TemporaryLikeAchilles / trainer_state.json
Ericu950's picture
Upload folder using huggingface_hub
2503bd8 verified
raw
history blame contribute delete
No virus
226 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.1752854264607118,
"eval_steps": 500,
"global_step": 14000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 0.30050432682037354,
"learning_rate": 9.995802552048355e-06,
"loss": 0.0015,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 0.02444104291498661,
"learning_rate": 9.99160510409671e-06,
"loss": 0.0012,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 0.30079442262649536,
"learning_rate": 9.987407656145065e-06,
"loss": 0.0022,
"step": 30
},
{
"epoch": 0.0,
"grad_norm": 0.11910504847764969,
"learning_rate": 9.98321020819342e-06,
"loss": 0.0044,
"step": 40
},
{
"epoch": 0.0,
"grad_norm": 0.025635888800024986,
"learning_rate": 9.979012760241774e-06,
"loss": 0.004,
"step": 50
},
{
"epoch": 0.01,
"grad_norm": 0.10340698063373566,
"learning_rate": 9.974815312290129e-06,
"loss": 0.0034,
"step": 60
},
{
"epoch": 0.01,
"grad_norm": 0.04860584810376167,
"learning_rate": 9.970617864338483e-06,
"loss": 0.0007,
"step": 70
},
{
"epoch": 0.01,
"grad_norm": 0.029318638145923615,
"learning_rate": 9.966420416386838e-06,
"loss": 0.0051,
"step": 80
},
{
"epoch": 0.01,
"grad_norm": 0.023802747949957848,
"learning_rate": 9.962222968435192e-06,
"loss": 0.0008,
"step": 90
},
{
"epoch": 0.01,
"grad_norm": 0.06850480288267136,
"learning_rate": 9.958025520483547e-06,
"loss": 0.0037,
"step": 100
},
{
"epoch": 0.01,
"grad_norm": 0.021825846284627914,
"learning_rate": 9.953828072531902e-06,
"loss": 0.0017,
"step": 110
},
{
"epoch": 0.01,
"grad_norm": 0.03794211894273758,
"learning_rate": 9.949630624580256e-06,
"loss": 0.0013,
"step": 120
},
{
"epoch": 0.01,
"grad_norm": 0.10325329005718231,
"learning_rate": 9.945433176628611e-06,
"loss": 0.0026,
"step": 130
},
{
"epoch": 0.01,
"grad_norm": 0.03597491234540939,
"learning_rate": 9.941235728676965e-06,
"loss": 0.0034,
"step": 140
},
{
"epoch": 0.01,
"grad_norm": 0.23858389258384705,
"learning_rate": 9.93703828072532e-06,
"loss": 0.0024,
"step": 150
},
{
"epoch": 0.01,
"grad_norm": 0.15489879250526428,
"learning_rate": 9.932840832773675e-06,
"loss": 0.0092,
"step": 160
},
{
"epoch": 0.01,
"grad_norm": 0.07030558586120605,
"learning_rate": 9.92864338482203e-06,
"loss": 0.0022,
"step": 170
},
{
"epoch": 0.02,
"grad_norm": 0.5954773426055908,
"learning_rate": 9.924445936870384e-06,
"loss": 0.0032,
"step": 180
},
{
"epoch": 0.02,
"grad_norm": 0.19402949512004852,
"learning_rate": 9.920248488918739e-06,
"loss": 0.0018,
"step": 190
},
{
"epoch": 0.02,
"grad_norm": 0.0862368643283844,
"learning_rate": 9.916051040967093e-06,
"loss": 0.0041,
"step": 200
},
{
"epoch": 0.02,
"grad_norm": 0.02818392589688301,
"learning_rate": 9.911853593015446e-06,
"loss": 0.0035,
"step": 210
},
{
"epoch": 0.02,
"grad_norm": 0.35951200127601624,
"learning_rate": 9.907656145063802e-06,
"loss": 0.0027,
"step": 220
},
{
"epoch": 0.02,
"grad_norm": 0.033146314322948456,
"learning_rate": 9.903458697112157e-06,
"loss": 0.0023,
"step": 230
},
{
"epoch": 0.02,
"grad_norm": 0.183016836643219,
"learning_rate": 9.899261249160512e-06,
"loss": 0.0029,
"step": 240
},
{
"epoch": 0.02,
"grad_norm": 0.010143009014427662,
"learning_rate": 9.895063801208866e-06,
"loss": 0.0038,
"step": 250
},
{
"epoch": 0.02,
"grad_norm": 0.01728486269712448,
"learning_rate": 9.890866353257221e-06,
"loss": 0.0016,
"step": 260
},
{
"epoch": 0.02,
"grad_norm": 0.016598215326666832,
"learning_rate": 9.886668905305575e-06,
"loss": 0.0021,
"step": 270
},
{
"epoch": 0.02,
"grad_norm": 0.010375693440437317,
"learning_rate": 9.88247145735393e-06,
"loss": 0.0028,
"step": 280
},
{
"epoch": 0.02,
"grad_norm": 0.0275142602622509,
"learning_rate": 9.878274009402285e-06,
"loss": 0.0058,
"step": 290
},
{
"epoch": 0.03,
"grad_norm": 0.3856583833694458,
"learning_rate": 9.874076561450638e-06,
"loss": 0.0029,
"step": 300
},
{
"epoch": 0.03,
"grad_norm": 0.01894170604646206,
"learning_rate": 9.869879113498994e-06,
"loss": 0.0016,
"step": 310
},
{
"epoch": 0.03,
"grad_norm": 0.33644765615463257,
"learning_rate": 9.865681665547347e-06,
"loss": 0.0016,
"step": 320
},
{
"epoch": 0.03,
"grad_norm": 0.015369775705039501,
"learning_rate": 9.861484217595703e-06,
"loss": 0.0018,
"step": 330
},
{
"epoch": 0.03,
"grad_norm": 0.027933355420827866,
"learning_rate": 9.857286769644058e-06,
"loss": 0.0024,
"step": 340
},
{
"epoch": 0.03,
"grad_norm": 0.006843829993158579,
"learning_rate": 9.853089321692412e-06,
"loss": 0.0096,
"step": 350
},
{
"epoch": 0.03,
"grad_norm": 0.008493180386722088,
"learning_rate": 9.848891873740767e-06,
"loss": 0.0031,
"step": 360
},
{
"epoch": 0.03,
"grad_norm": 0.6798232197761536,
"learning_rate": 9.84469442578912e-06,
"loss": 0.0049,
"step": 370
},
{
"epoch": 0.03,
"grad_norm": 0.0329626239836216,
"learning_rate": 9.840496977837476e-06,
"loss": 0.0016,
"step": 380
},
{
"epoch": 0.03,
"grad_norm": 0.015625745058059692,
"learning_rate": 9.83629952988583e-06,
"loss": 0.0008,
"step": 390
},
{
"epoch": 0.03,
"grad_norm": 0.024704214185476303,
"learning_rate": 9.832102081934186e-06,
"loss": 0.0054,
"step": 400
},
{
"epoch": 0.03,
"grad_norm": 0.6314619183540344,
"learning_rate": 9.827904633982538e-06,
"loss": 0.0021,
"step": 410
},
{
"epoch": 0.04,
"grad_norm": 0.46060019731521606,
"learning_rate": 9.823707186030895e-06,
"loss": 0.0057,
"step": 420
},
{
"epoch": 0.04,
"grad_norm": 0.05328000336885452,
"learning_rate": 9.819509738079248e-06,
"loss": 0.0009,
"step": 430
},
{
"epoch": 0.04,
"grad_norm": 0.22446390986442566,
"learning_rate": 9.815312290127604e-06,
"loss": 0.0011,
"step": 440
},
{
"epoch": 0.04,
"grad_norm": 0.035243839025497437,
"learning_rate": 9.811114842175959e-06,
"loss": 0.0009,
"step": 450
},
{
"epoch": 0.04,
"grad_norm": 0.06827638298273087,
"learning_rate": 9.806917394224312e-06,
"loss": 0.0004,
"step": 460
},
{
"epoch": 0.04,
"grad_norm": 0.08792345225811005,
"learning_rate": 9.802719946272668e-06,
"loss": 0.0031,
"step": 470
},
{
"epoch": 0.04,
"grad_norm": 0.00334025826305151,
"learning_rate": 9.79852249832102e-06,
"loss": 0.0043,
"step": 480
},
{
"epoch": 0.04,
"grad_norm": 0.03577791526913643,
"learning_rate": 9.794325050369377e-06,
"loss": 0.0009,
"step": 490
},
{
"epoch": 0.04,
"grad_norm": 0.10479724407196045,
"learning_rate": 9.79012760241773e-06,
"loss": 0.0009,
"step": 500
},
{
"epoch": 0.04,
"grad_norm": 0.006824735552072525,
"learning_rate": 9.785930154466086e-06,
"loss": 0.0044,
"step": 510
},
{
"epoch": 0.04,
"grad_norm": 0.003342829644680023,
"learning_rate": 9.78173270651444e-06,
"loss": 0.0029,
"step": 520
},
{
"epoch": 0.04,
"grad_norm": 0.010192741639912128,
"learning_rate": 9.777535258562796e-06,
"loss": 0.0022,
"step": 530
},
{
"epoch": 0.05,
"grad_norm": 0.08739089965820312,
"learning_rate": 9.773337810611148e-06,
"loss": 0.0032,
"step": 540
},
{
"epoch": 0.05,
"grad_norm": 0.3144104480743408,
"learning_rate": 9.769140362659503e-06,
"loss": 0.0015,
"step": 550
},
{
"epoch": 0.05,
"grad_norm": 0.0013409241801127791,
"learning_rate": 9.76494291470786e-06,
"loss": 0.0043,
"step": 560
},
{
"epoch": 0.05,
"grad_norm": 0.004873868077993393,
"learning_rate": 9.760745466756212e-06,
"loss": 0.0023,
"step": 570
},
{
"epoch": 0.05,
"grad_norm": 0.13740317523479462,
"learning_rate": 9.756548018804569e-06,
"loss": 0.0028,
"step": 580
},
{
"epoch": 0.05,
"grad_norm": 0.1480674296617508,
"learning_rate": 9.752350570852922e-06,
"loss": 0.0034,
"step": 590
},
{
"epoch": 0.05,
"grad_norm": 0.01920839212834835,
"learning_rate": 9.748153122901278e-06,
"loss": 0.0009,
"step": 600
},
{
"epoch": 0.05,
"grad_norm": 0.06275931745767593,
"learning_rate": 9.74395567494963e-06,
"loss": 0.0011,
"step": 610
},
{
"epoch": 0.05,
"grad_norm": 0.4199993312358856,
"learning_rate": 9.739758226997985e-06,
"loss": 0.0048,
"step": 620
},
{
"epoch": 0.05,
"grad_norm": 0.022090768441557884,
"learning_rate": 9.73556077904634e-06,
"loss": 0.0048,
"step": 630
},
{
"epoch": 0.05,
"grad_norm": 0.1497245728969574,
"learning_rate": 9.731363331094695e-06,
"loss": 0.0142,
"step": 640
},
{
"epoch": 0.05,
"grad_norm": 0.011996796354651451,
"learning_rate": 9.72716588314305e-06,
"loss": 0.0034,
"step": 650
},
{
"epoch": 0.06,
"grad_norm": 0.004964240826666355,
"learning_rate": 9.722968435191404e-06,
"loss": 0.0024,
"step": 660
},
{
"epoch": 0.06,
"grad_norm": 0.1952650249004364,
"learning_rate": 9.71877098723976e-06,
"loss": 0.0091,
"step": 670
},
{
"epoch": 0.06,
"grad_norm": 0.2123202085494995,
"learning_rate": 9.714573539288113e-06,
"loss": 0.0064,
"step": 680
},
{
"epoch": 0.06,
"grad_norm": 0.030770471319556236,
"learning_rate": 9.71037609133647e-06,
"loss": 0.0012,
"step": 690
},
{
"epoch": 0.06,
"grad_norm": 0.06245679035782814,
"learning_rate": 9.706178643384822e-06,
"loss": 0.0022,
"step": 700
},
{
"epoch": 0.06,
"grad_norm": 0.09754525870084763,
"learning_rate": 9.701981195433177e-06,
"loss": 0.0016,
"step": 710
},
{
"epoch": 0.06,
"grad_norm": 0.2677933871746063,
"learning_rate": 9.697783747481532e-06,
"loss": 0.0012,
"step": 720
},
{
"epoch": 0.06,
"grad_norm": 0.01193680427968502,
"learning_rate": 9.693586299529886e-06,
"loss": 0.0021,
"step": 730
},
{
"epoch": 0.06,
"grad_norm": 0.00777805782854557,
"learning_rate": 9.68938885157824e-06,
"loss": 0.0005,
"step": 740
},
{
"epoch": 0.06,
"grad_norm": 0.0012337628286331892,
"learning_rate": 9.685191403626595e-06,
"loss": 0.0026,
"step": 750
},
{
"epoch": 0.06,
"grad_norm": 0.009749867022037506,
"learning_rate": 9.68099395567495e-06,
"loss": 0.0089,
"step": 760
},
{
"epoch": 0.06,
"grad_norm": 0.05441896617412567,
"learning_rate": 9.676796507723305e-06,
"loss": 0.0051,
"step": 770
},
{
"epoch": 0.07,
"grad_norm": 0.2501550614833832,
"learning_rate": 9.67259905977166e-06,
"loss": 0.003,
"step": 780
},
{
"epoch": 0.07,
"grad_norm": 0.030330829322338104,
"learning_rate": 9.668401611820014e-06,
"loss": 0.0026,
"step": 790
},
{
"epoch": 0.07,
"grad_norm": 0.02893720008432865,
"learning_rate": 9.664204163868369e-06,
"loss": 0.0012,
"step": 800
},
{
"epoch": 0.07,
"grad_norm": 0.016928590834140778,
"learning_rate": 9.660006715916723e-06,
"loss": 0.0095,
"step": 810
},
{
"epoch": 0.07,
"grad_norm": 0.3214503824710846,
"learning_rate": 9.655809267965078e-06,
"loss": 0.0014,
"step": 820
},
{
"epoch": 0.07,
"grad_norm": 0.005326243583112955,
"learning_rate": 9.651611820013432e-06,
"loss": 0.0026,
"step": 830
},
{
"epoch": 0.07,
"grad_norm": 0.006249762140214443,
"learning_rate": 9.647414372061787e-06,
"loss": 0.0012,
"step": 840
},
{
"epoch": 0.07,
"grad_norm": 0.005905636586248875,
"learning_rate": 9.643216924110142e-06,
"loss": 0.0039,
"step": 850
},
{
"epoch": 0.07,
"grad_norm": 0.4915952682495117,
"learning_rate": 9.639019476158496e-06,
"loss": 0.0029,
"step": 860
},
{
"epoch": 0.07,
"grad_norm": 0.039309654384851456,
"learning_rate": 9.634822028206851e-06,
"loss": 0.0016,
"step": 870
},
{
"epoch": 0.07,
"grad_norm": 0.25026625394821167,
"learning_rate": 9.630624580255205e-06,
"loss": 0.0027,
"step": 880
},
{
"epoch": 0.07,
"grad_norm": 0.059997640550136566,
"learning_rate": 9.62642713230356e-06,
"loss": 0.0031,
"step": 890
},
{
"epoch": 0.08,
"grad_norm": 0.03755790740251541,
"learning_rate": 9.622229684351915e-06,
"loss": 0.0023,
"step": 900
},
{
"epoch": 0.08,
"grad_norm": 0.04567611590027809,
"learning_rate": 9.61803223640027e-06,
"loss": 0.0027,
"step": 910
},
{
"epoch": 0.08,
"grad_norm": 0.13617774844169617,
"learning_rate": 9.613834788448624e-06,
"loss": 0.0023,
"step": 920
},
{
"epoch": 0.08,
"grad_norm": 0.007796242367476225,
"learning_rate": 9.609637340496979e-06,
"loss": 0.0008,
"step": 930
},
{
"epoch": 0.08,
"grad_norm": 0.09431927651166916,
"learning_rate": 9.605439892545333e-06,
"loss": 0.0043,
"step": 940
},
{
"epoch": 0.08,
"grad_norm": 0.00464077852666378,
"learning_rate": 9.601242444593688e-06,
"loss": 0.0032,
"step": 950
},
{
"epoch": 0.08,
"grad_norm": 0.07417366653680801,
"learning_rate": 9.597044996642042e-06,
"loss": 0.004,
"step": 960
},
{
"epoch": 0.08,
"grad_norm": 0.07863020896911621,
"learning_rate": 9.592847548690397e-06,
"loss": 0.0038,
"step": 970
},
{
"epoch": 0.08,
"grad_norm": 0.06590215861797333,
"learning_rate": 9.588650100738752e-06,
"loss": 0.0051,
"step": 980
},
{
"epoch": 0.08,
"grad_norm": 0.5270329117774963,
"learning_rate": 9.584452652787106e-06,
"loss": 0.0081,
"step": 990
},
{
"epoch": 0.08,
"grad_norm": 0.005750787910073996,
"learning_rate": 9.580255204835461e-06,
"loss": 0.002,
"step": 1000
},
{
"epoch": 0.08,
"grad_norm": 0.4699457883834839,
"learning_rate": 9.576057756883816e-06,
"loss": 0.0025,
"step": 1010
},
{
"epoch": 0.09,
"grad_norm": 0.010430064052343369,
"learning_rate": 9.57186030893217e-06,
"loss": 0.0013,
"step": 1020
},
{
"epoch": 0.09,
"grad_norm": 0.006583033595234156,
"learning_rate": 9.567662860980525e-06,
"loss": 0.0027,
"step": 1030
},
{
"epoch": 0.09,
"grad_norm": 0.01882992312312126,
"learning_rate": 9.56346541302888e-06,
"loss": 0.0046,
"step": 1040
},
{
"epoch": 0.09,
"grad_norm": 0.06240439414978027,
"learning_rate": 9.559267965077234e-06,
"loss": 0.0054,
"step": 1050
},
{
"epoch": 0.09,
"grad_norm": 0.872185468673706,
"learning_rate": 9.555070517125589e-06,
"loss": 0.0046,
"step": 1060
},
{
"epoch": 0.09,
"grad_norm": 0.18814022839069366,
"learning_rate": 9.550873069173943e-06,
"loss": 0.003,
"step": 1070
},
{
"epoch": 0.09,
"grad_norm": 0.18875594437122345,
"learning_rate": 9.546675621222298e-06,
"loss": 0.0031,
"step": 1080
},
{
"epoch": 0.09,
"grad_norm": 0.03390906751155853,
"learning_rate": 9.542478173270652e-06,
"loss": 0.0027,
"step": 1090
},
{
"epoch": 0.09,
"grad_norm": 0.1369648426771164,
"learning_rate": 9.538280725319007e-06,
"loss": 0.0016,
"step": 1100
},
{
"epoch": 0.09,
"grad_norm": 0.02575293742120266,
"learning_rate": 9.534083277367362e-06,
"loss": 0.0011,
"step": 1110
},
{
"epoch": 0.09,
"grad_norm": 0.043303512036800385,
"learning_rate": 9.529885829415716e-06,
"loss": 0.0015,
"step": 1120
},
{
"epoch": 0.09,
"grad_norm": 0.0034562412183731794,
"learning_rate": 9.525688381464071e-06,
"loss": 0.0068,
"step": 1130
},
{
"epoch": 0.1,
"grad_norm": 0.10729283839464188,
"learning_rate": 9.521490933512426e-06,
"loss": 0.0017,
"step": 1140
},
{
"epoch": 0.1,
"grad_norm": 0.029321426525712013,
"learning_rate": 9.51729348556078e-06,
"loss": 0.0005,
"step": 1150
},
{
"epoch": 0.1,
"grad_norm": 0.26224708557128906,
"learning_rate": 9.513096037609135e-06,
"loss": 0.0056,
"step": 1160
},
{
"epoch": 0.1,
"grad_norm": 0.081400066614151,
"learning_rate": 9.50889858965749e-06,
"loss": 0.0012,
"step": 1170
},
{
"epoch": 0.1,
"grad_norm": 0.06388130784034729,
"learning_rate": 9.504701141705844e-06,
"loss": 0.0035,
"step": 1180
},
{
"epoch": 0.1,
"grad_norm": 0.03783797100186348,
"learning_rate": 9.500503693754199e-06,
"loss": 0.0028,
"step": 1190
},
{
"epoch": 0.1,
"grad_norm": 0.0943446159362793,
"learning_rate": 9.496306245802552e-06,
"loss": 0.0015,
"step": 1200
},
{
"epoch": 0.1,
"grad_norm": 0.4624231159687042,
"learning_rate": 9.492108797850908e-06,
"loss": 0.0035,
"step": 1210
},
{
"epoch": 0.1,
"grad_norm": 0.06161864846944809,
"learning_rate": 9.487911349899263e-06,
"loss": 0.0064,
"step": 1220
},
{
"epoch": 0.1,
"grad_norm": 0.027636563405394554,
"learning_rate": 9.483713901947617e-06,
"loss": 0.001,
"step": 1230
},
{
"epoch": 0.1,
"grad_norm": 0.06516788899898529,
"learning_rate": 9.479516453995972e-06,
"loss": 0.0044,
"step": 1240
},
{
"epoch": 0.1,
"grad_norm": 0.0022635594941675663,
"learning_rate": 9.475319006044326e-06,
"loss": 0.004,
"step": 1250
},
{
"epoch": 0.11,
"grad_norm": 0.016013437882065773,
"learning_rate": 9.471121558092681e-06,
"loss": 0.001,
"step": 1260
},
{
"epoch": 0.11,
"grad_norm": 0.018587607890367508,
"learning_rate": 9.466924110141034e-06,
"loss": 0.0022,
"step": 1270
},
{
"epoch": 0.11,
"grad_norm": 0.13551315665245056,
"learning_rate": 9.46272666218939e-06,
"loss": 0.0009,
"step": 1280
},
{
"epoch": 0.11,
"grad_norm": 0.016198860481381416,
"learning_rate": 9.458529214237743e-06,
"loss": 0.0034,
"step": 1290
},
{
"epoch": 0.11,
"grad_norm": 0.09964737296104431,
"learning_rate": 9.4543317662861e-06,
"loss": 0.0016,
"step": 1300
},
{
"epoch": 0.11,
"grad_norm": 0.01190153881907463,
"learning_rate": 9.450134318334452e-06,
"loss": 0.0023,
"step": 1310
},
{
"epoch": 0.11,
"grad_norm": 1.2369023561477661,
"learning_rate": 9.445936870382809e-06,
"loss": 0.01,
"step": 1320
},
{
"epoch": 0.11,
"grad_norm": 0.024809740483760834,
"learning_rate": 9.441739422431163e-06,
"loss": 0.001,
"step": 1330
},
{
"epoch": 0.11,
"grad_norm": 0.15385492146015167,
"learning_rate": 9.437541974479518e-06,
"loss": 0.0018,
"step": 1340
},
{
"epoch": 0.11,
"grad_norm": 0.12145934998989105,
"learning_rate": 9.433344526527873e-06,
"loss": 0.0018,
"step": 1350
},
{
"epoch": 0.11,
"grad_norm": 0.04905625805258751,
"learning_rate": 9.429147078576225e-06,
"loss": 0.0036,
"step": 1360
},
{
"epoch": 0.12,
"grad_norm": 0.03543395921587944,
"learning_rate": 9.424949630624582e-06,
"loss": 0.0015,
"step": 1370
},
{
"epoch": 0.12,
"grad_norm": 1.0014148950576782,
"learning_rate": 9.420752182672935e-06,
"loss": 0.005,
"step": 1380
},
{
"epoch": 0.12,
"grad_norm": 0.02649211883544922,
"learning_rate": 9.416554734721291e-06,
"loss": 0.0013,
"step": 1390
},
{
"epoch": 0.12,
"grad_norm": 0.03208385780453682,
"learning_rate": 9.412357286769644e-06,
"loss": 0.0007,
"step": 1400
},
{
"epoch": 0.12,
"grad_norm": 0.0064937155693769455,
"learning_rate": 9.408159838818e-06,
"loss": 0.0013,
"step": 1410
},
{
"epoch": 0.12,
"grad_norm": 0.294610857963562,
"learning_rate": 9.403962390866353e-06,
"loss": 0.002,
"step": 1420
},
{
"epoch": 0.12,
"grad_norm": 0.04225017875432968,
"learning_rate": 9.39976494291471e-06,
"loss": 0.0011,
"step": 1430
},
{
"epoch": 0.12,
"grad_norm": 0.17220620810985565,
"learning_rate": 9.395567494963064e-06,
"loss": 0.0017,
"step": 1440
},
{
"epoch": 0.12,
"grad_norm": 0.01658468507230282,
"learning_rate": 9.391370047011417e-06,
"loss": 0.004,
"step": 1450
},
{
"epoch": 0.12,
"grad_norm": 0.09207724779844284,
"learning_rate": 9.387172599059773e-06,
"loss": 0.0033,
"step": 1460
},
{
"epoch": 0.12,
"grad_norm": 0.0661146268248558,
"learning_rate": 9.382975151108126e-06,
"loss": 0.0016,
"step": 1470
},
{
"epoch": 0.12,
"grad_norm": 0.012653768993914127,
"learning_rate": 9.378777703156483e-06,
"loss": 0.0003,
"step": 1480
},
{
"epoch": 0.13,
"grad_norm": 0.038489799946546555,
"learning_rate": 9.374580255204835e-06,
"loss": 0.0009,
"step": 1490
},
{
"epoch": 0.13,
"grad_norm": 0.14739686250686646,
"learning_rate": 9.370382807253192e-06,
"loss": 0.004,
"step": 1500
},
{
"epoch": 0.13,
"grad_norm": 0.04394904151558876,
"learning_rate": 9.366185359301545e-06,
"loss": 0.0025,
"step": 1510
},
{
"epoch": 0.13,
"grad_norm": 0.2980542778968811,
"learning_rate": 9.3619879113499e-06,
"loss": 0.0049,
"step": 1520
},
{
"epoch": 0.13,
"grad_norm": 0.6049527525901794,
"learning_rate": 9.357790463398254e-06,
"loss": 0.0068,
"step": 1530
},
{
"epoch": 0.13,
"grad_norm": 0.2390802502632141,
"learning_rate": 9.353593015446609e-06,
"loss": 0.0014,
"step": 1540
},
{
"epoch": 0.13,
"grad_norm": 0.057861894369125366,
"learning_rate": 9.349395567494965e-06,
"loss": 0.0007,
"step": 1550
},
{
"epoch": 0.13,
"grad_norm": 0.010473741218447685,
"learning_rate": 9.345198119543318e-06,
"loss": 0.001,
"step": 1560
},
{
"epoch": 0.13,
"grad_norm": 0.6854666471481323,
"learning_rate": 9.341000671591674e-06,
"loss": 0.0037,
"step": 1570
},
{
"epoch": 0.13,
"grad_norm": 0.28555673360824585,
"learning_rate": 9.336803223640027e-06,
"loss": 0.0015,
"step": 1580
},
{
"epoch": 0.13,
"grad_norm": 0.05343065410852432,
"learning_rate": 9.332605775688383e-06,
"loss": 0.0011,
"step": 1590
},
{
"epoch": 0.13,
"grad_norm": 0.040986571460962296,
"learning_rate": 9.328408327736736e-06,
"loss": 0.0019,
"step": 1600
},
{
"epoch": 0.14,
"grad_norm": 0.09061778336763382,
"learning_rate": 9.324210879785091e-06,
"loss": 0.0009,
"step": 1610
},
{
"epoch": 0.14,
"grad_norm": 0.022660354152321815,
"learning_rate": 9.320013431833446e-06,
"loss": 0.0007,
"step": 1620
},
{
"epoch": 0.14,
"grad_norm": 0.019222592934966087,
"learning_rate": 9.3158159838818e-06,
"loss": 0.0012,
"step": 1630
},
{
"epoch": 0.14,
"grad_norm": 0.007360487245023251,
"learning_rate": 9.311618535930155e-06,
"loss": 0.001,
"step": 1640
},
{
"epoch": 0.14,
"grad_norm": 0.04865903779864311,
"learning_rate": 9.30742108797851e-06,
"loss": 0.0009,
"step": 1650
},
{
"epoch": 0.14,
"grad_norm": 0.04993268847465515,
"learning_rate": 9.303223640026866e-06,
"loss": 0.0011,
"step": 1660
},
{
"epoch": 0.14,
"grad_norm": 0.049482040107250214,
"learning_rate": 9.299026192075219e-06,
"loss": 0.0006,
"step": 1670
},
{
"epoch": 0.14,
"grad_norm": 0.08518658578395844,
"learning_rate": 9.294828744123573e-06,
"loss": 0.0018,
"step": 1680
},
{
"epoch": 0.14,
"grad_norm": 0.15694260597229004,
"learning_rate": 9.290631296171928e-06,
"loss": 0.0008,
"step": 1690
},
{
"epoch": 0.14,
"grad_norm": 0.2914963662624359,
"learning_rate": 9.286433848220282e-06,
"loss": 0.0046,
"step": 1700
},
{
"epoch": 0.14,
"grad_norm": 0.02077990211546421,
"learning_rate": 9.282236400268637e-06,
"loss": 0.0011,
"step": 1710
},
{
"epoch": 0.14,
"grad_norm": 0.13449931144714355,
"learning_rate": 9.278038952316992e-06,
"loss": 0.0037,
"step": 1720
},
{
"epoch": 0.15,
"grad_norm": 0.07630223780870438,
"learning_rate": 9.273841504365346e-06,
"loss": 0.0029,
"step": 1730
},
{
"epoch": 0.15,
"grad_norm": 0.19343329966068268,
"learning_rate": 9.269644056413701e-06,
"loss": 0.002,
"step": 1740
},
{
"epoch": 0.15,
"grad_norm": 0.03355003520846367,
"learning_rate": 9.265446608462056e-06,
"loss": 0.0006,
"step": 1750
},
{
"epoch": 0.15,
"grad_norm": 0.07213837653398514,
"learning_rate": 9.26124916051041e-06,
"loss": 0.0012,
"step": 1760
},
{
"epoch": 0.15,
"grad_norm": 0.11827400326728821,
"learning_rate": 9.257051712558765e-06,
"loss": 0.0014,
"step": 1770
},
{
"epoch": 0.15,
"grad_norm": 0.0025086326058954,
"learning_rate": 9.25285426460712e-06,
"loss": 0.0009,
"step": 1780
},
{
"epoch": 0.15,
"grad_norm": 0.05490359291434288,
"learning_rate": 9.248656816655474e-06,
"loss": 0.0012,
"step": 1790
},
{
"epoch": 0.15,
"grad_norm": 0.2864171862602234,
"learning_rate": 9.244459368703829e-06,
"loss": 0.0028,
"step": 1800
},
{
"epoch": 0.15,
"grad_norm": 0.18638712167739868,
"learning_rate": 9.240261920752183e-06,
"loss": 0.0027,
"step": 1810
},
{
"epoch": 0.15,
"grad_norm": 0.1295178234577179,
"learning_rate": 9.236064472800538e-06,
"loss": 0.0014,
"step": 1820
},
{
"epoch": 0.15,
"grad_norm": 0.4142198860645294,
"learning_rate": 9.231867024848892e-06,
"loss": 0.0017,
"step": 1830
},
{
"epoch": 0.15,
"grad_norm": 0.0262783020734787,
"learning_rate": 9.227669576897247e-06,
"loss": 0.0013,
"step": 1840
},
{
"epoch": 0.16,
"grad_norm": 0.030707404017448425,
"learning_rate": 9.223472128945602e-06,
"loss": 0.0009,
"step": 1850
},
{
"epoch": 0.16,
"grad_norm": 0.016312995925545692,
"learning_rate": 9.219274680993956e-06,
"loss": 0.0013,
"step": 1860
},
{
"epoch": 0.16,
"grad_norm": 0.036473799496889114,
"learning_rate": 9.215077233042311e-06,
"loss": 0.0031,
"step": 1870
},
{
"epoch": 0.16,
"grad_norm": 0.01363462395966053,
"learning_rate": 9.210879785090666e-06,
"loss": 0.0029,
"step": 1880
},
{
"epoch": 0.16,
"grad_norm": 0.06613222509622574,
"learning_rate": 9.20668233713902e-06,
"loss": 0.0037,
"step": 1890
},
{
"epoch": 0.16,
"grad_norm": 0.02161981351673603,
"learning_rate": 9.202484889187375e-06,
"loss": 0.0012,
"step": 1900
},
{
"epoch": 0.16,
"grad_norm": 0.013683819212019444,
"learning_rate": 9.19828744123573e-06,
"loss": 0.0019,
"step": 1910
},
{
"epoch": 0.16,
"grad_norm": 0.13119082152843475,
"learning_rate": 9.194089993284084e-06,
"loss": 0.001,
"step": 1920
},
{
"epoch": 0.16,
"grad_norm": 0.08101898431777954,
"learning_rate": 9.189892545332439e-06,
"loss": 0.0019,
"step": 1930
},
{
"epoch": 0.16,
"grad_norm": 0.0600835420191288,
"learning_rate": 9.185695097380793e-06,
"loss": 0.0014,
"step": 1940
},
{
"epoch": 0.16,
"grad_norm": 0.05432651564478874,
"learning_rate": 9.181497649429148e-06,
"loss": 0.0015,
"step": 1950
},
{
"epoch": 0.16,
"grad_norm": 0.06074979156255722,
"learning_rate": 9.177300201477503e-06,
"loss": 0.007,
"step": 1960
},
{
"epoch": 0.17,
"grad_norm": 0.44976115226745605,
"learning_rate": 9.173102753525857e-06,
"loss": 0.0025,
"step": 1970
},
{
"epoch": 0.17,
"grad_norm": 0.005641541443765163,
"learning_rate": 9.168905305574212e-06,
"loss": 0.0019,
"step": 1980
},
{
"epoch": 0.17,
"grad_norm": 0.006832782179117203,
"learning_rate": 9.164707857622566e-06,
"loss": 0.0013,
"step": 1990
},
{
"epoch": 0.17,
"grad_norm": 0.07477637380361557,
"learning_rate": 9.160510409670921e-06,
"loss": 0.0027,
"step": 2000
},
{
"epoch": 0.17,
"grad_norm": 0.07955386489629745,
"learning_rate": 9.156312961719276e-06,
"loss": 0.001,
"step": 2010
},
{
"epoch": 0.17,
"grad_norm": 0.009969334118068218,
"learning_rate": 9.15211551376763e-06,
"loss": 0.0015,
"step": 2020
},
{
"epoch": 0.17,
"grad_norm": 0.024671079590916634,
"learning_rate": 9.147918065815985e-06,
"loss": 0.0017,
"step": 2030
},
{
"epoch": 0.17,
"grad_norm": 0.23273734748363495,
"learning_rate": 9.14372061786434e-06,
"loss": 0.0012,
"step": 2040
},
{
"epoch": 0.17,
"grad_norm": 0.00974265392869711,
"learning_rate": 9.139523169912694e-06,
"loss": 0.0046,
"step": 2050
},
{
"epoch": 0.17,
"grad_norm": 0.5019146203994751,
"learning_rate": 9.135325721961049e-06,
"loss": 0.0024,
"step": 2060
},
{
"epoch": 0.17,
"grad_norm": 0.0385243222117424,
"learning_rate": 9.131128274009403e-06,
"loss": 0.0009,
"step": 2070
},
{
"epoch": 0.17,
"grad_norm": 0.005131669342517853,
"learning_rate": 9.126930826057758e-06,
"loss": 0.0014,
"step": 2080
},
{
"epoch": 0.18,
"grad_norm": 0.06231481209397316,
"learning_rate": 9.122733378106113e-06,
"loss": 0.0007,
"step": 2090
},
{
"epoch": 0.18,
"grad_norm": 0.057179391384124756,
"learning_rate": 9.118535930154467e-06,
"loss": 0.004,
"step": 2100
},
{
"epoch": 0.18,
"grad_norm": 0.06546686589717865,
"learning_rate": 9.114338482202822e-06,
"loss": 0.0018,
"step": 2110
},
{
"epoch": 0.18,
"grad_norm": 0.003714991034939885,
"learning_rate": 9.110141034251176e-06,
"loss": 0.0005,
"step": 2120
},
{
"epoch": 0.18,
"grad_norm": 0.06414463371038437,
"learning_rate": 9.105943586299531e-06,
"loss": 0.0005,
"step": 2130
},
{
"epoch": 0.18,
"grad_norm": 0.14546462893486023,
"learning_rate": 9.101746138347886e-06,
"loss": 0.0021,
"step": 2140
},
{
"epoch": 0.18,
"grad_norm": 0.05881655961275101,
"learning_rate": 9.09754869039624e-06,
"loss": 0.0015,
"step": 2150
},
{
"epoch": 0.18,
"grad_norm": 0.013162361457943916,
"learning_rate": 9.093351242444595e-06,
"loss": 0.0022,
"step": 2160
},
{
"epoch": 0.18,
"grad_norm": 0.017041141167283058,
"learning_rate": 9.08915379449295e-06,
"loss": 0.0018,
"step": 2170
},
{
"epoch": 0.18,
"grad_norm": 0.09564032405614853,
"learning_rate": 9.084956346541304e-06,
"loss": 0.0042,
"step": 2180
},
{
"epoch": 0.18,
"grad_norm": 0.014800913631916046,
"learning_rate": 9.080758898589657e-06,
"loss": 0.0009,
"step": 2190
},
{
"epoch": 0.18,
"grad_norm": 0.17374302446842194,
"learning_rate": 9.076561450638013e-06,
"loss": 0.0013,
"step": 2200
},
{
"epoch": 0.19,
"grad_norm": 0.035490747541189194,
"learning_rate": 9.072364002686368e-06,
"loss": 0.0038,
"step": 2210
},
{
"epoch": 0.19,
"grad_norm": 0.1741674244403839,
"learning_rate": 9.068166554734723e-06,
"loss": 0.0026,
"step": 2220
},
{
"epoch": 0.19,
"grad_norm": 0.04081317409873009,
"learning_rate": 9.063969106783077e-06,
"loss": 0.0014,
"step": 2230
},
{
"epoch": 0.19,
"grad_norm": 0.2914838194847107,
"learning_rate": 9.059771658831432e-06,
"loss": 0.0034,
"step": 2240
},
{
"epoch": 0.19,
"grad_norm": 0.02334975078701973,
"learning_rate": 9.055574210879786e-06,
"loss": 0.0041,
"step": 2250
},
{
"epoch": 0.19,
"grad_norm": 0.05057836323976517,
"learning_rate": 9.05137676292814e-06,
"loss": 0.0019,
"step": 2260
},
{
"epoch": 0.19,
"grad_norm": 0.16913112998008728,
"learning_rate": 9.047179314976496e-06,
"loss": 0.0023,
"step": 2270
},
{
"epoch": 0.19,
"grad_norm": 0.03450404852628708,
"learning_rate": 9.042981867024849e-06,
"loss": 0.0047,
"step": 2280
},
{
"epoch": 0.19,
"grad_norm": 0.007191210053861141,
"learning_rate": 9.038784419073205e-06,
"loss": 0.0058,
"step": 2290
},
{
"epoch": 0.19,
"grad_norm": 0.0037567466497421265,
"learning_rate": 9.034586971121558e-06,
"loss": 0.004,
"step": 2300
},
{
"epoch": 0.19,
"grad_norm": 0.6598537564277649,
"learning_rate": 9.030389523169914e-06,
"loss": 0.0025,
"step": 2310
},
{
"epoch": 0.19,
"grad_norm": 0.021132759749889374,
"learning_rate": 9.026192075218269e-06,
"loss": 0.002,
"step": 2320
},
{
"epoch": 0.2,
"grad_norm": 0.017818361520767212,
"learning_rate": 9.021994627266623e-06,
"loss": 0.0009,
"step": 2330
},
{
"epoch": 0.2,
"grad_norm": 0.016008947044610977,
"learning_rate": 9.017797179314978e-06,
"loss": 0.0005,
"step": 2340
},
{
"epoch": 0.2,
"grad_norm": 0.009955563582479954,
"learning_rate": 9.013599731363331e-06,
"loss": 0.0012,
"step": 2350
},
{
"epoch": 0.2,
"grad_norm": 0.11668864637613297,
"learning_rate": 9.009402283411687e-06,
"loss": 0.0032,
"step": 2360
},
{
"epoch": 0.2,
"grad_norm": 0.05549318715929985,
"learning_rate": 9.00520483546004e-06,
"loss": 0.0009,
"step": 2370
},
{
"epoch": 0.2,
"grad_norm": 0.026752743870019913,
"learning_rate": 9.001007387508396e-06,
"loss": 0.0018,
"step": 2380
},
{
"epoch": 0.2,
"grad_norm": 0.00796794705092907,
"learning_rate": 8.99680993955675e-06,
"loss": 0.0007,
"step": 2390
},
{
"epoch": 0.2,
"grad_norm": 0.02195807732641697,
"learning_rate": 8.992612491605106e-06,
"loss": 0.0031,
"step": 2400
},
{
"epoch": 0.2,
"grad_norm": 0.007499654311686754,
"learning_rate": 8.988415043653459e-06,
"loss": 0.001,
"step": 2410
},
{
"epoch": 0.2,
"grad_norm": 0.010224456898868084,
"learning_rate": 8.984217595701813e-06,
"loss": 0.0011,
"step": 2420
},
{
"epoch": 0.2,
"grad_norm": 0.031140731647610664,
"learning_rate": 8.98002014775017e-06,
"loss": 0.0007,
"step": 2430
},
{
"epoch": 0.2,
"grad_norm": 0.012480244040489197,
"learning_rate": 8.975822699798522e-06,
"loss": 0.0037,
"step": 2440
},
{
"epoch": 0.21,
"grad_norm": 0.13071498274803162,
"learning_rate": 8.971625251846879e-06,
"loss": 0.0008,
"step": 2450
},
{
"epoch": 0.21,
"grad_norm": 0.22285941243171692,
"learning_rate": 8.967427803895232e-06,
"loss": 0.0016,
"step": 2460
},
{
"epoch": 0.21,
"grad_norm": 0.2518530488014221,
"learning_rate": 8.963230355943588e-06,
"loss": 0.0026,
"step": 2470
},
{
"epoch": 0.21,
"grad_norm": 0.11510931700468063,
"learning_rate": 8.959032907991941e-06,
"loss": 0.0096,
"step": 2480
},
{
"epoch": 0.21,
"grad_norm": 0.010645145550370216,
"learning_rate": 8.954835460040297e-06,
"loss": 0.0013,
"step": 2490
},
{
"epoch": 0.21,
"grad_norm": 0.01338688749819994,
"learning_rate": 8.95063801208865e-06,
"loss": 0.0023,
"step": 2500
},
{
"epoch": 0.21,
"grad_norm": 0.01874951645731926,
"learning_rate": 8.946440564137005e-06,
"loss": 0.0031,
"step": 2510
},
{
"epoch": 0.21,
"grad_norm": 0.5476735234260559,
"learning_rate": 8.94224311618536e-06,
"loss": 0.0017,
"step": 2520
},
{
"epoch": 0.21,
"grad_norm": 0.01938123069703579,
"learning_rate": 8.938045668233714e-06,
"loss": 0.0003,
"step": 2530
},
{
"epoch": 0.21,
"grad_norm": 0.17968221008777618,
"learning_rate": 8.93384822028207e-06,
"loss": 0.0032,
"step": 2540
},
{
"epoch": 0.21,
"grad_norm": 0.07702717930078506,
"learning_rate": 8.929650772330423e-06,
"loss": 0.0012,
"step": 2550
},
{
"epoch": 0.21,
"grad_norm": 0.043315693736076355,
"learning_rate": 8.92545332437878e-06,
"loss": 0.002,
"step": 2560
},
{
"epoch": 0.22,
"grad_norm": 0.009717993438243866,
"learning_rate": 8.921255876427133e-06,
"loss": 0.0008,
"step": 2570
},
{
"epoch": 0.22,
"grad_norm": 0.0067781913094222546,
"learning_rate": 8.917058428475487e-06,
"loss": 0.0014,
"step": 2580
},
{
"epoch": 0.22,
"grad_norm": 0.011173662729561329,
"learning_rate": 8.912860980523842e-06,
"loss": 0.0018,
"step": 2590
},
{
"epoch": 0.22,
"grad_norm": 0.07794959843158722,
"learning_rate": 8.908663532572196e-06,
"loss": 0.0025,
"step": 2600
},
{
"epoch": 0.22,
"grad_norm": 0.005871940404176712,
"learning_rate": 8.904466084620551e-06,
"loss": 0.003,
"step": 2610
},
{
"epoch": 0.22,
"grad_norm": 0.06580822169780731,
"learning_rate": 8.900268636668906e-06,
"loss": 0.0054,
"step": 2620
},
{
"epoch": 0.22,
"grad_norm": 0.10812622308731079,
"learning_rate": 8.89607118871726e-06,
"loss": 0.0014,
"step": 2630
},
{
"epoch": 0.22,
"grad_norm": 0.08168531954288483,
"learning_rate": 8.891873740765615e-06,
"loss": 0.0023,
"step": 2640
},
{
"epoch": 0.22,
"grad_norm": 0.15115512907505035,
"learning_rate": 8.887676292813971e-06,
"loss": 0.0019,
"step": 2650
},
{
"epoch": 0.22,
"grad_norm": 0.04277985170483589,
"learning_rate": 8.883478844862324e-06,
"loss": 0.0011,
"step": 2660
},
{
"epoch": 0.22,
"grad_norm": 0.0317169725894928,
"learning_rate": 8.879281396910679e-06,
"loss": 0.0009,
"step": 2670
},
{
"epoch": 0.22,
"grad_norm": 0.01838303916156292,
"learning_rate": 8.875083948959033e-06,
"loss": 0.0011,
"step": 2680
},
{
"epoch": 0.23,
"grad_norm": 0.0044249724596738815,
"learning_rate": 8.870886501007388e-06,
"loss": 0.0009,
"step": 2690
},
{
"epoch": 0.23,
"grad_norm": 0.03765324875712395,
"learning_rate": 8.866689053055743e-06,
"loss": 0.0039,
"step": 2700
},
{
"epoch": 0.23,
"grad_norm": 0.29022645950317383,
"learning_rate": 8.862491605104097e-06,
"loss": 0.0055,
"step": 2710
},
{
"epoch": 0.23,
"grad_norm": 0.5368918776512146,
"learning_rate": 8.858294157152452e-06,
"loss": 0.0072,
"step": 2720
},
{
"epoch": 0.23,
"grad_norm": 0.0023665865883231163,
"learning_rate": 8.854096709200806e-06,
"loss": 0.0018,
"step": 2730
},
{
"epoch": 0.23,
"grad_norm": 0.07755130529403687,
"learning_rate": 8.849899261249161e-06,
"loss": 0.0015,
"step": 2740
},
{
"epoch": 0.23,
"grad_norm": 0.0149016622453928,
"learning_rate": 8.845701813297516e-06,
"loss": 0.0005,
"step": 2750
},
{
"epoch": 0.23,
"grad_norm": 0.0034178367350250483,
"learning_rate": 8.84150436534587e-06,
"loss": 0.0003,
"step": 2760
},
{
"epoch": 0.23,
"grad_norm": 0.11365914344787598,
"learning_rate": 8.837306917394225e-06,
"loss": 0.003,
"step": 2770
},
{
"epoch": 0.23,
"grad_norm": 0.11521656811237335,
"learning_rate": 8.83310946944258e-06,
"loss": 0.001,
"step": 2780
},
{
"epoch": 0.23,
"grad_norm": 0.10270581394433975,
"learning_rate": 8.828912021490934e-06,
"loss": 0.0015,
"step": 2790
},
{
"epoch": 0.24,
"grad_norm": 0.17976179718971252,
"learning_rate": 8.824714573539289e-06,
"loss": 0.0007,
"step": 2800
},
{
"epoch": 0.24,
"grad_norm": 0.07761932164430618,
"learning_rate": 8.820936870382809e-06,
"loss": 0.0029,
"step": 2810
},
{
"epoch": 0.24,
"grad_norm": 0.01468763779848814,
"learning_rate": 8.816739422431163e-06,
"loss": 0.0005,
"step": 2820
},
{
"epoch": 0.24,
"grad_norm": 0.009555538184940815,
"learning_rate": 8.812541974479516e-06,
"loss": 0.0018,
"step": 2830
},
{
"epoch": 0.24,
"grad_norm": 0.01784587651491165,
"learning_rate": 8.808344526527873e-06,
"loss": 0.0047,
"step": 2840
},
{
"epoch": 0.24,
"grad_norm": 0.35948890447616577,
"learning_rate": 8.804147078576226e-06,
"loss": 0.0044,
"step": 2850
},
{
"epoch": 0.24,
"grad_norm": 0.052703533321619034,
"learning_rate": 8.799949630624582e-06,
"loss": 0.0006,
"step": 2860
},
{
"epoch": 0.24,
"grad_norm": 0.171566441655159,
"learning_rate": 8.795752182672935e-06,
"loss": 0.0023,
"step": 2870
},
{
"epoch": 0.24,
"grad_norm": 0.3138861060142517,
"learning_rate": 8.791554734721291e-06,
"loss": 0.0007,
"step": 2880
},
{
"epoch": 0.24,
"grad_norm": 0.15738941729068756,
"learning_rate": 8.787357286769644e-06,
"loss": 0.0022,
"step": 2890
},
{
"epoch": 0.24,
"grad_norm": 0.03881895914673805,
"learning_rate": 8.783159838818e-06,
"loss": 0.004,
"step": 2900
},
{
"epoch": 0.24,
"grad_norm": 0.027307789772748947,
"learning_rate": 8.778962390866353e-06,
"loss": 0.0009,
"step": 2910
},
{
"epoch": 0.25,
"grad_norm": 0.009243845008313656,
"learning_rate": 8.774764942914708e-06,
"loss": 0.0021,
"step": 2920
},
{
"epoch": 0.25,
"grad_norm": 0.010441136546432972,
"learning_rate": 8.770567494963064e-06,
"loss": 0.0003,
"step": 2930
},
{
"epoch": 0.25,
"grad_norm": 0.053923387080430984,
"learning_rate": 8.766370047011417e-06,
"loss": 0.0033,
"step": 2940
},
{
"epoch": 0.25,
"grad_norm": 0.0529450997710228,
"learning_rate": 8.762172599059773e-06,
"loss": 0.0017,
"step": 2950
},
{
"epoch": 0.25,
"grad_norm": 0.017623068764805794,
"learning_rate": 8.757975151108126e-06,
"loss": 0.007,
"step": 2960
},
{
"epoch": 0.25,
"grad_norm": 0.06934964656829834,
"learning_rate": 8.753777703156483e-06,
"loss": 0.0017,
"step": 2970
},
{
"epoch": 0.25,
"grad_norm": 0.4888018071651459,
"learning_rate": 8.749580255204836e-06,
"loss": 0.0018,
"step": 2980
},
{
"epoch": 0.25,
"grad_norm": 0.3019413948059082,
"learning_rate": 8.74538280725319e-06,
"loss": 0.0078,
"step": 2990
},
{
"epoch": 0.25,
"grad_norm": 0.20448799431324005,
"learning_rate": 8.741185359301545e-06,
"loss": 0.0026,
"step": 3000
},
{
"epoch": 0.25,
"grad_norm": 0.15655213594436646,
"learning_rate": 8.7369879113499e-06,
"loss": 0.0007,
"step": 3010
},
{
"epoch": 0.25,
"grad_norm": 0.00702297780662775,
"learning_rate": 8.732790463398254e-06,
"loss": 0.0022,
"step": 3020
},
{
"epoch": 0.25,
"grad_norm": 0.03202832117676735,
"learning_rate": 8.728593015446609e-06,
"loss": 0.0048,
"step": 3030
},
{
"epoch": 0.26,
"grad_norm": 0.3602777123451233,
"learning_rate": 8.724395567494965e-06,
"loss": 0.0027,
"step": 3040
},
{
"epoch": 0.26,
"grad_norm": 0.035856008529663086,
"learning_rate": 8.720198119543318e-06,
"loss": 0.0011,
"step": 3050
},
{
"epoch": 0.26,
"grad_norm": 0.02676764689385891,
"learning_rate": 8.716000671591674e-06,
"loss": 0.0006,
"step": 3060
},
{
"epoch": 0.26,
"grad_norm": 0.06010160595178604,
"learning_rate": 8.711803223640027e-06,
"loss": 0.0023,
"step": 3070
},
{
"epoch": 0.26,
"grad_norm": 0.33468106389045715,
"learning_rate": 8.707605775688382e-06,
"loss": 0.0013,
"step": 3080
},
{
"epoch": 0.26,
"grad_norm": 0.07244027405977249,
"learning_rate": 8.703408327736736e-06,
"loss": 0.001,
"step": 3090
},
{
"epoch": 0.26,
"grad_norm": 0.1765681654214859,
"learning_rate": 8.699210879785091e-06,
"loss": 0.0021,
"step": 3100
},
{
"epoch": 0.26,
"grad_norm": 0.0064493464305996895,
"learning_rate": 8.695013431833446e-06,
"loss": 0.0021,
"step": 3110
},
{
"epoch": 0.26,
"grad_norm": 0.12435489892959595,
"learning_rate": 8.6908159838818e-06,
"loss": 0.0017,
"step": 3120
},
{
"epoch": 0.26,
"grad_norm": 0.05187215656042099,
"learning_rate": 8.686618535930155e-06,
"loss": 0.001,
"step": 3130
},
{
"epoch": 0.26,
"grad_norm": 0.01066290121525526,
"learning_rate": 8.68242108797851e-06,
"loss": 0.0011,
"step": 3140
},
{
"epoch": 0.26,
"grad_norm": 0.04758090898394585,
"learning_rate": 8.678223640026866e-06,
"loss": 0.0005,
"step": 3150
},
{
"epoch": 0.27,
"grad_norm": 0.02921927347779274,
"learning_rate": 8.674026192075219e-06,
"loss": 0.0021,
"step": 3160
},
{
"epoch": 0.27,
"grad_norm": 0.009303642436861992,
"learning_rate": 8.669828744123573e-06,
"loss": 0.0007,
"step": 3170
},
{
"epoch": 0.27,
"grad_norm": 0.5222365856170654,
"learning_rate": 8.665631296171928e-06,
"loss": 0.007,
"step": 3180
},
{
"epoch": 0.27,
"grad_norm": 0.005165881011635065,
"learning_rate": 8.661433848220283e-06,
"loss": 0.0018,
"step": 3190
},
{
"epoch": 0.27,
"grad_norm": 0.8817521333694458,
"learning_rate": 8.657236400268637e-06,
"loss": 0.002,
"step": 3200
},
{
"epoch": 0.27,
"grad_norm": 0.07392257452011108,
"learning_rate": 8.653038952316992e-06,
"loss": 0.0015,
"step": 3210
},
{
"epoch": 0.27,
"grad_norm": 0.039199940860271454,
"learning_rate": 8.648841504365346e-06,
"loss": 0.0007,
"step": 3220
},
{
"epoch": 0.27,
"grad_norm": 0.2725190222263336,
"learning_rate": 8.644644056413701e-06,
"loss": 0.0022,
"step": 3230
},
{
"epoch": 0.27,
"grad_norm": 0.012581940740346909,
"learning_rate": 8.640446608462056e-06,
"loss": 0.0073,
"step": 3240
},
{
"epoch": 0.27,
"grad_norm": 0.008940734900534153,
"learning_rate": 8.63624916051041e-06,
"loss": 0.0009,
"step": 3250
},
{
"epoch": 0.27,
"grad_norm": 0.1721634566783905,
"learning_rate": 8.632051712558765e-06,
"loss": 0.0017,
"step": 3260
},
{
"epoch": 0.27,
"grad_norm": 0.31307950615882874,
"learning_rate": 8.62785426460712e-06,
"loss": 0.0033,
"step": 3270
},
{
"epoch": 0.28,
"grad_norm": 0.18467161059379578,
"learning_rate": 8.623656816655474e-06,
"loss": 0.0029,
"step": 3280
},
{
"epoch": 0.28,
"grad_norm": 0.04656468331813812,
"learning_rate": 8.619879113498994e-06,
"loss": 0.0018,
"step": 3290
},
{
"epoch": 0.28,
"grad_norm": 0.0923081561923027,
"learning_rate": 8.615681665547347e-06,
"loss": 0.0011,
"step": 3300
},
{
"epoch": 0.28,
"grad_norm": 0.15977904200553894,
"learning_rate": 8.611484217595703e-06,
"loss": 0.0027,
"step": 3310
},
{
"epoch": 0.28,
"grad_norm": 0.10176800936460495,
"learning_rate": 8.607286769644058e-06,
"loss": 0.0024,
"step": 3320
},
{
"epoch": 0.28,
"grad_norm": 0.057700444012880325,
"learning_rate": 8.603089321692411e-06,
"loss": 0.0031,
"step": 3330
},
{
"epoch": 0.28,
"grad_norm": 0.03322164714336395,
"learning_rate": 8.598891873740767e-06,
"loss": 0.0015,
"step": 3340
},
{
"epoch": 0.28,
"grad_norm": 0.018284544348716736,
"learning_rate": 8.59469442578912e-06,
"loss": 0.0022,
"step": 3350
},
{
"epoch": 0.28,
"grad_norm": 0.02011759579181671,
"learning_rate": 8.590496977837476e-06,
"loss": 0.0008,
"step": 3360
},
{
"epoch": 0.28,
"grad_norm": 0.22747820615768433,
"learning_rate": 8.58629952988583e-06,
"loss": 0.0009,
"step": 3370
},
{
"epoch": 0.28,
"grad_norm": 0.08898269385099411,
"learning_rate": 8.582102081934186e-06,
"loss": 0.0018,
"step": 3380
},
{
"epoch": 0.28,
"grad_norm": 0.3748880922794342,
"learning_rate": 8.577904633982539e-06,
"loss": 0.0045,
"step": 3390
},
{
"epoch": 0.29,
"grad_norm": 0.007508324924856424,
"learning_rate": 8.573707186030893e-06,
"loss": 0.004,
"step": 3400
},
{
"epoch": 0.29,
"grad_norm": 0.08410963416099548,
"learning_rate": 8.569509738079248e-06,
"loss": 0.0049,
"step": 3410
},
{
"epoch": 0.29,
"grad_norm": 0.03300091251730919,
"learning_rate": 8.565312290127602e-06,
"loss": 0.0013,
"step": 3420
},
{
"epoch": 0.29,
"grad_norm": 0.030519753694534302,
"learning_rate": 8.561114842175959e-06,
"loss": 0.0035,
"step": 3430
},
{
"epoch": 0.29,
"grad_norm": 0.013716090470552444,
"learning_rate": 8.556917394224312e-06,
"loss": 0.0012,
"step": 3440
},
{
"epoch": 0.29,
"grad_norm": 0.07122453302145004,
"learning_rate": 8.552719946272668e-06,
"loss": 0.0023,
"step": 3450
},
{
"epoch": 0.29,
"grad_norm": 0.006021963898092508,
"learning_rate": 8.548522498321021e-06,
"loss": 0.0012,
"step": 3460
},
{
"epoch": 0.29,
"grad_norm": 0.007859396748244762,
"learning_rate": 8.544325050369377e-06,
"loss": 0.0005,
"step": 3470
},
{
"epoch": 0.29,
"grad_norm": 0.09434602409601212,
"learning_rate": 8.54012760241773e-06,
"loss": 0.0011,
"step": 3480
},
{
"epoch": 0.29,
"grad_norm": 0.01787838153541088,
"learning_rate": 8.535930154466085e-06,
"loss": 0.0011,
"step": 3490
},
{
"epoch": 0.29,
"grad_norm": 0.018500063568353653,
"learning_rate": 8.53173270651444e-06,
"loss": 0.0017,
"step": 3500
},
{
"epoch": 0.29,
"grad_norm": 0.0013119007926434278,
"learning_rate": 8.527535258562794e-06,
"loss": 0.0013,
"step": 3510
},
{
"epoch": 0.3,
"grad_norm": 0.02389811910688877,
"learning_rate": 8.523337810611149e-06,
"loss": 0.0011,
"step": 3520
},
{
"epoch": 0.3,
"grad_norm": 0.0021516175474971533,
"learning_rate": 8.519140362659503e-06,
"loss": 0.0056,
"step": 3530
},
{
"epoch": 0.3,
"grad_norm": 0.016399670392274857,
"learning_rate": 8.51494291470786e-06,
"loss": 0.001,
"step": 3540
},
{
"epoch": 0.3,
"grad_norm": 0.07569669187068939,
"learning_rate": 8.510745466756212e-06,
"loss": 0.0048,
"step": 3550
},
{
"epoch": 0.3,
"grad_norm": 0.002499897498637438,
"learning_rate": 8.506548018804569e-06,
"loss": 0.0015,
"step": 3560
},
{
"epoch": 0.3,
"grad_norm": 0.0031851092353463173,
"learning_rate": 8.502350570852922e-06,
"loss": 0.0002,
"step": 3570
},
{
"epoch": 0.3,
"grad_norm": 0.6716926693916321,
"learning_rate": 8.498153122901276e-06,
"loss": 0.0052,
"step": 3580
},
{
"epoch": 0.3,
"grad_norm": 0.026271577924489975,
"learning_rate": 8.493955674949631e-06,
"loss": 0.0017,
"step": 3590
},
{
"epoch": 0.3,
"grad_norm": 0.47044169902801514,
"learning_rate": 8.489758226997986e-06,
"loss": 0.0037,
"step": 3600
},
{
"epoch": 0.3,
"grad_norm": 0.175507590174675,
"learning_rate": 8.48556077904634e-06,
"loss": 0.0011,
"step": 3610
},
{
"epoch": 0.3,
"grad_norm": 0.04265287518501282,
"learning_rate": 8.481363331094695e-06,
"loss": 0.0057,
"step": 3620
},
{
"epoch": 0.3,
"grad_norm": 0.06122719869017601,
"learning_rate": 8.47716588314305e-06,
"loss": 0.0012,
"step": 3630
},
{
"epoch": 0.31,
"grad_norm": 0.09644156694412231,
"learning_rate": 8.472968435191404e-06,
"loss": 0.0008,
"step": 3640
},
{
"epoch": 0.31,
"grad_norm": 0.11276187747716904,
"learning_rate": 8.468770987239759e-06,
"loss": 0.0017,
"step": 3650
},
{
"epoch": 0.31,
"grad_norm": 0.02469436265528202,
"learning_rate": 8.464573539288113e-06,
"loss": 0.0046,
"step": 3660
},
{
"epoch": 0.31,
"grad_norm": 0.7709282040596008,
"learning_rate": 8.460376091336468e-06,
"loss": 0.008,
"step": 3670
},
{
"epoch": 0.31,
"grad_norm": 0.4117774963378906,
"learning_rate": 8.456178643384822e-06,
"loss": 0.0024,
"step": 3680
},
{
"epoch": 0.31,
"grad_norm": 0.07191318273544312,
"learning_rate": 8.451981195433177e-06,
"loss": 0.0011,
"step": 3690
},
{
"epoch": 0.31,
"grad_norm": 0.014883830212056637,
"learning_rate": 8.447783747481532e-06,
"loss": 0.001,
"step": 3700
},
{
"epoch": 0.31,
"grad_norm": 0.0047638434916734695,
"learning_rate": 8.443586299529886e-06,
"loss": 0.0012,
"step": 3710
},
{
"epoch": 0.31,
"grad_norm": 0.19078807532787323,
"learning_rate": 8.439388851578241e-06,
"loss": 0.0011,
"step": 3720
},
{
"epoch": 0.31,
"grad_norm": 0.07204841822385788,
"learning_rate": 8.435191403626596e-06,
"loss": 0.0018,
"step": 3730
},
{
"epoch": 0.31,
"grad_norm": 0.6154640913009644,
"learning_rate": 8.43099395567495e-06,
"loss": 0.0038,
"step": 3740
},
{
"epoch": 0.31,
"grad_norm": 0.02338569425046444,
"learning_rate": 8.426796507723305e-06,
"loss": 0.0016,
"step": 3750
},
{
"epoch": 0.32,
"grad_norm": 0.13649843633174896,
"learning_rate": 8.423018804566823e-06,
"loss": 0.0047,
"step": 3760
},
{
"epoch": 0.32,
"grad_norm": 0.0963369682431221,
"learning_rate": 8.41882135661518e-06,
"loss": 0.0006,
"step": 3770
},
{
"epoch": 0.32,
"grad_norm": 0.31827324628829956,
"learning_rate": 8.414623908663532e-06,
"loss": 0.0058,
"step": 3780
},
{
"epoch": 0.32,
"grad_norm": 0.006173072848469019,
"learning_rate": 8.410426460711889e-06,
"loss": 0.0014,
"step": 3790
},
{
"epoch": 0.32,
"grad_norm": 0.13730692863464355,
"learning_rate": 8.406229012760242e-06,
"loss": 0.0009,
"step": 3800
},
{
"epoch": 0.32,
"grad_norm": 0.004133482463657856,
"learning_rate": 8.402031564808596e-06,
"loss": 0.0033,
"step": 3810
},
{
"epoch": 0.32,
"grad_norm": 0.13467635214328766,
"learning_rate": 8.397834116856953e-06,
"loss": 0.0009,
"step": 3820
},
{
"epoch": 0.32,
"grad_norm": 0.07814080268144608,
"learning_rate": 8.393636668905305e-06,
"loss": 0.0016,
"step": 3830
},
{
"epoch": 0.32,
"grad_norm": 0.011251374147832394,
"learning_rate": 8.389439220953662e-06,
"loss": 0.0009,
"step": 3840
},
{
"epoch": 0.32,
"grad_norm": 0.001605527475476265,
"learning_rate": 8.385241773002015e-06,
"loss": 0.0013,
"step": 3850
},
{
"epoch": 0.32,
"grad_norm": 0.008475171402096748,
"learning_rate": 8.381044325050371e-06,
"loss": 0.0008,
"step": 3860
},
{
"epoch": 0.32,
"grad_norm": 0.018021291121840477,
"learning_rate": 8.376846877098724e-06,
"loss": 0.0004,
"step": 3870
},
{
"epoch": 0.33,
"grad_norm": 0.006848904769867659,
"learning_rate": 8.37264942914708e-06,
"loss": 0.0014,
"step": 3880
},
{
"epoch": 0.33,
"grad_norm": 0.025030123069882393,
"learning_rate": 8.368451981195433e-06,
"loss": 0.0014,
"step": 3890
},
{
"epoch": 0.33,
"grad_norm": 0.4155631959438324,
"learning_rate": 8.364254533243788e-06,
"loss": 0.0015,
"step": 3900
},
{
"epoch": 0.33,
"grad_norm": 0.005679064895957708,
"learning_rate": 8.360057085292142e-06,
"loss": 0.003,
"step": 3910
},
{
"epoch": 0.33,
"grad_norm": 0.009578896686434746,
"learning_rate": 8.355859637340497e-06,
"loss": 0.0048,
"step": 3920
},
{
"epoch": 0.33,
"grad_norm": 0.011455328203737736,
"learning_rate": 8.351662189388853e-06,
"loss": 0.0007,
"step": 3930
},
{
"epoch": 0.33,
"grad_norm": 0.02022560127079487,
"learning_rate": 8.347464741437206e-06,
"loss": 0.002,
"step": 3940
},
{
"epoch": 0.33,
"grad_norm": 0.013254129327833652,
"learning_rate": 8.343267293485563e-06,
"loss": 0.0008,
"step": 3950
},
{
"epoch": 0.33,
"grad_norm": 0.03954704850912094,
"learning_rate": 8.339069845533915e-06,
"loss": 0.0021,
"step": 3960
},
{
"epoch": 0.33,
"grad_norm": 0.03727477416396141,
"learning_rate": 8.334872397582272e-06,
"loss": 0.0021,
"step": 3970
},
{
"epoch": 0.33,
"grad_norm": 0.005523467902094126,
"learning_rate": 8.330674949630625e-06,
"loss": 0.0007,
"step": 3980
},
{
"epoch": 0.33,
"grad_norm": 0.022439954802393913,
"learning_rate": 8.32647750167898e-06,
"loss": 0.0021,
"step": 3990
},
{
"epoch": 0.34,
"grad_norm": 0.012325704097747803,
"learning_rate": 8.322280053727334e-06,
"loss": 0.0014,
"step": 4000
},
{
"epoch": 0.34,
"grad_norm": 0.12552155554294586,
"learning_rate": 8.318082605775689e-06,
"loss": 0.0011,
"step": 4010
},
{
"epoch": 0.34,
"grad_norm": 0.017489569261670113,
"learning_rate": 8.313885157824043e-06,
"loss": 0.0011,
"step": 4020
},
{
"epoch": 0.34,
"grad_norm": 0.052284810692071915,
"learning_rate": 8.309687709872398e-06,
"loss": 0.001,
"step": 4030
},
{
"epoch": 0.34,
"grad_norm": 0.05398709326982498,
"learning_rate": 8.305490261920754e-06,
"loss": 0.0014,
"step": 4040
},
{
"epoch": 0.34,
"grad_norm": 0.25828316807746887,
"learning_rate": 8.301292813969107e-06,
"loss": 0.0026,
"step": 4050
},
{
"epoch": 0.34,
"grad_norm": 0.01474198792129755,
"learning_rate": 8.297095366017462e-06,
"loss": 0.0017,
"step": 4060
},
{
"epoch": 0.34,
"grad_norm": 0.0038133654743433,
"learning_rate": 8.292897918065816e-06,
"loss": 0.0009,
"step": 4070
},
{
"epoch": 0.34,
"grad_norm": 0.021653324365615845,
"learning_rate": 8.288700470114171e-06,
"loss": 0.0008,
"step": 4080
},
{
"epoch": 0.34,
"grad_norm": 0.017369549721479416,
"learning_rate": 8.284503022162526e-06,
"loss": 0.0011,
"step": 4090
},
{
"epoch": 0.34,
"grad_norm": 0.18425977230072021,
"learning_rate": 8.28030557421088e-06,
"loss": 0.0015,
"step": 4100
},
{
"epoch": 0.35,
"grad_norm": 0.01438868883997202,
"learning_rate": 8.276108126259235e-06,
"loss": 0.0006,
"step": 4110
},
{
"epoch": 0.35,
"grad_norm": 0.0022420857567340136,
"learning_rate": 8.27191067830759e-06,
"loss": 0.0002,
"step": 4120
},
{
"epoch": 0.35,
"grad_norm": 0.12319325655698776,
"learning_rate": 8.267713230355944e-06,
"loss": 0.0051,
"step": 4130
},
{
"epoch": 0.35,
"grad_norm": 0.01060109119862318,
"learning_rate": 8.263515782404299e-06,
"loss": 0.0003,
"step": 4140
},
{
"epoch": 0.35,
"grad_norm": 0.001791100949048996,
"learning_rate": 8.259318334452653e-06,
"loss": 0.0016,
"step": 4150
},
{
"epoch": 0.35,
"grad_norm": 0.009030510671436787,
"learning_rate": 8.255120886501008e-06,
"loss": 0.0004,
"step": 4160
},
{
"epoch": 0.35,
"grad_norm": 0.1018250361084938,
"learning_rate": 8.250923438549362e-06,
"loss": 0.0018,
"step": 4170
},
{
"epoch": 0.35,
"grad_norm": 0.11478858441114426,
"learning_rate": 8.246725990597717e-06,
"loss": 0.0004,
"step": 4180
},
{
"epoch": 0.35,
"grad_norm": 0.08903089910745621,
"learning_rate": 8.242528542646072e-06,
"loss": 0.0018,
"step": 4190
},
{
"epoch": 0.35,
"grad_norm": 0.109068363904953,
"learning_rate": 8.238331094694426e-06,
"loss": 0.0007,
"step": 4200
},
{
"epoch": 0.35,
"grad_norm": 0.005220823921263218,
"learning_rate": 8.234133646742781e-06,
"loss": 0.0044,
"step": 4210
},
{
"epoch": 0.35,
"grad_norm": 0.09990601986646652,
"learning_rate": 8.229936198791136e-06,
"loss": 0.0012,
"step": 4220
},
{
"epoch": 0.36,
"grad_norm": 0.07386545091867447,
"learning_rate": 8.22573875083949e-06,
"loss": 0.0008,
"step": 4230
},
{
"epoch": 0.36,
"grad_norm": 0.1059875562787056,
"learning_rate": 8.221541302887845e-06,
"loss": 0.0007,
"step": 4240
},
{
"epoch": 0.36,
"grad_norm": 0.07418499141931534,
"learning_rate": 8.2173438549362e-06,
"loss": 0.0014,
"step": 4250
},
{
"epoch": 0.36,
"grad_norm": 0.01813233084976673,
"learning_rate": 8.213146406984554e-06,
"loss": 0.0007,
"step": 4260
},
{
"epoch": 0.36,
"grad_norm": 0.01308862492442131,
"learning_rate": 8.208948959032909e-06,
"loss": 0.0002,
"step": 4270
},
{
"epoch": 0.36,
"grad_norm": 0.6047486066818237,
"learning_rate": 8.204751511081263e-06,
"loss": 0.0054,
"step": 4280
},
{
"epoch": 0.36,
"grad_norm": 0.030174901708960533,
"learning_rate": 8.200554063129618e-06,
"loss": 0.0007,
"step": 4290
},
{
"epoch": 0.36,
"grad_norm": 0.07209368050098419,
"learning_rate": 8.196356615177972e-06,
"loss": 0.0011,
"step": 4300
},
{
"epoch": 0.36,
"grad_norm": 0.6904421448707581,
"learning_rate": 8.192159167226327e-06,
"loss": 0.0018,
"step": 4310
},
{
"epoch": 0.36,
"grad_norm": 0.11670655757188797,
"learning_rate": 8.187961719274682e-06,
"loss": 0.0028,
"step": 4320
},
{
"epoch": 0.36,
"grad_norm": 0.016770286485552788,
"learning_rate": 8.183764271323036e-06,
"loss": 0.0005,
"step": 4330
},
{
"epoch": 0.36,
"grad_norm": 0.021145779639482498,
"learning_rate": 8.179566823371391e-06,
"loss": 0.0018,
"step": 4340
},
{
"epoch": 0.37,
"grad_norm": 0.02627403475344181,
"learning_rate": 8.175369375419746e-06,
"loss": 0.0014,
"step": 4350
},
{
"epoch": 0.37,
"grad_norm": 0.01143514271825552,
"learning_rate": 8.1711719274681e-06,
"loss": 0.0007,
"step": 4360
},
{
"epoch": 0.37,
"grad_norm": 0.05208797752857208,
"learning_rate": 8.166974479516455e-06,
"loss": 0.0023,
"step": 4370
},
{
"epoch": 0.37,
"grad_norm": 0.08176781982183456,
"learning_rate": 8.16277703156481e-06,
"loss": 0.0028,
"step": 4380
},
{
"epoch": 0.37,
"grad_norm": 0.0852191299200058,
"learning_rate": 8.158579583613164e-06,
"loss": 0.0023,
"step": 4390
},
{
"epoch": 0.37,
"grad_norm": 0.007911001332104206,
"learning_rate": 8.154382135661519e-06,
"loss": 0.0032,
"step": 4400
},
{
"epoch": 0.37,
"grad_norm": 0.012264366261661053,
"learning_rate": 8.150184687709873e-06,
"loss": 0.0011,
"step": 4410
},
{
"epoch": 0.37,
"grad_norm": 0.012588711455464363,
"learning_rate": 8.145987239758228e-06,
"loss": 0.0003,
"step": 4420
},
{
"epoch": 0.37,
"grad_norm": 0.04339655116200447,
"learning_rate": 8.141789791806583e-06,
"loss": 0.0016,
"step": 4430
},
{
"epoch": 0.37,
"grad_norm": 0.009127240628004074,
"learning_rate": 8.137592343854937e-06,
"loss": 0.0029,
"step": 4440
},
{
"epoch": 0.37,
"grad_norm": 0.05993298068642616,
"learning_rate": 8.133394895903292e-06,
"loss": 0.0008,
"step": 4450
},
{
"epoch": 0.37,
"grad_norm": 0.0026275282725691795,
"learning_rate": 8.129197447951646e-06,
"loss": 0.0009,
"step": 4460
},
{
"epoch": 0.38,
"grad_norm": 0.04943431168794632,
"learning_rate": 8.125000000000001e-06,
"loss": 0.0014,
"step": 4470
},
{
"epoch": 0.38,
"grad_norm": 0.018763918429613113,
"learning_rate": 8.120802552048356e-06,
"loss": 0.0007,
"step": 4480
},
{
"epoch": 0.38,
"grad_norm": 0.5426563620567322,
"learning_rate": 8.11660510409671e-06,
"loss": 0.0021,
"step": 4490
},
{
"epoch": 0.38,
"grad_norm": 0.4148786962032318,
"learning_rate": 8.112407656145065e-06,
"loss": 0.0014,
"step": 4500
},
{
"epoch": 0.38,
"grad_norm": 0.479037344455719,
"learning_rate": 8.10821020819342e-06,
"loss": 0.0034,
"step": 4510
},
{
"epoch": 0.38,
"grad_norm": 0.005279048811644316,
"learning_rate": 8.104012760241774e-06,
"loss": 0.0015,
"step": 4520
},
{
"epoch": 0.38,
"grad_norm": 0.003773655276745558,
"learning_rate": 8.099815312290129e-06,
"loss": 0.0011,
"step": 4530
},
{
"epoch": 0.38,
"grad_norm": 0.0063260081224143505,
"learning_rate": 8.095617864338483e-06,
"loss": 0.0011,
"step": 4540
},
{
"epoch": 0.38,
"grad_norm": 0.030879395082592964,
"learning_rate": 8.091420416386836e-06,
"loss": 0.0092,
"step": 4550
},
{
"epoch": 0.38,
"grad_norm": 0.020290937274694443,
"learning_rate": 8.087222968435193e-06,
"loss": 0.0005,
"step": 4560
},
{
"epoch": 0.38,
"grad_norm": 0.007667260244488716,
"learning_rate": 8.083025520483545e-06,
"loss": 0.0006,
"step": 4570
},
{
"epoch": 0.38,
"grad_norm": 0.004535841289907694,
"learning_rate": 8.078828072531902e-06,
"loss": 0.0005,
"step": 4580
},
{
"epoch": 0.39,
"grad_norm": 0.011321828700602055,
"learning_rate": 8.074630624580256e-06,
"loss": 0.002,
"step": 4590
},
{
"epoch": 0.39,
"grad_norm": 0.3192904591560364,
"learning_rate": 8.070433176628611e-06,
"loss": 0.0031,
"step": 4600
},
{
"epoch": 0.39,
"grad_norm": 0.017921652644872665,
"learning_rate": 8.066235728676966e-06,
"loss": 0.0008,
"step": 4610
},
{
"epoch": 0.39,
"grad_norm": 0.0025262893177568913,
"learning_rate": 8.06203828072532e-06,
"loss": 0.0006,
"step": 4620
},
{
"epoch": 0.39,
"grad_norm": 0.013485982082784176,
"learning_rate": 8.057840832773675e-06,
"loss": 0.0016,
"step": 4630
},
{
"epoch": 0.39,
"grad_norm": 0.004218879155814648,
"learning_rate": 8.053643384822028e-06,
"loss": 0.0004,
"step": 4640
},
{
"epoch": 0.39,
"grad_norm": 0.1440621167421341,
"learning_rate": 8.049445936870384e-06,
"loss": 0.0023,
"step": 4650
},
{
"epoch": 0.39,
"grad_norm": 0.1656440794467926,
"learning_rate": 8.045248488918737e-06,
"loss": 0.0004,
"step": 4660
},
{
"epoch": 0.39,
"grad_norm": 0.05483277887105942,
"learning_rate": 8.041051040967093e-06,
"loss": 0.0022,
"step": 4670
},
{
"epoch": 0.39,
"grad_norm": 0.011686104349792004,
"learning_rate": 8.036853593015446e-06,
"loss": 0.0002,
"step": 4680
},
{
"epoch": 0.39,
"grad_norm": 0.23753832280635834,
"learning_rate": 8.032656145063803e-06,
"loss": 0.0022,
"step": 4690
},
{
"epoch": 0.39,
"grad_norm": 0.14190416038036346,
"learning_rate": 8.028458697112157e-06,
"loss": 0.001,
"step": 4700
},
{
"epoch": 0.4,
"grad_norm": 0.04893433302640915,
"learning_rate": 8.024261249160512e-06,
"loss": 0.0019,
"step": 4710
},
{
"epoch": 0.4,
"grad_norm": 0.06139785423874855,
"learning_rate": 8.020063801208866e-06,
"loss": 0.0023,
"step": 4720
},
{
"epoch": 0.4,
"grad_norm": 0.31157851219177246,
"learning_rate": 8.01586635325722e-06,
"loss": 0.0031,
"step": 4730
},
{
"epoch": 0.4,
"grad_norm": 0.01513377670198679,
"learning_rate": 8.011668905305576e-06,
"loss": 0.0032,
"step": 4740
},
{
"epoch": 0.4,
"grad_norm": 0.014633961953222752,
"learning_rate": 8.007471457353929e-06,
"loss": 0.0073,
"step": 4750
},
{
"epoch": 0.4,
"grad_norm": 0.03786788135766983,
"learning_rate": 8.003274009402285e-06,
"loss": 0.0005,
"step": 4760
},
{
"epoch": 0.4,
"grad_norm": 0.013805943541228771,
"learning_rate": 7.999076561450638e-06,
"loss": 0.0006,
"step": 4770
},
{
"epoch": 0.4,
"grad_norm": 0.423827201128006,
"learning_rate": 7.994879113498994e-06,
"loss": 0.0024,
"step": 4780
},
{
"epoch": 0.4,
"grad_norm": 0.15604566037654877,
"learning_rate": 7.990681665547347e-06,
"loss": 0.004,
"step": 4790
},
{
"epoch": 0.4,
"grad_norm": 0.1691051870584488,
"learning_rate": 7.986484217595702e-06,
"loss": 0.0007,
"step": 4800
},
{
"epoch": 0.4,
"grad_norm": 0.10865192860364914,
"learning_rate": 7.982286769644058e-06,
"loss": 0.0008,
"step": 4810
},
{
"epoch": 0.4,
"grad_norm": 0.002326708287000656,
"learning_rate": 7.978089321692411e-06,
"loss": 0.0047,
"step": 4820
},
{
"epoch": 0.41,
"grad_norm": 0.0631171315908432,
"learning_rate": 7.973891873740767e-06,
"loss": 0.001,
"step": 4830
},
{
"epoch": 0.41,
"grad_norm": 0.003491542534902692,
"learning_rate": 7.96969442578912e-06,
"loss": 0.0019,
"step": 4840
},
{
"epoch": 0.41,
"grad_norm": 0.023969637230038643,
"learning_rate": 7.965496977837476e-06,
"loss": 0.0013,
"step": 4850
},
{
"epoch": 0.41,
"grad_norm": 0.1500246375799179,
"learning_rate": 7.96129952988583e-06,
"loss": 0.0012,
"step": 4860
},
{
"epoch": 0.41,
"grad_norm": 0.05663210526108742,
"learning_rate": 7.957102081934186e-06,
"loss": 0.002,
"step": 4870
},
{
"epoch": 0.41,
"grad_norm": 0.004468943923711777,
"learning_rate": 7.952904633982539e-06,
"loss": 0.0006,
"step": 4880
},
{
"epoch": 0.41,
"grad_norm": 0.049168992787599564,
"learning_rate": 7.948707186030893e-06,
"loss": 0.0028,
"step": 4890
},
{
"epoch": 0.41,
"grad_norm": 0.017784954980015755,
"learning_rate": 7.944509738079248e-06,
"loss": 0.0034,
"step": 4900
},
{
"epoch": 0.41,
"grad_norm": 0.0007347997161559761,
"learning_rate": 7.940312290127602e-06,
"loss": 0.0009,
"step": 4910
},
{
"epoch": 0.41,
"grad_norm": 0.07714752852916718,
"learning_rate": 7.936114842175959e-06,
"loss": 0.0005,
"step": 4920
},
{
"epoch": 0.41,
"grad_norm": 0.24701713025569916,
"learning_rate": 7.931917394224312e-06,
"loss": 0.0027,
"step": 4930
},
{
"epoch": 0.41,
"grad_norm": 0.18050765991210938,
"learning_rate": 7.927719946272668e-06,
"loss": 0.0021,
"step": 4940
},
{
"epoch": 0.42,
"grad_norm": 0.10449690371751785,
"learning_rate": 7.923522498321021e-06,
"loss": 0.0026,
"step": 4950
},
{
"epoch": 0.42,
"grad_norm": 0.0794496163725853,
"learning_rate": 7.919325050369376e-06,
"loss": 0.002,
"step": 4960
},
{
"epoch": 0.42,
"grad_norm": 0.005070607643574476,
"learning_rate": 7.91512760241773e-06,
"loss": 0.0015,
"step": 4970
},
{
"epoch": 0.42,
"grad_norm": 0.23382118344306946,
"learning_rate": 7.910930154466085e-06,
"loss": 0.0014,
"step": 4980
},
{
"epoch": 0.42,
"grad_norm": 0.023488784208893776,
"learning_rate": 7.90673270651444e-06,
"loss": 0.001,
"step": 4990
},
{
"epoch": 0.42,
"grad_norm": 0.08839926868677139,
"learning_rate": 7.902535258562794e-06,
"loss": 0.0004,
"step": 5000
},
{
"epoch": 0.42,
"grad_norm": 0.11559606343507767,
"learning_rate": 7.898337810611149e-06,
"loss": 0.0061,
"step": 5010
},
{
"epoch": 0.42,
"grad_norm": 0.0005539359408430755,
"learning_rate": 7.894140362659503e-06,
"loss": 0.0017,
"step": 5020
},
{
"epoch": 0.42,
"grad_norm": 0.04608330875635147,
"learning_rate": 7.88994291470786e-06,
"loss": 0.0006,
"step": 5030
},
{
"epoch": 0.42,
"grad_norm": 0.029279248788952827,
"learning_rate": 7.885745466756213e-06,
"loss": 0.002,
"step": 5040
},
{
"epoch": 0.42,
"grad_norm": 0.055397383868694305,
"learning_rate": 7.881548018804567e-06,
"loss": 0.004,
"step": 5050
},
{
"epoch": 0.42,
"grad_norm": 0.014438183978199959,
"learning_rate": 7.877350570852922e-06,
"loss": 0.0007,
"step": 5060
},
{
"epoch": 0.43,
"grad_norm": 0.00333590479567647,
"learning_rate": 7.873153122901276e-06,
"loss": 0.002,
"step": 5070
},
{
"epoch": 0.43,
"grad_norm": 0.01975451223552227,
"learning_rate": 7.868955674949631e-06,
"loss": 0.0017,
"step": 5080
},
{
"epoch": 0.43,
"grad_norm": 0.2275656759738922,
"learning_rate": 7.864758226997986e-06,
"loss": 0.0007,
"step": 5090
},
{
"epoch": 0.43,
"grad_norm": 0.009497923776507378,
"learning_rate": 7.86056077904634e-06,
"loss": 0.0005,
"step": 5100
},
{
"epoch": 0.43,
"grad_norm": 0.12707078456878662,
"learning_rate": 7.856363331094695e-06,
"loss": 0.0022,
"step": 5110
},
{
"epoch": 0.43,
"grad_norm": 0.13071304559707642,
"learning_rate": 7.85216588314305e-06,
"loss": 0.0012,
"step": 5120
},
{
"epoch": 0.43,
"grad_norm": 0.12163705378770828,
"learning_rate": 7.847968435191404e-06,
"loss": 0.0025,
"step": 5130
},
{
"epoch": 0.43,
"grad_norm": 0.12689749896526337,
"learning_rate": 7.843770987239759e-06,
"loss": 0.0018,
"step": 5140
},
{
"epoch": 0.43,
"grad_norm": 0.24196821451187134,
"learning_rate": 7.839573539288113e-06,
"loss": 0.0024,
"step": 5150
},
{
"epoch": 0.43,
"grad_norm": 0.0556950718164444,
"learning_rate": 7.835376091336468e-06,
"loss": 0.001,
"step": 5160
},
{
"epoch": 0.43,
"grad_norm": 0.004124140832573175,
"learning_rate": 7.831178643384823e-06,
"loss": 0.0006,
"step": 5170
},
{
"epoch": 0.43,
"grad_norm": 0.03047008253633976,
"learning_rate": 7.826981195433177e-06,
"loss": 0.002,
"step": 5180
},
{
"epoch": 0.44,
"grad_norm": 0.6133390069007874,
"learning_rate": 7.822783747481532e-06,
"loss": 0.0014,
"step": 5190
},
{
"epoch": 0.44,
"grad_norm": 0.1056409627199173,
"learning_rate": 7.818586299529886e-06,
"loss": 0.0018,
"step": 5200
},
{
"epoch": 0.44,
"grad_norm": 0.02283824421465397,
"learning_rate": 7.814388851578241e-06,
"loss": 0.0018,
"step": 5210
},
{
"epoch": 0.44,
"grad_norm": 0.04858771339058876,
"learning_rate": 7.810191403626596e-06,
"loss": 0.0027,
"step": 5220
},
{
"epoch": 0.44,
"grad_norm": 0.001151357777416706,
"learning_rate": 7.80599395567495e-06,
"loss": 0.0011,
"step": 5230
},
{
"epoch": 0.44,
"grad_norm": 0.022891288623213768,
"learning_rate": 7.801796507723305e-06,
"loss": 0.0008,
"step": 5240
},
{
"epoch": 0.44,
"grad_norm": 0.2969476580619812,
"learning_rate": 7.79759905977166e-06,
"loss": 0.0008,
"step": 5250
},
{
"epoch": 0.44,
"grad_norm": 0.07708311080932617,
"learning_rate": 7.793401611820014e-06,
"loss": 0.003,
"step": 5260
},
{
"epoch": 0.44,
"grad_norm": 0.048849739134311676,
"learning_rate": 7.789204163868369e-06,
"loss": 0.0026,
"step": 5270
},
{
"epoch": 0.44,
"grad_norm": 0.04516473039984703,
"learning_rate": 7.785006715916723e-06,
"loss": 0.002,
"step": 5280
},
{
"epoch": 0.44,
"grad_norm": 0.0004609221068676561,
"learning_rate": 7.780809267965078e-06,
"loss": 0.0015,
"step": 5290
},
{
"epoch": 0.44,
"grad_norm": 0.04255954921245575,
"learning_rate": 7.776611820013433e-06,
"loss": 0.0022,
"step": 5300
},
{
"epoch": 0.45,
"grad_norm": 0.22186703979969025,
"learning_rate": 7.772414372061787e-06,
"loss": 0.0015,
"step": 5310
},
{
"epoch": 0.45,
"grad_norm": 0.014675665646791458,
"learning_rate": 7.768216924110142e-06,
"loss": 0.0005,
"step": 5320
},
{
"epoch": 0.45,
"grad_norm": 0.11927191913127899,
"learning_rate": 7.764019476158496e-06,
"loss": 0.0016,
"step": 5330
},
{
"epoch": 0.45,
"grad_norm": 0.24450215697288513,
"learning_rate": 7.759822028206851e-06,
"loss": 0.0009,
"step": 5340
},
{
"epoch": 0.45,
"grad_norm": 0.012072744779288769,
"learning_rate": 7.755624580255206e-06,
"loss": 0.0005,
"step": 5350
},
{
"epoch": 0.45,
"grad_norm": 0.11228229105472565,
"learning_rate": 7.75142713230356e-06,
"loss": 0.0031,
"step": 5360
},
{
"epoch": 0.45,
"grad_norm": 0.007494674064218998,
"learning_rate": 7.747229684351915e-06,
"loss": 0.0074,
"step": 5370
},
{
"epoch": 0.45,
"grad_norm": 0.020272986963391304,
"learning_rate": 7.74303223640027e-06,
"loss": 0.001,
"step": 5380
},
{
"epoch": 0.45,
"grad_norm": 0.041734930127859116,
"learning_rate": 7.738834788448624e-06,
"loss": 0.0011,
"step": 5390
},
{
"epoch": 0.45,
"grad_norm": 0.033374298363924026,
"learning_rate": 7.734637340496979e-06,
"loss": 0.0042,
"step": 5400
},
{
"epoch": 0.45,
"grad_norm": 0.007725745439529419,
"learning_rate": 7.730439892545333e-06,
"loss": 0.0008,
"step": 5410
},
{
"epoch": 0.46,
"grad_norm": 0.40455764532089233,
"learning_rate": 7.726242444593688e-06,
"loss": 0.0013,
"step": 5420
},
{
"epoch": 0.46,
"grad_norm": 0.0638241171836853,
"learning_rate": 7.722044996642043e-06,
"loss": 0.0012,
"step": 5430
},
{
"epoch": 0.46,
"grad_norm": 0.5149130821228027,
"learning_rate": 7.717847548690397e-06,
"loss": 0.0022,
"step": 5440
},
{
"epoch": 0.46,
"grad_norm": 0.041986189782619476,
"learning_rate": 7.71365010073875e-06,
"loss": 0.0028,
"step": 5450
},
{
"epoch": 0.46,
"grad_norm": 0.0023236775305122137,
"learning_rate": 7.709452652787106e-06,
"loss": 0.0006,
"step": 5460
},
{
"epoch": 0.46,
"grad_norm": 0.08948805928230286,
"learning_rate": 7.705255204835461e-06,
"loss": 0.0007,
"step": 5470
},
{
"epoch": 0.46,
"grad_norm": 0.03315718472003937,
"learning_rate": 7.701057756883816e-06,
"loss": 0.0018,
"step": 5480
},
{
"epoch": 0.46,
"grad_norm": 0.15775519609451294,
"learning_rate": 7.69686030893217e-06,
"loss": 0.0009,
"step": 5490
},
{
"epoch": 0.46,
"grad_norm": 0.00936898123472929,
"learning_rate": 7.692662860980525e-06,
"loss": 0.002,
"step": 5500
},
{
"epoch": 0.46,
"grad_norm": 0.00790819339454174,
"learning_rate": 7.68846541302888e-06,
"loss": 0.0011,
"step": 5510
},
{
"epoch": 0.46,
"grad_norm": 0.021303502842783928,
"learning_rate": 7.684267965077234e-06,
"loss": 0.0007,
"step": 5520
},
{
"epoch": 0.46,
"grad_norm": 0.0739978775382042,
"learning_rate": 7.680070517125589e-06,
"loss": 0.0019,
"step": 5530
},
{
"epoch": 0.47,
"grad_norm": 0.0839531272649765,
"learning_rate": 7.675873069173942e-06,
"loss": 0.0021,
"step": 5540
},
{
"epoch": 0.47,
"grad_norm": 0.6550424695014954,
"learning_rate": 7.671675621222298e-06,
"loss": 0.0017,
"step": 5550
},
{
"epoch": 0.47,
"grad_norm": 0.015562259592115879,
"learning_rate": 7.667478173270651e-06,
"loss": 0.0006,
"step": 5560
},
{
"epoch": 0.47,
"grad_norm": 0.29718390107154846,
"learning_rate": 7.663280725319007e-06,
"loss": 0.0016,
"step": 5570
},
{
"epoch": 0.47,
"grad_norm": 0.02291831001639366,
"learning_rate": 7.659083277367362e-06,
"loss": 0.0004,
"step": 5580
},
{
"epoch": 0.47,
"grad_norm": 1.3322148323059082,
"learning_rate": 7.654885829415716e-06,
"loss": 0.0037,
"step": 5590
},
{
"epoch": 0.47,
"grad_norm": 0.6336252689361572,
"learning_rate": 7.650688381464071e-06,
"loss": 0.0022,
"step": 5600
},
{
"epoch": 0.47,
"grad_norm": 0.17173808813095093,
"learning_rate": 7.646490933512426e-06,
"loss": 0.0007,
"step": 5610
},
{
"epoch": 0.47,
"grad_norm": 0.01680012419819832,
"learning_rate": 7.64229348556078e-06,
"loss": 0.0013,
"step": 5620
},
{
"epoch": 0.47,
"grad_norm": 0.07509610801935196,
"learning_rate": 7.638096037609133e-06,
"loss": 0.0054,
"step": 5630
},
{
"epoch": 0.47,
"grad_norm": 0.01283180620521307,
"learning_rate": 7.63389858965749e-06,
"loss": 0.0016,
"step": 5640
},
{
"epoch": 0.47,
"grad_norm": 0.03960631042718887,
"learning_rate": 7.629701141705842e-06,
"loss": 0.0005,
"step": 5650
},
{
"epoch": 0.48,
"grad_norm": 0.03129389509558678,
"learning_rate": 7.625503693754198e-06,
"loss": 0.0005,
"step": 5660
},
{
"epoch": 0.48,
"grad_norm": 0.01344276126474142,
"learning_rate": 7.621306245802553e-06,
"loss": 0.0012,
"step": 5670
},
{
"epoch": 0.48,
"grad_norm": 0.010192753747105598,
"learning_rate": 7.617108797850907e-06,
"loss": 0.0006,
"step": 5680
},
{
"epoch": 0.48,
"grad_norm": 0.029010513797402382,
"learning_rate": 7.612911349899263e-06,
"loss": 0.0031,
"step": 5690
},
{
"epoch": 0.48,
"grad_norm": 0.015769800171256065,
"learning_rate": 7.6087139019476164e-06,
"loss": 0.0005,
"step": 5700
},
{
"epoch": 0.48,
"grad_norm": 0.015835179015994072,
"learning_rate": 7.604516453995972e-06,
"loss": 0.0005,
"step": 5710
},
{
"epoch": 0.48,
"grad_norm": 0.009272762574255466,
"learning_rate": 7.600319006044326e-06,
"loss": 0.0018,
"step": 5720
},
{
"epoch": 0.48,
"grad_norm": 0.06365930289030075,
"learning_rate": 7.59612155809268e-06,
"loss": 0.0011,
"step": 5730
},
{
"epoch": 0.48,
"grad_norm": 0.011347809806466103,
"learning_rate": 7.591924110141035e-06,
"loss": 0.0011,
"step": 5740
},
{
"epoch": 0.48,
"grad_norm": 0.029438406229019165,
"learning_rate": 7.5877266621893895e-06,
"loss": 0.0018,
"step": 5750
},
{
"epoch": 0.48,
"grad_norm": 0.0030098084826022387,
"learning_rate": 7.583529214237743e-06,
"loss": 0.0026,
"step": 5760
},
{
"epoch": 0.48,
"grad_norm": 0.07430537790060043,
"learning_rate": 7.579331766286099e-06,
"loss": 0.0011,
"step": 5770
},
{
"epoch": 0.49,
"grad_norm": 0.2529279291629791,
"learning_rate": 7.5751343183344525e-06,
"loss": 0.001,
"step": 5780
},
{
"epoch": 0.49,
"grad_norm": 0.0054864161647856236,
"learning_rate": 7.570936870382808e-06,
"loss": 0.0004,
"step": 5790
},
{
"epoch": 0.49,
"grad_norm": 0.05911405757069588,
"learning_rate": 7.566739422431163e-06,
"loss": 0.0005,
"step": 5800
},
{
"epoch": 0.49,
"grad_norm": 0.03409591317176819,
"learning_rate": 7.562541974479517e-06,
"loss": 0.0005,
"step": 5810
},
{
"epoch": 0.49,
"grad_norm": 0.013324552215635777,
"learning_rate": 7.558344526527872e-06,
"loss": 0.0016,
"step": 5820
},
{
"epoch": 0.49,
"grad_norm": 0.012020808644592762,
"learning_rate": 7.5541470785762265e-06,
"loss": 0.0038,
"step": 5830
},
{
"epoch": 0.49,
"grad_norm": 0.2555389404296875,
"learning_rate": 7.549949630624581e-06,
"loss": 0.0011,
"step": 5840
},
{
"epoch": 0.49,
"grad_norm": 0.09523122757673264,
"learning_rate": 7.545752182672935e-06,
"loss": 0.0008,
"step": 5850
},
{
"epoch": 0.49,
"grad_norm": 0.07814309000968933,
"learning_rate": 7.54155473472129e-06,
"loss": 0.003,
"step": 5860
},
{
"epoch": 0.49,
"grad_norm": 0.018278853967785835,
"learning_rate": 7.537357286769644e-06,
"loss": 0.0016,
"step": 5870
},
{
"epoch": 0.49,
"grad_norm": 0.13258874416351318,
"learning_rate": 7.5331598388179996e-06,
"loss": 0.002,
"step": 5880
},
{
"epoch": 0.49,
"grad_norm": 0.008451174944639206,
"learning_rate": 7.528962390866353e-06,
"loss": 0.0016,
"step": 5890
},
{
"epoch": 0.5,
"grad_norm": 0.028404515236616135,
"learning_rate": 7.524764942914709e-06,
"loss": 0.0007,
"step": 5900
},
{
"epoch": 0.5,
"grad_norm": 0.46089375019073486,
"learning_rate": 7.520567494963063e-06,
"loss": 0.0011,
"step": 5910
},
{
"epoch": 0.5,
"grad_norm": 0.19054672122001648,
"learning_rate": 7.516370047011417e-06,
"loss": 0.0037,
"step": 5920
},
{
"epoch": 0.5,
"grad_norm": 0.7026522755622864,
"learning_rate": 7.512172599059773e-06,
"loss": 0.0026,
"step": 5930
},
{
"epoch": 0.5,
"grad_norm": 0.026187479496002197,
"learning_rate": 7.507975151108126e-06,
"loss": 0.0013,
"step": 5940
},
{
"epoch": 0.5,
"grad_norm": 0.0011275768047198653,
"learning_rate": 7.503777703156482e-06,
"loss": 0.0026,
"step": 5950
},
{
"epoch": 0.5,
"grad_norm": 0.007955752313137054,
"learning_rate": 7.499580255204836e-06,
"loss": 0.0011,
"step": 5960
},
{
"epoch": 0.5,
"grad_norm": 0.008177239447832108,
"learning_rate": 7.495382807253191e-06,
"loss": 0.0013,
"step": 5970
},
{
"epoch": 0.5,
"grad_norm": 0.01679716818034649,
"learning_rate": 7.491185359301545e-06,
"loss": 0.0008,
"step": 5980
},
{
"epoch": 0.5,
"grad_norm": 0.022488052025437355,
"learning_rate": 7.4869879113499e-06,
"loss": 0.0005,
"step": 5990
},
{
"epoch": 0.5,
"grad_norm": 0.00388827221468091,
"learning_rate": 7.482790463398254e-06,
"loss": 0.0054,
"step": 6000
},
{
"epoch": 0.5,
"grad_norm": 0.040717240422964096,
"learning_rate": 7.478593015446609e-06,
"loss": 0.001,
"step": 6010
},
{
"epoch": 0.51,
"grad_norm": 0.0023668904323130846,
"learning_rate": 7.474395567494964e-06,
"loss": 0.0004,
"step": 6020
},
{
"epoch": 0.51,
"grad_norm": 0.13439880311489105,
"learning_rate": 7.470198119543318e-06,
"loss": 0.007,
"step": 6030
},
{
"epoch": 0.51,
"grad_norm": 0.3206893503665924,
"learning_rate": 7.4660006715916734e-06,
"loss": 0.0048,
"step": 6040
},
{
"epoch": 0.51,
"grad_norm": 0.09794365614652634,
"learning_rate": 7.461803223640027e-06,
"loss": 0.0006,
"step": 6050
},
{
"epoch": 0.51,
"grad_norm": 0.3034574091434479,
"learning_rate": 7.457605775688383e-06,
"loss": 0.0022,
"step": 6060
},
{
"epoch": 0.51,
"grad_norm": 0.06691265106201172,
"learning_rate": 7.4534083277367364e-06,
"loss": 0.0008,
"step": 6070
},
{
"epoch": 0.51,
"grad_norm": 0.10203710198402405,
"learning_rate": 7.449210879785092e-06,
"loss": 0.0007,
"step": 6080
},
{
"epoch": 0.51,
"grad_norm": 0.014348863624036312,
"learning_rate": 7.445013431833446e-06,
"loss": 0.0008,
"step": 6090
},
{
"epoch": 0.51,
"grad_norm": 0.03633236512541771,
"learning_rate": 7.4408159838818e-06,
"loss": 0.002,
"step": 6100
},
{
"epoch": 0.51,
"grad_norm": 0.06054764240980148,
"learning_rate": 7.436618535930155e-06,
"loss": 0.0024,
"step": 6110
},
{
"epoch": 0.51,
"grad_norm": 0.021807057783007622,
"learning_rate": 7.4324210879785095e-06,
"loss": 0.0019,
"step": 6120
},
{
"epoch": 0.51,
"grad_norm": 0.024098997935652733,
"learning_rate": 7.428223640026865e-06,
"loss": 0.0008,
"step": 6130
},
{
"epoch": 0.52,
"grad_norm": 0.08782277256250381,
"learning_rate": 7.424026192075219e-06,
"loss": 0.0018,
"step": 6140
},
{
"epoch": 0.52,
"grad_norm": 1.0001189708709717,
"learning_rate": 7.419828744123574e-06,
"loss": 0.0021,
"step": 6150
},
{
"epoch": 0.52,
"grad_norm": 0.03228935971856117,
"learning_rate": 7.415631296171928e-06,
"loss": 0.0008,
"step": 6160
},
{
"epoch": 0.52,
"grad_norm": 0.014304124750196934,
"learning_rate": 7.411433848220283e-06,
"loss": 0.0062,
"step": 6170
},
{
"epoch": 0.52,
"grad_norm": 0.0017881994135677814,
"learning_rate": 7.407236400268637e-06,
"loss": 0.0013,
"step": 6180
},
{
"epoch": 0.52,
"grad_norm": 0.16372914612293243,
"learning_rate": 7.403038952316992e-06,
"loss": 0.0015,
"step": 6190
},
{
"epoch": 0.52,
"grad_norm": 0.04461158439517021,
"learning_rate": 7.3988415043653465e-06,
"loss": 0.0067,
"step": 6200
},
{
"epoch": 0.52,
"grad_norm": 0.09605967253446579,
"learning_rate": 7.394644056413701e-06,
"loss": 0.0008,
"step": 6210
},
{
"epoch": 0.52,
"grad_norm": 0.6841378211975098,
"learning_rate": 7.390446608462055e-06,
"loss": 0.005,
"step": 6220
},
{
"epoch": 0.52,
"grad_norm": 0.04636238515377045,
"learning_rate": 7.38624916051041e-06,
"loss": 0.0007,
"step": 6230
},
{
"epoch": 0.52,
"grad_norm": 0.5079997777938843,
"learning_rate": 7.382051712558766e-06,
"loss": 0.0018,
"step": 6240
},
{
"epoch": 0.52,
"grad_norm": 0.09513850510120392,
"learning_rate": 7.3778542646071196e-06,
"loss": 0.0024,
"step": 6250
},
{
"epoch": 0.53,
"grad_norm": 0.06880282610654831,
"learning_rate": 7.373656816655474e-06,
"loss": 0.0034,
"step": 6260
},
{
"epoch": 0.53,
"grad_norm": 0.18673266470432281,
"learning_rate": 7.369459368703829e-06,
"loss": 0.0037,
"step": 6270
},
{
"epoch": 0.53,
"grad_norm": 0.03851670026779175,
"learning_rate": 7.365261920752183e-06,
"loss": 0.0005,
"step": 6280
},
{
"epoch": 0.53,
"grad_norm": 0.021820688620209694,
"learning_rate": 7.361064472800537e-06,
"loss": 0.0009,
"step": 6290
},
{
"epoch": 0.53,
"grad_norm": 0.008911089040338993,
"learning_rate": 7.356867024848893e-06,
"loss": 0.0019,
"step": 6300
},
{
"epoch": 0.53,
"grad_norm": 0.016666969284415245,
"learning_rate": 7.3526695768972464e-06,
"loss": 0.0006,
"step": 6310
},
{
"epoch": 0.53,
"grad_norm": 0.0268320981413126,
"learning_rate": 7.348472128945602e-06,
"loss": 0.0018,
"step": 6320
},
{
"epoch": 0.53,
"grad_norm": 0.006555848754942417,
"learning_rate": 7.344274680993956e-06,
"loss": 0.0009,
"step": 6330
},
{
"epoch": 0.53,
"grad_norm": 0.014750578440725803,
"learning_rate": 7.340077233042311e-06,
"loss": 0.0011,
"step": 6340
},
{
"epoch": 0.53,
"grad_norm": 0.011571898125112057,
"learning_rate": 7.335879785090666e-06,
"loss": 0.0025,
"step": 6350
},
{
"epoch": 0.53,
"grad_norm": 0.014332173392176628,
"learning_rate": 7.33168233713902e-06,
"loss": 0.0006,
"step": 6360
},
{
"epoch": 0.53,
"grad_norm": 0.03584123030304909,
"learning_rate": 7.327484889187375e-06,
"loss": 0.001,
"step": 6370
},
{
"epoch": 0.54,
"grad_norm": 0.0009831301867961884,
"learning_rate": 7.323287441235729e-06,
"loss": 0.0016,
"step": 6380
},
{
"epoch": 0.54,
"grad_norm": 0.01324765756726265,
"learning_rate": 7.319089993284084e-06,
"loss": 0.0005,
"step": 6390
},
{
"epoch": 0.54,
"grad_norm": 0.2932877540588379,
"learning_rate": 7.314892545332438e-06,
"loss": 0.0017,
"step": 6400
},
{
"epoch": 0.54,
"grad_norm": 0.01268022321164608,
"learning_rate": 7.3106950973807935e-06,
"loss": 0.003,
"step": 6410
},
{
"epoch": 0.54,
"grad_norm": 0.04986048489809036,
"learning_rate": 7.306497649429147e-06,
"loss": 0.0007,
"step": 6420
},
{
"epoch": 0.54,
"grad_norm": 0.019490044564008713,
"learning_rate": 7.302300201477503e-06,
"loss": 0.0052,
"step": 6430
},
{
"epoch": 0.54,
"grad_norm": 0.7576485872268677,
"learning_rate": 7.2981027535258565e-06,
"loss": 0.0013,
"step": 6440
},
{
"epoch": 0.54,
"grad_norm": 0.1456100046634674,
"learning_rate": 7.293905305574212e-06,
"loss": 0.0009,
"step": 6450
},
{
"epoch": 0.54,
"grad_norm": 0.030114591121673584,
"learning_rate": 7.2897078576225665e-06,
"loss": 0.0015,
"step": 6460
},
{
"epoch": 0.54,
"grad_norm": 0.02888800948858261,
"learning_rate": 7.28551040967092e-06,
"loss": 0.0003,
"step": 6470
},
{
"epoch": 0.54,
"grad_norm": 0.003675349522382021,
"learning_rate": 7.281312961719276e-06,
"loss": 0.0048,
"step": 6480
},
{
"epoch": 0.54,
"grad_norm": 0.010422208346426487,
"learning_rate": 7.2771155137676296e-06,
"loss": 0.0021,
"step": 6490
},
{
"epoch": 0.55,
"grad_norm": 0.009601653553545475,
"learning_rate": 7.272918065815985e-06,
"loss": 0.0018,
"step": 6500
},
{
"epoch": 0.55,
"grad_norm": 0.00894622690975666,
"learning_rate": 7.268720617864339e-06,
"loss": 0.0005,
"step": 6510
},
{
"epoch": 0.55,
"grad_norm": 0.01711229234933853,
"learning_rate": 7.264523169912694e-06,
"loss": 0.0019,
"step": 6520
},
{
"epoch": 0.55,
"grad_norm": 0.25115489959716797,
"learning_rate": 7.260325721961048e-06,
"loss": 0.006,
"step": 6530
},
{
"epoch": 0.55,
"grad_norm": 0.028431080281734467,
"learning_rate": 7.256128274009403e-06,
"loss": 0.0004,
"step": 6540
},
{
"epoch": 0.55,
"grad_norm": 0.005961297079920769,
"learning_rate": 7.251930826057757e-06,
"loss": 0.0014,
"step": 6550
},
{
"epoch": 0.55,
"grad_norm": 0.017467642202973366,
"learning_rate": 7.247733378106112e-06,
"loss": 0.0008,
"step": 6560
},
{
"epoch": 0.55,
"grad_norm": 0.011652447283267975,
"learning_rate": 7.243535930154467e-06,
"loss": 0.0068,
"step": 6570
},
{
"epoch": 0.55,
"grad_norm": 0.0026287129148840904,
"learning_rate": 7.239338482202821e-06,
"loss": 0.002,
"step": 6580
},
{
"epoch": 0.55,
"grad_norm": 0.27473968267440796,
"learning_rate": 7.2351410342511766e-06,
"loss": 0.0024,
"step": 6590
},
{
"epoch": 0.55,
"grad_norm": 0.010974191129207611,
"learning_rate": 7.23094358629953e-06,
"loss": 0.0006,
"step": 6600
},
{
"epoch": 0.55,
"grad_norm": 0.0035286422353237867,
"learning_rate": 7.226746138347886e-06,
"loss": 0.0068,
"step": 6610
},
{
"epoch": 0.56,
"grad_norm": 0.09986690431833267,
"learning_rate": 7.22254869039624e-06,
"loss": 0.0006,
"step": 6620
},
{
"epoch": 0.56,
"grad_norm": 0.0497036911547184,
"learning_rate": 7.218351242444594e-06,
"loss": 0.0014,
"step": 6630
},
{
"epoch": 0.56,
"grad_norm": 0.0852247104048729,
"learning_rate": 7.214153794492949e-06,
"loss": 0.0039,
"step": 6640
},
{
"epoch": 0.56,
"grad_norm": 0.2507276237010956,
"learning_rate": 7.2099563465413034e-06,
"loss": 0.0084,
"step": 6650
},
{
"epoch": 0.56,
"grad_norm": 0.08247136324644089,
"learning_rate": 7.205758898589657e-06,
"loss": 0.0004,
"step": 6660
},
{
"epoch": 0.56,
"grad_norm": 0.010528313927352428,
"learning_rate": 7.201561450638013e-06,
"loss": 0.003,
"step": 6670
},
{
"epoch": 0.56,
"grad_norm": 0.014672808349132538,
"learning_rate": 7.197364002686368e-06,
"loss": 0.0015,
"step": 6680
},
{
"epoch": 0.56,
"grad_norm": 0.027063751593232155,
"learning_rate": 7.193166554734722e-06,
"loss": 0.0004,
"step": 6690
},
{
"epoch": 0.56,
"grad_norm": 0.007319859229028225,
"learning_rate": 7.1889691067830765e-06,
"loss": 0.0009,
"step": 6700
},
{
"epoch": 0.56,
"grad_norm": 0.17890995740890503,
"learning_rate": 7.184771658831431e-06,
"loss": 0.0006,
"step": 6710
},
{
"epoch": 0.56,
"grad_norm": 0.0006071716779842973,
"learning_rate": 7.180574210879786e-06,
"loss": 0.0091,
"step": 6720
},
{
"epoch": 0.56,
"grad_norm": 0.0499555803835392,
"learning_rate": 7.17637676292814e-06,
"loss": 0.0012,
"step": 6730
},
{
"epoch": 0.57,
"grad_norm": 0.1350906640291214,
"learning_rate": 7.172179314976495e-06,
"loss": 0.0011,
"step": 6740
},
{
"epoch": 0.57,
"grad_norm": 0.026632359251379967,
"learning_rate": 7.167981867024849e-06,
"loss": 0.0014,
"step": 6750
},
{
"epoch": 0.57,
"grad_norm": 0.011988186277449131,
"learning_rate": 7.163784419073204e-06,
"loss": 0.0044,
"step": 6760
},
{
"epoch": 0.57,
"grad_norm": 0.004848783370107412,
"learning_rate": 7.159586971121558e-06,
"loss": 0.0031,
"step": 6770
},
{
"epoch": 0.57,
"grad_norm": 0.1688844859600067,
"learning_rate": 7.1553895231699135e-06,
"loss": 0.0012,
"step": 6780
},
{
"epoch": 0.57,
"grad_norm": 0.05839391052722931,
"learning_rate": 7.151192075218268e-06,
"loss": 0.0014,
"step": 6790
},
{
"epoch": 0.57,
"grad_norm": 0.006566599011421204,
"learning_rate": 7.146994627266623e-06,
"loss": 0.0014,
"step": 6800
},
{
"epoch": 0.57,
"grad_norm": 0.028615593910217285,
"learning_rate": 7.142797179314977e-06,
"loss": 0.0019,
"step": 6810
},
{
"epoch": 0.57,
"grad_norm": 0.029592135921120644,
"learning_rate": 7.138599731363331e-06,
"loss": 0.0008,
"step": 6820
},
{
"epoch": 0.57,
"grad_norm": 0.027602553367614746,
"learning_rate": 7.1344022834116866e-06,
"loss": 0.0006,
"step": 6830
},
{
"epoch": 0.57,
"grad_norm": 0.012585037387907505,
"learning_rate": 7.13020483546004e-06,
"loss": 0.0016,
"step": 6840
},
{
"epoch": 0.58,
"grad_norm": 0.003249362576752901,
"learning_rate": 7.126007387508396e-06,
"loss": 0.0019,
"step": 6850
},
{
"epoch": 0.58,
"grad_norm": 0.6995984315872192,
"learning_rate": 7.1218099395567496e-06,
"loss": 0.0043,
"step": 6860
},
{
"epoch": 0.58,
"grad_norm": 0.02927136979997158,
"learning_rate": 7.117612491605105e-06,
"loss": 0.0014,
"step": 6870
},
{
"epoch": 0.58,
"grad_norm": 0.021853366866707802,
"learning_rate": 7.113415043653459e-06,
"loss": 0.002,
"step": 6880
},
{
"epoch": 0.58,
"grad_norm": 0.012909365817904472,
"learning_rate": 7.109217595701814e-06,
"loss": 0.0006,
"step": 6890
},
{
"epoch": 0.58,
"grad_norm": 0.09880993515253067,
"learning_rate": 7.105020147750169e-06,
"loss": 0.0012,
"step": 6900
},
{
"epoch": 0.58,
"grad_norm": 0.07496526092290878,
"learning_rate": 7.100822699798523e-06,
"loss": 0.0013,
"step": 6910
},
{
"epoch": 0.58,
"grad_norm": 0.34422990679740906,
"learning_rate": 7.096625251846878e-06,
"loss": 0.0015,
"step": 6920
},
{
"epoch": 0.58,
"grad_norm": 0.006235859822481871,
"learning_rate": 7.092427803895232e-06,
"loss": 0.0018,
"step": 6930
},
{
"epoch": 0.58,
"grad_norm": 0.03030078299343586,
"learning_rate": 7.088230355943587e-06,
"loss": 0.0011,
"step": 6940
},
{
"epoch": 0.58,
"grad_norm": 0.06541412323713303,
"learning_rate": 7.084032907991941e-06,
"loss": 0.001,
"step": 6950
},
{
"epoch": 0.58,
"grad_norm": 0.000910624279640615,
"learning_rate": 7.079835460040297e-06,
"loss": 0.0032,
"step": 6960
},
{
"epoch": 0.59,
"grad_norm": 0.004598297644406557,
"learning_rate": 7.07563801208865e-06,
"loss": 0.0007,
"step": 6970
},
{
"epoch": 0.59,
"grad_norm": 0.02102428674697876,
"learning_rate": 7.071440564137006e-06,
"loss": 0.0022,
"step": 6980
},
{
"epoch": 0.59,
"grad_norm": 0.17559894919395447,
"learning_rate": 7.06724311618536e-06,
"loss": 0.0029,
"step": 6990
},
{
"epoch": 0.59,
"grad_norm": 0.07612307369709015,
"learning_rate": 7.063045668233714e-06,
"loss": 0.0014,
"step": 7000
},
{
"epoch": 0.59,
"grad_norm": 0.25046348571777344,
"learning_rate": 7.05884822028207e-06,
"loss": 0.0009,
"step": 7010
},
{
"epoch": 0.59,
"grad_norm": 0.01964881271123886,
"learning_rate": 7.0546507723304234e-06,
"loss": 0.0009,
"step": 7020
},
{
"epoch": 0.59,
"grad_norm": 0.13353663682937622,
"learning_rate": 7.050453324378779e-06,
"loss": 0.003,
"step": 7030
},
{
"epoch": 0.59,
"grad_norm": 0.31505197286605835,
"learning_rate": 7.046255876427133e-06,
"loss": 0.0011,
"step": 7040
},
{
"epoch": 0.59,
"grad_norm": 0.05410172790288925,
"learning_rate": 7.042058428475488e-06,
"loss": 0.001,
"step": 7050
},
{
"epoch": 0.59,
"grad_norm": 0.015809668228030205,
"learning_rate": 7.037860980523842e-06,
"loss": 0.0031,
"step": 7060
},
{
"epoch": 0.59,
"grad_norm": 0.0416773185133934,
"learning_rate": 7.0336635325721965e-06,
"loss": 0.001,
"step": 7070
},
{
"epoch": 0.59,
"grad_norm": 0.018480846658349037,
"learning_rate": 7.029466084620551e-06,
"loss": 0.0023,
"step": 7080
},
{
"epoch": 0.6,
"grad_norm": 0.0960954800248146,
"learning_rate": 7.025268636668906e-06,
"loss": 0.0011,
"step": 7090
},
{
"epoch": 0.6,
"grad_norm": 0.08874082565307617,
"learning_rate": 7.02107118871726e-06,
"loss": 0.0005,
"step": 7100
},
{
"epoch": 0.6,
"grad_norm": 0.011005674488842487,
"learning_rate": 7.016873740765615e-06,
"loss": 0.001,
"step": 7110
},
{
"epoch": 0.6,
"grad_norm": 0.17913727462291718,
"learning_rate": 7.0126762928139705e-06,
"loss": 0.0005,
"step": 7120
},
{
"epoch": 0.6,
"grad_norm": 0.010488308034837246,
"learning_rate": 7.008478844862324e-06,
"loss": 0.0008,
"step": 7130
},
{
"epoch": 0.6,
"grad_norm": 0.03276985511183739,
"learning_rate": 7.00428139691068e-06,
"loss": 0.0034,
"step": 7140
},
{
"epoch": 0.6,
"grad_norm": 0.03501226380467415,
"learning_rate": 7.0000839489590335e-06,
"loss": 0.0004,
"step": 7150
},
{
"epoch": 0.6,
"grad_norm": 0.04926338791847229,
"learning_rate": 6.995886501007388e-06,
"loss": 0.0007,
"step": 7160
},
{
"epoch": 0.6,
"grad_norm": 0.00882528442889452,
"learning_rate": 6.991689053055743e-06,
"loss": 0.0009,
"step": 7170
},
{
"epoch": 0.6,
"grad_norm": 0.0117332823574543,
"learning_rate": 6.987491605104097e-06,
"loss": 0.0006,
"step": 7180
},
{
"epoch": 0.6,
"grad_norm": 0.019057177007198334,
"learning_rate": 6.983294157152451e-06,
"loss": 0.0038,
"step": 7190
},
{
"epoch": 0.6,
"grad_norm": 0.0012227203696966171,
"learning_rate": 6.9790967092008066e-06,
"loss": 0.0008,
"step": 7200
},
{
"epoch": 0.61,
"grad_norm": 0.03291918337345123,
"learning_rate": 6.97489926124916e-06,
"loss": 0.0095,
"step": 7210
},
{
"epoch": 0.61,
"grad_norm": 0.7484022378921509,
"learning_rate": 6.970701813297516e-06,
"loss": 0.0129,
"step": 7220
},
{
"epoch": 0.61,
"grad_norm": 0.04075862094759941,
"learning_rate": 6.96650436534587e-06,
"loss": 0.0012,
"step": 7230
},
{
"epoch": 0.61,
"grad_norm": 0.12988588213920593,
"learning_rate": 6.962306917394225e-06,
"loss": 0.0007,
"step": 7240
},
{
"epoch": 0.61,
"grad_norm": 0.04254931956529617,
"learning_rate": 6.95810946944258e-06,
"loss": 0.0004,
"step": 7250
},
{
"epoch": 0.61,
"grad_norm": 0.17507688701152802,
"learning_rate": 6.953912021490934e-06,
"loss": 0.0019,
"step": 7260
},
{
"epoch": 0.61,
"grad_norm": 0.018078705295920372,
"learning_rate": 6.949714573539289e-06,
"loss": 0.0056,
"step": 7270
},
{
"epoch": 0.61,
"grad_norm": 0.016659587621688843,
"learning_rate": 6.945517125587643e-06,
"loss": 0.0005,
"step": 7280
},
{
"epoch": 0.61,
"grad_norm": 0.05568535998463631,
"learning_rate": 6.941319677635998e-06,
"loss": 0.0047,
"step": 7290
},
{
"epoch": 0.61,
"grad_norm": 0.004235047847032547,
"learning_rate": 6.937122229684352e-06,
"loss": 0.0008,
"step": 7300
},
{
"epoch": 0.61,
"grad_norm": 0.017537878826260567,
"learning_rate": 6.932924781732707e-06,
"loss": 0.0019,
"step": 7310
},
{
"epoch": 0.61,
"grad_norm": 0.023402167484164238,
"learning_rate": 6.928727333781061e-06,
"loss": 0.0011,
"step": 7320
},
{
"epoch": 0.62,
"grad_norm": 0.05992670729756355,
"learning_rate": 6.924529885829417e-06,
"loss": 0.0021,
"step": 7330
},
{
"epoch": 0.62,
"grad_norm": 0.005670278798788786,
"learning_rate": 6.920332437877771e-06,
"loss": 0.0007,
"step": 7340
},
{
"epoch": 0.62,
"grad_norm": 0.03450972959399223,
"learning_rate": 6.916134989926126e-06,
"loss": 0.0044,
"step": 7350
},
{
"epoch": 0.62,
"grad_norm": 0.00952198263257742,
"learning_rate": 6.9119375419744805e-06,
"loss": 0.0023,
"step": 7360
},
{
"epoch": 0.62,
"grad_norm": 0.14748263359069824,
"learning_rate": 6.907740094022834e-06,
"loss": 0.0006,
"step": 7370
},
{
"epoch": 0.62,
"grad_norm": 0.008964456617832184,
"learning_rate": 6.90354264607119e-06,
"loss": 0.0025,
"step": 7380
},
{
"epoch": 0.62,
"grad_norm": 0.5854199528694153,
"learning_rate": 6.8993451981195435e-06,
"loss": 0.0018,
"step": 7390
},
{
"epoch": 0.62,
"grad_norm": 0.008997979573905468,
"learning_rate": 6.895147750167899e-06,
"loss": 0.0012,
"step": 7400
},
{
"epoch": 0.62,
"grad_norm": 0.013361281715333462,
"learning_rate": 6.890950302216253e-06,
"loss": 0.0006,
"step": 7410
},
{
"epoch": 0.62,
"grad_norm": 0.01578344777226448,
"learning_rate": 6.886752854264608e-06,
"loss": 0.0005,
"step": 7420
},
{
"epoch": 0.62,
"grad_norm": 0.011211412958800793,
"learning_rate": 6.882555406312962e-06,
"loss": 0.0047,
"step": 7430
},
{
"epoch": 0.62,
"grad_norm": 0.032785531133413315,
"learning_rate": 6.8783579583613166e-06,
"loss": 0.0004,
"step": 7440
},
{
"epoch": 0.63,
"grad_norm": 0.2810578942298889,
"learning_rate": 6.874160510409672e-06,
"loss": 0.0007,
"step": 7450
},
{
"epoch": 0.63,
"grad_norm": 0.012072360143065453,
"learning_rate": 6.869963062458026e-06,
"loss": 0.0006,
"step": 7460
},
{
"epoch": 0.63,
"grad_norm": 0.033594295382499695,
"learning_rate": 6.865765614506381e-06,
"loss": 0.0008,
"step": 7470
},
{
"epoch": 0.63,
"grad_norm": 0.09040901064872742,
"learning_rate": 6.861568166554735e-06,
"loss": 0.0004,
"step": 7480
},
{
"epoch": 0.63,
"grad_norm": 0.006718501914292574,
"learning_rate": 6.8573707186030905e-06,
"loss": 0.0094,
"step": 7490
},
{
"epoch": 0.63,
"grad_norm": 0.3409207761287689,
"learning_rate": 6.853173270651444e-06,
"loss": 0.0009,
"step": 7500
},
{
"epoch": 0.63,
"grad_norm": 0.017661171033978462,
"learning_rate": 6.8489758226998e-06,
"loss": 0.0006,
"step": 7510
},
{
"epoch": 0.63,
"grad_norm": 0.4969084560871124,
"learning_rate": 6.8447783747481535e-06,
"loss": 0.0021,
"step": 7520
},
{
"epoch": 0.63,
"grad_norm": 0.08554006367921829,
"learning_rate": 6.840580926796508e-06,
"loss": 0.0018,
"step": 7530
},
{
"epoch": 0.63,
"grad_norm": 0.0186705831438303,
"learning_rate": 6.836383478844863e-06,
"loss": 0.002,
"step": 7540
},
{
"epoch": 0.63,
"grad_norm": 0.020978650078177452,
"learning_rate": 6.832186030893217e-06,
"loss": 0.0045,
"step": 7550
},
{
"epoch": 0.63,
"grad_norm": 0.043272119015455246,
"learning_rate": 6.827988582941573e-06,
"loss": 0.0018,
"step": 7560
},
{
"epoch": 0.64,
"grad_norm": 0.05586528778076172,
"learning_rate": 6.823791134989927e-06,
"loss": 0.0005,
"step": 7570
},
{
"epoch": 0.64,
"grad_norm": 0.07728602737188339,
"learning_rate": 6.819593687038282e-06,
"loss": 0.0013,
"step": 7580
},
{
"epoch": 0.64,
"grad_norm": 0.23588360846042633,
"learning_rate": 6.815396239086636e-06,
"loss": 0.001,
"step": 7590
},
{
"epoch": 0.64,
"grad_norm": 0.016243470832705498,
"learning_rate": 6.8111987911349904e-06,
"loss": 0.0003,
"step": 7600
},
{
"epoch": 0.64,
"grad_norm": 0.048352282494306564,
"learning_rate": 6.807001343183345e-06,
"loss": 0.0008,
"step": 7610
},
{
"epoch": 0.64,
"grad_norm": 0.013593564741313457,
"learning_rate": 6.8028038952317e-06,
"loss": 0.001,
"step": 7620
},
{
"epoch": 0.64,
"grad_norm": 0.1125347837805748,
"learning_rate": 6.798606447280054e-06,
"loss": 0.0007,
"step": 7630
},
{
"epoch": 0.64,
"grad_norm": 0.040121354162693024,
"learning_rate": 6.794408999328409e-06,
"loss": 0.0023,
"step": 7640
},
{
"epoch": 0.64,
"grad_norm": 0.004880643915385008,
"learning_rate": 6.790211551376763e-06,
"loss": 0.0013,
"step": 7650
},
{
"epoch": 0.64,
"grad_norm": 0.03261597827076912,
"learning_rate": 6.786014103425118e-06,
"loss": 0.0048,
"step": 7660
},
{
"epoch": 0.64,
"grad_norm": 0.001294650137424469,
"learning_rate": 6.781816655473474e-06,
"loss": 0.0062,
"step": 7670
},
{
"epoch": 0.64,
"grad_norm": 0.013758821412920952,
"learning_rate": 6.777619207521827e-06,
"loss": 0.0004,
"step": 7680
},
{
"epoch": 0.65,
"grad_norm": 0.2742080092430115,
"learning_rate": 6.773421759570182e-06,
"loss": 0.0015,
"step": 7690
},
{
"epoch": 0.65,
"grad_norm": 0.09271080791950226,
"learning_rate": 6.769224311618537e-06,
"loss": 0.0025,
"step": 7700
},
{
"epoch": 0.65,
"grad_norm": 0.004653418902307749,
"learning_rate": 6.765026863666891e-06,
"loss": 0.0052,
"step": 7710
},
{
"epoch": 0.65,
"grad_norm": 0.0033459344413131475,
"learning_rate": 6.760829415715245e-06,
"loss": 0.0019,
"step": 7720
},
{
"epoch": 0.65,
"grad_norm": 0.24762499332427979,
"learning_rate": 6.7566319677636005e-06,
"loss": 0.0005,
"step": 7730
},
{
"epoch": 0.65,
"grad_norm": 0.2468775510787964,
"learning_rate": 6.752434519811954e-06,
"loss": 0.0014,
"step": 7740
},
{
"epoch": 0.65,
"grad_norm": 0.06739424914121628,
"learning_rate": 6.74823707186031e-06,
"loss": 0.0004,
"step": 7750
},
{
"epoch": 0.65,
"grad_norm": 0.01460706815123558,
"learning_rate": 6.7440396239086635e-06,
"loss": 0.0026,
"step": 7760
},
{
"epoch": 0.65,
"grad_norm": 0.0012919761938974261,
"learning_rate": 6.739842175957019e-06,
"loss": 0.0027,
"step": 7770
},
{
"epoch": 0.65,
"grad_norm": 0.16211023926734924,
"learning_rate": 6.7356447280053736e-06,
"loss": 0.0028,
"step": 7780
},
{
"epoch": 0.65,
"grad_norm": 0.9419199228286743,
"learning_rate": 6.731447280053728e-06,
"loss": 0.0155,
"step": 7790
},
{
"epoch": 0.65,
"grad_norm": 0.04390889033675194,
"learning_rate": 6.727249832102083e-06,
"loss": 0.004,
"step": 7800
},
{
"epoch": 0.66,
"grad_norm": 0.058596525341272354,
"learning_rate": 6.7230523841504366e-06,
"loss": 0.0041,
"step": 7810
},
{
"epoch": 0.66,
"grad_norm": 0.67715984582901,
"learning_rate": 6.718854936198792e-06,
"loss": 0.0014,
"step": 7820
},
{
"epoch": 0.66,
"grad_norm": 0.015570656396448612,
"learning_rate": 6.714657488247146e-06,
"loss": 0.0004,
"step": 7830
},
{
"epoch": 0.66,
"grad_norm": 0.07929646968841553,
"learning_rate": 6.710460040295501e-06,
"loss": 0.0036,
"step": 7840
},
{
"epoch": 0.66,
"grad_norm": 0.008636005222797394,
"learning_rate": 6.706262592343855e-06,
"loss": 0.0037,
"step": 7850
},
{
"epoch": 0.66,
"grad_norm": 0.1776953488588333,
"learning_rate": 6.7020651443922105e-06,
"loss": 0.0014,
"step": 7860
},
{
"epoch": 0.66,
"grad_norm": 0.03655342012643814,
"learning_rate": 6.697867696440564e-06,
"loss": 0.0003,
"step": 7870
},
{
"epoch": 0.66,
"grad_norm": 0.0143968490883708,
"learning_rate": 6.69367024848892e-06,
"loss": 0.0017,
"step": 7880
},
{
"epoch": 0.66,
"grad_norm": 0.137611523270607,
"learning_rate": 6.689472800537274e-06,
"loss": 0.002,
"step": 7890
},
{
"epoch": 0.66,
"grad_norm": 0.45386821031570435,
"learning_rate": 6.685275352585628e-06,
"loss": 0.001,
"step": 7900
},
{
"epoch": 0.66,
"grad_norm": 0.010178464464843273,
"learning_rate": 6.681077904633984e-06,
"loss": 0.0016,
"step": 7910
},
{
"epoch": 0.66,
"grad_norm": 0.039013270288705826,
"learning_rate": 6.676880456682337e-06,
"loss": 0.0015,
"step": 7920
},
{
"epoch": 0.67,
"grad_norm": 0.01565985381603241,
"learning_rate": 6.672683008730693e-06,
"loss": 0.0006,
"step": 7930
},
{
"epoch": 0.67,
"grad_norm": 0.017331531271338463,
"learning_rate": 6.668485560779047e-06,
"loss": 0.0005,
"step": 7940
},
{
"epoch": 0.67,
"grad_norm": 0.27378222346305847,
"learning_rate": 6.664288112827402e-06,
"loss": 0.0013,
"step": 7950
},
{
"epoch": 0.67,
"grad_norm": 0.29315513372421265,
"learning_rate": 6.660090664875756e-06,
"loss": 0.001,
"step": 7960
},
{
"epoch": 0.67,
"grad_norm": 0.37698185443878174,
"learning_rate": 6.6558932169241104e-06,
"loss": 0.0047,
"step": 7970
},
{
"epoch": 0.67,
"grad_norm": 0.009589233435690403,
"learning_rate": 6.651695768972465e-06,
"loss": 0.0012,
"step": 7980
},
{
"epoch": 0.67,
"grad_norm": 0.03333524242043495,
"learning_rate": 6.64749832102082e-06,
"loss": 0.0003,
"step": 7990
},
{
"epoch": 0.67,
"grad_norm": 0.00575454905629158,
"learning_rate": 6.643300873069175e-06,
"loss": 0.0007,
"step": 8000
},
{
"epoch": 0.67,
"grad_norm": 0.0640854686498642,
"learning_rate": 6.639103425117529e-06,
"loss": 0.0014,
"step": 8010
},
{
"epoch": 0.67,
"grad_norm": 0.007772125769406557,
"learning_rate": 6.634905977165884e-06,
"loss": 0.0004,
"step": 8020
},
{
"epoch": 0.67,
"grad_norm": 0.18589778244495392,
"learning_rate": 6.630708529214238e-06,
"loss": 0.0012,
"step": 8030
},
{
"epoch": 0.67,
"grad_norm": 0.027769211679697037,
"learning_rate": 6.626511081262594e-06,
"loss": 0.0005,
"step": 8040
},
{
"epoch": 0.68,
"grad_norm": 0.0029272972606122494,
"learning_rate": 6.622313633310947e-06,
"loss": 0.0006,
"step": 8050
},
{
"epoch": 0.68,
"grad_norm": 0.009183773770928383,
"learning_rate": 6.618116185359302e-06,
"loss": 0.001,
"step": 8060
},
{
"epoch": 0.68,
"grad_norm": 0.004283113870769739,
"learning_rate": 6.613918737407657e-06,
"loss": 0.001,
"step": 8070
},
{
"epoch": 0.68,
"grad_norm": 0.007778434548527002,
"learning_rate": 6.609721289456011e-06,
"loss": 0.0022,
"step": 8080
},
{
"epoch": 0.68,
"grad_norm": 0.0018987306393682957,
"learning_rate": 6.605523841504365e-06,
"loss": 0.0027,
"step": 8090
},
{
"epoch": 0.68,
"grad_norm": 0.12906016409397125,
"learning_rate": 6.6013263935527205e-06,
"loss": 0.0009,
"step": 8100
},
{
"epoch": 0.68,
"grad_norm": 0.007297148462384939,
"learning_rate": 6.597128945601076e-06,
"loss": 0.0007,
"step": 8110
},
{
"epoch": 0.68,
"grad_norm": 0.33462247252464294,
"learning_rate": 6.59293149764943e-06,
"loss": 0.0043,
"step": 8120
},
{
"epoch": 0.68,
"grad_norm": 0.8214085698127747,
"learning_rate": 6.588734049697784e-06,
"loss": 0.0021,
"step": 8130
},
{
"epoch": 0.68,
"grad_norm": 0.024006789550185204,
"learning_rate": 6.584536601746139e-06,
"loss": 0.0008,
"step": 8140
},
{
"epoch": 0.68,
"grad_norm": 0.02031201310455799,
"learning_rate": 6.5803391537944936e-06,
"loss": 0.0009,
"step": 8150
},
{
"epoch": 0.69,
"grad_norm": 0.08601708710193634,
"learning_rate": 6.576141705842848e-06,
"loss": 0.0009,
"step": 8160
},
{
"epoch": 0.69,
"grad_norm": 0.022639337927103043,
"learning_rate": 6.571944257891203e-06,
"loss": 0.0004,
"step": 8170
},
{
"epoch": 0.69,
"grad_norm": 0.02436881698668003,
"learning_rate": 6.5677468099395566e-06,
"loss": 0.0003,
"step": 8180
},
{
"epoch": 0.69,
"grad_norm": 0.02970285154879093,
"learning_rate": 6.563549361987912e-06,
"loss": 0.0022,
"step": 8190
},
{
"epoch": 0.69,
"grad_norm": 0.1934923380613327,
"learning_rate": 6.559351914036266e-06,
"loss": 0.001,
"step": 8200
},
{
"epoch": 0.69,
"grad_norm": 0.019940687343478203,
"learning_rate": 6.555154466084621e-06,
"loss": 0.0013,
"step": 8210
},
{
"epoch": 0.69,
"grad_norm": 0.025434071198105812,
"learning_rate": 6.550957018132975e-06,
"loss": 0.0008,
"step": 8220
},
{
"epoch": 0.69,
"grad_norm": 0.0027662869542837143,
"learning_rate": 6.5467595701813305e-06,
"loss": 0.0007,
"step": 8230
},
{
"epoch": 0.69,
"grad_norm": 0.2719932496547699,
"learning_rate": 6.542562122229685e-06,
"loss": 0.0006,
"step": 8240
},
{
"epoch": 0.69,
"grad_norm": 0.07692870497703552,
"learning_rate": 6.53836467427804e-06,
"loss": 0.0009,
"step": 8250
},
{
"epoch": 0.69,
"grad_norm": 0.0984962210059166,
"learning_rate": 6.534167226326394e-06,
"loss": 0.0021,
"step": 8260
},
{
"epoch": 0.69,
"grad_norm": 0.2010423243045807,
"learning_rate": 6.529969778374748e-06,
"loss": 0.0008,
"step": 8270
},
{
"epoch": 0.7,
"grad_norm": 0.047768231481313705,
"learning_rate": 6.525772330423104e-06,
"loss": 0.0006,
"step": 8280
},
{
"epoch": 0.7,
"grad_norm": 0.002377886790782213,
"learning_rate": 6.521574882471457e-06,
"loss": 0.0009,
"step": 8290
},
{
"epoch": 0.7,
"grad_norm": 0.04574467986822128,
"learning_rate": 6.517377434519813e-06,
"loss": 0.0004,
"step": 8300
},
{
"epoch": 0.7,
"grad_norm": 0.002358232159167528,
"learning_rate": 6.513179986568167e-06,
"loss": 0.0017,
"step": 8310
},
{
"epoch": 0.7,
"grad_norm": 0.04641246423125267,
"learning_rate": 6.508982538616522e-06,
"loss": 0.0017,
"step": 8320
},
{
"epoch": 0.7,
"grad_norm": 0.013856269419193268,
"learning_rate": 6.504785090664876e-06,
"loss": 0.0034,
"step": 8330
},
{
"epoch": 0.7,
"grad_norm": 0.08428198844194412,
"learning_rate": 6.5005876427132305e-06,
"loss": 0.0004,
"step": 8340
},
{
"epoch": 0.7,
"grad_norm": 0.07378539443016052,
"learning_rate": 6.496390194761586e-06,
"loss": 0.0007,
"step": 8350
},
{
"epoch": 0.7,
"grad_norm": 0.015434389002621174,
"learning_rate": 6.49219274680994e-06,
"loss": 0.0021,
"step": 8360
},
{
"epoch": 0.7,
"grad_norm": 0.014926637522876263,
"learning_rate": 6.487995298858295e-06,
"loss": 0.0011,
"step": 8370
},
{
"epoch": 0.7,
"grad_norm": 0.23808802664279938,
"learning_rate": 6.483797850906649e-06,
"loss": 0.0003,
"step": 8380
},
{
"epoch": 0.7,
"grad_norm": 0.013105678372085094,
"learning_rate": 6.479600402955004e-06,
"loss": 0.0008,
"step": 8390
},
{
"epoch": 0.71,
"grad_norm": 0.01477357279509306,
"learning_rate": 6.475402955003358e-06,
"loss": 0.0012,
"step": 8400
},
{
"epoch": 0.71,
"grad_norm": 0.1917484700679779,
"learning_rate": 6.471205507051714e-06,
"loss": 0.0009,
"step": 8410
},
{
"epoch": 0.71,
"grad_norm": 0.005253896117210388,
"learning_rate": 6.467008059100067e-06,
"loss": 0.0003,
"step": 8420
},
{
"epoch": 0.71,
"grad_norm": 0.028020091354846954,
"learning_rate": 6.462810611148422e-06,
"loss": 0.0009,
"step": 8430
},
{
"epoch": 0.71,
"grad_norm": 0.07543497532606125,
"learning_rate": 6.458613163196777e-06,
"loss": 0.0003,
"step": 8440
},
{
"epoch": 0.71,
"grad_norm": 0.21784624457359314,
"learning_rate": 6.454415715245131e-06,
"loss": 0.0007,
"step": 8450
},
{
"epoch": 0.71,
"grad_norm": 0.004629680421203375,
"learning_rate": 6.450218267293487e-06,
"loss": 0.0005,
"step": 8460
},
{
"epoch": 0.71,
"grad_norm": 0.05804940313100815,
"learning_rate": 6.4460208193418405e-06,
"loss": 0.0004,
"step": 8470
},
{
"epoch": 0.71,
"grad_norm": 0.008218007162213326,
"learning_rate": 6.441823371390196e-06,
"loss": 0.0024,
"step": 8480
},
{
"epoch": 0.71,
"grad_norm": 0.26576822996139526,
"learning_rate": 6.43762592343855e-06,
"loss": 0.0004,
"step": 8490
},
{
"epoch": 0.71,
"grad_norm": 0.05083778500556946,
"learning_rate": 6.433428475486904e-06,
"loss": 0.0018,
"step": 8500
},
{
"epoch": 0.71,
"grad_norm": 0.03845244273543358,
"learning_rate": 6.429231027535259e-06,
"loss": 0.0006,
"step": 8510
},
{
"epoch": 0.72,
"grad_norm": 0.22800813615322113,
"learning_rate": 6.425033579583614e-06,
"loss": 0.0011,
"step": 8520
},
{
"epoch": 0.72,
"grad_norm": 0.07355496287345886,
"learning_rate": 6.420836131631968e-06,
"loss": 0.0018,
"step": 8530
},
{
"epoch": 0.72,
"grad_norm": 0.14929671585559845,
"learning_rate": 6.416638683680323e-06,
"loss": 0.001,
"step": 8540
},
{
"epoch": 0.72,
"grad_norm": 0.009975256398320198,
"learning_rate": 6.412441235728677e-06,
"loss": 0.0004,
"step": 8550
},
{
"epoch": 0.72,
"grad_norm": 0.049514010548591614,
"learning_rate": 6.408663532572197e-06,
"loss": 0.0036,
"step": 8560
},
{
"epoch": 0.72,
"grad_norm": 0.17448657751083374,
"learning_rate": 6.404466084620551e-06,
"loss": 0.0029,
"step": 8570
},
{
"epoch": 0.72,
"grad_norm": 0.0019642149563878775,
"learning_rate": 6.400268636668906e-06,
"loss": 0.0007,
"step": 8580
},
{
"epoch": 0.72,
"grad_norm": 0.12057841569185257,
"learning_rate": 6.39607118871726e-06,
"loss": 0.0004,
"step": 8590
},
{
"epoch": 0.72,
"grad_norm": 0.08199835568666458,
"learning_rate": 6.391873740765615e-06,
"loss": 0.003,
"step": 8600
},
{
"epoch": 0.72,
"grad_norm": 0.008217026479542255,
"learning_rate": 6.3876762928139705e-06,
"loss": 0.0031,
"step": 8610
},
{
"epoch": 0.72,
"grad_norm": 0.0021252892911434174,
"learning_rate": 6.383478844862324e-06,
"loss": 0.0003,
"step": 8620
},
{
"epoch": 0.72,
"grad_norm": 0.5874733924865723,
"learning_rate": 6.379281396910679e-06,
"loss": 0.0028,
"step": 8630
},
{
"epoch": 0.73,
"grad_norm": 0.01447449903935194,
"learning_rate": 6.3750839489590335e-06,
"loss": 0.0016,
"step": 8640
},
{
"epoch": 0.73,
"grad_norm": 0.020882418379187584,
"learning_rate": 6.370886501007388e-06,
"loss": 0.0008,
"step": 8650
},
{
"epoch": 0.73,
"grad_norm": 0.03403160721063614,
"learning_rate": 6.366689053055743e-06,
"loss": 0.0008,
"step": 8660
},
{
"epoch": 0.73,
"grad_norm": 0.0914774239063263,
"learning_rate": 6.362491605104097e-06,
"loss": 0.0024,
"step": 8670
},
{
"epoch": 0.73,
"grad_norm": 0.02440485917031765,
"learning_rate": 6.358294157152451e-06,
"loss": 0.0008,
"step": 8680
},
{
"epoch": 0.73,
"grad_norm": 0.02295794151723385,
"learning_rate": 6.354096709200807e-06,
"loss": 0.0006,
"step": 8690
},
{
"epoch": 0.73,
"grad_norm": 0.005137599539011717,
"learning_rate": 6.34989926124916e-06,
"loss": 0.0007,
"step": 8700
},
{
"epoch": 0.73,
"grad_norm": 0.009145709685981274,
"learning_rate": 6.345701813297516e-06,
"loss": 0.004,
"step": 8710
},
{
"epoch": 0.73,
"grad_norm": 0.1550685167312622,
"learning_rate": 6.3415043653458705e-06,
"loss": 0.0009,
"step": 8720
},
{
"epoch": 0.73,
"grad_norm": 0.04273473471403122,
"learning_rate": 6.337306917394225e-06,
"loss": 0.0009,
"step": 8730
},
{
"epoch": 0.73,
"grad_norm": 0.01607104018330574,
"learning_rate": 6.33310946944258e-06,
"loss": 0.0018,
"step": 8740
},
{
"epoch": 0.73,
"grad_norm": 0.006609264761209488,
"learning_rate": 6.3289120214909335e-06,
"loss": 0.001,
"step": 8750
},
{
"epoch": 0.74,
"grad_norm": 0.6425864100456238,
"learning_rate": 6.324714573539289e-06,
"loss": 0.0061,
"step": 8760
},
{
"epoch": 0.74,
"grad_norm": 0.009361191652715206,
"learning_rate": 6.320517125587643e-06,
"loss": 0.0039,
"step": 8770
},
{
"epoch": 0.74,
"grad_norm": 0.06796202808618546,
"learning_rate": 6.316319677635998e-06,
"loss": 0.0021,
"step": 8780
},
{
"epoch": 0.74,
"grad_norm": 0.13239143788814545,
"learning_rate": 6.312122229684352e-06,
"loss": 0.0003,
"step": 8790
},
{
"epoch": 0.74,
"grad_norm": 0.4967783987522125,
"learning_rate": 6.307924781732707e-06,
"loss": 0.0011,
"step": 8800
},
{
"epoch": 0.74,
"grad_norm": 0.40928366780281067,
"learning_rate": 6.303727333781061e-06,
"loss": 0.0012,
"step": 8810
},
{
"epoch": 0.74,
"grad_norm": 0.06716176867485046,
"learning_rate": 6.299529885829417e-06,
"loss": 0.0005,
"step": 8820
},
{
"epoch": 0.74,
"grad_norm": 0.005962076131254435,
"learning_rate": 6.295332437877771e-06,
"loss": 0.0015,
"step": 8830
},
{
"epoch": 0.74,
"grad_norm": 0.03685719147324562,
"learning_rate": 6.291134989926125e-06,
"loss": 0.0004,
"step": 8840
},
{
"epoch": 0.74,
"grad_norm": 0.08257108926773071,
"learning_rate": 6.2869375419744805e-06,
"loss": 0.0004,
"step": 8850
},
{
"epoch": 0.74,
"grad_norm": 0.0008770768763497472,
"learning_rate": 6.282740094022834e-06,
"loss": 0.0003,
"step": 8860
},
{
"epoch": 0.74,
"grad_norm": 0.3550475537776947,
"learning_rate": 6.27854264607119e-06,
"loss": 0.0007,
"step": 8870
},
{
"epoch": 0.75,
"grad_norm": 0.007021765224635601,
"learning_rate": 6.2743451981195435e-06,
"loss": 0.0052,
"step": 8880
},
{
"epoch": 0.75,
"grad_norm": 0.012746850028634071,
"learning_rate": 6.270147750167899e-06,
"loss": 0.0005,
"step": 8890
},
{
"epoch": 0.75,
"grad_norm": 0.09132123738527298,
"learning_rate": 6.265950302216253e-06,
"loss": 0.0095,
"step": 8900
},
{
"epoch": 0.75,
"grad_norm": 0.004254049155861139,
"learning_rate": 6.261752854264607e-06,
"loss": 0.0008,
"step": 8910
},
{
"epoch": 0.75,
"grad_norm": 0.032813917845487595,
"learning_rate": 6.257555406312962e-06,
"loss": 0.001,
"step": 8920
},
{
"epoch": 0.75,
"grad_norm": 0.15052670240402222,
"learning_rate": 6.253357958361317e-06,
"loss": 0.001,
"step": 8930
},
{
"epoch": 0.75,
"grad_norm": 0.4991702139377594,
"learning_rate": 6.249160510409672e-06,
"loss": 0.0032,
"step": 8940
},
{
"epoch": 0.75,
"grad_norm": 0.0437382347881794,
"learning_rate": 6.244963062458026e-06,
"loss": 0.0008,
"step": 8950
},
{
"epoch": 0.75,
"grad_norm": 0.0036775097250938416,
"learning_rate": 6.240765614506381e-06,
"loss": 0.0003,
"step": 8960
},
{
"epoch": 0.75,
"grad_norm": 0.0415210947394371,
"learning_rate": 6.236568166554735e-06,
"loss": 0.0029,
"step": 8970
},
{
"epoch": 0.75,
"grad_norm": 0.005274785682559013,
"learning_rate": 6.2323707186030905e-06,
"loss": 0.0004,
"step": 8980
},
{
"epoch": 0.75,
"grad_norm": 0.001851436565630138,
"learning_rate": 6.228173270651444e-06,
"loss": 0.002,
"step": 8990
},
{
"epoch": 0.76,
"grad_norm": 0.027700498700141907,
"learning_rate": 6.223975822699799e-06,
"loss": 0.0022,
"step": 9000
},
{
"epoch": 0.76,
"grad_norm": 0.0017797650070860982,
"learning_rate": 6.2197783747481535e-06,
"loss": 0.0017,
"step": 9010
},
{
"epoch": 0.76,
"grad_norm": 0.2582283914089203,
"learning_rate": 6.215580926796508e-06,
"loss": 0.0009,
"step": 9020
},
{
"epoch": 0.76,
"grad_norm": 0.020876875147223473,
"learning_rate": 6.211383478844863e-06,
"loss": 0.0052,
"step": 9030
},
{
"epoch": 0.76,
"grad_norm": 0.07340060919523239,
"learning_rate": 6.207186030893217e-06,
"loss": 0.0018,
"step": 9040
},
{
"epoch": 0.76,
"grad_norm": 0.005077909212559462,
"learning_rate": 6.202988582941573e-06,
"loss": 0.0009,
"step": 9050
},
{
"epoch": 0.76,
"grad_norm": 0.0850280374288559,
"learning_rate": 6.198791134989927e-06,
"loss": 0.001,
"step": 9060
},
{
"epoch": 0.76,
"grad_norm": 0.22353577613830566,
"learning_rate": 6.194593687038281e-06,
"loss": 0.0012,
"step": 9070
},
{
"epoch": 0.76,
"grad_norm": 0.003676492255181074,
"learning_rate": 6.190396239086636e-06,
"loss": 0.0003,
"step": 9080
},
{
"epoch": 0.76,
"grad_norm": 0.1580864042043686,
"learning_rate": 6.1861987911349905e-06,
"loss": 0.0007,
"step": 9090
},
{
"epoch": 0.76,
"grad_norm": 0.1484569013118744,
"learning_rate": 6.182001343183345e-06,
"loss": 0.0004,
"step": 9100
},
{
"epoch": 0.76,
"grad_norm": 0.5588276982307434,
"learning_rate": 6.1778038952317e-06,
"loss": 0.0026,
"step": 9110
},
{
"epoch": 0.77,
"grad_norm": 0.36433857679367065,
"learning_rate": 6.1736064472800535e-06,
"loss": 0.0012,
"step": 9120
},
{
"epoch": 0.77,
"grad_norm": 0.009398641996085644,
"learning_rate": 6.169408999328409e-06,
"loss": 0.0055,
"step": 9130
},
{
"epoch": 0.77,
"grad_norm": 0.22657063603401184,
"learning_rate": 6.165211551376763e-06,
"loss": 0.0012,
"step": 9140
},
{
"epoch": 0.77,
"grad_norm": 0.2038237750530243,
"learning_rate": 6.161014103425118e-06,
"loss": 0.0011,
"step": 9150
},
{
"epoch": 0.77,
"grad_norm": 0.009869251400232315,
"learning_rate": 6.156816655473473e-06,
"loss": 0.001,
"step": 9160
},
{
"epoch": 0.77,
"grad_norm": 0.06920814514160156,
"learning_rate": 6.1526192075218274e-06,
"loss": 0.0009,
"step": 9170
},
{
"epoch": 0.77,
"grad_norm": 0.002625921508297324,
"learning_rate": 6.148421759570182e-06,
"loss": 0.0019,
"step": 9180
},
{
"epoch": 0.77,
"grad_norm": 0.00261191138997674,
"learning_rate": 6.144224311618537e-06,
"loss": 0.0005,
"step": 9190
},
{
"epoch": 0.77,
"grad_norm": 0.01784035935997963,
"learning_rate": 6.140026863666891e-06,
"loss": 0.0032,
"step": 9200
},
{
"epoch": 0.77,
"grad_norm": 0.013288385234773159,
"learning_rate": 6.135829415715245e-06,
"loss": 0.0011,
"step": 9210
},
{
"epoch": 0.77,
"grad_norm": 0.005483401007950306,
"learning_rate": 6.1316319677636005e-06,
"loss": 0.0006,
"step": 9220
},
{
"epoch": 0.77,
"grad_norm": 0.00357685424387455,
"learning_rate": 6.127434519811954e-06,
"loss": 0.0007,
"step": 9230
},
{
"epoch": 0.78,
"grad_norm": 0.025155210867524147,
"learning_rate": 6.12323707186031e-06,
"loss": 0.0093,
"step": 9240
},
{
"epoch": 0.78,
"grad_norm": 0.10848862677812576,
"learning_rate": 6.1190396239086635e-06,
"loss": 0.0009,
"step": 9250
},
{
"epoch": 0.78,
"grad_norm": 0.0211066585034132,
"learning_rate": 6.114842175957019e-06,
"loss": 0.0009,
"step": 9260
},
{
"epoch": 0.78,
"grad_norm": 0.4916308522224426,
"learning_rate": 6.110644728005374e-06,
"loss": 0.0042,
"step": 9270
},
{
"epoch": 0.78,
"grad_norm": 0.019028645008802414,
"learning_rate": 6.106447280053727e-06,
"loss": 0.0019,
"step": 9280
},
{
"epoch": 0.78,
"grad_norm": 0.143443763256073,
"learning_rate": 6.102249832102083e-06,
"loss": 0.0007,
"step": 9290
},
{
"epoch": 0.78,
"grad_norm": 0.017074372619390488,
"learning_rate": 6.098052384150437e-06,
"loss": 0.001,
"step": 9300
},
{
"epoch": 0.78,
"grad_norm": 0.0011280884500592947,
"learning_rate": 6.093854936198792e-06,
"loss": 0.0045,
"step": 9310
},
{
"epoch": 0.78,
"grad_norm": 0.015283357352018356,
"learning_rate": 6.089657488247146e-06,
"loss": 0.0007,
"step": 9320
},
{
"epoch": 0.78,
"grad_norm": 0.015474602580070496,
"learning_rate": 6.085460040295501e-06,
"loss": 0.0005,
"step": 9330
},
{
"epoch": 0.78,
"grad_norm": 0.0023845334071666002,
"learning_rate": 6.081262592343855e-06,
"loss": 0.0004,
"step": 9340
},
{
"epoch": 0.78,
"grad_norm": 0.40035849809646606,
"learning_rate": 6.0770651443922106e-06,
"loss": 0.002,
"step": 9350
},
{
"epoch": 0.79,
"grad_norm": 0.015063658356666565,
"learning_rate": 6.072867696440564e-06,
"loss": 0.0004,
"step": 9360
},
{
"epoch": 0.79,
"grad_norm": 0.10512324422597885,
"learning_rate": 6.068670248488919e-06,
"loss": 0.0054,
"step": 9370
},
{
"epoch": 0.79,
"grad_norm": 0.018911171704530716,
"learning_rate": 6.064472800537274e-06,
"loss": 0.0005,
"step": 9380
},
{
"epoch": 0.79,
"grad_norm": 0.01821274310350418,
"learning_rate": 6.060275352585628e-06,
"loss": 0.0019,
"step": 9390
},
{
"epoch": 0.79,
"grad_norm": 0.006468980107456446,
"learning_rate": 6.056077904633984e-06,
"loss": 0.0011,
"step": 9400
},
{
"epoch": 0.79,
"grad_norm": 0.0246189646422863,
"learning_rate": 6.051880456682337e-06,
"loss": 0.0003,
"step": 9410
},
{
"epoch": 0.79,
"grad_norm": 0.030287379398941994,
"learning_rate": 6.047683008730693e-06,
"loss": 0.0003,
"step": 9420
},
{
"epoch": 0.79,
"grad_norm": 0.14614807069301605,
"learning_rate": 6.043485560779047e-06,
"loss": 0.0007,
"step": 9430
},
{
"epoch": 0.79,
"grad_norm": 0.03680308535695076,
"learning_rate": 6.039288112827401e-06,
"loss": 0.0007,
"step": 9440
},
{
"epoch": 0.79,
"grad_norm": 0.020408429205417633,
"learning_rate": 6.035090664875756e-06,
"loss": 0.0042,
"step": 9450
},
{
"epoch": 0.79,
"grad_norm": 0.12720835208892822,
"learning_rate": 6.0308932169241105e-06,
"loss": 0.0008,
"step": 9460
},
{
"epoch": 0.79,
"grad_norm": 0.5571761727333069,
"learning_rate": 6.026695768972465e-06,
"loss": 0.0069,
"step": 9470
},
{
"epoch": 0.8,
"grad_norm": 0.287207692861557,
"learning_rate": 6.02249832102082e-06,
"loss": 0.0011,
"step": 9480
},
{
"epoch": 0.8,
"grad_norm": 0.19260409474372864,
"learning_rate": 6.018300873069175e-06,
"loss": 0.0012,
"step": 9490
},
{
"epoch": 0.8,
"grad_norm": 0.5683563351631165,
"learning_rate": 6.014103425117529e-06,
"loss": 0.0025,
"step": 9500
},
{
"epoch": 0.8,
"grad_norm": 0.0048590730875730515,
"learning_rate": 6.0099059771658844e-06,
"loss": 0.0016,
"step": 9510
},
{
"epoch": 0.8,
"grad_norm": 0.03889940306544304,
"learning_rate": 6.005708529214238e-06,
"loss": 0.0019,
"step": 9520
},
{
"epoch": 0.8,
"grad_norm": 0.09073323756456375,
"learning_rate": 6.001511081262593e-06,
"loss": 0.0041,
"step": 9530
},
{
"epoch": 0.8,
"grad_norm": 0.5227171182632446,
"learning_rate": 5.9973136333109474e-06,
"loss": 0.0047,
"step": 9540
},
{
"epoch": 0.8,
"grad_norm": 0.054277870804071426,
"learning_rate": 5.993116185359302e-06,
"loss": 0.0007,
"step": 9550
},
{
"epoch": 0.8,
"grad_norm": 0.0757797583937645,
"learning_rate": 5.988918737407657e-06,
"loss": 0.0008,
"step": 9560
},
{
"epoch": 0.8,
"grad_norm": 0.022691043093800545,
"learning_rate": 5.984721289456011e-06,
"loss": 0.0007,
"step": 9570
},
{
"epoch": 0.8,
"grad_norm": 0.010830945335328579,
"learning_rate": 5.980523841504365e-06,
"loss": 0.0004,
"step": 9580
},
{
"epoch": 0.81,
"grad_norm": 0.07802291959524155,
"learning_rate": 5.9763263935527205e-06,
"loss": 0.002,
"step": 9590
},
{
"epoch": 0.81,
"grad_norm": 0.03440363332629204,
"learning_rate": 5.972128945601076e-06,
"loss": 0.0009,
"step": 9600
},
{
"epoch": 0.81,
"grad_norm": 0.01543488446623087,
"learning_rate": 5.96793149764943e-06,
"loss": 0.0009,
"step": 9610
},
{
"epoch": 0.81,
"grad_norm": 0.014080232009291649,
"learning_rate": 5.963734049697784e-06,
"loss": 0.0008,
"step": 9620
},
{
"epoch": 0.81,
"grad_norm": 0.43045684695243835,
"learning_rate": 5.959536601746139e-06,
"loss": 0.0008,
"step": 9630
},
{
"epoch": 0.81,
"grad_norm": 0.006179173942655325,
"learning_rate": 5.955339153794494e-06,
"loss": 0.0011,
"step": 9640
},
{
"epoch": 0.81,
"grad_norm": 0.0018634877633303404,
"learning_rate": 5.951141705842847e-06,
"loss": 0.0002,
"step": 9650
},
{
"epoch": 0.81,
"grad_norm": 0.04732802137732506,
"learning_rate": 5.946944257891203e-06,
"loss": 0.0068,
"step": 9660
},
{
"epoch": 0.81,
"grad_norm": 0.028062958270311356,
"learning_rate": 5.942746809939557e-06,
"loss": 0.0021,
"step": 9670
},
{
"epoch": 0.81,
"grad_norm": 0.013063440099358559,
"learning_rate": 5.938549361987912e-06,
"loss": 0.0007,
"step": 9680
},
{
"epoch": 0.81,
"grad_norm": 0.008527678437530994,
"learning_rate": 5.934351914036266e-06,
"loss": 0.0003,
"step": 9690
},
{
"epoch": 0.81,
"grad_norm": 0.006061127409338951,
"learning_rate": 5.930154466084621e-06,
"loss": 0.0005,
"step": 9700
},
{
"epoch": 0.82,
"grad_norm": 0.05889621749520302,
"learning_rate": 5.925957018132975e-06,
"loss": 0.0008,
"step": 9710
},
{
"epoch": 0.82,
"grad_norm": 0.025425154715776443,
"learning_rate": 5.9217595701813306e-06,
"loss": 0.0009,
"step": 9720
},
{
"epoch": 0.82,
"grad_norm": 0.12469867616891861,
"learning_rate": 5.917562122229685e-06,
"loss": 0.0018,
"step": 9730
},
{
"epoch": 0.82,
"grad_norm": 0.0440966971218586,
"learning_rate": 5.913364674278039e-06,
"loss": 0.0014,
"step": 9740
},
{
"epoch": 0.82,
"grad_norm": 0.0034553324803709984,
"learning_rate": 5.909167226326394e-06,
"loss": 0.0014,
"step": 9750
},
{
"epoch": 0.82,
"grad_norm": 0.010510810650885105,
"learning_rate": 5.904969778374748e-06,
"loss": 0.0006,
"step": 9760
},
{
"epoch": 0.82,
"grad_norm": 0.01508377306163311,
"learning_rate": 5.900772330423104e-06,
"loss": 0.0026,
"step": 9770
},
{
"epoch": 0.82,
"grad_norm": 0.001469660666771233,
"learning_rate": 5.8965748824714574e-06,
"loss": 0.0012,
"step": 9780
},
{
"epoch": 0.82,
"grad_norm": 0.037131939083337784,
"learning_rate": 5.892377434519813e-06,
"loss": 0.0002,
"step": 9790
},
{
"epoch": 0.82,
"grad_norm": 0.00974042434245348,
"learning_rate": 5.888179986568167e-06,
"loss": 0.0003,
"step": 9800
},
{
"epoch": 0.82,
"grad_norm": 0.15985484421253204,
"learning_rate": 5.883982538616521e-06,
"loss": 0.0004,
"step": 9810
},
{
"epoch": 0.82,
"grad_norm": 0.17923088371753693,
"learning_rate": 5.879785090664876e-06,
"loss": 0.0004,
"step": 9820
},
{
"epoch": 0.83,
"grad_norm": 0.7428094148635864,
"learning_rate": 5.8755876427132305e-06,
"loss": 0.0025,
"step": 9830
},
{
"epoch": 0.83,
"grad_norm": 0.05475035682320595,
"learning_rate": 5.871390194761586e-06,
"loss": 0.0003,
"step": 9840
},
{
"epoch": 0.83,
"grad_norm": 0.0039265332743525505,
"learning_rate": 5.86719274680994e-06,
"loss": 0.0039,
"step": 9850
},
{
"epoch": 0.83,
"grad_norm": 0.02810598909854889,
"learning_rate": 5.862995298858295e-06,
"loss": 0.0017,
"step": 9860
},
{
"epoch": 0.83,
"grad_norm": 0.034790728241205215,
"learning_rate": 5.858797850906649e-06,
"loss": 0.0009,
"step": 9870
},
{
"epoch": 0.83,
"grad_norm": 0.17595210671424866,
"learning_rate": 5.8546004029550045e-06,
"loss": 0.0006,
"step": 9880
},
{
"epoch": 0.83,
"grad_norm": 0.12365200370550156,
"learning_rate": 5.850402955003358e-06,
"loss": 0.0027,
"step": 9890
},
{
"epoch": 0.83,
"grad_norm": 0.025172580033540726,
"learning_rate": 5.846205507051713e-06,
"loss": 0.0034,
"step": 9900
},
{
"epoch": 0.83,
"grad_norm": 0.03391372412443161,
"learning_rate": 5.8420080591000675e-06,
"loss": 0.0022,
"step": 9910
},
{
"epoch": 0.83,
"grad_norm": 0.0013367092469707131,
"learning_rate": 5.837810611148422e-06,
"loss": 0.0009,
"step": 9920
},
{
"epoch": 0.83,
"grad_norm": 0.2373090237379074,
"learning_rate": 5.833613163196777e-06,
"loss": 0.0013,
"step": 9930
},
{
"epoch": 0.83,
"grad_norm": 0.0014829274732619524,
"learning_rate": 5.829415715245131e-06,
"loss": 0.0004,
"step": 9940
},
{
"epoch": 0.84,
"grad_norm": 0.02379409596323967,
"learning_rate": 5.825218267293487e-06,
"loss": 0.0007,
"step": 9950
},
{
"epoch": 0.84,
"grad_norm": 0.01043054461479187,
"learning_rate": 5.8210208193418405e-06,
"loss": 0.0009,
"step": 9960
},
{
"epoch": 0.84,
"grad_norm": 0.5758562088012695,
"learning_rate": 5.816823371390196e-06,
"loss": 0.0076,
"step": 9970
},
{
"epoch": 0.84,
"grad_norm": 0.003396995598450303,
"learning_rate": 5.81262592343855e-06,
"loss": 0.0005,
"step": 9980
},
{
"epoch": 0.84,
"grad_norm": 0.004527990706264973,
"learning_rate": 5.808428475486904e-06,
"loss": 0.0003,
"step": 9990
},
{
"epoch": 0.84,
"grad_norm": 0.38935422897338867,
"learning_rate": 5.804231027535259e-06,
"loss": 0.0021,
"step": 10000
},
{
"epoch": 0.84,
"grad_norm": 0.007596536539494991,
"learning_rate": 5.800033579583614e-06,
"loss": 0.0026,
"step": 10010
},
{
"epoch": 0.84,
"grad_norm": 0.034029822796583176,
"learning_rate": 5.795836131631967e-06,
"loss": 0.0004,
"step": 10020
},
{
"epoch": 0.84,
"grad_norm": 0.8351893424987793,
"learning_rate": 5.791638683680323e-06,
"loss": 0.0024,
"step": 10030
},
{
"epoch": 0.84,
"grad_norm": 0.007966914214193821,
"learning_rate": 5.787441235728677e-06,
"loss": 0.0009,
"step": 10040
},
{
"epoch": 0.84,
"grad_norm": 0.039692457765340805,
"learning_rate": 5.783243787777032e-06,
"loss": 0.0011,
"step": 10050
},
{
"epoch": 0.84,
"grad_norm": 0.7721789479255676,
"learning_rate": 5.779046339825387e-06,
"loss": 0.0019,
"step": 10060
},
{
"epoch": 0.85,
"grad_norm": 0.01767701469361782,
"learning_rate": 5.774848891873741e-06,
"loss": 0.0008,
"step": 10070
},
{
"epoch": 0.85,
"grad_norm": 0.012562219053506851,
"learning_rate": 5.770651443922096e-06,
"loss": 0.0015,
"step": 10080
},
{
"epoch": 0.85,
"grad_norm": 0.34681907296180725,
"learning_rate": 5.766453995970451e-06,
"loss": 0.0045,
"step": 10090
},
{
"epoch": 0.85,
"grad_norm": 0.0029207984916865826,
"learning_rate": 5.762256548018805e-06,
"loss": 0.0017,
"step": 10100
},
{
"epoch": 0.85,
"grad_norm": 0.018310122191905975,
"learning_rate": 5.758059100067159e-06,
"loss": 0.0015,
"step": 10110
},
{
"epoch": 0.85,
"grad_norm": 0.06843991577625275,
"learning_rate": 5.7538616521155144e-06,
"loss": 0.0048,
"step": 10120
},
{
"epoch": 0.85,
"grad_norm": 0.03804745152592659,
"learning_rate": 5.749664204163868e-06,
"loss": 0.0025,
"step": 10130
},
{
"epoch": 0.85,
"grad_norm": 0.011987677775323391,
"learning_rate": 5.745466756212224e-06,
"loss": 0.0023,
"step": 10140
},
{
"epoch": 0.85,
"grad_norm": 0.0005486512091010809,
"learning_rate": 5.7412693082605774e-06,
"loss": 0.0008,
"step": 10150
},
{
"epoch": 0.85,
"grad_norm": 0.12931697070598602,
"learning_rate": 5.737071860308933e-06,
"loss": 0.0016,
"step": 10160
},
{
"epoch": 0.85,
"grad_norm": 0.014491924084722996,
"learning_rate": 5.7328744123572875e-06,
"loss": 0.0006,
"step": 10170
},
{
"epoch": 0.85,
"grad_norm": 0.008687620051205158,
"learning_rate": 5.728676964405641e-06,
"loss": 0.0007,
"step": 10180
},
{
"epoch": 0.86,
"grad_norm": 0.015420785173773766,
"learning_rate": 5.724479516453997e-06,
"loss": 0.0002,
"step": 10190
},
{
"epoch": 0.86,
"grad_norm": 0.0028673780616372824,
"learning_rate": 5.7202820685023505e-06,
"loss": 0.0005,
"step": 10200
},
{
"epoch": 0.86,
"grad_norm": 0.014829892665147781,
"learning_rate": 5.716084620550706e-06,
"loss": 0.0002,
"step": 10210
},
{
"epoch": 0.86,
"grad_norm": 0.051586270332336426,
"learning_rate": 5.71188717259906e-06,
"loss": 0.004,
"step": 10220
},
{
"epoch": 0.86,
"grad_norm": 0.0063542029820382595,
"learning_rate": 5.707689724647415e-06,
"loss": 0.0036,
"step": 10230
},
{
"epoch": 0.86,
"grad_norm": 0.014711436815559864,
"learning_rate": 5.703492276695769e-06,
"loss": 0.0021,
"step": 10240
},
{
"epoch": 0.86,
"grad_norm": 0.013807760551571846,
"learning_rate": 5.6992948287441245e-06,
"loss": 0.0004,
"step": 10250
},
{
"epoch": 0.86,
"grad_norm": 0.024736396968364716,
"learning_rate": 5.695097380792478e-06,
"loss": 0.0003,
"step": 10260
},
{
"epoch": 0.86,
"grad_norm": 0.6258419752120972,
"learning_rate": 5.690899932840833e-06,
"loss": 0.0026,
"step": 10270
},
{
"epoch": 0.86,
"grad_norm": 0.003710240125656128,
"learning_rate": 5.686702484889188e-06,
"loss": 0.0035,
"step": 10280
},
{
"epoch": 0.86,
"grad_norm": 0.653317391872406,
"learning_rate": 5.682505036937542e-06,
"loss": 0.0013,
"step": 10290
},
{
"epoch": 0.86,
"grad_norm": 0.02590692788362503,
"learning_rate": 5.6783075889858976e-06,
"loss": 0.0018,
"step": 10300
},
{
"epoch": 0.87,
"grad_norm": 0.006416289601475,
"learning_rate": 5.674110141034251e-06,
"loss": 0.0011,
"step": 10310
},
{
"epoch": 0.87,
"grad_norm": 0.02078012377023697,
"learning_rate": 5.669912693082607e-06,
"loss": 0.0004,
"step": 10320
},
{
"epoch": 0.87,
"grad_norm": 0.011740882880985737,
"learning_rate": 5.6657152451309606e-06,
"loss": 0.0013,
"step": 10330
},
{
"epoch": 0.87,
"grad_norm": 0.02781371772289276,
"learning_rate": 5.661517797179315e-06,
"loss": 0.0018,
"step": 10340
},
{
"epoch": 0.87,
"grad_norm": 0.011912044137716293,
"learning_rate": 5.65732034922767e-06,
"loss": 0.0008,
"step": 10350
},
{
"epoch": 0.87,
"grad_norm": 0.11006138473749161,
"learning_rate": 5.653122901276024e-06,
"loss": 0.0016,
"step": 10360
},
{
"epoch": 0.87,
"grad_norm": 0.03869205340743065,
"learning_rate": 5.648925453324379e-06,
"loss": 0.0008,
"step": 10370
},
{
"epoch": 0.87,
"grad_norm": 0.07503340393304825,
"learning_rate": 5.644728005372734e-06,
"loss": 0.0014,
"step": 10380
},
{
"epoch": 0.87,
"grad_norm": 0.001265190658159554,
"learning_rate": 5.640530557421089e-06,
"loss": 0.0003,
"step": 10390
},
{
"epoch": 0.87,
"grad_norm": 0.011775599792599678,
"learning_rate": 5.636333109469443e-06,
"loss": 0.0006,
"step": 10400
},
{
"epoch": 0.87,
"grad_norm": 0.11629336327314377,
"learning_rate": 5.632135661517798e-06,
"loss": 0.0006,
"step": 10410
},
{
"epoch": 0.87,
"grad_norm": 0.0015451526269316673,
"learning_rate": 5.627938213566152e-06,
"loss": 0.0038,
"step": 10420
},
{
"epoch": 0.88,
"grad_norm": 0.04855529963970184,
"learning_rate": 5.623740765614507e-06,
"loss": 0.0004,
"step": 10430
},
{
"epoch": 0.88,
"grad_norm": 0.01069630403071642,
"learning_rate": 5.619543317662861e-06,
"loss": 0.0006,
"step": 10440
},
{
"epoch": 0.88,
"grad_norm": 0.004674754571169615,
"learning_rate": 5.615345869711216e-06,
"loss": 0.0012,
"step": 10450
},
{
"epoch": 0.88,
"grad_norm": 0.025356121361255646,
"learning_rate": 5.611148421759571e-06,
"loss": 0.0006,
"step": 10460
},
{
"epoch": 0.88,
"grad_norm": 0.02232622541487217,
"learning_rate": 5.606950973807925e-06,
"loss": 0.0013,
"step": 10470
},
{
"epoch": 0.88,
"grad_norm": 0.497832328081131,
"learning_rate": 5.602753525856279e-06,
"loss": 0.0016,
"step": 10480
},
{
"epoch": 0.88,
"grad_norm": 0.057482216507196426,
"learning_rate": 5.5985560779046344e-06,
"loss": 0.0005,
"step": 10490
},
{
"epoch": 0.88,
"grad_norm": 0.003028752515092492,
"learning_rate": 5.59435862995299e-06,
"loss": 0.0004,
"step": 10500
},
{
"epoch": 0.88,
"grad_norm": 0.02643987536430359,
"learning_rate": 5.590161182001344e-06,
"loss": 0.0014,
"step": 10510
},
{
"epoch": 0.88,
"grad_norm": 0.02531200833618641,
"learning_rate": 5.585963734049698e-06,
"loss": 0.0005,
"step": 10520
},
{
"epoch": 0.88,
"grad_norm": 0.04974319785833359,
"learning_rate": 5.581766286098053e-06,
"loss": 0.0008,
"step": 10530
},
{
"epoch": 0.88,
"grad_norm": 0.005935164168477058,
"learning_rate": 5.5775688381464075e-06,
"loss": 0.0003,
"step": 10540
},
{
"epoch": 0.89,
"grad_norm": 0.00849095918238163,
"learning_rate": 5.573371390194761e-06,
"loss": 0.0003,
"step": 10550
},
{
"epoch": 0.89,
"grad_norm": 0.0163765549659729,
"learning_rate": 5.569173942243117e-06,
"loss": 0.0014,
"step": 10560
},
{
"epoch": 0.89,
"grad_norm": 0.003922377247363329,
"learning_rate": 5.5649764942914705e-06,
"loss": 0.0017,
"step": 10570
},
{
"epoch": 0.89,
"grad_norm": 0.008663144893944263,
"learning_rate": 5.560779046339826e-06,
"loss": 0.0005,
"step": 10580
},
{
"epoch": 0.89,
"grad_norm": 0.042090997099876404,
"learning_rate": 5.55658159838818e-06,
"loss": 0.0011,
"step": 10590
},
{
"epoch": 0.89,
"grad_norm": 0.018130453303456306,
"learning_rate": 5.552384150436535e-06,
"loss": 0.0002,
"step": 10600
},
{
"epoch": 0.89,
"grad_norm": 0.02145092561841011,
"learning_rate": 5.54818670248489e-06,
"loss": 0.0005,
"step": 10610
},
{
"epoch": 0.89,
"grad_norm": 0.016469236463308334,
"learning_rate": 5.5439892545332445e-06,
"loss": 0.0015,
"step": 10620
},
{
"epoch": 0.89,
"grad_norm": 0.07116610556840897,
"learning_rate": 5.539791806581599e-06,
"loss": 0.0003,
"step": 10630
},
{
"epoch": 0.89,
"grad_norm": 0.003410005010664463,
"learning_rate": 5.535594358629953e-06,
"loss": 0.0005,
"step": 10640
},
{
"epoch": 0.89,
"grad_norm": 0.0010806163772940636,
"learning_rate": 5.531396910678308e-06,
"loss": 0.0017,
"step": 10650
},
{
"epoch": 0.89,
"grad_norm": 0.0015185626689344645,
"learning_rate": 5.527199462726662e-06,
"loss": 0.0006,
"step": 10660
},
{
"epoch": 0.9,
"grad_norm": 0.016374943777918816,
"learning_rate": 5.5230020147750176e-06,
"loss": 0.0094,
"step": 10670
},
{
"epoch": 0.9,
"grad_norm": 0.00241836323402822,
"learning_rate": 5.518804566823371e-06,
"loss": 0.0004,
"step": 10680
},
{
"epoch": 0.9,
"grad_norm": 0.03313888981938362,
"learning_rate": 5.514607118871727e-06,
"loss": 0.0023,
"step": 10690
},
{
"epoch": 0.9,
"grad_norm": 0.1092633306980133,
"learning_rate": 5.5104096709200806e-06,
"loss": 0.0002,
"step": 10700
},
{
"epoch": 0.9,
"grad_norm": 0.03263653814792633,
"learning_rate": 5.506212222968435e-06,
"loss": 0.0069,
"step": 10710
},
{
"epoch": 0.9,
"grad_norm": 0.05272672697901726,
"learning_rate": 5.502014775016791e-06,
"loss": 0.0004,
"step": 10720
},
{
"epoch": 0.9,
"grad_norm": 0.007287841290235519,
"learning_rate": 5.4978173270651444e-06,
"loss": 0.0019,
"step": 10730
},
{
"epoch": 0.9,
"grad_norm": 0.006436245981603861,
"learning_rate": 5.4936198791135e-06,
"loss": 0.0042,
"step": 10740
},
{
"epoch": 0.9,
"grad_norm": 0.02231592871248722,
"learning_rate": 5.489422431161854e-06,
"loss": 0.0023,
"step": 10750
},
{
"epoch": 0.9,
"grad_norm": 0.013109694235026836,
"learning_rate": 5.485224983210209e-06,
"loss": 0.0044,
"step": 10760
},
{
"epoch": 0.9,
"grad_norm": 0.0014881811803206801,
"learning_rate": 5.481027535258563e-06,
"loss": 0.0074,
"step": 10770
},
{
"epoch": 0.9,
"grad_norm": 0.002896430203691125,
"learning_rate": 5.476830087306918e-06,
"loss": 0.0012,
"step": 10780
},
{
"epoch": 0.91,
"grad_norm": 0.009817220270633698,
"learning_rate": 5.472632639355272e-06,
"loss": 0.0028,
"step": 10790
},
{
"epoch": 0.91,
"grad_norm": 0.013071718625724316,
"learning_rate": 5.468435191403627e-06,
"loss": 0.0005,
"step": 10800
},
{
"epoch": 0.91,
"grad_norm": 0.14298699796199799,
"learning_rate": 5.464237743451981e-06,
"loss": 0.0003,
"step": 10810
},
{
"epoch": 0.91,
"grad_norm": 0.0034705400466918945,
"learning_rate": 5.460040295500336e-06,
"loss": 0.0019,
"step": 10820
},
{
"epoch": 0.91,
"grad_norm": 0.028280634433031082,
"learning_rate": 5.4558428475486915e-06,
"loss": 0.0009,
"step": 10830
},
{
"epoch": 0.91,
"grad_norm": 0.0049716150388121605,
"learning_rate": 5.451645399597045e-06,
"loss": 0.0003,
"step": 10840
},
{
"epoch": 0.91,
"grad_norm": 0.06866387277841568,
"learning_rate": 5.447447951645401e-06,
"loss": 0.0055,
"step": 10850
},
{
"epoch": 0.91,
"grad_norm": 0.001268894993700087,
"learning_rate": 5.4432505036937545e-06,
"loss": 0.0037,
"step": 10860
},
{
"epoch": 0.91,
"grad_norm": 0.4983358085155487,
"learning_rate": 5.43905305574211e-06,
"loss": 0.0123,
"step": 10870
},
{
"epoch": 0.91,
"grad_norm": 0.07618770748376846,
"learning_rate": 5.434855607790464e-06,
"loss": 0.0013,
"step": 10880
},
{
"epoch": 0.91,
"grad_norm": 0.054555345326662064,
"learning_rate": 5.430658159838818e-06,
"loss": 0.0014,
"step": 10890
},
{
"epoch": 0.92,
"grad_norm": 0.005016769748181105,
"learning_rate": 5.426460711887173e-06,
"loss": 0.0015,
"step": 10900
},
{
"epoch": 0.92,
"grad_norm": 0.02383585087954998,
"learning_rate": 5.4222632639355275e-06,
"loss": 0.0004,
"step": 10910
},
{
"epoch": 0.92,
"grad_norm": 0.032023441046476364,
"learning_rate": 5.418065815983881e-06,
"loss": 0.0009,
"step": 10920
},
{
"epoch": 0.92,
"grad_norm": 0.020678533241152763,
"learning_rate": 5.413868368032237e-06,
"loss": 0.0019,
"step": 10930
},
{
"epoch": 0.92,
"grad_norm": 0.04738430306315422,
"learning_rate": 5.409670920080592e-06,
"loss": 0.0008,
"step": 10940
},
{
"epoch": 0.92,
"grad_norm": 0.10293067991733551,
"learning_rate": 5.405473472128946e-06,
"loss": 0.001,
"step": 10950
},
{
"epoch": 0.92,
"grad_norm": 0.006902909372001886,
"learning_rate": 5.401276024177301e-06,
"loss": 0.0021,
"step": 10960
},
{
"epoch": 0.92,
"grad_norm": 0.09025087207555771,
"learning_rate": 5.397078576225655e-06,
"loss": 0.0015,
"step": 10970
},
{
"epoch": 0.92,
"grad_norm": 0.5670678615570068,
"learning_rate": 5.39288112827401e-06,
"loss": 0.0032,
"step": 10980
},
{
"epoch": 0.92,
"grad_norm": 0.08049952238798141,
"learning_rate": 5.3886836803223645e-06,
"loss": 0.0011,
"step": 10990
},
{
"epoch": 0.92,
"grad_norm": 0.009549125097692013,
"learning_rate": 5.384486232370719e-06,
"loss": 0.0005,
"step": 11000
},
{
"epoch": 0.92,
"grad_norm": 0.06387516856193542,
"learning_rate": 5.380288784419073e-06,
"loss": 0.0004,
"step": 11010
},
{
"epoch": 0.93,
"grad_norm": 0.00499248132109642,
"learning_rate": 5.376091336467428e-06,
"loss": 0.001,
"step": 11020
},
{
"epoch": 0.93,
"grad_norm": 0.04614701494574547,
"learning_rate": 5.371893888515782e-06,
"loss": 0.001,
"step": 11030
},
{
"epoch": 0.93,
"grad_norm": 0.2745211124420166,
"learning_rate": 5.367696440564138e-06,
"loss": 0.0042,
"step": 11040
},
{
"epoch": 0.93,
"grad_norm": 0.1052335873246193,
"learning_rate": 5.363498992612492e-06,
"loss": 0.0008,
"step": 11050
},
{
"epoch": 0.93,
"grad_norm": 0.025874366983771324,
"learning_rate": 5.359301544660847e-06,
"loss": 0.0022,
"step": 11060
},
{
"epoch": 0.93,
"grad_norm": 0.28177767992019653,
"learning_rate": 5.3551040967092014e-06,
"loss": 0.0015,
"step": 11070
},
{
"epoch": 0.93,
"grad_norm": 0.018476691097021103,
"learning_rate": 5.350906648757555e-06,
"loss": 0.0003,
"step": 11080
},
{
"epoch": 0.93,
"grad_norm": 0.015466156415641308,
"learning_rate": 5.346709200805911e-06,
"loss": 0.0009,
"step": 11090
},
{
"epoch": 0.93,
"grad_norm": 0.5496931076049805,
"learning_rate": 5.3425117528542644e-06,
"loss": 0.0071,
"step": 11100
},
{
"epoch": 0.93,
"grad_norm": 0.005472448188811541,
"learning_rate": 5.33831430490262e-06,
"loss": 0.0004,
"step": 11110
},
{
"epoch": 0.93,
"grad_norm": 0.02211419679224491,
"learning_rate": 5.334116856950974e-06,
"loss": 0.0007,
"step": 11120
},
{
"epoch": 0.93,
"grad_norm": 0.008153458125889301,
"learning_rate": 5.329919408999329e-06,
"loss": 0.0005,
"step": 11130
},
{
"epoch": 0.94,
"grad_norm": 0.04427899047732353,
"learning_rate": 5.325721961047683e-06,
"loss": 0.0006,
"step": 11140
},
{
"epoch": 0.94,
"grad_norm": 0.15745842456817627,
"learning_rate": 5.321524513096038e-06,
"loss": 0.0013,
"step": 11150
},
{
"epoch": 0.94,
"grad_norm": 0.10899525135755539,
"learning_rate": 5.317327065144393e-06,
"loss": 0.0025,
"step": 11160
},
{
"epoch": 0.94,
"grad_norm": 0.003296980634331703,
"learning_rate": 5.313129617192747e-06,
"loss": 0.0092,
"step": 11170
},
{
"epoch": 0.94,
"grad_norm": 0.00297130411490798,
"learning_rate": 5.308932169241102e-06,
"loss": 0.0005,
"step": 11180
},
{
"epoch": 0.94,
"grad_norm": 0.01586499996483326,
"learning_rate": 5.304734721289456e-06,
"loss": 0.0009,
"step": 11190
},
{
"epoch": 0.94,
"grad_norm": 0.03256227821111679,
"learning_rate": 5.3005372733378115e-06,
"loss": 0.0008,
"step": 11200
},
{
"epoch": 0.94,
"grad_norm": 0.015256262384355068,
"learning_rate": 5.296339825386165e-06,
"loss": 0.0018,
"step": 11210
},
{
"epoch": 0.94,
"grad_norm": 0.012666000984609127,
"learning_rate": 5.292142377434521e-06,
"loss": 0.0049,
"step": 11220
},
{
"epoch": 0.94,
"grad_norm": 0.009229323826730251,
"learning_rate": 5.2879449294828745e-06,
"loss": 0.0002,
"step": 11230
},
{
"epoch": 0.94,
"grad_norm": 0.07503992319107056,
"learning_rate": 5.28374748153123e-06,
"loss": 0.0006,
"step": 11240
},
{
"epoch": 0.94,
"grad_norm": 0.07024108618497849,
"learning_rate": 5.279550033579584e-06,
"loss": 0.0009,
"step": 11250
},
{
"epoch": 0.95,
"grad_norm": 0.0038027705159038305,
"learning_rate": 5.275352585627938e-06,
"loss": 0.0006,
"step": 11260
},
{
"epoch": 0.95,
"grad_norm": 0.006678453180938959,
"learning_rate": 5.271155137676294e-06,
"loss": 0.0008,
"step": 11270
},
{
"epoch": 0.95,
"grad_norm": 0.11047659069299698,
"learning_rate": 5.2669576897246476e-06,
"loss": 0.0011,
"step": 11280
},
{
"epoch": 0.95,
"grad_norm": 0.04119430482387543,
"learning_rate": 5.262760241773003e-06,
"loss": 0.0004,
"step": 11290
},
{
"epoch": 0.95,
"grad_norm": 0.0067368606105446815,
"learning_rate": 5.258562793821357e-06,
"loss": 0.0011,
"step": 11300
},
{
"epoch": 0.95,
"grad_norm": 0.08865031599998474,
"learning_rate": 5.254365345869712e-06,
"loss": 0.0028,
"step": 11310
},
{
"epoch": 0.95,
"grad_norm": 0.8145259618759155,
"learning_rate": 5.250167897918066e-06,
"loss": 0.0024,
"step": 11320
},
{
"epoch": 0.95,
"grad_norm": 0.10966548323631287,
"learning_rate": 5.245970449966421e-06,
"loss": 0.0013,
"step": 11330
},
{
"epoch": 0.95,
"grad_norm": 0.31640541553497314,
"learning_rate": 5.241773002014775e-06,
"loss": 0.0023,
"step": 11340
},
{
"epoch": 0.95,
"grad_norm": 0.01933100074529648,
"learning_rate": 5.23757555406313e-06,
"loss": 0.0006,
"step": 11350
},
{
"epoch": 0.95,
"grad_norm": 0.04006137698888779,
"learning_rate": 5.2333781061114845e-06,
"loss": 0.0004,
"step": 11360
},
{
"epoch": 0.95,
"grad_norm": 0.01764124073088169,
"learning_rate": 5.229180658159839e-06,
"loss": 0.0029,
"step": 11370
},
{
"epoch": 0.96,
"grad_norm": 0.02364155650138855,
"learning_rate": 5.224983210208195e-06,
"loss": 0.0004,
"step": 11380
},
{
"epoch": 0.96,
"grad_norm": 0.046450551599264145,
"learning_rate": 5.220785762256548e-06,
"loss": 0.0015,
"step": 11390
},
{
"epoch": 0.96,
"grad_norm": 0.802649736404419,
"learning_rate": 5.216588314304904e-06,
"loss": 0.0024,
"step": 11400
},
{
"epoch": 0.96,
"grad_norm": 0.0018641554052010179,
"learning_rate": 5.212390866353258e-06,
"loss": 0.0002,
"step": 11410
},
{
"epoch": 0.96,
"grad_norm": 0.027689019218087196,
"learning_rate": 5.208193418401612e-06,
"loss": 0.0014,
"step": 11420
},
{
"epoch": 0.96,
"grad_norm": 0.014141437597572803,
"learning_rate": 5.203995970449967e-06,
"loss": 0.0014,
"step": 11430
},
{
"epoch": 0.96,
"grad_norm": 0.04088863730430603,
"learning_rate": 5.1997985224983214e-06,
"loss": 0.0011,
"step": 11440
},
{
"epoch": 0.96,
"grad_norm": 0.26688361167907715,
"learning_rate": 5.195601074546675e-06,
"loss": 0.0016,
"step": 11450
},
{
"epoch": 0.96,
"grad_norm": 0.6439863443374634,
"learning_rate": 5.191403626595031e-06,
"loss": 0.0022,
"step": 11460
},
{
"epoch": 0.96,
"grad_norm": 0.009749379009008408,
"learning_rate": 5.1872061786433845e-06,
"loss": 0.0014,
"step": 11470
},
{
"epoch": 0.96,
"grad_norm": 0.04378553479909897,
"learning_rate": 5.18300873069174e-06,
"loss": 0.001,
"step": 11480
},
{
"epoch": 0.96,
"grad_norm": 0.01073415856808424,
"learning_rate": 5.1788112827400945e-06,
"loss": 0.001,
"step": 11490
},
{
"epoch": 0.97,
"grad_norm": 0.0023095139767974615,
"learning_rate": 5.174613834788449e-06,
"loss": 0.0004,
"step": 11500
},
{
"epoch": 0.97,
"grad_norm": 0.05889497324824333,
"learning_rate": 5.170416386836804e-06,
"loss": 0.0005,
"step": 11510
},
{
"epoch": 0.97,
"grad_norm": 0.018245402723550797,
"learning_rate": 5.166218938885158e-06,
"loss": 0.0006,
"step": 11520
},
{
"epoch": 0.97,
"grad_norm": 0.18495741486549377,
"learning_rate": 5.162021490933513e-06,
"loss": 0.0005,
"step": 11530
},
{
"epoch": 0.97,
"grad_norm": 0.002312073251232505,
"learning_rate": 5.157824042981867e-06,
"loss": 0.0005,
"step": 11540
},
{
"epoch": 0.97,
"grad_norm": 0.09305838495492935,
"learning_rate": 5.153626595030222e-06,
"loss": 0.0053,
"step": 11550
},
{
"epoch": 0.97,
"grad_norm": 0.11583968251943588,
"learning_rate": 5.149429147078576e-06,
"loss": 0.0057,
"step": 11560
},
{
"epoch": 0.97,
"grad_norm": 0.11358506232500076,
"learning_rate": 5.1452316991269315e-06,
"loss": 0.0004,
"step": 11570
},
{
"epoch": 0.97,
"grad_norm": 0.49590811133384705,
"learning_rate": 5.141034251175285e-06,
"loss": 0.0028,
"step": 11580
},
{
"epoch": 0.97,
"grad_norm": 0.016838036477565765,
"learning_rate": 5.136836803223641e-06,
"loss": 0.0007,
"step": 11590
},
{
"epoch": 0.97,
"grad_norm": 0.003618579125031829,
"learning_rate": 5.132639355271995e-06,
"loss": 0.0005,
"step": 11600
},
{
"epoch": 0.97,
"grad_norm": 0.00428624777123332,
"learning_rate": 5.128441907320349e-06,
"loss": 0.002,
"step": 11610
},
{
"epoch": 0.98,
"grad_norm": 0.08151665329933167,
"learning_rate": 5.1242444593687046e-06,
"loss": 0.0013,
"step": 11620
},
{
"epoch": 0.98,
"grad_norm": 0.006494326516985893,
"learning_rate": 5.120047011417058e-06,
"loss": 0.0011,
"step": 11630
},
{
"epoch": 0.98,
"grad_norm": 0.005522624123841524,
"learning_rate": 5.115849563465414e-06,
"loss": 0.0045,
"step": 11640
},
{
"epoch": 0.98,
"grad_norm": 0.0026271094102412462,
"learning_rate": 5.1116521155137676e-06,
"loss": 0.0023,
"step": 11650
},
{
"epoch": 0.98,
"grad_norm": 0.026925839483737946,
"learning_rate": 5.107454667562123e-06,
"loss": 0.0008,
"step": 11660
},
{
"epoch": 0.98,
"grad_norm": 0.019350741058588028,
"learning_rate": 5.103257219610477e-06,
"loss": 0.0006,
"step": 11670
},
{
"epoch": 0.98,
"grad_norm": 0.02016269974410534,
"learning_rate": 5.099059771658832e-06,
"loss": 0.0063,
"step": 11680
},
{
"epoch": 0.98,
"grad_norm": 0.07899802923202515,
"learning_rate": 5.094862323707186e-06,
"loss": 0.0009,
"step": 11690
},
{
"epoch": 0.98,
"grad_norm": 0.006061141844838858,
"learning_rate": 5.090664875755541e-06,
"loss": 0.0012,
"step": 11700
},
{
"epoch": 0.98,
"grad_norm": 0.0007106554112397134,
"learning_rate": 5.086467427803896e-06,
"loss": 0.0027,
"step": 11710
},
{
"epoch": 0.98,
"grad_norm": 0.08793941885232925,
"learning_rate": 5.08226997985225e-06,
"loss": 0.0011,
"step": 11720
},
{
"epoch": 0.98,
"grad_norm": 0.009847030974924564,
"learning_rate": 5.078072531900605e-06,
"loss": 0.0017,
"step": 11730
},
{
"epoch": 0.99,
"grad_norm": 0.1629151850938797,
"learning_rate": 5.073875083948959e-06,
"loss": 0.0007,
"step": 11740
},
{
"epoch": 0.99,
"grad_norm": 0.008842062205076218,
"learning_rate": 5.069677635997315e-06,
"loss": 0.0036,
"step": 11750
},
{
"epoch": 0.99,
"grad_norm": 0.0026031406596302986,
"learning_rate": 5.065480188045668e-06,
"loss": 0.0005,
"step": 11760
},
{
"epoch": 0.99,
"grad_norm": 0.00415264954790473,
"learning_rate": 5.061282740094024e-06,
"loss": 0.0013,
"step": 11770
},
{
"epoch": 0.99,
"grad_norm": 0.007113313302397728,
"learning_rate": 5.057085292142378e-06,
"loss": 0.0007,
"step": 11780
},
{
"epoch": 0.99,
"grad_norm": 0.024390025064349174,
"learning_rate": 5.052887844190732e-06,
"loss": 0.0012,
"step": 11790
},
{
"epoch": 0.99,
"grad_norm": 0.0008230148232541978,
"learning_rate": 5.048690396239087e-06,
"loss": 0.0004,
"step": 11800
},
{
"epoch": 0.99,
"grad_norm": 0.005063881631940603,
"learning_rate": 5.0444929482874415e-06,
"loss": 0.0004,
"step": 11810
},
{
"epoch": 0.99,
"grad_norm": 0.04501771554350853,
"learning_rate": 5.040295500335797e-06,
"loss": 0.0065,
"step": 11820
},
{
"epoch": 0.99,
"grad_norm": 0.007200128864496946,
"learning_rate": 5.036098052384151e-06,
"loss": 0.0054,
"step": 11830
},
{
"epoch": 0.99,
"grad_norm": 0.33267006278038025,
"learning_rate": 5.031900604432506e-06,
"loss": 0.0039,
"step": 11840
},
{
"epoch": 0.99,
"grad_norm": 0.21112149953842163,
"learning_rate": 5.02770315648086e-06,
"loss": 0.0006,
"step": 11850
},
{
"epoch": 1.0,
"grad_norm": 0.003047073259949684,
"learning_rate": 5.0235057085292145e-06,
"loss": 0.0003,
"step": 11860
},
{
"epoch": 1.0,
"grad_norm": 0.056417226791381836,
"learning_rate": 5.019308260577569e-06,
"loss": 0.0004,
"step": 11870
},
{
"epoch": 1.0,
"grad_norm": 0.0030665055382996798,
"learning_rate": 5.015110812625924e-06,
"loss": 0.0007,
"step": 11880
},
{
"epoch": 1.0,
"grad_norm": 0.021434858441352844,
"learning_rate": 5.010913364674278e-06,
"loss": 0.0013,
"step": 11890
},
{
"epoch": 1.0,
"grad_norm": 0.13436031341552734,
"learning_rate": 5.006715916722633e-06,
"loss": 0.0004,
"step": 11900
},
{
"epoch": 1.0,
"grad_norm": 0.012255342677235603,
"learning_rate": 5.002518468770987e-06,
"loss": 0.0025,
"step": 11910
},
{
"epoch": 1.0,
"grad_norm": 0.01314691361039877,
"learning_rate": 4.998321020819342e-06,
"loss": 0.0006,
"step": 11920
},
{
"epoch": 1.0,
"grad_norm": 0.17141364514827728,
"learning_rate": 4.994123572867697e-06,
"loss": 0.0015,
"step": 11930
},
{
"epoch": 1.0,
"grad_norm": 0.06440167874097824,
"learning_rate": 4.9899261249160515e-06,
"loss": 0.0004,
"step": 11940
},
{
"epoch": 1.0,
"grad_norm": 0.003272134577855468,
"learning_rate": 4.985728676964406e-06,
"loss": 0.0026,
"step": 11950
},
{
"epoch": 1.0,
"grad_norm": 0.10054084658622742,
"learning_rate": 4.981531229012761e-06,
"loss": 0.0003,
"step": 11960
},
{
"epoch": 1.0,
"grad_norm": 0.05302809551358223,
"learning_rate": 4.977333781061115e-06,
"loss": 0.0007,
"step": 11970
},
{
"epoch": 1.01,
"grad_norm": 0.023061759769916534,
"learning_rate": 4.97313633310947e-06,
"loss": 0.0026,
"step": 11980
},
{
"epoch": 1.01,
"grad_norm": 0.038504019379615784,
"learning_rate": 4.968938885157825e-06,
"loss": 0.0026,
"step": 11990
},
{
"epoch": 1.01,
"grad_norm": 0.0051516154780983925,
"learning_rate": 4.964741437206179e-06,
"loss": 0.0003,
"step": 12000
},
{
"epoch": 1.01,
"grad_norm": 0.05448487401008606,
"learning_rate": 4.960543989254534e-06,
"loss": 0.001,
"step": 12010
},
{
"epoch": 1.01,
"grad_norm": 0.048447802662849426,
"learning_rate": 4.9563465413028884e-06,
"loss": 0.0003,
"step": 12020
},
{
"epoch": 1.01,
"grad_norm": 0.14971762895584106,
"learning_rate": 4.952149093351243e-06,
"loss": 0.0004,
"step": 12030
},
{
"epoch": 1.01,
"grad_norm": 0.03111051581799984,
"learning_rate": 4.947951645399598e-06,
"loss": 0.0021,
"step": 12040
},
{
"epoch": 1.01,
"grad_norm": 0.035793572664260864,
"learning_rate": 4.943754197447952e-06,
"loss": 0.0002,
"step": 12050
},
{
"epoch": 1.01,
"grad_norm": 0.25612950325012207,
"learning_rate": 4.939556749496306e-06,
"loss": 0.0008,
"step": 12060
},
{
"epoch": 1.01,
"grad_norm": 0.015790291130542755,
"learning_rate": 4.935359301544661e-06,
"loss": 0.004,
"step": 12070
},
{
"epoch": 1.01,
"grad_norm": 0.00603965250775218,
"learning_rate": 4.931161853593015e-06,
"loss": 0.0005,
"step": 12080
},
{
"epoch": 1.01,
"grad_norm": 0.029880041256546974,
"learning_rate": 4.926964405641371e-06,
"loss": 0.0006,
"step": 12090
},
{
"epoch": 1.02,
"grad_norm": 0.14440520107746124,
"learning_rate": 4.922766957689725e-06,
"loss": 0.0005,
"step": 12100
},
{
"epoch": 1.02,
"grad_norm": 0.006907114293426275,
"learning_rate": 4.91856950973808e-06,
"loss": 0.0011,
"step": 12110
},
{
"epoch": 1.02,
"grad_norm": 0.08127205818891525,
"learning_rate": 4.914372061786435e-06,
"loss": 0.0013,
"step": 12120
},
{
"epoch": 1.02,
"grad_norm": 0.8747881650924683,
"learning_rate": 4.910174613834789e-06,
"loss": 0.0015,
"step": 12130
},
{
"epoch": 1.02,
"grad_norm": 0.0023702913895249367,
"learning_rate": 4.905977165883144e-06,
"loss": 0.0006,
"step": 12140
},
{
"epoch": 1.02,
"grad_norm": 0.0015660161152482033,
"learning_rate": 4.901779717931498e-06,
"loss": 0.0008,
"step": 12150
},
{
"epoch": 1.02,
"grad_norm": 0.016091827303171158,
"learning_rate": 4.897582269979852e-06,
"loss": 0.0006,
"step": 12160
},
{
"epoch": 1.02,
"grad_norm": 0.16924422979354858,
"learning_rate": 4.893384822028207e-06,
"loss": 0.0035,
"step": 12170
},
{
"epoch": 1.02,
"grad_norm": 0.0360410138964653,
"learning_rate": 4.8891873740765615e-06,
"loss": 0.0012,
"step": 12180
},
{
"epoch": 1.02,
"grad_norm": 0.2576335370540619,
"learning_rate": 4.884989926124916e-06,
"loss": 0.0007,
"step": 12190
},
{
"epoch": 1.02,
"grad_norm": 0.026088684797286987,
"learning_rate": 4.8807924781732716e-06,
"loss": 0.0004,
"step": 12200
},
{
"epoch": 1.03,
"grad_norm": 0.030685720965266228,
"learning_rate": 4.876595030221626e-06,
"loss": 0.0003,
"step": 12210
},
{
"epoch": 1.03,
"grad_norm": 0.043232034891843796,
"learning_rate": 4.872397582269981e-06,
"loss": 0.0004,
"step": 12220
},
{
"epoch": 1.03,
"grad_norm": 0.05359245836734772,
"learning_rate": 4.8682001343183346e-06,
"loss": 0.001,
"step": 12230
},
{
"epoch": 1.03,
"grad_norm": 0.0005476790247485042,
"learning_rate": 4.864002686366689e-06,
"loss": 0.0017,
"step": 12240
},
{
"epoch": 1.03,
"grad_norm": 0.005588450003415346,
"learning_rate": 4.859805238415044e-06,
"loss": 0.001,
"step": 12250
},
{
"epoch": 1.03,
"grad_norm": 0.07085375487804413,
"learning_rate": 4.855607790463398e-06,
"loss": 0.0011,
"step": 12260
},
{
"epoch": 1.03,
"grad_norm": 0.012516853399574757,
"learning_rate": 4.851410342511753e-06,
"loss": 0.0025,
"step": 12270
},
{
"epoch": 1.03,
"grad_norm": 0.010629468597471714,
"learning_rate": 4.847212894560108e-06,
"loss": 0.0007,
"step": 12280
},
{
"epoch": 1.03,
"grad_norm": 0.07004597038030624,
"learning_rate": 4.843015446608462e-06,
"loss": 0.0004,
"step": 12290
},
{
"epoch": 1.03,
"grad_norm": 0.009928259067237377,
"learning_rate": 4.838817998656817e-06,
"loss": 0.0055,
"step": 12300
},
{
"epoch": 1.03,
"grad_norm": 0.05139904469251633,
"learning_rate": 4.8346205507051715e-06,
"loss": 0.0008,
"step": 12310
},
{
"epoch": 1.03,
"grad_norm": 0.00505171250551939,
"learning_rate": 4.830423102753526e-06,
"loss": 0.0019,
"step": 12320
},
{
"epoch": 1.04,
"grad_norm": 0.22344523668289185,
"learning_rate": 4.826225654801881e-06,
"loss": 0.0101,
"step": 12330
},
{
"epoch": 1.04,
"grad_norm": 0.44625988602638245,
"learning_rate": 4.822028206850235e-06,
"loss": 0.002,
"step": 12340
},
{
"epoch": 1.04,
"grad_norm": 0.003404056653380394,
"learning_rate": 4.81783075889859e-06,
"loss": 0.0003,
"step": 12350
},
{
"epoch": 1.04,
"grad_norm": 0.05186540633440018,
"learning_rate": 4.813633310946945e-06,
"loss": 0.0011,
"step": 12360
},
{
"epoch": 1.04,
"grad_norm": 0.012058926746249199,
"learning_rate": 4.809435862995299e-06,
"loss": 0.0003,
"step": 12370
},
{
"epoch": 1.04,
"grad_norm": 0.17031264305114746,
"learning_rate": 4.805238415043654e-06,
"loss": 0.0003,
"step": 12380
},
{
"epoch": 1.04,
"grad_norm": 0.0036926474422216415,
"learning_rate": 4.8010409670920084e-06,
"loss": 0.0004,
"step": 12390
},
{
"epoch": 1.04,
"grad_norm": 0.0020336448214948177,
"learning_rate": 4.796843519140363e-06,
"loss": 0.0049,
"step": 12400
},
{
"epoch": 1.04,
"grad_norm": 0.009088937193155289,
"learning_rate": 4.792646071188718e-06,
"loss": 0.0005,
"step": 12410
},
{
"epoch": 1.04,
"grad_norm": 0.012697056867182255,
"learning_rate": 4.788448623237072e-06,
"loss": 0.0016,
"step": 12420
},
{
"epoch": 1.04,
"grad_norm": 0.05333826318383217,
"learning_rate": 4.784251175285427e-06,
"loss": 0.0014,
"step": 12430
},
{
"epoch": 1.04,
"grad_norm": 0.0574137419462204,
"learning_rate": 4.7800537273337815e-06,
"loss": 0.0007,
"step": 12440
},
{
"epoch": 1.05,
"grad_norm": 0.08394545316696167,
"learning_rate": 4.775856279382136e-06,
"loss": 0.0035,
"step": 12450
},
{
"epoch": 1.05,
"grad_norm": 0.014619513414800167,
"learning_rate": 4.771658831430491e-06,
"loss": 0.0006,
"step": 12460
},
{
"epoch": 1.05,
"grad_norm": 0.010346817784011364,
"learning_rate": 4.767461383478845e-06,
"loss": 0.0004,
"step": 12470
},
{
"epoch": 1.05,
"grad_norm": 0.010911854915320873,
"learning_rate": 4.7632639355272e-06,
"loss": 0.0007,
"step": 12480
},
{
"epoch": 1.05,
"grad_norm": 0.007937319576740265,
"learning_rate": 4.759066487575555e-06,
"loss": 0.0017,
"step": 12490
},
{
"epoch": 1.05,
"grad_norm": 0.004206392448395491,
"learning_rate": 4.754869039623909e-06,
"loss": 0.0004,
"step": 12500
},
{
"epoch": 1.05,
"grad_norm": 0.004093706142157316,
"learning_rate": 4.750671591672264e-06,
"loss": 0.0002,
"step": 12510
},
{
"epoch": 1.05,
"grad_norm": 0.015210299752652645,
"learning_rate": 4.746474143720618e-06,
"loss": 0.0019,
"step": 12520
},
{
"epoch": 1.05,
"grad_norm": 0.013090868480503559,
"learning_rate": 4.742276695768973e-06,
"loss": 0.0004,
"step": 12530
},
{
"epoch": 1.05,
"grad_norm": 0.006157224997878075,
"learning_rate": 4.738079247817328e-06,
"loss": 0.0003,
"step": 12540
},
{
"epoch": 1.05,
"grad_norm": 0.027267277240753174,
"learning_rate": 4.733881799865682e-06,
"loss": 0.0053,
"step": 12550
},
{
"epoch": 1.05,
"grad_norm": 0.28201332688331604,
"learning_rate": 4.729684351914037e-06,
"loss": 0.0005,
"step": 12560
},
{
"epoch": 1.06,
"grad_norm": 0.015838222578167915,
"learning_rate": 4.7254869039623916e-06,
"loss": 0.0023,
"step": 12570
},
{
"epoch": 1.06,
"grad_norm": 0.3681650459766388,
"learning_rate": 4.721289456010746e-06,
"loss": 0.0014,
"step": 12580
},
{
"epoch": 1.06,
"grad_norm": 0.0024659759365022182,
"learning_rate": 4.717092008059101e-06,
"loss": 0.0038,
"step": 12590
},
{
"epoch": 1.06,
"grad_norm": 0.024921996518969536,
"learning_rate": 4.7128945601074546e-06,
"loss": 0.0026,
"step": 12600
},
{
"epoch": 1.06,
"grad_norm": 0.018606165423989296,
"learning_rate": 4.708697112155809e-06,
"loss": 0.0006,
"step": 12610
},
{
"epoch": 1.06,
"grad_norm": 0.0071179852820932865,
"learning_rate": 4.704499664204164e-06,
"loss": 0.0002,
"step": 12620
},
{
"epoch": 1.06,
"grad_norm": 0.0048920102417469025,
"learning_rate": 4.7003022162525184e-06,
"loss": 0.001,
"step": 12630
},
{
"epoch": 1.06,
"grad_norm": 0.08117188513278961,
"learning_rate": 4.696104768300874e-06,
"loss": 0.0009,
"step": 12640
},
{
"epoch": 1.06,
"grad_norm": 0.002124218037351966,
"learning_rate": 4.6919073203492285e-06,
"loss": 0.0008,
"step": 12650
},
{
"epoch": 1.06,
"grad_norm": 0.013521471992135048,
"learning_rate": 4.687709872397583e-06,
"loss": 0.0026,
"step": 12660
},
{
"epoch": 1.06,
"grad_norm": 0.05654778704047203,
"learning_rate": 4.683512424445938e-06,
"loss": 0.0005,
"step": 12670
},
{
"epoch": 1.06,
"grad_norm": 0.007984794676303864,
"learning_rate": 4.6793149764942915e-06,
"loss": 0.0003,
"step": 12680
},
{
"epoch": 1.07,
"grad_norm": 0.0030824383720755577,
"learning_rate": 4.675117528542646e-06,
"loss": 0.0008,
"step": 12690
},
{
"epoch": 1.07,
"grad_norm": 0.0020946026779711246,
"learning_rate": 4.670920080591001e-06,
"loss": 0.0003,
"step": 12700
},
{
"epoch": 1.07,
"grad_norm": 0.03582073748111725,
"learning_rate": 4.666722632639355e-06,
"loss": 0.0014,
"step": 12710
},
{
"epoch": 1.07,
"grad_norm": 0.02298971451818943,
"learning_rate": 4.66252518468771e-06,
"loss": 0.0021,
"step": 12720
},
{
"epoch": 1.07,
"grad_norm": 0.028649525716900826,
"learning_rate": 4.658327736736065e-06,
"loss": 0.0002,
"step": 12730
},
{
"epoch": 1.07,
"grad_norm": 0.0038425384555011988,
"learning_rate": 4.654130288784419e-06,
"loss": 0.0009,
"step": 12740
},
{
"epoch": 1.07,
"grad_norm": 0.0029490068554878235,
"learning_rate": 4.649932840832775e-06,
"loss": 0.002,
"step": 12750
},
{
"epoch": 1.07,
"grad_norm": 0.022061439231038094,
"learning_rate": 4.6457353928811285e-06,
"loss": 0.0003,
"step": 12760
},
{
"epoch": 1.07,
"grad_norm": 0.001973432954400778,
"learning_rate": 4.641537944929483e-06,
"loss": 0.0011,
"step": 12770
},
{
"epoch": 1.07,
"grad_norm": 0.015841417014598846,
"learning_rate": 4.637340496977838e-06,
"loss": 0.0004,
"step": 12780
},
{
"epoch": 1.07,
"grad_norm": 0.005069794598966837,
"learning_rate": 4.633143049026192e-06,
"loss": 0.0002,
"step": 12790
},
{
"epoch": 1.07,
"grad_norm": 0.001450516632758081,
"learning_rate": 4.628945601074547e-06,
"loss": 0.0032,
"step": 12800
},
{
"epoch": 1.08,
"grad_norm": 0.01164446771144867,
"learning_rate": 4.6247481531229015e-06,
"loss": 0.0012,
"step": 12810
},
{
"epoch": 1.08,
"grad_norm": 0.027409467846155167,
"learning_rate": 4.620550705171256e-06,
"loss": 0.001,
"step": 12820
},
{
"epoch": 1.08,
"grad_norm": 0.020864328369498253,
"learning_rate": 4.616353257219611e-06,
"loss": 0.0012,
"step": 12830
},
{
"epoch": 1.08,
"grad_norm": 0.0067404527217149734,
"learning_rate": 4.612155809267965e-06,
"loss": 0.0003,
"step": 12840
},
{
"epoch": 1.08,
"grad_norm": 0.0015818218234926462,
"learning_rate": 4.60795836131632e-06,
"loss": 0.0013,
"step": 12850
},
{
"epoch": 1.08,
"grad_norm": 0.010182135738432407,
"learning_rate": 4.603760913364675e-06,
"loss": 0.0007,
"step": 12860
},
{
"epoch": 1.08,
"grad_norm": 0.025032339617609978,
"learning_rate": 4.599563465413029e-06,
"loss": 0.0005,
"step": 12870
},
{
"epoch": 1.08,
"grad_norm": 0.10075455904006958,
"learning_rate": 4.595785762256548e-06,
"loss": 0.0025,
"step": 12880
},
{
"epoch": 1.08,
"grad_norm": 0.03304919973015785,
"learning_rate": 4.591588314304903e-06,
"loss": 0.0021,
"step": 12890
},
{
"epoch": 1.08,
"grad_norm": 0.013808912597596645,
"learning_rate": 4.587390866353258e-06,
"loss": 0.0025,
"step": 12900
},
{
"epoch": 1.08,
"grad_norm": 0.0035981941036880016,
"learning_rate": 4.583193418401612e-06,
"loss": 0.001,
"step": 12910
},
{
"epoch": 1.08,
"grad_norm": 0.037712134420871735,
"learning_rate": 4.578995970449967e-06,
"loss": 0.0012,
"step": 12920
},
{
"epoch": 1.09,
"grad_norm": 0.03361295536160469,
"learning_rate": 4.5747985224983215e-06,
"loss": 0.0003,
"step": 12930
},
{
"epoch": 1.09,
"grad_norm": 0.016955135390162468,
"learning_rate": 4.570601074546676e-06,
"loss": 0.0005,
"step": 12940
},
{
"epoch": 1.09,
"grad_norm": 0.01527098473161459,
"learning_rate": 4.566403626595031e-06,
"loss": 0.0002,
"step": 12950
},
{
"epoch": 1.09,
"grad_norm": 0.011272264644503593,
"learning_rate": 4.562206178643385e-06,
"loss": 0.0002,
"step": 12960
},
{
"epoch": 1.09,
"grad_norm": 0.03093000315129757,
"learning_rate": 4.55800873069174e-06,
"loss": 0.0003,
"step": 12970
},
{
"epoch": 1.09,
"grad_norm": 0.019925663247704506,
"learning_rate": 4.553811282740095e-06,
"loss": 0.0008,
"step": 12980
},
{
"epoch": 1.09,
"grad_norm": 0.0032857642509043217,
"learning_rate": 4.549613834788449e-06,
"loss": 0.0007,
"step": 12990
},
{
"epoch": 1.09,
"grad_norm": 0.006109967827796936,
"learning_rate": 4.545416386836804e-06,
"loss": 0.0014,
"step": 13000
},
{
"epoch": 1.09,
"grad_norm": 0.007658987771719694,
"learning_rate": 4.541218938885158e-06,
"loss": 0.0003,
"step": 13010
},
{
"epoch": 1.09,
"grad_norm": 0.012553730048239231,
"learning_rate": 4.537021490933513e-06,
"loss": 0.0023,
"step": 13020
},
{
"epoch": 1.09,
"grad_norm": 0.0013994296314194798,
"learning_rate": 4.532824042981868e-06,
"loss": 0.0008,
"step": 13030
},
{
"epoch": 1.09,
"grad_norm": 0.44866248965263367,
"learning_rate": 4.528626595030222e-06,
"loss": 0.0068,
"step": 13040
},
{
"epoch": 1.1,
"grad_norm": 0.04373318701982498,
"learning_rate": 4.524429147078577e-06,
"loss": 0.0004,
"step": 13050
},
{
"epoch": 1.1,
"grad_norm": 0.00616934010758996,
"learning_rate": 4.5202316991269315e-06,
"loss": 0.0004,
"step": 13060
},
{
"epoch": 1.1,
"grad_norm": 0.12982237339019775,
"learning_rate": 4.516034251175286e-06,
"loss": 0.0005,
"step": 13070
},
{
"epoch": 1.1,
"grad_norm": 0.02142524905502796,
"learning_rate": 4.511836803223641e-06,
"loss": 0.0003,
"step": 13080
},
{
"epoch": 1.1,
"grad_norm": 0.054495833814144135,
"learning_rate": 4.5076393552719945e-06,
"loss": 0.0005,
"step": 13090
},
{
"epoch": 1.1,
"grad_norm": 0.2867399752140045,
"learning_rate": 4.503441907320349e-06,
"loss": 0.0008,
"step": 13100
},
{
"epoch": 1.1,
"grad_norm": 0.024905776605010033,
"learning_rate": 4.499244459368704e-06,
"loss": 0.0008,
"step": 13110
},
{
"epoch": 1.1,
"grad_norm": 0.0018473687814548612,
"learning_rate": 4.495047011417058e-06,
"loss": 0.0003,
"step": 13120
},
{
"epoch": 1.1,
"grad_norm": 0.1236455887556076,
"learning_rate": 4.490849563465413e-06,
"loss": 0.0046,
"step": 13130
},
{
"epoch": 1.1,
"grad_norm": 0.07682592421770096,
"learning_rate": 4.4866521155137685e-06,
"loss": 0.0006,
"step": 13140
},
{
"epoch": 1.1,
"grad_norm": 0.01288694143295288,
"learning_rate": 4.482454667562123e-06,
"loss": 0.0007,
"step": 13150
},
{
"epoch": 1.1,
"grad_norm": 0.0026057243812829256,
"learning_rate": 4.478257219610478e-06,
"loss": 0.0002,
"step": 13160
},
{
"epoch": 1.11,
"grad_norm": 0.012556037865579128,
"learning_rate": 4.4740597716588315e-06,
"loss": 0.0023,
"step": 13170
},
{
"epoch": 1.11,
"grad_norm": 0.01807723566889763,
"learning_rate": 4.469862323707186e-06,
"loss": 0.0003,
"step": 13180
},
{
"epoch": 1.11,
"grad_norm": 0.02373344823718071,
"learning_rate": 4.465664875755541e-06,
"loss": 0.0015,
"step": 13190
},
{
"epoch": 1.11,
"grad_norm": 0.0012779162498191,
"learning_rate": 4.461467427803895e-06,
"loss": 0.0001,
"step": 13200
},
{
"epoch": 1.11,
"grad_norm": 0.13116711378097534,
"learning_rate": 4.45726997985225e-06,
"loss": 0.0011,
"step": 13210
},
{
"epoch": 1.11,
"grad_norm": 0.0024926308542490005,
"learning_rate": 4.4530725319006046e-06,
"loss": 0.0011,
"step": 13220
},
{
"epoch": 1.11,
"grad_norm": 0.017771735787391663,
"learning_rate": 4.448875083948959e-06,
"loss": 0.0007,
"step": 13230
},
{
"epoch": 1.11,
"grad_norm": 0.002527177333831787,
"learning_rate": 4.444677635997314e-06,
"loss": 0.0009,
"step": 13240
},
{
"epoch": 1.11,
"grad_norm": 0.11742949485778809,
"learning_rate": 4.4404801880456684e-06,
"loss": 0.0003,
"step": 13250
},
{
"epoch": 1.11,
"grad_norm": 0.30960482358932495,
"learning_rate": 4.436282740094023e-06,
"loss": 0.0015,
"step": 13260
},
{
"epoch": 1.11,
"grad_norm": 0.039885230362415314,
"learning_rate": 4.432085292142378e-06,
"loss": 0.0005,
"step": 13270
},
{
"epoch": 1.11,
"grad_norm": 0.1734355092048645,
"learning_rate": 4.427887844190732e-06,
"loss": 0.0022,
"step": 13280
},
{
"epoch": 1.12,
"grad_norm": 0.003132781246677041,
"learning_rate": 4.423690396239087e-06,
"loss": 0.0053,
"step": 13290
},
{
"epoch": 1.12,
"grad_norm": 0.0005980546702630818,
"learning_rate": 4.4194929482874415e-06,
"loss": 0.0012,
"step": 13300
},
{
"epoch": 1.12,
"grad_norm": 0.004592574667185545,
"learning_rate": 4.415295500335796e-06,
"loss": 0.0006,
"step": 13310
},
{
"epoch": 1.12,
"grad_norm": 0.031002620235085487,
"learning_rate": 4.411098052384151e-06,
"loss": 0.0012,
"step": 13320
},
{
"epoch": 1.12,
"grad_norm": 0.09958618134260178,
"learning_rate": 4.406900604432505e-06,
"loss": 0.0006,
"step": 13330
},
{
"epoch": 1.12,
"grad_norm": 0.06991326808929443,
"learning_rate": 4.40270315648086e-06,
"loss": 0.004,
"step": 13340
},
{
"epoch": 1.12,
"grad_norm": 0.26911312341690063,
"learning_rate": 4.398505708529215e-06,
"loss": 0.0007,
"step": 13350
},
{
"epoch": 1.12,
"grad_norm": 0.35761767625808716,
"learning_rate": 4.394308260577569e-06,
"loss": 0.0023,
"step": 13360
},
{
"epoch": 1.12,
"grad_norm": 0.007256943266838789,
"learning_rate": 4.390110812625924e-06,
"loss": 0.0001,
"step": 13370
},
{
"epoch": 1.12,
"grad_norm": 0.16001386940479279,
"learning_rate": 4.3859133646742785e-06,
"loss": 0.0005,
"step": 13380
},
{
"epoch": 1.12,
"grad_norm": 0.01736878603696823,
"learning_rate": 4.381715916722633e-06,
"loss": 0.0003,
"step": 13390
},
{
"epoch": 1.12,
"grad_norm": 0.26081618666648865,
"learning_rate": 4.377518468770988e-06,
"loss": 0.001,
"step": 13400
},
{
"epoch": 1.13,
"grad_norm": 0.0361749492585659,
"learning_rate": 4.373321020819342e-06,
"loss": 0.0003,
"step": 13410
},
{
"epoch": 1.13,
"grad_norm": 0.028411924839019775,
"learning_rate": 4.369123572867697e-06,
"loss": 0.0005,
"step": 13420
},
{
"epoch": 1.13,
"grad_norm": 0.02852339670062065,
"learning_rate": 4.3649261249160515e-06,
"loss": 0.001,
"step": 13430
},
{
"epoch": 1.13,
"grad_norm": 0.0012577788438647985,
"learning_rate": 4.360728676964406e-06,
"loss": 0.0015,
"step": 13440
},
{
"epoch": 1.13,
"grad_norm": 0.23752494156360626,
"learning_rate": 4.356531229012761e-06,
"loss": 0.0006,
"step": 13450
},
{
"epoch": 1.13,
"grad_norm": 0.3116104304790497,
"learning_rate": 4.3523337810611146e-06,
"loss": 0.0012,
"step": 13460
},
{
"epoch": 1.13,
"grad_norm": 0.0024859828408807516,
"learning_rate": 4.34813633310947e-06,
"loss": 0.0008,
"step": 13470
},
{
"epoch": 1.13,
"grad_norm": 0.004907153081148863,
"learning_rate": 4.343938885157825e-06,
"loss": 0.0007,
"step": 13480
},
{
"epoch": 1.13,
"grad_norm": 0.0010646579321473837,
"learning_rate": 4.339741437206179e-06,
"loss": 0.0027,
"step": 13490
},
{
"epoch": 1.13,
"grad_norm": 0.2371465414762497,
"learning_rate": 4.335543989254534e-06,
"loss": 0.0012,
"step": 13500
},
{
"epoch": 1.13,
"grad_norm": 0.01433322299271822,
"learning_rate": 4.3313465413028885e-06,
"loss": 0.0015,
"step": 13510
},
{
"epoch": 1.13,
"grad_norm": 0.0013969101710245013,
"learning_rate": 4.327149093351243e-06,
"loss": 0.0008,
"step": 13520
},
{
"epoch": 1.14,
"grad_norm": 0.08360123634338379,
"learning_rate": 4.322951645399598e-06,
"loss": 0.0007,
"step": 13530
},
{
"epoch": 1.14,
"grad_norm": 0.009467093273997307,
"learning_rate": 4.3187541974479515e-06,
"loss": 0.0002,
"step": 13540
},
{
"epoch": 1.14,
"grad_norm": 0.07131925225257874,
"learning_rate": 4.314556749496306e-06,
"loss": 0.0003,
"step": 13550
},
{
"epoch": 1.14,
"grad_norm": 0.00770066911354661,
"learning_rate": 4.310359301544661e-06,
"loss": 0.0008,
"step": 13560
},
{
"epoch": 1.14,
"grad_norm": 0.011390848085284233,
"learning_rate": 4.306161853593015e-06,
"loss": 0.0006,
"step": 13570
},
{
"epoch": 1.14,
"grad_norm": 0.02216522768139839,
"learning_rate": 4.301964405641371e-06,
"loss": 0.0012,
"step": 13580
},
{
"epoch": 1.14,
"grad_norm": 0.1769302636384964,
"learning_rate": 4.2977669576897254e-06,
"loss": 0.001,
"step": 13590
},
{
"epoch": 1.14,
"grad_norm": 0.06368517130613327,
"learning_rate": 4.29356950973808e-06,
"loss": 0.0004,
"step": 13600
},
{
"epoch": 1.14,
"grad_norm": 0.022844741120934486,
"learning_rate": 4.289372061786435e-06,
"loss": 0.001,
"step": 13610
},
{
"epoch": 1.14,
"grad_norm": 0.009264915250241756,
"learning_rate": 4.2851746138347884e-06,
"loss": 0.0004,
"step": 13620
},
{
"epoch": 1.14,
"grad_norm": 0.004532854538410902,
"learning_rate": 4.280977165883143e-06,
"loss": 0.0011,
"step": 13630
},
{
"epoch": 1.15,
"grad_norm": 0.020466448739171028,
"learning_rate": 4.276779717931498e-06,
"loss": 0.0002,
"step": 13640
},
{
"epoch": 1.15,
"grad_norm": 0.0009050247026607394,
"learning_rate": 4.272582269979852e-06,
"loss": 0.0004,
"step": 13650
},
{
"epoch": 1.15,
"grad_norm": 0.011387677863240242,
"learning_rate": 4.268384822028207e-06,
"loss": 0.0005,
"step": 13660
},
{
"epoch": 1.15,
"grad_norm": 0.047608114778995514,
"learning_rate": 4.2641873740765615e-06,
"loss": 0.0003,
"step": 13670
},
{
"epoch": 1.15,
"grad_norm": 0.0906815230846405,
"learning_rate": 4.259989926124916e-06,
"loss": 0.0009,
"step": 13680
},
{
"epoch": 1.15,
"grad_norm": 0.06214312091469765,
"learning_rate": 4.255792478173272e-06,
"loss": 0.003,
"step": 13690
},
{
"epoch": 1.15,
"grad_norm": 0.580539345741272,
"learning_rate": 4.251595030221625e-06,
"loss": 0.0018,
"step": 13700
},
{
"epoch": 1.15,
"grad_norm": 0.08133112639188766,
"learning_rate": 4.24739758226998e-06,
"loss": 0.0008,
"step": 13710
},
{
"epoch": 1.15,
"grad_norm": 0.006161406170576811,
"learning_rate": 4.243200134318335e-06,
"loss": 0.0003,
"step": 13720
},
{
"epoch": 1.15,
"grad_norm": 0.11143741756677628,
"learning_rate": 4.239002686366689e-06,
"loss": 0.0004,
"step": 13730
},
{
"epoch": 1.15,
"grad_norm": 0.3503442406654358,
"learning_rate": 4.234805238415044e-06,
"loss": 0.005,
"step": 13740
},
{
"epoch": 1.15,
"grad_norm": 0.01972077041864395,
"learning_rate": 4.2306077904633985e-06,
"loss": 0.0002,
"step": 13750
},
{
"epoch": 1.16,
"grad_norm": 0.022888071835041046,
"learning_rate": 4.226410342511753e-06,
"loss": 0.0005,
"step": 13760
},
{
"epoch": 1.16,
"grad_norm": 0.12486062943935394,
"learning_rate": 4.222212894560108e-06,
"loss": 0.0006,
"step": 13770
},
{
"epoch": 1.16,
"grad_norm": 0.043477680534124374,
"learning_rate": 4.218015446608462e-06,
"loss": 0.0049,
"step": 13780
},
{
"epoch": 1.16,
"grad_norm": 0.02344905212521553,
"learning_rate": 4.213817998656817e-06,
"loss": 0.0018,
"step": 13790
},
{
"epoch": 1.16,
"grad_norm": 0.13969235122203827,
"learning_rate": 4.2096205507051716e-06,
"loss": 0.0007,
"step": 13800
},
{
"epoch": 1.16,
"grad_norm": 0.5832682847976685,
"learning_rate": 4.205423102753526e-06,
"loss": 0.0029,
"step": 13810
},
{
"epoch": 1.16,
"grad_norm": 0.004915925674140453,
"learning_rate": 4.201225654801881e-06,
"loss": 0.0007,
"step": 13820
},
{
"epoch": 1.16,
"grad_norm": 0.024911779910326004,
"learning_rate": 4.197028206850235e-06,
"loss": 0.0007,
"step": 13830
},
{
"epoch": 1.16,
"grad_norm": 0.0007527291309088469,
"learning_rate": 4.19283075889859e-06,
"loss": 0.0023,
"step": 13840
},
{
"epoch": 1.16,
"grad_norm": 0.008193068206310272,
"learning_rate": 4.188633310946945e-06,
"loss": 0.0006,
"step": 13850
},
{
"epoch": 1.16,
"grad_norm": 0.10630033910274506,
"learning_rate": 4.184435862995299e-06,
"loss": 0.001,
"step": 13860
},
{
"epoch": 1.16,
"grad_norm": 0.053735826164484024,
"learning_rate": 4.180238415043654e-06,
"loss": 0.0013,
"step": 13870
},
{
"epoch": 1.17,
"grad_norm": 0.008053253404796124,
"learning_rate": 4.1760409670920085e-06,
"loss": 0.0011,
"step": 13880
},
{
"epoch": 1.17,
"grad_norm": 0.012578310444951057,
"learning_rate": 4.171843519140363e-06,
"loss": 0.0005,
"step": 13890
},
{
"epoch": 1.17,
"grad_norm": 0.07252337783575058,
"learning_rate": 4.167646071188718e-06,
"loss": 0.0005,
"step": 13900
},
{
"epoch": 1.17,
"grad_norm": 0.004061713349074125,
"learning_rate": 4.163448623237072e-06,
"loss": 0.0016,
"step": 13910
},
{
"epoch": 1.17,
"grad_norm": 0.07049524784088135,
"learning_rate": 4.159251175285427e-06,
"loss": 0.0006,
"step": 13920
},
{
"epoch": 1.17,
"grad_norm": 0.00046717686927877367,
"learning_rate": 4.155053727333782e-06,
"loss": 0.0005,
"step": 13930
},
{
"epoch": 1.17,
"grad_norm": 0.021116163581609726,
"learning_rate": 4.150856279382136e-06,
"loss": 0.0017,
"step": 13940
},
{
"epoch": 1.17,
"grad_norm": 0.022102046757936478,
"learning_rate": 4.146658831430491e-06,
"loss": 0.0017,
"step": 13950
},
{
"epoch": 1.17,
"grad_norm": 0.007913138717412949,
"learning_rate": 4.1424613834788454e-06,
"loss": 0.0049,
"step": 13960
},
{
"epoch": 1.17,
"grad_norm": 0.022254914045333862,
"learning_rate": 4.1382639355272e-06,
"loss": 0.0019,
"step": 13970
},
{
"epoch": 1.17,
"grad_norm": 0.01675882749259472,
"learning_rate": 4.134066487575555e-06,
"loss": 0.0004,
"step": 13980
},
{
"epoch": 1.17,
"grad_norm": 0.004890616983175278,
"learning_rate": 4.1298690396239085e-06,
"loss": 0.0005,
"step": 13990
},
{
"epoch": 1.18,
"grad_norm": 0.5471255779266357,
"learning_rate": 4.125671591672263e-06,
"loss": 0.0041,
"step": 14000
}
],
"logging_steps": 10,
"max_steps": 23824,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 1000,
"total_flos": 0.0,
"train_batch_size": 12,
"trial_name": null,
"trial_params": null
}