learning_rate": 6.114091262832087e-06,
"loss": 0.0306,
"step": 6780
},
{
"epoch": 0.319593330430792,
"grad_norm": 0.2308640480041504,
"learning_rate": 6.0860164314084e-06,
"loss": 0.0233,
"step": 6790
},
{
"epoch": 0.3200640128025605,
"grad_norm": 0.2641204595565796,
"learning_rate": 6.05797798367173e-06,
"loss": 0.025,
"step": 6800
},
{
"epoch": 0.320534695174329,
"grad_norm": 0.2730650007724762,
"learning_rate": 6.029976180262431e-06,
"loss": 0.0222,
"step": 6810
},
{
"epoch": 0.3210053775460975,
"grad_norm": 0.2418140321969986,
"learning_rate": 6.0020112814802355e-06,
"loss": 0.0223,
"step": 6820
},
{
"epoch": 0.32147605991786593,
"grad_norm": 0.2687058448791504,
"learning_rate": 5.9740835472818145e-06,
"loss": 0.0289,
"step": 6830
},
{
"epoch": 0.3219467422896344,
"grad_norm": 0.370744913816452,
"learning_rate": 5.946193237278352e-06,
"loss": 0.0257,
"step": 6840
},
{
"epoch": 0.3224174246614029,
"grad_norm": 0.27342575788497925,
"learning_rate": 5.918340610733154e-06,
"loss": 0.0268,
"step": 6850
},
{
"epoch": 0.32288810703317133,
"grad_norm": 0.3032595217227936,
"learning_rate": 5.8905259265592315e-06,
"loss": 0.0236,
"step": 6860
},
{
"epoch": 0.32335878940493984,
"grad_norm": 0.33203062415122986,
"learning_rate": 5.8627494433168756e-06,
"loss": 0.0302,
"step": 6870
},
{
"epoch": 0.3238294717767083,
"grad_norm": 0.36624008417129517,
"learning_rate": 5.835011419211285e-06,
"loss": 0.0261,
"step": 6880
},
{
"epoch": 0.32430015414847674,
"grad_norm": 0.427299827337265,
"learning_rate": 5.807312112090129e-06,
"loss": 0.0245,
"step": 6890
},
{
"epoch": 0.32477083652024524,
"grad_norm": 0.27118441462516785,
"learning_rate": 5.779651779441192e-06,
"loss": 0.0282,
"step": 6900
},
{
"epoch": 0.3252415188920137,
"grad_norm": 0.33013179898262024,
"learning_rate": 5.752030678389948e-06,
"loss": 0.0265,
"step": 6910
},
{
"epoch": 0.32571220126378214,
"grad_norm": 0.21457287669181824,
"learning_rate": 5.724449065697182e-06,
"loss": 0.0185,
"step": 6920
},
{
"epoch": 0.32618288363555065,
"grad_norm": 0.24079112708568573,
"learning_rate": 5.696907197756598e-06,
"loss": 0.0293,
"step": 6930
},
{
"epoch": 0.3266535660073191,
"grad_norm": 0.3726034462451935,
"learning_rate": 5.669405330592457e-06,
"loss": 0.0209,
"step": 6940
},
{
"epoch": 0.3271242483790876,
"grad_norm": 0.2827111482620239,
"learning_rate": 5.6419437198571525e-06,
"loss": 0.0229,
"step": 6950
},
{
"epoch": 0.32759493075085605,
"grad_norm": 0.2810194194316864,
"learning_rate": 5.6145226208288875e-06,
"loss": 0.0223,
"step": 6960
},
{
"epoch": 0.3280656131226245,
"grad_norm": 0.27260082960128784,
"learning_rate": 5.587142288409262e-06,
"loss": 0.0209,
"step": 6970
},
{
"epoch": 0.328536295494393,
"grad_norm": 0.3105827569961548,
"learning_rate": 5.559802977120918e-06,
"loss": 0.0254,
"step": 6980
},
{
"epoch": 0.32900697786616145,
"grad_norm": 0.3558630347251892,
"learning_rate": 5.532504941105176e-06,
"loss": 0.0249,
"step": 6990
},
{
"epoch": 0.32947766023792996,
"grad_norm": 0.38981935381889343,
"learning_rate": 5.505248434119666e-06,
"loss": 0.0281,
"step": 7000
},
{
"epoch": 0.3299483426096984,
"grad_norm": 0.3139411211013794,
"learning_rate": 5.478033709535968e-06,
"loss": 0.024,
"step": 7010
},
{
"epoch": 0.33041902498146686,
"grad_norm": 0.2520444393157959,
"learning_rate": 5.4508610203372794e-06,
"loss": 0.0218,
"step": 7020
},
{
"epoch": 0.33088970735323536,
"grad_norm": 0.38872087001800537,
"learning_rate": 5.42373061911601e-06,
"loss": 0.0267,
"step": 7030
},
{
"epoch": 0.3313603897250038,
"grad_norm": 0.24046728014945984,
"learning_rate": 5.3966427580715044e-06,
"loss": 0.0293,
"step": 7040
},
{
"epoch": 0.3318310720967723,
"grad_norm": 0.259506493806839,
"learning_rate": 5.3695976890076375e-06,
"loss": 0.0224,
"step": 7050
},
{
"epoch": 0.33230175446854077,
"grad_norm": 0.26716184616088867,
"learning_rate": 5.3425956633305075e-06,
"loss": 0.0236,
"step": 7060
},
{
"epoch": 0.3327724368403092,
"grad_norm": 0.27387934923171997,
"learning_rate": 5.3156369320460796e-06,
"loss": 0.0227,
"step": 7070
},
{
"epoch": 0.3332431192120777,
"grad_norm": 0.23044610023498535,
"learning_rate": 5.2887217457578856e-06,
"loss": 0.0226,
"step": 7080
},
{
"epoch": 0.33371380158384617,
"grad_norm": 0.25945109128952026,
"learning_rate": 5.261850354664633e-06,
"loss": 0.0282,
"step": 7090
},
{
"epoch": 0.3341844839556147,
"grad_norm": 0.2804930508136749,
"learning_rate": 5.235023008557955e-06,
"loss": 0.0275,
"step": 7100
},
{
"epoch": 0.3346551663273831,
"grad_norm": 0.275107204914093,
"learning_rate": 5.20823995682003e-06,
"loss": 0.0254,
"step": 7110
},
{
"epoch": 0.3351258486991516,
"grad_norm": 0.26841190457344055,
"learning_rate": 5.1815014484212825e-06,
"loss": 0.0254,
"step": 7120
},
{
"epoch": 0.3355965310709201,
"grad_norm": 0.3452455699443817,
"learning_rate": 5.154807731918081e-06,
"loss": 0.0268,
"step": 7130
},
{
"epoch": 0.33606721344268853,
"grad_norm": 0.27676892280578613,
"learning_rate": 5.1281590554504095e-06,
"loss": 0.0231,
"step": 7140
},
{
"epoch": 0.33653789581445703,
"grad_norm": 0.32191967964172363,
"learning_rate": 5.101555666739563e-06,
"loss": 0.0251,
"step": 7150
},
{
"epoch": 0.3370085781862255,
"grad_norm": 0.2446443736553192,
"learning_rate": 5.074997813085873e-06,
"loss": 0.0229,
"step": 7160
},
{
"epoch": 0.33747926055799393,
"grad_norm": 0.32071539759635925,
"learning_rate": 5.048485741366351e-06,
"loss": 0.0247,
"step": 7170
},
{
"epoch": 0.33794994292976244,
"grad_norm": 0.2713768184185028,
"learning_rate": 5.0220196980324545e-06,
"loss": 0.0263,
"step": 7180
},
{
"epoch": 0.3384206253015309,
"grad_norm": 0.27324774861335754,
"learning_rate": 4.995599929107758e-06,
"loss": 0.0241,
"step": 7190
},
{
"epoch": 0.3388913076732994,
"grad_norm": 0.26016825437545776,
"learning_rate": 4.9692266801856815e-06,
"loss": 0.0234,
"step": 7200
},
{
"epoch": 0.33936199004506784,
"grad_norm": 0.22451132535934448,
"learning_rate": 4.942900196427195e-06,
"loss": 0.0272,
"step": 7210
},
{
"epoch": 0.3398326724168363,
"grad_norm": 0.3017019033432007,
"learning_rate": 4.916620722558568e-06,
"loss": 0.0258,
"step": 7220
},
{
"epoch": 0.3403033547886048,
"grad_norm": 0.268436074256897,
"learning_rate": 4.8903885028690454e-06,
"loss": 0.0258,
"step": 7230
},
{
"epoch": 0.34077403716037324,
"grad_norm": 0.2318979650735855,
"learning_rate": 4.864203781208632e-06,
"loss": 0.0225,
"step": 7240
},
{
"epoch": 0.34124471953214175,
"grad_norm": 0.32598090171813965,
"learning_rate": 4.838066800985786e-06,
"loss": 0.0236,
"step": 7250
},
{
"epoch": 0.3417154019039102,
"grad_norm": 0.39213067293167114,
"learning_rate": 4.811977805165174e-06,
"loss": 0.0265,
"step": 7260
},
{
"epoch": 0.34218608427567865,
"grad_norm": 0.3097468614578247,
"learning_rate": 4.7859370362654045e-06,
"loss": 0.0272,
"step": 7270
},
{
"epoch": 0.34265676664744715,
"grad_norm": 0.2696644067764282,
"learning_rate": 4.75994473635678e-06,
"loss": 0.0255,
"step": 7280
},
{
"epoch": 0.3431274490192156,
"grad_norm": 0.24730904400348663,
"learning_rate": 4.7340011470590415e-06,
"loss": 0.0259,
"step": 7290
},
{
"epoch": 0.34359813139098405,
"grad_norm": 0.2344730943441391,
"learning_rate": 4.708106509539134e-06,
"loss": 0.0292,
"step": 7300
},
{
"epoch": 0.34406881376275256,
"grad_norm": 0.26553988456726074,
"learning_rate": 4.682261064508944e-06,
"loss": 0.0221,
"step": 7310
},
{
"epoch": 0.344539496134521,
"grad_norm": 0.348336398601532,
"learning_rate": 4.656465052223079e-06,
"loss": 0.0279,
"step": 7320
},
{
"epoch": 0.3450101785062895,
"grad_norm": 0.2800534665584564,
"learning_rate": 4.630718712476628e-06,
"loss": 0.025,
"step": 7330
},
{
"epoch": 0.34548086087805796,
"grad_norm": 0.28959569334983826,
"learning_rate": 4.6050222846029315e-06,
"loss": 0.0227,
"step": 7340
},
{
"epoch": 0.3459515432498264,
"grad_norm": 0.2327575534582138,
"learning_rate": 4.5793760074713565e-06,
"loss": 0.023,
"step": 7350
},
{
"epoch": 0.3464222256215949,
"grad_norm": 0.2986283004283905,
"learning_rate": 4.553780119485093e-06,
"loss": 0.0259,
"step": 7360
},
{
"epoch": 0.34689290799336336,
"grad_norm": 0.3520296812057495,
"learning_rate": 4.528234858578894e-06,
"loss": 0.0271,
"step": 7370
},
{
"epoch": 0.34736359036513187,
"grad_norm": 0.2651362419128418,
"learning_rate": 4.502740462216919e-06,
"loss": 0.0239,
"step": 7380
},
{
"epoch": 0.3478342727369003,
"grad_norm": 0.2802233099937439,
"learning_rate": 4.477297167390487e-06,
"loss": 0.0251,
"step": 7390
},
{
"epoch": 0.34830495510866877,
"grad_norm": 0.21657630801200867,
"learning_rate": 4.451905210615889e-06,
"loss": 0.0217,
"step": 7400
},
{
"epoch": 0.3487756374804373,
"grad_norm": 0.34882038831710815,
"learning_rate": 4.426564827932185e-06,
"loss": 0.0223,
"step": 7410
},
{
"epoch": 0.3492463198522057,
"grad_norm": 0.2683546841144562,
"learning_rate": 4.401276254899014e-06,
"loss": 0.0225,
"step": 7420
},
{
"epoch": 0.3497170022239742,
"grad_norm": 0.26570001244544983,
"learning_rate": 4.3760397265943965e-06,
"loss": 0.0225,
"step": 7430
},
{
"epoch": 0.3501876845957427,
"grad_norm": 0.3013390898704529,
"learning_rate": 4.350855477612565e-06,
"loss": 0.0259,
"step": 7440
},
{
"epoch": 0.3506583669675111,
"grad_norm": 0.28702089190483093,
"learning_rate": 4.325723742061767e-06,
"loss": 0.0235,
"step": 7450
},
{
"epoch": 0.35112904933927963,
"grad_norm": 0.33644604682922363,
"learning_rate": 4.30064475356209e-06,
"loss": 0.0207,
"step": 7460
},
{
"epoch": 0.3515997317110481,
"grad_norm": 0.464994877576828,
"learning_rate": 4.275618745243301e-06,
"loss": 0.0243,
"step": 7470
},
{
"epoch": 0.3520704140828166,
"grad_norm": 0.32903018593788147,
"learning_rate": 4.2506459497426685e-06,
"loss": 0.0277,
"step": 7480
},
{
"epoch": 0.35254109645458503,
"grad_norm": 0.30534628033638,
"learning_rate": 4.225726599202808e-06,
"loss": 0.0237,
"step": 7490
},
{
"epoch": 0.3530117788263535,
"grad_norm": 0.28936251997947693,
"learning_rate": 4.200860925269519e-06,
"loss": 0.0224,
"step": 7500
},
{
"epoch": 0.353482461198122,
"grad_norm": 0.2381279468536377,
"learning_rate": 4.176049159089626e-06,
"loss": 0.0228,
"step": 7510
},
{
"epoch": 0.35395314356989044,
"grad_norm": 0.2866049110889435,
"learning_rate": 4.1512915313088505e-06,
"loss": 0.0251,
"step": 7520
},
{
"epoch": 0.35442382594165894,
"grad_norm": 0.23486320674419403,
"learning_rate": 4.126588272069645e-06,
"loss": 0.0255,
"step": 7530
},
{
"epoch": 0.3548945083134274,
"grad_norm": 0.24740472435951233,
"learning_rate": 4.101939611009059e-06,
"loss": 0.0203,
"step": 7540
},
{
"epoch": 0.35536519068519584,
"grad_norm": 0.3370332419872284,
"learning_rate": 4.077345777256614e-06,
"loss": 0.0279,
"step": 7550
},
{
"epoch": 0.35583587305696435,
"grad_norm": 0.23352529108524323,
"learning_rate": 4.052806999432161e-06,
"loss": 0.0244,
"step": 7560
},
{
"epoch": 0.3563065554287328,
"grad_norm": 0.22148875892162323,
"learning_rate": 4.028323505643762e-06,
"loss": 0.0198,
"step": 7570
},
{
"epoch": 0.3567772378005013,
"grad_norm": 0.2708054482936859,
"learning_rate": 4.003895523485575e-06,
"loss": 0.0209,
"step": 7580
},
{
"epoch": 0.35724792017226975,
"grad_norm": 0.2242029756307602,
"learning_rate": 3.979523280035723e-06,
"loss": 0.0294,
"step": 7590
},
{
"epoch": 0.3577186025440382,
"grad_norm": 0.2960135042667389,
"learning_rate": 3.955207001854197e-06,
"loss": 0.026,
"step": 7600
},
{
"epoch": 0.3581892849158067,
"grad_norm": 0.2580626904964447,
"learning_rate": 3.930946914980744e-06,
"loss": 0.0269,
"step": 7610
},
{
"epoch": 0.35865996728757515,
"grad_norm": 0.24140478670597076,
"learning_rate": 3.906743244932767e-06,
"loss": 0.0243,
"step": 7620
},
{
"epoch": 0.35913064965934366,
"grad_norm": 0.24541927874088287,
"learning_rate": 3.882596216703226e-06,
"loss": 0.0263,
"step": 7630
},
{
"epoch": 0.3596013320311121,
"grad_norm": 0.43830493092536926,
"learning_rate": 3.858506054758547e-06,
"loss": 0.0226,
"step": 7640
},
{
"epoch": 0.36007201440288056,
"grad_norm": 0.3424168825149536,
"learning_rate": 3.834472983036551e-06,
"loss": 0.0233,
"step": 7650
},
{
"epoch": 0.36054269677464906,
"grad_norm": 0.3170373737812042,
"learning_rate": 3.8104972249443417e-06,
"loss": 0.022,
"step": 7660
},
{
"epoch": 0.3610133791464175,
"grad_norm": 0.22543689608573914,
"learning_rate": 3.7865790033562532e-06,
"loss": 0.0219,
"step": 7670
},
{
"epoch": 0.361484061518186,
"grad_norm": 0.3022764027118683,
"learning_rate": 3.7627185406117707e-06,
"loss": 0.0267,
"step": 7680
},
{
"epoch": 0.36195474388995447,
"grad_norm": 0.25519615411758423,
"learning_rate": 3.738916058513462e-06,
"loss": 0.0223,
"step": 7690
},
{
"epoch": 0.3624254262617229,
"grad_norm": 0.25609269738197327,
"learning_rate": 3.7151717783249175e-06,
"loss": 0.0222,
"step": 7700
},
{
"epoch": 0.3628961086334914,
"grad_norm": 0.2746790051460266,
"learning_rate": 3.6914859207686916e-06,
"loss": 0.0236,
"step": 7710
},
{
"epoch": 0.36336679100525987,
"grad_norm": 0.34544268250465393,
"learning_rate": 3.6678587060242586e-06,
"loss": 0.0295,
"step": 7720
},
{
"epoch": 0.3638374733770283,
"grad_norm": 0.3708873689174652,
"learning_rate": 3.6442903537259556e-06,
"loss": 0.0244,
"step": 7730
},
{
"epoch": 0.3643081557487968,
"grad_norm": 0.3822401762008667,
"learning_rate": 3.6207810829609414e-06,
"loss": 0.0215,
"step": 7740
},
{
"epoch": 0.3647788381205653,
"grad_norm": 0.26645803451538086,
"learning_rate": 3.5973311122671695e-06,
"loss": 0.0226,
"step": 7750
},
{
"epoch": 0.3652495204923338,
"grad_norm": 0.2992480397224426,
"learning_rate": 3.5739406596313474e-06,
"loss": 0.024,
"step": 7760
},
{
"epoch": 0.36572020286410223,
"grad_norm": 0.3051004707813263,
"learning_rate": 3.5506099424869133e-06,
"loss": 0.0219,
"step": 7770
},
{
"epoch": 0.3661908852358707,
"grad_norm": 0.2562768757343292,
"learning_rate": 3.5273391777120136e-06,
"loss": 0.027,
"step": 7780
},
{
"epoch": 0.3666615676076392,
"grad_norm": 0.20431318879127502,
"learning_rate": 3.504128581627497e-06,
"loss": 0.024,
"step": 7790
},
{
"epoch": 0.36713224997940763,
"grad_norm": 0.2949235737323761,
"learning_rate": 3.480978369994885e-06,
"loss": 0.0251,
"step": 7800
},
{
"epoch": 0.36760293235117614,
"grad_norm": 0.28835704922676086,
"learning_rate": 3.4578887580143793e-06,
"loss": 0.0255,
"step": 7810
},
{
"epoch": 0.3680736147229446,
"grad_norm": 0.3093815743923187,
"learning_rate": 3.4348599603228584e-06,
"loss": 0.0266,
"step": 7820
},
{
"epoch": 0.36854429709471304,
"grad_norm": 0.35156458616256714,
"learning_rate": 3.411892190991882e-06,
"loss": 0.0238,
"step": 7830
},
{
"epoch": 0.36901497946648154,
"grad_norm": 0.2666817307472229,
"learning_rate": 3.3889856635257024e-06,
"loss": 0.0234,
"step": 7840
},
{
"epoch": 0.36948566183825,
"grad_norm": 0.32229962944984436,
"learning_rate": 3.366140590859276e-06,
"loss": 0.0237,
"step": 7850
},
{
"epoch": 0.3699563442100185,
"grad_norm": 0.2692343294620514,
"learning_rate": 3.343357185356284e-06,
"loss": 0.0225,
"step": 7860
},
{
"epoch": 0.37042702658178694,
"grad_norm": 0.2344328910112381,
"learning_rate": 3.3206356588071733e-06,
"loss": 0.0204,
"step": 7870
},
{
"epoch": 0.3708977089535554,
"grad_norm": 0.40593421459198,
"learning_rate": 3.2979762224271616e-06,
"loss": 0.0222,
"step": 7880
},
{
"epoch": 0.3713683913253239,
"grad_norm": 0.4432043433189392,
"learning_rate": 3.275379086854292e-06,
"loss": 0.0273,
"step": 7890
},
{
"epoch": 0.37183907369709235,
"grad_norm": 0.2933953106403351,
"learning_rate": 3.252844462147472e-06,
"loss": 0.0208,
"step": 7900
},
{
"epoch": 0.37230975606886085,
"grad_norm": 0.35729560256004333,
"learning_rate": 3.230372557784518e-06,
"loss": 0.0218,
"step": 7910
},
{
"epoch": 0.3727804384406293,
"grad_norm": 0.33724620938301086,
"learning_rate": 3.2079635826602053e-06,
"loss": 0.0234,
"step": 7920
},
{
"epoch": 0.37325112081239775,
"grad_norm": 0.3514053225517273,
"learning_rate": 3.185617745084343e-06,
"loss": 0.0207,
"step": 7930
},
{
"epoch": 0.37372180318416626,
"grad_norm": 0.31914830207824707,
"learning_rate": 3.163335252779811e-06,
"loss": 0.0237,
"step": 7940
},
{
"epoch": 0.3741924855559347,
"grad_norm": 0.2323969006538391,
"learning_rate": 3.1411163128806497e-06,
"loss": 0.0216,
"step": 7950
},
{
"epoch": 0.3746631679277032,
"grad_norm": 0.2753153443336487,
"learning_rate": 3.118961131930127e-06,
"loss": 0.0228,
"step": 7960
},
{
"epoch": 0.37513385029947166,
"grad_norm": 0.3089962303638458,
"learning_rate": 3.0968699158788185e-06,
"loss": 0.0203,
"step": 7970
},
{
"epoch": 0.3756045326712401,
"grad_norm": 0.29103320837020874,
"learning_rate": 3.0748428700826938e-06,
"loss": 0.0216,
"step": 7980
},
{
"epoch": 0.3760752150430086,
"grad_norm": 0.22521524131298065,
"learning_rate": 3.0528801993012056e-06,
"loss": 0.0224,
"step": 7990
},
{
"epoch": 0.37654589741477706,
"grad_norm": 0.22297009825706482,
"learning_rate": 3.0309821076953893e-06,
"loss": 0.0216,
"step": 8000
},
{
"epoch": 0.37701657978654557,
"grad_norm": 0.3377138078212738,
"learning_rate": 3.0091487988259684e-06,
"loss": 0.0212,
"step": 8010
},
{
"epoch": 0.377487262158314,
"grad_norm": 0.25053882598876953,
"learning_rate": 2.9873804756514513e-06,
"loss": 0.0221,
"step": 8020
},
{
"epoch": 0.37795794453008247,
"grad_norm": 0.32806795835494995,
"learning_rate": 2.965677340526254e-06,
"loss": 0.0236,
"step": 8030
},
{
"epoch": 0.378428626901851,
"grad_norm": 0.2913495600223541,
"learning_rate": 2.944039595198814e-06,
"loss": 0.0246,
"step": 8040
},
{
"epoch": 0.3788993092736194,
"grad_norm": 0.28053534030914307,
"learning_rate": 2.9224674408097207e-06,
"loss": 0.0231,
"step": 8050
},
{
"epoch": 0.3793699916453879,
"grad_norm": 0.23675855994224548,
"learning_rate": 2.900961077889837e-06,
"loss": 0.0262,
"step": 8060
},
{
"epoch": 0.3798406740171564,
"grad_norm": 0.21393704414367676,
"learning_rate": 2.879520706358446e-06,
"loss": 0.0245,
"step": 8070
},
{
"epoch": 0.3803113563889248,
"grad_norm": 0.25253769755363464,
"learning_rate": 2.8581465255213834e-06,
"loss": 0.0205,
"step": 8080
},
{
"epoch": 0.38078203876069333,
"grad_norm": 0.2491852045059204,
"learning_rate": 2.836838734069187e-06,
"loss": 0.0228,
"step": 8090
},
{
"epoch": 0.3812527211324618,
"grad_norm": 0.33427563309669495,
"learning_rate": 2.8155975300752524e-06,
"loss": 0.0256,
"step": 8100
},
{
"epoch": 0.3817234035042303,
"grad_norm": 0.200931578874588,
"learning_rate": 2.794423110993991e-06,
"loss": 0.027,
"step": 8110
},
{
"epoch": 0.38219408587599873,
"grad_norm": 0.30499860644340515,
"learning_rate": 2.7733156736589893e-06,
"loss": 0.0273,
"step": 8120
},
{
"epoch": 0.3826647682477672,
"grad_norm": 0.2738380432128906,
"learning_rate": 2.7522754142811957e-06,
"loss": 0.0265,
"step": 8130
},
{
"epoch": 0.3831354506195357,
"grad_norm": 0.25490108132362366,
"learning_rate": 2.731302528447063e-06,
"loss": 0.0256,
"step": 8140
},
{
"epoch": 0.38360613299130414,
"grad_norm": 0.30889907479286194,
"learning_rate": 2.710397211116774e-06,
"loss": 0.0215,
"step": 8150
},
{
"epoch": 0.3840768153630726,
"grad_norm": 0.23913100361824036,
"learning_rate": 2.6895596566223937e-06,
"loss": 0.0272,
"step": 8160
},
{
"epoch": 0.3845474977348411,
"grad_norm": 0.27032724022865295,
"learning_rate": 2.66879005866608e-06,
"loss": 0.0222,
"step": 8170
},
{
"epoch": 0.38501818010660954,
"grad_norm": 0.2597672641277313,
"learning_rate": 2.648088610318278e-06,
"loss": 0.0197,
"step": 8180
},
{
"epoch": 0.38548886247837805,
"grad_norm": 0.22353407740592957,
"learning_rate": 2.6274555040159265e-06,
"loss": 0.0216,
"step": 8190
},
{
"epoch": 0.3859595448501465,
"grad_norm": 0.38809898495674133,
"learning_rate": 2.606890931560667e-06,
"loss": 0.0265,
"step": 8200
},
{
"epoch": 0.38643022722191495,
"grad_norm": 0.2147957682609558,
"learning_rate": 2.5863950841170704e-06,
"loss": 0.0226,
"step": 8210
},
{
"epoch": 0.38690090959368345,
"grad_norm": 0.29343053698539734,
"learning_rate": 2.5659681522108428e-06,
"loss": 0.026,
"step": 8220
},
{
"epoch": 0.3873715919654519,
"grad_norm": 0.3017103374004364,
"learning_rate": 2.5456103257270693e-06,
"loss": 0.0229,
"step": 8230
},
{
"epoch": 0.3878422743372204,
"grad_norm": 0.2678447663784027,
"learning_rate": 2.5253217939084407e-06,
"loss": 0.0236,
"step": 8240
},
{
"epoch": 0.38831295670898885,
"grad_norm": 0.31499728560447693,
"learning_rate": 2.505102745353499e-06,
"loss": 0.0225,
"step": 8250
},
{
"epoch": 0.3887836390807573,
"grad_norm": 0.29174795746803284,
"learning_rate": 2.4849533680148787e-06,
"loss": 0.029,
"step": 8260
},
{
"epoch": 0.3892543214525258,
"grad_norm": 0.22995388507843018,
"learning_rate": 2.4648738491975745e-06,
"loss": 0.0243,
"step": 8270
},
{
"epoch": 0.38972500382429426,
"grad_norm": 0.32718899846076965,
"learning_rate": 2.4448643755571687e-06,
"loss": 0.0234,
"step": 8280
},
{
"epoch": 0.39019568619606276,
"grad_norm": 0.2130356878042221,
"learning_rate": 2.424925133098137e-06,
"loss": 0.0212,
"step": 8290
},
{
"epoch": 0.3906663685678312,
"grad_norm": 0.32778653502464294,
"learning_rate": 2.4050563071720867e-06,
"loss": 0.0234,
"step": 8300
},
{
"epoch": 0.39113705093959966,
"grad_norm": 0.23867756128311157,
"learning_rate": 2.3852580824760487e-06,
"loss": 0.0253,
"step": 8310
},
{
"epoch": 0.39160773331136817,
"grad_norm": 0.3283313512802124,
"learning_rate": 2.3655306430507563e-06,
"loss": 0.0235,
"step": 8320
},
{
"epoch": 0.3920784156831366,
"grad_norm": 0.32051002979278564,
"learning_rate": 2.345874172278939e-06,
"loss": 0.0244,
"step": 8330
},
{
"epoch": 0.3925490980549051,
"grad_norm": 0.2831832468509674,
"learning_rate": 2.326288852883607e-06,
"loss": 0.0194,
"step": 8340
},
{
"epoch": 0.39301978042667357,
"grad_norm": 0.32168397307395935,
"learning_rate": 2.306774866926377e-06,
"loss": 0.0254,
"step": 8350
},
{
"epoch": 0.393490462798442,
"grad_norm": 0.31431373953819275,
"learning_rate": 2.287332395805737e-06,
"loss": 0.025,
"step": 8360
},
{
"epoch": 0.3939611451702105,
"grad_norm": 0.3645617961883545,
"learning_rate": 2.26796162025541e-06,
"loss": 0.0231,
"step": 8370
},
{
"epoch": 0.394431827541979,
"grad_norm": 0.3034353256225586,
"learning_rate": 2.248662720342637e-06,
"loss": 0.0221,
"step": 8380
},
{
"epoch": 0.3949025099137475,
"grad_norm": 0.2703491747379303,
"learning_rate": 2.229435875466519e-06,
"loss": 0.0226,
"step": 8390
},
{
"epoch": 0.39537319228551593,
"grad_norm": 0.27526670694351196,
"learning_rate": 2.2102812643563455e-06,
"loss": 0.024,
"step": 8400
},
{
"epoch": 0.3958438746572844,
"grad_norm": 0.31563958525657654,
"learning_rate": 2.191199065069941e-06,
"loss": 0.0241,
"step": 8410
},
{
"epoch": 0.3963145570290529,
"grad_norm": 0.2942217290401459,
"learning_rate": 2.1721894549919863e-06,
"loss": 0.0223,
"step": 8420
},
{
"epoch": 0.39678523940082133,
"grad_norm": 0.25802069902420044,
"learning_rate": 2.1532526108324047e-06,
"loss": 0.0245,
"step": 8430
},
{
"epoch": 0.39725592177258984,
"grad_norm": 0.2779555022716522,
"learning_rate": 2.1343887086246893e-06,
"loss": 0.0239,
"step": 8440
},
{
"epoch": 0.3977266041443583,
"grad_norm": 0.3036048114299774,
"learning_rate": 2.1155979237242817e-06,
"loss": 0.0189,
"step": 8450
},
{
"epoch": 0.39819728651612674,
"grad_norm": 0.2793290317058563,
"learning_rate": 2.0968804308069324e-06,
"loss": 0.0289,
"step": 8460
},
{
"epoch": 0.39866796888789524,
"grad_norm": 0.41800734400749207,
"learning_rate": 2.0782364038670986e-06,
"loss": 0.021,
"step": 8470
},
{
"epoch": 0.3991386512596637,
"grad_norm": 0.26266881823539734,
"learning_rate": 2.0596660162162872e-06,
"loss": 0.02,
"step": 8480
},
{
"epoch": 0.3996093336314322,
"grad_norm": 0.2599759101867676,
"learning_rate": 2.041169440481493e-06,
"loss": 0.0252,
"step": 8490
},
{
"epoch": 0.40008001600320064,
"grad_norm": 0.22514712810516357,
"learning_rate": 2.022746848603543e-06,
"loss": 0.0209,
"step": 8500
},
{
"epoch": 0.4005506983749691,
"grad_norm": 0.21918708086013794,
"learning_rate": 2.0043984118355464e-06,
"loss": 0.0251,
"step": 8510
},
{
"epoch": 0.4010213807467376,
"grad_norm": 0.23551422357559204,
"learning_rate": 1.986124300741267e-06,
"loss": 0.0241,
"step": 8520
},
{
"epoch": 0.40149206311850605,
"grad_norm": 0.23872265219688416,
"learning_rate": 1.967924685193552e-06,
"loss": 0.0219,
"step": 8530
},
{
"epoch": 0.4019627454902745,
"grad_norm": 0.24538487195968628,
"learning_rate": 1.9497997343727513e-06,
"loss": 0.0269,
"step": 8540
},
{
"epoch": 0.402433427862043,
"grad_norm": 0.25393593311309814,
"learning_rate": 1.9317496167651563e-06,
"loss": 0.0224,
"step": 8550
},
{
"epoch": 0.40290411023381145,
"grad_norm": 0.29849711060523987,
"learning_rate": 1.9137745001613984e-06,
"loss": 0.0241,
"step": 8560
},
{
"epoch": 0.40337479260557996,
"grad_norm": 0.23868460953235626,
"learning_rate": 1.8958745516549382e-06,
"loss": 0.0194,
"step": 8570
},
{
"epoch": 0.4038454749773484,
"grad_norm": 0.29939624667167664,
"learning_rate": 1.8780499376404715e-06,
"loss": 0.0246,
"step": 8580
},
{
"epoch": 0.40431615734911686,
"grad_norm": 0.3117296099662781,
"learning_rate": 1.8603008238124043e-06,
"loss": 0.019,
"step": 8590
},
{
"epoch": 0.40478683972088536,
"grad_norm": 0.3051111102104187,
"learning_rate": 1.842627375163305e-06,
"loss": 0.0264,
"step": 8600
},
{
"epoch": 0.4052575220926538,
"grad_norm": 0.2971876263618469,
"learning_rate": 1.8250297559823716e-06,
"loss": 0.0223,
"step": 8610
},
{
"epoch": 0.4057282044644223,
"grad_norm": 0.20871317386627197,
"learning_rate": 1.8075081298539032e-06,
"loss": 0.0209,
"step": 8620
},
{
"epoch": 0.40619888683619076,
"grad_norm": 0.2546001970767975,
"learning_rate": 1.7900626596557924e-06,
"loss": 0.0201,
"step": 8630
},
{
"epoch": 0.4066695692079592,
"grad_norm": 0.27642807364463806,
"learning_rate": 1.7726935075579798e-06,
"loss": 0.0211,
"step": 8640
},
{
"epoch": 0.4071402515797277,
"grad_norm": 0.2520013749599457,
"learning_rate": 1.7554008350209862e-06,
"loss": 0.0252,
"step": 8650
},
{
"epoch": 0.40761093395149617,
"grad_norm": 0.26341691613197327,
"learning_rate": 1.7381848027943815e-06,
"loss": 0.0251,
"step": 8660
},
{
"epoch": 0.4080816163232647,
"grad_norm": 0.5791468620300293,
"learning_rate": 1.721045570915304e-06,
"loss": 0.0243,
"step": 8670
},
{
"epoch": 0.4085522986950331,
"grad_norm": 0.2677948772907257,
"learning_rate": 1.703983298706966e-06,
"loss": 0.0213,
"step": 8680
},
{
"epoch": 0.40902298106680157,
"grad_norm": 0.2771756052970886,
"learning_rate": 1.6869981447771876e-06,
"loss": 0.0242,
"step": 8690
},
{
"epoch": 0.4094936634385701,
"grad_norm": 0.2840181589126587,
"learning_rate": 1.670090267016895e-06,
"loss": 0.0205,
"step": 8700
},
{
"epoch": 0.4099643458103385,
"grad_norm": 0.2857266664505005,
"learning_rate": 1.653259822598683e-06,
"loss": 0.0225,
"step": 8710
},
{
"epoch": 0.41043502818210703,
"grad_norm": 0.25589776039123535,
"learning_rate": 1.6365069679753331e-06,
"loss": 0.0169,
"step": 8720
},
{
"epoch": 0.4109057105538755,
"grad_norm": 0.2455807477235794,
"learning_rate": 1.619831858878368e-06,
"loss": 0.0208,
"step": 8730
},
{
"epoch": 0.41137639292564393,
"grad_norm": 0.26690465211868286,
"learning_rate": 1.6032346503166007e-06,
"loss": 0.0255,
"step": 8740
},
{
"epoch": 0.41184707529741243,
"grad_norm": 0.26888808608055115,
"learning_rate": 1.5867154965746956e-06,
"loss": 0.0258,
"step": 8750
},
{
"epoch": 0.4123177576691809,
"grad_norm": 0.2304418683052063,
"learning_rate": 1.5702745512117323e-06,
"loss": 0.0244,
"step": 8760
},
{
"epoch": 0.4127884400409494,
"grad_norm": 0.2621494233608246,
"learning_rate": 1.553911967059788e-06,
"loss": 0.0233,
"step": 8770
},
{
"epoch": 0.41325912241271784,
"grad_norm": 0.276109904050827,
"learning_rate": 1.537627896222489e-06,
"loss": 0.0229,
"step": 8780
},
{
"epoch": 0.4137298047844863,
"grad_norm": 0.22495102882385254,
"learning_rate": 1.5214224900736375e-06,
"loss": 0.0234,
"step": 8790
},
{
"epoch": 0.4142004871562548,
"grad_norm": 0.23699739575386047,
"learning_rate": 1.5052958992557687e-06,
"loss": 0.0266,
"step": 8800
},
{
"epoch": 0.41467116952802324,
"grad_norm": 0.25966712832450867,
"learning_rate": 1.4892482736787717e-06,
"loss": 0.0237,
"step": 8810
},
{
"epoch": 0.41514185189979175,
"grad_norm": 0.30642253160476685,
"learning_rate": 1.4732797625184814e-06,
"loss": 0.0221,
"step": 8820
},
{
"epoch": 0.4156125342715602,
"grad_norm": 0.2032870650291443,
"learning_rate": 1.4573905142153134e-06,
"loss": 0.0233,
"step": 8830
},
{
"epoch": 0.41608321664332865,
"grad_norm": 0.28423812985420227,
"learning_rate": 1.44158067647285e-06,
"loss": 0.0282,
"step": 8840
},
{
"epoch": 0.41655389901509715,
"grad_norm": 0.3622647821903229,
"learning_rate": 1.4258503962565096e-06,
"loss": 0.0262,
"step": 8850
},
{
"epoch": 0.4170245813868656,
"grad_norm": 0.2506396174430847,
"learning_rate": 1.4101998197921352e-06,
"loss": 0.0267,
"step": 8860
},
{
"epoch": 0.4174952637586341,
"grad_norm": 0.320178747177124,
"learning_rate": 1.3946290925646788e-06,
"loss": 0.0239,
"step": 8870
},
{
"epoch": 0.41796594613040255,
"grad_norm": 0.24572508037090302,
"learning_rate": 1.379138359316814e-06,
"loss": 0.0245,
"step": 8880
},
{
"epoch": 0.418436628502171,
"grad_norm": 0.25783976912498474,
"learning_rate": 1.363727764047612e-06,
"loss": 0.0219,
"step": 8890
},
{
"epoch": 0.4189073108739395,
"grad_norm": 0.2832587659358978,
"learning_rate": 1.3483974500111907e-06,
"loss": 0.0249,
"step": 8900
},
{
"epoch": 0.41937799324570796,
"grad_norm": 0.2859082520008087,
"learning_rate": 1.3331475597153988e-06,
"loss": 0.019,
"step": 8910
},
{
"epoch": 0.41984867561747646,
"grad_norm": 0.24576634168624878,
"learning_rate": 1.3179782349204618e-06,
"loss": 0.027,
"step": 8920
},
{
"epoch": 0.4203193579892449,
"grad_norm": 0.2457434982061386,
"learning_rate": 1.3028896166377003e-06,
"loss": 0.0213,
"step": 8930
},
{
"epoch": 0.42079004036101336,
"grad_norm": 0.2660965025424957,
"learning_rate": 1.2878818451281939e-06,
"loss": 0.0205,
"step": 8940
},
{
"epoch": 0.42126072273278187,
"grad_norm": 0.295940101146698,
"learning_rate": 1.2729550599014862e-06,
"loss": 0.0205,
"step": 8950
},
{
"epoch": 0.4217314051045503,
"grad_norm": 0.26837435364723206,
"learning_rate": 1.2581093997142846e-06,
"loss": 0.0219,
"step": 8960
},
{
"epoch": 0.42220208747631877,
"grad_norm": 0.29017436504364014,
"learning_rate": 1.2433450025691807e-06,
"loss": 0.0237,
"step": 8970
},
{
"epoch": 0.42267276984808727,
"grad_norm": 0.3649994432926178,
"learning_rate": 1.2286620057133459e-06,
"loss": 0.0212,
"step": 8980
},
{
"epoch": 0.4231434522198557,
"grad_norm": 0.26739194989204407,
"learning_rate": 1.2140605456372856e-06,
"loss": 0.0222,
"step": 8990
},
{
"epoch": 0.4236141345916242,
"grad_norm": 0.3192472755908966,
"learning_rate": 1.1995407580735364e-06,
"loss": 0.0221,
"step": 9000
},
{
"epoch": 0.4240848169633927,
"grad_norm": 0.24230466783046722,
"learning_rate": 1.1851027779954373e-06,
"loss": 0.0251,
"step": 9010
},
{
"epoch": 0.4245554993351611,
"grad_norm": 0.30759841203689575,
"learning_rate": 1.1707467396158524e-06,
"loss": 0.028,
"step": 9020
},
{
"epoch": 0.42502618170692963,
"grad_norm": 0.21116474270820618,
"learning_rate": 1.1564727763859306e-06,
"loss": 0.0203,
"step": 9030
},
{
"epoch": 0.4254968640786981,
"grad_norm": 0.3194906413555145,
"learning_rate": 1.1422810209938627e-06,
"loss": 0.0253,
"step": 9040
},
{
"epoch": 0.4259675464504666,
"grad_norm": 0.2964280843734741,
"learning_rate": 1.1281716053636616e-06,
"loss": 0.0254,
"step": 9050
},
{
"epoch": 0.42643822882223503,
"grad_norm": 0.2164221853017807,
"learning_rate": 1.1141446606539063e-06,
"loss": 0.0236,
"step": 9060
},
{
"epoch": 0.4269089111940035,
"grad_norm": 0.21685759723186493,
"learning_rate": 1.1002003172565579e-06,
"loss": 0.0213,
"step": 9070
},
{
"epoch": 0.427379593565772,
"grad_norm": 0.2615600526332855,
"learning_rate": 1.086338704795722e-06,
"loss": 0.0206,
"step": 9080
},
{
"epoch": 0.42785027593754044,
"grad_norm": 0.32146474719047546,
"learning_rate": 1.0725599521264518e-06,
"loss": 0.0228,
"step": 9090
},
{
"epoch": 0.42832095830930894,
"grad_norm": 0.3794270157814026,
"learning_rate": 1.0588641873335558e-06,
"loss": 0.0211,
"step": 9100
},
{
"epoch": 0.4287916406810774,
"grad_norm": 0.24591699242591858,
"learning_rate": 1.0452515377303974e-06,
"loss": 0.0251,
"step": 9110
},
{
"epoch": 0.42926232305284584,
"grad_norm": 0.27612215280532837,
"learning_rate": 1.0317221298577163e-06,
"loss": 0.0193,
"step": 9120
},
{
"epoch": 0.42973300542461434,
"grad_norm": 0.22137640416622162,
"learning_rate": 1.0182760894824607e-06,
"loss": 0.0199,
"step": 9130
},
{
"epoch": 0.4302036877963828,
"grad_norm": 0.27747103571891785,
"learning_rate": 1.0049135415965926e-06,
"loss": 0.0198,
"step": 9140
},
{
"epoch": 0.4306743701681513,
"grad_norm": 0.3355626165866852,
"learning_rate": 9.916346104159602e-07,
"loss": 0.0213,
"step": 9150
},
{
"epoch": 0.43114505253991975,
"grad_norm": 0.23147238790988922,
"learning_rate": 9.784394193791169e-07,
"loss": 0.0206,
"step": 9160
},
{
"epoch": 0.4316157349116882,
"grad_norm": 0.2864983081817627,
"learning_rate": 9.653280911461837e-07,
"loss": 0.0206,
"step": 9170
},
{
"epoch": 0.4320864172834567,
"grad_norm": 0.2785286009311676,
"learning_rate": 9.523007475977064e-07,
"loss": 0.0183,
"step": 9180
},
{
"epoch": 0.43255709965522515,
"grad_norm": 0.24063687026500702,
"learning_rate": 9.393575098335339e-07,
"loss": 0.0247,
"step": 9190
},
{
"epoch": 0.43302778202699366,
"grad_norm": 0.34387263655662537,
"learning_rate": 9.264984981716663e-07,
"loss": 0.026,
"step": 9200
},
{
"epoch": 0.4334984643987621,
"grad_norm": 0.26749640703201294,
"learning_rate": 9.137238321471675e-07,
"loss": 0.0258,
"step": 9210
},
{
"epoch": 0.43396914677053056,
"grad_norm": 0.24192747473716736,
"learning_rate": 9.010336305110345e-07,
"loss": 0.0211,
"step": 9220
},
{
"epoch": 0.43443982914229906,
"grad_norm": 0.29791685938835144,
"learning_rate": 8.884280112290977e-07,
"loss": 0.0231,
"step": 9230
},
{
"epoch": 0.4349105115140675,
"grad_norm": 0.24515554308891296,
"learning_rate": 8.759070914809253e-07,
"loss": 0.0197,
"step": 9240
},
{
"epoch": 0.435381193885836,
"grad_norm": 0.30276063084602356,
"learning_rate": 8.634709876587344e-07,
"loss": 0.023,
"step": 9250
},
{
"epoch": 0.43585187625760446,
"grad_norm": 0.5064191222190857,
"learning_rate": 8.511198153663069e-07,
"loss": 0.0238,
"step": 9260
},
{
"epoch": 0.4363225586293729,
"grad_norm": 0.2317410558462143,
"learning_rate": 8.388536894179234e-07,
"loss": 0.0187,
"step": 9270
},
{
"epoch": 0.4367932410011414,
"grad_norm": 0.2904147803783417,
"learning_rate": 8.266727238372763e-07,
"loss": 0.0203,
"step": 9280
},
{
"epoch": 0.43726392337290987,
"grad_norm": 0.3585156202316284,
"learning_rate": 8.145770318564361e-07,
"loss": 0.023,
"step": 9290
},
{
"epoch": 0.4377346057446784,
"grad_norm": 0.32069194316864014,
"learning_rate": 8.025667259147773e-07,
"loss": 0.0244,
"step": 9300
},
{
"epoch": 0.4382052881164468,
"grad_norm": 0.23717960715293884,
"learning_rate": 7.906419176579416e-07,
"loss": 0.0203,
"step": 9310
},
{
"epoch": 0.43867597048821527,
"grad_norm": 0.2986515164375305,
"learning_rate": 7.788027179367997e-07,
"loss": 0.0222,
"step": 9320
},
{
"epoch": 0.4391466528599838,
"grad_norm": 0.2489549219608307,
"learning_rate": 7.670492368064275e-07,
"loss": 0.0237,
"step": 9330
},
{
"epoch": 0.4396173352317522,
"grad_norm": 0.2565132677555084,
"learning_rate": 7.553815835250644e-07,
"loss": 0.0248,
"step": 9340
},
{
"epoch": 0.44008801760352073,
"grad_norm": 0.23007354140281677,
"learning_rate": 7.437998665531221e-07,
"loss": 0.0248,
"step": 9350
},
{
"epoch": 0.4405586999752892,
"grad_norm": 0.2531014680862427,
"learning_rate": 7.323041935521502e-07,
"loss": 0.0267,
"step": 9360
},
{
"epoch": 0.44102938234705763,
"grad_norm": 0.3419587016105652,
"learning_rate": 7.208946713838638e-07,
"loss": 0.0214,
"step": 9370
},
{
"epoch": 0.44150006471882614,
"grad_norm": 0.24581468105316162,
"learning_rate": 7.095714061091241e-07,
"loss": 0.0212,
"step": 9380
},
{
"epoch": 0.4419707470905946,
"grad_norm": 0.2114059180021286,
"learning_rate": 6.983345029869681e-07,
"loss": 0.0255,
"step": 9390
},
{
"epoch": 0.44244142946236303,
"grad_norm": 0.2334326058626175,
"learning_rate": 6.871840664736251e-07,
"loss": 0.02,
"step": 9400
},
{
"epoch": 0.44291211183413154,
"grad_norm": 0.2780289053916931,
"learning_rate": 6.761202002215506e-07,
"loss": 0.0194,
"step": 9410
},
{
"epoch": 0.4433827942059,
"grad_norm": 0.23214717209339142,
"learning_rate": 6.65143007078447e-07,
"loss": 0.0215,
"step": 9420
},
{
"epoch": 0.4438534765776685,
"grad_norm": 0.3396427631378174,
"learning_rate": 6.542525890863338e-07,
"loss": 0.0212,
"step": 9430
},
{
"epoch": 0.44432415894943694,
"grad_norm": 0.28709450364112854,
"learning_rate": 6.434490474805743e-07,
"loss": 0.0236,
"step": 9440
},
{
"epoch": 0.4447948413212054,
"grad_norm": 0.2774445414543152,
"learning_rate": 6.327324826889469e-07,
"loss": 0.0252,
"step": 9450
},
{
"epoch": 0.4452655236929739,
"grad_norm": 0.43127578496932983,
"learning_rate": 6.221029943307099e-07,
"loss": 0.0216,
"step": 9460
},
{
"epoch": 0.44573620606474235,
"grad_norm": 0.2494528889656067,
"learning_rate": 6.115606812156749e-07,
"loss": 0.0249,
"step": 9470
},
{
"epoch": 0.44620688843651085,
"grad_norm": 0.2812323570251465,
"learning_rate": 6.01105641343287e-07,
"loss": 0.0219,
"step": 9480
},
{
"epoch": 0.4466775708082793,
"grad_norm": 0.2776971459388733,
"learning_rate": 5.907379719017181e-07,
"loss": 0.0235,
"step": 9490
},
{
"epoch": 0.44714825318004775,
"grad_norm": 0.20903664827346802,
"learning_rate": 5.804577692669533e-07,
"loss": 0.0252,
"step": 9500
},
{
"epoch": 0.44761893555181625,
"grad_norm": 0.2666015028953552,
"learning_rate": 5.702651290019112e-07,
"loss": 0.0218,
"step": 9510
},
{
"epoch": 0.4480896179235847,
"grad_norm": 0.2835904359817505,
"learning_rate": 5.601601458555406e-07,
"loss": 0.0202,
"step": 9520
},
{
"epoch": 0.4485603002953532,
"grad_norm": 0.21979349851608276,
"learning_rate": 5.501429137619452e-07,
"loss": 0.0203,
"step": 9530
},
{
"epoch": 0.44903098266712166,
"grad_norm": 0.3161912262439728,
"learning_rate": 5.402135258395114e-07,
"loss": 0.0199,
"step": 9540
},
{
"epoch": 0.4495016650388901,
"grad_norm": 0.29377660155296326,
"learning_rate": 5.303720743900475e-07,
"loss": 0.0192,
"step": 9550
},
{
"epoch": 0.4499723474106586,
"grad_norm": 0.2755526900291443,
"learning_rate": 5.206186508979083e-07,
"loss": 0.0195,
"step": 9560
},
{
"epoch": 0.45044302978242706,
"grad_norm": 0.24950221180915833,
"learning_rate": 5.109533460291694e-07,
"loss": 0.0228,
"step": 9570
},
{
"epoch": 0.45091371215419557,
"grad_norm": 0.22505171597003937,
"learning_rate": 5.01376249630764e-07,
"loss": 0.0195,
"step": 9580
},
{
"epoch": 0.451384394525964,
"grad_norm": 0.3186216652393341,
"learning_rate": 4.918874507296578e-07,
"loss": 0.0214,
"step": 9590
},
{
"epoch": 0.45185507689773247,
"grad_norm": 0.32421979308128357,
"learning_rate": 4.824870375320156e-07,
"loss": 0.022,
"step": 9600
},
{
"epoch": 0.45232575926950097,
"grad_norm": 0.3264558017253876,
"learning_rate": 4.731750974223892e-07,
"loss": 0.0221,
"step": 9610
},
{
"epoch": 0.4527964416412694,
"grad_norm": 0.29782626032829285,
"learning_rate": 4.639517169628971e-07,
"loss": 0.0241,
"step": 9620
},
{
"epoch": 0.4532671240130379,
"grad_norm": 0.34734734892845154,
"learning_rate": 4.548169818924275e-07,
"loss": 0.024,
"step": 9630
},
{
"epoch": 0.4537378063848064,
"grad_norm": 0.2604312002658844,
"learning_rate": 4.4577097712582897e-07,
"loss": 0.0205,
"step": 9640
},
{
"epoch": 0.4542084887565748,
"grad_norm": 0.40874001383781433,
"learning_rate": 4.3681378675313747e-07,
"loss": 0.0189,
"step": 9650
},
{
"epoch": 0.45467917112834333,
"grad_norm": 0.2917439341545105,
"learning_rate": 4.279454940387828e-07,
"loss": 0.0186,
"step": 9660
},
{
"epoch": 0.4551498535001118,
"grad_norm": 0.2822345197200775,
"learning_rate": 4.191661814208181e-07,
"loss": 0.0236,
"step": 9670
},
{
"epoch": 0.4556205358718803,
"grad_norm": 0.2076219916343689,
"learning_rate": 4.1047593051015245e-07,
"loss": 0.0218,
"step": 9680
},
{
"epoch": 0.45609121824364873,
"grad_norm": 0.25493815541267395,
"learning_rate": 4.018748220897994e-07,
"loss": 0.0211,
"step": 9690
},
{
"epoch": 0.4565619006154172,
"grad_norm": 0.2581843435764313,
"learning_rate": 3.933629361141078e-07,
"loss": 0.0208,
"step": 9700
},
{
"epoch": 0.4570325829871857,
"grad_norm": 0.4367718994617462,
"learning_rate": 3.849403517080452e-07,
"loss": 0.0253,
"step": 9710
},
{
"epoch": 0.45750326535895414,
"grad_norm": 0.24213773012161255,
"learning_rate": 3.7660714716643563e-07,
"loss": 0.0205,
"step": 9720
},
{
"epoch": 0.45797394773072264,
"grad_norm": 0.28450652956962585,
"learning_rate": 3.683633999532521e-07,
"loss": 0.0267,
"step": 9730
},
{
"epoch": 0.4584446301024911,
"grad_norm": 0.2230147123336792,
"learning_rate": 3.60209186700885e-07,
"loss": 0.0174,
"step": 9740
},
{
"epoch": 0.45891531247425954,
"grad_norm": 0.2249242216348648,
"learning_rate": 3.521445832094328e-07,
"loss": 0.021,
"step": 9750
},
{
"epoch": 0.45938599484602805,
"grad_norm": 0.2801191508769989,
"learning_rate": 3.441696644459969e-07,
"loss": 0.0233,
"step": 9760
},
{
"epoch": 0.4598566772177965,
"grad_norm": 0.2613255977630615,
"learning_rate": 3.362845045439911e-07,
"loss": 0.0223,
"step": 9770
},
{
"epoch": 0.460327359589565,
"grad_norm": 0.2005629688501358,
"learning_rate": 3.284891768024401e-07,
"loss": 0.0256,
"step": 9780
},
{
"epoch": 0.46079804196133345,
"grad_norm": 0.2479463666677475,
"learning_rate": 3.207837536853087e-07,
"loss": 0.0237,
"step": 9790
},
{
"epoch": 0.4612687243331019,
"grad_norm": 0.2626476585865021,
"learning_rate": 3.131683068208247e-07,
"loss": 0.0205,
"step": 9800
},
{
"epoch": 0.4617394067048704,
"grad_norm": 0.24201330542564392,
"learning_rate": 3.0564290700081044e-07,
"loss": 0.0199,
"step": 9810
},
{
"epoch": 0.46221008907663885,
"grad_norm": 0.28845342993736267,
"learning_rate": 2.9820762418002916e-07,
"loss": 0.0228,
"step": 9820
},
{
"epoch": 0.4626807714484073,
"grad_norm": 0.31786325573921204,
"learning_rate": 2.908625274755339e-07,
"loss": 0.0236,
"step": 9830
},
{
"epoch": 0.4631514538201758,
"grad_norm": 0.19718733429908752,
"learning_rate": 2.8360768516601745e-07,
"loss": 0.0168,
"step": 9840
},
{
"epoch": 0.46362213619194426,
"grad_norm": 0.25268790125846863,
"learning_rate": 2.764431646911947e-07,
"loss": 0.0216,
"step": 9850
},
{
"epoch": 0.46409281856371276,
"grad_norm": 0.3371817171573639,
"learning_rate": 2.693690326511533e-07,
"loss": 0.0267,
"step": 9860
},
{
"epoch": 0.4645635009354812,
"grad_norm": 0.273394912481308,
"learning_rate": 2.6238535480575533e-07,
"loss": 0.0221,
"step": 9870
},
{
"epoch": 0.46503418330724966,
"grad_norm": 0.19925937056541443,
"learning_rate": 2.55492196074012e-07,
"loss": 0.0225,
"step": 9880
},
{
"epoch": 0.46550486567901816,
"grad_norm": 0.32583102583885193,
"learning_rate": 2.4868962053348764e-07,
"loss": 0.0226,
"step": 9890
},
{
"epoch": 0.4659755480507866,
"grad_norm": 0.24440275132656097,
"learning_rate": 2.419776914196981e-07,
"loss": 0.0224,
"step": 9900
},
{
"epoch": 0.4664462304225551,
"grad_norm": 0.2490750402212143,
"learning_rate": 2.3535647112553295e-07,
"loss": 0.0269,
"step": 9910
},
{
"epoch": 0.46691691279432357,
"grad_norm": 0.20322208106517792,
"learning_rate": 2.288260212006599e-07,
"loss": 0.0203,
"step": 9920
},
{
"epoch": 0.467387595166092,
"grad_norm": 0.2916705310344696,
"learning_rate": 2.2238640235097032e-07,
"loss": 0.0258,
"step": 9930
},
{
"epoch": 0.4678582775378605,
"grad_norm": 0.34367409348487854,
"learning_rate": 2.1603767443799994e-07,
"loss": 0.0229,
"step": 9940
},
{
"epoch": 0.468328959909629,
"grad_norm": 0.38834792375564575,
"learning_rate": 2.097798964783826e-07,
"loss": 0.0251,
"step": 9950
},
{
"epoch": 0.4687996422813975,
"grad_norm": 0.24248725175857544,
"learning_rate": 2.0361312664329502e-07,
"loss": 0.0208,
"step": 9960
},
{
"epoch": 0.4692703246531659,
"grad_norm": 0.23434412479400635,
"learning_rate": 1.975374222579207e-07,
"loss": 0.0232,
"step": 9970
},
{
"epoch": 0.4697410070249344,
"grad_norm": 0.36066076159477234,
"learning_rate": 1.9155283980091366e-07,
"loss": 0.0222,
"step": 9980
},
{
"epoch": 0.4702116893967029,
"grad_norm": 0.18803060054779053,
"learning_rate": 1.8565943490387761e-07,
"loss": 0.0235,
"step": 9990
},
{
"epoch": 0.47068237176847133,
"grad_norm": 0.22799311578273773,
"learning_rate": 1.79857262350841e-07,
"loss": 0.0199,
"step": 10000
},
{
"epoch": 0.47115305414023984,
"grad_norm": 0.2661643326282501,
"learning_rate": 1.741463760777584e-07,
"loss": 0.0219,
"step": 10010
},
{
"epoch": 0.4716237365120083,
"grad_norm": 0.25226396322250366,
"learning_rate": 1.685268291719999e-07,
"loss": 0.0195,
"step": 10020
},
{
"epoch": 0.47209441888377673,
"grad_norm": 0.29328832030296326,
"learning_rate": 1.6299867387186363e-07,
"loss": 0.0296,
"step": 10030
},
{
"epoch": 0.47256510125554524,
"grad_norm": 0.2222321629524231,
"learning_rate": 1.5756196156608393e-07,
"loss": 0.0233,
"step": 10040
},
{
"epoch": 0.4730357836273137,
"grad_norm": 0.18176062405109406,
"learning_rate": 1.5221674279336408e-07,
"loss": 0.0214,
"step": 10050
},
{
"epoch": 0.4735064659990822,
"grad_norm": 0.3411014676094055,
"learning_rate": 1.4696306724189312e-07,
"loss": 0.0266,
"step": 10060
},
{
"epoch": 0.47397714837085064,
"grad_norm": 0.2579866945743561,
"learning_rate": 1.4180098374889429e-07,
"loss": 0.0253,
"step": 10070
},
{
"epoch": 0.4744478307426191,
"grad_norm": 0.3391454517841339,
"learning_rate": 1.367305403001673e-07,
"loss": 0.0248,
"step": 10080
},
{
"epoch": 0.4749185131143876,
"grad_norm": 0.33902114629745483,
"learning_rate": 1.3175178402964116e-07,
"loss": 0.0241,
"step": 10090
},
{
"epoch": 0.47538919548615605,
"grad_norm": 0.31108054518699646,
"learning_rate": 1.2686476121894e-07,
"loss": 0.0235,
"step": 10100
},
{
"epoch": 0.47585987785792455,
"grad_norm": 0.23655517399311066,
"learning_rate": 1.220695172969477e-07,
"loss": 0.024,
"step": 10110
},
{
"epoch": 0.476330560229693,
"grad_norm": 0.25435346364974976,
"learning_rate": 1.1736609683938749e-07,
"loss": 0.0201,
"step": 10120
},
{
"epoch": 0.47680124260146145,
"grad_norm": 0.24031013250350952,
"learning_rate": 1.1275454356841298e-07,
"loss": 0.0252,
"step": 10130
},
{
"epoch": 0.47727192497322996,
"grad_norm": 0.29191461205482483,
"learning_rate": 1.0823490035218986e-07,
"loss": 0.0233,
"step": 10140
},
{
"epoch": 0.4777426073449984,
"grad_norm": 0.2831117808818817,
"learning_rate": 1.038072092045117e-07,
"loss": 0.0179,
"step": 10150
},
{
"epoch": 0.4782132897167669,
"grad_norm": 0.21434128284454346,
"learning_rate": 9.947151128439692e-08,
"loss": 0.025,
"step": 10160
},
{
"epoch": 0.47868397208853536,
"grad_norm": 0.27633315324783325,
"learning_rate": 9.52278468957124e-08,
"loss": 0.0256,
"step": 10170
},
{
"epoch": 0.4791546544603038,
"grad_norm": 0.2525457739830017,
"learning_rate": 9.107625548679944e-08,
"loss": 0.0209,
"step": 10180
},
{
"epoch": 0.4796253368320723,
"grad_norm": 0.25609150528907776,
"learning_rate": 8.701677565010725e-08,
"loss": 0.0252,
"step": 10190
},
{
"epoch": 0.48009601920384076,
"grad_norm": 0.28637197613716125,
"learning_rate": 8.304944512182666e-08,
"loss": 0.0198,
"step": 10200
},
{
"epoch": 0.4805667015756092,
"grad_norm": 0.29926639795303345,
"learning_rate": 7.91743007815493e-08,
"loss": 0.0246,
"step": 10210
},
{
"epoch": 0.4810373839473777,
"grad_norm": 0.28080400824546814,
"learning_rate": 7.539137865192003e-08,
"loss": 0.0222,
"step": 10220
},
{
"epoch": 0.48150806631914617,
"grad_norm": 0.29619300365448,
"learning_rate": 7.17007138983028e-08,
"loss": 0.0212,
"step": 10230
},
{
"epoch": 0.48197874869091467,
"grad_norm": 0.2888430953025818,
"learning_rate": 6.810234082845313e-08,
"loss": 0.0217,
"step": 10240
},
{
"epoch": 0.4824494310626831,
"grad_norm": 0.2783377766609192,
"learning_rate": 6.459629289219838e-08,
"loss": 0.0236,
"step": 10250
},
{
"epoch": 0.48292011343445157,
"grad_norm": 0.2683919370174408,
"learning_rate": 6.118260268112908e-08,
"loss": 0.0232,
"step": 10260
},
{
"epoch": 0.4833907958062201,
"grad_norm": 0.2633776068687439,
"learning_rate": 5.786130192829809e-08,
"loss": 0.0218,
"step": 10270
},
{
"epoch": 0.4838614781779885,
"grad_norm": 0.24633152782917023,
"learning_rate": 5.4632421507916366e-08,
"loss": 0.0218,
"step": 10280
},
{
"epoch": 0.48433216054975703,
"grad_norm": 0.30980628728866577,
"learning_rate": 5.1495991435076555e-08,
"loss": 0.0242,
"step": 10290
},
{
"epoch": 0.4848028429215255,
"grad_norm": 0.26017606258392334,
"learning_rate": 4.845204086546984e-08,
"loss": 0.0213,
"step": 10300
},
{
"epoch": 0.48527352529329393,
"grad_norm": 0.32232603430747986,
"learning_rate": 4.5500598095110645e-08,
"loss": 0.0229,
"step": 10310
},
{
"epoch": 0.48574420766506243,
"grad_norm": 0.3198448121547699,
"learning_rate": 4.264169056008016e-08,
"loss": 0.0237,
"step": 10320
},
{
"epoch": 0.4862148900368309,
"grad_norm": 0.2460068017244339,
"learning_rate": 3.987534483626987e-08,
"loss": 0.0207,
"step": 10330
},
{
"epoch": 0.4866855724085994,
"grad_norm": 0.28588202595710754,
"learning_rate": 3.720158663913065e-08,
"loss": 0.0223,
"step": 10340
},
{
"epoch": 0.48715625478036784,
"grad_norm": 0.2417328953742981,
"learning_rate": 3.4620440823438517e-08,
"loss": 0.0214,
"step": 10350
},
{
"epoch": 0.4876269371521363,
"grad_norm": 0.19336606562137604,
"learning_rate": 3.2131931383059256e-08,
"loss": 0.0221,
"step": 10360
},
{
"epoch": 0.4880976195239048,
"grad_norm": 0.32787322998046875,
"learning_rate": 2.9736081450730813e-08,
"loss": 0.0256,
"step": 10370
},
{
"epoch": 0.48856830189567324,
"grad_norm": 0.198124498128891,
"learning_rate": 2.7432913297841256e-08,
"loss": 0.019,
"step": 10380
},
{
"epoch": 0.48903898426744175,
"grad_norm": 0.21610353887081146,
"learning_rate": 2.5222448334227822e-08,
"loss": 0.0239,
"step": 10390
},
{
"epoch": 0.4895096666392102,
"grad_norm": 0.2960192859172821,
"learning_rate": 2.3104707107974857e-08,
"loss": 0.0202,
"step": 10400
},
{
"epoch": 0.48998034901097864,
"grad_norm": 0.3331698179244995,
"learning_rate": 2.1079709305226183e-08,
"loss": 0.0192,
"step": 10410
},
{
"epoch": 0.49045103138274715,
"grad_norm": 0.2205742597579956,
"learning_rate": 1.914747374999304e-08,
"loss": 0.023,
"step": 10420
},
{
"epoch": 0.4909217137545156,
"grad_norm": 0.2562515437602997,
"learning_rate": 1.7308018403991988e-08,
"loss": 0.0217,
"step": 10430
},
{
"epoch": 0.4913923961262841,
"grad_norm": 0.2539699375629425,
"learning_rate": 1.556136036646838e-08,
"loss": 0.0174,
"step": 10440
},
{
"epoch": 0.49186307849805255,
"grad_norm": 0.26354366540908813,
"learning_rate": 1.3907515874042044e-08,
"loss": 0.0193,
"step": 10450
},
{
"epoch": 0.492333760869821,
"grad_norm": 0.27616196870803833,
"learning_rate": 1.2346500300551844e-08,
"loss": 0.0191,
"step": 10460
},
{
"epoch": 0.4928044432415895,
"grad_norm": 0.25880300998687744,
"learning_rate": 1.0878328156919139e-08,
"loss": 0.0229,
"step": 10470
},
{
"epoch": 0.49327512561335796,
"grad_norm": 0.36059385538101196,
"learning_rate": 9.503013091006763e-09,
"loss": 0.0247,
"step": 10480
},
{
"epoch": 0.49374580798512646,
"grad_norm": 0.2314724326133728,
"learning_rate": 8.220567887498033e-09,
"loss": 0.0228,
"step": 10490
},
{
"epoch": 0.4942164903568949,
"grad_norm": 0.28190183639526367,
"learning_rate": 7.031004467771274e-09,
"loss": 0.0212,
"step": 10500
},
{
"epoch": 0.49468717272866336,
"grad_norm": 0.226776123046875,
"learning_rate": 5.934333889794364e-09,
"loss": 0.0196,
"step": 10510
},
{
"epoch": 0.49515785510043187,
"grad_norm": 0.30392614006996155,
"learning_rate": 4.93056634801925e-09,
"loss": 0.0205,
"step": 10520
},
{
"epoch": 0.4956285374722003,
"grad_norm": 0.32477688789367676,
"learning_rate": 4.019711173289809e-09,
"loss": 0.0262,
"step": 10530
},
{
"epoch": 0.4960992198439688,
"grad_norm": 0.3085278570652008,
"learning_rate": 3.201776832749692e-09,
"loss": 0.0267,
"step": 10540
},
{
"epoch": 0.49656990221573727,
"grad_norm": 0.24356020987033844,
"learning_rate": 2.4767709297712772e-09,
"loss": 0.0282,
"step": 10550
},
{
"epoch": 0.4970405845875057,
"grad_norm": 0.23326337337493896,
"learning_rate": 1.8447002038779471e-09,
"loss": 0.0209,
"step": 10560
},
{
"epoch": 0.4975112669592742,
"grad_norm": 0.21290835738182068,
"learning_rate": 1.305570530686362e-09,
"loss": 0.0196,
"step": 10570
},
{
"epoch": 0.4979819493310427,
"grad_norm": 0.33896124362945557,
"learning_rate": 8.593869218487261e-10,
"loss": 0.028,
"step": 10580
},
{
"epoch": 0.4984526317028112,
"grad_norm": 0.24787376821041107,
"learning_rate": 5.061535250061589e-10,
"loss": 0.0209,
"step": 10590
},
{
"epoch": 0.4989233140745796,
"grad_norm": 0.2760564982891083,
"learning_rate": 2.45873623754278e-10,
"loss": 0.023,
"step": 10600
},
{
"epoch": 0.4993939964463481,
"grad_norm": 0.17310141026973724,
"learning_rate": 7.85496376076722e-11,
"loss": 0.0191,
"step": 10610
},
{
"epoch": 0.4998646788181166,
"grad_norm": 0.2157493531703949,
"learning_rate": 4.1831219799171044e-12,
"loss": 0.0208,
"step": 10620
},
{
"epoch": 0.5000058835296471,
"step": 10623,
"total_flos": 1.494988181940994e+18,
"train_loss": 0.03402505366156737,
"train_runtime": 46532.7758,
"train_samples_per_second": 1.826,
"train_steps_per_second": 0.228
}
],
"logging_steps": 10,
"max_steps": 10623,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"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": 1.494988181940994e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}